{ "0003/astro-ph0003451_arXiv.txt": { "abstract": "Integrated spectrophotometric properties of stellar systems are intrinsically dispersed due to the stochastic nature of the small numbers of bright stars they contain. Among clusters, only the most massive ones can be used individually for comparison with the mean properties predicted by population synthesis calculations. The appropriate minimal masses depend, among others, on the waveband or colour index studied and on age. Selected indices (near-IR CO and H$_2$O, EW(H$\\alpha$)) are used to illustrate the asymmetric nature of the probability distribution of observable properties and their dependence on cluster mass. ", "introduction": "Most population synthesis codes (those based on Monte Carlo simulations excepted) predict {\\em mean} properties of stellar populations as a function of fundamental model parameters such as the stellar initial mass function (IMF) and the star formation history (SFH). However, for a given model the number of stars in each area of the HR diagram is a statistical variable obeying Poisson statistics. The resulting intrinsic dispersion of integrated spectrophotometric properties is observed, both as pixel to pixel fluctuations in otherwise uniform objects (``surface brightness fluctuations\", Tonry \\& Schneider 1988) and as cluster to cluster variations among cluster samples restricted to similar SFHs (Ferraro et al. 1995, Girardi et al. 1995). Clusters are tempting targets for the tests and calibrations of population synthesis prediction because of the coeval nature of their stars. In this context, it is important to remember that the properties of {\\em individual} clusters are representative of the mean properties only in the limit of large star numbers, i.e. large cluster masses (assuming a universal IMF). Discussing how large these masses need to be in practice is the purpose of this paper. All results presented here are based on the population synthesis code { P\\'egase} (Fioc \\& Rocca-Volmerange 2000) and extensions thereof. They assume solar metallicity and a Salpeter IMF extending from 0.1 to 120\\,M$_{\\odot}$. ", "conclusions": "Stochastic fluctuations due to small numbers of bright stars need to be considered when stellar populations are compared to population synthesis models, be it using star counts or using integrated properties. The most probable spectrophotometric properties of small clusters are usually different from their expectation value (i.e. the properties in the large cluster limit), leading to systematic effects in the determination of age, metallicity or other fundamental parameters. The adequate definition of a {\\em massive cluster}, for which these effects would be negligible, depends strongly on the spectrophotometric property studied and on the star formation history. The cluster populations formed in galaxy mergers, thoroughly discussed during this workshop and known to contain objects of more than 10$^6$\\,M$_{\\odot}$, are becoming accessible to spectrographs on large telescopes and clearly represent important targets for population synthesis studies of the near future (e.g. Mouhcine \\& Lan\\c{c}on, this volume)." }, "0003/astro-ph0003384_arXiv.txt": { "abstract": "We present the analysis of a CCD survey of 31 nearby ($\\le$ 110 Mpc) edge-on spiral galaxies. The three-dimensional one-component best fit models provide their disk-scalelengths $h$ and for the first time their disk cut-off radii $R_{co}$. We confirm for this sample the existence of such sharp truncations, and find a significantly lower mean value of the distance independent ratio $R_{co}/h =2.9 \\pm 0.7$ than the standard value of $4.5$ often used in the literature. Our data show no correlation of these parameters with Hubble type, whereas we report a correlation between $R_{co}/h$ and the distance based scalelength in linear units. Compared to the Milky Way we find only lower values of $R_{co}/h$, explained either by possible selection effects or by the completely different techniques used. We discuss our data in respect to present models for the origin of the cut-off radii, either as a relict of the galaxy formation process, or as an evolutionary phenomenon. ", "introduction": "Although {\\it cut-off radii} of spiral galaxies are known for about 20 years no unique physical explanation has been given to describe this observational phenomenon. They were already mentioned by van der Kruit (\\cite{vdk79}), who stated, based on photographic material, that the outer parts of disks of spiral galaxies ''do not retain their exponential light distribution to such faint levels'', whereas the exponential behaviour of the radial light distribution for the inner part was well accepted (de Vaucouleurs \\cite{devau59}, Freeman \\cite{free}). For three nearby edge-on galaxies he claimed, that the typical radial scalelength $h$ steepens from 5 kpc to about 1.6 kpc at the edge of the disk. This is confirmed by modern deep CCD imaging (Abe et al. \\cite{abe}, Fry et al. \\cite{fry}, N\\\"aslund \\& J\\\"ors\\\"ater \\cite{naes}). In a fundamental series of papers van der Kruit \\& Searle (\\cite{vdk81a}, \\cite{vdk81b}, \\cite{vdk82a}, \\cite{vdk82b}) determined a three dimensional model for the luminosity density of the old disk po\\-pu\\-lation taking into account these sharp truncations at the {\\it cut-off radius} $R_{co}$. They applied their model of a locally isothermal, selfgravitating, and truncated exponential disk to a sample of seven edge-on galaxies and found that all disks show a relatively sharp cut-off where the scalelength $h$ suddenly drops below 1 kpc, starting at radii of $(4.2 \\pm 0.5) h$. The {\\it cut-off radius} of edge-on galaxies is detected at levels of 24-25 mag$/\\sq\\arcsec$ which is about 2-3 mag brighter compared to face-on disks due to the integration along the line of sight. Therefore van der Kruit \\& Shostak (\\cite{vdkshos82}) and Shostak \\& van der Kruit (\\cite{shos}) quote the only known cut-offs in the literature for face-on galaxies. In addition to the much lower brightness one has to deal with intrinsic deviations from the circular symmetry of the disk, for example from the young stellar population, hidden by an azimuthally averaged profile. In a subsequent paper van der Kruit (\\cite{vdk88}) stated that out of the 20 face-on galaxies observed by Wevers et al. (\\cite{wev}) only four did not show any sign for a drop off as judged from the last three contours. Barteldrees \\& Dettmar (\\cite{bdold}) confirmed for the first time the existence of these truncations for a larger sample of edge-on galaxies using CCD surface photometry refining the previous photographic measurements.\\\\ These truncations are not the boundary of the galactic baryonic mass distribution, but such 'optical edges' suggest dynamical consequences for the interpretation of observed rotation curves (Casertano \\cite{caser}), as well as for the explanation of warped disks (Bottema \\cite{bott}). Their sharpness restrict the radial velocity dispersion at the edge of the disk (van der Kruit \\& Searle \\cite{vdk81a}), and will therefore have important implication for viscous disk evolution (Thon \\& Meusinger \\cite{thon}). According to Zhang \\& Wyse (\\cite{zhang}) the disk cut-off radii constrain the specific angular momentum in a viscous galaxy evolution scenario.\\\\ In this letter we report the largest sample of well defined cut-off radii for edge-on galaxies derived by CCD surface photometry. Our sample (Pohlen et al. \\cite{pohl}, Paper II) comprises 31 galaxies, including the 17 galaxies of Barteldrees \\& Dettmar (\\cite{bd}, hereafter Paper I). Thereby we are able to derive first statistical conclusions and determine general correlations with other characteristic galaxy parameters in order to approach in the future a physical model explaining the observed phenomenon. ", "conclusions": "" }, "0003/hep-th0003066_arXiv.txt": { "abstract": "We numerically study classical time evolutions of Kaluza-Klein bubble space-time which has negative energy after a decay of vacuum. As the zero energy Witten's bubble space-time, where the bubble expands infinitely, the subsequent evolutions of Brill and Horowitz's momentarily static initial data show that the bubble will expand in terms of the area. At first glance, this result may support Corley and Jacobson's conjecture that the bubble will expand forever as well as the Witten's bubble. The irregular signatures, however, can be seen in the behavior of the lapse function in the maximal slicing gauge and the divergence of the Kretchman invariant. Since there is no appearance of the apparent horizon, we suspect an appearance of a naked singularity as the final fate of this space-time. ", "introduction": "It is likely that superstring or M-theory governs the physics of gravity or space-time in higher energy stages\\cite{String}. Such theories are naturally formulated in the higher dimensions than four. We expect a plausible scenario that such a higher dimensional space-time somehow evolves to the stable four dimensional space-time according to the history of the Universe \\cite{brane,Tess,recent}. The so-called brane world scenario\\cite{brane} is the most actively being investigated along to this line. This scenario is motivated by Horava and Witten's theory \\cite{Horava} which shows that an eleven dimensional supergravity theory on the orbihold ${\\bf R}^{10} \\times S^1 /Z_2$ is related to the ten-dimensional $E_8 \\times E_8$ heterotic string theory. Therein the matters are confined to the ten-dimensional space-time (three-brane) and gravitons are propagating in the full eleven dimensions. The brane world space-time should be stable. Although the brane world scenario may be plausible at the reduction from eleven to ten dimensions, the space-time will be still compactified to four dimensions in the normal Calabi-Yau's way. Regarding to these full scenario of the compactification, the stability of the space-time becomes the important issue to be investigated. The positive energy theorem guarantees the stability of the four dimensional asymptotically flat space-time in the framework of general relativity \\cite{PET}. Surprisingly, the existence of the extra dimensions can drastically change the situation. Witten showed that the five dimensional Minkowski space-time decays into the so-called Kaluza-Klein (KK) bubble space-time unless we assume the existence of the elementary fermion related to supersymmetry \\cite{Witten,Marika}, of which existence we can not expect generally. This also may indicate that the `bubble' appears somewhere at the bulk or on the brane in the brane world scenario and disturbed the three-brane where we are living. The metric of the KK bubble space-time given by Witten is written as\\cite{Witten} \\begin{eqnarray} ds_5^2 & = & -r^2dt^2+\\Bigl(1-\\frac{r^2_0}{r^2}\\Bigr)d\\chi^2+ \\Bigl(1-\\frac{r^2_0}{r^2}\\Bigr)^{-1}dr^2 +r^2 \\cosh^2 t \\, d\\Omega^2, \\end{eqnarray} where the $\\chi$-direction will be compactified and $r \\geq r_0$, and $d\\Omega^2=d\\theta^2+\\sin^2\\theta d\\varphi^2$. In general case, the metric has a conical singularity at $r_0$. However, if we carefully take a periodicity along the $\\chi$-direction, the metric can be regularized. More precisely to see this, we write the metric near $r=r_0$ as \\begin{eqnarray} ds_5^2 & \\simeq & -r_0^2dt^2+\\frac{2(r-r_0)}{r_0}d\\chi^2+ \\frac{r_0}{2(r-r_0)}dr^2 +r^2_0 \\cosh^2 t \\, d\\Omega^2 \\nonumber \\\\ & = & -r_0^2dt^2+2r_0\\Bigl[R^2d\\Bigl(\\frac{\\chi}{r_0}\\Bigr)^2 +dR^2 \\Bigr] +r^2_0 \\cosh^2 t \\, d\\Omega^2, \\end{eqnarray} where $R=\\sqrt{r-r_0}$. Then we realize that the period should be set to be $\\chi_p=2\\pi r_0$. As one can see, the `boundary of bubble' located at $r=r_0$ expands rapidly like $\\cosh t$ and the space-time does not have naked singularities. Here, we remind you that the total energy is zero. We are imaging the boundary of the space-time, $r=r_0$, as the surface of the `bubble'. Interestingly Brill and Pfister \\cite{Brill1} gave an initial data which has the negative total energy related to the size of the compactified dimension. The space-time with the negative energy may be favor in the aspect of energetics. One year later, Brill and Horowitz\\cite{Brill2} gave an initial data in a simple way. (We will briefly review their construction in Sec. \\ref{BH-initialdata}.) Contrasted to the `Witten bubble', their solution has arbitrary negative energy regardless of the size of the compactified space. This is too far from our intuition that the negative energy is proportional to the Casimir energy due to the boundary effect of the compactified space. Therefore it is difficult to imagine the classical evolution after the vacuum decay. Corley and Jacobson \\cite{Ted} discussed the subsequent evolution of Brill-Horowitz's initial data. They found that the positive acceleration of the bubble's surface area for the negative mass bubbles, and they conjectured that KK bubble with negative energy cannot collapse. However, their study is not sufficient to conclude the final fate of the bubble, as they already mentioned, because they considered only the initial behavior of the time-symmetric data and they did not do any dynamical studies. In this paper, we report our numerical analysis on this final fate problem of KK bubble, especially of the negative mass bubble. We start our numerical simulation from the Brill-Horowitz's initial data, and evolve the space-time using the standard Arnowitt-Deser-Misner formulation (but 4+1 dimensional decomposition). We will show that the space-time initially behaves as Corley-Jacobson's analysis, and expands forever, although the acceleration will be negative. Despite of the expanding, we will observe the irregular behavior of the curvature invariant. This paper is organized as follows. In Sec. \\ref{BH-initialdata}, we give a brief review of Brill-Horowitz's construction of their initial data. In Sec. \\ref{Numerical-method}, we describe numerical method and equations. The results of our simulations are shown in in Sec. \\ref{Results}. Finally, we summarize our results in Sec. \\ref{Summary}. ", "conclusions": "\\label{Summary} We numerically studied the dynamical evolution of the Brill and Horowitz's initial data which can have the negative energy. As the zero energy Witten's bubble space-time, we show that the `bubbles' with negative energy will expand by mean of area upto the time we stop the simulations. At first glance this result supports Corley and Jacobson's conjecture. However, from the facts that the curvature invariant blows up, and no appearance of the apparent horizon, we suspect that a formation of a naked singularity as the final fate of Kaluza-Klein negative energy bubble \\footnote{ An anonymous referee of this article pointed out the similarity of the positive and negative mass bubble results. However, from the results we obtained, we believe that there are qualitative differences between positive and negative mass cases in their dynamical behaviors. (We remark that our simulations are only up to a finite time in order to keep the resolution against the expansion of the spacetime.) }. Hence, we may have to consider seriously the decay problem from the Kaluza-Klein vacuum to the Witten-type `bubble' space-time. Possible resolution to this may be given by assuming the supersymmetry which may forbid the decay \\cite{Witten,Marika}, or by constructing quantum gravity theory which may smooth out singularities as normally been expected. Although the negative mass bubbles are expanding, we obtained the result that the bubble spacetime terminates at the singularity. At first glance, they are incompatible, because the naive picture, which the expanding keeps regularity. However, the picture may be based on the Raychaudhri-type equation and the equation does not hold in the present case. Moreover, the area % cannot properly describes whether the system will collapse or not. Properly speaking, we need the proper radius from the center which is absent in the present case. {}Finally, we would like to comment on the so-called brane world scenario \\cite{brane,Tess,recent}. The brane world is motivated by the reduction from the M-theory to the $E_8 \\times E_8$ heterotic superstring theory. This reduction drastically changes the picture of the reduction \\cite{brane} because `matters' are confined to the ten dimensions and gravitons are propagating in the eleven-dimensions. Here we call the timelike hypersurface, where matters are confined, by `brane'. A plausible history of the compactification are still quite actively discussed recently and we do not reach the consensus at this moment. Apart from this compactification scenario, the reduction from ten to four-dimensions follows the well-known Kaluza-Klein type or Calabi-Yau compactification. Thus, our present analysis is basically applicable to the space-time on the brane, because we supposed the well-known Kaluza-Klein compactification. More precisely, we may say that the Witten-type Kaluza-Klein `bubble' space-time on the 4-brane will be reduced from at least 6 dimensional space-time. As was recently reported \\cite{Tess}, the effective Einstein equations on the brane are different from the normal Einstein equations. Therefore, it might be worth re-asking what is the final fate of Kaluza-Klein bubble if we describe the space-time by such a modified Einstein equation when we take a brane world scenario." }, "0003/astro-ph0003101_arXiv.txt": { "abstract": "Combination frequencies are observed in the Fourier spectra of pulsating DA and DB white dwarfs. They appear at sums and differences of frequencies associated with the stellar gravity-modes. Brickhill (1992) proposed that the combination frequencies result from mixing of the eigenmode signals as the surface convection zone varying in depth when undergoing pulsation. This depth changes cause time-dependent thermal impedance, which mix different harmonic frequencies in the light curve. Following Brickhill's proposal, we developed {\\it analytical} expressions to describe the amplitudes and phases of these combination frequencies. The parameters that appear in these expressions are: the depth of the stellar convection zone when at rest, the sensitivity of this depth towards changes in stellar effective temperature, the inclination angle of the stellar pulsation axis with respect to the line of sight, and lastly, the spherical degrees of the eigenmodes involved in the mixing. Adopting reasonable values for these parameters, we apply our expressions to a DA and a DB variable white dwarf. We find reasonable agreement between theory and observation, though some discrepancies remain unexplained. We show that it is possible to identify the spherical degrees of the pulsation modes using the combination frequencies. ", "introduction": "} Long sequences of almost uninterrupted light curves, obtained during the Whole Earth Telescope (WET) campaign on the helium variable white dwarf GD358 (Winget \\etal 1994, for a more recent analysis, see Vuille \\etal 2000), disclosed not only the presence of a large number of stellar pulsational modes in the Fourier spectra, but also the presence of `combination frequencies', signals that lie at the sum or difference frequencies of the stellar eigen-modes (see Fig. 7 in the first paper). Combination frequencies have been observed in other pulsating white dwarfs with either hydrogen or helium atmospheres, e.g., ZZ Psc (aka G29-38, McGraw 1976; Kleinman \\etal 1998), GD154 \\cite{robinson78}, BPM31594 (McGraw 1976; O'Donoghue, Warner \\& Cropper 1992), G117-B15A \\cite{kepler82} and GD165 \\cite{bergeron93}. Indeed, every variable hydrogen white dwarf (class name ZZ Ceti, or DAV) that has been observed with sufficiently high signal-to-noise ratio exhibits combination frequencies (Brassard, Fontaine \\& Wesemael 1995). The same likely holds for helium variables (DBV). Combination frequencies are thought to result from nonlinear mixing of sinusoidal signals that are associated with the eigenmodes (named the `principal modes' in this article). This conclusion is based on the following arguments: combination frequencies are too numerous to be eigenmodes themselves (Winget \\etal 1994); amplitudes of the combination frequencies have been shown to correlate with those of their principal modes (for an early review, see McGraw 1978); combination frequencies tend to have more complicated fine structure than their principal modes, which can be explained naturally by a linear superposition of the principal modes' rotationally split multiplets \\cite{winget94}. Brickhill \\shortcite{brick92} showed that nonlinear mixing arises naturally in the context of his theory of convective driving (Brickhill 1983, 1990, 1991a, 1991b). He realized that the convective turn-over time scale in DA and DB variable white dwarfs is much shorter than the pulsation period. Thus one can safely assume that the surface convective region adjusts instantaneously during pulsation. Brickhill found that under this assumption, the photospheric flux variation is delayed and reduced relative to that entering the bottom of the convection zone, by an amount depending on the depth of the convection zone. Instantaneous adjustment of the convection zone also implies that the extent of the convection zone varies during the pulsation cycle, thus leading to variations in the reduction and delay of the flux variation. This distorts the shape of the light curve at the photosphere, and brings about the combination frequencies in the Fourier power spectrum. Using a numerical analysis, Brickhill found that for reasonable amplitudes of the principal modes, he could reproduce the observed amplitudes of the combination frequencies. Note that in this theory, the combination frequencies reflect distortion of the light curve by the nonlinear medium; they are not associated with physical displacements and velocities.\\footnote{Fast convection enforces uniform movement throughout the convective region \\cite{paperIII}. There is no distortion to the velocity signal.} This is indeed confirmed observationally (van Kerkwijk, Clemens \\& Wu 1999). In this paper, we use a perturbative analysis to derive analytical formulae for the strength and phase of the combination frequencies. The advantage of this analysis over Brickhill's numerical approach is that the dependence on stellar properties becomes explicit. We find that two parameters, namely, the depth of the surface convection zone when the star is at rest, and the sensitivity of this depth towards changes in stellar effective temperature, determine the efficiency of the mixing process. We also show that two geometric factors, the spherical degrees of the principal modes, and the inclination angle of the stellar pulsation axis, enter the analytical expressions. We compare our formulae with data on GD358 (a DBV) and G29-38 (a DAV), adopting appropriate values for the above stellar parameters. Despite imperfect agreement, we show that it is possible to infer the spherical degree for the principal modes. In \\S \\ref{subsec:othertype}, we briefly discuss the prospects of explaining the combination frequencies in other types of variable stars. ", "conclusions": "\\label{sec:conclusions} \\subsection{DA \\& DB Variables \\label{subsec:DADB}} Our analysis leads to two key formulae (eq. [\\ref{eq:theory-Rc}] and [\\ref{eq:general-phase}]) that describes the strength and the phase of the combination frequencies relative to their principal modes. A few stellar parameters enter these formulae. They are the thermal constant of the stellar convection zone at equilibrium ($\\tau_{c_0}$), the rate of deepening of the convection zone with cooling of the star (quantified by $|2 \\beta + \\gamma|$), and the inclination angle between the observer's line of sight and the stellar pulsation axis ($\\Theta_0$). It is becoming possible to use the combination frequencies to constrain these stellar parameters. We find that for both GD358 and ZZ Psc, the observed amplitude spectra can be roughly reproduced using reasonable choices of the above parameters. The same choices can also explain the values for the dimensionless numbers ($a,b$ and $c$) Brickhill (1992) summarized from his numerical study. The $\\ell$ and $m$ values of the eigenmodes also enter into these formulae. This presents the potential of determining the $\\ell$ values of the principal modes using the combination frequencies. An $\\ell = 2$ mode is expected to have a stronger harmonic than an $\\ell = 1$ mode, and this is indeed observed in ZZ Psc. When analyzing observed combination frequencies, we have ignored amplitude variability during a long observing run, or in the case of a short run, have assumed that modes are axisymmetric with respect to the pulsation axis ($m=0$). The failure of these assumptions may account for some of the discrepancies between observation and theory and may prevent us from accurately determining stellar parameters. More suitable data sets might yield more conclusive information. We find that theory over-predicts the amplitude in the harmonic of the strongest pulsation mode in the two stars we considered. We suspect that it results from the stronger nonlinearity associated with the largest mode. Combination frequencies are produced by the surface convection zone in a pulsating white dwarf. Photosphere in these stars is not thermally capable of distorting the light curve. We therefore expect equations \\refnew{eq:theory-Rc} and \\refnew{eq:general-phase} to hold for all wavelengths. \\subsection{Other Types of Variables \\label{subsec:othertype}} Combination frequencies have also been reported in two PG1159 variables (PG1707+427, Fontaine \\etal 1991; HS2324+3944, Silvotti \\etal 1999). Presumably, these hot white dwarfs do not have surface convection zones. What could be distorting the light curves? Could a radiative, partially ionising layer produce the distortions? Such a layer is believed to exist in the upper atmosphere of PG1159 variables and is believed to be responsible for driving the observed pulsations. It is similar to the surface convection zones in DA and DB variables in that it retains heat when warmer, and releases heat when cooler. However, unlike in the case of the convection zones, the amount of heat retained (or released) by the partial ionising region can not be significantly modulated throughout the pulsation cycle by the presence of other pulsation modes. This is because the reaction time of the ionising region is roughly the local thermal relaxation time, which is of the same order as periods of overstable modes. Thus, we can not explain the combination frequencies in PG 1159 stars. We note that in other types of small amplitude pulsators, e.g., $\\delta$-Scuti stars, sdB variables, and $\\gamma$-Doradus stars, combination frequencies have also been reported. We conjecture that a thin surface convection zone is present in these variables and is capable of exciting pulsation modes, as well as distorting the light curves. \\bigskip The author would like to acknowledge the many beneficial comments and suggestions by Drs. Peter Goldreich, Joerg Ising, Marten van Kerkwijk, Scot Kleinman and Francois Vuille. \\begin{table*} \\begin{center} \\begin{tabular}{c|ccc|} $\\ell \\setminus |m|$& $2$ & $1$ & $0$ \\\\ \\hline $0$\t& -- & -- & 0.28 \\\\ $1$\t& -- & $0.245\\sin\\Theta_0$ & $0.346\\cos\\Theta_0$ \\\\ $2$\t&$0.126\\sin^2\\Theta_0$ & $0.126\\sin(2\\Theta_0)$ &$0.05[1+3\\cos(2\\Theta_0)]$ \\\\ \\hline \\end{tabular} \\end{center} \\caption[]{Values of $g_{\\ell}^m$ as functions of $\\Theta_0$, the inclination angle between the observer's line-of-sight and the stellar rotation axis.} \\label{table:appendix-gvalue} \\end{table*} \\begin{table*} \\begin{center} \\begin{tabular}{c|ccc|} $m_j \\setminus m_i $ & $-1$ & $0$ & $+1$ \\\\ \\hline $-1$ & $0.65$ & $0.65$ & $0.63+{{0.90}\\over{\\sin^2 \\Theta_0}}$ \\\\ $0$ & $0.65$ & $0.65 + {{0.45}\\over{\\cos^2 \\Theta_0}}$ & $0.65$\\\\ $+1$ & $0.63 + {{0.90}\\over{\\sin^2 \\Theta_0}}$ & $0.65$ & $0.65$ \\\\ \\hline \\end{tabular} \\end{center} \\caption[]{Values of $G_{1\\,\\,\\,\\,\\, 1}^{m_i + m_j}/(g_{1}^{m_i} g_{1}^{m_j})$ as functions of $\\Theta_0$. Values of $G_{\\ell_i\\,\\,\\,\\, \\ell_j}^{m_i - m_j}/(g_{\\ell_i}^{m_i} g_{\\ell_j}^{m_j})$ are obtained by reversing the sign of $m_j$.} \\label{table:appendix-bigG-plus} \\end{table*}" }, "0003/astro-ph0003337_arXiv.txt": { "abstract": "We have undertaken a survey of H$\\alpha$ emission in a substantially complete sample of CGCG galaxies of types Sa and later within 1.5 Abell radii of the centres of 8 low-redshift Abell clusters (Abell 262, 347, 400, 426, 569, 779, 1367 and 1656). Some 320 galaxies were surveyed, of which 116 were detected in emission (39\\% of spirals, 75\\% of peculiars). Here we present previously unpublished data for 243 galaxies in 7 clusters. Detected emission was classified as `compact' or `diffuse'. From an analysis of the full survey sample, we reconfirm our previous identification of compact and diffuse emission with circumnuclear starburst and disk emission respectively. The circumnuclear emission is associated either with the presence of a bar, or with a disturbed galaxy morphology indicative of on-going tidal interactions (whether galaxy--galaxy, galaxy--group, or galaxy--cluster). The frequency of such tidally-induced (circumnuclear) starburst emission in spirals increases from regions of lower to higher local galaxy surface density, and from clusters with lower to higher central galaxy space density. The percentages of spirals classed as disturbed, and of galaxies classified as peculiar show a similar trend. These results suggest that tidal interactions for spirals are more frequent in regions of higher local density and for clusters with higher central galaxy density. The prevalence of such tidal interactions in clusters is expected from recent theoretical modelling of clusters with a non-static potential undergoing collapse and infall. Furthermore, in accord with this picture, we suggest that peculiar galaxies are predominantly on-going mergers. We conclude that tidal interactions are likely to be the main mechanism for the transformation of spirals to S0s in clusters. This mechanism operates more efficiently in higher density environments as is required by the morphological type--local surface density (T--$\\Sigma$) relation for galaxies in clusters. For regions of comparable local density, the frequency of tidally-induced starburst emission is greater in clusters with higher central galaxy density. This implies that, for a given local density, morphological transformation of disk galaxies proceeds more rapidly in clusters of higher central galaxy density. This effect is considered to be due to subcluster merging and could account for the previously considered anomalous absence of a significant T--$\\Sigma$ relation for irregular clusters at intermediate redshift. ", "introduction": " ", "conclusions": "Kennicutt (1998) notes that large-scale star formation takes place in two very distinct physical environments, viz. in the extended disks of spiral and irregular galaxies, and in compact, dense gas disks in the centres of galaxies. Line emission associated with star formation in the two regions have very different dependencies on galaxy morphology. In particular, circumnuclear emission has a strong dependence on a barred structure but weak dependence on Hubble type, while the opposite is true for disk emission. In addition, a clear causal link between strong nuclear starbursts and tidal interactions and mergers has been established by numerous observations (e.g. Keel et al. 1985; Bushouse 1987; Kennicutt et al. 1987; Wright et al. 1988) consistent with predictions of numerical simulations (e.g. Noguchi 1988; Hernquist 1989; Mihos \\& Hernquist 1996). For nearby samples of interacting galaxies, the H$\\alpha$ emission is typically 3--4 times stronger than for isolated spirals. In view of the above, the obvious and most compelling interpretation of the distinction between compact and diffuse emission in our survey sample is that of circumnuclear starburst and disk emission respectively. As has been seen, compact emission is generally centred on the nucleus of the galaxy and is of smaller spatial extent (median diameter $\\sim$ 2.5 kpc, Paper II), and correlates with a barred structure, all of which is typical of circumnuclear emission. Furthermore compact emission is strongly correlated with a disturbed morphology and with the presence of a nearby companion, strongly suggesting that much of this emission is indeed due to tidally-induced nuclear starbursts. Finally there is no significant dependence of compact emission on Hubble stage from Sa--Sc. By contrast, diffuse emission has a greater spatial extent closer to that expected for disk emission (median diameter $\\sim$ 7 kpc, Paper II) and is not correlated either with a barred structure or a disturbed morphology. There is no apparent dependence of emission on Hubble stage, but this is not considered significant because of uncertainties associated with detection of diffuse emission (cf. section \\ref{htype}). Despite the above considerations, could the compact emission be due to non-stellar emission? Ho, Filippenko \\& Sargent (1997a) have extended earlier studies to show that AGN and LINER (Heckman 1980) emission is very common, particularly in early-type spirals of which 60\\% show this non-stellar activity. However galaxies in their sample have low H$\\alpha$ luminosity (median $L_{H\\alpha} \\sim 10^{39}$ erg s$^{-1}$). By contrast the H$\\alpha$ luminosities of the ELGs in our survey sample are higher ($10^{40} \\la L_{H\\alpha} \\la 5 \\times 10^{41}$ erg s$^{-1}$), more typical of lower luminosity starburst emission (Balzano 1983). Moreover Ho, Filippenko \\& Sargent (1997b) show that whereas bars enhance nuclear star formation in their sample, there is no corresponding enhancement of AGN activity. Finally, despite the fact that a few of our ELGs are classed as Seyferts (viz. CGCG nos. 126-110, 522-081, 540-064, 540-103 and 160-148A) a spectral survey of ELGs in Abell 1367 (Moss \\& Whittle, unpublished) has confirmed that emission for most of these galaxies resemble HII regions and not AGN or LINERs. For these reasons, it is considered unlikely that most compact emission has a non-stellar origin. In what follows, we assume that both compact and diffuse emission are due to photoionisation by massive young stars, and investigate how the corresponding circumnuclear starburst and disk star formation varies within a cluster and field environment." }, "0003/astro-ph0003047_arXiv.txt": { "abstract": "\\label{abstract} % We present new near-infrared $J$ and $K$ imaging data for 67 galaxies from the Universidad Complutense de Madrid survey used to determine the SFR density of the local universe by Gallego et al. (1995). This is a sample of local star-forming galaxies with redshift lower than 0.045, and they constitute a representative subsample of the galaxies in the complete UCM survey. From the new data, complemented with our own Gunn-$r$ images and long-slit optical spectroscopy, we have measured integrated $K$-band luminosities, $r-J$ and $J-K$ colours, and H$\\alpha$ luminosities and equivalent widths. Using a maximum likelihood estimator and a complete set of evolutionary synthesis models, these observations have allowed us to estimate the strength of the current (or most recent) burst of star formation, its age, the star-formation rate and the total stellar mass of the galaxies. An average galaxy in the sample has a stellar mass of 5$\\times$10$^{10}$\\,M$_{\\sun}$ and is undergoing (or recently completed) a burst of star formation involving about 2~per cent of its total stellar mass. We have identified two separate classes of star-forming galaxies in the UCM sample: low luminosity, high excitation galaxies (HII-{\\it like}) and relatively luminous spirals galaxies (starburst disk-{\\it like}). The former show higher {\\it specific} star formation rates (SFR per unit mass) and burst strengths, and lower stellar masses than the latter. With regard to their {\\it specific} star formation rates, the UCM galaxies are intermediate objects between normal quiescent spirals and the most extreme HII galaxies. ", "introduction": "\\label{introduction} The study of the evolution of the Star Formation Rate (SFR) of individual galaxies and the SFR history of the Universe has experienced considerable progress recently (see, e.g., Madau, Dickinson \\& Pozzeti 1998 and references therein). These are key observables needed to extend our understanding of galaxy formation and evolution. In the last few years, the combination of very deep ground-based and HST multi-band imaging with deep spectroscopic surveys carried out with 4-m and 10-m class telescopes has allowed the sketching of the SFR history of the Universe up to $z>4$ (see, e.g., Madau et al. 1998 and references therein). A great deal of effort has been devoted to both observational and theoretical studies of star-forming objects and their evolution with look-back-time. Deep imaging and spectroscopy of faint galaxies at intermediate and high redhsifts have yielded vast amounts of quantitative information in this field (Lilly et al. 1995; 1998 and references therein; Driver, Windhorst \\& Griffiths 1995; Steidel et al. 1996; Lowenthal et al. 1997; Hammer et al. 1997; Hu, Cowie \\& McMahon 1998; see Ellis 1997 for a recent comprehensive review). Although substantial uncertainties still exist, a reasonably coherent picture is emerging. The Star Formation Rate density of the universe was probably about an order of magnitude higher in the past than it is now, perhaps peaking at $z\\sim1$--$2$ (e.g., Gallego et al. 1995; Madau et al. 1996; Connolly et al. 1997; Madau et al. 1998). These observational results seem to be in good agreement with the predictions of recent theoretical models of galaxy formation (Pei \\& Fall 1995; Baugh et al. 1998; Somerville, Primack \\& Faber 1999), although the question of whether the SFR density decreased beyond $z\\sim2$ is still a matter of intense debate (Hu et al. 1998; Somerville et al. 1999; Hughes et al. 1998; Barger et al. 1998; Steidel et al. 1999). Given the large redshift range covered by these studies, different SFR indicators have perforce been used, all of which have different calibrations, selection effects and systematic uncertainties. These indicators include emission line luminosities (e.g., H$\\alpha$, H$\\beta$, [OII]$\\lambda3727$\\AA), blue and ultraviolet fluxes, far-infrared and sub-mm fluxes, etc (see, e.g., Gallego et al. 1995; Rowan-Robinson et al. 1997; Tresse \\& Maddox 1998; Glazebrook et al. 1999; Treyer et al. 1998; Madau et al. 1996; Connolly et al. 1997; Hughes et al. 1998; Barger et al. 1998; see also Charlot 1998 and Kennicutt 1992). It would be highly desirable to use the same SFR indicator at all redshifts, so that the problems related to different selection effects and systematics could be avoided. It is widely accepted that the H$\\alpha$ is one of the most reliable measurements of the current star formation rate ({\\it modulo\\/} the IMF; see, e.g., Kennicutt 1992). Several groups have used the H$\\alpha$ line to estimate SFRs at different redshifts, from the local universe to beyond $z=1$ (Gallego et al. 1995; Tresse \\& Maddox 1998; Glazebrook et al. 1999), albeit with very different sample selection methods. Nevertheless, it is clear that it is now necesary to build sizeable samples of H$\\alpha$-selected star-forming galaxies at different redshifts {\\it and} study their properties. One would like to know the preferred sites of star formation in the local universe and beyond, and the main propeties of the star-forming galaxies and their evolution. Some questions that need to be answered include: does star formation mainly occur in dwarf, starbursting galaxies or in more quiescent, normal L$^*$ galaxies? how has that evolved with time? what fraction of the stellar mass of the galaxies is being built by their current star-formation episodes? Progress towards answering questions such as these requires, as a first step, a comprehensive study of the properties of the star-forming galaxies in the local universe. The Universidad Complutense de Madrid survey (UCM hereafter; Zamorano et al. 1994, 1996) is currently the most complete local sample of galaxies selected by their H$\\alpha$ emission (see section~\\ref{ucm}). It has been used to determine the local H$\\alpha$ luminosity function, the SFR function and the SFR density (Gallego et al. 1995). It is also widely used as a benchmark for high redshift studies (e.g., Madau et al. 1998 and references therein). Thus, the UCM survey provides a suitable sample of local star-forming galaxies for detailed studies. Both optical imaging (Gunn-$r$; Vitores et al. 1996a, 1996b) and spectroscopy of the whole UCM sample (Gallego et al. 1996; GAL96 hereafter; see also Gallego et al. 1997) are already available. The optical data provides information on the current star-formation activity, but is rather insensitive to the past star-formation history of the galaxies. In this paper we present new near-infrared imaging observations for a representative subsample of UCM galaxies. The near infrared luminosities are sensitive to the mass in older stars, and therefore provide a measurement of the integrated past star formation in the galaxies and their total stellar masses (see, e.g., Arag\\'on-Salamanca et al. 1993; Alonso-Herrero et al. 1996; Charlot 1998). Alonso-Herrero et al. (1996; AH96 hereafter) carried out a pilot study of similar nature with a very small sample. We will now extend the work to a galaxy sample that is large enough for statistical studies, and that is expected to represent the properties of the complete UCM sample and thus those of the local star-forming galaxy population. We will also improve the work of AH96 in two fronts: first, we will use up-to-date population synthesis models, and second, we will use a more sofisticated statistical technique when comparing observational data and model predictions. In section~\\ref{ucm} we briefly introduce the UCM sample. In section~\\ref{observations} the observations, reduction procedures, and data analysis are described. The evolutionary synthesis models are presented in section~\\ref{models}, and the results are described in section~\\ref{results}. Finally, section~\\ref{summary} contains a summary of this work. ", "conclusions": "\\label{summary} Using new nIR observations and published optical data for 67 galaxies from the Universidad Complutense de Madrid (UCM) survey, we have derived the main properties of their star-forming events and underlying stellar populations. This sample represents about 35~per cent of the UCM galaxies covering the whole range of absolute magnitudes, H$\\alpha$ luminosities and equivalent widths spanned by the survey. Burst strengths and ages, stellar masses, stellar mass-to-light ratios and, to a certain extent, metallicities, have been obtained by comparing the observed $r-J$ and $J-K$ colours, $K$-band magnitudes, and H$\\alpha$ equivalent widths and luminosities with those predicted by evolutionary synthesis models. The comparison of the observations with the model predictions was carried out using a maximum-likelihood estimator in combination with Monte Carlo simulations which take into account the observational uncertainties. Our main results are: \\begin{enumerate} \\item The star-forming galaxies in the UCM sample (used to determine the SFR density of the local universe), show typical burst strengths of about 2~per cent and stellar masses of 5$\\times$10$^{10}$\\,M$_{\\sun}$. The current star formation in these galaxies is taking place in discrete star formation events rather than in a continuous fashion. If this is typical of the past star-formation history in the galaxies, many of such star formation events would be necessary to build up their stellar mass. However, our observations provide very little information on star-formation episodes that took place before the current one. \\item We have identified two separate classes of star-forming galaxies in the UCM sample: SB-{\\it like} and HII-{\\it like} galaxies. Within the HII-{\\it like} class the DHIIH and BCD spectroscopic type galaxies, i.e. {\\it dwarfs}, constitute the most extreme case. The mean burst strength deduced for the SB-{\\it like} galaxies is about a 25~per cent lower than for the {\\it dwarf} HII-{\\it like} galaxies. The average stellar mass is an order of magnitude larger in the former than in the latter. The SB-{\\it like} galaxies are relatively massive galaxies where the current star formation episode is a minor event in the build up of their stellar masses, while HII-{\\it like} galaxies are less massive systems in which the present star formation could dominate in some cases their observed properties and contributes to a greater extent to their stellar population. \\item Because of the low burst strengths inferred, $K$-band luminosity is dominated by the old stellar populations, and the $K$-band stellar mass-to-light ratio is almost the same (within $\\sim$20~per cent) for all the galaxies. Thus, the $K$-band luminosity is a very good estimator of the stellar mass for typical star-forming galaxies. \\item The average SFR of the galaxies is log(SFR)$\\simeq$1.5, with the SFR expressed in M$_{\\sun}$\\,yr$^{-1}$, and it is similar for the SB-{\\it like} and the HII-{\\it like} galaxies. However, since the latter are typically less massive, their specific SFR (SFR per unit stellar mass) is significantly larger, in a factor 2.3, than that of the former. \\item The UCM galaxies represent a bridge in {\\it specific} SFR between relaxed spirals and extreme HII galaxies. The range in {\\it specific} star formation rate spanned by the local star-forming galaxies, (10--10$^{3}$)$\\times$10$^{-11}$\\,yr$^{-1}$, is very similar to that observed in higher redshift objects. \\end{enumerate}" }, "0003/astro-ph0003271_arXiv.txt": { "abstract": "{\\small A vital ingredient of DNA molecule named adenine may be produced by successive addition of HCN during molecular cloud collapse and star formation. We compute its abundance in a collapsing cloud as a function of the reaction rate and show that in much of the circumstances the resulting amount may be sufficient to contaminate planets, comets and meteorites. We introduce a $f$-parameter which may be used to study the abundance where radiative association takes place.} ", "introduction": "In recent papers, Chakrabarti [1], and Chakrabarti and Chakrabarti ([2], hereafter referred to as Paper I) explored the possibility of the formation of biomolecules in star formation region using gas-phase chemistry. Their conclusion was that even in frigid condition in interstellar matter some of the simplest amino acids such as glycine, alanine etc. could be produced even before the formation of stars and planets. Paper I also showed that with a choice of reaction rate constant $10^{-10}$, significant adenine may also be produced. Some preliminary results of amino acids are in [1]. Adenine is a simply produced vital component of the DNA molecule and its significant production may point to an important clue into the problem of origin of life on planets like ours. Because of this, it is essential to carry out careful analysis on the reaction rates during the adenine formation. In Paper I, we used the formation of adenine by successive addition of HCN [3] by using an `average' rate. In the normal circumstances, in gas-phase reaction $HCN+HCN \\rightarrow H_2C_2N_2$ rate would be small, since they must combine by radiative association, i.e., they must radiate a photon when combined together. This is a slow process and the probability of photon emission could be $1$ in a few thousand to a few million (T. Millar, private communication). However as the size of the molecule gets bigger, the process becomes faster. Thus, it is likely that for a large enough molecule, the radiative association may take place at every collision and at this stage, the collisional rate may be used. One possibility is therefore to assume that after every addition of HCN, the reaction rate goes up by a factor of $f$ ($f$ may be anywhere from $1$ to $100$ or more). Hence one may imagine that at the early stages, $HCN+HCN\\rightarrow H_2C_2N_2$ forms with a reaction rate of $10^{-16}$, but for $HCN+H_2C_2N_2$ the rate becomes $f\\times 10^{-16}$, for $HCN+H_3C_3N_3$ the rate becomes $f^2 \\times 10^{-16}$ and so on. It would be therefore interest to learn, whether significant adenine is formed and it is detectable when the radiative association process is taken into account. In the present paper, we do just that. It is possible that more favorable reactions take place on ice, but in view of little known reaction rates of ice chemistry we believe that the best we could do is to study the formation of these important molecules as a function of two parameters, namely, $\\alpha_{Ad}$ and $f$. It is quite possible that a suitable $f$ parameter we suggested above would take care of the ice-chemistry reaction rates as well. It is still possible that such an $f$ may actually be determined by actual detection of molecules in space. Similarly, constancy of $f$ is an assumption of our model. In reality it could vary with the size of the molecules. So far, there has been controversy whether glycine has been observed in interstellar matter. Miao et al. [4] tentatively detected glycine in the massive star forming region Sgr B2(N) though this was later challenged by Combes et al. [5] who suggested that with the sensitivity of the detector taken into account, the lines were really at the confusion limit and positive identification would require more sensitive instruments. It is not known if any attempts were made to detect adenine lines, however there may have been detection of adenine in meteoritic samples (M. Bernstein, private communication). ", "conclusions": "In presence of radiative association, adenine abundance $X_{Ad}$ in an interstellar cloud seems to be roughly proportional to $\\alpha_{Ad}^4 f^6$ for small $X_{Ad}$. This means that the measurements of both $\\alpha_{Ad}$ and $f$ must be made very accurately. We studied the $\\alpha-f$ parameter space and found that while some region could produce significant abundance, a smaller region produce detectable (with present day technology) amount, while the rest produces abundance insignificant enough to dismiss the contamination theory. One must wait for the technological advancements to improve laboratory experiments in extreme conditions and to improve the detectability limit in order to come to a firm conclusion. \\vspace{0.3cm} \\noindent {\\bf Acknowledgments}\\\\ \\noindent SC acknowledges the usages of the facilities of Centre for Space Physics for writing this article." }, "0003/astro-ph0003265_arXiv.txt": { "abstract": "In 1998 and 1999 the Whipple Observatory 10 m telescope was used to search for diffuse gamma ray emission from the Galactic Plane. At this time, the telescope was equipped with a large ({$\\rm4.8^o$}) field of view camera, well suited to detect diffuse $\\gamma$-ray emission. No signifiant evidence of emission was found. Assuming the TeV emission profile matches EGRET observations above 1~GeV with a differential spectral index of $2.4$, we derive an upper limit of { $\\rm {3.0\\cdot10^{-8}\\:cm^{-2}\\:s^{-1}\\:sr^{-1}}$} for the average diffuse emission above {{$\\rm500~GeV$}} in the galactic latitude range from {{$\\rm-2^o$ to $\\rm +2^o$}} at galactic longitude {{$\\rm 40^o$}}. Comparisons with EGRET observations provide a lower limit of 2.31 for the differential spectral index of the diffuse emission, assuming there is no break in the spectrum between 30~GeV and 500~GeV. This constrains models for diffuse emission with a significant inverse Compton contribution. ", "introduction": "High energy $\\gamma$-rays traverse the Galaxy without significant attenuation, and hence diffuse $\\gamma$-ray emission probes the galaxy as a whole. As was observed with OSO-3 \\cite{krau72}, emission from the galactic plane is the main feature of the high energy $\\gamma$-ray sky. Observations with the SAS-2 and COS-B satellites showed this emission was generally correlated with the spatial structures in the Galaxy seen at other wavelengths (\\cite{fich75}; \\cite{hart79}; \\cite{maye80}). The greater sensitivity and angular resolution of the EGRET instrument (\\cite{hugh80}; \\cite{camb88}; \\cite{camb89}) provided a much more detailed spectral and spatial picture of the diffuse emission from the Galaxy (\\cite{hunt97}). The spectral and spatial distributions of diffuse high energy emission can be interpreted using a detailed model (\\cite{bert93}) of the emission. This model is based on $\\gamma$-ray production by nucleon-nucleon and electron (bremsstrahlung and inverse Compton) scattering in the interstellar medium. The matter and photon distributions in the interstellar medium are inferred from radio, infrared, visible and ultra-violet observations. The cosmic rays are assumed to have the same composition and spectrum everywhere in the galaxy with a density that is correlated with the matter density. The model precisely reproduces the spatial and spectral data from EGRET up to energies of 100 MeV. However, for $\\rm E>1~GeV$ the predicted flux is 40\\% lower than observations indicate (\\cite{hunt97}). \\cite{pohl98} explained the observed excess at $\\rm E>1~GeV$ as inverse Compton (IC) emission by a cosmic-ray electron injection spectrum harder than that used by Hunter et al.(1997). The latter assumed a spectral index of 2.4, which was inferred from the local cosmic-ray electron spectrum when propagated through the galaxy. If shock acceleration in supernova remnants (SNRs) produces the electrons, a spectral index closer to 2.0 is expected. Furthermore, most SNRs have power law spectra at radio wavelengths with $\\alpha \\sim 0.5$ (\\cite{gree95}), indicating injection spectra with index near 2.0. \\cite{pohl98} showed that a change of spectral index of only 0.4 is sufficient to account for the observed excess. Indeed, it has been previously argued that if the electron cutoff energy is high enough, IC emission may be the dominant source of diffuse $\\gamma$-rays at TeV energies (\\cite{port97}). \\cite{pohl98} suggest that the discrepancy with the local electron spectrum is a statistical fluctuation: high energy cosmic-ray electrons tend to have a local origin because bremsstrahlung and IC energy losses prevent propagation over large distance. Therefore, they are subject to Poisson fluctuations in the number of supernovae accelerating electrons at a given time, whereas the unattenuated $\\gamma$-rays originate from broader parts of the Galaxy and reflect an electron spectral index closer to the galactic average. In examining possible mechanisms for the EGRET excess, \\cite{mos00} and Strong, Moskalenko and Reimer (1999, 2000) apply several detailed models to the full spectrum of diffuse gamma-ray emission from the Galaxy. These authors also conclude that IC emission from a hardened electron injection spectrum ($<2$) is the favored origin of the high energy EGRET excess. We have searched for diffuse emission from the Galactic Plane at higher energies than detected by EGRET using the Whipple Observatory 10 m imaging atmospheric Cherenkov telescope to observe the galactic plane region at ($b=0^o$, $l=40^o$). Although a simple extrapolation of the highest energy EGRET spectrum indicates the diffuse emission should be detectable, no signal was found. Using an analysis method specifically developed for this measurement we derive upper limits on the galactic diffuse $\\gamma$-ray emission at $\\rm E>500~GeV$. We compare this result to a power law extrapolation of EGRET's spectral points above 1~GeV and derive a lower limit on the spectral index. In sections 2 and 3 the observations and data analysis are described. In section 4, the results are presented and discussed in the context of the EGRET results summarized above. ", "conclusions": "The observations of the galactic plane did not yield a detection of the diffuse emission, and we derived an upper limit of $\\rm 3.0\\cdot10^{-8}s^{-1}cm^{-2}sr^{-1}$ for the flux above 500~GeV in the portion of galactic plane defined by {$\\rm 38.5^o 10^{6}$ K) post-shock stellar wind material. In the absence of thermal evaporation of material at the outer edge, the electron density within the MS bubble should be of order $10^{-3}$ cm$^{-3}$. After $4.5 \\times 10^{6}$ years the star evolves off the main sequence and into a red supergiant (RSG). Over its lifetime of $2 \\times 10^{5}$ years , the RSG wind fills the inner few parsecs of the MS bubble with cold dense material. The characteristic mass loss rate of the RSG phase is $ \\sim 10^{-4}$ M$_{\\odot}$ yr$^{-1}$ with terminal velocity 15 km s$^{-1}$. Between the free-streaming RSG wind and the MS bubble a thin dense shell of post-shock RSG material forms. Lastly the star enters its Wolf-Rayet phase, with mass loss rate $ \\sim 10^{-4.5}$ M$_{\\odot}$ yr$^{-1}$, with V$_{\\infty} \\sim$ 2000 km s$^{-1}$ over its lifetime of $2 \\times 10^{5}$ years. The WR wind progressively sweeps up the RSG material, eventually overtaking the outer RSG shell. The resulting collision fractures the outer shell and the WR bubble breaks out into the MS bubble material. At this point the nebular shell reaches maximum brightness due to its high density; according to their model, the nebula should be detectable over a period of order 10$^{4}$ years. The morphology and dynamics of WR shells are important not only as laboratories for studying wind interactions with the ISM, but also for the insights they provide into the mass-loss history of massive stars as well as the environments into which remnants of core collapse supernovae will propagate and interact. In this paper, we use observations of a region of NGC 6888 obtained with the {\\it Hubble Space Telescope} WFPC2 to study the ionization structure and infer physical conditions within the shell. ", "conclusions": "NGC 6888 is a bubble formed by the mass loss of the precursor to the central star WR 136. The internal pressure of the shocked WR stellar wind has swept up the ejecta from the previous RSG phase. This material collides with a thin RSG shell formed at the MS bubble interface, fragmenting into the clumps seen today. Models of spatial profiles of selected clumps appear to be fully ionized, resulting in a log Q$_{0} >$ 49.2 for the central star, and average abundances for the shell material of [N, O, S] = [8.1, 7.8, 6.8]. The oxygen depletion appears even more severe than that seen in previous spectroscopic studies. The internal pressure of the stellar wind drives a shock into the material external to the visible shell. The shock is visible in [\\ion{O}{3}] and H$\\alpha$ as a skin enveloping the clump network. The density of this material suggests that the MS bubble cooled efficiently. Combined with the mass of the visible shell compared to the total mass loss of the RSG phase, we argue that thermal evaporation of material from the outer RSG shell has already occurred. Comparing the pressure inside and outside the nebular shell suggests that a radical drop in pressure has recently occurred, and that the clumps have not yet come into equilibrium with the current pressure of the stellar wind." }, "0003/astro-ph0003323_arXiv.txt": { "abstract": "We present self-similar solutions for advection-dominated accretion flows with radial viscous force in the presence of outflows from the accretion flow or infall. The axisymmetric flow is treated in variables integrated over polar sections and the effects of infall and outflows on the accretion flow are parametrised for possible configurations compatible with the self-similar solution. We investigate the resulting accretion flows for three different viscosity laws and derive upper limits on the viscosity parameter $\\alpha$. In addition, we find a natural connection to non-rotating and spherical accretion with turbulent viscosity, which is assumed to persist even without differential rotation. Positive Bernoulli numbers for advection-dominated accretion allow a fraction of the gas to be expelled in an outflow and the upper limit on the viscosity predicts that outflows are inevitable for equations of state close to an ideal gas. ", "introduction": "Advection-dominated accretion flows (ADAFs) have been invented to explain low-luminosity black hole candidates like Sgr A$^*$ \\citep{na98} in our Galactic Center. The low-luminosity of this model is achieved in an optically thin plasma, where most of the energy is stored in hot ions, while electrons as potential radiators are inefficiently coupled to the heat source and remain relatively cold. The electrons become nonetheless mildly relativistic close to the central black hole and the inevitable synchrotron radiation is observed from most ADAF candidates. The presence of magnetic fields not far from energy equipartition with the gas is indicative of their origin in MHD-instabilities \\citep{bal91} leading to turbulence in the accretion flow and subsequent generation of an effective viscosity. On larger scales magnetic fields are likely to be responsible for the collimation of outflows from accretion disks into jets, seen in the cores of M87 \\citep{rey96} and NGC 4258 \\citep{lasota96,her98}, which are prototypical ADAF candidates. Furthermore the model can explain the accretion in some low-luminosity AGNs of elliptical galaxies \\citep{dim99} and in NGC 4258 \\citep{gam99}. In a recent X-ray survey \\citep{sam99} a few more examples have been found for low-luminosity core in radio-loud AGNs, which are candidates for advection-dominated accretion flows in their central engines, suggesting that ADAFs can be found even in radio-loud AGNs and that jet formation is a common feature. Outflow models for ADAFs have been investigated by \\citet{blan99} and applied to several candidates \\citep{dim99, qua99}. On the theoretical side \\citet{igu99} and \\citet{stone} have performed time-dependent 2D calculations of accretion flows, which in some cases resemble ADAFs for certain viscosity parameters $\\alpha \\approx 0.1$, but suggest the production of outflows for larger $\\alpha$ \\citep{igu99}. It is found that the $(rr)$ stress tensor component, which was not included in the original description of vertically integrated models for accretions flows, is important in the cited calculations for ADAFs. The existence of self-similar solutions with a radial viscous force has been shown in previous work \\citep{ny95} and discussed for 2D solutions with a separation of variables. In this paper we describe advection-dominated accretion flows in polar-integrated variables including the radial viscous braking force, which either produce outflows or are formed by wind infall. The wind infall is assumed to consist of free falling low angular momentum gas, which is cold with respect to the already existing accretion flow. One possible source for this gas are stellar winds from massive stars in young clusters as in the center of our galaxy. In the following we will refer to this wind infall into the accretion flow as infall. In our treatment the main difference between infall and outflow is the sign of the mass infall rate and we talk about winds, if it is not necessary to distinguish between infall and outflow. We restrict the discussion to an extension of the self-similar solutions given by \\citet{ny94} for the Newtonian limit. In \\S\\ref{WO} we present the equations, which describe the accretion flow including the reaction to winds. The role of $(rr)$ stresses and bulk viscosity is emphasised. We discuss angular momentum transport and viscosity in \\S\\ref{Vis} and specify the possible equations of state and the resulting energy equation in \\S\\ref{equStat}. General features of self-similar solutions are presented in \\S\\ref{SelfS} and in \\S\\ref{alphaD} detailed solutions for the $\\alpha$-viscosity law are shown. Consequences of the alternative $\\beta$-viscosity are discussed in \\S\\ref{BetaDisk} and ADAFs with an intermediate shear-limited viscosity law follow in \\S\\ref{shearADAF}. We compare the solutions and draw our conclusions in \\S\\ref{Conclud}. ", "conclusions": "We have investigated the importance of a radial viscous braking force for advection-dominated accretion flows (ADAFs) in the presence of infall or outflows. Under the assumption of almost isotropic turbulence, bulk viscosity and $(rr)$ component of the viscous stress tensor provide an efficient brake of the rapid radial accretion of hot gas in ADAFs and produce additional heat due to viscous friction. This opens a second channel for transfer of kinetic into internal energy and supports in part the radial motion so that the shear due to differential rotation and the centrifugal barrier is reduced in ADAFs with radial viscous braking. We derived self-similar solutions of ADAFs for three different viscosity laws. The standard $\\alpha$-viscosity produces solutions, which show more and more sub-keplerian rotation with increasing $\\alpha$. The solutions terminate at a critical $\\alpha_c$, which depends strongly on the ratio of specific heats so that the greatest possible $\\alpha$ tends to zero for a non-relativistic ideal gas. At the critical $\\alpha_c$ the rotation of the flow vanishes and a purely radial inflow appears, which is not only supported by a pressure gradient, but also by the viscous braking force, provided the flow is still turbulent in the absence of differential rotation. This limit is reminiscent of Bondi accretion \\citep{bond52} in the presence of effective turbulent viscosity. The transition from ADAFs to Bondi accretion nonetheless affords a suspicious fine tuning in $\\alpha$. The $\\beta$-viscosity law of \\citet{dus20} based on geometrical arguments in the absence of shock-limited turbulence allow ADAF solutions for all reasonable values of $\\beta$. The estimates of $\\beta$ as inverse of the critical Reynolds number of the flow suggests small $\\beta$s, for which the solutions do not differ from $\\alpha$-ADAFs. No transition to a Bondi like flow is possible in this case and no upper bound on $\\beta$ exists. We showed that it is possible to derive a shear-limited viscosity law from the $\\beta$-viscosity mentioned above, for which the transition to non-rotating Bondi like accretion with turbulent viscosity occurs naturally. The now familiar ADAF solutions with sub-keplerian rotation at $\\hat{\\alpha} \\rightarrow 0$ join with a second branch of solutions at a maximal sustainable $\\hat{\\alpha}$. Similar to the $\\alpha$-ADAFs no accretion flow with larger $\\hat{\\alpha}$ are possible. The second solution branch is a hot, slowly rotating, and rapidly accreting solution even for small values of $\\hat{\\alpha}$. The transition to viscous Bondi accretion occurs from the slowly rotating to the non-rotating solution in the limit $\\hat{\\alpha} \\rightarrow 0$, where $\\Omega$ is linear in $\\hat{\\alpha}$. A finite viscous force is present in these flows even in the limit $\\hat{\\alpha}=0$. In all cases the connection of ADAFs to non-rotating flows depends on the presence of a radial viscous force. The observation that ADAFs generally possess positive Bernoulli numbers lead to the idea \\citep{ny95, blan99} that ADAFs are good candidates for the production of outflows. In that way the accretion flow loses energy to the outflow and the remaining material is left with negative Bernoulli numbers and gravitationally bound to the central accreting mass. We confirm that statement in the presence of a radial viscous force for $\\alpha$- and $\\beta$-ADAFs and show the back-reaction of outflows on ADAFs for different outflow characteristics. Most noticeably is the cooling effect and the increased accretion velocity of the remaining ADAF as angular momentum and internal energy is carried away. Outflows with a minimal energy assumption for the extracted energy have to be fairly massive $p\\ga 0.35$ to lower the Bernoulli number and leave a bound flow in case of the shear-limited viscosity. It is much easier to get a bound flow, if the minimal energy assumption is violated and the terminal velocity of the outflow is the escape speed from the origin of the out-flowing material. If a natural choice of the viscosity parameter---either $\\alpha$ or $\\hat{\\alpha}$---exists and provided the strength of the radial viscous force can be estimated from isotropic turbulence, then outflows are inevitable for certain equations of state with ratios for specific heats close to $5/3$. This conclusion is independent of arguments based on the positiveness of Bernoulli numbers for ADAFs. The scenario of thin disk evaporation in binary systems \\citep{liu99} or the transition from cooling flows to ADAFs in low-luminosity cores of elliptical galaxies \\citep{qua99b} can explain the existence of ADAFs in these systems. The formation of an ADAF, which is the most promising model for the spectral energy distribution of Sgr A$^*$ in the Galactic Center, cannot proceed in either way. We suggest that an ADAF in the Galactic Center forms out of stellar wind infall \\citep{cok99} and the transfer of kinetic energy of the infall into internal energy of an advection-dominated flow. The even larger Bernoulli numbers produced in this way wound give rise to subsequent outflows and reduce the mass accretion rate inferred from infall calculations by \\citet{cok99} to the smaller accretion rates predicted from spectral fitting \\citep{qua99} of ADAF models to Sgr A$^*$. {" }, "0003/astro-ph0003115_arXiv.txt": { "abstract": "CCD photometry on the $uvbyCa$ system has been obtained for 521 candidate metal-poor stars from the northern hemisphere HK survey of Beers and colleagues. Reddening corrections, classifications of stars by luminosity class, and the derivation of metal abundances based on Str\\\"omgren indices $m_1$ and $hk$ are described and presented, along with preliminary spectroscopic estimates of abundance. A number of extremely metal-poor stars with [Fe/H]$ \\leq -2.50$ are identified. ", "introduction": "A common goal of Galactic structure studies is to understand the events and processes that have characterized the formation and ongoing evolution of the Milky Way. Surviving relics of early star formation in the Galaxy offer particularly rich rewards, but these rewards only come as a result of relatively large observational effort. Clues from the first epochs of star formation are scarce and challenging to decode, but they can be explored via examination of the stars of the halo field and in the populous globular clusters. Over the past few decades, researchers have probed the spatial, kinematic and chemical distribution of the stars formed over the halo's history. Since the seminal explorations of Eggen, Lynden-Bell, \\& Sandage (1962)\\markcite{els}, the amount and quality of available data has grown to the point that a rather more complicated picture of the formation of the Galaxy than that based on a simple, rapid, collapse of a pre-Galactic spheroid of gas now applies (see, e.g., Chiba, Beers, \\& Yoshii 2000\\markcite{chi}). While the pioneering work of Eggen et al. focused on the properties of a kinematically selected sample of nearby stars, one must concede that, in many ways, Galactic globular clusters are ideally suited for mapping out this history. The advantages of the globular clusters are well documented in the literature. Each globular cluster, comprised of $10^5$ to $10^6$ stars which share a common age, distance, and (usually) chemical composition permits determination of these fundamental properties with minimized statistical errors. Furthermore, globular clusters are readily observable at distances from the Galactic center out to many tens of kiloparsecs. Relative ages, precise to 1 Gyr or better, have been reported for halo clusters (Sarajedini, Chaboyer, \\& Demarque 1997\\markcite{sd}; Buonanno et al. 1998\\markcite{bcp}). The ``merger model'' for halo formation (Searle \\& Zinn 1978\\markcite{sz}), as well as the existence of a thick disk population of Galactic globulars is supported by the kinematic and spatial distribution within the globular cluster population (Zinn 1993\\markcite{zi}; Da Costa \\& Armandroff 1995\\markcite{dca}). There remains, however, one devastating weakness in the effort to reconstruct the history of Galactic evolution with globular cluster data alone -- there exist no more than $\\sim 200$ globulars in the Milky Way. Furthermore, there is no globular cluster located within one kpc of the Sun, thus restricting high-resolution stellar studies to the brighter giants, and requiring significant commitments of large telescope time. In addition, there are no known clusters with abundances below [Fe/H]$ = -2.5$, in sharp contrast with the field star sample (Laird et al. 1988\\markcite{lcl}). The exceptionally metal-poor field stars, according to all models of Galactic nucleosynthesis, should probe the first generations of star formation within the protogalactic cloud. Thus, if one seeks to study the {\\it earliest} phase of Galactic evolution, field stars offer the best option. The picture that is now emerging from examination of nearly primordial stars is fascinating but puzzling. Abundances of the light elements seem well-described by the standard model of Big Bang nucleosynthesis (Olive \\& Steigman 1995\\markcite{sch}; Ryan et al. 2000\\markcite{rya}). For the heavier elements, stars with [Fe/H] $> -2.5$ have, within broad tolerances, well-defined patterns of abundance ratios at a given [Fe/H], presumably reflecting their origin within a reasonably well-mixed interstellar medium by the time these stars were formed (Wheeler, Sneden, \\& Truran 1989 \\markcite{wst}; Edvardsson et al. 1993 \\markcite{e93}; McWilliam et al. 1995 \\markcite{mcw}; Ryan, Norris, \\& Beers 1996\\markcite{rnb}). Stars with [Fe/H] $< -2.5$ (McWilliam et al. 1995 \\markcite{mcw}; Ryan et al. 1996\\markcite{rnb}) exhibit a more chaotic distribution in a number of their elemental ratios, which may be the earmark of an poorly-mixed early Galaxy. Recent attempts to model this apparently complex early history have centered on supernova-induced star formation and chemical evolution in individual proto-Galactic clouds (Tsujimoto, Shigeyama, \\& Yoshii 1999\\markcite{tsy} and references therein). Continued progress in this area is assured only if the sample of extremely metal-poor stars is expanded -- by no means an easy task given that such stars make up a tiny fraction of the already sparsely populated halo component among stars in the solar neighborhood. Early attempts based upon objective-prism spectroscopy (Bond 1980\\markcite{b1}; Bidelman \\& MacConnell 1973\\markcite{bm73}) generated only a handful of stars which proved to exhibit metal abundances near [Fe/H] $\\sim -3.0$. The most effective survey technique to date has been that developed by Beers, Preston, \\& Shectman (1985, hereafter BPS)\\markcite{bps} in which the wide-field survey capability of a Schmidt telescope, used with an objective prism, is combined with a filter isolating the spectral region near the ultraviolet lines of singly ionized calcium. The strong H and K resonance lines of CaII should be visible in even the most metal-poor stars, so stars with nearly continuous spectra in this region are good candidates for extreme metal deficiency. Unfortunately, blended stellar images, high temperatures, and low surface gravity can also produce weak H and K lines, thus one requires medium-resolution spectroscopy and/or photometric follow-up to isolate the best candidates. For an example of how severe this winnowing process can be, one need only refer to the initial efforts of BPS. In their first attempt, plate material covering 1940 square degrees of sky was used to generate a list of 2000 candidate stars; from an initial subset of 450 candidates, roughly 30\\% proved to have [Fe/H] below --2.0, with only a few stars of extreme ([Fe/H] $< -3.5$) metal paucity being detected. Note, however, that prior to this study, only one star was known with an estimated metallicity near --4.0 and virtually every major analysis of the primordial halo over the past decade has included stars from the HK survey of Beers and colleagues. Beers (2000)\\markcite{b20} discusses how the ``effective yield'' of such survey efforts can be improved, and points out that one can roughly double the efficiency of searches for stars with [Fe/H] $< -2.0$ by conducting a photometric pre-selection of candidate stars. The HK survey has now been expanded to include a total of some 7000 square degrees in the both the northern and southern hemispheres (see Beers 1999\\markcite{b99} for a summary). Medium-resolution spectroscopic follow-up of the first portion of this expanded sample has been discussed by Beers et al. (1992b, hereafter BPS II)\\markcite{bps2}. Photometric data for subsets of the long lists of candidates have been presented in Doinidis \\& Beers (1990,1991) \\markcite{db1}\\markcite{db2}, Preston, Shectman, \\& Beers (1991)\\markcite{psb}, Beers et al. (1992a)\\markcite{bpsd}, Schuster et al. (1996, 1999)\\markcite{s0}\\markcite{s99}, and Norris, Ryan, \\& Beers (1999)\\markcite{nrb}. In this paper we analyze $uvbyCa$ CCD photometry for over 500 low metallicity candidates from the HK survey, the majority of which were selected from northern survey plates. Section 2 describes the photometric system and the detector; Section 3 describes the acquisition and processing of the CCD data. In Section 4 the photometric classification and reduction to the standard $uvbyCa$ system are discussed. In Section 5 we describe estimates of the reddening corrections for these stars, and in Section 6 the derivation of [Fe/H] from $m_1$ and $hk$ indices is considered, along with a comparison to spectroscopically estimated abundances. Finally, Section 7 presents a summary of the principal results. ", "conclusions": "A subset of the extensive HK survey of candidate metal-poor stars has been studied with extended Str\\\"omgren photometry and medium-resolution spectroscopy. Photometry for 521 candidates stars formed the basis for analysis of foreground reddenings, luminosity classification and metallicity estimates based on $m_1$ and $hk$ indices. Although a number of stars were outside calibrated color and index limits, and others would exceed the metallicity limits of the photometric calibrations, photometric abundances for 63 red giants, 289 dwarfs and 48 blue horizontal branch stars have been estimated and compared to abundances based on spectroscopic analysis of the Ca II K line. Among the red giants, nearly a third of the candidates appear to have metal abundances below [Fe/H] = --2.5 and a handful may have abundances below [Fe/H] = --3.0. This class of stars is favored with two well-developed photometric calibrations for metallicity based on both $m_1$ and $hk$. A preliminary calibration developed by Baird (1999)\\markcite{bai} for RR Lyrae stars has been applied to the $hk$ indices and $(b-y)$ colors for the FHB candidate stars, identifying two of these candidates as having abundances below [Fe/H] = --3.0. Although photometric calibrations for main sequence stars based on $m_1$ are well established, a preliminary calibration for metal-poor dwarfs based on $hk$ indices was developed for the present investigation that is tied to the medium-resolution abundances based on the K line. In part because the photometric calibrations do not extend to uniformly extreme low values of [Fe/H], the yield of extreme candidates from this unevolved population is somewhat lower, with approximately 30 stars having one or both photometric abundances below [Fe/H] = --2.5. Approximately 20 stars have photometric indices indicative of more extreme metal-paucity. We close by presenting in Table 6 an abridged list of identifications, equatorial coordinates, $V$ and $(b-y)$ along with luminosity classification, an averaged photometric abundance estimate and the [Fe/H]$_{K3}$ abundance estimate. These 42 stars fit the following criteria: any of the two or three methods implying [Fe/H] $ \\leq -3.0$; an estimate of [Fe/H] $\\leq -2.5$ that is the only abundance estimate available; two of three abundance estimates below [Fe/H $ = -2.5$." }, "0003/hep-th0003244_arXiv.txt": { "abstract": "We study instanton solutions in general relativity with a scalar field. The metric ansatz we use is composed of a particular warp product of general Einstein metrics, such as those found in a number of cosmological settings, including string cosmology, supergravity compactifications and general Kaluza Klein reductions. Using the Hartle-Hawking prescription the instantons we obtain determine whether metrics involving extra compact dimensions of this type are favoured as initial conditions for the universe. Specifically, we find that these product metric instantons, viewed as constrained instantons, do have a local minima in the action. These minima are then compared with the higher dimensional version of the Hawking-Turok instantons, and we argue that the latter always have lower action than those associated with these product metrics. ", "introduction": "Consistent unified field theories which include gravity appear to indicate that the Universe has more than four spacetime dimensions. An obvious problem which follows is how to interpret these unseen extra dimensions? One approach that has been followed is to postulate that only four of these are observable, the extra dimensions have managed to become compact and are unobservably small. Recently however there has been a tremendous amount of interest in the effective five dimensional cosmologies associated with Branes, in which the fifth dimension can be macroscopic in size, yet remain unobservable at low energies. In general, these compactified spaces are assumed as part of the initial metric ansatz, and the cosmology of such metrics is then determined. Although this is a natural approach to take, it does not address the issue of whether such an initial condition is to be expected in string or M -theory, for example. Is there any way in which we can calculate the probability of the Universe possessing such compact internal dimensions as an initial condition? It would be of great interest if it could be shown that quantum cosmology predicts a manifold with compact extra dimensions as the most likely initial configuration. Symmetry arguments usually provide a very powerful tool for determining which instanton solutions should provide the dominant contribution (i.e. those with lowest Euclidean action) to the Hartle Hawking path integral \\cite{hartle}, hence providing the most likely background spacetime. An example is the Hawking-Moss instanton, involving a scalar field $\\phi$ with potential $V(\\phi$)\\cite{moss}. Assuming the potential had a stationary point at some non-zero value they obtained in four spacetime dimensions an O(5) symmetric instanton solution where $\\phi$ is constant and the Euclidean manifold is a four sphere. However, Coleman and De Lucia \\cite{coleman80} obtained an instanton solution of lower action with O(4) symmetry which was non-singular and corresponded to the nucleation of a bubble of true vacuum in a sea of false vacuum deSitter space. It was used in the earliest versions of open inflation \\cite{bucher95}, because the interior of such a bubble is in fact an open universe. Hawking and Turok \\cite{hawking98a} took these solutions one step further, dropping the requirement for non-singular instanton configurations; they obtained solutions where the scalar field potential increased monotonically from a single minimum. These solutions also allowed for a natural continuation to an open universe which was inflating. Moreover, although the instanton solutions themselves were singular their action was finite. Indeed they demonstrated a family of solutions which had lower action than the more symmetric O(5) solution! The notion that the O(4) symmetry of the Hawking-Turok instanton was responsible for the low action was tested in \\cite{saffin99}. Treating the instanton as a foliation of squashed rather than round three spheres, it was found that the O(4) instanton was the lowest action solution within this family. In an interesting paper Garriga \\cite{garriga98} proposed a resolution to the problem of having a singularity in the solution; singular instantons can arise from compactifications of regular higher dimensional instantons when viewed as lower dimensional objects. In this paper we investigate the nature of instanton solutions for the largest range of cosmologically relevant higher dimensional metrics that have been studied to date. Our results will be of relevance for the study of any higher dimensional model which involves compactifications on Einstein metrics, i.e. models of string cosmology involving compactifications on tori, supergravity compactifications on spheres and string theories where the compactified dimensions are Calabi-Yau manifolds. In particular we will be investigating instanton solutions arising from the metric ansatz, \\begin{eqnarray} \\label{M_metric} \\rm{ds}^2&=&\\rm{d}\\xi^2+a_{(1)}^2(\\xi)\\rm{ds}_{(1)}^2 +a_{(2)}^2(\\xi)\\rm{ds}_{(2)}^2 +...+a_{(T)}^2(\\xi)\\rm{ds}_{(T)}^2. \\end{eqnarray} The only restriction on the $ds_{(i)}^2$'s is that they are Einstein metrics on compact manifolds; the Ricci tensor is proportional to the metric. Of the many solutions that exist, we will see how a class of these instantons may be continued to a four dimensional inflating universe, with a number of static extra dimensions. \\indent In general, because of the non-linear nature of the equations, the solutions for the scale factors $a_{i}$ are obtained numerically, and from these we can study the action of the (generically singular) instantons. The most important result we obtain is that the family of singular instantons of this type can provide a local minima of the action for non trivial extra dimensions. However, it turns out that in all the cases we examined the action of these local minima remains {\\it higher} than that of the corresponding higher dimensional Hawking-Turok instanton. The implication of such a result is important. The symmetry properties associated with the Hawking-Turok instanton appear to determine the most likely instanton configuration, at least for the cases involving Einstein metrics. \\indent The layout of the rest of the paper is as follows: In section II we derive the field equations and action associated with our metric. Section III contains the results of our numerical and analytical analysis and presents the nature of the local minima of the action. It also contains the comparison of these instantons with the equivalent higher dimensional Hawking-Turok case and shows how the latter always lead to a lower Euclidean action. Section IV presents exact solutions for the case of a cosmological constant replacing the scalar field potential. We also mention the analytical continuation of our solutions to a space-time with a lorentzian signature and demonstrate the existence of solutions where the internal dimensions remain static while the four dimensional spacetime is inflating. We draw our conclusions in section V. ", "conclusions": "In this paper, we have derived the equations of motion for a specific warp product of general Einstein metrics. The main conclusion we can draw is that instantons which continue to spaces with compact `extra' dimensions of the form considered here do not have lower action than the corresponding higher-dimensional Hawking-Turok instanton. However, non trivial minima of the action do occur if the Einstein metrics {\\it all} have positive $\\Lambda_i$. These results are significant: First, they seem to indicate that the symmetry arguments used by Hawking and Turok in their letter can be applied to higher dimensional cases. Secondly, our analysis applies to a wide range of metrics and cosmological scenarios. Our particular comparison of the instantons involved n-dimensional spheres as our internal compact dimensions. Bishop's theorem then implies that these will provide the lowest possible action for such spacetimes with compact internal dimensions, hence our results apply to any Einstein metric -- they will always be beaten by the corresponding Hawking-Turok instantons. This result strongly suggests to us that if the initial quantum state of the universe were to be described by the `Hartle Hawking proposal' then it would be difficult to explain the presence of extra compact dimensions." }, "0003/astro-ph0003359_arXiv.txt": { "abstract": "In this paper we present evidence that the \\hii\\ regions of high luminosity in disk galaxies may be density bounded, so that a significant fraction of the ionizing photons emitted by their exciting OB stars escape from the regions. The key piece of evidence is the presence, in the \\ha\\ luminosity functions (LFs) of the populations of \\hii\\ regions, of glitches, local sharp peaks at an apparently invariant luminosity, defined as the Stromgren luminosity ($L_{\\rm Str}$), $L_{{\\rm H}\\alpha}$ = $L_{\\rm Str}$ = 10$^{38.6}$ ($\\pm$ 10$^{0.1}$) erg s\\me\\ (no other peaks are found in any of the LFs) accompanying a steepening of slope for $L_{{\\rm H}\\alpha}$$>$ $L_{\\rm Str}$. This behavior is readily explicable via a physical model whose basic premises are: (a) the transition at $L_{{\\rm H}\\alpha}$ = $L_{\\rm Str}$ marks a change from essentially ionization bounding at low luminosities to density bounding at higher values, (b) for this to occur the law relating stellar mass in massive star-forming clouds to the mass of the placental cloud must be such that the ionizing photon flux produced within the cloud is a function which rises more steeply than the mass of the cloud. Supporting evidence for the hypothesis of this transition is also presented: measurements of the central surface brightnesses of \\hii\\ regions for $L_{{\\rm H}\\alpha}$ $<$ $L_{\\rm Str}$ are proportional to $L_{{\\rm H}\\alpha}^{1/3}$, expected for ionization bounding, but show a sharp trend to a steeper dependence for $L_{{\\rm H}\\alpha}$ $>$ $L_{\\rm Str}$, and the observed relation between the internal turbulence velocity parameter, sigma, and the luminosity, $L$, at high luminosities, can be well explained if these regions are density bounded. If confirmed, the density-bounding hypothesis would have a number of interesting implications. It would imply that the density-bounded regions were the main sources of the photons which ionize the diffuse gas in disk galaxies. Our estimates, based on the hypothesis, indicate that these regions emit sufficient Lyman continuum not only to ionize the diffuse medium, but to cause a typical spiral to emit significant ionizing flux into the intergalactic medium. The low scatter observed in $L_{\\rm Str}$, less than 0.1 mag rms in the still quite small sample measured to date, is an invitation to widen the data base, and to calibrate against primary standards, with the aim of obtaining a precise, $\\sim 10^5$ L$_{\\odot}$, widely distributed standard candle. ", "introduction": "The basic theory for modeling a gaseous region round a hot star was first given by Zanstra (1931) for planetary nebulae, and applied to \\hii\\ regions by Str\\\"{o}mgren (1939), who quantified the relation between the radius of the ionized zone and the temperature--luminosity of the central exciting star. He showed that, in a uniform medium, the transition layer between fully ionized and neutral gas will be thin compared to the radius, a structure since termed a Str\\\"{o}mgren sphere. This will occur in a large placental neutral cloud, of sufficient dimension to absorb all the ionizing photons (those in the Lyman continuum in the case of atomic hydrogen), in which case the \\hii\\ region is ``ionization bounded\". The case where the cloud is not big enough to absorb all the ionizing radiation was first dealt with by Hummer \\& Seaton (1964) for planetary nebulae, which are then termed optically thin. An \\hii\\ region formed in these circumstances, where the cloud radius is less than that of the Str\\\"{o}mgren sphere, is termed ``density bounded\" or ``matter bounded\". There is ample evidence that \\hii\\ regions are not homogeneous in density. Measurements of local electron densities via line ratios are typically two orders of magnitude higher than mean electron densities estimated using diametral emission measure (for a clear recent example see Rozas, Knapen, \\& Beckman 1996), explained if a region comprises knots of high density embedded in a lower density plasma. The fractional volume dense enough to contribute measurably in emission lines such as \\ha\\ has been termed the ``filling factor\" ({\\it e.g.}, Osterbrock 1989). An \\hii\\ region with this structure will form a Str\\\"{o}mgren sphere provided that the local density variations are on scales which are small compared with the \\hii\\ region diameter, and that the mean density varies little on the scale of the diameter. This condition appears to hold well for regions over a wide range of \\ha\\ luminosities, up to a critical value, which we will be able to quantify from observations, as described below. The regions in this range are ionization bounded. At higher luminosities our evidence, which will be described in this paper, points towards an increasing tendency for the photon output from the ionizing stars to overflow the cloud in which they have formed. One type of evidence comes from the measurements of a change in slope of the luminosity function in \\ha\\ of complete populations of \\hii\\ regions in the set of nearby spirals. Previous detections of this change have been reported in the literature, notably in earlier work by Hodge, and by Kennicutt and co-workers (Hodge 1987; Kennicutt 1984; Kennicutt, Edgar, \\& Hodge 1989); and most recently by McCall and co-workers (McCall, Straker, \\& Uomoto 1996; Kingsburgh \\& McCall 1998) as well as ourselves (Rozas, Beckman, \\& Knapen 1996). It is interesting to note, however, that the change in slope varies little from object to object. We will elaborate this point during the development of the present paper. Statistical studies of the relation between the \\ha\\ luminosities, $L_{{\\rm H}\\alpha}$, of complete samples of \\hii\\ regions in nearby large spirals, and their volumes, (Cepa \\& Beckman 1989, 1990; Knapen {\\it et al.} 1993; Rozas, Beckman, \\& Knapen 1996), show that to a first approximation these are proportional. Taken as a complete description of the observations this would imply, at least statistically, two properties: (1) The \\hii\\ regions obey a single physical regime, which should be that of ionization bounding, and (2) the densities of the clouds in which the \\hii\\ regions are formed differ little from cloud to cloud within a galaxy, and from galaxy to galaxy. We will see that at very high luminosities departure from this first order behavior is dominant. Illustration of these points may be found in a number of articles in the literature, where the \\ha\\ luminosities of individually measured \\hii\\ regions have been plotted, logarithmically, {\\it vs.} the cubes of their radii (Cepa \\& Beckman 1989, 1990; Knapen {\\it et al.} 1993; Rozas Knapen, \\& Beckman 1996; Rozas {\\it et al.} 1999). In these articles, it is shown that for eight galaxies there is a close to linear relation between the \\ha\\ luminosity and the volume of the \\hii\\ regions in the range 37.5 $<$ log $L_{{\\rm H}\\alpha}$ $<$ 39.5, as pointed out in Cepa \\& Beckman (1990). This implies that the product of the filling factor, the electron density and the ionized hydrogen density does not change very much over this range, although in all the galaxies there is a clear tendency for it to increase at the highest luminosities. It is not entirely ruled out that the filling factor and mean density might vary independently with a relation which leaves their product invariant, but this would amount to a ``conspiracy\", and it is more reasonable to assume that none of the cited parameters varies strongly. The general picture is of \\hii\\ regions as spongy structures, which may indeed have fractal characteristics as suggested by Elmegreen (1997), but with a degree of clumping, and an average density which do not depend strongly on luminosity. However in all the plots of log $L_{{\\rm H}\\alpha}$ {\\it vs.} $r^3$ cited above, there is a significant trend to steeper gradients at higher luminosities, which means that here the densities and/or the filling factors of the regions must be increasing. The evidence which we present below points to a rather sharp change in the properties of the \\hii\\ regions occurring at an \\ha\\ luminosity which appears to vary very little from galaxy to galaxy. We will show that one would in fact expect, on reasonable physical grounds, the most luminous regions to be producing more ionizing photons than can be absorbed in their placental clouds, whereas this should not be true globally for the less luminous regions. We can predict, from a model in which there is a transition between ionization bounding at lower luminosities, and density bounding at higher luminosities, that there should be an accumulation of \\hii\\ regions around the luminosity of the transition (which we have termed $L_{\\rm Str}$, \\ie\\ the Str\\\"{o}mgren transition), between the two regimes, and this accumulation is in fact found as a ``glitch'' in the \\ha\\ luminosity function of the galaxies observed. It is by no means easy to reproduce these observations with alternative hypotheses. Two aspects of the density-bounding phenomenon would be of particular interest. Since the most luminous regions are those from which we predict the highest proportion of Lyman continuum (Lyc) photons are escaping, these regions are candidates for the principal ionizing sources of the diffuse insterstellar medium in large spirals, and may contribute significantly to the intergalactic ionizing field. Also the transition between ionization bounding and density bounding appears to take place over a narrow range in $L_{{\\rm H}\\alpha}$, and while the sample of results presented here is not large, this point is worth following up as a high-luminosity feature of this kind would have its uses as a powerful, precise, and non-transient standard candle. However the detailed interpretation of the surface brightness data is not obvious and further, spectroscopic, observations will be important to follow up the arguments presented here. In Sect. 2 we present the evidence about the change in properties of the \\hii\\ regions, in Sect. 3 we present scaling relations which support the model of density bounding for the highest-luminosity regions, and a set of alternative models aimed at explaining the observed ``glitches\" in the luminosity functions of \\hii\\ regions in the galaxies observed. In Sect. 4 we discuss the implications of our models for the diffuse \\ha\\ emission from the warm ionized medium. In Sect. 5 we summarize our conclusions, outline the observations required to deepen our understanding of the transition phenomenon, and suggest applications to the correction of the global star formation rates (SFRs) in galaxies, and to the measurement of intergalactic distances. ", "conclusions": "We have used three types of observations to argue for a change in physical regime between ionization bounding and density bounding in \\hii\\ regions at a characteristic \\ha\\ luminosity close to log $L_{{\\rm H}\\alpha}$ = 38.6 erg s\\me\\, which we have termed the Str\\\"omgren luminosity, $L_{\\rm Str}$. They are the change of slope, accompanied by a ``glitch'' in the \\ha\\ LFs of the complete \\hii\\ region populations in a set of observed disk galaxies, the quantitative relation between internal velocity dispersion, $\\sigma$, and \\ha\\ luminosity, $L_{{\\rm H}\\alpha}$, of the \\hii\\ regions on the virial envelope of the log $L_{{\\rm H}\\alpha}$--log $\\sigma$ distribution for the complete population in a galaxy, and the change in behavior of the \\ha\\ surface brightness of regions in the range $L_{{\\rm H}\\alpha}$ $\\approx$ $L_{\\rm Str}$. Of these, the first two are presented as evidence of the kind of transition proposed, while the third, though of interest, is not as precisely defined by the observations, and its cause is at this stage fully open to other explanations. Indeed in order to sustain our basic hypothesis we will clearly require spectroscopic observations, as specified below. However, the narrow range in the luminosity of the peaks of the LFs presented here means that this feature is of considerable interest on purely empirical grounds, as a potential secondary standard candle. For this reason alone it is worth making a firm effort to understand the underlying physical behavior causing the peak and change of slope. In Sect. 2.1 we showed that two phenomena which could occur, and which at first thought might be able to account for the LF observations (the effect of the overlapping of regions on scales commensurate with the luminosities around $L_{{\\rm H}\\alpha}$ = $L_{\\rm Str}$, and the effect of dust extinction within and enveloping the regions) will not give rise to the observed LF change, and are not in fact viable as explanations for the observations. We ought also to consider an effect brought out by MW, who examine the relationship between the properties of \\hii\\ regions within the Galaxy, and their OB stellar components. They present models to explain the change in slope of the \\ha\\ LF at moderately high luminosities, which they attribute to an essentially statistical effect due to the discrete number of high-luminosity stars, and its relation with the total ionizing luminosity of an \\hii\\ region. In their paper, MW discuss the effects of envelopes of diffuse gas around giant \\hii\\ regions, which are observed for the majority of regions, and which correspond to the ``core--halo'' structure noted in Kennicutt, Edgar, \\& Hodge (1989). In the present study we have taken this structure as correct, and have not discussed it in further detail. The regions defined by our empirical limiting isophotal method contain the whole of the core and the halo, and are clearly bigger than regions defined only via their cores, or via their emission radii at radio wavelengths. It is important to note here that the change in the LF slope predicted by the MW models occurs at an \\ha\\ luminosity an order of magnitude less than that of the glitch we observe. Our observations here effectively disguise the MW break, because it occurs where our sample is incomplete, leading to the broad peak and decline to low luminosities which are statistical artifacts. Neither our models nor our observations contradict the results of MW, but their work does not examine the LF gradient change at $L_{{\\rm H}\\alpha}$ = $L_{\\rm Str}$. The changes which occur in the surface brightness and surface-brightness gradients of the regions in the luminosity range $L\\approx L_{\\rm Str}$ are clearly of interest physically, but it is not so easy for us to show that they are attributable to a transition to density bounding, although we do have significant technical advantages in resolution and signal to noise over the observers in the 1980s (for example the number of regions listed in Kennicutt, Edgar, \\& Hodge 1989 for NGC~7479 was 67, while the in data yielding the internal brightness gradients used in the present paper we catalogued over 1000 regions in this galaxy (Rozas {\\it et al.} 1999). Nor can the changes in parameters presented in Figs.~5a and 5b be readily attributed merely to angular resolution limitations, as suggested recently by Pleuss \\& Heller (2000), based on their study of M101. Their argument is based on a scaling of the results from that galaxy to a distance of 20 Mpc, but the distances of the galaxies in the present sample range from 7 to 38 Mpc ($H_{\\rm 0}$ = 65 km s\\me\\ Mpc\\me), a factor of over 5 in distance. The observed increase in surface brightness and surface-brightness gradients in the range $L_{{\\rm H}\\alpha}$ $>$ $L_{\\rm Str}$ cannot occur for regions with constant density and filling factor, as well pointed out by McCall (1999); density-bounded regions would not show surface-brightness gradients different from ionization-bounded regions if their mean densities and filling factors were invariant. Thus although we have used the approximation of constant density and filling factor in modeling the changes in properties across the transition, it cannot hold for regions at much higher luminosities than $L_{\\rm Str}$. The product of density and filling factor must rise in this range. We can examine a simple case in which two \\hii\\ regions have the same \\ha\\ luminosity but one is ionization-bounded and the other density bounded; the factor relating the total ionizing luminosity of the density-bounded region to the fraction which is down-converted to \\ha\\ we term $g$. In a self-consistent scenario, $g$ can represent the factor by which the product of the density and filling factor of the density-bounded region exceeds that in the ionization-bounded region. It is easy to show that the central surface brightness of the (assumed spherical) density-bounded region will be the greater by a factor $g^{2/3}$, and the radial-brightness gradient, parametrized as the central brightness divided by the radius, will be greater by a factor $g$. This exercise does not prove that density-bounded regions will necessarily have higher surface brightnesses, and brightness gradients, only that the assumption of density bounding is consistent with this condition, and hence consistent with our observations. In order to fulfill this requirement, either the density, or the filling factor, or both, must be rising parameters above $L_{{\\rm H}\\alpha}$ = $L_{\\rm Str}$. While it is perfectly plausible that more intense ionizing sources produce higher filling factors, and that the most massive clouds have higher mean densities, it is not obvious that these changes will not begin to occur below $L_{{\\rm H}\\alpha}$ = $L_{\\rm Str}$, \\ie\\ it is not obviously a physical necessity that the surface-brightness increase occur just at the transition. The fact that it appears to occur at, or near, what we believe to be the transition cannot at this stage be taken as strong support for the hypothesis of density bounding. Direct, spatially resolved measurements of in situ electron densities via line intensity ratios will be required to further our understanding of this point. We must also be prepared to examine our data in the light of non-isotropic models for luminous \\hii\\ regions such as the ``chimney'' hypothesis (Norman \\& Ikeuchi 1989; Heiles 1990) discussed in more physical detail by Dove \\& Shull (1994), in which at a critical luminosity a region can break out of the denser disk gas into the halo. These models appear, however, to predict reduced surface-brightness gradients for regions observed in face-on galaxies, and so are unlikely to offer an explanation of our observations. One of the most attractive aspects of our hypothesis is that it offers a promising scenario to account for the diffuse \\ha\\ from gas dispersed over the disk of a galaxy, outside the \\hii\\ regions. Two quite recent studies have pointed to the emission of ionizing flux from OB stars as a satisfactory explanation of the \\ha\\ emission from the warm ionized medium in disk galaxies. Oey \\& Kennicutt (1997) made a direct comparison of the rate of emission of ionizing photons from the OB stars in the Large Magellanic Cloud, with the rate of \\ha\\ emission from \\hii\\ regions, and concluded that up to 50 \\% of the flux emitted by the stars is not down-converted within the \\hii\\ regions, and is thus available in principle to ionized the warm diffuse medium. They compare this with an estimate of 35 \\% of the radiation emitted by the diffuse medium in \\ha\\, and conclude that the OB stars do put out sufficient flux in this galaxy. A study by Ferguson et al. (1996) of NGC~247 and NGC~7793 concluded that the integrated flux required to ionize the diffuse gas is some 40 \\% of the total \\ha\\ flux emitted by a galaxy, including its \\hii\\ regions. However, this calculation assumed, as have any such previous calculations, that the total flux of ionizing photons from an \\hii\\ region can be directly measured via its \\ha\\ flux. If the hypothesis presented in the present paper is correct, a significant fraction of the total ionizing flux liberated within the \\hii\\ region population of a galaxy is not converted to Balmer radiation within the \\hii\\ regions. This is true of the most luminous regions, with luminosities ranging up to 10$^{40}$ erg s\\me, and the escaping flux can range up to rather more than 10$^{40}$ erg s\\me\\ for the brightest regions. The integrated escaping flux for a galaxy can attain a few times 10$^{41}$ erg s\\me, easily enough to account for the measured diffuse \\ha\\ flux, according to the observations of Ferguson et al. (1996) and of our own (Zurita, Rozas, \\& Beckman 2000). Ferguson et al. (1996) pointed out that the emitting diffuse gas tends to surround \\hii\\ regions, and we can go further on this point, noting the geometrical correlation with the most luminous regions, which we will go further to quantify in a new study (Zurita, Rozas, \\& Beckman 2000). Finally if our hypothesis is valid, a significant fraction of all the ionizing photons emitted by the OB stars in a galaxy may escape completely from the disk into the halo and finally into the intergalactic medium. Two corollaries are of interest here: intergalactic clouds, even relatively far from a major galaxy, may have their surface layers ionized by this escaping flux, and global star formation rates in disk galaxies, estimated from integrated \\ha\\ fluxes, may in fact be significantly larger than these estimates. Both of these points are worth pursuing theoretically and observationally. We should treat the density bounding-hypothesis for the luminous \\hii\\ regions as speculative. Fundamentally missing pieces of the puzzle are line-ratio tests, of the sort performed by McCall, Rybski, \\& Shields (1985) comparing [\\oiii] to [\\oii] emission-line intensities (in a density-bounded region, the [\\oii] Str\\\"omgren sphere should under many conditions be larger in radius than the \\hii\\ region, so the [\\oiii]/[\\oii] ratio should be significantly enhanced over its value in ionization-bounded regions. McCall, Rybski, \\& Shields (1985) claimed one such detection, region(-606-1708) in NGC~598, but no others, and this detection has since been called into question (McCall 1999). Technical advances of the type represented by the TTF (Taurus Tuneable Filter, Bland-Hawthorn \\& Jones 1998) which permits imaging of complete galaxies in single emission lines with full redshift flexibility, will allow us to apply line-ratio tests over complete populations of fully imaged \\hii\\ regions in galaxies such as the ones measured only in \\ha\\ for the present paper. Another prospect, but only for the nearest galaxies due to limitations on angular resolution, is the use of two-dimensional fiber-fed spectrographs to sample the emission across the full face of an \\hii\\ region. Here we would trade complete spectral coverage for limitation on the number of \\hii\\ regions. In both cases we would hope to enhance understanding of the internal physics of the gas within the \\hii\\ regions, measuring as a function of position on a region its temperature and electron density, and applying tests such as the [\\oiii]/[\\oii] emission ratio to the question of density bounding. The apparent invariance in the luminosity of the ``glitch'' measured in the LF does offer a possible refined secondary standard candle of high luminosity and constant presence for use well into the Hubble flow, independently of whether the interpretation we have placed on it in the present article is fully, or even partially, valid. To demonstrate its scope will require amplification of the number of objects studied to determine reliably, using local galaxies, the scatter in the observed feature, including any possible dependence on galaxy luminosity and type, and to calibrate it using galaxies whose Cepheid distances have been determined in the {\\it HST\\/} key project." }, "0003/astro-ph0003390_arXiv.txt": { "abstract": "Star clusters -- open and globulars -- experience dynamical evolution on time scales shorter than their age. Consequently, open and globular clusters provide us with unique dynamical laboratories for learning about two-body relaxation, mass segregation from equipartition of energy, and core collapse. We review briefly, in the framework of star clusters, some elements related to the theoretical expectation of mass segregation, the results from N-body and other computer simulations, as well as the now substantial clear observational evidence. ", "introduction": "The dynamics of any stellar system may be characterized by the following three dynamical time scales: (i) the crossing time $t_{cr}$, which is the time needed by a star to move across the system; (ii) the two-body relaxation time $t_{rlx}$, which is the time needed by the stellar encounters to redistribute energies, setting up a near-maxwellian velocity distribution; (iii) the evolution time $t_{ev}$, which is the time during which energy-changing mechanisms operate, stars escape, while the size and profile of the system change. Several (different and precise) definitions exist for the relaxation time. The most commonly used is the half-mass relaxation time $t_{rh}$ of Spitzer (1987, Eq.~2-62), where the values for the mass-weighted mean square velocity of the stars and the mass density are those evaluated at the half-mass radius of the system (see Meylan \\& Heggie 1997 for a review). In the case of globular clusters, $t_{cr}$ $\\sim$ \\mily, $t_{rlx}$ $\\sim$ 100~\\mily, and $t_{ev}$ $\\sim$ 10~\\bily. Table~1 displays, for open clusters, globular clusters, and galaxies, some interesting relations between the above three time scales. For open clusters, crossing time $t_{cr}$ and relaxation time $t_{rlx}$ are more or less equivalent, both being significantly smaller than the evolution time $t_{ev}$. This means that most open clusters dissolve within a few gigayears. For galaxies, the relaxation time $t_{rlx}$ and the evolution time $t_{ev}$ are more or less equivalent, both being significantly larger than the crossing time $t_{cr}$. This means that galaxies are not relaxed, i.e., not dynamically evolved. It is only for globular clusters that all three time scales are significantly different, implying plenty of time for a clear dynamical evolution in these stellar systems, although avoiding quick evaporation altering open clusters. Consequently, star clusters -- open and globular -- represent interesting classes of dynamical stellar systems in which some dynamical processes take place on time scales shorter than their age, i.e., shorter than the Hubble time, providing us with unique dynamical laboratories for learning about two-body relaxation, mass segregation from equipartition of energy, stellar collisions, stellar mergers, and core collapse. All these dynamical phenomena are related to the internal dynamical evolution only, and would also happen in isolated glo\\-bular clusters. The external dynamical disturbances --- tidal stripping by the galactic gravitational field --- influence equally strongly the dynamical evolution of globular clusters. \\begin{table}[t] \\caption{ Dynamical time scales for open clusters, globular clusters and galaxies} \\begin{center} \\footnotesize \\begin{tabular}{|l|c|c|c|l|} \\hline & & \\\\ open clusters &t$_{cr}$ $\\sim$ t$_{rlx}$ $\\ll$ t$_{ev}$ & quickly dissolved\\\\ & & \\\\ globular clusters & t$_{cr}$ $\\ll$ t$_{rlx}$ $\\ll$ t$_{ev}$ & \\\\ & & \\\\ galaxies & t$_{cr}$ $\\ll$ t$_{rlx}$ $\\sim$ t$_{ev}$ ¬ relaxed\\\\ & & \\\\ \\hline \\end{tabular} \\end{center} \\end{table} ", "conclusions": "" }, "0003/astro-ph0003445_arXiv.txt": { "abstract": "Neutrino and axion production in neutron stars occurs mainly as bremsstrahlung from nucleon-nucleon ($NN$) scattering. The energy radiated via neutrinos or axions is typically very small compared to other scales in the two-nucleon system. The rate of emission of such ``soft\" radiation is directly related to the on-shell $NN$ amplitude, and thereby to the $NN$ experimental data. This facilitates the model-independent calculation of the neutrino and axion radiation rates which is presented here. We find that the resultant rates are roughly a factor of four below earlier estimates based on a one-pion-exchange $NN$ amplitude. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003244_arXiv.txt": { "abstract": "While model-independent limits are always interesting, a limit on neutron star radius as a function of mass attains special interest in light of recent interpretations of the periodic as well as quasi-periodic oscillations (QPOs) in brightness of X-rays emitted from neutron stars that are accreting matter from a low-mass companion. Here we derive such a limit based only on well accepted principles. We discuss our limit in connection with a recent interpretation of X-ray pulsations from SAX J1808.4-3658 as indicating a strange-star candidate, and show that this object can also be a normal neutron star, though one whose central core has very high density. The most plausible high-density phase of hadronic matter, which is also expected to be very compressible, is quark matter. So an alternative to the strange star interpretation of SAX J1808.4-3658 is that it is a hybrid neutron star. ", "introduction": "An accreting X-ray binary having millisecond pulses was recently discovered (\\cite{klis98:b,chakrabarty98:a}). Presumably it is one of a class of objects that represent the missing link between canonical and millisecond pulsars. If so, it confirms a long-standing conjecture that millisecond pulsars are formed by accretion onto canonical pulsars (\\cite{alpar82:a,bachus82:a,radhakrishnan82:a}). Many other accreting X-ray sources have been discovered consisting of a neutron star and a low-mass companion from which material is gathered, presumably, from an accretion disk (\\cite{klis96:a,strohmayer96:a,klis98:a,klis99:a}). An interpretation of quasi-periodic oscillations in X-ray brightness characteristic of these sources suggests that upper limits on mass and radius of the neutron star may possibly be deduced (\\cite{strohmayer96:a,lamb98:c}). However, the interpretation in such terms is the subject of some controversy and it remains to be seen how the models of the observations finally play out (\\cite{wasserman99:a}). Certainly there is much uncertainty concerning the magnetic field and accretion disk interaction which play an important role in the modeling of X-ray pulsations. Tentative limits on mass and radius deduced from models of quasi periodic oscillations in X-ray brightness (QPOs) have been employed recently to discriminate among models of the equation of state of dense nuclear matter (\\cite{schaab99:a}). The X-ray pulsar, Sax J1808.4-3658, is a particularly interesting object; it produces coherent X-ray emission with a 2.5 ms period as well as X-ray bursts. Based on an analysis of radiation from this object, a limiting mass-radius relationship was derived which is difficult to reconcile with existing neutron star models (\\cite{heuvel99:a}). The mass-radius relationship derived would be consistent with an interpretation of Sax J1808.4 as a strange star candidate as found by the above authors. Against this background our purpose is to derive a model-independent mass-radius constraint for neutron stars that depends only on minimal and well accepted principles. The limiting relation is analogous to a previously obtained lower limit on the Kepler period of a rotating star as a function of its mass (\\cite{glen92:a,friedman96:a}), and to an even earlier analysis of limits on the gravitational redshift from neutron stars (\\cite{lindblom84:a}). The most conservative minimal principles and constraints are:\\\\ 1. Einstein's general relativistic equations for stellar structure hold.\\\\ 2. The matter of the star satisfies $dp/d\\rho \\geq 0$ which is a necessary condition that a body is stable, both as a whole and also with respect to the spontaneous expansion or contraction of elementary regions away from equilibrium (Le Chatelier's principle).\\\\ 3. The \\eos satisfies the causal constraint for a perfect fluid; a sound signal cannot propagate faster than the speed of light, $v(\\epsilon) \\equiv\\sqrt{dp/d\\epsilon} \\leq 1 $, which is also the appropriate expression for sound signals in General Relativity (\\cite{curtis50:a,geroch91:a}).\\\\ 4. The high-density equation of state matches continuously in energy and pressure to the low-density equation of state of \\cite{baym71:b} and has no bound state at any density. The last condition assures that the $M-R$ relation obtained is for a neutron star and not some sort of exotic. We mean ``neutron star'' in the generic sense: it is made of charge neutral nuclear matter at low density, while at higher density in the interior, matter may be in a mixed or pure quark-matter or other high-density phase of nuclear matter. The last condition also implies that the star is bound by gravity as is a neutron star and is not a self-bound star such as a strange star. As we will see, a self-bound star can lie in a region of the $M-R$ plane that is forbidden to neutron stars. In referring to the constraints as conservative, we mean that we make no assumption about dense matter aside from the constraints mentioned and we specifically allow for a phase transition above a baryon density of $0.1625 {\\rm~fm^{-3}}$. We discuss this further in the Section, Caveats. We can adapt the results of our earlier search for a model-independent minimum Kepler period by searching for the radius at fixed mass that minimizes $P\\sim (R^3/M)^{1/2}$ (\\cite{glen92:a}). Several researchers found that the above classical result applies to relativistic stars to within a few percent accuracy with a suitable constant of proportionality (\\cite{haensel89:a,friedman89:a}). We use variational \\eoss subject to the above constraints and techniques as described in the above reference. Our earlier results for the Kepler period agree to six percent with the results of \\cite{friedman96:a}, who performed a numerical solution for rotating stars in place of the above approximation formula for the Kepler period in terms of mass and radius of the non-rotating counterparts. Our results are shown in Fig.\\ \\ref{rlimit}. Neutron stars at the mass limit can have radii as small as those shown by the line, and otherwise must lie in the shaded region marked for neutron stars. The region can be approximated in the interval illustrated by \\beqn R\\geq \\Bigl( 3.1125 - 0.44192 x + 2.3089 x^2 - 0.38698 x^3 \\Bigr) {\\rm~km},~~ \\nonumber \\\\(1\\leq x\\equiv \\frac{M}{\\msun} \\leq 2.5)\\, .~~~~~~~ \\eeqn Of course, for neutron stars (unlike white dwarfs), there is only one \\eos in nature; all neutron stars form a single family, and whatever the trajectory of the mass-radius relationship is for that family, the limiting mass star has the smallest radius, and it is greater or equal to the limit derived. For example, if the most massive neutron star that could exist in nature, independent of formation mechanism, is $2 \\msun$, the radius of all neutron stars would have to exceed 8.37 km. (Recall that measured masses tell us nothing about the maximum possible mass that can be supported by nature's \\eosp.) \\begin{figure}[htb] \\begin{center} \\leavevmode \\hspace{-.2in} \\psfig{figure=rlimit.eps,width=3in} \\parbox[t]{4.4 in} { \\caption{ \\label{rlimit} Allowed region of non-rotating neutron star radii as a function of limiting mass determined independent of specific models of dense matter, but rather by the minimal constraints (causality, etc) enumerated in the text. The variational results are indicated by the dots. Stars that are not at the mass limit will have larger radii. }} \\end{center} \\end{figure} Another limit of interest follows from the properties of General Relativity. Schartzs\\-child's limit $R>2M$ is actually less stringent than $R>9M/4$, which must be obeyed by any relativistic star (\\cite{buchdahl59:a,weinberg}). The latter is also plotted. If a star's mass and radius placed it in the region between the above described regions, it could be made of matter that is self-bound at high density, matter that would be bound in microscopic to stellar like-objects even in the absence of gravity (see Eq.\\ \\ref{den}). Strange stars, if the strange matter hypothesis is true, are examples. ", "conclusions": "" }, "0003/astro-ph0003134_arXiv.txt": { "abstract": "Relatively uniform light curves and spectral evolution of Type Ia supernovae (SNe Ia) have led to the use of SNe Ia as a ``standard candle'' to determine cosmological parameters. Whether a statistically significant value of the cosmological constant can be obtained depends on whether the peak luminosities of SNe Ia are sufficiently free from the effects of cosmic and galactic evolutions. Here we first review the single degenerate scenario for the Chandrasekhar mass white dwarf (WD) models of SNe Ia. We identify the progenitor's evolution and population with two channels: (1) the WD+RG (red-giant) and (2) the WD+MS (near main-sequence He-rich star) channels. In these channels, the strong wind from accreting WDs plays a key role, which yields important age and metallicity effects on the evolution. We then address the questions whether the nature of SNe Ia depends systematically on environmental properties such as metallicity and age of the progenitor system and whether significant evolutionary effects exist. We suggest that the variation of the carbon mass fraction $X$(C) in the C+O WD (or the variation of the initial WD mass) causes the diversity of the brightness of SNe Ia. This model can explain the observed dependences of SNe Ia brightness on the galaxy types and the distance from the galactic center. Finally, applying the metallicity effect on the evolution of SN Ia progenitors, we make a prediction of the cosmic supernova rate history as a composite of the supernova rates in different types of galaxies. ", "introduction": "\\vspace*{-1mm} Type Ia supernovae (SNe Ia) are good distance indicators, and provide a promising tool for determining cosmological parameters (e.g., \\cite{bra98}). SNe Ia have been discovered up to $z \\sim 1.32$ \\cite{gil99}. Both the Supernova Cosmology Project \\cite{per97,per99} and the High-z Supernova Search Team \\cite{gar98,rie98} have suggested a statistically significant value for the cosmological constant. However, SNe Ia are not perfect standard candles, but show some intrinsic variations in brightness. When determining the absolute peak luminosity of high-redshift SNe Ia, therefore, these analyses have taken advantage of the empirical relation existing between the peak brightness and the light curve shape (LCS). Since this relation has been obtained from nearby SNe Ia only \\cite{phi93,ham95,rie95}, it is important to examine whether it depends systematically on environmental properties such as metallicity and age of the progenitor system. High-redshift supernovae present us very useful information, not only to determine cosmological parameters but also to put constraints on the star formation history in the universe. They have given the SN Ia rate at $z \\sim 0.5$ \\cite{pai99} but will provide the SN Ia rate history over $0$1.5 Mpc from the centre of the cluster Cl0024+16 (Bonnet et al. 1994), and in a supercluster (Kaiser et al. 1998). Weak lensing by large-scale structure also produces small coherent distortions in the images of distant field galaxies (see Mellier 1999; Kaiser 1999; Bartelmann \\& Schneider 1999 for recent reviews). A measurement of this effect on various scales (defined as `cosmic shear') would provide invaluable cosmological information (Kaiser 1992; Jain \\& Seljak 1997; Kamionkowski et al. 1997; Kaiser 1998; Hu \\& Tegmark 1998; Van Waerbeke et al. 1998). In particular, it would yield a direct measure of the power spectrum of density fluctuations along the line of sight and thus provide an independent constraint on large scale structure models and cosmological parameters. Because of its small amplitude (a few percent on arcmin scales for favoured CDM models), cosmic shear has however been difficult to detect. In a pioneering paper, Mould et al. (1994) attempted to detect the coherent distortion of $R\\sim$26 field galaxies over a 9.6 arcmin diameter field and found an upper limit quoted in terms of the {\\it rms} shear at the 4\\% level. A search for this effect is the object of active observational effort (Van Waerbeke et al. 1998; Refregier et al. 1998; Seitz et al. 1998; Rhodes et al. 1999; Kaiser 1999). At present however, no unambiguous detections of cosmic shear have been reported (see however the limited results of Villumsen 1995; Schneider et al. 1998). A fundamental limitation of narrow field imaging as a probe of cosmic shear is that arising from cosmic variance, i.e. the fluctuation in the lensing signal measured with a limited number of pencil beam sight lines. Only through the analysis of image fields in many statistically-independent directions can this variance be overcome. Prior to such a measurement, it is important to demonstrate a reliable detection strategy, particularly in the presence of significant instrumental and other systematic effects. In this paper, we report the detection of a cosmic shear signal with 14 separated $16' \\times 8'$ fields observed with the 4.2m William Herschel Telescope (WHT). We provide a detailed treatment of systematic effects and of the shear measurement method. We test our results with numerical simulations of lensed images and quantify both our statistical and systematic errors. We discuss the consequence of our measurement for the normalisation of the mass power spectrum. Subsequent papers will extend this technique to a larger number of fields, reducing the limitations caused by cosmic variance. This paper is organised as follows. In \\S\\ref{theory}, we introduce the theory of weak lensing in the context of a cosmic shear survey. In \\S\\ref{survey} we discuss our observational strategy for detecting it and describe our observations taken at the WHT and the routine aspects of data reduction. In \\S\\ref{object}, we describe the generation of the object catalogue and how the image parameters were measured. In \\S\\ref{distortion} we discuss and characterise distortions introduced by the telescope optics. In \\S\\ref{shear_method} we discuss the point spread function and present our shear measurement method, alongside an important comparison with the same analysis conducted with simulated data (\\S\\ref{simulation}). In \\S\\ref{estimator}, we describe the estimator used for measuring the shear variance and the cross-correlation between adjacent cells. In \\S\\ref{results}, we present our results. Our conclusions are summarised in \\S\\ref{conclusion}. \\section[]{Theory} \\label{theory} \\subsection{Distortion Matrix} Gravitational lensing by large scale structure produces distortions in the image of background galaxies (see Mellier 1999; Kaiser 1999; Bartelmann \\& Schneider 1999 for recent reviews). These distortions are weak (about 1\\%) and can be fully characterised by the distortion matrix \\begin{equation} \\label{eq:psi_def_theory} \\Psi_{ij} \\equiv \\frac{\\partial (\\delta \\theta_{i})}{\\partial \\theta_{j}} \\equiv \\left( \\begin{array}{cc} \\kappa +\\gamma_{1} & \\gamma_{2}\\\\ \\gamma_{2} & \\kappa - \\gamma_{1} \\\\ \\end{array} \\right), \\end{equation} where $\\delta \\theta_{i}({\\mathbf \\theta})$ is the displacement vector produced by lensing on the sky. The convergence $\\kappa$ describes overall dilations and contractions. The shear $\\gamma_{1}$ ($\\gamma_{2}$) describes stretches and compressions along (at $45^{\\circ}$ from) the x-axis. The distortion matrix is directly related to the matter density fluctuations along the line of sight by \\begin{equation} \\label{eq:psi_dchi} \\Psi_{ij} = \\int_{0}^{\\chi_{h}} d\\chi ~g(\\chi) \\partial_{i} \\partial_{j} \\Phi \\end{equation} where $\\Phi$ is the Newtonian potential, $\\chi$ is the comoving distance, $\\chi_{h}$ is the comoving distance to the horizon, and $\\partial_{i}$ is the comoving derivative perpendicular to the line of sight. The radial weight function $g(\\chi)$ is given by \\begin{equation} g(\\chi) = 2 \\int_{\\chi}^{\\chi_{h}} d\\chi'~n(\\chi') \\frac{r(\\chi)r(\\chi'-\\chi)}{r(\\chi')}, \\end{equation} where $r$ is the comoving angular diameter distance, and $n(\\chi)$ is the probability of finding a galaxy at comoving distance $\\chi$ and is normalised as $\\int d\\chi n(\\chi) =1$. If the galaxies all lie at a single distance $\\chi_{s}$, $n(\\chi)=\\delta(\\chi-\\chi_{s})$ and \\begin{equation} g(\\chi) = 2 \\frac{r(\\chi) r(\\chi_{s}-\\chi)}{r(\\chi_{s})} \\end{equation} In practice, this approximation is accurate to within 10\\%, if $\\chi_{s}$ is set to the median distance of the galaxy sample. This is adequate given the median redshift of our galaxy sample is itself uncertain by about 25\\% (see \\S\\ref{zref}), yielding an uncertainty in the predicted rms shear of about 20\\% (see Eq.~[\\ref{eq:sigg_sig8}] below). \\subsection{Power Spectrum} The amplitude of the cosmic shear can be quantified statistically by computing its 2-dimensional power spectrum (Jain \\& Seljak 1997; Kamionkowski et al. 1997; Schneider et al. 1997; Kaiser 1998). For this purpose, we consider the Fourier transform of the shear field \\begin{equation} \\widetilde{\\gamma_{i}}({\\mathbf l}) = \\int d^{2}\\theta ~\\gamma_{i}({\\mathbf \\theta}) e^{i {\\mathbf l \\cdot \\theta}} \\end{equation} The shear power spectrum $C_{{\\mathbf l}}^{ij}$ is defined by \\begin{equation} \\label{eq:clij_def} \\langle \\widetilde{\\gamma_{i}}({\\mathbf l}) \\widetilde{\\gamma_{j}}({\\mathbf l'}) \\rangle = (2\\pi)^{2} \\delta^{(2)}({\\mathbf l} - {\\mathbf l'}) C_{{\\mathbf l}}^{ij} \\end{equation} where $\\delta^{(2)}$ is the 2-dimensional Dirac-delta function, and the brackets denote an ensemble average. It is also useful to define the scalar power spectrum $C_{l} = C_{{\\mathbf l}}^{11} + C_{{\\mathbf l}}^{22}$ for the shear amplitude by \\begin{equation} \\langle \\widetilde{\\gamma_{i}}({\\mathbf l}) \\widetilde{\\gamma_{i}}({\\mathbf l'}) \\rangle = (2\\pi)^{2} \\delta^{(2)}({\\mathbf l} - {\\mathbf l'}) C_{{\\mathbf l}}, \\end{equation} where the summation convention was used. Applying Limber's equation in Fourier space (Kaiser 1998) to Equation~(\\ref{eq:psi_dchi}) and using the Poisson equation, we can express the shear power spectrum $C_{l}$ in terms of the 3-dimensional power spectrum $P(k,\\chi)$ of the mass fluctuations $\\delta \\rho/\\rho$ and obtain \\begin{equation} C_{l} = \\frac{9}{16} \\left( \\frac{H_{0}}{c} \\right)^{4} \\Omega_{m}^{2} \\int_{0}^{\\chi_h} d\\chi~\\left[ \\frac{g(\\chi)}{a r(\\chi)} \\right]^{2} P\\left(\\frac{l}{r}, \\chi\\right), \\end{equation} where $a$ is the expansion parameter, and $H_{0}$ and $\\Omega_{m}$ are the present value of the Hubble constant and matter density parameter, respectively. After noting that $C_{l}$ is also equal to the power spectrum of the convergence $\\kappa$, we find that this expression agrees with that of Schneider et al. (1997). The component-wise power spectrum is given by \\begin{equation} C_{{\\mathbf l}}^{ij} = u_{i}(\\lambda) u_{j}(\\lambda) C_{l} \\end{equation} where $u_{i}(\\lambda)=\\{\\cos(2\\lambda),\\sin(2\\lambda)\\}$ and $\\lambda$ is the angle of the vector ${\\mathbf l}$, counter-clockwise from the $l_{1}$-axis. A measurement of the power spectrum enables differentiation between the different cosmological models listed in Table~\\ref{tab:models}. Standard Cold Dark Matter (SCDM) is approximately COBE-normalised (Bunn \\& White 1997), while the other variants are approximately cluster normalised ($\\sigma_{8} \\Omega_{m}^{.53} =0.6\\pm0.1$; Viana \\& Liddle 1996). For each model we compute the non-linear power spectrum using the fitting formula of Peacock \\& Dodds (1996). The resulting power spectra are shown in Figure~\\ref{fig:cl} for sources observed at $z_{s}=1$. \\begin{table*} \\centering \\begin{minipage}{140mm} \\caption{Cell-averaged statistics for each cosmological model (with $z_{s}=1$)} \\label{tab:models} \\begin{tabular}{crrrrrrrr} \\hline Model & $\\Omega_{m}$ & $\\Omega_{\\Lambda}$ & $\\sigma_{8}$ & $\\Gamma$ & $\\sigma_{\\gamma}$ (\\%) & $\\sigma_{\\times}$ (\\%) & $\\sigma_{\\times 1}$ (\\%) & $\\sigma_{\\times 2}$ (\\%) \\\\ \\hline SCDM & 1.0 & 0 & 1 & 0.50 & 2.60 & 1.62 & 1.23 & 1.05 \\\\ $\\tau$CDM & 1.0 & 0 & 0.6 & 0.25 & 1.25 & 0.86 & 0.64 & 0.58 \\\\ $\\Lambda$CDM & 0.3 & 0.7 & 1 & 0.25 & 1.15 & 0.71 & 0.54 & 0.46\\\\ OCDM & 0.3 & 0 & 1 & 0.25 & 1.04 & 0.62 & 0.48 & 0.39\\\\ \\hline \\end{tabular} \\end{minipage} \\end{table*} \\begin{figure} \\psfig{figure=figure1.ps,width=80mm} \\caption{Shear power spectrum for each cosmological model and for sources at $z_{s}=1$. Note that the SCDM spectrum is larger due to its higher normalisation.} \\label{fig:cl} \\end{figure} \\subsection{Cell-Averaged Statistics} For our measurement, we will consider statistics of the shear averaged over angular cells on the sky. This has the advantage of diminishing the impact of systematic effects (Rhodes et al. 1999) and allows extension in later surveys to minimise cosmic variance. The average shear $\\overline{\\gamma}_{i}$ in a cell can be written as \\begin{equation} \\overline{\\gamma}_{i} = \\int d^{2}\\theta~ W({\\mathbf \\theta}) \\gamma_{i}({\\mathbf \\theta}) \\end{equation} where $W({\\mathbf \\theta})$ is the cell window function and is normalised as $\\int d^{2}\\theta~ W({\\mathbf \\theta}) = 1$. It is convenient to define the Fourier transform of the window function as \\begin{equation} \\widetilde{W}_{{\\mathbf l}} = \\int d^{2}\\theta~ W({\\mathbf \\theta}) e^{i {\\mathbf l \\cdot \\theta}}. \\end{equation} For a square cell of side $\\alpha$, this is \\begin{equation} \\widetilde{W}_{{\\mathbf l}} = \\left( \\frac{\\sin(\\alpha l_{1})}{\\alpha l_{1}} \\right) \\left( \\frac{\\sin(\\alpha l_{2})}{\\alpha l_{2}} \\right), \\end{equation} To a good approximation, we can ignore the small azimuthal dependence of the window function and approximate \\begin{equation} \\widetilde{W}_{l} \\simeq \\left( \\frac{\\sin( \\alpha l / \\sqrt{2})} { \\alpha l/ \\sqrt{2}} \\right)^{2}. \\end{equation} Let us consider 2 cells separated by an angle ${\\mathbf \\theta}$. We are interested in the correlation function \\begin{equation} w_{ij}({\\mathbf \\theta}) \\equiv \\langle \\overline{\\gamma}_{i}(0) \\overline{\\gamma}_{j}({\\mathbf \\theta}) \\rangle \\end{equation} As is the case in our experiment, we take the separation vector ${\\mathbf \\theta}$ to lie along the $\\theta_{1}$-axis (or equivalently along the $\\theta_{2}$-axis). By taking Fourier transforms and using Equation~(\\ref{eq:clij_def}), we thus obtain \\begin{eqnarray} w_{ij}({\\mathbf \\theta}) & \\simeq & \\frac{1}{4\\pi} \\int_{0}^{\\infty} dl~ l C_{l} \\left| \\widetilde{W}_{l} \\right|^{2} \\times \\nonumber \\\\ & & \\left( \\begin{array}{cc} J_{0}(l \\theta)+J_{4}(l \\theta) & 0 \\\\ 0 & J_{0}(l \\theta)-J_{4}(l \\theta) \\\\ \\end{array} \\right). \\end{eqnarray} As noted above, we have ignored the azimuthal dependence of the window function $\\widetilde{W}_{l}$. In particular, the shear variance $\\sigma_{\\gamma}^{2} \\equiv \\langle \\overline{\\gamma}^{2} \\rangle = w_{11}(0)+w_{22}(0)$ is given by \\begin{equation} \\sigma_{\\gamma}^{2} = \\frac{1}{2\\pi} \\int_{0}^{\\infty} dl~ l C_{l} \\left| \\widetilde{W}_{l} \\right|^{2}. \\end{equation} We will denote the component-wise covariances between two adjacent cells by \\begin{equation} \\label{eq:sigma_cross} \\sigma_{\\times 1}^{2} \\equiv w_{11}(\\alpha), ~~~\\sigma_{\\times 1}^{2} \\equiv w_{22}(\\alpha), \\end{equation} and their modulus by $\\sigma_{\\times}^{2} \\equiv \\sigma_{\\times 1}^{2}+ \\sigma_{\\times 2}^{2}$. The values of these statistics for each model are listed in Table~\\ref{tab:models} for our cell size of $\\alpha=8'$. The rms shear is of the order of 1\\% for the cluster-normalised models and of about 2\\% for the COBE-normalised model. The cross-correlation rms is about half the zero-lag value (c.f. Schneider et al. 1997). Figure~\\ref{fig:sigmag} shows the dependence of $\\sigma_{\\gamma}$ on the source redshift $z_{s}$ and $\\sigma_{8}$ for the $\\Lambda$CDM model (again for $\\alpha = 8'$). The range chosen approximately reflects the likely uncertainty in these parameters for our experiment. Importantly, the rms shear is more sensitive to $\\sigma_{8}$. A 10\\% uncertainty in the source redshift results in a 8\\% uncertainty in $\\sigma_{\\gamma}$. For this model, the dependence of $\\sigma_{\\gamma}$ is very well approximated by \\begin{equation} \\label{eq:sigg_sig8} \\sigma_{\\gamma} \\simeq 0.0115 z_{s}^{0.81} \\sigma_{8}^{1.25}, \\end{equation} in agreement with the scaling laws of Jain \\& Seljak (1997). \\begin{figure} \\psfig{figure=figure2.ps,width=80mm} \\caption{Dependence of the rms shear on the source redshift $z_{s}$ and the power spectrum normalisation $\\sigma_{8}$. The cell was chosen to be a square of side $\\alpha=8'$.} \\label{fig:sigmag} \\end{figure} \\section[]{Data} \\label{survey} \\subsection{Survey Strategy} In order to detect and ultimately measure the cosmic shear, an array of deep imaging fields is required. These must be randomly placed on the sky to provide a fair sample, and should be well separated in order to be statistically independent, from the point of view of cosmic variance. As mentioned in \\S\\ref{intro}, it is expedient to distinguish between a {\\em detection} based on a careful analysis of a few fields, noting carefully the systematic effects, before embarking upon an exhaustive {\\em measurement} survey utilising a larger number of fields to beat down the uncertainties arising from cosmic variance. With these factors in mind, we now discuss our strategy and observations using the William Herschel Telescope (WHT). A bank of appropriate fields were selected for observation with the WHT prime focus CCD Camera (field of view 8' $\\times$ 16', pixel size 0.237'', EEV CCD) in the $R$ band. This photometric band offers the deepest imaging for a given exposure time with minimal fringing. Fields were selected using the Digital Sky Survey by choosing coordinates randomly within the range appropriate for the time of observations. Each field was retrospectively checked to see whether it contained large galaxies ($\\gg 5$ arcsec) (which would occult a significant fraction of the imaging field) or prominent groups/clusters (located using the NASA/IPAC Extragalactic Database) on a scale comparable to that under study ($\\simeq$8 arcmin). There is, of course, a danger of over-compensating by exclusion in this respect but, fortunately, none of the originally-chosen fields were discarded according to the above criteria. The fields were further required to be $>5^\\circ$ away from one other, in order to ensure statistical independence (c.f. figure \\ref{fig:cl}, where the power is small for $l < 10$). Using the APM and GCC catalogues, we ensured that the fields contained no stars with $R<11$ (in order to avoid large areas of saturation and ghost images). On the other hand, we required the fields to contain $\\simeq 200$ stars with $R<22$ in order to map carefully the anisotropic PSF and the camera distortion across the field of view. In order to achieve this, the fields were chosen to be at intermediate Galactic latitudes ($30^\\circ < b < 70^\\circ$; see Table \\ref{tab:coords}). A calibration of the stellar density at limits fainter than the APM and GCC catalogues was obtained from a test WHT image (see below). The final constraint on field position was our desire to observe each field within $20^\\circ$ of the telescope's zenith during the observing run; this reduces image distortion introduced by telescope and instrument flexure. This criterion was relaxed for the fields VLT1, CIRSI1 and CIRSI2 (see nomenclature below). Table \\ref{tab:coords} summarises the positions and Galactic latitude of the fields which are used in this paper. Two fields are in common with the VLT (Mellier et al., in preparation) and HST STIS (Seitz et al. 1998) cosmic shear programmes, allowing future comparisons with these programmes. A further two fields spanned the Groth Strip (Groth et al. 1998; Rhodes 1999) a deep survey conducted with HST, which has previously been studied for cosmic shear detection (Rhodes et al. 1999). Finally, two fields were chosen to be in common with the current CIRSI photometric redshift survey (Firth et al., in preparation) to give us clearer understanding of the redshift distribution of objects in our fields at a later date. \\begin{table*} \\centering \\begin{minipage}{170mm} \\caption{Field Coordinates (equinox 2000) and Properties} \\label{tab:coords} \\begin{tabular}{crrrrrrr} \\hline Field name & RA (h:m:s) & Dec (d:m:s) & Galactic & Seeing & Magnitude & Median & No. Survey Galaxies \\\\ & & & latitude & (arcsec) & limit & magnitude & $(8' \\times 16' field$)\\\\ & & & (deg) & & (imcat $5\\sigma)$ & of survey \\\\ & & & & & & galaxies\\\\ \\hline WHT0 & 02:03:09.31 & 11:30:20.0 & -47.6 & 0.59 & 26.2 & 23.1 & 1550\\\\ WHT3 & 14:00:15.00 & 10:13:40.0 & 66.6 & 0.82 & 26.2 & 23.3 & 2141\\\\ WHT5 & 14:50:46.67 & 20:18:03.2 & 61.9 & 0.76 & 26.5 & 23.7 & 2181\\\\ WHT7 & 15:13:40.86 & 36:31:30.8 & 58.6 & 0.83 & 25.9 & 23.0 & 1354\\\\ WHT11 & 16:31:44.28 & 27:56:30.0 & 41.6 & 0.85 & 26.0 & 23.3 & 1379\\\\ WHT12 & 16:37:20.00 & 20:46:30.0 & 38.4 & 0.90 & 26.0 & 23.3 & 1855\\\\ WHT14 & 16:51:15.38 & 25:46:44.0 & 36.8 & 0.99 & 25.9 & 23.2 & 1701\\\\ WHT16 & 17:13:40.00 & 38:39:19.0 & 34.9 & 0.78 & 25.8 & 23.4 & 2074\\\\ WHT17 & 14:24:38.10 & 22:54:01.0 & 68.5 & 0.63 & 27.3 & 24.5 & 2287\\\\ VLT1 & 12:28:18.50 & 02:10:05.0 & 64.4 & 0.71 & 26.4 & 23.6 & 1721\\\\ VLT2 & 15:28:43.00 & 10:14:20.0 & 49.3 & 0.79 & 26.1 & 23.4 & 2093\\\\ CIRSI1 & 12:05:35.01 & -07:43:00.0 & 60.1 & 1.14 & 25.4 & 22.6 & 1192\\\\ CIRSI2 & 15:23:37.00 & 00:15:00.0 & 60.4 & 0.76 & 26.3 & 23.5 & 1824\\\\ GROTH1 & 14:17:18.74 & 52:20:18.5 & 53.4 & 0.78 & 26.1 & 23.4 & 2237\\\\ GROTH2 & 14:15:35.00 & 52:08:48.0 & 44.7 & 0.89 & 26.1 & 23.6 & 1195\\\\ \\hline \\end{tabular} \\end{minipage} \\end{table*} An exposure time of 1 hour on the WHT enables the detection of $R$=26 objects with a signal-to-noise of 5.8 in 0.8'' seeing. This limit should correspond to a median redshift of about $z_s\\simeq$1.2. In our eventual analysis, we will introduce a brighter limit so as to keep only resolved galaxies (referred to as the survey sample). This serves to reduce the median redshift to about 0.8 (see $\\S$9.3). We however note from figure \\ref{fig:sigmag} that our expected shear signal is not very sensitive to median redshift ($\\sigma_{\\gamma} \\propto z_s^{0.8}$). In \\S\\ref{results}, we will show that the resulting depth is still sufficient to detect the lensing signal. \\subsection{Observations} We observed 14 selected fields with the WHT during the nights of 13-16 May 1999. For each field, a total of 4 exposures in $R$, each of 900s, was taken. All fields were observed as they passed through the meridian. Each exposure on a given field was offset by 10'' from its predecessor in order to remove cosmetic defects and cosmic rays, and to measure the optical distortion of the telescope and camera (see \\S\\ref{distortion} and \\S\\ref{shear_method}). All but two of the fields were observed with the long axis of the CCD pointing East-West; the exception being the two Groth fields, for which a $45^\\circ$ rotation (i.e. North-West orientation) was effected to align the WHT exposures with the HST survey (Groth et al. 1998; Rhodes 1999). Bias frames and sky flats were taken at the beginning and end of each night, and standard star observations were interspersed with the science exposures. The median seeing on our used exposures is 0.81''; one exposure with seeing $>1.2$'' was excluded. Table \\ref{tab:coords} lists the seeing and 'imcat' magnitudes corresponding to $5 \\sigma$ detections for each field. An 'imcat' signal-to-noise (see Section 7) of 5.0 corresponds to a median $R$=26.1; the median magnitude of galaxies on a field is $R$=25.2. To measure the shear, a number of cuts have to be applied to our object catalog (see \\S\\ref{shear_method}). Table \\ref{tab:coords} also lists the median magnitude of our final sample. At our final subsample limit of $R$=23.4, the median redshift is $\\simeq 0.8\\pm 0.2$ (Cohen et al 2000, $\\S$9.3). The median number density of adopted survey sources is 14.3 arcmin$^{-2}$ (see section \\ref{shear_method}). \\label{zref} In addition to the 13 useful fields, we had already obtained a test field (WHT0) in service time, and were also kindly given access to a suitable archival field, WHT17. Both were taken in good conditions: WHT0 is a 1 hour exposure in the $I$ band, whereas WHT17 is a 1.5 hour exposure in $R$ (chosen to include a known quasar). Removing these fields does not significantly alter our results. In terms of uniformity, apart from the deeper WHT17 field, the standard deviation in limiting magnitude is $\\simeq$0.2 mag which we consider acceptable for our survey. The error on magnitude zero point derived from standard stars is at the much lower level of 0.03 mag. \\subsection[]{Data Reduction} The reduction of these deep images proceeded along a standard route. A median-combined bias frame was subtracted from the skyflats and science exposures, and all such exposures were divided by a median unit-normalised sky-flat. Although the survey exposures were undertaken in the $R$ band to avoid fringing, fringing is still detected at a 0.5\\% sky level. In order to remove these fringes, which could potentially introduce structure into the image ellipticities, all long dithered exposures for a given night ($>15$ exposures per night) were stacked without offsetting with a sigma-clipping algorithm. This results in a fringe frame mapping the background fringes but devoid of foreground objects. The fringe frame for the relevant night was then subtracted from each science exposure individually, subtracting off the multiple of the fringe frame found to minimise the rms background noise. After applying this technique, the fringes are entirely imperceptible, any residue having an amplitude within the sky background noise. We experimented with automated and hand- subtraction of the fringes and verified that this had no noticeable effect on our shear analysis. The mean linear astrometric offset (in fractional number of pixels) between the four exposures was found by producing SExtractor (Bertin \\& Arnoults 1996) catalogues for each exposure, containing typically 2000-3000 objects. We used the mean offsets of the matched objects to align the fields. The images were shifted by the corresponding non-integer number of pixels using IRAF's imshift routine, taking linear combinations of neighbouring pixels to effect the non-integer pixel shifts. As discussed in \\S\\ref{distortion}, we find no need to rotate the exposures with respect to each other, or to make further astrometric distortions to compensate for the optical distortion of the instrument. The resulting four exposures for each science field were stacked with sigma-clipping. Since each exposure is 10'' away from the others, bad columns and cosmic rays were rejected. The images were examined visually and remaining defective pixels (e.g. a star just outside field of view leading to light leakage onto an area of the CCD; or highly saturated stars) were flagged as potentially unreliable. \\begin{figure} \\psfig{figure=figure5.ps,width=80mm,height=140mm} \\caption{Example reduced image (CIRSI2); the field of view is 8'$\\times$ 16'. Note that in our analysis, we divide each such field into two 8' $\\times$ 8' cells.} \\label{fig:cirsi2} \\end{figure} \\section[]{Image Analysis} \\label{object} We are now ready to measure the ellipticities of the galaxies on each field, and to apply the necessary corrections in order to take into account the smearing effect of the atmosphere (`seeing') plus tracking and other instrumental distortions introduced by the telescope and camera optics. Only then can we ascertain the true cosmic shear by averaging the ellipticity distributions of the corrected galaxies. If no shear were present on a given field, the mean ellipticity would be zero, within the noise expected from the non-circularity of galaxies and pixelisation effects. If a shear is present, the mean ellipticity will be significant, especially when results are combined from many fields. A number of methods have recently been proposed to derive the shear from galaxy shapes (Kaiser et al 1995; Rhodes et al. 1999; Kuijken 1999; Kaiser 1999). Here we choose the most documented method, namely the KSB formalism proposed by Kaiser et al (1995) and further developed in Luppino \\& Kaiser (1997) and Hoekstra et al. (1998). While this method is known to have a number of shortcomings (Rhodes et al. 1999; Kuijken 1999; Kaiser 1999), it is nevertheless the simplest and is readily available. As we will show in \\S\\ref{simulation} using simulations, the method is suitable for our purposes, after a number of precautions are taken (see Bacon et al. 2000 for more details). We therefore use this method as provided by the {\\tt imcat} software, a numerical implementation of Kaiser et al. 1995. The first task in this process is to detect all objects present on the fields down to the background noise level, and to measure their shapes. We then wish to measure their {\\em polarisabilities} i.e. measures of how each is affected by an isotropic smear (due principally to the atmosphere), an anisotropic smear (due to tracking errors at the telescope and local coaddition errors due to astrometric distortion) and shear (both the real gravitational shear and optical distortions due to the telescope and camera optics). One should note the distinction between smear and shear: a smear is a convolution of the image with a kernel, whereas a shear is a stretching of the image which conserves surface brightness. We will now describe the method for finding objects, and for measuring their ellipticities and their shear and smear polarisabilities. \\subsection{Object Detection} For the purpose of detecting cosmic shear, it is expedient to divide each of our fields into 2 $8'\\times 8'$ cells, since the signal is stronger on smaller scales (see Figure \\ref{fig:cl}). Furthermore, the mean shear correlation between two adjacent cells is expected to be about $0.7$\\% (see Section 2), and can thus be used to independently verify our results. We use the {\\tt imcat} software to find objects in each cell, and to measure their ellipticities, radii, magnitudes, and polarisabilities. The {\\tt hfindpeaks} routine convolves the cell with Mexican hat functions of varying size, and maximally significant peaks in surface brightness after convolution are designated objects. The radius of the hat giving the largest signal to noise $\\nu$ for a given galaxy is attributed to that galaxy as its filter radius $r_g$. The local sky background is estimated by the {\\tt getsky} routine, and aperture photometry is carried out on the objects, determining magnitude and half-light radius $r_h$ for all objects using the {\\tt apphot} routine. \\subsection{Shape Measurement} Using the {\\tt getshapes} routine, we then measure the weighted quadrupole moments of each object which are defined as \\begin{equation} I_{ij}\\equiv \\int d^2 x ~ w(x) x_i x_j I(\\mbox{\\boldmath $x$}) \\label{eq:qij} \\end{equation} where $I$ is the surface brightness of the object, $x$ is angular distance from object centre, and $w(x)$ is a Gaussian weight function of scale length equal to $r_g$. In this fashion we obtain ellipticity components \\begin{equation} \\label{eq:e_def} e_{i} \\equiv I_{i} / T, \\end{equation} where \\begin{equation} I_1 \\equiv I_{11}-I_{22}, \\;\\; I_2 \\equiv 2 I_{21}, \\;\\; T \\equiv I_{11}+I_{22}. \\end{equation} We can further define $e \\equiv (e_1^2+e_2^2)^{\\frac{1}{2}}$, where $e_1 = e \\cos{2 \\phi}$ and $e_2 = e \\sin{2 \\phi}$, where $\\phi$ is the position angle associated with the elongation direction of the object (anticlockwise from x-axis). The trace $T$ of the quadrupole moments provides a measure for the rms radius $d$ of the object, which we define as \\begin{equation} \\label{eq:d_def} d^{2} \\equiv \\frac{1}{2} (I_{11}+I_{22}) / I_{0}, \\end{equation} where $I_{0} \\equiv \\int d^2 x ~ w(x) I(\\mbox{\\boldmath $x$})$ is the flux of the object. \\subsection{Polarisability} The {\\tt imcat} software also enables us to calculate the smear and shear polarisabilities. In the following, we briefly review their function. It is possible (see e.g. KSB 95 Appendix) to calculate the effects of anisotropic smearing, by replacing the image $I(\\mbox{\\boldmath $x$})$ in (\\ref{eq:qij}) with a convolved (i.e. anisotropically smeared) image $I^\\prime(\\mbox{\\boldmath $x$})$ and by finding the effect on the original $e_i$. It is found that the galaxy ellipticity $e^g_{\\rm smeared}$ can be corrected for the smear as \\begin{equation} e^g_{\\rm corrected} = e^g_{\\rm smeared} - P^g_{sm} p, \\label{eq:ecorrect} \\end{equation} where the ellipticities are understood to denote the relevant 2-component spinor $e_i$, and $p$ is a measure of PSF anisotropy. The tensor $P^g_{sm}$ is the smear polarisability, a $2\\times 2$ matrix with components involving various moments of surface brightness. Since for stars $e^*_{\\rm corrected}=0$, we can set $p = (P^*_{sm})^{-1} e^*_{\\rm smeared}$, and find \\begin{equation} e^g_{\\rm corrected} = e^g_{\\rm smeared} - P^g_{sm} (P^*_{sm})^{-1} e^*_{\\rm smeared} \\label{eq:ecorrect2} \\end{equation} In this fashion, we can correct a galaxy ellipticity for the effect of anisotropic smearing, using the smear polarisability $P^g_{sm}$. In a similar manner, we can calculate the effect of a {\\em shear}, however it is induced. Replacing the image $I(\\mbox{\\boldmath $x$})$ in (\\ref{eq:qij}) with a weakly sheared image, we find that \\begin{equation} e^g_{\\rm sheared} = e^g_{\\rm initial} + P^g_{sh} \\gamma , \\label{eq:egsheared} \\end{equation} where $\\gamma$ denotes the two component shear (Eq.~[\\ref{eq:psi_def}]), and $P^g_{sh}$ is the shear polarisability, a $2\\times 2$ matrix with components involving various moments of surface brightness (different from $P^g_{sm}$ above). In practice, the lensing shear takes effect before the circular smearing of the PSF. Luppino and Kaiser (1997) showed that the {\\it pre}-smear shear $\\gamma$ averaged over a field can be recovered using \\begin{equation} \\langle {P_{\\gamma}}\\gamma \\rangle = \\langle {e^g_{\\rm corrected}} \\rangle \\label{eq:eoverpg} \\end{equation} where \\begin{equation} P_{\\gamma} = P_{sh}^g - \\frac{P_{sh}^*}{P_{sm}^*} P_{sm}^g. \\label{eq:pgamma} \\end{equation} Here, $e^g_{\\rm corrected}$ is the galaxy ellipticity corrected for smear, as in equation (\\ref{eq:ecorrect}), and $P_{sh}^*$ and $P_{sm}^*$ are the shear and smear polarisabilities calculated for a star interpolated to the position of the galaxy in question. The interpretation of the division in this equation is a matter of debate; our adopted procedure will be found in Section 7. With the smear and shear polarisabilities calculated by {\\tt imcat}, we can therefore find an estimator for the mean shear in a given cell. In summary, we can derive a catalogue of objects on a cell. For every object, we determine its centroid, magnitude, half-light and filter radii, ellipticity components and polarisabilities as defined above. We can now use these catalogues to understand and correct for systematic effects, particularly for instrumental distortion and PSF-induced effects. \\section[]{Instrumental Distortions} \\label{distortion} The instrumental distortion induced by the optical system of the telescope must be accounted for. If left uncorrected, this effect can indeed produce both a spurious shear and a smearing during the coadding process. In the following, we first present our method to measure the distortion using dithered astrometric frames. We then apply this method to our WHT fields and compare our measured distortion field to that predicted by the WHT Prime Focus manual (Carter \\& Bridges 1995). We then show how the coadding smear can be computed from the astrometric frames. We finally quantify the impact of these effects on our lensing measurement. \\subsection{Measurement of the Astrometric Distortion} The distortion field introduced by the telescope and camera optics can be measured from the astrometric shifts of objects observed in several frames offset by known amounts. Let ${\\mathbf x}$ be the true position of an object. Let ${\\mathbf x}^{f}$ be its position observed in frame $f$, without any correction for the camera distortion. The observed position can be written as \\begin{equation} {\\mathbf x}^{f} = {\\mathbf x} + {\\mathbf \\delta x}({\\mathbf x}-{\\mathbf \\overline{x}}^{f}) \\end{equation} where ${\\mathbf \\delta x}$ is the displacement produced by the distortion. The vector ${\\mathbf \\overline{x}}^{f}$ is the position of the centre of frame $f$, and can be measured as the average position of all the objects found in the image. We assume that the displacement field ${\\mathbf \\delta x}$ is the same for all frames. The position of this object observed in another frame $f'$ is ${\\mathbf x}^{f'} = {\\mathbf x} + {\\mathbf \\delta x}({\\mathbf x}-{\\mathbf \\overline{x}}^{f'})$. Here, ${\\mathbf \\overline{x}}^{f'}$ is the centre position of the new frame, which is assumed to be displaced from frame $f$ only by a translation. (This formalism can be easily extended to include a rotation of the frames about their centre, but this effect is negligible in our case). If the offset ${\\mathbf \\overline{x}}^{f} -{\\mathbf \\overline{x}}^{f'}$ is small compared to the scale on which ${\\mathbf \\delta x}$ varies, we can expand this last expression in Taylor series and get \\begin{equation} \\label{eq:delta_x} {\\mathbf x}^{f'} - {\\mathbf x}^{f} \\simeq {\\mathbf \\Psi} ({\\mathbf \\overline{x}}^{f} -{\\mathbf \\overline{x}}^{f'}), \\end{equation} where \\begin{equation} \\label{eq:psi_def} \\Psi_{ij} \\equiv \\frac{\\partial (\\delta x_{i})}{\\partial x_{j}} \\end{equation} is the distortion matrix at the location of the object as defined in Equation~(\\ref{eq:psi_def_theory}). Following the lensing conventions, the distortion matrix can be parametrised as \\begin{equation} \\label{eq:psi_params} {\\mathbf \\Psi} \\equiv \\left( \\begin{array}{cc} \\kappa +\\gamma_{1} & \\gamma_{2} + \\rho \\\\ \\gamma_{2} - \\rho & \\kappa - \\gamma_{1} \\\\ \\end{array} \\right), \\end{equation} where $\\kappa$ and $\\gamma_{i}$ are the spurious convergence and shear introduced by the geometrical distortion. We have included the rotation parameter $\\rho$, which, unlike the case of gravitational lensing, does not necessarily vanish. The 4 free parameters of the distortion matrix can thus be measured from the position of an object in 3 frames $f, f'$ and $f''$. This can be done by solving the system of 4 independent equations formed by equation~(\\ref{eq:delta_x}) and its counterpart for $f$ and $f''$. The system will not be degenerate, if the offsets ${\\mathbf \\overline{x}}^{f} -{\\mathbf \\overline{x}}^{f'}$ and ${\\mathbf \\overline{x}}^{f} -{\\mathbf \\overline{x}}^{f''}$ are not parallel. \\subsection{Distortion Field for the WHT Prime Focus} First we can compute the expected instrumental distortion using the specifications in the WHT Prime Focus manual (Carter \\& Bridges 1995). The displacement field is expected to be radial with an amplitude of ${\\mathbf \\delta x} = a r^{3} {\\mathbf \\hat{r}}$, where $r$ is the distance from the optical axis (located at $(1076.13,2010.7)$ pixels), ${\\mathbf \\hat{r}} = {\\mathbf r}/r$ is the associated unit radial vector, and $a \\simeq 4.27 \\times 10^{-10}$ pixels$^{-2}$. Using this expression in Equations~(\\ref{eq:psi_def}-\\ref{eq:psi_params}), we can compute the distortion parameters to be \\begin{eqnarray}\\ \\label{eq:dist_manual} \\kappa & = & 2 a r^{2} \\nonumber \\\\ \\gamma_{i} & = & a r^{2} \\hat{e}_{i}^{r} \\\\ \\rho & = & 0, \\nonumber \\end{eqnarray} where $\\hat{e}_{i}^{r} \\equiv \\{r_{1}^{2}-r_{2}^{2},2 r_{1}r_{2}\\}/(r_{1}^{2}+r_{2}^{2})$ is the unit radial ellipticity vector. This therefore predicts a radial instrumental shear with an amplitude growing like $r^{2}$, reaching $\\gamma \\sim 0.001$ at the edge of the chip. This expected shear pattern is shown on figure \\ref{fig:geom}. \\begin{figure} \\psfig{figure=figure6a.ps,width=80mm} \\psfig{figure=figure6b.ps,width=80mm} \\caption{Expected (top) and measured (bottom) instrumental shear pattern for the WHT Prime Focus. The expected pattern was derived from the distortion model given in the WHT Prime Focus manual (Carter \\& Bridges 1995). The observed pattern was measured using 3 astrometric frames in one of our fields.} \\label{fig:geom} \\end{figure} Figure~\\ref{fig:geom} also shows a typical instrumental shear pattern measured in one of our fields. This was derived using the method described above applied to 3 astrometric frames dithered by about 10'' and containing about 15 objects per square arcmin. The uncertainty for the mean shear component $\\gamma_{i}$ in each of the $2'\\times2'$ cell is of about 0.0005. Astrometric measurements thus allow us to measure the instrumental distortion with very high accuracy. The measured shear pattern is also approximately radial and agrees well with the expected pattern. More importantly, it also has an amplitude of at most 0.001 throughout the field. We have inspected all of our fields in this manner, and have found only small field-to-field variations (of about 0.002) for the shear patterns. In all fields, the maximum instrumental shear is only 0.003 in single $2'\\times2'$ cells. This number would be even smaller, for an average over a larger area. We also compared the convergence $\\kappa$ and $\\rho$ patterns to that expected from the WHT manual (Eq~[\\ref{eq:dist_manual}]). We again found good agreement with small field-to-field variations of about 0.002. The origin of these variations is unknown but could arise perhaps from telescope flexure. For our purposes, however, it is quite clear that the instrumental distortion is much smaller than the expected lensing signal. We therefore neglect this component in the subsequent analysis. \\subsection{Smear Arising from Co-Addition} If left uncorrected, instrumental distortions can also produce a systematic effect on the shapes of galaxies, during the coadding process. The images of a galaxy from each (distorted) frame will be slightly offset from one another, and will therefore combine into a blurred coadded image. Here, we show that this effect is equivalent to a convolution (or smear) by an additional kernel. Since this effect will equally affect the stars in the field, it will be corrected for by the PSF correction described in \\S\\ref{shear_method}. It is nevertheless important to estimate the amplitude of this effect, and to ensure that it does not dominate the dispersion of the PSF anisotropy. Let us consider the image of an object which appears on $N_{f}$ frames. As before, let ${\\mathbf x}^{f}$ be its centre position on frame $f$ (after correcting for a translation but not for the distortion). Let us choose the centre of our coordinate system to coincide with the centre-of-light ${\\mathbf x}^{o} \\equiv \\sum_{f} {\\mathbf x}^{f} / N_{f}$ of the coadded image. The coadded surface brightness is then \\begin{equation} I'({\\mathbf x}) = \\frac{1}{N_{f}} \\sum_{f=1}^{N_{f}} I({\\mathbf x}-{\\mathbf x}^{f}), \\end{equation} where $I({\\mathbf x})$ is the (undistorted) surface brightness of the object, and the factor of $N_{f}^{-1}$ was added for convenience. Note that the effect of the distortion on the object shape in individual frames was treated separately in the previous sections, and was thus ignored in this expression. It is easy to see that $I'$ can be written as a convolution of $I$ with the kernel \\begin{equation} Q({\\mathbf x}) = \\frac{1}{N_{f}} \\sum_{f=1}^{N_{f}} \\delta^{(2)}({\\mathbf x}-{\\mathbf x}^{f}), \\end{equation} where $\\delta^{(2)}$ is the 2-dimensional Dirac-Delta function. To estimate the amplitude of the effect, it is convenient and sufficient to consider the normalised unweighted quadrupole moments \\begin{equation} J_{ij} \\equiv \\int d^{2}x ~x_{i} x_{j} I(x) \\left/ \\int d^{2}x I(x) \\right., \\end{equation} (see Eq.~[\\ref{eq:qij}]) of the undistorted image, and similarly for the moments $J_{ij}'$ of the coadded image. The unweighted moments of the kernel $Q({\\mathbf x})$ are simply \\begin{equation} Q_{ij} = \\frac{1}{N_{f}} \\sum_{f=1}^{N_{f}} x_{i}^{f} x_{j}^{f} \\end{equation} Because $I$, $I'$ and $Q$ are simply related by a convolution, their respective quadrupole moments are related by $J'_{ij}=J_{ij}+Q_{ij}$ (see e.g. Rhodes et al. 1999). The rms radius $d'$ (Eq.~[\\ref{eq:d_def}]) of the coadded image is thus given by \\begin{equation} \\label{eq:dprime} d^{\\prime 2} = d^{2} + d_{q}^{2}, \\end{equation} where $d$ and $d_{q}$ are the rms radius for the undistorted image and for the kernel respectively. For simplicity, let us consider an object which is intrinsically circular. The ellipticity $e_{i}'$ of the coadded image (see Eq.~[\\ref{eq:e_def}]) is then given by ({\\it ibid}) \\begin{equation} \\label{eq:eprime} e_{i}'= \\frac{d_{q}^{2}}{d^{2}+d_{q}^{2}} e_{i}^{q}, \\end{equation} where $e_{i}^{q}$ is the ellipticity of the kernel. Turning to the specific case of the WHT observations, let us consider the ellipticity produced by the coadding smear on a star observed on 4 frames with a 0.7 arcsec circular seeing. Note that the effect will be smaller for galaxies which are extended, and so the following estimate should be considered as an upper limit. For simplicity, we conservatively assume that the seeing has a gaussian profile. We inspected all our fields and found that the astrometric offsets between the different frames was always smaller than 0.3 pixels. Using Equation~(\\ref{eq:dprime}) we calculated the change $(d'-d)/d$ in the radius of the star which is always less 2\\%, i.e. negligible compared to intrinsic changes in the seeing size. Using Equation~(\\ref{eq:eprime}) we also computed the induced ellipticity $\\epsilon'$ of the star and found it to be of the order of 0.01 and always less than 0.03. This is considerably smaller than the rms dispersion in the PSF ellipticity that we measure in our fields (about 0.07, see \\S\\ref{shear_method}) , which must therefore be due to other effects (tracking errors, atmospheric effects, etc). Again, we can conclude that smear arising via instrumental distortions during image coaddition is negligible. ", "conclusions": "" }, "0003/astro-ph0003193_arXiv.txt": { "abstract": "In this work, we raise the hypothesis that the density fluctuations field which originates the growth of large scale structures is a combination of two or more distributions, instead of assuming the widely accepted idea that the observed distribution of matter stems from a single Gaussian field produced in the very early universe. By applying the statistical analysis of finite mixture distributions to a specific combination of Gaussian plus non-Gaussian random fields, we studied the case where just a small departure from Gaussianity is allowed. Our results suggest that even a very small level of non-Gaussianity may introduce significant changes in the cluster abundance evolution rate. ", "introduction": "Generally, the problem of structure formation is associated to the gravitational growth of small density fluctuations generated by physical processes in the very early universe. Also, these fluctuations are supposed to build a Gaussian random field (GRF), where the Fourier components $\\delta_k$ have independent, random and uniformly distributed phases. Such a condition means that phases are non-correlated in space and assures the statistical properties of the GRF are completely specified by the two-point correlation function or, equivalently, by the power spectrum $P(k) = |\\delta_k|^2$, which contains information on the density fluctuation amplitude of each scale $k$. This makes the choice of a GRF the simplest initial condition for structure formation studies from the mathematical point of view. At the same time, the GRF simplicity is vindicated by a great number of inflationary models that predict a nearly scale-invariant spectrum of Gaussian density perturbations from quantum-mechanical fluctuations in the field that drives inflation (Guth \\& Pi 1982). Likewise, the central limit theorem guarantees a GRF if a wide range of random physical processes acts on the distribution of matter in the early universe. However, a number of mechanisms can generate non-Gaussian density fluctuations. For instance, they arise in some inflation models with multiple scalar fields (e.g. Salopek, Bond \\& Bardeen 1989); or after phase transitions when different types of topological defects can be formed (Kibble 1976); still, by any discrete, random distributed seed masses like primordial black holes and soliton objects (Sherrer \\& Bertschinger 1991); as well as in astrophysical processes during the non-linear regime where early generations of massive stars produce shocks which sweep material on to giant blast waves triggering formation of large-scale structure (Ostriker \\& Cowie 1981). Thus, in order to better understand the process of structure formation, it is necessary to investigate the possibility of the non-Gaussian statistics contribution to the density fluctuation field. Due to the difficulty to work with generic statistical models, the usual approach is to examine specific classes of non-Gaussian distributions. Examples of these efforts are the studies carried out by Weinberg \\& Cole (1992) that studied non-Gaussian initial conditions generated by a range of specific local transformations of an underlying Gaussian field; Moscardini et al. (1991) investigated whether non-Gaussian initial conditions can help to reconcile the CDM models with observations; and Kayama, Soda \\& Taruya (1999), who used data on the abundance of clusters at three different redshifts to establish constraints on structure formation models based on chi-squared non-Gaussian fluctuations generated during inflation. In this work, we propose a new approach to this problem, exploring the hypothesis that initial conditions for structure formation do not build a single GRF, but a combination of different fields, produced by different physical mechanisms, whose resultant effect presents an arbitrarily small departure from the strict Gaussianity. The paper is organized as follows: in Section 2, we introduce the statistical analysis of finite mixture distributions and present a two-component mixture model; in Sections 3, we apply the model to the cluster abundance evolution; in Section 4 we summarize and discuss our results. ", "conclusions": "We presented the first results of a study concerning small deviations from Gaussianity in the primordial density field. Using very simple arguments, we developed a model based on the combination of two random fields in order to take into account the non-Gaussianity effects. This model is physically motivated in the context of hybrid models, as well as in the framework of biased scenarios for structure formation. The weighted combined field involves a parameter $\\alpha$ which modulates the relative importance of its components. For $\\alpha\\approx 1$, we preserve the amplitude and shape of $P(k)$ and $\\sigma(R)$ almost the same as in the Gaussian case. At the same time, our results suggest that even very small values of $(1-\\alpha)$ can introduce a significant change in the cluster abundance evolution. This effect seems to be stronger in high density universes (at $z\\leq 1$) than in low density universes where the effect probably turns more important at higher redshifts. The model has some drawbacks. Firstly, it depends on the choice and amplitude of the second component of the combined field. Our choice of the lognormal function had a mathematical criterion of simplicity. A detailed investigation of the use of different distribution functions as the second component will be the subject of future works. However, the reasonable agreement between the model and the data gives some support to our arbitrary choice. Other possible limitation of this work comes from the use of the analytical approximation to the density of non-linear objects following Sasaki (1994). A more accurate description of the cluster abundance evolution requires the utilization of numerical methods. But Blain \\& Longair (1993), also working in the Press \\& Schecter framework, found results numerically similar to Sasaki's, so it seems that using this analytical approximation does not introduce any systematical error. Finally, we should keep in mind that our results are preliminary and both theoretical and observational efforts are necessary in order to confirm or disproof the hypothesis that the primordial density field can be described as a slightly non-Gaussian distribution." }, "0003/astro-ph0003470_arXiv.txt": { "abstract": "If mass and angular momentum are conserved in the cooling flows associated with luminous, slowly rotating elliptical galaxies, the flow cools onto extended, massive disks of rotationally supported cold gas in the equatorial plane. As the hot interstellar gas approaches the disks, its density and thermal X-ray emission increase, resulting in X-ray images that are considerably flattened toward the equatorial plane out to an optical effective radius or beyond. Remarkably, the flattening of X-ray images due to rotation is very small or absent at the spatial resolution currently available to X-ray observations. This is strong evidence that mass and angular momentum are not in fact conserved. In particular, if cooling flows are depleted by localized radiative cooling at numerous sites distributed throughout the flows, disks of cooled gas do not form and the X-ray images appear nearly circular. Even in this case, however, the spatial distribution of the cooled gas, and any young stars that may have formed from this gas, would be decidedly flatter than the old stellar population; if the young stars are optically luminous, the Balmer lines they contribute to the stellar spectra should be more elliptical than the total stellar light. In principle, X-ray images of galactic cooling flows can also be circularized by the turbulent diffusion of angular momentum away from the axis of rotation. But the effective viscosity of known processes -- stellar mass loss, supernovae, cooling site evolution, etc. -- is insufficient to circularize the X-ray images appreciably. Radial gradients in the interstellar iron abundance in elliptical galaxies similar to those observed are unaffected by the expected level of interstellar turbulence since these gradients are quickly re-established by Type Ia supernovae. ", "introduction": "Cooling flows in massive elliptical galaxies are expected to rotate since most of the interstellar gas within an optical effective radius $R_e$ is produced by mass loss from the slowly rotating stellar systems. If the cooling flow proceeds inward conserving angular momentum, the interstellar gas ultimately flows toward a large disk ($R_{disk} \\gta R_e$) and spins up to the local equatorial circular velocity, $\\sim 400$ km s$^{-1}$ (Kley \\& Mathews 1995; Brighenti \\& Mathews 1996, subsequently referred to as Paper I). As interstellar gas approaches the disk, the X-ray images are significantly flattened toward the equatorial plane on scales $\\sim R_e - 3R_e$ when viewed perpendicular to the axis of rotation. For bright elliptical galaxies at distance 17 Mpc in the Virgo cluster -- NGC 4472, NGC 4636, and NGC 4649 -- $R_e \\sim 7~{\\rm kpc}~\\sim 1.4$'. However, {\\it Einstein} X-ray images show no evidence of rotational flattening either at the resolution of the IPC ($\\sim 55'' \\sim 4.5$ kpc; Trinchieri et al. 1986) or at the much higher resolution or HRI ($\\sim 6'' \\sim 0.49$ kpc; Fabbiano, Kim \\& Trinchieri 1992). In a series of papers Buote \\& Canizares have detected large scale X-ray flattening in several massive elliptical galaxies which they attribute to a flattened dark matter potential, but the highly flattened inner isophotes anticipated from momentum-conserving cooling flows were not evident (see Buote \\& Canizares 1998 for a review). More recently, Hanlan \\& Bregman (1999) have found little or no rotationally enhanced X-ray ellipticity in ROSAT HRI and PSPC images of several bright elliptical galaxies. The implications of these observations are not widely appreciated among the community of astronomers interested in early type galaxies. Massive, X-ray luminous elliptical galaxies have two principal sources of hot interstellar gas: (1) ejection of stellar envelopes from an evolving population of old stars and (2) inflow of distant circumgalactic gas which accumulates in the galactic halo either by tidal exchange or secondary infall (Brighenti \\& Mathews 1999a). If the circumgalactic gas has a lower specific angular momentum than the stars, the X-ray flattening would be lessened as inflowing circumgalactic gas mixes with gas lost from the rotating stellar system within several $R_e$. But it is unlikely that X-ray disks can be eliminated by this means since most of the gas in the inner galaxy is produced by stellar mass loss and has not flowed in from the galactic halo region. Our recent models for the large elliptical galaxy NGC 4472 require that at least 60 - 70 percent of the hot gas in $r \\lta R_e$ originates from stellar mass loss. This rate of stellar mass ejection is expected from normal stellar evolution and is required to explain the radial variation of interstellar temperature and metallicity typically observed in massive elliptical galaxies (Brighenti \\& Mathews 1999a). In addition, it is unlikely that the absence of observed X-ray flattening can be understood simply by the inflow of halo gas into the optical centers of elliptical galaxies since it is plausible that the extended halo gas has an even {\\it larger} specific angular momentum than the stars. N-body simulations of galaxy formation typically produce elliptical galaxies and galaxy groups with increasing specific angular momentum at large radii (e.g. Barnes \\& Efstathiou 1987; Quinn \\& Zureck 1988). But the intrinsic angular momentum in distant halo gas is uncertain; gas acquired by tidal disruption is expected to have significant rotation, but secondary infalling gas that arrives after most mergers have occurred may have less net rotation than the stellar system. However, there is no doubt that the stellar systems in most large elliptical galaxies and the gas that they expel are rotating. Although the stellar systems in luminous elliptical galaxies (with $L_B \\gta 3 \\times 10^{10}$ $L_{B,\\odot}$) are not flattened by rotation, their rotation about the minor axis is not small, typically 50 - 100 km s$^{-1}$ (e.g. Binney, Davies, \\& Illingworth 1990), and the gas they expel must spin up further as it moves inward. These general considerations indicate that the net angular momentum of gas ejected from stars must be removed to circularize the X-ray images. This can be accomplished either by removing gas from the flow or by transporting angular momentum away from the axis of rotation by viscous interactions. In this paper we explore the importance of these two processes and determine their relative influence on the X-ray images. The first process, localized dropout cooling of interstellar gas in regions of low specific entropy, is also required to limit the masses of central black holes (or dark stellar nuclei) in elliptical galaxies to their observed values. Central black holes would be about ten times more massive if the cooling flow proceeded all the way to the galactic center with constant mass flux (Brighenti \\& Mathews 1999b). The presence of distributed optical emission lines from cooling gas also argues for cooling dropout (Mathews \\& Brighenti 1999a). However, as we show below, the strongest argument for the presence of localized mass dropout in the interstellar gas may be the circularization of X-ray images that results. We also explore the possibility that interstellar viscosity transports angular momentum outward, reducing the rotation and X-ray flattening in the inner galaxy. An effective viscosity can arise from several natural sources: plasma viscosity, interstellar ``turbulent'' viscosity driven by stellar mass loss or Type Ia supernovae, the viscosity implicit in regions of localized interstellar cooling as they sink in the galactic potential, and turbulent viscosity that may develop from shear instabilities in the flow. On several occasions Nulsen and Fabian have suggested that the turbulent viscosity is so large that the entire interstellar medium becomes approximately spherical (Nulsen, Stewart, \\& Fabian 1984; Nulsen \\& Fabian 1995), but no detailed calculations were provided to support this conjecture. In the following section we describe rotating cooling flows in which angular momentum is lost by cooling dropout. Then we study the outward transport of angular momentum by turbulent viscosity and the associated diffusion of interstellar metallicity. ", "conclusions": "In this paper we have addressed the curious absence of X-ray disks in rotating, luminous elliptical galaxies. In our previous hydrodynamic studies of rotating cooling flows (Paper I; Brighenti \\& Mathews 1997) both angular momentum and mass were conserved and the resulting X-ray images were dramatically flattened out to 1-2 effective radii when viewed perpendicular to the axis of rotation. These results appear to be inconsistent with the more circular X-ray images found by observations with {\\it Einstein} HRI (Fabbiano, Kim \\& Trinchieri 1992) and ROSAT HRI (Hanlan \\& Bregman 1999). Nevertheless, the X-ray shapes of elliptical galaxies are difficult to determine from the observations and we may need to wait for the next generation of X-ray telescopes with higher spatial resolution for definitive determinations of X-ray ellipticities. It is likely that interstellar rotational flattening will eventually be observed. In the preceding discussion we have explored the circularizing influence of mass dropout and turbulent transfer of angular momentum on the X-ray images of rotating galactic cooling flows. Strong theoretical and observational arguments support both of these possibilities. Localized radiative cooling leading to mass dropout throughout the cooling flow is indicated by the relatively low masses of central black holes and adjacent nuclear regions when compared to the mass of interstellar gas that is expected to cool over cosmic time. Further evidence for cooling dropout is provided by optical line emission within $R_e$ and the evidence in this same region for a young stellar population having masses that extend up to 1 - 2 $M_{\\odot}$ but not beyond (Ferland, Fabian, \\& Johnstone 1994; Mathews \\& Brighenti 1999a). Interstellar turbulence is also very likely in elliptical galaxies, generated by mass transport associated with stellar mass loss, Type Ia supernovae and mass dropout. Indirect evidence for cooling flow turbulence is provided by interstellar magnetic fields of several $\\mu$G at $\\sim 10 R_e$, which are typically observed in elliptical galaxies having double radio sources. Fields of this magnitude can be understood as originating from small seed fields ejected from mass-losing stars, followed by subsequent field amplification by an interstellar turbulent dynamo mechanism (Moss \\& Shukurov 1996; Mathews \\& Brighenti 1997). Due to the field concentration associated with the inward, converging motion of the cooling flow, interstellar magnetic fields are further magnified and are expected to be particularly strong, perhaps exceeding equipartition values, in the central regions of cooling flows. Owen and Eilek (1998) find fields of 10 - 100 $\\mu$G within $r = 50$ kpc in the bright cD galaxy NGC 6166. Elliptical galaxies of lower optical luminosity $L_B \\lta 3 \\times 10^{10}$ $L_{B,\\odot}$ differ in many qualitative ways from the more massive elliptical galaxies similar to the one we have studied here (Faber et al. 1997). Low luminosity elliptical galaxies rotate faster and are rotationally flattened. We have shown (Brighenti \\& Mathews 1997) that rotating, non-dropout cooling flows in these galaxies also form massive, extended disks of cooled gas. However, when $q = 1$ mass dropout is included in these calculations, disks of cold HI (or H$_2$) do not form from the rotating cooling flow. There is observational evidence for extended disks of cold HI gas in some low luminosity elliptical galaxies, but these HI disks often extend far beyond the optical images of the galaxies, indicating that they were created by a (possibly very old) merging event rather than by cooling flow dropout (Oosterloo, Morganti \\& Sadler 1999a; 1999b). In some cases the cold gaseous disks have central holes in neutral hydrogen replaced with HII emission (Oosterloo, Morganti \\& Sadler 1999), suggesting a more complete conversion into stars in this part of the disk where gas pressures and densities are highest and star formation (and ionization by PAGB stellar UV) should be most efficient. In general, systematically larger H$\\beta$ features are observed in the stellar spectra of elliptical galaxies with $L_B \\lta 3 \\times 10^{10}$ $L_{B,\\odot}$, indicating considerable recent star formation (de Jong \\& Davies 1997). Unfortunately, thermal X-ray emission from the hot interstellar gas in these low-luminosity, low-$\\Sigma_x$ ellipticals is masked by the collective emission from low mass X-ray binaries. A detection of X-ray rotational flattening in their cooling flows will require spectral separation of the hot gas and stellar X-rays with high spatial resolution, clearly a job for the next generation of X-ray telescopes. When mass dropout and turbulence are considered in rotating cooling flows inside luminous elliptical galaxies, we find that: \\begin{itemize} \\item[(1)]{ Mass dropout alone strongly circularizes the X-ray images in rotating cooling flows since gas is removed from the flow before it has moved very far from its point of origin (stellar mass loss) toward the axis of rotation. } \\item[(2)]{ Conversely, the absence of strong rotational flattening in X-ray images of elliptical galaxies is persuasive evidence for distributed mass dropout. } \\item[(3)]{ In the presence of mass dropout by radiative cooling with $q = 1$, no cold gaseous disks form on the equatorial plane having radii larger than our innermost grid size, $\\sim 150$ pc. This is due to the sensitivity of radiative cooling (and associated mass dropout) to the local density, $(\\partial \\rho / \\partial t)_{do} \\propto q \\rho/t_{do} \\propto \\rho^2$. } \\item[(4)]{The spatial distribution of cooled gas mass is markedly flatter than that of the old stellar population, especially near the galactic center. Any optical signature of a younger stellar population formed from the cooled gas, such as the H$\\beta$ index, should also exhibit a significantly higher ellipticity than that of the older, background stars. } \\item[(5)]{ The estimated viscosity from known sources of interstellar turbulence is dominated by motions induced by Type Ia supernovae; stellar mass loss and mass dropout are smaller sources of interstellar turbulent viscosity. } \\item[(6)]{ Supernova-induced turbulent viscosity is insufficient to circularize the X-ray appearance of rotating elliptical galaxies; the viscous effects of stellar mass loss and cooling dropout on the X-ray images are even smaller. } \\item[(7)]{ Much larger, spatially uniform turbulent viscosities can circularize the X-ray isophotes throughout the galaxy; such turbulent viscosities could result from rotational shear instabilities, but we have not demonstrated this instability here. } \\item[(8)]{ The spatial diffusion of interstellar iron due to turbulent viscosity does not appreciably reduce the interstellar iron abundance gradients observed in bright elliptical galaxies, provided the viscosity does not exceed $\\sim 100$ gm cm$^{-1}$ s$^{-1}$. This follows since the observed interstellar iron gradients are re-established on short time scales by Type Ia supernovae and, to a lesser extent, by stellar mass loss. } \\end{itemize}" }, "0003/astro-ph0003120_arXiv.txt": { "abstract": "\\noindent At sufficiently low frequencies, no ground-based radio array will be able to produce high resolution images while looking through the ionosphere. A space-based array will be needed to explore the objects and processes which dominate the sky at the lowest radio frequencies. An imaging radio interferometer based on a large number of small, inexpensive satellites would be able to track solar radio bursts associated with coronal mass ejections out to the distance of Earth, determine the frequency and duration of early epochs of nonthermal activity in galaxies, and provide unique information about the interstellar medium. This would be a ``space-space\" VLBI mission, as only baselines between satellites would be used. Angular resolution would be limited only by interstellar and interplanetary scattering. ", "introduction": "Ground-based radio interferometers are able to produce images of the sky at frequencies down to a few tens of MHz. Some important scientific goals, however, require imaging at even lower frequencies. Absorption and refraction by the ionosphere prevents imaging from the ground at frequencies of a few MHz and lower, so an interferometer array composed of inexpensive satellites will be needed. Suitable locations for a space-based array include very high Earth orbits, halo orbits about the Sun-Earth Lagrange points, Earth-trailing heliocentric orbits, the far side of the Moon, and (perhaps) lunar orbit. The optimal choice depends on financial considerations and the unavoidable tradeoff between a benign environment in which to maintain a multi-satellite array and the difficulty of getting enough data from the array to Earth. ", "conclusions": "" }, "0003/astro-ph0003299_arXiv.txt": { "abstract": " ", "introduction": "The shape and the intensity of the intergalactic UV background radiation are crucial factors in determining the ionization balance of the intergalactic medium and therefore influence the structure formation in the universe. Knowledge of this radiation field is thus necessary for the understanding of the early universe. AGNs are believed to be the major contributors to this background, though a significant contribution from star forming galaxies can not be ruled out. Detailed calculations of the propagation of AGN like ionizing radiation through intergalactic space, taking into account the absorption and reradiation by the galactic and intergalactic material, have been carried out by Haardt and Madau (1996, hereafter HM96). They have determined the frequency and redshift dependence of the background. Observationally, the intensity of the radiation has been determined in recent years, by studying the proximity effect in the Lyman alpha forest of the absorption lines in the spectra of QSOs (Bajtlik, Duncan \\& Ostriker 1988). This analysis is insensitive to the shape of the radiation (Bechtold, 1994 and Das and Khare, 1997). Values of the intensity of the background at the Lyman limit, J$_{\\rm \\nu_{LL}}$, obtained by Bechtold (1994) and Cooke et al (1996) are considerably higher than the value expected from the distribution of visible QSOs. Several sources of uncertainty in the value of the flux obtained by the proximity effect analysis have been considered by various authors (Bechtold, 1994, 1995, Srianand and Khare, 1995, Das and Khare, 1997). It has also been suggested (Fall \\& Pei 1993) that the actual number of QSOs may be larger than their observed number and that several QSOs may be rendered invisible due to dust extinction in the intervening absorbers. It is possible that the background radiation gets a significant contribution from star forming galaxies (Madau \\& Shull 1996, Giroux \\& Shapiro 1996, Khare and Ikeuchi, 1998). Here we try to obtain an independent estimate of the background flux by studying the ionization state of the QSO absorption systems for which an estimate of the particle density is available from the observations of the fine structure excited lines of C II. Where ever possible, we also try to estimate the contribution of the galactic flux to the total ionizing flux for the systems. In section II we present our analysis, the results are discussed in section III. ", "conclusions": "For 5 of the systems we could derive the range of flux values assuming the radiation to be AGN type. All of these are higher than the corresponding HM96 values by minimum factors ranging from 1.2 to 158. Note that we have taken into account the uncertainties in the column densities of all the ions which is the reason for obtaining large ranges for the flux values. For three of these systems we could obtain a minimum value for the flux of the AGN background. These values are 1.1, 3.8 and 98.8 times higher than the HM96 values at the appropriate redshifts. For these systems a large flux is needed from galaxies. For 4 other systems a lower limit to the flux could only be obtained with an assumption of the lower limit on the particle density to be 0.045 cm$^{-3}$, which is about half of the mean interstellar value of the particle density and which indicates that the actual C II column densities are higher than the observed lower limits by 0.64 to 1.37 dex. This is a reasonable lower limit as the systems being considered are Lyman limit or damped Lyman alpha systems and also as this value is considerably lower than the range of density values for systems for which the values could be obtained from the observations. For these systems, the required values of flux are higher than the HM96 values by minimum factors of 2.7 to 62. On the other hand, assuming the AGN flux to be that given by HM96, and assuming the rest of the required flux to be of local, galactic origin, very high galactic flux is required. For most of the systems, this high flux requires the absorption systems to be present with in 100 parsecs of typical O stars. The typical radius of the Stromgren spheres of these stars is of the same order, indicating that the absorption systems are inside the H II regions. Such conclusions have earlier been rejected on the basis of statistical arguments about the properties of the absorption systems (Srianand and Khare, 1994). The flux could come from QSOs which happen to lie close to the lines of sight at redshifts similar to the redshifts of the absorption systems. We have searched the catalogues for presence of any such QSOs near the line of sight to Q2231-00. However, no QSO is found to lie closer than 1000 Mpc to the line of sight within the required redshift range. The high values of the flux indicated by our analysis for almost all the systems, may be interpreted to indicate the presence of an unseen population of dust extinct QSOs. Note that in all our analysis we have assumed that all the ions producing absorption in a given velocity range in an absorption system are physically located in the same region (cloud). This may not be always valid. Kirkman and Tytler (1999) and Churchill and Charlton (1999) have found evidence for ions with the same velocity structure in their absorption lines belonging to a given redshift system, arising in physically different gaseous components. If C IV ions are from a more widely distributed component, then, the C II/C IV column density ratio in the region of interest will be smaller and may require lower values of the flux." }, "0003/astro-ph0003250_arXiv.txt": { "abstract": "K and M giants show variability on timescales from years down to days and possibly even hours. I discuss the contribution that can be made with high-precision photometry that will be obtained by the \\MONS{} Star Trackers. These include observations of flare-like events on Mira variables, and oscillation spectra for K~giants and short-period M~giants. ", "introduction": "The Star Trackers on the \\MONS{} satellite (Bedding \\& Kjeldsen, these Proceedings) should produce exquisite light curves for many hundreds of red giant stars. These observations, made over about 30 days with high duty cycle, will allow a number of questions to be addressed. Classes of stars are discussed in order of decreasing effective temperature, starting with the Mira variables. ", "conclusions": "" }, "0003/astro-ph0003066_arXiv.txt": { "abstract": "In this study we present two indicators that will reflect the difference between collapsars and normal collapse-driven supernovae. They are products of explosive nucleosynthesis and neutrino emission. In the collapsar model, it is natural to consider that the product of explosive nucleosynthesis depends on the zenith angle because the system becomes highly asymmetric in order to generate a fire ball. We also consider the detectability of the HNRs which is located nearby our Galaxy. As a result, the number of the HNRs is estimated to be 5 $\\times$ ($10^2$ -- $10^{-3}$), whose chemical composition can be spatially resolved. Using the optimistic estimate, more HNRs will be found and it will be possible to discuss on the chemical composition statistically. As for the energy spectrums of neutrinos, they are not thermalized in a collapsar because the density of the accretion disk is much lower than that of a neutron star. The energy spectrums of (anti-)electron neutrinos from hypernovae will be mainly determined by the process of electron (positron) capture on free proton (neutron). It is also noted that high energy tail is not dumped in the case of hypernovae because the density of emitting region is low. Total energy of neutrino from hypernovae will depend on a lot of physical parameters such as total accreting mass and mass accretion rate, which are quite contrary to the situation of the normal collapse-driven supernovae. Therefore there will be a large variety of total neutrino's energies among collapsars. In the case of SN 1998bw, we think that the matter around the equatorial plane might be ejected from the system, which resulted in the formation of relatively weak jets and faint GRB 980425. ", "introduction": "\\label{intro} It will be a big progress that the fact that at least a part of the gamma-ray bursts (GRBs) comes from the hypernova (HN) explosions is being supported by the observations. It was reported for the first time that there seems to be a physical connection between GRB 980425 and SN 1998bw~\\cite{galama98}. They discovered an optical transient within the BeppoSAX Wide Field Camera error box of GRB 980425. Then they reported that the optical transient can be interpreted to be the light curve of SN 1998bw. As for the explosion energy of SN 1998bw, it was estimated to be as high as (20-50)$\\times 10^{51}$ erg as long as we believe the explosion is spherically symmetric~\\cite{iwamoto98,woosley99}. This is the reason why SN 1998bw is called as a HN. The late afterglow of GRB 970228 also suggests the physical connection between GRB and HN~\\cite{reichart99}. It was shown that the optical light curve and spectrum of the late afterglow of GRB 970228 are well reproduced by those of SN 1998bw transformed to the redshift of GRB 970228. The afterglow of GRB 980326 is also believed to be the evidence for the GRB/HN connection due to the same reason~\\cite{bloom99}. If we believe that a part of the GRBs comes from the explosion of massive stars, the explosion must be a jet-induced one because spherical explosion model has a difficulty in avoiding the baryon contamination problem~\\cite{rees92}. In fact, some observations on GRBs are interpreted as evidence for the jet-induced explosion. For example, the breaks in the rate of decline of several afterglows can be explained by the beaming effect~\\cite{kulkarni99}. The light curve and spectrum of SN 1998bw also seem to suggest a jet-induced explosion~\\cite{hoflich99,nomoto_pri}. There are also some excellent numerical simulations on the jet-induced explosion of massive stars whose aim is to reproduce the fire ball~\\cite{khokhlov99,macfadyen99,aloy00}, although the fire ball has not been reproduced yet. Here we must note the following two points. (i) it is not determined that all of the GRBs come from HNe. (ii) the explosion energy of SN 1998bw may be small if the explosion is the jet-induced one. Taking these points into consideration, we can classify the relation of GRB, SN, and HN as shown in Figure~\\ref{fig1}. For example, region (a)/(d) means that HNe which didn't/ did generate GRBs. We note that HN $\\cap$ SN = $\\phi$ by definition. Here we defined that SN is the explosion of a massive star whose total explosion energy is about $10^{51}$ erg. HN is defined as the explosion of a massive star whose total explosion energy is significantly larger than $10^{51}$ erg. As for the region (b), other systems such as the merging neutron stars~\\cite{ruffert99} may belong to this region. One of the most famous model to realize a GRB from a death of a massive star is the collapsar model~\\cite{woosley93,macfadyen99,woosley99a}. The definition of the collapsar is written in~\\cite{woosley99a} as a massive star whose iron core has collapsed to a black hole that is continuing to accrete at a very high rate. Woosley also pointed out that there will be two types for collapsars. One (type I collapsar) is that the central core immediately forms a black hole with an accretion disk. The other (type II collapsar) is that the central core forms a neutron star at first, but the neutron star collapses to be a black hole with an accretion disk due to the continuous fall back. In both types, a strong jet, which is required to produce a GRB, is generated around the polar region due to the pair-annihilation of neutrinos that come from the accretion disk and/or MHD processes. The remnants of a collapsar will belong to the regions (a), (b), (c), and (d) in Figure~\\ref{fig1}. When the explosion energy of a collapsar is small, it will be classified as SNR. When the hydrogen envelope exists, a collapsar can not produce a GRB. Here we note that there are no observations that support directly the scenario of collapsars. This situation is a contrast to that of the scenario of collapse-driven SN, which is supported by the detection of neutrinos at Kamiokande~\\cite{hirata88} and IMB~\\cite{mathews88}. Thus we present in this study two observable indicators that reflect the mechanism of collapsars. These observations will affirm the difference between collapsars and collapse-driven SN clearly. These are products of explosive nucleosynthesis and neutrino emission. We will discuss these essential features in the following sections. We also discuss the possibility of detection of such observations taking the event rate into consideration. In section~\\ref{nucleosynthesis}, we consider the explosive nucleosynthesis in the collapsar model. The luminosity and spectrum of neutrino from collapsars are shown in section~\\ref{neutrino}. Summary and discussion are presented in section~\\ref{summary}. ", "conclusions": "\\label{summary} \\indent In this study, characteristic products of nucleosynthesis and neutrino emission have been proposed as two indicators that will reflect the features of the collapsars. We consider the detectability of the HNRs because we can not distinguish well whether it is the remnant of a collapsar or of a rotating collapse-driven supernova when we find an asymmetric SNR. As a result, the number of the HNRs is estimated to be 5 $\\times$ ($10^2$ -- $10^{-3}$), whose chemical composition can be spatially resolved. Using the optimistic estimate, more HNRs will be found and it will be possible to discuss on the chemical composition statistically. Due to such observations, we will be able to determine which model is realistic and which model is unrealistic. Such observations may also give a light on the occurrence frequency of the type I collapsar relative to the type II collapsar. Moreover, we can say that the HNR event rate seems larger than the lower estimate for the GRB rate if NGC 5471B and MF83 in M101 are really HNRs. Although other interpretations are possible for these highly luminous X-ray sources~\\cite{chu99}, search for the hypernova remnants nearby our Galaxy has a potential to reveal the mechanism of the GRB. Strictly speaking, there will be a little difference between the SNRs of collapsars and those of collapse-driven supernovae. We think that an extreme jet-induced explosion like collapsars will not happen in the case of SN. This is because almost all of the matter has to be ejected in order not to leave a black hole but to leave a neutron star at the center. That is, matter around the equatorial plane has to be also ejected, which will be observed as `jet-like` explosion like SN 1987A~\\cite{nagataki00}. On the other hand, an extreme jet-induced explosion is required in order to make fire balls for the model of the jet-induced HN. So, even if the matter around the equatorial plane is ejected due to some reasons in the case of HN too, the degree of jet-induced explosion will be very large and chemical composition will depend strongly on the zenith angle in the case of the type I collapsar. It is also noted that the mass accretion rate becomes low if the matter around the equatorial plane is ejected from collapsars. This will result in the decline of the total energy of neutrinos emitted from the accretion disk. As a result, total explosion energy may become small in that case. It is reported that the explosion energy of GRB 980425, which is said to be associated with SN 1998bw, is quite lower than that of the usual GRBs~\\cite{galama98}. In the case of SN 1998bw, we think that the matter around the equatorial plane might be ejected from the system, which resulted in the formation of relatively weak jets and faint GRB 980425. This means that SN 1998bw and GRB 980425 may be classified in the region (e) in Figure~\\ref{fig1}. Of course, this picture requires that the system of SN 1998bw and GRB 980425 is highly asymmetric, because the total explosion energy of SN 1998bw is estimated to be (20-50)$\\times 10^{51}$ when spherical explosion is assumed~\\cite{iwamoto98,woosley99}. As for the (anti-)electron neutrino emission from the collapsars, its energy spectrum is mainly determined by the emission rate due to electron (positron) capture on proton (neutron). As the temperature becomes higher, contribution of the process of electron-positron pair annihilation can not be negligible. It is also noted that high energy tail is not dumped in the case of the collapsar because the density of emitting region is low. These features on energy spectrum are quite different from that of SN. Total energy of neutrino depends on many physical quantum such as total accreting mass and mass accretion rate. It is noted the emission rate due to the electron capture on proton is proportional to $T^{6}$. So a little change in temperature results in great change in the neutrino flux. That is why there will be a variety of total luminosity of neutrino among collapsars, which is in striking contrast to the case of SN. As for the event rate, the detection probability of the collapsar can be as large as that of the collapse-driven SN if we use the optimistic event rate $\\sim \\; 5\\times10^{-2}$ per year at Super-Kamiokande. Finally, we stress again that these features on nucleosynthesis and neutrinos will reveal the mechanism of GRB quite well. We hope the increase of further observations in the near future." }, "0003/gr-qc0003021_arXiv.txt": { "abstract": "Gravitational wave experiments will play a key role in the investigation of the frontiers of cosmology and the structure of fundamental fields at high energies, by either setting stringent upper limits on, or by detecting the primordial gravitational wave background produced in the early-Universe. Here we discuss the impact of space-borne laser interferometric detectors operating in the low-frequency window $\\sim 10^{-6}$ Hz -- 1 Hz; the aim of our analysis is to investigate whether a primordial background characterized by a fractional energy density $\\Om \\sim 10^{-16}$ -- $10^{-15}$, which is consistent with the prediction of \"slow-roll\" inflationary models, might be detectable by the Laser Interferometer Space Antenna (LISA) or follow-up missions. In searching for stochastic backgrounds, the presently planned LISA mission suffers from the lack of two detectors with uncorrelated noise. We analyze the sensitivity improvements that could be achieved by cross-correlating the data streams from a pair of detectors of the LISA class; we show that this configuration is extremely compelling, leading to the detection of a stochastic background as weak as $\\Om \\simeq 5\\times 10^{-14}$. However, such instrumental sensitivity can not be fully exploited to measure the primordial component of the background, due to the overwhelming power of the signal produced by large populations of short-period solar-mass binary systems of compact objects. We estimate that the primordial background can be observed only if its fractional energy density $h_{100}^2\\,\\Omega$ is greater than $\\approx 5\\times 10^{-13}$. The key conclusion of our analysis is that the stochastic radiation from unresolved binary systems sets a fundamental limit on the sensitivity that can be achieved in searching for the primordial background in frequencies between $\\sim 10^{-6}$ Hz and 0.1 Hz, regardless of the instrumental noise level and the integration time. Indeed, the mHz frequency band, where LISA achieves optimal sensitivity, is not suitable to probe slow-roll inflationary models. We briefly discuss possible follow-up missions aimed at the frequency region $\\sim 0.1$ Hz -- 1 Hz, which is likely to be free from stochastic backgrounds of astrophysical origin: no fundamental limits seem to prevent us from reaching $\\Om \\sim 10^{-16}$, although the technological challenges are considerable and deserve careful study. ", "introduction": "\\label{sec:int} The Universe became \"thin\" to gravitational waves (GWs) at the Plank epoch, corresponding to the cosmic time $\\sim 10^{-43}$ sec.; the gravitons decoupled from the surrounding plasma at a temperature of the order of the Planck mass $\\sim 10^{19}$ GeV, and gravitational radiation produced at that epoch or later -- including the electro-weak and the Grand Unification (GUT) scale -- has travelled undisturbed to us, carrying full information about the state of the Universe, and the physical processes from which it took origin. Indeed GW experiments will open radically new frontiers for cosmology and high energy physics (see~\\cite{Maggiore99,Creighton99} and reference therein for an extensive discussion). In the time frame $\\sim$ 2002-2010 a large portion of the GW spectrum will progressively become accessible, mainly through large-scale laser interferometers. On the ground, the world-wide network of km-size interferometers -- LIGO, GEO600, VIRGO and TAMA -- sensitive in the frequency band $\\sim 10\\,{\\rm Hz} - 1\\,{\\rm kHz}$, will start carrying out \"science runs\" at the beginning of 2002, with the realistic goal of directly detecting GW's. Several instrumental upgrades, starting around 2005, will drive the sensitivity of the instruments to a GW stochastic background from $\\Om \\sim 10^{-6}$ (for the so-called initial generation) to $\\Om \\sim 10^{-10}$ (for the so-called advanced configuration). In space, a collaboration between ESA and NASA is carrying out the project called LISA (Laser Interferometer Space Antenna). This is a space-borne laser interferometer with arms of length $5\\times 10^6$ km, planned to fly by 2010~\\cite{LISA_ppa}. This instrument guarantees the detection of GW's at low frequencies ($\\sim 10^{-5}\\,{\\rm Hz} - 10^{-2}\\,{\\rm Hz}$). The purpose of this paper is to show the central role of the experiments in the low-frequency window $\\sim 10^{-6}$ Hz -- 1 Hz, with emphasis on instruments of the LISA class, in the search for the primordial GW background. Our aim is to identify the fundamental issues regarding the achievement of a sensitivity in the range $h_{100}^2\\,\\Omega \\sim 10^{-16}$ -- $10^{-15}$, which is set by the theoretical prediction of \"slow-roll\" inflationary models. \\subsection{The stochastic background spectrum} \\label{subs:spect} A stochastic GW background is a random process that can be described only in terms of its statistical properties. Without loss of generality, for the issues discussed in this paper, we assume it to be isotropic, stationary, Gaussian and unpolarized. The energy and spectral content of a stochastic background are described by the dimensionless function \\be \\Omega (f) \\equiv \\frac{1}{\\rho_{\\rm c}}\\,\\frac{d\\rho_{\\rm gw}(f)}{d\\ln f}\\,; \\label{omegagw} \\ee $\\rho_{\\rm gw}$ is the energy density carried by the background radiation, and \\ba \\rho_{\\rm c} = \\frac{3\\,H_0^2\\,c^2}{8\\pi\\,G_N} & \\approx & 1.6\\times 10^{-8}\\,h_{100}^2 \\,{\\rm erg}/{\\rm cm}^3\\,, \\nonumber\\\\ & \\approx & 1.2\\times 10^{-36} \\,h_{100}^2\\,{\\rm sec}^{-2} \\label{rhoc} \\ea is the {\\it critical energy density} required today to close the Universe. The value of the Hubble constant (today) is \\ba H_0 & = & 100\\,h_{100}\\,\\,{\\rm km}\\,{\\rm sec}^{-1}\\,{\\rm Mpc}^{-1}\\,,\\nonumber\\\\ & \\simeq & 3.2\\times 10^{-18}\\,h_{100}\\,\\,{\\rm sec}^{-1}\\,, \\label{H0} \\ea where $h_{100}$ is known from observations to be in the range $0.4 \\le h_{100} \\le 0.85$. $\\Omega(f)$ is therefore the ratio of the GW energy density to the critical energy density per unit logarithmic frequency interval; one usually refers to $h_{100}^2\\Omega(f)$, which is independent of the {\\it unknown} value of the Hubble constant. It is useful to introduce the {\\it characteristic amplitude} $\\hc(f)$ of the GW background: it is the dimensionless characteristic value of the total GW background-induced fluctuation $h(t)$ at the output of an interferometer per unit logarithmic frequency interval: \\be \\m h^2(t) \\M = 2 \\int_0^\\infty\\, d(\\ln f) \\hc^2(f)\\,; \\label{hc} \\ee here $\\m \\,\\M$ denotes the expectation value. The spectral density $S(f)$ of the background is related to $\\hc(f)$ by~\\cite{Maggiore99} \\be \\hc^2(f) = 2 f S(f)\\,, \\label{Sbg} \\ee and $\\Omega (f)$, $\\hc(f)$, and $S(f)$ satisfy the relation~\\cite{Maggiore99} \\be \\Omega(f) = \\frac{2\\pi^2}{3 H_0^2}\\,f^2\\,\\hc^2(f) = \\frac{4\\pi^2}{3 H_0^2}\\,f^3\\,S(f)\\,. \\label{OhS} \\ee The characteristic amplitude over a frequency band $\\Delta f$ is therefore: \\ba \\hc(f,\\Delta f) & = & \\hc(f)\\,\\left(\\frac{\\Delta f}{f}\\right)^{1/2}\\nonumber\\\\ & \\simeq & 7.1\\times 10^{-22}\\, \\left[\\frac{\\Om(f)}{10^{-8}}\\right]^{1/2}\\, \\left(\\frac{f}{1\\,{\\rm mHz}}\\right)^{-3/2}\\, \\left(\\frac{\\Delta_{\\rm b} f}{3.2\\times 10^{-8}\\,{\\rm Hz}}\\right)^{1/2}\\,, \\label{hc1} \\ea where $\\Delta_{\\rm b} f \\simeq 3.2\\times 10^{-8}\\,(1\\,{\\rm yr}/T)$ Hz is the width of the frequency bin for an observation time $T$. For comparison, the relevant characteristic amplitude of the LISA noise is $\\sim 10^{-24}$. \\subsection{Sources of stochastic backgrounds} \\label{subs:sources} The stochastic GW background can be divided into two broad classes, based on its origin: (i) the {\\it primordial GW background} (PGB), produced by physical processes in the early Universe, and (ii) the {\\it astrophysically-generated GW background} (GGB), generated by the incoherent superposition of gravitational radiation produced, at much later cosmic times, by a large number of astrophysical sources that can not be resolved individually. The emphasis of this paper is on the detectability of the PGB. In this paper we will use the following conventions: $\\Omega_p(f)$ and $\\Omega_g(f)$ identify the fractional energy density in GW's, Eq.~(\\ref{omegagw}), carried by the primordial and the generated component of the GW background, respectively. If no index is used, we refer to a general GW stochastic signal, with no assumption about its production mechanism. At present, there are three observational constraints on the PGB contribution to $\\Omega (f)$: \\begin{enumerate} \\item The high degree of isotropy of the cosmic microwave background radiation sets a limit at ultra-low frequencies~\\cite{Bennett96}: \\be h_{100}^2\\Omega_p(f) < 7\\times 10^{-11}\\,\\left(\\frac{f}{H_0}\\right)^{-2} \\quad\\,,\\quad 3\\times 10^{-18}\\,h_{100}\\,{\\rm Hz} \\simlt f \\simlt 10^{-16}\\,h_{100}\\,{\\rm Hz}\\,; \\ee \\item The very accurate timing of milli-second radio-pulsars constrains $\\Omega_p(f)$ in a frequency range of the order of the inverse of the observation time, typically of order of a few years~\\cite{KTR94}: \\be h_{100}^2\\Omega_p(f) < 10^{-8} \\quad\\quad f\\sim 10^{-8}\\,{\\rm Hz}\\,; \\ee \\item The standard model of big-bang nucleosynthesis constrains the total energy content in GWs over a wide frequency range~\\cite{KT90}: \\be \\int_{f = 10^{-8}\\,{\\rm Hz}}^{\\infty}\\,h_{100}^2\\Omega_p(f)\\,d(\\ln f) < 6\\times 10^{-6}\\,. \\ee \\end{enumerate} To foresee what physical processes could have produced a detectable GW background is an almost impossible challenge; nonetheless, it is enlighting to discuss some general principles and possible generation mechanisms to show the typical sensitivity that experiments should achieve in order to test different models. The main mechanisms that produce a PGB can be divided into two broad categories (for a recent detailed review see~\\cite{Maggiore99}): (i) Parametric amplifications of metric tensor perturbations that occurred during an inflationary epoch, and (ii) Some causal processes -- such as phase transitions -- that took place in the early Universe. Stochastic backgrounds produced by the parametric amplification of metric tensor perturbations that occurs during an inflationary epoch~\\cite{Gri93} extend over a huge range of frequencies, from $\\sim 3\\times 10^{-18}$ Hz up to a cutoff frequency in the GHz range. In the window $\\sim 10^{-16}\\,\\mbox{Hz} - 1\\,\\mbox{GHz}$, slow-roll inflationary models predict a quasi scale-invariant spectrum whose typical magnitude -- in order to satisfy the COBE bound -- cannot exceed $h_{100}^2\\,\\Omega_p\\sim 10^{-14}$ in the LISA frequency band, as well as in the Earth-based detectors observational window~\\cite{KW92}; a more refined analysis~\\cite{SRI} yields a more conservative upper limit: $h^2_{100}\\,\\Omega_p\\sim 10^{-16} - 10^{-15}$. Superstring-inspired cosmological models~\\cite{Veneziano91,GV93a,GV93b} predict a spectrum that, for suitable choices of the free parameters of the model, could reach $h_{100}^2\\Omega_p\\sim 10^{-7}$ at the frequencies accessible either to Earth-based or to space-borne experiments, while satisfying the existing observational bound~\\cite{BGGV95,BGV97,BMU97,AB97,UV00}. Stochastic backgrounds can also be produced by some classical causal processes that took place in the early Universe; for this class of signals, the characteristic frequency is related both to the time of emission and the corresponding temperature ${\\cal T}$. Non-equilibrium processes that occur at the reheating that takes place after inflation could provide a stochastic background with cutoff frequency in the range $\\sim 0.1\\,{\\rm mHz} - 1\\,{\\rm kHz}$, corresponding to reheating temperatures between $\\sim 1\\,{\\mbox TeV}$ and $\\sim 10^{9}\\,{\\mbox GeV}$. As an example, in hybrid and extended inflationary models, the exit towards a radiation-dominated era is characterized by a first-order phase transition, which, if strongly of the first order, generates a stochastic background with $h_{100}^2\\,\\Omega_p\\sim 10^{-6}$ at frequencies that can vary from the LISA observational window up to the sensitivity band of Earth-based interferometers ~\\cite{Hyb}. Phase transitions that inevitably occur at ${\\cal T}\\sim 10^2\\,{\\mbox MeV}$ (the QCD phase transition) and ${\\cal T}\\sim 10^2\\,{\\mbox GeV}$ (the electroweak phase transition) produce GWs. In particular, if the electroweak phase transition is strongly of the first order, the spectrum is approximately $h_{100}^2\\,\\Omega_p\\sim 10^{-11} - 10^{-9}$ at $f \\sim 1\\,{\\rm mHz}$~\\cite{KKT94}; the requirement of a strong first order phase transition, which is necessary in order to have baryogenesis at the electroweak scale (see~\\cite{RT99} and references therein for a recent review), is directly related, in a minimal supersymmetric extension of the standard model, to the mass of the super-partner of the top quark~\\cite{CQW96,CQW98,LR98}. Cosmic strings, which are topological defects formed during phase transitions, produce GW's whose typical frequency ranges from $f\\sim 10^{-8}$ Hz up to $f\\sim 10^{10}$ Hz with $h_{100}^2\\,\\Omega_p\\sim 10^{-9} - 10^{-8}$, see~\\cite{BCS97} and references therein for a review. Global phase transitions associated with some scalar field which acquires a non-zero vacuum expectation value (VEV) below a critical temperature would produce, via a quite general relaxation process, GW's whose energy content is very significant, $h_{100}^2\\,\\Omega_p\\sim 10^{-6}$, for VEV's near the Planck/string scale~\\cite{Krauss92,Hogan98}. Recently there has been a great amount of theoretical activity investigating higher dimensional \"brane-world\" scenarios, where gravity begins to probe the extra dimensions at energies as low as $10^3$ GeV; and an estimate of possible GW backgrounds in such models was presented recently in~\\cite{Hogan00}. These examples clearly show that the investigation of the primordial GW stochastic background in the low-frequency regime would provide us key information about the physics beyond the standard model and/or could allow us to discriminate between different inflationary cosmological models. \\subsection{Detecting a stochastic background} \\label{subs:det} A stochastic background is a random process which is intrinsically indistinguishable from the detector noise. In order to detect such a signal, the optimal signal processing strategy calls for correlations between two (or more pairs of) instruments, possibly widely separated in order to minimize the effects of {\\it common} noise sources. The relevant data analysis issues have been thoroughly addressed in~\\cite{Flanagan93,AR99}; here we simply review the main concepts, and refer to~\\cite{Flanagan93,AR99}, and references therein, for more details. The statistical analysis is based on the following assumptions: the signal and the detector noise are uncorrelated; the noise in each detector is stationary and Gaussian, and possible noise correlations between two detectors are negligible. We define the output (signal + noise) of the two instruments as $o_1(t)$ and $o_2(t)$; the cross-correlation signal $C$ that one constructs is therefore of the form: \\be C \\equiv \\int_{-T/2}^{T/2}\\,dt \\int_{-T/2}^{T/2}\\,dt'\\, o_1(t)\\, o_2(t')\\,Q(t-t')\\,, \\label{corr} \\ee where $Q(t-t')$ is a suitable filter function. In the general case, the filter function depends on $t$ {\\it and} $t'$ independently, that is $Q = Q(t,t')$; here we have used the property of the signal of being stationary, and therefore $Q(t,t') = Q(t-t')$. The SNR is defined as: \\be {\\rm SNR} = \\frac{\\mu}{\\sigma}\\,, \\label{snr} \\ee where $\\mu$ and $\\sigma$ are the mean value and the variance of the observable $C$: \\be \\mu \\equiv \\m C\\M = T\\,\\left(\\frac{3 H_0^2}{20\\pi^2}\\right)\\,\\left(\\tilde Q,\\tilde A\\right)\\,, \\label{mu} \\ee \\be \\sigma^2 \\equiv \\m C^2\\M - \\m C\\M^2 = \\frac{T}{4}\\,\\left(\\tilde Q,\\tilde Q\\right)\\,. \\label{sigma} \\ee Eqs.~(\\ref{mu}) and~(\\ref{sigma}) are written in terms of the usual {\\it inner product}~\\cite{AR99} \\be (a,b) \\equiv \\int_{\\rm -\\infty}^{+\\infty} \\,df\\,\\tilde a^*(f) \\tilde b(f) R(f)\\,, \\label{ip} \\ee where $\\tilde Q(f)$ is the Fourier transform of $Q(t-t')$. The functions $R(f)$ and $\\tilde A(f)$ are defined as follows: \\ba R(f) & \\equiv & S_n^{(1)}(f) S_n^{(2)}(f)\\times \\nonumber \\\\ &&\\left\\{ 1 + \\left(\\frac{3 H_0^2}{10\\pi^2}\\right)\\, \\frac{\\Omega(f)}{f^3}\\,\\left[\\frac{S_n^{(1)}(f) + S_n^{(2)}(f)}{S_n^{(1)}(f) S_n^{(2)}(f)} \\right]+ \\left(\\frac{3 H_0^2}{10\\pi^2}\\right)^2 \\frac{\\Omega^2(f)\\,\\left[1 + \\gamma(f)^2\\right]}{f^6\\,S_n^{(1)}(f)S_n^{(2)}(f)}\\right\\}\\,, \\label{Rf} \\ea \\be \\tilde A(f) \\equiv \\frac{\\gamma(f) \\Omega(f)}{f^3 R(f)}\\,. \\label{Af} \\ee In Eq.~(\\ref{Rf}), $S_n^{(k)}(f)\\,,k=1,2$ is the one-sided noise power spectral density of the $k-$th detector, and $\\gamma(f)$ is the so-called {\\it overlap reduction function}, which depends entirely on the relative orientation and location of the two detectors; it accounts for SNR losses that occur when the instruments are not optimally located and oriented, cf. Eq.~(\\ref{snr_w}) and Sec.~\\ref{sec:overlap}. Using Eqs.~(\\ref{mu}) and~(\\ref{sigma}), one can cast Eq.~(\\ref{snr}) in the form: \\be {\\rm SNR}^2 = T\\,\\left(\\frac{3 H_0^2}{10\\pi^2}\\right)^2\\, \\frac{\\left(\\tilde Q,\\tilde A\\right)^2}{\\left(\\tilde Q,\\tilde Q\\right)}\\,. \\label{snr1} \\ee The optimal choice of the filter $\\tilde Q$, is thus based on the maximizing the SNR, Eq.~(\\ref{snr1}), and is given by: \\be \\tilde Q(f) = ({\\rm const.}) \\times \\tilde A(f)\\,, \\label{Qf} \\ee where the overall normalization factor is arbitrary. Note that Eqs.~(\\ref{corr})-(\\ref{Qf}) are valid for a background of arbitrary energy density $\\Omega(f)$. In the case of a signal much weaker than the noise, $H_0^2\\Omega(f)/f^3 \\ll S_n^{(k)}(f)$, one can Taylor expand Eqs.~(\\ref{Rf}) and~(\\ref{Af}), retaining only the leading order term. As a consequence, Eq.~(\\ref{snr1}) reduces to: \\be {\\rm SNR} \\simeq \\frac{3 H_0^2}{\\sqrt{50}\\pi^2}\\,T^{1/2}\\, \\left[\\int_{0}^{\\infty}\\,df\\, \\frac{\\gamma(f)^2\\Omega^2(f)}{f^6\\,S_n^{(1)}(f)S_n^{(2)}(f)}\\right]^{1/2} \\quad\\quad ({\\rm signal} \\ll {\\rm noise} )\\,. \\label{snr_w} \\ee It is convenient to introduce the noise characteristic amplitude $\\hrms$, equivalent to $\\hc$, as follows: \\ba \\m n^2(t)\\M & = & \\int_0^\\infty\\, df S_n(f) \\nonumber\\\\ & = & 2 \\int_0^\\infty\\, d(\\ln f) \\hrms^2(f) \\,. \\label{nav} \\ea It is enlightening to write Eq.~(\\ref{snr_w}), using Eqs.~(\\ref{hc}) and~(\\ref{nav}), in the form: \\be {\\rm SNR} \\sim \\gamma(f_c) \\left(\\Delta f\\,T\\right)^{1/2}\\, \\left[\\frac{\\hc(f_c)}{\\hrms(f_c)}\\right]^2\\,; \\label{snr_appr} \\ee here we have assumed that the frequency band $\\Delta f$, which contains most of the SNR, is centered on the characteristic frequency $f_c$, and is sufficiently small; the noise spectral density of the two instruments, that for simplicity we assume identical, and the overlap reduction function can be therefore treated as roughly constant. If only one instrument is in operation, one could in principle detect a stochastic background with SNR$\\simgt 1$ when $\\hc \\simgt \\hrms$; with two instruments one can detect the signal when $\\hc \\simgt \\hrms/[\\gamma(f_c)\\,(\\Delta f\\,T)^{1/4}]$. Cross-correlation experiments are therefore highly desirable for both detection confidence and sensitivity. In fact, one can isolate the stochastic GW signal from all the spurious contributions which are uncorrelated between the two instruments. Common noise sources, which correlate on the same light-travel time scale, might, however, be present, degrading the overall sensitivity. Moreover, GW signals are expected to be very weak, well buried into the noise; using cross-correlations, through optimal filtering one increases the sensitivity by a factor $\\sim 10$ $(\\Delta f/1\\,{\\rm mHz})^{1/4}$ $(T/10^7\\,{\\rm sec})^{1/4}$, with respect to the single detector case. \\subsection{Summary of the results} \\label{subs:res} The goal of this paper is to explore the capability of space-borne laser interferometers, such as LISA and its successors, in searching for the primordial GW stochastic backgrounds.", "conclusions": "\\label{sec:concl} Gravitational wave experiments in the low-frequency window, together with high-frequency ground-based interferometers, are expected to improve our picture of the very early Universe, and the understanding of the behavior of fundamental fields at high energy, by detecting, or setting stringent upper limits on the primordial background of gravitational radiation. In this paper we have analyzed the sensitivity of space-borne laser interferometers of the LISA class, and possible succeeding missions. In order to set a reference frame for this discussion, we have regarded the detection of a GW background produced during the early Universe of energy $\\Om_p \\sim 10^{-16}$, consistent with the prediction of standard slow-roll inflation, as the goal of GW cosmology. We have assumed the operation of two space-detectors, in order to achieve the best sensitivity and detection confidence, and we have shown that the technology available for LISA already ensures the detection of a GW background as weak as $\\Om \\approx 5\\times 10^{-14}$. However, the strong stochastic signal in the mHz band due to short-period solar-mass binary systems that can not be resolved as individual sources prevents us from detecting a primordial background weaker than $\\Om_p \\approx 5\\times 10^{-13}$. Astrophysically generated stochastic backgrounds therefore set a fundamental limit in the mHz band that prevents us from achieving a sensitivity that goes beyond what is already guaranteed by the LISA technology. They also represent a guaranteed strong signal detectable at high signal-to-noise ratio, which enables the study of the distribution and merger rate of populations of binary compact objects in the Universe. Dedicated missions with optimal sensitivity in the window 0.1 Hz -- 1 Hz appear, at present, the only viable option in the search for very weak primordial backgrounds, and we have briefly discussed the technological challenges involved in probing slow-roll inflation. Our order-of-magnitude analysis strengthens the hope that a sensitivity level $\\Om_p \\sim 10^{-16}$ might be within the capability of future dedicated low-frequency detectors. Our analysis clearly indicates the key issues that deserve further investigation: a solid estimate of galactic and extra-galactic GW backgrounds produced by astrophysical sources, the investigation of the statistical issues that can lead to the discrimination of the PGB from the GGB, and a more rigorous analysis of the technical and conceptual problems for low-frequency experiments dedicated to GW cosmology. On the observational side, the presently designed single-instrument LISA mission is a fundamental step for the planning of more ambitious, multi-detector experiments: we will be able to measure directly the degree of anisotropy of the generated background, shedding light on the fundamental limiting factor of mHz experiments. In fact, while the present paper deals only with the detection of an isotropic stochastic signal, the remarkable sensitivity of LISA offers the chance of going far beyond: a detailed study of the anisotropy and angular dependence of stochastic signals, both of astrophysical and primordial origin. Such an investigation is currently in progress, and will be reported in a separate publication~\\cite{isotropy}." }, "0003/astro-ph0003389_arXiv.txt": { "abstract": "A new solution method is presented for steady-state, momentum-conserving, non-axisymmetric bow shocks and colliding winds in the thin-shell limit. This is a generalization of previous formulations to include a density gradient in the pre-shock ambient medium, as well as anisotropy in the pre-shock wind. For cases where the wind is unaccelerated, the formalism yields exact, analytic solutions. Solutions are presented for two bow shock cases: (1) that due to a star moving supersonically with respect to an ambient medium with a density gradient perpendicular to the stellar velocity, and (2) that due to a star with a misaligned, axisymmetric wind moving in a uniform medium. It is also shown under quite general circumstances that the total rate of energy thermalization in the bow shock is independent of the details of the wind asymmetry, including the orientation of the non-axisymmetric driving wind, provided the wind is non-accelerating and point-symmetric. A typical feature of the solutions is that the region near the standoff point is tilted, so that the star does not lie along the bisector of a parabolic fit to the standoff region. The principal use of this work is to infer the origin of bow shock asymmetries, whether due to the wind or ambient medium, or both. ", "introduction": "Supersonic stellar winds shock the surrounding gas and drive expanding bubbles into the interstellar medium. These shocks provide an opportunity to probe the properties of both the driving stellar wind and the ambient medium. If the star is moving with respect to the interstellar gas, the bubble will be distorted into a cometary shape. When the stellar motion is supersonic, we refer to these as stellar wind bow shocks (Baranov, Krasnobaev \\& Kulikovskii 1971; Dyson 1975). Since the discovery of such bow shocks around young B stars (Van Buren \\& McCray 1988), bow shocks have been found associated with many classes of objects, such as pulsars (Kulkarni et al.~1992) and cataclysmic variables (e.g. Vela X-1: Kaper et al.~1997); examples include well-known naked-eye stars (e.g. Betelgeuse: Noriega-Crespo et al.~1997). Bow shocks have been proposed as an explanation for cometary, ultracompact HII regions (Van Buren et al.~1990; Mac Low et al.~1991) and as a means of explaining the lifetimes of ultracompact HII regions. In a recent survey of the IRAS database using HiRes processing, Van Buren, Noriega-Crespo, \\& Dgani (1995) found 58 candidate bow shocks. Non-axisymmetric stellar wind bow shocks occur when a star with an anisotropic wind moves supersonically with respect to the local medium, or if the star has an isotropic wind but moves in an ambient medium containing a transverse density gradient. Models of non-axisymmetric bow shocks are relevant to cometary ultracompact HII regions due to wind-blowing O stars moving supersonically with respect to the surrounding molecular cloud, when the ambient material does not have a constant density. A non-axisymmetric bow shock has also been invoked to explain the morphology of Kepler's supernova remnant, where the supernova ejecta collide with a non-axisymmetric bow shock generated by the pre-supernova wind (Bandiera 1987; Borkowski, Blondin \\& Sarazin 1992). Another example is the bow shock due to the head of a jet propagating into a region with a density gradient. Non-axisymmetric, ram-pressure balance models of the collision between a stellar wind and the photoevaporating flow from an externally illuminated circumstellar disk have been given by Henney et al.~(1996). A formulation for steady-state non-axisymmetric bow shocks and colliding winds was given by Bandiera (1993). However, Bandiera's numerical method is sufficiently complicated that a simpler, analytic method is desirable. In this contribution, I present a method for solving the problem of steady-state, momentum-conserving, non-axisymmetric, thin-shell bow shocks and colliding winds. This is an extension of the previous analytic solution method of Wilkin (1996, hereafter Paper I) and of Cant\\'o, Raga, \\& Wilkin (1996, hereafter Paper II) to non-axisymmetric problems (see also Wilkin 1997a). An outline of the paper is as follows. In \\S~2, we formulate the problem of the steady-state collision of two winds, and in \\S~3 we treat the problem of a bow shock resulting from an isotropic wind interacting with a plane-parallel flow containing a transverse density gradient. In \\S~4, we allow for non-isotropic winds, especially an axisymmetric wind with random orientation of the symmetry axis with respect to the direction of stellar motion. The rate at which kinetic energy is thermalized for the bow shock is treated in \\S~5. Results and future directions of this research are summarized in \\S~6. ", "conclusions": "I have shown how to solve the problem of non-axisymmetric bow shocks and wind collisions with a simple formalism that requires only algebraic equations. Most often one considers constant wind speed for such problems, and in this case the method leads to exact, analytic (although possibly implicit) solutions. The availablility of simple analytic solutions makes it much easier to model observed sources and derive the properties of the driving winds. Among the principal applications of these solutions would be to determine the cause of the asymmetry in observed bow shocks, whether it be due to the ambient medium or an anisotropic wind, or both. Future improvements are needed to include non-radial and accelerating winds and shearing motions in the shell, in which case the fluid elements are not restricted to a plane. Also, non-axisymmetric bow shocks due to colliding winds in binary systems, including the orbital motion, require a formulation in a non-inertial frame. A forthcoming paper will describe how to solve the general problem of colliding winds from two stars, including both anisotropy and acceleration in the winds." }, "0003/astro-ph0003340_arXiv.txt": { "abstract": " ", "introduction": " ", "conclusions": "We have presented a theoretical formalism to predict the two-point clustering statistics on a light-cone in the cosmological redshift space. The present methodology will find two completely different applications. For relatively shallower catalogues like galaxy samples, the evolution of bias is not supposed to be so strong. Thus, one may estimate the cosmological parameters from the observed degree of the redshift distortion, as has been conducted conventionally. Most importantly, one can now correct for the systematics due to the light-cone and geometrical distortion effects, which affect the estimate of the parameters by $\\sim 10$\\%. Alternatively, for deeper catalogues like high-redshift quasar samples, one can extract information on the nonlinearity, scale-dependence and stochasticity of the object-dependent bias only by correcting the observed data on the basis of our formulae. In this case, although one should adopt a set of cosmological parameters a priori, those will be provided both from the low-redshift analysis described above and from precision data of the cosmic microwave background and supernovae Ia. In a sense, the former approach uses the light-cone and geometrical distortion effects as real cosmological signals, while the latter regards them as inevitable, but physically removable, noise. In both cases, the present methodology is essential in properly interpreting the observations of the universe at high redshifts. \\vspace{2mm} I thank Y.P.Jing, Tetsu Kitayama, Hiromitsu Magira, Takahiko Matsubara, Hiroaki Nishioka, and Kazuhiro Yamamoto for enjoyable collaborations on which the present talk is based. Numerical computations were carried out on VPP300/16R and VX/4R at the Astronomical Data Analysis Center of the National Astronomical Observatory, Japan, as well as at RESCEU and KEK (National Laboratory for High Energy Physics, Japan). This research was supported in part by the Grants-in-Aid by the Ministry of Education, Science, Sports and Culture of Japan to RESCEU (07CE2002)." }, "0003/astro-ph0003176_arXiv.txt": { "abstract": "Many lines of evidence suggest that nonbaryonic dark matter constitutes $\\sim$30\\% of the critical closure density, but the composition of this dark matter is unknown. One class of candidates for the dark matter is compact objects formed in the early universe, with typical masses $M\\sim 0.1-1\\,M_\\odot$ to correspond to the mass scale of objects found with microlensing observing projects. Specific candidates of this type include black holes formed at the epoch of the QCD phase transition, quark stars, and boson stars. Here we show that accretion onto these objects produces substantial ionization in the early universe, with an optical depth to Thomson scattering out to $z\\sim 1100$ of $\\tau\\approx 2-4 [f_{\\rm CO}\\epsilon_{-1}(M/M_\\odot)]^{1/2} (H_0/65)^{-1}$, where $\\epsilon_{-1}$ is the accretion efficiency $\\epsilon\\equiv L/{\\dot M}c^2$ divided by 0.1 and $f_{\\rm CO}$ is the fraction of matter in the compact objects. The current upper limit to the scattering optical depth, based on the anisotropy of the microwave background, is $\\approx 0.4$. Therefore, if accretion onto these objects is relatively efficient, they cannot be the main component of nonbaryonic dark matter. ", "introduction": "Observations of the rotation curves of galaxies and clusters, in addition to joint fits of Type~Ia supernova data and the power spectrum of the cosmic microwave background, suggest that the density of matter in the current universe is $\\sim$30\\% of the closure density, i.e., $\\Omega_m\\sim 0.3$. However, the success of big bang nucleosynthesis in explaining the primordial abundances of light elements, especially the primordial abundance ratio of D/H, requires that the contribution of baryons is only $\\Omega_b h^2=0.019\\pm 0.0024$ (95\\% confidence; Tytler et al.\\ 2000), where $h\\equiv H_0/100$~km~s$^{-1}$~Mpc$^{-1}$ and $H_0$ is the present day Hubble constant. The majority of the matter must be something else. One class of possibilities involves hypothesized exotic particles, from light particles such as axions (Peccei \\& Quinn 1977) to heavier particles such as the neutralino (e.g., Jungman, Kamionkowski, \\& Greist 1996) or even ultramassive particles such as ``WIMPZILLAs\" (Kolb, Chung, \\& Riotto 1998; Hui \\& Stewart 1999). Another class, which we focus on in this paper, involves dark matter that occurs primarily in $\\sim 0.1-1\\,M_\\odot$ clumps. This class, which has received recent attention because this is the mass scale of objects discovered by microlensing projects such as MACHO, EROS, and OGLE, has several specific candidates. For example, black holes may have formed during the QCD phase transition from quark matter to nucleonic matter (Jedamzik 1997, 1998; Niemeyer \\& Jedamzik 1999), during which the horizon mass was plausibly in the $0.1-1\\,M_\\odot$ range. Other suggestions involve quark stars (Banerjee et al.\\ 2000), boson stars (Colpi, Shapiro, \\& Wasserman 1986; Mielke \\& Schunck 2000), and stars formed of mirror matter (Mohapatra \\& Teplitz 1999). Here we consider those members of this class that involve primordial compact objects, specifically those objects which (1)~existed before the $z\\sim 1100$ epoch of decoupling, and (2)~have a mass to radius ratio of $GM/Rc^2\\gta 0.1$. These include black holes, quark stars, and boson stars, but not mirror matter stars, as they are envisioned to form at comparatively late times and to be comparable to ordinary stars in their compactness (Mohapatra \\& Teplitz 1999). Primordial compact objects will accrete from the ambient medium and will therefore generate substantial luminosity. This luminosity can ionize the surrounding medium. Unlike the energy spectra from ordinary stars, which drop off rapidly above the ground state ionization energy of hydrogen, the energy spectra from accreting compact objects are known observationally to be very hard, with substantial components above 1~keV and often extending above 100~keV. An important consequence of this is that whereas the Stromgren sphere of ionization around, say, an O or B star is extremely sharply defined, with an exponentially decreasing ionization fraction outside the critical radius, the ionization fraction produced by an accreting compact object dies off relatively slowly with radius, as $r^{-3/2}$ (Silk 1971; Carr 1981). Therefore, accretion onto a primordial object can produce ionization over a large volume in the early universe. If the resulting optical depth to Thomson scattering is too large, it will conflict with the upper limit to this optical depth derived from the observed anisotropy of the microwave background (Griffiths, Barbosa, \\& Liddle 1999). Conversely, the upper limit on the optical depth can be used to constrain the properties of primordial compact objects, if these are proposed as the dominant component of dark matter. Here we calculate the ionization produced by compact objects accreting in the early universe. We find that the ionization produced by secondary electrons, an effect not included in previous analyses of reionization by accretion, increases substantially the ionization fraction and hence the optical depth to Thomson scattering. In \\S~2 we show that the Thomson optical depth out to the $z\\approx 1100$ redshift of decoupling is $\\tau\\approx 2-4 [f_{\\rm CO}\\epsilon_{-1}(M/M_\\odot)]^{1/2} (H_0/65)^{-1}$, where $\\epsilon_{-1}$ is the accretion efficiency $L/{\\dot M}c^2$ divided by 0.1 and $f_{\\rm CO}$ is the fraction of matter in primordial compact objects. We compare this result to the current observational upper limit of $\\tau<0.4$, and show that either low accretion efficiency or low mass is required if dark matter is mostly composed of primordial compact objects. In \\S~3 we consider low-efficiency accretion such as flows dominated by advection or wind outflow. We show that the constraints from ionization are especially tight on objects without horizons. In \\S~4 we place this result in the context of previous constraints on, for example, primordial black holes as the main component of dark matter. We also discuss future improvements to our result. In particular, we show that the expected accuracy of optical depth measurements with MAP and Planck could decrease the upper bound on $f_{\\rm CO}\\epsilon_{-1} (M/M_\\odot)$ by a further factor of $\\sim$100. ", "conclusions": "Consideration of compact objects as components of dark matter has often been restricted to black holes, but many of the arguments apply more generally. Black holes with masses in excess of $\\sim 10^3\\,M_\\odot$ are ruled out as a significant component of galactic halos because their dynamical interactions with globular clusters would destroy the clusters (for a recent calculation see Arras \\& Wasserman 1999). The lack of an increase in the number of low equivalent width quasars with increasing redshift (expected to be caused by gravitational lensing) rules out a contribution $\\Omega\\gta 0.1$ from any objects with masses between $\\sim 10^{-2}\\,M_\\odot$ and $20\\,M_\\odot$ that are more compact than their Einstein radii (Dalcanton et al.\\ 1994). The lack of observed lensing of cosmological gamma-ray bursts also allows weak limits to be placed on the contribution of black holes of various sizes: $\\Omega<0.15$ at the 90\\% level for $M=10^{6.5}\\,M_\\odot$, $\\Omega<0.9$ at the $1\\,\\sigma$ level for $M=10^{-12.5}-10^{-9}\\,M_\\odot$, and $\\Omega<0.1$ ($z_{\\rm GRB}\\sim 1$) or $\\Omega<0.2$ ($z_{\\rm GRB}\\sim 2$) at the 95\\% level for $M=10^{-16}-10^{-13}\\,M_\\odot$ (Marani et al.\\ 1999). Here we show that ionization from compact object accretion in the early post-decoupling universe is more significant than had been thought previously, because of the effects of secondary ionization by electrons. The result is that, barring inefficient accretion ($\\epsilon<0.05$ for $M=0.1\\,M_\\odot$, $\\epsilon<0.005$ for $M=1\\,M_\\odot$), primordial compact objects in this mass range cannot compose a significant fraction of the mass of the universe, because they would ionize the universe enough to conflict with the measured small-scale anisotropies of the cosmic microwave background. If further analysis and numerical simulation of flows onto black holes demonstrates that the long-term time averaged accretion efficiency is $\\gta 0.1$, as might happen if matter tends to pile up as in a dwarf nova and then accrete quickly with efficient radiation, then all masses greater than $\\sim 0.1\\,M_\\odot$ are excluded from making a significant contribution. Future CMB missions such as MAP and Planck could strengthen these constraints considerably. The optical depth resolution of MAP is expected to be 0.022, and of Planck is expected to be 0.004 (Zaldarriaga, Spergel, \\& Seljak 1997; Bouchet, Prunet, \\& Sethi 1999; Eisenstein, Hu, \\& Tegmark 1999). Since the existence of a Ly$\\alpha$ emitter at $z=5.64$ (see Haiman \\& Spaans 1999) shows that reionization must have occurred before then, this means that both satellites, and especially Planck, will be able to detect the effects of ionization regardless of the actual redshift of reionization. If $z_{\\rm reion}\\sim 10$ then the redshift of reionization could even be determined directly with SIRTF or NGST via, e.g., analysis of the damping wing of the Gunn-Peterson trough (Miralda-Escud\\'e 1998) or detection of transmitted flux between Lyman resonances (Haiman \\& Loeb 1999). The upper limit on the product $\\epsilon_{-1}(M/M_\\odot)f_{\\rm CO}$ scales like $\\tau_{\\rm scatt}^2$ (or, for $z\\gg 1$, like $z_{\\rm reion}^3$), so if $z_{\\rm reion}\\sim 10$ this upper limit is decreased by almost a factor of 100. In this case, barring extremely inefficient accretion, dark matter must be composed of less compact objects or of WIMPs." }, "0003/astro-ph0003030_arXiv.txt": { "abstract": "The wide-band energy spectrum of NGC~6240 over the range 0.5--200~keV is investigated using the {\\it RXTE} and {\\it ASCA} data. The {\\it RXTE} data provide the spectrum beyond the {\\it ASCA} range (0.5--10 keV) with significant detection of signals up to 20~keV and the upper limits above 20~keV. The spectrum above 10~keV is found to be very flat. A strong iron-K emission line discovered in the previous {\\it ASCA} observation is also confirmed with the {\\it RXTE} PCA. These results provide further evidence for the dominance of a reflection component, i.e. emission from cool material illuminated by an AGN. By fitting the spectra obtained with {\\it RXTE} and {\\it ASCA} simultaneously, we satisfactorily modeled the AGN spectrum with a Compton reflection component and probably a transmitted AGN component penetrating through a thick absorber. The X-ray luminosity of the AGN is estimated to be in the range $4 \\times 10^{43} - 6 \\times 10^{44}$ ergs/s in the range 2--10~keV, which categorizes NGC~6240 among the most luminous Seyfert nuclei. The ratio of the 2--10~keV X-ray luminosity to the infrared luminosity, $L_{\\rm X}{\\rm (2-10~keV)}/L_{\\rm IR}$, is 0.01 -- 0.1, which implies a quite substantial, if not dominant, contribution of AGN to the infrared luminosity. ", "introduction": "The IRAS survey (Neugebauer et al. 1984) discovered many ultraluminous infrared galaxies (ULIRGs) that emit the bulk of their energy in infrared (IR) photons. Since their bolometric luminosity and the number density are as high as those of quasars, ULIRGs are among the most energetic objects in the universe. The most fundamental problem yet to be solved is the energy source of the extremely intense infrared emission. NGC~6240, a gravitationally interacting system with a complex optical morphology (Fosbury \\& Wall 1979; Fried \\& Schulz 1983), is a very interesting example of ULIRG. Its bolometric luminosity reaches $2.4\\times10^{12}$ $L_{\\odot}$ (Weight, Joseph, \\& Meikle 1984; z=0.0245 and $H_0=50$ km s$^{-1}$Mpc$^{-1}$ are assumed). NGC~6240 is outstanding in several respects. Its H$_2$ $1\\rightarrow0S(1)$ at 2.121$\\mu$m and [FeII] 1.644$\\mu$m line luminosities and the ratio of H$_2$ to bolometric luminosities are the largest currently known (e.g., van der Werf et al. 1993). Further, its stellar velocity dispersion of 360 km/s is among the highest values ever found in a galaxy centre (e.g., Doyon et al. 1994). The energy source of the huge IR luminosity is controversial. Many IR spectroscopic studies (e.g. Genzel et al. 1998; Ridgway et al. 1994; Rieke et al. 1985; Weight et al. 1984) have suggested that main energy source of the IR emission is starburst activity, which is presumably a super-starburst induced by a merger of two galaxies (Joseph \\& Wright 1985; Chevalier \\& Clegg 1985). The ground-based optical spectrum can be classified as LINER, and is interpreted as a result of shock heating (Heckman et al. 1987). On the other hand, a significant contribution from an active galactic nucleus (AGN) similar to Seyfert galaxies was also discovered from IR spectroscopy (DePoy et al. 1986). Another hint of an AGN in NGC~6240 is the presence of compact bright radio cores (Carral et al. 1990 but see Colbert et al. 1994). {\\sl HST} discovered a core that is excited higher than LINER (Rafanelli et al. 1997). X-ray observation provides an important tool for investigating both the starburst and AGN activity. The {\\it ROSAT} (Tr\\\"{u}mper 1990) observations showed that NGC~6240 is fairly bright in the soft X-ray band below 2~keV with a luminosity larger than $5\\times10^{42}$ ergs s$^{-1}$ (0.1-2.0~keV), and the bulk of the X-ray emission is extended in a scale of $\\sim 25''$ (Schulz et al. 1998; Komossa, Schulz, \\& Greiner 1998; Iwasawa \\& Comastri 1998). Detailed spectroscopic studies with {\\it ASCA} (Tanaka, Inoue, \\& Holt 1994) data of NGC~6240 have shown that the soft X-ray spectrum can be explained with two thermal components, a cooler component with a temperature of 0.2--0.6~keV and a hotter component of $\\sim1$~keV with an excess absorption of $\\sim10^{22}$ cm$^{-2}$ (Iwasawa \\& Comastri 1998). These results show that these soft X-rays are most likely originated from thermal processes, which may arise from starburst activities (e.g. Heckman et al. 1987). On the other hand, in the 3--10~keV band, another very hard continuum with a strong iron-K emission feature was observed with {\\it ASCA} (Mitsuda 1995; Iwasawa \\& Comastri 1998). The observed spectral feature can be accounted for in terms of Compton reflection from optically thick material (e.g. Lightman \\& White 1988; George, Nandra \\& Fabian 1990), and is generally accepted as evidence for the presence of an AGN in NGC~6240. The emission-line profile further indicates that a part of the reflector is highly ionized (Mitsuda 1995; Iwasawa \\& Comastri 1998), and both the {\\it ROSAT} and {\\it ASCA} spectra were modeled with a reflection component from warm material surrounding the AGN (Netzer, Turner, \\& George 1998; Komossa et al. 1998). As described above, many observational facts of NGC~6240, in particular X-ray spectroscopy, supports the presence of an AGN in NGC~6240, which may account for a significant fraction of the huge IR luminosity. However, the intrinsic power of the AGN is still uncertain, The hidden AGN could be visible as a strongly absorbed X-ray continuum above 10~keV, penetrating through a thick layer of the obscuring matter. Very recently, Vignati et al. (1999) has published the {\\it BeppoSAX} results of NGC~6240, concluding the detection of the direct X-rays from an AGN. As described in this paper, we performed an independent study of NGC~6240 with {\\it RXTE} which carries the PCA (Jahoda et al. 1996) and HEXTE (Rothschild et al. 1998) covering 2--250~keV. In the spectral analysis, we also utilized the {\\it ASCA} data covering 0.5--10~keV energy band with two Gas Imaging Spectrometers (GIS: Ohashi et al. 1996) and two X-ray CCD cameras (SIS: Burke et al. 1991). Simultaneous use of the {\\it ASCA} data provides more constraints in modeling the spectrum. Moreover, because {\\it RXTE} has no imaging capability, the {\\it RXTE} spectrum may be subject to contamination by nearby sources. In fact, we find from the {\\it ASCA} image that it is the case, and the {\\it ASCA} data are used to correct the {\\it RXTE} spectrum. We have examined various spectral models, and conclude the presence of a high-luminosity AGN in NGC~6240. Our results are in essential agreement with those of Vignati et al. (1999). ", "conclusions": "The 0.5--200~keV wide band energy spectrum of NGC~6240 obtained with {\\it RXTE} and {\\it ASCA} is accounted for in terms of a soft thermal component and a hard AGN component. The soft component is presumably due to star burst activity. The AGN component consists of a Compton reflection component accompanied by an intense Fe-K emission line and probably a transmitted component (a direct component penetrating through a thick absorber). The detection of the transmitted component is not conclusive from the fitting. However, without a transmitted component, the photon index is unusually small. Assuming that the solid angle factor $R$ does not exceed 1, we estimated the intrinsic X-ray luminosity of the AGN in the range 2--10~keV to be in the range $4 \\times 10^{43}-6\\times10^{44}$ ergs/s, which yields the ratio of the X-ray luminosity (2--10~keV) to the IR luminosity of $0.01-0.1$. The column density that obscures the central AGN is estimated to be larger than $1.0\\times 10^{24}$ cm$^{-2}$. These results show that NGC~6240 is among the most luminous Seyfert 2 galaxies." }, "0003/astro-ph0003426_arXiv.txt": { "abstract": "We study the spin evolution of X-ray neutron stars in binary systems, which are being spun up by mass transfer from accretion disks. Our investigation reveals that a quark phase transition resulting from the changing central density induced by the changing spin, can lead to a pronounced peak in the frequency distribution of X-ray neutron stars. This finding provides one of several possible explanations available in the literature, or at least a contributor to part of the observed anomalous frequency distribution of neutron stars in low-mass X-ray binaries (LMXBs), which lie in a narrow band centered at about 300 Hz, as found by the Rossi Explorer (RXTE). ", "introduction": "\\label{sec:deco} The density and pressure in the interior of neutron stars is high in comparison with nuclear density by a factor of some 5 to 10 depending on the particular models used to estimate it. At such densities it is quite plausible that the quark constituents of hadrons loose their association with particular hadrons---the deconfined quark matter phase replaces the normal phase. In a nonrotating star, the radial boundaries between quark core, mixed phase, and normal hadronic phase would remain fixed. However in a rotating star, because of the centrifugal distortion of the density in the interior, these boundaries will change as the rotational frequency of the star changes with time. A structural change occurs in such a star with changes in frequency (\\cite{glen97:a}). If there were no change in the nature of matter, the stell\\-ar fluid would res\\-pond simply under the action of the cent\\-rifugal force. However, the compressibility of the normal nuclear matter phase and the deconfined and relatively free Fermi gas of the quark matter phase, are different. The former must be less compressible than the latter. When a ms pulsar spins down, its central density may rise above the critical phase transition density and the central core will then change phase to softer quark matter; it is compressed both by its own gravitational attraction, and by the weight of the overlying nuclear phase. The reverse will be true in spinup due to accretion. In either case, the distribution of mass, radius and moment of inertia are changed by a phase transition beyond those changes that would take place in an immutable fluid under the action of a changing centrifugal force. These ideas were applied to the spin-down of a ms pulsar (\\cite{glen97:a,glen97:e,weber99:topr}). It was found that as the quark matter core grew in radial extent, the moment of inertia decreased anomalously, and could even introduce an era of spin-up lasting for $\\sim 10^7$ years (\\cite{glen97:e}). The response of the moment of inertia to changes in spin is very like the so-called ``backbending'' in nuclei predicted by Mottelson and Valatin (\\cite{mottelson82:a}) and discovered many years ago (\\cite{johnson72:a,stephens72:a}). Accreting X-ray neutron stars provide a very interesting contrast to the spin-down of isolated ms pulsars. The X-ray stars are being spun up by the accretion of matter from a low-mass, less-dense white dwarf companion. They are presumably the link between the canonical pulsars with mean period of $0.7$ sec and the ms pulsars (\\cite{klis98:b,chakrabarty98:a}). If the critical deconfinement density falls within the range spanned by canonical pulsars, quark matter will already exist in them but may be ``spun'' out of X-ray stars as their frequency increases during accretion. We can anticipate that in a certain frequency range, the changing radial extent of the quark matter phase will actually inhibit changes in frequency because of the increase in moment of inertia occasioned by the gradual disappearance of the quark matter phase. Accreters will tend to spend a greater length of time in the critical frequencies than otherwise. There will be an anomalous number of accreters that appear at or near the same frequency. This is what was found recently with the Rossi X-ray Timing Explorer (RXTE) (\\cite{klis00:a}). Presumably, accreters commence their evolution near the death line of active canonical pulsars with frequencies of $\\nu \\sim 1$ Hz and end as ms pulsars with $\\nu \\sim 200 {\\rm~to~} 600$ Hz. The spinup evolution of an accreting star is a more complicated problem than that of the spindown of an isolated ms pulsar of constant baryon number. It is complicated by the accretion of matter ($\\dot M > 10^{-10} M_\\odot~{\\rm yr}^{-1}$), a changing magnetic field strength (from $B \\sim 10^{12} {\\rm~to~} \\sim 10^8$~G), and the interaction of the field with the accretion disk. ", "conclusions": "\\label{ref:summary} We have traced the time evolution of the moment of inertia and rotational frequency for a neutron star accreting matter from a low-mass companion, under various assumptions about the accretion rate and for two stellar models, one an ordinary neutron star populated by nucleons, hyperons and leptons, and one in which phase equilibrium between ordinary and quark deconfined matter occurs within the density range found in canonical pulsars. In the second case the computed frequency distribution of X-ray neutron stars shows a spike, much as is observed in a recent compilation of data (\\cite{klis00:a}). There are various suggestions as to the cause of the spike, several of which we cite (c.f. \\cite{bildsten,andersson,levin}). A possible contributing mechanism which causes some accreters of suitable mass to resist spinup for a lengthy era is that discussed in this paper--- the ongoing reduction of quark matter cores in the centers of neutron stars as they are spun up. This occurs because, with increasing spin, the density of the inner region is centrifugally diluted until it falls below the threshold density at which quark matter can exist, first in the center, and then in an expanding region. As explained in the introduction, the conversion of quark mattter to confined hadronic matter manifests itself in an expansion of the star and a significant increase in its moment of inertia. As a consequence, the angular momentum added to a neutron star during this phase of evolution is then consumed by the star's expansion, inhibiting a further spin-up until the quark matter has been converted into a mixed phase of matter made up of hadrons and quarks." }, "0003/astro-ph0003432_arXiv.txt": { "abstract": "From a minimum of total energy of celestial bodies, their basic parameters are obtained. The steady-state values of mass, radius, and temperature of stars and white dwarfs, as well as masses of pulsars are calculated. The luminosity and giromagnetic ratio of celestial bodies are estimated. All the obtained values are in a satisfactory agreement with observation data. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003118_arXiv.txt": { "abstract": "We present a $V$ light curve of the eclipsing binary system \\object{HV\\,2543} (\\object{Sk--67${\\degr }$117}) in the Large Magellanic Cloud based on CCD images acquired between the years 1995 and 1998. We have analysed this light curve and published radial velocity data, finding that this system is semidetached, with the secondary (less massive and less luminous component) filling its Roche lobe. From our analysis with the Wilson-Devinney code, we estimated the following masses and radii for the components of HV\\,2543: $M_1=25.63 \\pm 0.7 \\, {\\rm M}_{\\sun}$, $R_1=15.54 \\pm 0.4 \\, {\\rm R}_{\\sun}$, $M_2=15.63 \\pm 1.0 \\, {\\rm M}_{\\sun}$ and $R_2=13.99 \\pm 0.4 \\, {\\rm R}_{\\sun}$. On the basis of $B$ and $V$ photometry of the field stars, we found that \\object{HV\\,2543} is member of an OB association, perhaps related to which the massive binary system \\object{Sk--67$ \\degr $105} belongs. ", "introduction": "The analysis of light curves of eclipsing binaries, in addition to radial velocity data, provides fundamental knowledge about the masses and physical dimensions of the stars. Studying massive binaries in the \\object{Magellanic Clouds} we can learn about the evolution of these systems at metallicities lower than that of our galaxy. The Harvard variable \\object{HV\\,2543} ($\\alpha=5^{\\rm h}27^{\\rm m}27^{\\rm s}$, $\\delta=-67\\degr 11\\arcmin 54\\arcsec$, J2000) is a hot binary star in the \\object{LMC}. It was catalogued as an OB star by Sanduleak (\\cite{sk}), who assigned it the identification --67$\\degr$117. The eclipsing nature of this binary was discovered by Gaposhkin (\\cite{gap70}), who published a photographic light curve fitting a period of 4.829052 days (see also Payne-Gaposhkin \\cite {pg}). Photoelectric photometry was performed by Isserstedt (\\cite{isser}) who found $V=12.92$, $(B-V)=-0.18$ and $(U-B)=-1.03$. Radial velocity orbit was obtained by Niemela \\& Bassino (\\cite{virpi}) who derived physical parameters of the binary components and concluded that HV\\,2543 was a semidetached system with the less massive component filling its equipotential Roche surface. On the basis of their spectroscopic data, they classified this system as O8V:+O9III. Smith Neubig \\& Bruhweiler (\\cite{neubig}) have published an UV spectral classification of LMC OB stars based on IUE data, assigning to HV\\,2543 the type O9III. In this paper we present a CCD $V$ light curve for HV\\,2543. By means of the combined analysis of these data and previously published radial velocities, we derive new values of the fundamental parameters of this system. The paper is organized as follows: in Sect. 2 we describe the observations, reductions and calibration steps. In Sect. 3 we describe the photometric results and light curve fitting. In Sect. 4 we discuss the results and in Sect. 5 we present our conclusions. ", "conclusions": "We note that the radius derived for the O8 component is substantially larger than the one obtained by Niemela \\& Bassino (\\cite{virpi}). In fact, the spectroscopic data suggest that the luminosity of the O9 star is larger than that of the O8 star, which caused Niemela \\& Bassino to refer to the O9 component as ``primary'', even though it is the one presenting the largest radial velocity amplitude, and consequently the less massive component of the binary system. Hence, we explored alternative solutions to the light curve that could yield a smaller radius for the O8 component. The values $R_1 \\sim 12.75$, $R_2 \\sim 15.31$, $M_1 \\sim 21.91$ and $M_2 \\sim 19.34$, in solar units, provide a good fit to the light curve, although they require a value of $q$ near 0.9, which is not compatible with the radial velocity data. Hence we have discarded this solution, which on the other hand, implies a rather low value of the distance modulus, namely 18.18, assuming $A_V = 0.62$ (see below) and adopting the temperature scale of Chlebowski \\& Garmany (\\cite{chg}). This value would result even smaller if a lower temperature scale is adopted. On the other hand, the temperature difference between the two components of HV\\,2543 resulting from the light curve analysis is larger than that suggested by the corresponding spectral types. This fact might be due to the difference between the spectral features of the unperturbed back sides and the heated inner sides of the stars. Also should we have in mind that the mass transfer history of the system might account for significant departures of the He abundances relative to those regarded as normal for LMC members, and this effect could influence some spectral lines of the secondary component. To solve these puzzles, high dispersion spectroscopy would be desirable. The semimajor axis and star dimensions of this system are alike those determined by Pritchard et al. (\\cite{prit}) for \\object{HV\\,2241}, but the masses are somewhat smaller. We presume that these systems have experienced case A mass transfer, being now in the slow stage of mass exchange. In these cases, the mass gainer should be indistinguishable of a normal star, excepting that it would stand on an isochrone corresponding to a shorter age (see Vanbeveren et al. \\cite{vbv}). From the stellar models of Schaerer et al. (\\cite{sch}) (for single stars) we found that a $25 {\\rm M}_{\\sun}$ star takes $\\sim 6.7$ Myr until its radius grows to $15{\\rm R}_{\\sun}$, but such star would have an effective temperature of only 30000\\,K. The radius and effective temperature derived for the O8 star are consistent whit those of a single star of some 3.7 Myrs and $\\sim 40 {\\rm M}_{\\sun}$, completely out of the range of masses compatible with the radial velocity data. We note that a similar problem arises from the analysis of \\object{AB Crucis} (Lorenz et al. \\cite{lor}). It is clear that there are a series of phenomena that we do not fully understand, and consequently, our analysis can not be considered definitive. The errors given for the derived parameters must be considered with caution. True errors are not easy to estimate analytically, since they depend on the importance of phenomena that are not properly accounted for, such as wind shocks, uncertainty of the adopted temperature scale, radiation pressure effects, etc. In fact, the sizes and positions of the spots used to model empirically the O'Connell effect are rather arbitrary, and an equally satisfactory solution could be found with other parameters, but these details do not affect meaningly the derived star dimensions. From our photometry we determine $(B-V)=-0.11 \\pm 0.015$ for HV\\,2543. This value is somewhat redder than $-0.18$, obtained by Isserstedt (\\cite {isser}). However, given that Isserstedt does not detect the variability of HV\\,2543, we think that both measures are still in reasonable agreement. If we assume a $(B-V)_0$ of -0.31 (Schmidt-Kaler \\cite{s-k}), $R=3.1$ (Koornneef \\cite{koo}) then it results an $A_V=0.62$ for HV\\,2543. This value does not depend on which temperature scale we adopt, since for the range of temperatures of the O-type stars the $(B-V)$ colours are degenerated. Estimating the bolometric corrections according to Massey \\& Hunter (\\cite{mh}), we derive a distance modulus of $(m-M)_0=18.31 \\pm 0.2$ to $18.40 \\pm 0.2$, depending on the adopted temperature scale. The error accounts for the uncertainties in the estimates of $R$ and the bolometric corrections. This distance modulus must be considered with caution, since this system has experienced strong mass transfer and exhibits the above mentioned anomalies." }, "0003/astro-ph0003097_arXiv.txt": { "abstract": "We report detection of giant cyclonic vortices in the gaseous disc of the spiral galaxy NGC 3631 in the reference frame rotating with the spiral pattern. A presence of such structures was predicted by the authors for galaxies, where the radial gradient of the perturbed velocity exceeds that of the rotational velocity. This situation really takes place in NGC 3631. ", "introduction": "In previous papers \\cite{1},\\,\\cite{2} authors have shown that besides the spiral arms, well known for more than one and a half century in spiral galaxies as structures in the brightness distribution, there also exist structures, revealing themselves in the velocity field. The latter structures have the appearance of giant anticyclonic vortices: their rotation is opposite to the rotation of the galactic disc. The centers of these vortices are located near the zero points of the force field (Lagrange points $L_4$ and $L_5$, see \\cite{3}). In spiral galaxies these points are fixed in the close vicinity of the corotation circle, where the velocity of the rigid-body rotating spiral pattern coincides with the velocity of the differentially rotating disc. Being in such a position, the centers of the anticyclones turn out to be fixed with respect to the spiral arms and stationary if the spiral structure is stationary. If the spiral structure develops due to some instability in the disc, then the arms and vortices arise simultaneously and grow with the same growth rate \\cite{4}. As a result a unified spiral--vortex structure forms. The growth of the perturbation amplitude leads rather often to the saturation of the instability, after which a spiral--vortex structure becomes stationary. From the very beginning this scenario implies that spiral arms in discs of galaxies have a wave nature, that satisfies modern theoretical conceptions \\cite{5} and agrees well with the results of the velocity field analysis in spiral galaxies (\\cite{2},\\,\\cite{6}). A typical rotation velocity field of a galactic disc is represented schematically in Fig.~1a. There is no way to close the trajectories of particles in the laboratory (inertial) reference frame by any radial component of the velocity (arrows radially directed in Fig.~1a). The situation is opposite in the rotating reference frame, where anticyclonic motion takes place under condition of appropriate azimuthal variation of the radial component (see Fig.~1b). Note that in the latter case it is not necessary to have large radial velocities to form a vortex near the circle corotating with the reference frame due to small values of the azimuthal velocities in this region. It is not accidental that the radial velocity was chosen as a periodic function of the azimuth: in the spiral density wave such periodicity should take place both for the radial $\\tilde V_r$ and for the azimuthal $\\tilde V_\\phi $ perturbed velocities. Arrows for $\\tilde {V_\\phi} $ are not shown in Fig.1. Taking into account the perturbed azimuthal velocity will not change the picture qualitatively, if \\begin{equation} \\label{1a} |\\partial\\tilde{ V_\\phi}/ \\partial r| < |d V_{circ} / dr|, \\end{equation} where $V_{circ}$ $\\equiv$ $V_{rot}-\\Omega_{rf}\\,r$ is the circular velocity in the reference frame rotating with the angular velocity $\\Omega_{rf}$. Note that even a weak inequality is quite enough to form only anticyclonic vortices like in Fig.\\,1b (for more details see \\cite{6aa}). Under the opposite condition: \\begin{equation} \\label{1b} |\\partial\\tilde {V_\\phi}/\\partial r| > |d V_{circ}/dr| \\end{equation} not only anticyclonic, but also cyclonic vortices can appear \\cite{6a}. In the next section we will consider both these cases in a schematic model of a galactic disc with a well-defined two-armed spiral structure (Grand-Design galaxy). ", "conclusions": "I. An analysis of velocity fields in Grand-Design galaxies shows that: 1) the field of residual velocities contains two cyclones and two anticyclones with centers on the corotation circle; 2) in the reference frame rotating with the spiral pattern the velocity field belongs to one of two types: \\noindent a) under the condition (\\ref{1a}) it contains only two anticyclones with centers near the corotation circle; \\noindent b) under the condition (\\ref{1b}) besides the anticyclones the field contains also either two or four cyclones. In the former case the cyclone centers lay either on, or outside, or inside the corotation circle. In the latter case, two pairs of cyclones appear with centers inside and outside the corotation circle. II. Our analysis of the velocity field data obtained by our team at the 6m telescope in the Special Astrophysical Observatory of the Russian Academy of Sciences shows that the Grand-Design galaxy NGC 3631 belongs to the type b)." }, "0003/gr-qc0003063_arXiv.txt": { "abstract": "% Exact \\ expressions \\ for \\ probability densities of \\ conjugate \\ pair \\ separation \\ in \\ euclidean \\ isometries are obtained, \\ for the cosmic crys- tallography.~These~are~the~theoretical~counter- parts \\ of \\ the \\ mean histograms~arising \\ from computer simulation of the isometries. \\\\ For completeness, \\, also the isometries with fixed points are examined, \\, as well as the orientation reversing isometries. ", "introduction": "} \\label{Intro} % \\setcounter{equation}{0} Various methods have been proposed to investigate the shape of the universe, and cosmic crystallography (CC) is one of them \\cite{lelalu}. As CC ponders, if the universe is multiply connected (MC) then multiple images of a same cosmic object may be seen in the sky. The separations between these images are correlated by the geometry and the topology of the spacetime; so if one selects a catalog of $n$ observed images of various cosmic objects and performs a histogram of the $n(n-1)/2$ separations between these $n$ images, then the existing correlations must somehow show up. It was recently examined \\cite{spikesI} in what respects the histogram for a multiply connected observed universe should differ {}from that of a simply connected (SC) one, with same geometry and radius. It was found that each isometry of the MC universe individually imprints either a small localized deformity on the histogram of the SC universe \\cite{spikesI}, or a sharp spike if the isometry is a Clifford translation \\cite{spikesII}. Since each observed universe model has a unique pair \\, separation \\, normalized \\, histogram, a \\, strategy to unveil the cosmic topology seems straightforward: one should compare the histogram obtained from observational astronomy with histograms obtained from computer simulated universe models with prescribed topologies. Both types of histograms (observational and simulated) are infected with statistical noises, and methods to reduce these noises are desirable. A suggestion was made, to replace the histogram related to the SC component of the simulated model by an {\\it exact} continuous probability density function. For each geometry with constant curvature the corresponding function was then derived \\cite{cc3mpf}; however, appropriate functions were still lacking, to replace the histograms related to each isometry component \\cite{tsct}. In the present study we derive some of these functions, namely the normalized pair separation probability densities for the {\\it euclidean} \\ isometries. Following the prescriptions of ref. \\cite{spikesI}, if we now merge these new functions with the functions already given in~\\cite{cc3mpf} then we obtain noiseless normalized probability densities more suitable for comparing with the normalized observational histograms. For completeness we also examine the euclidean isometries with fixed point, as well as the orientation reversing isometries. ", "conclusions": "\\label{sec4} \\setcounter{equation}{0} \\vspace*{-0.2cm} We initially aimed to write out one single expression for the probability density $\\PgBl$ for screw motions of solid balls, valid for whatever values of the {\\it four} parameters $a, b, t$\\,, and $\\omega$. However, we soon found that such expression would demand a quite large number of step functions to account for all sort of possibilities. Since in practice the isometries are dealt with one at each time, we found more appropriate to present a simple method to have the exact $\\PgBl$ for each individual screw motion with {\\it fixed} values of the four parameters. Nevertheless, for those isometric motions of solid balls described by {\\it three} or less free parameters the exact expression for $\\PgBl$ for {\\em any} euclidean isometry is short enough and was displayed. As promised in the Introduction, we exhibited the analytic counterpart $\\PgBl$ of the computer simulations of pair separations histograms of the euclidean isometries in cosmic crystallography thus far obtained in the literature. The graph of $\\PgBl$ in figure 10 corresponds to the isometries $b$ and $c$ in the Fagundes and Gausmann~\\cite{CESU} study, or equivalently the isometries $\\beta$ and $\\delta$ in Gomero~\\cite{FPGD}. The discontinuity in $l\\sim{0.7}$ is not observed in the two upper figures 1 (Universe E4) of~\\cite{CESU} due to the strong statistical noise present in these histograms; nevertheless it is clearly seen in the position $s=l^2=0.5$ in the {\\em mean} histogram 5b of~\\cite{spikesII} as well as in the position $l\\sim{0.7}$ in the mean histograms 1a, 1b, 2b, 4a, and 5 of \\cite{tsct}. Similarly, figure 11 corresponds to the isometry $a$ of \\cite{CESU} and $\\alpha$ in~\\cite{FPGD} and \\cite{spikesII}; the discontinuity in $l=1$ has their counterparts again in the histogram 5b of~\\cite{spikesII} and in the histograms 1a, 1b, 2a, 4a, and 5 of~\\cite{tsct}. In contrast with the screw motions (figures 10 and 11), the pure rotations (figures 13, 15, and 16, all with $t=0$) do not show discontinuity of $\\PgBl$\\,. Oppositely to figure 13, where $ab$\\,, so the ball $\\B$ now has fixed points and the graph of $\\PgBl$ effectively starts from $l=0$. The strange-looking graph in figure 15 was confirmed in a computer simulation; the irregular behavior near $l=0.7$ corresponds to the narrow $r$-interval where the pair-of-discs combined intersection eq.(\\ref{3.48}) occurs. In figure 16 we have chosen values for $a, b$, and $\\omega$ such that the three types of combined intersection (ring, pair of discs, and one disc) have equal range in the $l$ scale. {\\it En passant}, the pair-of-discs $l-$range is now wide, and does not originate a bump as did in figure 15. Figure 17 corresponds to rotation of the solid ball $\\B$ around a diameter ${\\cal D}$; from (\\ref{3.51}) we find that defining $l_{max}=2a\\sin\\omega/2$ then the graph of $l_{max}\\PgBl$ against $l/l_{max}$ does not depend on $l_{max}$. The points of $\\B$ along the diameter ${\\cal D}$ are fixed under the isometry, so the graph again effectively starts from the origin. Figure 20 corresponds to a glide reflection whose sections $x=const$ in the intersection $\\BUBg$ are either bicircular discs (for small $|x|$) or circular (for larger $|x|$). The minimum displacement $l$ occurs for the points of $\\B$ in the intersection with the reflector plane $x=0$, giving $l_{min}=t$, the translation. Since for all points near the reflector plane we have $l\\sim t(1+2|x|^2)$, then these points are displaced almost the same value $l\\sim t$; as a consequence, $\\PgBl$ diverges in the vicinity of $l=t$. Nevertheless the integrated area is finite, with value 1. Between $l=t=0.25$ and $l\\sim 0.5$ we have bicircular sections (\\ref{4.6}), while for $0.510^{22} {\\rm cm^{-3}}$) had not been studied. With these points in mind, we investigate in this paper the thermal and chemical evolution of collapsing spherical clouds as a function of metallicity. The outline of this paper is as follows. In \\S 2, we describe the method of our calculations. In \\S 3, results of our calculations are presented. We summarize our work in \\S 4. ", "conclusions": "We have investigated the thermal and chemical evolution of collapsing protostellar clouds with different metallicities. The varied range of metallicity spans the local interstellar value ($\\simeq 1 Z_{\\sun}$) to zero. The evolution of the clouds is summarized as follows. While the clouds are transparent to continuous radiation, the temperature of clouds with lower metallicity is higher since their radiative cooling rates are lower for the same density and temperature. However, after the clouds become opaque and begin adiabatic contraction, their evolutionary trajectories converge to a line that is determined only by physical constants. Thereafter, the trajectories coincide with each other regardless of their metallicity. Consequently, the physical dimension of the stellar core at the time of formation is the same for clouds with any composition. We have also discussed analytically the reasons for the convergence." }, "0003/astro-ph0003448_arXiv.txt": { "abstract": "A new pulsating X-ray source, AX J183220$-$0840, with a 1549.1-s period was discovered at R.A.$= 18^{\\rm h} 32^{\\rm m} 20^{\\rm s}$ and Dec.$=-8^{\\circ}40'30''$ (J2000, uncertainty$=0\\farcm 6$) during an ASCA observation on the Galactic plane. The source was observed two times, in 1997 and in 1999. A phase-averaged X-ray flux of $1.1\\times 10^{-11}$ {\\flux} and pulsation period of $1549.1\\pm 0.4$ s were consistently obtained from these two observations. The X-ray spectrum was represented by a flat absorbed power-law with a photon-index of $\\Gamma\\simeq 0.8$ and an absorption column density of $N_{\\rm H}\\simeq 1.3\\times10^{22}$ {\\col}. Also, a signature of iron K-shell line emission with a centroid of 6.7 keV and an equivalent width of approximately 450 eV was detected. From the pulsation period and the iron-line feature, AX J183220$-$0840 is likely to be a magnetic white dwarf binary with a complexly absorbed thermal spectrum with a temperature of about 10 keV. ", "introduction": "Since many X-ray sources in the Galaxy are considered to be compact objects, the population of X-ray sources % gives us information on the evolution and endpoints of stars. Thus, the study of faint X-ray sources in the Galaxy is an important subject in X-ray astronomy. However, because of the difficulty in hard X-ray optics, imaging observations with high sensitivities have been limited to the soft X-ray band below 3 keV. Since the soft X-rays with an energy below 3 keV are substantially absorbed by the interstellar medium (ISM) in the Galaxy, our knowledge of X-ray sources has always been limited to that of the local area neighboring our solar system. Recently, this situation has begun changing with the discoveries of faint X-ray pulsars by deep observations of ASCA and SAX (e.g., Sugizaki et al. 1997; Kinugasa et al. 1998; Torii et al. 1999; Oosterbroek et al. 1999; Sakano \\& Koyama 2000). ASCA is the first X-ray astronomical satellite which allows imaging observations with a high sensitivity in the X-ray band above 3 keV \\cite{Tanaka1994}. The ASCA Galactic plane survey, aimed at the systematic study of Galactic X-ray sources, was performed from March 1997 to April 1999. During the observation in the direction of $(l,b)=(23 \\fdg 0, 0 \\fdg 0)$ in the Galactic plane survey, a new variable X-ray source, AX J183220$-$0840, % was detected in the GIS field of view % \\cite{Sugizaki1999}. As a result of analyzing source variabilities, a signature of a coherent pulsation with a 1550-s period was discovered from the light curve. Since the exposure time of 6 ks in the survey observation was too short to investigate the variabilities with the 1550-s period and the SIS data could not be utilized for a quantitative analysis because of the bad source position, we conducted a follow-up observation and succeeded in confirming the periodicity and obtaining detailed spectral properties. In this paper, we report the discovery of a new X-ray pulsation source, AX J183220$-$0840, and the results of analyzing data obtained from the Galactic plane survey and the follow-up observation. We discuss the identification of the source based on these results. ", "conclusions": "We now discuss the nature of the new X-ray pulsating source, AX J183220$-$0840, based on the results of the analysis. We first estimate the distance to the source from the X-ray absorption column density, which has a correlation with the hydrogen column density within the RMS $\\lesssim$ 50\\% \\cite{Ryter1975,Savage1977}. We obtain the absorption column density for the source, $N_{\\rm H} \\simeq 1.3\\times 10^{22}$ {\\col}, from the spectral analysis. The hydrogen column density of the Galactic ISM in the direction to the source can be estimated from the H{\\sc i} and H$_2$ densities. % The H{\\sc i} density is derived from the radio 21-cm observation \\cite{Dickey1990}, which is $N_{\\rm H_I}=1.9\\times 10^{22}$ {\\col}. The H$_2$ density is estimated from the CO-line intensity \\cite{Dame1987} and CO-to-H$_2$ conversion factor \\cite{Hunter1997}, which is $N_{\\rm H_2}=2.1\\times 10^{22}$ {\\col}. Thus, the hydrogen column density in the direction of the source accumulated over the Galaxy is evaluated to be $N_{\\rm H_I} + 2N_{\\rm H_2}=6.1\\times 10^{22}$ {\\col}. Assuming that ISM is uniformly distributed on the Galaxy plane within the 10 kpc from the Galactic center and the solar system is 8 kpc from the Galactic center, the absorption column density of $N_{\\rm H} \\simeq 1.3\\times 10^{22}$ {\\col} corresponds to a distance of $4$ kpc. Thus, the luminosity of the source is estimated to be $L_{\\rm X} = 2.0\\times 10^{34} d_{\\rm 4kpc}^2$ {\\lumi} from the unabsorbed source flux, $1.1\\times 10^{-11}$ \\flux. The 1549.1-s pulsation period of AX J183220$-$0840 is in accord with typical rotation periods of white dwarfs (WDs) observed in cataclysmic variables (CVs). The broad, double-peak pulse profile and the X-ray spectrum represented by a multi-absorption, thin-thermal plasma emission with a temperature of about 10 keV also agree with those of typical magnetic CVs (e.g., Mukai et al. 1994; Ishida et al. 1997; Ezuka \\& Ishida 1999). The estimated luminosity of $2.0\\times 10^{34} d_{\\rm 4kpc}^2$ {\\lumi} seems to be relatively high compared with typical values of the magnetic CVs. However, it can be reasonably explained if the absorption is partially intrinsic to the source and the distance is less than the 4 kpc estimated by assuming that the absorption is responsible for the ISM. Thus, a magnetic WD binary is considered to be the most plausible candidate for the X-ray origin of AX J183220$-$0840. If AX J183220$-$0840 is a binary neutron star (NS) pulsar, the pulsation period is as long as that of the slowest known X-ray pulsar, RX J0146.9$+$6121, which has a 1407.3-s period \\cite{Habel1998}. The hard X-ray spectrum represented by a power-law with an index of $\\Gamma\\simeq 0.8$ agrees with that of typical NS binaries \\cite{Nagase1989}. However, the iron line feature with a centroid of $6.7$ keV indicating that it originates from He-like iron ion, disagrees with the typical NS binaries, many of which have a 6.4 keV line from neutral iron. This argues against the hypothesis of a NS pulsar. Recently, the number of faint X-ray pulsars with long periods exceeding 400 s has increased by the discoveries of ROSAT, ASCA, and SAX \\cite{Israel1998,Torii1999,Oosterbroek1999,Sakano2000}. It is not clear yet whether some of these are NS binaries or WD binaries. No evidence of binary systems such as eclipses, modulations of the pulsation period, flux changes, or binary counter parts, has been obtained so far from these sources including AX J183220$-$0840. Since the typical X-ray luminosity of the WD binaries ($L_{\\rm X}\\sim 10^{30-32}$ {\\lumi}; e.g., Ezuka \\& Ishida 1999), is less by five orders of magnitude than that of the NS binaries ($L_{\\rm X}\\sim 10^{35.5-37.5}$ {\\lumi}; e.g., Nagase 1989), the number of the WD binaries was only about a sixth of that of the NS binaries among bright Galactic X-ray sources with a flux above $2\\times 10^{-11}$ {\\flux} \\cite{Bradt1983}. However, the population of the WD binaries is expected to be greater by three orders of magnitude than that of the NS binaries in the Galaxy \\cite{Patterson1984}. Thus, AX J183220$-$0840 is considered to be a candidate of faint WD binaries that have been undetectable due to the large distance from the solar system but have now become detectable by the sensitive observations of ROSAT and ASCA. Further X-ray observations are required to confirm the identification of AX J183220$-$0840. Optical observations would also be useful to search for a counterpart, and possibly to measure binary parameters." }, "0003/astro-ph0003162_arXiv.txt": { "abstract": " ", "introduction": "Numerical simulations of three-dimensional self-gravitating fluids have become an indispensable tool in cosmology. They are now routinely used to study the non-linear gravitational clustering of dark matter, the formation of clusters of galaxies, the interactions of isolated galaxies, and the evolution of the intergalactic gas. Without numerical techniques the immense progress made in these fields would have been nearly impossible, since analytic calculations are often restricted to idealized problems of high symmetry, or to approximate treatments of inherently nonlinear problems. The advances in numerical simulations have become possible both by the rapid growth of computer performance and by the implementation of ever more sophisticated numerical algorithms. The development of powerful simulation codes still remains a primary task if one wants to take full advantage of new computer technologies. Early simulations \\citep[][among others]{Ho41,Pee70,Pre74,Wh76,Wh78,Aar79} largely employed the direct summation method for the gravitational N-body problem, which remains useful in collisional stellar dynamical systems, but it is inefficient for large $N$ due to the rapid increase of its computational cost with $N$. A large number of groups have therefore developed N-body codes for collisionless dynamics that compute the large-scale gravitational field by means of Fourier techniques. These are the PM, P$^3$M, and AP$^3$M codes \\citep{Eas74,Hohl78,Hock81,Ef85,Cou91,Ber91,Ma98}. Modern versions of these codes supplement the force computation on scales below the mesh size with a direct summation, and/or they place mesh refinements on highly clustered regions. Poisson's equation can also be solved on a hierarchically refined mesh by means of finite-difference relaxation methods, an approach taken in the ART code by \\citet{Kra97}. An alternative to these schemes are the so-called tree algorithms, pioneered by \\citet{App81,App85}. Tree algorithms arrange particles in a hierarchy of groups, and compute the gravitational field at a given point by summing over multipole expansions of these groups. In this way the computational cost of a complete force evaluation can be reduced to a ${\\cal O}(N\\log N)$ scaling. The grouping itself can be achieved in various ways, for example with Eulerian subdivisions of space \\citep{Ba86}, or with nearest-neighbour pairings \\citep{Press86,Je89}. A technique related to ordinary tree algorithms is the fast multipole-method \\citep[e.g.][]{Green87}, where multipole expansions are carried out for the gravitational field in a region of space. While mesh-based codes are generally much faster for close-to-homogeneous particle distributions, tree codes can adapt flexibly to any clustering state without significant losses in speed. This Lagrangian nature is a great advantage if a large dynamic range in density needs to be covered. Here tree codes can outperform mesh based algorithms. In addition, tree codes are basically free from any geometrical restrictions, and they can be easily combined with integration schemes that advance particles on individual timesteps. Recently, PM and tree solvers have been combined into hybrid Tree-PM codes \\citep{Xu95,Bag99,Bod99}. In this approach, the speed and accuracy of the PM method for the long-range part of the gravitational force are combined with a tree-computation of the short-range force. This may be seen as a replacement of the direct summation PP part in P$^3$M codes with a tree algorithm. The Tree-PM technique is clearly a promising new method, especially if large cosmological volumes with strong clustering on small scales are studied. Yet another approach to the N-body problem is provided by special-purpose hardware like the \\grape\\ board \\citep{Mak90,Ito91,Fuk91,Mak93,Ebi93,Okam93,Fuk96,Ma97,Kaw2000}. It consists of custom chips that compute gravitational forces by the direct summation technique. By means of their enormous computational speed they can considerably extend the range where direct summation remains competitive with pure software solutions. A recent overview of the family of \\grape-boards is given by \\citet{Hut99}. The newest generation of \\grape\\ technology, the \\grape-6, will achieve a peak performance of up to 100 TFlops \\citep{Mak2000}, allowing direct simulations of dense stellar systems with particle numbers approaching $10^6$. Using sophisticated algorithms, \\grape\\ may also be combined with P$^3$M \\citep{Brieu95} or tree algorithms \\citep{Fuk91,Mak91,At97} to maintain its high computational speed even for much larger particle numbers. In recent years, collisionless dynamics has also been coupled to gas dynamics, allowing a more direct link to observable quantities. Traditionally, hydrodynamical simulations have usually employed some kind of mesh to represent the dynamical quantities of the fluid. While a particular strength of these codes is their ability to accurately resolve shocks, the mesh also imposes restrictions on the geometry of the problem, and onto the dynamic range of spatial scales that can be simulated. New adaptive mesh refinement codes \\citep{No98,Kl98} have been developed to provide a solution to this problem. In cosmological applications, it is often sufficient to describe the gas by smoothed particle hydrodynamics (SPH), as invented by \\citet{Lu77} and \\citet{Gi77}. The particle-based SPH is extremely flexible in its ability to adapt to any given geometry. Moreover, its Lagrangian nature allows a locally changing resolution that `automatically' follows the local mass density. This convenient feature helps to save computing time by focusing the computational effort on those regions that have the largest gas concentrations. Furthermore, SPH ties naturally into the N-body approach for self-gravity, and can be easily implemented in three dimensions. These advantages have led a number of authors to develop SPH codes for applications in cosmology. Among them are {\\small TREESPH} \\citep{He89,Ka96}, {\\small GRAPESPH} \\citep{St96}, {\\small HYDRA} \\citep{Cou95,Pe97}, and codes by \\citet{Ev88,Na93,Hu97,Da97,Ca98}. See \\cite{Ka94} and \\cite{Fr98} for a comparison of many of these cosmological hydrodynamic codes. In this paper we describe our simulation code \\gadget\\ ({\\bf GA}laxies with {\\bf D}ark matter and {\\bf G}as int{\\bf E}rac{\\bf T}), which can be used both for studies of isolated self-gravitating systems including gas, or for cosmological N-body/SPH simulations. We have developed two versions of this code, a serial workstation version, and a version for massively parallel supercomputers with distributed memory. The workstation code uses either a tree algorithm for the self-gravity, or the special-purpose hardware \\grape, if available. The parallel version works with a tree only. Note that in principle several \\grape\\ boards, each connected to a separate host computer, can be combined to work as a large parallel machine, but this possibility is not implemented in the parallel code yet. While the serial code largely follows known algorithmic techniques, we employ a novel parallelization strategy in the parallel version. A particular emphasis of our work has been on the use of a time integration scheme with individual and adaptive particle timesteps, and on the elimination of sources of overhead both in the serial and parallel code under conditions of large dynamic range in timestep. Such conditions occur in dissipative gas-dynamical simulations of galaxy formation, but also in high-resolution simulations of cold dark matter. The code allows the usage of different timestep criteria and cell-opening criteria, and it can be comfortably applied to a wide range of applications, including cosmological simulations (with or without periodic boundaries), simulations of isolated or interacting galaxies, and studies of the intergalactic medium. We thus think that \\gadget\\ is a very flexible code that avoids obvious intrinsic restrictions for the dynamic range of the problems that can be addressed with it. In this methods-paper, we describe the algorithmic choices made in \\gadget\\, which we release in its parallel and serial versions on the internet\\footnote{{\\scriptsize GADGET}'s web-site is:\\\\ {\\tt http://www.mpa-garching.mpg.de/gadget}}, hoping that it will be useful for people working on cosmological simulations, and that it will stimulate code development efforts and further code-sharing in the community. This paper is structured as follows. In Section \\ref{secphys}, we give a brief summary of the implemented physics. In Section \\ref{secgrav}, we discuss the computation of the gravitational force both with a tree algorithm, and with \\grape. We then describe our specific implementation of SPH in Section \\ref{secsph}, and we discuss our time integration scheme in Section \\ref{sectime}. The parallelization of the code is described in Section \\ref{secpara}, and tests of the code are presented in Section \\ref{secres}. Finally, we summarize in Section~\\ref{secdis}. ", "conclusions": "\\label{secdis} We have presented the numerical algorithms of our code \\gadget, designed as a flexible tool to study a wide range of problems in cosmology. Typical applications of \\gadget\\ can include interacting and colliding galaxies, star formation and feedback in the interstellar medium, formation of clusters of galaxies, or the formation of large-scale structure in the universe. In fact, \\gadget\\ has already been used successfully in all of these areas. Using our code, \\citet{Sp98} have studied the formation of tidal tails in colliding galaxies, and \\citet{Spr99} has modeled star formation and feedback in isolated and colliding gas-rich spirals. For these simulations, the serial version of the code was employed, both with and without support by the \\grape\\ special-purpose hardware. The parallel version of \\gadget\\ has been used to compute high-resolution N-body simulations of clusters of galaxies \\citep{Spr99c,Spr99b,Yos2000,Yos2000b}. In the largest simulation of this kind, 69 million particles have been employed, with 20 million of them ending up in the virialized region of a single object. The particle mass in the high-resolution zone was just $\\sim 10^{-10}$ of the total simulated mass, and the gravitational softening length was $0.7\\lu$ in a simulation volume of diameter $140\\mlu$, translating to an impressive spatial dynamic range of $2\\times 10^5$ in three dimensions. We have also successfully employed \\gadget\\ for two `constrained-realization' (CR) simulations of the Local Universe (Mathis~et~al. 2000, in preparation). In these simulations, the observed density field as seen by IRAS galaxies has been used to constrain the phases of the waves of the initial fluctuation spectrum. For each of the two CR simulations, we employed $\\sim 75$ million particles in total, with 53 million high-resolution particles of mass $3.6\\times 10^{9} h^{-1}\\msun$ ($\\Lambda$CDM) or $1.2\\times 10^{10}h^{-1}\\msun$ ($\\tau$CDM) in the low-density and critical-density models, respectively. The main technical features of \\gadget\\ are as follows. Gravitational forces are computed with a Barnes~\\&~Hut oct-tree, using multipole expansions up to quadrupole order. Periodic boundary conditions can optionally be used and are implemented by means of Ewald summation. The cell-opening criterion may be chosen either as the standard BH-criterion, or a new criterion which we have shown to be computationally more efficient and better suited to cosmological simulations starting at high redshift. As an alternative to the tree-algorithm, the serial code can use the special-purpose hardware \\grape\\, both to compute gravitational forces and for the search for SPH neighbours. In our SPH implementation, the number of smoothing neighbors is kept exactly constant in the serial code, and is allowed to fluctuate in a small band in the parallel code. Force symmetry is achieved by using the kernel averaging technique, and a suitable neighbour searching algorithm is used to guarantee that all interacting pairs of SPH particles are always found. We use a shear-reduced artificial viscosity that has emerged as a good parameterization in recent systematic studies that compared several alternative formulations \\citep{Th98,Lo98}. Parallelization of the code for massively parallel supercomputers is achieved in an explicit message passing approach, using the MPI standard communication library. The simulation volume is spatially split using a recursive orthogonal bisection, and each of the resulting domains is mapped onto one processor. Dynamic work-load balancing is achieved by measuring the computational expense incurred by each particle, and balancing the sum of these weights in the domain decomposition. The code allows fully adaptive, individual particle timesteps, both for collisionless particles and for SPH particles. The speed-up obtained by the use of individual timesteps depends on the dynamic range of the time scales present in the problem, and on the relative population of these time scales with particles. For a collisionless cosmological simulation with a gravitational softening length larger than $\\sim 30\\lu$ the overall saving is typically a factor of $3-5$. However, if smaller softening lengths are desired, the use of individual particle timesteps results in larger savings. In the hydrodynamical part, the savings can be still larger, especially if dissipative physics is included. In this case, adaptive timesteps may be required to make a simulation feasible to begin with. \\gadget\\ can be used to run simulations both in physical and in comoving coordinates. The latter is used for cosmological simulations only. Here, the code employs an integration scheme that can deal with arbitrary cosmological background models, and which is exact in linear theory, i.e.~the linear regime can be traversed with maximum efficiency. \\gadget\\ is an intrinsically Lagrangian code. Both the gravity and the hydrodynamical parts impose no restriction on the geometry of the problem, nor any hard limit on the allowable dynamic range. Current and future simulations of structure formation that aim to resolve galaxies in their correct cosmological setting will have to resolve length scales of size $0.1-1\\lu$ in volumes of size $\\sim 100 \\Mlu$. This range of scales is accompanied by a similarly large dynamic range in mass and time scales. Our new code is essentially free to adapt to these scales naturally, and it invests computational work only where it is needed. It is therefore a tool that should be well suited to work on these problems. Since \\gadget\\ is written in standard ANSI-C, and the parallelization for massively parallel supercomputers is achieved with the standard MPI library, the code runs on a large variety of platforms, without requiring any change. Having eliminated the dependence on proprietary compiler software and operating systems we hope that the code will remain usable for the foreseeable future. We release the parallel and the serial version of \\gadget\\ publically in the hope that they will be useful for others as a scientific tool and as a basis for further numerical developments." }, "0003/astro-ph0003139_arXiv.txt": { "abstract": "We present a theoretical model for the alternating phase lags associated with QPO fundamental and harmonic frequencies observed in some Galactic black-hole candidates. We assume that the accretion flow exhibits a transition from an outer cool, optically thick accretion disk to a hot, inner advection-dominated accretion flow (ADAF), and that the QPOs are related to small-scale oscillations of the accretion rate and the transition radius. We present an analytical estimate of the expected phase lags at the fundamental and first harmonic frequencies of the QPOs and perform detailed time-dependent Monte-Carlo simualtions of the radiation transport in the oscillating ADAF / cool disk system. We find that this model is well suited to reproduce alternating phase lags between the fundamental and the first harmonic. It also naturally explains the trend observed in GRS~1915+105 that, as the soft X-ray luminosity increases, the QPO frequency increases and the phase lag associated with the QPO fundamental frequency changes sign from positive to negative. The relation between the disk temperature and the QPO frequency observed in GRS~1915+105 is consistent with a secular instability modulating the disk evaporation at the transition radius. ", "introduction": "The steadily increasing amount of high-quality, high time-resolution X-ray data from Galactic X-ray binaries, has stimulated vital interest in the characteristics of the rapid variability of these objects. The X-ray emission from both Galactic black-hole candidates (GBHCs) and accreting neutron stars is known to vary on a wide range of time scales, sometimes showing quasi-periodic oscillations (QPOs) (for reviews see, e.g., \\cite{vdk95,cui99a}). In GBHCs, such QPOs have been detected at frequencies ranging from several mHz to $\\lesssim 100$~Hz. These QPOs are most notable when the sources are in the low-hard or in the rare very high state. Recently, correlations between the amplitudes and centroid frequencies of several types of QPOs with the spectral characteristics of the source have been found in some objects (\\cite{rutledge99}, Markwardt, Swank, \\& Taam \\markcite{mst99}1999, \\cite{sobczak99}, Muno, Morgan \\& Remillard \\markcite{mmr99}1999). A general pattern emerging from these analyses is that the frequency of those types of QPOs which do show a correlation with spectral properties, seems to increase with both the power-law and the disk black-body flux from the source. A surprising property of some types of QPOs has recently been found in the 67 mHz QPO of GRS~1915+105 (\\cite{cui99b}), the 0.5 -- 10~Hz QPO of the same source (\\cite{reig00}, \\cite{lin00}), and in the 0.3 -- 3~Hz QPO in XTE~J1550-564 (Cui, Zhang, \\& Chen \\markcite{cui00}2000): While the phase lag at the QPO fundamental frequency is negative, the phase lag associated with the first harmonic was positive. In the case of the low-frequency QPO of GRS~1915+105, even three harmonics were detected, and the phase lags were found to alternate between subsequent harmonics (\\cite{cui99b}). The phase lag associated with the 0.5 -- 10~Hz QPO of GRS~1915+105 was found to change sign from positive to negative as the QPO frequency increases above $\\sim 2.5$~Hz (\\cite{reig00}, \\cite{lin00}). These peculiar patterns are apparently completely counter-intuitive in the light of currently discussed models for the hard phase lags in X-ray binaries. Models proposed to explain the hard phase lags are either based to the energy-dependent photon escape time in Compton upscattering scenarios for the production of hard X-rays (Kazanas, Hua \\& Titarchuk \\markcite{kht97}1997, B\\\"ottcher \\& Liang \\markcite{bl98}1998), or due to intrinsic spectral hardening during X-ray flares, e.g. due to decreasing Compton cooling in active regions pushed away from an underlying accretion disk in a patchy-corona model (\\cite{pf99}) or due to density perturbations drifting inward through an ADAF toward the event horizon (B\\\"ottcher \\& Liang \\markcite{bl99}1999). These models dealt only with the continuum variability and did not consider the effects of QPOs. In this paper we will explore the response of a two-phase accretion flow, consisting of an outer, cool, optically thick accretion disk and an inner, hot ADAF (\\cite{ny94}, \\cite{abramowicz95}, \\cite{cal95}), to a periodically varying soft photon input from the cool disk (\\cite{lb00}). A two-phase accretion flow with an inner ADAF has been found to produce good fits to the photon spectra of, e.g., several Galactic X-ray binaries (e.g., Narayan, McClintock \\& Yi \\markcite{nmy96}1996, \\cite{hameury97}, \\cite{esin98}), low-luminosity AGN (\\cite{quataert99}) and giant elliptical galaxies (\\cite{fr95}, \\cite{df97}, \\cite{dimatteo00}). In \\S \\ref{analytical}, we describe the basic model setup according to this two-phase accretion flow, and derive an analytical estimate for the expected phase lags associated with the QPO and the first harmonic applicable in some simplified cases. A short description of the Monte-Carlo simulations used to solve the time-dependent radiation transport problem follows in \\S \\ref{mc}. In \\S \\ref{grs1915} we describe a series of simulations designed specifically to explain the peculiar phase lag behavior associated with the 0.5 -- 10~Hz QPO in GRS~1915+105. We summarize in \\S \\ref{summary}. ", "conclusions": "" }, "0003/astro-ph0003413_arXiv.txt": { "abstract": "We present new results on the recently discovered 69 ms X-ray pulsar \\axs, the sixth youngest sample of all known pulsars. We have undertaken a comprehensive X-ray observing campaign of \\axs\\ with the \\asca, \\sax, and \\xte\\ observatories and follow its long term spin-down history between 1989 and 1999, using these, archival \\ginga\\ and \\asca\\ data sets, and the radio ephemeris. The spin-down is not simply described by a linear function as originally thought, but instead we find evidence of a giant glitch ($|\\Delta P/P| \\stackrel{>}{_{\\sim}} 10^{-6}$) between 1993 August and 1997 September, perhaps the largest yet observed from a young pulsar. The glitch is well described by steps in $P$ and ${\\dot P}$ accompanied by a persistent ${\\ddot P}$ similar to those in the Vela pulsar. The pulse profile of \\axs\\ presents a single asymmetric peak which is maintained over all observation epochs. The energy spectrum is also steady over time, characterized by a highly absorbed power-law with a photon index $\\Gamma = 1.4 \\ \\pm\\ 0.2$, consistent with that found for other young rotation powered pulsars. ", "introduction": "Radio pulsars are thought to be highly magnetized ($ \\sim 10^{12}$ G), rapidly spinning neutron stars whose luminosity is powered by rotational energy loss. The study of young ($<10^{5}$ yrs) rotation-powered pulsars provides an important laboratory for understanding the early evolution (thermal, spin, and magnetic) of these embers of stellar collapse. The most energetic of these pulsars are observable at X-ray wavelengths, which allow us to probe these extreme, but rare, examples. In addition to uniform spin-down corresponding to the rotational energy loss, rotation powered pulsars show sudden discontinuities in their rotation periods (see Lyne \\& Graham-Smith 1998, chapter 6). These rare phenomena, known as ``glitches'' are considered to arise from sudden changes in the configuration of super-dense material in the neutron star interior. To date, a total of 71 glitches with $|\\Delta P/ P| > 10^{-9}$ have been reported in 30 pulsars (Urama \\& Okeke 1999). Observation of pulsar glitches gives us insights into the structure and physical processes inside the neutron stars, such as the interactions of neutron superfluid and crust components (e.g., Anderson \\& Itoh 1975). In this article we report the detection of a giant glitch from \\axs\\ by using the newly acquired multi-mission X-ray data. This source is an unusual case of a young pulsar discovered first by its X-ray emission, revealed during the course of an archival X-ray study of the SNR \\rcw\\ (Gotthel, Petre, \\& Hwang 1997). Further analysis detected highly significant pulsations from photons attributed to this source (Torii \\etal\\ 1998). The 69 ms pulse period was consistent with that reported from a \\ginga\\ observation of the region (Aoki, Dotani, \\& Mitsuda 1992). Recent radio observation has confirmed \\axs\\ as a young energetic rotation-powered pulsar (Kaspi \\etal\\ 1998). The lack of evidence for this object in archival soft X-ray images of this field suggested that the source might be highly absorbed, making it difficult to determine whether it is associated with a SNR. ", "conclusions": "A $\\chi ^2$ fit to the all 14 data points as summarized in table 1 gives the mean spin-down rate of $\\dot{P}= 1.3611(1) \\times 10^{-13}$~s/s and $P=0.069347150(1)$ at MJD 50,000.0 (Figure 1). However, the quality of the fit is bad, with $\\chi ^2/d.o.f. = 4808/12$. The residual of the fit shows a jump of $\\Delta P \\simeq -1.2\\times 10^{-7}$~s between the observations of 1993 August (MJD 49,217.6) and 1997 September (MJD 50,696.0) (Figure 2, top panel). Within the observation span of 10 years, the residual is neither periodic nor smooth. A sudden change in the period between MJD 49,217.6 and MJD 50,696.0 is suggested. We consider the most likely explanation for these residuals is due to a glitch, similar to those observed in several young rotation-powered pulsars. Given the clear evidence of glitch activity, we next attempted to model the spin-down data with a single glitch followed by an exponential recovery. The data coverage is limited and we simply assumed the following relation for the spin-down. \\begin{equation} P(t) = P_0 + {\\dot P} \\cdot (t-t_0) + \\Delta P \\cdot \\exp (-\\frac{t-t_0}{\\tau}) \\end{equation} where $\\Delta P=0$ for $t0$) for $t \\geq t_0$. This model contains six parameters. Again, the time of the glitch, $t_0$, had to be assumed. The derived parameters are summarized in table 3. For the condition that $\\Delta P_0 < 0$, $t_0$ was restricted to $t_0 \\stackrel{<}{_\\sim} 50,205$. The residual for $t_0=50,000$ is shown in the bottom panel of figure 2. The quality of the fit is significantly improved to $\\chi^2 /d.o.f. = 17.1/9$. Compared to the fit to Equation (1), the f-test gives a chance probability of $\\sim 0.1$. This result suggests that transient effects had indeed ceased by MJD~50,696. The reduced $\\chi ^2$ is still larger than unity, suggesting the presence of timing noise and smaller glitches. We can estimate the expectancy of large ($|\\Delta P/P| > 10^{-7}$) glitches by using the semi-empirical relation based upon the superfluid vortex unpinning model (Alpar \\& Baykal 1994). Using this relation, the expected number of large glitches between the first \\ginga\\ observation and the last \\asca\\ observations is 3.7. Therefore, there should have been about 4 glitches of $|\\Delta P/P| > 10^{-7}$. Indeed the residuals to the fit of Equation (1) or (2) still hint a small jump of $\\Delta P = -5\\times 10^{-9}$~s between MJD~51,263.8 (1999 Mar.) and MJD~51,394.2 (1999 Aug). This may be another glitch of $\\Delta P/P= -7\\times 10^{-8}$, much smaller in magnitude than the one near MJD~50,000, but still relatively large compared to those seen for most radio pulsars (Shemar \\& Lyne 1996). We have thus found a giant glitch of $|\\Delta P/ P| \\stackrel{>}{_{\\sim}} 10^{-6}$ at $t_0 \\simeq 50,000$~MJD ($49,218 \\leq t_0 \\leq 50,696$) and possibly a glitch of $|\\Delta P/ P| \\simeq 7\\times 10^{-8}$ at $t_0 \\simeq 51,300$~MJD ($51,264 \\leq t_0 \\leq 51,394$). The detection of a giant glitch from \\axs\\ gives a rare sample for studying the origin of pulsar glitches and the interior structure of neutron stars. In this context, regular timing observation of this pulsar in the radio band is quite important to monitor the onset of a glitch and following transient effects. Long term timing observations for measuring the braking index are desired. Searches for the highly absorbed X-ray emission from the supernova remnant associated with this pulsar should be undertaken with the next generation X-ray observatories." }, "0003/astro-ph0003249_arXiv.txt": { "abstract": "The \\MONS{} satellite will have two Star Trackers to sense the spacecraft attitude, and we plan to use them as scientific instruments to perform high-precision photometry of thousands of stars. We briefly describe the current plans for the Star Trackers and their expected capabilities. ", "introduction": "Like many satellites that require precise attitude control, \\MONS{} will use star trackers to sense the spacecraft attitude. A star tracker is basically a wide-field CCD camera. In acquisition mode, images of the sky are compared with a star catalogue to determine the absolute orientation of the spacecraft. Once a target is acquired, the spacecraft attitude is continuously updated in tracking mode. There is clear scientific interest in using the photometry from a star tracker to observe variable stars. Despite the small aperture, the advantages of space (long observing periods and no atmospheric scintillation) make such a camera superior to ground-based telescopes for some applications. The only example so far has been the work by Buzasi~et~al. (2000 and these Proceedings), who are making impressive use of the 52-mm star camera on the {\\em WIRE\\/} satellite after the failure of the main instrument. As far as we know, \\MONS{} is the first mission to be designed from the start to use star trackers for science. Of course, we must keep in mind that the primary role of the star trackers is to sense spacecraft attitude with the precision required for the main camera. We describe here the plan for the \\MONS{} Star Trackers as it stood at the time of the Workshop (January 2000). This plan differs slightly from that described in the \\MONS{} Proposal (Kjeldsen, Bedding \\& Christensen-Dalsgaard 1999), and will undoubtedly change again before the final design is frozen\\footnote{One change is that we now use the term ``Star Tracker'' rather than ``Star Imager.''}. There are several factors driving the changes, including: (i)~performance of trackers as attitude sensors for the main camera, including redundancy issues; (ii)~volume, thermal and power constraints from the possible presence of Ballerina; and (iii)~performance of trackers as secondary science instruments. ", "conclusions": "" }, "0003/astro-ph0003280_arXiv.txt": { "abstract": "We compare $850\\mu$m SCUBA images of NGC 891 with the corresponding V-band optical depth predicted from radiation transfer simulations. These two tracers of dust show a very similar distribution along the minor axis and a reasonable agreement along the major axis. Assuming that the grains responsible for optical extinction are also the source of $850\\mu$m emission we derive a submillimeter emissivity (emission efficiency) for dust in the NGC 891 disk. This quantity is found to be a factor of 2-3 higher than the generally-accepted (but highly uncertain) values adopted for the Milky Way. It should be stated, however, that if a substantial fraction of dust in NGC 891 is clumped, the emissivity in the two galaxies may be quite similar. We use our newly-acquired emissivity to convert our $850\\mu$m images into detailed maps of dust mass and, utilizing 21cm and CO-emission data for NGC 891, derive the gas-to-dust ratio along the disk. We compute an average ratio of 260 -- a value consistent with the Milky Way and external spirals within the uncertainties in deriving both the dust mass and the quantity of molecular gas. The bulk of dust in NGC 891 appears to be closely associated with the molecular gas phase although it may start to follow the distribution of atomic hydrogen at radii $>$9 kpc (i.e.~$> \\frac{1}{2}R_{25}$). Using the optical depth of the NGC 891 disk, we quantify how light emitted at high redshift is attenuated by dust residing in foreground spirals. For B-band observations of galaxies typically found in the Hubble Deep Field, the amount of light lost is expected to be small ($\\sim$ 5\\%). This value depends critically on the maximum radial extent of cold dust in spiral disks (which is poorly known). It may also represent a lower limit if galaxies expel dust over time into the intergalactic medium. ", "introduction": "\\label{intro} The introduction of submillimeter (submm), imaging arrays such as SCUBA (Submillimeter Common User Bolometer Array; Holland et al 1999), heralds a revolution in our understanding of the dust properties of external spiral galaxies. For the first time, we can map cold (15-20K) interstellar grains with high sensitivity and good spatial resolution ($\\sim 10''$) (Chini et al 1995; Hughes et al 1997; Clements et al 1993). Previous studies of radiation emitted by dust have relied heavily on measurements carried out by the Infrared Astronomical Satellite (IRAS) at 60 and $100\\mu$m. These wavebands are sensitive primarily to warm ($\\sim$ 30K) dust, and as such, are unlikely to give an accurate impression of either the true quantity or distribution of interstellar grains in external disks (Alton et al 1998a; Devereux \\& Young 1990). Furthermore, satellite missions such as IRAS, and the recently completed Infrared Space Observatory (ISO) project, are constrained at present to using small mirrors (0.6-m c.f.~SCUBA 15-m dish). This limits their resolving capability to $\\sim 1'$ in the far-infrared (FIR) which is only adequate for probing disk structure in the very closest spiral galaxies (Alton et al 1998b).\\\\ One of the first studies of nearby galaxies, using SCUBA, was carried out by Alton et al (1998c) who mapped the edge-on galaxy NGC 891 at 450 and $850\\mu$m (see also Israel et al 1999). By carefully comparing the submm emission over the same region as the FIR flux ($60-100{\\mu}m$), they found that this galaxy contains about an order of magnitude more dust than would have been inferred from IRAS observations alone. This same conclusion was reached by Xilouris et al (1999) who fitted the optical and near-infrared surface photometry of both NGC 891 and 6 other nearby, edge-on spirals with a sophisticated radiation transfer simulation. Their model, which takes account of both scattering and absorption by interstellar grains, only produces a self-consistent solution across several wavebands if the extinction lane harbours about 10 times more dust than might be expected from IRAS measurements.\\\\ In this paper, we relate the extinction modelling carried out for NGC 891 (Xilouris et al 1998; hereafter XAD) with the corresponding submm images acquired from SCUBA (Alton et al 1998c; hereafter ABR). In doing so, we address some important issues, namely: \\begin{enumerate} \\item whether the submm emission, which accounts for the bulk of grain material in NGC 891, correlates closely with the dust evident in the extinction lane (\\S\\ref{rtm}); \\item in the case of a positive correlation, whether the submm emissivity in NGC 891 (or, equivalently, the grain mass-absorption coefficient) is comparable to that recorded for our own galaxy (\\S\\ref{emissivity}); \\item given that we recover most of the dust mass from our submm images, how the gas-to-dust ratio varies along the major axis of NGC 891(\\S\\ref{gdratio}); \\item how the submm emission is distributed with respect to the neutral gas phases (HI,H$_{2}$) and the stars in the disk (\\S\\ref{gdratio}). This information will prove crucial to the growing number of submm surveys, where sources are poorly resolved and subsidary observational data are limited. \\end{enumerate} We begin by summarizing the salient results from XAD and ABR (\\S\\ref{submm} and \\S\\ref{rtm}) before seeking to relate the submm to the extinction dust lane. ", "conclusions": "\\label{summary} We have compared the $850\\mu$m emission detected from the nearby, edge-on spiral NGC 891 (Alton et al 1998) with the corresponding optical depth predicted from a sophisticated radiation transfer model (Xilouris et al 1998). Morphologically, both model and submm emission show a very similar distribution along the minor axis, with most of the grain mass contained within a 600pc layer (FWHM). Profiles along the major axis show fair agreement between optical extinction and submm radiation. The $850\\mu$m emission is shown to trace the bulk of interstellar dust within NGC 891. Therefore, assuming both extinction and submm emission are attributable to the same grains, we can derive the emissivity (or, equivalently, mass absorption coefficient) at $850\\mu$m and, thence, the distribution of dust mass along the disk. We believe that this technique is more robust than taking estimates of FIR emissivity for Galactic grains (Hildebrand 1983; Draine and Lee 1984), and extrapolating them to submm wavelengths. The $850\\mu$m emissivity we derive for NGC 891 is $Q(850{\\mu}m)\\simeq\\frac{1}{10600}$, which is about a factor of 2-3 higher than the Draine and Lee (1984) model of Galactic dust. Uncertainty in our technique means that the emissivity in NGC 891 could be a factor of 2 higher or lower than the quoted value. Importantly, given that the employed radiation transfer model neglects dust-clumping, a {\\em decrease} of factor 2 seems quite likely.\\\\ We have computed the gas-to-dust ratio along the major axis of NGC 891 using our newly-acquired submm emissivity. We derive a value of 260, which is consistent, within the uncertainties, with estimates for the diffuse ISM of the Milky Way and nearby spirals. The dust in NGC 891 appears to be closely associated with the molecular gas phase, at least for the inner half of the disk. At radii $>$9 kpc ($> \\frac{1}{2} R_{25}$), it is possible the grains begin to follow the distribution of atomic hydrogen. We use the opacity of the NGC 891 disk to calculate the degree with which light emitted at high redshift is attenuated by foreground field spirals. For galaxies typically associated with the Hubble Deep Field (z=1-2), the fraction of light lost, in this way, is only 5\\%. Since this estimate is based on a single galaxy as a model, further submm measurements will be required to support our conclusions. In particular, observations at the periphery of spiral disks could establish the extent to which cold dust is associated with the expansive HI envelopes surrounding spiral galaxies (it is the maximum radial extent of the dust which primarily controls the level of foreground extinction). Futhermore, we point out that certain lines-of-sight through nearby clusters are likely to be much more optically thick (than our figure of 5\\% would suggest), particularly if grains are expelled into the intergalactic medium via starburst winds (Alton et al 1999c) or dispersed through tidal interactions (Yun et al 1994).\\\\ Finally, we note the strong correspondance between dust and molecular gas in NGC 891. Further submm imaging of face-on spirals (both H$_{2}$ rich and H$_{2}$ poor) would confirm the trend indicated by the current observations, that the bulk of galactic dust appears to be associated with the molecular gas phase." }, "0003/hep-ph0003256_arXiv.txt": { "abstract": "The neutrinos and antineutrinos of all the three flavors released from a galactic supernova will be detected in the water Cerenkov detectors. We show that even though the neutral current interaction is flavor blind, and hence neutrino flavor mixing cannot alter the total neutral current signal in the detector, it can have a non-trivial impact on the delay of massive neutrinos and alters the neutral current event rate as a function of time. We have suggested various variables of the neutral and charged current events that can be used to study this effect. In particular the ratio of charged to neutral current events can be used at early times while the ratio of the energy moments for the charged to the neutral current events can form useful diagnostic tools even at late times to study neutrino mass and mixing. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003233_arXiv.txt": { "abstract": "Gravitational potential energy of a homogeneous bispherical concavo-convex lens, with the equal absolute values of curvature radii of both surfaces, is studied. Compactness factor notion is introduced and calculated as function of central thickness of lens. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003005_arXiv.txt": { "abstract": "Binary merging constitutes a complementary mode of galaxy evolution to the canonical hierarchical clustering theory. This merger-driven evolution not only influences the galaxy mass distribution function, but may drive the main galactic activity cycles, such as starbursts and the activity in the nuclei (AGNs). In this paper, we use galaxy aggregation dynamics together with a possible merger-driven starburst and AGN phenomena, to study the effects of evolution of these consequential activities for the faint IRAS source counts and the infra-red background. We find that the strong evolution of IRAS 60$\\,\\mu m$ source count at flux range of $10\\,mJy \\,\\sim \\, 1\\,Jy$ is difficult to be explained only by the merger rate decrease with cosmic time. We need to assume a redshift dependent infrared burst phase of ultraluminous infrared galaxies (ULIGs) from the gas rich mergers at high redshift and the gas poor mergers at low redshift. The background intensity at 60$\\,\\mu m$ which we get from this aggregation evolution is a lower limit of only $1.9\\,nW\\,m^{-2}\\,sr^{-1}$, about half of those estimated by some previous models, but close to the lower end of the range derived by Malkan \\& Stecker (1998). ", "introduction": "Over the last few years, observational progress in discovering the faint Universe at ultraviolet (UV)/optical and infrared wavelength via the groundbased redshift surveys, Hubble Space Telescope (HST) multi-color images and the infrared satellites such as IRAS (Infrared Astronomical Satellite), ISO (Infrared Space Observatory) and COBE (Cosmic Background Explorer), challenges and renews our present understanding of cosmological structure formation and the evolution of luminous galaxies, which are usually based on two competing scenarios ``traditional monolithic collapsing'' versus ``hierarchical clustering theories''. The main difference between these two kinds of models is the formation mechanism and the first epoch of the formation of massive galaxies. More recently it has been shown that the emission history (e.g. Madau plot) derived from UV/optical detections agrees quite well with the star formation history indicated by hierarchical clustering theories (Madau 1997, Madau et al. 1996, 1998, Ellis et al. 1996). Since the FIR/submm unveiled a dust-shrouded early active phase and the recent detection of high redshift evolving galaxies by deep NICMOS/VLT images (Lilly et al. 1998, Benitez et al. 1998), the discussion of the structure formation and the scheme of galaxy evolution is going to require more subtlety and sophistication. For a clear picture of the star formation history and the galaxy evolution, we probably need more information at high redshift, especially from FIR/submm deep surveys. A successful interpretation of the excess of faint blue galaxies within the merger-trigger starburst scenario in a hierarchical clustering universe has been presented by Cavaliere \\& Menci (1997) using binary aggregation dynamics. The binary aggregation dynamical approach for the galaxy evolution presented by Cavaliere \\& Menci (1993, 1997), which includes more dynamics to describe a further step in galaxy-galaxy interactions within the scheme of direct hierarchical clustering (DHCs), probably can help to alleviate some intrinsic problems in the DHC scenario, such as the overproduction of small objects and the difficulty of a reconciliation between the excess faint counts and the flat local luminosity function. While, on the other hand, binary merging which plays a different and complementary role in structure formation, can actually continue to flatten the mass distribution function $N(m,z)$ and hence the luminosity function $N(L,z)$ till even moderate redshift $z<1$. Although there are many other possible evolutionary scenarios which could interpret the present observations in this or that way, the reason that we are encouraged to explore here a merger-driven galaxy evolution picture with the binary aggregation dynamics, is simply because the IRAS database does show that most of the luminous infrared sources are actually interacting/merging systems (Kleinmann \\& Keel 1987, Sanders \\& Mirabel 1996, Hutchings \\& Neff 1987, Vader \\& Simon 1987a,b), and many merger events can even occur till moderate redshift ($z<1$). In this paper, starting from the basic picture that galaxy evolution is driven by the galaxy-galaxy interaction complementary to the standard hierarchical clustering scenario (DHCs), we investigate a simple model with the basic concepts: 1) galaxy-galaxy interaction can even occur till quite moderate redshifts ($z<1$) within large scale structures that could offer the best combination of volume and density contrast, and it could erase the dwarf part of the mass distribution and produce the massive tail, thus flatten the luminosity functions; 2) starburst/AGN activities may be triggered by the merger events during the structure formation and evolution. This is actually the point different from other models where only the starburst is considered as the consequence of merger events (Cavaliere \\& Menci 1997, Somerville \\& Primack 1998, Somerville et al. 1998); 3) the gas rich merger events at high redshift may trigger drastic starburst and AGN activities in the central region of the merging galaxies and thus enhance dramatically the FIR luminosity because of the accumulating of dust grains from their progenitor galaxies as well as the dust newly formed in the star forming regions. This FIR luminosity burst phase is usually called ultraluminous infrared galaxies (ULIGs); 4) in our model, we assume a redshift dependent infrared burst phase for the gas rich merger events at high redshift and the gas poor mergers at low redshift, which means the enhancement of the infrared luminosities by the high redshift gas rich mergers is much higher than those of low redshift gas poor mergers. The motivation for this kind of thinking is: 1) direct observation of enhanced starburst-AGN activities in interacting galaxies, especially the ULIGs. This extremely infrared bright burst phase is believed due to the starburst merger events with far infrared luminosity $L_{fir}$ enhanced both by the accumulation of the dust mass $M_{d}$ and the increase of dust temperation $T_{d}$ as the relation $L_{fir} \\propto M_{d}\\,\\,T_{d}^5$. This burst phase could increase the infrared luminosity by a factor of about 20 over that of normal starburst galaxies. (Ashby et al. 1992, Terlevich \\& Melnick 1985, Heckman 1997, Perry \\& Williams 1993, Taniguchi \\& Ohyama 1998); 2) numerical simulation of the starburst/AGN evolutions by galaxy mergers, especially for their correlation and the burst phase (Wang \\& Biermann 1998, 1999, Wang 1999); 3) the successful interpretation of the excess faint blue counts by galaxy merging and the consequential starburst activities (Rocca-Volmerange \\& Guiderdoni 1990, Cavaliere \\& Menci 1997, Carlberg 1992, Calberg \\& Charlot 1992); 4) the pair production absorption of high energy $\\gamma$ rays by intergalactic low energy photons is expected to produce a high energy cutoff in extragalactic $\\gamma$ ray spectra. The new data of Mkn501 appear to show an extension of the TeV $\\gamma$ ray spectrum till about 24 TeV, which sets an upper limit for the intergalactic infrared emission history during the structure formation (Aharonian et al. 1999). Since the dust-shrouded geometry could strongly affect the AGN spectrum and cause significant radiation at infrared wavelength in the framework of a Unification Scheme, we thus add the statistics of nuclear activity in galaxies as one additional constraint in the context of the various models already in the literature. We know that nuclear activity is detected to very high redshift and provides a strong constraint on the cosmological evolution models. Therefore we are less conservative than Malkan \\& Stecker (1998), but use more constraints than, e.g., Somerville \\& Primack (1998). Obviously, any such modelling depends on the basic concepts used, and thus our results provide a strong lower limit to the far-infrared background. The outline of this paper is as follows: 1) in Sect. \\ref{dyn}, we introduce the binary aggregation dynamics, and the models of the interaction kernel; 2) we discuss the prescriptions of mass-to-light ratio in our model for the luminous infrared galaxies which is enhanced by the starburst and AGN activities in Sect. \\ref{mass-light}, especially modelling of the redshift dependent infrared burst phase for ultraluminous infrared galaxies (ULIGs) from gas rich mergers at high redshift and a suppressed burst phase for gas poor mergers at low redshift; 3) in Sect. \\ref{num}, we discuss the numerical simulations and compare the Monte Carlo results with the IRAS $60\\,\\mu m$ source counts from three major infrared sources (starburst galaxies, Seyferts and spirals). We check also the redshift distribution of these infrared sources which are brighter than $S_{60}\\sim 10\\,mJy$, calculate the integrated background level at wavelength of $60\\,\\mu m$, and make an extrapolation to $25\\,\\mu m$ and $100\\,\\mu m$ based on the model spectrum and souce counts. Finally, we give our conclusions. ", "conclusions": "We described in this paper a Monte Carlo simulation of the inverse-cascading process of a merging-driven galaxy evolution scenario, where the evolution of infrared luminous starburst galaxies and AGNs may be triggered by the galaxy-galaxy interactions till even moderate redshift (say, $z < 1$) in the Large Scale Structures. We assume in our model a redshift dependent infrared burst phase which is based on the concepts that starburst and AGN activities triggered by gas rich mergers at high redshift are more drastic than those of low redshift, thus the enhancement of the far infrared luminosities in these ULIGs from the high redshift merger events are higher than that of low redshift. We simulate this effect in our calculation by a power law suppression of the infrared burst phase since a transition redshift $z\\sim 1$. We adopt the transition redshift here at $z\\sim 1$ simply because the cosmic time scale of $z\\sim 1$ ($3\\times 10^9$ years) is approximately the disk evolution time scale (Lin \\& Pringle 1987). Varying any of the model parameters could influence the evolution of luminosity functions of infrared luminous sources and thus influence the souce counts more or less. But, no matter how we adjust the relevant parameters, such as $\\alpha$, $D$ and $\\eta$, in order to get a strong decrease of merger rate with cosmic time around redshift $1\\sim 2$ for the source count fitting, it seems that the quick fading and suppression of the infrared burst phase at redshift $z\\sim 1$ is a critical effect in our simulation to interpret the strong evolution of the IRAS $60\\,\\mu m$ source counts at flux range $10\\,mJy\\sim \\,1\\,Jy$. The impact from the variation of model parameters such as $\\alpha$, $D$ and $\\eta$, probably could be compensated by varying again other related parameters, thus their particular values in our model are not the critical points for such an evolutionary scenario; However, varying the transition redshift and the infrared enhancement $\\eta$ by a factor of two could strongly influence the final results. So, it appears that there is not much room for varying any of these parameters. Since our model is still too simple and crude, we probably need further information about the evolution of luminosity functions at high redshift for those infrared bright sources in order to give strong model constraints. We checked also the redshift distribution of the three major infrared sources (starburst galaxies, Seyferts and spirals). We see from Fig.\\ref{dis} that the mean redshift of starburst galaxies and AGNs which are brighter than $S_{60}\\sim 10\\,mJy$ is around $z\\sim 0.5$ and quickly diminish till $z\\sim 1$; a new population of high redshift ultraluminous infrared burst phase at mean redshift $z\\sim 1.5$ then takes over. This seems to be consistent with the present groundbased NIR and optical/UV HDF surveys, where they failed to have enough detection of starburst galaxies near $z\\sim 1$ (Ashby et al. 1992, Koo \\& Kron 1992). On the other hand, recent NICMOS/VLT and FIR/submm survey surely found a certain amount of infrared bright galaxies at high redshift, and the newest result of FIRBACK (Far Infrared Survey) with ISO shows that more than half of the ultraluminous infrared galaxies are at redshift $z>1$ (Lilly et al. 1998, Benitez et al. 1998, Dole et al. 1998). This provides further motivation to consider their contribution to the strong evolution of the IRAS faint source counts at flux range of $10\\,mJy\\sim 1\\,Jy$ in Fig.\\ref{fcount}. Meanwhile the new data of nearby blazars like Mkn501 by HEGRA team, appears to show an extension of the TeV $\\gamma$ ray spectrum till about 24 TeV, which sets an upper limit for the intergalactic infrared emission history during the structure formation (Aharonian et al. 1999). We calculated the background level at $60\\,\\mu m$ from a possible merger driven starbursts and AGNs scheme and a simple constant star formation history for the spiral galaxies. The infrared background level at $60\\,\\mu m$ is only $1.9\\,nW\\,m^{-2}\\,sr^{-1}$, which is about half of those estimated by some previous papers and consistent with the upper limit from the new data of TeV $\\gamma$ ray spectrum of Mkn501 (Stecker 1999, Funk et al. 1998). Clearly, this is a strong lower limit, because any variation on our model could produce a higher background. The forth-coming data on direct source counts at infrared wavelengths will allow to further constrain the evolution of both AGN and starbursts, as well as the absorption at gamma rays near 10 TeV photon energy." }, "0003/astro-ph0003143_arXiv.txt": { "abstract": "Radiatively inefficient (or advection dominated) disks are discussed at an introductory level. Ion supported and radiation supported flows are discussed, the different consequences of advection dominated flows onto black holes vs. solid surfaces (neutron stars, white dwarfs), hydrodynamics, the role of the ratio of specific heats, and the possible connection between ADAFs and outflows. To appear in `The neutron star black hole connection' (NATO ASI Elounda 1999, eds. C. Kouveliotou and V. Connaughton). \\keywords neutron stars, black holes, accretion: accretion disks ", "introduction": "In a thin accretion disk, the time available for the accreting gas to radiate away the energy released by the viscous stress is the accretion time, \\beq t_{\\rm acc}\\approx{1\\over\\alpha\\Omk}({r\\over H})^2, \\eeq where $\\alpha$ is the dimensionless viscosity parameter, $\\Omk$ the local Keplerian rotation rate, $r$ the distance from the central mass, and $H$ the disk thickness (see Frank et al. 1992 or Spruit, elsewhere in this volume). For a thin disk, $H/r\\ll 1$, this time is much longer than the thermal time scale $t_{\\rm t}\\approx 1/(\\alpha\\Om)$. There is then enough time for a local balance to exist between viscous dissipation and radiative cooling. For the accretion rates implied in observed systems the disk is then rather cool, and the starting assumption $H/r\\ll 1$ is justified. This argument is somewhat circular, however, since the accretion time is long enough for effective cooling only if the disk is assumed to be thin to begin with. Other forms of accretion disks may exist, even at the same accretion rates, in which the cooling is ineffective compared with that of standard (geometrically thin, optically thick) disks. Since radiatively inefficient disks tend to be thick, $H/r\\sim O(1)$, they are sometimes called `quasi-spherical'. However, this does {\\em not} mean that a spherically symmetric accretion model would be a reasonable approximation. The crucial difference is that the flow has angular momentum. The inward flow speed is governed by the rate at which angular momentum can be transferred outwards, rather than by gravity and pressure gradient. The accretion time scale, $t_{\\rm acc}\\sim 1/(\\alpha\\Om)$ is longer than the accretion time scale in the spherical case (unless the viscosity parameter $\\alpha$ is as large as $O(1)$). The dominant velocity component is azimuthal rather than radial, and the density and optical depth are much larger than in the spherical case. It turns out that there are two kinds of radiatively inefficient disks, the optically thin, and optically thick varieties. A second distinction occurs because accretion flows are different for central objects with a solid surface (neutron stars, white dwarfs, main sequence stars, planets), and those without (i.e. black holes). I start with optically thick flows. ", "conclusions": "" }, "0003/astro-ph0003157_arXiv.txt": { "abstract": "We show that the relativistic wind of the Crab pulsar, which is commonly thought to be invisible in the region upstream of the termination shock at $r \\leq r_{\\rm S} \\sim 0.1 \\, \\rm pc$, in fact could be directly observed through its inverse Compton (IC) $\\gamma$-ray emission. This radiation is caused by illumination of the wind by low-frequency photons emitted by the pulsar, and consists of two, {\\it pulsed} and {\\it unpulsed}, components associated with the nonthermal (pulsed) and thermal (unpulsed) low energy radiation of the pulsar, respectively. These two components of $\\gamma$-radiation have distinct spectral characteristics, which depend essentially on the site of formation of the kinetic-energy-dominated wind, as well as on the Lorentz factor and the geometry of propagation of the wind. Thus, the search for such specific radiation components in the spectrum of the Crab Nebula can provide unique information about the unshocked pulsar wind that is not accessible at other wavelengths. In particular, we show that the comparison of the calculated flux of the unpulsed IC emission with the measured $\\gamma$-ray flux of the Crab Nebula excludes the possibility of formation of a kinetic-energy-dominated wind within 5 light cylinder radii of the pulsar, $R_{\\rm w} \\geq 5 R_{\\rm L}$. The analysis of the pulsed IC emission, calculated under reasonable assumptions concerning the production site and angular distribution of the optical pulsed radiation, yields even tighter restrictions, namely $R_{\\rm w} \\geq 30 R_ {\\rm L}$. ", "introduction": "\\label{intro} The Crab pulsar is a powerful nonthermal machine, accelerating plasma in the form of a relativistic wind that carries off most of the rotational energy of the pulsar. At a distance of about $r=r_{\\rm S} \\sim 0.1 \\, \\rm pc$ the wind is terminated by a standing reverse shock, which accelerates the electrons up to energies $10^{15} \\, \\rm eV$, and randomizes their pitch angles \\cite{rees,kennel}. This results in formation of a bright synchrotron source in the region downstream of the shock. The synchrotron radiation of the Crab Nebula is well studied in a very broad frequency range, from radio to hard X-rays. Its general spectral and spatial characteristics are satisfactorily explained by the relativistic magnetohydrodynamics (MHD) model of Kennel \\& Coroniti \\shortcite{kennel}. Remarkably, the latter provides also a reasonable explanation, even in its simplified (spherically symmetric) form, for the detected very high energy (VHE) $\\gamma$-rays as a result of inverse Compton (IC) scattering of relativistic electrons in the ambient low-frequency photon fields. This implies that the study of the TeV IC radiation of the Crab Nebula, combined with synchrotron X-ray emission, can yield unambiguous information about the relativistic electrons and the nebular magnetic field in the downstream region of the shock \\cite{jager,stepanian,atoyan,hillas}. Although very important, this information unfortunately does not tell much about the origin and characteristics of the wind, i.e. about the region between the pulsar magnetosphere and the shock. It is generally believed that this region, where almost the whole rotational energy of the pulsar is somehow released in the form of kinetic energy of the wind, cannot be directly observed. This has a simple explanation. Although the wind electrons may have an energy as large as $\\sim 10^{13}$ eV, they move together with the magnetic field and thus do not emit synchrotron radiation. This explains the fact that the region upstream of the shock is {\\it underluminous} \\cite{kennel}. However, this statement is valid only for the synchrotron radiation of the wind. In fact, the wind could be {\\it directly} observed through its IC radiation. Indeed, the IC $\\gamma$-radiation of the wind electrons is unavoidable because of the illumination of the wind by external low-energy photons of different origin. There are three isotropic photon field populations that contribute effectively to the production of IC $\\gamma$-radiation of the nebula in the downstream region: the nonthermal (synchrotron) and thermal (dust) radiation of the Crab Nebula itself, and the 2.7 K microwave background radiation (Atoyan \\& Aharonian 1996). Formally, all these photon fields could also serve as a target for the IC scattering of the wind before the shock. However, the fluxes of $\\gamma$-rays emitted through these channels appear to be well below the sensitivity threshold of $\\gamma$-ray telescopes. Meanwhile, in the pulsar vicinity, namely within approximately 100 light cylinders, the efficiency of production of IC $\\gamma$-rays dramatically increases because of the existence of intense low energy radiation of the pulsar itself. The radiation consists of two, pulsed and unpulsed components, the latter being modulated at the period of the pulsar. In this paper we show that at certain circumstances concerning the position of formation of the particle dominated wind, the geometry of the flow and the Lorentz-factor of the bulk motion of the wind, the fluxes of the IC $\\gamma$-radiation of the wind could be sufficiently high to enable detection by present and forthcoming space-borne and ground-based $\\gamma$-ray telescopes. Moreover, the distinct spectral features of this radiation could allow effective separation of the ``wind'' component of radiation from the heavy background, which consists of unpulsed radiation of the Crab Nebula at very high (TeV) energies and pulsed radiation at low (GeV) energies. We argue, that even upper limits obtained in such a study could provide unique information about the origin of the pulsar wind. ", "conclusions": "The Crab Nebula is a unique cosmic laboratory with an unprecedentedly broad spectrum of the observed nonthermal radiation that extends throughout 21 decades of frequency -- from radio wavelengths to very high energy $\\gamma$-rays (for review see e.g. Aharonian \\& Atoyan 1998) It is commonly accepted that the synchrotron nebula is powered by the relativistic wind of electrons generated at the pulsar and terminated by a standing reverse shock wave at a distance $r_{\\rm s} \\sim 0.1 \\, \\rm pc$ \\cite{rees}. The relativistic MHD models, even in their simplified form (e.g. ignoring the axisymmetric structure of the wind and its interaction with the optical filaments), successfully describe the general characteristics of the synchrotron nebula, and predict realistic distributions of relativistic electrons and magnetic field in the downstream region behind the shock \\cite{kennel}. Meanwhile our knowledge of the unshocked wind, i.e. about the region between the pulsar magnetosphere and the shock is based only on theoretical speculations. Moreover, it is commonly thought that the wind could not be visible in the region upstream of the termination shock because the relativistic electrons and magnetic field in wind move together, thus the unshocked wind does not produce synchrotron radiation. In this paper we show that the kinetic energy dominated wind nevertheless could be directly observed through its IC radiation because of the illumination of the wind by low-energy radiation of the pulsar. The $\\gamma$-ray emission consists of two components, pulsed and unpulsed, associated with the nonthermal and thermal low-energy radiation components of the pulsar, respectively. The unpulsed component of $\\gamma$-ray emission associated with thermal radiation of the pulsar with temperature $\\simeq 2 \\cdot 10^6 \\, \\rm K$ is produced in deep the Klein-Nishina regime, and therefore has a very sharp (line-like) spectral feature which peaks at energy $E \\simeq \\gamma_w \\, \\rm m_{\\rm e} c^2$. Detection of this component would therefore result in unique information about the Lorentz-factor of the bulk motion of the wind. The nonthermal radiation of the pulsar has rather broad energy spectrum which extends to optical and infrared wavelengths, and therefore the IC $\\gamma$-ray emission associated with this component takes place, to a large extent, in the Thomson regime. This results in a broad $\\gamma$-ray spectrum with a sharp cutoff at $E \\simeq \\gamma_w \\, \\rm m_{\\rm e} c^2$. The absolute $\\gamma$-ray fluxes of both components depend strongly on the site of formation of the kinetic dominated wind, as well as the Lorentz-factor and the geometry of propagation of the wind. Thus even the flux upper limits of these $\\gamma$-ray components should provide important constraints on the wind parameters. In particular, we show that the comparison of the calculated flux of the unpulsed inverse IC emission with the measured $\\gamma$-ray flux of the Crab Nebula excludes the possibility of formation of the kinetic-energy-dominated wind within 5 light cylinder radii of the pulsar, $R_{\\rm w} \\geq 5 R_{\\rm L}$. The analysis of the pulsed IC emission, calculated under reasonable assumptions concerning the production site and angular distribution of the optical pulsed radiation, yields even tighter restrictions, namely $R_{\\rm w} \\geq 30 R_ {\\rm L}$. The mechanism of $\\gamma$-radiation of the wind of the Crab pulsar discussed in this paper should certainly take place in other pulsars as well. However, from the point of view of detection of this radiation, the Crab is a unique object due to its very powerful wind and relatively high luminosity of thermal and nonthermal low-energy radiation, which provides seed photons for the IC scattering. In other pulsars the IC $\\gamma$-ray fluxes of unshocked winds are expected to be below the detection threshold of current $\\gamma$-ray instruments, unless the kinetic energy dominated winds of pulsars are produced very close to the light cylinder. The situation could be different in binary systems containing a pulsar and luminous optical companion, the latter being an effective supplier of seed photons for IC scattering. For example, the pulsar/Be star binary system PSR 1259-63 seems to be a unique object for the search for IC TeV radiation from both shocked \\cite{kirk1} and unshocked \\cite{kirk2} winds of the pulsar. \\noindent {\\bf Acknowledgments} We thank the Astrophysics group of the MPI f\\\"ur Kernphysik, in particular, H. J. V\\\"olk, A.M. Atoyan, J.G. Kirk, as well as L. Ball and Yu. Lyubarskii for many fruitful discussions. SB thanks MPI f\\\"ur Kernphysik for warm hospitality and support during his work on this project." }, "0003/astro-ph0003361_arXiv.txt": { "abstract": "Recent observations and theoretical considerations have linked gamma-ray bursts with ultra-bright type Ibc supernovae (`hypernovae'). We here work out a specific scenario for this connection. Based on earlier work, we argue that especially the longest bursts must be powered by the Blandford-Znajek mechanism of electromagnetic extraction of spin energy from a black hole. Such a mechanism requires a high angular momentum in the progenitor object. The observed association of gamma-ray bursts with type Ibc supernovae leads us to consider massive helium stars that form black holes at the end of their lives as progenitors. In our analysis we combine the numerical work of MacFadyen \\& Woosley with analytic calculations in Kerr geometry, to show that about $10^{53}$\\,erg each are available to drive the fast GRB ejecta and the supernova. The GRB ejecta are driven by the power output through the open field lines threading the black hole, whereas the supernova can be powered both by the shocks driven into the envelope by the jet, and by the power delivered into the disk via field lines connecting the disk with the black hole. We also present a much simplified approximate derivation of these energetics. Helium stars that leave massive black-hole remnants can only be made in fairly specific binary evolution scenarios, namely the kind that also leads to the formation of soft X-ray transients with black-hole primaries, or in very massive WNL stars. Since the binary progenitors will inevitably possess the high angular momentum we need, we propose a natural link between black-hole transients and gamma-ray bursts. Recent observations of one such transient, GRO\\,J1655$-$40/Nova Scorpii 1994, explicitly support this connection: its high space velocity indicates that substantial mass was ejected in the formation of the black hole, and the overabundance of $\\alpha$-nuclei, especially sulphur, indicates that the explosion energy was extreme, as in SN\\,1998bw/GRB\\,980425. Furthermore, X-ray studies of this object indicate that the black hole may still be spinning quite rapidly, as expected in our model. We also show that the presence of a disk during the powering of the GRB and the explosion is required to deposit enough of the $\\alpha$ nuclei on the companion. \\medskip {\\it PACS codes:\\/} 98.70.Rz (Gamma-ray Bursts), 97.10.Cv (Stellar structure, Evolution, Nucleosynthesis), 97.10.Tk (Abundances), 97.60.Lf (Black holes), 95.30.Sf (Relativity and Gravitation) ", "introduction": "\\label{intro} The discovery of afterglows to gamma-ray bursts has greatly increased the possibility of studying their physics. Since these afterglows have thus far only been seen for long gamma-ray bursts (duration $\\gsim 2$\\,s), we shall concentrate on the mechanism for this subclass. The shorter bursts (duration $\\lsim 2$\\,s) may have a different origin; specifically, it has been suggested that they are the result of compact-object mergers and therefore offer the intriguing possibility of associated outbursts of gravity waves. (Traditionally, binary neutron stars have been considered in this category (Eichler et~al.\\ 1989, Janka et~al.\\ 1999). More recently, Bethe \\& Brown (1998)\\nocite{Bethe98} have shown that low-mass black-hole, neutron-star binaries, which have a ten times greater formation rate and are stronger gravity-wave emitters, may be the more promising source of this kind.) An important recent clue to the origin of long bursts is the probable association of some of them with ultra-bright type Ibc supernovae (Galama et~al.\\ 1998, Bloom et~al.\\ 1999, Galama et~al.\\ 2000)\\nocite{Galama98C,Bloom99C,Galama00A}. The very large explosion energy\\footnote{H\\\"oflich et~al.\\ (1999)\\nocite{Hoeflich99} have proposed that the explosion energy was not much larger than usual, but that the explosion was very asymmetric; this model also provides a reasonable fit to the light curve of SN\\,1998bw.} implied by fitting the light curve of SN\\,1998bw, which was associated with GRB\\,980425\\nocite{Galama98C}, indicates that a black hole was formed in this event (Iwamoto et~al.\\ 1998)\\nocite{Iwamoto98}. This provides two good pieces of astrophysical information: it implicates black holes in the origin of gamma-ray bursts, and it demonstrates that a massive star can explode as a supernova even if its core collapses into a black hole. In this paper, we start from the viewpoint that the gamma-ray burst is powered by electromagnetic energy extraction from a spinning black hole, the so-called Blandford-Znajek (1977)\\nocite{Blandford77} mechanism. This was worked out in detail by Lee, Wijers, \\& Brown (1999)\\nocite{Lee99A}, and further details and comments were discussed by Lee, Brown, \\& Wijers (2000)\\nocite{Lee99B}, who built on work by Thorne et~al.\\ (1986)\\nocite{Thorne86} and Li (2000)\\nocite{Li00}. They have shown that with the circuitry in a 3$+$1 dimensional description using the Boyer-Lindquist metric, one can have a simple pictorial model for the BZ mechanism. The simple circuitry which involves steady state current flow is, however, inadequate for describing dissipation of the black hole rotational energy into the accretion disk formed from the original helium envelope. In this case the more rapidly rotating black hole tries to spin up the inner accretion disk through the closed field lines coupling the black hole and disk. Electric and magnetic fields vary wildly with time. Using the work of Blandford \\& Spruit (2000) we show that this dissipation occurs in an oscillatory fashion, giving a fine structure to the GRB, and that the total dissipation should furnish an energy comparable to that of the GRB to the accretion disk. We use this energy to drive the hypernova explosion. Not any black-hole system will be suitable for making GRB: the black hole must spin rapidly enough and be embedded in a strong magnetic field. Moreover, the formation rate must be high enough to get the right rate of GRB even after accounting for substantial collimation of GRB outflows. We explore a variety of models, and give arguments why some will have sufficient energy and extraction efficiency to power a GRB and a hypernova. We argue that the systems known as black-hole transients are the relics of GRBs, and discuss the recent evidence from high space velocities and chemical abundance anomalies that these objects are relics of hypernovae and GRBs; we especially highlight the case of Nova Scorpii 1994 (GRO\\,J1655$-$40). The plan of this paper is as follows. We first show that it is reasonable to expect similar energy depositions into the GRB outflow and the accretion disk (Sect.~\\ref{circuit}) and discuss the amount of available energy to be extracted (Sect.~\\ref{energy}). Then we show the agreement of those results with the detailed numerical simulations by MacFadyen \\& Woosley, and use those simulations to firm up our numbers (Sect.~\\ref{compare}). We continue by presenting a simple derivation of the energetics that approximates the full results well (Sect.~\\ref{simple}). Finally, we discuss some previously suggested progenitors (Sect.~\\ref{progenitor}) and present our preferred progenitors: soft X-ray transients (Sect.~\\ref{sec7}). ", "conclusions": "\\label{conclu} Our work here has been based on the Blandford-Znajek mechanism of extracting rotational energies of black holes spun up by accreting matter from a helium star. We present it using the simple circuitry of {\\it ``The Membrane Paradigm\"} (Thorne et~al.\\ 1986)\\nocite{Thorne86}. Energy delivered into the loading region up the rotational axis of the black hole is used to power a GRB. The energy delivered into the accretion disk powers a SN Ib explosion. We also discussed black-hole transient sources, high-mass black holes with low-mass companions, as possible relics for both GRBs and Type Ib supernova explosions, since there are indications that they underwent mass loss in a supernova explosion. In Nova Sco 1994 there is evidence from the atmosphere of the companion star that a very powerful supernova explosion (`hypernova') occurred. We estimate the progenitors of transient sources to be formed at a rate of 300 GEM (Galactic Events per Megayear). Since this is much greater than the observed rate of GRBs, there must be strong collimation and possible selection of high magnetic fields in order to explain the discrepancy. We believe that there are strong reasons that a GRB must be associated with a black hole, at least those of duration several seconds or more discussed here. Firstly, neutrinos can deliver energy from a stellar collapse for at most a few seconds, and sufficient power for at most a second or two. Our quantitative estimates show that the rotating black hole can easily supply the energy as it is braked, provided the ambient magnetic field is sufficiently strong. The black hole also solves the baryon pollution problem: we need the ejecta that give rise to the GRB to be accelerated to a Lorentz factor of 100 or more, whereas the natural scale for any particle near a black hole is less than its mass. Consequently, we have a distillation problem of taking all the energy released and putting it into a small fraction of the total mass. The use of a Poynting flux from a black hole in a magnetic field (Blandford \\& Znajek 1977) does not require the presence of much mass, and uses the rotation energy of the black hole, so it provides naturally clean power. Of course, nature is extremely inventive, and we do not claim that all GRBs will fit into the framework outlined here. We would not expect to see all of the highly beamed jets following from the BZ mechanism head on, the jets may encounter some remaining hydrogen envelope in some cases, jets from lower magnetic fields than we have considered here may be much weaker and delivered over longer times, etc., so we speculate that a continuum of phenomena may exist between normal supernovae and extreme hypernovae/GRBs. This is why we call our effort ``A Theory of Gamma Ray Bursts\" and hope that it will be a preliminary attempt towards systematizing the main features of the energetic bursts." }, "0003/astro-ph0003011_arXiv.txt": { "abstract": "It is anticipated that the Pierre Auger Observatory can be used to detect cosmic neutrinos of $> 10^{19}$~eV that arrive at very large zenith angles. However showers created by neutrino interactions close to the detector must be picked out against a background of similar events initiated by cosmic ray nuclei. As a step towards understanding this background, we have made the first detailed analysis of air showers recorded at Haverah Park (an array which used similar detectors to those planned for the Auger Observatory) with zenith angles above $60^{\\circ}$. We find that the differential shower rate from $60^{\\circ}$ to $80^{\\circ}$ can be predicted accurately when we adopt the known primary energy spectrum above $10^{17}$~eV and assume the QGSJET model and proton primaries. Details of the calculation are given. ", "introduction": "\\label{intro.sec} Horizontal air showers have been studied for many years for several different reasons \\cite{Tokyo, EASTOP, CYGNUS}. Much of the relevance of horizontal air showers induced by cosmic rays is in the understanding of the background against which high energy neutrino showers could be detected. Although no neutrino events can be expected in the Haverah Park data set, even for the most optimistic neutrino flux predictions, it has recently been shown that relevant bounds could be obtained with the Auger Observatories \\cite{Auger,Capelle}. Horizontal cosmic ray showers are of great interest in their own right for two principle reasons. Firstly the acceptance of an air-shower array could be doubled if events above $60^{\\circ}$ can be adequately analysed. Secondly because, uniquely, they deal with surviving particles that are created very close to the shower core, they complement the information obtained in the study of near vertical cosmic ray showers. Moreover understanding the azimuthal asymmetries at large zenith angles [6,7,8,9,10 (hereafter Paper I)] can lead to significant improvement of ultra high energy shower analysis at moderate zenith angles ($30^{\\circ} - 60^{\\circ}$) \\cite{Dedenko}. The Haverah Park array, being made of 1.2~m deep water-\\v Cerenkov tanks \\cite{haverah}, is quite possibly the array detector so far constructed which is best suited on geometrical considerations for the analysis of very large inclined showers. Moreover it can be considered as an early prototype of the Auger Observatories which will employ water-\\v Cerenkov tanks of identical depth. The quantitative aspects of our results are very specific to the water-\\v Cerenkov technique. During the 14 years of operation of the Haverah Park array \\cite{Lawrence} which are considered here nearly 10000 air showers were recorded for $\\theta > 60^{\\circ}$. The analysis of these data was difficult because of the complex geomagnetic field effects which distort the circular symmetry of air showers, the relatively small size of the array (12~km$^{2}$) and the limited computing power then available, both for data analysis and simulation. A single remarkable event, perhaps of 10$^{20}$~eV, at $85^\\circ$ from the zenith which triggered 20 of the array detectors was extensively discussed \\cite{Hillas}, but no systematic study was made with those data for which the initial analysis gave a zenith angle $\\theta > 60^{\\circ} $. Here we present the results of the first systematic analysis of horizontal showers from primaries of energy greater than $10^{17}$~eV at Haverah Park with zenith angles exceeding $60^{\\circ}$, comparing them to those expected from the known cosmic ray spectrum. Preliminary results of this work were reported in \\cite{utah99}. For the calculation of the expected rates we make use of a parameterization for the muon number densities at ground level described elsewhere (Paper I). We take into account two possible compositions (proton and iron nuclei) and use alternative models for the high energy interactions namely QGSJET~\\cite{qgsjet} and SIBYLL 1.6~\\cite{sibyll}. % The article is organized as follows: In section \\ref{emag} we discuss the relative signals and fluctuations expected from the electromagnetic (electrons and photons) and muon components in horizontal air showers (HAS) induced by hadrons and in section \\ref{model.sec} we address the muon component of HAS. In section~\\ref{simulation} we discuss the detector simulation paying particular attention to the origin of the different parts of the signal from the individual tanks, stressing the differences between vertical and horizontal particle signals. In section \\ref{method} we describe in detail the procedure implemented to obtain a prediction for the horizontal shower rate and in section \\ref{results} we give the results obtained in the different cases. Our conclusions are presented in section \\ref{conclusions}. ", "conclusions": "\\label{conclusions} We have presented the first comparison of $>10^{17}$~eV air-shower data above $60^{\\circ}$ with simulation. The agreement is reasonable and allows us to state that the detection of HAS induced by hadronic primaries in an array of water-\\v Cerenkov tanks can be explained in terms of the muonic component. The effects of all the considered contributions to the tank signals have been shown to be extremely important in understanding the total rate. Our final results differ from the preliminary ones reported earlier \\cite{utah99} in several respects. We have taken into consideration the electromagnetic component from muon decay separately. Previously this contribution was taken together with the correction due to $\\delta-$rays. We have further improved the implementation of the PBN corrections taking into account their increase with the muon energy. We have also simulated the ground number density profiles in a completely different fashion making use of the model described in Paper I. The general shape agreement between data and simulation both for the zenith angle and azimuthal angle differential rates, together with agreement in the absolute normalization makes us confident that the geometrical considerations have been taken into account correctly, at least at the 20\\% level. This, together with the general agreement with our preliminary rate simulation, which used Monte Carlo shower simulation rather than the analytical approach described in Paper I, provides strong supporting evidence for this approach as an important tool for HAS studies. The normalization of the shower rate is sensitive to the model used for high energy hadronic interactions. For proton showers the absolute normalization of the rate differs by up to a factor of two for the two models considered. The absolute rate has also been shown to be very sensitive to the primary composition. There is about a factor 2.2 (3) enhancement, reasonably independent of zenith angle, for the differential rate normalization if the primary particles are assumed to be iron nuclei in the QGSJET (SIBYLL) model. It is interesting to note that current air-shower models are not able to reproduce the observations of high energy particles close to the shower axis above 10$^{16}$~eV \\cite{kascade}. This may be a possible explanation of the $\\sim$30\\% disagreement above 80 degrees (where the surviving muons all originate close to the shower core). These effects illustrate that HAS measurements may provide another tool with which to study both primary composition and hadronic interactions in ultra high energy cosmic rays. Future detectors such as the Auger Observatories should benefit significantly from the study of cosmic ray showers at large zenith angles. {\\bf Acknowledgements:} We thank Xavier Bertou for helping us with the angular reanalysis of the data described in \\cite{utah99} and Gonzalo Parente for suggestions after reading the manuscript. This work was partly supported by a joint grant from the British Council and the Spanish Ministry of Education (HB1997-0175), by Xunta de Galicia (XUGA-20604A98), by CICYT (AEN99-0589-C02-02) and by PPARC(GR/L40892)." }, "0003/astro-ph0003227_arXiv.txt": { "abstract": "The next generation low-frequency radio telescopes may \\hfil\\break probe cosmological models by means of observations of the cosmic microwave background (CMB). I discuss the prospects for observations of CMB imprints --- recombination lines from the epoch of recombination, $\\mu$ distortions and angular temperature anisotropies --- at low frequencies. A future low-frequency radio telescope, like the proposed SKA, may be capable of attempting some difficult CMB measurements because of the large collecting area and large element numbers; however, this will require a telescope design that will allow specialized calibration strategies and will give emphasis to the control of spurious responses. ", "introduction": "The coming decade may well see a monumental enhancement in observing capabilities at low radio frequencies if, for example, the proposed square-kilometre array (SKA) is constructed. The quantum leap in the total collecting area and in the numbers of antenna elements constituting the array could move certain key observational proposals from the realm of `dreams' to reality. In this context --- and while the telescope configuration and antenna element design are still being debated --- it is perhaps timely that key and challenging experiments that may be of relevance to the emerging opportunity are highlighted so that the instrument specifications may be tailored to eventually allow their endeavor. The discovery of the existence of the cosmic microwave background (CMB), followed by refinements in measurements of its characteristics leading to the precise $COBE$ observations of its spectrum and anisotropies, have been a major constraint and discriminant between models of cosmology and structure formation. Almost all measurements that have been valuable in constraining cosmology theories have been made to-date at cm, mm and sub-mm wavelengths. The long wavelength measurements of the CMB have been plagued by large errors owing to the bright Galactic background temperatures and high levels of extragalactic foreground discrete source confusion. In this article, I shall attempt to highlight some aspects of CMB research and discuss their relevance to low-frequency radio astronomy in the context of the changed scenario expected in the coming decade if the SKA is constructed. Possible observing strategies and calibration schemes are introduced; implications for the specifications for the SKA are touched upon. ", "conclusions": "" }, "0003/astro-ph0003294_arXiv.txt": { "abstract": "Motivated by the discovery of young, massive stars in the nuclei of some LINER/\\ion{H}{2} ``transition'' nuclei such as NGC 4569, we have computed photoionization models to determine whether some of these objects may be powered solely by young star clusters rather than by accretion-powered active nuclei. The models were calculated with the photoionization code CLOUDY, using evolving starburst continua generated by the the \\starb\\ code of \\citet{lei99}. We find that the models are able to reproduce the emission-line spectra of transition nuclei, but only for instantaneous bursts of solar or higher metallicity, and only for ages of $\\sim3-5$ Myr, the period when the extreme-ultraviolet continuum is dominated by emission from Wolf-Rayet stars. For clusters younger than 3 Myr or older than 6 Myr, and for models with a constant star-formation rate, the softer ionizing continuum results in an emission spectrum more typical of \\ion{H}{2} regions. This model predicts that Wolf-Rayet emission features should appear in the spectra of transition nuclei. While such features have not generally been detected to date, they could be revealed in observations having higher spatial resolution. Demographic arguments suggest that this starburst model may not apply to the majority of transition nuclei, particularly those in early-type host galaxies, but it could account for some members of the transition class in hosts of type Sa and later. The starburst models during the Wolf-Rayet-dominated phase can also reproduce the narrow-line spectra of some LINERs, but only under conditions of above-solar metallicity and only if high-density gas is present ($n_e \\gtrsim 10^5$ \\percucm). This scenario could be applicable to some ``Type 2'' LINERs which do not show any clear signs of nonstellar activity. ", "introduction": "Emission-line nebulae in galactic nuclei are generally considered to fall into three major categories: star-forming or \\ion{H}{2} nuclei, Seyfert nuclei, and low-ionization nuclear emission-line regions, or LINERs. The formal divisions between these classes are somewhat arbitrary, as the observed emission-line ratios of nearby galactic nuclei fall in a continuous distribution between LINERs and Seyfert nuclei and between LINERs and \\ion{H}{2} nuclei \\citep[e.g.,][]{hfs93}. Traditionally, LINERs have been defined as those nuclei having emission-line flux ratios which satisfy the relations [\\ion{O}{2}] \\lam3727/[\\ion{O}{3}] \\lam5007 $> 1$ and [\\ion{O}{1}] \\lam6300/[\\ion{O}{3}] \\lam5007 $> 1/3$ \\citep{h80}. It is possible to construct alternative but practically equivalent definitions, based on other line ratios, which can be applied to datasets that do not include the wavelengths of the [\\ion{O}{1}], [\\ion{O}{2}], or [\\ion{O}{3}] lines \\citep[e.g.,][]{hfs97a}. A sizeable minority of galactic nuclei has emission-line ratios which are intermediate between those of ``pure'' LINERs and those of typical \\ion{H}{2} regions powered by hot stars; these galaxies would be classified as LINERs except that their [\\ion{O}{1}] \\lam6300 line strengths are too small in comparison with other lines to meet the formal LINER criteria. Objects falling into this category have been dubbed ``transition'' galaxies by \\citet{hfs93}, and although this nomenclature is somewhat ambiguous we adopt it here for consistency with the spectroscopic survey of \\citet{hfs97a}. That survey defined the transition class in terms of the following flux ratios: \\begin{center} {[\\ion{O}{3}]} \\lam5007/\\hbeta\\ $<$ 3, \\\\ 0.08 $\\leq$ [\\ion{O}{1}] \\lam6300/\\hal\\ $<$ 0.17, \\\\ {[\\ion{N}{2}]} \\lam6583/\\hal\\ $\\geq$ 0.6, \\\\ {[\\ion{S}{2}]} \\lamlam6716, 6731 /\\hal\\ $\\geq$ 0.4. \\end{center} \\citet{ft92} have used the term ``weak-[\\ion{O}{1}] LINERs'' to refer to galaxies having [\\ion{N}{2}] \\lam6583/\\hal\\ $\\gtrsim 0.6$ (typical of LINERs) but which have [\\ion{O}{1}] \\lam6300/\\hal\\ $< 1/6$. This category is essentially identical to the transition class of \\citet{hfs93,hfs97a}, and we will refer to these galaxies as transition objects in this paper. According to the survey results of \\citet{hfs97b}, this transition class accounts for 13\\% of all nearby galaxies, making them about as numerous as Seyfert nuclei. The Hubble type distribution of transition galaxies is intermediate between that of LINERs, which are most common in E/S0/Sa galaxies, and that of \\ion{H}{2} nuclei, which occur most often in Hubble types later than Sb \\citep{hfs97b}. Roughly $20\\%$ of galaxies with Hubble types ranging from S0 to Sbc belong to the transition class. There is not a consensus, however, as to whether these transition objects should be regarded as star-forming nuclei, as accretion-powered active nuclei, or as composite objects powered by an AGN and by hot stars in roughly equal proportion. There is a large body of literature on the subject of the excitation mechanism of LINERs which is relevant to the similar transition class. A variety of physical mechanisms has been proposed to explain the emission spectra of LINERs, including shocks, photoionization by a nonstellar ultraviolet (UV) and X-ray continuum, and photoionization by hot stars. (See Filippenko 1996 for a review.) The possibility that LINERs (and Seyfert nuclei as well) might be photoionized by starlight was raised by \\citet{tm85}, who suggested that very hot ($\\teff \\sim 10^5$ K) Wolf-Rayet (W-R) stars in a metal-rich starburst could give rise to an ionizing continuum with a nearly power-law shape in the extreme-UV. More recent atmosphere models have indicated substantially lower temperatures for W-R stars, however, casting doubt on the Warmer hypothesis \\citep{lgs92}. Subsequent photoionization models have attempted to explain LINER and transition-type spectra as resulting from massive main-sequence stars. \\citet{ft92} found that the spectra of weak-[\\ion{O}{1}] LINERs could be explained in terms of photoionization by O3--O4 stars having effective temperatures of $\\gtrsim45,000$ K, at ionization parameters of $U \\approx 10^{-3.7}$ to $10^{-3.3}$. \\citet{shi92} carried this line of argument farther, proposing that genuine LINER spectra could be generated by early O stars with $\\teff \\approx 50,000$ K, provided that a high-density component ($n_e \\approx 10^{5.5}$ \\percucm) is present in the NLR; the high densities are needed to boost the strengths of high critical-density emission lines, most notably [\\ion{O}{1}] \\lam6300. Similar conclusions were reached by \\citet{sf94}, who explored the effects of absorption by ionized gas as a means to harden the effective ionizing spectrum. Recent observations, particularly in the UV and X-ray bands, have provided convincing evidence that many LINERs are in fact AGNs, particularly the ``Type 1'' LINERs which have a broad component to the \\hal\\ emission line \\citep[for a recent review see][]{ho99}. The possibility has remained, however, that some LINERs and transition nuclei are powered entirely by bursts of star formation. An important shortcoming of the model calculations performed by \\citet{ft92} and \\citet{shi92} is that the ionizing continua used as input were those of single O-type stars; these studies did not address the question of whether a LINER or transition-type spectrum could result from the the \\emph{integrated} ionizing continuum of a young stellar cluster. Compared with these single-star models, the contribution of late-O and B stars will soften the ionizing spectrum, making the emission-line ratios tend toward those of normal \\ion{H}{2} regions. W-R stars, on the other hand, will harden the ionizing spectrum during the period when these stars are present, roughly $3-6$ Myr after the burst. Another drawback of the O-star models is that they require the presence of stars with effective temperatures higher than are thought to occur in \\ion{H}{2} regions of solar or above-solar metallicity, in order to produce a LINER or transition-type spectrum rather than an \\ion{H}{2} region spectrum. Their applicability to galactic nuclei is therefore somewhat unclear. Other mechanisms have been proposed for generating LINER or transition-type spectra. Shock excitation by supernova remnants in an aging starburst may give rise to some transition objects; the nucleus of NGC 253 is a likely candidate for such an object \\citep{eng98}. Also, post-AGB stars and planetary nebula nuclei will produce a diffuse ionizing radiation field which could be responsible for the very faint LINER emission (with \\hal\\ equivalent widths of $\\sim1$ \\AA) observed in some ellipticals and spiral bulges \\citep{bin94}. An alternate possibility is that the transition galaxies may simply be composite systems consisting of an active nucleus surrounded by star-forming regions. For a galaxy at a distance of 10 Mpc, for example, a 2\\arcsec-wide spectroscopic aperture will include \\ion{H}{2} regions within 50 pc of the nucleus. Galaxies having emission lines both from a LINER nucleus and from surrounding star-forming regions, in roughly equal proportions, will appear to have a transition-type spectrum. This interpretation was advocated by \\citet{hfs93} as the most likely explanation for the majority of transition galaxies, and is consistent with the observed Hubble type distribution for the transition class. Other authors have similarly contended that transition galaxies are AGN/\\ion{H}{2} region composites, based on optical line-profile decompositions \\citep{vgv97,gvv99} and near-infrared spectra \\citep{hil99}. Two of the 65 transition nuclei observed in the \\citet{hfs97b} survey have a broad component to the \\hal\\ emission line, indicating the likely presence of an AGN, and it is probable that many more transition nuclei contain obscured AGNs which were not detected in the optical spectra. On the other hand, radio observations do not appear to support the composite AGN/starburst interpretation. In a VLA survey of nearby galactic nuclei, Nagar \\etal\\ (1999) find compact, flat-spectrum radio cores in more than 50\\% of LINER nuclei, but in only 6\\% (1 of 18) of transition objects. This discrepancy suggests that the simple picture of an ordinary LINER surrounded by star-forming regions may not apply to the majority of transition objects. Recent results from the \\emph{Hubble Space Telescope} (\\hst) have shed new light on the question of the excitation mechanism of transition nuclei. As shown by \\citet{mao98}, the UV spectrum of the well-known transition nucleus in NGC 4569 over 1200-1600 \\AA\\ is virtually identical to that of a W-R knot in the starburst galaxy NGC 1741, indicating that O stars with ages of a few Myr dominate the UV continuum. \\citet{mao98} find that the nuclear star cluster in NGC 4569 is producing sufficient UV photons to ionize the surrounding narrow-line region, a key conclusion which provides fresh motivation to study stellar photoionization models. The brightness of the NGC 4569 nucleus, and the consequently high S/N observations that have been obtained, make it one of the best objects with which to study the transition phenomenon. The recent availability of the \\starb\\ model set \\citep{lei99} has prompted us to reexamine the issue of ionization by hot stars in LINERs and transition nuclei. These models give predictions for the spectrum and luminosity of a young star cluster, for a range of values of cluster age, metal abundance, and stellar initial mass function (IMF) properties. Using the photoionization code CLOUDY \\citep{fer98} in combination with the \\starb\\ model continua, we have calculated the expected emission-line spectrum of an \\ion{H}{2} region illuminated by a young star cluster, to test the hypothesis that some LINERs and transition nuclei may be powered by starlight. Similar calculations have been performed by \\citet{sl96}, but for the physically distinct case of metal-poor objects representing \\ion{H}{2} galaxies. Other examples of photoionization calculations for \\ion{H}{2} regions using evolving starburst continua are presented by \\citet{gd94}, \\citet{gbd95}, and \\cite{bkg99}. ", "conclusions": "\\label{sectionconclusions} Our primary conclusion is that for standard starburst parameters and for nebular conditions which may be typical of galactic nuclei, the starburst models are able to reproduce the important diagnostic emission-line ratios for LINER/\\ion{H}{2} transition galaxies, otherwise known as weak-[\\ion{O}{1}] LINERs. The key ingredient needed to generate a transition-type spectrum is a UV continuum dominated by W-R stars, a condition which occurs during $t =$ 3--5 Myr after an instantaneous burst. A transition-type emission spectrum may thus be a phase in the evolution of some nuclear \\ion{H}{2} regions in which the ionizing continuum is generated by a single-burst stellar population. The models are also able to produce an [\\ion{O}{1}] / \\hal\\ ratio high enough to match LINER spectra, but only for conditions of above-solar metallicity combined with the presence of high-density ($\\gtrsim10^5$ \\percucm) clouds. A sensitive search for W-R spectral features in transition nuclei would provide a test of this starburst scenario. This model may apply only to a small fraction of LINERs and transition nuclei; many LINERs and some transition objects show clear signs of nonstellar activity, and the starburst models may not apply at all to objects in early-type host galaxies. Further multiwavelength observations of transition nuclei will be of great utility for determining what fraction of them contain genuine active nuclei, and what fraction appear to be purely the result of stellar phenomena." }, "0003/astro-ph0003407_arXiv.txt": { "abstract": "Gamma rays in the band from 20 MeV to 300 GeV, used in combination with data from radio and X-ray bands, provide a powerful tool for studying the origin of cosmic rays in our sister galaxies Andromeda and the Magellanic Clouds. Gamma-ray Large Area Space Telescope (GLAST) will spatially resolve these galaxies and measure the spectrum and intensity of diffuse gamma radiation from the collisions of cosmic rays with gas and dust in them. Observations of Andromeda will give an external perspective on a spiral galaxy like the Milky Way. Observations of the Magellanic Clouds will permit a study of cosmic rays in dwarf irregular galaxies, where the confinement is certainly different and the massive star formation rate is much greater. ", "introduction": "High-energy gamma rays are produced in interactions of high-energy cosmic rays with interstellar matter and photons. From the resulting diffuse emission of gamma rays, the properties of the cosmic rays can be inferred (e.g., \\cite{hunter97}). Gamma rays have proven to be a useful probe of cosmic rays in the Milky Way, but gamma-ray telescopes to date have lacked the sensitivity and angular resolution to permit the same kind of detailed study of cosmic rays in external galaxies. The Gamma-ray Large Area Space Telescope (GLAST) is the next generation high-energy (20 MeV--300 GeV) gamma-ray astronomy mission. It is part of the strategic plan of NASA's Office of Space Science and is currently planned for launch in 2005. GLAST will have a factor of 30 greater sensitivity than the Energetic Gamma-Ray Experiment Telescope (EGRET), launched in 1991 on the Compton Gamma-Ray Observatory. Derived performance parameters for our proposed design for the GLAST instrument, which was selected by NASA in February 2000, are presented in Table~\\ref{table1} and Figure~\\ref{fig1}. See the companion paper by Ormes {\\it et al.} for information about the design and instrumental response of GLAST, and the Web site http://glast.gsfc.nasa.gov for information about the mission. \\begin{table} \\begin{tabular}{lll} \\hline \\tablehead{1}{r}{b}{ } &\\tablehead{1}{c}{b}{EGRET} &\\tablehead{1}{c}{b}{GLAST} \\\\ \\hline Energy Range &\t0.02--30 GeV &\t0.02--300 GeV \\\\ Field of View &\t0.5 sr &\t2.4 sr \\\\ Peak Eff. Area\t& 1500 cm$^{2}$ &\t13,000 cm$^{2}$ \\\\ Point Source\\\\Sensitivity\\tablenote{Sensitivity at high latitude after a 2-year survey for a 5-$\\sigma$ detection, units 10$^{-8}$ cm$^{-2}$ s$^{-1}$, for $E > 100$ MeV.} & 5 & 0.16 \\\\ Source Location\\tablenote{Diameter of 95\\% confidence region; range: bright sources to sources of flux 10$^{-8}$ cm$^{-2}$ s$^{-1}$ ($E > 100$ MeV).} & 5$^{\\prime}$ -- 90$^{\\prime}$\t& 0.2$^{\\prime}$ -- 1$^{\\prime}$ \\\\ Mission Life & &\t\t5 years\\\\ & & (10-year goal) \\\\ \\hline \\end{tabular} \\caption{Selected Parameters for GLAST and EGRET} \\label{table1} \\end{table} \\begin{figure}[!t] \\includegraphics[scale=0.42]{glast_fig1_small} % \\caption{Expected performance of GLAST for localizing and resolving point sources. ($a$) Source localization at high latitudes. The position uncertainties for the brightest sources likely will be limited to 10--20$^{\\prime\\prime}$ by uncertainty in spacecraft pointing and instrument alignment. ($b$) Minimum flux required to resolve two closely-spaced sources of equal flux. For both figures, the sources are assumed to have $E^{-2}$ photon spectra and to be observed at high latitudes in a one-year sky survey. \\label{fig1}} \\end{figure} \\begin{figure}[!th] \\includegraphics[scale=0.60]{glast_fig2_small} % \\caption{$(a)$ Simulated map of the LMC in gamma rays (> 100 MeV) from a two-year sky survey with GLAST. The simulation is based on a model of the LMC by Sreekumar \\cite{kumar99} and also includes foreground diffuse emission from the Milky Way and an isotropic background consisting of a distribution of faint point sources. $(b)$ The LMC in $3/4$-keV X-rays, from a mosaic of pointed observations with ROSAT \\cite{snowden94}. The intense emission regions of 30 Doradus and LMC superbubble 2 are indicated. \\label{fig2}} \\end{figure} ", "conclusions": "For the first time, GLAST will enable spatial and spectral studies of diffuse gamma rays from external galaxies. Diffuse, high-energy gamma rays are diagnostic of cosmic-ray densities, which especially for the proton component are difficult to determine from observations at other wavelengths. When considered together with X-ray and radio observations, GLAST data promise a fairly complete understanding of the production, propagation, and confinement of cosmic rays in Local Group galaxies. IVM acknowledges support from an NAS/NRC Senior Associateship." }, "0003/astro-ph0003238_arXiv.txt": { "abstract": "We present images and high-resolution spectra of the planetary nebulae IC~2553 and NGC 5882. Spatio-kinematic modeling of the nebulae shows that they are composed of a markedly elongated inner shell, and of a less aspherical outer shell expanding at a considerably higher velocity than the inner one. Embedded in the outer shells of both nebulae are found several low-ionization knots. In IC~2553, the knots show a point-symmetric distribution with respect to the central star: one possible explanation for their formation is that they are the survivors of pre-existing point-symmetric condensations in the AGB wind, a fact which would imply a quite peculiar mass-loss geometry from the giant progenitor. In the case of NGC~5882, the lack of symmetry in the distribution of the observed low-ionization structures makes it possible that they are the result of { in situ} instabilities. ", "introduction": "Many planetary nebulae (PNe) are known to possess intriguing low-ionization small-scale structures which usually appear in the form of low-ionization knots, bullets, filaments, ansae, etc., embedded in the main bodies of the nebulae or outside them. These structures have received a great deal of attention in the last years (e.g., Balick et al. 1998, and references therein) because they might provide important insights into the processes governing PN formation and evolution, such as the role of dynamical instabilities and clumpiness, the collimation mechanisms of the mass outflow from single and binary AGB stars, or the occurrence of discrete mass-loss episodes in the post-AGB phase. Presently, we are still quite far from understanding their origin (Balick et al. 1998). We are carrying out an observational program aimed at investigating the properties of these low-ionization microstructures. Most of our targets were selected by Corradi et al. (1996) by computing (\\nii+\\ha)/\\oiii\\ ratio maps in the image catalogs of Schwarz, Corradi, \\& Melnick (1992) and Gorny et al. (1999). In previous papers (Corradi et al. 1997, 1999, 2000a), we have reported the results for several PNe, highlighting the occurrence of extended highly collimated structures (e.g., NGC 3918 and K~1-2), high-velocity symmetrical knots (K~4-47), precessing outflows (NGC 6337 and He~2-186), and multiple collimated ejecta (IC~4593 and Wray 17-1). From this work, it appears that different dynamical/radiative processes are needed to explain the variety of low-ionization structures observed. In this paper, we discuss the results for IC~2553 and NGC~5582, which present an additional type of low-ionization structures: systems of knots located in the outer shells of the nebulae. IC~2553 and NGC~5582 are two southern PNe which have not been extensively studied in the past, apart from several statistical studies which are referenced throughout the paper. A chemical study of the different morphological structures of NGC~5882 was recently presented by Guerrero \\& Manchado (1999). In this paper we present high-resolution long-slit spectra and narrow-band images which allow us to determine the geometry and kinematics of the main shells of IC~2553 and NGC~5582, with particular emphasis on the low-ionization structures contained therein. ", "conclusions": "The origin of the low-ionization microstructures which are observed in a large number of PNe is intriguing. From the data collected in this paper and in previous ones (see \\S\\ 1), it turns out that they appear with so wide a variety of morphological, kinematic, physical and topographical properties, that one can hardly think of a common physical mechanism able to account for all of them. In this article, we have focused our attention on the nature of low-ionization knots in the outer shells of PNe (see also Balick et al. 1998). Being found within nebular regions which, supposedly, have not yet been affected by the action of the fast AGB wind which shapes the bright rims of PNe, some of the mechanisms proposed to explain their origin can be excluded (such as the occurrence of instabilities in the fast vs. slow winds interaction). One attractive option is that they are the product of instabilities in the interaction of the AGB wind with the ionization front, the ISM, or previous AGB ejecta. This might well apply in the case of NGC~5882, but the point symmetry shown by the knots in IC~2553 indicates instead that they are the remnants of pre-existing, symmetric condensations in the AGB wind. Clumpy AGB winds have been observed in molecular lines (Olofsson et al. 2000), but the additional requirement of a point-symmetric distribution of the clumps would imply a very specific mass-loss geometry from the progenitor star. Understanding the physical processes causing this kind of mass loss in AGB stars clearly deserves further observational and theoretical study." }, "0003/astro-ph0003462_arXiv.txt": { "abstract": "One of a family previously proposed ``central engines'' for cosmic gamma-ray burst sources (Klu\\'zniak \\& Ruderman 1998) is considered in some detail. A steadily accreting $10^6$ Gauss magnetic white dwarf should ultimately collapse to a strongly differentially rotating, millisecond-rotation-period neutron star for a wide range of steady accretion rates and initial masses if the accreting white dwarf has an evolved O-Ne-Mg composition. A similar neutron star could also result from an initial C-O white dwarf but only for more constrained accretion rates. Because the collapsing white dwarf begins as a $\\gamma=4/3$ polytrope, the final neutron star's spin-rate increases strongly with cylindrical radius. A stable wind-up of the neutron star's poloidal magnetic field then produces buoyant magnetic toroids which grow, break loose, rise, and partly penetrate the neutron star surface to form a transient, $B\\approx 10^{17}$ G millisecond spin-period pulsar with a powerful pulsar wind (Usov 1992). This pulsar wind emission is then rapidly suppressed by the surface shear motion from the strong stellar differential rotation. This wind-up and transient pulsar formation can occur at other times on different cylinders and/or repeat on the same one, with (re-)wind up and surface penetration time scales hugely longer than the neutron star's millisecond spin period. In this way, differential rotation both opens and closes the doors which allow neutron star spin-energy to be emitted in powerful bursts of pulsar wind. Predictions of this model compare favorably to needed central engine properties of gamma-ray burst sources (total energy, duration, sub-burst fluctuations and time scales, variability among burst events, and baryon loading). ", "introduction": "Gamma-ray bursts (GRBs) are observed daily from sources at distances extending out to those of the oldest galaxies in our Universe. To account for details of these bursts, ``central engines'' (CEs) of the GRB sources should have the following properties (see Klu\\'zniak \\& Ruderman 1998, hereafter KR, for details and references). (a) {\\it Energy.} Some CEs must store and release of order $10^{53}$ ergs (assuming modest beaming of the energy outflow). (b) {\\it Fluctuations.} There are often large temporal variations in the CE power output. A CE should be capable of attaining peak power within tens of milliseconds and exhibiting large fluctuations thereafter. The main power emission is often in sub-bursts between which the CE is relatively dormant, typically for about $10$ seconds, but sometimes for as long as several $10^2$ seconds or as short as $10^{-1}$ seconds. (c) CE {\\it lifetimes}, typically seconds to tens of seconds, extend from less than a second to greater than $10^3$ seconds. (There is also some indication of an association of greater total energy release with longer CE lifetimes.) (d) {\\it Baryon loading.} The energy released from the CE of a GRB source carries with it at most only a tiny baryon load of mass $\\la 10^{-4} M_\\odot$. (e) The {\\it birth rate} of GRB sources $\\ga 10^{-6}$/galaxy/yr (see, for example, B\\\"ottcher \\& Dermer 2000). (f) There is a very great variability among observed GRB events: durations, time scales within a burst, and pulse shape structures, sub-burst numbers, etc., vary so much that one cannot really specify a typical GRB. The shortest time scales of (b) together with the total energy emission (a) suggest a CE formation involving stellar collapse to a neutron star or to a black hole, or a very tight binary of such collapsed objects, or as part of some exotic supernova which would form such objects. However, the lifetimes (c), baryon loading (d), the commonly observed repeated widely separated fluctuations (b), and perhaps the birthrate (e) may raise special problems for such CE models. Particularly significant is why, if the CEs are collapsed objects whose periods of rotation and vibration are expected to be milliseconds, energy emission from them so often involves several timescales which can be up to $10^6$ times longer. A promising way of constructing CE models based upon collapsed objects, which incorporates this needed family of relatively long time scales, begins by converting the most of the released collapse energy into rotational energy of the collapsed objects. The subsequent transfer of that energy to emitted power in a form useful for ultimate $\\gamma$-ray production may then be accomplished relatively slowly. It is generally necessary to have CE magnetic fields $B\\ga10^{15}$ G to extract the rotationally stored energy fast enough. Such a CE model was long ago proposed by Usov (1992). A millisecond spin-period pulsar with a magnetic field $B \\approx 10^{15}$ G was assumed to be formed from an accretion induced collapse of a strongly magnetized ($B \\approx 10^{9}$ G) white dwarf. This simple CE model would be expected to have the needed energy (a), lifetime (c) and baryon loading (d) properties, but a sufficiently high birthrate (e) may be questionable and the required fluctuation property (b) does not seem to be realized. It has been proposed more recently that very large differential rotation plays an essential role in CE models (KR). One such model has significant similarities to Usov's proposed millisecond-period ``magnetars'', but the initial white dwarf's pre-collapse history and magnetic field strengths differ, and there are essential differences in what happens within the neutron star and on its surface. This strongly differentially rotating CE would form and evolve in the following, quite different, way. 1) A common ``garden-variety'' magnetic white dwarf ($B \\approx 10^6$ G) in a tight binary is spun up to its equilibrium spin-period ($P \\approx 10^3$ s) by an accretion disk fed by its companion. 2) The accreting white dwarf is either an evolved one (O-Ne-Mg), or a canonical (C-O) dwarf, with accretion rates such that the accreting white dwarf increases its mass, implodes before its growing stellar mass reaches $1.4 M_\\odot$, and collapses to a neutron star. 3) A neutron star is then formed with an initial spin-period $P \\approx 10^{-3}$ seconds, a nearly canonical pulsar polar magnetic surface dipole component $B_p \\approx 10^{12}$ G, and, most importantly, a spin-rate which increases very greatly with distances from the star's spin-axis. It is this crucial last feature which is the reason for choosing here to discuss this particular CE model from among the previously suggested possibilities for CEs with large initial differential rotation (KR). 4) An interior toroidal field ($B_\\phi$) is then stably wound up from the poloidal field ($B_p$) by this differential rotation until $B_\\phi \\approx 10^{17}$ G. After that $B_\\phi$ is achieved, the wound-up (and probably slightly twisted) toroid's magnetic buoyancy for the first time exceeds interior anti-buoyancy forces (from compositional stratification). The buoyant toroid pushes up to the surface by moving parallel to the spin-axis up to, and then partly penetrating the stellar surface, within about $10^{-2}$ seconds after its initial release. 5) For as long as some of this magnetic field sticks out of the rapidly spinning neutron star's surface, this will be an extreme realization of an Usov pulsar, a hyper-magnetar powered by the star's spin energy. It is, however, extremely transient because of surface movements. 6) This surface dipole field (and higher multipoles) can survive for only a very brief time ($\\approx 10^{-2}$ seconds): it is continually smeared out around the spin-axis, and thus diminished by the strong on-going differential rotation shearing the surface below any protruding field. (There may also be considerable surface field reconnection after this.) 7) After the first break-out of wound-up toroidal field, surface penetration by some of it, and the resulting transient Usov pulsar, a similar wind-up of the $B_p \\approx 10^{12}$ G may begin again as in Step 4, around the same cylinder or a somewhat slower wind-up may exist on some other cylinder). In tiehr case, a new toroid grows until its $B_\\phi$ reaches $B_\\phi(max) \\approx 10^{17}$ G when another sub-burst occurs as in Steps 5-6. The characteristic interval between the first and second sub-burst would be \\begin{equation} \\tau_{sb} \\approx \\frac{2\\pi B_\\phi(max)}{(\\Delta\\Omega) B_p} \\approx 10 \\, \\, {\\rm seconds}, \\label{eq:subbursttime} \\end{equation} where $\\Delta\\Omega$ is the spin-frequency difference between the inner and outer parts of the differentially rotating neutron star. 8) The GRB source's CE finally turns off completely when either of two stages is reached by the engine: a) the differential rotation ($\\Delta\\Omega$) which drives the wind-up of $B_\\phi$ becomes so diminished by the conversion of the differential rotation energy into toroidal field energy that it can no longer cause build-up to the critical $B_\\phi\\approx 10^{17}$ G needed for a pulsar wind sub-burst, or b) the stellar spin ($\\Omega$ of the outer region) becomes so reduced in the transient pulsar phases sustained by it that pulsar wind emission is almost extinguished even if a huge protruding field were to survive. In this present note, we consider the above GRB source CE proposal in more detail and discuss why and how it should have all of the desired properties. ", "conclusions": "The required properties of GRB source CEs (summarized in the Introduction) are total energy stored and emitted (a), peak power and fluctuations within a given burst event (b), CE lifetimes (c), maximum baryon loading in the CE emission (d), CE birthrate (e) and very large variability among different CEs (f). None of these seem an embarrassing problem for the proposed model CE genesis, structure and dynamics outlined in the Introduction and described in Sections 2-9. Indeed each seems a rather expected consequence. However, a crucial point which should be considered further is the absence (so far) of any demonstrated instability in the wind-up of the toroidal field for of order $10^4$ turns (in about $10$ seconds) by the much more energetic initial differential rotations in the neutron star. A second related, but less crucial, question is the robustness of our presumption that during and after such toroidal wind up and release the initial much smaller poloidal field component of the differentially rotating neutron star is not hugely increased. If this does not turn out to be an adequate approximation, the often observed sub-burst multiplicity could still come from toroidal field wind-up and break-away in different cylindrical regions with different wind-up times rather than from long time delays for rewinding $B_\\phi$. Of course, because of the very great variability within the family of GRB events, neither mechanisms may hold in all, or perhaps not even in most cases, but at least one of them should certainly not be uncommon. Finally there is our unproven assumption % that large toroidal field bundles wound-up by differential rotation can overcome anti-buoyancy restraints and break free as a unit (or almost so). If, instead, buoyant toroidal field continually dribbled up and out to support a steady state in which increasing $B_\\phi$ from wind-up is balanced by a that loss, there would be no strong fluctuations in CE output. Instead a CE would be an Usov-like pulsar with emission decreasing monotonically after the first emission maximum is reached. This is a generic problem for many kinds of CE models. Why does the CE depart so far from a steady equilibrium that stored energy is released in huge sub-bursts (which are often separated by very many characteristic engine periods) rather than in smoother continuous steady way? Here too such a question needs further investigation." }, "0003/astro-ph0003148_arXiv.txt": { "abstract": "NGC 2355 is an old open cluster in the outer part of the galactic disk ($l=203\\fdg4$, $b=+11\\fdg8$) which has been little studied until now. This paper presents the first astrometric and spectroscopic investigation of this cluster. We have measured precise absolute proper motions from old Carte du Ciel plates, POSS-I plates and recent CCD observations obtained with the Bordeaux meridian circle. The proper motion data reveal 38 highly probable cluster members down to $\\Blim = 15$~mag within $7'$ of the cluster center. We have also obtained ELODIE high resolution spectra for 24 stars. Seventeen of them are confirmed to be members of the cluster on the basis of radial velocity. Eight of them are fast rotating turnoff stars for which the projected rotational velocity has been determined. The spectroscopic observations have also provided estimates of the physical parameters $\\Teff, \\logg, \\FeH, \\Mv $ of the 24 target stars. Two stragglers have been identified in the cluster. Combining our astrometric and spectroscopic results with previous UBV photometry and recent JHK$_s$ photometry from the 2MASS survey we have derived the fundamental properties of the cluster: metallicity, age, distance, size, spatial velocity and orbit. ", "introduction": "Old open clusters, with ages greater than the age of the Hyades ($\\sim$ 600 Myr), represent a minority of about 80 objects among 1200 known open clusters. Among their properties which enable to investigate both stellar physics and galactic structure (reviewed by Friel 1995), we are especially interested in orbits because they are related to the processes which have allowed them to survive tidal forces. The statistics are still poor but it seems that old open clusters follow orbits that keep them away from the plane and the disruptive effects of giant molecular clouds. The question is to know if these orbits result from special events or represent the tail of the distribution of clusters that have already been destroyed. Another relevant point to clarify is the relationship between orbits and metallicity [Fe/H] which traces the dynamical and chemical evolution of the Galaxy. The metallicity of old open clusters is intermediate between the disk and the thick disk, with a radial gradient, but a large dispersion that could indicate an inhomogeneous enrichment of the Galaxy. To answer such fundamental questions, new observations are needed to investigate in more details the old open cluster properties and their correlations. We have therefore undertaken a spectroscopic and astrometric program to obtain metallicities, distances and velocities of high quality for several poorly known old open clusters, NGC 2355 being our first target.\\ There are very few references on NGC 2355 in the astronomical literature. A photometric study in UBV down to $V\\sim 19.2$ was made by Kaluzny \\& Mazur (1991). In this study, the reddening of the cluster was estimated to be $E_{B-V}$=0.12 mag, the distance modulus $(m-M)_0=12.1$, the metallicity +0.13 and the age the same as Praesepe. In their search for old open clusters, Phelps et al. (1994) also report a photometric study of NGC 2355 in BV but the photometry of individual stars is not given. Their calibration of the index $\\delta V$, defined as the magnitude difference between the main-sequence turnoff and the giant clump leads to a Morphological Age Index corresponding to 0.9~Gyr, like Praesepe (Janes \\& Phelps 1994). More recently, Ann et al. (1999) examined this cluster as part of the BOAO survey (Bohyunsan Optical Astronomy Observatory, Korea) and determined from UBVI photometry : $\\FeH=-0.32, E_{B-V}=0.25, (m-M)_0=11.4$ and an age of 1 Gyr.\\ In Sect. 2 and 3 we present new data which are used to analyse the cluster in combination with the UBV photometry of Kaluzny \\& Mazur (1991) and the JHK$_s$ photometry which is available for the whole field in the 2MASS 1999 Spring Incremental Data Release. We describe the determination and analysis of proper motions from photographic plates and recent observations at the meridian circle of Bordeaux (Sect. 2). For 24 bright stars ($V \\le 13$) around the cluster's center, spectra were obtained with the echelle spectrograph ELODIE on the 193cm telescope at the Haute-Provence Observatory. The radial velocities of the red giants were obtained by standard on-line reduction directly at the telescope. The determination of the radial velocity and projected rotational velocity of the hot fast rotating turnoff stars required dedicated reduction tools (Sect. 3). In Sect. 4 we present our analysis of the spectra to estimate the atmospheric parameters $\\Teff, \\logg, \\FeH $ and the absolute magnitude $\\Mv$. For the latter, we developed a new version of the TGMET method (Katz et al. 1998 and Soubiran et al. 1998). We discuss the case of an unusual giant in NGC 2355 which is 2.3 magnitudes brighter than the giant clump for the same temperature. We also report the discovery of a blue straggler in the cluster and of a moving pair of field stars. Sect. 5 deals with the fundamental parameters of NGC 2355 resulting from our study. Our conclusions are reviewed in Sect. 6. For identifying individual stars we use as far as available the star numbers introduced by Kaluzny \\& Mazur (1991), preceded by the prefix \"KM\". ", "conclusions": "We have presented a detailed study of stars in the region of NGC 2355, combining new astrometric and spectroscopic data with recent photometric data from other sources. Our main results can be summarised as follows : - NGC 2355 is at 1.65~kpc of the Sun and 340~pc above the galactic plane in the direction of the anticenter, with a reddening of $E_{B-V}=0.16$ and $E_{V-K}=0.42$. - Its metallicity is $\\FeH = -0.07 \\pm 0.11$ and its age is 1 Gyr. - NGC 2355 has a core radius of about 0.7~pc, a central component with a radius of 3.3~pc and a halo out to 7.2~pc from the cluster center. - The turnoff stars of NGC 2355 are fast rotators, with a mean projected rotational velocity of $100\\kms$ and a mean $\\Teff$ of 7500 K. - The giant clump is well defined at $\\Teff = 5000 K$, $\\Mv = 0.51$. - Two stragglers have been identified in the cluster: a blue one, and a giant which has an unusual position in the HR diagram, 2.3 mag brighter than the giant clump. - NGC 2355 has a galactocentric space velocity vector $(U,V,W) = (-23.5, +206.2, -4.2)\\kms$ and an orbit which keeps it beyond the solar circle and with only brief passages through the galactic plane. As a by-product of the study of the cluster, we found a moving pair of field giants with a radial velocity of $50\\kms$." }, "0003/astro-ph0003132_arXiv.txt": { "abstract": "Absorption features from the Lyman and Werner bands of interstellar molecular hydrogen were recorded by the Interstellar Medium Absorption Profile Spectrograph (IMAPS) at $\\lambda/\\Delta\\lambda=80,000$ in the spectra of $\\delta$~Ori~A and $\\epsilon$~Ori. The objective was to find and study more examples of an unusual phenomenon found for one of the velocity components of H$_2$ in the spectrum of $\\zeta$~Ori by Jenkins \\& Peimbert (1997). Specifically, they detected a gradual shift in velocity and broadening for features arising from progressively higher rotational excitations $J$. This effect appears to be absent in the spectra of both $\\delta$ and $\\epsilon$~Ori, which are only a few degrees away in the sky from $\\zeta$~Ori. The absence of atomic material at a large negative velocity in the spectra of $\\delta$ and $\\epsilon$~Ori (and its presence in $\\zeta$~Ori) supports a proposal by Jenkins \\& Peimbert that the line of sight to $\\zeta$ intercepts a bow shock facing away from us, perhaps created by the collision of wind-like material with some foreground obstruction. One edge of the molecular cloud complex Lynds~1630 is situated close to $\\zeta$~Ori in the sky, but we present some evidence that seems to indicate that the cloud is more distant, in which case it could not serve as the obstruction. However it is possible that the outermost extension of a high-speed jet from a star forming within the cloud can explain the high-velocity material and the shock front created by it. For both stars, the H$_2$ absorption features are separated into two velocity components. Total H$_2$ column densities toward $\\delta$ and $\\epsilon$~Ori are $5.5\\times 10^{14}$ and $1.9\\times 10^{16}{\\rm cm}^{-2}$, respectively. When these values are compared to the column densities of H~I, the fractions of H atoms bound in molecular form $2N({\\rm H}_2)/[2N({\\rm H}_2)+N({\\rm H~I})]= 7\\times 10^{-6}$ for $\\delta$ and $1.3\\times 10^{-4}$ for $\\epsilon$. The rotation temperatures of the molecules with $J > 2$ toward $\\epsilon$~Ori indicate that the gas is in the general vicinity of the stars that emit UV fluxes capable of rotationally pumping the molecules. For the strongest component of H$_2$ toward $\\delta$~Ori, the pumping rate is lower and consistent with a general UV flux level in the plane of the Galaxy. ", "introduction": "The absorption bands of the Lyman and Werner systems arising from the ground vibrational state of H$_2$ provide a rich array of electronic transitions from different rotational levels. It is often true that these are the most conspicuous and plentiful features in the spectra of early-type stars below about 1100\\AA\\ (Morton 1975; Jenkins et al. 1989; Snow, Allen, \\& Polidan 1990) . In the interstellar medium, the rotational excitation of molecular hydrogen is driven by a number of processes, such as collisional excitation and de-excitation, optical pumping by starlight, and the introduction of newly formed molecules. For this reason, the relative rotational populations of H$_2$ have been used to gain insights on the characteristic molecule formation (and destruction) rates, temperatures, and densities in the H$_2$-bearing clouds (Spitzer \\& Cochran 1973; Spitzer et al. 1973; Spitzer, Cochran, \\& Hirshfeld 1974; Spitzer \\& Zweibel 1974; Jura 1975a, b; Morton \\& Dinerstein 1976; Shull 1977, 1979; Shull \\& York 1977). The UV data provide critical insights on the character of H$_2$ that are useful to other kinds of observations and our current understanding of how various environmental factors influence the rates of molecule formation and destruction (Shull \\& Beckwith 1982). In some cases, initial studies with the {\\it Copernicus} satellite revealed changes in the widths of the profiles for absorption features arising from levels with different values of the rotational quantum number $J$ (Spitzer \\& Cochran 1973; Spitzer, Cochran, \\& Hirshfeld 1974). More exacting analyses of high quality recordings of the H$_2$ profiles in several stars by Spitzer \\& Morton (1976) indicated that there were shifts in rotational temperatures for components with small changes in radial velocity, as one might expect for the gas behind shock fronts created by stellar winds. More recently, a significant improvement in the ability to study this phenomenon has come from the much better wavelength resolving power of the Interstellar Medium Absorption Profile Spectrograph (IMAPS) (Jenkins et al. 1988), which has flown on both sounding rockets and orbital missions (Jenkins 1993, 1995; Jenkins et al. 1996). Using data acquired by IMAPS on the ORFEUS-SPAS~I mission launched on the STS-51 Space Shuttle flight in 1993, Jenkins \\& Peimbert (1997) discovered that one of the velocity components of H$_2$ in front of $\\zeta$~Ori~A showed small, progressive velocity offsets and profile broadenings as the rotation states increased from $J=0$ to 5. They made the conjecture that this behavior could arise from H$_2$ forming in the collapsing column of partially ionized, cooling gas behind a nearly stationary bow shock facing away from us. This shock front was thought to be created by the abrupt deceleration of wind-like material from the star (or other stars nearby) as it collided with a dense cloud, perhaps similar to ones seen elsewhere in Orion that have comet-like tails facing away from prominent stars in the association (Bally et al. 1991; Cernicharo et al. 1992). This interpretation was supported by spectroscopic evidence for low-ionization material at $-94\\,{\\rm km~s}^{-1}$ (interpreted as the high speed material feeding the shock on its upstream side), and more highly ionized atoms at lower velocities (immediate post-shock gas at $-36\\,{\\rm km~s}^{-1}$). These velocities, relative to the observed pileup of cold material at about $0\\,{\\rm km~s}^{-1}$, implied a low compression ratio of 2.6. Jenkins \\& Peimbert concluded that this outcome could be produced by an embedded magnetic field in the pre-shock material that is oriented nearly parallel to the flow. The motivation of the present investigation of H$_2$ in front of the other two bright stars in the belt of the Orion constellation, $\\delta$~Ori~A and $\\epsilon$~Ori, was to gather independent evidence on the phenomenon studied by Jenkins \\& Peimbert (1997) for $\\zeta$~Ori~A. Since there are indications that the three stars are within a large cavity containing x-ray emitting gas (Burrows et al. 1993) and enclosed by an expanding shell of cooler material (Reynolds \\& Ogden 1979) powered by either stellar winds (Castor, McCray, \\& Weaver 1975; Weaver et al. 1977) or by one or more explosive events (Cowie, Songaila, \\& York 1979), it was not unreasonable to expect that we could witness additional manifestations of the unusual behavior of H$_2$. Tantalizing indications that a phenomenon similar to that found by Jenkins \\& Peimbert was occurring on the line of sight to $\\delta$~Ori~A were discussed by Spitzer \\& Morton (1976). The two additional Orion Belt targets were observed by IMAPS on its second flight conducted in late 1996 aboard the ORFEUS-SPAS~II mission on STS-80 (Hurwitz et al. 1998). We present here the results of our investigation of the absorption features arising from H$_2$. The paper is organized as follows. The section on observations (\\S\\ref{obs}) starts with a description of some fundamental properties of the two target stars (\\S\\ref{tgts}), $\\delta$ and $\\epsilon$~Ori, along with $\\zeta$~Ori~A which is featured as a contrasting case in later discussions. The IMAPS instrument and some properties of the data that are relevant to our investigation are described in the next subsection (\\S\\ref{inst}), and this is followed by a brief description of the exposures that were taken (\\S\\ref{exposures}). The section on analysis (\\S\\ref{analysis}) describes some general steps that were taken in the data reduction (\\S\\ref{general_steps}), outlines some special procedures to register the velocities of features from H$_2$ in different $J$ levels (\\S\\ref{vel_regs}), shows the results of an investigation to check on the relative $f$ values of the weakest lines out of $J=1$, 2 and 3 (\\S\\ref{validation}), and ends with a description of how we constructed profiles of column density vs. velocity for each $J$ level. The fundamental qualitative properties of the H$_2$ features in the spectra of $\\delta$ and $\\epsilon$~Ori are discussed in \\S\\ref{qualitative}. The distributions of H$_2$ over different $J$ levels are presented in \\S\\ref{distribution}, using the concept of a rotation temperature over specific $J$ levels as a convenient way to describe the relative populations. We use these rotational temperatures to synthesize a few approximate conclusions about the strengths of the UV pumping fields. In the section that describes our interpretation of the results (\\S\\ref{interpret}), we note that the unusual shifts and broadening of the H$_2$ profiles for different $J$ found by Jenkins \\& Peimbert (1997) seem to be correlated with the presence (for $\\zeta$~Ori~A) or absence (for $\\delta$ and $\\epsilon$~Ori) of high velocity gas (\\S\\ref{hi_vel}) or the possible proximity of $\\zeta$~Ori to a dense cloud with a high rate of star formation inside (\\S\\ref{structures}). Finally, in \\S\\ref{remarks} we present a few remarks that pull together some of the concepts presented earlier and discuss their possible meanings. ", "conclusions": "\\label{remarks} In studying the spectra of $\\delta$ and $\\epsilon$~Ori, we had high expectations of finding duplications of the unusual findings of Jenkins \\& Peimbert (1997) for one of the H$_2$ components toward $\\zeta$~Ori~A. They reported that the profile became broader and shifted toward more negative radial velocities as the rotational excitation increased from $J=0$ to 5. Our optimism about seeing more examples of this phenomenon was based on the premise that, in Jenkins \\& Peimbert's interpretation of a bow shock as its source, the stellar wind that may have created this shock front was probably a composite phenomenon common to all three stars in the Orion Belt. The apparent lack of this effect for $\\delta$ and $\\epsilon$~Ori deprives us of the chance to see additional manifestations of it, but opens the way for us to examine circumstantial evidence that could give us more insight on the validity of the shock interpretation. Aside from not showing the strange behavior of H$_2$, $\\delta$ and $\\epsilon$~Ori differ from $\\zeta$~Ori in two other respects that could have a bearing on the shock interpretation. First, $\\zeta$ shows absorption by singly and doubly ionized species at a velocity of about $-90\\,{\\rm km~s}^{-1}$, while the others do not. This high velocity material was a good candidate for the upstream gas that could have created the bow shock. Second, in projection $\\zeta$~Ori is very near the edge of the dense cloud Lynds~1630, an accumulation of gas that exhibits a bright edge of ionized material (IC~434) on the side towards $\\zeta$~Ori and holds within it many compact regions of star formation. The sight lines to $\\delta$ and $\\epsilon$ Ori are, by contrast, unremarkable. The real source of the gas at $-90\\,{\\rm km~s}^{-1}$ that is seen in the spectrum of $\\zeta$~Ori is uncertain. It may be some byproduct of the winds from the stars, but it is not easy to understand why similar components are not seen toward $\\delta$ and $\\epsilon$~Ori. An interesting alternative is the possibility that the $-90\\,{\\rm km~s}^{-1}$ component has nothing to do with stellar winds from the Belt stars. The complex of very dense gas near the $\\zeta$~Ori line of sight is known for its plentiful star forming regions. Such regions typically discharge gas at high velocities, starting with flows that are highly collimated. Perhaps one such flow could escape from its gas envelope and be responsible for the high velocity component that shows up for $\\zeta$~Ori in Figs.~\\ref{ciispec} and \\ref{niispec}. A well-directed flow could provide a reasonable way to explain the absence in the spectra of $\\delta$ and $\\epsilon$~Ori of any feature analogous to the $-90\\,{\\rm km~s}^{-1}$ component exhibited by $\\zeta$~Ori~A. New insights on the problem discussed in this paper may arise from observations by the Far Ultraviolet Spectroscopic Explorer (FUSE), which has the capability of examining a very much broader sample of cases in its lifetime. However its lower wavelength resolution will make it harder to study the detailed structures in the H$_2$ profiles, and productive research in this area may require the existence of even more extreme cases of profile variations with $J$ than those shown toward $\\zeta$~Ori~A." }, "0003/astro-ph0003304_arXiv.txt": { "abstract": "Significant adjustments to the values of the cosmological parameters estimated from high-redshift Type Ia Supernov\\ae~data are reported, almost an order of magnitude greater than previously found. They arise from the effects of weak gravitational lensing on observations of high-redshift sources. The lensing statistics used have been obtained from computations of the three-dimensional shear in a range of cosmological $N$-body simulations, from which it is estimated that cosmologies with an underlying deceleration parameter $q_0 = -0.51 +0.03/-0.24$ may be interpreted as having $q_0 = -0.55$ (appropriate to the currently popular cosmology with density parameter $\\Omega_M = 0.3$ and vacuum energy density parameter $\\Omega_{\\Lambda} = 0.7$). In addition, the standard deviation expected from weak lensing for the peak magnitudes of Type Ia Supernov\\ae~at redshifts of 1 is expected to be approximately 0.078 magnitudes, and 0.185 magnitudes at redshift 2. This latter value is greater than the accepted intrinsic dispersion of 0.17 magnitudes. Consequently, the effects of weak lensing in observations of high-redshift sources must be taken properly into account. ", "introduction": "\\subsection{Background} The weak gravitational lensing of light from distant sources by the large scale structure in the universe results in the appearance of shear and convergence in images. The application of `ray-tracing' methods, and other programmes, to cosmological $N$-body simulations has enabled weak lensing statistics to be recorded for imaginary sources at various redshifts. These statistics in general show significant differences between different cosmological models, and, of course, differences for sources at different redshifts. From weak lensing magnifications computed from `ray-tracing' in $N$-body simulations, Wambsganss et al. (1997) were able to estimate the likely magnification biases for sources at redshifts $z=0.5$ and $z=1$, and suggested that these biases should be applied to the observed magnitudes of high-redshift Type Ia Supernov\\ae. Their work was conducted in a cosmological simulation with density parameter $\\Omega_M = 0.4$ and vacuum energy density parameter $\\Omega_{\\Lambda} = 0.6$, for which the deceleration parameter is $q_0 = -0.4$. They found for this cosmology that if the computed median demagnification for $z = 1$ were applied to the observed Supernov\\ae~at this redshift it would shift their positions on the Hubble diagram indicative of a different cosmology with $q_0 = -0.395$. Riess et al. (1998) and Perlmutter et al. (1999) have independently used data sets of Type Ia Supernov\\ae~extending to a redshift of 0.97 to estimate values of the cosmological parameters from the Hubble diagram. In discussion of their results in the light of the work by Wambsganss et al. (1997), Riess et al. (1998) have stated that the effects on values of the cosmological parameters should be negligible. Perlmutter et al. (1999) have assumed that the effects of magnification or demagnification average out, and that the most overdense (or high magnification) lines of sight should be rare for their set of 42 high-redshift Supernov\\ae. However, Barber et al. (2000) have now applied a new algorithm developed by Couchman, Barber and Thomas (1999) for the three-dimensional shear to obtain weak lensing statistics in simulations with cosmological parameters similar to those estimated from the high-redshift Supernov\\ae~data. Their results show that the median demagnification values are considerably less and the dispersions in the probability distributions for the magnification are greater than found by Wambsganss et al. (1997) for redshifts of 1 and 0.5 (although from different cosmologies). This suggests that the work of Wambsganss et al. (1997) should be reviewed in the light of the new weak lensing data now available from a suitable cosmology. In this paper I apply the new data from Barber et al. (2000) in an attempt to re-estimate the effects of weak lensing on the cosmological parameters, and find the effects to be far from negligible. \\subsection{Outline of paper} A brief outline of this paper is as follows. In {\\bf Subsection 1.3} work by others on adjustments to the cosmological parameters resulting from weak gravitational lensing is summarised. {\\bf Section 2} summarises the method followed by Barber et al. (2000) for obtaining the weak lensing statistics. {\\bf Subsection 2.1} describes the cosmological $N$-body simulations; {\\bf Subsection 2.2} summarises the main features of the algorithm used for the determination of the shear in three-dimensions; {\\bf Subsection 2.3} outlines the method for integrating the shear values throughout the linked simulation volumes, and describes how these values are then used to determine the probability distributions for the magnifications of sources at high-redshift. {\\bf Section 3} summarises the relevant weak lensing statistics which are used in the present work. In {\\bf Section 4}, I describe the form of the Hubble diagram and its dependence on the cosmological parameters, and also summarise the important work by Riess et al. (1998) and Perlmutter et al. (1999) in estimating the cosmological parameters from high-redshift Type Ia Supernov\\ae~data. In {\\bf Section 5}, I discuss how the dispersions in the magnification distributions for different source redshifts may affect the observed distance moduli. For the specific case of the $\\Omega_M = 0.3$, $\\Omega_{\\Lambda} = 0.7$ cosmology, I then consider by how much the Supernov\\ae~data are displaced on the Hubble diagram, and how this translates into effective changes to the perceived values of the cosmological parameters. {\\bf Section 6} summarises the results obtained for the deceleration parameter, and discusses the findings in the light of work by others. \\subsection{Other work} Wambsganss et al. (1997) have used the same ray-tracing method as described by Wambsganss, Cen and Ostriker (1998) for cosmological simulations with $\\Omega_M = 0.4$, $\\Omega_{\\Lambda} = 0.6$, and normalisation $\\sigma_8 = 0.79$. The magnification values above and below which 97\\mbox{$\\frac{1}{2}$}\\% of all of their lines of sight fall were $\\mu_{\\mathrm{low}}=0.951$ and $\\mu_{\\mathrm{high}}=1.101$ for source redshifts of $z = 1$, and $\\mu_{\\mathrm{low}}=0.978$ and $\\mu_{\\mathrm{high}}=1.034$ for $z=0.5$. The median values for the magnification were 0.983 at $z = 1$, and 0.993 at $z = 0.5$. The authors claim that these values would give rise to observed values of $q_0 = -0.395 +0.125/-0.095$ for the $z = 1$ data, and $q_0 = -0.398 + 0.048/-0.077$ for the $z = 0.5$ data, rather than the assumed value of $-0.4$. (The quoted errors arise from the magnification ranges described above in the asymmetrical distributions.) They also report that the lensing-induced dispersions in their critical density cosmology are three times larger, but this cosmology uses a normalisation of $\\sigma_8 = 1.05$ which overproduces the present-day rich cluster abundances. For this cosmology, they report magnifications up to 100, and correspondingly highly dispersed distributions, very much larger than those of Barber et al. (2000) for $z=3$. The high magnifications almost certainly derive from the low value of their (fixed) softening scale. Fluke, Webster and Mortlock (1999) and Fluke, Webster and Mortlock (2000) have used a `ray-bundle' method in which a discrete bundle of light rays is traced. The method allows a direct comparison between the shape and size of the bundle at the observer and at the source plane, so that the magnification, ellipticity and rotation can be determined straightforwardly. The cosmological models investigated were similar to those of Barber et al. (2000). However, the particles were considered as point masses, so that very high magnification values could be achieved in principle; however, to alleviate this situation, the authors do not include bundles which pass within $\\sqrt{2}$ of the Einstein radius of any particle. Because of their use of point particles, they use the empty beam approximation which gives rise to magnification probability distributions with minimum magnifications of 1 (and therefore mean and median values greater than, or equal to, one), high magnification tails, and broad dispersions in magnification. For $z = 1$, the distribution in the magnifications for the SCDM cosmology ($\\Omega_M = 1,$ $\\Omega_{\\Lambda} = 0$) is clearly broader than that for the low density (LCDM) cosmology ($\\Omega_M = 0.3$, $\\Omega_{\\Lambda} = 0.7$), which in turn is broader than the open (OCDM) cosmology ($\\Omega_M = 0.3,$ $\\Omega_{\\Lambda} = 0$). This is also true of Barber et al.'s (2000) data, although Barber et al. (2000) find a completely different order for the cosmologies at higher source redshifts. Because of the high magnification tails in Fluke, Webster and Mortlock's (2000) distributions, they discard those high magnification lines of sight which occur with low probability before defining the $\\mu_{\\mathrm{low}}$ and $\\mu_{\\mathrm{high}}$ values. These magnification probability distributions are then used by Fluke and Webster (2000) to examine the effects of the weak lensing dispersion on measurements of $q_0$ in both the empty beam (more appropriate for point particles) and full beam limits. They consider whether the resulting shifts in the Hubble diagram can be fitted with various cosmological models. With input data from the $z=1$ magnification distribution for the flat LCDM simulations, they find that $q_0 = -0.53 +0.96/-0.02$ from the empty beam distribution. To convert their empty beam magnification values to filled beam values, Fluke and Webster (2000) use a simple scaling relationship. However, since the filled beam approach is inappropriate to their method, the resulting values should be treated with caution. They find the revised value $q_0 = -0.61~+0.89/-0.02$; i.e., a much larger departure than for the empty beam approximation, but a similar dispersion. Again the dispersion can be partly explained in terms of their use of point particles, although they do attempt to modify their distributions to account for the high magnification tails. It should be mentioned that Wambsganss et al. (1997) have assumed that the observed Supernov\\ae~have magnifications of unity, and then have assessed departures from the best fit cosmology, rather than assuming that the best fit cosmology comes from the most likely demagnification at the peak of the magnification distributions, which is the approach I take in this paper. Although Fluke and Webster (2000) use the median magnification values, the values are always close to unity and greater than, or equal to, unity for the empty beam case. ", "conclusions": "I have summarised the results for the magnification distributions for the different cosmologies for different source redshifts as obtained by Barber et al. (2000) in Table~\\ref{mu4C}. At high redshift, the LCDM cosmology produces the highest magnifications, the broadest distribution curves, and the lowest peak values. For sources at $z=3.6$ in the LCDM cosmology, 97\\mbox{$\\frac{1}{2}$}\\% of all lines of sight have magnification values up to 1.850. (The maximum magnifications, not quoted here, depended on the choice of the minimum softening in the code, although the overall distributions were very insensitive to the softening.) The rms fluctuations in the magnification (about the mean) were as much as 0.191 in this cosmology, for sources at $z=3.6$. Even for sources at $z=0.5$ there is a measurable range of magnifications in all the cosmologies. The immediate implication is the likely existence of a bias in observed magnitudes of distant objects, and a likely dispersion for standard candles, for example, Type Ia Supernov\\ae~at high redshift. In particular, the weak lensing dispersion from 97\\mbox{$\\frac{1}{2}$}\\% of the lines of sight expected in the peak magnitudes for Type Ia Supernov\\ae~at redshifts of 1 may be as much as 0.252 magnitudes. The standard deviations in the distributions are 0.078 magnitudes for $z = 1$ and 0.185 magnitudes for $z = 2$. These values are to be compared with the accepted dispersion of 0.17 magnitudes, so that we should expect to see an increasing dispersion for the peak magnitudes of Supernov\\ae~as they are discovered at higher redshifts. I have made use of the magnification statistics from the LCDM cosmology, to reanalyse the results for the cosmological parameters determined from the high-redshift Type Ia Supernov\\ae~data. (Both Riess et al., 1998, and Perlmuter et al., 1999, point to cosmologies with parameters close to those of the LCDM model, i.e., $\\Omega_M = 0.3$, $\\Omega_{\\Lambda} = 0.7$.) The dispersions in the magnification are monotonically increasing with redshift, and have rms fluctuations about the mean of 0.045 for $z = 0.99$, and 0.016 for $z = 0.49$. These have been translated into variations in distance modulus, which, in turn, suggest a different cosmological model for the data. On the assumption that $\\Omega_M = 0.3$, $\\Omega_{\\Lambda} = 0.7$ and $q_0 = -0.55$ is the cosmology which fits the observed data best, my results indicate that the true underlying cosmology could have $q_0 = -0.51$ based on the peak magnification values, with a spread of $+0.03/-0.24$ based on the computed magnification dispersion. This would represent a significant adjustment to the cosmological parameters. Wambsganss et al. (1997) found a median demagnification value much closer to unity (1.7\\% below for $z = 1$) in their cosmology with $\\Omega_M = 0.4,$ $\\Omega_{\\Lambda} = 0.6$ than Barber et al. (2000) have found (3.4\\% below for $z = 1$) in the $\\Omega_M = 0.3$, $\\Omega_{\\Lambda} = 0.7$ cosmology. Consequently, my adjustment to $q_0$ is considerably greater. However, the magnification dispersions give rise to similar dispersions in $q_0$. Wambsganss et al.'s (1997) calculated adjustment to $q_0$ is approximately an order of magnitude smaller than the departure I would suggest. Fluke, Webster and Mortlock's (2000) estimate for the revised value of the deceleration parameter ($q_0 = 0.53~+0.96/-0.02$ for the $\\Omega_M = 0.3,$ $\\Omega_{\\Lambda} = 0.7$ cosmology) in the empty beam approximation is only half the value of the departure I suggest. The value of their dispersion can not be easily compared with mine, because of their use of the empty beam approximation, which gives rise to a different form for the distribution curve with values of magnification which are always positive. Furthermore, their method to convert the empty beam magnification values (more appropriate to their method) to filled beam values was not rigorous, so that the resulting values should be treated with caution. Both Wambsganss et al. (1997) and Fluke, Webster and Mortlock (2000) have assumed that the observed Supernov\\ae~have magnifications of unity, so that the median demagnifications when using the filled beam approach are assumed to reposition the Supernov\\ae~at dimmer magnitudes. Fluke, Webster and Mortlock (2000) have to reposition them at brighter magnitudes when using the empty beam approximation. I have adopted the opposite viewpoint. I have assumed that the Supernov\\ae~are observed at their median demagnification values, and therefore have to be repositioned at brighter magnitudes (i.e., reduced distance moduli) to obtain the correct cosmological parameters. Consequently, the direction of my adjustment to $q_0$ is opposite to those of the above authors for the filled beam approximation." }, "0003/astro-ph0003074_arXiv.txt": { "abstract": "The dynamics and thermal structure of the surface layers of stars with outer convection zones can be studied in some detail by means of numerical simulations of time-dependent compressible convection. In an effort to investigate the properties of ``stellar granulation'' as a function of spectral type, we have carried out elaborate 2-dimensional radiation hydrodynamics calculations of surface convection for a variety of stellar parameters. The main features of these simulations are reviewed, with particular reference to standard mixing length models. ", "introduction": "Convection is a universal feature. Essentially all types of stars have either a convective core, a convective envelope, or both. Low-mass stars are fully convective, giants may accommodate several distinct convective shells. In the case of the Sun, the energy transport in the inner parts is entirely due to radiation, while in the outer 28.7\\% (in radius) it is primarily due to large-scale convective currents. At the surface, the solar granulation is the visible imprint of gas flows in the outermost layers of the convection zone. The role of stellar convection is far-reaching: Convective energy transport determines the internal temperature structure of a star and its radius (which decreases with increasing convective efficiency), and hence controls the star's global properties. Convective regions are chemically completely mixed, and overshooting convective flows lead to partial mixing of the adjacent radiative layers. ``Overshoot'' and similar mixing processes which are not confined to the convectively unstable regions are thought to be responsible for the existence of carbon stars, carbon-rich white dwarfs, and for the destruction of lithium in solar-type stars. Convective motions and concomitant temperature fluctuations exert a direct influence on stellar spectra, causing small but practically relevant changes in wavelength position, shape and strength of spectral lines. The stochastic convective motions can excite stellar oscillations (like in the Sun) and are a source of acoustic energy, contributing to the heating of stellar chromospheres. Finally, convection and rotation are necessary preconditions for the operation of the magnetic dynamo mechanism, and hence for stellar activity. Unfortunately, a closed analytical theory of stellar convection is lacking due to the complexity of the underlying hydrodynamical problem. So far, stellar structure models still rely on a phenomenological approximation, the so-called \\emph{mixing-length theory} (MLT). ", "conclusions": "Radiation hydrodynamics simulations of stellar surface convection have now reached a level of sophistication far beyond idealized numerical experiments. They are the key for a better understanding of the thermal structure and dynamics of stellar convection zones, including overshoot and its role for mixing. The comparison with ``classical'' mixing-length models reveals quantitative and qualitative differences. In the simulations, the dynamics of convection is dominated by fast, cool, narrow downdrafts, which form coherent structures extending over many pressure scale heights. A pronounced up/down asymmetry is a general feature of the numerical models (2D as well as 3D), seen for all types of stars investigated so far. Obviously, this result is in stark contrast with the symmetric picture of MLT where ``bubbles'' are assumed to travel for about one pressure scale height before dissolving. Since the ``jets'' are driven by radiative cooling at the stellar surface, convection is an extremely non-local process. The properties of the surface determine the dynamics and structure of the whole convection zone. Certainly, the assumption of locality is a major problem with MLT. Hydrodynamical model atmospheres can be used to study the formation of spectral lines in an inhomogeneous medium. For the Sun, we found that ``granulation corrections'' for spectroscopic abundance determinations can amount to -0.3 dex for the most temperature sensitive lines. These corrections are much larger than the well-known NLTE corrections that have been investigated so far. Work is under way to check whether these findings have notable consequences for the currently adopted chemical composition of the Sun and other stars. While is seems hopeless to successfully model the structure of the superadiabatic surface layers within the framework of MLT, it is possible to calibrate MLT through hydrodynamical simulations for application to stellar evolution. Based on 2D radiation hydrodynamics, our present calibration ultimately needs to be verified by 3D simulations. In the meantime, we are working to extend the calibration to metal-poor stars, down to [M/H]=-2. The hydrodynamical simulations for A-type stars (and White Dwarfs) demonstrate that the velocity field due to overshooting convective motions declines exponentially with distance from the Schwarzschild boundary and leads to diffusive mixing. Although this result is not directly applicable to strongly adiabatic conditions, we believe that the exponential depth-dependence of the diffusion coefficient is a general feature of overshoot and may have important implications for mixing and nucleosynthesis in stellar interiors." }, "0003/astro-ph0003256_arXiv.txt": { "abstract": "s{We apply statistical tests, based on the study of the coefficients in a wavelet decomposition, to a cosmological signal: the Cosmic Microwave Background (CMB) anisotropies. The latter represent the superposition of primary anisotropy imprints of the initial density perturbations and secondary anisotropies due to photon interactions after recombination. In an inflationary scenario with Gaussian distributed fluctuations, we study the statistical signature of the secondary effects. More specifically, we investigate the dominant effects arising from the Sunyaev-Zel'dovich effect of galaxy clusters. Our study predicts the non-Gaussian signature of these secondary anisotropies and its detectability in the context of the future CMB satellite Planck Surveyor.} ", "introduction": "One of the major goals of cosmology is to understand the origin of the initial density perturbations. Do they come from inflation\\cite{guth81} or from topological defects\\cite{vilenkin85}? One way to answer this question is to make a statistical analysis of the CMB anisotropies. In fact, inflation predicts a Gaussian distribution of the primary anisotropies whereas the topological defects predict a non-Gaussian distribution. However, not only topological defects produce non-Gaussian signals. There are other astrophysical sources of non-gaussianity such as gravitational lensing (cf. F. Bernardeau's contribution in this proceedings) and the secondary anisotropies such as the Sunyaev-Zel'dovich effect\\cite{sunyaev80} of galaxy clusters. \\\\ In this context, we have developed a method for the statistical analysis based on the wavelet decomposition of a signal. The method has been tested on an arbitrary set of Gaussian and non-Gaussian maps and then applied to CMB simulated maps to investigate the effect of the SZ contribution on the statistical signature of the anisotropies. ", "conclusions": "In the present study, we investigate the statistical signature induced by the SZ effect of galaxy clusters when the primary anisotropies result from an inflationary scenario and are thus Gaussian distributed. We use discriminators based on the statistical properties of the coefficients in a four level wavelet decomposition. In our study, we find that the SZ effect of clusters generates a very large non-Gaussian signature that dominates by far all other secondary anisotropies. We apply our statistical tests to a Planck-like configuration in order to estimate the capabilities of the satellite to detect the non-Gaussian signature of the SZ effect. In this case, we detect unambiguously the non-Gaussian signature at the third decomposition scale ($\\simeq 12$ arcminutes), the first ($\\simeq 3$ arcminutes) and second ($\\simeq 6$ arcminutes) scales being affected by the beam convolution." }, "0003/gr-qc0003027_arXiv.txt": { "abstract": "We present the first results in a new program intended to make the best use of all available technologies to provide an effective understanding of waves from inspiralling black hole binaries in time for imminent observations. In particular, we address the problem of combining the close-limit approximation describing ringing black holes and full numerical relativity, required for essentially nonlinear interactions. We demonstrate the effectiveness of our approach using general methods for a model problem, the head-on collision of black holes. Our method allows a more direct physical understanding of these collisions indicating clearly when non-linear methods are important. The success of this method supports our expectation that this unified approach will be able to provide astrophysically relevant results for black hole binaries in time to assist gravitational wave observations. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003310_arXiv.txt": { "abstract": "We present a catalog of photometry and photometric redshifts of 335 faint objects in the HDF--S NICMOS field. The analysis is based on (1) infrared images obtained with the Hubble Space Telescope (HST) using the Near Infrared Camera and Multi-Object Spectrograph (NICMOS) with the F110W, F160W, and F222M filters, (2) an optical image obtained with HST using the Space Telescope Imaging Spectrograph (STIS) with no filter, and (3) optical images obtained with the European Southern Observatory (ESO) Very Large Telescope (VLT) with {\\it U}, {\\it B}, {\\it V}, {\\it R}, and {\\it I\\/} filters. The primary utility of the catalog of photometric redshifts is as a survey of faint galaxies detected in the NICMOS F160W and F222M images. The sensitivity of the survey varies significantly with position, reaching a limiting depth of $AB(16,000) \\approx 28.7$ and covering 1.01 arcmin$^2$ to $AB(16,000) = 27$ and 1.05 arcmin$^2$ to $AB(16,000) = 26.5$. The catalog of photometric redshifts identifies 21 galaxies (or 6\\% of the total) of redshift $z > 5$, 8 galaxies (or 2\\% of the total) of redshift $z > 10$, and 11 galaxies (or 3\\% of the total) of best-fit spectral type E/S0, of which 5 galaxies (or 1\\% of the total) are of redshift $z > 1$. ", "introduction": "The Hubble Deep Field South (HDF--S) images are among the deepest images of the universe ever obtained at optical and infrared wavelengths. In this paper, we present a catalog of photometry and photometric redshifts of 335 faint objects in the HDF--S NICMOS field. The analysis is based on (1) infrared images obtained with the Hubble Space Telescope (HST) using the Near Infrared Camera and Multi-Object Spectrograph (NICMOS) with the F110W, F160W, and F222M filters, (2) an optical image obtained with HST using the Space Telescope Imaging Spectrograph (STIS) with no filter, and (3) optical images obtained with the European Southern Observatory (ESO) Very Large Telescope (VLT) with {\\it U}, {\\it B}, {\\it V}, {\\it R}, and {\\it I\\/} filters. The analysis is similar to the analyses of the Hubble Deep Field (HDF) described previously by \\citet[hereafter LYF96]{LYF96} and \\citet[hereafter FLY99]{FLY99}, although in detail the current analysis differs from the previous analyses in three important ways: First, objects are detected in the NICMOS F160W and F222M images, at central wavelengths of $\\lambda \\approx 16,000$ \\AA\\ and $\\lambda \\approx 22,200$ \\AA, respectively. The analysis is in principle sensitive to galaxies of redshift as large as $z \\approx 18$, beyond which the Ly$\\alpha$-forest absorption discontinuity is redshifted past the response of the NICMOS F222M filter. Second, the optical and infrared photometry is measured using a new quasi-optimal photometry technique that fits model spatial profiles of detected objects to the space- and ground-based images. The technique is based on but extends the spatial profile fitting technique described previously by FLY99. In comparison with conventional methods, the new technique provides higher signal-to-noise ratio measurements, and in contrast with conventional methods, the new technique accounts for uncertainty correlations between nearby, overlapping neighbors. Third, the photometric redshifts are measured using our redshift likelihood technique with a sequence of six spectrophotometric templates, including the four templates of our previous analyses (of E/S0, Sbc, Scd, and Irr galaxies) and two new templates (of star-forming galaxies). Inclusion of the two new templates eliminates the tendency of our previous analyses to systematically underestimate the redshifts of galaxies of redshift $2 < z < 3$ (by a redshift offset of roughly 0.3), in agreement with results found previously by \\citet{Benitez98}. Comparison with spectroscopic redshifts of galaxies identified in the HDF and HDF--S indicates that with the sequence of six spectrophotometric templates the photometric redshifts are accurate to within an RMS relative uncertainty of $\\Delta z/(1 + z) \\lesssim 7\\%$ at all redshifts $z < 6$ that have as yet been examined. The primary utility of the catalog of photometric redshifts is as a survey of faint galaxies detected in the NICMOS F160W and F222M images. The sensitivity of the survey varies significantly with position, reaching a limiting depth of $AB(16,000) \\approx 28.7$ and covering 1.01 arcmin$^2$ to $AB(16,000) = 27$ and 1.05 arcmin$^2$ to $AB(16,000) = 26.5$. Likewise, the survey reaches a limiting depth of $AB(22,200) \\approx 24.8$ and covering 0.79 arcmin$^2$ to $AB(22,200) = 24$ and 1.09 arcmin$^2$ to $AB(22,200) = 23$. The organization of the paper is as follows: In \\S\\ 2, the observations are described. In \\S\\ 3, the object identification, photometry, and photometric redshift measurements are described. The results are presented in \\S\\ 4, and the discussion is presented in \\S\\ 5 and \\S\\ 6. The summary and conclusions are given in \\S\\ 7. Scientific analysis of the catalog will be presented in forthcoming papers. ", "conclusions": "Here we present a catalog of photometry and photometric redshifts of 335 faint objects in the HDF--S NICMOS field. The analysis is based on (1) infrared images obtained with HST using NICMOS with the F110W, F160W, and F222M filters, (2) an optical image obtained with HST using STIS with no filter, and (3) optical images obtained with the ESO VLT with {\\it U}, {\\it B}, {\\it V}, {\\it R}, and {\\it I\\/} filters. The primary utility of the catalog of photometric redshifts is as a survey of faint galaxies detected in the NICMOS F160W and F222M images. The sensitivity of the survey varies significantly with position, reaching a limiting depth of $AB(16,000) \\approx 28.7$ and covering 1.01 arcmin$^2$ to $AB(16,000) = 27$ and 1.05 arcmin$^2$ to $AB(16,000) = 26.5$. Likewise, the survey reaches a limiting depth of $AB(22,200) \\approx 24.8$ and covering 0.79 arcmin$^2$ to $AB(22,200) = 24$ and 1.09 arcmin$^2$ to $AB(22,200) = 23$. The catalog of photometric redshifts identifies 21 galaxies (or 6\\% of the total) of redshift $z > 5$, 8 galaxies (or 2\\% of the total) of redshift $z > 10$, and 11 galaxies (or 3\\% of the total) of best-fit spectral type E/S0, of which 5 galaxies (or 1\\% of the total) are of redshift $z > 1$." }, "0003/astro-ph0003126_arXiv.txt": { "abstract": " ", "introduction": "Accretion-powered X-ray pulsars are binary systems consisting of a neutron star and a stellar companion (here, X-ray binary pulsars or XBPs in short). The gravitational energy of accreting matter is converted to X-ray radiation, hence its luminosity is generally variable, depending on the mass-accretion rate. Transient behavior is also observed from many XBPs, most of which have a Be star companion with an eccentric orbit (e.g.\\ Stella et al.\\ 1982; Bildsten et al.\\ 1997). XBPs have been observed mainly in the hard X-ray band ($\\sim 2$--40~keV) with non-imaging satellites (e.g.\\ Nagase 1989). Their spectra are generally described by a power-law with a high-energy exponential cut-off (hereafter, ECUT power-law) with a photon index of $\\Gamma \\sim 1$ and a cut-off energy at around 10--20~keV. A fluorescent emission line at 6.4~keV from neutral iron atoms has been observed in many bright XBPs, which serves as a probe of the circumstellar matter of the binary system. In the soft X-ray band (below $\\sim 2$~keV), however, their spectra have not been well studied due to a large interstellar absorption, because most of the XBPs are located in the galactic plane (e.g.\\ Bildsten et al.\\ 1997). In order to study soft X-rays from XBPs, therefore, the Magellanic Cloud sources have great advantages, owing to the low interstellar absorption and well-calibrated distances. Woo et al.\\ (1995, 1996) made broad-band spectroscopic studies in the 0.1--40~keV band of SMC X-1 and LMC X-4, respectively, and found a ``soft excess'' below $\\sim 2$~keV, in addition to the usual hard spectrum given by an ECUT power-law model. Such a soft excess was also discovered from a transient XBP XTE~J0111.2$-$7317 in the Small Magellanic Cloud (SMC), by an ASCA observation in the 0.5--10.0~keV band (Yokogawa et al.\\ 2000). However, since such a broad-band spectroscopy has been performed on only a few XBPs, whether or not the soft component is common among XBPs is still unclear, especially for transient XBPs. Hughes (1994) serendipitously discovered a new transient XBP, RX~J0059.2$-$7138, in the SMC with the soft X-ray band observation of ROSAT, and found that the spectrum is unusually soft, which is composed of a blackbody with a temperature of $kT \\sim 35$~eV and a steep power-law with a photon index of $\\Gamma \\sim 2.4$. A proposed candidate for the optical counterpart was revealed to be a Be star by Southwell and Charles (1996). In order to examine the nature of the unusually soft spectrum of RX~J0059.2$-$7138, observations in the higher energy band are essential. RX~J0059.2$-$7138 was also detected in a simultaneous observation of the SNR E~0102$-$72.2 with the ASCA satellite, which is sensitive in the $\\sim 0.7$--10.0~keV band. A preliminary short report on the ASCA results can be found in Kylafis (1996). In this paper, we combine the ROSAT and ASCA data, and perform a broad-band timing spectroscopy, covering $\\sim 0.1$--10.0~keV. We show that the spectrum of RX~J0059.2$-$7138 has a hard component, which resembles the spectra of usual XBPs well, and that the pulsations are due to the hard component. In addition, a non-pulsating soft component is present in the spectrum. We assume the distance to RX~J0059.2$-$7138 to be 60~kpc, the nominal value to the SMC (Mathewson 1985). ", "conclusions": "\\subsection{Comparison to the Previous Results} Hughes (1994) analyzed only the ROSAT data, and showed that the spectrum is fitted by a model having an extremely soft spectrum: a blackbody ($kT \\sim 35$~eV) plus a steep power-law ($\\Gamma \\sim 2.4$). However, we have shown that the spectrum in the ASCA band (the hard component) is as hard as those of usual XBPs, and the ROSAT spectrum consists of mostly the ``soft excess'' below $\\sim 2$~keV. For a consistency check, we first fitted our ROSAT spectrum with Hughes' model, and obtained best-fit parameters consistent with Hughes (1994). However, as can be also seen in figure 2 in Hughes (1994), systematic positive residuals were found above $\\sim 1.5$~keV, which implies the existence of a hard component. We then fitted the ASCA and ROSAT spectra simultaneously with a three-component model, consisting of the hard ECUT power-law to the Hughes' two-component model. In this model, the best-fit blackbody temperature became much lower than the original result. After corrections for absorption and detector efficiency, the bolometric luminosity of the soft black body has an unrealistic value of $\\sim 10^{45}~{\\rm erg~s^{-1}}$. On the other hand, as we found, the combined ASCA and ROSAT spectra can be fitted with two different components, with a more reasonable luminosity of $\\sim 10^{38}~{\\rm erg~s^{-1}}$ for both components. Also, the energy-dependent pulse fraction and energy-independent pulse profile are naturally explained by our two-component model. We thus infer that the present model is more probable, hence the spectrum of RX~J0059.2$-$7138 is not unusual for an XBP, at least in the hard energy band. \\subsection{Oxygen Over-Abundance} We found an edge-like structure at 0.5~keV, which would be a hint of over-abundance of oxygen in the absorbing matter. One may argue, however, that the ROSAT spectrum fitted by Hughes (1994) does not show any structure at 0.5~keV. We should note that the blackbody and power-law components in the Hughes's model cross at around 0.5~keV, which may produce an artificial dip, and hence would compensate for the edge-like structure. To check whether the structure is real or not, we fitted a narrow-band spectrum (0.2--0.8~keV) with a power-law absorbed by cool matter of solar abundance. Because the structure remained near 0.5~keV, we claim that the edge-like structure is real. Since the oxygen abundance of the SMC interstellar matter is less than the solar value (Russell, Dopita 1992), we can infer that there exists oxygen-rich matter around the binary system of RX~J0059.2$-$7138. The companion star, which should have ejected the oxygen-rich matter, may be in an evolved stage where the stellar surface becomes abundant only in oxygen, although such stars (e.g.\\ WO stars) are very rare. On the other hand, metals such as carbon, nitrogen, and oxygen could be brought to the surface of a massive star by turbulent diffusion, due to the rapid rotation or tides in a binary system (Maeder 1987). We thus fitted the spectra again, based on the assumption that abundances of carbon and nitrogen in the absorbing matter are the same as that of oxygen. The abundance was then determined to be $\\sim 6$ solar, which leads us to a more comfortable possibility that the surface of the companion star is enriched in CNO, which may be caused by tidal effects in the close binary system. We found that a thin-thermal plasma with very low abundance of 0.02 solar can fit the soft X-ray spectrum. As we discuss in the next subsection, the soft component is likely to originate from a large region surrounding the binary system. This creates a dilemma that the X-ray emitting circumstellar medium is extremely under-abundant, while the X-ray absorbing circumstellar gas is CNO over-abundant. Although we have no clear idea of how to solve this dilemma, we suggest that the thin-thermal plasma is not a physical model, but is a phenomenological model like a broken power-law for the soft component. \\subsection{Origin of the Hard and Soft Components} Since the pulsation was found to be mostly due to the hard component, we can conclude that the hard component originates from a small region, probably near to the polar cap of the neutron star. Hence, the origin would be the same as those of other XBPs. However, the cut-off energy is lower than that of the usual XBPs (10--20~keV; e.g.\\ Mihara 1995a). According to the correlation between the cut-off energy and magnetic field strength (Mihara 1995a, 1995b), the magnetic field of RX~J0059.2$-$7138 is estimated to be $6\\times10^{11}$ 30 m\\AA\\ in the H1821+643 spectrum is remarkably high, $dN/dz \\sim$ 48, with a high (90\\%) confidence that it is greater than 17. We conservatively estimate that the cosmological mass density of the \\ion{O}{6} systems $\\Omega _{b}({\\rm O \\ VI}) \\gtrsim 0.0008 h_{75} ^{-1}$. With an assumed metallicity of 1/10 solar and a conservative assumption that the fraction of oxygen in the \\ion{O}{6} ionization stage is 0.2, we obtain $\\Omega _{b}({\\rm O \\ VI}) \\gtrsim 0.004 h_{75} ^{-1}$. This is comparable to the combined cosmological mass density of stars and cool gas in galaxies and X-ray emitting gas in galaxy clusters at low redshift. ", "introduction": "The resonance line doublet of Li-like \\ion{O}{6} is a sensitive probe of hot collisionally ionized or warm very low density photoionized gas in the intergalactic medium and galaxy halos. The \\ion{O}{6} $\\lambda \\lambda$1031.92, 1037.62 doublet has been detected in absorption toward QSOs over a wide range of redshifts (see \\S 1 in Tripp \\& Savage\\markcite{ts2000} 2000). The lowest redshift \\ion{O}{6} absorbers are particularly interesting because the redshifts of galaxies near the QSO sight lines can be measured, and the relationship between the \\ion{O}{6} absorber properties and environment can be studied. Furthermore, cosmological simulations predict that a substantial fraction of the baryons in the universe are in a shock-heated phase at $10^{5} - 10^{7}$ K at low $z$ (e.g., Cen \\& Ostriker\\markcite{co99} 1999; Dav\\'{e} et al.\\markcite{dave99} 1999), and preliminary results indicate that low-$z$ \\ion{O}{6} systems may indeed be an important baryon reservoir (Tripp \\& Savage\\markcite{ts2000} 2000). In a previous paper, Savage, Tripp, \\& Lu\\markcite{stl98} (1998) studied an intervening \\ion{O}{6} absorber associated with two galaxies at $z \\approx$ 0.225 in the spectrum of the radio-quiet QSO H1821+643 using a combination of low resolution {\\it Hubble Space Telescope (HST)} spectra with broad wavelength coverage and a high resolution {\\it HST} spectrum with very limited wavelength coverage. We have re-observed this QSO with an echelle mode of the Space Telescope Imaging Spectrograph (STIS) on {\\it HST}, which provides a resolution of $\\sim$7 \\kms\\ (FWHM) with broad wavelength coverage. In this paper we present in \\S 2 and \\S 3 new results on one probable and four definite \\ion{O}{6} absorption line systems in the STIS H1821+643 spectrum. In \\S 4 we discuss the implications of the high rate of occurance of \\ion{O}{6} absorbers at low redshift. The direct information we obtain about the highly ionized state of the gas from the presence of \\ion{O}{6} allows us to estimate the baryonic content of these systems. We conclude that \\ion{O}{6} systems are likely to harbor an important fraction of the baryons at the present epoch. ", "conclusions": "" }, "0003/astro-ph0003107_arXiv.txt": { "abstract": " ", "introduction": " ", "conclusions": "There are alternative interpretations for the origin of the emission line in the northwestern companion: one is a star forming object and the other one is a photoionized gas illuminated by the quasar. Petitjean et al. (1996) discussed both possibilities by constructing model continuum spectra with emission lines. In the star forming object hypothesis, they assumed a star formation rate of 13 $M_{\\odot}$ yr$^{-1}$ and 0.1 solar abundance without dust extinction. (They used q$_0 =0.5$ and $H_0 = 75$ km s$^{-1}$ Mpc$^{-1}$.) A shortcoming for this model arises from the fact that contribution from Lyman $\\alpha$ emission in the model is smaller than the observed value to account for intensities of continuum and emission lines simultaneously. In the photoionization model, the model $V$-band flux is too small and the Lyman alpha emission is too strong (about one order of magnitude larger than the observed value). In the latter model, a strong NV$\\lambda 1240$ and CIV$\\lambda 1549$ emission line should be seen in the spectrum of the companion. However the spectrum of the companion by Petitjean et al. (1996) seems not to show the NV$\\lambda 1240$ emission, and Hu et al. (1997) reported the absence of CIV$\\lambda 1549$ emission line in the companion. Thus the identification of the northwestern companion with a star forming object seems to be more plausible. The star formation rate estimated from the [OII]$\\lambda 3727$ emission luminosity is 45 to 230 $M_{\\odot}$ yr$^{-1}$ depending on the adopted conversion factor from the [OII]$\\lambda 3727$ luminosity to the H$\\alpha$ luminosity (Gallagher et al. 1989; Kennicutt 1992) without extinction correction. The star formation rate estimated from Lyman $\\alpha$ luminosity is 28-47 $M_{\\odot}$ yr$^{-1}$ under the case B recombination assumption ($L$(Lyman $\\alpha$)/$L$(H$\\alpha) = 8.7$) (Hu et al. 1996; Petitjean et al. 1996). The star formation rate estimated from the rest UV continuum (around 1400 \\AA) (Madau et al. 1998) also gives a value of 22 $M_{\\odot}$ yr$^{-1}$. The star formation rate derived from the [OII] emission is larger than those estimated from the Lyman $\\alpha$ emission and the UV continuum by a factor of up to $\\sim 10$. The origin of the discrepancy must be due to the severer dust extinction to the Lyman $\\alpha$ emission and the UV continuum. By assuming that the extinction corrected star formation rate derived from the [OII] luminosity is equal to that from the Lyman $\\alpha$ luminosity, the amount of extinction in the companion can be estimated. The estimated visual extinction ($A_V$) ranges from $\\sim 0$ to 1.3 mag for the Milky Way type extinction curve (Scheffler and Els\\\"asser 1988), $\\sim 0$ to 0.5 mag for the SMC type extinction curve (Calzetti et al. 1994), and $\\sim 0$ to 1.1 mag for the Calzetti type extinction curve (Calzetti 1997), by assuming $A_V = 3 E(B-V)$. Thus the extinction at [OII]$\\lambda3727$ emission is estimated to be $\\sim 0$ to 2.5 mag, and the extinction corrected star formation rate amounts to 45 to 2300 $M_{\\odot}$ yr$^{-1}$. The emission line region is adjacent to the thermal dust emission and the CO emission line region which must also be an intense star forming region at which the star formation rate inferred from the CO luminosity and the radio continuum luminosity is at the order of 1000 $M_{\\odot}$ yr$^{-1}$. The [OII] emitting region may be a less obscured region of the starburst object. However, the huge value of the estimated star formation ratio ($\\sim 2000 M_{\\odot}$ yr$^{-1}$) for the companion alone may indicate that at least a part of the [OII] emission does not come from star formation, but from the ionized gas illuminated by the quasar. Deep spectroscopic observations of the companion are required to examine a fraction of the contribution from the ionization by the quasar. \\par \\vspace{1pc}\\par It is our great pleasure to thank Subaru telescope team for their enormous efforts for the construction of the telescope. K.O. is supported by grant-in-aid from the Ministry of Education, Science, Sports and Culture of Japan (117401230). \\clearpage" }, "0003/astro-ph0003457_arXiv.txt": { "abstract": "We have observed the central region of the nearby starburst galaxy NGC~4945 with the mid-infrared spectrometers SWS and PHT-S aboard ISO. We do not find any evidence for the existence of the powerful AGN, inferred from hard X-ray observations. The upper limits on our AGN tracers $[\\ion{Ne}{v}] $14.32$\\mu$m\\&24.3$\\mu$m and $[\\ion{Ne}{vi}]$ 7.65$\\mu$m imply an A$_V>$160 towards the NLR, assuming the NLR to be of equal strength as in the Circinus galaxy. Other possibilities are discussed. The starburst excitation indicators $[\\ion{Ne}{iii}]$15.56$\\mu$m/$[\\ion{Ne}{ii}]$12.81$\\mu$m and L$_{\\rm bol}$/L$_{\\rm lyc}$ suggest that the starburst in the central region is at least 5$\\times$10$^6$ yrs old, and that it accounts for at least half of the nuclear bolometric luminosity. The starburst might well power the entire bolometric luminosity, but the available constraints are also consistent with an up to 50\\% contribution of the embedded AGN. With PHT-S, at a resolution of $\\approx$90, we detect strong absorption features of water ice, and, for the first time in an external galaxy, of CO$_2$ and CO. The same PHT-S spectrum also reveals strong emission from the family of PAH features. Finally, we have observed and detected several pure rotational and ro-vibrational H$_2$ lines, two of which, the (0-0) S(0) \\& S(1) lines, allow us to determine the excitation temperature (160K) and warm H$_2$ mass (2.4$\\times$10$^7$M$_{\\odot}$). The low excitation temperature shows Orion-like shocks not to be representative for the entire emission of the central region of the galaxy and fairly normal PDRs to be perhaps more typical. ", "introduction": "NGC~4945 is a nearby, large (20$\\arcmin\\times$4$\\arcmin$) spiral galaxy seen nearly edge on (i $\\sim$ 78$\\degr$; Ott \\cite{Ott}). At a recession velocity of 560km/s it is at the mean radial velocity of the Centaurus group (Hesser et al. \\cite{Hesser}), of which it is believed to be a member. Distance estimates vary between 3.5 and 4.0Mpc (see Bergman et al. \\cite{Bergman} and Mauersberger et al. \\cite{Mauersberger} for discussions). In this paper we will adopt a distance of 3.9 Mpc (Bergman et al. \\cite{Bergman}), which implies that 1$\\arcsec$ is equivalent to 18pc. NGC~4945 is one of the brightest infrared galaxies in the sky: S[12]=24Jy, S[25]=43Jy, S[60]=588Jy, S[100]=1416Jy (Rice et al. \\cite{Rice}). The total infrared luminosity amounts to L(8-1000$\\mu$m)=2.95$\\times$10$^{10}$ L$_{\\sun}$, $\\sim$75\\% of which originates in the central 12$\\arcsec\\times$9$\\arcsec$ (Brock et al. \\cite{Brock}). Near infrared observations reveal the nuclear region to be the site of a powerful, yet visually obscured, starburst. Br$\\gamma$ (Moorwood et al. \\cite{Moorwood96a}) and Pa$\\alpha$ (Marconi et al. \\cite{Marconi}) recombination line maps show the starburst to be concentrated in a circumnuclear disk or ring $\\sim$200pc across (11$\\arcsec$). Further evidence for (a period of) strong star formation comes from the discovery of a conical structure, roughly perpendicular to the galaxy major axis. It is believed to be a cavity, vacated by a starburst-driven superwind (Heckman et al. \\cite{Heckman}; Moorwood et al. \\cite{Moorwood96a}). The non-detection of $[$\\ion{O}{iii}$]$ within the cone and the absence of coronal lines excludes an AGN as the driver of the outflow. Clear evidence for the presence of an AGN comes from hard X-ray observations (Iwasawa et al. \\cite{Iwasawa}; Guainazzi et al. \\cite{Guainazzi}). The AGN X-ray emission is however heavily absorbed by a column density of 10$^{24.7}$cm$^{-2}$, which obscures the AGN at all optical and infrared wavelengths. Previous authors have attributed most of the IR luminosity to the starburst (e.g. Moorwood \\& Oliva \\cite{Moorwood94}; Koornneef \\& Israel \\cite{Koornneef96}). Hard X-ray observations with {\\it BeppoSAX} indicate that the bolometric luminosity may as well be accounted for by the AGN alone (Guainazzi et al. \\cite {Guainazzi}). 3cm\\&6cm ATCA radio maps of the central region of NGC~4945 (Forbes \\& Norris \\cite{Forbes}) are dominated by strong nuclear emission, and emission extended along the disc of the galaxy. There is also evidence for some filamentry structure associated with the cavity cleared by the starburst superwind. VLBI observations by Sadler et al. (\\cite{Sadler}) reveal the existence of a compact radio core. This, as well as the presence of H$_2$O megamasers in a Keplerian disc about a $\\sim$10$^6$ M$_{\\odot}$ black hole (Greenhill et al. \\cite{Greenhill}), are taken as further evidence for the presence of an AGN. Near infrared observations of molecular hydrogen emission in NGC~4945 have been reported by several authors over the last 15 years (e.g. Moorwood \\& Glass \\cite{Moorwood84}; Moorwood \\& Oliva \\cite{Moorwood88}; Koornneef \\cite{Koornneef93}; Moorwood \\& Oliva \\cite{Moorwood94}; Koornneef \\& Israel \\cite{Koornneef96}; Moorwood et al. \\cite{Moorwood96a}; Quillen et al. \\cite{Quillen}; Marconi et al. \\cite{Marconi}). While fluxes are known for eight ro-vibrational transitions accesible from the ground (Koornneef \\& Israel \\cite{Koornneef96}), spatial information is available only for the (1-0) S(1) 2.1218$\\mu$m line. These observations show the H$_2$ emission to be associated with the hollow cone, {\\it not} with the starburst traced in hydrogen recombination emission. The absence of a correlation argues against photons as the source of excitation. Instead, the emission is attributed to shock heating of the molecular material at the face of the cavity (Moorwood et al. \\cite{Moorwood96a}; Marconi et al. \\cite{Marconi}). Mid-infrared spectroscopy is much less affected by intervening extinction than the UV and optical equivalents, with A($\\lambda$)/A$_V$ less than 0.1. Observations of the central region of the galaxy, using the mid-infrared spectrometer SWS (De Graauw et al. \\cite{deGraauw}) and the spectrophotometer PHT-S (Lemke et al. \\cite{Lemke}), both aboard ISO (Kessler et al. \\cite{Kessler}), are therefore very useful to study the nuclear components otherwise hidden by heavy extinction. In Sect.~3.1 we present the results of the search for high excitation emission from the AGN. In Sect.~3.2 we study the properties of the nuclear starburst. In Sect.~3.3 we discuss the dominant nuclear power source. Sect.~3.4 discusses the broad emission and absorption features, tracing the properties of the interstellar medium in and in front of the nucleus. Finally, in Sect.~3.5 we discuss the physical conditions and excitation of the warm molecular hydrogen. ", "conclusions": "The main results of this paper can be summarized as follows: \\begin{itemize} \\item The nuclear starburst is heavily obscured by 36$^{+18}_{-11}$ mag. of visual extinction, as infered from the $[\\ion{S}{iii}]$ 18.7$\\mu$m/33.5$\\mu$m ratio. \\item The excitation of the nuclear starburst is very low, as deduced from excitation indicators $[\\ion{Ne}{iii}]$15.56$\\mu$m/$[\\ion{Ne}{ii}]$12.81$\\mu$m and L$_{\\rm bol}$/L$_{\\rm lyc}$, consistent with an age of at least 5$\\times$10$^6$yrs. Comparison with starburst models implies that at least 50\\% of the bolometric luminosity is powered by the starburst. \\item The very low inferred black hole mass, the very cold mid-infrared to far-infrared colors, and the absence of any free line of sight to the NLR supports the conclusion that the starburst dominates the bolometric luminosity. \\item Our mid-infrared ISO spectroscopy does not provide any evidence for the existence of an AGN in the nucleus of NGC~4945. The only high excitation line detected, the 25.9$\\mu$m $[\\ion{O}{iv}]$ line, is most likely produced in shocks associated with the nuclear starburst. \\item The AGN, detected in hard X-rays, is unusual in not revealing itself at optical, near-infrared and mid-infrared wavelengths. Hence, either the NLR is extremely obscured (A$_V>$160), or UV photons from the AGN are absorbed close to the nucleus along all lines of sight, or the AGN is deficient in UV relative to its X-ray flux. \\item Many ISM solid state and molecular features have been observed with ISO-PHT-S in the 2.4--11.7$\\mu$m range. Most prominent in emission are the PAH features at 3.3, 6.2, 7.7 and 11.2$\\mu$m. The strongest absorption features are those of water ice, CO$_2$ and CO, seen against the nuclear spectrum. These features show striking similarities to the absorption features seen towards the Galactic center. \\item We have studied the physical conditions, excitation and mass of warm H$_2$, combining IRSPEC and ISO observations of 14 transitions. We derive a visual extinction of 20$^{+3}_{-3}$ mag. to the H$_2$ emitting region. From the (0-0) S(0)\\& S(1) lines, we compute a warm (160K) H$_2$ gas mass of 2.4$\\times$10$^7$M$_{\\odot}$, 9\\% of the total gas mass inferred from CO. The excitation diagram is best fitted by a power law of the form dM/dT=4.43$\\times$10$^{15}$ T$^{-4.793}$ M$_{\\sun}$/K. The low excitation temperature of 160K shows Orion-like shocks not to be representative for the entire emission, and fairly normal PDRs to be perhaps more typical. \\end{itemize}" }, "0003/astro-ph0003382_arXiv.txt": { "abstract": "In some scenarios, the peculiar gravitational potential of linear and mildly nonlinear structures depends on time and, as a result of this dependence, a late integrated Sachs-Wolfe effect appears. Here, an appropriate formalism is used which allows us to improve on the analysis of the spatial scales and locations of the main cosmological inhomogeneities producing this effect. The study is performed in the framework of the currently preferred flat model with cosmological constant, and it is also developed in an open model for comparisons. Results from this analysis are used to discuss the contribution of Great Attractor-like objects, voids, and other structures to the CMB anisotropy. ", "introduction": "In some scenarios where linear and mildly nonlinear structures create a time varying gravitational potential, the photons of the Cosmic Microwave Background (CMB) undergo a late Integrated Sachs-Wolfe (ISW) effect. In the absence of any cosmological constant, the partial time derivative of the gravitational potential tends to zero as the universe approaches a flat one and, consequently, the ISW effect tends also to zero. This paper is devoted to the study of some aspects poorly known of the ISW effect: our goal is a detailed analysis of the locations and scales of the subhorizon structures contributing to this effect. We are particularly interested in the scales corresponding to observable objects as voids, the Great wall, et cettera and, by this reason, we will only consider scales smaller than the horizon. We choose an adequate formalism to deal with this analysis. Two scenarios are considered: a flat universe with cold dark matter (CDM) and cosmological constant and an open universe with CDM, they are hereafter referred to as scenarios (or models) I and II, respectively. In both cases the spectrum has been first normalized by the condition $\\sigma_{8}=1$, to consider other normalizations (other $\\sigma_{8}$ values) when necessary. In case I, the spectrum corresponds to cold dark matter (CDM) with $\\Omega_{d}=0.25$, $\\Omega_{b}=0.05$, $\\Omega_{\\lambda}=0.7, $ $h=0.65$ and $n=1$, where $\\Omega_{b}$, $\\Omega_{d}$, and $\\Omega_{\\lambda}$ are the density parameters corresponding to baryonic matter, dark matter, and the cosmological constant, respectively, $h$ is the reduced Hubble constant ($h=H_{0}/100$, $H_{0}$ being the Hubble constant in units of $Km/s.Mpc$), and $n$ is the spectral index of the primordial scalar energy density fluctuations. The scenario II involves CDM and the relevant parameters are: $\\Omega_{d}=0.25$, $\\Omega_{b}=0.05$, $h=0.65$ and $n=1$. Model I is currently preferred according to recent observations of far Ia supernovae and the CMB spectrum (location of the Doppler peak), while model II can account for the abundances of rich clusters and Einstein's rings and, here, it is mainly used for comparisons. As it is well known, the normalization $\\sigma_{8}=1$ does not lead to a good normalization of the CMB angular power spectrum in most cases; in other words, when the $C_{\\ell}$ coefficients are calculated (for $\\sigma_{8}=1$), the resulting values do not fit well with the values observed by COBE, TENERIFE and other experiments. In each scenario, appropriate fits to the observed CMB spectrum correspond to $\\sigma_{8}$ values which are, in general, different from unity. The so-called bias parameter is $b= 1/ \\sigma_{8}$. Since our attention is focused on subhorizon scales, we will estimate the late ISW anisotropy in the $\\ell$--interval $(10, 40)$. For $\\ell < 10$, super-horizon scales would be important (see Kamionkowski and Spergel, 1994) and, then, the spatial curvature could be only neglected in model I; furthermore, the cosmic variance would lead to important uncertainties ($\\Delta C_{\\ell} / C_{\\ell}$ is proportional to $[2/(2 \\ell + 1)]^{1/2}$, see Knox, 1995) and the Sachs-Wolfe effect would be very important. For $\\ell > 40$, the Doppler effect starts its domination hidding other effects as the ISW one. An appropriate linear approach is used in next sections to estimate the $C_{\\ell}$ coefficients for $10 \\leq \\ell \\leq 40$. The method used to do this estimation should facilitate the separation of the ISW effect from other contributions to the angular power spectrum and, moreover, this method should give information about the sizes and locations of the main subhorizon structures contributing to the late ISW effect. The numerical integration of the Boltzmann equation or the computational strategy of Hu and Sugiyama (1994) could be used to perform the analysis of this paper; nevertheless, another appropriate approach --based on a certain approximation to the sources-- is described and used in next sections. In previous papers, it was claimed that some Great Attractor-Like (GAL) objects located between redshifts $2$ and $30$ in open enough universes (without cosmological constant) could account for an important part of the Integrated Sachs-Wolfe (ISW) effect. Arguments in those papers were based on the Tolman-Bondi (TB) solution of Einstein's equations, which was used to estimate both the anisotropy produced by a single GAL structure (Arnau, Fullana \\& S\\'aez, 1994; S\\'aez, Arnau \\& Fullana, 1995) and the abundance of these structures (S\\'aez \\& Fullana 1999). Unfortunately, our TB simulations have some features, as the spherical symmetry and a particular form of compensation, which could affect abundance and anisotropy estimations. By this reason, the mentioned claim should be discussed using a general formalism (not TB solution). It is done in Section 4 as a subsidiary application (in model II) of the formalism described along the paper. In section 2, the method used to compute the angular power spectrum inside the $\\ell$ interval [10, 40] is described. Results are presented in Section 3 and, Section 4 is a general discussion and a summary of conclusions. Finally, some words about notation: whatever quantity \"$A$\" may be, $A_{_{L}}$ and $A_{0}$ stand for the $A$ values on the last scattering surface and at present time, respectively. Simbols $x^{i}$, $\\phi$, $\\vec {v}$, $\\vec {n}$, $\\rho_{_{B}}$, $\\delta$, $a$, $t$, $G$, stand for the comoving coordinates, the peculiar gravitational potential, the peculiar velocity, the unit vector in the observation direction, the background mass density, the density contrast, the scale factor, the cosmological time, and the gravitational constant, respectively. Units are chosen in such a way that the speed of light is $C=1$. Quantities $\\Omega_{0}$ and $\\Omega_m$ are defined as follows: $\\Omega_{0} = \\Omega_{b} + \\Omega_{d} + \\Omega_{\\lambda}$ and $\\Omega_{m} = \\Omega_{b} + \\Omega_{d}$. The comoving wavenumber is $k_{c}$, while $k$ is the physical one. ", "conclusions": "" }, "0003/hep-ph0003212_arXiv.txt": { "abstract": "Recent evidence for neutrino oscillations has revolutionized the study of neutrino masses and mixing. This report gives an overview of what we are learning from the neutrino oscillation experiments, the prospects for the near term, and the bright future of neutrino mass studies. ", "introduction": "How many neutrinos are there? Neutrino counting at LEP of $Z\\to\\nu\\bar\\nu$ decays obtains $N_\\nu=3$ active flavors --- the expected $\\nu_e,\\ \\nu_\\mu,\\ \\nu_\\tau$. However, light isosinglet, right-handed ``sterile\" neutrinos with no gauge boson interactions could also exist. Big Bang Nucleosynthesis determines the equivalent number of massless neutrinos at the time of nucleosynthesis. The bounds inferred ($N_\\nu < 3.2$\\cite{burles}, $N_\\nu < 4$\\cite{lisi}, $N_\\nu < 5.3$\\cite{olive}) depend on which measurements of the $^4$He and D/H abundances are used in the analysis. Neutrino oscillation phenomenology for $N_\\nu=3$ and $N_\\nu=4$ is very different and both options need to be considered at present. ", "conclusions": "We have entered an exciting new era in the study of neutrino masses and mixing. From the SuperK evidence on atmospheric neutrino oscillations, we already have a surprising amount of information about the neutrino mixing matrix (near maximal $\\sin^22\\theta_{23}$ and near minimal $\\sin^22\\theta_{13}$). The SuperK, SNO, Borexino, KamLand, and ICARUS experiments are expected to differentiate among the candidate solar oscillation possibilities and determine $\\sin^22\\theta_{12}$. MiniBooNE will tell us whether a sterile neutrino is mandated. Neutrino factories will study the leading oscillations, determine the sign of $\\delta m_a^2$, measure $U_{e3}$, and possibly detect CP violation. The GENIUS $0\\nu\\beta\\beta$ experiment and the MAP and PLANCK satellite measurements of the power spectrum will probe the absolute scale of neutrino masses. There is a synergy of particle, physics, nuclear physics, and cosmology occurring in establishing the fundamental properties of neutrinos. A theoretical synthesis should emerge from these experimental pillars." }, "0003/hep-ph0003176_arXiv.txt": { "abstract": "We calculate the thermodynamic potential of bubbles of hadrons embedded in quark-gluon plasma, and of droplets of quark-gluon plasma embedded in hadron phase. This is a generalization of our previous results to the case of non-zero chemical potentials. As in the zero chemical potential case, we find that a quark-gluon plasma in thermodynamic equilibrium may contain stable bubbles of hadrons of radius $R \\simeq 1$~fm. The calculations are performed within the MIT Bag model, using an improved multiple reflection expansion. The results are of relevance for neutron star phenomenology and for ultrarelativistic heavy ion collisions. ", "introduction": "\\label{sec.intro} In a previous paper \\cite{us} we calculated the free energy of a bubble of hadrons embedded in an extended quark-gluon plasma (QGP), and of a droplet of QGP embedded in an extended hadron phase, for parameters in the vicinity of the cosmological quark-hadron transition, i.e.\\ the baryon chemical potential were set to zero. In Ref. \\cite{us} we found, as Mardor and Svetitsky in \\cite{marsve}, that the free energy of a hadron bubble of radius $R$ embedded in QGP possessed a minimum at radii of a few fm, even above the phase transition temperature. Thus, hadronization from QGP is strongly enhanced compared to the usual nucleation scenario, where an energy barrier has to be overcome before bubbles of hadrons can form and grow. An important ingredient in these calculations is the density of states of the relevant particles. In Ref.\\ \\cite{us} we advised a modification of the usual expressions arising from the multiple reflection expansion (MRE), and we saw that this modification (the MMRE) yielded more accurately the free energy than the MRE did when compared to a direct (but numerically heavier) sum-over-discrete-states calculation. Our modification of the multiple reflection expansion expressions is physically well motivated, in that it consists of a truncation of the density of states in such a way that we avoid the effects of a negative density of states, present in the usual MRE expressions. In this paper, we generalize the calculations to finite chemical potential, and we shall see that also in the case of non-zero chemical potential, the MMRE produces accurately the thermodynamic potential. We show that, as in the case of zero chemical potential, the thermodynamic potential of a hadron bubble embedded in QGP has a minimum at a radius of $R \\simeq 1$~fm, meaning that it is thermodynamically favorable for a QGP in equilibrium to spontaneously create bubbles of hadrons of about this size. However, the minimum in the thermodynamic potential found for non-zero chemical potential is less pronounced than in the zero chemical potential case considered in Ref.~\\cite{us}. The more general calculations presented in this paper should be relevant also to neutron star phenomenology as well as ultrarelativistic heavy ion collisions. In the following section we give an overview of the basic theory leading to the results of section~\\ref{sec.results}. Finally in Sec.~\\ref{sec.concl}, we summarize the conclusions. ", "conclusions": "\\label{sec.concl} We have seen that the thermodynamic potential of a hadron bubble embedded in quark-gluon plasma, exhibits a minimum at a radius $R \\simeq 1$~fm, even at temperatures somewhat above the bulk transition temperature. Thus, within the model described here, a homogeneous plasma of size larger than a few fermi is an impossibility. Regarding the relevance of these results in connection with current and forthcoming ultrarelativistic heavy ion collisions, we conclude that {\\em if} indeed a quark-gluon plasma is formed in the course of such a collision, then (according to the model considered here) this plasma phase will contain bubbles of hadrons of radius $R \\simeq 1$~fm. {\\em If} such hadronic bubbles form inside the plasma, the observable effects are likely to include some blurring of the plasma signatures, since this reduces the effective plasma volume. We emphasize that these conclusions may well be model dependent, inasmuch the minimum of the thermodynamic potential is of modest depth. However, the phenomenon of spontaneous creation of stable hadronic bubbles in a quark-gluon plasma, does seem to be well established within the the model discussed here \\cite{us,marsve,michael}. The main question is now, whether these hadronic bubbles are of physical nature, or merely an artifact of the model. Certainly, it would improve confidence in these results if other models of QCD, and eventually lattice calculations, were to yield similar results." }, "0003/astro-ph0003219_arXiv.txt": { "abstract": "In the construction of an X-ray selected sample of galaxy clusters for cosmological studies, we have assembled a sample of 495 X-ray sources found to show extended X-ray emission in the first processing of the ROSAT All-Sky Survey. The sample covers the celestial region with declination $\\delta \\ge 0\\deg $ and galactic latitude $|b_{II}| \\ge 20\\deg $ and comprises sources with a count rate $\\ge 0.06$ counts s$^{-1}$ and a source extent likelihood of 7. In an optical follow-up identification program we find 378 (76\\%) of these sources to be clusters of galaxies. It was necessary to reanalyse the sources in this sample with a new X-ray source characterization technique to provide more precise values for the X-ray flux and source extent than obtained from the standard processing. This new method, termed growth curve analysis (GCA), has the advantage over previous methods to be robust, easy to model and to integrate into simulations, to provide diagnostic plots for visual inspection, and to make extensive use of the X-ray data. The source parameters obtained assist the source identification and provide more precise X-ray fluxes. This reanalysis is based on data from the more recent second processing of the ROSAT Survey. We present a catalogue of the cluster sources with the X-ray properties obtained as well as a list of the previously flagged extended sources which are found to have a non-cluster counterpart. We discuss the process of source identification from the combination of optical and X-ray data. To investigate the overall completeness of the cluster sample as a function of the X-ray flux limit, we extent the search for X-ray cluster sources to the data of the second processing of the ROSAT Survey for the northern sky region between $9^h$ and $14^h$ in right ascension. We include the search for X-ray emission of known clusters as well as a new investigation of extended X-ray sources. In the course of this search we find X-ray emission from additional 85 Abell clusters and 56 very probable cluster candidates among the newly found extended sources. A comparison of the X-ray cluster number counts of the NORAS sample with the REFLEX Cluster Survey results leads to an estimate of the completeness of the NORAS sample of RASS I extended clusters of about 50\\% at an X-ray flux of $F_x(0.1-2.4 {\\rm keV}) = 3 \\times 10^{-12}$ erg s$^{-1}$ cm$^{-2}$. The estimated completeness achieved by adding the supplementary sample in the study area amounts to about 82\\% in comparison to REFLEX. The low completeness introduces an uncertainty in the use of the sample for cosmological statistical studies which will be cured with the completion of the continuing Northern ROSAT All-Sky (NORAS) cluster survey project. ", "introduction": "Galaxy clusters are important tracers of the large-scale structure of the matter distribution in the Universe. As the evolution of clusters is closely linked to the overall evolution of the cosmic large-scale structure, important tests of cosmological models can be performed with statistical data on the cluster population. The mass distribution and the spatial clustering of clusters are particularly interesting measures in such studies (e.g. Henry et al. 1992, Bahcall \\& Cen 1992, Eke et al. 1998, Thomas et al. 1998, Borgani et al. 1999). The construction of well defined cluster catalogues and the compilation of their properties is therefore an important task for observational cosmology. Galaxy clusters were first detected and are continued to be cataloged from optical observations of galaxy density enhancements in the sky (e.g. Abell 1958, Abell et al. 1989 (ACO), Zwicky et al. 1961 - 68, Shectman 1985, Dalton et al. 1992, Lumbsden 1992, Collins et al. 1995, Couch et al. 1991, Bower et al. 1994, Postman et al. 1996, Olsen et al. 1999, Scodeggio et al. 1999). X-rays have also successfully been used to detect galaxy clusters and to conduct clusters surveys (e.g. Picinotti et al. 1982, Kowalski 1984, Lahav et al. 1989, Gioia et al. 1984, 1990, Edge et al. 1990, Henry et al. 1992, Romer et al. 1994, Pierre et al. 1994, Ebeling et al. 1996, 1998, Castander et al. 1995, Rosati et al. 1995, 1998, Burns et al. 1996 Collins et al. 1997, Burke et al. 1997, Vihklinin et al. 1998, Scharf et al. 1997, Jones et al. 1998, B\\\"ohringer et al. 1998, De Grandi et al. 1999, Ledlow et al. 1999, Romer et al. 1999). The use of samples of clusters detected and characterized by their X-ray emission for cosmological studies has two major advantages over samples based on optical observations. First, the optical observations (without very extensive redshift measurements) provide only the projected galaxy distribution and not all galaxy density enhancements in the sky are bound, three-dimensional entities. In fact, in the course of the ESO Nearby Abell Cluster Survey (Katgert et al. 1996, Mazure et al. 1996) it was found that of the order of 10\\% of the rich clusters from the catalogue of Abell, Corwin, and Olowin (1989) in the nearby redshift range $z \\le 0.1$ were spurious clusters without obvious clustering peaks in redshift space. For the optical surveys the reliability has improved, however, with the advent of multi-color surveys and machine based matched-filter selection techniques (e.g. Postman et al. 1996, Olsen et al. 1999). Extended X-ray emission from the hot intra-cluster plasma of galaxy clusters is a more clear indication of the presence of a large gravitationally bound mass aggregate since otherwise the hot plasma would have been dispersed immediately. And secondly, the X-ray luminosity is a parameter much more tightly correlated with the mass of clusters than the usual richness parameter measured in the optical (e.g. Reiprich \\& B\\\"ohringer 1999). Thus X-ray emission gives evidence for the presence of galaxy clusters within a certain mass interval. (The correlation of the X-ray luminosity and cluster mass actually shows a dispersion of about a factor of 1.6 if one wishes to determine the mass for a given luminosity - Reiprich \\& B\\\"ohringer, in preparation). The one exception is the case where the X-ray emission is not clearly extended and where the cluster emission could be confused with the emission of an AGN within the cluster or with a possible foreground or background source. This confusion is a problem for a very small fraction of the cluster sources, but in general X-rays are a very useful indicator of a true cluster. In addition projection effects are minimized in X-ray surveys since the X-ray surface brightness is more centrally concentrated than the galaxy distribution. The ROSAT All-Sky Survey (RASS), the only large scale X-ray survey conducted with an X-ray telescope (Tr\\\"umper 1993, Voges et al. 1999), provides an ideal data base to detect large numbers of clusters and to compile an all-sky cluster catalogue with homogeneously applied selection criteria. To exploit this unique data base we are conducting an optical follow-up identification program and redshift survey of RASS X-ray clusters in the northern hemisphere, the Northern ROSAT All-Sky (NORAS) Cluster Survey project. A complementary survey, the REFLEX (ROSAT-ESO Flux Limited X-ray) Cluster Survey, is conducted for the southern part of the RASS (B\\\"ohringer et al. 1998, Guzzo et al. 1999). The NORAS identification program was started in 1992 in a first step with a list of extended X-ray sources extracted from the general source list of the first RASS processing (RASS I; Voges et al. 1992). Apart from the selection for extent the following extraction criteria were used: northern declination, a minimum distance of 20 degrees to the galactic plane, and a minimum count rate of 0.06 cts s$^{-1}$ in the ROSAT broad band (0.1 to 2.4 keV). The criterion of X-ray source extent was chosen for the selection of promising cluster candidates, because early tests have shown that such a sample would be highly enriched (by about 70 - 80\\%) in galaxy clusters. Contrary to the cluster selection scheme used for the REFLEX Survey which is based on the correlation of X-ray sources with galaxy overdensities, the present sample selection is purely based on X-ray information. With this different bias the NORAS Survey has also the potential to find more distant and possibly ``opitcally dark'' clusters. The identification of these sources is now complete. In this paper we present a catalogue of the X-ray properties of the 378 cluster sources and 117 non-cluster sources of the primary candidate list. An accompanying paper by Huchra et al. (1999) provides a detailed catalogue of the optical identifications and redshift measurements of this sample, and scientific aspects of this survey are discussed in a paper by Giacconi et al. (1999). While this survey was ongoing some of the brighter sources of this sample as well as some X-ray emitting Abell and Zwicky clusters were spectroscopically observed for the ''BCS program'' (Ebeling et al. 1998) by Allen et al. (1992) and Crawford et al. (1995, 1999). The region with the deepest exposure in the northern RASS, the north excliptic pole with exposure times ranging from 2000 to over 40000 sec, is also the subject of a dedicated survey which has identified all X-ray sources (c.f. Henry et al. 1995; Gioia et al. 1995; Bower et al. 1996; and Henry et al. 1997). Further studies on the X-ray properties of a sample of these extended sources in $2\\deg \\times 2\\deg$ sky fields extracted from the RASS revealed, that in the first standard processing of the RASS the count rate and the extent of the cluster sources are severely underestimated (see also Ebeling et al. 1996, DeGrandi et al. 1997). Therefore a detailed reanalysis of the sources in the present sample was necessary. Here we also report the results of the detailed reanalysis of the sources using a new X-ray source characterization technique. In 1996, a second revised processing of the RASS (RASS II; Voges et al. 1999) with greatly improved attitude quality and with a fully merged photon data base became available. Our reanalysis is based on these new data. Since the incomplete assessment of the X-ray count rate and extent in RASS I not only leads to an underestimate of the X-ray fluxes for extended sources but also to an incompleteness of the sample extracted from the data base with certain limiting parameters, we have also used the new RASS data base to explore the incompleteness of the present cluster sample in terms of a flux-limited X-ray selected sample of galaxy clusters. In this study we selected a subregion covering the right ascension range from 9$^h$ to 14$^h$. In a first step we use the Abell cluster catalogue to study the completeness provided by the RASS I extent criterion. We further study the prospects of finding more clusters with a more comprehensive extent criterion based on the new X-ray source analysis technique. This study also points the way to a more complete selection of galaxy clusters from the RASS X-ray sources. We now apply this algorithm in the ongoing NORAS cluster redshift survey. Since we will probably not be able tp rapidly complete the identifications of the newly found sources, we decided to publish the first part of the survey for which identifications are now complete and redshifts are nearly (all but 9) complete. The main emphasis here is not to publish a catalog of a complete, flux-limited sample, but to compile a cluster catalogue with reliable identifications based on a wealth of X-ray and optical data which are included in the identification process in a comprehensive way. The present sample contains many newly found objects, some of which are interesting targets for further astrophysical studies. The paper is organized as follows. In Section 2 we summarize the properties of the primary RASS I source list of extended sources, and in Section 3 we describe the techniques used to reanalyze the X-ray properties of the sample sources. In Section 4 we present the X-ray source catalogue with detailed X-ray properties of the 495 sources. Major X-ray properties of the sources which help in the identification of the objects are discussed in Section 5. The completeness of the sample is addressed in Sections 6 and 7 where we report the results of a rigorous search for X-ray emission from all ACO clusters and search for more extended X-ray sources with an improved analysis algorithm in a test region ranging in right ascensions from 9$^h$ to 14$^h$. In Section 8 we compare the present results to the previous ROSAT Bright Cluster Survey by Ebeling et al. (1998). Section 9 provides a summary and conclusions. Throughout the paper we are using a Hubble parameter of $H_0 = 50$ km s$^{-1}$ Mpc$^{-1}$ and $h_{50} = H_0 / (50$ km s$^{-1}$ Mpc$^{-1})$ and further $\\Omega _0 = 1$ for the density parameter and $\\Lambda _0 = 0$ for the cosmological parameter. \\section {The RASS I list of extended sources} During the ROSAT mission the first All-Sky Survey was conducted with an X-ray telescope (Tr\\\"umper 1992, 1993). The RASS was performed over a period of six months from August 1990 to January 1991 with two follow-up auxiliary survey missions carried out to fill the gaps in the survey in February and August 1991. The first processing of the survey (RASS I) provided a source list of 49441 sources (Voges et al. 1992, 1996). For this first analysis the survey data received were sorted into one of 90 2-degree wide strips (oriented in the direction of constant ecliptic longitude) while the satellite was still scanning the sky. As a consequence of this, strips are overlapping in regions outside the equator. Photons are exclusively sorted into only one of the overlapping regions. Therefore the exposure time across the strips is quite homogenous, but no advantage can be taken of the high total exposure in the ecliptic pole regions. The survey product resulting from this first processing of the RASS will be referred to as RASS I data base. (The situation is different in RASS II, the second processing, where the $6.4\\deg \\times 6.4\\deg$ sky regions contain the full exposure data from the RASS). The RASS was conducted with the X-ray telescope (Aschenbach et al. 1988) and Position Sensitive Proportional Counter (PSPC; Pfeffermann et al. 1986) providing a high sensitivity and a very low internal background. A histogram of the exposure time distribution of the NORAS survey area in RASS I is shown in Fig. 1 (where it is also compared to the exposure distribution in RASS II). For this statistic we use the maximum exposure in any of the strips for those cases where a sky pixel is covered by several survey strips. The mean and median exposure times are 397 and 402 sec, respectively. \\begin{figure} \\plotone{fig1.ps} \\caption{Exposure time distribution of the RASS I for the NORAS survey area. For comparison the distribution of the exposure time in the RASS II involving the merged photon data from all survey strips is also shown as broken line.} \\end{figure} The source detection procedure was based on detections using two alternative sliding window techniques plus a subsequent evaluation of the source detection significance and quality based on a maximum likelihood method (Voges et al. 1992, for aspects of the maximum likelihood method see also Cruddace et al. 1991). Only sources with a likelihood of detection larger than $L = 10$ were accepted into the RASS I source list. (The likelihood value here and throughout the paper is defined as $L = -\\ln P$, where P is the probability for a spurious source detection -- or a spurious extent detection in the case of the extent likelihood.) This threshold was chosen such that an estimated fraction of less than 1\\% spurious sources enter the source catalogue. The detections are based on the source counts in the broad ROSAT PSPC energy band covering the detector channels 11 -- 240 which roughly corresponds to an energy range of 0.1 to 2.4 keV. Further qualities of the sources evaluated during the maximum likelihood assessment in three energy bands comprise the source count rate, an estimated source extent in excess of the broadening of the sources due to the telescope-detector point spread function (PSF), and two hardness ratios based on the counts measured in the soft (channel 11 -- 40) and hard energy band (channel 52 -- 201) or in the hard band 1 (channel 52 -- 91) and hard band 2 (channel 92 -- 201), respectively. The source extent which is of special importance here was determined within the maximum likelihood analysis by assuming for the source image shape a convolution of two two-dimensional Gaussian functions for the PSF and the source shape (Voges et al. 1999), respectively. The result of this analysis is then a value for the excess extent in terms of a $\\sigma$-radius of the second Guassian. In this approximation the Gaussian wings are less extended than both the wings of the PSF and the wings of a King-type surface brightness model (e.g. Cavaliere \\& Fusco-Femiano 1976, Jones \\& Forman 1984). This is one, probably minor, reason for the effect that some of the X-ray flux in the outer X-ray halos is underestimated in the RASS standard analysis. \\begin{figure} \\plotone{fig2.ps} \\caption{A comparison of the distribution of the values for the extent and extent likelihood for a set of RASS sources identified with ACO clusters, stars, and AGN. The data are taken from the ROSAT Bright Source catalogue compiled from RASS II and the cross identifications with optical catalogues as described by Voges et al. (1999). The coincidence radius used for the cross-identification is 5 arcmin. The vertical and horizontal lines give the extent limit of 25 arcsec and the extent likelihood limit of 7. The fraction of sources for each category located in the different quadrants is given in each quadrant. The upper number in each quadrant gives the fraction of all ACO clusters falling into this quadrant, the middle number gives the corresponding fraction for stars and the lower number the fractions for AGN. The fractions for each class summed for all quadrant adds up to unity, thus, for example, less than 2-3 \\% of the non-cluster sources are found in the upper right quadrant. The selection criteria for the present sample correspond to the upper right quadrant which is highly enriched in ACO clusters. } \\end{figure} First tests of the source quality parameters in RASS I and ready identifications with existing source catalogues showed that many clusters of galaxies featured significantly extended X-ray emission in the RASS. It was known from previous X-ray surveys (e.g. the Einstein Medium Sensitivity Survey, e.g. Gioia et al. 1994, Stocke et al. 1994) that slightly more than about 10\\% of the X-ray sources at the depth of the RASS should be galaxy clusters emitting in X-rays (see also B\\\"ohringer et al. 1991). Therefore it was clear that an X-ray sample highly enriched in galaxy clusters could be obtained by selecting those RASS sources featuring a significant source extent. For example, this is demonstrated by a comparison of the extent properties of the X-ray sources for ACO clusters, stars, and AGN taken from the ROSAT Bright Source catalogue and the correlation with optical catalogues from Voges et al. (1999) shown in Fig. 2 for RASS II data (a comparable figure is also shown in Ebeling et al. 1996 for RASS I results). Even though the present study is concerned with RASS I results we are showing a statistic for RASS II in Fig. 2 because there is no principle difference and there is a larger data base of correlations with catalogued objects available for RASS II. About half of the galaxy clusters occupy an almost exclusive parameter space characterized by an excess source extent larger than 25 arcsec with a reasonably high extent likelihood (with a value of $L=7$). Only about 2-3 \\% of the non-cluster sources are found in this region. Since galaxy clusters account for about $10 - 15\\%$ of all X-ray sources we can expect a contamination of the order of $20 - 30\\%$ by non-cluster sources if the sample is selected from the upper right quadrant. This is approximately what is found below. That this small value of 25 arcsec for the excess extent radius threshold shows a significant effect is somewhat surprising, since the mean half power radius of the survey point spread function is 70 arcsec, much larger than this threshold. This can be explained by the fact that in the RASS analysis likelihoods are calculated separately for each photon before they are summed and therefore each photon can be weighted by its own PSF according to the place in the detector where it was registered. In this way photons registered in the central part of the detector with a half power radius of the PSF of 15 - 20 arcsec give a high weight to the maximum likelihood analysis. This makes the RASS source analysis very sensitive to the recognition of small deviations from the expected shape of point sources. In the following analysis we will not make use of the information on the detector positions of individual photons. This is a disadvantage when compared to the standard RASS maximum likelihood analysis, but other advantages more than compensate for this. A first search for X-ray selected clusters was made with RASS I sources flagged as extended. The selection criteria were as follows: For the extent parameters a minimum threshold of 25 arcsec for the extent radius and a minimum value of 7 for the extent likelihood was chosen, as indicated by the dividing lines in Fig. 2. Further a lower count rate limit of 0.06 cts s$^{-1}$ in the ROSAT broad band was set and the sky area was restricted to the region $\\delta \\ge 0\\deg$ and $|b_{II}| \\ge 20\\deg$. The sources at the count rate threshold are thus typically characterized by about 25 source photons. This leads to a source fraction of 76\\% galaxy clusters among the sources selected. The advantage of this approach is that it yields a low fraction of contaminating sources, provides an effective way to detect galaxy clusters, and involves relatively simple selection criteria. The disadvantage is that the selection by source extent is much more difficult to quantify and to model than for example a purely flux-limited selection technique. In total 537 sources matching the selection criteria were extracted from the RASS I data base. 40 of these sources have been found to be detections of secondary maxima or fragments of clusters which are already in the list due to a detection at the main maximum. The largest fraction of these fragment sources is located in the very extended, diffuse emission region of the Virgo cluster (B\\\"ohringer et al. 1994). The fragment sources were removed from the list after a careful check that they are not associated with another distinct X-ray source in the line-of-sight. We have also excluded from the present catalogue the two detections in the Virgo cluster at the position of M87 and M86, because a useful flux measurement in the Virgo region requires a more detailed approach. Thus we report results for these parts of Virgo separately. In the following we will therefore discuss the analysis and identification of the remaining 495 sources. Early 1996 a new ROSAT Survey product, RASS II, became available at MPE. This version which is based on a greatly revised attitude solution for the pointing of the satellite during the survey and also uses a much more stringent quality threshold for the times with acceptable attitudes was used to create a new RASS II source list from which the RASS Bright Source Catalogue (RASS BSC) was created (Voges et al. 1999). In this survey product the data are sorted in 1378 sky fields with sufficient overlap ($\\sim 0.23$ degrees) to guarantee an undiscriminating source assessment in the boundary regions. Each field now contains all the photons registered for this part of the sky during the entire survey. The resulting exposure distribution in the NORAS survey area is also shown in Fig. 1. As expected this distribution features a tail of high exposures up to about 40000 s. The reanalysis of the X-ray sources of the present sample makes almost exclusively use of the RASS II data base. \\section {Reanalysis of the X-ray source properties} To reanalyse the sources we apply a novel technique that is essentially based on measuring background-corrected source counts as a function of a growing circular aperture and checking for saturation to determine the observed source counts. The growth curve of the counts as a function of aperture radius is also used subsequently to analyse further source properties. We therefore term this method the growth curve analysis, GCA. We preferred to apply this method over techniques applied in earlier studies. The Steepness Ratio Technique used in De Grandi et al. (1997) has some similarity to the present analysis, but makes only restricted use of the available photon data in only extracting the source counts in aperture radii of 3 and 5 arcmin. A comparison shows that the uncertainties in the determined count rates are usually higher for that technique than for GCA. Our preference of the GCA method over Voronoi-Tesselation and Percolation (Ebeling et al. 1996, 1998) is due to the fact that the GCA technique is simple to reproduce in models and simulations, the resulting count rates are quoted for a known aperture radius for each source allowing a better assessment of the results in subsequent modeling, and the GCA technique provides a set of very essential diagnostic plots which make the interactive evaluation of the reliability of the GCA results easy and transparent. In addition the VTP technique needs two counteracting steps to correct for the unobserved flux. The first step relies on an extrapolation based on the assumption of spherically symmetric sources and leads to a significant overcorrection which is then compensated in a second step with an {\\it a posteriori} recalibration based on a comparison with pointed data. The present method achieves a good agreement with pointed data in one relatively minor {\\it ab initio} correction in a first step as shown below. The presentation of more details and tests of the GCA method is planned for a future publication (B\\\"ohringer et al. in preparation), while the essential features of this method are described in the following. The reanalysis of the X-ray properties was conducted for all 495 X-ray sources in the sample using RASS II data in fields of $1.5\\deg \\times 1.5\\deg$ centered on each X-ray source. For 17 nearby clusters featuring a large extent the analysis is performed in larger fields of $4\\deg \\times 4\\deg$ or $8\\deg \\times 8\\deg$. For 7 sources, where the exposure in RASS II is less than 70 sec, data fields were extracted from RASS I which features a higher exposure for these cases. The reason for the reduced exposure in RASS II is the tight quality constraint which leads to the rejection of some RASS photon data in RASS II as compared to RASS I. The 7 fields of RASS I used here were carefully checked and did not show any peculiarities as e.g. double or otherwise distorted images of bright sources which would indicate a problem with the attitude control during the observation. The primary data set used for each field consists of a photon event file containing all data of the photons registered in the field area and the corresponding exposure map, providing the exposure time as a function of sky position with a resolution of 45 arcsec pixels. The exposure maps include a broad band correction for vignetting and the effect of the shadowing of the support structure of the PSPC window. (The possible difference between the broad band vignetting correction and the proper correction for the specific source spectrum introduces an error no larger than 2\\% and no further correction is applied). The photon event files provide information on the sky position and energy channel as well as the time and the detector position for each registered photon. In the following analysis only the first two parameters are used. The three energy bands defined for our data reduction procedure are identical to those used in the RASS analysis (Voges et al. 1999): broad band, soft band, and hard band. For the source count rate determination and the shape characterization we use exclusively the photon counts in the hard energy band. In this band the soft X-ray background is reduced to about one fourth, while -- depending on the value of the interstellar column density -- 60 to 100\\% of the cluster emission is detected. Therefore the analysis in this energy band provides the highest signal-to-noise ratio and the most reliable count rates. Another quite important aspect of the choice of this energy band is that it minimizes the contribution of contaminating sources to the count rate. Since the majority of all sources in the RASS are softer than the cluster X-ray sources, their contribution to the hard band is usually less significant than the contribution to the broad band counts. \\subsection{Source position and count rate} Prior to the evaluation of the source count rate the source center position and the sky background brightness is determined. The input field centers chosen are the X-ray positions provided by the maximum likelihood technique of RASS II. These positions are not optimal for extended sources and in particular extended sources are sometimes multiply detected in the RASS. Therefore a redetermination of the source position is performed based on a moment method which determines the two-dimensional ``center of mass'' of the photon distribution within an aperture of 3 arcmin around the input value for the center. This procedure is iterated with the newly found center position until the process converges to a stable center position. The small aperture of 3 arcmin gives a high weight to local maxima. We check all the center positions interactively and correct those cases in which this method has settled on a secondary maximum or where the local maximum is obviously offset from the large scale symmetry of the cluster. Those 5 sources are marked in the catalogue. We have also applied the moment method for the determination of the center position using larger apertures (5 and 7.5 arcmin). In 7 cases we preferred to quote these centers in the catalogue. Also these cases are marked. \\begin{figure}[h] \\plottwo{fig3a.ps}{fig3b.ps} \\caption{Distribution of the photons detected by the ROSAT PSPC in the hard band ($\\sim $0.5-2.0 keV) in the regions of the X-ray cluster sources RXCJ0004.9+1142 and RXCJ0020.6+2840. The ring segments of the area used for the background estimate and the source region as determined from the analysis are marked. The photon field extracted from the ROSAT All-Sky Survey as shown here covers an area of 90 by 90 square arcmin. The ring segment of the source field of RXCJ0020.6+2840 marked by a cross is excluded from the background determination due to contamination.} \\end{figure} \\begin{figure}[h] \\plottwo{fig4a.ps}{fig4b.ps} \\caption{Radially integrated count rate for the cluster sources RXCJ0004.9+1142 and RXCJ0020.6+2840 in the hard band ($\\sim $0.5-2.0 keV). The two techniques used to determine the total observed source count rate are shown: the vertical dashed line shows the limiting radius outside which the $1\\sigma$ uncertainty of the signal increases faster than the signal itself; the horizontal line shows the fit to the saturation plateau (as explained in the text below). Also shown as dotted line is the best fit of a $\\beta$-model profile convolved with the PSF with a core radius of 0.5 arcmin and a normalization of 0.158 cts s$^{-1}$ for RXCJ0004.9+1142 and 2.5 arcmin and 0.360 cts s$^{-1}$ for RXCJ0020.6+2840, respectively. Note that the actual fits were performed to the differential count rate profiles and not to the cumulative ones as shown here. Thus the deviations of the fits at large radii have a low significance. They are often found and may indicate irregular structure at the outskirts of many clusters.} \\end{figure} The background of the field is then determined from a ring area centered on the source with an inner and outer ring radius of 20 arcmin and 41.3 arcmin, respectively, as shown in Fig. 3. The inner ring size has been chosen such that it is outside the outer radius of the X-ray emission for the majority of the clusters. The outer radius is chosen to make almost full use of the $1.5 \\times 1.5$ degree fields. This large background area ensures that the number of photons used for the determination of the background surface brightness is large and introduces an almost negligible Poissonian error into the source flux determination. There are 17 nearby clusters in the sample which exceed the inner background ring radius in size. For these clusters larger fields have been extracted from RASS II and larger background rings have been used in the analysis. The ring is subdivided in 12 sectors. The photons in each of the three energy bands in all sectors are counted and the exposure time for each photon position is obtained from the exposure map. The count rate in each sector and the surface brightness are then calculated by averaging in count rate: \\begin{equation} C = \\sum _i {1 \\over t_i} , \\end{equation} \\noindent where $C$ is the count rate, $t_i$ is the exposure time at each photon position, and the summation is over all photon events in the sector. The surface brightness is obtained by division with the sector area. An uncertainty for the surface brightness in each sector is calculated from Poisson statistics. To avoid that discrete sources located in the background ring are included in the background estimate, the median of the sector count rates is determined and sectors featuring a larger than $2.3\\sigma$ deviation from the median are discarded from the further calculations. Even though sources are only expected to cause large enhancements, we also exclude sectors that have count rates which are too low by more than $2.3\\sigma $, since some of the positive deviations are due to fluctuations, and to avoid a negative bias in the background the negative fluctuations have to be discarded for reasons of symmetry. The clipping threshold of 2.3 $\\sigma$ guarantees a successful removal of sources with typical count rates above about 0.04 s$^{-1}$ which would otherwise introduce a typical error $\\ge 1\\%$ in the background determination. The chosen clipping threshold leads in general to the clipping of not more than 1 - 3 sectors which preserves most of the background area for averaging resulting in a small photon statistical error in the background of typically about 5\\%. This also shows that the variations in the background on that scale are generally small and hardly larger than what is expected from Poisson statistics. The procedure of the background determination is illustrated in Fig. 3. For the data in Fig. 3a there is no interference of background sources and the background is smooth enough that all sectors were included in the background measurement, while for the data in Fig. 3b one of the sectors had to be discarded due to the presence of a significantly disturbing source. The two figures show the photon distribution of the hard band counts for each source field. The two sources, which have net source counts of 75.2 and 127.9, respectively, will be used in the following to illustrate the further analysis. (These two sources selected as the first sources in the list illustrating the features we like to show: RXCJ0004.9+1142 is the first source in the list with less than 100 photons which features a small but significant extent and RXCJ0020.6+2840 is the first source showing the search for the plateau in the count rate in the presence of steps in the plateau region, see below). The cumulative source count rate as a function of radius starting from the earlier determined central position is then found by integrating the source counts in concentric rings outwards while subtracting the background contribution. The integration is performed using a ring-width of 0.5 arcmin (a reasonable resolution for the given PSF of the instrument). The source count rate is determined for each ring by weighting each photon with the local exposure time according to eq.(1). The integration is performed for the three selected energy bands. The results are count rate profiles as shown in Figs. 4a and 4b. The uncertainty corridors in these count rate profiles resulting from Poisson photon statistics are indicated as dashed lines in the figures. They also include the Poisson error of the background determination. In most of these cumulative profiles the count rate levels off to a plateau value which gives the total observed source count rate. The total observed count rate is determined in the automated source characterization program in two alternative ways. In the first approach we determine the radius outside of which the source signal increases less than the $1\\sigma$ uncertainty in the count rate. This radius, which we call the outer radius of significant X-ray emission, $R_x$, is indicated by the vertical dashed line in Figs. 4a and 4b. The count rate with its statistical uncertainty at this radius provides the value of the significantly detected count rate of the source. The radius $R_x$ is shown for the two sources as the inner circle in Figs. 3a and 3b . Alternatively the total source count rate is measured by getting an estimate of the plateau level. In the simplest case it is the average of the flat plateau outside $R_x$. In practice we determine the mean value of the plateau as well as the slope by means of a linear regression method for the profile part outside $R_x$. If the slope of the plateau is less than $0.8$\\% of the total count rate per arcmin radius, the plateau value is accepted. If the plateau is decreasing the count rate is determined from the mean of three bins around $R_x$. If the plateau is increasing, another effort is made to find the best flat part of the plateau by iteratively excluding the outermost and in a second step also some of the innermost bins. This procedure helps in excluding an outer rise of the count rate profile due to a neighboring source or by skipping a few bins if the count rate curve has not completely saturated to a plateau at $R_x$. Both effects can be seen in Fig. 4b, where the outer radius of the considered plateau region before a secondary rise of the profile is indicated. The source analysis is checked in each case on the basis of diagnostic plots as those shown in Figs. 3 - 6. About 35\\% of the plateaus can be characterized by the first step and about 83\\% by the further iterative trials. There is a residual fraction of about 17\\% of the sources for which no satisfactory plateau can be established automatically often as a result of contaminating nearby sources. These cases are analysed individually. For all sources we determine the radius, $R_{out}$, out to which the plateau count rate was measured by simply following the profile until the plateau value is reached. For the count rates and their uncertainties quoted in the present catalogue we have adopted the following approach. For the count rates we take the results from the fitted plateau values, which are in general higher by a few percent than the count rates determined at $R_x$. This is mainly due to an insignificant further rise of the cumulative count rate profile beyond $R_x$. This rise is in most cases much smaller than the statistical error in the count rate. The error in the count rate is then calculated from the root mean square of the shot noise error and the deviation of the plateau value from the value at $R_x$. Taking the root mean square would be justified for errors which are statistically independent and Gaussian distributed. Since the second error is a systematic deviation, this does not apply in this case. Nevertheless this is for the present case a practical approach which integrates the two errors by putting a larger weight on the larger one of the two uncertainties. The correction of the measured count rate of the source to a total count rate is discussed below. \\subsection{Spectral hardness ratio and source extent} For the determination of the spectral hardness ratio the count rate in the soft band is also determined for the same radius as for the hard band value of $R_x$. The hardness ratio, HR, used here as well as in the RASS data base is defined as \\begin{equation} HR = {H - S \\over H + S} , \\end{equation} where $H$ is the hard band and $S$ the soft band source count rate. The expected values for the hardness ratio for cluster sources is roughly in the range $0 - 1$ as shown in Fig. 8. The source extent is addressed in two ways: quantifying the source size and testing the probability that the extent is real, respectively. In the first analysis a King profile with the surface brightness distribution, \\begin{equation} S_x(R) = S_0 \\left(1 + {R^2 \\over r_c^2} \\right)^{-1.5\\beta +0.5} , \\end{equation} (where $R$ is the projected radial distance from the soure center and $r_c$ is the core radius of the X-ray surface brightness distribution) convolved with the averaged survey PSF (as calculated by G. Hasinger from the ROSAT XRT/PSPC PSF averaged over the detector area with a correction for the vignetting effect) and azimuthally integrated is fitted to the differential count rate profile. For the profile parameters a fixed value $\\beta = 2/3$ is taken according to the most typical value found in X-ray cluster observations (e.g. Jones \\& Forman 1984). In the $\\chi^2$ fit the core radius is varied in steps of 0.5 arcmin, and the normalization is a free fitting parameter. Due to the very low count statistics we have thus limited the fitting parameters per step to one, the normalization, by choosing the most common value for $\\beta$. The results for the two example sources are shown in Figs. 5a and 5b, and the best fitting profiles are indicated by dotted lines in Figs. 4a and 4b. Note that, while the figures show fits to the cumulative profiles for a better diagnosis of the results, the actual calculations are conducted for the differential profiles to assure statistical independence of the count rates in rings as required by the $\\chi^2$ fitting method. Together with the best fitting value we also keep the minimum radius which is still consistent within the $2\\sigma$ uncertainty limit. Comparing the $\\chi^2$ values to the 1, 2, and 3$\\sigma$ limits shown in Fig. 5 we find that for the two examples the first source is only marginally extended ($2\\sigma$ result), while the second source features a clear and large extent. \\begin{figure} \\plottwo{fig5a.ps}{fig5b.ps} \\caption{Evaluation of the source extent by means of a $\\chi^2$ fit of a King profile convolved with the mean RASS PSF. The $\\chi^2$ value is determined for varying core radii in steps of 0.5 arcmin. Here the best fitting core radii are found to be 0.5 and 2.5 arcmin. Only the second source is found by this test to be clearly extended with high significance. The horizontal dashed lines show the 1, 2, and $3\\sigma$ uncertainty limits, respectively.} \\end{figure} The second, more sensitive method is used as a test for the probability that the source has an extent at all. For this we use a Kolmogorov-Smirnov test comparing the expected cumulative count rate profile of a point source including background with the given instrument PSF with the radially sorted, cumulative, and unbinned photon counts out to a radius of 6 arcmin. This radius is larger than the 90\\% power radius of the survey PSF and thus provides enough leverage to display the deviations of extended sources, but is small enough to minimize the influence of possible background errors. This test does not depend on the assumption that the uncertainties are Gaussian distributed. Since the pure photon counts also contain the background counts the previously determined background surface brightness has also to be added to the expected point source profile. Examples of the expected point source profile and measured curves are shown in Figs. 6a and 6b. The background contribution to the expected profile is also indicated and we can see that it is generally a minor contribution at these small source radii. Tests with known point sources have shown that the misclassification of point sources as extended sources is generally less than about 5\\% if we take an upper limit for the KS probability of 0.01 to classify a source as extended (see B\\\"ohringer 1999, in preparation). For the two examples we find probability values of $0.005$ and $1.3 \\times 10^{-14}$, respectively, and therefore both sources will be classified as extended in the catalogue. In the following we will use these results in the form of the extent parameter defined as $P_{ext} = -\\log _{10}$(KS probability). \\begin{figure} \\plottwo{fig6a.ps}{fig6b.ps} \\caption{Test for the probability of a source extent by a KS test comparing the cumulative count rate profile out to a radius of 6 arcmin with the expected profile for a point source plus background. Both profiles are normalized to 1 at the outer radius. The long dashed curve gives the expected point source profile plus background and the dotted curves show the effect of a 30\\% background variation. The dotted-dashed curve indicates the background contribution.} \\end{figure} \\subsection{Deblending and analysis of very extended sources} The visual inspection of the diagnostic plots of the GCA results for all the sources showed that in 19 cases the source analysis suffered from the blending of the cluster source with another nearby source probably not associated with the intracluster X-ray emission. These sources were scheduled for another analysis including a deblending technique. The correction by deblending was performed for all the sources where the contamination was clearly recognized as due to point sources. Tests show that a single contaminating source is usually easily recognized if its contribution to the total hard band count rate is larger than 5 - 10\\% and if the source is outside the central 3 arcmin radius. The deblending is not applied to irregular clusters or clusters with substructure where the non-symmetric emission region is most probably part of the diffuse intracluster X-ray emission (and not likely to be due to a point source). In this second analysis the source region is divided in two sets of twelve sectors for the radial region 3 to 8 and 8 to 15 arcmin. The variation of the surface brightness in the different ring sectors is analyzed in a similar way as done for the background ring sectors. Contaminating sources are best flagged by selecting those sectors with a more than $3.5\\sigma$ deviation from the median. For weaker sources this detection threshold corresponds roughly to sources with fluxes larger than $3 \\times 10^{-13}$ erg s$^{-1}$ cm$^{-2}$ and in general guarantees the deblending of sources with contributions larger than 10\\%. The clipping technique is illustrated in Figs. 7a and 7b. In the further analysis the marked sectors are interpolated, that is, they are assigned a value for the surface brightness equal to the mean of the remaining sectors. The automated clipping works well for most sources but in about 30\\% of the cases the clipping either did not remove the contaminating source completely or removed in addition other parts of the cluster. For these cases we preferred to determine the count rate of the contaminating source directly (with the analysis centered directly on the contaminating source) with the same deblending algorithm and subtracted it from the blended source count rate. The 19 sources that required a deblending are marked in the catalogue. Since the outer radius of the region in which the source profiles are analyzed is fixed in the automated source analysis routine, and since the cut out regions per source are restricted to a size of $1.5\\deg \\times 1.5\\deg$ , some nearby clusters are too extended to be covered completely by the analysis. For these clusters, 17 in total, larger files of photon data from RASS II were requested with fields covering $4\\deg \\times 4\\deg $ or even $8\\deg \\times 8\\deg$ around the source. To these data the same source analysis was applied as described above with an extended radial range. The clusters that required a reanalysis in a larger sky field are also marked in the catalogue and can also be recognized by their large values of $R_x$ and $R_{out}$. \\begin{figure} \\plottwo{fig7a.ps}{fig7b.ps} \\caption{The left figure shows the photon distribution of a cluster X-ray source which is clearly contaminated by a point source. The small circle near the point source marks the sector which was excised from the source analysis to deblend the contaminating source. The right figure shows the result for the count rate analysis. The upper dashed curve gives the cumulative source count rate without deblending. The solid curve shows the result after the sector with the source was cut out and interpolated. The two dotted curves show again the $1\\sigma$ Poisson error for the count rate including the effect of the interpolation. The horizontal short dashed curve indicates the plateau value determined and the vertical dashed curve shows the outer radius of significant X-ray emission from the source.} \\end{figure} \\subsection{Flux and luminosity determination} To determine the flux and the luminosity of a source we first obtain the value of the interstellar hydrogen column density as measured at 21cm (Dickey \\& Lockman 1990, Stark et al. 1992) for the direction of the source by means of the EXSAS software package routine (Zimmermann et al. 1994). The flux is then determined in a first step by calculating the conversion factor from count rate to flux for a source with a thermal spectrum and a temperature of 5 keV (based on a modern version of the radiation code by Raymond \\& Smith 1977), a metal abundance of 0.3 of the solar value (Anders \\& Grevesse 1989), a redshift of zero, and an interstellar absorption according to the measured 21 cm value (Dickey \\& Lockman 1990) and the absorption tables of Morrison \\& McCammon (1983). We calculate the flux in the nominal ROSAT energy band: 0.1 - 2.4 keV. All fluxes and luminosities quoted in this paper and in the catalogue refer to this energy band. Fig. 8a shows the conversion factors as a function of absorbing column density for the three plasma temperatures 2.5, 5 and 8 keV. We note that the main dependence is on the column density. The variation with temperature makes a difference of less than $7\\%$ in the temperature range (2 - 10 keV). Only for temperatures below 1.5 keV larger corrections occur, which applies for the smallest groups of galaxies in the sample. The dependence on the metal abundances is even less, about 1\\%. Prior to any further knowledge about the redshift and the nature of the cluster source the estimated flux value represents a good first approximation. Note, however, that for non-cluster sources a different spectral shape and thus a different conversion factor is expected. Once the redshift of the cluster source is known we can determine its luminosity. The luminosity is also calculated for the ROSAT energy band. The calculation is performed iteratively. In a first step we calculate a trial luminosity from the estimated flux value and the redshift and use its value to estimate a cluster plasma temperature using the X-ray luminosity-temperature relation of Markevitch (1998) \\begin{equation} T_x = 2.34~ L_{44}^{1/2} \\times h_{50}~~~~~ , \\end{equation} where $T_x$ is in keV, $L_{44}$ is the X-ray luminosity in units of $10^{44}$ erg s$^{-1}$ (in the 0.1 - 2.4 keV band). We make use of the relation which was derived by Markevitch without any correction for cooling flows in $L_{44}$ and $T_x$. This applies to our case since we are only dealing with integral count rates and average spectral properties. (Note that we have approximated the exponent of 1/2.02 found by Markevitch by 1/2). The temperature estimate allows the calculation of a new count rate-flux conversion factor for which we now also take the redshift of the source spectrum into account and calculate the source rest frame value for the X-ray luminosity (which involves the equivalent to the cosmic ``K-correction''). The K-correction term for example increases up to about 6\\% out to a redshift of $z = 0.3$ for clusters with a temperature of about 2 keV and up to about 15\\% for clusters with 10 keV. The iteration is repeated twice but found to actually converge to the final solution in the first step. The final rest frame X-ray luminosity is the value quoted in the catalogue. \\begin{figure} \\plottwo{fig8a.ps}{fig8b.ps} \\caption{Flux conversion factor and hardness ratio as a function of the galactic hydrogen column density, $N_H$, for three different plasma temperatures: 2.5, 5, and 8 keV. } \\end{figure} We have compared our results for the count rate to flux conversion based on EXSAS software and special programs using a similar data base and radiation code as EXSAS with results obtained from XSPEC, the online PIMMS software, and the conversion factors used in Ebeling et al. (1998). The differences are always less than $3\\%$. Thus the use of different flux evaluation software does not constitute a significant source of potential differences between different RASS cluster surveys. \\subsection{Estimates of the total flux} The X-ray fluxes determined from the observations may still be biased low compared to the total flux coming from the cluster, since part of the flux in the faint outer regions is lost in the background. We can use the fact that we obtained the flux within a well defined angular aperture to make an estimate of the flux that may lie outside this aperture. Note that we use this approach here as a tentative estimate of the flux lost and we will therefore make no further effort in this paper to use the results for a correction, since the underlying assumption that all clusters have the same self-similar shape is not realized precisely enough to make a case by case correction useful without further tests and justifications. The aperture radius that corresponds to the generally used plateau value of the count rate is $R_{out}$. To obtain a rough estimate of the flux possibly missed outside the aperture we adopt the following generic cluster model characterized by a $\\beta$-model surface brightness distribution as given in eq.(3) with $\\beta = 2/3$ to extrapolate the surface brightness profile outside $R_{out}$. For the core radius we are not using the results of the $\\chi^2$ fit, since they have too large uncertainties. We rather prefer to make a rough estimate of the cluster size from its X-ray luminosity. From the studies by Reiprich \\& B\\\"ohringer (1999) we find that the cluster mass is well correlated with the X-ray luminosity according to the relation \\begin{equation} L_x \\propto M_{grav}^{1.2} ~~~~~. \\end{equation} We further assume that the self-similar relation of the core radius and mass of the form $r_c \\propto M_{grav}^{1/3} \\propto L_x^{1/3.6}$ holds (see e.g. Kaiser 1986). Taking a Coma-type cluster with $L_x \\sim 7\\times 10^{44}$ erg s$^{-1}$ (0.1 - 2.4 keV) and a core radius of 300 kpc to normalize the relation we find \\begin{equation} r_c = 0.3~ {\\rm Mpc} \\left( {L_x \\over 7\\times 10^{44}{\\rm erg s}^{-1}} \\right)^{1/3.6}~~~ . \\end{equation} Rather than integrating the X-ray flux of the $\\beta$-model to infinite radius, we stop the integration at 12 core radii which is about as large as the virial radius of a Coma-type model cluster. The difference between an integration to infinity (as for example performed in Ebeling et al. 1998 and De Grandi et al. 1999) to the cut-off radius at 12 $r_c$ is about 8\\%. This overestimate of the flux for integration to infinity is in general smaller than the individual uncertainties but not negligible if one is concerned with the global bias of the sample. Applying this model to our cluster sample we can calculate the fraction of the X-ray flux missed for each of the sample sources. Fig. 9 shows these missing fractions as a function of the source luminosity and of the detected number of source photons. The mean missing flux is about 8.3\\%. (The most discrepant point with a missing flux of about 50\\% and $L_x \\sim 10^{44}$ erg s$^{-1}$ in Fig. 9a is for example a distant cluster observed at low flux ($0.5 \\cdot 10^{-12}$ erg s$^{-1}$ cm$^{-2}$) at low exposure which would be excluded in a proper flux limited sample). \\begin{figure} \\plottwo{fig9a.ps}{fig9b.ps} \\caption{Estimated missing flux of the NORAS cluster sources as a function of a) the source luminosity (in units of $10^{44}$ erg s$^{-1}$) and b) the number of detected photons per source.} \\end{figure} The missing flux fraction, $fr$, features only a very weak dependence on X-ray luminosity. The linear regression fit to the function $fr = f(\\log(L_x)$ shown in Fig. 9a decreases from 8.3\\% for $L_x = 10^{43}$ erg s$^{-1}$ to 7.9\\% for $L_x = 10^{45}$ erg s$^{-1}$. A more significant dependence is found for the number of source photons, as could be expected since this is the main parameter determining the significance of the source detection and how far out the count rate integration can be performed. Here the linear regression fit of Fig. 9b, $fr = f(\\log(N_{ph})$ shows a decrease of $fr$ from 9.5\\% for 30 source photons to 6.7\\% for 500 source photons. Also this dependence is weak. An exception to the relatively small values for the missing fraction constitute some of the low luminosity sources as displayed in Fig. 9a. These are elliptical galaxies or very small groups dominated by elliptical galaxies, which have much smaller core radii - as indicated by the GCA King-model fits - compared to the values assumed after eq.(6). Thus for these small objects the model assumption seems to break down and the actual values for the missing flux is much smaller than estimated here. The validity of this generic model for the extrapolation of the total flux will be pursued in more detail in a future publication. A comparison with pointed observations performed below gives already an encouraging confirmation of these estimates. \\begin{figure} \\plottwo{fig10a.ps}{fig10b.ps} \\caption{Estimated missing flux of the NORAS cluster sources as a function luminosity (left) and source photon number (right) if a constant core radius of 250 kpc is assumed. The straight lines show the results of linear regression fits to the data.} \\end{figure} Alternatively we explore a second approach to estimate the missing X-ray flux by fixing the core radius in the $\\beta $ model to 250 kpc. This approach was used in earlier studies of X-ray cluster samples (e.g. Henry et al. 1992). Figs. 10a and 10b show the results corresponding to the results of Fig. 9. One clearly notes a very steep increase in the missing flux for decreasing X-ray luminosity in Fig. 10a. The dependence on the photon number for which we would expect the strongest dependence is much less pronounced and not essentially different from the results in Fig. 9. One notes, however, that the scatter in Fig. 10b has approximately doubled. The obvious interpretation of these results is that the strong dependence on X-ray luminosity seen in Fig. 10a is artificial and results from an overestimate of the core radius for the less luminous objects. This inappropriate choice of the core radius also increases the scatter in Fig. 10b. Thus we conclude that this approach is clearly inappropriate for our study and the above used scaling of the core radius is a reasonable choice. \\subsection{Comparison with the results of RASS I} \\begin{figure} \\plottwo{fig11a.ps}{fig11b.ps} \\caption{Comparison of the count rates determined by the maximum likelihood method applied in RASS I and the results of the GCA. The straight line in the left pannel indicates equality. The sources found to be extended in the new analysis are shown as filled circles and the point-like sources as crosses. In the right pannel the extended sources are shown as small dots. The dotted line in the right pannel indicates the approximate count rate limit for the RASS I source results (the count rate limit in the plot has a small uncertainty because the count rate conversion from broad band to hard band also includes a dependence on the interstellar column density).} \\end{figure} A comparison of the count rates measured in RASS I for the 378 cluster sources listed in Table 1 with the results of the GCA reanalysis is shown in Figs. 11a and 11b. For the comparison the RASS I count rates measured in the broad band have been converted to hard band counts by means of the measured hardness ratio. Sources which feature a significant extent according to our new analysis are marked in the plot. There is a large fraction of sources for which the count rates measured in RASS I or RASS II are underestimated by up to an oder of magnitude, which are essentially the sources marked as extended by the GCA method. The pointlike sources scatter around the line of equal count rate with an increasing scatter with decreasing count rate. The increase of the RASS I to GCA count rate ratio for low count rates seen in Fig. 11b is most probably an artefact produced by the previously set count rate limit in the selection of the RASS I sources for this sample (as indicated in the figure by the dotted line). The source of disagreement for the extended sources results from the design of the source analysis technique used for the RASS which is tuned to work optimally for point sources. Two effects discriminate against the proper accounting of the count rate of extended sources: i) the Gaussian kernel of the source shape fitting of the maximum likelihood analysis is bound to miss the outer wings of a typical cluster surface brightness distribution and ii) part of the outer X-ray halos of extended clusters may be treated as background by the background spline fitting process as used in the RASS standard analysis. A source analysis technique tuned to process the extended sources properly is therefore required to avoid these problems and to obtain correct X-ray parameters for the objects in our sample. The reanalysis of all the X-ray sources in the sample was therefore a necessary prerequisite for the compilation of an X-ray cluster catalogue to be used for astronomical and cosmological studies. \\subsection{Comparison to pointed observations} To test the count rate determination of our new analysis technique against a more reliable standard we have analyzed 80 clusters of our sample in pointed observations in which the better photon statistics allows a more detailed and precise analysis. The results for the brighter sources were taken from the compilation of Reiprich \\& B\\\"ohringer (1999) of the ROSAT clusters with the highest flux. 13 of the clusters from this sample were also analyzed in very large RASS survey fields ($4\\deg \\times 4\\deg $ or $8\\deg \\times 8\\deg $) because of the large cluster sizes. In the analysis by Reiprich \\& B\\\"ohringer (1999) a more refined source analysis was performed (where contaminating sources are excised in a wide region in and around the cluster and a possibly badly measured background is iteratively corrected by parabolic fits to the azimuthally integrated background surface brightness profile outside the cluster). Therefore it is also interesting to keep these objects in the list for comparison. The other data were retrieved from the ROSAT archive. The analysis technique used is similar to the one described above with a main difference that contaminating sources are excised interactively. We are using the count rate at $R_x$. The values for $R_x$ found in the pointed observations are generally larger than $R_{out}$ found in the RASS data as the flux can usually be traced further out into the background in the deeper observations. Fig. 12 shows a comparison of the count rates found in the two data sets. The objects analyzed in large RASS fields are marked with open symbols. The mean deviation of the count rates determined for the pointed data and the present results is 8.6\\%. Thus we conclude that the missing flux is on average about $7 - 10\\%$ without a significant bias as a function of X-ray flux. This result is in excellent agreement with the estimates for the missing flux in section 3.5. The validity of the GCA approach is thus confirmed in two ways: the missing flux fraction is relatively small compared for example to the measurement errors and the {\\it ab initio} estimates for the missing fraction are approximately correct. \\begin{figure} \\plotone{fig12.ps} \\caption{Comparison of the count rates for 67 clusters as determined in the present analysis and in deeper pointed observations (full circles). In addition we also show the comparison between the present results and the interactive analysis by Reiprich an B\\\"ohringer (1999) for 13 clusters in the same RASS fields (open symbols). The solid line indicates equality and the two dashed lines indicate deviations of 20\\%.} \\end{figure} ", "conclusions": "To test the completeness of the catalogue, a comparison can be made to the southern RASS cluster survey project, the REFLEX Survey (B\\\"ohringer et al. 1998). This sample has been constructed in a different way making extensive use of the COSMOS data base to correlate X-ray sources with the galaxy distribution. Therefore this sample does not rely on existing cluster catalogues nor on the selection of X-ray sources featuring an extent. Internal statistical estimates for the REFLEX sample suggest a completeness larger than 90\\% for the flux limit of $F_X = 3 \\times 10^{-12}$ erg s$^{-1}$ cm$^{-2}$. The number counts of the REFLEX survey and the NORAS survey are compared in Fig. 19. We note that the NORAS sample reaches a fraction of 50\\% of the REFLEX number counts at the REFLEX flux limit. A large fraction of the missing clusters has been recovered by the supplementary sample in the study area. In Fig. 20 various subsamples of the combined cluster catalog in the 9$^h$ to 14$^h$ region are compared to the REFLEX Survey. While the cluster sample from Table 1 recovers with 40\\% an even smaller fraction of the sky density compared to the REFLEX Survey, 70\\% are reached if the ACO clusters are added and 82\\% are obtained for the combined sample. Thus there is still a fraction of clusters missing. This is easily understood, since the REFLEX sample contains a fraction of 22\\% of all objects which are not resolved as extended by the GCA analysis. We can expect that about 10 - 20\\% of the clusters for the REFLEX flux limit are neither listed as ACO clusters nor recognized as extended sources. Note that the combined sample in the study region completely recovers the previous RASS cluster sample by Ebeling et al. (1998) as discussed below. Most clusters not contained in Table 1 are easily found as ACO clusters and 7 further clusters show a clear extent. Therefore to achieve a higher completeness in our continuing northern cluster survey we are, in addition to including known catalogued clusters with X-ray emission and newly classified extended X-ray sources, conducting further imaging of promising non-identified X-ray sources to recover the compact X-ray clusters missed in the previous searches. \\begin{figure} \\plotone{fig19.ps} \\caption{ LogN-logS for the NORAS I sample (continous line) compared to the REFLEX Cluster Survey (dotted line).} \\end{figure} \\begin{figure} \\plotone{fig20.ps} \\caption{ LogN-logS for the NORAS 9-14 hour region. NORAS I main sample (lower dashed curve), NORAS I and ACO clusters (dashed-dotted curve), NORAS I, ACO, and supplementary extended cluster candidates (continuous curve), NORAS I sample for the whole northern sky (lower dotted curve), REFLEX Survey (upper dotted curve).} \\end{figure} Using the early results of the RASS we compiled an X-ray sample of galaxy clusters by selecting the RASS sources which featured a significant extent in the first standard processing of the RASS. The complete spectroscopic identification of these sources (as far as no identification was already available) leads to a compilation of a catalogue of 378 X-ray cluster sources. A reanalysis of the X-ray properties of these sources shows that the X-ray source properties can successfully be used in the source identification process. In particular we found that in many cases the hardness ratio can be used to flag sources which are severely contaminated by an AGN. In this way we could also exclude previously identified X-ray clusters -- as for example the two cases discussed in the previous section -- from identification as cluster sources. One of the major points of concern of such catalogs of X-ray clusters is the contamination of the observed X-ray luminosity by AGN. As commented in Section 4.1 there are all combinations of clusters and optically identified AGN or radio galaxies: AGN which do not provide a significant contribution to the cluster X-ray flux, AGN that contribute partially, and AGN which completely outshine the clusters. These cases can only definitely be distinguished if high resolution X-ray images or X-ray spectra of high quality are available. E.g. ROSAT HRI observations provide the means for this distinction, but they are only available for a small fraction of the sources for survey sizes of the present survey. Lack of additional X-ray data creates a twofold danger. Since X-ray sources are often identified with the most plausible nearby optical counterpart, there is a significant risk that the cluster ID will be discarded when an AGN is found near or within the error circle. Conversely, an AGN which is the primary X-ray emitter may not be recognized and the emission falsely associated to the cluster. Also in the present case this problem of AGN cluster associations cannot be solved in each case and we still expect some hidden misclassifications which will be hard to find until the whole sample is probed more deeply in X-rays. However, we believe that the remaining uncertainties are small and not harmful for the application of our sample for cosmological statistical studies and we will substantiate this believe in the following. First of all we note that $74.4 \\%$ of the sources in the catalog in Table 1 are sources recognized as significantly extended. This requires a major contribution to the X-ray emission from diffuse cluster sources. Thus for this source population we have a very high reliability that these sources are true X-ray clusters. We can now - as shown in Fig. 22 - compare the spectral properties of these extended cluster sources to those cluster sources where no significant extent could be established in terms of the parameter $P_{ext}$. We note that the two distributions are hardly different. Also both distributions are almost symmetric around unity, as would be ideally expected. For some reason the larger sample of extended sources shows a broader distribution of the $\\Delta HR$ parameter. This seems to be caused by clusters with a larger number of photons for which a small systematic error in the $HR$ estimate leads to a significant additional scatter in $\\Delta HR$ in units of $\\sigma$ and also the low temperature groups will add to this scatter. A comparison to the non-cluster sources shows a very large difference in the spectral parameter distribution, however. While the non-cluster sources have a median deviation of the parameter $\\Delta HR$ of $-4.6\\sigma$ the X-ray pointlike sources identified as clusters have a median $\\Delta HR \\sim -0.02\\sigma$. We can explore the contamination effect in somewhat more detail by considering the effect on a typical cluster source. The median number of source photons for the clusters listed in Table 1 is about 94 source photons. For such a source the typical uncertainty in the measured hardness ratio is about $\\delta HR \\sim 0.12$. An AGN with median properties easily shows a deviation of $3 - 4 \\sigma$ in the $\\Delta HR$ parameter. Even if the contamination fraction by AGN is only 50\\% or 25\\% the typical shift in the parameter of $\\Delta HR \\sim 1.7\\sigma$ and $\\sim 0.9\\sigma$ still constitutes a recognizable distortion of the $\\Delta HR$-distribution. Another test is shown in Fig. 22b. Here we compare the distribution of the $\\Delta HR$ values for the clusters with the sample of point like clusters artificially contaminated by removing 20\\% of the cluster sources and replacing them by a statistical sample of non-cluster sources. There is a clearly visible change in the distribution of the spectral parameters. A KS test shows that the two distributions are still only distinguishable at the $\\sim 90\\%$ level. A test with a similar contamination of 40\\%, however, leads to a clear difference with a KS probability for the two samples being statistically the same of $\\sim 10^{-4}$. Since a 20\\% contamination of the point like clusters corresponds to a contamination of about 5\\% of the total sample we expect that the real misclassification fraction is not larger than this percentage. \\begin{figure} \\plottwo{fig22a.ps}{fig22b.ps} \\caption{Distribution of the spectral parameter, $\\Delta HR$ in units of sigma, for different subsamples of the present study. The parameter $\\Delta HR$ is the difference of the measured from the estimated hardness ratio of the source for given interstellar absorption scaled by the 1$sigma$-uncertainty of the hardness ratio determination. The right pannel shows histograms of the parameter $\\Delta HR$ for pointlike cluster sources (thick line), extended cluster X-ray sources (thin line), and non-cluster sources (broken line). The right pannel shows normalized, cumulative histograms for (a) the pointlike cluster sources, (b) the extended cluster sources, (c) the sample of pointlike cluster sources artificially contaminated by 20\\% of non-cluster sources, and (d) the non-cluster sources. The samples (b) and (c) have an about 10\\% KS probability of being sampled from the same parent distribution as (a).} \\end{figure} The present X-ray cluster catalog provides a wealth of new data. 98 new cluster sources are listed in Table 1 and in addition new X-ray luminous groups associated with known giant ellipticals are reported in Table 1 and 8. Test searches for additional X-ray clusters in the 9$^h$ - 14$^h$ region have shown that both, the selection of extended sources by the RASS standard analysis and the finding of X-ray clusters by selecting extended sources is quite incomplete in terms of the compilation of flux-limited X-ray cluster catalogues - except for very high flux limits. We have shown that probably most of the missing clusters can be recovered by using a better search algorithm for extended sources and by screening the optically known clusters. This may not be sufficient to produce a high quality, highly complete catalogue. Further screening of RASS sources, including CCD imaging of sources not visibly extended in X-rays, is required. The completion of the NORAS sample thus requires additional imaging and spectroscopic observations." }, "0003/astro-ph0003396_arXiv.txt": { "abstract": "The response of a pair of differently polarized antennas is determined by their polarization states {\\it and} a phase between them which has a geometric part which becomes discontinuous at singular points in the parameter space. Some consequences are described.\\\\ ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003169_arXiv.txt": { "abstract": "We numerically test the mode-coupling model (Lubow 1991a) of tidal instability in SU~UMa systems. So far, all numercial models confirming it have been based on SPH codes and isothermal equation of state. In our paper we present Eulerians models, using both isothermal approximation and the full energy equation. We also investigate influence of different ways of mass transfer. While isothermal models behave similarly to SPH simulations, the behaviur of models with full energy equation is quite different, and the mode-coupling model is not confirmed in this case. ", "introduction": "SU UMa stars form a subclass of dwarf novae, which, in turn, are a subclass of cataclysmic variables (CVs). Like all CVs, the SU UMa stars are semidetached binary systems consisting of a white dwarf (the primary) and a low-mass, main-sequence star (the secondary). The secondary fills its Roche lobe, and a stream of gas flows from its surface toward the primary through the inner Lagrangian point. Because of excessive angular momentum, the stream is deflected from its original direction, and an accretion disk is formed around the primary. The disk may be subject to a thermal instability (Smak 1999), resulting in episodes of enhanced accretion rate. To a distant obsever, such an episode is visible as a temporary brightening of the star, commonly referred to as an outburst. As opposed to ordinary dwarf novae, the SU UMa stars exhibit a clearly bimodal distribution of outbursts. Normal outbursts have an amplitude of $\\sim$ 3 mag, and last from one to four days. Superoutbursts are by $\\sim$ 1 mag stronger, and last for up to several weeks. The recurrence time of normal outbursts (days to weeks) is not constant, and it varies substantially form one system to another. The superoutbursts repeat more regularly, and their recurrence time is much longer (months to years). In the extreme case of WZ Sge stars, superoutbursts are nearly exclusively observed, with a recurrence time of up to several tens of years. In superoutbursts, the light curve of a SU UMa system is modulated with a period a few per cent longer than the orbital period. Those modulations are referred to as superhumps. The superhump signal is known to originate from extended source(s) in the outer disk (Warner 1995). Superoutbursts are thought to be driven by a combination of thermal and tidal instability. During normal outbursts, the disk grows in size as it diffuses under the influence of increased viscosity. In systems with mass ratios $q\\lesssim0.25$, it eventually reaches up to the location of the 3:1 resonance, at which the orbital frequency of the disk gas is three times larger than the orbital frequency of the binary (we define $q$ as the ratio $M_2/M_1$, where $M_1$ and $M_2$ are primary's and secondary's mass, respectively). Subsequently, the tidal instability sets in, the disk becomes eccentric, and, seen in the inertial frame, it performs a slow, prograde precession. The tidal influence of the secondary on (and the viscous dissipation in) the outer disk, is largest when the bulk of the disk passes the secondary. The superhump period is then the beat period between the precession period and the orbital period of the binary (Osaki 1996). Disk precession and superhump phenomenon have been subject to rather intense theoretical investigations, largely based on numerical simulations. According to Lubow (1991a,b), the eccentricity builds up due to nonlinear interaction of waves, in which the m=3 component of the tidal field is a key factor. Heemskerk (1994) performed simulations using only that component, and he found that the disk became eccentric, but it precessed retrogradely. Moreover, with the full tidal potential, the accretion disk was kept away from the location of the resonance, and no significant eccentricity was produced. Heemskerk's results are the only ones obtained with an Eulerian (fluid) code. All remaining models presented in the literature have been based on Lagrangian (particle) codes. A detailed review of those calculations can be found in Murray (1998). While a qualitative agreement with observations of superhump systems was reached in several aspects, the models did not entirely agree with the analytical theory of the tidal instability. In particular, the measured eccentricity growth rates were much smaller than the predicted ones. The models themselves were often based on an extremely simple physical scenario (fully isothermal disk; gas from the secondary uniformly \"raining\" onto a circular orbit within the disk). In some of them the tidal instability was initiated by an arbitrary increase of viscosity in the disk by a factor of 10. Finally, even in the models which properly followed the stream of gas between the inner Lagrangian point up to its collision with the edge of the disk, the resolution in the collision region was too poor to resolve strong shock waves responsible for the hot spot phenomenon. In the present paper, we obtain Eulerian models of disks in SU Uma systems in order to isolate the influence of various approximations on the outcome of the simulations. The models can be directly compared to Lagrangian models of Murray (1996, 1998). We perform both isothermal simulations, and simulations in which the full energy equation with a realistic cooling term is solved. We also compare \"rainfall-type\" mass transfer models with those based on realistic modelling of the stream from the secondary. The physical assumptions on which our models are based are described in Sect.2 together with numerical methods emplyed to solve the equations of hydrodynamics. The models are presented in Sect. 3, and the results are discussed in Sect. 5. ", "conclusions": "In the present paper we reported a series of 4 simulations of disk in Cataclysmic Variables of SU UMa type. The least realistic model was obtained with an isothermal equation of state, and a \"rainfall-type\" approximation for mass transfer from the secondary. The most realistic simulation involved a full energy equation, and a stream of matter originating at the inner Lagrangian point. We found that only the isothermal disks exhibit a clear tendency toward elliptic distortions accompanied by precession. Within the framework of Lubow's (1992) theory, one could try to explain this result by assuming, that the strength of the tidally excited (2,2) mode is significantly larger in radiative models than in the isothermal ones. This is because, according to the theory, the (2,2) oscillations tend to keep the disk gas away from the 3:1 resonance responsible for the growth of the tidal instability. Such an assumption, however, is not confirmed by the analysis of our results: the (2,2) mode appears to be equally strong in all models. The phase of an exponential growth of the (1,0) mode, foreseen by the theory, was found only in the least realistic model. However, contrary to theoretical predictions, even in that case the time derivative of the strength of (1,0) mode was not proportional to the strength of the (2,3) mode. Oscillations with a period slightly longer than $P$, which may be tentatively associated with superhumps, were observed in isothermal models only. The period of oscillations found in radiative models was about 3 times shorter. The main results of this work may be summarized in three points: 1. As foressen by the tidal instability theory, isothermal develop an appreciable eccentricity, and begin to precess. The precession period is the same as the period of rapid fluctuations in viscous dissipation rate, and it is slightly longer than the orbital period of the binary. The behaviour of the (1,0) mode, however, is not consistent with the theory. 2. In radiative models an elliptic, precessing disk does not develop. The dominant oscillations in the viscous dissipation curve have periods of $\\sim 1/3P$. 3. In all models the two-armed (2,2) mode, excited and maintained by tidal forces of the secondary, is very clearly seen. Our conclusion is that the mechanism of superhump phenomenon is not yet entirely understood, and further research on this cubject is desirable. R\\'o\\.zyczka \\& Spruit (1993) simulated a viscosity-free disk with radiative losses, in which angular momentum was transported by spiral shocks. They found an eruptive instability, qualitatively similar to outbursts of Dwarf Novae. During the eruption, the radius of the disk increased substantially. We suggest that low-viscosity disks prone to this type of instability may develop eccentricity and exhibit superhump-like oscillations during outbursts. This issue is presently under investigation. {\\bf Acknowledgments} This research was supported by the Committee for Scientific Research through the grant 2.P03D.004.13." }, "0003/astro-ph0003443_arXiv.txt": { "abstract": "We present new high resolution ($R \\simeq 21,000$) near-infrared ($\\lambda = 2 \\mu$m) spectroscopic observations of a sample of Class I and flat-spectrum protostellar objects in the $\\rho$ Ophiuchi dark cloud. None of the five Class I spectra show CO v = 0 -- 2 absorption features, consistent with high $K$-band continuum veilings, $4 \\lesssim r_{k} \\lesssim 20$ and fast stellar rotation, assuming that the underlying protostellar photospheres are of late spectral type, as is suggested by the low luminosities of most of these objects. Two of the flat-spectrum protostellar objects also show no absorption features and are likely to be highly veiled. The remaining two flat-spectrum sources show weak, broad absorptions which are consistent with an origin in quickly rotating ($v$ sin $i \\approx 50$ km s$^{-1}$) late-type stellar photospheres which are also strongly veiled, $r_{k} \\simeq 3 - 4$. These observations provide further evidence that: 1)-Class I sources are highly veiled at near-infrared wavelengths, confirming previous findings of lower resolution spectroscopic studies; and 2)- flat-spectrum protostars rotate more rapidly than classical T Tauri stars (Class II sources) , supporting findings from a recent high resolution spectroscopic study of other flat-spectrum sources in this cloud. In addition our observations are consistent with the high rotation rates derived for two of the Class I protostellar objects in our sample from observations of variable hard X-ray emission obtained with the ASCA satellite. These observations suggest that certain Class I sources can rotate even more rapidly than flat-spectrum protostars, near breakup velocity. ", "introduction": "The physical natures, evolutionary states, and circumstellar disks of classical T Tauri stars are becoming better understood due to recent spectroscopic observations, high resolution imaging, and advances in the theory of pre-main-sequence (PMS) evolution. However, the natures of the central stars and inner circumstellar environments of protostars are still not very well known. This is primarily because they are so heavily extinguished that they are difficult to observe even with modern instruments and detectors. For example, it is not known whether the central stars in protostellar objects differ substantially in effective temperature, radius, or rotation properties from classical T Tauri stars (CTTSs). The presence of protostellar envelopes, significantly higher accretion rates, and more powerful outflows suggest that the photospheres of protostellar cores may indeed be physically different from CTTSs. Moreover, it is not known whether the physical natures of protostellar photospheres are consistent with the predictions of protostellar evolution and PMS stellar theory. Flat-spectrum and Class I young stellar objects (YSOs) are the best low-mass protostellar candidates for spectroscopic study because they have relatively well-developed central stars and are detectable at near-IR wavelengths. Several pioneering studies have recently been undertaken to begin investigating the physical natures of these objects. A few flat-spectrum protostars have been observed in low resolution spectroscopic surveys and thus far they appear to be characterized by late-type photospheres with high continuum veilings and sub-giant surface gravities \\citep[hereafter Paper I]{LR99,KBTB98,CM92,PaperI}. Even fewer have been observed at high spectroscopic resolution, and these observations suggest that the flat-spectrum protostars rotate significantly faster than the more evolved CTTSs \\citep[hereafter Paper II]{PaperII}. These investigations have provided some interesting clues to the natures of these objects, but more observations are clearly needed to make sense of these late-phase protostars. It would be most interesting to determine if the less evolved, more heavily embedded and veiled Class I protostars also show near-IR photospheric absorption lines when observed at high resolution with high signal-to-noise. Such observations could directly constrain the effective temperatures, gravities, veilings, and rotations of these objects, providing further evidence as to whether they are dominated by near-IR stellar, disk, or envelope emission and whether they are physically similar to flat-spectrum YSOs. More flat spectrum YSOs should also be observed with high-resolution near-IR spectroscopy to confirm that they are indeed late-type rapid rotators. In a sense the flat-spectrum YSOs provide a link to the well-known PMS stars, and this must be better developed so that they can in turn serve as a link to the less well-known Class I objects. Therefore we have undertaken a new high-resolution, near-IR spectroscopic study of flat-spectrum and Class I YSOs in the $\\rho$ Ophiuchi cloud core. We describe these new observations in \\S 2 and present a rotation and veiling analysis of these data in \\S 3. In \\S 4 we discuss the likely natures of each of these objects and suggest further observational work. ", "conclusions": "We now discuss how the results of this veiling / rotation analysis and pre-existing data constrain the possible physical natures of these sources. \\subsection{Flat-Spectrum Objects} The flat-spectrum YSOs IRS 63 and IRS 51 were both found to have broad, weak CO absorptions which matched those expected for late-type stellar photospheres rotating at $v$ sin $i \\simeq 50$ km s$^{-1}$. The weak CO absorptions of these two YSOs are consistent with their not being detected in our initial low-resolution survey (Paper I). We estimate the continuum veiling of IRS 63 to be $r_{k} \\simeq 4$ provided that it is a PMS YSO near M0 spectral type. \\citeauthor{LR99} find that the spectral type of IRS 51 is G5 -- K7, earlier than our M0 template VSSG 17. Thus their derived veiling $r_{k}$ = 1 -- 3 is lower than ours because a G5 -- K7 PMS star has less intrinsic CO absorption than a M0 one (see \\S 3.2). It is likely that IRS 51 is indeed an embedded low-mass YSO because its bolometric luminosity is only 1.4 L$_{\\odot}$ (WLY). The birthline mass for this luminosity is approximately 0.5 M$_{\\odot}$, corresponding to a spectral type of K5--7 and a true veiling of $r_{k} \\simeq 3$. Thus it is likely that both IRS 63 and IRS 51 are similar to the quickly rotating flat-spectrum YSOs which we analyzed in Paper II, but these new objects have even greater veiling (i.e. $r_{k}$ = 3 -- 4 versus $r_{k} \\simeq 1$ for the Paper II YSOs). \\citeauthor{LR99} found GSS 26 to have variable veiling, $r_{k}$ = 0.75 and $r_{k}$ = 4 at epochs of 1994 July and 1996 May, respectively. Our spectrum of GSS 26 (in Figure 1) was taken in 1997 May, one year after the latest \\citeauthor{LR99} spectrum. We estimate that $r_{k} \\gtrsim 11$ when our spectrum was acquired, and our assumption of a M0 spectral type is consistent with the \\citeauthor{LR99} determination of K5 -- M2. This rapid increase in veiling - a factor of 2 each year - is perhaps suggestive of a similarly rapid increase in accretion. \\citeauthor{LR99} also note that this source increased in brightness by $K \\simeq$ 1.2 mag between epochs. Our spectra are not photometrically calibrated, but comparisons with other objects support that this source was at least as bright as when observed by \\citeauthor{LR99} the previous year. This is one of the YSOs \\citeauthor{LR99} observed whose HI Br $\\gamma$ emission line flux increased as its veiling increased, implying that the excess continuum emission is associated with a circumstellar accretion disk if the Br $\\gamma$ emission arises from disk accretion. The rapid variability of this object's veiling also suggests that its $K$-band veiling is produced by accretion from inner disk distances (several AU) and not from an outer disk or an outer circumstellar envelope. The high $r_{k}$ values of these objects also constrain the physical origins of their veilings. In Paper I we showed that veilings $r_{k} > 1$ cannot be produced by a simple optically thick, geometrically thin reprocessing disk around a low-mass PMS star. Consequently we argued that these high veilings are most likely produced by either actively accreting circumstellar disks or circumstellar envelopes associated with these objects. Furthermore we found that veilings in the range measured for IRS 51 and IRS 63, $r_{k}$ = 3 -- 4, can be caused by luminous accretion disks ($L_{disk}/L_{*} \\leq 3$). Veilings in the range observed for GSS 26, $r_{k} \\simeq 5 - 10$, could be explained by extremely luminous accretion disks ($L_{disk}/L_{*} \\geq 3$). In either case, these accretion disks would have to have relatively large central holes to avoid producing strong CO absorption-line systems in the disk photosphere itself. However, because there is no obvious physical mechanism for producing central holes of the needed size, \\citet{CHS97} suggested that the veiling flux must originate in some other circumstellar structure such as the inner regions of the protostellar envelope. On the other hand, this seems to be inconsistent with the observation by \\citeauthor{LR99} that the $K$-band excesses of flat-spectrum YSOs are correlated with their HI Br $\\gamma$ line fluxes which in turn suggests that the veiling flux should originate in the disk. More detailed knowledge of the conditions required to produce CO absorption line systems in an accretion disk may be needed to resolve this issue. The flat-spectrum source YLW 13B was found to have H Br $\\gamma$ absorption by \\citeauthor{LR99}, who estimate its spectral type to be earlier than K0. However, they also find it to be significantly veiled with $r_{k} > 1$, so it is possibly an intermediate mass PMS cloud population member. Our non-detection of CO absorption does not constrain this source further. \\subsection{Class I Objects} We do not detect CO absorptions in any of the Class I YSOs which we observed (Figure 2), confirming earlier low resolution spectroscopic observations that found all these objects to be featureless and highly veiled (Paper I; \\citeauthor{LR99}). Consequently their spectral types are unknown, however all of these Class I sources exhibit HI Br $\\gamma$ emission (Paper I; \\citeauthor{LR99}). Analysis of our new data constrains the natures of these objects. Table 2 shows that they all have large veilings, $r_{k} > 4$, if they are late-type ($\\sim$ M0) stars. GSS 30, IRS 43, and WL 6, all have estimated minimum veilings of $r_{k} \\simeq 5-8$, overlapping with the flat-spectrum sample. These sources all have bolometric luminosities L$_{\\rm bol} \\leq 13$ L$_{\\odot}$ (WLY). This is consistent with their being low-mass (M $<$ 1 M$_{\\odot}$) protostars accreting matter at rates $\\dot{M} \\simeq 5 \\times 10^{-6}$ M$_{\\sun}$ yr$^{-1}$, the value expected for the T $\\simeq$ 20 K gas temperatures in the $\\rho$ Oph cloud \\citep[see][hereafter ALS]{ALS87}. Veilings in the observed range, $r_{k} > 4 - 10$, are also predicted by theoretical models of Class I circumstellar envelopes (Paper I; Calvet et al.). \\citeauthor{MGTK00} note that the maximum possible mass of IRS 43 (also known as YLW 15) is 2.2 M$_{\\odot}$ which is derived from PMS models given its bolometric luminosity (L $\\sim$ 10 L$_{\\odot}$) and assuming it is on the birthline. Likewise, they calculate that the maximum likely mass of WL6 is approximately 0.4 M$_{\\odot}$; this increases somewhat if the L$_{\\rm bol}$ = 2.4 L$_{\\odot}$ of WLY is adopted. These maximum calculated masses assume that essentially all luminosity is due to photospheric thermal radiation ($L = 4\\pi R^{2}\\sigma T^{4}$) and essentially none is due to accretion ($L = G M\\dot{M}/R$). These sources are discussed further in $\\S 4.3$. We estimate that Elias 29 and IRS 54 have very high veilings if they are late-type low-mass stars, $r_{k} > 14 - 34$. The bolometric luminosity of IRS 54 is estimated to be only 12 L$_{\\odot}$ (WLY), also consistent with this object being a low-mass protostar which is accreting its envelope at the rate prescribed by the $\\rho$ Oph cloud's gas temperature. However, its continuum veiling must be $r_{k} > 14 - 20$ if it has spectral type M0 and is rotating rapidly. This is about a factor of 2 higher than the model predictions of Calvet et al., but those calculations were done for a hypothetical $\\rho$ Oph Class I YSO with L = 5 L$_{\\odot}$. The model may predict greater veiling for IRS 54 if its higher luminosity is taken into consideration. It is also possible that this source may be a somewhat earlier type protostar which is less veiled. Elias 29 has the highest derived veiling of the sample, $r_{k} > 25 - 34$, assuming an intrinsic M0 photosphere and a high rotation rate. WLY estimate its luminosity to be L$_{\\rm bol}$ = 48 L$_{\\odot}$, and ALS have modeled it as a 1 M$_{\\odot}$ protostar which is accreting its circumstellar envelope. Such a star would have a spectral type of K3--4 if on the birthline \\citep[see][]{DM97, S88}, with an intrinsic CO absorption approximately 60\\% as strong as that of an M0 star (see \\S 3.2). Thus we revise our estimate of the likely veiling of this YSO to $r_{k} > 15 - 20$ if it is indeed a 1 M$_{\\odot}$ protostar. In Paper I we analyzed the ALS model for Elias 29 and showed that the predicted emission from the inner protostellar envelope of this source would produce a veiling of $r_{k} \\approx 20$, assuming that its disk luminosity is 0.75 L$_{\\rm bol}$. Thus our new measurement is consistent with our earlier prediction based on the ALS model. The high luminosity of Elias 29 also allows for it being a more massive, earlier spectral type YSO that has higher stellar luminosity and less accretion luminosity than assumed by the ALS model. However, it is unlikely to be very different because the observed near-to-far IR energy distribution and the 10 $\\micron$ silicate absorption of Elias 29 are fit well by the ALS model, and there are no clues which indicate that Elias 29 is an early-type object. For example, WL 16, which is likely an early A type star \\citep{BHCRL00}, has mid-IR aromatic hydrocarbon emission features which indicate a UV radiation field \\citep{HTG92}. However, Elias 29 shows no evidence for IR hydrocarbon emission \\citep{HBT95,B99} and thus no evidence for a UV radiation field. \\subsection{Protostellar Rotation} Our observations, specifically the broad band head shapes of IRS 51 and IRS 63, strengthen the earlier findings of Paper II which suggested that flat-spectrum protostars rotate more rapidly than Class II sources (CTTSs). Since flat-spectrum protostars are believed to be evolutionary precursors of CTTSs, then this finding may indicate that certain physical conditions characteristic of protostellar evolution (e.g., high accretion rates) may result in their higher rotation rates. It is therefore interesting to ask whether the less evolved Class I sources might rotate even more rapidly than the flat spectrum sources. Indeed, some models of protostar development predict that such objects should be rotating near breakup \\citep{S91}. Recently, strong hard X-ray flares have been observed with the ASCA satellite from two of the Class I protostars in our sample -- IRS 43 and WL 6 (\\citeauthor{MGTK00}). These remarkable observations reveal relatively rapid periodicities in the X-ray emission from these sources, enabling the derivation of their photometric rotation rates. \\citeauthor{MGTK00} find that WL 6 is rotating with a period of about 3 days ($v$ sin $i \\simeq 40$ km s$^{-1}$ for a 0.5 M$_{\\sun}$ star on the birthline), comparable to the rotation rates of the flat-spectrum sources observed here and in Paper II. This source also has a weak outflow \\citep{STUKTH97}, undetected millimeter emission from its envelope \\citep{AM94}, and an IR energy distribution which can be modeled as a highly extinguished flat spectrum YSO (\\citeauthor{MGTK00}). All of these properties indicate that WL 6 may indeed be very similar to the flat spectrum YSOs for which we have detected absorption lines and have found to be rotating more rapidly than CTTSs (Class II YSOs). \\citeauthor{MGTK00} also argue that the central star of IRS 43 has a 20 h rotation period, the observed period of its X-ray variability. This requires that its mass be greater than or equal to 1.8 M$_{\\odot}$ in order for it to be rotating below breakup velocity if it is on the birthline (\\citeauthor{MGTK00}). Thus the mass of IRS 43 is constrained to lie in the range 1.8 -- 2.2 M$_{\\odot}$ by its X-ray emission and bolometric luminosity (see $\\S 4.2$) if its rotation period is indeed equal to its X-ray variability period of 20 h. IRS 43 does have a more steep mid-IR energy distribution (clearly Class I) than WL 6, and it also has spatially-resolved (r $\\simeq$ 3000 AU) millimeter emission with a derived envelope mass of approximately 0.1 M$_{\\odot}$ (\\citeauthor{AM94}). Therefore it is likely to be in an earlier evolutionary state than WL 6. The slow rotation velocities of CTTSs have been explained by angular momentum regulation of these stars by magnetic coupling to their disks. \\citet{ESHSHHAMPG93} found that late-type T Tauri stars (TTSs) with large $H-K$ IR excesses (CTTSs) had slow rotation periods, P $>$ 4 d. They also found that TTS with small $H-K$ IR excesses had a broad range of periods, including a significant number with P $<$ 4 days. \\citeauthor{ESHSHHAMPG93} interpreted this correlation to arise because the magnetic fields of the CTTSs were were coupled to their disks, providing stellar angular momentum regulation and therefore long stellar rotation periods. The low-excess TTSs had already dissipated their disks and so were not subject to this regulation mechanism. More recent studies of larger TTS samples have both disputed that the correlation between IR excess and rotation period exists \\citep{SMMV99} and have provided evidence that it exists but is weak \\citep{HRHC00}. Our studies of YSO rotation (this paper and Paper II) have shown that flat-spectrum protostars rotate significantly more rapidly than Class II YSOs or CTTSs, suggesting that rotation velocities decrease as stars evolve past the protostellar state. This scenario has been bolstered and expanded further by the recent X-ray results of \\citeauthor{MGTK00}. Taken together (and along with the many rotation studies of optically visible CTTSs), these works suggest that heavily embedded protostars (Class I) rotate very rapidly, in some cases near breakup velocity, while less embedded ones (flat--spectrum YSOs) rotate somewhat less rapidly, at about 1/3 breakup velocity ($v$ sin $i \\simeq 50$ km s$^{-1}$), and Class II YSOs / CTTSs rotate slowly, $v$ sin $i < 20$ km s$^{-1}$. This finding would be strengthened considerably by further cross-checking of observational techniques; the X-ray protostars should be observed at higher signal-to-noise in the near-IR to search for rotationally broadened lines, while the flat spectrum and Class II YSOs with IR-derived rotation velocities should be observed for periodic X-ray variability. If the angular momenta of low-mass YSOs are indeed regulated by star--disk coupling, then the fact that flat-spectrum (and at least one Class I) YSOs rotate significantly more rapidly than CTTS implies that either the flat-spectrum / protostellar YSOs are coupled to faster rotating disk regions than CTTSs, or else that stars and disks do not become rotationally locked until the CTTS evolutionary phase. In the first case, flat-spectrum and Class I YSOs may couple to their disks at smaller radii (and hence have higher rotation velocities) because their accretion rates are much higher than CTTS. The veilings and luminosities of Class I and flat-spectrum YSOs are considerably higher (by about an order of magnitude) than CTTS, supporting the notion that they have higher accretion rates also. In support of the second case, \\citeauthor{MGTK00} have posited that protostars are not initially magnetically coupled to their disks but rather spin-down and become coupled over a magnetic braking time on the order of 10$^{5}$ yr which is nearly linearly proportional to stellar mass. This is comparable to the lifetime of the Class I and flat spectrum phases, so this would account for the higher rotation velocities of flat-spectrum and Class I protostars. This latter magnetic braking scenario of velocity evolution from Class I to flat-spectrum to CTTS YSOs may be somewhat complicated by mass effects; \\citeauthor{MGTK00} predict that at the same age more massive protostars will rotate more quickly than less massive ones. Protostellar masses must be measured much more accurately before this effect can be verified, however. Obtaining new high resolution, high signal-to-noise spectra over the entire 1.5 -- 2.4 $\\micron$ region will likely be the best method for obtaining more definitive information on the masses (spectral types) and rotational characteristics of Class I protostars. This wide spectral range is required in order to be sensitive to a wide range of spectral types. Even intermediate-to-high mass stars with HI line emission may show $H$-band HI Br absorption lines which may strongly constrain spectral types and masses (e.g., \\citeauthor{BHCRL00}), while $K$-band data are required to determine the properties of very red late-type YSOs. Detecting and resolving near-IR lines in WL 6 and IRS 43 would allow determination of their masses and photospheric rotation rates, providing a good test of the emerging scenario of protostellar rotational evolution." }, "0003/astro-ph0003325_arXiv.txt": { "abstract": "We report on follow-up spectroscopy and photo\\-metry of the cataclysmic variable candidate HS\\,0907+1902 selected from the Hamburg Quasar Survey. $B$, $V$, and $R$ photometry obtained during the first observed outburst of HS\\,0907+1902 ($V\\approx13$) reveals deep eclipses ($\\Delta V\\approx2.1-2.8$) and an orbital period $\\Porb=4.2$\\,h with the eclipse depth decreasing to the red. The outburst eclipse profiles are symmetric, indicating an axisymmetrical brightness distribution in the accretion disc. We derive an inclination $i\\approx73^{\\circ}-79^{\\circ}$ from the eclipse duration. The quiescent spectrum obtained with the Hobby-Eberly Telescope shows double peaked emission lines of \\Ion{H}{I}, \\Ion{He}{I} and \\Ion{Fe}{I,II} and clearly identifies the system as a dwarf nova. Absorption features of the secondary star are detected at red wavelengths from which a spectral class M$3\\pm1.5$ and a distance of $d=320\\pm100$\\,pc are derived. ", "introduction": "Cataclysmic variables (CVs) may be discovered by various means. Historically, most of them were found because of their {\\em cataclysmic} nature, i.e. strong variability. This is especially valid for dwarf novae, which show quasi-regular outbursts in the visual of up to 8 magnitudes. With the advent of space-based X-ray telescopes, a new class of CVs was discovered, containing magnetic white dwarfs as accretors. The ROSAT and EUVE missions were extremely successful in discovering this type of CVs (e.g. Beuermann 1998\\nocite{beuermann98-1}). However, a large number of CVs are neither prominent X-ray sources, nor strongly variable. In non-magnetic CVs with a constantly high mass transfer rate --~novalike variables~-- the accretion disc remains in a perpetual hot state, turning them into unspectacular blue objects. Similarly, dwarf novae with low outburst amplitudes or long outburst cycles are likely to slip the attention of sky patrols. As a result, the sample of known CVs \\cite{downesetal97-1} is skewed by selection effects, and the actual space density of CVs is a matter of serious debate (e.g. de Kool 1992\\nocite{dekool92-1} and Patterson 1998\\nocite{patterson98-1}). The Hamburg Schmidt objective prism survey (HQS, Hagen et al. 1995\\nocite{hagenetal95-1}), originally aimed at the detection of a magnitude-limited sample of bright quasars (V=13--17.5), provides a rich source of CV candidates selected because of their {\\em spectroscopic} properties. Up to date, only few CVs have been serendipitously identified from the HQS: HS\\,0551+7241 \\cite{dobrzyckaetal98-1}; HS\\,1023+3900 \\cite{reimersetal99-1}; and HS\\,1804+6753 (=EX\\,Dra) \\cite{billingtonetal96-1,fiedleretal97-1}. The latter two objects show the strength of the spectroscopic selection of CV candidates: HS\\,1023+3900 is a magnetic CV with a very low accretion rate and no X-ray emission, and HS\\,1804+6753 is a bright eclipsing dwarf nova with low-amplitude outbursts, both stars were unlikely to be detected with the ``classic'' selection mechanisms described above. We have initiated a search for new CVs selected from the HQS with follow-up observations of CV candidates detected also in the ROSAT Bright Source catalogue \\cite{vogesetal99-1}. They were identified as possible CVs by Bade et al. \\cite*{badeetal98-1} because of the Balmer line emission seen in their HQS prism spectra. HS\\,0907+1902 (\\,=\\,1RXS\\,J090950.6+184956) was independently confirmed spectroscopically as CV at the BAO (X. Jiang, private communication). Here we report on the first photometric and spectroscopic results for HS\\,0907+1902. ", "conclusions": "We have discovered a bright new eclipsing dwarf nova with an orbital period of 4.2\\,h. Eclipsing CVs offer the best means of deriving the system parameters such as stellar masses, mass transfer rates, and the structure of the accretion disc. With its long orbital period, HS\\,0907+1902 is only the fourth deeply eclipsing dwarf nova above the $2-3$\\,h period gap, the other ones being IP\\,Peg, HS\\,1804+6753 (=EX\\,Dra), and BD\\,Pav. With a quiescent and an outburst magnitude of $V\\approx16$ and $V\\approx12.5$, respectively, HS\\,0907+1902 is well suited for detailed follow-up studies." }, "0003/astro-ph0003113_arXiv.txt": { "abstract": "We use data from the Hipparcos catalogue to construct colour-magnitude diagrams for the solar neighbourhood, which are then treated using advanced Bayesian analysis techniques to derive the star formation history, $SFR(t)$, of this region over the last 3 Gyr. The method we use allows the recovery of the underlying $SFR(t)$ without the need of assuming any {\\it a priori} structure or condition on $SFR(t)$, and hence yields a highly objective result. The remarkable accuracy of the data permits the reconstruction of the local $SFR(t)$ with an unprecedented time resolution of $\\approx 50 $ Myr. A $SFR(t)$ having an oscillatory component of period $\\approx 0.5$ Gyr is found, superimposed on a small level of constant star formation activity. Problems arising from the non-uniform selection function of the Hipparcos satellite are discussed and treated. Detailed statistical tests are then performed on the results, which confirm the inferred $SFR(t)$ to be compatible with the observed distribution of stars. ", "introduction": "The problem of deducing the star formation rate history, $SFR(t)$, of the Milky Way has been generally attempted in terms of indirect inferences mostly through chemical evolution models. The validity of these methods relies on the soundness of the assignation of a ``chemical age'' to each of the studied stars. Generally a metallicity indicator is chosen, and used to measure the metal content of a number of stars which are then binned into age groups through the use of an age-metallicity relation derived from a chemical evolution model. For example, Rocha-Pinto \\& Maciel (1997) take a variety of age-metallicity relations (AMRs) from the literature and use a closed box chemical evolution model to translate AMRs into $SFR(t)s$, allowing for an intrinsic Gaussian spread in the AMR assumed to be constant in time. The advantages of these methods are that large samples of stars both in the solar neighbourhood and further away within the disk of the galaxy can be studied. An inferred $SFR(t)$ can be constructed over an ample time range and spatial extent within the Galaxy which is consistent with the metallicities of the sample studied, and the chemical evolution model proposed. However, the validity of the AMR assumption can not be checked independently of the proposed chemical evolution model, and is necessarily dependent on what is chosen for the mixing physics of the ISM, the possible infall of primordial non enriched gas, and the still largely unknown galactic formation scenario in general. This last problem also affects attempts at inferring the $SFR(t)$ from stellar kinematics (e.g. Gomez et al. 1990, Marsakov et al., 1990). With the recent availability of the Hipparcos satellite catalogue (ESA 1997) we are now in a position to attempt recovery of the local $SFR(t)$ directly, without the need of any model dependent assumptions. Previous direct approaches have been undertaken through the binning of observed stars into age groups according to the degree of chromospheric activity as measured through selected emission line features, with conflicting results depending on the assumed age-activity relation (e.g. Barry, 1988, and Soderblom et al., 1991). Using this technique, Rocha-Pinto et al. (1999) have derived a star formation history from the chromospheric activity-age distribution of a larger sample comprising 552 stars, founding evidence for intermittency in the $SFR(t)$ over 14 Gyr. The Hipparcos catalogue offers high quality photometric data for a large number of stars in the solar neighbourhood, which can be used to construct a colour-magnitude diagram (CMD) for this region. Once a CMD is available, it is in principle possible to recover the $SFR(t)$ which gave rise to the observed distribution of stars, assuming only a stellar evolutionary model in terms of a set of stellar tracks. In practice the most common approach to inverting CMDs has been to propose a certain parameterization for the $SFR(t)$, which is used to construct synthetic CMDs, which are statistically compared to the observed ones to select the values of the parameters which result in a best match CMD. Examples of the above are Chiosi et al. (1989), Aparicio et al. (1990) and Mould et al. (1997) using Magellanic and local star clusters, and Mighell \\& Butcher (1992), Smecker-Hane et al. (1994), Tolstoy \\& Saha (1996), Aparicio \\& Gallart (1995) and Mighell (1997) using local dSph's. We have extended these methods in Hernandez et. al. (1999) (henceforth paper I) by combining a rigorous maximum likelihood statistical approach, analogous to what was introduced by Tolstoy \\& Saha (1996), with a variational calculus treatment. This allows a totally non-parametric solution of the problem, where no {\\it a priori} assumptions are introduced. This method was applied by Hernandez et al. (2000) (paper II) to a set of HST CMDs of local dSph galaxies to infer the $SFR(t)$ of these interesting systems. The result differs from what can be obtained from a chemical evolution model in that a direct answer is available, with a time resolution which depends only on the accuracy of the observations. Limitations on the applicability of our method to the Hipparcos data appear in connection to the selection function of the catalogue. The need to work only with complete volume-limited samples limits the age range over which we can recover the $SFR(t)$ to 0--3 Gyr, with a resolution of $\\sim 0.05 $ Gyr. This makes it impossible to compare our results with those of chemical evolution models which typically sample ages of 0--14 Gyr, with resolutions of 0.5--1.5 Gyr. In Section 2 we give a summarized review of the method introduced in paper I, the sample selection and results are discussed in section 3. Section 4 presents a careful statistical testing of our results, and Section 5 our conclusions. ", "conclusions": "We have applied the method developed in our paper I to the data of the Hipparcos catalogue. An objective answer for the $SFR(t)$ of the solar neighbourhood over the last 3 Gyr was found, which can be shown to be consistent with the complete volume-limited Hipparcos samples relevant to this age range. A structured $SFR(t)$ is obtained showing a cyclic pattern having a period of about 0.5 Gyr, superimposed on some degree of underlying star formation activity which increases slightly with age. No random bursting behaviour was found at the time resolution of 0.05 Gyr of our method. A first order density wave model for the repeated encounter of galactic arms could explain the observed regularity." }, "0003/astro-ph0003439_arXiv.txt": { "abstract": "We have observed twenty two galaxies at 100 \\micron\\ with the Kuiper Airborne Observatory in order to determine the size of their FIR emitting regions. Most of these galaxies are luminous far-infrared sources, with $L_{FIR} > 10^{11} L_{\\sun}$. This data constitutes the highest spatial resolution ever achieved on luminous galaxies in the far infrared. Our data includes direct measurements of the spatial structure of the sources, in which we look for departures from point source profiles. Additionally, comparison of our small beam 100 \\micron\\ fluxes with the large beam IRAS fluxes shows how much flux falls beyond our detectors but within the IRAS beam. Several sources with point-like cores show evidence for such a net flux deficit. We clearly resolved six of these galaxies at 100 \\micron\\ and have some evidence for extension in seven others. Those galaxies which we have resolved can have little of their 100 \\micron\\ flux directly emitted by a point-like active galactic nucleus (AGN). Dust heated to $\\sim$40 K by recent bursts of non-nuclear star formation provides the best explanation for their extreme FIR luminosity. In a few cases, heating of an extended region by a compact central source is also a plausible option. Assuming the FIR emission we see is from dust, we also use the sizes we derive to find the dust temperatures and optical depths at 100 \\micron\\ which we translate into an effective visual extinction through the galaxy. Our work shows that studies of the far infrared structure of luminous infrared galaxies is clearly within the capabilities of new generation far infrared instrumentation, such as SOFIA and SIRTF. ", "introduction": "Data from the Kuiper Airborne Observatory, and later the Infra-Red Astronomical Satellite (IRAS) revealed that many galaxies emit much more flux in the far- infrared (FIR, $\\sim$40 - 120 \\micron) than in any other wavelength band (\\cite{Tel80}; \\cite{Hou84}; \\cite{Soi84}; \\cite{deJ84}). This FIR is usually attributed to thermal emission from warm ($\\sim$40 K) dust heated by starbursts (\\cite{Els85}; \\cite{Law86}). There may also be a direct nonthermal contribution to the FIR luminosity from an active galactic nucleus (AGN), but at least the energy distributions suggest that emission from warm dust grains likely dominates the FIR flux even if there is an obvious active core and the galaxy is not obviously dusty (\\cite{Bar92}; \\cite{Chi92a}; \\cite{Soi87}; \\cite{Ede87a}). While the most likely scenario for the FIR emission in galaxies is dust absorbing strong radiation from a burst of young massive stars, other scenarios are possible. Extended dust could also be heated directly by an AGN core. In addition, the dust could be reprocessing less energetic photons from an older population of stars in the galaxy, an energy source that is known to dominate in relatively low luminosity quiescent systems. It has been suggested that the dust could be heated by hot ($10^{8}$ K) intergalactic gas (\\cite{Dwe92}; \\cite{Bre90}), but this would only happen when such hot gas exists in the vicinity, as in the center of a large cluster of galaxies. Dust could also be heated by shocks in the interstellar medium during the collision or interaction of galaxies (\\cite{Har87}). Each of the dust heating mechanisms above should correspond to particular spatial distributions of FIR light. If the FIR emission is from the active core itself, the flux should appear point-like. If it is from dust heated by an active nucleus, we might expect color temperature gradients to point toward the nucleus. Starburst heated dust should have about the same scale size as the burst itself because the young stars are well mixed with the gas from which the stars are forming and the optical depth for their UV photons is very high in the interstellar medium. On the other hand, dust heated by nonionizing photons from a cooler population of stars that expel condensables might be expected to follow the smooth distribution of older stars in the galaxy. It is well understood that evidence for interactions, collisions, and mergers is almost invariably associated with the most luminous infrared galaxies, and comparison of the far infrared luminosity distribution with that of the interaction geometry is a vital clue to the mechanism by which this energy is produced. The question that motivated this observational project was a straightforward one. Does the far infrared emission in luminous galaxies, representing the bulk of the emitted energy, show any evidence for spatial structure using the highest spatial resolutions that are available to us? The distribution of luminosity in these sources can be a key to understanding their energy production mechanisms. Our observations address this question, and we develop relevant observational strategies for future high resolution studies. An associated question is whether we can use these measurements to help distinguish between the various plausible heating mechanisms. High spatial resolution observations in the far infrared are seriously handicapped compared with observations in most other spectral regions. This part of the spectrum, containing the peak of energy emission from luminous galaxies, is inaccessible from the ground, and the spatial resolution that can be achieved is completely diffraction limited. A large aperture above the terrestrial water vapor absorption is a necessary tool. While large ground-based telescopes can attack the problem on arcsecond scales using tracers of hot and cool dust in the mid-IR and submillimeter continuum repectively, the correspondence of this dust with the distribution of warm dust that dominates the energy budget of luminous galaxies is not completely understood. To date, the largest telescopes that routinely observe in this spectral region are airborne and balloon-borne facilities. While its high sensitivity allowed it to catalog many sources, IRAS resolved only the largest and nearest galaxies because of its comparatively large beam size of $\\sim$2-4 arcmin at 60 and 100 \\micron. Deconvolution efforts improved on this resolution somewhat, but could not approach the $\\sim$23 arcsec diffraction-limited beam of the KAO. While ISO too has made fine contributions to the study of luminous galaxies, the higher S/N ratios that could be achieved in the far infrared continuum with ISOPHOT could not offset the substantially larger diffraction-limited beamsize of that telescope, for example the $\\sim$43 arcsec pixel size of the C100 channel. We observed the spatial structure of a sample of luminous galaxies with the Kuiper Airborne Observatory at 100 \\micron. These observations were among the last made during the long mission of this highly successful platform, which concluded in 1996. With its 0.9m aperture, the KAO allowed the highest possible diffraction-limited resolution at this wavelength. This work is likely to be the highest spatial resolution study available near the peak of the energy emission from these galaxies until the commissioning of SOFIA and SIRTF. Determination of the distribution of far infrared emission in luminous galaxies enables several phenomenological and astrophysical insights into these objects. Many of these galaxies are interacting systems, and little direct information is available about precisely where in these multiple objects the luminosity actually originates. Such information would greatly illuminate the mechanism by which galactic collisions produce these luminous sources. The scale size over which the warm dust is distributed allows, through a simple, single-slab model calculation, the optical depth of the emitting dust. This information is of interest in our understanding of the global structure of these sources, and the extent to which higher spatial resolution information at shorter, more extinction-dependent wavelengths can be trusted to give an accurate picture. This analysis also allows independent estimation of the grain temperatures from the spectral color temperatures, which can be significantly different in sources with substantial optical depth. In Section 2 we will discuss the sample of galaxies that we observed. Section 3 will cover details about the observations and Section 4 will describe the analysis methodology. Section 5 presents the general results from our study, along with a discussion for each galaxy. Further discussion of our results follows in Section 6. ", "conclusions": "We have observed the distribution of 100 \\micron\\ continuum emission in 22 galaxies, most of which have $L_{FIR} > 10^{11} L_{\\sun}$. We clearly resolved the emission ($D_{g} > 20$\\arcsec, $D_{e} > 12$\\arcsec) in 6 of them. We also see some evidence for extension in 7 others. For every resolved and possibly resolved source in our sample except for MCG+02-04-025, NGC 4151 and NGC 1275, we are able to eliminate all possible methods for FIR production except for starburst heated dust. Of the galaxies we could not resolve, most show correspondingly small starburst regions, and our size limits are consistent with those regions dominating the energetics. In a few cases, most notably NGC 3110, the disk of the galaxy makes a substantial contribution to the FIR flux. The contribution of the disk of NGC 3110, and the extended emission outside the bright cores in MCG+02-04-025 and NGC 6286 suggest that it is imprudent to assume all of the FIR flux in more distant FIR luminous galaxies (that we do not resolve) is concentrated in the core. A better universal model for the FIR flux distribution for these galaxies would be a strong central core on top of an extended ``plateau\" of emission from the disk. For the systems we resolved, we used the core sizes in a simple, single slab emission model to estimate not only the dust temperature T$_{d}$, but also the optical depth at 100 \\micron\\, $\\tau_{100}$, assuming a dust emissivity exponent of n = 1.5 (emissivity Q $\\propto \\lambda ^{-n}$). In addition, when they were available, we used fluxes at longer wavelengths that we assumed were still associated with the thermal dust emission. The additional flux measurements when combined with the measurements at 60 and 100 \\micron\\ allow us to solve for the emissivity exponent, n, as well. These measured values of n are all equal to or greater than 1.5 and two far exceeded the expected range of 1-2 (e.g. \\cite{Car92}; see Table 8). With the values of $\\tau_{100}$, we calculated an estimate for the visual extinction (via $A_{V} / \\tau_{100} \\sim 750$ for our galaxy, \\cite{Mak85}). Assuming that the stars are well mixed with the dust, we can tell how much the visible light from the galaxy (and in particular the FIR flux producing region) is extinguished. The single slab model is incapable of dealing with geometrical situations that almost surely exist in all of these galaxies such as a central condensation of material. Therefore the A$_{V}$ estimates are only a rough indicator of the true extinction of the galaxy. The $A_{V}$ estimates vary from an insignificant value of $\\sim$0.2 in NGC 1275 to an extremely high value of $\\sim$35 in NGC 7469. Interestingly, and probably not surprisingly, some of the highest extinctions we obtained came from those spiral galaxies which we view edge-on. We do not have enough resolving power to determine how the q-value varies within these distant FIR luminous galaxies. We do, however, have enough resolution to separate galaxy pairs for which IRAS gives only one flux. We confirm that UGC 2369(south), NGC 3110, NGC 6286, UGC 10923(west), NGC 7469, NGC 7541, NGC 7771, and UGC 12915 dominate the FIR flux over their companions. Our observations confirm that the q-value for NGC 6286 is unusually low. We also derive a new q = 2.26 for UGC 12915. The new value confirms that it is the radio bridge and not low FIR flux from the disk of UGC 12915 that causes the low q-value for the system. Our project has developed strategies that will be of use for future missions, in particular SOFIA. The large aperture of SOFIA will immediately provide a factor of three improvement in resolution over KAO. New detector arrays deing developed for SOFIA instruments (e.g. HAWC) will better sample this diffraction spot. In addition, the higher sensitivity of SOFIA will not only provide sensitivity to the structure at small scales, but will more specifically make available a large number of asteroids for point source and flux calibration." }, "0003/astro-ph0003263_arXiv.txt": { "abstract": "We describe the European Large Area {\\em ISO\\/} Survey (ELAIS). ELAIS was the largest single Open Time project conducted by {\\em ISO\\/}, mapping an area of 12 square degrees at 15$\\mu$m with {\\em ISO-CAM\\/} and at 90$\\mu$ with {\\em ISO-PHOT\\/}. Secondary surveys in other {\\em ISO\\/} bands were undertaken by the ELAIS team within the fields of the primary survey, with 6 square degrees being covered at 6.7$\\mu$m and 1 square degree at 175$\\mu$m. This paper discusses the goals of the project and the techniques employed in its construction, as well as presenting details of the observations carried out, the data from which are now in the public domain. We outline the ELAIS ``Preliminary Analysis'' which led to the detection of over 1000 sources from the 15 and 90 $\\mu$m surveys (the majority selected at 15$\\mu$m with a flux limit of $\\sim$3 mJy), to be fed into a ground--based follow--up campaign, as well as a programme of photometric observations of detected sources using both {\\em ISO-CAM\\/} and {\\em ISO-PHOT\\/}. We detail how the ELAIS survey complements other {\\em ISO\\/} surveys in terms of depth and areal coverage, and show that the extensive multi--wavelength coverage of the ELAIS fields resulting from our concerted and on--going follow--up programme has made these regions amongst the best studied areas of their size in the entire sky, and, therefore, natural targets for future surveys. This paper accompanies the release of extremely reliable sub-sets of the ``Preliminary Analysis'' products. Subsequent papers in this series will give further details of our data reduction techniques, reliability \\& completeness estimates and present the 15 and 90 $\\mu$m number counts from the ``Preliminary Analysis'', while a further series of papers will discuss in detail the results from the ELAIS ``Final Analysis'', as well as from the follow--up programme. ", "introduction": "The {\\em Infrared Space Observatory} \\cite{Kessler et al. 1996} was the natural successor to the {\\em Infrared Astronomical Satellite} ({\\em IRAS}), and has primarily been used to undertake detailed studies of individual objects and regions. However, {\\em ISO} also provided an opportunity to perform survey work at sensitivities beyond the reach of {\\em IRAS}. The {\\em IRAS} survey was of profound significance for cosmology, extragalactic astrophysics and for the study of stars, star-forming regions and the interstellar medium in the Galaxy. The mapping of large-scale structure \\cite{Saunders et al. 1991} in the galaxy distribution, the discovery of ultra-luminous infrared galaxies (see the review by Sanders \\& Mirabel \\shortcite{Sanders et al. 1996}) and of hyper-luminous infrared galaxies like {\\em IRAS} F10214+4724 \\cite{Rowan-Robinson et al. 1991a}, and the detection of proto-planetary discs around fairly evolved stars, were all unexpected discoveries of the {\\em IRAS} survey. The $z = 2.3$ galaxy F10214+4724, was at the limit of detectability by {\\em IRAS} ($S_{60}\\gsimeq 0.2$Jy). Several other $z > 1$ galaxies and quasars have now been found from follow-up of faint {\\em IRAS} samples. Recent sub-mm surveys, in particular with SCUBA on the JCMT, (e.g. Smail et al. \\shortcite{Smail et al. 1997} Hughes et al.\\shortcite{Hughes et al. 1998} Barger et al. \\shortcite{Barger et al. 1998} Eales et al. \\shortcite{Eales et al. 1999} Blain et al. \\shortcite{Blain et al. 1999} ) are detecting sources which are probably very high redshift counterparts to these {\\em IRAS} sources. Pointed observations of high redshift quasars and radio galaxies produce detections at sub-millimetre wavelengths in continuum and line emission, but mostly lie below the limit of the {\\em IRAS} survey at far infrared wavelengths. While designed as an observatory instrument, the huge improvement in sensitivity provided by {\\em ISO} offered the opportunity to probe the galaxy population to higher redshift than {\\em IRAS} and to make progress in understanding the obscured star formation history of the Universe. A significant fraction of the mission time was thus spent on field surveys. In this paper we describe the ``European Large Area {\\em ISO} Survey'' ({\\em ELAIS}) which represents the largest non-serendipitous survey conducted with {\\em ISO}. This survey provides a link between the {\\em IRAS} survey, the deeper {\\em ISO} surveys and the sub-mm surveys. {\\em ELAIS} is a collaboration involving 25 European institutes, led from Imperial College. This project surveyed around 12 square degrees of the sky at 15$\\mu$m and 90$\\mu$m nearly 6 square degrees at 6.7$\\mu$m together with a further one square degree at 175$\\mu$m. The survey used the {\\em ISO} Camera \\cite{Cesarsky et al. 1996} at the two shorter wavelengths and the {\\em ISO} Photometer \\cite{Lemke et al. 1996} at the longer wavelengths. {\\em ELAIS} was the largest open time project undertaken by {\\em ISO}: a total of 375 hours of scientifically validated data have been produced. We have detected over 1000 extra-galactic objects and a similar number of Galactic sources. Around 200 of these objects have been re-observed with {\\em ISO} to provide detailed mid/far infrared photometry. This paper outlines the broad scientific objectives of this project and describes the selection of the observing modes and survey fields. It also details the execution of the {\\em ISO} observations and briefly outlines the data reduction and data products. Finally we show how this survey complements other {\\em ISO} surveys and summarise the extensive multi-wavelength programmes taking place in the {\\em ELAIS} fields. ", "conclusions": "In this paper we have described the motivation behind ELAIS, the largest non-serendipitous survey performed by {\\em ISO\\/}. Our primary goals in conducting the survey were to determine the relative importance and recent evolution of the dust--obscured mode of star formation in galaxies, and to constrain AGN unification models, and we detailed above how these influenced our selection of survey fields and observational parameters. The fields that have been covered by {\\em ISO\\/} are also being extensively mapped from radio to X--ray wavelengths as part of a concerted ground--based follow--up programme, whose multi--wavelength coverage will make the {\\em ELAIS\\/} regions fertile ground for undertaking future astrophysical investigations extending well beyond our initial survey aims. Subsequent papers in this series will discuss in detail the scientific results from the ELAIS ``Preliminary Analysis'' and ``Final Analysis''. The first of these papers will include: discussions of the extra-galactic counts from the ``Preliminary Analysis'' at 7 and 15 \\micron (Serjeant et al. 1999), and at 90$\\mu$m (Efstathiou et al., 1999, in preparation); discussion of the stellar calibration and counts (Crockett et al., in preparation); and a discussion of sources detected in the multiply--repeated areas (Oliver et al., in preparation). Preliminary ELAIS data products were released through our WWW page (\\verb+http://athena.ph.ic.ac.uk/+), which also contains further details on the programme and the follow-up campaign." }, "0003/astro-ph0003055_arXiv.txt": { "abstract": "Recently Murayama \\& Taniguchi proposed that a significant part of the high-ionization nuclear emission-line region (HINER) in Seyfert nuclei arises from the inner wall of dusty tori because type 1 Seyfert nuclei (S1s) show the excess HINER emission with respect to type 2 Seyfert ones (S2s). This means that the radiation from the HINER has the anisotropic property and thus statistical properties of the HINER emission can be used to investigate the viewing angle toward dusty tori for various types of Seyfert nuclei. In order to investigate viewing angles toward narrow-line Seyfert 1 galaxies (NLS1s) and intermediate types of Seyferts (i.e., type 1.5, 1.8, and 1.9 Seyfert galaxies; hereafter S1.5, S1.8, and S1.9, respectively), we apply this HINER test to them. We also apply the same test for S2s with/without the hidden broad line region. A sample of Seyfert nuclei analyzed here consists of 124 Seyferts compiled from the literature. Our main results and suggestions are as follows. (1) The NLS1s are viewed from a more face-on view toward dusty tori than the S2s. However, the HINER properties of the NLS1s are indistinguishable from those of the S1s. (2) The S1.5s appear to comprise heterogeneous populations; e.g., a) some of them may be seen from an intermediate viewing angle between S1s and S2s, b) some S1.5s are basically S1s but a significant part of the broad-line region (BLR) emission is accidentally obscured by dense, clumpy gas clouds, or c) some S1.5s are basically S2s but a part of the BLR emission can be seen from some optically-thin regions of the dusty torus. (3) The S1.8s, the S1.9s and the objects showing either a broad Pa$\\beta$ line or polarized broad Balmer lines are seen from a large inclination angle and the emission from the BLRs of such objects reaches us through optically-thin parts of dusty tori. These three results support strongly the current unified model of Seyfert nuclei. And, (4) the line ratios of [Fe {\\sc x}]$\\lambda$6374 to the low-ionization emission-lines have rather isotropic property than those of [Fe {\\sc vii}]$\\lambda$6087. Therefore it is suggested that the [Fe {\\sc x}]$\\lambda$6374 emission is not useful investigating the viewing angle toward the dusty torus in Seyfert nuclei. The most plausible reason seems that the [Fe {\\sc x}]$\\lambda$6374 emission is spatially extended and thus its strength tends to show less viewing angle dependence. ", "introduction": "Seyfert galaxies have been broadly classified into two classes based on the presence or absence of broad emission lines in their optical spectra (Khachikian \\& Weedman 1974): Seyfert galaxies with broad lines are type 1 (hereafter S1) while those without broad lines are type 2 (S2). According to the current unified model of Seyfert nuclei (Antonucci \\& Miller 1985; see for a review Antonucci 1993), this difference between S1 and S2 can be explained as follows. The broad-line region (BLR) is located in the very inner region (e.g., a typical radial distance from the central black hole is $r \\sim$ 0.01 pc; e.g., Peterson 1993) and is surrounded by a geometrically and optically thick dusty torus. Therefore, the visibility of the central engine as well as the BLR is strongly affected by the viewing angle toward the dusty torus and then the difference between S1s and S2s is naturally understood. Indeed, this unified scheme has been supported by various observational results, for example, obscured X-ray emission in S2s (Awaki et al. 1991; Rush et al. 1996), colors of mid-infrared (MIR) emission (Pier \\& Krolik 1992, 1993; Murayama, Mouri, \\& Taniguchi 2000), MIR luminosity distributions (Heckman, Chambers, \\& Postman 1992; Maiolino et al. 1995), polarized broad emission lines mentioned below, and the results of multi wavelength observational tests (Mulchaey et al. 1994). In order to understand Seyfert nuclei and active galactic nuclei (AGNs) more comprehensively, any new observational tests toward the unified model are very important. In addition to the traditional two types of Seyfert nuclei, it is known that some Seyfert nuclei show intermediate properties between S1 and S2; type 1.2 (S1.2), type 1.5 (S1.5), type 1.8 (S1.8), and type 1.9 (S1.9) (Osterbrock \\& Koski 1976; Osterbrock 1977, 1981b; Cohen 1983; Winkler 1992; Whittle 1992), which show both the narrow and broad components in the Balmer emission lines. It is also noted that the objects without BLR in their optical spectra (i.e., S2s) do not comprise a simple population. First, some S2s show a broad Pa$\\beta$ line (Goodrich, Veilleux, \\& Hill 1994; Hill, Goodrich, \\& Depoy 1996; Veilleux, Goodrich, \\& Hill 1997), providing evidence for highly reddened BLRs in these objects. Second, the hidden BLR is detected only in a part ($\\sim$ 20\\%) of S2s in the polarized optical spectra (Antonucci \\& Miller 1985; Miller \\& Goodrich 1990; Tran, Miller, \\& Kay 1992; Kay 1994; Tran 1995a, 1995b, 1995c); the survey promoted by Lick Observatory found 10 S2s with the hidden broad line among 50 S2s. Another important type of Seyfert nuclei is narrow-line Seyfert 1 galaxies (NLS1s; Davidson \\& Kinman 1978). Optical emission-line properties of the NLS1s are summarized as follows (e.g., Osterbrock \\& Pogge 1985). (1) The Balmer lines are only slightly broader than the forbidden lines such as [O {\\sc iii}]$\\lambda$5007 (typically less than 2000 km s$^{-1}$). This property makes NLS1s a distinct type of S1s. (2) The [O {\\sc iii}]$\\lambda$5007/H$\\beta$ intensity ratio is smaller than 3. This criterion has introduced to discriminate S1s from S2s by Shuder \\& Osterbrock (1981). (3) They present strong Fe {\\sc ii} emission lines which are often seen in S1s but generally not in S2s. And, (4) the soft X-ray spectra of NLS1s are very steep (Puchnarewicz et al. 1992; Boller, Brandt, \\& Fink 1996; Wang, Brinkmann, \\& Bergeron 1996) and highly variable (Boller et al. 1996; Turner et al. 1999a). Because of these complex properties, it has not yet been fully understood what NLS1s are in the context of the current unified model of Seyfert nuclei while various models for NLS1s have been proposed (see for reviews Boller et al. 1996; Taniguchi, Murayama, \\& Nagao 1999). Recently, Murayama \\& Taniguchi (1998a; hereafter MT98a) have found that S1s have excess [Fe {\\sc vii}]$\\lambda$6087 emission with respect to S2s. This means that a significantly large fraction of the high-ionization nuclear emission-line region (HINER; Binette 1985; Murayama, Taniguchi, \\& Iwasawa 1998) traced by [Fe {\\sc vii}]$\\lambda$6087 resides in a viewing-angle dependent region; i.e., the inner wall of dusty tori (Murayama \\& Taniguchi 1998b; see also Pier \\& Voit 1995). Accordingly, it turns out that the HINER provides the indicator of the viewing angle for dusty tori of Seyfert nuclei. In this paper, we report on our statistical analysis of the HINER in the various types of Seyfert nuclei. ", "conclusions": "The anisotropic property of the radiation from the HINER traced [Fe {\\sc vii}] reported by MT98a has been statistically confirmed using the larger sample. The line ratios of [Fe {\\sc x}] to the low-ionization emission lines show a rather isotropic property with respect to those of [Fe {\\sc vii}] to the low-ionization emission lines. This may be interpreted by an idea that a significant fraction of the [Fe {\\sc x}] emission arises from low-density ISM as suggested by Korista \\& Ferland (1989). We note that the [Fe {\\sc x}] emission is not suitable to investigate the viewing angle toward the dusty tori of Seyfert nuclei. We have also investigated the HINER properties of the intermediate-type of Seyfert nuclei. Using the frequency distributions of the line ratios of [Fe {\\sc vii}] to the low-ionization emission lines, we find the following suggestions. (1) The NLS1s are viewed from a more face-on orientation toward dusty tori than the S2s. (2) The line ratios of S1.5s are distributed in a wide range from the smallest value of the S2s to the largest value of the S1s. This suggests that the S1.5s are heterogeneous populations. (3) The HINER properties of the S1.8s, the S1.9s and the objects showing a broad Pa$\\beta$ line or polarized broad Balmer lines are considerably different from those of the S1s. These facts mean that the ``S2$^+$'' objects are those which are seen from a large inclination angle and their BLR emission comes through optically-thin line of sights toward the dusty tori." }, "0003/astro-ph0003373_arXiv.txt": { "abstract": "The phenomenology associated with line profile variations in $\\delta$ Scuti stars is reviewed. The three main techniques adopted to detect pulsation modes, i.e., the pixel--by--pixel analysis, the analysis of the moment time series and the Fourier Doppler Imaging are presented and discussed. Their application to the observational data has allowed the reliable detection of several pulsation modes, many of them not detectable with photometric observations. The need of better coordination of simultaneous spectroscopic and photometric campaigns in order to get the complete pulsation spectra of these stars is pointed out. The observation in different campaigns of the same objects has proved that dramatic changes in the mode amplitudes are quite common, even in a couple of years (and maybe less). In the last part of the paper a recent approach to the mode detection by direct fit of line profile variations in multiperiodic pulsators is described. ", "introduction": "While the presence of radial velocity variations in $\\delta$ Scuti itself was detected just at the beginning of the century (Wright 1900), the first hints about the possible presence in the same star of line profile variations (hereinafter LPV) were due to Struve about half a century later (Struve 1953). After a decade of quiescence the studies of the phenomena associated with LPV were resumed by several authors in the mid sixties and seventies. Several different phenomena were announced, such as variable emission in the cores of strong lines, large equivalent width variations and so on (for a review see the paper by Breger 1979). Unfortunately those findings were probably connected to the use of the photographic plate as a detector. As a matter of fact later studies of the same objects performed with photoelectric detectors were not able to confirm any of those findings. The first reliable observations of LPV in $\\delta$ Scuti stars were performed at the beginning of the eighties with the advent of the Reticon detector. Several stars were observed by Campos \\& Smith (1980) and Smith (1982) but unluckily most of them for a few hours in one night only, i.e., while the presence of line profile variations was clearly detected, the data were not sufficient to analyze the variations and to detect periodicities. Generally the few data were phased with the periods available from photometry or radial velocity curves (usually at that time only one or two periods per star were known). Among the stars studied by these authors only 28 Aql had a sufficient dataset (4 consecutive nights) which allowed the analysis of line width and radial velocity variations. A period--finding program confirmed the presence in both parameters of the two known periods derived from photometry (Smith 1982). The successive papers showed, beside the changing shape of the line profiles, the presence of moving sub-features in the lines (e.g., Yang \\& Walker 1986, Walker et al. 1987). In these cases the data were also not adequate for independent period searches, although two attempts were made by Kennelly et al. (1991, 1992), who analyzed 2.5 hours and 5 hours of observations of $k^2$ Boo and $\\gamma$ Boo, respectively. They did it by sampling at 10 km~s$^{-1}$ intervals the intensity residuals in the line profiles derived from the subtraction of the average profile, and computing the Discrete Fourier Transform. Obviously, given the extremely short temporal baseline, the frequency resolution was very poor, allowing only a rough confirmation of the photometric periods. In the following years it was realized that more intensive campaigns on a few selected objects were more fruitful than sporadic observations of many targets. This was in part the consequence of the photometric studies that were showing that usually many close frequencies were present in the pulsation spectra and consequently longer baselines were needed in order to resolve these frequencies. Now before discussing in detail the results concerning these campaigns we shall briefly review the phenomenology associated with the LPV, the constraints that it imposes on the observational programmes and the techniques we can adopt to analyze the data. ", "conclusions": "In this paper we have presented the main observational characteristics of LPV in $\\delta$ Scuti stars, and we have discussed how they can be observed and which techniques can be adopted to detect the pulsation modes. Moreover, a review of the main results obtained from the application of these techniques has been given. As we have seen, the number of stars carefully studied is at the moment scanty, so it is difficult to draw very general conclusions, although some facts, which can be useful as guidelines for future studies, can be pointed out: \\begin{itemize} \\item{} The careful observation and analysis of LPV allows several pulsation modes to be detected. If data with adequate $S/N$ and temporal baselines are available the number of spectroscopically detected periods rivals the number which are photometrically detected. Many of these detections are purely spectroscopic, especially for stars with high $v\\sin i$: photometric observations alone are not sufficient to get the complete pulsation spectrum. \\item{} The variability of the mode amplitudes is quite common. Sometimes some modes have amplitudes below the detectability threshold. Again, to get the complete pulsation spectrum, we need to observe the star at different epochs, and it may be desirable to get higher S/N data in order to better monitor the evolution of the amplitudes. \\item{} Period detections can be rather reliable: several of the spectroscopically detected modes have also been photometrically detected and moreover, for the stars with two spectroscopic campaigns, many of the periods have been independently detected in both datasets. Again, to do a good job, we need high $S/N$ with long baselines. \\item{} The three different analysis techniques presented in Section 4 tend to be complementary: the moment method is particularly suited for low-degree modes; FDI works preferably on high--degree ones and on stars with high $v\\sin i$; the pixel-by-pixel analysis works both on low- and high-degree modes, and, for the detection of the low-degree ones, it is preferable to study stars with low $v\\sin i$. For instance, with this technique we were able to detect several modes in FG Vir, which has a $v\\sin i$ of only 20 km/s. \\item{} Given the complexity of the pulsation spectra, in order to get adequate frequency resolution and to avoid aliasing ambiguities, we need relatively long baselines (possibly comparable to the photometric ones) as well as multi-site campaigns. \\item{} We have already pointed out the need for data with good $S/N$. Given the very small amplitudes of most of the modes (a few thousandths of the continuum intensity), in order not to to merely detect them, but also to proceed to their identification, a $S/N$ better than 500 is desirable. \\item{} Since together with high $S/N$ data we also need high spectral and temporal resolutions and short exposure times (unless large telescopes are used for extended periods of time, which is almost a hopeless possibility), it is important to develop and to adopt techniques that add the information from many spectral lines. This would entail the use of medium size telescopes, which are more accessible. Much work is still to be done in this direction, especially if the added information is used not merely for mode detection but for identification, too. \\item{} As an immediate consequence of the previous items, it is evident that the best strategy to get a complete picture of the pulsation behavior of a $\\delta$ Scuti star is to observe it simultaneously both photometrically and spectroscopically for several seasons and in the context of multi-site campaigns. In this respect the few researchers active in the field should coordinate their efforts in order to study a few carefully selected targets. \\end{itemize} Finally in the last part of this paper an approach to mode detection by LPV fitting in multiperiodic variables has been presented. This approach at the moment is not free of problems; for example, a better model of the LPV to fit to the data is needed, a model that at the same time is not too cumbersome, because the routine which uses it is called thousands of times during the minimization procedure. Another problem is the same one faced by the moment method described by Aerts \\& Eyer (this proceedings), i.e., the lack of confidence intervals for the minima of the discriminant, which makes it difficult to compare the competing modes. This second problem could be rather easily overcome once the first has been successfully addressed and the errors on the reconstruction of the LPV produced by each mode correctly estimated, because the discriminants (eqs. 5--7) have a $\\chi^2$ shape and therefore their statistical properties are well known." }, "0003/astro-ph0003145_arXiv.txt": { "abstract": "The results of a high-resolution interferometric campaign targeting the symbiotic long-period variable (LPV) R~Aqr are reported. With both near-infrared measurements on baselines out to 10\\,m and mid-infrared data extending to 32\\,m, we have been able to measure the characteristic sizes of regions from the photosphere of the LPV and its extended molecular atmosphere, out to the cooler circumstellar dust shell. The near-infrared data were taken using aperture masking interferometry on the Keck-I telescope and show R~Aqr to be partially resolved for wavelengths out to 2.2\\,$\\micron$ but with a marked enlargement, possibly due to molecular opacity, at 3.1\\,$\\micron$. Mid-infrared interferometric measurements were obtained with the U.C. Berkeley Infrared Spatial Interferometer (ISI) operating at 11.15~$\\micron$ from 1992 to 1999. Although this dataset is somewhat heterogeneous with incomplete coverage of the Fourier plane and sampling of the pulsation cycle, clear changes in the mid-infrared brightness distribution were observed, both as a function of position angle on the sky and as a function of pulsation phase. Spherically symmetric radiative transfer calculations of uniform-outflow dust shell models produce brightness distributions and spectra which partially explain the data, however limitations to this approximation are noted. Evidence for significant deviation from circular symmetry was found in the mid-infrared and more tentatively at 3.08\\,$\\micron$ in the near-infrared, however no clear detection of binarity or of non-LPV elements in the symbiotic system is reported. ", "introduction": "R~Aqr is a mass-losing long-period variable (LPV) star believed to be in a symbiotic system with an obscured hot companion whose presence is betrayed by nebular emission lines. Long noted for its peculiar elliptical visible nebulosity extending up to 2' (first reported in Lampland 1923), this star has recently been subject to intense scrutiny with the latest generation of astronomical instrumentation both on the ground and in space. Much of this attention can be traced to the detection in the optical (\\cite{WalGrens80}; \\cite{Herbig80}) and radio (\\cite{Sopka_et82}) of what is believed to be the nearest astrophysical jet to the earth ($\\sim 200$\\,pc; \\cite{VanL_et97}; \\cite{HPL97}). Thought to originate in the accretion disk around an unseen hot sub-dwarf, this jet has been studied extensively at UV, optical and radio wavelengths (e.g. \\cite{BVP92}; \\cite{PH94}; \\cite{LJ92}). A difference in location of the peak intensity of the $\\nu =1, J = 1-0$ SiO maser line and the nearby lineless continuum from 7\\,mm VLA maps was reported by Hollis, Pedelty and Lyon (1997). These authors interpreted the SiO peak as the location of the LPV and the continuum peak marking the accretion disk of the companion, and were thus able to derive a binary separation of $55\\pm2$\\,mas with a position angle of $18^\\circ \\pm 2^\\circ$. With its ability to penetrate layers of obscuring dust, the infrared might be thought an ideal wavelength to image the inner regions of this system, from photospheres of the stars to dust and jets in their immediate surroundings. However the high angular resolutions required and the presence of the enormously luminous LPV have limited the efficacy of this approach. Infrared interferometric measurements with the IOTA array (\\cite{vanB96}) yielded photospheric diameter measurements of $\\sim$14--15\\,mas for the LPV but no evidence for an additional component. Speckle observations at J band (\\cite{KMC94}) suggestive of an elongated image were interpreted in terms of a distorted inner dust shell, possibly due to the presence of the companion. Anandarao \\& Pottasch (1986) were able to match the near- to far-IR fluxes using a two-temperature dust shell model, while more recently Le Sidaner \\& Le Bertre (1996) were able to fit similar spectral energy distributions with a simple uniform outflow model. In the following section of this paper, we briefly describe our observational methods and apparatus for obtaining near- and mid-IR interferometric measurements. In Section~3 the experimental results are given, together with discussion of model fitting and physical interpretation. A brief summary of our conclusions is found in Section~4. ", "conclusions": "High-resolution interferometric studies of R~Aqr at narrow bandwidths within the J, H and K bands find no evidence for any significant departure from the best-fitting model of a marginally-resolved stellar disk, which we identify with the LPV component of the system. Any companion present and separated by more than $\\sim 15$\\,mas must exhibit a magnitude difference in excess of $\\Delta M $\\simge~5\\,mag in the near infrared. These results are consistent with those of Van Belle et al. (1996), but not with Karovska et al. (1994). An enlargement, by approximately a factor of 2, is reported for the apparent size at a wavelength of 3.08\\,$\\micron$ as compared with shorter near-infrared bands, which is attributed to molecular blanketing by an unidentified species in the atmosphere, or thermal emission from material in an extended halo. Tentative evidence for asymmetry at this wavelength is also reported. In the mid-IR, visibility data obtained with the ISI have been used to constrain simple uniform spherical outflow models. Using self-consistent radiative transfer calculations, good fits to most of the visibility data have been obtained. However, serious shortcomings were found when comparing the synthetic spectral energy distributions to measurements, many of which may be attributed to inadequate knowledge of the detailed dust shell parameters. Furthermore, mid-IR interferometry obtained at a different position angle on the sky points towards substantial departures both from spherical symmetry and from a simple radial distribution. These indicate that the shortcomings of the present models are probably in large part be due to the oversimplistic assumption of an isotropic uniform outflow. Further high-resolution studies, both in near-infrared with higher spatial resolution and dynamic range, and in the mid-infrared with more complete Fourier coverage, are important in order to fill in the gaps in understanding of this star." }, "0003/astro-ph0003235_arXiv.txt": { "abstract": "In evolved supernova remnants (SNRs) interacting with molecular clouds, a highly inhomogeneous structure consisting of a forward shock of moderate Mach number, a cooling layer, a dense radiative shell and an interior region filled with hot tenuous plasma is expected. We present a model of nonthermal electron injection, acceleration and propagation in that environment and find that these SNRs are efficient electron accelerators and sources of hard X- and \\gr emission. A forward shock of velocity $v_s \\gsim$ 100 km s$^{-1}$ with an ionized precursor propagating into the molecular cloud accompanied by magnetohydrodynamic turbulence provides a spatially inhomogeneous distribution of nonthermal electrons. The energy spectrum of the nonthermal electrons is shaped by the joint action of first and second order Fermi acceleration in a turbulent plasma with substantial Coulomb losses. Bremsstrahlung, synchrotron, and inverse Compton radiation of the nonthermal electrons produce multiwavelength photon spectra in quantitative agreement with the radio and the hard emission observed by {\\it ASCA} and {\\it EGRET} from IC~443. We distinguish interclump shock wave emission from molecular clump shock wave emission; particles reach higher energies in the interclump shock and that is the likely source of $\\gamma$-ray emission and radio synchrotron emission. Spatially resolved X- and \\gr spectra from the supernova remnants IC~443, W44, and 3C391 as might be observed with {\\it BeppoSAX, Chandra XRO, XMM, INTEGRAL} and {\\it GLAST} would distinguish the contribution of the energetic lepton component to the $\\gamma$-rays observed by {\\it EGRET}, constraining the cosmic ray nuclear component spectra in these SNRs. These data would provide a valuable tool for studying the complex structure of molecular clouds where SNR radiative shocks interact with dense molecular clumps. ", "introduction": "Supernovae and their remnants have long been identified as likely sites of galactic cosmic ray acceleration. Evidence for electron acceleration first came from observations of radio synchrotron radiation. The observation of nonthermal X-ray emission from SN 1006 by {\\it ASCA} has provided evidence for electron acceleration up to $\\sim 100$ TeV (Koyama \\etal 1995; Reynolds 1998). The proton cosmic ray component is more difficult to observe, but it has long been recognized that pion decays from collisions with interstellar gas could give an observable flux of $0.1-1$ GeV photons. Initial estimates of $\\gamma$-ray emission from supernova remnants (SNRs), concentrating on the component resulting from pion decays, were made by Chevalier (1977), Blandford \\& Cowie (1982), and Drury, Aharonian, \\& V\\\"olk (1994). When $\\gamma$-ray emission was apparently detected from SNRs by {\\it CGRO}, the $\\gamma$-ray spectrum could not be fitted by a pure pion decay spectrum and some other component was needed (Esposito \\etal 1996; Sturner \\etal 1997; Gaisser, Protheroe, \\& Stanev 1998). In addition to the pion decays, the relevant processes are bremsstrahlung emission of relativistic electrons and inverse Compton emission. Gaisser \\etal (1998) modeled these processes in detail in order to fit the observed $\\gamma$-ray spectra of the remnants IC~443 and $\\gamma$ Cygni. They assumed acceleration to a power law spectrum in the shock front and determined the spectral index, electron to proton ratio, and the upper energy cutoff. The evolution of the relativistic component in a supernova remnant has been modeled by a number of groups. Sturner \\etal (1997) assumed particle acceleration in the shock front with an $E^{-2}$ energy spectrum and followed the evolution of the particle spectrum. Their SNR model was homogeneous: a factor 4 density jump at the shock front, constant density in the interior, and a constant magnetic field in the interior. Their shock acceleration model had a diffusion length parameter that was important for the maximum energy reached by the particles. They modeled the nonthermal emission from IC~443 as synchrotron emission in the radio and bremsstrahlung in $\\gamma$-rays. de Jager \\& Mastichiadis (1997) dealt with the same processes in W44, as well as inverse Compton emission. They noted that the observed radio spectrum is flatter than would be expected from shock acceleration of newly injected particles and suggested that the particles originated from a pulsar in the supernova remnant. Recently, Ostrowski (1999) showed that second-order electron acceleration by the turbulent medium just after the shock could flatten the spectra to account for the radio observations of IC~443. Baring \\etal (1999) presented calculations of the broad-band emission from nonlinear shock models of shell-type SNRs. They used Sedov adiabatic shock dynamics in a homogeneous medium and a Monte Carlo simulation of the particle acceleration taking into account the nonlinear shock structure. The set of models considered by Baring \\etal (1999) covers the range of shock speeds 490 $\\leq v_S \\leq$ 4000 km s$^{-1}$ and ambient medium number densities 10$^{-3}\\leq n \\leq$ 1 cm$^{-3}$. Massive stars that are the likely progenitors of core collapse supernovae are expected to be spatially correlated with molecular clouds. The remnants that are likely to be $\\gamma$-ray sources in {\\it CGRO} observations (Esposito \\etal 1996) also show evidence for interaction with molecular gas. Chevalier (1999) recently studied the evolution of supernova remnants in molecular clouds and concluded that many aspects of the multiwavelength observations could be understood in a model where the remnants evolve in the interclump medium of a molecular cloud, which has density of $5-25$ H atoms cm$^{-3}$, and become radiative at radii $\\sim 6$ pc. The compression in the radiative shell is limited by the magnetic field. Molecular emission occurs when the radiative shell collides with molecular clumps. IC~443 is the remnant with the best evidence for high energy emission and also shows clear evidence for interaction of the shock with a molecular cloud (e.g., Burton \\etal 1990; van Dishoeck \\etal 1993; Cesarsky \\etal 1999). Most of the current models of particle acceleration in SNRs have dealt with adiabatic remnants in a homogeneous medium (see, however, the model of Boulares \\& Cox 1988 for the Cygnus Loop and the model of Jones \\& Kang 1993). In contrast, our aims here are to examine in detail the nonthermal emission of radiative supernova remnants in molecular clouds using a kinetic model of electron injection, acceleration and propagation, including the inhomogeneous structure deduced from multiwavelength SNR observations. We use global MHD (magnetohydrodynamic) models of supernova remnants interacting with molecular clouds as described by Chevalier (1999), including the theory of radiative shock structure by Shull \\& McKee (1979) and the kinetic model of electron injection and acceleration by shocks from Bykov \\& Uvarov (1999). The model described here is relevant to the mixed-morphology SNRs (Rho \\& Petre 1998), if they are interacting with molecular clouds. These remnants comprise a sizable fraction of the Galactic SNR population. The plan of our paper is as follows. In \\S~2, we treat the energy spectrum of energetic electrons as they pass through the regions associated with a radiative shock wave. We include the shock acceleration of particles from the thermal pool. In \\S~3, we present calculations of the nonthermal emission from the relativistic particles, including emission from shocks in clumps as well as emission from the radiative shock wave in the interclump region. Our model is specifically applied to the well-observed remnant IC 443, although we expect it to more generally apply to other supernova remnants in molecular clouds. We also examine the ionization by energetic particles and the energy involved in the relativistic particles. A discussion of results and future propects is in \\S~4. A detailed discussion of expected range of densities, shock velocities and magnetic fields determined from multiwavelength observations of IC 443, W44 and 3C391 was given in the paper by Chevalier (1999, and references therein). ", "conclusions": "For individual supernova remnants, the ambient density is an important parameter that can be estimated from multiwavelength observations (Chevalier 1999). Another important parameter for modeling the nonthermal particles in a SNR in a molecular cloud is the collisionless MHD turbulence spectrum, particularly in the radiative shock cooling layer. A substantial level of MHD collisionless turbulence could overcome Coulomb losses in the dense plasma downstream from the radiative shock. An accurate model of MHD turbulence in the postshock cooling layer is not available now. Thus we considered both limiting cases described in the \\S~2 and show in Fig. 1 the uncertainty introduced by the lack of data concerning the MHD turbulence properties. Coulomb losses are important in the postshock cooling region only for electrons with energies below $E_C \\approx$ 20 keV if the large scale turbulent velocity has a substantial longitudinal component of $\\sim 20 \\kms$. A higher value ($E_C \\gsim$ 8 MeV) is expected in the absence of a longitudinal component of large scale turbulence in the postshock cooling layer because the Coulomb losses are overcome by resonant interaction with Alfven waves. Since the Alfven velocity is relatively low ($\\sim 5\\kms$), the value of $E_C$ is much higher in that case. We used a conservative minimum value of $E_C$ = 120 keV for the radiative shock structure described above, although one could expect $E_C \\sim$ 20 keV in the most favorable case. The nonthermal emission of SNR in a molecular cloud (especially in the hard X-ray regime) and ionization structure of the radiative shock are sensitive to the MHD turbulence model. Spatially resolved observations of the nonthermal emission from SNRs may be able to constrain models of MHD turbulence. All of the SNR candidates from the {\\it EGRET} list: IC 443, $\\gamma$ Cyg, W44, and Monoceros (Esposito \\etal 1996) are old remnants interacting with molecular clouds. The \\gr emission from electrons accelerated by the radiative shock calculated for the IC 443 parameters presented in Figs. 1 and 2 is in good agreement with that observed by {\\it EGRET} (Esposito \\etal 1996), as well as with the upper limits established by {\\it Whipple} (Buckley \\etal 1998). The high energy $\\gamma$-rays ($\\gsim 50$ MeV) from the radiative shell should be spatially correlated with the radio emission. The {\\it EGRET} telescope detected an extended excess (95\\% confidence circle $\\sim 42^{\\prime}$ in radius) from the Monoceros SNR correlated with the 1.42 GHz radio emission (Esposito \\etal 1996). This is in accordance with the model of a leptonic origin of GeV emission from extended SNRs interacting with clouds (de Jager \\& Mastichiadis 1997). {\\it GLAST} (Gamma-ray Large Area Space Telescope) will be an excellent instrument for future $\\gamma$-ray observations, because it will have the capability of spatially resolving the $\\gamma$-ray emission from SNRs. Spatially resolved observations of W44, IC 443, 3C391 and some other mixed-morphology SNRs from the list given by Rho \\& Petre (1998) with {\\it BeppoSAX, Chandra XRO, XMM} would be very valuable tools to test our model. The hard X-ray detector (HXD) aboard the forthcoming {\\it ASTRO-E} mission could be used for observations of the hard 10--700 keV continuum with a field of view of 0.8$^{\\circ}$ FWHM at 60 keV and 2.8$^{\\circ}$ at 500 keV from the extended hard X-ray emission from IC 443, W44 and the Monoceros SNR predicted by the radiative shock model (see Fig. 1 and Fig. 3). Spatially resolved spectra are needed to distinguish the shell emission from the shocked clump emission. Molecular clumps interacting with moderately fast radiative shocks are also expected to be sources of hard X-rays and MeV $\\gamma$-rays (up to 100s of MeV). Nonthermal continuum radio emission (100s of MHz) with a sharply rising $\\nu F_{\\nu}$ spectrum (Fig. 3) and a time dependent cut-off frequency is expected in the fast shock model. Due to potentially substantial internal free-free absorption, the spectrum of radio emission from a localized clump might constrain the ionized gas density in the clump. The MeV \\gr spectrum of a localized clump may be resolved with the forthcoming \\gr missions {\\it INTEGRAL} (e.g., Sch\\\"onfelder 1999; Winkler 1999) and {\\it GLAST}. With an expected angular resolution about 12$^{\\prime}$ FWHM, imager {\\it IBIS} aboard {\\it INTEGRAL} would allow detection of hard X- ray ($\\gsim$ 50 keV) emission from molecular clouds within $\\sim$ 1 kpc. A comparison of hard emission spectra with the radio spectrum can provide valuable information about the density and magnetic field in a clump. A possible variability on a timescale of a few years for clump hard emission and radio emission (on a longer time scale) could further constrain the model. For low velocity (below $\\sim 30 \\kms$) shocks interacting with a dense ($\\sim 10^4 \\cmc$), magnetized ($B \\lsim$ 0.5 mG) molecular clump, one may expect hard X-ray emission below the MeV regime correlated with the regions of molecular emission of shocked gas. Nonthermal radio continuum is not expected in the case of low velocity shocks because of a lack of accelerated relativistic electrons at sufficiently high energies. Finally, we note that the high density of energetic particles in the vicinity of a shock wave in a molecular cloud can affect the ionization and thermal properties of the gas. Rich molecular spectra have been observed from IC 443 and 3C391 (Reach \\& Rho 1999). It will be interesting to see whether there is a signature of the presence of nonthermal particles that can be discerned from the molecular spectra." }, "0003/astro-ph0003286_arXiv.txt": { "abstract": "We study radiation-hydrodynamical normal modes of radiation-supported accretion disks in the WKB limit. It has long been known that in the large optical depth limit the standard equilibrium is unstable to convection. We study how the growth rate depends on location within the disk, optical depth, disk rotation, and the way in which the local dissipation rate depends on density and pressure. The greatest growth rates are found near the disk surface. Rotation stabilizes vertical wavevectors, so that growing modes tend to have nearly-horizontal wavevectors. Over the likely range of optical depths, the linear growth rate for convective instability has only a weak dependence on disk opacity. Perturbations to the dissipation have little effect on convective mode growth rates, but can cause growth of radiation sound waves. ", "introduction": "Shakura \\& Sunyaev (1973) predicted that the inner portions of accretion disks that extend into relativistically-deep gravitational potentials should be radiation pressure-dominated when the accretion rate is greater than a modest fraction of the Eddington rate. In that regime, they found that disks could achieve hydrostatic balance in the vertical direction if the local dissipation rate were proportional to the local mass density. Given that assumption, upward radiation flux could support the disk matter against gravity if the density were essentially constant as a function of height (falling sharply to zero at the top surface) and the radiation pressure fell gradually from the disk midplane to the surface. Soon after this equilibrium was discovered, it was found to suffer from several sorts of instabilities. Lightman \\& Eardley (1974) pointed out that if the viscous stress is proportional to the total pressure (in this case, dominated by radiation), perturbations with radial wavelengths long compared to the vertical thickness $h$, but short compared to a radius $r$, grow on the (comparatively long) viscous inflow timescale. Shakura \\& Sunyaev (1976) then observed that in these conditions perturbations in the same range of wavelengths would also grow on the (shorter) thermal timescale. Bisnovatyi-Kogan \\& Blinnikov (1977) noticed that if the radiation is locked to the gas even on short lengthscales (i.e., if, for the purpose of dynamics, the optical depth is treated as effectively infinite), such disks should be convectively unstable, for the specific entropy decreases outward; the linear growth rate for convective ``bubbles\" was worked out by Lominadze \\& Chagelishvili (1984). More recently, Gammie (1998) has demonstrated that a magnetic field in radiation-supported disks can catalyze a short-wavelength ($kh \\gg 1$) overstable wave mode. In view of these instabilities, it has long been a puzzle just what sort of equilibrium would actually be found in Nature when the accretion rate is high enough that radiation pressure-domination might be expected (see, e.g. Shapiro, Lightman \\& Eardley 1976; Liang 1977; Coroniti 1981; Svensson \\& Zdziarski 1994; Szuszkiewicz \\& Miller 1997; Krolik 1998). In this paper, we take a closer look at the nature of the short wavelength modes in radiation-supported disks without magnetic fields. Our goal (motivated by a companion work on radiation-hydrodynamics simulations of such disks: Agol \\& Krolik 2000b) is to examine more closely which modes can be expected to grow most quickly, what happens when finite optical depth permits some photon diffusion, and what role, if any, is played by associated perturbations in the local dissipation rate. ", "conclusions": "" }, "0003/astro-ph0003309_arXiv.txt": { "abstract": "A significant fraction of flat-spectrum, radio-loud quasars display most of the characteristics of relativistically beamed, high-optical polarization blazars, yet are weakly polarized in the optical regime ($m_{\\rm opt} \\lesssim 3\\%$). We have conducted a high-resolution polarization study with the VLBA at 22 and 43 GHz to look for differences in the parsec-scale magnetic field structures of 18 high- and low-optically polarized, compact radio-loud quasars (HPQs and LPRQs, respectively). We find a strong correlation between the polarization level of the unresolved parsec-scale radio core at 43 GHz and overall optical polarization of the source, which suggests a common (possibly co-spatial) origin for the emission at these two wavelengths. The electric vectors of the polarized 43 GHz radio cores are roughly aligned with the inner jet direction, indicating magnetic fields perpendicular to the flow. Similar orientations are seen in the optical, suggesting that the polarized flux at both wavelengths is due to one or more strong transverse shocks located very close to the base of the jet. In LPRQs, these shocks appear to be weak near the core, and gradually increase in strength down the jet. The LPRQs in our sample tend to have less luminous radio cores than the HPQs, and jet components with magnetic fields predominantly parallel to the jet. The components in HPQ jets, on the other hand, tend to have perpendicular magnetic field orientations. These differences cannot be accounted for by a simple model in which HPQs and LPRQs are the same type of object, seen at different angles to the line of sight. A more likely scenario is that LPRQs represent a quiescent phase of blazar activity, in which the inner jet flow does not contain strong shocks. Our high-resolution observations have shown that high rotation measures (up to 3000 $\\rm rad\\ m^{-2}$) previously seen in the nuclear regions of HPQs are present in LPRQs as well. The low-redshift quasars in our sample tend to have jet components with larger 43/22 GHz depolarization ratios than those found in the high-redshift sources. This may be due to small-scale magnetic field fluctuations in the Faraday screens that are being smeared out in the high-redshift sources by the poorer spatial resolution of the restoring beam. ", "introduction": "Since its introduction in 1978, the term ``blazar'' has been synonymous with radio-loud active galactic nuclei that have steep, smooth optical continua, highly core-dominated radio morphologies, and fluxes that are highly variable at all wavelengths. The bulk of their radiation is thought to be highly relativistically beamed synchrotron emission from plasma outflows in the form of jets. An important defining characteristic of blazars is their high degree of polarization at radio through optical wavelengths. This polarization is quite variable, often on short time-scales ($\\lesssim 1$ day), and indicates spatially small emitting regions with well-ordered magnetic fields. These are likely formed by relativistic shock fronts which re-order an originally tangled magnetic field in the jet. This shock model (e.g., \\citealt*{HAA85}; \\citealt{MG85}) has successfully reproduced many of the observed radio properties of blazar jets. One of the best indicators of blazar activity is the level of fractional polarization in the optical regime ($m_{\\rm opt}$). Studies of large AGN samples (e.g., \\citealt*{SMA84}; \\citealt{BSW90}) have shown that nearly all ``normal'' (i.e., radio-quiet) quasars have very weak, non-variable optical polarizations ($\\sim 0.5 \\%$). With the exception of a few sources such as OI 287, whose high polarizations can be attributed to scattering \\citep[e.g.,][]{RS88}, the optical polarizations of radio-quiet quasars rarely exceed $\\sim 3 \\%$. Those of blazars, on the other hand, span a large range, up to $\\sim 46$\\% in some cases \\citep{MBB90}, and are attributed to synchrotron emission from their relativistic jets. This dichotomy has led to the classification scheme ``high-optically polarized quasar'' (HPQ) for sources with $m_{\\rm opt} > 3\\%$ (i.e., blazars) and ``low-optically polarized quasar'' (LPQ) for those AGNs with $m_{\\rm opt} < 3\\%$. The connection between high-optical polarization and jet synchrotron emission might suggest that all core-dominated, radio-loud AGNs containing relativistically beamed jets should be HPQs, but this is not the case. Many blazars have optical polarizations that occasionally dip below $3\\%$, and there are many other radio-loud AGNs that display most of the characteristics of blazars, but have consistently low optical polarizations. Perhaps the most famous example of a low-optically polarized, radio-loud quasar (LPRQ) is that of 3C~273, a well-studied superluminal source, whose optical polarization has rarely exceeded 3\\%. High-sensitivity photo-polarimetry of this object by \\citet{IMT89} revealed a ``mini-blazar'' component, whose overall contribution to the optical flux is swamped by a strong optical continuum, possibly from a large, hot accretion disk. If its blazar component were significantly stronger, 3C~273 would in all likelihood have the properties of a typical HPQ. The reason why all radio-loud AGNs are not HPQs may lie with their relativistic jets, as there is ample evidence showing a link between the optical polarization and radio jet properties of blazars. For example, the optical electric polarization vectors of radio-loud AGNs are known to be well-aligned with their parsec-scale \\citep{RS85} and kiloparsec-scale \\citep*{SAM79} jets, indicating a shared radio and optical (and possibly co-spatial) emission mechanism. This view has been supported by variability studies such as that of \\citet{HB92} and \\citet{VVS91}, who found correlated flaring activity at optical and radio wavelengths. Also, \\citet*{GSS96} studied a small sample of blazars and found weak evidence that the levels of optical and parsec-scale radio polarization were correlated. Some authors \\citep[e.g.,][]{F88} have speculated that LPRQs and HPQs may represent the quiescent and active phases, respectively, of the same object. Others have suggested that orientation plays an important role, with the jets of LPRQs being oriented farther from the line of sight \\citep{VTU92}. The latter model is supported somewhat by observations showing that LPRQs are generally less variable (\\citealt{VTU92}), and have smaller misalignments between their parsec- and kiloparsec-scale jet directions \\citep{ILT91,XU94}. In this paper, we present observations that show a direct link between the optical polarization and parsec-scale radio properties of compact, radio-loud AGNs. We also show that there are intrinsic differences in the jets of LPRQs and HPQs that cannot be explained purely by differences in orientation. These intrinsic differences are associated with the magnetic field structure of the parsec-scale jet, which are in turn responsible for the optical-through-radio polarization properties of compact radio quasars. ", "conclusions": "In this paper we have used high-frequency VLBI and optical polarimetry to compare the parsec-scale magnetic field properties of a sample of high-optical polarization, compact radio quasars with a sample of similar objects having low optical polarizations. We summarize our results as follows: 1. We find a strong correlation between the level of optical polarization and radio core polarization at 43 GHz. The more optically polarized quasars also have higher core luminosities, core-to-extended flux ratios, and flatter integrated 22/43 GHz spectral indices. These trends strongly indicate that the optically polarized emission is synchrotron radiation, and is co-spatial with the radio core emission at 43 GHz. 2. The electric vectors of the highly polarized 43 GHz radio cores are roughly aligned with the inner jet, indicating magnetic fields perpendicular to the flow. A similar configuration is seen in the optical, suggesting that the polarized flux at both wavelengths is due to one or more strong transverse shocks located very close to the base of the jet. 3. There is a strong trend for the fractional polarizations of bright jet components to increase downstream from the core. The fact that all of the components with magnetic fields parallel to the jet are located near the core suggests that this trend is not due to an underlying longitudinal magnetic field of increasing strength. Rather, we believe it is either the result of the jet curving away from the line of sight, or an increase in shock strength along the jet. 4. We find evidence for large rotation measures (up to $\\sim 3000 \\ \\rm rad \\ m^{-2}$) in the nuclear regions of low-optically polarized radio quasars, which are indicative of parsec-scale Faraday screens with organized magnetic fields. The low-redshift quasars in our sample tend to have jet components with larger 43/22 GHz depolarization ratios than those found in the high-redshift sources. This might be due to small-scale magnetic field fluctuations in the Faraday screens that are being smeared out in the high-redshift sources by the poorer spatial resolution of the restoring beam. 5. We find that the parsec-scale jet properties of compact radio quasars are highly dependent on their level of optical polarization. Sources with optical fractional polarizations below $3\\%$ (low-optically polarized radio quasars: LPRQs) tend to have lower 43 GHz core polarizations, fainter cores, steeper total spectral indices, smaller pc/kpc-scale jet misalignments, and smaller core-to-extended flux ratios than high-optically polarized quasars (HPQs). Although the components in the jets of HPQs and LPRQs have similar fractional polarizations, those found in HPQs tend to have magnetic fields that are perpendicular to the jet, while those in LPRQ jets have mainly parallel orientations. 6. The observed differences in LPRQs and HPQs cannot be fully explained by a model in which LPRQs are seen at larger angles from the line of sight. Instead, our data are more consistent with an evolutionary scenario based on \\citet{F88}, in which flat-spectrum quasars go through episodic stages of blazar-like activity. During these phases, they emit strongly shocked components with transverse magnetic fields, which move at a variety of speeds down the jet. LPRQs may represent quiescent phases of blazars in which only weak shocks are generated in the flow." }, "0003/astro-ph0003415_arXiv.txt": { "abstract": "The deviation from the power-law decline of the optical flux observed in GRB~970228 and GRB~980326 has been used recently to argue in favor of the connection between GRBs and supernovae. We consider an alternative explanation for this phenomenon, based on the scattering of a prompt optical burst by $0.1M_{\\odot}$ of dust located beyond its sublimation radius $0.1-1\\,{\\rm pc}$ from the burst. In both cases, the optical energy observed at the time of the first detection of the afterglow suffices to produce an echo after $\\sim20-30$~d, as observed. Prompt optical monitoring of future bursts and multiband photometry of the afterglows will enable quantitative tests of simple models of dust reprocessing and a prediction of the source redshift. ", "introduction": "\\label{intro} The relationship between Gamma Ray Bursts (GRBs) and supernovae has become increasingly interesting over the past year. Though exploding massive stars have long been considered as possible progenitors of GRBs (e.g. \\fcitep{woo93}), no evidence existed to support these theories until observations of the afterglow of GRB~980425 suggested an association of the burst with an unusual supernova 1998bw (\\fcitep{gea98}, \\fcitep{kea98}). Later reanalysis of the optical afterglow lightcurves of two other bursts, GRB~970228 (\\fcitep{fea99}) and GRB~980326 (\\fcitep{bea99}) showed a deviation from the power-law decline expected if the emission is due to synchrotron radiation from electrons accelerated by the blast wave. In both cases a significant excess emission was observed around $\\sim 30$ days after the gamma-ray burst, with simultaneous reddening of the spectrum. \\fcitet{bea99}, \\fcitet{gea99}, and \\fcitet{rei99} attribute this excess to the emission from an underlying supernova event. The relationship of GRBs to SN explosions is a question of great importance, since it provides a powerful clue to the fundamental nature of these objects. However, the evidence presented so far is circumstantial -- the association of GRB~980425 with SN 1998bw is unproven and the excess emission seen from GRB~970228 and GRB~980326 is based upon relatively few actual measurements -- and possible alternative explanations need to be seriously considered, if only to strengthen the case for the SN explanation. In this spirit, \\fcitet{wad99} suggested that the red excess emission observed in GRB~970228 and GRB~980326 is due to dust in the vicinity of the burst progenitor absorbing and then re-radiating the optical/UV flash observed shortly after the recent GRB~990123 (\\fcitep{aea99}) and generally attributed to the reverse shock which propagates into the fireball ejecta (\\fcitep{mrp94}, \\fcitep{mer97}, \\fcitep{pam98}, \\fcitep{sap99}). However, the \\acite{wad99} scenario has two shortcomings. Firstly, the equilibrium temperature of dust is limited to $\\sim2300$~K and so the emission should peak at $\\sim 2\\,(1+z)\\,{\\rm \\mu m}$ (where $z$ is the GRB redshift), although a small amount of higher temperature emission may be produced by the dust as it is subliming. Secondly, the optical flash is so powerful that the sublimation radius lies beyond $\\sim10$~pc from the GRB. Thus, in this picture it is rather difficult to reproduce the observed flux in the $0.4-0.8\\,{\\rm \\mu m}$ band with a time delay of order a few weeks. In this letter we propose an alternative explanation, which relies on the {\\em scattering} of the direct optical transient emitted in the first day by dust as the primary source of excess optical radiation. The fundamental point is that in the two observed cases, assuming isotropic emission, the fluence of the {\\it observed transient} exceeds that of the reported excess and the {\\it unobserved transient} is even larger if we extrapolate to earlier times. A fraction of this emission scattered from a radius where dust can outlive the optical transient should therefore produce a delayed echo. As dust absorbs selectively as well as scatters, the echo is likely to be significantly redder than the original optical transient, as reported. In the next section, we describe our model for the dust scattering properties and then present the results in the context of the observed GRBs in \\S\\ref{obs}. Implications for future tests of our scenario are discussed in \\S\\ref{impl}. We assume $h=0.6$, $\\Omega_{\\rm M} = 0.3$, and $\\Omega_{\\Lambda} = 0.7$ so that the angular diameter distance of the GRB is $D_{\\rm A}=1.5-2$~Gpc for $0.5\\lsim z\\lsim3$. ", "conclusions": "\\label{impl} In this letter we present an alternative explanation for the reddened excess emission observed in GRB~970228 and GRB~980326, which we attribute to dust scattering of the early-time, afterglow emission. This scenario is predictive enough to be confirmed or ruled out with observations of future GRBs. In particular, in contrast to the supernova explanation (\\fcitep{bea99}; \\fcitep{gea99}; \\fcitep{rei99}), if the excess emission is due to dust scattering, then its properties will depend on the luminosity of the optical transient. HETE II (http://space.mit.edu/HETE/) scheduled to be launched in early 2000 and Swift (http://swift.gsfc.nasa.gov/homepage.html), scheduled for 2003, should provide real-time localization of GRB X-ray afterglows with sufficient precision to permit faster follow-up and better measurements of its total fluence. Infrared observations may discover the expected thermal emission from hot subliming dust ({\\it cf} \\fcitep{wad99}). In fact dust emission might be the correct explanation for the ``near-IR'' bump seen in the spectrum of the GRB~991216 afterglow (\\fcitep{fea00}). Note that as most GRBs are at redshifts $\\gsim 0.5$, 3~${\\rm\\mu m}$, (as as opposed to the more common 2~${\\rm\\mu m}$) photometry may be necessary to see this emission. In those GRBs, where it is also possible to measure a redshift, the the simplest model of dust-scattering is over-contrained and therefore refutable. Beaming and dust inhomogeneity introduce additional uncertainty but such models may also be excludable. For example, if ROTSE (\\fcitep{aea99}) were to detect another optical flash in a GRB as luminous as that seen in GRB~990123, which had an isotropic luminosity $L\\sim10^{51}$~erg s$^{-1}$, then dust should be physically sublimed out to a distance $R_{\\rm sub}\\sim100$~pc along the line of sight. Unreasonably large beaming would then be required to explain a dust echo with a delay of only a few weeks. Alternatively, if the radio light curve in an afterglow tracked the optical light curve, then this would be incompatible with both dust scattering and a supernova. A further prediction of the dust echo model is that, unless the dust and OT are both arranged axisymmetrically with respect to the line of sight, we expect there to be linear polarization associated with dust echos and this may be measurable in bright examples. (1.7 percent polarization has been reported in the optical transient associated with GRB 990510 by \\fcitet{cov99} but this is unlikely to be due to scattering.) In conclusion, we have demonstrated that dust scattering can account for the excess optical emission observed in the afterglows of two GRBs as an alternative to an underlying supernova explosion. Future observations should be able to rule out or confirm this explanation." }, "0003/astro-ph0003401_arXiv.txt": { "abstract": "We present the evolutionary properties of a set of massive stellar models (namely 13, 15, 20 and 25 $\\rm M_\\odot$) from the main sequence phase up to the onset of the iron core collapse. All these models have initial solar chemical composition, i.e. $Y=0.285$ and $Z=0.02$. A 179 isotope network, extending from neutron up to $\\rm ^{68}Zn$ and fully coupled to the evolutionary code has been adopted from the Carbon burning onward. Our results are compared, whenever possible, to similar computations available in literature. ", "introduction": "This is the second paper in a series devoted to a detailed study of the evolutionary properties of massive stars up to the onset of the final collapse. In the first paper of this series (\\cite{CLS98} 1998 - hereinafter Paper I) we have discussed in some detail the relevant literature on the subject and the improvements progressively made in the computation of the advanced burning phases in these last two decades: such a discussion will not be repeated here. We discussed also the latest version of our evolutionary code FRANEC (version 4.2) together with the evolutionary properties of a $\\rm 25~M_\\odot$ star, from the Pre Main Sequence up to the onset of the collapse. That evolution was computed by adopting a network which included 12 isotopes for the H burning, 25 isotopes for the He burning and 149 isotopes for all the more advanced burning phases. Though such a network was well suited to present that first test model, an accurate tracing of the chemical evolution of the matter requires a significantly larger nuclear network and hence we now adopt in all our computations a much more extended network (see section \\ref{code}) with respect to the one adopted in Paper I. In this paper we present the hydrostatic evolution of a first set of massive stars in the mass range $\\rm 13-25~M_\\odot$ while in the next one of the series we will present the evolution of metal poor ($Z=10^{-3}$ and $Z=0$) massive stars together with the explosive yields. For the moment we did not extend the computations outside of this mass range because a) stars less massive than $\\rm 13~M_\\odot$ are close enough to the limit of semidegenerate C ignition that the computer time required to follow them becomes exceedingly large for the computers we have at present and b) stars more massive than $\\rm 25~M_\\odot$ begin to be significantly affected by mass loss, phenomenon which we do not want to include yet. Each subsection of section \\ref{results} will be devoted to the analysis of a specific burning (both central and in shell) and to its dependence on the initial mass. Section \\ref{disc} is devoted to a general discussion of the results together with a comparison with similar computations available in literature. A last summary and conclusions follows. Since at the end of the writing of this paper the computation of all the other masses of lower metallicity (i.e. $Z=0$ and $Z=10^{-3}$) were completed together with their explosive nucleosynthesis, since our tables of explosive yields have been already distributed to many researchers, and since at present they are also available at the web site \\hfill\\break \\centerline{\\url{http://www.mporzio.astro.it/$^\\sim$mandrake/orfeo.html}} we decided to add our full set of the elemental explosive yields as appendix A to the present paper. ", "conclusions": "In this paper, which is the second of the series, we have presented and discussed in some detail the presupernova evolution of four stellar masses having solar chemical composition. All the various burning phases have been studied by adopting a very extended network fully embedded into the stellar evolutionary code (FRANEC). A comparison of our results with similar ones available in the literature shows that, in spite of an overall similarity, the existing differences are significant and at present not completely understood." }, "0003/astro-ph0003017_arXiv.txt": { "abstract": " ", "introduction": "The terminology of {\\it narrow-line} Seyfert 1 galaxies (NLS1s) originates from their relatively $narrow$ optical Balmer-line emission with $v_{\\rm BLR} \\lsim 2000$km s$^{-1}$ compared with those of usual Seyfert 1 galaxies with broad-line emission, BLS1s (see a concise review by Brandt 1999 and references therein). Their optical line ratios and remarkable Fe{\\sc ii} lines (Osterbrock, Pogge 1985; Halpern, Oke 1987; Grupe et al. 1999) are also distinct from those of BLS1s. NLS1s exhibit unique X-ray properties ($e.g.$, Grupe et al. 1998); they are characterized by large soft X-ray excess (Pounds et al. 1996; Otani et al. 1996; Leighly 1999b) and a good correlation is known to exist between the strength of soft excess and FWHM of optical Balmer lines (Boller et al. 1996; Laor et al. 1997). Rapid soft/hard X-ray variability is another signature characterizing NLS1s (Otani et al. 1996; Boller et al. 1997; Leighly 1999a). NLS1s are not rare but comprise a significant part (say, 20\\%) of the Seyfert 1 galaxies and it seems that a group of NLS1s is smoothly connected to the class of BLS1s (Brandt, Boller 1998). It is often suggested that the strength of the soft X-ray excess relative to the hard X-ray power law appears to be directly related to the `primary eigenvector' of Boroson and Green (1992), which represents the strongest set of optical emission line correlations ($e.g.$, Brandt 1999). In this sense, NLS1s are located at the extreme end of the primary eigenvector. The UV line properties of NLS1s also fits this scheme (Wills et al. 1999). Then, what is the control parameter which drives the primary eigenvector? What physical factor distinguishes NLS1s from BLS1s? There are several models proposed to account for the narrowness of otherwise broad-line emission (see Boller et al. 1996 for extensive discussion). Successful models should account for a smooth continuation between NLS1s and BLS1s. Among them, one fascinating and probably most promising explanation is that NLS1s contain relatively less massive black holes (with $M \\sim 10^{5-7}M_\\odot$). Moderate brightness of NLS1s, $L \\sim 10^{43-45}$erg s$^{-1}$, implies similar mass-flow rates, $\\dot M$, to those of BLS1s. It then follows that the ratio of $L/L_{\\rm E}$ (or $\\dot M/M$) is relatively large for NLS1s. The control parameter which drives the eigen vector could be the fraction of the Eddington rate at which the supermassive black hole is accreting, ${\\dot M}/(L_{\\rm E}/c^2)$. For small black-hole masses, narrow Balmer line emission of NLS1s can be understood, provided that the broad-line clouds (which emit broad-line emission in BLS1s) are bound in the potential by a central black hole (Laor et al. 1997). Since the radius of the broad-line clouds are roughly scaled as $r_{\\rm BLR} \\propto L^{0.5}$ from the reverberation mapping (Kaspi et al. 1996), the circular velocity of broad-line clouds, $\\sim \\sqrt{GM/r_{\\rm BLR}}$, is systematically smaller for smaller $M$ at a constant $L$ (and thus constant $r_{\\rm BLR}$), yielding narrow Balmer emission. For a luminosity close to the Eddington, \\begin{equation} L_{\\rm E} \\simeq 1.2\\times 10^{43} M_5~{\\rm erg~s}^{-1}, \\end{equation} where $M_5 \\equiv M/(10^5M_\\odot)$, what do we then expect theoretically for the disk structure and its spectra? It is known that for such a high luminosity advective energy transport dominates over radiative cooling (Abramowicz et al. 1988; section 10.3 of Kato et al. 1998). Such a disk is named as the slim disk, since it is moderately geometrically thick. The slim disk is an optically thick version of ADAF (advection-dominated accretion flow) and should not be confused with optically thin ADAF proposed for low-luminosity AGNs (see section 10.2 of Kato et al. 1998 and references therein). Unlike the optically thin ADAF the observable features of the slim disk have been poorly investigated. Szuszkiewicz et al. (1996) was the first to start investigation along this line (see also Wang et al. 1999), and we give more detailed discussion on the observational consequences in the present study. Watarai et al. (2000a) studied the super-critical accretion flow in the context of Galactic black-hole candidates (GBHC), showing some unique observational features of the slim disk. We focus on the AGN case in the present study. Methods of calculations are given in section 2. We then discuss the non-standard emission properties of the slim disk in the context of NLS1s in section 3. We next turn to the subject of large variability and discuss its origin in relation to magnetic-field activity in section 4. Outstanding issues will be discussed in section 5. The final section is devoted to conclusions. Throughout the present study we use the normalization of $\\dot M$ with $\\dot M_{\\rm crit} \\equiv L_{\\rm E}/c^2$; i.e., \\begin{eqnarray} {\\dot m} \\equiv {\\dot M}/{\\dot M}_{\\rm crit} &\\!\\!\\sim\\!\\!& {\\dot M}/(1.3\\times 10^{22} M_5~ {\\rm g~s}^{-1}) \\cr &\\!\\!\\sim\\!\\!& {\\dot M}/(2.5\\times 10^{-4}M_5M_\\odot{\\rm ~yr}^{-1}). \\end{eqnarray} ", "conclusions": "To summarize, we discussed various aspects of the slim disk in the context of NLS1s in comparison with the other disk models (see table 3). The followings are the main conclusions. \\vspace{-6pt} \\begin{enumerate} \\item The disk begins to exhibit the slim-disk properties, when the luminosity moderately exceeds $L_{\\rm E}$. Then, substantial amount of radiation is expected to originate from inside the marginally stable last circular orbit at $3r_{\\rm S}$. Therefore, a small $r_{\\rm bb} < 3r_{\\rm S}$ does not necessarily mean the presence of a Kerr hole. \\item If we fit the spectrum of the slim disk with the disk blackbody model, the derived size of the X-ray emitting region will be $r_{\\rm in} \\lsim 3 r_{\\rm S}$ and the maximum temperature will be $kT_{\\rm in} \\sim 0.2M_5^{-1/4}$keV. Further, $p$ $(\\equiv d\\ln T_{\\rm eff}/d\\ln r)$ decreases from $0.75$ to $\\sim 0.5$ as $\\dot M$ increases. \\item All the ASCA data of NLS1s fall on the region with supercritical accretion, $\\dot m > 10$, supporting that the disks in NLS1s are likely to be slim disks. \\item The slim disk can also produce large fluctuations because release of magnetic-field energy (which will produce fluctuations via magnetic flares) dominates over the energy release due to persistent emission. \\end{enumerate} \\vspace{1pc} \\par We are grateful for A. Laor and an anonymous referee for useful comments. This work was supported in part by the Grants-in Aid of the Ministry of Education, Science, Sports, and Culture of Japan (10640228, SM)." }, "0003/astro-ph0003221_arXiv.txt": { "abstract": "We prove the existence of the old and metal-rich moving group HR 1614. This is done using the new Hipparcos parallaxes in combination with metallicities derived from Str\\\"omgren photometry, supported by dynamical simulations of the evolution of old moving groups in the galactic potential. A new selection criterion for this moving group is presented as well as a new, extended sample of probable member stars. In particular we find that the HR 1614 moving group has an age of about 2 Gyr (using Bertelli et al. 1994 isochrones) and a [Fe/H] $\\simeq 0.19 \\pm 0.06$ dex. We also revisit and apply our new selection criterion to the samples in Eggen (1992) and Eggen (1998b). It is found that, when binaries and stars with too low metallicity have been removed, 15 of his stars fulfill our criteria. ", "introduction": "Globular and open clusters provide useful probes of the longterm chemical and dynamical evolution of the Milky Way. The globular clusters probe the formation and early evolution of the spheroidal components of the Milky Way while the open clusters provide a useful tool to study the evolution of the galactic disk. However, the paucity of very old (old is here taken to mean $\\gtrsim 10^8$ years, the time scale on which an open cluster will be dissolved, Spitzer 1958) open clusters in the disk forces us to consider the considerably more loosely arranged moving groups to probe the earlier evolution of the disk. It was Olin Eggen who first introduced the concept of moving or stellar kinematic groups, of which the Hyades is a well known example. The basic idea behind the moving groups is that stars form in clusters and thus with similar space motion, on top of which the random motions of single stars are added, resulting in a modest velocity dispersion within the group. Through the orbital motion within the galactic potential the group will be stretched out into a tube-like structure and finally, after several galactic orbits, dissolve. The result of the stretching is that the stars will appear, if the Sun happens to be inside the tube, all over the sky but may be identified as a group through their common space velocity. Thus the moving groups may provide the essential, and so far largely un-utilized, link between cluster and field stars. These are the assumptions, but are moving groups observable realities? A large stumbling block for assessing the reality of moving groups has been the lack of large numbers of reliable parallaxes. This has now been largely overcome by the observations from the Hipparcos satellite (ESA 1997). This has, in fact, resulted in a small burst of recent papers studying, mainly young, moving groups, e.g. Asiain et al. (1999), Barrado y Navascu\\'es (1998), Odenkirchen et al. (1998), Skuljan et al. (1997), and Dehnen (1998). Eggen defined moving groups as stars that all share the same velocity in the direction of galactic rotation, i.e. V-velocity. Specifically the velocities required to be constant were corrected for the stars differing radial distance from the Sun in order to make the circular orbits iso-periodic (Eggen 1998b). However, using the Hipparcos parallaxes it is noted that firstly the groups get more compact and secondly that stars identified as group members do not form flat bars or ellipses with small $\\sigma_V$ but are in fact structures tilted in the $UV$-plane (Skuljan et al. 1997 Fig. 1). Part of this shape can be attributed to the errors in the parallaxes themselves and their transformation into errors in the $UV$-plane. However, through dynamical simulations Skuljan et al. (1997) show that all of the tilt cannot be attributed to the errors in parallaxes but also has a physical basis. In view of these new possibilities it is now appropriate to re-asses the reality and membership criteria for the HR 1614 moving group. Eggen (1998b) has compared Hipparcos and cluster parallaxes for stars in his sample (Eggen 1992) of HR 1614 moving group member stars. However, he disregards the Hipparcos parallaxes in favour of cluster parallaxes, also when the discrepancies are large, without further discussion. The article is organized as follows: Sects. 2, 3, and 4 describe the search for the HR 1614 moving group in the Hipparcos catalogue, as well as dynamical simulations of old moving groups and their characteristics today. Sect. 5 reviews previous work on the HR 1614 moving group. A new selection criterion for HR 1614 moving group is developed in Sect. 6 and used to derive its age. In Sect. 7. we derive, from data in the literature, a metallicity for the moving group. The Eggen (1992, 1998b) sample is revisited in Sect. 8 and discussed in detail. Sect. 9 contains a discussion primarily of possible sources of contamination in our sample. Sect. 10 provides a brief summary of the main results of this paper. ", "conclusions": "\\label{findconcl.sect} We close the first part of our paper with a few remarks on the reality of old moving groups. Before the Hipparcos mission the numbers of reliable parallaxes were too small to address the reality of most proposed old moving groups successfully. Recent studies (e.g. Barrado y Navascu\\'es 1998, and Skuljan et al. 1997) have shown that well known moving groups such as the Pleiades and the Hyades but also several other young moving groups, e.g. Castor ($200\\pm100$ Myr, Barrado y Navascu\\'es 1998), are well identified as physical entities using the new data from the Hipparcos mission. Are the proposed old moving groups a reality? We conclude that at least one old moving group exists and that it's possible to find other ones using velocity information in combination with metallicities based on Str\\\"omgren photometry. The study by Dehnen (1998) further supports our findings. He recovered many maxima, using a maximum likelihood solution, in the velocity distribution of nearby stars using the Hipparcos catalogue. Several of these are identifiable with known moving groups. In particular he found several maxima that exclusively contained red stars, indicating an old age. He identifies one of these maxima with the HR 1614 moving group. However, HR 1614 might be a rather special case. It is particularly metal-rich compared to the majority of stars in the part of the $UV$-plane it resides in. In other parts of the $UV$-plane (e.g. close to the local standard of rest) the group would have been completely obscured by other metal-rich stars. In summary, we conclude that at least one old moving group exists and it's possible to find others using our simple method if they stand out in terms of metallicity and/or age from the ambient background of stars in their space of the $UV$-plane." }, "0003/astro-ph0003151_arXiv.txt": { "abstract": "During more than 3 years (August 1996--October 1999) monitoring of a $40\\degr \\times 40\\degr$ sky region around the Galactic Centre by the Wide Field Cameras on board {\\em BeppoSAX}, a total of 8 type-I bursts have been detected from a sky position consistent with that of 4U~1812$-$12, a likely neutron-star low-mass X-ray binary. We present the results of a detailed study of the bursts of 4U~1812$-$12, about 15 years after the last reported observations of X-ray bursts from this source (\\cite{Mura83}). Clear evidence for photospheric radius expansion due to Eddington-limited burst luminosity is present in most of the observed events, allowing an accurate estimate of the source distance ($\\sim 4$ kpc) and its burst parameters. ", "introduction": "Since its first {\\em Uhuru} detections (\\cite{Form76,Form78}), 4U~1812$-$12 was observed by several satellite X-ray experiments: {\\em OSO~7} (1M~1812-121, \\cite{Mark79}), {\\em Ariel~V} (3A~1812-121, \\cite{Warw81}), {\\em HEAO}~1 (1H~1815$-$121, \\cite{Wood84}), and {\\em EXOSAT} (GPS~1812-120, \\cite{Warw88}). From these observations, it is clear that 4U~1812$-$12 is a persistent, though variable, source. {\\em Uhuru} found a 2--10 keV maximum intensity of $\\sim 5\\times 10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$ and a variability of a factor of at least 2 (\\cite{Form78}). Similar variability characteristics were observed by {\\em Ariel~V}, as the source varied in the range $\\sim 3-6\\times 10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$ in the same energy band (\\cite{Warw81}), while lower intensities of $\\sim 2\\times 10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$ and $\\sim 3\\times 10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$ were measured by {\\em HEAO}~1 (2--10 keV, \\cite{Wood84}) and {\\em EXOSAT} (2--6 keV, \\cite{Warw88}). The 3--10 keV source spectrum as obtained by the {\\em EXOSAT} GSPC was best fitted by a power law (\\cite{Gott95}). 4U~1812$-$12 is being monitored by {\\em RXTE}-ASM since February 1996 \\footnote{the ASM measurements are publicly available at URL {\\tt http://www.space.mit.edu/XTE}}, confirming its previously reported characteristics. The source is always detected, with an average 2--10 keV flux of $\\sim 3.8\\times 10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$ ($\\sim 20$ mCrab) and a variability of a factor $\\la 3$ on $\\sim 1$ week time scale. Three X-ray bursts were detected from this source in 1982 by {\\em Hakucho} (\\cite{Mura83}). Two of the events showed clear evidence for photospheric radius expansion, and reached a maximum 1--22 keV intensity of $\\sim 1.7\\times 10^{-7}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$. The burst spectra were consistent with a $\\sim 2.5$ keV blackbody emission, and showed evidence for softening during the exponential decay ({\\em e}-folding time $\\tau \\sim 20$ s). This indicated the bursts to be type-I, i.e. thermonuclear flashes originating on the hot surface of a neutron star, and the source to be likely located in a low-mass X-ray binary. During the observation no persistent emission was detected above 20 Uhuru flux units ($\\sim 5\\times 10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$ in 2-10 keV). Anyway this is not in disagreement with the identification of the {\\em Hakucho} burster with the persistent source 4U~1812$-$12. \\cite{Mura83} also proposed the association of the burst source with the transient Ser~X-2, observed once in 1965 (\\cite{Frie67}). More recently, a single-peaked burst from 4U~1812$-$12 was observed by the {\\em BeppoSAX}-WFC instrument on 1997 Mar 12.2209 UT (\\cite{Burd97}). The event had a peak intensity of 1.2 and 0.6 Crab in the 1.5--10 keV and 10$-$26 keV band respectively, and a decay time of $\\sim 20$ s. 4U~1812$-$12 is classified as an atoll source, which is common among the type-I X-ray bursters, and shows band-limited noise and a $\\sim 0.8$ Hz QPO (\\cite{Wijn99}). In this paper we investigate the burst characteristics of 4U~1812$-$12, about 15 years after the {\\em Hakucho} observations of its type-I bursts. The event observed by Burderi et al. (1997) is also re-analysed. Through a homogeneus sample of Eddington-limited type-I X-ray bursts, we accurately estimate the source distance and test the reliabilty of near-Eddington bursts as a standard candle. In the next section we briefly introduce the Wide Field Cameras telescopes and report on the observations of 4U~1812$-$12. Time-resolved spectroscopy of the bursts is presented in Section 3, while the scientific implications of our results are discussed in Section 4. \\begin{figure}[hbt] \\psfig{figure=4u_fig2.ps,width=8.3cm,clip=t} \\caption{ Time profiles of burst C in three energy bands. The double-peaked structure becomes more evident at higher energies. } \\label{Fig1} \\end{figure} \\begin{table*}[hbt] \\caption{Summary of the characteristics of the observed bursts. Average spectral parameters are calculated for the first 30 s of burst data. } \\protect\\label{t:ae} \\begin{flushleft} \\begin{tabular}{lcccccccc} \\hline \\hline \\noalign{\\smallskip} & burst A & burst B & burst C & burst D & burst E & burst F & burst G & burst H \\\\ \\hline \\noalign{\\smallskip} date & 1996 Aug.22 & 1996 Aug.29 & 1996 Sep.11 & 1997 Mar.12 & 1997 Mar.30 & 1998 Sep.30 & 1998 Oct.7 & 1999 Sep.22 \\\\ UT (h) & 16.9158 & 5.6625 & 7.3972 & 5.3021 & 21.2650 & 21.9400 & 1.8097 & 7.7229 \\\\ \\noalign{\\smallskip} $\\tau_{2-28~{\\rm keV}}$ & $15.0\\pm 1.3$ & $14.5\\pm 1.2$ & $15.1\\pm 0.5$ & $14.1\\pm 0.6$ & $13.9\\pm 1.0$ & $14.2\\pm 1.2$ & $10.6\\pm 1.1$ & $12.1\\pm 1.0$ \\\\ $\\tau_{2-8~{\\rm keV}}$ & $16.8\\pm 1.7$ & $18.0\\pm 1.4$ & $20.1\\pm 0.6$ & $16.4\\pm 0.8$ & $17.9\\pm 1.6$ & $18.0\\pm 1.7$ & $15.1\\pm 1.8$ & $19.0\\pm 2.1$ \\\\ $\\tau_{8-28~{\\rm keV}}$ & $ 7.3\\pm 1.2$ & $ 5.8\\pm 0.9$ & $ 6.9\\pm 0.4$ & $ 6.3\\pm 0.7$ & $ 4.1\\pm 0.6$ & $ 4.4\\pm 0.7$ & $ 6.8\\pm 1.1$ & $ 6.1\\pm 0.7$ \\\\ \\noalign{\\smallskip} {\\em k}T (keV) & $1.98^{+0.07}_{-0.06}$ & $1.85^{+0.04}_{-0.05}$ & $1.98^{+0.03}_{-0.02}$ & $1.96^{+0.03}_{-0.04}$ & $2.10\\pm 0.06$ & $1.99^{+0.06}_{-0.05}$ & $2.26\\pm 0.07$ & $2.33^{+0.08}_{-0.07}$ \\\\ ${\\rm R}_{bb}^{(a)}$ & $20.1^{+1.4}_{-1.3}$ & $21.8^{+1.3}_{-1.1}$ & $20.2^{+0.5}_{-0.4}$ & $20.9^{+0.8}_{-0.7}$ & $15.6\\pm 0.9$ & $19.5^{+1.1}_{-1.0}$ & $15.0^{+1.0}_{-0.9}$ & $13.6^{+0.9}_{-0.8}$ \\\\ $\\chi^{2}_{r}$ & 1.57 & 0.89 & 1.53 & 1.13 & 1.69 & 1.36 & 1.18 & 0.75 \\\\ \\noalign{\\smallskip} $I_{peak}^{(b)}$ & $4.69\\pm 0.32$ & $4.21\\pm 0.27$ & $4.36\\pm 0.14$ & $4.87\\pm 0.21$ & $4.66\\pm 0.28$ & $4.81\\pm 0.33$ & $4.32\\pm 0.37$ & $4.68\\pm 0.33$ \\\\ fluence$^{(c)}$ & $2.91\\pm 0.17$ & $2.81\\pm 0.14$ & $3.14\\pm 0.10$ & $2.93\\pm 0.12$ & $2.26\\pm 0.12$ & $3.02\\pm 0.16$ & $2.37\\pm 0.15$ & $2.25\\pm 0.13$ \\\\ \\noalign{\\smallskip} \\hline\\noalign{\\smallskip} \\noalign{$^{(a)}$ blackbody radius, for a 10 kpc distance; $^{(b)}$ peak intensity, in Crab units, 2-28 keV band; $^{(c)}$ bolometric, in $10^{-6}~{\\rm erg~cm}^{-2}$. The bursts {\\em e}-folding times $\\tau$ are given in s.} \\end{tabular} \\end{flushleft} \\end{table*} ", "conclusions": "Following the classification proposed by \\cite{Hoff78}, X-ray bursts are classified in two main types ({\\em type-I}, {\\em type-II}, see Lewin, van Paradijs \\& Taam (1993) for a comprehensive review). On the basis of the spectral and timing properties of the eight bursts observed by the WFCs, it is apparent that 4U~1812$-$12 is a type-I burster. In fact, the blackbody emission and the measured colour temperatures of $\\sim 2$ keV are consistent with type-I bursting. Moreover, spectral softening is observed in the time resolved spectra of the bursts (Table 2 and Figure 3), and the bursts time profiles can be fitted with exponential decays whose characteristic times are significantly shorter at higher energies (see e.g. Table 1). These results confirm the measurements obtained 15 years earlier by {\\em Hakucho} (\\cite{Mura83}), indicating 4U~1812$-$12 is a neutron-star low-mass X-ray binary. The photospheric radius expansion derived from the time resolved spectral analysis of most of the observed bursts can be interpreted as adiabatic expansion during a high luminosity (Eddington-limit) type-I burst. Actually, the double-peaked profiles observed in the high energy (above 8 keV) time histories of the bursts (with the only exception of burst H) are typical of super-Eddington events (e.g. \\cite{Lewi95}). Even though burst H is not double-peaked, its peak luminosity is consistent with those of the other observed events. Moreover, its 8--28 keV time profile could be associated with the flat-top profiles of some observed Eddington-limited type-I bursts (\\cite{Lewi95}), so burst H too can be regarded as an event with peak luminosity close to the Eddington limit. We also notice that the two bursts which show less clear evidence for photospheric radius expansion, namely bursts E and H, are the less energetic ones, their fluences being the lowest observed (Table 1). We can regard their total energy release ($\\sim 5\\times 10^{39}$ erg assuming a 4 kpc distance, see later) as the minimum needed to drive the expansion of the photosphere of the neutron star in 4U~1812$-$12. Eddington-luminosity X-ray bursts can lead to an estimate of the source distance, assuming the burst emission to be isotropic and the peak flux to be very close to the Eddington luminosity. Actually, the peak intensities observed for the eight events are all consistent with a constant value of $4.53 \\pm 0.09$ Crab (2--28 keV); the associated reduced $\\chi^2$ is 1.03 for 7 d.o.f.. This average peak intensity extrapolates to an unabsorbed bolometric flux of $(15.03 \\pm 0.29)\\times 10^{-8}~{\\rm erg~cm}^{-2}~{\\rm s}^{-1}$. The consistency of the peak luminosities of all the bursts with a constant value supports the adoption of the peak bolometric intensities of super-Eddington bursts as a standard candle. An average luminosity of Eddington-limited bursts was empirically calculated by Lewin, van Paradijs and Taam (1995) on a sample of bursters whose distance was estimated with other methods: a luminosity value of ($3.0\\pm 0.6\\times 10^{38}~{\\rm erg~s}^{-1}$) was obtained. The adoption of this standard luminosity leads to a distance value $d = 4.1\\pm 0.5$ kpc for 4U~1812$-$12. On the other hand, assuming the theoretical Eddington luminosity for a typical $1.4~{\\rm M}_{\\sun}$ neutron star ($\\sim 2\\times 10^{38}~{\\rm erg~s}^{-1}$) we obtain $d \\sim 3.3$ kpc. For the calculated distance of $\\sim 4$ kpc, and with the simple assumptions on the burst emission made in Section 3, an average radius of $8\\pm 1$ km for the blackbody emitting region during the bursts is obtained. This value, which supports the neutron-star nature of the collapsed object, should indeed be regarded as a lower limit for the actual neutron star radius, according to Ebisuzaki, Hanawa \\& Sugimoto (1984). Assuming the source's persistent spectrum to be consistent with the one suggested by Barret et al. (2000) for the X-ray bursters in low state, i.e. a Comptonized spectrum with electron temperature $k{\\rm T}_{e} \\sim 25$~keV and $\\tau \\sim 3$, the bolometric luminosity of 4U~1812$-$12 can be extrapolated. For a distance of $\\sim 4$ kpc and an average 2--10 keV persistent intensity of $\\sim 4\\times10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$, we obtain $L_{bol} \\sim 5.6\\times 10^{36}~{\\rm erg~s}^{-1}$. We also derive, for a canonical $1.4~{\\rm M}_{\\odot}$ neutron star with a radius of 10 km, an average accretion rate of $\\sim 5\\times10^{-10}~{\\rm M}_{\\odot}{\\rm y}^{-1}$. These values are common among low-mass X-ray binaries. Due to the non-continuous WFCs coverage of the Galactic Centre region and to the {\\em BeppoSAX} orbit characteristics, we cannot accurately establish the burst occurrence rate of 4U~1812$-$12. Anyway the minimum observed intervals are 6.53 d and 6.16 d for bursts A-B (August 1996) and F-G (October 1998) respectively, and such intervals are of the same order of magnitude than the one measured by {\\em Hakucho} in 1982 (4.61 d). Under the hypothesis that the above values are the actual wait times for bursts B and G, we can calculate the ratio $\\alpha = E_{p}/E_{b}$, where $E_{b}$ and $E_{p}$ are the bolometric fluences of the burst and of the persistent emission between two contiguous bursts, respectively. Average 2--10 keV intensities of $(4.4\\pm 0.4) \\times 10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$ and $(3.5 \\pm 0.9) \\times 10^{-10}~{\\rm erg~cm}^{-2}{\\rm s}^{-1}$ were measured by {\\em RXTE}-ASM between bursts A-B and F-G respectively. Again, with the former assumptions on the spectrum of the persistent emission, we determine $\\alpha = (6.4\\pm 1.0)\\times 10^{2}$ and $\\alpha = (5.8\\pm 1.9)\\times 10^{2}$ for the two events. These values are consistent with each other and are within the observed range ($10-10^{3}$, distribution peaking at $\\sim 10^{2}$) for the $\\alpha$ parameter of known X-ray bursters, even if on the higher side. This is suggestive of helium-burning with no spare fuel left for the next burst, and possibly of steady burning of part of the accreted matter ({\\cite{Lewi93, vanP88}). As pointed out above, the eight bursts we analysed show very similar features. Moreover, their characteristics are also consistent with those of the bursts detected by {\\em Hakucho} (\\cite{Mura83}). For the {\\em Hakucho} bursts, 1--22 keV peak intensities of $\\sim 4.6$ Crab were measured. Also the event Burderi et al. (1997) reported to have very different burst parameters, i.e. single peak profile and much lower luminosity (1.2 Crab in 1.5--10 keV), is actually very similar to the others, according to the results of our re-analysis (burst D). The above consistencies, together with the $\\sim 5-6$ d burst wait times observed by both {\\em Hakucho} and {\\em BeppoSAX}, suggest the burst characteristics of the binary 4U~1812$-$12 to be remarkably stable in observations spanning $\\sim 15$ years apart." }, "0003/astro-ph0003198_arXiv.txt": { "abstract": "We present preliminary source counts at $6.7\\mu$m and $15\\mu$m from the Preliminary Analysis of the European Large Area ISO survey, with limiting flux densities of $\\sim2$mJy at $15\\mu$m and $\\sim1$mJy at $6.7\\mu$m. We separate the stellar contribution from the extragalactic using identifications with APM sources made with the likelihood ratio technique. We quantify the completeness and reliability of our source extraction using (a) repeated observations over small areas, (b) cross-IDs with stars of known spectral type, (c) detections of the point spread function wings around bright sources, (d) comparison with independent algorithms. Flux calibration at $15\\mu$m was performed using stellar identifications; the calibration does not agree with the pre-flight estimates, probably due to effects of detector hysteresis and photometric aperture correction. The $6.7\\mu$m extragalactic counts are broadly reproduced in the Pearson \\& Rowan-Robinson model, but the Franceschini et al. (1997) model underpredicts the observed source density by $\\sim0.5-1$ dex, though the photometry at $6.7\\mu$m is still preliminary. At $15\\mu$m the extragalactic counts are in excellent agreement with the predictions of the Pearson \\& Rowan-Robinson (1996), Franceschini et al. (1994), Guiderdoni et al. (1997) and the evolving models of Xu et al. (1998), over $7$ orders of magnitude in $15\\mu$m flux density. The counts agree with other estimates from the ISOCAM instrument at overlapping flux densities (Elbaz et al. 1999), provided a consistent flux calibration is used. Luminosity evolution at a rate of $(1+z)^3$, incorporating mid-IR spectral features, provides a better fit to the $15\\mu$m differential counts than $(1+z)^4$ density evolution. No-evolution models are excluded, and implying that below around $10$ mJy at $15\\mu$m the source counts become dominated by an evolving cosmological population of dust-shrouded starbursts and/or active galaxies. ", "introduction": "\\label{sec:introduction} The IRAS mission enjoyed huge successes, including the sensational discoveries of ultra- and hyper-luminous galaxies and of an enormous population of evolving starbursts. However, the survey had several drawbacks. For example, the bright limiting flux densities restricted the samples to low redshifts ($z\\stackrel{<}{_\\sim}0.3$) for all but a few ultraluminous objects. Also, only $\\sim1000$ galaxies were detected at $12\\mu$m over the whole sky. These deficiencies restricted the study of IR-luminous galaxies at all redshifts. The Infrared Space Observatory (ISO) offered $\\sim\\times 1000$ improvements in sensitivity in the mid-IR over IRAS, and the large allocations of guaranteed and discretionary time for deep surveys on ISO will greatly improve on the IRAS surveys in the mid-IR. For instance, ISO observations of the northern Hubble Deep Field (Serjeant et al. 1997, 1999, Goldschimdt et al. 1997, Oliver et al. 1997, Aussel et al. 1999, D\\'{e}sert et al. 1999) reached the $15\\mu$m confusion limit ($\\sim0.1$ mJy) over $17$ square arcminutes, while the CAM-Deep and CAM-Shallow surveys (Elbaz et al. 1998a,b) were slightly less sensitive but had wider areal coverage ($0.5$ mJy over $0.3$ square degrees and $0.8$ mJy over $0.41$ square degrees). These have also been complemented by deep ISO photometry of selected high-$z$ galaxies (e.g. Flores et al. 1999). The European Large Area ISO Survey (ELAIS, Oliver et al. 1999 (paper I), Rowan-Robinson et al. 1998) was the largest open time project on ISO, complementing the deep ISO samples by surveying $\\sim12$ square degrees to a depth of $\\sim2$ mJy at $15\\mu$m and $\\stackrel{<}{_\\sim}100$ mJy at $90\\mu$m. Around half the area was also mapped at $6.7\\mu$m to $\\sim1$ mJy. Three fields in the Northern hemisphere (N1, N2, N3) collectively comprised around two-thirds of the $15\\mu$m areal coverage, with the remaining area taken by the Southern S1 field and several small areas in both hemispheres. The ambitious cosmological aims include tracing the extinguished star formation history of the Universe to $z\\sim1-2$, orientation-independent selection of dust-shrouded quasars, and the potential discovery of hyperluminous galaxies (with comparable intrinsic luminosities to IRAS FSC 10214+4724) out to redshifts $z\\stackrel{<}{_\\sim}5$. A more detailed discussion of the diverse scientific aims of ELAIS, the selection of areas and observational parameters can be found in the ELAIS survey paper (Oliver et al. 1999); in summary, the survey areas were selected to have low galactic cirrus emission, high visibility by ISO, high ecliptic latitude and avoiding $12\\mu$m IRAS sources brighter than $0.6$ Jy. In another companion paper, Efstathiou et al. 1999, we discuss the $90\\mu$m source counts from the Preliminary Analysis of the ELAIS ISOPHOT data, and in Crockett et al. (1999) we discuss the stellar mid-infrared source counts. The ELAIS areas have also been the subject of intensive multi-wavelength follow up, summarised to date in Oliver et al. (1999) and presented in detail in other papers (e.g. Ciliegi et al. 1999, Gruppioni et al. 1999). Here we present the completeness, reliability and extragalactic source counts from our initial Preliminary $6.7\\mu$m and $15\\mu$m ISOCAM catalogues. A future paper will present the Final Analysis products from the ISOCAM ELAIS data, which is expected to improve on the Preliminary Analysis presented here. This paper is structured as follows. In section \\ref{sec:method} we describe the Preliminary Analysis CAM pipeline, explaining the artefacts in the data (section \\ref{sec:method_intro}), and the pipeline algorithm (section \\ref{sec:cam_pipe}). The results from the Preliminary Analysis catalogue are presented in section \\ref{sec:results}. Our various completeness and reliability estimates are discussed in section \\ref{sec:comp_rel}, and the segregation of extragalactic from stellar sources in section \\ref{sec:optid}. Section \\ref{sec:counts} presents the source counts in both wavebands. These results are compared with source count models and previous results in section \\ref{sec:discussion}, where we also discuss the implications for the evolution of star forming galaxies and on the star formation history of the Universe. ", "conclusions": "\\label{sec:conclusions} The extragalactic source counts agree extremely well with all evolving model predictions (Franceschini et al. 1994, Pearson \\& Rowan-Robinson 1996, Guiderdoni et al. 1997, Xu et al. 1998) over seven orders of magnitude in $15\\mu$m flux density. The Pearson \\& Rowan-Robinson (1996) models can broadly reproduce the $6.7\\mu$m extragalactic counts, but the observations are in excess of the Franceschini et al. (1997) predicted counts at this wavelength using our preliminary $6.7\\mu$m flux calibration. All no-evolution models are clearly excluded, and imply a cosmologically evolving population of obscured starbursts and/or active galaxies dominates below $\\sim10$ mJy at $15\\mu$m, independent of K-correction assumptions. Source confusion appears to have been underestimated in the WIRE and SIRTF missions, but will not impact significantly on the NGST." }, "0003/astro-ph0003420_arXiv.txt": { "abstract": "The study of the statistical distribution of short quiescent times ($\\le$ 30 minutes) between solar flares has been investigated on a 1D MHD model. A power law behaviour is found which indicates the existence of sympathetic flaring. This prediction is supported by recent observations. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003085_arXiv.txt": { "abstract": "We present interferometric measurements of the Sunyaev-Zel'dovich (SZ) effect towards the galaxy cluster Abell 370. These measurements, which directly probe the pressure of the cluster's gas, show the gas distribution to be strongly aspherical, as do the x-ray and gravitational lensing observations. We calculate the cluster's gas mass fraction in two ways. We first compare the gas mass derived from the SZ measurements to the lensing-derived gravitational mass near the critical lensing radius. We also calculate the gas mass fraction from the SZ data by deprojecting the three-dimensional gas density distribution and deriving the total mass under the assumption that the gas is in hydrostatic equilibrium (HSE). We test the assumptions in the HSE method by comparing the total cluster mass implied by the two methods and find that they agree within the errors of the measurement. We discuss the possible systematic errors in the gas mass fraction measurement and the constraints it places on the matter density parameter, \\omegam. ", "introduction": "Clusters of galaxies, by virtue of being the largest known virialized objects, are important probes of large scale structure and can be used to test cosmological models. Rich clusters are extremely massive, $\\sim$$\\ 10^{15} M_{\\odot}$, as indicated by the presence of strongly gravitationally lensed background galaxies and by the deep gravitational potential necessary to explain both the large velocity dispersion ($>$ 1000 km s$^{-1}$) in the member galaxies and the high measured temperature ($> 5$ keV) of the ionized intracluster gas. Dynamical mechanisms for segregating baryonic matter from dark matter on these mass scales are difficult to reconcile with observations and standard cosmological models, and so within the virial radius the mass composition of clusters is expected to reflect the universal mass composition. Under the fair sample hypothesis, a cluster's gas mass fraction, which is a lower limit to the its baryonic mass fraction, is then a lower limit to the universal baryon fraction, i.e., $f_{gas} \\le f_B$. The luminous baryonic content of galaxy clusters is mainly contained in the gaseous intracluster medium (ICM). The gas mass is nearly an order of magnitude larger than the mass in optically observed galaxies (\\eg, White \\etal\\ 1993, Forman \\& Jones 1982). Hence, the gas mass is not only a lower limit to the cluster's baryonic mass, it is a reasonable estimate of it. The intracluster medium has largely been studied through observations of its x-ray emission. The ICM is hot, with electron temperatures, $T_e$, from $\\sim$5 to 15 keV; rarefied, with peak electron number densities of $n_e \\simeq 10^{-3}$ ${\\rm cm}^{-3}$; and cools slowly ($t_{cool}> t_{Hubble}$), mainly via thermal Bremsstrahlung in the x-ray band. The x-ray surface brightness is proportional to the emission measure, $S_x \\propto \\int n_e^2 \\Lambda(T_e) dl$, where the integration is along the line of sight, and so, under simplifying assumptions, the gas mass can be calculated from an x-ray image deprojection and the measured gas temperature. Since the sound crossing time of the cluster gas is much less than the dynamical time, one may reasonably assume that, in the absence of a recent merger, the cluster gas is relaxed in the cluster's potential. The total binding mass can be extracted from the gas density and temperature distribution under this assumption. A significant body of work exists in which the gas mass fraction, $f_g$, is measured in this way, with $f_g$ measurements out to radii of 1 Mpc or more (White \\& Fabian 1995; David \\etal\\ 1995; Neumann \\& Bohringer, 1997; Squires \\etal\\ 1997; Mohr, Mathiesen \\& Evrard 1999). The mean cluster gas mass fraction within approximately the virial radius was calculated in Mohr \\etal\\ (1999) to be (0.0749 $\\pm 0.0005) h^{-3/2}$. Here, and throughout the paper, we assume the value of the Hubble constant to be $H_\\circ=100 h$ km s$^{-1}$ Mpc$^{-1}$. To derive the gas mass fraction from x-ray imaging data, one is required to deproject the surface brightness into a model for the density distribution. As the x-ray emission is proportional to the square of the gas density, the gas mass measurement can be biased by clumped, multi-phase gas, should it be present. Also, the emission from the cores of relaxed clusters may be dominated by cooling flows, which complicate the interpretation of the x-ray data and may bias the result strongly if not taken into account (Allen 1998; Mohr \\etal\\ 1999). In addition, the x-ray surface brightness is diminished in proportion to its distance; $S_x \\propto 1/(1+z)^4$, where $z$ is the redshift of the cluster, neglecting experiment-specific K-corrections, and so it becomes increasingly difficult to make sensitive x-ray measurements of the ICM as the cluster redshift increases. We present a scheme for measuring the gas mass fractions with the Sunyaev-Zel'dovich effect which is different from the x-ray method in a number of ways, and also provides an independent measurement of $f_g$. The Sunyaev-Zel'dovich (SZ) effect is a spectral distortion of Cosmic Microwave Background (CMB) radiation due to scattering of CMB photons by hot plasma (Sunyaev \\& Zel'dovich 1970). The SZ effect can be detected significantly in galaxy clusters, where the ionized intracluster gas serves as the scattering medium. A small fraction, $\\leq 1\\%$, of CMB photons are inverse-Compton scattered and, on average, gain energy. At frequencies less than about 218 GHz, the intensity of the CMB radiation is diminished as compared to the unscattered CMB, and the SZ effect is manifested as a brightness temperature decrement towards the cluster. This decrement, $\\Delta T_{SZ}$, has a magnitude proportional to the total number of scatterers, weighted by their associated temperature, ${\\Delta T_{SZ} \\over T_{CMB}} \\propto \\int n_e T_e dl$, where $n_e$ is the number density of electrons, $T_e$ is the electron temperature, $T_{CMB}$ is the temperature of the CMB, and the integration is again along the line of sight. Note that this is simply proportional to the integrated electron pressure. Also, the magnitude of the SZ decrement is independent of redshift, so as the long as the cluster is resolved (the experiment's characteristic beamsize is not larger than the angle subtended by the cluster), the SZ effect can be measured towards arbitrarily distant clusters. A cluster's gas mass is directly proportional to its integrated SZ effect if the gas is isothermal. So under the isothermality condition, an image deprojection is not strictly required to obtain the gas mass. The cluster's gas mass fraction can be calculated by comparing the integrated SZ decrement, in effect a surface gas mass, to the total cluster mass in the same volume. The total mass can be measured with strong or weak gravitational lensing, for example. The SZ images may also be deprojected to infer the three-dimensional gas mass and the HSE mass. Since the SZ decrement is directly propotional to the electron density, the SZ image deprojection will not be affected strongly by clumped gas. Thus, the cluster's gas mass fraction can then be measured as a function of cluster radius as well. Recent cluster gas mass fraction measurements from SZ effect observations are presented in Myers \\etal\\ (1997). In this work, the integrated SZ effect is measured using a single radio dish operating at centimeter wavelengths. The integrated SZ effect is used to normalize a model for the gas density from published x-ray analyses, and this gas mass is compared to the published total masses to determine the gas mass fraction. For three nearby clusters, A2142, A2256 and the Coma cluster, Myers \\etal\\ find a gas mass fraction of $(0.061\\pm0.011) h^{-1}$ at radii of 1-1.5 $h^{-1}$ Mpc; for the cluster Abell 478, they report a gas mass fraction of $(0.16\\pm0.014)h^{-1}$. In this work, we describe a method to calculate cluster gas mass fractions from interferometric SZ observations. Here, the shape parameters are derived directly from the SZ dataset rather than from an x-ray image. We apply this method to the SZ effect measurements towards the cluster Abell 370, which were made as part of an SZ survey of distant clusters, the first results of which were reported in Carlstrom, Joy \\& Grego (1996), and are further reported in Carlstrom \\etal\\ (1997). We choose this primarily because it has been studied at optical and x-ray wavelengths, and so allows a comparison of the SZ data with other observations. Upon detailed investigation, it is apparent that this cluster is one of the most difficult in our sample to analyze, with significant ellipticity and complicated optical and x-ray structure; as such, it serves as a test of the gas mass fraction analysis method, which we plan to use on a large sample of clusters. With the interferometric SZ measurements, complemented by observations at other wavelengths, we measure the cluster's gas mass fraction in two ways. First, we calculate the surface gas mass from the SZ measurements and the surface total mass from strong gravitational lensing observations and models. Second, we measure the gas mass fraction in a manner similar to x-ray analyses. The gas mass is inferred from a deprojected model; the total mass is determined from the spatial distribution of the gas under the isothermal hydrostatic equilibrium (HSE) assumption. We test the assumptions made in the HSE analysis by comparing the total cluster masses derived in the two methods. The optical and x-ray observations of this cluster are discussed in Section~\\ref{sec:optxray}, and the SZ observations are discussed in Section~\\ref{sec:szobs}. The method for modeling the SZ data is presented in Section~\\ref{sec:model}, and the gas mass fraction results and the systematic uncertainties are discussed in Section~\\ref{sec:analysis}. The cosmological implications of the results and plans for future work are discussed in Section~\\ref{sec:discussion}. ", "conclusions": "\\label{sec:discussion} \\subsection{Gas Mass Fractions at $r_{500}$} To compare the $f_g$ we have measured within a fixed angular radius to $f_g$ measurements in clusters with different sizes and redshift, we extrapolate our measured $f_g$ to a fiducial radius. The ICM in nearby clusters has been observed to be distributed more uniformly than the dark matter (\\eg, David \\etal\\ 1995), which is to be expected if energy has been added to the intracluster medium before collapse or from galactic winds. If this is generally true, the gas mass fraction measured depends on the radius within which the measurement is made. As suggested in Evrard (1997) and Metzler, Evrard, \\& Navarro (1998), we choose this radius to be that within which the average density of the cluster is 500 times the critical density, $\\rho_c = {{3 H^2} \\over {8 \\pi G}}$. These numerical simulations suggest that within this radius, $r_{500}$, the cluster's baryon fraction should closely reflect the universal baryon fraction, if the current physical models of hierarchical structure formation are correct. We use the analytical expression of Evrard (1997) which describes the expected variation of $f_g$ with overdensity. This variation is found to be consistent with the $f_g$ variation reported in the David \\etal\\ (1995) sample. \\begin{equation}{f_{g}(r_{500}(T_e)) = f_{g}(r_X)\\left ({r_{500}(T_e) \\over r_X} \\right )^{\\eta}},\\label{eq:fextrap} \\end{equation} where $\\eta$ = 0.17, $f_{g}(r_{500}(T_e))$ is the gas mass fraction at $r_{500}$, and $r_X$ is the radius within which the gas mass fraction is measured. We modify Evrard's expression for $r_{500}$, derived for low redshift clusters, to include the change in the value of $\\rho_c$ with redshift; $\\rho_c(z) = \\rho_c(z=0)(H/H_\\circ)^2$, where $H^2=H_\\circ^2[(1+z)^3\\Omega_M+(1+z)^2(1-\\Omega_M-\\Omega_\\Lambda) + \\Omega_\\Lambda]$ \\begin{equation}{r_{500}(T_e) = (1.24 \\pm 0.09) \\left ( {T_e \\over 10\\ {\\rm keV} (H/H_\\circ)^2} \\right ) ^{1/2} h^{-1} {\\rm Mpc}}.\\label{eq:r500} \\end{equation} The gas mass fraction values at $r_{500}$, estimated from those measured at 65$''$, are summarized in Table 1. This experiment best measures $f_g$ at a given angular scale, which corresponds in Abell 370 to an overdensity of $\\sim 5000\\rho_c$. This is not the optimal radius at which to compare with numerical simulations, since resolution is limited in the cores of the clusters, and the gas in the core may also be sensitive to additional physics not yet included in the models, \\eg, magnetic fields and cooling. For these reasons, the corrections should be taken with some caution. \\subsection{Constraints on \\omegam\\ from $f_g$} Under the fair sample hypothesis, A370's gas mass fraction within $r_{500}$, a lower limit to the cluster baryon fraction, should reflect the universal baryon fraction: \\begin{equation} f_g \\le f_B = {\\Omega_B \\over \\Omega_M},\\end{equation} where $f_B$ is the cluster's baryon fraction, \\omegam\\ is the ratio of the total mass density to the critical mass density, and $\\Omega_B$ is the ratio of baryon mass density in the universe to the critical mass density. The cluster gas mass fraction measurements can then be used within the Big Bang Nucleosynthesis (BBN) paradigm to constrain \\omegam: \\begin{equation}\\Omega_M \\le \\Omega_B/f_g.\\end{equation} The value of $\\Omega_B$ is constrained by BBN calculations and the measurements of light element abundances. The relative abundance of deuterium and hydrogen provides a particularly strong constraint on the baryonic matter density. A firm upper limit to $\\Omega_B$ is set by the presence of deuterium in the local interstellar medium. This constrains the value of $\\Omega_B$ to be less than $0.031 h^{-2}$ (Linsky \\etal\\ 1995). Measurements of the D/H ratio in metal-poor Lyman-$\\alpha$ absorption line systems in high-redshift quasars put a tighter constraint on the baryonic mass density. Such measurements made by Burles \\& Tytler (1998) predict a value of $\\Omega_B = (0.019\\pm 0.002)h^{-2}$ at 95\\% confidence. The gas mass fractions measured for A370 from both lensing and HSE methods range from $5-13\\%h^{-1}$. We consider the simplest $f_g$ measurement, that in the spherical model, and compare this gas mass fraction at $r_{500}$ to the Burles \\& Tytler (1998) value for $\\Omega_B$. This gives an upper limit to the matter density parameter, $\\Omega_M \\le 0.19^{+0.10}_{-0.10}h^{-1}$, at 68\\% confidence. However, the bimodal model gives a surface gas mass fraction at angular radius 40$''$ of 0.048$h^{-1}$, a value which permits $\\Omega_M$ to be as high as 0.40$h^{-1}$ in this scheme. (We note again that the dependence of $f_g$ through the angular diameter distance is weak at this redshift, with a change in $f_g$ of 5-10\\% when a wide range of cosmological parameters is used.) These values are consistent with the limits on \\omegam\\ from observations of supernovae, which are derived from geometrical arguments, rather than the $\\Omega_B/f_g$ ratio. Depending on the method used to calibrate the sample, for a spatially flat universe, Garnavich \\etal\\ (1998) find \\omegam\\ $< 0.4-0.5$ at 68\\% confidence. \\subsection{Conclusion \\& Future Work} We have measured the Sunyaev-Zel'dovich effect in the galaxy cluster Abell 370 and present spatially filtered images from these data. The optical and x-ray observations of this cluster show a complicated and perhaps bimodal mass distribution. The SZ effect image, however, looks smoothly distributed and significantly aspherical. We have fit both one-component and two-component \\be-models to the data and find that the two-component model does not fit significantly better. For both models, we calculate the gas mass fraction for the cluster using measurements of the total cluster mass from both the gravitational lens model (the ``surface'' gas mass fraction) and from the hydrostatic equilibrium assumption. When integrated in the same volume, the HSE masses are consistent with the mass derived from the gravitational lensing model for both the one- and two-component models, lending support to the HSE assumption. The surface gas mass fraction measurement is made within an angular radius of 40$''$ and the HSE gas mass fraction is made within a radius of 65$''$. The gas mass fraction near the virial radius is derived from the gas mass fractions at 65$''$ using a correction factor derived from numerical simulations. For the range of methods and models used, we find gas mass fraction values of $\\sim (5-13)h^{-1}\\%$. Constraints on the Hubble parameter \\Ho\\ can in principle be derived from the SZ and x-ray measurements of a cluster. The SZ and x-ray observables depend on different moments of the electron density, and so the characteristic length scale of the cluster along the line of sight can be measured and the angular diameter distance of the cluster inferred. This is useful not only as a distance ladder-independent measurement of \\Ho, but, when compared with other \\Ho\\ measurements, can be used to explore possible systematic effects in the $f_g$ calculation, \\eg, to constrain the deprojection of the gas distribution. However, the quality of the x-ray imaging data in this case and the apparent disagreement between the SZ and x-ray fitted models do not permit putting a strong constraint on the Hubble constant. A long observation with the Chandra x-ray observatory towards this cluster is planned, and should help resolve these issues. The value of the baryonic mass fraction in any one cluster will be susceptible to systematic uncertainties which may be difficult to estimate. In order to use cluster gas mass fractions as a cosmological tool, one wants to ameliorate the effect of these errors by studying a large sample of clusters. A SZ effect survey in galaxy clusters is being carried out by this group at the BIMA and OVRO observatories, and an analysis of the gas mass fraction of the sample is in preparation. Many thanks are due to the staff at the BIMA and OVRO observatories for their contributions to this project, especially Rick Forster, John Lugten, Steve Padin, Dick Plambeck, Steve Scott, and Dave Woody. Many thanks to Cheryl Alexander for her work on the system hardware. Thanks also to Jack Hughes and Doris Neumann for valuable discussions concerning the x-ray analysis, to Jean-Paul Kneib concerning his lensing analysis, and to Naomi Ota and collaborators for sharing their reduced ASCA data. This work is supported by NASA LTSA grant NAG5-7986. LG, EDR, and SKP gratefully thank the NASA GSRP program for its support. Radio astronomy with the OVRO and BIMA millimeter arrays is supported by NSF grant AST 93-14079 and AST 96-13998, respectively. The funds for the additional hardware for the SZ experiment were from a NASA CDDF grant, a NSF-YI Award, and the David and Lucile Packard Foundation. \\newpage" }, "0003/astro-ph0003200_arXiv.txt": { "abstract": "The redshifted $1420\\, \\rm MHz$ emission from the HI in unresolved damped Lyman-$\\alpha$ clouds at high z will appear as a background radiation in low frequency radio observations. This holds the possibility of a new tool for studying the universe at high-$z$, using the mean brightness temperature to probe the HI content and its fluctuations to probe the power spectrum. Existing estimates of the HI density at $z \\sim 3$ imply a mean brightness temperature of $1 \\, {\\rm mK}$ at $320 \\, {\\rm MHz}$. The cross-correlation between the temperature fluctuations across different frequencies and sight lines is predicted to vary from $10^{-7} \\, \\rm K^2$ to $10^{-8} \\, \\rm K^2$ over intervals corresponding to spatial scales from $10 \\, {\\rm Mpc}$ to $40 \\, {\\rm Mpc}$ for some of the currently favoured cosmological models. Comparing this with the expected sensitivity of the GMRT, we find that this can be detected with $\\sim 10 \\, \\rm hrs$ of integration, provided we can distinguish it from the galactic and extragalactic foregrounds which will swamp this signal. We discuss a strategy based on the very distinct spectral properties of the foregrounds as against the HI emission, possibly allowing the removal of the foregrounds from the observed maps. ", "introduction": "The problem of determining the distribution of matter on large scales in the universe and understanding the large scale structure (LSS) formation is of prime importance in modern cosmology. Observing the angular positions and redshifts of galaxies has been the most straightforward method of mapping the LSS in the present day universe (Peebles 1993, Peebles 1980), but the interpretation of these observations is complicated by the fact that the relation between the distribution of galaxies and the distribution of underlying matter is not fully understood (Bardeen \\etal 1986). Other direct methods use galaxy clusters or super-clusters as tracers to map the large scale structures. A large variety of techniques have been developed and applied to quantify the distribution of galaxies, and amongst these the correlation functions (two-, three-point correlation functions, etc) and their Fourier counterparts (power spectrum, bi-spectrum, etc.) have been most popular (Peebles 1980). Much of the work comparing the observations with different theories has been based on these. An alternative approach is to use observations of fluctuations in the diffuse extragalactic background radiation at different wavelengths to probe the large scale structure. Here, the observations of anisotropies in the cosmic microwave background radiation (CMBR) have been most important. These observations probe the clustering of matter at the last scattering surface (e.g. Bond 1996), and combined with the information from the galaxy surveys, they have been successful in narrowing down the allowed class of theoretical models of LSS formation (e.g., Peacock 1999 and reference therein). The study of diffuse background at other wavelengths has been used to infer the clustering properties of matter at more recent epochs. Shectman (1974) observed fluctuations in the optical background; the results of this observation are consistent with predictions from galaxy counts (Peebles 1980). This method has also been applied to far-infrared background (Kashlinsky {\\it et al. \\/} 1997) and recently the first detection of fluctuations in this background has been reported (Kashlinsky {\\it et al. \\/} 1999). Gorjian, Wright and Chary (1999) have recently reported a tentative detection of a near-infrared background. There are similar predictions for fluctuations in the X-ray background (Barcons \\etal 2000, Barcons \\etal 1998). In this paper we investigate the possibility of using the extragalactic background radiation at low radio frequencies (meter wave) to probe the LSS. This is largely motivated by the fact that the Giant Meter-wave Radio Telescope (GMRT; Swarup \\etal 1991) which is designed to observe in several frequency bands in the interval $150 \\,{\\rm MHz}$ to $1420 \\, {\\rm MHz}$ has recently started functioning. This frequency range corresponds to the $1420 {\\rm MHz}$ emission from HI in the redshift range $0 \\le z \\le 8.5$. Absorption studies along sight lines to quasars indicate that most of the HI in the redshift range $0 \\le z \\le 3$ is in damped Lyman-$\\alpha$ (DLA) clouds and the density of HI in DLAs has been determined reasonably well from absorption studies (Lanzetta, Wolfe, \\& Turnshek 1995). Although the HI emission from individual DLAs at high $z$ is too faint to detect using presently available telescopes, the redshifted HI emission from unresolved DLAs will contribute to the background radiation at low frequencies. In this paper we investigate how the brightness temperature of this radiation is related to the density and peculiar velocity of the HI. We consider the possibility of detecting correlations in the fluctuations in this component of the background radiation and using this to probe LSS at high $z$. The possibility of observing the redshifted 21 cm emission from HI at high $z$ has been discussed earlier by many authors in a large variety of contexts. One of the first papers (Sunyaev \\& Zel'dovich, 1975) considers the possibility of meter-wave observations to detect protogalaxies and protoclusters at $z \\le 10$. There have been several attempts to detect the HI in proto-clusters and proto-super clusters (see Subrahmanyan \\& Anantharamaiah 1990 and reference therein). More recently Subramanian and Padmanabhan (1993) have calculated the abundance of protoclusters as a function of the redshifted HI flux density for various redshifts for both the CDM and HDM models. In a later paper Kumar, Padmanabhan and Subramanian (1995) have calculated the line profile of the HI emission from a spherically symmetric protocluster. Bagla, Nath and Padmanabhan (1997) and Bagla (1998) have used a combination of N-body simulations and a model for the behaviour of the baryons to calculate the abundance and the expected flux from the HI in structures like protoclusters at high redshifts. A major uncertainty in all of these works is in the assumptions about the HI content of the universe at high redshifts. The main focus of all of these works has been on individual peaks of the density fluctuations (protoclusters) which will manifest themselves as {% detectable features in low-frequency radio maps. Subramanian and Padmanabhan (1993) have also studied the possibility of detecting the excess variance in radio images due to the density fluctuations in the HI at high $z$. Katz, Weinberg and Hernquist (1998) have used smoothed particle hydrodynamic simulations to study the distribution of HI gas at high $z$ and they consider the possibility of detecting HI emission from galaxies at $z>2$. Tozzi \\etal (1999) and Shaver \\etal (1999) have studied the possibility of detecting the HI in the IGM at $z>5$. The state of the HI at these redshifts is unknown and these works are based on different scenarios for the reionization of the universe. The work presented here differs significantly from the previous papers in that: \\begin{itemize} \\item[(1)] It is restricted to $z<3.5$ where the HI content of the universe is well determined from DLA absorption studies. We use the results of these observations as inputs to our calculations. \\item[(2)] Instead of looking at the possibility of detecting individual features (as has been the focus of a large number of previous papers) we have studied the statistical properties of the fluctuations in the brightness temperature in low frequency radio maps. The statistical quantity we have studied is the cross-correlation between the temperature fluctuations along different lines of sight in radio map made at different frequencies. Individual features corresponding to protoclusters are rare events and protoclusters with flux in the range $1.5-3 \\, {\\rm m \\, Jy}$ are predicted to occur with abundances in the range $10^{-8} - 10^{-7} \\, {\\rm Mpc}^{-3}$ in the CDM model (Subramanian and Padmanabhan, 1993). Even small density fluctuations will contribute to the cross-correlation signal and our proposal has the advantage of simultaneously using the signal in all the pixels in all the frequency channels across the bandwidth of observation. \\end{itemize} The structure of the paper is as follows: in \\S 2 we discuss the basic formalism of calculating the brightness temperature and fluctuations in the redshifted 21 cm radiation, in \\S 3 we discuss the numerical results for two currently favoured cosmological models. In \\S 4, the observational difficulties in the presence galactic and extragalactic foregrounds are presented, and we discuss a possible strategy for overcoming these. \\S 5 gives a summary of our main results. ", "conclusions": "We have investigated the contribution from the HI in unresolved damped Lyman-$\\alpha$ clouds at high redshifts to the background radiation at low frequency radio waves (meter waves). The isotropic part of this radiation depends on the density of HI and the background cosmological model, while the fluctuations in this component of the background radiation have an added dependence on the fluctuations in the distribution of the damped Lyman-$\\alpha$ clouds and their peculiar velocities. We have used estimates of the HI density available from absorption studies to calculate the brightness temperature of this radiation. We find that this has a value $\\sim 1 \\, {\\rm mK}$ at $320 \\, {\\rm MHz} $ which corresponds to $z \\sim 3$. The distribution of damped Lyman-$\\alpha$ clouds is assumed to trace the underlying dark matter distribution which also determines the peculiar velocities. Using this and the linear theory of density perturbations, we have calculated the relation between the fluctuation in this component of the background radiation and the density perturbations at high $z$. Observations of the cross-correlations of the fluctuations at different sight lines across images produced at different frequencies holds the possibility of allowing us to probe the two point correlation function (or power spectrum) at high redshifts. We have calculated the expected cross-correlations for two currently acceptable CDM models and find it to be in the range $10^{-7} {\\rm K^2}$ to $10^{-8} {\\rm K^2}$ at $\\nu \\sim 320 {\\rm MHz}$ for separations in sight lines and frequencies such that the corresponding spatial separation is in the range $10 \\, {\\rm Mpc}$ and $40 \\, {\\rm Mpc}$. The cross-correlations are expected to be larger at smaller scales where the linear theory cannot be applied. Our results show that both the isotropic background (Figure~\\ref{fig:2}) and its fluctuations (Figure~\\ref{fig:3} and ~\\ref{fig:4}) can be detected by GMRT which is the largest telescope operating at meter waves at present, provided this signal can be distinguished from other sources which contribute to the low frequency background radiation. The biggest obstacles in detecting the HI contribution are the galactic and extra-galactic foregrounds, both of which are many orders larger than the signal we want to detect. The fact that both those sources of contamination emit continuum radiation while the HI contribution is from individual damped Lyman-$\\alpha$ clouds each of which emits a spectral line with a relatively small velocity width keeps alive the possibility of being able to distinguish this signal from the contamination. We have, in this paper, considered one possible approach which might allow us to model and subtract the foreground along any line of sight. More work is needed in this direction and work is currently underway in investigating other viable possibilities for foreground removal." }, "0003/astro-ph0003036_arXiv.txt": { "abstract": "I show here that the time--delay between the gamma ray burst proper and the onset of the afterglow is a sensitive function of the surrounding density, going from $\\approx 10\\; s$ for typical disk densities to several hours for IGM densities. Since bulk Lorenz factor, explosion energy and environment density can be simultaneously determined from observations of the time--delay and of the afterglow, afterglow observations alone may establish which fraction, if any, of all bursts resides inside galaxies, in their outer haloes, or outright in the IGM medium, subjecting the NS/NS merger hypothesis to a definitive test. If bursts are due to collapse of massive stars, no matter how short the burst, the onset of the afterglow should be immediate. ", "introduction": "There are currently three favored models for the formation of GRBs: mergers of neutron star binaries (Narayan, Paczynski and Piran 1992), collapsars (or hypernovae, Woosley 1993, Paczynski 1998), and SupraNovae (Vietri and Stella 1998, 1999), but a clearcut test to which each of these could be subjected is still missing. This is so because the details of the fireball model are to a large extent independent of the details of the energy deposition mechanism. The situation has also a paradoxical side: even if we managed to observe the event that deposits the energy, the enormous expected optical depths (exceeding $10^{10}$) would prevent us from deriving any information on the nature of the source. Gravitational waves can probably provide a signature of at least some of the scenarios, but detection of these from these distant sources does not look like an immediate prospect. For these reasons, many authors (B\\\"ottcher \\etal 1999, M\\'esz\\'aros and Rees 1998) have proposed that interaction of the burst with surrounding material might provide detectable signatures of star--forming environments. However, the possible discovery of iron emission lines (Piro \\etal, 1999, Yoshida \\etal, 1999) from two bursts has highlighted the weak side of this possibility, \\ie, that any conclusion is subject to the usual uncertainties associated with modeling complex and incomplete sets of observations (Lazzati \\etal, 1999, Vietri \\etal, 1999). I show in this paper that a surprisingly simple and robust test exists, for both the neutron binary merger and the collapsar scenarios, resulting from a trivial, but so far overlooked property of fireballs. ", "conclusions": "The relevance of the above arguments to the problem of the identification of the bursts' progenitors is clear. Bloom, Sigurdsson and Pols (1999) have shown that there is a well--defined distribution of expected distances of GRBs from their parent galaxies, if they originate from the merger of neutron stars. In particular, they showed that $\\approx 15\\%$ of all burst progenitors may be able to escape from their parent galaxy altogether, and that about $50\\%$ of all bursts will be located more than $8\\; kpc$ from their site of origin, nearly in the galactic halo. This feature is potentially verifiable also with optical investigations, but we know that even the most powerful ground--based telescopes are often inadequate to find the host galaxy, while HST searches, which up to now have scored a remarkable $5/6$ success rate (Fruchter 2000), are slow, cumbersome, and, for some of the most distant bursts, cannot go deep enough in the luminosity function. The test I propose is instead simple, and a statistically meaningful database will be secured in the near future, since several space missions (AXAF, Hete II, Integral, XMM, SWIFT) will observe hundreds of bursts for several hours after the burst. Especially important among them is SWIFT, which will be responsible for most of these observations, and will provide continuous coverage of each detected bursts in both X--ray and UV/optical wavebands. From this database of perhaps $300$ bursts observed continuously over several hours after the burst, we may expect $\\approx 50$ bursts located in the IGM medium, and another $\\approx 100$ located in the haloes of their parent galaxies. Thus, within this scenario we expect about a third of all bursts to have delays of several minutes, and a few tens to have delays of a few hours. The flux should then peak again to a factor $\\approx t_{b}/\\xi t_d$ below that of the burst proper. The very same test can also be used {\\it a converso} to check whether bursts are surrounded by the massive winds expected in all scenarios involving massive stars. In particular, it is expected that no time--delay will ever be observed, except perhaps for the shortest bursts, $t_{b} \\ll 0.03\\;s$. One may wonder what happens when only incomplete data are available, \\ie, when the burst disappears shortly after $t_b$ and is re--observed when already in the afterglow regime. This is the case of BeppoSAX which, because of instrumentational limitations, is incapable of providing significant information on the key silent period described here. It should be notice that the above estimates for the luminosity ($L_b$ at $t_b$, and $L_b t_d/\\xi t_d$ at $2 t_d$) fall precisely on a $t^{-1}$ slope. So, if further observations of the afterglow find an afterglow slope close to $-1$, extrapolation of the afterglow luminosity to early time will match closely with the burst luminosity, regardless of the existence of the silent period. Thus, when only incomplete data are available, only bursts with slopes significantly different from $-1$ can be helpful in deciding whether a silent period exists. A corollary of the above tests comes from the observations of flares in the radio lightcurves. In fact, these flares are expected only from afterglow sources with radii smaller than $R_c$ (Goodman 1997), where \\begin{equation} R_c = 10^{17}\\; cm \\frac{\\nu_{10}^{6/5}}{d_{sc,kpc}h_{75}} \\left(\\frac{SM}{10^{-2.5} m^{-20/3} kpc}\\right)^{-3/5}\\;. \\end{equation} Here $\\nu_{10}$ is the observing frequency in units of $10\\; GHz$, $d_{sc,kpc}$ the distance of the Galactic scattering medium, assumed to be a simple screen, in units of $1\\; kpc$, and $SM$ is the Galactic scattering measure, scaled in terms of a typical Galactic value. It can be seen from the above that $R_c \\ll R_{ag}$ every time that the burst is not located inside a galactic disk. Thus, for the $\\approx 150$ sources mentioned above, which are expected to show measurable time--delays, we expect {\\it a fortiori} no radio flares. Radio flares are instead allowed if bursts originate from collapsars. In summary, I have shown here that the X--ray luminosity has a silent period between the burst and the power--law afterglow, of duration \\begin{equation} t_d = \\left\\{ \\begin{array}{ll} 15 \\; s & \\mbox{ISM} \\\\ 5 \\; min & \\mbox{galactic halo} \\\\ 4\\; h & \\mbox{IGM} \\end{array} \\right. \\end{equation} after which we expect a resurgence in the bolometric flux to a level a factor $\\approx t_{b}/\\xi t_d$ below that of the burst proper, from which the easily recognizable, power--law afterglow begins. Afterglow intensity at peak frequency depends only on the environmental density and total explosion energy, while the peak frequency depends on total energy, $n$ and $\\gamma$, allowing a simultaneous solution which fixes $n$ (Eqs. \\ref{td}, \\ref{num}, \\ref{fnum}). {}" }, "0003/astro-ph0003346_arXiv.txt": { "abstract": "We are developing two-dimensional $16 \\times 25$ pixel detector arrays of both unstressed and stressed Ge:Ga photoconductive detectors for far-infrared astronomy from SOFIA. The arrays, based on earlier $5 \\times 5$ detector arrays used on the KAO, will be for our new instrument, the Far Infrared Field Imaging Line Spectrometer (FIFI LS). The unstressed Ge:Ga detector array will cover the wavelength range from 40 to 120~$\\mu$m, and the stressed Ge:Ga detector array from 120 to 210~$\\mu$m. The detector arrays will be operated with multiplexed integrating amplifiers with cryogenic readout electronics located close to the detector arrays. The design of the stressed detector array and results of current measurements on several prototype 16 pixel linear arrays will be reported. They demonstrate the feasibility of the current concept. {\\bf This paper does not include Figures due to astro-ph size limitations. Please download entire file at http://fifi-ls.mpe-garching.mpg.de/fifils.ps.gz.} ", "introduction": "For the Far Infrared Field Imaging Line Spectrometer (FIFI~LS)\\cite{gei98,raa99,loo00}, we need two-dimensional $16 \\times 25$ pixel detector arrays which cover the wavelength range from 40 to 210~$\\mu$m. Gallium-doped germanium detectors are proven to be very sensitive in the wavelength range of 40 to 120~$\\mu$m. Application of $\\approx$~600~Nmm$^{-2}$ of stress along the [100] crystallographic axis shifts the long wavelength cutoff from 120~$\\mu$m to approximately 220~$\\mu$m\\cite{kaz77}. Thus, we will use two Ge:Ga detector arrays, one stressed and one unstressed, to cover the desired wavelength range. The concentrations of Gallium dopants will be $1 \\times 10^{14}$~cm$^{-3}$ and $2 \\times 10^{14}$~cm$^{-3}$ for the stressed and unstressed arrays, respectively. The expected dark detector NEP is $\\leq 5 \\times 10^{-18}$~WHz$^{-1/2}$, which has been reached with a similar design in a balloon-borne experiment\\cite{hir89}. Some of the operational parameters of the detector arrays are listed in Tab.~\\ref{tab:para}. ", "conclusions": "" }, "0003/astro-ph0003170_arXiv.txt": { "abstract": "By using the flux ratio [Fe\\,II] $\\lambda$8617/[O\\,I] $\\lambda$6300, we demonstrate that most of the interstellar dust grains survive in shocks associated with supernova remnants and Herbig-Haro objects. The [Fe\\,II]/[O\\,I] flux ratio is sensitive to the gas-phase Fe/O abundance ratio, but is insensitive to the ionization state, temperature, and density of the gas. We calculate the [Fe\\,II]/[O\\,I] flux ratio in shocks, and compare the results with the observational data. When only 20\\% of iron is in the gas phase, the models reproduce most successfully the observations. This finding is in conflict with the current consensus that shocks destroy almost all the grains and $\\sim 100$\\% of metals are in the gas phase. We comment on previous works on grain destruction, and discuss why grains are not destroyed in shocks. ", "introduction": "For interstellar dust grains, the predominant destruction process is shocks driven by supernova explosions. In the postshock cooling gas, a charged grain is accelerated around the magnetic field line (betatron acceleration), collides with other grains and gas particles, and thereby loses its mass. Generally, it is believed that almost all the grains are destroyed in a single shock. The references are summarized in Savage \\& Sembach (1996) and Jones (2000). However, we would like to argue that the actual efficiency of grain destruction is as low as $\\sim 20$\\% by mass in representative shock-heated nebulae, i.e., supernova remnants (SNRs) and Herbig-Haro (HH) objects. The relative intensity of the emission lines [Fe\\,II] $\\lambda$8617 and [O\\,I] $\\lambda$6300 is used to estimate the gas-phase Fe/O abundance ratio. In the usual interstellar medium, iron is depleted into grains by a factor of $\\ge$ 100 as a major dust constituent, while oxygen is largely undepleted (Savage \\& Sembach 1996). Thus the gas-phase Fe/O ratio is proportional to the mass fraction of destroyed grains. The [Fe\\,II]/[O\\,I] flux ratio is sensitive to the gas-phase Fe/O ratio, but is insensitive to the ionization state, temperature, and density of the gas. This is because the same physical conditions are required to generate the [Fe\\,II] and [O\\,I] emissions. They are excited by electron collisions. Since the ionization potentials of Fe$^+$ and O$^0$ are only 16.2 and 13.6 eV, both the [Fe\\,II] and [O\\,I] emissions are generated in a partially ionized zone. The excitation energies of the [Fe\\,II] and [O\\,I] lines are 19,000 and 23,000 K. Their critical densities for collisional de-excitation at $10^4$ K are 3.5 $\\times$ $10^5$ and 1.8 $\\times$ $10^6$ cm$^{-3}$, which are well above the typical electron density in shocks. Moreover, the [Fe\\,II] and [O\\,I] lines are prominent in shocks. The grain destruction is expected to have been completed in their emission region, which is far downstream from the shock front. We calculate the [Fe\\,II]/[O\\,I] flux ratio in shocks, and compare the results with the observational data of SNRs and HH objects. The analysis and subsequent discussion employ the same values for atomic constants and interstellar abundances, the references of which are given below. ", "conclusions": "Though shocks destroy grains in SNRs and HH objects, the destruction is far from complete. Typically, 80\\% of iron is still locked into grains. However, many observational studies of shock-heated nebulae conclude that the grain destruction is almost complete, as reviewed by Savage \\& Sembach (1996) and Jones (2000; see also references for the data used in Fig. 1). In the usual interstellar gas, heavy metals such as Fe and Ca are depleted by factors of $10^2$--$10^4$ (Savage \\& Sembach 1996). If only a small fraction of the grains is destroyed, emission and absorption lines of those metals are greatly enhanced (Fesen \\& Kirshner 1980). The observer is easily misled to consider that a large fraction of the grains is destroyed. Moreover, owing to wide variations of physical quantities across the gas (Fig. 2), it is generally difficult to determine elemental abundances in shocks. Nevertheless, conclusions similar to ours were obtained in some of the past observations of SNRs. Phillips \\& Gondhalekar (1983) and Jenkins et al. (1998) observed ultraviolet absorption lines of stars behind S147 and Vela SNR, estimated column densities of gas-phase ions across these SNRs, and found depletion of Al. Raymond et al. (1988, 1997) observed ultraviolet and optical emission lines of Cygnus Loop and Vela SNR, compared their relative strengths with predictions of shock models, and found depletions of Fe and of C and Si. Oliva, Moorwood, \\& Danziger (1989) observed near-infrared emission lines of RCW~103, compared their strengths with model predictions, and found depletion of Fe. Reach \\& Rho (1996) detected continuum emission from grains in the far-infrared spectrum of W44. It should be noted that our result is more reliable than these previous ones. The Fe/O abundance ratio estimated from the [Fe\\,II] and [O\\,I] fluxes is robust with respect to the shock velocity and preshock density (Fig. 3). Since we studied the Fe/O abundance ratio alone, the grain survival probability estimated here is applicable, strictly speaking, only to Fe-bearing grains. There could exist several types of grains and subgrains which have different survival abilities. Of importance is a careful analysis of emission and absorption lines of the other elements. The present conclusion is nonetheless general. Observations of various Galactic interstellar clouds indicate that Mn, Cr, Ni, and Ti always have the same dust-phase fraction as Fe (Savage \\& Sembach 1996; Jones 2000). Though the major dust constituents Mg and Si appear to be more easily liberated to the gas phase than Fe, large fractions of Mg- and Si-bearing grains survive in shocks. The above observations indicate that, when 80\\% of Fe is locked into grains, $\\sim 50$\\% of Mg and Si are in the dust phase.\\footnote{ These dust-phase fractions were adapted from Savage \\& Sembach (1996). Their gas-phase Mg abundance was scaled by a factor of 2, in order to allow for the revised Mg$^+$ oscillator strengths (Fitzpatrick 1997). Since Savage \\& Sembach (1996) used the Zn abundance to normalize the Fe, Mg, and Si abundances, we made no correction for the difference in the assumed total (gas $+$ dust) abundances. Noticeably, in Galactic clouds, the observed $\\delta_{\\rm Fe}$ value is always less than 0.3. This fact supports our conclusion for SNRs and HH objects.} The present conclusion is, at least quantitatively, consistent with theoretical models. Jones, Tielens, \\& Hollenbach (1996) predicted that 60\\% (by mass) of silicate grains survive in a shock with $v_s$ = 100 km s$^{-1}$, $n_{{\\rm H},0}$ = 25 cm$^{-3}$, and $B_0$ = 3 $\\mu$G. This predicted probability of grain survival is somewhat low, but we could increase it. The above model assumes that grains are solid and homogeneous. If the grains are porous, e.g., consisting of several types and sizes of subgrains, they undergo less destruction (Jones et al. 1994). This is because their effective cross section is large. The resultant large gas drug prevents efficient betatron acceleration. Such porous grains are the natural result of coagulation of small grains into larger ones, and are found as interplanetary dust particles. The presence of porous grains is also suggested by the recent finding that the Sun is overabundant in heavy elements (Snow \\& Witt 1996). The amount of metals available for grains in the interstellar space is much less than that had been estimated from the solar abundances. However, the observed interstellar extinction per unit length puts a lower limit to the volume fraction of space occupied by grains. This situation calls for porous grains which have high volume-to-mass ratios (see also Jones et al. 1996; Mathis 1990, 1998). The destruction efficiency in shocks determines lifetime of interstellar grains. We estimate the grain lifetime in our Galaxy (Tielens 1998). The gas mass shocked by a supernova to a velocity equal to or greater than $v_s$ is 2500 ($v_s$/100 km s$^{-1}$)$^{-9/7}$ $M_\\sun$. Since the effective supernova rate is 8 $\\times$ 10$^{-3}$ yr$^{-1}$ and the mass of diffuse gas is 5 $\\times$ $10^8$ $M_\\sun$, the time interval for a grain to experience a supernova-driven shock with $v_s$ $\\ge$ 100 km s$^{-1}$ is 3 $\\times$ 10$^7$ yr. If each of the shocks destroys 20\\% of the grains, their mean lifetime is 2 $\\times$ 10$^8$ yr. The lifetime of the gas parcel itself is much longer, i.e., 2 $\\times$ 10$^9$ yr, which is estimated from the total gas mass 8 $\\times$ 10$^9$ $M_\\sun$ and the star formation rate 5 $M_\\sun$ yr$^{-1}$. Since interstellar grains are actually present, there has to exist some growth process, e.g., accretion of gas particles onto grains in dense gas (Jones et al. 1994). Tielens (1998) obtained a similar grain lifetime, from the metal depletions observed in diffuse and dense gases and the timescale for cycling the material between them. Finally, we underline that dust depletion is crucial to understanding spectra of shock-heated nebulae. Their gas-phase abundances are often assumed to represent the total (gas $+$ dust) abundances of the preshock gas. This assumption could be wrong. For example, near-infrared [Fe\\,II] emission lines at 1.257 and 1.644 $\\mu$m are more prominent by factors of $\\sim 500$ in SNRs than in H\\,II regions (Graham, Wright, \\& Longmore 1987). This fact is often explained by shock destruction of Fe-bearing grains. However, Mouri, Kawara, \\& Taniguchi (2000) found with the code MAPPINGS III that the flux ratio [Fe\\,II] 1.257 $\\mu$m/Pa$\\beta$ observed in SNRs is reproduced only when the gas-phase iron abundance is as low as in the H\\,II region M42. The flux ratio predicted for the solar abundance is too high. This finding motivated the present work. We adopted the flux ratio [Fe\\,II] $\\lambda$8617/[O\\,I] $\\lambda$6300 as a more reliable diagnostic, conducted numerical calculations in more detail for gas-phase metallicity, and thereby determined more precisely the gas-phase iron abundance." }, "0003/astro-ph0003164_arXiv.txt": { "abstract": "We have refined the estimate of the primordial level of $^7$Li abundance to an accuracy better than 10\\%, based on high-precision Li abundances for metal-poor halo stars, and a recent model of post-BBN (Big Bang Nucleosynthesis) chemical evolution that provides a quantitative explanation of the detected gentle ascent of the Spite Plateau for stars with metallicities [Fe/H]$ > -3$. Our maximum likelihood analysis obtains an estimate for the primordial Li abundance of $A({\\rm Li})_p=2.07^{+0.16}_{-0.04}$, after taking into account possible systematic errors in the estimation of Li abundances, with the exception of a still-controversial issue regarding stellar depletion. The inferred value of $\\eta$ (the baryon-to-photon number-density ratio in the universe) based on this estimate is more consistent with that derived from the set of reported ``low He'' + ``high D'' from extragalactic sites than that derived from reported ``high He'' + ``low D'' measurements. Since, within current models of stellar depletion processes, it is difficult to account for the observed very small scatter of Li abundance in metal-poor stars, our estimate of $A({\\rm Li})_p$ should be taken as an independent constraint on the baryonic mass density parameter in the universe, giving $\\Omega_b h^2=(0.64-1.4)\\times 10^{-2}$ with $h=H_0/100$ km s$^{-1}$Mpc$^{-1}$. ", "introduction": "The absolute abundances of $^4$He, $^2$D, and $^7$Li synthesized in the first three minutes following the hot Big Bang provide the key to a determination of the universal baryon density via its relationship to the $\\eta$ parameter. In order to refine observational estimates of the primordial levels of these light elements, previous attempts have been made to measure $^4$He and $^2$D abundances in extragalactic sites, where potential problems associated with correction for the effects of post-BBN chemical evolution could be minimized or avoided. However, the ``low'' value of $^{4}$He/H reported for metal-poor extragalactic HII regions by Pagel et al. (1992) and Olive, Skillman, \\& Steigman (1997) stands in contrast to the ``high'' value of $^{4}$He/H reported by Izotov \\& Thuan (1998). The ``high'' value for $^2$D/H reported for high-redshift intergalactic HI clouds by Songaila et al. (1994), Carswell et al. (1994), and Rugers \\& Hogan (1996) appears at odds with the ``low'' values for $^2$D/H from Tytler et al. (1996), Burles \\& Tytler (1998ab), and Burles et al. (1999). Detection of a high $^2$D/H abundance (Webb et al. 1997; Tytler et al. 1999) for a gas cloud at rather low redshift ($z\\simeq$0.7) has made the problem more complicated, because this is opposite to the expectation, based on one-zone models of chemical evolution, that the amount of $^2$D is should decrease following BBN as the result of various destruction processes. Given that two different values for both the primordial levels of $^4$He and $^2$D, leading to two distinguishable values of $\\eta$, remain tenable at present, there is a pressing need for accurate estimate of the primordial level of $^7$Li, which provides an independent constraint on $\\eta$. Obtaining such a constraint has proven difficult for two reasons. First, the prediction of the primordial $^{7}$Li abundance is a non-linear function of $\\eta$, which formally permits the assignment of two values for $\\eta$ at each level of measured primordial $^7$Li. Second, $^7$Li is several orders of magnitude less abundant than the other two light elements, so that high-precision observations of surface Li abundances are not possible except for nearby stars, which might have experienced the effects of chemical evolution in the Galaxy. Despite these apparent difficulties, ever since the discovery of a roughly constant value of Li ($^{6}$Li + $^{7}$Li) abundance in a small sample of metal-poor dwarf stars by Spite \\& Spite (1982, the so-called ``Spite Plateau''), many groups (e.g., Ryan et al 1996; hereafter RBDT; Bonifacio \\& Molaro 1997) have attempted to better determine the appropriate primordial level of this element (see Spite 2000 for a recent review). The recent high precision (and homogeneously analyzed) data of Ryan, Norris, \\& Beers (1999, hereafter RNB) showed that the Spite plateau is in fact incredibly ``thin,'' with an intrinsic star-to-star scatter in derived Li abundance $\\sigma < 0.02$ dex. RNB also claimed the existence of a statistically significant slope of $A$(Li) {\\it versus} [Fe/H] in the Spite Plateau at low metallicity (first detected by Thorburn 1994), apparently due to the influence of early Galactic chemical evolution. Ryan et al. (2000, hereafter RBOFN) have shown that the observed slope in the Spite plateau can be used to {\\it empirically} constrain the total expected contribution from Galactic Cosmic Rays (GCRs) and supernovae (SNe). These authors showed that a simple one-zone model for chemical evolution produces a slope which is of similar magnitude to that which is observed. However, their evaluation is necessarily tied to the presently uncertain relationship between O and Fe abundances in the early Galaxy. Furthermore, their evaluation is based on a model which ignores the expected stochastic nature of early star formation which is likely to apply during the first 10$^7$ to 10$^9$ years of chemical evolution in the early Galaxy. In this paper we employ the SN-induced chemical evolution model presented by Tsujimoto, Shigeyama, \\& Yoshii (1999; hereafter TSY) and Suzuki, Yoshii, \\& Kajino (1999; hereafter SYK), which has the great advantage of treating the production of heavy and light elements {\\it consistently} in the inhomogeneous early Galaxy (\\S 2). We recover the observed slope of $A$(Li) {\\it versus} [Fe/H], and obtain an estimate of the primordial level of lithium (\\S 3). We discuss the impact of our results in the context of the standard BBN model (\\S 4). ", "conclusions": "The importance of our new estimates of the primordial level of $^7$Li is that they are derived, for the first time, from a self-consistent model which explains both the $^6$LiBeB observations and the small, but {\\it real}, increasing trend of Li appearing at [Fe/H]$ > -3$. Figure 2 shows our preferred value of $A({\\rm Li})_p=2.07^{+0.15}_{-0.04}$ with a horizontal line, and the allowed 95\\% confidence intervals by a box, along with previous estimates of $^4$He and $^2$D. The theoretical prediction of standard BBN (Thomas et al. 1994; Fiorentini et al. 1998) is superposed. It is interesting to note that our preferred value lies at the very bottom of the valley of the function of $^7$Li abundance against $\\eta$, which means that we can assign a single value for $\\eta$, independent of the results from $^4$He and $^2$D. Our analysis indicates $\\eta({\\rm Li})=(1.7-3.9)\\times 10^{-10}$, which corresponds to a universal baryonic density parameter\\footnote{If taking into accout uncertainties ($1\\sigma$ errors) of nuclear reaction rates for the theoretical BBN calculation, the constraints become $\\eta({\\rm Li})=(1.4-4.5)\\times 10^{-10}$ and $\\Omega_b h^2=(0.53-1.7)\\times 10^{-2}$.} $\\Omega_b h^2=(0.64-1.4)\\times 10^{-2}$ with the Hubble constant expressed as $h=H_0/100$ km s$^{-1}$Mpc$^{-1}$. The range of $\\eta({\\rm Li})$ we obtain appears to agree best with that inferred from the reported ``low $^4$He'' + ``high $^2$D'' measured from extragalactic sites, rather than the pair of reported ``high $^4$He'' + ``low $^2$D''. This range is also consistent with that inferred from the primordial value of $^4$He obtained from recent observations of HII regions in the Magellanic Clouds (Peimbert \\& Peimbert 2000). However, the possible effects of stellar depletion, which are not taken into account here, might still play a role. If one adopts the reported ``high $^4$He'' and ``low $^2$D'' values as the correct ones with which to estimate $\\eta$, then it follows that $^7$Li has been destroyed through stellar evolution during the long lifetimes of metal-poor halo stars. From our Li constraint one might conclude that Li has been depleted by a factor of $2.7^{+0.7}_{-1.4}$ as a result of stellar processing. The required depletion factor, $\\sim$ 2$-$3, is much larger than the prediction ($ < 1.2$; Deliyannis et al. 1990) of so-called standard stellar evolution models which only take into account classical surface convection as the origin of mixing in stellar interiors. In order to destroy lithium in deeper and hotter regions extra mixing processes are necessary. Among such processes, stellar rotation is thought to be one of the most effective candidates, but depletion factors $\\sim 1$ dex inferred by rotation-induced mixing depend sensitively on the initial conditions of stellar rotation (Pinsonnealt et al. 1992). So, in general, this process predicts that one might expect to see a scatter about the Spite Plateau as large as the depletion factor itself, reflecting star-to-star differences in stellar rotation and other properties. Although recent models of rotation-induced mixing obtain more moderate depletion factors $\\sim 0.2-0.4$ dex (Pinsonnealt et al. 1999), the expected scatter in the Spite Plateau would still seriously contradict the very small intrinsic scatter ($\\sigma < 0.02$ dex) observed by RNB. All the above considerations indicate that, unless a novel process which is capable of significant {\\it and uniform} depletion of stellar Li abundance is identified, our preferred value of $A({\\rm Li})_p$ should be taken as a stringent constraint on $\\eta$ or $\\Omega_b$. The inferred value of the baryonic contribution to the density parameter is $\\Omega_b\\sim 0.01-0.02$ ($h=0.75$). This work has been supported in part by the Grant-in-Aid for the Center-of-Excellence research (07CE2002) of the Ministry of Education, Science, Sports, and Culture of Japan. TCB acknowledges partial support for this work from grant AST 95-29454 from the National Science Foundation." }, "0003/astro-ph0003022_arXiv.txt": { "abstract": "We present Li abundances for 14 solar--type stars in the intermediate age ($\\sim 2$~Gyr) clusters IC~4651 and NGC~3680. The \\nli~vs. effective temperature distributions are compared with those of the similar age cluster NGC~752, of the younger Hyades (600~Myr) and of the older M~67 (4.5~Gyr) and NGC~188 (6--7~Gyr) clusters. Neither IC~4651 nor NGC~3680 show the dispersion in Li which is observed in M~67. The 2~Gyr clusters have very similar Li vs. \\teff~distributions; in addition, stars in the upper envelope of the M~67 distribution have the same Li content as stars in the 2~Gyr clusters, suggesting that either they have not suffered any significant depletion between $\\sim$~2 and 4.5~Gyr or they had a much slower Li depletion. Mechanisms that lead to Li depletion on the main sequence are discussed in the light of these observations. None of the existing models seem to reproduce well the observed features. ", "introduction": "Understanding the processes that lead to lithium destruction on the main sequence (MS) provides powerful diagnostics of stellar structure and evolution and of mixing mechanisms in stars. \\begin{center} \\begin{table*} \\caption{The sample} \\begin{tabular}{rccccl}\\\\ \\hline & & & & & \\\\ \\multispan{6}{{\\bf IC~4651}\\hfill}\\\\ & & & & & \\\\ name & (B$-$V)$_0$ & EW(Li~{\\sc i}) & \\teff & \\nli & RV \\\\ (1) & & (m\\AA) & (K) & & info. (2) \\\\ EG 7 & 0.56 & 70 $\\pm$ 15 & 6061 & 2.7 $\\pm 0.12$ & S \\\\ EG 45 & 0.57 & 44 $\\pm$ ~~7 & 6016 & 2.4 $\\pm 0.12$ & S \\\\ AT 38 & 0.54 & 35 $\\pm$ ~~8 & 6122 & 2.4 $\\pm 0.06$ & \\\\ AT 39 & 0.54 & 49 $\\pm$ 10 & 6114 & 2.6 $\\pm 0.06$ & \\\\ AT 1108 & 0.53 & 47 $\\pm$ 10 & 6189 & 2.6 $\\pm 0.06$ & SB \\\\ AT 1109 & 0.62 & 61 $\\pm$ 15 & 5814 & 2.4 $\\pm 0.08$ & \\\\ AT 1225 & 0.51 & 64 $\\pm$ ~~7 & 6235 & 2.8 $\\pm 0.08$ & S \\\\ AT 2207 & 0.59 & 44 $\\pm$ 10 & 5908 & 2.3 $\\pm 0.08$ & \\\\ AT 2105 & 0.54 & 48 $\\pm$ 10 & 6110 & 2.5 $\\pm 0.08$ & S \\\\ AT 3226 & 0.58 & 49 $\\pm$ ~~8 & 5967 & 2.4 $\\pm 0.06$ & NM \\\\ AT 4226 & 0.62 & 31 $\\pm$ 10 & 5795 & 2.1 $\\pm 0.08$ & \\\\ & & & & & \\\\ \\multispan{6}{{\\bf(1)} From Eggen (\\cite{eg71}) and Anthony-Twarog et al. (\\cite{at88}). Photometry comes from the latter source.\\hfill}\\\\ \\multispan{6}{{\\bf(2)} Information on radial velocities (membership and binarity) was kindly provided by Dr. B. Nordstr\\\"om. \\hfill}\\\\ & & & & &\\\\ \\multispan{6}{{\\bf NGC~3680}\\hfill}\\\\ & & & & &\\\\ name & (B$-$V)$_0$ & EW(Li~{\\sc i}) & \\teff & \\nli & RV \\\\ & & (m\\AA) & (K) & & info. \\\\ 23 & 0.58 & 46$\\pm$ 5 & 5971 & 2.41 $\\pm$ 0.04 & SB1 \\\\ 60 & 0.59 & 48$\\pm$ 6 & 5924 & 2.38 $\\pm$ 0.05 & M? \\\\ 70 & 0.60 & 42$\\pm$ 5 & 5884 & 2.28 $\\pm$ 0.04 & M \\\\ 4114 & 0.58 & 51$\\pm$ 6 & 5951 & 2.44 $\\pm$ 0.05 & M? \\\\ & & & & &\\\\ \\multispan{6}{{\\bf(1)} Star numbers, colors, and information on membership/binarity were taken from Nordstr\\\"om et al. (1997). \\hfill}\\\\ \\hline \\end{tabular} \\end{table*} \\end{center} Surveys of Li among cluster and field stars have shown that standard models (i.e., those that take into account convecting mixing only) are in contradiction with many of the observational features (see Deliyannis \\cite{del00}; Jeffries \\cite{jef00}; Pasquini \\cite{pas00}, and references therein for recent reviews). Focusing on solar--type stars, at least two results in strong disagreement with model predictions have been found: {\\it i)} these stars appear to deplete Li while on the MS, in spite of the fact that their convective zones bases are too cool to burn Li; {\\it ii)} even more surprisingly, old, otherwise similar solar--type stars show different amounts of Li depletion. Solar--type stars in the solar age, solar metallicity cluster M~67 (4.5~Gyr) show a dispersion in Li abundances larger than a factor of 10 (Spite et al. \\cite{spi87}; Garc\\'\\i a L\\'opez et al. \\cite{gar88}; Pasquini et al. \\cite{pas97}; Jones et al. \\cite{jon99}), in contrast with the tight relationship between Li abundance and effective temperature (or mass) observed for single stars in the younger Hyades (age 600 Myr). The evolution from the Hyades to the age of the Sun seems to be `bimodal', with a fraction ($\\sim$ 60 \\%) of stars depleting a relatively small amount of lithium, and another fraction, virtually similar to the other one as far as mass and chemical composition are concerned, that undergoes a severe Li depletion. It is important to mention that, whereas no Li data for other clusters coeval to M~67 are available, a similar behavior is observed among old solar-type stars in the field (e.g., Pasquini et al. \\cite{pas94}). The Sun with $\\log$ n(Li) = 1.1 belongs to the class of Li-poor stars. In summary, not only we do not understand the mechanism(s) that lead to Li depletion on the MS, but these mechanism(s) seem to work differentially for otherwise similar stars. \\begin{figure} \\resizebox{8.8cm}{!}{\\includegraphics{fig1.eps}} \\vspace{-3.7cm} \\caption[]{\\nli~vs. \\teff~for our sample stars in IC~4651 and NGC~3680 (filled circles and triangles, respectively), the Hyades (stars), M~67 (open circles), NGC~752 (open squares), and NGC~188 (open triangles). M~67 data have been taken from Jones et al. (\\cite{jon99}). The SB2 binary S1045 is not included in the figure. Possible non--members in NGC~752 and NGC~188 are indicated as crossed symbols. Errors in \\nli~ take into account only errors in EWs. The Sun is also shown in the figure.} \\end{figure} \\begin{figure} \\resizebox{8.8cm}{!}{\\includegraphics{fig2.eps}} \\vspace{-3.7cm} \\caption[]{\\nli~vs. \\teff~for IC~4651, NGC~3680, NGC~752, and M~67. Symbols are the same as in Fig.~1. Dotted and dashed curves represent the 1.7~Gyr and 4~Gyr isochrones for models with rotationally induced mixing; upper and lower curves at the same age denote models with initial rotational velocities of 10 and 30 km/sec, respectively. } \\end{figure} As a possible solution for the dispersion observed among M~67 stars, Jones et al. (\\cite{jon99}) speculated that the mixing is driven by MS spin--down and angular momentum loss; in this case, the dispersion in Li would be due to different initial rotation rates. If this were the case, a spread in Li should be seen already at ages of 2 Gyr or younger (e.g., Pinsonneault \\cite{pin97}). More in general, in order to explain the observational evidences, several models have been developed in the last decade, involving more complex physics or/and additional mixing processes; the validation of these models still needs strong observational support, in particular at ages intermediate between the Hyades and M~67 where very few data are available. Determining Li in a significant sample of stars in intermediate age clusters is thus extremely important: so far Li measurements for clusters of this age have been obtained only for very few objects in NGC~752 (Hobbs \\& Pilachowski, \\cite{hp86}). We present here Li abundances for a sample of solar--type stars in the open clusters IC~4651 and NGC~3680. These clusters, for which metallicities close to solar have been derived (Friel \\cite{fr95}), have estimated ages of 1.5 -- 2 Gyr (e.g., Meynet et al. \\cite {mey93}; Friel \\cite{fr95}); these data, therefore, allow us to address the problem of Li depletion from the age of the Hyades to that of M~67 and to put observational constraints on the mechanisms responsible for MS Li depletion and on their timescales. ", "conclusions": "" }, "0003/astro-ph0003214_arXiv.txt": { "abstract": "Cosmic-ray production in young supernova remnant (SNR) shocks is expected to be efficient and strongly nonlinear. In nonlinear, \\DSA, compression ratios will be higher and the shocked temperature lower than \\TP , Rankine-Hugoniot relations predict. Furthermore, the heating of the gas to X-ray emitting temperatures is strongly coupled to the acceleration of cosmic-ray electrons and ions, thus nonlinear processes which modify the shock, influence the emission over the entire band from radio to gamma-rays and may have a strong impact on X-ray line models. Here we apply an algebraic model of nonlinear acceleration, combined with SNR evolution, to model the radio and X-ray continuum of Kepler's SNR. \\vskip12pt {Proceedings of the ACE-2000 Symposium on {\\it The Acceleration and Transport of Energetic Particles Observed in the Heliosphere}, January 5 - 8, 2000, Indian Wells, CA} ", "introduction": "More than twenty years of spacecraft observations in the heliosphere have proven that collisionless shocks can accelerate particles with high efficiency, \\iec 10-50\\% of the ram energy can go into superthermal particles (\\egc Eichler \\cite{Eich81}; Gosling \\etal\\ \\cite{GoslingEtal81}; Ellison \\etal\\ \\cite{EMP90}; Terasawa \\etal\\ \\cite{Terasawa99}). Energetic particles exist throughout the universe and shocks are commonly associated with them, confirming that shock acceleration is important beyond the heliosphere as well. In fact, shocks in supernova remnants (SNRs) are believed to be the main source of Galactic cosmic rays, and these shocks are expected to be much stronger than those in the heliosphere and can only be more efficient and nonlinear. The conjecture that collisionless shocks are efficient accelerators is strengthened by results from plasma simulations, which show efficient shock acceleration consistent with spacecraft observations (\\egc Scholer, Trattner, \\& Kucharek \\cite{STK92}; Giacalone \\etal\\ \\cite{GBSE97}), and other indirect evidence comes from radio emission from SNRs (see Reynolds \\& Ellison \\cite{RE92}) and equipartition arguments in AGNs and $\\gamma$-ray bursts (see Blandford \\& Eichler \\cite{BE87} for an early review). There is also clear evidence that shocks can produce strong self-generated turbulence. This has long been seen in heliospheric shocks (\\egc Lee \\cite{Lee82, Lee83}; Kennel \\etal\\ \\cite{KennelEtal84}; Baring \\etal\\ \\cite{BaringEtal97}) and there is evidence that it occurs at SNRs as well (\\iec Achterberg, Blandford, \\& Reynolds \\cite{ABR94}). While the importance of nonlinear (NL) shock acceleration is evident, NL solutions to \\DSA\\ are complicated and results are often unwieldy and difficult to use for astrophysical applications. Therefore, we have developed a simple, algebraic model of \\DSA, based on more complete studies, which includes the essential nonlinear effects (Berezhko \\& Ellison \\cite{BEapj99}; Ellison, Berezhko, \\& Baring \\cite{EBB2000}). This technique is computationally fast and easy-to-use, yet includes (i) the modification of particle spectra when the backpressure from energetic ions smooths the shock structure, and (ii) the influences on the shock dynamics when the magnetic turbulence is strongly amplified by wave-particle interactions. The complications of NL shock acceleration and the many parameters required to characterize it are offset somewhat by the fact that the entire particle distribution function, from thermal to the highest energies, is interconnected and must be accounted for self-consistently with a nonthermal tail connecting the quasi-thermal population to the energetic one. Because energy is conserved, a change in the production efficiency of the highest energy particles {\\it must} impact the thermal properties of the shock heated gas and vice versa. If more energy goes into relativistic particles, less is available to heat the gas. In contrast, the power laws assumed by test-particle models have no connection with the thermal gas, energy conservation does not constrain the normalization of the power law, and the spectral index can be changed with no feedback on the thermal plasma. Furthermore, there is a direct linkage between protons and electrons (which produce most of the photon emission associated with shocked gas) in nonlinear models, so the entire emission from radio to gamma-rays, plus cosmic-ray observations, can, in principle, be used to constrain the models. Here we describe some of the nonlinear features expected to occur in young SNRs and investigate some implications of efficient cosmic-ray production on the broad-band continuum from Kepler's SNR. We refer to Berezhko \\& Ellison \\cite{BEapj99} and Ellison, Berezhko, \\& Baring \\cite{EBB2000} for details of the NL shock model and its application to particle and photon production in SNRs. Work on the NL X-ray line emission from Kepler is in progress, i.e., Decourchelle, Ellison, \\& Ballet \\cite{DEB2000}. Previous test-particle calculations of the X-ray emission from Kepler have been reported by Borkowski, Sarazin, \\& Blondin \\cite{BSB94}, who used a two-dimensional hydrodynamic simulation, and by Rothenflug \\etal\\ \\cite{RMCB94}, who investigated the emission from the reverse shock. The work of Decourchelle \\& Ballet \\cite{Decour94} is mentioned below. \\vspace{-6 mm} ", "conclusions": "X-ray line and continuum emission contains a vast amount of information on supernova (SN) ejecta elemental composition, ISM density and elemental composition, the SN explosion energy, and the mass of ejecta. In addition to heating the plasma, the forward and reverse shocks accelerate some fraction of the shocked material to cosmic-ray energies and this acceleration is believed to be quite efficient, removing energy from the thermal plasma (\\egc Kang \\& Jones \\cite{KJ91}; Dorfi \\& B\\\"ohringer \\cite{DorfiB93}; Berezhko, Ksenofontov, \\& Petukhov \\cite{BereKP99}). Despite the expected efficiency of shock acceleration, virtually all current X-ray line models assume that the shocks that heat the gas {\\it do not} place a significant fraction of their energy in cosmic rays (exceptions to this are Chevalier \\cite{Chev83} and Dorfi \\cite{Dorfi94}). Here, we investigated the broad-band continuum emission expected in Kepler's SNR from efficient shock acceleration, by coupling self-similar hydrodynamics (Chevalier \\cite{Chev83}; Truelove \\& McKee \\cite{Truelove99}) with nonlinear diffusive shock acceleration (\\egc Berezhko, Ksenofontov, \\& Petukhov \\cite{BereKP99}). We were able to show that the radio and X-ray continuum can be fit with reasonable parameters in a way that allows the reverse shock to contribute substantially to the total X-ray emission. This constraint is required since X-ray line models of Kepler (\\egc Decourchelle \\& Petre \\cite{Decour99}) require emission from the shock-heated, metal-rich ejecta material. This preliminary calculation is not fully self-consistent for several reasons. Most importantly, we use self-similar results (\\iec Truelove \\& McKee \\cite{Truelove99}) to model the SNR evolution. These solutions neglect the effects of energetic particle escape from the FS and assume that the ratio of cosmic-ray pressure to total pressure at the shock front is a constant. They also assume that $\\gameff = 5/3$. As indicated in the bottom panel of Figure~\\ref{fig:tp_nl}, the total shocked pressure doesn't differ much with or without cosmic-ray production and we have demonstrated (\\iec Decourchelle, Ellison, \\& Ballet \\cite{DEB2000}) that NL shock results do not change dramatically over most of the age of Kepler for typical values of the injection parameter, $\\etainjP > 10^{-4}$. However for lower $\\etainjP$, the nonlinear solutions can have test-particle, unmodified solutions at very high sonic Mach numbers with a rapid transition to the NL solution as the Mach number drops (see Fig.~\\ref{fig:tp_nl} and Berezhko \\& Ellison \\cite{BEapj99} for a detailed discussion). The self-similar solutions we use are still approximate, however, because we have not yet modified them for the change in $\\gameff$ that results when a substantial fraction of the shocked pressure is in cosmic rays. We also neglect cosmic-ray diffusion and assume they are spatially coupled to the gas -- an excellent approximation for all but the highest energy particles. However, the highest energy electrons produce the X-ray \\syn\\ photons so there may be differences that are not modeled in the emission volumes and other important parameters between the radio and X-ray bands. Finally, we have not included absorption in our models which is probably not important for the Kepler radio observations, but will be required to model the low energy X-rays.\\footnote{Note that we have only plotted the X-ray observations above 2 keV in Figure~\\ref{fig:kepler} to avoid conflict with the absorbed low energy end of the X-ray distribution.} Besides providing a more self-consistent model of photon production, predictions from NL shock models provide a test of the fundamental assumption that SNRs are the primary source of galactic cosmic-ray ions. If this is so, the acceleration is almost certainly nonlinear since 5-30\\% of the total ejecta kinetic energy is required to replenish cosmic rays as they escape from the galaxy. Since shocks put more energy into accelerated ions than electrons, nonlinear effects seen in X-ray emission will be evidence for the efficient shock acceleration of ions as well as electrons. X-ray observations potentially provide {\\it in situ} information on cosmic-ray ion production, complementing observations of pion-decay $\\gamma$-rays in this regard. Any inference of nonthermal tails on electron distributions in X-ray observations will provide information on electron injection, the least well understood aspect of shock acceleration. Such {\\it in situ} information on high Mach number shocks is available nowhere else. Our modeling of Kepler's SNR suggests that typical source parameters produce large nonlinear effects in the broad-band spectrum and suggest that the test-particle approximations that are almost universally used are inadequate for SNRs as young as this. Besides the differences discussed above, one might expect that the growth rate of the Rayleigh-Taylor instability will be greater in a cosmic-ray modified shock because of the larger spatial gradients of density, pressure, etc. Furthermore, the high compression ratios result in a considerably thinner region between the forward and reverse shocks than predicted in the TP case (Decourchelle, Ellison, \\& Ballet \\cite{DEB2000}). This places the contact discontinuity closer to the shock and may make it easier for the Rayleigh-Taylor ``fingers'' to distort or overtake the FS, a situation that appears difficult with normal TP parameters (Chevalier, Blondin, \\& Emmering \\cite{CBE92}; Chevalier \\& Blondin \\cite{CB95}). Another important difference concerns electron heating and equilibration. The higher densities in the NL models mean that electron heating is much more efficient than in TP shocks. Our initial calculations \\cite{DEB2000} suggest that it may be possible to obtain full equipartition between electrons and ions, at least for high values of $\\etainjP$, in the shocked ejecta. \\vspace{-6 mm}" }, "0003/astro-ph0003091_arXiv.txt": { "abstract": "NGC~3532 is a very rich southern open cluster of age $\\sim 200-350$ Myr; it is therefore a good candidate to investigate the X-ray activity--age--rotation relationship at ages intermediate between the Pleiades and the Hyades, where, to our knowledge, X-rays studies exist for only one cluster (NGC~6475). We have performed an X-ray study of NGC~3532 using HRI observations retrieved from the $ROSAT$ archive. The observations have a limiting sensitivity $L_{\\mathrm x} \\sim 4 \\times 10^{28}$ erg~sec$^{-1}$ in the center of the field. We detected $\\sim 50$ X-ray sources above a $4 \\sigma$ threshold, half of which have a known optical counterpart within 10 arcsec; 15 of the X-ray sources have at least one cluster member as optical counterpart. A comparison of NGC~3532 with the nearly coeval cluster NGC~6475 indicates that the former cluster is considerably X-ray underluminous with respect to NGC~6475. However, because of the existence of possible selection effects, additional X-ray and optical observations are needed before definitively concluding that the X-ray properties of NGC~3532 and NGC~6475 are significantly different. ", "introduction": "The $ROSAT$ PSPC and HRI detectors have provided X-ray images for a large number of open clusters sampling the age range from $\\sim$~20 to 600~Myr (e.g., Randich \\cite{randich00} and references therein; Jeffries \\cite{jeffries99} and references therein; see also Belloni \\cite{belloni97}, for a review on older open clusters). The data have allowed investigating in great detail the dependence of X-ray activity on mass, age, rotation, and, in particular, to check the validity of the rotation--activity--age paradigm. The overall picture emerging from $ROSAT$ generally confirms that there is a tight dependence of X-ray activity on rotation (or on the so called Rossby number, the ratio of the rotation period over the convective turnover time -- e.g., Noyes et al. \\cite{noyes84}) and, through rotation, on age: the level of X-ray activity increases with increasing rotation and, since stars spin down as they age, the average or median X-ray luminosity decays with increasing age. However, the X-ray luminosity (or X-ray over bolometric luminosity) does not depend simply on some power of the rotational rate, and the activity--age dependence cannot be described by a Skumanich--type power law. In addition, a few puzzling results have arisen from $ROSAT$ data. For example, the finding that the bulk of the population of Praesepe solar-type stars have a significantly lower X-ray luminosity than the coeval Hyades and Coma Berenices clusters (Randich \\& Schmitt \\cite{randich95}; Randich et al. \\cite{randich96}) has casted doubts on the common thinking that a unique activity--age relationship holds, and, consequently, that the X-ray properties of a cluster of a given age are representative of all clusters of the same age. A study by Barrado y Navascu\\'es et al. (\\cite{barrado98}) seems to exclude that this result is due to a strong contamination of the Praesepe sample by cluster non-members; at the same time, $ROSAT$ observations of NGC~6633 suggest that this cluster, which is coeval to the Hyades and Praesepe, is more Praesepe--like than Hyades--like (Franciosini et al. \\cite{francio00}; Totten et al. \\cite{totten00}). We also mention that the comparison of the Pleiades (120~Myr) with NGC~6475 (200~Myr) and with other clusters with ages of the order of 100--200 Myr also suggests that a tight/unique age--activity relationship may not hold (e.g. Randich \\cite{randich00}). The issue of the uniqueness of the activity--age relationship is therefore not at all settled. In addition to optical studies that should ascertain cluster membership and provide complete (or close to completeness) lists of members and better defined cluster ages, additional, and possibly deeper, X-ray surveys of samples of coeval clusters are clearly required to further address this problem. We present here a $ROSAT$ study of the NGC~3532 cluster: NGC~3532 is a very rich southern open cluster with an estimated age $\\sim 200-350$ Myr (Fernandez \\& Salgado \\cite{fs80}; Johansson \\cite{johan81}; Eggen \\cite{eggen81}; Koester \\& Reimers \\cite{koester93}; Meynet et al. \\cite{meynet93}); it is therefore a good candidate to investigate the X-ray activity--age--rotation relationship at ages intermediate between the Pleiades and the Hyades, where, to our knowledge, X-rays studies exist for only one cluster (NGC~6475). The most likely value for the reddening of NGC~3532 is $E(B-V) = 0.04$ (Fernandez \\& Salgado \\cite{fs80}; Eggen \\cite{eggen81}; Schneider \\cite{schneid87}; Meynet et al. \\cite{meynet93}); the metallicity of the cluster has been estimated to be close to solar ([Fe/H] $\\sim 0.02$; Clari\\'a \\& Lapasset \\cite{claria88}). The cluster is located at very low galactic latitude ($b=+1.43$ ~deg). Distance determinations range from 405$^{+76}_{-55}$~pc (from Hipparcos; Robichon et al. \\cite{robich99}) to 500 pc (Eggen \\cite{eggen81}); in this paper the most recent value of 405 pc by Robichon et al. (\\cite{robich99}) has been adopted. \\begin{figure} \\resizebox{\\hsize}{!}{\\includegraphics{fig1.ps}} \\caption{ $V$ vs. $(B-V)$ diagram for the probable and possible members of NGC~3532 included in the HRI field of view. Filled circles indicate stars detected in X-rays } \\label{cmd} \\end{figure} ", "conclusions": "We have analyzed $ROSAT$ archive data of the open cluster NGC~3532. The comparison of the X-ray properties of solar-type stars in the cluster with those of the supposedly coeval NGC~6475 cluster indicates that NGC~3532 is considerably X-ray underluminous with respect to NGC~6475. If this result is not due to selection effects and biases in the two cluster samples, it would provide an additional piece of evidence that the X-ray activity--age relationship is not unique and that other parameters, in addition to rotation, determine the level of coronal emission. However, before such a conclusion can be accepted, additional X-ray and optical observations should be performed. Namely, {\\bf I.} an additional X-ray survey of NGC~6475 should be carried out; the survey should be deeper than the $ROSAT$ ones so that, if present, an X-ray faint population of cluster members could be detected; {\\bf II.} additional photometric and spectroscopic studies of NGC~3532 should be carried out in order to confirm cluster membership for the optical candidates known at present and to detect still unidentified solar-type and lower mass stars in the cluster. These studies would also provide information on rotation for cluster members; {\\bf III.} If possible, an effort should also be done, once more low-mass cluster members are known, to provide a definitive estimate of the cluster age using also low main-sequence fitting. Besides the 15 cluster members, the X-ray survey resulted in the detection of 13 foreground/background stars -- which is not surprising given the low cluster galactic latitude -- and of 21 objects without any known optical counterparts. Priority should be given to optical observations aimed at determining the nature of these sources, and, in particular, at ascertaining whether they are cluster members or not." }, "0003/astro-ph0003058_arXiv.txt": { "abstract": "Global solutions of optically thick advective accretion disks around black holes are constructed. The solutions are obtained by solving numerically a set of ordinary differential equations corresponding to a steady axisymmetric geometrically thin disk. We pay special attention to consistently satisfy the regularity conditions at singular points of the equations. For this reason we analytically expand a solution at the singular point, and use coefficients of the expansion in our iterative numerical procedure. We obtain consistent transonic solutions in a wide range of values of the viscosity parameter $\\alpha$ and mass accretion rate. We compare two different form of viscosity: one takes the shear stress to be proportional to the pressure, while the other uses the angular velocity gradient-dependent stress. We find that there are two singular points in solutions corresponding to the pressure-proportional shear stress. The inner singular point locates close to the last stable orbit around black hole. This point changes its type from a saddle to node depending on values of $\\alpha$ and accretion rate. The outer singular point locates at larger radius and is always of a saddle-type. We argue that, contrary to the previous investigations, a nodal-type inner singular point does not introduce multiple solutions. Only one integral curve, which corresponds to the unique global solution, passes simultaneously the inner and outer singular points independently of the type of inner singular point. Solutions with the angular velocity gradient-dependent shear stress have one singular point which is always of a saddle-type and corresponds to the unique global solution. The structure of accretion disks corresponding to both viscosities are similar. \\noindent {\\em Subject headings:} Accretion, accretion disks --- black hole physics --- hydrodynamics ", "introduction": "Accretion discs are formed when the matter with a large angular momentum is falling into a black hole or another gravitating body. The well known objects where the accretion disks are found are protostar nebulae, binary X-ray sources, cataclysmic variables, active galactic nuclei and others. In this paper we discuss accretion disks around black holes. The standard model of geometrically thin accretion disk has been developed by Shakura (1972), Novikov \\& Thorne (1973) and Shakura \\& Sunyaev (1973), and has played a significant role in the development of accretion theory (see Pringle 1981; Frank, King, \\& Rain 1992; Kato, Mineshige, \\& Fukue 1998 for reviews). The standard model bases on the vertically averaged approach to equilibrium, and a suggestion of the local thermal balance in which the viscous heating of the gas is balanced by the local radiative cooling. Non-local effects, like the radial advection of thermal energy and the transonic nature of accretion flow, are neglected in the standard model. This simplified approach allows to reduce the general problem to a set of algebraical equations. Such a simple description becomes possible due to an approximate parameterization of the viscosity stress tensor with one non-zero component, $$ t_{r\\phi}=-\\alpha P, \\eqno(1) $$ suggested by Shakura (1972). The standard model gives a satisfactory appropriate solution of the problem at low accretion rates $\\dot{M}\\la 16 L_{Edd}/c^2$, where $L_{Edd}$ is the Eddington luminosity. Simplified solution with inclusion of the advective terms into equations described the vertically integrated models of accretion disks was obtained by Paczy\\'nski \\& Bisnovatyi-Kogan (1981). This approach with some modifications have been used by many researchers to study transonic accretion flows around black holes (Muchotrzeb \\& Paczy\\'nski 1982; Muchotrzeb 1983; Matsumoto et al. 1984; Fukue 1987; Abramowicz et al. 1988; Chen \\& Taam 1993; Beloborodov 1998). The importance of the transonic nature of the accretion flows on the disk structure has been emphasized by H\\\"oshi \\& Shibazaki (1977), Liang \\& Thompson (1980) and Abramowicz \\& Zurek (1981), and later studied in more details by Abramowicz \\& Kato (1989). Despite a significant progress in the study of optically thick accretion disks obtained during almost three decades there are a number of unsolved problems still posed in the theory. The problems are connected with a possible non-uniqueness of a solution at $\\alpha\\ga 0.01$ and a non-standard behavior of a singular point type. It was reported by Matsumoto et al. (1984), Muchotrzeb-Czerny (1986) and Abramowicz et al. (1988) that in the case of viscosity prescription (1) the singular point changes its type from a saddle to node when one increases $\\alpha$. The presence of the nodal-type singular point leads to creating of a possibility of multiple solutions as the authors have claimed. A similar change of the singular point type was reported by Chen \\& Taam (1993), who used the angular velocity gradient-dependent viscous stress, $$ t_{r\\phi}=\\rho\\nu r {d\\Omega\\over dr}, \\eqno(2) $$ where $\\nu$ is the kinematic viscosity coefficient defined by (14). Narayan, Kato, \\& Honma (1997) have compared two forms of viscosity (1) and (2) in the case of radiatively inefficient advection-dominated accretion flows. They concluded that the structures of flows corresponded to both viscosities are similar at $\\alpha<0.15$. In this paper we show that the mentioned problems have been created by several inconsistencies in the preceding studies. Some problems are connected with an inaccurate averaging of the equations over a disk thickness (Chen \\& Taam 1993), another ones appear due to an incomplete investigation of the singular points (Abramowicz et al. 1988). We have found that in the case of viscosity prescription (1) a set of equations describing the vertically averaged advective accretion disks has {\\it two} singular points, independed of $\\alpha$ and accretion rate. Note, that multiplicity of singular points in solutions for accretion flows in Paczy\\'nski-Wiita potential (3) was revealed by Fukue (1987), Chakrabarti \\& Molteni (1993) and Chakrabarti (1996) in a somewhat different context. We have shown, that at $\\alpha\\la 0.01$ the inner and outer (with respect to the black hole location) singular points are of the saddle type, and only one integral curve (``separatrix'') which crosses the inner point simultaneously crosses the outer one. This separatrix corresponds to the unique global solution which is determined by two parameters, $\\alpha$ and $\\dot{m}=\\dot{M}c^2/L_{Edd}$, for a given black hole mass. In Figure~1a the structure of integral curves is schematically represented in the vicinity of the global solution which is shown by the thick line. At larger $\\alpha\\ga 0.1$ the inner singular point is changed to a nodal-type one, while the outer point remains of a saddle-type. There is still one integral curve which goes continuously through both singular points providing a unique global solution, as it is shown in Figure~1b. In the case of viscosity prescription (2) we have found that there is only one singular point which is always a saddle, and only one physical solution which passes through this point exists. Solutions which correspond to both forms of viscosity (1) and (2) are very close at low $\\alpha$ limit, $\\alpha\\la 0.1$. We have developed a numerical method to solve the set of equations describing the vertically averaged advective accretion disks. The method is based on the standard relaxation technique and explicitly uses conditions at the inner singular point and its vicinity. We have obtained these conditions by expanding a solution into power series around the singular point. Such a modification of the method allows to construct solutions which smoothly pass the singular points and satisfy the regularity conditions at these points with high computer precision in wide range of parameters $\\alpha$ and $\\dot{m}$. The paper is organized as follows. In \\S2 we formulate a mathematical approach to the problem, write a set of equations, and formulate boundary conditions. In \\S3 we investigate critical points and discuss behavior of physical values in their vicinity. In \\S4 we describe our numerical results and discuss them in \\S5. Details of the numerical method and explicit expansion of physical quantities in the vicinity of the critical points are represented in Appendixes~A and B, respectively. ", "conclusions": "We have obtained unique solutions for structure of advective accretion disk in a wide range of accretion rates and $\\alpha$-parameters. Both viscosity prescriptions (1) and (2) have been investigated. The solutions corresponding to both prescriptions are very close for $\\alpha \\la 0.1$, and begin to differ at larger $\\alpha$. This is connected, probably with larger deviation of the angular velocity $\\Omega$ from the Keplerian one, $\\Omega_{K}$, leading to larger difference between $t_{r\\phi}$ in both prescriptions. Unfortunately, our comparison of the prescriptions is not complete due to technical problems in calculation of the high viscosity models in the case of viscosity prescription (2). The main difference of the present study from the previous ones is in using more sophisticated numerical technique which accurately treats the regularity conditions in the inner singular point of equations. We have performed an analytical expansion at the singular point to calculate the derivatives of physical quantities. These derivatives help us to find the proper integral curve passing through the singular point. The approach allow us to avoid numerical instabilities and inaccuracies, appearing when only variables at the singular point, but not its derivatives, are included into a numerical scheme. We have found different behaviour of integral curves depending on used viscosity prescription. In the case of viscosity prescription (1) there are two singular points located at $(r_s)_{in}$ and $(r_s)_{out}$. The inner point, $(r_s)_{in}$, locates close to the last stable black hole orbit (see Figure~2), and is an analogy of the singular point in spherical flow, where the point divides the subsonic and supersonic regions. The location of the outer point, $(r_s)_{out}$, is determined by the accretion rate (see Figure~3). At low $\\alpha\\la 0.1$ both points are of a saddle-type. Only one integral curve (``separatrix'') simultaneously crosses two saddle-type points, as it is shown in Figure~1a, and corresponds to the global solution which smoothly connects the supersonic innermost region of the accretion disk and the subsonic outer (formally at $r=\\infty$) parts. For larger $\\alpha\\ga 0.1$ the inner singular point changes its type to a node. There was suggestion by Muchotrzeb-Czerny (1986) and Abramowicz et al. (1988) that there is no unique solution in this case, because of all integral curves cross the node. Existence of a unique separatrix crossed simultaneously both singular points preserves a uniqueness of the solution in this case (see Figure~1b). The conclusion of Muchotrzeb-Czerny (1986) and Abramowicz et al. (1988) is probably connected with their neglection of outer singular points inherited to the problem. Matsumoto et al. (1984) used slightly different form of equations (6) and (8), and they found that in this case only one singular point exists and changes type from a saddle to node. The difference with respect to our results arises because of using different form of the pressure gradient force [the first term on the right hand side of equation (6)]. We use the pressure and density taken at the equatorial plane in this term, whereas Matsumoto et al. (1984) used the vertically averaged quantities in it. In the latter case the term has the following form, $$ {1\\over\\Sigma}{{\\cal P}\\over dr}, \\quad {\\rm where} \\quad \\Sigma=2\\int_0^h\\rho dz \\quad {\\rm and} \\quad {\\cal P}=2\\int_0^h P dz. \\eqno(33) $$ The vertically integrated approach (33) introduces the difference because in this case the free terms with $\\alpha^2$ in (20) are absent. Formally, it corresponds to location of the outer singular point at infinity, where ${\\cal M}_2=0$. Such a visible difference is not qualitatively important for the physical solution, because conditions at the outer singular point are only shifted to infinity, and the integral curve itself has little changes. Thus, similar to our results, in the approach by Matsumoto et al. (1984) the inner critical points of both types, a saddle or node, correspond to a unique solution. In the case of viscosity prescription (2) we have found that one singular point exist. The point is always of a saddle-type and determines a unique solution. Another results had been obtaining by Chen \\& Taam (1993). They also found that equations have one singular point, but the point changes its type from a saddle to nodal one, depending on $\\alpha$ and the accretion rate. It is not clear why such a result was obtained. There are two main differences in our equations (6) and (8), and those used by Chen \\& Taam (1993). First, they used the same vertical averaging for the equation of motion as Matsumoto et al. (1984). Second, they used the vertically averaged energy equation which corresponds to the inappropriate polytropic relation, ${\\cal P}\\propto\\Sigma^\\gamma$, when one neglects terms corresponding to the viscous heating and radiative cooling. Here $\\gamma$ is the effective adiabatic index, and other notations are similar to those used in (33). Our equation (8) corresponds to the correct polytropic relation, $P\\propto\\rho^\\gamma$. It could be that one of the mentioned differences results in the change of critical point type. \\noindent{\\it Acknowledgments.} This work was supported in part by RFBR through grant 99-02-18180, the Royal Swedish Academy of Sciences, the Danish Natural Science Research Council through grant No 9701841, Danmarks Grundforskningsfond through its support for establishment of the Theoretical Astrophysics Center. \\appendix" }, "0003/astro-ph0003434_arXiv.txt": { "abstract": " ", "introduction": "There is strong observational and theoretical evidence that dark, nonbaryonic matter accounts for about a third of the critical density of the Universe (for a recent review see \\cite{mturner} and references therein). Many candidates have been proposed and some of them (cosmions, heavy Dirac neutrinos) have already been rejected. Slow thermal relics born in an early phase of the Universe, stable or very long lived, are excellent candidates for nonbaryonic dark matter \\cite{kolbturner}. These weakly interacting ($\\sigma \\leq \\sigma_{weak})$, massive (1 GeV - 1 TeV) particles (WIMPs) arise independently from cosmological considerations in supersymmetric models as neutralinos - the lightest supersymmetric particles. Direct detection of neutralinos can occur in very low background experiments, where the elastic neutralino scattering off target nuclei \\cite{goodwitt} is exploited. In the following, after giving a short overview on the principles of direct detection, we will describe existing or planned experimental techniques aiming at direct WIMP detection, illustrated by examples of currently running experiments and of future projects. We do not intend a complete coverage of all direct detection experiments, but rather present some of the most promising ones. ", "conclusions": "Doubtless the present situation in the field of direct dark matter detection is exciting. A plenitude of experiments using very different detection techniques have reached sensitivities which for the first time start to probe the supersymmetric parameter space. A well motivated WIMP candidate from supersymmetry, the neutralino, might thus reveal its presence in the near future. While the report of evidence for a WIMP signal by one experiment has raised an equal amount of interest and criticism, confirmation or exclusion by other experiments is expected soon. The goal of future projects has shifted from merely setting limits on the WIMP interaction strength to the ability of detecting a distinctive WIMP signature. The experimentally explored WIMP signatures up to now, leading to unambiguous WIMP identification, are caused by the Earth`s motion with respect to the galactic rest frame. The hope is that future dark matter experiments will not only unveil the major composition of matter in the Universe, but will possess the aptitude to study its inherent features, thus opening the avenue towards discerning between different galactic halo models. Needless to say, direct detection of dark matter would have striking implications for particle physics and cosmology." }, "0003/astro-ph0003328_arXiv.txt": { "abstract": "We present high--resolution $J$--, $H$--, and $K$--band observations and the first $H-K$ color image of the carbon star \\object{IRC +10 216}. The images were reconstructed from 6 m telescope speckle interferograms using the bispectrum speckle interferometry method. The $H$ and $K$ images with resolutions between 70~mas and 92~mas consist of several compact components within a 0\\farcs2 radius and a fainter asymmetric nebula. The brightest four components are denoted with A to D in the order of decreasing brightness in the 1996 image. A comparison of our images from 1995, 1996, 1997, and 1998 gives --- almost like a movie of five frames --- insight into the dynamical evolution of the inner nebula. For instance, the separation of the two brightest components A and B increased from 191 mas in 1995 to 265 mas in 1998. At the same time, component B is fading and the components C and D become brighter. The {\\sf X}--shaped bipolar structure of the nebula, most prominently present in the $J$--band image, implies an asymmetric mass--loss. Such asymmetries are often present in protoplanetary nebulae but are unexpected for AGB stars. \\object{IRC +10 216} is thus likely to be very advanced in its AGB evolution, shortly before turning into a protoplanetary nebula. The cometary shapes of A in the $H$ and $J$ images and in the 0.79 $\\mu$m and 1.06 $\\mu$m HST images suggest that the core of A is not the central star, but the southern lobe of a bipolar structure. The position of the central star is probably at or near the position of component B, where the $H-K$ color has a value of 4.2. If the star is at or near B, then the components A, C, and D are likely to be located at the inner boundary of the dust shell. ", "introduction": "\\object{IRC +10 216} (\\object{CW Leo}) is the nearest and best--studied carbon star and one of the brightest infrared sources. It experiences a strong mass loss at a rate of $\\dot{M} \\approx 2-5\\times10^{-5}$M$_{\\odot}\\,$yr$^{-1}$ (see e.g.\\ Loup et~al.\\ \\cite{LoupForveilleEtAl93}). The central star of \\object{IRC +10 216} is a long--period variable star (LPV) with a period of approximately 649\\,days (Le~Bertre \\cite{LeBertre92}). Recent distance estimates of 110~pc to 135~pc (Groenewegen \\cite{Groenewegen97}) and 150~pc (Crosas \\& Menten \\cite{CrosasMenten97}) were reported. \\object{IRC +10 216}'s initial mass can be expected to be close to 4\\,M$_\\odot$ (Guelin et~al.\\ \\cite{GuelinForestiniEtAl95}, Weigelt et~al.\\ \\cite{WeigeltBalegaEtAl98}). The bipolar appearance of the nebula around this object was already reported by Christou et~al.\\ (\\cite{ChristouRidgwayEtAl90}) and Kastner \\& Weintraub (\\cite{KastnerWeintraub94}). The non-spherical structure is consistent with the conjecture that \\object{IRC +10 216} is in a phase immediately before entering the protoplanetary nebula stage. High--resolution observations of this object and its circumstellar dust shell were reported by McCarthy et~al.\\ (\\cite{McCarthyMcLeodEtAl90}), Christou et~al.\\ (\\cite{ChristouRidgwayEtAl90}), Dyck et~al.\\ (\\cite{DyckBensonEtAl91}), Danchi et~al.\\ (\\cite{DanchiBesterEtAl94}), Osterbart et~al.\\ (\\cite{OsterbartBalegaEtAl97}), Weigelt et~al.\\ (\\cite{WeigeltBalegaEtAl97}, \\cite{WeigeltBalegaEtAl98} \\cite{WeigeltBloeckerEtAl99}), Skinner et~al.\\ (\\cite{SkinnerMeixnerEtAl98}), and Haniff \\& Buscher (\\cite{HaniffBuscher98}). The results of Dyck et~al.\\ (\\cite{DyckBensonEtAl91}) and Haniff \\& Buscher (\\cite{HaniffBuscher98}) showed that the structure of the dust shell of \\object{IRC +10 216} has been changing for some years. Detailed radiative transfer calculations for \\object{IRC +10 216} were recently performed by Ivezi\\'{c} \\& Elitzur (\\cite{IvezicElitzur96}), Crosas \\& Menten (\\cite{CrosasMenten97}), and Groenewegen (\\cite{Groenewegen97}) using a large amount of spectroscopic and visibility data. The aim of this paper is to discuss the properties of the inner dust shell of \\object{IRC +10 216} on the basis of a series of high--resolution observations. In a second paper (Men'shchikov et~al., in prep.) we will present a detailed two--dimensional radiative transfer model for this object. ", "conclusions": "We have presented high-resolution $J$--, $H$--, and $K$--band observations of \\object{IRC +10 216} with the highest resolution so far at $H$ of 70~mas. A series of $K$--band images from five epochs between October 1995 and November 1998 shows that the inner nebula is non-stationary. The separations of the four dominant resolved components increased within the 3 years by up to $\\sim35\\%$. For the two brightest components a relative velocity within the plane of the sky of about 23~mas/yr or 14~km/s was found. Within these 3 years the rather faint components C and D became brighter whereas component B faded. The general geometry of the nebula seems to be bipolar. We find that the most promising model to explain the structures and changes in the inner nebula is to assume that the star is at or near the position of component B. The star is then strongly but not totally obscured at $H$ and $K$. Consistently component B is very red in the $H-K$ color while A and the northern $J$--band components are relatively blue. The polarization pattern with strong polarization in the northern arms and also a significant polarization in the peak supports this model. The inner nebula and the apparent motions seem to be rather symmetric around this position and the observed changes are consistent with the assumption of an enhanced mass loss becoming apparent at least in 1997. \\object{IRC\\,+10\\,216} is without doubt in a very advanced stage of its AGB evolution. The observed bipolarity of its dust shell even reveals that it has possibly entered the phase of transformation into a protoplanetary nebula." }, "0003/hep-th0003186_arXiv.txt": { "abstract": "Calculations of reaction rates for the third-order QED process of photon splitting \\teq{\\gamma\\to\\gamma\\gamma} in strong magnetic fields traditionally have employed either the effective Lagrangian method or variants of Schwinger's proper-time technique. Recently, Mentzel, Berg and Wunner \\cite{mbw94} presented an alternative derivation via an S-matrix formulation in the Landau representation. Advantages of such a formulation include the ability to compute rates near pair resonances above pair threshold. This paper presents new developments of the Landau representation formalism as applied to photon splitting, providing significant advances beyond the work of \\cite{mbw94} by summing over the spin quantum numbers of the electron propagators, and analytically integrating over the component of momentum of the intermediate states that is parallel to field. The ensuing tractable expressions for the scattering amplitudes are satisfyingly compact, and of an appearance familiar to S-matrix theory applications. Such developments can facilitate numerical computations of splitting considerably both below and above pair threshold. Specializations to two regimes of interest are obtained, namely the limit of highly supercritical fields and the domain where photon energies are far inferior to that for the threshold of single-photon pair creation. In particular, for the first time the low-frequency amplitudes are simply expressed in terms of the Gamma function, its integral and its derivatives. In addition, the equivalence of the asymptotic forms in these two domains to extant results from effective Lagrangian/proper-time formulations is demonstrated. ", "introduction": "\\label{sec:intro} The third-order quantum electrodynamical process of photon splitting \\teq{\\gamma\\to\\gamma\\gamma} in a strong magnetic field, currently popular in several astrophysical models of different neutron star sources, was first studied over three decades ago. Due to analytic complexities encountered when investigating this interaction, it was not until the beginning of the 1970s that a body of correct and uncontroversial results emerged. These early splitting calculations used either effective Lagrangian \\cite{Adl70,Bial70,Adler71} or variations of Schwinger's proper-time techniques \\cite{PR72,Stone79,bms86}, the expediency of which yielded compact analytic forms for the rates \\teq{R} when specializing to low energy (\\teq{R\\propto \\omega^5}) or low field (\\teq{R\\propto B^6}) cases. After a hiatus of nearly two decades, photon splitting became of interest again in the literature \\cite{bms96,as96,bh97apj,wbm98} following the publication of an S-matrix calculation in the Landau representation of its rates by Mentzel, Berg and Wunner\\cite{mbw94}, specifically because of their contention that the earlier works cited above had seriously underestimated the strength of this process. The rates computed in \\cite{mbw94} were later retracted in \\cite{ww97}, with a sign error in their numerical coding having been discovered and corrected. Mentzel et al.'s analytic derivation was the first comprehensive presentation of the application of a Landau representation technique specifically to magnetic photon splitting, though the QED formalism presented by Melrose and Parle \\cite{mpII,mpIII} virtually provided an equivalent enunciation of such S-matrix forms for splitting amplitudes. More recently, Weise, Baring \\& Melrose \\cite{wbm98} confirmed the analytic derivation of \\cite{mbw94}. The Landau representation calculations and most of the earlier effective Lagrangian and proper-time presentations were generally applicable to non-dispersive regimes below the pair creation threshold (\\teq{\\hbar\\omega =2mc^2}), where the momentum vectors of the initial and final photons are collinear, and arbitrary field strengths. Below pair threshold, the effective Lagrangian approach of \\cite{Adl70,Bial70,Adler71} and the proper-time calculations in \\cite{PR72,Stone79,bms86,bms96} appear much more amenable for the purposes of numerical evaluation than the S-matrix formulation in the Landau representation. This arises because effective Lagrangian and proper-time (collectively referred to by the label ELP here) methods produce results that involve triple integrals over relatively simple (hyperbolic and exponential) functions, while the S-matrix amplitudes integrate over the parallel momentum \\teq{p_z} and include a triple summation over the Landau level quantum numbers of the intermediate pair states. Both techniques start from different but equivalent \\cite{mpI} forms of the electron propagator, and hence S-matrix computations \\cite{mbw94,wbm98} should yield identical results to proper-time numerics \\cite{Adl70,Bial70,Adler71,PR72,Stone79,bms86,bms96}. For the specific case of magnetic pair creation \\teq{\\gamma\\to e^+e^-}, such an equivalence of the S-matrix and proper-time methods has been demonstrated\\cite{te74,dh83}, but only via continuous asymptotic approximations that smoothly average out the exact ``sawtooth'' resonance structure. Yet the S-matrix Landau representation approach explicitly retains the resonances in the scattering amplitudes above pair threshold, whereas the ELP methods eliminate such information early during developments. Photon splitting becomes effectively first-order in \\teq{\\fsc} at any one of a multitude of pair resonances, generated when the intermediate states become ``on-shell.'' Hence it is quite possible that splitting can compete effectively with pair creation as a photon absorption mechanism above pair threshold. Ascertaining whether this is true is an interesting physics question. Moreover, if splitting is approximately as probable as pair creation above threshold, then it manifestly changes the character of vacuum dispersion, so that quadratic (and by inference perhaps higher order) contributions to the vacuum polarization tensor become significant relative to the standard linear ones used in the derivation \\cite{Adler71} of kinematic selection rules for splitting. Hence, the generation of exact and compact expressions for the rates for \\teq{\\gamma\\to\\gamma\\gamma} valid both below and above pair creation threshold is clearly a worthwhile enterprise from a physics perspective. Developed expressions for the rates for photon splitting are also important for astrophysical applications of this process, particularly to effect efficient and accurate computations of such rates. These applications have so far focused on neutron star magnetospheres, primarily on models of soft gamma repeaters (SGRs) and strongly-magnetized pulsars, both being extremely topical in the astrophysics community at present. The potential importance of splitting in neutron star environments was suggested by \\cite{Adler71,mitrof86,b88}. Possible formation of splitting cascades has been explored in models of SGR transient outbursts as a means of softening the spectrum efficiently with no production of pairs \\cite{b95apjl,bh95apss,hb96aip,ccfh96aip}. If both polarizations can split, or if polarization switching is active during SGR outbursts, then the properties of the splitting cross-section guarantee emergent spectra in the observed range (20--150 keV) and of the observed shape for all fields in excess of around \\teq{10^{14}}Gauss \\cite{b95apjl,bh95apss,hb96aip}, provided that the emission region is not concentrated near the polar cap. The spectral properties of SGRs in quiescent emission appear to be distinct from those during outburst. Pulsations and temporal increases of their periods (i.e. spin-down) have now been observed \\cite{kouv99nat,hurl99nat,kouv99apjl} for two of the four confirmed SGRs (SGR 1806-20 and SGR 1900+14), leading to inferences of fields in the vicinity of \\teq{10^{15}}Gauss. The connection between these pulsars of extremely high magnetization, so-called {\\it magnetars}, and conventional radio/X-ray/gamma-ray pulsars is not well-understood. Baring \\& Harding \\cite{bh98apjl} postulated that radio quiescence, a property of the SGRs, may be common in magnetars due to the efficient action of photon splitting and other effects in suppressing the creation of pairs. Photon splitting also has spectral implications for such pulsars with more modest fields: \\cite{hbg97apj} demonstrated that the unusual absence of \\teq{>30}MeV emission in the gamma-ray pulsar PSR 1509-58 (whose spin-down field is \\teq{\\sim 3\\times 10^{13}}Gauss) can naturally be explained by the operation of \\teq{\\gamma\\to\\gamma\\gamma} in the intense magnetic and gravitational fields near its surface. Several desirable goals are immediately identifiable on the basis of this historical path for the study of the physics of photon splitting, and the needs of the astrophysics community. It would be satisfying (i) to obtain analytic expressions for rates that are valid above pair creation threshold using the Landau representation methodology, (ii) to know whether the analytic formalism of Mentzel et al. \\cite{mbw94} can be developed and simplified, and (iii) to demonstrate a formal equivalence between this S-matrix Landau representation approach and extant results from proper-time/effective Lagrangian techniques. This paper addresses these issues, using the verified analytic formalism of Mentzel et al. as the starting point for mathematical developments. The analysis here considers all the polarization modes that are permitted by the CP invariance symmetry (\\teq{\\pll}, \\teq{\\ppp} and \\teq{\\lpl}), and applies for collinear momenta of the incoming and outgoing photons, i.e. when the effects of vacuum dispersion are neglected. A significant development provided in this paper is the dramatic simplification incurred by algebraically performing the summation over the spin states that are incorporated in the electron propagators. The resulting expressions in Section~\\ref{sec:reduce} (first stated in \\cite{wbm98}) are relatively compact, and of an appearance familiar to Landau representation/S-matrix theory applications to magnetized environments (i.e. including associated Laguerre functions). Furthermore, here the integrations over the momentum parallel to the field are performed analytically for the first time in Section~\\ref{sec:pzint}, rendering the splitting rates in most amenable forms (see Eqs.~[\\ref{eq:totratefin}] and [\\ref{eq:Mform}]) that are optimal for numerical applications: the analytic forms presented consist of just triple summations over Landau level quantum numbers of the intermediate states. These general results are valid both below and above pair threshold at non-resonant photon energies, and provide substantial advances over the work of \\cite{mbw94}; they are much more suitable for numerical evaluation since many cancellations have been eliminated algebraically. Two specializations are discussed in Section~\\ref{sec:limits}, primarily to (partially) demonstrate equivalence of the Landau representation formalism presented here with extant proper-time/effective Lagrangian limiting forms for splitting rates, and simultaneously to serve as a check on the mathematical manipulations of this paper. Results are presented for all three polarization modes permitted by CP invariance in the limit of zero dispersion. The first asymptotic regime is (see Section~\\ref{sec:Bgg1}) for highly supercritical fields, \\teq{B\\gg B_{\\rm c}=m^2c^3/e\\hbar}, where in the case of \\teq{\\pll}, the limit was found to concur with a recent analytic result that was obtained by Baier et al. \\cite{bms96}, while new results were obtained for the other two modes. In the second specialization, in Section~\\ref{sec:omegall1}, asymptotic results for energies \\teq{\\omega\\ll mc^2} well below pair creation threshold were obtained, reproducing the cubic energy dependence of the amplitudes obtained by other QED techniques. Moreover, new and compact expressions for the scattering amplitudes in this low energy limit are derived in terms of the logarithm of the Gamma function, its integral and their derivatives. These simplified forms in Eqs.~(\\ref{eq:Mperptoparparwll1}) and~(\\ref{eq:Mperptoperpperpwll1}) are also produced from extant integral forms for splitting matrix elements derived first in \\cite{Adl70,Adler71}, thereby facilitating the first analytic demonstration of the equivalence of splitting rates obtained by the S-matrix formulation in the Landau representation and those derived using Schwinger-type techniques. ", "conclusions": "\\label{sec:conclusion} This paper has provided a detailed development of the S-matrix formulation of the QED process of magnetic photon splitting in the Landau representation, focusing on the case of zero dispersion where photon propagation is collinear. The formalism in Section~\\ref{sec:formalism} rederives and extends the exposition of Mentzel, Berg \\& Wunner \\cite{mbw94}. The two principal general developments offered here are an analytic reduction via the summation over the spins of the intermediate pair states, discussed briefly in \\cite{wbm98}, and the analytic integration over the momenta parallel to the field incorporated in the electron propagators. This latter accomplishment is presented here for the first time. The cumulative product of these developments is a satisfyingly simple and elegant form in Eq.~(\\ref{eq:Mform}) for the scattering amplitude for each of the polarization modes permitted by CP invariance. These amplitudes possess products of generalized Laguerre polynomials that are common to QED processes in external magnetic fields, and elementary functions involving the photon energies and the various pair thresholds associated with the propagators. Moreover, the analytic forms presented consist of just triple summations over Landau level quantum numbers of the intermediate states, and are eminently suitable for accurate numerical computations both below and above pair creation threshold \\teq{\\omega =2}. The applicability of these results to regimes above pair threshold is a benefit of the S-matrix expansion in the Landau representation that is not afforded by effective Lagrangian and proper-time calculations: while these (latter) Schwinger-type techniques elegantly formulate splitting rates below pair threshold, they eliminate the resonance structure early on in their mathematical developments, a severe limitation above \\teq{\\omega =2}. As an embellishment to these general results, specializations in two significant domains have been obtained. The first is for highly supercritical fields, \\teq{B\\gg 1}, reproducing in particular the result of \\cite{bms96} for the \\teq{\\pll} mode, and deriving new results for the other two modes permitted by CP invariance in the limit of zero dispersion. The second group of asymptotic results are for energies \\teq{\\omega\\ll 1} well below pair creation threshold, where new and compact expressions for the scattering amplitudes have been derived in Eqs.~(\\ref{eq:Mperptoparparwll1}) and~(\\ref{eq:Mperptoperpperpwll1}) in terms of the logarithm of the Gamma function, its integral and their derivatives. These two domains of specialization herein have facilitated the first analytic demonstration of the equivalence of splitting rates obtained by the S-matrix formulation in the Landau representation and those derived using Schwinger-type effective Lagrangian/proper-time techniques." }, "0003/hep-ex0003014_arXiv.txt": { "abstract": " ", "introduction": " ", "conclusions": "Astronomy using very high energy particles ($\\gamma$-rays, cosmic rays, and neutrinos) is a diverse and rapidly developing field. It is currently a field largely driven by experimental results where significant scientific progress can be made on a relatively short time scale. For example, in the last few years there have been several exciting discoveries. Using $\\gamma$-rays, we are probing remarkable and unexpected phenomena in objects such as active galaxies and gamma ray bursts. We are also searching for the origins of the cosmic radiation. At the very highest energies, we are discovering cosmic ray particles that probably should not be there, but are. Discoveries in this field often raise as many questions as they answer. For the future, there will be an expanding interest in this field, both to understand astrophysics under extreme conditions and to search for evidence of physics beyond the standard models of elementary particles and cosmology. Future next-generation experiments in space and on the ground will greatly expand our discovery horizon. For $\\gamma$-rays, the new projects include more powerful atmospheric Cherenkov telescopes and a new $\\gamma$-ray satellite, GLAST. Larger air shower detectors, such as Fly's Eye HiRes and Auger, will explore questions relating to the highest energy cosmic rays. In the world of neutrinos, the AMANDA experiment has demonstrated the ability to detect neutrinos produced in the atmosphere. Future experiments, such as IceCube and ANTARES, will greatly expand the possibility of detecting high energy astrophysical neutrino sources. We can only hope that our overall knowledge of high energy particles from the Universe advances at the same rate as new instruments are being constructed. \\bigskip I wish to acknowledge the help and encouragement of many people in the particle physics and astrophysics communities. The contributions of the following people were particularly important: Katsushi Arisaka, Steve Barwick, Michael Catanese, Corbin Covault, Jim Cronin, Francis Halzen, Charles Jui, Tadashi Kifune, Peter Leonard, Eckart Lorenz, John Matthews, Masaki Mori, Gus Sinnis, David Smith, Simon Swordy, Steve Ritz, Pierre Sokolsky, Masahiro Takeda, Masahiro Teshima, Trevor Weekes, and Heinz V\\\"olk. I also thank the organizers of the Lepton-Photon Symposium (especially John Jaros, Helen Quinn, and Michael Peskin) for their encouragement and patience. Any inaccuracies are my fault alone. This research is supported in part by the National Science Foundation." }, "0003/astro-ph0003322_arXiv.txt": { "abstract": "A theoretical light curve is constructed for the quiescent phase of the recurrent nova U Scorpii in order to resolve the existing distance discrepancy between the outbursts ($d \\sim 6$ kpc) and the quiescences ($d \\sim 14$ kpc). Our U Sco model consists of a very massive white dwarf (WD), an accretion disk (ACDK) with a flaring-up rim, and a lobe-filling, slightly evolved, main-sequence star (MS). The model properly includes an accretion luminosity of the WD, a viscous luminosity of the ACDK, a reflection effect of the MS and the ACDK irradiated by the WD photosphere. The $B$ light curve is well reproduced by a model of 1.37 $M_\\odot$ WD $+$ 1.5 $M_\\odot$ MS (0.8---2.0 $M_\\odot$ MS is acceptable) with an ACDK having a flaring-up rim, and the inclination angle of the orbit $i \\sim 80 \\arcdeg$. The calculated color is rather blue ($B-V \\sim 0.0$) for a suggested mass accretion rate of $2.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$, thus indicating a large color excess of $E(B-V) \\sim 0.56$ with the observational color of $B-V = 0.56$ in quiescence. Such a large color excess corresponds to an absorption of $A_V \\sim 1.8$ and $A_B \\sim 2.3$, which reduces the distance to 6---8 kpc. This is in good agreement with the distance estimation of 4---6 kpc for the latest outburst. Such a large intrinsic absorption is very consistent with the recently detected period change of U Sco, which is indicating a mass outflow of $\\sim 3 \\times 10^{-7} M_\\odot$ yr$^{-1}$ through the outer Lagrangian points in quiescence. ", "introduction": "U Scorpii is one of the best observed recurrent novae, the outbursts of which were recorded in 1863, 1906, 1936, 1979, 1987, and the latest in 1999. Especially, the 1999 outburst was well observed from the rising phase to the cooling phase by many observers (e.g., \\cite{mun99}; \\cite{kah99}; \\cite{lep99}) including eclipses (Matsumoto, Kato, \\& Hachisu 2000). Based on Matsumoto et al.'s (2000) observation, Hachisu et al. (2000) have constructed a theoretical light-curve model for the 1999 outburst of U Sco and obtained various physical parameters of the recurrent nova. Their main results are summarized as follows: (1) A direct light-curve fitting of the 1999 outburst indicates a very massive white dwarf (WD) of $M_{\\rm WD}= 1.37 \\pm 0.01 M_\\odot$. (2) The envelope mass at the optical maximum is estimated to be $\\Delta M \\sim 3 \\times 10^{-6} M_\\odot$. (3) Therefore, the mass accretion rate of the WD is $\\dot M_{\\rm acc} \\sim 2.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$ during the quiescent phase between 1987 and 1999. (4) An optically thick wind blows from the WD and plays a key role in determining the nova duration because it reduces the envelope mass (\\cite{kat94}). About 60\\% of the envelope mass is carried away in the wind, which forms an expanding shell as observed in T Pyx (e.g., \\cite{shr89}). The residual 40\\% ($1.2 \\times 10^{-6} M_\\odot$) is added to the helium layer of the WD. (5) As a result, the WD can grow in mass at an average rate of $\\sim 1 \\times 10^{-7} M_\\odot$ yr$^{-1}$. \\par The above physical pictures are exactly the same as proposed by Hachisu et al. (1999b) as a progenitor system of Type Ia supernovae (SNe Ia). However, the distance to U Sco is still controversial because the direct light-curve fitting results in a relatively short distance of $\\sim 6$ kpc (\\cite{hac2000}), which is incompatible with the distance of $\\sim 14$ kpc at the quiescent phase (e.g., \\cite{web87}; \\cite{war95}; \\cite{kah99}, for a summary). If the distance of $\\sim 14$ kpc is the case, it could be hardly consistent with the results (1) to (5) mentioned above. \\par Our purpose in this Letter is to construct a light-curve model for the quiescent phase and to rectify the distance to U Sco. Our numerical method to obtain light curves has been described both in Hachisu \\& Kato (1999) to explain the second peak of T CrB outbursts and in Hachisu et al. (2000) to reproduce the light curve for the 1999 outburst of U Sco. Therefore, we mention only new parts of our numerical method in \\S 2. In \\S 3, by directly fitting our theoretical light curve to the observations, we derive the distance to U Sco. Discussions follow in \\S 4, especially in relation to the recently detected orbital-period change of U Sco and a systemic mass loss through the outer Lagrangian points. We also discuss the relation to a progenitor system of SNe Ia. ", "conclusions": "Matsumoto et al. (2000) observed a few eclipses during the 1999 outburst and, for the first time, detected a significant period-change of $\\dot P / P = (-1.7 \\pm 0.7) \\times 10^{-6}$ yr$^{-1}$. If we assume the conservative mass transfer, this period change requires a mass transfer rate of $\\gtrsim 10^{-6} M_\\odot$ yr$^{-1}$ in quiescence. Such a mass transfer for 12 years is too high to be compatible with the envelope mass on the white dwarf, thus implying a non-conservative mass transfer in U Sco. \\par We have estimated the mass transfer rate for a non-conservative case by assuming that matter is escaping from the outer Lagrangian points and thus the specific angular momentum of the escaping matter is $1.7 a^2 \\Omega_{\\rm orb}$ (\\cite{saw84}; \\cite{hac99a}), where $a$ is the separation and $\\Omega_{\\rm orb} \\equiv 2 \\pi /P$. Then the mass transfer rate from the companion is $\\dot M_{\\rm MS}= (-5.5 \\pm 1.5) \\times 10^{-7} M_\\odot$ yr$^{-1}$ for $M_{\\rm MS}= 0.8$---2.0 $M_\\odot$ under the assumption that the WD receives matter at a rate of $\\dot M_{\\rm acc} = 2.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$. The residual ($\\sim 3 \\times 10^{-7} M_\\odot$ yr$^{-1}$), which is escaping from the system, forms an excretion disk outside the orbit of the binary. Such an extended excretion disk/torus may cause a large color excess of $E(B-V)= 0.56$. \\par Kahabka et al. (1999) reported the hydrogen column density of (3.1---4.8)$\\times 10^{21}$ cm$^{-2}$, which is much larger than the Galactic absorption in the direction of U Sco (1.4$\\times 10^{21}$ cm$^{-2}$, \\cite{dic90}), indicating a substantial intrinsic absorption. It should also be noted here that Barlow et al. (1981) estimated the absorption toward U Sco by three ways: (1) the Galactic absorption in the direction of U Sco, $E(B-V) \\sim 0.24$ and $A_V \\sim 0.7$, (2) the line ratio of He~II during the 1979 outburst ($t \\sim$ 12 days after maximum), $E(B-V) \\sim 0.2$ and $A_V \\sim 0.6$, and (3) the Balmer line ratio during the 1979 outburst ($t \\sim$ 33---34 days after maximum), $E(B-V) \\sim 0.35$ and $A_V \\sim 1.1$. The last one is significantly larger than the other two estimates. They suggested the breakdown of their case B approximation in high density regions. However, we may point out another possibility that the systemic mass outflow from the binary system has already begun at $t \\sim$ 33 days and, as a result, an intrinsic absorption is gradually increasing. \\par The mass of the companion star can be constrained from the mass transfer rate. Such a high transfer rate as $\\dot M_{\\rm MS} \\sim 5.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$ strongly indicates a thermally unstable mass transfer (e.g., \\cite{heu92}), which is realized when the mass ratio is larger than 1.0---1.1, i.e., $q= M_{\\rm MS}/ M_{\\rm WD} >$ 1.0---1.1 for zero-age main-sequence stars (\\cite{web85}). This may pose a requirement $M_{\\rm MS} \\gtrsim 1.4 M_\\odot$. We estimate the most likely companion mass of 1.4---1.6 $M_\\odot$ from equation (11) in Hachisu et al. (1999b). \\par If the distance to U Sco is $\\sim$ 6.0---8.0 kpc, it is located $\\sim$ 2.3---3.0 kpc above the Galactic plane ($b=22\\arcdeg$). The zero-age masses of the progenitor system to U Sco are rather massive (e.g., $8.0 ~M_\\odot + 2.5 ~M_\\odot$ from Hachisu et al. 1999b) and it is unlikely that such massive stars were born in the halo. Some normal B-type main-sequence stars have been found in the halo (e.g., PG0009+036 is located $\\sim$ 5 kpc below the Galactic disk, \\cite{smt96}), which were ejected from the Galactic disk because of their relatively high moving velocities $\\sim$100---200 km s$^{-1}$. The radial velocity of U Sco is not known but it is suggested that the $\\gamma$-velocity is $\\sim$50---100 km s$^{-1}$ from the absorption line velocities (\\cite{joh92}; \\cite{sch95}). If so, it seems likely that U Sco was ejected from the Galactic disk with a vertical velocity faster than $\\sim 20$ km s$^{-1}$ and has reached at the present place within the main-sequence lifetimes of a $\\sim 3.0 M_\\odot$ star ($\\sim 3.5\\times 10^8$ yr). \\par Now, we can understand the current evolutionary status and a further evolution of U Sco system. The white dwarf has a mass $1.37 \\pm 0.01 M_\\odot$. It is very likely that the WD has reached such a large mass by mass accretion. In fact the WD is currently increasing the mass of the helium layer at a rate of $\\dot M_{\\rm He} \\sim 1.0 \\times 10^{-7} M_\\odot$ yr$^{-1}$ (\\cite{hac2000}). We then predict that the WD will evolve as follows. When the mass of the helium layer reaches a critical mass after many cycles of recurrent nova outbursts, a helium shell flash will occur. Its strength is as weak as those of AGB stars because of the high mass accretion rate (\\cite{nom82}). A part of the helium layer will be blown off in the wind, but virtually all of the helium layer will be burnt into carbon-oxygen and accumulates in the white dwarf (\\cite{kat99h}). Therefore, the WD mass can grow until an SN Ia explosion is triggered (\\cite{nom84})." }, "0003/astro-ph0003114_arXiv.txt": { "abstract": "We calculate for the first time the probability density functions (PDFs) $P$ of burst energy $e$, duration $T$ and inter-burst interval $\\tau$ for a known turbulent system in nature. Bursts in the earth-sun component of the Poynting flux at 1 AU in the solar wind were measured using the MFI and SWE experiments on the NASA WIND spacecraft. We find $P(e)$ and $P(T)$ to be power laws, consistent with self-organised criticality (SOC). We find also a power law form for $P(\\tau)$ that distinguishes this turbulent cascade from the exponential $P(\\tau)$ of ideal SOC, but not from some other SOC-like sandpile models. We discuss the implications for the relation between SOC and turbulence. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003264_arXiv.txt": { "abstract": "We calculate light-curves for Gamma-Ray Burst afterglows when material ejected in the explosion is confined to a jet which propagates in a medium with a power-law density profile. The observed light-curve decay steepens by a factor of $\\Gamma^2$ when an observer sees the edge of the jet. In a uniform density medium the increase in the power-law index ($\\beta$) of the light-curve as a result of this {\\it edge effect} is $\\sim0.7$ and is completed over one decade in observer time. For a pre-ejected stellar wind ($\\rho \\propto r^{-2}$) $\\beta$ increases by $\\sim0.4$ over two decades in time due to the edge effect and the steepening of the light-curve due to the jet sideways expansion takes about four decades in time. Therefore, a break in the light-curve for a jet in a wind model is unlikely to be detected even for very narrow jets of opening angle of a few degrees or less, in which case the lateral expansion occurs at early times when the afterglow is bright. The light-curve for the afterglow of GRB 990510, for which an increase in $\\beta$ of approximately 1.35 was observed on a time scale of 3 days, cannot be explained only by the sideways expansion and the edge effects in a jet in a uniform ISM -- the increase in $\\beta$ is too large and too rapid. However, the passage of the cooling or synchrotron peak frequencies through the observing band at about 0.1 -- 1 day together with jet edge effect explains the observed data. The jet opening angle is found to be $\\sim 5^o$ and the energy in the explosion to be less than about $10^{50}$ erg. ", "introduction": "In a recent paper Chevalier and Li (1999) pointed out that some of the GRB afterglow light-curves are best modeled when the density of the circum-burst medium is taken to fall off as $r^{-2}$ (this is referred to as the {\\it wind model}). These afterglows show no evidence for a jet, i.e. their light-curves follow a power-law decline without any break. This is puzzling since collimated outflows are expected in the collapsar model for GRBs (MacFadyen, Woosley \\& Heger 2000). We offer a possible explanation for this puzzle by showing that the light-curve resulting from the interaction of a jet with a pre-ejected wind falls off as a power-law whose index changes very slowly with time. We carry out a detailed modeling of the multi-wavelength afterglow flux data for GRB 990510, which provides the best evidence for a jet propagation in a uniform density medium (Harrison et al. 1999, Stanek et al. 1999), to show that effects associated with a finite jet opening-angle are insufficient to explain the observed rapid steepening of the light-curve. In \\S\\ref{dynamics} we calculate the propagation of a jet in a stratified medium and in \\S\\ref{synchrotron} we describe the calculation of the synchrotron emission and afterglow light-curve. ", "conclusions": "One of the main results of this work is to show that afterglows from well collimated Gamma-Ray Burst remnants going off in a medium with density decreasing as $r^{-2}$ show little evidence for light-curve steepening due to jet edge and sideways expansion. This could explain the lack of breaks in the afterglows of GRB 980326 and GRB 980519, which Chevalier \\& Li (1999) found to offer support for the {\\it wind model}. Jets can perhaps be detected by the measurement of time dependent polarization. In a collimated outflow the sharpest break in the light-curve is produced in a uniform density circum-stellar medium, and is associated with the edge of the jet coming within the relativistic beaming cone (the {\\it edge effect}). The magnitude of this break is $\\sim0.7$ (0.4) for a uniform ISM (wind model) and occurs over about 1 decade (2 decades) in time. Further steepening of the light-curve, associated with the sideways expansion of the jet, occurs on a much longer time scale of $R_{t_e}\\sim$10$^2$ (10$^4$), i.e. weeks to months. The power-law index for the light-curve of GRB 990510 increased between days 0.8 and 3 by about 1.35. This is too large and too fast to result from jet edge \\& sideways expansion effects. However, the observations can be explained if either the cooling or the synchrotron peak frequency passed through the observing band at about 1 or 0.1 day, respectively. Models that are consistent both the optical and radio data of this afterglow have an opening angle of $\\sim 5^o$ and energy in the explosion is $\\lta 10^{50}$ erg (see Figure 3). For the afterglow of GRB 990123 the power-law index of the light-curve increased by 0.55 between days 1.5 and 3, which can be explained by the {\\it edge effect} alone (\\Meszaros \\& Rees 1999)." }, "0003/astro-ph0003052_arXiv.txt": { "abstract": "We present simulations of the propagation of magnetized jets. This work differs from previous studies in that the cross-sectional distributions of the jets's state variables are derived from analytical models for magneto-centrifugal launching. The source is a magnetized rotator whose properties are specfied as boundary conditions. The jets in these simulations are considerably more complex than the ``top-hat''constant density etc. profiles used in previous work. We find that density and magnetic field stratification (with radius) in the jet leads to new behavior including the separation of an inner jet core from a low density collar. We find this {\\it jet within a jet} structure, along with the magnetic stresses, leads to propagation behaviors not observed in previous simulation studies. Our methodology allows us to compare MHD jets from different types of sources whose properties could ultimately be derived from the behavior of the propagating jets. ", "introduction": "Highly collimated supersonic jets are a ubiquitous phenomena occurring in many astrophysical environments. These jets are observed propagating from sources as diverse as Active Galactic Nuclei (AGN, \\cite{Leahy91}), Young Stellar Objects (YSOs, \\cite{Reipurth97}) and Planetary Nebulae (PNe, \\cite{SokLiv94}). While considerable progress has been made in understanding the nature of jets from AGN and YSOs, there remains considerable debate concerning the nature of the more recently discovered PNe jets. The ubiquity of jets in astrophysics has made them a popular subject for study. They are excellent laboratories for the study of basic astrophysical processes (shocks, instabilities, etc.). Their long dynamical or ``look-back'' times, $t_{dyn} = L_j/V_j$, also make them ideal astrophysical fossils for studying the evolution of the obscured and often unobservable central sources, \\ie there is the hope in jet studies that the physics of the central engine can be revealed by studying the exhaust. Given the diversity of jet producing environments there also exists the hope that an underlying unity can be be found in terms of the fundamental processes which create jets. Articulating these processes is one of the critical issues facing astrophysical jet studies. Accretion disks are believed to play a key role in the physics of both YSOs and AGN. In-falling, rotating matter is stored in these disks until dissipation allows material to spiral inward and feed the central, gravitating object. Both YSO and AGN disks are believed to support strong, well ordered magnetic fields. The current consensus holds that these fields are the agents for producing jets in a process known as {\\it Magneto-centrifugal} launching. In this mechanism, plasma in the disk is loaded on to co-rotating field lines. If conditions in the disk are favorable (\\ie field strength and orientation) the plasma is centrifugally flung outward along the field lines. Strong toroidal field components are generated in the flow as the field is dragged backwards by the plasma inertia leading to collimation of the wind into into a narrow jet. We note, however, that the external medium might also help focus the outflow. Magneto-centrifugal launching has been studied in detail by many authors both analytically (\\cite{HevNor89}, \\cite{Pudritz91}, \\cite{Shuea94}, \\cite{ostriker}, \\cite{Leryea99}) and through numerical simulations (\\cite{OP97a}, \\cite{Romanova98}, \\cite{Kudohea98}). In the YSO community two principle flavors of the Magneto-centrifugal launching model exist. The first is a pure disk wind model (\\cite{Pudritz91}) in which the jet is generated at the surface of a Keplerian disk. The second, called ``X-Winds'' (\\cite{Shuea94}), produces a jet from the boundary layer between the disk and the central star's magnetosphere. Other models exist as well (\\cite{Goodsonea97}) and there remains considerable debate as too which mechanisms are obtained in real YSO flows. While there is an exhaustive literature concerning jet {\\it launching} and {\\it collimation}, there has also been considerable study of jet {\\it propagation}. Propagation studies focus on scales many orders of magnitude larger (\\cite{Reipurth97}) than the region where collimation occurs. For example in the work of (\\cite{OP97a}) the collimation of the jet was followed out to a height above the disk of $H = 80 R_i$ where $R_i$ is the inner disk radius. Since $R_i \\le 10 R_*$ ($R_*$ is the stellar radius, \\cite{Hartmann98}), the scale of the simulation was at least 10 times smaller than the smallest scales on which jets have been resolved and at least $10^3$ times smaller than the typical scale of observational jet studies. Much of the propagation work has been numerical and for both YSOs and AGN much of it has been have been purely hydrodynamic. For YSOs only a handful of MHD studies of jet propagation have been carried out to date (\\cite{Todo92}, \\cite{Cerqueira98}, \\cite{Frankea98}, \\cite{Cenquria99}, \\cite{Gardinerea99}, \\cite{StoneHardee99}). If, however, strong magnetic forces produce the jets then these forces should effect their propagation downstream. Unless the fields are somehow removed, Maxwell stresses should alter at least some characteristics of the jet's propagation. Recently \\cite{Frankea99} have shown that ambipolar diffusion may be operative in YSO jets in some part of the flow. However the time-scales involved are such that changes in jet magnetic fields will only occur for parsec-scale jets. Flows on time-scales less than $\\tau \\approx 10^3$ y will not lose their fields. In the case of AGN, the ambipolar time-scales are even larger. Thus, a proper accounting for the MHD forces in the propagation of both YSO and AGN jets is needed. In this paper we focus mainly on YSOs but our results will be applicable to AGN jets as well. To date all radiative MHD jet simulations of steady, constant density ``top-hat'' jets have been performed using simple field geometries. \\cite{Cenquria99} showed that jets with purely poloidal $\\vec{B} = B_z \\bf{k}$ topologies did not have propagation characteristics which differed significantly from pure hydrodynamic jets. \\cite{Gardinerea99} have also found similar results for pulsed ``top-hat'' MHD jets with poloidal fields. \\cite{Frankea98} however, found that if the field had a strong toroidal ($B_\\phi$) component then the jet head could be strongly effected by the Maxwell stresses leading to the production of so-called ``nose-cones''. Nose-cones form when post-shock gas is restricted from lateral expansion by the axially directed ``hoop stresses'' associated with strong toroidal fields. Instead of back-flowing to form a cocoon, the shocked gas is confined to the head of the beam in the region downstream of the jet-shock. The hoop stresses lead to a conical streamlined configuration for the head i.e. a nose-cone. Such structures were also seen in the early MHD simulations of AGN jets (\\cite{Lindea89}). In \\cite{Frankea98} the addition of radiative losses, appropriate for YSO jets, caused the nose-cones to narrow significantly. In a more extensive set of calculations \\cite{StoneHardee99} found that MHD effects on jet propagation is strongly dependent on initial field topology. While these results were promising, there still remains considerable distance to be traveled in the study of MHD jets. The principle issue that must be addressed is that all the simulations carried out to date is the use of ad-hoc field topologies. Unless a force-free configuration is adopted, $J\\times{\\bf B} = 0$, Maxwell stresses will act on the jet beam independent of propagation effects. Thus some effort must be expended in developing equilibrium configurations for MHD jet simulation initial conditions. With little to guide them, all modelers have chosen simple topologies which allow for a simple specification of the required equilibrium. Frank \\ea (1999) used a pure toroidal geometry. \\cite{Gardinerea99} used a pure poloidal geometry. \\cite{Cenquria99} used both toroidal and poloidal as well as force free helical configurations which had to extend throughout the entire computational domain (jet + ambient medium). \\cite{StoneHardee99} used helical pressure matched beams in a variety of configurations. None of the configurations used in these papers deviated from the simple constant velocity, constant density model for the jet beam. These efforts were necessary for articulating the basic role of MHD forces in jets, but they do not help establish a connection between conditions in the jet and the protostellar source (a protostar and rotating magnetized accretion disk). What is needed for use by the broader community is to begin the simulations with jet cross-sections derived directly from magneto-centrifugal flow models. That is the goal of the work presented here. In what follows we present models of MHD jet propagation with initial configurations in the jet taken directly from the solution of force balance perpendicular (the Grad-Shafranov equation) and parallel (the Bernoulli equation) to magnetic surfaces generated by a magnetized rotator. Our simulations follow the evolution of jets composed of helical fields embedded in hypersonic plasmas whose density and velocity vary with radius. Thus our models constitute a further step towards realism in the theoretical description of magnetized astrophysical jets. The goal of this paper is to articulate the basic physics which can occur in these kinds of jets and to look for differences between the propagation of jets forming from different kinds of rotators. We note that the parameter space of solutions is quite large and in this paper we present only the first results of this project. In future papers we will present a more systematic exploration of parameter space. The plan of the paper is as follows. In section 2 we describe the methods used to construct the initial equilibria and numerically simulate the flows. In section 3 we present results of our simulations focusing on adiabatic, isothermal and radiative cases. The next section compares the results with observations. Finally, in section 5 we present and discuss our conclusions. ", "conclusions": "We have carried out a series of simulations intended to address the issue of MHD jet propagation. Whereas previous studies have used ad-hoc initial conditions we inject flows into our computational grid derived from models of collimated jets driven by magneto-centrifugal launching. This strategy allows us to compare the propagation characteristics of jets driven by different types of outflows. In particular we have studied the propagation of jets driven by: (1) a purely Keplerian rotator (a disk) exterior to a solid body rotator (a star); (2) a Keplerian rotator with a sub-Keplerian boundary layer both of which are exterior to a solid body rotator. The former model we refer to as a {\\em Keplerian} jet, the latter is called a {\\em Multi-component} model. Our simulations follow the jets out to observable scales. In the Keplerian jet simulations the jet radius is $R_j = 1.5x10^{15}$ cm making the grid extend out to $3000 ~AU$. For the multi-component jet $R_j$ is almost a factor of ten larger and the the grid extends out $.1 ~pc$. The width of the multi-component jet is interesting in that it yields a model with a very narrow, dense core (a jet) surrounded by a wider lower density outflow. Both models were calculated under the assumption that the jets are launched under isothermal conditions. We have carried out simulations of the propagation of both {\\em Keplerian} and {\\em Multi-component} jets under adiabatic, isothermal and radiative conditions in order to determine the behavior of the resulting flows with, and without, radiative losses. We note again that our adiabatic and isothermal simulations have low magneto-sonic Mach numbers $M_{ms} = 2 ~-~ 4$. While these values are small compared with the values used in previous numerical studies of MHD jets $(M_{ms} >10$, \\cite{StoneHardee99}) they is quite similar to what has been obtained in other studies MHD collimation of jets ($M_{ms} \\approx 3$,\\cite{cam182}). Our simulations show significant differences in the propagation characteristics for the two types of rotators. In addition, features are seen in both classes of jet which have not been seen in previous models of either pure hydrodynamic or MHD jet propagation. In all cases it appears that the most important aspect of the flow behavior seen in the simulations can be traced back to the annular stratification of the jets. In particular, the radial distributions of density, velocity, and toroidal magnetic field appear to be the principle causes of the new behavior seen in the simulations. Both {\\em Keplerian} and {\\em Multi-component} jets exhibit a core/collar structure such that a high density core region exists near the axis surrounded by one or more lower density annuli (collar) extending out to the jet boundary. The strongest toroidal fields exist at the boundary between the core and collar. Since the momentum in the core is higher than that in the collar the propagation characteristics of the jets are dominated by the core pulling ahead of the collar. The strong field surrounding the core ensures that the two regions remain fairly distinct in terms of their dynamics. As the jets propagate we see the core acting as a {\\it jet within a jet}. In the {\\em Keplerian} case the relatively low density contrast between core and collar keeps the two propagating at relatively similar velocities. The stratification of the magnetic fields produces strong dynamical differences between core and collar. All plasma flowing into the cocoon comes from the lower field strength regions of the collar. In the {\\em Multi-component} case there exists an extremely low density inner collar (which also has higher velocity than the surrounding regions) and this leads to a complete separation of core and collar. The ``peel-off'' of the collar in the {\\em Multi-component} models is quite dramatic and occurs in both the adiabatic and isothermal simulations. Our results have bearing on a number of issues. The simplest conclusion that can be drawn is that the structure imposed on a YSO jet by the launching and collimation process can lead to fairly complex propagation characteristics. Thus our models build on and extend the previous works which utilized only ``top-hat'' jets as initial conditions. Our results also indicate that jets launched from different classes of rotators may have different propagation characteristics. It is likely that in real jet systems the dynamics is too complex to make an isomorphic identification of a given class of rotators with a set of observed jet morphologies. There is however the possibility that as these kinds of studies mature one might be able to distinguish between different classes of MHD launching models via consideration of the way the jets from these models would appear on the sky. Finally we note that given the large parameter space of initial conditions for both the Given Geometry Model and for the jet propagation simulations, the work described here which focuses only on two instances must be seen as preliminary. It does however point to the fact that the jets produced by magnetized rotators are likely to be more complex in their structure and, furthermore, that this complexity will be reflected in the observed jet morphologies. In future studies we will attempt to build a larger catalog of jet propagation characteristics through a more thorough exploration of parameter space of the Given Geometry Model." }, "0003/astro-ph0003028_arXiv.txt": { "abstract": "We consider the dynamics of a protostellar disk surrounding a star in a circular-orbit binary system. Our aim is to determine whether, if the disk is initially tilted with respect to the plane of the binary orbit, the inclination of the system will increase or decrease with time. The problem is conveniently formulated in the binary frame in which the tidal potential of the companion star is static. We may then consider a steady, flat disk that is aligned with the binary plane and investigate its linear stability with respect to tilting or warping perturbations. The dynamics is controlled by the competing effects of the $m=0$ and $m=2$ azimuthal Fourier components of the tidal potential. In the presence of dissipation, the $m=0$ component causes alignment of the system, while the $m=2$ component has the opposite tendency. We find that disks that are sufficiently large, in particular those that extend to their tidal truncation radii, are generally stable and will therefore tend to alignment with the binary plane on a time-scale comparable to that found in previous studies. However, the effect of the $m=2$ component is enhanced in the vicinity of resonances where the outer radius of the disk is such that the natural frequency of a global bending mode of the disk is equal to twice the binary orbital frequency. Under such circumstances, the disk can be unstable to tilting and acquire a warped shape, even in the absence of dissipation. The outer radius corresponding to the primary resonance is always smaller than the tidal truncation radius. For disks smaller than the primary resonance, the $m=2$ component may be able to cause a very slow growth of inclination through the effect of a near resonance that occurs close to the disk center. We discuss these results in the light of recent observations of protostellar disks in binary systems. ", "introduction": "The existence of disks around young stars was spectacularly confirmed by direct images from the {\\it Hubble Space Telescope\\/} ({\\it HST\\/}) (McCaughrean \\& O'Dell 1996; Burrows et~al. 1996). Observations suggest that young stars are usually found in binary systems and that young binaries typically interact strongly with the disks that surround the stars (Ghez, Neugebauer, \\& Matthews 1993; Mathieu 1994; Osterloh \\& Beckwith 1995; Jensen, Mathieu, \\& Fuller 1996). There is growing evidence that disks within a binary are sometimes inclined with respect to the binary orbital plane. Such a case may have been seen in {\\it HST\\/} and Keck images of a disk in the young binary HK Tau (Stapelfeldt et~al. 1998; Koresko 1998). Suppose that a protostellar disk surrounds a star in a circular-orbit binary system, and that the disk is tilted with respect to the binary orbital plane. The evolution of the disk is affected by the tidal field of the companion star, as has been considered by Papaloizou \\& Terquem (1995). Some features of their analysis were confirmed in three-dimensional numerical simulations by Larwood et~al. (1996). The basic physics involved may be summarized as follows (see also Bate et al. 2000). In a non-rotating frame of reference centered on the star about which the disk orbits, the companion star orbits at the binary frequency $\\Omega_{\\rm b}$ and exerts a time-dependent tidal torque on the disk. This torque may be decomposed into a steady component and an oscillatory component with a frequency of $2\\Omega_{\\rm b}$, and their effects may be considered separately. Consider first the steady torque. If the disk were composed of non-interacting circular rings, the steady torque would cause each ring to precess, about an axis perpendicular to the binary plane, at a rate that depends on the radius of the ring, resulting in a rapid twisting of the disk. However, if the disk is able to maintain efficient radial communication, whether by wave propagation, viscosity, or self-gravitation, it may be able to resist this differential precession by establishing an internal torque in the disk. This can be arranged so that the net torque on each ring is such as to produce a single, uniform precession rate. However, to establish this internal torque, the disk must become warped. The concomitant dissipation changes the total angular momentum of the disk, tending to bring it into alignment with the binary plane in addition to causing accretion. Consider now the oscillatory torque. Applied to a single ring, this would cause a modulation of the precession rate and also a nutation (Katz et~al. 1982). However, in the presence of radial communication, the oscillatory torque drives a bending wave (with azimuthal wavenumber $m=1$) in the disk. Papaloizou \\& Terquem (1995) showed that, if the wave is subject to dissipation, it too may change the total angular momentum of the disk and tend to increase its inclination. The net effect of the steady and oscillatory torques determines whether an initially coplanar disk will acquire a tilt over time or whether an initially inclined disk will evolve towards coplanarity. The purpose of this paper is to determine this outcome, which could provide clues to the origin of misaligned disks in systems such as HK Tau. The basic mechanism suggested by Papaloizou \\& Terquem for generating a tilt by the oscillatory torque can be related to earlier work by Lubow (1992), who showed that an aligned, Keplerian disk in a circular binary may be linearly unstable to tilting if it contains a local resonance at which the orbital angular velocity $\\Omega(r)$ satisfies \\begin{equation} \\Omega=\\left({{m_*}\\over{m_*-2}}\\right)\\Omega_{\\rm b}. \\label{resonance} \\end{equation} Here $m_*$ is the azimuthal wavenumber of the component of the tidal potential that is involved in the instability cycle. The cycle works through a mode-coupling process as follows: given a perturbation with $m=1$ (a tilt), the tidal potential interacts with it to a drive a wave with $m=m_*-1$ at the resonant radius. This in turn interacts with the tidal potential to produce a stress with $m=1$, which can influence the tilt. The role of dissipation is subtle, since some dissipation is required to provide a change in the angular momentum of the disk if instability is to occur, yet the associated damping can compete with the intrinsic growth rate of the instability. In particular, if the disk extends to the $3:1$ resonance ($\\Omega=3\\Omega_{\\rm b}$) it may be unstable to tilting through the $m=3$ component of the tidal potential. This resonance has the smallest $m_*$ for which equation (\\ref{resonance}) can be satisfied (for a prograde disk) and is the closest resonance to the central star. It is difficult for disks to extend even as far as the $3:1$ resonance, because of the effects of tidal truncation (Paczy\\'nski 1977; Papaloizou \\& Pringle 1977). Superhump binary disks might extend to the $3:1$ resonance because of their extreme binary mass ratios, the secondary companion having less than $1/5$ the mass of the primary about which the disk orbits (see the review by Osaki 1996). This instability is related to, and occurs at the same position as, the eccentric instability that is believed to be responsible for superhumps in cataclysmic variable disks (Lubow 1991). However, the growth rate is invariably much smaller for tilting than for eccentricity, and the weak tilt instability may be suppressed by the effects of viscous damping and accretion (Murray \\& Armitage 1998). The same instabilities had been previously identified in the context of planetary rings for higher $m_*$ (Goldreich \\& Tremaine 1981; Borderies, Goldreich, \\& Tremaine 1984). More fundamentally, free-particle orbits undergo even stronger, parametric instabilities at these resonant locations (Paczy\\'nski 1977), although free particles fail to model properly the behavior of a fluid disk at resonances. We relate this theory to the suggestion of Papaloizou \\& Terquem (1995) by noticing in equation (\\ref{resonance}) that, for $m_*=2$, a near resonance is obtained in the inner part of the disk where $\\Omega\\gg\\Omega_{\\rm b}$. Indeed, Papaloizou \\& Terquem rely on the $m=2$ component of the tidal potential to drive an $m=1$ bending wave in the tilted disk. The resulting response is a slowly rotating $m=1$ bending wave, with frequency $2\\Omega_{\\rm b}$ in the inertial frame. Such a wave is close to resonance in the inner part of a nearly Keplerian disk because of the near coincidence of the effective wave driving frequency $\\Omega-2\\Omega_{\\rm b}$ and the frequency of vertical oscillations $\\Omega_z\\approx\\Omega$; this is indeed the origin of equation (\\ref{resonance}) with $m_*=2$. An additional resonant effect occurs owing to the near coincidence of the driving frequency and the epicyclic frequency of horizontal oscillations $\\kappa\\approx\\Omega$. We describe the instability cycle associated with the oscillatory torque as a mode-coupling process in Fig.~1. However, because the resonance is not exact, and because of the importance of resonantly induced horizontal motions, a proper treatment requires a distinct analysis from that of Lubow (1992). \\placefigure{fig1} In this paper, we therefore examine whether a flat, aligned disk in a binary is linearly unstable to tilting even if it does not extend to the $3:1$ resonance. This problem is most conveniently analyzed in the binary frame where the tidal potential is static, since the disk can then be considered to be steady and to admit normal modes. These modes do not have a pure azimuthal wavenumber because the disk is non-axisymmetric as a result of tidal distortions. However, the tilting instability, if present, may be expected to appear as a modification of the rigid-tilt mode, which is trivial in the absence of the companion star. This mode may be followed continuously as the mass of the companion is increased, in order to determine whether it acquires a net rate of growth or decay. In general, the analysis of a normal mode of a tidally distorted disk is very difficult owing to the non-axisymmetric distortions of the disk. We therefore adopt the following simple approach, which is appropriate when only $m=1$ bending waves are involved. We start by writing down the reduced equations for linear bending waves in a protostellar disk subject to an axisymmetric external potential (Section~2). These can be derived formally without great effort (see the Appendix). We then give a physical interpretation of these equations and use this insight to see how to modify them in the presence of a non-axisymmetric potential (Sections~3 and~4). We present a simple disk model (Section~5) and describe the results of numerical calculations of normal modes (Section~6). Some further analysis illuminates the underlying physics and helps to explain the numerical results (Section~7). Finally, we summarize our findings (Section~8). ", "conclusions": "In this paper, we have considered the linear stability of a coplanar protostellar disk that surrounds a star in a circular-orbit binary system. We have determined whether a slight tilt introduced into the disk would grow or decay in time. The outcome depends on the size of the disk. For disks that are truncated by standard tidal torques, typically resulting in an outer disk radius of about $0.3$ times the binary separation, we find that the disk tilt generally decays in time. For smaller disks, tilt growth is possible. As seen in Fig.~2, a disk undergoes a strong, `primary' resonance with the tidal field when its outer radius is a certain fraction of the binary separation. This characteristic radius, which we denote by $r_{\\rm p}$, is approximately $0.118$ times the binary separation for the parameters we have considered (see Table~1), but would be smaller still for thinner disks with $H/r<0.1$. In such a resonance, the disk experiences a growing tilt and becomes significantly warped (see Fig.~6). This resonance occurs when the frequency of the lowest-order global bending mode in the disk matches the tidal forcing frequency, which is here twice the binary orbital frequency. Weaker resonances occur at a series of discrete resonances corresponding to radii greater than $r_{\\rm p}$. There is also a near resonance that occurs close to the disk center. For disks smaller than radius $r_{\\rm p}$, this resonance causes a very slight tilt growth if $\\alpha$ is sufficiently small (but non-zero), and any initial tilt would be retained. For disks with radii larger than $r_{\\rm p}$, including disks truncated by standard tidal torques, the tilt will decay on approximately the viscous time-scale of the disk, or roughly $10^3$ binary orbits for $\\alpha=0.01$ (see Fig.~5). For disks with large tilts, nonlinear effects may shorten the time-scale to reach small tilts, perhaps to the precessional time-scale of the disk, or about $20$ binary orbits (Bate et~al. 2000). The net outcome of growth or decay of the disk tilt is determined by the competition of two torques. As seen in the inertial frame, the tidal torque acting on a tilted disk may be decomposed into a steady component and an oscillatory component with twice the binary orbital frequency. The steady torque, resulting from the $m=0$ component of the tidal field, causes the disk to become aligned with the binary orbit in the presence of dissipation, while the oscillatory torque, resulting from the $m=2$ component of the tidal field, causes misalignment. The steady torque produces an intuitively simple result because it causes the disk to settle to a state of coplanarity, where it experiences a minimum tidal potential energy, as a result of dissipation. The effect of the oscillatory torque is somewhat counterintuitive, but can be understood in terms of a mode-coupling model (see Fig.~1). Provided that $\\alpha$ is sufficiently small, the oscillatory torque slightly dominates for smaller disks because material in such disks is generally closer to the near resonance that occurs in the vicinity of the disk center (see eq.~[\\ref{resonance}]). A major issue is the origin of the tilt in observed protostellar disks. In the case of HK Tau, the disk surrounds the secondary star, but the two stars are similar in spectral type (Monin, M\\'enard, \\& Duch\\^ene 1998). Although there are considerable uncertainties in the system parameters, the disk could extend to its standard tidal truncation radius, as suggested by Stapelfeldt et~al. (1998). In that case, the results of this paper imply that tidal effects may cause decay of the primordial tilt, but in any case would not cause tilt growth. On the other hand, the existence of the tilt means that the decay time-scale cannot be much shorter than the binary age, estimated as $5\\times10^5$~yr. This places some constraints on both the theory and the binary parameters, although there are considerable uncertainties. For example, consider the case that the binary separation is close to its projected value of 340 AU. For $\\alpha=0.01$, the linear tilt decay time-scale (based on Figure 2) would be several times longer than the estimated system age. On the other hand, the nonlinear decay time-scale estimate of Bate et al (2000) suggests a decay time-scale substantially shorter than the estimated age. The nonlinear time-scale estimate would be more compatible with a somewhat larger binary separation. The predicted shape of a tilted, tidally truncated disk with $H/r\\approx0.1$ is not strongly warped (see Fig.~7), in accord with the observations (Stapelfeldt et~al. 1998; Koresko 1998). The lack of an observed warp cannot be used as evidence against binarity. On the other hand, a slight warp does occur for thinner disks such as in Fig.~7 case b, which could be observed as a small asymmetry. Note that the decay time-scale of proper bending modes of this disk (based on Table 3) is of order $10^4$~yr if $\\alpha=0.01$, much shorter than the linear tilt decay time-scale. If the disk were tilted and warped in an arbitrary way as a result of its formation process, we would expect it to evolve rapidly to a tilted but essentially unwarped shape, then the tilt itself would decay on a longer time-scale. However, the nonlinear effects discussed by Bate et~al. (2000) are likely to speed up both stages considerably. Similar considerations apply to a recent observational test of coplanarity among a sample of T Tauri binaries by Donar, Jensen, \\& Mathieu (2000). The data show some evidence for approximate coplanarity on a statistical basis. It is possible that some tidal evolution of the tilt towards coplanarity may be have occurred, if the tilt decays as rapidly as a disk precessional time-scale. It is important to understand whether disk truncation could occur close to the resonant radius $r_{\\rm p}$, so that the disk would be unstable to tilting. For disks with a substantial tilt, Terquem (1998) has shown that a disk of radius close to $r_{\\rm p}$ is sometimes subject to a strong resonant torque that is parallel to its spin axis. This resonant torque can exceed the viscous torque in the disk for sufficiently small values of $\\alpha$, $\\alpha\\la10^{-3}$. If this torque could truncate an initially tilted disk at radius $r_{\\rm p}$, the disk might become strongly warped (as seen in Fig.~6) and tilted further. The disk radius would be less than half of the standard tidal truncation radius. However, it is unclear that this torque would lead to disk truncation at $r_{\\rm p}$ because it is smoothly distributed over the disk rather than being concentrated near $r_{\\rm p}$. This is because the resonance is global rather than local. The lack of a strong warp in HK Tau argues against this process in that system. Disks in cataclysmic binaries are expected to be much colder than protostellar disks, having a smaller value of $H/r$. Consequently, such disks are even less likely to be unstable to tilting as a result of the $m=2$ component of the tidal field (see eq.~[\\ref{stab}]). In several X-ray binaries, most notably Her X-1, there is evidence for a tilted, precessing disk (see Wijers \\& Pringle 1999 and references therein). The tilting mechanism we have described is very unlikely to operate in such disks, which are expected to be tidally truncated and to have $H/r\\ll0.4$. Therefore, it appears that tidal torques are not responsible for the tilting of disks in X-ray binaries (cf.~Larwood 1998). Possible mechanisms for tilting these disks include wind torques (Schandl \\& Meyer 1994) and radiation torques (Wijers \\& Pringle 1999). Another possible application of this work is to nearly Keplerian disks that surround black holes in active galactic nuclei. If the disk is subject to a bar potential from the galaxy and the disk radius is sufficiently smaller than the corotation radius of the bar, then the disk will be subject to this tilt instability. The results in this paper have implications to protostellar disks perturbed by inclined planets. A secular resonance occurs where the precession frequency of a planet matches the local precession frequency of an orbiting particle. The resonant radius changes as the nebula disperses and the resonance sweeps across a major portion of the solar nebula (Ward 1981). However, the current results suggest that the effects of such resonances on the gaseous nebula are mild and are distributed over the disk. Further analysis can be carried out through the methods described in this paper." }, "0003/astro-ph0003358_arXiv.txt": { "abstract": "How does the clustering of galaxies depend on their inner properties like morphological type and luminosity? We address this question in the mathematical framework of marked point processes and clarify the notion of luminosity and morphological segregation. A number of test quantities such as conditional mark--weighted two--point correlation functions are introduced. These descriptors allow for a scale--dependent analysis of luminosity and morphology segregation. Moreover, they break the degeneracy between an inhomogeneous fractal point set and actual present luminosity segregation.\\\\ Using the Southern Sky Redshift Survey~2 (\\citealt{dacosta:southern}, SSRS2) we find both luminosity and morphological segregation at a high level of significance, confirming claims by previous works using these data {}\\citep{benoist:biasing,willmer:southern}. Specifically, the average luminosity and the fluctuations in the luminosity of pairs of galaxies are enhanced out to separations of 15\\hMpc. On scales smaller than 3\\hMpc\\ the luminosities on galaxy pairs show a tight correlation. A comparison with the random--field model indicates that galaxy luminosities depend on the spatial distribution and galaxy-galaxy interactions. Early--type galaxies are also more strongly correlated, indicating morphological segregation. The galaxies in the PSCz catalog {}\\citep{saunders:pscz} do not show significant luminosity segregation. This again illustrates that mainly early--type galaxies contribute to luminosity segregation. However, based on several independent investigations we show that the observed luminosity segregation can not be explained by the morphology--density relation alone. ", "introduction": "The geometrical properties of the large--scale structure in the Universe are a common test for cosmic structure formation theories. However, comparisons between analytical models and observational data suffer from the fact, that theoretical predictions refer to mass correlations whereas in galaxy catalogs only luminous matter is observed. This gap gives rise to the bias problem and is usually filled using biasing schemes. Mostly, these schemes relate properties of the density contrast field to the distribution of the galaxies, thus combining descriptors of a random field with point process characteristics. Due to the nature of the {\\em dark} matter, only indirect methods are feasible to address the bias problem empirically. In this line of thought, it seems promising to ask whether the clustering properties of galaxies depend on their mass, luminosity or morphological type. The idea behind this search for {\\em luminosity and morphology segregation} is that different galaxy subpopulations may trace the dark matter distribution on a different level.\\\\ Empirical investigations concerned with this problem were mainly carried out in two directions: \\begin{itemize} \\item The two--point correlation function was calculated for a series of volume--limited subsamples from galaxy surveys. A difference in the amplitude of the two--correlation function between such samples was interpreted as an indication of luminosity or morphology--segregation. For luminosity segregation see e.g., {}\\citet{ostriker:luminosity,hamilton:evidence,tenreiro:multidimensional,benoist:biasing,willmer:southern}. The void probability and cross--correlation functions have been used by {}\\citet{maurogordato:void} and {}\\citet{valotto:dependence}. For morphology segregation see e.g., {}\\citet{tenreiro:morphology-segregation,hermit:morphology-segregation}. These investigations are sensitive to segregation effects on scales roughly between 1 and 10\\hMpc. However, {}\\citet{coleman:fractal} gave an alternative explanation of the rising amplitude in terms of a fractal galaxy distribution, without any luminosity--dependent clustering. \\item {}\\citet{dressler:galaxy} showed that in clusters of galaxies the morphological type of a galaxy is depending on the local (surface) density; this is called the morphology--density relation. For mainly spherical clusters, where the local density is closely related to the radial distance from the cluster center, this translates into the {}\\citet{butcher:evolution} effect. For more recent accounts of the morphology--density relation see {}\\citet{caon:morphology-segregation,dressler:evolution,andreon:morphologyIII}. Most of these investigations focussed on the morphology--density relation {\\em inside} clusters, hence on scales smaller than 1.5\\hMpc. But the morphology--density relation can be observed also in groups of galaxies {}\\citep{postman:morphology-density,maia:groups} and for dwarf galaxies in the field {}\\citep{binggeli:abundance}. \\end{itemize} With the first method, one compares two--point correlation functions, whereas with the second, one considers the relation between the local number density and the local morphology, i.e., a comparison of one--point densities. Both methods rely on unweighted descriptors. The observations of luminosity segregation or the mor\\-phology--density relation were complemented by theoretical considerations. Motivated by the offset between the galaxy--galaxy and the cluster--cluster correlation functions, {}\\citet{kaiser:onspatial} and {}\\citet{bardeen:gauss} suggested that clusters may be understood as peaks in the density field. Starting from a Gaussian random field they showed how the amplitude of the correlation function increases with the threshold imposed on the initial density field, i.e., with the height of the peaks in the density field. This also provided an explanation for the morphology--density relation {}\\citep{evrard:morphology}.\\\\ Other authors developed a conceptual framework to describe the bias (see e.g., {}\\citealt{coles:galaxy}, {}\\citealt{dekel:stochastic}, and refs.\\ therein). Within these biasing schemes, characteristics of the galaxy point pattern are connected with descriptions of the density field -- often the mass density contrast and the galaxy over--density are compared. The relation is assumed to be (non--) linear and either deterministic or stochastic {}\\citep{dekel:stochastic}. More involved biasing schemes were considered to facilitate the extraction of reasonable galaxy catalogs from $N$-body simulations (see e.g., {}\\citealt{kates:highres}, {}\\citealt{weiss:highres}, {}\\citealt{kauffmann:galaxy}). In this paper, we introduce a new method to handle the bias problem. Our approach complements both the more observational methods and the analytical and theoretical treatments. We understand the galaxies with their intrinsic properties as a realization of a marked point process. Using conditional weighted correlation functions, we put an intermediate step in between the pure point process statistics and the statistics of random fields. In our description stochasticity is present from the very beginning. It provides us with stochastic models which enable us to exclude certain families of models for the luminosity distribution of galaxies. \\noindent More precisely, the aim of our paper is twofold: On the one hand, we want to clarify the notion of luminosity/morphology--dependent clustering by discussing this task in the mathematical framework of marked point processes (Sect.~\\ref{sect:maths}). This allows us to introduce a new class of indicators sensitive to luminosity segregation (Subsect.~\\ref{sect:mark-weighted}) and to discuss models for marked point patterns (Subsect.~\\ref{sect:marked-poisson} and Subsect.~\\ref{sect:two-species}). Methods similar in spirit are the cross--correlation function and luminosity--weighted correlation functions considered by {}\\citet{alimi:cross-correlation}, {}\\citet{boerner:correlation}, {}\\citet{vallsgabaud:luminosity}, and {}\\citet{tegmark:observational}. Our methods allow for a study of the interplay between the spatial clustering and the luminosity and morphology distribution of the galaxies, complementing the characterization of the purely spatial distribution of the galaxies. On the other hand, we address the empirical question, whether the luminosities or morphological types of galaxies depend on their spatial distribution by analyzing the SSRS2 catalog {}\\citep{dacosta:southern} in Sect.~\\ref{sect:lum-morph-seg-gal}. Our results show a significant scale--dependent luminosity and morphological segregation. To understand the data more closely we compare our results with the random field model. The comparison with galaxy samples from the IRAS~1.2Jy {}\\citep{fisher:irasdata} and the PSCz {}\\citep{saunders:pscz} strengthens our conclusions. In Sect.~\\ref{sect:usual} we will discuss the usual way of looking for luminosity segregation via the amplitude of the correlation function in the framework of marked point processes. The criticism by {}\\citet{coleman:fractal} is reviewed and we show that this degeneracy between a fractal spatial distribution and luminosity segregation is not encountered if one uses the mark--correlation functions we proposed. This strengthens the conclusions of our empirical work in Sect.~\\ref{sect:lum-morph-seg-gal}. Investigations inside clusters of galaxies gave clear evidence for the morphology--density relation {}\\citep{dressler:galaxy}. In Sect.~\\ref{sect:morphology-density} we however show that the observed luminosity segregation may not be explained by the spatial interaction of early-- and late--type galaxies {\\em alone}. Luminosity segregation is already present in the subsample consisting only of early--type galaxies. In Sect.~\\ref{sect:summary} we summarize and provide an outlook. Technicalities concerning the estimation of mark--correlation functions are left to Appendix {}\\ref{sect:estimators}. ", "conclusions": "\\label{sect:summary} The investigation of luminosity and morphology segregation of galaxies has been a scientific task for many years. Our results allow for a new perspective and suggest that both the {\\em methodology} and the {\\em physical interest} should shift slightly. Methodologically, we discussed luminosity and morphological segregation in the framework of marked point processes. This perspective provides us with a unifying view on morphology and luminosity segregation. Moreover, the mathematical theory of marked point processes provides us with test quantities and models to be compared with the data. In this line we discussed the mark--weighted conditional correlation functions. These functions are not only easy to estimate, but also offer a clear interpretation. They may be applied to a single volume limited sample, a sequence of volume limited samples is not necessary. As a consequence, they break the degeneracy between a fractal spatial structure and luminosity segregation. We suggest that the $k_{mm}$, $\\var$, and $\\cov$ functions are of special interest for a first test on luminosity segregation. Since several bias--models assume scale--dependent bias, we need quantities like $k_{mm}$, $\\var$, and $\\cov$ which can unfold the scales at which mass or luminosity segregation is relevant. This is not possible by looking at the amplitude of the two--point correlation function $\\xi_{R}(r)$ alone. Moreover our method allows for a ``built in'' significance test, by randomly re--shuffling the marks. The conditional cross--correlation functions seem to be useful if mark segregation has already been shown to be present and is to be understood more closely. However, they are based on a division of the whole sample into subpopulations, a division that has to be done carefully. The conditional mark--correlation functions are rather flexible. With the peculiar velocities or the orientations of galaxies treated as marks, the conditional mark correlation functions will allow for a fresh look at the pairwise velocity dispersion and on alignment effects. Our methods can be easily extended to higher--order correlations. In a forthcoming work we will study the mark correlations using higher--order statistics as the $J$--functions ({}\\citealt{vanlieshout:j}, {}\\citealt{kerscher:regular}). Concerning the physical results, we were not only able to assess luminosity segregation as well as morphological segregation. Rather, our perspective allowed us to ask the question: What is the luminosity and morphological segregation like? Our main results obtained from the SSRS2 survey are: \\begin{itemize} \\item The average luminosity of pairs of galaxies and the fluctuations in the luminosity on each galaxy is enhanced for pairs closer than 15\\hMpc. Hence, luminosity correlations are scale--dependent, and they are significant even outside clusters of galaxies. On scales larger than 15\\hMpc\\ our results indicate that neither luminosity nor morphological segregation is present. \\item The luminosities of galaxies in pairs closer than 3\\hMpc\\ show an increased covariance -- close galaxies preferably have similar luminosities. \\item The luminosity segregation is not compatible with the random field model. Thus, the luminosity does not trace an underlying {\\em independent} random field. The luminosity of a galaxy depends on the local clustering and on interactions with other galaxies. \\item There is an interesting feature, a small peak, in $k_{mm}$, $\\var$ and $\\cov$ for galaxy pairs with a separation of approximately $10\\hMpc$, which is currently beyond an explanation. \\item We observe morphological segregation between early-- and late-- type galaxies for scales smaller than 10\\hMpc. This effect is mainly due to highly luminous galaxies. Especially the luminous early--type galaxies seem to play an important role, both for luminosity and morphology segregation. \\item The importance of early--type galaxies for luminosity segregation is confirmed by our analysis of the IRAS samples. These infrared samples exhibit a deficit in early--type galaxies and consequently show no luminosity segregation. \\item An inhomogeneous, scale--invariant galaxy distribution, but without luminosity segregation, can not account for the signal seen in $k_{mm}$, $\\var$, and $\\cov$. The lowered correlation of the dim galaxies, and the enhanced correlation of the luminous galaxies we found, explains at least in part why the amplitude of the correlation function rises if deeper, i.e., more luminous, galaxy samples are considered. \\item With several independent tests we could show that it is not possible to explain the observed luminosity segregation from the morphology--density relation alone. \\end{itemize} Nevertheless, a couple of question remain open. \\\\ Concerning the data, it seems important to confirm our results using other galaxy surveys. Also the influence of redshift space distortions and of galaxy clusters should be investigated beyond the simple error--estimates presented in Subsect.~\\ref{sect:errors} and Sect.~\\ref{sect:morphology-density}.\\\\ Our methods are directly applicable to volume--limited samples, similar to the usual way of assessing luminosity segregation, where one needs a series of volume--limited samples. Using models for the conditional mark density $\\CM_2$ or the mark--correlation functions one may determine the parameters of such models from magnitude--limited surveys directly. Similarly, the influence of mark segregation on the two-- and $N$--point correlations estimated from magnitude--limited surveys can be estimated.\\\\ Closely related is the question how strongly the deprojected two-- and $N$--point correlation functions, determined from 2-dimensional galaxy catalogs, are influenced by luminosity segregation. With models for the mark--correlations a refined Limber's equation may be constructed (see e.g., {}\\citealt{gardini:limber}). Both, the concerns about magnitude--limited surveys and deprojection formulas will be addressed in future work. In this article we focused on clarifying the mathematical framework, on the data--analysis, and on the interpretation of the observed luminosity and morphological segregation. The relation to the peak--formalism {}\\citep{bardeen:gauss} and other biasing schemes will be investigated in future work. Understanding the luminosity distribution on the galaxies from dynamical models is the major goal." }, "0003/astro-ph0003391_arXiv.txt": { "abstract": "For the eclipsing binary system CM Draconis, eclipse minimum times have been monitored with high precision between 1994 and 1999. Periodic deviations of minimum times from a linear ephemeris may indicate the presence of an orbiting third body. Individual measurements of 41 eclipse minimum times result in a standard deviation from linear ephemeris of 5.74 seconds. A power spectral analysis of the residuals reveals only one periodicity with more then 2 seconds amplitude. This feature, with a periodicity between 750 and 1050 days has an amplitude of 2.8$\\pm$0.5 seconds, and is also present with similar phases if the power spectral analysis is performed independently for primary and secondary eclipses. It would be compatible with a circumbinary planet of 1.5 -3 Jupiter masses at an orbital distance of 1.1-1.45 AU to the binary barycenter. The assignation of a planet to the CM Dra system can however only be upheld if this periodicity can be followed in future observations for several years. For low-mass eclipsing binary stars, the method of eclipse minimum timing allows one to reach mass limits for the detection of third bodies well below that feasible by radial velocity measurements. ", "introduction": "It has long been known that the presence of a third body orbiting both components of an eclipsing binary system will offset the binary from a common binary/third-mass barycenter thereby causing a periodic shift in the observed times of the binary eclipses. The amplitude of this shift is given by \\[ \\delta T = M_\\mathrm{P} a_\\| / M_\\mathrm{B} c ,\\] where $M_\\mathrm{P}$ is the third body's mass, $a_\\| = a \\sin i$ the third body's semi-major axis along the line of sight, $M_\\mathrm{B}$ the mass of the binary system, and $c$ the speed of light. As pointed out by \\cite{schn95} and \\cite{doyl98}, eclipse timings with a precision of a few seconds could detect the presence of an orbiting Jovian-mass object around a low-mass eclipsing binary system. Also, \\cite{doyl98} gives a sample of 250 eclipsing binaries for which Jupiter-mass planets may be detectable by such studies. In this paper we report on an analysis of eclipse timings that were obtained as part of a photometric search for extrasolar planetary transits undertaken during the six years 1994 - 1999 around the M4.5/M4.5 binary CM Dra by the TEP project (\\cite{obspap} --~further TEP1~--~; \\cite{TEP2} --~further TEP2~--~) ", "conclusions": "The power spectra (Fig.\\,2) indicate that there are no periodic O-C minimum time variations with amplitudes of larger then 3-4 seconds present, for all periods less than 2000 days. This absence of amplitude variations allows us to \\emph{exclude} the presence of very massive planets around the CM Dra system, as indicated by the hatched region in the search-space diagram (Fig.\\,3). Excluding the peak around 1000 days period, the power spectra from all data (lower panel of Fig.\\,2) is relatively flat with an amplitude of about 2 seconds. This white-noise like flatness indicates an intrinsic imprecision in our data of about 2 seconds. This is most likely the results of the precision of the eclipse minima times being limited by the photometric noise of the eclipse lightcurves. O-C deviations of about 2 seconds constitute therefore a lower detection limit. Finally, the peak in the power spectra between 750 and 1100 days with an amplitude of $2.5 \\pm 0.5 $ seconds and a good match of phases from primary and secondary eclipses may be the consequence of a third body, but is close to the observational noise. If this amplitude variation is caused by a third body, it would correspond to a circumbinary planet of 1.5-3 Jupiter masses at an orbital distance from CM Dra of 1.1 - 1.45 AU. We note that such a body would cause a periodic variation in the radial velocity of CM Dra with an amplitude of $65 \\pm 20 \\mathrm{m s}^{-1}$. Though sufficient precision to detect such radial velocities amplitudes has routinely been obtained in planetary detection programs, these program are always concerned with single stars. For eclipsing binaries, the mutual orbiting of the binary components causes large radial velocity amplitudes on the order of km/s, which obstruct the separation of the much smaller radial velocity amplitudes from a third body. In the case of CM Dra, the velocities of the binary components reported by \\cite{metcalfe96} are 72 and 78 km/s, and the precision of these data would only allow the separation of third body amplitudes of more than 200 m/s (\\cite{latham}). Finally, the limited time-baseline of our observations does not allow the detection of periodicities longer then about 2000 days. The absence of very heavy third bodies with periods up to a few times longer is however rather certain due to the good general adherence of the O-C times to a linear ephemeris. Influences from third bodies {\\emph within} the Solar System onto the heliocentric eclipse minimum times are not of consequence. The strongest influence, by Jupiter, causes a 12-yearly deviation, but due to the high ecliptic latitude of CM Dra ( $76.3 \\deg$) its amplitude is limited to 0.58 seconds. The \\emph{absence} of periodicities above 3-4 seconds amplitude - and the exclusion of corresponding massive planets - may be stated with certainty, even if the data analysis may not have accounted for every factor that may introduce spurious periodicities. The claim by \\cite{guin98} of a periodicity in minimum times of 70 days with an amplitude of 18 seconds, corresponding to a third body with a mass of $0.01 M_\\odot$, is clearly invalidated (see also \\cite{iauc6875}). The \\emph{presence} of apparent periodicities with $\\approx $ 3 seconds may however also be a consequence of slowly changing starspots which distort the symmetry of the eclipses. Although we have not been able to find any relevant variations in lightcurves of CM Dra through the different observing seasons 1994-1999, the possibility of starspots can not be entirely excluded. In any case, further monitoring with high precision minimum timing of the CM Dra system is needed to ascertain the continuing presence of the 700-1050 day periodicity. Fig.~\\ref{fig:figdiscover} shows the search space of exoplanets around CM Dra covered by the eclipse timing observations described here and by the observations of transits from TEP1 and TEP2. The transit observations covered coplanar planets ($\\sin i \\approx 1$) on short period orbits, between 7 days (the shortest stable orbit around CM Dra) and 60 days (as a limit where observational coverage gets sparse), with a maximum detectable periodicity of 100 days (the limit where even coplanar planets would not cause transits because of the 89.82$\\degr$ inclination of the system). We assumed a mass limit of $m/m_\\mathrm{Earth} \\approx 10$, corresponding to the lower size limit of about 2.5 Earth Radii for detections with transits. The lower mass limit from the O-C timing method is derived from the absence of amplitudes over 2.5 seconds, except between 700 and 1050 days, were a planet candidate is indicated. The two methods employed do cover rather complementary regimes: Whereas the strength of the transit method is the detection of relatively small planets on close orbits, O-C minimum timing is best for the detection of long period planets with at least Jupiter-like masses. The usefulness of the radial velocity method is limited in binary systems, though it might also lead to the discovery of massive third bodies around them. To verify the persistence of the 700-1050 day periodicity and the possibility of a planet, observations of CM Dra's eclipse minimum times need to be continued during the the next several years. \\begin{figure} \\resizebox{\\hsize}{!}{\\includegraphics{Cc172_4.ps}} \\caption{Search space covered by the O-C timing observations and by transit observations reported in TEP1, TEP2. The hatched regions are those where planets can be excluded from the eclipse minimum timing and the transit search reported in TEP2. The small rectangular gray region corresponds to the planet candidate from the power-spectra based on O-C minimum timing. Regions left blank are those where these detection methods have not had sufficient sensitivity. Also indicated is a line where an orbiting third body would cause a radial velocity variation of 50 m/s in the spectrum of the eclipsing binary.} \\label{fig:figdiscover} \\end{figure}" }, "0003/astro-ph0003450_arXiv.txt": { "abstract": "We present new modelling of the spectrophotometric properties of intermediate age stellar populations in the near-infrared (NIR). We take into account the evolutionary and spectroscopic properties of the Asymptotic Giant Branch (AGB) stars which dominate the integrated NIR emission of those populations. The predictions of spectrophotometric narrow-band molecular indices require the use of an effective temperature -- colour (or spectrum) calibration for AGB stars; synthetic indices show a strong dependence upon this calibration. Preliminary results of NIR observations of a supermassive 500\\,Myr old stellar cluster in the prototypical merger remnant NGC 7252 are presented. The spectra are indeed consistent with intermediate age models dominated by light from AGB stars, a significant fraction of which may be carbon rich. Implications for the stellar inputs used in the modelling of intermediate age stellar populations and their NIR emission are discussed. ", "introduction": "Space and ground based observations have shown that the star formation in starburst regions appears to be biased toward compact clusters. This process seems to be common to all galaxy types: merging galaxies (NGC 3597: Lutz 1991, NGC 1275: Holtzman et al. 1992, NGC 4038/39 Whitmore et al. 1999), starburst galaxies (NGC 1705: Meurer et al. 1992, M82: Gallagher \\& Smith 1999, Smith \\& Gallagher, this volume), dwarf galaxies (Henize 2-10: Conti \\& Vacca 1994, ESO338-IG04, \\\"Ostlin et al. 1998) and even barred galaxies (NGC 1097: Barth et al. 1995). Colours, luminosities and spatial properties of those clusters are consistent with their being globular clusters with effective radii of a few parsecs (Schweizer 1999). It is established that the AGB stars are the principal contributors to the near infrared (NIR) luminosity of intermediate age stellar populations\\footnote{We define intermediate age populations as those having ages between 10$^8$ and 10$^9$\\,yr.} (Persson et al. 1983). The AGB evolutionary phase translates into major changes of the HR diagram properties of stellar populations over small time scales. Traditionally, the spectroscopic diagnostic tools that are used in the NIR for age dating are principally based on CO bands (2.3$\\,\\mu$m) or broad-band colours. The accuracy of those methods is questionable. After the disappearance of massive supergiants that display very deep CO bands, it becomes hard to discriminate between AGB stellar populations and the red giant branch (RGB) stellar populations on the basis of CO bands only. Changes in the NIR colours are ambiguous because of degeneracies with the effects of metallicity and extinction. Evidence for strong colour changes due to the AGB stars is found principally in observations of the Magellanic Cloud globular clusters. However, the spectrophotometry of those clusters is affected by large stochastic fluctuations due to the small numbers of luminous cool stars. The interpretation of those observations is not very secure and larger samples (not available in the LMC) are required. Recently, Lan\\c{c}on et al. (1999) have proposed new diagnostic tools to identify the intermediate age populations using other molecular absorption bands than the CO bands. They distinguish between the oxygen rich and carbon rich AGB stars contributing to the NIR light of post-starburst populations based on the spectroscopic features of the AGB stars that are absent in the RGB star spectra. ", "conclusions": "The use of massive globular clusters in merging galaxy remnants gives us the opportunity to overcome the stochastic fluctuations affecting the NIR spectrophotometric properties of the intermediate age population due to the small number of bright AGB stars and to test, both qualitatively and quantitatively, the population synthesis models. We have observed the most massive cluster of NGC 7252 in the NIR. These observations allow us to report the first observations of a stellar population dominated by AGB stars beyond the Local Group. Comparisons with our model cluster spectra lead us to argue that the use of an effective temperature scale calibrated on static giant branch stars for TP-AGB stars is, at best, questionable. The observed spectrum is more consistent with model cluster spectra calculated using an effective temperature scale calibrated on LPV stars. The good fit obtained when a significant fraction of the 500\\,Myr old TP-AGB stars are carbon rich shows the importance of taking surface abundance evolution into account in models for the spectroscopic evolution of intermediate age populations." }, "0003/astro-ph0003385_arXiv.txt": { "abstract": "The Euler equations of ideal gas dynamics posess a remarkable nonlinear involutional symmetry which allows one to factor out an arbitrary uniform expansion or contraction of the system. The nature of this symmetry (called by cosmologists the transformation to supercomoving variables) is discussed and its origin clarified. It is pointed out that this symmetry allows one to map an explosion problem to a dual implosion problem and vice versa. The application to laboratory simulations of supernova remnants is considered; in principle this duality allows the complete three-dimensional evolution of highly structured explosion ejecta to be modelled using a static target in an implosion facility. ", "introduction": "There is much interest at present in the possible use of the new generation of high-power laser facilities (in particular the National Ignition Facility at Livermore and the Laser MegaJoule in Bordeaux) to simulate astrophysical phenomena such as supernovae. At first sight this programme appears to suffer from one obvious drawback. The phenomena one wishes to simulate generally involve {\\em explosions} while the laser facilities are designed to produce {\\em implosions}. Remarkably, as we will show, this is not a problem. Under certain, not too restrictive, conditions there exists an exact mathematical duality which allows one to transform an explosion problem to an implosion problem and vice versa. Thus it is possible, in a precise sense, to use {\\em implosion} experiments to simulate {\\em exploding} systems. ", "conclusions": "" }, "0003/astro-ph0003046_arXiv.txt": { "abstract": "It has been shown by many independent studies that the cold dark matter scenario produces singular galactic dark halos, in strong contrast with observations. Possible remedies are that either the dark matter is warm so that it has significant thermal motion or that the dark matter has strong self interactions. We have combined these ideas to calculate the linear mass power spectrum and the spectrum of cosmic microwave background (CMB) fluctuations for self-interacting warm dark matter. Our results indicate that such models have more power on small scales than is the case for the standard warm dark matter model, with a CMB fluctuation spectrum which is nearly indistinguishable from standard cold dark matter. This enhanced small-scale power may provide better agreement with the observations than does standard warm dark matter. ", "introduction": "Dark matter is a necessary ingredient in the standard Big Bang model of the universe. Its presence has an impact from subgalactic dynamics to the global evolution of the universe. However, the nature of the dark matter remains unknown. So far, the cold dark matter model has been very successful in explaining how structure forms \\cite{peacock,gross}. In this model the dark matter consists of weakly interacting massive particles (WIMPs) which are extremely non-relativistic when structure formation begins. Because they are so massive they do not free stream and perturbations on small scales are preserved. In the 1980s it was realised that CDM produces too much small-scale structure, and that some modification of the model is needed. Several possibilities exist: there could be a large component of hot dark matter damping small scale fluctuations or there could be a non-zero cosmological constant. Recent data from type Ia supernovae indeed suggest that the energy density of the universe is dominated by a cosmological constant \\cite{SNIa}. Thus, the problem with CDM is at first sight remedied. However, in the past few years very high resolution N-body simulations of structure formation have shown that any type of CDM model produces far too much substructure on galactic scales, compared with observations. The halo of a galaxy like our own should contain of the order 1000 distinct subhaloes, a factor of ten more than is found by observations \\cite{moore,ghigna}. Another, related problem is that galaxies are predicted to have singular cores. Navarro, Frenk and White \\cite{NFW96} found that N-body simulations predicted a universal core profile of halos where $\\rho \\propto r^{-1}$. Later simulations with higher resolution find an even steeper profile \\cite{FP94,N99,NS99,moore2}. At the same time galactic rotation curves indicate dark matter halos with finite cores, i.e. constant core density \\cite{paolo}. This problem is very severe and is consistently found in all simulations. If the details of star formation and feedback do not solve the problem, then physics at a more fundamental level possibly could. One option is that the primordial power spectrum has a sharp drop at subgalactic scales so that substructure is prevented from forming \\cite{KL99}. Another option along this line is that the dark matter is not cold, but warm \\cite{SS88,SLD99}. In this model the dark matter particle mass should be around 1 keV so that the dark matter has significant thermal motion and perturbations on small scales are erased. However, the cut-off scale needed for the correct core radius of halos to be produced is so large that it is difficult to form the correct number of dwarf galaxies \\cite{HD}. A radically different explanation was suggested by Spergel and Steinhardt \\cite{SS99}, namely that the dark matter could be cold, but have significant self-interactions. If the mean free path of the dark matter particles is of the order the size of the collapsing system, then the core singularity would form much more slowly, while the outer parts of the halo would remain unchanged. Recently, a large number of papers have appeared which investigate this possibility numerically \\cite{H99,burkert,firmani,yoshida,MGJPQ,ostriker}. The conclusion is that if the interactions are very strong, the model does not fit observations \\cite{burkert,firmani,yoshida,MGJPQ,ME}. The halos become completely spherical apart from a small rotational deformation, and a singular core develops. However, it seems that models where the dark matter mean free path is similar to the system size produce halos closely resembling the observed ones \\cite{burkert,firmani}. It has also been suggested that the self-interacting matter could be in the form of a scalar field \\cite{peebles}. That dark matter could have self-interactions is an old idea. It was originally suggested by Raffelt and Silk \\cite{RS} that HDM neutrinos could have strong self interactions. In this way free streaming would be suppressed and fluctuations only washed out via diffusion. The scenario was elaborated on by Atrio-Barandela and Davidson \\cite{AD} who did a numerical study of this model. The possibility of number changing self interactions has also been considered \\cite{CMH92,machacek,LSS}. In the present paper we wish to explore the possibility that dark matter has both significant thermal motion and self-interactions. The self-interactions are assumed to consist only of two-particle scattering. In general, the inclusion of self interactions leads to less small scale suppression of perturbations because the small scale cut-off in power is given by the Jeans scale which is smaller than the free-streaming scale. We find that self-interacting hot dark matter, as suggested by Refs.~\\cite{RS,AD}, is clearly ruled out because it produces far too little small-scale structure. However, self interacting warm dark matter may be a viable possibility. Strong self interactions push the power spectrum towards smaller scales by roughly a factor of 1.6, which may make it consistent with observations. ", "conclusions": "We have performed a quantitative calculation of the linear behaviour of warm dark matter models with possible self interactions. As expected, power on small scales is generally increased in self-interacting models because free streaming is suppressed. In collisionless models, power is suppressed on the free streaming scale, whereas in strongly self-interacting models the cut-off is at the Jeans scale. This increase in the amplitude of the fluctuations on small scales has the effect of pushing the cut-off in the power spectrum down to smaller scales by approximately a factor of 1.6. This may allow warm dark matter to better fit the dwarf galaxy observations for masses which are small enough to explain the core structure of dark matter halos, a result which could make warm dark matter a more viable dark matter candidate. Our CMB results indicate that, like standard warm dark matter, self-interacting warm dark matter is indistinguishable from standard cold dark matter in terms of the CMB fluctuation spectrum. Thus, it is one of the few variants on the standard model which will not be probed by future CMB experiments. Any constraints on this model must therefore come from large-scale and galactic structure considerations. For instance, analysis of high-$z$ structure like damped Ly-$\\alpha$ systems might lead to interesting constraints. Note that the cross section for scattering of dark matter particles would have to be of the order $10^{-36} {\\rm cm}^2$ in order to change the matter power spectrum significantly. This is orders of magnitude more than the cross sections typical in weak interactions, and at present there are no obvious candidates for such dark matter particles. However, it could well be that warm dark matter with relatively strong self-interactions could be in a mirror sector, in which case there are no real restrictions \\cite{mirror}." }, "0003/astro-ph0003336_arXiv.txt": { "abstract": "We discuss the early evolution of beryllium and oxygen in our Galaxy by comparing abundances of these elements for halo and disk metal-poor stars. Both, O and Be rise as we go progressively to more metal-rich stars, showing a slope $0.41\\pm 0.09$ ([Be/O] vs [Fe/H]) for stars with [Fe/H]$\\leq -1$. This relationship provides an observational constraint to the actually proposed Galactic Cosmic Ray theories. ", "introduction": "First attempts to measure beryllium abundances in metal-poor stars by Molaro \\& Beckman (1984) and Molaro, Beckman \\& Castelli (1984) demonstrated that stars in the early Galaxy formed with much lower Be abundances than in the present epoch. First detection of Be in metal-poor stars was achieved by Rebolo et al. (1988) and further studies by S. Ryan, G. Gilmore, A. Boesgaard, P. Molaro, R. J. Garc\\'\\i a L\\'opez and their respective collaborators revealed a clear linear correlation with iron. Accelerated protons and $\\alpha$-particles in cosmic rays interact with ambient CNO in ISM and create Be. According to the standard Galactic Cosmic Ray (GCR) theory, these interactions in the general ISM should have given a quadratic relation between Be and O. Alternatively, spallation of cosmic ray CNO nuclei accelerated out of freshly processed material could account for the primary character of the observed early galactic evolution of Be. Another production site is the collective acceleration by SN shocks of ejecta-enriched matter in the interiors of superbubbles. In these two cases, the evolution of Be should reflect the production of CNO from massive stars. Oxygen is mostly produced by Type II SNe while iron is produced in both, Type II and in Type Ia SNe. The fact that Type Ia SNe have longer lifetime progenitors has been commonly used to argue that oxygen must be overabundant in very old stars. Observational evidence for high [O/Fe] ratios in many metal-poor stars has been reported over the last two decades. Based on the study of [O\\,{\\sc i}] lines at 6300 and 6363 \\AA\\ in evolved stars (though the second line at 6363 \\AA\\ is not visible in very metal-poor stars and the analysis is based {\\it only on one line}), several authors have found that [O/Fe]$=0.3-0.4$ dex at [Fe/H]$< -1$ and is constant until [Fe/H]$\\sim -3$ (e.g. Barbuy 1988 and Kraft et al. 1992). In contrast with this result, oxygen abundances derived in unevolved stars using the O\\,{\\sc i} IR triplet at 7774 \\AA\\ (Abia \\& Rebolo 1989; Tomkin et al. 1992; King \\& Boesgaard 1995; and Cavallo, Pilachowski, \\& Rebolo 1997) point towards linearly increasing [O/Fe] values with decreasing [Fe/H] and reaching a ratio $\\sim 1$ for stars with [Fe/H]$\\sim -3$. This may suggest a higher production of oxygen during the early Galaxy. We discuss in this paper the comparison of these abundances with those derived from OH lines located in the near-UV part of the spectra of metal-poor stars, and their relation with beryllium abundances consistently derived from the same spectra. ", "conclusions": "" }, "0003/astro-ph0003471_arXiv.txt": { "abstract": "A theoretical light curve for the 1987 outburst of V394 Coronae Austrinae (V394 CrA) is modeled to obtain various physical parameters of this recurrent nova. We then apply the same set of parameters to a quiescent phase and confirm that these parameters give a unified picture of the binary. Our V394 CrA model consists of a very massive white dwarf (WD) with an accretion disk (ACDK) having a flaring-up rim, and a lobe-filling, slightly evolved, main-sequence star (MS). The model includes irradiation effects of the MS and the ACDK by the WD. The early visual light curve ($t \\sim 1-10$ days after the optical maximum) is well reproduced by a thermonuclear runaway model on a very massive WD close to the Chandrasekhar limit ($1.37 \\pm 0.01 ~M_\\odot$). The ensuing plateau phase ($t \\sim 10-30$ days) is also reproduced by the combination of a slightly irradiated MS and a fully irradiated flaring-up disk with a radius $\\sim 1.4$ times the Roche lobe size. The best fit parameters are the WD mass $\\sim 1.37 ~M_\\odot$, the companion mass $\\sim 1.5 M_\\odot$ ($0.8-2.0 M_\\odot$ is acceptable), the inclination angle of the orbit $i \\sim $65---68$\\arcdeg$, and the flaring-up rim $\\sim 0.30$ times the disk radius. The envelope mass at the optical peak is estimated to be $\\sim 6 \\times 10^{-6} M_\\odot$, which indicates an average mass accretion rate of $\\sim 1.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$ during the quiescent phase between the 1949 and 1987 outbursts. In the quiescent phase, we properly include an accretion luminosity of the WD and a viscous luminosity of the ACDK as well as the irradiation effects of the ACDK and MS by the WD. The observed light curve can be reproduced with a disk size of 0.7 times the Roche lobe size and a rather slim thickness of 0.05 times the accretion disk size at the rim. About 0.5 mag sinusoidal variation of the light curve requires the mass accretion rate higher than $\\sim 1.0 \\times 10^{-7} M_\\odot$ yr$^{-1}$, which is consistent with the above estimation from the 1987 outburst. These newly obtained quantities are exactly the same as those predicted in a new progenitor model of Type Ia supernovae. ", "introduction": "Type Ia supernovae (SNe Ia) are one of the most luminous explosive events of stars. Recently, SNe Ia have been used as good distance indicators which provide a promising tool for determining cosmological parameters because of their almost uniform maximum luminosities (\\cite{rie98}; \\cite{per99}). These both groups derived the maximum luminosities ($L_{\\rm max}$) of SNe Ia completely empirically from the shape of the light curve (LCS) of nearby SNe Ia, and assumed that the same $L_{\\rm max}$--LCS relation holds for high red-shift SNe Ia. To be sure of any systematic biases, the physics of SNe Ia must be understood completely. By far, one of the greatest problems facing SN Ia theorists is the lack of a real progenitor (e.g., \\cite{liv99} for a recent review). Finding a reliable progenitor is urgently required in SN Ia research. Recurrent novae are probably the best candidate for this target (e.g., Starrfield, Sparks, \\& Truran 1985; Hachisu et al. 1999b; Hachisu, Kato, \\& Nomoto 1999a). Recently, the recurrent nova U Sco underwent the sixth recorded outburst on February 25, 1999. For the first time, a complete light curve has been obtained from the rising phase to the final fading phase toward quiescence through the mid-plateau phase (e.g., Matsumoto, Kato, \\& Hachisu 2000). Constructing a theoretical light curve of the outburst, Hachisu et al. (2000a) have estimated various physical parameters of U Sco: (1) The early linear phase of the outburst ($t \\sim 1-10$ days after the optical maximum) is well reproduced by a thermonuclear runaway model on a $1.37 \\pm 0.01 M_\\odot$ white dwarf (WD). (2) The envelope mass at the optical maximum is estimated to be $\\sim 3 \\times 10^{-6} M_\\odot$, which results in the mass transfer rate of $\\sim 2.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$ during the quiescent phase between the 1987 and 1999 outbursts. (3) About 60\\% of the envelope mass has been blown off in the outburst wind but the residual 40\\% ($\\sim 1.2 \\times 10^{-6} M_\\odot$) of the envelope mass has been left and accumulated on the white dwarf. Therefore, the net mass increasing rate of the white dwarf is $\\sim 1.0 \\times 10^{-7} M_\\odot$ yr$^{-1}$, which meets the condition for SN Ia explosions of carbon-oxygen cores (\\cite{nom91}). Thus, Hachisu et al. (2000a, 2000b) have concluded that the white dwarf mass in U Sco will reach the critical mass ($M_{\\rm Ia}= 1.378 M_\\odot$, taken from Nomoto, Thielemann, \\& Yokoi 1984) in quite a near future and explode as an SN Ia. Therefore, we regard that U Sco is a very strong candidate for the immediate progenitor of SNe Ia. It has been suggested that the recurrent nova V394 CrA is a twin system of U Sco because of its almost same decline rate of the early light curve and spectrum feature during the outburst (e.g., \\cite{sek89}). It is very likely that the physical parameters obtained for U Sco are common to V394 CrA. In this paper, we derive various physical quantities of V394 CrA, both during the 1987 outburst and in quiescence, by constructing the same theoretical light curve models as for U Sco, and examine whether or not V394 CrA is an immediate progenitor of SNe Ia. In \\S 2, we briefly describe our light curve model during the outburst and present the fitting results for the 1987 outburst of V394 CrA. In \\S 3, based on the physical parameters obtained in \\S 2, we construct theoretical light curves for the quiescent phase of V394 CrA and confirm that the parameters during the outburst are consistent with those in quiescence. Discussion follows in \\S 4, especially for relevance to SN Ia progenitors. ", "conclusions": "Even for much different mass accretion rates, the distance to V394 CrA has been estimated not to be so much different from 4.2 kpc as tabulated in Table \\ref{v394cra_quiescence}. For lower mass accretion rates such as $\\dot M_{\\rm acc}= 1.0 \\times 10^{-7} M_\\odot$ yr$^{-1}$, however, we need the 100\\% irradiation efficiency of the MS ($\\eta_{\\rm ir,MS}=1.0$) or an intrinsic luminosity of the WD as large as $L_{\\rm WD,0} \\sim 300 L_\\odot$ for the 50\\% efficiency ($\\eta_{\\rm ir,MS}=0.5$), in order to reproduce the $\\sim 0.5$ mag sinusoidal variation. We also need an intrinsic luminosity of the WD as large as 300 $L_\\odot$ both for $\\dot M_{\\rm acc}= 0.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$ and for $\\dot M_{\\rm acc}= 0.25 \\times 10^{-7} M_\\odot$ yr$^{-1}$, and 200 $L_\\odot$ for $\\dot M_{\\rm acc}= 0.1 \\times 10^{-7} M_\\odot$ yr$^{-1}$, as summarized in Table \\ref{v394cra_quiescence}. The brightness of the system depends on various model parameters adopted here, that is, the efficiency of the irradiations $\\eta_{\\rm ir, DK}$ and $\\eta_{\\rm ir, MS}$, the intrinsic luminosity of the WD $L_{\\rm WD,0}$, the power of the disk shape $\\nu$. However, the distance estimation itself is hardly affected even if we introduce the different values of the parameters, as clearly shown in Table \\ref{tbl-2}. Thus, we may conclude that the determination of the distance to V394 CrA in quiescence is rather robust as has already been shown in U Sco (\\cite{hkkm00}, 2000b). About 0.5 mag sinusoidal variation of the $B$ light curve during the quiescence needs a relatively large reflection of the companion star as calculated in Figure \\ref{mix_lum_bv}, thus indicating a relatively large luminosity of the WD photosphere. If the intrinsic luminosity of the WD is negligibly small compared with the accretion luminosity (e.g., the nuclear burning is smaller than the accretion luminosity), the mass accretion rate should be higher than $\\dot M_{\\rm acc} \\gtrsim 1.0 \\times 10^{-7} M_\\odot$ yr$^{-1}$ because the efficiency of the irradiation effect must be smaller than 100\\%, which is consistent with our estimation of $\\dot M_{\\rm acc}= 1.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$ derived from the envelope mass at the optical maximum. These systems with relatively high mass accretion rates are exactly the same as those proposed by Hachisu et al. (1999b) as a progenitor system of SNe Ia (see also \\cite{lih97}). Using the same simplified evolutional model as described in Hachisu et al. (1999b), we have followed binary evolutions for various pairs with the initial sets of ($M_{1,i}$, $M_{2,i}$, $a_i$), i.e., for the initial primary masses of $M_{1,i}=4$, 5, 6, 7, and $9 M_\\odot$, the initial secondary masses of $M_{2,i}=1.7$---$3.0 M_\\odot$ by $\\Delta M_{2,i}=0.1 M_\\odot$ step, and the initial separations of $a_i=80$---$600 R_\\odot$ by $\\Delta \\log a_i= 0.01$ step. Starting from the initial set ($7 M_\\odot$, $2.0 M_\\odot$, $150 R_\\odot$), for example, we have obtained a binary system of $M_{\\rm WD,0}= 0.9 M_\\odot$, $M_{\\rm MS,0}= 2.2 M_\\odot$, and $P_0= 1.375$ days, after the binary underwent the first common envelope evolution and then the primary naked helium star evolved to a helium giant and had transferred helium to the secondary MS. \\placetable{tbl-2} Then, the secondary MS has slightly evolved to expand and filled its Roche lobe. Mass transfer begins from the MS to the WD. We have further followed evolution of the binary until the binary reaches $M_{\\rm WD}= 1.37 M_\\odot$ and $P= 0.7577$ days at the same time, that is, we regard the binary as V394 CrA when both the conditions, $M_{\\rm WD}= 1.37 M_\\odot$ and $P= 0.7577$ days, are satisfied at the same time. Then, we obtain the present state of V394 CrA having the secondary mass of $M_{\\rm MS}= 1.39 M_\\odot$ and the mass transfer rate of $\\dot M_2= 1.6 \\times 10^{-7} M_\\odot$ yr$^{-1}$. In our evolutionary model, this binary system will soon explode as an SN Ia when the WD mass reaches $M_{\\rm Ia}= 1.378 M_\\odot$. The mass transfer rate of our evolutionary model is consistent with $\\sim 1.5 \\times 10^{-7} M_\\odot$ yr$^{-1}$ estimated from the light curve fitting. Finally, we may conclude that V394 CrA is the second strong candidate for Type Ia progenitors, next to U Sco (Hachisu et al. 2000a, 2000b)." }, "0003/astro-ph0003298_arXiv.txt": { "abstract": "We present low resolution spectrophotometric and imaging ISO observations of a sample of 57 AGNs and one non-active SB galaxy over the 2.5--11~$\\mu$m range. The sample is about equally divided into type I ($\\leq\\,1.5$; 28 sources) and type II ($>\\,1.5$; 29 sources) objects. The mid-IR (MIR) spectra of type I (Sf1) and type II (Sf2) objects are statistically {\\em different\\/}: Sf1 spectra are characterized by a strong continuum well approximated by a power-law of average index $\\langle \\alpha \\rangle\\,=\\,-0.84\\pm0.24$ with only weak emission features from Polycyclic Aromatic Hydrocarbon (PAH) bands at 3.3, 6.2, 7.7 and 8.6~$\\mu$m. In sharp contrast to Sf1s, most Sf2s display a weak continuum but very strong PAH emission bands, with equivalent widths (EW) up to 7.2~$\\mu$m. On the other hand, Sf1s and Sf2s do not have statistically different PAH luminosities while the 7~$\\mu$m continuum is on the average a factor $\\sim$~8 less luminous in Sf2s than in Sf1s. Because the PAH emission is unrelated to the nuclear activity and arises in the interstellar medium of the underlying galactic bulge, its EW is a sensitive nuclear redenning indicator. These results are consistent with unification schemes and imply that the MIR nuclear continuum source of Sf2s is, on the average, extinguished by $92\\pm37$ visual magnitudes whereas it is directly visible in Sf1s. The dispersion in Sf2's PAH EW is consistent with the expected spread in viewing angles. Those Sf2s with ${\\rm EW(PAH)\\,>\\,5\\,\\mu m}$ suffer from an extinction ${\\rm A_{v}\\,>\\,125}$ magnitudes and are invariably extremely weak X-ray sources. Such Sf2s presumably represent the highly inclined objects where our line of sight intercepts the full extent of the molecular torus. Conversely, about a third of the Sf2s have PAH EW $\\leq\\,2\\mu$m, in the range of Sf1s. Among them, those which have been observed in spectropolarimetry and/or in IR spectroscopy invariably display ``hidden'' broad lines. As proposed by Heisler et al (\\cite{heisler}), such Sf2s are most likely seen at grazing incidence such that one has a direct view of {\\em both\\/} the ``reflecting screen'' {\\em and\\/} the torus inner wall responsible for the near and mid-IR continuum. Our observations therefore constrain the screen and the torus inner wall to be spatially co-located. Finally, the 9.7~$\\mu$m Silicate feature appears weakly in emission in Sf1s, implying that the torus vertical optical thickness cannot significantly exceed ${\\rm 10^{24}\\,cm^{-2}}$. \\footnote{Tables 2,3,4 \\& 5 are only available in electronic form at the CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or via http://cdsweb.u-strasbg.fr/Abstract.html.} ", "introduction": "According to ``unified models'' of Active Galactic Nuclei (AGN), Seyfert 1 and Seyfert 2 galaxies (hereafter Sf1 and Sf2) are essentially the same objects viewed at a different angle: Sf1s are observed close to face-on such that we have a direct view to the Broad emission Line Region (BLR) and the accretion disk responsible for the strong UV-Optical-X-ray continuum, whereas Sf2s are seen at an inclination such that our view is blocked by an optically thick dusty torus which surrounds the disk and the BLR (e.g. Antonucci \\cite{antonucci}). This model makes specific predictions. In particular, the UV photons from the disk which are absorbed by the grains in the torus should be re-emitted as thermal radiation in the IR. Several arguments constrain the torus inner radius to be of the order of $\\sim$ 1 pc in which case the dust temperature should peak to about 700--1000~K and give rise to an emission ``bump'' between $\\simeq$ 2 and 15~$\\mu$m (Pier \\& Krolik \\cite {pier}). The model also predicts that the silicate 9.7~$\\mu$m feature should appear preferentially in absorption in Sf2s and in emission in Sf1s. In order to test these predictions and better constrain the model, we initiated a program of mid-IR (MIR) observations of a large sample of AGNs. Throughout this paper, we use ${\\rm H_{0}\\,=\\,75\\,km\\,s^{-1};\\,q_{0}\\,=\\,0}$. Unless otherwise stated, all quoted uncertainties correspond to 1-$\\sigma$ errors. ", "conclusions": "A sample of 57 AGNs and one normal SB galaxy (NGC~701) were observed with the ISOPHOT-S spectrometer and the ISOCAM imaging camera. The sample is about equally divided into Sf1s (28) and Sf2s (29), where we define Sf1s as all objects of type $\\leq 1.5$ and Sf2s those whose type is $>1.5$. The observations show that: \\begin{enumerate} \\item{Forty-four of the 57 AGNs in the sample appear unresolved at the $\\simeq 4-5\\arcsec$ resolution of ISOCAM. Of the 13 resolved sources, 12 are sufficiently compact to ensure that all of the flux falls into the $24\\arcsec\\times\\,24\\arcsec$ ISOPHOT-S spectrograph aperture. Moreover, even in these resolved sources, nuclear/bulge emission contributes for at least 3/4\\,th of the light recorded with ISOPHOT-S.} \\item{The spectrum of Sf1s is characterized by a strong continuum and weak Polycyclic Aromatic Hydrocarbon (PAH) emission bands at 3.3, 6.2, 7.7 and 8.6~$\\mu$m. The continuum is well described by a power-law of average index $\\langle \\alpha \\rangle\\,=-0.84\\pm0.24$.} \\item{In sharp contrast with Sf1s, Sf2s generally have a weak continuum with very strong PAH emission bands.} \\item{The distribution of PAH equivalent widths (EW) is statistically different in Sf1s and Sf2s. The average EW for the strongest band at 7.7~$\\mu$m is $0.53\\pm0.47~\\mu$m in Sf1s versus $2.86\\pm1.95~\\mu$m in Sf2s. Moreover, the distribution of PAH EW in Sf1s is confined to values smaller than 2.0~$\\mu$m whereas that of Sf2s extends from 0.24~$\\mu$m up to 7.2~$\\mu$m.} \\item{There are however no statistical differences in the PAH luminosity distribution of Sf1s and Sf2s.} \\item{The 7~$\\mu$m continuum is on the average a factor $\\simeq$~8 less luminous in Sf2s than in Sf1s.} \\item{The PAH emission is not related to the activity in the nucleus and originates in the interstellar medium of the underlying galactic bulge. The PAH EW can therefore be used as a nuclear redenning indicator.} \\item{The above results are consistent with unification schemes and imply that the MIR continuum of Sf2s suffers from an average extinction of $92\\pm37$ visual magnitudes. This corresponds to an average hydrogen absorbing column ${\\rm N_{H}\\,=\\,2.0\\pm0.8\\,cm^{-2}}$, in good agreement with X-ray measurements. The large dispersion in the Sf2s EW is consistent with the expected spread in viewing angles.} \\item{The spectrum of Sf2s whose 7.7~$\\mu$m PAH band EW exceeds 5~$\\mu$m is indistinguishable from that of a normal non-active galaxy, implying that the MIR continuum is completely obscured in these sources (${\\rm A_{v}\\,>\\,125}$ magnitudes). Without exception, these Sf2s are also heavily absorbed in the X-rays and probably ``compton thick''. These large PAH EW Sf2s are presumably those where the torus is seen edge-on.} \\item{Ten Sf2s have 7.7~$\\mu$m PAH EW $\\leq\\,2.0~\\mu$m, in the range of Sf1s. Of these ten, four have been observed in spectropolarimetry and all four display ``hidden'' broad lines. Conversely, none of the three Sf2s with PAH EW $>\\,2\\,\\mu$m which have been observed in spectropolarimtery display ``hidden'' broad lines. This confirms the finding of Heisler et al. (\\cite{heisler}) that those Sf2s with a ``hidden'' BLR are those for which our line-of-sight grazes the upper surface of the torus. In these sources, we have a direct view of {\\em both\\/} the reflecting mirror {\\em and\\/} of the torus inner wall responsible for the MIR continuum. Thus, our observations strongly favour a model where the ``mirror'' and the torus inner wall are spatially co-located. It is in fact conceivable that the mirror is the torus inner wall itself or a wind of hot electrons boiled-off its surface by radiation pressure.} \\item{The silicate 9.7~$\\mu$m feature appears weakly in emission in Sf1s. This implies that the torus cannot be extremely thick and the average silicate EW ($0.25\\pm0.01\\,\\mu$m) suggests that the total hydrogen column integrated along the torus vertical axis lies in the range ${\\rm 10^{23}\\,\\leq\\,N_{H}\\,\\leq\\,10^{24}\\,cm^{-2}}$, consistent with our previous estimate based on the Sf2 PAH EW.} \\item{As far as their MIR properties are concerned, AGNs of intermediate types 1.8 and 1.9 are indistingishable from genuine SF2s, whereas Narrow Line Seyfert 1 (NLS1) and Broad-Line Radio-Galaxies (BLRG) behave as normal SF1s.} \\end{enumerate} The sketch outline in this paper makes specific predictions. First, Sf2s which have 7.7~$\\mu$m PAH EW in excess of $\\sim\\,5\\,\\mu$m should never exhibit broad-lines in spectropolarimetry. Second, these sources should always be heavily absorbed in the X-rays, possibly up to 10~keV. Third, Sf2s whose PAH EW $\\leq\\,2\\,\\mu$m should exhibit broad lines when observed in spectropolarimetry and/or direct IR spectroscopy. This last prediction seems to be borne by the few existing ISO observations of Sf2 with a ``hidden'' BLR" }, "0003/astro-ph0003251_arXiv.txt": { "abstract": "The new transient X-ray pulsar XTE~J0111.2--7317 was observed with {\\it Advanced Satellite for Cosmology and Astrophysics} ({\\it ASCA}) on 1998 November 18, a few days after its discovery with the Proportional Counter Array onboard the {\\it Rossi X-ray Timing Explorer}. The source was detected at a flux level of $3.6\\times10^{-10}$~erg~cm$^{-2}$~s$^{-1}$ in the 0.7--10.0~keV band, which corresponds to the X-ray luminosity of $1.8\\times10^{38}$~erg~s$^{-1}$, if a distance of 65 kpc for this pulsar in the Small Magellanic Cloud is assumed. Nearly sinusoidal pulsations with a period of $30.9497 \\pm 0.0004$~s were unambiguously detected during the {\\it ASCA} observation. The pulsed fraction is low and slightly energy dependent with average value of $\\sim27$\\%. The energy spectrum shows a large soft excess below $\\sim 2$~keV when fitted to a simple power-law type model. The soft excess is eliminated if the spectrum is fitted to an ``inversely broken power-law'' model, in which photon indices below and above a break energy of 1.5~keV are 2.3 and 0.8, respectively. The soft excess can also be described by a blackbody or a thermal bremsstrahlung when the spectrum above $\\sim 2$~keV is modeled by a power-law. In these models, however, the thermal soft component requires a very large emission zone, and hence it is difficult to explain the observed pulsations at energies below 2~keV. A bright state of the source enables us to identify a weak iron line feature at 6.4~keV with an equivalent width of $50\\pm14$~eV. Pulse phase resolved spectroscopy revealed a slight hardening of the spectrum and marginal indication of an increase in the iron line strength during the pulse maximum. ", "introduction": "The transient X-ray pulsar XTE~J0111.2--7317 was discovered with the Proportional Counter Array (PCA) of the {\\it Rossi X-ray Timing Explorer} ({\\it RXTE}) in 1998 November (Chakrabarty et al. 1998a\\markcite{Chakrabarty1998a}) and was simultaneously detected in hard X-rays (Wilson \\& Finger 1998\\markcite{Wilson1998}) with the Burst and Transient Source Experiment (BATSE) onboard the {\\it Compton Gamma-Ray Observatory} ({\\it CGRO}). Public data of {\\it CGRO}/BATSE and {\\it RXTE}/ASM (All Sky Monitor) revealed that both the hard and the soft X-ray intensities of the source showed two outbursts as shown in Figure \\ref{fig:initial}. Initial BATSE measurements also found the pulsar to be spinning-up with a very short time scale of $P/\\dot{P} \\sim 20$~yr, which is also a confirmation that the compact object is a neutron star. The spin-up rate shows a positive correlation with the flux, which is similar to other transient X-ray pulsars. Following its discovery, a {\\it Target of Opportunity} observation was made with the {\\it Advanced Satellite for Cosmology and Astrophysics} ({\\it ASCA}) and the position of the X-ray source was determined precisely (Chakrabarty et al. 1998b\\markcite{Chakrabarty1998b}). Previous observations of the field with {\\it EINSTEIN} and {\\it ROSAT} did not detect any source at the position determined with {\\it ASCA} (Wang \\& Wu 1992\\markcite{Wang1992}; Kahabka \\& Pietsch 1996\\markcite{Kahabka1996}; Cowley et al. 1997\\markcite{Cowley1997}; Kahabka et al. 1999\\markcite{Kahabka1999}; Schmidtke et al. 1999\\markcite{Schmidtke1999}), confirming its true transient nature. Since transient pulsars in our Galaxy are located in the Galactic plane whereas XTE~J0111.2--7317 is located in the direction of the Small Magellanic Cloud (SMC), i.e. at high Galactic lattitude ($\\sim -43^\\circ$), it is very likely that XTE~J0111.2--7317 lies in the SMC. A candidate for the optical counterpart was detected in the {\\it ASCA} error circle, and was found to be emitting strong H$\\alpha$ and H$\\beta$ lines (Israel et al. 1999\\markcite{Israel1999}; Coe et al. 1999\\markcite{Coe1999}). Coe et al. (1999\\markcite{Coe1999}) also determined the velocity shift of optical emission lines to be $166 \\pm 15$~km~s$^{-1}$, which indicates that this object is in the SMC. Therefore the distance would be large (65~kpc is assumed in this paper), which indicates a very high luminosity, in excess of 10$^{38}$~erg~s$^{-1}$ in the 2.0--10.0~keV band during the transient phase. The {\\it ASCA} observation of the source was made to study the pulsations and the soft X-ray spectrum extending upto 10~keV. Though the {\\it ASCA} observation happened to coincide with the quiescent period between the two outbursts (Figure \\ref{fig:initial}), the source was bright enough for us to make a detailed study of its energy spectrum. We present here results of our temporal and spectral study of this source made with the {\\it ASCA} data. The principal result from this study is discovery of a soft excess in the spectrum which is modeled as an ``inversely broken power-law model.'' In the following sections, we present the details of the observation and the temporal and spectral analysis, and discuss the results suggesting that this source is a Be/X-ray binary. ", "conclusions": "We have presented the {\\it ASCA} observation of the newly discovered X-ray pulsar XTE~J0111.2--7317 in its bright transient phase. Pulsations at 30.95~s were unambiguously detected, and the pulse shape and the pulsed fraction are found to have a slight energy dependence. The energy spectrum in the 0.7--10.0~keV band has a soft excess above a simple power-law type model, and can be described by an ``inversely broken power-law'' model, which is a broken power-law with a larger/smaller photon index below/above a break energy. \\clearpage" }, "0003/astro-ph0003067_arXiv.txt": { "abstract": "This paper analyzes the integrated ultraviolet spectra taken with the {\\it International Ultraviolet Explorer} ({\\it IUE}) and optical ground-based spectra taken with the William Herschel Telescope (WHT) of the giant H\\,{\\sc ii} region NGC 604. These data are complemented with ultraviolet (WFPC2 through F170W) and H$\\alpha$ (WFPC2 through F656N) images retrieved from the {\\it Hubble Space Telescope} ({\\it HST}) archive. The ultraviolet resonance wind stellar lines, the nebular optical emission lines and the higher order terms of the Balmer series and HeI absorption lines detected in the spectra of NGC 604 are interpreted using evolutionary models optimized for young star forming regions. The evolutionary state and the massive stellar content of the region is derived in a self-consistent way. The three techniques applied suggest that the central ionizing cluster in NGC 604 is very young, $\\simeq3$ Myr old, and that the stars in the cluster were formed in an instantaneous burst following a Salpeter ($\\alpha$=2.35) or flatter ($\\alpha$=1.5) initial mass function (IMF), having stars more massive that 80 M$\\odot$. The stellar cluster is able to provide most of the ionizing photons needed to photoionize the whole nebula, and the wind power to form the central shell structure where the cluster core is located. The stellar cluster is affected by an extinction similar to the average extinction that affects the ionized gas. The estimated number of massive stars in the cluster is also in agreement with that derived from previous studies based on the detection of individual stars. The results that we present here support the use of these techniques for the interpretation of the integrated light of more distant star forming regions. ", "introduction": "Giant extragalactic H\\,{\\sc ii} regions (RH\\,{\\sc ii}s) are amongst the brightest objects in galaxies. RH\\,{\\sc ii}s have been studied extensively because they are the best indicators of the conditions that lead to massive star formation, and they show the cloud properties immediately after the moment when stars form. Equally important, they are favourable places where to measure the chemical abundances and star formation rates in galaxies, and hence can be used to trace the chemical and star formation history of the universe, when observed in galaxies at different redshift (Madau et al. 1996; Kobulnicky \\& Zaritsky 1999). RH\\,{\\sc ii}s are characterized to have a size larger than 100 pc and H$\\alpha$ luminosity brighter than 10$^{39}$ erg s$^{-1}$ (Kennicutt 1984). Therefore, the nebula requires an ionizing photon luminosity larger than 10$^{51}$ s$^{-1}$; this is provided by a stellar cluster that contains more than 100 young massive stars. These characteristics are very similar to those of starburst galaxies. However, they are less luminous than prototypical starbursts, and thus they are referred to as mini-starbursts (Walborn 1991). Like starbursts, RH\\,{\\sc ii}s show a nebular emission-line spectrum at optical wavelengths, and an absorption-line spectrum at wavelengths shorter than the Balmer jump (Leitherer 1997; Rosa, Joubert \\& Benvenuti 1984). This spectral morphology reflects the fact that RH\\,{\\sc ii}s are powered by massive stars. These stars emit photons with energies of tens of eV which are absorbed and re-emitted in their stellar winds, producing ultraviolet resonance transitions. However, the stellar wind is optically thin to most of the ultraviolet photons, that can travel tens of parsec from the star before they are absorbed and photoionize the surrounding interstellar medium. Subsequently, this ionized gas cools down via an emission spectrum. This spectral dichotomy picture allows to derive the stellar content and the evolutionary state of the cluster, through the analysis of the ultraviolet (e.g. Vacca et al. 1995; Leitherer et al. 1996), or the optical light (e.g. Cid Fernandes et al. 1992; Garc\\'\\i a-Vargas \\& D\\'\\i az 1994; Stasi\\'nska \\& Leitherer 1996) using evolutionary synthesis and photoionization models. The two techniques have been applied to the prototypical starburst nucleus NGC 7714, giving similar results (Gonz\\'alez Delgado et al. 1999a). However, around the Balmer jump the spectra of starbursts (Gonz\\'alez Delgado et al. 1998) and some H\\,{\\sc ii} regions (Terlevich et al. 1996) show the higher order terms of the Balmer series and He\\,{\\sc i} lines in absorption, formed in the photospheres of massive stars. These stellar lines can be detected in absorption because the strength of the gaseous Balmer lines in emission decreases rapidly with decreasing wavelength, whereas the equivalent width of the stellar absorption lines is almost constant with wavelength (Gonz\\'alez Delgado, Leitherer \\& Heckman 1999b). Evolutionary synthesis models that predict the profiles of the higher order terms of the Balmer series and He\\,{\\sc i} lines in absorption can also be used to estimate the evolutionary state of the stellar cluster in H\\,{\\sc ii} regions. The second more luminous and brightest H\\,{\\sc ii} region in the Local Group of galaxies, after 30 Dor in the LMC, is NGC 604 in M 33. Its distance, 840 kpc (Freedman et al. 1991), allows detailed studies of the individual stars, and also to obtain integrated properties. Thus, NGC 604 is an excellent laboratory in which to explore questions about the effect of star formation, the slope and upper mass limit of the IMF, and to test the consistency between the three techniques described above, that allow to determine the stellar content and the evolutionary state of the ionizing stellar cluster from the spatially integrated ultraviolet and optical spectra of the H\\,{\\sc ii} region. NGC 604 has been intensively studied in the past (e.g. Peimbert 1970; Israel \\& van der Kruit 1974; Smith 1975; Hawley \\& Grandi 1977; Kwitter \\& Aller 1981; Conti \\& Massey 1981; Rosa \\& D'Odorico 1982; Viallefond \\& Goss 1986). A precise determination of the chemical composition of the gas was made by D\\'\\i az et al. (1987); they give an oxygen abundance 12+log(O/H)=8.5. VLA observations indicate that the gas is very tenuous (average electron density, $rms$ $N_{\\rm e}$, between a few and $\\sim100$ cm$^{-3}$), and not very dusty, with a mean visual extinction $A_V\\sim0.5$ mag (Churchwell \\& Goss 1999). The morphology of the ionized gas is very complex, showing many filaments and shell structures that are expanding (e.g. Hippelein \\& Fried 1984; Rosa \\& Solf 1984; Clayton 1988; Sabalisck et al. 1995; Mu\\~noz-Tu\\~n\\'on et al. 1995; Yang et al. 1995; Medina Tanco et al. 1997). This morphology is a consequence of the violent star formation activity in NGC 604. {\\it HST} imaging photometry has revealed that the stellar cluster is resolved into $\\sim$200 massive stars in an area of $\\sim$10000 pc$^2$ (Drissen, Moffat \\& Shara 1993; Hunter et al. 1996). Evidences that the stellar cluster is evolved come from the detection of Wolf-Rayet stars (Conti \\& Massey 1981; D'Odorico \\& Rosa 1981; Drissen, Moffat \\& Shara 1990, 1993), one supernova remnant (D'Odorico et al. 1980), and one candidate to red supergiant (Terlevich et al. 1996). The existence of these stars as members of the stellar cluster of NGC 604 suggests that the age of the region is in the range 3-5 Myr. This paper presents spatially integrated ultraviolet and optical spectra of NGC 604. The goal is to derive the massive stellar content and the evolutionary state of the ionizing cluster, by means of evolutionary synthesis and photoionization models applied to the ultraviolet resonance wind stellar lines, to the nebular emission lines and to the higher-order terms of the Balmer series and He\\,{\\sc i} lines in absorption. The consistency between the results obtained with the three techniques, and the similarity with the results obtained from studies based on the detection of individual stars, strengthens the reliability and power of these techniques when they are applied to determine the stellar content of more distant star-forming regions. Section 2 presents the observations. Section 3 describes the photoionization models that fit the emission-line spectrum of the nebula. In section 4, the higher-order terms of the Balmer and He\\,{\\sc i} lines in absorption are analyzed. Evolutionary synthesis models of the ultraviolet resonance wind stellar lines are in section 5. In section 6, the massive stellar content is derived. The summary and conclusions are in section 7. ", "conclusions": "" }, "0003/astro-ph0003317_arXiv.txt": { "abstract": "We present $JHK_s$ photometry, far red spectra, and spectral classifications for an additional 67 L dwarfs discovered by the Two Micron All Sky Survey. One of the goals of this new search was to locate more examples of the latest L dwarfs. Of the 67 new discoveries, 17 have types of L6 or later. Analysis of these new discoveries shows that H$\\alpha$ emission has yet to be convincingly detected in any L dwarf later than type L4.5, indicating a decline or absence of chromospheric activity in the latest L dwarfs. Further analysis shows that 16 (and possibly 4 more) of the new L dwarfs are lithium brown dwarfs and that the average line strength for those L dwarfs showing lithium increases until type $\\sim$L6.5 V then declines for later types. This disappearance may be the first sign of depletion of atomic lithium as it begins to form into lithium-bearing molecules. Another goal of the search was to locate nearer, brighter L dwarfs of all subtypes. Using absolute magnitudes for 17 L dwarf systems with trigonometric parallax measurements, we develop spectrophotometric relations to estimate distances to the other L dwarfs. Of the 67 new discoveries, 21 have photometric distances placing them within 25 parsecs of the Sun. A table of all known L and T dwarfs believed to lie within 25 parsecs -- 53 in total --- is also presented. Using the distance measurement of the coolest L dwarf known, we calculate that the gap in temperature between L8 and the warmest known T dwarfs is less than 350K and probably much less. If the transition region between the two classes spans a very small temperature interval, this would explain why no transition objects have yet been uncovered. This evidence, combined with model fits to low-resolution spectra of late-M and early-L dwarfs, indicates that L-class objects span the range 1300K $\\lesssim T_{eff} \\lesssim$ 2000K. The near-infrared color-color diagram shows that L dwarfs fall along a natural, redder extension of the well known M dwarf track. These near-infrared colors get progressively redder for later spectral types, with the L dwarf sequence abruptly ending near ($J-H$, $H-K_s$, $J-K_s$) $\\approx$ (1.3, 0.8, 2.1). ", "introduction": "In 1993, the first spectrum of what would later be known as an L dwarf was published (Kirkpatrick, Henry, \\& Liebert 1993). This object, GD 165B, had been discovered earlier by Becklin \\& Zuckerman (1988) as a resolved companion to a nearby white dwarf. For several years GD 165B remained in a class by itself. Then, beginning in 1997, an explosion of discoveries proved that L dwarfs are quite common in the solar neighborhood (Reid et al. 1999; see also Delfosse et al. 1997; Ruiz, Leggett, \\& Allard 1997; Rebolo et al.\\ 1998; Kirkpatrick et al.\\ 1999; Goldman et al.\\ 1999; Mart{\\'{\\i}}n et al. 1999b; Fan et al.\\ 2000). Providing a historical parallel to the L dwarfs are the T dwarfs, even cooler objects spectroscopically defined as those showing methane at $K$-band (Kirkpatrick et al.\\ 1999; hereafter referred to as Paper I). In 1995, the first spectrum of a T dwarf was published (Oppenheimer et al.\\ 1995). This object, Gl 229B, had been discovered as a companion to a nearby M dwarf by Nakajima et al.\\ (1995) and remained in a class by itself for several years. Then beginning in 1999, an explosion of discoveries proved that observable T dwarfs have a space density comparable to that of L dwarfs (Strauss et al.\\ 1999, Burgasser et al.\\ 1999, Cuby et al.\\ 1999, Burgasser et al.\\ 2000a, Tsvetanov et al.\\ 2000, Burgasser et al.\\ 2000c). Despite the implied space density and the subsequent profusion of cooler, T dwarf discoveries, the number of known L dwarfs is still small. Additional examples, including a larger number of late-L dwarfs and nearer (brighter) examples of all L subtypes, are needed for further studies including parallax measurement, luminosity and temperature determination, kinematics, binarity, and detailed spectroscopic analyses related to magnetic activity, lithium frequency, atmospheric abundances, dust formation, etc. To this end, we present here another 67 L dwarfs found during follow-up of candidates selected from Two Micron All Sky Survey (2MASS) data. ", "conclusions": "We present spectra for another 67 L dwarfs discovered during follow-up of sources identified by 2MASS. These together with L dwarfs from Paper I, from Gizis et al.\\ (2000), and from other surveys such as SDSS and DENIS, bring the total of known L dwarfs to well over 100. This sample can be used for a variety of follow-up investigations. The presence of H$\\alpha$ emission is seen to decline rapidly from early- to mid-L dwarfs, and H$\\alpha$ emission is not seen in any L dwarf later than type L4.5. If the lack of H$\\alpha$ emission is an indicator of youth and/or substellarity as Gizis et al.\\ (2000) suggest, then the fraction of H$\\alpha$-emitters to non-emitters at any given L subtype may reflect the fraction of stars to brown dwarfs at that class. Lithium absorption, when detected, is seen to increase in strength from early- to mid-L types, but then declines markedly after type L6.5 V. This turnover in lithium strength may herald the depletion of atomic lithium into lithium-bearing molecules and as such would provide a vital clue to the temperature scale for L dwarfs because these reactions are expected roughly around 1500K. The difference in temperature between the latest L dwarf and Gl 229B is calculated at $\\sim$350K. This means that the gap in temperature between L8 and the warmest of the known T dwarfs must be significantly less that 350K as several of the known T dwarfs are suspected of being warmer than Gl 229B itself. This also means that L dwarfs span the likely temperature range 1300K $\\lesssim T_{eff} \\lesssim$ 2000K. The locus of L dwarfs in near-infrared color space is also shown, and distances estimates are made for all L and T dwarfs lacking trigonometric parallax measurements. Even at this early stage in our investigations, researchers have identified 53 L and T dwarfs known (or suspected) to be within 25 parsecs of the Sun, clearly indicating that this previously hidden population of cool objects is very large." }, "0003/astro-ph0003121_arXiv.txt": { "abstract": "Helioseismic observations have detected small temporal variations of the rotation rate below the solar surface corresponding to the so-called `torsional oscillations' known from Doppler measurements of the surface. These appear as bands of slower and faster than average rotation moving equatorward. Here we establish, using complementary helioseismic observations over four years from the GONG network and from the MDI instrument on board SOHO, that the banded flows are not merely a near-surface phenomenon: rather they extend downward at least 60 Mm (some 8\\% of the total solar radius) and thus are evident over a significant fraction of the nearly 200 Mm depth of the solar convection zone. ", "introduction": "The intensely turbulent state of the solar convection zone is revealed by the patterns of granulation, mesogranulation and supergranulation evident in its surface layers (e.g., Brummell, Cattaneo \\& Toomre 1995). Yet accompanying such turbulent and seemingly chaotic small-scale dynamics are also signs of ordered large-scale behavior. Most notably the solar differential rotation involves a relatively smooth decrease in angular velocity from equator to pole, both in the surface layers (e.g., Snodgrass 1984) and within the convection zone as inferred from helioseismic measurements (e.g., Thompson et al. 1996; Schou et al. 1998a). On the largest scales, the magnetic activity similarly exhibits well-defined rules as the 22-year cycle progresses. An enticing link between the latitudes of field emergence and small variations in the rotation rate of the surface layers is provided by bands of slightly faster and slower than average zonal flows, called torsional oscillations, that were observed from direct Doppler measurements to migrate towards the equator in a manner similar to the zones of solar activity (e.g., Howard \\& LaBonte 1980; Snodgrass, Howard \\& Webster 1985; Ulrich 1998). Helioseismic analysis of data from the Michelson Doppler Imager (MDI) instrument (e.g., Scherrer et al. 1995) on the Solar and Heliospheric Observatory (SOHO) spacecraft has confirmed the presence of such bands of weak zonal flow, and their drift towards the equator, for the present solar cycle (Kosovichev \\& Schou 1997; Schou et al. 1998a,b; Schou 1999). Although the causal relation between these banded flows and the zones of magnetic activity is still unclear, it is important to understand whether the flows are confined to the layer of rotational shear just below the solar surface. In this letter, we address such questions using two extensive helioseismic data sets, covering slightly over four years, obtained with MDI and with the ground-based Global Oscillation Network Group (GONG) project (e.g., Harvey et al. 1996). We establish the consistency of the independent determinations of the flow from the two data sets, and infer that the zonal banding signature extends to depths of about 60 Mm (or about 8\\% in radius) below the solar surface. Thus these are not superficial features, and provide evidence of ordered rotational responses as the magnetic cycle is progressing. More extensive accounts of such analyses of zonal flows are provided for GONG data by Howe, Komm \\& Hill (2000) and for MDI data by Toomre et al. (2000). \\vskip-20truept ", "conclusions": "Analysis of extended series of GONG and MDI data has revealed coherent banded flow structures in the solar convection zone. These correspond to the torsional oscillations detected in direct Doppler observations of the solar surface. We have demonstrated that the flows are likely to extend to a depth of at least 60 Mm, a substantial fraction of the total 200 Mm depth of the convection zone, and considerably more than the depth (about 35 Mm) at which the rotation rate attains its maximum in the subsurface radial shear layer at low latitudes (cf. Schou et al. 1998a). In addition, there appear to be other systematic variations with time of the residual rotation rate, with different signatures at low and high latitudes (cf. Fig.~\\ref{fig:F3}). Inversions of global oscillation frequency splittings sample the component of rotation symmetric around the equator. The actual flows will exhibit some level of asymmetry which will depend on the time scale used in the analysis. Indeed, local analyses by means of the time-distance and ring-diagram techniques (Giles, Duvall, \\& Scherrer 1998; Haber et al. 2000) have shown features similar to those found here, but with clear differences between the two hemispheres. {\\center\\epsfig{file=fig4.ps,width=\\linewidth}} \\figcaption{ Sample artificial data tests to study residual rotation rate inferred from OLA inversions at various selected radii. The imposed flow used in the calculation of the frequency splittings extends from the surface to $0.95 R$ in the top row, to $0.92 R$ in the middle row, and to $0.80 R$ in the bottom row. \\label{fig:F4} } \\vspace{5mm} \\placefigure{fig:F4} The link between the evolving latitudinal positioning of the faster zonal bands and of the sites of sunspot emergence suggest that the dynamics are related, yet how this is accomplished is uncertain. The strong magnetic fields most likely originate from deep within the Sun, probably formed by dynamo action near the base of the convection zone (e.g., Spiegel \\& Zahn 1992; Parker 1993; Weiss 1994; Charbonneau \\& MacGregor 1997). Field bundles ascending from this region through the convection zone, before erupting into the atmosphere as large-scale magnetic loops, could well lead to significant perturbations in velocity and thermal fields there. This is likely to be accompanied by some redistribution of angular momentum, given that the magnetic structures will attempt to conserve their original angular momentum (e.g., Brummell, Cattaneo \\& Toomre 1995). The coupling of a highly turbulent medium with ascending magnetic structures, and their mutual feedbacks, have not yet been assessed in recent flux-tube models. Global simulations of turbulent convection in rotating spherical shells (e.g., Elliott et al. 2000; Miesch et al. 2000; Miesch 2000) to study the resulting differential rotation have revealed intrinsic variability in zonal flows over intervals of several rotation periods, some of which may be inertial oscillations (e.g., Gunther \\& Gilman 1985), but such modelling has not included large-scale magnetic fields. Obtaining propagating bands and time scales of variation of order the solar cycle seems problematic unless there is some selective coupling to magnetic processes. Adding to the puzzle is that the evolving zonal bands are present at the higher latitudes even before the prominent large-scale magnetic eruptions begin (e.g. Ulrich 1998), as within this cycle. Continued helioseismic observations as this magnetic cycle is proceeding may help to provide clues about such aspects of solar internal dynamics, for we now have the ability to probe hitherto unseen flows well below the solar surface." }, "0003/astro-ph0003135_arXiv.txt": { "abstract": "ROSAT/HRI observations of the powerful radio-loud galaxy 3C382 reveal extended X-ray emission associated with the source. On the basis of this new spatial component, a previous ROSAT/PSPC spectral analysis of the source is revised. Allowing for the presence of an additional thermal component in the PSPC spectrum, the non-thermal component is found to be compatible with the extrapolation of the well defined 3C~382 ---2 - 10 keV--- power-law spectrum into the soft X-rays. The thermal --extended-- component would then account for the soft excess emission previously reported for this source. The origin of this thermal component is not clear. Its luminosity compares with that of rich Abell clusters; yet, the galaxy environment in 3C382 appears of moderate optical richness. An alternative is that it is due to a massive extended gaseous atmosphere sustained by the deep gravitational potential well of 3C382.\\\\ ", "introduction": "Soft X-ray excess emission above a simple extrapolation of the hard energy spectrum is found in a considerable number of AGNs, mainly radio-quiet sources (see Mushotzky, Done, \\& Pounds 1993 and references therein). There is a growing body of evidence for its spectral ubiquity below $\\sim$2 keV or so in Seyferts (e.g. Pounds et al. 1994) as in highly luminous quasars (e.g. Saxton et al. 1993). Most previous studies before ROSAT converge toward the idea that the soft excess is a rather common feature among radio-quiet quasars, whereas it is almost absent in their radio-loud counterparts. {\\em ROSAT}/PSPC data of radio loud sources have shown that a soft excess component is also present in radio loud sources (e.g. Buehler et al. 1995; Prieto 1996; Siebert et al. 1998). Possible interpretations for the soft X-ray excess in AGNs include thermal emission from the inner regions of an accretion disk, scattering by highly ionized material in its vicinity (Pounds et al. 1986; Ross \\& Fabian 1993), or thermal emission due to shock-heated gas in the close vicinity of the nucleus (Viegas \\& Contini 1994). The poor spatial resolution of the {\\em ROSAT}/PSPC makes difficult the separation between possible components of the observed emission. Indeed, the large PSPC resolution beam, $\\sim$ 25 arcsec at 1 keV, makes plausible that an important part of the observed emission to be due to an extended gas component surrounding the AGN. In the particular case of radio-loud galaxies which are characterized by large radio sizes, a hot surrounding medium becomes a necessary component for providing the working surface for the radio emission. In the analysis of the 3CRR sample by Prieto, a first attempt to fit the PSPC spectra of sources with extended emission --mostly in Fanaroff \\& Riley (1974) type I sources (FRI)-- with a single power-law leaded to extreme step spectral index, the reason being due to the dominant contribution of the gaseous medium in which those sources usually reside. In the case of FRII, a single power-law fit provided a fair representation of the PSPC spectrum but with average spectral index about -1.1, and so above the extrapolation of the canonical hard-energy spectrum into the soft X-rays. Clustering of galaxies about FRII sources is less common than in FRI, in particular at low redshift; yet, FRII could contain their own extended gaseous atmosphere which may directly translate into a steepening of the PSPC spectrum. This component however may prove to be elusive with present X-ray instrumentation. This paper presents deep {\\em ROSAT}/HRI observations of the powerful X-ray radio-loud source 3C~382. This is one of the few nearby broad-line galaxies ($z= $0.0578) that show extremely bright and broad permitted lines (FWZI$> 25000 km~s{-1}$; Tadhunter, Perez \\& Fosbury 1986) and a strong continuum, with with a X-ray luminosity in 0.2-2.4 keV band of $L_{\\rm x} \\sim 7.10^{44}$ erg~s$^{-1}$ (Prieto 1996), and a radio power at 178 MHz of $L_{\\rm 178MHz} \\sim 3.10^{33}$ erg~s$^{-1} Hz^{-1}$ (Laing et al. 1983). {\\em EXOSAT} monitoring of the source (1983--1985) tightly constrains the high-energy (above 2 keV) spectral index of 3C382 to $\\alpha=-0.7\\pm 0.1$ (Ghosh \\& Soundararajaperumal 1992). This is also confirmed by more recent ASCA data (Wozniak et al. 1998). However, the ROSAT/PSPC spectral analysis of the source shows compatible with a power-law model with spectral index $\\alpha= -1.2\\pm0.3$, and absorbed by a column density, $N(H)= 0.78\\times 10^{21} cm^{-2}$, that is in agreement with the Galactic value. Thus, 3C382 shows a soft excess emission below $\\sim$ 2keV (Prieto 1996). Independently, the presence of a soft excess is also inferred from the analysis of the EXOSAT (Ghosh \\& Soundararajaperumal), ASCA (Wozniak et al.) and Ginga (Kaastra et al 1991) data. Extended soft X-ray emission associated with this source is detected in the ROSAT/HRI data. On the basis of that new component a re-evaluation of the PSPC spectrum is presented. Throughout this work $H_{0}$ = 50 km s$^{-1}$ Mpc$^{-1}$. 1 arcsecond corresponds to $\\sim$ 1.7 kpc at the source. \\section {Analysis of HRI data} The HRI observations of 3C~382 were conducted in 1996 October and 1997 April (WG900720H and WG900720H-1 datasets respectively) The corresponding total accepted times were 4514 s and 13310 s, respectively. The counts are integrated from channels 1 to 8 which enclose most of the energy accumulated in the ROSAT band, yielding for both dataset count rates of $\\sim$1 cts s$^{-1}$. The nominal resolution of the {\\em ROSAT}/HRI is $\\sim$ 5--6 arcsec (FWHM). Residual errors in the ROSAT aspect solution are known to give rise to elongated images (David et al. 1996), the shape of the surface brightness profile dramatically departing from the expected point response function (PRF). In the case of very bright sources, improvement of the HRI spatial resolution becomes feasible by using speckle interferometric techniques such as the ``shift-and-add'' method. If one constructs images over short time intervals, the PRF becomes symmetric and therefore the elongation in the image appears as an apparent residual motion of the X-ray source in the sky. Such residual motion can be corrected for by de-speckling. 3C~382 is particularly suitable for that technique as it shows very bright in the {\\em ROSAT} band, with $\\sim$ 2 cts s$^{-1}$ in the PSPC and $\\sim1$ cts s$^{-1}$ in the HRI. The procedure used follows the same criteria and approach originally presented in Schmitt, G\\\"udel \\& Predehl (1994). Basically, each event file is divided into time bins of typically 50 s. For each bin, the apparent X-ray position of the source (i.e., RA and $\\delta$) is determined as a function of the observing time. A spline function is fitted to these data points, all recorded photons being then corrected with the appropriate time-dependent correction in RA and $\\delta$. To validate the correction, new measurements of the source centroid are repeated on the corrected event file. The correction is considered as satisfactory if the new centroid positions cluster during the time period of the observation about an average constant value. The uncertainty in the final source centroid is about 2.5 arcsec in RA and 1.5 arcsec in $\\delta$. The X-ray spatial analysis was then performed on the corrected HRI event files. Because of the much higher statistic of the April event file, reliable results from the de-speckle procedure are only found from that dataset. Thus, the following analysis focuses only on this dataset. ", "conclusions": "Extended X-ray emission associated with the very bright, nearby radio-loud galaxy 3C~382 is detected. The analysis of the ROSAT/HRI data shows that about 10\\% of the total 0.2 -2.4 keV emission is compatible with the presence of an extended component. Assuming that component to be due to hot gas emitting via bremsstrahlung, and allowing for the contribution of such thermal component into the PSPC spectrum, it is found that the non-thermal component of 3C~382 emission becomes consistent with the extrapolation of the ``well established'' 3C~382 high-energy power-law spectrum ---above 2 keV--- into the soft X-ray regime. The temperature of the gas component as formally derived from the PSPC fit is $0.6^{+0.4}_{-0.1}$ keV. This low temperature contrasts with the high luminosity of the gas component, $\\sim3\\times 10^{44} erg~sec^{-1}$. Taken together the spectral limitation of the PSPC and the complexity of the model fit, the uncertainty in the temperature could however be larger. There is furthermore the possibility that a temperature gradient dominates the gas emission --- a cooling flow process. In this case, the PSPC spectrum may be mostly sampling the central gas region, where the coolest and more dense gas is located. There are however other factors that could also have lead to an overestimation of the gas luminosity. The analysis by Markevitch (1998) on clusters with strong cooling flows indicates moderate temperature increase of up to 20\\% but luminosity decrease of up to 40\\% for the cluster gas after excising the cooling flow regions. On the other hand, if cooling by metals is considered -- for sake of simplicity, pure bremsstrahlung has been assumed -- the luminosity of the gas could decrease by about 40\\%, assuming a Raymond-Smith model with metal abundance Z=0.35. Still, the derived thermal luminosity in 3C382 is about two order of magnitude larger than that found in isolated, normal elliptical galaxies (Canizares, Fabbiano \\& Trincheri 1987), and in low-power radio galaxies (Worral \\& Birkinshaw, 1994) but it is in the range found in powerful radio sources (Worrall et al. 1994; O'Dea et al. 1996; Hardcastle et al. 1999; Crawford et al. 1999). Also, the estimated core radius for 3C382, $\\sim$50 kpc, is within the range found by Crawford et al. and Hardcastle et al. in their respective samples of 3CRR radio-loud sources. Large X-ray halos are often seen in FRI sources, those being associated with the cluster environment in which they often reside (e.g. M~87, Perseus, 3C465). These halos largely dominate the ROSAT emission from these sources. Besides the outstanding case of Cygnus A, evidence for clustering is less obvious in classical double radio sources, particularly at low redshift (cf. Hill \\& Lilly 1991; Miller et al. 1999). Unambiguous extended X-ray emission in powerful FRII radio galaxies and quasars has mostly being found in sources with redshift larger than 0.1 (Hardcastle and Worall, 1999; Crawford et al 1999; O'Dea et al. 1996); yet, a few low-redshift FRII are reported to show extended X-ray emission (cf. Hardcastle and Worral). In most of these cases, the large X-ray luminosities are found compatible with thermal emission from a moderately rich cluster environment. Comparing with Cyg A, the archetypal double radio source at z= 0.0574, 3C382 is also one of the few very bright double sources at low redshift with extended X-ray emission. Contrarily to Cyg A which presents an optical narrow line spectrum, 3C382 presents an extreme, in width and strength, broad permitted line spectrum. If this difference is interpreted as due to obscuration of the AGN region in Cyg A, it may explain why the dominant X-ray feature in Cyg A is emission from a hot diffuse gas --the AGN component is obscured at X-ray waves-- whereas in 3C382, the unresolved X-ray nuclear component --presumably associated with the AGN-- dominates the total X-ray emission, making more difficult the detection of any extended gas component. The X-ray luminosity of 3C382, of about $10^{44}erg~s^{-1}$, compares with that of rich Abell clusters (this is also the case of Cyg A; yet Cyg A is at the center of a poor cluster of galaxies). Longair \\& Seldner (1979) derived however a rather poor environment in the vicinity of 3C~382 on the basis of their cross-correlation analysis between the radio position and galaxy counts. HST/WFPC2 images of 3C382 collected in parallel mode show an elliptical galaxy with a very bright unresolved nucleus and a halo very smooth (Martel et al. 1999). Yet, within the 2.5 arcminutes field of view (Fig 1b), several small galaxies can easily be distinguished in the 300 seconds exposure; the WFPC2 images also show a bright galaxy at 85 arcsec Northeast from 3C382, presenting two at least extended gaseosus tails of material in the direction of 3C382; two additional difuse regions located close to 3C382 and in the direction of the bright galaxy are also apparent. Judging from the HST images, 3C382 may be residing in a relatively poor cluster environment; also, it may be in interaction with that gas-rich galaxy companion. Such interaction could have brought plenty of gas into 3C382. The luminosity of the halo component in 3C382 would imply a large mass of gas, of about $10^{11}$Mo, assuming it concentrated in a sphere of about 50 kpc --the estimate core radius derived from the HRI spatial analysis-- and a temperature in the 0.6 -1 keV range. An alternative to the cluster environment is that 3C382 it may consist of a self-contained gravitational potential deep enough to restrain such large amount of gas. This could also be the case of the radio-louds 3C48 and 3C273, for which extended X-ray emission is found but the evidence for a cluster environment from optical images is minor (Crawford et al. 1999). Evidence for a massive dark halo in 3C382 comes from the velocity measurements on the extended ionized gas surrounding this galaxy. Tadhunter et al. (1986) detected ionized gas up to 25 kpc from the galaxy center. The gas follow a a rotation curve which extend flat up to those distances with velocities of about 400 km/s relative to the systemic velocity. Assuming a spherical potential, the estimated mass within a 25 kpc radius would be $\\sim8\\times 10^{11}Mo$. This is about the gravitating mass needed to keep the X-ray gas binded to the galaxy. Following Fornan, Jones and Tucker (1985) formalism, the total gravitating mass within a 50 kpc radius is estimated between $8-15 \\times10^{11}Mo$ for gas temperatures between 0.6 and 1 keV, which is in the order of magnitude of the mass derived from the ionized gas kinematics. The results so far derived show compatible with a cooling flow process being dominating the gas emission. If the extended gas emission is modeled as that of a uniform sphere of hot gas emitting via bremsstrahlung, for a maximum radius of about 170 kpc (the size of the radio structure) and a temperature in the 0.6 - 1 keV range, the implied density would be $simeq 6 \\times 10^{-3}cm^{-3}$. This yields a cooling time of about $4\\times 10^{9} yr$, considerable smaller than the Hubble time. Thus, a cooling flow process could be operating in 3C~382. A better characterization of the extended X-ray emission in 3C382 would demand much larger spatial resolution but also deeper observations. \\\\ \\bigskip {\\bf Acknowledgments: } It is a pleasure to thank Peter Predehl for his support regarding the de-speckling procedure of the HRI data and G\\\"unther Hasinger and Paddy Leahy for critical reading of early versions of the manuscript. \\bigskip" }, "0003/astro-ph0003303_arXiv.txt": { "abstract": "We present a progress report on the design and construction of the Field-Imaging Far-Infrared Line Spectrometer (FIFI LS) for the SOFIA airborne observatory. The design of the instrument is driven by the goal of maximizing observing efficiency, especially for observations of faint, extragalactic objects. Thus, FIFI LS utilizes an integral field technique that uses slicer mirrors to optically re-arrange the two-dimensional field into a single slit for a long slit spectrometer. Effectively, a 5 $\\times$ 5 pixel spatial field of view is imaged to a 25 $\\times$ 1 pixel slit and dispersed to a 25 $\\times$ 16 pixel, two-dimensional detector array, providing diffraction-limited spatial and spectral multiplexing. In this manner, the instrument employs two parallel, medium resolution (R $\\sim$2000) grating spectrometers for simultaneous observations in two bands: a short wavelength band (42 to 110 $\\mu$m) and a long wavelength band (110 to 210 $\\mu$m). Overall, for each of the 25 spatial pixels, the instrument can cover a velocity range of $\\sim$1500 km/s around selected far-infrared lines with an estimated sensitivity of $2\\times 10^{-15}$ W Hz$^{1/2}$ per pixel. This arrangement provides good spectral coverage with high responsivity. {\\bf This paper does not include Figures due to astro-ph size limitations. Please download entire file at http://fifi-ls.mpe-garching.mpg.de/fifils.ps.gz.} ", "introduction": "\\label{sect:intro} % Astronomers will soon have access to unprecedented spatial resolution and sensitivity in the far-infrared with the Stratospheric Observatory For Infrared Astronomy (SOFIA). Far-infrared astronomical observations, which are impossible from the ground due to water absorption, are necessary to understand fully a number of important astronomical problems and issues. Many astrophysical conditions require far-infrared probes because the areas of interest are mostly inaccessible at other wavelengths due to severe extinction from interstellar dust or the physics of interest is only manifest at far-infrared wavelengths. In particular, far-infrared spectroscopy, pioneered and developed on SOFIA's predecessor the Kuiper Airborne Observatory (KAO) and greatly extended with the Infrared Space Observatory (ISO), will provide an injection of important data into astrophysical issues. Building upon the success of our previous imaging Fabry-Perot far-infrared spectrometer (FIFI\\cite{fifi1,fifi2}) that was designed for the KAO, we are developing a successor instrument for SOFIA: the Field-Imaging Far-Infrared Line Spectrometer (FIFI LS\\cite{us1,us2}). FIFI LS will utilize integral field spectral imaging in two wavelength bands: 42 to 110 $\\mu$m and 110 to 210 $\\mu$m. This will allow the instrument to simultaneously obtain dual wavelength band, spectral imaging of a 5 $\\times$ 5 pixel field, without scanning a Fabry-Perot or multiple pointings with a long-slit spectrometer. Thus, FIFI LS is designed as a ``major-step'' forward and will take advantage of the unique benefits that SOFIA offers. With the increased sensitivity and resolution provided by SOFIA, the main scientific targets for FIFI LS will include the detailed morphological studies of: (1) the heating and cooling of galaxies, (2) star formation and the interstellar matter under low-metalicity conditions, as found in dwarf galaxies, (3) active galactic nuclei and their environment, (4) merging and interacting galaxies, and (5) large surveys of nearby galaxies. To reach our scientific goals, very high observing sensitivities and efficiencies are essential, requiring a compromise of spectral resolution; however, for the science objectives listed above, a comparably low spectral resolution (R $\\sim$2000) is more than sufficient. Overall, FIFI LS on SOFIA will be more sensitive than the ISO Long-Wavelength Spectrometer and have much higher spatial resolution and mapping capabilities. As a future option, an extension of the instrument to the 25-42 $\\mu$m range is planned upon availability of the Si:Sb detector arrays developed for SIRTF. ", "conclusions": "" }, "0003/astro-ph0003073_arXiv.txt": { "abstract": "\\noindent In this paper I outline a new strategy for the routine detection and imaging of faint (sub-mJy and microJy) radio sources with VLBI and SVLBI. The strategy relies on a combination of in-beam phase-referencing, wide-field VLBI imaging and simultaneous correlation of multiple field centres. A combination of these techniques, together with the steeply rising radio source counts observed at cm wavelengths, permit routine high resolution observations of radio sources previously considered too faint for conventional VLBI. ", "introduction": "VLBI is sensitivity limited. Most sources that can be robustly detected by conventional self-calibration techniques have peak fluxes in excess of 10~mJy. The success of phase-referencing techniques, as applied to mJy and a few sub-mJy radio sources, are often limited (particularly in terms of image fidelity) to the brighter sources for which subsequent self-calibration (over much longer solution intervals) is then possible. So far, few attempts have been made to detect sub-mJy sources, despite the fact that with a coherent integration time of 24 hours, global VLBI arrays can routinely produce images with $1\\sigma$ rms noise levels better than $30\\mu$Jy/beam. Nevertheless, the focus of VLBI over the last 3 decades (and in particular Space VLBI -- SVLBI) has been directed towards the study of the brightest and most compact radio sources in the sky. At these flux levels ($> 10$mJy), the radio sky is virtually empty, with most radio sources associated with relatively distant AGN. As a result the overlap with other wave-bands is sometimes limited. In this paper, I suggest a new strategy for the {\\it routine\\/} detection and imaging of faint sub-mJy and $\\mu$Jy radio sources. The strategy relies on a combination of in-beam phase-referencing (with obvious advantages for SVLBI but also VLBI generally - see Fomalont et al. 1999), wide-field VLBI imaging (see Garrett et al. 1999) and simultaneous correlation of multiple field centres. These techniques, together with the steeply rising radio source counts at $\\lambda$cm wavelengths, should permit high resolution, VLBI investigations of the faint sub-mJy and microJy source populations to begin. ", "conclusions": "" }, "0003/astro-ph0003245_arXiv.txt": { "abstract": "A blind survey for \\ion{H}{1} bright galaxies in the southern Zone of Avoidance, ($212^\\circ \\le \\ell \\le 36^\\circ$, $|b|$~$\\leq$~5$^\\circ$), has been made with the 21~cm multibeam receiver on the Parkes 64~m radiotelescope. The survey, sensitive to normal spiral galaxies to a distance of $\\sim$ 40 Mpc and more nearby dwarfs, detected 110 galaxies. Of these, 67 have no counterparts cataloged in the NASA/IPAC Extragalactic Database. In general, the uncataloged galaxies lie behind thicker obscuration than do the cataloged objects. All of the newly-discovered galaxies have \\ion{H}{1} flux integrals more than an order of magnitude lower than the Circinus galaxy. The survey recovers the Puppis cluster and foreground group (Kraan-Korteweg \\& Huchtmeier 1992), and the Local Void remains empty. The \\HI mass function derived for the sample is satisfactorily fit by a Schechter function with parameters $\\alpha$ = 1.51 $\\pm$ 0.12, $\\Phi^*$ = 0.006 $\\pm$ 0.003, and log M$^*$ = 9.7 $\\pm$ 0.10. ", "introduction": "The obscuration due to dust and the high stellar density in our Galaxy varies from place to place within the Milky Way. Overall, it blocks our optical view of the extragalactic Universe over $\\sim$20\\% of the sky, somewhat less in the infrared. This ``Zone of Avoidance\" (ZOA) was recognized even before the nature of the spiral nebulae themselves was understood. This sky coverage limitation does not pose a problem for the study of galaxies themselves, as there is no reason to believe that the population of obscured galaxies should differ from those in optically unobscured regions. However, to understand the Local Group's motion requires mapping the surrounding mass inhomogeneity, measured in practice by galaxy over- and under-densities, across the entire sky. In particular, the lack of a full census of nearby, hidden galaxies is troublesome, since the local galaxies should play a significant role in the Milky Way's motion with respect to the microwave background (Kraan-Korteweg 1993). The ZOA has been successfully narrowed by deep searches in the optical and infrared (see Kraan-Korteweg \\& Woudt 1999 for a comprehensive review of the various efforts). However, optical searches fail where the extinction exceeds 4 - 5 magnitudes, within about $|b| \\la 5^\\circ$ of the Galactic plane. Near-infrared surveys, e.g. 2MASS (Skrutskie \\etal 1997) and DENIS (Epchtein 1997), will eventually produce catalogs of galaxies closer to the plane than is possible with optical searches, but they do not recover the most heavily obscured galaxies, or galaxies of low surface brightness (Schr\\\"{o}der, Kraan-Korteweg, \\& Mamon 1999). Far-infrared surveys become confusion limited by Galactic sources at low latitudes, and the remaining ZOA still covers $\\sim10~\\%$ of the sky. Galaxies which contain \\ion{H}{1} can be found in the regions of thickest obscuration and IR confusion. The technique was pioneered over a decade ago by Kerr \\& Henning (1987) who showed through a small, pilot survey that completely optically-hidden galaxies could be readily uncovered through the detection of their 21-cm emission. Since then, a spatially complete survey for spirals out to 4000 \\kms has been conducted over the northern ZOA ($30^\\circ \\le \\ell \\le 220^\\circ$, $|b|$~$\\leq$~5$^\\circ$; rms noise 40 mJy beam$^{-1}$) with the 25~m Dwingeloo telescope (Henning \\etal 1998, Rivers \\etal 1999). The survey uncovered no massive, ``Andromeda\"-type galaxy in the ZOA, indeed the nearest, previously unknown galaxy revealed by the survey was Dwingeloo 1, a likely member of the IC342/Maffei group (Kraan-Korteweg \\etal 1994). The census for nearby, \\ion{H}{1}-bearing galaxies in the northern ZOA is complete, at least for those galaxies whose redshifts or blueshifts are sufficient to separate their \\ion{H}{1} signals from Galactic 21-cm emission, at $0$~$\\pm$~$\\sim$100 \\kms. We report here on a somewhat deeper survey (rms noise 15 mJy beam$^{-1}$) for \\HI galaxies in the southern ZOA, conducted with the new multibeam receiver on the 64~m Parkes radiotelescope \\footnote{The Parkes telescope is part of the Australia Telescope which is funded by the Commonwealth of Australia for operation as a National Facility managed by CSIRO.} . The angular coverage ($212^\\circ \\le \\ell \\le 36^\\circ$, $|b|$~$\\leq$~5$^\\circ$) completes the survey of the great circle of the ZOA for relatively nearby, dynamically important HI galaxies. The present survey discussed here (the ``shallow survey\") represents the first stage of an ongoing deeper search of the area with the multibeam system. The shallow survey is comprised of the first two scans of a planned 25 scans of the southern ZOA, estimated to be completed in mid-2000. (This full sensitivity survey will be sensitive to spirals to a redshift of $\\sim$ 10,000 \\kms). In addition to the astronomical motivation outlined above, the shallow survey serves as a testbed of techniques for the full sensitivity survey. An intermediate-depth survey consisting of four scans of the region $308^\\circ \\le \\ell \\le 332^\\circ$; $|b|$~$\\leq$~5$^\\circ$, has been conducted (Juraszek \\etal 2000). This region is of particular interest as it contains the predicted position of the core of the Great Attractor (Kolatt, Dekel, \\& Lahav 1995). In \\S~2, the observations and data reduction will be described. The search method and galaxy \\HI parametrization procedure will be outlined in \\S~3. The resulting catalog is presented in \\S~4. Discussion of the galaxy distribution at low Galactic latitudes, the \\HI mass function derived for the sample, and predictions for the full sensitivity survey are contained in \\S~5. ", "conclusions": "\\subsection{Nearby, \\HI Bright Galaxies in the Southern ZOA} The shallow survey's rms noise of 15 mJy is equivalent to a 5$\\sigma$ HI mass detection limit of $4 \\times 10^6$ d$^2\\!\\!_{\\rm Mpc}$ M$_{\\odot}$ (for a galaxy with the typical linewidth of 200 \\kms). Thus, the sensitivity to normal spirals falls rapidly beyond about 40 Mpc, and the survey is not well suited to discuss large-scale structure behind the southern Milky Way beyond a few tens of Mpc. The full survey will be sensitive to spirals to a much larger redshift, $\\sim$10,000 \\kms, and will be able to address issues such as the Great Attractor, predicted to lie at (l, b, v) $\\sim$ (320$^\\circ$, 0$^\\circ$, 4500 \\kms) (Kolatt, Dekel, \\& Lahav 1995), and other features which may lie hidden (see Kraan-Korteweg \\& Woudt 1999 for a review of these structures, and Kraan-Korteweg \\& Juraszek 2000 for preliminary analysis of multibeam full-sensitivity survey detections in the Great Attractor region). However, the shallow survey does now complete the census of nearby, HI bright galaxies which lie at redshifts away from Galactic HI. Sixteen of the 110 galaxies lie at redshifts less than 1000~\\kms and are, therefore, fairly nearby. However, all of the shallow survey galaxies have HI flux integrals an order of magnitude or more below that of the Circinus galaxy, a nearby, massive, low-Galactic latitude galaxy, which is cataloged here as HIZSS 086. One can estimate the dynamical importance of each hidden galaxy without knowledge of its inclination or precise distance by realizing that if the galaxies in the catalog have similar M$_{\\rm HI}$ / M$_{\\rm total}$, the gravitational force on the Milky Way due to one of these galaxies is proportional to its HI flux integral. The closest rival in the catalog to Circinus is ESO 494-G026 (Lauberts 1982), listed here as HIZSS 032. It has only about 10\\% of Circinus's HI flux. (Dynamical analysis of this object to estimate its total mass will be conducted using HI synthesis data obtained with the VLA). All of the newly discovered galaxies have even lower HI flux integrals. No single, previously unknown, dynamically important HI galaxy was found. Five of the 110 galaxies are extended objects at the multibeam resolution, and of these, three were previously cataloged: Circinus, ESO 494-G026, and IRAS 15109-5248. 21-cm follow-up synthesis mapping has been done for IRAS 15109-5248 and the two uncataloged, extended objects, HIZSS 097 and HIZSS 102 by Staveley-Smith \\etal (1998). Rotation curve analysis indicates the IRAS galaxy is a massive disk system, estimated M$_{\\rm tot}$ = 6~$\\times$ 10$^{11}$ M$_\\odot$. The other two are moderate to low mass systems, which under the higher resolution scrutiny of the synthesis observations (synthesized beam FWHM $\\sim$ 2.2 - 4.5 arcmin) were seen to break up into apparently interacting systems of low \\HI column density. \\HI synthesis observations of all of the shallow survey detections using the Australia Telescope Compact Array and the NRAO Very Large Array have been completed or are planned. Identifications on the currently available strips of the southern sky NIR survey DENIS (DEep NIR southern sky Survey; Epchtein 1997) show that many of the shallow survey sources are visible on the NIR images (Schr\\\"{o}der, Kraan-Korteweg, \\& Mamon 1999; Schr\\\"{o}der \\etal in prep.). \\subsection{Large-Scale Structure in the Shallow Survey} Figure 2 shows the distribution on the sky of the shallow survey detections, and galaxies from the literature within a redshift of 4000 \\kms. \\begin{figure} \\psfig{file=henning.fig2.eps,width=17cm} \\caption{Distribution on the sky of cataloged galaxies from the LEDA database within 4000 \\kms (small dots) and shallow survey detections (larger dots). The shallow survey search area is delineated by the dashed rectangle.} \\label{fig-2} \\end{figure} The shallow survey now fills in the southern ZOA within about 40 Mpc for the first time, with minimal dependence of detection-rate on Galactic latitude (Fig. 3). \\begin{figure} \\psfig{file=henning.fig3.eps,width=17cm} \\caption{Number of shallow survey detections as a function of Galactic latitude.} \\label{fig-3} \\end{figure} The Local Void is clearly evident in the distribution of both the optically-cataloged objects, and the shallow survey detections. It is also apparent in Figure 4, the longitude-velocity distribution of the HI detected objects. \\begin{figure} \\psfig{file=henning.fig4.eps,width=17cm} \\caption{Distribution in Galactic longitude and redshift of the 110 shallow survey detections.} \\label{fig-4} \\end{figure} The three shallow survey objects on the border of the Local Void at l $\\sim$ 30$^\\circ$ all have redshifts $\\sim$ 1500 \\kms, consistent with their being members of the group proposed by Roman \\etal (1998). Two of these galaxies lie within the survey boundary of the Dwingeloo Obscured Galaxies Survey, and were also detected by this survey (Rivers, Henning, \\& Kraan-Korteweg 1999). The most obvious overdensity of the low latitude galaxies apparent in Figure 2 occurs at l $\\sim$ 245$^\\circ$. Figure 4 reveals the overdensity is caused by two groupings of galaxies, at (l, v) $\\sim$ (245$^\\circ$, 800 \\kms) and (l, v) $\\sim$ (245$^\\circ$, 1500 \\kms). The latter corresponds to the location of the moderately obscured, nearby cluster in Puppis (l, b, v) $\\sim$ (245$^\\circ$, 0$^\\circ$, 1500 \\kms) (Kraan-Korteweg \\& Huchtmeier 1992). This overdensity of galaxies was deduced by Scharf \\etal (1992) through a spherical harmonic analysis of IRAS galaxies. The nearer group was also recovered by Kraan-Korteweg \\& Huchtmeier (1992). There is no evidence of any other nearby, unrecognized clusters of galaxies in the survey region. Except near the Galactic center where continuum emission decreased the sensitivity of the survey somewhat, the survey was uniformly sensitive to spirals with M$_{\\rm HI}$ = $2 \\times 10^9$ M$_{\\odot}$ at 1500 \\kms (H$_0$ = 75 \\kms Mpc$^{-1}$) across the entire southern ZOA, so any hidden overdensities within this redshift range should have been easily detected. Structures at v $\\sim$ 4000 \\kms are not well probed by this survey, however we can see the relative overdensity of galaxies toward the general Great Attractor region, and galaxy underdensity behind the Puppis cluster, broadly consistent with the theoretical mass density reconstructions of Kolatt, Dekel, \\& Lahav (1995) and Webster, Lahav, \\& Fisher (1997). The deep survey should definitively confirm or refute the predictions of these and other mass density reconstructions in the ZOA. \\subsection{\\HI Mass Function} To determine the number density of galaxies as a function of \\HI mass (the \\HI mass function), the sensitivity of the survey must be carefully determined to be able to apply appropriate volume corrections. Near strong continuum sources, galaxies which would otherwise be detected are hidden in the increased noise. Over the search area, $212^\\circ \\le \\ell \\le 36^\\circ$, $|b|$~$\\leq$~5$^\\circ$, 7.7\\% of the data were disturbed by strong continuum emission, and had rms noise fluctuations a factor of three or more above the quoted sensitivity of 15 mJy. In the volume correction calculations, we take the survey area to be 92.3\\% of the area covered by the telescope. The galaxy selection function for the shallow survey is not based simply on peak flux, but is determined empirically to involve both total flux and linewidth in the following way: $\\int$ S dv $\\times$ W$_{50}^{-0.7}$ $>$ 0.3. Schneider, Spitzak, \\& Rosenberg (1998) find a very similar functional form for the completeness limit of two previous blind 21-cm surveys done at Arecibo, $\\int$ S dv $\\times$ W$_{50}^{-0.75}$. With this description, the maximum volume in which each galaxy could be detected is calculated. The average value of V / V$_{max}$ = 0.58, is quite close to the theoretical value of 1/2 for a correctly determined completeness. The resulting number density of \\HI masses with each galaxy weighted by the inverse of its maximum detectable volume is shown in Figure 5. \\begin{figure} \\psfig{file=fig5preprint.ps,width=17cm} \\caption{\\HI mass function determined from the shallow survey detections. The errorbars on the points are determined by counting statistics. The function is satisfactorily fit by a Schechter function with $\\alpha$ = 1.51 $\\pm$ 0.12, $\\Phi^*$ = 0.006 $\\pm$ 0.003, and log M$^*$ = 9.7 $\\pm$ 0.10, shown by the solid curve. The dashed curve shows an \\HI mass function derived by Zwaan \\etal (1997) for a smaller \\HI -selected sample of galaxies.} \\label{fig-5} \\end{figure} This \\HI mass function is well fit by a Schechter function (Schechter 1976) with parameters $\\alpha$ = 1.51 $\\pm$ 0.12, $\\Phi^*$ = 0.006 $\\pm$ 0.003, and log M$^*$ = 9.7 $\\pm$ 0.10. Note that the Schechter function was not used as an {\\it a priori} assumption of the shape of the \\HI mass function. The low-mass slope is significantly steeper than the $\\alpha$ = 1.20 suggested by Zwaan \\etal (1997). The densities at the lowest mass bins are consistent with the higher densities found by Schneider, Spitzak, \\& Rosenberg (1998). The main difficulty remains small number statistics, with the three lowest mass bins of Figure 5 containing a total of six galaxies. With the increased search volume of the HIPASS and full sensitivity ZOA surveys, the statistical robustness of the mass function parametrization will be improved. \\subsection{Predictions for the Full Sensitivity Survey} We now estimate how many galaxies should be uncovered by the full sensitivity ZOA survey, to be completed in the year 2000. The deep survey will consist of 435 scans at each longitude, compared with the 34 of the shallow survey. This factor of 12.5 increase in integration time will lead to a $\\sqrt{12.5}$ = 3.5 improvement in sensitivity. For a galaxy of a given M$_{\\rm HI}$ and velocity linewidth, the distance to which it could be detected will increase by a factor of $\\sqrt{3.5}$. Thus, the volume increase of the deep survey over the shallow survey is about a factor of 6.5, leading to a rough estimate of $6.5 \\times 110 \\approx 700$ galaxies to be detected by the deep survey. Indeed, a portion of the deep survey is completed, and the four data cubes in the region of the supposed Great Attractor have been inspected, and about 300 galaxy candidates noted (S. Juraszek, private communication). Extrapolation over the full spatial extent of the survey leads to an estimate of about 1700 galaxies. However, this region of space contains a significant overdensity of galaxies, and much of the final survey volume will contain the Local Void, so the total tally may be closer to 1000 galaxies. Efforts are underway to develop software tools which model and remove strong continuum sources from the data, which should decrease the effective noise even further, increasing the survey sensitivity and the number of detected objects." }, "0003/hep-ph0003016_arXiv.txt": { "abstract": "Following an argument by Kirzhnits we rederive an exact expression for the energy loss of a fast charged particle in a relativistic plasma using the quantum field theoretical language. We compare this result to perturbative calculations of the collisional energy loss of an energetic electron or muon in an electron-positron plasma and of an energetic parton in the quark-gluon plasma. ", "introduction": " ", "conclusions": "" }, "0003/hep-th0003052_arXiv.txt": { "abstract": "We study a Randall-Sundrum cosmological scenario consisting of a domain wall in anti-de Sitter space with a strongly coupled large $N$ conformal field theory living on the wall. The AdS/CFT correspondence allows a fully quantum mechanical treatment of this CFT, in contrast with the usual treatment of matter fields in inflationary cosmology. The conformal anomaly of the CFT provides an effective tension which leads to a de Sitter geometry for the domain wall. This is the analogue of Starobinsky's four dimensional model of anomaly driven inflation. Studying this model in a Euclidean setting gives a natural choice of boundary conditions at the horizon. We calculate the graviton correlator using the Hartle-Hawking ``No Boundary'' proposal and analytically continue to Lorentzian signature. We find that the CFT strongly suppresses metric perturbations on all but the largest angular scales. This is true independently of how the de Sitter geometry arises, i.e., it is also true for four dimensional Einstein gravity. Since generic matter would be expected to behave like a CFT on small scales, our results suggest that tensor perturbations on small scales are far smaller than predicted by all previous calculations, which have neglected the effects of matter on tensor perturbations. ", "introduction": "\\label{app:A} \\setcounter{equation}{0} This appendix demonstrates how a metric perturbation on the boundary of a ball of AdS is decomposed into vector, scalar and tensor components. \\medskip Consider a ball of perturbed AdS with a spherical boundary. Let $\\bar{l}$ be the AdS length scale. Gaussian normal coordinates are introduced by defining $\\bar{l} y$ to be the geodesic distance of a point from the origin. The surfaces of constant $y$ are spheres on which we introduce coordinates $x^i$. In these coordinates the metric takes the form \\be ds^2 = \\bar{l}^2(dy^2 + \\sinh^2 y \\hat{\\gamma}_{ij}(x) dx^i dx^j) + h_{ij}(y,x) dx^i dx^j. \\ee The ball of AdS has been perturbed, so the boundary will be at a position $y=y_0 + \\xi(x)$. Let the induced metric perturbation on the boundary be $\\hat{h}_{ij} (x)$. This can be decomposed into scalar, vector and tensor perturbations with respect to the round metric on the sphere \\cite{stewart}: \\be \\label{eqn:decomp} \\hat{h}_{ij}(x) = \\hat{\\theta}_{ij} + 2\\hat{\\nabla}_{(i} \\hat{\\chi}_{j)} + \\hat{\\nabla}_i \\hat{\\nabla}_j \\hat\\phi + \\hat{\\gamma}_{ij} \\hat{\\psi}, \\ee where we use hats to denote quantities defined on the sphere (i.e. quantities that depend only on $x$). $\\hat{\\theta}_{ij}$ is a transverse traceless tensor on the sphere and $\\hat{\\chi}_i$ is a transverse vector on the sphere. $\\hat{\\phi}$ and $\\hat{\\psi}$ are scalars on the sphere. $\\hat{\\chi}_i$ and $\\hat{\\phi}$ can be gauged away by infinitesimal coordinate transformations on the sphere of the form $x^i = \\tilde{x}^i - \\eta^i(\\tilde{x})-\\partial^i \\eta(\\tilde{x})$ where $\\eta^i$ is transverse. Therefore we shall assume that $\\hat{\\chi}$ and $\\hat{\\phi}$ vanish. Note that it is not possible to gauge away $\\hat{\\psi}$ or $\\xi$. This paper only deals with tensor perturbations so we shall assume that the scalars $\\hat{\\psi}$ and $\\xi$ are vanishing. The induced metric perturbation is then transverse and traceless and can be extended into the bulk as described in section \\ref{sec:CFT}. The scalars will be discussed in our next paper. ", "conclusions": "" }, "0003/astro-ph0003465_arXiv.txt": { "abstract": "We report on a new survey for $z\\approx 4.5$ \\lya\\ sources, the Large Area Lyman Alpha (LALA) survey. Our survey achieves an unprecedented combination of volume and sensitivity by using narrow-band filters on the new $8192^2$ pixel CCD Mosaic Camera at the 4 meter Mayall telescope of Kitt Peak National Observatory. Well-detected sources with flux and equivalent width matching known high redshift \\lya\\ galaxies (i.e., observed equivalent width $\\eqw > 80$\\AA, $2.6 < \\hbox{(line + continuum flux)}/ (10^{-17}\\ergcm2s) < 5.2$, and $\\delta(\\hbox{EW}) / \\hbox{EW} < 0.25$) have an observed surface density corresponding to $11000 \\pm 700$ per square degree per unit redshift at $z=4.5$. Spatial variation in this surface density is apparent on comparison between counts in $6561 \\pm 40$\\AA\\ and $6730 \\pm 40$\\AA\\ filters. Early spectroscopic followup results from the Keck telescope included three sources meeting our criteria for good \\lya\\ candidates. Of these, one is confirmed as a $z=4.52$ source, while another remains consistent with either $z=4.55$ or $z=0.81$. We infer that $30$ to $50\\%$ of our good candidates are {\\it bona fide\\/} \\lya\\ emitters, implying a net density of $\\sim 4000 $ \\lya\\ emitters per square degree per unit redshift. ", "introduction": "More than three decades ago Partridge and Peebles (1967) predicted that galaxies in their first throes of star-formation should be strong emitters in the \\lya\\ line. Their predictions were optimistic, based on converting roughly 2\\% of gas into stars in $3 \\times 10^7$ years in Milky Way sized galaxies, which translates into a luminosity of $6\\times 10^{44} \\erg\\, \\sec^{-1}$. These objects are also expected to be common - if all the $L^*$ galaxies have undergone a phase of rapid star-formation one should see a surface density of about $3 \\times 10^3 \\times (\\Delta t/(3 \\times 10^7 \\year)) \\deg^{-2}$ (Pritchet 1994). Searches based on these expectations did not detect \\lya\\ emitters (LAEs). (See review by Pritchet 1994; Koo \\& Kron 1980; Pritchet \\& Hartwick 1987, 1990; Cowie 1988; Rhea et al 1989; Smith et al 1989; Lowenthal et al 1990; Wolfe et al 1992; De Propris et al 1993; Macchetto et al 1993; M{\\o}ller \\& Warren 1993; Djorgovski \\& Thompson 1992; Djorgovski, Thompson, \\& Smith 1993; Thompson, Djorgovski, \\& Trauger 1992; Thompson et al 1993; Thompson, Djorgovski, \\& Beckwith 1994; Thommes et al 1998.) Only recently have \\lya\\ emitters been observed, albeit at luminosity levels roughly a hundred times lower than the original prediction. These \\lya\\ emitters have been found from both deep narrow band imaging surveys (Cowie \\& Hu 1998; Hu, Cowie \\& McMahon 1998; Hu, McMahon, \\& Cowie 1999; Kudritzki et al 2000), and from deep spectroscopic surveys (Dey et al 1998; Manning et al 2000; but see Stern et al 2000). Weak \\lya\\ emitters have also been found through targeted spectroscopy of Lyman break objects (e.g., Steidel et al 1996, Lowenthal et al 1997). The lower luminosity in the \\lya\\ line may be because of attenuation by dust if chemical enrichment is prompt; or because the star-forming phase is more protracted; or because the star-formation happens in smaller units which later merge. The first two scenarios will give a smaller equivalent width than early predictions, while the last scenario results in low luminosities but high equivalent width. Dust effects are expected to be severe--- even a small amount of dust can greatly attenuate this line, because it is resonantly scattered. However, two factors can help the \\lya\\ photons escape. If \\lya\\ photons are produced in diffuse regions of a clumpy interstellar medium, they can simply scatter off the dense clumps and escape (Neufeld 1991), and some geometries can even lead to an increase in the equivalent width of the line. Secondly, energetic winds are seen in low-$z$ \\lya\\ emitters (Kunth et al 1998). These can displace the neutral gas and doppler-shift the peak wavelength of the resonant scattering, thereby reducing the amount of scattering and the path length for interaction with dust. Detailed predictions for luminosities and surface densities of LAEs using a Press-Schechter formalism and exploring a range of dust obscuration and star-formation time scales have been explored by Haiman and Spaans (1999), who are able to reproduce the surface densities of LAEs reported by Hu et al (1998) with a wide range of models. In order to narrow down the range of possibilities and characterize the high redshift \\lya\\ population, better statistics over a wide range of flux and source density are needed. ", "conclusions": "By combining our imaging survey with these spectroscopic results, we can estimate the source density of \\lya\\ emitters passing our selection cut. Our spectra included three sources fulfilling the criteria given above for good candidates. Of these, one was confirmed as a $z=4.52$ \\lya\\ source. A second remains a candidate $z=4.55$ source, but is more conservatively interpreted as a $z=0.81$ [O II] emitter on the basis of a rather strong continuum on the blue side of the line. The third is a clear $z=0.34$ [O III] emitter. We therefore estimate that roughly $1/3$ to $1/2$ of the good candidates will be confirmed as \\lya\\ sources, yielding $\\sim 4000$ emitters per square degree per unit redshift. This is compatible with earlier measurements from smaller volumes (Hu et al 1998) after accounting for differences in flux threshold. Our measurement is distinct from previous efforts in the field for its basis in a large number of candidate emitters. Poisson errors in our source counts are of order $\\pm 7\\%$. This is smaller than the variations observed in the comparison of two filters (of order $\\pm 40\\%$). By combining observations in multiple fields, we will be able to average over local fluctuations in number densities effectively. When completed, the LALA survey will yield comoving volume of $\\sim 2 \\times 10^6 \\Mpc^3$ (\\S \\ref{obsdesc}) and a sample of several hundred LAEs, and will allow the luminosity function, equivalent width distribution, and correlation function of this population to be determined for the first time." }, "0003/astro-ph0003379_arXiv.txt": { "abstract": "We report the detection of linear polarization from Sgr~A$^{\\star}$ at 750, 850, 1350 and 2000$\\mu$m which confirms the contribution of synchrotron radiation. From the lack of polarization at longer wavelengths it appears to arise in the millimetre/sub-millimetre excess. There are large position angle changes between the millimetre and sub-millimetre results and these are discussed in terms of a polarized dust contribution in the sub-millimetre and various synchrotron models. In the model which best explains the data the synchrotron radiation from the excess is self-absorbed in the millimetre region and becomes optically thin in the sub-millimetre. This implies that the excess arises in an extremely compact source $\\sim$ 2 Schwarzschild radii. ", "introduction": "Observations of stellar proper motions in the vicinity of Sgr~A$^{\\star}$ (Ghez et al. 1998), the non-thermal radio source at the apparent centre of the Galaxy, show that its mass of $\\sim 2.5 \\times 10^{6}~$M$_{\\odot}$ is highly compact on a scale $<0.01$pc, reinforcing its claim to be a massive black-hole candidate. The spectrum of Sgr~A$^{\\star}$ extends as a rough power law $\\sim \\nu^{\\alpha}$, where $\\alpha$ lies between $\\onequarter$ and $\\onethird$, from a low-frequency turn-over at a few GHz up to $\\sim$ 100GHz, (e.g. Mezger, Duschl and Zylka 1996 and references therein), and above this frequency there is evidence of a mm to sub-mm excess over this power law extending almost to the atmospheric cut-off near 1000 GHz (Serabyn et al. 1997, Falcke et al. 1998). The nature of the mm/sub-mm excess has been discussed by Serabyn et al. (1997) and the possibility that it is due to dust is effectively eliminated: at 1.3mm its size is less than 1 arcsecond with an implied brightness temperature in excess of 100 K, inconsistent with the shorter wavelength data, and the implication is that there are two separate synchrotron components. The radio spectrum of Sgr~A$^{\\star}$ has for some years been modelled in various ways in terms of synchrotron radiation but searches for linear polarization, the characteristic signature of this mechanism, have only recently been reported. Bower et al. (1999a,b) used the VLA at 4.8, 8.4, 22 and 43~GHz in its spectropolarimetric mode and the BIMA array at 86~GHz finding upper limits of $<$0.2\\% at the lower frequencies and $<$1\\% at 86~GHz. These small upper limits on linear polarization, given the sensitivity of the observations to large rotation measures due to Faraday rotation (RM $\\sim 10^{7}$ rad m$^{-2}$ at 8.4 GHz), are difficult to account for and this is discussed at length by Bower et al. (1999a). Apart from this there has been only a marginal detection at 800 \\micron~ of $4.9 \\pm 3.2\\%$ at 129 $\\pm 19^{\\circ}$ (Flett and Murray 1991) and a report of circular polarization at 4.8 GHz by Bower et al. (1999c). In mm to sub-mm polarimetric imaging studies of the central 15 pc of the Galaxy we have detected linear polarization in Sgr~A$^{\\star}$ at 750, 850, 1350 and 2000 \\micron~ which we present here. Discussion of the field distribution in the circum-nuclear disk (CND) and neighbouring molecular clouds as revealed by these observations will be presented separately. ", "conclusions": "We report the observation of millimetre and sub-millimetre polarization from Sgr~A$^{\\star}$, confirming the role of synchrotron radiation. The polarization is a property of the mm/sub-mm excess, demonstrating that the excess is real and not an artefact of variability and that it and the power-law spectrum arise in distinct structures. There is a large position angle shift between the mm and sub-mm observations. This can be explained by a transition between optically thin and self absorbed synchrotron radiation near 1~mm. Such a high self-absorption frequency implies a very compact source $\\sim$ 2R$_{\\rm S}$." }, "0003/astro-ph0003096_arXiv.txt": { "abstract": "{diffuse radiation -- intergalactic medium -- radiative transfer} \\ni In popular cosmological scenarios, some time beyond a redshift of 10, stars within protogalaxies created the first heavy elements; these systems, together perhaps with an early population of quasars, generated the ultraviolet radiation and mechanical energy that reheated and reionized the cosmos. The history of the Universe during and soon after these crucial formative stages is recorded in the all-pervading intergalactic medium (IGM), which contains most of the ordinary baryonic material left over from the big bang. Throughout the epoch of structure formation, the IGM becomes clumpy and acquires peculiar motions under the influence of gravity, and acts as a source for the gas that gets accreted, cools, and forms stars within galaxies, and as a sink for the metal enriched material, energy, and radiation which they eject. ", "introduction": "\\ni At epochs corresponding to $z\\sim 1000$ the intergalactic medium (IGM) is expected to recombine and remain neutral until sources of radiation and heat develop that are capable of reionizing it. The detection of transmitted flux shortward of the \\lya wavelength in the spectra of sources at $z\\sim 5$ implies that the hydrogen component of this IGM was ionized at even higher redshifts. There is some evidence that the double reionization of helium may have occurred later, but this is still controversial. It appears then that substantial sources of ultraviolet photons and mechanical energy were already present when the Universe was less than 7\\% of its current age, perhaps quasars and/or young star-forming galaxies: an episode of pre-galactic star formation may provide a possible explanation for the widespread existence of heavy elements (like carbon, oxygen, and silicon) in the IGM, while the integrated radiation emitted from quasars is likely responsible for the reionization of the intergalactic helium. Establishing the epoch of reionization and reheating is crucial for determining its impact on several key cosmological issues, from the role reionization plays in allowing protogalactic objects to cool and make stars, to determining the small-scale structure in the temperature fluctuations of the cosmic microwave background. Conversely, probing the reionization epoch may provide a means for constraining competing models for the formation of cosmic structures, and of detecting the onset of the first generation of stars, galaxies, and black holes in the Universe. ", "conclusions": "" }, "0003/astro-ph0003433_arXiv.txt": { "abstract": "The peculiar elliptical galaxy IC 1459 ($M_V = -21.19$, $D = 16.5 h^{-1}\\Mpc$) has a fast counterrotating stellar core, stellar shells and ripples, a blue nuclear point source and strong radio core emission. We present results of a detailed HST study of IC 1459, and in particular its central gas disk, aimed a constraining the central mass distribution. We obtained WFPC2 narrow-band imaging centered on the {\\HalphaNII} emission lines to determine the flux distribution of the gas emission at small radii, and we obtained FOS spectra at six aperture positions along the major axis to sample the gas kinematics. We construct dynamical models for the {\\HalphaNII} and {\\Hbeta} kinematics that include a supermassive black hole, and in which the stellar mass distribution is constrained by the observed surface brightness distribution and ground-based stellar kinematics. In one set of models we assume that the gas rotates on circular orbits in an infinitesimally thin disk. Such models adequately reproduce the observed gas fluxes and kinematics. The steepness of the observed rotation velocity gradient implies that a black hole must be present. There are some differences between the fluxes and kinematics for the various line species that we observe in the wavelength range 4569 {\\AA} to 6819 {\\AA}. Species with higher critical densities generally have a flux distribution that is more concentrated towards the nucleus, and have observed velocities that are higher. This can be attributed qualitatively to the presence of the black hole. There is some evidence that the gas in the central few arcsec has a certain amount of asymmetric drift, and we therefore construct alternative models in which the gas resides in collisionless cloudlets that move isotropically. All models are consistent with a black hole mass in the range $\\Mbh=1$---$4 \\times 10^8 \\Msun$, and models without a black hole are always ruled out at high confidence. The implied ratio of black holes mass to galaxy mass is in the range $0.4$--$1.5 \\times 10^{-3}$, which is not inconsistent with results obtained for other galaxies. These results for the peculiar galaxy IC 1459 and its black hole add an interesting data point for studies on the nature of galactic nuclei. ", "introduction": "\\label{s:intro} Supermassive central black holes (BH) have now been discovered in more than a dozen nearby galaxies (e.g., Kormendy \\& Richstone 1995; Ford \\etal 1998; Ho 1998; Richstone 1998, and van der Marel 1999a for recent reviews). BHs in quiescent galaxies were mainly found using stellar kinematics while the BHs in active galaxies were detected through the kinematics of central gas disks. Other techniques deployed are VLBI observations of water masers (e.g., Miyoshi \\etal 1995) and the measurement of stellar proper motions in our own Galaxy (Genzel \\etal 1997; Ghez \\etal 1998). The BH masses measured in active galaxies are all larger than a few times $10^8 \\Msun$, while the BH masses in quiescent galaxies are often smaller. The number of accurately measured BHs is expected to increase rapidly in the near future, especially through the use of STIS on board HST. This will establish the BH `demography' in nearby galaxies, yielding BH masses as function of host galaxy properties. In this respect two correlations in particular have been suggested in recent years. First, a correlation between BH mass and host galaxy (spheroid) optical luminosity (or mass) was noted (e.g., Kormendy \\& Richstone 1995; Magorrian \\etal 1998; van der Marel 1999b). However, this correlation shows considerable scatter (a factor $\\sim 10$ in BH mass at fixed luminosity). The scatter might be influenced by selection effects (e.g., it is difficult to detect a low mass BH in a luminous galaxy) and differences in the dynamical modeling. Second, a correlation between BH mass and either core or total radio power of the host galaxy was proposed (Franceschini, Vercellone, \\& Fabian 1998). However, the available sample is still small and incomplete. Establishing the BH masses for a large range of optical and radio luminosities is crucial to determine the nature of galactic nuclei. An accurate knowledge of BH demography will put constraints on the connection between BH and host galaxy formation and evolution and the frequency and duration of activity in galaxies harboring BHs. In this paper we measure the BH mass of IC 1459. IC 1459 is an E3 giant elliptical galaxy and member of a loose group of otherwise spiral galaxies. It is at a distance of $16.5 h^{-1}\\Mpc$ with $M_V=-21.19$ (Faber \\etal 1989). Williams \\& Schwarzschild (1979) noted twists in the outer optical stellar isophotes. Stellar spiral `arms' outside the luminous stellar part of the galaxy were detected in deep photographs (Malin 1985). Several stellar shells at tens of kpc from the center were discovered by Forbes \\& Reitzel (1995). A remarkable feature is the counter-rotating stellar core (Franx \\& Illingworth 1988) with a maximum rotation of $\\sim 170 \\kms$. IC 1459 also has an extended emission gas disk (diameter $\\sim 100''$) with spiral arms (Forbes \\etal 1990, Goudfrooij \\etal 1990) aligned with the galaxy major axis. The disk rotates in the same direction as the outer part of the galaxy (Franx \\& Illingworth 1988). The nuclear region of IC 1459 has line ratios typical of the LINER class (see e.g., Heckman 1980, Osterbrock 1989 for the definition of LINERS). A warped dust lane is also present. It is misaligned by $\\sim 15\\grad$ from the galaxy major axis and some dust patches are observed at a radius of $2''$ (Carollo \\etal 1997). IC 1459 has a blue nuclear optical source with $V=18.3$ (Carollo \\etal 1997; Forbes \\etal 1995) which is unresolved by HST. It also has a variable compact radio core (Slee \\etal 1994). There is no evidence for a radio-jet down to a scale of $1''$ (Sadler \\etal 1989). IC 1459 has a hard X-ray component, with properties typical of low-luminosity AGNs (Matsumoto \\etal 1997). Given the abovementioned properties, IC 1459 might best be described as a galaxy in between the classes of active and quiescent galaxies. This makes it an interesting object for extending our knowledge of BH demography, in particular since there are only few other galaxies similar to IC 1459 for which an accurate BH mass determination is available. We therefore made IC 1459, and in particular its central gas disk, the subject of a detailed study with the Hubble Space Telescope (HST). We observed the emission gas of IC 1459 with the Second Wide Field and Planetary Camera (WFPC2) through a narrow-band filter around {\\HalphaNII} and took spectra with the Faint Object Spectrograph (FOS) at six locations in the inner $1''$ of the disk. In Section~\\ref{s:wfpc2} we discuss the WFPC2 observations and data reduction. In Section~\\ref{s:spec} we describe the FOS observations and data reduction, and we present the inferred gas kinematics. To interpret the data we construct detailed dynamical models for the kinematics of the {\\Hbeta} and {\\HalphaNII} emission lines in Section~\\ref{s:modelH}, which imply the presence of a central BH with mass in the range $1$---$4 \\times 10^8 \\Msun$. In Section~\\ref{s:species} we discuss how the kinematics of other emission line species differ from those for {\\Hbeta} and {\\HalphaNII}, and what this tells us about the central structure of IC 1459. In Section~\\ref{s:starkin} we present dynamical models for ground-based stellar kinematical data of IC 1459, for comparison to the results inferred from the HST data. We summarize and discuss our findings in Section~\\ref{s:discon}. We adopt $H_0 = 80 \\kms \\Mpc^{-1}$ throughout this paper. This does not directly influence the data-model comparison for any of our models, but does set the length, mass and luminosity scales of the models in physical units. Specifically, distances, lengths and masses scale as $H_0^{-1}$, while mass-to-light ratios scale as $H_0$. ", "conclusions": "\\label{s:discon} We have presented the results from a detailed HST study of the central structure of IC 1459. The kinematics of the gas disk in IC 1459 was probed with FOS observations through six apertures along the major axis. In our modeling of the observed kinematics we took into account the stellar mass density in the central region by fitting WFPC2 broad-band imaging, and we determined the flux distribution of the emission-gas from WFPC2 narrow-band imaging. From the models we have determined that IC1459 harbors a black hole with a mass in the range $1 \\times 10^8 \\Msun$ -- $4 \\times 10^{8} \\Msun$, with the exact value depending somewhat on whether we model the gas as rotating on circular orbits, or as an ensemble of collisionless cloudlets. While the dynamical models that we have constructed provide good fits to the observations, the true structure of IC 1459 could of course be more complex than our models. Below we discuss several aspects of this. Ground-based observations (Goudfrooij \\etal 1994; Forbes \\& Reitzel 1995) indicate an ellipticity $\\epsilon = 0.5$ for the gas disk at radii larger than a few arcseconds, implying an inclination angle $i=60\\grad$. By contrast, from our HST emission-line image we found a monotonic increase in ellipticity from $\\epsilon = 0.17$ to $\\epsilon = 0.37$ between $r=0.25''$ to $1.0''$. In our modeling we have assumed that these rounder inner isophotes are due to a thickening of the gas disk caused by asymmetric drift, and we estimated the effect of this on the inferred $\\Mbh$ (see Sections~\\ref{ss:CTIOcomp} and \\ref{ss:asymdrift}). However, an alternative interpretation would be to assume that the disk is warped. The presence of a dust lane slightly misaligned with the gas disk, a counter-rotating stellar core, and stellar shells and ripples in the outer galaxy make it plausible that the central gas and dust were accreted from outside, displaying warps as it settles down. In this interpretation we infer an increase in inclination angle from $32.9\\grad$ to $47.9\\grad$ between $r=0.25''$ to $1.0''$. The BH mass in the models of Section~\\ref{ss:dyn_mod} scales as $\\Mbh \\propto \\sin^{2}i$ due to the projection of the rotational velocities. Hence the inferred differences in inclination angle would amount to an increase in $\\Mbh$ by only a factor $\\sim 1.5$, which would not significantly change our results. IC 1459 hosts an unresolved blue nuclear point source. Carollo \\etal (1997) estimated a luminosity of $L_V \\approx 1.5 \\times 10^7 \\Lsun$. So far, we have assumed that this light is non-stellar radiation from the active nucleus. However, one could assume alternatively that the blue light is emitted by a cluster of young stars. If the cluster were to have a mass equal to the mass $\\Mbh$ that we have inferred from our models, this would require $\\Upsilon_V \\gta 10$. Depending on age and metallicity, stellar evolution models typically predict $\\Upsilon_V \\lta 2$ for a young cluster (e.g., Worthey 1994). Thus the assumption that the point source is stellar in origin cannot lift the need for a central concentration of non-luminous matter, most likely a BH. The data show differences between the fluxes and kinematics for the various line species. The main characteristics of the observed kinematics are similar for all species, but we see that for the species with higher critical densities the flux distribution is generally more concentrated towards the nucleus and the observed velocities are higher. This can be well understood qualitatively in the framework of a single $\\Mbh$. However, to actually verify quantitatively whether each species implies the same value for $\\Mbh$ would require information on the flux distributions for each species at high spatial resolution as well as detailed knowledge on the ionization structure of the gas, neither of which is available. An extremely conservative upper limit on the differences in the $\\Mbh$ implied by the different line species is obtained by assuming that all species have the same flux distribution (not actually correct), in which case $\\Mbh$ values are obtained that are up to $\\sim 4$ times larger than inferred from {\\Hbeta} and {\\HalphaNII}. Irregularities in the velocity dispersion profiles of {\\OIII}, {\\OI} and {\\SII} suggest localized turbulent motions. We incorporated turbulent motion in our models in a very simple manner (cf.~equation~\\ref{eq:turbdef}), using a parametrization that fits the main trend of an increase of the velocity dispersion toward the nucleus. Nevertheless, going to the extreme, one could assume that all observed motions have a non-gravitational origin. The overall kinematics would then be due to in- or outflows. There are several objections to this interpretation. A spherical in- or outflow could not produce any net mean velocity. A bi-directional flow is unlikely since no hint of this is seen in the HST {\\HalphaNII} emission image. Next we consider the location of IC 1459 with respect to the correlations between $\\Mbh$ and host total optical and radio luminosity, mentioned in the Introduction. The ratio of $\\Mbh$ and galaxy mass is in the range $0.4$--$1.5 \\times 10^{-3}$. This is somewhat lower than the average value of $\\sim 2 \\times 10^{-3}$ seen for other galaxies (Kormendy \\& Richstone 1995), but still comfortably within the observed scatter. As discussed in the Introduction, IC 1459 is probably the end-product of a merger between two galaxies. Apparently, this merger history has not moved IC 1459 to an atypical spot in the $\\Mbh$ vs.~galaxy mass scatter diagram. The fact that IC 1459 has a LINER type spectrum and core radio emission, but no jets, makes it interesting to see where it is located on a $\\Mbh-L_{\\rm radio}$ plot. IC 1459 has a radio luminosity of $5.5 \\times 10^{22} {\\rm WHz^{-1}}$ at 5 GHz (Wright \\etal 1996). Interestingly, this puts IC 1459 within the scatter observed around the correlation between $\\Mbh$ and total radio luminosity inferred for a small sample of galaxies with available BH mass determinations, but quite off the correlation with core radio luminosity (see Figs.~3 and 4 of Franceschini, Vercellone, \\& Fabian 1998). However, one should be weary of beaming and variability of core radio sources, and possible resolution differences among the observations. Our dynamical modeling has used the gas disk in IC 1459 as a diagnostic tool to constrain $\\Mbh$. A logical next step to improve on our work would be to obtain better two-dimensional coverage of the gaseous and stellar kinematics. A definite advance in our understanding of nuclear gas disks would also be obtained if we had a better knowledge of properties of the gas disk such as the electron density, metallicity, and the ionization structure and ionization mechanism. One could then try to simultaneously understand these properties and the gas dynamics. We now know that quite likely most, or even all, bright ellipticals host a massive central BH. More detailed knowledge of the chemical properties and kinematics of the dust and gas surrounding the BH could tell us about the origin of this material: accretion of small satellites, stripping of companions, or internal mass loss from stars. This would immediately constrain the frequency and probability with which any bright elliptical hosts this material. A better understanding of the kinematics, such as the importance of dissipative shocks generated by turbulence, could then help to determine the accretion rate of the black hole. These pieces of information together could tell us what fraction of the observed $\\Mbh$ could have come from this process over the life time of a galaxy. Knowledge on the ratio of black hole and stellar mass as function of time would be valuable for understanding the formation and evolution of early-type galaxies in general." }, "0003/astro-ph0003119_arXiv.txt": { "abstract": "\\noindent The physical conditions in the inner parsec of accretion disks believed to orbit the central black holes in active galactic nuclei can be probed by imaging the absorption of background radio emission by ionized gas in the disk. We report high angular resolution observations of the nearby galaxy NGC 4261 which show evidence for free-free absorption by a thin, nearly edge-on disk at several frequencies. The angular width, and probably the depth, of the absorption appears to increase with decreasing frequency, as expected. Because free-free absorption is much larger at lower frequencies, the longest possible baselines are needed to provide adequate angular resolution; observing at higher frequencies to improve resolution will not help. ", "introduction": "The nearby FR-I radio galaxy NGC 4261 (3C270) is a good candidate for the detection of free-free absorption by ionized gas in an inner accretion disk. The galaxy is known to contain a central black hole with a mass of $5 \\times 10^{8} \\ {\\rm M}_{\\odot}$, a nearly edge-on nuclear disk of gas and dust with a diameter of $\\approx 100$ pc, and a large-scale symmetric radio structure which implies that the radio axis is close to the plane of the sky. At an assumed distance of 40 Mpc, 1 milliarcsecond (mas) corresponds to 0.2 pc. Previous VLBA observations of this galaxy revealed a parsec-scale radio jet and counterjet aligned with the kpc-scale jet (see Figure~\\ref{Fig1}). The opening angle of the jets is less than $20^{\\circ}$ during the first 0.2 pc and $< 5^{\\circ}$ during the first 0.8 pc. At 8.4 GHz we found evidence for a narrow gap in radio brightness at the base of the parsec-scale counterjet, just east (left) of the brightest peak which we identified as the core based on its inverted spectrum between 1.6 and 8.4 GHz (see the left part of Figure~\\ref{Fig2}, from Jones and Wehrle 1997). We tentatively identified this gap as the signature of free-free absorption by a nearly-edge on inner disk with a width $<< 0.1$ pc and an average electron density of $10^{3}-10^{8}\\ {\\rm cm}^{-3}$ over the inner 0.1 pc. \\begin{center} \\begin{figure} \\vspace{46mm} \\special{psfile=VSOP-symp-fig1.ps hoffset=43 voffset=152 hscale=32 vscale=32 angle=270} \\caption{VLBA image of NGC 4261 at 8.4 GHz. The contours increase in steps of $\\sqrt{2}$ starting at $\\pm 0.75$\\% of the peak, which is 99 mJy/beam. The restoring beam is $1.86 \\times 0.79$ mas with major axis PA = $-1.3^{\\circ}$.} \\label{Fig1} \\end{figure} \\end{center} ", "conclusions": "Our observations at 1.6 and 4.9 GHz appear to confirm the free-free absorption explanation for the sub-parsec radio morphology in NGC 4261. Measurements of the optical depth in the absorbed region and the distance between the absorption and the core as a function of frequency will allow the radial distribution of electron density in the inner parsec of the disk to be determined." }, "0003/astro-ph0003449_arXiv.txt": { "abstract": "The detection of unpulsed TeV emission from PSR~B1706-44 raised the question if this VHE emission results from inverse Compton scattering of multi-TeV electrons on the CMBR and other soft photon fields. This hints at the existence of an unseen synchrotron nebula, which is larger than the observed $1'$ compact X-ray nebula. In a search for the synchrotron counterpart of the TeV nebula, we have taken V-band CCD images of the region around the PSR~B1706-44 pulsar position. By rejecting all point source contributions down to the diffuse sky limit, we were able to search for such extended plerionic synchrotron emission on scale sizes limited by the angular resolution of the CANGAROO imaging TeV observations. By combining the V-band flux limits with the observed TeV flux, we were able to obtain upper limits for the magnetic field strength as a function of the radius of the assumed TeV nebula. Assuming that the $2'$ radio plerion also defines the size of the TeV plerion, we constrain the steady state particle spectral index to values between 4 and 5.5, with a field strength limit of about 3~$\\mu$G. This compares with an index of $3.6\\pm 0.6$ inside the 1' compact X-ray nebula. ", "introduction": "PSR B1706-44 is a young Vela-like pulsar (spin-down age 17.5~kyr), with a period of 102 ms and a large spin-down power (\\.{E}~=~3.4~x~10$^{36}$~ergs/s). It was discovered during a 20~cm radio pulsar survey of the southern Galactic plane (Johnston et al. 1992), and later detected in soft X-rays during the ROSAT mission (Becker et al. 1992, 1995) and identified as a pulsed GeV source by EGRET (Thompson et al. 1992). Very high energy (VHE) $\\gamma$-ray observations above 1~TeV from CANGAROO (Kifune et al. 1995; Kifune 1997) and above 0.3~TeV from Durham (Chadwick et al. 1997) detected and confirmed the existence of unpulsed radiation of statistical significance from this source. The CANGAROO detection is consistent with a point source, with angular extend not exceeding $\\sim 0.^{\\circ}12$ -- the pixel size of the imaging camera. A dispersion based distance measure of Taylor \\& Cordes (1993) places PSR B1706-44 at $\\sim$1.8~kpc. A possible SNR G~343.1-2.3 association with the pulsar was proposed by McAdam et al. (1993), but this was later found to be unlikely by Frail et al. (1994) and Nicastro et al. (1996). Chakrabarty \\& Kaspi (1998) gave a red-band upper limit of $R\\gaeq18$ to the pulsar. A 3$\\sigma$ upper limit to the pulsar magnitude of $V=24.5$ was given by Lundqvist et al. (1999), and this is consistent with the theoretical prediction of $V=24.12$ (Urama \\& Okeke 1998, and references therein). Mignani et al. (1999), using the same data as Lundqvist et al., obtained an upper limit of $V\\gaeq 27.5$. The photon flux above 1~TeV from this unpulsed source is only two times smaller compared to the flux of the Crab Nebula at the same energy. This is remarkable in the context of a synchrotron-inverse Compton (on external photon fields) interpretation for the TeV $\\gamma$-rays from PSR B1706-44, since the synchrotron nebula of this source is very weak compared to the Crab Nebula's synchrotron intensity. Whereas the Crab Nebula's intense synchrotron emission is the result of a large $B$ ($\\sim 10^{-4}$~G) field, we must have a much weaker field for this plerion to avoid a bright synchrotron nebula. De Jager (1995) speculated that a Vela-like compact synchrotron nebula ($<1'$, Harnden et al. 1985) may be present, which would account for most of the unpulsed X-ray emission from this source, since the pulsar wind magnetic field (which scales as $B\\propto 1/R$) would predict a compact synchrotron nebula, if the conditions are similar to Vela. The corresponding particle density required for the synchrotron nebula would however be too low to produce a detectable inverse Compton compact nebula, since the target photon density from the CMBR and Galactic disk would be too small. A compact nebula was indeed discovered (Finley et al. 1998), which confirms the abovementioned interpretation. The scenario of Harding \\& de Jager (1997), with a detailed analysis by Aharonian et al. (1997) may apply in this case: electrons are streaming away from the compact X-ray nebula into a low-B extended plerion. If the diffusion coefficient for electrons in the extended plerion is small enough, it may be possible to trap enough electrons to account for the TeV emission by IC scattering. Frail et al. (1994) detected a radio synchrotron nebula associated with the pulsar, which has a radius of $\\sim 2'$ at 20~cm wavelength. This may represent the low energy counterpart of the TeV nebula, since the associated $B$ in the radio nebula is smaller than the pulsar driven field in the compact X-ray nebula, and the size is just smaller than the angular resolution of the CANGAROO telescope. If the TeV $\\gamma$-rays originate from the extended (radio?) nebula, we will expect a bright extended nebula in optical and/or X-rays, if the field strength is large enough. No extended X-ray nebula was however seen, resulting in an upper limit of four times the flux of the point source/compact X-ray nebula at 1~keV, for $R=2'$ (K. Brazier 1998, personal communication). This is consistent with the expectation that the pulsar field strength (which drops as $1/R$ outside the light cylinder) should start to drop below the ambient field strength of a few $\\mu$G at a distance of a few arcminutes (de Jager \\& Harding 1998). The consequence of this is that both the synchrotron brightness and synchrotron characteristic frequencies should drop with increasing $R$. Scaling from the Durham detection above 0.3~TeV, and the expected $B\\sim3\\;\\mu$G in the extended nebula (de Jager \\& Harding 1998), we may use equation (5) of Aharonian et al. (1997) to calculate the synchrotron frequency which corresponds to the Durham detected $\\gamma$-rays near 0.3~TeV, assuming that the CMBR is mostly responsible for the inverse Compton scattering. This gives a frequency of $$\\nu\\sim 1.5\\times 10^{15} \\left(\\frac{E_{\\gamma}}{0.3\\; TeV}\\right) \\left(\\frac{B}{3\\;\\mu G}\\right)\\; Hz,$$ which is within a factor of 3 from the V-band. The non-detection of the unpulsed component at 20 GeV (Thompson et al. 1996) hints at a marginal turnover in the $\\gamma$-ray spectrum (de Jager \\& Harding 1998), which corresponds to a turnover frequency as low as $10^{14}$ Hz, which is well below the V-band. At a distance of 1.8 kpc to the source, we expect significant interstellar absorption in the blue and UV bands. We therefore take wide field V-band images as a compromise between interstellar absorption and avoiding a spectral turnover at too low frequencies. By taking overlapping images, we cover a total field-of-view of about $10'$, which should include the expected TeV source. Even if the full extend of the TeV source is $10'$, we still expect a typically centrally brightened image, resulting in a radial gradient from the source. Failure to detect the optical counterpart will allow us to set upper limits on the magnetic field strength in the extended nebula, since any stronger field should have resulted in a brighter optical nebula. ", "conclusions": "Our search for an extended nebula synchrotron counterpart to the TeV nebula did not reveal any nebular structure around PSR~B1706-44. The TeV observations do hint at the existence of such a plerionic structure, but our results constrain the nebular magnetic field strength if we assume that the TeV $\\gamma$-rays originate from the IC scattering of energetic pulsar wind electrons on the CMBR. At $R=2'$, we find that the optical and X-ray limits are complementary: the spectral index is constrained to $\\alpha = 1.5$ to 2.3, whereas $B\\leq 3$~$\\mu$G. This limit is barely above the pulsar wind solution, but below the average Galactic field values shown in Figure 4. Furthermore, our constraints on $\\alpha$ are marginally consistent with $\\alpha = 1.3\\pm 0.3$ measured for the compact X-ray nebula (Finley et al. 1998). Our optical observations and the information gathered from other wave bands (mainly from X-ray and $\\gamma$-ray energies) show that better quality observations (dark moon) should improve these limits, or, result in a detection. Polarimetry of the ``extended nebula\" of PSR~B1706-44 may also help to clarify the status of synchrotron emission." }, "0003/astro-ph0003480_arXiv.txt": { "abstract": "Recent observations of \\rcwx\\, the neutron-star candidate at the center of the supernova remnant RCW 103, show that a component of its emission varies sinusoidally with a period of approximately six hours. We argue that this period is what one would expect for a freely precessing neutron star with a spin period of about one second. We produce light curves for a freely precessing neutron star with a hotspot. By a suitable choice of parameters, we obtain light curves which are constant with rotational phase when the flux from the star reaches a maximum. At other phases of the precession, the flux varies as the star rotates but the total flux decreases by a factor of several. These models can explain the behavior observed from \\rcwx\\ and predict that the spin period should be detectable at minimum flux from sufficiently sensitive measurements. ", "introduction": "\\jcite{Garm00} have noted that the x-ray source, \\rcwx , located near the center of the supernova remnant RCW 103, has a sinusoidal light curve with a period of approximately six hours. The authors suggest that this period implies that the x-ray source has a low-mass companion with a six hour orbital period. In this paper, we examine an alternative possibility, that \\rcwx\\ is a freely precessing neutron star with a precession period of about six hours. The varying spectrum from \\rcwx\\ is similar to a blackbody yielding an effective area of a small fraction of a square kilometer (Garmire, private communication), much less than the surface area of a neutron star; therefore, the situation could be well approximated by a point source (a hotspot) on the surface of a freely-precessing neutron star. Free precession has often been invoked to explain long-period variability in neutron stars and neutron-star systems. \\jcite{1975Natur.257..203B} attributed the 35-day cycle of Her X-1 (\\cite{1972ApJ...174L.143T}) to the free precession of the neutron star secondary. \\jcite{1988MNRAS.235..545J} explained timing residuals in the Crab pulsar as arising from small amplitude free precession of the neutron star. \\jcite{1996A&A...306..443C} found a small amplitude modulation in the optical flux from the Crab pulsar which \\jcite{1997A&A...324.1005C} cite as evidence of free precession. In \\S \\ref{sec:freep} we outline the kinematics of free precession, and in \\S \\ref{sec:lensing}, we review the equations that determine the trajectory of light leaving the surface of the neutron star. \\S \\ref{sec:lightc} presents the light curves as a function of precessional and rotational phase for Newtonian, relativistic and ultracompact neutron stars with a emission from a hotspot. Finally, \\S \\ref{sec:disc} places the results in a greater context. ", "conclusions": "\\label{sec:disc} We have explored several possible light curves of a freely precessing neutron star with a hotspot, and focussed on those whose flux is constant with rotational phase when the flux reaches a maximum value. These possibilities indicate that \\rcwx\\ may be a freely precessing neutron star. However, as the mean flux decreases from its maximum it also begins to vary with the rotation of the star; therefore, in the context of this model, we would expect that subsequent observations of \\rcwx\\ may uncover its rotational period which we would expect to be on the order of several seconds. The variation of the flux with the rotational phase of the star during some portion of the precession appears generic to freely precessing stars with a hotspot. The Earth undergoes a free precession with a period of about 433~days, known as the \\jcite{1891AJ.....11...59C} wobble. This is significantly longer than one would expect from the asymmetry of the Earth's figure due to dissipation inside the Earth (\\cite{Burs93}). Without excitation the wobble would disappear within about a century, and the avenue for its excitation is still unclear. In neutron stars, precession has been proposed to explain long-term variations in their spin and pulse profiles (\\eg \\cite{Davi70,Gold70,Rude70,Brec72,Pine72,Pine73,Pine74}). If neutron stars rotate as rigid bodies the precessional period would be $P/\\epsilon$ (\\cite{Pine72,Pine74}) where $P$ is the rotational period. \\jcite{1974ApJ...190..137R} proposed that neutron stars contain a superfluid component in their cores. \\jcite{1977ApJ...214..251S} explored how the pinning of the superfluid vortices affects the free precession of a neutron star. He argued that the dissipation timescale for the precessional mode ($\\tau_w$) is of the order of the postglitch relaxation time ($\\tau$) times the ratio of the rotational to the precessional frequency. $\\tau$ ranges from a week for the Crab to nearly a century for 1641-45 (\\cite{Shap83}); therefore, depending on the nature of the superfluid coupling the precessional mode may last for millennia. \\jcite{1977ApJ...214..251S} also found that if the star is triaxial the geometry of the precession is more complicated that for the purely free precession considered here. Additionally, the precessional frequency in this case is given by angular velocity of the superfluid component of the star times the fractional contribution of the superfluid to the total moment of inertia of the star (about one percent). \\jcite{1999ApJ...524..341S} have recently reexamined the precession of multicomponent neutron stars with imperfect vortex pinning and found several possibly long-lasting precessional modes with long periods like the precession described here. The excitation and decay of precessional motions in neutron stars are still uncertain. Evidence has been found for free precession in some radio pulsars (\\eg\\ \\cite{1997A&A...324.1005C,1988MNRAS.235..545J}), but it is not generic (\\eg\\ \\cite{1995AAS...187.1603M}); therefore, the question arises as to which properties of a neutron star would allow or prevent it from precessing and would they correlate with its radio emission. \\jcite{1999ApJ...519L..77M} and \\jcite{2000MNRAS.313..217M} argue that precession is characteristic of strongly magnetized neutron stars (\\cf\\ \\eqref{bdistort}). \\jcite{Usov96} and \\jcite{Aron98} have proposed that strongly magnetized neutron stars are unlikely to produce radio emission collectively due to the formation of bound electron-positron pairs. Alternatively, \\jcite{Bari97b} suggest that in sufficiently strong fields ($B \\gtrsim B_c$) the QED process of photon splitting (\\cite{Adle71}; \\cite{Heyl97hesplit}) can dominate one-photon pair production. This will effectively quench the pair cascade, making coherent pulsed radio emission impossible. Since the timescales for both the excitation and decay of precessional motion in neutron stars are unknown, one can appeal to the relative youth of \\rcwx\\ and the other members of the AXP class. They are all several thousand years old, much younger than vast majority of radio pulsars (\\cite{Tayl93}). The appropriate timescales for precession may simply be shorter than the ages of most radio pulsars while longer than those of AXPs. Furthermore, the hints of precession seen in the Crab pulsars (\\cite{1997A&A...324.1005C}) point toward this explanation. We have examined the light curves of free precessing neutron stars with a hotspot and focussed on those geometries which exhibit an epoch during each precessional period where the flux does not vary as the star rotates. These geometries account for about three percent of a random sample and may provide an explanation for the emission from \\rcwx\\ . If this is the case, further observations of the light curve from \\rcwx\\ should reveal a pulse period of the order of $10^{-4}$ times the precessional period of six hours. Free precession may be a hallmark of young or highly magnetized neutron stars, and it is a direct probe of the structure of the crust and interior of the neutron star and the coupling between them." }, "0003/astro-ph0003163_arXiv.txt": { "abstract": "ZZ Ceti stars exhibit small amplitude photometric pulsations in multiple gravity modes. As the stars cool their dominant modes shift to longer periods. We demonstrate that parametric instability limits overstable modes to amplitudes similar to those observed. In particular, it reproduces the trend that longer period modes have larger amplitudes. Parametric instability is a form of resonant 3-mode coupling. It involves the destabilization of a pair of stable daughter modes by an overstable parent mode. The 3-modes must satisfy exact angular selection rules and approximate frequency resonance. The lowest instability threshold for each parent mode is provided by the daughter pair that minimizes $(\\delta\\omega^2+\\gamma_d^2)/\\kappa^2$, where $\\kappa$ is the nonlinear coupling constant, $\\delta\\omega$ is the frequency mismatch, and $\\gamma_d$ is the energy damping rate of the daughter modes. Parametric instability leads to a steady state if $|\\delta\\omega|>\\gamma_d$, and to limit cycles if $|\\delta\\omega|<\\gamma_d$. The former behavior characterizes low radial order ($n\\leq 3$) parent modes, and the latter those of higher $n$. In either case, the overstable mode's amplitude is maintained at close to the instability threshold value. Although parametric instability defines an upper envelope for the amplitudes of overstable modes in ZZ Ceti stars, other nonlinear mechanisms are required to account for the irregular distribution of amplitudes of similar modes and the non-detection of modes with periods longer than $1,200\\s$. Resonant 3-mode interactions involving more than one excited mode may account for the former. Our leading candidate for the latter is Kelvin-Helmholtz instability of the mode-driven shear layer below the convection zone. ", "introduction": "\\label{sec:nl-intro} Within an instability strip of width $\\Delta T_{\\rm eff}\\approx 10^3\\K$ centered at $T_{\\rm eff}\\approx 1.2\\times 10^3\\K$, hydrogen white dwarfs exhibit multiple excited gravity modes with $10^2\\lesssim P\\lesssim 10^3\\s$. Convective driving, originally proposed by Brickhill (\\cite{nonad-brick90}, \\cite{nonad-brick91}), is the overstability mechanism (Goldreich \\& Wu \\cite{nl-paperI}, hereafter Paper I). Individual modes maintain small amplitudes; typical fractional flux variations range from a few mma to a few tens of mma.\\footnote{$1$ mma of light variation is approximately $0.1\\%$ fractional change in flux.} The nonlinear mechanism responsible for saturating mode amplitudes has not previously been identified. We demonstrate that parametric resonance between an overstable parent g-mode and a pair of lower frequency damped daughter g-modes sets an upper envelope to the parent modes' amplitudes.\\footnote{The g-mode dispersion relation allows plenty of good resonances.} Moreover, the envelope we calculate reproduces the broad trends found from observational determinations of mode amplitudes in ZZ Ceti stars. Our investigation follows pioneering work by Dziembowski \\& Krolikowska (\\cite{nl-dziem85}) on overstable acoustic modes in $\\delta$-Scuti stars. They showed that parametric resonance with damped daughter g-modes saturates the growth of the overstable p-modes at approximately their observed amplitudes. This paper is comprised of the following parts. In \\S \\ref{sec:para-para} we introduce parametric instability for a pair of damped daughter modes resonantly coupled to an overstable parent mode. We evaluate the parent mode's threshold amplitude and describe the evolution of the instability to finite amplitude. \\S 3 is devoted to the choice of optimal daughter pairs. We discuss relevant properties of 3-mode coupling coefficients, and the constraints imposed by frequency resonance relations and angular selection rules. Evaluation of the upper envelope for parent mode amplitudes set by parametric resonance is the subject of \\S 4. Numerical results are interpreted in terms of analytic scaling relations and compared to observations. \\S 5 contains a discussion of a variety of issues leftover from this investigation. Detailed derivations are relegated to a series of Appendices. The stellar models used in this investigation were provided by Bradley (\\cite{nl-bradley96}). Their essential characteristics are $M_{*} = 0.6 M_{\\odot}$, $\\log(g/\\cm\\s^{-2}) = 8.0$, hydrogen layer mass $1.5\\times 10^{-4} M_{*}$, and helium layer mass $1.5\\times 10^{-2} M_{*}$. ", "conclusions": "\\label{sec:nl-discussion} \\subsection{Turbulent Saturation} \\label{subsec:nl-turbulent} Turbulent convection severely reduces the vertical gradient of the horizontal velocity of g-modes in the convection zone. As a result, a shear layer forms at the boundary between the bottom of the convection zone and the top of the radiative interior (Goldreich \\& Wu \\cite{nl-paperIII}, hereafter Paper III). Kelvin-Helmholtz instability of this layer provides a nonlinear dissipation mechanism for overstable modes. An overstable mode's amplitude cannot grow beyond the value at which nonlinear damping due to the Kelvin-Helmholtz instability balances its linear convective driving. Equation (44) of Paper III provides an estimate for the value at which this mechanism saturates the surface amplitude of $(\\delta p/p)$; \\be \\left({\\delta p\\over p}\\right)\\sim {0.1\\over C_D}{[(\\omega\\tau_c)^2+1]^{1/2}[(\\omega\\tau_c)^2-1]\\over \\omega\\tau_c}{L z_\\omega^2\\over R z_b}. \\label{eq:na-vis-dpnl} \\ee Our ignorance of the complicated physics involved in a nonlinear shear layer is covered by the range of possible values of the dimensionless drag coefficient, $C_D$. Terrestrial experiments indicate that $C_D$ falls between $10^{-3}$ and $10^{-1}$. The dashed lines in Figure \\ref{fig:ampli-clemens} show the effect of including nonlinear turbulent damping in addition to parametric instability on limiting mode amplitudes. Amplitudes of overstable modes saturated by the Kelvin-Helmholtz instability are uncertain because $C_D$ is poorly constrained. \\subsection{Granddaughter Modes} \\label{subsec:granddaughters} Here we answer the following questions. Under what conditions do daughter modes excite granddaughter modes by parametric instability?\\footnote{In this subsection subscripts $p$, $d$, and $g$ refer to parent, daughter, and granddaughter modes.} What are the consequences if they do? In \\S \\ref{subsec:para-time} we show how parametric instability of linearly damped daughter modes maintains the amplitude of a parent mode close to its equilibrium value. In order to dispose of the energy they receive from the parent mode, the time averaged energies of the daughter modes must be close to their equilibrium values. This raises a worry. Suppose the daughters are prevented from reaching their equilibrium amplitudes by parametric instability of granddaughter modes Then they would not be able to halt the amplitude growth of the parent mode. To answer the first of these questions, we calculate the ratio, denoted by the symbol ${\\cal S}$, between the threshold amplitude for a daughter mode to excite granddaughter modes and its equilibrium amplitude under parametric excitation by the parent mode. The former is obtained from equation \\refnew{eq:thresh}, and the latter from equations \\refnew{eq:nl-amplisteady2}-\\refnew{eq:nl-amplisteady3}. We make a few simplifying assumptions to streamline the discussion. Resonances between daughters and granddaughters are taken as exact; individual members of daughter and granddaughter pairs are treated as equivalent. Equations \\refnew{eq:nl-kappamax} and \\refnew{eq:fastgamma} are combined to yield \\begin{equation} \\kappa^2\\sim \\frac{\\gamma_p}{n_p^2L}. \\label{eq:nl-kappagamma} \\end{equation} It is then straightforward to show that \\be {\\cal S}\\approx \\frac{\\omega_p\\omega_d}{\\omega_g^2} \\left(\\frac{n_d}{n_p}\\right)^2 \\left(\\frac{\\gamma_g^2}{\\gamma_d^2+\\delta \\omega^2}\\right)\\approx 32\\left(\\frac{\\gamma_g^2}{\\gamma_d^2+\\delta\\omega^2}\\right). \\label{eq:nl-calSp} \\ee The factor $32$ is an approximation based on taking $\\omega_g/\\omega_d = \\omega_d/\\omega_p=1/2$ and $n_p/n_d=1/2$. In general ${\\cal S}\\gg 1$, so the excitation of granddaughter modes requires the daughter modes to have energies in excess of their equilibrium values. However, the equilibrium solution is unstable if the best daughter pair corresponds to $\\gamma_d>|\\delta\\omega|$, and then the daughter mode energies episodically rise far above their equilibrium values. At such times, granddaughter modes may be excited by parametric instability and consequentially limit the amplitude growth of the daughter modes. This slows the transfer of energy from parent to daughter modes, but it does not prevent the daughter modes from saturating the growth of the parent mode's amplitude at the level described by equation \\refnew{eq:thresh}. For the few lowest order parent modes, we typically find $|\\delta \\omega| \\geq \\gamma_d$. This may reduce ${\\cal S}$ to below unity with the consequence that the parent mode amplitude may rise above that given by equation \\refnew{eq:thresh}. \\subsection{Additional 3-Mode Interactions} \\label{subsec:threemode} Parametric instability sets reasonable upper bounds on the photospheric amplitudes of overstable modes. In a given star this upper bound rises with increasing mode period except possibly for the lowest few modes. However, the observed amplitude distributions are highly irregular. This mode selectivity may arise from 3-mode interactions which involve more than one overstable mode. We investigate a particular example of this type. It is closely related to parametric instability, the only difference being that the daughter modes of the overstable parent mode are themselves overstable. Acting in isolation, resonant mode couplings tend to drive mode energies toward equipartition. They conserve the total energy, ${\\dot E}_p+{\\dot E}_{d_1}+{\\dot E}_{d_2}=0$, and transfer action according to ${\\dot E}_{d_1}/\\omega_{d_1} = {\\dot E}_{d_2}/\\omega_{d_2}=-{\\dot E}_p/\\omega_p$. In this context it is important to note that the energies of overstable modes limited by parametric instability decline with increasing mode period (Fig. \\ref{fig:mode-energy}). Therefore nonlinear interactions transfer energy from the parent mode to its independently excited daughters. As shown below, this transfer may severely suppress the parent mode's amplitude. We start from equations \\refnew{eq:nl-amplieqnfulla}-\\refnew{eq:nl-amplieqnfullc}. These may be manipulated to yield \\begin{equation} {d E_p\\over dt} =\\gamma_p E_p+3\\sqrt{2}\\omega_p\\kappa(E_p E_{d_1} E_{d_2})^{1/2}\\sin\\Phi, \\label{eq:nl-Eone} \\end{equation} where $\\Phi = \\theta_{d_1} +\\theta_{d_2} - \\theta_p$. For $E_p\\gg E_{d_1}$ and $E_{d_2}$, nonlinear interactions transfer energy from the parent mode to its daughter modes. In particular, if we ignore phase changes in the overstable daughter modes due to their interactions with granddaughter modes, we find that $\\Phi$ satisfies \\begin{equation} \\frac{d\\Phi}{dt}=\\delta\\omega-\\frac{3}{\\sqrt{2}}\\kappa(E_p E_{d_1} E_{d_2})^{1/2} \\left[\\left(\\frac{\\omega_{d_1}}{E_{d_1}}+\\frac{\\omega_{d_2}}{E_{d_2}}\\right)- \\frac{\\omega_p}{E_p}\\right]\\cos\\Phi, \\label{eq:nl-Phi} \\end{equation} with a stable solution at $\\Phi=-\\pi/2$ when $\\delta \\omega = 0$. We denote the ratio of the nonlinear term to the linear term in equation \\refnew{eq:nl-Eone} by the symbol ${\\cal T}$; \\begin{equation} {\\cal T}\\approx \\frac{3\\sqrt{2}\\omega_p\\kappa} {\\gamma_p} \\left(\\frac{E_{d_1}E_{d_2}}{E_p}\\right)^{1/2}. \\label{eq:nl-calT} \\end{equation} Using the magnitudes of $E_i$ set by parametric instability of their respective daughters, and adopting the same approximations made in \\S \\ref{subsec:granddaughters}, we arrive at \\begin{equation} {\\cal T}\\approx \\frac{\\omega_p\\omega_d}{\\omega_g^2} \\left(\\frac{n_d}{n_p}\\right)^2 \\left(\\frac{\\gamma_g^2}{\\gamma_d^2 + \\delta \\omega^2}\\right)\\approx 32\\left(\\frac{\\gamma_g^2}{\\gamma_d^2+\\delta\\omega^2}\\right). \\label{eq:nl-calTp} \\end{equation} Comparing equations \\refnew{eq:nl-calSp} and \\refnew{eq:nl-calTp} we see that ${\\cal S=T}$. A little thought reveals that this is not a coincidence. For ${\\cal T} \\gg 1$, overstable daughter modes can suppress a parent mode's energy below the value set by parametric instability. We expect this suppression to be important in cool ZZ Ceti stars whose overstable modes extend to long periods. It may render their intermediate period modes invisible. In a similar manner, the amplitudes of high frequency overstable modes with $\\ell=2$ and $3$ may be heavily suppressed by interactions with their lowest $\\ell$ overstable daughters. The irregular amplitude distributions among neighboring modes may be partially accounted for by this type of resonance. Mode variability may also play a role. We explore this in the next subsection. \\subsection{Mode Variability} \\label{subsec:variability} Excited g-modes in ZZ Ceti stars exhibit substantial temporal variations. Parametric instability may at least partially account for these variations. When $|\\delta \\omega| < \\gamma_d$, parametric instability gives rise to limit cycles in which the amplitudes and phases of parent and daughter modes vary on time scales as short as $\\gamma^{-1}_d$. Stable daughter modes may briefly attain visible amplitudes. Temporal amplitude variations may contribute to the irregular mode amplitude distribution seen in individual stars. Phase variations of a parent mode obey the equation \\begin{equation} {{d\\theta_p}\\over{dt}} = \\omega_p - {3\\over{\\sqrt{2}}} \\omega_p \\kappa {{|A_{d_1}| |A_{d_2}|}\\over{|A_p|}} \\cos \\Phi. \\label{eq:phaseevol} \\end{equation} At the equilibrium given by equations \\refnew{eq:nl-amplisteady1}-\\refnew{eq:nl-amplisteadyphi}, the parent mode's frequency is displaced from its unperturbed value such that \\begin{equation} \\omega_p^\\prime = {{d\\theta_p}\\over{dt}} = \\omega_p + {{\\delta \\omega \\gamma_p}\\over{\\gamma_{d_1} + \\gamma_{d_2} - \\gamma_p}} \\sqrt{1+\\left({{2\\delta\\omega}\\over{\\gamma_{d_1} + \\gamma_{d_2} - \\gamma_p}}\\right)^2}. \\label{eq:equilidomega} \\end{equation} This constant frequency shift is of order $10^{-9} s^{-1}$ for the $n=1$, $\\ell=1$ mode and of order $10^{-7} s^{-1}$ for the $n=2$, $\\ell=2$ mode. Frequency shifts in higher order overstable modes which are involved in limit cycles are predicted to be larger and time variable. During brief intervals of length $\\sim \\gamma^{-1}_d$, when the daughter mode energies are comparable to that of the parent mode, $|\\omega_p^\\prime - \\omega_p| \\sim \\gamma_d$, which is of order a few times $10^{-5} s^{-1}$, or a few $\\mu Hz$ in angular frequency. These shifts might account for the time-varying rotational splittings reported by Kleinman \\etal (\\cite{kleinmanZZPsc}) provided different $m$ components of the overstable modes are involved in different limit cycles. \\subsection{Miscellany} \\label{subsec:trapping} We briefly comment on two relevant issues. Gravitational settling produces chemically pure layers between which modes can be partially trapped. Modes that are trapped in the hydrogen layer have lower mode masses, and therefore higher growth rates and larger maximum values of $|\\kappa|$ than untrapped modes of similar frequency. This implies lower threshold energies for parametric instability. Nevertheless, trapping does not affect predicted photospheric amplitudes of $\\delta p/p$, and hence $v_h$ and $\\delta F/F$, since these are proportional to $A$ divided by the square root of the mode mass. In circumstances of small rotational splitting, the simple limit-cycles depicted in Figure \\ref{fig:nl-detail} are unlikely to be realistic. In such cases, different $m$ components of an overstable parent mode share some common daughter modes. This leads to more complex dynamics." }, "0003/astro-ph0003355_arXiv.txt": { "abstract": "We present observations of the young, oxygen-rich supernova remnant \\oursnr\\ taken by the {\\it Chandra X-ray Observatory\\/} during its orbital activation and checkout phase. The boundary of the blast-wave shock is clearly seen for the first time, allowing the diameter of the remnant and the mean blast-wave velocity to be determined accurately. The prominent X-ray bright ring of material may be the result of the reverse shock encountering ejecta; the radial variation of \\ovii\\ versus \\oviii\\ emission indicates an ionizing shock propagating inward, possibly through a strong density gradient in the ejecta. We compare the X-ray emission to Australia Telescope Compact Array 6 cm radio observations (Amy and Ball 1993) and to archival {\\it Hubble Space Telescope\\/} {\\oiii} observations. The ring of radio emission is predominantly inward of the outer blast-wave, which is\\break consistent with an interpretation of synchrotron radiation originating behind the blast-wave but outward of the bright X-ray ring of emission. Many (but not all) of the prominent optical filaments are seen to correspond to X-ray bright regions. We obtain an upper limit of $\\mysim 9\\times 10^{33}\\; \\mathrm{erg}\\; \\mathrm{s}^{-1}$ ($3\\sigma$) on any potential pulsar X-ray emission from the central region. ", "introduction": "\\label{sec:introduction} The supernova remnant \\oursnr\\ (hereafter E0102) in the Small Magellanic Cloud (SMC) is an interesting example of a young O-rich supernova remnant. It was first identified as a likely supernova remnant by \\citet{sm81} based on {\\it Einstein\\/} soft X-ray images of the SMC. \\citet{dtm81} found a filamentary \\oiii\\ shell ($\\mysim 24\\arcsec$ diameter); the remnant was not visible in \\halpha\\ (\\oiii/\\hbeta\\ $\\gtrsim\\kern-0.3em 60$), and they classified the remnant as O-rich. They noted diffuse \\oiii\\ and \\halpha\\ emission surrounding the \\oiii\\ filaments, with an emission ``hole'' of diameter $\\mysim 35$--$40\\arcsec$ surrounding the \\oiii, and suggested that the diameter of the hole could correspond to that of the blast-wave. In a follow-up spectroscopic study \\citep{td83}, O-rich material was found to have a full width velocity dispersion of $\\mysim 6500\\;\\kms$. If the O-rich material is assumed to be undecelerated ejecta, the 6.9 \\break pc diameter of the \\oiii\\ emission (assuming a distance \\break of 59 kpc) implies an age of $\\mysim 1000$ yr. More recently, the \\break remnant has been studied with the {\\it Hubble Space Telescope\\/} \\citep{b2000}. The remnant has also been observed in the UV \\citep{brdm89}, radio \\citep{ab93} and numerous times in the X-ray (Hayashi et al. 1994 and references therein). The X-ray spectrum of the remnant has proved difficult to understand; \\citet{hkmmth94} were unable to obtain a formally acceptable fit to the {\\it ASCA\\/} data, even with a rather complicated nonequilibrium ionization model, and concluded that the abundances in the plasma are inhomogeneous. By providing high spatial resolution, the {\\it Chandra X-ray Observatory\\/} will help us to solve these problems by reducing the spatial confusion that hampered earlier X-ray studies. ", "conclusions": "\\label{sec:discussion} It is plausible to identify the sharp outer edge of the X-ray emission as the location of the direct blast-wave interacting with the ambient medium and the inner edge of the bright ring as the reverse shock. The scalloping in the region between the bright ring and the rim might then be the result of Rayleigh-Taylor instabilities of the contact surface. The outer shock is brightest in the northeast, and at least 5 times fainter in the southwest. The nonthermal radio emission would be synchrotron emission originating behind the main blast-wave. The bright radio ridge in the northeast, corresponding to the brightest section of the outer X-ray shock, together with the slight elongation northeast to southwest elongation of the remnant suggests that the ambient medium may be slightly stratified in that direction. In the X-ray difference maps (e.g., Fig.~\\figthree), the remnant shows a clear layered morphology, in which the inner edge of the bright ring is enhanced in \\ovii\\ emission, while the outer part is enhanced in \\oviii. The \\neix\\ and \\nex\\ emission shows a similar pattern with \\neix\\ more prominent near the inner edge. The Ne emission is relatively more prominent outward of the peak in the \\ion{O}{7}-dominated band. This pattern suggests that we are seeing the ionization of the gas which has passed through the reverse shock. The southeast portion of the ring is particularly interesting in this regard, being much brighter and narrower than the rest of the ring; Figure \\figfour\\ shows radial profiles through this portion of the remnant. The steep rise at the inner edge and the lag of \\oviii\\ relative to \\ovii\\ are suggestive of a reverse shock, possibly propagating through a steep density gradient. In summary, the {\\it Chandra\\/} observations have allowed us to detect the outer blast wave for the first time, and to determine accurately the size of the remnant; the ring of radio emission is predominantly inward of the X-ray blast wave, which is consistent with an interpretation as synchrotron radiation originating behind the blast wave, but outward of the bright X-ray ring of emission. Complex variations are seen in the bright X-ray ring. The X-ray spectra vary throughout the remnant. Difference images centered on the prominent H-like and He-like lines of O and Ne suggest that the bright ring corresponds to an ionizing reverse shock. Many (but not all) of the prominent optical filaments are seen to correspond to X-ray bright regions." }, "0003/astro-ph0003025_arXiv.txt": { "abstract": "The relation between the cosmological evolution of the jet-disk symbiosis model for GRBs and the cosmic rays energy distribution is presented. We used two different Star Formation Rates (SFR) as a function of redshift and a Luminosity Function (LF) distribution to obtain the distribution in fluence of GRBs in our model and compare it with the data. We show a good agreement between the fluence distribution we obtain and the corrected data for the 4B BATSE catalogue. The results we obtain are generally valid for models that use jet physics to explain GRB properties. The fluence in the gamma ray band has been used to calculate the energy in cosmic rays both in our Galaxy and at extragalactic distances as a function of the redshift. This energy input has been compared with the Galactic and extragalactic spectrum of cosmic rays and neutrinos. Using our jet disk symbiosis model, we found that in both cases GRBs cannot give any significant contribution to cosmic rays. We also estimate the neutrino background, obtaining a very low predicted flux. We also show that the fit of our model with the corrected fluence distribution of GRBs gives strong constraints of the star formation rate as a function of the redshift. ", "introduction": "More than 30 years after their di\\-sco\\-ve\\-ry, thanks to the Burst and Transient Source Experiment (BATSE) and the Italian-Dutch satellite BeppoSAX, the scientific community now knows that Gamma Ray Bursts (GRBs) are isotropically distributed in the sky (Fishman $\\&$ Meegan 1995) and that at least some of them are at cosmological distances (GRB970228: Djorgovski et al. 1999b, GRB970508: Metzger et al. 1997, GRB971214: Kul\\-kar\\-ni et al. 1998, GRB980613: Djorgovski et al. 1999a, GRB980703: Djorgovski et al. 1998, GRB990123: Hjorth et al. 1999, GRB990510: Vreeswijk et al. 1999, GRB990712: Galama et al. 1999). But the present data available for redshift position and host galaxy localization are still too few to give us good statistics to study the evolution of GRBs and their redshift distribution. Before the discovery of GRB afterglows by BeppoSAX, the only way to study their distributions was to compare some GRB properties (like for example the intensity), with some parametric models (Fenimore $\\&$ Bloom 1995, Cohen $\\&$ Piran 1995, Kommers et al. 1999). Because of this lack of information, it is still necessary to assume that GRBs follow the statistical distribution of some other better known objects to obtain the GRBs fluence or flux distribution itself. The origin of GRBs is still controversial. According to different models, their progenitor can be identified with the merging of two neutron stars, or with the collapse of a massive star. In the model presented by Pugliese et al. (1999), GRBs are created inside a pre-existing jet in a binary system formed by a neutron star and an O/B/WR companion, where the input energy comes from the collapse of the neutron star into a black hole and the emission is due to synchrotron radiation from the ultrarelativistic shock waves that propagate along the jet with a low-energy cut-off in the electron distribution. Following this scenario, the birth of GRBs cannot happen too far from the region where the progenitor formed, and this implies that their rate should be connected with the Star Formation Rate (SFR). Already other authors studied the connection between the SFR and GRBs flux distribution. For example, Wijers et al. (1998) showed that the assumption that the GRB rate is proportional to the SFR in the universe is consistent with the GRB flux distribution. In Sect. 2 we calculate the cumulative distribution of GRB fluences using two SFR distributions as a function of redshift, the one by Miyaji (Miyaji et al. 1998), and the other by Madau (Madau et al. 1996). We compared it with the data from the BATSE catalogue. In Sect. 3 we calculate the maximum energy available in our model to obtain high energy cosmic rays. In Sect. 4 we present our results for the contribution of GRBs to the cosmic rays distribution, both Galactic and extragalactic and in Sect. 5 the eventual contribution from GRBs to the neutrino flux. ", "conclusions": "In the first part of this article we calculated the GRB rate and compared the corresponding cumulative distribution in fluence with the observational data. There were two main points to decide on: a) which luminosity function distribution and b) which star formation rate were the best to reproduce the data. We checked if in our jet model GRBs were standard candles. But we did not obtain any good fit, therefore we tried a power law for the luminosity distribution, $\\phi(f) \\propto f^{-\\beta}$. Together with this function we used the SFR from Miyaji et al. (1998), in which the SFR grows linearly from $z=0$ up to about $z=2$ and after this redshift is flat up to $z=6$. There was not a good agreement between the distribution in fluence we obtained and the corrected data by Petrosian for the 4B BATSE catalogue. Therefore we used the SFR model from Madau et al. (1996), in which the rate follows the same behavior of Miyaji's up to a redshift $z^{\\star}$, and after this redshift the SFR begins to decrease. There are still some uncertainties about the shape of this second part of the distribution, so we considered different curves with different slopes (depending on a constant $a_2$) for this decreasing part. We have only two free parameters that we can change to fit the data, the power law index $\\beta$, and the exponential index $a_2$. The redshift $z^{\\star}$, and the upper limit $f_{\\star \\rm{b}}$ of the interval in which the LF is defined are not really free parameter because their range is limited by the observations. We can reproduce the 4B BATSE corrected data by Petrosian using the following values: redshift $z^{\\star}=1.7$, $\\beta=1.55$, $f_{\\star \\rm{b}}=2.2 \\times 10^{-4}$ and $a_2$ in the range [0.8,1.3]. These parameters are remarkably constrained. Thus, given a final model for GRBs and a cosmological model, we may be able to derive strong limits on cosmological parameters. The key point of the second part of the article is the calculation of the GRB rate inside and outside our Galaxy. The results we obtained depends mainly on three other parameters, the value of the Hubble constant $H_{\\rm o}$, the opening angle of the jet $\\theta_{\\rm j}$, and the power law index $\\alpha$ we assumed for the electron distribution in our model (see Eq.~\\ref{plato}). The choice of $\\alpha=2$ has been done in the first version of our model, where even if we simplified in many places the physics used, we obtained a good agreement with the data. Any possible small changes in this interval will influence the results obtained in this work in a marginal way, because the dependence on $\\alpha$ in the expression of $N$ is not strong. However, putting $\\alpha = 3$, which is not suggested by the data, would strongly influence our results. On the other hand, changes in $H_{\\rm o}$ and $\\theta_{\\rm j}$, because of the strong dependence in $N$ will influence the results. It is interesting to note that both, a lower value of the Hubble constant and a smaller opening angle of the jet, go into the direction of decreasing $N$. But these changes cannot influence our results because the interactions of cosmic rays with the microwave background and the large difference between the Galactic GRBs and the leakage time of CR's rule out that {\\it in our jet model} GRBs can give any contribution beyond $10^{18}$ eV. In the calculation of the contribution from GRBs to the neutrino flux, it is important to define the parameters that characterize the neutrino production. The number of hits in the $p \\gamma$ collisions has been obtained considering only the major photohadronic interaction channel, the one that gives a single pion. At higher energies it is possible to have the channels for the production of multi-pions, $p \\gamma \\to n 2 \\pi^{+} \\pi^{-}$ and $p \\gamma \\to n 3 \\pi^{+} 2 \\pi^{-}$, but as the energy increases the corresponding cross sections decrease. We used only the first channel with the highest cross section, because the energies involved are not high enough to require secondary channels. In the context of our {\\it jet-disk symbiosis model} for GRBs, there is only one way to make the extragalactic contribution to CRs significant at the highest energies, and that is to drastically increase the CR energies per GRB deposited. We consider this implausible in the context of our model because of energy conservation requirements. We use $10 \\%$ of the entire energy available, and so the CR contribution may be increased over our simple calculation by a factor of a few, but not more. This implies that the contributions given by each GRB to CR and neutrino flux cannot be much bigger than the energy emitted in the $\\gamma$-ray band. Using a relatively small set of parameters, the jet-disk symbiosis model applied to GRBs, a tested star formation rate and the fundamental physics of the photohadronic interactions we arrive at the conclusion that GRBs are not able to give any significant contribution to the high energy cosmic ray spectrum both inside and outside our Galaxy and predict only a low flux of neutrinos. A main conclusion of this work is that fitting the corrected fluence distribution of GRBs with the jet-model is well possible. The fit allows strong constraints of the star formation rate as a function of redshift. \\vspace{0.3cm}" }, "0003/astro-ph0003207_arXiv.txt": { "abstract": "A southern hemisphere survey of methanol emission sources has been carried out using the ATNF Mopra millimetre telescope. 85 sources, the majority of them masers, have been detected in the $8_0-7_1A^+$ transition of methanol at 95~GHz. Together with a similar northern hemisphere survey this completes the search for 95-GHz methanol emission from the Galactic Plane. The previously found correlation between intensity of methanol emission at 44 and 95~GHz is confirmed here with the larger sample of sources. The results of LVG statistical equilibrium calculations confirm the classification of these sources as class~I methanol masers pumped through collisional excitation. ", "introduction": "Methanol, OH and \\water\\/ masers are all frequently associated with massive star formation regions, however, methanol masers offer more possibilities for the study of star forming regions than either OH or \\water\\/, because there are numerous transitions in the microwave region of spectrum. One of the most widespread methanol masers is the $7_0-6_1A^+$ transition at 44~GHz. About 50 masers from this transition have been detected in the northern hemisphere (Morimoto, Ohishi \\& Kanzawa 1985, Haschick, Menten \\& Baan 1990, Bachiller et al. 1990, Kalenskii et al. 1992, Kalenskii et al. 1994) and a similar number have been found in the southern hemisphere (Slysh et al. 1994). According to the empirical scheme of Menten (1991), the 44~GHz $7_0-6_1A^+$ transition is a class~I methanol maser. Class~I methanol masers differ from class~II methanol masers in that they are not directly associated with compact H{\\sc ii} regions and OH masers, and the two classes also emit in different transitions. The $8_0-7_1A^+$ transition at 95~GHz is an analogue of the 44-GHz masing transition (see Fig.~1 of Val'tts et al. (1999)). The upper ($8_0$) energy level of the 95-GHz transition is 18.5K higher than the upper ($7_0$) energy level of the 44-GHz transition and modelling (Cragg et al. 1992) shows that strong maser emission is also expected from this transition. Val'tts et al. (1995) carried out a search for 95-GHz methanol masers with the Onsala radiotelescope which detected a large number, mostly at the position of 44-GHz masers and with spectra similar to the spectra of the 44-GHz masers. The intensity of 95-GHz masers was found to correlate with intensity of 44-GHz masers, and was on average about 0.5 of the intensity 44-GHz masers. In this paper we present the results of the first search for 95-GHz methanol masers in the southern hemisphere, carried out with the ATNF Mopra telescope in Australia. This search completes a whole sky survey of class~I methanol masers source in two transitions, at 44 and 95~GHz. ", "conclusions": "The spectra of the 95-GHz $8_0-7_1A^+$ methanol emission sources found in this survey are in general similar to the spectra of the corresponding 44-GHz $7_0-6_1A^+$ sources. The emission in the two transitions typically covers the same velocity range, has approximately the same number of spectral features with very similar radial velocities, and in some cases even the same relative intensities of the components. In Table~\\ref{MopraP} we list the single strongest feature in each spectrum at 95~GHz and the corresponding spectral features at 44~GHz of the sources detected by Slysh et al. (1994). One can see that there is always a corresponding spectral feature at 44~GHz to every 95~GHz spectral feature from Table~\\ref{t4}, and their radial velocities agree in general to within 0.1~\\ks. The line width of the 95-GHz components is in general somewhat larger than the line width of the corresponding 44-GHz features, partly due to a lower spectral resolution in the 95~GHz observations, but nevertheless there are many very narrow 95-GHz features with a line width less than 1~\\ks. The peak flux density of the 95-GHz components is generally lower than the flux density of the 44-GHz features. Fig.~\\ref{MopraOnsala} shows a comparison between the flux densities of spectral features with the same radial velocities from the two transitions. In constructing this plot data on all available sources were used, including the results of this work and of the observations at Onsala (Val'tts et al. 1995). The straight line (with a correlation coefficient $r$=0.73) shows the best fit linear dependence which was found to be : \\begin{figure*} \\resizebox{\\hsize}{!}{\\includegraphics{fig11.eps}} \\caption{A correlation between 95 GHz and 44 GHz peak flux density. Straight line is a best fit: $y$=(0.32$\\pm$0.08)$x$-(8.1 $\\pm$ 2.7), correlation coefficient - $r$=0.73. } \\label{MopraOnsala} \\end{figure*} \\begin{equation} S(95)=(0.32 \\pm 0.08) \\times S(44) - (8.1 \\pm 2.7) \\end{equation} Although the scatter is quite large, on average the 95-GHz methanol masers are a factor of 3 weaker than the 44-GHz masers. This result is consistent with the findings of Val'tts et al. (1995) who found a linear dependence between integrated fluxes at two frequencies. We have used peak flux densities rather than luminosities since our survey is flux density limited by the sensitivity of the instrument, and luminosities would give a spurious correlation due to the multiplication of the flux densities at the two frequencies by the same distance squared. The slope of the dependence between integrated flux densities found by Val'tts et al. (1995) is $0.52\\pm0.05$, which is larger than the slope $0.32\\pm0.08$ found in this paper for the peak flux densities. This difference may be due to the larger average line width of 95-GHz masers mentioned above. The correlation between the peak flux density and the observed similarity in the spectra of the two transitions is strong evidence in favour of the suggestion that the emission from both transitions arises from the same spatial location. A comparison of published high resolution maps of the 44- and 95-GHz class I methanol masers in DR21(OH) and W33MetC shows that their images are very alike and consist of the same number of isolated maser spots (Plambeck \\& Menten 1990, Pratap \\& Menten 1992, Kogan \\& Slysh 1998, Slysh et al. 1999), consistent with this hypothesis. The two transitions belong to the class~I methanol masers (Menten 1991), which are thought to be pumped through collisional excitation. The difference between the two transitions is that the upper level of the 95-GHz $8_0-7_1A^+$ transition is 18.5~K above the upper level of the 44-GHz $7_0-6_1A^+$ transition. Therefore the population of the former is expected to be lower than the population of the latter, resulting in the lower intensity of 95-GHz emission compared to the intensity of the 44-GHz transition, although it is difficult to estimate the difference without any knowledge of the kinetic temperature and particle density in the source. \\begin{table} \\caption{LVG calculation results: ratios of 44/95 intensities for four models. Methanol density, divided by velocity gradient, for all models is $0.67\\times 10^{-2}$ cm$^{-3}$/(km s$^{-1}$/pc).} \\label{LVGtable} \\begin{tabular}{lrrr} \\hline \\cthead{Model}&\\cthead{$T_{kin}$}&\\cthead{n$_{H_2}$}&\\cthead{44/95 GHz}\\\\ & (K) &\\cthead{} &\\cthead{intensity ratio}\\\\ \\hline 1 & 20 & 0.56E+5 & 3.3 \\\\ 2 & 50 & 0.56E+5 & 1.7 \\\\ 3 & 100 & 0.56E+5 & 0.4 \\\\ 4 & 20 & 0.56E+6 & 1.2 \\\\ \\hline \\end{tabular} \\end{table} We used LVG code to calculated the intensity ratios of the $7_0-6_1A^+$ and $8_0-7_1A^+$ transitions in a collisional excitation model for four different parameter sets. The model parameters and the intensity ratios are presented in Table~\\ref{LVGtable}. The collisional selection rules are based on the paper by Lees \\& Haque (1974) and imply that $\\Delta K=0$ collisions are preferred by a factor of four. For model 1 with a gas temperature 20~K and density $0.56\\times 10^5$ cm$^{-3}$, the ratio of the 44- and 95-GHz intensities is 3.3, i.e., close to the mean observed ratio. The 95~GHz intensity is lower due to the lower population of the $8_0A^+$ level relative to that of the $7_0A^+$ level and due to a weaker inversion at 95~GHz. Increasing either the gas temperature or the density decreased the model ratio below the observed value. Thus, our results favour class I maser model with gas temperature about 20 K and density less than 10$^6$ cm$^{-3}$." }, "0003/astro-ph0003031_arXiv.txt": { "abstract": "We have used the two spectrometers on the Infrared Space Observatory to observe the HH\\,7-11 flow, its red-shifted counterpart, and the candidate exciting source SVS\\,13, in the star formation region NGC\\,1333. We detect atomic (\\oia, \\oib, \\sii, \\cii) and molecular (\\hii, CO, H$_2$O) lines at various positions along the bipolar flow. Most of the observed lines can be explained in terms of shock-excited emission. In particular, our analysis shows that dissociative (J-type) and non-dissociative (C-type) shocks are simultaneously present everywhere along both lobes of the flow. We confirm the low-excitation nature of the Herbig-Haro nebulosities, with shock velocities \\mvs$\\lsim 40-50$ \\kms. Toward both lobes of the outflow we find pre-shock densities of $n_0 \\sim 10^4$ \\cmthree\\ for both the J and C components, implying \\mbo$\\sim$100\\mug\\ for \\mbo$\\propto n_0^{0.5}$. In the central region of the flow, close to the exciting source, the pre-shock density deduced for the C-shock component is $n_0\\sim10^5$\\cmthree, suggesting a magnetic field $\\sim$3 times stronger. We propose that the deficiency of gas-phase water in the post C-shock regions is due to freezing onto warm grains processed through the J-shock front and traveling along the magnetic field lines. The total observed cooling from the dissociative shock components is consistent with the power lost by a slow molecular outflow accelerated by a fast neutral H{\\sc i} wind. Finally, the skin of the cloud seen in projection toward the flow appears to be weakly photo-ionised by BD +30$^{\\circ}$\\,549, the dominant illuminating source of the NGC\\,1333 reflection nebula. ", "introduction": "\\label{intro} Atomic and molecular outflows trace the mass loss from protostellar objects, which is a fundamental characteristic of the formation and evolution of low mass stars. These outflows are often traced by the optical Herbig-Haro (HH) objects, shock-excited nebulosities which mark the interface between outflowing and circumstellar material. One of these systems, which since its discovery (Herbig~\\cite{H74} ; Strom et al.~\\cite{Setal74}) has been subjected to a detailed multi-wavelength analysis, is HH\\,7-11. The system is relatively bright and lies in the very active star forming NGC\\,1333 region (Aspin et al. ~\\cite{ASR94}; Bally et al.~\\cite{Betal96}). A distance to the outflow of 350pc (Herbig \\& Jones~\\cite{HJ83}) is widely adopted in the literature, although Cernis (\\cite{C93}) proposes 200pc. Early optical spectroscopic studies show that the HH\\,7-11 outflow has a complex velocity field and low excitation (Solf \\& B\\\"ohm~\\cite{SB87}; B\\\"ohm \\& Solf ~\\cite{BS90}), a conclusion further supported by near infrared studies (Hartigan et al.~\\cite{HCR89}; Carr~\\cite{Carr93}). Early on, a well defined CO bipolar outflow was detected associated with this system (Snell \\& Edwards~\\cite{SE81}; Bachiller \\& Cernicharo~\\cite{BC90}), which also is observed in some other molecules like HCO$^+$ and H$_2$O (Mehringer~\\cite{M96}; Cernicharo et al.~\\cite{CBG96}). The HH\\,7-11 outflow has an unusual morphology; the blue outflow lobe is made up of an arc-shaped chain of knots, while the red one is invisible at optical wavelengths. The red lobe is, however, detected in the (1-0)S(1) \\hii\\ line at 2.12\\um\\ having a very ragged appearance (see Fig~\\ref{map}). Both lobes have a total extension of $\\sim $ 2\\amin. It has been thought that the driving source of the outflow is the infrared star SVS\\,13 (Strom et al.~\\cite{Setal76}), a conclusion partially supported by the proper motion measurements of the knots (Herbig \\& Jones\\cite{HJ83}) and the source observed outbursts (Goodrich~\\cite{G86}; Eisl\\\"offel et al. ~\\cite{Eetal91}; Liseau et al.~\\cite{LLM92}). The source has a luminosity of $\\sim 85$ \\lsun\\ (Molinari et al.~\\cite{MLL93}). Recent high angular resolution ($\\sim$ 0.\\asec 3) VLA continuum observations at 3.6 cm suggest another nearby embedded source (VLA\\,3) as a likely candidate, based on its better alignment with the HH string. Interferometric observations (Bachiller et al. ~\\cite{Betal98}) at 1.3 and 3.5 mm with better than $0.\\arcsec 2$ resolution, however, have not confirmed this. The interferometric observations have led to the discovery of a second jet emanating from a more deeply embedded source 14.\\asec 5 away from SVS\\,13, named SVS\\,13B (Grossman et al.~\\cite{Getal87}). One of the reasons why the HH\\,7-11 flow has been so intensively studied (see e.g. Reipurth~\\cite{R94}), is that it was the first system showing clear signatures of a high velocity outflow in both neutral and molecular gas tracers (Lizano et al.~\\cite{Letal88}; Rodriguez et al.~\\cite{Retal90}; Giovanardi et al.~\\cite{Getal92}). This was a major step forward in the interpretation of molecular outflows as being driven by faster but more tenuous (than the outflow entrained gas) atomic stellar winds (Masson \\& Chernin~\\cite{MC93}; Raga et al.~\\cite{Retal93}). This has led to a more careful analysis of the energetics and shock conditions associated with the ionic/atomic and molecular gas outflows (Raga~\\cite{Ra91}). HH\\,7-11 is also one of the few examples where it is possible to disentangle the contributions of shock excited and fluorescent emission from its near infrared H$_2$ spectra (Gredel~\\cite{G96}; Fernandes \\& Brand~\\cite{FB95}; Everett~\\cite{E97}). In the present study, we take advantage of the capabilities of the Infrared Satellite Observatory (ISO, Kessler et al.~\\cite{Ketal96}) spectrometers to study the mid- and far infrared emission line spectra from the HH\\,7-11 red and blue lobes and around the driving source SVS\\,13. The observations are described in Sect.~\\ref{obs}; the results are presented and discussed in Sect.~\\ref{res} and following, and the main conclusions are summarised in Sect.~\\ref{summary}. ", "conclusions": "\\label{summary} The HH\\,7-11 flow, together with its red-shifted counterpart and SVS\\,13 (the candidate exciting source) have been studied via atomic, ionic and molecular spectroscopy. A complex scenario emerges, where: \\begin{enumerate} \\item we have detected atomic (\\oia, \\oib), ionic (\\cii, \\sii) and molecular (\\hii, CO and \\ho) lines along the flow (both lobes) and toward SVS\\,13. \\item the low-excitation shock nature of the HH nebulosities along the flow is confirmed. Spectral signatures of C and J shocks are ubiquitously found along the HH\\,7-11 flow and its red-shifted counterpart. Our estimates for the shock velocities are \\mvsj$\\lsim40-50$ \\kms\\ and 15$\\lsim$\\mvsc$\\lsim$20 \\kms. The pre-shock density is $\\sim10^4$ \\cmthree\\ toward the blue and the red lobe; for the C component only, we find $n_0\\sim 10^5$ \\cmthree\\ at the location of SVS\\,13. \\item there is indirect evidence for an ordered \\bb\\ field oriented parallel to the direction of the flow. The magnitude of the magnetic field is \\mbo$\\sim100$\\mug\\ on the lobes, increasing to $\\sim300$\\mug\\ at the position of the flow origin; these figures, however, can vary of a factor 3 in each direction. \\item the gas-phase in the post C-shock region is deficient in \\ho. We presented evidence that this may be due to freezing onto warm grains processed through the J-shock front and traveling downstream along the magnetic field lines. \\item the asymmetry in optical and NIR properties among the two lobes of the outflow is probably not caused by different pre-shock densities or shock velocities, supporting the hypothesis of higher extinction values toward the red lobe. \\item the total J-shock cooling is compatible with the molecular outflow being accelerated by the fast neutral wind detected in H{\\sc i} and CO. \\item the whole flow area appears to be associated with a faint PDR illuminated by BD +30$^{\\circ}$\\,549, the source responsible for the illumination of the whole NGC\\,1333 nebula. \\end{enumerate} We thank L. Testi and M. Cecere for their assistance with the observations and data reduction of the \\hii\\ 2.12\\um\\ image presented in Fig.~\\ref{map}. The staff of the Mt. Palomar 60\\asec\\ telescope is also acknowledged. We also thank an anonymous referee whose comments improved the paper, and L.F. Rodriguez for his comments on an early version of this manuscript. The ISO Spectral Analysis Package (ISAP) is a joint development by the LWS and SWS Instrument Teams and Data Centers. Contributing institutes are CESR, IAS, IPAC, MPE, RAL and SRON." }, "0003/astro-ph0003177_arXiv.txt": { "abstract": "We have compiled the near infrared Point Source Catalogue (PSC) towards the Magellanic Clouds (MCs) extracted from the data obtained with the Deep Near Infrared Survey of the Southern Sky -- DENIS (Epchtein et al. \\cite{ept1}). The catalogue covers an area of of $19.87\\times 16$ square degrees centered on $(\\alpha,\\delta)=(5^h27^m20^s$,$-69\\degr00\\arcmin00\\arcsec)$ for the Large Magellanic Cloud (LMC) and $14.7\\times 10$ square degrees centered on $(\\alpha,\\delta)=(1^h02^m40^s$,$-73\\degr00\\arcmin00\\arcsec)$ for the Small Magellanic Cloud (SMC) at the epoch J2000. It contains about $1\\,300\\,000$ sources towards the LMC and $300\\,000$ sources towards the SMC each detected in at least 2 of the 3 photometric bands involved in the survey ($I$, $J$, $K_s$). $70\\%$ of the detected sources are true members of the Magellanic Clouds, respectively and consist mainly of red giants, asymptotic giant branch stars and super-giants. The observations have all been made with the same instrument and the data have been calibrated and reduced uniformly. The catalogue provides a homogeneous set of photometric data. ", "introduction": "The DENIS project aims to survey the entire southern hemisphere simultaneously in three photometric bands, $I$ (Gunn--i, $0.8\\, \\mu m$), $J$ ($1.25\\, \\mu m$) and $K_s$ ($2.15\\, \\mu m$) with a spatial resolution of $1\\arcsec$ in $I$ and $3\\arcsec$ in the $J$ and $K_s$ bands, and limiting magnitudes of $I=18$, $J=16$, $K_s=14$. See Epchtein et al. (\\cite{epal}) for the first general release of DENIS data. Here we present a catalogue of DENIS Point Sources towards the Magellanic Clouds, requiring that objects are detected in at least two of the three photometric bands. At the distance of the Magellanic Clouds, $(m-M)=18.45\\pm0.1$ for the LMC and $(m-M)=19.0\\pm0.1$ for the SMC according to Westerlund (\\cite{west}), our catalogue contains: (1) all Asymptotic Giant Branch stars (AGB), except those with shells optically thick at $2\\mu m$ and the faintest stars at the very beginning of the Early AGB branch (E-AGB), (2) upper Red Giant Branch stars (RGB), (3) most of the super-giants except those brighter than $I=10.5$, $J=8.0$, $K_s=6.5$ because they saturate the detectors, (4) relatively bright post-AGB stars. The catalogue will thus be a major tool for statistical studies of the post-main sequence stellar populations of the Magellanic Clouds. Dwarfs and giants are the main galactic sources seen in front of the Magellanic Clouds (Ruphy et al. \\cite{ruph}). Compared to earlier spectroscopic and photometric surveys of the Magellanic Clouds for red giants and super-giants, and for stars on the AGB, probably we find a few hundreds times more sources, for several reasons: (1) previous surveys were not sensitive enough (Westerlund \\cite{west0}, \\cite{west1}; Sanduleak \\& Philip \\cite{saph}; Westerlund et al. \\cite{weal1}, \\cite{weal2}; Rebeirot et al. \\cite{reb0}), (2) they were spatially limited (see e.g. Blanco, McCarthy, and Blanco \\cite{bmc}; Blanco \\& McCarthy \\cite{blamc}), (3) they were restricted to a peculiar type of objects (e.g. Hughes \\cite{hug} in his search for Miras variables, Rebeirot et al. \\cite{reb} in their search for carbon stars). About $1/4$ of the sources discovered in these surveys were later observed in the $JHK(L)$ infrared photometric bands (e.g. Hughes \\& Wood \\cite{hugwo}, Costa \\& Frogel \\cite{cofrog}). DENIS provides simultaneous $IJK_s$ observations of the entire Clouds, with a good sensitivity, and connecting for the first time the traditional optical and infrared wavelengths domains by simultaneous observations. Sect. $2$ describes the instrument characteristics and the observing technique. Sect. $3$ describes the data reduction procedure in the two ``data analysis centers`` with particular attention to: flat and bias subtraction, point spread function, and astrometric and photometric calibration. Sect. $4$ discusses the quality of the data with regard to the selection criteria applied and to the completeness reached. Sect. $4.2$ discusses in particular the foreground sources belonging to our Galaxy. Finally, Sect. $5$ describes the content of the catalogue and Sect. $6$ gives conclusive remarks. The catalogue is available through the Strasbourg Astronomical Data Center (CDS); it carries the number II/228. ", "conclusions": "The catalogue is a suitable tool for the study of late--type stars in the Magellanic Clouds. These studies may involve the statistical separation of various species of stars, i.e. RGB and AGB (both O--rich and C--rich); the characterization of the mass loss properties of these stars, when combined with measurements in the mid and far--IR; the relations of infrared colours and magnitudes with variability, when combined with measurements of light curves (EROS, MACHO) or comparable photometric data (2MASS); the interpretation of the Hertzsprung--Russel diagram through theoretical evolutionary models; the investigation of metallicity effects inside the Magellanic Clouds and in comparison with our own Galaxy; the study of the history of star formation." }, "0003/hep-ph0003112_arXiv.txt": { "abstract": "We compute the energy spectra for massless Kalb-Ramond axions in four-dimensional anisotropic string cosmological models. We show that, when integrated over directions, the four-dimensional anisotropic model leads to infra-red divergent spectra similar to the one found in the isotropic case. ", "introduction": "The pre-big-bang (PBB) model of cosmology~\\cite{PBB} inspired by the duality properties of string theory, is faced, on the phenomenological side, with the question whether or not it can reproduce the amplitude and slope of the observed temperature anisotropy spectrum~\\cite{cmb} and of large-scale density perturbations. Within the PBB model, the inflationary expansion due to the dilaton field in the low-energy effective action of string theory, leads to an amplification of metric fluctuations as well as of quantum fluctuations of all the fields present in PBB cosmology. Such fields, which are not part of the homogeneous background whose perturbations we study, are for example the gauge fields and the pseudo-scalar partner of the dilaton field in the string theory effective action. \\par At first, it was thought that the PBB scenario could not lead to the observed scale-invariant Harrison-Zel'dovich spectrum of perturbations at large-scales. First-order scalar and tensor metric perturbations were found to lead to primordial spectra that grow with frequency~\\cite{stp}, with a normalization imposed by the string cut-off at the shortest amplified scales. These blue spectra have too little power at scales relevant for the observed anisotropies in the cosmic microwave background (CMB). In contrast, the axion energy spectra were found to be diverging at large scales, red spectra, leading to very large CMB anisotropies, in conflict with observations. \\par These results already rule out four-dimensional isotropic PPB cosmology. However, if one allows for internal contracting dimensions in addition to the three expanding ones, the situation is different. The axion field can lead to a flat Harrison-Zel'dovich spectrum of fluctuations for an appropriate relative evolution of the external and the compactified internal dimensions~\\cite{axion1,axion2}. Thus, it is possible that the amplification of quantum fluctuations of fields which are present in the PBB scenario, can generate via the seed mechanism~\\cite{seed} the observed anisotropy of the CMB radiation. \\par Considering an isotropic PBB model with extra dimensions, the amplification of electromagnetic vacuum fluctuations and of Kalb-Ramond axion vacuum fluctuations lead to interesting observational consequences within the context of primordial magnetic fields~\\cite{mmg} and large-scale temperature anisotropies~\\cite{rmmg}. In particular, massless axions as well as very light axions can exhibit a flat or slightly tilted blue spectrum which may reasonably fit the observational data~\\cite{rmmg,afrg}. (Even though an acoustic peak at $\\ell\\sim 350$ is excluded by experiments published after Ref.~\\cite{afrg} was completed, it is possible to shift this peak to $\\ell\\sim 220$ by closing the universe with a cosmological constant. More details about this model can be found in Ref.~\\cite{farg}.) \\par Recently it has been suggested that four-dimensional string cosmology models which expand anisotropically can also lead to blue or flat energy spectra for axionic perturbations~\\cite{aniso}. According to Ref.~\\cite{aniso}, one can instead of assuming internal extra dimensions~\\cite{rmmg}, consider an anisotropic four-dimensional background. This has become especially interesting in view of new results which show that the pre-big-bang phase may generically be homogeneous but anisotropic~\\cite{kr}. In Ref.~\\cite{aniso}, the axion spectrum is only computed for the part of phase space where the longitudinal component of the wave vector is sufficiently large. In this work we correct the result of Ref.~\\cite{aniso} and complete the computation to contain all directions in phase space. We then integrate the obtained spectrum over directions and compare it with the result for the isotropic PBB. We find that the anisotropic spectrum, when averaged over directions agrees roughly with the isotropic one. Therefore, anisotropic expansion during the pre-big-bang phase cannot solve the axion problem of four-dimensional string cosmology. ", "conclusions": "In total we can summarize the calculated spectrum by \\beq \\rho(\\o,s) \\simeq {\\o_1^4\\over 2\\pi^3}\\left\\{\\begin{array}{l} s^{-2\\mu} \\left({\\o\\over \\o_1}\\right)^{3-2\\mu } \\mbox{ if } k_Tk_T(k_T/k_1)^{-\\g/2/(1+\\gamma/2)}$, the spectrum of the produced Kalb-Ramond axions is flat. \\par To estimate the total energy density per logarithmic frequency interval we integrate the axion density $\\O_\\s(\\o,s)$ over $s$. For this we use \\beq d^3k=2\\pi k_Tdk_L\\wedge dk_T =4\\pi k^2ds\\wedge dk~, \\eeq where we have used $dk_L=kds+sdk$ and \\[ dk_T={-s\\over \\sqrt{1-s^2}}kds + \\sqrt{1-s^2}dk~. \\] Hence, we have \\bea \\lefteqn{\\O_\\s(\\o,\\eta) = \\int\\O_\\s(\\o,s,\\eta)ds} \\nonumber \\\\ &\\simeq& {1\\over \\rho_c}\\left[ \\int_0^{s_c(\\o)}\\rho_T(\\o,s)ds + \\int_{s_c(\\o)}^1\\rho_L(\\o,s)ds\\right] \\label{gen} \\\\ &\\simeq& g_1^2\\O_\\gamma(\\eta)\\left[ s_c \\left({\\o_1\\over \\o}\\right)^{9/13} + {0.5\\over s_c^2} \\right]. \\label{Oo} \\eea This spectrum is shown in Fig.~2. \\begin{figure}[ht] \\centerline{\\psfig{figure=fig2.eps,width=72mm}} \\caption{The energy density $\\O_\\s(\\o)/g_1^2\\O_\\g$, integrated over directions $s$ is shown as a function of $\\o$ (solid line). Comparing it with the isotropic result (dashed line) we conclude that, the two spectral indices are the same and, within our accuracy, also the amplitudes are comparable.} \\end{figure} Using $s_c \\simeq (\\o/\\o_1)^{3/13}$, which is a good approximation as long as $\\o\\le 0.1\\o_1$ it can also be seen directly from Eq.~(\\ref{Oo}) that the isotropic energy spectrum is nearly reproduced. The isotropic spectral index, $3-2\\sqrt{3}\\sim -0.464$ is actually replaced by $-6/13\\sim -0.463$. Inserting reasonable values for the string scale, $0.01\\le g_1<1$, we see that also in the anisotropic case axions are over-produced in unacceptable amounts. Even if the spectrum of the axions from wave vectors directed sufficiently far from the plane $k_L=0$, is scale-invariant, the enhancement of the spectrum in the vicinity of the plane $k_L=0$ leads to a total contribution which agrees with the one obtained in the isotropic case. Therefore, the model is excluded (see Ref.~\\cite{axion1}). So far we have mainly considered the case $\\a=-7/9$ and $\\b=-4/9$, but our results apply quite generically, as long as $\\g<0$ and thus the $k_L$-term dominates at sufficiently early times. But also if $\\g>0$, Eq.~(\\ref{vacT}) is an approximate solution on sub-horizon scales. In this situation, however the $k_T$-term dominates at sufficiently early times and continues to do so until the perturbation becomes super-horizon if the inequality given in Eq.~(\\ref{kTkL}) is violated. For $\\g>0$ this is the case outside a narrow cylinder around the $k_T=0$ axis. Therefore, the generic formula given in Eq.~(\\ref{gen}) always applies, but $s_c \\ll 1$, if $\\g<0$ and $s_c \\simeq 1$, if $\\g>0$. For general values of $\\a$ and $\\b$ we obtain \\bea \\lefteqn{\\O_\\s(\\o,\\eta) = \\int\\O_\\s(\\o,s,\\eta)ds} \\nonumber \\\\ &\\simeq& g_1^2\\O_\\gamma(\\eta) \\left[\\left({\\o\\over\\o_1}\\right)^{3-2\\mu q} \\int_0^{s_c(\\o)}(1-s^2)^{-\\mu q}ds \\right. \\nonumber \\\\ && \\left. ~~~~~~~ + ~~ \\left({\\o\\over\\o_1}\\right)^{3-2\\mu}\\int_{s_c(\\o)}^1 s^{-2\\mu}ds\\right]~. \\label{genint} \\eea The transition value of $s$ is given by \\beq \\sqrt{1-s_c^2}=s_c^{1+\\g/2}\\left({\\o\\over \\o_1}\\right)^{\\g/2}~. \\eeq If $\\g<0$ (i.e. $\\a<\\b$), the factor $\\left({\\o\\over \\o_1}\\right)^{\\g/2}$ is very large in most of phase space and hence $s_c\\ll 1$. On the other hand, if $\\g>0$ (i.e. $\\a>\\b$), the above factor is very small for the relevant frequencies, $\\o\\ll\\o_1$ and $s_c \\simeq 1$. A reasonable approximation is \\bea s_c &\\simeq \\left({\\o\\over \\o_1}\\right)^{q-1} & \\mbox{ if } ~~~\\g<0 \\\\ 1-s_c^2 &\\simeq \\left({\\o\\over \\o_1}\\right)^{2/q-2} & \\mbox{ if } ~~~\\g > 0~, \\eea where we have used the relation $ q=1/(1+\\g/2)$. Inserting these results in Eq.~(\\ref{genint}), the integrals can be approximated by \\beq \\O_\\s(\\o,\\eta) \\sim g_1^2\\O_\\g(\\eta)\\left({\\o\\over \\o_1}\\right)^n~, ~~~\\mbox{ where } \\eeq \\bea n &= 2+q-2\\mu q & = {1+\\a+2\\b\\over 1-\\b} ~\\mbox{ if } \\a<\\b~, \\\\ n &= 1+2/q-2\\mu & = {1+\\a+2\\b\\over 1-\\a} ~\\mbox{ if } \\a > \\b~. \\eea Clearly, since $\\a^2+2\\b^2=1$ and $\\a,\\b\\le 0$ it is $\\a+2\\b\\le -1$. This shows that the spectrum is never blue and becomes scale invariant only in the degenerate case with two static dimensions, $\\b=0$. This is also shown in Fig.~3, where the above approximation for the spectral index plotted as a function of $\\a$: the spectrum always remains red with a spectral index relatively close to the isotropic value, $n_{\\rm iso}=3-2\\sqrt{3}\\sim -0.46$, except in the extremal case, when two dimensions are frozen and $\\a=-1$. If one relaxes the condition that both $a$ and $b$ be expanding and just asks for volume expansion, $\\a +2\\beta<0$, there is another pair of values for the Kasner indices leading to a flat spectrum, namely $\\a= -1/3$ and $\\beta=-2/3$. However, if we want expansion in all three dimensions the spectrum is always red. \\begin{figure}[ht] \\centerline{\\psfig{figure=fig3.eps,width=72mm}} \\caption{The spectral index $n$ is shown as a function of the exponent $\\a$ of the expansion law. For $\\a = -1/\\sqrt{3}$, the isotropic case, our approximation is not very good since there $\\a=\\b$. This is reflected in the unphysical kink at this value of $\\a$. Clearly, the resulting spectrum is always red ($n<0$), with $ -0.4 > n\\ > -0.5$ except close to the degenerate case $\\a \\rightarrow -1$.} \\end{figure} To summarize, we find that anisotropic expansion has very little influence on the overall axion production and cannot cure the axion problem of four-dimensional pre-big-bang models. Only by allowing for extra dimensions one can escape this conclusion and obtain a scale invariant spectrum of axions as described in Refs.~\\cite{rmmg,afrg}. A 'realistic' string cosmology with a Kalb-Ramond axion can therefore be realized only in models with extra dimensions." }, "0003/astro-ph0003427_arXiv.txt": { "abstract": "We report on {\\it Hubble Space Telescope} Faint Object Spectrograph observations of the double QSO HS 1216+5032 AB ($z_e({\\rm A})=1.455$; $z_e({\\rm B})=1.451$, and angular separation $\\theta=9\\farcs1$). The spectral coverage is 910 \\AA{} to 1340 \\AA{} in the QSO rest-frame. An unusual broad-absorption-line (BAL) system is observed only in the B component: maximum outflow velocity of $\\sim5\\,000$ \\kms; probably a mixture of broad and narrow components. Observed ions are: \\ion{H}{i}, \\ion{C}{ii}, \\ion{C}{iii}, \\ion{N}{iii}, \\ion{N}{v}, \\ion{O}{vi}, and possibly \\ion{S}{iv} and \\ion{S}{vi}. We also discuss two outstanding intervening systems: (1) a complex \\ion{C}{iv} system at $z=0.72$ of similar strength in A and B, with a velocity span of $1500$ \\kms{} along the lines of sight (LOSs; LOS separation: $S_{\\bot}[z=0.72]\\approx 75~h_{50}^{-1}$ kpc); and (2) a possible strong \\ion{Mg}{ii} system at $z=0.04$ observed in B only, presumably arising in a damped Ly$\\alpha$ system. We assume \\hs{} is a binary QSO but discuss the possibility of a gravitational lens system. The size of Ly$\\alpha$ forest clouds is constrained using $S_{\\bot}\\approx 80~h_{50}^{-1}$ kpc at redshifts between $z=1.15$ and $1.45$. Four Ly$\\alpha$ systems not associated with metal lines and producing lines with $W_0>0.17$ \\AA{} are observed in both spectra, while five appear in only one spectrum. This sample, although scarce due to the redshift path blocked out by the BALs in B, allows us to place upper limits on the transverse cloud sizes. Modelling the absorbers as non-evolving spheres, a maximum-likelihood analysis yields a most probable cloud diameter $D=256~h_{50}^{-1}$ kpc and $2 \\sigma$ bounds of $17212$ $h_{50}^{-1}$ kpc could be set for the transverse sizes (spherical clouds, no evolution). A similar result (Smette et al.~\\cite{Smette}) was found for HE 1104$-$1805 AB ($\\theta=3\\arcsec$) but here the position of the lens $z_{\\rm l}$ was unknown. A transverse size of $D>60~h_{50}^{-1}$ kpc was estimated for $z_{\\rm l}\\approx 1$. Surprisingly, QSO pairs at very large projected distances seem to still show Ly$\\alpha$ lines common to both spectra. In the spectra of LB 9605 and LB 9612 ($\\theta=1\\farcm65$), Dinshaw et al. (\\cite{Dinshaw1}) find 5 such lines within $400$ \\kms{} and derive a most probable diameter of $1520$ $h_{50}^{-1}$ kpc at $1160$ $h_{50}^{-1}$ kpc at $1.70.32$ \\AA{} are expected in the wavelength interval covered by the FOS on the {\\it HST} (Weymann et al.~\\cite{Weymann1}). Therefore, the size-estimate uncertainties for low-redshift Ly$\\alpha$ absorbers are large. Studies using FOS spectra of QSO pairs have been made by Dinshaw et al. (\\cite{Dinshaw2}; $\\theta=1\\farcm29$, $D>600$ $h_{50}^{-1}$ kpc at $0.50.17$ \\AA{} yields four lines common to both spectra and five lines without counterpart in the other spectrum. For $W_0>0.32$ \\AA{} lines the numbers are two and four, respectively. Using a maximum likelihood technique, most probably diameters for spherical clouds of $192$ and $256$ \\hkpc{} are found for $W>0.32$ \\AA{} and $W>0.17$ \\AA{} lines, respectively. The $2 \\sigma$ limits derived using the cumulative distribution of the probability function are $136=1.3$. Assuming that the absorbers are filamentary structures lying perpendicular to the LOSs, transverse dimensions almost $50 \\%$ smaller than for spherical clouds are found. In both cases, the results of our analysis do not confirm the claim that the characteristic size of the Ly$\\alpha$ absorbers increases with decreasing redshift. Independently of the cloud models used, we note that there are significant equivalent width differences between lines in A and B. Also, there appears to be a trend of larger equivalent width differences with increasing line strength, while no velocity differences between common lines is found. This provides evidence that the absorbers are coherent entities. The results for each line sample suggest that the absorbers must have a smooth gas density distribution, with lower density gas being more extended. } \\end{enumerate}" }, "0003/astro-ph0003082_arXiv.txt": { "abstract": "Numerical results for particle acceleration at multiple oblique shocks are presented. We calculate the steady state spectral slope of test particles accelerated by the first order Fermi process. The results are compared to analytical treatments, for parameters, where the diffusion approximation does apply. Effects of injection and finite shock extend are included phenomenologically. We find the spectrum of accelerated particles to harden substantially at multiple oblique shocks and discuss the influence of the number of shocks compared to the obliquity itself. ", "introduction": "\\label{intro} The theory of first order Fermi acceleration is commonly used to understand non thermal particle spectra or source distributions of synchrotron radiation in various astrophysical objects (for reviews see Drury~\\cite{Drur83}; Blandford \\& Eichler~\\cite{BlEi87}; Kirk et al.~\\cite{KiMePr94}). In the diffusion approximation the steady-state spectral index of accelerated test particles is solely a function of the compression ratio of the shocked plasma flow. For strong shocks, the compression ratio is given by $r=\\rho'/\\rho=4$, where $\\rho'$ and $\\rho$ are the downstream and upstream densities respectively. This leads to a phase space distribution of accelerated particles $f(p)\\propto p^{-s}$ with the canonical spectral index $s=\\sca:=3r/(r-1)=4$. The synchrotron emission of such an (e.g.) electron distribution as a function of frequency is given by $\\epsilon(\\nu)\\propto \\nu^{-\\alpha}$, where $\\alpha=(s-3)/2$. Unless the injection into the acceleration process is very effective, (leading to a strongly modified shock structure), this is the hardest spectrum which can be produced through first order Fermi acceleration by a single shock for which the diffusion approximation is valid. In complex large scale structures like the galaxy or active galactic nuclei (AGN) particles may encounter several shocks. If the transport time between shocks is larger than the acceleration time at a single shock, the problem can be described by subsequent shocks, where (in the case of planar geometry) the downstream distribution is transported (and decompressed) upstream of the next shock. It is well known from the theory of diffusive shock acceleration, that the supply of an upstream power law distribution with the canonical spectral index leads to an amplification of the distribution, and the spectral index is not changed. However, the increased number of high energy particles is accompanied by a decrease of the number density at the low energy cutoff, leading to a flattening of the distribution at intermediate energies. This can be seen by calculating the spectrum as it is processed trough a number of shocks, suppressing new injection, and considering adiabatic decompression, as shown by Melrose \\& Pope~(\\cite{MePo93}). At sufficiently high energies, a power law with the canonical spectral index is always revealed. This applies for multiple identical shocks without losses, which are subject of this work.\\protect\\footnote{For a consideration of synchrotron losses in the framework of multiple shocks, see e.g. Marcowith \\& Kirk~(\\protect\\cite{MaKi99}) and the references therein.} In the limit of an infinite number of subsequent shocks with injection at each shock, the flattening of the spectrum (compared to a single shock) extends even to the highest energy particles, with a momentum dependence of $f(p)\\propto p^{-3}$ (White~\\cite{Whit85}; Achterberg~\\cite{Acht90}). Since the modification of the particle spectrum evolves from the low energy part of the spectrum due to further acceleration and adiabatic decompression, effects of injection can be very important. Including a theory of the injection process is well beyond the scope of this work. However, given a momentum $p_0$ at which particles are injected, we assume, that this is the momentum that divides particles which are able to diffuse across the shock from the thermal pool. Adiabatic decompression can shift the momentum of some particles to $p$60 spectral resolution over 5.3 to 40 microns and two modules providing R$\\sim$600 spectral resolution over 10 to 37 microns. \\begin{figure}[!ht] \\centering \\includegraphics[width=0.48\\textwidth]{houckj1_2.ps} \\caption[]{The 4 IRS modules on their common base plate.} \\label{irs} \\end{figure} The IRS instrument has no moving parts (``bolt-and-go'' philosophy). Each module has its own entrance slit in the focal plane. The low-resolution modules employ long slit designs that allow both spectral and one-dimensional spatial information to be acquired simultaneously on the same detector array. Two small imaging sub-arrays (``peak-up cameras'') in the short-low module (SL) will also allow infrared objects with poorly known positions to be accurately placed into any of the IRS modules' entrance slits. The high-resolution modules use a cross-dispersed echelle design that gives both spectral and spatial measurements on the same detector array. \\begin{figure}[!ht] \\centering \\includegraphics[width=0.48\\textwidth]{houckj2_2.ps} \\caption[]{Details on the IRS slits} \\label{slits} \\end{figure} \\begin{figure}[!ht] \\centering \\includegraphics[width=0.48\\textwidth]{houckj3_2.ps} \\caption[]{The IRS peakup camera} \\label{peakup} \\end{figure} ", "conclusions": "More information on the IRS and on upcoming deadlines for observing opportunities with SIRTF can be found at the following web sites:\\\\ \\\\ http://www.astro.cornell.edu/SIRTF\\\\ http://sirtf.caltech.edu" }, "0003/astro-ph0003156_arXiv.txt": { "abstract": "We report on photometric observations of V 1493 Aql during the early decline and highlight some uncommon aspects of the light curve . V 1493 Aql was hotter at maximum light than in the following phases, and was characterized by the presence of a long lasting secondary maximum, that, unlike in other novae, was quite red in color. The mean of three distance estimates yields d $\\sim 18.8\\pm 3.6$ Kpc. Such a large distance would place V 1493 Aql at the extreme outskirts of our Galaxy or even in an external Local Group galaxy. ", "introduction": "V 1493 Aql (= Nova Aql 1999 n. 1) was discovered by Tago (1999) as an 8.8 mag object on two films taken on July ~13.56 UT with a 55-mm f/3 camera lens. Nothing was visible on a film taken 4 days before. The precursor of the nova was too faint in quiescence to be recorded by the 1.2 m Palomar Schmidt, which sets the amplitude of the outburst in the $\\delta {\\rm m} \\ge 12$ mag range (Moro et al, 1999). Low-resolution spectra taken on July 14.6 UT by Ayani and Kawabata (1999) showed strong, broad Balmer lines with FWHM about 3400 ${\\rm kms^{-1}}$. Similar values are reported by Tomov et al (1999) on July ~15.9 UT who, in addition, remark the presence of FeII emission lines and the lack of absorption components in the Balmer emission lines. The decline of the nova was very fast and already on July 17.02 UT was fainter than 11.0, thus indicating a $t2$ of the order of 3 days. On August 3.89 the nova was already at $V=13.0$ (Lehky, 1999). We took advantage of an observing run at the Catania Observatory to monitor the following stages in the decline of the nova. The object underwent a secondary outburst which began shortly before our observations. In this letter we report our and other photometric observations of V 1493 Aql and discuss the implication of these data in the context of the classical nova phenomenon. ", "conclusions": "The very fast character of the light curve of V 1493 Aql associated with the conspicuous re-brightening (peak increase in $V \\sim 2.5$ mag with respect to the smoothed V curve and total duration of about 40 days) is not common in novae. This, together with the fact that V 1493 Aql is hotter at maximum light ($(B-V)<0.4$) than in the following phases: $(B-V)\\sim 1$ near the secondary maximum and $\\sim 0.7$ in the following decline) makes V 1493 Aql quite peculiar among novae. We considered the possibility that the rapid reddening of V 1493 Aql in the early phases might be due to the formation of dust, however we regard this unlikely because dust formation is associated with the presence of a ``deep minimum'' in the light curve, which is not observed. Moreover dust is generally found in slow novae. The large distance we have derived requires a comment. Clearly $d$ would be significantly reduced if the apparent magnitude at maximum were much brighter than 8.8. If $V_{max}\\sim 6.0$ the distance would be reduced to a more comfortable 6.2 Kpc. However, we consider it very unlikely that a $V = 6$ object might have escaped the attention of vigilant sky-watchers during a few days. The lack of detection on films taken on 9.9 July UT (Tago 1999), implies that even if the nova managed to escape the attention of the astronomers, it had to decline by $\\sim 3$ mag in an extremely short time. Considering that the observed decline already suggests a $t2$ less than 3 days, this would imply a decay by about 5 magnitudes in less than 6 days, thus making of V 1493 Aql the fastest known nova. The photometric and spectroscopic behaviour of V 1493 Aql is that of a nova; the presence of an emission line spectrum near the maximum of the outburst (Ayani \\& Kawabata 1999; Tomov et al 1999; Lynch et al 1999) is generally associated with the very fast character in the light curve. We are therefore reluctant to accept that this is a unique object which does not obey the empirical MMRD-like relations found for other members of its class. The three distance estimates are compatible, within 1.7 $\\sigma$. Although the distances estimated from the magnitude at maximum are uncertain, to the extent that the ``true'' maximum may have been missed, the fact that assuming $V_{max}=8.3$ we obtain distances which are compatible with the estimate from $V$ at day 15, suggests that the maximum may not have been missed by more than a few tenths of magnitude. A distance of $18.8 $ Kpc places V 1493 Aql at over 14 Kpc from the Galactic centre, thus the star is at the outskirts of our own galaxy. If however the distance is 25.5 Kpc, as derived from the magnitude at day 15, then the distance from the Galactic centre is about 20 Kpc and thus the nova is outside the Galaxy. In this case it would fall in a Local Group galaxy, which could be called the Aquila galaxy. That a Local Group galaxy at such low Galactic latitudes could have gone undetected is not implausible, so the fact that no Local Group galaxy is known in that direction does not allow to rule out this possibility. The existence of the Aquila galaxy may be disproved by number counts and/or radial velocity surveys." }, "0003/astro-ph0003360_arXiv.txt": { "abstract": "We describe the morphological composition of a sample of 518 galaxies in the field of CL1358+62 at z=0.33, drawn from a large {\\it HST} mosaic covering 53 sq.~arcmin. The sample is complete to $I$=22, corresponding to $M_V$=-18.5 in the rest frame. The galaxy morphologies have been independently classified by the authors of this paper and by Alan Dressler. Dressler's classifications place our work in context with the previous MORPHS study, and allow us to estimate the scatter between different sets of visual classifications. We restrict most of our analysis to the brighter part of the sample, $I<21$ ($M_V<-19.5$), where the scatter between the two sets of classifications is $\\sim$1 in morphological type. The scatter doubles at $I=22$, presumably due to the lower signal-to-noise and poorer sampling of faint, small galaxy images. To $I$=21 the two sets of classifiers agree on the fraction of early type galaxies (elliptical+S0): 72\\%. We conclude that CL1358+62 does not contain the large population of spiral galaxies found in other studies of clusters at $z\\sim0.3$, and that there is probably a significant spread in the degree of cluster evolution at intermediate redshift. The two groups of classifiers differ on the relative fraction of S0 and elliptical galaxies. We show that the distributions of ellipticities and bulge/total light cannot resolve this discrepancy. Nonetheless, we can derive significant constraints on physical models for the evolution of the galaxy population in CL1358+62. The higher ratio of S0 to elliptical galaxies (1.6) found by DF/MF/PvD requires that the evolution preserve the relative fraction of elliptical, S0 and spiral galaxies. Alternately, the lower ratio (1.1) found by AD requires that the evolution preserve the early-type to spiral ratio while increasing the S0 to elliptical ratio. In the latter case, a possible evolutionary mechanism is accretion of galaxies that predominantly evolve to S0's between $z$=0.33 and the present. We use our large body of spectra to make the correspondence between spectral and morphological type. Our data follow the pattern seen in the field at low redshift: emission line spectra are more prevalent among the later morphological types. The 11 identified k+a galaxies (absorption line spectra with strong Balmer lines) have S0--Sb morphologies. ", "introduction": "The WFPC2 camera on the Hubble Space Telescope ({\\it HST}) has made it possible to determine the morphologies of galaxies at intermediate redshift and beyond. It has been known for some time that the photometric and spectral properties of galaxies in intermediate redshift clusters differ from galaxies in nearby clusters; the population of blue, star-forming galaxies and post-starburst galaxies is larger at intermediate redshift, see e.g., \\cite{bo84}, \\cite{dr87}, \\cite{gd88}, \\cite{cs87}, and \\cite{dr92}. Given the correlation between the spectral and morphological properties of galaxies (\\cite{mm57}, \\cite{mo69}, \\cite{ke92}), we might be able to detect a corresponding evolution of galaxy morphology in intermediate redshift clusters. However, the evolution of morphology is likely to be more subtle than spectral evolution, since galaxies of the same morphological type can have significantly different star formation rates (e.g. \\cite{ja99}). In the most ambitious study of this sort with WFPC2 to date, the MORPHS group have classified over 1200 galaxies in 10 clusters at 0.37$<$z$<$0.56 (\\cite{sm97}). They find that S0 galaxies are less common than in low redshift clusters and that the ratio of S0's to E's within a radius of $\\sim$600 kpc (for H$_0$=50, q$_0$=0.5) decreases with redshift, falling from 2 in low redshift clusters to less than 0.5 at $z=0.5$. \\cite{an97} (see also \\cite{an98}) have independently classified galaxies in a WFPC2 image of Cl0939+4713, one of the less concentrated clusters in the MORPHS sample. They find a ratio of S0's to E's of $\\sim$2, quite comparable to a low-$z$ reference sample in the Coma Cluster. However, they classify 40-50\\% of the galaxies in Cl0939+4713 as spirals (S), in contrast with 20-30\\% S in a comparable region of the Coma Cluster. The sample of galaxies in Cl0939+4713 is relatively small ($\\sim$70), and redshifts are available for less than one third of these. \\cite{co98} present a study of the morphological types in three clusters at $z$=0.31, also using WFPC2 images. There is substantial overlap between the authors of this paper and the MORPHS group, and the two groups have attempted to adopt a consistent morphological system. At $z$=0.31, Couch et al.~find an excess of S's, with an abundance at small radii ($\\sim$400 kpc for $H_0$=50, $q_0$=0.5) approximately twice that in low-$z$ reference clusters. However, averaged over the three clusters, within 400 kpc, they find a ratio of S0's to E's at most slightly depressed relative to regions of comparable galaxy surface density in low-$z$ clusters. \\footnote{It is important to account for the morphology-density relation when we consider the morphological content of clusters. At high galaxy densities, typically found in the cores of clusters, the low-$z$ reference population becomes increasingly dominated by E's, \\cite{dr97}}. Note, however, that the average $z$ of the MORPHS clusters is larger (0.46). \\cite{lu98} report the morphological types in two more distant clusters at $z\\sim$0.9. One cluster, CL0023+04, appears to be composed of two low velocity dispersion groups, and contains predominantly S's. The other, CL1604+43, with a velocity dispersion of $\\sim$1200 km s$^{-1}$, contains $\\sim$76\\% early-types. In the latter case, the S0/E ratio is found to be 1.7$\\pm$0.9. This result is sensitive to the assumed morphological composition of the foreground/background population, but is evidence that the S0/E ratio does not decline smoothly with $z$. Our approach is complementary to the \\cite{dr97}, \\cite{an97}, and \\cite{co98} studies which predominantly describe the galaxy morphologies in cluster cores. We use mosaics of WFPC2 fields to study the galaxy population in a larger region (allowing larger galaxy samples per cluster), and we have acquired large numbers of spectra of cluster galaxies. The spectra remove ambiguity about cluster membership and allow us to directly connect the morphological and spectral properties of the galaxies. Our sample of clusters is x-ray selected, with x-ray luminosities exceeding 4$\\times$10$^{44}$ erg s$^{-1}$ in the 0.2--4.5 keV band. In CL1358+62 at $z$=0.33, we have drawn a complete sample of 518 galaxies to a magnitude limit $I$=22 from a WFPC2 mosaic image of CL1358+62 covering 53 square arcminutes. Spectra for 276 of the 518 galaxies in the morphological sample were previously obtained at the Multiple Mirror and William Herschel Telescopes. The color-magnitude relation of the 194 spectroscopically confirmed cluster members in the {\\it HST} mosaic (3 are fainter than $I$=22) has been previously described in \\cite{vd98}. The spectroscopic properties of 232 cluster members (some outside the {\\it HST} mosaic), as well as the cluster dynamics have been described in \\cite{fi98}. Our objectives in this paper are fourfold. (1) We introduce the morphological classification techniques that we will apply to our entire sample of clusters. (2) We classify the galaxies in CL1358+62 at z=0.33, comparing our classifications with those of an experienced external researcher, Alan Dressler. Our deep sample with two independent classifications provides a useful assessment of the scatter between WFPC2 visual morphological classifications at intermediate redshifts. (3) We describe the robust, classifier-independent conclusions and explore the physical implications of the differences between the two sets of classifications. (4) We connect the spectral and morphological types of the cluster galaxies. The paper is organized in the following fashion. In $\\S$ 2, we describe the photometric catalog from which the morphological sample was drawn. The two set of morphological classifications are discussed in $\\S$ 3. The morphological composition of the cluster and evidence for morphological evolution are presented in $\\S$ 4. The connection between the morphological and spectral properties of the galaxies is made in $\\S$ 5. $\\S$ 6 contains a brief discussion and conclusions. ", "conclusions": "For CL1358+62, we have acquired a unique data set including a large mosaic of HST fields and extensive spectroscopy that allows us to unambiguously determine cluster membership for galaxies with $M_V < -20$. We have directly compared the morphological classifications of two sets of classifiers for the galaxies in CL1358+62. The two sets of classifiers agree that (to a limit of $M_V$=-20) the fraction of early type galaxies (and therefore spirals) in this cluster at $z$=0.33 is indistinguishable from the fraction in comparable low-$z$ clusters. In contrast, previous workers, \\cite{an97} and \\cite{co98}, who also studied WFPC2 images of clusters at z$\\sim$0.3, found an elevated population of spirals compared with low-$z$ reference samples. Because our work confirms the reliability of early-type/spiral classifications from intermediate $z$ WFPC2 observations, we conclude that this is evidence for a dispersion in the evolution of intermediate-$z$ clusters. The two groups of classifiers differ on ratio of E to S0 galaxies in CL1358+62. DF/MF/PvD find a population of S0 galaxies (S0/E=1.6$\\pm$0.3) that is within 1.4$\\sigma$ of the low-$z$ reference sample, while AD finds a significantly smaller ratio (1.1$\\pm$0.2). This systematic difference is most likely related to the fact that the transition between S0's and intermediate luminosity E's is rather gradual. Many of the intermediate luminosity E's are thought to have disks, e.g.~\\cite{sc98}, \\cite{ri90}, and \\cite{jo94}. It may only be possible to resolve this issue by direct model fitting to images at low and intermediate $z$. Even though we conclude that we have not reliably determined the ratio of S0's to E's among the early types, our work significantly restricts possible evolutionary models. If we accept the MF/DF/PvD classifications, evolution must preserve the fraction of E's, S0's and S's as well as the morphology-density relation. If the AD classifications are correct, evolution must decrease the fraction of E's and increase the fraction of S0's while maintaining the fraction of S's. A possible mechanism for driving the evolution of the morphological mix in this latter fashion is accretion of additional galaxies from the spiral-rich infall region that become predominantly S0's. The cluster population within an $\\sim$1 Mpc radius must increase by a minimum of 50\\% from $z$=0.5 to the present day in order convert an intermediate $z$ population rich in E's to a low-$z$ population rich in S0's. It will be interesting to see whether such accretion can be produced in simulations of cluster formation. In most cluster formation scenarios, massive clusters form by the merging of pre-existing massive clusters with (presumably) similar populations of early type galaxies. It may therefore be difficult to double the ratio of S0 to E galaxies. We compare our morphological classifications to our previous spectral classifications and conclude that the morphologies of the spectrally ``active'' galaxies are as might be expected from the low-$z$ field population: the galaxies with emission lines are predominently spirals and the k+a (or E+A) post-starburst galaxies are typically early type disk galaxies (S0--Sb). We wish to acknowledge the generous contributions of Alan Dressler to this paper, including his independent classifications and insightful comments. We thank Gus Oemler for checking the classifications and for thoughtful comments on the manuscript, which led to several improvements. We thank Margaret Geller for a critical reading of an earlier version of the manuscript and helpful comments. Our referee, Ian Smail, and our editor, Greg Bothun, both made insightful comments that helped us clarify the paper. \\appendix \\newpage \\bigskip \\centerline{APPENDIX} \\centerline{SYSTEMATIC EFFECTS OF MEASUREMENT ERRORS} The effects of classification errors on the distribution of types can be important. Quantifying these effects is difficult because morphological classification is a subjective procedure, but we can gain some insight by considering simple models for the errors. We begin by assuming that the numerical type is based on a one dimensional measurement with a simple, constant error. This simple model would imply that any peaks in the distribution of types would be softened. If we adopt nominal intrinsic fractions of S:S0:E of 0.20:0.53:0.27, as is approximately correct for the inner 600 kpc of low redshift clusters (\\cite{dr97}), the errors will automatically decrease the fraction of S0's, and enhance the fraction of E's and S's. We can calculate an upper limit to the loss of S0's for our nominal 20:53:27 (S:S0:E) population with this model for the errors. We assume that the types are scattered with a mean absolute deviation (MAD) of 1, and all intermediate types are divided equally between adjacent types. The outcome of this experiment is a distribution of 0.27:0.38:0.34 (S:S0:E). In this case, all galaxies scattered beyond the normal type boundaries were assigned to the boundary type (i.e.~E). A MAD of 1 may be an overestimate as the difference in the types assigned by the two groups of classifiers is of this order. The intrinsic errors are $\\sqrt{2}$ smaller if the errors are independent. For these smaller errors the resulting distribution would be 0.24:0.43:0.33 (S:S0:E). In either case, the results serve to illustrate that the systematic effects can be very significant. Let us consider a second, more physical model for the visual classification errors. Here, we assume that morphological type is based on two independent variables with continuous distributions: bulge-to-total light fraction ($f_b$), and asymmetric features due to spiral arms ($A$). This is very similar to the quantitative classification devised by \\cite{ab96}. Galaxies with low $A$ will be classified as early type ($t < 0$) and then divided into E's or S0's based on whether $f_b$ is above or below a critical value. It has been argued that most $L*$ ellipticals have faint disks, e.g.~\\cite{ri90} and \\cite{jo94}. Similarly, the spiral classification will be based on a combination of $f_b$ and $A$. If the intrinsic distribution of $f_b$ is flat, then errors in $f_b$ will not change the ratio of S0's to E's. However, if the intrinsic distribution of $f_b$ is peaked, the errors will have a systematic effect. The sign depends on the details of the intrinsic $f_b$ distributions and the size of the errors. Contributing to this uncertainty, the errors can also be asymmetric, if for example, faint extended disks are missed in noisy data. The situation is similar for errors in $A$, where E's and Sa's now share a boundary. Extensive simulations are required to estimate the systematic effects of limited S/N on classification errors, taking into account the the point spread function of WFPC. We will undertake this effort in a future paper." }, "0003/astro-ph0003406_arXiv.txt": { "abstract": "The standard model of Gamma-Ray Bursts afterglows is based on synchrotron radiation from a blast wave produced when the relativistic ejecta encounters the surrounding medium. We reanalyze the refreshed shock scenario, in which slower material catches up with the decelerating ejecta and reenergizes it. This energization can be done either continuously or in discrete episodes. We show that such scenario has two important implications. First there is an additional component coming from the reverse shock that goes into the energizing ejecta. This persists for as long as the re-energization itself, which could extend for up to days or longer. We find that during this time the overall spectral peak is found at the characteristic frequency of the reverse shock. Second, if the injection is continuous, the dynamics will be different from that in constant energy evolution, and will cause a slower decline of the observed fluxes. A simple test of the continuously refreshed scenario is that it predicts a spectral maximum in the far IR or mm range after a few days. ", "introduction": "The standard model for GRB afterglows assumes that relativistic material is decelerating due to interaction with the surrounding medium. A shock wave is formed heating the surrounding matter to relativistic temperatures. It is assumed that both magnetic fields and accelerated electrons aqcuire an energy density which is a significant fraction of the equipartition value. In the simplest case, which will be referred to as the standard scenario, a single value of the energy and the bulk Lorentz factor is injected either as delta or as a top-hat function, of duration short respect to the afterglow. The total energy is fixed in time and equals the initial energy of the explosion. Slower moving material is essential to all models which use density gradients as a means of acceleration. Actually, if this is indeed the mechanism, most of the system's energy is carried by the slower material. This scenario overcomes the need for a clean environment. The fact that one now needs a substantially higher energy input can be addressed by a very energetic source such as a massive star. A similar situation exists in some cases of supernova. When a shock wave propagates through the envelope of the star arrives at the edge it accelerates, and higher and higher velocities are being imparted to a smaller fraction of the mass. In GRB afterglows, if such slower material with significant energy is ejected, it will affect the evolution in two major ways: first, the system becomes more energetic as time passes (refreshed shock scenario), therefore the temporal decay of the afterglow will be slower (Rees \\& \\Mesz 1998). Second, since the reverse shock will last for as long as the energy supply continues, it adds an additional long-living (reverse) emission component, typically at low frequencies. The emission from such a reverse shock was considered by \\cite{kumpir99} for the discrete injection case. With accurate enough observations of the afterglow temporal decay or good spectral sampling, especially at radio to mm frequencies, both of these features may be detected, and could therefore constrain the possibility of additional energy injection. ", "conclusions": "The dynamics and emission of the forward and reverse shocks is controlled by several factors, including the continuity and nature of the energy and mass input, the possible existence of external density gradients, and the strength of the magnetic fields in these regions. A contineous injection of energy with a lower Lorentz factor has as its main consequence that it tends to flatten the decay slopes of the afterglow, after it has gone through the maximum. An external density gradient (e.g. as in a wind with $\\rho_{ext} \\propto r^{-g}$, where $g\\sim 2$ might be typical) has the property of steepening the decay. This applies in general both to the forward and the reverse shocks. We have suggested ways in which one can attempt to discriminate between the standard straight top-hat injection of energy and momentum with a single $\\gamma_o$ and $E_o$, which then remains constant throughout the afterglow phase, and a refreshed scenario, where the injection is also brief (e.g. comparable to the gamma-ray burst duration and therefore instantaneous compared to the afterglow timescale) but in which there is a varying distribution of $\\gamma$ and of energy during that injection, so that matter ejected with low Lorentz factor cateches up with the bulk of the flow on long timescales. The afterglow energy then increases with time. Under the simple assumption that both the forward and the reverse magnetic fields are equal ($\\epsilon_{B,r}=\\epsilon_{B,f}$) a remarkable prediction is that in all regimes (both shocks are fast cooling, reverse is slow cooling and forward shock is fast cooling or both are slow cooling) the reverse shock spectrum joins seamlessly, or with only a very modest step $\\propto \\gamma^{p-2}$, onto the forward shock spectrum, extending it to lower frequencies. (This could be modified if, for instance, $\\epsilon_{Br} \\ll \\epsilon_{Bf}$, which would give a spectrum with a more pronounce through separating the reverse and forward components). Specifically, in the case of GRB 970508, if a descrete episode of injection that produced refreshed shocks at $t\\sim 1.5$ days is the explanation of the step in the X-ray and optical flux at this time, then one would expect (for equal forward and reverse $\\epsilon_B$) an even more dramatic rise of the 20 GHz and 100GHz flux at 1.5-2 days. This should be about a factor of 5 and 8, respectively, larger than expected from the forward shock values extrapolated back to 2 days. In the case of GRB 990123 a reverse shock appears to have been responsible for the prompt optical flash, and the decay indices (for reasonable spectral slopes) are compatible with monoenergetic impulsive (or straight top-hat) injection. For future GRB afterglow observations, the main prediction from having comparable values of $\\epsilon_B$ in the forward and reverse shocks of baryon loaded fireballs is that the peak flux is found at the peak frequency of the reverse, rather than of the forward shock, i.e. at lower frequencies than typically considered. The IR, mm and radio fluxes would therefore be expected to be significantly larger than for simple (forward shock) standard afterglow models (e.g. Fig 1). This holds whether the injection in contineous or descrete. The two contributions continue to evolve as a pair of smoothly joined components, the ratio of the two peak frequencies $\\nu_{m}^r/\\nu_{m}^f\\propto \\gamma^{-2}$ and peak fluxes $F_{\\nu,mr}/F_{\\nu,mf} \\propto \\gamma$ gradually approaching each other until they coincide at the transition to the non-relativistic case $\\gamma\\sim 1$." }, "0003/astro-ph0003295_arXiv.txt": { "abstract": "The Trifid nebula (M20) is a well-known prominent optical HII region trisected by obscuring dust lanes. Radio continuum VLA observations of this Nebula show free-free emission at $\\lambda$3.6 and 6cm from three stellar sources lying close to the O7V star at the center of the nebula. We argue that neutral material associated with these stars are photoionized externally by the UV radiation from the hot central star. We also report the discovery of a barrel-shaped SNR G7.06--0.12 at the northwest rim of the nebula and two shell-like features G6.67--0.42 and G6.83--0.21 adjacent to W28 and M20. We discuss the nature of these features and their possible relationship to the pulsar PSR 1801-2306 and W28 OH (1720 MHz) masers. ", "introduction": "The Trifid Nebula (M20), one of the most spectacular optical HII regions in the sky, is centered on a small cluster of hot stars which include components A through G of HD 164492. M20 is located at a distance of $\\approx$1.7 kpc in the Sagittarius spiral arm (but see also Kohoutek, Mayer \\& Lorenz 1999); it's angular size of 6$'$ corresponds to about 3pc at this distance. The ionizing flux of $10^{48.8}$ s$^{-1}$ required to maintain the HII region (Chaisson and Wilson 1975) is supplied by the O7.5III star HD 164492A (Wallborn 1973), which has M$_v$=--5.3 for A$_v\\approx$1.3 towards the central stars (Lynds and O'Neil 1985). This nebula is associated with the young star cluster NGC 6514 and a molecular cloud to the southwest (Ogura and Ishida 1975). The recent detection of several molecular condensations associated with protostellar sources in the HII region and HH 399, a remarkable jet-like structure, suggest a new generation of star formation induced by the nebula (Cernicharo et al.\\ 1998). Hester et al.\\ (1999) have recently shown the high resolution images of the SE corner of the nebula based on observations made with WFPC2 of the HST. The emission from this corner of the HII region is dominated by a photoionized photoevaporative flow. M20 shows many similarities to M42 such as its interaction with its parent molecular cloud and its inhomogeneous nebular structure, but the Trifid is thought to be significantly younger than the Orion Nebula ($\\sim 10^{5}$ yr rather than $\\sim 10^{6}$ yr) and the protostellar molecular condensations associated with massive star formation are even younger (10$^4$ yr) (Cernicharo et al.\\ 1998). Here we present near IR and radio observations of M20. We have detected a number of near-IR stellar sources within the central star cluster in J, H, K and L images. All seven of the components of HD 164492 (A--G) appear to varying degrees in these images. Three components -- B, C and D -- coincide with compact radio continuum sources within 12 arcseconds of HD 164492A. These detections suggest that the central massive star is photoionizing the envelope of cool stars in its immediate vicinity similar to the ``proplyds'' in the Orion Nebula (Churchwell et al.\\ 1987; Garay et al. 1987; O'Dell et al. 1993). Radio images of the nebula at 20 and 6 cm show dark features suggesting the presence of cold and dense regions of dust and gas clouds within the HII region shadowing the UV radiation from HD 164492A. Lastly, on a scale of tens of arcminutes, we report the discovery of a new candiate barrel-shaped supernova remnant (SNR) lying adjacent to M20 and two shell-type features to the north and east of SNR W28. ", "conclusions": "\\subsection{Central Stars of M20} Figure 1 shows contours of 3.6cm emission with a resolution of 0.63$''\\times0.45''$ (PA=80$^0$) with rms$\\approx33 \\mu$Jy. Radio continuum peaks whose flux densities and sizes are listed in Table 1 coincide with the positions of HD164492 B, C and D. Optical positions of HD164492 A, B and C are also listed in Table 1 using the Hipparcos coordinates which are based on a fit to a multiple source model. The quality of the model solution for \"C\" using the Hipparcos database was poor, so we did not include it. However, there is an excellent agreement between the absolute position of radio and optical sources to within 0.1$''$, especially for sources B and C. The grayscale images in Figure 2 show the components of multiple star HD 164492 and are identified on the J image, which is very similar to visible images. The central star A coincides with the optical star HD 164992A but is not detected at a level of rms$\\approx33 \\mu$Jy at 3.6cm. Source C, the brightest radio source, has a flux of 2.43 mJy at 3.6cm. HD 164492D lies only 2 arcsec west of star C and was found to be a strong H$\\alpha$ emission source by Herbig, who classified it as a Be-star and named it LkH$\\alpha$ 123 (Herbig 1957). It was also included as H$\\alpha$-emission star number 46 in a survey by Velge (1957) and star number 145 in the survey of Ogura and Ishida (1975). The D star is the brightest point source in recent 12.5 and 17.9 $\\mu$m images that were taken with the JPL MIRLIN camera at the IRTF (Ressler and Shure 1995). Notice several objects in the L image which are either unseen or are very much dimmer shortward of K. Among these newly discovered sources is one roughly 3 arcsec NE and the other only 2 arcsec N of star A. If they are physically associated with this star (2$''$ = 3400 AU at 1.7 kpc), they would represent some of the first low-mass companions to high-mass stars. All three sources B, C and D are also detected at 6cm based on our high resolution BnA array data. Accurate spectral index measurements between 6 and 3.6cm using similar {\\it uv} coverage and spatial resolution of 0.99$''\\times0.7''$ (PA=--87$^0$) showed $\\alpha\\approx$0 for sources B and C and $\\alpha$=--0.19, where F$_\\nu\\propto\\nu^\\alpha$, for source D. The D star as listed in Table 1 is the only star resolved in our 3.6cm measurements with a deconvolved size of 0.19$''\\times0.12''$ (PA=129$^0$). \\subsection{Dense neutral gas associated with stars B,C, and D} What is the origin of the emission from B,C, and D? By analogy with the Orion nebula, one suspects that these sources are externally ionised neutral condensations being photoevaporated by the intense UV field of the O7.5III star HD 164492 A. The argument that this is so proceeds similarly to those of Garay et al.\\ (1987) Churchwell et al.\\ (1987) applied to the proplyds first detected in radio continuum in the Orion nebula. The flat spectra of sources B,C, and D indicate that their radio continuum arises from optically thin free-free emission. For $T\\approx8000$\\,K, the volume emission measures are $n_e^2 V \\sim 1 \\ee 57 \\ut cm -3 $, and for the emission to be optically thin the characteristic scale $R$ of the emission region $\\ga 50 \\u AU $. If this region is a roughly constant-density compact H\\textsc{ii} region, this scale represents the outer boundary. However, the gas is then gravitationally unbound to the central star and the H\\textsc{ii} region would expand on a time scale of $\\sim 30$ yr. On the other hand, if the emitting region is an ionized wind with $n_e \\propto r^{-2}$, the emission is dominated by the innermost radii and $R$ corresponds to the \\emph{inner} boundary. An ionized stellar wind can be discounted, because then $R\\ll 50 \\u AU $, and the mix of optically-thin and optically-thick contributions at any given frequency produces a $\\nu^{0.6}$ spectrum; further the source would be much weaker than observed. Thus we conclude that the emitting region is an ionized wind that is photoevaporated from a reservoir of neutral material near the star. The neutral reservoir cannot be too large as sources B and C are unresolved and source D is barely resolved. Adopting a distance of 1.7 kpc to M20, the geometric mean of the semi-major and semi-minor axes of source D (see Table 1) is $R \\approx 130 AU$, and the corresponding upper limit on sources B and C is 80 AU. The neutral reservoir will be even smaller, and is therefore clearly associated with the star. The O7.5III star HD 164492 A is likely to be the dominant source of the ionizing photons. An O7.5III star emits $9.6\\ee 49 $ ionizing photons $\\ut s -1 $ (Panagia 1973), so that the total ionizing flux incident on a 130 AU radius target at the projected distance of star D ($\\approx 0.095 \\u pc $) is $1.1\\ee 45 \\ut s -1 $. This is comparable to the hydrogen recombination rate of source D, $6.6\\ee 44 \\ut s -1 $. The recombinations in source B and C are also consistent with this hypothesis. Note, however, that stars HD 164492 B,C, and D are of spectral type B (Kohoutek et al 1999) and produce $\\sim 10^{45}$ Lyman continuum photons $\\ut s -1 $ or more, so they may contribute significantly to the ionization if enough photons can intercept the nearby neutral material. Our observations do not determine the distribution of the neutral material associated with stars B,C, and D, but we speculate that it is in circumstellar disks, as for the ``proplyds'' in Orion. \\subsection{Embedded EW Dust Lane in M20} A $\\lambda$6cm grayscale image with resolution of 2.3$''\\times1.1''$ (PA=--80$^0$) and an optical image based on the Palomar Sky Survey are compared in Figures 3a and b. The prominent elongated dust lanes to the SW, SE and NW of the optical image have no counterpart in radio, indicating that these dust features lie in front of the HII region. However, there is a remarkable EW dark radio feature which closely mimics the shape of the optical dust lane seen to the SW of the central hot star in Figure 3b near $\\alpha, \\delta(1950) = 17^h 59^m 17^s, -23^0 02' 50''$. Figure 4 shows a NS slice cut across this feature in the 6cm radio image. The radio continuum emission is depressed by a factor of 3 where the optical nebula appears to be crossed by an EW dark dust lane with a thickness of about 40$''$ (0.3 pc). The correlation between reduced radio emissivity and the optical dust lane is evidence for the dust lane being embedded within the nebula. Other examples of dark features are also apparent in the inhomogeneous large-scale distribution of ionized gas beyond the inner region shown in Figure 3. These dark features are particularly noticeable as broken shell-like structures surrounding the 6$'$ size of the nebula and correlate with the distribution of the HCO$^+$ J=1--0 emission from the nebula presented by Cernicharo et al. (1998). We also note a column of dark feature labelled as dark shadow within the nebula to the east of TC1. These dark features are unlikely to be produced by a lack of short {\\it uv} spacing data but are instead due to dense 10$^4$ cm$^{-3}$ column of gas arising from the surface layer of the molecular cloud and causing the HII region to become ionization bounded. The dearth of emission from a series of dark features including the EW feature, as best represented in Figure 3a, are interpreted to be the peaks of dense gas shielding the ionizing flux arising from the central hot star. These columns of dense gas are responsible to reduce the emission measure $n_e^2 L$ where $n_e$ and L are the electron density and the path length. \\subsection{Ionized Rims of TC1 and TC2} Figure 5 shows total intensity contours over the central part of the HII region at 2.3$''\\times1.1''$ resolution (PA$=-80^0$). The two extended ionized features to the NW and SE are associated with two bright point-like condensations of dust emission at 1.3mm denoted TC1 and TC2 by Cernicharo et al.\\ (1998). High velocity broad wings in the HCO$^+$ emission from TC1 and a jet-like HH feature associated with TC2 led Cernicharo et al.\\ to suggest that these condensations are associated with protostars. The age of these condensations are estimated to be about 10$^4$ yrs and therefore formed after the birth of the HII region. The extended photoionized features in Figure 5 delineate the ionized rims of TC1 and TC2 facing the central hot star. High resolution WFPC2 observations of TC2 using a number of spectral lines was recently reported by Hester et al. (1999) who interpret the ionized layer of TC2 as a photoionizing photevaporative flow. The typical flux density of the ionized rims of TC1 and TC2 is about 0.5 mJy/beam which corresponds to $n_e \\approx2\\times10^3$ cm$^{-3}$." }, "0003/astro-ph0003248_arXiv.txt": { "abstract": "We use Gauss-Hermite functions to study the line of sight velocity distributions in simulated merger remnants. Our sample contains sixteen remnants; eight produced by mergers between disk galaxies of equal mass and eight produced by mergers between disk galaxies with mass ratios of 3:1. The equal-mass mergers display a wide range of kinematic features, including counterrotation at large radii, orthogonally rotating cores, and misaligned rotational axes. Most of the unequal-mass remnants exhibit fairly regular disk-like kinematics, although two have kinematics more typical of the equal-mass remnants. Our results may be compared to observations of early type objects, including ellipticals with misaligned kinematic axes, counterrotating systems, and S0 galaxies. ", "introduction": "Kinematic studies of early-type galaxies have revealed a remarkable variety of interesting behavior; some galaxies have rotation axes ``misaligned'' with respect to their minor axes (Franx, Illingworth, \\& de Zeeuw 1991), while in others the inner regions counterrotate with respect to the rest of the galaxy (Statler, Smecker-Hane, \\& Cecil 1996; Bender \\& Surma 1992; van der Marel \\& Franx 1993). Such intriguing kinematics could plausibly result if these galaxies are the end-products of disk-galaxy mergers (Toomre \\& Toomre 1972), and N-body simulations have gone some ways toward showing that mergers can indeed produce remnants with distinctive kinematics (Hernquist \\& Barnes 1991; Barnes 1992, 1998; Balcells \\& Gonz\\'alez 1998). However, other theories have been put forward for such kinematic features, particularly in the case of counterrotation (Kormendy 1984; Bertola, Buson, \\& Zeilinger 1988). Distinguishing between major mergers and other explanations for distinctive kinematics in galaxies has been especially difficult. The projected luminosity profiles and isophotal shapes of simulated disk galaxy mergers are reasonably good matches to those of elliptical galaxies (eg.~Barnes 1988; Hernquist 1992, 1993; Governato, Reduzzi, \\& Rampazzo 1993; Heyl, Hernquist, \\& Spergel 1994), but few workers have investigated the projected {\\it kinematics\\/} of simulated merger remnants. Hernquist (1992, 1993) described principal-axis profiles of projected mean velocity and velocity dispersion for several disk-disk merger remnants, and Heyl, Hernquist, \\& Spergel (1996) studied line of sight velocity distributions for a somewhat larger sample of objects. These studies showed that kinematic misalignments of merger remnants are observable, and indicated that skewness of line profiles could provide information on the initial orientations of the merging disks. However, while systematically exploring different projections, these studies were limited to equal-mass mergers, and did not examine the structure of line profiles in detail or map velocity fields in two dimensions. Therefore, we studied line of sight velocity distributions for a larger sample of simulated merger remnants. We examined eight mergers between disk galaxies with mass ratios of 1:1 and another eight mergers between disk galaxies with mass ratios of 3:1. We limited our analysis to a single projection along the intermediate axis of each remmant, but we complement a extensive presentation of major-axis kinematics with detailed examinations of individual line profiles and with two-dimensional maps of key kinematic parameters. This work extends the studies described above to unequal-mass mergers, clarifies the connection between initial conditions and line profile, and provides predictions to be compared with kinematic studies of early-type galaxies using the next generation of integral-field spectrometers. The outline of this paper is as follows. The rest of Section~1 describes the merger simulations and the methods we use to extract line of sight velocity distributions and represent the distributions with Gauss-Hermite parameters. Sections~2 and~3 present the results for the equal-mass and unequal-mass mergers, respectively. Section~4 compares our results to observational studies and summarizes our conclusions. \\subsection{Merger simulations} The remnants analyzed here came from a modest survey of parabolic encounters between model disk galaxies (Barnes 1998). Each model had three components: a central bulge with a shallow cusp (Hernquist 1990), an exponential/isothermal disk with constant scale height (Freeman 1970; Spitzer 1942), and a dark halo with a constant-density core (Dehnen 1993; Tremaine et al.~1994). Density profiles for these components are \\begin{eqnarray} \\rho_{\\rm b} &\\propto& r^{-1} (r + a_{\\rm b})^{-3} \\, , \\\\ \\rho_{\\rm d} &\\propto& \\exp(-R/R_{\\rm d}) \\, {\\rm sech}^2(z/z_{\\rm d}) \\, , \\\\ \\rho_{\\rm h} &\\propto& (r + a_{\\rm h})^{-4} \\, , \\end{eqnarray} where $r$ is spherical radius, $R$ is cylindrical radius in the disk plane, and $z$ is distance from the disk plane. Adopting simulation units with $G = 1$, the model used in the equal-mass mergers has a bulge mass of $M_{\\rm b} = 0.0625$, a disk mass of $M_{\\rm d} = 0.1875$, and a halo mass of $M_{\\rm h} = 1$. The bulge scale length is $r_{\\rm b} = 0.0417$, the disk scale radius and scale height are $R_{\\rm d} = 0.0833$ and $z_{\\rm d} = 0.007$, and the halo scale radius is $r_{\\rm h} = 0.1$. With these parameter choices, the model has a half-mass radius $r_{1/2} \\simeq 0.28$, and the circular velocity and orbital period at this radius are $v_{1/2} \\simeq 1.5$ and $t_{1/2} \\simeq 1.2$. The model may be roughly scaled to the Milky Way by equating our units of length, mass, and time to $40 {\\rm\\,kpc}$, $2.2 \\times 10^{11} {\\rm\\,M_\\odot}$, and $2.5 \\times 10^8 {\\rm\\,yr}$, respectively. In the unequal-mass mergers, the larger model had the same parameters as listed above, while the small model was scaled down by a factor of $3$ in mass and $\\sqrt{3}$ in radius in rough accord with the standard luminosity-rotation velocity relation for disk galaxies. Each experiment used a total of $131072$ particles, $65536$ assigned to the luminous components, and $65536$ assigned to the dark halos. The models were run with a tree code using a spatial resolution of $\\epsilon = 0.01$ and a time-step $\\Delta t = 1/128$. With these integration parameters, total energy was conserved to within $0.5$\\% peak-to-peak. All eight equal-mass merger simulations used the same initial orbit, leading in each case to close ($r_{\\rm p} \\simeq 0.2$) parabolic encounter. Disk angles for these experiments are listed in Table~\\ref{eqmass-angles}; $i$ and $\\omega$ are the inclination and argument of pericenter (Toomre \\& Toomre 1972), while the subscripts $1$ and $2$ label the two disks. After merging, remnants were evolved for several more dynamical times before being analyzed. \\begin{table} \\caption{Initial disk angles for equal-mass merger simulations.} \\label{eqmass-angles} \\begin{tabular}{lrrrr} ID & $i_1$ & $\\omega_1$ & $i_2$ & $\\omega_2$ \\\\ \\noalign{\\smallskip} A & 0 & & 71 & 30 \\\\ B & -109 & 90 & 71 & 90 \\\\ C & -109 & -30 & 71 & -30 \\\\ D & -109 & 30 & 180 & \\\\ E & 0 & & 71 & 90 \\\\ F & -109 & -30 & 71 & 30 \\\\ G & -109 & 30 & 71 & -30 \\\\ H & -109 & 90 & 180 & \\\\ \\end{tabular} \\end{table} The eight unequal-mass merger simulations generalize the equal-mass simulations A, B, C, and D by allowing the mass of either galaxy to vary by a factor of three. Like their equal-mass counterparts, these experiments adopted a parabolic initial orbit with pericentric separation $r_p = 0.2$. Table~\\ref{uneqmass-angles} lists the inclinations and pericentric arguments for each simulation; here $i_1$ and $\\omega_1$ are the angles for the larger disk, while $i_2$ and $\\omega_2$ are the angles for its smaller companion. \\begin{table} \\caption{Initial disk angles for unequal-mass merger simulations.} \\label{uneqmass-angles} \\begin{tabular}{lrrrr} ID & $i_1$ & $\\omega_1$ & $i_2$ & $\\omega_2$ \\\\ \\noalign{\\smallskip} A$_1$& 0 & & 71 & 30 \\\\ A$_2$& 71 & 30 & 0 & \\\\ B$_1$& -109 & 90 & 71 & 90 \\\\ B$_2$& 71 & 90 & -109 & 90 \\\\ C$_1$& -109 & -30 & 71 & -30 \\\\ C$_2$& 71 & -30 & -109 & -30 \\\\ D$_1$& 180 & & -109 & 30 \\\\ D$_2$& -109 & 30 & 180 & \\\\ \\end{tabular} \\end{table} Some salient properties of these merger remnants are summarized here; for a more detailed discussion, see Barnes (1998). All sixteen remnants are ellipsoidal objects with luminosity profiles generally following a de~Vaucouleurs law. The projected half-light radii $R_{\\rm e}$ of the equal-mass remnants range from $0.133$ to $0.157$, while for the unequal-mass remnants $R_{\\rm e}$ ranges from $0.099$ to $0.123$. Fig.~\\ref{shapes} shows axial ratios determined from the inertia tensor for the most tightly-bound half of the luminous particles in each object. On the whole, the remnants of equal-mass mergers are triaxial or prolate, while those produced by unequal-mass mergers tend to be more oblate. \\begin{figure} \\begin{center} \\epsfig{figure=shape_eq.ps, width=2in} \\end{center} \\begin{center} \\epsfig{figure=shape_un.ps, width=2in} \\end{center} \\caption{Axial ratios for remnants of (a) equal-mass mergers, and (b) unequal-mass mergers. Solid, dashed, and dotted lines are contours of triaxiality $T = 1$, $\\frac{2}{3}$, and $\\frac{1}{3}$, respectively, where $T \\equiv (1 - b^2)/(1 - c^2)$.} \\label{shapes} \\end{figure} \\subsection{Gauss-Hermite analysis} For all sixteen simulated remnants we extracted five frames, each containing $65536$ luminous particles -- that is, particles from the bulges and disks of the progenitor galaxies. Each set of five frames is equally-spaced over a total of $0.5$ time units; this interval is long enough that individual particles are sampled at effectively random orbital phases, but not long enough for the remnant to undergo significant evolution. We shifted each frame to place the potential minimum at the origin and rotated it to diagonalize the moment of inertia tensor for all particles with a potential less than $0.8$ times the minimum potential. In what follows, we use $X$, $Y$, and $Z$ for the major, intermediate, and minor axes of the remnants. To measure line of sight velocity distributions as a function of position along a given axis, we created a two dimensional grid, with one dimension representing position and the other dimension representing velocity, thus simulating a slit in a spectrometer. Typically, we placed the slit along the major ($X$) axis and projected along the intermediate ($Y$) axis, although other options were used in preliminary investigations. The width of the slit was set at $0.03$, which is roughly $20$\\% of the projected half-light radius. For each frame from each remnant, the particles falling within the slit were binned in position and velocity. The grid spacing along the slit was set to a minimum of $0.02$ and increased as necessary to keep the total number of particles falling within the range above a minimum. The width of the velocity bins was set to a fixed value of $0.2$, spanning the velocity range $|v| \\le 4$ with $40$ bins. To map the line of sight velocity distributions across the plane of the sky, we used a generalization of the above procedure. However, two adjustable bin dimensions were created instead of one along the given slit. After binning the data, the velocity distribution at each location was fit with a parameterized Gauss-Hermite series (van der Marel \\& Franx 1993). The value of each parameter was determined by combining the five frames and performing a least-squares fit; uncertainties were estimated by comparing fits of individual frames. Gauss-Hermite functions are modified Gaussians with additional skewness and kurtosis parameters; they provide an effective way to parameterize the moderately non-Gaussian distributions which arise in systems which have undergone incomplete violent relaxation. The formula for the fitting function is \\begin{eqnarray} P(v) &=& \\gamma \\frac{\\alpha(w)}{\\sigma} [1 + h_3 H_3(w) + h_4 H_4(w)] \\, , \\end{eqnarray} where $w \\equiv (v - v_0) / \\sigma$ and \\begin{eqnarray} \\alpha(w) &\\equiv& \\frac{1}{\\sqrt{2 \\pi}} e^{-w^2/2} \\, , \\\\ H_3(w) &\\equiv& \\frac{1}{\\sqrt{6}}(2 \\sqrt{2} w^3 - 3 \\sqrt{2} w) \\, , \\\\ H_4(w) &\\equiv& \\frac{1}{\\sqrt{24}}(4 w^4 - 12 w^2 + 3) \\, . \\end{eqnarray} This function has five parameters: $\\gamma$, $v_0$, $\\sigma$, $h_3$, and $h_4$. The normalization factor $\\gamma$ has little physical significance in our study. The mean velocity $v_0$ and velocity dispersion $\\sigma$ have dimensions of velocity, while the $h_3$ and $h_4$ parameters represent the skewness and kurtosis of the velocity distribution and are dimensionless. When $h_3 = h_4 = 0$, the Gauss-Hermite series produces a normal Gaussian profile. When $h_3$ has the same sign as $v_0$ the distribution's leading wing is broad and the trailing wing is narrow, while when $h_3$ and $v_0$ have opposite signs the trailing wing is broad and the leading wing is narrow. When $h_4 > 0$, the distribution has a narrow peak with broad wings, and when $h_4 < 0$, the distribution has a broad peak with narrow wings. \\subsection{Orbit classification} In order to discover which orbital families are responsible for various features in the velocity distributions, we assigned each particle to an orbit family using the algorithm described in Fulton \\& Barnes (submitted). This algorithm follows each particle for a number of radial periods and classifies its orbit by examining the sequence of principal plane crossings. To save time and slightly reduce the effects of discreteness, we calculated the trajectories using a quadrupole-order expansion of the gravitational field (White 1983). For the present purpose all ``boxlet'' orbits were counted as boxes; thus the major orbital families recognized here are Z-tubes, which rotate about the minor axis, X-tubes, which rotate about the major axis, and boxes, which do not rotate. ", "conclusions": "During a merger, stellar orbits are scattered by the fluctuating gravitational potential. However, the potential settles down long before orbits can be completely randomized; consequently, merger remnants preserve significant ``memories'' of their progenitors (eg.~Barnes 1998 and references therein). In this study we have shown that such memories can be partly recovered from the line of sight velocity profiles of merger remnants. \\subsection{Comparison with observations} Observations of early-type galaxies reveal a wide variety of kinematic phenomena similar to those seen in our sample of remnants. Very briefly, we will touch on some of these similarities. \\subsubsection{Misaligned rotation} As pointed out in sections 2.1 and 2.2.3 as well as in previous studies, kinematic misalignments are expected in merger remnants, and especially in equal-mass mergers. Franx, Illingworth, and de Zeeuw (1991) present a study of kinematic misalignment in elliptical galaxies; most of the galaxies they observed have small misalignments. While remnant~C (Fig.~\\ref{mapC}) is dramatically misaligned, as a whole the equal-mass remnants described here are better aligned than samples reported in earlier work (Barnes 1992). The incidence of severe misalignment probably depend on several factors; for example, central density profile can have a significant impact on the phase-space available to major-axis tube orbits. Until the factors which favor misalignment are better understood, it's not clear if the observed scarcity of severe kinematic misalignment can constrain the role of equal-mass mergers in the formation of elliptical galaxies. \\subsubsection{Kinematically decoupled cores} Hernquist and Barnes (1991) presented a dissipational simulation showing that the core of a merger remnant could decouple and counterrotate. We have examined the quantitative effect that counterrotation can have on the observed kinematics of merger remnants. Several galaxies, such as NGC 1700 (Statler, Smecker-Hane, \\& Cecil 1996), NGC 4365, NGC 4406, NGC 5322 (Bender \\& Surma 1992), IC 1459, NGC 1374, NGC 4278 (van der Marel \\& Franx 1993), NGC 4816, and IC 4051 (Mehlert et al. 1998) all show line of sight kinematics similar to the line of sight velocity distributions of our models (though some of these galaxies are strong candidates for other scenarios that create counterrotation). In particular, we find amplitudes of $h_3$ and $h_4$ similat to those reported in the observational studies. This shows that major merger can produce remnants with the degree of skewness and kurtosis observed in counterrotating systems. We expect that as more galaxies are observed further examples with line of sight velocity distributions similar to ours will be found. Also worth noting are the observations of NGC 253 by Anantharamaiah \\& Goss (1996) which found an orthogonally rotating core which was suspected to be caused by a merger event. One of our 1:1 merger models also produced an orthogonally rotating core (see Fig.~\\ref{mapG}). \\subsubsection{Counterrotating populations} Early-type galaxies with extended counterrotating populations are rare but not unknown. Some of these systems may have formed by episodic galaxy building (Thakar \\& Ryden 1996), but others are harder to explain in this way. For example, NGC~4550 (Rubin et al.~1992) has counterrotating disks of comparable radial extent and luminosity; Pfenniger (1999) has proposed this galaxy formed by an in-plane merger of two disk galaxies. Our analysis of remnant~H shows that a somewhat wider range of merger scenarios can produce counterrotating populations. \\subsubsection{Rapid rotators} Barnes and Hernquist (1992) and Schweizer and Seitzer (1992), among others, have suggested that S0 galaxies could be made by mergers. Fisher (1997) has collected a sample of S0 galaxies with line of sight velocity distributions fit using Gauss-Hermite parameters. Comparing his observations to our simulations, we find a good match between Fisher's parameters and the parameters for our disky 3:1 mergers. The overall shapes of the Gauss-Hermite parameters plotted on the major axis are remarkably similar, except that our rotation curve near the origin is less steep than the observed S0 galaxies, and some details near the center of our simulations (such as the first twist in the $h_3$ parameters) are not apparent in Fisher's data. \\begin{figure} \\begin{center} \\epsfig{figure=f2201_vs.ps,width=1.6in,bbllx=26,bblly=6,bburx=210,bbury=129} \\epsfig{figure=f2202_vs.ps,width=1.6in,bbllx=26,bblly=6,bburx=210,bbury=129} \\end{center} \\begin{center} \\epsfig{figure=f2204_vs.ps,width=1.6in,bbllx=26,bblly=6,bburx=210,bbury=129} \\epsfig{figure=f2203_vs.ps,width=1.6in,bbllx=26,bblly=6,bburx=210,bbury=129} \\end{center} \\begin{center} \\epsfig{figure=f2206_vs.ps,width=1.6in,bbllx=26,bblly=6,bburx=210,bbury=129} \\epsfig{figure=f2205_vs.ps,width=1.6in,bbllx=26,bblly=6,bburx=210,bbury=129} \\end{center} \\begin{center} \\epsfig{figure=f2207_vs.ps,width=1.6in,bbllx=26,bblly=6,bburx=210,bbury=129} \\epsfig{figure=f2208_vs.ps,width=1.6in,bbllx=26,bblly=6,bburx=210,bbury=129} \\end{center} \\caption{Ratio of $v_0 / \\sigma$ as a function of major-axis position for all 3:1 remnants. Different plotting symbols refer to opposite sides of the major axis.} \\label{vsigma} \\end{figure} Based on measurements of the ratio of mean velocity to velocity dispersion for a set of faint elliptical galaxies, Rix, Carollo, and Freeman (1999) have argued that these galaxies rotate too rapidly to be products of dissipationless mergers. When we compare these measurements to $v/\\sigma$ ratios we measured in our simulations (Fig.~\\ref{vsigma}), we find that unequal-mass mergers can not only produce the same peak $v/\\sigma$ ratios but also produce the same relations between $v/\\sigma$ and radius. Moreover, both our results and the results of Rix, Carollo, and Freeman show similar ranges of maximum $v/\\sigma$ ratios, with values in the range of $\\sim 1$ to $\\sim 4$. We conclude that unequal mass mergers can produce remnants with the dynamics, including the $v/\\sigma$ ratios, characteristic of these faint ellipticals. This result complements a recent study by Naab, Burkert, \\& Hernquist (1999), which finds that unequal-mass mergers can also produce the disky isophotes characteristic of faint ellipticals and S0 galaxies. \\subsection{Summary} We have used Gauss-Hermite expansions to measure the line of sight velocity profiles of simulated merger remnants. Even relatively modest values of $N$ provide enough data to obtain significant detections of non-Gaussian profiles. Some key results are listed below. 1. Equal-mass merger remnants exhibit a variety of kinematic features rather than any single unique ``merger signature''. However, certain features seem common to most of the remnants in our sample; these include slowly rotating inner regions, relatively flat dispersion profiles, off-axis dispersion peaks, and velocity distributions with broad leading and narrow trailing wings. 2. Unequal-mass merger remnants show much less variation in kinematic properties; instead, the larger disk often survives with only moderate damage. Such disk-dominated remnants are characterized by relatively rapid rotation, falling dispersion profiles, and velocity distributions with narrow leading and broad trailing wings. For mass ratios of 3:1, between half and three-fourths of the remnants in our study had strong disk-like kinematics. 3. Simulated remnants have many kinematic characteristics similar to those observed in early-type galaxies. For example, we described counterrotating populations, misaligned rotation, and kinematically decoupled cores resembling those reported in some elliptical galaxies, and rapid rotation consistent with faint ellipticals and S0 galaxies. However, our simulations don't always match observed galaxies. For example, the mean velocity and $h_3$ parameters usually have opposite signs in luminous elliptical galaxies (eg.~Bender, Saglia, \\& Gerhard 1994), while these parameters often have the same sign in our simulated remnants. More work needs to be done to examine the connections between simulated remnants and real galaxies; in particular, the effects of random viewing angles must be taken into account before definitive comparisons of models and observations are possible. We thank Hans-Walter Rix and Andreas Burkert for stimulating discussions, and the referee for a prompt and helpful report. JEB acknowledges partial support from NASA grant NAG 5-8393." }, "0003/physics0003043_arXiv.txt": { "abstract": "The lowest bound states of the hydrogen negative ion and negative donor systems in a homogeneous magnetic field are investigated theoretically via a full configuration interaction approach with an anisotropic Gaussian basis set. The broad magnetic field regime $\\gamma=8\\cdot10^{-4}-4 \\cdot 10^{3}$ is covered. Nonrelativistic total energies, electron detachment energies and transition wavelengths are presented assuming an infinite nuclear mass. The binding mechanisms are discussed in detail. The accuracy for the energies is enhanced significantly compared to previously published data. ", "introduction": "The term ``strong field'' characterizes a situation for which the Lorentz force is of the order of magnitude or greater than the Coulomb binding force. For a hydrogen atom in the ground state the corresponding field strength cannot be reached in the laboratory, but only in astrophysical objects like white dwarfs ( B$\\approx$ $10^2$--$10^5$T) or neutron stars ( B$\\approx$ $10^7$--$10^9$T). Astrophysicists possess therefore a vivid interest in the behavior and properties of matter in strong magnetic fields: theoretically calculated data of magnetized atoms can be used for the determination of the decomposition and magnetic field configuration of astrophysical objects \\cite{Wickramasinghe:1984_1,Ruder:1994:_1,Henry:1984_1,Jordan:1998_1}. On the other hand the strong magnetic field regime is accessible in the laboratory if one considers highly excited Rydberg states of e.g. atoms \\cite{Friedrich:1989_1,Schmelcher:1998_1}. In solid state physics donor states in semiconductors with parabolic conduction bands are systems which possess a Hamiltonian equivalent to the one of hydrogen within an effective mass approximation. Due to screening effects the Coulomb force is much weaker than in the case of hydrogen. The regime where the ground state of the system is dominated by magnetic forces can therefore be reached for certain semiconductors in the laboratory. As an example we mention GaAs for which the effective mass is $m^*=0.067~m_e$ and the static dielectric constant $\\epsilon_s=12.53~\\epsilon_0$. Since the Hamiltonian of the atomic ion and the negative donor are connected through a scaling transformation the values for the energies given in the present work hold for both systems equally. The reader should however keep in mind that they are given in differently scaled units. Apart from the above atoms and molecules in strong magnetic fields are also of interest from a pure theoretical point of view. Due to the competition of the spherically symmetric Coulomb potential and the cylindrically symmetric magnetic field interaction we encounter a nonseparable, nonintegrable problem. Perturbation theory, which is possible in the weak and in the ultrastrong field regime, breaks down in the intermediate field regime. It is therefore necessary to develop new techniques to solve such problems. The neutral hydrogen atom in a strong magnetic field is now understood to a high degree (see \\cite{Friedrich:1989_1,Ruder:1994_1} and references therein). Recently Kravchencko has published an ``exact'' solution which provides an infinite double sum for the eigenvalues \\cite{Kravchenko:1996_1}. With the presented method all energy values of bound states could in principle be calculated to arbitrary precision. For two electron atoms the situation is significantly different. The problems posed by the electron-electron interaction and the non-separability on the one-particle level have to be solved simultaneously, which is much harder. The H$^-$ ion provides an additional challenge since correlation plays an important role for its binding properties. Without a field it possesses only one bound state \\cite{Hill:1977_1}. In the presence of a magnetic field and for the assumption of an infinitely heavy nucleus it could be shown \\cite{Avron:1981_1} that there exists an infinite number of bound states. For laboratory field strengths these states are, due to the binding mechanism via a one dimensional projected polarization potential, very weakly bound \\cite{Bezchastnov:unpub}. Some finite nuclear mass effects can be included via scaling relations\\cite{Ruder:1994_1,Becken:1999_1,Pavlov-Verevkin:1980_1}. However, the influence of the center of mass motion has not been investigated in detail so far. In the present work we assume an infinitely heavy nucleus which represents a good approximation for the slow H$^-$ atomic ion in strong magnetic fields and describes simultaneously the situation of negatively charged donors D$^-$ in the field. Relativistic corrections were neglected since they are assumed to be small compared to the electron detachment energy of the system. We will use in the following the spectroscopic notation $^{2S+1}M$ for the electronic states of the ion where $M$ and $S$ are the total magnetic and spin quantum numbers. Since states with negative z-parity are not considered here we omit the corresponding label in our notation (see also section \\ref{subsec:sym_ham}). Many authors have tackled the quantum mechanical problem of H$^-$ in a strong magnetic field. One of the first, who pursued a variational approach to this problem, were Henry et al.\\cite{Henry:1974_1}. They give first qualitative insights into the weak and intermediate field regime. Mueller et al. \\cite{Mueller:1975} qualitatively described the strong field ground state $^3(-1)$ and the $^10$ state for high fields ($\\gamma\\approx4$ to $\\gamma\\approx20\\,000$, where $\\gamma=1$~a.~u. corresponds to $2.3554\\cdot10^5$T). Larsen has published a number of papers on this problem \\cite{Larsen:1979_1,Larsen:1979_2,Larsen:1981_1}. On the one hand he created very simple and physically motivated trial functions with only a small number of variational parameters. On the other hand his energies were ``state of the art'' in variational calculations for a long time. In \\cite{Larsen:1979_1} he provides binding energies of the lowest $^10$ state in the field regime $\\gamma=0-5$ and of the $^3(-1)$ state in the regime $\\gamma=0-3$. He also presents figures showing the binding energies of the singlet and triplet state for $M=-2$ and $M=-3$. Later \\cite{Larsen:1981_1} he presents total and electron detachment energies for the lowest $^10$, $^3(-1)$ and $^3(-2)$ state in the high field regime. More specifically the regime $\\gamma=20-1\\,000$ for the $^3(-1)$ state and $\\gamma=20-200$ for the other states were investigated. Furthermore Park and Starace \\cite{Park:1984_1} provided upper and lower bounds for energies and binding energies of the ground state $^10$ for weak fields. In the nineties several authors \\cite{Vincke:1989_1,Larsen:1992_1,Larsen:1993_1,Blinowski:1994_1} improved the accuracy of the binding energies and total energies by new techniques. Vincke and Baye \\cite{Vincke:1989_1} report total ionization energies for the lowest singlet and triplet states with $M=0,-1$ and $-2$ for a few field strengths in the regime $\\gamma=4-400$. They are to our knowledge the first who reported that the $^1(-1)$ state becomes bound for sufficiently high field strengths and realized that the $^1(-2)$ state is slightly stronger bound than the corresponding triplet state in the high field regime. Larsen and McCann present in \\cite{Larsen:1992_1} one-particle binding energies for the $^10$ state in the broad magnetic field regime $\\gamma=0- 200$. In \\cite{Larsen:1993_1} the same authors consider furthermore the singlet and triplet states of $M=-1,-2$. The triplet states are calculated for $\\gamma=0.5-200$, the $^1(-1)$ state in the field regime $\\gamma=55-2\\,000$ and the $^1(-2)$ state is calculated for a few field strength in the range $\\gamma=1-100$. Blinowski and Szwacka \\cite{Blinowski:1994_1} have subsequently used a Gaussian basis set, similar to the one used in our calculation. They present results for the $^10$ state, which are less accurate than those of ref. \\cite{Larsen:1992_1}. We also mention some Hartree--Fock calculations: very early Virtamo \\cite{Virtamo:1976_1} has investigated the ground state energies from $\\gamma\\approx20$ to $\\gamma\\approx20\\,000$. Thurner et al.\\cite{Thurner:1993_1} (results published in \\cite{Ruder:1994_1}) have calculated triplet states for $M$=$-1$,$-2$ and $-3$ for many field strength in the broad range $\\gamma=2\\cdot10^{-4}- 2\\cdot10^{3}$. However since they use spherical wave functions for weak fields and cylindrical ones for high fields, there remains a gap of inaccurate results in the intermediate field regime. In the present investigation we provide lower variational energies and higher one-particle binding energies for the atomic H$^-$ problem and respectively the negatively charged donor center D$^-$ problem in a strong magnetic field compared to all other published data sofar. An exception is the field free situation: the calculation by Pekeris \\cite{Pekeris:1962_1} gives $-0.52775$~a.u. for the ground state binding energy whereas we obtain $-0.5275488$~a.u. Clearly the field-free situation is much better understood than the case of a strong field. The paper is organized as follows: in section \\ref{sec:sym_ham_basis} we consider the symmetries of the Hamiltonian and the basis set we use in our calculations. In section \\ref{sec:select} we will report on the strategy we employed for the selection of basis functions in order to obtain accurate results. Section \\ref{sec:res_dis} contains the discussion of our results and a comparison with the literature. ", "conclusions": "\\label{sec:res_dis} As already mentioned the H$^-$ ion possesses only one bound state in the absence of the magnetic field \\cite{Hill:1977_1}. Turning on the field it has been shown \\cite{Avron:1981_1} that there exists (for infinite nuclear mass) for any nonzero field an infinite number of bound states. The corresponding proof \\cite{Avron:1981_1} relies on the physical picture \\cite{Bezchastnov:unpub} that the external electron is for weak fields far from the neutral atomic core and experiences therefore to lowest order a polarization potential due to the induced dipole moment of the core. Perpendicular to the field the motion of the external electron is dominated by the field and it occupies approximately Landau orbitals whereas parallel to the field it is weakly bound due to the projection of the mentioned polarization potential on the Landau orbitals which yields an one-dimensional binding along the field. For typical strong laboratory fields the corresponding binding energies are of the order of $10^{-6}$~eV for the hydrogen atom negative ion and significantly larger for more electron atoms with a larger polarizability. To investigate theses states in the weak field regime goes clearly beyond the feasibility of the present method. Instead we will investigate a number of states, starting from the value of the field strength for which they become significantly bound, which means that the outer electron is already relatively close to the core and possesses a binding energy of at least a few meV. Clearly in that case the picture of the polarization potential is no more valid since exchange and correlation effects rule the binding properties of the ion. Within our approach we could find one bound state for each negative magnetic quantum number of the ion considered ($-3\\leq M\\leq 0$) for both singlet and triplet states, except the $^30$ state, which is unbound. Their behavior has been studied for the complete range of field strengths $0.01\\leq\\gamma\\leq4000$. The one bound state of the H$^-$ ion in the absence of the field represents, in the above sense, an exception since it is already significantly bound without the field. All these states possess positive z-parity and \\emph{no bound states could be found for negative z-parity}. \\subsection{Threshold energies} The electron detachment energy is defined to be the energy we need to remove one electron from the atom without changing the quantum numbers of the total system. The corresponding lowest threshold energy $E_T$ for the H$^-$ ion can be expressed as: \\begin{equation} E_T=\\frac{\\gamma}{2}\\left(|M|+M+2+g_e M_s\\right)-I(\\textnormal{H}) \\end{equation} where $I($H$)$ is the binding energy of the ground state of the neutral hydrogen atom in a magnetic field. The term $\\gamma/2(|M|+M+2)$ is the energy of an electron in the lowest Landau level with magnetic quantum number $m=M$ where the spin part is omitted. This means that the free electron carries the whole angular momentum of the state. For magnetic quantum numbers $M\\leq0$ the threshold energy $E_T$ is independent of the angular momentum $M$, i.e. there is only a singlet and a triplet threshold. The threshold energy is then $E_T=\\gamma-I($H$)$ for singlet states and $E_T=-I($H$)$ for triplet states. We denote the electron detachment energy by $I(H^-)$ which is given by $I($H$^-)=E_T-E_{tot}$ where $E_{tot}$ is the total energy of the considered state of H$^-$. \\subsection{Total, electron detachment and transition energies} Before we discuss the individual states and their properties let us describe some general features of the states considered here. The total energy of the singlet states is monotonically increasing with increasing field strength. This fact is caused by the increase of the field-dependent kinetic energy. In contrast to this the total energy of the triplet states is monotonically decreasing with increasing field strengths. This is a consequence of the additional spin Zeeman term (we consider here only the $S_z=-1$ component of the spin triplet states). {\\em The electron detachment energies are monotonically increasing with increasing field strength for all states considered here, i.e. both singlet and triplet states.\\/} This has to be seen in view of the above-mentioned fact that the zero-point kinetic (Landau) energy of the electrons is raised in the presence of the magnetic field and therefore the threshold energy for loosing one electron is raised in the same way. For the $^10$ state the total energy raises from $-0.52754875$ at $\\gamma=0$ to $3986.49870$ at $\\gamma=4000$. This state is the most tightly bound state for all field strengths. The detachment energy increases from $0.027549$ a.u. at $\\gamma=0$ to $2.29805$ a.u. at $\\gamma=4000$. There are two reason which give rise to the fact, that this state is the most tightly bound one. On the one hand the electrons are in this state much closer to the nucleus than in other states. This increases the binding due to the attractive nuclear potential energy. On the other hand correlation has an important impact on the binding energy. Both effects are reinforced with increasing field strength as the electrons become more an more confined in the x-y plane perpendicular to the magnetic field. These effects overcome the influence of the static electron-electron repulsion. The total energies and the detachment energies of the $^10$ state are presented in table \\ref{tab:0Sbindcomp}. It can be seen that the detachment energies for this most tightly bound state could be improved by 1-2\\% for all field strengths compared to the existing literature. This is not correct for a vanishing field, where much more efficient basis sets like the Hylleraas basis set are available. For numerical reasons the relative accuracy for the detachment energies is largest in the intermediate field regime. The $^30$ state is not bound for all considered field strengths. This can be understood in an effective particle picture as follows: for triplet states the spatial two-particle wave function is antisymmetric with respect to particle exchange and therefore the two particles have to occupy different spatial orbitals, i.e. we are exclusively dealing with excited configurations. For $M\\not=0$ it is (see later) possible to obtained tightly bound triplet states in a strong magnetic field by occupying different orbitals of the hydrogenic series ($1s, 2p_{-1}, 3d_{-2}, \\ldots$) which yields the one-particle excited configurations of the type $1s2p_{-1}, 1s3d_{-2}, \\ldots$. For the case of the $^30$ state, however, we have $M=0$ and only configurations constructed from pairs of two orbitals with $(m,-m)$ are allowed which are either of doubly excited character ($m\\not=0$) or a singly excited configuration with $m=0$. Therefore no magnetically tightly bound configurations are allowed for the $^30$ state which illuminates its unbound character for any field strength. All singlet and triplet electron detachment energies of all the considered bound states are presented also graphically: Figure \\ref{fig:sinbind} shows the singlet detachment energies and figure \\ref{fig:tripbind} the corresponding energies for the triplet states. It is important to mention that the global ground state of the ion undergoes a crossover with respect to its symmetry with increasing field strength. For weak fields the $^10$ state is the ground state of the system, whereas in strong fields the $^3(-1)$ state becomes the ground state which was first shown in ref.\\cite{Henry:1974_1}. This is caused by the spin Zeeman term, which lowers the total energy of the triplet states. The crossover takes place at $\\gamma_c\\approx0.05$ which corresponds to approximately $10^4$~T for the H$^-$ ion. The $^3(-1)$ state is very weakly bound when it becomes the ground state (at $\\gamma_c$ the detachment energy is $\\approx 3\\cdot10^{-4}$ a.u.). This prevents us from localizing more exactly the field strength at which the crossover takes place. The $^3(-1)$ state, being the ground state of the anion for $\\gamma>\\gamma_c$ never becomes the most tightly bound state. At $\\gamma=4\\,000$ its electron detachment energy is $1.25$ a.u. and therefore much less than the detachment energy of the $^10$ state. This is due to the fact that the tightly bound states are formed by occupying the hydrogenic series $1s, 2p_{-1}, 3d_{-2}, \\ldots$ (as mentioned above) and the $^10$ states allows for the $1s^2$ configuration yielding the strongest binding although it represents an excited state for $\\gamma > \\gamma_c$ due to its spin character. The singlet state $^1(-1)$ is not bound for weak fields. It becomes bound in the regime $\\gamma\\approx1-5$ which is an unexpected behavior. The $^1(-1)$ state lies higher in the spectrum than the bound $^1(-2)$ and $^1(-3)$ states for the intermediate field region. In the high field region it however crosses both states. The crossing with the $^1(-3)$ takes place at $\\gamma\\approx300$, the crossing with the $^1(-2)$ state is at $\\gamma\\gtrsim4\\,000$. Unfortunately the accuracy of our method is not sufficient to provide a closer look at this crossing. The fact that the $^1(-1)$ state is not bound for weak fields but bound for strong fields is a consequence of the complicated interplay of the different interactions. The Coulomb repulsion of the two electrons is much weaker for the spatially antisymmetric triplet states compared to the singlet states. The electron-electron repulsion is higher for the states with $M=-1$ compared to the states with $M<-1$. This pushes the $|M|=1$ singlet states for weak fields beyond the threshold energy, i.e. makes them unbound. The total ionization and the detachment energies of the singlet and triplet states with $M=-1$ are presented in table \\ref{tab:singtrip1}. The suppression of the binding for the singlet state can clearly be seen from this table: the detachment energy of the singlet is $100$ times lower than for the triplet at $\\gamma=10$, but at $\\gamma=4000$ the ratio is of the order $2$. The comparison with the literature (see table \\ref{tab:singtrip1}) shows that our detachment energies are variationally lower by several percent than the best available data. For the situation of weakly bound states the improvement is significantly larger. Let us now consider the energies for the states with $M=-2$ which are presented in table \\ref{tab:singtrip2}. Focusing on the detachment energies we realize that for weak fields the triplet state possesses a larger detachment energy than the singlet state, but for intermediate and high fields the singlet state is stronger bound than the triplet one, i.e. we encounter a crossover which is presented in figure \\ref{fig:bindcross2s2t}. Compared to the data of ref.\\cite{Larsen:1993_1}, our method yields $5-10$\\% higher variational detachment energies for the triplet state and several times higher detachment energies for the singlet one. If we consider the singlet-triplet splitting which is the difference of the total energies between the singlet and the triplet state, where the spin-Zeeman shift is omitted, it can be observed, that for all states this splitting behaves monotonically increasing with increasing field strength in the weak field regime. The splitting for the states with $M=-2$ and $M=-3$ are shown in figure \\ref{fig:singtripsplitee}. The splitting for the $M=-2$ states increases in weak fields, but for high fields this splitting decreases and becomes negative above some critial field strength. It seems that the Coulomb repulsion, due to antisymmetrization of the wave function is dominated by correlation effects. That the above observation is in fact a consequence of correlation is supported by Vincke and Baye \\cite{Vincke:1989_1}: the reversed order concerning the detachment energies (see figure \\ref{fig:bindcross2s2t}) occurs if they include so-called transverse mixing, which simulates correlations in their approach. For states with $M=-3$ only a few published data are available. These states are only weakly bound, although they are stronger bound for $\\gamma\\gtrsim300$ than the $^1(-1)$ state. The singlet state has for $\\gamma=0.2$ a detachment energy of $7.1~10^{-5}$~a.u.\\ and at $\\gamma=1000$ its detachment energy is $0.19$. The electron detachment energies of the triplet state are of the same order of magnitude and the absolute value of the singlet triplet splitting is the lowest of the states considered here. As a consequence a careful convergence study of the results (detachment energy) is indispensable. Our data are given in table \\ref{tab:singtrip3}. The wavelengths of the transitions of the singlet states are presented in figure \\ref{fig:singtrans}. The wavelengths are monotonically decreasing with incresing field strength except for the transition from the $^1(-1)$ state to the $^1(-2)$ state. As mentioned above these states cross at $\\gamma\\gtrsim4\\,000$. Therefore the corresponding wavelength for this transition diverges at the crossing field strength. The transition wavelengths for the triplet states shown in figure \\ref{fig:triptrans} are also monotonically decreasing with increasing field strength. Finally we comment on corrections due to the finite nuclear mass. There are two kinds of corrections, which are relevant here. One, which is special for ions in strong magnetic fields and which describes the coupling between the center of mass motion and the electronic motion. This coupling is due to a motional electric field of intrinsic dynamical origin seen by the moving ion in a magnetic field \\cite{Schmelcher:1991}. Second there are corrections due to the replacement of the naked masses by reduced ones which can be easily included in our data by performing the corresponding shifts \\cite{Becken:1999_1,Becken:1999_2,Pavlov-Verevkin:1980_1} . A full dynamical treatment of the atomic ion including the collective motion goes clearly beyond the scope of the present investigation. It is important to note that for the case of the fixed negative donors there naturally occur no such corrections." }, "0003/astro-ph0003138_arXiv.txt": { "abstract": "We present wide-band radio observations spanning from 1.4 GHz to 350 GHz of the afterglow of \\grb, taken from 1 to 80 days after the burst. The optical and X-ray afterglow of this burst were fairly typical and are explained by a jet fireball. In contrast, the radio afterglow is unusual in two respects: (a) the radio light curve does not show the usual rise to maximum flux on timescales of weeks and instead appears to be declining already on day 1 and (b) the power law indices show significant steepening from the radio through the X-ray bands. We show that the standard fireball model, in which the afterglow is from a forward shock, is unable to account for (b) and we conclude that the bulk of the radio emission must arise from a different source. We consider two models, neither of which can be ruled out with the existing data. In the first (conventional) model, the early radio emission is attributed to emission from the reverse shock as in the case of GRB 990123. We predict that the prompt optical emission would have been as bright (or brighter) than 8th magnitude. In the second (exotic) model, the radio emission originates from the forward shock of an isotropically energetic fireball ($10^{54}$ erg) expanding into a tenuous medium (10$^{-4}$ cm$^{-3}$). The resulting fireball would remain relativistic for months and is potentially resolvable with VLBI techniques. Finally, we note that the near-IR bump of the afterglow is similar to that seen in GRB 971214 and no fireball model can explain this bump. ", "introduction": "} The radio afterglow from \\grb\\ is unusual on two counts. First, the radio afterglow in the centimeter band does not show the usual rise to a peak value $f_m$ (at epoch $t_m$) before undergoing a power law decay. The radio flux appears to decline continuously starting from the epoch of the first observation. Thus $t_m<1.49$ d as compared to the 10--100 d seen in other bursts (e.g. \\cite{fwk00}, \\cite{fkb+99}, \\cite{fks+99}). Second, the temporal decay indices ($\\alpha_\\nu$) in the radio, optical and X-ray bands are markedly different from each other. Proceeding from radio to higher frequencies, $\\alpha_{\\nu}$ steepens by $\\sim$0.4 every four decades in frequency. In contrast, the optical and X-ray afterglow appears to find a straightforward explanation in the standard afterglow model in which a jet geometry is invoked (\\cite{hum+00}). Below we show that the radio observations cannot be reconciled with a standard jet (or sphere) afterglow model. We then explore possible modifications to the standard model. The simplest afterglow model is one in which the broad-band afterglow emission arises from the forward shock of a relativistic blast wave propagating into a constant density medium (\\cite{spn98}). It is assumed that the electrons in the forward shock region are accelerated to a power law distribution for $\\gamma_e>\\gamma_m$, $dN/d\\gamma_e\\propto \\gamma_e^{-p}$; here $\\gamma_e$ is the Lorentz factor of the electrons, $p$ is the power law index and $\\gamma_m$ is the minimum Lorentz factor. Gyration of these electrons in strong post-shocked magnetic fields gives rise to broad-band afterglow emission. Two modifications to this picture are routinely considered. (1) An inhomogeneous circumburst medium (specifically, $\\rho(r)\\propto r^{-2}$; here $\\rho$ is the density at distance $r$ from the source). Such a circumburst medium is expected should GRBs originate from massive stars (\\cite{cl99b}). (2) A jet-like geometry for the blast wave (\\cite{sph99}). This modification is motivated by the propensity of jets in astrophysical sources. Regardless of these modifications, the broad band spectrum is composed of three characteristic frequencies: $\\nu_a$, the synchrotron self-absorption frequency; $\\nu_m$, the frequency at which the emission peaks (and attributed to the electrons with Lorentz factor $\\gamma_m$), and $\\nu_c$, the cooling frequency. Electrons radiating photons with frequency $>\\nu_c$ cool on timescales faster than the age of the blast wave. The evolution of these frequencies is determined by the dynamics of the blast wave. The usual ordering of these frequencies at epochs relevant to the discussion here is (going from low to high frequencies) $\\nu_a$, $\\nu_m$ and $\\nu_c$. For \\grb\\ the early radio decay implies that $\\nu_m$ is already below the centimeter radio band at 1.49 days. The steepening of the afterglow emission from optical to X-ray can be explained by placing $\\nu_c$ between the optical and X-ray bands. The expected steeping $\\Delta\\alpha$ is 1/4 which is marginally consistent with $\\alpha_o-\\alpha_x=0.54^{+0.18}_{-0.10}$. However, even if we ignore this, we are simply unable to explain the decay in the radio band, since no additional steepening is expected between $\\nu_m$ and $\\nu_c$. The standard afterglow model can be made to agree with the light curves by postulating an energy slope $p$ which gradually steepens with increasing electron energy $\\gamma_e$. We use the spherical, constant density afterglow model (\\cite{spn98}) to convert, in each band, the observed decay index to $p$ and obtain: $p=2.09\\pm0.03$ (radio), $p=2.43^{+0.23}_{-0.11}$ (optical), and $p=2.81\\pm0.08$ (if $\\nu_c$ is below the X-ray band) or $p=3.15$ (if $\\nu_c$ above the X-ray band). We are justified in applying the spherical model for early times ($tt_J$ (\\cite{hum+00}). Curvature is both observed and modeled in the synchrotron spectra of the non-relativistic shocks from supernova remnants which are accelerating cosmic rays (e.g. Baring et al.~1999\\nocite{ber+99}). To date, models of ultra-relativistic shocks favor a universal value of $p$, independent of energy (\\cite{vie00}, \\cite{gakg00}), but non-linear effects have yet to be treated. Nonetheless, the invocation of curvature in the energy distribution of the electrons cannot explain the observed broad-band spectrum (Figure~\\ref{fig:broadband}) of the afterglow on December 18 (corresponding to 1.33 days after the burst). A plausible fit to the entire data is obtained with $\\nu_a=1.3$~ GHz, $\\nu_m=270$~GHz and $\\nu_c=7\\times 10^{16}$~Hz and $f_m=3.4$ mJy; this fit is displayed by the dashed line in Figure~\\ref{fig:broadband}. As the blast wave slows down, $\\nu_m$ moves to lower values while preserving $f_m$ and thus we expect the flux in the centimeter band to rise, whereas the observed flux falls. If the afterglow has a jet-like geometry then the radio afterglow is expected to rise until the epoch $t_J$, and subsequently decay very slowly ($f_\\nu\\propto t^{-1/3}$) until $\\nu_m$ passes through the centimeter band, after which we expect to see a decline similar to that seen in the optical ($f_\\nu \\propto t^{-2.2}$) (\\cite{hbf+99}). As can be seen from Figure~\\ref{fig:broadband}, the radio observations are grossly inconsistent with these expectations, particularly the decay is much faster than $t^{-1/3}$. To summarize, while the optical and X-ray observations can be accounted for by a jet model, the radio observations are inconsistent with the standard model. This forces us to consider afterglow models in which the radio emission (at least in bulk) arises from a source other than the usual forward shock. ", "conclusions": "" }, "0003/astro-ph0003374_arXiv.txt": { "abstract": " ", "introduction": "Dwarf novae are cataclysmic binary stars which exhibit repetitive outbursts of several magnitudes. They contain a Roche-lobe-filling cool dwarf star that loses mass through the inner Lagrangian point, and a white-dwarf star accreting it (Warner 1995). The SU\\,UMa stars form a sub-class of dwarf novae, showing two types of outburst, namely, a short ``normal'' outburst and a long ``superoutburst''. According to theories for the superoutburst mechanism (e.g.\\,Osaki 1996), after the accretion disk grows over a critical radius it becomes tidally unstable due to a gravitational interaction with the secondary. In this model the precession of an eccentric disk can explain the ``superhump'' modulation present in the superoutburst. Eclipsing systems provide a unique opportunity to reconstruct the brightness distribution of an accretion disk from the observed integrated light (Horne 1985; Baptista, Steiner 1991, 1993). There are only five known SU\\,UMa stars which exhibit deep eclipses, indicating occultation of the accretion disk and the white dwarf by the secondary star. Of these systems, HT\\,Cas (Zhang et al.\\,1986), OY\\,Car (Krzeminski, Vogt 1985), and Z\\,Cha (Bailey 1979) have long been studied; these limited samples have historically provided almost all of our knowledge concerning the spatial structure and time-evolution of accretion disks in SU\\,UMa stars. Although two more eclipsing SU UMa stars, DV UMa (Nogami et al., in preparation) and V2051 Oph (Kiyota, Kato 1998), have recently been discussed, the low frequency of superoutbursts and the small number of the known eclipsing systems still make it difficult to directly clarify the eccentric disk, itself, and its evolution with time by an observational approach. In this letter we report on the discovery of a new deeply eclipsing northern SU\\,UMa-type dwarf nova, IY\\,UMa (= TmzV85), along with the results of our photometric monitoring and time-resolved photometry. A more detailed analysis of the eclipses, including the time-evolution of the accretion disk during this superoutburst and the subsequent rapid fading phase, will be presented in a separate paper. ", "conclusions": "We have derived some of the physical parameters of the newly discovered SU\\,UMa-type dwarf nova with a deep eclipse, IY\\,UMa, and summarize them in table 2 along with those of the other eclipsing SU\\,UMa stars. As shown in this table, the orbital period of IY\\,UMa is quite similar to those of Z\\,Cha and HT\\,Cas. Because the normal outburst of IY\\,UMa has historically not been detected, continuous observations are essential to determine the frequency of a normal outburst and the part of accretion disk where the disk-instability begins. IY\\,UMa is potentially the most valuable northern SU\\,UMa-type dwarf nova with deep eclipses, the bright quiescence magnitude, and relatively frequent superoutbursts compared to HT Cas, which has not been observed to undergo a superoutburst since 1985. The star thus provides a unique opportunity for SUBARU telescope to study the structure of accretion disks of SU\\,UMa-type dwarf novae in quiescence through eclipse-timing spectroscopic observations.\\par \\vspace{1pt}\\par We are pleased to acknowledge comments by D. Nogami, which lead to several improvements in this paper. This research has been supported in part by a Grant-in-Aid for Scientific Research (10740095) of the Japanese Ministry of Education, Science, Sports, and Culture. KM has been financially supported as a Research Fellow for Young Scientists by the Japan Society for the Promotion of Science. PS's observations were made with the Iowa Robotic Observatory, and he wishes to thank Robert Mutel and his students. \\onecolumn \\begin{figure} \\centerline{ \\epsfbox{fig3.ps}} \\end{figure} \\begin{fv}{3} {0pc} {Eclipses and superhumps. Upper panel: the light curve on HJD 2451561 (intermediate phase in superoutburst). Lower panel: on HJD 2451567 (late phase in superoutburst). The eclipses become deeper with time.} \\end{fv} \\begin{table}[t] \\begin{center} Table~2. \\hspace{4pt}Physical parameters of the known eclipsing SU UMa stars.\\\\ \\bigskip \\begin{tabular}{cccccc} \\hline \\hline Object Name & $V_{\\rm quies}$ & $V_{\\rm super}$ & $T_{\\rm super}$ & $P_{\\rm orb}$ & $P_{\\rm sh}$\\\\ \\hline Z Cha & 15.3 & 11.9 & 287 & 0.074499 & 0.07740 \\\\ OY Car & 15.3 & 11.4 & 318 & 0.063121 & 0.064544 \\\\ V2051 Oph & 15.0 & 11.7$^*$ & 430$^*$ & 0.062428 & 0.06423$^\\dag$ \\\\ HT Cas & 16.4 & 11.9$^\\ddag$ & & 0.073647 & 0.076077 \\\\ DV UMa & 18.6 & 14.0$^*$ & 970$^*$ & 0.08587$^\\S$ & 0.08867$^\\S$ \\\\ IY UMa & 18.4$^*$ & 13.0 & 400 or 800 & 0.073913 & 0.07588 \\\\ \\hline \\end{tabular} \\end{center} \\begin{center} \\footnotesize \\vspace{6pt}\\par\\noindent ${\\rm V_{\\rm quies}}$: magnitude at quiescence, ${\\rm V_{\\rm super}}$: maximum magnitude during superoutburst, \\\\ ${\\rm T_{\\rm super}}$: supercycle, ${\\rm P_{\\rm orb}}$: orbital period, and ${\\rm P_{\\rm sh}}$: superhump period. \\\\ Data without symbols from Ritter and Kolb (1998)\\\\ $*$ from VSNET data, $\\dag$ IBVS 4644, $\\ddag$ IAU Circ. 4027, $\\S$ Uemura et al. 2000 in preparation \\end{center} \\end{table} \\twocolumn" }, "0003/astro-ph0003142_arXiv.txt": { "abstract": "We present ISOPHOT\\_S, ISOSWS and 8 to 13 micron ground based observations of Centaurus A that show prominent PAH and silicate features. These and other data are used to construct a model for the infrared continuum. We find that in a nuclear sized aperture (\\verb!~!4 arcsec,\\verb!~!60 pc) the SED is characteristic of emission from a starburst and AGN torus; in larger apertures an additional component of cirrus emission is required. Based on our model, the torus diameter is estimated to be 3.6 pc and the best fitting inclination angle of the torus is 45 degrees. This result has implications for the detectability of tori in low power AGN and in particular for the use of the IRAS 60/25 micron flux ratio as an indicator of the torus inclination. \\vspace {5pt} \\\\ Key~words: active-galaxies; individual; Cen A; nuclei-galaxies; galaxies-radiative transfer. ", "introduction": "Rowan-Robinson and Crawford (1989) attribute the infrared (IR) spectra of galaxies to a mixture of up to 3 components: i) general disc emission from grains heated by the interstellar radiation field (cirrus), ii) a Seyfert component peaking in the mid-infrared and iii) a starburst component peaking at about 60 microns. Essentially all of these sources arise from the thermal reprocessing of ultra-violet and other high frequency photons by the dust within these objects. The dust grains re-radiate the absorbed energy in the IR, with the resultant spectrum dependent on the distribution of dust grain temperatures. For the starburst and Seyfert components, the clouds of dust are optically thick, even to IR photons, so radiative transfer effects are important. Centaurus A, the famous southern radio (FRI) counterpart to NGC5128, identified by Bolton, Stanley and Slee (1949), at a distance of approximately 3.1 Mpc (Tonry and Schechter, 1990) is the closest active galaxy to us. It is a multi-faceted object, showing evidence of a merger (see Mirabel et al, these proceedings), starburst and AGN activity, with HII regions, shells, jets, optical filaments and a warped dust lane. Evidence of star formation is most apparent in the dust lane that intersects the host galaxy NGC5128. Marston and Dickens (1988) found that their 12 micron IRAS DSD observations followed the H$\\alpha$ emission, along the dust lane, and hence the regions of star formation. They modelled their 12, 25, 60 and 100 micron observations as a cirrus spectrum of small and large grains heated by the interstellar radiation with two grain temperatures - hot (240K) small grains and cooler (30K) large grains. Evidence of starburst activity comes from enhanced far-IR and sub-mm nuclear emission that shows a structure offset from the dust lane (e.g.\\ Hawarden et al, 1993) and from the high star formation rate, which is typical of a starburst galaxy (Eckart et al, 1990). Evidence of AGN activity comes from the radio jets/lobes (e.g.\\ Clarke, Burns and Norman, 1992) and variable x-ray emission (e.g.\\ Morini, Anselmo and Molteni, 1989) whilst evidence of a Seyfert-like dusty torus comes from the high optical depth to the nucleus (e.g.\\ Blanco, Ward and Wright, 1990). ", "conclusions": "The complete model includes contributions from torus, starburst and cirrus in all apertures. The model fit in the nuclear aperture is quantitatively similar to that found by Laurent et al (these proceedings), with ISOCAM CVF observations and AGN and starburst observational templates. The opening half-angle of the torus cone is very similar to that of the 'cones' in the recently published HST nuclear Pa$\\alpha$ image (Schreier et al, 1998), and if it is assumed that these observations are of ionisation cones, rather that the suggested warped accretion disc, the cone opening half-angle is constrained between 30 and 40 degrees. Extinction to the 2.2 micron region can be estimated by comparing the flux for the face-on torus to the 45 degree inclined torus and is 2.6 mags, equating to a visual extinction of 23 mags. This degree of extinction is consistent with the analysis of Meadows and Allen (1992) and suggests that the near-IR emission region is considerably less extincted than the x-ray region (70 mags, Blanco, Ward and Wright, 1990). Based on this model, the torus diameter is calculated to be 3.6 pc, assuming a distance of 3.1 Mpc, placing it easily within the nuclear aperture. This is significantly smaller than that actually observed in other galaxies (e.g.\\ \\verb!~!1 kpc for MG0414+0534 (Oya et al, these proceedings),\\verb!~!200pc for NGC1068 (Young et al, 1996), \\verb!~!100 pc for NGC4261 (Jaffe et al, 1993) and \\verb!~!50 pc for NGC4151 (Mundel et al, 1995)). However, although of a small diameter, this model does not support the even smaller diameter tori of the Pier and Krolik (1992) model. Possible observational support for the small torus size estimate comes from the 2.3 GHz radio observations of the nuclear sub-arcsecond scale jet and counter jet (Jones et al, 1996). These images show the core to be completely absorbed between the jet and counter jet, attributed to a gaseous disc or torus of 0.4 to 0.8 pc. The model torus, with an inclination of 45 degrees would present a 1.4 pc absorption band; very similar to the size of the radio core absorption. Within the context of unified theories the inclination of the torus is a crucial parameter in determining whether an AGN is a Type 1 or Type 2. Various studies have suggested that the IRAS 60/25 micron flux ratio is an indicator of torus inclination with high or low ratios implying edge-on to face-on torus inclinations. The combination of all the model components gives a 60/25 micron flux ratio of 9.4, although the actual torus 60/25 micron flux ratio is only 0.7. The torus inclination implied from this study is identical to that found for NGC1068 (Efstathiou, Hough and Young, 1995) and yet the IRAS 60/25 micron flux ratio for NGC1068 is only 2.1. If the torus was as powerful as that in NGC1068, or a smaller 'nuclear' sized aperture was used, the 60/25 micron flux ratio would be more indicative of the torus inclination but within the large IRAS sized aperture, the torus is dominated by other more powerful IR emitting components. A comparison of the mean IRAS 60/25 micron flux density ratios of the extended 12 micron sample of galaxies, which is considered statistically complete on 12 micron flux, gives a mean of 4.1$\\pm$2.9 for the Seyfert 1s and 4.5$\\pm$3.0 for the Seyfert 2s (Andy Thean, private communication). This suggests that this colour ratio cannot distinguish between Type 1 and Type 2 AGN and so cannot be a reliable indicator of the torus inclination. The IRAS 60/25 micron flux density ratio is more likely to provide a ratio of the AGN to star formation components. Evidence for this comes from the 60 micron to 6 cm flux correlation (Wilson, 1988) which works only if the radio jet/lobe AGN emission is removed (Peter Bartel, private communication), thereby strongly suggesting that the 60 micron flux is coming from star formation and starburst regions. Additional evidence comes from the ISOLWS observations of NGC1068 (Spinoglio et al, these proceedings) where the far-IR continuum is almost indistinguishable to that of the archetypal starburst galaxy M82." }, "0003/astro-ph0003468_arXiv.txt": { "abstract": "We apply the dynamical modeling approach of Statler (\\markcite{Statler1994b}1994b) to 13 elliptical galaxies from the Davies and Birkinshaw (\\markcite{DB1988}1988) sample of radio galaxies to derive constraints on their intrinsic shapes and orientations. We develop an iterative Bayesian algorithm to combine these results to estimate the parent shape distribution from which the sample was drawn, under the assumption that this parent distribution has no preferred orientation. In the process we obtain improved estimates for the shapes of individual objects. The parent shape distribution shows a tendency toward bimodality, with peaks at the oblate and prolate limits. Under minimal assumptions about the galaxies' internal dynamics, 35\\% of the objects would be strongly triaxial ($0.2 < T < 0.8$). However, the parent distribution is sensitive to the assumed orbit populations in the galaxies. Dynamical configurations in which all galaxies rotate purely about either their long or short axes can be ruled out because they would require the sample to have a strong orientation bias. Configurations in which the mean motion about the short or long axis is either ``disklike''---dropping off away from the symmetry planes---or ``spheroidlike''---remaining roughly constant at a given radius---are equally viable. Spheroidlike rotation in the long-axis or short-axis tube orbits significantly lowers the abundance of prolate or oblate galaxies, respectively. If rotation in ellipticals is generally disklike, then triaxiality is rare; if spheroidlike, triaxiality is common. ", "introduction": "Over the years a number of attempts have been made to derive the intrinsic shape distribution of elliptical galaxies from observations (Hubble \\markcite{Hubble1926}1926, Sandage et al.\\ \\markcite{Sandage1970}1970, Noerdlinger \\markcite{Noerdlinger1979}1979, Marchant \\& Olson \\markcite{Marchant1979}1979, Richstone \\markcite{Richstone1979}1979, Binggeli \\markcite{Binggeli1980}1980, Binney \\& de Vaucouleurs \\markcite{Binney1981}1981, Olson \\& de Vaucouleurs \\markcite{Olson1981}1981; for a review see Statler \\markcite{Statler1996}1996). Generally the results of these efforts have been ambiguous, and interest in the problem waned somewhat in the 1980s. But more recent developments have sparked renewed attempts to crack this classic chestnut. Among these developments are the recognition that halo shapes may serve as a diagnostic of galaxy formation physics (Dubinski \\& Carlberg \\markcite{Dubinski1991}1991, Weil \\& Hernquist \\markcite{Weil1996}1996), and indications that Hamiltonian chaos, dissipation, or both may either force triaxial equilibrium configurations to evolve slowly toward axisymmetry or render them altogether impossible (Dubinski \\markcite{Dubinski1994}1994, Merritt \\& Fridman \\markcite{Merritt1996}1996, Merritt \\& Quinlan \\markcite{Merritt1998}1998). Studies of central surface brightness profiles using {\\sl HST\\/} suggest that the fundamental properties of elliptical galaxies may be bimodally distributed. There appears to be a dichotomy between high-luminosity, slowly rotating systems with shallow central cusps and boxy isophotes, and lower luminosity, rotationally supported systems with steeper cusps and a tendency for diskiness (Lauer et al.\\ \\markcite{Lauer1995}1995; see also Kormendy \\& Bender \\markcite{Kormendy96}1996). Since rapid rotation and triaxiality are generally regarded as being incompatible, one might anticipate a bimodal distribution of triaxialities. Tremblay \\& Merritt \\markcite{Tremblay1996}(1996) find that low- and high-luminosity elliptical galaxies have different distributions of apparent ellipticity, which would imply different distributions of true shapes. Merritt \\& Tremblay's work joins that of Fasano \\& Vio \\markcite{Fasano1991}(1991), Ryden (\\markcite{Ryden1992}1992, \\markcite{Ryden1996}1996), and Fasano \\markcite{Fasano1996}(1996) as successors to the classical photometric approaches pioneered by Hubble, Sandage, and others. However, photometric methods, while effective in constraining the distribution of overall flattenings, reveal little about the frequency of axisymmetry {\\em vs.\\/} triaxiality in the population. Uncovering this information requires the use of kinematic data and dynamical models to connect the kinematics to the shape of the gravitational potential. In the rare cases where well-defined, equilibrium gas disks are present, emission-line kinematics can yield excellent constraints on the shape of the potential if one assumes that the gas is on closed orbits (Bertola et al.\\ \\markcite{Bertola1991}1991). But for the majority of ellipticals, methods relying primarily on stellar kinematics are essential. Approaches of this type were originated by Binney \\markcite{Binney1985}(1985) and enlarged upon by Franx et al.\\ \\markcite{Franx1991}(1991) and Tenjes et al.\\ \\markcite{Tenjes1993}(1993). Statler (\\markcite{Statler1994a}1994a, \\markcite{Statler1994b}1994b) introduced major refinements, including improved dynamical models and a Bayesian approach to model fitting. This method has great potential to place quite narrow constraints on the triaxialities of individual galaxies for which very high quality stellar kinematic data are available. Unfortunately, the number of such galaxies is still very small, and is likely to increase at only a modest pace in the short term. Our goal in this paper is to see what can be learned from the larger sample of galaxies with stellar kinematic data of less-than-ideal quality already in the literature. We focus on the Davies \\& Birkinshaw \\markcite{DB1988}(1988, hereafter DB) sample of radio ellipticals, all of which have kinematic data on multiple position angles and are photometrically well studied. In the process, we extend the statistical methods of Statler \\markcite{Statler1994b}(1994b) and show how to estimate the parent shape distribution from a sample of galaxies for which the data may be very inhomogeneous. Our method will thus continue to be generally applicable as new data are obtained. In the next section of the paper we describe the general statistical approach for determining the parent distribution of a set of intrinsic quantities from measurements of related, but different, observable quantities, and show the particular application of this approach to the shape problem. In \\S\\ 3, we discuss our treatment of the data and define a subsample of the DB galaxies which we are able to model reliably. Section 4 presents the results for the parent distribution, and examines systematic effects relating to unknown aspects of the stellar dynamics. Section 5 compares our results to those of previous studies, and \\S\\ 6 sums up. ", "conclusions": "By combining photometric and kinematic data with dynamical models using the method of Statler (\\markcite{Statler1994b}1994b), we have derived constraints on the intrinsic shapes and orientations of 13 ellipticals from the Davies and Birkinshaw (\\markcite{DB1988}1988) sample of radio galaxies. Using an iterative Bayesian approach we have then combined those results to estimate the parent shape distribution from which they were drawn, under the assumption that this parent distribution has no preferred orientation. In the process we have obtained improved constraints on the shapes of the individual objects. We have found that the parent shape distribution shows a tendency toward bimodality, with peaks at the oblate and prolate limits. In the distribution derived under minimal assumptions about the galaxies' internal dynamics, only about a one-third of the objects would be strongly triaxial ($0.2 < T < 0.8$). However, the parent distribution does depend on dynamical assumptions. Some of these assumptions can be ruled out because they would require the sample to have a strong orientation bias; configurations in which all galaxies rotate purely about either their long axes or their short axes can be excluded on these grounds. On the other hand, configurations in which the mean motions in the short-axis and long-axis tube orbits are either disklike---dropping off away from the symmetry planes---or spheroidlike---staying approximately constant at a given radius---cannot be distinguished at this point. Whether the rotation is disklike or spheroidlike has a strong effect on the inferred shape distribution. Spheroidlike rotation in the long-axis or short-axis tubes, respectively, significantly reduces the fraction of nearly prolate or nearly oblate galaxies; bimodality is completely eliminated if the long-axis tubes are spheroidlike and the short-axis tubes disklike. In a nutshell, {\\em if rotation in ellipticals is generally disklike, then triaxiality is rare; if spheroidlike, triaxiality is common.} This inferential link between diskiness and axisymmetry complements the intuitive physical notion that the two ought to go hand in hand. There is evidence from the width of the Tully-Fisher relation that the disks of spiral galaxies are very nearly circular (Franx \\& de Zeeuw \\markcite{FdZ92}1992), and indications from numerical experiments that growing even a weak disk in a triaxial halo can render the latter axisymmetric (Dubinski \\markcite{Dubinski1994}1994). Whether weak disks in elliptical galaxies are detectable is another long-standing issue receiving renewed attention (Magorrian \\markcite{Mag99}1999). High-accuracy, multi-position-angle kinematic mapping may be able to reveal hidden disks, but the expected signatures are subtle. Some support is lent to the possibility that weak disks may be common by the kinematic similarities that the ``standard elliptical'' NGC 3379 shares with the S0 galaxy NGC 3115 (Statler \\& Smecker-Hane \\markcite{SSH99}1999). Theoretically, however, the origin of these particular kinematic features is not understood. As we have stressed, a physical understanding of the processes that may establish disklike or spheroidlike rotation in a hot stellar system is sorely needed." }, "0003/astro-ph0003232_arXiv.txt": { "abstract": "s{I review the status of cosmic velocity analysis as of January 2000, with an emphasis on two key questions: (1) What is the scale of the largest bulk flows in the universe? and (2) What is the value of $\\beta\\equiv\\Omega_m^{0.6}/b$ indicated by cosmic velocities, and what does this tell us about $\\Omega_m$ itself? These are the most important issues for cosmic flow analysis, and each has been controversial in recent years. I argue that a preponderance of the evidence at present argues against very large scale ($\\simgt 100\\,\\hmpc$) bulk flows, and favors $\\beta \\simeq 0.4$--$0.5,$ corresponding to a low-density ($\\Omega_m \\simeq 0.2$--$0.3$) universe. } ", "introduction": " ", "conclusions": "I have argued that the two major controversies in cosmic flow analysis have been largely resolved in the last few years. With regard to bulk flows, most recent surveys show convergence to the CMB frame by a distance of $\\sim 60\\hmpc.$ A corollary is that the observed bulk motions do not require more large-scale power than is provided by COBE-normalized CDM density fluctuation spectra. With regard to the value of $\\beta_I=\\Omega_m^{0.6}/b_I,$ a number of independent analyses now suggest a low value, $\\beta_I \\simeq 0.4$--$0.5.$ If the IRAS galaxies are nearly unbiased with respect to mass, a reasonable if not airtight hypothesis, these $\\beta$-values imply a density parameter $\\Omega_m \\simeq 0.2$--$0.3.$ A word of caution is in order, however. The above conclusions represent a consensus view, not a unanimous one. The bulk flow detections listed in the first three rows of Table~1 have not been in any sense ``refuted,'' which is to say as far as we know there is nothing wrong with the data. New surveys, such as FP200 (see http://astro.uwaterloo.ca/$\\sim$mjhudson/fp200 for details) will, it is hoped, settle the issue definitively. % Similarly, while a majority of recent velocity-density comparisons favor low $\\beta_I,$ the reason for the discrepant POTENT result, $\\beta_I \\approx 0.9,$ is not well understood. Tests of methods such as POTENT and VELMOD using N-body simulations are under way, and may clarify things. Moreover, the coming decade will bring much larger TF data sets from the DENIS and 2MASS infrared surveys. As always, these new data sets, if they live up to their promise, are our best hope for putting any remaining contoversy to rest." }, "0003/astro-ph0003004_arXiv.txt": { "abstract": "We estimate the empirical R parameter in 26 Galactic Globular Clusters (GGCs) covering a wide metallicity range, imaged by WFPC2 on board the Hubble Space Telescope. The improved spatial resolution permits a large fraction of the evolved stars to be measured and permits accurate assessment of radial populaton gradients and completeness corrections. In order to evaluate both the He abundance and the He to metal enrichment ratio, we construct a large set of evolutionary models by adopting similar metallicities and different He contents. We find an absolute He abundance which is lower than that estimated from spectroscopic measurements in HII regions and from primordial nucleosynthesis models. This discrepancy could be removed by adopting a \\cdo nuclear cross section about a factor of two smaller than the canonical value, although also different assumptions for mixing processes can introduce systematical effects. The trend in the R parameter toward solar metallicity is consistent with an upper limit to the He to metal enrichment ratio of the order of 2.5. Detailed calculations of central He-burning times as a function of the HB morphology suggest that He lifetimes for hot HB stars are on average $\\approx$ 20\\% longer than for RR Lyrae and red HB stars. Therefore, the increase in the empirical R values of metal-poor clusters characterized by blue HB morphologies is due to an increase in the HB lifetime and not due to an increase in the He abundance. ", "introduction": "The He abundance is fundamental in several astrophysical problems. Big bang nucleosynthesis models supply tight predictions on the primordial He content, and therefore empirical estimates of this parameter are crucial for constraining their plausibility (Hogan, Olive, \\& Scully 1997). At the same time, stellar evolutionary and pulsational models do require the assumption of a He to metal enrichment ratio \\dydz in order to reproduce the observed properties of both metal-poor and metal-rich stellar structures (Bono et al. 1997a). Observational constraints on this parameter can improve the accuracy of several theoretical observables, and in particular of stellar yields predicted by Galactic chemical evolution models (Tsujimoto et al. 1997; Pagel \\& Portinari 1998). One of the most widely used methods for {\\em measuring} the He abundance is to measure fluxes of nebular emission lines in planetary nebulae (Peimbert 1995) or in extragalactic H~II regions (Pagel et al. 1992; Izotov, Thuan, \\& Lipovetsky 1997; Olive, Steigman, \\& Skillman 1997). Independent estimates based on high signal-to-noise measurements of He abundance give very similar results ($Y=0.23-0.24$), thus suggesting that the empirical uncertainties are quite small. However, by adopting detailed radiative transfer calculations of H and He, Sasselov \\& Goldwirth (1995) supported the evidence that current He measurements could be affected by large systematic errors. The He content can also be obtained by direct spectroscopic measurements in hot Horizontal Branch (HB) stars. Unfortunately, these stars are affected by gravitational settling and by radiation levitation (Michaud et al. 1983; Moehler et al. 1999), and therefore they might present peculiar abundance patterns. Nevertheless, there seems to be a consensus that the primordial He abundance should not be lower than Y=0.22 (Olive, Steigman, \\& Walker 1999). Both absolute and/or relative He abundances can also be {\\em estimated} because the evolution of population II stars is sensitive to the primordial helium abundance. The first helium-sensitive indicator to be identified was the R parameter, defined as the ratio between the number of stars along the HB and the number of red giant branch (RGB) stars brighter than the HB luminosity ($R=N_{HB}/N_{RGB}$, Iben 1968). Additional parameters also use the helium burning stars of the horizontal branch as abundance indicators. The $\\Delta$ parameter is the magnitude difference between HB stars and main sequence (MS) stars (Carney 1980) and the $A$ parameter is the mass-luminosity exponent of RR Lyrae stars (Caputo, Cayrel, \\& Cayrel de Strobel 1983). The fine structure of the main sequence locus of population II stars (Faulkner 1967) is also a potentially powerful method to constrain \\dydz. On the basis of Hipparcos parallaxes Pagel \\& Portinari (1998) investigated the fine structure of solar neighborhood MS stars and found that the current estimates of \\dydz are still affected by large uncertainties. The other three methods have been recently applied by Sandquist (1999, hereinafter S99) to a sample of 42 Galactic Globular Clusters (GGCs). Sandquist also comprehensively discusses the pros and cons of these abundance indicators and in particular the statistical and systematic errors affecting both absolute and relative He estimates. The results by S99 support the evidence that both the $\\Delta$ and the $A$ parameter can only give reliable relative He abundances, due to current uncertainties on the metallicity scale and on the RR Lyrae temperature scale. At the same time, S99 brought out that absolute He abundances based on the R parameter ($Y\\approx0.2$) could also be affected by additional systematic errors, and that both relative and absolute estimates do not show, within current uncertainties, a clear evidence of a trend with metallicity. The latter finding does not support the results by Renzini (1994), Minniti (1995), Bertelli et al. (1996), and Desidera, Bertelli, \\& Ortolani (1998) who suggest that in the Galactic bulge the He abundance scales with metallicity according to a slope ranging from 2 to 3.5. Moreover detailed comparisons between solar standard models and accurate helioseismic data (Ciacio, Degl'Innocenti \\& Ricci 1997; Degl'Innocenti et al. 1997; Christensen-Dalsgaard 1998) is more consistent with \\dydz$\\approx2$. The large spread in the empirical values suggests that current He estimates are still hampered by large uncertainties which do not allow us to disentangle the intrinsic variation, if any, from systematic effects. The empirical evaluation of the R parameter relies only on star counts. Nevertheless, misleading effects can be introduced by the method adopted for fixing the Zero Age Horizontal Branch (ZAHB) luminosity, by differential reddening, as well as by the occurrence of population gradients inside the cluster (Buzzoni et al. 1983; Caputo, Martinez Roger, \\& Paez 1987; Djorgovski \\& Piotto 1993; Bono et al. 1995; S99). The He abundance is estimated by comparing observed values with the ratio of HB and RGB evolutionary times, which relies on evolutionary predictions characterized by a negligible dependence on stellar age (Iben \\& Rood 1969). The He burning lifetimes do depend on input physics such as equation of state, opacity, and nuclear cross sections (Brocato, Castellani, \\& Villante 1998; Cassisi et al. 1998) adopted to construct HB models as well as on the algorithm adopted for treating the mixing processes (Sweigart 1990, and references therein). The main aim of this investigation is to derive the R parameter for a sample of 26 GGCs, and to compare empirical values with theoretical predictions in order to gather information on both the He content and its trend with metallicity. In order to accomplish this goal we specifically calculate a large set of HB models, adopting the most up-to-date input physics. We rely on the high number of stars sampled in each cluster, on the wide metallicity range covered by clusters, and on the high homogeneity of theoretical predictions and data to constrain the behavior of both R and \\dydz parameters. ", "conclusions": "We have measured the helium-sensitive R parameter in 26 Galactic globular clusters imaged with WFPC2 on board the Hubble Space Telescope. Our calculated R values are based on star counts that are corrected for completeness and tested for radial variations within each cluster. The high quality HST photometry also permits more clear separation of the HB, AGB, and RGB stars. The comparison between predicted and empirical R values appears to be consistent with the absolute He abundance being lower than that found from the observations of HII regions and from the primordial nucleosynthesis models. One approach to overcome this discrepancy is to adopt a \\cdo nuclear cross section about a factor of two smaller than current canonical values. We note that HB lifetimes depend not only on nuclear reaction rates, but also on the efficiency of mixing processes and on the algorithms adopted for handling these physical mechanisms. In fact, as recently suggested by Cassisi et al. (2000), current algorithms adopted for quenching the \"breathing pulses\" introduce a $\\approx 5 \\%$ uncertainty on $t_{HB}$. As a consequence, the R parameter cannot presently be absolutely calibrated in terms of a helium abundance. The only trend in our data set is an unphysical trend toward higher helium abundance in the clusters of lowest metallicity. These clusters tend to have blue horizontal branches, and we argue that longer HB lifetimes in high temperature HB stars likely account for this trend. In fact, the global trend in R with metallicity is well accounted for by changes in HB lifetime as a function of metal abundance. The trend in R values of the metal-rich globular clusters in our sample is consistent with an upper limit of 2.5 for the helium to metal enrichment ratio (\\dydz). The increased dispersion in R for the intermediate metallicity clusters may be caused by the RGB bump fading below the HB luminosity at [Fe/H]$\\approx -1$, causing a drop in the calculated RGB lifetime. We conclude that these factors make the R values of low and intermediate globular clusters less useful in constraining the helium abundance. Accurate photometric data for metal-rich globular clusters, however, do place an interesting constraint on \\dydz . The GGCs in the Galactic bulge are certainly a key target to accomplish these measurements. It is not a trivial effort to collect high quality data for such clusters, since they are often affected by high absolute and differential extinction (Ortolani et al. 1999), but the new NIR detectors should allow us to overcome these difficulties and to secure accurate data for a sizable fraction of cluster stars. A reconsideration of the bulge clusters and field population would also be in order given the results of Minniti (1995); when his counts are corrected for the contribution of AGB stars, the bulge fields have R values of 1.7 to 2, higher than those of the most metal-rich clusters in our sample. The promise of the R parameter being a constraint on the primordial helium abundance remains frustratingly unfulfilled. Two further thorny problems affect sound empirical estimates of this parameter, namely the foreground contamination and the radial population gradients. The former can be overcome by including only the innermost regions, whereas the latter is an open problem, since we still lack a systematic and quantitative estimate of this effect in GGCs from the very center to the tidal radius (Walker 1999). The precise comparison of synthetic H-R diagrams with observational data may be required to achieve further progress." }, "0003/astro-ph0003311_arXiv.txt": { "abstract": "We examine standard methods of measuring nebular chemical abundances, including estimates based on direct \\te\\ measurements, and also emission-line diagnostics. We use observations of the LMC \\hii\\ regions DEM L199, DEM L243, DEM L301, and DEM L323, whose ionizing stars have classifications ranging from O7 to WN3. Following common practice, we assume a two-zone \\te\\ structure given by \\thi\\ and \\tlo\\ to compute ionic abundances. We compare with photoionization models tailored to the observed properties of the individual objects, and emphasize the importance of correctly relating \\te\\ in the two zones, which can otherwise cause errors of $\\sim$0.2 dex in abundance estimates. The data show no spatial variations or local metallicity enhancements to within 0.1 -- 0.15 dex in any of the objects, notably including DEM L199, which hosts three WR stars. Our data agree well with both the modeled $R23$ and $S23$ abundance diagnostics for O and S. We present the first theoretical tracks for $S23$, which are in excellent agreement with a larger available dataset. However, contrary to earlier suggestions, $S23$ is much {\\it more} sensitive to the ionization parameter ($U$) than is $R23$. This occurs because $S23$ does not sample \\ion{S}{4}, which is often a significant population. We therefore introduce $S234\\equiv$(\\sii$+$\\siii$+$\\siv)/\\Hb, and demonstrate that it is virtually independent of $U$. Predicted and observed spatial variations in $S234$ are thus dramatically decreased in contrast to $S23$. The intensity of \\siv\\lam10.5$\\mu$ can be easily estimated from the simple correspondence between \\siv/\\siii\\ and \\oiii/\\oii. Using this method to estimate $S234$ for data in the literature yields excellent agreement with our model tracks, hence we give a theoretical calibration for $S234$. Our models show that the double-valued structure of $S23$ and $S234$ remains an important problem as for $R23$, and presently {\\it we consider calibrations of these S diagnostics reliable only at $Z\\lesssim 0.5\\Zsol$}. However, the slightly larger dynamic range and excellent compatibility with theoretical predictions suggest the S parameters to be more effective abundance diagnostics than $R23$. ", "introduction": "The emission-line signatures of \\hii\\ regions are a powerful and widely-used indicator of galactic abundances. This is especially true in distant galaxies where most stellar abundance probes cannot be employed. Furthermore, the spectral properties of \\hii\\ regions also give important diagnostics of the ionizing stellar population, such as effective temperature and numbers of stars. Given the wide use of such nebular diagnostics, it is vital to test and calibrate them using \\hii\\ regions with known characteristics and ionizing stellar populations. Along with a companion paper (Oey {\\etal}2000; hereafter Paper~I), we report here on a detailed study for this purpose, of four \\hii\\ regions in the Large Magellanic Cloud (LMC). The OB associations in all four of these \\hii\\ regions have been examined in detail and classified, thereby strongly constraining the ionizing energy distributions. In addition, the LMC's proximity and orientation with respect to the Galaxy also permit a detailed understanding of the nebular morphology. Thirdly, the abundances can be accurately determined. We therefore have high-quality information on the three principal parameters that determine the nebular emission: stellar effective temperature (\\tstar), the ionization parameter ($U$) that relates ionizing flux to gas density, and metallicity ($Z$). Table~\\ref{sample} gives a brief summary of our objects, which can be examined in greater detail in Paper~I. The first two columns in Table~\\ref{sample} give the \\hii\\ region identification in the Davies, Elliott, \\& Meaburn (1976) and Henize (1956) \\Ha\\ catalogs, respectively; the third column identifies the parent OB association from Lucke \\& Hodge (1970). Column four shows the spectral type of the dominant ionizing stars, as classified by the references in column 8. Column five lists the nebular \\Ha\\ luminosity from Oey \\& Kennicutt (1997), and column 6 shows the adopted value from Paper~I, of the inner radius of the gas distribution, as a fraction of the Str\\\"omgren radius $R_{\\rm S}$. This gives some indication of the nebular morphology. Finally, column 7 indicates the presence of significant shock excitation: DEM L243 includes an embedded or superimposed supernova remnant (SNR), and DEM L301 also shows evidence of significant shock activity (Paper I). While shocks may be present in the other objects as well, their contribution to the nebular emission is unimportant. The detailed presentation of the objects is given in Paper~I, including narrow-band images in \\Ha, \\oiii, and \\sii, and spectroscopic observations over the wavelength range 3500 -- 9200 \\AA. For each object, we observed two to three stationary, spatially-resolved slit positions. For all of the objects except DEM L301, we also obtained spatially integrated observations by scanning the long slit across the nebulae. The scanned data should therefore resemble typical observations of such \\hii\\ regions at distances of 10 -- 20 Mpc. As seen in Table~\\ref{sample}, the sample spans a range in dominant stellar spectral types, from O7 to early Wolf-Rayet (WR). There is also variety in the morphology of the objects, ranging from near perfect Str\\\"omgren sphere (DEM L323), to extreme shell structure (DEM L301), to highly complex (DEM L199). Paper~I provides a detailed analysis of the spatially resolved emission-line ratios with respect to the sequence in \\tstar\\ represented by these objects, and a comparison with photoionization models using the current generation of stellar atmosphere models for both O stars and early WR stars. In general we found a gratifyingly high level of agreement, largely supporting the CoStar energy distributions of Schaerer \\& de Koter (1997) and early WR atmospheres of Schmutz, Leitherer, \\& Gruenwald (1992) and Hamann \\& Koesterke (1998). In conjunction with data from Kennicutt {\\etal}(2000), we provide a first, empirical calibration of nebular diagnostics for the dominant \\tstar. In addition to the well-known $\\eta^\\prime$ radiation softness parameter of V\\'\\i lchez \\& Pagel (1988), we introduce [\\ion{Ne}{3}]\\lam3869/\\Hb\\ as a similar diagnostic, which is more robust to nebular conditions and is sensitive for a higher range of \\tstar. We also presented in Paper~I the spatially-resolved behavior of the O abundance parameter (Pagel {\\etal}1979), \\begin{equation}\\label{eqR23} R23 \\equiv \\rm\\frac{[O\\thinspace II]\\lambda3727 + [O\\thinspace III]\\lambda\\lambda4959,5007}{\\hb} \\quad ; \\end{equation} and S abundance parameter (V\\'\\i lchez \\& Esteban 1996; Christensen {\\etal}1997; D\\'\\i az {\\etal}1999), \\begin{equation}\\label{eqS23} S23 \\equiv \\rm\\frac{[S\\thinspace II]\\lambda6724 + [S\\thinspace III]\\lambda\\lambda9069,9532}{\\hb} \\quad , \\end{equation} where we designate the \\sii\\ lines $\\lambda6716 + \\lambda6732$ as \\lam6724, analogous to \\oii\\lam3727. We confirmed that both observationally and theoretically, $R23$ remains spatially uniform across the nebulae. In contrast, for uniform abundances, models of $S23$ vary across the nebulae, showing lower values in central regions. The observations clearly reflect this pattern, which is caused by the missing ionization stage of \\ion{S}{4}. We address this issue in detail below in this paper. As is well-known, $R23$ and $S23$ are also sensitive to \\tstar, and these effects are also shown in Paper~I for our objects. In this paper, we present the abundance determinations for our sample, for both the spatially-resolved and scanned longslit observations. We examine conventional assumptions for the nebular electron temperature (\\te) structure and resulting ionic and total abundance determinations. We then explore the metallicity diagnostics $R23$ and $S23$ in more detail, and introduce another diagnostic, $S234$. As before, we use the photoionization code {\\sc Mappings~II} (Sutherland \\& Dopita 1993) in conjunction with CoStar stellar atmosphere models (Schaerer \\& de Koter 1997) for O stars. \\bigskip\\bigskip ", "conclusions": "We have carried out a detailed investigation of elemental abundance derivations using four \\hii\\ regions in the LMC. We use tailored photoioinzation models to examine standard abundance analyses based on measured values of \\te. Our data (Paper~I) are derived from both spatially-resolved observations extracted from stationary long slit positions, and scanned, spatially-integrated slit observations. We also examine the bright-line abundance diagnostics for O and S, in light of the direct abundance determinations and photoionization models. Our abundance determinations are based on measurements of $T$\\oiii, which we take to represent \\thi, and we assume a two-zone temperature structure for the nebulae, represented by \\thi\\ and \\tlo. We use standard ionic abundance relations to then determine the total elemental abundances for He, N, O, Ne, S, and Ar, with respect to H. Comparison with tailored {\\sc Mappings} photoionization models highlights the importance of choosing a relation between \\thi\\ and \\tlo\\ that adequately represents the nebular temperature structure. Failure to do so can result in metallicity estimates that are discrepant by at least 0.2 dex from values indicated by tailored photoionization models. Abundance measurements for the stationary slit positions show high spatial uniformity, with no evidence of variations or gradients to within 0.1 -- 0.15 dex. Thus it is unlikely that there are systematic biases resulting from the strong ionization gradients seen in these objects. The adopted two-zone \\te\\ structure therefore appears to be highly reliable for estimating ionic abundance estimates even through ``pencil-beam'' apertures that sample only a small nebular area, at least for our fairly isothermal \\hii\\ regions. No areas of local enrichment were detected in DEM L199, in spite of the presence of two WN3 stars and one WC4 star. The stellar products may be hidden in hot, coronal gas within the central superbubble, or the stars may not have produced enough enriched material to be readily detectable. The results show that self-enrichment by WR stars is likely to be a complex phenomenon, empirically. DEM L243 and DEM L301, both showing evidence of recent SNR activity, also do not show local enrichments, although with poorer constraints. Abundance measurements from the scanned, spatially integrated apertures are consistent with those obtained from the spatially resolved observations. Our results are $\\sim 0.2$ dex lower than average LMC \\hii\\ region measurements (Dufour 1984; Garnett 1999), probably resulting in part from different descriptions for the \\te\\ structure. The spatially-integrated measurements are also consistent with there being no variation between the four \\hii\\ regions, although, interestingly, they are also consistent with the marginal abundance gradient suggested by Pagel {\\etal}(1978). While the presence of the SNR in DEM L243 did not affect the resulting abundances from the spatially integrated observation, the derived abundances for DEM L301 are on the low end of the distribution, hinting at spurious effects caused by the shock activity in that superbubble. We computed the $R23$ O abundance parameters (Pagel {\\etal}1979) for the spatially integrated data, and compared these with model tracks constructed with {\\sc Mappings}. The models assume the Costar C2 (Schaerer \\& de Koter 1997) stellar atmosphere corresponding to an O6 -- O7 spectral type, and an inner nebular radius of 0.4 $R_{\\rm S}$. As has historically been the case, the models do not agree well with the locus of observations in the literature, although our LMC data do agree well, coincidentally, with both. Similarly, we examined the $S23$ abundance parameter for S (e.g., Christensen {\\etal}1997; D\\'\\i az \\& P\\'erez-Montero 2000). Our models reveal that, contrary to previous suggestions, $S23$ is {\\it more} sensitive to the ionization parameter than is $R23$. \\ion{S}{4} is produced by the same radiation that ionizes \\ion{O}{3}, and is a significant ion of S in many \\hii\\ regions, but it is not sampled by $S23$. Its omission therefore causes $S23$ to be much more sensitive to $U$ than $R23$. The spatially resolved observations confirm this by showing, in agreement with model predictions, lower values of $S23$ in the central nebular regions where \\ion{S}{4} is important. As shown in Paper~I, this spatial variation is not predicted or observed in $R23$. Our models also suggest that the maximum in $S23$ occurs at only $\\sim 0.2$ dex higher in $Z$ than in $R23$. Nevertheless, this appears to significantly alleviate the effect of the double-valued structure of log(S/H) vs $S23$ when inferring abundances, as shown by D\\'\\i az \\& P\\'erez-Montero (2000). It is highly encouraging that the data, both from our sample and from the literature, are in excellent agreement with the models, in contrast to the behavior of $R23$. We offer a theoretical calibration of $S23$ (equation~\\ref{calibS23}) which appears to be fully compatible with the data in the literature thus far. However, we caution that the locus of the available data may well be deceptive in suggesting that a power-law relation can be used at $Z\\gtrsim 0.5\\Zsol$. To overcome the limitations of $S23$ in $U$-sensitivity and spatial variation, we introduce a similar S abundance parameter, $S234$. This is the same as $S23$ with the added emission of \\siv\\lam10.5$\\mu$. Although this mid-IR line is not readily observable with most conventional ground-based spectrographs, it is straightforward to estimate its intensity from the simple correspondence between \\siv/\\siii\\ and \\oiii/\\oii\\ (equation~\\ref{eqS4O3}). Our models show that $S234$ is less dependent on $U$ than is even $R23$. $S234$ for our objects and for the larger sample of Dennefeld \\& Stasi\\'nska (1983) are in excellent agreement with the models. Likewise, the spatial variations for both models and observations are dramatically reduced for $S234$ in contrast to $S23$. We provide a theoretical calibration for log(S/H) vs $S234$ at $Z\\lesssim 0.5\\Zsol$ (equation~\\ref{calibS234}). Finally, we reiterate some caveats for the use of $R23$, $S23$, and $S234$. We find that the presence of shock excitation increases the value of these parameters; for our objects, the effect is about 0.1 dex in magnitude. Secondly, significant variations in the S/O ratio dictate caution in inferring O abundances using $S234$ and $S23$ (Figure~\\ref{DSfig}). It is also important to bear in mind the double-valued structure for all three of these parameters. Lastly, we emphasize the deviation between the data and models above $0.5\\Zsol$, and we therefore consider the calibrations presented thus far for $S23$ and $S234$ to be reliable only for $Z \\lesssim 0.5\\Zsol$. Further empirical investigation is needed to understand the behavior of these parameters at higher metallicity. Bearing in mind these caveats, the excellent correspondence between the modeled $S234$, $S23$, and the available data, together with the more highly monotonic behavior of these parameters, promises greater effectiveness as metallicity indicators than $R23$. With improving access to the \\siv\\lam10.5$\\mu$ line, it should be possible to confirm the behavior of $S234$ directly." }, "0003/astro-ph0003257_arXiv.txt": { "abstract": "The CCD magnitudes in Cousins $R$ and $I$ photometric passbands are determined for GRB 991216 and GRB 991208 afterglows respectively $\\sim$ 1 and $\\sim$ 3 day after trigger of the corresponding $\\gamma-$ray bursts. Light curves of the afterglow emissions are obtained by combining the published data with the present measurements in $R$ and $I$ passbands for GRB 991208 and in $R$, Gunn $i$ and $J$ passbands for GRB 991216. They indicate that the flux decay constants of a GRB are almost the same in each passband with values $\\sim 2.2$ for GRB 991208 and $\\sim 1.2$ for GRB 991216 indicating very fast optical flux decay in the case of former which may be due to beaming effect. However, cause of steepening by $0.23\\pm0.06$ dex in the $R$ light curve of GRB 991216 afterglow between 2 to 2.5 day after the burst is presently not understood. Redshift determinations indicate that both GRBs are at cosmological distance with a value of 4.2 Gpc for GRB 991208 and 6.2 Gpc for GRB 991216. The observed fluence $>$ 20 keV indicates, if isotropic, release of energy $\\sim 1.3\\times 10^{53}$ erg for GRB 991208 and $\\sim 6.7\\times 10^{53}$ erg for GRB 991216 by these bright $\\gamma-$ray flashes. The enormous amount of released energy will be reduced, if the radiation is beamed which seems to be case for GRB 991208 afterglow. The quasi-simultaneous broad-band photometric spectral energy distributions of the afterglows are determined $\\sim$ 8.5 day and $\\sim$ 35 hour after the bursts of GRB 991208 and GRB 991216 respectively. The flux decreases exponentially with frequency. The value of spectral index in the optical-near IR region is $-0.75\\pm$0.03 for GRB 991208 and $-1.0\\pm$0.12 for GRB 991216. ", "introduction": "For GRB 991208, prompt $\\gamma-$ray emissions were detected by KONUS and NEAR while in the case of GRB 991216, they were detected by BATSE and NEAR. We have downloaded the light curves from the archive and shown them in Figs. 2 and 3 for GRB 991208 and GRB 991216 respectively. Presence of multi-peaked spiky temporal profile in the energy distributions of both GRBs is an unambiguous indicator of a series of internal shocks within a relativistic flow. Further discussions on the $\\gamma-$ray light curves of each GRB are given below. \\subsection {GRB 991208} Fig. 2 shows the light curve in two energy bands accumulated by KONUS in 50 -- 200 keV and by NEAR in 100 -- 1000 keV. The burst profile is dominated by two strong peaks, separated by about 55 s. Almost identical temporal as well as intensity structures in the light curves at both energy bands indicate that perhaps, dominant emission is in the 100 -- 200 keV range as it is common in both observed energy bands. The burst began with a strong pulse which lasted for $\\sim$ 6 s. It has a sharp rise and a relatively slow decline. This is followed by a relatively weak pulse starting at $\\sim$ 40 s after the trigger. It is a relatively broad profile. Almost at the end of this pulse and $\\sim$ 52 s after trigger of the burst, the strongest pulse of this GRB started and lasted for $\\sim$ 20 s. This has almost the same rise and decline time, though the profile is multi-peaked, asymmetric and irregular. As expected, the spiky nature of the profile is clearly visible only on the 64 ms time resolution light curve. Duration (full width at half maxima) of the profile at trigger of the burst is only $\\sim$ 2 s while that of the strongest one is more than 5 s. They also differ in temporal structures. The GRB is a long duration burst as it lasted for more than 70 s. \\subsection {GRB 991216} We show in Fig. 3 the light curves of GRB 991216 in four energy bands, obtained by the BATSE on board the Compton Gamma-Ray Observatory satellite. The light curve obtained by NEAR is not shown here as it resembles to the BATSE highest energy light curve and also has poor time resolution. The burst has a complicated and irregular time profile. The event began with a weak precursor pulse lasting about 2 s, followed $\\sim$ 15 s later by an intense multi-peaked complex. The main emission lasted for $\\sim$ 17 s followed by a fainter tail that persisted for another $\\sim$ 20 s (Fig. 3). The $T_{50}$ and $T_{90}$ durations of the burst, as measured in the 50 -- 300 keV energy range, are 6.272$\\pm$0.09 s and 15.168$\\pm$0.11 s respectively (Kippen 1999) indicating that it is a long duration burst. The burst lasted for $\\sim$ 60 s and it has peaks of width $\\sim$0.5 s, yielding a value of the variability index as 120. \\medskip The overall shape of the GRB 991216 in all the energy bands can be described as a fast rise starting at $\\sim$ 17 s; arrived maximum $\\sim$ 21 s and then decayed slowly. Each phase of the burst profile contains a number of well-defined short duration (full width at half-maxima generally $\\le$ 0.5 s) sub-pulse or spikes within the burst. There are 14 such spikes. We list in Table 2 their time of occurrence and relative counts with respect to the first spike which has 24.3, 38.9, 49.1 and 8.5 K count/s above the background in the energy bands 20 -- 50 keV, 50 -- 100 keV, 100 -- 300 keV and $>$ 300 keV respectively. Spikes 1, 2 and 3 are during ascending phase; 4, 5 and 6 are during maxima phase and others are during descending phase of the burst. The peak of spikes occurred almost simultaneously in all the four energy bands. The variations in relative count rates of a spike from higher to lower energy bands are similar for spikes of a phase but differ from the spikes of other phases. The relative count rates are nearly the same in all energy bands for the spikes during ascending phase; they are more in $>$ 100 keV energy range than in 20 -- 100 keV range for spikes around maxima and they decrease systematically with increase in energy for spikes of descending phase. The light curves show that the low-energy emission persists longer and peaks later than high-energy emission. Particularly striking is the paucity of $>$ 300 keV emission during the shoulder about 32 s after trigger of the burst. In the highest energy band emission is maximum in spike 4 and is relatively much reduced at lower energy bands. {\\bf Hard-to-soft} spectral evolution thus observed in GRB 991216 is the typical one usually (but not always) seen throught the GRB and also in sub-pulses within a burst (Fishman et al. 1999 and references therein). \\vspace{0.5cm} \\noindent {\\bf Table 2.}~List of well-defined $\\gamma-$ray spikes during GRB 991216. Time of its occurrence ($T_p$) in 20 -- 50 keV energy band after trigger of the burst at 1999 December 16.671 UT along with their relative count rate with respect to first spike in all the four $\\gamma-$ray energy bands are given. Spikes are identified with the ascending, maxima and descending parts of the GRB 991216 burst. \\begin{tabular}{cccccc l} \\hline Spikes&$T_p$&\\multicolumn{4}{c}{Relative count rate in keV energy band}&Phase \\\\ &\t(s) & 20 -- 50 & 50 -- 100 & 100 -- 300 & $>$ 300 \\\\ \\hline 1 & 17.3 & 1.000 & 1.000 & 1.000 & 1.000 & Ascending \\\\ 2 & 19.3 & 1.693 & 1.591 & 1.554 & 1.611 & Ascending \\\\ 3 & 19.7 & 1.771 & 1.804 & 1.810 & 1.705 & Ascending \\\\ 4 & 19.9 & 1.775 & 1.809 & 2.197 & 3.526 & Maxima \\\\ 5 & 20.6 & 2.381 & 2.356 & 2.561 & 2.586 & Maxima \\\\ 6 & 21.9 & 2.216 & 2.089 & 2.368 & 2.452 & Maxima \\\\ 7 & 22.4 & 1.610 & 1.360 & 1.212 & 0.847 & Descending \\\\ 8 & 22.7 & 2.402 & 2.402 & 2.470 & 1.752 & Descending \\\\ 9 & 23.0 & 2.072 & 1.719 & 1.187 & 0.589 & Descending \\\\ 10 & 23.7 & 1.313 & 0.978 & 0.601 & 0.307 & Descending \\\\ 11 & 25.6 & 1.124 & 0.836 & 0.575 & 0.313 & Descending \\\\ 12 & 26.8 & 1.033 & 0.679 & 0.369 & 0.154 & Descending \\\\ 13 & 29.5 & 1.887 & 1.614 & 1.291 & 0.918 & Descending \\\\ 14 & 31.1 & 1.940 & 1.616 & 1.244 & 0.713 & Descending \\\\ \\hline \\end{tabular} \\medskip In the energy band 50 -- 300 keV, peak flux at 64, 256 and 1024 ms intervals are 19.93$\\pm$0.24, 17.80$\\pm$0.11 and 14.55$\\pm$0.13 $\\mu$erg/cm$^2$/s respectively while the fluence in the BATSE energy channels 1, 2, 3 and 4 are 17.2637$\\pm$0.053, 22.9693$\\pm$0.054, 65.0656$\\pm$0.136 and 150.204$\\pm$1.167 $\\mu$erg/cm$^2$ respectively (Kippen 1999). The hardness ratio $\\frac{f_{100-300}}{f_{50-100}}$ is thus = 2.83$\\pm$0.02. ", "conclusions": "" }, "0003/astro-ph0003061_arXiv.txt": { "abstract": " ", "introduction": "The last five years have seen great progress in the detection of brown dwarfs in the Local Neighborhood, young Galactic Clusters and star formation regions, starting with the near simultaneous discovery of the first clearly confirmed brown dwarfs, Teide 1 in the Pleiades (Rebolo, Zapatero-Osorio \\& Martin 1995) and Gl229b in the Local Neighborhood (Nakajima et al.1995). Star formation regions offer the advantage that substellar objects are 3 orders of magnitude more luminous at an age of a few Myr than at an age of a few Gyr. Early photometric and spectroscopic work (Comeron et al.1993,1996; Williams et al.1995) indicated that brown dwarfs are probably very common in star formation regions. However, confirmation of substellar status is problematic in star formation regions, owing to the ubiquity of Lithium in young objects and the complicating effects of extinction on both photometry and spectroscopy. Recently, high quality spectroscopy (Luhman \\& Rieke 1998; Luhman et al. 1998, Wilking, Greene \\& Meyer 1999) and the publication of theoretical evolutionary models for young substellar objects (Burrows 1997;D'Antona \\& Mazzitelli 1998, hereafter B97 and DM98) have provided convincing evidence that photometric identification of young brown dwarf candidates is reliable. In photometric studies, masses of candidate objects are derived by comparison of the observables (luminosity and temperature) with the evolutionary tracks. The isochrones of B97 and DM98 are in fairly good agreement in regard to the mass-luminosity relation at an age of about 1 Myr but there is some disagreement about the mass-temperature relation (HR diagrams are compared by Luhman \\& Rieke 1998). Even if the theoretical effective temperatures were without flaw, there is considerable uncertainty in the derivation of temperatures from photometry or spectroscopy, at the level of $\\pm 200~K$ in M and L dwarfs. Hence, we use luminosity, which is more easily measured, to derive masses for our sources. In this paper we report the results of a deep infrared photometric survey of the Trapezium Cluster in Orion. A large population of substellar objects is discovered, including the first free-floating objects of planetary mass. We note that the IAC group (Bejar et al.1999, not yet published) has simultaneously reported a similar discovery of planetary mass objects in the adjacent $\\sigma$ Orionis cluster. The Trapezium has been intensively studied for many years and we have been able to draw upon a large body of publications to aid in our work. We selected the Trapezium for several reasons. (1) Its very high stellar density allowed photometry of several hundred sources in a fairly small survey. (2) False positive detections are essentially eliminated because the dense backdrop of OMC-1 obscures all background stars even at K band, as shown by Hillenbrand \\& Hartmann (1998) through optical-infrared comparison of the cluster stellar density profile. (3) The extinction within the cluster is relatively low ($0 < A_{V} < 15$ for most sources), permitting reasonably precise dereddening. (4) Star formation is essentially complete in the cluster and the age range is thought to be 0.3-2~Myr, so the age-luminosity degeneracy is not large. ", "conclusions": "" }, "0003/astro-ph0003477_arXiv.txt": { "abstract": "We present new grids of pre-main sequence (PMS) tracks for stars in the mass range 0.1 to 7.0\\msun. The computations were performed for four different metallicities ($Z$=0.01, 0.02, 0.03 and 0.04). A fifth table has been computed for the solar composition ($Z$=0.02), including a moderate overshooting. We describe the update in the physics of the Grenoble stellar evolution code which concerns mostly changes in the equation of state (EOS) adopting the formalism proposed by Pols et al. (1995) and in the treatment of the boundary condition. Comparisons of our models with other grids demonstrate the validity of this EOS in the domain of very low-mass stars. Finally, we present a new server dedicated to PMS stellar evolution which allows the determination of stellar parameters from observational data, the calculation of isochrones, the retrieval of evolutionary files and the possibility to generate graphic outputs. ", "introduction": "Interests in modeling of very low-mass stars (VLMS) has grown steadily over the past decade due to big efforts in both theory and observations. Brown dwarfs and very low-mass stars are now frequently discovered and the abundant literature in this domain attests of the vitality of this field. From these observations, we now realize that very low-mass stars represent a large fraction of the stellar population. However, the precise determination of the initial mass function, and more generally of stellar parameters and evolutionary states is still limited by the accuracy of stellar models. It is therefore desirable that different tracks be available in order to understand the theoretical discrepancies and also to estimate the uncertainties associated with the derivation of stellar parameters. On the theoretical side, the computation of the structure of very low-mass stars is very challenging since it involves the micro-physics associated with dense, cold and partially degenerate matter. In the interior of VLMS, collective effects due to large densities become important and an accurate treatment of pressure ionization and Coulomb interactions is required. Also, in the cool atmosphere of VLMS, several molecules form and produce strong absorption bands that considerably modify the emergent spectrum. As a consequence, the stellar surface cannot be considered as a black body and stellar atmosphere models must also be used. Since our last grids (Siess et al. 1997, hereafter SFD97), substantial improvements have been made in the Grenoble stellar evolution code that led us to the production of new grids of PMS tracks. The main modification to the code concerns the equation of state (EOS) which treats more accurately the behavior of cold and degenerate matter. In particular, we now take into account the effects associated with the pressure ionization and treat more carefully degeneracy conditions in stellar interior. We also updated the opacity tables using the latest release of the OPAL group and modified our surface boundary conditions using analytic fits to stellar atmosphere models. The updates to the code are describe in Sect. \\ref{code} followed by a brief description of the grids. Then in Sect. \\ref{comp} we compare our models to grids recently published by other groups. Finally, we describe a new server dedicated to PMS evolution. ", "conclusions": "We presented new pre-main sequence evolutionary tracks for low- and intermediate-mass stars. Comparisons of our models with other PMS tracks indicate rather strong discrepancies in the regime of very low-mass stars ($M\\la 0.5$\\msun). Our tracks are very similar, in morphology and effective temperature, to those computed by Baraffe et al. (1998) and the Geneva group (Charbonnel et al, 1999). This similarity is a strong indicator that our treatment of the EOS and boundary conditions are correct. Conversely, we report strong deviations of these tracks with the D'Antona and Mazzitelli (1997) ones, especially below 0.3 \\msun. For higher mass stars, the morphology of the tracks are similar and effective temperature differences $\\la 200$K are noted, partly due to different mixing length parameter values and chemical compositions. Comparisons of the isochrones indicate age determination remains, in any case, very uncertain below $10^6$yr. Finally, we present our internet server which provides several facilities to use and take advantage of our large database of PMS stellar models. \\begin{acknowledgement} The authors wish to thank C. Tout and O. Pols for their numerous and very helpful interactions during the implementation of the EOS in the code. LS also wants to thank F. Roch for her help in building the server. The computations presented in this paper were performed at the ``Centre de Calcul de l'Observatoire de Grenoble'' and at ``IMAG'' on a IBM SP1 computer financed by the MESR, CNRS and R\\'egion Rh\\^one-Alpes. \\end{acknowledgement}" }, "0003/astro-ph0003351.txt": { "abstract": "During the last decade, the understanding of fine features of the structure and evolution of stars has become possible as a result of enormous progress made in the acquisition of high-quality observational and experimental data and of new developments and refinements in the theoretical description of stellar plasmas. The confrontation of high-quality observations with sophisticated stellar models has allowed many aspects of the theory to be validated, and several characteristics of stars relevant to Galactic evolution and cosmology to be inferred. This paper is a review of the results of recent studies undertaken in the context of the Hipparcos mission, taking benefit of the high-quality astrometric data it has provided. Successes are discussed, as well as the problems that have arisen and suggestions proposed to solve them. Future observational and theoretical developments expected and required in the field are also presented. ", "introduction": "\\label{intro} Stars are the main constituents of the observable Universe. The temperatures and pressures deep in their interiors are out of reach for the observer, while the description of stellar plasmas requires extensive knowledge in various domains of modern physics such as nuclear and particle physics, atomic and molecular physics, thermo- and hydrodynamics, physics of the radiation and of its interaction with matter, and radiative transfer. The development of numerical codes to calculate models of stellar structure and evolution began more than forty years ago with the pioneering works of Schwarzschild (1958) and Henyey et al (1959). These programs have allowed at least the qualitative study and understanding of numerous physical processes that intervene during the various stages of stellar formation and evolution. During the last two decades, observational data of increasingly high accuracy have been obtained as a result of 1) the coming of modern ground-based or space telescopes equipped with high-quality instrumentation and with detectors giving access to almost any possible range of wavelengths and 2) the elaboration of various sophisticated techniques of data reduction. Ground-based astrometry has progressed, while space astrometry was initiated with Hipparcos. In the meantime, CCD detectors on large telescopes opened the era of high-resolution, high signal-to-noise ratio spectroscopy while multi-color filters were designed for photometry. New fields have appeared or are under development, such as helio- and asteroseismology or interferometry. On the other hand, stellar models have been enriched by a continuously improved physical description of the stellar plasma, while the use of increasingly powerful computers has led to a gain in numerical accuracy. The confrontation of models with observations allows testing and even validation of the input physics of the models if numerous observations of high quality are available. Fundamental returns are expected in many domains that make use of quantitative results of the stellar evolution theory such as stellar, Galactic, and extragalactic astrophysics as well as cosmology. Because of their positions, movements, or interactions with the interstellar medium stars are actors and tracers of the dynamical and chemical evolution of the Galaxy. Astrophysicists aim to determine their ages and chemical compositions precisely. For example, the firm determination of the ages of the oldest stars, halo stars or members of globular clusters, is a long-standing objective because it is one of the strongest constraints for cosmology. Although great progress has been made, a number of observations cannot be reproduced by stellar models, which raises many questions regarding both the observations and the models. In the last few years, two scientific meetings have been explicitly devoted to unsolved problems in stellar structure and evolution (Noels et al 1995, Livio 2000). A major point of concern is that of transport processes at work in stellar interiors (transport of the chemical elements, angular momentum or magnetic fields by microscopic diffusion and/or macroscopic motions). Observations show that transport processes are indeed playing a role in stellar evolution but many aspects remain unclear (sometimes even unknown) and need to be better characterized. Another crucial point concerns the atmospheres, which link the stellar interior model to the interstellar medium and are the intermediate agent between the star and the observer. Uncertainties and inconsistencies in atmospheric descriptions generate errors in the analysis of observational data and in model predictions. This paper is the third of the series in ARAA dedicated to the results of the Hipparcos mission; Kovalevsky (1998) presented the products of the mission and the very first astrophysical results obtained immediately after the release of the data, while Reid (1999) reviewed the implications of the Hipparcos parallaxes for the location of the main sequence (MS) in the Hertzsprung-Russell (H-R) diagram, the luminosity calibration of primary distance indicators, and the Galactic distance scale. Also, van Leeuwen (1997) presented the results of the mission, and Baglin (1999) and Lebreton (2000) discussed the impact of Hipparcos data on stellar structure and evolution. Hipparcos has provided opportunities to study rather large and homogeneous samples of stars sharing similar properties, for instance, in terms of their space location or chemical composition. I review studies based on Hipparcos observations which ({\\it 1}) confirmed several elements of stellar internal structure theory, ({\\it 2}) revealed some problems related to the development of stellar models, and ({\\it 3}) yielded more precise characteristics of individual stars and clusters. In Sections 2 and 3, I discuss the recent observational (including Hipparcos) and theoretical developments from which new studies could be undertaken. In Section 4, I concentrate on the nearest stars, observed with highest precision (A-K disk and halo single or binary field stars, and members of open clusters). In Section 5, I review recent results on variable stars, globular clusters and white dwarfs based on Hipparcos data. The stars considered are mostly of low or intermediate mass, and except for white dwarfs, the evolutionary stages cover the main sequence and subgiant branch. Throughout this paper, I emphasize that the smaller error bars on distances that result from Hipparcos make the uncertainties on the other fundamental stellar parameters more evident; fluxes, effective temperatures, abundances, gravities, masses and radii have to be improved correspondingly, implying in many cases the need for progress in atmospheric description. ", "conclusions": "" }, "0003/astro-ph0003463_arXiv.txt": { "abstract": "We present compact, analytic expressions for the age-redshift relation $\\tau(z)$ for standard Friedmann-Lema\\^\\i tre-Robertson-Walker (FLRW) cosmology. The new expressions are given in terms of incomplete Legendre elliptic integrals and evaluate much faster than by direct numerical integration. ", "introduction": "Since Type Ia supernova observations \\cite{SB,PS} have favored a Universe with $\\Lambda \\neq 0$, interest in FLRW cosmologies has grown. For this family of models most observational relations, \\eg the Hubble curve, are given by integral expressions; however, \\textcite{kantk2000} recently succeeded in giving useful analytic distance-redshift relations for them. In this paper we extend these analytic results to include the age-redshift relation $\\tau(z)$. These new expressions are useful for any computation that requires a transformation $\\tau(z)$ from the observed redshift variable $z$ to the age $\\tau$ of the Universe at that $z$. \\textcite{FB} provides related light travel times in terms of Legendre elliptic integrals; however, his expressions are not easy to make use of. A presentation closer to what we give appears in \\cite{ED}. There, light travel time as a function of redshift was given for the $\\OO=1$ universe, see (\\ref{age_B1}). For all other cases, \\textcite{ED} was only able to give $\\tau$ and $z$ parametrically as Jacobi elliptic functions of conformal time $\\omega\\equiv \\int dt/R(t)$.\\footnote{ The error detected when numerical checks were performed by \\textcite{CL} was evidently caused by an error in equation 361.54 of \\textcite{BF}, see footnote 3 of \\cite{kantk2000}} We have concentrated on giving useful and valid expressions for $\\tau(z)$ appropriate for \\underline{all} big bang models in the first quadrant of the $\\OM$--\\,$\\OL$ plane. Because the incomplete Legendre elliptic integrals have branch points, more than one expression is necessary to completely cover this domain, \\eg see (\\ref{age_A1new}) and (\\ref{age_A1}). These new expressions should be quite useful for everything from gravitational lensing to high $z$ evolution studies. As an example \\cite{MP,SR,RP} and \\cite{DT} all estimate event rates of supernovae at very high ($z>1$) redshifts, given an observed star formation rate. Computing such event rates for any choice of ($\\Omega_m$,$\\Omega_\\Lambda$) requires the transformation $\\tau(z)$. We have made similar estimates of event rates and find that our computations are reduced from hours down to minutes when our analytic $\\tau(z)$ is used.\\footnote{ FORTRAN 90 and {\\it Mathematica} implementations of the results presented here are available at http://www.nhn.ou.edu/\\~{ }thomas/z2t.html.} In \\S 2 we present our results and in \\S 3 some conclusions. ", "conclusions": "We have given valid analytic expressions for $\\tau(z)$ in FLRW, the age of the Universe as a function of redshift, which are relatively simple and are quite useful when a fast computer implementation is needed.\\footnote{FORTRAN 90 implementation of the results presented here and available at http://www.nhn.ou.edu/$\\sim$thomas/z2t.html are 20-40 times faster than a traditional Bulirsch-Stoer integrator \\cite{PTVF}.} These expressions completely cover the big bang models of the first quadrant of the ($\\OM$,\\,$\\OL$) plane. If lookback times are wanted they can additionally be obtained from results given here by simply evaluating $\\tau(0)-\\tau(z)$. Readers that are interested in adding radiation pressure as a source of gravity should see \\cite{AA} and \\cite{DM} and cited references. Even though we give several expressions for $\\tau(z)$, most of the ($\\OM$,\\,$\\OL$) plane, which includes currently favored values, is covered by case A1, \\ie result (\\ref{age_A1new}) or (\\ref{age_A1}). If $\\tau(z)$ for the flat model, $\\OO=1$, is wanted, the simpler result (\\ref{age_B1}) should be used. Results for $\\OO\\ne 1$, (\\ref{age_A1new}), (\\ref{age_A1}), and (\\ref{age_A2}), appear complicated because of the presence of extra constants, \\eg $A$ and $y_1$ that have been retained to compactify formulas. The reader should keep in mind that these are simply constants that depend on ($\\OM$,\\,$\\OL$) through the single combination $b$ of (\\ref{b}). We could have eliminated these auxiliary constants and given $\\tau(z)$ directly in terms of the two parameters $\\OM$ and $\\OL$; however, such expressions would take up more than a page. Expressions (\\ref{age_A1new}) and (\\ref{age_A1}) for $\\tau(z)$ remain real but as presented can contain imaginary terms because of branch points. The threshold is defined by $1-\\alpha^2\\sin^2\\phi_z=0$ in $\\Pi(\\phi_z,\\alpha^2,{\\rm k})$. If $1-\\alpha^2\\sin^2\\phi_z<0$ canceling imaginary terms appear in $\\Pi(\\phi_z,\\alpha^2,{\\rm k})$ and the logarithm. For expressions that avoid this imaginary complication the reader simply replaces the argument of the logarithm with its magnitude and $\\Pi(\\phi_z,\\alpha^2,{\\rm k})$ with its principal part. At threshold points where $1-\\alpha^2\\sin^2\\phi_z=0$, canceling infinities appear in $\\Pi(\\phi_z,\\alpha^2,{\\rm k})$ and the logarithm. The infinity problem is avoided by switching between (\\ref{age_A1new}) and (\\ref{age_A1}). If both have infinities then (\\ref{age_B3}) gives the correct result." }, "0003/astro-ph0003180_arXiv.txt": { "abstract": "During several observations in 1999 August with RXTE of the low-mass X-ray binary GX\\,3+1, we found a single short and strong X-ray burst. This is the first burst from GX\\,3+1 which clearly shows evidence for radius expansion of the neutron-star photosphere during the thermo-nuclear runaway. We show that the cooling phase of the neutron star photosphere starts already just before the end of the contraction phase. Considering the fact that the radius expansion is due to the burst luminosity being at the Eddington luminosity, assuming standard burst parameters and accounting for gravitational redshift effects we derive a distance to the source of $\\sim$4.5\\,kpc, although relaxing these assumptions may lead to uncertainties up to $\\sim$30\\%. By comparing the persistent flux with that observed at the peak of the burst we infer that near the time of the X-ray burst the persistent luminosity of GX\\,3+1 is $\\sim$0.17\\,L$_{\\rm edd}$, confirming predictions from theoretical modeling of X-ray spectra of bright sources like GX\\,3+1. ", "introduction": "The overall X-ray intensity of the low-mass X-ray binary (LMXB) GX\\,3+1 varies slowly on time scales of months to years (Makishima et al.\\ 1983; Asai et al.\\ 1993, see also Fig.~1). X-ray bursts in GX\\,3+1 were discovered by {\\em Hakucho}, at a time when the persistent X-ray flux was about half of that seen previously (Makishima et al.\\ 1983). During that time roughly one burst per day was observed. The bursts from this source were shown to be thermonuclear flashes on the neutron star surface, i.e.\\ being of type I (Makishima et al.\\ 1983; Asai et al.\\ 1993: {\\em Ginga}; Molkov et al.\\ 1999: {\\em GRANAT/ART-P}), but none of them showed evidence for photospheric radius expansion. GX\\,3+1 is one of the four brightest so-called ``atoll'' sources (Hasinger \\&\\ van der Klis 1989). The sources in this group (including GX\\,13+1, GX\\,9+1 and GX\\,9+9) hardly show any X-ray bursts (if at all), and display properties like those of other atoll sources when these are in their high accretion rate state: their tracks in X-ray colour-colour diagrams are long, diagonal and slightly curved, while their fast timing properties are at all times dominated by a relatively weak (1--4\\%\\ rms) power-law shaped noise component. Detailed X-ray spectral modeling seems to suggest that they accrete with rates near 10\\%\\ of the Eddington mass accretion rate, i.e.\\ intermediate between the more frequently bursting atoll sources and that of the so-called ``Z'' sources (Psaltis \\&\\ Lamb 1998). At low accretion rates (and therefore probably low intensities) such sources are predicted to display the properties characteristic of the more frequently bursting atoll sources, which in view of the Hakucho result (see above) at least GX\\,3+1 seems to satisfy. During one of our series of target of opportunity observations with RXTE aimed at observing GX\\,3+1 at low intensities, we observed a strong ($\\sim$2.3\\,Crab [2--10\\,keV] at maximum) and short (15--20\\,s) X-ray burst. The burst onset occurred on 1999 August 10, 18:35:53.5~UTC. In this paper we discuss its properties. ", "conclusions": "The light curve and X-ray spectral behaviour of the X-ray burst in GX\\,3+1 observed with RXTE show clear evidence for radius expansion of the neutron star photosphere due to near-Eddington luminosities during a themonuclear runaway on the neutron star surface (for a review see e.g.\\ Lewin et al.\\ 1993). The total time for the expansion and contraction phase is only $\\sim$1.5\\,s, during which the radius expanded only by a factor of $\\sim$2. Such short bursts with small expansion phases have been seen in other bright X-ray sources, such as Cyg\\,X-2 (Smale 1998). The gradual softening at the end of the burst is attributed to cooling of the neutron star surface, which is characteristic for type-I bursts (Hoffman et al.\\ 1978). During the burst our derived black-body temperatures are smaller, whereas our inferred black-body radii (all at the same assumed distance) are larger, than reported for previous GX\\,3+1 bursts (Makishima et al.\\ 1983, see also Inoue et al.\\ 1981; Asai et al.\\ 1993; Molkov et al.\\ 1999). The burst parameter values for $\\gamma$ ($\\sim$0.10--0.20) and $\\tau$ (4--8\\,s) quoted by Asai et al.\\ (1993), and inferred from the observations presented by Makishima et al.\\ (1983) and Molkov et al.\\ (1999), are similar to our findings. We note (see also Asai et al.\\ 1993) that $\\gamma$, $\\tau$ and our estimate of $\\alpha$ fall on the extreme end of relations between $\\tau$ vs.\\ $\\gamma$ and $\\alpha$ vs.\\ $\\gamma$ as presented by Van Paradijs et al.\\ (1988) for typical type~I bursters. This shows that if bright sources burst, the burst duration tends to be short (order of 10\\,s; note however, that some bursts in the bright ``Z'' source GX\\,17+2 have a duration of the order of minutes, see e.g.\\ Kuulkers et al.\\ 1997 and references therein). It is interesting to note that our X-ray burst from GX\\,3+1 is very similar to the radius expansion burst seen in Cyg\\,X-2 with RXTE in most of its facets, except notably for the $\\gamma$ being a factor 4.3 larger for Cyg\\,X-2 (Smale 1998). Note also that during the burst from Cyg\\,X-2 no evidence for pulsations was reported, similar to what we conclude for GX\\,3+1, both with upper limits on the modulation strength which are significantly lower than for bursts during which oscillations were seen (see Strohmayer 1998, 2000). \\begin{figure} \\psfig{figure=Cb211_f4.ps,bbllx=50pt,bblly=75pt,bburx=566pt,bbury=626pt,angle=-90,width=7.2cm} \\caption{Bolometric black-body flux (F$_{\\rm bol}$) versus black-body temperature (T$_{\\rm bb}$) for the first 14\\,s of the burst. Data points are connected for clarity. The dotted line represents the fit to the cooling track of the burst, see text. Note that T$_{\\rm bb}$ runs from right to left.} \\end{figure} A convenient way to display the burst properties as they vary in time, is a flux-temperature diagram, see Fig.~4. In such a diagram the phases of expansion/contraction and subsequent cooling of the neutron star photosphere are distinguished by two separate tracks (see e.g.\\ Lewin et al.\\ 1993). GX\\,3+1 moves from the middle bottom to top left (rising phase), top middle (expansion phase), back towards top left (contraction phase), and finally to the lower right part of the diagram (cooling phase). We can adequately fit ($\\chi^2_{\\rm red}$/dof = 0.9/34) $\\log{{\\rm F}_{\\rm bol}}$ versus $\\log{{\\rm T}_{\\rm bb}}$ during the cooling phase of the burst by a straight line with a slope of 3.97$\\pm$0.15 (dotted line in Fig.~4). This means that F$_{\\rm bol}$ is consistent with being proportional to T$_{\\rm bb}^4$, which indicates that the neutron star photosphere radiates as a black-body during the cooling phase, at a constant radius R$_{\\rm bb}$. We note that burst spectra are generally not described by pure black-body radiation, especially near the Eddington limit (see Lewin et al.\\ 1993, and references therein). Instead the black-body radiation is modified mainly at energies below $\\sim$3\\,keV and above $\\sim$10\\,keV. Since our burst spectra are analysed in the 2.9--20\\,keV energy range, we are, therefore, not greatly affected by modified black-body radiation. Note that we then probably underestimate our bolometric fluxes. At the start of the expansion phase the black-body bolometric flux and temperature values do not match those at the end of the contraction phase. Our estimated emission areas are the same at these instants; the above then means that the photosphere is cooler at the end of the contraction phase with respect to the start of the expansion phase. From Fig.~4 we see that F$_{\\rm bol}$ drops below the constant peak flux before the end of the contraction phase. We infer that the cooling phase therefore already started before the end of the contraction phase. Using the fact that during the expansion and contraction phase of the neutron star photosphere the burst luminosity equals the Eddington luminosity one can get an estimate of the distance (see e.g.\\ Lewin et al.\\ 1993). Assuming standard burst paramaters (isotropy, cosmic abundances and a canonical neutron star mass of 1.4\\,M$_{\\sun}$) and taking into account gravitational redshift effects we find $d=4.5\\pm 0.1$\\,kpc. If we assume a neutron star mass of 2.0\\,M$_{\\sun}$ we instead find $d=5.1\\pm 0.1$\\,kpc. For bright sources like GX\\,3+1 most of the hydrogen content is being burned persistently, so during the expansion/contraction phase the neutron star atmosphere is likely to lack hydrogen. Using the Eddington luminosity appropriate for hydrogen-poor matter then leads to $d=6.1\\pm 0.1$\\,kpc. Dropping only our assumption that the burst radiates isotropically, and assuming anisotropy values of $0.5< \\xi <2$ (e.g.\\ van Paradijs \\&\\ Lewin 1987), we derive distances between 3--7\\,kpc. On the other hand, if the peak luminosities during radius expansion bursts are standard candles we can use the mean peak luminosity for such bursts seen in globular cluster sources for which the distances are known, i.e.\\ $3.0 \\times 10^{38}$\\,erg\\,s$^{-1}$ (Lewin et al.\\ 1993). In this case we derive $d\\sim 5.6$\\,kpc. These distance estimates show that in principle one can get an idea of the distance to the source, but the exact value still remains rather uncertain by $\\sim$30\\%. The persistent flux during our observations is the same within a factor of $\\sim$2 with respect to the previous reports when GX\\,3+1 was bursting, i.e.\\ low ($\\sim$0.2\\,Crab). Using the fact that during the peak of the burst the observed (net-burst) luminosity is at near Eddington values we can now for the first time estimate the persistent flux in terms of the Eddington luminosity for the bright atoll sources like GX\\,3+1 (i.e.\\ GX\\,13+1, GX\\,9+1 and GX\\,9+9). For GX\\,3+1 we find L$_{\\rm pers}\\simeq 0.17$\\,L$_{\\rm edd}$ (assuming the burst and persistent emission is radiated in the same directions). This is consistent with that inferred from models of X-ray spectra, i.e.\\ $\\sim$0.1\\,L$_{\\rm Edd}$ (Psaltis \\&\\ Lamb 1998). GX\\,13+1 has been seen to burst sporadically (Matsuba et al.\\ 1995), whereas no bursts have been reported for GX\\,9+1 and GX\\,9+9. This may mean that GX\\,3+1 and GX\\,13+1 are accreting near to the critical mass accretion rate at which bursts cease to occur, whereas GX\\,9+9 and GX\\,9+1 accrete above this limit. However, this does not explain the fact that some sources that are accreting at even higher rates (near Eddington), i.e.\\ Cyg\\,X-2 and GX\\,17+2, also irreglularly show bursts." }, "0003/astro-ph0003149_arXiv.txt": { "abstract": "I will discuss the presence of massive star clusters in starburst galaxies with an emphasis on low mass galaxies outside the local group. I will show that such galaxies, with respect to their mass and luminosity, may be very rich in young luminous clusters. ", "introduction": "During the last decade, the study of young massive stellar clusters, sometimes referred to as super star clusters (SSCs) has seen rapid progress. From a few known examples of ``blue populous clusters'' in the Large Magellanic Cloud (LMC) and SSCs in the star forming dwarf galaxies NGC~1705 (Melnick, Moles, \\& Terlevich 1985) and NGC~1569 (Arp \\& Sandage 1985), the numbers have grown, largely thanks to HST. SSCs have now been found in a variety of different environments, as have globular clusters (GCs) which are {\\it old} massive clusters. In this paper I will review the massive cluster content of, in particular, low mass starburst galaxies like blue compact galaxies (BCGs). A review of massive clusters in starbursts naturally becomes biased towards young SSC like objects. However, such galaxies may indeed contain also rich populations of older clusters. The study of SSCs in starburst galaxies gained momentum with the advent of the HST. Ultraviolet imaging with the aberrated HST/FOC (e.g. Meurer et al. 1995, Conti and Vacca 1994) of starbursts revealed that a significant fraction of the star formation activity took place in SSCs. Meurer et al. (1995) studied nine galaxies, finding SSCs in most of them. Several other starbursts have been imaged with the aberrated HST/FOC (e.g. Conti et al., unpublished) and SSCs are frequently encountered. Optical imaging with the aberrated HST has also discovered SSCs in many low mass starburst galaxies (e.g. Hunter et al. 1994). The greater sensitivity of WFPC2 as compared to FOC has multiplied the number of detected clusters in ESO~338-IG04 (\\\"Ostlin, Bergvall, \\& R\\\"onnback 1998) and He~2-10 (Johnson et al. 2000). In giant galaxies there seems to be several ways to form SSCs/GCs, e.g. mergers (see Miller 2000), and bars (e.g. Kristen et al. 1997) and circum-nuclear rings in spiral galaxies (e.g. Barth et al. 1995). There are even indications of SSC formation in the discs of normal spirals (Larsen \\& Richtler 1999). In low mass galaxies some of these mechanisms, e.g. formation of SSCs in bar and resonant induced density enhancements, are not available. Mergers are certainly producing SSCs in some low mass galaxies (e.g. ESO~338-IG04, ESO~350-IG38, ESO~185-IG13), but there might be other mechanisms too. The origin of active star formation in galaxies like NGC~1569 and NGC~1705 are not yet well understood. There are also dwarf stabursts which do not contain luminous SSCs (e.g. IC10, see Grebel 2000). Cluster destruction mechanisms (e.g. due to tidal shocks) are weaker in low mass galaxies giving SSCs a greater chance of survival. In ESO~338-IG04 the dominant GC population is $\\sim 3 $ Gyr old, and this population alone is enough to classify the galaxy as GC rich in terms of specific frequency. Low mass galaxies are in general metal-poor (typically [O/H]$\\sim -1$, Kunth \\& \\\"Ostlin 2000), which make them suitable for comparison with high redshift conditions and early GC formation. Dwarf galaxies have certainly been important ingredients in the hierarchial buildup of massive galaxies. Another virtue is that internal extinction in general is small, which for instance makes age dating more secure. Thus, dwarf starbursts are good places to investigate the formation and evolution of massive star clusters. Even if one cannot be sure whether a SSC will evolve into a bona fide GC or dissolve, a proto-GC must look very much like a SSC (Kennicutt \\& Chu 1988) and the collective formation of a few SSCs will, by necessity, be associated with a starburst. Thus populations of GCs and survived SSCs trace former starbursts, and may be used to study the evolution of galaxies. For example, if the excess of faint blue galaxies seen in deep optical counts is due to starbursts originating in merging galaxies at intermediate redshift, one would expect these to form significant numbers of SSCs/GCs which should be visible as intermediate age GC populations in local galaxies. ", "conclusions": "Among dwarf and low mass galaxies we encounter both galaxies that appear to be very efficient formers of massive clusters, and galaxies that appear totally devoid of such objects. Some luminous BCGs have specific frequncies of luminous ($M_V \\le -11$) SSCs $S_{11} \\ge 1$, which is an order of magnitude larger than most of the prototypical SSC factories: the ``Antennae'', NGC~7252 and NGC~1275. Thus there is a tendency for $S_{11}$ to increase when going to low mass starburst galaxies. A similar trend has been found for the specific frequency of GCs among dwarf ellipticals (Miller et al. 1998). The purpose of comparing $S_{11}$ values was to show that low mass, metal-poor, starburst galaxies are excellent hunting grounds for luminous SSCs, and in addition problems with extinction are much smaller than in giant mergers. The BCGs with the highest $S_{11}$ are believed to be the product of dwarf galaxy mergers (\\\"Ostlin et al. 1999, 2000). A possible explanation to the higher $S_{11}$ values is that the starburst timescales are shorter in systems with lower mass, whereas in a giant merger one expects a more extended starburst. There are also BCGs and low mass starbursts which do not contain luminous SSCs. Although the expected number of SSCs in galaxies of very low luminosity will always be small and subject to statistical fluctuations, there migh be a connection to the triggering mechanism of the starbursts. Merging dwarfs might be more efficient SSC formers than non-merging ones. A couple of BCGs (ESO~338-IG04 and Mrk~996) in addition contains rich populations of older GCs. There is no a priori reason to believe that old GC systems are intrinsically rare among BCGs. Rather few BCGs have been studied at sufficient depth and spatial resolution to unveil faint old GCs. The properties of relatively old GCs in BCGs may provide important information about the nature of the host galaxy. An unbiased survey of star forming dwarf galaxies with HST to characterize the frequency of star clusters, their colors and host galaxy properties, would allow to quantitatively study cluster formation in low mass galaxies. The Malkan et al. (1998) survey do not fullfil these criteria but show that such a program could be very rewarding. A better understanding of the ultimate fate of SSCs is also required. Dynamical mass estimates are still rare and often result in masses of the right order of magnitude but surprisingly low mass to light ratios (see Smith \\& Gallagher 2000). If this is due to flat or top heavy IMFs it would make it harder for young SSCs to survive and become GCs." }, "0003/astro-ph0003075_arXiv.txt": { "abstract": "Based on detailed 2D radiation hydrodynamics (RHD) simulations, we have investigated the effects of photospheric temperature inhomogeneities induced by convection on spectroscopic determinations of the lithium abundance. Computations have been performed both for the solar case and for a metal-poor dwarf. NLTE effects are taken into account, using a five-level atomic model for Li~I. Comparisons are presented with traditional 1D models having the same effective temperature and gravity. The net result is that, while LTE results differ dramatically between 1D and 2D models, especially in the metal-poor case, this does not remain true when NLTE effects are included: 1D/2D differences in the inferred NLTE Li abundance are always well below 0.1 dex. The present computations still assume LTE in the continuum. New computations removing this assumption are planned for the near future. ", "introduction": "It would be an offense to the audience here to pretend to explain why it is important to determine accurately the abundance of $^6$Li and $^7$Li in the oldest stars. In this respect, we have nothing to add to the exposition by Fran\\c{c}ois Spite (this volume). We will right away mention the two major problems which may cast doubts on our real knowledge of the actual initial abundance of Li in the oldest stars, and consequently in the primordial matter. (i) While standard models of the internal structure of metal-poor dwarfs do not deplete $^7$Li, more sophisticated models including rotationally induced mixing (Pinsonneault et al.\\ 1992) have predicted that the measured abundance in the photosphere is 5 to 10 times less than the initial abundance representative of Big Bang material. (ii) On top of that, Kurucz (1995) claimed that the hot and cold convective structures produce large effects in metal-poor stellar photospheres, where the convection zone reaches the line formation layers. The claimed effect is an overionization of Li by a factor of 10, leading to an underestimation of the abundance of Li when derived from the resonance line of Li~I ($\\lambda$~670.8~nm) in the usual way. If these two statements are correct, the true abundance of Li in primordial matter is 50 to 100 times higher than the value derived from 1D, LTE models of halo subdwarfs so far. The first factor of 5 to 10 has been discussed in a previous paper by Ryan (this symposium), and shown to be likely much smaller, of the order of 1 to 1.4. We shall not come back to this point, which we consider as very well treated. Before this symposium, a single paper (Asplund et al.\\ 1999) has dealt with the question of the other factor of 10 claimed by Kurucz (1995), whose arguments were based on a simplified two-column model. In contrast, the work by Asplund et al.\\ relies on realistic 3D hydrodynamical models, similar to the simulations of the solar granulation (Stein \\& Nordlund 1998), but with parameters appropriate for two metal-poor stars: HD 140283 and HD 84937, both subgiants. The computation of the lithium resonance line was made under the assumption of LTE, and the correction to be applied to the Li abundance derived from standard 1D models was found to be large, of the order of -0.2 to -0.35 dex. Note that these corrections have the opposite sign as Kurucz's prediction! However, Kiselman (1997, 1998) had shown, in the solar case, that NLTE and LTE computations lead to significantly different values of equivalent widths of the Li~I $\\lambda$~670.8~nm line over hot and cold structures (see Fig.~3 of his 1997 paper, top panel). For this reason, we decided to undertake NLTE radiation hydrodynamics computations for the case of a metal-poor star, and we report here on the results of this investigation. In the next section we recall former work related to simulations of the solar granulation, a useful benchmark for checking the theory, but not directly applicable to metal-poor stars. In section~3 we describe the assumptions underlying the construction of the 2D RHD models used for the spatially resolved computation of the lithium resonance line. Section~4 gives the description of the NLTE treatment of the Li atom, and section~5 summarizes our results and compares them to those presented by M. Asplund (this symposium). Finally, our conclusions are listed in section~6. ", "conclusions": "1. The statement of Kurucz (1995) that abundances of lithium derived from standard 1D models of metal-poor stellar atmospheres is too small by a factor of 10 is not supported by actual multidimensional NLTE computations. Even the sign of the correction is doubtful, and the error is well below 0.1 dex, both according to our investigation and the one presented by M. Asplund on this symposium. \\\\ \\noindent 2. LTE abundance determinations based on inhomogeneous atmospheres are strongly discouraged. They produce large ``granulation abundance corrections'' due to non-linear effects in the direction opposite to Kurucz' prediction, but the actual NLTE line formation mechanism couples the population of the atomic levels more closely to the mean radiation field than to the local temperature. \\\\ \\noindent 3. The combination of multidimensional models with NLTE line formation for the Li~I $\\lambda$~670.8 nm resonance line leads to the same lithium abundance as that derived from NLTE analysis with flux-constant 1D models, abundance differences being less than 0.1 dex. However, this result must not be hastily generalized to other atoms with a different atomic structure. \\\\ \\noindent 4. An obvious future improvement is to extend the NLTE analysis to the continuum, which has be assumed here to be in LTE. We plan to do that in the near future. If it turns out that the H$^-$ ion is affected by NLTE, this will raise a new question: should such effects be included already in the radiation hydrodynamics code, which determines the amplitude of the thermodynamical fluctuations?" }, "0003/astro-ph0003243_arXiv.txt": { "abstract": "For all sources in which the phenomenon of kilo-Hertz quasi-periodic oscillation (kHz QPO) is observed, the QPOs disappear abruptly when the inferred mass accretion rate exceeds a certain threshold. Although the threshold cannot at present be accurately determined (or even quantified) observationally, it is clearly higher for bright Z sources than for faint atoll sources. Here we propose that the observational manifestation of kHz QPOs {\\em requires} direct interaction between the neutron star magnetosphere and the Keplerian accretion disk and that the cessation of kHz QPOs at high accretion rate is due to the lack of such an interact when the Keplerian disk terminates at the last stable orbit and yet the magnetosphere is pushed farther inward. The threshold is therefore dependent of the magnetic field strength --- the stronger the magnetic field the higher the threshold. This is certainly in agreement with the atoll/Z paradigm, but we argue that it is also generally true, even for individual sources within each (atoll or Z) category. For atoll sources, the kHz QPOs also seem to vanish at low accretion rate. Perhaps the ``disengagement'' between the magnetosphere and the Keplerian disk also takes place under such circumstances, because of, for instance, the presence of quasi-spherical advection-dominated accretion flow (ADAF) close to the neutron star. Unfortunately, in this case, the estimation of the accretion rate threshold would require a knowledge of the physical mechanisms that cause the disengagement. If the ADAF is responsible, the threshold is likely dependent of the magnetic field of the neutron star. ", "introduction": "The detection of kilohertz quasi-periodic oscillation (kHz QPO) in low-mass X-ray binaries is arguably the greatest discovery that the {\\it Rossi X-ray Timing Explorer} (RXTE) has made to date. Such signals almost certainly originate in the immediate vicinity of central neutron stars, given that the fastest oscillations are observed to occur on dynamical timescales near such objects (review by van der Klis 2000). The prospect of using kHz QPOs to probe the effects of strong gravity near neutron stars is therefore very exciting. Various models (e.g., Klein et al. 1996; Miller et at. 1998; Stella \\& Vietri 1999; Osherovich \\& Titarchuk 1999) have been proposed to explain kHz QPOs that, in nearly all cases, come in pairs (note that we choose not to discuss the QPOs observed during thermonuclear bursts). Except for the ``photon bubble model'' (Klein et al. 1996), all other models invariably associate one of the pair to the Keplerian motion of clumps of matter or ``hot spots'' at the inner edge of a geometrically thin accretion disk (although the ``sonic point model'' differs in detail, e.g., Miller et at. 1998; however, see Lai 1998). The inner edge of the disk is determined by the pressure balance between the accreted matter and the magnetic field of the neutron star. As the mass accretion rate increases, the ram pressure of the accreted matter increases, which squeezes the magnetosphere more and thus the accretion disk extends farther toward the neutron star. Therefore, we expect that the frequency of the ``Keplerian QPO'' increases with the accretion rate, which agrees with the observations (van der Klis 2000). Since no Keplerian flow can exist inside the last stable orbit, we expect that there is an upper limit to the frequency of this QPO --- any further increase in the accretion rate cannot result in an increase in the frequency of the QPO. Observing this limit would provide strong observational evidence for the presence of the last stable orbit around neutron stars, a natural consequence of strong gravity. However, due to the lack of detailed knowledge on how the signals are produced in the first place, it is still not clear how kHz QPOs would behave when the accretion disk reaches the last stable orbit. Without considering any physical mechanisms responsible for modulating the X-ray emission, one might take it for granted that the frequency of the Keplerian QPO would saturate at sufficiently high accretion rate (e.g., Zhang et al. 1998b; Lai 1998). On the other hand, Cui et al. (1998) emphasized the importance of disk-magnetosphere interaction in producing the QPOs, based on a detailed study of the evolution of a kHz QPO observed of Aquila X-1, a transient atoll source, throughout the rising phase of an X-ray outburst. In this Letter, we generalize the ideas proposed by Cui et al. to all kHz QPO sources. ", "conclusions": "The observations of the kHz QPO phenomenon seem to suggest that the interaction between the Keplerian accretion disk and the magnetosphere of the neutron star is directly responsible for modulating the X-ray emission. For a given source, the presence (or absence) of such an interaction dictates the appearance (or disappearance) of the QPOs. At high mass accretion rate, we argue that the presence of the last stable orbit manifests itself in the disappearance of the QPOs, as opposed to the saturation in the QPO frequency. The difference may only seem semantic since, quantitatively, both interpretations require that the neutron star is inside the last stable orbit. However, we feel that it is imperative for the models to begin to address such critical issues as modulation mechanisms for kHz QPOs, in light of the ever improving quality of the data. One critical question is whether the neutron star magnetosphere can be disengaged from the Keplerian disk at the last stable orbit. Studies have shown that the evolution of the magnetic field configuration is very complicated as the disk approaches the last stable orbit (e.g., Lai 1998), but the exact solution is not known at present. Intuitively, as the accretion process proceeds from the inner edge of the Keplerian disk onto the surface of the neutron star, the ram pressure of the accreted matter continues to push the magnetosphere inward. In this case, the magnetosphere only interacts with the non-Keplerian accretion flow in the ``gap'' between the last stable orbit and the neutron star surface, but {\\em not} directly with the Keplerian flow in the disk. The importance of such gap accretion has been studied extensively (Klu\\`{z}niak \\& Wagoner 1985; Klu\\`{z}niak \\& Wilson 1991). The situation is less certain at low mass accretion rates. In fact, observationally it can still be argued whether the QPOs actually disappear, given that in nearly all cases the upper limits derived are comparable to the fractional rms amplitudes of the QPOs measured at high accretion rates (M\\'{e}ndez 2000, private communication). In the case of 4U 0614+09, however, the 95\\% upper limit is only about half of the measured amplitude when the source is bright (M\\'{e}ndez et al. 1997). Therefore, we have at least one source in which the QPOs, if present at all, are definitely much weaker at low accretion rates (i.e., below $\\dot{M}_l$). Moreover, we note that often the kHz QPOs {\\em strengthen} (relative to the average source intensity) as the accretion rate {\\em decreases} (Wijnands et al. 1997; Wijnands \\& van der Klis 1997; Smale, Zhang, \\& White 1997). If the QPOs do disappear at low accretion rate, some physical process, like the ADAF, could be present to truncate the Keplerian disk at large distance from the neutron star under such circumstances. This would destroy the disk-magnetosphere interaction and thus the kHz QPOs for atoll sources. The process may also operate in Z sources: the persistence of the kHz QPOs in such cases can be attributed to a lower accretion rate threshold that is due to a stronger magnetic field of the neutron star. Alternatively, the process bears no relevance to Z sources. The disk-magnetosphere interaction is always present at low accretion rates, so are the QPOs." }, "0003/astro-ph0003419_arXiv.txt": { "abstract": "A simple model of chemical enrichment in cluster early-type galaxies is presented where the main parameters driving the formation of the stellar component are reduced to four: infall timescale ($\\tau_f$), formation epoch ($z_F$), star formation efficiency ($\\ceff$) and fraction of gas ejected in outflows ($\\bout$). We find that only variations in $\\bout$ or $\\ceff$ can account for the colour-magnitude relation, so that the most luminous galaxies had low values of ejected gas and high efficiencies. Less massive galaxies can be either related to a lower star formation efficiency ($\\ceff$-sequence) or to an increased outflow rate ($\\bout$-sequence). The combination of chemical enrichment tracks with population synthesis models (Bruzual \\& Charlot 2000) is used to explore the correlation between mass-to-light ratios and masses. A significant slope mismatch is found between stellar and total $M/L$ ratios, which cannot be explained by an age spread and implies a non-linear correlation between total and stellar mass: $M_{\\rm TOT} \\propto M_{ST}^{1.2}$. The sequences driven by star formation efficiency ($\\ceff$) and outflows ($\\bout$) are shown to predict different trends at high redshift. The variation with redshift of the slope of the fundamental plane will increase significantly in the efficiency sequence --- driven by age --- and will slightly decrease in the outflow sequence --- driven by metallicity. The evolution of the zero point is similar in both cases and within the observational errors of current observations. Measurement of the dependence of the tilt of the fundamental plane on redshift will break the degeneracy between outflows and star formation efficiency, which will enable us to determine whether the colour-magnitude relation is controlled by age or metallicity. ", "introduction": "The process of galaxy formation and evolution can be explored in two complementary ways: A ``forwards'' approach takes into account the physics underlying the most basic processes of structure evolution and star formation and --- after finding a suitable set of initial conditions --- evolves the system forward so that the final output is compared with observations. This is the philosophy behind N-body simulations or semi-analytic modelling (e.g. Baugh et al. 1998; Kauffmann \\& Charlot 1998). On the other hand, a ``backwards'' approach simplifies the physics behind galaxy formation and evolution to a phenomenological problem comprising a reduced set of parameters, using local observations as constraints. Then the system is evolved backwards so that the predictions for a given set of parameters are compared with observations at moderate-to-high redshifts (e.g. Bouwens, Broadhurst \\& Silk 1998a,b). This ``brute force'' method allows one to search a reasonable volume of parameter space, throwing light on processes, such as star formation, which are otherwise too complicated to tackle from basic principles. Cluster early-type galaxies are ideal candidates for a comparison between models and observations. Significant samples of these galaxies can be found over a large redshift range. Furthermore, the tight observed correlations such as the colour-magnitude relation or the fundamental plane can be used as powerful constraints on a phenomenological backwards approach. The list of observed clusters at moderate to high redshift is quite large (e.g. Dressler et al. 1999 ; Stanford, Eisenhardt \\& Dickinson 1998; Oke, Postman \\& Lubin 1998; Van Dokkum 1999) and ever increasing (e.g. Yee et al. 1999). Clusters observed at redshifts $z\\sim 1-1.5$ yield valuable information about the epoch of star formation, pushing it to very high redshift ($z_F\\ga 3$). Unfortunately, the direct spectrophotometric determination of the star formation history is hampered by the age-metallicity degeneracy (Worthey 1994; Ferreras, Charlot \\& Silk 1999) which allows the variations of most of the spectrophotometric observables to be explained either by a range of ages or metallicities. Broadband photometry is strongly dependent both on age and metallicity but even spectral indices targeting single lines such as Balmer absorption or magnesium abundance can change both with age and metallicity. Not surprisingly, a similar age estimation technique based on spectral indices applied to similar sets of elliptical galaxies yields contradictory results: Kuntschner (2000) and Kuntscher \\& Davies (1998) find coeval stellar populations in Fornax cluster ellipticals so that the colour range is explained by a metallicity sequence. On the other hand, the sample of field and group ellipticals observed by Gonz\\'alez (1993) and further analysed by Trager et al. (2000) presents a relatively large spread in ages. Hence, the issue of the stellar age distribution in galaxies still requires the aid of modelling. We will show that incorporating chemical enrichment allows one to potentially solve the age-metallicity degeneracy problem. The next three sections describe our chemical enrichment model and the meaning of the reduced set of parameters used to trace the star formation history in cluster ellipticals. \\S5 deals with the comparison of predicted and observed mass-to-light ratios and its use at high redshift to discriminate between a mass sequence driven by age or metallicity. Finally in \\S6 we discuss the predictions and list the main conclusions. ", "conclusions": "A simple phenomenological treatment is described in this paper, where the mechanisms underlying the evolution of galaxies are reduced to a set of four parameters. The star formation rate is assumed to follow a linear Schmidt law whose proportionality constant is used to describe a varying star formation efficiency ($\\ceff$). The supply of primordial gas fuelling star formation is controlled by gaussian infall characterised by the epoch at which the infall rate is maximum ($z_F$), and the width of the gaussian profile gives a characteristic infall timescale ($\\tau_f$). The model is allowed to eject a fraction ($\\bout$) of the enriched gas, thereby lowering the effective yield. A first stage in this analysis involves finding a suitable pair of infall parameters ($z_F$,$\\tau_f$) which are capable of generating the restframe $U-V$ colours of the brightest cluster galaxies. The colour constraints imposed by the reddest (and brightest) galaxies in clusters at moderate and high redshift allow us to discard long infall timescales and recent star formation epochs. One could argue that the constraint on the star formation history of the brightest systems need not be the same as for less massive ellipticals. However, the current most plausible scenarios for galaxy formation assume either a simultaneous process of star formation regardless of galaxy mass, or a hierarchical structure where the most massive galaxies might have undergone the {\\sl latest} bursts of star formation. Furthermore, the possibility of an ``inverted-hierarchical'' scenario can still be accomodated in this model, as long as the constraints imposed ($z_F\\ga 3$,$\\tau_f\\ga 1$ Gyr) are held even for the less massive galaxies. Out of the four parameters considered in the model, we have found that --- within the framework of this model --- only the SF efficiency ($\\ceff$) and the ejected fraction in outflows ($\\bout$) help determine the mass sequence in cluster early-type galaxies. Infall parameters ($z_F$,$\\tau_f$) do not appreciably change much the output unless very recent stages of star formation are included. However, this will result in restframe $U-V$ colours that are in contradiction with observations. The efficiency parameter generates a significant spread in the age distribution of stars, although mixed with a metallicity range (any model with a reasonable IMF must include this range of abundances). Alternatively, a range of outflow rates result in a range of metallicities with no significant spread in ages. Both mechanisms are degenerate in local clusters because the average age of the stellar populations predicted for any model at $z=0$ are too old to be able to disentangle the effects of age and metallicity. However, the predictions of age-sensitive observables at high redshift differ noticeably for these two sequences. Unfortunately, present data is not capable of ruling out one model against the other, but the continuing flow of data from clusters at moderate-to-high redshift will eventually enable us to break this degeneracy. A sequence driven by efficiency ($\\ceff$ sequence) predicts a steepening of the slope of the fundamental plane, whereas a $\\bout$ sequence --- driven by outflows --- predicts no change or a slight decrease of this slope. High precision observations of the dynamical and spectrophotometric properties of cluster galaxies at high redshift will confirm the importance of either age or metallicity in the mass range of cluster ellipticals." }, "0003/astro-ph0003305_arXiv.txt": { "abstract": "In this paper we present a two-dimensional, time dependent, numerical simulation of a reconnection current layer in incompressible resistive magnetohydrodynamics with uniform resistivity in the limit of very large Lundquist numbers. We use realistic boundary conditions derived consistently from the outside magnetic field, and we also take into account the effect of the back pressure from flow into the the separatrix region. We find that within a few Alfv{\\'e}n times the system evolves from an arbitrary initial state to a steady state consistent with the Sweet--Parker model, even if the initial state is Petschek-like. ~\\\\ \\noindent PACS Numbers: 52.30.Jb, 96.60.Rd, 47.15.Cb. ", "introduction": "\\label{sec-intro} Magnetic reconnection is of great interest in many space and laboratory plasmas \\cite{Kulsrud-1998,MRX-Yamada}, and has been studied extensively for more than four decades. The most important question is that of the reconnection rate. The process of magnetic reconnection, is so complex, however, that this question is still not completely resolved, even within the simplest possible {\\it canonical} model: two-dimensional (2D) incompressible resistive magnetohydrodynamics (MHD) with uniform resistivity $\\eta$ in the limit of $S\\rightarrow \\infty$ (where $S=V_A L/\\eta$ is the global Lundquist number, $L$ being the half-length of the reconnection layer). Historically, there were two drastically different estimates for the reconnection rate: the Sweet--Parker model \\cite{Sweet-1958,Parker-1963} gave a rather slow reconnection rate ($E_{\\rm SP} \\sim S^{-1/2}$), while the Petschek \\cite{Petschek-1964} model gave any reconnection rate in the range from $E_{\\rm SP}$ up to the fast maximum Petschek rate $E_{\\rm Petschek} \\sim 1/\\log S$. Up until the present it was still unclear whether Petschek-like reconnection faster than Sweet--Parker reconnection is possible. Biskamp's simulations \\cite{Biskamp-1986} are very persuasive that, in resistive MHD, the rate is generally that of Sweet--Parker. Still, his simulations are for $S$ in the range of a few thousand, and his boundary conditions are somewhat tailored to the reconnection rate he desires, the strength of the field and the length of layer adjusting to yield the Sweet--Parker rate. Thus, a more systematic boundary layer analysis is desirable to really settle the question. In particular, one needs an elaborate and detailed picture of the reconnection current layer --- namely, a picture that features a realistic model for the variation of the outside magnetic field along the layer, and realistic 2D profiles of the plasma parameters inside the layer. The development of such a framework is the main goal of the present paper. We believe that the methods developed in this paper are rather universal and can be applied to a very broad class of reconnecting systems that include more realistic physics. However, for definiteness and clarity we keep in mind a particular global geometry, that presented in Fig.~\\ref{fig-global} (although we do not use it explicitly in our present analysis). This Figure shows the situation somewhere in the middle of the process of merging of two plasma cylinders. Regions~I and II are ideal MHD regions: regions~I represent unreconnected flux, and region~II represents reconnected flux. The two regions~I are separated by the very narrow {\\it reconnection current layer}. Plasma from regions~I enters the reconnection layer and gets accelerated along the layer, finally entering the {\\it separatrix region}, also between regions~I and II. In general, both the reconnection layer and the separatrix region require resistive treatment. The plasma entering the separatrix from the reconnection layer is traveling at nearly the Alfv{\\'e}n speed. It crashes into the plasma at rest on the separatrix lines, that has not passed through the reconnection layer, but which got there by direct ${\\bf E \\times B}$ motion across the lines as the position of the separatrix changed by reconnection. This crash generates considerable heat, and hence back pressure on the reconnection layer. However, the separatrix region is continually in transition since different plasma occupies it as the reconnection proceeds so that the heated plasma is moved into the reconnected region. Thus, the plasma encountered by the outflowing reconnected plasma is continually refreshed and can always be taken initially at rest. Therefore, there is a time delay before the back reaction sets in. In our paper, we attempt to model this dynamical behavior of the separatrix plasma as accurately as possible. \\begin{figure} [tbp] \\centerline {\\psfig{file=global.eps,height=3 in,width=3 in,angle=90}} \\caption[The global geometry] {The global geometry.} \\label{fig-global} \\end{figure} In the limit $S\\rightarrow \\infty$ the reconnection rate is slow compared with the Alfv{\\'e}n time $\\tau_A=L/V_A$, which allows one to break the whole problem into the global problem and the local problem. In a previous paper Ref.~\\cite{Uzdensky-1996}, we argued that on the {\\it global scale} (i.e., on the scale of order the half-length of the layer~$L$) the time evolution of the reconnecting system can be described as a sequence of magnetostatic equilibria. In paper~\\cite{Uzdensky-1997} we explained that the role of the global solution is to give the general geometry of the reconnecting system, the position and the length of the reconnection layer and of the separatrix, and the boundary conditions for the local problem (which, in turn, determines the reconnection rate). These boundary conditions are expressed in terms of the outside magnetic field~$B_{y,0}(y)$, where $y$ is the direction along the layer. In particular, $B_{y,0}(y)$ provides the characteristic global scales: the half-length of the layer~$L$, defined as the point where $B_{y,0}(y)$ has minimum, and the global Alfv{\\'e}n speed, defined as $V_A=B_{y,0}(0)/\\sqrt{4\\pi\\rho}$. It is important to understand that the global solution is essentially independent of the local reconnection physics. In this paper we study the local problem concerning the reconnection layer itself. Our main goal here is to determine the internal structure of a steady state reconnection current layer (i.e., to find the 2D~profiles of plasma velocity and magnetic field), and the reconnection rate represented by the (uniform) electric field~$E$. Our most important result is that, in the case of uniform resistivity, there is a unique stable steady-state solution, and this is essentially the Sweet--Parker solution. We show that Petschek's solution (which is not unique and supposedly encompasses a variety of reconnection solutions from that of Sweet and Parker to a solution reconnecting at almost the Alfv{\\'e}n speed) actually relaxes to the single Sweet-Parker solution. First, in Section~\\ref{sec-model} we discuss the assumptions of our physical model of the layer in some detail. Then, in Section~\\ref{sec-rescaled} we introduce the rescaled equations representing the mathematical model of our problem. In Section~\\ref{sec-simulations} we present our numerical simulations. And, finally, in Section~\\ref{sec-conclusions} we give our conclusions. ", "conclusions": "\\label{sec-conclusions} To summarize, in this paper we present a definite systematic solution to a particular clear-cut, mathematically consistent problem concerning the {\\it internal structure of the reconnection layer} within the canonical framework (incompressible 2D MHD with uniform resistivity) with the outside field $B_{y,0}(y)$ varying on the global scale along the layer. We have first derived a system of rescaled equations that should be valid in the limit $S\\rightarrow \\infty$. Then, we have developed a 2D resistive MHD code that followed the time evolution of the system in order to achieve the steady state. We conclude that, under the assumptions of our model, {\\it the Petschek-like solutions are unstable} and the system quickly evolves to the only stable steady-state solution corresponding to the {\\it Sweet--Parker reconnection layer}. Thus, the Petschek mechanism for fast reconnection does not work in our model. The steady-state reconnection rate in our model problem is remarkably close to the Sweet--Parker value $E_{\\rm SP}=B_{y,0}(0)V_A/ \\sqrt{S}$. This main result is consistent with the results of simulations conducted by Biskamp\\cite{Biskamp-1986} and also those by Ugai\\cite{Ugai-Tsuda-1977} and by Scholer\\cite{Scholer-1989}. It also agrees with the experimental results in the MRX experiment \\cite{Ji-1998}. Finally, even though we draw our conclusions (about Petschek-like structures being unstable) only for this very specific model, this result is fundamentally important, because this model is the canonical framework typical of most models of magnetic reconnection, including both Sweet--Parker and Petschek. This framework is the simplest possible framework for a reconnection problem, and thus provides the necessary foundation on top of which one can add more complicated physical processes. Because the Sweet--Parker model with the classical (Spitzer) resistivity is known to be too slow to explain the very fast time scale for the energy release in solar flares, one has to look for physics beyond resistive MHD with the Spitzer resistivity. The inclusion of some new physical processes into the model (for example, locally enhanced anomalous resistivity is probably the most suitable candidate) would create a very different situation in which some Petschek-like structure with fast reconnection may be possible." }, "0003/astro-ph0003397_arXiv.txt": { "abstract": "We present a systematic numerical study of two-dimensional axisymmetric accretion flows around black holes. The flows have no radiative cooling and are treated in the framework of the hydrodynamical approximation. The models calculated in this study cover the large range of the relevant parameter space. There are four types of flows, determined by the values of the viscosity parameter $\\alpha$ and the adiabatic index $\\gamma$: convective flows, large-scale circulations, pure inflows and bipolar outflows. Thermal conduction introduces significant changes to the solutions, but does not create a new flow type. Convective accretion flows and flows with large-scale circulations have significant outward-directed energy fluxes, which have important implications for the spectra and luminosities of accreting black holes. \\noindent {\\em Subject headings:} Accretion, accretion disks --- conduction --- convection --- hydrodynamics --- turbulence ", "introduction": "In this paper we present the numerical study of global properties of hydrodynamical black hole accretion flows with a very inefficient radiative cooling. Such flows, coined ADAFs by Lasota (1996, 1999), are thought to be present in several astrophysical black hole candidates, in particular in low mass X-ray binaries and in some active galactic nuclei. Observed properties of ADAFs may be directly connected to black hole physics, and for this reason ADAFs have recently attracted a considerable attention (for reviews see e.g. Kato, Fukue \\& Mineshige 1998; Abramowicz, Bj\\\"ornsson \\& Pringle 1998; Narayan 1999). The radiation losses are unimportant for dynamics as well as for thermal balance of ADAFs, and therefore the details of radiative processes are not crucial. Radiative feed-back into hydrodynamics is negligible and may be treated as a small perturbation\\footnote{Objects that are {\\it external} to ADAFs may change the balance: for example, if there is an external source of soft photons, they may provide an additional, possibly efficient, Compton cooling of optically thin plasma (Shapiro, Lightman \\& Eardley 1976).}. Abramowicz et al. (1995) and in more details Chen et al. (1995) described accretion disks solutions in the parameter space ($\\dot{m}$, $\\tau$, $\\alpha$), where $\\dot{m}=\\dot{M}/\\dot{M}_{Edd}$ is the accretion rate expressed in the Eddington units, $\\tau$ is the optical depth and $\\alpha$ is the viscosity parameter. In this space ADAFs exist in two regimes (as anticipated by Rees et al. 1982): \\noindent (1) ADAFs with $\\tau\\gg 1$ have super-Eddington accretion rates, $\\dot{m} > 1$. Radiation is trapped inside the accretion flow (Katz 1977; Begelman 1978). To this category belong slim accretion disks (Abramowicz et al. 1988) which have a vertical scale comparable the corresponding radius. \\noindent (2) ADAFs with $\\tau\\ll 1$ have very sub-Eddington accretion rates, $\\dot{m}\\ll 1$. These flows have been first investigated by Ichimaru (1977), but the recent interest in them was generated mostly by the works of Narayan and his collaborators, after important aspects of nature of the flows had been explained by Narayan \\& Yi (1994, 1995b) and Abramowicz et al. (1995). The most important parameters for the physics of ADAFs are the viscosity parameter $\\alpha$ and the adiabatic index $\\gamma$. The latter parameter determines the regime of ADAFs through the equation of state. The parameters $\\tau$ and $\\dot{m}$ are not important: ADAFs have either $\\tau\\gg 1$ and $\\dot{m}> 1$, or $\\tau\\ll 1$ and $\\dot{m}\\ll 1$. There are no strong observational or theoretical limits for $\\alpha$ and $\\gamma$, and therefore, at present, one needs to construct models in wide ranges of them. In this paper we have constructed models for $10^{-2}\\le\\alpha\\le 1$ and $\\gamma = 4/3$, $3/2$, $5/3$. Our models are time dependent and fully 2-D: all components of forces and all components of viscous stresses are included in the calculations. To minimize the influence of the outer boundary condition onto the flow structure, we consider the solutions in the large radial range, $3r_g \\le r \\le 8\\times 10^3r_g$, where $r_g = 2GM/c^2$ is the gravitational radius of the central black hole with mass $M$. We have found that the properties of ADAFs depend mainly on the viscosity, i.e. on $\\alpha$, and also, but less strongly, on the adiabatic index $\\gamma$. Four types of accretion flows can be distinguished (see Figure 1). \\noindent (i) Convective flows. For a very small viscosity, $\\alpha\\la 0.03$, ADAFs are convectively unstable, as predicted by Narayan \\& Yi (1994) and confirmed in numerical simulations by Igumenshchev, Chen \\& Abramowicz (1996), Stone, Pringle \\& Begelman (1999) and Igumenshchev \\& Abramowicz (1999). Axially symmetric convection transports the angular momentum {\\it inward} rather than outward, for a similar reason as that described by Stone \\& Balbus (1996) in the context of turbulence: if there are no azimuthal gradients of pressure, turbulence tries to erase the angular momentum gradient. This property of convection governs the flow structure, as it was shown for ADAFs by Narayan, Igumenshchev \\& Abramowicz (2000) and Quataert \\& Gruzinov (2000). There are small scale circulations, with the matter fluxes considerably greater than that entering the black hole. Convection transports a significant amount of the dissipated binding energy outward. No powerful outflows are present. \\noindent (ii) Large-scale circulations. For a larger, but still small viscosity, $\\alpha\\sim 0.1$, ADAFs could be both stable or unstable convectively, depending on $\\alpha$ and $\\gamma$. The flow pattern consists of the large-scale ($\\sim r$) meridional circulations. No powerful unbound outflows are present. In some respect this type of flow is the limiting case of the convective flows in which the small scale motions are suppressed by larger viscosity. \\noindent (iii) Pure inflows. With an increasing viscosity, $\\alpha\\simeq 0.3$, the convective instability dies off. Some ADAFs (with $\\gamma\\simeq 3/2$) are characterized by a pure inflow pattern, and agree in many aspects with the self-similar models (Gilham 1981; Narayan \\& Yi 1994). No outflows are present. \\noindent (iv) Bipolar outflows. For a large viscosity, $\\alpha\\simeq 1$, ADAFs differ considerably from the simple self-similar models. Powerful unbound bipolar outflows are present. Effects of turbulent thermal conduction have been studied in several simulations. The conduction has an important influence to the flow structure, but it does not introduce a new type of flow. The paper is organized as follows. In \\S2 we describe equations, numerical method and boundary conditions. In \\S3 we present numerical results for models with and without thermal conduction. In \\S4 we discuss the properties of the solutions and their implications. In \\S5 we give the final conclusions. ", "conclusions": "We have performed a systematic study of 2D axisymmetric viscous rotating accretion flows into black holes in which radiative losses are neglected. Assumptions and numerical technique adopted here are similar to those used by Igumenshchev \\& Abramowicz (1999); a few modifications are connected to inclusion of thermal conduction. The thermal conduction flux was chosen to be proportional to the specific entropy gradient. We assumed that mass is steadily injected within an equatorial torus near the outer boundary of the spherical grid. The injected mass spreads due to the action of viscous shear stress and accretes. We set an absorbing inner boundary condition for the inflow at $r_{in}=3 r_g$. We study the flow structure over three decades in radius using a variety of values of the viscosity parameter $\\alpha$ and adiabatic index $\\gamma$. Our models without thermal conduction cover an extended region in the ($\\alpha$, $\\gamma$) plane, $0.01\\le\\alpha\\le 1$ and $4/3\\le\\gamma\\le 5/3$. We have found four types of pattern for the accretion flow, which had been found earlier in two-dimensional simulations by Igumenshchev \\& Abramowicz (1999), Stone et al. (1999) and Igumenshchev (2000). The type of flow mainly depends on the value of $\\alpha$, and is less dependent on the value of $\\gamma$. The high viscosity models, $\\alpha\\simeq 1$, form powerful bipolar outflows. The pure inflow and large-scale circulations patterns occur in the moderate viscosity models, $\\alpha\\simeq 0.1-0.3$. The low viscosity ($\\alpha\\la 0.03$) models exhibit strong convection. All models with bipolar outflows and pure inflow are steady. The models with large-scale circulations could be either steady or unsteady depending on the values of $\\alpha$ and $\\gamma$. All convective models are unsteady. Some of our pure inflow and convective models do show a self-similar behaviour. In particular, the pure inflow model ($\\alpha=0.3$, $\\gamma=3/2$) reasonably well satisfies the predictions of the self-similar solutions of Gilham (1981) and Narayan \\& Yi (1994) in which $\\rho\\propto r^{-3/2}$. The convective accretion flows show a good agreement with the self-similar solutions recently found by Narayan et al. (2000) and Quataert \\& Gruzinov (2000) in which $\\rho\\propto r^{-1/2}$. The latter solutions have been constructed for the convection transporting angular momentum towards the gravitational center. The self-similar solutions for accretion flows with bipolar outflows (ADIOs) proposed by Blandford \\& Begelman (1999) have not been confirmed in our numerical simulations. The most interesting feature of the flows with large-scale circulations and convective accretion flows is the non-zero outward energy flux, which is equivalent to the effective luminosity $\\sim 10^{-2}\\dot{M}_0 c^2$, where $\\dot{M}_0$ is the black hole accretion rate. This result has important implications for interpretation of observations of accreting black hole candidates and neutron stars. The accretion flows with thermal conduction have not been studied as completely as the non-conductive flows. The conductive models show only two types of laminar flow patterns: pure inflow (with $\\alpha=0.3$) and global circulation (with $\\alpha\\simeq 0.03-0.1$). The thermal conduction mainly acts as a cooling agent in our models, it suppresses bipolar outflows and convective motions. \\noindent{\\it Acknowledgments.} The authors gratefully thank Ramesh Narayan for help with interpretation of the numerical results and comments on a draft of the paper, Ed Spiegel for pointing out the importance of thermal conduction in viscous accretion flows, Rickard Jonsson for useful comments, and Jim Stone and Eliot Quataert for discussions. The work was supported by the Royal Swedish Academy of Sciences. \\clearpage \\begin{deluxetable}{lccccc} \\footnotesize \\tablecaption{Parameters of the models. \\label{tbl-1}} \\tablewidth{0pt} \\tablehead{ \\colhead{Model} & \\colhead{~~$\\alpha$~~} & \\colhead{~~$\\gamma$~~} & \\colhead{~~${Pr}$~~} & \\colhead{stability} & \\colhead{outflow(s)\\tablenotemark{a}} } \\startdata A & $1$ & $5/3$ & $\\infty$ & stable & bipolar \\nl B & $1$ & $3/2$ & $\\infty$ & stable & bipolar \\nl C & $1$ & $4/3$ & $\\infty$ & stable & bipolar \\nl D & $0.3$ & $5/3$ & $\\infty$ & stable & bipolar \\nl E & $0.3$ & $3/2$ & $\\infty$ & stable & --- \\nl F & $0.3$ & $4/3$ & $\\infty$ & unstable & unipolar \\nl G & $0.1$ & $5/3$ & $\\infty$ & stable & unipolar \\nl H & $0.1$ & $3/2$ & $\\infty$ & unstable & --- \\nl I & $0.1$ & $4/3$ & $\\infty$ & unstable & --- \\nl J & $0.03$ & $5/3$ & $\\infty$ & unstable & --- \\nl K & $0.03$ & $3/2$ & $\\infty$ & unstable & --- \\nl L & $0.03$ & $4/3$ & $\\infty$ & unstable & --- \\nl M & $0.01$ & $5/3$ & $\\infty$ & unstable & --- \\nl N & $0.3$ & $5/3$ & $1$ & stable & --- \\nl O & $0.3$ & $3/2$ & $1$ & stable & --- \\nl P & $0.3$ & $4/3$ & $1$ & stable & --- \\nl Q & $0.1$ & $5/3$ & $1$ & stable & unipolar \\nl R & $0.1$ & $3/2$ & $1$ & stable & unipolar \\nl S & $0.03$ & $5/3$ & $1$ & stable & unipolar \\nl \\enddata \\tablenotetext{a}{Only powerful outflows are indicated} \\end{deluxetable} \\clearpage \\begin{deluxetable}{lc} \\footnotesize \\tablecaption{The ratio of accretion rate to outflow rate. \\label{tbl-2}} \\tablewidth{0pt} \\tablehead{ \\colhead{Model~~~~~~} & \\colhead{~~~~$\\dot{M}_{0}/\\dot{M}_{out}$\\tablenotemark{a}~~~~} } \\startdata A & $0.015$ \\nl C & $0.079$ \\nl D & $0.063$ \\nl E & $1.95$ \\nl F & $0.126$ \\nl G & $0.006$ \\nl I & $0.021$ \\nl J & $0.001$ \\nl K & $0.002$ \\nl L & $0.003$ \\nl \\enddata \\tablenotetext{a}{$\\dot{M}_{0}$ and $\\dot{M}_{out}$ are the accretion and outflow rates measured at the inner and outer numerical boundaries, respectively.} \\end{deluxetable} \\clearpage" }, "0003/astro-ph0003168_arXiv.txt": { "abstract": "Seyfert galaxies have been shown to exhibit a large variety of features in their X-ray spectra from which the environment of the central engine can be deduced. We focus on the two following aspects: the Warm Absorber, mainly responsible of the soft X-ray properties, and the reprocessing/reflecting plasma medium at the origin of the iron K$\\alpha$ fluorescent line. The physical parameters and the location of the so-called Warm Absorber (WA), a photoionized medium along the line of sight to the nuclear region, are more strongly constrained by optical coronal lines than by the oxygen edges observed in the soft X-rays and produced by the WA. The photoionization models also predict the intensities of the X-ray emission lines which are going to be detected with the new generation of X-ray satellites. An alternative model to the relativistic accretion disc is proposed to explain the profile of the X-ray iron K$\\alpha$ line observed in the Seyfert 1. This line can be formed in the framework of a quasi-spherical accretion of optically thick clouds. An optically thick photoionization code coupled with a Monte-Carlo code has been developed to compute the entire spectrum from the IR to the hard X-rays for a close geometry with a large covering factor. The multiple Compton reflections allow to reproduce the redshifted broad iron line as detected in several Seyfert 1 galaxies and in MCG-6-30-15. ", "introduction": "All types of AGN are bright X-ray emitters and a large number has been discovered after their optical identification with an X-ray source. A large fraction of their high luminosity, due to the release of gravitational energy in the central regions, is radiated in the X-rays ($\\sim$ 20 to 30 \\% for Seyfert 1 and 10\\% for quasars (see review by Mushotzky et al. 1993)). In this paper, X-ray properties of Seyfert 1 only are discussed. The shape of the X-ray spectrum above 2\\,keV is schematically described by a power law (photon index of -1.7) with a high energy cut-off at 100 keV, and below 1\\,keV it is absorbed by a warm medium. It also exhibits a hump at 20~keV and a soft X-ray excess, particularly strong in narrow line Seyfert 1 galaxies. A broad emission iron line at 6.4\\,keV is also often seen (Fabian et al. 1995). The X-ray emission is rapidly variable: variations on a timescale of a few hours are often detected in Seyfert 1 and down to 100~s in MCG-6-30-15 (e.g. Yaqoob et al. 1997). Such short timescales indicate emitting regions smaller than 10$^{13}$\\,cm. The detection of a strong correlation between the UV and the X-rays with a small or even null lag might favor an origin of the UV due to the reprocessing of the X-rays from a source close to the black hole. However this simple scenario is in conflict with recent observations of the two sources NGC 7469 (Nandra et al. 1998) and NGC 5548 (Chiang et al. 1999). On longer timescales several objects including some Narrow Line Seyfert 1 galaxies have exhibited two brightness states with different spectral shape, similarly to galactic black hole candidates (f.i. 1H0419-577, Guainazzi et al. 1998). Several phenomenological models have been proposed to account for the bulk of the X-ray emission such as the disk-corona model, the standard irradiated model, the dense blob model and the dilute cloud model (see review by Collin \\& Dumont 1996). ", "conclusions": "" }, "0003/astro-ph0003442_arXiv.txt": { "abstract": "A comparative study of simulated air shower longitudinal profiles is presented. An appropriate thinning level for the calculations is first determined empirically. High statistics results are then provided, over a wide energy range, ($10^{14.0}$ to $10^{20.5}$~eV), for proton \\& iron primaries, using four combinations of the \\mocca\\ \\& \\corsika\\ program frameworks, and the \\sibyll\\ \\& \\qgsjet\\ high energy hadronic interaction models. These results are compared to existing experimental data. The way in which the first interaction controls \\xm\\ is investigated, as is the distribution of \\xm. ", "introduction": "The energy spectrum of cosmic rays is a power law with the flux falling by three orders of magnitude for each decade increase in energy. At $\\sim 10^{14}$~eV the flux becomes so low that current balloon and satellite experiments lack the exposure required to detect a significant sample of events. This is unfortunate as the nature of the primaries remains of great astrophysical interest. Where direct measurements are possible the cosmic rays are known to be mostly protons and atomic nuclei. The most plausible acceleration site is at the shock fronts produced by supernova explosions. However, theoretical considerations predict a maximum energy from this process of $\\sim 10^{15}$~eV, whereas the energy spectrum is observed to continue with only small deviations up to $>10^{19}$~eV. The origin of the particles at $>10^{15}$~eV is somewhat mysterious. It has long been supposed that insight would result if the composition of the primaries could be measured. Due to the extremely low flux the only way to get information on these particles is to study the extensive atmospheric cascades which they initiate. When a cosmic ray enters the atmosphere it collides with the nucleus of an air atom, producing a number of secondaries. These go on to make further collisions, and the number of particles grows. Eventually the energy of the shower particles is degraded to the point where ionization losses dominate, and their number starts to decline. It is a coincidence that at the energy where direct detection of the cosmic rays becomes impractical, the resulting air showers become big enough to be easily detectable at ground level. The number of particles in the cascade also becomes large enough that the longitudinal profile, or number of particles versus atmospheric depth, becomes a smooth curve, with a well defined maximum. This maximum depth, referred to as \\xm, is often regarded as the most basic parameter of an air shower, and much effort has been expended to measure and interpret it. The depth of maximum increases with primary energy as more cascade generations are required to degrade the secondary particle energies. For given total energy \\xm\\ is related to the energy per nucleon of the primary. To first order the interactions occur between individual nucleons of the primary, and the target air nuclei. Therefore a shower initiated by a compound nucleus can be thought of as the superposition of many proton initiated showers, with correspondingly lower energy. Unfortunately, of course, the detail is not so simple. For a number of reasons extracting information on the nature of the cosmic ray primaries from the air showers they produce has proved to be exceedingly difficult. The most fundamental problem is that the initial interactions are subject to large inherent fluctuations. This limits the event-by-event mass resolution of even an ideal detector. However, progress can still be made by looking at mean parameter values, or better, their distributions. The second major problem is that sophisticated modeling is required to predict the absolute value of an observable parameter which is expected for a primary of given type and energy. Nucleus-nucleus interactions at the energies of the first few cascade steps are well beyond the reach of accelerator experiments. Therefore it is necessary to rely on hadronic interaction models which attempt to extrapolate from the available data using different mixtures of theory and phenomenology. The lower energy part of the cascade can be modeled using well known physics, although the programs are complex with corresponding scope for errors. The depth of shower maximum has been determined by a number of experiments. In the energy range $10^{14}$ to $10^{16}$~eV it has been measured with varying degrees of directness using \\v{C}erenkov light~\\cite{dice,airobicc,blanca}. The range $10^{17}$ to $10^{19}$~eV has been observed rather directly by the Fly's Eye detector through fluorescence light~\\cite{Bird93_1}. Finally the region above $10^{19}$~eV is the focus of the HiRes Fly's Eye~\\cite{HiRes}, Auger Project~\\cite{Auger}, and others. In the past the experimental resolution and statistics have often been so poor that the mean value of \\xm\\ has been discussed rather than its distribution --- this is changing. The simulations required to interpret the data from any given experiment have usually been performed only for the energy range accessible to it. This is unfortunate since checking the consistency of experiments in adjacent energy ranges is critically important, given the uncertainty of the high energy hadronic interaction models. Additionally the exact value of \\xm\\ for a given model can depend on the way in which the longitudinal profiles are recorded and fit. The purpose of this paper is to provide \\xm\\ values with good statistical precision, over a wide energy range, and computed in a consistent way using several hadronic models and two different cascade ``framework'' programs; for a more detailed discussion see~\\cite{GAP-98-035}. The process of air shower simulation can be broken up into several parts. A framework program is required which handles the mechanics of the process and calls appropriate subroutines to model the interaction and propagation of the particles. Some fraction of the required transport and interaction modeling may be provided using third party code. In this paper two air shower simulation packages which have been heavily used in the literature are considered. The first is \\mocca\\ written by Hillas~\\cite{MOCCA}. This originally used a simple, built-in hadronic interaction model, but has also been linked to \\sibyll~\\cite{Sibyll}; all other modeling is handled internally. The second program is \\corsika, a well documented and thorough program prepared originally for the Kascade experiment~\\cite{CORSIKA}. It is linked to a number of high energy hadronic models, two of which are suitable for use over the very wide energy range of this study; \\sibyll\\ and \\qgsjet~\\cite{QGSjet}. An attractive feature of this program is the use of the well established High Energy Physics codes \\egs~\\cite{EGS4} and \\geisha~\\cite{GEISHA}, for the electromagnetic, and lower energy hadronic modeling respectively. See Table~\\ref{tab:simprogs} for a summary. \\begin{table} \\begin{tabular}{|l|l|l|l|l|} \\hline Program frame & High E Hadronic & Low E Hadronic & Electromagnetic \\\\ \\hline \\mocca & Internal & Internal & Internal \\\\ & \\sibyll && \\\\ \\hline \\corsika & \\sibyll & \\geisha & \\egs \\\\ & \\qgsjet && \\\\ \\hline \\end{tabular} \\label{tab:simprogs} \\caption{Summary of the four program-frame / interaction-model combinations.} \\end{table} Due to the inherent limitations of air shower fluctuations, and also because of poor experimental resolution and statistics, \\xm\\ data is often compared only to simulated values for proton and iron nuclei primaries. These are generally regarded as the extreme ends of the possible range. At lower energies the composition of cosmic rays tracks the general abundances of solar system material, with some modifications due to propagation spallation effects. Iron is the heaviest significantly abundant element. ", "conclusions": "When running shower simulations to study \\xm\\ it is better to generate heavily thinned showers, with explicit low energy hadronic and electromagnetic cascades, than to rely on analytic approximations for these parts of the calculation. The latter has frequently been done in the past, leading to concerns that the results are biased to an unknown extent. When working with an explicit, but thinned, cascade simulation it is possible to determine an appropriate thinning level empirically, by comparing against more lightly thinned results. Carefully calculated \\xm\\ results have been presented, over a wide energy range, for proton \\& iron primaries, using four combinations of framework program and high energy hadronic interaction model. It is hoped that these will be of use for future comparisons with experimental data, and with other simulation results. The way in which the first interaction controls \\xm\\ has been investigated. The influence is strong --- if one were to use model $A$ for the first few interactions, and model $B$ thereafter, the mean \\xm\\ results would be close to using model $A$ throughout. \\qgsjet\\ predicts that the separation between proton and iron \\xm\\ declines at the highest energies. If this is true it is unfortunate from an experimental perspective. It would be very useful if a common reference set of showers were made available by the authors of new, or modified, hadronic interaction models. For the purposes of longitudinal profile comparison the set used here seems adequate; the raw and processed output is available online~\\cite{data}. The Fermilab computing department are thanked for the use of their machines. \\begin{table} \\begin{tabular}{|l|l|c|c|c|c|} \\hline Primary & $\\log_{10}(E (eV))$ & \\multicolumn{4}{c|}{ $<\\!X_{max}\\!>$, $\\sigma X_{max}$ (g~cm$^{-2}$)} \\\\ \\cline{3-6} & & \\multicolumn{2}{c|}{MOCCA} & \\multicolumn{2}{c|}{CORSIKA} \\\\ \\cline{3-6} & & Internal & SIBYLL & SIBYLL & QGSjet \\\\ \\hline Proton & 14.0 & 537 , 96 & 525 , 91 & 517 , 97 & 484 , 96 \\\\ & 14.5 & 570 , 86 & 559 , 92 & 560 , 99 & 525 , 92 \\\\ & 15.0 & 605 , 78 & 597 , 86 & 589 , 97 & 560 , 88 \\\\ & 15.5 & 637 , 72 & 625 , 81 & 621 , 84 & 587 , 76 \\\\ & 16.0 & 674 , 71 & 651 , 72 & 651 , 81 & 618 , 72 \\\\ & 16.5 & 704 , 62 & 676 , 59 & 679 , 70 & 646 , 70 \\\\ & 17.0 & 738 , 60 & 713 , 65 & 699 , 63 & 679 , 75 \\\\ & 17.5 & 767 , 54 & 733 , 57 & 729 , 63 & 705 , 68 \\\\ & 18.0 & 802 , 53 & 765 , 60 & 762 , 59 & 728 , 69 \\\\ & 18.5 & 839 , 50 & 793 , 53 & 790 , 62 & 755 , 67 \\\\ & 19.0 & 867 , 44 & 829 , 55 & 817 , 52 & 779 , 68 \\\\ & 19.5 & 908 , 60 & 852 , 49 & 847 , 52 & 804 , 68 \\\\ & 20.0 & 940 , 50 & 883 , 53 & 875 , 54 & 825 , 59 \\\\ & 20.5 & 972 , 42 & 908 , 47 & 900 , 45 & 849 , 59 \\\\ \\hline Iron & 14.0 & 370 , 36 & 366 , 38 & 357 , 32 & 346 , 32 \\\\ & 14.5 & 420 , 34 & 414 , 35 & 407 , 33 & 390 , 30 \\\\ & 15.0 & 461 , 34 & 464 , 34 & 452 , 31 & 432 , 29 \\\\ & 15.5 & 507 , 32 & 508 , 34 & 493 , 28 & 471 , 26 \\\\ & 16.0 & 545 , 29 & 542 , 31 & 533 , 27 & 509 , 25 \\\\ & 16.5 & 582 , 25 & 577 , 27 & 568 , 27 & 544 , 26 \\\\ & 17.0 & 619 , 23 & 610 , 29 & 600 , 25 & 577 , 24 \\\\ & 17.5 & 652 , 22 & 641 , 26 & 631 , 25 & 609 , 22 \\\\ & 18.0 & 688 , 21 & 668 , 24 & 660 , 25 & 640 , 24 \\\\ & 18.5 & 720 , 18 & 697 , 21 & 689 , 24 & 669 , 23 \\\\ & 19.0 & 754 , 18 & 725 , 22 & 717 , 21 & 701 , 20 \\\\ & 19.5 & 787 , 17 & 754 , 21 & 744 , 20 & 730 , 21 \\\\ & 20.0 & 820 , 16 & 783 , 19 & 774 , 20 & 759 , 21 \\\\ & 20.5 & 853 , 14 & 809 , 19 & 802 , 20 & 791 , 22 \\\\ \\hline \\end{tabular} \\label{tab:xmax} \\caption{Mean and standard deviation values of \\xm\\ for proton and iron primaries, and four program-frame / interaction-model combinations. Each pair of numbers comes from a 500 shower set.} \\end{table}" }, "0003/astro-ph0003218_arXiv.txt": { "abstract": "We describe the possible electromagnetic signals expected from the magnetospheric interactions of a neutron star binary prior to merger. We find that both radio and X-ray signals of detectable strength are possible. We discuss possible links with the phenomenon of gamma-ray bursts and describe the prospects for direct detection of these signals in searches for radio and X-ray transients. ", "introduction": "The gravitational wave-induced merger of binary neutron stars has evoked considerable interest in recent years due to their importance as a source of gravitational radiation (Thorne 1987 and references therein) and potentially also gamma-ray bursts (Blinnikov et al 1984; Paczynski 1986; Eichler et al 1989; Paczynski 1991). The goal of identifying electromagnetic signatures of the merger events is an important one, even if such manifestations are not gamma-ray bursts. Given the considerable information processing required to infer the presence of a gravitational wave burst (Cutler et al 1993), the presence of another signature will be invaluable. In this paper we examine the magnetospheric interactions in merging neutron star binary systems and describe their pre- and post-merger signatures. In particular, we consider systems containing one low field $B_r \\sim 10^{9-11} G$, rapidly spinning ($P \\sim 1-100 ms$) recycled pulsar and one high field ($B_m \\sim 10^{12-15} G$), slowly spinning ($P \\sim 10-1000 s$) non-recycled pulsar, as expected on both empirical and evolutionary grounds. We shall examine how energy is extracted from the spin and orbital motion of the pulsar and in what manner it is radiated. Aspects of this calculation have been considered before by Vietri (1996), who considered magnetospherically induced gamma-ray bursts, and Lipunov \\& Panchenko (1996), who considered the the far-field dipolar and quadrupolar configurations of a dipole merging with a superconducting sphere. Our default estimates will be for systems in which the high field pulsar has a field $\\sim 10^{15}$~G (sometimes called a magnetar), which has the potential for the strongest signal. Recent work suggests that such pulsars may constitute $\\sim 10\\%$ of the young pulsar population (Kulkarni \\& Thompson 1998). In section~\\ref{Extract} we will describe the magnetospheric interactions that remove energy from the orbit and which allow it to emerge in electromagnetic form. This section will draw heavily on concepts from pulsar electrodynamics and also the field of satellite-magnetosphere electrodynamics, such as in the Io-Jupiter system. One of the primary results is that much of the energy is released as a pair plasma into the magnetosphere. Section~\\ref{Cooling} describes the physical state and evolution of this plasma, drawing on concepts developed to describe Soft Gamma Repeaters and section~\\ref{Discuss} reviews the state of the observations appropriate to this phenomenon. ", "conclusions": "If neutron star mergers are not associated with GRB, then any additional electromagnetic signatures will be invaluable when the search begins in earnest for the gravitational wave signal. Li \\& Paczynski (1999) have suggested one such signature; namely a post-merger mini-supernova powered by radioactive decay of disrupted neutron star material. We have demonstrated the possibility of additional {\\em precursor} signals in the radio and X-ray regimes, driven by the magnetospheric interactions of the neutron star and their magnetic fields. Our results differ somewhat from those of Vietri (1996) who considered a related model. We ascribe this to the much more localized interaction in our scenario, the result of a more realistic choice of parameters, and to our more complete description of the electrodynamics of the accelerated plasma. To conclude we re-iterate the properties of what we would consider a prime candidate for an electromagnetic counterpart to a neutron star merger. Estimates of the merger rate suggest that the events typically observed would be at distances $\\sim 100$~Mpc, suggesting X-ray fluxes $\\sim 3 \\times 10^{-9} {\\rm ergs \\, cm^{-2} s^{-1}}$ with effective temperatures progressing upwards through the 10-100~keV range preceding the gamma-ray event on timescales of order seconds or less. Associated radio fluxes could be as much as $\\sim 5$~Jy at this distance, although the ability of the radio waves to propagate in the late-time plasma shroud is rather uncertain. The coincidence of the radio signal could be influenced by dispersion in both the host galaxy and ours. Dispersion in the inter-Galactic medium will be of the order of $\\sim 1 {\\rm cm^{-2} pc} (D/100 {\\rm Mpc})$ for an ionized IGM mass fraction $\\sim 10^{-2}$ of the critical density and thus is unlikely to contribute significantly for any detectable events. There are also several possible signatures of the merger event itself, depending on how the orbital and binding energies of the binary and components is disbursed between remnant and ejecta. We thank Steve Thorsett, Vicky Kaspii \\& Jackie Hewitt for information regarding the FLIRT and STARE radio transient programs and Vladimir Lipunov and Andrei Gruzinov for discussions." }, "0003/astro-ph0003054_arXiv.txt": { "abstract": "We investigate the rates of production and thermalization of $\\nu_\\mu$ and $\\nu_\\tau$ neutrinos at temperatures and densities relevant to core-collapse supernovae and protoneutron stars. Included are contributions from electron scattering, electron-positron annihilation, nucleon-nucleon bremsstrahlung, and nucleon scattering. For the scattering processes, in order to incorporate the full scattering kinematics at arbitrary degeneracy, the structure function formalism developed by Reddy et al. (1998) and Burrows and Sawyer (1998) is employed. Furthermore, we derive formulae for the total and differential rates of nucleon-nucleon bremsstrahlung for arbitrary nucleon degeneracy in asymmetric matter. We find that electron scattering dominates nucleon scattering as a thermalization process at low neutrino energies ($\\varepsilon_\\nu\\lesssim 10$ MeV), but that nucleon scattering is always faster than or comparable to electron scattering above $\\varepsilon_\\nu\\simeq10$ MeV. In addition, for $\\rho\\gtrsim 10^{13}$ g cm$^{-3}$, $T\\lesssim14$ MeV, and neutrino energies $\\lesssim60$ MeV, nucleon-nucleon bremsstrahlung always dominates electron-positron annihilation as a production mechanism for $\\nu_\\mu$ and $\\nu_\\tau$ neutrinos. ", "introduction": "\\label{sec:intro} The cores of protoneutron stars and core-collapse supernovae are characterized by mass densities of order $\\sim 10^{10}-10^{14}$ g cm$^{-3}$ and temperatures that range from $\\sim 1$ to $50$ MeV. The matter is composed predominantly of nucleons, electrons, positrons, and neutrinos of all species. For $\\nu_\\mu$ and $\\nu_\\tau$ types (collectively `$\\nu_\\mu$s'), which carry away 50$-$60\\% of the $\\sim 2-3\\times 10^{53}$ ergs liberated during collapse and explosion, the prevailing opacity and production processes are $\\nu_\\mu$-electron scattering, $\\nu_\\mu$-nucleon scattering, electron-positron annihilation ($e^+e^-\\leftrightarrow \\nu_\\mu\\bar{\\nu}_\\mu$), and nucleon-nucleon bremsstrahlung. While all of these processes contribute for the electron types ($\\nu_e$s and $\\bar{\\nu}_e$s), the charged-current absorption processes $\\nu_e n\\leftrightarrow p e^-$ and $\\bar{\\nu}_ep\\leftrightarrow ne^+$ dominate their opacity so completely that in this paper we address only $\\nu_\\mu$ production and thermalization. Supernova theorists had long held \\cite{lamb_pethick} that $\\nu_\\mu$-nucleon scattering was unimportant as a mechanism for neutrino equilibration. While this process was included as a source of opacity \\cite{bruenn_1985,bhf_1995}, it served only to redistribute the neutrinos in space, not in energy. In contrast, $\\nu_\\mu$-electron scattering was thought to dominate $\\nu_\\mu$ neutrino thermalization. In addition, the only $\\nu_\\mu\\bar{\\nu}_\\mu$ pair production mechanisms employed in full supernova calculations were $e^+e^-\\leftrightarrow \\nu_\\mu\\bar{\\nu}_\\mu$ and plasmon decay ($\\gamma_{pl}\\leftrightarrow \\nu_\\mu\\bar{\\nu}_\\mu$) \\cite{bruenn_1985}; nucleon-nucleon bremsstrahlung was neglected as a source. Recent developments, however, call both these practices into question and motivate a re-evaluation of these opacities in the supernova context. Analytic formulae have recently been derived \\cite{reddy_1998,burrows_sawyer,bs_1999} which include the full nucleon kinematics and Pauli blocking in the final state at arbitrary nucleon degeneracy. These efforts reveal that the average rate of energy transfer in $\\nu_\\mu$-nucleon scattering may surpass previous estimates by an order of magnitude \\cite{burrows_sawyer,hannestad,keil_1995,janka_1996,raffelt_seckel,sigl_1997}. Hence, this process may compete with $\\nu_\\mu$-electron scattering as an equilibration mechanism. Similarly, estimates for the total nucleon-nucleon bremsstrahlung rate have been obtained \\cite{hannestad,burrows_1999,friman,fsb_1975} which indicate that this process might compete with $e^+e^-$ annihilation. These results suggest that the time is ripe for a technical study of the relative importance of each process for production or thermalization. To conduct such a study, we consider $\\nu_\\mu$ neutrinos in an isotropic homogeneous thermal bath of scatterers and absorbers. In this system, the full transport problem is reduced to an evolution of the neutrino distribution function (${\\cal F}_\\nu$) in energy space alone. Although this is a simplification of the true problem, it provides a theoretical laboratory in which to analyze the rates both for equilibration of an initial neutrino distribution function with dense nuclear matter and for production of the neutrinos themselves. From these rates we determine the importance and particular character of each process, and discover in which energy, temperature, or density regime each dominates. We employ a general prescription for solving the Boltzmann equation in this system with the full energy redistribution collision term. We compare quantitatively, via direct numerical evolution of an arbitrary neutrino distribution function, the rates for thermalization and production by each process, at all neutrino energies. Furthermore, we present the total nucleon-nucleon bremsstrahlung rate for arbitrary nucleon degeneracy and derive the single $\\nu_\\mu$ and $\\bar{\\nu}_\\mu$ production spectra. This facilitates a more comprehensive evaluation of its relative importance in neutrino production than has previously been possible. In \\S \\ref{sec:genform}, we discuss the general form of the Boltzmann equation and our use of it to study $\\nu_\\mu$ equilibration and production rates. In \\S \\ref{sec:interactions}, we provide formulae for each of the four processes we consider: $\\nu_\\mu$-nucleon scattering, $\\nu_\\mu$-electron scattering, and $\\nu_\\mu\\bar{\\nu}_\\mu$ pair production via both nucleon-nucleon bremsstrahlung and $e^+e^-$ annihilation. In \\S \\ref{sec:results}, we present the results of our equilibration calculations, showing the time evolution of $\\nu_\\mu$ distribution functions as influenced by each of these processes individually. We include plots of thermalization and production rates for each process as a function of neutrino energy and time. For the scattering interactions we include figures of the time evolution of the net energy transfer to the medium as a function of incident neutrino energy. We repeat this analysis at points in temperature, density, and composition space relevant to supernovae and protoneutron stars, taken from snapshots of a stellar profile during a realistic collapse calculation \\cite{bhf_1995}. Using these results, we discuss the relative importance of each process in shaping the emergent $\\nu_\\mu$ spectrum. In \\S \\ref{sec:conclusions}, we recapitulate our findings and conclusions. ", "conclusions": "\\label{sec:conclusions} Our results for equilibration via $\\nu_\\mu$-electron scattering and $\\nu_\\mu$-nucleon scattering indicate that the latter competes with or dominates the former as a thermalizer for neutrino energies $\\gtrsim10$ MeV for $\\rho\\gtrsim1\\times10^{11}$ g cm$^{-3}$ at all temperatures. At neutrino energies $\\gtrsim30$ MeV the difference at all densities and temperatures is approximately an order of magnitude. For the production and absorption processes, we find that nucleon-nucleon bremsstrahlung, at the average energy of an equilibrium Fermi-Dirac distribution at the local temperature, is 5 and 2 orders of magnitude faster than $e^+e^-$ annihilation at StarA ($T\\sim10$ MeV, $\\rho\\sim10^{14}$ g cm$^{-3}$) and StarB ($T\\sim15$ MeV, $\\rho\\sim10^{13}$ g cm$^{-3}$), respectively. Only for $\\rho\\sim10^{12}$ g cm$^{-3}$ and $T\\sim6$ MeV does $e^+e^-\\leftrightarrow\\nu_\\mu\\bar{\\nu}_\\mu$ begin to compete with bremsstrahlung at all energies. We conclude from this study that the emergent $\\nu_\\mu$ and $\\nu_\\tau$ spectrum is (1) brighter and (2) softer than previously estimated. The former results from the inclusion of the new pair emission process, nucleon-nucleon bremsstrahlung. The latter is a consequence of both the increased energy coupling between the nuclear and neutrino fluids through $\\nu_\\mu$-nucleon scattering and the fact that bremsstrahlung dominates $e^+e^-$ annihilation near the neutrinospheres at the lowest neutrino energies. While the full transport problem, including $\\nu_\\mu$-nucleon scattering energy redistribution and nucleon-nucleon bremsstrahlung, must be solved in order to delineate precisely what consequences these processes have for the emergent $\\nu_\\mu$ spectrum, these calculations demonstrate that they should not be omitted." }, "0003/astro-ph0003324_arXiv.txt": { "abstract": "We present a measurement of the cluster X-ray luminosity-temperature relation out to high redshift ($z \\sim 0.8$). Combined $ROSAT$ PSPC spectra of 91 galaxy clusters detected in the Wide Angle $ROSAT$ Pointed Survey (WARPS) are simultaneously fit in redshift and luminosity bins. The resulting temperature and luminosity measurements of these bins, which occupy a region of the high redshift L-T relation not previously sampled, are compared to existing measurements at low redshift in order to constrain the evolution of the L-T relation. We find a best fit to low redshift ($z<0.2$) cluster data, at $\\rm T > 1 \\, keV$, to be $\\rm L \\propto T^{3.15\\pm0.06}$. Our data are consistent with no evolution in the normalisation of the L-T relation up to $z \\sim 0.8$. Combining our results with $ASCA$ measurements taken from the literature, we find $\\eta=0.19\\pm0.38$ (for $\\Omega_0=1$, with $1\\sigma$ errors) where $\\rm L_{Bol} \\propto {(1+\\it z)}^{\\eta} \\rm T^{3.15}$, or $\\eta=0.60\\pm0.38$ for $\\Omega_0=0.3$. This lack of evolution is considered in terms of the entropy-driven evolution of clusters. Further implications for cosmological constraints are also discussed. ", "introduction": "Clusters of galaxies, as the largest virialised objects in the Universe, allow us a unique insight into the formation and evolution of mass clustering on cosmological time-scales. The X-ray emission from galaxy clusters originates from hot ($\\rm T \\sim 10^{7-8}$ K) intra-cluster gas which, in relaxed systems, is in hydrostatic equilibrium with the cluster's total gravitational field. Since the temperature of the gas is proportional to its mass, the observed correlation between X-ray luminosities and temperatures implies a relationship between cluster baryon mass and total mass. Thus, studying the evolution of the L--T relation probes the interrelated evolution of these mass components. One of the first attempts at modelling these properties assumed self-similar evolution of both the gravitational potential and the baryonic intracluster medium (ICM) at varying cosmological epochs (\\pcite{kai86}). These self-similar predictions are in conflict with observational evidence. The strong positive evolution of the cluster X-ray luminosity function (XLF) originally predicted (assuming realistic cosmological parameters) is in stark contrast to the negative evolution reported by \\scite{gio90} and \\scite{hen92} for X-ray luminous clusters. The prediction is also in conflict with the results of more recent cluster surveys which find no significant evolution, at any luminosity, out to moderate redshift ($z \\sim 0.3$, \\pcite{ebe97}; \\pcite{deg99}), and again no significant (or mild negative) evolution at low and intermediate luminosities and high redshifts (\\pcite{ros95}; \\pcite{bur97}; \\pcite{vik98a}; \\pcite{jon98b}). To account for such discrepancies, modifications were suggested to the self-similar theory. \\scite{kai91} and \\scite{evr91} modelled a 'pre-heated ICM', in which an initial injection of energy broke the direct self-similar scaling by introducing an entropy floor in cluster cores. It also brought theoretical predictions of cluster properties into better agreement with observational constraints. Tests of gravitational self-similar scaling are thus fundamental to the understanding of cluster evolution. One such test used in this analysis is the study of the cluster X-ray luminosity--temperature (L--T) relation, which was predicted by the original self-similar model to be described by an $\\rm L \\propto T^2$ law. However, a relationship closer to $\\rm L \\propto T^3$, has been found by several cluster surveys and compilations (e.g. \\pcite{edg91}; \\pcite{dav93}; \\pcite{whi97}), whilst even further steepening of the relation is required to agree with galaxy group measurements (\\pcite{pon96}; \\pcite{hel00}). In addition to predicting the slope of the L--T relation, self-similar theory also predicts significant evolution in the normalisation of this relation at different redshifts. However, the evolution predicted is reduced in low density cosmologies, or by significant entropy injection at early epochs (e.g. \\pcite{kai91}; \\pcite{cen94}; \\pcite{kay99}). Attempts at constraining evolution in the L--T relation, at redshifts above $z \\sim 0.1$, are complicated by the inherent difficulty of temperature estimation from often poor signal-to-noise data. \\scite{mus97}, using $ASCA$ data from a sample of 38 clusters at $z > 0.14$, found no significant evolution out to $z \\sim 0.3$. Prior to this work, the only attempt at quantifying evolution of the L--T relation to $z \\sim 0.3$ was that of \\scite{hen94}, whose results were also consistent with no evolution. \\scite{hen94} used low signal-to-noise spectra of 67 clusters observed with the $Einstein$ observatory, each individually spectrally fitted, and then averaged together in redshift bins. Recently \\scite{don99} have used $ASCA$ measurements of clusters drawn from the $Einstein$ Medium Sensitivity Survey (EMSS, \\pcite{gio90}; \\pcite{hen92}) to extend the conclusion of no evolution to $z > 0.5$. In this paper we present an L--T relation based on $ROSAT$ spectra of WARPS clusters, in various redshift and luminosity bins. Although this procedure is analogous to the one adopted by \\scite{hen94}, a more sophisticated simultaneous fitting method is adopted here. We discuss the data analysis techniques used to obtain temperature estimates from often low signal-to-noise $ROSAT$ PSPC data, and present results from simulated spectra to demonstrate their validity. In order to allow self-consistent discussion of any evolutionary effects in our relation, we compare our results with a local L--T relation over a wide range of temperatures and luminosities. ", "conclusions": "We have presented a cluster X-ray luminosity--temperature relation out to high redshifts ($z \\sim 0.8$), based on $ROSAT$ PSPC spectra of WARPS clusters. Due to the low signal-to-noise of many of our $ROSAT$ PSPC spectra, a multiple fitting technique was adopted in order to constrain temperatures. The 91 clusters analysed were binned into redshift and luminosity bins, and spectrally fit to a characteristic temperature. In order to analyse the evolution of the L--T relation, a large sample of low redshift ($z<0.2$) clusters was compiled from the literature. The best-fitting power law to the data at low redshift and at T$>$1 keV was found to be a power law of index $\\rm L \\propto T^{3.15\\pm0.06}$ (90 per cent confidence). This relation was then compared with the high redshift measurements to constrain possible evolution in the normalisation of the L--T relation. The binned $ROSAT$ measurements of WARPS clusters were consistent with no evolution of the L--T relation to $z \\sim 0.8$. When combined with previous cluster measurements from the literature they agree with the no evolution result of \\scite{don99}, but give stronger constraints; $\\eta=0.19\\pm0.38$ for $\\Omega_0=1$, or $\\eta=0.60\\pm0.38$ for $\\Omega_0=0.3$ where $\\rm L_{Bol} \\propto {(1+\\it z)}^{\\eta} \\rm T^{3.15}$. Using the entropy driven evolution model of \\scite{bow97} and \\scite{kay99}, a limit to the implied entropy evolution of cluster cores was found. Whilst the level of entropy evolution is slightly dependent on the value of the primordial density fluctuation spectrum, $n$, low amounts of evolution were favoured. The pure self-similar scaling expectations of \\scite{kai86} were found to be strongly excluded. The lack of evolution in the WARPS L--T relation, in conjunction with other recent results, such as a non-evolving XLF, implies either an open cosmology ($\\Omega_0<1$) or one closed by a cosmological constant, assuming $n<-1$. The forthcoming X-ray missions Chandra and XMM will allow more accurate temperature estimates of clusters, out to high redshifts, which will greatly improve our understanding of the L--T relation and its cosmological implications." }, "0003/astro-ph0003438_arXiv.txt": { "abstract": "We present a tabulation of 46 transient x-ray sources monitored with the All-Sky Monitor (ASM) on the Rossi X-ray Timing Explorer (RXTE). They fall into four broad categories: short ($\\sim1$~d), intermediate, and long ($>500$~d) duration of outbursts, and long period binary systems that flare up at periastron (e.g., Be systems). The mixture of outburst/quiescent cycles and low-level persistent emission in a few systems could indicate conditions are near the limit for stable mass flow in the accretion disk. The two short-time-scale systems, CI~Cam and V4641~Sgr, are within 1~kpc of the sun, and hence many more such systems may await discovery. ", "introduction": "The All-Sky Monitor (ASM; Levine et~al. 1996) on RXTE has been monitoring the entire sky for new (uncataloged) transient x-ray sources while also recording the intensities of the known sources. The current catalog contains about 325 source positions of which about 180 have yielded positive detections on some occasion. The monitoring has been reasonably continuous except for times when the sun is relatively close to a source and except for a period of $\\sim$7 weeks shortly after launch when the detectors were turned off due to a temporary breakdown problem. The detected sources include many well known persistent sources as well as a substantial number of transient sources. Some of these are recurrent and others are in their first known outburst. Most of the latter were discovered in the RXTE era, either with other satellites, {\\it e.g.} CGRO and BSAX, or with RXTE. Some were discovered prior to the launch of RXTE. Of the 180 positive detections, approximately 150 reached 15~mCrab on at least some occasion and 30 are detected at levels 2 to 15~mCrab in averages over long periods, up to 6~months. For sources with known positions, the detection threshold (3~sigma) away from the galactic center is about 30 mCrab in a single sweep of the ASM cameras across the source. A sweep usually consists of four 90-s integrations or ``snapshots'' as the cameras step across the source. The one-day threshold (typically 5 -- 8 sweeps) can reach down to $\\sim10$~mCrab. The data are routinely searched for new ({\\it i.e.,} not in the ASM catalog) sources with a cross-correlation search of the entire FOV. Confidence in the detection of a new persistent source arises through multiple detections that yield crossed lines of position. In one day, a 50-mCrab source is solidly established. Fainter sources to about 7~mCrab can be retrieved from cross-correlation maps that integrate one week of data. These thresholds apply to positions reasonably removed from bright sources. The list of detections include about 50 sources we call ``transients''. Another 23 objects are extragalactic (14~Sy1 and QSOs, 4~BL Lacs, and 5~clusters). About 40 objects exhibit periodicities in the ASM data from the spin period of X~Per (837 s) to the 164-d precession period of SS~433. ", "conclusions": "" }, "0003/astro-ph0003330_arXiv.txt": { "abstract": "We generalize to elliptical models the argument of Kuijken (1997), which connects the microlensing optical depth towards the Galactic bulge to the Galactic rotation curve. When applied to the latest value from the MACHO collaboration for the optical depth for microlensing of bulge sources, the argument implies that the Galactic bar cannot plausibly reconcile the measured values of the optical depth, the rotation curve and the local mass density. Either there is a problem with the interpretation of the microlensing data, or our line of sight to the Galactic centre is highly atypical in that it passes through a massive structure that wraps only a small distance around the Galactic centre. ", "introduction": "Searches for gravitational microlensing events, over 500 of which have now been observed, provide an important constraint on the mass in the inner Galaxy. Deriving a mass from an observed optical depth is not straightforward because one usually has only limited knowledge of the distances to the stars that are lensed, and even less information about the distances to the deflecting objects. Kuijken (1997) showed that the {\\it minimum\\/} mass required to generate a given optical depth towards the Galactic centre from an axisymmetric distribution of matter can be determined without any knowledge of the location of the lenses provided one knows the optical depth to a source on the Galaxy's symmetry axis. He showed, further, that the minimum mass required by the then available microlensing data was barely compatible with the measured rotation curve and local mass density. Here we generalize Kuijken's argument to the case of elliptical distributions of matter, which require less matter for a given optical depth. The apparent magnitude of a red-clump star at the Galactic centre is relatively bright, so the microlensing optical depth to such objects can be determined without significant uncertainty due to blending. Moreover, the red-clump stars must follow the general distribution of near infra-red light quite closely, because they are part of the population of evolved stars that are responsible for most of the Galaxy's near-IR luminosity (McWilliam \\& Rich, 1994). Given this, it proves possible to deduce from the measured mean optical depth of the clump giants the optical depth to a source on the Galaxy's axis. From 13 lensing events Alcock et al.\\ (1997) inferred $\\tau = 3.9^{+1.8}_{-1.2}\\times 10^{-6}$ for clump giants centred on $(l,b) = (2.55\\deg, -3.64\\deg)$. From a difference image analysis of observations that include 99 events, Alcock et al.\\ (2000b) recently measured $\\tau=2.9^{+0.47}_{-0.45}\\times10^{-6}$ for a mix of stars centred on $(l,b)=(2.68\\deg,-3.35\\deg)$, and from this measurement deduced for the same direction $\\tau=(3.88\\pm0.6)\\times10^{-6}$ for bulge sources after making allowance for the lower optical depth for lensing of disk stars. Here we use our results to show that such large values of the optical depth for microlensing of bulge sources cannot be reconciled with measurements of the rotation curve and local mass density, even by elliptical Galaxy models. The paper is organized as follows. In \\S2 we rederive Kuijken's result, generalize it to elliptical systems, and demonstrate that it is applicable to the measured clump-giant optical depth. In \\S3 we show that axisymmetric models are very clearly excluded. In \\S4 we show that barred models can be excluded, albeit with somewhat less confidence. ", "conclusions": "" }, "0003/astro-ph0003089_arXiv.txt": { "abstract": "Spectroscopic observations of a sample of 42 flat-spectrum radio sources from the Cosmic Lens All-Sky Survey (CLASS) have yielded a mean redshift of $ = 1.27$ with an RMS spread of $0.95$, at a completeness level of $64\\%$. The sample consists of sources with a 5-GHz flux density of $25-50$ mJy, making it the faintest flat-spectrum radio sample for which the redshift distribution has been studied. The spectra, obtained with the Willam Herschel Telescope (WHT), consist mainly of broad-line quasars at $z>1$ and narrow-line galaxies at $z<0.5$. Though the mean redshift of flat-spectrum radio sources exhibits little variation over more than two orders of magnitude in radio flux density, there is evidence for a decreasing fraction of quasars at weaker flux levels. In this paper we present the results of our spectroscopic observations, and discuss the implications for constraining cosmological parameters with statistical analyses of the CLASS survey. ", "introduction": "\\label{sec:intro} The number of gravitational lens systems found in systematic surveys provides a strong constraint on cosmological models, especially those with a cosmological constant (Kochanek 1996a; Falco, Kochanek and Mu\\~noz 1998, hereafter FKM; Helbig et al.\\ 1999). Statistical analyses of lens surveys currently give a 2$\\sigma$ upper limit on the cosmological constant in flat models ($\\Omega_{\\Lambda}$+$\\Omega_{M}$=1) of $\\Omega_{\\Lambda} < 0.62$ (FKM). These constraints are consistent with those obtained from Type Ia supernovae (eg.\\ Perlmutter et al.\\ 1999). Conducting a lens survey at radio frequencies has several advantages. Most importantly, radio lens searches do not suffer from the observational biases that plague optical surveys, particularly those due to dust extinction in the lensing galaxy. In addition, a survey of compact radio sources with the Very Large Array (VLA) provides maps with consistently high resolution ($0\\farcs25$) and dynamic range ($\\sim 100$), allowing lens systems with even large component flux density ratios ($\\sim 10:1$) to be easily identified in most snapshot observations. The short integration time required for such sources ($30$ seconds) means that large samples can be mapped in any given observing run. The drawback of radio lens surveys is that they are conducted in flux density ranges where there is little information on the source redshift distribution. Complete redshift surveys of radio sources exist only for those with $S_{\\rm{5~GHz}}>$ 300 mJy (eg.\\ the CJ samples, Henstock et al.\\ 1997; and the Parkes Half-Jansky Flat-Spectrum Survey (PHFS), Drinkwater et al.\\ 1997). The Cosmic Lens All-Sky Survey (CLASS; Myers et al.\\ 2000), the largest and most successful lens survey to date, has a flux density limit of $S_{\\rm{5~GHz}}\\geq30$ mJy. Kochanek (1996b) showed that a lack of redshift information at these flux levels will lead to serious systematic uncertainties in the derived cosmological constraints. FKM attempted to address this problem with three samples of flat-spectrum radio sources selected from the MIT-Greenbank Survey (Stern et al.\\ 1999) and the Jodrell-VLA Astrometric Survey (JVAS; Patnaik et al.\\ 1992; Browne et al.\\ 1998; Wilkinson et al.\\ 1998; King et al.\\ 1999). The 5-GHz flux density ranges of these samples were $50-100$, $100-200$ and $200-250$ mJy, respectively. They found a mean source redshift of $\\sim1.2$, with a decreasing fraction of identified quasars at weaker flux levels, falling to less than $50\\%$ of sources at 100 mJy. To extend the spectroscopic study of flat-spectrum radio sources to even weaker flux densities, a program of redshift determination for a small sample of CLASS sources was undertaken. The aim was to obtain information on the redshift distribution of the unlensed sources in CLASS. The same source population comprises the ``parent'' sample of lensed sources in the JVAS survey. This paper describes the selection and observation of the CLASS optical sample and discusses some preliminary results. ", "conclusions": "The number of gravitational lens systems found in a well-defined sample of sources can place strong limits on the cosmological constant. This is done by comparing the observed lensing rate with the predictions of various cosmologies. The optical depth for lensing depends both on the cosmological parameters and the redshifts of the sources being studied. Systematic uncertainties in the statistical analyses of radio lens surveys are therefore created by the paucity of data on the redshift distribution of unlensed sources fainter than 300 mJy. Recently Helbig et al.\\ (1999) analyzed the lensing rate in the JVAS statistical sample, which includes four lenses and approximately 2300 radio sources. At $95\\%$ confidence, the lower and upper limits on $\\Omega_{\\Lambda}$ in a flat universe are $-0.85$ and $0.84$, respectively. These results are consistent with other recent measurements of $\\Omega_{\\Lambda}$, both from lensing statistics and alternate cosmological probes such as Type Ia supernovae (Perlmutter et al.\\ 1999). Helbig et al.\\ assumed that the redshift distribution of the JVAS survey was identical to that of the CJ surveys, which were done at a higher flux density level. A similar assumption is harder to justify for CLASS, which searches for lensing among sources that are fainter than the CJ sample by more than an order of magnitude. Future statistical analyses of CLASS require a determination of the redshift distribution of flat-spectrum radio sources down to the $30$ mJy level. Our current results only partially address this issue. The measured redshifts of the CLASS sources appear to be consistent with those of brighter radio samples ($ \\approx 1.2 \\pm 0.1$). However, the mean redshift may change once completeness is improved, and the nature of the currently unidentified sources is determined. An underestimate of the mean redshift of the unlensed source population would lead to a overestimate of the cosmological constant, and vice versa. Consequently, the uncertainty in the redshift distribution of the CLASS sample permits large variations in the expected number of lenses for a given cosmological model. This is nicely illustrated by Kochanek (1996b). For radio luminosity functions constrained to produce four JVAS lenses, the mean redshift of a 50 mJy source varies from 0.4 for $\\Omega_{M}=0$, to 1.9 for $\\Omega_{M}=1$, to almost 4.0 for $\\Omega_{M}=2$ in a flat universe. Although these are extreme values, they underscore the importance of understanding the source redshift distribution when using the observed lensing rate as a probe of the cosmology. At this time, the constraints on cosmological parameters from the radio and optical lens surveys (and other tests) appear consistent, suggesting that the observational uncertainties are not overwhelming. The CLASS survey, however, pushes down into a lower flux density regime in which very little is known about the source redshift distribution. Therefore, a reliable and complete spectroscopic study of flat-spectrum radio sources for the entire flux density range from 30--300 mJy must be obtained if CLASS is to fulfill its promise as a cosmological tool. To this end, a new program of spectroscopic observations is currently being undertaken by the CLASS/CERES collaboration at La Palma." }, "0003/astro-ph0003276_arXiv.txt": { "abstract": " ", "introduction": "Neutron stars are among the most fascinating bodies in our universe. They contain over a solar mass of matter within a radius of $\\sim$ 10 km at densities of order $10^{15}$ g/cc. They probe the properties of cold matter at extremely high densities, and have proven to be fantastic test bodies for theories of general relativity. In a broader perspective, neutron stars and heavy ion collisions provide access to the phase diagram of matter at extreme densities and temperatures, that is basic for understanding the very early Universe and several other astrophysical phenomena. The discovery of the neutron by Chadwick in 1932 prompted Landau \\cite{rosenfeld} to predict the existence of neutron stars. The first theoretical calculations of neutron stars were performed by Oppenheimer and Volkoff \\cite{Oppe39} in 1939 assuming that they are gravitationally bound states of neutron Fermi gas. The calculated stars had a maximum mass of $\\sim$ 0.7 $M_\\odot$, central densities up to $\\sim 6 \\times 10^{15}$~g/cm$^3$ and radii $\\sim10$~km. For comparison the density of nuclear matter inside a large nucleus like $^{208}$Pb is $\\sim$ 0.16 nucleons/fm$^3$, {\\em i.e.} $\\simeq$ 2.7 $\\times 10^{14}$~g/cm$^3$ \\cite{BohrMo}. Their predicted maximum mass was less than the Chandrasekhar mass limit of $\\sim$ 1.4 $M_\\odot$ for white dwarfs made up of iron group nuclei, and having densities up to $\\sim 10^9$~g/cm$^3$ \\cite{ShapTu}. The pressure to balance the gravitational attraction in white dwarfs and Oppenheimer-Volkoff neutron stars is supplied by degenerate electron and neutron Fermi gases respectively. In 1934 Baade and Zwicky \\cite{Baad34} suggested that neutron stars may be formed in {\\it supernovae} in which the iron core of a massive star exceeds the Chandrasekhar limit and collapses. The large amount of energy released in the collapse blows away the rest of the star and the collapsed core may form a neutron star. For efficient production of neutron stars with this mechanism, the maximum mass of neutron stars should exceed 1.4 $M_\\odot$. In the 60's, using schematic models of nuclear forces, Tsuruta and Cameron \\cite{camsol} showed that they could increase the neutron star masses beyond 1.4 $M_\\odot$. Bell and Hewish discovered {\\it radio pulsars} in 1967, and they were soon identified as rotating neutron stars by Gold \\cite{Gold69}. The subsequent detection of the Crab pulsar in the remnant of the Crab supernova, observed in China in 1054 A.D., confirmed the link to supernovae, and initiated the present efforts to better understand neutron stars. \\subsection{A Brief Overview of Observations} Almost 1200 pulsars have been discovered by the turn of this millennium. In these stars the magnetic and rotational axes are misaligned, thus they emit dipole radiation in the form of radio waves that appear to pulse on and off like a lighthouse beacon as the pulsar beam sweeps across the Earth. The rotational energy loss due to dipole radiation is \\beq \\dot{E} = I\\Omega\\dot{\\Omega} = - \\frac{B^2R^6\\Omega^4\\sin^2\\theta}{6c^3} \\,, \\label{dE} \\eeq where the moment of inertia for a typical neutron star is $I\\sim 10^{45}$~g~cm$^2$. Pulsars have magnetic fields $B$ of $\\sim 10^{12}$ G, deduced from the observed $\\dot{\\Omega}$, and independently confirmed by cyclotron absorption lines found in X-ray spectra. Their periods, $P = 2 \\pi / \\Omega$, ranging from 1.5~ms to 8.5~s, are increasing with derivatives $\\dot{P}\\sim 10^{-12}-10^{-21}$. The pulsar age is approximately given by $ P/2\\dot{P}$ \\cite{Lorimer}; most pulsars are old and slowly rotating with relatively small period derivatives, except for a few young pulsars, e.g., those found in the Crab and Vela nebulae. In 1969 the Crab and the Vela pulsars were observed to {\\it \"glitch\"}, {\\em i.e.} to suddenly speedup with period changes $\\Delta P/P$ of the order of $10^{-8}$ and $10^{-6}$ respectively \\cite{Boyn72}. In post-glitch relaxation most of the period increase $\\Delta P$ decays. These pulsars have glitched several times since then. The glitches suggest that the neutron stars have a solid crust containing superfluid neutrons. The interesting structure of their crust has been recently reviewed \\cite{prannr}, and we discuss it rather briefly in this report. The first {\\it binary} of two {\\it pulsars} was found by Hulse and Taylor in 1973 and they could determine many of its parameters including both masses, orbital period and period derivative, orbital distance and inclination. General relativity could be tested to an unprecedented accuracy by measuring the inward spiralling of the neutron stars in the Hulse-Taylor binary PSR 1913+16 \\cite{HulseT}. The periastron advance in PSR 1913+16 is 4.2$^\\circ$ per year as compared to 43'' per century for Mercury, which originally was used by Einstein to test his theory of general relativity. Six double neutron star binaries are known so far, and neutron stars in all of them have masses in the range $1.36\\pm 0.08 M_\\odot$ \\cite{Thor99}. They confirm that nuclear forces have a large effect on the structure of neutron stars and increase their maximum mass beyond 1.4 $M_\\odot$. Neutron stars are estimated to have a binding energy of $\\sim 10 \\%$ of their mass. Thus $\\sim 1.5$ $M_\\odot$ of nuclei are needed to obtain a $1.35$ $M_\\odot$ star. A distinct subclass of radio pulsars are {\\it millisecond pulsars} with periods $\\la 100$ ms. The fastest pulsar known has a period of 1.56~ms \\cite{fastes}. The period derivatives of millisecond pulsars are very small corresponding to low magnetic fields $\\sim 10^8-10^{10}$G. They are believed to be recycled pulsars, i.e.\\ old pulsars that have been spun up by mass accretion whereby the magnetic fields have decayed. About 80\\% of the millisecond pulsars are in binaries whereas less than 1\\% of normal radio pulsars are in binaries. About 20 - almost half of the millisecond pulsars - are found in binaries where the companion is either a white dwarf or a neutron star. With X-ray detectors on board satellites since the early 1970's about two hundred {\\it X-ray pulsars and bursters} have been found of which the rotational period has been determined for about sixty. The X-ray pulsars and bursters \\cite{BildStroh} are believed to be neutron stars accreting matter from high ($M\\ga 10M_\\odot$) and low mass ($M\\la 1.2M_\\odot$) companions respectively. The X-ray pulses are attributed to strong accretion on the magnetic poles emitting X-rays (as northern lights). The observed radiation is pulsed with the rotational frequency of the accreting star. {\\em X-ray bursts} are thermonuclear explosions of accreted matter on the surface of neutron stars. After accumulating hydrogen on the surface for hours, pressure and temperature become sufficient to trigger a runaway thermonuclear explosion seen as an X-ray burst that lasts a few seconds \\cite{BildStroh}. Masses of these stars are less accurately measured than for binary pulsars. We mention recent mass determinations for the X-ray pulsar Vela X-1: $M=1.87^{+0.23}_{-0.17} \\ M_\\odot$ \\cite{Barziv}, and the burster Cygnus X-2: $M=1.8\\pm0.4)M_\\odot$ \\cite{Orosz}. They are larger than the typical $1.36 \\pm 0.08 M_\\odot$ masses found in pulsars binaries, presumably due to accreted matter. A subclass of half a dozen {\\em anomalous X-ray pulsars} has been discovered. They are slowly rotating, $P\\sim 10$~sec, but rapidly slowing down. This requires huge magnetic fields of $B\\sim10^{14}$~G and they have appropriately been named ``magnetars'' \\cite{Duncan}. Four gamma ray repeaters discovered so far are also believed to be slowly rotating neutron stars. The magnetars and likely also the gamma ray repeaters reside inside supernova remnants. Recently, {\\em quasi-periodic oscillations} (QPO) have been found in 12 binaries of neutron stars with low mass companions. If the QPO originate from the innermost stable orbit \\cite{zss97,miller} of the accreting matter, their observed values imply that the accreting neutron stars have masses up to $\\simeq 2.3M_\\odot$. In this case the QPO's also constrain the radii of the accreting star. Non-rotating and non-accreting neutron stars are virtually undetectable but the Hubble space telescope has observed one thermally radiating neutron star \\cite{Walter}. Its surface temperature is $T\\simeq 6\\times 10^5$~K$\\simeq 50$~eV and its distance is less than 120~pc from Earth. Circumstantial evidence indicate a distance of $\\sim80$~pc which leads to a radius of 12-13~km for this star. In recent years much effort has been devoted to measuring pulsar temperatures, especially with the Einstein Observatory and ROSAT. Surface temperatures of a few pulsars have been measured, and upper limits have been set for many \\cite{pethick}. >From the human point of view supernova explosions are rare in our and neighboring galaxies. The predicted rate is 1-3 per century in our galaxy and the most recent one was 1987A in LMC. No neutron star associated with this explosion has been detected; however, 19 neutrinos were detected on earth from 1987A \\cite{1987an}, indicating the formation of a ``proto-neutron star''. It has been suggested by Bethe and Brown \\cite{Bebhole} that an upper limit to the mass of neutron stars can be obtained assuming that the remnant of SN 1987A collapsed into a black hole. Astrophysicists expect a large abundance of $\\sim 10^8$ neutron stars in our galaxy. At least as many supernova explosions, responsible for all heavier elements present in our Universe today, have occurred. The scarcity of neutron stars in the solar neighborhood may be due to production of black holes or other remnants in supernovae, or due to a high initial velocity (asymmetric ``kick'') received during their birth in supernovae. Recently, many neutron stars have been found far away from their supernova remnants; and of the $\\sim1200$ discovered radio pulsars only about $\\sim10$ can be associated with the 220 known supernova remnants. Neutron stars thrown out of the galactic plane may be detected by gravitational microlensing experiments \\cite{gmicro} designed to search for dark massive objects in the galactic halo. The recent discovery of afterglow in {\\em Gamma Ray Bursters} (GRB) allows determination of their very high redshifts ($z\\ge 1$). They imply that GRB occur at enormous distances. Evidence for beaming has been observed \\cite{Kulkar}, and the estimated energy output is $\\sim 10^{53}$ ergs. Such enormous energies can be produced in neutron star mergers eventually forming black holes. From abundance of binary pulsars one can estimate the rate of neutron star mergers; it is compatible with the rate of GRB of approximately one per day. Another possible mechanism, is a special class of type Ic supernova ({\\it hypernovae}) where cores collapse to black holes \\cite{hypernovae}. {\\em The future} of neutron star observations looks bright as new windows are about to open. A new fleet of X- and Gamma-ray satellites have and will be launched. With upgraded ground based observatories and detectors for neutrinos and gravitational waves \\cite{LIGO} our knowledge of neutron star properties will be greatly improved. \\subsection{Theory of Neutron Star Matter} Neutron stars are made up of relatively cold, charge neutral matter with densities up to $\\sim 7$ times the equilibrium density $\\rho_0$ = 0.16 nucleons/fm$^3$ of charged nuclear matter in nuclei. The matter density is $> \\rho_0$ over most of the star, apart from the relatively thin crust \\cite{prannr}. The Fermi energy of neutron star matter is in excess of tens of MeV, and hence, at typical temperatures of $\\la$ KeV, thermal effects are a minor perturbation on the gross structure of the star. Matter at such densities has not yet been produced in the laboratory, its properties must be theoretically deduced from the available terrestrial data with guidance from observed neutron star properties. The quantities of interest are the phase and composition of cold catalyzed neutral dense matter, its energy density $\\epsilon (\\rho)$ and pressure $P(\\rho)$, where $\\rho$ denotes the baryon number density. The baryon number is conserved in all known interactions, therefore it is convenient to find the composition by minimizing the total energy $E_T(\\rho)$ per baryon, including rest mass contributions. This gives: \\beq \\epsilon(\\rho)=\\rho E_T(\\rho), \\ \\ \\ \\ P(\\rho)= \\rho^2 \\frac{\\partial E_T(\\rho)} {\\partial \\rho}. \\label{eq:eandp} \\eeq The equation of state (EOS) $P(\\epsilon)$ is found by eliminating $\\rho$ from the above two. The gravitational equilibrium of a nonrotating star is described by the Tolman-Oppenheimer-Volkoff (TOV) \\cite{ShapTu} Eq: \\beq \\frac{dP(r)}{dr} = - \\frac{G(\\epsilon(r)+P(r)/c^2)(m(r)+4 \\pi r^3 P(r)/c^2)} {r^2 (1-2Gm(r)/rc^2)} \\ , \\label{eq:tov} \\eeq where $G$ is the gravitational constant, $P(r)$ and $\\epsilon(r)$ are the pressure and mass density at radius $r$ in the star, and \\beq m(r) = \\int_0^r 4 \\pi r^{\\prime 2} \\epsilon (r^{\\prime}) dr^\\prime \\ , \\label{eq:minr} \\eeq is the mass inside $r$. If we neglect the general relativistic corrections of order $1/c^2$ the TOV Eq. reduces to the Newtonian hydrodynamic equation. The TOV Eq. can be easily integrated starting from the central density $\\epsilon_c$ ar $r=0$ to find the density profile $\\epsilon(r)$. At the radius $R$ of the star $P(R) = 0$, and $m(R) = M$ is the mass of the star as seen from outside. The stability of the star can be deduced from the $M(\\epsilon_c)$ as discussed in \\cite{ShapTu}, and the equations for rotating stars are given by \\cite{rotats}. The effect of rotation on the structure of most observed neutron stars seems to be rather small, however, it could be significant at periods less than a millisecond \\cite{cookstu}. At densities $< 2 \\times 10^{-3} \\rho_0 $ matter is believed to have the form of a lattice of nuclei in a relativistic degenerate electron gas \\cite{prannr}, qualitatively similar to that of metals. The main focus of the theory reviewed here has been on determining the properties and EOS of matter in the density range $2 \\times 10^{-3} \\rho_0 < \\rho < 10 \\rho_0$ from terrestrial data. In the lower part of this range we expect to find nucleon matter (NM) composed of nucleons and electrons. In contrast to matter in nuclei, it has mostly neutrons with a small fraction of protons and equal number of electrons to maintain charge neutrality. The large Fermi energy, $\\mu_e \\sim$ 100 MeV, of the electron gas limits the fraction of protons in NM. At higher densities there are several possibilities including condensation of negatively charged pions and kaons, occurrence of hyperons, and the transition from hadronic to quark matter. All these possibilities exploit the large electron Fermi energy of NM, therefore only one of these, if any, may occur and lower the $\\mu_e$. In addition, neutron star matter can have interesting mixed phase regions in which the mixing phases are charged but the matter is overall neutral \\cite{prannr}. We begin with a review of NM, and later consider the more exotic possibilities. In the last sections the range of neutron star structures predicted by theory is presented along with a comparison with the observational data. ", "conclusions": "Since the discovery of pulsars a significant effort has been devoted to accurately calculate properties of dense NM from realistic models of nuclear forces. Exact calculations of NM are still out or reach, however the new AFDMC methods (sect. 3.7) may eventually succeed. The present variational upper bounds seem to be above the true energies by $\\sim$ 12 \\%. Such an error does not have serious consequences on the predicted properties of neutron stars. For example, an EOS obtained by reducing the variational energies, without rest mass terms, by 12 \\% reduces the maximum mass of A18+$\\delta v$+UIX$^*$ model by 2.3 \\% to 2.14 $M_{\\odot}$, and the radius of 1.4 $M_{\\odot}$ star by 2.9 \\% to 11.2 km. Larger uncertainties stem from the fact that the double $\\pi^0$ and $\\pi^-$ condensation scenario illustrated in Fig.11 has not yet been calculated with realistic interactions, though it appears unlikely that it will influence the NM EOS by much more than 10 \\%. Local models of two-nucleon interaction seem to be now converging. The predictions based on the 1988 calculations with Argonne 14 interaction are not too different from those of the 1998 calculations with the more accurate A18. It also seems likely that the local models give a fairly accurate description of two-nucleon interaction. A concern is that the present models of TNI are based on fits to a rather limited set of data, and are not as precise as the NN-interaction models. However, addition of the UIX$^*$ TNI to the A18+$\\delta v$ increases the maximum mass by $\\sim$ 20 \\% and $R(1.4)$ by 13 \\% (Table 2). These changes may be important but they are not very large. The present models of kaon-nucleon and hyperon-nucleon interactions are based on very limited data, and we have none on $K^-NN$ and $\\Sigma^-NN$ three-body forces. These could have significant effect on the threshold densities for kaons and hyperons to appear in dense matter. Hopefully advances in QCD and quark-models will provide a more rigorous framework to describe these interactions, and calculate properties of quark matter. The bag model estimates of QM EOS may have significant corrections at densities of interest in neutron stars. >From present observations there seem to be three possible scenarios for the limiting mass of neutron stars. If QPO's are indeed due to accretion from the innermost stable orbit, then the NM predictions of $M_m \\sim 2.2 M_{\\odot}$ are reasonable, and strange baryons and quark drops do not soften the EOS of matter at $\\rho < 7 \\rho_0$ significantly. If the Vela X-1 mass measurement is correct, but QPO's have some other origin, then $M_m$ could be $\\sim$ 1.8 $M_{\\odot}$, indicating some softening of the NM EOS. However, if the present interpretation of QPO's and Vela X-1 mass measurements are both faulty, and $M_m$ is as small as 1.56 $M_m$ as estimated from the absence of a neutron star in SN 1987A, then a significant softening of the NM EOS by phase transitions is indicated. Further observations will hopefully clear this situation. Phase transitions such as NM to QM, can soften the EOS significantly. Fortunately these can have a measurable effect on the spin down of a rapidly rotating star in favorable cases, as has been recently pointed out \\cite{gpw97,rot}. Consider the case of a rapidly rotating star whose central density is close to a first order phase transition. As the star slows and the central pressure increases due to decrease of the centrifugal force, the core matter will change its phase and become more dense at a critical angular velocity $\\Omega_c$. This decreases the moment of inertia, which assumes the characteristic form: \\begin{equation} I = I_0\\left( 1 +c_1\\Omega^2-c_2(\\Omega_c^2-\\Omega^2)^{3/2} + ... \\right) . \\label{Igen} \\end{equation} for $\\Omega < \\Omega_c$. Here, $c_1$ and $c_2$ are small parameters proportional to the density difference between the two phases, and $c_2=0$ for $\\Omega>\\Omega_c$. In order to make contact with observation, the temporal behavior of angular velocities must be considered. The pulsars slow down at a rate given by the loss of rotational energy, believed to be given by: $d(\\frac{1}{2}I\\Omega^2)/dt\\propto -\\Omega^{n+1}$, where $n=3$ for dipole radiation, Eq. (\\ref{dE}) and $n=5$ for gravitational radiation. With the moment of inertia given by Eq. (\\ref{Igen}) the angular velocity can be calculated. The corresponding braking index, $n(\\Omega)=\\ddot{\\Omega}\\Omega/\\dot{\\Omega}^2$, depends on the second derivative of the moment of inertia, $I''=dI/d^2\\Omega$. Using Eq. (\\ref{Igen}) we obtain: \\begin{eqnarray} n(\\Omega) \\simeq n - c_1\\Omega^2 +c_2\\frac{\\Omega^4}{\\sqrt{\\Omega_c^2-\\Omega^2}} \\,.\\label{n} \\end{eqnarray} which exhibits a characteristic $(\\Omega_c-\\Omega)^{-1/2}$ singularity as $\\Omega$ approaches $\\Omega_c$ from below. Observations of the braking index of a rapidly rotating, new born pulsar would be very interesting. All realistic NM EOS predict that the radius of neutron stars with a mass of 1 to 1.5 $M_{\\odot}$ is $\\sim$ 11 to 12 km. Future high resolution Chandra and XMM space observatories will hopefully be able to measure black-body spectra and detect gravitationally redshifted spectral lines from several stars. Such observations will help determine masses, radii and temperatures uniquely if the distance of the star is known. It is important to know the radius of a 1.4 $M_{\\odot}$ star, because that would test the EOS in the $\\rho \\la 3 \\rho_0$ region in which large modifications of NM EOS are not expected on the basis of our present, naive estimates of kaon-nucleon and $\\Sigma^-$-nucleon interactions." }, "0003/hep-ex0003006_arXiv.txt": { "abstract": "Based on the unified analyses of KASCADE, AKENO, EAS-TOP and ANI EAS size spectra, the approximations of energy spectra of different primary nuclei have been found. The calculations were carried out using the SIBYLL and QGSJET interaction models in 0.1-100 PeV primary energy range. The results point to existence of both rigidity-dependent steepening energy spectra at $R\\simeq200-400$ TV and an additional proton (neutron) component with differential energy spectrum $(6.1\\pm0.7)\\cdot10^{-11}(E/E_{k})^{-1.5}$ (m$^2\\cdot$s$\\cdot$sr$\\cdot$TeV)$^{-1}$ before the knee $E_{k}=2030\\pm130$ TeV and with power index $\\gamma_{2}=-3.1\\pm0.05$ after the knee. ", "introduction": "High statistical accuracy in modern EAS experiments in the knee region encouraged the investigation of the fine structure of EAS size spectra. Although the origin of the knee is still a matter of debate, recently the series of publications \\cite{EW} appeared, where the sharpness and the spectral structure in the knee region were interpreted by the contribution of heavy nuclei from a single local supernova. Along with this, the absolute differential EAS size spectra measured at different atmosphere depths and different zenith angles are not explained yet from the point of view of a single $A-A_{Air}$ interaction model and a single model of primary energy spectra and elemental composition. Such an attempt has been made in \\cite{KHR} based on the QGS interaction model and rigidity-dependent steepening primary energy spectra \\cite{Peters} for a description of vertical MSU, AKENO and Tien-Shan EAS size spectra. In the present work we worked out a formalism of the inverse problem solution - reconstruction of the primary energy spectrum and elemental composition based on the known EAS size spectra of KASCADE \\cite{KAS}, AKENO \\cite{AKE}, EAS-TOP \\cite{TOP} and ANI \\cite{ANI} measured at different zenith angles. The calculations were done in the frames of QGSJET \\cite{QGS} and SIBYLL \\cite{SIB} interaction models. As a primary spectrum we have tested the modified rigidity-dependent steepening primary energy spectra and the hypothesis of the additional component in the knee region \\cite{JK,JL,TG,PB2}. In this case the type of nucleus of the additional component was considering as unknown and determining by the best fit of the fine structure of EAS size spectra in the knee region. ", "conclusions": "" }, "0003/astro-ph0003383_arXiv.txt": { "abstract": "We present deep \\hst\\ imaging, as well as ground-based imaging and spectroscopy, of the optical afterglow associated with the long-duration gamma-ray burst GRB~990712 and its host galaxy. The data were obtained 48--123 days after the burst occurred. The magnitudes of the host ($R=21.9$, $V=22.5$) and optical afterglow ($R=25.4$, $V=25.8$, 47.7 days after the burst) favor a scenario where the optical light follows a pure power-law decay with an index of $\\alpha \\sim -1.0$. We find no evidence for a contribution from a supernova like SN1998bw. This suggests that either there are multiple classes of long-duration gamma-ray bursts, or that the peak luminosity of the supernova was $> 1.5$ mag fainter than SN1998bw. The \\hst\\ images and EFOSC2 spectra indicate that the gamma-ray burst was located in a bright, extended feature (possibly a star-forming region) 1.4 kpc from the nucleus of a $0.2 L^{\\ast}_B$ galaxy at $z= 0.434$, possibly a Seyfert~2 galaxy. The late-time afterglow and host galaxy of GRB~990712 bear some resemblance to those of GRB~970508. ", "introduction": "The spatial association of GRB~980425 with the unusual Type Ib/c supernova SN1998bw at $z=0.0085$ provided the first tantalizing evidence that some gamma-ray bursts (GRBs) are related to the end-stages of the lives of massive stars \\citep{GVV1998}. Recent evidence for similar supernova (SN) signatures in the late ($\\sim 15(1+z)$ days) light curves of the genuine cosmological GRBs GRB~970228 \\citep{D1999,R1999,GNV1999} and GRB~980326 \\citep{CTG1999,BKD1999} indicates that at least some long-duration GRBs are related to SN explosions. A GRB--SN association suggests that the progenitors of GRBs are short-lived and that GRBs die where they were born---in the star-forming regions of their host galaxies \\citep{P1998}. \\citet{HH1999} found evidence from {\\sl Hubble Space Telescope\\/} ({\\hst}) Space Telescope Imaging Spectrograph (STIS) imaging for a spatial coincidence between GRB~990123 and a star-forming region in its host galaxy. The association of GRBs with star-forming regions is important for models of their progenitors and can be used to probe the physics of star formation and the global star-formation history of the Universe \\citep{MM1998,T1999,BN1999}. GRB~990712 was first localized by {\\sl BeppoSAX\\/} and detected as having the strongest $X$-ray afterglow observed to date \\citep{HIT1999}. \\citet{BSM1999} discovered a bright, decaying optical afterglow (OA) ($R = 19.4 \\pm 0.1$) four hours after the burst. \\citet{GVR1999} measured a preliminary redshift of $z = 0.430\\pm0.005$ from a set of absorption and emission lines, which makes it the nearest GRB with a secure redshift that has been observed to date (apart from SN1998bw). ESO New Technology Telescope images obtained 3.7 days after the burst led \\citet{HCC1999} to hypothesize the existence of a bright host galaxy with $R = 22$ on the grounds of an apparent leveling off of the light curve relative to the suspected power-law decline ($\\alpha = -1.05$) \\citep{KH1999} of the OA\\@. Subsequent ESO Very Large Telescope imaging \\citep{HFD1999} confirmed the leveling off of the light curve, and yielded evidence for the existence of an extended object contributing to the flux at the position of the GRB\\@. \\citet{HFD1999} predicted that the existence of a SN would lead to a bump in the light curve around 1 August 1999, and that a SN model could be distinguished from a no-SN model in late \\hst\\ and ground-based imaging, as the OA would be brighter and the host fainter in the SN scenario than in the no-SN scenario. These predictions are presented in \\citet{S1999}, which reports the discovery and early light curve of the OA of GRB~990712. In this {\\paper} we present late {\\hst} imaging, as well as ground-based imaging and spectroscopy, aimed at testing these predictions. At the time of the {\\hst} observations the $R$-band magnitudes of the host galaxy and the OA are predicted to be $22.25\\pm0.05$ and $<23.91\\pm0.05$ in the SN scenario, and $21.75\\pm0.05$ and $25.39\\pm0.1 $ in the no-SN scenario. We assume a standard Friedman cosmology with $H_0 = 65\\ \\ho$, $\\Omega_0 = 0.2$, and $\\Lambda = 0$. At $z = 0.4337$ this corresponds to a scale of 5.6 proper kpc per arcsecond, a luminosity distance of 2.37 Gpc, a distance modulus of 41.88, and a look-back time of 4.9 Gyr. Including a cosmological constant of $\\Lambda = 0.8$ increases these values by $\\sim 10$\\%. ", "conclusions": "" }, "0003/hep-ph0003163_arXiv.txt": { "abstract": " ", "introduction": "There is now strong evidence in support of an oscillatory behavior of neutrinos. The results on atmospheric neutrinos from SuperKamiokande \\cite {ska} find a comprehensive explanation in terms of oscillations, indicating a non-zero neutrino mass. This result and the solar neutrino problem, which has been held as a signal for neutrino oscillations for long, have created much excitement in the community. In this paper we examine the latest solar neutrino data from SuperKamiokande \\cite{sk1117} along with those from the radiochemical Chlorine and Gallium experiments assuming that MSW resonant flavor conversion \\cite{msw} is operative. The above experiments have presented the measured arrival rates \\cite{solar} which, in all the cases, are less than the predictions from the Standard Solar Model (SSM). In addition, SuperKamiokande (SK) has provided the observed electron energy distribution \\cite{sk1117,sksolar}. We use these data to test the consistency of the MSW mechanism taken together with the SSM predictions in a two flavor scenario. There are several recent MSW analyses of the SK solar neutrino data \\cite{sk825} - \\cite{gg4}. Although the data fitting method followed in all of them is to minimize a $\\chi^2$ function, the details of the statistical procedure used vary among the different groups. In this work, we indicate two different ways of performing the statistical analysis for the spectrum and the combined rate and spectrum data and check the consistency of the best-fit values of the mass squared differences and the mixing angles so obtained. For the analysis of the spectrum results we explore the possibility of using its moments as variables for fitting the data. The main advantage of these moments is that they probe the shape of the spectrum in a manner independent of the $^8$B flux normalization uncertainties. For the combined analysis of the rates and the spectrum, apart from the standard procedure of varying the $^8$B flux normalization and treating these two sets of data as independent, we also adopt a second method which takes into account the correlations among the rates and the spectrum data. In addition to the best-fit values of the oscillation parameters, we present the 99\\% C.L. and 90\\% C.L. allowed regions and the goodness of fit (g.o.f) of a particular solution. By g.o.f. is meant the probability that the $\\chi^2$ will exceed $\\chi^2_{min}$. When presenting the allowed region we take $\\chi^2_{min}$ to be the value at the {\\em global} minimum in that region\\footnote{The other approach is to present the allowed regions with respect to the {\\em local} minimum.}. For the neutrino fluxes and the neutrino production positions within the sun we use the BP98 solar model \\cite{bp98}. We consider oscillation of $\\nu_e$ to a sequential ($\\nu_\\mu$ or $\\nu_\\tau$) neutrino. This paper is organized as follows. In the next section we present the formulae for oscillation of neutrinos with the inclusion of matter effects both in the sun and during their passage through the earth. In section 3 we use the data on the total solar neutrino rates as measured at the Chlorine, Gallium, and SuperKamiokande (1117-day data) detectors to obtain the best-fit values of the neutrino mass splitting and the mixing angles. In section 4 we consider the electron energy spectrum observed at SuperKamiokande. Using the MSW predictions, we obtain the best-fit values from a direct fit to the data as well as from a fit to the normalized moments. In section 5 we use both the total rates data and the SK electron energy spectrum data to make a combined fit. As noted earlier, here we allow the normalization of the $^8$B spectrum to vary and compare the results with those obtained when the SSM prediction for this normalization is used, allowing the inclusion of correlations between the total rates and the observed spectrum {\\em via} astrophysical uncertainties. We end in section 7 with a summary, discussions, and conclusions. ", "conclusions": "In this paper we have performed a detailed $\\chi^2$-analysis of the latest SK solar neutrino data together with the results from the Cl and Ga experiments in terms of two-generation MSW conversions of $\\nu_e$ to sequential ($\\nu_\\mu$, $\\nu_\\tau$) neutrinos. Compared to the recent analyses in the literature \\cite{sk825}-\\cite{gg4} there are two new features in our analysis. \\begin{itemize} \\item We fit the observed electron energy spectrum data in two different ways, exploring for the first time, the use of moments of the energy spectrum in a $\\chi^2$-analysis. \\item The combined fits to the total rates and spectral data are also performed in two different manners. In the first, the $^8$B flux normalization is used as a free parameter while in the other the SSM normalization is chosen for it and correlations between the rates and spectral data due to astrophysical uncertainties of the $^8$B flux are included. \\end{itemize} We find that the two-generation MSW scenario can well explain the data on total rates. The solution in the SMA (Small Mixing Angle) region is preferred over the other possibilities although the quality of the fit is poorer as compared to the one obtained using the 825-day SK data. The best-fit from the spectrum data comes in a region disallowed from the total rates. In this region the $^8$B neutrinos are suppressed much more than required by the rates data. For the analysis of the spectrum, the absolute normalization of the $^8$B flux, $X_B$, has been permitted to be greater than unity, thus effectively compensating the shortfall. We have explored the use of normalized moments of the observed electron energy spectrum to signal MSW resonant flavour conversion. These variables are independent of the absolute normalization of the $^8$B flux and probe the effect of oscillations on the spectral shape. This procedure is somewhat handicapped by the large errors on the moments calculated from the present data. However, the best-fit values obtained by the two methods are more or less in agreement. Similarly, for the two methods followed in the combined $\\chi^2$ analysis of the rates and time averaged spectrum data, the best-fit values are not much different. The first approach gives a better fit because we utilise the freedom of varying the $^8$B flux normalization. We remark that in the combined analysis, where the $^8$B normalization is held fixed at the SSM value, the correlations between the rates and the spectrum data are found to be important and thus one should use caution regarding results obtained treating these as independent. For both methods, the best-fit from the combined analysis falls in the LMA region. Compared to the rates analysis the goodness of fit of the LOW(SMA) region increases(decreases). With the inclusion of the day-night dependence of the data the goodness of fit in the SMA region worsens further \\cite{sk1117}. In this work we have not included the new GNO result \\cite{gno} which is consistent with the Gallex and SAGE data. Thus its inclusion is not expected to affect the conclusions drastically. For illustration we give below the results of the global analysis of rates and spectrum including the GNO data. We take the weighted average of Gallex and GNO and treat SAGE as a separate experiment. The best-fit values and $\\chi^2_{min}$ that we get are: \\begin{itemize} \\item $\\sin^2 \\theta = 5.26 \\times 10^{-4},~~ \\Delta m^2 = 5.28 \\times 10^{-6}$ eV$^2,~~ X_{B} = 0.61,~~ \\chi^2_{min} = 12.73$,~~ g.o.f = 85.21\\% ~~~~(SMA) \\item $\\sin^2 \\theta$ = 0.18,~~ $\\Delta m^2 = 2.48 \\times 10^{-5}$ eV$^2$,~~ $X_{B}$ = 1.39,~~ $\\chi^2_{min}$ = 11.55,~~ g.o.f = 90.39\\%~~~~(LMA) \\item $\\sin^2 \\theta$ = 0.41,~~ $\\Delta m^2 = 9.39 \\times 10^{-8}$ eV$^2$,~~ $X_{B}$ = 0.89,~~ $\\chi^2_{min}$ = 19.85,~~ g.o.f = 40.34\\%~~~~(LOW) \\end{itemize} Thus the global best-fit continues to be in the LMA region. In conclusion, we have probed the most recent solar neutrino data on total rates and the observed electron energy spectrum at SK from various angles within the framework of MSW flavour conversion. We find good fits in some situations but a degree of uncertainty still remains since different fits do not prefer the same values of the parameters. More data from the running and new experiments, it is hoped, will further sharpen the results in the near future.\\\\ \\parindent 0pt {\\large{\\bf {Acknowledgements}}}\\\\ D.M. and A.R. are partially supported by the Eastern Centre for Research in Astrophysics, India. A.R. also acknowledges a research grant from the Council of Scientific and Industrial Research, India. We would like to thank Sandhya Choubey for pointing out an error in one of our computer codes and J.W.F. Valle for drawing our attention to the updated analysis in \\cite{valle}. S.G. would like to thank Plamen Krastev for many helpful correspondences. \\vskip 3mm" }, "0003/hep-ph0003025_arXiv.txt": { "abstract": "We show by studying the Abelian Higgs model with numerical lattice simulations that non-thermal phase transitions arising out of preheating after inflation are possible in gauge-Higgs models under rather general circumstances. This may lead to the formation of gauged topological defects and, if the scale at which inflation ends is low enough, to electroweak baryogenesis after preheating. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003293_arXiv.txt": { "abstract": "We present the results of a 0.5-0.9$\"$ FWHM imaging survey at K (2.2$\\mu$m) and H (1.6$\\mu$m) covering $\\sim$5.1'$\\times$5.1' centered on $\\theta^1C$ Ori, the most massive star in the Orion Nebula Cluster (ONC). At the age and distance of this cluster, and in the absence of extinction, the hydrogen burning limit (0.08 M$_\\odot$) occurs at K$\\approx$ 13.5 mag while an object of mass 0.02 M$_\\odot$ has K$\\approx$ 16.2 mag. Our photometry is complete for source detection at the 7$\\sigma$ level to K$\\approx$17.5 mag and thus is sensitive to objects as low-mass as 0.02$M_\\odot$ seen through visual extinction values as high as 10 magnitudes. We use the observed magnitudes, colors, and star counts to constrain the shape of the inner ONC stellar mass function across the hydrogen burning limit. After determining the stellar age and near-infrared excess properties of the optically visible stars in this same inner ONC region, we present a new technique that incorporates these distributions when extracting the mass function from the observed density of stars in the K--(H-K) diagram. We find that our data are inconsistent with a mass function that rises across the stellar/sub-stellar boundary. Instead, we find that the most likely form of the inner ONC mass function is one that rises to a peak around 0.15 M$_\\odot$, and then declines across the hydrogen-burning limit with slope N(log M) $\\propto$ M$^{0.57}$. We emphasize that our conclusions apply to the inner 0.71 pc x 0.71 pc of the ONC only; they may not apply to the ONC as a whole where some evidence for general mass segregation has been found. ", "introduction": "The Orion Nebula is one of the most famous objects in the sky, and has been the target of innumerable astronomical observations at virtually all wavelengths over the past 100 years. Yet it is only within the past few years that we have begun to discover the extent of the young stellar population just emerging from the ambient molecular cloud, and to characterize its nature. Work by Herbig \\& Terndrup (1986), Prosser et al. (1994) and Hillenbrand (1997) has established that the mean age of stars projected within $\\sim$2 pc of the massive Trapezium stars is $<$1 Myr. The mass distribution derived for $\\sim$1000 ONC stars located on a theoretical HR diagram by Hillenbrand (1997) rises to $\\sim0.2~M_\\odot$ and shows some evidence for flattening or turning over towards lower masses (see, however, the reinterpretation of these data in Figure~\\ref{fig:optical.newvsold} of the current paper using updated tracks/isochrones and updated transformations from observational to theoretical quantities). Our former study was complete to just above the hydrogren burning limit and did not constrain the mass function across the stellar/sub-stellar boundary into the brown dwarf regime. Existence of brown dwarfs in the ONC has been discussed previously by McCaughrean et al. (1995). Star forming regions like the ONC provide one of the best environments for investigating the shape of the stellar mass function into the brown dwarf regime. Unlike the case in older clusters and associations, star-forming regions are essentially unperturbed by dynamical evolution that selectively remove the lowest mass objects. Further, contracting low-mass pre-main sequence stars and brown dwarfs are 2-3.5 orders of magnitude more luminous that their counterparts on the main sequence and hence can be readily detected, especially in the near-infrared. Star forming regions are also less affected by field star contamination compared to older clusters due to their small angular extent and their association with obscuring molecular material. The ONC cluster in particular affords several distinct advantages compared to any other young stellar cluster for measuring the initial mass function. First, since it is located at high galactic latitude toward the outer Galaxy, contamination from field stars is minimized. Further, the winds and ionization from the central OB stars have dispersed much of the surrounding gas and dust, drastically reducing the extinction to the cluster members. A high-column density of obscuring molecular material does remain intact behind the stellar cluster. Foremost, however, as the nearest massive star-forming region to the Sun and the most populous young cluster within at least 2 kpc, the ONC is the one region where one can assemble a statistically robust assessment of the mass distribution well into the brown dwarf regime. In this contribution we investigate whether the distribution of stars in the \\khk color-magnitude diagram for the ONC is consistent with a mass function that rises across the stellar/sub-stellar boundary and into the brown dwarf regime, or if the data demand that the mass function turns over. After describing the observations, image analysis, construction of the point source list, and extraction of photometry, we present a new approach for constraining the stellar/sub-stellar mass function. We consider that the location of a particular star in the \\khk diagram depends on four parameters: stellar mass, stellar age, presence and properties of a circumstellar disk, and extinction. De-reddening the stars along a known reddening vector in the K-(H-K) diagram enables us to compute the probability that a star could be of a certain mass given the distributions in age and near-infrared excess that characteristize the ONC cluster. Summing of these individual mass probability distributions yields the mass function for the entire cluster. We believe that our technique produces the most rigorously derived constraint yet from photometry alone on the inner ONC initial mass function. ", "conclusions": "We have introduced a new method for constraining the stellar/sub-stellar mass distribution for optically invisible stars in a star-forming region. A comparative review of the various techniques already in use for measuring mass functions in star-forming regions is presented by Meyer et al. (2000). These techniques range from studies of observed K-magnitude histograms (e.g. Muench et al. 2000), to discrete de-reddening of infrared color-magnitude diagrams (e.g. Comeron, Rieke, \\& Rieke 1996), to the assembly of photometric and spectroscopic data from which HR diagrams are created (e.g. Luhman \\& Rieke 1998). Our method is a variation on and an improvement to the discrete de-reddening of color-magnitude diagrams since we fully account for distributions in the relevant parameters instead of assuming a mean value for them. However, our method is not as good as a complete photometric - plus - spectroscopic survey since we produce only a mass probability distribution for each star, not a uniquely determined mass. Nonetheless, we believe that the statistical nature of our method does provide the most rigorously established constraint to date from photometry alone on the stellar mass function in a star-forming region. We have used information from previous studies of optically visible stars in the ONC to derive plausible functional forms for the stellar age and the circumstellar near-infrared excess distributions in the innermost regions studied here. We assume that these distributions apply equally well to the optically invisible population. We find a mass function for the inner 0.71 pc x 0.71 pc of the ONC which rises to a peak around 0.15 M$_\\odot$ and then declines across the stellar/sub-stellar boundary as N(log M) $\\propto$ M$^\\Gamma$ with slope $\\Gamma=0.57$. This measurement is of the primary star/sub-star mass function only, and should be adjusted by the (currently unknown) companion mass function in order to derive the ``single star mass function,'' if desired. We find strong evidence that the shape of the mass function for this inner ONC region is different from that characterizing the ONC as a whole, in the sense that the flattening and turning over of the mass function occurs at higher mass in the inner region than in the overall ONC. In fact, the shape of mass function for the overall ONC is currently unconstrained across the stellar/sub-stellar boundary, and appears now based on the most recent theoretical tracks and conversions between the theory and observables used in this paper, to continue to rise to at least 0.12 M$_\\odot$." }, "0003/astro-ph0003400_arXiv.txt": { "abstract": "{Intergalactic medium, galaxy formation, dark matter} The last few years have seen a dramatic improvement in our understanding of the origin of Lyman $\\alpha$ absorption systems. Hydrodynamic numerical simulations of cold dark matter dominated universes have shown that the many properties of the Lyman $\\alpha$ absorption systems can be explained by a photoionized, space-filling, intergalactic medium. Lyman $\\alpha$ lines offer promising probes of the photoionizing background, the amplitude of the mass fluctuations at high redshift and the evolution of the equation of state of the intergalactic medium. ", "introduction": "The existence of a forest of absorption lines blueward of the Ly$\\alpha$ emission line in quasar spectra has been known for over 30 years (Bahcall and Salpeter 1965; Lynds 1971). These lines arise from Ly$\\alpha$ absorption by neutral hydrogen from intervening structure along the line-of-sight. Early theoretical models interpreted this structure as absorption caused by discrete gas clouds in the intergalactic medium (IGM), either pressure confined by a hot IGM (Sargent \\etal\\ 1980; Ostriker and Ikeuchi 1983) or confined by the gravity of dark matter `mini-halos' ({\\it e.g.} Rees 1986). Over the last few years our understanding of the Ly$\\alpha$ forest has undergone a transformation for at least two reasons. Firstly, observations with the Keck telescope have produced almost noise-free spectra of quasars at high spectral resolution over the redshift range $2 \\simlt z \\simlt 4$. The exquisite quality of Keck spectra has allowed observers to resolve Ly$\\alpha$ absorption lines at low column densities ($ \\sim 10^{12.5}\\; {\\rm cm}^{-2}$) and to study their evolution. Secondly, hydrodynamic numerical simulations of structure formation in cold dark matter (CDM) universes with high spatial resolution are now possible and have proved remarkably successful in reproducing many observed properties of the Ly$\\alpha$ forest (Cen \\etal\\ 1994; Zhang , Anninos and Norman 1995, 1997; Miralda-Escud\\'e \\etal\\ 1996; Hernquist \\etal\\ 1996, Theuns \\etal\\ 1998a). These simulations have shown that most of the Ly$\\alpha$ lines at column densities $\\simlt 10^{14.5} {\\rm cm}^{-2}$ arise from modest fluctuations in the baryon density in a space filling photoionized IGM, rather than from distinct clouds. The properties of the Ly$\\alpha$ lines can therefore be used to probe the structure and thermal history of the diffuse IGM and of the background UV radiation that determines its ionization state. The key characteristics of the numerical simulations are described in the next Section. Section 3 summarizes a number of results from these simulations and describes how the Ly$\\alpha$ lines can be used to study the IGM. In this paper we discuss only the properties of the low column density Ly$\\alpha$ lines. For a discussion of damped Ly$\\alpha$ systems and metal lines see Pettini's contribution to these proceedings. For a recent review of observations and theoretical models of Ly$\\alpha$ absorption lines see Rauch (1998). ", "conclusions": "The work reviewed in this article provides a powerful case that the Ly$\\alpha$ forest arises from a space-filling, highly photoionized diffuse IGM that contains most of the baryonic material in the Universe at high redshift. This model is a natural outcome of CDM theories of structure formation and can account for many observed properties of the Ly$\\alpha$ forest in quantitative detail. The general features of the model thus seem to us to be reasonably secure. However, a more detailed analysis of the thermal history of the IGM requires simulations that incorporate radiative transfer and a model for the spatial distribution of ionizing sources. Such calculations are now being done (Abel, Norman and Madau 1999; Gnedin 2000; Madau these proceedings) but the computational problems are formidable. Some outstanding problems that deserve further attention include: \\noindent $\\bullet$ detailed simulations of the inhomogeneous reionization of hydrogen and helium; \\noindent $\\bullet$ extending the analysis of Ly$\\alpha$ line widths to redshifts $\\simgt 4$, perhaps leading to constraints on the epoch of reionization of hydrogen; \\noindent $\\bullet$ analysis of inhomogeneities in the temperature of the IGM. Are there, for example, regions in the spectra of quasars in which Ly$\\alpha$ line-widths are systematically broader or narrower than in other regions? \\noindent $\\bullet$ further observations of absorption gaps in HeII Ly$\\alpha$ absorption (reported by Heap \\etal\\ (2000) and others) and the development of a model to understand their sizes; \\noindent $\\bullet$ searching for signatures of outflows around protogalaxies in the Ly$\\alpha$ forest; \\noindent $\\bullet$ determining the mean metallicity of the IGM and understanding how the metals were transported from protogalaxies." }, "0003/nucl-th0003008_arXiv.txt": { "abstract": "We study the equation of state of kaon-condensed matter including the effects of temperature and trapped neutrinos. Several different field-theoretical models for the nucleon-nucleon and kaon-nucleon interactions are considered. It is found that the order of the phase transition to a kaon-condensed phase, and whether or not Gibbs' rules for phase equilibrium can be satisfied in the case of a first order transition, depend sensitively on the choice of the kaon-nucleon interaction. To avoid the anomalous high-density behavior of previous models for the kaon-nucleon interaction, a new functional form is developed. For all interactions considered, a first order phase transition is possible only for magnitudes of the kaon-nucleus optical potential $\\gord100$ MeV. The main effect of finite temperature, for any value of the lepton fraction, is to mute the effects of a first order transition, so that the thermodynamics becomes similar to that of a second order transition. Above a critical temperature, found to be at least 30--60 MeV depending upon the interaction, the first order transition disappears. The phase boundaries in baryon density versus lepton number and baryon density versus temperature planes are delineated, which are useful in understanding the outcomes of protoneutron star simulations. We find that the thermal effects on the maximum gravitational mass of neutron stars are as important as the effects of trapped neutrinos, in contrast to previously studied cases in which the matter contained only nucleons or in which hyperons and/or quark matter were considered. Kaon-condensed equations of state permit the existence of metastable neutron stars, because the maximum mass of an initially hot, lepton-rich protoneutron star is greater than that of a cold, deleptonized neutron star. The large thermal effects imply that a metastable protoneutron star's collapse to a black hole could occur much later than in previously studied cases that allow metastable configurations. ", "introduction": "It is believed that a neutron star begins its life as a proto-neutron star (PNS) in the aftermath of a supernova explosion. The evolution of the PNS depends upon the star's mass, composition, and equation of state (EOS), as well as the opacity of neutrinos in dense matter. Previous studies \\cite{bigus,tpl,kj} have shown that the PNS may become unstable as it emits neutrinos and deleptonizes, so that it collapses into a black hole. The instability occurs if the maximum mass that the equation of state (EOS) of lepton-rich, hot matter can support is greater than that of cold, deleptonized matter, and if the PNS mass lies in between these two values. The condition for metastability is satisfied if ``exotic'' matter, manifested in the form of a Bose condensate (of negatively charged pions or kaons) or negatively charged particles with strangeness content (hyperons or quarks), appears during the evolution of the PNS. Even if collapse to a black hole does not occur, the appearance of exotic matter might lead to a distinguishable feature in the PNS's neutrino signature ({\\it i.e.}, its neutrino light curve and neutrino energy spectrum) that is observable from current and planned terrestrial detectors. This was investigated recently by Pons {\\it et al.} \\cite{pons} who studied the evolution of a PNS in the case where hyperons appeared in the star during the latter stages of deleptonization. Although the possibility of black hole formation was first discovered in the context of kaon condensation in neutron star matter \\cite{tpl}, a full dynamical calculation of a PNS evolution with consistent EOS and neutrino opacities in kaon condensed matter has not been performed so far. One of the objectives of this paper is to investigate $K^-$ condensation in finite temperature matter, including the situation of trapped neutrinos in more detail. An impetus for this study is the recent suggestion that a mixed phase of kaon-condensed and normal matter might exist which could greatly affect the structure \\cite{gs} and its neutrino opacity \\cite{rbp}. Another objective of our study is to identify differences in thermodynamic quantities such as the pressure, entropy or specific heat that might produce discriminating features in the star's neutrino emission. In separate works, we will examine neutrino interactions in kaon-condensed matter and neutrino signals from PNS evolution calculations in a consistent fashion. Since we wish to isolate the aforementioned effects due to kaons in this paper, we deliberately exclude consideration of hyperons. Hyperons and kaons were considered together in Refs. \\cite{kpe} and \\cite{schaffner}. Hyperons tend to delay the appearence of kaons in matter, especially if the $\\Sigma^-$ appears first. However, the $\\Sigma^-$ couplings are not as well determined as those of the $\\Lambda$ and even in this case the data are restricted to nuclear or subnuclear densities. Relatively small variations in the coupling constants can lead to a situation where the threshold density for the appearance of $\\Sigma^-$ particles is larger than that for kaons. These uncertainties remain unresolved; further hyper-nuclear experiments are needed to pin down their couplings. The original investigations of kaon condensation in neutron star matter ({\\it e.g.} Refs. \\cite{kapnel,pw,bkrt} and its astrophysical conseqences \\cite{tpl,mfmt}) employed a chiral $SU(3)_L\\times SU(3)_R$ model in which the kaon-nucleon interaction occurs directly via four point vertices. However, one can also employ an indirect, finite-range interaction which arises from the exchange of mesons. Several studies have been performed along these lines \\cite{gs,kpe,schaffner,mti,ty}. Ref. \\cite{kpe} found that the chiral and meson exchange approaches give similar results provided that the kaon-nucleon couplings are chosen to yield similar optical potentials in nuclear matter. Allowing kaons to interact via the exchange of mesons has the advantage that it is more consistent with the Walecka-type effective field-theoretical models usually used to describe nuclear matter \\cite{sew}. In most studies of kaon condensation it has been found that the transition to a phase in which kaons condense is second order for modest values of the kaon optical potential, $U_K$, of order -100 MeV. For magnitudes of $U_K$ well in excess of 100 MeV, however, the phase transition becomes first order in character. Even when the transition is first order, it is not always possible to satisfy Gibbs' criteria for thermal, chemical and mechanical equilibrium, so a Maxwell construction, which satisfies only thermal and mechanical equilibrium, was sometimes employed to construct the pressure-density relation. Recently, Glendenning and Schaffner-Bielich (GS) \\cite{gs} modified the meson exchange Lagrangian in such a way that the Gibbs criteria for thermal, chemical and mechanical equilibrium in a first order phase transition was possible. The extended mixed phase of kaon-condensed and normal matter which results produces a qualitative difference for the structure of a neutron star, since the EOS is softened over a wider region than in the case in which there is no mixed phase. This has implications for the mass-radius relation and the maximum mass, among other properties of the star. In this paper, we investigate the phase transition involving kaon-condensed matter and its influence upon the equation of state. We find that the precise form assumed for the scalar interactions (particularly, their density dependence), both for baryon-baryon and kaon-baryon interactions, determines whether or not the transition is first or second order, and, in the case of a first order phase transition, establishes whether or not a Gibbs construction is possible. Since the form of the scalar interactions is not experimentally well constrained at present, we have explored several different models in this study of the effects of kaon condensation on the EOS and the structure of a PNS. For each model, we have performed a detailed study of the thermal properties which are summarized in terms of phase diagrams in the density-lepton content and density-temperature planes. In Sec. II we present the various Lagrangians and derive exressions for the thermodynamic properties of each. We also develop the theoretical formalism necessary to describe baryons and kaon condensed matter in both the pure and mixed phases. This is followed by a discussion of the determination of the various coupling constants. Section III contains a comparison of the results for the EOS and for the structure of neutron stars for typical values of entropy and lepton content in a proto-neutron star as it evolves. Our conclusions and outlook for evolution of a proto-neutron simulation are presented in Sec. IV. In Appendix A, the extent of the correspondence between a meson exchange formalism and a chiral model to describe kaon condensation in matter is examined. The role of higher order kaon self-interactions in determining the order of the phase transition to a kaon condensed state is studied in Appendix B. ", "conclusions": "\\subsection{Zero temperature case} The effects of kaon condensation on the EOS are more pronounced at zero temperature than at finite temperature, since the fraction of thermally excited kaons increases with temperature relative to the fraction of kaons residing in the condensate. We therefore begin by examining results for the zero temperature case. We have considered two different nucleon Lagrangians, GM and ZM, and three different kaon Lagrangians, KPE, GS and TW. Below densities of about $0.5n_0$, matter is composed of neutron-rich nuclei immersed in a neutron sea. For this regime, we use the potential model results of Negele and Vautherin \\cite{NV} in the range $0.001 < n < 0.08~{\\rm fm}^{-3}$ and those of Baym, Bethe, and Pethick \\cite{BBP} for $n < 0.001~{\\rm fm}^{-3}$. For cold stars, the EOS in this regime has little effect on maximum masses or stellar radii. Furthermore, since the entropy in the stellar mantle $(n0$, the quantity $(\\mu+X_0)$, and hence $m_K^*$, has to be positive definite. In the ZM+TW model the kaon effective mass is relatively large so that $X_0$ is positive and therefore $\\theta$ is relatively small. On the other hand in the GM+GS model $m_K^*$ is quite small so that $X_0$ is negative and $\\theta$ has to be large. The negative value of $X_0=g_{\\omega K}\\omega_0+g_{\\rho K}b_0$ implies a large negative value of \\begin{equation} g_{\\rho K}b_0=\\frac{g_{\\rho K}g_\\rho}{2m_\\rho^2}\\left(n_p-n_n- \\frac{2g_{\\rho K}}{g_\\rho}n_K\\right)\\;, \\end{equation} which is clearly sensitive to the value of $g_{\\rho K}$. In fact, if this coupling is reduced by more than about 15\\% from our chosen value it is no longer possible to satisfy the Gibbs criteria. By comparing the pure phase results in Figs. \\ref{fig2} and \\ref{fig3} with the mixed phase results of Figs. \\ref{fig4} and \\ref{fig5}, it is clear that substantial modifications of the various fields are required to satisfy Gibbs' criteria. We examine now the KPE model for which it was not possible to satifsfy the Gibbs' criteria. In this case, Eq.~(\\ref{alKPE}) and the last of Eqs.~(\\ref{hhyp5}) leads to the condensate condition \\begin{equation} \\mu(\\mu+2X_0)={(m_K^*)}^2\\,, \\label{culprit} \\end{equation} whereas the functional form of the number density of kaons is identical to that in Eq.~(\\ref{kden}). Eq.~(\\ref{culprit}) differs in important ways from Eq.~(\\ref{lucky}). For the KPE model, even if $\\mu+2X_0$ is positive, $\\mu$ has the proclivity to turn negative for large $\\mu_n$ (or equivalently, for large baryon densities), leading to $(m_K^*)^2 < 0$ or imaginary values of the kaon effective mass $m_K^*$. This may be seen in Fig. \\ref{fig5a} where we show the electron chemical potential $\\mu$ as a function of the (negative) charge density in pure phase II for a typical value of the neutron chemical potential $\\mu_n = 1250$ MeV. It is now possible to understand qualitatively why a mixed phase cannot occur in the case of the kaon Lagrangian KPE. In comparison with the GM+GS and ZM+TW models, a distinctive feature of the KPE model is that $\\mu$ decreases rapidly with the (negative) charge density. In constructing a mixed phase, we are attempting to balance the positive charge in phase I with the negative charge in the dense phase II in which the electron chemical potential, and hence the charge content in leptons, is rapidly decreasing towards zero. The balance never occurs, hence the failure to meet the Gibbs' criteria. In terms of compositions, the GS or TW Lagrangians introduce negative charges in matter by increasing the number density of kaons, while keeping the electron density nearly constant or even slightly increasing with the charge density. The KPE Lagrangian, however, rapidly substitutes electrons by kaons, which is detrimental to meeting the Gibbs' criteria. For these reasons, we will concentrate on results with the other two kaon Lagrangians in the remainder of this paper. The influence of the condensate on neutron star structure (at zero temperature) is shown in Fig. \\ref{fig6} in which the gravitational mass is displayed as a function of the star's central baryon number density (left panel) and its radius (right panel). For the models shown, the transition is first order and Gibbs equations for mechanical and chemical equilibrium are utilized. For all cases shown the central densities of the maximum mass stars lie in the mixed phase. The effects of the condensate are more evident in the case of the GM+GS model in which the mixed phase occurs over a wider region of density than in the ZM+TW model. When the effects of the softening induced by the occurrence of the condensate are large, the limiting mass and the radius at the limiting mass are reduced significantly from their values when the condensate is absent. Note, however, that the softening effects are limited by the constraint that the maximum mass must exceed that of the binary pulsar PSR 1913+16, 1.442 M$_\\odot$. In the case of GM+GS, this constraint limits $|U_K^{\\rm {lin}}|$ to be smaller than about 125 MeV. In such a case, the minimum radius achieved is not as small as in the case $U_K^{\\rm {lin}}=-140$ MeV, as shown in Fig. \\ref{fig6}. The radii of stars with masses less than 1.2 M$_\\odot$ are not affected by the choice of the kaon Lagrangian or the kaon optical potential, since the condensation threshold is not reached in these cases. \\subsection{Comparison with other works} The density dependence of $m_K^*/m_K$, $-U_K$ and $\\omega_K$ have been investigated in other works \\cite{ww,ro,bgn,ppt,chp}, but for the most part either in isospin symmetric nuclear matter or pure neutron matter. In general, our results for $m_K^*/m_K$ with $-U_K^{\\rm lin}=80$ MeV are consistent with those of Refs.~\\cite{ro,bgn} (for an appropriate comparison, our results are to be compared with results obtained without in-medium pion contributions in Ref.~\\cite{ro}) and those of Ref.~\\cite{ww} for nuclear matter at both $n/n_0=1$ and 3. There is a relatively small change produced in going from nuclear matter to beta-equilibrated neutron star matter to pure neutron matter for the quantities $m_K^*/m_K$ and $-U_K$. Note that a direct comparison of the real parts of the optical potentials between different calculations must also account for the fact that in obtaining fits to data, the imaginary parts are often found to be as large as the real parts, which indicates fragmentation of strength in the quasi-particle spectral function. Relatively larger variations are found in the kaon energies in matter with varying amounts of isospin as can be seen from Fig.~\\ref{newfig}. In this figure, the top panel provides a comparison of results for beta-equilibrated neutron-star matter for the GM+KPE, GM+GS, and ZM+TW models, respectively, for values of $-U_K^{\\rm lin}$ at the extreme ends considered here, namely, 80 and 120 MeV. The bottom panel shows results for the ZM+TW model for $-U_K^{\\rm lin}=80$ MeV, for pure neutron matter, neutron-star matter, and isospin symmetric nuclear matter, respectively. At nuclear density where the models are calibrated, $\\omega$ decreases by about a few MeV in going from pure neutron matter to neutron star matter and by about a few tens of MeV in going from neutron star matter to nuclear matter. With increasing density, these differences become progressively larger. This trend is chiefly due to the behavior of the vector fields in matter with different amounts of isospin. At this time, our results for the density dependence of $\\omega$ can be compared with those of the potential models in Refs. \\cite{ppt,chp}. For values of $-U_K^{\\rm lin}$ near the lower end of the range we explored, in the neighborhood of 80 MeV, the behavior of $\\omega$, for example, is quite similar to the potential model results. As the authors in Refs. \\cite{ppt,chp} indicated, kaon condensation may be unlikely in this case. However, the relevant comparision must also include the electron chemical potential $\\mu_e$, since the density where $\\omega=\\mu_e$ determines the onset of kaon condensation. As demonstrated in Ref.~\\cite{lpph}, the behavior of $\\mu_e$ for neutron star matter at high densities is determined by the density dependence of the nuclear symmetry energy (see also a similar discussion in Ref.~\\cite{ppt}). Potential model calculations (see, for example Ref.~\\cite{apr}) tend to have a relatively weak density dependence of the symmetry energy, which generally results in an onset of kaon condensation that is at a rather large density. In field-theoretical and Dirac-Brueckner-Hartree-Fock \\cite{Eng} models, however, the symmetry energy varies relatively rapidly with density. These lead to smaller densities where kaon condensation occurs, for a given behavior of the kaon energy $\\omega$. Furthermore, the calculations of Ref. \\cite{ppt} have been performed only for pure neutron matter which further enhaces the values of $\\omega$ and discourages kaon condensation. In addition, as $|U_K^{\\rm lin}|$ is increased in magnitude in field-theoretical models, the role of kaons increases and $\\omega$ becomes progressively smaller as a function of density. Nevertheless, the lack of effective constraints at high density preclude choosing any model over another at this time. In summary, choosing values of $-U_K^{\\rm lin}$ near the lower end of the range we explored either lead to a second order phase transition or no transition at all in a neutron star, in which case the gross properties of the star are relatively unaffected from the case without kaons. On the other hand, values near the higher end of this range lead to a first order phase transition at a relatively low density, depending on the form of the interaction chosen, and a more pronounced effect on the star. Our aim has been to provide benchmark calculations in which both possibilities are entertained in order to consider their impact on thermodynamics and their astrophysical implications. \\subsection{Finite temperature case} We now investigate results at finite temperature and values of the lepton content characteristic of those likely to be encountered in the evolution of a PNS. We choose three representative sets of PNS conditions which correspond to: the initial conditions within a PNS (entropy/baryon $s=1$, trapped neutrinos with a lepton fraction $Y_L=0.35$), a time after several seconds when the interior is maximally heated ($s=2$, no trapped neutrinos so $Y_\\nu=0$), and a very late time when the PNS has cooled ($s=0, Y_\\nu=0$ -- identical to the zero temperature case discussed above). For a detailed explanation of the evolution of a cooling PNS see Pons {\\it et al.} \\cite{pons}. The contribution of the nucleons to the entropy per baryon $s_{\\cal N}\\equiv S_{\\cal N}/n_B$, with $n_B=n_n+n_p$ denoting the total nucleon density, in degenerate situations ($T/E_{F_i}\\ll1$) can be written \\begin{eqnarray} s_{\\cal N} = \\pi^2T ~ \\frac {\\sum_{i=n,p} k_{F,i} {\\sqrt {M^{*^2}+k_{F_i}^2} }} {\\sum_{i=n,p} k_{F,i}^3 } \\,, \\label{nucentropy} \\end{eqnarray} where $M^*$ and $k_{F,i}$ are the effective mass and the Fermi momentum of species $i$, respectively. For the temperatures of interest here, and particularly with increasing density, the above relation provides an accurate representation of the exact results for entropies per baryon even up to $s_{\\cal N}=s_n+s_p\\simeq2$. The behavior with density of both the Fermi momenta and the effective mass controls the temperatures for a fixed $s_{\\cal N}$. For kaons it is straightforward to show that the contribution to the entropy from $K^+$ mesons can be ignored since it is exponentially suppressed in comparison to the $K^-$ contribution. For the latter, keeping the leading temperature dependence of the simplest approximation scheme for bosons given in Ref. \\cite{jel}, the kaon entropy per baryon is \\begin{equation} s_K\\equiv\\frac{S_K}{n_B}=\\left[\\fivequar(2-y)-\\psi\\right] \\frac{n_K^{TH}}{n_B}\\quad{\\rm where} \\quad \\psi T=\\mu+X_0-\\sqrt{\\alpha+X_0^2}\\;, \\end{equation} and $y$ is determined from $\\psi$ by solving the equation \\begin{equation} \\psi=1-y+\\ln y\\;. \\end{equation} Below the kaon condensation threshold as the temperature becomes very small $\\psi\\rightarrow-\\infty$ so $y\\rightarrow0$. Above the kaon condensation threshold the last of Eqs. (\\ref{hhyp5}) implies that $\\psi=0$ in which case $y=1$. This simple approximation provides quite an accurate account of the kaon entropy per baryon which is fairly small for the scenarios examined here since it involves just the thermal contribution and the condensate plays no role. The total entropy per baryon $s_{\\rm tot}=s_{\\cal N}+s_K+(S_e+S_\\mu+S_\\nu)/n_B$ also includes the lepton contributions; $s_{\\rm tot}$ is dominated, however, by $s_{\\cal N}$. In Figs. \\ref{fig7} and \\ref{fig8}, the relative concentrations of various particles are displayed versus baryon number density for our three PNS conditions in the cases GM+GS and ZM+TW, respectively. The cases shown allow the Gibbs equations to be solved, and the boundaries of the mixed phase regions are indicated by vertical lines. The effect of finite temperature is to allow the existence of $\\mu^-$ and $K^-$ particles at all densities, although kaons become relatively abundant only within the mixed phase region. In the third set of diagrams, trapped neutrinos are present at all densities and the appearance and abundances of the negatively charged particles $\\mu^-$ and $K^-$ are suppressed. Furthermore, the critical density for kaon condensation is shifted to higher density. In Fig. \\ref{fig9} the pressure is displayed as a function of baryon number density for these two Lagrangians and the three PNS conditions. Two choices for the kaon optical potential are shown to highlight differences between cases in which kaons condense in second or first order phase transitions. The reduction of the pressure when kaons condense is obvious. For conditions in which the phase transition is first order, the result of applying the Gibbs conditions and the result of assuming pure phases (thin line) are both shown. The application of the Gibbs conditions leads to further softening of the pressure over a wider density range. In the case of model ZM+TW, a first order phase transition occurs only for very low temperatures and low neutrino concentrations. In Fig. \\ref{fig10} we show the matter temperature as a function of the baryon density for these two Lagrangians for the two PNS conditions with $s>0$ (the kaon optical potentials are as in the previous figure). The appearance of the kaon condensate generally leads to a reduction in specific heat which is indicated by the abrupt temperature increase which persists to high densities. In the case of first order transitions, applying the Gibbs conditions leads to a further enhancement of the temperature in the mixed phase regime. This behavior is in marked contrast to the case in which additional fermionic degrees of freedom, such as hyperons or quarks, are excited \\cite{pons} causing the temperature to drop and the specific heat of the matter to be increased. The latter follows from Eq. (\\ref{nucentropy}) where, in the absence of any variation of $M^*$, a system with more components at a given baryon density has a smaller temperature than a system with fewer components (recall that $\\sum_i Y_i =1$). In the present case the dropping of the effective mass is the dominant effect and this leads to larger temperatures. Figure \\ref{fig11} shows the phase diagram of kaon condensed matter, for the case GM+GS with $U_K^{\\rm {lin}}=-120$ MeV. The left panel displays results for zero temperature in the density--lepton concentration plane. The dashed lines show the minimum lepton concentration allowed at zero temperature (with $Y_\\nu=0$) for each density. Note that the minimum lepton concentration increases with density until the phase transition begins; above this density, the minimum lepton concentration decreases with increasing density. Also note that the phase transition to a kaon-condensed phase is pushed to higher densities when trapped neutrinos are present. This implies that in the initial PNS core material, in which $Y_L\\approx0.35-0.4$ and the central density is less than 3.5 times the nuclear saturation density, a kaon condensate phase likely does not exist. However, as neutrinos leak from the star the transition density decreases and a kaon condensate eventually forms. The right panel displays results in the density versus temperature plane, assuming no trapped neutrinos ($Y_\\nu=0$). The phase diagram for kaon condensed matter for the case ZM+TW with $U_K^{\\rm {lin}}=-140$ MeV is shown in Fig. \\ref{fig12}; the results are qualitatively similar to the GM+GS case in which $U_K^{\\rm{lin}}=-120$ MeV in Fig. \\ref{fig11}. This is understandable from the perspective that the actual optical potential for these two models are nearly the same. The boundary between phases I and the mixed-phase region are nearly the same for the two cases. The major difference is the much smaller extent of the mixed-phase region for the case ZM+TW. Note that for both cases the density at which the phase transition begins is relatively independent of temperature, so that the heating which initially occurs in the PNS has little effect on the eventual appearance of a kaon condensate. Also note that the density range of the mixed phase decreases with increasing temperature, and the mixed phase persists to high temperatures. It appears that the mixed phase exists up to temperatures exceeding 60 MeV, for the case GM+GS and $U_K^{\\rm {lin}}=-120$ MeV, or 30 MeV for the case ZM+TW with $U_K^{\\rm {lin}}=-140$ MeV. It becomes increasingly difficult to determine the properties of a mixed phase near the temperature at which it disappears. In Fig. \\ref{fig13} the gravitational mass is plotted as a function of central baryon number density for these models. Results are shown for our three PNS conditions which correspond to snapshots of the PNS evolution. The initial configuration (dotted curves) has the largest maximum mass. The progression to the dashed and solid curves indicates the evolution with time and we see that the maximum masses decrease. The effect of temperature upon the structure of the PNS is significant. Thermal kaons play a significant role here, since the net negative charge they contribute to the system partially inhibits the appearance of the condensate which allows hot neutrino--free stars to reach higher masses than cold stars. The net decrease in maximum mass during the evolution for either case is seen to be of order 0.2--0.3 M$_\\odot$. Thus there is an appreciable range of masses for the PNS which will result in metastability with the star ultimately collapsing to a black hole. The central density of the maximum mass, zero temperature star is smaller for the GM+GS case than for the ZM+TW case. This is in spite of the apparently ``softer\" GM+GS equation of state in which the kaon condensed mixed-phase region extends over a wider density range. Ultimately, the smaller maximum mass of the GM+GS EOS leads to a smaller central density at the maximum mass." }, "0003/astro-ph0003150_arXiv.txt": { "abstract": "I review the understanding of bulges that emerged from the lively discussions and presentations during the meeting, and emphasize areas for future work. The evidence is for a diversity of `bulges', and of formation mechanisms. ", "introduction": "Classical bulges are centrally-concentrated, high surface density, three-dimensional stellar systems. Their high density could arise either because significant gaseous dissipation occurred during their formation, or could simply reflect formation at very high redshift (or some combination of these two, depending on the density). For illustration, equating the mean mass density within the luminous parts of a galaxy (assumed to have circular velocity $v_c$ and radius $r_c$) with the cosmic mean mass density at a given redshift, $z_f$, gives (e.g. Peebles 1989) $$z_f \\sim 30 \\, {1 \\over f_c \\Omega^{1/3}}\\, ({v_c \\over 250 {\\rm km/s}})^{2/3}\\, ({10 {\\rm kpc} \\over r_c})^{2/3},$$ where $f_c$ is the collapse factor of the proto-galaxy, being at least the factor 2 of dissipationless collapse, and probably higher so that bulges, as observed, are self-gravitating, meaning that they have collapsed relative to their dark haloes. The majority view at the meeting, consistent with the observations, is that indeed proto-bulges radiated away binding energy, but also at least their stars formed at relatively high redshift. One must always be careful to distinguish between the epoch at which the stars now in a bulge formed, and the epoch of formation of the bulge system itself (as emphasized by Pfenniger, this volume). Of course if the bulge formed with significant dissipation, meaning gas physics dominated, then the star formation and bulge formation probably occured together. The small length-scale of bulges, combined with their modest rotation velocity, leads to a low value of their angular momentum per unit mass. Indeed, in the Milky Way Galaxy, the angular momentum distribution of the bulge is similar to that of the slowly-rotating stellar halo, and different from that of the disk, strongly suggestive of a bulge--halo connection, perhaps via gas ejection from halo star-forming regions (e.g. Wyse \\& Gilmore 1992). One can appeal to bulges forming from the low angular momentum regions of the proto-galaxy, a variant on the Eggen, Lynden-Bell \\& Sandage (1967) `monolithic collapse' scenario, explored further by van den Bosch (1998 and this volume). Or one can posit angular momentum transport prior to the formation of the bulge, taking angular momentum away from the central regions, and depositing it in the outer regions. Such transport of angular momentum could perhaps occur during hierarchical merging, by dynamical friction and gravitational torques, although one must be careful not to end up with too small a disk due to over-efficient angular momentum re-arrangement (e.g. Zurek, Quinn \\& Salmon 1988; Navarro \\& Benz 1991; Navarro \\& Steinmetz 1997). More modest amounts of angular momentum transport may be achieved by some viscosity in the early disk (e.g. Zhang \\& Wyse 1999). A recurring theme of the meeting was that large bulges (of early-type disk galaxies?) are related to ellipticals while small bulges (intermediate--late-type disk galaxies?) are more closely allied to disks. We need to be very clear about the observational selection criteria used in the definition of samples, and how this could bias our conclusions. As we will see below, the Milky Way bulge shows characteristics of {\\it both\\/} early- and late-type bulges, and will feature in {\\it both\\/} bulge--elliptical connections and bulge--disk connections. \\subsection{The elliptical--bulge connection} There has been remarkably little new kinematic data for representative samples of bulges (as opposed to detailed study of particular individual bulges, chosen for their unusual characteristics) since the pioneering work of the 1970s and 1980s. As demonstrated by Davies \\etal\\/ (1983), the bulges of early-type spirals are like ellipticals of equal luminosity in terms of rotational support, and are consistent with being isotropic oblate rotators i.e. with having an isotropic stellar velocity dispersion tensor, and being flattened by rotation about their minor axis. This sample was biased towards early-type spirals to facilitate bulge--disk decomposition, by observing edge-on systems with a prominent bulge. The bulge of the Milky Way Galaxy can be observed to match the techniques employed in the study of the bulges of external galaxies and, also then has stellar kinematics consistent with being an isotropic rotator (Ibata \\& Gilmore 1995a,b; Minniti 1996), as shown in Figure 1 here. \\begin{figure} \\hskip0.5in \\psfig{file=wyse_fig1.ps,width=4in,angle=270} \\caption{The level of rotational support as measured by (V/$\\sigma$)*, which has the value unity for an isotropic oblate rotator, against absolute magnitude for elliptical galaxies (plus symbols) and bulges of early-type spirals (open squares); data from Davies {\\it et al.} 1983. The bulge of the Milky Way, with kinematic quantities and flattening estimated in a similar manner as for the external galaxies, is indicated by the point with error bars.} \\end{figure} The trend apparent in Figure~1, and discussed more fully in Davies \\etal\\/ (1983), is that the level of rotational support in ellipticals increases as the luminosity of the elliptical decreases. The surface brightness of ellipticals also increases with decreasing luminosity, at least down to the luminosity of M32 (the dwarf spheroidal galaxies are another matter), as noted by Kormendy (1977), Wirth \\& Gallagher (1984) and many subsequent papers. These two relations are consistent with an increasing level of importance of dissipation in ellipticals with decreasing galaxy luminosity (Wyse \\& Jones 1984). Further, the bulges of S0-Sc disk galaxies follow the general trend of the Kormendy (1977) relations, in that smaller bulges are denser (de Jong 1996; Carlberg, this volume; see Figure~3 below for details). Thus one interpretation of Figure 1 is then that (some) bulges too formed with significant dissipation. As discussed by several speakers, the bulges of S0-Sc disk galaxies have approximately the same Mg2 -- velocity dispersion relation as do ellipticals (Jablonka \\etal\\/ 1996; Idiart \\etal\\/ 1997; see Renzini this volume), although the actual physics behind this correlation is not uniquely constrained. The properties of line-strength gradients in ellipticals of a range of luminosities are consistent with lower luminosity ellipticals forming with more dissipation than the more luminous ellipticals (Carollo, Danziger \\& Buson 1993). Again these results are suggestive that bulges are similar to low-luminosity ellipticals, and that gas dissipation was important. The detailed interpretation of the line-strength data in terms of the actual age and metallicity distributions of the stars is extremely complex and as yet no definitive statements can be made. There is a clear need for more data, including radial gradients, and for more models (see Trager, this volume). The broad-band colors of (some) bulges are consistent with those of the stellar populations in early-type galaxies in the Coma cluster (Peletier \\& Davies, this volume). We still need better models to interpret even broad-band colors. \\subsection{The disk--bulge connection} The surface-brightness profiles of bulges in later-type disk galaxies are better fit by an exponential law than by the steeper de Vaucouleurs profile, which in turn is a better fit for the bulges of early-type disk galaxies (Andredakis, Peletier \\& Balcells 1995; de Jong 1996). The sizes of bulges are statistically related to those of the disks in which they are embedded, and indeed the (exponential) scale-lengths of bulges are around one-tenth that of their disk; this correlation is better for late-type spirals than for early types (Courteau, de Jong \\& Broeils 1996). The projected starlight of the bulge of the Milky Way can be reasonably well-approximated by exponentials (vertically and in the plane); the Milky Way then fits within the scatter of the correlation of the external galaxies. The optical colors of bulges are approximately the same as those of the inner disk, for the range of Hubble types S0-Sd (Balcells \\& Peletier 1994; de Jong 1996), but as ever the decomposition of the light curves is difficult, as is correction for dust. This correlation implies similar stellar populations in bulges and their disks, as may be expected if bulges form from their disks (see Pfenniger, this volume). Thus, should there be a variation of mean stellar age from disk to disk, as may be expected from the range of colors observed, and indeed from observations of gas fraction etc., together with models of star formation in disks, one would expect a corresponding range in the mean stellar age of the different bulges. However, Peletier \\& Davies (this volume) find only a narrow range in bulge ages for their sample, based on optical--IR colors. More data are clearly needed. Figure~1 demonstrated the similarity in their kinematics between bulges and ellipticals of the same luminosity; Figure~2 (taken from Franx 1993) illustrates some of the complexity of bulge kinematics, and emphasizes the need to be aware of the selection criteria -- not all bulges are the same. The left-hand panel shows that in terms of the ratio of stellar velocity dispersion to true circular velocity (not the rotational streaming velocity), bulges scatter below ellipticals. Further, the right-hand panel shows that bulges of late-type disk galaxies have values of this ratio similar to that typical of inner disks (from Bottema 1993). The Milky Way bulge in this plot is quite typical ($\\sigma/{\\rm V_c} \\sim 0.5$, B/T$ \\sim 0.25$). \\begin{figure} \\vskip -2.75truein \\psfig{file=wyse_fig2.ps,width=5.75in} \\vskip -2.75 truein \\caption{{\\it a}) The central velocity dispersion of stellar tracers, $\\sigma$, against dark halo circular velocity, $v_c$. Open symbols represent bulges; closed symbols represent ellipticals. Circular velocities for the ellipticals are derived from models, as described by Franx (1993). ({\\it b}) The ratio of velocity dispersion in the bulge to dark halo circular velocity, $\\sigma/v_c$, taken from Franx (1993), plotted as a function of bulge-to-total luminosity (B/T) ratio, for the entire range of Hubble Type. The triangle at left is valid for the inner regions of pure disks, the square at right for ellipticals. Note that systems with low B/T have kinematics almost equal to those of inner disks.} \\end{figure} Complexity in the relationship between surface brightness and scale-length for bulges is illustrated in Figure~3, based on {\\tt WFPC2} data from Carollo (1999). The plot shows that while the large, $R^{1/4}$-law bulges follow the same scaling as ellipticals, the smaller, exponential-profile bulges are offset to lower surface brightnesses and occupy the extension to smaller scalelengths (by about a factor of ten, as noted above) of the locus of late-type disks. This strengthens the disk--bulge connection for these small bulges. However, Carollo (1999) finds both R$^{1/4}$ and exponential bulges in apparently very similar disks, so some additional parameter is important. \\begin{figure} \\vskip -1.5truein \\hskip -0.3truein \\psfig{file=wyse_fig3.ps,width=5.75in} \\vskip -1.25 truein \\caption{The mean $V$-band surface brightness $\\mu_e$ within the half-light radius $R_e$, as a function of log$R_e$ (in pc). The {\\tt WFPC2} measurements are shown with pentagons for the exponential bulges and large circles for the `classical' $R^{1/4}$-law bulges. Comparison data from the literature are shown for the $R^{1/4}$ bulges from Bender \\etal\\/ (1992; small circles) and Scd-Sm disks from Burstein \\etal\\/ (1997; crosses). The solid line is the best fit to the elliptical galaxy sequence (data from Bender \\etal\\/ 1992 and from Burstein \\etal\\/ 1987). The typical 1-$\\sigma$ error bar for the {\\tt WFPC2} masurements is shown in the upper-right corner.} \\end{figure} Association of `peanut' bulges with bars, which are essentially a disk phenomenon, was made in several contributions, using both gas and stellar kinematics (Kuijken; Bureau). However, the pronounced `peanut' in the early COBE images of the Milky Way was apparently largely an artefact of patchy dust, and the amplitude of such a morphology in the bulge of the Milky Way is not reliably established (Binney, Gerhard \\& Spergel 1997). As emphasized by Pfenniger (this volume), the kinematical and dynamical effects of bars are 3-dimensional; they can scatter stars by resonances, and/or themselves go unstable, fatten and dissolve, leading to a bulge. Which process dominates? There is a wealth of fascinating physics to explore. The modellers need to make more contact with observations, including predictions for direct comparison with the stellar kinematics, ages of stars, surface brightness profiles etc. M33 has neither a bulge nor a bar, but does have a central nucleus, and of course a substantial disk. Such systems need to be discussed in this context. The central nucleus of the Milky Way contains a black hole and star clusters of mass fraction well below the 1\\% or so estimated to destroy a bar (Norman, Sellwood \\& Hasan 1996), if we associate all the $10^{10}$M$_\\odot$ of the bulge with the bar. Indeed it is somewhat of a curiosity that the Milky Way does not fit the relationship between black hole mass and bulge mass found by Magorrian \\etal\\/ (1998). ", "conclusions": "Bulges are diverse in their properties, and probably in their formation mechanisms, or at least in the dominant physics at the epochs of star formation and/or assembly. Perhaps the differences are just a matter of degree, since, for example, even `monolithic collapse' involves fragmentation, with subsequent star formation in the fragments. A centrally-concentrated profile appears to match `maximum entropy' arguments (Tremaine, Henon \\& Lynden-Bell 1986) for the end-point of violent relaxation of a cold, clumpy system, independently of the details of the evolution to that end-point. The overall trends of the observations are that small bulges, of late-type disk galaxies, show a strong connection to their disk, while big bulges, of early-type disk galaxies, are more like the low-luminsity extension of the elliptical galaxy sequence. The bulge of the Milky Way appears to straddle these two generalities, having an affinity for its disk in terms of structure, but having the old, metal-rich population associated with `spheroids'. What does this mean? Even the casual reader should have noted the not-infrequent occurrence of the sentiment `more data and models are needed' in the text above. We are at the stage of requiring robust quantitative results from both theory and observations. More specifically, for the Milky Way, we require good HST color-magnitude diagrams for more lines-of-sight towards the Milky Way bulge, following the work of Feltzing \\& Gilmore (1999) in establishing the association of a younger stellar population with foreground disk. We also require good reddening maps and metallicity data to aid the interpretation of these color-magnitude diagrams. The inner disk of the Milky Way is remarkably under-studied, and again age and metallicity distributions -- and stellar kinematics -- are obviously crucial in determining the similarity or otherwise of inner disk and bulge. Further, we need to understand the relationship between the `bulge' globular clusters and the bulge field population; present models of globular-cluster formation appeal to pre-enrichment to provide the uniform enrichment within a given cluster, so it is not obvious that the enrichment signatures of cluster stars and field stars should be the same. Elemental abundances for statistically-significant samples of unbiased tracers of the field in a variety of lines-of-sight are required to understand the history of star formation. A combination of HST and ground-based (to probe both small- and large-scale structure) broad-band optical and IR colors, and surface brightness profiles, are still lacking for large samples, including the whole range of spiral Hubble types. These data should allow a robust quantification of the correlations between morphologies. Basic kinematic data, including gradients, should be obtained for a representative sample of bulges and disks. While we may lack the means at present for a unique interpretation of absorption line-strength data, the straightforward test for continuity in the line strengths from bulges to their disks is meaningful. The redshift of statistically-significant samples of galaxies is being continually pushed back (at what point will this pose a real problem for CDM?) and HST and the next generation of telescopes should provide robust morphological classifications. We will no doubt see evolution, but need to have the model predictions to be able to distinguish the underlying physics behind the evolution. `Secular-evolution' models are their early stages of development, but several key questions may be posed. While it may be reasonable to comment that a correlation between bulge scale-length and disk scale-length points to a connection between bulge and disk, can the models `post'-dict the factor of ten that is observed? Can they predict the frequency with which one should see barred spirals today, even ones with big bulges? Are all bars the same? Are there too many bars and/or central concentrations observed for the models of bar dissolution? Or is the dominant mechanism of bulge-building in this scenario actually scattering of disk stars through resonant coupling, rather than bar dissolution? How can this be compatible with uniformly old bulges? But are exponential bulges (apart from the Milky Way bulge) composed of old stars? Cold-dark-matter dominated cosmologies gained popularity partially because of their robust predictive power, a requirement for a good theory, in terms of the large-scale structure formed by the dissipationless dark haloes, (e.g. Davis, Efstathiou, Frenk \\& White 1992). The predictions for the luminous components, the galaxies as we observe them, have not yet achieved the same level of maturity. Advocates of merging and hierarchical clustering should quantify further the ages of stars now in bulges, and the epoch of assembly into bulges. What is predicted for the age spread within a typical bulge like the Milky Way? What fraction of bulges should have angular momentum vector misaligned with their disk? Should colors of bulge and disk be correlated? If bulges form in a `star-burst', what is the role of a supernova-driven wind? In this context, the X-ray properties of bulges, including the Milky Way, should constrain the ability of the bulge potential well to retain hot gas. Where do we stand? -- inspired to get to work!" }, "0003/astro-ph0003199_arXiv.txt": { "abstract": "In the generic CDM cosmogony, dark-matter halos emerge too lumpy and centrally concentrated to host observed galactic disks. Moreover, disks are predicted to be smaller than those observed. We argue that the resolution of these problems may lie with a combination of the effects of protogalactic disks which would have had a mass comparable to that of the inner dark halo and be plausibly non-axisymmetric, and of massive galactic winds, which at early times may have carried off as many baryons as a galaxy now contains. A host of observational phenomena, from quasar absorption lines and intracluster gas through the G-dwarf problem point to the existence of such winds. Dynamical interactions will homogenize and smooth the inner halo, and the observed disk will be the relic of a massive outflow. The inner halo expanded after absorbing energy and angular momentum from the ejected material. Observed disks formed at the very end of the galaxy formation process, after the halo had been reduced to a minor contributor to the central mass budget and strong radial streaming of the gas had died down. ", "introduction": "High resolution simulations of galaxy formation, incorporating realistic CDM initial conditions of dark halo formation, generally confirm the existence of a universal density (NFW) profile in the outer regions of galaxies \\cite {nfw}. Moreover, some groups are now reporting significant central dark matter density cusps that are as steep as $\\propto r^{-\\beta}$ with $\\beta\\approx 1.5$. The existence of even a more modest cusp ($\\beta\\approx 1$, as in the original NFW result) implies that at the current epoch $L_*$ galaxies have two to three times too much dark matter within a 2 to 2.5 disk scale lengths \\cite{nav}. This conclusion applies both to the Milky Way, where the mass of the disk can be dynamically estimated from the motions of stars near the Sun, and to an ensemble of nearby spirals, for which the Tully--Fisher relation effectively measures a $M/L$ ratio that can be compared with values predicted by stellar-synthesis models. The Tully-Fisher slope and dispersion are accounted for by the high resolution simulations, but the normalization is discrepant, by about a factor of 3 in $M/L$ at given surface brightness, rotation velocity and luminosity \\cite{nav}. Two further problems encountered with the cold dark matter hypothesis are (i) that the scale-lengths of disks are predicted to be too small by a factor $\\sim 5$ \\cite{ste}, and (ii) an order of magnitude more satellites are predicted than are observed \\cite{moo}. Both of these problems are closely related to the persistence of substructure in high-resolution N-body simulations of hierarchical models of dark halo formation. There are two possible avenues for resolution of these problems. One approach is to tinker with the particle physics. One may abandon the idea that CDM is weakly interacting. There are CDM particle candidates for which annihilation rates are of order the weak rate but for which scattering crossections are of the order the strong interaction \\cite{car,mah}. Such dissipative CDM may erase both the CDM cusps and clumpiness \\cite {spe}, but at the price of introducing an unacceptably spherical inner core in massive clusters \\cite{jor}. One may suppress the small-scale power on subgalactic scales, either by invoking broken scale-invariance \\cite {kam} or warm dark matter \\cite {som}, in the hope that the structure of massive dark halos will be modified. Here we adopt the less radical approach of exploring astrophysical alternatives. We accept the fundamental correctness of the CDM picture, and ask (i) could excess dark matter be ejected from the optical galaxy? and (ii) why do baryons in galaxies currently have more specific angular momentum than predicted by the simple CDM picture. We argue that these questions are connected, and that both may be resolved if galaxies have first absorbed and then ejected a mass of baryons that is comparable to their current baryonic masses. An earlier paper argued that baryonic winds can imprint cores within dwarf galaxy dark halos \\cite{nef}. Here we propose that energy and angular momentum surrendered by the ejected baryons have profoundly modified the dark halo within the current optical massive galaxy. In this picture most protogalactic material remained gaseous until the period of mass ejection was substantially complete -- this conjecture is tenable because we have no reliable knowledge of either the rate at which, or the efficiency with which, stars form in a protogalactic environment. In Section 2 we argue for massive galactic outflows. In Section 3 we ask how the dark halo was modified as a result of processing the material prior to ejection. Section 4 is concerned with the implications for the star-formation rate within a gaseous bar. Section 5 sums up our arguments. ", "conclusions": "Two serious problems currently plague the CDM theory of galaxy formation: an excess of dark matter within the optical bodies of galaxies, and disks that are too small. The second problem reflects the low angular momentum of infalling matter, and is made worse when one accepts that infalling baryons will surrender much of their angular momentum to the dark halo. In consequence, galaxies start with more low-angular-momentum baryons than they currently hold in their bulges and central black holes. We have argued that the surplus material was early on ejected as a massive wind. Many direct and indirect observational arguments point to the existence of such winds. Although the angular momentum of the first baryons to fall in was inadequate for the formation of the disk, it was not entirely negligible, and caused the inner halo to expand when the latter absorbed it. Similarly, the angular momentum of the baryons that are now in the disk was originally larger than it now is, and the surplus angular momentum further expanded the inner halo. In short, through relieving perhaps twice the baryonic mass of the current galaxy of angular momentum and energy, the dark halo has become substantially less centrally concentrated than it was originally, and it now contributes only a small fraction of the mass within the visible galaxy. During this refashioning of its inner parts, substructure is likely to have been erased, leaving the final inner halo smoother both locally and globally. This picture requires the baryonic mass to remain gaseous until the dark halo has been reduced to a minor contributor to the central mass, and a disk has formed in which most material is on circular orbits. This conjecture is plausible for two reasons: (i) the dark halo will be unresponsive to the collective modes of a gaseous disk, so the disk will not have growing modes until it dominates the gravitational potential in which it sits, and (ii) the enhanced orbital shear that is characteristic of closed orbits in a barred potential cannot be conducive to star formation. In any case we have little understanding of what controls the rate of star formation in a protogalaxy, and we know from the fragility of disks \\cite{toth} that disks formed at the end of the formation process, after merging had all but ceased and the largest substructure had been erased from bulge and inner halo. Existing numerical simulations of the interactions of baryons and dark matter during galaxy formation (e.g., Navarro \\& Steinmetz, 2000; Benson et al., 2000) lack both the mass resolution and some of the physics that is required to realise the essential ideas employed here. For example, in the simulations of Benson et al.\\ the gravitational softening length is $10h^{-1}\\kpc$, and the basic baryonic resolution element has mass $\\sim4\\times10^{10}\\msun$ and spurious discreteness effects will be present on mass scales several times larger. Such simulations neglect magnetic fields (which are believed to drive winds off accretion disks) and energy input by both supernovae and the central massive object. In summary, a considerable mass of low-angular momentum baryons must have been ejected. This prediction is a priori plausible, given observations of winds from star-burst galaxies and outflows from Lyman-break galaxies, and given the prevalence of outflows in star-formation regions. The heavy element abundances of hot gas in clusters of galaxies and in cool, low-density gas observed at redshifts $z\\sim2$ through quasar absorption lines are likely to arise through the mixing of metal-rich ejecta with primordial gas. The low-angular-momentum material having been ejected, the current disks formed from the higher-angular-momentum baryons that fell in later. Since the ejection stage commences only once $M_{\\rm baryon}\\sim M_{\\rm dm}$, in order for self-gravity to drive gas flows and the ensuing winds, thereby causing the dark-matter distribution to expand and the baryons to further contract, the visible galaxy is inevitably baryon-dominated but has a circular speed that is required, via the baryonic mass-loss and the protogalactic dynamical coupling, to approximately match that of the embedding halo. Thus the so-called `disk--halo conspiracy' \\cite {bah} is not really a coincidence but a consequence of dynamical evolution." }, "0003/astro-ph0003016_arXiv.txt": { "abstract": "Four quiescent elliptical galaxies were imaged with the NRAO VLA at 8.5~GHz. Within the context of canonical advection-dominated accretion flows (ADAFs), these VLA images plus published black hole masses constrain the accretion rates to be $ <1.6\\times10^{-4}$, $ <3.6\\times10^{-4}$, $\\le7.8\\times10^{-4}$, and $\\le7.4\\times10^{-4}$ of the Eddington rates. These ADAF accretion rates derived at 8.5~GHz have important implications for the levels of soft and hard X-rays expected from these quiescent galaxies. ", "introduction": "Evidence is accumulating that nearby galactic nuclei commonly harbour massive dark objects \\citep{mag98}. These objects are probably black holes, because such remnants from the QSO era should be common \\citep{fab95} and because some nuclear star clusters with the requisite mass and size would be improbably younger that their host galaxies \\citep{mao98}. Given a black hole mass, $M_{\\rm BH}$, a next important step is to constrain the rate, $\\dot{M}$, at which material is being accreted onto the black hole. Associated with any black hole is its Eddington accretion rate, $\\dot{M}_{\\rm E}$, which is the rate necessary to sustain the Eddington luminosity. For a 10\\% radiative efficiency and $M_{\\rm BH} = 10^8 m_8~M_{\\sun}$, $\\dot{M}_{\\rm E} = 2.2 m_8~M_{\\sun}~{\\rm yr}^{-1}$. For giant elliptical galaxies hosting black holes, accretion-rate estimates have long been available for comparison with Eddington rates \\citep{fab88,mah97}. These estimates are based on \\citet{bon52} accretion from an interstellar medium with temperature $T = 10^7 T_7$~K and pressure $P = 10^6 P_6$~cm$^{-3}$~K, with the accretion being characterized by a Bondi radius $r_{\\rm B} \\sim 4.3 m_8 T_7$~pc and a Bondi rate $\\dot{M}_{\\rm B} \\sim 1.9\\times10^{-4} m_8 P_6 T_7 ~M_{\\sun}~{\\rm yr}^{-1}$. Then a black hole with $m_8 = 1$ in a medium with $P_6 =1$ and $T_7 = 1$ will accrete at a Bondi rate $\\dot{M}_{\\rm B} \\sim 1.9\\times10^{-4}~M_{\\sun}$~yr$^{-1}$, which is almost four orders of magnitude less than the associated Eddington rate of $\\dot{M}_{\\rm E} = 2.2~M_{\\sun}$~yr$^{-1}$. The Bondi rate estimates are therefore extremely sub-Eddington. Still, such low rates could be pervasive among nearby ellipticals and, moreover, could define a minimum level of activity for ellipticals hosting massive black holes This Letter examines the consequences, in the radio and X-ray regimes, of such low accretion rates in nearby elliptical galaxies. The approach is to obtain deep radio continuum images of four ellipticals studied by \\citet{mag98} and then interpret those images within the context of canonical advection-dominated accretion flows (ADAFs), reviewed recently by \\citet{nar98}. The radio continuum from an ADAF is thermal synchrotron emission from a magnetized plasma, and the conversion between the ADAF radio power, $P_\\nu$, and the ADAF accretion rate, $\\dot{M}_{\\rm A}$, is relatively straightforward if the black hole mass is known \\citep{mah97,yi98}. Recent efforts along these lines have focused on galaxies with significant radio emission from nonthermal synchrotron jets \\citep{dim00}. In constrast, this study examines four quiescent elliptical galaxies previously undetected at radio wavelengths \\citep{wro91a,wro91b}, thereby minimizing one of the largest potential complications in an ADAF analysis - the role of jet outflows. Further evidence for the quiesence of these four galaxies comes from their use as absorption-line templates in optical spectroscopic studies \\citep{ho97}, plus their weak or undetected X-ray emission \\citep{beu99}. For these four quiescent ellipticals, the new radio imaging constrains the ADAF accretion rates to be $ <1.6\\times10^{-4}$, $ <3.6\\times10^{-4}$, $\\le7.8\\times10^{-4}$, and $\\le7.4\\times10^{-4}$ of the Eddington rates. These ADAF accretion rates derived at 8.5~GHz have important implications for the levels of X-ray emission expected from these quiescent galaxies. ", "conclusions": "" }, "0003/astro-ph0003220_arXiv.txt": { "abstract": "{We describe a code designed for hot media {(T $\\ge$ } a few 10$^4$ K), optically thick to Compton scattering. It computes the structure of a plane-parallel slab of gas in thermal and ionization equilibrium, illuminated on one or on both sides by a given spectrum. Contrary to the other photoionization codes, it solves the transfer of the continuum and of the lines in a two stream approximation, without using the local escape probability formalism to approximate the line transfer. We stress the importance of taking into account the returning flux even for small column densities (10$^{22}$ cm$^{-2}$), and we show that the escape probability approximation can lead to strong errors in the thermal and ionization structure, as well as in the emitted spectrum, for a Thomson thickness larger than a few tenths. The transfer code is coupled with a Monte Carlo code which allows to take into account Compton and inverse Compton diffusions, and to compute the spectrum emitted up to MeV energies, in any geometry. Comparisons with {\\sc{cloudy}} show that it gives similar results for small column densities. Several applications are mentioned.} ", "introduction": "In {\\sc{titan}} the physical state of the gas (temperature, ion abundances and level populations of all ionic species) is computed at each depth, assuming stationary state, i.e. local balance between ionizations and recombinations of ions, excitations and deexcitations, local energy balance ({equality} of heating due to absorption and cooling due to local emission), and finally total energy balance (equality between inward and outward fluxes). Due to the large range of density and temperature inside the medium, many physical processes play a role at some place and should therefore be taken into account. Ionization equilibrium equations include radiative ionizations by continuum and line photons, collisional ionizations and recombinations, radiative and dielectronic recombinations, charge transfer by H and He atoms, the Auger effects, and ionizations by high energy electrons arising from ionizations by X-ray photons. Energy balance equations include free-free, free-bound and line cooling, and Compton heating/cooling. The emission-absorption mechanisms for the continuum include free-free and free-bound processes, two-photon process, and Thomson scattering. Special care is given to recombinations to ground state, which are very important from an observational point of view in the X-ray range. They are treated differently according to the relative values of $kT$ to the photon energy bin, in order to get an accurate frequency dependence. Hydrogen and hydrogen-like ions are treated as 6-level atoms. Levels 2s and 2p are treated separately, while full l-mixing is assumed for higher levels. All processes including collisional and radiative ionizations and recombinations are taken into account for each level (cf. Mihalas 1978). Recombinations onto levels $n>5$ are not taken into account, which amounts to assuming that the higher levels are in LTE with the continuum, which is generally true in the conditions for which this code is presently used. In the future we plane to add several other levels, and to sum the contributions of the higher levels as it is done for instance in {\\sc{cloudy}}. Level populations are then obtained as usual by matrix inversion. In order to save computation time, non H-like ions are presently treated with a rough approximation: interlocking between excited levels is neglected and populations of the excited levels are computed separately using a two-level approximation. This approximation does not predict correctly the details of the line spectrum, since it neglects subordinate lines. Nevertheless recombinations onto excited states are taken into account in the ionization equilibrium and the transfer of these photons is treated in an approximative way as proposed by Canfield \\& Ricchiazzi (1980). We assume also that each recombination produces a resonant photon after cascades. The gas composition include 10 elements (H, He, C, N, O, Ne, Mg, Si, S, Fe), and all their ionic species are taken into account. Photoionization cross sections are fitted from Reilman \\& Manson (1979) and Band L.M. et al. (1990), these values being correct as far as neutral and once ionized ions are not concerned, and it is the case in hot media. For total radiative and dielectronic recombination rates, we use Aldrovandi \\& P\\'equignot's data (1973). When possible, collisional excitation rates are taken from the Daresbury Report (1985). Most of data for iron come from Arnaud \\& Raymond (1992), Kaastra \\& Mewe (1993), and from Fuhr et al. (1988). Ionizations by high energy electrons arising from ionizations by X-range photons are taken from Bergeron \\& Souffrin (1973). Inverse processes (except dielectronic recombinations) are computed through the equations of detailed balance. In the case of a gas close to LTE, we neglect dielectronic recombinations for consistency to insure the balance for each process. All induced processes are taken into account. The equations are not recalled as they have been given in previously quoted papers. \\medskip {\\bf Iron K lines} \\medskip These lines require special attention as they are intense in Seyfert nuclei, and they will be observed in detail in the future with Chandra and XMM. Though the iron K lines constitute a complex system described in Band D.L. et al. (1990), we assume presently only one ``mean\" line per ion with an oscillator strength equal to 0.4, as suggested by Band. More detailed computations are not required as far as the Doppler broadening of the lines (or Compton broadening as well, see Abrassart 2000) is much larger than the distance between the lines, as it is the case in AGN. In other computations of the line fluxes it is generally assumed that resonant trapping of K$\\alpha$ photons of FeXVII to Fe XXIII (Ross \\& Fabian 1993) suppress completely these lines when the Thomson thickness of the emitting medium is larger than a given value, of the order of 0.02 (Zycki \\& Czerny 1994). Here the transfer of these lines is handled in a standard way, with an additional term included in the statistical and ionization equations to take into account the competition between radiative deexcitation and the Auger process. The population $N_i^k$ of the upper level of a transition K$\\alpha$ of the ion $i$ is given by: \\begin{equation} N_i^kA_{ki}=y_{i-1}(N_{i-1}K_{i-1} + N_iB_{ik}J_{ik}) \\label{eq-kalfa1} \\end{equation} \\noindent while the ionization equation for the same ion $i$ writes: \\begin{eqnarray} \\label{eq-kalfa2} N_i[(P_i&+&K_i)+B_{ik}J_{ik}(1-y_{i-1})] \\\\ \\nonumber &=&N_{i+1}\\alpha_i\\ -\\ N_{i-1}K_{i-1}(1-y_{i-1}). \\end{eqnarray} $A_{ki}$ and $B_{ik}$ are the Einstein coefficients of the K$\\alpha$ line, $J_{ik}$ is the mean intensity integrated over the line profile, $y_i$ is the fluorescent yield, $\\alpha_i$ is the recombination coefficient, and $P_i$ (respt. $K_i$) is the photoionization rate of the ground state (respt. of the K-shell) of the ion $i$. The second term on the left side of Eq. \\ref{eq-kalfa2} is generally only of the order of $10\\%$ of the total photoionization rate. ", "conclusions": "We have shown that the coupling of {\\sc{titan}} and {\\sc{noar}} allows to compute the structure and the emission of hot Compton thick irradiated media in an unprecendented way. First it solves consistently both {\\it the global and the local} energy balance, which is impossible in a thick medium with codes handling the line transfer with the escape probability approximation, as all present photoionization codes do. We have also shown the importance of the returning flux (which is neglected in photoionization codes) even for relatively low column densities. Second it takes into account in an exact way inverse and direct Compton scattering, both in the energy balance and in the computation of the emitted spectrum. Finally, it allows to treat any geometry, open or closed. Although the problem of the convergence process is not as drastic as in stellar atmospheres, since we are able to get complete converged structure and spectrum in a still reasonable computing time, a most urgent improvement of the code is to accelerate the convergence process through the use of the Accelerated Lambda Iteration method. This will not only allow to get the results in a much smaller time, but also to get convergence for the few lines which are still not converged after about 10$^3$ iterations. Then the following improvements of {\\sc{titan}} will be to take into account subordinate lines in solving a multi-level atom for all ions and to bring up to date the atomic data. More detailed line spectra for some abundant ions will be obtained through coupling with the code {\\sc{iris}}. The L shells of Iron which are already taken into account for FeXVII to FeXXII will be implemented as well as a better representation of the Iron K lines. A few elements will be added to the already ten existing ones." }, "0003/astro-ph0003234_arXiv.txt": { "abstract": "Blue compact galaxies (BCG) are ideal objects in which to derive the primordial $^4$He abundance because they are chemically young and have not had a significant stellar He contribution. We discuss a self-consistent method which makes use of all the brightest He I emission lines in the optical range and solves consistently for the electron density of the He II zone. We pay particular attention to electron collision and radiative transfer as well as underlying stellar absorption effects which may make the He I emission lines deviate from their recombination values. Using a large homogeneous sample of 45 low-metallicity H II regions in BCGs, and extrapolating the Y-O/H and Y-N/H linear regressions to O/H = N/H = 0, we obtain Y$_p$ = 0.2443$\\pm$0.0015, in excellent agreement with the weighted mean value Y$_p$ = 0.2452$\\pm$0.0015 obtained from the detailed analysis of the two most metal-deficient BCGs known, I Zw 18 and SBS 0335--052. The derived slope dY/dZ = 2.4$\\pm$1.0 is in agreement with the value derived for the Milky Way and with simple chemical evolution models with homogeneous outflows. Adopting $Y_p$ = 0.2452$\\pm$0.0015 leads to a baryon-to-photon ratio of (4.7$^{+1.0}_{-0.8}$)$\\times$10$^{-10}$ and to a baryon mass fraction in the Universe $\\Omega_b$$h^2_{50}$ = 0.068$^{+0.015}_{-0.012}$, consistent with the value derived from the primordial D abundance of Burles \\& Tytler (1998). ", "introduction": "Blue compact galaxies (BCG) are low-luminosity (M$_B$ $\\ga$ -18) systems which are undergoing an intense burst of star formation in a very compact region (less than 1 kpc) which dominates the light of the galaxy (Figure 1) and which shows blue colors and a HII region-like emission-line optical spectrum (Figure 2). BCGs are ideal laboratories in which to measure the primordial $^4$Helium abundance because of several reasons: 1) With an oxygen abundance O/H ranging between 1/50 and 1/3 that of the Sun, BCGs are among the most metal-deficient gas-rich galaxies known. Their gas has not been processed through many generations of stars, and thus best approximates the pristine primordial gas. Izotov \\& Thuan (1999) have argued that BCGs with O/H less than $\\sim$ 1/20 that of the Sun may be genuine young galaxies. Their argument is based on the observed constancy and very small scatter of the C/O and N/O ratios in extremely metal-deficient BCGs with 12 + log O/H $\\la$ 7.6, which they interpret as implying that the C and N in these galaxies have been made in the same massive stars (M $\\ga$ 9 M$_\\odot$) which manufactured O, and that intermediate-mass stars (3 M$_\\odot$ $\\la$ M $\\la$ 9 M$_\\odot$) have not had time to release their nucleosynthetic products. Since the main-sequence lifetime of a 9 M$_\\odot$ star is $\\sim$ 40 Myr, Izotov \\& Thuan (1999) suggest that very metal-deficient BCGs are younger than $\\sim$ 100 Myr. Thus the primordial Helium mass fraction Y$_p$ can be derived accurately in very metal-deficient BCGs with only a small correction for Helium made in stars. \\begin{figure} \\vspace{3.5cm} \\vspace{2.cm} \\caption{Slit orientations superposed on {\\sl HST} archival $V$ images of I Zw 18 and SBS 0335--052. The slit orientation of I Zw 18 is chosen in such a way as to get spectra of the SE and NW components as well as of the C component to the NW of the main body of the galaxy. The spatial scale is 1\\arcsec\\ = 49 pc in the case of I Zw 18 and is 1\\arcsec\\ = 257 pc in the case of SBS 0335--052.} \\end{figure} 2) Because of the relative insensitivity of $^4$He production to the baryonic density of matter, Y$_p$ needs to be determined to a precision better than 5\\% to provide useful cosmological constraints. This precision can in principle be achieved by using BCGs because their optical spectra show several He I recombination emission lines and very high signal-to-noise ratio emission-line spectra with moderate spectral resolution of BCGs can be obtained at large telescospes (4 m class or larger) coupled with efficient and linear CCD detectors with a relatively modest investment of telescope time. The theory of nebular emission is well understood and the theoretical He I recombination coefficients calculated by Brocklehurst (1972) and Smits (1996) are well known enough to allow to convert He emission-line strengths into abundances with the desired accuracy. ", "conclusions": "" }, "0003/astro-ph0003002_arXiv.txt": { "abstract": "We cross-identified the {\\it ROSAT} Bright Source Catalog (RBSC) and the NRAO VLA Sky Survey (NVSS) to construct the RBSC-NVSS sample of the brightest X-ray sources ($\\geq 0.1$ counts s$^{-1} \\sim 10^{-12}$ ergs cm$^{-2}$ s$^{-1}$ in the 0.1--2.4 keV band) that are also radio sources ($S \\geq 2.5$ mJy at 1.4 GHz) in the 7.8 sr of extragalactic sky with $|b| > 15^\\circ$ and $\\delta > -40^\\circ$. The sky density of NVSS sources is low enough that they can be reliably identified with RBSC sources having rms positional uncertainties $\\geq 10''$. We used the more accurate radio positions to make reliable X-ray/radio/optical identifications down to the POSS plate limits. We obtained optical spectra for many of the bright identifications lacking published redshifts. The resulting X-ray/radio sample is unique in its size ($N \\sim 1500$ objects), composition (a mixture of nearly normal galaxies, Seyfert galaxies, quasars, and clusters), and low average redshift ($\\langle z\\rangle \\sim 0.1$). ", "introduction": "The {\\it ROSAT} All-Sky Survey Bright Source Catalogue \\citep[][RBSC revision 1RXS]{voges99} contains the first large all-sky sample of the brightest X-ray sources, analogous in many respects to the optical NGC catalog. It was derived from the soft (0.1--2.4 keV) X-ray survey performed during the first half year of the {\\it ROSAT} mission in 1990/91. The catalog sky coverage is 92\\%, and there are 8,547 sources above its 0.1 counts s$^{-1}$ ($\\sim 10^{-12}$ ergs s$^{-1}$ cm$^{-2}$) completeness limit. \\citet{bade98} found that about one third of the RBSC sources can be reliably identified with galactic stars, while most of the rest are extragalactic. The extragalactic content of the RBSC comprises a diverse blend of (1) normal spiral galaxies whose X-ray emission is powered by stars and stellar remnants, (2) elliptical galaxies with hot gaseous halos, (3) AGN in Seyfert galaxies, elliptical galaxies, quasars, and BL Lac objects, and (4) clusters of galaxies. The large number of sources in this catalog easily permits statistical analyses of {\\it each} type of X-ray object. However, the essential properties of these X-ray sources cannot be determined from the X-ray data alone---we need observations in optical and other wavebands to measure their distances, identify their energy sources, etc. Such observations are possible only for those RBSC sources whose optical counterparts have been identified. In this paper we present reliable radio and optical identifications for sources in the RBSC complete sample. Most RBSC sources have rms positional uncertainties $\\geq 10''$ and the sky density of faint optical objects is high, so only the nearest extragalactic X-ray sources can be optically identified by position coincidence alone \\citep[cf.][]{bade98}. Fortunately, most extragalactic RBSC sources are also radio sources in the 1.4 GHz NRAO/VLA Sky Survey \\citep[][NVSS]{condon98a}, whose sky density is low enough for identification with RBSC sources. Since the radio positions are significantly more accurate, the radio sources may be optically identified, yielding optical identifications for the corresponding X-ray sources as well. The NVSS covers the 10.3 sr of sky north of $\\delta = -40^\\circ$ and contains over $1.8 \\times 10^6$ sources stronger than its 2.5 mJy beam$^{-1}$ completeness limit. Since the NVSS was made with relatively low resolution ($45\\arcsec$ FWHM), it does not discriminate against moderately extended radio sources in nearby galaxies and clusters. Its rms positional uncertainties range from $< 1\\arcsec$ for the $N \\approx 4 \\times 10^5$ sources stronger than 15 mJy to 7$\\arcsec$ for the faintest (S = 2.3 mJy) detectable sources, allowing us to make optical identifications with objects as faint as R $\\approx 21$. We optically identified the RBSC-NVSS sources with objects in the United States Naval Observatory catalog A2.0 \\citep[][USNO]{monet98}. The USNO catalog contains 526,280,881 objects detected by the Precision Measuring Machine on the Palomar Optical Sky Survey I (POSS-I) blue O and red E plates, the UK Science Research Council SRC-J survey plates, and the European Southern Observatory ESO-R survey plates. The catalog was compiled from the blue/red overlaps (within 2$\\arcsec$) of the detection lists generated from scans of POSS-I O and E plates centered on $\\delta > -18^\\circ$ and SRC-J and ESO-R plates centered on $\\delta < -20^\\circ$. The stated astrometric and photometric errors are about $0\\farcs25$ and 0.5 mag rms, respectively. The USNO catalog covers the entire sky and probes as deep as B=21 (O plates), R=20 (E plates), J=22, and F=21 for objects with appropriate colors. Section~\\ref{cid} explains our method for making the identifications and assessing their reliabilities. The results are presented in Section~\\ref{results}. ", "conclusions": "" }, "0003/astro-ph0003372_arXiv.txt": { "abstract": "We report on CCD photometry in the Johnson $B$, $V$ and $I$ passbands for 146 stars in a $9^{\\prime} \\times 9^{\\prime}$ region around the southern aggregate NGC~6994 (C 2056-128), which appears in the Lyng{\\aa} (1987) catalogue of open star clusters. We argue that this object is not really an open cluster, but simply a random enhancement of four bright stars above the background level. This stars sample includes HD~358033 and GSC~05778--0082, together with M~73, which is referred to as a binary or multiple star, but actually represents the whole asterism. Since NGC~6994 is not the first case (see for instance Carraro \\& Patat 1995), this raises the possibility that other open clusters may have been misclassified. We also suggest that NGC~6994 is unlikely to be an open cluster remnant (OCR). ", "introduction": "The disk of the Milky Way harbors about 1200 open clusters. The youngest ones trace the spiral pattern in the disk, whereas the oldest ones are ideal templates to study the chemical and dynamical evolution of the disk, together with the Star Formation History (Friel 1995).\\\\ Open star clusters are usually studied by analyzing the Color Magnitude Diagram (CMD, Chiosi et al 1992).\\\\ Despite the importance of these objects for our understanding of the disk properties, CMDs are at present available only for a small fraction of the clusters population, say 40$\\%$. However the situation is rapidly changing thanks to dedicated surveys like for instance the PLCON (Palomar Las Campanas Observatory NOAO) Open Clusters Survey (Phelps 1999).\\\\ One of the crucial problem with open clusters is that they are highly contaminated by field stars in the disk, so that it is rather difficult to obtain precise estimates of their fundamental parameters.\\\\ This way many clusters remain completely unstudied but for the identification, which in most cases is done on a by eye basis, by inspecting at several different sky charts.\\\\ In some cases the identification is made by recognizing a peak of star concentration in the field. This is naturally a good criterion, which however deserves further studies to confirm the real nature of the star enhancement.\\\\ This is not only a semantic question. A star cluster is doubtless a star enhancement in the field. \\\\ Nonetheless the real nature can be unraveled with higher degree of confidence by looking at the CMD, and recognizing typical features, like a Main Sequence of H-burning stars, a Red Giant Branch, and/or a clump of He-burning stars (Chiosi et al 1992). \\\\ For instance Carraro \\& Patat (1995) analyzed photometry for the presumed old cluster Ruprecht~46, and concluded that it is not a cluster, but only a random enhancement of bright stars.\\\\ \\noindent Recently, de la Fuente Marcos (1998) has performed numerical simulations of open clusters evolution, suggesting that many star concentrations in the sky may be OCRs. The basic criterion is the evidence of a star concentration higher than the field stars. So in principle an enhancement of stars could be just the final stage of the evolution of an open cluster. \\noindent \\begin{figure*} \\centerline{\\psfig{file=9522_fig1.ps,height=14cm,width=14cm}} \\caption{A DSS $10 \\times 10$ arcmin$^{2}$ image in the region of the open cluster NGC~6994. North is on the top, east on the left. The two brighter stars in the asterism are GSC~05778--0082 and HD~358033, respectively.} \\end{figure*} In this paper we present $B$, $V$ and $I$ photometry for NGC~6994, an object which appears in the Lyng{\\aa} (1987) open clusters catalogue and that it is classified as open cluster also by SIMBAD. \\\\ The aim is to provide some photometric data for this cluster, complementing the very poor informations we have, which basically consists of the cluster identification and diameter, which according to Lyng{\\aa} (1987) is about $1.0$ arcmin. Its coordinates for the 2000.0 equinox are: $\\alpha~=~20^{h}~59^{m}~00^{s}$, $\\delta~=~-12^{o}~38{\\prime}$, $l~=~35^{o}.73$, $b~=~-33^{o}.95$.\\\\ NGC~6994 is expected to lie close to HD~358033, GSC~05778--0082, and M~73. This latter might be a binary or multiple system according to SIMBAD, but more probably represents all the asterism. \\\\ From the acquired data, it turns out that NGC~6994 is not a cluster, but just a random enhancement of the four mentioned bright stars.\\\\ Section~2 is dedicated to a brief description of the data acquisition and reduction; Section~3 deals with the cluster structure and CMD. Finally Section~4 summarizes the results. \\begin{table*} \\tabcolsep 0.08truecm \\caption{Basic parameters of the two brighter stars in NGC~6994 region. Magnitudes, proper motions and trigonometric parallax are from the Tycho catalogue. } \\begin{tabular}{cccccccccc} \\hline \\multicolumn{1}{c}{Name} & \\multicolumn{1}{c}{$\\alpha$(2000.0)} & \\multicolumn{1}{c}{$\\delta$(2000.0)} & \\multicolumn{1}{c}{$B$}& \\multicolumn{1}{c}{$V$}& \\multicolumn{1}{c}{$\\mu_{\\alpha}$} & \\multicolumn{1}{c}{$\\mu_{\\delta}$} & \\multicolumn{1}{c}{Spectral type} & \\multicolumn{1}{c}{$\\pi$} & \\multicolumn{1}{c}{dist} \\\\ & hh:mm:ss &($^{o}$)($\\prime$)($\\prime\\prime$)& & &$\\prime\\prime/yr$& $\\prime\\prime/yr$& & mas & pc\\\\ \\\\ HD~358033 & 20:58:57.9&-12:37:45.9& 11.915 & 11.170& 0.0190&-0.006&F5& 7.40&135\\\\ GSC~05778-00802 & 20:58:56.7&-12:38:30.1& 11.688 & 10.427& 0.0075&-0.011& &23.80&42\\\\ \\hline \\end{tabular} \\end{table*} \\section[]{Data Acquisition and Reduction } Observations were conducted at La Silla on 1999 October 8, using the Tektronix 2024$\\times$ 2024 pixel CCD $\\#$ 36 mounted in the red EMMI arm of the 3.6-m ESO NTT telescope. The reader is referred to Maris et al (2000) for any detail about data reduction and acquisition.\\\\ Magnitudes and colors of all the observed stars are available upon request, together with the frame coordinates ($X$ and $Y$) and the instrumental ALLSTAR rms errors $\\sigma$. \\section[]{The cluster} \\subsection{Preliminary considerations} NGC~6994 is a poorly known object. It is classified as a Trumpler (1930) {\\it IV~1~p} cluster, say a poorly populated (4 stars?) and compact system. This classification however is doubtful. Many observers detected this stars concentration, which was baptized also as M~73, Cr~426, C~2056-128, and OCL~89 (Ruprecht 1966). Its angular diameter is estimated to range between 1 to 2.8 arcmin, which means that it should be a rather compact cluster (Collinder (1931) suggested it is a globular).\\\\ It seems that all these suggestions converge to the conclusion that we have to consider NGC~6994 as a group of four bright stars shown in Fig.~1. \\subsection{The structure} NGC~6994 is a cluster located at relatively high galactic latitude ($b=-33.95$). It appears as a concentration of four bright stars in the DSS image presented in Fig~1, although the barycenter of these stars is offset with respect to the commonly reported cluster center, which corresponds to the center of the image. These four stars are rather close, justifying the small diameter reported by Lyng{\\aa} (1987), who presumably referred to these stars when describing NGC~6994 properties. In details, the two brighter stars are HD~358033 and GSC~05778--0082, and their parameters are listed in Table~1, where magnitudes ($B$ and $V$), distance and proper motions are from Tycho catalogue (H$\\o$g et al 1998). Typical errors affecting these magnitudes are 0.3-0.4 mag. These two stars are well inside the solar neighborhood, and lie 135 and 42 pc from the Sun, respectively. M~73 according to Simbad is a double or multiple star, but most probably represents the whole asterism, and it could be considered as a duplicate name for NGC~6994. Moreover we noticed that the faintest star in the asterism might be a visual binary, or simply the two stars are projected in the same sky direction. It is however rather unlikely to imagine that this binary system actually represents M~73. Apart from these four stars, no many other stars are visible close to the cluster center which could justify the classification of this object as an open cluster. Indeed the surrounding field (see Fig.~1) appears very smoothly populated. Looking at the cluster structure it is reasonable to suggest that these four stars are responsible for the by eye identification of this aggregate as an open cluster. \\begin{figure*} \\centerline{\\psfig{file=9522_fig2.ps,height=14cm,width=14cm}} \\caption{ The CMD of the measured stars in the region of NGC~6994 in the V vs (V-I) (right panel) and V vs (B-V) (left panel) planes.} \\end{figure*} \\subsection{The CMD} The measured stars in the plane $V$ versus $(B-V)$ and $V$ versus $(V-I)$ are shown in Fig.~2 (left and right panel, respectively), and represent a region somewhat smaller ($9 \\times 9$ arcmin$^{2}$) than that shown in Fig.~1. They define a broad vertical sequence which gets wider at increasing magnitude. The distribution of stars does not exhibit any distinguishing feature, but resembles a typical field stars CMD. The straightforward conclusion is that NGC~6994 is not an open cluster, but the stars in this region define a rather smooth field population, with some voids and less rich in stars than the typical galactic disk fields. We must stress that we are looking at a relatively high galactic latitude, where the thickness of the disk is rather small. \\subsection{Is NGC~6994 an open cluster remnant?} de la Fuente Marcos (1998) studied the dynamical evolution of open star clusters, suggesting that the final stage of their evolution consists of a handful of stars which emerges from the general galactic field. The life-time, number of remaining stars and dimension of the OCR depends on the initial cluster size and on the distance to the galactic center. As for NGC~6994, we find that there are 11 stars which significantly emerge from the field (see Fig.~2). This would mean that NGC~6994 was a rich open cluster with an initial population of about 700 stars, and an age of almost a billion yrs, or a younger (half a billion yrs) initially less rich (250 stars) open cluster if the binary population was significant ($30\\%$). Since NGC~6994 does not have bright stars (the two brightest ones are probably dwarfs), the most plausible conclusion would be that it was initially a rich populated cluster, with an age of 1 Gyr or more. There are however not enough arguments leading to this scenario. In fact the lack of any feature in the CMD is a strong argument against the classification of this object as an open cluster. Indeed moving groups and OCR actually maintain in time some structures resembling a cluster CMD (see de la Fuente Marcos, 1998, Fig.~3). \\\\ The most reasonable conclusion is that we are looking at the general galactic field, with stars at any distance from the Sun. ", "conclusions": "We reported on photometry in $B$, $V$ and $I$ passbands for NGC~6994, an object previously classified as open cluster.\\\\ The analysis of the observed cluster field, and the distribution of the stars in the CMD seems to favor the suggestion that this object is not really an open cluster, but simply an enhancement of four bright stars in the general galactic field.\\\\ \\noindent The conclusion of this work can be summarized as follows: \\begin{description} \\item[$\\bullet$] NGC~6994 is an asterism of four stars and not an open cluster; \\item[$\\bullet$] NGC~6994 is unlikely to be an OCR; \\item[$\\bullet$] M~73 should not be considered a binary system; instead it can be used as a duplicate name for NGC~6994; \\end{description} \\noindent Carraro \\& Patat (1995) found another object classified as open cluster - Ruprecht~46 - which was proved to be just a random fluctuation in the field of the galactic disk. We wonder whether the possibility exists that some other unstudied or poorly studied open clusters might not be real open clusters, but OCRs or just random star concentration fluctuations in the galactic disk." }, "0003/astro-ph0003144_arXiv.txt": { "abstract": "In this lecture the basic theory of accretion disks is introduced, with emphasis on aspects relevant for X-ray binaries and Cataclysmic Variables. To appear in `The neutron star black hole connection' (NATO ASI Elounda 1999, eds. C. Kouveliotou and V. Connaughton). \\keywords neutron stars, black holes, Caraclysmic Variables, accretion: accretion disks ", "introduction": "Accretion disks are inferred to exist in objects of very different scales: km to millions of km in low Mass X-ray Binaries (LMXB) and Cataclysmic Variables (CV), solar radius-to-AU scale in protostellar disks, and AU-to-parsec scales for the disks in Active Galactic Nuclei (AGN). An interesting observational connection exists between accretion disks and jets (such as the spectacular jets from AGN and protostars), and outflows (the `CO-outflows' from protostars and possibly the `broad-line-regions' in AGN). Lacking direct (i.e. spatially resolved) observations of disks, theory has tried to provide models, with varying degrees of success. Uncertainty still exists with respect to some basic questions. In this situation, progress made by observations or modeling of a particular class of objects is likely to have direct impact for the understanding of other objects, including the enigmatic connection with jets. In this lecture I concentrate on the more basic aspects of accretion disks, but an attempt is made to mention topics of current interest, such as magnetic viscosity, as well. Emphasis is on those aspects of accretion disk theory that connect to the observations of LMXB and CV's. For other reviews on the basics of accretion disks, see Pringle (1981), Treves et al. (1988). For a more in-depth treatment, see the textbook by Frank et al. (1992). ", "conclusions": "" }, "0003/astro-ph0003308_arXiv.txt": { "abstract": "We analyzed the power density spectra of a sample of 9 neutron star and 9 black hole binaries in the low/hard spectral state. In the power density spectra of accreting neutron stars with a weak magnetic field a significant power is contained at frequencies close to one kHz. At the same time, most Galactic accreting black holes demonstrate a strong decline in the power spectra at the frequencies higher than 10--50 Hz. We propose to use this empirical fact as a method to distinguish the accreting neutron stars from black holes in X-ray transients. The X-ray transients that demonstrate significant noise in their X-ray flux at frequencies above $\\sim500$ Hz should be considered neutron stars. We propose to explain the observed difference as a result of the existence of a radiation dominated spreading layer on the neutron star surface (Inogamov\\& Sunyaev 1999). The possible very high frequency variabilities of this layer are discussed. ", "introduction": "Among the Galactic X-ray sources black holes (BH) distinguish themselves by the shape of their X-ray spectrum (see e.g. \\cite{tananbaum72}, \\cite{shef_trump}, \\cite{white_swank}, \\cite{tanaka_shi}). In the low spectral state BHs emit a significant part of their luminosity at energies of hundreds keV (see e.g. \\cite{two_src}), while neutron stars (NS) radiate much smaller part of their total luminosity in this energy range (see e.g. \\cite{barret_review}). A soft component that is present in the spectra of BH binaries in the high(soft) state has a characteristic temperature which can be significantly lower than that in the spectra of NS binaries with similar luminosity. Besides, the BH binaries in this soft spectral state demonstrate a hard power law tail (likely without a high energy cutoff up to 500-600 keV, see e.g. Sunyaev et al. 1988, 1992; \\cite{1655_osse}), whereas bright accreting NSs have never yet shown such spectra (e.g. \\cite{tanaka_shi}). These spectral properties were frequently used as a criterion to determine the nature of the compact object. We will not discuss here the widely accepted methods of establishing neutron star systems through the presence of pulsations or X-ray bursts. The detection of coherent pulsations indicates the presence of the strong magnetic field and rotation of the NS. X-ray bursts (type I) demonstrate that nuclear explosions occur in the matter that was collected at the surface of the NS during the accretion. However there is a significant number of sources, mostly transients, for which neither X-ray pulsations nor X-ray bursts have been observed. Below we propose another method of determining the nature of the compact object based on its power density spectrum (PDS) at the high frequencies f$>$10--100 Hz. ", "conclusions": "" }, "0003/astro-ph0003414_arXiv.txt": { "abstract": "One of the major goals of observational cosmology is to acquire empirical data that has the diagnostic power to develop the theoretical modelling of the high-redshift universe, ultimately leading to an accurate understanding of the processes by which galaxies and clusters form and subsequently evolve. New bolometer arrays operating on the world's largest submillimetre telescopes now offer a unique view of the high-redshift universe through unbiassed surveys with unprecedented sensitivity. For brevity, except when there is a need to be more specific, the FIR to millimetre wavelength regime ($\\rm 100\\mu m < \\lambda < 6000\\mu m$) will be referred to as the ``submillimetre'' (submm). One of the major challenges in this field is to accurately quantify the star-formation history of submm-selected galaxies, particularly those at redshifts $> 1$, and determine their contribution to the submm extragalactic background. The field of observational cosmology will be revolutionized during the course of the next 10 years due to the variety of powerful new ground-based, airborne and satellite facilities, particularly those operating at FIR to millimetre wavelengths. This review summarises the results from the recent blank-field submm surveys, and describes the future observations that will provide accurate source-counts over wider ranges of wavelength and flux-density, constrain the spectral energy distributions of the submm-selected galaxies and accurately constrain the redshift distribution and submm luminosity function by removing the current ambiguities in the optical, IR and radio counterparts. ", "introduction": "In addressing the question `what is the main epoch of metal production in the universe?', or equivalently, `when did the cosmic star-formation rate reach its peak value?', a number of separate lines of evidence suggest that a high-rate of star-formation ($\\gg 100 M_{\\odot} \\rm yr^{-1}$) must have occurred in massive systems at $z \\simeq 3$. This evidence includes (i) the demonstration by Renzini (1998), using clusters of galaxies as probes of the past star-formation and metal production history, that 30--50\\% of the present-day baryons are currently locked up in massive structures which appear to have formed at $z > 3$; (ii) the peak in the co-moving number density of AGN (radio galaxies and quasars) at $z \\sim 2$, AGN whose counterparts at low-redshift are hosted in luminous, massive elliptical galaxies ($> 2L^{\\star}$ - Taylor \\etal\\ 1996, McClure \\etal\\ 1999). At $z \\sim 2$ the universe is only 3--4\\,Gyrs old, which implies that a sustained star-formation rate (SFR) $ > 200 M_{\\odot} \\rm yr^{-1}$ is required to build a massive elliptical galaxy by $z \\sim 2$ (assuming that galaxies hosting high-z AGN have already converted the major fraction of their mass into stars); (iii) the recent discovery of elliptical galaxies at $z \\sim 1.5$ which contain stellar populations with ages of 3--4 Gyrs (Dunlop {\\it et al.} 1996, Peacock {\\it et al.} 1998). Regardless of the cosmological model this requires an extreme formation redshift ($z > 5$) for the initial starburst in these galaxies; (iv) the dramatic increase in the number of star-forming galaxies at high-redshift identified in ground-based and HST faint galaxy samples. Using a Lyman-break colour selection technique (Steidel \\etal\\ 1996), more than 3000 galaxies now have photometric redshifts with $\\sim$ 700 galaxies already spectroscopically confirmed at $z \\simeq 2$ (Adelberger priv.comm.), with SFRs $\\sim 1-5 \\, h^{-2} M_{\\odot} \\rm yr^{-1}$. However the attenuating effects of dust, inevitably associated with star-formation, means that SFRs estimated from these rest-frame UV luminosities must be treated as strict lower-limits. Near-IR observations of rest-frame Balmer-line emission suggest an upward correction factor to the SFRs of $2-15 \\times$ (Pettini \\etal\\ 1998), whilst more robust measurements of SFRs, derived from rest-frame FIR luminosities, imply SFRs 600 times greater than that estimated from the UV luminosity (Hughes {\\it et al.} 1998, Cimatti {\\it et al.} 1998); (v) similar evolution seen in both the radio-source population and the local starburst population, implying that radio source evolution is a good tracer of the star-formation history of the Universe, suggests that the SFR density derived from Lyman-limit galaxies at $z \\simeq 3-4$ is under-estimated by a factor of $\\simeq 5$ (Dunlop 1998), and therefore that once again the star-formation activity in the Universe peaked at $z > 2$; (vi) the small, but increasing number of submm continuum and CO detections of high-$z$ quasars and radio galaxies, indicate that the host galaxies of these powerful AGN contain large quantities of metal-enriched molecular gas ($1-10 \\times 10^{10} M_{\\odot}$, after correcting for gravitational amplification) which can fuel massive bursts of star-formation (Omont \\etal\\ 1996, Hughes \\etal\\ 1997, Combes \\etal\\ 1999). Taken together, the observational evidence suggests that much of the on-going star-formation in the young Universe may be {\\em hidden} by dust from optical surveys and possibly also from IR surveys. Hence the {\\em transparent} view of the Universe provided by submm observations, which now have the instrumental sensitivity to detect high-$z$ dust-enshrouded galaxies forming stars at a rate $> 100 M_{\\odot} \\rm yr^{-1}$, and the preliminary evidence that galaxies (particularly massive spheroidal systems) exhibit strong luminosity evolution at submm wavelengths, demonstrate that comprehensive submm surveys will provide an important alternative measurement of the star-formation history of high-$z$ galaxies unhindered by the effects of dust. ", "conclusions": "" }, "0003/astro-ph0003287_arXiv.txt": { "abstract": "This work studies the optical emission line properties and physical conditions of the narrow line region (NLR) of seven narrow-line Seyfert 1 galaxies (NLS1) for which high signal-to-noise spectroscopic observations were available. The resolution is 340 \\kms\\ (at \\halfa) over the wavelength interval 3700 -- 9500\\,\\AA, enabling us to separate the broad and narrow components of the permitted emission lines. Our results show that the flux carried out by the narrow component of \\hbeta\\ is, on average, 50\\% of the total line flux. As a result, the [OIII] \\lb5007/\\hbeta\\ ratio emitted in the NLR varies from 1 to 5, instead of the universally adopted value of 10. This has strong implications for the required spectral energy distribution that ionizes the NLR gas. Photoionization models that consider a NLR composed of a combination of matter-bounded and ionization-bounded clouds are successful at explaining the low [OIII] \\lb5007/\\hbeta\\ ratio and the weakness of low-ionization lines of NLS1s. Variation of the relative proportion of these two type of clouds nicely reproduce the dispersion of narrow line ratios found among the NLS1 sample. Assuming similar physical model parameters of both NLS1s and the normal Seyfert 1 galaxy NGC\\,5548, we show that the observed differences of emission line ratios between these two groups of galaxies can be explained, to a first approximation, in terms of the shape of the input ionizing continuum. Narrow emission line ratios of NLS1s are better reproduced by a steep power-law continuum in the EUV -- soft X-ray region, with spectral index $\\alpha \\sim -2$. Flatter spectral indices ($\\alpha \\sim -1.5$) match the observed line ratios of NGC\\,5548 but are unable to provide a good match to the NLS1 ratios. This result is consistent with ROSAT observations of NLS1s, which show that these objects are characterized by steeper power-law indices than those of Sy1 galaxies with strong broad optical lines. ", "introduction": "Narrow-Line Seyfert 1 Galaxies (hereafter NLS1) are a peculiar group of AGNs where the permitted optical lines show full width half-maximum (FWHM) not exceeding 2000 \\kms, the [\\ion{O}{3}] \\lb5007/\\hbeta\\ ratio is $<$ 3 and the UV-VIS spectrum is usually very rich in high ionization lines and \\feii\\ emission multiplets. In the soft X-ray band, NLS1s have generally much steeper continuum slopes and rapid variability (Boller, Brandt \\& Fink 1996, hereafter BBF96). Recently, Leighly (1999) found that the hard X-ray photon index is significantly steeper in NLS1s compared with that of normal Seyfert 1s, and that soft excess emission appears considerably more frequently in NLS1s than in Seyfert 1 (hereafter Sy1) galaxies with broad optical lines. It is not known at present the origin of the narrowness of broad permitted lines in NLS1s. Osterbrock \\& Pogge (1985); Ulvestad, Antonucchi \\& Goodrich (1985) and Stephens (1989) suggest that if the velocities in the BLR of Seyfert 1s were largely confined to a plane, the NLS1 galaxies could be understood as cases in which the line of sight is nearly perpendicular to this plane. BBF96 state, on the other hand, that if the gravitational force from the central black hole is the dominant cause of the motions of Seyfert BLR clouds, narrower optical emission lines will result from smaller black hole masses provided the characteristics BLR distance from the central source does not change strongly with black hole masses. However, Rodr\\'{\\i}guez-Pascual, Mass-Hesse \\& Santos-Ll\\'eo (1997) report the detection in NLS1 galaxies of broad components with FWHM around 5000 \\kms\\ for the high ionization UV permitted lines such as Ly$\\alpha$, \\ion{C}{4} \\lb1550 and \\ion{He}{2} \\lb1640. This result indicates that gas moving at velocities comparable to those found in typical Sy1 galaxies does indeed exist in NLS1s. In the optical region, they found ``broad'' components with FWHM less than 3000 \\kms, narrower than the broadest UV component in the same objects. Nonetheless, deblending the optical permitted lines in NLS1s is difficult because no transition between the narrow and broad components is observed. This shortcoming has strong influences in, for example, the analysis of the narrow line region (NLR) due to the large uncertainties in determining the fraction of \\halfa\\ and \\hbeta\\ which originates from low ionization material. Up to now, most studies of the NLR in NLS1s assume that the flux emitted by the narrow \\hbeta\\ equals 10\\% of the flux of [\\ion{O}{3}] \\lb5007 (Osterbrock \\& Pogge 1985; Leighly 1999). This assumption is based on the results obtained from Seyfert 2 and intermediate Sy1 galaxies (e.g. Koski 1978; Cohen 1983). But in recent years, growing observational evidence points out to the existence of differences between the NLR of normal Sy1 and Sy2 galaxies (Schmitt \\& Kinney 1996; Schmitt 1998), making the above assumption highly uncertain. In addition, fixing the [\\ion{O}{3}] \\lb5007/\\hbeta\\ ratio to 10 implies ignoring the large scatter in the value of this ratio observed in normal Sy1s (2 to 19, see for example Rodr\\'{\\i}guez-Ardila, Pastoriza \\& Donzelli 1999, hereafter Paper\\,I) and overlook the influences that this ratio could have in the energetics and physical conditions of the NLR of these objects. Due to the above reason, the main purpose of this paper is to seek additional constrains in order to estimate the actual contribution of the narrow \\hbeta\\ flux to the total \\hbeta\\ emission line and study the implications that the newly adopted values could have on narrow line ratios and the physics of the NLR of NLS1s. The present work is organized as follows. In Section~\\ref{observ} we describe the sample of NLS1s used in this paper. Section~\\ref{gauss} presents the decomposition into narrow and broad components carried out in the Balmer lines of the NLS1 galaxies. Photoionization models that successfully reproduce the observed line ratios of NLS1s are presented in Section~\\ref{MBIB}. A discussion of the main results appears in Section~\\ref{bla} and the conclusions are presented in Section~\\ref{fin}. ", "conclusions": "\\label{fin} We have analyzed long-slit spectral data of a sample composed of seven NLS1 galaxies. A decomposition of the \\halfa\\ and \\hbeta\\ emission line profiles into Gaussian components allowed us to separate the flux contribution of the NLR from the total flux of the line. Our results show that, on average, 50\\% of the total \\hbeta\\ flux is due to emission from the NLR. Using the [\\ion{O}{3}] \\lb5007 line profile as a template for the narrow lines in order to subtract this contribution in the permitted lines give very similar results to those obtain throught the Gaussian decomposition. This confirms the presence of a broad component in the permitted lines. The FWHM of the broad components of \\halfa\\ and \\hbeta\\ in the NLS1s studied here seems to be rather uniform within the same galaxy after comparing \\halfa\\ and \\hbeta\\ and throughout the sample (2250 \\kms\\ and 2560 \\kms\\ for \\halfa\\ and \\hbeta, respectively). No evidence of Lorentzian profiles was observed neither in the narrow nor the broad lines. The narrow components of \\halfa\\ and \\hbeta\\ present FWHM comparable to those of the forbidden lines which are typical of the NLR in any Seyfert. The resulting [\\ion{O}{3}] \\lb5007/\\hbeta\\ ratios fall in the interval 1--5, significantly lower than the value currently assumed ($\\sim$10). This entails that the emission line ratios from the NLR are different in NLS1s from those observed in normal and intermediate Sy1 galaxies. We test photoionization models that consider a NLR composed of a combination of matter-bounded clouds and ionization-bounded clouds. The former, with typical densities $\\sim 10^{6}$ \\cm\\ and photoionized by the intrinsic continuum from the central source, are responsible for the emission of most of the [\\ion{O}{3}] and high ionization lines. This component should be located in the inner regions of the NLR. The latter, located farther out than the MB clouds and characterized by a lower density ($n_{e} \\sim 10^{3}$ \\cm), are photoionized by the continuum filtered from the MB clouds and emits most of the low ionization lines. Assuming similar physical parameters in the NLR of NLS1s and normal Sy1s, we show that the observed differences in emission line ratios between these two groups of galaxies can be explained in terms of differences in the form of the input ionizing spectra. NLS1s ratios are better reproduced with a steep power-law continuum, with spectral index $\\alpha < -2$ while flatter spectral indices ($\\alpha \\sim -1.5$) match the observed line ratios in normal Sy1s. This scenario reproduces with very good agreement the line ratios of NLS1s. It is furthermore consistent with ROSAT observations of NLS1s, which show that these objects are characterized by steeper power-law indices than those of Sy1 galaxies with broad optical lines. Our modeling therefore support the view that the NLR is directly photoionized by the unaltered SED distribution emitted by the central engine. \\clearpage" }, "0003/astro-ph0003078_arXiv.txt": { "abstract": "A new method, based on the simulated annealing algorithm and aimed at the inverse problem in the analysis of intergalactic (interstellar) complex spectra of hydrogen and metal lines, is presented. We consider the process of line formation in clumpy stochastic media accounting for fluctuating velocity and density fields (mesoturbulence). This approach generalizes our previous Reverse Monte Carlo and Entropy-Regularized Minimization methods which were applied to velocity fluctuations only. The method allows one to estimate, from an observed system of spectral lines, both the physical parameters of the absorbing gas and appropriate structures of the velocity and density distributions along the line of sight. The validity of the computational procedure is demonstrated using a series of synthetic spectra that emulate the up-to-date best quality data. H\\,{\\sc i}, C\\,{\\sc ii}, Si\\,{\\sc ii}, C\\,{\\sc iv}, Si\\,{\\sc iv}, and O\\,{\\sc vi} lines, exhibiting complex profiles, were fitted simultaneously. The adopted physical parameters have been recovered with a sufficiently high accuracy. The results obtained encourage the application of the proposed procedure to the analysis of real observational data. ", "introduction": "QSO absorption line spectroscopy being a major activity at many observatories for the last two decades is now developing into a powerful tool for extragalactic research thanks to the new generation of large telescopes. The steady improvement in sensitivity and resolution of spectroscopic instrumentation opens new fields in the study of QSO absorption systems. It is now becoming possible to investigate the intensity fluctuations within the line profiles and thus to estimate hydrodynamic characteristics of the absorbing gas. The majority of the narrow QSO absorption lines represents intervening systems and allows us to probe the properties of diffuse matter at very high redshifts. Resolved profiles of hydrogen lines and especially lines of heavier elements (`metals') show a diversity of shapes and structures. Up to now, their analysis is based on the assumption that the observed complexity is caused by individual `clouds' randomly distributed along the line of sight with slightly different radial velocities. It is also a basic assumption that the hydrodynamic (`bulk' or `turbulent') velocity distribution inside each cloud is Gaussian and completely uncorrelated ({\\it microturbulence}). This model implies that each subcomponent of the complex profile being resolved should have a symmetrical profile and its shape should not alter with higher spectral resolution. Observations show, however, that the complexity of the line profiles increases with higher resolution, a tendency expected for correlated bulk motions which have, in general, non-Gaussian distributions along a given line of sight (Levshakov \\& Kegel 1997; Levshakov, Kegel \\& Mazets 1997; Levshakov, Kegel \\& Takahara 1999; Papers~I, II, and III hereafter, respectively). It follows that the microturbulent approximation is not appropriate in this case because it does not account for all the relevant physical processes involved in the radiative transfer. Moreover, being applied to real data, the microturbulent analysis leads to a loss of valuable information contained in the observations and may even yield unphysical results (Levshakov \\& Kegel 1999; Levshakov, Takahara \\& Agafonova 1999; LTA hereafter). The need for more sophisticated procedures of data analysis becomes therefore obvious. In recent years, it has been shown that accounting for the correlations in the velocity field ({\\it mesoturbulence}) may change the interpretation of the line measurements substantially (Papers~I and II). A mesoturbulent approach has been already successfully applied to the study of the deuterium and hydrogen absorption in Q~1937--1009 (Levshakov, Kegel \\& Takahara 1998a), Q~1718+4807 (Levshakov, Kegel \\& Takahara 1998b), and Q~1009+2956 (Levshakov, Tytler \\& Burles 2000). For all three QSOs about the same value for the D/H ratio was derived in contrast to the previously announced microturbulent results. Our first inversion codes, -- the Reverse Monte Carlo (Paper~III), and the Entropy-Regularized Minimization (LTA), -- have been developed for a model of a stochastic velocity field neglecting any density fluctuations. They have been applied to the analysis of the H\\,{\\sc i} and D\\,{\\sc i} lines and/or to the metal absorption lines with similar profiles when species trace the same volume elements independently on the density fluctuations. In the present paper, we extend this study to the inverse problem for a model of compressible turbulence when one observes non-similar profiles of different atoms and/or ions from the same absorption-line system. As in our previous papers, we use the term `turbulence' in a wider sense as compared with hydrodynamic turbulence to label the unknown nature of the line broadening mechanism. In this regard we consider any kind of bulk motions (infall, outflows, tidal flows etc.) of more or less stochastic nature leading to fluctuating velocity and density (temperature) fields as continuous random functions of the space coordinate along a given line of sight within the intervening absorbing region. Two noteworthy works have been recently carried out aiming at the recovery of the physical intergalactic structure from the Lyman-$\\alpha$ forest lines. Nusser \\& Haehnelt (1999a,b) developed an inverse procedure based on the relation between density and velocity Fourier coefficients. The quality of their recovery is, however, restricted by the assumption that the Lyman-$\\alpha$ forest structure traces mainly the matter density distribution and that the amplitude of the peculiar velocities is rather small to affect the local absorption coefficient significantly. This assumption is questionable since there is no simple way to distinguish {\\it observationally} whether the density or the velocity fluctuations are the main source of the `line-like' structure observed in the Lyman-$\\alpha$ forest (Levshakov \\& Kegel 1998). Moreover, recent studies of nearby large-scale motions in the universe indicate that the Hubble flow is considerably perturbed. Peculiar velocities in the range from 300 to 500 km~s$^{-1}$ have been found in a sample of galaxies complete out to a distance of 60~Mpc (e.g., Watkins 1997; Gramann 1998; Giovanelli et al. 1998), a fact which should be taken into account in the inverse procedures. The method described in the present paper is quite flexible and equally accounts for the density and velocity fluctuations. It is based on a stochastic optimization approach similar to that developed in Paper~III. We estimate simultaneously the physical parameters {\\it and} appropriate realizations of the density $n(s)$ and velocity $v(s)$ distributions along the line of sight to reproduce hydrogen and metal lines from a given absorption system. In this regard, the more spectra of different elements are incorporated in the analysis the higher accuracy of the estimation can be obtained. In $\\S 2$ our model and the underlying basic assumptions are specified. The inversion code is described in $\\S 3$. The validity of the method is tested in $\\S 4$ using simulated sets of noisy line profiles (H\\,{\\sc i}, C\\,{\\sc ii}, Si\\,{\\sc ii}, C\\,{\\sc iv}, Si\\,{\\sc iv}, and O\\,{\\sc vi}). Finally, the main conclusions are outlined in $\\S 5$. ", "conclusions": "We have developed a new method to solve the inverse problem in the analysis of intergalactic (interstellar) hydrogen and metal lines arising from clumpy stochastic media. In the method, the random velocity and density configurations along the line of sight are approximated by Markovian processes. The global optimization method based upon simulated annealing is then used to fit theoretical line profiles to a set of `observational' data. The proposed procedure allows us to estimate the physical parameters of the absorbing gas such as column densities, metal abundances, mean (density-weighted) kinetic temperatures for each ion, and mean ionization parameter together with the hydrodynamic characteristics -- the radial velocity dispersion and the dispersion of the density fluctuations. The computational scheme has been tested on a variety of synthetic spectra that emulate modern observational data~: the absorption lines of H\\,{\\sc i}, C\\,{\\sc ii}, Si\\,{\\sc ii}, C\\,{\\sc iv}, Si\\,{\\sc iv}, and O\\,{\\sc vi} which are usually observed in the Lyman limit systems ($N_{{\\rm H}\\,{\\sc i}} \\la 3\\times10^{17}$ cm$^{-2}$). The ionization structure of the absorbing region was calculated using the standard photoionization model of Donahue \\& Shull (1991) with a background ionizing spectrum given by Mathews \\& Ferland (1987). The inversion procedure proved to be very effective and robust allowing us to recover the physical parameters with reasonable accuracy albeit the structure of the random velocity and density fields cannot be restored with a pixel-to-pixel conformity. However, the integral characteristics of these random fields, namely, the density-weighted velocity distribution, can be estimated quite precisely. Thus we can conclude that our procedure provides reliable results and can be applied to the analysis of real data. Note that while performing the inversion of absorption lines, one has to take into account the following. All our computational tests have been carried out under the assumption that the spectrum of the ionizing radiation is known, i.e. we used the same Mathews \\& Ferland spectrum to generate `observational' data and to fit them with our theoretical profiles. In reality, the characteristics of the ionizing radiation are not known exactly. Therefore in real applications several types of the background photoionizing spectra should be tried. The problem how the computational results are affected by different types of the background ionizing radiation will be studied in detail elsewhere. The proposed method has been successfully applied to the analysis of QSO high resolution spectral data with possible deuterium absorption at $z_{\\rm a} = 3.514$ towards APM~08279+5255 (Levshakov, Agafonova \\& Kegel 2000). It has been demonstrated that the blue-side asymmetry of the hydrogen Ly$\\alpha$ line can be explained quite naturally by an asymmetric configuration of the velocity field only. The results obtained revealed a considerably lower neutral hydrogen column density as compared with the VPF measurements performed by Molaro et al. (1999). In contrast to Molaro et al., we have managed to fit simultaneously all absorption lines observed in this system. These results can be considered as encouraging and favor the application of the developed computational procedure to the analysis of other high quality observational data. \\medskip\\noindent" }, "0003/hep-ph0003298_arXiv.txt": { "abstract": "We provide a general overview of the velocity-dependent one-scale model for cosmic string evolution and discuss two further extensions to it. We introduce and justify a new ansatz for the momentum parameter $k$, and also incorporate the effect of radiation backreaction. We thus discuss the evolution of the basic large-scale features of cosmic string networks in all relevant cosmological scenarios, concentrating in particular on the `scaling' solutions relevant for each case. In a companion paper, we show, by comparing with numerical simulations, that this model provides an accurate description of the large-scale features of cosmic string networks. ", "introduction": "\\label{sint} The velocity-dependent one-scale (VOS) model provides the most convenient and reliable method by which to calculate the large-scale quantitative properties of a string network in cosmological and other contexts\\cite{ms1,ms2,thesis}. It is widely used for making quantitative predict ions of the potential observational implications of cosmic strings\\cite{vsh}. Given its simplicity, it is remarkable how well the VOS model performs when tested against high resolution numerical simulations of string networks\\cite{test}. It is well-known that string evolution is a complex physical process with a build-up of small-scale structure on the strings, which is very computationally demanding to model accurately\\cite{at,bb,as}. Analytic approaches like the VOS model, abandon the possibility of describing the statistical physics of the string network accurately and concentrate instead on its thermodynamics. In other words, a small number of macroscopic quantities are selected and the microscopic string equations of motion are used to derive evolution equations for these averaged quantities. The price to be paid in this approach is that the averaging process introduces phenomenological parameters whose values are not specified by the model itself. Instead, one must still fix these parameters by direct comparison with numerical simulations. The VOS model is a generalization of the `one-scale' model pioneered by Kibble\\cite{kib} (see also ref.\\cite{Ben}) which describes string motion in terms of a single correlation length $L$. By incorporating a variable rms velocity $v$, the VOS model extends its validity into early regimes with frictional damping and across the important matter-radiation transition, thus giving a quantitative picture of the complete history of a cosmic string network. Other analytic approaches to string evolution have attempted to incorporate the additional small-scale structure seen in numerical simulations. This includes a `kink-counting' model\\cite{AllCal}, a functional approach \\cite{emb}, a `three-scale' model\\cite{ack} and a `wiggly' model \\cite{prep3}. While these are important for characterising detailed network features, they introduce a significant number of further phenomenological parameters which must be fixed by simulations (and which remain rather uncertain). Nevertheless, for describing the large-scale properties of a long-string network, the VOS model has proved to be sufficient for a good quantitative fit using only a single parameter, the loop chopping efficiency $\\tilde c$. The purpose of the present paper is, first, to provide a concise exposition of the VOS string evolution model. We summarise how it can be applied to describe cosmological string evolution, including late times with a cosmological constant, and we present the very different histories of both GUT- and electroweak-scale strings. Secondly, we propose an improvement of the VOS model by presenting a new ansatz for the momentum parameter $k$, which we justify both analytically and numerically. Thirdly, we present a further extension incorporating radiation backreaction, which provides small corrections to the cosmological scaling laws and which also compares favourably with published results of global string simulations. Finally, we review generalizations in a curved FRW spacetime, giving some further asymptotic scaling solutions. We report on detailed comparisons between numerical string simulations and the VOS model elsewhere \\cite{test,frac,prep2}. ", "conclusions": "\\label{sdsc} In this paper we have presented a modified version of the velocity-dependent one-scale (VOS) model \\cite{ms1,ms2,ms3,thesis} which depends on a single free parameter, the loop chopping efficiency ${\\tilde c}$. We have tested it against the largest and most accurate numerical simulations to date \\cite{frac,prep2}, and we find that it provides a good fit to the large-scale scaling properties of the string network in both the radiation and matter epochs, as well as in the transition between the two eras---we will describe these tests elsewhere \\cite{test}. These facts and its intrinsic simplicity make this model particularly suited for any analytic or semi-analytic study of cosmic strings where one is only interested in the large-scale features of the network. We have re-analysed some simple evolutionary properties of cosmic string networks in the light of the VOS model and corresponding numerical simulations. An important conclusion to note is that any realistic cosmic string network is {\\it not} scaling at any time from just before the epoch of equal matter and radiation through to the present day. This is something that must be properly taken into account particularly when discussing string-seeded structure formation scenarios with GUT-scale strings. The extended VOS model is also valid when deviations from scaling are even larger at late times in a universe which becomes dominated by curvature or a cosmological constant. Finally, we considered the effects of radiation back-reaction on the scaling properties of the long string network, and we have shown that although the existence (or otherwise) of a scale-invariant attractor solution will not be affected, the quantitative scaling properties can be. In some cases, the suppression of string density can be quite dramatic (as we saw for small-scale global string simulations), although the string velocities always remain relativistic. For the most part, however, the density of a cosmic string network, whether local or global is only affected slightly by radiation backreaction effects. Despite the many virtues of the VOS model, we are aware, of course, that the small number of available degrees of freedom means that this model is unable to provide a proper description of the small-scale properties of the network; these are important in a number of cosmological scenarios (and sometimes even crucial). Nevertheless, we believe that the phenomenological parameter $k$ does encode some important small-scale structure effects, though clearly a more detailed analytic and numerical study is still required. A number of possible approaches to the problem of string small-scale structure have been suggested in the literature \\cite{emb,ack}, and our own analysis using Carter's elastic string model \\cite{carter} will be discussed in a forthcoming publication \\cite{prep3}." }, "0003/nucl-th0003066_arXiv.txt": { "abstract": "The mixed phase of quarks and hadrons which might exist in the dense matter encountered in the varying conditions of temperature and trapped neutrino fraction in proto-neutron stars is studied. The extent that the mixed phase depends upon the thermodynamical parameters as well as on the stiffness of matter in the hadronic and quark phases is discussed. We show that hadronic equations of state that maximize the quark content of matter at a given {\\it density} generally minimize the extent of the mixed phase region in a neutron star of a given mass, and that only in extreme cases could a pure quark star result. For both the Nambu Jona-Lasinio and MIT bag quark models, neutrino trapping inhibits the appearance of a mixed phase which leads to possible proto-neutron star metastability. The main difference between the two quark models is the small abundance of strange quarks in the former. We also demonstrate that $\\partial T/\\partial n<0$ along adiabats in the quark-hadron mixed phase, opposite to what is found for the kaon condensates-hadron mixed phase. This could lead to core temperatures which are significantly lower in stars containing quarks than in those not containing quarks. \\bigskip\\noindent PACS: 97.60.Jd, 21.65.+f, 12.39.Fe, 26.60.+c \\\\ ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003023_arXiv.txt": { "abstract": "We present the results of ROSAT High Resolution Imager (HRI) observations and the survey data of the radio-loud Narrow-Line Seyfert 1 galaxy (NLS1) PKS~0558--504. We find strong and persistent X-ray variability on both short and medium time-scales. The most extreme amplitude variations require a radiative efficiency exceeding the theoretical maximum for a Schwarzschild black hole, suggesting the presence of a rotating black hole or the influence of relativistic beaming effects. The spatial analysis rules out the possibility that the high luminosity and the strong variability are related to a nearby source. ", "introduction": "Narrow-line Seyfert 1 galaxies are identified by their optical emission line properties: the ratio [O III]/H$\\beta$ is less than 3 and FWHM H$\\beta$ is less than 2000${~\\rm km~s^{-1}}$ (Osterbrock \\& Pogge 1985, Goodrich 1989). Their optical spectra are also characterized by the presence of strong permitted Fe II, Ca II, O I $\\lambda$ 8446 lines (Persson 1988). NLS1 exhibit characteristic features at other wavelengths as well: they are seldom radio loud (Ulvestad et al. 1995, Siebert et al. 1999, Grupe et al. 1999, 2000) and they are usually strong infrared emitters (Moran et al. 1996). In X-rays NLS1 have been generally found to have extreme spectral and variability properties that might be related to an extreme value of a fundamental physical parameter, originating from the vicinity of a supermassive black hole (e.g. Brandt \\& Boller 1998). \\begin{table} \\caption{Long term X-ray variability} \\begin{center} \\begin{tabular}{llll} \\hline \\noalign{\\smallskip} Satellite & {\\rm \\G}& ${\\rm L}_{\\rm 0.2-2.4~keV}$ & Reference\\\\ \\noalign{\\smallskip} & & $({\\rm erg~s^{-1}})$ & \\\\ \\noalign{\\smallskip} \\hline \\hline \\noalign{\\smallskip} \\noalign{\\smallskip} EINSTEIN & $2.21^\\dag$ & $3.2 \\times10^{45}$ & Elvis et al. (1992)\\\\ \\noalign{\\smallskip} \\hline \\noalign{\\smallskip} EXOSAT & 2.21& $2.8 \\times10^{45}$ & Lawson et al. (1992)\\\\ \\noalign{\\smallskip} \\hline \\noalign{\\smallskip} & $2.24^{+0.08}_{-0.08}$ & $1.5\\times10^{45}$ \\\\ GINGA$^\\ddag$ & & & Remillard et al. (1991)\\\\ & $1.92^{+0.12}_{-0.12}$ & $2.0\\times10^{45}$ \\\\ \\noalign{\\smallskip} \\hline \\noalign{\\smallskip} ROSAT & $3.1^{+0.05}_{-0.06}$& $5.4\\times10^{45}$& Brinkmann et al. (1997)\\\\ \\noalign{\\smallskip} \\hline \\noalign{\\smallskip} ASCA & $2.25^{+0.03}_{-0.03}$& $3.1 \\times10^{45}$ &Leighly (1999b)\\\\ \\noalign{\\smallskip} \\hline \\end{tabular} \\end{center} \\dag~ Photon index taken from the EXOSAT observations.\\\\ \\ddag~ The values refer to quiescence and peak, respectively. \\end{table} PKS~0558--504 ($z=0.137, m_{\\rm B}=14.97$) is one of the few radio-loud NLS1 galaxies ($R_{\\rm L}=f_{\\rm 5 GHz}/f_{\\rm B}\\simeq 27$, Siebert et al. 1999). It was optically identified on the basis of X-ray positions from the High Energy Astronomy Observatory (HEAO-1, Remillard et al. 1986). A Ginga observation (Remillard et al. 1991) showed an increase of the X-ray flux by 67\\% in 3 minutes, implying that the apparent luminosity must be enhanced by relativistic beaming. Further X-ray observations with different satellites have confirmed the steep X-ray spectrum and high luminosity of this source, but no more relativistic flares have been presented in the literature. It is important to search for such flares with an X-ray imaging detector to definitively rule out the possibility that the Ginga data suffered from source confusion. Tab. 1 summarizes the luminosities, observed by previous X-ray instruments, converted to the ROSAT soft X-ray band. The conversion to luminosities in the 0.2--2.4 keV energy band was performed using PIMMS, assuming Galactic absorption (\\NH = $4.39\\times10^{20}{~\\rm cm^ {-2}}$, Dickey \\& Lockman 1990) and a power law spectral model with photon index {\\rm\\G} ranging between 2.1 (Remillard et al. 1991) and 3.1 (Brinkmann et al. 1997). However, possible deviations from a single power law or long term spectral changes can lead to systematic uncertainties. The measured soft X-ray spectrum is rather steep and the medium energy power laws are considerably flatter, but the sparse data do not allow determination of whether the source shows spectral steepening towards lower energies or whether long term spectral changes occur during intensity variations. The luminosities were calculated by assuming a Friedman cosmology with $H_0=70 {~\\rm km ~s^{-1}~Mpc^{-1}}$, $q_0=0.5$ and isotropic emission. In this paper we report the results of two ROSAT HRI observation campaigns taken five months apart (in November 1997 and April 1998) and the survey PSPC data (September 1990), with the purpose to check whether the strong X-ray variability is persistent and whether a nearby source contributes to the X-ray flux. In section 2 we present the observations and the spatial analysis. Section 3 deals with the variability of PKS~0558--504. Section 4 contains the main conclusions. \\begin{figure} \\psfig{figure=Cb081f1.ps,width=8.7cm,height=8.7cm,% bbllx=88pt,bblly=299pt,bburx=530pt,bbury=744pt,clip=} \\caption {Contour plot of the X-ray emission of PKS~0558--504 overlaid on the optical image of a $16\\farcm67 \\times 16\\farcm67$ region. The contours correspond to 3, 6, 15, 25, 50 and 3000 \\s~ above background.} \\end{figure} ", "conclusions": "We have presented ROSAT HRI observations of the radio-loud NLS1 galaxy PKS~0558--504. The main results can be summarized as follows: From the spatial analysis, no other strong X-ray sources have been detected in the neighborhood of PKS~0558--504, therefore external contributions to the high luminosity and to the strong variability from a nearby source are ruled out. By comparing the X-ray observations throughout the last decade, it is evident that the strong X-ray variability of PKS~0558--504 occurs persistently. During the ROSAT HRI observations, PKS~0558--504 shows strong variability, both on medium (months) and short (days, hours) time scales. The most extreme variation implies a radiative efficiency larger than the theoretical maximum for accretion onto a Schwarzschild black hole, and our findings generally support those of Remillard et al. (1991) where a relativistic $(\\eta\\simgt 2)$ flare was discovered. As PKS~0558--504 is a radio-loud object, beamed emission from a jet could be the cause for the brightness and variability in X-rays. However, it is worth noting that the radio-quiet NLS1 PHL~1092 has also shown a relativistic flare $(\\eta\\simgt 0.6)$, and radio-quiet NLS1 more generally show enhanced X-ray variability. The soft X-ray spectrum is rather steep with a power law photon index of $\\Gamma \\sim 3.0$ and shows no strong indications for spectral breaks. The obtained medium energy power laws are considerably flatter ($\\Gamma \\sim 2.2$) but the sparse data and the limited energy bands of the different instruments do not allow determination of whether the source shows a spectral steepening towards lower energies or whether long term spectral changes occur during intensity variations. An answer to these vital questions can only be given by the current broad band X-ray missions like SAX, XMM-Newton or Chandra." }, "0003/astro-ph0003486_arXiv.txt": { "abstract": "We report new observations of the spectrum of Ganymede in the spectral range 1160 -- 1720 \\AA\\ made with the Space Telescope Imaging Spectrograph (STIS) on HST on 1998 October 30. The observations were undertaken to locate the regions of the atomic oxygen emissions at 1304 and 1356 \\AA, previously observed with the Goddard High Resolution Spectrograph on HST, that \\citet{hal98} claimed indicated the presence of polar aurorae on Ganymede. The use of the $2''$ wide STIS slit, slightly wider than the disk diameter of Ganymede, produced objective spectra with images of the two oxygen emissions clearly separated. The \\Oone\\ emissions appear in both hemispheres, at latitudes above $|40|$\\deg, in accordance with recent Galileo magnetometer data that indicate the presence of an intrinsic magnetic field such that Jovian magnetic field lines are linked to the surface of Ganymede only at high latitudes. Both the brightness and relative north-south intensity of the emissions varied considerably over the four contiguous orbits (5.5 hours) of observation, presumably due to the changing Jovian plasma environment at Ganymede. However, the observed longitudinal non-uniformity in the emission brightness at high latitudes, particularly in the southern hemisphere, and the lack of pronounced limb brightening near the poles are difficult to understand with current models. In addition to observed solar \\Hone\\ Lyman-$\\alpha$ reflected from the disk, extended Lyman-$\\alpha$ emission resonantly scattered from a hydrogen exosphere is detected out to beyond two Ganymede radii from the limb, and its brightness is consistent with the Galileo UVS measurements of \\citet{bar97}. ", "introduction": "Recent observations from both the Galileo spacecraft and the Hubble Space Telescope (HST) have considerably altered our knowledge of the atmospheres of the Jovian satellites Europa and Ganymede. Both are now known to have tenuous atmospheres (column density $\\sim5\\times10^{14}$\\,cm$^{-2}$) with both molecular oxygen (Europa and Ganymede) and atomic hydrogen (Ganymede) components. The Galileo UV spectrometer detected \\Hone\\ \\LA\\ emission at Ganymede from a hydrogen exosphere \\citep{bar97}, and charged particle measurements indicated that there is also an outflow of protons, implying ongoing gas production \\citep{fra97}. The oxygen component was detected through HST/Goddard High Resolution Spectrograph (GHRS) observations of the atomic oxygen multiplets \\Oone\\ \\lam1304 and \\Oone]\\,\\lam1356 \\citep{hal95,hal98}. The intensity ratio of these emissions implies that the primary source is electron dissociative excitation of molecular oxygen. The source of both the hydrogen and O$_2$ is thought to be sputtering of surface water ice by Io plasma torus ions. Beyond the fact that they exist, very little is known about these atmospheres, including their vertical structure, areal coverage, and variability, which could be significant if the dominant source is surface sputtering because of the asymmetric nature of the plasma bombardment. Plasma bombardment of the surface is also supported by the recent detection of ozone and O$_2$ embedded in the surface ice of Ganymede, and SO$_2$ embedded in the surface ice of Europa \\citep{spe95,cal96,nol95,nol96}. Galileo magnetometer measurements have also shown strong perturbations in the Jovian magnetic field near Ganymede \\citep{kiv96,kiv97}. The measured perturbations indicate that the satellite possesses a magnetic field sufficiently strong ($\\sim$1500\\,nT) to overpower Jupiter's ambient field, and that Ganymede's magnetic and spin axes are roughly aligned \\citep{kiv96}. Near Ganymede closest approach the plasma wave experiment also detected a significant population of trapped, charged particles \\citep{gur96}, implying that Ganymede possesses a ``magnetosphere within a magnetosphere.\" Consistent with these results, the HST/GHRS observations \\citep{hal98} have raised the intriguing possibility that Ganymede exhibits polar aurora. In those spectra the Ganymede \\Oone]\\,\\lam1356 emission line exhibits a doubly-peaked profile that is inconsistent with that of a diffuse source filling the aperture, or with emission from a uniform disk. \\citeauthor{hal98} postulated that the double-peaked profile implies the existence of a similarly double-peaked structure in the spatial distribution of the emission source within the aperture, with the strongest emissions coincident with Ganymede's north and south polar regions. In this paper, we report ultraviolet objective grating images of Ganymede made with the Space Telescope Imaging Spectrograph (STIS) (installed in HST in February 1997) which confirm the spatial distribution inferred from the earlier observation and raise new questions about the interaction of Ganymede's atmosphere with the Jovian magnetosphere. ", "conclusions": "Objective grating images of Ganymede obtained with HST/STIS show clearly separated \\Oone\\ emissions confirming the result of \\citet{hal98} that the emissions are confined to polar regions (latitudes above 45\\deg). The total fluxes are consistent with those reported by \\citeauthor{hal98} but appear to vary in time and in the relative intensities between northern and southern hemispheres. The \\Oone]\\,\\lam1356/\\Oone\\ \\lam1304 ratio is consistent with the primary excitation mechanism being electron impact on O$_2$, as postulated by \\citeauthor{hal98} While the spatial distribution of the emissions is consistent with current models of the magnetic field of Ganymede, expected longitudinal uniformity and limb brightening are not observed. In addition, \\LA\\ limb emission from a hydrogen exosphere is detected and the measured brightness is found to be in good agreement with the Galileo UVS observations of Barth \\etal\\ (1997)." }, "0003/astro-ph0003165_arXiv.txt": { "abstract": "The neutral hydrogen line profiles along the major axis of the nearby spiral galaxy NGC~2403 show a wing towards the systemic velocity. This asymmetry can be explained with the presence of an abnormally thick HI disk (FWHM $\\sim$ 5 kpc) or with a two-component structure: a thin disk and a slowly rotating, thicker (1-3 kpc) HI layer. The latter model gives a better representation of the observations. These results throw a new light on the disk-halo connection in spiral galaxies. In particular, the decrease of rotational velocity with height above the plane may be the result of a galactic fountain flow. A vertically extended, slowly rotating HI layer may be common among spiral galaxies with high levels of star formation. ", "introduction": "In recent years there has been a growing interest for the vertical structure of the HI disks of spiral galaxies and for the disk-halo connection. Evidence for HI gas flows between disk and halo comes from the detection of large vertical motions in several galaxies viewed close to face-on (Dickey et al.~1990; Kamphuis 1993; Schulman \\& Bregman 1994) and the connection of this high velocity gas with HI holes and star formation activity in the disk (Kamphuis \\& Sancisi 1993). More evidence comes from the study of edge-on galaxies like NGC~891 in which the HI has been found to extend up to at least 5 kpc into the halo, where it seems to rotate about 25 km ${\\rm s}^{-1}$ more slowly than in the plane (Swaters et al.~1997). The combination of the results from face-on and edge-on galaxies leads to the picture of effervescent galaxies (Sancisi et al.~1996), consistent with galactic fountains models (Bregman 1980; Spitzer 1990). \\begin{figure*}[t] \\center{\\scalebox{.8}{\\includegraphics{figure1.ps}}} \\parbox{16.5cm}{\\caption{\\small Optical image (top left; from the Palomar Digitized Sky Survey), HI column density map (top right; the column densities range from $7.5 \\cdot 10^{19}$ to $3.3 \\cdot 10^{21}$ atoms/cm$^2$) and HI velocity field (bottom left; the contours run from 14 to 254 km ${\\rm s}^{-1}$ in steps of 15 km ${\\rm s}^{-1}$.). The beam ($45'' \\times 45''$) is shown in the lower left corner. The bottom right panel shows the HI position-velocity map along the major axis (PA=$124^{\\circ}$). Contours are -2.5 (dashed), 2.5 (1.8$\\sigma$), 10, 20, 40, 60 and 100 mJy/beam. The black dots indicate the rotation curve derived by Sicking (1997). The angular and velocity resolutions ($45'', 8.2$ km ${\\rm s}^{-1}$) are indicated by the cross in the lower left corner. \\label{f: n2403panel}}}\\end{figure*} The galaxy studied here, NGC~2403, has an intermediate inclination (i=$61^{\\circ}$). The consequence is that the measured line-of-sight velocities are a combination of rotational, radial and vertical motions and the column densities are integrated along an oblique line-of-sight. Therefore, the interpretation is less straightforward than in face-on or edge-on galaxies. However, there is the advantage that information is obtained on both the vertical density structure and the vertical kinematics of the HI for the same object. For this study we have used the HI observations obtained by Sicking (1997) with the Westerbork Radio Telescope, which have about a factor of 2 better sensitivity and a higher velocity resolution than those of Wevers et al.~(1986). Fig.~1 shows the optical image, the total HI density distribution, the velocity field and the HI position-velocity map along the major axis. The latter shows that the HI line profiles at any position along the major axis are not symmetrical with respect to the rotation velocities, as they would be if they were determined by random motions only. Instead, they are systematically more extended towards the systemic velocity. This striking asymmetry is particularly obvious in the lowest contour and extends systematically over almost the whole major axis. It was already noticeable in the maps produced by Wevers et al.~(1986) and by Begeman (1987) and the puzzle presented by those early observations has motivated this study. The presence of such a `beard' is remarkable considering the size of the beam (see lower left in fig.~1) with respect to the size of the galaxy. Usually such an asymmetry is seen in edge-on galaxies or in galaxies which are not well resolved. In such cases the telescope beam `sees' not only the emission from a small area on the major axis but also larger areas away from it which have lower line-of-sight velocities and therefore cause the observed asymmetry. NGC~2403 is neither highly inclined nor poorly resolved. What is the origin of the `beard'? It is clear that it cannot simply be explained by gas moving perpendicularly away from or towards the disk, because that would produce extensions symmetric with respect to the rotation velocity. Neither can it be the result of deviations from axial symmetry or circular motion. These would affect the kinematics of the disk and produce visible effects in the velocity field, but not a low density asymmetry as observed. A likely explanation for the observed asymmetry is that not all of the HI is concentrated in a thin disk, but that part of it is in a vertically extended component. In this case, given the inclination of 61 degrees, a line-of-sight to a point on the major axis will also intercept the HI located above and below the plane which has line-of-sight velocities lower than those in the plane. This will produce a systematic broadening of the HI profiles towards the systemic velocity of the kind seen in fig.~1. It is this effect that we study here with 3-D models of the density distribution and kinematics of the HI in NGC~2403. \\begin{figure*} \\center{\\scalebox{.85}{\\includegraphics{figure2.ps}}} \\parbox{16.5cm}{\\caption{\\small Position-velocity maps (receding side, see fig.~1) along the major axis of NGC~2403 (PA=$124^{\\circ}$) of the observations (top right panel, contour levels are -2.5, 2.5 (1.8$\\sigma$), 10, 20, 40, 60 and 100 mJy/beam) and of different models. The types of models are indicated at the top of each column. The FWHM values for the thin and thick components and their velocity differences are given in the panels. The angular and velocity resolutions are $45''$ and $8.2$ km ${\\rm s}^{-1}$.}} \\label{f: lvmodels} \\end{figure*} ", "conclusions": "From section 3 it is clear that all three types of models can explain the observed asymmetry to some extent. However, the first two models require unrealistic amounts of HI above the plane. The third model has a much more realistic vertical distribution, with less than 20\\% of the gas in the thick component, and it requires that the gas above the plane rotates more slowly than the gas in the disk, by about 25 km ${\\rm s}^{-1}$. This model also appears to reproduce the observed position-velocity diagram better than the other models, in particular near the center, where the HI has a narrow and peaked distribution, and at large radii, where low level wings are visible. Note that in the observed position-velocity diagram (fig.~1) the wings near the center extend almost to the systemic velocity. Clearly, such features cannot be reproduced in models with corotation, and not even in those with a velocity decrease as high as 35 km ${\\rm s}^{-1}$. To explain these extended wings a much larger velocity decrease is needed, probably at least 50 km ${\\rm s}^{-1}$. What produces such a thick HI layer and what causes its slower rotation? Galactic fountain models (Bregman 1980; Spitzer 1990) may provide the framework for an answer. The fountain is formed by hot gas rising from the disk, its energy derived from stellar winds and supernova explosions. In the halo region the gas cools and condenses into clouds that fall back onto the disk. As the gas moves up, the centrally directed gravitational force decreases, and the gas moves outwards. Due to conservation of angular momentum the azimuthal velocity decreases. This decrease will be most pronounced in the central parts of the galaxy, where a fixed radial displacement will result in a larger velocity decrease than in the outer regions. Using thermohydrodynamic models, Struck \\& Smith (1999) have recently shown that a reduced circular velocity above and below the plane is expected to be present in turbulent disks as a result of radial motions driven by star formation activity. NGC~2403 appears to have sufficient star formation activity to drive a galactic fountain. This is for instance indicated by its large number of HII regions (Sivan et al.~1990). Four of these are exceptionally bright, comparable to the most massive starburst region in the Local Group, the 30 Doradus complex (Drissen et al.~1999). Furthermore, Thilker et al.~(1998) have found that the surface of NGC~2403 is covered by shells and fragmentary structures which are likely to have formed as a result of star formation activity. They have also found that these structures are part of a diffuse component of neutral hydrogen extending at least 400 pc from the plane. The overall picture of NGC~2403, as suggested by the modelling, is reminiscent of that of NGC~891. For this galaxy it was found (Swaters et al. 1997) that the thick component has a FWHM of about 4 kpc, and this gas appears to rotate more slowly than the gas in the disk by about 25 km ${\\rm s}^{-1}$. In the central parts the velocity decrease was found to be much larger, perhaps up to 100 km ${\\rm s}^{-1}$. In conclusion, we have presented evidence that the thin hydrogen disk of NGC~2403 is surrounded by a vertically extended layer of HI, which rotates more slowly than the disk. The observational picture is very similar to that found for NGC~891. These results suggest that a vertically extended, slowly rotating HI layer may be common among spiral galaxies, at least among those with high levels of star formation. Such extended HI layers have, however, very low surface densities and therefore very sensitive observations are needed to detect them." }, "0003/astro-ph0003353_arXiv.txt": { "abstract": "We present new, high resolution, infrared spectra of the T dwarf Gliese 229B in the $J$, $H$, and $K$ bandpasses. We analyze each of these as well as previously published spectra to determine its metallicity and the abundances of NH$_3$ and CO in terms of the surface gravity of Gl 229B, which remains poorly constrained. The metallicity increases with increasing gravity and is below the solar value unless Gl 229B is a high-gravity brown dwarf with $\\log g \\,({\\rm cgs}) \\approx 5.5$. The NH$_3$ abundance is determined from both the $H$ and the $K$ band spectra which probe two different levels in the atmosphere. We find that the abundance from the $K$ band data is well below that expected from chemical equilibrium, which we interpret as strong evidence for dynamical transport of NH$_3$ in the atmosphere. This is consistent with the previous detection of CO and provides additional constraints on the dynamics of the atmosphere of this T dwarf. ", "introduction": "Gliese 229B is not only the first brown dwarf recognized as genuine \\citep{naka95,oppen95}, but it is also the brightest and best-studied T dwarf known. With an effective temperature of $\\Teff \\sim 950\\,$K, it lies squarely between the latest L dwarfs ($\\Teff \\sim 1500\\,$K, \\citet{kirk98}) and the giant planets of the solar system ($\\Teff \\sim 100\\,$K). Indeed, its near infrared spectrum shows the strong H$_2$O absorption bands characteristic of very-low mass stars and the strong CH$_4$ bands seen in the spectra of Jupiter, Saturn and Titan. The transitional nature of the spectrum of Gl 229B is remarkable and hints at the spectral appearance of extrasolar giant planets which have effective temperatures in the range 200 -- 1600$\\,$K \\citep{guillot99}. A wealth of data on Gl 229B has accumulated since its discovery five years ago. Broad band photometry from $R$ through $N$ and an accurate parallax \\citep{matth96, golim98, legg99, hipparcos} allow an accurate determination of its bolometric luminosity. Spectroscopic observations \\citep{oppen98,geb96,schultz} covering the range from 0.8 to 5.0$\\,\\mu$m have revealed a very rapidly declining flux shortward of 1$\\,\\mu$m, the unmistakable presence of CH$_4$, H$_2$O, and Cs, and demonstrated the {\\it absence} of the CrH, FeH, VO and TiO features characteristic of late M and early L dwarfs. Finally, Noll, Geballe \\& Marley (1997) and \\citet{oppen98} have detected CO with an abundance well above the value predicted by chemical equilibrium, a phenomenon also seen in the atmosphere of Jupiter. Model spectra for Gl 229B \\citep{mar96,allard96,tsuji96b} reproduce the overall energy distribution fairly well and all agree that 1) $\\Teff \\sim 950\\,$K, 2) compared to gaseous molecular opacity, the dust opacity is small if not negligible in the infrared, 3) the gravity of Gl 229B is poorly constrained at present. The rapid decline of the flux at wavelengths shortward of 1$\\,\\mu$m is interpreted as caused by an absorbing haze of complex hydrocarbons (Griffith, Yelle \\& Marley 1998) or alternatively by the pressure-broadened red wing of the K I resonance doublet at 0.77$\\,\\mu$m (Tsuji, Ohnaka \\& Aoki 1999; Burrows, Marley \\& Sharp 1999). In this paper, we present new high-resolution spectra in the $J$, $H$, and $K$ bands. With the inclusion of the ``red'' spectrum of \\citet{oppen98}, we analyze each part of the spectrum separately to obtain independent measures of the H$_2$O abundance of Gl 229B -- broadly interpreted as the metallicity index -- to detect for the first time the presence of NH$_3$ in its spectrum, and to estimate the NH$_3$ abundance at two different depths in the atmosphere. Our results are expressed in terms of the surface gravity which cannot be determined from the data presented here. Nevertheless, we identify a reduced set of acceptable combinations of $\\Teff$ and gravity, using the observed bolometric luminosity of Gl 229B \\citep{legg99}. The observations and the near infrared spectra are presented in \\S2. Section 3 shows how an accurate parallax, a well-sampled spectral energy distribution and evolutionary models greatly reduce the possible range of combinations of $\\Teff$ and gravity without having to resort to spectrum fitting. The synthetic spectrum calculation and our method of analysis are described in \\S4. The results concerning several molecules of interest which are at least potentially detectable are presented in \\S5, followed by a discussion in \\S6. Finally, a summary of the results and directions for future study are given in \\S7. ", "conclusions": "With the availability of extensive photometric, astrometric, and spectroscopic data, our picture of the atmosphere of Gl 229B is gradually becoming more exotic and more complex. The initial discovery of CH$_4$ in its spectrum set it appart and has prompted the creation of a new spectral class, the T dwarfs. H$_2$O, CO, Cs I, and K I have also been detected. There is good evidence that the rapid decrease of the flux at visible wavelengths is caused by unprecedently broad lines of atomic alkali metals \\citep{liebert00}. The presence of condensates may also play a role in shaping the spectrum of Gl 229B. It is unfortunate that the surface gravity of Gl 229B remains poorly constrained. We have not been able to further restrict the allowed range with our new $J$, $H$, and $K$ spectroscopy. As a result, all our results are expressed as a function of gravity. This is the most significant obstacle to further progress in elucidating the astrophysics of this T dwarf. The surface gravity can probably be determined from the study of the pressure broadened shape of molecular lines. We have found good evidence for the presence of NH$_3$ in the spectrum of Gl 229B, which was expected from chemical equilibrium calculations. We have been able to determine its abundance at two different levels in the atmosphere, and we find a significant deviation from chemical equilibrium. A similar situation has been found with CO previously \\citep{noll} and this abundance pattern can be explained by vertical mixing in the atmosphere. The extent of the convection zone is not sufficient to account for the abundances we find and the mixing may be due to overshooting or to less efficient eddy diffusion. We find that NH$_3$ photolysis is not important in shaping the spectrum of Gl 229B. Because NH$_3$ can be observed in three different bands corresponding to three distinct depths in the atmosphere, an accurate determination of its abundance in each band provides information on the time scale of mixing as a function of depth. This is an unusual and powerful diagnostic tool which can provide valuable clues for modeling the vertical distributions of possible condensates. In principle, any absorber with a large abundance gradient through the visible part of the atmosphere can be used to infer the details of the mixing process. Among detected and abundant molecules, only CO and NH$_3$ satisfy this criterion. Chemical equilibrium calculations with rainout of condensates \\citep{lodders99,bms99} show that we can expect significant vertical gradients in the abundances of atomic K, Rb, Cs, and Na as they become bound in molecules (KCl, RbCl, CsCl and Na$_2$S, respectively) in the cooler, upper reaches of the atmosphere. Cesium and potassium have been detected in the spectrum of Gl 229B, and resonance doublets of K I and Na I appear to shape the visible spectrum. However, the chemical timescales for alkali metals are so short that they should always remain in thermodynamic equilibrium (Lodders 1999). Therefore, they cannot serve as probes of vertical mixing in Gl 229B. Further progress in understanding the atmosphere of Gl 229B requires better opacities for CH$_4$ and NH$_3$, and, to a lesser extent, of H$_2$O. A more accurate determination of the CO abundance from 4 -- 5$\\,\\mu$m spectroscopy is very desirable and will require higher signal-to-noise spectroscopy than is currently available. Similarly, 10$\\,\\mu$m spectroscopy to determine the NH$_3$ abundance for $P \\wig < 1\\,$bar, while difficult, is important. The issue of vertical mixing and departures from chemical equilibrium gains importance when we consider that the observed departure of CO from chemical equilibrium implies a significantly reduced CH$_4$ abundance, by conservation of the abundance of elemental carbon. Similarly, our results imply that NH$_3$ absorption in the 10$\\,\\mu$m region is reduced. Because CH$_4$ is a significant absorber in the near infrared, as is NH$_3$ in the 10$\\,\\mu$m range, departures from equilibrium must be taken into account when accurate modeling of the atmosphere and spectrum of Gl 229B is desired. This new level of complexity compounds the exoticism and the challenges posed by T dwarfs. The astrophysics of Gl 229B is far richer than has been originally anticipated. Gl 229B is currently the only T dwarf known to be in a binary system. There is no evidence that the illumination from the primary star has a significant effect on the state of its atmosphere and Gl 229B is most likely typical of isolated T dwarfs. It remains the brightest and by far the best studied of the seven T dwarfs currently known, but the list should expand to several dozens during the next 2 -- 3 years \\citep{burgasser, strauss}. The existing body of work on Gl 229B points to the most rewarding observations to conduct on T dwarfs. The possibility of studying trends in the physics of T dwarf atmospheres as a function of effective temperature is a fascinating prospect." }, "0003/astro-ph0003435_arXiv.txt": { "abstract": " ", "introduction": "The study of neutrinos coming from the Sun is a very active area of research. Results from five solar neutrino experiments are now available. These experiments measure the solar neutrino flux with different energy thresholds and using very different detection techniques. All of them, the Chlorine experiment at Homestake \\cite{chlor}, the radiochemical Gallium experiments, GALLEX \\cite{gallex} and SAGE \\cite{sage}, the water Cerenkov detectors Kamiokande \\cite{kamiok} and Super-Kamiokande \\cite{SuperK}, measure a deficit of the neutrino flux compared to the predictions of the standard solar models (SSM) \\cite{SSM,dar}. Recently it has been stated that it is impossible to construct a solar model which would reconcile all the data \\cite{hiroshi}. Moreover, a global analysis of the data of all the experiments does not leave any room for the $^7$Be neutrinos \\cite{Bah98b}. On the other hand the predictions of the SSM have been confirmed by helioseismology \\cite{basu} to a high precision. An explanation of the results of solar neutrino experiments seems to require new physics beyond the standard model of electroweak interaction. If neutrinos have non-zero masses and if they mix in analogy to the quark sector, then conversions between different neutrino flavours become possible. Flavour conversions can occur in different physical scenarios, depending on certain parameters on neutrino masses and mixing angles. One oscillation scenario makes use of the MSW-mechanism \\cite{msw85}, where the solar $\\nu_e$ transform into other neutrino flavours or into sterile neutrinos as they pass through a thin resonance region near the solar core. The other scenario assumes that the neutrinos oscillate in the vacuum between the Sun and the Earth \\cite{gla87}, which means that the oscillation length `just so` matches the Earth-Sun distance. ", "conclusions": "GENIUS could be the first detector to detect the solar pp neutrinos in real-time. Although this imposes very strong purity restrictions for all the detector components, with a liquid nitrogen shielding of 13 m in diameter and production of the Germanium detectors below ground, it should be feasible to achieve such a low background level. The advantages are the well understood detection technique (ionization in a HPGe detector), the excellent energy resolution (1 keV at 300 keV), low energy threshold (about 11 keV) and the measurement of the recoiling electrons in real-time. The good energy resolution for detecting the recoiling electrons would allow for the first time to measure the 1.3 keV predicted shift of the average energy of the beryllium neutrino line. This shift is a direct measure of the central temperature of the Sun \\cite{bah93}." }, "0003/astro-ph0003329_arXiv.txt": { "abstract": "The average near-infrared ($K$-band) luminosity of 238 {\\it Hipparcos} red clump giants is derived and then used to measure the distance to the Galactic center. These {\\it Hipparcos} red clump giants have been previously employed as $I$-band standard candles. The advantage of the $K$-band is a decreased sensitivity to reddening, and perhaps also a reduced systematic dependence on metallicity. In order to investigate the latter, and also to refer our calibration to a known metallicity zero-point, we restrict our sample of red clump calibrators to those with abundances derived from high-resolution spectroscopic data. The mean metallicity of the sample is [Fe/H]~=~$-0.18$~dex ($\\sigma = 0.17$~dex). The data are consistent with no correlation between $M_{K}$ and [Fe/H], and only weakly constrain the slope of this relation. The luminosity function of the sample peaks at $M_{K} = -1.61 \\pm 0.03$~mag. Next, we assemble published optical and near-infrared photometry for $\\sim$ 20 red clump giants in a Baade's Window field with a mean metallicity of [Fe/H]~=~$-0.17 \\pm 0.09$~dex, which is nearly identical to that of the {\\it Hipparcos} red clump. Assuming that the average $(V-I)_{0}$ and $(V-K)_{0}$ colors of these two red clumps are the same, the extinctions in the Baade's Window field are found to be $A_{V}$/$A_{I}$/$A_{K}$ = 1.56/0.87/0.15, in agreement with previous estimates. We derive the distance to the Galactic center: $(m-M)_{0} = 14.58 \\pm 0.11$~mag, or $R = 8.24 \\pm 0.42$~kpc. The uncertainty in this distance measurement is dominated by the small number of Baade's Window red clump giants examined here. ", "introduction": "The horizontal branch ``red clump'' is a landmark feature in color-magnitude diagrams of intermediate-age ($\\sim$1-10 Gyr) or very old ($\\simgt$10 Gyr) but metal-rich stellar populations. It has been suggested that red clump giants are standard candles, like their very old and metal-poor cousins, the RR~Lyrae variable stars (Cannon 1970). Red clump giants have recently gained popularity as standard candles because the {\\it Hipparcos} color-magnitude diagram of the solar neighborhood displays a very prominent red clump (Perryman et al.~1997). Paczy\\'{n}ski and Stanek (1998) measured the distance to the Galactic center using the {\\it Hipparcos} $I$-band calibration of the red clump luminosity. Distance estimates to the Large Magellanic Cloud (LMC) and M31 soon followed (Udalski et al.~1998; Stanek et al.~1998; Stanek \\& Garnavich 1998). At first, the {\\it Hipparcos} red clump yielded a ``short'' distance to the LMC, consistent with earlier measurements based on the red clump (Butcher 1977; Seidel et al.~1987). However, Romaniello et al.~(1999) have since obtained a long LMC distance also using the red clump. Extant distance measurements made with the LMC red clump do not agree. Recent theoretical and observational studies of the effects of mean population age and metallicity on the red clump luminosity have been made by Cole (1998), Girardi et al.~(1998), Udalski (1998, 1998b, 1999), Alves and Sarajedini (1999), and others. Earlier studies include Lattanzio (1986), Seidel et al.~(1987b), and Olszewski et al.~(1987). Variations in mean population age and metallicity are theoretically predicted to change the red clump luminosity at the $\\sim$20\\% level. Finally, issues of extinction that potentially affect recent distance measurements made with the red clump have been discussed by Romaniello et al.~(1999), Zaritsky~(1999), and Popowski~(2000). In the spirit of McGonegal et al.~(1982), who first discussed the advantages of the near-infrared wavelength regime for distance scale work with Cepheid variable stars, we assemble $K$-band photometry data for 238 {\\it Hipparcos} red clump giants from the literature, derive a $K$-band luminosity calibration, and apply it to published $K$-band photometry of Baade's Window in order to measure the distance to the Galactic center. The advantage of our $K$-band calibration compared to the Paczy\\'{n}ski and Stanek (1998) $I$-band calibration is a decreased sensitivity to reddening (e.g.~Cardelli et al.~1989), and perhaps also a reduced systematic dependence on metallicity. In order to investigate the latter, and also to refer our calibration to a known metallicity zero-point, we restrict our sample of {\\it Hipparcos} red clump calibrators to those with abundances derived from high-resolution spectroscopic data (McWilliams 1990). Udalski (1999) made an important analysis of the $M_{I}$--[Fe/H] correlation with these same data from McWilliams (1990). ", "conclusions": "{\\it Hipparcos} measured accurate parallaxes for several hundred red clump giants in the solar neighborhood. These stars may be useful standard candles. We have assembled $K$-band photometry data from the literature for 238 of these {\\it Hipparcos} red clump giants and calculated the mean absolute magnitude, $M_{K} = -1.61 \\pm 0.03$~mag. The advantage of our $K$-band calibration compared to the previously employed $I$-band calibration is a decreased sensitivity to reddening, and perhaps also a reduced systematic dependence on metallicity. Using published $K$-band photometry of Baade's Window, we derive the distance to the Galactic center, $R = 8.24 \\pm 0.42$ kpc. The error in this distance measurement is dominated by the small number of Baade's Window red clump giants examined here. Suggestions for future work include refining the distance to the Galactic center by using a larger sample of red clump giants with known metallicities, reddenings, and $K$-band photometry. Our $K$-band red clump calibration may also be used to determine the distance to the LMC, which currently serves as the zero-point for the extragalactic distance scale." }, "0003/astro-ph0003059_arXiv.txt": { "abstract": "The separation $\\del$ between two peaks in the gamma-ray pulse profile is calculated as a function of energy for several polar cap models with curvature-radiation-induced cascades. The Monte Carlo results are interpreted with the help of analytical approximations and discussed in view of the recent data analysis for the Vela pulsar \\cite{kanbach}. We find that the behaviour of $\\del$ as a function of photon energy $\\varepsilon$ depends primarily on local values of the magnetic field, $B_{\\rm local}$, in the region where electromagnetic cascades develop. For low values of $B_{\\rm local}$ ($< 10^{12}$ G), $\\del(\\varepsilon)$ is kept constant. However, for stronger magnetic fields ($\\ga 10^{12}$ G) in the hollow-column model $\\del$ decreases with increasing photon energy at a rate dependent on maximum energy of beam particles as well as on viewing geometry. There exists a critical photon energy $\\et$ above which the relation $\\del(\\varepsilon)$ changes drastically: for $\\varepsilon > \\et$, in hollow-column models the separation $\\del$ increases (whereas in filled-column model it decreases) rapidly with increasing $\\varepsilon$, at a rate of $\\sim 0.28$ of the total phase per decade of photon energy. The existence of critical energy $\\et$ is a direct consequence of one-photon magnetic absorption effects. In general, $\\et$ is located close to the high-energy cutoff of the spectrum, thus photon statistics at $\\et$ should be very low. That will make difficult to verify an existence of $\\et$ in real gamma-ray pulsars. Spectral properties of the Vela pulsar would favour those models which use low values of magnetic field in the emission region ($B_{\\rm local} \\simless 10^{11}$ G) which in turn implies a constant value of the predicted $\\del$ within \\egret\\ range. ", "introduction": "Two prominent peaks are a characteristic feature of gamma-ray pulse shapes in the three brightest out of seven gamma-ray pulsars detected so far: Crab (PSR B0531+21), Vela (PSR B0833-45), and Geminga (J0633+1746). Phase separation between the two peaks is very large in each case, in the range between 0.4 and 0.5 (e.g. Fierro, Michelson \\& Nolan 1998). The separation, which we denote by $\\Delta^{\\rm peak}$, was determined with photons from the entire energy range of \\egret. Kanbach \\shortcite{kanbach} suggested that the separation $\\Delta^{\\rm peak}$ in Vela might be energy dependent. The effect would be of the order of a few percent or less. The plot of the phase separation against energy (fig.2 of Kanbach 1999) shows that $\\Delta^{\\rm peak}$ decreases by about $5\\%$ over 20 energy intervals covering the range between $\\sim 50\\MeV$ and $\\sim 9\\GeV$. The scatter of points is, however, large enough for this result still to be consistent with the separation staying at a constant level of $0.43$, especially when one rejects two energy intervals: of the lowest and the highest value. Such effects as suggested by Kanbach can be justified qualitatively, at least within polar cap scenarios. Their origin may be different at different energy ranges, and their magnitude may vary as well. For example, Miyazaki \\& Takahara \\shortcite{miyazaki} found dramatic changes in peak-to-peak phase separation due to magnetic absorption effects in their attempts to model the Crab pulse shapes. Their numerical calculations were performed with low photon energy resolution for a model with homogeneous polar cap, and instant acceleration. This new aspect of studying the HE properties of pulsars is potentially attractive. The problem of poor photon statistics should become less essential with future high-sensitivity missions like \\glast. Then any well established empirical relation between the phase separation $\\Delta^{\\rm peak}$ and the photon energy $\\varepsilon$ (including $\\Delta^{\\rm peak} = const$) may help to discriminate in favour of some particular models of pulsar activity. In this context we present a model of the peak-to-peak phase separation in the gamma-ray lightcurve of Vela and confront it with the results of Kanbach \\shortcite{kanbach}. Our aim is to present properties of $\\Delta^{\\rm peak}(\\varepsilon)$ predicted by the polar cap model with curvature (CR) and synchrotron (SR) radiation being dominant emission mechanisms. This is an extension of the preliminary results of Dyks, Rudak \\& Bulik \\shortcite{bulik}. In section 2 we outline the model and introduce the input parameters for which Monte Carlo simulations were performed. Section~3 describes the numerical results and offers their interpretation; conclusions follow in section 4. ", "conclusions": "In this paper we addressed a recent suggestion of Kanbach \\shortcite{kanbach} that peak separation $\\del$ in the double-peak gamma-ray pulses of the Vela pulsar may monotonically decrease with increasing photon energy at a rate $\\sim 0.025$ phase per decade in energy over the range $50\\MeV$ to $9\\GeV$, We calculated gamma-ray pulses expected in polar-cap models with magnetospheric activity induced by curvature radiation of beam particles. Two types of geometry of magnetospheric column above the polar cap were assumed: a hollow-column associated with an outer rim of the polar cap and a filled column associated with a uniform polar cap. Four models were considered with two scenarios for the acceleration of beam particles. Pulsed emission in the models was a superposition of curvature radiation due to beam particles and synchrotron radiation due to secondary $e^\\pm$ pairs in magnetospheric cascades. The changes in the peak separation were investigated with Monte Carlo numerical simulations. We found that regardless the differences in the models, the peak separation $\\del$ below a few GeV, where the emission is dominated by synchrotron component, is either a weak decreasing function of photon energy $\\varepsilon$, or remains constant. Both variants may be considered to be in agreement with the results of Kanbach \\shortcite{kanbach} for the latter are affected by large statistical errors. A particular behaviour of $\\del$ depends on a combination of several factors, including strength of magnetic field in the region of pair formation and model of electron acceleration (both of which determine spectral and directional properties of the radiation at different altitudes), as well as viewing geometry. Essentially, in strong fields, $B_{\\rm local} \\ga 10^{12}\\G$, $\\del$ decreases with increasing photon energy $\\varepsilon$, whereas for $B_{\\rm local} < 10^{12}\\G$, the peak separation $\\del$ stays at a constant level. Moreover, we found that due to the magnetic absorption ($\\gamma \\bld B \\rightarrow e^\\pm$) there exists a critical energy $\\et$ at which the peak separation $\\del$ makes an abrupt turn and then changes dramatically for $\\varepsilon > \\et$. It increases in the hollow-column models (A, B, and C) and decreases in the filled-column model (D), at a rate $\\sim 0.28$ phase per decade of photon energy. The numerical behaviour of $\\Delta^{\\rm peak}$ in this regime in the hollow-column models was easily reproduced to high accuracy with a simple analytical model of magnetospheric transparency for a photon of energy $\\varepsilon$, and its momentum tangential to local dipolar magnetic field line at a site of its origin. An exact value of $\\varepsilon_{\\rm turn}$ is model-dependent but it is confined to a range between $\\sim 0.9\\GeV$ and $\\sim 4.5\\GeV$. To find such a hypothetical turnover of $\\Delta^{\\rm peak}$ in real observational data would require, however, high-sensitivity detectors, since for $\\varepsilon > \\varepsilon_{\\rm turn}$ the expected flux of gamma-rays drops significantly. If detected, this turnover would be an important signature of polar cap activity in gamma-ray pulsars. It would support the notion that high-energy cutoffs in gamma-ray spectra of pulsars are due to magnetic absorption. The CR-induced cascades models, like those considered in this work, are not the only possibility for nearly aligned rotators to produce double-peak pulses with large phase separations. There exists an alternative class of models - with pair cascades above polar cap induced by magnetic inverse Compton scatterings (ICS) of primary electrons in the field of soft photons from the stellar surface - proposed in a series of papers (e.g. Sturner \\& Dermer 1994, Sturner et al. 1995). In particular, Sturner et al. (1995) present the detailed Monte Carlo model spectra of the Vela pulsar. They also present pulse profiles at a fixed energy of 50 MeV (for several viewing angles) but no word of comment is given regarding the problem of $\\Delta^{\\rm peak}$ versus photon energy. We expect the outcome to be qualitatively similar to our results. First, the scatterings take place mostly within a very limited height above the polar cap surface (below $h \\sim R_{\\rm pc}$) and the preferred directions of propagation of the ICS photons will be fixed by magnetic field lines just above the surface. Therefore, $\\Delta^{\\rm peak}$ due solely to ICS photons should stay constant for a wide range of energy. Inclusion of synchrotron photons due to pairs is unlikely to notably affect $\\Delta^{\\rm peak}$ unless the pair formation front is vertically more extended than for CR-induced cascades. Second, some turnover point at $\\varepsilon_{\\rm turn}$ not exceeding 1~GeV should be present due to magnetic absorption. The behaviour of $\\del$ for $\\varepsilon > \\et$ should roughly follow the dashed lines in Fig. 2 (upper panel) and Fig.3 as long as the assumption about photons (which are to be absorbed) propagating tangentially to local dipolar magnetic field line at their site of origin remains valid for majority of ICS photons.\\hfill\\break To verify this qualitative picture would, however, require detailed numerical calculations." }, "0003/astro-ph0003084_arXiv.txt": { "abstract": "The brightest, or first-ranked, galaxies (BCGs) in rich clusters show a very small dispersion in luminosity, making them excellent standard candles. This small dispersion raises questions about the nature of BCGs. Are they simply the extremes of normal galaxies formed via a stochastic process, or do they belong to a special class of atypical objects? If they do, are all BCGs special, or do normal galaxies compete for the first rank? To answer these questions, we undertake a statistical study of BCG magnitudes using results from extreme value theory. Two-population models do better than do one-population models. A simple model where a random boost in the magnitude of a fraction of bright normal galaxies forms a class of atypical galaxies best describes the observed distribution of BCG magnitudes. ", "introduction": "Among the most luminous bodies in the universe are the brightest, or first-ranked, galaxies in rich clusters. These galaxies have absolute magnitudes between -21.5 and -23.3 and are among the farthest observable objects. In addition, the magnitudes of these brightest cluster galaxies (BCGs) are highly uniform, with a dispersion of 0.32 magnitudes (Hoessel \\& Schneider 1985). Their uniformity and large luminosity make BCGs excellent standard candles. The uniformity of BCG magnitudes raises a particularly important question regarding their nature (Peebles 1968; Sandage 1972). Are BCGs simply the brightest of a statistical set of galaxies or do they belong to a special class of objects? If a special class of galaxies exists, do all clusters have special galaxies and are they always first-ranked (Bhavsar 1989)? We investigate these questions using extreme value theory (Fisher \\& Tippett 1928). ", "conclusions": "For more than thirty years, cosmologists have debated the nature of the magnitude-distribution of brightest cluster galaxies. Peebles (1968) and Sandage (1972, 1976) \\& Peach (1969) reach markedly different conclusions. More recently, Bhavsar (1989) and Postman \\& Lauer (1995) differ regarding the population(s) that comprise the first-ranked galaxies. In light of this controversy, we have conducted a new examination of the distribution of BCG magnitudes. We consider the BCGs as $a$ $class$ $of$ $objects$ to which we may apply well established results from extreme value theory. We find that there are a number of models that perform well in describing the HGT and LP data sets. Though a Gaussian fits both data sets, the confidence limits warrant further investigation of two-population models. Tables 2 \\& 3 clearly show that we should reject the Gumbel (model B) as a fit, i.e., the hypothesis that all BCGs are statistical extremes. The Gaussian (model A) is marginally acceptable but without physical basis. Two-population models, in particular, the three combinations of $f_{G}$ and $f_{g}$, describe the data very well. Tables 2 \\& 3 show their relative merits. Model E stands out as giving the best overall fit and is motivated by a physical basis. Therefore, it is most likely that there are two populations of BCGs: the extremes of a normal population and a class of atypical galaxies with a brighter average mean. \\\\\\\\ We thank Marc Postman for sending us the LP data. This research was supported by an ANN grant from the US Department of Education and the Kentucky Space Grant Consortium." }, "0003/astro-ph0003421_arXiv.txt": { "abstract": "We present a mechanism related to the migration of giant protoplanets embedded in a protoplanetary disc whereby a giant protoplanet is caught up, before having migrated all the way to the central star, by a lighter outer giant protoplanet. This outer protoplanet \\add{may} get captured into the 2:3 resonance with the more massive one, \\delete{and}\\add{in which case} the gaps that the two planets open in the disc overlap. Two effects arise, namely a \\add{squared} mass weighted torque imbalance and an increased mass flow through the overlapping gaps from the outer disc to the inner disc, which both play in favour of an outwards migration. Indeed under the conditions presented here, which describe the evolution of a pair of protoplanets respectively Jupiter and Saturn sized, the migration is reversed, while the planets semi-major axis ratio is constant and the eccentricities are confined to small values by the disc material. The long-term behaviour of the system is briefly discussed, and could account for the high eccentricities observed for the extrasolar planets with semi-major axis $a>0.2$~AU. ", "introduction": "In the past few years a number of extrasolar giant planets have been discovered around nearby solar--type stars. These objects masses range from $0.17$~\\MJ~to $11$~\\MJ~(where \\MJ~is Jupiter's mass) and their orbital semi-major axis range from $0.038$~AU to $3.3$~AU (Marcy, Cochran \\& Mayor, 1999). Although many uncertainties remain about planet formation, it is now commonly accepted that planets have formed in and from protoplanetary discs. Necessarily, there must be some time interval over which a giant planet and the surrounding gaseous disc material coexist. The planet and the disc exchange angular momentum through tidal interactions which generally make the planet lose angular momentum This mechanism is called migration. It can roughly be divided in two regimes: \\begin{itemize} \\item If the planet mass is small enough, the disc response is linear. The migration rate, in that regime, is proportional to the planet \\add{and disc} mass\\add{es}, is independent of the viscosity and weakly dependent of the disc surface density and temperature profiles. This is the so-called type~I migration (Ward, 1997). \\item When the protoplanet mass is above a certain threshold, the torques acting locally on the surrounding disc material open a gap (Papaloizou \\& Lin, 1984), whose width and depth are controlled by the balance between the tidal torques, which tend to open the gap, and the viscous torques which tend to close it. The disc response is \\delete{markedly} \\add{significantly} non-linear, and most of the protoplanet Lindblad resonances fall in the gap and therefore cannot contribute to the planet-disc angular momentum exchange. The migration rate slows down dramatically compared to type~I migration. Furthermore, the tidal truncation splits the disc in\\add{to} two parts and the planet is locked \\delete{in} \\add{to} the disc viscous evolution (Nelson et al. 2000). This is the type~II migration, \\add{which describes the orbital evolution of giant protoplanets.} \\end{itemize} \\delete{Type~II migration occurs on the disc viscous time-scale, about $ 10^5$~yr at 5~AU. This time-scale is still one or two orders of magnitude shorter than the disc estimated lifetime. Unless the planet has formed much further out in the disc, in which case one has to face time-scale problems again, or unless the disc is cleared by some process before the planet has migrated all the way to the central star, there is clearly a conflict between type~II migration theoretical predictions and observational facts. Indeed, roughly speaking only half of the detected extrasolar planets are ``hot Jupiters'', which orbit very close to the central star ($a<0.2$~AU). There is another problem with the type~II migration of one giant protoplanet. All the extrasolar planets with a semi-major axis larger than $0.2$~AU have large eccentricities ($e>0.16$, except 47~UMa for which $e=0.096$, which is already twice as large as Jupiter's eccentricity). Now the migration of a giant protoplanet embedded in a disc occurs on a nearly circular orbit, since the eccentricity is damped by the coorbital material (Ward, 1988).} In this letter we consider the coupled evolution of a system of giant protoplanets consisting of two non-accreting cores with masses $1$~\\MJ~and $0.29$~\\MJ, which we are going to call from now on respectively ``Jupiter'' and ``Saturn''. Attempts have already been made to describe the behaviour of a system of planets embedded in a disc. Melita \\& Woolfson (1996) and Haghighipour (1999) considered an embedded Jupiter and Saturn system orbiting a solar mass star, and showed how resonance trapping would affect their evolution. However the dissipative force in these works was due to the dynamical friction with a uniform density interplanetary medium, hence type~II migration effects were not taken into account. Resonance trapping of planetesimals by a fixed orbit Jupiter sized protoplanet has also been investigated by Beaug\\'e et al. (1994), and shown to be able to build up a single planetary core with orbital characteristics close to Saturn's ones. Kley (2000) studied the orbital evolution of two maximally accreting giant cores embedded in a minimal mass protosolar disc, and showed how the migration of the inner core could be halted by the presence of the outer one, and how the eccentricity of the inner core is pumped up by the outer one.\\delete{and hence could account for observations of high eccentricity extrasolar planets.} ", "conclusions": "\\add{We have performed a series of restart runs (see section~\\ref{sec:restart}) in order to check for a variety of behaviours. \\subsection{Differential Lindblad torque sign} The one sided Lindblad torque has been shown to be proportional to $h'^{-3}$ (Ward 1997). We have performed two restart runs ($h'=0.04\\rightarrow 0.03$ and $h'=0.04\\rightarrow 0.05$) in order to check that the migration rate variation is consistent with this dependence. This is indeed the case. We note in passing that the migration rate varies as $h'^{-3}$, and not as $h'^{-2}$ as it would be the case in a one planet problem, since the Outer/Inner Lindblad torque asymmetry does not vanish as the disk thickness tends to zero (the OLRs would pile-up at Saturn's orbit, whereas the ILRs would pile-up at Jupiter's orbit). These results confirm that the behaviour we observe occurs mainly due to the differential Lindblad torque and shows as well that this latter quantity is positive, as expected from Eq.~(\\ref{eq:raplim}).} \\subsection{$\\alpha$-viscosity vs. uniform viscosity} So far we have only considered a uniform viscosity. Switching to a uniform-$\\alpha$ viscosity of the form $\\nu=\\alpha c_sH$ \\delete{would} makes $\\nu$ scale here as $r^{1/2}$, so the viscosity at the outer edge of the common gap \\delete{would be}\\add{is} higher, whereas it \\add{is} \\delete{would be} smaller in the inner disc. This \\delete{would} has \\delete{two}\\add{the following} effect, which \\delete{would both} plays in favour of enhancing the migration reversal mechanism: \\delete{(i) The mass flow through the gap, which is thought to be linked to the viscosity, is likely to increase as the viscous stress increases (since the tidal truncation is weaker and more material is allowed to penetrate the gap on Saturn's side);} the viscous time-scale of the inner disc \\delete{would be}\\add{is} higher and therefore its surface density \\delete{would} increases \\delete{as well}\\add{accordingly}, since the material brought through the gap \\delete{would} piles-up in the inner disc for a longer time before being accreted on the primary. \\add{This has been checked with a restart run}. \\subsection{Accretion on to the planets} The cores considered above do not accrete gas from the disc. One can wonder what would be the effects of accretion. \\delete{Part of the material flowing inwards from the outer disc would eventually be accreted on the planets and therefore would lose less angular momentum than the amount it would lose if it went all the way to the inner disc, and if the accreted fraction was important this would weaken the main source of the migration reversal mechanism. In particular, only a tiny fraction of the gap penetrating material should be accreted on Saturn. Accretion on to Jupiter is less constraining, since if the inwards flowing material was accreted on to Jupiter one would have to cancel $R_H^j$ in Eq.~(\\ref{eq:dyn}), and the value of $\\overdot J_{\\rm dyn}$ would not change dramatically. In the runs presented here, the mass flow through the gap amounts to about $2.5\\cdot 10^{-3}$ \\MJ.orbit$^{-1}$, which corresponds roughly to $7.5\\cdot 10^{-2}$ ${\\rm M}_\\oplus.{\\rm yr}^{-1}$. Kley (1999), using a maximally accreting scheme to describe the growth of a protoplanet, found an accretion rate about $10^{-2}\\;{\\rm M}_\\oplus.{\\rm yr}^{-1}$ for a one Jupiter mass protoplanet embedded in a tidally truncated disc with the same viscosity as the one used here. This accretion rate is maximal and corresponds roughly to a Bondi rate, or a half emptying time of the Roche lobe of about $\\tau_{1/2}^{min}\\sim 1/4$ of the orbital time. Now the material flowing into the Roche lobe does have angular momentum in the non-rotating frame centered on Jupiter, and needs to get rid of it before being accreted. An upper limit for the circumjovian disc time-scale is $\\tau^j_v\\sim (R_H^j)^2/3\\nu\\sim25$~orbits, about one hundred times larger than $\\tau_{1/2}^{min}$. A realistic value of $\\tau_{1/2}$ is probably between these two extreme values, is not known at the present time, and probably deserves a full self-consistent 3D treatment to be properly evaluated. Lubow, Seibert and Artymowicz (1999), using high resolution numerical 2D simulations, give an estimate for $\\overdot M_j$ under similar conditions which agree reasonably well with Kley's values. They also claim that the mass flow reaching the inner disc is about twice as large as the mass accreted on to Jupiter. From the considerations above, it is reasonable to expect that most of the inwards flow does reach Jupiter orbit and that a sizable amount reaches the inner disc. } \\add{We have performed a number of restart runs in order to investigate the effect of accretion on the mechanism presented here. We have only considered accretion on to Jupiter, as it is likely that the accretion rate on Saturn can be regarded as being negligible (i.e. its mass doubling time is much longer than the timescale of the outwards migration, see e.g. Pollack et al. 1996). The prescription we used to model accretion on to Jupiter consists in removing a proportion of the material which lies in the inner Roche lobe (i.e. a sphere with a radius of half the Hill radius). The amount which is removed in one timestep is calculated from the half emptying time of the inner Roche lobe $\\tau_{1/2}$. We have performed four different restart runs, corresponding to the following values of $\\tau_{1/2}$: $\\tau_{1/2}=T_0$ (maximally accreting core, see Kley 1999), $\\tau_{1/2}=3T_0$, $\\tau_{1/2}=10T_0$ and $\\tau_{1/2}=30T_0$, where $T_0=2\\pi/\\Omega_J$ is Jupiter's orbital time. In each of these cases, turning on accretion had no impact on the system migration rate, at least in the early stages: in the first case, the mass doubling time for Jupiter is relatively short, and when Jupiter's mass is significantly larger than its initial mass some additional effects, which will be presented in much greater detail elsewhere, affect the migration rate which then differ from the non-accreting case. } \\subsection{\\delete{Grid resolution and }Smoothing} \\delete{The mechanism presented above relies on an increased mass flow through the overlapping gaps. No analytical theory, at the present date, is able to predict the flux of mass going from the outer disc to the inner disc, and one can wonder whether the observed flux is not linked to numerical aspects, in particular the finite zone size. To answer this question, we performed a test run with twice as high a resolution as the geometric radial spacing run, i.e. with $N_r=244$ and $N_\\theta=600$, all the other parameters being unchanged. No difference has been found on the mass flow through the gap. } The smoothing parameter of the potential can have a dramatic impact on \\delete{the initial} Saturn's \\add{initial} migration rate. \\delete{Indeed} This rate is controlled by a subtle balance between outer disc and inner disc torques. In the case of Saturn, all the Lindblad resonances play a role, since there is no gap. Many prescriptions for the smoothing are unable to give trustworthy results for the balance between the outer and inner torques since, depending on the prescription, these two quantities are affected in a different way. On the other hand Jupiter's migration rate is much more robust, since the presence of the gap prevents high-$m$ Lindblad resonances \\delete{to} play\\add{ing} a role in the migration, which is \\add{therefore} controlled only by remote, low $m$ resonances and \\delete{therefore}\\add{thus} almost insensitive to the smoothing parameter. For this reason we have adopted an approach which \\delete{consists in}\\add{involves} choosing a smoothing prescription which endows Saturn with a migration velocity of the order of magnitude of the linear analytical predictions (type~I migration), which is needed to give correct results for the capture in\\add{to} resonance. Once Saturn is trapped into resonance with Jupiter, it is dynamically slaved by \\delete{this}\\add{the} latter and the system evolution is \\add{only} very weakly affected by the exact value of the outer disc torque exerted on Saturn. We have found that using \\add{either of} the \\add{two} prescription\\add{s} below satisfactorily preserves the analytical torque imbalance on Saturn and therefore gives it a type~I migration rate: \\begin{itemize} \\item The potential of a planet acting on the disc is smoothed over the length $\\varepsilon=0.4R_H$ where $R_H$ is the Hill radius of the planet under consideration, whereas the potential of the disc acting back on the planet is smoothed over $\\varepsilon'=\\sqrt{H^2+d^2}$ where $H$ and $d$ are respectively the local disc thickness and zone diagonal. Since $\\varepsilon'\\ne\\varepsilon$ the action-reaction law is not fulfilled and the numerical biases which arise favour an inwards migration, as can be easily checked. \\item The potential of a planet acting on the disc and the potential of the disc acting on the same planet are smoothed over $\\varepsilon = 0.4R_H$. This prescription does fulfill the action-reaction law. \\end{itemize} In \\add{both} these two cases, as in any other which gives Saturn a type~I migration rate, including runs performed with a uniform radial spacing, the migration gets reversed. The run presented here corresponds to the first prescription. \\subsection{Impact of mass ratio and Long-term behaviour} One can wonder \\delete{on which}\\add{about the size of the} interval of ``Saturn'''s mass \\add{which causes} the migration \\delete{can}\\add{to} be significantly slowed down or reversed. \\delete{Indeed} If ``Saturn'' is not massive enough it will not significantly affect Jupiter's evolution (\\add{the common ``gap'' will be too full on Saturn's side, and therefore Jupiter's OLR torques will not be shut off}), whereas if it is too massive, the torque imbalance will \\add{be negative again}.\\delete{strongly play in favour of an inwards migration and the steepest tidal truncation at the common gap outer edge is likely to reduce the inwards mass flow.} Work is in progress to accurately determine which range of parameters leads to a migration reversal. \\add{It should be noted that the results presented here depend on the artificial initial conditions. We have performed other runs in which Saturn is initially very close either to the $1:2$ or $3:5$ resonance, and it turns out that neither of these resonances is able to struggle against the strong Lindblad torques on Saturn: no resonance angle can be found which provides a resonant torque on Saturn which counteracts the tide. Therefore a trapping into the $2:3$ resonance is the most likely outcome when the system is still embedded in a massive disc, whatever the initial conditions: catching-up of ``Saturn'' or {\\em in-situ} assembling from smaller, type~I migrating bodies. } \\delete{, and to properly understand the physics of the high inwards mass flow (funnel effect due to the weak tidal truncation of the outer edge, or a possible link with the stochastic behaviour of a test-particle which arises in the coorbital region of a pair of resonantly orbiting massive planets). } The long-term behaviour of the system is twofold: \\begin{itemize} \\item \\delete{It should be noted that the outwards migration is a slowly accelerating process (the torques scale as $a$ and the inwards mass flow increases as well for geometrical reasons) which requires the gap outer edge to be able to follow the migration rate $\\overdot a/a$. Now the torque Saturn exerts on the outer disc scales as $a$, whereas the angular momentum content of the outer disc scales as $a^{5/2}$, so for a sufficiently high value of $a$, the gap outer edge will not be able to follow the global outwards motion, and a runaway inwards mass flow can result. The most important consequence of this runaway flow is that Saturn will leave the 2:3 resonance with Jupiter and will move further in the disc. The details and consequences of this mechanism will be presented in a forthcoming paper.} \\add{The system is locked in\\add{to} resonance as long as~: \\begin{itemize} \\item The two-planet system can adjust its resonance angle in order to prevent the planets being ``pushed'' towards each other by the Lindblad torques exerted by the disk on each of them. In all our runs we have never observed this behaviour. Now, given the small eccentricities involved here, and given the fact that the adiabatic criterion threshold increases as $j(j+1)$, the most probable outcome is that Saturn would then be captured in the next order resonance, that is to say $3:4$, and all the physics exposed in this paper would still be valid (presence of a common gap, sharing of the coorbital material by the two planets, mass-weighted torque imbalance, etc.) \\item The planets are not pulled apart by any other torques. Now we have mentioned the possibly important role of the coorbital corotation torque in this problem, which may be sufficient to move the planets apart at some stage, in which case we may ultimately get a low eccentricity double giant planet system when the disc disappears. This will be presented in greater detail elsewhere. \\end{itemize} } \\item If the planets happen to be \\add{locked} in\\add{to} resonance at the time \\add{that} the gas effects become negligible, then the system is likely to be unstable (we mentioned already that at least two angles librate simultaneously, which strongly suggests a possible chaotic behaviour; see also Kley 2000), and the most likely outcome is that one planet will be ejected whereas the other \\delete{one}\\add{planet} will end up on an eccentric orbit. This could account for the observed eccentricities of the extrasolar planets which are not orbiting close to their host star, i.e. which have not migrated all the way to the star. \\end{itemize}" }, "0003/nucl-th0003029_arXiv.txt": { "abstract": "We calculate the longitudinal response of ${}^{208}$Pb using a relativistic random-phase approximation to three different parameterizations of the Walecka model with scalar self-interactions. From a nonspectral calculation of the response---that automatically includes the mixing between positive- and negative-energy states---we extract the distribution of strength for the isoscalar monopole, dipole, and high-energy octupole resonances. We employ a consistent formalism that uses the same interaction in the calculation of the ground state as in the calculation of the response. As a result, the conservation of the vector current is strictly maintained throughout the calculation. Further, at small momentum transfers the spurious dipole strength---associated with the uniform translation of the center-of-mass---gets shifted to zero excitation energy and is cleanly separated from the sole remaining physical fragment located at an excitation energy of about 24~MeV; no additional dipole strength is observed. The best description of the collective modes is obtained using a ``soft'' parameterization having a compression modulus of $K\\!=\\!224$~MeV. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003171_arXiv.txt": { "abstract": "Recently, eigenmodes of rotating fluids, namely inertial modes, have received much attention in relation to their destabilization when coupled to gravitational radiation within neutron stars. However, these modes have been known for a long time by fluid dynamicists. We give a short account of their history and review our present understanding of their properties. Considering the case of a spherical container, we then give the exact solution of the boundary (Ekman) layer flow associated with inertial r-modes and show that previous estimations all underestimated the dissipation by these layers. We also show that the presence of an inner core has little influence on this dissipation. As a conclusion, we compute the window of instability in the temperature/rotation plane for a crusted neutron star when it is modeled by an incompressible fluid. ", "introduction": "Recently much work has been devoted to the study of the rotational instability of neutron stars resulting from a coupling between gravitational radiation and the so-called ``r-modes\" of a rotating star \\cite{anderss98,FM98,LOM98,KS99}. Such an instability may indeed play a key role in the distribution of rotation periods of neutron stars as well as it may be an important source of gravitational radiation. In this paper, we shall first clarify a point of history concerning ``r-modes\" which are in fact a special class of inertial modes; we shall then review their singular properties which have been clarified only very recently in \\cite{RV97}, \\cite{RGV99} and \\cite{RGV00b}. The last section will present the analytical derivation of the damping rate of inertial r-modes in a neutron star with a crust and/or a core through the boundary layer analysis within the framework of newtonian theory. We conclude on the stability of crusted neutron stars when modeled by an incompressible viscous fluid in a rotating sphere. ", "conclusions": "Let us now apply these results to the case of rapidly rotating neutron stars. We take the viscosity from \\cite{BU00}, $\\nu=1.8\\,f/T^2_8$ m$^2$/s where $f$ is a dimensionless parameter taking into account the different transport mechanisms in the fluid (superfluid phases for instance) and $T_8$ is the temperature in 10$^8$K unit. Using a radius of 12.53~km and an angular frequency of $2\\pi\\times$1kHz, we find an Ekman number $\\sim 10^{-12}$ which is indeed very small and thus boundary layer theory applies. \\begin{figure}[t] \\centerline{\\psfig{figure=./critical.ps,width=8cm}} \\caption[]{Curves of critical angular velocity, normalized by $\\Omega_K=\\frac{2}{3}\\sqrt{\\pi G\\overline{\\rho}}$, for different models. The solid line shows the result of the present work, the dashed-dotted one is that of \\cite{AJKS00} and the dashed one is for \\cite{BU00}. The dotted line is the critical curve for a non-crusted star. No core has been included ($\\eta=0$).} \\label{crit} \\end{figure} We may now estimate the charateristic time scale for the damping of the $m=2$-mode. We find \\beq T_{d} = 26.7 {\\rm s}\\; \\frac{T_8}{\\sqrt{f}}\\lp\\frac{R}{10\\;\\rm km}\\rp \\lp\\frac{1\\;\\rm kHz}{\\nu_s}\\rp^{1/2} \\eeq which is a somewhat smaller value than the previous estimate of \\cite{BU00} and \\cite{AJKS00} who find a characteristic time of 100~s and 200~s respectively. Our disagreement with these authors comes from their approximate evaluation of the boundary layer dissipation and from the resulting functional dependence with respect to mass and density. Let us first evaluate the damping rate according to \\cite{LL7189}; it turns out that \\beq 2\\gamma = -\\lp\\frac{\\omega E}{2}\\rp^{1/2}\\frac{\\lp \\int_{4\\pi} \\vu^2\\sth d\\theta d\\varphi\\rp(r=1)}{\\intvol \\vu^2 dV} \\eeq where we used our non-dimensional units. Since the radial dependence of the modes is in $r^m$ and $\\omega=1/(m+1)$, we easily find that \\[ \\gamma = -\\frac{2m+3}{2\\sqrt{2m+2}} \\sqrt{E} \\] When this expression is applied to the $m=2$-mode, we find that $\\gamma=-1.429\\sqrt{E}$ which is a factor $1.74$ weaker than the correct result. If we use, as previous authors, a step function for describing the density difference between that of the crust and the mean density, we find that the damping rate reads \\beq \\gamma_{Ek} = -2.4876\\sqrt{E}\\frac{\\rho_b}{\\overline{\\rho}}\\, 2\\Omega = -0.0346 \\frac{\\rho_b}{\\overline{\\rho}} \\frac{\\sqrt{f\\Omega_\\star}}{T_8} {\\rm s}^{-1}\\eeqn{damp_rate} where $\\rho_b$ is the density of the fluid just below the crust and $\\Omega_\\star=\\Omega/\\sqrt{\\pi G\\overline{\\rho}}$. Our calculation therefore shows that the window of instability in the $\\Omega, T$ plane is smaller than previously estimated for crusted neutron stars. Considering a 1.4~M${_\\odot}$ neutron star with a radius of 12.53~km as a test case, the growth rate of the mode due to gravitational radiation is $\\gamma_{gw} = 0.658 {\\rm s}^{-1} \\Omega_\\star^6$ \\cite[we use the expression given in][]{LOM98}; although, it is not relevant for an incompressible fluid, we take into account the damping rate due to bulk viscosity in order to ease comparison with previous work; from \\cite{LMO99}, we find $\\gamma_{bulk} = -2.2\\, 10^{-12}\\;{\\rm s}^{-1}\\; T_9^6\\Omega_\\star^2$. From \\eq{damp_rate}, we have $\\gamma_{Ek}= -1.53\\, 10^{-3}\\;{\\rm s}^{-1}\\; \\Omega_\\star^{1/2}/T_9$ where we took $\\rho_b=1.5\\,10^{17}$kg/m$^3$; solving the equation \\[ \\gamma_{gw}+\\gamma_{Ek}+\\gamma_{Bulk} = 0 \\] for different values of the temperature yields the curves displayed in figure~\\ref{crit}. As expected, we see that the window of instability narrows compared to \\cite{AJKS00}: for a given temperature, the critical angular velocity raises by $\\sim$10\\% typically. Another interesting conclusion of this work is that the presence of a solid inner core does not change the damping rates very much unless its radius is close to unity. The reason for that is to be found in the shape of the inertial r-modes whose amplitudes are concentrated near the outer boundary. Therefore, the rotating instability of rapidly rotating stars is quite insensitive to the presence of a solid core and more generally to any phase transition which does not occur close to the surface." }, "0003/astro-ph0003201_arXiv.txt": { "abstract": "We present the preliminary results of a detailed theoretical investigation on the hydrodynamical properties of Red Supergiant (RSG) stars at solar chemical composition and for stellar masses ranging from 10 to 20 $M_\\odot$. We find that the main parameter governing their hydrodynamical behaviour is the effective temperature, and indeed when moving from higher to lower effective temperatures the models show an increase in the dynamical perturbations. Also, we find that RSGs are pulsationally unstable for a substantial portion of their lifetimes. These dynamical instabilities play a key role in driving mass loss, thus inducing high mass loss rates (up to almost 10$^{-3}$ \\mloss) and considerable variations of the mass loss activity over timescales of the order of 10$^4$ years. Our results are able to account for the variable mmass loss rates as implied by radio observations of type II supernovae, and we anticipate that comparisons of model predictions with observed circumstellar phenomena around SNII will provide valuable diagnostics about their progenitors and their evolutionary histories. ", "introduction": "More than 20 years of radio observations of supernovae (SNe) have provided a wealth of evidence for the presence of substantial amounts of circumstellar material (CSM) surrounding the progenitors of SNe of type II and Ib/c (see Weiler \\etal, this Conference, and references therein). Also, the radio measurements indicate that $(a)$ the CMS density falls off like $r^{-2}$, suggesting a constant velocity, steady wind, and that $(b)$ the density is so high as to require a ratio of the mass loss rate, $\\dot{M}$, to the wind velocity, $w$, to be higher than $\\dot{M}/w\\sim 10^{-7}$ \\mloss \\kms. These requirements are best satisfied by red supergiants (RSG), with original masses in the range 8-30 \\msun, that indeed are the putative progenitors of SNII. Note that in the case of SNe Ib/c, the stellar progenitor cannot provide such a dense CSM directly and that a wind from a binary companion must be invoked to explain the observations (Panagia and Laidler 1988, Boffi and Panagia 1996, 2000). This scenario is able to account for the basic properties of all radio SNe. However, the evolution of SN 1993J indicated that the progenitor mass loss rate had declined by almost a factor of 10 in the last few thousand years before explosion (Van Dyk \\etal~ 1994). In addition, there are SNe, such as SN 1979C (Montes \\etal~ 2000), SN 1980K (Montes \\etal~ 1998), and SN 1988Z (Lacey \\etal~ 2000), that have displayed relatively sudden changes in their radio emission evolution about 10 years after explosion, which also cannot be explained in term of a constant mass loss rate. Since a SN shock front, where the radio emission originates, is moving at about 10,000~\\kms~ and a RSG wind is typically expanding at 10~\\kms, a sudden change in the CSM density about ten years after explosion implies a relatively quick change of the RSG mass loss rate about 10,000 years before it underwent the SN explosion. These findings are summarized in Figure 1 that, for several well studied RSNe, displays the mass loss rate implied by radio observations as a function of the look-back time, calculated simply as the actual time since explosion multiplied by a factor of 1000, which is the ratio of the SN shock velocity to the RSG wind velocity. Additional evidence for enhanced mass loss from SNII progenitors over time intervals of several thousand years is provided also by the detection of relatively narrow emission lines with typical widths of several 100~\\kms~ in the spectra of a number of SNII (\\eg SN 1978K: Ryder \\etal~ 1993, Chugai, Danziger \\& Della Valle 1995, Chu \\etal~1999; SN 19997ab: Salamanca \\etal~1998; SN 1996L: Benetti \\etal~ 1999), that indicate the presence of dense circumstellar shells ejected by the SN progenitors in addition to a more diffuse, steady wind activity. We note that a time of about 10,000 years is a sizeable fraction of the time spent by a massive star in the RSG phases and implies a kind of variability which is not predicted by standard stellar evolution. In particular, a time scale of $\\sim 10^4$ years is considerably shorter than the H and He burning phases but is much longer than any of the successive nuclear burning phases that a massive star goes through before core collapse (\\eg Chieffi et al. 1999). Therefore, some other phenomenon is to be sought to properly account for the observations. \\begin{figure} \\centerline{\\psfig{figure=panagia_1.ps,width=8.5cm}} \\caption{Mass loss rates as a function of look-back time as measured for a number radio supernovae (adapted from Weiler \\etal~2000; schematic)} \\label{radiomloss} \\end{figure} \\begin{figure} \\centerline{\\psfig{figure=panagia_2.ps,width=8.5cm}} \\caption{Mass loss rates predicted from canonical stellar evolution theory by using Reimers' formula.} \\label{doesntwork} \\end{figure} \\begin{figure} \\centerline{ \\psfig{figure=panagia_3.ps,width=8.5cm}} \\caption{Predicted mass loss rates as a function of the position in the HR diagram adopting Reimers' formula. The size of the circles are proportional to the logarithm of the mass loss rate: the highest values are several 10$^{-6}$ \\mloss~ and the lowest ones about 10$^{-7}$ \\mloss. The dashed lines represent the analytical relations for the fundamental blue and red edges of the Cepheid instability strip defined by Bono \\etal~ (2000a).} \\label{Reimers_mloss} \\end{figure} Another problem which needs to be addressed is the actual rate of mass loss for red supergiants. The observational evidence is that mass loss rates in the range 10$^{-6}$--10$^{-4}$ \\mloss are commonly found in RSG, with a relatively steep increase in mass loss activity for the coolest stars (\\eg Reid, Tinney \\& Mould 1990, Feast 1991). On the other hand, there is no statisfactory theory to predict mass loss rates in these phases of stellar evolution, and current parametrizations fall short from describing the phenomenon in detail. For example, let us consider the classical formula by Reimers (1975), $$log(\\dot{M})~ =~ -12.6~ +~ log(\\frac{LR}{GM})~ + ~log(\\eta)$$ which can be rewritten as: $$\\dot{M}~ \\propto ~\\eta ~\\frac{L^{1.5}}{M~T_{eff}^2} $$ \\vskip .1in \\noindent This formula was devised to dimensionally account for the mass loss from low-mass red giants, but has also been widely adopted for evolutionary track calculations. We see that the predicted mass loss rate varies rather slowly when a star is moving from the blue to the red region (i.e. during H-shell burning and/or He-core burning) of the HR diagram, the main functional dependence being a 1.5 power of the luminosity. The corresponding mass loss rates, computed using the evolutionary tracks by Bono \\etal~ (2000b) for stars in the mass range 10--20\\msun, are shown in Figures 2 and 3. It is apparent that not only the rates are not as high as suggested by spectroscopic observations of RSGs (this aspect alone could easily be ''fixed\" by increasing the efficiency factor $\\eta$) but, more importantly, are very slowly varying with time and, therefore, cannot account for radio observations of SNe, either. Other parametrizations of the mass loss rate in the HR diagram have been proposed by different authors (\\eg De Jager, Nieuwenhuijzen \\& van der Hucht 1988, Salasnich, Bressan and Chiosi 1999), but insofar for RSGs the main dependence of $\\dot{M}$ is a power of $\\sim 2$ of the luminosity, they all are unable to reproduce appreciable mass loss variations over a timescale of roughly 10$^4$ years. Actually, one notices that for masses above 10 \\msun, the last phases of the RGS evolution fall within the extrapolation of the Cepheid instability strip (see Figure 4), as calculated by Bono \\etal~ (1996), and therefore, one may expect that pulsational instabilities could represent the additional mechanism needed to trigger high mass loss rates. Indeed, the pioneering work of Heber \\etal~ (1997), based on both linear and nonlinear pulsation models, demonstrated that RSG stars are pulsationally unstable. In particular, they found that, for periods approaching the Kelvin-Helmotz time scale, these stars display large luminosity amplitudes, which could trigger a strong enhancement in their mass loss rate before they explode as supernovae. According to these authors this pulsation behaviour should take place during the last few $10^4$ yrs before the core collapse, due to the large increase in the luminosity to mass ratio experienced by RSG stars during these evolutionary phases. However, the nonlinear calculations performed by Heber \\etal~ (1997) were hampered by the fact that their hydrodynamic code could not properly handle pulsation destabilizations characterized both by small growth rates due to numerical damping, and by large pulsation amplitudes due to the formation and propagation of strong shock waves during the approach to limit cycle stability. Also, as Heber \\etal~ (1997) pointed out, their main theoretical difficulty in dealing with the dynamical instabilities of RSG variables resided in the coupling between convection and pulsation. In fact, they constructed the linear models by assuming that the convective flux is frozen in, and the nonlinear ones by assuming that the convective flux is instantaneously adjusted. However, this treatment does not account for the driving and/or quenching effects caused by the interaction between pulsation and convection: this shortcoming may explain why their nonlinear models could not approach a stable limit cycle. It is clear that a more general approach must be adopted to solve the problem. This motivated us to start a systematic study of the pulsational properties of massive stars. In the following we shall illustrate briefly the procedures adopted and the first results obtained (Section 2), and will present and discuss our findings on the mass loss rates in the late phases of the evolution of massive stars (Sections 3 and 4). \\begin{figure} \\centerline{ \\psfig{figure=panagia_4.ps,width=10cm}} \\caption{Evolutionary tracks for stars in the range 3.3--25 \\msun. The marked area is the portion of the HR diagram shown in more detail in Figure 5. The dashed lines are the extrapolation of the fundamental boundaries of the Cepheid instability strip according to the analytical relations provided by Bono \\etal~(2000a), which are based on a detailed investigation of Cepheid models at solar chemical composition and stellar masses ranging from 5 to 11 \\msun. Note that a substantial portion of both H-shell and He burning phases for stars with masses higher than about 10 \\msun~ occur within the instability strip.} \\label{hrdiag99} \\end{figure} \\begin{figure} \\centerline{ \\psfig{figure=panagia_5.ps,width=9cm}} \\caption{Evolutionary tracks and pulsation models at solar chemical composition in the HR diagram. Evolutionary models constructed by adopting different stellar masses are plotted by using different line styles, while pulsation models characterized by different limiting amplitude behaviours are plotted using different symbols. See text for further details.} \\label{rosse} \\end{figure} ", "conclusions": "The computed mass loss rates, for stars in the range 12-20 M$_\\odot$, as a function of look-back time are displayed in Figures 10-12 for short (0-30,000 years), medium (0-150,000 years) and long time-scales (0-1 Myrs), respectively. We see that the mass loss rates may be as high as almost 10$^{-3}$ \\mloss, \\ie similar to what is measured for extreme red supergiants, and may vary by an order of magnitude over relatively short times, say, 10,000 years or less. In other words the predicted mass loss rates are able to account, at least qualitatively, for all of the features observed in radio supernovae. Moreover, since the predicted mass loss history is a critical function of how a massive star evolves within the pulsation instability strip, a comparison between observations and theory should lead to an accurate determination of the stellar progenitor mass. For example, the mass loss decline of a 20 \\msun star may be used to represent the apparent drop of emission of SN~1988Z about 9 years after explosion (cf. Figure 1). Similarly, the quick increase found for our 14~\\msun model closely resembles the behaviour observed for SN~1993J. Of course, detailed comparisons will be meaningful only we will have a fully self-consistent set of evolutionary tracks (see below). The mid- and long-term behaviour of the mass loss rate as a function of look-back time is also interesting because allows one to make predictions about the radio emission, as well as on {\\it any} other phenomenon linked to a SN shock front and/or ejecta interaction with a dense circumstellar medium, such as relatively narrow optical emission lines and X-ray emission. As we can see in Figures 11 and 12, massive stars are expected to display rather sudden variations of their mass loss rates of all time scales, both because of pulsational instabilities which arise with crossing the instability strip (\\eg the 12~\\msun star in the time range 20-60$\\times 10^3$ years) and because of the so-called blue loops (an effect clearly apparent at look-back times around 0.4--1 Myrs) that are determined by a combination of core He-burning and shell H-burning (\\eg Brocato \\& Castellani 1993, Langer \\& Maeder 1995). Because of these effects, one may expect that in some cases, a SN may drop below detection limit for a while but still may have a renaissance, in the X-ray, optical and radio domains, several tens or hundreds of years later. \\begin{figure} \\centerline{ \\psfig{figure=panagia_11.ps,width=8.5cm}} \\caption{Pulsationally enhanced mass loss rates as a function of look-back time (medium time-scale variation).} \\label{doeswork2} \\end{figure} \\begin{figure} \\centerline{ \\psfig{figure=panagia_12.ps,width=8.5cm}} \\caption{Pulsationally enhanced mass loss rates as a function of look-back time (long time-scale variation).} \\label{doeswork3} \\end{figure} Also, we note in passing that our findings support the empirical evidence recently brought out by van Loon et al. (1999) on the basis of ISO data on RSG stars in the LMC. In fact, they found that the mass loss rates increase with increasing luminosities and decreasing effective temperatures and range from $10^{-6}$ up to $10^{-3} M_\\odot yr^{-1}$. A strong dependence of the mass loss rate on the effective temperature in Tip-AGB stars was recently suggested by Schr\\\"oder, Winters and Sedlmayer (1999) on the basis of theoretical evolutionary models which account for carbon-rich wind driven by radiation pressure on dust. Another interesting consequence of our results is that a more efficient mass loss in the RSG phase implies a lower mass cutoff to produce Wolf-Rayet stars and, therefore, one has to expect a more efficient mass return into the ISM than commonly adopted in galactic evolution calculations. Still there are improvements and refinements to apply to our models, because the calculations we presented here are not fully self-consistent in that we adopted evolutionary tracks computed either with no mass loss whatsoever, or with modest mass loss rates, and on them we performed our pulsational stability analysis and, thus, determined our new mass loss rates. Moreover, our models were constructed by adopting the diffusion approximation even in optically thin layers and therefore we neglected the dust formation processes (Arndt et al. 1997). A macroscopic example of the shortcomings of our current approach is that if we integrate the mass loss rates over time, in many cases we find that the star looses a substantial fraction of their mass before reaching its evolutionary end. Although this is close to what one should expect on the basis of observations, definitely it is at variance with the assumptions that went into the adopted evolutionary model calculations. It is clear that what we need to do now is to follow an iterative procedure in which we first use our present prescriptions to compute new evolutionary tracks, then we repeat our pulsational stability analysis, then we compute new mass loss rates, and we iterate the procedure until adequate convergence is achieved. This work is in progress and will be presented in future papers. For the time being, our conclusions can be summarized as follow: \\\\ -- We have defined a new theoretical scenario for pulsation induced mass loss in RSGs.\\\\ -- RSGs are pulsationally unstable for a substantial portion of their lifetimes.\\\\ -- Dynamical instabilities play a key role in driving mass loss.\\\\ -- Bright, cool RGSs undergo mass loss at considerably higher rates than commonly adopted in stellar evolution.\\\\ -- Comparisons of model predictions with observed CSM phenomena around SNII will provide valuable diagnostics about their progenitors and their evolutionary history.\\\\ -- More efficient mass loss in the RSG phase implies a lower mass cutoff to produce Wolf-Rayet stars and a more efficient return of polluted material into the ISM, thus affecting the expected chemical evolution of galaxies." }, "0003/gr-qc0003070_arXiv.txt": { "abstract": "A vector-tensor theory of gravity that was introduced in an earlier publication is analyzed in detail and its consequences for early universe cosmology are examined. The multiple light cone structure of the theory generates different speeds of gravitational and matter wave fronts, and the contraction of these light cones produces acausal, superluminary inflation that can resolve the initial value problems of cosmology. ", "introduction": "In previous work~\\cite{Clayton+Moffat:1999,Clayton+Moffat:1999a}, we have introduced a new kind of vector-tensor and scalar-tensor theory of gravity, which exhibits a bimetric structure and contains two or more light cones. This type of model has attracted some interest recently~\\cite{Bassett+:2000,Liberati+:2000,Avelino+Martins:2000}, and similar effects have been noted elsewhere~\\cite{Herdeiro:2000}. The motivation for considering these models is derived form earlier work of one of the present authors~\\cite{Moffat:1993a}, which provided a scenario in which some of the outstanding issues in cosmology can be resolved. The present line of work provides a specific class of models that realize these ideas, for it provides a fundamental dynamical mechanism for varying speed of light theories and generates a new mechanism for an inflationary epoch that could solve the initial value problems~\\cite{Moffat:1993a} of early universe cosmology. In this article, we focus on clarifying the role that these models can play in the early universe, demonstrating how matter that satisfies the strong energy condition can nevertheless contribute to the cosmic acceleration. Recently, an analysis of a similar class of theories has appeared~\\cite{Bassett+:2000,Liberati+:2000} which, while introducing some interesting ideas, unfortunately claimed that we had made an algebraic error in our previous work~\\cite{Clayton+Moffat:1999}. This can be attributed to a misunderstanding of part of the construction that was perhaps not dealt with in sufficient detail in our initial publication. We will rectify this situation here, developing the model in additional detail and showing explicitly that the error attributed to us in~\\cite{Bassett+:2000} is in fact an error on their part. We will also show that our cosmological model can be mapped to a model with varying fundamental constants~\\cite{Albrecht+Maguiejo:1999,Barrow+Maguiejo:1998,Barrow:1999}, albeit not uniquely and requiring some care in the interpretation of the varying constants that appear. It is hoped that the models can shed some light on the new observational data that suggests the expansion of the universe at present is undergoing an acceleration~\\cite{Perlmutter+etal:1999,Garnavich+etal:1998,Bachall+etal:1999}. Although there has been some success in understanding the latter problem by the inclusion of a class of very particular scalar field potentials~\\cite{Caldwell+Dave+Steinhardt:1998}, it is fair to say that not all issues have been resolved. In this article, we will not have much to say about this issue since, as we shall see, the vector field that produces a superluminary expansion in the early universe must vanish at some scale, and standard cosmology results afterwards. Using the scalar field version of the model, we expect that not only will we be able to generate sufficient inflation, but that a quintessence-like solution should be achieveable. We shall concentrate our efforts on the vector-tensor model by providing a more detailed analysis of its consequences and postpone a fuller analysis of the scalar-tensor model to a future article. ", "conclusions": "One of our purposes here was to give a more complete description of the model than was given in~\\cite{Clayton+Moffat:1999}. In doing so we showed that the universe generically accelerates ($\\hat{q}<0$) during some period in the early universe, and that in the same period the physical importance of spatial curvature diminishes ($\\vert\\Omega-1\\vert$ is decreasing). This can occur even when the matter fields satisfy the strong energy condition. This conclusion is the opposite of that which appeared recently~\\cite{Bassett+:2000,Liberati+:2000}, who take a somewhat different point of view on the interpretation of the metrics appearing in these theories. Nevertheless, we have demonstrated conclusively that the claim appearing in~\\cite{Bassett+:2000} that we had made an algebraic error in~\\cite{Clayton+Moffat:1999} is false. The model that we have considered generalizes that which appeared in~\\cite{Clayton+Moffat:1999} in a way that more closely follows the scenario discussed in~\\cite{Moffat:1993a}. In the very early universe, matter and gravitational fields propagate with different and approximately constant velocities. During a period during which the matter light cone, originally much larger than the light cone of gravity, contracts, material observers will see an acausal expansion of the universe similar to inflation. Because the light cone of gravity does not undergo the same contraction, we expect there to be an observable difference in the scalar versus tensor contributions to the cosmic microwave background anisotropies." }, "0003/astro-ph0003037_arXiv.txt": { "abstract": "{We report the first detection of CO in M\\,31's bulge. The $^{12}$CO (1-0) and (2-1) lines are both detected in the dust complex D395A/393/384, at 1$\\farcm$3 ($\\sim 0.35$ kpc) from the centre. From these data and from visual extinction data, we derive a CO-luminosity to reddening ratio (and a CO-luminosity to H$_2$ column density ratio) quite similar to that observed in the local Galactic clouds. The (2-1) to (1-0) line intensity ratio points to a CO rotational temperature and a gas kinetic temperature $> 10$ K. The molecular mass of the complex, inside a 25$\\arcsec$ (100 pc) region, is 1.5 10$^4$ $M_\\odot$.} ", "introduction": "The bulk of gas of M\\,31 {lies between 6 and 18 kpc and follows a pattern of thin spiral arms (e.g. Dame et al.\\ 1993, Neininger et al. 1998) \\nocite{Dame:1993, Neininger:1998}. Despite the presence of dust lanes, massive stars, and evolved stars (e.g. O'Connell et al. 1992; Davidge et al.\\ 1997)\\nocite{O'Connell:1992, Davidge:1997}, there is little HI atomic gas in the inner bulge and, up to now, molecular gas had escaped detection: Sofue \\& Yoshida (1993)\\nocite{Sofue:1993} reported the detection the $^{12}$CO(1-0) line in the dust complex D395/393 at less than 1~kpc from the centre, but this detection was invalidated by Loinard, Allen \\& Lequeux (1996)\\nocite{Loinard:1996} who re-observed the same complex with a much better sensitivity. So, the detection of CO in the dust complexes D~478 and D~268, at 2-4~kpc from the centre\\footnote{We assume a distance to M\\,31 of 780~kpc (e.g. Holland 1998; Stanek \\& Garnavich 1998\\nocite{Holland:1998,Stanek:1998c}), i.e.\\ 1$\\farcs$=3.8pc.}, by Allen \\& Lequeux (1993) and Loinard \\& Allen (1998)\\nocite{Allen:1993,Loinard:1998}, was the closest to the centre ever reported. Noting that the CO (1-0) to (2--1) intensity ratio was lower in these clouds than in the Galactic Giant Molecular Clouds, these authors concluded that CO may be so cold in the {\\em inner disc} of M\\,31, that its excitation temperature barely exceeds the cosmic background temperature. They thus suggested that although molecular gas and CO could be abundant, the CO mm lines are so weak that they escape detection. In order to better understand the gas budget at the centre of M\\,31, we have embarked on a reanalysis of the optical data. We found that the positions observed by Loinard, Allen \\& Lequeux (1996)\\nocite{Loinard:1996} were not centred on the darkest dust patches, which prompted us, taking advantage of the installation of new generation mm-wave receivers on the IRAM 30-m telescope, to make new CO observations.} ", "conclusions": "Emission of the (1-0) and (2-1) { lines} of the $^{12}$CO molecule has been detected near the centre of M\\,31 at $15\\sigma$ and $10\\sigma$ respectively. The CO centroid velocity of $-155$~km~s$^{-1}$ suggests that this molecular gas could be located in the ionised gas disc detected in these regions. { The observed velocity dispersion of the molecular gas in the beam is compatible with the velocity gradient of the ionised gas. Together with the patchy appearance of the extinction map, this indicates the presence of several clouds distributed over the beam.} The I$_{\\rm CO}$/$E(B-V)$ ratio is remarkably similar to the values observed in molecular clouds in the Galaxy and M\\,31's disc. Based on a standard gas-to-dust ratio, the mass of the molecular complex is of order 10$^4$~M$_{\\odot}$. { In order to reach densities compatible with the excitation of the CO lines,} the complex must be highly clumped. As previous studies with poorer resolution have shown, the mass inferred is small compared to the mass content of the Galactic centre (Morris \\& Serabyn 1996)\\nocite{Morris:1996}. With such a mass the line-width of the CO line is by far too broad for a gravitationally bound complex, which explains previous inconsistencies based on this assumption. { Accordingly, we modelled the extinction with a random distribution of small spherical clumps, all identical with a size $r$ and density $\\rho$. They all lie in a sphere of radius 13$\\arcsec$, tracing a total mass of $10^4$ M$_\\odot$. Models with $\\sim$250 clumps reproduce the main features of the histogram of measured extinctions. With the previous assumption of a common location for the molecular, ionised gas and the dust, the configuration (r,$\\rho$) which reproduces best the measured extinction correspond to clumps with r$\\sim$2.6$\\arcsec$ and $\\rho\\sim$400~H~cm$^{-3}$. Following the LVG homogeneous cloud models of Garc\\'\\i a-Burillo, Gu\\'elin $\\&$ Cernicharo (1993)\\nocite{Garcia-Burillo:1993}, the measured line ratio corresponds to CO (1-0) and (2-1) line excitation temperatures and to a kinetic temperature all $\\geq$10~K. This modelling will be further discussed in a subsequent paper. The ``standard'' CO-luminosity to visual extinction ratio and the relatively high CO (1-0) excitation temperature seem to rule out the presence of large amounts of hidden CO and H$_2$ in the inner bulge of M\\,31, in apparent disagreement with the result of Loinard et al. (1996)\\nocite{Loinard:1996} in the inner disc.}" }, "0003/astro-ph0003070_arXiv.txt": { "abstract": "We report on searches of the globular cluster Terzan 5 for low luminosity and accelerated radio pulsars using the 64-m Parkes radio telescope. One new millisecond pulsar, designated PSR~J1748$-$2446C, was discovered, having a period of 8.44~ms. Timing measurements using the 76-m Lovell radio telescope at Jodrell Bank show that it is a solitary pulsar and lies close to the core of the cluster. We also present the results of timing measurements which show that the longer-period pulsar PSR~J1748$-$2444 (formerly known as PSR~B1744$-$24B) lies 10 arcmin from the core of the cluster and is unlikely to be associated with the cluster. We conclude that there are further pulsars to be detected in the cluster. ", "introduction": "Globular clusters were proposed as good hunting grounds for rapidly rotating pulsars because they were known to contain a relatively high proportion of potential progenitor systems such as low mass X-ray binaries \\cite{hhb85}. Since the first globular cluster pulsar was found in M28 \\cite{lbm+87}, several dozen have been found in a number of clusters \\cite{lyn95}. This paved the way for new understanding of the formation and evolution of millisecond pulsars \\cite{ka96}, as well as providing opportunities for other applications such as measuring cluster gravitational potentials \\cite{phi92b,clf+99}. In an early search of the cluster Terzan 5 \\cite{lmd+90}, two radio pulsars were found, PSR~B1744$-$24A and PSR~B1744$-$24B. PSR~B1744$-$24A (also known as PSR~J1748$-$2446A) is clearly associated with the cluster and lies $\\sim 30\\arcsec$ from its core. The pulsar is in a 1.7-hour orbit and is often eclipsed by its companion star \\cite{lmd+90,nt92}. PSR~B1744$-$24B is rather weak and, until now, it has not had a phase connected timing solution, so that its precise position was not known and any association with the cluster was unclear. Radio continuum observations at the VLA have revealed the presence of a number of steep spectrum radio sources, within, or close to, the core of the cluster, hinting at the possibility of as yet undiscovered pulsars \\cite{fg90}. Since PSR~J1748$-$2446A is not in the core of the cluster and we show here that PSR~B1744$-$24B is not associated with the cluster at all, these two pulsars account for none of the flux density detected by Fruchter \\& Goss (1990)\\nocite{fg90}. This suggests that there may be further undiscovered pulsars within the core of the cluster. Any other pulsars may remain undetected as a result of broadening of the pulses due to dispersion in the receiver filterbank channels, broadening of the pulses due to acceleration in tight orbits, very short pulsation period or insufficient sensitivity. In this paper, we report on a search which was undertaken to try to resolve these issues and present the one new pulsar that was found, which we have designated PSR~J1748$-$2446C. We also report on timing observations of both this pulsar and PSR~B1744$-$24B in order to establish their positions and associations with the cluster. ", "conclusions": "The new single millisecond pulsar, PSR~J1748$-$2446C (Ter5C), lies only about $\\sim 10\\arcsec$ from the cluster centre ($\\alpha$(J2000): $17^{\\rm h}48^{\\rm m}04\\fs9$, $\\delta$(J2000): $-24^{\\circ}46\\arcmin 45\\arcsec$ \\cite{fg00}) and has a dispersion measure close to that of PSR~J1748$-$2446A (Ter5A), suggesting with high probability that it also is associated with the cluster. \\begin{figure} \\plotone{fig1.ps} \\caption{Positions of sources associated with the globular cluster Terzan 5. A: PSR~J1748$-$2446A, C: PSR~J1748$-$2446C, circle: cluster core, radius = $7\\farcs9$ (Trager et al. 1993), N,S: northern and southern radio sources identified by Fruchter \\& Goss (2000), and X: XB 1745--25 (Johnston et al. 1995). } \\end{figure} \\nocite{jvh95} \\nocite{tdk93} \\nocite{fg00} The position for PSR~J1748$-$2444 given in Table 1 is $10\\arcmin$ from the cluster centre, indicating that it is unlikely that this pulsar is associated with the cluster. As Table 1 shows, this pulsar has a dispersion measure which differs from the values for the other two pulsars, Ter5A and Ter5C, by about 30 cm$^{-3}$ pc, adding weight to the conclusion that PSR~J1748$-$2444 is not associated with the cluster. The period of 0.443~s and the measured period derivative indicates that it is an unremarkable pulsar, having a surface magnetic field of $2.3\\times10^{11}$G which is near to the lower end of the distribution of normal pulsars \\cite{tml93}, and a characteristic age of 63~My. Figure 1 shows the positions of the two cluster pulsars and other sources around Terzan 5 relative to the cluster centre. The two radio sources (N and S) were found by Fruchter \\& Goss (2000)\\nocite{fg00} in a continuum image of the region at around 1400~MHz. Although the position of source N agrees within the errors with that of the 8.4-ms pulsar PSR~J1748$-$2446C, the flux density of the pulsed emission from this pulsar has a mean value of only $0.5\\pm0.2$~mJy. This is significantly less than the flux density of source N, 1.5 mJy. Neither the flux density of the pulsar or of source N vary significantly between observations, so that such variations cannot explain the discrepancy. We conclude therefore that other, rapidly-rotating pulsars or pulsars in short-period binary orbits may exist in Terzan 5. Searches with a more sensitive receiving system having better time and frequency resolution may find them. Johnston, Verbunt \\& Hasinger (1995)\\nocite{jvh95} identified a transient X-ray source close to the cluster core; its positional accuracy is not sufficient to confirm an identification with the radio source S. The negative period derivatives of Ter5A and Ter5C indicate that they lie behind the cluster and are experiencing a gravitational acceleration towards the cluster core in the manner described by Phinney (1992)\\nocite{phi92b}. If we assume that the intrinsic period derivatives are greater than zero and follow the procedures of Phinney (1992) and Camilo et al. (1999)\\nocite{clf+99}, we can place a lower limit on the central mass density $\\rho(0)$ of the cluster. Using the cluster core radius $r_{c} = 7\\farcs9$ \\cite{tdk93}, a cluster distance of 7.6 kpc \\cite{jvh95}, and the cluster position of Fruchter \\& Goss (2000)\\nocite{fg00}, we find lower limits on $\\rho(0)$ of $0.85 \\times 10^5$ and $5.0 \\times 10^5$ M$_{\\odot}$pc$^{-3}$ for Ter5A and Ter5C, respectively. We note that the latter is considerably lower than the value of the density quoted by Webbink (1985) \\nocite{web85} of $24 \\times 10^5$ M$_{\\odot}$pc$^{-3}$." }, "0003/astro-ph0003246_arXiv.txt": { "abstract": "Assuming an adiabatic evolution of a Gamma-Ray Burst (GRB) remnant interacting with an external medium, we calculate the injection, cooling, and absorption break frequencies, and the afterglow flux for plausible orderings of the break and observing frequencies. The analytical calculations are restricted to a relativistic remnant and, in the case of collimated ejecta, to the phase where there is an insignificant lateral expansion. Results are given for both a homogeneous external medium and for a wind ejected by the GRB progenitor. We compare the afterglow emission at different observing frequencies, for each type of external medium. It is found that observations at sub-millimeter frequencies during the first day provide the best way of discriminating between the two models. By taking into account the effect of inverse Compton (IC) scatterings on the electron cooling, a new possible time-dependence of the cooling break is identified. The signature of the up-scattering losses could be seen in the optical synchrotron emission from a GRB remnant interacting with a pre-ejected wind, as a temporary mild flattening of the afterglow decay. The up-scattered radiation itself should be detected in the soft X-ray emission from GRB remnants running into denser external media, starting few hours after the main event. ", "introduction": "One of the most important issues regarding Gamma-Ray Bursts (GRBs) is the nature of the object that releases the relativistic ejecta generating the high energy emission of the main event and the lower frequency emission during the ensuing afterglow. Some insight about the GRB progenitor can be obtained from the properties of the circum-burst medium, which can be inferred from the features of the afterglow emission. If the ejecta is expelled during the merging of two compact objects (\\Mesz \\& Rees 1997b), it is expected that the medium surrounding the GRB source is homogeneous. However, if a collapsing massive star (Woosley 1993, Paczy\\'nski 1998) is the origin of the relativistic fireball, the circum-burst medium is the wind ejected by the star prior to its collapse, whose density decreases outwards. The two models differ in the dependence on radius of the particle density of the circum-burst medium which the GRB remnant interacts with, and in the value of this density at the deceleration length-scale. The former modifies the rate of decline of the afterglow, while the latter determines the overall afterglow brightness. Therefore it is possible to correlate afterglow emission features to a specific type of external medium. Significant work in this direction has been done by many researchers. The two afterglows that exhibited breaks consistent with the effects arising from strong collimation of ejecta -- GRB 990123 (Kulkarni \\etal 1999a) and GRB 990510 (Stanek \\etal 1999, Harrison \\etal 1999) -- indicate that the external gas was homogeneous (recent work by Kumar \\& Panaitescu 2000 shows that jets interacting with winds cannot produce sharp breaks in the afterglow light-curve). The optical emission of three afterglows had a steeper than usual decline. GRB 970228 decayed as $T^{-1.7}$ after the subtraction of an underlying supernova emission (Reichart 1999, Galama \\etal 2000). The light-curve of GRB 980326 fell off as $T^{-2.1}$ (Groot \\etal 1998) and an emission in excess of the early time extrapolation was detected $\\sim 20$ days after the main event, indicating a supernova contribution (Bloom \\etal 1999). A $T^{-2}$ decay was observed for the afterglow of GRB 980519 (Halpern \\etal 1999). Such steep declines can be produced either by a fireball interacting with a pre-ejected wind (Chevalier \\& Li 1999) and an electron index around 3, or by a narrow jet expanding laterally in a homogeneous external medium and an electron index slightly larger than 2. Chevalier \\& Li (1999) found that the radio emission of the afterglow of GRB 980519 is consistent with an external wind; however Frail \\etal (2000) point out that the interstellar scintillation present in the radio data does not allow ruling out the jet model. Nevertheless, the existence of supernovae associated with GRB 970228 and GRB 980326 points toward a massive star as the origin of these bursts, implying a pre-ejected wind as the external medium. From the analysis of the optical radio emission of the afterglow of GRB 970508, Chevalier \\& Li (2000) conclude that the surrounding medium was a wind. Frail, Waxman \\& Kulkarni (2000) argue that the same radio afterglow can be explained by a homogeneous external medium. In this work we investigate the differences between the light-curves of afterglows arising for each type of external medium, with the aim of finding ways for distinguishing between the two models. This study is done within the usual framework of a relativistic remnant interacting with a cold external gas. As the fireball is decelerated, a shock front sweeps up the external gas, accelerating relativistic electrons and generating a magnetic field in the shocked gas. We ignore the emission from electrons accelerated by the reverse shock which propagates through the ejecta at very early times. At optical wavelengths this emission is short lived, lasting up to few tens of seconds after the main event (Sari \\& Piran 1999), but it could be important for the radio emission until few days (Kulkarni \\etal 1999b). Analytical afterglow light-curves for spherical remnants interacting with homogeneous external media have been previously published by Sari, Piran \\& Narayan (1998). Features of afterglows from spherical fireballs, such as peak flux, break frequencies, and time evolution of fluxes at a fixed frequency, have been studied by \\Mesz \\& Rees (1997a), Waxman (1997b), Wijers \\& Galama (1999), and Dai \\& Lu (2000) for homogeneous media, and by Chevalier \\& Li (2000) for pre-ejected winds. In this work we present and compare analytical and numerical light-curves at various observing frequencies, covering all cases of interest, for both types of external media, taking into account the differential arrival-time delay and Doppler boosting due to the spherical shape of the source. We take into account first order IC scattering, calculate its effect on the electron cooling and on the afterglow synchrotron emission, and study briefly the high-energy emission resulting from the up-scattering of synchrotron photons. The possible importance of IC scatterings for the early afterglow emission was pointed out by Waxman (1997a) and Wei \\& Lu (1998). \\vspace*{5mm} ", "conclusions": "Using the analytical results given in equations (\\ref{F01}) -- (\\ref{F28}), the afterglow light-curve can be calculated at any frequency and at observing times up to the onset of the non-relativistic phase. As illustrated in Figure 3, the largest differences between the afterglow emission in the two models for the external medium is seen at low frequencies (lower panels). However, the scintillation due to the local interstellar medium (Goodman 1997), may hamper the use of the radio light-curves to identify the type of external medium and geometry of the ejecta (Frail \\etal 2000). Figure 4 shows that, for various model parameters, the rate of change of the afterglow emission at $\\nu \\sim 10^{12}$ Hz and at early times (when the jet effects are negligible, provided that the jet is initially wider than a few degrees) exhibits a strong dependence on the type of external medium. If the external medium is homogeneous the sub-millimeter afterglow should rise slowly at times between $\\sim 1$ hour and $\\sim 1$ day, while for a pre-ejected wind the emission should fall off steeply, followed by a plateau\\footnotemark . \\footnotetext{The light-curves presented in the upper left panel of Figure 4 show that this criterion for determining the type of external medium fails only if the particle density of the homogeneous medium exceeds $\\sim 10\\;{\\rm cm^{-3}}$. In this case the X-ray emission may help to distinguish between the two models of external medium, as the absence of a flattening of the high energy emission is compatible only with a pre-ejected wind with $A_*$ less than a few.} Therefore observations made at sub-millimeter frequencies with the SCUBA (James Clerk Maxwell Telescope) or with the MAMBO (IRAM Telescope) instruments would be very powerful in determining if the medium which the remnant runs into is homogeneous or follows a $r^{-2}$ law. We note that if turbulence in the shocked fluid does not lead to a significant mixing, then the inhomogeneous electron distribution will alter the afterglow spectrum below the absorption frequency $\\nu_a$ as described by Granot, Piran \\& Sari (2000). The result is that the afterglow emission at $\\nu < \\nu_a$ rises more slowly than calculated here. For instance, the $T^2$ rise exhibited by the $\\nu = 10^{12}$ Hz light-curves shown in Figure 4 for the wind model at early times becomes a $T^1$ rise. Nevertheless, the basic difference mentioned above between the temporal behaviors of the sub-millimeter light-curves at $10^{-2}\\,{\\rm day} \\siml T \\siml 1\\,{\\rm day}$ remains unchanged. The IC losses alter the evolution of the cooling break $\\nu_c$ if the electrons injected with minimal energy are adiabatic and if the Compton parameter is above unity (\\ie the magnetic field parameter is weaker than that given in \\eq [\\ref{epsB0}]). In the case of a homogeneous external medium, the cooling break frequency decreases as $T^{-1/2}$ if the electrons are radiative. When the electrons become adiabatic, this break evolves as $T^{\\frac{3p-8}{8-2p}}$ for $p < 3$ and increases as $T^{1/2}$ for $p > 3$. For an external wind the change is from $\\nu_c \\propto T^{1/2}$ to $\\nu_c \\propto T^{\\frac{3p-4}{8-2p}}$ for $p < 3$, and to $\\nu_c \\propto T^{5/2}$ for $p > 3$. Consequently the power-law decay of the afterglow emission at frequencies above the cooling break flattens by up to $1/2$ if the external medium is homogeneous and by up to $1$ if the medium is a wind. For an electron index $p < 3$, the flattening is mild and likely to be seen only in the optical emission from a remnant interacting with a pre-ejected wind. The IC emission itself is generally weaker than the synchrotron emission. Nevertheless, if the external medium is sufficiently dense (\\ie $n_* \\simg 10\\,{\\rm cm^{-3}}$ or $A_* \\simg$ few), a flattening of the soft X-ray light-curve should be seen few hours after the main event, at fluxes well above the threshold of BeppoSAX (see Figure 3, upper left panel). The flattening of the afterglow emission due to the up-scattered radiation is a chromatic feature, appearing only at high frequencies, and its strength is moderately dependent on the remaining model parameters. Finally, another possible signature of the interaction with the wind ejected by a Wolf-Rayet star should be found during the GRB emission in the form of smooth pulses of increasing duration. Such pulses are generated in internal shocks when the decelerating outermost shell is hit from behind by shells ejected at later times. This phenomenon is more likely to be seen in the wind model, for which the deceleration radius is smaller than for a homogeneous medium." }, "0003/astro-ph0003136_arXiv.txt": { "abstract": "We have found the stellar populations of early-type galaxies are homogeneous with no significant difference in color or Mg2 index, despite the dichotomy between X-ray extended early-type galaxies and X-ray compact ones. Since the X-ray properties reflect the potential gravitational structure and hence the process of galaxy formation, the homogeneity of the stellar populations implies that the formation of stars in early-type galaxies predates the epoch when the dichotomy of the potential structure was established. ", "introduction": "ASCA X-ray observations of NGC 4636 (Matsushita et al. 1998) and some other giant early-type galaxies (Matsushita 1997) show that early-type galaxies can be classified into two categories in terms of X-ray extent. Some early-type galaxies have a very extended dark matter halo characterized by X-ray emission out to $\\sim$ 100~kpc from the galaxy center, while others have a compact X-ray halo. The galaxies with an extended X-ray emission can be interpreted as sitting in larger scale potential structure, such as galaxy groups, subclumps of clusters, or clusters themselves, as well as sitting their own potential well associated with each galaxy. Potential structure must have played a big role in the cource of galaxy formation. If the difference in potential structure had been already established before the bulk of stars formed, we would expect some differences in stellar populations, such as mean age or metallicity, as well. A deeper potential well would keep the gas more effectively against the thermal energy input by supernova (SN) explosion, and the chemically enriched gas can be recycled more efficiently, and the galaxy would end up with higher mean stellar metallicity (Larson 1974). Therefore we would expect that the X-ray extended galaxies have higher metallicities than the X-ray compact ones at a given stellar mass. Furthermore, considering that the higher density peaks collapse earlier in the Universe, which is likely to be the case for the X-ray extended galaxies sitting in the local density peaks, we would also expect their older ages than those of the X-ray compact ones. Both of these effects would make the colors of the X-ray extended galaxies redder. The central question of this paper is, therefore, how this dichotomy in X-ray properties hence the potential structure is related to the optical properties which trace the stellar populations in galaxies. Another interesting issue is whether the number of globular clusters per unit optical luminosity of galaxy correlates with the X-ray extent of galaxy. This is because, if the X-ray extended galaxies are the products of galaxy mergers as they are located in the center of larger scale potential structure, and if a considerable number of new globular clusters form during galaxy mergers as suggested by Zepf and Ashman (1993), we would expect more globular clusters in the X-ray extended galaxies for a given optical luminosity. Matsushita (2000; hereafter M2000) has recently compiled the X-ray properties of 52 nearby early-type galaxies from ROSAT data. Together with the archival data of various optical properties, we now compare the optical properties with the X-ray properties to examine the correlation between them. The structure of this paper is the following. In \\S~2 we summarize the X-ray properties of our sample of early-type galaxies, highlighting the dichotomy of the potential structure. In \\S~3 we present their optical properties, including integrated colors and Mg$_2$ index. We show the homogeneity of the stellar populations despite the dichotomy in X-ray properties. We discuss the impact of this result on the formation of early-type galaxies in \\S~4, and conclude the paper in \\S~5. We use $H_0=75$ km s$^{-1}$ Mpc$^{-1}$ throughout this paper. ", "conclusions": "The stellar population makeup in early-type galaxies does not correlate with their present-day global potential structure. Early-type galaxies form stars at early epoch in their own potential wells independently, and some of the galaxies become incorporated into larger scale potential structures (clusters/groups) later on. This idea naturally explains the homogeneity of the stellar populations despite the variety of X-ray properties." }, "0003/astro-ph0003300_arXiv.txt": { "abstract": "The OH 1667 MHz maser in the circumstellar shell around the Mira variable U~Her has been observed with the NRAO Very Long Baseline Array (VLBA) at 6 epochs, spread over 4 years. Using phase referencing techniques the position of the most blue-shifted maser spot was monitored with respect to two extra-galactic radio sources. The absolute radio positions of the maser can be compared with the stellar optical position measured by the Hipparcos satellite to 15 mas accuracy. This confirms the model in which one of the maser spots corresponds to the stellar continuum, amplified by the maser. The stellar proper motion and the annual parallax ($\\pi_{\\rm VLBI} = 5.3 \\pm 2.1$ mas) were measured. ", "introduction": "\\label{intro} Until recently, distances to Mira variables were mostly based on the Period -- Luminosity relation. Primary distance measurements are important to discuss the calibration and origin of this relation, which bears on the understanding of the structure and evolution of stars on the AGB. Prior to the Hipparcos mission there were very few measurements of parallaxes of Mira stars. For a number of nearby Miras the distances are now better known; however, in several cases and for U Her in particular, the optical parallax and hence the distance is very uncertain. In this paper we show that it is possible to obtain fundamental measurements of AGB star properties by monitoring the OH maser positions with Very Long Baseline Interferometry (VLBI). The Mira variable U~Her is a well known source for studies of circumstellar masers. Both its main line OH and water masers have been the target of a number of VLBI studies (e.g. Chapman et al. \\cite{Chapman}, Yates \\& Cohen \\cite{YatCoh}). It has been assumed to be relatively close; Chapman et al.\\ \\cite{Chapman} give a value of 385 pc based on the Mira Period -- Luminosity relation. Using a revised P -- L relation Alvarez \\& Mennesier (\\cite{AlMe}) find a value of 280 pc. In order to use the maser positions to monitor the trajectory of U~Her, an assumption has to be made about the motion of the masers with respect to the underlying star. If the most blue-shifted circumstellar maser spot corresponds to a special condition, maser action in the shell initiated by the radio continuum radiation from the photosphere, then this most blue-shifted spot should be a bright beacon, necessarily fixed on the true stellar position. VLBI observations by Sivagnanam et al.\\ (\\cite{SivaAmpl}) provided strong evidence for this. They showed that in U~Her the dominant OH 1665 and 1667 MHz VLBI features at the most blue-shifted side of the spectrum are coincident, in accordance with such a model. Observations presented here support the amplified stellar image model, although some questions still remain. Even with considerable errors on the parallax, the Hipparcos data on U~Her allow a comparison between the optical and maser positions with unprecedented accuracy. Traditionally the position of the star with respect to the maser features, as well as the relative positions of maser features of different species, had to be assumed, for instance by fitting a shell and assuming the star is in the center. It has now become possible to overlay these positions directly to 15 mas accuracy. \\begin{figure*} \\resizebox{\\hsize}{!}{\\includegraphics{uher.f1}} \\hfill \\caption{The motion of the most blue-shifted 1667~MHz maser spot at $v_{\\rm lsr}=-20.6$ km/s with respect to one of the extra-galactic reference sources. Images are 300x300 mas with a flux scale of 0 to 500 mJy} \\label{motion} \\end{figure*} ", "conclusions": "One motivation to attempt the detection of the U~Her motion by means of the OH maser, was the prediction that on the line of sight to the star the maser will amplify radio photons from the stellar surface (Norris et al.\\ \\cite{Norris}; Van~Langevelde \\& Spaans \\cite{Spaans2}). In a spherically symmetric shell with a constant outflowing velocity, the most blue-shifted part of the masering shell would then be amplifying a background source rather than spontaneous emission. Hence this would result in a spot with more maser beaming and thus brighter, more compact emission. The spot would be fixed to the stellar radio-photosphere (which could be as large as $\\approx 20$ mas for U~Her, while the visible photosphere is perhaps twice smaller) and probably be persistent over the years. It would also be observable in different maser lines, as the mechanism would work in a similar way for all transitions. It could still vary in flux with stellar cycle through changing pumping conditions. This idea is confirmed by our results, although some issues remain unexplained. We find that the brightest and most compact features are all from the blue-shifted side of the shell. We find no high brightness red-shifted maser spots. Our wide-band data confirm the observation by Sivagnanam et al.\\ (\\cite{SivaAmpl}) that the most blue-shifted 1667~MHz spot matches one of the 1665 MHz spots exactly. This spot at $-20.6$ km/s also seems the most persistent. We managed to follow this 1667~MHz and corresponding 1665~MHz features for over 4 years. The size of the most blue-shifted spot is consistent with the estimated size of the stellar radio-sphere. Finally, and most importantly, we have been able to compare the radio and optical positions. The position of the most blue-shifted maser spot has been shown to match the Hipparcos optical position within the errors. However, contrary to what would be expected, the most blue-shifted spot is not always the brightest or most compact; other spots at different velocities are dominant at different epochs. And not just the most blue-shifted 1667~MHz maser spot coincides with a 1665~MHz spot; almost $75\\%$ of the other bright spots coincide with one as well. Furthermore, it should have been possible to describe the motion of the $-20.6$ km/s spot, presumably fixed on the stellar position, with a simple trajectory. So, although our data is consistent with the amplified stellar image theory, the residuals of the fit may be indicative of more complex kinematics. A fixed maser spot could still show some random motion, as it drifts over the large surface of the radio-photosphere. The compact maser spot is a result of high beaming so small motions could result from minor changes of the maser path length due to variations in the pumping mechanism. This could also explain some of the relative motion between the spots. A correlation between stellar phase and these motions however cannot be determined. One other possibility is that U~Her itself does not follow a simple trajectory because it is part of a binary or multiple system. All these mechanisms could give rise to residuals of the order of those observed. ($3.7$, $2.5$ mas rms) Another issue that remains is the high surface brightness of the multiple maser spots. The most blue-shifted spot can be explained by amplified emission from the stellar radio-photosphere, but it is not always the brightest spot. The brightness of these other spots cannot easily be explained by normal self-amplifying masering regions. Since they are also only seen at the blue-shifted side of the shell, and since they move along with the star, they can not be amplified sources from beyond the shell. They lie too far from the star to be amplifying part of a very extended stellar radio-photosphere. At least two other spots are too bright to be self-amplifying. Should U~Her really be a multiple system, the high brightness spots could be the other components, amplified like U~Her itself. The motion of a multiple system should also be visible in the separation of the spots at different epochs. There might be an indication of such systematic motion between two of the spots. The occurrence of multiple bright spots is a subject of further study in a larger sample of OH-masering stars and will also be studied in the 1612 MHz maser transition. Although the precise nature of the additional compact blue-shifted emission features remains debatable, amplified stellar emission seems indeed to be the cause of the compact most blue-shifted maser spot. We are able to measure a proper motion derived from the VLBI maser spot which is consistent with the results from the Hipparcos satellite. The fit to the observed trajectories improves significantly when a parallax is included. The distance of $189 (+123, -54)$ pc determined with VLBI from a fit to the most blue-shifted OH spot is significantly smaller than the Chapman et al.\\ \\cite{Chapman} $P-L$ estimate of 385 pc. It is somewhat smaller than the revised $P-L$ estimate of 280 pc by Alvarez \\& Mennesier (\\cite{AlMe}). {\\it Acknowledgments:} We thank Craig Walker for help verifying the correlator model used, Jim Brauher for carrying out some initial data reduction and Jean-Francois Lestrade and Bob Campbell for providing software to check our fitting procedures. This project is supported by NWO grant 614-21-007." }, "0003/astro-ph0003466_arXiv.txt": { "abstract": "We report multifrequency radio observations of \\groj1655 obtained with the Australia Telescope Compact Array, the Molonglo Observatory Synthesis Telescope and the Hartebeesthoek Radio Astronomy Observatory at the time of the major hard X-ray and radio outbursts in 1994 August-September. The radio emission reached levels of the order of a few Jy and was found to be linearly polarized by up to 10\\%, indicating a synchrotron origin. The light curves are in good agreement with those measured with the VLA, but our closer time sampling has revealed two new short-lived events and significant deviations from a simple exponential decay. The polarization data show that the magnetic field is well ordered and aligned at right angles to the radio jets for most of the monitoring period. The time evolution of the polarization cannot be explained solely in terms of a simple synchrotron bubble model, and we invoke a hybrid `core-lobe' model with a core which contributes both synchrotron and free-free emission and `lobes' which are classical synchrotron emitters. ", "introduction": "The X-ray transient \\groj1655 (Nova Sco 1994) was first detected with the {\\em Burst and Transient Source Experiment} (BATSE) on board the Compton Gamma-Ray Observatory on 1994 July 27 (\\cite{nan94}). Significant flux was observed up to 200 keV and the source reached a luminosity of 1.1 Crab (20--100 keV) on Aug 1. \\groj1655 remained in outburst until about Aug 15, and after a period of quiescence flared again on Sept 6 (\\cite{harmon95}). The detection of a steeply-rising radio counterpart was first reported by Campbell-Wilson \\& Hunstead (1994a) following observations with the Molonglo Observatory Synthesis Telescope (MOST) at 843~MHz on 1994 Aug 6 and 11. The flux density continued to increase, reaching 4.2 Jy on 1994 Aug 14 and 5.5 Jy on 1994 Aug 15 (\\cite{cw&h94b}). High resolution radio observations using the Very Large Array (VLA), the Very Long Baseline Array (VLBA) (Hjellming \\& Rupen 1995, hereafter HR95) and the Southern Hemisphere VLBI Experiment (SHEVE) array (\\cite{tin95}) showed repeated episodes of relativistic ejections. Three major ejection events on TJD~9577.5, 9584 and 9596 (TJD defined as JD $-$ 2440000.5) were observed with the VLA and three on TJD~9574$\\pm$1, 9605$\\pm$2 and 9668$\\pm$5 were recorded with the VLBA (\\cite{h&r95}). The ejection velocity, after correction for inclination to the line of sight, was inferred to be $v\\sim0.92c$ in the rest frame of \\groj1655, at a position angle of $47^{\\circ}\\pm1^{\\circ}$ (\\cite{h&r95}); wiggles about the jet axis were interpreted as precession with a period of $3.0\\pm0.2$d. Photometric observations showed \\groj1655 to be a high-inclination binary system (\\cite{bailyn95}). Subsequent optical observations led to the determination of a precise mass for the primary, $M_1=7.02\\pm0.22\\,M_{\\odot}$ (\\cite{o&b97}), which is well above the theoretical upper limit for a neutron star and direct evidence for it being a black hole. In addition, Soria et al.\\ (1998) found a 95\\% confidence limit of $M_1\\,>\\,5.1\\,M_{\\odot}$ for the mass of the primary based on measurements of velocity variations in the He~{\\sc ii} disk emission lines, thought to reflect the orbital motion of the primary, and hence confirming \\groj1655 to be a black hole binary. The mass of the secondary, which is classified as {\\rm F}3 {\\rm IV}--{\\rm F}6 {\\rm IV}, is $M_2=2.34\\pm0.12\\,M_{\\odot}$, and the spectroscopic period of the binary is $P=2^d.62157\\pm0^d.00015$ (\\cite{o&b97}). During the 1994 ejection events, the radio outbursts from \\groj1655 were monitored in the southern hemisphere by the Molonglo Observatory Synthesis Telescope (MOST), the Australia Telescope Compact Array (ATCA), and the Hartebeesthoek Radio Astronomy Observatory (HartRAO). We discuss below the results of these observations. Sections \\ref{obs+dr} and \\ref{lightcurves} describe the MOST, ATCA and HartRAO observations and the resulting light curves. In Section \\ref{linpol} we present the linear polarization data from the ATCA, and we discuss the evolution of the radio spectra in Section \\ref{spectralstuff}. We briefly compare our results with those of HR95 in Section \\ref{compare}. Finally, we interpret the polarization data in terms of a `core-lobe' model in Section \\ref{physint}. ", "conclusions": "\\label{summary} As Figure~\\ref{fig-most+batse} shows, there is a relationship between the hard X-ray and radio emission from \\groj1655. The ejection episodes traced by the VLBA both originated during enhanced activity in the hard X-rays, and preceded the radio outbursts recorded with the MOST, suggesting a connection between activity near the event horizon of the black hole and the production of relativistic electrons. One possible explanation for the decline in intensity of the radio outbursts with time could be that the first ejection occurred in an environment that was relatively undisturbed, whereas the subsequent outbursts will have taken place in an environment already disrupted by previous activity (\\cite{hjell96}). The implications of the X-ray/radio correlations have been discussed elsewhere (e.g. \\cite{harmon95}), with a general consensus that the hard X-rays may be indicating enhanced accretion near the black hole which, through processes still not well understood, triggers the formation of relativistic radio jets. In general the ATCA, MOST and HartRAO flux density light curves agree well with the VLA light curves and ejection epochs reported in HR95. However, the better time sampling in Figures \\ref{fig-most+rao} and \\ref{fig-atca} reveals the presence of short-lived events that were not recorded by the VLA or VLBA, and shows that the light curves do not decay as smooth exponentials. The radio spectra, especially the linear polarization spectra, have proved to be valuable diagnostics of the timing of plasmon ejection events and their subsequent evolution. The ATCA polarization data show that the magnetic field is aligned at right angles to the radio jets (jet ${\\rm PA} = 47 \\pm 1^{\\circ}$), except towards the end of the monitoring period when core contributions may have become important. The rotation measure initially shows a contribution local to the \\groj1655 system, but after TJD 9582.5, the rotation measure is roughly constant at $\\sim$60~rad~m$^{-2}$ which must correspond to the interstellar value. After examining the time evolution of the total and polarized flux density of \\groj1655 we conclude that there are specific aspects of the behavior which cannot be explained by the simple synchrotron bubble model. We therefore invoke a hybrid `core-lobe' model, with a core which emits by non-thermal (or maybe thermal) free-free emission and lobes which are classical synchrotron emitters. We suggest that a similar model may apply to the other Galactic superluminal jet X-ray binary, GRS 1915+105." }, "0003/astro-ph0003185_arXiv.txt": { "abstract": "The recent discovery that the close-in extrasolar giant planet, HD209458b, transits its star has provided a first-of-its-kind measurement of the planet's radius and mass. In addition, there is a provocative detection of the light reflected off of the giant planet, $\\tau$ Boo b. Including the effects of stellar irradiation, we estimate the general behavior of radius/age trajectories for such planets and interpret the large measured radii of HD209458b and $\\tau$ Boo b in that context. We find that HD209458b must be a hydrogen-rich gas giant. Furthermore, the large radius of a close-in gas giant is not due to the thermal expansion of its atmosphere, but to the high residual entropy that remains throughout its bulk by dint of its early proximity to a luminous primary. The large stellar flux does not inflate the planet, but retards its otherwise inexorable contraction from a more extended configuration at birth. This implies either that such a planet was formed near its current orbital distance or that it migrated in from larger distances ($\\geq$0.5 A.U.), no later than a few times $10^7$ years of birth. ", "introduction": "The recent indirect detections of extrasolar giant planets (EGPs) by Doppler spectroscopy have taught us that such planetary systems can be very much unlike our own. To date, nearly $\\sim$30 EGPs have been discovered around stars with spectral types from M4 to F7 ({\\it e.g.}, \\cite{mq95}; \\cite{marcy99a}; \\cite{fischer99}). The planets themselves have minimum masses (\\mp $\\sin (i)$) between $\\sim$0.42 \\mj and $\\sim$10 \\mj, orbital semi-major axes from $\\sim$0.042 A.U. to $\\sim$3.8 A.U., and eccentricities as high as $\\sim$0.71. Such variety vastly expands the parameter space within which both theorists and observers must operate in defining the physical character of extrasolar planetary systems. The most interesting, unexpected, and problematic subclass of EGPs are those found within $\\sim$0.1 A.U. of their primaries, 50--100 times closer than Jupiter is to our Sun. At such orbital distances, due to prodigious stellar irradiation alone, an EGP can have an effective temperature (\\teff) greater than 600 K. Indeed, the EGPs HD187123b, HD209458b, $\\tau$ Boo b, HD75289b, 51 Peg b, $\\upsilon$ And b, and HD217107b likely all have \\teffs above 1000 K. This is to be compared with \\teffs for Jupiter and Saturn of 125 K and 95 K, respectively. The minimum masses of these close-in EGPs imply that they occupy the high-\\teff/low-gravity corner of parameter space for which the compositions and composition profiles may be unique (\\cite{seager98}; \\cite{gouk99}). However, high stellar fluxes on a close-in EGP can have profound structural consequences for the planet. In particular, stellar insolation can be responsible for maintaining the planet's radius at a value 20\\% to 80\\% larger than that of Jupiter itself (\\cite{guillot96}). This prediction has recently been verified by the observation of the transit by HD209458b of its primary. The depth of the transit yields values for its radius that range from $\\sim$1.27 \\rj (\\cite{char99b}) to $\\sim$1.7 \\rj (\\cite{henry99}), with a best value near $\\sim$1.4 \\rj (\\cite{mazeh99}). Importantly, since HD209458b transits its primary, astronomers can derive $\\sin (i)$, from which the planet's mass can be directly determined. This is a major advance in the emerging study of extrasolar planets. In this paper, we focus on the HD209458 system and what broadly can be concluded theoretically from these new transit data. We do not provide detailed models from which one can extract bulk or atmospheric composition. Rather, we show that the large radius of HD209458b is a consequence of the retardation by stellar irradiation of the otherwise natural cooling of the convective core of the planet and that such a radius requires that the planet did not dwell for long, if at all, at large orbital distances after its formation. Hence, the large radius of a close-in EGP is not due to the thermal expansion of its atmosphere, but to the high residual entropy that remains throughout its bulk as a consequence of its early proximity to a luminous primary. Recently, Cameron \\etal (1999), using spectral deconvolution, claim to have seen $\\tau$ Boo b in reflection. To investigate this exciting possibility, we calculate a range of theoretical radii for the $\\tau$ Bo\\\"otis planet. As we show, the Cameron \\etal value of 1.6--1.8 \\rj for $\\tau$ Boo b's radius, if verified, is a challenge to the still embryonic theory of massive, close-in EGPs. Nevertheless, with the discovery of a transiting planet, HD209458b, with the maturation of the technique of spectral deconvolution, and with the anticipated development of adaptive optics and interferometry for the direct study of extrasolar planets (\\cite{angel94}), we are clearly entering a new phase in extrasolar planetary research. ", "conclusions": " \\begin{enumerate} \\item HD209458b is a real object, made predominantly of hydrogen. \\item HD209458b's radius is a consequence of the retardation of contraction by stellar irradiation and is not due to atmospheric expansion by stellar heating. \\item A large radius such as that of HD209458b requires early proximity to its central star. \\end{enumerate} Curiously, given L$_{\\ast}$, $a$, and \\rp, HD209458b's total luminosity is $\\sim 1.5\\times 10^{-4}$ \\lo, about twice that of a star with $\\sim$100 times the mass at the very bottom of the stellar main sequence. Given the large inferred mass of $\\tau$ Boo b, its large radius is less easy to explain theoretically. However, the inherent difficulties of close-in EGP modeling may yet be responsible for theoretical surprises of a qualitative nature." }, "0003/astro-ph0003228_arXiv.txt": { "abstract": "Examining the nature of the relative clustering of different galaxy types can help tell us how galaxies formed. To measure this relative clustering, I perform a joint counts-in-cells analysis of galaxies of different spectral types in the Las Campanas Redshift Survey (LCRS). I develop a maximum-likelihood technique to fit for the relationship between the density fields of early- and late-type galaxies. This technique can directly measure nonlinearity and stochasticity in the biasing relation. At high significance, a small amount of stochasticity is measured, corresponding to a correlation coefficient $r\\approx 0.87$ on scales corresponding to 15 $h^{-1}$ Mpc spheres. A large proportion of this signal appears to derive from errors in the selection function, and a more realistic estimate finds $r\\approx 0.95$. These selection function errors probably account for the large stochasticity measured by \\citet{tegmark99a}, and may have affected measurements of very large-scale structure in the LCRS. Analysis of the data and of mock catalogs shows that the peculiar geometry, variable flux limits, and central surface-brightness selection effects of the LCRS do not seem to cause the effect. ", "introduction": "\\label{c5_motiv} Galaxies of different morphologies have different spatial distributions, as first noted by \\citet{hubble36a}. Early-type galaxies, such as ellipticals and S0s, are highly clustered and account for 90\\% of galaxies in the cores of rich clusters; late-type galaxies, such as spirals and irregulars, are less clustered and make up 70\\% of galaxies in the field (\\citealt{dressler80a}; \\citealt{postman84a}; \\citealt{whitmore93a}). A general way of expressing the relationship between the density fields of galaxies of different types on any scale $R$ is with the joint probability distribution $f(\\delta_e, \\delta_l)$; that is, the probability at any location of finding an overdensity $\\delta_e$ of early-type galaxies and an overdensity $\\delta_l$ of late-type galaxies. This quantity is analogous to the joint probability distribution of galaxy and mass density introduced by \\citet{dekel99a}. The traditional method of measuring the properties of $f(\\delta_e,\\delta_l)$ has been to compare the amplitude of the fluctuations in each density field, using the correlation functions or the power spectra. By these measures, the level of fluctuations in ellipticals is stronger than that of spirals by a factor of 1.3--1.5 (\\citealt{davis76a}; \\citealt{giovanelli86a}; \\citealt{santiago92a}; \\citealt{loveday96a}; \\citealt{hermit96a}; \\citealt{guzzo97a}). These relative clustering properties are successfully reproduced by current models of galaxy formation. For example, \\citet{blanton99a} examined hydrodynamical simulations and identified galaxies as dense, rapidly cooling clumps of gas. Older galaxies, which correspond to early-types, turned out to be clustered more strongly than younger galaxies, which correspond to late-types. The relative bias factor, $b\\equiv\\sigma_e/\\sigma_l$, where $\\sigma^2\\equiv\\avg{\\delta^2}$, is approximately 1.5 between these populations. Semi-analytic models, which follow halos in collisionless $N$-body simulations and use simple models for star-formation and feedback inside each halo and for the effect of halo mergers, find similar results (\\citealt{somerville99a}). However, \\citet{blanton99a} also found that there was considerable scatter between the two density fields; that is, that there was not a one-to-one relationship between the number of old galaxies in a region to the number of young galaxies. A measure of this scatter is the correlation coefficient $r\\equiv\\avg{\\delta_e\\delta_l}/\\sigma_e\\sigma_l$ between the early-type overdensity field $\\delta_e$ and the late-type overdensity field $\\delta_l$. In the simulations, $r \\sim 0.5$--$0.8$. On the other hand, the semi-analytic models of \\citet{somerville99a} find that the correlation coefficient $r\\sim 0.9$; that is, they find very little scatter. The essential difference between the predictions of this model and that of the hydrodynamic model is the effect of the temperature history of the gas in the hydrodynamic simulations and its relationship with large-scale structure. Thus, one can use the correlation coefficient between different galaxy types to distinguish between these models of galaxy formation. Measuring this scatter requires a probe of $f(\\delta_e|\\delta_l)$ which differs from the traditional statistics mentioned above. For example, two completely unrelated density fields ($r=0$) can have the same correlation function. To detect the scatter, one must compare the density fields point by point, not just compare the overall levels of the fluctuations. A direct approach to constraining the properties of $f(\\delta_e,\\delta_l)$ is to measure the related joint probability distribution $P(N_e, N_l)$ of finding $N_e$ early-type and $N_l$ late-type galaxies in a single cell of size $R$. After all, this latter probability is simply $f(\\delta_e, \\delta_l)$ convolved with Poisson distributions. If one notes that \\begin{equation} \\label{c5_bayes} f(\\delta_e,\\delta_l) = f(\\delta_l|\\delta_e) f(\\delta_e), \\end{equation} then one can write \\begin{eqnarray} \\label{c5_prob2} &P(N_e, N_l)& = \\int d\\delta_e \\frac{{N}_{e,\\mathrm{exp}}^{N_e} (1+\\delta_e)^{N_e}}{N_e!} e^{-{ N_{e,\\mathrm{exp}}}(1+\\delta_e)} f(\\delta_e) \\cr &&\\times \\int d\\delta_l \\frac{{N}_{l,\\mathrm{exp}}^{N_l} (1+\\delta_l)^{N_l}}{N_l!} e^{-{ N_{l,\\mathrm{exp}}}(1+\\delta_l)} f(\\delta_l| \\delta_e), \\end{eqnarray} where ${N}_{e,\\mathrm{exp}}$ and ${N}_{l,\\mathrm{exp}}$ are the average number of galaxies of each type expected in a cell of a given volume (and given selection criteria). Naturally, one can integrate Equation (\\ref{c5_prob2}) over $N_l$ to obtain: \\begin{equation} \\label{c5_prob1} P(N_e) = \\int d\\delta_e \\frac{{N}_{e,\\mathrm{exp}}^{N_e} (1+\\delta_e)^{N_e}}{N_e!} e^{-{ N_{e,\\mathrm{exp}}}(1+\\delta_e)} f(\\delta_e), \\end{equation} the probability distribution of counts of early-type galaxies. As I show below, one can use Equation \\ref{c5_prob2} to devise a maximum likelihood method to fit for $f(\\delta_l| \\delta_e)$, and Equation \\ref{c5_prob1} to fit for $f(\\delta_e)$. Equation (\\ref{c5_prob2}) provides a direct probe of the relationship between galaxy density fields $f(\\delta_l|\\delta_e)$, including its nonlinearity and scatter. Consider for contrast the work of \\citet{benoist99a}, who infer nonlinearity in the relative bias of galaxies of different luminosities in the Southern Sky Redshift Survey from the scale dependence of the higher-order moments of the density fields. Using the same data, one could instead compare the observed $P(N_e, N_l)$ to models and detect nonlinearity more directly. Furthermore, the joint distribution contains more information than a comparison of the moments of each density field. While moments of the density field yield averaged information about the fluctuations, $P(N_e, N_l)$ yields a point-by-point comparison of two density fields, which can be much more powerful (\\citealt{santiago92a}). For instance, one can use this comparison to determine whether the effects detected by \\citet{benoist99a} are actually due to nonlinearity (as they propose), or perhaps due properties of the scatter in the relationship between low luminosity and high luminosity galaxies. In this paper, I perform an maximum-likelihood analysis of this joint distribution for different spectral types of galaxies in the Las Campanas Redshift Survey (LCRS), using cells with volumes approximately equal to that of cubes 25 $h^{-1}$ Mpc on a side. A similar analysis has been performed on the LCRS by Tegmark \\& Bromley (1999; hereafter TB99), to which I will compare my results throughout. Essentially, their method calculates the second moments of $f(\\delta_e,\\delta_l)$, namely $\\sigma_l^2\\equiv\\avg{\\delta_l^2}$, the variance of the density field of late-type galaxies, $\\sigma_e^2\\equiv\\avg{\\delta_e^2}$, the variance of early-type galaxies, and $r\\equiv\\avg{\\delta_e\\delta_l}/\\sigma_e\\sigma_l$, the correlation coefficient between the two fields, which is unity if the fields are perfectly correlated and zero if the fields are completely uncorrelated. As I will show below, calculating second moments is probably not sufficient on the scales which TB99 probe ($\\sim 5$--$10$ $h^{-1}$ Mpc), because it does not correctly account for the fact that density fields cannot be negative. For this reason, the resulting $r$ may overestimate the degree of scatter in the relationship between the two fields. On larger scales where $\\sigma\\ll 1$, these differences would of course be much reduced. Furthermore, the moments method also yields no information on how nonlinear the relationship between the density fields of the two galaxy types is, which the maximum likelihood method described here will. Finally, I have detected important effects concerning the galaxy selection function which affect the results of this analysis and have consequences for the interpretation of TB99 and other measurement of large-scale structure in the LCRS. This paper is organized as follows. In Section \\ref{c5_data}, I describe the details of the LCRS. In Section \\ref{c5_selfunc}, I describe the method used to calculate the selection function for the survey. In Section \\ref{c5_method}, I describe in detail the maximum likelihood analysis of the counts-in-cells. In Section \\ref{c5_results}, I present the results of fitting for the relationship between the different galaxy types and demonstrate the presence of systematic errors in the selection of galaxies in this survey. In Section \\ref{c5_mock}, I describe the results of an analysis of mock catalogs, in order to quantify a number of possible statistical and systematic effects as well as to evaluate the importance of cosmic variance. I conclude in Section \\ref{c5_conclusions}. ", "conclusions": "\\label{c5_conclusions} I have presented a straightforward maximum likelihood method to determine the relationship between the density fields of different galaxies types on a point-by-point basis by looking at the joint counts-in-cells distribution $P(N_e, N_l)$. Using mock catalogs, I have demonstrated the reliability of the method. I have applied the method to the LCRS in an attempt to constrain the nature of the segregation of different galaxy spectral types (as classified by \\citealt{bromley98a}). At most a small amount of stochasticity affects the relationship between early- and late-type galaxies in the LCRS, corresponding to $r\\sim 0.87$, a larger correlation coefficient than found using the simple moments method of TB99. In addition, it is likely that even this result is low because of poorly understood selection effects in the survey, and that the true value of $r$ is closer to $\\sim 0.95$. In either case, the large scatter predicted by \\citet{blanton99a} from hydrodynamic simuations does not seem to exist, and the results are more consistent with the semi-analytic predictions of \\citet{somerville99a}. It is not clear yet what the implications of this result are, but there are at least three possibilities. First, because the survey is selected in the $R$ band and is surface-brightness limited, there may not be a sufficient range of galaxy types represented to reveal the predicted stochasticity. The fact that the relative bias $b$ between early- and late-type galaxies is also smaller than predicted (1.2 instead of 1.5) is consistent with this explanation. Second, since the simulations of \\cite{blanton99a} are low resolution and cannot resolve galactic disks, it may be that the simulations are not modelling important effects on subgrid scales which would considerably reduce the stochasticity. Third and most interesting (though probably least likely), is the possibility that the fundamental principles behind the way galaxy formation is approximated in the simulations are flawed, and need to be revised. Improved simulations and the analysis of new, larger, and more complete redshift surveys such as the SDSS will help answer these questions. In addition to the main result, I have found suspicious behavior of the selection function derived for the sample (both my own and that of \\citealt{bromley98a}). A thorough investigation of possible causes of these errors, using the data itself as well as mock catalogs, has turned up no likely cause of this effect, including surface-brightness selection effects, the use of isophotal magnitudes, and errors in the $K$-correction. On the other hand, it is possible that some of the mock catalog experiments presented here are misleading because the model I used for galaxy profiles was inadequate (for instance, if I used an inaccurate distribution of galaxy sizes). Analysis of larger surveys with better quality images and spectra, such as the SDSS, may thus be more useful than the mock catalogs in understanding the effect. I must note that the distinct, though unlikely, possibility remains that the low redshift portion of the LCRS is indeed an unusual section of the universe, either due to a rapid evolution of galaxy properties between $z\\approx 0.2$ and today or because of some peculiar local phenomenon. The inadequacy of the selection function may have consequences for other results based on the LCRS. First, I have shown here that the low correlation coefficients measured by TB99 may be in doubt. Second, the excess large-scale power in this survey claimed by \\citet{landy96a} may be due to this effect. In fact, the largest amplitude wave in the survey that those authors detect is in the ``outward,'' redshift direction, which might indicate that redshift dependent selection effects could be contaminating their results; on the other hand, they also detect large waves tangent to the redshift direction, which might not be so readily explained. In any case, the method presented here is applicable to any comparison of counts-in-cells of different galaxy populations. It may be most useful in surveys which are volume-limited and have simpler geometries. Such surveys would also make it easier to explore the scale-dependence of the relative bias of galaxies; this task is difficult in the LCRS, since looking at larger scales forces one to change the geometry of one's cells, which as I have shown affects the results. In particular, the Sloan Digital Sky Survey (SDSS; \\citealt{gunn95a}) and the Two-Degree Field (2DF; \\citealt{colless98a}) would allow one to make powerful tests of the nature of morphological segregation. It is possible, of course, to compare the galaxy densities in different surveys using this method (\\citealt{seaborne99a}). For example, one might use the future $K$-selected redshift survey based on the Two-Micron All Sky Survey (2MASS; \\citealt{beichman98a}) to compare in the appropriate volume to the SDSS or 2dF. In conclusion, the details of morphological segregation contain much information about how galaxies formed. This paper has attempted to extract some of this information by measuring the stochasticity in the relative clustering of galaxy types. Future redshift surveys and more sophisticated galaxy formation models will be able to make much more powerful and informative tests." }, "0003/astro-ph0003191_arXiv.txt": { "abstract": "We present a low-flux extension of the X-ray selected ROSAT Brightest Cluster Sample (BCS) published in Paper I of this series. Like the original BCS and employing an identical selection procedure, the BCS extension is compiled from ROSAT All-Sky Survey (RASS) data in the northern hemisphere ($\\delta \\geq 0^{\\circ}$) and at high Galactic latitudes ($|b| \\geq 20^{\\circ}$). It comprises 100 X-ray selected clusters of galaxies with measured redshifts $z \\leq 0.3$ (as well as seven more at $z> 0.3$) and total fluxes between $2.8\\times 10^{-12}$ erg cm$^{-2}$ s$^{-1}$ and $4.4\\times 10^{-12}$ erg cm$^{-2}$ s$^{-1}$ in the 0.1--2.4 keV band (the latter value being the flux limit of the original BCS). The extension can be combined with the main sample published in 1998 to form the homogeneously selected extended BCS (eBCS), the largest and statistically best understood cluster sample to emerge from the ROSAT All-Sky Survey to date. The nominal completeness of the combined sample (defined with respect to a power law fit to the bright end of the BCS $\\log N-\\log S$ distribution) is relatively low at 75 per cent (compared to 90 per cent for the high-flux sample of Paper I). However, just as for the original BCS, this incompleteness can be accurately quantified, and thus statistically corrected for, as a function of X-ray luminosity and redshift. In addition to its importance for improved statistical studies of the properties of clusters in the local Universe, the low-flux extension of the BCS is also intended to serve as a finding list for X-ray bright clusters in the northern hemisphere which we hope will prove useful in the preparation of cluster observations with the next generation of X-ray telescopes such as {\\sl Chandra} or {\\sl XMM-Newton}. An electronic version of the eBCS can be obtained from the following URL: {\\em www.ifa.hawaii.edu/$\\sim$ebeling/clusters/BCS.html.} ", "introduction": "Until recently, the compilation of large statistical samples of clusters of galaxies was a task accomplishable only at optical wavelengths where photographic plates provide both all-sky coverage and sufficient depth to detect clusters at redshifts of $z\\la 0.3$ (e.g., Abell 1958, Zwicky et al.\\ 1961--1968, Abell, Corwin \\& Olowin 1989). Only with the completion of the {\\sl ROSAT} All-Sky Survey (RASS) in 1991 (Voges 1992, Tr\\\"umper 1993) did unbiased large compilations of X-ray detected clusters become a feasible alternative. To date, three X-ray flux limited cluster samples have been published from RASS data. The all-sky sample of the 242 X-ray Brightest Abell-type Clusters (XBACs) of Ebeling et al.\\ (1996) was the first statistical sample of X-ray bright clusters to emerge from the RASS. However, although X-ray flux limited, the XBACs sample is, by design, limited to Abell clusters and thus still affected by the biases inherent in optical cluster surveys. The other two large-scale RASS cluster samples are truly X-ray selected though: the ROSAT Brightest Cluster Sample (BCS, Ebeling et al.\\ 1998, Paper I) comprises 203 X-ray selected clusters in the northern hemisphere, and the RASS1 Bright Sample (RASS1-BS, De Grandi et al.\\ 1999) consists of 130 such clusters in the southern hemisphere. A fourth RASS cluster sample covering most of the southern extragalactic sky is under compilation (B\\\"ohringer et al, in preparation). In the following we briefly summarize the key features of the BCS. ", "conclusions": "" }, "0003/astro-ph0003472_arXiv.txt": { "abstract": "{Compactness of some stars is explained if they are strange stars (SS) as shown by Dey et al. (1998) (D98) and Li et al. (1999a). One of these compact star candidates is the SAX~J1808.4$-$3658 (SAX in short) believed to be an important link in the genesis of radio pulsars. SS have also been suggested for bursting X-ray pulsars (GRO~J1744$-$28, Cheng et al. 1998), from quasi-periodic oscillations (QPO) of X-ray binaries (4U~1728$-$34, Li et al. 1999b) and from peculiarity of properties of radio pulsars (PSR~0943$+$10, Xu et al.1999; Kapoor et al. 2000). We now extend the calculation to include high temperatures upto $T$~=~70~MeV~$\\sim~8\\times 10^{11}~^o$K and find that the nature of the mass (M) and radius (R), derived from astrophysical data, is still retained. The entropy is calculated and matches onto that calculated from hadronic models thus supporting the idea that the quark-hadron transition may be continuous.} ", "introduction": "A calculation for cold strange stars (D98) enabled us to draw conclusions about chiral symmetry restoration in QCD when the EOS was used to get SS fitting definite mass-radius (MR) relations (D98; Li et al. 1999a; Li et al. 1999b; Ray et al. 2000b) . The empirical MR relations were derived from astrophysical observations like luminosity variation and some properties of quasi-periodic oscillations from compact stars. The calculations are compared to these stars which emit X-rays, generated presumably due to accretion from their binary partner. During the genesis of these stars higher $T$ may be encountered and in this paper we deal with a generalized case of an object at a uniform $T$. We show at upto $T$ = 70 MeV a self sustained system can be supported by the parameter set eos1 (D98) (called SS1 in Li et al. 1999a). The changes in star masses are shown in the range of $T$ mentioned above. The conclusions of D98 and Li et al. (1999a, 1999b) are still valid for the finite $T$ cases. Secondly, parameters of the single particle potential at finite $T$ are tabulated. In particular the entropy is studied and this can be compared with hadronic models. The comparison shows that the entropy may indeed be continuous supporting the idea of a continuous phase transition between hadrons and quarks. ", "conclusions": "" }, "0003/astro-ph0003158_arXiv.txt": { "abstract": "As part of a survey of \\HI\\ $\\lambda$21-cm emission in the Southern Milky Way, we have detected two large shells in the interstellar neutral hydrogen near $l=279^{\\circ}$. The center velocities are +36 and +56 km s$^{-1}$, which puts the shells at kinematic distances of 7 and 10 kpc. The larger shell is about 610 pc in diameter and very empty, with density contrast of at least 15 between the middle and the shell walls. It has expansion velocity of about 20 km s$^{-1}$ and swept up mass of several million solar masses. The energy indicated by the expansion may be as high as $2.4 \\times 10^{53}$ ergs. We estimate its age to be 15 to 20 million years. The smaller shell has diameter of about 400 pc, expansion velocity about 10 km s$^{-1}$ and swept up mass of about $10^6$ solar masses. Morphologically both regions appear to be shells, with high density regions mostly surrounding the voids, although the first appears to have channels of low density which connect with the halo above and below the \\HI\\ layer. They lie on the edge of the Carina arm, which suggests that they may be expanding horizontally into the interarm region as well as vertically out of the disk. If this interpretation is correct, this is the first detection of an \\HI\\ chimney which has blown out of both sides of the disk. ", "introduction": "\\label{sec:intro} Studies of external galaxies, in particular, recent studies of the Large and Small Magellanic Clouds, indicate large populations of shells and supershells, which dominate the structure of the interstellar medium (ISM) (Staveley-Smith \\etal 1997; Kim \\etal 1998). By injecting large quantities of energy into the ISM, these shells reshape galaxies on size scales of tens to hundreds of parsecs and trigger new star formation. In our own Galaxy, surveys have found many small shells and ``worms'', (eg. Heiles 1979, 1984; Koo, Heiles, \\& Reach 1992 (KHR)) but the number of large supershells and chimneys which have been identified is still relatively small (Normandeau, Taylor, \\& Dewdney 1996; Heiles 1998). These exceptionally large structures are most often identified as dramatic voids in the Galactic neutral hydrogen, observed with the \\HI\\ line at 1420 MHz. Unfortunately, in the inner Galaxy, where they are most likely to occur, they prove difficult to detect due to distance ambiguities. As a result, our knowledge of how dramatically the Galaxy has been shaped by shells is limited. These \\HI\\ voids range in size from tens of parsecs to kiloparsecs, and are found with a variety of morphologies from nearly spherical to chimney-like. The dominant paradigm suggests that these structures are caused by the combined pressures of stellar winds and sequential supernovae (SNe) in OB associations (Heiles 1984). It has also been suggested, however, that the largest of these structures, with energies in excess of $10^{53}$ ergs, may be caused by impacts of high velocity clouds (HVCs) with the Galactic disk (cf. Heiles 1984 for a discussion of both formation methods), or more recently, that they are the remnants of hypernovae and/or gamma ray bursts (Loeb \\& Perna 1998). Younger supershells are often associated with some ionized emission in the shell interior in the form of hot x-ray emitting gas, or an H$\\alpha$ emitting inter-rim (Points \\etal 1999). For the oldest supershells it is likely that the hot x-ray emitting medium has diffused and the massive stars of the OB association have expired, leaving only a evacuated region in the Galactic \\HI. For the largest, and therefore the oldest, of these shells, expansion can exceed the scale height of the \\HI\\ layer of the Galaxy. Such expansion will elongate along the axis perpendicular to the Galactic plane, as predicted by theories of expansion into a stratified medium (eg. Kompaneets 1960). In this case we expect to see chimneys where the polar regions of the shell become Rayleigh-Taylor unstable and break through into the Galactic halo, providing a source of ionized hydrogen and thermal support for the halo. The Galactic Plane near $l=280\\deg$, $v=+35$ \\kms\\ (LSR)\\footnote{All velocities are quoted with respect to the Local Standard of Rest (LSR).} is a dynamic place, with very dramatic brightness temperature fluctuations over relatively small scales, and the edge of the Carina arm. Positive velocities in this direction are beyond the solar circle, corresponding to a unique distance. As a result, it is somewhat easier to unravel the Galactic structure in this region than it is in the inner Galaxy. The $l=280\\deg$ line-of-sight is tangent to the Carina spiral arm. The region between $l=275\\deg$ and $l=280\\deg$ and $v=+25$ \\kms\\ and $v=+50$ \\kms\\ is in between spiral arms, with the Carina arm towards greater longitudes and the Perseus arm towards lesser longitudes or higher velocities. In the \\HI\\ and CO longitude-velocity ({\\em l-v}) diagram the Carina arms forms a loop with the apex at $l=280\\deg$, $v=0$ \\kms. Towards greater longitudes the Carina arm is seen to extend along $v\\sim 35$ \\kms\\ to $l=330\\deg$ and beyond (Grabelsky \\etal 1987). At lesser longitudes the Vela supernova remnant dominates radio continuum and X-ray emission towards $l=265\\deg$. In this paper we report on the discovery of two large Galactic \\HI\\ shells near the Carina tangent. One, GSH 277+0+36, is centered at a galactic longitude of $l=277$\\deg, latitude of $b=0$\\deg\\, and velocity $v=+36$ \\kms\\ with an angular diameter of 5\\deg.5. The second, smaller shell, GSH 279+0+59 is centered on $l=280\\deg$, $b=0\\deg.1$, $v=+59$ \\kms\\ with an angular diameter of 2\\deg.7. We will explore the possibility that the shells are interarm voids as previously suggested (Grabelsky \\etal\\ 1987), and present arguments in favor of a shell interpretation. In section~\\ref{sec:obs} we describe the observations and analysis. In sections~\\ref{subsec:sh1-morph} and \\ref{subsec:sh1-props} we discuss the morphology and physical properties of GSH 277+0+36. In section~\\ref{subsec:sh2} we discuss the morphology and properties of GSH 279+0+59. In section~\\ref{subsec:otherwaves} we compare the \\HI\\ emission with other wavebands, including far-infrared, 2.4 GHz continuum, and CO. Finally, in section~\\ref{sec:disc} we discuss possible formation methods. ", "conclusions": "\\label{sec:concl} In conclusion, we have found two large HI shells in the outer Galaxy. The first and most dramatic, GSH 277+0+36, can be classified as a supershell on the basis of its large size and expansion energy. Prior interpretation of this large void as an interarm region now seems inappropriate on the basis of the supershell's chimney and shell-like morphology. The supershell most probably exists in the region between spiral arms, though it was not necessarily formed there. The strong arm-interam contrast previously noticed has undoubtedly been enhanced by the supershell edges. We find evidence for molecular clouds along the supershell's edges, indicating that star formation may have been initiated by the supershell's expansion. Because of the shell's unusual position between spiral arms, and its large formation energy requirements we have considered several formation theories for this shell. We have considered the conventional formation method of stellar winds and supernovae, an HVC collision with the Galactic disk, and finally we have raised the possibility that the shell formed in the Carina arm and expanded into the interarm region. We believe that the latter is the most likely scenario, as it decreases the energy requirements and is consistent with theories of Galactic structure which predict higher star formation rates and therefore higher supernova rates in the spiral arms. The second shell, GSH 280+0+59, though smaller than the first shell, is large by Galactic standards with $R_{sh} \\sim 220$ pc. It also appears to have blown out of the Galactic plane. While there is no definitive interaction between the two shells, it is possible that they may be interacting if one has a systemic velocity which departs from its local of standard of rest by $\\sim 20$ \\kms. They appear, however, to be distinct shells which presumably formed independently. The energy requirements for the smaller shell are much more reasonable, indicating that the shell could have been created by $\\sim 20$ supernovae, or equivalent stellar winds, over several million years. The effects of these shells on their local ISM is dramatic. Regardless of whether they are associated and whether or not GSH 277+0+36 formed in the Carina arm, they have significantly reshaped the large scale structure of the Galaxy in that region on the timescale of millions of years. Because other galaxies are so dramatically influenced by shells, supershells, and chimneys, it is reasonable to expect that the Milky Way has been similarly influenced. However, the catalogued shells and chimneys have not revealed the level of influence on the structure of the Milky Way as those seen in the Large and Small Magellenic Clouds. In addition, the relatively few chimneys seen cannot support the halo. We expect, therefore, that there are many more supershells and chimneys to be detected as we probe deeper into the Galaxy with the Southern Galactic Plane Survey. It is imperative to understanding the structure of the Galaxy that we have a complete catalog of supershells and chimneys, particularly in the inner Galaxy." }, "0003/astro-ph0003314_arXiv.txt": { "abstract": "Aligned grains provide one of the easiest ways to study magnetic fields in diffuse gas and molecular clouds. How reliable our conclusions about the inferred magnetic field depends critically on our understanding of the physics of grain alignment. Although grain alignment is a problem of half a century standing recent progress achieved in the field makes us believe that we are approaching the solution of this mystery. I review basic physical processes involved in grain alignment and show why mechanisms that were favored for decades do not look so promising right now. I also discuss why the radiative torque mechanism ignored for more than 20 years looks right now the most powerful means of grain alignment. ", "introduction": "Magnetic fields are extremely important for star formation, galactic feedback processes etc. and polarized radiation arising from absorption and emission by aligned grains provides an important means for studying magnetic field topology. However, the interpretation of polarimetry data requires clear understanding of processes of grain alignment and the naive rule of thumb that dust grains are aligned everywhere and with longer axes perpendicular to magnetic field may be misleading (see Goodman et al. 1995, Rao et al. 1998). Physics of grain alignment is deep and exciting. It is enough to say that its study resulted in a discovery of a few new solid state effects. However, let us start by recalling a few simple facts. The grain alignment in interstellar medium always happens in respect to magnetic field. It is fast Larmor precession of grains that makes magnetic field the reference axis. Note, that grains may align with their longer axes {\\it perpendicular} or {\\it parallel} to magnetic field direction. Similarly, magnetic fields may change their configuration and orientation in space (e.g. due to Alfven waves), but if the time for such a change is much longer than the Larmor period the alignment of grains {\\it in respect to the field lines} persists as the consequence of preservation of the adiabatic invariant. The alignment of grain axis is described by the Rayleigh reduction factor: \\begin{equation} R\\equiv \\langle G(\\cos^2\\theta) G(\\cos^2\\beta)\\rangle \\end{equation} where angular brackets denote ensemble averaging, $G(x) \\equiv 3/2 (x-1/3)$, $\\theta$ is the angle between the axis of the largest moment of inertia (henceforth the axis of maximal inertia) and the magnetic field $\\bf B$, while $\\beta$ is the angle between the angular momentum $\\bf J$ and $\\bf B$. One may see (e.g. Hildebrand 1988) that $R$ is directly related to the degree of polarization. To characterize $\\bf J$ alignment in grain axes and in respect to magnetic field, the measures $Q_X\\equiv \\langle G(\\theta)\\rangle$ and $Q_J\\equiv \\langle G(\\beta)\\rangle$ are used. Unfortunately, these statistics are not independent and therefore $R$ is not equal to $Q_J Q_X$ (see Roberge \\& Lazarian 1999). This considerably complicates the treatment of grain alignment. This review attempts to cover the recent advancements of our understanding of grain alignment and places them in the context of the earlier works done by giants of E. Purcell and L. Spitzer caliber. It happened that several times the problem of grain alignment seemed to be solved and theorists got satisfied. However, accumulation of new observational facts and deeper insights into grain physics caused the changes of paradigms. Thus in what follows we describe three periods of grain alignment theory. A more detailed treatment of various aspects of grain alignment the interested reader can find in earlier reviews (e.g. Hildebrand 1988, Roberge 1996, Lazarian, Goodman \\& Myers 1997). ", "conclusions": "Let us summarize what we learned about the dynamics of grain alignment. For a $10^{-5}$~cm grain in cold diffuse interstellar medium the fastest motion is the grain rotation, which happens on the time scale less than $10^{-4}$~s. The grain tumbling and rotation of angular velocity about $\\bf J$ happens on approximately the same time scale. The alignment of $\\bf J$ with the axis of maximal inertia happens as a matter of hours due to the very efficient nuclear relaxation. On the time scale of days $\\bf J$ rotates about $\\bf B$ due to its magnetic moment (Dolginov \\& Mytrophanov 1976), while gaseous damping time takes $t_{gas}\\sim 10^{5}$ years. An alignment mechanism is efficient if the alignment time is a fraction of $t_{gas}$ for thermally rotating grains, but it may be many $t_{gas}$ if grains rotate suprathermally. In the latter case the dynamics of crossovers is all-important. At the moment radiative torques look as the most promising means of aligning dust. Due to thermal trapping the Purcell alignment is suppressed. The superparamagnetic hypothesis looks OK (see Goodman \\& Whittet 1996), but the mechanism faces the problem with driving grain rotation. The same thermal trapping makes grain alignment less efficient in molecular clouds where grain rotational temperature approaches its body temperature. It is likely that the radiative torques are still required to drive grain rotation. The most challenging problem right now is to understand the radiative torque mechanism. For this purpose it is necessary to describe crossovers induced by radiative torques and include the recently discovered flipovers into existing codes. It looks necessary to understand why grains align (not always, but {\\it very} frequently) $\\bf J$ with $\\bf B$ when subjected to anisotropic radiation. My experiments with slightly irregular grains (using the code kindly provided to me by Bruce Draine) interacting with anisotropic monochromatic radiation made me believe that it is possible to get a theoretical insight into the underlying physics. However, whatever theory says, observational tests are necessary. Inversion of the polarimetric data (see Kim \\& Martin 1995) allows to find {\\it for different environments} the critical grain size starting with which grains are aligned. Comparing this size with predictions calculated for radiative torques should enable testing the mechanism. Whatever the success of the radiative torques, it is necessary to proceed with further development of alternative alignment mechanisms. Some of them, e.g. the mechanism of mechanical alignment is suspected to cause alignment at least in some regions (see Rao et al. 1998). Ward-Thompson et al. (2000) reported 850 $\\mu$m polarization from dense pre-stellar cores, where radiative torques should be inefficient. Could the grain larger than $a_c$ and aligned via modified Purcell mechanism (LD97) be responsible? Or should we apeal to Alfven waves or ambipolar diffusion? Further research will provide us with the answer. In general, the variety of Astrophysical conditions allows various mechanisms (see Lazarian, Goodman \\& Myers 1997) to have their niche. Clear understanding of grain alignment will make polarimetry much more informative. Although so far grain alignment theory was applied only to interstellar environments, it is clear that its potential is great for circumstellar and interplanetary studies (see Lazarian 2000)." }, "0003/astro-ph0003408_arXiv.txt": { "abstract": " ", "introduction": "Coherent radio Cherenkov emission is a remarkably effective method for detecting high energy particles. The history of the effect goes back to Jelly, who first asked whether cosmic ray air showers might produce a radio signal \\cite{Allan}. Askaryan \\cite{Askaryan} subsequently predicted a net charge imbalance in air showers, and coherent radio power scaling like the energy of the shower squared. Substantial radio emission from atmospheric electromagnetic cascades was observed more than 30 years ago. \\cite{Jelly, Allan}. Progress in ultra-high energy air showers has sparked renewed interest, and new observations of radio pulses have been reported recently \\cite{Rosner}. The current pilot project RICE \\cite{fmr, RICE} uses radio Cherenkov emission to detect $100 TeV$ and higher energy neutrinos in Antarctic ice. The radio Cherenkov signal is the most efficient known mechanism \\cite{fmr,mkr, RICE, Provorov, Price} for detecting neutrinos of $100TeV$ and above in solid media, yielding detection volumes of order $ 1km^{3}$ per radio detector for $PeV$ neutrinos on ice targets. At $PeV$ energies and above, the neutrino interaction cross sections offer fascinating new tests of Standard Model physics \\cite {sigma} and new physics \\cite {gzknus}. Tomography of the Earth is also possible with $PeV$-scale neutrinos\\cite {tomo}. Radio Cherenkov signals have also been used to search recently for neutrinos and cosmic rays with energies upwards of $10^{20}eV$ impinging on the Moon\\cite{Moon}. Cherenkov radiation is also an intrinsically interesting and beautiful physical phenomenon. Coherence is a basic feature of electrodynamics, and the coherent enhancement of Cherenkov radiation in the microwave region has been observed in the laboratory \\cite{Takahashi, Wake}. The Argonne wake-field acceleration project\\cite{Wake} has successfully generated extremely large microwave field-strengths by manipulating coherent radiation from an intense electron beam. Despite a long history, the previous literature apparently does not contain a careful treatment of evolving charge distributions, such as those of electromagnetic showers in air or ice, which incorporates all important features of the problem. The problem is intricate because of a multitude of scales. When an electromagnetic shower evolves, it produces a pancake of charge with a finite thickness, a finite width, probed at a finite wavelength of radiation, and for a finite distance over which the shower is big. All this occurs at a finite distance from the detector. Results on evolving and finite-sized charge distributions are few. Tamm \\cite{Tamm} grappled with the problem of a charged particle on a track of limited length in the early days of the theory. Askaryan \\cite{Askaryan} anticipated a coherence cut-off in air showers at high frequencies of order the inverse pancake size, imposed somewhat by hand. Allan\\cite{Allan} gave physical arguments and order of magnitude estimates based on one of Feynman's electrodynamic formulas. Kahn and Lerche \\cite{KL} attempted to resolve the coherence issue using superpositions of infinite tracks. McKay and Ralston \\cite{mkrlpm} and Alvarez-Mu\\~{n}iz and Zas ($AZ$) \\cite{AZ} considered the influence of the $LPM$ effect at ultra-high energies. Zas, Halsen, and Stanev ($ZHS$) \\cite{ZHS}, and $AZ$ reported results from summing asymptotic far-fields track-by-track in Monte Carlo calculations of great complexity. We present an approach which incorporates all the scales and allows a general analysis. Main results include an expression for the electric field in a factorized form. The ``factorization\" occurs when distance scales can be separated: the characteristic size of the moving charge distribution must be substantially smaller than the scale over which the charge develops. This condition is well satisfied for all cosmic ray applications we have examined. A {\\it form factor} characterizes the moving charge distribution, which multiplies a charge-evolution integral. Not all the scales decouple: subtleties coming under the classic description of {\\it Fraunhofer} and {\\it Fresnel} zones need careful treatment. Finally, the generic situation can be summarized by analytic formulae. This is indispensable given the large parameter space. For example, the numerous and varied numerical plots obtained from immense Monte Carlos can be summarized by a few parameters. With the parameters fixed, predictions can be made for any number of circumstances. ", "conclusions": "" }, "0003/astro-ph0003252_arXiv.txt": { "abstract": "{We present new measurements of the ground state fine-structure line of atomic carbon at 492 GHz in a variety of nearby external galaxies, ranging from spiral to irregular, interacting and merging types. In comparison with CO(1-0) emission observed at the same spatial resolution, the CI(1-0) line intensity stays fairly comparable in the different environments, with an average value of the ratio of the line integrated areas in Kkms$^{-1}$ of CI(1-0)/CO(1-0) = $0.2 \\pm 0.2$. However, some variations can be found within galaxies, or between galaxies. Relative to CO lines (J=2-1, 3-2, 4-3), CI(1-0) is weaker in galactic nuclei, but stronger in disks, particularly outside star forming regions. Also, in NGC~891, the CI(1-0) emission follows the dust continuum emission at 1.3mm extremely well along the full length of the major axis where molecular gas is more abundant than atomic gas. Atomic carbon therefore appears to be a good tracer of molecular gas in external galaxies, possibly more reliable than CO. Atomic carbon can contribute significantly to the thermal budget of interstellar gas. The cooling due to C and CO are of the same order of magnitude for most galaxies. However, CO is generally a more important coolant in starburst galaxies. Cooling due to C and CO amounts typically to $2 \\times 10^{-5}$ of the FIR continuum or 5\\% of the CII line. However, C and CO cooling reaches $\\sim$ 30\\% of the gas total, in Ultra Luminous InfraRed Galaxies (ULIRG) like Arp~220, where CII is abnormally faint. Together with CII/FIR, the emissivity ratio CI(1-0)/FIR can be used as a measure of the non-ionizing UV radiation field in galaxies. The plots of CII/CI or CII/FIR versus CI/FIR show good correlations, in agreement with PDR models, except for two remarkable galaxies Arp~220 and Mrk~231, where high opacities of the CII line and possibly the dust thermal emission may be factors reducing the CII strength below the predictions of the current PDR models.} ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003064_arXiv.txt": { "abstract": "The relation between the gravitational potential energy, the central potential, and the mass is considered for various self-gravitating bodies. ", "introduction": "For homogeneous triaxial ellipsoid with semiaxes $A,\\,B,\\,C\\,$ and density $\\rho$ the next relations are valid (Landau and Lifshits 1975).% The gravitational potential at the inner point ($X,\\,Y,\\,Z\\,$), $-A\\leq X \\leq A$, $-B\\leq Y \\leq B$, $-C\\leq Z \\leq C$, is \\be \\label{Uel} \\ba{l}U_{ell}(X,Y,Z)=\\pi\\,\\rho\\,G\\,A\\,B\\,C\\,\\int_0^\\infty \\, \\biggl(1 -{X^2 \\over A^2+s}-{Y^2 \\over B^2+s}-{Z^2\\over C^2+s} \\biggr) {\\,d\\,s \\over Q_s}; \\\\Q_s=\\sqrt{(A^2+s)(B^2+s)(C^2+s)}. \\ea \\ee The potential energy of the homogeneous triaxial ellipsoid is: \\be \\label{Wel} W_{ell}={3\\over 10}G\\,M^2\\int_0^\\infty{d\\,s\\over Q_s}.\\ee Here $M_{ell}=4/3\\,\\pi\\,\\rho\\,A\\,B\\,C $ is the ellipsoid's mass and $G$ stands for Newtonian constant of gravitation. Note that gravitational energy of self-gravitating body is of negative sign but we loosely write all $W$s with positive sign. In general case $A\\neq B \\neq C$ the integrals in (\\ref{Uel}) and (\\ref{Wel}) are expressed only in terms of the incomplete elliptic integrals and this precludes any detailed analysis. \\ However, if we consider only potential at the center of ellipsoid, $U_{ell}(0,0,0)$, then we get the remarkable relation: \\be \\label{WUMel} WUM={W_{ell} \\over U_{ell}(0,0,0)\\,M_{ell}}= {2\\over 5}, \\ee valid for any values of semi-axes. We shortly refer to this relation (\\ref{WUMel}) as $WUM$-ratio.\\\\ Recently Seidov and Skvirsky (2000a) presented the gravitational potential and the potential energy for the homogeneous rectangular parallelepiped (hereafter RP) which allows to analyse this $WUM$-ratio for the new class of the homogeneous self-gravitating bodies.\\\\ In this paper, we show in sections II and III that the value of $WUM$ for RPs has minimal value $.395437$ (see Eq. (\\ref{WUMC})) for the cube (all three dimensions of RP equal to each other), tends to 1/2 for one dimension of RP far larger larger than two others (long thin \"stick\" with square cross-section), and tends to {${1\\over 2}-{\\sqrt{2}-1\\over 6\\,\\ln(\\sqrt{2}+1)}=.421673$} as one dimension is far less than two others (thin square \"plate\"), see Fig. 1.\\\\ Also, in section \\ref{KANT} we discuss the $WUM$-ratio for homogeneous gravitating bodies studied recently by Kondrat'ev and Antonov (1993). We show that values of $WUM$-ratio for the homogeneous summetrical lenses are in the interval from $128/105\\,\\pi$, for infinitesimally thin symmetrical lens, to $17/20$, for two homogeneous equal spheres just touching each other.\\\\ In section \\ref{poly} we analyse $WUM$-ratio for spherical polytropic stars, and show that $WUM$-ratio varies from $2/5$ to $3/32 \\,\\pi$ for polytropic index $n$ varying from $0$ to $5$.\\\\ In the section \\ref{2ph} we discuss the interesting class of two-phase spheres and show that unlike the polytropes, in this case the $WUM$-ratio's interval is larger: it is possible to get very small values of $WUM$ if the ratio of two densities $q=\\rho_2/\\rho_1$ is large enough and if the relative value of core's radius is not too small. \\\\ At last, in the sections \\ref{step} and \\ref{alpha} we consider another two simple classes of {\\it non-homogeneous} bodies both allowing analytical treatment. ", "conclusions": "There is a rather classic problem of looking for general theorems of stellar structure, see e.g. chapter 3 in the classic text Chandrasekhar (1957). The problem considered in this paper may be also referred to as that dealing with general structure of celestial self-gravitating bodies.\\\\ We start from interesting observation on one constant ratio, namely, (potential energy W)/((central potential U ) x (total mass M)), in homogeneous ellipsoids and then try to look for behavior of this ratio for another homogeneous bodies: rectangular parallelelepipeds and symmetrical lenses. We found that in both cases $WUM$-ratios are confined in rather narrow interval. Suprisingly, dependence of $WUM$ for homogeneous rectangular parallelepipeds (RP) on edge lengths ratio is non-monotonic: it has minimal value $.395437$ for cube while any deviation from cube form to prolate RP (one dimension being smaller than two others) or elongated RP (one dimension being larger that two others) leads to the increase of value of $WUM$. In this respect the behavior of homogeneous rectangular parallelepipeds is quite unlike the behavior of homogeneous ellipsoids and there is still some mystery even to authors.\\\\ As to the homogeneous symmetrical lenses (SL) by Kondrat'ev and Antonov (1993), here the dependence of $WUM$ on parameters of SL is monotonic, however in this case there is also some suprise in the sense that in the limiting case of thin symmetrical $spherical$ lens the $WUM$-ratio's value, ($128/105\\,\\pi$), differs radically from the case of the infinitesimally thin $quadratic$ plate with $WUM=1/2$.\\\\ Then we look for the non-homogeneous however spherically symmetric bodies and found that for the polytropes with polytropic index $n$ in the interval $0-5$, $WUM$-ratio again lies in narrow interval from $2/5$ to $3/32\\,\\pi$. However for two-phase sphere with large ratio of densities $q=\\rho_2/\\rho_1$ it is possible to get very small values of WUM. The physical reason of it is that if we put in the center of any spherical symmetric star a (very) small but dense spherical body then central potential may be very large while total potential energy of star, being integral value, increases not so drastically. The effect of the strong variation of density in the center of star, the \"first-order phase-transition\", is known since pioneer works of W.H. Ramsey (1950).\\\\ In last two paragraphs of paper we consider pure mathematical toy models in further attempts to understand the behavior of the $WUM$-ratio. We conclude this discussion with notice that central-to-surface potential ratio $U(0)/U(R)$ (among other \"global\" characteristics of the celestial self-gravitating configurations) is also worth studying. For polytropes, $U(0)/U(R)=1+\\xi_1/\\mu_1$, see section \\ref{poly}." }, "0003/astro-ph0003209_arXiv.txt": { "abstract": "\\ifpreprint\\noindent\\fi The tree code for the approximate evaluation of gravitational forces is extended and substantially accelerated by including mutual cell-cell interactions. These are computed by a Taylor series in Cartesian coordinates and in a completely symmetric fashion, such that Newton's third law is satisfied by construction and hence momentum exactly conserved. The computational effort is further reduced by exploiting the mutual symmetry of the interactions. For typical astrophysical problems with $N$=$10^5$ and at the same level of accuracy, the new code is about four times faster than the tree code. For large $N$, the computational costs are found to scale almost linearly with $N$, which can also be supported by a theoretical argument, and the advantage over the tree code increases with ever larger $N$. ", "introduction": "\\ifpreprint\\noindent\\fi The tree code (cf.\\ Barnes \\& Hut 1986, hereafter B\\&H) has become an invaluable tool for the approximate but fast computation of the forces in studies of collisionless gravitational dynamics. It has been applied to a large variety of astrophysical problems. The gravitational potential generated by $N$ bodies of masses $\\mu_n$ and at positions $\\bX_{\\!n}$ is \\beq \\Phi(\\bX) = - \\sum_{n=1}^N\\, \\mu_n\\; g(|\\bX-\\bX_{\\!n}|), \\eeq where $g(r)$ denotes the greens function, i.e.\\ for un-softened gravity $g(r)=G/r$. The essence of the tree code is to approximate this sum over $N$ terms by replacing any partial sum over all bodies within a single cell which is well-separated from \\bX\\ by just one term. The inner structure of the cell is partly taken into account using its multipole moments. This method reduces the overall costs for the computation of all forces from ${\\cal O}(N^2)$ to ${\\cal O}(N\\log N)$. The tree code, however, does not exploit the fact that the force due to the contents of some cell is very similar at nearby positions (even though one may use the fact that nearby bodies tend to have very similar interaction lists, cf.\\ Barnes 1990). Exploiting this is the idea of the fast multipole method (FMM) \\citep{FMM-a}. The FMM employs a (usually) non-adaptive structure of hierarchical grids and considers only interactions between nodes on the same grid level according to their geometrical neighbourhood. The gravitational field due to some source cell and within some sink cell is approximated by a multipole expansion in spherical harmonics, the order of which is adapted to meet predefined accuracy limits. This method has been claimed to reduce the overall amount of operations to ${\\cal O}(N)$, but the tables given by \\citet{FMM-b} do not support this claim. \\citet{CDM} find that the FMM needs ${\\cal O}(N\\log N)$ operations, and is significantly {\\em slower\\/} for astrophysical applications than the tree code at comparable accuracy. Instead of using a spherical multipole expansion of adaptive order, it is actually more efficient to use a Cartesian expansion of fixed order. Moreover, by preserving the symmetry of the gravitational interaction for mutual cell-cell interactions, one can (i) reduce the computational effort and (ii) obtain a code that satisfies Newton's third law by construction and hence results in exact conservation of momentum, a property not shared by the traditional tree code. ", "conclusions": "\\ifpreprint\\noindent\\fi A new code for the approximate evaluation of gravitational forces has been presented, tested, and compared to the tree code. This new code is substantially faster than the tree code. Moreover, unlike the latter, it satisfies Newton's third law by construction, such that any $N$-body code based on it will not introduce spurious net-accelerations. The new code is based on a Taylor expansion of the greens function in Cartesian coordinates and incorporates mutual cell-cell interactions. The simple algorithm is well suited for implementation on parallel computers: different mutual interactions (MIs) can be passed to different CPUs. The scaling of the CPU time required for the mutual forces of a number $N$ of bodies becomes essentially linear at $N\\ga10^5$, so that with ever larger $N$ the new code is increasingly faster than the tree code, allowing for a substantial improvement in simulations employing large number of bodies. The only disadvantage is the increased requirement of memory compared to the standard tree code: 20 floating point numbers per cell are needed to hold the Taylor expansion coefficients. (By using a tree-walking algorithm instead of that given in \\S\\ref{sec:algo}, one can avoid this at the price of enhanced CPU time consumption.) In spirit, the new code is similar to Greengard \\& Rokhlin's (1987) fast multipole method, but is more efficient because it (i) uses a Cartesian instead of a spherical harmonic multipole expansion and (ii) fixes the order of the expansion while controlling the accuracy via the interaction condition, rather than fixing the interactions and adapting the expansion order to the accuracy. A concern with codes based on cell-cell interactions is their performance in the presence of individual time steps. Clearly, when not all the forces are to be computed, such codes fare less favorably. However, when the forces for all bodies within some domain are desired, the new code is still a significant improvement over the tree code. The new code has been written in C++ and \\ifpreprint will be \\else is \\fi electronically available from the author upon request." }, "0003/astro-ph0003179_arXiv.txt": { "abstract": " ", "introduction": "Cosmic neutrinos with energy $ > 10^{12}\\,$ eV are a new and exciting observational window for cosmic ray physics and astrophysics. Indeed, many interesting models predict the existence of cosmic neutrinos with an energy up to $10^{18}-10^{20}\\,$ eV ({\\em extreme energy}), where charged particles are expected to be cut-off by the so called GZK mechanism~\\cite{GZK}. The aforementioned threshold of $10^{12}\\,$ eV is necessary for a detection of diffuse flux, because of the large atmospheric neutrino background.\\\\ The detection of such extreme energy neutrinos is challenging as it demands a very large sensitive area. The experiments of the upcoming generation plan to use underwater-underice Cerenkov detectors~\\cite{gai95} and large field-of-view atmospheric fluorescence detectors~\\cite{lin97}. One of the best evidence of cosmic neutrinos would be the detection of upstream showers/particles emerging from the Earth. For this kind of events, the atmospheric muon and primary charged cosmic ray background would be completely suppressed. On the other hand, this signature can hardly be observed at extreme energy because the rise of weak cross sections entails the opacity of the Earth with respect to neutrino propagation~\\cite{nau98,gand}.\\\\ The problem of neutrino propagation through the Earth can be understood on the basis of neutrino interactions with matter. For neutrino energy well above few GeV, interactions with conventional matter are described by neutrino-nucleon deep inelastic scattering and neutrino-electron scattering~\\cite{gand}. The former can be classified in two main categories: charged current (CC) $\\nu_{l}\\,N\\rightarrow\\l\\,+hadrons$, where the neutrino disappears and a charged lepton $l$ is created, and neutral current (NC) $\\nu_{l}\\,N\\rightarrow\\nu_{l}\\,+hadrons$, where the neutrino survives the interaction. The neutrino-electron scattering cross section is several order of magnitude less than neutrino-nucleon cross section and does not affect significantly neutrino propagation in matter. The only exception is the $W^-$ resonant scattering for the reaction $\\bar{\\nu}_{e}\\,e\\rightarrow W^-+anything$, which, however, hardly affects neutrino attenuation length in matter for $\\bar{\\nu}_{e}$ energy in the range $2\\cdot\\,10^{15}$~eV$\\leq\\,E_{\\nu}\\,\\leq\\,2\\cdot\\,10^{16}$~eV~\\cite{gand}.\\\\ The neutrino-nucleon total cross section is approximately $\\sigma_{tot}\\approx 10^{-37}$~cm$^2$ at $E_{\\nu}=10$~GeV and rises roughly linearly with energy. As a consequence, while in the GeV energy range the Earth is completely transparent to neutrinos, it is expected to become opaque for sufficently high energy. As to muon neutrinos, the CC interaction lenght inside Earth equals Earth's diameter for energy around $40$~TeV~\\cite{gand}. In order to accurately describe the change in energy spectrum of a muon neutrino flux traversing the Earth, both the effect of neutrino-nucleon CC and NC must be properly taken into account. The problem of neutrino propagation in the Earth can be solved accurately either by Monte-Carlo simulation or by approximate iterative solution of the relevant transport equation~\\cite{nau98}.\\\\ It has been recently pointed out~\\cite{far97, hal98,bot99,Iyer} that the behaviour of $\\tau$-neutrinos, whose existence should be guaranteed in a neutrino-oscillation scenario, should be significantly different from $\\nu_{\\mu}$ and $\\nu_e$. Whilst muon and electron neutrinos are practically absorbed after one CC interaction, the $\\tau$ lepton created by the $\\nu_{\\tau}$ CC scattering may decay in flight before losing too much energy, thereby generating a new $\\nu_{\\tau}$ with comparable energy. Hence, ultra high energy $\\tau$-neutrinos should emerge from the Earth instead of being absorbed. For a correct evaluation of energy spectrum of $\\nu_{\\tau}$'s emerging from the Earth, one has to properly take into account $\\nu_{\\tau}$ interactions as well as $\\tau$ energy loss and decay. An analytical approach similar to that used in~\\cite{nau98} for $\\nu_{\\mu}$ has been proposed by S. Iyer et al.~\\cite{Iyer} neglecting $\\tau$ energy loss. This approach holds as long as energy does not exceed $10^{16}$ eV because above this energy $\\tau$ interaction lenght becomes comparable with $\\tau$ decay lenght (see next section).\\\\ In the present work, a detailed Monte Carlo calculation of $\\nu_{\\tau}-\\tau$ system propagation through the Earth has been performed for energy up to $10^{20}$~eV including the $\\tau$ energy loss contribution, with special emphasis on the initial $\\nu_{\\tau}$ spectrum deformation as a function of the zenith angle of the emerging particle. ", "conclusions": "Whilst, as it is well known, the Earth start to be opaque for muon neutrino propagation at energies around $10$ TeV, a regeneration mechanism relevant to \\nut prevents them from being absorbed. This mechanism is ultimately owing to the shorter, with respect to muon, $\\tau$ lepton decay length. Nevertheless, the unavoidable radiative $\\tau$ energy loss sets an upper bound on the energy of emerging \\nut's to $10^{17}\\,$ eV. Therefore, for underwater-underice neutrino telescopes, whose sensitivity will cover the whole neutrino spectrum, \\nut's could be the major source of neutrino events above the PeV region. In such detectors the charactestics of Fig.~\\ref{fig:fin1} could be also used as one of the signatures of tau neutrinos. For atmospheric fluorescence detectors the effective cut-off of $10^{17}\\,$ eV could be below their energy threshold. If this is the case, those detectors should be able to detect only almost horizontal neutrinos." }, "0003/astro-ph0003453_arXiv.txt": { "abstract": "Among the galaxies which were included in the Atlas and Catalogue of Interacting galaxies by Vorontsov-Vel'yaminov (hereafter VV) as multiple systems (``nests'', ``chains'' and similarly looking systems), there are many objects, where the interaction is not evident. Some of them are single objects, including low-mass galaxies with active star formation (SF). In this work we present the description of observations and results of the long-slit spectrophotometry with the Russian 6\\,m telescope of three VV-galaxies looking like double or multiple systems, and H{\\sc i} observations of one of them in order to elucidate their nature, determine their metallicity, kinematic properties and the evolution status. Galaxies VV~432 and VV~747 are found to be dwarf systems with low oxygen abundance (O/H $\\approx$~1/22 and 1/12 of the solar value, respectively). Their velocity curves indicate quite slow rotation with respective maximum velocities of about 60 and 80 km~s$^{-1}$, in agreement with their low luminosities. The distance to VV~432 is rather uncertain. If it is a member of Virgo Cluster, this is the most metal-deficient known galaxy of this aggregate. For galaxy VV~543 the measured emission-line redshift 0.047 appeared ten times larger than it was given in the original paper and is cited in databases. This ``system'' evidently represents an optical pair of two galaxies with large velocity difference. The fainter western component is an H{\\sc ii}-galaxy, while the brighter one is an absorption-line early-type galaxy with the radial velocity being 1600 km~s$^{-1}$ lower. ", "introduction": "Among the definitely interacting galaxies in close pairs or groups, many of which were discovered by Vorontsov-Vel'yaminov and included in his Atlas and Catalog of interacting galaxies (Vorontsov-Vel'yaminov, 1959, 1977) there are objects having VV numbers, in which the interaction is not so evident. Such objects were usually classified by Vorontsov-Vel'yaminov as the \"nests\" or often similarly looking systems called \"chains\" or \"pairs in contact\". He considered these objects as compact fragmenting systems, giving birth to young galaxies, but very soon observations showed that their nature may be different. Such systems can present both galaxies of strange and unusual shapes (either single objects or mergers), and really multiple systems, where it is hard to guess the number of components without detailed studies. In many cases only spectral measurements of their gas velocity distribution enable to prove or to disprove their solitude. Earlier spectral observations confirmed that some VV systems actually present single dwarf galaxies with clumpy inner structure, resembling dwarf irregulars with multiple regions of active star formation (SF), or blue compact galaxies (BCGs) (Vorontsov-Vel'yaminov 1979a,b; Afanasiev et al.\\, 1980; Arkhipova et al.\\, 1981, 1987a,b,c). Only a few single nearby VV galaxies have been studied in detail (e.g. VV~556~$\\equiv$~GR~8, VV~499~$\\equiv$~DDO~053). The important questions on the chemical abundances in the studied VV galaxies left outside the scope of these early observations. New opportunities appeared due to new CCD-detectors, which enable to realize high sensitivity and large dynamical range. It induced authors to return back to the study of \"nest-like\" VV galaxies. Besides, many new observational data have appeared on many of these objects. For the current study we selected more than twenty VV galaxies looking like multiple systems or singular irregular systems in the POSS/DSS \\footnote {DSS is Digital Sky Survey distributed by Hubble Space Telescope Science Institute} images, for which such indicators of interaction as well defined tidal tails or bridges are absent. The main objectives of our investigation are: -- to clarify the distances, total luminosities and masses of those systems, systemic velocities of which are not known or badly known; -- to carry out high S/N ratio spectrophotometry in order to address problems of chemical abundances and evolutional status of these objects, -- to analyse the inner gas kinematics and structural properties of the objects. Since many of VV galaxies in question are dwarfs with recent or current SF burst, it is also important to check possible companions which could exert strong enough tidal action. \\begin{table*} \\begin{center} \\caption{\\label{Tab1} Journal of observations} \\begin{tabular}{lllcrccc} \\\\ \\hline \\MC{1}{c}{ Galaxy } & \\MC{1}{c}{ Date } & \\MC{1}{c}{ Instrument } & \\MC{1}{c}{ Grating } & \\MC{1}{c}{ Exposure } & \\MC{1}{c}{ Wavelength } & \\MC{1}{c}{ Dispersion } & \\MC{1}{c}{ PA } \\\\ \\MC{1}{c}{ Name } & & & \\MC{1}{c}{ [grooves/mm] } & \\MC{1}{c}{ time [s] } & \\MC{1}{c}{ Range [\\AA] } & \\MC{1}{c}{ [\\AA/pixel] } & \\MC{1}{c}{ [Degree] } \\\\ \\MC{1}{c}{ (1) } & \\MC{1}{c}{ (2) } & \\MC{1}{c}{ (3) } & \\MC{1}{c}{ (4) } & \\MC{1}{c}{ (5) } & \\MC{1}{c}{ (6) } & \\MC{1}{c}{ (7) } \\\\ \\hline \\\\[-0.3cm] VV~432& 12.02.1999 & LSS+PMCCD & 325 & 900 & $3700-8000$ & 4.6 & ~~27 \\\\ VV~432& 23.04.1999 & LSS+PMCCD & 1302& 2$\\times$1200 & $4000-5200$ & 1.2 & ~~27 \\\\ VV~432& 23.04.1999 & LSS+PMCCD & 1302& 2$\\times$1200 & $6000-7200$ & 1.2 & ~~27 \\\\ VV~543& 11.02.1999 & LSS+PMCCD & 325 & 900 & $3700-8000$ & 4.6 & 101 \\\\ VV~747& 12.02.1999 & LSS+PMCCD & 325 & 1200 & $3700-8000$ & 4.6 & ~~56 \\\\ \\hline \\\\[-0.2cm] \\end{tabular} \\end{center} \\end{table*} \\begin{table*}[hbtp] \\centering{ \\caption{\\label{Tab2} Main parameters of studied VV-galaxies} \\begin{tabular}{lrrrr} \\hline \\rule{0pt}{10pt} Parameter & VV~432 & VV~543~W & VV~543~E & VV~747 \\\\ \\hline $\\alpha_{2000}$ & 12$^h$17$^m$34.7$^s$ & 13$^h$42$^m$22.2$^s$ & 13$^h$42$^m$23.5$^s$& 10$^h$57$^m$46.9$^s$ \\\\ $\\delta_{2000}$ & +12$^\\circ$23$^{\\prime}$46$^{\\prime\\prime}$ & +29$^\\circ$49$^{\\prime}$33$^{\\prime\\prime}$ & +29$^\\circ$49$^{\\prime}$30$^{\\prime\\prime}$ & +36$^\\circ$15$^{\\prime}$38$^{\\prime\\prime}$ \\\\ B$_{tot}^L$ & 14.73$\\pm$0.15 & 17.7$\\pm$0.3 & 15.20$\\pm$0.20 & 15.52$\\pm$0.71 \\\\ A$_B^N$ & 0.03 & 0.00 & 0.00 & 0.04 \\\\ V$_{HeI}$ (km/s) & $-160\\pm$6 & 14100$\\pm$20 & 12480$\\pm$90 & 629$\\pm$7$^5$ \\\\ Dist$_{Vir}$(Mpc) & 20.7$\\pm$0.8 & 188.4$\\pm$0.3 & 166.8$\\pm$1.2 & 9.2$\\pm$0.1 \\\\ M$_{B}$$^{1}$ & $-16.8$ & $-18.6$ & $-20.9$ & $-14.4$ \\\\ D$_{25}$ (arcsec) & 111 & $\\approx$16 & 39 & 50 \\\\ D$_{25}$ (kpc) & 11.1 & $\\approx$14.6 & 31.5 & 2.2 \\\\ Axis ratio b/a$^L$ & 0.31 & 0.50 & 1.00 & 0.85 \\\\ 12+log(O/H) \\ & 7.58$\\pm$0.06 & 8.5$\\pm$0.1 & --- & 7.85$\\pm$0.05 \\\\ H{\\sc i} flux$^{2}$ & 9.04$\\pm$1.99$^3$ & --- & --- & 4.03$\\pm$0.39 \\\\ W$_{20}$~km s$^{-1}$ & 123$^3$ & --- & --- & 117$\\pm$11 \\\\ M(H{\\sc i}) (10$^{8}$M$_{\\odot}$) & 9.1$\\pm$2.0 & --- & --- & 0.81$\\pm$0.08 \\\\ M(H{\\sc i})/L$_{B}$$^{4}$ & 1.2 & --- & --- & 1.0 \\\\ \\hline \\multicolumn{4}{l}{B$_{tot}$ --- total blue magnitude; M$_{B}$ --- absolute blue magnitude} \\\\ \\multicolumn{4}{l}{D$_{25}$ --- Diameter at surface brightness $\\mu_{B}$ = 25~$mag/\\Box^{\\prime\\prime}$} \\\\ \\multicolumn{4}{l}{A$_{B}$ --- Galactic extinction; $^{L}$\\,\\ Data from LEDA; $^{N}$\\,\\ Data from NED} \\\\ \\multicolumn{4}{l}{$^{1}$\\,\\ With Galactic extinction correction; $^{2}$\\,\\ Units of (Jy$\\cdot$km/s)} \\\\ \\multicolumn{4}{l}{$^{3}$\\,\\ from Schneider et al. (\\cite{Schneider91}); $^{4}$\\,\\ In (M/L$_{B}$)$_{\\odot}$; $^{5}$\\,\\ From H{\\sc i}-profile} \\\\ \\end{tabular} } \\end{table*} In this paper, the first in the series, we present the results of recent long-slit spectroscopy for three VV objects: VV~432, VV~543 and VV~747 and the observation in HI-line of VV~747. In section~\\ref{observations} we describe observations, data reduction, abundances determination and the measurements of velocity distribution of ionized gas along the slit. Observations of VV~747 in the HI-line 21 cm and their results are presented in section~\\ref{HI_observations}. In section~\\ref{Individual_prop} we consider the individual properties of studied galaxies. Discussions and preliminary conclusions are presented in section ~\\ref{Conclusions}. We adopt throughout the paper H$_0 =$75 km sec$^{-1}$. ", "conclusions": "\\label{Conclusions} One of the important questions concerning the nature of the ``nest'' and ``chains'' of VV-galaxies is their evolution status. Current study as well as several previous publications demonstrate that many of the low-luminosity VV-galaxies are relatively nearby irregular galaxies with several bright knots of enhanced SF. Their position in two-colour diagram indicates a presence of SF burst in many of them (Zasov \\& Arkhipova \\cite{Zasov00}). They roughly follow so called luminosity -- metallicity relation: the least luminous galaxies show in general smaller values of $O/H$. VV~432 has a very low heavy-element abundance. Its $O/H$ is among the lowest ten values of the most metal-deficient BCGs out of more than one thousand BCGs/H{\\sc ii}-galaxies known up-to-now. It may be considered as an example of non-evolved galaxy. Analysis of the empirical correlations suggests that dwarf galaxies with 12+log(O/H) $<$ 7.6 can currently experience only the first in their history episode of SF (Izotov \\& Thuan \\cite{Izotov99}). Therefore VV~432 with 12+log(O/H) = 7.58 is very good candidate for more detailed study. If it is situated in Virgo cluster, it will be the most metal deficient galaxy of this aggregate, being even less chemically evolved than another well-known metal-poor H{\\sc ii}-galaxy in the direction of Virgo cluster H{\\sc i}~1225+01 with 12+log(O/H) = 7.66 (Salzer et al. \\cite{Salzer91}; Chengalur et al. \\cite{Chengalur95}). One of the possible ways to resolve the dilemma of radial distance to VV~432 is a detection of brightest stars and construction of their color-magnitude diagram. The galaxies we discuss have rather close neighbours. Icke (\\cite{Icke85}) first has drawn attention to the importance of relatively weak interactions to trigger gravitational instability in gas disks via generation of shocks. Many observational evidences for the important role of weak interactions to trigger SF were obtained since that time, including the detection of low mass H{\\sc i}-companions of nearby H{\\sc ii}-galaxies (Chengalur et al. {\\cite{Chengalur95}; Taylor et al. \\cite{Taylor93}, \\cite{Taylor95}; Taylor \\cite{Taylor97}) and optical faint companions of BCGs (Pustilnik et al. {\\cite{Pustilnik97}). Recent results on late spirals by Reshetnikov \\& Combes (\\cite{Reshet97}) and Rudnick \\& Rix (\\cite{Rudnick98}) also suggest the importance of weak interactions to modulate SF history in these galaxies. From the observational data discussed above some preliminary conclusions can be drawn: \\begin{itemize} \\item Spectrophotometry of VV-galaxies shows that low luminosity representatives of this sample are in general metal-deficient objects, and in this aspect they are similar to dwarf irregular galaxies. \\item The extremely metal-deficient galaxy VV~432 (12+log(O/H) = 7.58) is probably the least evolved known member of Virgo cluster. \\item The system VV~543 radial velocity cited in the RC3 catalog and other databases is wrong. This object consists of two galaxies with discordant redshifts (a unique example among the galaxies of this type!), and probably presents an optical pair. VV~543W is an H{\\sc ii}-galaxy with the radial velocity 1620 km~s$^{-1}$ higher than that of absorption-line E-type galaxy VV~543E. \\item VV~747 is probably a single dwarf galaxy rich of neutral hydrogen. \\item The presence of massive companion galaxies at the distances of few hundred kpc from the studied VV-objects with enhanced SF rate is probably indicative of the important role of weak interactions to trigger SF activity at least in some fraction of low mass VV galaxies. \\end{itemize}" }, "0003/astro-ph0003386_arXiv.txt": { "abstract": "The status of the hydrodynamical modelling of nonlinear multi-mode stellar pulsations is discussed. The hydrodynamical modelling of steady double-mode (DM) pulsations has been a long-standing quest that is finally being concluded. Recent progress has been made thanks to the introduction of turbulent convection in the numerical hydrodynamical codes which provide detailed results for individual models. An overview of the modal selection problem in the HR diagram can be obtained in the form of bifurcation diagrams with the help of simple nonresonant amplitude equations that capture the DM phenomenon. ", "introduction": "Self-excited multi-mode pulsations are quite common among the less evolved luminous stars, such as the delta Scuti stars, but they mostly involve nonradial modes of oscillation. In contrast, the classical variable stars, \\viz the Cepheids and RR~Lyrae stars, are believed to be radially pulsating, and two independent frequencies are typically identified in the Fourier spectrum. These pulsating stars are referred to as double-mode even though, in principle, there could be more than two modes involved with locked frequencies. (We should add that there is evidence that suggests the presence for nonradial modes as well (Kov\\'acs \\etal, Moskalik, Olech \\etal in Szabados \\& Kurtz 2000) . In the Galaxy, the Beat Cepheids, also called double-mode (DM) Cepheids are relatively rare. Only about a good dozen are known. On the other hand, the recent observations of the Magellanic Clouds (Beaulieu \\etal 1995, 1997, Welch \\etal 1995, Udalski \\etal 1999) have shown that in these galaxies Beat Cepheids are quite common. Beat Cepheid pulsations occur either in the fundamental/first overtone (F/O$_1$) modes, or in the first/second overtone (O$_1$/O$_2$) modes. In the Galaxy there is only one O$_1$/O$_2$ known DM pulsator (CO~Aur), but in the SMC Udalski \\etal 1999 claim 70 candidates out of 93 DM Cepheids. The occurrence of DM RR~Lyrae stars (RRd stars), depends on the type of cluster. For example, not a single RRd star is observed in $\\omega$ Centauri (\\cf Kov\\'acs in this Volume) while in M68, 9 RRd stars are known out of 37 RR Lyrae stars. The DM pulsators are important for pulsation theory in that they impose very stringent requirements on the numerical modelling. Double periodicity and indeed the overall view of the modal selection picture provide many more observational constraints than their single-mode siblings. Despite the frequent occurrence of DM pulsations in nature, the numerical modelling of this type of pulsation had remained a serious challenge until recently. In fact, it had become abundantly clear that purely radiative models, \\ie models that disregarded convective transport, were not capable of yielding DM pulsations, except on a purely transient basis, \\ie they were switching from one mode to another, but too fast to account for the observed fairly steady nature of DM pulsators. Actually, some purely radiative models of RRd stars had been found already by Kov\\'acs and Buchler (1993), but those results were not satisfactory. However, recently, DM behavior has been found simultaneously, and fully independently, in RR~Lyrae models by Feuchtinger (1998) and in Cepheid models by Koll\\'ath \\etal (1998), this with different numerical methods, \\viz the Vienna and the Florida codes. The breakthrough came as a result of the inclusion of time-dependent turbulent convection in the models. ", "conclusions": "" }, "0003/astro-ph0003103_arXiv.txt": { "abstract": "We present a revised catalog of 2106 Galactic stars, selected without kinematic bias, and with available radial velocities, distance estimates, and metal abundances in the range $0.0 \\le \\feh \\le -4.0$. This update of the Beers \\& Sommer-Larsen (1995) catalog includes newly-derived homogeneous photometric distance estimates, revised radial velocities for a number of stars with recently obtained high-resolution spectra, and refined metallicities for stars originally identified in the HK objective-prism survey (which account for nearly half of the catalog) based on a recent re-calibration. A subset of 1258 stars in this catalog have available proper motions, based on measurements obtained with the {\\it Hipparcos} astrometry satellite, or taken from the updated Astrographic Catalogue (AC 2000; second epoch positions from either the Hubble Space Telescope Guide Star Catalog or the {\\it Tycho} Catalogue), the Yale/San Juan Southern Proper Motion (SPM) Catalog 2.0, and the Lick Northern Proper Motion (NPM1) Catalog. Our present catalog includes 388 RR Lyrae variables (182 of which are newly added), 38 variables of other types, and 1680 non-variables, with distances in the range 0.1 to 40 kpc. ", "introduction": "Studies of the kinematics of various stellar populations in the Galaxy, in particular the thick disk and the nascent halo, have long been limited by the availability of large samples of stars with measurements of velocities, distances, and metallicities. Such a database is required in order to constrain plausible scenarios for the formation and evolution of the Milky Way and other large spiral galaxies like it. Current issues which might be addressed with such data include: (a) the rotational character of the thick disk and halo (see, e.g., Beers \\& Sommer-Larsen 1995, hereafter BSL, and references therein), (b) the existence and the observed lower limit on the metal abundance of stars in the so-called metal-weak thick disk (MWTD) (Morrison, Flynn, \\& Freeman 1990, hereafter MFF; BSL; Carney et al. 1996; Chiba \\& Yoshii 1998; Martin \\& Morrison 1998) (c) a dual-component (flattened plus spherical) halo population in the Galaxy (Hartwick 1987; Sommer-Larsen \\& Zhen 1990; Norris 1994; Kinman, Suntzeff, \\& Kraft 1994; Sommer-Larsen et al. 1997), (d) quantitative estimates of the local density of thick disk and halo stars (Yoshii 1982; Preston, Shectman, \\& Beers 1991; Morrison 1993), (e) tests for the existence of a putative ``counter-rotating'' halo component beyond 4--5 kpc from the plane (Majewski 1992; Wilhelm 1995; Carney et al. 1996; Wilhelm et al. 1996; Zinn 1996; Carney 1999), (f) measures of the local halo velocity ellipsoid for comparison with the derived ellipsoid for more distant halo stars (Sommer-Larsen et al. 1997), and (g) derivation of a reliable RR Lyrae absolute magnitude estimate based on statistical parallax analyses (Layden et al. 1996; Fernley et al. 1998; Popowski \\& Gould 1998). As more extensive searches are carried out for evidence of past (and present) mergers of smaller galaxies with the Milky Way (e.g., Preston, Beers, \\& Shectman 1994; Harding et al. 1998; Helmi \\& White 1999; Majewski 1999), it is of equal importance to obtain secure knowledge of the ``global'' Galactic kinematic properties, so that deviations from the expected behavior can be reliably assessed. To confidently address the above issues (and many others) stars chosen as kinematic tracers should be identified in a manner which does {\\it not} depend on a kinematic selection criterion itself. Although it may be possible to statistically correct for an input selection bias of this nature (Bahcall \\& Casertano 1986; Ryan \\& Norris 1991a; Carney et al. 1994; Carney 1999), one is left with lingering doubt concerning the derived kinematic parameters based on post-facto modifications of the results. It is similarly important that the tracer stars cover a wide range of metallicities and distances (over both northern and southern Galactic hemispheres), so that correlations of kinematics as a function of these parameters can be investigated. Although hints of the impact of a kinematic selection criterion are evident in the work of Yoshii \\& Saio (1979), the first {\\it large} database suitable for exploration of many of these issues was that of Norris (1986), which included some 400 spectroscopically and/or photometrically selected stars with abundances $\\feh \\le -0.6$, and with available radial velocities and distance estimates. In Paper I of this series BSL extended the Norris catalog by inclusion of some 900 stars identified from the HK object objective-prism survey of Beers \\& colleagues (Beers, Preston, \\& Shectman 1985; Beers, Preston, \\& Shectman 1992a; Beers et al. 1992b), as well as some 600 additional stars from other smaller samples which appeared in the literature subsequent to the publication of the Norris catalog, obtaining a total sample of 1936 stars. In this paper we present a revision of the BSL catalog of Paper I, based on additional observational information which has recently become available. In addition to revisions of metallicities for the published HK survey stars, new photometric distance estimates have been made for the entire BSL catalog based on an internally self-consistent methodology, and in some cases, new photometry. Radial velocities have been updated based on recently-obtained high-resolution data for a number of stars. Much more accurate positional information for the stars in our revised catalog has been obtained by comparison with astrometric positions available for many of the brighter stars in our sample, plus improved information from automated scans of wide-field photographic plates, such as compiled in USNO-A V2.0 (Monet et al. 1998), NPM1 (Klemola, Hanson, \\& Jones 1993), and SPM 2.0 (Platais et al. 1998). We have also added 182 RR Lyrae variables from the recent work of Layden (1994), Layden et al. (1996), and Fernley et al. (1998), so that useful comparisons of the kinematics of these stars with the non-variables (which presumably sample the same Galactic phase space) can be carried out. The {\\it major} difference between the present catalog and the BSL catalog is the addition of proper motions, from a variety of sources, for over half of the stars in our catalog. In \\S 2 we discuss the assemblage of the present catalog. Revisions of radial velocities and abundances, and in particular, distance estimates, are discussed in \\S 3. In \\S 4 we present the new proper motion information, and discuss the averaging we have carried out in order to minimize statistical errors. In \\S 5 we discuss the observed characteristics of the stars in the revised catalog, and derive estimates of space motions and orbital parameters for the subset of stars with complete kinematical information . In the accompanying analysis paper (Chiba \\& Beers 2000; Paper III) we use this wealth of new information to consider many of the questions put forth above. ", "conclusions": "" }, "0003/astro-ph0003429_arXiv.txt": { "abstract": "We present near-infrared and optical observations of the afterglow to the Gamma-Ray Burst (GRB) 991216 obtained with the F. L. Whipple Observatory 1.2-m telescope and the University of Hawaii 2.2-m telescope. The observations range from 15 hours to $3.8$~days after the burst. The temporal behavior of the data is well described by a single power-law decay $t^{-1.36\\pm 0.04}$, independent of wavelength. The optical spectral energy distribution, corrected for significant Galactic reddening of $E(B-V)=0.626$, is well fitted by a single power-law with $\\nu^{-0.58\\pm 0.08}$. Combining the IR/optical observations with a Chandra X-ray measurement gives a spectral index of $-0.8\\pm 0.1$ in the synchrotron cooling regime. A comparison between the spectral and temporal power-law indices suggest that a jet is a better match to the observations than a simple spherical shock. ", "introduction": "The BeppoSAX (Boella et al.~1997) and RXTE (Levine \\etal\\ 1996) satellites have brought a new dimension to gamma-ray burst (GRB) research, by providing rapid localizations of several bursts per year. This has allowed many GRBs to be followed up at other wavelengths, ranging from the X-ray (Costa et al.~1997) and optical (van Paradijs et al.~1997) to the radio (Frail et al.~1997). Precise positions have also allowed redshifts to be measured for a number of GRBs (e.g. GRB 970508: Metzger et al.~1997), providing definitive proof of their cosmological origin. The extremely bright gamma-ray burst GRB 991216 was detected by BATSE (Kippen, Preece, \\& Giblin~1999) on December 16.671544 UT, with its peak flux (fluence) ranking it as the 2nd (13th) of all BATSE bursts detected so far. The RXTE PCA search for the X-ray afterglow of GRB 991216 started about four hours after the burst (Takeshima et al.~1999) and detected a strong, decaying X-ray afterglow, providing much improved burst position. It should be noted that the X-ray afterglow of GRB 991216 was also detected by much less sensitive RXTE ASM instrument as early as one hour after the burst (Corbet \\& Smith~1999), providing a measurement of the X-ray afterglow at times which have previously not been studied. In addition, observations of GRB 991216 by the Chandra Observatory resulted in the first arcsecond position determination for an X-ray afterglow (Piro et al. 1999). The optical afterglow of GRB 991216 was identified by Uglesich et al.~(1999) with data taken about $12$~hours (December 17.142 and 17.372 UT) after the burst, using the MDM 1.3-m telescope. It was recognized as a bright variable object ($R\\approx18.8$ at Dec. 17.142), not present in the digitized POSS II plate, declining with a temporal decay index of $\\approx -1.4$. Numerous independent confirming observations of the fading optical transient (OT) have followed, starting with Henden et al.~(1999) and Jha et al.~(1999). Near-infrared observations were also reported by Vreeswijk \\etal\\ (1999a) and Garnavich \\etal\\ (1999b). Absorption lines at $z=1.02$ seen in the optical spectrum of GRB~991216 taken with the VLT-UT1 8-m telescope by Vreeswijk et al.~(1999b) provide a lower limit to the redshift of the GRB source. Given the gamma-ray fluence (Kippen 1999), the isotropic energy from the burst was more than 8$\\times 10^{53}$~ergs ($H_o=65$ km$\\;$s$^{-1}\\;$Mpc$^{-1}$, $\\Omega_m=0.3$, $\\Omega_\\Lambda=0.7$), or nearly half a Solar rest mass radiated away in under 10 seconds. This exceedingly large energy requirement can be reduced if the burst emission is beamed. To date, evidence for jets has been found in only a handful of GRB afterglows (Sari, Piran \\& Halpern 1999; Kulkarni et al. 1999; Stanek et al. 1999) and it remains to be shown whether anisotropy is ubiquitous. We present optical and near-IR photometry of GRB~991216 from observations obtained at the Hawaii 88-inch and the Fred L. Whipple 1.2m telescopes. We describe the data and the reduction procedure in Section 2. In Section 3 we discuss the multiband temporal behavior of the GRB OT. In Section 4 we describe the broad-band spectral properties of the afterglow deduced from our IR/optical data. ", "conclusions": "We present well-calibrated $RJK$ observations of the GRB~991216. Our data indicates that the decay of the optical afterglow is well represented by a single power-law with index $\\alpha =1.36\\pm 0.04$ from 0.5 days to four days after the burst. Combining published late-time $R$-band observations with our data suggests a single power-law is a good fit out to 20 days after the burst. The optical spectral energy distribution, corrected for significant Galactic reddening, is well fitted by a single power-law with an index of $\\beta =0.58\\pm 0.08$. However, when the possible systematic error in the SFD extinction map is considered, the index may be somewhat steeper ($\\beta =0.87\\pm 0.08$). A Chandra X-ray observation obtained near the time of our photometry provides a spectral index between the near IR and X-rays of $\\beta =0.8\\pm 0.1$. A comparison between the spectral and temporal power-law indices suggest that the GRB is not consistent with a simple spherical shock model. The IR/optical light curve and colors are better matched by a shock produced from a collimated jet." }, "0003/astro-ph0003335_arXiv.txt": { "abstract": "s{ For those angular multipoles where cosmic variance is an issue, non-Gaussianities in the Cosmic Microwave Background (CMB) anisotropies will be hard to detect. Here, we construct explicitly the best unbiased estimator for the CMB angular bispectrum.} While the statistical properties of the CMB anisotropies are a powerful means to discriminate amongst the possible cosmological scenarios, actually measuring non-Gaussianity in the data is a very difficult task\\cite{waynhu}. The typically small signal should be compared to the noise and the key quantity is the signal to noise ratio. The noise creeps into the dataset through instrumental errors, foregrounds contamination or incomplete sky coverage. Add to this the so-called `cosmic variance': the fact that we only have access to one realization of the temperature anisotropies ${\\Delta }({\\bf e}) \\equiv { \\delta T / T} ({\\bf e})$ whereas theoretical predictions are expressed through ensemble averages. It can dominate the other sources of error and therefore, if one wants to unveil non-Gaussianity, it is necessary to address the cosmic variance problem for those quantities characterizing a possible non-Gaussian CMB temperature anisotropy distribution. For that one constructs estimators by performing spatial averages on the celestial sphere and finds the one which has the smallest possible variance. We here show the {\\em best unbiased estimator} for the angular bispectrum ${\\cal C}_{\\ell_1 \\ell_2 \\ell_3 }$ and we display the corresponding cosmic variance as well. Recall that ${\\cal C}_{\\ell_1 \\ell_2 \\ell_3 }$ contains all the information available in the three-point correlation function, or its variants, like the skewness, collapsed and equilateral configurations\\cite{Ganetal94}. The present analysis borrows from recent joint work with J\\'er\\^ome Martin\\cite{GanMar00,GanMar01}, to whom I am greatly {\\em reconnaissant}. Expanding the anisotropies over the microwave sky as usual \\begin{equation} \\label{expSW} \\Delta ({\\bf e})=\\sum _{\\ell m} a_{\\ell }^{m}\\Ylm{}({\\bf e}) \\end{equation} the first three moments can be written as \\begin{equation} \\label{propa} \\bigl\\langle a_{\\ell }^{m} \\bigr\\rangle =0, \\quad \\bigl\\langle a_{\\ell _1 }^{m_1}a_{\\ell _2 }^{m_2*} \\bigr\\rangle = {\\cal C}_{\\ell_1}\\delta _{\\ell_1\\ell_2}\\delta _{m_1m_2}, \\quad \\bigl\\langle a_{\\ell_1 }^{m_1} a_{\\ell_2}^{ m_2} a_{\\ell_3}^{ m_3} \\bigr\\rangle = \\left(^{\\ell_1~\\,\\;\\ell_2~\\,\\;\\ell_3}_{m_1~m_2~m_3}\\right) {\\cal C}_{\\ell_1 \\ell_2 \\ell_3 }, \\end{equation} where $\\left(^{\\ell_1~\\,\\;\\ell_2~\\,\\;\\ell_3}_{m_1~m_2~m_3}\\right)$ is a Wigner 3$j$-symbol. The second equation ensures the isotropy of the CMB. The quantity $\\bigl\\langle a_{\\ell _1 }^{m_1}a_{\\ell _2 }^{m_2*} \\bigr\\rangle$ is the second moment of the $\\almn$'s and ${\\cal C}_\\ell$ is usually called the angular spectrum. In the third equation, the quantity $\\bigl\\langle a_{\\ell_1 }^{m_1} a_{\\ell_2}^{ m_2} a_{\\ell_3}^{ m_3} \\bigr\\rangle$ is the third moment while ${\\cal C}_{\\ell_1 \\ell_2 \\ell_3 }$ is called the angular bispectrum. The presence of the 3$j$-symbol guarantees that the third moment vanishes unless $m_1+m_2+m_3=0$ and $|\\ell _i-\\ell_j| \\le \\ell _k \\le \\ell _i+\\ell _j$. Moreover, invariance under spatial inversions of the three-point function implies the additional rule \\cite{luo94,GanMar00} $\\ell _1 +\\ell _2 +\\ell _3=\\mbox{even}$, in order for the third moment not to vanish. Finally, from this last relation and using standard properties of the 3$j$-symbols, it follows that the angular bispectrum is left unchanged under any arbitrary permutation of the indices $\\ell_i$. Let us call ${\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})$ the estimator for the angular bispectrum ${\\cal C}_{\\ell _1 \\ell_2 \\ell_3}$. The most general definition reads \\begin{equation} \\label{defE3} {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})\\equiv \\int \\int \\int {\\rm d}\\Omega _1 {\\rm d}\\Omega _2 {\\rm d}\\Omega _3 E_{\\rm S}^{\\ell _1 \\ell_2 \\ell_3}({\\bf e}_1,{\\bf e}_2,{\\bf e}_3) \\Delta({\\bf e}_1)\\Delta ({\\bf e}_2)\\Delta({\\bf e}_3). \\end{equation} where $E_{\\rm S}^{\\ell _1 \\ell_2 \\ell_3}$ is the weight function. The angular bispectrum is a real quantity and so is its estimator. Therefore, the weight function can be taken real. It is also symmetric under arbitrary permutations of directions ${\\bf e}_i$. In addition, like ${\\cal C}_{\\ell _1 \\ell_2 \\ell_3}$, the weight function satisfies $E_{\\rm S}^{\\ell _1 \\ell_2 \\ell_3} = E_{\\rm S}^{\\ell _2 \\ell_1 \\ell_3}$, as well as for any other arbitrary permutation of the indices $\\ell_i$. The weight function can be expressed on the basis of the spherical harmonics as \\begin{equation} \\label{defd3} E_{\\rm S}^{\\ell _1 \\ell_2 \\ell_3}({\\bf e}_1,{\\bf e}_2,{\\bf e}_3)= \\sum _{\\ell_1' m_1'} \\sum _{\\ell_2' m_2'} \\sum _{\\ell_3' m_3'} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'} {\\ell_1}{\\ell_2}{\\ell_3} Y_{\\ell_1'}^{m_1'}({\\bf e}_1) Y_{\\ell_2'}^{m_2'}({\\bf e}_2)Y_{\\ell_3'}^{m_3'}({\\bf e}_3). \\end{equation} The properties of the weight function imply that the coefficients $d$ must satisfy \\begin{equation} \\label{propd3} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'} {\\ell_1}{\\ell_2}{\\ell_3 *} = (-1)^{m_1'+m_2'+m_3'} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{-m_1'}{-m_2'}{-m_3'} {\\ell_1}{\\ell_2}{\\ell_3} \\quad , \\quad \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'} {\\ell_1}{\\ell_2}{\\ell_3} = \\coefdd{\\ell_2'}{\\ell_1'}{\\ell_3'}{m_2'}{m_1'}{m_3'} {\\ell_1}{\\ell_2}{\\ell_3}, \\end{equation} where the last relation is in fact valid for arbitrary permutations of any two columns of the collective subindex. Like the weight function, $d$ is also left invariant under arbitrary permutations of indices $\\ell_i$ (not primed). The estimator can be expressed in terms of the coefficients $d$ and the $a_{\\ell }^m$'s only: inserting the expansion of the weight function in the above expression for the estimator and using standard properties of the spherical harmonics one obtains \\begin{equation} \\label{esti3ad} {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3}) = \\sum _{\\ell_1' m_1'} \\sum _{\\ell_2' m_2'} \\sum _{\\ell_3' m_3'} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'} {\\ell_1}{\\ell_2}{\\ell_3 *} a_{\\ell _1'}^{m_1'}a_{\\ell _2'}^{m_2'}a_{\\ell _3'}^{m_3'} . \\end{equation} In practice, CMB observational settings are devised such that both the monopole and the dipole are subtracted from the anisotropy maps. This means that the coefficients $d$ in the last equation are only non-vanishing for indices $\\ell_i' \\ge 2$ in the collective subindex. Moreover, the coefficients $d$ satisfy $\\ell _1 +\\ell _2 +\\ell _3=\\mbox{even}$. We must now require that our general estimator given by Eq. (\\ref{esti3ad}) be unbiased, i.e. $\\langle {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})\\rangle ={\\cal C}_{\\ell _1 \\ell_2 \\ell_3}$. This forces the coefficients $d$ to fulfill the following constraint \\begin{equation} \\label{cons3} \\sum _{m_1'm_2'm_3'} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3 *} \\wjma{\\ell _1'}{\\ell _2'}{\\ell _3'}{m_1'}{m_2'}{m_3'} = \\delta _{\\rm S}^{\\ell_{i} \\ell_{j}'}, \\end{equation} where we have defined a symmetrized Kr\\\"onecker symbol for the $\\ell$ multipole indices only, as follows $ \\delta _{\\rm S}^{\\ell_{i} \\ell_{j}'} \\equiv \\frac{1}{6} ( \\delta_{\\ell_1\\ell_1'}\\delta_{\\ell_2\\ell_2'}\\delta_{\\ell_3\\ell_3'} +\\mbox{ 5 additional permutations } ). $ It is easy to check that the constraint equation satisfies the conditions imposed by Eqns. (\\ref{propd3}) on the coefficients $d$. Using the previous properties for $d$, relabelling the indices $m_1' \\leftrightarrow m_2'$ in Eq. (\\ref{cons3}) and finally noting that $\\ell_1'+\\ell_2'+\\ell_3'= \\ell_1+\\ell_2+\\ell_3=\\mbox{even}$, which allows us to permute any two columns of the Wigner 3$j$-symbol, one verifies that the left hand side of the constraint is invariant under $\\ell_1'\\leftrightarrow\\ell_2'$. The same applies for any pair of $\\ell$ multipole indices and this explains the presence of the symmetrized $\\delta_{\\rm S}^{\\ell_{i} \\ell_{j}'}$ in Eq. (\\ref{cons3}). We see from this that all coefficients $d$ that do not satisfy $\\ell_1'+\\ell_2'+\\ell_3'=\\mbox{even}$ do not enter the constraint. These terms only increase the variance (which we want to minimize) and as a consequence one can take them equal to zero. We are now in a position to calculate the variance of the estimator. Looking at Eq. (\\ref{esti3ad}) we see that this requires the computation of the sixth moment of the $\\almn$'s. After having made use of the properties of the coefficients $d$ and rearranging the resulting 15 terms into two groups, straightforward algebra yields \\bea \\label{var3} \\langle \\left[ {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})\\right]^2 \\rangle &=& \\sum _{\\ell_1' m_1'} \\sum _{\\ell_2' m_2'} \\sum _{\\ell_3' m_3'} {\\cal C}_{\\ell_1'}{\\cal C}_{\\ell_2'}{\\cal C}_{\\ell_3'} \\\\ &\\times & \\biggl[ 6 \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3 *} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3} + 9(-1)^{m_1'+m_2'} \\coefdd{\\ell_1'}{\\ell_1'}{\\ell_3'}{m_1'}{-m_1'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3 *} \\coefdd{\\ell_2'}{\\ell_2'}{\\ell_3'}{m_2'}{-m_2'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3} \\biggr] . \\nonumber \\eea The square of the variance of ${\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})$ is given by $\\sigma ^2 _{{\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})} = \\la \\left[ {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})\\right]^2 \\ra - \\la {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3}) \\ra^2$ . Since departures from Gaussianity are expected to be small (specially on large angular scales), higher moments will be calculated in the mildly non-Gaussian approximation. Within this approximation we can write $a_{\\ell}^{m} = a_{\\ell}^{m (0)} + \\epsilon \\, a_{\\ell}^{m (1)} + {\\cal O}(\\epsilon^2)$ where $a_{\\ell}^{m (0)}$ is a Gaussian random variable and the expansion parameter $\\epsilon$ is small. [The `$^{(0)}$' label will be dropped out hereafter] Now, the term $\\la \\left[ {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})\\right]^2 \\ra$ is of order $\\epsilon^0$ whereas the lowest non-vanishing order of $\\la {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3}) \\ra^2$ is $\\epsilon^2$. Therefore, the latter one will not enter the minimization procedure and the variance squared will be written as $\\sigma ^2 _{{\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})} \\approx \\la \\left[ {\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})\\right]^2 \\ra$. This does not occur for the two-point correlator\\cite{GriM}; in that case both terms contributing to the square of the variance are of the same order in $\\epsilon$ . With a bit of effort one can see that the various contributions of the imaginary part of the coefficients $d$ to the two terms, $6 d^* d$ and $9 d^* d$, only increase the variance. Since we know that a vanishing imaginary part does satisfy the constraint Eq. (\\ref{cons3}), it can be disregarded in the sequel. Therefore, Eq. (\\ref{var3}) can then be written solely in terms of {\\em real} coefficients $d$ as follows \\[ \\sigma ^2 _{{\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})} = \\sum _{\\ell_1' m_1'} \\sum _{\\ell_2' m_2'} \\sum _{\\ell_3' m_3'} {\\cal C}_{\\ell_1'}{\\cal C}_{\\ell_2'}{\\cal C}_{\\ell_3'} \\biggl[6 \\biggl( \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3 } \\biggr)^2 +9(-1)^{m_1'+m_2'} \\coefdd{\\ell_1'}{\\ell_1'}{\\ell_3'}{m_1'}{-m_1'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3} \\coefdd{\\ell_2'}{\\ell_2'}{\\ell_3'}{m_2'}{-m_2'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3} \\biggr]. \\] Our {\\em next move} now is to minimize this variance with respect to the coefficients $d$, taking into account the constraint of Eq. (\\ref{cons3}) \\begin{equation} \\label{mini3B} \\delta \\biggl\\{\\sigma ^2 _{{\\cal E}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})} +\\sum _{\\ell _1'\\ell _2'\\ell _3'} \\lambda _{\\ell _1'\\ell _2'\\ell _3'}^{\\ell _1\\ell _2\\ell _3} \\biggl[ \\sum _{m_1'm_2'm_3'} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'} {\\ell _1}{\\ell_2}{\\ell _3} \\wjma{\\ell _1'}{\\ell _2'}{\\ell _3'}{m_1'}{m_2'}{m_3'} - \\delta_{\\rm S}^{\\ell_{i} \\ell_{j}'} \\biggr]\\biggr\\}=0 . \\end{equation} Performing the variation ${\\rm \\delta }$ having in mind that the symmetries of the coefficients $d$ must be respected, we get \\begin{eqnarray} \\label{mini3resu} &&12{\\cal C}_{\\ell _1'}{\\cal C}_{\\ell _2'}{\\cal C}_{\\ell _3'} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'}{\\ell _1}{\\ell _2}{\\ell _3} + \\lambda _{\\ell _1'\\ell _2'\\ell _3'}^{\\ell _1\\ell _2\\ell _3} \\wjma{\\ell _1'}{\\ell _2'}{\\ell _3'}{m_1'}{m_2'}{m_3'} \\\\ &+& 6(-1)^{m_2'}{\\cal C}_{\\ell_2'}{\\cal C}_{\\ell_3'} \\delta_{\\ell_1'\\ell_2'}\\delta_{m_1'-m_2'} \\sum _{\\ell m}{\\cal C}_{\\ell} (-1)^{m} \\coefdd{\\ell}{\\ell}{\\ell_3'}{\\ m \\ }{\\ -m \\ }{\\ m_3' \\ } {\\ell_1 }{\\ell _2 }{\\ell _3 } + {\\tiny \\Big[\\begin{array}{c} 1'\\to 2' \\\\ 2'\\to 3' \\\\ 3'\\to 1' \\end{array} \\Big]} + {\\tiny \\Big[\\begin{array}{c} 1'\\to 3' \\\\ 2'\\to 1' \\\\ 3'\\to 2' \\end{array} \\Big]} =0 . \\nonumber \\end{eqnarray} [$\\dots$] terms are shorthand for the first one on the second line. This formula, together with Eq. (\\ref{cons3}), form a set of equations which completely determines the best unbiased estimator. {}From this last equation and using the constraint Eq. (\\ref{cons3}) we can get the general expression for the Lagrange multipliers. Thus, we multiply Eq. (\\ref{mini3resu}) by the appropriate 3$j$-symbol and we sum over the three indices $m_i'$. The first term is exactly the constraint and produces a $\\delta_{\\rm S}^{\\ell_{i} \\ell_{j}'}$. Using the fact that a triple sum over the $m_i$'s of the squared of a 3$j$-symbol gives unity, the second term yields the Lagrange multipliers themselves. Unfortunately, I don't have enough space to show that the last three terms vanish. Then, the Lagrange multipliers are given by \\begin{equation} \\label{LagMul3} \\lambda _{\\ell _1'\\ell _2'\\ell _3'}^{\\ell _1\\ell _2\\ell _3}= -12{\\cal C}_{\\ell _1'}{\\cal C}_{\\ell _2'}{\\cal C}_{\\ell _3'} \\delta_{\\rm S}^{\\ell_{i} \\ell_{j}'} . \\end{equation} Plugging this into Eq. (\\ref{mini3resu}), one has \\begin{eqnarray} \\label{mini3resubis} &&12{\\cal C}_{\\ell _1'}{\\cal C}_{\\ell _2'}{\\cal C}_{\\ell _3'} \\biggl[ \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'}{\\ell _1}{\\ell_2}{\\ell _3} - \\delta_{\\rm S}^{\\ell_{i} \\ell_{j}'} \\wjma{\\ell _1'}{\\ell _2'}{\\ell _3'}{m_1'}{m_2'}{m_3'} \\biggr] \\\\ &+& 6(-1)^{m_2'}{\\cal C}_{\\ell_2'}{\\cal C}_{\\ell_3'} \\delta_{\\ell_1'\\ell_2'}\\delta_{m_1'-m_2'} \\sum _{\\ell m}{\\cal C}_{\\ell} (-1)^{m} \\coefdd{\\ell}{\\ell}{\\ell_3'}{\\ m \\ }{\\ -m \\ }{\\ m_3' \\ } {\\ell_1 }{\\ell _2 }{\\ell _3 } + {\\tiny \\Big[\\begin{array}{c} 1'\\to 2' \\\\ 2'\\to 3' \\\\ 3'\\to 1' \\end{array} \\Big]} + {\\tiny \\Big[\\begin{array}{c} 1'\\to 3' \\\\ 2'\\to 1' \\\\ 3'\\to 2' \\end{array} \\Big]}=0 . \\nonumber \\eea This is the final equation to be solved in order to determine the best unbiased estimator. A solution is \\begin{equation} \\label{sold3} \\coefdd{\\ell_1'}{\\ell_2'}{\\ell_3'}{m_1'}{m_2'}{m_3'}{\\ell_1 }{\\ell_2 }{\\ell_3 } =\\wjma{\\ell _1'}{\\ell _2'}{\\ell _3'}{m_1'}{m_2'}{m_3'} \\delta_{\\rm S}^{\\ell_{i} \\ell_{j}'} , \\end{equation} which leads to \\begin{equation} \\label{solest3} {\\cal E}_{\\rm Best}({\\cal C}_{\\ell_1 \\ell_2 \\ell_3 }) = \\sum _{m_1' m_2' m_3'} \\wjma{\\ell_1 }{\\ell_2 }{\\ell_3 }{m_1'}{m_2'}{m_3'} a_{\\ell_1 }^{m_1'}a_{\\ell_2 }^{m_2'}a_{\\ell_3 }^{m_3'} . \\end{equation} Seems familiar? An estimator restricted to the diagonal case $\\ell _1=\\ell _2=\\ell _3$ (and then extended to $\\ell_2 = \\ell_1+2$ and $\\ell_3 = \\ell_1-2$) has been proposed \\cite{Feretal98,newturn} for ${\\cal B}_{\\ell }\\equiv {\\cal C}_{\\ell \\ell \\ell}$. ~\\footnote{${\\cal B}$ like ${\\cal B}$ispectrum (like spe${\\cal C}$trum) $\\ldots$ or ${\\cal C}$, ${\\cal B}$, ${\\cal A}$?, $\\ldots$ . Hope the ${\\cal T}$rispectrum will be called ${\\cal T}$. \\\\ Still, since no ambiguity arises I stick to ${\\cal C}_{\\ell_1 \\ell_2 \\ell_3 }$ (for now).} The aim of these authors was not to seek the best estimator, but to use the corresponding, say, ${\\cal E}({\\cal B}_{\\ell })$ to analyse the non-Gaussian features of the 4-yr COBE-DMR data (see also\\cite{Hea98,hobson98,BroTeg99,BZG}). While their estimator is {\\em not} unbiased, for it does not satisfy the constraint (\\ref{cons3}), by just removing an overall prefactor one gets our best unbiased estimator ${\\cal E}_{\\rm Best}({\\cal C}_{\\ell_1 \\ell_2 \\ell_3 })$, Eq. (\\ref{solest3}). We now know the best unbiased estimator for ${\\cal C}_{\\ell_1 \\ell_2 \\ell_3 }$ and then we can compute its variance, the smallest one amongst all possible estimator variances, which yields \\be \\label{vari2gm} \\sigma ^2 _{{\\cal E}_{\\rm Best}({\\cal C}_{\\ell _1 \\ell_2 \\ell_3})} = {\\cal C}_{\\ell_1} {\\cal C}_{\\ell_2} {\\cal C}_{\\ell_3} (1+\\delta_{\\ell_1\\ell_2}+\\delta_{\\ell_2\\ell_3}+\\delta_{\\ell_3\\ell_1} + 2 ~ \\delta_{\\ell_1\\ell_2}\\delta_{\\ell_2\\ell_3}) . \\ee We like to dub this (the square of) the `bispectrum cosmic variance' in perfect analogy with $\\sigma^2_{{\\cal E}_{\\rm Best}({\\cal C}_{\\ell})} = 2 {\\cal C}_{\\ell}^2 / (2\\ell+1)$, which is (the square of) the variance of the best unbiased estimator for the angular spectrum, commonly known as the `cosmic variance'. \\noindent {\\em Acknowledgments} \\noindent I'd like to thank my collaborator J. Martin for extensive discussions and A. Heavens for useful correspondence. This work was partially financed with funds from the World Lab. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003045_arXiv.txt": { "abstract": "HD\\,89744 is an F7 V star with mass 1.4 \\Msun, effective temperature 6166~K, age 2.0 Gy and metallicity [Fe/H]$=0.18$. The radial velocity of the star has been monitored with the AFOE spectrograph at the Whipple Observatory since 1996, and evidence has been found for a low mass companion. The data were complemented by additional data from the Hamilton spectrograph at Lick Observatory during the companion's periastron passage in fall 1999. As a result, we have determined the star's orbital wobble to have period $P = 256$d, orbital amplitude $K = 257$ m/s, and eccentricity $e=0.7$. From the stellar mass we infer that the companion has minimum mass \\msini $= 7.2$ \\Mjup\\ in an orbit with semi-major axis $a_2 = 0.88$ AU. The eccentricity of the orbit, among the highest known for extra-solar planets, continues the trend that extra-solar planets with semi-major axes greater than about 0.15 AU tend to have much higher eccentricities than are found in our solar system. The high metallicity of the parent star reinforces the trend that parent stars of extra-solar planets tend to have high metallicity. ", "introduction": "We report on the detection of a massive (\\msini $= 7.2$ \\Mjup) planet in a highly elliptical ($e=0.7$), 256 day orbit about the star HD\\,89744 (HR\\,4067, HIP\\,50786), from radial velocity variations which reveal Keplerian motions of the star. Observations were carried out from 1996 through 1999 using the Advanced Fiber Optic Echelle (AFOE) spectrograph \\citep{BrownEtal94, NisensonEtal98}, a bench-top spectrograph located at the Whipple Observatory 1.5m telescope, and also with the Hamilton spectrograph at the Lick Observatory CAT and Shane telescopes, in November and December of 1999. The AFOE spectrograph is designed primarily for precise radial velocity studies of the seismology of bright stars, and of reflex motions of stars due to planetary companions. Long term stability of the velocity reference is provided by use of an iodine ($I_2$) cell \\citep{ButlerEtal96}. The AFOE determines radial velocity variations induced by planetary companions with a precision and long-term accuracy of approximately 10 m/s. On the order of 100 relatively bright stars ($m_v \\leq 7$) have been monitored for this purpose since 1995. Since 1995 when the planetary candidate oribiting the star 51 Pegasus was detected \\citep{MayorQueloz95}, some 29 additional candidates have been detected by several groups, all from Doppler shifts measured using precise radial velocity techniques \\citep{MarcyButler98, MayorEtal98, NoyesEtal97, CochranEtal97}. HD\\,89744 (F7 V) was added to the AFOE observing list in early 1996, based on its relatively low chromospheric emission as measured with the Mt.\\ Wilson ``HK'' chromospheric activity monitoring program \\citep{BaliunasEtal95}. AFOE observations have been obtained regularly since then, and indicated the presence of a planet with a highly eccentric orbit. However data near the companion's periastron, critical to an accurate determination of the orbital parameters, were not obtained until late 1999. Between October and December 1999, while the companion was near periastron, observations were made at Lick Observatory as well as with the AFOE, to ensure good phase coverage. The data points taken with the Lick CAT and Shane telescopes agree extremely well with the AFOE data, and thus provide a confirmation of the detection along with a precise determination of the ellipticity of the planet's orbit. ", "conclusions": "The radial velocity data shown in Figure~\\ref{fig:phasePlot} are unambiguous in revealing a periodic radial velocity variation of HD\\,89744, which can be fit well by a Keplerian orbit. Since observations obtained with two different radial velocity instruments at two different telescopes yield exactly the same radial velocity variations within their respective instrumental errors, the evidence is compelling that the measured velocities are real variations on the star. There is no known way a stellar signal in a late F-type main sequence star could mimic a Keplerian orbital signature with such a long period and large amplitude, and with such a large eccentricity. Hence we are driven to the interpretation that the star is orbited by a low-mass companion, HD\\,89744~b, (\\msini $= 7.2 M_{\\odot}$), in an orbit with semi-major axis 0.88 AU and eccentricity 0.70. The residuals to the orbital fit are larger than would be expected from internal errors in the data; this is true both for the AFOE data and the Hamilton echelle data. However, \\citet{SaarEtal98} conclude that for a typical F stars with this rotation period, the velocity jitter induced by stellar magnetic activity and inhomogeneous convection is approximately 10 m/s. Adding this jitter in quadrature to the uncertainties lower the reduced $\\chi$ to 1.2. The residuals display a long term trend ($\\approx 15$ m/s/year) that is marginally significant. While it might be caused by a residual instrumental drift in the AFOE data, we can not rule out that it might be due to a distant companion; further observations over a longer baseline are required. The orbital eccentricity of the companion to HD\\,89744 is among the highest planetary eccentricities known. Only two other planets, 16\\,Cyg\\,B~b \\citep[$e = 0.68$, $a_2 = 1.70$ AU;][]{CochranEtal97} and HD\\,222582~b \\citep[$e = 0.71$, $a_2 = 1.35$ AU;][]{VogtEtal99} have comparable eccentricities. HD\\,89744~b, with $a_2 = 0.88$ AU, has the smallest semi-major axis of the three. At periastron it dips to within 0.26 AU, still well outside a periastron distance which could lead to tidal circularization within the stellar lifetime. The discovery of the first highly eccentric planet, 16\\,Cyg\\,B~b, led to the suggestion \\citep{MazehEtal97, HolmanEtal97} that its eccentricity may have been ``pumped up'' by the influence of a nearby companion star, 16\\,Cyg\\,A. However, neither HD\\,89744 nor HD\\,222582 is orbited by a stellar companion. Thus a different explanation is required for their high eccentricities, an explanation that might also apply to 16\\,Cyg\\,B~b. The high orbital eccentricity of the HD\\,89744 system continues the trend that planetary-mass companions whose semi-major axes are greater than about 0.15 AU tend to have a broad range of eccentricities, with no apparent trends of eccentricity with mass or semi-major axis. This circumstance must be explained by any successful planetary formation and migration scenario. As noted by others \\citep[\\eg,][]{VogtEtal99, MarcyEtal99}, this causes difficulties with a number of proposed mechanisms for planetary formation and migration. The values of $v$sin$i$, $P_{\\rm rot}$, and $R$ for HD\\,89744 given in Table~\\ref{table:stellarProps} imply an inclination of the stellar rotational equator of $i= 42^{\\circ}$ (\\ie, \\sini = 0.66). Moreover, if we assume that the orbit is coplanar with the star's equatorial plane, we infer that $m_2$ = 10.8 \\Mjup. The astrometric orbital amplitude would be 0.17 mas, too small for Hipparcos detection but within the range of next-generation astrometric missions. The mass of 10 \\Mjup\\ suggested by the stellar rotational data is near the upper limit of masses associated with extra-solar giant planets \\citep[\\eg,][]{MarcyEtal00}. If such large masses hold up to further investigations, then theoretical understanding of the origin and evolution of extra-solar giant planets must be able to accommodate a mass range spanning at least values between 0.5 \\Mjup\\ and 10 \\Mjup. The metallicity of HD\\,89744, [Fe/H]= 0.18, is substantially higher than the mean for nearby sun-like stars \\citep{FavataEtal97, Gonzalez98}. HD\\,89744 was placed on the AFOE observing list without reference to its metallicity; therefore the association of its high metallicity with the presence of a planet is not a selection effect. This association continues the trend, already noted \\citep[\\eg,][ and references therein]{Gonzalez+Laws99} and references therein that the metallicity of stars with planets tends to be higher than that of stars without planets." }, "0003/astro-ph0003273_arXiv.txt": { "abstract": "We report optical spectroscopic identifications of 10 hard (2--10 keV) X-ray selected sources discovered by {\\it Chandra}. The X-ray flux of the sources ranges between 1.5 and 25 $\\times10^{-14}$ \\cgs, the lower value being 3 times fainter than in previous BeppoSAX and ASCA surveys. Their R band magnitudes are in the range 12.8--22. Six of the {\\it Chandra} sources are broad line quasars with redshifts between 0.42 and 1.19, while the optical identification of the remaining four is quite varied: two are X-ray obscured, emission line AGN at $z$=0.272 and $z$=0.683, one is a starburst galaxy at $z$=0.016 and one, most unusually, is an apparently normal galaxy at $z$=0.158. These findings confirm and extend down to fainter X--ray fluxes the BeppoSAX results, in providing samples with a wide range of X-ray and optical properties. The ratio between the soft X-ray and the optical luminosity of the $z$=0.158 galaxy is a factor at least 30 higher than that of normal galaxies, and similar to those of AGN. The high X--ray luminosity and the lack of optical emission lines suggest an AGN in which either continuum beaming dominates, or emission lines are obscured or not efficiently produced. ", "introduction": "The {\\it Chandra} X-ray Observatory was launched on July 23 1999, carrying on board a revolutionary high resolution mirror assembly, with a Point Spread Function of 0.5 arcsec (half power radius) over the broad 0.1 to 10 keV band (Van Speybroeck, et al. 1997). This, together with the aspect camera which at the moment provides attitude solutions with errors of the order of 1-2 arcsec\\footnote{When the data are definitively reprocessed, the aspect for image reconstruction should be $\\sim0.5$ arcsec and the source positions should be better than 1 arcsec}, allows the study of spatial extent of X-ray sources on similar scales, i.e. smaller or similar to the size of a L$^*$ galaxy at any redshift; and gives X-ray source positions at least as good as 2--3 arcsec, immediately allowing the unambiguous identification of the optical counterparts of faint X-ray sources. Consequently, the determination of the source redshifts via optical spectroscopy becomes highly efficient. The improvement provided by {\\it Chandra} is especially significant in the hard (2--10 keV) X-ray band. Surveys of the hard X-ray sky have been performed in the past by ASCA and BeppoSAX (Ueda et al. 1998, Della Ceca et al. 1999, Fiore et al. 2000a, Giommi et al. 2000, Comastri et al. 2000). However, the large error boxes (1-2 arcmin) limited the optical identification process to classes of objects with low surface density, at a given optical magnitude limit. As a result, most of the identified sources are emission line AGN (Fiore et al. 1999, Akiyama et al. 2000, and La Franca et al. in preparation). About 30 \\% of the BeppoSAX HELLAS survey sources studied spectroscopically down to R=20.5 have escaped a secure identification (Fiore et al. 2000b, La Franca et al. in preparation), although many normal galaxies and stars have been observed in these error-boxes. {\\it Chandra}'s unprecedented capabilities make identifications unambiguous, and open up the possibility of searching for and studying classes of sources not previously recognized as strong hard X-ray emitters, and of assessing their contribution to the hard X-ray cosmic background (XRB; e.g. Griffiths \\& Padovani 1990). In particular, it will be possible for the first time to begin studying normal galaxies at $z>0.1$, as well as possible ``minority'' hard X-ray source populations (Kim \\& Elvis 1999). We have started a pilot project of spectroscopic identification of {\\it Chandra} sources in two medium--deep fields that were visible from La Silla in January 2000, with the aim of verifying the feasibility of such studies with 4m class telescopes. The results on 10 X-ray sources are very encouraging and are described in the following. ", "conclusions": "We have carried out a pilot program to study the faint hard X-ray source population using the revolutionary capabilities of the {\\it Chandra} satellite. We identified ten 2--10 keV selected sources from two 4-chip, medium deep {\\it Chandra} fields covering about 0.14 deg$^2$ of sky at fluxes in the range $1.5-25\\times10^{-14}$ \\cgs, a factor of 3 fainter than previous ASCA and BeppoSAX surveys. Recently, Mushotzky et al. (2000) reported detection of faint X-ray sources from a single 1--chip field (0.0175 deg$^2$) at fluxes $0.3-3\\times10^{-14}$ \\cgs. Our results fill the gap between the shallow BeppoSAX and ASCA surveys and the deep {\\it Chandra} field. Almost all sources have an optical counterpart within 2 arcsec. Optical spectra allow us to measure their redshifts, and assess their optical classification. We find six broad line quasars, two emission line AGN (LAR6 and P4), one starburst galaxy (LAR5), and one apparently normal galaxy at $z$=0.158 (P3). LAR6 and P4 are likely to be obscured in X-ray by a column densities of $\\approx10^{23}$ cm$^{-2}$ and $\\approx10^{22}$ cm$^{-2}$ respectively. The X-ray source in the $z$=0.158 normal galaxy P3 may be covered by a column density of about $10^{22}$ cm$^{-2}$ too. The spatial extension of all X-ray sources is smaller than a few arcsec in all cases, and it is roughly consistent with the {\\it Chandra} PSF. The X-ray sources associated with the $z$=0.016 starburst galaxy LAR5 and the $z$=0.158 normal galaxy P3 appear coincident with the galaxy nuclei. The X-ray to optical luminosity ratio of P3 is higher by a factor of at least 30 than those of normal galaxies, while it is similar to those of AGN. The high X-ray luminosity and the lack of optical emission lines suggest an AGN in which either continuum beaming is important or emission lines are absorbed or not efficiently produced. In any case, objects like P3 would be missed or ignored in optical surveys or in X-ray surveys with large error boxes. It is only thanks to the new revolutionary capabilities of {\\it Chandra} that this kind of sources {\\it can} be detected and identified. Based on the ASCA and BeppoSAX surveys at fluxes $>5\\times10^{-14}$ \\cgs and on our first {\\it Chandra} identifications, which push the flux limit down to $\\sim2\\times10^{-14}$ \\cgs, the hard X-ray sky appears populated by a large fraction of broad line AGN (about 50\\%), by a mixture of intermediate AGN (type 1.8-2.0 and composite starburst/AGN) and, most intriguingly, also by X-ray luminous apparently normal galaxies. These populations span ranges of X-ray and optical properties wider than previously thought. \\bigskip {\\it Acknowledgements} The results presented in this paper are made possible by the successful effort of the entire {\\it Chandra} team. In particular we thank the XRT and ACIS teams for building and calibrating the high resolution mirror and the CCD camera, the CXC team, and in particular A. Fruscione, for the quick data reduction and archiving. We thank P. Giommi, S. Molendi, G. Fabbiano, E. Giallongo, M. Mignoli, L. Stella, M. Vietri and H. Tananbaum for useful discussions and L.A. Antonelli for help in the ESO observation preparation. We also thank the referee, Gianni Zamorani, for his detailed and constructive comments which improved the quality of the paper. This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. This work is partly supported by the Italian Space Agency, contract ARS--99--75 and by the Ministry for University and Research (MURST) under grant COFIN--98--02--32. M.E., F.N, and M.C. acknowledge support from NASA contract NAS8--39073." }, "0003/astro-ph0003051_arXiv.txt": { "abstract": "Using a simple combinatorial algorithm for generating finite and discrete events as our numerical cosmology, we predict that the baryon/photon ratio at the time of nucleogenesis is $\\eta= 1/256^4$, $\\Omega_{DM}/\\Omega_B= 12.7$ and (for a cosmological constant of $\\Omega_{\\Lambda}=0.6\\pm 0.1$ predicted on general grounds by E.D.Jones) that $0.325 > \\Omega_M > 0.183 $. The limits are set not by our theory but by the empirical bounds on the renormalized Hubble constant of $0.6 < h_0 < 0.8$. If we impose the additional empirical bound of $t_0 < 14 \\ Gyr$, the predicted upper bound on $\\Omega_M$ falls to $0.26$. The predictions of $\\Omega_M$ and $\\Omega_{\\Lambda}$were in excellent agreement with Glanz' analysis in 1998, and are still in excellent agreement with Lineweaver's recent analysis despite the reduction of observational uncertainty by close to an order of magnitude. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003098_arXiv.txt": { "abstract": "We present $UBVRIz^\\prime$ photometry of the gra\\-vitational lens candidate CLASS \\object{B1152+119} obtained with the Nordic Optical Telescope. The two QSO components are resolved in the $B$, $V$, $R$, $I$ and $z^\\prime$ bands confirming the lensing nature of the system. The $z=0.44$ lens galaxy is clearly detected in $B$, $R$, $I$ and $z^\\prime$ and its position is found to be almost coincident with the faint QSO image which is heavily extincted (relative to the brighter QSO image) by dust in the lens galaxy. The extinction curve of the lens galaxy derived from the relative photometry is well fitted by a Galactic extinction law with $1.3 \\lesssim R_V \\lesssim 2.0$ and $E(B-V) \\approx 1$. From a simple model of the system we predict a time delay of $\\sim$ 60 days. ", "introduction": "Dust is ubiquitous in the Universe and is responsible for extinction of the light from distant sources. The type of dust (composition, grain sizes and shape) determines the amount of extinction as a function of wavelength -- the extinction curve $A_{\\lambda}$. Extinction curves almost certainly evolve with redshift since the metallicity, elemental abundance ratios, mean star-formation rate, and energy-injec\\-tion rates that determine the structure and evolution of the dust are all strong functions of redshift. The extinction in our Galaxy has been well studied (Cardelli et al. 1989), but very little is known about extinction in other galaxies, especially at higher redshifts. This is unfortunate as dust plays an increasingly important role in contemporary cosmology. Recently the use of multiply imaged QSOs has been explored as a means of inferring differential extinction curves for distant galaxies (Nadeau et al.~\\cite{nadeau}; Malhotra et al.~\\cite{malhotra}; Falco et al.~\\cite{Falco}). The idea behind this approach is that a lensed QSO can be used as a `standard beacon', shining through different paths of the lensing galaxy. If one of the images suffers negligible extinction, or if the extinction curve is the same throughout the galaxy, then the relative intensity ratios of any two images as a function of wavelength is a direct measure of the extinction curve. As a part of the CLASS radio survey, Myers et al. (\\cite{Myers}) reported the discovery of a new gravitational lens candidate, CLASS B1152+199, with an image separation of $\\theta =1\\farcs 56$ and a flux ratio of $3.03\\pm 0.03$ at 8.46 GHz. Spectra obtained with the Keck II telescope revealed a background quasar at $z=1.019$ and a foreground galaxy at $z=0.439$. Optical follow-up observations revealed a bright unresolved object ($g\\simeq 16.5$ and $i\\simeq 16.6$) located at the radio position, presumably the brighter of the two lensed images. The non-detection of the weak image in $g$ and $i$ is indicative of a large color difference between the two images. If \\object{B1152+199} is indeed a gravitationally lensed system this could be attributed to differential extinction caused by dust in the lens, making \\object{B1152+199} a well-suited system for studying the extinction law of the lensing galaxy. ", "conclusions": "Extinction curves have been derived for the Galaxy and the Magellanic Clouds, mainly through multi-color photometry of individual stars, for which the intrinsic luminosities is known from their spectral types (Cardelli et al. \\cite{Cardelli}). The extinction curves of the LMC and SMC can be fitted by the Galactic extinction law (see Mathis~\\cite{mathis}; Fitzpatrick~\\cite{fitz}). The main difference is the 2175~\\AA\\ bump (cf. Fig.~\\ref{ext}) which is not observed in the LMC and SMC. At distances $\\gtrsim$ 10 Mpc study of individual stars is difficult and other methods must be considered. In this paper we have shown that accurate extinction curves of high-$z$ galaxies can be determined by studying gravitationally lensed systems. We have found that the extinction curve of \\object{B1152+199} is well fitted by the Galactic extinction law. It is interesting to note that the inferred $1.3 \\lesssim R_V \\lesssim 2.1$ in principle is outside the validity of the parameterization of Galactic extinction curve by Cardelli et al.~(\\cite{Cardelli}). Note also that the differential extinction of \\object{B1152+199} $0.9 \\lesssim E(B-V) \\lesssim 1.1$, is one of the largest observed to date (Falco et al.~\\cite{Falco}). The evidence for a Galactic extinction curve is intriguing, but $U$ and $B$ band observations with longer exposure times and better seeing are needed to rule out other extinction curves. These will sample the restframe UV radiation which is highly sensitive to dust extinction, and thus strongly constrain the shape of the extinction curve, and allow for the possible detection of the 2175 \\AA\\ bump. The predicted time delay of 59 days makes \\object{B1152+199} an interesting candidate for monitoring." }, "0003/astro-ph0003321_arXiv.txt": { "abstract": "s#1#2#3{{ \\centering{\\begin{minipage}{4.5in}\\footnotesize\\baselineskip=10pt \\parindent=0pt #1\\par \\parindent=15pt #2\\par \\parindent=15pt #3 \\end{minipage}}\\par}} \\def\\keywords#1{{ \\centering{\\begin{minipage}{4.5in}\\footnotesize\\baselineskip=10pt {\\footnotesize\\it Keywords}\\/: #1 \\end{minipage}}\\par}} \\newcommand{\\bibit}{\\nineit} \\newcommand{\\bibbf}{\\ninebf} \\renewenvironment{thebibliography}[1] {\\frenchspacing \\ninerm\\baselineskip=11pt \\begin{list}{\\arabic{enumi}.} {\\usecounter{enumi}\\setlength{\\parsep}{0pt} \\setlength{\\leftmargin 12.7pt}{\\rightmargin 0pt}% \\setlength{\\itemsep}{0pt} \\settowidth {\\labelwidth}{#1.}\\sloppy}}{\\end{list}} \\newcounter{itemlistc} \\newcounter{romanlistc} \\newcounter{alphlistc} \\newcounter{arabiclistc} \\newenvironment{itemlist} {\\setcounter{itemlistc}{0} \\begin{list}{$\\bullet$} {\\usecounter{itemlistc} \\setlength{\\parsep}{0pt} \\setlength{\\itemsep}{0pt}}}{\\end{list}} \\newenvironment{romanlist} {\\setcounter{romanlistc}{0} \\begin{list}{$($\\roman{romanlistc}$)$} {\\usecounter{romanlistc} \\setlength{\\parsep}{0pt} \\setlength{\\itemsep}{0pt}}}{\\end{list}} \\newenvironment{alphlist} {\\setcounter{alphlistc}{0} \\begin{list}{$($\\alph{alphlistc}$)$} {\\usecounter{alphlistc} \\setlength{\\parsep}{0pt} \\setlength{\\itemsep}{0pt}}}{\\end{list}} \\newenvironment{arabiclist} {\\setcounter{arabiclistc}{0} \\begin{list}{\\arabic{arabiclistc}} {\\usecounter{arabiclistc} \\setlength{\\parsep}{0pt} \\setlength{\\itemsep}{0pt}}}{\\end{list}} \\newcommand{\\fcaption}[1]{ \\refstepcounter{figure} \\setbox\\@tempboxa = \\hbox{\\footnotesize Fig.~\\thefigure. #1} \\ifdim \\wd\\@tempboxa > 5in {\\begin{center} \\parbox{5in}{\\footnotesize\\smalllineskip Fig.~\\thefigure. #1} \\end{center}} \\else {\\begin{center} {\\footnotesize Fig.~\\thefigure. #1} \\end{center}} \\fi} \\newcommand{\\tcaption}[1]{ \\refstepcounter{table} \\setbox\\@tempboxa = \\hbox{\\footnotesize Table~\\thetable. #1} \\ifdim \\wd\\@tempboxa > 5in {\\begin{center} \\parbox{5in}{\\footnotesize\\smalllineskip Table~\\thetable. #1} \\end{center}} \\else {\\begin{center} {\\footnotesize Table~\\thetable. #1} \\end{center}} \\fi} \\def\\@citex[#1]#2{\\if@filesw\\immediate\\write\\@auxout {\\string\\citation{#2}}\\fi \\def\\@citea{}\\@cite{\\@for\\@citeb:=#2\\do {\\@citea\\def\\@citea{,}\\@ifundefined {b@\\@citeb}{{\\bf ?}\\@warning {Citation `\\@citeb' on page \\thepage \\space undefined}} {\\csname b@\\@citeb\\endcsname}}}{#1}} \\newif\\if@cghi \\def\\cite{\\@cghitrue\\@ifnextchar [{\\@tempswatrue \\@citex}{\\@tempswafalse\\@citex[]}} \\def\\citelow{\\@cghifalse\\@ifnextchar [{\\@tempswatrue \\@citex}{\\@tempswafalse\\@citex[]}} \\def\\@cite#1#2{{$\\null^{#1}$\\if@tempswa\\typeout {IJCGA warning: optional citation argument ignored: `#2'} \\fi}} \\newcommand{\\citeup}{\\cite} \\def\\pmb#1{\\setbox0=\\hbox{#1} \\kern-.025em\\copy0\\kern-\\wd0 \\kern.05em\\copy0\\kern-\\wd0 \\kern-.025em\\raise.0433em\\box0} \\def\\mbi#1{{\\pmb{\\mbox{\\scriptsize ${#1}$}}}} \\def\\mbr#1{{\\pmb{\\mbox{\\scriptsize{#1}}}}} \\def\\fnm#1{$^{\\mbox{\\scriptsize #1}}$} \\def\\fnt#1#2{\\footnotetext{\\kern-.3em {$^{\\mbox{\\scriptsize #1}}$}{#2}}} \\def\\fpage#1{\\begingroup \\voffset=.3in \\thispagestyle{empty}\\begin{table}[b]\\centerline{\\footnotesize #1} \\end{table}\\endgroup} \\def\\runninghead#1#2{\\pagestyle{myheadings} \\markboth{{\\protect\\footnotesize\\it{\\quad #1}}\\hfill} {\\hfill{\\protect\\footnotesize\\it{#2\\quad}}}} \\headsep=15pt \\font\\tenrm=cmr10 \\font\\tenit=cmti10 \\font\\tenbf=cmbx10 \\font\\bfit=cmbxti10 at 10pt \\font\\ninerm=cmr9 \\font\\nineit=cmti9 \\font\\ninebf=cmbx9 \\font\\eightrm=cmr8 \\font\\eightit=cmti8 \\font\\eightbf=cmbx8 \\font\\sevenrm=cmr7 \\font\\fiverm=cmr5 \\newtheorem{theorem}{\\indent Theorem} \\newtheorem{lemma}{Lemma} \\newtheorem{definition}{Definition} \\newtheorem{corollary}{Corollary} \\newcommand{\\proof}[1]{{\\bf Proof.} #1 $\\Box$.} \\textwidth=5truein \\textheight=7.59truein \\def\\qed{\\hbox{${\\vcenter{\\vbox{\t % \\hrule height 0.4pt\\hbox{\\vrule width 0.4pt height 6pt \\kern5pt\\vrule width 0.4pt}\\hrule height 0.4pt}}}$}} \\renewcommand{\\thefootnote}{\\fnsymbol{footnote}} % \\begin{document} \\setlength{\\textheight}{7.7truein} % \\runninghead{C. Palomba} {Detectability of gravitational radiation $\\ldots$} \\normalsize\\textlineskip \\thispagestyle{empty} \\setcounter{page}{1} \\copyrightheading{}\t\t% \\vspace*{0.88truein} \\centerline{\\bf DETECTABILITY OF GRAVITATIONAL RADIATION FROM} \\vspace*{0.035truein} \\centerline{\\bf PROMPT AND DELAYED STAR COLLAPSE TO A BLACK HOLE} \\vspace*{0.37truein} \\centerline{\\footnotesize CRISTIANO PALOMBA} \\vspace*{0.015truein} \\centerline{\\footnotesize\\it Dipartimento di Fisica ``G. Marconi'', Universit\\`a di Roma ``La Sapienza''} \\baselineskip=10pt \\centerline{\\footnotesize\\it and Sezione INFN ROMA1, p.le A. Moro 5} \\baselineskip=10pt \\centerline{\\footnotesize\\it Roma, I-00185, Italy} \\vspace*{0.225truein} \\publisher{(received date)}{(revised date)} \\vspace*{0.21truein} \\abstracts{We consider the emission of gravitational waves in the two proposed models for the collapse of a massive star to a black hole: the {\\em prompt} collapse, in which nearly all the star collapses to a black hole in a dynamical time scale, and the {\\em delayed} collapse, in which a light black hole, or a neutron star, which subsequently accretes matter, forms due to the fall-back achieving, in the neutron star case, the critical mass for black hole formation. Recent simulations strongly support this last scenario. We show that, due to the slowness of fall-back, in the {\\em delayed} collapse the main burst of gravitational radiation is emitted depending on the parameters, mass and angular momentum, of the initial, light, black hole. We estimate, under different assumptions, the detectability of the emitted gravitational waves showing that such kind of collapse is not particularly suited for detection by forthcoming interferometric detectors. Detectors with high sensitivity at frequencies greater than $\\sim 4\\div 5~ kHz$ would be better suited for this kind of sources. We calculate also the final mass distribution function of single black holes.}{}{} \\vspace*{1pt}\\textlineskip\t% \\vspace*{-0.5pt} \\noindent ", "introduction": "\\noindent Gravitational collapse of a stellar core to a black hole has been studied since many years. Efforts have been also devoted to the calculation of the gravitational radiation emitted in this process. Most studies have been based on a perturbative approach,\\cite{cun1}$^,$\\cite{cun2}$^,$\\cite{cun3}$^,$\\cite{seid1}$^,$\\cite{seid2} others on the numerical solution of the full Einstein equations \\cite{stark}$^,$\\cite{stark2}. All these papers consider the collapse of a ``naked'' stellar core, described by a ``dust'' of particles or, at most, by a politropic equation of state, without taking into account the presence of the outer layers of the star, which are involved in the process. In particular, depending on the ratio between the energy released in the final explosion of a massive star and the binding energy of the ejected material, two different kinds of collapse have been outlined \\cite{woos}$^,$\\cite{kalo}: the {\\em prompt} collapse, in which a large fraction of the star collapses on a dynamical time-scale forming a massive black hole, and the {\\em delayed} collapse, in which a low mass black hole or, alternatively, a neutron star forms at the beginning and later accretes matter, due to fall-back. If a neutron star is the initial outcome of the collapse, fall-back pushes its mass above the critical mass and the formation of a black hole takes place. In both cases, this light black hole continues to slowly accrete matter until the final mass is reached. In the gravitational wave community the {\\em prompt} collapse has been considered for a long time as representative of realistic collapse processes. The estimated mass (several solar masses) of the first black hole candidates (like $Cygnus~ X-1$) has led to the idea that black holes should often be born with a \"typical\" mass of $\\sim 10M_\\odot $. This assumption appears no more justified now: we know that the evolution of massive stars in binary systems (to which all observed black hole candidates belong) is different from that of single stars; in addition, more refined observational techniques have allowed to find black hole candidates of few solar masses. In the light of these results and of the recent numerical simulations which we will describe, the {\\em prompt} birth of such massive black holes ($\\sim 10M_\\odot $ or more) should be considered as a very rare event. In this paper, we will discuss the {\\em delayed} collapse model, the expected mass distribution of isolated black holes and the detectability of the emitted signal, by forthcoming interferometric detectors. Our main aim is to understand how our perspectives of detection change with respect to the \"naive\" {\\em prompt} collapse. The plan of the paper is as follows. In Sec.2 we will shortly describe the {\\em delayed} collapse scenario. In Sec.3 we will estimate the contribution of fall-back to the total gravitational emission in the {\\em delayed} collapse discussing the consequences. In Sec.4 we will discuss the detectability of the emitted signal by forthcoming interferometric detectors, and compare with the {\\em prompt} collapse which could happen for progenitor stars above about $40M_\\odot $, if stellar winds were much less important in the evolution of massive stars than it is currently believed. In Sec.5 we will derive the theoretical final mass distribution function for isolated black holes, using the results of recent simulations of the collapse of massive single stars. Finally, in Sec.6 the results and their implications will be discussed. ", "conclusions": "In this paper we have considered the collapse of a massive star to a black hole, exploring the models of {\\em prompt} and {\\em delayed} collapse. According to recent simulations, the formation of stellar mass black holes should take place through a {\\em delayed} collapse. The {\\em prompt} collapse could happen for very massive progenitors (mass greater than $\\sim 40M_\\odot$) only if stellar winds were negligible, an assumption which appears to be not very reliable. In the {\\em delayed} collapse all black holes were born with a mass equal to its minimum value, or just a little greater, and then slowly accrete matter up to their final mass. We have shown that the main burst of gravitational radiation is emitted when the black hole forms so that the gravitational energy spectrum is peaked in the range $\\sim 4.5\\div 9~kHz$, depending on the initial mass and the angular momentum of the black hole. Such frequencies do not match very well with the sensitivity curve of ground-based interferometers which reach their best sensitivity in the band $\\sim 60\\div 1000~Hz$. We have estimated the detectability of the emitted gravitational radiation in the {\\em delayed} case, the most reliable scenario, and in the {\\em prompt} case, for progenitor stars more massive than $40M_\\odot $. For each collapse model, we have considered different values of black hole minimum mass and angular momentum and different laws for progenitors mass distribution. We have also derived the theoretical black hole final mass distribution function. {\\em Delayed} collapses are detectable only inside the Local Group of galaxies by interferometers of the first generation. Obviously, {\\em delayed} collapses are less detectable than the {\\em prompt} ones, due to the lower matching of their characteristic frequencies to the sensitivity curve of the detectors. On the other hand, the initial stage of the {\\em delayed} collapse, with the formation of a neutron star following a strongly asymmetric explosion (if $m_{min}=2~M_\\odot $), could be a promising source of gravitational waves. Detection perspectives of {\\em delayed} collapses are not so much better for adavnced interferometers because, if distances up to the Virgo Cluster are considered (where we expect $\\sim 1\\div 5~ev/yr$), the signal-to-noise ratio is much lower than one, unless a very high degree of asimmetry is produced. On the contrary, {\\em prompt} collapses could be detected with large enough $SNR$. {\\em Delayed} collapses to a black hole belong to a class of high frequency sources of gravitational waves, which comprises various processes involving compact objects, as for instance the excitation of neutron star w-modes \\cite{kokko}, the coalescence of two neutron stars with the formation of a light black hole \\cite{baum2}, dynamical instabilities in neutron stars \\cite{centr}, and some kinds of secular instability in neutron stars \\cite{lai}. Such sources cannot be efficiently detected by present resonant detectors and forthcoming interferometers because they are expected to emit at frequencies higher than those at which both interferometric and resonant detectors have their best sensitivity. In past years {\\em local arrays} of small resonant detectors, which are particularly suited for the detection of high frequency gravitational radiation, have been proposed\\cite{frasca1}$^,$\\cite{frasca2}. A detailed study of their detection performances, considering different geometries, dimensions and materials, has been done \\cite{palo}. The non-continuous background of gravitational waves produced by the ensemble of the star collapses to a black hole, which occurred at a higher rate in the early phases of the Universe, has recently been calculated \\cite{fersch}. It would be interesting to repeat the calculation in the case of {\\em delayed} collapse. In such a case we have the superposition of energy spectra nearly with the same shape and all peaked at nearly the same frequency, in the range $4.5\\div 9kHz$ (this holds also for star metallicity $Z=0$). Roughly speaking, as most of the collapses take place at redshift $z\\sim 2$, we expect to have, at the detector, a background spectrum strongly peaked somewhere in the band $1.5\\div 3~kHz$. \\nonumsection{Aknowledgements} I want to thank S. Frasca and M. A. Papa for the useful discussions and suggestions, T. Piran for his encouragement, V. Ferrari for her careful reading of the manuscript and the anonymous referees for their comments aiming to an improvement of this paper. \\nonumsection{References}" }, "0003/astro-ph0003117_arXiv.txt": { "abstract": "{% \\if@twocolumn We present the results of a series of axisymmetric time-dependent magnetohydrodynamic (MHD) simulations of the propagation of cooling, overdense jets. Our numerical models are motivated by the properties of outflows associated with young stellar objects. A variety of initial field strengths and configurations are explored for both steady and time-variable (pulsed) jets. For the parameters of protostellar jets adopted here, even apparently weak magnetic fields with strengths $B \\gapprox 60 \\mu$G in the pre-shocked jet beam can have a significant effect on the dynamics, for example by altering the density, width, and fragmentation of thin shells formed by cooling gas. Strong toroidal fields ($\\geq 100\\mu$G) with a radial profile that peaks near the surface of the jet result in the accumulation of dense shocked gas in a ``nose cone'' at the head of jet. We suggest that this structure is unstable in three-dimensions. A linear analysis predicts that axisymmetric pinch modes of the MHD Kelvin-Helmholtz instability should grow only slowly for the highly supermagnetosonic jets studied here; we find no evidence for them in our simulations. Some of our models appear unstable to current-driven pinch modes, however the resulting pressure and density variations induced in the jet beam are not large, making this mechanism an unlikely source of emission knots in the jet beam. In the case of pulsed jets, radial hoop stresses confine shocked jet material in the pulses to the axis, resulting in a higher density in the pulses in comparison to purely hydrodynamic models. In addition, if the magnetic field strength varies with radius, significant radial structure is produced in the pulses (the density is strongly axially peaked, for example) even if the density and velocity in the jet follow a constant ``top-hat\" profile initially. ", "introduction": "The most promising mechanism for the production of supersonic, highly collimated jets from low mass young stellar objects is by magnetic forces associated either directly with an accretion disk (K\\\"{o}nigl \\& Pudritz 1999), or with the interaction between an accretion disk and a magnetized central star (Shu et al.\\ 1999). Magnetic fields are also thought to contribute to the collimation of the jets on larger scales, (although only slowly so that the observed jet may only be a part of a much broader wind, e.g., Ostriker 1997). Thus, unless the outflowing material is highly resistive (which seems unlikely), protostellar jets should contain a dynamically important magnetic field which may affect both the propagation and stability of the outflow. Observation of the magnetic field strength associated with protostellar jets is difficult. However, by fitting one-dimensional radiative shock models to the observed line ratios in the bow shock of HH47, Morse et al.\\ (1993) inferred an upper limit to the magnetic field in the ambient medium upstream of the jet of $\\sim 30 \\mu$G, a value which they argued was too small to affect the dynamics except by increasing the cooling lengths behind radiative shocks. More recently, Ray et al.\\ (1997) have reported direct evidence for strong fields in the outflow associated with the source T~Tau~S through the detection of opposite circular polarization in the two spatially resolved outflow lobes. These authors infer very high field strengths within the lobes; several Gauss at a distance of few tens of AU from the central star. If protostellar jets contain strong magnetic fields as expected from theory, there may be a signature of such fields in their dynamics. While hydrodynamical studies of cooling, dense protostellar jets are widely available in the literature (e.g., see Raga 1995; Cabrit 1997 for reviews), MHD models are less well explored. MHD studies of extragalactic (i.e., {\\em underdense} and {\\em non-radiative}) jets have been reported by Clarke, Norman, \\& Burns (1986), and Lind et al.\\ (1989) for the case of toroidal fields, and by K\\\"{o}ssl et al.\\ (1990a; 1990b) for the case of toroidal and axial fields (see also Clarke 1996 for a recent review). For strong fields, characterized by a small value of $\\beta \\equiv 8 \\pi P/B^{2}$ where B is the field strength and P the thermal pressure (although note such fields may still have an energy density small compared to the kinetic energy density in the flow), it is found that the cocoon formed by lateral expansion of hot, shocked gas from the head of the jet is strongly inhibited. Instead, a magnetically confined ``nose cone\" of shock processed jet material is formed between the bow shock and Mach disk. In addition, the stability properties of the jet beam itself are strongly altered by the presence of a magnetic field. MHD studies of the propagation of {\\em overdense}, non-radiative jets into a uniform ambient medium which contains a helical magnetic field everywhere have been reported by Todo et al.\\ (1992). They describe results for a number of field strengths and approximate the effects of radiative cooling with an equation of state with $\\gamma=1.2$. The formation of a nose cone, as well as suppression of the cocoon, is evident for strong fields. In subsequent three-dimensional simulations (Todo et al.\\ 1993), it was found that a magnetic field strength of $70\\mu$G was sufficient to produce nonaxisymmetric kink instabilities in the jet beam which the authors suggest may be relevant to bends observed in the jets from some protostellar systems. This field strength is approaching the value inferred by Morse et al. (1993). More recently, several authors have reported results from MHD studies of dense jets in which optically thin radiative cooling is directly coupled to the dynamics. Frank et al.\\ (1998, hereafter FRJN-C) report axisymmetric studies which confirm the formation of nose cones, and (for the particular form of the initial field distribution adopted in the jet) the presence of pinch modes driven by magnetic tension. Cerqueira et al.\\ (1997, hereafter CGH; see also de Gouveia Dal Pino \\& Cerqueira 1996) have used an SPH code which has been extended with an algorithm designed to represent magnetic forces to study the propagation of dense, cooling jets in three-dimensions. They find that the fragmentation of the dense shell formed by cooling at the head of the jet is strongly affected by magnetic forces, and they use these results to conclude that the clumpy structure observed in jets provides evidence that near the head of the jet the field must be primarily axial rather than helical. In their latest simulations (Cerqueira \\& de Gouveia Dal Pino 1999, hereafter CG), these authors also report evidence for both modes of the MHD Kelvin-Helmholtz (KH) and a current-driven pinch instability in models of protostellar jets. Finally, Frank et al. (1999a; see also Lery \\& Frank 1999) have studied the propagation of MHD jets launched by Keplerian accretion disks using analytic models for magneto-centrifugal outflows to describe the jet structure at the nozzle. A variety of complex behavior is seen in this case, some of which results from MHD effects and some which results from the non-uniform radial structure of the jet beam. Given the apparent sensitivity of jet dynamics to the assumed initial orientation and strength of the magnetic field within the jet beam suggested by published numerical results, it is clearly useful to survey different initial field profiles. In this paper, we present the results from a large number of two-dimensional, axisymmetric MHD simulations of dense, cooling jets. To emphasize the effects produced solely by the magnetic field, we study jets with a uniform (``top-hat\") radial profile propagating into an initially uniform ambient medium. Since the theory of the production of such jets suggests a primarily toroidal field within the outflow far from the source, we focus on toroidal fields with different strengths and radial profiles. In most of our models, the ambient medium into which the jet propagates is unmagnetized, although in some cases we add a poloidal field which threads both the jet and ambient gas initially. Given the importance of nonaxisymmetric instabilities to the dynamics of magnetized jets (Todo et al.\\ 1993), fully three-dimensional MHD simulations of cooling jets are important (CGH; CG), however in order to cover parameter space efficiently we confine the models described in this study to axisymmetry. Moreover, for the highly supermagnetosonic jets studied here, non-axisymmetric modes of, e.g., the K-H instability grow slowly and will not affect the jet in regions close to the nozzle. Of course, precession of the jet beam, a clumpy ambient medium, or a much lower magnetosonic Mach number could introduce three-dimensional effects much earlier. Our results extend previous studies in that we evolve the jet for much longer (using a grid roughly three times larger than used by FRJN-C or CG), and we focus our attention on strongly magnetized jet beams propagating into an unmagnetized ambient medium. Where there is overlap, we compare and contrast our results to previous studies throughout this paper. The paper is organized as follows. We describe our numerical techniques in \\S2. As an aid to the interpretation of our results, we review the relevant stability properties of MHD jets in \\S3. Our results for steady are presented in \\S4, and for pulsed jets in \\S5. In \\S6 we summarize and conclude. ", "conclusions": "We have studied the propagation of both steady and time-variable (``pulsed\") protostellar jets using numerical MHD simulations. Although we focus our attention on models in which the jet beam contains a purely toroidal magnetic field peaked near its surface, we have also studied the effect of varying both the strength and geometry of the field on the dynamics. We find that even a weak magnetic field ($B \\leq 60 \\mu$G) in the jet beam can lead to important effects on the structure and dynamics of steady jets (FRJN-C; CGH; CG). In particular, such fields can alter the density structure and fragmentation of dense shells formed in cooling jets. However, the details of the effects depend sensitively on the geometry of the field. For example, while magnetic pressure can limit compression in such shells, resulting in {\\em lower} densities and less fragmented shells, radial hoop stresses associated with purely toroidal fields can confine shocked gas towards the axis, resulting in {\\em higher} densities there. In some cases, hoop stresses lead to the formation of nose cones ahead of the Mach disk. We see no evidence for significant structure induced by the Kelvin-Helmholtz instability (such modes were seen for the parameters and field geometry adopted by CG). We do see evidence for current driven pinch modes but the induced pressure pulses in the jet interior do not appear strong enough to explain the emission knots that are observed in protostellar jets (see also CG). Thus, models invoking jet velocity fluctuation appear to remain the most viable explanation for the knots observed in protostellar outflows. We do not see any evidence for disruption of the jet by pinch modes. In the case of pulsed cooling jets, we find the primary effect of a toroidal magnetic field is to confine shocked jet material to the axis, preventing it from begin ejected into the cocoon, and leading to higher postshock densities in the pulses in comparison to purely hydrodynamic models. This result is in contrast to the expectation of planar radiative shock models, in which the addition of a magnetic field leads to {\\em lower} postshock densities. Our results indicate it is important to account for multidimensional effects in the study of magnetized cooling jets. Toroidal confinement also leads to radial variation of quantities in the pulses, even if the density and velocity initially are constant with radius. The radial variation of, e.g. the density, in the pulses is large enough (a factor of 80 in a strongly magnetized jet) that it will likely affect the computation of the resulting emission properties. While there are many uncertainties in the magnetic field strength and topology implied by the observations, our present results make it clear that MHD models of protostellar jets need to be seriously investigated. This is particularly true since we find that average values of the magnetic field which correspond to a plasma $\\beta$ much larger than one (and which therefore one would infer to be too weak to be important) can still make a significant difference in the dynamics and physical conditions associated with the jet, and therefore its emission properties as well. It is thus important to understand the asymptotic structure of the field produced by the mechanism which drives the outflow. \\vspace{0.5cm} We thank E. de Gouveia Dal Pino for comments on an earlier version of the manuscript, and an anonymous referee for suggested improvements. JS acknowledges support from the DOE through grant DFG0398DP00215. PH acknowledges support from the National Science Foundation through grant AST-9318397 and AST-9802955 to the University of Alabama. \\newpage" }, "0003/astro-ph0003392_arXiv.txt": { "abstract": "The MACHO project is a search for dark matter in the form of massive compact halo objects (MACHOs). The project has photometrically monitored tens of millions of stars in the Large Magellanic Cloud (LMC), Small Magellanic Cloud (SMC), and Galactic bulge in search of rare gravitational microlensing events caused by these otherwise invisible objects. In 5.7 years of observations toward the LMC some 13--17 microlensing events have been observed by the MACHO survey, allowing powerful statements to be made about the nature of the dark population in the halo of our Galaxy. A critical component of these statements is an accurate determination of the survey's detection efficiency. The detection efficiency is a complicated function of temporal sampling, stellar crowding (the luminosity function), image quality, photometry, time-series analysis, and criteria used to select the microlensing candidates. Such a complex interdependence is most naturally solved using a Monte Carlo approach. Here we describe the details of the Monte Carlo used to calculate the efficiency presented in the MACHO 5.7-year LMC results. A similar calculation was performed for MACHO's 1-year and 2-year results. Here we correct several shortcomings of these past determinations, including (1) adding fainter source stars (2.5 magnitudes below our faintest detected ``stars\"), (2) an up-to-date luminosity function for the LMC, (3) better sampling of real images in both stellar density and observing conditions, (4) an improved scheme for adding artificial microlensing onto a random sample of real lightcurves, and many other improvements. The Monte Carlo technique presented here realistically simulates the negative effects of severe crowding (blending) that is a fact of microlensing surveys. ", "introduction": "\\label{sec-intro} Gravitational microlensing has become an important tool for the discovery and characterization of dark populations. There are now as many microlensing groups (\\cite{erossmc2,lmc2,AGAPE,MEGA99,udalski-erosII}) as there are potential lines of sight out of the Galactic halo (LMC, SMC, Galactic bulge, M31). Surveys toward the LMC have observed an excess in the number of microlensing events over what is expected from known populations of stars. We have recently reported 13--17 microlensing events (\\cite{lmc5}; hereafter \\yrfive) in 5.7 years of observations toward the LMC and recently EROSII has reported two new events that they interpret as limiting the amount of halo dark matter (\\cite{eros-2}), but are consistent with the results of \\yrfive. In addition, a number of candidates have been observed towards the SMC (\\cite{smc1,smc2,planetsmc2,erossmc1,erossmc2,smc-binary}). One natural explanation for this excess rate is a population of massive compact halo objects of mass $\\sim\\!0.6 M_{\\odot}$ that contribute $\\sim\\!20$\\% to the mass of our Galaxy's halo. However, previously unknown (or underestimated) populations of stellar lenses (\\eg, in an LMC halo) are a distinct possibility. In order to make quantitative statements about such a population an accurate determination of the survey's microlensing detection efficiency is required. Here we describe the MACHO project's pipeline for calculating its detection efficiency. More details on the experiment, microlensing terminology, analysis, and interpretation can be found in the companion paper \\yrfive, and detailed reviews of microlensing in general are given by Roulet \\& Mollerach 1997 and Paczy{\\'n}ski 1996. The detection probability for individual events depends on many factors, \\eg, the 3 event parameters $\\Amax$, $\\that$, $\\tmax$ (maximum magnification, Einstein-diameter crossing time and time of peak magnification, respectively), and the unlensed stellar magnitude, as well as the observing strategy and weather conditions. Such a complicated dependence is most naturally solved using a Monte Carlo technique. We may simplify the dependence by averaging over the known distributions in $\\Amax$, $\\tmax$, the stellar magnitudes, and the known time-sampling and weather conditions, to derive our efficiency as a function only of event timescale, $\\Ethat$. Given an efficiency $\\Ethat$ we may compute the observed optical depth as, \\begin{equation} \\tau_{meas} = {{\\pi}\\over{4}}{{1}\\over{E}}\\sum_i{{\\that_i}\\over{{\\cal E}(\\that_i)}}, \\label{equation2} \\end{equation} \\noindent where $E$ is the survey's exposure in object-years, $\\that_i$ is the duration of event $i$ and $\\Ethat$ is the efficiency for detecting microlensing with duration $\\that$ (see also Alcock \\etal 1996 and 1997a; hereafter \\yrone\\ and \\yrtwo, respectively). The optical depth is a function of the amount and distribution of mass along the line of sight and is independent of the lens masses or velocities. However, we may extract more information if we assume a model of the Galactic halo. With the use of a model, a likelihood analysis may be performed to determine the most likely lens mass $m$ and mass fraction $f$ of the halo model. A model yields the distribution of event durations $d\\Gamma/d\\that$ (\\cite{explore}), which, when combined with the efficiency and integrated over all possible durations, predicts the number of events one expects to observe from such a Galactic halo composed entirely of MACHOs. \\begin{equation} N_{exp} = E \\int_{0}^{\\infty} {\\cal E} (\\that)\\,{d\\Gamma \\over d\\that}\\,d\\that. \\label{equation3} \\end{equation} \\noindent An identical method may be outlined for stellar populations as in \\yrfive. A likelihood estimator may be constructed from the observed events and the model--predicted number of events to determine simultaneously the most likely lens mass $m$ and mass fraction $f$ of the halo. Note the vital role the efficiency plays in connecting the world of models to that of the survey. The efficiency is a strong function of the temporal sampling of the survey. Events with very short durations (or very long durations) are unlikely to be detected as they easily fall `in-between' observations (or for long durations, extend through the data window). A straightforward way of simulating this ``sampling efficiency\" while retaining realistic behavior of the data is to use a random sample of real lightcurves from the survey in which to inject artificial microlensing. By randomly generating a number of events and running the same time-series analysis and selection criteria used in the real analysis, this Monte Carlo can be used to determine the survey's detection efficiency (\\cite{lmc1,lmc2,pratt}). How does one add microlensing onto a lightcurve? A simple method is outlined in \\S~\\ref{samplinglc} for the case where one assumes that each lightcurve represents a single resolved star. In this case all of the flux represented by the lightcurve is magnified by the microlensing model $A(t)$ (Equation~\\ref{equation6} below). The rarity of microlensing, however, demands crowded stellar fields in order to detect even a handful of events, and this in turn results in the blending\\footnote{ Blending is when two or more star's point-spread functions overlap to such an extent that the photometry code can no longer identify individual stars, but rather recovers one photometric `blob' or object.} of stellar light. The issue now becomes what fraction of the flux in the lightcurve (or object) is lensed? Our solution was first described briefly in \\yrone\\ (see Pratt 1997 for a more thorough description) and is outlined briefly as follows. The method involves the use of a large set of artificial star tests on a sample of MACHO survey images that are seeded with a color-magnitude diagram of the LMC. Each artificial star was added to an entire series of observing conditions and over 15 peak magnifications. The MACHO survey's dedicated photometry code \\sodophot\\ was then run on the resulting images to determine how each star behaved over a range of seeing, sky, and magnification. The resulting photometry and photometric flags were stored in binary files, which we refer to as Photometric Response DataBases (PRDBs). In this ``photometric efficiency\" technique each artificial star (hence forth referred to as a Photometric Response Function or PRF) is used as a model for adding artificial microlensing onto lightcurves. In most cases only a fraction of the observed flux in the lightcurve is actually lensed, and the PRF supplies us this fraction as a function of seeing and sky. In this way both blending and systematic photometry effects are mimicked in the resulting artificial data. Blending introduces several problems that can affect microlensing surveys in serious ways and has been extensively investigated (\\cite{lmc1,wozniak,pratt,hanhubble,han}). In addition to diluting the true peak magnification $\\Amax$, blending also biases the measured durations $\\that$ to shorter values, since blended events spend less time above a given threshold. This biasing of $\\that$ is particularly important as the optical depth estimate is proportional to the average duration of the events. Another effect of blending (which has received less attention, but is equally important) is that the survey's exposure in star-years is typically quite a bit larger than is estimated by naively counting photometered objects. One can view this (for a fixed exposure in object-years) as a net increase in the efficiency which in part balances out a decrease in efficiency due to blending. All of these effects must be taken into account if one desires an accurate detection efficiency known at least to the level of, if not better than, the shot noise inherent in the low number of events seen toward the Magellanic Clouds. We make several corrections and improvements to the Monte Carlo presented in \\yrone\\ and \\yrtwo. Most notably these are: (1) we add source stars to $V=24.5$, more than 2.5 magnitudes below our faintest detected stars (and 2.5 magnitudes fainter than was used in \\yrone\\ and \\yrtwo), (2) we use an up-to-date luminosity function of the LMC, constructed with our ground-based photometry for bright stars ($V < 19$) and HST photometry for dim stars ($V > 19$), (3) we compute luminosity function normalizations separately for each MACHO field, (4) 10 fields of size $5' \\times 5'$ with widely differing stellar density were used to simulate photometry of artificial stars over an average of 69 different observing conditions (only 2 fields with $\\sim\\!20$ observing conditions were used in \\yrone\\ and \\yrtwo), (5) an improved scheme for adding artificial microlensing onto real lightcurves, and (6) we describe a robust method of statistically correcting for the $\\that$ bias that can be used to estimate the optical depth. In \\S~\\ref{sec-sodo} we briefly review the MACHO telescope, photometry system and time-series analysis in order to introduce some MACHO specific vocabulary used throughout the rest of the paper. \\S~\\ref{sec-sampling} outlines the sampling efficiency approach as a primer to the photometric efficiency discussion in \\S~\\ref{sec-photometric}. The results of the photometric efficiency analysis is reserved for \\S~\\ref{sec-results}, and we conclude in \\S~\\ref{sec-summary}. ", "conclusions": "\\label{sec-summary} The results of \\yrfive\\ for 5.7 years of photometric data toward the LMC rely critically on how well we understand the detection efficiency. In this paper we calculated these efficiencies, correcting the most important systematics effects with realistic models. To account for the wide range of issues due to blending we perform artificial star tests on a broad range of images. These artificial stars tests allowed us to empirically account for blending and to model our photometry code's systematics. Correcting for blending also required an accurate knowledge of the LMC's LF because microlensing magnifies flux and increases the survey's sensitivity to dim, unresolved stars. We found that our sensitivity ${\\cal E}(\\that)$ to dim, unresolved stars vanishes for magnitudes fainter than $V\\sim\\!24$ and durations less than $\\that\\sim\\!300$ days. Our sensitivity in previous results (\\yrone\\ and \\yrtwo) was somewhat underestimated for durations greater than $\\that\\sim\\!100$ days. We also found the $\\that$ bias, due to blending, is of the order of 20\\% and we presented a method for statistically correcting this bias in our optical depth estimates. The method is complimentary to using blended $\\that$ fits." }, "0003/hep-ph0003034_arXiv.txt": { "abstract": "Taken at face value, current experimental data indicate the existence of a new particle, the sterile neutrino, which must be a singlet under the Standard Model gauge group. Although they are not detectable through traditional means, such particles have interesting {\\it observable} consequences for particle astrophysics and cosmology. Here we examine these implications and discuss, in particular, sterile neutrino dark matter and the relationship between matter-enhanced active-sterile neutrino transformation and the synthesis of heavy elements in supernovae. ", "introduction": "In 1930, after inferring the existence of the neutrino from the continuous electron spectrum of nuclear $\\beta$-decay, W.~Pauli remarked \\cite{pauli}, ``I have done a terrible thing. I have postulated a particle that cannot be detected.'' Twenty-three years later, F.~Reines and C.~L.~Cowan \\cite{reinescowan} reported the first detection of neutrinos via inverse $\\beta$-decay. Encouraged by these events, the propensity of history to repeat itself, and a steady stream of positive experimental data, contemporary particle physicists and astrophysicists have recently explored the ramifications of so-called ``sterile'' neutrinos, denoted $\\nu_s$. The existence of such Standard Model-singlet fermions, which couple to the conventional (or ``active'') neutrinos $\\nu_e$, $\\nu_\\mu$, and $\\nu_\\tau$ solely through effective mass terms, is implied by the confluence of several neutrino oscillation experiments: \\begin{itemize} \\item {\\bf Atmospheric neutrinos} ~The Super-Kamiokande Collaboration has reported convincing evidence for the suppression of the flux of $\\nu_\\mu$ and $\\bar{\\nu}_\\mu$ produced by cosmic ray collisions with the Earth's upper atmosphere \\cite{superk}. (The measured flux of $\\nu_e$ and $\\bar{\\nu}_e$ is within expectation.) In particular, there is a statistically significant zenith angle dependence of the high energy muon-like events which is consistent with neutrino oscillations.\\footnote{Continued observations and future long base-line accelerator experiments will help rule out other explanations for the anomaly. See Ref. \\cite{pakvasa} for a catalog and discussion of these ``non-standard'' solutions.} A two-neutrino vacuum mixing fit yields $\\delta m^2 \\approx 10^{-3}-10^{-2}\\ {\\rm eV}^2$ and $\\sin^2(2\\theta) \\approx 1$, if the neutrino mixing maximally with $\\nu_\\mu$ is $\\nu_\\tau$ \\cite{superk}. There are also matter-enhanced (Mikheyev-Smirnov-Wolfenstein or MSW \\cite{msw}) solutions $\\delta m^2 \\approx \\pm 5\\times 10^{-3}\\ {\\rm eV}^2$ and $\\sin^2(2\\theta) \\approx 1$ if the mixing partner is $\\nu_s$ \\cite{yasuda}.\\footnote{More recent data favors the $\\nu_\\mu \\rightleftharpoons \\nu_\\tau$ channel over $\\nu_\\mu\\rightleftharpoons\\nu_s$ \\cite{superkrecent}.} \\item {\\bf Solar neutrinos} ~An array of solar neutrino experiments \\cite{bahcall} has observed an energy-dependent deficit of $\\nu_e$ emitted by nuclear reactions in the sun. The Kamiokande and Super-Kamiokande experiments observe about one-half of the expected flux of the highest energy solar neutrinos. The Homestake chlorine experiment, sensitive to intermediate and higher energy neutrinos, sees approximately one-third of the expected flux. Further, the Soviet/Russian-American Gallium Experiment (SAGE) and Gallex record roughly one-half of the expected flux integrated over nearly the entire solar spectrum. The combined result is a distorted spectrum which is extremely difficult to reconcile with the standard solar model. Global two-neutrino fits to these observations yield \\cite{bahcall} the large angle ($\\delta m^2 \\approx 10^{-5}\\ {\\rm eV}^2$, $\\sin^2(2\\theta) \\approx 1$) and small angle ($\\delta m^2 \\approx 10^{-5}\\ {\\rm eV}^2$, $\\sin^2(2\\theta) \\approx 5\\times 10^{-3}$) MSW solutions and the ``just-so'' vacuum oscillation solution ($\\delta m^2 \\sim 10^{-10}\\ {\\rm eV}^2$, $\\sin^2(2\\theta) \\approx 1$). The neutrino mixing with $\\nu_e$ may be $\\nu_\\mu$, $\\nu_\\tau$, or $\\nu_s$, depending on the solution.\\footnote{Of course, requiring compatibility with other experiments and astrophysical constraints (see below) restricts the allowed oscillation channels.} \\item {\\bf Accelerator neutrinos} ~The Los Alamos Liquid Scintillator Neutrino Detector (LSND) experiment has recorded an excess of $\\nu_e$ and $\\bar{\\nu}_e$ events in accelerator-produced beams of $\\nu_\\mu$ and $\\bar{\\nu}_\\mu$ respectively \\cite{athana}. There are several allowed regions in the oscillation parameter space, all of which fall in the ranges $0.2\\ {\\rm eV}^2 \\lsim \\delta m^2 \\lsim 8\\ {\\rm eV}^2$ and $10^{-3} \\lsim \\sin^2(2\\theta) \\lsim 10^{-1}$, assuming two-neutrino mixing. The agreement \\cite{athana} between the regions for the neutrino ($\\nu_\\mu\\rightleftharpoons\\nu_e$) and antineutrino ($\\bar{\\nu}_\\mu\\rightleftharpoons\\bar{\\nu}_e$) channels reinforces the oscillation interpretation. The Karlsruhe Rutherford Medium Energy Neutrino (KARMEN) experiment has searched for excess events in the same channels \\cite{karmen}, and despite a null result, a joint analysis of the KARMEN and LSND data preserves some of the LSND solution space \\cite{eitel}. \\end{itemize} The mutual incompatibility of these disparate sets of results is {\\it prima facie} evidence for the existence of a {\\it light} sterile neutrino $\\nu_s$, since the number of light weakly interacting neutrino species is known to be three (namely, $\\nu_e$, $\\nu_\\mu$, and $\\nu_\\tau$) \\cite{pdg}, an effectively three-neutrino mass matrix yields at most two independent $\\delta m^2$'s, and the active neutrinos are known to be very light compared to their charged leptonic counterparts.\\footnote{If the additional neutrino is not a Standard Model (SM) singlet, then it must have {\\it very weak} interactions with the SM particles in order to meet these constraints. See Ref. \\cite{kolbmohatep} for an astrophysical limit on such interactions.} Global fits of the data to three-neutrino mass matrices have borne out this conclusion, and several authors have indicated how to accommodate all of the data in a four-neutrino mixing matrix \\cite{bgg}. Assuming that the current data is explained by oscillations and that future neutrino experiments confirm the reality of an additional light neutrino species, a number of workers have begun constructing theoretical models which yield the naturally small Dirac {\\it and} Majorana neutrino masses required for appreciable active-sterile neutrino mixing. Some of these models are generalizations of the traditional see-saw mechanism \\cite{seesaw} for generating light active neutrinos in the presence of very heavy sterile neutrinos \\cite{mohapatra}. Others involve restricted or extended couplings in the context of the Standard Model (SM) \\cite{mclaughlinng}. Many methods rely on new symmetries to ensure that both active and sterile neutrino masses are small and comparable. All of them can be classified as simple extensions of the SM gauge group and matter content (including Grand Unified Theories or GUTs), supersymmetric models, or superstring-inspired scenarios. The phenomenological consequences and uses of sterile neutrinos are equally interesting. In particular, resonant transitions among active and sterile neutrinos can alter significantly the dynamics of early universe cosmology and various astrophysical venues. Indeed, significant consequences are almost guaranteed in phenomena such as Big Bang nucleosynthesis (BBN) and core-collapse supernovae, whose outcome is determined or dominated by neutrino physics. Transitions to and from sterile neutrinos can distort severely the active neutrinos' energy spectra, resulting, for example, in nucleosynthetic abundances markedly different from the commonly accepted values. These effects are not invariably unfavorable. Indeed, sterile neutrinos have been variously invoked to explain the origin of pulsar kicks \\cite{kusenko}, provide a new dark matter candidate \\cite{shifuller}, account for the diffuse ionization in the Milky Way galaxy \\cite{diffuse}, resolve the ``crisis'' in BBN \\cite{kev}, and help enable the synthesis of heavy elements in Type II supernovae \\cite{gail,mitesh}. In these proceedings, we describe the interesting phenomenological implications of two of these scenarios. We recapitulate in Sec. II the physics of sterile neutrino dark matter, concentrating on the production of the cold, non-thermal variety. In Sec. III, we summarize our recent work on and the status of matter-enhanced active-sterile neutrino transformation solutions to heavy-element nucleosynthesis. We give conclusions in Sec. IV. ", "conclusions": "The present solar, atmospheric, and accelerator neutrino experiments suggest the existence of a light sterile neutrino. While this is a radical departure from the folklore that sterile neutrinos (if they exist) are very heavy, future experiments \\cite{bgg} may confirm their reality and force theorists to modify or discard cherished models of neutrino mass. As we have indicated in this paper, the mixing of sterile and active neutrinos has potentially far-reaching consequences for cosmology and astrophysics. They may account for much of the dark matter of the universe. They may even be the reason why we have gold rings, tin cans, atomic bombs, and lead shielding! MP is supported in part by a NASA GSRP fellowship. This work was partially supported by NSF grant PHY98-00980." }, "0003/hep-ph0003172_arXiv.txt": { "abstract": "We study the parametric amplification of super-Hubble-scale scalar metric fluctuations at the end of inflation in some specific two-field models of inflation, a class of which is motivated by hybrid inflation. We demonstrate that there can indeed be a large growth of fluctuations due to parametric resonance and that this effect is not taken into account by the conventional theory of isocurvature perturbations. Scalar field interactions play a crucial role in this analysis. We discuss the conditions under which there can be nontrivial parametric resonance effects on large scales. ", "introduction": "It was recently suggested \\cite{BKM1} that parametric resonance during the reheating phase of an inflationary Universe \\cite{TB90,KLS,STB95} may lead to an exponential amplification of super-Hubble scale gravitational fluctuations. If true, this would affect the usual predictions of inflationary models for observables such as the matter power spectrum and the spectrum of cosmic microwave anisotropies. In particular, it would require the coupling constants in the particle physics model of inflation to be exponentially smaller than previously thought in order that the theory does not generate a too large amplitude for the fluctuations. In Ref. \\cite{FB1} it was shown that, although there are no causality constraints which prohibit the amplification of super-Hubble (but sub-horizon) modes during reheating, the effect does not occur in a simple massive scalar field model of chaotic inflation based on the potential $V(\\phi) = m^2 \\phi^2 / 2$ (Here, $\\phi$ is the inflaton field). This is true even beyond the linear analysis \\cite{PE1}. Similarly, there is no effect for a quartic potential \\cite{EP2} $V(\\phi) = \\lambda \\phi^4 / 4$, nor for a potential containing both quadratic and quartic terms \\cite{LMZ99}. These results agree with the earlier analyses in Refs. \\cite{NT97} and \\cite{KH96}. It was then suggested \\cite{BTKM2} that the amplification of super-Hubble-scale modes would occur for two field models of inflation, e.g. for a model with potential \\begin{equation} \\label{KLSmodel} V(\\phi, \\chi) \\, = \\, {1 \\over 2} m^2 \\phi^2 + {1 \\over 2} g^2 \\phi^2 \\chi^2 \\, , \\end{equation} where as before $\\phi$ is the inflaton field and $\\chi$ is a second scalar matter field. This model had earlier been analyzed by Taruya and Nambu \\cite{TN98} who claimed that the isocurvature mode of the fluctuations will be parametrically amplified during reheating. However, as was shown in Refs. \\cite{JS} and \\cite{Ivanov} (and more recently in \\cite{LLMW}), the fluctuations in the $\\chi$ field are exponentially suppressed during inflation for values of the coupling constant for which the equation of motion of the metric fluctuations corresponds to broad resonance, thus rendering the effect studied in Ref. \\cite{BTKM2} completely inefficient. The suppression of fluctuations in the $\\chi$ field which renders the parametric amplification of gravitational fluctuations ineffective in the model given by (\\ref{KLSmodel}) occurs since during inflation the induced mass $m_{\\chi}$ of the $\\chi$ field which is given by $m_{\\chi} = g |\\phi|$ is larger than the Hubble expansion parameter $H$, and hence, as can be easily seen by considering the equation of motion \\begin{equation} \\label{chieom} {\\ddot \\delta \\chi} + 3 H {\\dot \\delta \\chi} + \\bigl( {{k^2} \\over {a^2}} + g^2 \\phi^2 \\bigr) \\delta \\chi \\, = \\, 0 \\end{equation} for the linearized fluctuation of the $\\chi$ field with comoving wave number $k$ (the scale factor is denoted by $a(t)$), $\\delta \\chi$ undergoes damped oscillatory motion. A model in which $m_{\\chi} < H$ during the stage of inflation when scales of cosmological interest today exit the Hubble radius was recently studied by Bassett and Viniegra \\cite{BV99}. It is a two field model given by the potential \\begin{equation} \\label{GKLSmodel} V(\\phi, \\chi) \\, = \\, {1 \\over 4} \\lambda \\phi^4 + {1 \\over 2} g^2 \\phi^2 \\chi^2 \\, . \\end{equation} In the absence of metric fluctuations, this model was studied in detail in Ref. \\cite{GKLS} (see also \\cite{Kaiser}), where it was shown that for values of the coupling constants satisfying \\begin{equation} g^2 \\, \\simeq \\, 2 \\lambda \\end{equation} long wavelength modes ($k \\simeq 0$) are in the first broad instability band of the Floquet-type equation of motion derived from (\\ref{chieom}) after field rescaling which describes the parametric resonance of matter fluctuations in an unperturbed expanding space-time. Bassett and Viniegra \\cite{BV99} showed that in this model the quantity $\\zeta$ \\cite{BST83} increases exponentially during the initial stages of reheating. Note that $\\zeta$ is a measure of the curvature fluctuations and is believed to be conserved on super-Hubble scales in the absence of isocurvature fluctuations (see e.g. \\cite{MFB92} for a review of the theory of cosmological fluctuations). However, since the model given by (\\ref{GKLSmodel}) admits isocurvature fluctuations, a growth of $\\zeta$ on super-Hubble modes is expected also in the ``usual\" analysis of the evolution of fluctuations in inflationary cosmology. In this paper we take a closer look at the theory given by the potential (\\ref{GKLSmodel}). Subject to certain assumptions on initial conditions of the background field dynamics we recover results similar to Bassett and Viniegra: exponential growth of $\\zeta$ during the initial stages of reheating. Furthermore, we demonstrate that this effect is not taken into account by the conventional theory of isocurvature perturbations. We then discuss some criteria which an inflationary Universe model must satisfy in order to have substantial parametric growth of $\\zeta$ during reheating via scalar field interactions. We argue that these conditions are naturally satisfied in some models of hybrid inflation, and we study a couple of concrete examples in which super-Hubble-scale gravitational fluctuations grow exponentially during reheating (some other examples where exponential growth of super-Hubble-scale modes could occur are given in \\cite{BGMK}). \\vskip 0.4cm ", "conclusions": "We have studied the parametric amplification of long wavelength gravitational fluctuations during reheating in two field inflationary Universe models. We have partially confirmed the results of Bassett and Viniegra \\cite{BV99} and shown that this effect is possible for certain models. We have established criteria under which an exponential increase in the amplitude of cosmological perturbations during the period when the inflaton field oscillates should be expected. It is crucial that there must be either broad-band parametric instability or negative coupling instability in the matter sector of the theory (i.e., in the absence of gravitational perturbations). This will excite isocurvature fluctuations during reheating. It is important that there be no net exponential damping in the amplitude of the isocurvature fluctuations before reheating. The resonance in the matter sector then induces a resonance in the gravitational sector provided that the background values of the matter fields do not vanish. Since large coupling constants are not necessary in order to have efficient resonance, the effect is stable against perturbative coupling constant renormalizations. We have shown that in this case the resulting increase in the amplitude of the adiabatic mode, conveniently tracked in terms of the variable $\\zeta$, and of isocurvature fluctuations, tracked in terms of the non adiabatic pressure $p \\Gamma$, is exponential and is due to the oscillations in the inflaton field. This means that the effect is absent if the phase transition is modelled by a monotonic change in $w = p / \\rho$. We then argue that the conditions under which parametric amplification of long wavelength gravitational fluctuations occurs are naturally satisfied in a class of models of hybrid inflation. The presence of a complex matter scalar field enhances the resonance, since it ensures the existence of a field which is massless in the true vacuum of the theory, but it is not crucial if there is negative coupling instability. However, note that the existence of massless modes is helpful for the effect to occur. Such massless modes arise quite generically in string theory (see e.g. \\cite{Dine} for a recent review). Thus, the parametric amplification of long wavelength fluctuations may be also present in models of inflation based on string theory. \\centerline{\\bf Acknowledgements} We are grateful to Bruce Bassett, Serguei Khlebnikov, Lev Kofman and Bill Unruh for stimulating discussions, and Jim Zibin for comments on the draft. R.B. wishes to thank Bill Unruh for hospitality at the University of British Columbia where this work was completed. F. F. wishes to thank Brown University for hospitality. The research was supported in part (at Purdue) by the U.S. Department of Energy under Contract DE-FG02-91ER40681, TASK B, and (at Brown) by DE-FG02-91ER40688, TASK A." }, "0003/hep-ph0003312_arXiv.txt": { "abstract": "We propose a new scenario of baryogenesis in the context of theories with large extra dimensions. The baryon number is almost conserved at zero temperature by means of a localization mechanism recently analyzed by Arkani-Hamed and Schmaltz: leptons and quarks are located at two slightly displaced positions in the extra space, and this naturally suppresses the interactions which ``convert'' the latter in the former. We show that this is expected to be no longer true when finite temperature effects are taken into account. The whole scenario is first presented in its generality, without referring to the bulk geometry or to the specific mechanism which may generate the baryon asymmetry. As an example, we then focus on a baryogenesis model reminiscent of GUT baryogenesis. The Sakharov out of equilibrium condition is satisfied by assuming nonthermal production of the bosons that induce baryon number violation. ", "introduction": "Despite the great success of Quantum Field Theory, a consistent scenario where gravity is also included still lacks. The most promising framework that could help in this task is string theory, whose consistency requires additional dimensions beyond the standard $3+1\\,$. This extra space is usually assumed to be compact, with a small compactification radius of order $M_p^{-\\,1}\\,$. However, it has been observed in ref. \\cite{add2} that, having no test of gravity below the millimeter scale, we do not really need such a tiny compactification radius, provided the extra dimensions are accessible only to gravitational interactions. The Standard Model degrees of freedom must indeed be localized on a $3$ dimensional wall whose inverse thickness does not exceed the scale of energy, of order $\\mbox{TeV}\\,$, we currently probe in accelerator experiments. The choice of such large compactification scale has the main goal of solving (or at least of weakening) the hierarchy problem. Denoting by $V_n$ the volume of the compact space -- assumed in \\cite{add2} to have a trivial metric -- and by $M$ the fundamental scale of gravity, the observed Planck mass is obtained by the relation $M_p^2 = M^{n+2} V_n \\,$. Under the condition $n \\geq 2\\;$, $M$ can be safely assumed to be very close to the electroweak scale, without conflicting with either cosmological, astrophysical, or laboratory bounds. Although considerably improving the standard situation, the above scenario retains however some degree of fine tuning, connected to the largeness of the quantity $V_n \\cdot M^n\\,$. A better result in this regard is provided by the more recent work \\cite{rs1}, where, due to the presence of cosmological constants in the bulk and on two walls, the metric is nonfactorizable with an exponential scaling in the extra space. This fact allows the achievement of both a (phenomenologically) acceptable theory with just one extra dimension (that could be even infinite in extension \\cite{rs2}) and a more satisfactory solution to the hierarchy problem. There are some other aspects common to all of these theories besides the ones discussed above. In particular, both proton stability and baryogenesis may be problematic in models with very low fundamental masses. For what concerns proton stability in Grand Unified Theories, the standard way to achieve it is to increase the mass of the additional bosons up to about $10^{15} - 10^{16}\\:$GeV $\\,$. In the framework of theories with extra--dimensions, an interesting mechanism has been suggested in ref. \\cite{ahs}. \\footnote{See also \\cite{aahdd,st,bd} for alternative suggestions.} In this paper, a dynamical mechanism for the localization of fermions on the wall \\cite{rs} is adopted: leptons and quarks are however localized at two slightly displaced positions in the extra space, and this naturally suppresses the interactions which ``convert'' the latter in the former. However, the observed baryon asymmetry requires baryon number ($B$) violating interactions to have been effective in the first stages of the evolution of the Universe. In this paper we thus wonder how this last requirement can be satisfied in a theory which adopts the idea of \\cite{ahs}, to ensure proton stability {\\it now} and baryon production {\\it in the past}. Our proposal is that thermal corrections, which are naturally relevant at early times, may modify the localization of quarks and leptons so to weaken the mechanism that suppresses the $B$ violating interactions. \\footnote{There exist other proposals for baryogenesis in these theories \\cite{bd,dg,p}: in the work \\cite{bd}, after considering several bounds on baryogenesis with large extra dimensions, a mechanism based on nonrenormalizable operators is proposed; in ref. \\cite{dg} baryon number is violated by ``evaporation'' of brane bubbles that carry a net baryonic charge into the bulk, and the matter--antimatter asymmetry can be due to a primordial collision of our brane with another one, that carried away the missing antimatter; in ref. \\cite{p} baryogenesis is obtained via leptogenesis, the latter being due to the existence of sterile neutrinos in the bulk.} The plan of the work is the following. In section $2$ we review the mechanism used in ref. \\cite{ahs} to localize chiral fermions on a domain wall and to suppress the rate of baryon number violating interactions. In our work, however, we also take into account the finite thickness of the wall: this enforces on the parameters of the model some bounds which are stronger than the ones reported in ref. \\cite{ahs}. In section $3$ we estimate the thermal corrections to the parameters of the theory, and in particular to the function that measures the suppression of the B violating interactions. Being our model nonrenormalizable, a perturbative treatment can be meaningful only at low energies. Anyhow, it is conceivable that the results we present in section $3$ can be a hint for the behavior of the theory at higher temperatures. The issue of baryogenesis is faced in section $4$, where we consider a very simple example reminiscent of GUT baryogenesis. In this mechanism, the baryon asymmetry is achieved through the dacay of a boson, whose interactions violate baryon number. In order for the model to work, the boson must be out of equilibrium before decaying, and this is not obvious in theories with low fundamental masses. In those theories, the Hubble parameter is indeed very low at energies below the fundamental scale of gravity, that sets the natural cutoff of the theory. We will thus consider a model where the bosons responsible for the baryon asymmetry are produced nonthermally (for instance, at the end of inflation) fulfilling thus naturally the out of equilibrium requirement. After considering other bounds, such as the stability of the kink under thermal corrections, we finally calculate the baryon asymmetry in a $B - L$ conserving scheme. In the conclusions we discuss our results and their possible future extensions. ", "conclusions": "The present work concerns the important issue of baryogenesis in theories with large extra--dimensions. Since the observed proton stability requires to a very high degree of accuracy baryon conservation at zero temperature, this task may be problematic within the above theories, which have very low fundamental scales. Our proposal relies on the localization mechanism for fermions discussed in ref. \\cite{ahs}. While in this work the present proton stability is due to a different localization (in the transverse direction) of leptons and quarks, we believe that thermal corrections may activate early baryon violating interactions. In our work we first provide a general discussion of the above scheme, without referring to any particular mechanism of baryogenesis. We find indeed that the first thermal corrections are in the direction of increasing the rate of baryon violations. We then consider a very specific example, where the matter--antimatter asymmetry is achieved through the decay of a (relatively) heavy boson in a $B - L$ conserving context. In this situation the Sakharov out of equilibrium condition can be obtained in the simplest way by considering nonthermal production of the bosons responsible for $B + L$ violation. Several bounds apply to the whole mechanism. The most general ones concern the localization procedure (we have found that the limits given in ref. \\cite{ahs} become more stringent once the thickness of the wall is considered) and its stability against thermal corrections. In addition, there are some other constraints which hold in the particular scheme of baryogenesis we adopted. The temperature of the heat bath right after the production of the baryon asymmetry cannot be too high, to avoid thermal regeneration of the bosons that induced baryogenesis. Moreover, this temperature has not to exceed the electroweak scale, in order not to activate the sphaleron transitions that would erase the $B + L$ asymmetry produced at some higher energy. Of course, this last bound can be easily overcome by considering some $B - L$ nonconserving process. We have found that the observed baryonic asymmetry can be accomplished quite naturally in our example, and we believe this should be the case in a more general context as well. Possible extensions of the present work are related to the generality of the scenario we discussed. Our idea indeed relies only on the localization mechanism adopted in \\cite{ahs} and not on the geometry of the bulk, nor on the details of the interactions responsible for the baryon asymmetry. Future works could thus proceed in two directions. Firstly, one could try to embed the scheme here described in a more complete cosmological setting. Secondly, some other baryogenesis mechanisms, for example {\\it \\`a la} Affleck--Dine (which does not require very high energy scales), may be explored." }, "0003/astro-ph0003411_arXiv.txt": { "abstract": "The first complete set of unified cross sections and rate coefficients are calculated for photoionization and recombination of He- and Li-like ions using the relativistic Breit-Pauli R-matrix method. We present total, unified (e~+~ion) recombination rate coefficients for $(e~+~C~VI \\longrightarrow C~V)$ and $(e~+~C~V \\longrightarrow C~IV)$ including fine structure. Level-specific recombination rate coefficients up to the n = 10 levels are also obtained for the first time; these differ considerably from the approximate rates currently available. Applications to recombination-cascade coefficients in X-ray spectral models of K$\\alpha$ emission from the important He-like ions are pointed out. The overall uncertainty in the total recombination rates should not exceed 10-20\\%. Ionization fractions for Carbon are recomputed in the coronal approximation using the new rates. The present (e~+~ion) recombination rate coefficients are compared with several sets of available data, including previous LS coupling results, and `experimentally derived' rate coefficients. The role of relativistic fine structure, resolution of resonances, radiation damping, and interference effects is discussed. Two general features of recombination rates are noted: (i) the non-resonant (radiative recombination) peak as $E,T \\rightarrow 0$, and the (ii) the high-T resonant (di-electronic recombination) peak. ", "introduction": "Electron-ion recombination with H- and He-like ions is of particular interest in X-ray astronomy (Proc. X-ray Symposium, 2000). X-ray emission in the K$\\alpha$ complex of He-like ions, such as C~V, from the $n = 2 \\rightarrow 1$ transitions yields perhaps the most useful spectral diagnostics for temperature, density, ionization balance, and abundances in the plasma source (Gabriel 1972, Mewe and Schrijver 1978, Pradhan and Shull 1981, Pradhan 1985). Li-like C~IV is of considerable importance in UV emission spectra from active galactic nuclei and quasars (e.g. Laor \\etal 1994), as well as absorption in AGN (Crenshaw and Kraemer 1999). In addition, the C~IV and other Li-like ionization states are valuable tracers of the plasma in the `hot interstellar medium' (Spitzer 1990, Spitzer and Fitzpatick 1993, Martin and Bowyer 1990, Bregman and Harrington 1986). The primary sets of atomic data needed for accurate calculations of ionization fractions are for photoionization and recombination. Theoretical models of spectral formation also require excitation cross sections and transition probabilities. A considerable amount of atomic data is being computed for these atomic processes under the Iron Project (IP; Hummer \\etal 1993) for electron impact excitation and radiative transition probabilities for astrophysically abundant elements using the Breit-Pauli R-matrix (BPRM) method that includes relativistic fine structure in intermediate coupling (Berrington \\etal 1995). The present work is an extension of the IP work to photoionization and recombination. The ionization balance equations usually correspond to photoionization equilibrium \\begin{equation} \\int_{\\nu_0}^{\\infty} \\frac{4 \\pi J_{\\nu}}{h\\nu} N(X^{z}) \\sigma_{PI}(\\nu,X^{z}) d\\nu = \\sum_j N_e N(X^{z+1}) \\alpha_R(X_j^{z};T), \\end{equation} and collisional equilibrium \\begin{equation} C_I(T,X^{z}) N_e N(X^{z}) = \\sum_j N_e N(X^{z+1}) \\alpha_R(X_j^{z};T), \\end{equation} \\noindent where the $\\sigma_{PI}$ are the photoionization cross sections, and the $\\alpha_R(X_j^{z};T)$ are the total electron-ion recombination rate coefficients of the recombined ion of charge $z$, $X_j^{z}$, to state j at electron temperature T. The $C_I$ are the rate coefficients for electron impact ionization that can be reliably obtained from experimental measurements (Bell \\etal 1983). On the other hand, the (e~+~ion) recombination cross sections and rates are difficult to compute or measure. However, several experimental measurements of electron-ion recombination cross sections using ion storage rings have been carried out in recent years (e.g. Wolf \\etal 1991, Kilgus \\etal 1990,1993; Mannervik \\etal 1997). The experimental cross sections exhibit detailed resonance structures observed at very high resolution in beam energy, and measure absolute cross sections. Therefore they provide ideal tests for theoretical methods, as well as the physical effects included in the calculations. Many of these experimental measurements have been for recombination with H- and He-like C and O. Among the recent theoretical developments is a self-consistent method for calculations for photoionization and (e~+~ion) recombination, as described in previous papers in this series. An identical eigenfunction expansion for the ion is employed in coupled channel calculations for both processes, thus ensuring consistently accurate cross sections and rates in an ab initio manner. The theoretical treatment of (e~+~ion) recombination subsumes both the non-resonant recombination (i.e. radiative recombination, RR), and the resonant recombination (i.e. di-electronic recombination, DR) processes in a unified scheme. In addition to the total, unified recombination rates, level-specific recombination rates and photoionization cross sections are obtained for a large number of atomic levels. The calculations are carried out in the close coupling approximation using the R-matrix method. Although the calculations are computationally intensive, they yield nearly all photoionization and recombination parameters needed for astrophysical photoionization models with higher precision than hitherto possible. Previous calculations of unified (e~+~ion) recombination cross sections and rates, reported in the present series on photoionization and recombination, were carried out in LS coupling (Nahar and Pradhan 1997, paper I; Nahar 1999). There were two reasons. First, the calculations are extremely complex and involve both radiative photoionization and collisional electron-ion scattering calculations; the full intermediate coupling relativistic calculations are many times more computationally intensive than the LS coupling ones. Second, the effect of relativistic fine structure was expected to be small for these light elements. For the highly charged H- and He-like ions, however, subsequent calculations showed that results including the relativistic effects are signficantly more accurate not only in terms of more detailed resonance structure, but also to enable a full resolution of resonances necessary to include radiative damping (Pradhan and Zhang 1997, Zhang \\etal 1999, and references therein). The relativistic Breit-Pauli R-matrix (BPRM) method is now extended to calculate the total and level-specific recombination rate coefficients in the self consistent unified manner. In this paper we describe the first of a series of full-scale BPRM calculations of photoionization and photo-recombination, as inverse processes, and DR, to obtain total, unified (e~+~ion) recombination rates of He- and Li-like Carbon, C IV and C V. ", "conclusions": "New relativistic calculations are presented for the total, unified (e~+~ion) rates coefficients for C~IV and C~V of interest in X-ray astronomy. As the photo-recombination cross sections in the dominant low-energy region have earlier been shown to be in very good agreeement with experiments (Zhang \\etal 1999), it is expected that the present rates should be definitive, with an uncertainty that should not exceed 10--20\\%. The unified theoretical formulation and experimental measurements both suggest that the unphysical and imprecise division of the recombination process into 'radiative recombination (RR)' and 'di-electronic recombination (DR)' be replaced by 'non-resonant' and 'resonant' recombination, since these are naturally inseparable. Further calculations are in progress for Oxygen (O~VI and O~VII) and Iron (Fe~XXIV and FeXXV). The available data includes: (A) Photoionization cross sections for bound fine structure levels of C~IV and C~V up to n = 10 -- both total and partial (into the ground level of the residual ion). !!!! (B) Total, unified recombination rates for C~IV and C~V, and level-specific recombination rate coefficients for levels up to n = 10. All photoionization and recombination data are available electronically from the first author at: nahar@astronomy.ohio-state.edu. The total recombination rate coefficients are also available from the Ohio State Atomic Astrophysics website at: www.astronomy.ohio-state.edu/$\\sim$pradhan." }, "0003/astro-ph0003282_arXiv.txt": { "abstract": "The hypothesis that AGN jets might be the sources of the ultra-high energy cosmic rays has originally motivated the venture of TeV gamma ray astronomy. Surprisingly, after the discovery of TeV emission from blazars the attention has shifted to more traditional explanations which do not involve energetic hadrons, and there is even common believe that a hadronic interpretation is disfavored by observations. It is shown here that this is not the case, and that the currently observed spectra and variability features of blazars can be perfectly understood within hadronic blazar models. I also discuss how hadronic models might be observationally distinguished from common leptonic models, and point out some interesting aspects which could be relevant for the understanding of the differences between blazar classes. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003231_arXiv.txt": { "abstract": "{We consider the evolution of domain walls produced during an anisotropic phase in the very early universe, showing that the resulting network can be very anisotropic. If the domain walls are produced during an inflationary era, the network will soon freeze out in comoving coordinates retaining the imprints of the anisotropic regime, even though inflation makes the universe isotropic. Only at late times, when the typical size of the major axis of the domain walls becomes smaller than the Hubble radius, does the network evolve rapidly towards isotropy. Hence, we may hope to see imprints of the anisotropic era if by today the typical size of the major axis of the domain walls is of the order of the Hubble radius, or if the walls re-entered it only very recently. Depending on the mass scale of the domain walls, there is also the possibility that they re-enter at earlier times, but their evolution remains friction-dominated until recently, in which case the signatures of the anisotropic era will be much better preserved. These effects are expected to occur in generic topological defect models.} ", "introduction": "\\label{intro} It is well know that the `Hot Big Bang' model \\cite{kolb}, despite its numerous successes, is plagued by a number of `initial conditions' problems, of which the horizon, flatness and unwanted relic ones are the best known. The standard way to solve them is to invoke an epoch of cosmological inflation \\cite{guth,linde,lyth}, a relatively brief period of exponential (or quasi-exponential) cosmological expansion. The way inflation solves these problems is, loosely speaking, by erasing all traces of earlier epochs and re-setting the universe to a rather simple state. Indeed, inflation is so efficient in this task that a number of people have wondered if one can ever hope to probe the physics of a pre-inflationary epoch. There are, however, a small number of possible pre-inflationary relics. For example, the recent work by Turok and collaborators \\cite{curv} shows that curvature can, in some sense, survive inflation. Another class of inflationary survivors are topological defects \\cite{vsh}, formed at phase transitions either before or during inflation \\cite{shafi,yoko,hodges}). It is known (see, e.g. \\cite{kolb,linde,acm2}) that one needs about 20 e-foldings of inflation\\footnote{The exact number is of course model-dependent.} to solve the monopole problem. One can reverse the argument and say that monopoles can survive about 20 e-foldings of inflation. The inflationary epoch itself will obviously push the monopoles outside the horizon, but the subsequent evolution of the universe tends to make them come back inside, so if the inflationary epoch is not too long they can still have important cosmological consequences. Cosmic strings are even more successful, being able to survive about 50 e-foldings. The reason for this difference is that their non-trivial dynamics \\cite{ms1,ms2,ms3,acm2} makes them come back inside the horizon faster than one might naively have expected. The above two numbers are typical, but there are specific models where defects can survive even longer. One example is that of open inflation scenarios \\cite{bgt}. In this case the universe undergoes two different inflationary epochs---roughly speaking, a period of `old inflation' followed by one of `new inflation'. As pointed out by Vilenkin \\cite{vil}, one can expect that defects will form between the two inflationary epochs. In this case, a collaboration \\cite{acm2} including the present authors has recently shown that not only will cosmic strings survive the entire second inflationary epoch, regardless of how long it lasts\\footnote{Note that in these models the duration of the second inflationary epoch is fixed by the present value of the density of the universe \\cite{bgt}.}, but they will in fact be back inside the horizon by the time of equal matter and radiation densities. In such models, monopoles can survive up to about 30 e-foldings. Now, given that defects seem to be so successful surviving inflation, and that one expects them to be frozen out while they are outside the horizon, one can think of a further interesting possibility. For the best-studied case of cosmic strings, it is well known that the scaling properties of the network depend on the background cosmology \\cite{ms1,ms2,ms3}. Moreover, in some cases (typically when their evolution is friction-dominated) they can retain a `memory' of the initial conditions, or the general properties of the cosmology in which they find themselves at early times, for quite a large number of orders of magnitude in time \\cite{ms1,ms3}. It is therefore conceivable that if such an imprint of an early cosmological epoch is retained by a defect network which manages to survive inflation, we might still be able to observe it today. We believe that this is a general feature of defect models, and a number of non-trivial pieces of information about the very early universe can probably be preserved in this way. In the present paper we will restrict ourselves to a simple example. We will discuss the possibility of a domain wall network retaining information about an early anisotropic phase of the universe. There are very strong constraints \\cite{zeld} on the mass of domain walls formed after inflation, due to the fact that their density decays more slowly than the radiation and matter densities. However, these can be evaded by walls forming before or during inflation. In a subsequent paper, we will discuss the more interesting, but also more complicated, case of cosmic strings. The plan of the paper is as follows. In section \\ref{evequations} we briefly describe our background (Bianchi I) cosmology and the basic evolutionary properties of the domain walls. In particular, we focus on the approach to isotropy during inflation, which is discussed through both analytic arguments and numerical simulations. We emphasise that these simulations {\\it do not} include the defects. However, they serve an important purpose, as they are used in the subsequent discussion to show that the timescale needed for isotropization is compatible with the `survival' on anisotropic defect networks. We provide a description of our numerical simulations of domain wall evolution in section \\ref{nums}. These are analogous to those of Press, Ryden \\& Spergel \\cite{PRS}, and the interested reader is referred to this paper for a more detailed discussion of some relevant numerical issues. Here defect networks are evolved in an {\\it isotropic}, matter-dominated (ie, post-inflationary universe), and their main purpose is to show that isotropic and anisotropic networks will evolve in different ways, so two such networks can in principle be observationally distinguished as they re-enter the horizon. Our main results are presented and discussed in section \\ref{redi}, and finally we present our conclusions and discuss future work in section \\ref{conc}. Throughout this paper we will use fundamental units in which $c=1$. ", "conclusions": "\\label{conc} In this paper we have discussed a simple example of what we believe to be a rather generic feature of topological defect models, namely that they can easily retain information about the properties of the very early universe. This information is encoded in the scaling (ie, `macroscopic') and statistical (ie, `microscopic') properties of the defect networks. This is even more relevant given the fact that defects can survive significant amounts of inflation. Hence, they can provide a unique probe of the pre-inflationary universe. The two crucial scales in the problem are the defect mass scale and the epoch when the defects come back inside the horizon. Specifically, we have discussed the role of domain walls. We have highlighted the existence of two scaling regimes for the domain wall network, in agreement with previous work \\cite{PRS}. Furthermore, we have shown that an anisotropic network re-entering the horizon will take longer to approach scaling than an isotropic one. Hence, if the very early universe had an anisotropic phase which was erased by an inflationary epoch, and if domain walls are present, then the walls can retain an imprint of the earlier phase, and this can have important observational consequences, eg for structure formation scenarios. As is well known, there are quite strong constraints \\cite{zeld,vsh} on the mass of domain walls formed after inflation. These are basically due to the fact that their density will decay more slowly than the radiation and matter densities. However, essentially all of these can be evaded (or at least significantly relaxed) by walls forming before or during inflation (and also by walls evolving in a friction-dominated regime). Having said this, how could these anisotropies be detected? The most naive answer would be through their imprint on CMB, but this is only true if their energy density is not too low, and such models are constrained in a variety of other ways (not only from the cosmology side, but also from the high-energy physics side). The case of `light' walls is therefore more interesting: note that just like in the case of `light strings' \\cite{ms3}, these are expected to be friction-dominated throughout most of the cosmic history. Here the observational detection of the effects we have described becomes somewhat non-trivial. The best way of doing it should be through observations of numbers of objects as a function of redshift in different directions (assuming that one has a reliable understanding of other possible evolutionary effects). Two specific examples would be large-scale velocity flows \\cite{zeh} and gravitational lensing statistics of extragalactic surveys \\cite{quast}. Finally, there is also an important implication of our work if at least one of the minima of the scalar field potential has a non-zero energy density, which is an anisotropic non-zero vacuum density. In a subsequent, more detailed publication, we shall discuss this scenario in more detail, as well as the analogous one for cosmic strings. To conclude, we have shown that the importance of topological defects as a probe of cosmological physics goes well beyond structure formation. Even if defects turn out to be unimportant for structure formation they can still (if detected) provide us with extremely valuable information about the physical conditions of the very early universe." }, "0003/astro-ph0003377_arXiv.txt": { "abstract": "name}{} \\newcommand{\\grad}{\\mathop{\\rm grad}\\nolimits} \\def\\dfrac#1#2{{\\displaystyle#1\\over\\displaystyle#2}} \\def\\slantfrac#1#2{\\hbox{$\\,^#1\\!/_#2$}} \\begin{document} \\title{Driven-disk model for binaries with precessing donor star. Three-dimensional simulations} \\author{ Bisikalo D.V.$^1$, Boyarchuk A.A.$^1$,\\\\ Kuznetsov O.A.$^2$, Chechetkin V.M.$^2$\\\\[0.3cm] $^1$ {\\it Institute of Astronomy of the Russian Acad. of Sci., Moscow}\\\\ {\\sf bisikalo@inasan.rssi.ru; aboyar@inasan.rssi.ru}\\\\[0.3cm] $^2$ {\\it Keldysh Institute of Applied Mathematics, Moscow}\\\\ {\\sf kuznecov@spp.keldysh.ru; chech@int.keldysh.ru}\\\\[0.3cm] } \\date{} {\\bf Abstract}---We present the results of three-dimensional numerical simulations of mass transfer in semi-detached binary with a donor star whose rotation vector precesses around the orbital rotation axis of the binary in the observer's coordinate frame. The calculations support our previous model of flow without a `hot spot'. Characteristic features of the flow in this model, such as the formation of an circumbinary envelope, the absence of a `hot spot' at the edge of the accretion disk, and the formation of a shock wave along the edge of the stream, are also present in the solution for a binary with precessing donor star. The parameters of accretion disk and of the structure of the near-disk regions recur with the precessional period of the rotation axis of donor star. ", "introduction": "Observations of binary systems over the past ten years indicate that in a number of close binaries, the best known being Her~X-1 (HZ~Her) and SS433, long-period variations are detected on characteristic timescales substantially longer than the orbital period. To explain these variations, precession of the accretion disk in the binary is widely assume (see, e.g., [1--3] and references therein). Possible reasons for precession of an accretion disk have been analyzed in a number of studies. As early as in 1972 N.I.Shakura [4] noted that the accretion disk can precess if its plane does not coincide with the orbital plane of a binary. Among various mechanisms that can lead to formation of an accretion disk inclined to the orbital plane, two are usually considered to be most probable: influence of magnetic field of the accretor or violation of the symmetry of the donor-star outflow due to rotation of the star. The question of formation an inclined disk under the action of the accretor's magnetic field remains open: control of the disk orientation would require a strong magnetic field which, in turn, could inhibit the formation of the disk itself. An inclined accretion disk is more likely to form due to a change of the position of the stream flowing from the inner Lagrangian point $L_1$. In particular W.Roberts [1] and A.M.Cherepashchuk [5,6] proposed a scenario in which a minor asymmetry of the explosion of a supernova (resulting in the formation of a relativistic object in the binary) can decline the orbital plane of a binary relative to the rotation axis of the normal component of the system. In this case, after the supernova explosion and the formation of a relativistic object, the rotation axis of the normal component of the system might become oriented not perpendicular to the orbital plane of the binary. For systems where mass is transfer from the normal star to the relativistic object, the disturbance of the symmetry of the outflow of matter from the donor-star might result in the formation of an accretion disk not aligned with the orbital plane. Precession of such a disk could be caused either by induced precession of the disk itself under the action of the gravitational attraction of donor-star [7,8], or by the stream oscillations due to precession of the rotation vector of the donor-star (the `slaved disk' model, [1,4]), or by some other mechanism (see, e.g., [9]). The description of the gas dynamics of mass transfer binaries of this type calls for the use of three-dimensional models because the rotation vectors of the donor-star is not perpendicular to the orbital plane and is engaged in precession. This means that this problem cannot be reduced to a two-dimensional one. Further, there is an additional complexity in describing such systems connected with the periodic time dependence of the boundary conditions, that is with the absence of a steady-state flow of matter in the system, thus necessitating the consideration of the structure of flow over long periods of time (few times longer than maximum characteristic period of the system). Until very recently, these circumstances, together with insufficient computational resources, made numerical studies of flow structures in binary systems of this type difficult. a Few attempts to consider the formation of an accretion disk in such systems were made in substantially simplified formulation [10, 11]. Here we consider for the first time numerical, self-consistent solution for the three-dimensional flow structure of mass transfer in semidetached binary systems in which the donor's rotation axis precesses. The obtained results support the model of `driven accretion disk', based on the idea that the oscillations of the disk relative to the equatorial plane reflect variations of the stream of matter from the inner Lagrangian point $L_1$. ", "conclusions": "Our three-dimensional simulations of mass transfer in semidetached binaries with precession of rotation vector of donor star reveal the `driven' nature of forming accretion disk. For typical parameters of numerical viscosity adopted for the numerical model ($\\alpha \\sim 0.1\\div0.5$ in terms of $\\alpha$-disk), the change of the flow pattern, the parameters of accretion disk, and parameters of the near-disk regions reflect the variations of boundary conditions on the donor star. In turn, the periodicity of boundary conditions is determined by the precessional velocity. Analysis of the flow pattern indicates that the basic features of the solution are qualitatively similar to those for calculations previously obtained for the cases of synchronous, aligned asynchronous, and non-aligned asynchronous rotation of the donor-star [12], and, in turn, indicates to the universal character of the model without a `hot spot' proposed in [15--18]. The `driven' character of the solution implies that the emission properties of the accretion disk and intercomponent gaseous structures recur with the precessional period of the rotation axis of donor-star. In binary systems where observed long-period variations can be explained by the precession of the donor-star the periodicity of the solution obtained here can be used to interpret the observational data." }, "0003/astro-ph0003141_arXiv.txt": { "abstract": "After the positive detection by \\bsax of hard X-ray radiation up to $\\sim$80 keV in the Coma cluster spectrum, we present evidence for nonthermal emission from A2256 in excess of thermal emission at a 4.6$\\sigma$ confidence level. In addition to this power law component, a second nonthermal component already detected by \\asca could be present in the X-ray spectrum of the cluster, not surprisingly given the complex radio morphology of the cluster central region. The spectral index of the hard tail detected by the PDS onboard \\bsax is marginally consistent with that expected by the inverse Compton model. A value of $\\sim 0.05~\\mu G$ is derived for the intracluster magnetic field of the extended radio emission in the northern regions of the cluster, while a higher value of $\\sim 0.5~\\mu G$ could be present in the central radio halo, likely related to the hard tail detected by \\asca. ", "introduction": "Nonthermal hard X-ray (HXR) radiation has been detected for the first time in the Coma cluster by \\bsax (Fusco-Femiano \\etal 1999) and {\\it RXTE} (Rephaeli, Gruber \\& Blanco 1999), while marginal evidence is reported for A2199 (Kaastra \\etal 1999). These observations are only first steps towards assessing the general existence of this new component in the X-ray spectra of clusters of galaxies. The search for nonthermal emission in more clusters is of high importance as it will allow to derive additional informations on the physical conditions of the intracluster medium (ICM) environment, which cannot be obtained by studying the thermal plasma emission only. Various interpretations of the HXR emission have been presented since its discovery in the Coma cluster spectrum. The most direct explanation is inverse Compton (IC) scattering of cosmic microwave background (CMB) photons by the relativistic electrons responsible of the extended radio emission present in the central region of Coma (Willson 1970). The combined radio synchrotron and IC HXR fluxes (e.g., Rephaeli 1979) allow to estimate a volume-averaged intracluster magnetic field of $\\sim 0.16~\\mu G$ (Fusco-Femiano \\etal 1999). One of the problems with the IC model is that this value of the magnetic field in the ICM seems to be at odd with the value determined from Faraday rotation of polarized radiation toward the head tail radio galaxy NGC4869 that gives a line-of-sight $B\\sim 6~\\mu G$ (Feretti \\etal 1995), and with the equipartition value in the radio halo, which is $\\sim 0.4~h_{50}^{2/7}~\\mu G$ (Giovannini \\etal 1993). We note, however, that Feretti \\etal (1995) also inferred the existence of a weaker and larger scale magnetic field component in the range of $0.1-0.2~ h^{1/2}_{50}~\\mu G$, and therefore the $\\sim 6~\\mu G$ field could be local. A low average magnetic field is also consistent with the model developed by Brunetti \\etal (1999), which predicts a magnetic field strenght decreasing with the distance from the cluster centre. An alternative explanation is nonthermal bremsstrahlung (NTB) emission from suprathermal electrons currently accelerated at energies greater than $\\sim$10 keV by shocks or turbulence (Kaastra \\etal 1998; Ensslin, Lieu, \\& Biermann 1999; Sarazin \\& Kempner 1999). Another and more trivial possibility is that the HXR radiation is due to a hard X-ray source present in the external regions of the field of view of the \\bsax PDS (FWHM=$1.3^{\\circ}$, hexagonal), as for example a highly obscured Seyfert 2 galaxy like the Circinus galaxy (Matt \\etal 1999). In the central region ($\\sim 30^{\\prime}$ in radius), the MECS image does not show evidence of this kind of sources (Fusco-Femiano 1999). Hovewer, the detection of a hard nonthermal component in other clusters should strongly reduce the probability of this last interpretation. In this letter we present the results of a long observation of A2256, exploiting the unique capabilities of the PDS, onboard \\bsax, to search for HXR emission (Frontera \\etal 1997). The cluster was also observed with the MECS, an imaging instrument working in the 1.5-10 keV energy range (Boella \\etal 1997). The galaxy cluster A2256 is similar to the Coma cluster in many X-ray properties, as luminosity and presence of substructures. The \\rosat PSPC observations showed that A2256 is a double X-ray cluster (Briel \\etal 1991), suggesting that a subcluster may be merging with a larger cluster, although there is no strong evidence in the temperature map in favour of an advanced merger (Markevitch \\& Vikhlinin 1997), as it is for Coma. The average gas temperature is $\\sim$7 keV, as measured by several X-ray instruments (David \\etal 1993; Hatsukade 1989; Markevitch \\& Vikhlinin 1997; Henriksen 1999). Both clusters show a radio halo in the central and periferal regions. However, the radio emission from A2256 is notably complex. The region around the cluster centre is occupied by an unusual concentration of radio galaxies: at least five discrete sources have been identified with cluster galaxies, but there also two extended emission regions which have linear sizes $\\leq$1 Mpc (Bridle \\& Formalont 1976; Bridle \\etal 1979; Rottgering \\etal 1994). Throughout the Letter we assume a Hubble constant of $H_o = 50~km~s^{-1}~Mpc^{-1}~h_{50}$ and $q_0 = 1/2$, so that an angular distance of $1^{\\prime}$ corresponds to 92 kpc ($z_{A2256} = 0.0581$; Struble \\& Rood 1991). Quoted confidence intervals are at $90\\%$ level, if not otherwise specified. \\section {PDS and MECS Data Reduction} The total effective exposure time was $\\sim 1.3\\times 10^5$ sec for the MECS and $\\sim 7.1\\times 10^4$ sec for the PDS in the two observations of February 1998 and February 1999. The observed count rate for A2256 was 0.497$\\pm$0.002 cts/s for the 2 MECS units and 0.27$\\pm$0.04 cts/s for the PDS instrument. Since the source is rather faint in the PDS band (approximately 1.5 mCrab in 15-150 keV) a careful check of the background subtraction must be performed. The background sampling was performed using the default rocking law of the two PDS collimators that samples ON, +OFF, ON, -OFF fields for each collimator with a dwell time of 96\" (Frontera et al. 1997). When one collimator is pointing ON source, the other collimator is pointing toward one of the two OFF positions. We used the standard procedure to obtain PDS spectra (Dal Fiume et al. 1997), which consists in extracting one accumulated spectrum for each unit for each collimator position. We then checked the two independently accumulated background spectra in the two different +/-OFF sky directions, offset by 210' with respect to the on-axis pointing direction. The comparison between the two accumulated backgrounds ([+OFF] - [-OFF]) shows a difference with a marginal excess below ~30 keV in the [+OFF] pointing. This excess is much lower than the signal from the source, but it must not be neglected. The total excess in the first two equalized energy channels (15-33.5 keV) is $0.048\\pm 0.024~cts~s^{-1}$, i.e. approximately 2$\\sigma$. This concentration in only the lowest energy channels implies that the excess is likely due to contamination by a point source rather than to a statistical fluctuation. The total source spectrum was therefore obtained using only the uncontaminated background accumulated pointing at the [-OFF] field. Hovewer, in Section 3 we report the confidence level of the nonthermal emission in excess of the thermal one considering the average of the two background measurements. The background level of the PDS is the lowest obtained thus far with high-energy instruments on board satellites thanks to the equatorial orbit and is very stable again thanks to the favorable orbit. No modeling of the time variation of the background is required. MECS data preparation and linearization was performed using the {\\sc Saxdas} package under {\\sc Ftools} environment. We have extracted a MECS spectrum from a circular region of 8$^{\\prime}$ radius (corresponding to about 0.8 Mpc) centered on the primary emission peak. From the ROSAT PSPC radial profile, we estimate that about 70\\% of the total cluster emission falls within this radius. The background subtraction has been performed using spectra extracted from blank sky event files in the same region of the detector as the source. A numerical relative normalization factor among the two instruments has been included in the fitting procedure (see next Section) to account for: a) the fact that the MECS spectrum includes emission out to $\\sim$0.8 Mpc from the X-ray peak, while the PDS field of view (1.3 degrees FWHM) covers the entire emission from the cluster; b) the slight mismatch in the absolute flux calibration of the MECS and PDS response matrices employed (September 1997 release; Fiore, Guainazzi \\& Grandi 1999); c) the vignetting in the PDS instrument, (the MECS vignetting is included in the response matrix). The estimated normalization factor is $\\sim$1.1. In the fitting procedure we allow this factor to vary within 15\\% from the above value to account for the uncertainty in this parameter. \\section {PDS and MECS Data Analysis and Results} The spectral analysis of the MECS data alone, in the energy range 2-9.7 keV and in the central $\\sim$0.8 Mpc region, gives a gas temperature of $kT=7.41\\pm 0.23$ keV ($\\chi^2$=154.5 for 162 degrees of freedom; hereafter dof), using an optically thin thermal emission model (MEKAL code on the XSPEC package), absorbed by a galactic line of sight equivalent hydrogen column density, $N_H$, of 4.01$\\times 10^{20}~cm^{-2}$. This value of the temperature is consistent with the \\asca GIS measurement (6.78-7.44 keV; Henriksen 1999), and with the values obtained by previous observations : the \\einstein MPC (6.7-8.1 keV; David \\etal 1993) and \\ginga (7.32-7.70 keV; Hatsukade 1989). Also the flux of $\\sim 5.3\\times 10^{-11}$\\erg in the 2-10 keV energy range is consistent with the previous measurements. The iron abundance is $0.26\\pm 0.03$, in agreement with the \\asca results (Markevitch \\& Vikhlinin 1997). The analysis of the PDS data with a thermal bresstrahlung component gives a temperature of $\\sim$30 keV. Fitting the data with two thermal components, one of these at the fixed temperature of 7.4 keV, we obtain a temperature greater than $\\sim$90 keV for the second component. These unrealistic high values for the gas temperature obtained in both the fits are interpreted as a strong indication that the detected hard excess is due to a nonthermal mechanism. Figure 1 shows the simultaneous fit to the MECS and PDS data with a thermal component at the temperature of 7.47$\\pm$0.35 keV and a normalization factor of $\\sim$1.2 for the two data sets. The $\\chi^2$ is 180.5 for 167 dof. Hard X-ray radiation at energies greater than $\\sim$20 keV is in excess with respect to the thermal component at a level of $\\sim 4.6\\sigma$ and this value is rather stable against variation of the normalization factor. It results slightly lower ($\\sim 4.5\\sigma$) considering the average of the two background measurements. Besides, also fitting the PDS data alone with a thermal component at the fixed temperature of 7.47 keV we obtain an excess at a level of 4.3$\\sigma$. If we introduce a second nonthermal component, modeled as a power law, we obtain the fit shown in figure 2. The $\\chi^2$ is 156.6 for 165 dof. The improvement with respect to the previous model is significant at more than the 99.99\\% confidence level, according to the F-test. The confidence contours of the parameters $kT$ and photon spectral index ($\\alpha_X$) show that, at 90\\% confidence level, the temperature is well determined, 6.8-7.5 keV, while $\\alpha_X$ describes a large interval 0.3-1.7. The presence of the nonthermal component has the effect to slightly decrease the best fit value of the temperature ($6.95^{+0.45}_{-0.35}$ keV), with respect to the temperature obtained considering only the MECS data. The flux of the nonthermal component is rather stable, $\\sim 1.2\\times 10^{-11}$\\erg in the 20-80 keV energy range, against variations of $\\alpha_X$. The contribution of the nonthermal component to the thermal flux in the 2-10 keV energy range is $\\leq 10\\%$ for $\\alpha_X\\leq$1.70. The analysis of the two observations with effective exposure times of $\\sim$23 ksec (February 1998) and $\\sim$48 ksec (February 1999) for the PDS does not show significant flux variations. These results and the fact that the two clusters with a detected hard X--rays excess (Coma and A2256) both have radio halos, strongly support the diffuse nonthermal mechanism as responsible for the excess, as discussed in the next section. ", "conclusions": "A2256 is the second cluster, after Coma (Fusco-Femiano \\etal 1999), which shows hard X-ray radiation up to $\\sim$80 keV in the PDS spectrum, with a clear excess above the thermal intracluster emission. (A2199 shows only a marginal evidence in the external region of the MECS detectors, Kaastra \\etal 1999). We have investigated the possibility that the observed excess in A2256 could be due by a confusing source in the field of view of the PDS. The most qualified candidate is the QSO 4C +79.16 observed by \\rosat PSPC with a count rate of $\\sim$0.041 c/s (WGA Catalogue). With a typical photon index of 1.8 (\\rosat reports a steeper index of $\\sim$2.5), about 1.2 c/s are necessary to account for the observed HXR emission of $\\sim 1.2\\times 10^{-11}$\\erg in the 20-80 keV energy range of the PDS. Considering that the QSO is $\\sim 52^{\\prime}$ off-axis, an unusual variability of about two orders of magnitude is required. There is still the possibility that an obscured source, like Circinus (Matt \\etal 1999), be responsible of the detected HXR radiation. Unless the obscured source is within 2$^{\\prime}$ of the central bright core of A2256, our analysis of the MECS image excludes the presence of this kind of sources in the central region ($\\sim 30^{\\prime}$ in radius) of the cluster. The application of the inverse Compton model, based on the scattering of relativistic electrons with the 3K background photons, appears less straightforward in A2256 than in the Coma cluster. The radio morphology is remarkably complex (Bridle \\& Fomalont 1976; Bridle \\etal 1979; Rottgering \\etal 1994). There are at least four radio sources classified as head-tail radio galaxies, an ultra steep spectrum source and a diffuse region in the north with two diffuse arcs ($G$,$H$ according to Bridle \\etal 1979), at a distance of $\\sim 8^{\\prime}$ from the cluster centre. The extent of this diffuse region is estimated to be 1.0$\\times$0.3 Mpc, with a total flux density of 671 mJy at 610 MHz and a rather uniform spectral index of 0.8$\\pm$0.1 between 610 and 1415 MHz (Bridle \\etal 1979). The percentage polarization is uniform with an average value of 20\\%. The alignement of the electric field vectors suggests a well ordered magnetic field. The equipartition magnetic field is 1-2$~\\mu G$ (Bridle \\etal 1979). A fainter extended emission permeates the cluster centre (diffuse emission around $D$ in Bridle \\etal 1979) with a steeper radio spectral index of $\\sim$1.8 as estimated by Bridle \\etal (1979) and in agreement with the 327 MHz data from the Westerbork Northern Sky Survey (Rengelink \\etal 1997). The total flux density is 100 mJy at 610 MHz and no polarized emission have been detected from this region. We note that the physical and morphological properties of the diffuse $D$ emission are consistent with those of central halo sources while those in the $G-H$ region are consistent with the properties of peripheral relic sources as 1253+275 in the Coma cluster. In addition to the thermal emission, a second component in the X-ray spectrum of A2256 was noted by Markevitch \\& Vikhlinin (1997) in their spectral analysis of the \\asca data in the central r=$3^{\\prime}$ spherical bin. Although they were not able to firmly establish the origin of this emission, their best fit is a power law model with photon index 2.4$\\pm$0.3, therefore favouring a nonthermal component. The contribution of this component to the total flux is not reported in the paper. Considering that there are no bright point sources in the \\rosat HRI image, they argued for an extended source. Also the joint \\asca GIS \\& \\rxte PCA data analysis is consistent with the detection of a nonthermal component in addition to the thermal component. The contribution of this nonthermal component to the total X-ray flux in the 2-10 energy range is $\\leq 4\\%$. However, a second thermal component (0.75-1.46 keV), instead of a nonthermal one, provides a better description of the data (Henriksen 1999). The MECS data do not show evidence of this steep nonthermal component in the central bin of $2^{\\prime}$ because the energy range is truncated to a lower limit of 2 keV (Molendi, De Grandi, \\& Fusco-Femiano 2000), while a joint fit to the LECS \\& MECS data within $4^{\\prime}$ does not show a significant evidence for an additional component at energies lower than 2 keV. The power-law component (slope $2.4\\pm 0.3$) found in the analysis of the \\asca data (Markevitch \\& Vikhlinin 1997), and the upper limit of 1.7 for $a_{X}$, determined by \\bsax data, suggest that two tails could be present in the X-ray spectrum of A2256. The former might due to the central diffuse radio source with the steep index $\\alpha_R\\sim$1.8, and the last to the more extended radio emission in the northern region of the cluster with the flatter energy spectral index of 0.8$\\pm$0.1. Assuming that the contribution of the power-law component, detected by \\asca, to the total X-ray flux ($F_X(2-10 keV)\\sim 5\\times 10^{-11}$\\erg) is $\\sim 5\\%$, we obtain a negligible contribution at PDS energies ($\\sim 4\\times 10^{-13}$\\erg) and a magnetic field in the central radio region of $\\sim 0.5~\\mu G$. For the external radio region, with spectral index 0.8, the nonthermal X-ray flux $f_X(20-80 keV)\\simeq 1.2\\times 10^{-11}$\\erg, derived by the PDS excess, leads to a low value of $\\sim 0.05~\\mu G$. Even assuming that a large fraction (say 50$\\%$) of the HXR flux is due to the several point radio sources in the central region and/or to the contribution of different mechanisms, we obtain only a slightly greater value of $\\sim 0.08~\\mu G$. The combined fit of \\asca GIS and \\rxte PCA data (Henriksen 1999) gives an upper limit of $2.64\\times 10^{-12}$\\erg in the 2-10 keV energy range for the nonthermal component that corresponds to a lower limit for the volume-averaged intracluster magnetic field, $B$, of 0.36$~\\mu G$ ($\\alpha_R=1.8$). Considering that the HXR flux detected by the PDS is in agreement with the above value, we would obtain a value for $B$ consistent with that derived by the GIS \\& PCA data, but the fit to the MECS \\& PDS data is unacceptable for $\\alpha_X = 1+\\alpha_R = 2.8$. The previous scenario of a decreasing intracluster magnetic field from the cluster center would be difficult to reconcile with the stronger periferal radio region and higher equipartition magnetic field with respect to the central radio halo. Therefore, we could consider the possibility, recently suggested by Brunetti \\etal (1999), that the HXR IC spectrum may be flatter than the synchrotron radio spectrum because of the acceleration and energy loss processes that produce an electron spectrum with different slopes. A different electron spectrum index for HXR and radio emissions is more likely for low magnetic fields which require higher electron energies for synchrotron than for IC radiation. This could explain the better fit to the PDS data of A2256 with $\\alpha_X < 1+\\alpha_R$=1.8. Besides, this model suggests an alternative interpretation of the HXR excess of A2256. We can consider that a single hard tail is present in the X-ray spectrum of the cluster with index $\\alpha_X\\leq$1.7, as detected by the PDS. The electron spectrum responsible of this HXR IC emission can produce radio emission with spectral index $\\alpha_R > \\alpha_X$-1=0.7 with a resulting mean volume-averaged intracluster magnetic field higher than the one we derive from the classical IC model. A different mechanism which may produce HXR radiation is given by nonthermal bremsstrahlung. Sarazin \\& Kempner (1999) suggest that all or part of the HXR emission detected in the Coma cluster might be NTB from suprathermal electrons formed through current acceleration of the thermal gas, either by shocks or turbulence in the ICM. For A2256 the MECS \\& PDS measurements determine a power-law momentum spectrum of the electrons with index $\\leq 2\\alpha_X-1=2.4$ (90\\%). The consequence is that an accelerating electron model with flat spectrum produce more IC HXR emission than the NTB mechanism, unless the electron spectrum cuts-off or steepens at high energies. Besides, these models produce more radio emission than observed if $B$ is $\\geq 1~\\mu G$." }, "0003/astro-ph0003188_arXiv.txt": { "abstract": "We have extended a simple model of nonlinear \\DSA\\ (Berezhko \\& Ellison \\cite{BEapj99}; Ellison \\& Berezhko \\cite{EBicrc99a}) to include the injection and acceleration of electrons and the production of photons from \\brem, \\syn, \\IC, and \\pion\\ processes. We argue that the results of this model, which is simpler to use than more elaborate ones, offer a significant improvement over test-particle, power-law spectra which are often used in astrophysical applications of \\DSA. With an evolutionary supernova remnant (SNR) model to obtain shock parameters as functions of ambient interstellar medium parameters and time, we predict broad-band continuum photon emission from supernova remnants in general, and SN1006 in particular, showing that our results compare well with the more complete time-dependent and spherically symmetric nonlinear model of \\BKP\\ (\\cite{BereKP99a}). We discuss the implications nonlinear shock acceleration has for X-ray line emission, and use our model to describe how ambient conditions determine the TeV/radio flux ratio, an important parameter for $\\gamma$-ray observations of radio SNRs. ", "introduction": "\\label{sec:intro} More than twenty years of spacecraft observations in the heliosphere have proven that collisionless shocks can accelerate particles with high efficiency, \\iec 10-50\\% of the ram energy can go into superthermal particles (\\egc Eichler \\cite{Eich81}; Gosling \\etal\\ \\cite{GoslingEtal81}; Ellison \\etal\\ \\cite{EMP90}). A recent example comes from Terasawa \\etal\\ (\\cite{Terasawa99}) who report on an interplanetary shock, observed by GEOTAIL, where the pressures in thermal and accelerated particles, and the magnetic field are in approximate equipartition in the foreshock. They observed that the shock transition was smoothed by the energetic particle backpressure, giving unambiguous evidence that nonlinear acceleration occurred in an interplanetary shock, weak by astrophysical standards (for this exceptionally strong interplanetary shock, the sonic Mach number was $\\MSZ \\sim 5-10$). Energetic particles exist throughout the universe and shocks are commonly associated with them, confirming that shock acceleration is important beyond the heliosphere. In fact, shocks are believed to be the main source of Galactic cosmic rays, and energetic particles from shocks may produce spectacular events such as $\\gamma$-ray bursts and X-ray and gamma-ray flaring in blazars. The shocks in these objects are expected to be much stronger than those in the heliosphere and can only be more efficient and nonlinear. The conjecture that collisionless shocks are efficient accelerators is strengthened by results from plasma simulations which show efficient shock acceleration consistent with spacecraft observations (\\egc Scholer, Trattner, \\& Kucharek \\cite{STK92}; Giacalone \\etal\\ \\cite{GBSE97}). Other indirect evidence comes from the implied efficiency of galactic cosmic ray production, radio emission from supernova remnants (SNRs) (see Reynolds \\& Ellison \\cite{RE92}), equipartition arguments in AGNs and $\\gamma$-ray bursts, etc. (see Blandford \\& Eichler \\cite{BE87} for an early review). There is also clear evidence that shocks can produce strong self-generated turbulence. This has long been seen in heliospheric shocks (\\egc Lee \\cite{Lee82}, \\cite{Lee83}; Kennel \\etal\\ \\cite{KennelEtal84}; Baring \\etal\\ \\cite{BaringEtal97}) and there is evidence that it occurs at SNRs as well (\\iec Achterberg, Blandford, \\& Reynolds \\cite{ABR94}). Despite this compelling evidence for efficient, nonlinear shock acceleration, many astrophysical applications of shock acceleration continue to use the test-particle theory proposed more than 20 years ago by Axford, Leer, \\& Skadron (\\cite{ALS77}), Krymskii (\\cite{Krym77}), Bell (\\cite{Bell78}), and Blandford \\& Ostriker (\\cite{BO78}). We believe it is possible that test-particle models are used in situations where nonlinear shocks are clearly expected because the numerical, nonlinear results are complicated and unwieldy. None of the current techniques modeling nonlinear shocks have simple analytic approximations and this makes it difficult to perform parameter searches or compare nonlinear results to observations. Here, in an attempt to remedy this situation, we present a simple model of \\DSA, based on more complete studies, which includes the essential physics of nonlinear acceleration when the backpressure from energetic ions modifies the shock structure and induces spectral curvature. This model is computationally fast and easy-to-use, but complete enough to be a valuable tool for interpreting observations. We also investigate some implications of efficient acceleration and the associated nonlinear effects on the modeling and interpretation of particle and photon observations of SNRs, and include a detailed study of the broad-band continuum emission from the forward shock in SN1006. Our basic model is given in \\BEname\\ (\\cite{BEapj99}) where only protons are considered. Here, we extend the model to include electrons and alpha particles, and calculate the broad-band, continuum photon emission from the ion and electron distributions (Ellison \\& Berezhko \\cite{EBicrc99a}). For protons, the injection process whereby some fraction of the shock heated plasma becomes accelerated to superthermal energies is, of necessity, parameterized but the model allows for the investigation of parameters against observations. Unfortunately, the theory of electron injection and acceleration in shocks is on a less secure footing than for protons, so we are forced to use additional parameters for electron injection. Helium is included without accounting for the enhancement effect known to occur for high mass to charge number ions (\\egc Baring \\etal\\ \\cite{BaringEtal99}), but this approximation has little effect on the results presented here. We claim our nonlinear model, in spite of its approximations and parameters, is far more physically meaningful than the test-particle power laws that are still routinely assumed by many workers as the outcome from shock acceleration. A crucial property of efficient, nonlinear shock acceleration is the interconnection of the entire particle distribution from thermal to the highest energies, and the linkage between protons and electrons. Because energy is conserved, a change in the production efficiency of the highest energy cosmic rays {\\it must} impact the thermal properties of the shock heated gas and vice versa. If more energy goes into relativistic particles, less is available to heat the gas. In contrast, the power laws assumed by test-particle models have no connection with the thermal gas, energy conservation does not constrain the normalization of the power law, and the spectral index can be changed with no feedback on the thermal plasma. The availability of an easy-to-use nonlinear shock acceleration model will have important implications for interpreting the broad-band emission from cosmic sources. In particular, the model presented here can efficiently explore parameter space to determine which quantities are the most dominant in controlling the spectral character and detectability of a remnant's emission in each waveband. The density of the ambient interstellar medium and the environmental magnetic field strength are the most interesting and critical parameters in this regard. It is clear that adjusting parameters to fit one component, say radio, changes the predicted fluxes at all other frequencies and most significantly in the X-ray band. Based on constraints imposed by radio and $\\gamma$-ray observations, the simple model can distinguish, as a function of source model and environmental parameters, the relative contributions of \\synch\\ (from TeV electrons) and non-thermal \\brems\\ to the X-ray continuum. Our model predicts the shape of the full, nonthermal electron spectrum and, for sources where non-thermal \\brems\\ is significant, X-ray line models may need to consider these non-thermal distributions. Conversely, the model possesses the ability to predict TeV fluxes given detections at other energies. One of the most perplexing developments of cosmic ray physics is the lack of clear detections of pion-decay photons from SNRs. The original predictions of Drury, Aharonian, \\& V\\\"olk (\\cite{DAV94}) are well above current upper limits from EGRET and ground-based TeV telescopes. It is almost universally assumed that supernova blast waves accelerate cosmic ray ions at least up to the ``knee'' near $10^{15}$ eV, and if so, these high energy ions will interact with the ambient gas and produce pions which will decay into GeV and TeV photons. The distinctive pion-decay bump was expected to be seen with current sensitivities and the lack of detection means there is still no direct evidence that SNRs produce cosmic ray ions and is something of a concern for both theorists and builders of $\\gamma$-ray telescopes. Since the detectability depends on source parameters, observers critically need selection criteria which reliably predict which SNRs, observed at other frequencies, are likely to be bright at $\\gamma$-ray energies. This is particularly true for ground-based air \\v{C}erenkov telescopes which must devote weeks or even months observing a single position in the sky to obtain good statistics. As we show in our survey below, the model can provide these selection criteria. For most of the cases we have studied, supernov\\ae\\ which explode in low-density, low magnetic field regions have the largest TeV to radio flux ratios. ", "conclusions": "\\label{sec:conclusion} We have developed a computationally fast and easy-to-use model of nonlinear diffusive shock acceleration along with the accompanying photon emission from the resultant electron and ion spectra. Using this model, we showed how the emission depends on both model and environmental parameters and identified ambient density and magnetic field as being the most important in determining the broad-band emission. We also showed that our simple model is in excellent agreement with a more complete and complex model of the supernova remnant SN1006. A particularly important aspect of nonlinear acceleration is that shock {\\it heating} is linked to particle acceleration and thus to the broad-band photon emission. As shown in Figures~\\ref{fig:fpdjde}, \\ref{fig:paraModel}a, and \\ref{fig:testpart}, shocks which accelerate particles efficiently have lower postshock temperatures and higher postshock densities than test-particle predictions. Nonlinear shocks also produce electron spectra with superthermal tails (Figures~\\ref{fig:fpdjde}, \\ref{fig:SNRcomp}, and \\ref{fig:testpart}). These factors are important for X-ray line emission (\\egc Dorfi \\& B\\\"ohringer \\cite{DorfiB93}) and open up the possibility of using radio, X-ray continuum, and $\\gamma$-ray observations to constrain X-ray line models and vice versa. We believe the model we present here is a first step in this process. To describe SNRs, we use a model of an expanding, spherical shock wave (Truelove \\& McKee \\cite{Truelove99}) to obtain shock parameters as a function of SNR parameters and time. Using these parameters, we calculate nonlinear particle soectra and produce ``snapshot'' continuum photon spectra from \\syn, \\brems, \\IC, and \\pion, spanning the range from radio to TeV $\\gamma$-rays. More realistic models would include inhomogeneous ejecta, emission from dense clumps, pre-SN winds, a reverse shock, volume integrated emission from regions undergoing adiabatic cooling, oblique shock geometry, and effects on the SNR evolution from particle escape (or dilution). Some of these additions have already been performed by Berezhko, Ksenofontov, \\& Petukhov (\\cite{BereKP99a}) and our results are in excellent agreement with theirs for SN1006 (Figure~\\ref{fig:SN1006} and Section~\\ref{sec:BereCompare}), some can be added easily to the simple model, but others (such as oblique geometry) are much more difficult. Despite this, we believe this nonlinear model is accurate enough to approximate expected emission fluxes and clearly track important trends as we describe in detail in Section~\\ref{sec:keyparameters}. Nonlinear shock acceleration unavoidably involves a large number of model and environmental parameters and it is essential to know how they relate to each other and which ones have the greatest impact on the results. Of the model parameters, the injection efficiency, $\\etainjP$, has the greatest influence on the solutions since it sets the overall efficiency and determines whether the acceleration is nonlinear or can be treated as test particle (Figures~\\ref{fig:fpdjde} and \\ref{fig:paraModel}a). Unfortunately, $\\etainjP$ is not well constrained by theory and heliospheric observations remain limited; in fact, there's virtually no constraining information on the efficiency of high Mach number shocks like those expected in young SNRs. However, the differences between test-particle and nonlinear predictions are so large in both the broad-band continuum and the X-ray line emission, that we believe it will be possible to set strong constraints on the acceleration efficiency once nonlinear models of X-ray line emission become available. Other model parameters, such as the ratios of shocked electron to proton temperature and of electron to proton acceleration efficiency, are less fundamental but produce characteristic differences in the emission which can be constrained with sufficiently complete models and observations (Figures~\\ref{fig:paraModel}b and c). The environmental parameters include the ambient interstellar proton density $\\ProDenUpS$ and density profile, the magnetic field $B_0$, the ejecta mass and density profile, supernova explosion energy, etc. Of these, the ambient density and magnetic field have, by far, the largest effect on the broad-band emission. As we show in Figures~\\ref{fig:SNRcomp}, \\ref{fig:paraInputSens}, and \\ref{fig:TeVtoRadio}, varying $\\ProDenUpS$ and $B_0$ produce a complicated set of changes in the emission which cannot be simply characterized. If density is held constant, the intensity in the radio band scales roughly as $B_0^{3/2}$, but at photon energies $E>$ MeV, the intensity can decrease with increasing $B_0$ due to nonlinear effects (Figure~\\ref{fig:paraInputSens}b) and a weakening of the shock (\\iec decrease in the \\alf\\ Mach number). Likewise, when $B_0$ is fixed, an increase in density generally causes an increase in emission in the MeV range, but the emission in the radio band is relatively insensitive to density (Figure~\\ref{fig:paraInputSens}a). At TeV energies, the maximum photon energy increases as $\\ProDenUpS$ is decreased and, in general, the highest photon energies are obtained with the largest $B_0$ and the lowest $\\ProDenUpS$. The TeV/radio flux ratio, however, is greatest for low ambient magnetic fields (Figure~\\ref{fig:TeVtoRadio}) and this is an important parameter if radio SNRs are selected for hard $\\gamma$-ray searches. As we show in Figure~\\ref{fig:ic_pp}, the distinctive pion-decay bump will be most prominent for remnants in high $\\ProDenUpS$ and high $B_0$ environments; given the flux levels predicted, we anticipate positive detections of such spectral features (generally below EGRET sensitivities for $n\\lesssim 10$ cm$^{-3}$), by the GLAST experiment in the not too distant future. While TeV experiments cannot probe such spectral features, they can detect the upper range of pion decay emission in remnants in high $B_0$ surroundings; this situation may already be realized in the recent announcement (V\\\"olk et al. 2000) of a positive detection of Cas A by HEGRA. The nature of the X-ray emission also depends importantly on density and magnetic field. As density is increased at a given $B_0$, the keV X-ray emission goes from being totally dominated by \\syn\\ from relativistic electrons to quasi-thermal \\brem\\ emission. Note that the \\brem\\ dominates even in the radio band for $\\ProDenUpS=10$ \\pcc\\ and $B_0=3$ \\muG (Fig.~\\ref{fig:paraInputSens}a). Similarly, at a given density (Figure~\\ref{fig:paraInputSens}b), the X-ray emission goes from being thermal \\brem\\ to \\syn\\ as $B_0$ increases. If \\syn\\ is dominant, X-ray lines will be weak or absent so these differences are readily distinguishable observationally. Finally, we suggest that the most important aspect of modeling photon emission from astrophysical shocks depends on whether or not the acceleration is efficient and nonlinear or inefficient and test-particle in nature. These situations can be quite different and if the acceleration is nonlinear, all parameters are interconnected and observations in any energy band limit the freedom to vary parameters in all other bands. Since there are a large number of parameters, understanding the nonlinear interactions requires a model that can efficiently map parameter space. We believe the model presented here, while not a replacement for more complete models, can do this expediently and accurately and be an aid in interpreting the vast amount of information expected from current and future space and ground-based telescopes. \\placefigure{fig:ic_pp} % \\vskip 1cm \\noindent {\\bf Acknowledgments:} We wish to thank John Raymond for providing information concerning SN1006. We also thank Rod Lessard for providing Veritas flux sensitivity data and Anne Decourchelle for helpful discussions. This work was performed, in part, with support from NASA's Space Physics Theory Program. \\vskip24pt" }, "0003/astro-ph0003155_arXiv.txt": { "abstract": "We compare the distribution of stars of different spectral types, and hence mean age, within the central SMC and find that the asymmetric structures are almost exclusively composed of young main sequence stars. Because of the relative lack of older stars in these features, and the extremely regular distribution of red giant and clump stars in the SMC central body, we conclude that tides alone are not responsible for the irregular appearance of the central SMC. The dominant physical mechanism in determining the current-day appearance of the SMC must be star formation triggered by a hydrodynamic interaction between gaseous components. These results extend the results of population studies (cf. Gardiner and Hatzidimitriou) inward in radius and also confirm the suggestion of the spheroidal nature of the central SMC based on kinematic arguments (Dopita \\etal; Hardy, Suntzeff \\& Azzopardi). Finally, we find no evidence in the underlying older stellar population for a ``bar'' or ``outer arm'', again supporting our classification of the central SMC as a spheroidal body with highly irregular recent star formation. \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ \\hfill\\break \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ ", "introduction": "Galaxy interactions play a key role in current understanding of galaxy formation and evolution. The dominant physical effect of an interaction is generally thought to arise from the tidal forces exerted between galaxies. Those tidal forces remove angular momentum from the gas, the gas falls toward the center of the galaxies, and that inflow results in a nuclear, or at least centrally concentrated, episode of star formation (cf. Mihos, Richstone, \\& Bothun 1992). However, for small, gaseous galaxies within a larger dark matter halo (for which the encounter velocity is greater than the internal velocity dispersion), the dominant effect of a close interaction or collision could be hydrodynamic as the gaseous components of the two galaxies interact. Perhaps this type of interaction is more typical in the protogalactic environment, where small sub-galactic stellar aggregates are coalescing into a larger galaxy, and among interacting satellite galaxies within the halos of current galaxies. The nearest interacting satellite galaxies for study are the Small and Large Magellanic Clouds (SMC and LMC, respectively). The Magellanic system is highly complex and clearly interacting. It is an ideal laboratory for examining the effects of interactions, such as the development of tidal material and the triggering of star formation. On the largest scales, the Magellanic Stream (Mathewson, Cleary, \\& Murray 1974) is evidently a relic of an interaction, although its origin as tidal, rather than hydrodynamic, is still debated because of the lack of stars found in the stream (cf. Guhathakurta \\& Reitzel 1998). Stellar structures in the Magellanic system that appear to be tidal in origin were identified by Shapley (1940; the eastern SMC wing), by de Vaucouleurs (1955; the LMC tidal tails, although later de Vaucouleurs (de Vaucouleurs \\& Freeman 1972) partially retracted his claim due to potential confusion with galactic diffuse emission), by Hindman, Kerr, \\& McGee (1963; the H I bridge between the LMC and SMC), by Irwin, Demers, \\& Kunkel (1990; the stellar bridge between the LMC and SMC), and by Putman \\etal (1998; the leading Magellanic Stream). Recent numerical simulations (only including gravity; Gardiner and Noguchi 1996) show that such dynamics reproduce many of the observed features, including the SMC wing and the line-of-sight depth of the outer regions of the SMC (Hatzidimitriou and Hawkins 1989). Our study of the distribution of stars in the SMC is complementary to that of the outer SMC by Gardiner and Hatzidimitriou (1992). The structure of the central SMC is currently somewhat more uncertain than that of the outer regions (cf. Hatzidimitriou, Cannon, and Hawkins 1993 for a discussion). We demonstrate, based on the distribution of stars of different ages within the central SMC (4\\degs $\\times$ 4\\degs), that although the existence of tidal forces on the SMC is not in dispute, the visible appearance of the central SMC arises primarily from hydrodynamic effects. The underlying, old stellar population of the central SMC is relatively undisturbed (as concluded by Dopita \\etal 1985; Hardy \\etal 1989 on the basis of kinematic arguments). These results provide direct evidence of star formation triggered by the interaction of the SMC with another gaseous system (presumably, but not demonstrated to be, the LMC). Several common features, such as the ``bar'' and the ``outer arm'' are not seen in the underlying population and so are not true dynamical structures. ", "conclusions": "There are several aspects of the observations that appear in conflict with the hypothesis of a tidal origin for the morphology of the system. First, the recent star formation is more extended than the older stellar system, rather than being more concentrated as most generic interaction models suggest. Second, the older stars do not follow what initially appear to be the tidal ``tails'' as noticeably as the younger stars. There are old stars at large radius from the SMC, but their distribution is much more regular than that of the younger stars (Gardiner and Hatzidimitriou 1992). One must conclude either that the tides predominantly affected the outer part of the SMC, which presumably contained large amounts of gas, that then formed the young stars --- or that star formation in the central SMC was triggered when the SMC interacted directly with a gaseous component (either the LMC gaseous halo or perhaps a third body). The outer tidal hypothesis has several difficulties. First, the stellar extensions appear to go in several directions rather than the standard tail/bridge geometry. Second, if the gaseous material was originally at large radius and of insufficient density to form stars, one would expect that tidal forces would further lower the densities at large radii where the gas has no means to lose angular momentum and dissipate energy. In its favor, at least as an explanation for the wing structure, is the continuity between the wing and bridge (both in stars and gas) that join the SMC and LMC. But if the dominant tidal force created the LMC-SMC bridge, what formed the NE-SW structure of young stars in the SMC? We speculate that the NE-SW geometry arises either from shocking of the gas in the central SMC as the SMC moves in a perpendicular direction (i.e. NW or SE) to a second gaseous object (\\eg speculatively, a hot Milky Way halo or the outer gaseous envelope of the LMC), or that the infall of a gaseous cloud along the NE-SW axis formed stars that have orbits aligned with the infall axis. We see no evidence for shocking in the H I map of Stanimirovic \\etal 1999, although the H I morphology is highly irregular. Furthermore, the H I kinematics contain a large velocity gradient along the ``bar\" axis, which would not arise naturally in a ram pressure model but which could arise in a model of the collision of two gaseous components along the ``bar\" axis. The latter hypothesis is also supported by the presence of the stellar ``shell\" feature (upper left, Figure 2), which could arise if young stars are on radial orbits along the major axis of the distribution. This discussion is highly speculative and a definite conclusion awaits numerical simulations that include both gravity and hydrodynamics, and additional measurements of the stellar kinematics. Despite the disturbed visual appearance of the SMC and the observations supporting large line-of-sight depths in the outer SMC, the bulk of the stars in the SMC's central region evidently form a spheroidal population. This result confirms previous observations of the kinematics of older stars. Dopita \\etal 1985 from PN observations and Hardy \\etal 1989 from C star observations found that the older stellar component in the central SMC region has the kinematics of a spheroidal component, with no clear signs for multiple kinematic components or strong rotation. In contrast, the H I shows a large kinematic gradient along the NE-SW axis (Stanimirovic \\etal 1999). Furthermore, we find no evidence for a bar in the underlying older stellar distribution and the ``outer arm'' is entirely a product of the young main sequence stars. The discrepancies between morphological classifications based on images, which inappropriately weigh the younger populations, and stellar catalogs, from the which the numerically dominant, underlying population can be extracted, highlight the inherent difficulties in morphological studies from integrated photometry. These difficulties are exacerbated in studies at higher redshifts, where the younger stars are even more disappropriately weighted due to K-corrections and where the spatial resolution is poorer. We conclude, 1) that the central SMC is principally a spheroidal system, 2) that its visual morphology is dominated by highly irregular recent star formation, and 3) that hydrodynamics, rather than tidal forces, must play the key role in SMC star formation over the last several hundred million years. \\vskip 1in \\noindent ACKNOWLEDGMENTS: DZ acknowledges partial financial support from an NSF grant (AST-9619576), a NASA LTSA grant (NAG-5-3501), and fellowships from the David and Lucile Packard Foundation and the Alfred P. Sloan Foundation. EKG acknowledges support from NASA through grant HF-01108.01-98A from the Space Telescope Science Institute. \\vskip 1cm \\noindent \\vfill\\eject" }, "0003/astro-ph0003363_arXiv.txt": { "abstract": "s{Structure formation with topological defects is described. The main differences from inflationary models are highlighted. The results are compared with recent observations. It is concluded that all the defect models studied so far are in disagreement with recent observations of CMB anisotropies. Furthermore, present observations do not support 'decoherence', a generic feature of structure formation from topological defects. } ", "introduction": "Even if the big bang is an ``irreproducible experiment'', we want to learn from it as much as possible about the physics at high energies. We have reasons to hope that it may have left traces from energies much higher than those reached in any astrophysical event or terrestrial experiment. Therefore, even if it is irreproducible and hence not as controllable as we might want, we simply cannot afford to ignore the information it may have left. The initial fluctuations in the cosmic matter density and geometry may represent one such trace. In fact, presently there are two relatively worked out ideas for cosmic initial fluctuations, both relying on the physics at very high energies. In the first model, cosmic initial perturbations are due to quantum fluctuations which 'freeze in' as classical fluctuations when they become super-horizon during an {\\em inflationary era}. The second possibility is that {\\em topological defects} which may have formed during a phase transition in the early universe have induced structure formation. This second possibility is the topic of this talk. Here, a pedagogical remark may be in order: Often these two alternatives have been represented as 'inflation versus defects'. This is of course not quite correct, as topological defects have nothing to say about the flatness, the horizon and the monopole or moduli (or whatever unwanted relicts) problems which inflation also solves. It is, however, easy to construct inflationary models where the amplitude of initial fluctuations is much too small to be relevant for structure formation. Therefore, in a model, where cosmic structure is due to topological defects, one needs either inflation prior to defect formation or another mechanism to solve the flatness, horizon and relict problems. The reminder of this talk is organized as follows: In Section~2 I give a short overview on the formation of topological defects during cosmological phase transitions. In Section~3 I discuss the problem of structure formation with topological defects. I will first describe some generic insights and then discuss results for specific models. Conclusions are presented in Section~4. ", "conclusions": "All the defect models studied in detail are in disagreement with current observations. They exhibit no acoustic peaks (global $O(N)$ models) or only one broad hump on too small scales (cosmic strings). Decoherence, which is inherent to the non-linear evolution of the defect source term smears out the distinguished series of acoustic peaks expected in inflationary models. The width of the first peak measured by the the Toco\\cite{toco} and BoomerangNA\\cite{Boom} experiments is relatively narrow, which already clearly disfavors a model where decoherence is important. Secondary peaks in the CMB anisotropy spectrum will finally be a unambiguous sign for a (quasi-)linear process of structure formation like, {\\em e.g.}, inflation. It has been shown, however, that linearly evolving causal scaling seeds might mimic an inflationary CMB and dark matter power spectrum~\\cite{Tu,DS}. Nevertheless, due to causality they differ from inflation in the CMB polarization spectrum\\cite{SZ}. Clearly, such seeds are not topological defects and there is so far no convincing physical motivation to introduce them." }, "0003/astro-ph0003225_arXiv.txt": { "abstract": "We present mid-infrared (5--16$\\mu m$) images and spectra of a sequence of interacting galaxies, observed by ISOCAM. The galaxies were selected as being at progressive stages in the time evolution of a merging event, following what is known as Toomre's ``merger sequence'', and having no detected contribution from an active galactic nucleus (AGN) in their mid-infrared spectrum. To trace the intensity of the global star formation in those galaxies, we use the ratio of the 15$\\mu m$ to 7$\\mu m$ flux. Our analysis indicates that this ratio increases from $\\sim$ 1 to $\\sim$ 5 as galaxies move from the pre-starburst to the merging/starburst phase only to decrease to $\\sim$ 1 again in the post-starburst phase of the evolved merger remnants. Moreover, we find that the variation of this ratio is well correlated with the one of the IRAS 25$\\mu m$/12$\\mu m$ and 60$\\mu m$/100$\\mu m$ flux ratios. The potential to test and improve upon these results using the Infrared Spectrograph (IRS) on board SIRTF is discussed. ", "introduction": "One of the major steps in the understanding of galaxy evolution was the realization that tails and bridges are the result of galaxy interactions (\\cite{tt}). It was also proposed by Toomre (1977) to use the morphology of the observed tidal features and the separation between the galaxies in order to create a ``merging sequence'' of 11 peculiar NGC galaxies, also found in the Arp atlas. Ever since, improvements in numerical modeling of the stellar and gaseous component in galaxies have clearly demonstrated that galaxy interactions cause large scale instabilities in the galactic disks leading to the formation of transient bars which drive the gas into the center of the galaxies (\\cite{bn}). Furthermore, numerous multi-wavelength studies of those systems (see Hibbard 1995; Schweizer 1998 and references therein) have been performed in effort to better understand phenomena such as starburst and AGN activity, as well as mass transfers and morphological transformations associated with interacting galaxies. One of the major quests in those studies remained the identification of observational characteristics which could be used as alternatives of assigning an ``age'' to the event of the interaction (i.e. \\cite{ss92}). The discovery by IRAS of the class of luminous IR galaxies which harbor of obscured massive starbursts (\\cite{soifer}) and the revelation later on that they are also interacting/merging systems (\\cite{sanders}), attracted further attention to this problem (see \\cite{sm} for a review). In this paper we examine the global star formation activity in a sample of interacting galaxies as it becomes evident in the mid-infrared via the heating of the dust. ", "conclusions": "\\subsection{The ISO results} \\begin{figure*}[!ht] \\begin{center} \\resizebox{0.8\\hsize}{!}{\\includegraphics{figure1.ps}} \\end{center} \\caption{In this figure we present four of the eight galaxies of our sample found in increasing stages of interaction. From NGC\\,4676 at the top to NGC\\,3256 at the bottom. For each galaxy we include an optical DSS image on the left, marked with the box imaged by ISOCAM, a 7$\\mu$m image in the middle and the mid-infrared spectrum of the galaxy on the right. Note how the flux beyond 10$\\mu$m progressively increases comparing to the strength of the UIB features. The horizontal bars indicate the width of several of the broad band filters used in the observations. NGC\\,4676 and NGC\\,3632 as well as NGC\\,3921 (see Fig. 2) were observed only in broad band filter mode. \\label{fig1}} \\end{figure*} \\begin{figure*}[!ht] \\begin{center} \\resizebox{0.8\\hsize}{!}{\\includegraphics{figure2.ps}} \\end{center} \\caption{ Same as in Fig. 1 for the remaining four galaxies. The mid-infrared continuum reaches its peak emission in Arp220 and progressively decreases in NGC7252 and NGC3921. The solid circle in the mid-infrared 7$\\mu$m images indicates the FWHM of the point spread function. \\label{fig2}} \\end{figure*} In Figures \\ref{fig1} and \\ref{fig2} we present mid-infrared images and integrated spectra of the galaxies in our sample. All galaxies show evidence of star formation activity as it's indicated by the presence of the Unidentified Infrared Bands (UIBs) in their spectra (\\cite{leger}). Two trends became apparent from those figures: As we move from early stage interactions to mergers the continuum at 12-16\\,$\\mu$m is rising very steeply. This continuum is attributed to Very Small Grains (VSGs) with radius less than 10\\,nm (\\cite{desert}), and is prominent in regions actively forming stars. It reaches its peak in NGC6240 and Arp220, which host massive starbursts, and progressively becomes flatter in post-starburst systems. The fraction of the mid-infrared flux associated with the UIB features decreases when we reach the starburst face. This can be easily seen by observing the strength of the 7.7$\\mu$m feature. This could be due to the fact than in massive starbursts one has numerous young stars and their associated HII regions. As a result the filling factor of the photodissociation regions where UIBs form would decrease as well as the corresponding UIB emission. \\begin{figure}[h] \\resizebox{\\hsize}{!}{\\includegraphics{figure3.ps}} \\caption{A comparison of the variation of the ISOCAM LW3/LW2 flux ratio along the merging sequence, with the well known IRAS flux ratios. Note how well the ISOCAM starburst diagnostic follows the evolution of the star forming activity/merger age of the sequence. One may effectively consider the LW3/LW2 ratio as a tracer of the location of FIR peak of the bolometric luminosity. The IRAS 12$\\mu m$ and 25$\\mu m$ fluxes, corrected for the extent of the galaxies, have been kindly provided by D.B. Sanders (Univ. Hawaii). \\label{fig3}} \\end{figure} One can attempt to quantify this phenomenon by observing the variation of the global mid-infrared colors of the galaxies. We present the flux ratio the total broad band LW3(12-18\\,$\\mu m$)/LW2(5-8.5\\,$\\mu$m) for our galaxies in Figure 3. This has been proposed as an indicator of the fraction of the VSG continuum to the UIB feature emission and it is close to 1 for quiescent star formation (\\cite{boselli}). As we clearly see this ratio presents a monotonic variation with the intensity of the star formation activity in the galaxies. We can also examine how the IRAS colors vary across the same sequence of galaxies. Of particular interest is the IRAS60$\\mu m$/IRAS100$\\mu m$ ratio since this indicates the location of the peak of the spectral energy distribution. The correlation of the ISOCAM LW3/LW2 diagnostic ratio with the IRAS colors is apparent. The only discrepant point is NGC4676, but this can be understood since the one of the galaxies has an old stellar population which can contribute to the mid-infrared emission (\\cite{hibbard}). This result suggests that {\\it even though the bolometric luminosity of interacting luminous infrared galaxies is found at $\\lambda \\geq 40 \\mu m$, the study of the mid-infrared spectral energy distribution can be used to trace the location of the far-infrared peak}. Moreover, the higher spatial resolution one can achieve in the mid-infrared (\\cite{keck}) would further facilitate the identification of the most active regions in the galaxies (ie. \\cite{felix}) \\subsection{The SIRTF/IRS contribution} Strong absorption by dust can distort the apparent morphology of interacting galaxies, hiding the main heating source and revealing to us only reprocessed radiation. Consequently, ISO estimates of the absorption using measurements of line strengths (\\cite{lutz}) may be biased towards lower limits. The use of the depth of the 9.6$\\mu$m silicate absorption feature could be an alternative, but as seen in Arp220 this feature can often be nearly saturated and the underlying continuum is poorly determined. IRS, the infrared spectrograph (\\cite{houck}) on board SIRTF, with a spectral coverage from 5.3 to 40 $\\mu$m will enable to address this issue. Improved estimates on the absorption could be obtained using the depth of both silicate absorption bands (at 9.6 and 18 $\\mu$m) and the shape of the mid-infrared continuum (\\cite{dudley}). Moreover, the superb sensitivity and good spatial and spectral resolution of IRS will allow us to further expand the diagnostic of Fig.3 in fainter more distant systems." }, "0003/hep-ph0003183_arXiv.txt": { "abstract": "Recently we have made considerable progress in our understanding of the behavior of QCD in extreme conditions of high temperature or large baryon number density. Among the highlights are the prediction of a well-characterized true critical point, and the discovery that the ground state of three-flavor QCD at asymptotically high densities exhibits color-flavor locking. The critical point occurs at the unique temperature and density where a sharp distinction between an ionized plasma of quarks and gluons and the hadronic phase first appears. It appears to be accessible both to numerical and to laboratory experiments. Color-flavor locking provides a calculable, weak-coupling realization of confinement and chiral symmetry breaking. It also provides a microscopic realization of Han-Nambu charge assignments for quark quasiparticles, and of Yang-Mills theory for the physical vector mesons. Here I provide a self-contained introduction to these developments. ", "introduction": "In some ways, QCD is a mature subject. Its principles are precisely defined, and they have been extensively confirmed by experiment. QCD specifies unambiguous algorithms, capable of transmission to a Turing machine, that supply the answer to any physically meaningful question within its domain -- any question, that is, about the strong interaction. I believe there is very little chance that the foundational equations of QCD will require significant revision in the foreseeable future. Indeed, as we shall soon review, these equations are deeply rooted in profound concepts of symmetry and local quantum field theory, which lead to them uniquely. So one cannot revise the equations without undermining these concepts. Granting that the foundations are secure, we have the task -- which is actually a wonderful opportunity -- of building upon them. Due to the peculiarities of QCD, this is a particularly interesting and important challenge. Interesting, because while the foundational equations are conceptually simple and mathematically beautiful, they seem at first sight to have nothing to do with reality. Notoriously, they refer exclusively to particles (quarks and gluons) that are not directly observed. Less spectacular, but more profound, is the fact that equations exhibit a host of exact or approximate symmetries that are not apparent in the world. One finds that these symmetries are variously hidden: confined in the case of color, anomalous in the cases of scale invariance and axial baryon number, spontaneously broken in the case of chirality. It is fascinating to understand how a theory that superficially appears to be ``too good for this world'' actually manages to describe it accurately. Conversely it is pleasing to realize how the world is, in this profound and very specific way, simpler and more beautiful than it at first appears. Important, because there are potential applications to which the microscopic theory has not yet rendered justice. An outstanding, historic, challenge is to derive the principles of nuclear physics. At present meaningful connections between the microscopic theory of QCD and the successful practical theory of atomic nuclei are few and tenuous, though in principle the former should comprehend the latter. This seems to be an intrinsically difficult problem, probably at least as difficult as computing the structure of complex molecules directly from QED. In both cases, the questions of interest revolve around small energy differences induced among valence structures, after saturation of much larger core forces. If one starts calculating from the basic equations, unfocussed, then small inaccuracies in the calculation of core parameters will blur the distinctions of interest. A related but simpler class of problems is to calculate the spectrum of hadrons, their static properties, and a variety of operator matrix elements that are vital to the planning and interpretation of experiments. This is the QCD analogue of atomic physics. Steady progress has been made through numerical work, exploiting the full power of modern computing machines. Other significant applications appear more accessible to analytic work, or to a combination of analytics and numerics. The behavior of QCD at high temperature and low baryon number density is central to cosmology. Indeed, during the first few seconds of the Big Bang the matter content of the Universe was almost surely dominated by quark-gluon plasma. There are also ambitious, extensive programs planned to probe this regime experimentally. The behavior of QCD at high baryon number density and (relatively) low temperature is central to extreme astrophysics -- the description of neutron star interiors, neutron star collisions, and conditions near the core of collapsing stars (supernovae, hypernovae). Also, we might hope to find -- and will find -- insight into nuclear physics, coming down from the high-density side. The special peculiarity of QCD, that its fundamental entities and abundant symmetries are well hidden in ordinary matter, lends elegance and focus to the discussion of its behavior in extreme conditions. Quarks, gluons, and the various symmetries will, in the right circumstances, come into their own. By tracing symmetries lost and found we will be able to distinguish sharply among different phases of hadronic matter, and to make some remarkably precise predictions about the transitions between them. In these 5 lectures I hope to provide a reasonably self-contained introduction to the study of QCD in extreme conditions. The first lecture is a rapid tour of QCD itself. I've organized it as a cluster of related stories narrating how apparent symmetries of the fundamental equations are hidden by characteristic dynamical mechanisms. Lectures 2-3 are mainly devoted to QCD at high temperature, and lectures 4-5 to QCD at high baryon number density. I have structured the lectures so that they head toward two climaxes: the prediction of a true critical point in real QCD, that ought to be accessible to numerical and laboratory experiments, in Lecture 3, and the prediction that at asymptotic densities QCD goes over into a color-flavor locked phase with remarkable properties including fully calculable realizations of confinement and chiral symmetry breaking, in Lecture 5. These are, I believe, remarkable results, and they bring us to frontiers of current research. A wide variety of tools will be brought to bear, including three different renormalization groups (the usual `asymptotic freedom' version toward bare quarks and gluons at high virtuality, the usual `Kadanoff-Wilson-Fischer' version toward critical modes at a second-order phase transition, and the `Landau-Anderson' version toward quasiparticles near the Fermi surface), perturbative quantum field theory, effective field theory, instantons, lattice gauge theory, and BCS pairing theory. Of course I won't be able to give full-blown introductions to all these topics here; but I'll try to give coherent accounts of the concepts and results I actually need, and to supply appropriate standard references. In a few months, I hope, a comprehensive reference will be appearing. ", "conclusions": "" }, "0003/astro-ph0003296_arXiv.txt": { "abstract": "Expansion distances (or expansion parallaxes) for classical novae are based on comparing a measurement of the shell expansion velocity, multiplied by the time since outburst, with some measure of the angular size of the shell. We review and formalize this method in the case of prolate spheroidal shells. For such shells there is no unique angular size except when the shell is seen pole-on, and several different measures of angular size have been used in the literature. We present expressions for the maximum line-of-sight velocity from a complete, expanding shell and for its projected major and minor axes, in terms of the intrinsic axis ratio and the inclination of the polar axis to the line of sight. For six distinct definitions of ``angular size'', we tabulate the error in distance that is introduced under the assumption of spherical symmetry (i.e., without correcting for inclination and axis ratio). The errors can be significant and systematic, and can affect studies of novae whether considered individually or statistically. Each of the six estimators overpredicts the distance when the polar axis is close to the line of sight, and most underpredict the distance when the polar axis is close to the plane of the sky. Use of the straight mean of the projected semimajor and semiminor axes gives the least distance bias for an ensemble of randomly oriented prolate shells, and we recommend this method when individual inclinations and axis ratios cannot be ascertained. The best individual expansion distances, however, result from a full spatio-kinematic modeling of the nova shell. We discuss several practical complications that affect expansion distance measurements of real nova shells. We recommend that nova shell expansion distances be based on velocity and angular size measurements made contemporaneously if possible, and that the same ions and transitions be used for the imaging and velocity measurements. We emphasize the need for complete and explicit reporting of measurement procedures and results, regardless of the specific method used. ", "introduction": "The distance to a classical nova in the Galaxy is best inferred by comparing the angular size of the resolved nova shell with the size calculated from its rate of expansion and the time since the shell was ejected. However, if nova shells are ellipsoids of revolution (spheroids) rather than spherical, then the concept of ``angular size'' is ambiguous, and the expansion velocity along the line of sight does not correspond to the transverse expansion velocity. Thus the use of formulas that are valid in the spherical case will lead to erroneous distance estimates. Individual distance estimates may be too large or too small, depending on the true axis ratio of the nova shell and its inclination to the line of sight. Furthermore, these errors do not necessarily average toward zero when ensemble averages are taken. In this paper, we consider systematic errors in estimates of nova shell expansion distances, and recommend procedures to minimize the errors. \\subsection{The Usefulness of Nova Distances} As with all classes of astronomical objects, our understanding of the classical nova phenomenon depends on having accurate estimates of the distances of these objects. In turn, having a well-founded understanding of the distances and luminosities of novae allows them to be studied as astrophysical objects and to be exploited for other purposes. The need for accurate distances is evident, both for novae taken individually and for novae used collectively, i.e., in a statistical fashion. An accurate distance to an individual nova, combined with good coverage of the outburst light curve and a knowledge of the interstellar extinction, can allow the theory of the nova outburst to be verified and further developed. For example, it is possible to check whether there is a phase after maximum light during which the nova's luminosity is close to the Eddington limit. All inferences about the mass of the shell that is ejected during the outburst depend on some power of the distance, through establishing the volume occupied by the emitted gas. At late stages in the evolution of a nova, when the shell can be resolved from the central binary star, the distance is needed to convert the angular size of the shell into a linear size, so that the physical conditions in the ejected gas (ionization, excitation) can be related to the ionizing flux from the central white dwarf (the post-nova) and its accretion disk. An accurate distance to a classical nova also allows the modeling of the accretion process in the post-nova system; without a distance constraint (leading to a constraint on the luminosity), it has proven impossible to infer uniquely the mass accretion rate onto the white dwarf (Wade 1988, Wade \\& Hubeny 1998). Treated collectively, novae have the possibility of providing a secondary or even primary distance indicator for the extragalactic distance scale. Since novae are present in galaxies of all Hubble types, they have the potential to be used directly to compare and unite the distance scales of spiral and elliptical galaxies. The so-called maximum magnitude -- rate of decline (MMRD) relation gives the visual absolute magnitude at maximum light, from a measurement of the rate of decline after maximum (or equivalently the time taken to decline 2 or 3 magnitudes). The shape of the mean MMRD curve and the dispersion around this mean relation have been found from observations in external galaxies (e.g., Della Valle \\& Livio 1995). For the MMRD relation to be a primary distance indicator, however, the zero-point calibration must be provided by Galactic novae. Another proposed distance indicator is the absolute magnitude at 15 days past maximum light ($M_{15}$), where the dispersion in absolute magnitude is small for all novae taken without regard to speed class. The same remarks about zero-point calibration apply to this method.\\footnote{As one step in calibrating the MMRD relation, Cohen (1985) adjusted $M_V$ at maximum for her best observed novae, assuming that they had {\\em identical} $M_{15}$; this step should be replaced by actually measuring the dispersion in $M_{15}$.} Even the nearest Galactic novae are generally too distant for direct trigonometric parallax measurements. Instead, indirect methods of distance estimation are often used, based on the Galactic rotation curve, the total amount of interstellar reddening, the presence or absence of discrete components (``clouds'') in interstellar absorption lines, etc. The only geometrical (hence fundamental) method is that of ``expansion distance'' (also referred to as ``expansion parallax''), in which the measured angular size of the resolved nova shell is compared with the linear size of the shell; the latter is calculated from the expansion speed of the shell gas and the known time since the outburst. In her work on the MMRD relation, Cohen (1985) had only eleven novae with well-observed expansion distances and suitable coverage of their light curves. This is largely because the surface brightness of a nova shell declines rapidly with time since outburst, so that by the time the shell is large enough to be resolved from the ground, it is often too faint to observe. Since Cohen's study, several additional novae have had good light curve coverage, and expansion distances may become available for these from ground-based observations and especially from the {\\it Hubble Space Telescope\\/}\\footnote{Narrow-band imagery of recent novae has been carried out with the HST Wide Field/Planetary Camera 2, used in ``snapshot'' mode, as part of program 7386; these images are public. HST imagery of somewhat older nova shells is discussed in Gill \\& O'Brien (2000).}, since the latter can resolve some shells within a few months or years of the outburst. The simplest way to derive an expansion distance is to assume (often implicitly) that the nova shell is expanding spherically symmetrically, hence that the transverse velocity of gas in the plane of the sky is the same as the radial velocity of gas moving directly along the line of sight. What if the ejection of the shell is asymmetric? To be specific, suppose the shell expands as a spheroid, the simplest generalization from the spherically symmetric case, and one that suffices to describe many actual nova shells. First, the projected image of the nova shell will not be circular for most orientations, and thus there will be an ambiguity in what is meant by angular size. Second, the maximum velocity along the line of sight will usually not correspond to {\\em either} the ``polar'' or ``equatorial'' expansion velocity. For example, suppose that the angular size is taken to be the largest projected ``radius'' of the nova shell, which is perhaps the easiest size parameter to estimate on a barely resolved image. If all nova shells were oblate, then the calculated expansion distance (based on the assumption of spherical symmetry) would always be less than the true distance, because the line-of-sight velocity would be smaller than the transverse expansion velocity. The resulting nova distance scale would be too short. On the other hand, if all nova shells are prolate, than the distance to an individual object may be underestimated or overestimated, depending on the orientation and the ratio of major and minor axes. While it is clear that the distance to an individual nova can be in error as a result, it was not made clear until the work of Ford \\& Ciardullo (1988; hereinafter FC88) that in the prolate case, a systematic error might remain, even after averaging over an ensemble of novae that are taken to be randomly oriented in space. In their analysis, FC88 made the assumption stated above as an example, that the angular size of the nova shell is taken to be the major axis of the projected image. However, all workers do not make this identification. For example, Cohen \\& Rosenthal (1983) did use the projected semimajor axis for the angular size, but Cohen (1985) used an angle-averaged radius. What way is best? A goal of this paper is to extend the FC88 analysis to include consideration of six distinct yet plausible ways of defining the angular size of the shell. As the number of Galactic novae with well observed light curves increases, and with the much greater resolving power provided by adaptive optics and HST, it is likely that the calibration of the MMRD and $M_{15}$ relations will be improved, but the question of possible systematic errors in the distances becomes more important. This is especially so, if shell morphology is related to nova speed class, as has been suggested by Slavin, O'Brien, \\& Dunlop (1995). Likewise, as more expansion distances for individual novae become available, it is important to have clearly in mind whether and how much these distances may be in error, as the result of measuring uncertainties and modeling assumptions. \\subsection{Prolate or Oblate?} Theoretical arguments have been made favoring both oblate and prolate geometries for nova shells (Porter, O'Brien, \\& Bode 1998 and references therein). Empirically, it is now the consensus that, to the extent that nova shells can be described by spheroids, they are either prolate or spherical, but not oblate. FC88 discussed the few cases of resolved nova shells known at the time, in terms of whether they were elongated along one axis (prolate spheroids) or compressed along one axis (oblate spheroids). FC88 noted that most data on the shapes of nova shells were consistent with spherical or prolate geometries, but categorized the shell of nova HR Del 1967 as oblate. For this object, early models by Hutchings (1972) and Soderblom (1976) indeed suggested an oblate symmetry. A spatio-kinematic model by Solf (1983), however, has clearly shown that the resolved shell of HR Del is consistent with a prolate geometry, and not consistent with being oblate. Slavin, O'Brien, \\& Dunlop (1995) carried out imaging of nova shells using narrow band filters; in particular they have obtained images at several different tilts of an interference filter with nominal wavelength 6560 \\AA\\ (17 \\AA\\ FWHM), which allowed them to distinguish crudely between gas approaching or receding from the observer. Their data are clearly consistent with the shells being prolate, not oblate, if they depart detectably from spherical symmetry. Other spatio-kinematic studies, for example of the shell around nova DQ Her 1934 (e.g. Herbig \\& Smak 1992) also indicate prolate symmetry. Therefore we proceed with the assumption that to first approximation, nova shells are prolate spheroidal shells, with their properties in projection specified by their axis ratio and the inclination of the polar axis to the line of sight. In Section 2 of this paper, we investigate several different ways of defining the angular size of the resolved nova shell. We derive the projected size and shape of the shell and the maximum radial velocity of gas in the shell, as functions of the intrinsic axis ratio and the inclination of the polar axis. We then derive analytic expressions that give the inferred distance in terms of the true distance, as a function of axis ratio and inclination. We tabulate results for a variety of cases, using six definitions of ``angular size.'' In Section 3 we investigate how these various definitions of expansion distance behave, both for individual objects and when averaged over an ensemble of nova shells oriented randomly in space. We also discuss some practical matters relating to the measurement of the expansion speed and angular size of nova shells. We summarize our findings in Section 4. ", "conclusions": "We have reviewed and formalized the method of nova shell expansion distances as a means of estimating the distances to classical novae. This method combines a measurement of the shell expansion velocity (multiplied by the time since outburst) with some measure of the angular size. Expansion distances for novae underlie the calibration of the MMRD and $M_{15}$ relations and also form the basis for astrophysical studies of individual novae and their remnants. It is therefore important to adopt methods of measurement that minimize any possible bias in the distances that results from incomplete information about the shape or orientation of the nova shells. Many resolved shells exhibit significant prolate symmetry, so that there is no unique angular size except when the shell is seen pole-on. We developed analytic expressions for the maximum line-of-sight velocity from a complete, expanding prolate spheroidal shell and for its projected major and minor axes, as functions of the intrinsic axis ratio and the inclination of the polar axis to the line of sight. For six definitions of ``angular size'', we then computed the error introduced by deriving a distance using the assumption of spherical symmetry (i.e., without correcting for inclination and axis ratio). The errors can be significant and possibly systematic, affecting studies of novae whether considered individually or statistically. The definition of angular size that results in the least errors at the extremes is $\\rho_6$, the harmonic mean of the projected semimajor and semiminor axes. However, the definition that results in the least bias when an ensemble of randomly oriented prolate shells is considered is $\\rho_3$, the straight mean of the projected semimajor and semiminor axes, and we recommend this method when individual inclinations and axis ratios cannot be ascertained. The $\\rho_3$--based method is always as good or better than than the $\\rho_1$ method (projected semimajor axis alone). The best individual expansion distances result from a full spatio-kinematic modeling of the nova shell, using spectroscopy of emission lines at multiple locations across the resolved shell. We have discussed practical issues and made recommendations for observers who make measurements of either the maximum line-of-sight velocity and the angular size of a resolved nova shell. The velocity measurement may be complicated by the presence of line emission from the central cataclysmic binary star, and if the spheroidal shell is not complete, the theoretical maximum velocity may not be observed at all. The correct application of angular size measurements can be compromised by convolution with the image point spread function, by improper technique, or by incomplete reporting. For best results, velocity and angular size measurements should be made contemporaneously, and must refer to the same features of the shell. Observers are encouraged to report as completely as possible the measurements they have made. Estimates of nova distances by the shell expansion method (or any other method) should be accompanied by a discussion of both random and systematic errors, including possible effects due to unaccounted-for departures from spherical symmetry, as discussed in this paper." }, "0003/astro-ph0003405_arXiv.txt": { "abstract": "We present the results from both the timing and spectroscopic analysis of the transient X--ray pulsar \\G1 observed by the \\B satellite on 1997 April 4, when the source was at a luminosity of $\\sim10^{37}$~\\ergs. \\G1 shows a very hard spectrum that is well fitted by an absorbed power law (N$_H$ $\\sim 2.3 \\times 10^{22}$ cm$^{-2}$) modified by a high energy cut--off above 6~keV. The source shows a small pulse amplitude in the whole energy band. The pulse profile evolves with energy from a double--peaked to single--peaked shape. The barycentric pulse period is $29.477\\pm0.001$ s. ", "introduction": "The transient X--ray pulsar \\G1 was discovered on 1988 April 3 during a galactic plane scan observation near the Scutum region by the {\\em Ginga \\rm} satellite (\\cite{Makino88a}; \\cite{Makino88b}). The Large Area Counter (LAC, \\cite{Turner89}) on board {\\em Ginga \\rm} (\\cite{Makino87}) detected at a J2000 position of ${\\rm \\alpha = 18^{h} 45^{m}\\pm1^{m}}$, ${\\rm \\delta = 0^\\circ 55'\\pm7'}$, a coherent pulsation with a period of 29.5 s and an 2--37~keV X--ray intensity of 50 mCrab (\\cite{Koyama90a}). On 1988 April 19 and 20 {\\em Ginga \\rm} carried out a pointed observation of \\G1, measuring a highly variable X--ray flux on a wide range of time scales, ranging from 30 to 60 mCrab. In addition to a coherent oscillation with P$= 29.5056 \\pm 0.0002$~s, an energy--dependent aperiodic variation was found (Koyama et~al. \\cite{Koyama90b}). On 1997 March 3, the Burst and Transient Source Experiment ({\\em BATSE}) on board CGRO detected a new outburst from this peculiar source (\\cite{Wilson97}). The mean 20--50~keV {\\it rms} pulsed flux was $37\\pm2$ mCrab while the mean barycentric pulse period at an epoch of March~6.0 was P$ = 29.5631 \\pm 0.0003$~s. The P variation during this observation implies a spin--up rate, ${\\rm \\dot P}$, of $(-3.65\\pm0.11)\\times10^{-8}$ s~s$^{-1}$. The data confirmed the low pulsed fraction ($\\sim$7\\%) observed by {\\em Ginga}. Between 1997 February~1 and March~19 the All Sky Monitor (ASM) on board the Rossi X--ray Timing Explorer ({\\em RXTE}) observed the source to be at a flux level of ${\\rm F_{2-10} \\sim 15-30 }$~mCrab (\\cite{Takeshima97}). A pointed observation carried out on 1997 March~5 with the {\\em RXTE} Proportional Counter Array (PCA) detected the source at a 2--60~keV flux level of 62 mCrab, measuring a barycentric pulse period of $29.565\\pm0.002$~s at an epoch of March 5.1712 UT (\\cite{Takeshima97}). On 1997 April~4 the \\B Narrow Field Instruments (NFIs) performed a pointed observation of \\G1 (\\cite{Piraino98}). The source flux was $(2.9\\pm0.3)\\times10^{-9}$erg cm$^{-2}$s$^{-1}$ in the 0.3--100~keV energy range. Using the capability of \\B imaging instruments, the 90\\% confidence J2000 position of \\G1 was constrained to be within a $30''$ radius circular error region centered on ${\\rm \\alpha = 18^{h} 45^{m} 34^{s}}$, ${\\rm \\delta = 0^{\\circ} 52\\farcm5}$ (\\cite{Santangelo97}). On the same day, the source was also observed by the {\\em ROSAT} High Resolution Imager (HRI) which found a J2000 position for the source of ${\\rm \\alpha = 18^{h} 45^{m} 36\\fs9}$, ${\\rm \\delta = 0^{\\circ} 51' 45''}$ (90 \\% confidence error radius $10''$, \\cite{Dennerl97}). In this {\\it paper} we present the results of both a timing and spectroscopic analysis of the \\B observation of \\G1. ", "conclusions": "After its discovery in 1988, \\G1 was detected again in 1997 March as a bright (0.3--100~keV) $\\sim 2.9\\times10^{-9}$ erg cm$^{-2}$s$^{-1}$ X--ray source. Due to the spatial capabilities of the \\B imaging instruments an improved position was obtained. The \\B position is within the {\\it Ginga} (Koyama et~al. \\cite{Koyama90b}) and RXTE (\\cite{Chakrabarty97}) error boxes, and is also consistent with that measured by the {\\it ROSAT} HRI (\\cite{Dennerl97}). Accurate measurement of the position of the source is important in order to carry out a systematic search for the still unidentified optical counterpart. Pulsations with a period $P=29.477\\pm 0.001$~s together a mean pulse period change ${\\rm \\dot{P}/P =-4.1\\times10^{-2}}$ yr$^{-1}$, which is in good agreement with the one measured by {\\it Ginga} were found. Koyama et~al. (\\cite{Koyama89}) suggested that such a high spin up rate could be due, at least partly, to an orbital Doppler motion. Pulse period variations observed in the 30 days monitoring obtained by combining data from {\\em BATSE}, {\\em RXTE--PCA} and \\B, confirmed the presence of a high intrinsic spin--up rate. Moreover, assuming a Be transient system having an orbital period between 50 and 60 days, inferred from the pulse--orbital periods relation of Corbet (\\cite{Corbet86}), a possible Doppler effect may be overlapped to this intrinsic spin--up rate. The source spectrum, which is well described by an absorbed power law with high energy cut--off, is typical of accreting X--ray pulsars. The very high absorption, ${\\rm N_{H} = 2.3 \\times 10^{22} cm^{-2}}$ is consistent with that reported by Koyama et~al. (\\cite{Koyama90b}). The hypothesis that the absorption is mainly interstellar rather then circumstellar (\\cite{Koyama90a}) is supported by the marginal detection of a fluorescent K$_\\alpha$ iron line in the source spectrum. Assuming a distance of 10 kpc (\\cite{Koyama90a}; \\cite{Hayakawa77}) the 0.3--100~keV luminosity is $\\sim$$3\\times10^{37}$~erg~ s$^{-1}$. It is unclear if cyclotron resonance scattering features are present in the hard X--ray spectrum of the source. Koyama et~al. (\\cite{Koyama90b}) suggested that the cut--off in the spectrum observed at $\\sim$18~ keV could be related to a very intense magnetic field typical of this class of source. Moreover, Mihara (\\cite{Mihara}), fitting the phase resolved spectrum with an absorption--like feature at $\\sim$20~keV, classified \\G1 as a possible cyclotron source. Although the spectrum is observed with good statistics up to $\\sim$100~keV, no evidence of any cyclotron feature is observed in the \\B pulse phase averaged spectrum of \\G1. Also the \"crab--ratio technique\"(\\cite{DalFiume98}), successfully exploited in detecting Resonance Cyclotron Features (RCFs) in other X--ray pulsars, does not display any sign of cyclotron features. Moreover no evidence of cyclotron absorption features was found in the phase resolved spectra below 100 keV. However, we found an upper limit on the depth of 0.15 for the possible 20 keV feature. This value is compatible with that found by Mihara (\\cite{Mihara}). Manchanda (\\cite{Manchanda99}), using data from the LASE experiment, a balloon--born large area scintillation counter, recently suggested the possibility of an absorption feature around 100 keV or an emission at 140 keV. Unfortunately, statistics of \\B spectra is quite low at that energy and a much deeper analysis, which is underway, is required. There are $\\sim$80 known accreting X--ray pulsars (see \\cite{Bildsten97} for a recent review). Until recently only the relatively bright nearby pulsars were visible due to the limited sensitivity of previous detectors. This is changing with the discovery by {\\em ASCA}, {\\em ROSAT}, \\B and {\\em RXTE} of a population of faint, absorbed pulsars (e.g., \\cite{Angelini98}; \\cite{Kinugasa98}, \\cite{Torii98}). The search for faint pulsars is one of the main scientific objectives of the {\\em ASCA} galactic plane survey (e.g., \\cite{Sugizaki97}; \\cite{Torii98})." }, "0003/astro-ph0003342_arXiv.txt": { "abstract": "We report on radio continuum and \\HI\\ observations of the radio source \\g328\\ using the Australia Telescope Compact Array. Our results confirm \\g328\\ to be a filled-center nebula with no surrounding shell, showing significant linear polarization and an almost flat spectral index. These results lead us to conclude that \\g328\\ is a Crab-like, or ``plerionic'', supernova remnant (SNR), presumably powered by an unseen central pulsar. \\HI\\ absorption towards \\g328\\ puts a lower limit on its distance of $17.4\\pm0.9$~kpc, making it the largest ($D=25$~pc) and most luminous ($L_R = 3\\times10^{35}$~erg~s$^{-1}$) Crab-like SNR in the Galaxy. We infer \\g328\\ to be significantly older than the Crab Nebula, but powered by a pulsar which is fast spinning ($P<20$~ms) and which has a comparatively low magnetic field ($B<10^{12}$~G). We propose \\g328, G74.9+1.2 and N157B as a distinct group of large-diameter, high-luminosity Crab-like SNRs, all powered by fast-spinning low-field pulsars. ", "introduction": "\\label{sec_intro} When a massive star ends its life in a supernova explosion, the outer layers of the star are expelled at high velocity, and interact with the ambient interstellar medium (ISM) to produce a supernova remnant (SNR). ``Shell'' SNRs are usually characterized by radio synchrotron emission with a limb-brightened morphology, roughly centered on the site of the supernova explosion. Some SNRs are classified as ``composite'', indicating that as well as a shell they have an additional central component, characterized by a filled-center morphology, a flat spectral index ($-0.3 \\la \\alpha \\la 0$; $S_\\nu \\propto \\nu^\\alpha$) and significant linear polarization. This extra component is usually interpreted as a synchrotron nebula, powered by a pulsar also formed in the supernova explosion (\\cite{mgh+79}; \\cite{wp80}; \\cite{rc84}). Thus even when a pulsar itself is not detected (as is the case much more often than not), the mere existence of one of these synchrotron nebulae tells us that an energetic pulsar must be located within. The best known example of a pulsar-powered nebula is the Crab Nebula. However, it has long been apparent that the Crab is very different from most other SNRs, in that it consists of a synchrotron nebula and associated pulsar, but has {\\em no}\\ surrounding shell corresponding to the supernova blast wave. It is now recognized that about 5\\% of all SNRs similarly lack shells; these sources are generally referred to as ``Crab-like'' SNRs, or ``plerions'' (\\cite{wp78}). It is still an open question as to whether Crab-like SNRs have surrounding shells which are simply not detectable (\\cite{che77}; \\cite{sh97}), or whether they are fundamentally different from other SNRs in that they have no shell, invisible or otherwise (\\cite{nom87}; \\cite{wlt97}). Crab-like SNRs are thus important to identify and study, both for understanding the nature and evolution of SNRs, and because they are an unambiguous indication of the presence of a young pulsar. The source \\g328\\ (MSH~15--5{\\em 7}; \\cite{msh61}) has been been the subject of several radio studies (\\cite{sg70c}; \\cite{chmw80}; \\cite{wg96}), which have indicated a featureless filled-center appearance and a comparatively flat spectral index. These properties have resulted in \\g328\\ being proposed as a member of the Crab-like class of SNRs, and indeed is the only Galactic Crab-like SNR easily accessible to Southern hemisphere observers. \\HI\\ measurements with a two-element interferometer have shown \\HI\\ absorption in this direction to a distance in excess of 20~kpc (\\cite{cmr+75}). This large distance would make the SNR extraordinary luminous, even brighter than the Crab Nebula itself. However, previous images of this source have been of only intermediate resolution (the best observations to date had only six beams across the SNR), while the lack of imaging capability in the \\HI\\ measurements of Caswell \\etal\\ (1975\\nocite{cmr+75}) mean that the absorption they saw could have been coming from another bright source in the vicinity, rather than from the SNR itself. We therefore present new radio observations of \\g328\\ both in polarimetric continuum and in the \\HI\\ line, aimed at confirming the Crab-like nature of this source, verifying the large distance claimed for it, and studying it at much higher resolution and sensitivity than in previous work. Our observations and analysis are described in \\S\\ref{sec_obs}, and our results are presented in \\S\\ref{sec_results}. In \\S\\ref{sec_discuss} we confirm \\g328\\ as a Crab-like SNR, and infer the properties of the central pulsar presumed to be powering it. A complementary study of this source in X-rays is reported in a companion paper by Hughes, Slane \\& Plucinsky (2000\\nocite{hsp00}). ", "conclusions": "Our radio observations of \\g328\\ confirm it to be a Crab-like SNR at a distance of $>$17~kpc, making it the largest and most radio-luminous such object in our Galaxy. \\g328, together with the Galactic SNR G74.9+1.2 and N157B in the LMC, appear to form a small subset of Crab-like SNRs with both high radio luminosities and large diameters. The high luminosities of these remnants demand a high value of $\\dot{E}$ for their central pulsars, while these remnants' large extent, even assuming an energetic pulsar and a low ambient density, requires them to be significantly older than the Crab Nebula. Since $\\dot{E}\\tau_c \\propto P^{-2}$ and $\\dot{E}\\tau_c^2 \\propto B^{-2}$, this combination of a high $\\dot{E}$ and large age can only be produced by a pulsar which is spinning at least twice as fast as the Crab Pulsar, but which has a magnetic field $\\sim$5 times weaker. Specifically, we infer \\g328\\ to be $\\sim$7000~yrs old and powered by a pulsar with period $P\\approx11$~ms and dipole magnetic field $B=6\\times10^{11}$~G. Similar properties are predicted for G74.9+1.2 and for N157B, which in the latter case agree with those of the pulsar recently detected within this SNR. Using the distribution of pulsar initial magnetic fields proposed by Stollman (1987\\nocite{sto87b}), we can estimate that the birth-rate of pulsars with magnetic fields $B \\sim 6\\times10^{11}$~G is a fraction 0.05 of that of pulsars with magnetic fields comparable to the Crab Pulsar. Thus of the $\\sim$35 Crab-like and composite SNRs known in the Galaxy, we can expect $\\sim$2 to be powered by such low-field pulsars, and \\g328\\ and G74.9+1.2 may thus represent the complete sample of such sources. The radio beaming fraction for young pulsars is estimated to be in the range 50--70\\% (\\cite{fm93}; \\cite{bj98}). Thus the failure to detect radio pulsations from \\g328, G74.9+1.2 and N157B (\\cite{kmj+96}; \\cite{llc98}; \\cite{ckm+98}) is unlikely to be solely due to beaming, and more likely results from the large distance to these objects. With continuing improvements in sensitivity, searches for pulsations towards these sources should eventually be successful." }, "0003/astro-ph0003174_arXiv.txt": { "abstract": "{ We present an axisymmetric numerical model of a dynamo active accretion disc. If the dynamo-generated magnetic field in the disc is sufficiently strong (close to equipartition with thermal energy), a fast magneto-centrifugally driven outflow develops within a conical shell near the rotation axis, together with a slower pressure driven outflow from the outer parts of the disc as well as around the axis. Our results show that a dynamo active accretion disc can contribute to driving an outflow even without any external magnetic field. The fast outflow in the conical shell is confined by the azimuthal field which is produced by the dynamo in the disc and advected to the disc corona. This part of the outflow has high angular momentum and is cooler and less dense than its surroundings. The conical shell's half-opening angle is typically about $30^\\circ$ near the disc and decreases slightly with height. The slow outflow is hotter and denser. ", "introduction": "The importance of magnetic fields for disc accretion is widely recognized (e.g.,\\ Mestel 1999), and the turbulent dynamo is believed to be a major source of magnetic fields in accretion discs (Pudritz 1981a, 1981b; Stepinski \\& Levy 1988; Brandenburg et al.\\ 1995; Hawley et al.\\ 1996). A magnetic field anchored in the disc is further considered to be a major factor in launching and collimating a wind in young stellar objects and active galactic nuclei (Blandford \\& Payne 1982; Pelletier \\& Pudritz 1992); see K\\\"onigl \\& Pudritz (2000) for a recent review of stellar outflows. Yet, most models of the formation and collimation of jets rely on an externally imposed poloidal magnetic field and disregard any field produced in the disc. Our aim here is to study outflows in connection with dynamo-generated magnetic fields. We discuss parameters of young stellar objects in our estimates, but the model also applies to systems containing a compact central object after rescaling and possibly other modifications such as the appropriate choice of the gravitational potential. Extensive numerical studies of collimated disc winds have been performed using several types of model. Uchida \\& Shibata (1985), Matsumoto et al.\\ (1996) and Kudoh et al.\\ (1998) consider an ideal MHD model of an accretion disc embedded in a non-rotating corona, permeated by an external magnetic field initially aligned with the disc rotation axis. Intense accretion develops in the disc (Stone \\& Norman 1994), accompanied by a centrifugally driven outflow. The wind is eventually collimated by toroidal magnetic field produced in the corona by winding up the poloidal field. Bell (1994), Bell \\& Lucek (1995) and Lucek \\& Bell (1996, 1997) use two- and three-dimensional ideal MHD models, with a polytropic equation of state, to study the formation and stability of pressure driven jets collimated by poloidal magnetic field, which has a minimum inside the jet. The general structure of thermally driven disc winds was studied analytically by, e.g.,\\ Fukue (1989) and Takahara et al.\\ (1989). Another type of model was developed by Ustyugova et al.\\ (1995), Romanova et al.\\ (1997, 1998), Ouyed et al.\\ (1997) and Ouyed \\& Pudritz (1997a, 1997b, 1999) who consider ideal MHD in a polytropic corona permeated by an external poloidal magnetic field and subsume the physics of the accretion disc into boundary conditions at the base of the corona. The disc is assumed to be in Keplerian rotation (with any accretion neglected). The corona is non-rotating initially, and the system is driven by the injection of matter through the boundary that represents the disc surface. These models develop a steady (or at least statistically steady) state consistent with the analytical models by Blandford \\& Payne (1982), showing a magneto-centrifugal wind collimated by toroidal magnetic field, which again is produced in the corona by the vertical shear in the angular velocity. Three-dimensional simulations suggest that the resulting collimated outflow does not break due to the kink instability (Ouyed et al.\\ 2003; Thomsen \\& Nordlund 2003). It is not quite clear how strong the external magnetic field in accretion discs can be. Dragging of an external field from large radii in a viscous disc requires that magnetic diffusivity is much smaller than the kinetic one (Lubow et al.\\ 1994), i.e.\\ that the magnetic Prandtl number is significantly larger than unity which would be difficult to explain in a turbulent disc (Heyvaerts et al.\\ 1996). This argument neglects, however, the effect of magnetic torques which could produce significant field line dragging even when the magnetic Prandtl number is of order unity (Shalybkov \\& R\\\"udiger 2000). On the other hand, the efficiency of trapping an external magnetic field at initial stages of the disc formation is questionable because only a small fraction of the external flux can be retained if the density contrast between the disc and the surrounding medium is large. In that case the magnetic field will be strongly bent and reconnection will remove most of the flux from the disc. Furthermore, the ionization fraction is probably too small in the inner disc to ensure sufficient coupling between the gas and the magnetic field (Fromang et al.\\ 2002). For these reasons it seems appropriate to explore whether or not magnetic fields generated by a dynamo within the disc can produce an outflow with realistic properties. The models of Ustyugova et al.\\ (1995) and Ouyed \\& Pudritz (1997a, 1997b, 1999) have a distributed mass source and imposed poloidal velocity at the (unresolved) disc surface. As a result, their system develops a persistent outflow which is not just a transient. On the other hand, the models of Matsumoto et al.\\ (1996) resolve the disc, but have no mass replenishment to compensate losses via the outflow and accretion, and so the disc disappears with time. Our model is an attempt to combine advantages of both of these approaches and also to add dynamo action in the disc. Instead of a rigidly prescribed mass injection, we allow for self-regulatory replenishment of matter within a resolved disc. Instead of prescribing poloidal velocity at the disc surface, we resolve the disc and prescribe an entropy contrast between the disc and the corona, leaving more freedom for the system. Such an increase of entropy with height is only natural to expect for a disc surrounded by a hot corona, and we parameterize the coronal heating by a (fixed) entropy contrast. We further add self-sustained, intrinsic magnetic field to our system, as opposed to an external field used in the other models. Since our model goes beyond ideal MHD, the magnetic field must be maintained against decay. A simple form of mean-field dynamo action is included for this purpose. Like in the model of Ouyed \\& Pudritz (1997a, 1997b, 1999), the hot, pressure supported corona does not rotate initially. The disc is cool and is therefore centrifugally supported, so its rotation is nearly Keplerian. The corresponding steady-state solution is used as our initial condition. This solution is, however, unstable because of the vertical shear in the angular velocity between the disc and the corona (Urpin \\& Brandenburg 1998). Angular momentum transfer by viscous and magnetic stresses also leads to a departure from the initial state. As a result, a meridional flow develops, which exchanges matter between the corona and the disc surface layers. Mass losses through the disc surface and to the accreting central object are then replenished in the disc where we allow local mass production whenever and wherever the density decreases below a reference value. Matter is heated as it moves into the corona; this leads to an increase in pressure which drives the wind. Another efficient driver of the outflow in our model is magneto-centrifugal acceleration. A strong toroidal magnetic field produced by the dynamo in the disc is advected into the corona and contributes to confining the wind. Altogether, our model contains many features included in a range of earlier models. For example, the disc is essentially resistive to admit dynamo action, whereas magnetic diffusion turns out to be relatively unimportant in the corona, similarly to the models of, e.g.,\\ Wardle \\& K\\\"onigl (1993), Ferreira \\& Pelletier (1995) and Casse \\& Ferreira (2000a, 2000b). The synthetic nature of the outflow, driven by both pressure forces and magneto-centrifugally, is characteristic of the self-similar solutions of Ferreira \\& Pelletier (1995) and Casse \\& Ferreira (2000b). The latter authors also stress the r\\^ole of the vertical entropy gradient in enhancing the mass flux in the outflow and prescribe a vertical profile of the entropy production rate. Structured outflows outwardly similar to our results have been discussed by Krasnopolsky et al.\\ (1999) and Goodson et al.\\ (1997). However, in the former paper the structure in their solutions is due to the boundary conditions imposed on the disc surface, and in the latter model the structure is due to the interaction between the stellar magnetic field and the disc under the presence of differential rotation between star and disc. On the contrary, the structure in our model results from the dominance of different driving mechanisms in the inner and outer parts of the disc. While we had initially expected to find collimated outflows similar to those reported by Ouyed \\& Pudritz (1997a, 1997b), the outflow patterns we obtained turned out to be of quite different nature. However, a significant difference in our model is the resolved disc, which is necessary to model dynamo action in the disc, and the relatively small extension into the corona, where collimation is not yet expected. Figure~\\ref{Fig-structure} shows a sketch of the overall, multi-component structure of the outflows in the presence of magnetic fields and mass sink, which needs to be compared to the individual Figures in the rest of the paper. \\begin{figure} \\centering \\includegraphics[height=5cm]{fig1} \\caption{General structure of the outflows typically obtained in the magnetic runs with mass sink presented in this paper. The cool, dense disc emits (a) a hardly collimated, thermally driven wind (slow, hot, dense, magnetized, rotating), (b) a magneto-centrifugally driven outflow in a conical shell (faster, cooler, less dense, magnetized, quickly rotating), and (c) a thermally driven outflow near the axis (slow, hot, dense, weakly magnetized and weakly rotating). } \\label{Fig-structure} \\end{figure} \\bigskip The plan of the paper is as follows. We introduce our model in Sect.~\\ref{TM}, and then consider a range of parameters in Sect.~\\ref{RS} to clarify and illustrate the main physical effects. In Sect.~\\ref{FinalModel} we explore the parameter space, and our conclusions are presented in Sect.~\\ref{Discu}. ", "conclusions": "\\label{Discu} If the disc dynamo is sufficiently strong, our model develops a clearly structured outflow which is fast, cool and rarefied within a conical shell near the rotation axis where most of the angular momentum and magnetic energy is carried, and is slower, hotter and denser in the region around the axis as well as in the outer parts of the domain. The slower outflow is driven mostly by the entropy contrast between the disc and the corona, but the faster wind within the conical shell is mostly driven magneto-centrifugally. Without a central mass sink, the flow near the axis is faster, but otherwise the flow structure is similar to that with the sink. The half-opening angle of the cone with hot, dense gas around the axis is about $20^\\circ$--$30^\\circ$; this quantity somewhat changes with model parameters but remains close to that range. The outflow in our models does not show any signs of collimation. It should be noted, however, that not all outflows from protostellar discs are actually collimated, especially not at such small distance from the source. An example is the Becklin--Neugebauer/Kleinmann--Low (BN/KL) region in the Orion Nebula (Greenhill et al.\\ 1998), which has a conical outflow with a half-opening angle of $30^\\circ$ out to a distance of 25--60\\,AU from its origin. Therefore, collimation within a few AU (the size of our computational domain) is expected to be only weak. The region around the fast, cool and rarefied conical shell seen in Fig.~\\ref{FRun7} is similar to the flow structure reported by Krasnopolsky et al.\\ (1999); see their Fig.~1. In their model, however, the thin axial jet was caused by an explicit injection of matter from the inner parts of the disc which was treated as a boundary. In our reference model the fast outflow is sub-Alfv\\'enic because of the presence of a relatively strong poloidal field, whereas in Krasnopolsky et al.\\ (1999) the outflow becomes super-Alfv\\'enic at smaller heights. Outside the conical shell the outflow is mainly pressure driven, even though the criterion of Blandford \\& Payne (1982) is fulfilled. However, as Casse \\& Ferreira (2000b) pointed out, pressure driven outflows might dominate over centrifugally driven outflows if thermal effects are strong enough. In our model, matter is replenished in the resolved disc in a self-regulatory manner where and when needed. We believe that this is an improvement in comparison to the models of Ouyed \\& Pudritz (1997a, 1997b, 1999) and Ustyugova et al.\\ (1995), where mass inflow is prescribed as a boundary condition at the base of the corona. If we put $q_\\varrho^{\\rm disc}=0$ in Eqs~(\\ref{Cont}) and (\\ref{momentum}), the disc mass soon drops to low values and the outflow ceases. This is qualitatively the same behaviour as in the models of, e.g.,\\ Kudoh et al.\\ (1998). We should stress the importance of finite magnetic diffusivity in the disc: although poloidal velocity and poloidal magnetic field are well aligned in most of the corona, dynamo action in the disc is only possible in the presence of finite magnetic diffusivity, and the flow can enter the corona only by crossing magnetic field lines in the disc. An outflow occurs in the presence of both dipolar and quadrupolar type magnetic fields, even though fields with dipolar symmetry seem to be more efficient in magneto-centrifugal driving (cf.\\ von Rekowski et al.\\ 2000). The effects of the magnetic parity on the outflow structure deserves further analysis. The dynamo active accretion disc drives a significant outward Poynting flux in our model. Assuming that this applies equally to protostellar and AGN discs, this result could be important for understanding the origin of seed magnetic fields in galaxies and galaxy clusters; see Jafelice \\& Opher (1992) and Brandenburg (2000) for a discussion. We note, however, that the pressure of the intracluster gas may prevent the magnetized plasma from active galactic nuclei to spread over a significant volume (Goldshmidt \\& Rephaeli 1994). Our model can be improved in several respects. In many systems, both dynamo-generated and external magnetic fields may be present, so a more comprehensive model should include both. We used an $\\alpha^2\\Omega$ dynamo to parameterize magnetic field generation in the disc because we have restricted ourselves to axisymmetric models. As argued by Brandenburg (1998), dynamo action of turbulence which is driven by the magneto-rotational instability can be roughly described as an $\\alpha^2\\Omega$ dynamo. But this parameterization can be relaxed in three-dimensional simulations where one may expect that turbulence will be generated to drive dynamo action. Such simulations will be discussed elsewhere (von Rekowski et al.\\ 2002). Since our model includes angular momentum transport by both viscous and magnetic stresses, it is natural that the accreted matter is eventually diverted into an outflow near the axis; this is further facilitated by our prescribed entropy gradient at the disc surface. We believe that this picture is physically well motivated (Bell \\& Lucek 1995), with the only reservation that we do not incorporate the (more complicated) physics of coronal heating and disc cooling, but rather parameterize it with a fixed entropy contrast. We include a mass sink at the centre which could have prevented the outflow, and indeed the sink strongly affects nonmagnetized outflows. We have shown, however, that the magnetic field can efficiently shield the sink and thereby support a vigorous disc wind. The assumption of a prescribed entropy distribution is a useful tool to control the size of the disc and to parameterize the heating of the disc corona. However, it should be relaxed as soon as the disc physics can be described more fully. The energy equation, possibly with radiation transfer, should be included. This would lead to a self-consistent entropy distribution and would admit the deposition of viscous and Ohmic heat in the outflow. In the simulations by von Rekowski et al.\\ (2002), entropy is evolved. We believe that a mass source is a necessary feature of any model of this kind if one wishes to obtain a steady state. In the present paper the mass source is distributed throughout the whole disc to represent replenishment of matter from the midplane of the disc. Alternatively, a mass source could be located near to or on the domain boundary." }, "0003/astro-ph0003204_arXiv.txt": { "abstract": "We derive structural parameters and evidence for extended tidal debris from star count and preliminary standard candle analyses of the Large Magellanic Cloud based on Two Micron All Sky Survey (2MASS) data. The full-sky coverage and low extinction in $K_s$ presents an ideal sample for structural analysis of the LMC. The star count surface densities and deprojected inclination for both young and older populations are consistent with previous work. We use the full areal coverage and large LMC diameter to Galactrocentric distance ratio to infer the same value for the disk inclination based on perspective. A standard candle analysis based on a sample of carbon long-period variables (LPV) in a narrow color range, $1.6 1$ (\\cite{pbw96}), it is likely that $p_{\\rm ram} > p_{\\rm gas}$, and that the smaller of the two sizes we have just estimated, corresponding to a bow-shock morphology, should apply. We note that in such a case, it is still reasonable to model the PWN as a circular disk, since in observed bow-shock nebulae most of the emission is concentrated close to the head of the nebula. In any case, regardless of the dominant source of confining pressure, the expected extent of a PWN powered by \\rxj\\ is small. Although we believe the sizes derived above to be robust, we conservatively adopt a maximum angular size for any PWN of 30~arcsec to take into account possible uncertainties in $V$, $n$, $T$, or the distance to the source. From Fig~\\ref{fig_sens}, it can be seen that at all three frequencies, the upper limit on the flux density for such a source is $\\sim$7~mJy. Assuming a typical PWN spectral index of $\\alpha = -0.3$ ($S_\\nu \\propto \\nu^{\\alpha}$), an upper limit of 7~mJy at 1.4~GHz corresponds to a broad-band radio luminosity (integrated between 10~MHz and 100~GHz) of $L_R = 2 \\times 10^{30}$~erg~s$^{-1}$. Defining $\\epsilon \\equiv L_R/\\dot{E}$ to be the ratio between a PWN's broad-band radio luminosity and its spin-down luminosity, we find that for $\\dot{E} = 1.4\\times 10^{37}$~erg~s$^{-1}$ as reported by PZT99 and ZTP99, we can derive an upper limit $\\epsilon < 10^{-7}$. This is a more stringent limit on $\\epsilon$ than has been derived for almost any other pulsar (cf.\\ \\cite{fs97}; \\cite{gsf+00}). In particular, this upper limit is sharply at odds with the values of $\\epsilon$ seen for other young ($\\la 20$~kyr) pulsars, almost all of which produce radio PWN or have upper limits consistent with $\\epsilon \\ge 10^{-4}$ (\\cite{fs97}; \\cite{gsf+00}). The glaring exception to this is PSR~B1509--58 in the SNR~G320.4--1.2 (MSH~15--5{\\em 2}), which powers an X-ray PWN but for which no radio PWN has yet been detected (\\cite{gbm+98}). However, this can be understood in terms of the low ambient density ($n < 0.01$~cm$^{-3}$), which results in severe adiabatic losses and a consequently underluminous radio PWN (\\cite{bha90}). This condition is not satisfied for \\rxj, and so cannot be considered as a possible explanation for the non-detection of a PWN.\\footnote{Furthermore, PSR~B1509--58 powers a bright X-ray PWN (e.g.\\ \\cite{shss84}; \\cite{bb97}), while no X-ray nebula is seen around \\rxj\\ (\\cite{psz00}).} We thus find that any PWN in \\pup\\ has a radio luminosity three orders of magnitude fainter than expected for the spin parameters derived by PZT99 and ZTP99. Nevertheless, if we assume, as Brazier \\& Johnston (1999\\nocite{bj98}) have argued, that \\rxj\\ is a rotation-powered pulsar, what spin parameters can we infer for it? If we require that $\\epsilon \\ge 10^{-4}$ as seen for other young pulsars, the maximum value of $\\dot{E} \\equiv 4\\pi^2 I \\dot{P}/P^3$ which is consistent with our non-detection of a radio PWN is $\\sim10^{33}$~erg~s$^{-1}$. Meanwhile, it is unlikely that the characteristic age, $\\tau \\equiv P/2\\dot{P}$, of the pulsar is more than 50~kyr, $\\sim$10 times the true age of the system. These upper limits on $\\dot{E}$ and $\\tau$ correspond to lower limits $P>3.5$~s, $\\dot{P} > 1.1 \\times 10^{-12}$~s~s$^{-1}$ and $B>6.4 \\times 10^{13}$~G, parameters which are very similar to those seen for the SGRs/AXPs (\\cite{kcs99}) and for the young radio pulsar PSR~J1814--1744 (\\cite{pkc00}; \\cite{ckl+00}), but quite different than those of other young pulsars in SNRs, for which typically $P<0.2$~s, $\\dot{E} > 10^{36}$~erg~s$^{-1}$ and $B\\approx 10^{12}$~G. Whether \\rxj\\ indeed has a long initial period and high magnetic field, or has some other properties such that it does not produce a detectable radio nebula or radio pulsations, the lack of a PWN around this source (and around other RQNS such as 1E~1207.4--5209 in the SNR~G296.5+10.0; \\cite{mbc96}; \\cite{gdg+00}), argues that at least some RQNS have drastically different properties from young radio pulsars. Brazier \\& Johnston (1999\\nocite{bj98}) list six RQNS which are younger than 20~kyr and nearer than 3.5~kpc. Excluding two RQNS from their list which do power PWN and thus may well be radio pulsars beaming away from us, but including the recently-discovered RQNS in the young and nearby SNR~Cassiopeia~A (\\cite{tan99}; \\cite{pza+00}; \\cite{cph+00}), this implies a Galactic birth-rate for such sources of at least once every $\\sim$200 years, comparable to or even in excess of the birth-rate for radio pulsars (e.g.\\ \\cite{lml+98}). Radio-quiet neutron stars thus point to the possibility that pulsars like the Crab are not the most common manifestation of neutron star." }, "0003/astro-ph0003081_arXiv.txt": { "abstract": "The Dominion Radio Astrophysical Observatory's Synthesis Telescope provides the highest resolution data (1 arcmin and 0.82 \\kms) to date of an H~I worm candidate. Observed as part of the Canadian Galactic Plane Survey, mushroom-shaped GW 123.4-1.5 extends only a few hundred parsecs, contains $\\sim10^5 \\Mo$ of neutral hydrogen, and appears unrelated to a conventional shell or chimney structure. Our preliminary Zeus-2d models use a single off-plane explosion with a modest ($\\sim 10^{51}$ ergs) energy input. These generic simulations generate, interior to an expanding outer blast wave, a buoyant cloud whose structure resembles the morphology of the observed feature. Unlike typical model superbubbles, the stem can be narrow because its width is not governed by the pressure behind the blast wave nor the disk scale height. Using this type of approach it should be possible to more accurately model the thin stem and other details of GW 123.4-1.5 in the future. Subject Headings: ISM: bubbles --- ISM: individual (GW 123.4-1.5) --- ISM: structure --- Galaxy: halo --- Galaxy: structure ", "introduction": "\\label{intro} \\nocite{heil84}Heiles (1984) has identified atomic hydrogen (HI) gas filaments ``crawling'' away from the plane of the inner Galaxy. These so-called ``worms'' were proposed to be parts of larger HI shells blown by the energetic stellar winds or supernovae in stellar associations. Such open, or blown out, shells would serve as conduits for hot gas and radiation to escape into the galactic halo, as recently observed in the superbubble/chimney reported by \\nocite{NorTayDew96} Normandeau et~al.~1996. Koo et al. (1992) later produced a catalogue of 118 Galactic worm candidates, defining as a worm any dusty, H~I structure perpendicular to the Galactic Plane. The Canadian Galactic Plane Survey (CGPS) is currently mapping a 70 degree longitude segment of the northern Galaxy at high resolution in HI (\\nocite{taylor99} Taylor et al. 1999). One Galactic worm candidate, GW 123.4-1.5, is within the early regions surveyed. These observations reveal that GW 123-1.5 is an unusual, mushroom-shaped cloud, hundreds of pc in size, apparently unrelated to a conventional shell or chimney structure. In this paper we present the observations, derive the observed properties of the mushroom cloud, and discuss possible scenarios for its origin. ", "conclusions": "\\label{cfmodel} The mushroom cloud shape and mass distribution of GW123.4-1.5 pose a challenge to conventional superbubble scenarios. In these models (e.g., see \\nocite{macnor89} Mac Low, McCray, \\& Norman 1989; \\nocite{GTTrozbod90} Tenorio-Tagle, Rozyczka, \\& Bodenheimer 1990) the lower part (the stem) retains the bulk of the mass, even though the upper part of the bubble (the cap) may expand to a large radius. This is not the case with GW123.4-1.5, where we estimate that the cap contains about four times the mass in the stem. Additionally the greatest cap to stem width ratio in the superbubble models is about 3:1 \\nocite{GTTrozbod90} (Tenorio-Tagle, Rozyczka, \\& Bodenheimer 1990) while the mushroom cloud's ratio is 6:1. Furthermore, the radius of the {\\em model} stem is typically equal to $2 H$, where $H$ is the exponential scale length of the local Gaussian density distribution; the stem is the cavity created by a blowout from a stratified atmosphere into a uniform low-density halo (at $\\sim$500 pc). If the local scale length were equal to the global average of the HI disk $H \\simeq 135$ pc (\\nocite{1990ARA&A..28..215DiLoc} Dickey \\& Lockman~1990), then the width of the mushroom's stem would be 500 pc rather than the $\\sim$ 40 pc observed. The above discrepancies lead us to consider alternate scenarios in which a stem plus cap morphology can be realized. For example, an H~I jet could be ejected from the disk and a wide lobe could be created where it has stalled, possibly falling back to the disk in a fountain-like manner. One possible origin for such an event is the passage of a High Velocity Cloud through the galactic disk and the subsequent emergence of gas on the other side (\\nocite{GTTfrabodroz87}Tenorio-Tagle et~al. 1987). However, here we consider the rise of buoyant gas. A rising fireball after a terrestrial nuclear explosion creates a structure strikingly similar to that of GW123.4-1.5. In the interstellar context, mushroom-shaped clouds can arise from the interaction between different components in a multiphase medium (e.g. \\nocite{RosBreg95} Rosen \\& Bregman 1995, \\nocite{avil99} Avillez 1999), including a cloud-cloud collision (Miniati et al. 1997). For our model, we focus on the rise of buoyant hot gas resulting from a single supernova event. Jones (1973) has investigated the early evolution of a supernova remnant and found signs of buoyant rise. Here, we follow to late times the evolution of a remnant which does not have enough energy to blow out of the disk atmosphere. The initially pressure driven hot bubble stalls at radius smaller than the scale height of the medium\\footnote{The outer shock front continues to propagate up through the stratified atmosphere, but has little dynamic importance since the inner hot bubble has stalled and is no longer compressing matter into a thin shell behind the shock front.}. Buoyancy forces lift the low density bubble out of the galactic plane and through the stratified atmosphere. In analogy with the nuclear fireball, the quasi-vacuum produced under the rising cap may pull in surrounding material, entraining it to form the stem. In this case, the stem width is not determined by the scale height of the medium and can be quite narrow. However the existence of cold neutral material in the cap must still be explained since the rise of a buoyant plume in pressure equilibrium does not physically move much material from the galactic disk to high latitudes. To examine the plausibility of the buoyant bubble scenario we have begun modeling a single 10$^{51}$ ergs explosion at 60 pc above the midplane using a modified Zeus 2-d code (e.g. \\nocite{StoNor92} Stone \\& Norman~1992). Our simulations include the effects of radiative cooling, heat conduction, and the vertical gravitational field. We also use artificially low quiescent gas density to reduce the influence of ``numerical diffusion\", a computational artifact resulting from steep gradients. For example, we use a scale height 60 pc\\footnote{The stalling radius is fixed for a given energy input and ambient density in the plane. So for any given source the buoyancy condition, that the stalling radius is less than $H$, is more likely to exist if the z-distribution of the ambient gas is $H \\simeq 135$ pc .} and a midplane number density of $\\rm n_o = 1 \\ atom \\ cm^{-3}$ to create an ambient medium distribution which allows bubbles to form and evolve. Our preliminary low-resolution models are intentionally generic and detailed comparisons with GW123.4-1.5 will be left to a later paper. In both Gaussian and exponential atmospheres, a bubble forms interior to the blastwave and, rather than elongating vertically as in a conventional superbubble, the bubble rises buoyantly. Since the rise is supersonic relative to the cooled gas above (interior to the blastwave), the bubble accumulates cold ambient gas in a snow-plow mode along its shock front. Gas also flows upwards in a column following the bubble, contributing to its evolution into a mushroom-shaped cloud. By 8 Myr the interior bubble has the kinetic temperature distribution shown in Fig.~\\ref{figmodel}a. The models show a cold gas stem which is narrower than the primary bubble (stem) in conventional superbubble scenarios in which the stem width is dictated by the gas scale height. Our modeled stem is expected to become narrower and more obvious than shown here when a higher ambient density is used and cooling becomes more efficient. The coolest temperatures trace a curlyqued cap structure like that observed in GW 123.5-1.5. These structures are also evident in the column density map, plotted for gas $<$7500 K, in Fig.~\\ref{figmodel}b. Similar to some models of supernova evolution (\\nocite{SlaCox92}Slavin \\& Cox~1992) the hottest gas in our simulation decreases to fill a small volume which eventually collapses. Residing just inside the skin of the lobes are remnants of a warm gas envelope, formed by heat conduction, which had been pushed away from the bubble by the upward flow of cold, dense gas from the stem. Their temperature of a few $\\times 10^4$ K is consistent with the lack of soft x-ray emission. The cool, H~I by-product of the buoyant bubble should be observable for substantially longer than the cooling time ($\\sim 2 \\times 10^5$~yr) of this hot gas. We have carried out a preliminary search in other wavebands for evidence of a possible energy source, but no obvious candidates can be identified. For example, there are no IRAS sources near the base with CO emission in the FCRAO database (between the midplane and - 3$^o$ galactic latitude) within the FWHM velocity range of the stem. The projected position of the H~II emitting reflection nebula Sharpless~185 lies near the base of the stem. However, it is associated with the Be X-ray emitting star $\\gamma$ Cass at a distance of only about two hundred pc (see \\nocite{bloudew97}Blouin et~al.~1997)." }, "0003/astro-ph0003048_arXiv.txt": { "abstract": "We present new radio and optical observations of the colliding-wind system WR\\,146 aimed at understanding the nature of the companion to the Wolf-Rayet star and the collision of their winds. The radio observations reveal emission from three components: the WR stellar wind, the non-thermal wind-wind interaction region and, for the first time, the stellar wind of the OB companion. This provides the unique possibility of determining the mass-loss rate and terminal wind velocity ratios of the two winds, independent of distance. Respectively, these ratios are determined to be $0.20\\pm0.06$ and $0.56\\pm0.17$ for the OB-companion star relative to the WR star. A new optical spectrum indicates that the system is more luminous than had been believed previously. We deduce that the ``companion'' cannot be a single, low luminosity O8 star as previously suggested, but is either a high luminosity O8 star, or possibly an O8+WC binary system. ", "introduction": "Wolf-Rayet stars are surrounded by dense stellar winds giving rise to free-free emission extending from IR to radio wavelengths. Typically, this emission is characterised by a power-law spectrum of the form $S_\\nu\\propto \\nu^{\\alpha}$, with values of the spectral index $\\alpha \\sim +0.7-+0.8$, and radio brightness temperatures $\\sim10^4$~K. A small number of WR stars have radio emission that exhibits quite different properties: negative spectral indices and brightness temperatures $\\sim10^6$~K or higher, properties that are characteristic of non-thermal emission. WR\\thinspace146 is a member of this group, which includes WR\\thinspace125, WR\\thinspace140 and WR\\thinspace147. The radio emission from WR\\thinspace146 was first resolved in high resolution observations with MERLIN (Dougherty et al. 1996, hereafter Paper I). These 5-GHz observations revealed two components, N$_5$ and S$_5$, separated by $\\sim120$ milli-arcseconds (mas). The flux of S$_5$ was consistent with that estimated from extrapolation of the IR-millimetre spectrum arising from the free-free emitting envelope around the WR star. The brightness temperature of N$_5$ ($\\sim10^6$K) identified the nature of the emission from this component as non-thermal. An optical spectrum showed evidence for absorption lines at H$\\delta$ and H$\\gamma$, which we attributed to an early-type companion to the WR star. This led us to hypothesize in Paper I that the non-thermal emission arose from a population of relativistic electrons, accelerated in a wind-wind collision region where the wind of the WR star and the companion interacted (e.g. Eichler \\& Usov 1993). To be consistent with such a model, we suggested the companion lay at the same position angle as N$_5$ from S$_5$, but slightly further away from the WR star. The presence of a companion was confirmed in optical imaging with the Hubble Space Telescope (HST) by Niemela et al. (1998). They observed two stars, WR146A and B (hereafter S$_{\\rm O}$ and N$_{\\rm O}$ respectively), at the same position angle as the radio sources but separated by $\\sim 168$ mas. Under the assumption that the southern sources in both the HST and MERLIN images are coincident, these observations place the non-thermal source between the two stellar images, strongly supporting wind-wind collision as the origin of the non-thermal emission. Its position relative to the two stellar components ($\\sim 120$ mas from S$_O$ and $\\sim$ 48 mas from N$_O$) is where the dynamical pressure of the two stellar winds is balanced. This indicates that the momentum of S$_O$'s stellar wind is $\\sim$ 0.1 times that of N$_O$. With the wind velocity of the WC star in WR\\,146 ($\\sim$ 2900 km s$^{-1}$, Eenens \\& Williams 1994) being greater than that of a typical OB star, and the expectation that the mass-loss rate of a WR star would be greater than that of an OB star, this strongly supports the identification of S$_O$ with the WC6 star in Paper I and identifies N$_O$ with an OB companion having a lesser wind momentum. The momentum ratio lead Niemela et al. to infer that the wind momentum of the companion was more appropriate to a star of early O or Of type than a late O-type main-sequence star. Taking account of the photometry, they suggested an O6-O5 V-III spectral type for the companion. The photometry by Niemela et al. showed the two stars to be equally bright in B but that S$_{\\rm O}$ was redder than N$_{\\rm O}$ in (B--V) and (U--B). The (near-zero) magnitude difference between the two blue images falls between two very different estimates of the WR:O light ratio deduced from spectra of WR\\,146 which included both stars. On the one hand, in Paper I, we measured equivalent widths of H$\\delta$ and H$\\gamma$ in our blue spectrum of WR\\,146 to be $W_{\\lambda} \\sim $0.9\\AA. Comparison of these with those ($\\sim 2.5$\\AA) typical of mid-to-late O stars in the Walborn \\& Fitzpatrick (1990) atlas indicated that the O-star spectrum was diluted and suggested that the continuum of the WC star was twice as bright as that of the O companion in the blue. On the other hand, Willis et al. (1997) used a spectrum of lower resolution but longer wavelength coverage than presented in Paper I to determine the WR:O light ratio from comparison of the equivalent widths of the emission lines with those of two single WC6 stars. They found that the O star was {\\em brighter} than the WC star, reporting a continuum light ratio (WR:O) of 1:(2$\\pm$1). The discrepancy in the light ratios of the WC6 and O components derived from the dilution of O star absorption lines and WC6 star emission lines is significant and probably too great to arise purely from the observational uncertainties. Leaving aside the question of whether one of the stars varied between the two observations, the principal uncertainties in two spectroscopic light-ratio determinations come from whether the intrinsic line strengths of the components are indeed equal to those of the comparisons adopted. The discrepancy could be reduced if the equivalent widths of H$\\delta$ and H$\\gamma$ in the O star were weaker than the 2.5\\AA\\ adopted or if the emission lines in the WC6 component were weaker than those of other WC6 stars. Both possibilities are plausible. The strengths of the emission lines in WR\\,146 may well be atypical for its WC6 type: both Eenens \\& Williams (1992) and Willis et al. found anomalously low C/He abundances for this star. On the other hand, the strengths of H$\\gamma$ and H$\\delta$ in O-type stars do depend on spectral type and luminosity class, which are not known directly for the companion in WR\\thinspace146. We note that the ``O8.5V'' adopted by Willis et al. comes not from the star's spectral lines but from its luminosity inferred from that of the WC6 star and the continuum light ratio. We therefore re-observed the blue spectrum of WR\\thinspace146 with the William Herschel Telescope in an attempt to determine the spectral type and, hopefully, luminosity of the early-type companion to the WR star. These data, together with the photometry by Niemela et al., will give us a better idea of the intrinsic properties of the companion for comparison with the radio observations. We also re-observed WR\\thinspace146 at radio wavelengths, extending the frequency coverage of Paper I, to search for the wind flux from the companion, and study the characteristics of the two previously observed radio components with the aim of furthering our understanding of the colliding wind phenomenon. \\begin{figure} \\vspace{8.0cm} \\special{psfile=figure1.ps hoffset=-15 voffset=-46 hscale=40 vscale=40 angle=0} \\caption[]{ a) 5-GHz MERLIN observations obtained on 1992 December 26 The $1\\sigma$ rms uncertainty is $250\\mu$Jy, with contour levels at $-3,3,6,9,12,24,48\\sigma$. The synthesized beam is $53\\times53$~mas$^2$.} \\label{fig:MERLIN} \\end{figure} ", "conclusions": "From high-resolution observations with the VLA at 22 GHz and with MERLIN at 5 GHz we have observed all three components of the WR\\,146 system: the OB and WC6 stellar winds and the non-thermal source where they collide. The source geometry and ratio of stellar wind fluxes allow us to determine the ratios of mass-loss rates and wind velocities independent of distance to the system. From these ratios and the observations of the WC6 star, we derive the wind velocity of the OB star to be $1600\\pm480$ km\\,s$^{-1}$ and its mass-loss rate to be one quarter that of the WC6 star. If the WC6 star has an ``average'' WR-star mass-loss rate of $\\sim 4\\times10^{-5}$~M$_\\odot$\\,y$^{-1}$, that of the OB star would be $\\sim 8\\times 10^{-6}$~M$_\\odot$\\,y$^{-1}$, suggesting a very high luminosity object. Support for high luminosity comes from the optical spectrum of WR\\,146, which includes both stars. This shows absorption-line ratios formed in the OB companion suggesting it to be a high-luminosity O8 star. If it is a single star, the inferred luminosity places WR\\,146 at the distance of the Cyg OB2 association. This gives an anomalously high luminosity for the WC6 star but, given its other anomalies, does not rule out this distance. However, the mass-loss rates determined using this distance ($1.3 \\times 10^{-4}$ M$_\\odot $y$^{-1}$ for the WC6 star and $2.6 \\times 10^{-5}$ M$_\\odot $y$^{-1}$ for the O8 star) are awkwardly high and the latter is probably inconsistent with the spectroscopy. Many, if not all, of the observations could be explained if the companion was itself a binary comprising an O8 star and another WC star. This needs to be tested by separate spectra of the two visual components of WR\\thinspace146. The presence of an unresolved companion to the O8 star could modulate its wind so as to cause the 3.38-y variability in the non-thermal emission reported by Setia Gunawan et al. (2000). We may not have reached a firm conclusion as to the nature of the stellar companion(s) to the WC star from the new radio and optical observations. However, the ability to study all three radio components of the WR\\thinspace146 system separately will ensure that this system becomes an archetype for studying the wind-collision phenomenon." }, "0003/astro-ph0003338_arXiv.txt": { "abstract": "{ We present estimates of the gravitational lensing shear variance obtained from images taken at the CFHT using the UH8K CCD mosaic camera. Six fields were observed for a total of 1 hour each in V and I, resulting in catalogs containing $\\sim 20,000$ galaxies per field, with properly calibrated and optimally weighted shear estimates. These were averaged in cells of sizes ranging from $1'.875$ to $30'$ to obtain estimates of the cosmic shear variance $\\langle \\overline{\\bgamma}^2\\rangle$, with uncertainty estimated from the scatter among the estimates for the 6 fields. Our most reliable estimator for cosmic shear is provided by the cross-correlation of the shear measured in the two passbands. At scales $\\lsim 10'$ the results are in good agreement with those of \\citeN{vwme+00}, \\citeN{bre00} and \\citeN{wtk+00} and with currently fashionable cosmological models. At larger scales the shear variance falls below the theoretical predictions, and on the largest scales we find a null detection of shear variance averaged in $30'$ cells of $\\langle \\overline{\\bgamma}^2 \\rangle = (0.28 \\pm 1.84) \\times 10^{-5}$. } ", "introduction": "Weak lensing provides a potentially powerful probe of mass fluctuations in the Universe (\\citeNP{gunn67}; \\citeNP{mellier99} and references therein). Three independent groups have recently presented estimates of the shear variance from deep `blank-field' CCD imaging surveys. \\citeNP{vwme+00} (hereafter vWME+) measured the shear variance in circular cells of radii ranging from $0'.7$ to $3'.5$; \\citeNP{bre00} (hereafter BRE) measured the shear variance in square cells of side $8'.0$ and \\citeNP{wtk+00} (hereafter WTK+) have provided estimates of the shear-shear correlation function at separations $3'.25$, $8'.5$ and $22'.0$. Here we present shear variance measurements from $\\simeq 1.5$ square degrees of deep photometry obtained as part of our ongoing weak lensing survey. We find results which are broadly in good agreement with the recently published estimates. ", "conclusions": "For an effective background galaxy redshift of $z_{\\rm eff} \\simeq 1.0$ these measurements probe mass fluctuations in a shell peaked at $z \\simeq 0.4$. At this redshift the $30'$ field size corresponds to a comoving distance of about $6 h^{-1}$Mpc, so the cell variances presented here probe scales in the range $0.4-6 h^{-1}$Mpc. On the smaller end of this scale we find very good agreement with recently announced estimates from other groups, and also with canonical cosmological theory predictions. It is hard to definitively rule out the possibility that the small angle measurements are inflated by systematic errors, but one can safely rule out theories such as light-traces mass high density models which predict shear variance a factor $\\sim 5$ higher than our results. On larger scales our measurements are extremely precise, yet we find only a null detection for our largest cells. These results show that on large scales the rms shear is at most a fraction of a percent. The apparent discrepancy between these results and the theoretical predictions is quite interesting, and suggests a steepening of the mass correlation function at scales $\\sim 1-2 h^{-1}$Mpc. More data are needed however to definitively confirm this." }, "0003/astro-ph0003424_arXiv.txt": { "abstract": "The eclipsing novalike cataclysmic variable star V348 Pup exhibits a persistent luminosity modulation with a period 6 per cent longer than its 2.44 hour orbital-period ($P_{\\rmn{orb}}$). This has been interpreted as a ``positive superhump'' resulting from a slowly precessing non-axisymmetric accretion disc gravitationally interacting with the secondary. We find a clear modulation of mid-eclipse times on the superhump period, which agrees well with the predictions of a simple precessing eccentric disc model. Our modelling shows that the disc light centre is on the far side of the disc from the donor star when the superhump reaches maximum light. This phasing suggests a link between superhumps in V348 Pup and late superhumps in SU UMa systems. Modelling of the full lightcurve and maximum entropy eclipse mapping both show that the disc emission is concentrated closer to the white dwarf at superhump maximum than at superhump minimum. We detect additional signals consistent with the beat periods between the implied disc precession period and both $\\frac{1}{2}P_{\\rmn{orb}}$ and $\\frac{1}{3}P_{\\rmn{orb}}$. ", "introduction": "V348 Pup (1H 0709--360, Pup 1) is a novalike cataclysmic variable (CV): a system with a high mass transfer rate which maintains its accretion disc in the hot, ionized, high viscosity state reached by dwarf novae in outburst. It exhibits deep eclipses in its optical and infrared lightcurves (Tuohy et al. 1990): it is a high inclination system with orbital period $P_{\\rmn{orb}}=2.44$ hours (Baptista et al. 1996). \\subsection{Superhumps} Modulations in luminosity with a period a few per cent longer than the orbital period have been observed in many short period CVs (see reviews in Molnar \\& Kobulnicky 1992, Warner 1995, Patterson 1998a). These modulations typically take the form of a distinct increase in luminosity, or superhump. The standard explanation of this phenonemon is that the system contains an eccentric precessing accretion disc. If the accretion disc extends out far enough, the outermost orbits of disc matter can resonate with the tidal influence of the secondary star as it orbits the system. A 3:1 resonance can occur which results in the disc becoming distorted to form an eccentric non-axisymmetric shape. The tidal forces acting on this eccentric disc will cause it to precess slowly in a prograde direction. The superhump period, $P_{\\rmn{sh}}$, is then the beat period between the disc precession period, $P_{\\rmn{prec}}$, and the orbital period, $P_{\\rmn{orb}}$ (Osaki 1996): \\[ \\frac{1}{P_{\\rmn{sh}}} = \\frac{1}{P_{\\rmn{orb}}} - \\frac{1}{P_{\\rmn{prec}}}. \\] $P_{\\rmn{sh}}$ is the period on which the relative orientation of the line of centres of the two stars and the eccentric disc repeats. Possible models for the light modulation on $P_{\\rmn{sh}}$ are described below. This paper considers these models in relation to our observations. In the tidal model the superhump is a result of tidal stresses acting on the precessing eccentric disc (Whitehurst 1998b). The light may be due to a perturbation of the velocity field in the outer disc, leading to azimuthal velocity gradients and crossing or converging particle trajectories. Thus extra dissipation modulated on the superhump period arises when the secondary sweeps past the eccentric disc. In addition, the superhump-modulated tidal stress would lead to a superhump-modulated angular momentum loss from the disc which would facilitate a variation in the mass transfer rate through the disc, and hence a modulation in disc luminosity. The bright spot model arises from noting that the energy gained by material in the accretion stream will depend on how far it falls before impacting on the disc (Vogt 1981). The energy dissipated at impact will be modulated on the superhump period since the non-axisymmetric disc radius causes a stream-disc impact region at varying depths in the white dwarf potential well. Recent SPH simulations of accretions discs in AM CVn stars lead to a third, more realistic, model in which the disc shape changes from nearly circular to highly eccentric over the course of a superhump period (Simpson \\& Wood 1998). Superhumps arise from viscous energy production as the distorting disc is tidally stressed. Other SPH simulations (e.g. Murray 1996, 1998) also reveal a disc whose shape changes, with Murray (1996) predicting superhump modulations from both the periodic compression of the eccentric disc and the varying depth in the primary Roche potential at which the stream impacts the disc. Dwarf novae in super-outburst exhibit two distinct positive superhump phenomena (Vogt 1983, Schoembs 1986). Normal superhumps appear early in the super-outburst and fade away towards the end of the outburst plateau to be replaced with `late' superhumps which persist into quiescence. These late superhumps are roughly anti-phased with the normal superhumps, and are more likely to be analogous to the persistent superhumps seen in novalikes (Patterson 1998b), where the system has had sufficient time to settle into a steady state. Our extensive photometry (Section \\ref{observations}) reveals similarities between superhumps in V348 Pup and late superhumps in dwarf novae. In Section \\ref{periodsearch} we present power spectra revealing the superhump period and additional signals close to orbital period harmonics. In Section \\ref{qi} we estimate the orbital parameters, $q$ and $i$, for V348 Pup using the average orbital lightcurve and the superhump period. The waveform of the superhump modulation is discussed in Section \\ref{superhumpmod}. Section \\ref{averages} considers average orbital lightcurves grouped according to superhump phase. In Section \\ref{ecpar} we fit our lightcurves with a precessing eccentric disc model, hence deducing the location of light centre of the disc. We consider the results of maximum entropy eclipse mapping in section \\ref{eclipsemapping}. Our results and their implications are discussed in Section \\ref{discussion}. ", "conclusions": "\\label{discussion} The phase of the superhump relative to the conjunction of the line of centres of the system and the semi-major axis of the disc should make it possible to determine whether the bright spot model or the tidal heating model better explains the source of the superhump. The simplest tidal model predicts that the superhump light should peak when (or slightly after) the largest radius part of the disc coincides with the line of centres. This is because the tidal interaction is strongly dependent on distance from the secondary, and so will be most significant in regions where the disc extends out close to the L1 point. However, if the bright spot model is to be believed, then the the superhump light source will be brightest when the accreting material has the furthest to fall. In other words, the superhump should occur when the stream impacts on the disc at its smallest radius. \\begin{figure} \\psfig{file=phasing.ps,width=8.4cm,angle=270} \\caption{The first row of Figure \\ref{eclipsestuff} shows the eclipses to be earliest at superhump phase $\\phi_{\\rm{sh}}\\sim$ 0.75, which implies orientation (a), from which we deduce the relative phasing of disc and secondary star at superhump maximum ($\\phi_{\\rm{sh}}=0$) shown in (b).} \\label{phasing} \\end{figure} The mid-eclipse times shown in the top row of panels in Figure \\ref{eclipsestuff} show the eclipses to be earliest around superhump phase 0.75 in all cases. Assuming that the centre of light of the eccentric disc is offset from the white dwarf in the direction of the largest radius, we can deduce the disc orientation during these eclipses to be as shown in Figure \\ref{phasing}a. A quarter of a superhump period later, the orientation of the disc has barely changed, the secondary will be lined up with the smallest radius part of the disc and the superhump phase will be 0.0 (Figure \\ref{phasing}b). Therefore superhump maximum occurs when the secondary star is lined up with the smallest radius part of the disc. The values of $\\phi_{disc}$ in Table \\ref{fits} agree with this deduction. This phasing is consistent with the bright spot model for the superhump emission but is inconsistent with the simple tidal heating model. In Section \\ref{averages}, we noted that the superhump light appears not to be emitted isotropically: the superhump is strongest when it occurs around orbital phase 0. This is easily explained if the major contribution to superhump light is the bright spot: the bright spot is most visible when it is on the nearside of the accretion disc. Schoembs (1986) observed late superhumps in the eclipsing SU UMa dwarf nova OY Car, also a high inclination system. When a superhump was coincident with a pre-eclipse orbital hump, the combined amplitude was greater than that predicted for a linear superposition of the individual amplitudes i.e. OY Car's late superhumps were strongest around orbital phase 0. However, van der Woerd et al. (1988) studied the dwarf nova VW Hyi, concluding that there was no correlation between the orbital phase and amplitude of late superhumps. Since VW Hyi has an intermediate inclination, $\\sim60^\\circ$ (Schoembs \\& Vogt 1981), the bright spot visibility need not vary with phase, so if the bright spot is the main superhump light source we expect no variation in superhump amplitude with orbital phase. Krzeminski \\& Vogt (1985) studied OY Car during a super-outburst and through variations in the O-C eclipse timings deduced the presence of an eccentric disc with phasing similar to that in V348 Pup. Krzeminski \\& Vogt's definition of O-C time was criticized by Naylor et al. (1987), with Naylor et al. concluding that the O-C evidence was weaker than originally thought. Schoembs (1986) followed OY Car from early in a super-outburst almost until the return to quiescence, observing the $\\sim 180^{\\circ}$ phase change from normal superhumps around the height of the outburst to late superhumps during the decline of the super-outburst. Patterson et al. (1995) observed the same change in superhump phase late in a super-outburst of V1159 Ori. Hessman et al. (1992) studied OY Car at the end of a super-outburst. By looking at the varying hot spot eclipse ingress times, and considering the trajectory of the accretion stream, they concluded that the disc was eccentric. The orientation of the disc at superhump maximum was very similar to that which we find in V348 Pup. The broad waveform of these late superhumps in OY Car (Hessman et al.) was also similar to the superhump modulation in V348 Pup. Such similarity between late superhumps in OY Car and the superhumps in V348 Pup is not surprising. Late superhumps in dwarf novae appear towards the end of the superoutburst, after the disc has had time to adjust to its high state. V348 Pup is persistently in a high state. Superhumps in a novalike system might resemble those to which the superhumps in a superoutbursting dwarf nova would tend if it remained in superoutburst for a long time. It seems likely that the mechanism responsible for late superhumps in SU UMa systems is the same mechanism responsible for superhumps in V348 Pup. However, Skillman et al. (1998) observed strong superhumps in the nova-like TT Ari throughout 1997 whose waveform is triangular like those of normal superhumps in dwarf novae. There are many other studies of the disc structure in SU UMa stars during superoutburst. Vogt (1981) and Honey et al. (1988) found evidence for an eccentric precessing disc in Z Cha from the radial velocity variations of various absorption and emission lines respectively. The very prominent normal superhump in Z Cha made it possible for Warner \\& O'Donoghue (1988) to study the location of the superhump light source. They found strong departures from axisymmetry in the superhump surface-brightness. O'Donoghue (1990) employed a modified eclipse mapping technique to Z Cha lightcurves and found the normal superhump light coming from three bright regions of the disc rim, located near the L1 point and the leading and trailing edges of the disc, concluding that the superhumps are tidal in origin, and that a highly eccentric disc with a smooth brightness distribution is not necessary to explain superhump behaviour. One anomalous eclipse did confine the superhump light source in Z Cha to the region of the quiescent bright spot. van der Woerd et al. (1988) concluded that the late superhumps in VW Hyi come from an optically thin plasma and could be a result of tidal interaction. In the SPH simulations of Murray (1996, 1998) pseudo-lightcurves are produced by assuming the heat produced by viscous dissipation to be radiated away where it is generated. Murray (1996, 1998) reveals an extended superhump light source in the outer disc, while Murray (1996) reveals an additional superhump modulation which arises from the impact of the accretion stream with the edge of the disc occuring at a varying depth in the primary Roche potential. This additional weaker superhump modulation is approximately 180$^\\circ$ out of phase with the modulation due to tidal stressing, another similarity between late superhumps in dwarf novae, the persistent superhumps in V348 Pup and the bright spot model. If we consider the stream to impact the disc at radius $r$ in a $\\frac{1}{r}$ potential, then the luminosity, $L$, of the bright spot should vary roughly as $\\Delta(\\frac{1}{r})$. Considering the change in $r$ as the disc with eccentricity $e$ precesses, we get $\\frac{\\Delta L}{L} \\sim 2e$. The eccentricities we find are in the range 0.035--0.15 predicting superhump fractional amplitude in the range 0.07--0.3. This is consistent with the measured superhump amplitudes in V348 Pup (Table \\ref{shper}). While we limit the conclusions drawn from our eclipse maps in Section \\ref{eclipsemapping}, there are a number of points deserving consideration. Our eclipse maps do not show evidence of a bright spot, but this does not rule out the possibility that a bright spot is the source of the superhump light, for the following reasons. First, we subtracted the superhump modulation from the lightcurves before performing the eclipse mapping, which should reduce the contribution of the bright spot in the maps if it is the primary source of the superhump light. Also, our maps are fixed in the precessing disc frame, rather than the orbital frame of the system, so the hot spot should be blurred azimuthally in our maps by $\\sim70^\\circ$ corresponding to the eclipse width of the system of $\\sim0.2$ in orbital phase. There will be additional azimuthal blurring since the eclipses contributing to each map have a spread of disc orientations at mid-eclipse corresponding to the values of $\\sigma_\\phi$ in Table \\ref{groups}. Azimuthal structure in the maps is also suppressed by looking for the maximally axisymmetric solution. The eclipse maps tell us that the azimuthally averaged radial extent of the emission is lowest at superhump maximum, shown in Figure \\ref{mapprofiles}. If this change in extent of the emission region is interpreted as a result of a changing disc size, the the smaller disc radius at superhump maximum is consistent with the bright spot model for the superhump light source. In the SPH models of Simpson \\& Wood (1998), the symmetry axis of the disc is aligned roughly perpendicular to the line of centres of the system when the superhump reaches maximum intensity. Inspecting their plots suggests that this model would lead to eclipses being earliest and widest at superhump phase 0 contrary to our findings. Simpson \\& Wood stress that their pseudo-lightcurves should be treated cautiously since no radiative processes were explicitly considered. The difference in the mass ratio between V348 Pup and the values considered in Simpson \\& Wood's simulations may affect the phasing of the early eclipses and superhump maximum, given that the predicted superhump waveform is sensitive to $q$. Furthermore, the spiral density waves in their simulations complicate the structure, so that simulated maps of the intensity may in fact produce reasonable agreement with our findings. SPH simulations (Murray 1996 \\& 1998, Simpson \\& Wood 1998) show the behaviour of tidally distorted discs to be more complicated than a simple eccentric disc, and with treatment of radiative processes the predictions are likely to become even more complicated. Once such models are developed further, comparisons with observation should provide a more complete understanding of superhump phenomena." }, "0003/astro-ph0003430_arXiv.txt": { "abstract": "The characteristics of the star model with the \"prescribed\" density distribution $\\rho=\\rho_c [1-(r/R)^\\alpha]$ are analytically studied. The model has been first considered briefly in our 30-year old note of a restricted circulation \\cite{pseu}. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003095_arXiv.txt": { "abstract": "We present simultaneous optical polarimetry and X-ray data of the near synchronous polar RX J2115--5840. We model the polarisation data using the Stokes imaging technique of Potter et al. We find that the data are best modelled using a relatively high binary inclination and a small angle between the magnetic and spin axes. We find that for all spin-orbit beat phases, a significant proportion of the accretion flow is directed onto the lower hemisphere of the white dwarf, producing negative circular polarisation. Only for a small fraction of the beat cycle is a proportion of the flow directed onto the upper hemisphere. However, the accretion flow never occurs near the upper magnetic pole, whatever the orientation of the magnetic poles. This indicates the presence of a non-dipole field with the field strength at the upper pole significantly higher. We find that the brightest parts of the hard X-ray emitting region and the cyclotron region are closely coincident. ", "introduction": "Polars (or AM Her systems) are amongst the most suitable objects with which to study the interaction between an accretion flow and a magnetic field. This is because the magnetic field of the accreting white dwarf is strong enough ($B\\sim$10--200MG) to prevent the formation of an accretion disk. Therefore, the dominating emission source at all wavelengths is the post-shock region above the surface of the white dwarf. In polars, the spin period of the white dwarf and the binary orbital period are generally synchronised and the accretion flow from the dwarf secondary star threads onto magnetic fields lines which have an unchanging orientation with respect to the white dwarf (see Cropper 1990 and Beuermann \\& Burwitz 1995 for general reviews of polars). However, four systems are known to be slightly ($\\sim1\\%$) asynchronous (the near synchronous polars) and the accretion flow will therefore attach onto different field lines as the flow rotates around the white dwarf on the timescale of the spin-orbit beat period. Until very recently, observations covering a beat period have been difficult to obtain because the beat period is weeks or more (V1432 Aql: Watson et al 1995, Friedrich et al 1996, Geckeler \\& Staubert 1997; BY Cam: Silber et al 1997, Mason et al 1998) or the system is faint (V1500 Cyg: Stockman, Schmidt \\& Lamb 1988). Now a fourth system (RX J2115--58) which is reasonably bright ($V\\sim$17) has been discovered with a beat period of 6.3 days (Schwope et al 1997, Ramsay et al 1999) which allows a detailed study of these systems to be undertaken for the first time. The observations of Ramsay et al (1999) made over 13 nights in 1997 provide the most direct evidence that the accretion flow is directed onto one magnetic pole and then the other as the flow rotates around the white dwarf on the timescale of the beat period. This is most apparent in the circular polarisation data which shows a change of sign when the accreting flow is directed onto opposite magnetic poles. From these observations they determined the spin period of the white dwarf to be $P_\\omega$=109.55 mins and the binary orbital period to be $P_\\Omega$=110.89 mins. The observations of Ramsay et al (1999) were difficult to reconcile with simple views of how the accretion stream attaches onto the magnetic field of the white dwarf. To investigate this in greater detail we have obtained simultaneous optical polarisation and X-ray data obtained using {\\sl RXTE} in July 1998. ", "conclusions": "" }, "0003/gr-qc0003061_arXiv.txt": { "abstract": "\\baselineskip = 16 true pt Dirac equation is separable in curved space-time and its solution was found for both spherically and axially symmetric geometry. But most of the works were done without considering the charge of the black hole. Here we consider the spherically symmetric charged black hole background namely Reissner-Nordstr\\\"om black hole. Due to presence of the charge of black-hole charge-charge interaction will be important for the cases of incoming charged particle (e.g. electron, proton etc.). Therefore both gravitational and electromagnetic gauge fields should be introduced. Naturally behaviour of the particle will be changed from that in Schwarzschild geometry. We compare both the solutions. In the case of Reissner-N\\\"ordstrom black hole there is a possibility of super-radiance unlike Schwarzschild case. We also check this branch of the solution. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003210_arXiv.txt": { "abstract": "Starting from the observation that kilohertz Quasi Period Oscillations (kHz QPOs) occur in a very narrow range of X--ray luminosities in neutron star low mass X--ray binaries, we try to link the kHz QPO observability to variations of the neutron star magnetospheric radius, in response to changing mass inflow rate. At low luminosities, the drop off of kHz QPOs activity may be explained by the onset of the centrifugal barrier, when the magnetospheric radius reaches the corotation radius. At the opposite side, at higher luminosities, the magnetospheric radius may reach the neutron star and the vanishing of the magnetosphere may led to the stopping of the kHz QPOs activity. If we apply these constraints, the magnetic fields of atoll ($B\\sim 0.3-1\\times 10^8$ G for Aql X-1) and Z ($B\\sim 1-8\\times 10^8$ G for Cyg X-2) sources can be derived. These limits naturally apply in the framework of beat frequency models but can also work in the case of general relativistic models. ", "introduction": "Accretion phenomena onto a fast spinning, weakly magnetic neutron star three radii which do not vary with the mass inflow rate can be identified: \\begin{itemize} \\item{the neutron star radius, $R$, which depends on the neutron star mass and equation of state. We scale here $R=10\\,R_6$ km for a neutron star of mass $M=2\\,M_2\\msole$; } \\item{ the marginally stable orbit radius, $r_{\\rm ms}$. In the case of a rapidly rotating neutron star a small correction to the Schwarzschild value has to be included. As a first order in the specific angular momentum $j$ [$j=2\\,\\pi\\,c\\,I/(G\\,M^2\\,P)$, with $c$ the speed of light, $I$ and $P$ the neutron star moment of inertia and spin period and $G$ the gravitational constant], $r_{\\rm ms}\\simeq (6-2.31\\,j)\\,G\\,M/c^2$ (Klu\\'zniak 1998). In the case of `soft' equations of state this radius is larger than $R$, whereas for stiff equation of states it is smaller and does not play any role. In the case of a spin frequency of 1.8 ms (as in the case of Aql X-1; Zhang et al. 1998a), and for a $2\\msole$ neutron star with $I=2\\times10^{45}$ g cm$^{-2}$, $r_{\\rm ms}\\sim 16\\,M_2$ km; } \\begin{figure*}[!htb] \\centerline{\\psfig{figure=atoll31.ps,width=11.cm}} \\caption{Color-color diagram of Aquila X--1. The soft and hard colors are defined as the ratio of count rates in the bands $3.5 - 6.0$ keV and $2.0 - 3.5$ keV, and $9.7 - 16.0$ keV and $6.0 - 9.7$ keV, respectively. Points in the banana branch are marked with closed circles, in the lowest island state with filled squares, in the middle island with open circles and in the extreme island with open squares. Typical error bars in the banana and the island states are shown. Black and grey symbols indicate segments with and without kHz QPOs, respectively. Taken from Reig et al. (2000). } \\label{uno} \\end{figure*} \\item{the corotation radius, $r_{\\rm cor}$, at which a test particle in Keplerian orbit corotates with the neutron star, $r_{\\rm cor}=\\Bigr({{G\\,M\\,P^2}\\over {4\\,\\pi^2}}\\Bigl)^{1/3}$. In the case of a spin frequency of 1.8 ms ($P=1.8\\,P_{\\rm 1.8 ms}$ ms) we have $r_{\\rm cor}=28\\,M_2^{1/3}\\,P_{\\rm 1.8 ms}^{2/3}$ km.} \\end{itemize} As the mass inflow rate changes the limiting radius at which the neutron star magnetic field pressure balances the ram pressure of the mass inflow towards the neutron star (i.e. the magnetospheric radius, $r_{\\rm m}$) changes too. In the case of spherical accretion with a dipolar magnetic field $r_{\\rm m}\\sim\\Bigr({{\\mu^4}\\over{2\\,G\\,M\\,\\mdot^2}}\\Bigl)^{1/7}$ (where $\\mdot$ is the mass inflow rate onto the neutron star magnetosphere and $\\mu=B\\,R^3$ is the magnetic moment). A parameter $\\xi$ is usually considered to account for the case of disk accretion, with $\\xi\\sim 0.5-1$ (Ghosh \\& Lamb 1979, 1992). Wang (1995) pointed out that $\\xi$ is in the range 0.87--0.95. In the following, we assume $\\xi=1$. For an accretion luminosity of $L=G\\,M\\,\\mdot/R=10^{37}\\,L_{37}\\ergs$, we derive a magnetospheric radius of $r_{\\rm m}=20\\,B_8^{4/7}\\,R_6^{10/7}\\,M_2^{1/7}\\,L_{37}^{-2/7}\\,\\xi$ km (where $B=B_8\\,10^8$ G is the magnetic field). For higher luminosities the radiation pressure starts playing a role (region A in the standard disk model) and the magnetospheric radius changes to $r_{\\rm m}^{\\rm A}=33\\,B_9^{20/39}\\,L_{38}^{-2/13}\\,M_{2}^{1/39} \\,R_6^{18/13}\\,\\xi$ km (cf. Ghosh \\& Lamb 1992; here we neglect the weak dependence on the viscosity parameter). ", "conclusions": "We interpret the onset and the end of kHz QPOs in terms of the variable extent of the neutron star magnetosphere. The turn-on of kHz QPOs at low fluxes is interpreted as the starting of accretion onto the neutron star surface, overcoming the effects of the centrifugal barrier (here we do not claim that this barrier is working perfectly, but if it can stop the large fraction of the infalling matter, the flux reduction will prevent to detect kHz QPOs). The disappearance of the kHz QPOs corresponds instead to the `disruption' of the magnetosphere on the neutron star (see also Cui et al. 1998). Given these assumptions, we estimate in the case of the atoll source Aql X-1 a magnetic field of $B\\sim 0.3-1\\times 10^8$ G. Similar results can be obtained for other atoll sources. Interpreting the saturation of the kHz QPO frequency for increasing luminosities in 4U 1820--30, leads to a magnetic field estimate of $\\sim 2\\times 10^8$. In the case of Z sources we derive a rough estimate based on luminosity obtained with EXOSAT. In the case of Cyg X-2, we obtain a very conservative lower limit of $B\\sim 1-8\\times 10^8$ G. This is consistent with the idea that Z sources have higher magnetic field than atoll sources. This scenario allows for clear predictions on the luminosities at which the kHz QPOs set in and out, which can be verified with RossiXTE monitoring of low mass X--ray binaries." }, "0003/astro-ph0003483_arXiv.txt": { "abstract": "\\noindent \\rightskip=0pt We study the evolution of Hernquist profile ``galaxies'' in the presence of self-interacting dark matter (SIDM), where the properties of the dark matter can be parameterized by one number, $\\hat{\\sigma}_{DM}=\\sigma_{DM} M_T/a^2$ for a halo of mass $M_T$ and break radius $a$. While the halos form constant density cores of size $\\sim a/2$ on the core radius relaxation time scale ($t_{rc} \\simeq 1.7\\tdyn/\\hat{\\sigma}_{DM}$) core collapse begins shortly thereafter and a steeper $1/r^2$ central density cusp starts forming faster than predicted by 2-body relaxation. The formation of the steeper central cusp is accelerated if the cooling baryons adiabatically compress the dark matter. The natural consequence of SIDM is to exacerbate rather than to mitigate astrophysical problems created by dark matter density cusps. ", "introduction": "A model based on the growth of small fluctuations through gravitational instability in a universe with cold dark matter (CDM) provides an excellent fit to a wide range of observations on large scales ($\\gg 1$Mpc). However the nature and properties of the CDM, apart from its being cold and dark, remain mysterious. It is one of the major goals of cosmology to further constrain the properties of the dark matter and determine its nature. Recently much attention has been refocused on this question because of a suggestion by Spergel \\& Steinhardt~(\\cite{SpeSte}) that possible discrepancies with observations on kpc scales could be probes of dark matter properties, particularly its self-interaction rate. In its simplest incarnation, the self-interacting dark matter (SIDM) model provides a 1-parameter family of models with standard CDM as a limiting case, and it is thus interesting to examine this model in some detail. In this paper we calculate the quantitative effects of SIDM on the evolution of isolated halos using an N-body code with particle scattering. We review the astrophysical arguments motivating SIDM in \\S\\ref{sec:motivate} and describe our implementation of the scattering algorithm in \\S\\ref{sec:method} (our method is very similar to that of Burkert~(\\cite{Bur}), but uses a more accurate approximation to the scattering rate integral). The impact of self-interaction on various astrophysical processes is discussed in \\S\\ref{sec:core} and we finish with some conclusions in \\S\\ref{sec:conclusion}. ", "conclusions": "\\label{sec:conclusion} We combined an N-body code with a Monte Carlo model of scattering to examine the effects of self-interacting dark matter (SIDM) on the properties of dark matter halos. Our models form a one parameter family with ``standard'' CDM as a limiting case. We first examined the time evolution of Hernquist models with $1/r$ central density cusps in the presence of SIDM. Our simulations are similar to those of Burkert~(\\cite{Bur}), but use a better approximation to the scattering integrals. The models form a constant density core with a radius $\\sim 0.4 a$ on the core radius relaxation time scale $t_{rc}=1.7\\tdyn/\\hat{\\sigma}_{DM}$, where the SIDM cross section per unit mass is $\\hat{\\sigma}_{DM}=\\sigma_{DM} M_T/a^2$. The initial growth of the core radius is very rapid because the relaxation time is much shorter in the inner regions (see Quinlan~\\cite{Qui}). Shortly after reaching this maximum core size the core begins to shrink. We find that the recollapse proceeds on a faster time scale than expected from 2-body relaxation. The central density evolution evolves on time scales of $t_{rc}\\hat{\\sigma}_{DM}^{1/2}\\propto \\hat{\\sigma}_{DM}^{-1/2}$ rather than $t_{rc}\\propto\\hat{\\sigma}_{DM}^{-1}$, probably because SIDM is not a diffusive process when the mean free paths are large. Thus, the dark matter briefly has a large, finite core radius which could be fine tuned to address the problem of dwarf galaxy rotation curves at the expense of predicting core collapsed dark matter halos with $\\rho \\propto 1/r^2$ cusps in most other galaxies. In particular, it is probably impossible to use SIDM to eliminate the dark matter cusps in the dwarf galaxies while preserving them in the low surface brightness galaxies. Even with SIDM, however, we must also consider the role of the baryons in modifying the structure of the dark matter halo. In particular, as the baryons cool and form a disk they adiabatically compress the dark matter (Blumenthal et al.~\\cite{BFFP}; Dubinski~\\cite{Dub}). The adiabatic compression raises the central dark matter density, increases the SIDM scattering rate, and reduces the time scale for core collapse. Moreover, adiabatic compression of a density profile with a finite core radius produces a steep central density cusp (Young~\\cite{Young}) which relaxes to form a $\\rho \\propto 1/r^2$ density cusp even faster than the shallower $\\rho \\propto 1/r$ cusp we considered initially (Quinlan~\\cite{Qui}). As a result, the development of a core radius due to the SIDM scattering is first reversed by the compression, and then core collapse begins on a far shorter time scale. Such a further acceleration of the time scale for producing a final density distribution with a steep $\\rho \\propto 1/r^2$ density cusp will make it even harder for SIDM to maintain a core radius in the dark matter profile for periods comparable to the presumed lifetimes of dwarf galaxies. All of the above simulations supposed that the halo collapsed and virialized before scattering became important, which is appropriate to ``low'' scattering rates. Finally, we considered the effects of SIDM on the collapse of cold top-hat perturbations with $\\hat{\\sigma}$ large enough that scattering and virialization could occur at the same time. We found that the scattering allowed higher phase space densities at fixed dynamical time than could be found in the collisionless systems. It thus appears as though SIDM exacerbates rather than solves the ``central density cusp problem''. Turning this around we can say that astrophysics almost certainly requires SIDM cross sections small enough to avoid significant scattering over the age of the universe for all galaxies, or mean free paths $\\ga 1$~Mpc. \\bigskip" }, "0003/astro-ph0003160_arXiv.txt": { "abstract": "Is gravitational growth responsible for the observed large scale structure in the universe? Do we need non-gaussian initial conditions or non-gravitational physics to explain the large scale features traced by galaxy surveys? I will briefly revise the basic ideas of non-linear perturbation theory (PT) as a tool to understand structure formation, in particular through the study of higher order statistics, like the skewness and the 3-point function. Contrary to what happens with the second order statistics (the variance or power-spectrum), this test of gravitational instability is independent of the overall amplitude of fluctuations and of cosmic evolution, so that it does not require comparing the clustering at different redshifts. Predictions from weakly non-linear PT have been compared with observations to place constraints on our assumptions about structure formation, the initial conditions and how galaxies trace the mass. ", "introduction": "Where does structure in the Universe come from? The current paradigm is that it comes from gravitational growth of some small initial fluctuations. The self-gravity of an initial overdensed region increases its density contrast so that eventually the region collapses. For a flat Universe in the linear regime, the local density contrast $\\delta\\equiv \\rho/\\bar{\\rho}-1$ grows as the expansion factor, eg $D=a$, so that since decoupling linear gravitational growth has the potential of amplifying fluctuations by at least a factor of a thousand. But Gravity is not linear and when objects start collapsing the growth could be much larger. On galactic scales one also has to consider other forces such as hydrodynamics, heating and cooling by friction, dissipation, feedback mechanism from stars, such as nova and supernova explosions, interaction with the CMB and so on. To test if the above picture of gravitational growth is correct we need to deal with a classical initial condition problem. Because gravitational time scales are very slow, we have no way to measure the growth of individual large scale structures and we need to resort to the statistical study of mean quantities. One can imagine, for example, measuring the rms fluctuations (at a given scale) at different cosmic times to see if this agrees with the predicted amount of gravitational growth, $D$. Observationally this corresponds to finding the clustering properties of some tracer of structure (eg galaxies) at different redshifts. If the tracer is not perfect, we will have some statistical {\\it biasing}. The problem with this approach is that by the time the rms fluctuations change significantly there typically has also been a substantial cosmic evolution of the corresponding tracers. Thus, it is difficult to disentangle the effects of the underlying cosmological model (which sets the rate $D$ of gravitational growth) from galaxy evolution. It is therefore important to have a way of testing the gravitational growth paradigm at a single cosmic time or redshift. Higher order correlations and weakly non-linear clustering allows us to do just this. This is because one can construct ratios of higher order correlations to powers of the two point amplitude which are independent of cosmic time or cosmological parameters, but still contain information of the underlying dynamics. ", "conclusions": "The values of $S_N = \\xibar_N/\\xibar_2^{N-1}$ can be measured as traced by the large scale galaxy distribution (eg Bouchet \\etal 1993, Gazta\\~naga 1992, 1994, Szapudi el at 1995, Hui \\& Gazta\\~naga 1999 and references therein), and also the weak-lensing (Bernardeau \\etal 1997, Gazta\\~naga \\& Bernardeau 1998) or the Ly-alpha QSO absorptions (Gazta\\~naga \\& Croft 1999). These measurements of the skewness $S_3$, kurtosis $S_4$, and so on, can be compared with the predictions from weakly non-linear perturbation theory (see Figure 1) to place constraints on our assumptions about gravitational growth, initial conditions or biasing at a given redshift (see Mo, Jing \\& White 1997). Contrary to what happens with the second order statistics (eg the variance), this test of gravitational instability is quite independent of the overall amplitude of fluctuations and other assumptions of our model for cosmological evolution, and does not require comparing the clustering at different redshifts. As shown in Gazta\\~naga \\& Lobo (2000), one can also use the $S_N$ measurements to constraint non-standard cosmologies. Frieman \\& Gazta\\~naga (1999) have presented new results for the angular 3-point galaxy correlation function in the APM Galaxy Survey and its comparison with theoretical expectations (see also Fry 1984, Scoccimarro et al. 1998, Buchalter, Jaffe \\& Kamionkowski 2000). For the first time, these measurements extend to sufficiently large scales to probe the weakly non-linear regime (see previous work by Groth \\& Peebles 1977, Fry \\& Peebles 1978, Fry \\& Seldner 1982). On large scales, the results are in good agreement with the predictions of non-linear perturbation theory, for a model with initially Gaussian fluctuations (see Figure 1). This reinforce the conclusion that large-scale structure is driven by non-linear gravitational instability and that APM galaxies are relatively unbiased tracers of the mass on large scales; they also provide stringent constraints upon models with non-Gaussian initial conditions (eg see Gazta\\~naga \\& Mahonen 1996; Peebles 1999a,b; White 1999; Scoccimarro 2000)." }, "0003/astro-ph0003356_arXiv.txt": { "abstract": "By introducing a hitherto ignored $^6$Li producing process, due to accelerated $^3$He reactions with $^4$He, we show that accelerated particle interactions in solar flares produce much more $^6$Li than $^7$Li. By normalizing our calculations to gamma-ray data we demonstrate that the $^6$Li produced in solar flares, combined with photospheric $^7$Li, can account for the recently determined solar wind lithium isotopic ratio, obtained from measurements in lunar soil, provided that the bulk of the flare produced lithium is evacuated by the solar wind. Further research in this area could provide unique information on a variety of problems, including solar atmospheric transport and mixing, solar convection and the lithium depletion issue, and solar wind and solar particle acceleration. ", "introduction": "The solar wind lithium isotopic ratio, ($^6$Li/$^7$Li)$_{\\rm sw}$ =0.032$\\pm$0.004, has recently been determined from measurements in lunar soil (Chaussidon \\& Robert 1999). As these authors point out, this value greatly exceeds the expected photospheric ratio, based on the fact that $^7$Li in the photosphere is depleted by over a factor of 100 relative to its protosolar value (i.e. the photospheric vs. the meteoritic abundance, Grevesse, Noels, \\& Sauval 1996), and that this depletion, due to burning at the bottom of the convection zone (Brun, Turck-Chieze, \\& Zahn 1999), should lead to a much more severe depletion of $^6$Li, which burns at a lower temperature than $^7$Li. In addition, there exist observational upper limits on the photospheric ratio, ($^6$Li/$^7$Li)$_{\\rm ph}$$\\le$0.01 (M\\\"uller, Peytremann, \\& de la Reza 1975) and ($^6$Li/$^7$Li)$_{\\rm ph}$$\\le$0.03 (Ritzenhoff, Schr\\\"oter, \\& Schmidt 1997). Chaussidon \\& Robert (1999) thus suggest that the measured solar wind $^6$Li must be solar flare produced. However, they only consider $^6$Li production by spallation from C, N and O. The demonstration that solar flares can indeed account for the $^6$Li in the solar wind has very important implications on many problems in solar physics. Light element production by accelerated particle interactions was treated in detail (e.g. Ramaty et al. 1997). In non-solar settings, and for accelerated particles of predominantly low energy, the dominant reactions are $^4$He($\\alpha$,p)$^7$Li, $^4$He($\\alpha$,n)$^7$Be (with $^7$Be decaying to $^7$Li) and $^4$He($\\alpha$,x)$^6$Li (where x stands for either a proton and a neutron, or a deuteron). In solar flares, however, the reaction $^4$He($^3$He,p)$^6$Li is also very important (Mandzhavidze, Ramaty, \\& Kozlovsky 1997a), both because of its very low threshold energy and because for solar energetic particles $^3$He/$^4$He can be as large as 1 or even larger (e.g. Reames 1998). Such $^3$He/$^4$He enhancements are one of the main characteristics of the acceleration mechanism responsible for impulsive solar energetic particle events, as distinguished from gradual events, based on the duration of the accompanying soft X-ray emission. The $^3$He enrichment is thought to be due to stochastic acceleration through gyroresonant wave particle interactions which preferentially accelerate the $^3$He (Temerin \\& Roth 1992; Miller \\& Vi\\~nas 1993). Concerning the particles which interact at the Sun, evidence for accelerated $^3$He enrichment was obtained from the detection (Share \\& Murphy 1998) of a gamma-ray line at 0.937 MeV produced by the reaction $^{16}$O($^3$He,p)$^{18}$F$^*$ (Mandzhavidze, Ramaty, \\& Kozlovsky 1997b; 1999). Using gamma-ray data from 20 flares, Mandzhavidze et al. (1999) showed that for essentially all of these flares $^3$He/$^4$He can be as large as 0.1, while for some of them values as high as 1 are possible. In addition, they showed that for the particles that interact and produce gamma rays, $^3$He enrichments are present for both impulsive and gradual flares. Thus, we can expect $^3$He/$^4$He$\\gsim$0.1 for most flares that produce gamma rays and isotopes at the Sun. In the present Letter we carry out new calculations of Li production and re-calculate (see Ramaty \\& Simnett 1991) the average accelerated ion irradiation of the Sun, to show that flare accelerated particle interactions produce enough $^6$Li which, combined with photospheric $^7$Li, can account for the solar wind $^6$Li/$^7$Li measured in lunar soil. ", "conclusions": "We demonstrated that it is possible to produce enough $^6$Li by flare accelerated particles to account for the measured $^6$Li/$^7$Li in lunar soil that is thought to originate from solar wind implantation. The presence of enriched accelerated particle $^3$He is essential for the production of sufficient $^6$Li. We note that the radioactive $^{26}$Al in the early solar system is thought to be produced in $^3$He induced reactions (Lee et al. 1998). This raises the possibility that some of the meteoritic $^6$Li could also be of local early solar system origin. Kotov et al. (1996) claimed that flare accelerated particle interactions could account for all the photospheric lithium. If this were true, since the solar wind acceleration is not expected to significantly alter the lithium isotopic ratio, the solar wind $^6$Li/$^7$Li should exceed 0.2 (Figure~2), contrary to the observed value of 0.03. This confirms the previous result of Mandzhavidze et al. (1997a) that production in flares does not make a significant contribution to the average photospheric lithium. But the fact that as much as 10$^{30}$ Li atoms are produced in large solar flares, suggests that flare produced lithium may be detected in a small area of the solar surface near the foot points of the flaring loops shortly after the time of the flare (see Livshits 1997). In this connection, it is interesting to point out that Ritzenhoff et al. (1997) don't rule out the presence of $^6$Li near a sunspot at a value close to their reported upper limit $^6$Li/$^7$Li$\\le$0.03, which in fact coincides with the measured solar wind value. Further research in this area requires direct measurement of lithium and its isotopic ratio in the solar wind, spectroscopic measurements of $^6$Li in the photosphere, and the detection of gamma rays from small flares that would lead to a more precise determination of the proton irradiation of the Sun. All of these should lead to new insights into the processes of transport and mixing in the solar atmosphere and of the acceleration of the solar wind." }, "0003/astro-ph0003043_arXiv.txt": { "abstract": "Some topics in the theory of jets are reviewed. These include jet precession, unconfined jets, the origin of knots, the internal shock model as a unifying theme from protostellar jets to Gamma-ray bursts, relations between the Blandford-Znajek and MHD disk-wind models, and jet collimation in magnetic acceleration models. ", "introduction": "Precession is measured directly in the jets of SS 433, whose direction varies with an approximate 165d period. Indirect evidence is seen in the morphology of the hot spots of AGN jets. An example is as Cyg A, where the radio lobes show `fossil' hot spots, offset from the present (most luminous) hot spot position by rotation over angles of some 10 degrees. This gives an approximate point symmetric appearance to the lobes of. If the central engine causes the direction of the jet to change with time (precession), its path in space at a given time appears curved, like the spray of water from a rotating garden sprinkler. At each point along the instantaneous path of the jet, there is a slight difference between the direction of fluid motion and the tangent to the jet's path. In many cases, this may be the simplest explanation for apparent bending in FRII jets. Alternatives like redirection by clouds in the path of the jet have would be called for only if there is supporting evidence like the dissipation and decollimation that accompanies the redirection of supersonic flows by external obstacles (observe this by directing the jet from a garden hose at the tiles on your garden path). If jets are produced by accreting compact objects, their flow direction is plausibly determined either by the rotation axis of the disk, or that of the accreting object. The rate at which the rotation axis of the compact object can change is limited by the rate $\\dot M/M$ at which its angular momentum can change by accretion. The disk itself can change direction more rapidly, for example if the angular momentum vector of the gas supplied to the disk changes in time. It then takes only the viscous time in the disk for this change to propagate to the inner region where the jet originates. If the jet is caused by the disk itself, as in the magnetic wind model, its direction will naturally follow the orientation of the inner disk. If, on the other hand, the jet is caused by the rotation of the compact object, as in the Blandford-Znajek (1977) model (see also Blandford, 1993), one might at first sight expect the jet direction to be given also by the axis of the rotating hole. In this mechanism, however, a disk must be present to supply the magnetic field to the horizon of the hole and extract the rotation energy. Since the mechanism by itself does not produce a highly collimated jet, it is possible that the collimation of a Blandford-Znajek outflow is also provided by the disk. In this case, the direction of the jet could follow the disk axis even though it is powered by the hole. In either case, we arrive at changes of orientation of a disk as a likely explanation for jet precession (cf. van den Heuvel et al. 1982). The most promising cause for such changes proposed so far is an instability due to irradiation of the disk by the central source. Such irradiation can cause the outer parts of the disk to develop a radiation-heated atmosphere which drives a wind (Begelman et al.\\ 1983). Schandl and Meyer (1994) have shown that the momentum flux in such a wind can cause the disk to become unstable to bending out-of-the plane, i.e. warping. As soon as the disk is warped, the radiation intercepted by one side of the disk is larger than the other, and the wind pressure on that side larger. The net torque on the disk due to the difference in wind pressure causes the irradiated part of the disk to precess. At the same time, the warp propagates radially by viscous diffusion, and grows in amplitude with time. Shadowing of parts of the disk by warps in regions closer in makes the nonlinear development of the warps quite complicated. Schandl and Meyer show how this irradiation-driven wind instability can explain the precessing tilted disk in Her X-1. A radiation-driven wind is expected to be important in the outer regions of a disk, where the Compton temperature corresponding to the incident X-ray spectrum is of the order of the escape velocity from the disk. Closer to the compact object, wind losses by this process are small. Pringle (1996) studied the same instability without an irradiation-induced wind, using only the effect of radiation pressure on the disk. Pringle (1997) follows the evolution of such warps to arbitrary tilt angles, including the self-shadowing effects, and concludes that the inner regions of AGN disks can tilt over more than 90\\deg. Apart from a time-dependent jet direction, this means that one should expect little correlation between jet axis and the plane of the host galaxy. The equations for the evolution of warps in thin accretion disks have been corrected with respect to previous treatments, and put on a firm mathematical basis by Ogilvie (1999). He also presents a practical scheme for the numerical treatment of evolving warps of arbitrary amplitude. A final cause for precession could be the momentum carried by a magnetically accelerated jet. As in the case of irradiation- and wind -induced warping, the reaction of the jet thrust on the disk may make the disk unstable if the thrust depends on the disk inclination. This possibility has apparently not been studied much, so far. ", "conclusions": "" }, "0003/astro-ph0003455_arXiv.txt": { "abstract": "We investigate the properties of clusters of galaxies in the $\\Lambda$CDM models with a steplike initial power spectrum. We examine the mass function, the peculiar velocities and the power spectrum of clusters in models with different values of the density parameter $\\Omega_0$, the normalized Hubble constant $h$ and the spectral parameter $p$, which describes the shape of the initial power spectrum. The results are compared with observations. We also investigate the rms bulk velocity in the models, where the properties of clusters are consistent with the observed data. We find that the power spectrum of clusters is in good agreement with the observed power spectrum of the Abell-ACO clusters, if the spectral parameter $p$ is in the range $p=0.6-0.8$. The power spectrum and the rms peculiar velocity of clusters are consistent with observations only if $\\Omega_0<0.4$. The $\\Omega_0=0.3$ models are consistent with the observed properties of clusters, if $h=0.50-0.63$. For $h=0.65$, we find that $\\Omega_0=0.20 -0.27$. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003349_arXiv.txt": { "abstract": "We present mid and far IR ISO data of the CfA Seyfert galaxy sample. These data allow a detailed study of the far IR Spectral Energy Distribution (SED) of these galaxies. A Bayesian inversion method has been used to invert the SED of these sources yielding two fundamental results, namely, that the mid and far IR SED of Seyfert galaxies can be explained solely through thermal reradiation of high energy photons by dust, and that this thermal emission is made up of two or three different independent components, a warm, a cold and a very cold dust component. These thermal components have been readily explained as produced respectively by warm dust heated by either the active nucleus or by circumnuclear starburts, cold dust heated by star forming region in the galaxy disk, and very cold dust heated by the general interstellar radiation field. Comparisons between the parameters obtained from the analysis of the IR SEDs (fluxes, temperatures, luminosities) have been made. Our results suggest that the emission in the mid IR is anisotropic and the differences found between Seyfert 1 and Seyfert 2s can be explained with thin molecular tori models. ", "introduction": "The origin of the IR emission from Seyfert galaxies has been a matter of discussion since the early work of Rieke \\& Low (1972); Stein (1975); Rieke (1978); Neugebauer (1978), and others. These authors found an emission excess between 3 and 5~$\\mu$m in Seyfert galaxies, that originated a strong controversy between those who supported that the emission was non thermal and those who defended that the excess was emission from dust heated by the nucleus. The use of efficient instruments to observe at 10~$\\mu$m, and especially the pioneering work at the far IR carried out from the Kuiper Airborn Observatory (KAO) showed the importance of the mid and far IR emission to quantify the bolometric luminosity of Seyfert galaxies (Telesco \\& Harper 1980; Smith et al. 1983). However, it was the IRAS satellite that provided an extensive set of IR data for a large number of galaxies, from which it was shown that Seyfert galaxies are indeed strong far IR emitters (Rodr\\'{\\i}guez Espinosa, Rudy \\& Jones 1987; Edelson, Malkan \\& Rieke 1987; Spinoglio et al. 1995). The IRAS satellite was key to the understanding of the importance of the IR emission to the total luminosity of active galaxies. However the IRAS data alone are not sufficient to clarify the nature of the IR emission, as there are measurements only at a limited number of wavebands preventing a good definition of the shape of the Spectral Energy Distribution (SED) at mid and far IR wavelength (see, e.g., Telesco 1988; 1990; Bregman 1990). A proper characterization of the mid and far SED is essential to understand the emission mechanisms that produce the high output of Seyfert galaxies in the IR. Recent studies have discussed the origin of the IR emission suggesting that it is of thermal origin. For example Giuricin, Mardirossian \\& Mezzetti (1995) have studied a complete sample of Seyfert galaxies at 10~$\\mu$m and propose that the emission is due to thermal reemission by dust. Bonatto \\& Pastoriza (1997), based on color studies of IRAS data from diverse Seyfert samples, find that the colors obtained can be explained with a combination of dust heated by the nucleus plus cold dust in the host galaxy. Siebenmorgen et al. (1997) show that the SED at IR and milimetric wavelengths can be modeled assuming that the dust heated by a central source dominates the luminosity output of these objects. Maiolino et al. (1998) confirmed this last result finding, in their high resolution IR images of the Circinus galaxy, an unresolved source, with size $<$~1~pc, that is reprocessing the nuclear output via dust reradiation. Rigopoulou et al. (1997) have observed a sample of AGN in the CO mm line suggesting that the far IR emission of Seyfert galaxies is thermal, based on three different evidences: the correlation found between the far IR and the CO emission, the dependence of the far IR emission to hydrogen molecular mass ratio with dust temperature, and the similarity of the profile shape of the CO and HI lines. Another important issue is the understanding of the differences between the two Seyfert types. According to the unified models, Seyfert 2 nuclei are intrinsically similar to Seyfert 1 nuclei, the differences observed being due solely to geometrical effects. In Seyfert 2, neither the broad line region nor the optical, UV and soft X ray continuum can be observed directly because the central region is obscured by intervening material in the line of sight. Some authors argue that this material forms a sort of disc or torus of molecular material. This disc or torus is thought to be reponsible for the collimation of radiation and the observed anisotropies, i.e., biconic structures in emission line images (Simpson et al. 1997; Wilson et al. 1993) or highly collimated jets. It is expected that at sufficiently long wavelengths the optical depth of the torus would decrease and the differences between the two Seyfert types should disappear. Several tests have been made to ascertain the presence of these molecular tori (Heckman 1995; Pier \\& Krolik 1993). These obscuring tori have been theoretically modeled by Pier \\& Krolik (1992) and Granato \\& Danese (1994) among others, predicting that the mid IR optical depth is still considerable, thus it should be expected that Seyfert 1s are more luminous than Seyfert 2s in the mid IR. In this work, we make an attempt to understanding the origin of the mid and far IR emission from Seyfert galaxies through the study of their spectral energy distributions (SED). We present ISO data of the entire CfA Seyfert sample (Huchra \\& Burg 1992), consisting of 25 Seyfert 1 and 22 Seyfert 2, plus a LINER. Section 2 describes the observations. Section 3 presents the separation of the Spectral Energy Distributions in thermal components by means of the Inverse Planckian Transform. Section 4 describes the thermal emission components obtained from the inversion and in Section 5 we perform a statistical analysis of the parameters obtained and discuss the differences between the two Seyfert types. As the CfA Seyfert sample is a complete sample of Seyfert galaxies we expect that the results obtained here are statistically significant for all Seyfert galaxies and certainly suggestive for other classes of AGNs. ", "conclusions": "We have presented far IR photometry with ISO of the CfA Seyfert sample. The data have allowed a detailed study of the far IR SED of these sources using a Bayesian inversion method. It has been shown that the mid and far IR emission of Seyfert galaxies can be explained by the emission of three thermal components, a warm component, associated with dust heated by the nucleus and circumnuclear starformation regions; a cold dust component heated by star forming region in the galaxy disk, and very cold dust component heated by the general interstellar radiation field. The mid to far IR output from Seyfert galaxies does not have a simple origin but different ingredients play an important role in it. The comparison of cold far IR fluxes with H$\\alpha$ data confirms that the cold emission component that appeared naturally from the inversion process is related to radiation from dust in starforming regions in the galaxy discs. We find that the mid IR emission (warm component) is larger in Seyfert 1 than in 2s, suggesting the presence of obscuring material in Seyfert 2s. The median value obtained for the optical depth is in the range predicted by the thin and extended tori models." }, "0003/astro-ph0003380_arXiv.txt": { "abstract": "In this paper we study the accuracy of photometric redshifts computed through a standard SED fitting procedure, where SEDs are obtained from broad-band photometry. We present our public code {\\it hyperz\\/}, which is presently available on the web. We introduce the method and we discuss the expected influence of the different observational conditions and theoretical assumptions. In particular, the set of templates used in the minimization procedure (age, metallicity, reddening, absorption in the Lyman forest, ...) is studied in detail, through both real and simulated data. The expected accuracy of photometric redshifts, as well as the fraction of catastrophic identifications and wrong detections, is given as a function of the redshift range, the set of filters considered, and the photometric accuracy. Special attention is paid to the results expected from real data. ", "introduction": "The estimate of redshifts through photometry is one of the most promising techniques in deep universe studies, and certainly a key point to optimize field surveys with large-field detectors. It is in fact an old idea of Baum (1962), who originally applied it to the measure of redshifts for elliptical galaxies in distant clusters. It was later used by several authors in the eighties (Couch et al. 1983, Koo 1985) on relatively low-redshift samples, observed in the $\\sim 4000$ to $8000$\\,\\AA\\ domain. Later in the nineties, the interest for this technique has increased with the development of large field and deep field surveys, in particular the Hubble Deep Field North and South (HDF-N and HDF-S). Basically two different photometric redshift techniques can be found in the literature: the so-called empirical training set method, and the fitting of the observed Spectral Energy Distributions (hereafter SED) by synthetic or empirical template spectra. The first approach, proposed originally by Connolly et al. (1995, 1997), derives an empirical relation between magnitudes and redshifts using a subsample of objects with measured spectroscopic redshifts, i.e. the training set. A slightly modified version of this method was used by Wang et al. (1998) to derive redshifts in the HDF-N by means of a linear function of colours. This method produces small dispersions, even when the number of filters available is small, and it has the advantage that it does not make any assumption concerning the galaxy spectra or evolution, thus bypassing the problem of our poor knowledge of high redshift spectra. However, this approach is not flexible: when different filter sets are considered, the empirical relation between magnitudes and redshifts must be recomputed for each survey on a suitable spectroscopic subsample. Moreover, the training set is constituted by the brightest objects, for which it is possible to measure the redshift. Thus, this kind of procedure could in principle introduce some bias when computing the redshifts for the faintest sources, because there is no guarantee that we are dealing with the same type of objects from the spectrophotometrical point of view. Also, the redshift range between $1.4$ and $2.2$ had been hardly reached by spectroscopy up to now, because of the lack of strong spectral features accessible to optical spectrographs. Thus, no reliable empirical relation can be found in this interval. The SED fitting procedure, described in detail in the following section, bases its efficiency on the fit of the overall shape of spectra and on the detection of strong spectral properties. The observed photometric SEDs are compared to those obtained from a set of reference spectra, using the same photometric system. The photometric redshift of a given object corresponds to the best fit of its photometric SED by the set of template spectra. This method is used mainly for applications on the HDF, using either observed or synthetic SEDs (e.g. Mobasher et al. 1996, Lanzetta et al. 1996, Gwyn \\& Hartwick 1996, Sawicki et al. 1997, Giallongo et al. 1998, Fern\\'andez-Soto et al. 1999, Arnouts et al. 1999, Furusawa et al. 2000). A crucial test in all cases is the comparison between the photometric and the spectroscopic redshifts obtained on a restricted subsample of relatively bright objects. A combination of this method with the Bayesian marginalization introducing a prior probability was proposed by Ben\\'{\\i}tez (2000). The aim of this paper is to explain in a straightforward way the expected performances and limitations of photometric redshifts computed from broad-band photometry. This study has been conducted with our public code called {\\it hyperz\\/}, which adopts a standard SED fitting method, but most results should be completely general in this kind of calculations. This program was originally developed by Miralles (1998) (see also Pell\\'o et al. 1999), and the present version of the code {\\it hyperz\\/} is available on the web at the following address: \\begin{center} {\\tt http://webast.ast.obs-mip.fr/hyperz\\/} . \\end{center} The plan of the paper is the following. In Section~\\ref{method} we present the method used by {\\it hyperz\\/} and the involved set of parameters. The accuracy of the redshift determinations and the expected percentage of catastrophic identifications, as a function of the filter set and the photometric errors, are studied through simulations in Section~\\ref{simul}. The influence of the different parameters on the accuracy of photometric redshifts is investigated in Section~\\ref{param}, using both simulations and spectroscopic data from the HDF. Section~\\ref{realsim} is devoted to the analysis on the expected accuracy and possible systematics when exploring real data, coming from deep photometric surveys. A general discussion is given in Section~\\ref{discuss} and conclusions are listed in Section \\ref{conclu}. ", "conclusions": "\\label{conclu} We have presented the characteristics and the performances of our public code {\\it hyperz}, available on the web, which make use of the template SED fitting technique. We can summarize the main conclusions as follows: \\begin{enumerate} \\item Simulations of ideal catalogues have shown the main trends of the accuracy on $z_{\\rm phot}$ calculations. In particular, $z_{\\rm phot}$ estimates are improved when the filters set spans a wide wavelength range, including near-IR and $U$ filters, and when the photometric errors become small. \\item We have investigated the weight of the different parameters on the final results, using both a spectroscopic subsample of HDF and simulations. In particular, the templates, the flux decrement by Lyman forest, the age of the stellar population, the reddening, the cosmology, the metallicity, the IMF and the presence of emission lines have been discussed. According to these results, the $z_{\\rm phot}$ preciseness seems to be more sensitive to the photometric accuracy rather than to the detailed set of parameters. Nevertheless, a subset of these parameters (reddening, age of the stellar population and Lyman forest blanketing) has to span a sufficiently wide range of values to obtain accurate $z_{\\rm phot}$s. \\item The robustness of the method has been illustrated through realistic deep field simulations, aiming to reproduce the redshift distribution, photometric accuracy and limiting magnitudes encountered in deep field surveys. \\item We have pointed out some of the manifold applications of the photometric redshift machinery in present and future projects. \\item We plan to include AGN SEDs in the present scheme of {\\it hyperz\\/}, as well as stellar templates, in order to automatically classify objects in a photometric survey through a unique pipeline. This particular application is presently under development (Hatziminaoglou et al. 2000). \\end{enumerate}" }, "0003/astro-ph0003039_arXiv.txt": { "abstract": " ", "introduction": "A decade after launch of the ROSAT satellite, narrow and wide angle surveys using All-Sky Survey (RASS) data or pointed observations become complete. In order to characterize the source content of the X-ray sky as seen by ROSAT, such surveys were performed at high galactic latitudes by different teams. The most important surveys based on pointed observations are the deep surveys in the Lockman Hole (Hasinger et al.~1998, Schmidt et al.~1998), the UK deep survey (McHardy et al.~1998), and the medium deep RIXOS survey (Mason et al.~2000). Appenzeller et al.~(1998) presented a catalogue of sources in selected areas of the sky down to the RASS limit. Thomas et al.~(1997) performed a high-galactic latitude survey ($|b| > 20\\degr$) of bright, point-like, soft sources (HR1 + $\\Delta$HR1 $ < $ 0.0), the corresponding survey of bright, point-like, hard sources (HR1 $ > $ 0.5) was presented by Fischer et al.~(1998, referred to a Paper I). For definition of the hardness ratio HR1 see Sect.~\\ref{s:tab} Based on their Schmidt plate survey run at Calar Alto, Bade et al.~(1995, 1998) published a catalogue of northern AGNs detected in the RASS. The missing piece, an unbiased survey of all bright, CR $ > 0.2$\\,s$^{-1}$, high-galactic latitude sources detected in the RASS without further selection, e.g.~for X-ray extent or X-ray color, is presented here. It is termed the ROSAT Bright Survey RBS. The survey area of the RBS is the sky above galactic latitude $30\\degr$ and below $-30\\degr$ excluding the Virgo cluster and the Magellanic clouds. The 1RXS-catalogue by Voges et al.~(1999) contains 2072 high-galactic latitude sources brighter than our count rate limit.% We refer to these as RBS0001 \\dots RBS2072 with increasing rightascension. After exclusion of the Virgo and MC regions, approximated as circles with radius $5\\degr$ centred on $(\\alpha,\\delta) = (16\\fd44,-73\\fd27), (83\\fd80,-68\\fd00)$, and $(188\\fd30,1\\fd70)$ the remaining survey area is 20391 deg$^2$ and contains 2012 sources. These were identified almost completely by catalogue work and a dedicated identification program. One of the main results of this survey is a catalogue of all 2012 sources summarizing their main properties at X-ray and optical wavelengths which is presented in this paper. A description of the source content highlighting the extremes and the statistics of the survey is in preparation and will be published in the near future. At that time the catalogue will go online, i.e.~finding charts and identification spectra will become electronically accessable. One of the major results of the RBS was the determination of the soft X-ray AGN luminosity function and, combined with deeper surveys, the assessment of its cosmological evolution and its contribution to the soft X-ray background (Miyaji et al.~2000). Some of the new interesting galactic sources found in the RBS, magnetic cataclysmic variables and isolated neutron stars, were presented in Schwope et al.~(1997, 1999a, 1999b) and Motch et al.~(1999). ", "conclusions": "" }, "0003/astro-ph0003394_arXiv.txt": { "abstract": "Recent studies of nearby spiral galaxies suggest that photodissociation regions (PDRs) are capable of producing much of the observed \\HI\\ in galaxy disks. In that case, measurements of the observed \\HI\\ column density and the far--ultraviolet (FUV) photon flux responsible for the photodissociation process provide a new probe of the volume density of the local underlying molecular hydrogen. We develop the method and apply it to the giant Scd spiral M101 (NGC~5457). The \\HI\\ column density and amount of FUV emission have been measured for a sample of 35 candidate PDRs located throughout the disk of M101 using the Very Large Array and the Ultraviolet Imaging Telescope. We find that, after correction for the best-estimate gradient of metallicity in the ISM of M101 and for the extinction of the ultraviolet emission, molecular gas with a narrow range of density from 30--1000 cm$^{-3}$ is found near star-forming regions at all radii in the disk of M101 out to a distance of $12' \\approx 26$ kpc, close to the photometric limit of $R_{25} \\approx 13.5'$. In this picture, the ISM is virtually all molecular in the inner parts of M101. The strong decrease of the \\HI\\ column density in the inner disk of the galaxy at $R_G < 10$ kpc is a consequence of a strong increase in the dust-to-gas ratio there, resulting in an increase of the \\Htwo\\ formation rate on grains and a corresponding disappearance of hydrogen in its atomic form. ", "introduction": "\\label{sec:intro} Massive gas clouds in the ISM are considered to be the progenitors of young stars in the conventional picture of star formation in galaxy disks. The gas may initially be either in an atomic or in a molecular state \\citep{elm95} depending on the ambient physical conditions (UV flux and gas pressure), but as the volume densities increase through self-gravity, the gas becomes mostly molecular on its way to forming stars. Once the gas is molecular and at densities above $\\approx 100\\:{\\rm cm}^{-3}$, the cooling rate is so high (e.g.~\\citet{gol78}) that rapid collapse follows. The rate-determining step in the conventional picture is then the aggregation of \\HI\\ or diffuse \\Htwo\\ and the formation of Giant Molecular Clouds. \\citet{elm93} has discussed this problem in detail. For many late-type systems, the rate-determining step along the path to star formation appears to be the conversion of \\HI\\ clouds into \\Htwo. The fraction of the ISM in atomic form is thought to increase as one progresses along the Hubble Sequence towards the late-type galaxies (see e.g.\\ the review by \\citet{rob94}). Furthermore, the atomic fraction is thought to dominate the molecular by an order of magnitude or more for a significant number of actively-star-forming galaxies of types Sc and later \\citep{you89}. However, in recent years evidence has been accumulating that a significant fraction of the \\HI\\ in the ISM of galaxies is in fact not a precursor to the star formation process, but instead is produced by photodissociation of the \\Htwo\\ by UV photons emanating from nearby newly-formed massive stars. In that case, either the way we compute the molecular fraction in galaxies via the CO(1-0) emission is wrong, or our ideas of how to control the otherwise runaway collapse of molecular clouds to form stars need further modification. \\citet{shu97} has summarized the view that magnetic fields may be a controlling factor in cloud collapse. The dissociation process was first described by \\citet{ste67}. The \\Htwo\\ absorbs photons emitted at 1108 \\AA\\ and 1008 \\AA\\ via electronic transitions in the Lyman ($X ^1\\Sigma_g^+ \\rightarrow B ^1\\Sigma_u^+$) and Werner ($X ^1\\Sigma_g^+ \\rightarrow C ^1\\Pi_u^+ $) bands. In the subsequent decay to the vibrationally excited levels of the ground electronic state, $\\sim 10$\\% of the \\Htwo\\ molecules will dissociate. Photons with wavelengths as long as $\\approx 1850$ \\AA\\ can continue to create \\HI\\ by dissociating the ``pumped'' ($X ^1\\Sigma_g^+, 0 < v < 14$) \\Htwo\\ via additional Lyman and Werner band transitions. The UV fluorescence spectrum predicted from this process was first observed by \\citet{wit89} in the nebula IC~63. The initial evidence that \\HI\\ may be a \\textit{product} of the star formation process rather than a precursor to it was discovered in a comparison of the morphology of dust lanes and \\HI\\ ridges in a prominent spiral arm of the galaxy M83 (NGC 5236) by \\citet{all85,all86}. This comparison showed that the ridges defining the \\HI\\ spiral arms peaked not on the dust lanes, as would have been expected in the conventional picture of density-wave-triggered star formation \\citep{rob69}, but instead they were co-linear with the chain of \\HII\\ regions created by young massive stars forming \\textit{downstream} from the dust lane. Other studies followed on M83 and other nearby spiral galaxies and have generally reached the same", "conclusions": "\\label{sec:imp} Figures \\ref{fig:tau1} and \\ref{fig:tau2} are our current best estimates for the radial dependence of the total gas volume density in M101. This gas should be essentially all \\Htwo. In the process of correcting the ``raw'' values of volume density (Table \\ref{tab:data}) for the radial variations of dust-to-gas fraction (resulting in Figures \\ref{fig:volcorr} and \\ref{fig:volcorrb}) and UV extinction (finally producing Figures \\ref{fig:tau1} and \\ref{fig:tau2}), the radial variations have disappeared. The molecular gas shows a surprisingly narrow range of values from 30 - 1000 \\Htwo\\ molecules cm$^{-3}$ for the Edmunds \\& Pagel type metallicity gradient, with no clear trend from the inner \\HI\\ --deficient regions near $R_G \\lesssim 5$ kpc, through the \\HI\\ --rich ``main body'' of the galaxy near $R_G \\approx 15$ kpc, all the way out to the \\HI\\ --poor outer parts at $R_G \\gtrsim 25$ kpc. We note that the Edmunds \\& Pagel type metallicity gradient is a good representation for the mean of an ensemble of nearby galaxies \\citep{pan00}, which leads us to favor Figure \\ref{fig:tau1} as the most likely final result. One concern is that the narrow range of values in Figure \\ref{fig:tau1} could result from observational selection. For example, regions of high volume density located near typical FUV sources may be missed since the \\HI\\ would have a relatively low column density and a low filling factor. This is not an issue for this study since only 3\\% of our initial sample of FUV sources were rejected due to a lack of 21 cm line emission. To confirm this statement, we examined an additional 10 distinct FUV sources. Measurable \\HI\\ is found within 800 pc for all ten sources. This is consistent with the spatial correlation between FUV sources and 21 cm line emission observed in Figures \\ref{fig:overlay2}, \\ref{fig:fuvhi2}, and \\ref{fig:fuvhi3}. Virtually all FUV sources are associated with 21 cm line emission, with the exception of the innermost regions of M101, where the \\HI\\ column density is supressed due to the strong increase in the dust-to-gas ratio. What kind of observational selection could operate on the low-density side of Figure \\ref{fig:tau1}? According to Equation \\ref{eqn:nhi}, \\Htwo\\ with low volume density can produce large columns of \\HI\\ which may become optically thick, leading us by Equation \\ref{eqn:voldens} to overestimate the \\Htwo\\ density and thus depopulate the lower part of Figure \\ref{fig:tau1}. However, gas of density e.g.~$n \\lesssim 10$ cm$^{-3}$ with \\HI\\ columns in the optically-thick regime of N(\\HI) $\\approx 10^{22}$ cm$^{-2}$ would be typically $\\gtrsim 300$ pc in size. Examination of Figure \\ref{fig:overlay2} and the zoomed Figures \\ref{fig:fuvhi2} and \\ref{fig:fuvhi3} shows that the highest-brightness \\HI\\ regions are unresolved, however. NGC~5447 in Figure \\ref{fig:fuvhi3} is a good example of this, as are NGC~5461 and NGC~5462 in Figure \\ref{fig:fuvhi2}. This means the path lengths are $\\lesssim 200$ pc, which implies values of N(\\HI) $\\lesssim 6 \\times 10^{21}$ cm$^{-2}$ in Figure \\ref{fig:nhichi}, where optical depth in the \\HI\\ is not likely to be a problem. Therefore, the paucity of points at low volume densities in Figure \\ref{fig:tau1} is very likely real. We conclude that the 35 regions studied do indeed comprise a representative sample of PDRs, and that observational selection is not likely to be the cause of the narrow range of values for $n$(\\Htwo) in Figure \\ref{fig:tau1}. Our results refer only to the ISM in the immediate vicinities of young stars and clusters of young stars. We do not know from the present observations if the results are representative of the ISM as a whole. Furthermore, the measurement of the \\HI\\ emission does not immediately provide a total gas mass without additional assumptions about the geometry of the ISM since the \\HI\\ emission is a surface phenomenon in the PDR picture. Further discussion of the implications of our results for the ISM content of galaxies in general is beyond the scope of this paper; we hope to return to these questions in a future publication. \\citet{wal96} have proposed an interpretation for the radial distribution of the bright \\HI\\ peaks in M101 and other nearby spirals in terms of variations in the hydrostatic pressure of the ISM. Their picture is insightful and deserves a more thorough discussion in the light of the present results than space here permits. UV photons clearly dominate the physics of the ISM in the immediate neighborhoods of young stars. The production of \\HI\\ from \\Htwo\\ by photodissociation is a natural and inevitable consequence of this physics. We have described how photodissociation can explain the geometrical structure of the \\HI\\ on scales of $\\sim 100$ pc in nearby galaxies and have discussed how the physics of photodissociation can be used to provide a new probe of the density of the underlying molecular hydrogen. In addition, we have presented a straightforward explanation for the disappearance of \\HI\\ in the inner parts of galaxies. Acceptance of this picture requires a shift in parts of the paradigm for star formation from the ISM, from viewing \\HI\\ as a precursor, to seeing it as a \\textit {product} of the star formation process. While the present results do not by themselves demand discarding the old paradigm, they further strengthen the viability of the new one." }, "0003/hep-th0003076_arXiv.txt": { "abstract": "{ We study the long distance behaviour of brane theories with quasi-localized gravity. The 5D effective theory at large scales follows from a holographic renormalization group flow. As intuitively expected, the graviton is effectively four dimensional at intermediate scales and becomes five dimensional at large scales. However in the holographic effective theory the essentially 4D radion dominates at long distances and gives rise to scalar anti-gravity. The holographic description shows that at large distances the GRS model is equivalent to the model recently proposed by Dvali, Gabadadze and Porrati (DGP), where a tensionless brane is embedded into 5D Minkowski space, with an additional induced 4D Einstein-Hilbert term on the brane. In the holographic description the radion of the GRS model is automatically localized on the tensionless brane, and provides the ghost-like field necessary to cancel the extra graviton polarization of the DGP model. Thus, there is a holographic duality between these theories. This analysis provides physical insight into how the GRS model works at intermediate scales; in particular it sheds light on the size of the width of the graviton resonance, and also demonstrates how the holographic RG can be used as a practical tool for calculations. } \\preprint{{\\tt hep-th/0003076} \\\\ HUTP-00/A008} \\begin{document} ", "introduction": "Randall and Sundrum (RS) have recently shown that it is possible to localize gravity to a brane in five dimensional anti-de Sitter space \\cite{RS,RS2}. In this model the theory on the brane reproduces four dimensional Einstein gravity at large distances even though the size of the extra dimension is infinitely large. This idea has sparked a flurry of activity in this field \\cite{Nima,CS, junction,cvetic,Gremm,US,Gremm2,Linde,skenderis,Kostas,CHR,GT,GKR, Japanese,CGS2,other} (see \\cite{noncompact,oldCvetic} for related earlier work). Gregory, Rubakov and Sibiryakov (GRS) \\cite{GRS} proposed a modified version of the RS model in which gravity appears five dimensional both at short and at long distances (see \\cite{Kogan} for a related idea), while at intermediate scales four dimensional gravity is reproduced. The GRS model has a positive tension brane, as in the RS model, and negative tension branes at a large proper distance away from the positive tension brane on either side of it. These negative tension branes have half the tension of the central brane. The brane tensions and the negative bulk cosmological constant are tuned such that the background is static, and the geometry between the branes is a slice of AdS$_5$, while beyond the negative tension branes the spacetime is ordinary 5D Minkowski space. The fact that 4D gravity is reproduced at intermediate scales has been explained in \\cite{GRS,CEH,Dvali}, and the reason for this is that in this model the 4D graviton is replaced by a resonance with a finite lifetime, which can decay into the bulk; thus gravity is only quasi-localized. It was also suggested in \\cite{Dvali,Witten} that there may be a connection in these theories to bulk supersymmetry and vanishing of the cosmological constant. It has been shown in \\cite{CEH2,GRS2}, that at intermediate distance scales the theory indeed reproduces the results of ordinary general relativity due to the bending of the brane in the presence of inhomogenous matter on the brane. The reason for this is that the effect of the bending of the brane exactly cancels the effects of the extra polarization in the massive graviton propagator (up to corrections that can be made arbitrary small by adjusting the width of the resonance). We should stress that one of the essential features of these models is that they do not have a 4D low-energy effective field theory description (see also \\cite{Witten}). Instead of an effective 4D theory, these models have a an effective 5D theory at large distances which can be derived through a holographic renormalization group (RG) flow. This holographically renormalized theory will play the role of the low-energy effective theory. In order to simply perform a calculation at a given energy scale on the brane one performs the RG running to that scale in the theory on the brane. From the conjectured AdS/CFT correspondence the RG flow corresponds to moving the brane a finite distance into the bulk. This procedure is referred to as the holographic RG and will be the key to understanding the theory at large distances. At intermediate energies, the RG flow corresponds to moving the brane inside the AdS slice. Since an exactly AdS bulk corresponds to a conformal field theory, the brane tension of the effective theory remain unchanged. However, at distances large enough that effectively the branes have crossed, the effective low-energy theory will be that of a tensionless brane in 5D Minkowski space. This procedure can also be implemented for models which are smooth versions of the GRS model (see \\cite{CEH}), with the difference that there will be a continuous running in the brane tension of the theory. However, as long as the asymptotic metrics are equivalent to those of GRS the asymptotic form of the low energy effective theory will be the same as for GRS. In fact, a detailed analysis below will show that one finds that at intermediate distances (more precisely at a scale $k^{-1} e^{ky_0}$, where $y_0$ is the location of the negative tension brane in the GRS model) the model becomes equivalent to the model recently proposed by Dvali, Gabadaze and Porrati \\cite{DGP3}, where a tensionless brane is embedded into 5D Minkowski space, but there is an additional induced four dimensional curvature term on the brane present. The induced operator is a consequence of the holographic renormalization. In addition, the radion mode of the GRS model (which corresponds to fluctuations of the distance between the two branes) will also be localized on the tensionless brane, as predicted in \\cite{PRZ}. This radion field will have a wrong-signed kinetic term, which is needed to cancel the effects of the extra graviton polarization in the DGP model. However, at very long distances, where the graviton mode becomes 5 dimensional, the radion will start to dominate and give rise to a peculiar 4D scalar antigravity, as discussed in \\cite{GRS2,PRZ}. Due to the negative kinetic term of the radion these theories are probably not internally consistent at large scales; but from a purely phenomenological point of view it is still interesting to use these results to see how the cosmology of these models deviates from the ordinary FRW expansion of the Universe at large scales. The paper is organized as follows: in Section 2 we explain the basic idea behind the holographic renormalization group, and calculate the effective brane tension and induced curvature term on the brane. In Section 3 we review the calculation of the induced radion kinetic term on the effective brane. In Section 4 we use the effective holographic theory obtained in Sections 2 and 3 to calculate the graviton propagator at large distances. We speculate on the cosmology of these models in Section 5, and conclude in Section 6. ", "conclusions": "We have examined a class of 5D metrics with embedded 3-branes. Along the extra dimension these models are asymptotically Minkowski, and gravitons are quasi-localized on a brane. One would expect that at sufficiently large distances the details near the brane are irrelevant and effectively there is a tensionless brane in 5D Minkowski space (a brane with non-zero tension would curve the space around it). We find that a holographic renormalization group analysis confirms this intuitive picture for the graviton. The renormalization group analysis also shows how the radion effectively gives rise to scalar ``anti-gravity'' at long distances by an induced radion coupling on the brane with negative kinetic term. In fact, at intermediate and large distances the holographic effective theory is equivalent to the recently proposed model of Dvali et al., where the tensionless brane in 5D Minkowski space also has an induced 4D curvature term on the brane. Thus the behavior of quasi-localized gravity in GRS-type models at different length scales is as follows: at very short distances the theory is five dimensional (both scalar potential and tensor structure); at intermediate scales it is given by ordinary 4D gravity with corrections that can be arbitrarily small; and at ultra-large distances the graviton is again five dimensional and the 4D radion dominates. Thus these models do not seem to be internally consistent; however, if a generalized model could eliminate the radion from the light degrees of freedom (as can happen in RS models) they might produce viable cosmologies which decelerate after a late epoch. To be consistent with current observations this epoch must be later than the current epoch." }, "0003/astro-ph0003327_arXiv.txt": { "abstract": "We present CCD photometry and proper motion studies of the two open star clusters NGC\\,1960 (M\\,36) and NGC 2194. Fitting isochrones to the colour magnitude diagrams, for NGC 1960 we found an age of $t=16$ Myr and a distance of roughly $d=1300$ pc and for NGC 2194 $t=550$ Myr and $d=2900$ pc, respectively. We combined membership determination by proper motions and statistical field star subtraction to derive the initial mass function of the clusters and found slopes of $\\Gamma=-1.23 \\pm 0.17$ for NGC\\,1960 and $\\Gamma=-1.33 \\pm 0.29$ for NGC\\,2194. Compared to other IMF studies of the intermediate mass range, these values indicate shallow mass functions. ", "introduction": "The shape of the initial mass function (IMF) is an important parameter to understand the fragmentation of molecular clouds and therefore the formation and development of stellar systems. Besides studies of the Solar neighbourhood (Salpeter \\cite{salpeter}, Tsujimoto et al. \\cite{tsuji}), work on star clusters plays a major role (Scalo \\cite{scalo1}) in this field, as age, metallicity, and distance of all stars of a star cluster can generally be assumed to be equal. Restricted to certain mass intervals, the IMF can be described by a power law in the form \\begin{equation} \\mbox{d} \\log N(m) \\sim m^\\Gamma \\mbox{d} \\log m. \\end{equation} In this notation the ``classical'' value found by Salpeter (\\cite{salpeter}) for the Solar neighbourhood is $\\Gamma=-1.35$. Average values for $\\Gamma$ from more recent studies, mostly of star clusters, can be found, e.g., in Scalo (\\cite{scalo2}): \\begin{eqnarray} \\Gamma=-1.3 \\pm 0.5 & \\mbox{\\ for } & m > 10 M_\\odot, \\nonumber \\\\ \\Gamma=-1.7 \\pm 0.5 & \\mbox{\\ for } & 1 M_\\odot < m < 10 M_\\odot, \\mbox{\\ and}\\\\ \\Gamma=-0.2 \\pm 0.3 & \\mbox{\\ for } & m < 1 M_\\odot, \\nonumber \\end{eqnarray} where the ``$\\pm$'' values refer to a rough range of the slopes derived for the corresponding mass intervals, caused by empirical uncertainties or probable real IMF variations. Knowledge of membership is essential to derive the IMF especially of open star clusters, where the contamination of the data with field stars presents a major challenge. Two methods for field star subtraction are in use nowadays: separating field and cluster stars by means of membership probabilities from stellar proper motions on one hand, statistical field star subtraction on the other hand. Our work combines these two methods: The proper motions are investigated for the bright stars of the clusters, down to the completeness limit of the photographic plates used, whereas the fainter cluster members are selected with statistical considerations. From the cleaned data we derive the luminosity and mass functions of the clusters. Including the proper motions, we expect to receive a more reliable IMF, since the small number of bright stars in open clusters would lead to higher uncertainties, if only statistical field star subtraction were applied. This is the second part of a series of studies of open star clusters, following Sanner et al. (\\cite{n0581paper}). Here we present data on two clusters of the northern hemisphere, NGC\\,1960 (M\\,36) and NGC\\,2194. NGC\\,1960 (M\\,36) is located at $\\alpha_{2000}=5^{\\rm h} 36^{\\rm m}6^{\\rm s}$, $\\delta_{2000}=+34^\\circ 8 \\arcmin$ and has a diameter of $d=10 \\arcmin$ according to the Lyng{\\aa} (\\cite{lynga}) catalogue. Morphologically, NGC\\,1960 is dominated by a number of bright ($V \\ga 11 \\mbox{\\ mag}$) stars, whereas the total stellar density is only marginally enhanced compared to the surrounding field. The cluster has not yet been studied by means of CCD photometry. Photographic photometry was published by Barkhatova et al. (\\cite{barkhatova}), photoelectric photometry of 50 stars in the region of the cluster by Johnson \\& Morgan (\\cite{johnsmorg}). The most recent proper motion studies are from Meurers (\\cite{meurers}) and Chian \\& Zhu (\\cite{chianzhu}). As their epoch differences between first and second epoch plates (36 and 51 years, respectively) are smaller than ours and today's measuring techniques can be assumed to be more precise we are confident to gain more reliable results. Tarrab (\\cite{tarrab}) published an IMF study of 75 open star clusters, among them NGC\\,1960, and found an exteme value for the slope of (in our notation) $\\Gamma=-0.24 \\pm 0.05$ for this object. Her work includes only 25 stars in the mass range $3.5 M_\\odot \\la m \\la 9 M_\\odot$, so that a more detailed study covering more members and reaching towards smaller masses is necessary. For NGC\\,2194 (located at $\\alpha_{2000}=6^{\\rm h} 13^{\\rm m}48^{\\rm s}$, $\\delta_{2000}=+12^\\circ 48 \\arcmin$, diameter $d=9 \\arcmin$), our work is the first proper motion study according to van Leeuwen (\\cite{vanleeuwen}). The {\\it RGU} photographic photometry of del\\,Rio (\\cite{delrio}) is the most recent publication on NGC\\,2194 including photometric work. The cluster is easily detectable as it contains numerous intermediate magnitude ($13 \\mbox{\\ mag} \\la V \\la 15 \\mbox{\\ mag}$) stars, although bright stars $V \\la 10 \\mbox{\\ mag}$ are lacking. In Sect. \\ref{cadata}, we present the data used for our studies and the basic steps of data reduction and analysis. Sects. \\ref{n1960disc} and \\ref{n2194disc} include the proper motion studies, an analysis of the colour magnitude diagrams (CMDs), and determination of the IMF of the clusters. We conclude with a summary and discussion in Sect. \\ref{caconcl}. ", "conclusions": "\\label{caconcl} With our work we found NGC\\,1960 to be a young open star cluster with an age of 16 Myr. It is located at a distance of 1300 pc from the Sun. These results confirm the findings of Barkhatova et al. (\\cite{barkhatova}) obtained with photographic photometry. We derived proper motions of 404 stars in the region of the cluster down to $14 \\mbox{\\ mag}$. 178 of those can be considered members of NGC\\,1960. Despite the problems with our proper motion determination (see Sect. \\ref{n1960eb}), we are able to state that our results do not support the values given as the absolute proper motion of NGC\\,1960 by Glushkova et al. (\\cite{glush}) on the base of the ``Four Million Star Catalog'' (Gulyaev \\& Nesterov \\cite{4M}): They found $\\mu_\\delta=-8.2 \\pm 1 \\mbox{\\ mas yr}^{-1}$ which is in agreement with our study, but $\\mu_\\alpha \\cos \\delta=14.7 \\pm 1 \\mbox{\\ mas yr}^{-1}$ which differs from our result by more than $10 \\mbox{\\ mas yr}^{-1}$. Our study of the IMF of NGC\\,1960 led to a power law with a slope of $\\Gamma=-1.23 \\pm 0.17$. This value is very high (i.e. the IMF is shallow) compared to other studies, however, it still matches the interval for $\\Gamma$ suggested by Scalo (\\cite{scalo2}) for intermediate mass stars ($-2.2 \\leq \\Gamma \\leq -1.2$). Although we should stress that we cannot say anything about the shape of the IMF in the very low mass range ($m \\ll M_\\odot$), we do not see any evidence for a flattening of the IMF of NGC 1960 below $1 M_\\odot$. NGC\\,2194 --- with an age of 550 Myr --- belongs to the intermediate age galactic open star clusters. Our findings from the photometric study are in good agreement with the photographic $RGU$ photometry published by del\\,Rio (\\cite{delrio}). As the cluster is located at a distance of almost 3 kpc we could only cover its mass spectrum down to $1 M_\\odot$. Nevertheless, we were able to determine the IMF on the base of 623 main sequence stars which led to a slope of $\\Gamma=-1.33 \\pm 0.29$, almost Salpeter's (\\cite{salpeter}) value, but still close to the shallow end of the interval given by Scalo (\\cite{scalo2}). In our previous paper (Sanner et al. \\cite{n0581paper}), we studied the open star cluster NGC\\,581 (M\\,103) for which we found the same age of $16 \\pm 4$ Myr as for NGC\\,1960, but a much steeper IMF slope of $\\Gamma=-1.80 \\pm 0.19$. We therefore can state that our method of IMF determination does not systematically lead to steep or shallow mass functions. With our yet very small sample, it is not possible to find evidence for the dependence of the IMF of open star clusters on any parameter of the cluster. Therefore, we will have to investigate further clusters and also compare our results with other studies." }, "0003/astro-ph0003261_arXiv.txt": { "abstract": "Pre-white dwarf (PWD) evolution can be driven by energy losses from neutrino interactions in the core. Unlike solar neutrinos, these are not the by-product of nuclear fusion, but instead result from electron scattering processes in the hot, dense regions of the PWD core. We show that the observed rate of period change in cool PWD pulsators will constrain neutrino emission in their cores, and we identify appropriate targets for future observation. Such a measurement will tell us whether the theories of lepton interactions correctly describe the production rates and therefore neutrino cooling of PWD evolution. This would represent the first test of standard lepton theory in dense plasma. ", "introduction": "In general, stars are too remote---and observables too few---to make them practical experimental physics test-beds: our data are spent in simply describing the dimensions of the objects under study. In many cases we must extrapolate experimental data over many orders of magnitude, or resort to untested calculations from first principles, to reach the regions of phase space that apply to stellar interiors. If we hope to overcome these problems and pursue ``experimental'' astrophysics, we can either attempt to increase the number of observables or find simpler stars. As first realized by Mestel (1952), the evolution of white dwarfs and pre-white dwarfs (PWDs) is primarily a simple cooling problem. In general, our growing understanding of white dwarf interiors and evolution has paralleled advances in the theory of dense plasmas, with the recognition of important influences like electron degeneracy (Chandrasekar 1939), Coulomb interactions (Salpeter 1961), crystallization (Kirzhnitz 1960; Abrikosov 1960; Salpeter 1961; Stevenson 1980) and neutrino cooling effects (Chin, Chiu, \\& Stothers 1966; Winget, Hansen, \\& Van Horn 1983; Kawaler, Hansen, \\& Winget 1985). Iben \\& Tutukov (1984) summarize the various mechanisms which dominate white dwarf evolution from the planetary nebula nucleus (PNN) stage to the coolest white dwarfs. On the observational side, the discovery of white dwarf pulsation in the 1960s, and pre-white dwarf pulsation in the 1970s, greatly increased the observable parameters available for comparison with theoretical models. These are short period, multiperiodic, $g$-mode variables, showing anywhere from a few to over a hundred separate periodicities on timescales of 100-3000~s. The pulsating PWD stars are divided into two classes: the planetary nebula nucleus variables (PNNV stars), and the slightly more evolved GW~Virginis (or simply GW~Vir stars) which lack observed nebulae. With high surface gravities (log~$g \\sim 6-7.5$), and effective temperatures between 80,000~K and 170,000~K, they occupy a region of the H-R diagram between the high-\\teff~end of the PNN branch and the top of the white dwarf cooling track. There are eight known PNNV stars, and four GW~Vir stars (Ciardullo \\& Bond 1996). The evolutionary timescale of PWD stars is of order $10^{6}$ years. During this short transition from PNN star to hot white dwarf, stellar radius and photon luminosity decrease by one and three orders of magnitude, respectively. High core density and temperature allow electron scattering processes to produce a large neutrino flux which remains roughly constant during this time. As photon luminosity plummets, neutrinos contribute an increasing fraction of the total energy losses. Neutrino emission eventually comes to dominate the overall evolution of the star. Unlike photon energy, which must diffuse relatively slowly through the entire star before emerging into space, neutrinos created near the center of the PWD escape directly. This neutrino luminosity cools the center of the star, maintaining a temperature inversion similar to that within stars at the tip of the red giant branch. Calculations of the relevant reaction rates were performed initially by Beaudet, Petrosian, \\& Salpeter (1967) based on the theory of weak interactions proposed by Feynman \\& Gell-Mann (1958). Later, Dicus (1972) and Dicus et al.~(1976) recalculated these rates in the unified electroweak theory of Weinberg and Salam (Weinberg 1967, Salam 1968). All of these calculations are theoretical, however. We have no direct experimental or observational confirmation of neutrino production rates under conditions appropriate to PWD interiors. The cooling of a GW~Vir interior tends to increase the periods of each given pulsation mode. Their high luminosity (log~$L \\sim 0 - 3$) means they cool much more rapidly than cooler white dwarf variables. GW~Vir period changes are therefore expected to be more rapid also. Winget, Hansen \\& Van Horn (1983) show that the $e$-folding time for period changes in GW~Vir stars should be of the same order as the evolutionary timescale---$10^{6}$ years; such rapid changes are measurable in $1-3$ years time. This is an exciting prospect: to measure directly, on human timescales, the rate of evolution of a star, and specifically to place strict constraints on the mechanisms which regulate the evolution of a stellar interior. Over 30 years ago, Chin, Chiu, and Stothers (1966) predicted that at some point in PWD evolution neutrino losses should dominate all other cooling processes. Asteroseismological analysis can tell us which stars these are, and then measurement of period changes can tell us if our neutrino physics is right. Such a test has implications far beyond the study of PWD evolution. For instance, one of the fundamental questions of stellar astrophysics is the length of time stars spend on the main sequence. Answering this question requires precise knowledge of the p-p and CNO nuclear reaction rates. Currently, the best laboratory for measuring these rates is our own Sun, since terrestrial labs cannot in general reproduce the conditions of the stellar interior. However, models which successfully reproduce the known structure of the Sun predict a neutrino flux two to three times that measured by earthly detectors (Bahcall \\& Pinsonneault 1996, and references therein). For a long time, it was thought the problem might reside in our incomplete knowledge of conditions in the solar interior. Recently, helioseismology projects such as the Global Oscillation Network Group (GONG) have resulted in the measurement of millions of solar pulsation frequencies (Harvey et al.~1996). With so many parameters to constrain model properties, the possibility that the solar neutrino problem can be solved through variations in the thermodynamics or mechanics seems to be excluded (Bahcall \\& Pinsonneault 1996). The problem, then, almost certainly lies with the way we handle the nuclear physics. Under the most intense scrutiny is the standard theory of lepton interactions. Our calculations of neutrino emission from PWDs are based on this same theory. In PWDs, however, the energy loss rate due to neutrinos is thousands of times greater than in the Sun. Measurement of the effects of neutrino interactions in PWDs would afford a critical independent test not only of the standard lepton theory but also of non-standard theories brought forward to solve the solar neutrino problem. To explore this possibility, we calculated PWD evolutionary tracks using different neutrino production rates. In the next section we describe the calculation of those rates and summarize the basic interactions that lead to neutrino emission in PWD interiors. Section~3 describes PWD sequences with varied neutrino production rates and examines effects on measurable quantities such as $T_{\\rm eff}$, surface gravity, and rate of period change. Finally, in \\S~4 we discuss prospects for placing observational constraints on neutrino physics, and we identify appropriate targets for future observation. ", "conclusions": "We have shown that the predicted rates of period change in GW~Vir stars near the cool end of the instability strip are sensitive to the neutrino production rates used in stellar models. The persistence of the solar neutrino problem has made the standard model of neutrino interactions one of the most intensely scrutinized theories in all of physics. Determination of $\\dot{\\Pi}$ in the GW~Vir stars PG~0122 and PG~2131 will provide an important test of the standard model and of any new theories put forward to replace it." }, "0003/astro-ph0003057_arXiv.txt": { "abstract": " ", "introduction": "Hard X-rays have opened a new window to see the Galactic center (GC) region. One of the remarkable discoveries is that a high-temperature ($\\sim$ 10~keV) thin-thermal plasma is prevailing over an $\\sim$ 100 pc-radius region around the GC (Koyama et al.\\ 1986, 1989, 1996; Yamauchi et al.\\ 1990). The presence of the large-scale hot plasma suggests violent activities in the GC region in the past, although the origin is not yet clear. One of the relevant processes to produce such a high-temperature diffuse plasma would be multiple supernova explosions. In this context, X-ray observations of individual supernova remnants (SNRs) near to the GC may provide useful information not only on the SNR physics, but also on the origin of the GC plasma. Since X-rays from SNRs are usually dominated in the low-energy band ($\\leq 2$ keV), and are easily absorbed by interstellar gas, an X-ray study of the SNRs near to the GC region is rather limited. High sensitivity and imaging capability in the hard X-ray band are required to detect SNRs behind a large absorbing medium, and to separate individual SNRs from the GC X-ray emissions, a complex diffuse plasma, binary sources, and the other stellar objects. X-ray imaging spectroscopy also provides direct information on the nuclear synthesis, the total explosion energy and the age of SNRs, physical parameters of the surrounding interstellar environment, such as the density, its chemical compositions, and other related subjects: the star forming rate, the structure and the evolution in the central region of the Galaxy. The ASCA satellite, having high sensitivity in the hard X-ray band ($\\geq 2$ keV) and high energy resolution, enables us to study more elaborate imaging spectroscopy than was possible with previous detectors. We conducted survey and pointing observations near to the GC region with the ASCA satellite, and found X-rays from radio SNRs, including new candidates. Among them, this paper reports on the first detailed X-ray information and analyses of two radio SNRs, G~359.1$-$0.5 and G~359.0$-$0.9. G~359.1$-$0.5 was first identified as an SNR by a 4.9~GHz observation (Downes et al.\\ 1979) and by a 10.55~GHz observation (Sofue et al.\\ 1984). Uchida et al.\\ (1992b) found a shell-like structure at 1.4~GHz surrounded by a $^{12}$CO ring. Comparing the 21~cm absorption feature of the $^{12}$CO ring with the Galactic rotation curve, they concluded the location of this SNR to be near to the GC. Although and Egger, Sun (1998) discovered X-rays from G~359.1$-$0.5 with ROSAT, the spectral parameters, such as the temperature and the chemical composition, are not well constrained, due to the poor statistics and limited energy resolution. Preliminary results of the ASCA observation on G~359.1$-$0.5 are found in Yokogawa et al.\\ (1999). G~359.0$-$0.9 was first identified as an SNR by a 10.55~GHz observation (Sofue et al.\\ 1984) and by a 2.7~GHz observation (Reich et al.\\ 1990), and was later found to have an incomplete shell at the 843~MHz (Gray 1994). Leahy (1989) first detected a partial shell of soft X-rays from G~359.0$-$0.9 with the Einstein satellite, but no spectral information was reported. This paper presents more comprehensive ASCA results and analyses of these two SNRs. Particular care concerning the background subtraction was made to exclude any possible contamination of near-by bright X-ray sources and the GC plasma's contribution, of which the X-ray flux differs from position to position. We describe the observations and the method of data reduction in section~2, and the analyses in subsection~3.1 and subsection~3.2, for G~359.1$-$0.5 and G~359.0$-$0.9, respectively. Section~4 is devoted to results and discussions on the distances, chemical compositions and morphology of these SNRs, and also on a relevant subject, the origin of the GC plasma. ", "conclusions": "} \\subsection{G~359.1$-$0.5} G~359.1$-$0.5 is found to exhibit a large absorption column of $\\sim 5.9 \\times 10^{22}$ H cm$^{-2}$. Since G~359.1$-$0.5 is reported to be surrounded by the $^{12}$CO ring for a total mass of about $2.5\\times 10^6 M_{\\odot}$ (Uchida et al.\\ 1992b), local absorption due to the $^{12}$CO ring may not be ignored. Assuming that the $^{12}$CO ring is a homogeneous shell with nearly the same shape of the G~359.1$-$0.5 radio shell, the absorption column due to the $^{12}$CO ring is estimated to be $\\sim 3\\times 10^{22}$~H~cm$^{-2}$. Therefore, we infer that the column density of the foreground interstellar matter is about $\\sim 3\\times 10^{22}$~H~cm$^{-2}$. This value is equal to that of other X-ray sources near to the GC with the same Galactic coordinate of this SNR (Sakano et al \\ 1999b); hence, this SNR would really be located near to the GC region with a distance of about 8.5~kpc. Thus, the diameter of the radio shell is estimated to be $\\sim$ 57~pc, while that of the X-ray emitting central sphere is $\\sim$ 28~pc. We found that G~359.1$-$0.5 has at least two temperature plasmas; the cooler plasma (component 1) is abundant in Si, whereas the hotter one (component 2) is extremely over abundant in S. The center-filled thermal X-rays imply that these plasmas originated from the ejecta. Assuming a $\\sim$28~pc-dimameter of spherical plasma of uniform density with a filing factor of 0.1, we estimate the total mass of Si and S to be about $0.1$~$M_\\odot$ and $0.3$~$M_\\odot$, respectively. However, no current theory of nucleosynthesis in supernova explosions predicts such a large mass of S compared to Si (see e.g. Thielemann et al.\\ 1996). This problem can be solved by assuming a smaller filling factor of S than that of Si; a smaller filling factor of S, less than 0.1, reduces the total mass of S to be acceptable for the model of Thielemann et al.\\ (1996). The context of the very small filling factor and the extreme richness of S lead us to suspect that the S-rich plasma is a ``shrapnel'' ejected from the massive progenitor of G~359.1$-$0.5, in analogy of the Vela SNR (Aschenbach et al.\\ 1995). However, the narrow band image including only the S-line (2.1--3.2 keV) shows no spatial structure like a ``shrapnel'', mainly due to a lack of photon statistics. Uchida et al.\\ (1992a) argued that the $^{12}$CO ring surrounding the SNR shell was created by stellar winds and/or multiple supernovae of O-type stars, and that several radio sources clustered at the center of the SNR are possibly O-type stars. The X-ray spectrum of G~359.1$-$0.5 contains no clear emission line of Fe (see figure~2), and is thus consistent with the proposed O-star origin. However, we could not quantitatively predict on the mass of the progenitor because of a lack of photon statistics. We found no shell-like X-ray from G~359.1$-$0.5, although the radio morphology shows an almost complete shell. Shell-like X-rays may originate either by (1) a synchrotron mechanism in a shell, or (2) a thermal plasma made by the shock wave. The lifetime of X-ray emission in case (1) depends on the magnetic field. Generally, the GC region is known to have a stronger magnetic field than the other regions of our Galaxy. In particular, Robinson et al.\\ (1996) observed Zeeman effects of three OH masers near to the shell of G~359.1$-$0.5, then directly estimated the magnetic field to be 0.4--0.6 mG, which is more than one order of magnitude larger than that in usual shell-likes SNRs ($\\sim 10~\\mu$G). High-energy electrons to emit synchrotron X-rays have a lifetime of $\\sim 10^3$~yr in an $\\sim 10~\\mu$G field (e.g. Reynolds 1996); hence, those in the shell of G~359.1$-$0.5 should be much shorter than $\\sim 10^3$~yr. From the large diameter of the G~359.1$-$0.5 shell of 60~pc, this SNR would be middle age, or typically $\\gtsim$ $10^4$~yr. Thus, no synchrotron X-ray from the shell would be expected from this SNR. The evolved age of G~359.1$-$0.5 makes a shock-heated shell rather cool, with a typical temperature of less than a few 100~eV. Furthermore, if the shell interacts with the dense $^{12}$CO ring, the X-ray emitting shell density becomes larger, and hence the cooling time is much shorter. Therefore, X-rays should be very soft, and should be entirely absorbed by the large interstellar gas column. Thus, the apparent lack of an X-ray shell of G~359.1$-$0.5 would be reasonable. Rho and Petre (1998) proposed that SNRs with center-filled X-rays and a shell-like radio structure should be called ``mixed morphology'' (MM) supernova remnants. They show that the scale height of the MM SNRs' distribution from the Galactic plane is smaller than that of shell-like SNRs in both radio and X-ray bands, and that many MM SNRs are located near to molecular clouds or H~I clouds. G~359.1$-$0.5 is located near to the Galactic plane, and is surrounded by the $^{12}$CO ring (Uchida et al.\\ 1992b). Thus, G~359.1$-$0.5 shares the common features of MM SNRs. \\subsection {G~359.0$-$0.9} The best-fit column density of $\\sim 1.8\\times 10^{22}$ H cm$^{-12}$ is smaller than that of the GC region (Sakano et al.\\ 1999b). In fact, using the Galactic interstellar mass model by Olling and Merrifield (1998), and by Dame et al.\\ (1987), we can estimate the distance to be 6 kpc. Hence, G~359.0$-$0.9 would not be in the GC region, but would be a foreground SNR. The shell-like morphology and a thin-thermal spectrum in X-rays suggest that the shell is a shock-heated plasma. The plasma, however, shows a partial shell extending only 1/4 of a full circle. With the reasonable assumption that the supernova explosion was spherically symmetric, the apparent partial shell is attributable to an inhomogeneity of the interstellar medium; the direction to the X-ray emitting shell would have a higher density than the other directions, and would hence have a larger surface brightness than the others. For simplicity, we assume that the dense region covers $\\pi /2$ str toward the partial shell; thus, 1/8 of the total explosion energy would be given in this direction. Using the 6-kpc distance, the diameter and X-ray luminosity of the partial shell are estimated to be 38 pc and $1\\times 10^{34}$erg s$^{-1}$, respectively. Together with the observational temperature of 0.4~keV, we could solve the Sedov equation, and found that the total explosion energy ($E$), density ($n$) of the interstellar medium (dense region) and the age ($t$) are $1.2\\times 10^{51}$~erg, 0.5~cm$^{-3}$, and 1.8$\\times10^4$~yr, respectively. Because the total mass of the X-ray emitting shell is about 40$M_\\odot$, most of them are attributable to the swept-up interstellar matter. Therefore, the result of the spectral fitting (in table~3) implies that the interstellar matter near to the G~359.0$-$0.9 is either over-abundant in Mg, or under-abundant in Fe. \\subsection {Comments on the Galactic Diffuse Plasma} The GC region is surrounded by the large scale hot plasma, which emits fairly strong X-rays with many emission lines from He-like and H-like Si, S, Ar, Ca and Fe. The co-existence of highly ionized light elements (such as Si and S) and heavier elements (such as Ca and Fe) implies that the plasma is not a single temperature. Kaneda (1996) and Kaneda et al.\\ (1997) confirmed the two-temperature structure of the Galactic ridge plasma, which shows a very similar spectrum to that of the GC region. In fact, Maeda(1998) found that the GC plasma has two-temperature components. The low-temperature component ($\\ltsim$1~keV) would be the same as that found with ROSAT (Snowden et al.\\ 1997). The plasma has a large-scale height of 1.9~kpc and a temperature of 0.3$-$0.4~keV. Kaneda et al.\\ (1997) suggested that the soft components of the Galactic ridge originated from a multiple supernova explosion. From the spectral similarity, this scenario may be applied to the GC soft component. Then, can we find many individual SNRs which have a similar spectral shape? Because G~359.0$-$0.9 has a 0.4~keV temperature and strong emission line of Mg and Si, it is a possible candidate. The GIS flux at 0.5--10.0~keV of G~359.0$-$0.9 is $2.4\\times 10^{-12}$~ erg~s$^{-1}$cm$^{-2}$, while that of the GC plasma in a 25$^\\prime$ diameter field is $8.3\\times 10^{-12}$~erg~s$^{-1}$cm$^{-2}$ (Kaneda 1996). We thus require about four G~359.0$-$0.9-like SNRs in this field. However, at this moment, the number density of the resolved X-ray SNRs or its candidates is far less than the requirement. A more difficult issue is the origin of the high-temperature component. The temperature of 10~keV and the size of about 100-pc diameter are inferred by the observation of Ginga and ASCA GC surveys (Koyama et al.\\ 1989; 1996). As far as we know, no SNR exhibits such a high temperature. In this sense, the high-temperature sulfur-rich plasma of G~359.1$-$0.5 is suggestive. If there are many G~359.1$-$0.5-like SNRs, with an enriched abundance of not only S, but also other elements in high-temperature plasmas, the integrated emission may contribute, at least, some fractions of the GC hot plasma. At present, we are still lacking any quantitative information of individual X-ray sources, like SNRs, near to the GC region. Thus, further systematic study is highly encouraged.\\par \\vspace{1.5pc}\\par We would like to thank Dr.Y.\\ Maeda for useful comments and discussion about G~359.1$-$0.5. We also thank the members of the ASCA team. This work was supported by the Research Fellowship of the Japan Society for the Promotion of Science for Young Scientists. \\small" }, "0003/astro-ph0003460_arXiv.txt": { "abstract": "We report the results of a long-term campaign of radio, soft- and hard- X-ray observations of the galactic black hole candidate \\gx. In the Low-Hard X-ray state the system displays a strong 3-way linear correlation between soft- and hard-X-rays and radio emission, implying a coupling between the Comptonising corona and a radio-emitting compact jet. In this state the radio emission is linearly polarised at a level of around 2\\%, with an almost constant polarisation angle, indicative of a favored axis in this system probably related to the compact jet and/or black hole spin axis. In the Off X-ray state the radio emission declines with the X-ray emission to below detectable levels, suggesting that it is simply a lower-luminosity version of the Low-Hard state. In the High-Soft state both the hard-X-ray and radio emission are suppressed. We also note that the transitions from the Low-Hard state to the High-Soft state (and the reverse) are possibly associated with discrete ejection(s) of expanding relativistic plasma. ", "introduction": "The X-ray source \\gx\\ is believed to harbor a black hole, based on the similarity of its X-ray spectral and timing properties with the black hole system \\cx\\ (e.g. Samimi et al. 1979; Maejima et al. 1984; Ilovaisky et al. 1986; Makishima et al. 1986). Black hole candidates are known to exhibit transitions between various X-ray states, distinguished by their different spectral and timing properties. Five distinct X-ray states have been reported, namely the Off state, the Low-Hard state, the Intermediate state, the High-Soft state and the Very High state. See Tanaka \\& Lewin (1995) for a review of these properties in the Low-Hard and High-Soft states. \\gx\\ and GS~1124--683 (\\cite{ebi94}) are the only X-ray sources observed in all these states [see Miyamoto et al. (1991) for the Very High state and M\\'endez \\& van der Klis (1997) for the Intermediate state in \\gx]. Two new X-ray transient sources have also been observed in the Very High state, i.e XTE~J1748--288 (Revnivtsev et al. 2000) and XTE~J1550--564 (e.g. Sobczak et al. 1999, Homan et al. 2000). The superluminal sources GRO~J1655--40 and GRS~1915+105 may also display behaviour related to X-ray spectral state changes (e.g. M\\'endez et al. 1998, Belloni et al. 2000). Spectral and timing properties of \\gx\\ in the Low-Hard state can be found in Wilms et al. (1999) and Nowak et al. (1999). In the Low-Hard state, a peak in the low-frequency power spectrum is seen at a period of 240 days, and is possibly related to a precessing accretion disk (Nowak et al. 1999). Black hole candidates can be divided into two different classes of sources: the soft X-ray transients (SXTs) and the persistent black hole candidates. SXTs (\\cite{cha98}) are, by definition, transient and are usually associated with a low mass and late type stellar companion. \\gx\\ is persistent in soft-X rays (as are the other galactic black hole candidates \\cx, 1E~1740.7--2942 and GRS~1758--258), but at hard X-rays \\gx\\ shows behaviour resembling that of the transient sources (\\cite{har94,rub98}). We note that LMC~X--1 and LMC~X--3 can also be considered as persistent black hole candidates in the Large Magellanic Cloud (\\cite{tan95}). Due to the faintness of the stellar companion relative to the accretion disk, the spectral type of the stellar companion of \\gx\\ has not been identified, but it is believed to be a low mass star with a possible orbital period of 14.8 hours (\\cite{cow87,cal92}). Soria et al. (1999) suggested the true orbital period may be twice this value. The distance to \\gx\\ is about 4 kpc (\\cite{zdz98}) with an optical extinction of 3.5 magnitudes (\\cite{cor99}). Optical emission in \\gx\\ also displayed the state transitions (e.g. \\cite{mot85}). \\gx\\ is usually bright and variable (V $\\approx$ 15-17 mag.) in the Low-Hard state, faint in the Off state (V $\\approx$ 20 mag.) and at an intermediate level in the High-Soft state (V $\\approx$ 16.5 mag.). Most of the optical emission is believed to be dominated by the accretion flow, but the physical process behind its origin has not been fully understood (\\cite{fab82,mot83}, Ilovaisky et al. 1986). Although \\gx\\ has been studied extensively at high energies, little has been known until recently about its properties in the radio regime. It was discovered as a radio source in 1994 (Sood \\& Campbell-Wilson 1994) and a possible jet like feature has been reported by Fender et al. (1997a). In the Low-Hard state, Hannikainen et al. (1998) found a correlation between the radio, soft and hard X-rays emission on timescale of the order of 5 days over a period of one year. During the High-Soft state, we have shown that the radio emission from \\gx\\ disappeared until the return to the standard Low-Hard state (\\cite{fen99b}). This behaviour is reminiscent of that observed in the 1996 High-Soft (or Intermediate) state of \\cx\\ (Zhang et al. 1997). This paper presents the results from a long term campaign of radio observations of \\gx\\ with the Australia Telescope Compact Array and the Molonglo Observatory Synthesis Telescope. Following the description of the characteristics of radio emission (light curve, spectra, variability and polarisation) from \\gx, we present evidence for the existence of a compact jet in this system. We then focus on the behaviour of \\gx\\ in the various X-ray states, as observed in radio, soft and hard X-rays. We demonstrate a strong coupling in the Low-Hard state of the compact jet with the Comptonising corona. ", "conclusions": "The following list summarizes the conclusions we have been able to draw from this multi-wavelength analysis of the behaviour of \\gx: \\begin{itemize} \\item {In the Low-Hard X-ray state, we observe a very strong 3-way correlation between the radio, soft and hard X-rays emission, believed to be the result of a coupling between the Comptonising corona and a compact jet. Models of persistent accreting black holes would need to take into account this coupling in the Low-Hard state and the quenching of the radio jet in the High-Soft state. The Off X-ray state is consistent with being a lower-luminosity Low-Hard state, in all three energy bands.} \\item {The radio spectrum is flat or slightly inverted. A similar behaviour is observed in the black hole \\cx. The radio emission can be understood as synchrotron emission from a compact jet, with quasi-continuous injection of relativistic plasma. } \\item{Transitions from the Low-Hard state to the High-Soft state (or vice versa) are possibly associated with discrete ejection(s) of expanding relativistic plasma.} \\item{The pattern of behaviour observed in radio, soft- and hard-X-rays is sufficient to explain in broad terms all the observations over several years.} \\item{Linear polarisation has been detected from the radio emission of \\gx\\ with a nearly constant polarisation angle, which points to a favored axis in the system. This direction is almost certainly related to the direction of the outflow, and in turn to the inner disc and/or black hole rotation axes.} \\end{itemize}" }, "0003/astro-ph0003183_arXiv.txt": { "abstract": "We present a numerical Hydro+N-body model of A3266 whose X-ray surface brightness, temperature distribution, and galaxy spatial and velocity distribution data are consistent with the A3266 data. The model is an old ($\\sim$3 Gyr), off-axis merger having a mass ratio of $\\sim$2.5:1. The less massive subcluster in the model is moving on a trajectory from southwest to northeast passing on the western side of the dominant cluster while moving into the plane of the sky at $\\sim$45 degrees. Off-axis mergers such as this one are an effective mechanism for transferring angular momentum to the intracluster medium (ICM), making possible a large scale rotation of the ICM. We demonstrate here that the ICM rotation predicted by our fully 3-dimensional model of A3266 is observable with current technology. As an example, we present simulated observations assuming the capabilities of the high resolution X-ray spectrometer (XRS) which was to have flown on {\\it Astro-E}. ", "introduction": "A3266 is a nearby (z=0.059; Quintana, Ramirez \\& Way 1996), X-ray luminous cluster which exhibits optical and X-ray substructure. Two models have recently been proposed to explain the substructure in this cluster. Flores, Quintana, \\& Way (1999) have proposed that A3266 experienced a major merger {\\it into} the plane of the sky while Henriksen, Donnelly, \\& Davis (1999) have proposed a minor merger parallel to the plane of the sky. Using a large sample of galaxy redshifts (387 galaxies), Quintana \\etal (1996) suggested that A3266 might have experienced a merger 1-2 Gyr ago. Flores et al. (1999) found support for this interpretation using a simple N-body model. They noted an enhancement of galaxies north of the X-ray core similar, both visually and statistically, to the N-body particle spray found in their numerical simulations. They also noted a similar enhancement of emission-line galaxies in the same region. It has long been suggested that galaxies passing through cluster cores could be spectroscopically altered (Dressler \\& Gunn 1983), although it now appears that this would be mostly due to the tidal force rather than ram pressure by the ICM (\\eg Moore \\etal 1996; Bekki 1999; Fujita \\etal 1999). Burns \\etal (1994) have suggested that the Coma cluster E+A galaxies distributed in the core and SW toward the NGC 4839 group are the result of a burst of star formation induced by a merger about 2 Gyr ago, which appears consistent with their starburst age (Caldwell \\etal 1996). Similarly, the emission-line galaxies in A3266 could be the tail end of the disrupted less massive cluster in this model. Evidence of recent dynamical evolution is also apparent in the X-ray surface brightness (XSB; Fig. \\ref{a3266}) and temperature distributions. For example, the XSB was shown to exhibit a systematic centroid shift by Mohr, Fabricant \\& Geller (1993). Also, the XSB exhibits changing ellipticity and isophotal twisting between 4\\amin and 8\\amin (Mohr \\etal 1993) as well as a large ($\\sim$500 kpc) core radius (Mohr, Mathiesen \\& Evrard 1999), as would be expected in the case of a recent merger (Roettiger \\etal 1996). Peres \\etal (1998) find no evidence of a cooling flow. Several researchers (\\eg McGlynn \\& Fabian 1984) have suggested that mergers will disrupt cooling flows. G\\'omez \\etal (1999) have demonstrated using numerical simulations that the time scale for re-establishment of the disrupted cooling flow in the post-merger environment can be greater than several billion years, depending on the initial cooling flow and merger parameters. Markevitch \\etal (1998; hereafter MFSV98) and Henriksen \\etal (1999; hereafter HDD99) have produced temperature maps based on {\\it ASCA} data that show significant temperature variations across the cluster. MFSV98 shows a radially decreasing temperature profile ranging from 12 keV in the central 3\\amin to $\\sim$6 keV at radii greater than 8\\amin. De Grandi \\& Molendi (1999) find a similar radial temperature gradient using {\\it BeppoSAX} data. The HDD99 temperature map exhibits a comparable range in ICM temperatures with a gradient increasing from NE to SW across the cluster. In this paper, we extend the N-body model of Flores \\etal (1999) by including the hydrodynamics of the ICM. We then demonstrate using a fully 3-dimensional numerical model that the current A3266 data are consistent with an old off-axis merger occurring largely {\\it into} the plane of the sky. Off-axis mergers are a natural consequence of large-scale tidal torques, the latter being a generic feature of hierarchical clustering (Peebles 1969). The model of Flores et al. (1999) resulted in an off-axis merger as a result of the global angular momentum imposed at the protocluster stage, which in terms of the standard dimensionless angular momentum $\\lambda$ (Peebles 1969) corresponded to $\\lambda = 0.07$. This amount is consistent with tidal torquing and is expected to be largely independent of mass (Barnes \\& Efstathiou 1987). Therefore the characteristics of the merger are not sensitively dependent on extending the region that was simulated around the cluster. In sufficient quantity, angular momentum can significantly alter the internal structure of clusters, which can then influence our interpretation of other cluster observations. As an example, numerical simulations (Inagaki \\etal 1995; Roettiger \\etal 1997) have shown that the shape of clusters (\\eg oblateness being a consequence of rotation) can have significant systematic effects on determinations of H$_\\circ$ based on the Sunyaev-Zeldovich effect (see Birkinshaw (1999) for a review). We describe our model and make direct comparisons to the data in \\S\\ref{model}. In \\S\\ref{strategy}, we present detailed models of proposed {\\it Astro-E} observations based on the line-of-sight (LOS) ICM density, temperature and velocity structure provided by the numerical model. Section \\ref{summary} is a summary of our results. We assume \\Ho=70 \\kms Mpc$^{-1}$ when scaling the simulation to physical dimensions. ", "conclusions": "\\label{summary} We have presented a 3-dimensional numerical Hydro +N-body model of A3266 which is consistent with a wide range of observed properties. We believe A3266 represents an old ($\\sim$3 Gyr), off-axis merger that is occurring into the plane of the sky at a 45$^\\circ$ angle. The model is consistent, within resolution limits, with the {\\it ROSAT} PSPC image, current {\\it ASCA} temperature maps, {\\it BeppoSAX} hard X-ray flux limits, the galaxy spatial and velocity distributions, and the existing radio data. We have also checked that the projected mass distribution agrees with gravitational lensing data. In this model, the off-axis merger has imparted significant angular momentum on the ICM of the merger remnant which we predict should be observable with the XRS on {\\it Astro-E}. The signature of rotation will appear as two opposing bulk flows ($\\Delta$V$\\sim$800\\kms) located on either side of the cluster's X-ray core along a line of constant declination. The degree of circulation present in the core of our model is consistent with high resolution Hydro+N-body simulations of galaxy cluster formation from cosmological initial conditions (Norman \\& Bryan 1998). Neither the galaxy redshift data nor the N-body particle distribution show similar evidence of rotation. Unlike the ICM interaction which can be characterized as `collisional', the interaction between the subcluster galaxy components is `collisionless'. Therefore, while the transfer of angular momentum between ICM components is very efficient, angular momentum is not transferred between the galaxy components. It is difficult to assess the overall uniqueness of our model at the present time. Individually, no single observation places a strong constraint on the model. Even taken together, there is considerable flexibility in the merger parameters and viewing geometry, and we cannot rule out the possibility that recent mergers with significantly less massive subclusters have played some role (HDD99). Although limitations in the current data set certainly contribute to this problem, it is also indicative of an older merger that no distinctive features currently exist. In the event ICM rotation is not detected, the observations described here will nonetheless provide important new constraints on the model and on ICM dynamics in general. In addition, these data will result in highly accurate temperature and metalicity measurements for two widely separated regions within a single cluster thus giving clues to the degree of spatial variation in both quantities. Inhomogeneities in the distribution of metals may help quantify the rate of mixing between gas components within the merger." }, "0003/astro-ph0003130_arXiv.txt": { "abstract": "We have studied the effect of adopting different values of the total baryonic mass surface density in the local disk at the present time, $\\Sigma(R_\\odot, t_{Gal})$, in a model for the chemical evolution of the Galaxy. We have compared our model results with the G-dwarf metallicity distribution, the amounts of gas, stars, stellar remnants, infall rate and SN rate in the solar vicinity, and with the radial abundance gradients and gas distribution in the disk. This comparison strongly suggests that the value of $\\Sigma(R_\\odot, t_{Gal})$ which best fits the observational properties should lie in the range 50\\,--\\,75 $M_\\odot$ pc$^{-2}$, and that values of the total disk mass surface density outside this range should be ruled out. ", "introduction": "Despite a number of estimates, obtained by means of several different methods, the value of the local column density of the Galactic disk in stars and gas, $\\Sigma_{b}(R_\\odot \\simeq$ 8 kpc) = $\\Sigma_{stars} + \\Sigma_{gas}$, has been until recently very uncertain. Direct HST star counts imply a stellar column density of $\\simeq 27$ $M_\\odot$ pc$^{-2}$ (Flynn, Gould, \\& Bahcall 1999), while the (H {\\small I} + H$_2$) column density amounts to 7\\,--\\,13 $M_\\odot$ pc$^{-2}$ (Dickey 1993; Flynn et al. 1999). Therefore, the surface density of the directly seen baryonic matter amounts to 35\\,--\\,40 $M_\\odot$ pc$^{-2}$. However, the actual column density of the ``luminous'' disk could be higher because the above detection does not account for dead stars and brown dwarfs. \\par The total disk surface density $\\Sigma_{z}(R_\\odot)$, including also any eventual disk dark matter component, can be determined from the motion of stars in the solar neighborhood perpendicular to the Galactic plane (e.g. Binney \\& Tremaine 1987; Gilmore, Wyse, \\& Kuijken 1989). From the kinematics of a sample of K-giants Bahcall, Flynn, \\& Gould (1992) derived $\\Sigma_{z}(R_\\odot) = 84^{+30}_{-25}$ $M_\\odot$ pc$^{-2}$ (at 1 $\\sigma$ level). Notice that such a measure is a complex one and it is strongly dependent on the underlying assumptions. By employing different samples and/or analysis, lower values for $\\Sigma_{z}(R_\\odot)$ have been claimed: $\\Sigma_{z}(R_\\odot) = 46 \\pm 9$ $M_\\odot$ pc$^{-2}$ (Kuijken \\& Gilmore 1989) --- revised to $48 \\pm 9$ $M_\\odot$ pc$^{-2}$ (Kuijken \\& Gilmore 1991), $\\Sigma_{z}(R_\\odot) = 52 \\pm 13$ $M_\\odot$ pc$^{-2}$ (Flynn \\& Fuchs 1994). The local column density of stars and gas could be smaller, because these dynamical estimates may include some non-baryonic matter. Notice however that it is quite unlikely that there is non-baryonic matter in the disk, because non-baryonic dark matter is probably non dissipational. Any non-baryonic matter present in the disk is likely to be part of the dark halo, and this component is always subtracted from the dynamical estimates of the local column density. \\par Fitting the Galaxy rotation curve (RC) with a mass model provides a further way of estimating the local column density of disk matter $\\Sigma_{V}$. In fact, such a fit indicates what is the fraction $\\beta$\\footnote{It should be noted that the value for $\\beta$ depends on the assumed profiles for the disk and the halo.} of the circular velocity due to the disk component (at the Sun's position). Then, $\\beta\\,V^2_\\odot$ is compared with the circular velocity of an exponential thin disk, $V^2_{disk} = G^{-1}\\,\\Sigma_{V}\\,R_{D}\\,f(R_\\odot/R_{D})$, with $f$ a known function of the Galactocentric distance of the Sun expressed in terms of disk length scales (see Freeman 1970), to yield $\\Sigma_{V}$, providing that $R_\\odot$ and $R_{D}$ are known. The results are far from unique, and reflect the presence of uncertainties in the mass modelling, in the actual shape of the Galactic RC, and in the basic structural parameters. The values found range between $\\Sigma_{V} \\sim 50$ $M_\\odot$ pc$^{-2}$ and $\\sim 100$ $M_\\odot$ pc$^{-2}$ (Olling \\& Merrifield 1998; Dehnen \\& Binney 1998), and moreover include any dark matter distributed like the stellar disk. In external galaxies of luminosity similar to the Galaxy $(L_{B} \\sim 10^{11}$ $L_\\odot)$, values of $\\Sigma_{V}$ as high as $\\simeq 90$ $M_\\odot$ pc$^{-2}$ at $R$ = 8 kpc cannot be ruled out (Persic, Salucci, \\& Stel 1996). \\par Finally, let us note that if the dark matter halo in the Galaxy follows the standard cold dark matter (CDM) universal density profile (Navarro, Frenk, \\& White 1997), and therefore at the Sun position $\\rho_{SCDM} \\propto R^{-2}$, then the local disk column density is required to be towards the low end of the above-cited values (e.g. Courteau \\& Rix 1999). On the other hand, low-$\\Omega$ CDM density profiles are not inconsistent with a baryon dominated region inside $R_\\odot$, and then with much higher values for the local disk column density (Navarro 1998). \\par Summarizing: all the above observations, often entangled with heavy theoretical assumptions, poorly constrain the value of the column density of the baryonic matter in the local disk, that may lie between 35 and 100 $M_\\odot$ pc$^{-2}$. \\par However, very recently the results of the European Space Agency's Hipparcos mission have allowed a drastical reduction of this permitted range, by making available direct distances for the tracer stars. In particular, Cr\\'ez\\'e et al.\\,(1998) and Holmberg \\& Flynn (1998) have analyzed the A and F stars in the Hipparcos data set (some 10,000 stars) and derived greatly improved estimates of the total gravitating mass: the gravitating disk mass seems to be now firmly established at 50 to 60 $M_\\odot$ pc$^{-2}$. \\par In this paper we investigate how the gravitating mass of the disk influences the Galactic chemical evolution. Successful models for the chemical evolution of the Galaxy, in fact, require a star formation law depending on the total mass surface density (Tosi 1988; Matteucci \\& Fran\\c cois 1989; Chiappini, Matteucci, \\& Gratton 1997), and the star formation rate (SFR) is a fundamental parameter in the evolution of galaxies. ", "conclusions": "We used a detailed model of galactic chemical evolution applied to our own galaxy in order to put constraints on the total amount of disk baryonic matter. Observational estimates of this quantity in the local disk span a wide range of values (from 35 to 100 $M_\\odot$ pc$^{-2}$), depending on the underlying theoretical assumptions and on the adopted methods. We show that by means of chemical evolution models it is possible to substantially restrict the observed range of the total mass surface density. \\par We show that the value of $\\Sigma(R_\\odot)$ should be restricted to the range 50\\,--\\,75 $M_\\odot$ pc$^{-2}$, in order to guarantee the best fit to all the observational constraints available for the solar neighborhood and the overall Galactic disk. This is well consistent with the mass range which Cr\\'ez\\'e et al.\\,(1998) and Holmberg \\& Flynn (1998) advocate from their analysis of A and F stars with parallaxes and proper motions from the Hipparcos satellite." }, "0003/astro-ph0003076_arXiv.txt": { "abstract": "Multi-wavelength VLA observations of two extended radio galaxies, $0235-197$ and $1203+043$ are presented. There is some evidence from earlier studies that these two sources exhibit low frequency ($<1$\\,GHz) variability. This work shows that both sources have linear polarizations, if any, below the detection limits at 320 MHz, so we cannot explain the variability as being due to instrumental polarization effects as has been suggested for 3C159. Refractive scintillation may be the cause of the variability in 0235$-$197. This would require the existence of a bright, compact component in one of the hot spots seen in these observations. This is not implausible but the resolution of this observational program is insufficent to address that question. The radio source 1203$+$043 lacks any bright compact component thereby ruling out a refractive scintillation mechanism for its variability. Consequently, it is possible that claims of variability in this source are spurious. However, the 320\\,MHz VLA observations show that 1203$+$043 has an `X'-shaped radio structure. This is a rare morphology for the brightness distribution of a radio galaxy; the implications of this are examined. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003254_arXiv.txt": { "abstract": "Recent studies of galactic nuclei suggest that most galaxies are seeded by super-massive black holes which power the central nucleus. In this picture, the proto-galactic object is likely to have undergone a very active phase during which the surrounding medium was shocked and heated up to very high temperatures. We predict the cosmological effects of this scenario in terms of the thermal and kinetic Sunyaev-Zel'dovich distortions induced on galactic scales by a population of proto-galaxies. These predictions are compared to the observational limit on the mean Compton distortion set by the COBE-FIRAS instrument. This enables us to derive tight constraints on the fraction of proto-galaxies seeded by super-massive black holes as well as on the black hole-to-spheroid mass ratio. Finally, we estimate the contribution of such a population to the angular power spectrum of the Cosmic Microwave Background temperature anisotropies on very small angular scales ($l\\simeq 10^4-10^5$). ", "introduction": "Most theories of hierarchical structure formation are based on the study of the evolution of density perturbations under their own gravity. A density fluctuation, which represents an over- (or under-) density with respect to the mean matter distribution, contains both baryonic and dark matter (DM). The baryonic component sinks into the gravitational potential of the DM halo. It collapses and cools, resulting in star formation. In these scenarios, after the gravitational collapse of the DM halo, stars are assumed to be the first objects to form. A structure will thus end up as an emitting object after virialisation has occurred. An alternative picture involves the formation of a super-massive black hole (BH) that powers the central regions of galaxies \\cite[]{lynden-bell69}. Numerous studies have been performed that relate the quasar luminosity function to galaxy formation scenarios by assuming that the formation of quasars (i.e., BH) in the potential well of the DM halos constitutes one of the phases in the galaxy formation process \\cite[]{efstathiou88,haehnelt93,nusser93,haiman97}. Recent observations even suggest that a super-massive BH may be present in the centres of all galaxies with spheroidal components \\cite[]{kormendy95}.\\par Several authors have looked at several consequences of the presence of massive BHs on galaxy formation and evolution \\cite[]{haiman97,natarajan98,silk98}. In this paper, we investigate the cosmological implications of such an alternative scenario for the Cosmic Microwave Background (CMB) anisotropies and spectral distortions. More specifically, we study the effects of the outflows, driven by the BH activity, on the gas within the seeded proto-galaxy. In fact, the outflow expands and shock-heats the ambient medium (proto-galactic gas), and then interacts with the inter-galactic medium (IGM). Three regimes of interest may be considered: 1) the high density region of the proto-galaxy, 2) the low density IGM and 3) the thin compressed layer (four times denser than the IGM) induced by the front shock. The second and third regimes give results very similar to those computed in \\cite{aghanim96}. We thus focus on the first regime, i.e. the localised effects of the BH-driven shock on the gas within the seeded proto-galaxy. This shock-heated gas will Compton scatter the CMB photons and induce spectral distortions and temperature anisotropies through the so-called Sunyaev-Zel'dovich effect \\cite[]{suniaev80}. The thermal SZ effect depresses the CMB brightness in the Rayleigh-Jeans region and increases it above a frequency of about 219 GHz. Its amplitude represents the integral along the line of sight of the electron pressure. It is proportional to the electron density $n_e$ and is characterised by the Compton parameter $y$ defined by: \\begin{equation} y=2\\int_0^R\\frac{kT}{m_ec^2}\\sigma_Tn_e(l)\\,dl, \\label{eq:y} \\end{equation} where $T$ is the temperature of the gas, $R$ the physical size of the structure, $m_e$ the electron mass, $k$ the Boltzmann constant, $c$ the speed of light and $\\sigma_T$ the Thomson cross section. An additional secondary anisotropy arises due to the first-order Doppler effect of the CMB photons when they scatter on a structure moving with respect to the Hubble flow, with radial peculiar velocity $v_r$. This interaction is called the SZ kinetic effect. It generates an anisotropy, with no specific spectral signature, whose amplitude is given by: \\begin{equation} \\frac{\\delta T}{T}=\\frac{v_r}{c}\\times\\left( 2\\int_0^R\\sigma_Tn_e(l)\\,dl \\right). \\label{eq:dt} \\end{equation} Previous work on galaxy formation has evaluated the global distortion of a population of galaxies in the virialised regime. The global Compton parameter was found to be much smaller than the constraints set by the FIRAS instrument (Far InfraRed Absolute Spectro-photometer) on board COBE (COsmic Background Explorer) \\cite[]{fixsen96} on the global SZ distortion of the universe. By contrast, our model focuses on a regime in which proto-galaxies undergo a BH formation phase that induces larger distortions. The paper is organised as follows: in \\S 2, we model the shock in an individual structure and give its physical characteristics (size and temperature). In \\S 3, we compute the predicted number density of primordial galaxies, using the \\cite{press74} mass function. In \\S 4, we generalise the description of the shock to the whole population of proto-galaxies and we simulate maps of the induced secondary anisotropies. We estimate this contribution to the CMB anisotropies. We also compare our predicted global $y$ parameter to the COBE-FIRAS value and derive constraints on the model. Conclusions are given in the last section. ", "conclusions": "Previous studies on galaxy formation have evaluated the global distortion of a population of galaxies in the virialised regime. In these studies, the global Compton parameter was found to be very small, and smaller than the COBE-FIRAS value. In contrast, our model focuses on a regime in which proto-galaxies undergo a BH formation phase. During this phase, the proto-galactic matter is shock-heated up to a few $10^8$ K and cools down to $10^5$ K. CMB photons undergo inverse Compton scattering on the heated gas. In addition, galaxy peculiar motions induce temperature anisotropies through the SZ kinetic effect. We have estimated the global Compton parameter due to a population of proto-galaxies and the expected power spectrum of the induced secondary anisotropies. We find that there are four main parameters that control our model: the fraction $f$ of BH-seeded proto-galaxies, the fraction $\\epsilon_{BH}$ of the spheroid mass in the BH, the steepness of the density profile $\\alpha$ and the gas core radius $p=R_{vir}/R_c$. The comparison between our predictions and the COBE-FIRAS observation constrains these parameters. Given the observed fraction of seeded galaxies, $f=95$\\%, our results put rather strong constraints on the density profile and on $\\epsilon_{BH}$. Indeed, our predictions agree with the observations whatever $p$ if the density profile is an approximation to a King profile. On the contrary, if the density profile is isothermal, then the core radius must be at least 30 times smaller than the virial radius and the BH-to-spheroid mass ratio has to be small, of the order of $10^{-3}$. The computations in the two extreme cosmological models show that the global Compton parameter due to proto-galaxies is not very sensitive to $\\Omega_0$. We compare the power spectra of the different contributions to the temperature anisotropies. Our results show that the SZ effect of the very early shock-heated proto-galaxies could constitute the major source of CMB distortions on very small scales (arcsecond and sub-arcsecond scales). The anisotropies are likely to be detected and measured by future long baseline interferometers such as ALMA. The shock heating is likely to contribute to the re-heating of the proto-galactic gas, which plays a role in galaxy formation theory. \\cite{blanchard92} used preheating to modify the galaxy luminosity function, suppressing and finally delaying dwarf galaxy formation. We do not take into account this effect in our model, therefore, our results should be taken as an upper limit to the proto-galaxy contribution in terms of secondary anisotropies." }, "0003/astro-ph0003062_arXiv.txt": { "abstract": "Phase transitions in the early universe are prime settings for the production of primordial black holes, since they can break the relatively quiescent homogeneity and isotropy of Friedmann--Robertson--Walker (FRW) cosmologies. These epochs of ``symmetry breaking,'' moreover, can affect the subsequent development of spacetime by changing the evolution of some FRW parameters, including the present age and density of the universe. We discuss the relative importance of such effects on constraining mechanisms of black hole formation. ", "introduction": "Although they have not yet been observed, primordial black holes (PBHs) already deserve a special place in the temple of modern theoretical physics, for they have spawned many creative ideas at the intersection of cosmology, astrophysics, and particle physics. In cosmology, for example, they may affect the outcome of Big Bang nucleosynthesis (BBN) \\cite{BBNlimits}. Primordial black holes may be of astrophysical interest through their evaporative production of the highest energy cosmic rays \\cite{cosmicrays}. Their formation during cosmic phase transitions (e.g., the electroweak (EW) and quantum chromodynamic (QCD) transitions) also may help constrain the relevant particle physics. Further examples of this rich spectrum of applications abound in these proceedings. We consider below the example of Massive Compact Halo Object (MACHO) black holes and discuss two bulk cosmological constraints on the production of such holes. ", "conclusions": "" }, "0003/astro-ph0003312_arXiv.txt": { "abstract": "We show that the energy-level splitting arising from grain rotation ensures that paramagnetic dissipation acts at its maximum rate, i.e., the conditions for paramagnetic resonance are automatically fulfilled. We refer to this process as ``resonance relaxation''. The differences between the predictions of classical Davis-Greenstein relaxation and resonance relaxation are most pronounced for grains rotating faster than 1~GHz, i.e., in the domain where classical paramagnetic relaxation is suppressed. This mechanism can partially align even very small grains, resulting in linearly polarized microwave emission which could interfere with efforts to measure the polarization of the cosmic microwave background. ", "introduction": "} Experiments to study the cosmic background radiation have stimulated renewed interest in diffuse galactic emission. Recent maps of the microwave sky brightness have revealed a component of the 10-100 GHz microwave continuum which is correlated with 100 $\\mu$m thermal emission from interstellar dust (see review by Draine \\& Lazarian 1999a). Draine \\& Lazarian (1998a,b, henceforth DL98a,b) attributed this emission to electric dipole radiation from small ($<10^{-7}$~cm) rapidly rotating grains. Recent observations by de Oliveira-Costa et al.\\ (1999) support this interpretation. The question now is whether these small grains are aligned and their emission polarized. One process that might produce alignment of the ultrasmall grains is the paramagnetic dissipation mechanism suggested by Davis and Greenstein (1951) to explain the polarization of starlight. The Davis-Greenstein mechanism is straightforward: the component of interstellar magnetic field perpendicular to the grain angular velocity varies in grain coordinates, resulting in time-dependent magnetization, energy dissipation, and a torque acting on the grain. As a result grains tend to rotate with angular momenta parallel to the interstellar magnetic field. Although recent research (Draine \\& Weingartner 1996, 1997, Lazarian \\& Draine 1999a,b) suggests that paramagnetic alignment may not be the dominant alignment mechanism for $a\\gtsim 10^{-5}\\cm$ grains, it may be effective for small ($a\\ltsim 5\\times10^{-6}$~cm) grains. In the present paper we claim that the traditional picture of paramagnetic relaxation is incomplete, as it disregards the splitting of energy levels that arises within a rotating body. Unpaired electrons having spin parallel and antiparallel to the grain angular velocity have different energies resulting in the Barnett effect (Landau \\& Lifshitz 1960) -- the spontaneous magnetization of a paramagnetic body rotating in field-free space. Therefore the implicit assumption in Davis \\& Greenstein (1951) -- that the magnetization within a {\\it rotating grain} in a {\\it static} magnetic field is equivalent to the magnetization within a {\\it stationary grain} in a {\\it rotating} magnetic field -- is clearly not exact. In what follows we show that a very important effect due to rotation has thus far been overlooked. This effect, which we term ``resonance relaxation'', leads to energy dissipation -- and grain alignment -- which is much more rapid than the classical Davis-Greenstein estimate when the grain rotates very rapidly. ", "conclusions": "We have discussed a new gyromagnetic effect -- resonance relaxation -- which is closely related to normal paramagnetic resonance, and arises naturally whenever a body rotates in a weak magnetic field. The standard assumption of the equivalence of relaxation when the magnetic field rotates about a grain or a grain rotates in a static magnetic field is incorrect; the difference is directly related to the spontaneous magnetization due to the Barnett effect. Although present for all grains, resonance relaxation is most prominent for the smallest ones. When grains rotate very rapidly, as is the case for very small grains, the resonance relaxation effect ensures that $\\chi^{\\prime\\prime}$ does not plunge as the rotation frequency increases. As a result, we conclude that small grains (e.g. $a\\leq 10^{-7}$~cm) should be paramagnetically aligned. The degree of their aligment depends on the particular phase of the interstellar medium and on the efficiency of spin-lattice relaxation. The latter factor is unfortunately uncertain for very small grains for which the existing laboratory data is not applicable. If the ultrasmall grains are partially aligned, the implications are as follows: (1) The microwave radiation described in DL98ab will be polarized -- by a few \\% -- and could have dramatic consequences for experiments -- such as MAP or PLANCK -- designed to measure polarization of the cosmic microwave background. (2) If the grain body axes are aligned with $\\bJ$, then absorption by these small grains will contribute to starlight polarization in the ultraviolet, and (3) the infrared emission following absorption of starlight photons by these small grains will also be polarized. However, the contribution to starlight polarization is expected to be small due to only partial alignment of the grain body axes with $\\bJ$. The infrared emission will be even less polarized, due to disorientation of the grain axes (Lazarian \\& Roberge 1997) during the thermal spike following a photon absorption, i.e., while the infrared emission is taking place." }, "0003/astro-ph0003124_arXiv.txt": { "abstract": "Two-point correlation function of galaxy distribution shows that the structure in the present Universe is scale-free up to a certain scale (at least several tens Mpc), which suggests that a fractal structure may exist. If small primordial density fluctuations have a fractal structure, the present fractal-like nonlinear structure below the horizon scale could be naturally explained. We analyze the time evolution of fractal density perturbations in Einstein-de Sitter universe, and study how the perturbation evolves and what kind of nonlinear structure will come out. We assume a one-dimensional collisionless sheet model with initial Cantor-type fractal perturbations. The nonlinear structure seems to approach some attractor with a unique fractal dimension, which is independent of the fractal dimensions of initial perturbations. A discrete self-similarity in the phase space is also found when the universal nonlinear fractal structure is reached. ", "introduction": "The present Universe shows a variety of structures. The galaxies are not distributed randomly in the Universe. Totsuji \\& Kihara (1969) and Peebles (1974) showed that the observed two-point correlation function $\\xi(r)$ is given by a power law with respect to a distance $r$ as $\\xi(r) \\sim r^{-\\gamma}$ with $\\gamma \\sim 1.8$. The recent galaxy surveys also agree with this result, i.e. the power $\\gamma$ is nearly equal to $1.8$ (CfA (Geller \\& Hachra (1989)), LCRS (Jing, Mo, \\& B\\\"{o}rner (1998)), and ESP (Guzzo et al. (1998, 1999)). This may imply that the present distribution of galaxies is fractal. Sylos Labini, Montuori, \\& Pietronero (1998) have also claimed that all available data are consistent with a fractal structure with the dimension $D \\sim 2$ up to the deepest observed scale ($1000 h^{-1}$[Mpc]). However, the observation of Cosmic Microwave Background Radiation (CMBR) has revealed that the Universe in the recombination era is homogeneous and isotropic at least in very large scale. Although CMBR observation seems to be more reliable, we should not decide yet whether the large scale structure of the Universe is really fractal up to the horizon scale or not. To answer this question more definitely, we should await forthcoming next galaxy survey projects (Colless (1995), Maddox (1997), Loveday \\& Pier (1998), Knapp et al. (1999)). However, since it seems true that the galaxy distribution is really fractal up to a certain scale, one may ask how such a structure is formed in the evolution of the Universe. One of the most plausible explanations is that the nonlinear dynamics of the perturbations will provide such a scale-free structure during the evolution of the Universe. The pioneering work to explain the power-law behavior in nonlinear stage has been done by Davis \\& Peebles (1977). They assume a self-similar evolution of density fluctuation and some additional condition, i.e. a physical velocity $\\dot{r}$ vanishes in nonlinear regime. Then they showed a relation between the power index $\\gamma$ of two-point correlation function and that of initial power spectrum $n$ as $\\gamma=3(n+3)/(n+5)$. If we have $n=0$, then we find that $\\gamma =1.8$. Since their additional condition is not trivial and might not be appropriate, Padmanabhan (1996) and Yano \\& Gouda (1998a) extended their model to the case with non-vanishing $\\dot{r}$. They found that the relation between $n$ and $\\gamma$ is $\\gamma=[3h(n+3)]/[2+h(n+3)]$, where, $h\\equiv -a\\left<\\dot{x}\\right>/\\dot{a}x$, which is a ratio of a peculiar velocity to the Hubble expansion. With this result, $\\gamma$ can vary from 0 to 2 for $n=1$ (Harrison-Zel'dovich spectrum) and $0\\leq h \\leq 1$ ($h=1$ corresponds to the Davis-Peebles solution). Since we do not know the stability of those solutions, in order to find which value of $\\gamma$ is most likely, we should study the dynamics of density fluctuations in other methods, e.g. $N$-body simulation. Several groups in fact showed that a power-law behavior in two-point correlation function is obtained by $N$-body simulation with appropriate primordial density fluctuations (Miyoshi \\& Kihara (1975), Efstathiou (1979), Aarseth, Gott III, \\& Turner (1979), Frenk, White, \\& Davis (1983), Davis et al. (1985), Jing (1998)). The question is whether those power-law behaviors mean that we have a fractal structure in the present Universe. Peebles (1985) and Couchman \\& Peebles (1998) showed how to proceed with a high resolution analysis in the $N$-body simulation using a kind of renormalization method. They have used Davis-Peebles solution as a scaling relation. Without such an ansatz, we do not know whether usual $N$-body simulation is suitable to discuss the formation of a fractal structure. With the present state of computers, it may not be possible to obtain high enough resolution to analyze a fractal structure. As for a fractal structure in the Universe, one may ask another question. Did the Universe not have any non-trivial structure such as a fractal in the initial density fluctuations? In the conventional approaches, initial density perturbations are usually assumed to be given by a power-law (or a power-law-like) spectrum with random Gaussian phase. Although such initial conditions may provide the presently observed nonlinear scale-free structure via nonlinear dynamics, no one has shown whether such a structure is fractal or not, and if yes, what kind of fractal structure comes out. To provide a fractal structure in the present Universe, we may adopt an alternative scenario, in which primordial density fluctuations have already a fractal-like structure in the beginning. Note that a background spacetime is assumed to be a smooth universe, which is described by the Einstein-de Sitter universe, but not a fractal universe. The properties of an initial fractal may be preserved during the evolution of the Universe, then nonlinear fractal structure will be formed. In fact, De Gouveia Dal Pino et al. (1995) reported that the temperature fluctuation of CMBR has a fractal relation, and recently, Pando \\& Fang (1998) and Feng \\& Fang (2000) also reported that non-Gaussianity was detected in the distribution of Ly$\\alpha$ forest lines in the QSO absorption spectra. In this scenario, several natural questions may arise. How does such a primordial fractal perturbation evolve into nonlinear regime? Will any properties of the initial fractal be preserved during the evolution of the Universe, or not? If not, what kind of nonlinear structure will come out at present? Is there any fundamental difference in the structure formation process between a conventional density perturbation and the present fractal one? In order to answer those questions, we study the time evolution of the initial density fluctuations with a fractal structure in Einstein de-Sitter universe. Since we are interested in a fractal structure, a quite high resolution is required in our calculation. As we discussed, $N$-body simulation may not have enough resolution in the present state of computer development, unless we develop some skillful method. So, in this paper, we consider only a very simple toy model, which is a one-dimensional (1-D) sheet model, in order to get some insight into the questions raised in the above. To set up primordial fractal density perturbations, we distribute $N$ sheets initially by some systematic rule, i.e. we apply a Cantor set or random Cantor-type set (see below). Mathematically, in order to construct a Cantor set, the procedure must be repeated an infinite number of times, but it is not practically possible to set up such initial data. We therefore stop the procedure at a certain point, i.e. the initial set is given by several times removing line segments with a given ratio (Falconer (1990)). This could be justified because an infinite scale-free structure never exists in the real Universe. In order to construct the initial density perturbations, we set that the remaining segments have small positive density perturbations, while the removed ones correspond to small negative ones. Since we study a 1-D sheet-model, the motion of each sheet is described by an analytic solution (Zel'dovich (1970)), which guarantees enough resolution to analyze a fractal structure. In $\\S 2$, we present our formalism and initial setting. As for the initial data, we consider three cases: regular Cantor set, random Cantor-type set, and random white noise. Comparing those time evolutions, we show our results in $\\S 3$. In $\\S 4$, we focus particularly on the phase space. The conclusion and discussion follow in $\\S 5$. ", "conclusions": "We have studied the nonlinear evolution of primordial fractal fluctuations by using a 1-D sheet model. We have analyzed 7 models with initial fluctuations constructed by a regular Cantor set, 3 models with initial fluctuations constructed by a random Cantor-type set, and 2 models with white noise fluctuations. For all models except for the case with white noise, we find a kind of attractor with a universal fractal dimension ($\\sim 0.9$) as the fluctuations evolve into nonlinear regime. In the case with white noise fluctuations, the estimated dimension becomes stable around $0.7$, but the error in the estimation is larger than the other cases and the power-law behavior in a box-counting is also not completely fitted. Then, it may not contain a fractal structure. From the phase space analysis, we find a hierarchical structure, that is, the large vortex consists of some similar small vortices, and such small vortices again consist of similar but much smaller vortices. In particular, we find a discrete self-similarity for the model with a regular Cantor set. Why is the fractal dimension close to $0.9$? Is it really universal? Is the present fractal structure really an attractor? Although we need more analysis to answer this question, we have some hints in previous work. Gouda and Nakamura studied the present 1-D sheet model for the initial power law spectrum. They found two types of generic singularities when we have a shell crossing (Gouda \\& Nakamura 1988, 1989). When a first shell crossing appears, the relation between Eulerian and Lagrangian coordinates must be \\begin{equation} x=q_c +\\beta (q-q_c)^3 + \\cdots, \\end{equation} while that after a shell crossing turns out to be \\begin{equation} x=q_c +\\beta (q-q_c)^2 + \\cdots. \\end{equation} Following Arnold's classification, the former and latter cases are classified into A3 and A2, respectively. A3 is structurally unstable and may appear transiently in the expanding Universe. A2 is structurally stable and appears universally for the initial power-law spectrum. The latter case gives \\begin{eqnarray} \\delta_k &=& \\int \\delta (x) e^{ikx} dx \\nonumber \\\\ & \\propto & \\int (\\beta x)^{-1/2}e^{ikx} dx \\nonumber \\\\ & =& (\\beta k)^{-1/2} \\int \\eta^{-1/2} e^{i\\eta} d\\eta, \\end{eqnarray} i.e. $P(k) \\sim k^{-1}$. This predicts $\\gamma = 0$, i.e. $D_F=1$, which is rather close to our ``universal\" dimension 0.9. Although one may wonder that these are essentially the same, we have another result which suggests that there seems to exist a new type of stable phase. Recently, Yano and Gouda analyzed a more realistic case, i.e. the initial power law spectrum with a cut-off and found 5 characteristic regions in Fourier space (Yano \\& Gouda 1998a, b). The regime 1 is the linear one and then it is just an initial power spectrum. In the regime 2, they found $P(k) \\sim k^{-1}$, which is the single-caustic regime (Gouda \\& Nakamura (1989)). The regime 3 is called the multi-caustic regime, in which the power spectrum depends on the initial power-law index. Beyond the cut-off scale, two regimes appear, one gives $P(k) \\sim k^{-1}$ (regime 5), which may correspond to A2 type stable solution. In the intermediate wave number $k$ between the regime 3 and regime 5, they find $k^\\nu$, which $\\nu$ is independent of initial power index and close to 1, but a little less. They called it the virialized regime. This seems to be a new transient region, which may appear in some specific initial conditions. We would conjecture that the fractal structure with a universal dimension 0.9 corresponds to this virialized regime (regime 4) and the dimension 0.7 found in the case with white noise would be the regime 3. By reanalyzing the Yano-Gauda model in the case of $k=0$, we have confirmed that $\\nu =\\sim 0.9$. We also find a small tail with index 0.7 in the size-number relation in the Cantor set model with $n_D =15, 20$ (Fig. 19). This conjecture is also supported by the analysis for a self-gravitating 1-D sheet model without the background expansion of the Universe (Tsuchiya, Konishi \\& Gouda 1994). They found two time scales; one is a micro relaxation time ($t_{\\rm micro} = N t_c$) and the other is a global relaxation time ($t_{\\rm global} = 4 \\times 10^4 N t_c$), where $t_c =\\sqrt{L/4\\pi G N m}$ is a crossing time. After $t_{\\rm micro} $, some equilibrium state is reached by exchanging particle energy, but the global relaxation is not achieved, i.e. the partition function is not yet described by an equilibrium state such as an ergodic state (Tsuchiya, Konishi \\& Gouda 1994). In the present model, we can speculate that the fractal structure is obtained after this micro relaxation time but before the global relaxation time. In fact, if we estimate the above time scales in the present models, we find that $t_{\\rm micro}$ corresponds to $a=5\\times 10^3$, while $t_{\\rm global}$ corresponds to $a=5\\times 10^6$. The time when we find a stable fractal structure ($a=(1 \\sim 3) \\times 10^4$) is between those two time scales. If this speculation is true, our fractal structure is temporal. In the future of the Universe, it will evolve into more relaxed and ergodic state. Since we analyze the simplest case, we have to extend our analysis to more generic cases. First, we should study different types of fractal in order to check whether the present results are universal for any fractal distributions or not. Secondly, we need to analyze the case with scale-dependent fluctuations. In the present analysis, we set $\\delta_+=$ constant and $\\delta_-=$ constant. In the realistic case, there must be a scale dependence to the fluctuations. In the conventional perturbations, we usually assume a power-law spectrum with some cutoff. Even if the primordial fluctuations contain a fractal structure, their amplitude may depend on the scale. Its dependence may change the present results. In particular, in the present model, the scenario of structure formation could be different from either TOP-DOWN or BOTTOM-UP for some scale-dependence. The primordial fractal fluctuations will evolve directly into a hierarchical nonlinear structure. But, it will definitely depend on the scale dependence of the fluctuations. Secondly, we need to extend the present analysis to other cosmological models, i.e. the open Universe model and $\\Lambda \\neq 0$ flat Universe model. For the 1-D sheet model, the solutions are still exact, and the growth and decay rates in these models are different from those in the Einstein-de Sitter Universe model. We expect that the structure formation after a shell crossing is not the same as that in the present cosmological model, and then the fractal dimension would be different. For more realistic cases, we must study either the 2-D or 3-D model. Since the Zel'dovich solution is no longer exact, we have to explore a new method. In order to preserve a high resolution, we may develop a kind of renormalization method in $N$-body simulation as Couchman \\& Peebles (1998). Finally, it would also be interesting to look for the origin of such a primordial fractal density perturbation. The inflationary scenario may provide the origin of primordial fluctuations. One may wonder whether such a fractal primordial fluctuation is expected in some inflationary models. If we have more than two scalar fields, then the system is not integrable and may show a chaotic behavior or a fractal property (Easther \\& Maeda (1999)). Such a model might show up a kind of fractal density perturbation." }, "0003/astro-ph0003197_arXiv.txt": { "abstract": "We explore the viability of the unification of BL Lacs and FR~I radio galaxies by comparing the core emission of radio galaxies with those of BL Lacs of similar extended radio power, taking advantage of the newly measured optical nuclear luminosity of FR~I sources. The spectral properties of complete samples are also studied in the radio-optical luminosity plane: starting from the Spectral Energy Distribution (SED) of BL Lacs, we calculate the predicted luminosity of FR~I nuclei in the frame of a simple one--zone model, by properly taking into account the relativistic transformations. We find that the bulk Lorentz factors required by the spread in the observed luminosities in all bands are significantly smaller than those implied by other, both observational and theoretical, considerations. This discrepancy is also reflected in the fact that FR~I nuclei are over--luminous by a factor of 10-10$^4$, with respect to the predictions, both in the radio and in the optical band. In order to reconcile these results with the unification scheme, velocity structures in the jet are suggested, where a fast spine is surrounded by a slow (but still relativistic) layer so that the emission at different angles is dominated by different velocity components: the fast one dominates the emission in BL Lacs while the slow layer dominates the emission in misaligned objects. Furthermore for the lowest luminosity BL Lacs it has to be also postulated that their beaming factor in the radio band is lower than in the optical (and X--ray), as would result from deceleration of the jet. The self--consistency of the unification model therefore requires that both intrinsic differences in the SED and different beaming properties play a substantial role in characterizing the phenomenology of these sources. ", "introduction": "Unification models adduce the main differences between the observed properties of different classes of AGNs to the anisotropy of the radiation emitted by the active nucleus (see Antonucci \\cite{anto} and Urry \\& Padovani \\cite{urrypad} for reviews). In particular, for low luminosity radio-loud objects, namely BL Lacs and FR~I radio galaxies (Fanaroff \\& Riley \\cite{fr74}), it is believed that this effect is mainly due to relativistic beaming. In fact, there is growing evidence that obscuration does not play a significant role in these objects, contrary to other classes of AGNs. This is indicated by optical (Chiaberge et al. \\cite{paperI} hereafter Paper~I), radio (Henkel et al. \\cite{henkel}), and X-ray information (e.g. Fabbiano et al. \\cite{fabbiano}, Worral and Birkinshaw \\cite{wb94}, Trussoni et al. \\cite{edo}). Within this scenario, the emission from the inner regions of a relativistic jet dominates the observed radiation in BL Lacs, while in FR~I, whose jet is observed at larger angles with respect to the line of sight, this component is strongly debeamed. Evidence for this unification scheme includes the power and morphology of the extended radio emission of BL Lacs (e.g. Antonucci \\& Ulvestad \\cite{antoulve}, Kollgaard et al. \\cite{koll92}, Murphy et al. \\cite{murphy}) and the properties of their host galaxies (e.g. Ulrich \\cite{ulrich89}, Stickel et al. \\cite{stickel91}, Urry et al. \\cite{urry99}), which are similar to those of FR~I. Furthermore, there is a quantitative agreement among the amount of beaming required by different observational properties (e.g. Ghisellini et al. \\cite{gg93}), the number densities and luminosity functions of the parent and beamed populations in different bands (e.g. Urry \\& Padovani \\cite{urrypad}, Celotti et al. \\cite{celo93}) and the comparison of the radio core emission of beamed and unbeamed objects with similar total radio power (Kollgaard et al. \\cite{koll96}). Despite this global agreement, it should be stressed that beaming factors inferred from the broad band spectral properties of blazars, more specifically superluminal motions, transparency to the $\\gamma$--ray emission, shape of the SED and time--lags among variations at different frequencies, are significantly and systematically larger than those suggested by radio luminosity data (Dondi \\& Ghisellini \\cite{dondi}, Ghisellini et al. \\cite{gg98}, Tavecchio et al. \\cite{taold}). Thanks to the Hubble Space Telescope (HST), faint optical nuclear components have been recently detected in FR~I galaxies (Chiaberge et al. \\cite{paperI}). A strong linear correlation is found between this optical and the radio core emission which strongly argues for a common non-thermal origin. This suggests that the optical cores can be identified with synchrotron radiation produced in a relativistic jet, qualitatively supporting the unifying model for FR~I and BL Lacs. These information offer a new possibility of verifying the unification scheme, by directly comparing the properties of the optical and radio cores of radio galaxies with their putative aligned (beamed) counterparts, analogously to the procedure followed for the radio cores. X-ray observations also provide useful constraints to the nuclear emission of FR~I sources (e.g. Hardcastle \\& Worrall \\cite{hard99}). The main advantage of using multifrequency data is the possibility of directly comparing the full broad band spectral distributions of these two classes of sources and eventually shed light on the apparent discrepancy in the Lorentz factors inferred from different approaches. The paper is organized as follows. The (complete) samples of BL Lacs and radio galaxies are presented in Sect. \\ref{samples}. In Sect. \\ref{corext} we compare separately the core radio and optical emission of beamed and unbeamed objects with similar extended radio power. From this we infer the Lorentz factors requested by the unification scheme within the simplest scenario in which the radiation is emitted by a single uniform region of the relativistic jet. In Sect. \\ref{lrlo} the radio and optical data are considered together and, starting from the observed SED of BL Lacs, we derive the expected properties of the nuclear emission of FR~I, by taking into account the spectral dependence of the relativistic transformations. As the single--region picture does not account for the observed properties, in Sect. \\ref{multijet} we explore a (simple) alternative scenario and test it also against the X-ray information. Summary and conclusions are presented in Sect. \\ref{summary}. ", "conclusions": "\\label{summary} With the aim of exploring the viability of the unification scenario between (HBL, LBL) BL Lacs and FR~I radio galaxies we have compared their nuclear emission in the radio, optical and X--ray bands. We have firstly considered these spectral regions separately, comparing the nuclear emission of the two classes of objects for similar extended radio power. As the core radiation of BL Lacs is enhanced by relativistic beaming, we derived the bulk Lorentz factors requested to account for the observed distribution. The values of $\\Gamma$ thus inferred are not compatible with the higher bulk velocities requested by theoretical arguments, such as the pair production opacity and the spectral modeling of the SED of BL Lacs. We then examined the core emission of three samples in the $L_r-L_o$ plane. In the frame of the simplest one-zone emission model, we calculated debeaming trails of the BL Lac broad band emission as predicted by the relativistic transformation for an increasing angle of sight. We found that the model does not account for the observed spectral properties of FR~I, as expected from the above inconsistency of the Lorentz factors. The simplest and rather plausible hypothesis to account for this discrepancy within the unification scenario is to assume a structure in the jet velocity field, in which a fast spine is surrounded by a slow layer. Note however that the slower jet component must be relativistic in order to explain the anisotropic radiation of radio galaxy cores (e.g. Capetti \\& Celotti \\cite{ac2}). The observed flux is dominated by the emission from either the spine or the slower layer, in the case of aligned and misaligned objects, respectively. Interestingly, the existence of velocity structures in the jet has been suggested by various authors (Komissarov \\cite{komiss}, Laing \\cite{laing93}) in order to explain some observed properties of FR~I (and FR~II) jets, such as the structure of the magnetic field in FR~I which appears to be longitudinal close to the jet axis and transverse at the edges. Swain et al. \\cite{swain} obtained VLA images of \\object{3C~353} (an FR~II with straight jets), finding that a model consisting in a fast relativistic spine ($\\beta >0.8$) plus a slower outer layer ($\\beta <0.5$, but still relativistic in order to produce the observed jet-counterjet intensity asymmetry) could account for the apparently lower emissivity near the jet axis. Similar behaviours have been inferred for the two low luminosity radio galaxies \\object{M~87} (Owen et al. \\cite{owenm87}) and \\object{B2~1144+35} (Giovannini et al. \\cite{gg299}). Furthermore, Laing et al. (\\cite{laing99}) showed that the jet asymmetries in FR~I can be explained by means of a two-speed model. As a consequence, they argued that the lower velocity component dominates in the cores of the edge-on sources, while the fast spine emission dominates the end-on ones. This possibility might be also supported by recent numerical simulations of relativistic jets (Aloy et al. \\cite{aloy}). The same indication has been found through different approaches. Capetti \\& Celotti \\cite{ac2} reveal a trend in the radio galaxy/BL Lac relative powers with the line of sight, which is consistent with a slower (less beamed) component dominating at the largest angles. Capetti et al. \\cite{hstx} consider the same issue by examining the more detailed SED of five radio galaxies and consider their beamed counterparts. They found that while the spectral shapes of \\object{3C~264} and \\object{3C~270} can be reconducted to those of BL Lacs, the required ratio of beaming factors, i.e. $\\delta_{\\rm BL Lac}/ \\delta_{\\rm FR~I}\\sim 10-100$, implies that the corresponding BL Lacs would be overluminous. The inclusion of a slower (less beamed) jet component seems to be a plausible explanation. We found that Lorentz factors of the layer $\\Gamma_{layer}\\sim 2$ can account for the unification of FR~I (of the 3CR) with LBL and intermediate luminosity BL Lacs. Instead the debeaming trails for the lowest luminosity HBL do not cross the FR~I region in the $L_r-L_o$ plane. While the HBL behavior should be compared with that of radio galaxies with which they share the extended radio power (e.g. those of the B2 catalogue), our simple two-component jet model could not account for the observed properties if the cores of such low-power FR~I radio galaxies lied on the extrapolation of the 3CR radio-optical correlation. The properties of such weak sources can be instead reproduced if their radio emitting region is less beamed than the optical one, as could be expected if the jet decelerates after the higher energy emitting zone. Finally, the presence of velocity structures in jets of course affects the number counts of beamed and unbeamed sources: for example, the lack of BL Lacs in clusters (Owen et al. \\cite{owen96}) could be attributed to values of typical bulk Lorentz factors higher than those derived from statistical arguments (Urry et al. \\cite{ups91}). Intriguingly, the very latter authors had to require a wide distribution of Lorentz factors to account for the number densities of FR~I and BL Lacs in the radio band. Much has still to be understood on the dynamics and emitting properties of relativistic jets. Multifrequency studies of the nuclear properties of beamed sources and their parent populations and their comparison -- according to unification scenarios which are well supported by other independent indications -- constitute a new and powerful tool to achieve that, both for well studied individual sources as well as complete samples. Near IR observations by HST, mm data and higher resolution and sensitivity by Chandra in X--rays will further open this possibility. Concluding, the radio, optical and X-ray nuclear emission of FR~I and BL Lacs strongly indicate the presence of a velocity structure in the jet if indeed these sources are intrinsically identical. In other words, by considering the indications of trends in the SED of blazars emerged in the last few years (Giommi \\& Padovani \\cite{giopad94}, Fossati et al. \\cite{gfos}) together with the constraints derived from their unification with radio galaxies, it appears that the phenomenology of these sources is characterized and determined by differences {\\it both} in the intrinsic SED {it and} in beaming properties." }, "0003/astro-ph0003422_arXiv.txt": { "abstract": "s{We report preliminary results from a first season of photometric monitoring of 600 quasars behind the Virgo galaxy cluster with the aim of detecting microlensing by the cluster dark matter. Our project is sensitive to dark objects of surface mass densities down to $\\sim 20 \\,{\\rm g \\; cm}^{-2}$. We are thus capable of detecting diffuse objects, such as cold molecular clouds, unlike all Galactic microlensing surveys whose surface mass density limits are $> 10^{4} \\; {\\rm g \\; cm}^{-2}$. The average optical depth to microlensing of quasars through the central 30 sq. deg. of Virgo is $\\sim 1 \\times 10^{-3}$. We report a null detection which implies that less than half the dark matter in Virgo is in objects of mass $\\sim 10^{-5}$ solar masses, of surface mass density $> 20 \\,{\\rm g \\; cm}^{-2}$, at $90\\%$ confidence.} ", "introduction": " ", "conclusions": "From our null detection, and the fact that we would have expected $5$ events for lenses of mass $1 \\times 10^{-5}$ \\Msun, we can use Poisson statistics to put constraints on the mass in Virgo in the form of dark objects of this mass. The result of this pilot project is that less than $1/2$ the mass in Virgo can be in the form of $10^{-5}$ \\Msun objects at $90\\%$ confidence. The most important thing we have learnt from the pilot project is that a quasar monitoring microlensing project is technically feasible and we hope, over the next 2 years, to acquire of order $100$ more plates of Virgo. Our simulations indicate that this will allow us to improve our constraints on the nature of dark matter in Virgo by a factor $5-10$. A second season of monitoring Virgo started in February 2000. We would also like to extend the experiment by monitoring other clusters - especially the Perseus cluster in the North as this is the most massive nearby cluster. Alternatively, another strategy would be to monitor several more distant clusters (thereby avoiding the need for a very wide field of view). In the long term we would also like to monitor a control field of quasars with no intervening massive cluster in order to precisely quantify quasar variability." }, "0003/astro-ph0003278_arXiv.txt": { "abstract": "We discuss the evolution of linear perturbations about a Friedmann--Robertson--Walker background metric, using only the local conservation of energy--momentum. We show that on sufficiently large scales the curvature perturbation on spatial hypersurfaces of uniform-density is conserved when the non-adiabatic pressure perturbation is negligible. This is the first time that this result has been demonstrated independently of the gravitational field equations. A physical picture of long-wavelength perturbations as being composed of separate Robertson--Walker universes gives a simple understanding of the possible evolution of the curvature perturbation, in particular clarifying the conditions under which super-horizon curvature perturbations may vary. ", "introduction": "Structure in the Universe is generally supposed to originate from the quantum fluctuation of the inflaton field. As each scale leaves the horizon during inflation, the fluctuation freezes in, to become a perturbation of the classical field. The resulting cosmological inhomogeneity is commonly characterized by the intrinsic curvature of spatial hypersurfaces defined with respect to the matter. This metric perturbation is a crucial quantity, because at approach of horizon re-entry after inflation it determines the adiabatic perturbations of the various components of the cosmic fluid, which seem to give a good account of large-scale structure~\\cite{LL}. To compare the inflationary prediction for the curvature perturbation with observation, we need to know its evolution outside the horizon, through the end of inflation, until re-entry on each cosmologically relevant scale. The standard assumption is that the curvature perturbation is practically constant. This has recently been called into question in the context of preheating models \\cite{KLS97} at the end of inflation where non-inflaton perturbations can be resonantly amplified \\cite{Betal,LLMW}. The purpose of the present paper is to investigate the circumstances under which the curvature perturbation may vary. Using only the local conservation of energy--momentum, we show that the rate of change of the curvature perturbation on uniform-density hypersurfaces\\footnote{The ``conserved quantity'' $\\zeta$ was originally defined in Bardeen, Steinhardt and Turner \\cite{BST}, but constructed from perturbations defined in the uniform Hubble-constant gauge.}, $\\zeta$, on large scales is due to the non-adiabatic part of the pressure perturbation. This result is independent of the form of the gravitational field equations, demonstrating for the first time that the curvature perturbation remains constant on large scales for purely adiabatic perturbations in {\\em any} relativistic theory of gravity where the energy--momentum tensor is covariantly conserved, $T^\\mu_{~\\nu;\\mu}=0$. We also show that for adiabatic perturbations produced during single field inflation the curvature perturbation on uniform-density hypersurfaces, $\\zeta$ \\cite{BST,Bardeen88,MS98}, can be identified with the comoving curvature perturbation, ${\\cal R}$ \\cite{LL,David+Tony}. The pressure perturbation must be adiabatic if there is a definite equation of state for the pressure as a function of density, which is the case during both radiation domination and matter domination. On the other hand, a change in $\\zeta$ on super-horizon scales will occur during the transition from matter to radiation domination if there is an isocurvature matter density perturbation~\\cite{ks87,David+Tony}. We give a simple derivation of this effect in terms of the curvature perturbations on uniform-radiation and uniform-matter hypersurfaces which remain constant throughout. A simple intuitive understanding of how the curvature perturbation on large scales changes, due to the different integrated expansion in locally homogeneous but causally-disconnected regions of the universe, can be obtained within the `separate universes' picture which we describe in section~\\ref{sepsect}. This enables one to model the evolution of the large-scale curvature perturbation using the equations of motion for an unperturbed Robertson--Walker universe. In section~\\ref{sectinfl} we use this approach to discuss the evolution of the curvature perturbation in single- and multi-field inflation models. ", "conclusions": "In this paper, we have identified the general condition under which the super-horizon curvature perturbation on spatial hypersurfaces can vary as being due to differences in the integrated expansion along different worldlines between hypersurfaces. As long as linear perturbation theory is valid, then, when spatial gradients of the perturbations are negligible, such a situation can be described using the separate universes picture, where regions are evolved according to the homogeneous equations of motion. In particular, the curvature perturbation on uniform-density hypersurfaces, $\\zeta$, can vary only in the presence of a significant non-adiabatic pressure perturbation. The result follows directly from the local conservation of energy--momentum and is independent of the gravitational field equations. Thus $\\zeta$ is conserved for adiabatic perturbations on sufficiently large scales in any metric theory of gravity, including scalar--tensor theories of gravity or induced four-dimensional gravity in the brane-world scenario. Multi-component inflaton models are an example where non-adiabatic perturbations may cause the curvature perturbation to evolve on super-horizon scales." }, "0003/astro-ph0003087_arXiv.txt": { "abstract": "We present a detailed analysis of the space motions of 1203 solar-neighborhood stars with metal abundances $\\feh \\le -0.6$, on the basis of a recently revised and supplemented catalog of metal-poor stars selected without kinematic bias (Beers et al. 2000). This sample, having available proper motions, radial velocities, and distance estimates for stars with a wide range of metal abundances, is by far the largest such catalog to be assembled to date. We show that the stars in our sample with [Fe/H] $\\le -2.2$, which likely represent a ``pure'' halo component, are characterized by a radially elongated velocity ellipsoid $(\\sigma_U,\\sigma_V,\\sigma_W)=(141 \\pm 11,106 \\pm 9,94 \\pm 8)$ km~s$^{-1}$ and small prograde rotation $=30$ to 50 km~s$^{-1}$, consistent with previous analysis of this sample based on radial velocity information alone (Beers \\& Sommer-Larsen 1995). In contrast to the previous analysis, we find a decrease in $$ with increasing distance from the Galactic plane for stars which are likely to be members of the halo population ($\\Delta/\\Delta|Z|=-52 \\pm 6$ km~s$^{-1}$ kpc$^{-1}$), which may represent the signature of a dissipatively formed flattened inner halo. Unlike essentially all previous kinematically selected catalogs, the metal-poor stars in our sample exhibit a diverse distribution of orbital eccentricities, $e$, with no apparent correlation between [Fe/H] and $e$. This demonstrates, clearly and convincingly, that the evidence offered by Eggen, Lynden-Bell, and Sandage (1962) for a rapid collapse of the Galaxy, an apparent correlation between the orbital eccentricity of halo stars with metallicity, is basically the result of their proper-motion selection bias. However, even in our non-kinematically selected sample, we have identified a small concentration of high-$e$ stars at [Fe/H] $\\sim-1.7$, which may originate, in part, from infalling gas during the early formation of the Galaxy. We find no evidence for an additional thick disk component for stellar abundances [Fe/H] $\\le-2.2$. The kinematics of the intermediate-abundance stars close to the Galactic plane are, in part, affected by the presence of a rapidly rotating thick disk component with $ \\simeq 200 $ km~s$^{-1}$ (with a vertical velocity gradient on the order of $\\Delta/\\Delta|Z|=-30 \\pm 3$ km~s$^{-1}$ kpc$^{-1}$), and velocity ellipsoid $(\\sigma_U,\\sigma_V,\\sigma_W)=(46 \\pm 4,50 \\pm 4,35 \\pm 3)$ km~s$^{-1}$. The fraction of low-metallicity stars in the solar neighborhood which are members of the thick disk population is estimated as $\\sim 10\\%$ for $-2.2 < $ [Fe/H] $ \\le-1.7$ and $\\sim 30\\%$ for $-1.7< $ [Fe/H] $\\le-1$. We obtain an estimate of the radial scale length of the metal-weak thick disk of $4.5 \\pm 0.6$ kpc. We also analyze the global kinematics of the stars constituting the halo component of the Galaxy. The outer part of the halo, which we take to be represented by local stars on orbits reaching more than 5 kpc from the Galactic plane, exhibits no systematic rotation. In particular, we show that previous suggestions of the presence of a ``counter-rotating high halo'' are {\\it not} supported by our analysis. The density distribution of the outer halo is nearly spherical, and exhibits a power-law profile that is accurately described as $\\rho \\propto R^{-3.55 \\pm 0.13}$. The inner part of the halo is characterized by a prograde rotation and a highly flattened density distribution. We find no distinct boundary between the inner and outer halo. We confirm the clumping in angular-momentum phase space of a small number of local metal-poor stars noted by Helmi et al. (1999). We also identify an additional elongated feature in angular-momentum phase space extending from the clump to regions with high azimuthal rotation. The number of members in the detected clump is not significantly increased from that reported by Helmi et al., even though the total number of the sample stars we consider is almost triple that of the previous investigation. We conclude that the fraction of halo stars that may have arisen from the precursor object of this clump may be smaller than 10\\% of the present Galactic halo, as previously suggested. The implications of our results for the formation of the Galaxy are discussed, in particular in the context of the currently favored Cold Dark Matter theory of hierarchical galaxy formation. ", "introduction": "Over the past few decades, studies of the luminous halo population of metal-deficient field stars and globular clusters have provided an increasingly detailed picture of the formation and evolution of the Galaxy. Because the time required for mixing of the initial phase-space distribution of these objects, via exchange of energies and angular momenta, is thought to exceed the age of the Galaxy, kinematic information obtained at the present enables one to elucidate the initial dynamical conditions under which these objects were born. To the extent one is able to estimate ages of these halo population objects, either directly (which is difficult at present), or indirectly (by postulating that the {\\it ensemble} metallicities of these objects increases with time), their formation history is obtainable as well. Thus, the dynamical and chemical state of these halo-population objects provides important information on how the Galaxy has developed its characteristic structures during the course of its evolution. Almost forty years ago, the ``canonical'' scenario of the early dynamical evolution of the Galaxy was put forward by Eggen, Lynden-Bell, \\& Sandage (1962, hereafter ELS) to explain what they believed to be an observed correlation between the orbital characteristics and metal abundances of stars in the solar neighborhood. Focusing on the lack of metal-poor stars with low eccentricity orbits in their (proper-motion selected) sample, ELS argued that the Galaxy must have undergone a rapid collapse, then formed a rotationally supported disk. Although criticism of the ELS model has been levied because of the potential influence of their kinematic selection bias, especially the extent to which this might alter the derived collapse timescale of the early Galaxy (Yoshii \\& Saio 1979; Norris, Bessell, \\& Pickles 1985; Norris 1986; Norris \\& Ryan 1991; Beers \\& Sommer-Larsen 1995, hereafter BSL; Chiba \\& Yoshii 1998, hereafter CY), the ELS collapse picture has long been influential for studies of disk galaxies like our own, and for elliptical galaxies as well (e.g., Larson 1974; van Albada 1982). An alternative picture for the formation of the Galactic halo was proposed by Searle \\& Zinn (1978, hereafter SZ), who noted a number of difficulties in reconciling several observed properties of the halo globular cluster system with predictions of the ELS model. Among these, the existence of a large (several Gyr) spread in the inferred ages of the Galactic globulars, and the lack of an abundance gradient with distance from the Galactic center were thought to be the most crucial. SZ suggested that, in its earliest epochs, the halo component of the Galaxy may have experienced prolonged, chaotic accretion of subgalactic fragments. More recent studies of halo field stars also provide evidence which may support the SZ picture, including a possible age spread among halo subdwarfs (e.g., Schuster \\& Nissen 1989; Carney et al. 1996), a gradient in the inferred ages of field horizontal-branch (FHB) stars and RR Lyrae variables, in the sense that the outer halo FHB stars and RR Lyrae variables appear several Gyr younger than those of the inner halo (Preston, Shectman, \\& Beers 1991; Lee \\& Carney 1999), a report of the apparent clustering of FHB stars in the halo (Doinidis \\& Beers 1989), possible kinematic substructure in the halo (e.g., Doinidis \\& Beers 1989; Majewski, Munn, \\& Hawley 1994; 1996), and distinct changes in the kinematics of the field populations as one moves from the inner to outer halo (e.g., Majewski 1992; Carney et al. 1996; Sommer-Larsen et al. 1997). In order to assess which picture, ELS or SZ (or both, e.g., Norris 1994; Freeman 1996; Carney et al. 1996), more correctly describes the early history of the Galaxy, we require a large and reliable set of data for halo-population objects chosen with criteria that do not unduly influence the subsequent analysis. As we show, the analysis of stars chosen with a kinematic selection bias can be particularly troublesome. Interestingly, in order to obtain adequately large samples of stars exhibiting a range of metallicities in the solar neighborhood, an abundance bias is actually {\\it required}, otherwise the exceedingly rare very low-metallicity stars of the halo population will not be represented in sufficient numbers. One must exercise caution, however, that abundance estimates for stars in the sample under consideration are as accurate as possible, due to the presence of an overlap of the local halo with the relatively high density thick-disk population (e.g., Anthony-Twarog \\& Twarog 1994; BSL; Ryan \\& Lambert 1995; Twarog \\& Anthony-Twarog 1996; CY). It is similarly important to assemble a large and homogeneously analyzed sample, both to minimize statistical fluctuations in the derived kinematic quantities, and to reduce the effects of other systematic errors (such as might arise in estimates of stellar distances). In this paper we present an analysis of the kinematics of metal-deficient field stars in the solar neighborhood, based on a large catalog of stars selected without kinematic bias (Beers et al. 2000, hereafter Paper II). This catalog, consisting of 2041 stars from the published literature with abundances [Fe/H]$ \\le -0.6$, includes updated stellar positions, newly derived homogeneous distance estimates, revised radial velocities, and refined metal abundance estimates. Moreover, a subset of some 1200 stars in the catalog now have available proper motions, taken from a variety of recently completed proper motion catalogs. We note that this catalog is (by far) the largest sample of metal-deficient field stars with available proper motions among any previously assembled non-kinematically selected samples. Thus, it is now possible to draw a much more definitive picture of the early kinematic evolution of the Galaxy. Our paper is organized as follows. In \\S 2 we present a discussion of the detailed velocity distributions of our sample stars, concentrating on those presently located in the solar neighborhood. In \\S 3 we analyze the orbital motions of these stars. In \\S 4 we consider the global character of the halo of the Galaxy, as deduced from the kinematics of a local sample. In \\S 5, we further examine evidence for kinematic substructure in the phase-space distribution of the halo. Finally, in \\S 6, the results of the present work are summarized, and their implications for the formation and evolution of the Galaxy are discussed. ", "conclusions": "We have analyzed both the local and global kinematics of 1203 metal-poor stars in the Galaxy with [Fe/H] $\\le-0.6$, based on a large, revised, catalog of stars selected without kinematic bias (Paper II). All of these stars have available distance estimates, radial velocities, proper motions, and abundance estimates over the full applicable range in the Milky Way. This is the largest non-kinematically selected sample yet assembled, so the derived kinematic properties are the least affected by systematics as well as statistical fluctuations. We summarize our results below, and discuss them in the context of the formation of the Galaxy. \\subsection{Summary of the Results} The local kinematics of the halo population, based on the stars with [Fe/H] $ \\le-2.2$ and $|Z|<1$ kpc, are characterized by a radially elongated velocity ellipsoid $(\\sigma_U,\\sigma_V,\\sigma_W)=(141 \\pm 11,106 \\pm 9,94 \\pm 8)$ km~s$^{-1}$ and a small prograde rotation $=30 \\sim 50$ km~s$^{-1}$ (assuming $V_{LSR}=220$ km~s$^{-1}$). When additional halo stars at larger $|Z|$ are taken into account, the velocity ellipsoid remains essentially unchanged, but $$ exhibits a marked decrease (Figs. 2 and 3). We find no evidence for an increase of $\\sigma_W$ at the lowest abundances, as had been previously suggested. At higher metallicities, the stars in our sample exhibit disk-like kinematics, and a higher mean rotation. Specifically, for stars in the abundance interval $-0.7\\le$ [Fe/H] $<-0.6$ and with $|Z|<1$ kpc, we have obtained $(\\sigma_U,\\sigma_V,\\sigma_W) = (46 \\pm 4,50 \\pm 4,35 \\pm 3)$ km~s$^{-1}$ and $=200$ km~s$^{-1}$, which characterize the kinematic parameters of the thick disk. We have also confirmed previous results that there exists a remarkable discontinuity of the rotational properties of the Galaxy at [Fe/H] $\\simeq -1.7$ (Fig. 3). Analysis of a large sample of non-kinematically selected stars provides clear evidence, supporting earlier suspicions based on much smaller samples, that there exists {\\it no correlation between metal abundances and orbital eccentricities} for metal-poor stars of the Milky Way (Fig. 5). Even at the lowest abundances explored in our sample, [Fe/H] $\\le-2.2$, about 20\\% of the stars have $e<0.4$. In addition, there is a small concentration of high-$e$ stars at [Fe/H] $\\sim-1.7$, which is possibly responsible for the near zero $$ at the same metallicity. We found that the fraction of the low-eccentricity stars with [Fe/H] $\\le-2.2$ remains the same, even as one changes the range of $|Z|$ (Fig. 6a), so such stars belong to the halo, not the MWTD component. On the other hand, stars in intermediate abundance ranges above $-2.2$ dex exhibit a decrease of low-$e$ stars with increasing $|Z|$, and the fraction of such stars appears to converge at high $|Z|$ to that found for [Fe/H] $\\le-2.2$ (Fig. 6b). Both a KS test and a Monte Carlo simulation enable a determination of the structural parameters of the disk component in these abundance ranges. Specifically, the fraction of the disk component is about 30\\% for $-1.7<$ [Fe/H] $\\le-1$, but is less than 10\\% for more metal-poor ranges (Fig. 8). The global kinematics of the halo stars are summarized as follows. In contrast to the claims of Majewski (1992), and Carney et al. (1996), stars in our sample do {\\it not} show a net retrograde rotation at large $Z_{max}$, but rather exhibit a near zero systematic rotation (Fig. 10). The difference between our result and that of Carney et al. (1996) probably arises from the (unavoidable) kinematic bias inherent in their sample selection criteria. The observed decrease of $$ with increasing $Z_{max}$ is continuous, so that it is not possible to conclude that the inner ``contracted'' population (with a positive $$) is distinct from an outer ``accreted'' population, based on the rotational properties of the metal-poor stars alone. Our analysis of the global density distribution of halo stars, based on the reconstruction method developed by SLZ, confirms SLZ's conclusion that the outer halo is quite round (Fig. 12). However, we see no evidence of an overlapping flattened component in addition to the main, nearly spherical one, as was claimed by SLZ. Rather, the density distribution of the halo is better described as nearly spherical in the outer region (beyond $R=15-20$ kpc) and highly flattened in the inner region. We have confirmed a recent detection of kinematic substructure in the solar neighborhood by HWdZZ, based on a small number of stars which cluster together in the halo angular momentum diagram. We have also found an additional elongated ``trail'' which appears to connect between HWdZZ's ``clump'' and the high $L_z$ region (Fig. 14). Further analysis, using several integrals of motion for the ``clump,'' does not result in a dramatic increase in the numbers of stars associated with it, even though the total number of our sample stars is three times as large as that available to HWdZZ. \\subsection{Implications for the Formation of the Galaxy} The local and global kinematics of metal-poor stars provide valuable clues for understanding the formation process of the halo and thick disk components in the Galaxy, as well as in disk-type galaxies in general. If the primordial collapse from the halo to the disk occurred in a monolithic manner, starting from an overdense homogeneous spheroid, one might expect (as predicted by the ELS model) to observe a continuous increase of $$ for the stars born from the infalling gas, as well as a continuous decrease of their orbital eccentricities with increasing [Fe/H] as the spheroid spins up in order to conserve angular momentum. The fact that we observe no correlation between [Fe/H] and $e$, and basically no change of $$ with abundance for stars with [Fe/H] $\\le-1.7$ conflicts with this scenario. The lack of an abundance gradient in the halo stars (Carney et al. 1990; CY) is also difficult to interpret in this context. The outer halo, if formed from a monolithic collapse, might be expected to be dominated by radially elongated motions of the stars, but this is actually opposite to the inferred tangentially anisotropic velocity ellipsoid at large distance from the Sun (see Sommer-Larsen et al. 1997). We also note that a small portion of the metal-poor stars having [Fe/H] $\\sim-1.7$ may have been formed from the infalling part of gas, so as to explain both the nearly zero $$ and the excess number of high-$e$ stars found at [Fe/H] $\\sim-1.7$. If the halo is assembled from merging and/or accretion of numerous fragments falling into the Galaxy (SZ), one might expect little or no correlation between kinematic and chemical properties, as each fragment has its own chemical history, and the merging process may proceed in a chaotic manner. Our results for the $$ vs. [Fe/H] and [Fe/H] vs. $e$ relations are basically in agreement with this scenario. The SZ scenario is also consistent with a several Gyr age spread in globular clusters in the outer halo (see, e.g., Rosenberg et al. 1999), and even in field stars (Schuster \\& Nissen 1989), because the initiation and duration of star formation may not be coherent from fragment to fragment\\footnote{Harris et al. (1997) showed that the most metal-poor globular clusters, such as M92, have essentially the same age everywhere in the halo. As they argued, this result could also be explained within the precepts of the SZ scenario if all of the ``SZ fragments'' began building the first generation of clusters in the same time period.}. However, the original SZ scenario seems unlikely to explain our observed vertical gradient of $$ for halo stars, as well as the highly flattened density distribution of the inner halo, in contrast to the nearly spherical outer halo. Totally incoherent, chaotic merging of SZ fragments would not be expected to produce these ``internal'' kinematic structures in the halo. It is also unclear as to how the rapidly rotating disk component subsequently formed out of the aftermath of merging (see also Freeman 1996). SZ first suggested that at least the inner part of the halo may have undergone a coherent contraction in a manner similar to the ELS hypothesis, an idea which has been invoked by subsequent workers to explain the duality of the density, kinematics, and ages of the halo field stars (e.g., SLZ; Norris 1994; Carney et al. 1996; Sommer-Larsen et al. 1997), as well as the age difference between outer and inner globular clusters (Zinn 1996; Rosenberg et al. 1999). This sort of hybrid picture, combining aspects of both the ELS and SZ scenarios, proposes that the outer halo is made up from merging and/or accretion of subgalactic objects, such as dwarf-type satellite galaxies, whereas the inner part of the halo has undergone a dissipative contraction on relatively short timescales. This hybrid model might explain our identification of the inner, flattened, slowly rotating component of the halo with a finite spatial gradient in $$. An alternative hypothesis to explain an observed ``duality'' of the Galactic halo relies on the existence of a thick-disk population of stars even at rather low abundances (MFF; Norris 1994; BSL). If stars with disk-like kinematics have a metallicity distribution which extends below [Fe/H] $=-2$, then a finite fraction of their orbits would be characterized by low $e$, as found in our present investigation. One possible origin of this MWTD component may be the heating of the pre-existing thin disk triggered by the dissipationless merging of a satellite falling into the disk (Quinn, Hernquist, \\& Fullagar 1993). Under this hypothesis, the currently observed thin disk, with a vertical scale height of $\\sim 350$ pc, could only have formed {\\it after} the merging event was completed. However, our finding that few thick-disk stars exist with [Fe/H] $\\le-1.7$, and no observed increase of $\\sigma_W$ with decreasing [Fe/H], belies the existence of a dynamically hot, proto-disk population at the lowest abundances (see also Norris 1994; Ryan \\& Lambert 1995; Twarog \\& Anthony-Twarog 1996). Furthermore, following the results presented in \\S 4.2, we see no evidence for an overlapping flattened component of the halo in addition to a nearly spherical component. An indication that there might exist a significant vertical gradient in $$ for [Fe/H] $\\le-1.7$, compared with a much smaller gradient observed for the thick disk itself, also conflicts with this hypothesis. Thus, we conclude that the formation of the inner flattened halo possibly involves a dissipative contraction, not a dissipationless heating of the proto-disk. If a hybrid halo formation picture, based on dissipationless merging in the outer halo and dissipative contraction in the inner halo, applies, the question arises as to whether there is a clear boundary distinguishing the two regions. The results presented in \\S 4 suggest no clear distinction between the outer and inner regions of the halo, at least as seen from inspection of the $$ vs. $Z_{max}$ relation, and the inferred globafl density distribution of the halo. Furthermore, there presently exists no reasonable theoretical explanation for the existence of two distinct populations of stars in the halo. Thus, our current analysis implies that both dissipationless and dissipative processes in the outer and inner halo, respectively, may have occurred more or less in a simultaneous manner. We now ask whether the above hybrid scenario is a natural consequence of the currently favored theory of galaxy formation based on hierarchical assembly of cold dark matter (CDM) halos (e.g., Peacock 1999). The CDM model postulates that initial density fluctuations in the early Universe have larger amplitudes on smaller scales. Thus, the initially overdense regions that end up forming large galaxies such as our own contain large density fluctuations on subgalactic scales. As a protogalaxy collapses from the general cosmological expansion, these small-scale fluctuations develop into numerous clumps of CDM particles, into which the interstellar gas falls from gravitational attraction. The protogalaxy is thus made up of numerous clumps comprised of a mixture of primordial gas and dark matter, interacting with one another via their mutual gravitational attraction. According to numerical simulations by Steinmetz \\& M\\\"uller (1994; 1995) and Bekki \\& Chiba (1999), most of the metal-poor stars which presently occupy the outer halo of our Galaxy form in these local, small-scale density fluctuations. Once star formation initiates, the gas inside of these small fragments quickly escapes due to energy feedback from supernovae. Later, these clumps begin to merge with one another, and the aftermath of these essentially dissipationless merging processes exhibits a nearly spherical density distribution with no abundance gradient. The subsequent evolution of the system may be described in the following way (Bekki \\& Chiba 1999). As a consequence of the merging of low-mass fragments, a smaller number of more massive clumps develops -- within each of these merged clumps one expects to find previously formed metal-poor stars as well as newly born stars. These large clumps continue to accrete gas from their immediate surroundings. These clumps gradually move toward the central region of the system due to both dynamical friction, and dissipative merging with smaller clumps. Then, when the last merging event between the largest clumps occurs, the stars which have been confined inside the clumps are disrupted and spread over the inner part of the halo, whereas a large fraction of the disrupted gas appears to end up in the center of the Galaxy and may form a bulge. As a consequence, the inner part of the halo should have a flattened density distribution with a finite prograde rotation, as reported here, and its angular momentum distribution may be similar to that of the bulge (Wyse \\& Gilmore 1992). Also, the stars born from this infalling stage of gas may explain the existence of high-$e$ stars at [Fe/H] $\\sim-1.7$. The simulations conducted to date imply that the thick disk component is partially composed of debris stars, but it is mainly made from diluted gas which has been accreted from the outer part of the halo (Sommer-Larsen et al. 1997). Therefore, although more detailed simulation work is required, CDM models appear to reproduce, at least qualitatively, the overall kinematic properties of the metal-poor stars via both dissipationless merging in the outer halo, and dissipative merging in the inner halo. It is unknown whether or not evidence for merging events of CDM clumps during the early evolution of the Galaxy might be still preserved as kinematic substructures at the current epoch. Within the currently available precision of space velocities for stars in our sample, typically of the order of 10 to 20 km~s$^{-1}$, the main body of the halo appears to be well mixed in phase space; higher precision measurements of proper motions by the planned astrometric satellites missions (e.g., {\\it FAME, SIM, GAIA}) may be able to disentangle this complex mixture of halo stars (Helmi, Zhao \\& de Zeeuw 1999). Alternatively, the confirmed kinematic clumping of halo stars presented in \\S 5 may originate from the recent accretion of a satellite galaxy, which has fallen into the Galaxy after the major part of the halo was formed. Firmer conclusions on the formation of the Galaxy require the assembly and analysis of still larger numbers of stars with accurate distances and proper motions, especially at larger distances from the Sun. Exploration along this line is now in progress. More elaborate numerical modeling of the formation of large spiral galaxies such as the Milky Way is also needed in order to clarify the physical processes that lead to the currently observed dynamics and structure of the halo and disk components. It is of particular importance to model and understand the chemo-dynamical evolution of the system of subgalactic fragments in the course of the Galaxy's collapse. Once a fundamental understanding of the formation and evolution of {\\it our} Galaxy is established, it will then be possible to obtain additional insights into formation of disk-type galaxies in general, by combining our refined picture with the rapidly growing observational database of young galaxies becoming available in the deep realm of the Universe." }, "0003/astro-ph0003202_arXiv.txt": { "abstract": "We analyzed nine X-ray bright Virgo early-type galaxies observed by both ASCA and ROSAT. Through spatially resolved spectroscopy, we determined the radial temperature profile and abundances of Mg, Si and Fe for six galaxies. The temperature profiles are consistent with isothermal temperatures outside of a cooler region at the galaxy center. We present new evidence for iron abundance gradients in NGC4472 and NGC4649 and confirm the previous results on NGC4636. Mg and Si abundance gradients on average are flatter compared to those of iron and correspond to an underabundance of alpha-process elements at high Fe values, while at low iron, the element ratios favor enrichment by type II SNe. We explain the observed trend by the metallicity dependence of SN~Ia metal production and present constraints on the available theoretical modeling for low-metallicity inhibition of SNe~Ia (Kobayashi \\etal 1998). Our results imply a cut-off metallicity in the range 0.07--0.3 solar and require a lower limit of 0.3 solar on the Fe contribution of SN~Ia. We estimate an SN~Ia rate at the centers of the brightest galaxies in our sample of $\\sim0.08h_{75}^{3}$ SNU (supernova units). The rates inferred from optical searches should be corrected for the presence of ``faint'' SN Ia events, since these release limited metals and therefore do not contribute significantly to the measured metallicity in the X-ray gas. With this correction the present-epoch SN~Ia rate in early-type galaxies is $0.10\\pm0.06$ $h_{75}^{2}$ SNU (Cappellaro \\etal 1997) and is therefore comparable with the X-ray estimates. A simple comparison shows the X-ray abundances we derive are still discrepant from optically determined values. We attribute this difference to the low spatial resolution of our X-ray measurements, radial gradients in the abundances and the importance of hydrodynamical effects, particularly the inflow of cooling gas, on the measured X-ray abundances. ", "introduction": "X-ray observations of the hot interstellar gas in early-type galaxies provide a unique tool for the study of stellar metallicities and SN~Ia rates within distances of a few effective radii from the galaxy's center. Detailed comparison of the metallicities determined from X-ray and optical observations, as was done for NGC5846 (Finoguenov \\etal 1999, hereinafter F99), demonstrates good agreement. The abundance pattern detected through X-ray spectroscopy in low-mass systems like groups and early-type galaxies favors a dominance of SNe~Ia in the enrichment of the hot gas in these systems (F99, Finoguenov \\& Ponman 1999, hereinafter FP99). During the early stages of elliptical galaxy formation, SN~II products are likely to escape the galaxy, but be contained by the potentials of clusters of galaxies (see Fukazawa \\etal 1998 for a recent compilation of X-ray data, hereinafter Fu98). The hot gaseous halos of ellipticals contain elements released via stellar mass loss, plus elements synthesized in SN~Ia events, {\\it after} the cessation of early star formation (\\eg\\ Renzini \\etal 1993), which is used to constrain the present-day SN~Ia rate. In previous X-ray studies, though, limits on SN~Ia activity generally were derived without regard to the metallicity of the progenitor stars. While optical searches find only a handful of SNe~Ia in early-type galaxies (Cappellaro \\etal 1997), through X-ray observations we can study the rate of SN~Ia events as a function of the metallicity of the host galaxy. This is the {\\it first Paper} in our project to study the X-ray properties of Virgo galaxies, observed by ASCA. The main goal is to present the results of gas temperature and element abundance measurements, and to compare these with current observations and theoretical models. Our approach provides a full treatment of the instrumental effects, including PSF, as well as gas projection effects, on the derivation of the spatially resolved temperature and abundance structure. We consider separately the behavior of iron and some of the $\\alpha$-process elements, that enables us to separate the contribution from different types of SN. Discussion of a number of important quantities, including the luminosity of the hard galactic component is postponed for the next paper, while we await the completion of our observational program on low-luminosity Virgo early-types, that should allow us to present the results for a complete optically selected sample. This {\\it Paper} is organized as follows: in section \\ref{sec:dr} we describe the X-ray analysis, where in subsections we discuss our measurements and comment on individual galaxies; in section \\ref{sec:disc-n} we attempt a detailed study of the abundance pattern, measured for our sample. Through a comparison of our results with abundance measurements for the stars in our Galaxy, as well as for the stellar population of cluster early-type galaxies, we propose a progenitor star metallicity dependence for SN~Ia metal yields. In this context, we discuss the model for low-metallicity inhibition of SN~Ia's (Kobayashi \\etal 1998). ", "conclusions": "" }, "0003/astro-ph0003458_arXiv.txt": { "abstract": "s{ We have evaluated the observational constraints on the spectral index $n$, in the context of a $\\Lambda$CDM model. For $n$ scale-independent, as predicted by most models of inflation, present data require $n\\simeq 1.0 \\pm 0.1$ at the 2-$\\sigma$ level. We have also studied the two-parameter scale-dependent spectral index, predicted by running-mass inflation models. Present data allow significant variation of $n$ in this case, within the theoretically preferred region of parameter space.} ", "introduction": "It is generally supposed that structure in the Universe originates from a primordial gaussian curvature perturbation, generated by the quantum fluctuations of the inflaton field during slow-roll inflation. Then the spectrum of the curvature perturbation $\\delta_H (k)$ is determined by the inflaton potential $V(\\phi)$. In this paper we will consider the scale--dependence of the primordial spectrum, defined by the spectral index $n$: \\be n(k)-1\\equiv 2 {\\partial \\ln \\delta_H\\over \\partial \\ln k} = 2M_{Pl}^2 (V''/V)-3M_{Pl}^2 (V'/V)^2 \\, , \\ee where\\footnote{$M_{Pl}=2.4\\times 10^{18}\\mbox{GeV}$ is the Planck mass, $a$ is the scale factor and $H=\\dot a/a$ is the Hubble parameter, and $k/a$ is the wavenumber.} the potential and its derivatives are evaluated at the epoch of horizon exit $k=aH$. The value of $\\phi$ at this epoch is given by $ \\ln(k_{end}/k)=N(\\phi) =M_{Pl}^{-2}\\int^\\phi_{\\phi_{end}} (V/V') d\\phi\\,$ , where $k_{end}$ is the scale leaving the horizon at the end of slow roll inflation and $N(\\phi)$ is the number of e-folds. In the majority of the inflation models, $n$ is practically scale-independent so that $\\delta_H^2\\propto k^{n-1}$, but we shall also discuss an interesting class of models giving significant scale dependence. ", "conclusions": "We have evaluated the observational constraints on the spectral index $n$, using a range of data, and we find, for constant $n$ at 2-$\\sigma$ level, $0.88\\leq n\\leq 1.11$ for $0 \\leq z_R \\leq 20$. We have also investigated the running mass models, parameterized by $c$ and $\\sigma$. For $c$ and $\\sigma$ with the same sign, we have found that indeed $n$ can vary by about $0.05$ between the COBE scale and $8 h^{-1} \\mbox{Mpc}$. Moreover, if $c$ is positive as it would be for a gauge coupling, $n-1$ {\\em can change sign between the COBE and $8h^{-1}\\,\\mbox{Mpc}$ scales}. It will be very interesting to see how the present situation changes with the advent of better data." }, "0003/astro-ph0003172_arXiv.txt": { "abstract": "It is proposed that a rearrangement of single-particle degrees of freedom may occur in a portion of the quantum fluid interior of a neutron star. Such a rearrangement is associated with the pronounced softening of the spin-isospin collective mode which, under increasing density, leads to pion condensation. Arguments and estimates based on fundamental relations of many-body theory show that one realization of this phenomenon could produce very rapid cooling of the star via a direct nucleon Urca process displaying a $T^5$ dependence on temperature. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003485_arXiv.txt": { "abstract": "s{ Cosmic ray antiprotons have been detected for over 20 years and are now measured reliably. Standard particle and astrophysics predict a conventional spectrum and abundance of secondary antiprotons consistent with all current measurements. These measurements place limits on exotic Galactic antiproton sources and non-standard antiproton properties. Complications arise, particularly at low energies, with heliospheric modulation of cosmic ray fluxes and production of standard secondaries from $A >$ 1 nuclear targets. Future experiments and theoretical developments are discussed.} ", "introduction": " ", "conclusions": "} \\eightrm \\begin{tabular}{|lc|c|c|c|c|c|c|c|} \\hline Experiment& &Field&Flight&KE Range&Cand-&Back-&Observed &Predict-\\\\ & &Pol.$^{\\rm a}$&Date&(GeV)&idates&ground&$\\pbar /p$ Ratio& tion$^{\\rm b}$\\\\ \\hline Golden et al. 1979$^\\dag$&&$+$&June 1979&5.6 -- 12.5&46&18.3& $(5.2\\pm 1.5)\\times 10^{-4}$& -- \\\\ \\hline Bogomolov et al. 1979$^\\dag$&&$+$&1972-1977&2.0 -- 5.0&2&--& $(6\\pm 4)\\times 10^{-4}$& -- \\\\ Bogomolov et al. 1987$^\\ddag$&&$-$&1984-1985&0.2 -- 2.0&1&--& $(6^{+14}_{-5})\\times 10^{-5}$& -- \\\\ Bogomolov et al. 1990$^\\ddag$&&$-$&1986-1988&2.0 -- 5.0&3&--& $(2.4^{+2.4}_{-1.3})\\times 10^{-4}$& -- \\\\ \\hline MASS91~\\protect\\cite{MASS91a}&&$+$&Sep. 1991&3.70--19.08&11&3.3& $(1.24^{+0.68}_{-0.51})\\times 10^{-4}$&$1.3\\times 10^{-4}$\\\\ \\hline IMAX$^{\\ddag}$~\\protect\\cite{IMAX}&&$+$&July 1992&0.25 -- 1.0&3&0.3& $(3.14^{+3.4}_{-1.9})\\times 10^{-5}$&$1.5\\times 10^{-5}$\\\\ IMAX~\\protect\\cite{IMAX}&&$+$&July 1992&1.0 -- 2.6 &8&1.9& $(5.36^{+3.5}_{-2.4})\\times 10^{-5}$&$6.5\\times 10^{-5}$\\\\ IMAX~\\protect\\cite{IMAX}&&$+$&July 1992&2.6 -- 3.2 &5&1.2& $(1.94^{+1.8}_{-1.1})\\times 10^{-4}$&$1.1\\times 10^{-4}$\\\\ \\hline BESS93$^{\\ddag}$~\\protect\\cite{BESS93}&&$+$&July 1993&0.20 -- 0.60&7&$ \\sim 1.4$&$(5.2^{+4.4}_{-2.8})\\times 10^{-6}$&$8.9\\times 10^{-6}$\\\\ \\hline CAPRICE~\\protect\\cite{CAPRICE94}&&$+$&Aug. 1994&0.6 -- 2.0 &4&1.5& $(2.5^{+3.2}_{-1.9})\\times 10^{-5}$&$3.5\\times 10^{-5}$\\\\ CAPRICE~\\protect\\cite{CAPRICE94}&&$+$&Aug. 1994&2.0 -- 3.2 &5&1.3& $(1.9^{+1.6}_{-1.0})\\times 10^{-4}$&$1.1\\times 10^{-4}$\\\\ \\hline BESS95$^{\\ddag}$$^\\ast$~\\protect\\cite{BESS95}&&$+$&July 1995&0.175 -- 0.3&3&0.17& $(7.8^{+8.3}_{-4.8})\\times 10^{-6}$&$-$\\\\ BESS95$^{\\ddag}$$^\\ast$~\\protect\\cite{BESS95}&&$+$&July 1995&0.3 -- 0.5&7&0.78& $(7.4^{+4.7}_{-3.3})\\times 10^{-6}$&$1.1\\times 10^{-5}$\\\\ BESS95$^{\\ast}$~\\protect\\cite{BESS95}&&$+$&July 1995&0.5 -- 0.7&7&1.4& $(7.7^{+5.3}_{-3.7})\\times 10^{-6}$&$5.5\\times 10^{-6}$\\\\ BESS95$^{\\ast}$~\\protect\\cite{BESS95}&&$+$&July 1995&0.7 -- 1.0&11&2.8& $(1.01^{+5.7}_{-4.3})\\times 10^{-5}$&$1.3\\times 10^{-5}$\\\\ BESS95$^{\\ast}$~\\protect\\cite{BESS95}&&$+$&July 1995&1.0 -- 1.4&15&3.5& $(1.99^{+0.91}_{-0.73})\\times 10^{-5}$&$3.1\\times 10^{-5}$\\\\ \\hline \\end{tabular} \\end{center} \\vspace{0.25cm} \\indent{${}^{\\rm a}$~Northern hemisphere heliomagnetic polarity: + = outward field.}\\\\ \\indent{${}^{\\rm b}$~ILDM prediction with $V_W({\\rm eq/polar})$ = 400/750 km sec$^{-1}$, $B_\\odot$ = 4.5 nT.}\\\\ \\indent{${}^\\dag$~Not shown in Figure~\\ref{fig:unmodRatio} or used in analysis. ${}^\\ddag$~Not used in analysis. ${}^\\ast$~Statistical and systematic uncertainties on ratio added in quadrature.} \\label{balloon_tab} \\end{table} \\begin{figure} \\epsfxsize=25pc % \\epsfbox{Hep00-3Fig7.eps} % \\caption{Measured $\\pbar /p$ spectral flux ratios, compared with {\\em unmodulated} IS CR $\\pbar$ flux predicted by ILDM (see text).~\\protect\\cite{GeerKenn}\\label{fig:unmodRatio}} \\end{figure} \\begin{figure} \\epsfxsize=25pc % \\epsfbox{Hep00-3Fig8.eps} % \\caption{Measured $\\pbar /p$ spectral flux ratios corrected with heliospheric modulation to common epoch (July 1995), compared with modulated prediction of IS CR $\\pbar$ flux in ILDM (see text).~\\protect\\cite{GeerKenn}\\label{fig:modRatio}} \\end{figure} The analysis presented here in based on all refereed and published measurements not contradicted by later measurements with better detectors (Table~\\ref{tab:Summary}).~\\cite{GeerKenn} Figure~\\ref{fig:unmodRatio} shows the selected measurements compared with the ILDM prediction for IS fluxes. The disagreement evident in the figure is explicable by heliospheric modulation. Figure~\\ref{fig:modRatio} compares the modulated ILDM predictions with the measured fluxes. This figure makes the comparison by renormalizing the measured fluxes to a single epoch (July 1995, chosen as roughly the most recent heliomagnetic minimum) and using the prediction for that epoch. Our analysis did not use measurements with $K <$ 500 MeV because of the large and difficult-to-calculate diffusion modulation in that energy range. \\subsection{Implications~\\protect\\cite{GeerKenn,KennCPT}} The most basic result implied by Table~\\ref{tab:Summary} and Figure~\\ref{fig:modRatio} is the standard $\\pbar$ secondary flux alone, from a realistic ILDM, can account for the observed flux in the relevant energy range, within uncertainties. If variant Galactic transport mechanisms (such as reacceleration or shrouded sources~\\cite{Gaisser}) or exotic $\\pbar$ sources are at work in this $K$ range, their effects are too small to see at this time. (A hint of reacceleration may be visible in the range $K\\simeq$ 2--5 GeV by distortion of the spectrum evident in Figure~\\ref{fig:modRatio}, but the effect is not significant within uncertainties.) A second, less obvious, result is a limit on the intrinsic decay lifetime of the antiproton: $\\tau_{\\pbar} >$ 0.8 Myr, the best limit currently feasible. While the exclusion of the $K <$ 500 MeV spectrum does {\\em not} significantly affect the $\\tau_{\\pbar}$ limit, it does limit conclusions about the absence of exotic $\\pbar$ sources, as these would have their largest effect relative to the standard secondaries precisely at such low $K$. \\begin{table} \\caption{$CPT$-- and $B$--violating scale limits associated with $p$ lifetime $\\taup$ = $10^{32}$ yr and $\\pbar$ lifetime $\\tpbar$ = $10^7$ yr (see text).~\\protect\\cite{KennCPT} \\label{tab:CPTscales}} \\vspace{0.2cm} \\begin{center} \\footnotesize \\begin{tabular}{|c|c|c|c|} \\hline $n$ & $M_X$ (GeV) & $k$ & $M_Y$ (GeV) \\\\ \\hline & & & \\\\ 5 & $2\\times 10^{19}$ & 1 & $2\\times 10^{63}$\\\\ 6 & $4\\times 10^9$ & 2 & $5\\times 10^{31}$\\\\ 7 & $3\\times 10^6$ & 3 & $1\\times 10^{21}$\\\\ 8 & $6\\times 10^4$ & 4 & $7\\times 10^{15}$\\\\ 9 & $7\\times 10^3$ & 5 & $5\\times 10^{12}$\\\\ 10 & $2\\times 10^3$ & 6 & $4\\times 10^{10}$\\\\ \\hline \\end{tabular} \\end{center} \\end{table} A short $\\pbar$ lifetime $\\tau_{\\pbar}\\lesssim$ 10 Myr (Galactic CR storage time) would of course indicate CPT violation. The two pictures of CPT violation introduced in subsection~\\ref{sec:cpt} are: modification of LRQFT within ordinary quantum mechanics, and non-standard quantum mechanics (NSQM) with non-unitary time evolution. If only one new mass scale is relevant to the CPT violation, lower limits can be placed on such scales. In Table~\\ref{tab:CPTscales}, the limiting CPT-violating scales associated with modified QFT $(M_X)$ and NSQM $(M_Y)$ are shown, assuming $\\tau_{\\pbar}$ = 10 Myr. The $\\pbar$ lifetime is assumed related to each scale by simple mass dimensions. For modified QFT, $\\Gamma_{\\pbar}$ = $m_p(m_p/M_X)^n$; while for NSQM, $\\Gamma_{\\pbar}$ = $(m_p/2)(m_p/M_Y)^k$. It is interesting to note that the {\\em largest} $M_X$ lower bound is ${\\cal O}(M_{\\rm Pl})$, while the scales of order the ``intermediate'' scale $(10^8$--$10^{12}$ GeV) are possible, as well as scales $\\sim$ TeV. The last scale may not be unreasonably low in the context of ``large'' extra dimensional gravity.~\\cite{LargeExtraDim}" }, "0003/astro-ph0003166_arXiv.txt": { "abstract": "Gamma-ray emission from classical novae is dominated, during the first hours, by positron annihilation resulting from the beta decay of radioactive nuclei. The main contribution comes from the decay of \\fo\\ and hence is directly related to \\fo\\ formation during the outburst. A good knowledge of the nuclear reaction rates of production and destruction of \\fo\\ is required to study \\fo\\ synthesis in novae and the resulting gamma-ray emission. The rates relevant for the main mode of \\fo\\ destruction (i.e, through proton captures) have been the object of many recent experiments. However, subsequent analyses were focused on providing rates for X-ray burst nucleosynthesis not valid at nova temperatures (lower than 3.5\\power{8}~K). Accordingly, it is crucial to propose and discuss new reaction rates, incorporating all new experimental results, down to the domain of nova nucleosynthesis. We show that in this temperature regime, the \\pg\\ and \\pa\\ reaction rates remain uncertain and deserve further experimental and theoretical efforts. Our hydrodynamic calculations including the new nuclear rates demonstrate that their impact on \\fo\\ synthesis in nova explosions is quite large and, consequently, the early gamma-ray emission from classical novae is also affected. ", "introduction": "Classical novae emit gamma-ray radiation at and below 511 keV during the early epochs after the explosion. This emission is produced by electron-positron annihilation in the expanding envelope, and the subsequent Comptonization of the resulting gamma-ray photons, and it shows a line, at 511 keV, and a continuum, between 20 and 511 keV (\\cite{Gom98}). The positrons responsible for this emission come mainly from the disintegration of \\fo\\ (\\cite{Lei87}; \\cite{Gom98}), because its lifetime ($\\tau$=158 min) is such that positrons are emitted at the ``right time\", i.e., when the expanding envelope starts to be transparent to gamma-ray radiation. Therefore, the amount of radiation emitted strongly depends on the \\fo\\ content of the nova envelope. The synthesis of \\fo\\ in novae depends largely on some key nuclear reaction rates of \\fo\\ destruction and production which are far from being well known. This is the case, in particular, of the \\pg\\ and \\pa\\ reactions. Recent experimental studies (\\cite{Gra97}; \\cite{Utk98}) drastically improved the knowledge of these reaction rates with respect to previous studies (\\cite{WK82}). In a recent paper (\\cite{Her99}), we have analyzed the influence of these rates (\\cite{Utk98}) of \\pg\\ and \\pa\\ on the final yields of \\fo\\ for different models of CO and ONe novae. The effect of the new rates was important, since a factor of 10 reduction in the yields and in the resulting gamma-ray fluxes was obtained for all the models. Therefore, we concluded that a more detailed analysis of the reaction rates was necessary, in order to predict the gamma-ray emission of classical novae. The rates proposed by Utku et al. (1998) were limited to relatively high resonance energies and temperature domains, more appropriate for temperatures typical of X--ray bursts than for those of classical novae. The reason is that for their determination neither the influence of uncertainties on low energy resonance strengths (based on assumed reduced width) nor the effect of {\\em all} low energy resonance tails were considered. At higher temperatures the rates are more reliable since their main contributions come from two directly measured resonances. The purpose of this paper is to provide rates for the $^{18}$F(p,$\\gamma)^{19}$Ne and $^{18}$F(p,$\\alpha)^{15}$O reactions valid in the domain of temperature of nova nucleosynthesis, incorporating the latest experimental data. In addition to the nominal rate, we provide upper and lower limits. Since other nuclear reactions also affect \\fo\\ synthesis in novae, a global analysis is done, including results from the recent NACRE compilation (\\cite{NACRE}). With the new rates affecting \\fo\\ synthesis, new nova models have been computed, in order to determine the mass of \\fo\\ they eject and the impact of the new yields on their early gamma-ray emission. The organisation of this paper is the following. In section \\ref{s:synt} we describe the nucleosynthesis of \\fo. In section \\ref{s:frates} we discuss in detail the \\fo+p rates and the corresponding uncertainties while in the following section (\\ref{s:orates}), we briefly discuss other recently published rates. In section \\ref{s:yields} we present new results that show the influence of these new rates on \\fo\\ production. Following the conclusion, an appendix gives some analytical approximations to the rates. ", "conclusions": "We have investigated the \\fo\\ formation and destruction in nova outbursts, identified the key reactions (proton capture on \\fo\\ and \\ox) and analysed their rates. The proton capture rates on \\fo\\ are higher than the Wiescher \\& Kettner (1982) ones at nova temperatures due to the contribution of the tail of the 659 keV resonance whose large measured width (\\cite{Cos95}; \\cite{Gra97}) has been indirectly confirmed (\\cite{But98}). Another important contribution comes from the tail of the 38~keV resonance which was neglected in previous studies. Its {\\em nominal} contribution is larger than the 659 keV one but is proportional to its {\\em assumed} reduced proton width $\\theta^2_p$. The strengths (proton widths) of the low lying resonances are unknown and induce large uncertainties (factors of 100 to 1000) in the rates at nova temperatures. We have provided updated nominal rates for the two capture reactions together with upper and lower limits. The \\ox+p rates also display some large uncertainties at nova temperatures according to the recent compilation of Angulo et al. (1999). We have used these new nuclear physics results in a fully hydrodynamical nova code to calculate the \\fo\\ yields in novae for different rates : our low, nominal and high \\fo+p rates, and the low, recommended and high \\ox+p rates from Angulo et al. (1999). These results have been compared with models computed with the old Wiescher \\& Kettner (1982) rates (\\cite{Jos98}), and with more recent models with the Utku et al. (1998) rates (\\cite{Her99}). Two important results have been obtained. First, there is always a reduction of the amount of \\fo\\ synthesized in a nova explosion, with the nominal rates for the \\fo+p reactions both alone and combined with the recommended rates for the \\ox+p reactions. The nominal \\fo\\ yield (nominal \\fo+p and recommended for \\ox+p) is 4.84\\power{-5} by mass, which is 60 times smaller than the one obtained with Wiescher \\& Kettner (1982) rates and 6 times smaller than the one with Utku et al. (1998) rates (and Caughlan \\& Fowler (1988) and Landr\\'e et al. (1989) for \\ox+p). The impact on the early gamma-ray spectrum of the nova is a reduction of the flux by the same amount (with respect to G\\'omez-Gomar et al. (1998) and Hernanz et al. (1999), respectively). Second, the yields are found to be very sensitive to the rates with resulting combined (\\fo+p and \\ox+p) uncertainties of more than three orders of magnitude. This supports the need of new experimental and theoretical studies to improve the knowledge of the \\fo+p and \\ox+p rates and, consequently, allow for a larger reliability of the predictions of annihilation gamma--ray fluxes from novae, to be observed by current and future instruments." }, "0003/astro-ph0003020_arXiv.txt": { "abstract": "Spectral fits to X-ray data from both NGC\\,253 and M\\,82 provide ambiguous results. The so-called ``best fit'' results depend on the instrument with which the data were obtained and obviously on the choice of spectral model composition. We show that different spectral models can be fit equally well to {\\it BeppoSAX} data of both galaxies. Metallicities are unreliable in general, with a strong dependence on the choice of model. Preference to one particular spectral model can only be given by combining spectroscopic and imaging X-ray data from all available satellites ({\\it ROSAT}, {\\it ASCA}, and {\\it BeppoSAX}). Based on spectra of NGC\\,253, we demonstrate that a model consisting of two or more thermal plasma components plus a hard power law continuum and Fe K$\\alpha$ line emission can explain all observations. These model components represent the integral spectrum of thermal gas and compact sources in starburst galaxies that are most likely supernova remnants and X-ray binaries. The same model can fit the X-ray data of M\\,82, but there the evidence, from {\\it ROSAT}\\ imaging, of the existence of compact sources which might represent high-mass X-ray binaries is weaker. This implies that its hard X-ray emission, which is extended in {\\it ROSAT}\\ images, might--if truly diffuse--be dominated by a very hot (several keV energy) thermal gas component. ", "introduction": "\\label{par:intro} There have been a number of papers recently discussing {\\it ASCA} and {\\it ROSAT}\\ X-ray observations of nearby starburst galaxies, in particular the two best-studied systems NGC\\,253 and M\\,82, with different interpretations of the resulting spectral fits (Moran \\& Lehnert 1997; Ptak \\etal\\ 1997; Strickland \\etal\\ 1997; Tsuru \\etal\\ 1997; Vogler \\& Pietsch 1999). All authors agree on the general complexity of the X-ray properties, but--depending on the data and on the spectral models used for the spectral analysis--they reached different conclusions. Especially the choice of spectral model components and the resulting best-fitting element abundances are under debate. By treating all available {\\it imaging and spectroscopy} data from {\\it ASCA} and {\\it ROSAT}\\ (both HRI and PSPC) in a self-consistent manner, Dahlem \\etal\\ (1998; hereafter \\pone) and Weaver \\etal\\ (2000; hereafter \\ptwo) were able to reconcile the apparent discrepancies based on a mini-survey of 5 nearby edge-on starburst galaxies. The results from \\pone\\ and \\ptwo\\ suggest that the combined {\\it ASCA} and {\\it ROSAT}\\ PSPC integral spectra of NGC\\,253 and M\\,82 can be fit with comparable values of $\\chi^2$ by different combinations of spectral components, which means that there is an ambiguity in the choice of the best-fitting spectral model. By cross-checking the spectral results with {\\it ROSAT}\\ PSPC and HRI imaging data, a spectral composition of (at least) two thermal plasmas, with temperatures in the ranges 0.1--0.4 keV and 0.6--0.8 keV, respectively, plus a hard power law component turns out to be the only model combination that can explain {\\it all} observational data of all galaxies in the sample simultaneously (\\ptwo). This is in contrast with the recent findings by Cappi \\etal\\ (1999; hereafter C99), based on {\\it BeppoSAX} data of NGC\\,253 and M\\,82. These authors claim that there is ``compelling evidence for the presence of an extended hot thermal gas'' of several keV temperature in these two galaxies. The purpose of the current letter is to investigate this apparent discrepancy between their results and ours by re-analyzing the {\\it BeppoSAX} observations of NGC\\,253 and M\\,82, taking into account earlier results based on {\\it ROSAT}\\ and {\\it ASCA} observations. ", "conclusions": "\\subsection{NGC\\,253} We could reproduce the results by C99 within the uncertainties, using a spectral model with 2 Mekal plasma components (2M; see their Fig.~4). The fit to the {\\it BeppoSAX} data of NGC\\,253 following the model preferred by us, with a Mekal and a power law (M+P) component, is displayed in Fig.~\\ref{fig:kimfit}. This is evidently also an acceptable fit. The goodness of fit for our preferred model (\\ptwo) is $\\chi^2$ = 261.9 for 265 d.o.f. ($\\chi^2/\\nu$ = 0.99), while that for the 2 Mekal (hereafter ``2M'') model favored by C99 is $\\chi^2$ = 282.5 for 268 degrees of freedom (d.o.f.) and thus $\\chi^2/\\nu$ = 1.05. The M+P model fits the data better than the 2M model at the highest and lowest energies of the passband. The results of the two spectral fits to the {\\it BeppoSAX} data of NGC\\,253 are tabulated in Table~\\ref{tab:fits}. All uncertainties are given at the 90\\% confidence level for one interesting parameter; note that these only apply under the assumption that the choice of model components represents the different contributing emission mechanisms correctly. The softest thermal emission component found in the {\\it ROSAT}\\ data is not required. When including a thermal plasma of 0.26 keV energy, $\\chi^2$ is improved, but not significantly. Therefore it was left out in the fits to the {\\it BeppoSAX} data. The hard part of the X-ray spectrum can be fit with a power law that is compatible with those of Galactic X-ray binaries (XRBs) and can thus be explained naturally as the continuum emission from HMXRBs (\\ptwo). The integral spectrum of all the point sources from the {\\it ROSAT}\\ PSPC observations is consistent with this interpretation, as is the contribution of this spectral component to the total X-ray flux (\\pone). There is no reason why all {\\it compact} sources should emit a thermal spectrum. The claim that there is a hot thermal gas at a few keV energy (C99) hinges only on the assumption that this is the only mechanism that could explain the observed Fe line around 6.7 keV. However, supernova remnants (SNRs) and XRBs, including high-mass XRBs (HMXRBs), {\\it can} produce both fluorescent and thermal Fe line emission, i.e., at 6.4 keV and 6.7 keV, respectively (e.g., Nagase 1989; White, Nagase, \\& Parmar 1995; Liedahl \\etal\\ 1999), while C99 assume that all line emission is of thermal origin. Thus, the observed Fe line might well be a superposition of emission from hot gas and from X-ray binaries. The fitted equivalent width of the Fe line (Table~\\ref{tab:fits}) is in agreement with this interpretation. It, too, might be a composition of a (broad) thermal component and a (narrow) XRB contribution. However, a composite line fit cannot be performed based on the current data, because the line is only marginally resolved. The above finding that different models can fit the data equally well proves that the ``optimal'' fit is model-dependent, as already stated in \\ptwo. Therefore, there is no reason to reject the M+P model. Moreover, it is, as argued in \\ptwo, the physically most plausible model choice. Taking into account the {\\it ROSAT}\\ imaging results, which indicate clearly that there is a considerable number of unresolved compact sources in the central part of the disk of NGC\\,253 (\\pone), the most likely identification of these sources--based on their spectral properties and soft X-ray luminosities, $L_{\\rm X}$--is that they constitute a population of HMXRBs (\\pone). Thus, part of the emission distribution seen by {\\it BeppoSAX} is not truly ``diffuse'', but smeared out by its broad point-spread function. These point sources detected by {\\it ROSAT}\\ contribute about 50\\% of the flux from the central disk (\\pone\\ and \\ptwo). Thus, they are very significant contributors to the measured total flux, especially in the hard part of the X-ray spectrum. On the other hand, the spectral model preferred by C99 does {\\bf not} take into account the presence of HMXRBs and their spectral signature. Given the luminosities of emission mechanisms tracing the presence of high-mass stars in galaxies like NGC\\,253 and M\\,82, especially far-infrared radiation, a large number of HMXRBs must be expected to be present in them. It is still unclear how the previously detected X-ray emitting thermal plasma (with temperatures in the range of a few tenths of a keV) is heated, especially in the galaxy halos, up to several kpc away from the disk planes of the starbursts. The presence of another, extremely hot medium of several keV energy contributing of order 2/3 of the total 2--10 keV flux, as suggested by C99, would further excruciate the problem of energy supply. When taking into account the trade-off between metallicities and absorbing \\hi\\ column densities in fitting the softest part of X-ray spectra (which cannot be resolved by {\\it BeppoSAX} data only, but requires the low-energy response of {\\it ROSAT}), extreme subsolar metallicities, $Z$, are {\\bf not} required to obtain a good fit (\\ptwo). This $N_{\\rm H}$ vs. $Z$ dichotomy is another, independent ambiguity in the minimum $\\chi^2$ space of the spectral fits. Low metallicities in starburst galaxies, i.e., the galaxies with the highest star formation rates in the local Universe, would be hard to understand because of the proven presence of large numbers of massive stars, which are the most prolific producers of metals. \\subsection{M\\,82} The same ambiguities are present in fits to the {\\it BeppoSAX} data of M\\,82. We could fit almost equally well the model by C99 and ours from \\pone\\ and \\ptwo. Just as for NGC\\,253, the M+P model fits the data points at the very highest energies slightly better than a 2M model. The goodness of fit is 517.4/446 d.o.f. = 1.16 (2M model) and 466.9/442 d.o.f. = 1.06 (M+P model), respectively. Note that, just as for the combined {\\it ROSAT}\\ + {\\it ASCA} data, the {\\it BeppoSAX} data require another, soft thermal component to be added to the M+P model. With the M+P spectral model composition, we obtain almost equally good fits with two very different metallicities. In one case, $Z = 17 Z_\\odot$ (constrained to be $>2.2$ at the 90\\% confidence level), in the other $Z = 0.13 Z_\\odot$. In the high-metallicity case the flux at $\\sim 1$ keV is modeled primarily as Ne and Fe-L line emission, while in the low-metallicity case it is modeled as a peak in the thermal distribution. The energy resolution of the LECS of $\\sim 200$ eV ({\\it FWHM}) at 1 keV is insufficient to discriminate between the two options. Note that these two fits do not yet take into account the additional information obtained with {\\it ASCA} and {\\it ROSAT}\\ requiring an additional soft thermal component (\\pone\\ and \\ptwo). There is less evidence from {\\it ROSAT}\\ imaging for the existence of large numbers of compact sources in M\\,82. Instead, there appears to be a spatially extended, hard spectral component. Part of this might be truly diffuse, in which case the most likely interpretation is that of a very hot gaseous component, as suggested by C99. Only recently the {\\it Chandra} image by Griffiths et al. (2000) showed that there is indeed a population of compact sources in M\\,82, surrounded by diffuse emission. The compact sources in the central part of M\\,82 could not be resolved by {\\it ROSAT}, because they are too close to each other. The most likely identification is again that they are HMXRBs (Griffiths et al. 2000). Individual HMXRBs could also explain the observed X-ray variability in the hard part of the spectrum, while there is no evidence in the {\\it Chandra} data for the presence of an AGN (Ptak \\& Griffiths 1999, Matsumoto \\& Tsuru 1999, Gruber \\& Rephaeli 1999, C99). The measured position and equivalent width of the Fe line in M\\,82 of $6.63\\pm0.21$ keV and $60\\pm40$ eV, respectively, leaves open whether the line emission comes from either binaries or diffuse hot gas or a superposition of both. In both M\\,82 the width of the Fe line near 6.6 keV is unresolved. Thus, except for the (poorly constrained) position of the line centroid, no further information on the relative contribution of thermal or fluorescent line emission can be made based on the existing data. Both model compositions tested above fit the data (statistically) so well that no useful constraint can currently be derived on the possible contribution of both a hot thermal plasma and HMXRBs to the 2--10 keV flux of M\\,82, when fitted simultaneously." }, "0003/hep-ph0003045_arXiv.txt": { "abstract": "{Non-perturbative production of fermions after chaotic inflation has been the object of several studies in the very recent past. However, the results in the most interesting case of production of massive fermions in an expanding Universe were so far known only numerically. We provide very simple and readable analytical formulae, both for the spectra of the created fermions and for their total energy density. Their derivation is closely related to the one adopted for bosons and exploits the fact that the production occurs during very short intervals of non-adiabatical change of the fermionic frequency. Our formulae show the presence of resonance bands if the expansion of the Universe is neglected, and their disappearance when the latter is included. As in the bosonic case, this last effect is due to the stochastic character that the expansion gives to the whole process. Backreaction is considered in the last part of the work. All our analytical results are in excellent agreement with the previous numerical ones in the regime of validity of the latter. However, a more accurate scaling for the energy density of the produced fermions is here found.} ", "introduction": "\\label{sec:1} One of the key ingredients for our understanding of the early Universe is the mechanism of inflation~\\cite{guth}, which constitutes a very elegant solution to several cosmological problems. Despite of the simplicity of the general idea, the details of the physical processes which govern it are still somehow unclear and matter of intense work. The two main aims of these studies are \\begin{list}{}{\\setlength{\\labelwidth}{13pt}} \\item[(i)] to embed inflation in a context more motivated by particle physics (for a review see~\\cite{lr}) and \\item[(ii)] to understand the reheating phase which converts the energy density that drives inflation into the matter and radiation that we presently see. \\end{list} This second issue has been deeply influenced in the last decade by the possibility of particle creation through parametric resonance~\\cite{tb}. The application of this phenomenon to creation of matter after inflation has been called \\emph{preheating} in the paper~\\cite{kls}, since (with the exception of some very recent versions~\\cite{fkl}) it is usually followed by a stage of (ordinary) perturbative reheating. Preheating of bosons~\\cite{kls,stb} is characterized by a very efficient and explosive creation, due to the coherent effect of the oscillations of the inflaton field. This allows significant production even when single particle decay is kinematically forbidden. It has been very recently noticed~\\cite{gprt} that preheating of fermions can also be very efficient despite the production is in this case limited by Pauli blocking. Parametric creation of spin $1/2$ fermions has been the subject of some works in the past. Pure gravitational production has been examined in refs.~\\cite{mmf, kt}, while creation by an oscillating background field is instead considered in the works~\\cite{gmm,bhp,gk,fkl,gprt,bp}. References~\\cite{gmm,bhp} report results for creation in a Minkowski space. Reference~\\cite{gk} studies the production of massless fermions after a $\\lambda \\phi^4$ inflation, exploiting the fact that this case can also be reconducted to a static one. In this work, production in a static Universe after chaotic inflation is also considered, and some conjectures on the effects of the expansion are made. Moreover, the full calculation of preheating of massive fermions after chaotic inflation in an expanding Universe has been performed numerically in ref.~\\cite{gprt}. \\looseness=1 These last works had a great impact on the most recent studies. For example, their results turned also useful to the study of gravitinos production at preheating~\\cite{gra}. This issue is particularly important, since gravitinos can easily overclose the Universe (if they are stable) or (if they decay) spoil the successful predictions of primordial nucleosynthesis through photodissociation of the light elements. Gravitinos can be thermally produced during the stage of reheating. To avoid this overproduction, the reheating temperature $T_{\\mathrm{RH}}$ after inflation cannot be larger than $\\sim (10^{8}-10^{9})$~\\,~GeV~\\cite{grath}. However, it has been realized~\\cite{gra} that the non-thermal production of helicity $\\pm 1/2$ gravitinos (whose equation of motion can be reconducted to the one of an ordinary spin $1/2$ Dirac particle) can easily be more efficient than the thermal one, and this in general leads to more stringent upper bounds on $T_{\\mathrm{RH}}$. Several papers related to the works~\\cite{gra} have recently appeared~\\cite{gra2,grahyb}. Another important implication of preheating of fermions is constituted by leptogenesis, as the work~\\cite{gprt} and the related papers~\\cite{lepto} show. In this scheme~\\cite{fy}, a leptonic asymmetry is first created from the decay of right-handed neutrinos, and then partially converted to baryon asymmetry through sphaleronic interactions. Since leptogenesis is very sensitive to both the mechanism of creation of the heavy neutrinos and to the neutrino mass matrices, it could constitute an interesting link between preheating and the experiments on neutrino oscillations. Other phenomenological implications of these works appear in refs.~\\cite {high}, with preheating as a possible mechanism for creating superheavy relic particles responsible for the ultrahigh energy cosmic rays, and in ref.~\\cite{ckrt}, where the possible impact of fermions produced \\emph{during} inflation on the microwave background anisotropies and on the large-structure surveys is considered. Finally, fermionic production can play an interesting role in hybrid inflationary models~\\cite{grahyb, hyb2}. Due to the large number of these studies, it may be worth to reconsider the basic mechanism of fermionic preheating. In particular, it should be important to give an analytical confirmation to the results of production of massive fermions in the expanding Universe, which are so far known only numerically from the analysis~\\cite{gprt}. This is the aim of the present work. In the next section we revise the basic formalism for preheating of fermions. We consider creation of very massive particles right after chaotic inflation. The coupling of the fermions to the oscillating inflaton gives them a time varying mass. As it is known, this can cause a non-adiabatical change of the frequency of the fermions and their consequent creation. In case of very massive fermions, the non-adiabaticity condition can be satisfied only when their total mass vanishes, and the production occurs at discrete intervals, until the inflaton oscillations become too small for the total fermionic mass to vanish. \\looseness=-1 In section~\\ref{sec:3} we derive analytical formulae for the spectra of the fermions after a generic production. Our derivation follows the one developed in ref.~\\cite{kls} for preheating of bosons. It exploits the fact that the production occurs in very short intervals around the zeros of the total fermionic mass: the calculation is made possible from the fact that the occupation number can be considered as constant outside these small regions, and that the expansion of the Universe can be neglected inside them. As a result, the only physical quantities relevant for the creation are the time derivative~$\\phi'$ of the inflaton field and the value of the scale factor $a$ at each production. The derivation of the ``fermionic counterpart'' of the formulae obtained for preheating of bosons in ref.~\\cite{kls} has also been done to a certain extent in the work~\\cite{ckrt}, where the results of a single production during inflation is given. However, when one is interested in the successive productions, a more detailed study is necessary, as our analysis shows. As may be expected, the final results that we obtain closely resemble the ones of the work~\\cite{kls}. What is most surprising is their excellent agreement with the numerical results, as some figures provided manifest. In section~\\ref{sec:4} we consider the production in a non-expanding Universe. In this case our analytical formulae considerably simplify and agree with the ones of refs.~\\cite{bhp,gk}. In particular, they show the presence of resonance bands which are anyhow limited by Pauli blocking. \\looseness=1 In section~\\ref{sec:5} we study the more interesting case of production in an expanding Universe. As indicated in ref.~\\cite{gk} and as confirmed by the numerical results of the work~\\cite{gprt}, the creation is now very different with respect to the previous case. The expansion removes the resonance bands and the production (almost) saturates a Fermi sphere up to a maximal momentum. Our analytical results confirm this behavior. In section~\\ref{sec:5} we also calculate the total energy density $\\rho_X$ of produced fermions, which may be the quantity of most physical relevance. We compare our results with the ones of ref.~\\cite{gprt}, where it is shown that the final value of $\\rho_X$ (normalized to the inflaton energy density) scales linearly with the parameter $q \\equiv( g^2 \\phi_0^2) / ( 4 m_\\phi^2)$,\\footnote{In this expression $g$ is the Yukawa coupling between the inflaton and the fermions, $\\phi_0$ the initial value of the inflaton, and $m_\\phi$ its mass.} while it depends very weakly on the fermion bare mass $m_X$. However, these numerical results are valid only in a limited range for $m_X$, and it has been wondered if the density $\\rho_X$ decreases at values of $m_X$ below this range. Our analytical results give a positive answer in this regard. To see this, a proper average of the analytical formulae must be done, exploiting the fact that the expansion of the Universe gives the production a stochastic character. In this way one can get a ``mean'' function that interpolates very well between the maxima and the minima of the spectra of produced particles. Again we derived it in close analogy with what is done in the bosonic case, and again the results that we get are in very good agreement with the numerical ones of ref.~\\cite{gprt} in the region of validity of the latter. All this analysis neglects the backreaction of the produced fermions on the evolution of the inflaton field and of the scale factor. Despite the difficulty of a more complete treatment, backreaction effects can be understood at least in the Hartree approximation. This was done numerically in ref.~\\cite{gprt}. In section~\\ref{sec:6} we see that the analytical formulae here provided allow to understand the effects of backreaction observed in the numerical simulations. ", "conclusions": "\\label{sec:7} Preheating of fermions and bosons present several analogies. In both cases, especially when one is interested in the production of very massive particles, the creation occurs for very short intervals of time, during which the frequency of the particles varies non adiabatically and their occupation number cannot be defined. This strong similarity can be exploited to extend to the fermionic case the formalism developed in ref.~\\cite{kls} for the analytical study of preheating of bosons. This was first done in ref.~\\cite{ckrt}, where the result for the particles created in the first production is reported. \\pagebreak[3] When one wants to consider the successive productions, a more detailed analysis is necessary. This has been the object of the first part of the present work. The formulae that we obtained, valid for an arbitrary number of productions, give a very good agreement with the numerical results, as some examples provided manifest. Another strong similarity between the two phenomena is that in both cases the production would occur only through resonance bands (in momentum space) was not for the expansion of the Universe. For what concerns fermions, this feature was first studied in ref.~\\cite{bhp,gk}, where production in a static Universe is considered. As it is well known, the expansion of the Universe removes the resonance bands and, as a result, this allows the fermionic production to saturate the whole Fermi sphere up to a maximal momentum $k_{\\max}$. Our analytical results confirm both the presence of the resonance bands in the static case (where our formulae considerably simplify) and their disappearance when the expansion is considered. Rather than the detailed form of the spectrum, a quantity which may be of more physical interest is the total energy of the produced particles. The analogy with the bosonic case turns useful also in this regard. In the final spectra one notices the presence of some maxima and minima, whose exact position is determined by the detailed knowledge of the phases of the Bogolyubov coefficients (adopted in the analytical derivation) after each production. However, due to the expansion of the Universe, these phases are effectively uncorrelated among themselves. This gives the production a stochastic character which is responsible for the disappearance of the resonance bands. Averaging over these phases, it is possible to get an analytical function for the mean occupation number. For what concerns preheating of bosons, the results are simplified by the fact that, in the region of very efficient production, the mean occupation number after a given production is proportional to the mean occupation number of the previous one. This assumption is not possible in the fermionic case, since the Pauli principle prevents the occupation number to exceed $1$. However, it is remarkable that also for fermions the final result can be cast in a very simple and readable form. The analytical formula for the mean occupation number confirms the saturation of the Fermi sphere (times the factor $1/2$ which comes from the average) up to a momentum $k_{\\max}$. In order to get this quantity, it is sufficient to know the derivative of the inflaton field and the value of the scale factor at the points where the productions occur. These values can be calculated numerically evolving the equations for the inflaton field alone (at least when backreaction is neglected), or analytically. The former possibility is much more rapid than a full numerical computation, since one does not have to consider the equations for the fermions. Moreover, its results agree very well with the full numerical ones~\\cite{gprt}. Thanks to the increase in the rapidity, this evaluation (that is called ``semi-analytical'' in the present work) allows to get results with a more extended range of validity with respect to the previous full numerical~ones. The results provided in this way can also be achieved with a full analytical study. From both these methods, we deduce that the ratio $\\rho_X / \\rho_\\phi$ scales as \\begin{equation} \\label{final} \\frac{\\rho_X}{\\rho_\\phi} \\sim q m_X^{1/2} \\left[ \\log \\frac{q^{1/2}}{m_X} \\right]^{3/2}, \\end{equation} up to $\\big( m_X \\big)_{\\mathrm{th}}\\sim \\sqrt{q}/2$. For $m_X$ not too smaller than $\\big( m_X \\big)_{\\mathrm{th}}$ this scaling is in good agreement with the one $\\rho_X / \\rho_\\phi \\sim q$ of the numerical work~\\cite{gprt}. However, we see that the density of produced fermions decreases at smaller values of $m_X$. We would like to conclude our work showing that the scaling~(\\ref{final}) can be also achieved from very immediate considerations. As we said, preheating of fermions saturates a Fermi sphere in momentum space up to a maximal momentum $k_{\\max}$. From the analytical formula~(\\ref{matrix1}), we notice that at high momenta $k$ the occupation number is well approximated by \\begin{equation} N_n ( k ) \\simeq \\sum_{i=1}^n e^{- k^2/z_i^2} \\,, \\end{equation} where we remember $z_i \\propto q^{1/4} a^{1/2} \\big(\\eta_{*i}\\big) | \\phi'\\big(\\eta_{*i}\\big) |^{1/2}$. \\looseness=-1 In this last equation, we replace all the parameters $z_i$ with a mean value ${\\bar z}$, so that $N_n \\sim n\\, \\mathrm{exp} \\big( - k^2 / {\\bar z}^2 \\big)$. The scaling of ${\\bar z}$ with the physical parameters $q$ and $m_X$ follows from the scaling of all the $z_i$. The maximal momentum $k_{\\max}$ is thus expected to scale as the quantity $z_i ( \\log n)^{1/2}$. Considering now the evolution of the inflaton field in physical time $t$, we notice that both the number $n$ of productions and the times $t_{*i}$ at which they occur are proportional to the parameter $R = q^{1/2} / \\big( 2 m_X)$. Moreover, we see that the $z_i$'s scale as \\begin{equation} z_i \\propto q^{1/4} a_{*i} \\left[ \\frac{d \\phi}{ d t} |_{*i} \\right]^{1/2} \\propto q^{1/4} t_{*i}^{2/3} \\frac{1}{t_{*i}^{1/2}} \\propto q^{1/4} R^{1/6} \\,. \\end{equation} We thus get $k_{\\max} \\propto q^{1/4} R^{1/6} [ \\log R]^{1/2}$, from which (remember $\\rho_X \\simeq m_X k_{\\max}^3$) the scaling~(\\ref{final}) simply follows." }, "0003/astro-ph0003436_arXiv.txt": { "abstract": "Planetary Nebulae (PNe) in the Large Magellanic Cloud (LMC) offer the unique opportunity to study both the Population and evolution of low- and intermediate-mass stars, by means of the morphological type of the nebula. Using observations from our LMC PN morphological survey, and including images available in the {\\it HST} Data Archive, and published chemical abundances, we find that asymmetry in PNe is strongly correlated with a younger stellar Population, as indicated by the abundance of elements that are unaltered by stellar evolution (Ne, Ar, S). While similar results have been obtained for Galactic PNe, this is the first demonstration of the relationship for extra-galactic PNe. We also examine the relation between morphology and abundance of the products of stellar evolution. We found that asymmetric PNe have higher nitrogen and lower carbon abundances than symmetric PNe. Our two main results are broadly consistent with the predictions of stellar evolution if the progenitors of asymmetric PNe have on average larger masses than the progenitors of symmetric PNe. The results bear on the question of formation mechanisms for asymmetric PNe, specifically, that the genesis of PNe structure should relate strongly to the Population type, and by inference the mass, of the progenitor star, and less strongly on whether the central star is a member of a close binary system. ", "introduction": " ", "conclusions": "" }, "0003/quant-ph0003014_arXiv.txt": { "abstract": "We propose a nanostructure switch based on nuclear magnetic resonance (NMR) which offers reliable quantum gate operation, an essential ingredient for building a quantum computer. The nuclear resonance is controlled by the magic number transitions of a few-electron quantum dot in an external magnetic field. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003371_arXiv.txt": { "abstract": "By means of N-body/hydrodynamical simulations we model the evolution of a primordial $10^{8} M_{\\odot}$ density peak which ends up in an object closely resembling the present day globular cluster $\\omega$~Centauri.\\\\ We succeed to reproduce the main features of the cluster, namely the structure, kinematics and metallicity distribution.\\\\ We suggest that $\\omega$~Centauri might be a cosmological dwarf elliptical, formed at high redshift, evolved in isolation and self-enriched, and eventually fallen inside the potential well of the Milky Way, in agreement with the Searle-Zinn (1978) paradigm for galactic globular cluster formation.\\\\ We finally suggest that $\\omega$~Centauri is probably surrounded by an extended Dark Matter (DM) halo, for which no observational evidence is at present available. We expect that signatures, if any, of the DM halo can be found in the kinematics of stars outside about 20 arcmin. ", "introduction": "$\\omega$~Centauri is one of the most interesting globular cluster in the Milky Way, and maybe the most studied one (Majewski et al. 1999, Lee et al. 1999, Pancino 1998). A compilation of its main properties is given in Table~1.\\\\ The most striking feature of this cluster is the measured metallicity spread, which has been interpreted as the evidence of a multiple stellar population inside it (Norris et. al 1996, Suntzeff \\& Kraft 1996). The precise metallicity distribution function (MDF) is nonetheless still disputed. Norris et al (1996) suggest the presence of a secondary peak at $[Ca/H] \\approx -0.9$ roughly 5 times smaller that the main peak at $[Ca/H]\\approx -1.4$. This trend is not confirmed by Suntzeff \\& Kraft (1996), who claim for a more regular MDF.\\\\ On the base of a large photometric survey Lee et al. (1999) (but see also Ortolani et al 1999) analyzed the color distribution of a sample of bright stars, showing that on the average it has an e-folding trend, with the presence of several significant metallicity peaks. However Majewski et al. (1999) arrive at a somewhat different conclusion, showing that the MDF has a gaussian shape with the maximum at $[Fe/H] \\approx -1.7$, and with some evidences of a secondary peak. Although different, all these analyses point to the common picture of an object which experienced an irregular self-enrichment over its evolution.\\\\ Putting together the chemical and kinematical properties Majewski et al. (1999) claim that $\\omega$~Centauri might be a possible dwarf galaxy relict.\\\\ In this paper we propose a N-body/gasdynamical model for the formation and evolution of $\\omega$~Centauri, suggesting that this globular cluster can actually be the remnant of a dwarf elliptical galaxy, formed and evolved avoiding strong mergers, and eventually captured by the Milky Way.\\\\ To this aim, the plan of this paper is as follows. In Sect.~2 we describe our model and the initial conditions setup; Sect.~3 to 5 are dedicated to the analysis of the structure, chemistry and internal kinematics, respectively, of our model, and the comparison with $\\omega$~Centauri; in Sect.~6 we investigate about the possible presence of an extended DM halo around the cluster. Finally Sect.~7 summarizes our results. ", "conclusions": "We have presented a N-body/gasdynamical simulation of the formation and evolution of the globular cluster $\\omega$~Centauri. We are able to reproduce the bulk properties of the cluster, namely structure, kinematics and chemistry assuming that it formed and evolved in isolation, and eventually fell inside the Milky Way potential well.\\\\ According to our results and to the dwarf galaxies taxonomy proposed by Roukema (1999), $\\omega$~Centauri can actually be a cosmological {\\it dwarf by mass}, formed in a high redshift low mass halo, which escaped significant merging up to the present time. Finally we stress that in order to obtain these results, an extended DM halo should surround the present day $\\omega$~Centauri. \\subsection*{\\bf Acknowledgments } G.C. expresses his gratitude to Prof. S. Ortolani for fruitful discussions. The referee, George Meylan, is acknowledged for important suggestions which led to the improvement of the paper. This work has been financed by italian MURST and ASI." }, "0003/astro-ph0003237_arXiv.txt": { "abstract": "We have carried out a precise energy spectral analysis of the super-luminal jet source GRS1915+105 observed with ASCA six times from 1994 to 1999. The source was so bright that most SIS data suffered from event pileup. We have developed a new technique to circumvent the pileup effect, which enabled us to study the spectrum in detail and at high resolution ($\\Delta E/E \\approx$ 2 \\%). In the energy spectra of 1994 and 1995, resonant absorption lines of Ca {\\sc xx} K$\\alpha$, Fe {\\sc xxv} K$\\alpha$, Fe {\\sc xxvi} K$\\alpha$, as well as blends of the absorption lines of Ni {\\sc xxvii} K$\\alpha$ + Fe {\\sc xxv} K$\\beta$ and Ni {\\sc xxviii} K$\\alpha$ + Fe {\\sc xxvi} K$\\beta$, were observed. Such absorption lines have not been found in other objects, except for iron absorptions lines from GRO J1655$-$40, another super-luminal jet source (Ueda et al.\\ 1998). We carried out a ``curve of growth'' analysis for the absorption lines, and estimated column densities of the absorbing ions. We found that a plasma of moderate temperature (0.1--10 keV) and cosmic abundance cannot account for the observed large equivalent widths. The hydrogen column density of such plasma would be so high that the optical depth of Thomson scattering would be too thick ($N_H\\gtrsim10^{24}$cm$^{-2}$). We require either a very high kinetic temperature of the ions ($\\gtrsim 100$ keV) or extreme over-abundances ($\\gtrsim 100$ $Z_\\odot$). In the former case, the ion column densities have reasonable values of $10^{17} - 10^{18}$ cm$^{-2}$. We modeled the absorber as a photo-ionized disk which envelops the central X-ray source. Using a photo-ionization calculation code, we constrain physical parameters of the plasma disk, such as the ionization parameter, radius, and density. Estimated parameters were found to be consistent with those of a radiation-driven disk wind. These absorption-line features may be peculiar to super-luminal jet sources and related to the jet formation mechanism. Alternatively, they may be common characteristics of super-critical edge-on systems. ", "introduction": "GRS 1915+105 was discovered as a transient X-ray source with WATCH/Granat (Castro-Tirado, Brandt, \\& Lund 1992), and later recognized as a super-luminal Galactic jet source at a distance of 12.5 kpc (Mirabel \\& Rodr\\'{\\i}guez 1994). The velocity of the jet and its inclination were estimated by Mirabel \\& Rodr\\'{\\i}guez (1996) to be 0.92 $c$ and 70$^\\circ$, respectively . The optical type of the companion star and binary parameters are as yet unknown. The formation of the relativistic jet is considered to be related to instabilities in a super-critical accretion disk (Belloni et al.\\ 1997; Mirabel et al.\\ 1998), but the precise formation mechanism is not yet clear. GRS 1915+105 is extremely important because it is a potential key to understanding the production of jets, which is relevant to astrophysical systems ranging from binaries to AGNs. To obtain information on the environment of the central engine of jet sources, X-ray spectroscopy with a fine resolution is desirable. GRS 1915+105 was observed with the X-ray astronomy satellite ASCA, and surprisingly, absorption features of calcium and iron K structure were found in the spectrum (Ebisawa 1997a; Ebisawa 1997b; Kotani et al.\\ 1997a; Ebisawa et al.\\ 1998). These features were interpreted as absorption lines due to K$\\alpha$-resonant scattering of helium-like and hydrogen-like ions (Kotani et al.\\ 1999a; Kotani et al.\\ 1999b). Such X-ray absorption lines have never been found in other X-ray binaries except for GRO J1655$-$40, another super-luminal jet source (Ueda et al.\\ 1997; Ueda et al.\\ 1998). Ueda et al.\\ (1997; 1998) identified helium-like and hydrogen-like iron-K$\\alpha$ absorption lines, and estimated the physical condition of the plasma scattering iron-K$\\alpha$ photons. Since two of the super-luminal jet sources show absorption lines, it is natural to consider the lines in the context of jet production. The absorption lines may be a byproduct of the unknown jet formation mechanism. In this paper, we discuss the identification and origin of the X-ray absorption lines in GRS 1915+105 in detail. ", "conclusions": "\\subsection{Interpretation of the Absorption Lines} We notice the center energies of the negative Gaussians shown in Table~\\ref{kotani:tbl:bestfitsis} are consistent with or very close to the K-line energies of iron or nickel ions. Candidate resonant and inter-combination lines are listed in Table~\\ref{kotani:tbl:candidates} (forbidden lines are not included). The energies of the negative Gaussians A and B coincide with those of Fe {\\sc xxv} K$\\alpha$ ($1s^12p^1$ $^1{\\rm P}_1$ and $1s^12p^1$ $\\rm^3P_{1,2}$ in Table~\\ref{kotani:tbl:candidates}) and Fe {\\sc xxvi} K$\\alpha$ ($2p^1$ $\\rm^2P_{1/2,3/2}$), respectively. The negative Gaussian C is considered to be Fe {\\sc xxv} K$\\beta$ ($1s^12p^1$ $\\rm^1P_1$) or Ni {\\sc xxvii} K$\\alpha$ ($1s^12p^1$ $\\rm^1P_1$ and $1s^12p^1$ $\\rm^3P_{1,2}$), although nickel is less abundant than iron in a plasma of cosmic abundance. The negative Gaussian D resides just between the Ni {\\sc xxviii} K$\\alpha$ ($2p^1$ $\\rm^2P_{1/2,3/2}$) and Fe {\\sc xxvi} K$\\beta$ ($3p^1$ $\\rm^2P_{1/2}$) energies. These coincidences strongly support the interpretation that the absorption features consist of iron or nickel resonant absorption lines. Hereafter, we discuss the physics of the system assuming that the negative Gaussians A, B, C, and D are absorption lines of Fe {\\sc xxv} K$\\alpha$, Fe {\\sc xxvi} K$\\alpha$, Fe {\\sc xxv} K$\\beta$ + Ni {\\sc xxvii} K$\\alpha$, and Fe {\\sc xxvi} K$\\beta$ + Ni {\\sc xxviii} K$\\alpha$, respectively, \\placetable{kotani:tbl:candidates} \\subsection{Curve of Growth Analysis} The observed spectral features are obviously different from the P-Cygni profile which accompanies strong emission lines, and thus the absorption is probably not caused by a spherical wind. As for the origin of the absorption lines, we will consider line-scattering material anisotropically irradiated or anisotropically distributed around the central source. To estimate amount of the line-scattering matter in the line of sight, we calculate the ``curve of growth'', namely, the equivalent widths of an absorption line as function of the column density of scattering ions. We summarize here the basic equations we used to calculate the curves of growth. Let us assume particles with the column density $N$ producing an absorption line at frequency $\\nu_0$ by resonant scattering. The optical depth $\\tau(\\nu)$ is expressed as \\begin{equation} \\tau(\\nu) = N s \\phi_a(\\nu), \\end{equation} and the normalized equivalent width $W_\\nu/\\nu_0$ is \\begin{equation}\\label{ew} \\frac{W_\\nu}{\\nu_0} = \\frac 1 {\\nu_0} \\int (1-\\exp[-\\tau(\\nu)])d\\nu, \\end{equation} respectively, where $s \\phi_a(\\nu)$ is the averaged cross section of scattering particles, normalized as $\\int \\phi_a(\\nu) d\\nu =1$ (see Spitzer 1978, section 3.4). In CGS unit, the transition probability $s$ is expressed as \\begin{equation} s = \\frac{\\pi f_{lu}e^2}{m_{\\rm e}c}(1-\\frac{b_u}{b_l}\\exp[-h \\nu_0/kT]), \\end{equation}\\label{kotani:eq:transition} where $f_{lu}$ is the oscillator strength, and the subscripts $u$ and $l$ denote values in the upper and lower states, respectively. The factor $(b_u/b_l)\\exp[-h \\nu_0/kT]$, expressing the effect of induced emission, is defined to be equal to the actual rate of the density of ions in the upper state to that in the lower (Spitzer 1978, section 2.4). We assumed that the density of the ions in the upper state is negligible and omit the factor from the following calculation. The validity of the assumption is checked in \\S~\\ref{kotani:subsec:validity}. If scattering particles obey the Maxwell-Boltzmann distribution with temperature $T$, $\\phi_a(\\nu)$ is written as a Voigt function $H(a, u)$ with \\begin{eqnarray} \\phi_a(\\nu) &= &\\frac c {\\sqrt\\pi\\nu_0b} H(a,u)\\\\ H(a,u) &= &\\frac a \\pi \\int \\frac{\\exp[-y^2]dy}{a^2+(u-y)^2}\\\\ a &= &\\frac {c}{4\\pi\\nu_0b}\\sum_{l}A_{ul}\\\\ u &= &\\frac c {\\nu_0b}(\\nu-\\nu_0)\\\\ b &= &\\sqrt\\frac{2kT}{m\\rm_i}, \\end{eqnarray} where $A_{ul}$ are the Einstein coefficients related to the oscillator strengths as \\begin{equation} A_{ul} = \\frac{8\\pi^2\\nu_0^2e^2}{m_{\\rm e}c^3} \\frac {g_l}{g_u}f_{lu} \\end{equation} (see Rybicki \\& Lightman 1979, section 10.6). Even if the energy distribution of particles is dominated by turbulence or other bulk motion rather than thermal motion, these equations are still applicable as a good approximation. In that case, $T$ represents the velocity dispersion within the scattering gas. With these formulas and constants in Table~\\ref{kotani:tbl:candidates}, the equivalent widths are calculated according to equation (\\ref{ew}), and the curves of growth are plotted in Fig.~\\ref{kotani:fig:curve}. Using the curves of growth thus calculated, the column densities of the ions $N_{\\rm ion}$ responsible for the observed absorption lines may be obtained. Since Fe {\\sc xxv} K$\\alpha$ and Fe {\\sc xxvi} K$\\alpha$ lines are considered not contaminated by other ion species, $N_{\\rm Fe\\, XXV}$ and $N_{\\rm Fe \\, XXVI}$ are directly derived from the observed equivalent widths, assuming a temperature. From these column densities, equivalent widths of Fe {\\sc xxv} K$\\beta$ and Fe {\\sc xxvi} K$\\beta$ are estimated using the curves of growth. Then, these estimated equivalent widths are subtracted from the observed equivalent widths of the blends of Ni {\\sc xxvii} K$\\alpha$ + Fe {\\sc xxv} K$\\beta$, or of Ni {\\sc xxviii} K$\\alpha$ + Fe {\\sc xxvi} K$\\beta$, to obtain the equivalent widths of Ni {\\sc xxvii} K$\\alpha$ or Ni {\\sc xxviii} K$\\alpha$. Column densities of $N_{\\rm Ni\\, XXVII}$ and $N_{\\rm Ni\\, XXVIII}$ are obtained using the curves of growth again. Ion column densities thus derived are shown in Table~\\ref{kotani:tbl:column} for two extreme temperatures, 1000 keV and 0.1 keV. If the temperature is as high as 1000 keV (high-temperature limit), all the observed values of the equivalent width will be found in the ``linear part'' of the curves of growth, while if temperature is as low as 0.1 keV (low-temperature limit), they will be in the ``square-root section'' of the curves. It also should be noted that the column densities shown in the Table~\\ref{kotani:tbl:column} are lower limits for both temperatures, since line photons scattered from matter located out of the line of sight would reduce the absorption line equivalent widths. That effect is not taken into account here. The derived column densities for the low-temperature limit are larger than those for the high-temperature limit by several orders of magnitude. If the column densities are so large as the low-temperature calculation suggests, and if abundances of the absorbing plasma is not very different from the solar value, the optical depth of the plasma for the Thomson scattering will exceed unity. This is unlikely, since the modification of the absorption line features that would be present due to Thomson scattering is not. It is plausible that either iron and nickel are significantly over-abundant, or that the kinematic temperature is as high as the high-temperature limit suggests. However, at such high temperature, iron and nickel would be fully ionized and absorption lines should not be observed. It may be that the ionization temperature of the absorbing plasma is lower than the kinematic temperature, or that the absorbing matter consists of several parts with different bulk velocities whose dispersion is comparable to thermal velocity of iron atoms at $kT>$ 100 keV\\@. In following discussion, we first consider the high-temperature limit in Table~\\ref{kotani:tbl:column}, which presumably gives the lower limit of column density. \\placefigure{kotani:fig:curve} \\placetable{kotani:tbl:column} \\subsection{Photo-ionized Plasma Model} We note that the spectral parameters changed between the observations of 1994 and 1995. The luminosity in the 1995 observation was higher than in 1994, and the ratio $N_{\\rm Fe\\, XXVI}/N_{\\rm Fe \\, XXV}$ became larger, suggesting that the absorbing matter was more ionized. This behavior is consistent with characteristics of a photo-ionized plasma. Thus, we assume that the source of absorption lines in 1994 and 1995 is a photo-ionized plasma irradiated by the central source. In 1996, the system was considered to be in a different state, because, not only were absorption lines absent, but also the photon index and folding energy of the spectral continuum were different. We calculated electron temperature and ion-population distribution of the photo-ionized plasma with XSTAR (Kallman \\& Krolik 1996), and searched for parameters which are consistent with those in the high-temperature limit in Table~\\ref{kotani:tbl:column}. To simplify the calculation, the plasma geometry was assumed to be isotropic around the source. It should be noted that assumption of isotropy is adopted only in the calculation, and that the realistic geometry of the line-absorbing plasma must be anisotropic, since otherwise absorption lines would be canceled by emission lines from out of the line of sight. In the calculations, abundance was assumed to be solar. The absorption-corrected 2--10 keV luminosity was estimated from Table~\\ref{kotani:tbl:bestfitgis} assuming a distance of 12.5 kpc, and used as total luminosity of the model. Continuum spectral parameters were drawn from Table~\\ref{kotani:tbl:bestfitgis}. We arbitrarily fixed the ionization parameter $\\xi = L/(nr^2)$ at the innermost boundary of the plasma to be $10^{5.5}$. Our results depend only weakly on this value, because iron and nickel is fully ionized with such a high ionization parameter at the innermost boundary, and thus column density of hydrogen-like or helium-like ions is not sensitive to the innermost condition. The assumed geometry is schematically shown in Fig.~\\ref{kotani:fig:geometry}. We searched a combination of plasma density and outer radius to give $N_{\\rm Fe \\, XXVI}$ and $N_{\\rm Fe \\, XXV}$ consistent to observation. The determined parameters are shown in Table~\\ref{kotani:tbl:photo}, and expected column density of ions are shown in Table~\\ref{kotani:tbl:calccolumn}. Obviously, the expected column density of nickel ions in Table~\\ref{kotani:tbl:calccolumn} is much lower than the observation by one or two orders of magnitude. Therefore, we conclude that the relative abundance of nickel to iron is larger than that of solar value. We next searched for a combination of plasma parameters to reproduce the column density in the low-temperature limit in Table~\\ref{kotani:tbl:column}. Since the metal abundance was found to be non-solar, there was no reason to fix them to the solar value. We found that if both the iron and nickel abundances are multiplied by $\\sim 100$, the resultant photo-ionized plasma gives the column densities in the low-temperature limit. Other parameters, such as the electron density, the ionization parameter at the outer boundary, and the outer radius of the plasma were hardly affected. Although the column densities are different between the high-temperature-limit and low-temperature-limit cases, the column-density ratios of Fe {\\sc xxv} to Fe {\\sc xxvi} are not much different between the two cases, and thus both yield similar values of $\\xi_1$. (Subscript``0'' and ``1'' denote values at the innermost and outermost boundary, respectively.) Thus we consider that $\\xi_1$ and $r_1$ in Table~\\ref{kotani:tbl:photo} are reliable, regardless of the kinematic temperature, the metal abundances, the electron density, or the innermost radius of the plasma. \\placefigure{kotani:fig:geometry} \\placetable{kotani:tbl:photo} \\placetable{kotani:tbl:calccolumn} \\subsection{Geometry} To produce absorption lines, the solid angle of the line-scattering material from the source must be small, otherwise the absorption lines would be offset by an emission line counterpart. From a moderate assumption that the solid angle is less than $4\\pi\\exp[-1]$, the disk half thickness, $h$, at the radius $r_1$ is constrained as $h/r_1 \\lesssim 0.4$, where $h/r_1$ is related to the solid angle as $\\Omega = 1/\\sqrt{1+(h/r_1)^2}$. Therefore, the half opening angle of the plasma, $\\theta$, is constrained to be $\\theta < 20^\\circ$. The inclination of the jets is $70^\\circ$ to our line of sight (Mirabel \\& Rodor\\'{\\i}guez 1993). We might reasonably assume that the immediate neighborhood of the central source (e.g., the accretion disk) is also so inclined. From these two constraints, we conclude that the half thickness of the reprocessing plasma is $\\sim 0.4 r_1$. Thus, total mass of the reprocessing plasma is estimated as $\\sim 4.4 \\times 10^{-11}$ M$_\\odot$ (1994) and $\\sim 1.5 \\times 10^{-10}$ M$_\\odot$ (1995). Note that $n\\rm_p$ in Table~\\ref{kotani:tbl:photo} is dependent on assumed geometry and has a large uncertainty, while $\\xi_1$ and $r_1$ are rather model independent and thus reliable. For example, a toroidal plasma distribution around the source can also explain the observed properties, and in this case the path in the plasma along the line of sight will be shorter. Therefore the proton density $n\\rm_p$ will be larger, conserving column density $(r_1-r_0) \\times n_{\\rm p}$. The other parameters $\\xi_1$ and $r_1$ would not be changed much, because the former is determined from the ratio $N_{\\rm Fe\\, XXVI}/N_{\\rm Fe\\, XXV}$, and the latter is determined from the relation $\\xi_1 = L/(N_{\\rm e}r_1^2)$. \\subsection{Nature of the Reprocessing Plasma} From the discussion above, we can construct a picture of the plasma producing the absorption lines. It is located at $\\sim 10^{11}$ cm from the source, constrained to be within $\\theta<20^\\circ$ of the disk plane, and has a total mass of $10^{-11}$ to $10^{-10}$ M$_\\odot$. The kinetic temperature of iron ion is either high or low. If the metal abundance is comparable to the solar value, then the kinematic temperature must be as high as or higher than 100 keV, which corresponds to a velocity of $\\sim10^8$ cm s$^{-1}$. If the reprocessing plasma is hotter than 100 keV, it must be observed within 10 s of its production to avoid full ionization of iron and nickel. A post-shock flow within 10 s after the shock is a candidate for such a hot plasma. Both the thermal motion of each ion and the bulk motion of the plasma may give such a high kinematic temperature. An inward or outward bulk flow with a velocity gradient or randomly moving blobs may be the X-ray reprocessor. As a candidate of such a flow, we suggest a radiation-driven disk wind. If gas at rest exists in the vicinity of the super-Eddington X-ray source, it would be blown out by Thomson scattering. The terminal velocity $v_{\\rm wind}$ would be \\begin{equation} v_{\\rm wind} \\sim \\sqrt{ \\frac {L\\sigma_{\\rm Th}}{4\\pi rcm_{\\rm p}}}, \\end{equation} where $L$ is the luminosity of the source, $\\sigma_{\\rm Th}$ is the cross section of Thomson scattering, $r$ is the initial distance of the gas from the source, and $m\\rm_p$ is the mass of a proton. If such a gas is supplied from the accretion disk, we would observe a stream with a radial-velocity gradient from 0 to $v_{\\rm wind}$. Substituting $r = 10^{11}$ cm and $L = 4\\times 10^{38}$ erg s$^{-1}$, a terminal velocity of $v_{\\rm wind} = 10^8$ cm s$^{-1}$ is obtained, which coincides with the velocity inferred in the curve-of-growth discussion. It should be noted that most of the centroid energies of the absorption lines in Table~3 are blue-shifted from those in Table~5, which would be expected from the absorption lines from an outward flow with a velocity of $10^8$ cm s$^{-1}$. Thus a radiation-driven disk wind at $ \\sim 10^{11}$ cm can well explain the observed absorption line features. Because the disk wind would be refreshed in every $\\sim 10^3$ s, the outflow rate is estimated as $10^{-7} - 10^{-6}$ M$_\\odot$ yr$^{-1}$. On the other hand, if the plasma is very metal rich, the assumption of high-temperature is no longer necessary. The high column density of ions may be explained by low-temperature, metal rich plasma, if the metal abundance is $100$ times higher than the solar value. Even in that case, the conclusion on the geometry of the plasma would not be changed appreciably. The nickel abundance of the jet material of SS~433 has been found to be a few tens times higher than the solar value (Kotani et al.\\ 1997b), which suggests that anomalous abundances may not be rare in binary systems with strong jets. Thus they may be possible that iron and nickel are abundant in the environment of GRS~1915+105, and that the reprocessing plasma has a velocity dispersion smaller than $10^8$ cm s$^{-1}$ or a lower temperature than 100 keV\\@. \\subsection{Alternative situations}\\label{kotani:subsec:validity} In the analysis above, the term of the induced emission $(b_u/b_l)\\exp[-h \\nu_0/kT]$ in equation~(\\ref{kotani:eq:transition}) was neglected. The validity of the assumption can be confirmed as follows. Suppose that a reprocessing plasma is illuminated by an external photon source. An ion in the plasma receives photons with a mean interval of $1/(\\sigma_{\\rm res}F(\\nu_0))$, and emits each photon after a time $\\sim 1/A_{ul}$, where $F(\\nu_0)$ is the photon flux of the source and $\\sigma_{\\rm res}$ is the resonant-scattering cross section. Therefore the ratio of the number density of ions at upper state to those at lower state is $\\sigma_{\\rm res}F(\\nu_0)/A_{ul}$. This is an overestimation because photons at $\\nu_0$ are reflected by the plasma surface due to resonant scattering. If the ratio is much less than unity, the term of induced emission is negligible. Substituting $\\sigma_{\\rm res} = c^2/(8\\pi\\nu_0^2)$, the ratio is estimated as \\begin{equation} \\frac {b_u}{b_l}\\exp[-h\\nu_0/kT] \\sim \\frac {c^2F(\\nu_0)}{8\\pi\\nu_0^2A_{ul}} \\sim 10^{-10} \\left ( \\frac L {10^{39}\\rm \\; erg\\; s^{-1}} \\right ) \\left ( \\frac R {10^{11}\\rm \\; cm} \\right )^{-2},\\label{kotani:eq:ratio} \\end{equation} which is much less than unity. Thus induced emission is shown to be negligible in the case of the photo-ionized plasma treated here. There is another case where the induced-emission term is important. The ratio~\\label{kotani:eq:ratio} approaches unity on the surface of a neutron star ($R = 10^6$ cm). Under the presence of the induced emission, more ions are necessary to produce an absorption line. As an estimation of column density of Fe {\\sc xxvi}, we adopt $10^{20.4}$ cm$^{-2}$ from Table~6. The corresponding hydrogen column density would be $10^{25}$ cm$^{-2}$, assuming solar abundance. To avoid full ionization of iron, the ionization parameter must be $\\xi \\leq 10^{4}$, and thus the density of the plasma must be $\\geq 10^{22}$ cm$^{-3}$, and the height of the atmosphere must be $\\stackrel{<}{\\sim}10^{3}$ cm. Although the scale height of the atmosphere at $kT \\sim 100$ is larger than $10^{3}$ cm by an order of magnitude, such an atmosphere may produce absorption lines. However, the absorption lines made by the atmosphere would be gravitationally red-shifted, which was not observed. The atmosphere of a neutron star is unlikely to be the origin of the absorption lines. Thus we conclude that induced emission can be neglected in the usual situations where equation~(\\ref{kotani:eq:ratio}) is applicable." }, "0003/astro-ph0003001_arXiv.txt": { "abstract": "We explore the Tully-Fisher relation over five decades in stellar mass in galaxies with circular velocities ranging over $30 \\lesssim V_c \\lesssim 300\\kms$. We find a clear break in the optical Tully-Fisher relation: field galaxies with $V_c \\lesssim 90\\kms$ fall below the relation defined by brighter galaxies. These faint galaxies are however very gas rich; adding in the gas mass and plotting baryonic disk mass $\\mass_d = \\mass_* + \\mass_{gas}$ in place of luminosity restores a single linear relation. The Tully-Fisher relation thus appears fundamentally to be a relation between rotation velocity and total baryonic mass of the form $\\mass_d \\propto V_c^4$. ", "introduction": "The relation between luminosity and rotation velocity for galaxies is well known (Tully \\& Fisher 1977\\markcite{TF}). It has been used extensively in estimating extragalactic distances (e.g., Sakai et al.\\ 2000\\markcite{HSTkey}, Tully \\& Pierce 2000\\markcite{TP}), and it provides a critical constraint on galaxy formation theory (Dalcanton, Spergel, \\& Summers 1997\\markcite{D97}; McGaugh \\& de Blok 1998\\markcite{MdB98a}; Mo, Mao, \\& White 1998\\markcite{MMW}; Steinmetz \\& Navarro 1999\\markcite{SN}; van den Bosch 1999\\markcite{vdB}). However, the physical basis of the Tully-Fisher relation remains unclear. The requirements of the empirical Tully-Fisher relation are simple, but the steep slope and small scatter are difficult to understand. Luminosity must trace total (dark plus luminous) mass, which in turn scales exactly with circular velocity. Considerable fine-tuning is required to obtain these strict proportionalities (McGaugh \\& de Blok 1998\\markcite{MdB98a}). The intrinsic properties of dark halos are not expected to be as tightly correlated as observed (Eisenstein \\& Loeb 1995\\markcite{EL}). The mapping from the properties of dark matter halos to observable quantities should introduce more scatter, not less. Somehow the baryons ``know'' precisely how many stars to form. Let us suppose that, for whatever fundamental reason, there does exist a universal relationship between total mass and rotation velocity of the form $\\mass_{tot} \\propto V_c^b$. The empirical Tully-Fisher relation then follows if luminosity traces mass: \\begin{equation} L = \\Upsilon_*^{-1} f_* f_d f_b \\mass_{tot}, \\end {equation} where $f_b$ is the baryon fraction of the universe, $f_d$ is the fraction of the baryons associated with a particular galaxy halo which reside in the disk, $f_*$ is the fraction of disk baryons in the form of stars, and $\\Upsilon_*$ is the mass-to-light ratio of the stars. Each of the pieces which intervene between $L$ and $\\mass_{tot}$ must be a nearly universal constant shared by all disks in order to maintain the strict proportionality the Tully-Fisher relation requires. Cast in this form, the traditional luminosity-linewidth relation is a sub-set of a more fundamental relation between {\\it baryonic mass\\/} and {\\it rotational velocity}. In this context, one would expect to find galaxies which deviate from the luminosity-linewidth relation because much of their baryonic mass is not in the form of stars. For example, a gas rich galaxy should appear underluminous for its circular velocity, but would, after correction for the gas content, fall on the underlying ``Baryonic Tully-Fisher relation'' (cf.\\ Freeman 1999\\markcite{Ken}). In this paper, we specifically test this premise by constructing the luminosity-linewidth and Baryonic Tully-Fisher relations for a sample of late type galaxies that span a much larger range of luminosities than any previously available sample. Section 2 describes the data we employ. Section 3 discusses the results and \\S 4 explores some of their implications. A summary is given in \\S 5. All distance dependent quantities assume $H_0 = 75\\kms\\;{\\rm Mpc}^{- 1}$. ", "conclusions": "We have explored the Tully-Fisher relation over five decades in luminous mass. This is a considerable increase in dynamic range over previous studies. We find clear evidence for a break in the optical Tully-Fisher relation around $V_c \\approx 90\\kms$. Galaxies with rotation velocities less than this are underluminous relative to the extrapolation of the fit to more rapidly spinning galaxies. However, these faint galaxies are very gas rich. Considering both stellar and gas mass restores a linear relation over the entire observed range. These observations strongly suggest that the Tully-Fisher relation is fundamentally a relation between rotation velocity and total baryonic disk mass. This relation has the form \\begin{displaymath} \\mass_d = {\\cal A} V_c^4 \\end{displaymath} with ${\\cal A} \\approx 35\\, (\\Upsilon_*^{K'}/0.8)\\, h_{75}^{-2}\\;\\mass_{\\sun}\\;{\\rm km}^{-4}\\;{\\rm s}^4$. The well known optical Tully-Fisher relation is an approximation to this more fundamental relation in the limit of galaxies dominated by stars. The existence of the Baryonic Tully-Fisher relation has a number of important implications. That it works means that stars in spiral galaxies have mass-to-light ratios which are reasonable for composite stellar populations. The modest amount of scatter indicates that the IMF must be nearly universal in order to yield such uniform mass-to-light ratios. Only corrections for gas content are necessary to obtain the Baryonic Tully-Fisher relation. The data do not allow much room for any further significant baryonic mass components. Any component of dark baryons which does exist must do so in strict proportionality to the observed baryons, with effectively zero scatter. This unlikely situation argues against a significant mass in dark baryons in any form (be it very cold molecular gas in the disk, very hot ionized gas in the halo, or baryonic MACHOs). Any model which supposes a large mass of such baryons must explain why it contributes so little to the scatter in the Baryonic Tully-Fisher relation. The results presented here make sense in terms of a simple interpretation of the Tully-Fisher relation in which the mass of observed baryons is directly proportional to the total mass which in turn scales with the observed rotation velocity. This potentially includes the case where the mass observed in baryons {\\it is\\/} the total mass (Milgrom 1983\\markcite{MOND}). Matching these observations is a substantial challenge for modern structure formation theories based on cold dark matter. These predict a slope which is too shallow (3 rather than 4, different by $8\\sigma$), and fail to anticipate that effectively all the baryons associated with a halo have cooled into the disk." }, "0003/astro-ph0003284_arXiv.txt": { "abstract": " ", "introduction": "Two results of the last years in the statistics of the gamma-ray bursts (GRBs) are doubtlessly remarkable. The first one concerns the number of subclasses. Recently, two different articles (\\cite{muk}; \\cite{hor}) simultaneously suggest that the earlier separation \\cite{kou} of GRBs into short and long subclasses is incomplete. (It is a common practice to call GRBs having $T_{90} <2$ s ($T_{90} > 2$ s) as short (long) GRBs, where $T_{90}$ is the time during which 90\\% of the fluence is accumulated \\cite{kou}.) These articles show that, in essence, the earlier long subclass alone should be further separated into a new \"intermediate\" subclass ($2$ s $< T_{90} < 10$ s) and into a \"truncated long\" subclass ($T_{90} > 10$ s). (In what follows, the long subclass will contain only the GRBs with $T_{90} > 10$ s, and the intermediate subclass will be considered as a new subclass.) The second result concerns the angular distribution of GRBs. At the last years several attempts (\\cite{hart}, \\cite{briggs}; $\\;$ \\cite{teg}; [Bal\\'azs et al. 1998]; \\cite{ba99}) were done either to confirm or to reject the randomness in the angular sky distribution of GRBs being collected at BATSE Catalog (\\cite{fish}; \\cite{mee}). Theoretically, if the intrinsic distribution of GRBs is actually random, an observation of some non-randomness is still expected due to the BATSE non-uniform sky-exposure function (\\cite{fish}; \\cite{mee}). Hartmann et al. (1991), Briggs et al. (1996) and Tegmark et al. (1996b) did not find any statistically significant departure from the randomness. On the other hand, the existence of {\\it some} non-randomness was confirmed on the $>99.9\\%$ confidence level by Bal\\'azs et al. (1998). This behavior can be caused either purely by instrumental effects or the instrumental effects alone do not explain fully the detected behavior and some intrinsic non-randomnesses should also exist. Bal\\'azs et al. (1998, 1999) suggest the second possibility. This conclusion follows from the result that while the short subclass shows a non-randomness, the intermediate + long subclasses do not indicate it. It is difficult to explain such behavior of subclasses by the instrumental effects alone. In this article we will again investigate the angular distribution of GRBs. Trivially, after the discovery of the new intermediate subclass, it is highly required to test the intrinsic randomness in the angular distribution of this new subclass, too. In addition, of course, new different tests, which exactly eliminate the effect of the sky-exposure function, are also required in order to complete the results of Bal\\'azs et al. (1998, 1999). The aim of this article is to test the {\\it intrinsic} randomness in the angular distribution of all GRBs and of the three subclasses separately, too. We will use a modification of the well-known counts-in-cells method. This is a standard and simple statistical test (see, cf., M\\'esz\\'aros (1997) and references therein). The advantage of this method is given by the fact that it allows to eliminate quite simply and exactly the sky-exposure function. The main result of paper will be the surprising conclusion that the intermediate subclass and only this subclass alone suggests a non-randomness on the 96.4\\% confidence level; its \"dimmer\" half even on the 99.3\\% confidence level. The paper is organized as follows. In Section 2 the method is described. In Section 3 the results of test are presented. Section 4 discusses and summarizes the results of the article. ", "conclusions": "The most surprising result of paper concerns the intermediate subclass. The intrinsic non-randomness is confirmed on the confidence level $>95\\%$. This confidence level, as discussed in Section 2, is \"remarkable\", but is not enough to reject the null-hypothesis of randomness. The results concerning the $339$ short GRBs should also be mentioned. Nevertheless, the $92\\%$ confidence level is clearly not enough to reject the confidence null-hypothesis. On the other hand, this result, together with Bal\\'azs et al. (1998, 1999), suggest that also for the short subclass itself the rejection of null-hypothesis of intrinsic randomness can also occur by further tests. In the case of $764$ long GRBs, and also of the $1284$ all GRBs, there are no indications for the non-randomnesses. All this seems to be in accordance with the results of Bal\\'azs et al. (1998, 1999). We think that the result concerning the intermediate subclass is highly surprising, because just this new subclass, having the smallest number of GRBs, has a remarkable \"proper\" behavior. A short further investigation of this subclass fully supports this conclusion. There are 181 GRBs in this intermediate subclass. Be divided this subclass into two further subclasses; into the \"dim\" and \"bright\" ones. By chance the peak flux = 2 photons/(cm$^2$s) (on $0.256$s trigger) is practically identical to the medium of peak flux for this subclass. Therefore, we consider the GRBs having smaller (bigger) peak flux 2 photons/(cm$^2$s) as the \"dim\" (\"bright\") subclass of the intermediate subclass. There are 92 GRBs at the \"dim\" subclass, and 89 GRBs at the \"bright\" one. We provide the 105 \"var/mean\" tests for these two parts, too. We obtain the surprising result that the \"dim\" subclass has an intrinsic non-randomness on the 99.3\\% confidence level ($\\omega = 7$). Contrary this, the \"bright\" subclass can still be random ($\\omega = 662$). The sky distribution of $92$ intermediate dim GRBs is shown on Figure 2. We mean that the behavior of the intermerdiate subclass of GRBs, quite independently, supports the correctness of the introduction of this new subclass (\\cite{muk}; \\cite{hor}). Further investigations of this new subclass are highly required. Three notes are still needed. First, purely from the statistical point of view, it must be precised that even the rejection of null-hypothesis of the intrinsic randomness would not mean a pure intrinsic non-randomness in the spatial angular distribution of GRBs. This is given by the fact that, up to now, it cannot be fully excluded that GRBs (or some part of them) are not unique phenomenons, and there can occur some repetitions, too. This question is studied intensively by several papers (\\cite{mee95}, \\cite{qua95}, \\cite{qua96}, \\cite{tega}, \\cite{graz}, \\cite{hak}) con\\-clu\\-ding that repetition can still play a role. Second, strictly speaking, the statistical counts-in-cells test is testing the \"complete spatial randomness\" (shortly the \"randomness\") of the distribution of GRB on the celestial sphere (\\cite{diggle}, Chapt.1.3). Therefore, in this paper we have kept this terminology. In cosmology, on the other hand, the word \"random\" (\"non-random\") is rarely used, and the word \"isotropic\" (\"anisotropic\") is usual (for the exact definition of isotropy in cosmology see, e.g., Weinberg (1972), Chapt. 14.1). Of course, here we will not go into the details of these terminology questions (see, e.g., Peebles (1980) for more details concerning these questions). We note only that the \"random-isotropic\" (\"non-random-anisotropic\") substitution is quite acceptable on the biggest angular scales; on smaller angular scales the situation is not so clear. Therefore, in Bal\\'azs et al. (1998, 1999), where only the angular scales $\\sim 90$ degrees and higher were studied, the words \"isotropy\" and \"anisotropy\" were quite usable. In this article, going down up to the scales $\\sim (20-25)$ degree, the used terminology is more relevant. Third, trivially, further studies are needed. They should test - by other different statistical methods - again the intrinsic randomnesses (more generally: the intrinsic spatial distributions \\cite{lamb}), both for all GRBs and for the subclasses. In addition, a test of the repetition alone, i.e. a test not being influenced by positions, is highly required. As the conclusion, the results of paper may be summarized as follows. \\begin{itemize} \\item We developed a method, which can verify quite simply the intrinsic randomness alone in the angular distribution of GRBs, because the method eliminates exactly the non-zero sky-exposure function. \\item We rejected the null-hypothesis of the intrinsic randomness in the angular distribution of 181 intermediate GRBs on the $96.4\\%$ confidence level. \\item We rejected the null-hypothesis of the intrinsic randomness in the angular distribution of 92 \"dim\" intermediate GRBs on the $99.3\\%$ confidence level. \\item We did not reject the null-hypotheses of the intrinsic randomnesses in the angular distribution of the remaining two subclasses and of the all GRBs, respectively, on the $>95\\%$ confidence levels; the \"bright\" intermediate GRBs seem to be distributed randomly, too. \\end{itemize} \\bigskip \\bigskip We thank the valuable discussions with Drs. Michael Briggs, Peter M\\'e\\-sz\\'a\\-ros, L\\'aszl\\'o P\\'asztor, Dennis Sciama, G\\'abor Tusn\\'ady and anonymous referee. One of us (A.M.) thanks for the hospitality at Konkoly Observatory and E\\\"{o}tv\\\"{o}s University. This article was partly supported by GAUK grant 36/97, by GA\\v{C}R grant 202/98/0522, by Domus Hungarica Scientiarium et Artium grant (A.M.), by OTKA grant T024027 (L.G.B) and by OTKA grant F029461 (I.H.). \\newpage" }, "0003/astro-ph0003417_arXiv.txt": { "abstract": "Observations of Young Star Cluster ({\\bf YSC}) systems in interacting galaxies are reviewed with particular emphasis on their Luminosity Functions ({\\bf LF}) and colour distributions. A few spectroscopic abundance measurements are available. They will be compared to YSC abundance predictions from spiral galaxy models. Evolutionary synthesis models allow to derive ages for individual YSCs on the basis of their broad band colours. With individual YSC ages models predict the future colour and luminosity evolution of the YSC systems that will be compared -- after a Hubble time -- to observations of old Globular Cluster ({\\bf GC}) systems. Using model M/L ratios as a function of age, YSC masses can be estimated. Age spread effects in young systems can cause the shape of the LF to substantially differ from the shape of the underlying mass function. Major sources of uncertainty are the metallicity, dust reddening, and observational colour uncertainties. ", "introduction": "GCs are conventionally believed to be (among) the oldest objects in the Universe, dating back to the times of galaxy formation. They are used to constrain the age of the Universe. LFs of GC systems are believed to be universal enough for determinations of distances to $> 20$ Mpc and of the Hubble constant. On the other hand, ``present-day GCs are the hardiest survivors of a larger original population'' (Harris 1991). Hence, their observed LF is not only their LF at formation shifted by stellar evolutionary fading, but might be additionally modified by cluster destruction processes. Dynamical modelling of cluster systems in the Galactic potential shows that destruction processes and timescales strongly depend on cluster masses. While destruction by dynamical friction is more efficient for high mass clusters, tidal shocking and evaporation preferentially destroy low mass clusters. Bright YSCs are observed in large numbers in interacting galaxies and merger remnants and a burning question with far-reaching implications is if these YSCs are young GCs. Masses derived for these YSCs are much higher than those of open clusters in the Milky Way. Effective radii of YSCs typically are a few pc similar to GC radii. The mass function ({\\bf MF}) of YSC systems in comparison with the MFs of molecular clouds or molecular cloud cores tells us about star and cluster formation processes. The MF of YSCs, as first pointed out by Meurer (1995), may differ in shape from their LF since M/L varies rapidly at young ages and the age spread within a YSC system is not much smaller than its age. The LF of open clusters in the Milky Way, the MFs of molecular clouds and molecular cloud cores, the observed LFs of YSCs, e.g. in NGC 4038/39, NGC 7252, NGC 3256, and of giant HII regions all are power laws with slopes in the range $\\alpha \\sim -1.5 \\, . \\, . \\, . \\, -1.8$ (cf. Solomon \\etal 1987, Lada \\etal 1991, Kennicutt 1989, and reviews by Harris \\& Pudritz 1994, Elmegreen \\& Efremov 1997). Yet the LF of {\\bf old} GCs is Gaussian with typical parameters ${\\rm \\langle M_V \\rangle \\sim -7.3 \\, mag, ~\\sigma (M_V) \\sim 1.3 \\, mag}$, their MF is log-normal with typically ${\\rm \\langle Log(M/M_{\\odot}) \\rangle \\sim 5.5, ~ \\sigma \\sim 0.5}$ (e.g. Ashman \\etal 1995). Hence the question as to the MF of YSCs has profound implications. If the MF of YSCs were a power law like their LF and if YSC systems are to evolve into something similar to old GC systems, dynamical destruction processes would have to transform the power law MF into a log-normal MF over a Hubble time. If, on the other hand, the MF of YSCs were log-normal (and their LF distorted to a power law by age spread effects), then the star/cluster formation process would have to transform the power law MF of the molecular clouds into the log-normal MF of YSCs. Or, else, might already the MF of molecular clouds/cloud cores in violently star forming mergers (where to my knowledge it has not yet been observed) be different from what it is in quietly star forming `normal galaxies' (where it is observed)? ", "conclusions": "" }, "0003/astro-ph0003129_arXiv.txt": { "abstract": "We reconsider the problem of gravitational structure formation inside and outside General Relativity (GR), both in the weakly and strongly non-linear regime. We show how these regimes can be explored observationally through clustering of high order cumulants and through the epoch of formation, abundance and clustering of collapse structures, using Press-Schechter formalism and its extensions. We address the question of how different are these predictions when using a non-standard theory of Gravity. We study examples of cosmologies that do not necessarily obey Einstein's field equations: scalar-tensor theories (STT), such as Brans-Dicke (BD), parametrized with $\\omega$, a non-standard parametrisation of the Hubble law, $H^2= a^{-3(1+\\epsilon)}$, or a non-standard cosmic equation of state $p=\\gamma\\rho$, where $\\gamma$ can be chosen irrespective of the cosmological parameters ($\\Omega_M$ and $\\Omega_\\Lambda$). We present some preliminary bounds on $\\gamma$, $\\omega$ and $\\epsilon$ from observations of the skewness and kurtosis in the APM Galaxy Survey. This test is independent of the overall normalization of rms fluctuations. We also show how abundances and formation times change under these assumptions. Upcoming data on non-linear growth will place strong constraints on such variations from the standard paradigm. ", "introduction": "\\footnotetext[1]{gazta@inaoep.mx, gaztanaga@ieec.fcr.es} \\footnotetext[2]{lobo@ffn.ub.es} In Cosmology the standard picture of gravitational growth, and also many aspects of fundamental physics, are extrapolated many orders of magnitude, from the scales and times where our current theory of gravity (General Relativity, GR) has been experimentally tested, into the distant universe. In particular, current limits on the (parametrized) Post Newtonian formalism mostly restrict to our very local Universe (see Will 1993). It is important to evaluate how much our predictions and cosmological picture depend on the underlying hypothesis (see Peebles 1999 for insightful comments on the state of this subject). The other side of this argument is that cosmology can be used to test fundamental physics, such as our theory of gravity. One aspect of GR that could be questioned or tested without modifying the basic structure or symmetry of the theory are Einstein's field equations, relating the energy content ($T_{\\mu\\nu}$) to the curvature ($R_{\\mu\\nu}$). One such modification, which will be considered here, is scalar-tensor theories (STT), such as Brans-Dicke (BD) theory. A more generic, but also more vague, way of testing the importance of Einstein's field equations is to model independently the geometry and the matter content, thus allowing for the possibility of other relations between them. Some simple aspects of this idea will be illustrated here by studying structure formation in a flat, matter dominated universe but with a more general growth law for the Hubble rate ---see section \\ref{sec32} below. Similarly, we will also consider results for a generic equation of state: $p=\\gamma \\rho$, where $\\gamma$ can be chosen independently of the cosmological parameters ($\\Omega_M$, $\\Omega_k$ and $\\Omega_\\Lambda$). Our aim in this paper is to explore certain variations of the standard model to see how they affect structure formation. The idea is to find a way to parameterize variations from GR that might produce differences large enough to be observable. The variations considered could have other observable consequences (eg in the local universe or in the radiation dominated regime) which might rule them out as a viable new theory. But even if this were the case, we still would have learn something about how structure formation depends on the underlying theory of Gravity or the assumptions about the equation of state. This aspect of the theory has hardly been explored and it therefore represents an important step forward in analyzing alternatives to the current paradigm, eg non-baryonic matter (see Peebles 1999), and could also help to set limits on variations of GR or the equation of state at high red-shifts. Here we consider two main regimes for structure formation in non-standard gravity/cosmology: weakly non-linear and strongly non-linear large scale clustering. We study the shear-free or spherical collapse (SC) model, which corresponds to the spherically symmetric (or local) dynamics (see below). This approximation works very well at least in two different contexts, that will be explored here. The first one is the growth of the smoothed 1-point cumulants of the probability distribution for large scale density fluctuations: the SC model turns out to reproduce exactly the leading order perturbation theory predictions (Bernardeau 1992), and turns out to be an excellent approximation for the exact dynamics as compared to N-body simulations both with Gaussian (Fosalba \\& Gazta\\~naga 1998a, 1998b) and non-Gaussian initial conditions (Gazta\\~naga \\& Fosalba 1998). The measured 1-point cumulants in galaxy catalogues have been compared with these predictions (eg Bouchet \\etal 1993, Gazta\\~naga 1992,1994, Gazta\\~naga \\& Frieman 1994, Baugh, Gazta\\~naga \\& Efstathiou 1995, Gazta\\~naga 1995, Baugh \\& Gazta\\~naga 1996, Colombi etal 1997, Hui \\& Gazta\\~naga 1999). The second one is the study of the epoch of formation and abundance of structures (such as galaxies and clusters), using the Press \\& Schechter (1974) formalism and its extensions (eg Bond \\etal 1991, Lacey \\& Cole 1993, Sheth \\& Lemson 1999, Scoccimarro \\etal 2000). Given some Gaussian initial conditions, this formalism can predict the number of structures (halos) of a given mass that will form at each stage of the evolution. One can use the SC model to predict the value of the critical linear over-density, $\\delta_c$, that will collapse into virialized halos. It turns out that the analytical predictions for the halo mass function and formation rates are remarkably accurate as compared to N-body simulations (Lacey \\& Cole 1994). One can also use this type of modeling to predict clustering properties of halos (eg Mo \\& White 1996, Mo, Jing \\& White 1997), cluster abundances (White, Efstathiou \\& Frenk 1993, Bahcall \\& Fan 1998) or weak lensing through mass functions (Jain \\& Van Waerbeke 2000). The observed cluster abundances have been used as a strong discriminant for cosmological models and also as a way to measure the amplitude of mass fluctuations, $\\sigma_8$ (see White, Efstathiou \\& Frenk 1993, Bahcall \\& Fan 1998). In summary, we propose to address a very specific question here: how different are the above non-linear predictions when using a non-standard cosmology and non-standard theory of Gravity? To answer this question we will consider two non-standard variations: scalar-tensor models and some examples of a cosmology that do not obey Einstein's field equations. The paper is organized as follows: In \\S2 we give a summary on how non-linear structure formation relates to the underlying theory of Gravity (see Weinberg 1972, Peebles 1993, Ellis 1999 and references therein, for a review on the relation between gravitational theory and cosmology). This section covers old ground with some detail as an introduction to later sections and for the reader that is not familiar with this subject or notation. We also present the more general case of an ideal (relativistic) fluid. As far as we know, some of the non-linear results presented here are new. In \\S3 we show how these predictions change in the two examples of non-standard gravity. Observational consequences are explored in \\S4. In \\S5 we present a discussion and the conclusions. ", "conclusions": "We have reconsidered the problem of non-linear structure formation in two different contexts that relate to observations: 1-point cumulants of large scale density fluctuations and the epoch of formation and abundance of structures using the Press \\& Schechter (1974) formalism. We have use the the shear-free or spherical collapse (SC) model, which is very good approximation for the above applications. We have addressed the question of how different are the predictions when using a non-standard theory of Gravity, such as BD model, or non-standard cosmological model (eg a different equation of state or Hubble law). Note that these are slight variations on the standard theme in the sense that they preserved the main ingredients of GR, such as the covariance and the geometrical aspects of the theory, including the same metric, with only slight changes in the field equations. We have also presented some preliminary bounds on $\\gamma$, $\\omega$ and $\\epsilon$ from observations of the skewness and kurtosis in the APM Galaxy Survey, eg Eq.[\\ref{bounds1}]-[\\ref{bounds2}]. These bounds are optimistic given the current data, but the situation is going to change rapidly, and one can hope to find much better bounds form upcoming data (such as 2DF or SDSS projects). In terms of the equation of state the bounds in Eq.[\\ref{bounds2}] would indicate that our Universe is neither radiation ($\\gamma=1/3$) or vacuum dominated ($\\gamma=-1$), but somewhere in between (eg matter dominated). In terms of the Gravitational constant, the bounds on $\\omega$ from Eq.[\\ref{bounds2}] would say that $G$ has not changed by more than $\\simeq 5\\%$ from $z \\simeq 1.15$, or by distances of $\\simeq 400 \\Mpc$. Clustering at higher red-shift would probe much larger scales and times. In terms of $\\epsilon$ the bounds Eq.[\\ref{bounds2}], would say that the Hubble law does not differ by more than $7\\%$ from the standard result (assumed here to be $\\epsilon=0$). We have also shown how halo and cluster abundances and formation times could change in these non-standard cases. The above bounds on $\\gamma$, $\\omega$ and $\\epsilon$ from observations of the skewness and kurtosis in the APM still allow significant changes (of several tens to hundreds of percent) on formation red-shifts $z_f$ and the corresponding abundances (see \\S\\ref{sec:abundances}). In the context of BD models the limits we find for $\\omega$ are less restrictive than the solar system limits $\\omega \\ga 100$. However, BD models allow $\\omega=\\omega(\\phi)$ so that $\\omega$ can increase with cosmic time, $\\omega=\\omega(z)$, in such a way that it could approach the general relativity predictions ($\\omega \\rightarrow \\infty$) at present time and still give significant deviations at earlier cosmological times. It is important to recall that our theory of gravity has only be tested on stellar distances (a.u.) while we want to use it on cosmological scales ($Mpc$). Our working example shows, for the first time, how non-linear effects are changed in such a model and sets the framework to study non-linear effects of more complicated (or realistic) Scalar-Tensor theories of gravity. It is straightforward to combine several of the changes proposed here to explore more general situations. One could for example parameterize theories in the ($\\gamma, \\omega$) plane, eg different equations of state with different BD parameters, or consider the whole ($\\gamma,\\omega,\\Omega_M,\\Omega_\\Lambda$) space. One could also consider a different equation of state for the $\\Lambda-$component, as in quintessence cosmologies (Caldwell, Dave, Steinhardt 1998), such models have already been used to predict cluster abundances within the ``standard'' cosmology (see Haiman, Mohr, Holder 2000 and references therein). This would obviously allow for a wider set of possible solutions and degeneracies. One should also consider other observational consequences of these variations, in particular relating to BD theory, such as the age of the Universe, the effects on CMB (eg see Chen \\& Kamionkowski 1999), radiation-matter transition (Liddle, Mazumdar \\& Barrow 1998), or the constraints from nucleosynthesis (Santiago \\etal 1997). These considerations could rule out some aspects of the proposed variations on the standard model, or might require more elaborate solutions (eg $\\omega=\\omega(\\phi)$ which implies $\\omega =\\omega(z)$). But even if this were the case, we still have learn a few new things about how structure formation depends on the underlying theory of Gravity, which is a first step towards further analysis of these issues. Throughout this paper we have assumed Gaussian initial conditions and no biasing. Both biasing (eg Fry \\& Gazta\\~naga 1993) and non-Gaussianities in the initial conditions (Gazta\\~naga \\& Fosalba 1998) would provide an additional source of degeneracy as they might produce similar effects as the non-standard variations presented here. This is the case for example when we have non-zero initial skewness or kurtosis, which could produced quite different values of $S_3$ and $S_4$ (eg see Gazta\\~naga \\& Mahonen 1996; Peebles 1999a,b; White 1999; Scoccimarro 2000), and therefore to the inferred values of $\\nu_2$ and $\\nu_3$. Biasing can have a very similar effect (eg see Mo, Jing \\& White 1997). One would also expect some level of degeneracy with biasing and initial conditions for cluster abundances or formation times (see Robinson, Gawiser \\& Silk 1998, Willick 1999). Rather than proposing an alternative theory of gravity or cosmological model, the aim of this paper was to show that some small deviations from the current paradigm have significant and measurable consequences for non-linear structure formation. This could eventually help explaining some of the current puzzles confronting the theory, such as the need of non-baryonic dark matter. Alternatively, current and upcoming observations of non-linear clustering and mass functions can be used to explore our assumptions and place limits on the theory of gravity at large ($\\ga 1 \\Mpc$) scales. This provides an interesting test for gravity as the driving force for structure formation and for our knowledge of the cosmological equation of state. A more comprehensive comparison with particular scenarios is left for future work." }, "0003/astro-ph0003403_arXiv.txt": { "abstract": "We perform direct 3-dimensional numerical simulations for magnetohydrodynamic (MHD) turbulence in a periodic box of size $2\\pi$ threaded by strong uniform magnetic fields. We use a pseudo-spectral code with hyperviscosity and hyperdiffusivity to solve the incompressible MHD equations. We analyze the structure of the eddies as a function of scale. A straightforward calculation of anisotropy in wavevector space shows that the anisotropy is scale-{\\it independent}. We discuss why this is {\\it not} the true scaling law and how the curvature of large-scale magnetic fields affects the power spectrum and leads to the wrong conclusion. When we correct for this effect, we find that the anisotropy of eddies depends on their size: smaller eddies are more elongated than larger ones along {\\it local} magnetic field lines. The results are consistent with the scaling law $\\tilde{k}_{\\parallel} \\sim \\tilde{k}_{\\perp}^{2/3}$ proposed by Goldreich and Sridhar (1995, 1997). Here $\\tilde{k}_{\\|}$ (and $\\tilde{k}_{\\perp}$) are wavenumbers measured relative to the local magnetic field direction. However, we see some systematic deviations which may be a sign of limitations to the model, or our inability to fully resolve the inertial range of turbulence in our simulations. ", "introduction": "Many astrophysical plasmas, including the interstellar medium and the solar wind, often show magnetic fields whose energy density is greater than or equal to the local kinetic energy density. In these plasmas the magnetic fields play a dominant dynamical role, mediated by magnetohydrodynamic (MHD) waves. In the incompressible limit, there are only two types of linear modes: shear Alfv\\'{e}n waves and pseudo Alfv\\'{e}n waves. While these two modes have different polarization directions, they have the same dispersion relation and propagate along the magnetic field lines at the Alfv\\'{e}n speed. Therefore, the nonlinear interactions of wave packets moving along the magnetic field lines at the Alfv\\'{e}n speed determine the dynamics of incompressible magnetized plasmas with a strong background field. In this paper, we study the anisotropy of the MHD turbulence in this regime. We will refer to this turbulence as incompressible Alfv\\'{e}nic turbulence. Nonlinear processes and the corresponding energy spectrum of incompressible Alfv\\'{e}nic turbulence are still among the most controversial problems in MHD. Since the pioneering works of Iroshnikov (1963) and Kraichnan (1965), the Iroshnikov-Kraichnan (IK) theory has been widely accepted as a model for incompressible, highly conducting MHD turbulence. The IK theory predicts $E_M(k) \\sim E_K(k) \\sim k^{-3/2}$ from a Kolmogorov-like dimensional analysis. Here, $E_M(k)$ and $E_K(k)$ are the magnetic and kinetic energy spectra respectively. In this framework, two counter-traveling eddies (i.e. Alfv\\'{e}n wave packets) interact and transfer energy to smaller spatial scales only when they collide, as they move in opposite directions along the magnetic field lines. Since the duration of such a collision is shorter than the conventional eddy turnover time by a factor of $t_v(l)/t_A(l)$, this collisional process is inefficient and the spectral energy transfer time as a function of scale $l$ ($=t_{cas}(l)$) increases by the same factor compared to the eddy turnover time ($l/v_l$) in ordinary hydrodynamic turbulence. Here $t_v(l)=l/v_l$ and $t_A(l)=l/V_A$ are eddy turnover time and Alfv\\'{e}n time respectively, $V_A \\equiv B/\\sqrt{4\\pi \\rho}$, and $B$ is the rms magnetic field strength. When the external field is strong, as assumed in IK theory, this quantity is usually set to $B_0$, the strength of the uniform background field. If the spectral energy cascade rate \\be \\epsilon \\sim \\frac{ v_l^2 }{ t_{cas}(l) } \\sim \\frac{ v_l^3 }{l} \\frac{ t_A(l) }{ t_v(l) } \\ee is scale-independent and $E_M(k) \\approx E_K(k)$, then we obtain the IK energy spectra. The IK theory assumes an isotropic distribution of energy in ${\\bf k}$-space. However, many researchers have argued that anisotropy is an important characteristic in MHD turbulence (for example, Shebalin et al 1983, Montgomery and Matthaeus 1995). This anisotropy results from the resonant conditions for 3-wave interactions (or 4-wave interactions, when 3-wave interactions are null). The resonant conditions for the 3-wave interactions are \\begin{eqnarray} {\\bf k}_1 + {\\bf k}_2 & = & {\\bf k}_3, \\\\ \\omega_1 + \\omega_2 & = & \\omega_3, \\end{eqnarray} where ${\\bf k}$'s are wavevectors and $\\omega$'s are wave frequencies. The first condition can be viewed as momentum conservation and the second as energy conservation. Alfv\\'{e}n waves satisfy the dispersion relation: $\\omega = V_A |k_{\\|}|$, where $k_{\\|}$ is the component of wavevector parallel to the background magnetic field. Since only opposite-traveling wave packets interact, ${\\bf k}_1$ and ${\\bf k}_2$ must have opposite signs. Then from equations (2) and (3), either $k_{\\|,1}$ or $k_{\\|,2}$ must be equal to 0. That is, zero frequency modes are essential for energy transfer. If $k_{\\|,2}=0$, we have \\begin{eqnarray} k_{\\|,1} &=& k_{\\|,3},\\\\ k_{\\|,2} &=& 0 \\end{eqnarray} (Shebalin et al 1983). Therefore, in the wavevector space, 3-wave interactions make energy cascade in directions perpendicular to the mean magnetic field. Since the energy cascade is strictly perpendicular to the mean magnetic field, the actives modes in wavevector space have a slab-like geometry with a constant width. The implication is that the nonlinear cascade of energy works against isotropy in ${\\bf k}$ space. Furthermore, it is important to note that equations (2) and (3) are true only when wave amplitudes are constant. In reality, nonlinear interactions provide a natural broadening mechanism, following the uncertainty relation, $\\Delta t \\cdot \\Delta \\omega \\sim 1$. In particular, if a wave has a frequency less than or comparable to the nonlinear interaction rate, it is effectively a zero frequency mode. Goldreich and Sridhar (1995, 1997) showed that in the strong incompressible shear Alfv\\'{e}nic turbulence regime (i.e. $\\tau_{NL}^{-1} \\sim \\omega$), these arguments lead to a new scaling law with a scale-dependent anisotropy. In this model smaller eddies are more elongated. Their arguments are based on the assumption of a {\\it critically balanced} cascade, $k_{\\parallel}V_A \\sim k_{\\perp}v_l$, where $k_{\\perp}$ and $k_{\\|}$ are wave numbers perpendicular and parallel to the external dc field\\footnote{Later in this paper, we will use $\\tilde{k}_{\\perp}$ and $\\tilde{k}_{\\|}$, instead of $k_{\\perp}$ and $k_{\\|}$, to represent their scaling relation in a slightly different, yet we believe equivalent, viewpoint. Here, $\\tilde{k}_{\\perp}$ and $ \\tilde{k}_{\\|}$ are wavenumbers with respect to the direction of the local magnetic field, not the external field.}. The argument given above for 3-wave interactions makes it clear that $k_{\\perp}$ will tend to increase until it becomes important in the plasma dynamics. The assumption of strong nonlinearity implies that wave packets lose their identity after they travel one wavelength along the field lines. Consequently the eddy turnover time ($(k_{\\perp}v_l)^{-1}$) is actually the same as Alfv\\'{e}nic time ($(k_{\\parallel}V_A)^{-1}$). In this model, the cascade time, $t_{cas}(l)$ can be determined without ambiguity: $t_{cas} \\approx (k_{\\perp}v_l)^{-1} \\approx (k_{\\parallel}V_A)^{-1}$. Since the cascade time is comparable to the period of the Alfv\\'{e}n wave, the 3-wave resonant condition can be violated according to the uncertainty relation $\\Delta \\omega \\cdot \\Delta t \\sim V_A k_{\\|} \\cdot t_{cas} \\sim 1$. The quantity $k_{\\|}$ is the width of the active region in wavevector space. {}Finally the assumption of a scale-independent cascade rate $\\epsilon \\sim v_l^2/t_{cas}(l)\\sim E_{waves}V_A/L$ gives \\bea k_{\\parallel} \\sim k_{\\perp}^{2/3} L^{-1/3} \\left({E_{waves}\\over V_A^2}\\right)^{1/3}, \\\\ v_l \\sim V_A(k_{\\perp}L)^{-1/3} \\left({E_{waves}\\over V_A^2}\\right)^{1/3}, \\eea where $E_{waves}$ is the wave energy per mass. These formulae assume that all scales, from $L$ on down, are within the inertial range of MHD turbulence. Here the typical $k_{\\|}$ should be interpreted as the size of the range of parallel wavevectors, corresponding to a given $k_{\\perp}$, that contain significant energy. Matthaeus et al. (1998) recently tested this model numerically, and showed that the anisotropy of low frequency MHD turbulence scales linearly with the ratio of perturbed and total magnetic field strength ($b/B$), a result which seems inconsistent with Goldreich and Sridhar's model. To explain this scaling relation, they suggested that the region of Fourier space where the energy transfer takes place actively is given by \\be |{\\bf k}\\cdot\\frac{{\\bf B}_0}{\\sqrt{4\\pi\\rho}} |< \\frac{1}{ \\tau_{NL} }, \\ee where $\\tau_{NL}$ is the eddy turnover time of the energy containing length $L$. Consequently, ``the region of the wave number space where spectral transfer is most vigorous'' has a slab-like geometry with a constant width proportional to $1/(\\tau_{NL}B_0)$. All these theories (except the IK theory) share a common prediction for anisotropy: anisotropy should be more pronounced on smaller scales. Oughton et al. (1994) and Ghosh and Goldstein (1997) already reported this scale-dependent anisotropy through numerical simulations. The former extended Shebalin et al. (1983)'s 2-D calculations to 3-D cases. To measure anisotropy, they introduced the Shebalin angles, $\\theta_Q$, defined by $ \\tan^2{\\theta_Q}=( \\sum k_{\\perp}^2 |{\\bf Q}({\\bf k},t)|^2 )/ ( \\sum k_{\\|}^2 |{\\bf Q}({\\bf k},t)|^2 ), $ where ${\\bf Q}$ is vector potential {\\bf A}, magnetic field {\\bf B}, or current {\\bf J}, etc. Greater $\\theta_Q$ means greater anisotropy. They found that $\\theta_A < \\theta_B < \\theta_J$. If the energy spectrum of {\\bf B} scales as $E_M(k)\\propto k^{-s}$, then the spectra of vector potential and current scale as $E_A(k)\\propto k^{-s-2}$ and $E_J(k)\\propto k^{-s+2}$, respectively. The spectra of vector potential has the steepest slope among the three spectra. This means that the vector potential is least strongly dependent on small scales (and the current is most strongly dependent on small scales). Therefore they concluded that anisotropies are more pronounced at smaller scales\\footnote{ The width of the active region in Fourier space, $k_{\\|}$, is a function of $k_{\\perp}$. If $$ \\left[\\frac{ k_{\\parallel} }{ k_{\\perp} } \\right]_{\\mbox{small $k_{\\perp}$}} >\\left[\\frac{ k_{\\parallel} }{ k_{\\perp} } \\right]_{\\mbox{large $k_{\\perp}$}},$$ then we will have the ordering of Shebalin angles as observed by Shebalin et al.\\ and Oughton et al. However, if $k_{\\|} \\propto k_{\\perp}$, then we do not expect any ordering among the angles. In \\S3, we will find that, no matter what the true functional form $k_{\\|} = k_{\\|}(k_{\\perp})$ is, Fourier transformation smooths out the {\\it true} relation and leads to a {\\it fake} linear relationship between $k_{\\|}$ and $k_{\\perp}$. This suggests that their results are in contradiction to our discussion in \\S3. However, the apparent linear relationship between $k_{\\|}$ and $k_{\\perp}$ in \\S3 is not perfectly linear. Instead, we expect $k_{\\parallel}=c_1k_{\\perp}+c_2$, where $c_1$ is a decreasing function of $B_0$ and $c_2$ depends on the $k_{\\|}$ of forcing terms (or initial values of $k_{\\|}$ for decaying turbulence). The presence of $c_2$ does not seem to be important in our Fig. 3 in \\S3. However, it does affect the calculation of Shebalin angles. That is, because of $c_2$, the ratio $k_{\\|}/k_{\\perp}=c_1 + c_2/k_{\\perp}$ becomes a function of $k_{\\perp}$. At the largest energy containing eddy scale, $k_{\\perp} \\sim c_2$ and hence $k_{\\|}/k_{\\perp}\\sim c_1 +O(1)\\sim O(1)$. But, at small scales, the ratio can be much smaller than unity. Therefore, because of $c_2$, we can obtain a scale-dependent anisotropy: $(k_{\\|}/k_{\\perp})$ at small $k_{\\perp}$ is greater than that at large $k_{\\perp}$. Note that this is a result of the initial conditions, rather than a true scaling relation. This will lead the ordering of the angles as observed by Shebalin et al.\\ and Oughton et al.. This interpretation is qualitatively consistant with their results. For example. the ratio $k_{\\|}/k_{\\perp}=c_1 + c_2/k_{\\perp}^{peak}$, therefore $(\\tan{\\theta_Q})^{-1}$, approaches to a constant value $c_2/k_{\\perp}^{peak}$ as $B_0$ becomes strong. Here $k_{\\perp}^{peak}$ is the wavenumber of the peak of the energy spectrum. It is also possible to explain the increased anisotropy at high magnetic Reynolds numbers. If the magnetic Reynolds number increases, then $k_{\\perp}^{peak}$ increases and, therefore, the ratio decreases. }. They also found a similar ordering for velocity field (and vorticity, etc). On the other hand, Ghosh and Goldstein (1997) calculated the Shebalin angles as a function of wavenumber bin. They found that the Hall-MHD\\footnote{Hall MHD includes the Hall term, which is important at ion-cyclotron scales. In this paper, we consider only the standard MHD equations.} simulations show increased anisotropy at small scales (i.e. greater Shebalin angles at smaller scales). However their standard MHD simulations do not show increased anisotropies at small scales. We refrain from comparing their work with ours because they used different physics (the Hall effect) and their simulations are $2\\slantfrac{1}{2}$ dimensional. The simulations given in this paper are 3-D standard MHD simulations. We note that none of the previous papers quantitatively compared their results with particular theories of anisotropy. In this paper, we examine the scaling law for Alfv\\'{e}nic MHD turbulence numerically and resolve the controversy concerning the anisotropic structure of the turbulence. Our results are consistent with Goldreich and Sridhar's model. We describe our numerical methods in \\S 2. In \\S 3, we describe our results for anisotropy in wavevector space. In this section, we demonstrate that none of the scaling laws mentioned above agrees with our data, and explain why a straightforward evaluation of the distribution of spectral power does not correspond to a physically meaningful set of scaling laws. We attribute this to the systematic effects of large scale curvature of the magnetic fields. In \\S 4, we determine the shape of individual eddies, avoiding the systematic error described above. We compare our results to Goldreich and Sridhar's model and give our conclusions in \\S 5. ", "conclusions": "Here we rederive the scaling law $ \\tilde{k}_{\\parallel} \\propto \\tilde{k}_{\\perp}^{2/3}$ in the framework of 3-wave interactions. Except for the use of the uncertainty principle, the work in this section is independent of Goldreich and Sridhar's derivation. As noted by Goldreich and Sridhar (1995), 3-wave interactions are an adequate proxy for wave-wave interactions of all orders in a strong MHD turbulence. As long as we assume the locality of interactions, it is pointless to distinguish $k_{\\|}$ and $k_{\\perp}$ from $\\tilde{k}_{\\|}$ and $\\tilde{k}_{\\perp}$. Hence, for simplicity, we use $k_{\\|}$ and $k_{\\perp}$ instead of $\\tilde{k}_{\\|}$ and $\\tilde{k}_{\\perp}$ during the derivation. Suppose the 3-dimensional energy spectrum is given by \\begin{equation} E_3(k_{\\perp},k_{\\parallel})\\equiv |\\hat{{\\bf V}}({\\bf k})|^2 \\propto k_{\\perp}^{-2\\alpha} f, \\end{equation} where $\\hat{{\\bf V}}({\\bf k})$ is the amplitude of the mode whose wavevector is ${\\bf k}$ and $f(u)$ is a positive, symmetric function of $u$ (cf. equation (7) of Goldreich and Sridhar (1995)) which describes the power distribution as a function of eddy shape. If the $width$ (or, $thickness$) of the energy spectrum in the direction of $k_{\\parallel}$ is $k_{\\perp}^{\\beta}$, then we can write \\be E_3(k_{\\perp},k_{\\parallel}) \\propto k_{\\perp}^{-2\\alpha} f(k_{\\parallel}/k_{\\perp}^{\\beta}). \\ee If the $width$ is caused by the uncertainty principle ($\\Delta t \\cdot \\Delta \\omega \\approx 1$ with $\\Delta t \\propto t_{cas}(l)$ and $\\Delta \\omega \\propto k_{\\parallel}$), then \\be t_{cas}(l) \\propto k_{\\perp}^{-\\beta}. \\ee Suppose the energy cascade rate $\\epsilon \\sim v_l^2/t_{cas}(l)$, where $l = 2\\pi/k_{\\perp}$, is scale-independent. Because $v_l^2 \\sim k_{\\perp}^{-2\\alpha} k_{\\perp}^2 k_{\\perp}^{\\beta}$ ($\\sim k_{\\perp} E(k_{\\perp})$, $E(k_{\\perp})=$ 1-dimensional spectrum) and $t_{cas}(l) \\sim k_{\\perp}^{-\\beta}$, we have \\be k_{\\perp}^{-2\\alpha+2+2\\beta} = const. \\ee Therefore, \\begin{equation} 1-\\alpha + \\beta = 0. \\label{c1} \\end{equation} Now, let's pick up a mode at a wavevector ${\\bf p}$ and consider nonlinear interactions with other wave modes. First, the strength of the interaction with another mode at ${\\bf q}$ is $ \\propto p |\\hat{ {\\bf V} }({\\bf p})| |\\hat{ {\\bf V} }({\\bf q})|$. (This comes from the $(\\nabla \\times {\\bf V}) \\times {\\bf V}$ term in equation (9)) Hereafter we assume $p\\equiv |{\\bf p}|\\approx p_{\\perp}$. Second, the number of interactions is $\\propto p^2 p^{\\beta}$. This is the number of modes $near$ ${\\bf p}$. Here we use locality of 3-wave interactions. If the interactions are random, the net change of amplitude per unit time will be the strength of the interaction times the square root of the number of interactions, or \\be |\\Delta \\hat{ {\\bf V} }({\\bf p})| \\propto p |\\hat{ {\\bf V} }({\\bf p})| |\\hat{ {\\bf V} }({\\bf q})|\\cdot (p^2 p^{\\beta})^{1/2}. \\ee Therefore, we have \\be t_{cas} \\propto |\\hat{ {\\bf V} }({\\bf p})|/|\\Delta \\hat{ {\\bf V} }({\\bf p})| \\propto p^{-2} p^{\\alpha} p^{-\\beta/2}, \\ee where we assumed $p\\propto q$. Equating this with $t_{cas}\\propto p^{-\\beta}$, we can write \\begin{equation} \\alpha - 2 = -\\beta/2. \\label{c2} \\end{equation} {}From equations (\\ref{c1}) and (\\ref{c2}), we have \\begin{equation} \\alpha=5/3, \\beta=2/3, \\end{equation} which is just the result of Goldreich and Sridhar (1995): \\be k_{\\parallel} \\propto k_{\\perp}^{2/3}. \\ee As a consequence, the 3-D energy spectrum becomes \\be E_3(k_{\\perp},k_{\\parallel})\\propto k_{\\perp}^{-10/3} f(k_{\\parallel}/k_{\\perp}^{2/3}) \\ee and the corresponding 1-D spectrum is given by \\be E(k) \\propto k^{-5/3}. \\ee In summary, we have shown that the anisotropy of Alfv\\'{e}nic turbulence depends on the spatial scales of eddies. In particular, our results confirm the claim by Goldreich and Sridhar (1995, 1997) that smaller eddies are relatively more elongated along the direction of the local magnetic field lines than larger ones. Quantitative measurements of the anisotropy using the velocity fields show good agreement with their proposed scaling law, $k_{\\parallel}\\sim k_{\\perp}^{2/3}$ as long we interpret these wavenumbers as referring to the {\\it local} magnetic field direction. However, when the external magnetic field is very strong, magnetic fields scale somewhat differently, showing a slightly more rapid increase in anisotropy at smaller scales. It is important to note that the correct scaling laws depend on comparing the eddy shape to the {\\it local} magnetic field direction. As a final note, we wish to stress that our results are not in agreement with the IK theory. The IK theory is based on the assumption of isotropy in wavenumber space, which may be true when the external magnetic field is very weak or zero. However, even in these cases, the turbulence is globally isotropic, but locally very anisotropic. In this paper, we showed that eddies do show anisotropy and that the anisotropy is scale-dependent when there is a strong large scale field (which should apply to very small scales within any MHD turbulence cascade). On the other hand, our results are consistent with Goldreich and Sridhar's theory of strong MHD turbulence. More precisely, if we consider the ratio of hydrodynamic to Alfv\\'{e}nic rates, that is $(k v_k/ \\tilde{k}_{\\|} V_A)$, we find from equations (18) and (19) that \\begin{equation} {k v_k\\over \\tilde{k}_{\\|} V_A}\\propto \\tilde{k}_{\\perp}^{0.3-0.5} v_k, \\end{equation} where $k\\approx \\tilde{k}_{\\perp}$. {}From Fig. 2 we see that for the inertial range this implies a ratio which is either constant or increasing with wavenumber. A constant ratio is predicted by Goldreich and Sridhar's model. The IK model predicts a slow decline." }, "0003/astro-ph0003290_arXiv.txt": { "abstract": "We present new near-infrared images of $z>0.8$ radio galaxies from the flux-limited 7C-{\\sc iii} sample of radio sources for which we have recently obtained almost complete spectroscopic redshifts. The 7C objects have radio luminosities $\\approx 20$ times fainter than 3C radio galaxies at a given redshift. The absolute magnitudes of the underlying host galaxies and their scale sizes are only weakly dependent on radio luminosity. Radio galaxy hosts at $z\\sim 2$ are significantly brighter than the hosts of radio-quiet quasars at similar redshifts and the model AGN hosts of Kauffmann \\& Haehnelt (2000). There is no evidence for strong evolution in scale size, which shows a large scatter at all redshifts. The hosts brighten significantly with redshift, consistent with the passive evolution of a stellar population that formed at $z\\stackrel{>}{_{\\sim}} 3$. This scenario is consistent with studies of host galaxy morphology and submillimeter continuum emission, both of which show strong evolution at $z\\stackrel{>}{_{\\sim}}2.5$. The lack of a strong ``redshift cutoff'' in the radio luminosity function to $z>4$ suggests that the formation epoch of the radio galaxy host population lasts $\\stackrel{>}{_{\\sim}}1$ Gyr from $z\\stackrel{>}{_{\\sim}}5$ to $z\\sim 3$. We suggest these facts are best explained by models in which the most massive galaxies and their associated AGN form early due to high baryon densities in the centres of their dark matter haloes. ", "introduction": "At low redshifts, FRI and FRII radio sources with radio luminosities at 151 MHz of $L_{R (151)} \\stackrel{>}{_{\\sim}} 10^{24} {\\rm WHz^{-1}}$ are associated almost exclusively with giant elliptical host galaxies. If this continues to be the case out to high redshift, then radio galaxies can give us a unique insight into the formation and evolution of a single class of massive galaxy. This is particularly exciting in the light of submillimetre detections of $z\\sim 4$ radio galaxies (e.g.\\ Archibald et al.\\ 2000) which may indicate that we can see these objects during their major bursts of star formation [although see also Willott, Rawlings \\& Jarvis (2000)]. Furthermore, the similarity of radio galaxy hosts, in contrast to the wide range in luminosity of radio-quiet quasar hosts (McLure et al.\\ 1999; Ridgway et al.\\ 2000), suggests that one of the conditions necessary for producing powerful radio jets is the presence of a massive spheroidal component, and therefore a supermassive ($\\stackrel{>}{_{\\sim}} 10^9 M_{\\odot})$ black hole (Lacy, Ridgway \\& Trentham 2000). Studies of high redshift radio galaxy hosts have traditionally concentrated on the $K-z$ Hubble Diagram. Work on the 3C and 1 Jy samples by Lilly (1989 and refs.\\ therein) initially pointed to a passively-evolving stellar host formed at high redshift which evolved into the giant elliptical radio galaxy hosts seen today, but this was challenged when some high redshift radio galaxies were found to have significant emission line contributions to their $K$-band light (Eales \\& Rawlings 1993, 1996). Only by finding low AGN-luminosity radio galaxies at high redshift could the controversy be resolved. Eales et al.\\ (1997) used the 6C sample, a factor of five fainter in radio flux than the 3C sample, to show that there did seem to be a radio luminosity dependence of host galaxy magnitude. Further work by Roche, Eales \\& Rawlings (1998) indicated that the hosts of 6C radio galaxies were not only significantly fainter than their 3C counterparts, but also had smaller scale sizes. We have used the 7C-{\\sc iii} radio galaxy redshift survey of Lacy et al.\\ (1999b) to select a complete sample of $z>0.8$ radio galaxies. The 7C redshift surveys are a factor of 4--5 lower still in luminosity at a given redshift than the 6C sample of Eales et al.\\ (1997), and thus allow the study of high redshift radio galaxy hosts over a wide range in radio luminosity (Willott et al.\\ 1999). The 7C-{\\sc iii} sources were imaged in the near-infrared on the 3-m NASA Infrared Telescope Facility (IRTF) and the 3-m Shane Telescope at Lick Observatory. These data have allowed us to further investigate the radio luminosity dependence of host properties at $z\\sim 1$ and, because our 7C objects at $z\\sim 2$ have similar radio luminosities to $z\\sim 1$ 6C radio galaxies and $z\\sim 0.3$ 3C radio galaxies, we can also investigate host galaxy evolution over a wide range in redshift. We assume a cosmology with $\\Omega_{\\rm M} =1, \\Omega_{\\Lambda}=0$ and $H_0=50 \\, {\\rm kms^{-1}Mpc^{-1}}$ except where otherwise stated. \\begin{deluxetable}{llclll} \\footnotesize \\tablecaption{Observing log \\label{tbl-1}} \\tablewidth{0pt} \\tablehead{ \\colhead{Object} & \\colhead{Telescope} & \\colhead{$z$} & \\colhead{Filter} & \\colhead{$t_{\\rm int}$/min} & \\colhead{FWHM PSF/$^{''}$}} \\startdata 7C 1733+6719 &IRTF &1.84 & $K^{'}$ & 18 & 0.95\\\\ 7C 1740+6640 &IRTF &2.10 & $K^{'}$ & 18 & 0.65\\\\ 7C 1741+6704 &IRTF &1.05 & $J$ & 13.5& 0.65\\\\ 7C 1742+6346 &IRTF &1.27 & $H$ & 18 & 0.90\\\\ 7C 1748+6703 &Shane& ? & $J$ \\& $K^{'}$& 56 & 1.3\\\\ & & & & & \\\\ 7C 1748+6657 &IRTF &1.05 & $J$ & 13.5& 0.90\\\\ 7C 1751+6809 &IRTF &1.54 & $H$ & 18 & 0.56\\\\ 7C 1753+6311 &Shane&1.96?& $J$ \\& $K^{'}$& 82 & 1.5\\\\ 7C 1754+6420 &IRTF &1.09 & $J$ & 13.5& 1.07\\\\ 7C 1756+6520 &IRTF &1.48?& $K^{'}$ & 18 & 0.78\\\\ & & & & & \\\\ 7C 1758+6719 &IRTF &2.70 & $K^{'}$ & 82 & 0.42\\\\ 7C 1802+6456 &IRTF &2.11 & $K^{'}$ & 18 & 0.74\\\\ 7C 1804+6313 &IRTF &? & $K^{'}$ & 18 & 0.80\\\\ 7C 1805+6332 &IRTF &1.84 & $K^{'}$ & 36 & 0.63\\\\ 7C 1807+6841 &IRTF &0.82 & $J$ & 13.5& 0.72\\\\ & & & & & \\\\ 7C 1807+6719 &IRTF &2.78 & $K^{'}$ & 36 & 0.75\\\\ 7C 1812+6814 &IRTF &1.08 & $J$ & 13.5& 0.93\\\\ 7C 1814+6702 &IRTF &4.05?& $K^{'}$ & 54 & 0.78\\\\ &Shane& & $K^{'}$ & 54 & 1.3\\\\ 7C 1814+6529 &IRTF &0.96 & $J$ & 13.5& 0.84\\\\ 7C 1820+6657 &IRTF &2.98 & $K^{'}$ & 36 & 0.69\\\\ & & & & & \\\\ 7C 1816+6605 &IRTF &0.92 & $J$ & 13.5& 1.02\\\\ 7C 1825+6602 &IRTF &2.38 & $K^{'}$ & 18 & 0.81\\\\ \\enddata \\end{deluxetable} ", "conclusions": "\\subsection{Radio-luminosity and redshift dependence of absolute magnitudes} In Fig.\\ 3 we plot the rest-frame $R$-band magnitudes against redshift for several samples of radio galaxies. We plot the $z>0.8$ 7C-{\\sc iii} objects with spectroscopic redshifts and infrared imaging, $z>0.8$ 6C radio galaxies with photometry from Eales et al.\\ (1997), $z>0.8$ 3C radio galaxies in the Laing, Riley \\& Longair (1983; LRL) complete sample with photometry from Best, Longair \\& R\\\"{o}ttgering (1998) [apart from 3C22 which is a lightly-reddened quasar (Rawlings et al.\\ 1995) and therefore excluded from the sample, 3C~175.1 ($z=0.92$) which has photometry from Ridgway \\& Stockton (1997) and 3C~263.1 ($z=0.824$) which has photometry from Eales (personal communication)]. We have also added local radio galaxies in LRL from Owen \\& Laing (1989) and the HZ sample of galaxies of Table 3. All these objects, with the exception of three FRIs out of the 24 objects in the Owen \\& Laing LRL sample are either FRII or compact steep-spectrum sources. All the $z>0.8$ 7C-{\\sc iii} sources at $z>0.8$ are well above the FRI/FRII boundary in radio luminosity, with $L_{R (151)} \\stackrel{>}{_{\\sim}} 10^{26} {\\rm WHz^{-1}sr^{-1}}$ compared to the FRI/FRII boundary at $L_{R (151)} \\sim 10^{25} {\\rm WHz^{-1}sr^{-1}}$. There is a clear trend for redshift and absolute magnitude to correlate, even if the incomplete HZ sample is excluded. The 3C, 6C and 7C samples are all complete samples, selected on the basis of low frequency radio flux only, and are nearly completely identified. Thus the only selection effect which needs to be considered for these samples is the tendency for redshift and luminosity to correlate within each flux limited sample. With the wide range in radio luminosities in the complete samples at $z\\sim 1$, however, we can separate out the luminosity dependence. This is illustrated in Fig.\\ 4, where we have plotted absolute magnitude against radio luminosity for 3C, 6C and 7C galaxies in the redshift range $0.8}{_{\\sim}}3$ has long been used as an argument for a high redshift of radio galaxy formation (e.g.\\ Lilly 1989). This low scatter is thought to come about because radio galaxies form at $z>3$ then evolve along similar passive evolution tracks to end at radio galaxy hosts today. With a few modifications this simple picture still seems to be basically valid. Fig.\\ 3 and Table 4 show that the dispersion in absolute magnitudes to $z\\sim 4$ is indeed very low. After correction for AGN-contamination, however, the dispersion in $L_0$ for the $z>1.8$ objects seems to be higher than for the uncorrected objects (Table 4). As discussed above, the correlation of absolute magnitude with redshift is also less tight after correction. At first sight it seems paradoxical that correction for luminosity-dependent effects should {\\em increase} the scatter in the absolute magnitudes. However, this can be be understood if the luminosity-dependent contributions $(\\alpha + \\beta)L_{R (151)}^{0.8}$ are approximately equal to the highest values of $L_0$. (The $z>1.8$ objects have a very narrow range in $L_R$ as the range in flux of these objects is low, all except VLA 123642+621331 having 151 MHz fluxes between 0.5 and 10 Jy.) The addition of the luminosity-dependent contributions can then boost the objects with low $L_0$ by a significant factor, whereas the addition to the objects with high $L_0$ results in only a relatively small fractional increase in the luminosity (see also de Vries 1999). The result of this is a reduction of the scatter in the total luminosities. We have used the F-test for variances to examine the statistical significance of the increase in scatter of the $L_0$ values. Comparing the objects with $0.82.8$, our sample contains two upper limits on the magnitudes, so we give a lower limit to the scatter of $\\sigma_{n-1}>0.97$. This is again significantly higher than that in the range $0.8$-24.3& $>$-25.1&$>$0.7 & $>$1.0\\\\ \\tableline \\end{tabular} \\tablenotetext{a}{Number of objects in each sub-sample} \\tablenotetext{b}{Standard deviation of the sample.} \\tablecomments{$M_{R}$ is measured in a 63.9 kpc metric aperture. Uncorrected magnitudes have $k$-corrections only, no corrections for emission line or AGN luminosity have been applied. Corrected magnitudes have, in addition to $k$-corrections, emission-line contamination corrections and a correction for AGN luminosity as detailed in Section 4.1.} \\end{table} \\twocolumn \\begin{figure} \\plotone{figure6.ps} \\caption{ The dependence of half-light radius on host absolute magnitudes. Errors are shown for the 7C objects only. The dot-dashed line represents the correlation observed for local ellipticals (equation (2) with $\\Delta R = 0$); the dashed line the same relationship corrected to $z\\sim 1$ using the mean properties of the 3C hosts (i.e.\\ with $\\Delta R = 0.68$).} \\end{figure} \\begin{figure} \\plotone{figure7.ps} \\caption{Half-light radius in arcseconds versus redshift. The lines show the angular size of a standard rod of length equal to the mean of the half-light radii of local FRII host galaxies (13 kpc) as a function of redshift in an $\\Omega_{M}=1, \\Omega_{\\Lambda}=0$ cosmology (dotted line), an $\\Omega_{M}=0.3, \\Omega_{\\Lambda}=0.7$ cosmology (long-dashed line) and an $\\Omega_{M}=0, \\Omega_{\\Lambda}=0$ cosmology (short-dashed line).} \\end{figure} \\subsection{The scale sizes of the hosts} Roche et al.\\ (1998) have shown that the scale sizes of the $1$ 250 GeV) gamma rays and, along with Mrk 421, one of only two that have been confirmed as VHE sources (for a review, see Catanese \\& Weekes 1999). Mrk 501 has also been detected as a source of gamma rays at GeV energies with the Energy Gamma-Ray Experiment Telescope (EGRET; Kataoka et al. 1999) and at a few hundred keV by the Oriented Scintillation Spectrometer Experiment (OSSE; Catanese et al. 1997) on the {\\it Compton Gamma-Ray Observatory}. Just as all other blazars, Mrk 501 exhibits rapid, large amplitude variability over a wide range of wavelengths. In X-rays, variations of from 30\\% to 300\\% were observed on time scales of days in 1997 during a high emission state \\citep{Pian98,Lamer98,Catanese99b} but no sub-day scale flares were seen. The fastest variation observed in X-rays was a flux increase of approximately 20\\% in about 12 hours seen with EXOSAT in 1986 \\citep{Giommi90}. Spectral variability in X-rays from Mrk 501 has been both moderate, with changes in the spectral index of $\\sim$0.1-0.3 on several day scales (e.g., Pian et al. 1998), and rapid, with spectral variations of $\\sim$0.5 on 2-3 day time-scales, observed in 1998 June \\citep{Sambruna99a}. The Whipple Observatory has observed VHE gamma-ray variations spanning a factor of $>$70 in flux in four years of observations and has observed significant variability with time-scales as short as 2 hours \\citep{Quinn99}. Similar variability ranges are observed by other VHE telescopes \\citep{Hayashida98,Aharonian99a,Djannati99}. In the R-band, \\citet{Miller99} reported the detection of a flare in which the flux increased by 4\\% (from $V_R = 13.90$ to 13.80) in 15 minutes with a decay to the previous level in the same amount of time. Multi-wavelength observations have revealed correlations between VHE gamma rays and X-rays in this object \\citep{Catanese97} and in 1997, the synchrotron spectrum of Mrk 501 was observed to extend up to approximately 100 keV \\citep{Catanese97,Pian98,Catanese99b}, the highest seen in any blazar and a 50-fold increase over what was observed only one year before \\citep{Kataoka99}. This behavior has established Mrk 501 as the prototype for a subset of BL Lacs that exhibit large shifts in the peak of their synchrotron spectra during flares. In this paper, we report on observations of Mrk 501 taken with the {\\it Rossi X-ray Timing Explorer} ({\\it RXTE}) in 1998 May as part of a multi-wavelength campaign. The full multi-wavelength results will be reported in a future work. Here, we concentrate on the observation of a very rapid flare and discuss its implications. ", "conclusions": "\\label{discuss} The variation in the spectral index during the course of the flare can provide insights into the dominant flaring timescales and acceleration process. As discussed by \\citet{Kirk99}, for a flare in which the variability and acceleration time-scales are much less than the cooling time-scale a plot of the spectral index versus flux should follow a clockwise pattern, i.e., the harder energies vary first. For a flare where the variability, acceleration, and cooling time-scales are similar, the spectral index versus flux diagram should move in a counter-clockwise direction, i.e., the softer energies vary first because the number of particles changes due to the acceleration process which proceeds from low energy to high energy. Clockwise patterns are most commonly observed in the TeV sources Mrk 421 (e.g., Takahashi et al. 1996) and PKS 2155-304 (e.g., Kataoka et al. 2000) but counter-clockwise patterns have been recently observed from these objects \\citep{Fossati99,Sambruna99b}. Because the data do not have sufficient statistics to plot spectral index versus flux on such short time scales, we instead plot the hardness ratio (10-15 keV count rate/2-10 keV count rate) versus flux for the flare on May 25 in Figure~\\ref{hard_flux}. The numbers in the plot indicate the development of the hardness ratio in time during the flare. The large cluster of filled circles represents the observations before the onset of the flare. The filled triangles represent the rising part of the flare. The point indicated by the tail of the arrow marked with the ``1'' is the last low flux point before the flare starts. The filled squares are data taken during the decay of the flare. During the rise of the flare, the hardness ratio increases steadily. During the decay of the flare, there is a slight trend for the hardness ratio to increase even further. Thus, these observations are consistent with a counter-clockwise pattern. A clockwise pattern seems precluded by the significantly harder spectrum during the decay of the flare than the rise (see Table~\\ref{spec_fits}) but other patterns in the hardness ratio versus flux diagram cannot be ruled out given the statistical errors in this observation. The counter-clockwise pattern and the large shift in the synchrotron peak imply that the acceleration process dominates the development of the flare, accelerating a fresh population of high energy electrons which causes the flare. In summary, {\\it RXTE} observations of Mrk 501 in 1998 May have revealed a flux state which was approximately one-fourth as strong as observed in 1997, and an average spectrum with peak power output at approximately 6 keV. This is a decrease of more than a factor of 15 from the 100 keV peak seen in 1997. During these observations, a very rapid flare was observed in which the location of peak power output increased from $\\lesssim$3 keV to $gtrsim$30 keV. This large shift in peak power output energy is similar to the behavior of Mrk 501 in 1997 and in 1998 June. The evolution of the hardness ratio of the flare is consistent with the flare development being dominated by the acceleration process but the lack of simultaneous multi-wavelength observations prohibits further detailed testing of emission models. Though rapid variations have been seen from Mrk 501 at other wavelengths before, it was generally regarded as a more slowly varying object than the other TeV sources, Mrk 421 and PKS 2155-304. Thus, one could conduct less dense observations of Mrk 501 and still sample the variations with adequate coverage to resolve the shape of the variations and the correlations between wavelengths. These observations show that very dense multi-wavelength observations are required for Mrk 501 as well since it can vary on time-scales comparable to the fastest seen in the other TeV sources. They also indicate that, as has been seen in Mrk 421 \\citep{Gaidos96}, these very rapid flares can occur when the source is not in a particularly high emission state, so dense observations must be used regardless of the flux level observed from these objects. Multi-wavelength observations of such rapid flares will provide stringent tests of emission models for these TeV sources and may lead to a better understanding of the acceleration process that occurs in their jets." }, "0003/astro-ph0003365_arXiv.txt": { "abstract": "Cold dark matter (CDM) models predict small-scale structure in excess of observations of the cores and abundance of dwarf galaxies. These problems might be solved, and the virtues of CDM models retained, even without postulating {\\it ad hoc} dark matter particle or field interactions, if the dark matter is composed of ultra-light scalar particles ($m\\sim 10^{-22}$eV), initially in a (cold) Bose-Einstein condensate, similar to axion dark matter models. The wave properties of the dark matter stabilize gravitational collapse providing halo cores and sharply suppressing small-scale linear power. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003190_arXiv.txt": { "abstract": "The energy spectrum and primary composition of cosmic rays with energy between $3\\times 10^{14}$ and $3\\times10^{16}\\unit{eV}$ have been studied using the CASA-BLANCA detector. CASA consisted of 957 surface scintillation stations; BLANCA consisted of 144 angle-integrating Cherenkov light detectors located at the same site. CASA measured the charged particle distribution of air showers, while BLANCA measured the lateral distribution of Cherenkov light. The data are interpreted using the predictions of the CORSIKA air shower simulation coupled with four different hadronic interaction codes. The differential flux of cosmic rays measured by BLANCA exhibits a knee in the range of 2--3\\,PeV with a width of approximately 0.5 decades in primary energy. The power law indices of the differential flux below and above the knee are $-2.72\\pm0.02$ and $ -2.95\\pm0.02$, respectively. We present our data both as a mean depth of shower maximum and as a mean nuclear mass. A multi-component fit using four elemental species suggests the same composition trends exhibited by the mean quantities, and also indicates that QGSJET and VENUS are the preferred hadronic interaction models. We find that an initially mixed composition turns lighter between 1 and 3\\,PeV, and then becomes heavier with increasing energy above 3\\,PeV\\@. ", "introduction": "\\label{sec.intro} The all-particle energy spectrum of cosmic rays can be described by a steeply falling power law over many decades of energy. The smoothness of this drop in intensity with energy is broken by a change in index of the power law just above $10^{15}\\unit{eV}$. While the origin of this break (referred to as the ``knee'') is not yet fully understood, the prevailing theoretical models describe the knee as a result of the energy limit for particle acceleration in supernova shocks~\\cite{Cesarsky,Drury94}. Further, these models predict that the composition of the primary cosmic rays should change from proton (or ``light''-nuclei) dominated to iron (or ``heavy''-nuclei) dominated as energy increases through the region of the knee. Measuring a composition trend from light to heavy would lend support to the supernova shock acceleration picture. Determining the composition of cosmic rays with energies greater than $10^{15}\\unit{eV}$ is a notoriously difficult problem. To detect primary cosmic rays above this energy directly by satellite experiments requires an unacceptable launch payload volume. Similarly, stratospheric balloon-borne experiments are limited by volume and flight time in their collection of primary particles. Thus, to investigate the composition of cosmic rays at the knee, we must rely on ground-based detection of air showers generated by the primary cosmic rays. The Cherenkov light emission from the charged particle component of an air shower provides an integrated measurement of the longitudinal development~\\cite{Brennan58,Chudakov60}. One approach is to sample the Cherenkov lateral distribution, the photon density as a function of distance from the air shower core. The Cherenkov intensity is proportional to the primary energy, while the slope of the lateral distribution is related to the depth of maximum shower development --- and hence to the mass of the primary cosmic ray nucleus. Therefore measuring a large number of Cherenkov lateral distributions can provide information on how the composition changes with energy~\\cite{patterson_hillas}. Previous attempts to exploit this fact at the knee include~\\cite{Dawson89,airobicc,vulcan}. Optical photons suffer little absorption as they travel through the atmosphere. This means that the Cherenkov lateral distribution is much broader than that of charged particles. Additionally their numerical density is much higher. Thus it is possible to make high signal-to-noise measurements of Cherenkov lateral distributions using an array of detectors with smaller area and wider spacing than would be required for equivalent measurements of charged particles. To obtain high quality Cherenkov lateral distribution data the Broad Lateral Non-imaging Cherenkov Array (BLANCA) was built at the Chicago Air Shower Array (CASA) installation in Dugway, Utah. Using CASA as the cosmic ray trigger, BLANCA operated on clear, moonless nights in 1997 and 1998. In the following analysis we use CASA to find the shower core position and arrival direction and BLANCA to make a precision measurement of the Cherenkov lateral distribution. This paper details the results obtained through these measurements on the energy spectrum and composition of cosmic rays in the energy range between $3\\times 10^{14}$ and $3\\times10^{16}\\unit{eV}$. ", "conclusions": "\\label{sec.conclusions} The CASA-BLANCA experiment has studied cosmic rays in the energy range 0.3--30\\,PeV\\@. The primary energy and mass are found by measuring the Cherenkov lateral distribution for each air shower. In an effort to understand how results depend on the unknown physics of high energy nuclear interactions, we have interpreted the data using the CORSIKA air shower Monte Carlo program with four different hadronic interaction models: QGSJET, VENUS, SIBYLL, and HDPM. The BLANCA energy spectrum agrees well with previous measurements and exhibits a smooth knee near 2--3\\,PeV in primary energy. The model dependence of the energy scale is less than the absolute calibration uncertainty. \\begin{figure} \\bigfig{xmax_all.eps} \\caption {The CASA-BLANCA measurement of \\Xmax\\ compared with other results. All experiments operating near the knee use atmospheric Cherenkov light, including DICE~\\cite{dice}, the AIROBICC array of HEGRA~\\cite{airobicc}, and the VULCAN array at the South Pole~\\cite{vulcan}. The high energy measurements ($>10^{17}\\unit{eV}$) by Fly's Eye use the atmospheric fluorescence technique~\\cite{flys_eye_result}. The ``direct'' point estimates the mean \\Xmax\\ that would be expected on the basis of direct balloon measurements~\\cite{dice}. The Monte Carlo lines use CORSIKA with QGSJET~\\cite{clem_corsika}.} \\label{fig.xmax_all} \\end{figure} We find the transformation from measured Cherenkov lateral distribution slope to the depth of shower maximum \\Xmax\\ to be essentially model independent. In Figure~\\ref{fig.xmax_all} our results are compared to previous experiments over a wide energy range. The BLANCA data are well within the physically reasonable range bounded by the pure proton and iron curves; furthermore they are consistent at low energy with those expected from direct measurements and at high energy with the Fly's Eye result~\\cite{flys_eye_result}. We have also interpreted our data as a mean nuclear mass. This is essentially equivalent to the \\Xmax\\ analysis but is a quantity of more direct astrophysical interest. It has been a long held goal in the air shower field to choose an adequate hadronic interaction model \\emph{and} determine the nuclear composition of the primary cosmic rays simultaneously. With the advent of the powerful simulation tool provided by the CORSIKA group, and high collecting power arrays such as CASA-BLANCA, it seems that this ambition may be becoming a reality. A multi component fit of the type described in Section~\\ref{sec.multi} is a much more efficient use of the available data than simply considering the mean value and spread of a quantity, and the experimental statistics are starting to justify this approach. The agreement between data and simulation in Figure~\\ref{fig.multi_example} is impressive. On the basis of our data we favor the QGSJET and VENUS models and reject SIBYLL and HDPM. At the same time, both of the ways in which we have analyzed our data indicate that the cosmic ray composition is lighter near 3\\,PeV than it is at either 300\\,TeV or 30\\,PeV. The trend towards heavier primary mass above 3\\,PeV agrees with the canonical model of Galactic production and a rigidity-dependent time for escape, and is not consistent with acceleration at sites such as AGN, which require a pure proton composition well above the knee~\\cite{agn_protheroe}. The trend shown in our data to a lighter composition approaching the knee is puzzling but not without theoretical precedent. Swordy, arguing that there must be a minimum path length in the Galaxy even for the highest energy cosmic rays, predicts a light composition at the knee~\\cite{swordy_model}. We acknowledge the invaluable assistance of the CASA-MIA collaboration, as well as the University of Utah High-Resolution Fly's Eye (HiRes) group and the command and staff of the U.S.\\ Army Dugway Proving Ground. We thank D.\\ Heck and the rest of the CORSIKA team for providing and maintaining their excellent program, and the authors of the hadronic interaction models to which it is linked. We thank C.\\ Cassidy, J.\\ Jacobs, J.\\ Meyer, M.\\ Pritchard, and K.\\ Riley for helping with BLANCA's construction and K.\\ Anderson and C.\\ Eberhardy for calibration work. We especially wish to thank M.\\ Cassidy for his essential contributions as our technician. JF and CP acknowledge fellowships from the William Grainger Foundation and the Robert R.\\ McCormick Foundation, respectively. This work was supported by the U.S.\\ National Science Foundation. We would also like to thank S.\\ Swordy for useful conversations." }, "0003/astro-ph0003473_arXiv.txt": { "abstract": "We use the results from recent computations of updated non-linear convective pulsating models to constrain the distance modulus of Galactic globular clusters through the observed periods of first overtone ($RR_c$) pulsators. The resulting relation between the mean absolute magnitude of RR Lyrae stars $$ and the heavy element content $[Fe/H]$ appears well in the range of several previous empirical calibrations, but with a non linear dependence on $[Fe/H]$ so that the slope of the relation increases when moving towards larger metallicities. On this ground, our results suggest that metal-poor ($[Fe/H]<$-1.5) and metal-rich ($[Fe/H]>$-1.5) variables follow two different linear $-[Fe/H]$ relations. Application to RR Lyrae stars in the metal-poor globular clusters of the Large Magellanic Cloud provides a LMC distance modulus of the order of 18.6 mag, thus supporting the ``long\" distance scale. The comparison with recent predictions based on updated stellar evolution theory is shortly presented and discussed. ", "introduction": "The intrinsic luminosity of RR Lyrae variables has been for a long time a very popular way to give reasonable estimates of the distance to globular clusters (GCs) both in the Milky Way and in Local Group galaxies (Magellanic Clouds, M31) and, in turn, to constrain the age of these very old stellar systems. However, notwithstanding the large body of work, a general consensus on a precise evaluation of such a luminosity has been not yet achieved. One may notice that a firm knowledge of RR Lyrae luminosities would be of paramount relevance, since it would provide an independent test of the Cepheid distance scale as well as a reliable calibration of several secondary distance indicators (as, e.g., the GC luminosity function or the Red Giant Tip) for external galaxies, thus providing important clues on the value of the Hubble constant $H_0$. On these grounds, RR Lyrae variables could represent relevant milestones on the path to set both a lower and an upper limit to the age of the Universe, playing a fundamental role in several astrophysical problems ranging from stellar evolution to cosmological models. From the observational side, studies dealing with the absolute magnitude $M_V(RR)$ of RR Lyraes and with the dependence of these magnitudes on the heavy element content $[Fe/H]$ has yielded to the well known debate between the so-named \"short\" and \"long\" distance scales. As recently reviewed by Cacciari (1999), empirical estimates of $M_V(RR)$ for RR Lyrae stars at $[Fe/H]$=-1.6 actually range from about 0.4 mag to 0.7 mag, thus leaving an uncertainty of $\\sim\\pm$ 0.2 mag on the derived distance moduli (see also Popowski \\& Gould 1999). Different estimates have been also given for the dependence of these magnitudes on the star metallicity. As a matter of the fact, for the often assumed linear relation $$=a+b[Fe/H]$$ \\noindent one finds in the literature evaluations of the coefficient $\"b\"$ {\\bf mainly in the range} $b\\sim 0.18\\pm 0.03$ to $\\sim$0.30, where the former value is based on the Baade-Wesselink method (see, e.g., Fernley et al. 1998b [Fn98b]) and the latter value has been early suggested by Sandage (1993 [Sa93]) when discussing the period-metallicity relation for field and GC RR Lyrae pulsators. {\\bf However, an even milder slope has been suggested by Fusi Pecci et al. (1996), who investigated eight globular clusters in M31 to derive, over the range -1.8$<{[Fe/H]}<$-0.4, $$=(0.13\\pm0.07)*[Fe/H]+(0.95\\pm0.09)$$}. The recent release of HIPPARCOS statistical and trigonometric parallaxes for halo RR Lyraes ($[Fe/H]\\le$-1.30) has not clarified the issue: one may indeed recall that Fernley et al. 1998a [Fn98a] and Groenewegen \\& Salaris (1999 [GS99]), both assuming the same slope $b$=0.18, give a zero-point of $1.05\\pm0.15$ mag and 0.77$\\pm0.26$ mag, respectively, as derived from an identical sample of variables but using different approaches (statistical parallaxes or reduced parallaxes, respectively). In the meantime, McNamara 1999 [MN99] claims that Baade-Wesselink results for variables with $[Fe/H]>-1.5$ yield a quite different relation as given by $$=1.06+0.32[Fe/H] \\label{[MN99]}$$ On the theoretical side, the literature already contains several sets of horizontal branch (HB) evolutionary models computed for wide ranges of the overall metallicity ($Z$ in the range of 0.0001 to 0.02) which provide the \"theoretical route\" to the calibration of the $M_V(RR)$ versus $[Fe/H]$ relation. One finds that almost all the recent theoretical predictions concerning the absolute magnitude $M_V(ZAHB)$ of the zero age horizontal branch (ZAHB) sequence at the RR Lyrae instability strip confirm the non-linear dependence of $M_V(ZAHB)$ on $\\log Z$ formerly suggested by Castellani, Chieffi \\& Pulone (1991 [CCP]). However, the scaling of the overall metallicity $Z$ to the measured $[Fe/H]$ values could be a tricky matter since the classical assumption of solar-scaled chemical mixtures is likely inappropriate to GC stars. There is indeed a growing observational evidence for a significant enhancement of $\\alpha$-elements with respect to iron ($[\\alpha/Fe] \\sim 0.3$) in GC and field metal-poor stars (see Carney et al. 1997, Gratton et al. 1997). Moreover, one has to bear in mind that observed RR Lyrae samples do contain stars evolved off their original ZAHB position. Thus, realistic predictions on the average magnitude $$ require the evaluation of the evolutionary effects, possibly through synthetic HB simulations (SHB). The wide grids of SHBs so far published (e.g., Lee, Demarque \\& Zinn 1990, Lee 1991, Bencivenni et al. 1991, Caputo et al. 1993) have already shown that the predicted mean magnitude of RR Lyrae stars $$ significantly depends, with everything else being constant, on the HB morphology. Simulations based on slightly modified CCP models yielded Caputo (1997) to suggest $$=1.19+0.19 \\log Z$$ \\noindent for RR Lyrae-rich metal-poor GCs with $\\log Z\\le -3.0$, whereas for larger metallicities the theory gives $$=1.57+0.32 \\log Z$$ \\noindent Similar results have been more recently found by Demarque et al. (1999), who definitively reject the existence of a unique linear relation covering the metallicities spanned by GCs, confirming that the slope of the predicted $M_V(RR)-\\log Z$ relation depends on the metallicity range and that the HB morphology of each cluster must be taken in the due account when using RR Lyrae stars as distance indicators. On these grounds, one is tempted to conclude that theoretical and observational investigations do show a sort of consistency: the former give warnings against a \"universal\" linear $M_V(RR)-[Fe/H]$ relation, the latter fail to reach an agreement on both its slope and zero-point! Within such a confusing scenario, one has to mark the seminal attempts made by Sandage (Sa93 and references therein) to use RR Lyrae periods to constrain the luminosity of these stars. This appears a quite relevant approach, since periods are firm and safe observational parameters, independent of distance and reddening. To discuss Sandage's philosophy, one has to recall that since the pioneering work by Christy (1966) and Stellingwerf (1975, 1984) pulsating models have suggested the existence within the instability strip of a region where both fundamental ($RR_{ab}$) and first overtone ($RR_c$) modes are stable (see Bono \\& Stellingwerf 1994, Bono et al. 1997a, Bono et al. 1997c). The boundaries of this \"either-or\" region, namely the fundamental blue edge (FBE) at the higher temperature side and the first overtone red edge (FORE) at the lower temperature side, encompass for each given luminosity the range of temperatures (or colours) where the mode-shift (i.e. the transition from $RR_{ab}$ to $RR_c$) may occur. Assuming that for both Oosterhoff type I (OoI) and Oosterhoff type II (OoII) globular clusters this transition occurs at the blue edge for fundamental pulsation and using periods and $B-V$ colours of the shortest period $RR_{ab}$ in clusters and in the field, Sa93 derives the star luminosity from the well established period-mass-luminosity-temperature relation. In this way he obtains the relation $$M_V(RR)=0.94+0.30[Fe/H] \\label{[Sa93]}$$ \\noindent which accounts for the Oosterhoff dichotomy in Galactic globular clusters as mainly due to a luminosity effect. However, a re-analysis by Fernley (1993 [Fn93]), using $V-K$ colours and a limited sample of clusters with low and well-known reddening, yields $$M_V(RR)=0.84+0.19[Fe/H] \\label{[Fn93]}$$ More recently the assumption of a unique $RR_{ab}/RR_c$ transition line has been questioned by Bono, Caputo \\& Marconi (1995), who concluded that the Oosterhoff dichotomy is largely the result of different transition lines in OoI (near FBE) and OoII (near FORE) clusters, as early suggested by van Albada and Baker (1973). However, the pulsation theory predicts the limits of the whole instability strip, as given by the first overtone blue edge (FOBE) and the fundamental red edge (FRE), without any ambiguity about the actual pulsation mode. On this basis, Caputo (1997) used a preliminary set of pulsating models to show that theoretical predictions on the pulsator distribution in the period-absolute magnitude $M_V-\\log P$ plane can constrain the distance to RR Lyrae-rich globular clusters. In recent times, the RR Lyrae pulsating models have been updated and extended to wide ranges of mass and chemical composition, shedding light on the dependence of the instability strip on the metal content. In this paper we will take advantage of these improvements to reconsider the Caputo (1997) analysis. The updated $M_V-\\log P$ relations at FOBE and FRE are discussed in the following Sect. 2, while Sect. 3 presents the comparison with observation and the derived \"pulsational\" distance moduli for a selected sample of well-studied GCs. The resulting dependence of our $$ values on the cluster metallicity is discussed in Sec. 4 in comparison with both empirical relations and recent theoretical HB models. Some concluding remarks will close the paper. ", "conclusions": "In this paper we have used results from the most recent and updated computations of non-linear convective pulsating models to constrain the distance modulus of Galactic globular clusters through the observed periods of $RR_c$ pulsators. The resulting $-[Fe/H]$ relation appears in the range of several empirical linear calibrations, but with evidence for a non linearity which suggests that the slope of the relation increases when moving towards the metal-richer variables. On observational grounds, a similar behavior seems present among RR Lyrae stars in $\\omega$ Centauri (Rey et al. 2000). Moreover, we notice that over the range of metal-poor stars ($[Fe/H]<$-1.5) our pulsational calibration is in good agreement with the relations given by Fernley (1993), Groenewegen \\& Salaris (1999) and Carretta et al. (1999), while with $[Fe/H]>$-1.5 it agrees with MacNamara (1999) results. Application of our results to RR Lyrae stars of the metal-poor globular clusters in the Large Magellanic Cloud (see data in GS) would give a distance modulus of 18.61$\\pm$0.12 mag ($f$=1) and 18.56$\\pm$0.12 mag ($f$=3), thus supporting the ``long\" distance scale (see also Romaniello et al. 1999). By relying on the present pulsational RR Lyrae absolute magnitudes, one derives that the non linearity of our $-[Fe/H]$ relation is well reproduced by current predictions based on stellar evolution theory. However, in the case of solar-scaled chemical compositions, none of the evolutionary predictions published in the recent literature appears in satisfactory agreement, supporting observational evidence for $\\alpha$-enhanced chemical mixtures in metal-poor stars. With the $\\alpha$-elements enhanced by a factor of 3 with respect to iron, the predictions by CS and Fr99 agree with our pulsational magnitudes even though with a tendency of overestimating the luminosity of metal-rich pulsators. Interesting enough, one finds that the Cs99 relation is well reproducing the general dependence of $M_V(RR)$ on log$Z$, but with an overluminosity of about 0.08 mag. Holding CS99 results, a beautiful agreement with our data would be achieved by sistematically increasing the cluster metallicity by $\\sim$ 0.2 dex, an occurrence hardly to be accepted. To further discuss this point, one has to remind that differences in stellar models are mainly, if not only, the result of differences in the adopted input physics. Discussing RR Lyrae stars in the globular cluster M5 (Caputo et al. 1999) we have already reported pulsational evidence suggesting that models with the \"most updated\" input physics (as in Cs99) give too luminous HB stars. Such an evidence has been further supported by independent estimates based on HIPPARCOS parallaxes for clumping field He burning stars (Castellani et al. 1999). Data in the previous Fig. 4 reinforce such an evidence, suggesting that the \"most updated physics\" is probably far from being the most adequate one. As a whole, we remain with the tantalising evidence that Cs99 models give the rightest metal dependence but not right luminosities, whereas those by Fr99 and CS give much better luminosities but slightly worst slope. The role played by the various physical ingredients in determining the predicted luminosity of HB structures has been recently discussed in several papers (see Cassisi et al. 1999, Castellani and Degl'Innocenti 1999, Castellani 1999) and cannot be repeated here. However, one may notice that the most recent theoretical predictions displayed in Fig. 4 all agree within a range of luminosity of about $\\pm$ 0.05 mag. This in our feelings should be regarded as an evidence of the high standard reached by evolutionary theories, as well as a warning that better precision should require a corresponding level of accuracy in the input physics not yet reached by currently available evaluations. {\\bf Acknowledgment:} It is a pleasure thank the referee, B. Carney, for his valuable report. We deeply thank B. Carney and M. Corwin for providing us with data on NGC 5466 before publication. Thanks are also due to Santi Cassisi for several warm discussions with one of us (F.C.) during an icy week in Teramo. Financial support for this work was provided by the italian Ministero dell'Universit\\`a e della Ricerca Scientifica e Tecnologica (MURST) under the scientific project ``Stellar Evolution''." }, "0003/astro-ph0003159_arXiv.txt": { "abstract": "One of remarkable features of the gamma ray blazar Markarian 501 is the reported shape of the TeV spectrum, which during strong flares of the source remains essentially stable despite dramatic variations of the absolute $\\gamma$-ray flux. I argue that this unusual behavior of the source could be explained assuming that the TeV emission is a result of synchrotron radiation of extremely high energy ($E \\geq 10^{19} \\, \\rm eV$) protons in highly magnetized ($B \\sim 30-100 \\, \\rm G$) compact regions of the jet with typical size $R \\sim 10^{15} - 10^{16} \\, \\rm cm$ and Doppler factor $\\delta_{\\rm j} \\simeq 10-30$. It is shown that if protons are accelerated at the maximum possible rate, i.e. $t_{\\rm acc}=\\eta (r_{\\rm g}/c)$ with so-called gyro-factor $\\eta \\sim 1$, the synchrotron cooling of protons could not only dominate over other radiative and non-radiative losses, but could also provide good fits (within uncertainties introduced by extragalactic $\\gamma$-ray extinction) to the $\\gamma$-radiation of two firmly established TeV blazars - Markarian 501 and Markarian 421. Remarkably, if the proton acceleration takes place in the regime dominated by synchrotron losses, the spectral shape of the Doppler-boosted $\\gamma$-radiation in the observer's frame is determined essentially by the self-regulated ``synchrotron cutoff'' at $\\epsilon_0 \\simeq 0.3 \\ \\delta_{\\rm j} \\eta^{-1} \\ \\rm TeV$. The hypothesis of the proton-synchrotron origin of TeV flares of BL Lac objects inevitably implies that the energy contained in the form of magnetic field in the $\\gamma$-ray emitting region exceeds the kinetic energy of accelerated protons. ", "introduction": "Blazars are Active Galactic Nuclei (AGN) dominated by a highly variable component of non-thermal radiation produced in relativistic jets close to the line of sight (e.g. Begelman et al. 1984, Urry \\& Padovani 1995). The dramatically enhanced fluxes of the Doppler-boosted radiation, coupled with the fortuitous orientation of the jets towards the observer, make these objects ideal laboratories to reveal the underlying physics of AGN jets through multi-wavelength studies of temporal and spectral characteristics of radiation from radio to very high energy $\\gamma$-rays (Ulrich et al. 1997) . First of all this concerns the BL Lacertae (BL Lac) objects - a sub-population of blazars of which two prominent representatives, Markarian 421 and Markarian 501, are firmly established as TeV $\\gamma$-ray emitters. The flux variability on different time-scales (and, plausibly, of different origin) is a remarkable feature of TeV radiation of BL Lac objects. It ranges from the spectacular 1996 May 15 flare of Markarian 421 with duration less than 1 h to the extraordinary high state of Markarian 501 in 1997 lasting several months (for review see Aharonian 1999; Catanese \\& Weekes 1999). The recent multi-wavelength campaigns revealed that the TeV flares of both objects, Markarian 501 (Pian et al. 1998; Catanese et al. 1997; Krawczynski et al. 2000; Sambruna et al. 2000) and Markarian 421 (Buckley et al. 1996; Maraschi et al. 1999; Takahashi et al. 1999), correlate with X-radiation on time-scales of hours or less. This is often interpreted as a strong argument in favor of the so-called synchrotron-Compton jet emission models in which the same population of ultra-relativistic electrons is responsible for production of both X-rays and TeV $\\gamma$-rays via synchrotron radiation and inverse Compton scattering, respectively (see, e.g., Ulrich et al. 1997). However, the very fact of correlation between X-ray and TeV $\\gamma$-ray fluxes does not yet rule out other possibilities, in particular the so-called hadronic models which assume that the observed $\\gamma$-ray emission is initiated by accelerated protons interacting with ambient gas or low-frequency radiation. Generally, the hadronic models do not offer efficient $\\gamma$-ray production mechanisms in the jet. For example, for any reasonable acceleration power of protons, $L_{\\rm p} \\leq 10^{45} \\, \\rm erg/s$, the density of the thermal plasma in the jet should exceed $10^{6} \\, \\rm cm^{-3}$ in order to interpret the reported TeV flares of Markarian 501 by $\\pi^0$-decay $\\gamma$-rays produced at $p$-$p$ interactions. Therefore this mechanism could be effectively realized only in a scenario like ``relativistic jet meets target'' (Morrison et al. 1984), i.e. assuming that $\\gamma$-radiation is produced in dense gas clouds that move across the jet (e.g. Dar \\& Laor 1997) Recently a novel, ``non-acceleration'' version of $\\pi^0$-decay $\\gamma$-ray production by blazar jets was suggested by Pohl \\& Schlickeiser (2000). The {\\em Proton Induced Cascade} (PIC) model (Mannheim 1993; Mannheim 1996) is another attractive possibility for production of high energy $\\gamma$-rays. This model relates the observed $\\gamma$-radiation to the development of pair cascades in the jet triggered by secondary ``photo-meson'' products ($\\gamma$-rays and electrons) produced at interactions of accelerated protons with low-frequency synchrotron radiation. The efficiency of this model significantly increases with energy of accelerated protons, therefore the postulation of an existence of extremely high energy (EHE; $E \\geq 10^{19} \\, \\rm eV$) protons is a key assumption for the PIC model. In a compact $\\gamma$ production region of the jet with characteristic size less than $10^{16} \\, \\rm cm$, the protons could be accelerated to such high energies only in the presence of large magnetic field, $B \\gg 1 \\, \\rm G$. Even so, below I will show that the very fact of observations of multi-TeV $\\gamma$-rays from Markarian~421 and Markarian~501 allow a rather robust lower limit on the ``photo-meson'' cooling time of protons, $t_{\\rm p \\gamma} \\geq 10^7 \\, \\rm s$. This implies uncomfortably high luminosity in EHE protons which would be required to match the observed TeV $\\gamma$-ray fluxes. Meanwhile, at such conditions the synchrotron radiation of the EHE protons becomes a very effective channel of production of high energy $\\gamma$-rays. In this paper I show that for a reasonable set of parameters, which characterize the small-scale (sub-parsec) jets in Markarian 421 and Markarian 501, the synchrotron radiation of EHE protons not only may dominate over other possible radiative and non-radiative losses, but also could provide adequate fits to the observed TeV spectra of both objects\\footnote{The synchrotron radiation of protons was actually included in the overall PIC code of Mannheim (1993), but the effect of this process was somehow disregarded. The importance of the proton synchrotron radiation in the PIC scenario was recently recognized, independent of the present paper, by M\\\"ucke \\& Protheroe (2000).}. Moreover, this hypothesis could naturally explain one of the remarkable features of TeV flares of Markarian 501 - its essentially stable spectral shape despite spectacular variations of the absolute TeV flux up to factor of 10 or more on time-scales less than 1 day. ", "conclusions": "Although there is little doubt that the bulk of the highly variable X-ray emission of the BL Lac objects has synchrotron origin and is produced in the relativistic jets pointed to the observer (Urry \\& Padovani 1995), the TeV radiation of Markarian 421 and Markarian 501 is the only model-independent and unambiguous indicator of acceleration of ultra-relativistic particles in these small-scale (sub-pc) jets. In the currently popular synchrotron-self-Compton (SSC) models of BL Lac objects both the X-ray and $\\gamma$-ray components are attributed to the radiation of directly accelerated TeV electrons. For a certain, physically well justified combination of parameters (the blob size, magnetic field, Doppler factor, etc.), these models (see, e.g., Inoue \\& Takahara 1996, Mastichiadis \\& Kirk 1997, Pian et al. 1998, Bednarek \\& Protheroe 1999, Coppi \\& Aharonian 1999a, Wagner et al. 1999; Kataoka et al. 1999, Maraschi et al. 1999, Takahashi et al. 1999, Krawczynski et al. 2000) give quite satisfactory explanation of the observed spectral and temporal characteristics of non-thermal radiation over more than ten decades of frequencies from $10^{16} \\, \\rm Hz$ to $10^{27} \\, \\rm Hz$. In the SSC models the strength of the magnetic field in the $\\gamma$-ray emitting regions typically is less than $1 \\, \\rm G$, the most likely value being close to $0.1 \\, \\rm G$, which directly follows from the comparable energy fluxes released in X-rays and $\\gamma$-rays during the TeV flares of Markarian 421 and Markarian 501 (e.g. Tavecchio et al. 1998). At such a low magnetic field the maximum energy of accelerated protons cannot exceed $E_{\\rm p, max} \\sim 10^{18} \\Delta t_{\\rm 3h} \\delta_{10} \\ (B/1 \\, \\rm G) \\, \\eta^{-1} \\rm eV$, and therefore the contribution of protons in $\\gamma$-ray production through both the photo-meson and the synchrotron channels is negligible; the production of TeV $\\gamma$-rays in the jet is strongly dominated by the inverse Compton scattering of electrons directly accelerated up to energies $\\sim 10 \\, \\rm TeV$. The increase of the magnetic field leads to reduction ($\\propto B^{-2}$) of the flux of the inverse Compton $\\gamma$-rays; for $B \\geq 1 \\, \\rm G$ the bulk of energy of accelerated electrons is channeled into the synchrotron radiation. Also, the increase of the magnetic field shifts ($\\propto B$) the synchrotron peak to higher energies. But for any reasonable Doppler factor of the jet, the observed TeV fluxes of Markarian 421 and Markarian 501 cannot be explained by the electron synchrotron radiation because of the self-regulated synchrotron cutoff at $\\epsilon_{\\rm max} \\simeq 1.6 \\ \\delta_{10} \\, \\rm GeV$, which inevitably appears if the electron acceleration and $\\gamma$-ray production take place in the same region of the jet. Formally, we may avoid this limit assuming that the regions of the electron acceleration and the $\\gamma$-ray production are separated, i.e. the electrons are accelerated up to energies $\\sim 10^3 \\, \\rm TeV$ in a region with rather small magnetic field, $B \\leq 0.01 \\, \\rm G$, but release all their energy in a form of GeV/TeV synchrotron radiation later, after entering the region(s) of strongly compressed/amplified magnetic field, $B \\geq 10 \\, \\rm G$. This scenario seems, however, rather artificial. Large magnetic fields, $B \\sim 100 \\, \\rm G$ or so, coupled with effective acceleration of protons at the maximum rate, may create very favorable conditions for TeV $\\gamma$-ray production. Indeed, at such conditions the synchrotron cooling of protons not only well dominates over other radiative and non-radiative losses (see Fig.~4), but also provides good fits to the observed spectra of TeV radiation of Markarian 421 and Markarian 501. Remarkably, if the proton acceleration takes place in the {\\em synchrotron-loss-dominated} regime, the spectral shape of radiation depends only on the power-law index of accelerated protons, $\\alpha_{\\rm p}$ and the parameter $\\rho=\\delta_{10}/\\eta$, but not on the magnetic field and the size of the $\\gamma$-ray emitting blobs, which generally endure significant time-evolution. Meanwhile, any change in the size and/or magnetic field of the evolving blobs should lead to significant variation in the absolute flux of $\\gamma$-rays. This effect could give a natural explanation for one of the remarkable features of strong flares of Markarian 501 (and, perhaps, also Markarian 421) - the essentially stable spectral shape of TeV radiation despite strong variation of the absolute flux observed on time scales less than 1 day. During the strong flares of Markarian 501 with a flat spectrum around 1 TeV, the parameter $\\rho=\\delta_{10}/\\eta \\sim 1$. The exact value of $\\rho$ depends on the level of the diffuse extragalactic background and on the power-law index of accelerated protons. The uncertainty in this parameter, despite significant uncertainties in the intergalactic extinction of TeV radiation, as well as the lack of adequate information about the $\\gamma$-ray fluxes at MeV/GeV energies, does not exceed a factor of three or so. Since the Doppler factor in Markarian 501 is believed to be within 10 to 30, the acceleration rate should be very high, $\\eta=\\rho^{-1} \\ \\delta_{10} \\sim 1-3$. The {\\em unusually high} GeV flux observed from Markarian 501 during the multiwavelength campaign in March 1996, when the source was in a {\\em very low} TeV state, could be explained by a transition of the source from the {\\em synchrotron-loss-dominated} regime to the {\\em escape-loss-dominated} regime, caused, for example, by dramatic drop of the magnetic field. In this regime we should expect steeper TeV $\\gamma$-ray spectra with a slope depending on the magnetic field and the size of the source. Therefore, an important test of this hypothesis could be detection of spectral variability of $\\gamma$-radiation in a low state of the source. The significantly steeper TeV spectrum of Markarian 421 requires early synchrotron cutoff, $\\epsilon_0 \\sim 0.25 - 1 \\, \\rm TeV$. This could be interpreted as a result of formation of the synchrotron spectrum in the regime dominated by the escape of protons from the acceleration region. For a magnetic field of about 100~G this would require a compact source with $R < 10^{15} \\, \\rm cm$. Interestingly, such a small linear size of the $\\gamma$-ray production region is supported independently by the observed dramatic variations of the TeV flux of Markarian 421 on timescales $\\sim 15$ minutes. Within the model of proton synchrotron radiation, we may expect also synchrotron radiation produced by {\\em directly accelerated} electrons - the counterparts of EHE protons. If the particle acceleration takes place in the synchrotron-loss-dominated regime, the self-regulated synchrotron cutoff of this component depends only on the parameter $\\rho=\\delta_{10}/\\eta$, namely $\\epsilon_0 \\simeq 1.6 \\rho \\, \\rm GeV$, thus it correlates with the position of the cutoff in the proton synchrotron spectrum at $ \\epsilon_0 \\simeq 3 \\rho \\, \\rm TeV$. The ratio of the energy fluxes of these two components is determined simply by the ratio of non-thermal energy channeled into the accelerated protons and electrons, $\\dot{W_{\\rm p}}/\\dot{W_{\\rm e}}$. In a strong magnetic field of about 100 G, the synchrotron cooling time of electrons is shorter, almost at all relativistic energies, than the typical dynamical (e.g. light-crossing) times. This results in a well-established steady-state spectrum of electrons ${\\rm d}N/{\\rm d}E \\propto E^{-(\\alpha_0+1)}$, provided that the power-law index of acceleration spectrum $\\alpha_0 \\geq 1$. Consequently, a pure power law synchrotron spectrum with a photon index $(\\alpha_0+2)/2$ would be formed. In particular, for $\\alpha_0=2$, we should expect a flat synchrotron SED ($\\nu S_\\nu=const$) from the optical/UV wavelengths to MeV/GeV $\\gamma$-rays. The broad-band spectra of both Markarian 421 and Markarian 501 do not agree with such a pure power-law behavior; in fact, the SED of both objects show pronounced synchrotron X-ray peaks. Therefore this (theoretically possible) population of directly accelerated electrons - counterparts of the EHE protons - cannot be responsible for the bulk of X-ray emission. The latter could be referred to electrons, produced, most probably, in a different way, and/or in other region(s) of the jet. The X-ray light curves of both Markarian~421 (e.g. Takahashi et al. 1999) and Markarian~501 (e.g. Sambruna et al. 2000) show so-called ``soft'' and ``hard'' lags. A possible interpretation of this effect in terms of competing acceleration, radiative cooling, and escape timescales of synchrotron-emitting electrons (Takahashi et al. 1996; Kirk et al. 1998), would require that all these timescales are comparable with the light crossing time, $t=R/c \\sim 10^{4} - 10^{5}$ s. The cooling time of an electron responsible for a synchrotron photon of energy $\\epsilon$ is $t^{\\rm (e)}_{\\rm sy} \\simeq 1.5 \\times 10^3 (B/1 \\, \\rm G)^{-3/2} \\ (\\epsilon/1 \\, \\rm keV)^{1/2} \\, \\rm s$. Therefore in the X-ray production region the magnetic field cannot, independent of specific model assumptions, significantly exceed $0.1 \\, \\rm G$. Thus, the hypothesis of proton-synchrotron origin of TeV radiation implies that the production regions of TeV $\\gamma$-rays ($B \\sim 100 \\, \\rm G$) and and synchrotron X-rays ($B \\sim 0.1 \\, \\rm G$) are essentially different. Although in any reasonable scenario we may expect a non-negligible correlation between the X-ray and TeV $\\gamma$-ray fluxes, the spatial separation of the X-ray and TeV $\\gamma$-ray production regions does not allow definite predictions for such a correlation. In the proton-synchrotron model of TeV radiation of BL Lac objects, two more components of X-radiation are expected. Firstly, in the field of about 100~G the accelerated protons of energy $E \\sim 10^{15} \\, \\rm eV$ {\\em themselves} produce synchrotron X-rays. However, the contribution of this component to the observed X-ray flux is negligible. A much more prolific channel for X-ray production connected (indirectly) with the EHE protons, can be provided by electrons of non-acceleration origin, namely by secondary electrons produced at interactions of the primary TeV $\\gamma$-rays with the ambient low-frequency radiation. Actually, for a spectrum of EHE protons containing sharp spectral pile-up, we {\\em must} assume an essential ($\\tau_{\\gamma \\gamma} \\sim 1$) internal absorption of $\\gamma$-rays in order to match the observed spectrum of Markarian 501 (see Fig.~10). The appearance of secondary electrons in the jet results in production of a hard synchrotron X-ray component. Apparently, for an optical depth $\\tau_{\\gamma \\gamma} \\sim 1$, the luminosity of this component would be comparable to to the luminosity of their ``grandparents'' - TeV $\\gamma$-rays. The photo-produced electrons have a rather specific, significantly different from the directly accelerated particles, shape. For example, the spectrum of electrons produced at interactions of high energy $\\gamma$-rays with a photon index $\\Gamma$ and field photons with a narrow (e.g. Planckian) spectral distribution with a characteristic energy $\\overline{h \\nu}$, has the following characteristic form: starting from the minimum (allowed by kinematics) energy at $E_{\\ast}=m_{\\rm e}^2 c^4/4 \\overline{h \\nu}$, the electron spectrum sharply rises reaching the maximum at $E_{\\rm m} \\simeq 2.4 E_{\\ast} \\simeq 0.15 (\\overline{h \\nu}/1 \\, \\rm eV)^{-1} \\, \\ \\rm TeV$, and then at $E \\gg E_{\\rm m}$ it decreases as $q_{\\pm} \\propto E^{-(\\Gamma+1)} \\ln E$. Within an accuracy better than 20 per cent, the spectrum of secondary pairs could be approximated in a simple analytical form (Aharonian \\& Atoyan 1991): \\begin{equation} q_{\\pm}(E)\\ {\\rm d}E=f(\\Gamma)\\frac{\\exp{[-(1/(x-1)]}} {E_{\\ast} x (1+0.07 \\ x^{\\Gamma}/\\ln x)} \\ {\\rm d} E \\, , \\end{equation} where $x=E/E_{\\ast} \\geq 1$, $f(\\Gamma)=(1.11-1.60 \\Gamma + 1.17 \\Gamma^2)$, and $\\int q_{\\pm}(E)\\ {\\rm d}E=2$ (two electrons per interaction). The intensive synchrotron losses in the strong magnetic field $B \\sim 100 \\, \\rm G$ quickly establish a steady-state spectrum of electrons proportional to $E^{-2}$ below $E_{\\rm m}$, and approximately as $E^{-(\\Gamma+2)}$ (if we ignore the weak logarithmic term) above $E_{\\rm m}$. Correspondingly, the synchrotron spectrum of the secondary pair-produced electrons is characterized by a smooth transition, through the energy around $\\epsilon_{\\rm b}=120 B_{100} (\\overline{h \\nu}/1 \\, \\rm eV)^{-2} \\, \\rm keV$, from $s=1.5$ to $s \\approx (\\Gamma+3)/2$. Assuming, for example, that the TeV $\\gamma$-rays with a photon index $\\Gamma \\simeq 1.4$ are absorbed in an external infrared photon field with characteristic energy in the jet frame $\\overline{h \\nu}=1.5 \\delta_{10} \\, \\rm eV$ (like in Fig.~10), one should expect {\\em in the observer frame} a hard X-ray component of radiation with a SED $\\nu S_\\nu \\propto \\epsilon^{0.5}$ and $\\nu S_\\nu \\propto \\epsilon^{-0.2}$ below and above $\\epsilon_{\\rm b} \\approx 500 B_{100} \\ \\delta_{10}^{-1} \\, \\rm keV$, respectively. Finally, at energies above several MeV the spectrum becomes very steep due to the cutoff in the spectrum of primary (proton-synchrotron) $\\gamma$-rays. We may speculate that such a component of hard X-rays was observed during the strong TeV flares of Markarian 501, in particular in April 1997 (Pian et al. 1998). Obviously, this component of radiation should strongly correlate with the flux of TeV emission. However, because the synchrotron radiation produced by directly accelerated electrons significantly contributes to the observed X-ray flux as well, the TeV/X-ray correlation may have a rather complicated and non-standard behavior. The absence (or suppression) of such a hard X-ray component in the spectrum of Markarian 421 could be explained, within the framework of this model, by the steep spectrum of TeV $\\gamma$-rays caused by an early synchrotron cut-off. The quantitative study of this question within a detailed time-dependent treatment of the problem will be discussed elsewhere. In BL Lac objects like Markarian 501, the photo-meson processes do not play, most probably, dominant role in the production of high energy $\\gamma$-rays. Indeed, the TeV $\\gamma$-ray transparency condition puts a robust lower limit on the characteristic time of this process, $t_{\\rm p \\gamma} \\sim 10^{7} \\, \\rm s$ which is almost 3 orders of magnitude larger than the characteristic synchrotron cooling time of protons. The severe synchrotron losses of protons cannot be avoided since the requirement of extremely high energy protons in the blob - a {\\em key} assumption in the PIC model - automatically implies a large magnetic field of an order of 100 G. On the other hand, the characteristic photo-meson cooling time cannot be arbitrary reduced, e.g. by assuming extremely high density of radiation in the blob. For any reasonable spectrum of the latter, and any reasonable geometry of the $\\gamma$-ray production region, this would lead to an unacceptably large optical depth for TeV $\\gamma$-rays. Therefore, one may conclude that the very fact of observations of TeV $\\gamma$-rays from Markarian 501 and Markarian 421 (in fact, from {\\em any} TeV blazar) almost rules out the ``photomeson'' origin of the bulk of the observed TeV radiation. This also implies very low TeV neutrino fluxes. The hypothesis of proton-synchrotron origin of TeV $\\gamma$-flares from Markarian 421 and Markarian 501 requires a very large amount of energy contained in the form of magnetic field \\begin{equation} W_{\\rm B}\\approx \\frac{1}{6} R^3 \\ B^2 = 5.6 \\times 10^{49} \\ \\Delta t_{\\rm 3h}^3 \\delta_{10}^3 \\ B_{100}^2 \\ \\rm erg \\, . \\end{equation} For comparison, the energy of the magnetic field allowed by the SSC models ($B \\sim 0.1 \\, \\rm G$) is less by six orders of magnitude. On the other hand, the estimates of the kinetic energy in $\\geq 10^{19} \\, \\rm eV$ protons (within the proton-synchrotron radiation model) and $\\geq 10^{12} \\, \\rm eV$ electrons (within the SSC model) are comparable, if we assume that in both models the $\\gamma$-rays are produced with high efficiency. Indeed, in this case the inverse Compton cooling time of electrons and the synchrotron cooling time of protons responsible for TeV emission in the SSC and proton-synchrotron models, respectively, are equal or less than the observed variability time-scale $t^{\\ast} \\simeq 1.08 \\times 10^5 \\ \\Delta t_{\\rm 3h} \\delta_{10} \\ \\rm s$, thus \\begin{equation} W^{\\rm (SSC)}_{\\rm e} \\sim W^{\\rm (PSR)}_{\\rm p} \\leq 4 \\pi d^2 \\ f_\\gamma \\delta_{\\rm j}^{-4} \\ t^{\\ast} \\simeq 1.4 \\times 10^{46} \\ \\Delta t_{\\rm 3h} \\delta_{10}^{-3} \\ \\rm erg \\, . \\end{equation} In this estimate the average flux of Markarian 501 in 1997, corrected for extragalactic extinction for ``low-CBR'' model is assumed. Higher CBR fluxes, as well as the possible internal absorption of $\\gamma$-rays, would increase this estimate by a factor of 3 or so. If the spectrum of accelerated particles extends, e.g. as $E^{-2}$, down to energies $E \\sim m c^2$, the above estimate could be increased by an order of magnitude. In fact, this simplified estimate only reflects the average level of the energy content of accelerated particles during the high state of Markarian 501 in 1997. For the strongest flares with duration $\\leq 1 \\, \\rm day$, the energy in accelerated protons could be several times larger than it follows from Eq.~(25). But in any case, these uncertainties do not prevent us to conclude that in the proton-synchrotron model we deal with a highly magnetized condensation of $\\gamma$-ray emitting clouds of EHE protons, where the magnetic pressure dominates over the pressure of relativistic protons\\footnote{The minimum energy budget condition is achieved in the case of approximate equipartition between the protons and magnetic field, $ W_{\\rm B} \\approx W_{\\rm p}$. However, for Markarian 501 this condition would imply magnetic field of about $10 \\, \\rm G$, which only marginally could provide effective production of the proton-synchrotron radiation in the TeV regime.}. The total magnetic energy given by Eq.~(24) of a single blob formally is sufficient to supply the necessary energy in TeV $\\gamma$-rays released during the entire high activity state of Markarian 501 in 1997. Thus speculating that an effective acceleration mechanism operates in such magnetized condensations, stimulated presumably by interactions of the latter with the surrounding plasma, a single (or a few) energetic blob(s) ejected from the central source towards the observer could, in principle, explain the extraordinary high TeV flux of Markarian 501 in 1997. In the SSC models the situation is exactly opposite - the pressure of relativistic particles is significantly larger than the pressure of the magnetic field. Even so, the total energy contained in the relativistic electrons is sufficient to support the observed X-ray and $\\gamma$-ray emission during only several hours, therefore this model requires quasi-continuous injection of energy into a single blob, or, most likely, a ``multi-blob'' scenario in which the observed radiation is a result of superposition of many short-live blobs continuously ejected from the central source. Then, the high state of the source like the extraordinary long outburst of Markarian 501 in 1997 could be associated with a dramatic increase of the rate of ejection/formation of $\\gamma$-ray emitting blobs. \\ack I thank the anonymous referee for her/his valuable comments, as well as S. Wagner, M. Sikora, M. Ostrowski, H. Krawczynski, P. Coppi, A. Atoyan, L. Drury, and S. G. Rowell for fruitful discussions." }, "0003/astro-ph0003229_arXiv.txt": { "abstract": "We present an analysis of the full set of {\\it Rossi X-Ray Timing Explorer} (\\rxte) observations of \\ksxrb\\ made between 1996~August and 1999~February, concentrating on the nine type~I X-ray bursts that were observed. We find that the bursts divide naturally into two populations: ``fast bursts'' occur on the Banana Branch when the accretion rate is high and exhibit short decay times, high peak fluxes, and radius expansion episodes. ``Slow bursts'' occur in the Island State at lower accretion rates, have lower peak fluxes, higher fluences, longer decay times, and show no evidence of radius expansion. All five of the fast bursts, and none of the four slow bursts, show coherent oscillations near 524~Hz. Thus the mechanism that produces the burst pulsations may well be related to the helium-rich burning process indicated by the other properties of the fast bursts. We perform in-burst phase connection of the burst pulsations, which allows us to unambiguously characterize their frequency evolution. That evolution exhibits a variety of behaviors, including a sharp spin-down during one burst. The frequency evolution exhibited by two bursts that occurred 2.6~years apart is consistent with nearly the same exponential-relaxation model; in particular, the asymptotic frequencies of the two bursts differ by only $0.13\\pm0.09$~Hz. The evolution during another burst, not modeled as exponential, shows a maximum frequency which is $0.6\\pm0.2$~Hz greater than the smaller of these asymptotic frequencies. Applying our phase models, we find that the pulsations are spectrally harder than the burst emission, with the strength of the pulsations increasing monotonically with photon energy. Coherently summing all of the burst signals, we find upper limits of \\twid{5\\%} relative to the power of the main pulse on any modulation at 1/2 or 3/2 the main pulse frequency. We discuss our results in the context of models in which the burst pulsations are produced by a hot spot on the surface of a spinning neutron star. ", "introduction": "\\ksxrb\\ is a transient X-ray binary system located near the galactic center, discovered in August~1989 using the imaging spectrometer aboard the {\\it Mir-Kvant\\/} observatory \\citep{sun89}. Three type~I X-ray bursts were observed from the source during 24 1000~s observations spanning 15~days \\citep{sun90}, which established that the source was a neutron star low mass X-ray binary (LMXB). Subsequent ROSAT observations confirmed that \\ksxrb\\ was a typical neutron star LMXB \\citep{bar98}. The source has been detected once in hard X-rays (35--150 keV) by SIGMA, making \\ksxrb\\ one of several neutron star LMXBs with states similar to the low-hard states of black hole binaries, but during several other observations no hard tail was detected \\citep{bar92}. Recently, attention has focused on \\ksxrb\\ because it is one of only four sources displaying both coherent oscillations during a type~I X-ray burst (at 524~Hz; \\citet{smb97}) and twin kilohertz quasi-periodic oscillations with an approximately constant separation in Fourier frequency; in \\ksxrb\\ the measured frequency difference is close to half of the burst frequency \\citep{wk97}. In this paper we report on a comprehensive analysis of the nine X-ray bursts that were observed during observations of \\ksxrb\\ with the {\\it Rossi X-ray Timing Explorer} (\\rxte) between 1996~August and 1999~February. The bursts occurred while the persistent luminosity varied systematically, which allows us to study how the properties of the bursts correlate with the spectral states of the source. Five of the bursts exhibited coherent oscillations, and we systematically investigate what conditions correspond to the presence of these oscillations. Finally, we have tracked the evolution of the phases of the burst oscillations, which allows us to place tight constraints on any long-term change in oscillation frequency, to derive upper limits on the strength of sub-harmonics and harmonics of the observed burst frequency, and to determine the strength and energy spectrum of the pulsations. To place this work in context, we begin by reviewing briefly the phenomenology and theory of type~I X-ray bursts. In Section~1.1, we examine how the properties of bursts have been observed to correlate with the properties of the persistent emission in several sources, and how theories address these correlations. In Section~1.2, we discuss the observations of nearly coherent oscillations during type~I X-ray bursts and their possible connection with the spin period of the underlying neutron star. The remainder of the paper concerns the analysis (Section~2) and interpretation (Section~3) of \\rxte\\ observations of \\ksxrb. \\subsection{The diversity of type~I X-ray bursts} Type~I X-ray bursts have been observed from more than 45 low-mass X-ray binary systems (LMXBs; see \\citep{lvt93} for a review, and \\citep{coc98} for some recent discoveries). These bursts occur when unstable nuclear burning of helium or hydrogen ignites matter accreted onto the surface of the neutron star primary. The models reproduce the observed rise times (\\twid{1}~s), durations (\\twid{1}~minute), and recurrence times (\\twid{hours}) of bursts, spectral softening in the burst tails, and total energies of the bursts ($10^{39}-10^{40}$ ergs). Detailed models predict variations in the frequency and strength of bursts from a single source due to changes in the composition of the burning material, which is in turn determined by the metallicity of the matter accreted onto the neutron star, the amount of hydrogen burned during the time between bursts, and the amount of fuel left-over from the previous burst \\citep{fhm81, aj82, fl87, fuj87, bil97}. In addition, variations from source to source are expected because of variations in the core temperatures of the neutron stars and accretion rate \\citep{aj82, fl87}. The theoretical models that are most readily compared to observations of burst sources are those that predict how burst properties in an individual system should change as the accretion rate onto the neutron star varies. Of the forty-odd bursting LMXBs, only about ten have had their burst and persistent emission properties correlated; for the majority of sources, either few bursts are seen, or several bursts are seen in a single luminosity state \\citep{lvt93}. Nonetheless, studies of these ten sources reveal some intriguing relationships between the persistent emission and the recurrence times, durations, and energetics of the bursts which can be compared to theoretical models. The model of \\citet{fhm81} predicts that X-ray bursts should occur in three regimes depending on the accretion rate ($\\dot M$), which they label Cases~1, 2, and 3. We denote these accretion rates in units of the Eddington accretion rate, \\mdotedd. At the lowest accretion rates ($< 0.01$\\mdotedd, Case~3), the temperature in the burning layer is too low for stable hydrogen burning; the hydrogen ignites unstably, in turn triggering helium burning, which produces a type~I X-ray burst in a hydrogen-rich environment. At higher accretion rates ($>0.01$\\mdotedd, Case~2), hydrogen burns stably via the hot CNO cycle and is converted to helium as quickly as material is accreted onto the neutron star. A pure helium layer develops at the base of the accreted material, and heats steadily until a pure helium burst is triggered. At higher accretion rates ($> 0.1$\\mdotedd, Case~1), material is accreted faster than it can be consumed by hydrogen burning (which is limited by the rate of $\\beta$-decays in the CNO cycle), so that the helium ignites unstably in the H-rich environment (see Bildsten 1998 for dependences of these accretion rates on the metallicities). At higher accretion rates ($>$\\mdotedd), stable helium burning becomes important on the surface of the neutron star, which depletes the primary fuel reserves and causes bursts to occur less frequently. \\citet{hf82} point out that the accretion rate inferred from the luminosity of any given neutron star LMXB that exhibits bursts can range from $\\sim 0.01-0.1$ \\mdotedd, so that an individual source that varies in luminosity should exhibit changes in its bursting behavior between Case~2 and Case~1. This model provides three predictions. First, as the accretion rate increases, the column of material above the burning layer builds more quickly, and thus the time required to reach the critical temperature for unstable helium burning decreases. This is consistent with the increase in burst rate with persistent flux observed in \\uxrbs{1728}{34} \\citep{bas84} and \\uxrbs{1820}{30} \\citep{cla77}. At the highest accretion rates, steady helium burning reduces the amount of fuel for X-ray bursts, which causes bursts to occur less often, as seen in EXO~0748$-$676 \\citep{got86} and \\uxrbs{1705}{44} \\citep{lan87}, or not at all, as observed in \\uxrbs{1820}{30} \\citep{cla77} and GX~3$+$1 \\citep{mak83}. However, no correlation was found between persistent flux and burst recurrence times in Ser~X-1 \\citep{szt83}, \\uxrbs{1735}{44} \\citep{lew80, par88}, and \\uxrbs{1636}{53} \\citep{lew87}, which suggests either that an additional mechanism may control how frequently bursts occur, or that the persistent flux is not a good measure of the accretion rate in these sources (see below). Second, the ratio of the energy released in the bursts to that released by accretion between bursts should be larger when large amounts of hydrogen are burned during a burst, and should decrease as more helium is added to the nuclear burning layer. Assuming that all of the accreted material is eventually consumed in a burst, this ratio is simply the amount of energy released by the nuclear reactions --- the CNO cycle for hydrogen burning, and the triple-$\\alpha$ process for helium burning --- divided by the gravitational energy released by the accreted material. Historically, the inverse of this ratio is used, and defined as $\\alpha$. The expected values for $\\alpha$ are $\\sim25-100$, where the lower value corresponds to pure hydrogen burning, which releases about four times the energy per nucleon than helium burning does. The value of $\\alpha$ was observed to increase as expected (from 12 to 75) as the persistent flux from EXO 0748$-$763 increased by a factor of five \\citep{got86}. However, several sources exhibit $\\alpha$ values as large as $10^3$, which suggests that both hydrogen and helium burn steadily during the periods of steady accretion between bursts \\citep{ppl88, bil95}. Third, the larger the fraction of helium in a burst, the faster nuclear energy will be released (via the triple-alpha process) during the helium flash, so that the peak flux of the burst will be higher \\citep{fhm81, aj82}. Conversely, if more hydrogen is present in the burning material, as in Cases~1 and 3, then helium will be diluted, slowing the rise of the burst. The hydrogen will, however, burn slowly in the tail of the burst via proton capture onto the products of the helium burning. Thus the fastest, most intense bursts should occur under Case~2, which corresponds to low accretion rates. Surprisingly, this is precisely contrary to the sense of the majority of the observations. The decay time scales of bursts from \\uxrbs{1608}{52} \\citep{mur80}, \\uxrbs{1636}{53} \\citep{lew87}, and \\uxrbs{1705}{44} \\citep{lan87} have all been reported to decrease with increasing flux. The peak fluxes of bursts were found to increase as the persistent flux increased in EXO~0748$-$676 \\citep{got86}, \\uxrbs{1608}{52} \\citep{mur80}, and Ser~X-1 \\citep{szt83}, although \\uxrbs{1705}{44} \\citep{lan87} and \\uxrbs{1735}{44} \\citep{lew80, par88} exhibit no apparent correlation between the peak fluxes of bursts and the persistent flux. A survey of ten LMXBs by \\citet{ppl88} reveals a global decrease in burst duration with increasing flux. The sense of these observations is that the fastest, most intense bursts occur at high accretion rates. The origin of this discrepancy is uncertain (see Bildsten 2000 for a recent discussion). It is apparent that for each correlation mentioned above, only a few sources can be cited as examples. This is largely because two systematic difficulties interfere with observation of these effects. First, recurrence times and $\\alpha$ values for bursts are difficult to measure because most satellites observe from low Earth orbit, where the Earth occults most sources for a significant fraction of every hour. Since bursts recur on a time scale of hours, many bursts are missed during occultation. Only EXOSAT, with its 91~hour orbit, was relatively free from this difficulty. Second, the persistent flux from an LMXB is not a monotonic function of accretion rate. As the trends predicted by the model of Fujimoto, Hanawa, \\& Miyaji (1981) are a function of the accretion rate, it is not surprising that correlations between burst properties and persistent flux would not be evident in every source. A better measure of accretion rate is provided by the color-color diagram \\citep{vdk95}. Indeed, in the atoll source \\uxrbs{1636}{53} both the burst duration and the apparent temperature of the emission in the burst tails were found to be correlated with the position on the color-color diagram, but not with the flux \\citep{vdk90}. Observations using instruments prior to the EXOSAT mission did not provide sufficient timing and/or spectral information to deduce the spectral state of LMXBs on a color-color diagram unambiguously. Fortunately, the excellent time and energy resolution available with \\rxte\\ allows us to examine the relationship between burst properties and accretion rate in sources such as \\ksxrb. \\subsection{Burst pulsations and kilohertz quasi-periodic oscillations} Although type~I X-ray bursts are useful measures of the energetics, time scales and recurrence frequencies of nuclear burning on neutron stars, the bursts are only observed after the energy has propagated through the neutron star envelope, and so they are not direct probes of the burning layer itself. The discovery of near-coherent oscillations during type~I bursts from several sources using data from the Proportional Counter Array (PCA; \\citep{jah96} of the {\\it Rossi X-ray timing Explorer} ({\\it RXTE}; \\citet{brs93}) has opened a new and extremely valuable window onto X-ray bursts and the neutron stars underlying them (see \\citep{ssz98} for a review). The oscillations have been seen in both rising and cooling phases of the bursts and have extremely high coherence values ($\\nu/\\Delta\\nu > 300$), although the pulse frequency often drifts by up to 2~Hz over the course of the burst \\citep{str98b, sm99}. In particular, the oscillations sometimes relax towards an asymptotic frequency which is identical in bursts separated by several years --- stable enough (to less than one part in one thousand) that it has been suggested it might be used to constrain the binary mass function \\citep{str98b}. The obvious candidate for a pulsation of this high coherence and long-term stability is the spin period of the neutron star \\citep{str96}. Under this interpretation, it is usually proposed that the modulation is the result of localized burst emission (a ``hot spot'') on the neutron star surface. Several lines of argument support this hypothesis. First, the fractional amplitudes of the oscillations in the rise of bursts are generally strongest on the leading edge of the bursts, and are undetectable at the burst peak when a large portion of the neutron star surface is in conflagration \\citep{ssz98}. Second, the RMS strength of the oscillations in the rise of the bursts is very high, up to $\\sim$30\\% in \\slowb\\ \\citep{szs97} and $\\sim$50\\% in \\uxrbs{1636}{53} \\citep{str98c}; note that in both cases the authors report the half-amplitude, which is a factor of 1.4 larger than the RMS amplitude). This would be expected if the initial burning is confined to a small area on the neutron star surface \\citep{ml98}. Finally --- although it is not understood how a hot spot could survive the initial conflagration on the neutron star --- pulse phase-resolved spectroscopy of an oscillation seen in the tail of a burst from \\uxrbs{1636}{53} indicates that the flux modulation is accompanied by a modulation of the blackbody temperature of the spectrum, again as expected (Strohmayer \\etal\\ 1998a). A similar conclusion has been reached for a burst from \\aqlx\\ through the detection of soft lags in the burst photons, which are attributed to Doppler shifts of the light from the fast-moving hot spot \\citep{for99}. The favored model for the drifts in frequency is that they indicate an expansion of the burning layer by 10--30~m at the burst start. During the expansion, the rotation of the burning layer slows, only to relax back to the neutron star surface and spin up again later in the burst \\citep{str98b}. Cumming \\& Bildsten (2000) have shown that the hydrostatic expansion expected during the burst is consistent with that observed, and that the shear layers are likely stabilized by the strong stratification in the atmosphere. Consistent with this hydrostatic model, \\citet{str99} observed a decrease in frequency of the burst pulsations during the tail of a burst from \\uxrbs{1636}{53} which was coincident with a re-heating episode. Still, there are several unanswered questions regarding the hot spot interpretation. Even in sources where the burst pulsations are seen, not all bursts show them, which suggests that the creation of a hot spot requires special conditions. (Until now, no systematic search for correlations between the presence of coherent oscillations and the properties of the bursts or the persistent emission has been carried out.) Moreover, as mentioned above, it is not known how a hot spot could remain on the surface of the neutron star after the initial burning phase, which presumably envelops the whole neutron star \\citep{szs97}. Finally, \\citet{mil00} finds that in one burst from \\uxrbs{1636}{53} (his Burst C) the pulsations have an ``asymptotic'' frequency that is significantly less than the highest frequency observed during the burst, which is hard to reconcile with the simple picture of a burning layer contracting and re-coupling to the underlying neutron star. The importance of the burst oscillations has been further enhanced by the detection of quasi-periodic oscillations (QPOs) in the 200--1200~Hz frequency range in power density spectra (PDS) of the persistent emission from numerous LMXBs. These so-called kHz QPOs (for reviews, see \\citet{vdk97}, \\citet{vdk00}) may reflect the relativistic motions of material at the inner edge of the accretion disk in these systems \\citep{str96, mlp98, sv99}. The relative ubiquity of pairs of kHz QPOs --- with the twin peaks separated by a frequency difference \\nud\\ that stays roughly constant as the frequencies of the peaks increase \\citep{str96, ketal97} --- leads naturally to beat-frequency models which identify \\nud\\ as the spin frequency of the neutron star, \\nus \\citep{str96, mlp98}. Initially this interpretation was supported by the observation that \\nud\\ was consistent (within errors) with the burst frequency \\nub, or half that, in sources where both phenomena were observed. This coincidence was evident in \\slowb\\ and \\uxrbs{1702}{42}\\ ($\\nud\\approx\\nub$), and in \\uxrbs{1636}{53}\\ and \\ksxrb\\ ($\\nud\\approx\\nub/2$). (Note that in the burst-pulsation sources \\aqlx\\ and \\mxbgc\\ twin kHz peaks have not been detected to date.) The detection of a sub-harmonic signal in \\uxrbs{1636}{53}\\ \\citep{mil99} has lent the idea further support by demonstrating that the strongest signal in the bursts from this source is actually at the first harmonic of the spin period, $\\nub=2\\nus$, so that $\\nud\\approx\\nus$. However, it has also since been observed that (1) \\nud\\ is not constant in \\scox\\ \\citep{ketal97}, \\uxrbs{1608}{52}\\ \\citep{men98}, and \\slowb\\ \\citep{mk99}; (2) kHz QPO data from all other sources are equally consistent with a changing or a constant \\nud\\ \\citep{psa98}; and, (3) the ``equalities'' in \\slowb\\ and \\uxrbs{1636}{53}\\ do not in fact hold ($\\nud<\\nub$ in \\slowb, \\citep{mk99}; and $\\nud<\\nub/2$ in \\uxrbs{1636}{53}, \\citep{mkp98}. One response to this dilemma is proposed by \\citet{ot99}, who suggest that the decreasing difference frequency is that of a differentially rotating magnetosphere. In their model, the lower kHz QPO represents the Keplerian frequency at the inner disk, and the upper kHz QPO occurs at an upper hybrid frequency of Keplerian blobs in elliptical orbits, with the frequency determined by the Coriolis force acting in the frame of reference rotating with the neutron star magnetosphere. Another set of models states that the upper and lower kHz QPOs are the general relativistic Keplerian and periastron precession frequencies near the inner edge of the accretion disk \\citep{sv99}. The difference frequency in these latter models is unrelated to the spin period of the neutron star, and the apparent relationship between \\nud\\ and \\nub\\ is taken to be a coincidence. It is therefore critical to determine the nature of the coherent oscillations observed during bursts in order to distinguish between models for the kHz QPOs. \\ksxrb\\ provides crucial observational evidence for testing the current models, as it is one of only four sources for which both the burst frequency and the difference frequency have been measured. ", "conclusions": "We have observed nine type~I X-ray bursts from \\ksxrb\\ with the \\rxte\\ PCA. We find that the bursts can be separated into two categories (Section~2.2, 2.3): fast bursts occur on the Banana Branch when the accretion rate is high, and generally exhibit short decay time scales, high peak flux, radius expansion episodes, and coherent oscillations. Slow bursts occur in the Island State at lower accretion rates, and have lower peak fluxes, higher fluences, and longer decay time scales, but exhibit no evidence of either radius expansion or coherent oscillations. The fast and slow bursts may occur when the nuclear fuel is rich in helium and hydrogen respectively (Sections~1.1 and 3.1). We have analyzed the frequency evolution of the pulsations during the bursts by applying the technique of pulsar phase connection, which allows us to ``count pulses'' and thereby characterize the evolution uniquely (Section~2.5). We find that the frequency evolution during the peak and decay of the bursts exhibits a variety of behaviors, including a sharp spin-down in one burst (Burst~8) which challenges the attribution of the frequency evolution during bursts to expansion or contraction of an expanded burning layer. The frequency evolution during Bursts~1 and 9, separated by 2.6~years, extends for over 9~s in both cases and is consistent with nearly the same exponential-relaxation model, with a best-fit frequency difference of 0.13$\\pm$0.09~Hz. The evolution of the oscillations in Burst~2 is more complicated, and the maximum frequency of the oscillation is greater than those modeled with exponentials by 0.6$\\pm$0.2~Hz. Phase connection of the burst pulsations allows us to make several precision tests of their properties (Section~2.6). We demonstrate the coherence of these pulsations (after accounting for their gross frequency evolution) for $\\gtrsim$5000~cycles. We find that the pulsations are spectrally hard in comparison to the burst emission, with the strength of the pulsations increasing monotonically with photon energy. Coherently summing the signals from all the bursts with detected pulsations, we find marginal (2.1$\\sigma$) evidence for modulation at the first harmonic of the main pulse frequency at 524~Hz, with an average power equivalent to 1.9$^{+1.8}_{-0.9}$\\% that of the main signal. In combination with our upper limits on modulation at 1/2 or 3/2 the main pulse frequency (5.5\\% and 4.3\\% the power of the main signal, respectively), this suggests that the spin frequency in \\ksxrb\\ is more likely to be \\twid{524~Hz} than half that, as the measured frequency difference of the kHz QPOs ($\\nud\\approx 260$~Hz) might otherwise suggest." }, "0003/hep-ph0003285_arXiv.txt": { "abstract": "We discuss the evolution of $U(1)$ symmetric scalar field at the inflation epoch with a pseudo Nambu--Goldstone tilt revealing after the end of exponential expansion of the Universe. The $U(1)$ symmetry is supposed to be associated with baryon charge. It is shown that quantum fluctuations lead in natural way to baryon dominated Universe with antibaryon excess regions. The range of parameters is calculated at which the fraction of Universe occupied by antimatter and the size of antimatter regions satisfy the observational constraints, survive to the modern time and lead to effects, accessible to experimental search for antimatter. ", "introduction": "Phase Distribution for NG Field at The Inflation Period} We start our consideration with the discussion of evolution of $U(1)$ symmetric scalar field which coexists with inflaton at the inflation epoch. The quantum fluctuations of such field during the inflation stage cause the perturbations for the phase marking the Nambu--Goldstone vacuum. In our model this phase determines the sign and value of baryon excess, so the size distribution of domains containing the appropriate phase values, caused by that fluctuations, coincide with the size distribution of antizillahs. Thus to estimate the number density of antimatter regions with sizes exceeding the critical survival size $l_c$ in the baryogenesis model under consideration we have to deal with long -- wave quantum fluctuations of the NG boson field at the inflation period. Various aspects of this question have been examined in the numerous papers \\cite{kofm,lindekofm,spokyok,lindelyth,lindeax,lindebook,lyth,lythstewart,lindebig} in the connection with cosmology of invisible axion. Also the de--Sitter quantum fluctuations have been analyzed in the framework of AD baryogenesis \\cite{b2,b3}. The effective potential of the complex field is taken in the usual form \\beq \\label{3phase} V(\\chi )=-m_{\\chi}^2\\chi^*\\chi+\\lambda_{\\chi}(\\chi^*\\chi )^2+V_0, \\eeq where the field $\\chi$ can be represented in the form \\beq \\label{phase} \\chi =\\frac{f}{\\sqrt{2}}\\exp{\\left(\\frac{i\\alpha}{f}\\right)} \\eeq The $U(1)$ symmetry breaking implies that the radial component of the field $\\chi$ acquires a nonvanishing classical part, $f=m_{\\chi}/\\sqrt{\\lambda_{\\chi}}$ and field $\\alpha$ in eq. (\\ref{phase}) becomes a massless NG scalar field with a vanishing effective potential, $V(\\alpha )=0$. In this case, $\\chi$ has the familiar Mexican--hat potential, and the degenerated vacua correspond to the circle of radius $f$. Throughout present paper we deal with dimensionless angular field $\\theta =\\alpha /f$. We concern here the possibility to store appropriate phase value in the domain with the size exceeding the critical survival size. Such value of phase plays the role of starting point for clockwise movement, which is going to generate antibaryon excess when the tilt of potential breaking $U(1)$ explicitly, will turn to be topical. We assume that the Hubble constant varies slowly during inflation. Also we use well established behavior of quantum fluctuations on the de Sitter background \\cite{lindebook}. It implies that vacuum fluctuations of every scalar field grow exponentially in the inflating Universe. When the wavelength of a particular fluctuation becomes greater than $H^{-1}$ the average amplitude of this fluctuation freezes out at some nonzero value because of the large friction term in the equation of motion of the scalar field, whereas its wavelength grows exponentially. In the other words such a frozen fluctuation is equivalent to the appearance of classical field that does not vanish after averaging over macroscopic space intervals. Because the vacuum must contain fluctuations of every wavelengths, inflation leads to the creation of more and more new regions containing the classical field of different amplitudes with scale greater than $H^{-1}$. The averaged amplitude of such NG field fluctuations generated during each time interval $H^{-1}$ is given by \\cite{alstar} \\beq \\label{b11} \\delta\\alpha =\\frac{H}{2\\pi} \\eeq During such time interval the universe expands by a factor of $e$. Since the NG field is massless during inflation period (the PNG tilt is vanish yet), one can see that the amplitude of each frozen fluctuation is not changed in time at all and the phases of each wave are random. Thus the quantum evolution of NG field looks like one--dimensional Brownian motion \\cite{lindebook,lindebig} along the circle valley corresponding to the bottom of NG potential. This statement implies that the values of the phase $\\theta$ in different regions become different, and the corresponding variance grows as \\cite{lindebig47} \\beq \\label{b12} \\langle (\\delta\\theta )^2\\rangle =\\frac{H^3t}{4\\pi^2f^2} \\eeq that means that dispersion grows as $\\sqrt{\\langle (\\delta\\theta )^2\\rangle}=\\frac{H}{2\\pi f}\\sqrt{N}$, where N is the number of e--folds. In the other words the phase $\\theta$ makes quantum step with the scale $\\frac{H}{2\\pi f}$ at each e--fold, and the total number of steps during some time interval $\\Delta t$ is given by $N=H\\Delta t$. Let us consider the scale $k^{-1}=H_0^{-1}=3000h^{-1}Mpc$ which is the biggest cosmological scale of interest. We suppose that Universe is baryon asymmetric in this scale which leaves the horizon at definite e--fold $N=N_{max}$. On the other side this scale is the one entering the horizon now, namely $a_{max}H_{max}=a_0H_0$ where the subscript $0$ indicates the contemporary epoch. This implies that: \\beq \\label{b13} N_{max}=\\ln{\\frac{a_{end}H_{end}}{a_0H_0}}- \\ln{\\frac{H_{end}}{H_{max}}} \\eeq the subscript $end$ denotes the epoch at the end of inflation. The slow-roll paradigm tells us that the last term of (\\ref{b13}) is usually $\\le 1$. The first term depends on the evolution of scale factor $a$ between the end of slow-roll inflation and the present epoch. Assuming that inflation ends by short matter dominated period, which is followed by RD stage lasting until the present matter dominated era begins, one has \\cite{LythLiddlePhysRep} \\beq \\label{b14} N_{max}=62-\\ln{\\frac{10^{16}GeV}{\\sqrt{H_{end}M_p}}}- \\frac{1}{3}\\ln\\frac{\\sqrt{H_{end}M_p}}{\\rho_{reh}^{1/4}}, \\eeq where $\\rho_{reh}^{1/4}$ is the reheating temperature when the RD stage is established. With $H_{end}\\simeq 10^{13}GeV$ and instant reheating this gives $N_{max}\\approx 62$, the largest possible value. However, if one has to invoke supersymmetry to prevent the flatness of the inflation potential, for example like as in the case of AD baryogenesis, the $\\rho_{reh}^{1/4}$ should not exceed then $10^{10}GeV$ to avoid too many gravitino overproduction \\cite{KhlopovLinde}, and one have $N_{max}=58$, perhaps the biggest reasonable value. Through the paper we will use $N_{max}=60$. The smallest cosmological scale of antizillah that is survived after annihilation is $k^{-1}_{c}=l_{c}\\approx 8h^2kpc$ \\cite{we}. It is $9$ order of magnitude smaller then $H^{-1}_0$, that corresponds to \\beq \\label{b15} N_c\\approx N_{max}-13-3\\ln{h}\\approx 45 \\eeq Thus the $l_c$ should left horizon at 45--folds before the end of inflation. \\begin{figure}[t] \\centering \\leavevmode\\epsfysize=5.1cm \\epsfbox{fff.ps}\\\\ \\ \\caption[fig2]{\\label{fig1} Baryosynthesis in the spontaneous baryogenesis mechanism. The sign of baryon asymmetry depends on the starting point of phase oscillations.} \\end{figure} Let us assume that the phase value $\\theta =0$ corresponds to South Pole of NG field circle valley, and $\\theta =\\pi$ corresponds to the opposite pole. The positive gradient of phase in this picture is routed as anticlockwise direction, and the dish of PNG potential would locate at the South Pole of circle (see fig.\\ref{fig1}). It will be shown below (see section III) that the antibaryon production corresponds to the regions that would contain phase values caused anticlockwise rolling of PNG field $\\alpha$ during the first half period of oscillation. If the field $\\alpha$ rolls clockwise towards the dish of tilted potential just after the start of first oscillation then baryon production will take place. Now we are in the position to estimate the fraction of the Universe containing antizillahs. To ensure that the Universe would be baryon asymmetric as a whole it is necessary to suppose that the phase average value $\\theta =\\theta_{60}$ within biggest cosmological scale of interest emerging at the $N_{max}=60$ e--folds before the end of inflation is located in the range $[0,\\pi ]$. The $\\theta_{60}$ is the starting point for Brownian motion of the phase value along the circle valley during inflation. As it has been mentioned above, the phase makes Brownian step $\\delta\\theta =\\frac{H}{2\\pi f}$ at each e--fold. Because the typical wavelength of the fluctuation $\\delta\\theta$ generated during such timescale is equal to $H^{-1}$, the whole domain $H^{-1}$, containing $\\theta_{60}$, after one e--fold effectively becomes divided into $e^3$ separate, causal disconnected domains of radius $H^{-1}$. Each domain contains almost homogeneous phase value $\\theta_{60-1}=\\theta_{60}\\pm\\delta\\theta$. In half of these domains the phase evolves towards $\\pi$ (the North Pole) and in the other domains it moves towards zero (the South Pole). To have antizillah with appropriate sizes to avoid full annihilation one should require that the phase value crosses $\\pi$ or zero not later then after $15$ steps. Only in this case the antizillahs would have the sizes larger than $l_c$ and are conserved up to the modern era. This means that one of the two following inequality must be satisfied \\beq \\label{b16} \\pi -\\frac{15H}{2\\pi f}\\le\\theta_{60}\\le\\frac{15H}{2\\pi f} \\eeq Consider initially the case of exact equalities in expression (\\ref{b16}) when the main part of antimatter is contained in the antizillahs of size $l_c$. The number of domains containing the equal values of phase at the $45$ e--folds before the end of inflation is given by the following expression \\beq \\label{b17a} n_{45}\\approx (e^3/2)^{15}\\approx 10^{15}. \\eeq Then the probability that every domain of size $l_c$ would not be separated into $e^3$ domains with size one order of magnitude less then $l_c$ at the next e--fold is given by $P_s\\approx (1/2)^{e^3}\\approx 10^{-6}$. Thus the number of domains serving as the prototypes for antizillahs of size $l_c$ looks like \\beq \\label{b18} \\bar n=n_{45}P_s\\approx 10^9 \\eeq There are about $10^{11}$ galaxies in the Universe. Thus, according to such simple consideration, we reveal that $1\\%$ of volume boxes corresponding to each galaxy contains the region of size $l_c$ filled with antimatter of highest possible antibaryonic density if the $\\theta_{60}$ coincides with left side of inequality (\\ref{b16}) or lowest one in the case if the opposite equality is held. We are able also to find the size distribution for antizillahs. For this purpose it is necessary to study the inhomogeneities of phase induced by (\\ref{b11}). It has been well established that for any given scale $l=k^{-1}$ large scale component of the phase value $\\theta$ is distributed in accordance with Gauss's law \\cite{alstar,lindebig47,lindebook,lindebig}. The quantity which will be especially interesting for us is the dispersion (\\ref{b12}) for quantum fluctuations of phase with moments from $k=H^{-1}$ to $k_{min}=l^{-1}_{max}$ (where the $l_{max}$ is the biggest cosmological scale that corresponds to $60$ e--folds). This quantity can be expressed in the following manner \\beq \\label{b19} \\sigma^2_l=\\frac{H^2}{4\\pi^2}\\int\\limits_{k_{min}}^kd\\ln{k}=\\nonumber\\\\ \\frac{H^2}{4\\pi^2}\\ln{\\frac{l_{max}}{l}=\\frac{H^2}{4\\pi^2f^2}(60-N_l)}, \\eeq where $N_l$ is the number of e--folds which relates the biggest cosmological scale to the given scale $l$. This means that the distribution of phase has the Gaussian form \\beq \\label{b21} P(\\theta_l ,l)=\\frac{1}{\\sqrt{2\\pi}\\sigma_l} \\exp{\\left\\{ -\\frac{(\\theta_{60}-\\theta_l)^2}{2\\sigma^2_l}\\right\\}} \\eeq Suppose that at e--fold $N_t$ before the end of inflation the volume $V(\\bar\\theta ,N_t)$ has been filled with phase value $\\bar\\theta$. Then at the e--fold $N_{t+\\Delta t}=N_t-\\Delta N$ the volume filled with phase $\\bar\\theta$ will follow iterative expression \\begin{eqnarray}\\label{iteration} V(\\bar\\theta , N_{t+\\Delta t})=e^3V(\\bar\\theta ,N_t)+\\nonumber\\\\ +(V_U(N_t)-e^3V(\\bar\\theta ,N_t)P(\\bar\\theta , N_{t+\\Delta t})h. \\end{eqnarray} Here the $V_U(N_t)\\approx e^{3N_t}H^{-3}$ is the volume of the Universe at $N_t$ e--fold. Expression (\\ref{iteration}) makes it possible to calculate the size distributions of domains filled with appropriate value of phase numerically. In order to illustrate quantitatively the number distribution of domains, we present here the numerical results for definite values of $\\theta_{60}$, $\\bar\\theta$ and $h=\\frac{H}{2\\pi f}$. The table contains the results concerning to number of domains with average phase $\\bar\\theta$ at e-fold number $N$, \\begin{table} \\caption{The sample of distribution of proto--antizillahs by sizes and numbers of e--folds at $ \\theta _{60}=\\frac{\\pi }{6}$; $\\overline\\theta =-0$; $h=0.026$} \\begin{tabular}{lcc} $N$&$N_{antizillahs}$&$L_{antizillah}h$ \\\\ \\tableline $59$&$0$&$1103$Mpc\\\\ $55$&$5.005 \\, \\times \\, 10^{-14}$&$37.7$Mpc\\\\ $54$&$7.91 \\, \\times \\, 10^{-10}$&$13.9$Mpc\\\\ $52$&$1.291 \\, \\times \\, 10^{-3}$&$1.9$Mpc\\\\ $51$&$0.499$&$630$kpc\\\\ $50$&$74.099$&$255$kpc\\\\ $49$&$8.966 \\, \\times \\, 10^{3}$&$94$kpc\\\\ $48$&$8.012 \\, \\times \\, 10^{5}$&$35$kpc\\\\ $47$&$5.672 \\, \\times \\, 10^{7}$&$12$kpc\\\\ $46$&$3.345 \\, \\times \\, 10^{9}$&$4.7$kpc\\\\ $45$&$1.705 \\, \\times \\, 10^{11}$&$1.7$kpc\\\\ \\end{tabular} \\end{table} The fraction of the Universe filled with phase $\\bar\\theta$ appears to be equal to $7.694\\times 10^{-9}$. Thus we see that the distribution of domains with size is very abrupt and should be peaked at smallest value of size. Adjusting the free parameters $\\theta_{60}$ and $h$ we are able to achieve the situation that volume box corresponding to each galaxy contains ($1\\div 10$) regions with appropriate phase $\\bar\\theta$. The sizes of such regions are larger or equal to critical surviving size. In spite of the sufficiently large total number of antizillahs only the small part of our Universe will be occupied by antizillahs (see the last line in the presented table). The nontrivial question on the actual forms of astrophysical objects antizillahs can have in the modern Universe needs spacial analysis, which, in general, strongly depends on the assumed form of the nonbaryonic dark matter, dominating in the period of galaxy formation. However, based on the early analysis \\cite{we,khlop,ams} the two extreme cases can be specified, when the evolution of antizillahs is not strongly influenced by the dark matter content. In the first case, the antibaryon density within the antizillah is by an order of magnitude higher than the average baryon density, so that the over-density inside this region can exceed the dark matter density and rapid evolution of such an antizillah with the size exceeding the surviving scale can provide formation of compact antimatter stellar system (globular cluster (see for review \\cite{glob})) which can survive in galaxy \\cite{khlop,ams}. The other extreme case is antizillah with extremely low internal antibaryon density $\\Omega_{\\bar B}<10^{-5}$. Then the diffused antiworld is realized, when no compact antimatter objects are formed and antizillahs evolve into low density antiproton-positron plasma regions in voids outside the galaxies \\cite{we,khlop}. \\section {\\label{BARYONS} Spontaneous Baryogenesis Mechanism } The following element of our scenario of inhomogeneous baryogenesis should contain a conversion of the phase $\\theta$ into baryon/antibaryon excess. We consider the ansatz of spontaneous baryogenesis mechanism \\cite{cohen}. The basic feature of this mechanism is that the sign of baryon charge created by relaxation of energy of PNG field critically depends on the direction that the phase is rotated on the bottom of Mexican heat potential. It provides us to convert the domains containing the appropriate phase value, caused by fluctuations, to the antizillahs at the period when the NG potential gets the tilt. The one of reasonable issue to the spontaneous baryogenesis \\cite{cohen} has been considered in the work \\cite{dolgmain}. Let us briefly discuss it. It was assumed that in the early Universe a complex scalar field $\\chi$ coexists with inflaton $\\phi$ responsible for inflation. This field $\\chi$ has non vanishing baryon number. The possible interaction of $\\chi$ that violates lepton number can be described by following Lagrangian density (see e.g. \\cite{dolgmain}) \\begin{eqnarray} \\label{1} L=-\\partial_{\\mu}\\chi^*\\partial^{\\mu}\\chi -V(\\chi )+i\\bar Q\\gamma^{\\mu}\\partial_{\\mu}Q + i\\bar L\\gamma^{\\mu}\\partial_{\\mu}L - \\nonumber\\\\ -m_Q\\bar QQ- m_L\\bar LL + (g\\chi\\bar QL+h.c.) \\end{eqnarray} The fields $Q$ and $L$ could represent heavy quark and lepton, coupled to the ordinary quark and lepton matter fields. Since fields $\\chi$ and $Q$ possess baryon number while the field $L$ does not, the couplings in the (\\ref{1}) violate lepton number \\cite{dolgmain}. The $U(1)$ symmetry that corresponds to baryon number is expressed by following transformations \\beq \\label{2} \\chi\\to\\exp{(i\\beta )}\\chi ,\\qquad Q\\to\\exp{(i\\beta )}Q,\\qquad L\\to L \\eeq The effective Lagrangian density for $\\theta$, $Q$ and $L$ eventually has the following form after symmetry breaking \\cite{dolgmain} \\begin{eqnarray} \\label{5} L=-\\frac{f^2}{2}\\partial_{\\mu}\\theta\\partial^{\\mu}\\theta +i\\bar Q\\gamma^{\\mu}\\partial_{\\mu}Q + i\\bar L\\gamma^{\\mu}\\partial_{\\mu}L - \\nonumber\\\\ -m_Q\\bar QQ -m_L\\bar LL + (\\frac{g}{\\sqrt{2}}f\\bar QL+h.c.)+\\partial_{\\mu}\\theta\\bar Q\\gamma^{\\mu}Q \\end{eqnarray} At the energy scale $\\Lambda <0$ if $\\theta >0$ during classical movement of phase $\\theta $ to zero. Thus the anticlockwise rotation gives rise to antibaryon excess while the clockwise rotation to the baryon excess one. During reheating, the inflaton energy converts into the radiation. It is assumed that reheating takes place when the Mexican--hat potential is not sensitive to the PNG tilt yet. This implies that the total decay width of inflaton $\\Gamma_{tot}$ into light degrees of freedom exceeds the mass $m_ {\\theta}$. In the other words the reheating is going on under the condition (\\ref{cond}). The relaxation of $\\theta$ field starts when $H\\approx m_{\\theta}$ and converts to the baryons or antibaryons. Baryonic charge is converted inside a comoving volume after reheating owing to very effective decay during the cosmological time. This means that the baryon--to--entropy ratio in $n_{B(\\bar B)}/s=Const$ in the course of expansion. The entropy density after thermalization is given by \\beq \\label{entropy} s=\\frac{2\\pi^2}{45}g_*T^3 \\eeq where $g_*$ is the total effective massless degrees of freedom. Here we concern with the temperature above the electroweak symmetry breaking scale. At this temperature all the degrees of freedom of the standard model are in equilibrium and $g_*$ is at least equal to $106.75$. The temperature is connected with expansion rate as follow \\beq \\label{habble} T=\\sqrt{\\frac{m_pH}{1.66g_*^{1/2}}}=\\frac{\\sqrt{m_pm_{\\theta}}} {g_*^{1/4}} \\eeq The last part of expression (\\ref{habble}) takes into account that the relaxation starts under the condition $H\\approx m_{\\theta}$. Using the formulas (\\ref{bar1}), (\\ref{entropy}), (\\ref{habble}) we are able to get the baryon/antibaryon asymmetry \\beq \\label{asym} \\frac{n_{B(\\bar B)}}{s}=\\frac{45g^2}{16\\pi^4g_*^{1/4}}\\left(\\frac{f}{m_p}\\right)^{3/2} \\frac{f}{\\Lambda}Y(\\theta_i) \\eeq The function $Y(\\theta)=\\theta ^2 \\int\\limits_{-\\theta /2}^{\\theta /2}d\\omega\\frac{\\sin^2{\\omega}}{\\omega^2}$ takes into account the dependence of amplitude of baryon asymmetry and its sign on the initial phase value in the different space regions during inflation. The expression (\\ref{asym}) allows us to get the observable baryon asymmetry of the Universe as a whole $n_B/s\\approx 3\\cdot 10^{-10}$. In the model under consideration we have supposed initially that $f\\ge H\\simeq 10^{-6}m_p$. The natural value of coupling constant is $g\\le 10^{-2}$. We are coming to observable baryon asymmetry at quite reasonable condition $f/\\Lambda\\ge 10^5$ (see e.g. \\cite{freese}). ", "conclusions": "Discussion} In this paper we have proposed a model for inhomogeneous baryosynthesis on the base of the spontaneous baryogenesis mechanism \\cite{cohen}. The model predicts the generation of {\\it antizillahs} with sizes exceeding the critical surviving size. The antibaryon number density relative to background baryon density in the resulting antizillahs and its number depends on the value of phase established at the beginning and on the parameters of PNG field potential. It is possible to have one or several antizillahs the volume box corresponding to every galaxy depending on the parameter values. The observational consequences of existence of antizillahs and the restrictions on their number and sizes have been analyzed in papers \\cite{we,khlop,ams} Of course we may in general expect that some region with size exceeding $l_c$ would contain antibaryon excess after the annihilation of small primordial domains and antidomains contained in this region is completed. However the probability to have such region is suppressed exponentially. Therefore to have observational acceptable number of antimatter regions \\cite{khlop} with the size exceeding the critical survival size, a superluminous cosmological expansion in the formation of primordial antimatter proto--domain seems necessary. As we have mentioned, the additional problem for the most models of inhomogeneous baryogenesis invoking phase transitions at the inflation epoch is prediction of the large scale unwanted topological defects. Our scheme contains the premise for existence of domain walls too. Such walls are not formed when the only minimum of PNG potential exists, what corresponds in the considered model to the fluctuations around $\\theta =0$, when the North pole ($\\theta =\\pi$) is not crossed. But in the case, when such crossing takes place the multiple degeneracy of vacua appears (e.g. vacua with $\\theta =0$ and $\\theta =2\\pi$). The equation of motion that correspondes to potential (\\ref{tilt}) admits kink--like, domain wall solution, which interpolates between two adjacent vacua. Thus, when the PNG tilt is significant, domain wall is formed along the closed surface (e.g. $\\theta =\\pi$) \\cite{kim}. In the other words every antizillah with high relative antibaryon density will be encompassed by domain wall bag. The wall stress energy $\\Delta\\approx 8f\\Lambda^2$ \\cite{kim,sik} leads to the oscillation of wall bag after the whole bag enters the cosmological horizon. During the oscillations the energy stored in the walls is released in the form of quanta of NG field and gravitational waves. As we are taken $0<\\theta_{60}<\\pi$, the wall's bag will have the scale of the order of modern horizon, if the dispersion $\\sigma_{l_{max}}$ is large as $\\pi -\\theta_{60}$. Owing to very large oscillation period such big wall bag would presumably survive to the present time, which would be cosmological disaster \\cite{lyth,lythstewart}. Thus the upper limit on the dispersion will be $\\sigma_{60}<\\pi$. From the other hand this condition should be valued if we want to have parameters of antizillah population that do not contradict to direct and indirect observational constraints \\cite{1}. It means that we will have wall bags with the sizes less then cosmological horizon and that walls had to decay until present time. The mechanisms of their decay is a subject of separate paper, in which we also plan to obtain additional constraints on the model, which follow from the condition that walls do not dominate within the cosmological horizon before the bag decays. If the energy density of walls is sufficiently high to give local wall dominance in the border region before the bag enters the horizon, it is of interest to analyze the role of superluminous expansion in the border regions in the bag evolution (see e.g. \\cite{vil}). The interesting question on the wall interaction with baryons in the course of wall contraction and decay will be also studied separately. In general all baryogenesis models that are able to generate some amount of antimatter regions look like radical limit of models with local baryon number density fluctuations so called isocurvature fluctuations \\cite{LythLiddlePhysRep,iso}. It is known that the contribution of isocurvature fluctuations to the cosmic microwave background (CMB) anisotropy obeys to $\\frac{\\delta T}{T}=-\\frac{1}{3}\\frac{\\Omega_B}{\\Omega_0}\\delta_{B_i}$, where $\\delta_{B_i}$ is the amplitude of initial baryon number fluctuations and ${\\Omega_0}$ (${\\Omega_B}$) are the total (baryon) density (in units of critical density). As it follows from our numerical illustration (see \\ref{PERT} and expression (\\ref{bar1})) we must have quite large amplitude of initial baryon number fluctuations $\\delta_{B_i}\\sim h/\\theta_{60}\\simeq 10^{-2}$ at the biggest cosmological scales, and consequently we would have large amplitude of isocurvature fluctuations at large scales that contradicts with COBE measurements \\cite{iso}. To be keeping away of the problem of large--scale isocurvature fluctuations, we can, for example, prevent the fluctuations of phase at largest cosmological scales. The point is that to have antizillah with size exceeding few kpc. we do not need to start phase fluctuations at the e--folds that correspond to the biggest cosmological scales. It is sufficiently to start fluctuations of phase from the moment, for instance, when the scale $8h^{-1}Mpc$ leaves Habble horizon during inflation, namely after the $6.2$ e--folds from the beginning of inflation. We took this scale, because it is known that at the scale less then $8h^{-1}Mpc$ we could be generated initial baryon number fluctuations at the level $\\delta_{B_i}\\simeq 10^{-2}\\div 10^{-3}$ without any contradictions with observations. One of the natural way to prevent the phase fluctuations at the early inflation is to keep $U(1)$ symmetry restored during first $7$ e--folds. The mechanism that is able to restore symmetry during inflation has been consider in the works \\cite{lindekofm,lindeax,lindebook,sakhlop}. According to that works we can introduce interaction between inflaton field $\\phi$ and field $\\chi$. The simple potential of such kind may be chosen as $ V(\\phi ,\\chi )= \\frac{1}{4}\\lambda_{\\phi}\\phi^4+V(\\chi )+\\nu\\phi^2\\chi^*\\chi $, where $\\nu =m_{\\chi}^2/cM_p^2$, and $c\\simeq 1$. The effective mass of the field $\\chi$ depends on $\\phi$ directly $m_{\\chi}^2(\\phi )= m_{\\chi}^2+\\nu\\phi^2$. One considers here for simplicity the case $\\nu =m_{\\chi}^2/cM_p^2$. This implies that the effective value of mass $m_{\\chi}^2(\\phi )$ during inflation is given by $\\nu (\\phi^2-cM_p^2)$ and is positive because of very large value of the inflation field. It means that our $U(1)$ symmetry is restored during the period when the amplitude of the inflaton field exceeds $\\phi_c=\\sqrt{c}M_P$, and the field $\\chi$ settles into the minimum of its symmetric potential. During this period there was no NG boson valley and phase fluctuations. After the moment that inflaton field turns to be less then $\\phi_c$ the symmetry breaking takes place and the NG potential has the radius $f_{eff}=\\sqrt{\\nu(cM_p^2-\\phi^2)/\\lambda_{\\chi}}$ and fluctuations are started. To keep symmetry restored during first $7$ e--folds we should have $\\phi_c=4M_p$. After the moment of symmetry breaking it is allowed to start the fluctuations of phase with appropriate dispersion to create antizillahs, without any contradictions with observed CMB anisotropy. Of course to evaluate the distribution of antizillahs by sizes we have to take another parameters then we have used in our numerical example, but it does not change the main result of this paper. Another story will take place if we would like to consider the AD baryogenesis as a basis for generation of antizillahs. As it was discussed in the introduction the dynamics of the AD field is more complicated that in the case of spontaneous baryogenesis. Moreover it depends on the fact, D-- or F-- term inflation takes place. Also some details depend on the dimension ($d=4,6..$) of non--renormalizable term lifting the flat direction \\cite{b3,lisa}, but it is enough for the brief discussion to circumscribe ourself with the minimal AD baryogenesis \\cite{b3}, where $d=4$. Thus in the case of D-- term inflation, when the coherent slow rolling of AD field and inflaton are already established, the maximal radius $f_{eff}^{AD(D)}\\simeq 10^{16}GeV$ of effectively massless angular direction can be obtained from the requirement that radial de Sitter fluctuations of AD field would not disturb significantly the spectral index of primordial adiabatic density perturbations \\cite{b3} measured by COBE. Thereby, it is possible to get dispersion of phase fluctuations at the level $h\\simeq 10^{-2}$ that is required for successful generation of antizillahs. The similar situation we could have in the case of F-- term inflation \\cite{b3,lisa} because the AD potential gets an order of $H^2$ negative mass squared term during inflation, which causes the effective minimum at $f_{eff}^{AD(F)}\\simeq C_F\\sqrt{Hm_p}\\simeq 10^{16}GeV$ (the $C_F$ is a constant of order of one). The isocurvature fluctuations in the model of inhomogeneous AD baryogenesis with dispersion of phase fluctuations appropriate for antizillahs generation should be already observed by COBE \\cite{b3}. Moreover this fluctuations can get some amplification owing to possible transformation of fluctuations of AD condensate into the isocurvature fluctuations of neutralinos \\cite{b2}. The exact solution of the problem of isocurvature fluctuations for the AD based antimatter generation is the subject of separate investigation. Here we can only present some speculations, how to avoid the large isocurvature fluctuations at large cosmological scales, which are based on the similar strategy that has been chosen in the case of spontaneous baryogenesis. As it has been mentioned in the Introduction, to organize the angular effectively massless direction in the AD potential we should accept the condition of the absence of order $H$ correction to the A-- term both during and after inflation \\cite{b3}. This condition gets automatically satisfied in the case of D-- term inflation \\cite{mr}, while it is not true if the inflation is F-- term dominated (see for example \\cite{lisa}). According to this observation we can hope to find the such kind of trajectory of inflaton in field space that corresponds to the F-- term dominated inflation in the beginning and then goes into D-- term dominated regime. It implies that during the F-- term dominated inflation the angular direction gets a mass of order $H$ and imaginary component of AD field is dumped and exponentially close to the minimum caused by this effective mass term. In such situation there are no de Sitter fluctuations of the phase. The fluctuations start only at the moment when the inflation goes to the D-- term dominated regime and the angular direction turns to be effectively massless, because there is no correction of order $H$ to the A-- term anymore. As we estimated before, to put the maximal scale of isocurvature fluctuations far below the modern cosmological horizon the transition from F-- term to D-- term inflation should take place 5--10 e--folds after the beginning of inflation. How to organize such transition is the subject of separate publication, but it seems that it could appear, for example, in the context of a realistic supergravity theory deriven from the weak coupled supestring \\cite{string}, which is already beyond the MSSM. There is some possibility to generate the F-- term from a Fayet--Iliopoulos D-- term \\cite{d-f}. It could preserve the flatness of F-- term direction during the first 5--10 e--folds of inflation causing the F-- term domination firstly and subsequent trasformation of the vacuum energy into the D-- term domination mode when it is allowed to begin phase fluctuations of AD field with dispersion appropriate for generation of antizillahs and without contradictions with COBE measurements. We would like to notice in conclusion that the regions with antimatter in matter--dominated Universe could arise naturally in the variety of models. The main issue, that is needed, is a valley of potential. It is the valleys that are responsible for formation of causally separated regions with different values of field which in its turn give rise to antimatter domains. Many extensions of standard model based on supersymmetry possess this property, what strongly extends the physical basis for cosmic antimatter searches. \\bigskip" }, "0003/astro-ph0003253_arXiv.txt": { "abstract": "s{We propose a $\\Lambda$-inflation model which explains a large fraction of the COBE signal by cosmic gravitational waves. The primordial density perturbations fulfil both the contraints of large-scale microwave background and galaxy cluster normalization. The model is tested against the galaxy cluster power spectrum and the high-multipole angular CMB anisotropy.} ", "introduction": "The observational reconstruction of the {\\it cosmological density perturbation} (CDP) spectrum is a key problem of the modern cosmology. It provids a dramatic challenge after detecting the primordial CMB anisotropy as the signal found by DMR COBE at $10^0$ has appeared to be few times higher than the expected value of $\\Delta T/T$ in the most simple and best developed cosmological standard CDM model. During recent years there were many proposals to improve sCDM (in the simplest term, to remove the discrepancy between the CDP amplitudes at $8h^{-1} \\rm{Mpc}$ as determined by galaxy clusters, and at large scales, $\\sim 1000h^{-1} \\rm{Mpc}$, according to $\\Delta T/T$) by adding hot dark matter, a $\\Lambda$-term, or considering non-flat primordial CDP spectra. Below, we present another, presumably more natural way to solve the sCDM problem based on taking into account a possible contribution of {\\it cosmic gravitational waves} (CGWs) into the large-scale CMB anisotropy; we will also try to preserve the original near-scale-invariant CDP spectrum. Thus, the problem is reduced to the construction of a simple inflation producing near {\\it Harrison-Zel'dovich} (HZ) spectrum of CDPs ($n_S\\simeq 1$) and a considerable contribution of CGWs into the large-scale $\\Delta T/T$. A simple model of such kind is $\\Lambda$-inflation, an inflationary model with an effective metastable $\\Lambda$-term\\cite{lm1}${}^,$ \\cite{lm3}. This model produces both S (CDPs) and T (CGWs) modes which have a non-power-law spectra, with a shallow minimum in the CDP spectrum, located at a scale $k_{cr}$ (there the $\\Lambda$-term and the scalar field have equal energies while slowly-rolling at inflation) where the S-slope is exact HZ locally. The S-spectrum is 'red' for $kk_{cr}$; around the $k_{cr}$ scale T/S is close to its maximum, it is of the order unity depending on the model parameters. ", "conclusions": "We have presented a new inflationary model predicting a near scale-invariant spectrum of density perturbations and large amount of CGWs. Our model is consistent with COBE $\\Delta T/T$ and cluster abundance data. The perturbation spectra depend on one free scale-parameter, $k_{cr}$, which can be found in further analysis by fitting other observational data. At the location of $k_{cr}$, the CDP spectrum transfers smoothly from the red ($kk_{cr}$) parts. Today we seriously discuss a nearly flat shape of the dimensionless CDP spectrum within the scale range encompassing clusters and superclusters, \\begin{equation} \\Delta_k^2\\sim k^{(0.9^+_-0.2)}\\;,\\;\\;\\;\\;k \\in (0.04, 0.2) h\\; \\rm{Mpc}^{-1}, \\end{equation} (with a break towards the HZ slope on higher scales) which stays in obvious disagreement with the sCDM prediction. The arguments supporting eq.(13) came from the analysis of large-scale galaxy distribution \\cite{guzzo} and the discovery of large quasar groups \\cite{kl}${}^,$ \\cite{kkl}, a higher statistical support was brought by recent measurements of the galaxy cluster power spectrum \\cite{tadros}${}^,$ \\cite{retz}. A possible explanation of eq.(13) could be a fundamental red power spectrum established on large scales, then the transition to the spectrum (13) at $\\sim 100 \\rm{Mpc}/h$ would be much easier understood with help of a traditional modification of the transfer function $T(k)$ (e.g. for mixed hot+cold dark matter). The rediness may be not too high, remaining in the range (0.9, 1). A way to enhance the power spectrum at Mpc scale could be the identification of $k_{cr}$ within a cluster scale ($k_{cr}\\sim k_1$). Notice that one of the problems for the matter-dominated models is a low number of $\\sigma_8$: if $\\sigma_8<0.6$, then the first acoustic peak in $\\Delta T/T$ cannot be as high as ${}^>_\\sim 70\\;\\mu$K. \\subsection*" }, "0003/astro-ph0003315_arXiv.txt": { "abstract": "We present an approximation for the numerical calculation of \\v Cerenkov radio pulses in the Fraunhofer limit from very high energy showers in dense media. We compare it to full Montecarlo simulations in ice studying its range of applicability and show how it can be extended with a simple algorithm. The approximation reproduces well the angular distribution of the pulse around the \\v Cerenkov direction. An improved parameterization for the frequency spectrum in the \\v Cerenkov direction is given for phenomenolgical applications. We extend the method to study the pulses produced by showers at distances at which the Fraunhofer limit does not apply, and give the ranges of distances and frequencies in which Fraunhofer approximation is good enough for interpreting future experimental data. Our results are relevant for the detection of very high energy neutrinos with this technique. ", "introduction": "The confirmed detection of cosmic rays above the Greisen-Zatsepin-Kuz'min cutoff gives confidence in the existence of neutrinos of energies reaching the EeV scale and above. Such neutrinos are expected both in models in which the protons are accelerated to the highest energies \\cite{nufluxes}, such as in Active Galactic Nuclei \\cite{mannheim} or Gamma Ray Bursts \\cite{bahcall} and in ``top bottom\" scenarios \\cite{Berezinsky,sigl} in which cosmic rays are basically produced through quark fragmentation in events such as the decay of long lived heavy relic particles \\cite{Sarkar} or the annihilation of topological defects \\cite{TD}. If the highest energy component of the cosmic rays are protons, as suggested by increasing experimental evidence \\cite{ave00,vazquez,yakutsk}, they are expected to produce neutrinos in their interactions with the cosmic microwave background \\cite{stecker}. Neutrino detection would provide extremely valuable information on fundamental questions, both in astrophysics, such as the origin of the highest energy cosmic rays and in particle physics. Detecting high energy neutrinos may be a reality in the immediate future as many efforts are being made to develop large scale \\v Cerenkov detectors under water or ice \\cite{AABN}, designed to challenge the low neutrino cross section exploiting the long range of the high energy muons produced in charged current muon neutrino interactions. For EeV neutrinos these detectors are also capable of detecting light from high energy showers produced by neutrinos of any flavor in both neutral and charged current interactions, but the effective acceptance of the detector is reduced because the shower must be produced very close or within the instrumented volume. It has been known for long that the development of showers in dense media produces an excess charge which generates a coherent \\v Cerenkov pulse in the radiowave frequency when it propagates through the medium \\cite{askaryan}. The detection of these pulses provides a possible alternative to neutrino detection particularly appropriate for very high energies \\cite{ralston,provorov,price} because the signal scales with the square of the primary energy \\cite{markov,zas92}. The method is attractive because of the good transmission properties of large natural volumes of ice and sand and because much information about the charge distribution in the shower is preserved in the frequency and angular distribution of the pulses. This last property can be used to extract information about shower energy and neutrino flavor \\cite{alz99}. The technique faces a number of technical difficulties however \\cite{jelley} and several attempts are currently being made to test the theoretical predictions \\cite{Argonne} and to study the feasibility of the technique in Antarctic ice \\cite{RICE}. Theoretical calculations are also difficult because a complete interference calculation calls for simulations capable of following electrons and positrons to the \\v Cerenkov threshold ($\\sim 100~$keV). For high energy showers this is unfortunately out of question because of the large number of particles involved and approximations have been specifically deviced to study the radio emission of high energy showers in ice. The calculation of radio pulses from EeV showers has been possible in the {\\sl one dimensional} (1-D) {\\sl approximation} which consists on neglecting both the lateral distribution and the subluminal velocity of shower particles \\cite{alz99,alz97,alz98}. All the calculations of radio pulses have been made so far in the Fraunhofer limit. In this limit the dependence of the electric field on distance to shower is trivial and the characterization of the angular distribution of the radio pulse at a given frequency is effectively only dependent on one variable, namely the angle between the shower axis and the observation direction what simplifies the simulations \\cite{zas92}. Clearly Fresnel type interference will take place if the showers are close enough to the detectors, but the calculation of these effects becomes even more time consuming. In this paper we firstly give a brief introduction to coherent radio emission in Section II (fuller details can be found in Refs.~ \\cite{zas92,Allan}) accounting for the approximations made. In Section III we make extensive tests and explore the validity of the 1-D approximation in the Fraunhofer limit by direct comparison with complete simulations, and we discuss the approximation pointing out the connections between the radio emission and shower fluctuations, what gives new and useful insight into the radioemission processes. In Section IV we use the 1-D approximation without taking the Fraunhofer limit to study the radiopulse as a function of the distance to observation point. In Section V we summarize and conclude, commenting on the implications of our results for neutrino detection. ", "conclusions": "We have shown that the calculation of coherent \\v Cerenkov radio pulses from high energy showers in ice in the Fraunhoffer limit can be well approximated by neglecting the lateral distributions of the particles assuming that they travel at constant speed ($c$). The electric field amplitude simply becomes the one dimensional Fourier transform of the excess charge depth distribution. For the most relevant region around the \\v Cerenkov direction, the approximation is correct for frequencies below 100~MHz. At higher frequencies the approximation is still relatively good but systematically overestimates the pulse in the \\v Cerenkov direction. We have shown that the model can be made to agree at least up to 1~GHz by subtracting a simple ad-hoc gaussian correction that is proportional to the shower energy and otherwise only dependent on frequency. We have reported the relevant parameters for the correction and have presented an improved parameterization for the electric field amplitude in the \\v Cerenkov direction. We have also shown that instead of the actual charge excess distribution one can use the shower size longitudinal development curve which is more conventional than the excess charge, scaling the amplitude of the central peak by the excess tracklength fraction $0.25$. We have developped a similar approximation for the region in which the Fraunhofer limit ceases to be valid. We have finally studied the behavior of the radiopulses of long electromagnetic showers in this region. Our results are again suggesting to use low frequencies for EeV showers as concluded in Ref\\cite{alz99}. These frequencies have a number of advantages because they are less attenuated, they allow observation of the angular structure with less detectors, and they have milder Fresnel effects at a given distance. Because of Fresnel corrections, the possibility of extracting the mixed character of electron neutrino interactions suggested in \\cite{alz99} requires frequencies below 100~MHz if the electron initiated subshower exceeds about 10~EeV. Lowering the frequency implies a higher threshold for detection because the \\v Cerenkov spectrum increases with frequency but for EeV showers this should not be a problem. It has been estimated that the threshold for detecting showers at 1~km distance with 1~GHz broadband antennas is in the 10~PeV range \\cite{zas92}. Since the signal to noise roughly scales with the square root of the bandwith which directly relates to the central frequency, a factor of 100 reduction in frequency will only call for about a factor of 10 enhancement of the threshold still giving a very large signal to noise ratio for EeV showers. Although our tests of the 1-D approximation rely heavily on a specific simulation program \\cite{zas92}, our claim on the validity of the 1-D approximation is model independent. For testing purposes we used the charge excess distribution and the emitted radiopulses as obtained by the same routine. Numerically our results only apply for ice but it is only natural to expect that the same procedures can be applied to calculate the radiation in other materials. \\vskip 0.5 cm {\\bf Acknowledgements:} We thank P.~Gorham for many early discussions about Fresnel corrections and D.~Besson, D.W. McKay, J.P. Ralston, S. Razzaque, D. Seckel and S. Seunarine for constructive criticism of the Montecarlo and many discussions. This work was supported in part by CICYT (AEN99-0589-C02-02) and by Xunta de Galicia (XUGA-20602B98). J. A. thanks the Department of Physics, University of Wisconsin, Madison and the Fundaci\\'on Caixa Galicia for financial support. E. Z. thanks the Department of Physics, University of Wisconsin, Madison, where this work was finished for its hospitality, and the Xunta de Galicia for partially supporting this trip. \\begin{center} \\bf{APPENDIX A: The ZHS Montecarlo} \\end{center} The simulation program used described in \\cite{zas92} is a specifically deviced program for calculating radio-pulses from electromagnetic showers that follows particles to $\\sim 100$~keV, taking into account low energy processes and timing. The depth development results have been compared to analytical parameterizations given in the Particle Data Book \\cite{pdb}, with which they agree to a few percent. The calculation of the radio emission uses Eq.~\\ref{t1t2} for electron and positron tracks. Several approximations can be made according to different choices in the subdivision of the individual charged particle tracks. In Ref.~\\cite{radiorome} three different choices, named approximations $a$, $b$, and $c$ have been compared, testing for convergence as the subtracks become smaller. Approximation $a$ is the standard that has been used in Refs.~\\cite{zas92,zas91}. It corresponds to taking the end points of all the tracks, and it just uses the average velocity for the corresponding effective track in Eq.~\\ref{t1t2}. This is the standard reference calculation used throughout in this article except for Fig.~\\ref{diff-freq}. Note that this approximation gives the correct result provided the particle velocity is constant along the track. Approximation $b$ subdivides the electron tracks according the different interaction points found along the track, (multiscattering is not considered as an interaction here). This approximation subdivides the track in finer subintervals as the energy becomes smaller, because the low energy electron scattering cross sections exceed bremsstrahlung and pair production. For each subtrack the average velocity is calculated between the corresponding end points of the track. Finally approximation $c$ subdivides each interaction according to a convenient algorithm for spliting the propagation of particles designed to better calculate the multiple scattering at low energies. The three approximations are compared in Fig.~\\ref{abc} illustrating the convergence of the method and how the approximation $a$ is valid in the \\v Cerenkov cone to a precision better than about $10\\%$ for frequencies below 1~GHz. Full simulations in approximation $c$ are much more time consuming and have to be done for shower energies below $\\sim 100~$TeV. At low energies fluctuations from shower to shower are more important so that these tests are inevitably subject to larger uncertainties because of such fluctuations. \\begin{center} \\bf{APPENDIX B: The gaussian approximation} \\end{center} For electromagnetic (hadronic) showers below 10~PeV (10~EeV), that is having no important deviations from Greisen behavior, the electric fied around the \\v Cerenkov cone can be accurately determined with a gaussian approximation. The precise width of the cone inversely relates to the width (in $z$) of the excess charge depth distribution, $Q(z)$. As $p$ is directly related to the observation angle $\\theta$ with an expression that involves the frequency as an overall factor, the width of the angular distribution of the ''central peak'' becomes inversely proportional to $\\omega$. For small deviations from the \\v Cerenkov angle ($\\Delta \\theta$) the expression for $p$ to first order is \\cite{alz99}: \\begin{equation} p = {\\omega \\over {\\rm c}} \\sqrt{n^2-1}~\\Delta \\theta+O(\\Delta \\theta^2) \\simeq 30.8 \\left[ {\\nu \\over 1~{\\rm GHz}} \\right]~ \\Delta \\theta~ ({\\rm m}^{-1}) \\label{papprox} \\end{equation} The numerical value given in this expression corresponds to showers in ice with $n=1.78$. Defining the gaussian width by the points in which the amplitude drops by a factor $\\sqrt{e}$ a gaussian of half-width $\\sigma_z$ transforms to another gaussian of half-width $\\sigma_p = (\\sigma_z)^{-1}$. We can fit a gaussian to the excess charge depth development curve identifying the {\\sl shower length} by the width $l=2 \\sigma_z$ and the angular full width of the radiopulse is then: \\begin{equation} \\sigma_{\\theta} \\simeq 3.72^\\circ \\left[{1~{\\rm GHz} \\over \\nu} \\right] \\left[{1~{\\rm m} \\over l} \\right] \\end{equation} using approximation given by Eq.~~\\ref{papprox}. For a typical shower length of 8 radiation lengths ($\\sim 3.1~$m in ice) the angular width of the pulse is about $1^{\\circ}$ at 1~GHz, in agreement with Ref.~\\cite{zas92}. \\newpage" }, "0003/astro-ph0003409_arXiv.txt": { "abstract": " ", "introduction": "Although photoionization and recombination are inverse processes as they occur in nature, they are usually treated in independent theoretical frameworks. This basic inconsistency, directly related to ionization balance in radiatively ionized media, and consequent inaccuracies, propagate through to the photoionization models employed in astrophysics. A further division, largely artificial, is made in theoretical methods used to compute electron-ion recombination rates. Two sets of data are usually calculated: (i) 'radiative recombination' (RR), calculated using background, or non-resonant, photoionization cross sections, and (ii) 'di-electronic recombination' (DR) representing the contribution of autoionizing resonances, first shown to be important by Burgess (1964). That this procedure is not only theoretically unsatisfactory, but also impractical in most cases, is seen from both theoretical calculations and experimental measurements of photoionization and recombination cross sections. The simple reason is that the resonances are inseparable from the background. The cross sections contain, in general, extensive and interacting Rydberg series of resonances; the non-resonant and resonant contributions are not accurately separable (except, possibly, for few-electron, highly charged ions). The large number of photoionization cross sections computed under the Opacity Project exhibit these features in detail, particularly for many electron systems ({\\it The Opacity Project Team} 1995, 1996 - compiled publications and data) In addition, the cross sections for photoionization and recombination of excited states, particularly metastable levels, may contain even more complicated resonances than the ground state (Luo et al 1990). Experimentally, of course, the measurements {\\it always} yield a combined '(RR + DR)' cross section (albeit in limited energy ranges usually accessible in experimental devices). Therefore a theoretical method that accounts for both the resonant and the non-resonant recombination in a unified manner is desirable, and has been developed (e.g. Nahar and Pradhan 1994, Zhang et al 1999), based on the close coupling (CC) approximation using the R-matrix method (Burke \\& Seaton 1984, Berrington et al 1987, Hummer et al 1993) as used in the Opacity Project and the Iron Project (hereafter OP and IP). Photoionization cross sections may be computed essentially for all bound states, level of excitation ($n,\\ell,SL\\pi,SLJ\\pi)$, energy range, and with resolution of resonances. In principle, the cross section for the inverse photo-recombination process is given by detailed balance. However, since recombination takes place to an infinite number of bound states of (e~+~ion) system, it becomes impractical (and as it turns out, unnecessary) to do so for the very higly excited levels above a certain n-value (chosen to be 10 in practice). For recombination into levels with n$>$10, the non-resonant contribution, relative to the resonant contribution per unit energy is negligible owing to the density of resonances as n $ \\longrightarrow \\infty $. In that range we employ a precise theoretical treatment of DR based on multi-channel quantum defect theory and the CC approximation (Bell \\& Seaton 1985, Nahar \\& Pradhan 1994) to compute the recombination cross section. Among the problems that manfiest themselves in the CC photoionization/recombination calculations are: the accuracy and convergence of the eigenfunction expansion for the ion, relativistic fine structure effects, the contribution from non-resonant recombination into high-n levels as $E \\longrightarrow 0; n \\longrightarrow \\infty$, and resolution of narrow resonances with increasing n and/or $\\ell$, and radiation damping thereof. Experimental work is of importance in ascertaining the accuracy of theoretical calculations and the magnitude of various associated effects, since most of the photoionization/recombination data can only be calculated theoretically. In recent years there has also been considerable advance in the measurements of (e~+~ion) recombination cross sections on ion storage rings (e.g. Kilgus et al, 1990, 1993, Wolf et al. 1991), and photoionization cross sections using accelerator based photon light sources (R. Phaneuf et al., private communication). We compare the CC calculations for both atomic processes with the latest experimental data. ", "conclusions": "" }, "0003/astro-ph0003123_arXiv.txt": { "abstract": "We have derived azimuthally-averaged radial temperature profiles of the X-ray gas contained within 11 clusters of galaxies with redshift $z=0.03-0.2$ observed with {\\it BeppoSAX}. Each of the 11 clusters have had their radial temperature profiles previously determined with {\\it ASCA}. We find that the temperature profiles of these clusters are generally flat or increase slightly out to $\\sim$30\\% of the virial radius, and that a decline in temperature of 14\\% out to 30\\% of the virial radius is ruled out at the 99\\% confidence level. This is in accordance with a previous {\\it ROSAT} PSPC study and an {\\it ASCA} study by White (1999), but in disagreement with an {\\it ASCA} study by Markevitch et al.\\ (1998) that found on average that cluster temperature profiles decreased significantly with radius. ", "introduction": "\\label{sec:intro} Knowledge of the radial temperature profile of the hot gas contained within galaxy clusters is a crucial element in determining the total gravitational mass of clusters. Through the equation of hydrostatic equilibrium, the total mass of the cluster can be derived if the density gradient, temperature, and temperature gradient of the gas is known. The latter quantity is the most difficult quantity to obtain, and it has generally been assumed that the gas is isothermal in most previous approaches, since collimated X-ray instruments such as {\\it Ginga} and {\\it EXOSAT} could not determine the temperature structure of clusters. The assumption of isothermality of the hot gas has been called into question in recent years, mainly as a result of studies done with {\\it ASCA}. The ability of {\\it ASCA} to perform spatially-resolved spectroscopy over the 1--10 keV energy range made it the first X-ray instrument capable of addressing the issue of temperature structure in hot clusters. Many {\\it ASCA} studies have found that the gas within clusters is not isothermal, but decreases with increasing radius, in some cases up to a factor of two (e.g., Markevitch 1996; Markevitch et al.\\ 1998; Markevitch et al.\\ 1999). However, other studies of clusters using {\\it ASCA} data have come to the conclusion that the gas is largely isothermal (e.g., White 1999; Fujita et al.\\ 1996; Ohashi et al.\\ 1997; Kikuchi et al.\\ 1999), at least outside of the cooling radius of cooling flow clusters. A likely cause of this discrepancy is the handling of the large, energy-dependent point spread function (PSF) of {\\it ASCA} that preferentially scatters hard X-rays. This creates an artificial increase in the temperature profile with radius if not dealt with properly. The PSF-correction method applied by Markevitch et al. consistently leads to significantly decreasing temperature profiles, while other methods (most notably the method of White 1999) lead to isothermal profiles. The discrepancy among the different PSF-correction methods prompted an analysis of {\\it ROSAT} PSPC data by Irwin, Bregman, \\& Evrard (1999). Although {\\it ROSAT} was only sensitive to photon energies up to 2.4 keV and was therefore not the most ideal instrument with which to study hot clusters, large (factor of two) differences in temperature should have been detected, but were not. The composite X-ray ``color\" profiles for 26 clusters in the Irwin et al.\\ (1999) survey indicated isothermality outside of the cooling radius. In fact, a 20\\% temperature drop within 35\\% of the virial radius was ruled out at the 99\\% confidence level. In this paper, we attempt to resolve the temperature profile discrepancy using {\\it BeppoSAX} data. {\\it BeppoSAX} is sensitive to photon energies up to 10.5 keV, and has a half-power radius that is one-half that of the {\\it ASCA} GIS instrument. In addition, the PSF of {\\it BeppoSAX} is only weakly dependent on energy. Thus, {\\it BeppoSAX} is better-suited for determining temperature profiles for clusters of galaxies than previous X-ray telescopes. Using a sample of 11 clusters found in the {\\it BeppoSAX} archive, we derive radial temperature profiles for each cluster. In a future paper, we will discuss the abundance profiles of the 11 clusters. Throughout this paper, we assume $H_0=50$ km s$^{-1}$ Mpc$^{-1}$ and $q_0=0.5$. ", "conclusions": "} \\end{figure*} \\subsection{Temperature Profiles Determined With {\\it ROSAT}} \\label{ssec:rosat} Although the limited bandpass of the {\\it ROSAT} PSPC (0.2-2.4 keV) precludes tight constraints to be put on the temperatures of hot clusters, large (factor of two) temperature changes should be detectable with large enough signal-to-noise ratios. Unfortunately, most clusters observed with {\\it ROSAT} did not have good enough statistics to accomplish this on an individual basis. To circumvent this problem, Irwin et al.\\ (1999) averaged together the radial color profiles (ratio of counts in various bands covering the {\\it ROSAT} PSPC bandpass) of 26 clusters observed with the PSPC. If large-scale deviations from isothermality were common in clusters, such a feature lost in the noise for an individual cluster would become apparent when the clusters were added together. Although a drop in temperature was found in the center of cooling flow clusters (indicating that the method could indeed detect changes in temperatures even for hot clusters), the temperature profiles were flat outside of the cooling region out to 35\\% of the virial radius. It was found that a 20\\% temperature drop within 35\\% of the virial radius was ruled out at the 99\\% confidence level. This is in agreement with the {\\it BeppoSAX} data presented here, where a decline in temperature of 14\\% out to 30\\% of the virial radius is ruled out at the 99\\% confidence level. \\subsection{Temperature Profiles Determined With {\\it BeppoSAX}} \\label{ssec:bepposax} Several of the clusters in our sample have been analyzed by other authors: A2199 (Kaastra et al.\\ 1998), A2029 (Molendi \\& De Grandi 1999), A2319 (Molendi et al.\\ 1999), and A3266 (De Grandi \\& Molendi 1999). The profile of A2199 is fully consistent with ours. For the other three clusters, somewhat different steps were taken in the data reduction process than what we did. Whereas we excluded data below 3.0 keV, other authors have included data down to 2.0 keV. In addition, they have frozen the redshift whereas we have let it be a free parameter. We have re-analyzed the {\\it BeppoSAX} data for these three clusters, this time including data down to 2.0 keV and freezing the redshift. Although inclusion of channels below 3.0 keV lowered the derived temperatures somewhat (see \\S~\\ref{ssec:temp_global}), it did not change the overall shape of the profiles. Our new profile of A2319 was consistent with that of Molendi et al.\\ (1999). The profiles of A2029 and A3266 differed somewhat from Molendi \\& De Grandi (1999) and De Grandi \\& Molendi (1999). However, the differences occurred only in the 2$^{\\prime}-4^{\\prime}$ bins, with our 1$\\sigma$ error bars overlapping their 1$\\sigma$ error bars for the other three spatial bins. For both A2029 and A3266 our 2$^{\\prime}-4^{\\prime}$ temperature was $\\sim$1.5 keV lower than those obtained by Molendi \\& De Grandi (1999) and De Grandi \\& Molendi (1999), and the difference was significant at the 3.3$\\sigma$ and 1.9$\\sigma$ confidence levels for A2029 and A3266, respectively. The cause of this discrepancy may be the choice of the assumed surface brightness profile, especially for the case of A2029, which possesses a large cooling flow. As mentioned in \\S~\\ref{sec:sample}, we have used the double-beta model profile obtained by Mohr et al.\\ (1999) from {\\it ROSAT} PSPC data. It is not stated where the other authors obtained their assumed surface brightness profiles. However, the agreement between the other three spatial bins (especially the first and fourth spatial bins) is encouraging, indicating a flat temperature profile out to $9^{\\prime}$ for A2029 and A2319, and a modestly decreasing profile for A3266. It should be noted that Molendi \\& De Grandi (1999) and De Grandi \\& Molendi (1999) find significant temperature drops in A2029 and A3266, respectively, at very large radii, where we have truncated our profiles because of the presence of the strongback, and the increasing asymmetry of the PSF at larger radii." }, "0003/astro-ph0003137_arXiv.txt": { "abstract": "This paper presents and gives the COP (COP: CFHT Optical PDCS; CFHT: Canada-France-Hawaii Telescope; PDCS: Palomar Distant Cluster Survey) survey data. We describe our photometric and spectroscopic observations with the MOS multi-slit spectrograph at the CFH telescope. A comparison of the photometry from the PDCS (Postman et al. 1996) catalogs and from the new images we have obtained at the CFH telescope shows that the different magnitude systems can be cross-calibrated. After identification between the PDCS catalogues and our new images, we built catalogues with redshift, coordinates and V$_{PDCS}$, I$_{PDCS}$ and R$_{COP}$ magnitudes. We have classified the galaxies along the lines of sight into field and structure galaxies using a gap technique (Katgert et al. 1996). In total we have observed 18 significant structures along the 10 lines of sight. ", "introduction": "One of the main goals of the study of distant rich clusters of galaxies is to understand their origin and evolution. Clusters are invaluable cosmological probes, since the evolution of cluster abundances is strongly dependent on the underlying cosmology and therefore can constrain cosmological models (e.g. Bahcall et al. 1997, Oukbir \\& Blanchard 1992 and 1997, Reichart et al 1999, Nichol et al. 1999). In order to be able to exploit this potential, large statistically representative spectroscopic samples are, however, required. There are two types of samples that are being developed: those based on optical selection criteria and those based on X-ray detections. Below z of about 0.1, optically selected studies (e.g. ENACS: Katgert et al. 1996) have had about as many clusters in them as those selected via their X-ray fluxes. Up until recently, however, the higher redshift work has been dominated by X-ray selection techniques (e.g. CNOC: Carlberg et al. 1996 SHARC: Romer et al. 2000; RDCS: Rosati et al. 1998; WARPS: Jones et al. 1998; Vikhlinin et al. 1998). To greatly enlarge the sample of detailed studies of redshift about 0.4 optically selected clusters, we have embarked on a photometric and redshift campaign based on the Palomar Distant Cluster survey (Postman et al. 1996, see also Holden et al. 1999). We have observed a significant number of regions on the sky (10) and obtained about 70 redshifts per line of sight. These pointings were known to contain candidate clusters of galaxies based on the PDCS studies (e.g. Postman et al. 1996, Holden et al. 1997). The main purpose of this paper is to publish the COP survey data and to describe the data reduction so as to lay a foundation for future papers. The interpretation of the results is, therefore, given in later papers (e.g. Holden et al. 2000). The outline of this paper is as follows. In Section 2, we give the observational strategy. In Section 3, we describe the way we have reduced and analyzed our photometry. In Section 4, we describe the way we have reduced and analyzed our spectroscopy. In the last section, we give an analysis of the redshift and spatial distribution of the galaxies in our sample. The data are given in tables 6-15. ", "conclusions": "We have presented and given the data gathered in the COP survey. The spectroscopic and photometric observations were performed with the MOS/STIS2 instrument during 6 nights at the CFH telescope with the grism O300 and 2 blocking filters to enhance the multiplex gain of MOS. We have used a method based on the MST theory to optimize the number of slits per mask. This allowed us to measure 636 redshifts for 10 PDCS lines of sight. These lines of sight were selected to hold PDCS candidate clusters, with significant peaks in the galaxy density distribution. The success rate (percentage of targeted galaxies with a successfully measured redshift) was close to 70$\\%$ for the primary targets (typically brighter than V$_{PDCS}$=22.). The completeness level (percentage of all galaxies with a measured redshift) was about 35$\\%$ down to V$_{PDCS}$=20.5. The galaxies with a redshift were proved to be a spatially representative sub-sample down to V$_{PDCS}$=20.5 (no significant spatial selection effects). Finally, the percentage of false redshifts was about 5$\\%$, based on 19 galaxies observed twice. A comparison of the photometry from the PDCS (Postman et al. 1996) catalogs and from the new images we have obtained at the CFH telescope shows that the different magnitude systems can be cross-calibrated. This confirmation is important for the reliability of future works based on the multi-color photometry of COP. After identification between the PDCS catalogues and our new images, we built catalogues with redshift, coordinates and V$_{PDCS}$, I$_{PDCS}$ and R$_{COP}$ magnitude (Tab. 6-15). We have classified the galaxies along the lines of sight into field and structure galaxies using a gap technique (Katgert et al. 1996). In total we have observed 18 significant structures along the 10 lines of sight (Tab. 4). As noted in the introduction, the interpretation of the results is given elsewhere (e.g. Holden et al. 2000)." }, "0003/astro-ph0003301_arXiv.txt": { "abstract": "Results of a careful analysis of the absorption systems with \\zabs$\\simeq$\\zem seen toward the bright, $z_{\\rm em}$~$\\sim$3.91, gravitationally lensed quasar APM~08279+5255 are presented. \\par\\noindent Two of the narrow-line systems, at \\zabs~=~3.8931 and \\zabs~=~3.9135, show absorptions from singly ionized species with weak or no \\nv~ and \\ovi~ absorptions at the same redshift. Absorption due to fine structure transitions of \\cii~ and \\siii~ (excitation energies corresponding to, respectively, 156$\\mu$m and 34$\\mu$m) are detected at \\zabs~=~3.8931. Excitation by IR radiation is favored as the column density ratios are consistent with the shape of APM~08279+5255 IR spectrum. The low-ionization state of the system favors a picture where the cloud is closer to the IR source than to the UV source, supporting the idea that the extension of the IR source is larger than $\\sim$200~pc. The absence of fine structure lines at \\zabs~=~3.9135 suggests that the gas responsible for this system is farther away from the IR source. Abundances are $\\sim$0.01 and 1~$Z_{\\odot}$ at \\zabs~=~3.913 and 3.8931 and aluminum could be over-abundant with respect to silicon and carbon by at least a factor of two and five. All this suggests that whereas the \\zabs = 3.8931 system is probably located within 200~pc from the QSO and ejected at a velocity larger than 1000~km~s$^{-1}$, the \\zabs = 3.9135 system is farther away and part of the host-galaxy. \\par\\noindent Several narrow-line systems have strong absorption lines due to \\civ, \\ovi~ and \\nv~ and very low neutral hydrogen optical depths. This probably implies metallicities $Z\\geq ~Z_\\odot$ although firm conclusion cannot be drawn as the exact value depends strongly on the shape of the ionizing spectrum. \\par\\noindent The \\civ~ broad absorption has a complex structure with mini-BALs (width $\\le 1000$~\\kms) and narrow components superposed on a continuous absorption of smaller optical depth. The continuous absorption is much stronger in \\ovi~ indicating that the corresponding gas-component is of higher ionization than the other components in the flow and that absorption structures in the BAL-flow are mainly due to density inhomogeneities. There is a tendency for mini-BALs to have different covering factors for different species. It is shown that a few of the absorbing clouds do not cover all the three QSO images, especially we conclude that the \\zabs~=~3.712 system covers only image C.\\par\\noindent Finally we identify {\\sl narrow} components within the BAL-flow with velocity separations within 5~km~s$^{-1}$ of the \\ovi, \\nv~ and \\siiv~ doublet splittings suggesting that line driven radiative acceleration is an important process to explain the out-flow. ", "introduction": "The gravitationally lensed high-redshift Broad Absorption Line (BAL) QSO APM~08279+5255 has been given tremendous interest since its discovery by Irwin et al. (1998) as it is one of the most luminous objects in the universe even after correction for gravitational amplification. Based on the position of the emission lines, Irwin et al. (1998) derived a redshift $z_{\\rm em}$~=~3.87. A probably better estimate of the systematic redshift comes from the detection of CO(4--3) emission at $z_{\\rm em}$~=~3.9114$\\pm$0.0003 by Downes et al. (1999). Imaging of the field reveals two main components (Irwin et al. 1998, Ledoux et al. 1999) separated by 0.378$\\pm$0.001~arcsec as measured on HST/NICMOS data (Ibata et al. 1999) and of relative brightness $f_{\\rm B}/f_{\\rm A}$~=~0.773$\\pm$0.007. The HST images reveal the presence of a third object C with $f_{\\rm C}/f_{\\rm A}$~=~0.175$\\pm$0.008, located in between A and B and almost aligned with them. The PSF fits on the three objects are consistent with the three components being point-sources and the colors are similar within the uncertainties suggesting that C is a third image of the quasar (Ibata et al. 1999, Egami et al. 1999). A high-resolution high signal-to-noise ratio spectrum of APM 08279+5255, covering the wavelength range 4400--10000\\AA~ was obtained using the Keck telescope and made available to the Astronomy community for analysis (Ellison et al. 1999a,b). This spectrum, though complicated by the combination of light traveling along three different sight lines, is a unique laboratory for studying the intervening and associated absorption systems. It is well known that the origin of associated systems (systems with $z_{\\rm abs}$~$\\sim$~$z_{\\rm em}$) cannot be inferred directly from their position in the spectrum. Indeed, absorption can arise from (i) gas ejected by the central engine at velocity as high as 60000~km~s$^{-1}$ and nonetheless physically located very close to the source of ionizing photons (e.g. the \\zabs = 2.24 \"mini-BAL\" towards Q~2343+125 at \\zem~=~2.515; Hamann, Barlow \\& Junkkarinen 1997a) or from (ii) gas associated with the host-galaxy or with members of a galaxy cluster surrounding the quasar. The distinction can be made in terms of physical properties. The systems belonging to the first class are characterized by high metal enrichment, high-ionization parameters, broader line profiles, partial coverage and time variability (Barlow et al. 1992, Petitjean et al. 1994, Savaglio et al. 1994, Hamann 1997, Hamann et al. 1997b, Barlow \\& Sargent 1997, Ganguly et al. 1999, Papovich et al. 1999). The second class of absorber is characterized by classical properties of intervening systems such as low metallicities (typically 0.01 to 0.1 of solar) and undisturbed kinematics. In this study we investigate the nature and physical properties of \\zabs$\\simeq$ \\zem systems toward APM~08279+5255. In Section 2, we describe the data and the grids of photoionization models we use to infer ionizing conditions in the absorbing gas. We analyse a probably intervening metal line system very close to the emission redshift in Section 3. The nature of narrow-line systems with low-ionization conditions are investigated in Section 4. Section 5 describes the high-ionization narrow-line systems. In Section 6, we analyse the nature of the BAL outflow and in Section 7 we suggest the presence of \"line-locking\" among narrow components in the BAL flow. A summary is given in Section 8. \\begin{figure*} \\centerline{\\vbox{ \\psfig{figure=9340.f1,height=11.5cm,width=18.cm,angle=0} }} \\caption[]{Results of photoionization models with $Z=Z_\\odot$, log~$N$(H~{\\sc i}) = 16.0, constant density and plane parallel geometry. The dashed line shows the column density of O~{\\sc vi}. For models in panel(a), the Mathews \\& Ferland (1987) spectrum has been used; the same ionizing spectrum has been attenuated by a slab of gas with perpendicular neutral hydrogen column density log~$N$(H~{\\sc i}) = 17.0 and log~$N$(H~{\\sc i}) = 18.0 and ionization parameter $U$~=~0.1 in panels (b) and (c) respectively. Panel (d) gives the ionizing spectra in each case: solid line is the Mathews \\& Ferland spectrum; dashed and dotted lines are the attenuated spectra for log~$N$(H~{\\sc i}) = 17.0 and 18.0 respectively. } \\label{lis} \\end{figure*} ", "conclusions": "" }, "0003/gr-qc0003036_arXiv.txt": { "abstract": "It is known that the gravitational collapse of a dust ball results in naked singularity formation from an initial density profile which is physically reasonable. In this paper, we show that explosive radiation is emitted during the formation process of the naked singularity. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003071_arXiv.txt": { "abstract": "Large scale, deep survey missions such as GAIA will collect enormous amounts of data on a significant fraction of the stellar content of our Galaxy. These missions will require a careful optimisation of their observational systems in order to maximise their scientific return, and will require reliable and automated techniques for parametrizing the very large number of stars detected. To address these two problems, I investigate the precision to which the three principal stellar parameters (\\teff, \\logg, \\met)\\ can be determined as a function of spectral resolution and signal-to-noise (SNR) ratio, using a large grid of synthetic spectra. The parametrization technique is a neural network, which is shown to provide an accurate three-dimensional physical parametrization of stellar spectra across a wide range of parameters. It is found that even at low resolution (50--100\\,\\AA\\ FWHM) and SNR (5--10 per resolution element), \\teff\\ and \\met\\ can be determined to 1\\% and 0.2 dex respectively across a large range of temperatures (4000--30\\,000\\,K) and metallicities ($-$3.0 to $+$1.0 dex), and that \\logg\\ is measurable to $\\pm 0.2$ dex for stars earlier than solar. The accuracy of the results is probably limited by the finite parameter sampling of the data grid. The ability of medium band filter systems (with 10--15 filters) for determining stellar parameters is also investigated. Although easier to implement in a unpointed survey, it is found that they are only competitive at higher SNRs ($\\geq 50$). ", "introduction": "An understanding of the origin, properties and evolution of our Galaxy requires a careful census of its constituents, in particular its stellar members. Of special importance are the intrinsic physical properties of these stars. The fundamental properties are mass, age and abundances, as these determine a star's history and future development. However, ages are not observable, and masses can only be directly obtained from some multiple systems. Thus we must indirectly gain this information via the stellar spectrum, and a number of atmospheric parameters have been defined for this purpose. The main ones are the effective temperature, \\teff, the surface gravity, \\logg, and the metallicity, \\met. To these can also be added the alpha abundances, \\abun\\ (which measure the devations away from the `standard' abundance ratios), the photospheric microturbulence velocity, \\vmicro, and the extinction by the interstellar medium, \\extinct (although not intrinsic to the star, it is necessary for determining its luminosity). Masses and ages can then be determined from stellar structure and evolution models and with calibration via binary systems. It is important to realise that this modelling is complex, and a number of assumptions have to be made. There is, therefore, a limit to the precision with which we can determine physical properties. Historically, spectroscopic parameters have been measured indirectly through the MK classification system (Morgan et al.\\ \\cite{morgan_43a}) or via colour-magnitude and colour-colour diagrams. In the MK system, the two parameters {\\it spectral type} and {\\it luminosity class} act as proxies for \\teff\\ and \\logg. Originally a qualitative system relying on a visual match between observed spectra and a system of standards, much progress has been made in quantifying it with automated techniques (e.g.\\ Weaver \\& Torres-Dodgen \\cite{weaver_97a}; Bailer-Jones et al.\\ \\cite{bailerjones_98a}). The most commonly used classification techniques have been neural networks and $\\chi^2$ matching to templates (or more generally, minimum distance methods). A summary of recent progress in this area is given by von Hippel \\& Bailer-Jones (\\cite{vonhippel_00a}). Despite this focus on the MK system, it is not well suited to classifying data from the deep surveys which will be central to the future development of Galactic astrophysics. This is for a number of reasons, but in particular because it lacks a measure of metallicity. Although MK does make allowance for various `peculiar' stars, these are defined as exceptions, and the notation is not suited to a statistical, quantifiable analysis. This is problematic given the significance of metal poor halo stars in a deep survey. There is also now no good reason why we should not determine physical parameters directly from the observational data. Some attempts have been made to determine the physical parameters of real spectra directly by training neural networks on synthetic spectra. Gulati et al.\\ (\\cite{gulati_97a}) used this approach to determine the effective temperatures of ten solar metallicity G and K dwarfs. Taking the ``true'' effective temperature of these stars as those given by Gray \\& Corbally \\cite{gray_94a}, they found a mean ``error'' in the network-assigned temperatures of 125\\,K. Bailer-Jones et al.\\ (\\cite{bailerjones_97a}) determined \\teff\\ for over 5000 dwarfs and giants in the range B5--K5, and also showed evidence of sensitivity of the parametrization models to metallicity. The accuracy with which physical parameters can be determined from a stellar spectrum depends upon, amongst other things, the wavelength coverage, spectral resolution and signal-to-noise ratio (SNR). From the point of view of designing a stellar survey project it is essential to know how well the stellar parameters can be determined for a given set of these observational parameters. Moreover, given that there is always a limit to the collecting area and integration time available, there is always a trade-off between spectral resolution, sensitivity and sky coverage. The goal of this paper is to determine the accuracy with which physical stellar parameters can be determined from spectroscopic data at a range of SNRs and resolutions which could realistically be achieved in a {\\it deep} survey mission. This specification rules out high resolution spectra. The parametrization work has been carried out using neural networks (Sect.~\\ref{networks}) because they have been shown to be one of the best approaches for this kind of work. This is not to presuppose, however, that some other approach may not ultimately be better. The simulations have been made using a large database of synthetic spectra generated from Kurucz atmospheric models (Sect.~\\ref{synspec}). While these spectra do not show the full range of variation in real stellar spectra, they are adequate for a realistic demonstration of what is possible as a function of SNR and resolution. The results are presented in Sect.~\\ref{results} and summarised and discussed in Sect.~\\ref{discussion}. Finally, the requirements for a complete survey-oriented classification system are given in Sect.~\\ref{system}. ", "conclusions": "\\label{discussion} The results demonstrate that a fully automated neural network can accurately determine the three principal physical parameters from spectroscopic or photometric stellar data, something which has not previously been demonstrated. Moreover, this work has used spectra of considerably lower resolution than have been used before in automated classifiers. Even at low resolution (50--100\\,\\AA\\ FWHM) and SNR (5--10 per resolution element), neural networks can yield good determinations of \\teff\\ and \\met, and even for \\logg\\ for stars earlier than solar. Still lower resolutions permit good results provided the SNR is high enough ($\\geq 50$). That good \\teff\\ can be achieved even at low resolution and SNR is perhaps not surprising when we consider that the spectra have absolute fluxes, which will be the case with high precision parallax missions such as GAIA. However, the more distant objects will have lower precision parallaxes and hence errors in the {\\em mean} flux level. But even if we completely ignore distance information (and flux normalise the spectra), the shape of the spectrum is still a strong indicator of \\teff: For example, Bailer-Jones et al.\\ (\\cite{bailerjones_98a}) obtained an MK spectral type precision of 0.8 subtypes ($\\Delta\\log$\\teff=0.010--0.015) across a wide range of spectral types (B2--M7) using flux normalised spectra. This is similar to what can be achieved from broad band photometry, implying that \\teff\\ determination only requires very low resolution. The good performance of `high' resolution spectroscopy (25\\,\\AA) at very low SNR ($\\sqrt 5$ per pixel) was not expected. It seems to imply that for a given amount of integration time it may be better to sacrifice SNR for resolution. It is noteworthy that while the filters provide good \\teff, their ability to determine \\met\\ and especially \\logg\\ is very limited at low SNR. How do these results compare with classical parametrization methods? Gray (\\cite{gray_92a}) compiles results showing that with photometric errors below 0.01 magnitudes, the B$-$V colour calibrates \\teff\\ to 2--3\\% (4\\% for hotter stars) in the absence of reddening. Slightly better precision can be obtained from the slope of the Paschen continuum and size of the Balmer discontinuity. The latter may also be used to measure \\logg\\ to $\\pm 0.2$ dex. With spectra at a few \\AA\\ resolution over a similar wavelength range to that used here, Cacciari et al.\\ (\\cite{cacciari_87a}) obtained uncertainties in log~\\teff\\ and \\logg\\ of 0.01 and 0.04 respectively. Sinnerstad (\\cite {sinnerstad_80a}) made uvby,$\\beta$ photometric measurements of B stars, and for uncertainties of 0.005 in $\\beta$ and of 0.01 in u$-$b (i.e.\\ SNR $\\sim$ 200), infers errors in log~\\teff\\ and \\logg\\ of 0.004 and 0.08 respectively. These are similar to or slightly better than the results for similar stars in Tables~\\ref{restab1}--\\ref{restab3} ($\\epsilon_3$) at the highest resolutions. High resolution ($r \\leq 0.1$\\,\\AA)\\ spectra have generally been used to determine metallicity, and in a review, Cayrel de Strobel (\\cite{cayreldestrobel_85a}) notes that metallicity can be determined to $\\pm 0.07$ dex at SNR=250 (but only $\\pm 0.2$ dex at SNR=50) provided the effective temperature and gravity are approximately known. At lower SNR (10--20), Jones et al.\\ (\\cite{jones_96a}) could determine [Fe/H] to $\\pm 0.2$ dex for G stars using a set of spectroscopic indices measured at 1\\,\\AA\\ resolution in the range 4000--5000\\,\\AA, again using a known effective temperature. More recently, Katz et al.\\ (\\cite{katz_98a}) have used a minimum distance method to parametrize spectra by finding the closest matching template spectrum. The template grid consisted of 211 flux calibrated spectra (3900--6800\\,\\AA, $r \\simeq 0.1$\\,\\AA) with 4000\\,K\\,$\\leq$\\,\\teff\\,$\\leq$\\,6300\\,K, $-0.29 \\leq$\\,[Fe/H]\\,$\\leq +0.35$, and \\logg\\ for dwarfs and giants. The {\\em internal accuracy} of the method for log~\\teff, \\logg\\ and \\met\\ was 0.008, 0.28 dex, and 0.16 dex respectively at SNR=100, and 0.009, 0.29 dex and 0.17 dex at SNR=10. As expected, their results for \\logg\\ are much better than those in this paper at the similar temperature range ($\\epsilon_1$ in Table~\\ref{restab2}), presumably due to their much higher resolution. In contrast, their performance for \\met\\ is similar and for \\teff\\ somewhat worse than that in this paper at 500 times lower resolution. Their results also confirm that at high resolution a lower SNR leads to very little loss in performance. Snider et al.\\ (\\cite{snider_00a}) trained and tested neural networks on a set of 182 real F,G and K spectra over the range 3630--4890\\,\\AA\\ at intermediate resolution ($\\sim$1\\AA), and achieved 1$\\sigma$ errors in log~\\teff, \\logg\\ and \\met\\ of 0.015, 0.41 dex and 0.22 dex respectively, based on training and testing a network with a set of 182 real F,G and K spectra. When judging the relative values of the different resolution/SNR combinations in this paper, we must also take account of their implementation `costs', specifically the relative integration times required. Usually for a survey, a fixed total amount of integration time is available for all filters/spectra. In the case of GAIA -- which is continuously rotating -- a star moves across a focal plane covered with a mosaic of CCDs which are clocked at the rotation rate. The different filters are fixed to different CCDs, so that as a star moves across the mosaic it is recorded in different wavelength ranges. Thus less numerous and/or broader filters would achieve a higher SNR than more or narrower filters. Some filters could be replaced with a slitless spectrograph (e.g.\\ a prism or grism). This disperses every point on the sky and thus gives the full integration time for all wavelengths, but at the expense of increased sky noise and object confusion. These could be reduced by using one or more dichroics to redirect the light to two or more focal planes. (Confusion would be reduced further with GAIA by the fact that each area of sky is observed at many different position angles over the mission life.) An alternative approach is a set of many medium band filters ($\\sim 100$ for $r$=100\\,\\AA\\ over the complete wavelength range, although omission of some filters could be achieved). While this avoids the two principal disadvantages of the slitless spectrograph, the integration time per wavelength interval is dramatically reduced." }, "0003/astro-ph0003247_arXiv.txt": { "abstract": "Ultraviolet light curves are calculated for several thermonuclear supernova models using a multifrequency radiation hydrodynamic code. It is found that Chan\\-d\\-ra\\-se\\-khar-mass models produce very similar light curves both for detonation and deflagration. Sub-Chandrasekhar-mass models essentially differ from ``normal'' Chandrasekhar ones regarding behaviour of their UV fluxes. Differences in absolute brightness and in shape of light curves of thermonuclear supernovae could be detectable up to 300~Mpc with modern UV space telescopes. ", "introduction": "Early ultraviolet (UV) emission from Type Ia supernovae (SNe~Ia) is poorly known by now. Only a very few brightest events have been observed a decade ago with the International Ultraviolet Explorer (IUE). Although observational data were sometimes quite fascinating, as in the case of SN 1990N (Leibundgut~et~al. \\cite{leietal}, Jeffery~et~al. \\cite{jefetal}), an amount of observed UV light curves and spectra remained too small to reveal what is typical for the UV emission from SNe~Ia and how individual features of the explosion can be displayed. The Hubble Space Telescope (HST) has slightly improved the situation. More data of better quality were obtained. This allowed theorists to make a comparison between the predictions of explosion models and the observational results. Reproduction of supernova UV emission is a good test for an explosion model because this spectral region reflects more directly the distribution of \\element[][56]{Ni} synthesized during the explosion and the conditions in the exploding star. Several models were already used to fit the observed spectra. Analyses of early and late emission from SNe~Ia by Kirshner~et~al.~\\cite{kiretal}, Ruiz-Lapuente~et~al.~\\cite{rulaetal}, Nugent~et~al.~\\cite{nugetal} show that Chandra\\-se\\-khar-mass models DD4 (Woosley \\& Weaver~\\cite{wwmods}) and W7 (Nomoto et al.~\\cite{nthyo}) and sub-Chandrasekhar-mass helium detonation models (see Livne~\\cite{liv}; Livne \\& Glasner~\\cite{ligl}; Woosley \\& Weaver~\\cite{wwlowm}; H\\\"oflich \\& Khokhlov~\\cite{hoekho}) can reproduce some features of UV spectra of SNe~Ia quite well. In this Letter we calculate the light curves of the similar models and discuss how they differ from each other in several UV wavelength ranges. It is quite probable that more observational data will soon be available with the HST, and that the Far-Ultraviolet Spectroscopic Explorer (FUSE), operating at shorter wavelengths (Sembach~\\cite{sem}), will be able to obtain light curves and spectra of SNe~Ia in the range where they were not observed so far. The analysis proposed here can help to distinguish which mode of explosion is actually realized in the SNe~Ia. ", "conclusions": "The main results of our calculations are presented in Figs.~\\ref{iue},~\\ref{fuse}. The light curves of our models are shown in near and far UV ranges. The fluxes are plotted as they would be seen for supernovae at distance of 10~Mpc. Declared sensitivity of FUSE is $\\sim 3\\cdot 10^{-15}$~ergs~s${}^{-1}$~cm${}^{-2}$~\\AA${}^{-1}$, and sensitivity of HST (at working range of wavelengths almost equal to that of IUE) is roughly $10^{-16}$~ergs~s${}^{-1}$~cm${}^{-2}$~\\AA${}^{-1}$. This allows us to estimate that SNe~Ia could be observed up to 300~Mpc in the near UV and up to 30~Mpc in the far UV (with HST and FUSE, respectively). Yet it should be noticed that in the far UV SNe~Ia are bright enough only during several days, and their flux declines very quickly after the maximum light, so the probability of discovering for them is quite low, unless they are very close to us (a few Mpc or even less). \\begin{table} \\caption{Parameters of SN Ia models, rise time to the maximum of the UV light curves and UV fluxes at maximum light in FUSE and IUE ranges. Fluxes are calculated under the assumption that a supernova is at distance of 10~Mpc from the observer.} \\label{uvfluxes} \\begin{tabular}{lllll} \\hline \\hline Model & DD4 & W7 & LA4 & WD065 \\\\ \\hline $M_{\\rm WD}{}^{\\rm a}$ & 1.3861 & 1.3775 & 0.8678 & 0.6500 \\\\ $M_{{}^{56}{\\rm Ni}}{}^{\\rm a}$ & 0.63 & 0.60 & 0.47 & 0.05 \\\\ $E_{51}{}^{\\rm b}$ & 1.23 & 1.20 & 1.15 & 0.56 \\\\ \\hline \\multicolumn{5}{c}{FUSE 905--1187\\AA \\rule{0mm}{4mm}} \\\\ $t_{\\rm max}{}^{\\rm c}$ & 5.7 & 5.3 & 1.4 & 7.7 \\\\ $F_\\lambda{}^{\\rm d}$ % & 3.28 & 11.0 & 1.97 & $4.83 \\cdot 10^{-9}$ \\\\ \\hline \\multicolumn{5}{c}{SWP 1150--1975\\AA \\rule{0mm}{4mm}} \\\\ $t_{\\rm max}{}^{\\rm c}$ & 6.7 & 6.4 & 1.8 & 7.7 \\\\ $F_\\lambda{}^{\\rm d}$ % & 41.7 & 51.8 & 9.21 & $7.45 \\cdot 10^{-3}$ \\\\ \\hline \\multicolumn{5}{c}{LWPshort 1975--2500\\AA \\rule{0mm}{4mm}} \\\\ $t_{\\rm max}{}^{\\rm c}$ & 7.5 & 7.4 & 2.0 & 8.0 \\\\ $F_\\lambda{}^{\\rm d}$ % & 53.2 & 50.4 & 6.90 & $4.22 \\cdot 10^{-2}$ \\\\ \\hline \\multicolumn{5}{c}{LWPmiddle 2500--3000\\AA \\rule{0mm}{4mm}} \\\\ $t_{\\rm max}{}^{\\rm c}$ & 8.7 & 9.1 & 7.5 & 8.1 \\\\ $F_\\lambda{}^{\\rm d}$ % & 46.0 & 45.8 & 17.7 & 0.447 \\\\ \\hline \\multicolumn{5}{c}{LWPlong 3000--3500\\AA \\rule{0mm}{4mm}} \\\\ $t_{\\rm max}{}^{\\rm c}$ & 13.5 & 13.4 & 8.5 & 8.2 \\\\ $F_\\lambda{}^{\\rm d}$ % & 52.7 & 48.9 & 38.3 & 2.87 \\\\ \\hline \\multicolumn{5}{c}{Total IUE luminosity 1150--3300\\AA \\rule{0mm}{4mm}} \\\\ $t_{\\rm max}{}^{\\rm c}$ & 7.9 & 7.4 & 7.6 & 8.2 \\\\ $L{}^{\\rm e}$ % & 11.0 & 11.4 & 3.72 & 0.161 \\\\ \\hline \\multicolumn{5}{l}{${}^{\\rm a}$in $M_{\\sun}$} \\\\ \\multicolumn{5}{l}{${}^{\\rm b}$in $10^{51}$~ergs~s${}^{-1}$} \\\\ \\multicolumn{5}{l}{${}^{\\rm c}$in days} \\\\ \\multicolumn{5}{l}% {${}^{\\rm d}$in $10^{-14}$~ergs~s${}^{-1}$~cm${}^{-2}$~\\AA${}^{-1}$} \\\\ \\multicolumn{5}{l}{${}^{\\rm e}$in $10^{42}$~ergs~s${}^{-1}$} \\\\ \\end{tabular} \\end{table} \\begin{figure} \\resizebox{\\hsize}{\\hsize}{\\includegraphics{f1.ps}} \\caption{Near-UV light curves of the models DD4 (solid), W7 (dots), LA4 (long dash), and WD065 (dash-dot). Fluxes $F_\\lambda$ in units ~ergs~s${}^{-1}$~cm${}^{-2}$~\\AA${}^{-1}$ averaged in the four IUE spectral bands are reduced to distance of 10~Mpc.} \\label{iue} \\end{figure} \\begin{figure} \\resizebox{\\hsize}{\\hsize}{\\includegraphics{f2.ps}} \\caption{Far-UV light curves for the same models as in Fig.~\\protect\\ref{iue}, but in the FUSE working wavelength range (905--1187\\AA)} \\label{fuse} \\end{figure} Differences in the shapes and absolute brightness of SNe~Ia light curves become more pronounced in this spectral range, especially in far UV, for different modes of explosion. As we have found in our paper (Sorokina~et~al.~\\cite{sbb}), the light curves of W7 and DD4 are very similar in {\\it B} band close to the maximum light and differ drastically several days after it. We can see likely behaviour in near-UV wavelengths (Fig.~\\ref{iue}). In far UV, differences grow up, so that even at maximum light one can distinguish between two Chandrasekhar-mass models. Such a behaviour can most probably be explained by different distribution of \\element[][56]{Ni} inside debris of exploded star. In our case, the fraction of \\element[][56]{Ni} decreases more sharply in the model W7, and this perhaps leads to the faster decline of its light curve. The shape of the UV light curve of WD065 is conformal to those of Chandrasekhar-mass models, though it shows much lower absolute flux (due to an order-of-magnitude lower \\element[][56]{Ni} mass). The light curve maxima of the Chandrasekhar-mass models are shifted progressively towards later epochs for longer wavelengths remaining almost equal in brightness, while the WD065 maxima occur at nearly the same epoch for all of the IUE and FUSE ranges, and the emission virtually disappears at the shorter edge of the spectrum (see Table~\\ref{uvfluxes} and Fig.~\\ref{iue}), since such a small amount of \\element[][56]{Ni} as present in this model is not able to maintain high temperature inside the ejecta. The emission of WD065 in far UV becomes so weak that it could be detected only if supernova exploded in our neighbourhood (not farther than a few hundreds of parsecs from us). The model LA4 (helium detonation in outer layers) is apparently distinguished by its rise time in the shortest spectral bands. The most interesting feature of this light curve is its clear two-maxima structure in the short LWP range. The earliest spike of the far-UV radiation is due to the outer \\element[][56]{Ni} layer specific for this model. It is well known that those helium detonation models are too blue near visual maximum (H\\\"oflich \\& Khokhlov~\\cite{hoekho}; Ruiz-Lapuente~et~al.~\\cite{rulaproc}). This is also confirmed by our {\\it UBVRI} computations (Sorokina~et~al.~\\cite{sbb}). In far UV, LA4 looks out not so hot, yet it can be detected in far UV earlier than in visual light. One should be cautious applying our results directly to observations in UV range. A large fraction of SNe~Ia shows a significant correlation with star-forming regions (Bartunov~et~al. \\cite{btsf}; McMillan \\& Ciardullo~\\cite{mcmici}; Bartunov \\& Tsvetkov \\cite{batsv}). The circumstellar medium in those regions can absorb radiation, especially in UV band. In this case our predictions should be used as an input for calculations of reprocessing of UV photons to redder wavelengths. Certainly, more thorough investigation of the UV emission from SNe~Ia has to be done. It is still necessary to calculate UV light curves of wider range of SNe~Ia models and to predict their UV spectra. This work is worth doing because, as it is seen from this Letter, near-UV and far-UV observations with modern UV space telescopes, when combined with standard {\\it UBVRI} study, could be used as an efficient means to distinguish modes of explosion of thermonuclear supernovae leading us to better understanding of these phenomena." }, "0003/astro-ph0003467_arXiv.txt": { "abstract": "The triple-lined spectroscopic triple system HD 109648 has one of the shortest periods known for the outer orbit in a late-type triple, 120.5 days, and the ratio between the periods of the outer and inner orbits is small, 22:1. With such extreme values, this system should show orbital element variations over a timescale of about a decade. We have monitored the radial velocities of HD 109648 with the CfA Digital Speedometers for eight years, and have found evidence for modulation of some orbital elements. While we see no definite evidence for modulation of the inner binary eccentricity, we clearly observe variations in the inner and outer longitudes of periastron, as well as in the radial velocity amplitudes of the three components. The observational results, combined with numerical simulations, allow us to put constraints on the orientation of the orbits. ", "introduction": "The number of triple systems with well-determined orbital elements is still small (Fekel 1981; Tokovinin 1997, 1999). In particular, the number of spectroscopic triples in which the wide orbit is also known from radial-velocity observations is very small. Part of the problem is that the velocity amplitude of the outer binary is usually small compared to the amplitude of the inner binary. Moreover, after a binary orbit has been solved, the natural reaction is to discontinue observing it, and checks for longer-term variations are rarely made \\cite{may87}. This series of papers is aimed at increasing our knowledge of triples by investigating systems where the inner and outer orbits can both be determined from spectroscopic observations. The first paper of the series (Mazeh, Krymolowski and Latham 1993, hereinafter Paper I) examined the halo triple G38-13, while the second paper (Krymolowski and Mazeh 1998, hereinafter Paper II) derived an analytic technique which allows for fast simulation of orbital modulations of a binary induced by a third star. In the present paper we analyse the triple-lined spectroscopic triple system HD 109648 (HIP 61497, $\\alpha = 12^{\\rmn{h}}35^{\\rmn{m}}59 \\fs 8$, $\\delta = +36\\degr 15\\arcmin 30\\arcsec$ (J2000); $V = 8.8$). HD 109648 was identified as one (star 6) of a handful of stars belonging to the remnant of a nearby old open cluster, Upgren~1 \\cite{upg65}, but subsequent studies have weakened the interpretation that all of the stars originally identified are physically associated (Upgren, Philip \\& Beavers 1982; Gatewood et al.\\ 1992; Stefanik et al.\\ 1997; Baumgardt 1998). The triple-lined nature of HD 109648 was noticed soon after we began observing it, because the one-dimensional correlations of some of the spectra clearly showed three peaks. A periodicity analysis revealed periods at $\\sim 5.5$ and $\\sim 120$ days. Triple systems tend to be hierarchical, usually with a close binary and a more distant third star, as other configurations are generally unstable and are unlikely to persist and be detected. To first-order, a hierarchical triple system can be separated into an inner orbit (comprising the two close stars) and an outer orbit (comprising the third star and the centre-of-mass of the inner pair). This approximation is most valid when the distance to the third star far exceeds the separation between the inner two stars. One of the goals of this study is to investigate the interaction of these three stars (through the variation of the inner and outer orbits) to higher order. A preliminary version of this work was presented at a conference entitled `Thirty Years of Astronomy at the Van Vleck Observatory: A Meeting in Honor of Arthur R. Upgren' \\cite{jha97}. This paper updates the orbital solutions presented there and adds a significantly more detailed analysis of the system, partly through the use of numerical simulations. In Section 2 we summarise the analysis of the observations, including the derivation of the radial velocities, orbital solutions, and additional parameters such as the mass ratios and constraints on the orbital inclinations. We discuss the theoretically expected modulations of orbital elements in Section 3. In Section 4 we describe our efforts to search for such variations and present our results. Further constraints for the system via numerical simulation are derived in Section 5. Finally, in Section 6 we discuss our results and relate them to previous and future work. ", "conclusions": "We have shown that HD 109648 is a hierarchical triple system, with an outer period and outer:inner period ratio conducive to modulations of orbital elements on timescales of about a decade. Indeed, our observations clearly indicate an advance of the inner longitude of periastron corresponding roughly to this timescale. We also found strong evidence for variations of the radial velocity amplitudes of the inner orbit, most naturaly accounted for by the precession of the nodes. Furthermore, the inner eccentricity is small but significant, presumably due to the interaction with the outer star. Such effects have been predicted theoretically for hierarchical triples for a number of years. However, there have been few observational confirmations. Mayor and Mazeh \\shortcite{may87} have looked for evidence of the precession of nodes in a number of close binaries, and have reported several significant changes, based on observations made at two widely-spaced epochs. Mazeh and Shaham \\shortcite{maz76} also suggested a few systems where the effect may have had a role, but none of these were confirmed triples. The inner eccentricity modulation has been less conclusively observed. Mazeh and Shaham (1977; 1979) have postulated it to be the cause of long-period phenomena, such as episodic accretion, in some close binaries. More convincing evidence has been offered for the interaction of a third star with the tidal circularization of the inner binary. Mazeh \\shortcite{maz90} has looked for eccentric orbits in samples of short-period binaries that should have been circularized, as a fingerprint for a third star in the system. Three such examples were found, and the hypothesis of a triple system was confirmed in each case. In addition, one system (HD 144515) showed evidence for a variation in the inner eccentricity, again based on observations from two epochs. Ford, Kozinsky, and Rasio (1999) also provide instances of some other triple systems where these interactions may have played a role. HD 109648 provides the best observational evidence so far of these predicted modulations. This system is a confirmed hierarchical triple, a direct result of analysing the triple-lined spectra. Furthermore, the evidence for variations in the elements comes from observations in a homogeneous set of data, rather than relying on two-epoch observations. Finally, we see evidence both for the precession of the nodes and for the apsidal advance in the same system. More data are needed to strengthen this case. With better information on the variation of the elements with time, we should be able to derive better constraints on the orientation of the system. For example, if we are able to fit the variation of the inner and outer inclination angles, we can determine the various angles between the total angular momentum and its inner and outer binary components, as well as the angle between the total angular momentum and the line of sight. If the inner eccentricity modulation becomes clearer with additional data as well, it should provide stronger constraints on the relative inclination. In addition to continued spectroscopic observations, there may be some hope that interferometric observations of HD 109648 will be able to help clarify the orientation of the system. The Hipparcos parallax for HD 109648 of 4 mas may potentially be in error due to the outer orbit, particularly because the outer period is nearly commensurate with one year. Nevertheless, the separation of the third star from the inner binary is on the order of a few mas, allowing for the possibility of a visual orbit in the future. We can also hope that additional triple systems will be discovered, perhaps ones that are even better than HD 109648 for this type of study. Indeed, Saar, N\\\"{o}rdstrom \\& Andersen \\shortcite{saa90} have noted a promising system with a modulation timescale perhaps shorter even than that of HD 109648. Determining the geometry and orientation of such systems will be a great advance in our understanding of them." }, "0003/astro-ph0003184_arXiv.txt": { "abstract": "The twin peaks in the nucleus of M31 have been interpreted by Tremaine as a thick, eccentric, disk of stars orbiting a massive dark object; the required alignment of the apoapsides of the stellar orbits could be maintained by self--gravity, and the whole structure might be a discrete, nonlinear eigenmode. The pattern speed of this mode could, in principle, be determined by the Tremaine--Weinberg (TW) method, which requires measurements of the surface brightness, and radial velocity along a strip parallel to the line of nodes. However, spectroscopic observations along the line of nodes are not available. We propose a variant of the TW method, which exploits a basic feature of the eccentric disk model, to extract estimates of the pattern speed from {\\it Hubble Space Telescope} spectroscopic data, taken along the line joining the two peaks. Within limitations imposed by the data, we estimate that the pattern rotates in a prograde manner and, for an assumed disk inclination of $77^{\\circ}$, the pattern speed $\\left|\\Omega_p\\right| < 30\\;{\\rm km\\,s^{-1}\\,pc^{-1}}$, or period more than $200,000$~years. ", "introduction": "The nucleus of M31 was first resolved by the Stratoscope~II balloon--borne telescope (Light, Danielson, and Schwarzschild~1974), which showed that the peak brightness was displaced relative to the center, as inferred from the isophotes of the outer parts of the galaxy. This was confirmed, and extended by {\\it Hubble Space Telescope} ({\\it HST}) observations, which revealed two peaks in the brightness, separated by $0''.49\\,$~(Lauer et al.~1993, King, Stanford, and Crane~1995). Ground--based, as well as {\\it HST} spectroscopy, have probed the structure of the radial velocities and velocity dispersions, in increasing detail, along many strips across the nuclear region (Dressler and Richstone~1988, Kormendy~1988, Bacon et al.~1994, van der Marel et al.~1994, Gerssen, Kuijken, and Merrifield~1995, Statler et al.~1999, Kormendy and Bender~1999). These provide evidence for the presence of a massive dark object (MDO), which could be a supermassive black hole, of mass $\\sim 3\\times 10^7\\,{\\rm\\,M_\\odot}\\,$, located very close to the fainter peak (P2). The dynamical center of the nucleus is believed to coincide with the center of the isophotes of the bulge of M31; this point has been estimated to lie between the two peaks. Tremaine~(1995) proposed that the nucleus could be a thick eccentric disk, composed of stars on nearly Keplerian orbits around the MDO, with their apoapsides aligned in the direction toward the brighter peak (P1); the brightness of P1 is then explained as the increased concentration of stars, resulting from their slow speeds near their apoapsides. He also showed that this model is broadly consistent with the kinematics, as inferred from the spectroscopic observations of Kormendy~(1988), and Bacon et al.~(1994). Recent work has not only produced further support for Tremaine's model (Statler et al.~1999, Kormendy and Bender~1999), but has stimulated variations on the basic theme (Statler~1999). Tremaine also suggested that the alignment could be maintained by the self--gravity of the disk, wherein the eccentric distortion could arise as a discrete, nonlinear eigenmode, with some nonzero pattern speed ($\\Omega_p$), equal to the common apsidal precession rate. The dynamical question is yet to be resolved in a self--consistent manner, although explorations of orbits in model potentials have identified a family of resonant, aligned loop orbits, which could serve as building blocks (Sridhar and Touma~1999, Statler~1999); reasonably faithful reproduction of the nuclear rotation curve adds some degree of confidence (Statler~1999). If the nuclear disk is indeed a steadily rotating, nonlinear eigenmode, what is $\\Omega_p\\,$? Tremaine and Weinberg~(1984, hereafter TW) invented a method of estimating the pattern speed of a barred disk galaxy, that uses measurements of the surface brightness, and radial velocity along a strip parallel to the line of nodes (defined as the line of intersection of the disk and sky planes). That this methods works was proved when the pattern speed of the bar in NGC~936 was estimated by Merrifield and Kuijken~(1995). Unfortunately, the radial velocity measurements of the nucleus of M31 (see references above on spectroscopy) are available on strips that, either do not coincide with the line of nodes, or possess too poor an angular resolution for direct application of the TW method. In this Letter, we show that the {\\it HST} observations of Statler et al.~(1999, hereafter SKCJ), together with Tremaine's (1995) model, can be used to estimate $\\Omega_p\\,$. ", "conclusions": "Our estimate of the pattern speed of the nuclear disk of M31 should be qualified by a discussion of possible sources of errors, most of which are difficult to estimate quantitatively. SKCJ calibrate velocities relative to an average over an $8\\,{\\rm arcsec}^2$ aperture centered on the nucleus (Ho, Filippenko, and Sargent~1993), and assure us that the errors are likely to be small. More significant, perhaps, are the systematic errors in $\\radvel$, mentioned by SKCJ; these are shown as open squares in Figure~2a. SKCJ used a slit of width $0''.063$, and this will introduce contributions to $\\radvel$ from nonzero values of $v_x'$ (see equation~(\\ref{radvel})). This effect is somewhat mitigated by cancellation between positive and negative values of $v_x'$, and the fact that the width of $0''.063$ is of much smaller scale than the minor axis, projected onto the sky plane, of the smallest ellipse used by Tremaine~(1995; see Figure~2a of his paper) to represent the nuclear disk. The limits of integration are necessarily finite in numerical computation, and we have resisted the temptation to include corrections by extrapolation of the $\\Sigma_s$ and $\\radvel$ profiles. A basic assumption underlying the application of a TW--like method is that the surface brightness obeys a continuity equation, which would be valid for a stellar disk in the absence of star formation (or death). We expect the numbers of stars to be conserved, except possibly in the vicinity of P2, where the observed UV excess has been interpreted as contributions from early--type stars (King, Stanford and Crane~1995, Lauer et al.~1998). However, these stars do not contribute much to the photometric and kinematic data we have used. Tremaine~(1995) argues that two--body relaxation is expected to thicken the disk, whereas we assumed that the disk was razor--thin. The original TW method is applicable to thick disks, so long as the streaming velocity normal to the disk plane is zero. In addition to the assumption of zero normal streaming velocity, let us suppose that the three dimensional density, $\\rho$, is symmetric about the midplane of the disk. The contribution to $\\radvel$ from $v_x'$ arise from an integral along the (inclined) line of sight that runs through the thick disk. Consider two points along this line of sight that are equally displaced about the midplane of the disk. Aligned, nearly Keplerian orbits have flows such that $\\rho$ is equal, whereas $v_x'$ is equal and opposite at these two points; in this ideal picture, there is pair--wise perfect cancellation, and no net contribution to $\\radvel$ from $v_x'\\,$. In practice there should be some cancellation, because $v_x'$ will have opposite signs at two corresponding points, but there could be a net contamination from the unequal values of $|v_x'|$ and $\\rho$\\,. Tremaine's original model, which was a reasonable fit to the then available photometry and kinematics, considered a non rotating disk, and it would be appropriate to inquire about the implications of a non zero pattern speed. A pattern that is prograde with angular speed, say, of $20\\;{\\rm km\\,s^{-1}\\, pc^{-1}}$ would contribute only about $35\\;{\\rm km\\,s^{-1}}$ to the radial velocity at P1, which is about $250\\;{\\rm km\\,s^{-1}}\\,$, according to the measurements of SKCJ. The maximum radial velocity quoted by Tremaine~(1995) is less than $200\\;{\\rm km\\,s^{-1}}\\,$, so a non zero pattern speed could still be accomodated. Our estimates do not rule out a non rotating disk, but we would like to offer a physical argument in support of a non zero pattern speed. For the disk plus MDO to be in a steady, non rotating state, the gravitational force on the MDO should necessarily vanish. Our (unpublished) numerical computations indicate that the force is indeed non zero. A limitation of our method is that it uses, in an essential manner, the assumption that most of the contribution to $\\radvel$ comes from orbits which intersect the measurement strip at right angles. In comparison, the original TW method does not rely on assumptions about the geometry of the mean flow; averaging over several strips, all parallel to the line of nodes, will improve estimates of the pattern speed, as Merrifield and Kuijken~(1995) demonstrated. Thus it is necessary to verify our estimates of $\\Omega_p$, by using the TW method on future observations of $\\radvel$, together with better photometry such as Lauer et al.~(1998), along strips parallel to the line of nodes. An extremely useful set of observations that could be performed would be two--dimensional spectroscopy, similar to the work of Bacon et al.~(1994), with the increased angular resolution that should be available in the near future." }, "0003/astro-ph0003320_arXiv.txt": { "abstract": "We present the results of spectrophotometry and $BVR$ CCD photometry, as well as integrated H{\\sc i} radio measurements of a new blue compact galaxy (BCG) HS~0822+3542 with extremely low oxygen abundance: 12 + log(O/H) = 7.35, or 1/36 of solar value. The galaxy is the third most metal-deficient BCG after I~Zw~18 and SBS~0335--052. Its very high mass fraction of gas ($\\approx$ 95\\% of all visible mass) and blue colours of underlying nebulosity are also similar to those of SBS~0335--052. This suggests that HS~0822+3542 is one of the nearest and dimmest galaxies experiencing a recently-started first star formation (SF) episode. Its properties imply that for such galaxies there is a linear scaling of the main parameters, at least for the baryon mass range (0.3--20)$\\times$10$^{8}$ $M_{\\odot}$. The total mass estimate indicates that the galaxy is dynamically dominated by a dark matter (DM) halo, which itself is one of the least massive for galaxies. ", "introduction": "Since the Searle \\& Sargent (1972) paper identifying blue compact galaxies (BCGs), that is, low-mass galaxies showing emission line spectra characteristic of H{\\sc ii} regions, intense star formation (SF), and oxygen abundances of 1/50 -- 1/3 solar\\footnote{12+log(O/H)$_{\\odot}$ = 8.92 (Anders \\& Grevesse \\cite{Anders89}).}, such objects have been considered as young galaxies undergoing one of their first star formation bursts. I~Zw~18, a BCG with the lowest known oxygen abundance among the galaxies (O/H $\\sim$ 1/50 (O/H)$_{\\odot}$, Searle \\& Sargent \\cite{Searle72}; Izotov \\& Thuan \\cite{IT99}), has been suggested as a candidate to be a truly-local young galaxy, experiencing its first short SF episode. The second candidate young galaxy, SBS~0335--052E, with an oxygen abundance of 1/41 (O/H)$_{\\odot}$ (Melnick et al. \\cite{Melnick92}; Izotov et al. \\cite{Izotov97a}; Lipovetsky et al. \\cite{Lipovetsky99}) was discovered 18 years later by Izotov et al. (\\cite{Izotov90}). With only two probable examples, we must be extremely lucky to be witnessing local galaxy formation. The proximity of these probable young galaxies allows one to study their properties in detail and to set important constraints on models of galaxy formation. Such studies are important for understanding the nature of very faint and compact probable primeval galaxies at high redshifts. Most of such galaxies at $z = 3-5$ were discovered only recently (e.g. Steidel et al. \\cite{Steidel96}; Dey et al. \\cite{Dey98}), and it seems that the majority of them are already rather evolved systems. Moreover, the local candidate young galaxies are at least one order of magnitude less massive than the faintest candidate young galaxies at high redshifts, and represent the range of baryon mass (10$^{8}$--10$^{9}$ $M_{\\odot}$) within which possibly most of primeval galaxies have formed (e.g. Rees \\cite{Rees88}). Evidence for the existence of old low-mass stellar populations was obtained in the last 25 years for most of the studied BCGs (Thuan \\cite{Thuan83}; Loose \\& Thuan~\\cite{Loose86}). Moreover, no conclusive answer has been reached yet about the youth of the few most metal-poor BCG. However, some observational data have been collected lately, which apparently support young ages for these BCGs. Among them we point out: a) Extremely low abundances of heavy elements in H{\\sc ii} regions surrounding young clusters, consistent with theoretical expectations of ``metal'' yield during a first SF event ($Z$ $<$ 1/20 $Z_{\\odot}$) (e.g., Pilyugin \\cite{Pil93}); b) Very blue colours outside the location of the current SF burst, consistent with a lack of stars older than 100 Myr (Hunter \\& Thronson \\cite{Hunter95}; Papaderos et al. \\cite{Papa98}). While the recent analysis of HST data for I~Zw~18 by Aloisi et al. (\\cite{Aloisi99}) suggests an age of 1 Gyr for the underlying stellar population of the galaxy, Izotov et al. (\\cite{Izotov2000}) argue that a self-consistent treatment of all data favours a significantly larger distance to I~Zw~18 then adopted by Aloisi et al., and a 100 Myr stellar population; c) A large amount of neutral gas, making up 99\\% of all baryonic (luminous) mass (van Zee et al. \\cite{vanZee98}; Pustilnik et al. \\cite{Pus2000}); d) Practically zero metallicity for this H{\\sc i} gas, e.g., (O/H) $<$ 3$\\times$10$^{-5}$(O/H)$_{\\odot}$, as reported for SBS~0335--052E (Thuan \\& Izotov \\cite{TI97}). This emphasizes either an extremely slow evolution on these systems, or a very recent onset of metal production. The latter suggests that the neutral gas clouds in these galaxies are composed of pregalactic material not yet polluted by stellar nucleosynthesis products. It was suggested recently by Izotov \\& Thuan (\\cite{IT99}), from the analysis of carbon and nitrogen abundances, that several BCGs with O/H $<$ 1/20 (O/H)$_{\\odot}$ in H{\\sc ii} regions are good candidate galaxies with a recent first SF episode. Until now, less than ten such galaxies with good abundance determinations are known. Even though the point on the existence of truly young local galaxies is debatable (see, e.g., Kunth \\& \\\"Ostlin~\\cite{Kunth99}), the importance of studies of extremely metal-poor galaxies is undoubtful, since they best approximate the properties of primeval galaxies at large redshifts. In this paper we describe the data obtained for the third most metal-deficient galaxy, HS~0822+3542 with O/H = 1/36 (O/H)$_{\\odot}$. This is one of the {\\it nearest}, and at the same time the {\\it dimmest} candidate young galaxy known. \\begin{figure} \\psfig{figure=9558_f1.ps,angle=0,width=8.5cm,bbllx=70pt,bblly=300pt,bburx=520pt,bbury=750pt} \\caption{$R$-band image of HS~0822+3542. North is to the top, east is to the left. Low-contrast filaments NW of the main bright body indicate gas structures typical of high-velocity ejecta. At the adopted distance of 12.5 Mpc, 1$^{\\prime\\prime}$ = 60.6 pc. Surface brightness isophotes (white contours) are superposed on the central bright region to show the morphology of supergiant H {\\sc ii} region.} \\label{fig:HS_direct} \\end{figure} \\begin{table}[h] \\centering{ \\caption{Main parameters of the young galaxy candidates} \\label{tab:Main_par} \\begin{tabular}{lrrr} \\hline\\hline Parameter & 0822+3542$^{1}$ & IZw18 & 0335--052E \\\\ \\hline $B_{\\rm tot}$ &17.92$\\pm$0.07~~~ & 16.21$^{2}$ & 17.00$\\pm$0.02$^{3}$ \\\\ $(B-V)_{\\rm tot}$ &0.32$\\pm$0.08~~~ & 0.18$^{2}$ & 0.31$\\pm$0.05$^{3}$ \\\\ $(V-R)_{\\rm tot}$ &0.17$\\pm$0.09~~~ & 0.41$^{4}$ & 0.15$\\pm$0.05$^{3}$ \\\\ $V_{\\rm Hel}$ (km s$^{-1}$) & 732$\\pm$6~~~ & 751$\\pm$2$^{5}$ & 4043$\\pm$5$^{6}$ \\\\ $D_{\\rm Vir}$(Mpc) & 12.5~~~ & 14.2$^{1}$ & 52.8$^{6}$ \\\\ $E(B-V)$ & 0.047$^{11}$ & 0.032$^{11}$ & 0.047$^{11}$ \\\\ $M_{B}$$^{\\dag}$ & --12.7~~~ & --14.64~~ & --16.8$^{3}$ \\\\ Angular size (\\arcsec)$^{\\ddag}$ & 14.8$\\times$7.4~~~ & 22$\\times$15$^{1}$ & 14$\\times$10$^{3}$ \\\\ Optical size (kpc) & 0.90$\\times$0.45~~~& 1.5$\\times$1.0~ & 3.7$\\times$2.6$^{3}$ \\\\ 12+log(O/H) \\ & 7.35~~~ & 7.16$^{7}$ & 7.29$^{8}$~ \\\\ $T_{\\rm e}$(O {\\sc iii}) (K)\\ & 20,350~~~ & 19,600$^{7}$~& 19,300$^{8}$~ \\\\ H {\\sc i} flux$^{*}$ & 0.68$\\pm$0.07~~~ & 2.97$^{9}$ & 2.46$^{6}$ \\\\ $W_{50}$ (km s$^{-1}$) & 42$\\pm$5~~~ & 49$^{9}$ & 83$^{5}$ \\\\ $W_{20}$ (km s$^{-1}$) & 58$\\pm$8~~~ & 84$^{9}$ & 105$^{5}$ \\\\ $M$(H{\\sc i}) (10$^{8}$$M_{\\odot}$) & 0.24~~~ & 1.41$^{9}$ & 16.2$^{6}$ \\\\ $M$(H{\\sc i})/$L_{B}$$^{**}$ & 1.40~~~ & 1.40~ & 2.3~~ \\\\ SFR ($M_{\\odot}$/year) & 0.007~~~ & 0.04$^{10}$ & 0.4$^{10}$ \\\\ \\hline\\hline \\multicolumn{4}{l}{$B_{\\rm tot}$ -- total blue magnitude; $M_{B}$ -- absolute blue mag.} \\\\ \\multicolumn{4}{l}{$L_{B}$ -- total blue luminosity. $^{*}$\\ Units of Jy km s$^{-1}$;}\\\\ \\multicolumn{4}{l}{$^{**}$ In units of ($M$/$L_{B}$)$_{\\odot}$; $^{\\dag}$ With the Galaxy extinction} \\\\ \\multicolumn{4}{l}{~$A_{B}$ = 0.20, 0.14, 0.20, respectively, } \\\\ \\multicolumn{4}{l}{{corresponding to $E(B-V)$ in the previous line};} \\\\ \\multicolumn{4}{l}{$^{\\ddag}$\\ $a \\times b$ at surface brightness $\\mu_{B}$=25 mag arcsec$^{-2}$.} \\\\ \\multicolumn{4}{l}{{\\bf References}: $^1$This paper; $^{2}$Mazzarella \\& Boroson (\\cite{Maz93});} \\\\ \\multicolumn{4}{l}{ $^{3}$Papaderos et al. (\\cite{Papa98}); $^{4}$Huchra (\\cite{Huchra77}); $^{5}$Thuan et al.}\\\\ \\multicolumn{4}{l}{(\\cite{TLMP99}); $^{6}$Pustilnik et al. (\\cite{Pus2000}); $^{7}$Izotov \\& Thuan (\\cite{IT98});} \\\\ \\multicolumn{4}{l}{$^{8}$Izotov \\& Thuan (\\cite{IT99}); $^{9}$van Zee et al. (\\cite{vanZee98});} \\\\ \\multicolumn{4}{l}{$^{10}$ Thuan et al. (\\cite{TIL97}); $^{11}$ Schlegel et al. (\\cite{Schlegel98}) } \\\\ \\end{tabular} } \\end{table} ", "conclusions": "From the data and discussion above we reach the following conclusions: \\begin{enumerate} \\item HS~0822+3542 is a new nearby ($D$ = 12.5 Mpc) galaxy with oxygen abundance 12 + log(O/H) = 7.35. After I~Zw~18 and SBS~0335--052 this is the third lowest metallicity object among Blue Compact Galaxies. \\item Its very low metallicity, very small stellar mass fraction (0.05 relative to the entire baryon mass) and blue colours of the underlying disc [($B-V$)$_0$ = 0\\fm05, ($V-R$)$_0$ = 0\\fm26] imply that this is one of the few candidates to be a local young galaxy, forming its first generation of stars. \\item HS~0822+3542 is 50--60 times less luminous and massive than another candidate young galaxy SBS~0335--052. This implies a broad range of global parameters for the candidate young galaxies. A linear scaling between several important parameters of such galaxies probably exists, including parameters related to the SF burst. \\item The dynamical mass estimate using the width of the H{\\sc i} profile and a typical H{\\sc i} gas extent relative to the optical size, leads to the conclusion that HS~0822+3542 is dynamically dominated by a dark matter halo. \\item Higher S/N long-slit spectra than presented here, and deep H$\\alpha$ images are needed to follow the ionized gas extent. Resolved H{\\sc i} maps will be very helpful to study the dynamics of its ISM and the parameters of its DM halo. \\end{enumerate}" }, "0003/astro-ph0003116_arXiv.txt": { "abstract": "We present $K-$band spectra of more than 110 M giants in Galactic bulge fields interior to $-$4 degrees and as close as 0.2 degrees of the Galactic Center. From the equivalent widths of three features in these spectra, EW(Na), EW(Ca), and EW(CO) we calculate [Fe/H] for the stars with a calibration derived from globular clusters \\citep{ste00}. The mean [Fe/H] for each field is in good agreement with the results from \\citet{fro99} based on the slope of the giant branch method. We find no evidence for a metallicity gradient along the minor or major axes of the inner bulge ($R < $ 0.6 kpc). A metallicity gradient along the minor axis, found earlier, arises when fields located at larger galactic radius are included. However, these more distant fields are located outside of the infrared bulge defined by the COBE/DIRBE observations. We compute the [Fe/H] distribution for the inner bulge and find a mean value of -0.21 dex with a full width dispersion of 0.30 dex, close to the values found for Baade's Window (BW) by \\citet{sad96} and to a theoretical prediction for a bulge formed by dissipative collapse \\citep{mol00}. ", "introduction": "Baade's Window (BW, $l=1^{\\circ}, b=-4^{\\circ}$) is the most studied region in the Galactic bulge. In the late 1970s, \\citet{whi78} demonstrated that the integrated optical spectrum from BW closely resembles that from bulges of spiral galaxies and from moderate luminosity E and S0 galaxies. At the same time \\citet{fro78} found that the near infrared light from these galaxies is dominated by cool giant stars. Shortly thereafter, Blanco and his collaborators determined that BW contains an unusually high percentage of middle and late type M giants compared to other regions of the Galaxy \\citep{bla84}. Detailed studies of the M giants in BW revealed that they have photometric and spectroscopic properties significantly different from those of M giants in the field \\citep{fro87,fro88,fro90,ric83,ric88,ter90,mcw94}. The accurate determination of stellar metallicities is essential for constraining models of star formation and chemical evolution in the bulge. \\citet{fro90} used $JHK$ colors and CO and {$\\rm H_{2}O$} photometric indices to determine metallicities of stars at latitudes between $-3^{\\circ}$ and $-12^{\\circ}$ along the minor axis. \\citet{tie95} used the relation between the slope of the upper giant branch and [Fe/H] \\citep{kuc95a,kuc95b} to estimate metallicities for the same stars studied by \\citet{fro90}. \\citet{tys91} used Washington photometry on stars in similar fields for the same purpose. These three studies agreed that there is a small metallicity gradient along the minor axis of the bulge, with values ranging from -0.04 dex/deg \\citep{fro90} to -0.09 dex/deg \\citep{tys91}. Finally, \\citet{min95} discussed evidence for a metallicity gradient in the Galactic bulge based on compiled observations of ten fields, eight of them exterior to BW. Stars in most of the inner bulge ($ |b| \\leq 3^{\\circ}$) can be studied only with near infrared observations because of high reddening and extinction. Recently, \\citet{fro99} have studied 11 fields in the inner Galactic bulge using $JHK$ photometry. Seven of these fields are on the minor axis; five are at a latitude of $-1.3^{\\circ}$ parallel to the major axis. They estimated the reddening of each field from their CMDs and the mean metallicity of each field with the giant branch slope method. They combined their results with those of \\citet{tie95} and derived a gradient of -0.064 $\\pm$ 0.012 dex/degree in the range $ -0.2^{\\circ} \\leq b \\leq -10.25^{\\circ}$ along the minor axis. Our main goal is to obtain independent values for the metallicity of the stars in the same inner bulge fields studied by \\citet{fro99} but with spectroscopic techniques. We use the strength of three absorption features present in the $K-$band of cool stars: Na, Ca, and CO. The calibration is based on similar observations of giants in globular clusters by \\citet{ste00}. ", "conclusions": "We present $K-$band spectra of giant stars in fields interior to $-$4 degrees and as close as 0.2 degrees of the Galactic Center. We measure equivalent widths of the strongest features present in the $K-$band spectra, EW(Na), EW(Ca), and EW(CO), and also dereddened photometric indices $M_{K_{0}}$ and $(J-K)_{0}$. We use these indices to compute [Fe/H] for the individual stars, using the calibration derived for globular clusters by \\citet{ste00}. The mean [Fe/H] for each field is in good agreement with the results obtained with the slope of the giant branch method \\citep{fro99}. We find no evidence for a metallicity gradient along the minor or major axis of the bulge for $ R < $ 0.6 kpc. We also show that metallicity gradients found in earlier works only arise when fields located at larger galactic radii are included. Those higher galactic radii fields are located outside the infrared bulge defined by the COBE/DIRBE outline. We compute the [Fe/H] distribution for the inner bulge, finding a mean value of -0.21 dex with a full width dispersion of 0.30 dex, which are very similar to the mean and width of the BW's [Fe/H] distribution from \\citet{sad96} and to the theoretical distribution of a bulge formed by dissipative collapse \\citep{mol00}." }, "0003/astro-ph0003446_arXiv.txt": { "abstract": "High resolution FUV echelle spectra showing absorption features arising from \\CI\\ and CO gas in the \\bp\\ circumstellar (CS) disk were obtained on 1997 December 6 and 19 using the Space Telescope Imaging Spectrograph (STIS). An unsaturated spin-forbidden line of \\CI\\ at 1613.376 \\AA\\ not previously seen in spectra of \\bp\\ was detected, allowing for an improved determination of the column density of \\CI\\ at zero velocity relative to the star (the stable component), \\NP$~=~(2 - 4) \\times 10^{16}$ cm$^{-2}$. Variable components with multiple velocities, which are the signatures of infalling bodies in the \\bp\\ CS disk, are observed in the \\CI\\ $\\lambda\\lambda$1561 and 1657 multiplets. Also seen for the first time were two lines arising from the metastable $^{1}$D level of carbon, at 1931 \\AA\\ and 1463 \\AA. The results of analysis of the CO $A-X$ (0-0), (1-0), and (2-0) bands are presented, including the bands arising from $^{13}$CO, with much better precision than has previously been possible due to the very high resolution provided by the STIS echelle gratings. Only stable CO gas is observed, with a column density $N(\\rm{CO})~=~(6.3 \\pm 0.3) \\times 10^{14} \\ \\rm{cm}^{-2}$. An unusual ratio of the column densities of $^{12}$CO to $^{13}$CO is found ($R$~=~$15 \\pm 2$). The large difference between the column densities of \\CI\\ and CO indicates that photodissociation of CO is not the primary source of \\CI\\ gas in the disk, contrary to previous suggestion. ", "introduction": "\\bp\\ is the most extensively studied of the young planetary systems discovered in the last decade and a half. It is a bright Southern hemisphere star (type A5 V), located about 19.3 pc distant from the Sun, with a systemic radial velocity of 20 km s$^{-1}$ (for a review of the \\bp\\ system, see \\citet{vid98}). It was observed in 1983 by the IRAS satellite to have a large excess of emission at infrared wavelengths. This was referred to as the Vega-like phenomenon and was identified as arising from an edge-on circumstellar (hereafter CS) dust disk, presumed to be associated with planetary formation \\citep{smi84}. It was soon determined through absorption spectroscopy that there was CS gas associated with the dust as well. A large body of evidence has accumulated indicating that there are comet-like bodies present in the \\bp\\ CS disk. Collisions between dust particles are expected to produce submicron fragments which should be expelled from the system by radiation pressure on time scales much shorter than any plausible stellar age. Thus there must be a secondary source of particles; one model for the production of dust and gas in the CS disk focusses on evaporating comets and is called the Orbiting Evaporating Bodies model (OEB). In this picture, the comets orbit the star at several tens of AU, and thus, the \\bp\\ CS disk is a kind of ``gigantic multi-cometary tail with its natural constituents: gas and dust'' \\citep{lec96}. Spectra of \\bp\\ show variable redshifted absorption features arising from gas infalling toward the star at high velocities (and infrequently, blueshifted features as well); these features are best attributed to the evaporation of star-grazing comets, called the Falling Evaporating Bodies scenario (FEB) \\citep{beu90}. Also, gas at close to 20 km s$^{-1}$ (zero radial velocity relative to the star) is identified in all observations and is called the stable gas component. This gas is difficult to understand, as modeling indicates that it should be expelled from the system by radiation pressure; a continuous source for this gas is required. Neutral carbon and carbon monoxide have been observed in HST-GHRS UV absorption spectra of \\bp; carbon monoxide is the only molecule detected in the CS disk to date \\citep{vid94, jol98}. Since CO and \\CI\\ can be dissociated and ionized by interstellar UV photons on time scales of the order of 200 years, both must be continuously replenished. Carbon monoxide, in particular, is difficult to reform after dissociation in the \\bp\\ environment. Thus, the presence of these species indicates that a secondary source for this gas should exist, just as for the CS dust \\citep{vid94}. \\citet{jol98} found the column densities of \\CI\\ and CO to be comparable, around $10^{15}$ cm$^{-2}$; since their rates of destruction are also comparable, this was taken as evidence that the \\CI\\ is produced by photodissociation of CO, which evaporates from comets orbiting at various distances and velocities. However, the \\CI\\ column density was determined from a heavily saturated multiplet and is therefore quite uncertain. In the hopes of further constraining the characteristics of \\CI\\ and CO in the \\bp\\ system, we have reinvestigated the transitions observed in the GHRS data, as well as some that were not seen due to the relatively low spectral resolution of GHRS compared with that of the STIS high resolution echelle. ", "conclusions": "The very high resolution, low scattered light contamination, and good order separation of this STIS echelle data set has provided some clear advantages over previous observations of \\bp. The rotational lines of CO have been resolved, allowing for a much more precise determination of the physical parameters of the gas. The column density of CO is $N(\\rm{CO})~=~(6.3 \\pm 0.3) \\times 10^{14} \\ \\rm{cm}^{-2}$ and the ratio $R(^{12}\\rm{CO}/^{13}\\rm{CO})~=~15 \\pm 2$ is found. The absence of transient red or blueshifted components in the high resolution CO spectra supports the suggestion that this gas evaporates from cometary bodies orbiting far (several tens of AU) from the star. But the fact that the column density of CO is only about 2\\% of the total column density of \\CI\\ in the $^{3}$P ground term implies that photodissociation of this CO is not the primary source for \\CI\\ gas. It could perhaps be produced directly from infalling comets close to the star, but the mechanism by which it comes to zero velocity relative to the star and accumulates before being photoionized is unclear (although see \\citet{lag98}). The \\CID\\ gas may not have a stable component at 20 km s$^{-1}$; this unique species could prove to be a valuable tracer of FEB activity in the \\bp\\ CS disk. Despite the advantages of this data set, our lack of success in modeling the heavily saturated \\CI\\ multiplets indicates that in order to really determine the characteristics of the variable components of the \\CI\\ gas, we need to observe an unsaturated line or multiplet, with an oscillator strength between that of the 1561 \\AA\\ multiplet and that of the spin-forbidden 1613.376 \\AA\\ line. A number of suitable multiplets and lines lie in the FUV, shortward of $\\sim$ 1300 \\AA. Also, an unsaturated line arising from the $^{1}$D level would allow us to confirm the velocity structure in this gas and to determine if the velocities of the $^{1}$D gas components correspond with any of the velocities of the variable components in the $^{3}$P gas. Three likely lines lie between 1311 \\AA\\ and 1359 \\AA. Measurement of the densities of \\ion{C}{2} and \\ion{O}{1} would greatly help to unravel the carbon chemistry of the \\bp\\ disk. Again, potentially useful multiplets of these species lie in the FUV below 1340 \\AA. Thus, although this data set has vastly increased our knowledge about the important species \\CI\\ and CO in the \\bp\\ disk, our understanding would probably benefit greatly from investigation of \\bp\\ at shorter ultraviolet wavelengths." }, "0003/astro-ph0003452_arXiv.txt": { "abstract": "name}{} \\renewcommand{\\figurename}{Fig.} \\newcommand{\\grad}{\\mathop{\\rm grad}\\nolimits} \\def\\dfrac#1#2{{\\displaystyle#1\\over\\displaystyle#2}} \\def\\slantfrac#1#2{\\hbox{$\\,^{#1}\\!/_{#2}$}} \\def\\apgt{\\ {\\raise-.5ex\\hbox{$\\buildrel>\\over\\sim$}}\\ } \\def\\aplt{\\ {\\raise-.5ex\\hbox{$\\buildrel<\\over\\sim$}}\\ } \\newcommand{\\Msyr}{$M_\\odot~\\mbox{yr}^{-1}$} \\def\\Ms{M_{\\odot}} \\def\\Porb{P_{orb}} \\def\\GWR{{\\mbox{GWR}}} \\def\\MSW{{\\mbox{MSW}}} \\def\\LOSS{{\\mbox{LOSS}}} \\def\\EXCH{{\\mbox{EXCH}}} \\newcommand{\\myv}{\\vphantom{\\frac11}} \\newcommand{\\diver}{\\mathop{\\rm div}\\nolimits} \\newcommand{\\dd}{\\mathop{\\rm ~d}\\nolimits} \\newcommand{\\ddd}{\\mathop{\\rm \\raisebox{0pt}{d}}\\nolimits} \\begin{document} \\language=0 \\title{Non-conservative Evolution\\\\ of Cataclysmic Variables} \\author{ A.V.Fedorova$^1$, D.V.Bisikalo$^1$, A.A.Boyarchuk$^1$,\\\\ O.A.Kuznetsov$^2$, A.V.Tutukov$^1$, L.R.Yungelson$^1$\\\\[0.3cm] $^1$ {\\it Institute of Astronomy of the Russian Acad. of Sci., Moscow}\\\\ {\\sf afed@inasan.rssi.ru; bisikalo@inasan.rssi.ru; aboyar@inasan.rssi.ru;}\\\\ {\\sf atutukov@inasan.rssi.ru; lry@inasan.rssi.ru}\\\\[0.3cm] $^2$ {\\it Keldysh Institute of Applied Mathematics, Moscow}\\\\ {\\sf kuznecov@spp.keldysh.ru}\\\\[0.3cm] } \\date{} {\\large\\bf Abstract}---We suggest a new mechanism to account for the loss of angular momentum in binaries with non-conservative mass exchange. It is shown that in some cases the loss of matter can result in increase of the orbital angular momentum of a binary. If included into consideration in evolutionary calculations, this mechanism appreciably extends the range of mass ratios of components for which mass exchange in binaries is stable. It becomes possible to explain the existence of some observed cataclysmic binaries with high donor/accretor mass ratio, which was prohibited in conservative evolution models. ", "introduction": "Cataclysmic variables (CVs) are close binary systems consisting of a low-mass main-sequence (MS) star that fills its Roche lobe and a white dwarf. Main-sequence star (mass-donating star, usually referred to as star 2 or secondary) loses matter through the vicinity of the inner Lagrangian point $L_1$. White dwarf (accretor star, referred to as star 1 or primary) accretes at least a fraction of this matter via accretion disk or via accretion columns in the polar zones in the case when white dwarf has a strong magnetic field. Physics and evolution of CVs as well as evolution of similar to them low-mass X-ray binaries were investigated starting from the late sixties (see, e.g., [\\ref{RJW82}--\\ref{fed97}] and references therein). These studies were motivated by the recognition of the fact that evolution of CVs is determined by losses of orbital angular momentum due to radiation of gravitational waves [\\ref{kmg62},\\ref{p67}] and magnetically coupled stellar wind [\\ref{sch62}--\\ref{vz81}]. A number of studies investigated the influence of the loss of angular momentum associated with the material outflow from the system on the CV evolution (see, e.g., [\\ref{yu73}--\\ref{kk95}]). However, in the absence of gas dynamical simulations of the mass transfer in the binaries, these processes were considered under the parametric approach. Recent three-dimensional (3D) gas dynamical simulations of the structure of gaseous flows in semi-detached binaries proved an important role of a circumbinary envelope (see, e.g. [\\ref{bbkc98b},\\ref{bbkc98c}]). These calculations also show that during mass exchange a significant fraction of matter leaves the system. The present work is mainly devoted to the numerical investigation of the evolution of CVs using the data of 3D gas dynamical calculations on the losses of mass and angular momentum from close binaries. The main attention is paid to the stability of mass exchange against runaway mass loss and its dependence on the donor to accretor mass ratio $q=M_2/M_1$. It is known that the loss of matter by the donor star results in the violation of its hydrostatic and thermal equilibrium. Hydrostatic equilibrium is restored adiabatically, i.e. in dynamical time scale, while thermal equilibrium is restored in thermal time scale. This transition to a new equilibrium state leads to a change of donor radius. The sign of radius variation depends on the convective and radiative stability of outer envelope of the star. For stars with masses $M\\aplt \\Ms$ with deep convective envelope as well as for white dwarfs mass loss results in increase of the radius, while for stars with radiative envelopes it leads to the shrinkage of the star. Mass exchange in close binary is unstable when in the course of evolution mass-losing star tends to overfill the Roche lobe. It can occur when radius of the donor $R_2$ increases faster (or decreases slower) than the effective radius of Roche lobe $R_{RL}$.\\footnote{The effective radius of the Roche lobe $R_{RL}$ is determined as the radius of a sphere with a volume equal to the volume of the Roche lobe.} The case when the radius of the donor increases while the radius of the Roche lobe decreases is also possible. Thus, the question of stability of mass exchange against runaway is determined by the balance of derivatives $\\partial R_2/\\partial M_2$\\ and $\\partial R_{RL}/\\partial M_2$. It means that unstable mass exchange is possible even for stars which contract when losing matter. The study of conditions of stable mass exchange is motivated by two related problems: i) it is necessary to explain why $\\sim$ 10\\% of CVs with well-determined masses of components have combination of donor mass and mass ratio of components which is ``forbidden'' in the evolutionary models based on conventional assumptions on the conservation of angular momentum of a binary; ii) for population synthesis studies of CVs it is necessary to distinguish progenitors of CVs among all binaries containing a white dwarf and a low-mass companion (i.e., specify in which binaries of this type stable mass exchange is possible). A problem similar to the last one appears in the studies of low-mass X-ray binaries. The paper is organized as follows. In Section~2 we describe major factors affecting the evolution of CVs and conditions of stable mass exchange in binaries. The reasons which motivated us to rule out the conservative approximation for mass exchange are considered in Section~3. Sections~4 and~5 describe some results of 3D gas dynamical simulations of the flow structure in semi-detached binaries and introduce the model for description of angular momentum losses due to non-conservative mass exchange. Results of evolutionary calculations for conservative and ``non-conservative'' models are compared in Section~6. Criteria of stable mass exchange for ``non-conservative'' model with mass and angular momentum loss are determined in Section~7. Section~8 summarizes the main results of the work. ", "conclusions": "Results of three-dimensional gas dynamical simulations of the flow structure in semi-detached binaries show that during mass exchange a significant fraction of the matter leaves the system. On the other hand, a number of observed CVs have a combination of donor mass and mass ratio of components forbidden in the evolutionary models based on conventional assumptions on the loss of angular momentum of binaries, since in these cases mass exchange should be unstable. This contradiction can be resolved in the framework of the model suggested here. This model takes into account losses of mass and angular momentum from the system according to the results of gas dynamical simulations. It is shown that approximation for outflowing angular momentum as the difference of specific angular momentum of the donor matter in $L_1$ and specific angular momentum of accreted matter (being multiplied by appropriate mass flows) permits to explain observations satisfactorily. Note, that the new ``non-conservative'' model practically does not change such parameters of evolutionary tracks as limits of period gap and minimal period. Thus, the main conclusion of our study is: if we use a ``non-conservative'' approximation for CV evolution, we obtain a new boundary of the region of stable mass exchange in $q-{M_2}$ plane which explains the distribution of observed CVs better than conventional models." }, "0003/astro-ph0003044_arXiv.txt": { "abstract": "The theory of stellar structure can be used to identify the most plausible mechanisms for the irradiance variations associated with the solar cycle. Changes in surface emissivity, i.e. the reduced cooling in spots and enhanced emission by small scale magnetic fields, are the most effective mechanisms and account for most of the observed variation. Helioseismology will soon be able to test the consequences of changes in surface emissivity, and distinguish them from other scenarios for irradiance variability. \\keywords Sun: irradiance, luminosity radius, sunspots, magnetic fields ", "introduction": "When a sunspot group appears on the surface of the Sun, it reduce the brightness locally. Does this lead to an actual reduction of the Sun's brightness as measured on earth (the so-called irradiance)? Or is the `missing flux' re-emitted elsewhere on the solar surface? Faculae are bright areas, visible especially near the limb of the Sun. Do they contribute to the irradiance? Does the irradiance vary only because of spots and faculae, or are there other effects as well, and can these be large enough to influence climate on earth? Some of these questions have been answered by actual measurements with space-base radiometers over the past 20 years (figure 1). The data show a clear variation, obviously correlated with the solar cycle: the solar magnetic field does indeed cause the Sun's brightness to vary a bit, by about 0.1\\% between minimum and maximum of the cycle. The record shows prominent sharp dips; these are associated with the passage of individual sunspot groups across the disk. When a large group appears, the Sun actually does become dimmer by a tiny bit. The smoothed data (thick line in the figure) shows, however, that the {\\it average} trend is the opposite: at maximum activity, the Sun is brighter. The level of variation seen in figure 1 is not large, and by general opinion in the climate modeling community, insufficient to cause variations in the earth's climate, by a factor of 10 or so (see the contributions in Nesme-Ribes et al., 1994, Friis-Christensen et al. 2000). Even if the Sun were to settle into a state of activity as seen during the minima in figure 1, the average reduction of irradiance by 0.05\\% would have negligible effects, and probably would not explain events like the so-called `little ace age'\\footnote{A period during the 17th century when the climate in Europe was colder by a degree C or so. This cold spell appears to have been limited mostly to Western Europe, however.}. \\begin{figure} \\centerline{\\psfig{file=frohlich.ps,width=0.71\\textwidth,angle=90,clip=}} \\caption{Solar irradiance record, showing the variation due to the solar cycle. From Fr\\\"ohlich and Lean, 1998 (see also {\\tt http://www.pmodwrc.ch/solar\\_irr/sola\\_irr.html})} \\end{figure} For climate modeling, the record of figure 1 is too short. One would like to know what the irradiance has been in the past, during the last 10000 years, for example, so comparisons can be made with climate records such as have been obtained from ice cores in Greenland and the Antarctic (see the contributions Friis-Christensen et al. 2000). Is it conceivable, for example, that figure 1 shows only a fraction of what the solar magnetic field is capable of doing, and that variations of, say, 10 time larger amplitude could have been caused by the solar magnetic field, or something else that has not yet been discovered on the Sun? These are questions that can excite the imagination, including that of funding agencies. In the absence of reliable independent indicators of the past brightness of the Sun, it is of some importance to understand how the variation seen in figure 1 come about. If theoretical understanding is sufficient, and accepted by the community, it may be possible to assess with more certainty whether irradiance variations larger than those seen so far are likely to have occurred in the past, and perhaps influence the climate. Until the advent of helioseismology (see Dziembowski, elsewhere in this volume), our knowledge of the internal structure of the Sun was based on the theory of stellar structure and evolution. Since this theory is reasonably simple, confidence in the models was high even before helioseismological tests became possible. The main uncertainty had to do with the exact values of the opacity of stellar plasma. Increasingly accurate values for this were computed from first principles (atomic physics) by extensive numerical work. The basic correctness of the theory was then confirmed by helioseismology, to such an extent even that one is now confident that the `solar neutrino problem' has its cause in neutrino physics, not inaccuracies in the solar models (e.g. Bahcall 1996, Richard et al. 1996, Brun et al. 1998). Though these solar models are spherical and only slowly evolving in time, the theory behind them also apply to time dependent and non-spherical disturbances. In this form, the theory makes definite predictions about the thermal behavior of the Sun, that is, the evolution in time and position in the star of perturbations that might be related to the observed irradiance variations. In the following I present my view how this theory can be used to locate the most plausible source of irradiance variations. After this, I discuss how the new heliosmological measurements currently being made can be used to test the predictions of proposed mechanisms of irradiance variation. Conceptually, one can separate the thermal effects of magnetic fields into three types: 1. `Sources and sinks'. The generation of a magnetic field involves the conversion of energy of motion into magnetic energy. Since the motions in the solar envelope are thermally driven, this ultimately means conversion of thermal into magnetic energy: building up a magnetic field produces a thermal sink somewhere. The opposite happens when the field decays: magnetic energy is converted into heat. These thermal effects exist only during {\\it changes} in the magnetic energy content of the envelope. 2. `Shadows', or changes in the heat transport coefficient. Magnetic fields interfere with convection, causing a reduction in the efficiency of heat transport in the envelope. In contrast to (1), these changes last as long as the magnetic field itself is present. Both these sources and sinks, and changes in the transport coefficient cause thermal perturbations, varying with magnetic activity, which propagate through the envelope and cause variations in surface energy flux. Related to the second class of per\\-tur\\-ba\\-tions are: 3. The effects of magnetic fields at the surface of the star. Sunspots, being dark, radiate less than the surrounding photosphere, while the small elements that make up plages and the network have an excess emission. In addition, it is conceivable that the magnetic elements have an indirect effect surface by modifying the convective flow in their surroundings slightly (for which there is some observational evidence, see section \\ref{magmod}). ", "conclusions": "" }, "0003/gr-qc0003003_arXiv.txt": { "abstract": "Using approximate techniques we study the final moments of the collision of two (individually non-spinnning) black holes which inspiral into each other. The approximation is based on treating the whole space-time as a single distorted black hole. We obtain estimates for the radiated energy, angular momentum and waveforms for the gravitational waves produced in such a collision. The results can be of interest for analyzing the data that will be forthcoming from gravitational wave interferometric detectors, like the LIGO, GEO, LISA, VIRGO and TAMA projects. ", "introduction": "The phrase ``collision of black holes'' has an aura of a mysterious and exotic happening that is not far from the reality of such an event. A black hole is not an ordinary object defined by the amount and properties of the material of which it is made. Rather it is a region from which no signal can escape. The surface, the black hole horizon, bounding this region is defined by the formal ``no escape'' property. Unlike the surface of an ordinary object the horizon has no local properties that would be sensed by an observer with the bad fortune to fall inward through it. A collision of two holes is the process in which two no-escape regions merge to become a single, larger, region of no escape. In the last few years such mergers have become the focus of much research attention, for two not entirely independent reasons. The first reason is the development of numerical relativity \\cite{Segr}. General relativity, Einstein's theory of gravity, sets the dynamics of space-time via a set of nonlinear partial differential equations of such complexity that analytic solutions have been limited to two classes: solutions of high symmetry, or solutions based on approximation techniques, such as linearized weak field theory. The study of Einstein's equations on computers has been viewed as the key to finding more general asymmetric strong field solutions and it was natural for this key to be be applied to black hole collisions. Black holes are incontrovertibly strong field regions, but single isolated black holes are stationary solutions of Einstein's theory, and the simplifying symmetry of time independence allows for closed form well-understood solutions \\cite{Sch}. Collisions of black holes, on the other hand, are necessarily nonstationary as well as being crucially strong-field events. It is known that the collision will result in a single final black hole and in the generation of gravitational waves carrying off some of the mass energy originally associated with the holes. But this is all that is known with certainty. The nature of the merging of the horizons, in the general collision, is not even qualitatively understood. A reasonably complete understanding awaits progress in numerical relativity, and the wait has been longer than anticipated. The solution of general black hole collisions on computers has proved to be remarkably difficult. There is, however, a class of cases in which reliable answers are available. If the collision is a 'head-on' collision along a straight line, then there is rotational symmetry about the line of the collision. Though the collision is still highly dynamic and nonlinear, the simplifications afforded by this symmetry reduce the computational demands sufficiently that the collision could successfully be simulated even in the mid 1970's, and run with good reliability in the mid 1990's\\cite{NCSA}. The simplification of head-on collisions, however, masks some of the physics of the most interesting types of collisions, the fully three dimensional collisions at the end point of the inspiral of a mutually orbiting pair of black holes. The second development that directed attention to black hole collisions is the advent of sensitive gravitational wave detectors. In the next few years, several interferometric gravitational wave observatories (the LIGO project in the US, the VIRGO and GEO projects in Europe and the TAMA project in Japan \\cite{gw}) may be capable of detecting gravitational waves. Whether near term searches are successful will depend more than anything else on the strength of astrophysical sources. Attributes of a good generator of gravitational waves include strong gravitational fields and high velocities, so black hole processes are a natural source to consider. It is astrophysically plausible that black holes form binary associations with other objects, including other black holes \\cite{nyt}. Due to the loss of energy by the emission of gravitational radiation, the separation and period of the binary orbits would decrease. If the binary consists of two black holes, the inspiral would end with a rapid strong field merger that has the potential to be a powerful source of detectable gravitational waves \\cite{Tho}. The whole process of inspiral generates gravitational radiation, but in the early large-separation stages the radiation is relatively weak and is reasonably well described by Newtonian gravity theory and Post-Newtonian extensions of it \\cite{Cliff}. It is only the final strong field merger that could in principle produce a powerful burst of gravitational waves, but at this point only one parameter of the burst is reliably known. The characteristic frequency of the waves is inversely proportional to the mass of the final black hole formed, and works out to be on the order of $10^{3}$\\,Hz for a 10$M_{\\odot}$ hole, a typical expected mass of a ``stellar'' sized hole. For supermassive holes of mass $\\geq 10^{6}M_{\\odot}$ typical of galactic nuclei, the waves would be less than 1\\,Hz. The maximum sensitivity of the next generation of gravitational wave detectors occurs at frequency around 100\\,Hz and the detectors will be ideally suited to waves from a black hole with mass of several hundred $M_{\\odot}$. Some recent observations\\cite{midweight} offer indirect evidence that black holes in this range may exist. If they do not, then the detection of the collision of black holes may require the deployment of space-based detectors\\cite{LISA} sensitive to the low frequency waves produced by supermassive holes. The ratio of the masses in a binary determines both how difficult it is to analyze, and how exciting it is as a potential source. If the mass of a black hole $M_{1}$ is much larger than the mass of its binary companion $M_{2}$, then the smaller mass object can be treated as a perturbation to the well understood spacetime of the larger mass black hole. The equations that describe perturbations are linear, and hence relatively easily dealt with in general. In the specific case of perturbations to black hole spacetimes, the techniques of calculation were worked out in the 1970s and resulted in the Regge-Wheeler and Zerilli equations \\cite{ReWh,Z} for perturbations of Schwarzschild (nonrotating) black holes, and in the Teukolsky \\cite{Te} equation for perturbations of Kerr (rotating) black holes. The relatively easily analyzed \\cite{dripp} ``particle limit'' case $M_{2}\\ll M_{1}$ may be of interest in connection, say, with neutron stars merging with supermassive black holes, but this process cannot give the hoped for high power. It is easy to show the gravitational wave power generated scales in the masses as $(M_{2}/M_{1})^{2}$. High power requires roughly equal masses, and this means the simplifications of the particle limit do not apply to the most interesting sources. If not directly applicable to equal mass inspiral, the clarity of the particle limit can, at least, help us to formulate questions about the nature of the endpoint of inspiral, like the existence of a last stable circular orbit. As a particle orbits a black hole it reaches a radius at which it can no longer stably orbit with slowly decreasing radius and it begins a rapid inward plunge. For the inspiral of two roughly equal mass holes it can be imagined that the binary gradually spirals inward or that it reaches a point at which a discontinuous plunge begins. If the late orbits are being degraded rapidly enough by the emission of gravitational radiation, there might not even be any meaning to late stage ``stability.'' This uncertainty about even the qualitative nature of the late stage of the inspiral is related to an important, but totally unresolved, question: How does the inspiraling binary shed enough angular momentum to form a black hole? In a relativist's units in which $c=G=1$, a black hole must have a total angular momentum $J$ that is limited by the maximum angular momentum $J= M^{2}$ that a rotating (Kerr) black hole can possess. Until the binary pair is close, its angular momentum will be above this limit, but technical considerations\\cite{angmom} limit the rate at which angular momentum can be shed in gravitational waves at very late stages. If both black holes of the pair are rapidly rotating with angular momentum in the same direction the shedding appears to present a barrier to the formation of the final single black hole. It is possible that even the qualitative details of the late stage inspiral depend on the angular momentum of the inspiraling binary. The set of possibilities is considerable and the answers are important both to our understanding of nonlinear gravitational interactions and to an understanding of gravitational wave sources. Real answers will require advances in numerical relativity that will be several years in coming, but interest in the questions justifies approximation methods that can help, even slightly, to close some of the wide open questions. We take such an approach here. We offer an estimate of the gravitational radiation generated during the late stage of inspiral of two black holes. Our method involves a number of assumptions and limitations that constrain its applicability and reliability, but for all its shortcomings it is one step towards a complete understanding. The approximation method we use, the ``close limit,'' \\cite{Pujap} takes advantage of the property of a black hole horizon. Late in the merger of the binary the single horizon of the final black hole engulfs the entire binary. All the complex structure of the binary will be inside that final horizon, and cannot influence spacetime outside the horizon. It is only what is outside the horizon that can generate gravitational waves that can be detected by distant observers. Since the ultimate fate of the merger is a stationary black hole, it follows that sufficiently late in the merger what is outside the hole will be a perturbation of the final stationary hole. Thus the gravitational waves generated during the latest stage of inspiral can be computed using the techniques of perturbations of black hole spacetimes, with the Zerilli, Regge-Wheeler, and Teukolsky equations. To understand how the close limit method is to be used, it is necessary to consider the general problem addressed by numerical relativity. Einstein's field equations are divided into ``initial value'' equations and equations of time evolution \\cite{York}. The initial value equations determine the nature of spacetime at a chosen initial moment. The solutions of these initial value equations are the initial values for the remaining differential equations of Einstein's theory, the equations that determine the spacetime (including its gravitational wave content) to the future of the initial time. The two tasks of numerical relativity are first to find an initial value solution representing a moment in the life of the colliding holes, and second to find the future spacetime for those initial values. The more computationally difficult task is that of evolving to the future and the codes that accomplish this task tend to be unstable for long time evolutions. For long evolutions to be avoided, the initial value solutions must be chosen to be a moment late in the life of the inspiral. If that moment is late enough, the close limit method can be brought to bear and evolution can be carried out with the stable linearized equations of perturbation theory. But choosing too late a starting moment for evolution creates a new difficulty. The connection of an initial value solution to a ``sensible'' physical configuration for the binary is reasonably secure only if the binary pair is well separated. At close separations, the gravitational field of each of the binary holes strongly affects the other hole, and the individual mass, individual angular momentum, and physical separation of the holes lose clear meaning. The problem then requires navigating between the Scylla of numerical instabilities for evolution, and the Charybdis of uncertain initial conditions. By using a very late initial moment and linearized evolution, the close limit method completely avoids the former hazard. There are reasons beyond speculation to believe that close limit evolutions give useful answers. Numerical relativity results are available for axisymmetric head-on collisions\\cite{NCSA}. These represent evolution of a number of initial value solutions, in particular the closed form solution due to Misner\\cite{Mi}, containing a single parameter representing the initial separation of equal mass holes in units of the mass of the spacetime (in $c=G=1$ units). This separation index defines a parameterized family of initial value solutions. Choices of this parameter can be made corresponding to large or small initial separation. When numerical relativity and close limit results are compared it is seen that agreement is excellent for small initial separations, and is surprisingly good even when the initial configuration is not close enough for a horizon to engulf the entire binary\\cite{PrPu,etal}. Arguments can be made also, that the gravitational waves calculated in the late stage of inspiral are not highly sensitive to details of initial data. Particularly interesting in this regard is work by Abrahams and Cook\\cite{AbCook}. In the past several years the close limit method has been extensively studied for head-on collisions of boosted and spinning holes and compared with the results of numerical relativity. Most notably, second order perturbation theory has been developed for the close limit method\\cite{2nd}. In this process of comparison much has been learned about the strengths and limitations of the close limit method, with the goal of applying the method to problems that cannot yet be handled with numerical relativity. The present work represents the first example of this. We report here the results of the application of the close limit method for the three dimensional problem of the late stage inspiral of two black holes. We will use the close limit method for the initial data families constructed by the Bowen and York \\cite{BoYo} method and the associated ``punctures'' families \\cite{BB}. It is known that these families possess an artificial radiation content when one considers black holes that are close, but such content is also known to be moderate \\cite{GlNiPrPuspin}. An important advantage of these methods is that they are typically the starting point for numerical relativity, and thus close limit evolution of these starting points can be compared with the numerical evolution of these same initial data when such evolutions become available. The most important disadvantage, for our purposes, is that the Bowen-York family does not include the Kerr solution, the solution for a rotating hole. This precludes finding a family of initial value solutions that goes, in the limit of small initial separation, to a Kerr black hole. With Bowen-York initial value solutions, then, we cannot consider a collision that will result in a rapidly rotating hole. Rather, we limit our attention to collisions involving a modest amount of total angular momentum and consider the angular momentum as well as the initial separation to be a perturbation of a nonrotating final hole. It it should also be mentioned that currently fashionable astrophysical scenarios suggest that the individual holes might not carry a significant amount of spin \\cite{nyt} in realistic black hole collisions. The organization of this paper is as follows: in the next section we review the method for obtaining the initial data and describe the approximations involved. In the following two sections we discuss how to set up the perturbative formalism geared towards evolution. Since the collisions have net angular momentum we will evolve them {\\em both} as a perturbation of a rotating and a non-rotating black hole. The comparison of both approaches is given in the subsequent section and we will see that insight is gained by treating the problem in two different ways. We end with a discussion of the results in terms of waveforms and radiated energies and we describe a puzzle in the calculation of the angular momentum radiated. For the reader who wishes to be spared all the details, we summarize our results in a brief punchline: the final ringdown of the inspiraling collision of two non-spinning black holes is unlikely to radiate more than $1\\%$ of the mass of the system or more than $0.1\\%$ of its angular momentum in gravitational waves. ", "conclusions": "We have used the ``close limit'' to estimate the radiation in the collision at the end of the inspiral of two equal mass nonrotating black holes. The assumptions and restrictions were: (i) only the ``ringdown'' radiation was computed; (ii) we assumed that a simple initial data set gave an adequate representation of appropriate astrophysical conditions; (iii) we assumed that the final hole is not near the extreme Kerr limit; (iv) we used close limit estimates of the evolution. Our main conclusion is that the energy radiated in ringdown is probably not more than 1\\% of the total mass of the system, and the angular momentum radiated is not more than 0.1\\% of the initial angular momentum. The most serious uncertainty in this result is the possibility that the radiation from the early merger stage of coalescence is very much larger than the ringdown radiation. With our 1\\%$Mc^{2}$ estimate, collisions of black holes of $100M_{\\odot}$ would be detectable with signal to noise of 6 out to distances on the order of 200Mpc by the initial LIGO configuration and to distances of 4Gpc with the advanced LIGO detector." }, "0003/astro-ph0003272_arXiv.txt": { "abstract": "\\noindent Numerical experiments have been performed to investigate the thermal behavior of a cooled down white dwarf of initial mass $M_{\\rm WD} = 0.516 M_{\\sun}$ which accretes hydrogen-rich matter with $Z = 0.02$ at the rate $\\dot{M}=10^{-8}$ \\msun \\yrm1, typical for a recurrent hydrogen shell flash regime. The evolution of the main physical quantities of a model during a pulse cycle is examined in detail. From selected models in the mass range $M_{\\rm WD} = 0.52\\div 0.68$ \\msunend, we derive the borders in the $M_{\\rm WD}$ - $\\dot{M}$ plane of the steady state accretion regime when hydrogen is burned at a constant rate as rapidly as it is accreted. The physical properties during a hydrogen shell flash in white dwarfs accreting hydrogen-rich matter with metallicities $Z = 0.001$ and $Z = 0.0001$ are also studied. For a fixed accretion rate, a decrease in the metallicity of the accreted matter leads to an increase in the thickness of the hydrogen-rich layer at outburst and a decrease in the hydrogen-burning shell efficiency. In the $M_{\\rm WD}$-$\\dot{M}$ plane, the borders of the steady state accretion band are critically dependent on the metallicity of the accreted matter: on decreasing the metallicity, the band is shifted to lower accretion rates and its width in $\\dot{M}$ is reduced. ", "introduction": "\\noindent The knowledge of the physical consequences of the accretion of hydrogen-rich matter onto a white dwarf has played an important role in understanding the main properties of several types of eruptive stars such as slow and fast novae and symbiotic stars (e.g., Starrfield 1971, Starrfield, Truran, \\& Sparks 1978, Sparks, Starrfield, \\& Truran 1978). In addition, the scenario in which a red giant star transfers mass to its carbon-oxygen (CO) white dwarf companion (Whelan \\& Iben 1973) is regarded by some as one of the most plausible precursor candidates for Type Ia supernovae (e.g., Hachisu, Kato, \\& Nomoto 1996). Fairly extensive surveys of the dependence of behavior on white dwarf mass and accretion rate have been conducted (see, e.g., Iben \\& Tutukov 1996; Cassisi, Iben, \\& Tornamb\\'e 1998 [hereinafter CIT] and references therein). For a low mass CO white dwarf of typical mass in the range $0.5\\div 0.8$ \\msunend, the consequences of the accretion of hydrogen-rich matter of solar metallicity can be summarized as follows: \\begin{itemize} \\item for sufficiently large mass-accretion rates (say, $10^{-7}$ \\msun \\yrm1 or larger, depending on the white dwarf mass), the accreted layer adopts an expanded configuration similar to that of the envelope of a red giant star; \\item for intermediate mass-accretion rates (say, in the range $4\\div 10 \\times 10^{-8}$ \\msun \\yrm1), the accretor burns hydrogen in a steady state regime at the same rate as it accretes hydrogen; \\item for small mass-accretion rates (say, in the range $1\\div 4 \\times 10^{-8}$ \\msun \\yrm1), the accretor experiences recurrent mild flashes; \\item for even smaller mass-accretion rates (say, smaller than $10^{-9}$ \\msun \\yrm1), the accretor experiences very strong nova-like hydrogen shell flashes. \\end{itemize} Although considerable attention has been paid to models that experience recurrent mild hydrogen-burning flashes (e.g., Paczy\\'nski \\& $\\dot{Z}$ytkow 1978, Iben 1982; Jos\\'e, Hernanz, \\& Isern 1993, CIT), a deep insight into the evolution of the main physical properties in the accreting models over an outburst-cooling cycle is still missing. Analytical studies of hydrogen-burning shells by Sugimoto \\& Fujimoto (1978) and by Fujimoto (1982a,b) have established the general properties of the hydrogen-burning shell as a function of the fundamental parameters of the accreting star, but have not provided profiles of structural and chemical variables in the shell itself. Finally, the extant numerical experiments do not explore systematically how the outcome of the accretion process depends on the abundances of heavy elements in the accreted matter. To investigate in detail the evolution of the main physical characteristics of the hydrogen-burning shell during a flash episode, we have adopted as an initial model a white dwarf of mass 0.516 \\msun which has accreted matter at the rate $\\dot{M} = 10^{-8}$ \\msun \\yrm1 for $7.6\\times 10^{5}$ yr, at which point the interior has been cooled to a temperature of $8.6\\times 10^{6}$ K and the density at the center is $2.56 \\times 10^{6}$ g cm$^{-3}$. For the composition of accreted matter, we have adopted a helium abundance by mass of $Y=0.28$ and three different metallicities ($Z=0.02$, 0.001, and 0.0001). In \\S 2 we discuss the input physics and the assumptions. In \\S 3, the thermal properties of the hydrogen-burning shell are presented and discussed in detail for the Z=0.02 case. In \\S 4, analytical relations in the ($M_{\\rm WD}$-$\\dot{M}$) plane are obtained for the case Z=0.02 and, in \\S 5, the dependence on metallicity of the evolutionary behavior of the accreting models is analyzed. Conclusions and a brief discussion follow in \\S 6. ", "conclusions": "We have investigated and discussed in detail the evolutionary behaviour of a white dwarf accreting hydrogen-rich matter of three different metallicities: $Z=0.02, 0.001, \\hbox{and } 0.0001$. An analysis of the evolutionary behavior of several physical characteristics of the models has shown that, for fixed values of $M_{WD}$ and $\\dot{M}$, lowering the metallicity causes the recurrence period to become longer because, in order to achieve the larger temperatures and densities necessary to offset the reduction of CNO catalysts in the accreted matter, the thickness of the hydrogen-rich accreted layer must increase. For the steady-state burning regime, we have been able to derive borders in the $M_{WD}$-$\\dot{M}$ plane as they depend on the metal content of the accreted matter. In agreement with earlier estimates, we find that the area of the region in this plane in which steady-state burning takes place becomes narrower as the white dwarf mass is increased. In addition, the location and the extension of the steady-state burning regime have been found to depend critically on the metallicity of the accreted matter, as shown clearly in Figure 10, where the topology of the steady-state region in the $M_{WD}-\\dot{M}$ plane is provided for the three metallicities considered. Reducing the metallicity, the steady-state burning region drops to smaller accretion rates and its extension is drastically decreased. The consequences of our results for the final behavior of real low metallicity accretors are not easy to predict. On the one hand, as metallicity is decreased, the hydrogen-burning shell becomes hotter. This means that the underlying helium-burning layer is hotter and less degenerate when a helium-burning thermonuclear runaway is initiated. On the other hand, the fact that, for fixed core mass and accretion rate, the power of hydrogen-burning flash decreases as the metallicity is reduced suggests that low metallicity accretors may experience relatively mild helium shell flashes for a range of helium layer masses more extended than in the solar metallicity case. However, as extensively discussed in Piersanti et al.(1999), in the mild pulse regime, there is a parameter region in which the effects of the hydrogen-burning shell on the helium layer are negligible, and a model with characteristics in theis region behaves as if pure helium is accreted. These models lead to a sub-Chandrasekhar explosion if the initial mass of the white dwarf and the accretion rate are within a given range (see Fig. 1 in Tornamb\\'e et al. 1998 for the solar metallicity case). Therefore, over the long term evolution, once the helium-burning layer becomes thermally decoupled from the hydrogen-burning shell, the accreted layer behaves in a way that depends only on the accretion rate and not on the metallicity. Due to the prohibitively long computing time required, we have only partially studied the long term evolution of models accreting hydrogen-rich matter of metallicities $Z=0.001$ and $Z=0.0001$ at the mass-accretion rate of $\\dot{M}=10^{-8}$ \\msun\\yrm1. Both models show that the helium layer and the hydrogen-rich layer become decoupled as in the case of accretion of hydrogen-rich matter of solar metallicity (Piersanti et al. 1999). Such models will likely experience similar outcomes independent of the metallicity of the hydrogen-rich accreted matter. On the basis of the results obtained so far, we suggest that, on lowering the metallicity, the area in the $M_{WD} - \\dot{M}$ plane suitable for sub-Chandrasekhar dynamical outcomes is shifted toward slightly lower values of $\\dot{M}$, remaining almost unchanged in extention, as indicated in Tornamb\\'e et al. (1998) for the solar metallicity. It has to be finally considered that metallicity could even affect other parameters of the binary system (as, for instance, initial white dwarf masses, accretion rates, etc) with the consequence that in the real world, this scenario could be also significantly changed. \\newpage" }, "0003/astro-ph0003102_arXiv.txt": { "abstract": "We present a conceptual design for a space based Galactic Exoplanet Survey Telescope (GEST) which will use the gravitational microlensing technique to detect extra solar planets with masses as low as that of Mars at all separations $\\simgt 1\\,$AU. The microlensing data would be collected by a diffraction limited, wide field imaging telescope of $\\sim 1.5\\,$m aperture equipped with a large array of red-optimized CCD detectors. Such a system would be able to monitor $\\sim 2\\times 10^8$ stars in $\\sim 6$ square degrees of the Galactic bulge at intervals of 20-30 minutes, and it would observe $\\sim 12000$ microlensing events in three bulge seasons. If planetary systems like our own are common, GEST should be able to detect $\\sim 5000$ planets over a 2.5 year lifetime. If gas giants like Jupiter and Saturn are rare, then GEST would detect $\\sim 1300$ planets in a 2.5 year mission if we assume that most planetary systems are dominated by planets of about Neptune's' mass. Such a mission would also discover $\\sim 100$ planets of an Earth mass or smaller if such planets are common. This is a factor of $\\sim 50$ better than the most ambitious ground based programs that have been proposed. GEST will also be sensitive to planets which have been separated from their parent stars. ", "introduction": "\\label{sec-intro} The main strength of the gravitational microlensing planet search technique is that it is sensitive to lower mass planets than other techniques. Observed from space, signals for planets down to the mass of Mars are detectable, but they are much rarer and have a shorter duration than higher mass planetary signals. Thus, a large number of stars must be followed with a high sampling frequency in order to detect low mass planets. With GEST, we are able to monitor $\\sim 2 \\times 10^8$ stars once every 30 minutes or so with a photometric accuracy of $\\sim 1\\,$\\%. The microlensing event will not repeat, so high quality photometric data must be obtained while the event is in progress. Our proposed GEST mission will accomplish this. The planetary systems studied by the microlensing technique are located $1-8\\,$kpc away towards the Galactic center rather than in the local neighborhood. The planetary signals will usually be detected as a modifications of the single lens light curve due to the gravitational effect of the planet. Microlensing is most sensitive to planets near the Einstein ring radius which corresponds to a distance of $1-10\\,$AU from the lens star, but because GEST does not require the discovery of stellar microlensing event to begin intensive monitoring for planets, GEST will be able to detect planets at arbitrarily large separations from their host stars. One such isolated planet may have already been observed (Bennett et al 1997) by the MACHO Collaboration. ", "conclusions": "\\label{sec-sample} \\begin{figure} \\plottwo{lcmarsorb15.ps}{vlcmarsorb15.ps} \\caption{A comparison of the same planetary microlensing event as seen by the GEST and VST simulations. The planet has the same mass fraction as the Earth, $\\epsilon = 3\\times 10^{-6}$ with a separation of $a = 0.97$ Einstein ring radii. GEST detects the planet with a signal of $\\Delta\\chi^2 = 3100$, while The VST survey signal is $\\Delta\\chi^2 = 160$. The VST survey data are probably insufficient to determine the planetary parameters. \\label{fig-lcmarsorb15}} \\end{figure} \\begin{figure} \\plottwo{lcsatorb22.ps}{vlcsatorb22.ps} \\caption{The most significant detection ``fully sampled\" Earth mass fraction planet detection from the simulated VST survey is compared with the same event as seen in the GEST simulation. The separation is $a=1.31$ and the detection signal strengths are $\\Delta\\chi^2 = 39000$ and $\\Delta\\chi^2 = 5400$ for the GEST and VST simulations, respectively. Note that despite passing our criteria, the VST data does not fully sample the planetary deviation. \\label{fig-lcsatorb22}} \\end{figure} \\begin{figure} \\plottwo{lcsatorb14.ps}{vlcsatorb14.ps} \\caption{A large amplitude planetary signal due to a $\\epsilon = 3\\times 10^{-6}$, $a=2.47$ planet is seen in a low magnification stellar microlensing event for the GEST simulation, but both lensing event is entirely missed in the VST simulation. \\label{fig-lcsatorb14}} \\end{figure} In addition to detecting the planetary perturbation to the microlensing lightcurve, it is also important to determine the characteristics of the planet that has been detected. Microlensing generally allows the determination of the planetary mass fraction, $\\epsilon$, and the transverse separation of the planet from the lens star in units of the Einstein radius which is typically about $3(M_{\\rm star}/\\msun)$AU. Gaudi and Gould (1997) have shown that these parameters can be accurately determined if the lightcurve deviations are well sampled. For caustic crossing planetary microlensing events which comprise a large fraction of the low mass planet detections, it is also possible to determine the planetary mass to about a factor of 2 or 3. The gravitational microlensing planet search technique has previously been been considered for ground based observations (Peale 1997; Sackett 1997; Rhie et al 2000; Albrow et al 2000), but ground based surveys face some difficulties due to the requirement of continuous lightcurve monitoring. This can be accomplished with a network of microlensing follow-up telescopes spanning the globe at southern latitudes, but this requires that the survey rely upon observing sites that often have poor weather or seeing conditions. Sackett (1997) has argued that a dedicated 2.5m telescope like the VST with a wide field camera at an excellent site like Paranal could efficiently search for low mass planets where the deviations are expected to last only a few hours. We have done a simulation of such a survey (optimistically) assuming observations in consistent 0.7\" seeing for 8 hours every night for 3 bulge seasons, and we find that low mass planets are not easily detected by such a ground based survey because the highest signal-to-noise events generally last longer than a few hours. If we demand that more than 90\\% of the $\\Delta\\chi^2$ signal occur in the 8 hour observing window for an event to have measurable parameters, then we find that this ``VST\" survey is 30-50 times less sensitive to low mass planets than GEST as shown in Figure 2. Also, even the ``best\" planet detection in the VST survey misses a significant part of the planetary deviation as seen in Figure 6. A study by Peale (1997) of a ground based microlensing planet search program with follow-up telescopes in Chile, Australia, and South Africa, also only manages a handful of low mass planet detections after eight years of observations and is not sensitive to isolated or Mars mass planets. So, the proposed GEST mission is $\\simgt 50$ times more sensitive to low mass planets than both types of propsed ground based surveys." }, "0003/astro-ph0003334_arXiv.txt": { "abstract": "We present new ISO-SWS data for a sample of 27 starburst galaxies, and with these data examine the issues of formation and evolution of the most massive stars in starburst galaxies. Using starburst models which incorporate time evolution, new stellar atmosphere models for massive stars, and a starburst model geometry derived from observations of the prototypical starburst M82, we model the integrated mid-infrared line ratio [Ne~III](15.6\\um)/[Ne~II](12.8\\um). This line ratio is sensitive to the hardness of the stellar energy distribution and therefore to the most massive stars present. We conclude from our models, with consideration of recent determinations of the stellar census in local, high-mass star forming regions, that the [Ne~III]/[Ne~II] ratios we measure are consistent with the formation of massive ($\\sim$ 50-100\\msolar) stars in most starbursts. In this framework, the low nebular excitation inferred from the measured line ratios can be attributed to aging effects. By including estimates of the ratio of infrared-to-Lyman continuum luminosity for the galaxies in our sample, we further find that most starbursts are relatively short-lived (10$^6$-10$^7$ years), only a few O-star lifetimes. We discuss a possible cause of such short events: the effectiveness of stellar winds and supernovae in destroying the starburst environment. ", "introduction": "\\label{Sintro} Starburst galaxies constitute an important class of extragalactic objects. They contribute a significant fraction of the total high-mass star formation in the local universe ({\\em e.g.} Soifer \\etal 1987; Gallego \\etal 1995), and about 25\\% of the high-mass star formation within 10 Mpc occurs in four starburst galaxies (Heckman 1998). At intermediate and high redshifts, important populations of galaxies are observed which exhibit properties indicating intense star formation activity ({\\em e.g.} Colless \\etal 1994; Babul \\& Ferguson 1996; Steidel \\etal 1996; Lowenthal \\etal 1997). A strong correlation between the most prodigious starbursts and interaction or merger events ({\\em e.g.} Condon \\etal 1982; Kennicutt \\etal 1987; Telesco, Wolstencroft \\& Done 1988; Joseph 1990) further emphasizes the importance of starbursts in galaxy evolution. Despite the obvious importance of starbursts, a canvas of recent starburst galaxy studies suggests that a full ``prescription'' for these events has yet to be written. Discrepancies exist even for the most massive stars formed, which represent the largest energy contributions. For instance, it has been suggested by some authors that the production of the highest mass stars may be suppressed in starburst galaxies, based on low measured values of diagnostic line ratios such as He I (2.06\\um)/ HI(Br$\\gamma$) and [Ne II]/[Ar III] detected in well-known starburst sources. These results have been interpreted as possibly indicating severe upper mass cutoffs to the initial mass function (IMF), some as low as $\\sim$25-30\\msolar~({\\em e.g.} Puxley \\etal 1989; Doyon \\etal 1994; Doherty \\etal 1995; Achtermann \\& Lacy 1995; Beck, Kelly, \\& Lacy 1997). A lack of massive stars in starbursts is difficult to reconcile with a growing body of evidence for the formation of very massive stars in nearby regions of active star-formation. In recent studies of local high-mass star-forming regions at both high and low metallicity, stars up to at least 100~\\msolar~are observed; these regions include the Galactic Center (e.g. Krabbe \\etal 1995; Najarro \\etal 1997; Serabyn, Shupe, \\& Figer 1998; Figer \\etal 1998), the Galactic star-forming region NGC\\,3603 (Drissen \\etal 1995; Eisenhauer \\etal 1998), and the R136 cluster at the center of 30 Doradus (e.g. Hunter \\etal 1995, Massey \\& Hunter 1998). There is also a growing body of indirect evidence for the presence of very massive stars in starbursts. {\\em HST} optical imaging has revealed the presence of young compact ``super star clusters'' in several nearby starburst galaxies, including M\\,82 (O'Connell \\etal 1995), NGC\\,4038/4039 (Whitmore \\& Schweizer 1995), NGC 5253 (Meurer \\etal 1995), He 2-10 (Conti \\& Vacca 1994), NGC\\,1569 and NGC\\,1705 (O'Connell, Gallagher, \\& Hunter 1994; Ho \\& Filippenko 1996; Sternberg 1998), and NGC\\,1140 (Hunter, O'Connell \\& Gallagher 1994). These super star clusters have extreme luminosities, and in some cases bright emission features characteristic of Wolf-Rayet stars (e.g., He II 4696\\AA~emission); such properties are difficult to explain without a large contribution from very massive stars (M$\\ge$60\\msolar; see e.g., Gonzalez Delgado \\etal 1997; Schaerer \\etal 1997). Several authors have recently argued that observed low-excitation nebular line ratios in starbursts may be due to aging effects rather than a severe upper mass cutoff ({\\em e.g.} Rieke \\etal 1993; Genzel, Hollenbach, \\& Townes 1994; Achtermann \\& Lacy 1995; Satyapal \\etal 1997; Engelbracht \\etal 1998; see also Vanzi \\& Rieke 1997). Studying a sufficiently large sample of starbursts is therefore valuable in breaking the degeneracy between age and upper mass cutoff. As we are not likely to catch all starbursts at a ``late'' age, and the ultraviolet, optical, and far-infrared luminosities decline with time after the peak star formation activity, we can determine the relative importance of variations in star formation parameters and aging effects by studying a larger sample. In this work, we present results of a starburst modeling program to interpret mid-infrared (MIR) spectroscopy from the Short Wavelength Spectrometer (SWS; de Graauw \\etal 1996) aboard the Infrared Space Observatory (ISO; Kessler \\etal 1996). To this end, we have gathered atomic fine structure line fluxes (in particular for the [Ne~II] 12.8$\\mu$m and [Ne~III] 15.6$\\mu$m lines) for 27 starburst galaxies. This project is ideally suited for investigating the properties of the high-mass stellar population in heavily obscured star-forming regions of starburst galaxies, since the extinction at MIR wavelengths is only a few percent of the optical extinction. This investigation thus constitutes an important contribution to the understanding of star formation in starburst systems, a field that has been dominated by studies in the optical and ultraviolet regimes in the intervening years between the IRAS and ISO missions. As recent studies of the cosmic infrared background suggest a significant contribution from dusty starburst systems (e.g., Puget \\etal 1996; Hughes \\etal 1998; Hauser \\etal 1998; Elbaz \\etal 1998), the understanding of dusty starbursts is important in defining the picture of the most powerful star-forming events in the universe. We present the sample data in \\S\\ref{Sdata}. We outline our modeling procedure and compare our ISO-SWS observations with model predictions in \\S\\ref{Smodels}. We examine additional constraints on this starburst modeling study and examine the robustness of the modeling analysis in \\S\\ref{Sanalysis}. We discuss the implications of our findings and summarize this work in \\S\\ref{Ssummary}. ", "conclusions": "\\label{Ssummary} The cluster models that we have presented in this paper support the formation of very massive stars (50-100\\msolar) in starburst galaxies. While the quantitative estimate of \\mup~for each galaxy is model-dependent, it is clear that the formation of very massive stars is necessary to explain the ionized line diagnostics observed in this starburst sample. This result suggests that while starbursts produce prodigious amounts of energy and stars, the high-mass stellar populations in starburst galaxies are not radically different than those in high-mass star-forming regions observed locally. \\subsection{Short timescales for starburst activity} As illustrated in Figures \\ref{Fsbmods} and \\ref{Fsbmodsllyc}, the combination of the neon line ratio with the \\lir/\\llyc\\ ratio strongly favors the scenario for starburst activity where very massive stars form, as in local smaller-scale starburst templates, and where the burst last typically a few million to $\\sim$10$^7$ years. Such timescales are shorter than previously thought (10$^{7} - 10^{8}~{\\rm yr}$; e.g., Thronson \\& Telesco 1986; Heckman 1998). It is clear that detailed modeling is required to secure this result, including additional constraints ({\\em e.g.} $K$-band luminosity, the rate of supernova explosions, the depth of the near-infrared CO bandheads) and spatially resolved information. It is nonetheless in agreement with other recent detailed studies of a few starburst galaxies, some of which are also included in our sample ({\\em e.g.} M\\,82, Schreiber 1999; NGC\\,253, Engelbracht \\etal 1998). As a result of instrumental progress, there is now growing evidence that starburst activity occurs in individual burst sites on physical scales of a few tens of parsecs or less. Short timescales are therefore naturally understandable locally. Our data, in conjunction with the other studies cited above, provide evidence for short timescales on much larger scales, suggesting that starburst activity also occurs {\\em globally} on short timescales, presumably as a result of one brief gas compression event, or of successive episodes of such events separated by more than one typical timescale. Short burst durations thus imply strong negative feedback effects of starburst activity, globally as well as locally. A simple argument can be invoked to explain the physical arguments behind this result. We can compare the cumulative kinetic energy injected in the ISM by the supernova explosions over time ($E_{\\rm kin}$) with the binding energy of the gas ($E_{\\rm grav}$), and assume the starburst activity will stop when $E_{\\rm kin}$ just balances $E_{\\rm grav}$. This is a simplistic way of expressing the conditions for a starburst wind to break out of the galaxy ({\\em e.g.} Heckman, Armus \\& Miley 1990), but it is sufficient for order-of-magnitude estimates. In order to relate $E_{\\rm kin}$ to observed quantities, we have considered the relationship between the rate of supernova explosions \\snrate\\ and the bolometric luminosity \\lbol. Model predictions obtained with STARS for a variety of star formation histories and upper mass cutoffs of the IMF indicate that $$ 10^{12}\\,\\left(\\frac{\\nu_{\\rm SN}}{\\rm yr^{-1}}\\right)\\, \\left(\\frac{L_{\\rm bol}}{\\rm L_{\\odot}}\\right)^{-1} \\sim 1 \\eqno(4) $$ as soon as the massive stars start to explode as supernovae, and as long as substantial star formation takes place. It thus seems reasonable to assume that Equation (4) holds for the bulk of the sample, likely having a range in age and timescale but all exhibiting signs of significant, recent starburst activity. For simplicity, we here assume a spherical geometry for the starbursts, with uniform mass distribution. In addition, we assume that each supernova explosion liberates $E_{\\rm mech}^{\\rm SN} = 10^{51}~{\\rm erg}$ of mechanical energy, transferred as kinetic energy to the ISM with an efficiency $\\eta$. The timescale $\\tau$ for our condition above satisfies : $$ \\eta\\,\\left(\\frac{{\\rm d}E_{\\rm mech}^{\\rm SN}}{{\\rm d}t}\\right)\\,\\tau \\simeq \\frac{G\\,M_{\\rm dyn}^{2}}{R}, \\eqno(5) $$ where ${\\rm d}E_{\\rm mech}^{\\rm SN}/{\\rm d}t$ is the rate of mechanical energy injection from the supernovae, $G$ is the gravitational constant, $M_{\\rm dyn}$ is the dynamical mass of the system, and $R$ is the radius of the starburst region. Equation (5) can be re-written as $$ \\frac{\\tau}{\\rm Myr} \\simeq \\left(\\frac{8.5}{\\eta}\\right)\\, \\left(\\frac{M_{\\rm dyn}}{\\rm 10^{9}\\,M_{\\odot}}\\right)^{2}\\, \\left(\\frac{R}{\\rm kpc}\\right)^{-1}\\, \\left(\\frac{L_{\\rm IR}}{\\rm 10^{10}\\,L_{\\odot}}\\right)^{-1} \\eqno(6) $$ where we have substituted \\lir\\ for \\lbol, appropriate for our sample galaxies. Application of Equation (6) to M82 ($M_{\\rm dyn} = 8 \\times 10^{8}~{\\rm M_{\\odot}}$, $R = 0.25~{\\rm kpc}$, $L_{\\rm IR} = 4 \\times 10^{10}~{\\rm L_{\\odot}}$), yields $\\tau \\simeq 5\\,\\eta^{-1}~{\\rm Myr}$, so for efficiencies $\\ga 10\\%$, the estimated timescales are $\\sim 10^{6} - 10^{7}~{\\rm yr}$. Our argument above is based on ``gas-disruption timescale'' estimates. This differs from the conventional ``gas-consumption'' arguments, which compare the star formation rates with the mass of the gas reservoir. In such estimates, the star formation rates are based on comparison of absolute fluxes ({\\em e.g.} H$\\alpha$ fluxes, \\lir) with predictions from starburst models. The estimates are thus very sensitive to the assumed age and history of the starburst. Our estimates of the gas-disruption timescales are also model-dependent, but have the advantage of being based on a quantity ($10^{12}\\,\\nu_{\\rm SN}/L_{\\rm bol}$) which varies by smaller factors. Neither point of view accounts for further fueling processes, or dynamical evolution of the systems ({\\em e.g.} starbursts in barred galaxies, interacting/merging systems, etc.). However, the discussion presented here gives an alternative perspective to the issue of global burst timescales, and provides a plausible explanation for our results. \\subsection{Summary} Starburst models predicting the [Ne~III]/[Ne~II] ratio from ISO-SWS spectra of 27 starburst galaxies show that the observed data are consistent with the formation of very massive stars in starbursts, thus precluding the need for the restrictive upper mass cutoffs suggested by some earlier studies (\\mup$\\sim$25-30\\msolar). Combining the neon line ratios with starburst modeling and the consideration of the stellar content measured in local star forming regions, we find that starburst events may be generally described as short bursts of star formation which produce very massive stars, and which exhibit relatively soft integrated line ratios as a result of aging the stellar population. In particular, our modeling of neon and \\lir/\\llyc\\ ratios, together with results on local high-mass star-forming regions, suggest: $\\bullet$ very massive stars (\\mup$\\ga$50\\msolar) form in typical starbursts. $\\bullet$ starbursts have short global timescales, \\tsc$\\la$10$^7$ years. These results suggest strong negative feedback from starburst activity; the galactic superwinds frequently observed in starburst galaxies are particularly striking examples of the consequences of such feedback. In our analysis, we have examined the degeneracy between aging effects and model parameter variations in the assessment of upper mass cutoffs to the IMF. There is still room for significant improvements in modeling the properties of starbursts: determination of metallicities and the radiation environment (e.g., for measurements of the ionization parameter U) compete with the characterization of stellar properties (SEDs, evolutionary tracks) as the largest contributors to uncertainty in the modeling of star formation properties such as the upper mass cutoff. Other datasets, such as additional MIR line ratios (e.g., Kunze \\etal 1996; Rigopoulou \\etal 1996; Engelbracht \\etal 1998) or K-band luminosities and near-infrared spectroscopy (e.g., Forbes \\etal 1993; van der Werf \\etal 1993; Genzel \\etal 1995; Tacconi-Garman \\etal 1996; B\\\"oker, F\\\"orster-Schreiber, \\& Genzel 1997; Engelbracht \\etal 1998; Schreiber 1999) would be very useful in further constraining the properties of starbursts. However, it will be important to compile such additional data for a large sample in order to proscribe further, {\\em general} constraints on the way in which starbursts form stars. Observations with higher spatial resolution would better isolate regions of active star formation, making it possible to confirm whether high- and low-excitation lines arise from the same region; the spectroscopic capabilities of SIRTF will be well-suited to addressing this issue. By accounting for a reasonable range of uncertainties which constrain the present observations, we find that the observed MIR neon ratios are generally consistent with the formation of very massive stars in starburst events; we offer this hypothesis up to future datasets for increasingly rigorous testing. We would like to thank A. Pauldrach and R.-P. Kudritzki for providing model atmospheres, Tal Alexander for assistance in introducing low-metallicity SEDs into STARS, and Eckhard Sturm for providing the ISO-SWS spectra of Arp 220. We would also like to thank Jack Gallimore and Dan Tran for interesting discussions. MDT would like to thank the Alexander von Humboldt-Stiftung and the NRAO\\footnote{The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.} for support during the production of this work. This research also received support from the German-Israeli Foundation under grant I-0551-186.07/97. SWS and the ISO Spectrometer Data Center at MPE are supported by DARA under grants 50 QI 8610 8 and 50 QI 9402 3." }, "0003/astro-ph0003012_arXiv.txt": { "abstract": "We present a morphological analysis of the feature-rich 2MASS LMC color-magnitude diagram, identifying Galactic and LMC populations and estimating the density of LMC populations alone. We also present the projected spatial distributions of various stellar populations. The conditions that prevailed when 2MASS observed the LMC provided $10\\sigma$ limiting sensitivity for $J\\simlt16.3$, $H\\simlt15.3$, $K_s\\simlt14.7$. Major populations are identified based on matching morphological features of the color-magnitude diagram with expected positions of known populations, isochrone fits, and analysis of the projected spatial distributions. 2MASS has detected a significant population of asymptotic giant branch (AGB) stars ($\\simgt 10^4$ sources) and obscured AGB stars ($\\simgt 2 \\times 10^3$ sources). The LMC populations along the first-ascent red giant branch (RGB) and AGB are quantified. Comparison of the giant luminosity functions in the bar and the outer regions of the LMC shows that both luminosity functions appear consistent with each other. The luminosity function in central (bar) field has well-defined drop-off near $K_s=12.3^m$ corresponding to the location of RGB tip; the same feature is seen in the luminosity function of the entire LMC field. Isochrone fits for the corresponding giant branches reveal no significant differences in metallicities and ages between central and outer regions of the LMC. This may be evidence for strong dynamical evolution in the last several gigayears. In particular, the observed LMC giant branch is well-fit by published tracks in the CIT/CTIO system with a distance modulus of $\\mu=18.5\\pm0.1$, reddening $E_{B-V}=0.15- 0.20$, metallicity $Z=0.004^{+0.002}_{-0.001}$ and age $3-13$ Gyr. Analysis of deep 2MASS engineering data with six times the standard exposure produces similar estimates. ", "introduction": "\\label{sec:introduction} Large and homogeneous data sets of near-infrared photometry for the entire LMC, a by-product of large-scale infrared sky surveys such as 2MASS \\cite{skr97} and DENIS \\cite{epc97}, have become available to astronomical community only recently. Cioni \\etal{} (2000) introduced the DENIS Point Source Catalog towards the Magellanic Clouds. Here, we present the LMC data from The Two Micron All Sky Survey (2MASS). The 2MASS has observed the entirety of the Large Magellanic Cloud and much of these data are included in the recent second incremental data release. Empirically, the photometry has signal-to-noise (SNR) ratio 10 at J, H, K$_s$ magnitudes of 16.3, 15.3, 14.7, respectively, slightly better than the nominal survey limit. At these limits, we can observe all of the thermally-pulsing asymptotic giant branch (AGB) populations and part way down the red giant branch (RGB). The red clump, representing helium burning giants, lies $\\sim 2$ mag below the sensitivity limit of these data. The extinction in near-infrared (NIR) is small throughout the LMC and negligible on average everywhere but the inner degree of arc. Together with the high quality of 2MASS photometry ($\\sigma_m \\sim 0.03^m$), overall zero-point stability (better than $0.01^m$) and with reliable identification of LMC populations, the survey is ideal for studies of spatial structure of the LMC or its evolution (see the companion paper, Weinberg \\& Nikolaev 2000; hereafter WN). We describe data selection, cross-correlation with a few well-known populations, and comparison of the 2MASS color system in \\S\\ref{sec:data}. We present a morphological analysis the color-magnitude diagram in \\S\\ref{sec:cmd}. In particular, we make identifications of the Galactic and LMC populations corresponding to all features in the color-magnitude diagram, correlate these with spatial distribution, and estimate the density of LMC populations alone. The LMC giant branch is well determined and we separately identify the AGB, first-ascent red giant branch tip (TRGB), and the carbon star sequence. The luminosity function of RGB and AGB populations is derived (\\S\\ref{sec:lf}) and compared with the galactic giant-branch luminosity function. We explore the feasibility of determining the spatial dependence of metallicity using giant-branch morphology. Finally, \\S\\ref{sec:summary} summarizes our results and discusses the implications and opportunity for future study. ", "conclusions": "\\label{sec:summary} In this paper, we analyzed the near-infrared CMD of the Large Magellanic Cloud and identified the major stellar populations. The populations are identified based on isochrone fitting and matching the theoretical CMD colors of known populations to the observed CMD source density. Tables~\\ref{table:regions} and \\ref{table:populations} summarize the contents of the CMD regions. \\begin{table*}[t] \\caption{Regions of interest in the 2MASS CMD of the Large Magellanic Cloud. \\label{table:regions}} \\begin{tabular} {ccccr} \\hline Region & $N_{src}$ & $f_{Gal}$$^{\\rm a}$ & Region boundaries & Dominant Spectral Types$^{\\rm b}$ \\\\ \\hline A & 6,659 & $0.15$ & $112$ keV; $K_* = (0.2\\pm 0.1) \\times 10^{34}$ erg cm$^2$ g$^{-5/3}$ for groups and clusters with average temperatures $k_BT<2$ keV. These values correspond to different excess energy per particle of $k_BT \\geq 0.1 (K_*/0.4\\times 10^{34}) $ keV. The dependence of $K_*$ on the mass scale can be well reproduced by an epoch dependent external entropy: the relation $K_* = 0.8(1+z)^{-1}\\times 10^{34}$ erg cm$^2$ g$^{-5/3}$ fits the data over the whole temperature range. The model can be extended to include internal heating, but in this case the energy budget required to fit the X--ray properties would be much higher. Observations of both local and distant clusters can be used to trace the distribution and the evolution of the entropy in the cosmic baryons, and to constrain the typical epoch and the source of the heating processes. The X--ray satellites Chandra and XMM can add to our knowledge of the history of the cosmic baryons, already derived from the high redshift, low density gas observed in the QSO absorption-line clouds, by imaging the hot, higher density plasma observed in groups and clusters of galaxies. ", "introduction": "Clusters of galaxies are the largest virialized objects in the Universe, and are usually considered a canonical data set for testing cosmology. They are the largest collections of diffuse, highly ionized baryons that are directly observable in X--rays mostly through thermal bremsstrahlung emission. The strong dependence of X--ray emission on density $L\\propto \\rho^2$ allows one to select clusters and define complete samples much better than in the optical band. X-ray observations of cluster number counts, luminosity functions and temperature distributions indicate little apparent evolution in clusters back to redshifts as high as $\\sim 0.7$ (e.g., Henry 1997, 2000; Rosati et al. 1998; Schindler 1999), with the exception of very high luminosity objects or very high redshifts (Gioia et al. 1990; Rosati et al. 2000). This set of results provides one of the strongest challenges to high--density cosmological models in which cluster evolution is expected to be detectable even at redshifts as low as $z\\simeq 0.3$. However, these tests are highly dependent on the thermodynamic evolution of the ICM (e.g. see Borgani et al. 1999 and references therein; Bower 1997). The best--fit cosmological parameters are degenerate with the phenomenological parameters used to describe the evolutionary properties of the ICM. In fact, the diffuse baryons in clusters do not simply follow the dark matter, as would be the case if they were driven only by gravity as in self--similar models (Kaiser 1986). Significant efforts have been devoted recently to building a physical model for the ICM including an energy scale at which baryons and dark matter effectively decouple and the self--similarity is broken. The presence of a minimum entropy in the pre-collapse IGM has been advocated for some time as a way to naturally break the self-similar behaviour (Kaiser 1991, Evrard \\& Henry 1991). Such an extra entropy is the key ingredient in reproducing the observed luminosity--temperature relation $L\\propto T^{n}$ with $n\\simeq 3$ (David et al. 1993, Mushotzky \\& Scharf 1997, Allen \\& Fabian 1998; Arnaud \\& Evrard 1999; Markevitch 1998), which is at variance with the self--similar prediction $L \\propto T^2$. Such an entropy minimum bends the relation from self--similar $L\\propto T^2$ behaviour at very large scales ($\\sim 10^{15} M_\\odot$) towards a steeper slope on the scale of groups ($\\sim 10^{13}- 10^{14}M_\\odot$) which is actually observed (Ponman et al. 1996; Helsdon \\& Ponman 2000). The average $L\\propto T^3$ relationship is essentially produced by the flattening of the density distribution in the cores of the X--ray halos; such cores grow larger as the mass scale decreases, and the luminosity steepens further on the scale of groups, where the gas is only adiabatically compressed (see Balogh, Babul \\& Patton 1999; Cavaliere, Menci \\& Tozzi 1997, hereafter CMT97; Cavaliere, Menci \\& Tozzi 1999). The picture has been reinforced by the net change observed in the chemical properties and the spatial distribution of the ICM on the scale of groups, below the observed temperature of 1 keV (Renzini 1997, 1999) where the effects of the entropy excess are expected to be strongest. Another piece of evidence can be obtained from the observed mass--temperature relation (see Horner, Mushotzky \\& Scharf 1999). Recently, an excess of entropy (with respect to the self similar scaling) has been directly detected in the central regions of small clusters with temperatures between $1$ and $3$ keV (Ponman, Cannon \\& Navarro 1999, hereafter PCN; see also Lloyd--Davies, Ponman \\& Cannon 2000), pointing to the role of the entropy as the key ingredient determining the different properties of clusters and groups. Independent hints come from the extragalactic X--ray background: without a substantial entropy injection at early epochs, its level and correlation function would exceed the observed limits, due to the widespread cooling phenomena that would radiate most of the gravitational energy of the collapsing baryons in the soft X--ray band (Pen 1999; Wu, Fabian \\& Nulsen 1999). However, even if there are many hints pointing towards a comprehensive picture, there is a large uncertainty on the amount of {\\sl extra--energy} that effectively generates the entropy excess. It can be shown that it is the final entropy distribution that determines both the spatial distribution of the ICM and its evolutionary properties, irrespective of the total energy released in the past. A given entropy level can be reached through different thermodynamic histories, so that it is not possible to relate the ICM properties directly to a given energy excess without knowing the detailed physics of the heating processes. As we will show in this paper, the first question to answer is not: {\\sl how much energy has been released in the ICM}? but rather: {\\sl what is the sequence of adiabats through which the baryons evolve}? It is difficult to predict {\\sl a priori} the entropy excess of the cosmic baryons, since most of the processes regulating nuclear activity, star and galaxy formation, and the transfer of energy to the surrounding baryons, are out of reach of present--day techniques. Thus, at present there is no general consensus on the production mechanism of such extra--entropy. For example, it is not clear whether the entropy minimum has been established in the IGM before it has been accreted --the external scenario, or in the high density ICM after accretion--the internal scenario. A different energy budget is required in the two different scenarios: a few tenths of a keV per particle are needed if the entropy is generated early enough to keep the baryons on a high adiabat, which prevents them from reaching high densities and cooling massively; much higher energy excess ($> 1$ keV per particle) is required if the entropy is generated later, when the cooling process is eventually already widespread and most of the gas is already at high densities (Tozzi, Scharf \\& Norman 2000, hereafter TSN00). The external scenario, which we will assume as a reference model, is provided by a ubiquitous entropy {\\sl floor} in the diffuse gas, which is entirely due to non--gravitational processes and is assumed to be in place before the onset of gravitational collapse of massive halos. The initial extra entropy is ineffective in large mass systems, where most of the entropy is due to strong shocks, but is more important in smaller mass systems, where the entropy production via shocks is strongly reduced. Eventually a large part of the baryons are merely adiabatically compressed and retain full memory of the initial entropy level. The non--gravitational origin of the excess entropy is crucial, since its level is independent of the mass scale and it breaks the self--similarity, while gravitational processes always scale self--similarly with mass. We present a detailed model to relate the thermodynamic properties of the ICM in groups and clusters of galaxies to an initial entropy excess in the IGM, taking into account the transition between the adiabatic and the shock regime in the growth of X--ray emitting halos. The effect of radiative cooling is also included. We show that, despite the many complexities involved, the entropy is always a convenient synthetic quantity to describe the thermodynamic history of the cosmic baryons at least on the scale of groups and clusters. In particular, we show that in many circumstances the entropy track of a shell of baryons being accreted onto dark matter halos goes through three major regimes: (1) adiabatic compression, during which both heating and cooling are negligible and the entropy is constant; (2) step--like discontinuities due to gravitationally induced shocks; and (3) slow decrease when cooling becomes efficient for baryons in the inner regions of large halos. The entropy jump, the onset of cooling, and the final spatial distribution of the ICM, depend on the initial entropy. Such an external, initial entropy level can be reconstructed from the observation of a large number of distant clusters, or from the spatially and spectrally resolved profiles of nearby halos (see TSN00). Even if the knowledge of the entropy does not resolve the details of the underlying heating history and determine unambiguously the energy budget, the combination of data in the X-ray band with data in the optical and infrared bands can help to identify the major source of heating. In principle, this allows a detailed reconstruction of the energetic processes that affect the cosmic baryons over a wide range of scales and cosmic epochs. The paper is organized as follows. In \\S 2 we establish a one--to--one correspondence between the entropy level and the distribution of the ICM in halos in equilibrium. In \\S 3 we present a generalized spherical infall model to follow the entropy track of each shell. In \\S 4 we derive the average density and temperature profiles and the related global properties such as luminosity, emission weighted temperature and core radius, as a function of mass scale, cosmology, epoch and dark matter profile. In \\S 5 we widen the parameter space, and investigate a time--dependent background entropy to show how the evolution in the entropy reflects in the X--ray properties of clusters of galaxies. In \\S 6 we discuss the limitation of the present approach. Finally, our conclusions and future perspectives are presented in \\S 7. ", "conclusions": "We have presented a detailed model to relate the X--ray properties of diffuse baryons in clusters of galaxies to the entropy history of the cosmic baryons, after including adiabatic compression, shock heating and cooling. Our aim is to build a useful tool to reconstruct the entropy history of the universe from the observations of local and distant clusters. In particular, a major goal is to identify and follow in time the processes that generate the entropy excess. This entropy excess is now probed by many observations and it is connected with many scaling properties of X--ray halos. Even if a given entropy excess does not translate into a unique heating history, the comparison of X--ray data with observations in other bands may allow identification of the major heating sources. Favoured candidates are star formation processes and nuclear activity. At present, however, neither the epoch, nor the source of the related heating process have been identified. In this paper we have limited the investigation to a scenario in which the excess entropy is present since very high $z$ and is uniform throughout the IGM. A case with an external entropy decreasing with redshift, mimicking the rise of a population of heating sources, is also presented. In both the constant and time--evolving case, the scaling properties of local clusters of galaxies are reproduced on a large range of scales, with an appropriate choice of the free parameter $K_*$. The properties of distant X--ray halos are predicted to be generally similar to properties of the local population, but significative differences can be actually observed by the present--day X--ray satellites, shedding light on the thermodynamics history of the ICM. We recall here the general results on density and temperature profiles, together with the results on the evolution of the global X--ray properties, especially luminosity and emission weighted temperatures. The bending of the $L$--$T$ relation with respect to the self--similar case $L\\propto T^2$, is due to the flatter profiles of the ICM going from large mass to small mass halos. Good fits are obtained for a background entropy in the range $K_* = (0.2\\pm 0.1) \\times 10^{34}$ erg cm$^2$ g$^{-5/3}$ for $k_BT_{ew}<2$ keV, and $K_* = (0.4\\pm 0.1) \\times 10^{34}$ erg cm$^2$ g$^{-5/3}$ for $k_BT_{ew}>2$ keV. This scale dependence can be introduced by an evolution in the effective value of $K_*$. In particular, $K_{34} = 0.8 (1+z)^{-1}$ gives a good fit over the whole range of observed temperatures. The central regions of groups and clusters, which dominate the X--ray emission, are formed during the initial stages of accretion. In these early phases, if a significant background entropy is present, the accretion is adiabatic, and the gas is compressed in a flat, low density profile with steep temperature gradients. This is relevant for the smallest halos, where the gravity does not overcome the pressure support of the baryons for the majority of the subsequent accretion of gas. In clusters the infall velocities rapidly become larger than the sound speed, and the shock regime takes over. In the outer regions of clusters the entropy is entirely due to gravitational processes, and the entropy profile is a featurless power law approaching $K\\propto r^{1.1}$. This mechanism is particularly efficient if cooling is neglected. However, it is known that the cooling is an important ingredient in the history of the ICM. The main effect is that the isentropic cores expected in the constant entropy scenario, are partially erased by the process of cooling. Still, if $K_* > 0.1 \\times 10^{34}$ erg cm$^2$ g$^{-5/3}$, the cooling processes are significantly suppressed and the inner regions of the halos keep the imprint of the initial entropy level. Cooling processes appear again only in massive halos, where the gravity dominates the energy of the system and the excess entropy is no longer able to keep the gas at low density. In the extreme case of negligible $K_*$, it is worth noting that the cooling processes alone would have a dramatic effect on both small and large mass halos. In small mass halos ($10^{13} M_\\odot$) most of the gas is expected to cool and recombine, causing a central {\\sl baryonic catastrophe}. Other important characteristics are found in the temperature structure especially of smaller halos. Temperature gradients are commonly expected both in clusters and in groups. The polytropic index is predicted to be $\\gamma_p\\simeq 0.9 -1.2$ in the region where the gas is shock heated. The polytropic index can be higher if the dark matter profile is centrally peaked (e.g., with a power law with index $\\sim -1.4$, see Moore et al. 1998). Another relevant observable (for local halos) is the position of the final shock radius, which is expected to be close to the virial one at large mass scale, while it migrates to larger radii in small groups. In the smallest halos, in fact, the shock is vanishingly small. As a function of epoch, for a given object, the shock/accretion radius is initially quite distant from the virial radius. It is very close to the virial radius when the mass accretion rate reaches its maximum and the shock regime is well developed. Eventually, the mass accretion rate decreases (especially in the $\\Lambda$CDM universe) and the shock radius relaxes again to larger positions. A consequence of the above picture is that the ratio of the baryonic mass included in the virial radius to the total mass, is always lower but still close to unity; it can be significantly lower ($1/3$) only for small mass halos (corresponding to emission--weighted temperatures of $0.3-1$ keV). It is remarkable that the simple presence of an initial excess entropy in the diffuse IGM can reproduce many of the scaling properties of the observed X--ray halos, without the contribution of any internal heating. It is interesting to discuss the implications of this simple scenario for the energetic budget and the past cosmic star formation history. The minimum excess energy associated with an initial background entropy $K_{34}$ is about \\begin{equation} k_BT\\simeq 0.1 \\Big( {{K_{34}}\\over{0.4}}\\Big) \\, \\, {\\rm keV}, \\end{equation} where the gas is assumed to be at the background density at the epoch of the heating. However, we can speculate on the energy budget when the entropy excess is generated after the collapse, at much larger densities (the internal scenario). Following PCN, we can establish a relation between the epoch of heating and the energy released. Under the assumption that the heating process can be described with a single epoch and a typical overdensity, we have: \\begin{equation} 1+z_h = \\Big( {{k_BT_h}\\over{3.2 \\times 10^{-2} K_{34}}}\\Big)^{1/2} \\delta^{-1/3} \\, , \\end{equation} where $k_BT_h$ is the average energy per particle released in the IGM by non--gravitational processes. If we adopt the conservative scenario in which the gas is heated at a typical virial density ($\\delta \\simeq 200$), to have an entropy level in the range $K_* = (0.4\\pm 0.2)\\times 10^{34}$ erg cm$^2$ g$^{-5/3}$, we obtain: \\begin{equation} 1+z_h \\simeq (1.5\\pm 0.3)\\Big( {{k_BT_h}\\over{1 {\\rm keV}}} \\Big)^{1/2}\\, . \\end{equation} Thus, if we want heating at $z>1$ in order to avoid the overcooling catastrophe, the energy budget must be larger than $1$ keV per particle. The above estimate would give even larger values after the inclusion of cooling. In fact, if the gas is heated at high densities, most of the extra energy is likely to be re--emitted soon, and this would raise the energetic budget for a given final entropy excess. In this respect, the relation between the epoch of heating and the energy released is strongly dependent on the physical process. Of course, a scenario in which the extra entropy is provided by the contributions of several different sources, active at different epochs, is a likely possibility. In this perspective, the measure of metallicities as a function of the entropy of the baryons in different systems, from Ly$\\alpha$ clouds to rich clusters, may be useful in determining whether the excess entropy is linked to star formation processes. The assumption of an initial excess entropy uniformly diffused in the IGM, offers new perspectives in the approach to cluster formation, but also galaxy formation. Such an entropy background, once established, may affect the star formation itself, since the cooling processes on all scales are virtually inhibited. This is the mechanism which is expected to solve the {\\sl cooling catastrophe} (see White \\& Rees 1978, Blanchard, Valls Gabaud \\& Mamon 1992; Prunet \\& Blanchard 1999) and in this view X--ray clusters and galaxy formation processes are intimately related. Current attempts to model {\\sl ab initio} the physics of the heating process, and then link the entropy history of the cosmic baryons to galaxy formation, must include the well known plethora of ingredients that has been already mentioned several times: feedback from star formation processes and SNe explosions, radiative and mechanical heating from active galactic nuclei, radiative heating from hard X--ray background, gravitational heating on large scale filaments (see Menci \\& Cavaliere 2000; Valageas \\& Silk 2000; Wu, Fabian \\& Nulsen 2000; Madau \\& Efstathiou 1999; Cen \\& Ostriker 1999). Such different scenarios allow for different entropy histories of the universe, determining both the spatial distribution and the evolution of the entropy in the diffuse gas. A promising strategy for the near future is to look directly for the distribution of the entropy in the ICM (TSN00). A direct consequence of assuming a uniform entropy everywhere in the gas, is that the groups are expected to be surrounded by large halos of low surface brightness gas, spread out over radii much larger than the virial radius of dark matter halos. This low--density gas may have been missed by observations with the ROSAT satellite, but can be detected by the XMM satellite. Its emission can enhance the total luminosity of the groups by more than a factor of $3$, including the lowest energy bins of $\\simeq 0.1$ keV. Another promising observational channel is the absorption from metals in the gas seen against bright X or UV sources. If the source of the background entropy is star formation, significant pollution by metals is expected. The model presented here is to be considered a useful tool to interpret the observations of high redshift clusters, that will be provided especially by the Chandra and XMM satellites. Our aim is to build a solid link between the thermodynamics of the diffuse cosmic baryons and the emitting properties of X--ray halos, in order to be able to {\\sl reconstruct the entropy history of the universe, at high and low redshifts, from spectral and imaging X--ray observations}. This will help in understanding the source of the entropy excess and the time evolution of the corresponding heating process. We acknowledge discussions with S. Borgani, N. Menci, and P. Rosati. We thank T.J. Ponman for discussions and for providing the data in Figure \\ref{fig17}. We thank R. Giacconi for discussions and continuous encouragement. We thank the referee, Greg Bryan, for detailed comments. PT thanks ESO Garching for hospitality during the completion of this work. This work has been supported by NASA grant NAG 8-1133. \\appendix" }, "0003/astro-ph0003068_arXiv.txt": { "abstract": "A Keck~II low resolution spectrum shortward of one micron is presented for SDSS~1624+0029, the first field methane or T dwarf discovered in the Sloan Digital Sky Survey. Significant flux is detected down to the spectrum's short wavelength limit of 6200\\AA. The spectrum exhibits a broad absorption feature centered at 7700\\AA, which we interpret as the K~I 7665/7699 resonance doublet. The observed flux declines shortward of 7000\\AA, due most likely to the red wing of the Na~I doublet. Both Cs~I doublet lines are detected more strongly than in an earlier red spectrum. Neither Li~I absorption nor H$\\alpha$ emission are detected. An exploratory model fit to the spectrum suggests that the shape of the red spectrum can be primarily accounted for by the broad wings of the K~I and Na~I doublets. This behavior is consistent with the argument proffered by Burrows, Marley and Sharp that strong alkali absorption is principally responsible for depressing T dwarf spectra shortward of 1$\\mu$m. In particular, there seems no compelling reason at this time to introduce dust or an additional opacity source in the atmosphere of the SDSS object. The width of the K~I and strengths of the Cs~I lines also indicate that the Sloan object is warmer than Gl~229B. ", "introduction": "Methane or T dwarfs are substellar objects cooler than L and M dwarfs, and have near-infrared (1-2$\\mu$m) spectra dominated by molecular absorption due to water, methane, and pressure-induced molecular hydrogen. Methane is expected to remain an important atmospheric constituent down to the temperature of Jupiter ($\\sim$125~K), where it also is prominent in the infrared spectrum. The prototype of the class is the companion to the nearby M dwarf star Gl~229 (Nakajima et al. 1995, Oppenheimer et al. 1995). Model atmosphere analyses fitting synthetic spectra to detailed spectrophotometric and photometric observations indicate a temperature for this object near or slightly below 1,000~K (Marley et al. 1996, Allard et al. 1996). This past year has seen the discovery of several, similar field T dwarfs, found first in the Sloan Digital Sky Survey (Strauss et al. 1999, S99; Tsvetanov et al. 2000), shortly thereafter in the Two Micron All Sky Survey (Burgasser et al. 1999) data sets, and also in an ESO survey (Cuby et al. 1999). All of the new objects also have 1--2$\\mu$m spectra characterized by the very strong molecular absorbers listed above. Unfortunately, at least at low spectral resolution, the differences among their spectra appear somewhat subtle. The current field surveys by SDSS and 2MASS are magnitude limited, and therefore likely to identify the warmest (highest luminosity) T dwarfs. Still, the known parallax for Gl~570D shows that this object must be substantially cooler than the prototype, yet its infrared spectrum is similar to the others (Burgasser et al. 2000). The relative strengths of the molecular bands are not strongly dependent on the effective temperature. The molecular absorbers also are effective in hiding the weaker atomic line transitions which might be useful discriminants of temperature. At least initially, it is proving difficult to establish spectral types and a temperature sequence for the T dwarfs at the wavelengths where they are easiest to observe. It is possible to observe the brightest of the new T dwarfs at wavelengths significantly shortward of 1$\\mu$m, where the atmospheres may prove to be more transparent. Model calculations indicate that there may be few molecular and atomic opacity sources, and those that are present may be more sensitive to temperature. In particular, the behavior of the alkali resonance doublet features which reside generally in this red part of the spectrum (0.5-0.9$\\mu$m) could be particularly useful in diagnosing the temperature and testing for the formation of dust grains (Burrows and Sharp 1999; Lodders 1999; Tsuji, Ohnaka \\& Aoki 1999; Burrows, Marley \\& Sharp 2000; hereafter, BMS). These papers generally predict that different alkalis should precipitate out as sulfides, salts or other condensates over a range of \\Teff\\ below about 1,500~K, in the order (with decreasing \\Teff) according to BMS of Li first, then Cs, K and Na. Indeed, the latter two alkali features are the most prominent features in the red spectra of the somewhat-warmer late L dwarfs (Kirkpatrick et al. 1999, Mart{\\'{\\i}}n et al. 1999, Reid et al. 2000). The red spectra of field T dwarfs can be relevant to another controversy regarding the spectrum of Gl~229B: The companion object shows too little red flux relative to the model predictions of Marley et al. (1996) and Allard et al. (1996). These authors concluded that an additional opacity source exists shortward of 1$\\mu$m. Golimowski et al. (1998) suggested as the solution that TiO returned to gaseous form (it precipitates out in M-L dwarfs above 2,000~K) in Gl~229B. However, this hypothesis predicts TiO band absorption at the wavelengths seen in M dwarfs, but these are seen neither in late L dwarfs nor in Gl~229B. Griffith et al. (1998) turned to solar system physics for an intriguing answer: they hypothesized a population of small photochemical haze particles analogous to the red Titan Tholins (Khare and Sagan 1984), heated by ultraviolet radiation from the primary M dwarf to temperatures at least 50\\% higher than the \\Teff. Dust is invoked in two other proposed solutions to the Gl~229B spectral slope: Tsuji, Ohnaka \\& Aoki (1999) describe a hybrid atmospheric model with a warm dust layer that effectively blocks short-wavelength flux. Pavlenko, Zapaterio Osorio \\& Rebolo (2000) attempt to fit the red spectrum with a scattering dust opacity which increases sharply to shorter wavelengths. Finally, BMS suggest that the alkali opacity alone -- in particular the broad wings of K and Na -- is the agent which depresses the emergent flux out to 1$\\mu$m. Provided K and Na still exist in atomic form at the relevant temperature, they argued that there is no need to invoke dust or some additional absorber at short wavelengths. SDSS~1624+0029 (hereafter, SDSS~1624), the first field T dwarf, was found in preliminary Sloan Digital Sky Survey data (S99). The discovery paper includes a red spectrum with the Apache Point 3.5-m telescope, with detected flux down to 8000\\AA. The accessible Cs~I features at 8521\\AA\\ and 8943\\AA\\ appeared weak or absent, while both are distinct features in the Gl~229B red spectrum (Oppenheimer et al. 1998). The apparent weakness of the Cs~I features and shallower red spectral slope of SDSS~1624 led BMS to the preferred conclusion that this object ``is tied to a lower core entropy,'' which would normally imply a lower \\Teff\\ than that of Gl~229B. In contrast, the Sloan source showed somewhat shallower methane absorption in the infrared spectrum, suggesting to Nakajima et al. (2000) that it is warmer than the prototype. The red spectrum also shows that a field object can show a similar excess below 1$\\mu$m as the companion object Gl~229B, thus demonstrating that the excess does not depend upon the presence of a nearby source of potential ultraviolet photons (ie., to produce the ``Titan Tholins''). We discuss here a red spectrum of SDSS~1624 obtained with the Keck~II LRIS spectrograph that extends the detection of flux down to 6200\\AA, and has improved signal-to-noise ratio at longer red wavelengths. We believe this observation allows us to test the roles of the alkali metals and the need for dust and/or an additional short-wavelength absorber. ", "conclusions": "" }, "0003/astro-ph0003297_arXiv.txt": { "abstract": "The results of a 7-epoch interferometric imaging study, at wavelengths in the near-infrared K-band, of the carbon star IRC\\,+10216 are presented. The use of non- and partially-redundant aperture masking techniques on the 10-m Keck-I telescope has allowed us to produce images of the innermost regions of the circumstellar dust envelope with unprecedented detail. With roughly twice the resolving power of previous work (\\cite{Weigelt97}; \\cite{Weigelt98}; \\cite{HB98}), the complex asymmetric structures reported within the central \\mbox{0 \\farcs 5 ($\\sim$20\\,R$_\\star$)} have been imaged at the size scale of the stellar disk itself ($\\sim$50\\,mas). A prominent dark lane at a position angle of approximately $120^\\circ$ is suggested to be an optically thick disk or torus of dust which could help to explain IRC\\,+10216's well-known bipolarity at a position angle of $\\sim20^\\circ$. Observations spanning more than a pulsational cycle ($\\sim$638\\,days) have revealed significant temporal evolution of the nebula, including the outward motion of bright knots and clumps. Registering these displacements against the compact bright core, which we tentatively identify as marking the location of the star, has allowed us to determine the apparent angular velocity at a number of points. The magnitudes of the proper motions were found to be in agreement with current estimates of the stellar distance and radial velocity. Higher outflow speeds were found for features with greater separation from the core. This is consistent with acceleration taking place over the region sampled by the measurements, however alternate interpretations are also presented. Although a number of changes of morphology were found, none were clearly interpreted as the condensation of new dust over the pulsation cycle. Unfortunately, ambiguities associated with the true three-dimensional nature of the nebula weaken a number of our quantitative and qualitative conclusions. ", "introduction": "The extreme carbon star IRC\\,+10216 is a classic example of a red giant caught in the act of evolving into a planetary nebula. Its relative proximity, high infrared luminosity, and abundance of molecules found in its dense outflow has resulted in a barrage of observations by astronomers, working across the spectrum, but particularly in the infrared and millimeter/sub-millimeter. Despite all this attention, a good model of what is happening in the innermost regions where the stellar outflow is born and accelerated is still sorely lacking. Numerous studies of molecular lines in the outer envelope (e.g. \\cite{BT93}) have revealed a spherically expanding outflow, a finding which was beautifully confirmed with deep $B$ and $V$ band images of the dust shell in ambient scattered galactic light (\\cite{Mauron99}). However this spherical symmetry, a characteristic of most red giant winds, will likely be broken as the IRC\\,+10216 evolves into a planetary nebula, most of which are elongated or bipolar (e.g. \\cite{ZA86}). The pronounced asymmetry in the innermost regions of the envelope of IRC\\,+10216 reported by numerous high-resolution imaging experiments (most recently \\cite{Weigelt97}; \\cite{coolstars98}; \\cite{Weigelt98}; \\cite{HB98}) and also polarization studies (\\cite{TDW94}; \\cite{KW94}) suggests that the onset of this aspherical flow has already begun, probably within the last few hundred years. With a privileged vantage onto such a brief yet important period in the evolution of a low to intermediate mass (initial mass $\\sim 3$\\,--\\,5\\,M$_\\odot$ \\cite{Guelin95}) star, high resolution observations are crucial in distinguishing between the many competing models for the physical mechanisms underlying the onset of asymmetry in the birth of a planetary nebula. In this paper, we present a 7-epoch diffraction-limited imaging study of the inner dust shell of IRC\\,+10216 in the near-infrared K-band. Although some interpretation of the morphology of the images is given, full radiative transfer modelling results are beyond the scope of this report, and will be presented in a second paper. Instead, we emphasize here the detection and measurement of the motion of features presumably embedded in the outflow. Although proper motions have been reported for near-infrared images of dusty Wolf-Rayet shells (\\cite{wr104}; \\cite{wr98a}), the two order-of-magnitude slower winds around Asymptotic Giant Branch (AGB) stars result in the requirement of longer time bases, and extremely high fidelity mapping schemes. ", "conclusions": "Diffraction-limited images recovered using interferometric techniques from a multi-epoch study spanning more than two years at the Keck~I telescope are presented. Taken in the near-infrared K~Band, the maps have revealed an asymmetric and clumpy structure at angular resolutions exceeding the expected diameter of the stellar photosphere ($\\sim$ 50\\,mas). The most likely morphology for the circumstellar environment is an optically thick circumstellar disk or torus, possibly tilted towards the line of sight in the South revealing the hot inner cavity and emission from the stellar photosphere. The angular separations of clumps of material thought to be in the Northern bipolar lobe have been followed over time, revealing increasing separation from the compact core to the South. Outflow velocities derived from this motion are consistent with estimates of the radial outflow velocity (from CO measurements) and the expected distance. Clumps at greater distances from the Core were found to show increasing velocities, which may be taken as evidence for acceleration in the inner regions; the effects of geometrical projection; or the result of a past event which ejected material with a range of velocities. In addition to the changing separations of components, the appearance of the inner nebula was found to be evolving in other ways, however none were interpreted as evidence for new dust condensation over the pulsation cycle. Further modelling of this system is currently underway, and will be presented in a subsequent paper." }, "0003/astro-ph0003283_arXiv.txt": { "abstract": "Unbiased analysis shows that IRAS data reliably differentiate between the early and late stages of stellar evolution because objects at these stages clearly segregate in infrared color-color diagrams. Structure in these diagrams is primarily controlled by the density distribution of circumstellar dust. The density profile around older objects is the steepest, declining as $r^{-2}$, while young objects have profiles that vary as $r^{-3/2}$ and flatter. The different density profiles reflect the different dynamics that govern the different environments. Our analysis also shows that high mass star formation is strongly concentrated within \\about\\ 5 kpc around the Galactic center, in support of other studies. ", "introduction": "The IRAS all-sky survey provides a unique opportunity to classify the infrared properties of astronomical objects from a homogeneous data set obtained with a single facility. Cross-correlations of various catalogues with the IRAS Point Source Catalogue (PSC) showed that certain Galactic objects tend to cluster in well defined regions of IRAS \\ccds. Notable examples include \\HII\\ regions (e.g. Hughes \\& MacLeod 1989; Wood \\& Churchwell 1989 [WC]) and AGB stars (van der Veen \\& Habing 1988; VH). The reason for this clustering was not understood, nor was it clear whether such biased analysis based on pre-selection implies reliable selection criteria. From detailed modeling of dusty winds we were able to validate the VH selection criterion proposed for AGB stars and to explain its origin (Ivezi\\'c \\& Elitzur 1995; IE95). Here we extend this approach to all Galactic PSC sources in an unbiased analysis of IRAS fluxes without prior selections. ", "conclusions": "Our modeling confirms that the primary reason for different IRAS classes is different dust density distributions. Only 5\\% of the sources require an additional variation of the dust temperature on the shell inner boundary. Spherical shells were employed here as the simplest method to model extended, three dimensional dust distributions. Power laws were used just for illustration purposes and although their actual values should not be taken literally, they provide a good indication of the overall behavior of the density distributions in the four IRAS classes. The presence of disks, expected in various sources, should not significantly affect our conclusions. The standard disk spectrum $F_\\nu \\propto \\nu^{1/3}$ produces a single point in \\ccds\\ and thus cannot effect the observed scatter. This spectrum is modified if the disk is embedded in an extended shell, but then the disk is expected to dominate only at sub-mm and mm wavelengths, longer than those observed by IRAS (Miroshnichenko et al 1999). Flared disk emission can be shown equivalent to that from a flat disk embedded in an appropriate spherical shell. Therefore, such a configuration, too, cannot modify our conclusions. We have queried the SIMBAD database about our sample sources. SIMBAD identifications are occasionally ambiguous (``maser\"), sometimes less than informative (``part of cloud\"), and reliability is not always certain. Nevertheless, they provide useful clues when there are major trends in the data. Class A returned the most decisive results --- 88\\% of its members have possible optical identifications, of which roughly 90\\% are commensurate with AGB stars. Since class A obeys the VH criterion, this corroborates our earlier finding (IE95) that this criterion is both sufficient and necessary for AGB selection. In IE95 we present a thorough analysis of the IRAS fluxes of AGB stars, including color tracks geared specifically for these objects (single chemistry grains, appropriate stellar temperature, etc.). While that analysis was necessary to verify the VH criterion, the $p$ = 2 track presented here captures the essence of the more detailed study, demonstrating that the density distribution is the leading factor in controlling IRAS colors. These circumstellar shells have the steepest density distribution, setting them apart in the \\ccds. In addition, the color tracks are primarily controlled by a single parameter, \\tV, hence the compactness of this class color region. Class B had the same identification rate but its composition not as homogeneous. Planetary nebulae comprise 40\\% of positive and possible identifications. At 13\\%, the only other significant group is ``emission-line stars\", a classification consistent with planetary and reflection nebulae. The remaining identifications span a variety of objects, indicating that detached shells may occasionally form under different circumstances. Significantly, the optical depths required for class B are lower than for the others, the dashed-line tracks in fig.\\ 3 terminate at \\tV\\ = 10. SIMBAD identification rates for the two other classes are much lower --- only 32\\% for class C and 21\\% for D. Among the 38 class C identifications, \\HII\\ regions comprise the only significant group with 15 (40\\%). In class D, too, \\HII\\ regions comprise the single largest group with 23\\% of the identifications, followed by young stellar objects at 11\\% and planetary nebulae at 10\\%. These two classes are clearly dominated by star formation and early stages of stellar evolution, in agreement with their Galactic distributions and with previous attempts to associate star-forming regions with IRAS colors. In the most extensive study of this kind, 83\\% of 1302 IRAS selected sources were found to be embedded in molecular clouds and thus trace star formation, and the selection criterion (\\Ca\\ $\\ge$ 0, \\Cb\\ $\\ge$ 0.4) essentially identifies classes C and D (Wouterloot \\& Brand 1989; Wouterloot et al 1990). The IRAS colors of classes C and D imply dust density distributions flatter than for AGB stars. These colors are spread over large regions, reflecting variation in density profiles in addition to optical depth. The spread in \\Cc\\ colors is smaller than in \\Cb\\ because all shells are optically thin at both 100 and 60 \\mic\\ while their 25 \\mic\\ optical depth can become significant (IE97). Class C colors occupy a sub-region of class D and are produced by the optically-thick end (\\tV\\ \\ga\\ 1) of flat density distributions. Objects whose colors fall in that region can belong to either class C or D. However, among all sources with class D colors, those with high fluxes ($>$ 1000 Jy at both 60 and 100 \\mic) concentrate in a compact color region, hence the separate class C. Since all class C sources have $L >$ \\E3 \\Lo, they are high-mass objects and their concentration in the inner \\about\\ 5 kpc of the Galaxy shows that the high-mass star formation rate decreases with distance from the Galactic center. This result is in agreement with studies of the initial mass function inside and outside the solar circle (Garmany et al 1982; Wouterloot et al 1995; Casassus et al 1999). Our results explain the findings of all earlier studies that were based on object pre-selection, and reveal the limitations of that approach. By example, consider the WC study. After identifying IRAS counterparts of known ultracompact \\HII\\ regions, WC proposed the corresponding colors as a necessary and sufficient selection for all ultracompact \\HII\\ regions and proceeded to estimate the birthrate of O stars. Codella et al (1994) then found that most \\HII\\ regions in a more extended catalog indeed obeyed the WC color criterion. However, that included also diffuse, not just compact \\HII\\ regions, therefore WC overestimated the O star birthrate. This clearly demonstrates the shortcomings of any classification based on a pre-selected population. The unbiased analysis presented here shows that IRAS colors reflect primarily the dust density profiles of circumstellar shells and provide a unique indication of the underlying object only in the case of AGB stars. IRAS data in itself is sufficient for differentiating young and old stellar objects; apart from a limited number of detached shells, IRAS sources belong to two distinct groups as is evident from both the \\ccds\\ and the Galactic distributions: (1) class A sources are at the late stages of stellar evolution and (2) class C and D sources are objects at the early evolutionary stages. This differentiation occurs because the density distributions of dust around young stellar objects have flatter profiles, reflecting the different dynamics that govern the different environments." }, "0003/astro-ph0003410_arXiv.txt": { "abstract": "Relativistic atomic structure calculations for electric dipole ($E1$), electric quadrupole ($E2$) and magnetic dipole ($M1$) transition probabilities among the first 80 fine-structure levels of Fe~VI, dominated by configurations $3d^3, 3d^24s$, and $3d^24p$, are carried out using the Breit-Pauli version of the code SUPERSTRUCTURE. Experimental energies are used to improve the accuracy of these transition probabilities. Employing the 80-level collision-radiative (CR) model with these dipole and forbidden transition probabilities, and Iron Project R-matrix collisional data, we present a number of [Fe~VI] line ratios applicable to spectral diagnostics of photoionized H~II regions. It is shown that continuum fluorescent excitation needs to be considered in CR models in order to interpret the observed line ratios of optical [Fe~VI] lines in planetary nebulae NGC~6741, IC~351, and NGC~7662. The analysis leads to parametrization of line ratios as function of, and as constraints on, the electron density and temperature, as well as the effective radiation temperature of the central source and a geometrical dilution factor. The spectral diagnostics may also help ascertain observational uncertainties. The method may be generally applicable to other objects with intensive background radiation fields, such as novae and active galactic nuclei. The extensive new Iron Project radiative and collisional calculations enable a consistent analysis of many line ratios for the complex iron ions. \\footnote{The complete tables of transition probabilities are available in electronic form at the CDS via anonymous ftp 130.79.128.5} ", "introduction": "The EIE calculations are described in detail in CP99a,b, and compared with previous works. Below, we briefly summarise the qualitative aspects of those results. The next subsection discusses in detail the present calculations for the forbidden and allowed A-values for Fe~VI. \\subsection{Electron Impact Excitation of Fe~VI} With the exception of two calculations two decades ago, there were no other calculations until the recent work reported in CP99a,b. It is difficult to consider the relativistic effects together with the electron correlation effects in this complex atomic system, and the coupled channel calculations necessary for such studies are very computer intensive. The two previous sources for the excitation rates of Fe VI are the non-relativistic close coupling (CC) calculations by Garstang \\etal (1978) and the distorted-wave (DW) calculations by Nussbaumer and Storey (1978). Although the Garstang \\etal (1978) calculations were in the CC approximation, they used a very small basis set and did not obtain the resonance structures; their results are given only for the averaged values. The Nussbaumer and Storey (1978) calculations were in the DW approximations that does not enable a treatment of resonances. Therefore neither set of calculations included resonances or the coupling effects due to higher configurations. Owing primarily to these factors we find that the earlier excitation rates of Nussbaumer and Storey are lower by up to factors of three or more when compared to the new Fe~VI rates presented in CP99b. \\subsection {Radiative transition probabilities} The target expansions in the present work are based on the 34-term wave function expansion for Fe~VI developed by Bautista (1996) using the SUPERSTRUCTURE program in the non-relativistic calculations for photoionization cross sections of Fe~V. The SUPERSTRUCTURE calculations for Fe~VI were extended to include relativistic fine structure using the Breit-Pauli Hamiltonian (Eissner \\etal 1974; Eissner 1998). The designations for the 80 levels (34 LS terms) dominated by the configurations $3d^3$, $3d^24s$ and $3d^24p$ and their observed energies (Sugar and Corliss, 1985), are shown in Table~1. These observed energies were used in the Hamiltonian diagonalization to obtain the R-matrix surface amplitudes in stage STGH (Berrington \\etal 1995). This table also provides the key to the level indices for transitions in tabulating dipole-allowed and forbidden transition probabilities and the Maxwellian-averaged collision strengths from CP99a,b. Examining the new Breit-Pauli SUPERSTRUCTURE calculations we deduce that the computed energy values for levels 7 and 13; levels 11 and 12 in Table~1 of CP99a should be reversed given the level designations, respectively. An indication of the accuracy of the target eigenfunctions may be obtained from the calculated energy levels in Table~1 of CP99a, and from the computed length and velocity oscillator strengths for some of the dipole fine structure transitions given in their Table~2. The agreement between the length and velocity oscillator strengths is generally about 10\\%, an acceptable level of accuracy for a complex iron ion. \\subsubsection{Dipole allowed fine-structure transitions} The weighted oscillator strength $gf$ or the Einstein $A$-coefficient for a dipole allowed fine-structure transition is proportional to the generalised line strength (Seaton 1987) defined, in either length form or velocity form, by the equations \\begin{equation} S_L=\\mid <\\Psi_j\\mid\\sum^{N+1}_{k=1}z_k\\mid\\Psi_i>\\mid ^2 \\end{equation} and \\begin{equation} S_V=\\omega^{-2}\\mid\\left<\\Psi_j\\mid\\sum^{N+1}_{k=1} \\frac{\\partial}{\\partial z_k}\\mid\\Psi_i\\right >\\mid ^2 \\end{equation} where $\\omega$ is the incident photon energy in Ry, and $\\Psi_i$ and $\\Psi_j$ are the wave functions representing the initial and final states, respectively. Using the transition energy, $E_{ij}$, between the initial and final states, $g_if_{ij}$ and $A_{ji} $for this transition can be obtained from S as \\begin{equation} g_if_{ij}=\\frac{E_{ij}}3S \\end{equation} and \\begin{equation} A_{ji}(a.u.)=\\frac 12\\alpha ^3\\frac{g_i}{g_j}E_{ij}^2f_{ij} =2.6774\\times 10^9(E_j-E_i)^3S^{E1}_{ij}/g_j \\end{equation} where $\\alpha=1/137.036$ is the fine structure constant in a.u., and $g_i$, $g_j$ are the statistical weights of the initial and final states, respectively. In terms of c.g.s unit of time, \\begin{equation} A_{ji}(s^{-1})=\\frac {A_{ji}(a.u.)}{\\tau _0} \\end{equation} where $\\tau _0=2.4191^{-17}s$ is the atomic unit of time. We can use experimental transition energy $E_{ij}^{exp}$ to obtain refined $g_if^e_{ij}$ and $A^e_{ji}$ values through \\begin{equation} g_if^e_{ij}=g_if_{ij}\\frac{E_{ij}^{exp}}{E_{ij}^{cal}} \\end{equation} \\begin{equation} A^e_{ji}=A_{ji}(\\frac{E_{ij}^{exp}}{E_{ij}^{cal}})^3 \\end{equation} Computed $gf_L$ and $gf_V$ values, in both the length and the velocity formulations, for 867 $E1$ (dipole allowed and intercombination) transitions within the first 80 fine structure levels are tabulated in Table~2 (a partial table is given in the text; the complete Table~2 is available electronically from the CDS library). Transition probabilities $A_L$ are also given in the length formulation, which is generally more accurate than the velocity formulation in the present calculations. Experimental level energies are used to improve the accuracy of the calculated $gf$ and A-values. All of these $E1$ transition probabilities of Fe~VI were incorporated in the calculation of line ratios when accounting for the FLE effect by the UV continuum radiation field (details below). \\subsubsection{Forbidden electric quadrupole ($E2$) and magnetic dipole ($M1$) transitions} The Breit-Pauli mode of the SUPERSTRUCTURE code was also used to calculate the $E2$ and the $M1$ transitions in Fe VI. The configuration expansion was adapted from that used to optimise the lowest 34 LS terms by Bautista (1996). The spectroscopic configurations,% the correlation configurations and the scaling parameters $\\lambda _{nl}$ for the Thomas-Fermi-Dirac-Amaldi type potential of orbital $nl$ are listed in Table~3 and Table~4 of CP99a. Much effort was devoted to choosing the correlation configurations to optimise the target wavefunctions, within the constraint of computational constraints associated with large memory requirements for many of the $3p$ open shell configurations. The primary criteria in this selection are the level of agreement with the observed values for (a) the level energies and fine structure splittings within the lowest LS terms, and (b) the f-values for a number of the low lying dipole allowed transitions. Another practical criterion is that the calculated $A$-values should be relatively stable with minor changes in scaling parameters. Like the procedure used in the calculation of the dipole allowed and intercombination $E1$ $gf$-values, the experimental level energies are also used to improve the accuracy of the computed $E2$ and $M1$ transition probabilities $A^{E2}$ and $A^{M1}$, given as \\begin{equation} g_jA^{E2}_{ji}=2.6733\\times 10^3(E_j-E_i)^5S^{E2}(i,j) (s^{-1}) \\end{equation} and \\begin{equation} g_jA^{M1}_{ji}=3.5644\\times 10^4(E_j-E_i)^3S^{M1}(i,j) (s^{-1}) \\end{equation} The computed $A^{E2}$ and $A^{M1}$ for all 130 transitions among the first 19 levels are given in Table~3. The results calculated by Garstang \\etal (1978) and by Nussbaumer and Storey (1978) are also given for comparison, where available. For 70 transitions in Table~3 the $A^{E2}$ are much smaller than the $A^{M1}$, by up to several orders of magnitude for some transitions. While for the other 60 transitions, $A^{E2}$ are greater than the $A^{M1}$. The computed $A^{E2}$ and $A^{M1}$ for all the other 1101 transitions within the first 80 levels are given in Table~4 (a partial table is given in the text; the complete Table~4 is available electronically from the CDS library). There are no other calculations in literature for these transitions for comparisons. There are many cases where one of the two transition probabilities is negligible, usually the $A^{E2}$. But the case of Fe~VI is somewhat different from that of Fe~III (Nahar and Pradhan 1996), where the $A^{E2}$ are greater than $A^{M1}$ for nearly half the total number of transitions, especially those with large excitation energies. ", "conclusions": "An extensive calculation of fine structure transition probabilities of Fe~VI is presented for the allowed $E1$ and the forbidden $E2$, $M1$ transitions. An indication of the uncertainties in the computed gf-values is given in the plot of length $gf_L$ vs. the velocity $gf_V$ for 867 $E1$ transitions computed in this work (Fig. 8). It shows an agreement at about 10\\% level for most of the transitions, with no more than about 5\\% of the transitions lying outside that range even for gf-values less than 10$^{-4}$. Combined with previously calculated data for electron impact excitation, a 80-level CR spectral model for line ratios diagnostics is used to predict the effect of collisional and fluorescent excitation (FLE) in planetary nebulae. An illustrative and limited analysis of line ratios is carried out as an example of the use of the atomic data and the model proposed herein. Some of the diagnostics procedures developed earlier by Chen and Pradhan (2000, CP00) are employed to analyse observed line intensties from three planetary nebulae: NGC 6741, IC 351 and NGC 7662. The detailed model and analysis yields a consistent set of diagnostics, for example, for the electron density and effective temperature of the source. It shows that fluorescence effects should be included in CR models of these objects. The method may be used to determine the physical conditions in PN, especially when it is difficult to do so by other methods. For example, we have determined, within certain error limits, both the effective temperature and the emission region distance (via a dilution factor) for IC 351. By combining the line ratios that are independent of the physical conditions of PN (like cases in Table~5), and our method, we are able to estimate possible errors in the observed intensities, both individually and in term of consistency among the set of observed lines. It is expected that the method and procedures described in this paper would be generally applicable to spectral diagnostis of other radiative plasma sources, such as novae and AGN. Finally, some possible uncertainties in our model and procedure, as employed in the present calculations, are as follows. 1) Static conditions are assumed in the CR model, independent of photoionization eqilibrium in the [Fe~VI] regions of PN's. Because [Fe~VI] region is almost the highest ionization state in PN's considered here, this assumption should not carry a large uncertainty. 2) We have assumed inherently in our procedures that there are nearly constant N$e$ and T$e$ in [Fe~VI] regions. This assumption is not entirely true because both N$e$ and T$e$ are a function of the distance from the central star to the [Fe~VI] emission region. However, according to 1), if photoionization equilibrium prevails then there should not be large variations in N$e$ and T$e$. 3) The radiation field should in principle simulate the ionizing stellar radiation. A further refinement of the model proposed herein would be (a) to include a radiation field with proper allowance for the Helium and Hydrogen opacities in various ionization and excitation steps, and (b) in addition to the radiation flux from the central star, resonance fluorescence from H~I and He~II Ly$\\alpha$ should be considered in the model. However as noted earlier, Fe~VI is likely to be in the fully ionized He~III zone, and therefore not greatly susceptible to the effects. 4) Even though the most advanced R-matrix codes are employed in generating atomic data, the atomic data still have some uncertainties, estimated at about 10--20\\%. All data tables are available electronically from the CDS, or via ftp from the authors at: chen@astronomy.ohio-state.edu." }, "0003/astro-ph0003376_arXiv.txt": { "abstract": "s{ We study heterogeneous quantities that efficiently cross-correlate the lensing information encoded in \\comic polarization and large-scale structures recovered from weak lensing galaxy surveys. These quantities allow us to take advantage of the special features of weak lensing effect on CMB \\( B \\)-polarization and of the high (40\\%) cross-correlation between the two data sets. We show that these objects are robust to filtering effects, have a low intrinsic cosmic variance (around 8\\% for small 100 square degrees surveys) and can be used as an original constraint on the vacuum energy density. } ", "introduction": "Secondary CMB anisotropies offer new windows to constrain cosmological models. Lens effects\\cite{TlensEffects} are particularly attractive since they are expected to be one of the dominant effect. Methods to detect the lens effects on CMB\\cite{T4pt} have been proposed recently. Unfortunately all of them suffer from a high sensitivity to cosmic variance. However, this problem can be solved if one considers \\comic polarization instead of temperature anisotropies. Standard cosmological models predict that at small scale, the so called \\( B \\) component of the polarization can be significant only if CMB-lens couplings are present\\cite{B.eq.lent,B2E}. This feature of CMB polarization will allow us to present tools that enhance the detection of lens effect in \\comic by mixing them with galaxy survey \\cite{survlens}. ", "conclusions": "We showed that weak lensing effect on the \\comic \\(B\\) polarization can be embedded in a simple, real space, first order expression. This expression can be used to create mathematical objects that compare the lensing effects up to the last scattering surface to the one up to our galaxy surveys probes. We showed that this objects should produce a significant information, even in realistic (i.e. filtered) situations, with a low intrinsic statistical error. These objects are also expected to be good, unbiased, cosmic parameters tracers. Fig. \\ref{rparam} presents their behavior in the \\((\\Omega_{0},\\Lambda)\\) plane which exhibit a high sensitivity to the vacuum energy density. This is not surprising, since we are probing the length of the optical bench we are working in, which is rather sensitive to \\(\\Lambda\\) \\cite{CalcKappa,LeBon}. \\begin{figure}[h] {\\centering \\begin{tabular}{cc} \\resizebox*{0.45\\textwidth}{!}{\\includegraphics{rparam.ps}} & \\resizebox*{0.45\\textwidth}{!}{\\includegraphics{rGparam.ps}} \\\\ \\end{tabular}\\par} \\caption{\\label{rparam}\\protect\\( \\Cross_{\\bo} \\protect \\) (left figure) and \\protect\\( \\Cross_{\\pabo} \\protect \\) for a CDM model. The filtering beam is 2 arc minutes for all fields. } \\end{figure}" }, "0003/astro-ph0003140_arXiv.txt": { "abstract": " ", "introduction": "Discovery of X-ray emitting hot Interstellar Medium (ISM) with the {\\it Einstein observatory} has drastically renewed the view of early-type galaxies (Forman et al.\\ 1985; Trinchieri et al.\\ 1986; Canizares et al.\\ 1987; Fabbiano et al.\\ 1988; Fabbiano 1989). Since the hot ISM is considered to be gravitationally confined to a galaxy, X-ray observations provide useful knowledge about the depth and the shape of gravitational potential. Furthermore, chemical composition of the ISM tells us about the past history of the galaxy evolution, since stellar mass-loss products and supernova ejecta are accumulated in the ISM\\@. The ISM is metal-enriched by stellar mass loss and Type Ia SNe that are currently observed in elliptical galaxies. The standard supernova rates predict the metallicity of ISM to be as high as several times the solar value (e.g.\\ Loewenstein and Mathews 1991; Ciotti et al. 1991; Renzini et al. 1993). However, the previous measurements of ISM with ASCA have show that the metallicity was less than half a solar (Awaki et al. 1994; Loewenstein et al. 1994; Mushotzky et al. 1994; Matsushita et al. 1994; Arimoto et al. 1997; Matsumoto et al. 1997). Giant early-type galaxies are estimated to have stellar iron abundance of about 1 solar, when observed strong radial gradient of Mg$_2$ index is taken into account (Arimoto et al.\\ 1997). Thus, the X-ray measured abundances of ISM implied that they were even lower than the stellar metallicity. The very low metal abundances in ICM cast some doubts in the current scenario of supernova enrichment and chemical evolution of galaxies. Arimoto et al.\\ (1997) discussed various astrophysical aspects of the inferred low iron abundances in the ISM, concerning the chemical evolution of galaxies and cluster of galaxies, the evolution of gas flows in ellipticals, and the heating of the intra-cluster medium. In the attempt of interpreting the observed results in a consistent way, they explored possibilities of hiding or diluting iron in the ISM\\@. However, none of the mechanisms appeared astrophysically plausible, and they alternatively questioned the reliability of using iron-L lines as a diagnostic tool to infer abundances from X-ray spectra. Recently, Buote et al. (1998), Buote et al. (1999) and Buote (1999) reported that some X-ray luminous galaxies have about 1 solar abundance, employing a multi-temperature plasma model. The high quality X-ray spectra of NGC~4636 (Matsushita et al.\\ 1997) enabled us to study systematic error in our abundance determination, such as the Fe-L atomic physics and abundance ratios. Based on this study, we conclude that the ISM abundance of NGC~4636 is in fact $\\sim 1$ solar. This value is significantly higher than the previous ASCA results. {\\it Einstein} observations have shown that the X-ray luminosities of early-type galaxies scattered by nearly two orders of magnitude, even for the objects with very similar optical properties (e.g.\\ Canizares et al.\\ 1987; Fabbiano et al.\\ 1992). A clue to this long-standing problem has been obtained recently in the deep ASCA observation of the X-ray luminous galaxy NGC 4636, in which a largely extended X-ray emission around the galaxy has been detected. This indicates that NGC 4636 is sitting in the bottom of a large and deep potential structure ($M \\sim10^{13} M_\\odot$) filled with hot tenuous plasma (Matsushita et al.\\ 1998). This is the main reason why the system has acquired its high X-ray luminousity. ASCA and ROSAT observations of other X-ray luminous galaxies also indicate greatly extended halos (Matsushita 1997; Matsushita 2000). Therefore, the presence or absence of such an extended X-ray halo is considered as the main origin of the large scatter in the X-ray luminosity. Later, Davis and White (1996) discovered a correlation between the metal abundance and temperature in ISM\\@. Loewenstein et al.\\ (1994) report that some X-ray fainter galaxies show significantly lower metal abundances. These features suggest that the problem of low metal abundance may be related with the scatter of X-ray luminosity. In this paper, we re-examine ISM metallicity in many early-type galaxies including NGC~4636 and discuss the origin of metals based on the relation between ISM metallicity and ISM luminosity. In section 2, we summarize ASCA observations of relevant galaxies. Section 3 deals with conventional data analysis and results, and in section 4, we examine systematic uncertainties in the abundance measurements such as abundance ratios and Fe-L atomic physics. In section 5, we improve the spectral analysis based on the result of section 4 and study correlation between the ISM metallicity and the ISM luminosity. Section 6 gives discussion of the obtained results. In this paper we adopt for the solar iron abundance the `meteoritic' value, Fe/H $=3.24\\times 10^{-5}$ by number (Anders and Grevesse 1989). ", "conclusions": "We have analyzed ASCA data for 27 giant ($L_B>10^{10}L_\\odot$) early-type galaxies. % X-ray luminosities of these galaxies are $10^{40\\sim 42} \\rm{~ergs^{-1}}$. \\subsection{Summary of results} The spectra of the early-type galaxies have been fitted reasonably well with a $ 0.3-1.0$ keV thermal model (the soft component) and a 10 keV bremsstrahlung~model (the hard component). The hard component should mostly consist of emission from LMXBs such as in early-type spirals and in galactic bulges. % By separating the pure ISM component, we showed that the large scatter in the overall X-ray luminosity of these objects was due to fluctuations in the ISM luminosity. Unusually low abundances ($\\sim 0.3$ solar) were obtained with the R-S plasma model assuming solar-abundance ratios. We carefully examined whether these results are artifacts of the data analysis. We showed that multi-temperature models could not change the results of the most X-ray luminous galaxies, alghouth higher abundances were allowed for X-ray fainter galaxies simply because of poor data statistics. Assumption of non-ionization equilibrium could not solve the problem either. We found that the overlap among emission lines and free-bound continuum cause strong coupling among the abundances of various elements. As a result, when models with variable abundance ratios are fitted to the data, the allowed range of abundance increases dramatically while keeping strong correlations among different elements. Another difficulty is the large uncertainty in the Fe-L atomic physics. Different plasma emission codes give significantly different abundances, and none of them provide acceptable fits to the Fe-L spectra of X-ray luminous galaxies. In contrast, when abundances of $\\alpha$-elements ($A_{\\rm{\\alpha}}$) were fixed, all the plasma codes give similar Fe-abundances ($A_{Fe}$) within 20--30\\%. Therefore, the strong dependence of the abundance on the emission code is mainly attributed to the strong coupling between Fe-L and $\\alpha$-element lines (particularly O-K and Ne-K lines). In order to solve the above coupling problem, we have assumed $\\alpha$-elements to take the solar abundance ratios based on the results by Matsushita et al.\\ (1997). Effect of the free-bound emission of Fe-L on the Si abundance was estimated, and the Si abundance was concluded to be roughly 1 solar within a factor of 2 in X-ray luminous galaxies. Finally, we included additional systematic errors by 20\\% in the Fe-L region of the spectra and relaxed the strong coupling between Fe and $\\alpha$-elements (particularly O and Ne). This has allowed the Si-K lines to make a significant contribution to the $\\chi^2$ minimization for several X-ray luminous galaxies, and as a consequence, the derived $A_{\\rm{\\alpha}}$ and $A_{\\rm{Fe}}$ both became $\\sim 1$ solar within a factor of 2. In addition, discrepancy among the plasma emission codes has been reduced, and the spectral fits have become mostly acceptable with all plasma models. In contrast to the luminous objects, the ISM metallicity in X-ray faint galaxies remained highly uncertain because of very large statistical errors in the Si line intensity. However, strong correlation was again seen between Fe and $\\alpha$-element abundances, and the $A_{\\rm{Fe}}$ vs.\\ $A_{\\rm{\\alpha}}$ confidence regions showed statistically significant dependences on the X-ray luminosity. \\subsection{ISM metallicity and chemical evolution} The hot ISM is considered to be a mixture of the stellar mass-loss products and the supernova ejecta. Therefore, elemental abundances in the ISM measured with X-ray spectra give almost unique information on the contribution from these enrichment processes. More specifically, the ISM abundances can be expressed as, \\begin{equation} z^i =\\frac{\\alpha_*y^i_{*}+\\alpha_{\\rm SN}y_{\\rm SN}^i}{\\alpha_*+\\alpha_{\\rm SN}} \\sim y_{*}^i+\\langle \\frac{\\alpha_{\\rm SN}}{\\alpha_{*}} \\rangle y_{\\rm SN}^i \\end{equation} (Loewenstein and Mathews 1991; Ciotti et al.\\ 1991), where $z^i$ is the mass fraction of the $i$th element, $\\alpha_*$ and $\\alpha_{\\rm SN}$ are the mass loss rates of stars and SNe respectively, and $y_{*}^i$ and $y_{\\rm SN}^i$ are their yields. \\subsubsection{Is the ISM Diluted by the ICM?} We first check the validity of equation (1) in which we assume that the whole ISM in a galaxy consists of matter supplied by the stellar component of the galaxy itself. Elliptical galaxies are often surrounded by a hot ICM, which can be accreted if an inflow is established, and dilute the indigenous ISM (Renzini et al.\\ 1993). Eskridge et al. (1995) discovered a strong correlation between $L_X/L_B$ and potential depth, or stellar velocity dispersion. Later, the existence of largely extended X-ray emission around some X-ray luminous galaxies such as NGC 4636 (Matsushita et al.\\ 1998) and NGC 1399 (Ikebe et al.\\ 1996) suggests that these galaxies are located in the bottom of a largely extended potential well filled with a tenuous plasma. If surrounding metal poor gas is accreting onto a galaxy, the galaxy becomes overluminous in X-rays, as the additional $PdV$ work by the infalling material needs to be radiated away in a quasi-stationary cooling flow and become metal poor. Actually, the most luminous objects, NGC 507, which is surrounded by diffuse extended group ICM (Trinchieri et al.\\ 1997 ), and NGC 1399 tends to have slightly lower ISM abundance than the other X-ray luminous galaxies. The similarity of $A_{\\rm{\\alpha}}$ to $A_{\\rm{Fe}}$ ratio among the X-ray luminous objects suggests that the low metallicity in the most luminous objects may be slightly more diluted by metal poor gas than other X-ray luminous galaxies. X-ray fainter objects show small scatter of $A_{\\rm{Fe}}$ for a given level of $A_{\\rm{\\alpha}}$, as shown in figure 13. The sample contains objects surrounded by dense ICM, such as NGC 1404 and NGC 4374, and also relatively isolated galaxies, such as NGC 720 and IC 1459. This suggest that the ISM in the X-ray faint objects are not significantly diluted by ICM\\@. As summarized in figure 19, X-ray luminous objects show higher metal abundance than X-ray fainter ones. Therefore, dilution may well occur in the X-ray luminous objects, and in this case the abundance difference between the luminous and faint galaxies becomes even larger. \\subsubsection{Constraints on the SNe Ia rate} From the equation (1), combining the latest observed SNe Ia rate with the Hubble constant of $75~ {\\rm km s^{-1} Mpc^{-1}}$ (Cappellaro et al.\\ 1993; Cappellaro et al.\\ 1997; Tammann et al.\\ 1995; van den Bergh et al.\\ 1991), with a nucleosynthesis calculation by Thielemann et al.\\ (1996) and a stellar mass loss rate by Ciotti et al.\\ (1991), the abundance of Fe enriched only by the SNe Ia is predicted to be 2 solar. The measured Fe abundance actually consists of contributions from SNe II and SNe Ia; the former one is supplied via stellar mass loss, while the latter are by SN mass ejection and by stellar mass loss, since we do not know whether stars contain SNe Ia products or not. If we would explain the observed Fe abundance in terms of SNe Ia only, then it gives an upper limit of the SNe Ia rate. The SNe II ejecta are estimated to have an abundance ratio $A_{\\rm{\\alpha}}/A_{\\rm{Fe}}\\sim 3$ (e.g.\\ Thielemann et al.\\ 1996), while SNe Ia products have $A_{\\rm{\\alpha}} \\ll A_{\\rm{Fe}}$. Si and S can be supplied by both SNe II and SNe Ia, but the latter contribution should be negligible as long as the SN Ia rate is low. Thus, contribution from SNe Ia to $A_{\\rm{Fe}}$ (${A_{\\rm{Fe}}}_{\\rm SN Ia}$) is estimated as \\begin{equation} {A_{\\rm{Fe}}}_{\\rm SN Ia}=A_{\\rm{Fe}}-1/3A_{\\rm{\\alpha}}. \\end{equation} On the $A_{\\rm{\\alpha}}$ vs.\\ $A_{\\rm{Fe}}$ plane, we draw lines of $A_{\\rm{Fe}} = \\rm{const.} + 1/3 A_{\\rm{\\alpha}}$ as shown in figure 19 by dotted lines. Each line indicates a constant contribution from SNe Ia. The large uncertainty in $A_{\\rm{\\alpha}}$ in figure 19 indicates that the contribution of SN II is highly ambiguous. However, the confidence contours for $A_{\\rm{\\alpha}}$ vs.\\ $A_{\\rm{Fe}}$ are rather narrow along the lines of constant SN Ia contribution. This is most clearly seen for X-ray fainter galaxies. This means that the contribution of SN Ia is well constrained by the data. By extrapolating these contours along the dotted lines in figure 19 to a small $A_{\\rm{\\alpha}}$ value, we can estimate the SN Ia contribution to $A_{\\rm{Fe}}$. X-ray fainter galaxies indicate the iron abundance attributable to SN Ia, ${A_{\\rm{Fe}}}_{\\rm SN Ia}$ inferred from equation (1), to be less than 0.2 solar for R-S, MEKA, and MEKAL models, and $0.2 \\sim 0.5$ solar for Masai model, respectively. In contrast, X-ray luminous galaxies show ${A_{\\rm{Fe}}}_{\\rm SN Ia}$ to be about 0.5 solar for RS, MEKA, and MEKAL and $\\sim 1$ solar for Masai model, respectively. With the slight difference among the plasma emission codes, we can conclude that the SN Ia contribution to $A_{\\rm{Fe}}$ is positively correlated with the X-ray luminosity. Taking the result by Masai model, which gives the highest Fe abundance, the upper limit of the inferred SNe Ia rate is a factor of 2 and 5 smaller than the Cappellaro's value for the X-ray luminous and fainter galaxies, respectively. As mentioned in section 6.2.1, the ISM in X-ray luminous objects may be simply diluted by ICM\\@. However, the low Fe abundance in X-ray fainter galaxies is a severe problem which contradicts with the SN rate. \\subsubsection{Constraint on stellar metallicity} Optical measurements of stellar metallicity in giant early-type galaxies have been performed only within $\\sim 1 r_e$. In our sample galaxies, the average stellar metallicity over an entire galaxy is typically $0.5\\sim 1 $ solar taking into account the observed metallicity gradient beyond $\\sim 1r_e$ (Arimoto et al.\\ 1997; Kobayashi et al.\\ 1999). Stellar metal ratios are reported to be SNe II-like at least in the center of galaxies (Worthey et al.\\ 1992; Kobayashi et al.\\ 1999). X-ray observations are able to constrain stellar metallicity over an entire galaxy, which is hardly performed with optical observations. Based on equation (1), $A_{\\rm{Fe}}$ gives an upper limit of the stellar Fe abundance. $A_{\\rm{\\alpha}}$, on the other hand, directly reflects the stellar $\\alpha$-element abundance, since SNe Ia contribution is negligible even for Si and S\\@. X-ray luminous galaxies show both $A_{\\rm{Fe}}$ and $A_{\\rm{\\alpha}}$ to be $\\sim 1.0$ solar, which is fairly close to the stellar levels. A few galaxies such as NGC 507 indicate somewhat lower ISM abundance. Thus, the previous severe discrepancy, that the ISM metallicity is even lower than the stellar metallicity (e.g.\\ Awaki et al.\\ 1994; Arimoto et al.\\ 1997), has been mostly removed based on our improved analysis. Since Mg$_2$ indices and colors are similar for all the observed galaxies (Kodama and Matsushita 2000), it is reasonable to assume that X-ray fainter galaxies have similar stellar metallicity to those in X-ray lumious galaxies. In this case, $A_{\\rm{\\alpha}}$, which directly reflects the stellar $\\alpha$-element abundance, should be $\\sim 1.0$ solar in these galaxies. If ISM is diluted in X-ray luminous objects, the $\\alpha$-element abundance in the fainter galaxies may be even higher. Now, figure 19 implies a larger SN II contribution to ISM in more luminous systems, which should mainly come from stellar mass loss. This feature leads us to predict that stellar metallicity in giant early-type galaxies should be SNeII-like even when averaged over the whole galaxy. We can also constrain stellar metallicity in S0 galaxies. There are 5 S0 galaxies in the X-ray fainter group, and we do not detect any difference in the ISM abundance between S0 and elliptical galaxies (figure 13). These S0 galaxies show Mg$_2$ indices close to those of elliptical galaxies. Therefore, abundances of $\\alpha$-elements in the stars in S0 galaxies should be similar within a few tens of \\% to those in elliptical galaxies. If stellar iron abundance is different between S0 and elliptical galaxies, it should be reflected as a significant difference in the ISM abundance, contrary to the observed feature. Therefore, stellar metallicity in S0 galaxies is considered to be similar to that in elliptical galaxies. This conclusion gives an important constraint in modeling the origin and evolution of S0 galaxies. \\subsubsection{Relation to metals in the ICM} It has been recognized that the hot ICM in clusters of galaxies contains a large amount of iron (Hatsukade 1989; Tsuru 1993; Fukazawa et al.\\ 1998). Clearly, this iron must have been supplied by member galaxies, with early-type galaxies as main contributors (Arnaud et al.\\ 1992). In evaluating the metal enrichment process independently of the amount primordial gas, it is convenient to introduce ``iron mass to light ratio'' (IMLR), which is the ratio of the total iron mass in a cluster ICM over the total optical luminosity of cluster galaxies. The observed value of IMLR appears to be remarkably constant among the clusters at $0.01-0.02$ (Tsuru 1993; Arnaud et al.\\ 1992; Renzini 1993). This IMLR level is too high to be explained by the stellar mass loss from member galaxies (Tsuru 1993). The total released mass from stars over the Hubble time is $10 \\sim20\\%$ of the stellar mass in a galaxy, and its contribution to the cluster IMLR is estimated to be $(1-2) \\times 10^{-3}$ assuming the metallicity of the mass-loss gas to be about 1 solar. This is only $ \\sim 1/10$ of the observed IMLR in the ICM\\@. Therefore, the supernova enrichment is definitely required to explain the observed IMLR\\@. The IMLR values are calculated for the ISM, based on our updated abundance determination, and compared with the cluster level in figure 20. The ISM mass within $4 r_{e}$ is estimated from a single $\\beta$ model, assuming $\\beta=0.5$ and a core radius = 1 kpc which are typical values obtained by {\\it Einstein} (Forman et al.\\ 1985; Trinchieri et al 1986). Individual galaxies indicate low IMLR values compared with clusters. This is not surprising in a sense, because the ISM mass is at most a few percent of the stellar mass while the ICM is 2--5 times more massive than the stellar content in clusters. An important feature seen in figure 20 is that the stellar mass-loss and SNe Ia can release only $ \\sim 20$\\% of the total stellar mass in the Hubble time (e.g.\\ Renzini et al.\\ 1993). Therefore, with the present ISM metallicity in the outflow, the accumulated IMLR would reach at most 10 times higher level and unable to explain the observed cluster metals. In this view, the ISM metallicity is too low regardless of the estimated supernova rate. This consideration suggests that the ISM metallicity may not directly reflect the SN Ia rate, even though the measured values themselves are secure quantities. \\subsection{Interpretation of the remaining problem, low SNe Ia contribution} The discrepancy between stellar and ISM abundances has been mostly removed. There are various models which try to explain how a large amount of Fe in ICM could have been synthesized even with the present SNe Ia rate, including the Capperaro's value (e.g.\\ Renzini et al.\\ 1993; Arimoto et al.\\ 1997; Ishimaru 1997). Considering the large Fe contribution from SNe Ia to the ICM (Fukazawa et al.\\ 1998), the SNe Ia rate must have been higher in the past. We need to examine if this picture can explain why SNe Ia contribution is lower in the X-ray fainter galaxies, as shown in figure 19. Since X-ray luminous and X-ray fainter galaxies show similar optical properties (Kodama and Matsushita 2000), the difference in the metal abundance cannot be simply explained by the present SN Ia frequency. The larger amount of ISM in X-ray luminous galaxies suggests a longer accumulation time. In this case, the higher metal abundance may reflect a gradual change in the SN Ia rate in the past. However, as shown in figure 20, within the optical radii, the gas mass in X-ray luminous galaxies is only a few \\% of the stellar mass. This means that it takes only a few Gyr to build up the ISM with the present stellar mass loss rate (e.g.\\ Ciotti et al.\\ 1991) even in the luminous galaxies. A quick decline in the type Ia supernova rate in such a short time scale is very unlikely. We propose here that the low SN Ia contribution in the ISM of X-ray fainter galaxies is due to a loss of SN Ia products to the intergalactic space (see also Fujita 1997). This is mainly caused by the shallow gravitational potential. The stellar mass-loss component may also escape from these systems in the form of a mild outflow (Ciotti et al.\\ 1994). The SN Ia products would escape together with the mass-loss gas or may have higher escaping probability than it. In X-ray luminous galaxies, the extended gravitational potential (Matsushita 1997; Matsushita et al.\\ 1998) would trap considerable fraction of the supernova product. Therefore, the spatial extent of SN Ia products probably reflect the overall potential structure beyond the optical scale. In closing, we should note that the dynamical motion of supernova bubbles in early-type galaxies may be significantly different from that in spirals. Progenitors follow a random velocity field in early-type systems instead of the disk rotation, and the ambient pressure is dominated by the hot plasma (i.e.\\ ISM) rather than by cold interstellar gas. We hope future high resolution imaging and spectroscopy in X-rays will bring us rich information about the motion and structure of metal-rich gas in early galaxies. \\bigskip We would like to thank Nobuo Arimoto and Kuniaki Masai for valuable suggestions on this work. This work was supported by the Japan Society for the Promotion of Science (JSPS) through its Postdoctoral Fellowship for Research Abroad and Research Fellowships for Young Scientists. \\clearpage" }, "0003/astro-ph0003189_arXiv.txt": { "abstract": "We investigate the influence of an exotic fluid component (``quintessence\") on the angular size-redshift relation for distant extragalactic sources. Particular emphasis is given for the redshif $z_{m}$ at which the angular size takes its minimal value. We derive an analytical closed form which determines how $z_m$ depends on the parameter of the equation of state describing the exotic component. The results for a flat model dominated by a ``quintessence\" are compared in detail with the ones for the standard open model dominated by cold dark matter. Some consequences of systematic evolutionary effects on the values of $z_{m}$ are also briefly discussed. It is argued that the critical redshift, for all practical purposes, may completely be removed if such effects are taken into account. ", "introduction": "Recent data from SNe Ia have provided strong evidence for an expanding Universe speeding up, rather than slowing down (Riess et al. 1998; Perlmutter et al. 1998). These observational evidences have stimulated great interest in a more general class of cosmological models driven by nonrelativistic matter and a ``quintessence\" component, i.e., an exotic fluid with an arbitrary equation of state $p_{x} = \\omega_x \\rho_x$ ($\\omega_x \\geq -1$), which probably dominates the bulk of matter in the observed Universe. Examples of these models include the evolving scalar field (Ratra \\& Peebles 1988; Frieman et al. 1998; Caldwell et al. 1998), the smooth noninteracting component (xCDM) (Turner \\& White 1997; Chiba et al. 1997), and still the frustated network of topological defects in which $\\omega_x = - \\ \\frac{n}{3}$, being $n$ the dimension of the defect (Spergel \\& Pen 1997). Some observational aspects of these models have extensively been analyzed in the literature. For example, Waga \\& Miceli (1999), combining statistics of gravitational lenses and SNe Ia data have found $\\omega_x < -0.7$ ($68\\%$ cl) for a spatially flat Universe. Efstathiou (1999), by using high-z Type Ia supernovae and cosmic microwave background anisotropies, has found $\\omega_x < -0.6$ (2$\\sigma$) if the Universe is assumed to be spatially flat, or $\\omega_x < -0.4$ (2$\\sigma$) for universes of arbitrary spatial curvature. Perlmutter et al. (1999) constrained $\\omega_x < -0.6$ ($95\\%$ cl) using large-scale structure and SNe Ia in a spatially flat geometry. However, although carefully investigated in many of their theoretical and observational aspects, the influence of a ``quintessence\" component in some kinematic tests like the angular size-redshft relation still remains to be analyzed. In principle, the lensing effect of the expanding Universe may provide strong limits on the free parameter describing this exotic component. Therefore, it is interesting to explore how uncertaints in distance measures of extragalactic objects and their underlying evolutionary effects may alter the standard cold dark matter results. On the other hand, the existing angular size data for distant objects are until nowadays somewhat controversial, specially because they envolve at least two kinds of observational dificulties. First, any high redshift object may have a wide range of proper sizes, and, second, evolutionary and selection effects probably are not negligible. Indeed, the $\\Theta(z)$ relation for some extended sources samples seems to be quite imcompatible with the predictions of the standard FRW model when the latter effects are not taken into account (Kapahi 1987;1989). There have also been some claims that the best fit model for the observed distribution of high redshifts extended objects is provided by the standard Einstein-de Sitter universe ($q_o={1 \\over 2}$, $\\Omega_\\Lambda=0$) with no significant evolution (Buchalter et al. 1998). However, all these results are in contradiction with the recent observations from type Ia supernovae. Indeed, such data seem to ruled out world models filled only by baryonic matter, and more generally, any model with positive deceleration parameter. The same happens with the corresponding bounds using the ages of old high redshift galaxies (Dunlop et al. 1996; Krauss 1997; Alcaniz \\& Lima 1999). The case for compact radio sources is also of great interest. These objects seem to be less sensitive to evolutionary effects since they are short-lived ($\\sim 10^{3} yr$) and much smaller than their host galaxy. Initially, the data from a sample of 82 objects gave remarkable suport for the Einstein-de Sitter Universe (Kellerman 1993). However, some analysis suggest that, although compatible with an Einstein-de Sitter Universe, the Kellerman data cannot rule out a significant part of the $\\Omega_{M}-\\Omega_{\\Lambda}$ plane (Kayser 1995). Some authors have also argued that models where $\\Theta(z)$ diminishes and after a given $z$ remains constant may also provide a good fit to Kellerman's data. In particular, by analysing a subset of 59 compact sources within the same sample, Dabrowski et al. (1995) found that no useful bounds on the value of the deceleration parameter $q_o$ can be derived. Indeed, even considering that Euclidean angular sizes ($\\Theta \\sim z^{-1}$) are excluded at 99$\\%$ confidence level, and that the data are consistent with $q_o=1/2$, they apparently do not rule out higher values of the deceleration parameter (Stephanas \\& Saha 1995). More recently, based in a more complete sample of data, which include the ones originally obtained by Kellermann, it was argued that the $\\Theta(z)$ relation may be consistent with any model of the FRW class with deceleration parameter $\\leq 0.5$ (Gurvits et al. 1999). In this context, we discuss the influence of a ``quintessence\" component (Q-model) on the angular size-redshift relation. Particular emphasis is given for the critical redshift at which the angular size of an extragalactic source takes its minimal value. In the limiting case ($\\omega_x=-1$), the results previously derived by Krauss \\& Schramm (1993) for a flat universe with cosmological constant ($\\Lambda$CDM) are recovered. For comparison, we also consider the case of an open model dominated by nonrelativistic matter (OM). ", "conclusions": "" }, "0003/astro-ph0003006_arXiv.txt": { "abstract": "We model the evolution of the galactic bulge and of the bulges of a selected sample of external spiral galaxies, via the multiphase multizone evolution model. We address a few questions concerning the role of the bulges within galactic evolution schemes and the properties of bulge stellar populations. We provide solutions to the problems of chemical abundances and spectral indices, the two main observational constraints to bulge structure. ", "introduction": " \\begin{itemize} \\item{} it is not needed to require the merging and/or accretion of external material to reproduce the main observed characteristics of bulges; the self--collapse of the protogalaxy seems to be sufficient to give the correct time scales, enrichment history and populations to reproduce observations not only for the Milky Way but also for external spiral galaxies. \\item{} The observed properties of bulge stellar populations are well reproduced in detail if a correct evaluation of the contribution from the various regions involved in the bulge formation and evolution are considered, we mean the halo, bulge and core populations. \\item{} There is no evidence of an analogy with elliptical galaxies: the apparent geometrical similarity does not correspond to any functional and physiological equivalence between these two families of astronomical objects. \\item{} The mean age of bulges varies according to the galactic type of the parent galaxy: there is a strong correlation between the bulge populations and the whole galactic evolution. \\end{itemize} In conclusion, not only it makes sense to figure bulge and disc evolution as two joint components of an intrinsically unitary structure, along the Hubble sequence, but it appears to be the only realistic way to obtain a good comprehension of bulge structure.", "conclusions": "The multiphase model, inserted in the dissipative collapse scenario following ELS picture, is able to reproduce the observed properties of bulges. The determination of the history of star formation as the principal ingredient of the evolution is useful in the understanding of the bulge variations in the Hubble sequence. In massive bulges corresponding to early type galaxies, the intensity of the star formation maximum is large enough to consume the whole gas, by preventing the subsequent star formation. The young stellar populations do not exist or they exist in very small proportions in comparison with the bulk of stars formed in earlier times. In later type galaxies, less massive, this maximum is not so large, and the star formation rate has almost a constant intensity. These results are in agreement with observations in the K-band from Seigar \\& James (1998a,b), whose results seem to indicate that the morphological types must be controlled by the star formation ratio between past and present and not by the ratio bulge/disc, because it does not present any correlation with the morphological type T. Chemical results show bulges very similar in their abundances at the present time and for the time when the star formation maximum occurs, in agreement with estimations obtained from colors data (Peletier \\& Balcells, 1996; deJong, 1996a,b). Mean abundances have also been calculated by showing subsolar values and similar for all bulges independently of the Hubble type, the arm class and/or the luminosity of their host galaxies. With a method reverse to the one adopted by Jablonka et al. (1996) and other authors, we compute the spectral indices Mg$_{2}$ and $<\\rm Fe>$ from our stellar generation results, by using SSP synthesis models. Our results are in agreement with a set of observations for bulges in the plane Mg$_{2}-<\\rm Fe>$. We may explain why the first of this indices depend on the Hubble type or the total luminosity of the bulge, while the index $<\\rm Fe>$ does not show a similar trend. The evolution and the star formation histories of our bulge models are crucial for this respect: the mean age of the stellar populations in bulges are different, even if the formation of the bulge himself starts at the same moment. The star formation is more continuous in the central regions of the late type galaxies than in the early type ones, by producing more iron in the stars of the first ones. This mixing of the old metal poor populations and young and iron rich populations in the late type galaxies produce the same result for $<\\rm Fe>$ than is produced by the old metal poor population created in the early type galaxies in the first phases of star formation. The magnesium is produced instead very quickly in these last galaxies, by increasing the abundance of the stars. Therefore the index Mg$_{2}$ show a clear trend with the morphological type because both ages and abundances have the same effect, by increasing the value of Mg$_{2}$. The model shows that it is not needed the consider bars or radial flows to explain the features of bulges or their formation. Theoretical results for intermediate type bulges show the highest ratio H$_{2}$/HI. These galaxies present the strongest observed bars, as from K band data; this suggests that bars may appear and disappear in the evolution of the bulge. They would contribute to produce a new starburst phase in bulges, but it may occurs several Gyr after the formation of the bulge. In this case, the gas flows into the central region, and younger stars will be created in the bulge. Their effects would be measured if the quantity of inflow gas is large enough or comparable to the mass of stars already created in the bulge. Otherwise, their influence will be negligible in the indices Mg$_{2}$ and $<\\rm Fe>$ and we could only reveal them if we use other kind of data such the calcium triplet or H$\\beta$. We conclude presenting our answer to the questions addressed in the" }, "0003/astro-ph0003094_arXiv.txt": { "abstract": "We present arguments which show that in all likelihood mesogranulation is not a true scale of solar convection but the combination of the effects of both highly energetic granules, which give birth to strong positive divergences (SPDs) among which we find exploders, and averaging effects of data processing. The important role played by SPDs in horizontal velocity fields appears in the spectra of these fields where the scale $\\sim$4~Mm is most energetic; we illustrate the effect of averaging with a one-dimensional toy model which shows how two independent non-moving (but evolving) structures can be transformed into a single moving structure when time and space resolution are degraded. The role of SPDs in the formation of the photospheric network is shown by computing the advection of floating corks by the granular flow. The coincidence of the network bright points distribution and that of the corks is remarkable. We conclude with the possibility that supergranulation is not a proper scale of convection but the result of a large-scale instability of the granular flow, which manifests itself through a correlation of the flows generated by SPDs. ", "introduction": "In the traditional view of solar convection seen at the sun's surface, three scales play the main roles: {\\it granulation} (1~Mm) which shows up as an intensity pattern most probably first seen by Herschel in 1801 \\cite{BLD84}, {\\it supergranulation} ($15 - 30$~Mm) which appears in (but not only) dopplergrams of the full disk of the sun as a pattern of horizontal velocities and which was first noticed by Hart\\cite*{Hart56b}\\nocite{Hart56a} and confirmed by Leighton et al. \\cite*{LNS62}, and {\\it mesogranulation} ($3 - 10$~Mm) observed by November et al. \\cite*{NTGS81} on Doppler measurements of vertical velocities. While the dynamics of granulation is rather well understood, its scale being controlled by the balance of radiative diffusion of heat and convection, the origin of the two other scales remains largely mysterious. The ionization of helium was often invoked to explain these two scales since the first and second ionizations of this atom occur at depths similar to the mesogranular and supergranular scale respectively. However, this geometrical explanation is based on a ``laminar view\" of solar convection which is, on the contrary, very strongly turbulent. The aim of the present paper is to investigate the dynamics of scales larger than that of the granulation, using the horizontal flows given by granular motions determined by the new algorithms described in Roudier et al. (1999). Briefly, these algorithms, Local Correlation Tracking on binarized images (LCT$_{\\rm bin}$) or Coherent Structure Tracking (CST), allow us to increase noticeably the spatial and temporal resolutions of the surface velocity fields; typically, we can bring the spatial grid size down to 0\\arcsec7 and the time step down to 5~mn. Thus, we first concentrate on \\MG\\ (Sect.~2) and show the major role played at this scale by strong positive divergences (SPDs) and by averaging procedures. It turns out, indeed, that what has been described in previous work as \\MG\\ results from a combination of a physical phenomenon (SPDs among which are found exploding granules) and a data processing effect applied to a turbulent flow (averaging). Since each author had his own technique for averaging data, results have been rather confusing and no clear-cut description of \\MG\\ has emerged. We show here that when averaging is properly controlled, no quasi-steady flow can be detected in the mesoscale range. We then proceed (Sect.~3) with the investigation of the transport properties of the mesoscale flows and show that the \\SG\\ scale appears when the positions of concentrations of corks are compared to the positions of network bright points. We conclude the paper with a discussion of a model which seems to explain many of the observations and the interactions of the three scales. ", "conclusions": "To conclude this paper we would like to emphasize the point that the picture of surface solar convection which has been popular in the literature and where supergranulation advects \\MG\\ which in turn advects granulation is too simple and misleading. It is obviously too simple because it tentatively describes turbulent convection with three scales instead of a continuous spectrum of scales. It is misleading as it reduces the nonlinear interaction between scales to a simple advection as for instance the kinematic models of Simon et al. 1991. We have shown here that no quasi-steady flow could be identified at the \\MG\\ scale and that after a three hour averaging the mean flow shows a component at the \\SG\\ scale while it keeps a small-scale (5\\arcsec) component. This latter component seems to be the source of the former large-scale flow. Our results therefore suggest a scenario where the large-scale supergranular flow is generated directly by the granular flow through a large-scale instability which fixes the scale, in space and time, of \\SG. We thus conjecture that nonlinear interaction between flows at the granulation scale, in other words Reynolds stresses, are sufficient to drive flows at the supergranulation scale and that the energy released by the recombination of ionized helium plays no part. This scenario needs now to be tested for its various implications, theoretical as well as observational." }, "0003/astro-ph0003431_arXiv.txt": { "abstract": "We investigated the central gas systems of E/S0 galaxies by making use of the WFPC2 images of the Hubble Space Telescope archive. We searched the gas systems that were traced by the dust with a new method of making color excess images in $F555W - F814W$ ($V-I$). Out of 25 sample galaxies, we detected gas system in 14 galaxies. The dust was newly detected in two galaxies that were thought to contain no dust based on single band, pre-refurbishment data. The full extents of the gas systems are 0.1 to 3.5~kpc, and the masses of the gas, log~$M_{\\rm gas}$~[$M_{\\odot}$], are 4.2 to 7.2. The AGN activity is well correlated with existence of the gas systems. None of galaxies without the gas systems show the AGN activity. On the other hand, some galaxies with the gas systems show the AGN activity; optical AGN activities are shown in 5 out of 11 galaxies of which AGNs are optically studied, and radio activities are shown in 6 out of 14 galaxies. This shows that the AGN activity is driven with the gas system. ", "introduction": "Early-type galaxies, E/S0 galaxies, are in general gas poor. However, some of them contain significant gas and dust in their central regions; it may be related to AGN activity, which make it be an interesting object. \\cite{jaf93} (1993, 1996) presented an example; they showed an HST image of central dust ring in NGC~4261, and they discussed the dust ring in connection with AGN activity. Similar dust ring structures were found in NGC~1439 and NGC~4494 by \\cite{for95} (1995) and \\cite{car97} (1997), and in NGC~7052 by \\cite{vdm98} (1998). \\cite{vdo95} (1995) studied statistically the central gas system which were traced by the dust in early-type galaxies using the HST archives. However, they used data obtained by pre-COSTAR WF/PC camera with PC mode, and they used only one band data of $F555W$ ($V$ band) and without color data. Therefore, their estimates are limited by a spherical aberration of pre-refurbishment of the HST, and a detailed morphology may be missed. Though they pointed out a correlation between the gas systems and the AGN activity using radio data, we thought further discussions with post-COSTAR HST images are necessary. Though \\cite{car97} (1997) studied extensively central regions of elliptical galaxies with $F555W-F814W$ color images by WFPC2, they did not discuss the relation of gas systems with AGN phenomena in detail. Recently, \\cite{ver99} (1999) investigated the central dust in early-type galaxies based on new HST observations with multibands for a radio-loud sample. We made use of WFPC2 archival images with two bands, $F555W$ ($V$ band) and $F814W$ ($I$ band), and investigated the central gas system traced by dust more precisely by making and analyzing color excess images for early-type galaxies with or without AGN activities. In section~2 we describe sample and a new method of data analysis. We give results in section~3; how frequent the gas systems are observed, and how much are those masses, and what are those morphologies. We discuss the characteristics of the gas system in connection with AGN activity in section~4. Conclusions are given in section~5. ", "conclusions": "We investigated the gas systems in the central regions of 25 E/S0 galaxies by analyzing the color excess images made from HST archives. The conclusions are: \\noindent 1. We detect the central gas systems that can be traced by the dust in 14 out of 25 sample galaxies. We find the gas system in two galaxies that were reported to be dust-free by \\cite{vdo95} (1995) which was based on single band, pre-refurbishment data. \\noindent 2. The morphologies of the gas systems are grouped into three: small disk, large disk, and irregular system. For small disk, full extent of $d$ is 0.1 to 0.5~kpc and gas mass of log~$M_{\\rm gas}$~[$M_{\\odot}$] is 4.3 to 5.9. For large disk, $d$ is about 2~kpc and log~$M_{\\rm gas}$~[$M_{\\odot}$] is about 7. For irregular system, $d$ is 0.1 to 3.5~kpc and log~$M_{\\rm gas}$~[$M_{\\odot}$] is 4.2 to 7.1. \\noindent 3. The AGN activity is well correlated with existence of the gas system which can be found in the HST WFPC2 archival image. \\noindent 4. The AGN activity is triggered by an infalling gas system with a probability of about 50\\%, and the probability is higher if the morphology of the gas system is seen as irregular type, though the statistics are poor." }, "0003/astro-ph0003482_arXiv.txt": { "abstract": "We have developed a generalization of the Zeldovich approximation (ZA) that is exact in a wide variety of situations, including plannar, spherical and cilyndrical symmetries. We have shown that this generalization, that we call complete Zeldovich approximation (CZA), is exact to second order at an arbitrary point within any field. For gaussian fields, the third order error have been obtained and shown to be very small. For statistical purposes, the CZA leads to results exact to the third order. ", "introduction": " ", "conclusions": "" }, "0003/astro-ph0003161_arXiv.txt": { "abstract": "The duration of starbursts is important for determining how they are regulated and the impact they have on their environment. Starbursts contain numerous prominent star clusters which typically comprise in total $\\sim$ 20\\%\\ of the ultraviolet light, embedded in a diffuse glow of recently formed stars responsible for the dominant $\\sim 80$\\%. Hubble Space Telescope images have been obtained for four starburst galaxies in order to determine their burst duration from the ages of their star clusters. Preliminary results for NGC~3310 are reported here. The $UBI$ colors of its clusters and diffuse light give consistent results. The clusters have a broad range of colors, consistent with a population of instantaneous bursts with ages ranging from 0 to a few 100 Myr. The diffuse light has a narrow range of color consistent with continuous star formation over timescales ranging from $\\sim 10$ to 100 Myr. Hence NGC~3310's starburst has lasted $\\sim 100$ Myr, or about 10 times the crossing time. Other results noted in the literature also indicate bursts lasting several times longer than the crossing time. These results suggest that starbursts are not self-extinguishing flashes. Rather they are sustainable, perhaps self-regulated. ", "introduction": "Over the past few decades starbursts, brief intense episodes of massive star formation, have been recognized as important agents of galaxy evolution. However, despite many journal pages per year of attention, there is much we do not know about the inner workings of starbursts. In particular: how long do they last? Estimates in the literature have been obtained from a variety of techniques and range from Myr to Gyr time scales (Mas-Hesse \\&\\ Kunth \\&\\ 1999, and Coziol, Barth \\&\\ Demers 1995). {\\em A priori\\/} the minimum expected duration is the crossing time \\tcross: the time it would take for a disturbance to travel from one end of the starburst to the other. Local starbursts typically have size scales of $\\sim 0.2 - 2$ Kpc and velocity dispersions of $\\sim 30 - 300$ km/s, and hence crossing times ranging from $\\sim 1$ Myr to $60$ Myr, with 10 Myr being typical. Starburst duration is important for determining the efficiency that metals are released into the IGM, and for determining the total number of bursts that a galaxy can undergo. Knowing the duration of starbursts will give us a better understanding about how bursts evolve. If the durations are similar to \\tcross\\ it would suggest that they are self-extinguishing explosions, destroyed by their energy output produced in a non-equilibrium fashion. Conversely, durations much longer than \\tcross\\ would indicate that starbursts are sustainable, perhaps even being in equilibrium: i.e.\\ ISM inflow $\\approx$ star formation rate (SFR) + outflow. ", "conclusions": "Other researchers have also reported extended star formation durations within starbursts. The work of Whitmore \\etal\\ (1999) on the spectacular merging ``Antenna'' system (NGC~4038/4039; discussed in these proceedings by Miller, 2000) reveals clusters with ages up to $\\sim 200$ Myr within the starburst. This is largely consistent with the merger timescale of 200 Myr, and a few times the disk crossing timescale of $\\sim 50$ Myr (Barnes 1988). Calzetti \\etal\\ (1997; 2000) report clusters in the blue compact dwarf (BCD) galaxy NGC~5253 with photometric and spectroscopic ages up to $\\sim$ 10 Myr, a few times larger than the starburst crossing timescale of $\\sim 3$ Myr. Walborn \\&\\ Blades (1997) discuss in detail the complex age distribution of stars and star clusters in the nearest external starburst: 30 Doradus. Grebel \\&\\ Chu (2000) expand on a case in point, reporting an age of 25 Myr for Hodge 301, a neighbor of R136a which has an age of 2.5 Myr. The projected turbulent velocity dispersion crossing time between these clusters is $\\la 1$ Myr. Clearly even small starbursts have an extended history. Ground based observations of BCDs and H{\\sc ii} galaxies yield similar results. Their strong emission line spectra indicate that they must contain a substantial young ionizing population. However, the colors of their starbursts (determined after carefully subtracting the underlying populations, and correcting for extinction), are clearly too red to have been produced by a young ($\\la 10$ Myr) instantaneous burst. Instead they are consistent with continuous star formation over $\\sim$ 10 to 100 Myr timescales (Telles \\&\\ Terlevich 1997; Marlowe \\etal\\ 1999). This amounts to at least a few crossing times in most starbursts. Contrary claims for relatively short burst durations have been made, particularly with regard to Wolf-Rayet galaxies (Schaerer, Contini, \\&\\ Kunth, 1999) and BCDs (Mas-Hesse \\& Kunth 1999) which have star formation durations estimated to be $\\la 4$ Myr\\footnote{Mas-Hesse \\&\\ Kunth note that in several cases BCD data are consistent with $\\sim$ 20 Myr continuous star formation.}. In these studies timescales are often spectroscopically constrained by the presence of WR features, which are washed out for durations $\\ga 5$ Myr. However, the spectroscopy can be misleading as suggested by Fig.~4. An observer would usually center the slit on the brightest source within the galaxy, which is likely to be a cluster, and preferably a young one. Such a cluster could dominate the entire spectrum, the majority of the starburst being outside of the slit. This would give the impression that the whole starburst is dominated by a short duration burst. Cases in point are NGC~5253 (Walsh \\&\\ Roy, 1987; Schaerer \\etal\\ 1997) and NGC~3310 (Pastoriza \\etal\\ 1993), both of which show WR spectra, but only over a fraction of the starburst. \\begin{figure} \\plotone{meurergf4.eps} \\caption{Cartoon view of a typical slit orientation (large rectangle) on a starburst. The contours represent the distribution of the smooth light. The various filled symbols are the clusters with symbol shape representing age, and size representing brightness. } \\end{figure} Another strong voice for short burst durations is Elmegreen (2000) who claims that star formation occurs over a timescale of only one or two crossing times, citing various results in the Galaxy and LMC. The difference here may in part be due to the velocity used to calculate the crossing time. Elmegreen uses the sound velocity whereas I have quoted measured velocity dispersions of the ISM of the starbursts. The latter is supersonic, resulting in lower crossing times, and a better representation of the true speed that a disturbance travels within the star forming medium. If starbursts do last many crossing times there are some important implications. Firstly, starbursts are sustainable, perhaps even self-regulating. The upper-limit to the effective surface brightness of starbursts reported by Meurer \\etal (1997) and Lehnert \\&\\ Heckman (1996) is further evidence of this regulation, although it is not clear what the regulating mechanism is. Second, burst durations may be longer than the typical ISM expansion timescales seen in galactic winds (Martin 1998). Hence, galactic winds are likely to occur in a previously fractured ISM. This should increase the efficiency of metal ejection into the intergalactic medium compared to simulations which usually are modeled in smooth undisturbed media. {\\bf Acknowledgements}. I gratefully acknowledge Tim Heckman and Claus Leitherer, my collaborators on this project." }, "0003/astro-ph0003357_arXiv.txt": { "abstract": "Long slit spectra of the molecular outflow Herbig-Haro (HH) 46/47 has been taken in the J and K near-infrared bands. The observed \\hh\\ line emission confirms the existence of a bright and extended redshifted counter-jet outflow southwest of HH46. We show that this outflow seems to be composed of two different emission regions which have distinct heliocentric velocities, in contrast with the optical appearance, and which implies an acceleration of the counter-jet. The observed \\iron\\ emission suggests an average extinction of 7--9 visual magnitudes for the region associated with the counter-jet. Through position-velocity diagrams, we show the existence of different morphologies for the \\hh\\ and \\iron\\ emission regions in the northern part of the HH\\,46/47 outflow. We have detected for the first time high velocity (-250\\kms) \\iron\\ emission in the region bridging HH46 to HH47A. The two strong peaks detected can be identified with the optical positions B8 and HH47B. The \\hh\\ excitation diagrams for the counter-jet shock suggest an excitation temperature for the gas of $T_{\\rm{ex}} \\approx 2600$. The lack of emission from the higher energy \\hh\\ lines such as the 4-3~S(3) transition, suggests a thermal excitation scenario for the origin of the observed emission. Comparison of the \\hh\\ line ratios to various shock models yielded useful constraints about the geometry and type of these shocks. Planar shocks can be ruled out whereas curved or bow-shocks (both J and C-type) can be parametrised to adjust our data. ", "introduction": "For its tight spatial alignment of the emission knots and its clear visibility against the progenitor dark globule, HH\\,46/47 is one of the most remarkable examples of the classical Herbig-Haro object. Being optically bright, this HH complex has been the subject of many studies of its spectral, photometric and kinematical properties after its discovery by Schwartz (1977). The morphology of the Gum nebula region (Dopita et~al. 1982) shows the existence of a very collimated jet that is composed of several components which extend about 0.39 pc on the sky at the assumed distance of 350 pc (Eisl{\\\"o}ffel and Mundt 1994). A young T Tauri star still obscured by the surrounding dust envelope sits at the center of this system. This star illuminates a cloud of dust revealing a reflection nebula from which the optically visible HH\\,46 and HH\\,47(A,B,C,D) jet components emerge. The bipolar nature of this outflow was first ascertained by Graham and Elias (1983), on discovering blue and redshifted outflowing shocked gas to either side of the central source. Schwartz et al. (1984) reported the first proper motions for some dense knots along the flow suggesting the occurrence of episodic ejections of gas material from the energy source. Combining the radial velocity measurements (Meaburn \\& Dyson 1987, Hartigan et al. 1993, Morse et al 1994) with a detailed study of the proper motions of the knots in the HH\\,46/47 jet (Eisl{\\\"o}ffel \\& Mundt 1994), a space velocity of about 300\\kms is found for the jet. While several imaging, spectroscopic and proper motion studies have been carried out in the optical for the blueshifted part of HH\\,46/47, there has been considerably less work done for the redshifted part of the flow. This well-collimated counter-jet flow located southwest of the source was discovered by Reipurth and Heathcote (1991) in [SII] $\\lambda\\lambda$6717/6731 line emission and found to be redshifted with a heliocentric radial velocity of about +125\\kms. K-band imaging of the redshifted counter-flow has been performed by Eisl\\\"offel et al. (1994), together with \\hh\\ 1-0~S(1) (2.1218\\mic) and \\iron\\ (1.644\\mic) line imaging. More recently, proper motion studies of the \\hh\\ emission have been carried out for the \\hh\\ knots in HH\\,46/47 (Micono et~al. 1998). ", "conclusions": "The physical conditions along the redshifted counter-jet and the blueshifted HH46 to 47A outflow were investigated through observation of velocity-position diagrams and spectra in the J (\\ironsix\\ and \\irontwo) and K (several \\hh\\ transitions) infrared bands. The observed \\hh\\ line emission confirms the existence of a bright and extended redshifted counter-jet outflow southwest of HH46. Moreover, there seems to be compelling evidence for acceleration of the counter-jet from 27\\kms\\ to 55\\kms\\ along the southwestern emission lobe as it travels away from the source. The observed \\iron\\ emission suggests a near uniform extinction in the HH46 and the counter-jet region of $\\approx 7$--9 visual magnitudes, whereas the extinction calculated for the blueshifted knot B8 and HH47B objects is negligible. The \\hh\\ excitation temperatures derived from the 1-0/2-1~S(1) ratio are in the range 1800--2100~K for all the objects, with the exception of the HH46 object for which the excitation temperature is 1200~K. The unusual low 2-1~S(1) line flux in this object was probably caused by the limited region covered by our slit in this object which may have sampled only the wing sections of the bow-shock structure. The inclusion of the 3-2~S(3) in the \\hh\\ excitation diagrams for the counter-jet shock yields an excitation temperature for the gas of $T_{\\rm{ex}} \\approx 2600$K. The non-detection of higher energy transitions of \\hh\\ in the K band, precludes several non-thermal excitation mechanisms such as fluorescence or \\hh\\ reformation. The observed 2-1/1-0~S(1) and 3-2/1-0~S(1) line ratios were found consistent with models of J and C-type bow shocks, whilst ruling out planar shocks. In the counter-jet, the shape parameter for C-type bows can be constrained to the range $1.5 < s < 2.0$ whereas J-type bows need to be made extremely wide, $1.2 < s < 1.5$, to adjust well our data." }, "0003/astro-ph0003211_arXiv.txt": { "abstract": "We have used a multiwavelength data set from the Canadian Galactic Plane Survey (CGPS) to study the Galactic \\h region KR~140, both on the scale of the nebula itself and in the context of the star forming activity in the nearby W3/W4/W5 complex of molecular clouds and \\h regions. From both radio and infrared data we have found a covering factor of about 0.5 for KR~140 and we interpret the nebula as a bowl-shaped region viewed close to face on. Extinction measurements place the region on the near side of its parent molecular cloud. The nebula is kept ionized by one O8.5~V(e) star, VES~735, which is less than a few million years old. CO data show that VES~735 has disrupted much of the original molecular cloud for which the estimated mass and density are about 5000~$M_{\\odot}$ and 100~cm$^{-3}$, respectively. KR~140 is isolated from the nearest star forming activity, in W3. Our data suggest that KR~140 is an example of spontaneous (i.e., non-triggered) formation of, unusually, a high mass star. ", "introduction": "\\label{sec:intro} Massive OB stars are almost always found in clusters. In fact, accumulating observational evidence suggests that most stars, regardless of mass, actually form as members of some kind of group, cluster, or association. While there has been a lot of work to understand the processes involved in the formation of a single star (e.g., Shu, Adams, \\& Lizano 1987), theories of cluster formation are still in their infancy (see the recent reviews by Elmegreen \\etal\\ 2000, and Clarke, Bonnell \\& Hillenbrand 2000). Nevertheless, many important issues have already been identified. Foremost among these is the question of whether the formation of clusters, particularly ones with OB stars, are triggered by an external agent (Elmegreen 1992) such as expanding \\h regions (Elmegreen \\& Lada 1977), or colliding molecular clouds (Loren 1976, 1977; Scoville, Sanders \\& Clemens 1986; Usami, Hanawa, \\& Fujimoto 1995). Despite triggered or sequential star formation being theoretically and intuitively appealing, a major problem is to determine unambiguously a cause and effect relationship, because of the long time scales (and time lags) of the processes involved. There is observational evidence for regions that have had triggered star formation, both on large scales (e.g., IC~1396, Patel \\etal\\ 1998) and on small scales (e.g., IC~1805, Heyer \\etal\\ 1996). However, there are other young star forming regions, like Taurus, where no evidence of a trigger can been found. These are often in modest-sized molecular clouds and contain only lower mass stars. The Perseus Arm star forming regions W3/W4/W5 (Westerhout 1958) have been studied extensively over the last twenty years (e.g., Lada \\etal\\ 1978; Braunsfurth 1983; Digel \\etal\\ 1996; Normandeau, Taylor \\& Dewdney 1997; Heyer \\& Terebey 1998). They are often considered to be the archetypical examples of how the formation of massive star clusters can be triggered by the influence of other nearby clusters. For example, W3 is thought to be have been triggered by the expansion of W4 (Dickel 1980; Thronson, Lada, \\& Hewagama 1985; van der Werf \\& Goss 1990) and there is evidence that the expansion of W5 is also triggering star formation (Vall\\'{e}e, Hughes, \\& Viner 1979; Wilking \\etal\\ 1984). In new high resolution multiwavelength (radio and mid-infrared) data of the W3/W4/W5 complex from the Canadian Galactic Plane Survey (CGPS; English \\etal\\ 1998) we have identified a star forming region containing a single O star. Figure~\\ref{fig:1420all} shows a 1420~MHz continuum image from the CGPS pilot project (Normandeau \\etal\\ 1997). The circled area is the \\h region in question, KR~140 ($l=133.425^{\\circ}$, $b=0.055^{\\circ}$; Kallas \\& Reich 1980). This region appears to be completely separate from the vigorous star formation going on nearby in W3, although it is in the same Perseus arm molecular complex (see context in CO images in Heyer \\etal\\ 1998). What is unusual is that this massive star seems to have been formed spontaneously. In this paper we present and analyze the multiwavelength data on KR~140 in order to quantify the properties of this region of spontaneous massive star formation. As described in \\S~\\ref{sec:data} these data have sufficient resolution (1$^{\\prime}$) to resolve this \\h region for the first time and reveal a fairly symmetrical structure. Based on first impressions, we thought that KR~140 might prove to be a ``textbook'' spherical \\h region. Instead, we find that KR~140 is likely a bowl-shaped \\h region (an example of a blister geometry: Israel 1978, Yorke \\etal\\ 1989; \\S~\\ref{sec:blist} and \\S~\\ref{subsub:ion}, \\S~\\ref{sub:irmod}). We have found that the \\h region is kept ionized by an O8.5V(e) star, VES~735, and is at a distance of 2.3$\\pm$0.3~kpc from the Sun (Kerton \\etal\\ 1999). We analyze the ionized component of KR~140 in \\S~\\ref{sec:kr} and make various estimates of the age in \\S~\\ref{sec:age}. The dust, molecular, and atomic components of KR~140 are examined in \\S~\\ref{sec:ir}, \\S~\\ref{sec:mol-cld}, and \\S~\\ref{sec:hi}, respectively. In \\S~\\ref{sec:discuss} we discuss star formation in KR~140 in the context of the Perseus arm and the possible accompanying cluster. ", "conclusions": "\\label{sec:concl} We have utilized our multiwavelength data set (all at a resolution of about 1$^{\\prime}$) to study not only the physics of the \\h region itself, but also, since the data are from the larger CGPS survey, to study KR~140 in the context of the overall picture of star formation in the Perseus spiral arm. We find no evidence for a mechanism that triggered the formation of the O8.5V(e) star, VES~735, and its (largely unseen) cluster. We therefore conclude that this region formed spontaneously out of its parent molecular cloud, independent of the more vigorous star formation in W3 and W4 nearby. Our data of KR~140 are consistent with the model of a bowl-shaped region viewed close to face on. Extinction measurements to the exciting star, VES~735, and nebula show that the \\h region is quite likely on the near side of its molecular cloud. We have not observed any champagne flow, and cannot rule out other geometries. KR~140 has an age less than a few million years. We have estimated that the original molecular cloud had a mass of $4.9 \\times 10^3~M_{\\odot}$ and an average density about 100~cm$^{-3}$, which classifies it as a dwarf molecular cloud. This makes KR~140 even more unusual as it is a rare example of an O star that has formed in a cloud with a mass less than 10$^4~M_{\\odot}$. There is tentative evidence that the molecular material is undergoing a large scale flow towards KR~140. Follow-up observations are needed to pursue this idea. There are four IRAS point sources associated with the KR~140 complex, one of which is a possible protostar candidate and another a molecular core. Near infrared observations of KR~140 are needed to find and study the young cluster that likely formed along with VES~735." }, "0003/astro-ph0003205_arXiv.txt": { "abstract": "Weak gravitational lensing observations probe the spectrum and evolution of density fluctuations and the cosmological parameters which govern them but are currently limited to small fields and subject to selection biases. We show how the expected signal from large-scale structure arises from the contributions from and correlations between individual halos. We determine the convergence power spectrum as a function of the maximum halo mass and so provide the means to interpret results from surveys that lack high mass halos either through selection criteria or small fields. Since shot noise from rare massive halos is mainly responsible for the sample variance below 10$'$, our method should aid our ability to extract cosmological information from small fields. ", "introduction": "Weak gravitational lensing of faint galaxies probes the distribution of matter along the line of sight. Lensing by large-scale structure (LSS) induces correlation in the galaxy ellipticities at the percent level (e.g., \\cite{Mir91} 1991; \\cite{Blaetal91} 1991; \\cite{Kai92} 1992). Though challenging to measure, these correlations provide important cosmological information that is complementary to that supplied by the cosmic microwave background and potentially as precise (e.g., \\cite{JaiSel97} 1997; \\cite{Beretal97} 1997; \\cite{Kai98} 1998; \\cite{Schetal98} 1998; \\cite{HuTeg99} 1999; \\cite{Coo99} 1999; \\cite{Vanetal99} 1999; see \\cite{BarSch00} 2000 for a recent review). Indeed several recent studies have provided the first clear evidence for weak lensing in so-called blank fields (e.g., \\cite{Vanetal00} 2000; \\cite{Bacetal00} 2000; \\cite{Witetal00} 2000). Weak lensing surveys are currently limited to small fields which may not be representative of the universe as a whole, owing to sample variance. In particular, rare massive objects can contribute strongly to the mean power in the shear or convergence but not be present in the observed fields. The problem is compounded if one chooses blank fields subject to the condition that they do not contain known clusters of galaxies. Our objective in this {\\it Letter} is to quantify these effects and to understand what fraction of the total convergence power spectrum should arise from lensing by individual massive clusters as a function of scale. In the context of standard cold dark matter (CDM) models for structure formation, the dark matter halos that are responsible for lensing have properties that have been intensely studied by numerical simulations. In particular, analytic scalings and fits now exist for the abundance, profile, and correlations of halos of a given mass. We show how the convergence power spectrum predicted in these models can be constructed from these halo properties. The critical ingredients are: the Press-Schechter formalism (PS; \\cite{PreSch74} 1974) for the mass function; the NFW profile of \\cite{Navetal96} (1996), and the halo bias model of \\cite{MoWhi96} (1996). Following \\cite{Sel00} (2000), we modify halo profile parameters, specifically concentration, so that halos account for the full non-linear dark matter power spectrum and generalize his treatment to be applicable through all redshifts relevant to current galaxy ellipticity measurements of LSS lensing. This calculational method allows us to determine the contributions to the convergence power spectrum of halos of a given mass. Throughout this paper, we will take $\\Lambda$CDM as our fiducial cosmology with parameters $\\Omega_c=0.30$ for the CDM density, $\\Omega_b=0.05$ for the baryon density, $\\Omega_\\Lambda=0.65$ for the cosmological constant, $h=0.65$ for the dimensionless Hubble constant and a scale invariant spectrum of primordial fluctuations, normalized to galaxy cluster abundances ($\\sigma_8=0.9$ see \\cite{ViaLid99} 1999) and consistent with COBE (\\cite{BunWhi97} 1997). For the linear power spectrum, we take the fitting formula for the transfer function given in \\cite{EisHu99} (1999). \\begin{figure*}[t] \\centerline{\\psfig{file=fig1.eps,width=0.9\\textwidth}} \\caption{Halo model power spectra. (a) Dark matter power spectrum at redshift of 0.5. (b) Convergence power spectrum with $z_{s}=1$. The sum of the Poisson (dotted line) and correlation (dashed) contribution (solid line) compares well with that predicted by the non-linear power spectrum based on the PD fitting function (long-dashed line). In (a), the linear matter density power spectrum is shown with a dot-dashed line.} \\label{fig:pd} \\end{figure*} ", "conclusions": "" }, "0003/astro-ph0003033_arXiv.txt": { "abstract": "We present Keck HIRES spectra of 6 late-M dwarfs and 11 L dwarfs. Our goal is to assign effective temperatures to the objects using detailed atmospheric models and fine analysis of the alkali resonance absorption lines of \\cs and \\rb. These yield mutually consistent results ($\\pm$ 150 K) when we use ``cleared-dust'' models, which account for the removal of refractory species from the molecular states but do not include dust opacities. We find a tendency for the \\rb line to imply a slightly higher temperature, which we ascribe to an incomplete treatment of the overlying molecular opacities. The final \\teff we adopt are based on the \\cs fits alone, though the \\rb fits support the \\cs temperature sequence. This work, in combination with results from the infrared, hints that dust in these atmospheres has settled out of the high atmosphere but is present in the deep photosphere. We also derive radial and rotational velocities for all the objects, finding that the previously discovered trend of rapid rotation for very low mass objects is quite pervasive. To improve on our analysis, there is a clear need for better molecular line lists and a more detailed understanding of dust formation and dynamics. ", "introduction": "M dwarfs are among the faintest and coolest stellar objects. Their optical spectra are dominated by molecular band absorption. The major opacity sources in the optical regions are TiO and VO bands, which produce a pseudo-continuum with the atomic lines superposed. Recent surveys (e.g. DENIS, 2MASS) have turned up even cooler objects, including brown dwarfs (\\cite{Delfosse99}; \\cite{Kirkpatrick99a}). The optical spectra of these objects are still characterized by molecular band absorption; however, the TiO and VO bands, which constitute the defining characteristic of the M spectral class, become very weak or are absent, while the metal hydrides begin to dominate. This is presumably because of the depletion of heavy elements in their cool photospheres through the formation (and perhaps subsequent gravitational settling) of dust grains, especially solid VO and perovskite (CaTiO$_3$) (\\cite{Sharp90}; \\cite{Tsuji96a}; \\cite{Allard98}). For this reason, a new spectral class, `L', has been proposed for these very cool low-mass objects (\\cite{Martin97}; \\cite{Kirkpatrick99a}; \\cite{Martin99b}). To understand these very cool, low mass stellar and substellar objects, it is necessary to assign to them an effective temperature scale. The construction of \\teff sequences has been attempted in recent years by comparing the spectra of these objects to synthetic ones generated by atmospheric models. Comparisons have been made on the basis of spectral energy distributions (SEDs) (\\cite{Tinney98}; \\cite{Ruiz97}) and infrared colors (\\cite{Leggett98}). Attempts have also been made to construct a temperature scale using the molecular lines in the observed spectra, since these are sensitive to photospheric temperature. \\cite{Tinney98} do this in the optical by ranking the objects in order of TiO, VO, CrH and FeH equivalent widths; \\cite{Delfosse99} pursue a similar program in the infrared by using H$_2$O indices. \\cite{Tokunaga99} do so with a spectral color index based on moderate dispersion spectroscopy in the K band. \\cite {Kirkpatrick99a} have performed a detailed analysis of low dispersion optical spectra as a first step towards defining the subclasses of the L spectral type. It is found that the models fit the observed SEDs and colors fairly well overall, and an effective temperature sequence is indeed derivable from the fits. However, as \\teff decreases from early to late spectral types, some molecular bands which comprise most of the low resolution spectral features of these cool objects first increase and then decrease in strength due to dust formation (both through depletion from the gas phase and heating effects). The broadband infrared color indices (eg. J-K or I-K) saturate at very cool temperatures, and eventually also reverse direction (though it seems that I-Z or I-J might be monotonic with temperature). The molecular line lists in the models are not fully satisfactory, which could affect the temperature scale. It is therefore useful to have an alternative and complementary way of estimating effective temperatures. We present such a method here: fitting synthetic spectra to the profiles of strong atomic alkali lines. Their extremely low ionization potential allows the alkali metals (Li, Na, K, Cs and Rb) to remain neutral only at temperatures below 3000K. More importantly, their atomic resonance lines occur in the red and are thus prominent in very cool photospheres, displaying deep cores and broad wings in high-resolution spectra. Finally and most importantly, the alkali metals are relatively undepleted by dust formation (\\cite{Burrows99}), and their resonance lines are apparently formed in photospheric layers located above the region affected most by dust. For these reasons, the resonance features of the alkali metals are excellent temperature indicators in cool photospheres, and grow monotonically in strength with decreasing temperature in the range of interest. We therefore attempt to assign effective temperatures to 17 late-M and L dwarfs by fitting model spectra to their \\cs and \\rb high-resolution line-profiles. In principle, the \\na and \\pot lines can be also be used for this purpose. However, we have not considered them in this paper, for the following reasons. The \\pot line in some of our sample objects is very broad, extending beyond one whole echelle order. To fit models to it, neighboring orders must be patched together, a task fraught with sources of error and thus considered too risky for the fine-analysis we undertake here. We also do not have observations of the \\na lines for a number of our sample objects. Therefore, we consider only the \\cs and \\rb lines in this paper. We should point out that this method has its own weak points. The treatment of dust, or of the background molecular opacities around the lines, do influence the apparent strength of the line by changing the nearby continuum. While it is unlikely that the molecular opacities are wildly off, we show that dust can have a major effect. The synthetic spectra we use are generated by the atmospheric models of Allard and Hauschildt (hereafter AH). The models come with 3 treatments: ``standard (no-dust)'', ``dusty'' and ``cleared-dust'' (discussed in \\S3). The first does not account for dust formation at all, while the second takes dust formation into account by considering both dust grain opacities and photospheric depletion of dust-forming elements (physically, this implies a haze of dust that is mixed throughout the atmosphere). The third ignores dust opacities but accounts for photospheric depletion, simulating the state of affairs when dust forms and then gravitationally settles below the photosphere. All 3 treatments treat line broadening in the same fashion. We find that the ``standard'' and ``dusty'' models do a poor job of reproducing the optical observations -- the alkali lines in the first are too narrow and in the second they are too weak to fit most observations. The ``cleared-dust'' models, however, fit the data much better. ", "conclusions": "In general we find that the cleared-dust model profiles fit the observed \\cs and \\rb resonance lines reasonably well (with errors of $\\pm$ 50K in the fitting). We further find that, in a given object, the effective temperatures derived independently from \\cs and \\rb agree to within $\\pm$ 150K, with the \\rb fit consistently giving the higher temperature. We think this is due to the JOLA modeling of overlying molecular opacities, and the overlap of a JOLA band with the \\rb line. We find that the \\cs and \\rb analyses independently imply the same ordering of objects by effective temperature. The temperature scale derived from these lines may be a little hotter than that which seems to be coming out of consideration of low dispersion spectral energy distributions. This could be due to our approximate treatment of dust or variations in gravity and/or metallicity, or perhaps the infrared treatment needs adjustment. Our final temperatures are based on the \\cs lines. On the whole, we find good agreement between the models and the data with log[g]=5.0 and [M/H]=0.0. With these values, we estimate our errors in \\teff to be $\\pm$50K, since our model grid has a spacing of 100K. However, an analysis of ``no-dust'' AH models indicates that gravity variations in the log[$g$]=4.5-5.5 regime may lead us to infer a \\teff greater by $\\lesssim$ 300K than the real value, for low-gravity objects. Metallicity effects, which act in an opposite sense to gravity ones, may also affect our \\teff values. ``Cleared-dust'' models with varying gravities and metallicites are needed to resolve this issue. In the cases where the models do not reproduce the resonance lines very well, various effects may be responsible - collisional broadening effects due to metallicity and/or surface gravity variations, imperfectly modeled molecular opacity overlapping the line, or a low S/N ratio in the data. We find that the molecular lines are not modeled as well as the resonance lines are, and that the fit to the molecular lines generally worsens as one moves lower in effective temperature. Both effects are expected, given the comparative paucity of well-determined parameters for many of the observed molecules. Obviously it is desirable to improve our modeling of these lines. These affect the analysis of the objects at both high and low spectral dispersions (and in both the optical and near-IR). The treatment of dust is still problematic. Although it must be true that there is rather little dust opacity in the far red (or we would not observe such strong atomic lines), it is not clear that our assumption of $no$ opacity is valid. We must also reconcile the results in the far red with those in the near infrared, which apparently require more dust opacity. We suggest that the dust has largely settled out of the portion of the atmosphere sampled by the alkali resonance lines, but is present in the lower photosphere where the near infrared is formed. \\cite{Tsuji99} have made a similar suggestion in the context of Gl 229B. It will be important to study line profiles in the near infrared to help sort this out. The suggestion is that stratified dust models are worth pursuing. At the moment, the height or extent of such dust stratification can probably be better informed by observations than theory. This is the weak point in our analysis; the addition of some dust opacity would tend to reduce the inferred temperatures. It is not clear whether the good fits to the line shapes can be preserved as dust opacity is added (and a significant amount is needed to affect these very strong lines). The temperature scale found here led \\cite{Martin99b} to propose a subclass designation scheme for the L spectral class in which L0 occurs at about 2200K, and each subclass is 100K cooler. For the coolest L dwarf in common with \\cite{Kirkpatrick99a} (2MASSW 1632+1904), there is a disagreement in temperature. We predict that there is still a gap of a few hundred degrees between such objects and the T dwarfs, while \\cite{Kirkpatrick99a} believe the gap is quite small. On the other hand, variations in gravity and/or metallicity may be causing us to infer an artificially high \\teff for this object. Depending on the resolution of this issue, either the \\cite{Martin99b} or \\cite{Kirkpatrick99a} L star classification scheme should be modified in order that the L stars extend down to the appearance of methane in the K band. One clear and fairly model independent result from our analysis is that the average rotation velocity of very low mass stars gets higher and higher as one moves down through the bottom of the main sequence. It is clear that as hydrogen burning begins to turn off near the substellar boundary, the magnetic braking which affects all higher mass convective stars is also weakening. Our sample is entirely from the field, and some objects are several hundred Myr old (although there is certainly an observational bias against finding very old objects, particularly if they are brown dwarfs). It appears that there is relatively little angular momentum evolution among these objects, and that they are typically born with relatively rapid rotation. Their lack of magnetic activity is seen directly (through a lack of \\hal emission), as well as indirectly in their rapid rotation. The effort to understand these very low mass objects has just begun." }, "0003/astro-ph0003343_arXiv.txt": { "abstract": "We present an analytic model for the fully nonlinear power spectrum $P$ and bispectrum $Q$ of the cosmological mass density field. The model is based on physical properties of dark matter halos, with the three main model inputs being analytic halo density profiles, halo mass functions, and halo-halo spatial correlations, each of which has been well studied in the literature. We demonstrate that this new model can reproduce the power spectrum and bispectrum computed from cosmological simulations of both an $n=-2$ scale-free model and a low-density cold dark matter model. To enhance the dynamic range of these large simulations, we use the synthetic halo replacement technique of Ma \\& Fry (2000), where the original halos with numerically softened cores are replaced by synthetic halos of realistic density profiles. At high wavenumbers, our model predicts a slope for the nonlinear power spectrum different from the often-used fitting formulas in the literature based on the stable clustering assumption. Our model also predicts a three-point amplitude $Q$ that is scale dependent, in contrast to the popular hierarchical clustering assumption. This model provides a rapid way to compute the mass power spectrum and bispectrum over all length scales where the input halo properties are valid. It also provides a physical interpretation of the clustering properties of matter in the universe. ", "introduction": "Two conceptual pictures of galaxy clustering have been examined in the literature, the continuous hierarchical clustering model and the power-law cluster model (Peebles 1980, \\S 61). In the hierarchical clustering model, which has emerged as the accepted model over the past two decades, galaxy clustering is characterized by power-law correlation functions: the $N$-point correlation function $\\xi_N$ scales with configuration size as $\\xi_N \\propto r^{-\\gamma_N} \\propto \\xi_2^{(N-1)}$, where $ \\gamma_N = (N-1) \\gamma $ and the two-point correlation function goes as $\\xi_2=\\xi\\propto r^{-\\gamma} $. The hierarchical model is motivated by the observed power-law behavior $\\gamma\\approx 1.8$ of galaxy correlations (Groth \\& Peebles 1977; Fry \\& Peebles 1978), with a theoretical basis in a self-similar, scale-invariant solution to the equations of motion (Davis \\& Peebles 1977). The alternative power-law cluster model has an even longer history (Neyman \\& Scott 1952; Peebles 1974, 1980; McClelland \\& Silk 1977; Scherrer \\& Bertschinger 1991; Sheth \\& Jain 1997; Valageas 1998; Yano \\& Gouda 1999). In this model, galaxies are placed in spherical clumps that are assumed to follow a power-law density profile $\\rho(r) \\propto r^{-\\epsilon}$, with the centers of the clumps distributed randomly. The resulting two-point correlation function is also a power law with a logarithmic slope $\\gamma = 2\\epsilon - 3$. While it is possible to reproduce the observed two-point function by an appropriate choice of the power index $\\epsilon = (3+\\gamma)/2 \\approx 2.4$, Peebles and Groth (1975) pointed out that this model produces a three-point function that is too steep to be consistent with observations in the Zwicky and Lick catalogs. In an earlier paper (Ma \\& Fry 2000a), we have shown that in the nonlinear regime, the three-point correlation function $\\zeta=\\xi_3$ of the cosmological mass density field does not exactly follow the prediction $\\zeta\\propto \\xi^2$ of the hierarchical clustering model. These conclusions are drawn from study of high resolution numerical simulations of a cold dark matter (CDM) model with cosmological constant and of a model with scale-free initial conditions $P(k)\\sim k^n$ with $n=-2$. In experiments replacing simulation dark matter halos with power-law density profiles, $\\rho(r)\\sim r^{-\\epsilon}$, we have demonstrated that the behavior of the correlation functions in the nonlinear regime are determined by the halo profiles, but that it is not possible to match both the two- and three-point correlations with a single slope $\\epsilon$. These results differ from the predictions of both of these two conceptual models. In this paper, we expand our previous study of the nonlinear two- and three-point correlation functions by investigating a new prescription that takes into account the non-power-law profiles of halos, the distribution of halo masses, and the spatial correlations of halo centers. Each of these ingredients has been well studied in the literature. We find that this halo model provides a good description of the two- and three-point correlation functions in both the $n=-2$ and CDM simulations over the entire range of scales from the weak clustering, perturbative regime on large length scales, to the strongly nonlinear regime on small length scales. Our result is approximately hierarchical over an intermediate range of scales, thus uniting the two pictures. An independent recent study by Seljak (2000), which appeared during completion of this work, has also examined the two-point power spectrum in a similar construction and has found that this type of approach can reproduce the power spectrum in the CDM model. The analytic model proposed here can be used to compute the two- and three-point correlation functions and their Fourier transforms, the power spectrum and bispectrum, over any range of scale where the input halo properties are valid. In a subsequent paper (Ma \\& Fry 2000c), we study the predictions of this analytic halo model for the asymptotic nonlinear behavior of the $N$-point correlation functions and the pairwise velocities and examine the conditions required for stable clustering. The outline of this paper is as follows. In \\S 2 we describe the three input ingredients of the model: halo density profiles, halo mass functions, and halo-halo correlations. In \\S 3 we assemble these ingredients and construct analytic expressions for the two-point correlation function $\\xi(r)$ and the power spectrum $P(k)$. In \\S 4 we do the same for the three-point correlation function $\\zeta(r_1,r_2,r_3)$ and its Fourier transform, the bispectrum $B(k_1,k_2,k_3)$. In \\S 5 we test the validity of this new model by comparing its predictions with results from numerical simulations of an $n=-2$ scale free model and a low-density CDM model with a cosmological constant ($\\Lambda$CDM). We also present results of the synthetic halo replacement technique used to enhance the numerical resolution. In \\S 6 we discuss further the physical meanings and implications of the model. In particular, we elaborate on two important implications of this model: deviations from the common assumptions of stable clustering and hierarchical clustering. Section~7 is a summary. ", "conclusions": "We have constructed a physical model for the correlation functions of the mass density field in which the correlations are derived from properties of dark matter halos. We have described in detail the input, construction, and results of this model in \\S 2 -- \\S 5. We now examine more closely its physical meanings and implications in three separate regimes. On scales larger than the size of the largest halo, the contributions from separate halos dominate, and (by design) the model reproduces the results of perturbation theory. On intermediate scales, $ 1/R_* \\la k \\la 1/R_s(M_*) $, because of the exponential cutoff in the mass function $dn/dM$ at the high mass end, the contribution to the volume integrals in \\eq{xi2} is dominated by the large-$r$ regime where the halo profiles are roughly $r^{-3}$. The correlation functions therefore behave approximately as predicted by the power-law model with $\\epsilon=3$, i.e., $\\Delta \\propto k^{2\\epsilon-3} \\sim k^3$ and $Q \\propto k^{3-\\epsilon} \\sim $ constant. This is why $Q$ exhibits an approximately flat plateau at intermediate $k$ in the bottom panels of Figures~3 and 4. On the smallest and most nonlinear scales, the correlation functions probe the innermost regions of the halos. Intriguingly, the halo model predicts on these scales a behavior that is different from either the frequently-assumed stable clustering result of $\\Delta(k) \\propto k^\\gamma $ with $ \\gamma = (9+3n)/(5+n)$ (Davis \\& Peebles 1977), or the power-law profile result of $\\gamma = 2\\epsilon - 3$. The implication of departure from stable clustering is significant because all the fitting formulas for the nonlinear $P(k)$ in the literature (see \\S 5.2) have been constructed to approach the stable clustering limit at high $k$. A more detailed study on the criteria for stable clustering in this model is given in a separate paper (Ma \\& Fry 2000c). The origin of the deviation from stable clustering in the model at high-$k$ can be understood as follows. For the two-point function, as $k$ becomes large, the one-halo integral $P_{1h}(k)$ in \\eq{Pk} converges before the exponential cutoff, and is dominated by contributions near the mass scale for which $kR_s=1 $. The behavior now depends on the mass distribution function. The various mass functions discussed in \\S 2.2 have the same general behavior of $dn/dM \\propto M^{-2}\\,\\nu^\\alpha\\, e^{-\\nu^2/2}$, where $\\nu=\\delta_c/\\sigma$. The Press-Schechter form assumes $\\alpha=1$ (see eq.~[\\ref{PS}]), while others (e.g., Sheth \\& Tormen 1999; Jenkins et al. 2000) suggest a flatter slope of $\\alpha\\approx 0.4$ for the lower mass halos. Since the scale radius $R_s$ depends on mass as $R_s = R_{200}/c \\propto M^{1/3} / M^{-(3+n)/6} \\propto M^{(5+n)/6}$, and $R_s^3 \\deltabar \\propto M $ (up to logarithmic factors), we find from \\eq{Pk} that the power spectrum at high $k$ goes as \\begin{equation} \\Delta(k) \\approx \\Delta_{1h}(k) \\propto k^3\\,\\int dM\\,\\nu^\\alpha \\, \\ut^2(kR_s) \\,. \\end{equation} Changing variables to $y = kR_s \\propto k\\,(M/M_*)^{(5+n)/6}$, we see that \\begin{equation} \\Delta(k)\\propto k^\\gamma\\,,\\qquad \\gamma = \\left( {9+3n\\over 5+n} \\right) - \\alpha \\left( {3+n\\over 5+n} \\right) \\,, \\label{g2halo} \\end{equation} where the first term in $\\gamma$ is the prediction of stable clustering. The departure arises from the factor $\\nu^\\alpha$ in the mass function, and would vanish only if $\\alpha=0$ or $n=-3$. This is the origin of the difference in $\\Delta(k)$ at high $k$ between the model prediction (solid curves) and the fitting formula (dotted curves) shown in Figures 3 and 4. For the three-point function, the one-halo integral $B^{\\rm eq}_{1h}$ in \\eq{Bk} converges (barely, for $ p = {3 \\over 2} $ and $ n =-2 $), giving \\begin{eqnarray} B^{\\rm eq}(k) &\\propto & k^{\\gamma_3 -6}\\,,\\qquad \\gamma_3 = 2 \\left( {9+3n \\over 5+n} \\right) - \\alpha \\left( {3+n\\over 5+n} \\right) \\nonumber\\\\ Q^{\\rm eq}(k) &\\propto & k^{\\alpha (3+n)/(5+n)} \\label{g3halo} \\end{eqnarray} This again disagrees with the prediction of stable clustering that $Q$ is constant, but it appears to be consistent with numerical simulations as shown in Figures~3 and 4. For yet higher order correlations, details of the halo profile begin to matter. For $p=1$, the pattern of equations (\\ref{g2halo}) and (\\ref{g3halo}) persists to all orders, but for $ p = {3 \\over 2} $ they apply only for the two- and three-point functions; for four-point and higher functions the nonlinear scale $M_* $ and $ \\gamma_n = np - 3 $ for $ n \\ge 4 $. Thus there seems to be some potentially interesting behavior that is tested only in the four-point function and higher." }, "0003/astro-ph0003175_arXiv.txt": { "abstract": "Shells are faint arc-like stellar structures, which have been observed around early type galaxies and are thought to be the result of an interaction. HI gas has recently been detected in shells, a surprising result in view of the theoretical predictions that most of the gas should decouple from stars and fall into the nucleus in such interactions. Here we report the first detection of molecular gas (CO) in shells, found 15\\,kpc away from the center of NGC\\,5128 (CenA), a giant elliptical galaxy that harbors an active nucleus (AGN). The ratio between CO and HI emission in the shells is the same as that found in the central regions, which is unexpected given the metallicity gradient usually observed in galaxies. We propose that the dynamics of the gas can be understood within the standard picture of shell formation if one takes into account that the interstellar medium is clumpy and hence not highly dissipative. The observed metal enrichment could be due to star formation induced by the AGN jet in the shells. Furthermore our observations provide evidence that molecular gas in mergers may be spread out far from the nuclear regions. ", "introduction": "Early type galaxies (E/S0) are often found to be surrounded by faint arc-like stellar structures, called shells or ripples (\\cite{mc}), due to the accretion and subsequent merging of a smaller companion galaxy. It is widely accepted that the high frequency of galaxies with shells ($\\sim$\\,50\\%) attests to the importance of merging in galaxy formation (Schweizer \\& Seitzer 1988, 1992). Simulations of the {\\em stellar component} have shown that the shells or ripples are created either by ``phase-wrapping'' of the tidal debris of the accreted companion on nearly radial orbits (\\cite{quinn}), or by ``spatial-wrapping'' of matter in thin disks (\\cite{dupraz}, \\cite{hq}). CenA is a giant elliptical galaxy with strong radio lobes on either side of a prominent dust lane situated along its minor axis (\\cite{clarke}). Additionally, optical and HI observations (\\cite{dufour}, \\cite{vangorkom}) show a warped gaseous disk which has been accreted along the minor axis of this apparently prolate elliptical galaxy. CO mapping suggests that the disk contains 2$\\times 10^8$ M$_{\\odot}$ of molecular gas (\\cite{eckart}). Recently, mid-IR observations revealed the presence of a bisymmetric bar-like distribution of hot dust in the inner disk (\\cite{mirabel}). High contrast optical images of the galaxy show stars distributed in a large number of faint narrow shells around the galaxy (\\cite{malin}). The presence of the warped gas disk and shells suggests that CenA has accreted one (or more) smaller disk galaxy(ies) approximately $\\sim 10^8$ yrs ago (\\cite{quillen}). Schiminovich et al. 1994 detected 4$\\times 10^8$ M$_{\\odot}$ of HI gas associated with the stellar shells, having the same arc-like curvature but displaced 1 arcmin to the outside of the stellar shells. This result is intriguing since in general it is thought that the dynamics of the gas and stellar components are decoupled during a merging event. Detailed numerical modeling of the infall of a small companion on a massive elliptical has demonstrated that aligned and interleaved shells are formed through phase wrapping of the companion's stars on almost radial orbits (\\cite{quinn}, \\cite{dupraz}). When gas is taken into account, due to its dissipation it rapidly concentrates in the nucleus and does not form any shell (\\cite{weil}). Another possibility is that shells result from space wrapping when the relative angular momentum of the two galaxies is high. In this case the stellar shells do not have the same regular structure (\\cite{prieur}) and as they rotate around the central potential they dissolve more rapidly. The gas could remain associated with the stellar shells during a few dynamical times before condensing to the center. The morphology of the shells in CenA, though, suggests a combination of both phase and space wrapping since there are both a number of shells aligned with the major axis of the prolate giant elliptical and there are a few which are irregular. Furthermore, the HI is mostly associated with what Malin et al. (1983) called the diffuse shells. To explain the presence of gas in phase-wrapped shells, one should consider the interstellar medium as multiphase: a large fraction of the ISM could be composed of dense clumpy material with low dissipation. During a galaxy merger the dense gas behaves almost as collisionless particles and can orbit through the center as the stars do. To trace this dense component we attempted to detect molecular gas from the shells in CenA. The results were positive beyond our expectation, as described now. \\begin{figure*} \\resizebox{\\hsize}{!}{\\includegraphics{ca141.f1}} \\caption{a) A Digitized Sky Survey optical image of CenA with the contours of HI gas (from \\cite{david1}) superimposed in white. The HI contour levels are 1, 4, 7, 10, 15, 30, 35, 40\\,$\\times$\\,10$^{20}$ cm $^{-2}$. North is up and east is to the left, while the image scale is shown by the horizontal bar. The positions observed in CO are marked with the red circles whose size corresponds to the SEST 44$''$ beam of CO(1-0). The type of each map (half-beam spacing or simple pointing) is evident by the placement of the circles. The locations of the outer stellar shells are underlined by the yellow solid lines (see also Fig. 1a of Schiminovich et al 1994) . The inner 6cm radio continuum lobes (from \\cite{clarke}) are depicted by the blue contours (contour levels 0.01, 0.05, 0.1 Jy/beam). Note the jet alignment with the location of the CO detections. The outer radio lobes are far more extended. b) CO(1-0) and CO(2-1) spectra towards the northern shell S1 with the temperature scale in main beam T$_{\\rm mb}$, smoothed to 18\\,km\\,s$^{-1}$. c) Same as in b) but for the southern shell S2. } \\label{fig} \\end{figure*} ", "conclusions": "We have detected CO molecules in two shells aligned along the major axis of Centaurus A. The molecular gas is globally associated with the HI and stellar shells but with a radial shift, the HI being the more external component, and the stars the more internal. The presence of molecular clouds in these distant shells is compatible with the dynamical scenario of phase-wrapping, following the merger of a spiral galaxy with Centaurus A. Part of the interstellar medium of the spiral is clumpy, with very low collision rate and dissipation, and can follow nearly radial orbits during the merger, like the stellar component, without accumulating towards the center. The differential dynamical friction experienced by the gas and stellar components, that are unbound from the spiral companion at different epochs, can explain the radial shifts between the different shells. The detection of CO emission far from the center implies the presence of H$_2$ molecules as far as 1.16\\,R$_{25}$. The present detections, taking into account that only a small fraction of the shells was mapped in CO, suggest that more than 50\\% of the gas in the outer regions of CenA is in molecular form, and at least 10\\% of the total molecular gas detected in CenA is not in the nucleus. Moreover the derived HI/H$_2$ mass ratio is nearly constant with radius. This requires a metallicity enrichment in the most external gas, that could be due to the interaction between the gaseous shells and the radio jet. This prototypical example of a gaseous accretion suggests that the molecular gas is not always confined in the nuclear regions in merger remnants." }, "0003/astro-ph0003339_arXiv.txt": { "abstract": "s{ The Cosmic Microwave Background (CMB) anisotropies measurements can provide many clues about the Universe. Although the common belief is that they will allow a very precise measurement of the cosmological parameters (that is, the current state of the Universe), they will alternatively give interesting informations about the state of the initial perturbations (that is, the state of the Universe at the end of inflation). In this paper, we study the observational consequences on the CMB anisotropies of some wide set of inital conditions, with a correlated mixture of adiabatic and isocurvature perturbations.} ", "introduction": "The CMB anisotropies can indirectly measure the cosmological parameters by looking at the evolution of the cosmological perturbations between the end of inflation and the recombination epoch. In order to do so, one must implicitly assume a simple form for the initial perturbations. Usually, one considers only adiabatic fluctuations, with a power law spectrum. Adiabatic fluctuations arise in the context of the simplest inflationary scenario. However, as soon as one has a multiple inflation scenario, isocurvature fluctuations can be generated, the amplitude of which, as well as their correlation with the adiabatic part, depend on the parameters of the model. In many of the models already studied~\\cite{iso}, the isocurvature and the adiabatic parts of the fluctuations are uncorrelated, but it is possible to have a correlated mixture of such perturbations, as was already stressed by one of us~\\cite{l99} in the study of a specific inflation model with two massive non interacting scalar fields. In this communication, we study this issue in a phenomenological way, and look at the observable consequences of a correlated mixture of adiabatic and isocurvature cosmological perturbations. These perturbations happen to have a richer structure than the more usual uncorrelated adiabatic and isocurvature ones. ", "conclusions": "" }, "0003/astro-ph0003425_arXiv.txt": { "abstract": "The presence of a substantial number of hot stars in the {\\it extremely metal-rich} open cluster NGC\\,6791 has been a mystery. If these hot stars are in their core helium burning phase, they are significantly bluer (hotter) than predicted by canonical stellar evolution theory. No obvious explanation is available yet. We consider the effects of mass loss {\\em during the evolution of horizontal branch} (HB) stars as their possible origin. We find that the addition of mass loss causes HB stars to evolve to be {\\it hotter} and {\\it fainter}. Mass loss has a more pronounced effect for less massive stars and thus naturally widens the temperature (and color) distribution on the HB. If mass loss rates are higher for more metal-rich stars, this phenomenon would be even more pronounced in the metal-rich populations, such as NGC\\,6791. We find that mass loss on the HB may be a viable method of forming subdwarf B (sdB) stars both in the field and in clusters, especially when the metallicity is high. ", "introduction": "Understanding the origin and evolution of hot horizontal branch (HB) stars would have a significant impact on several areas of astronomy. For example, it is important in the study of globular cluster HB morphology, which constrains the formation history of the Galaxy (Searle \\& Zinn 1978). The global picture of HB morphology appears to be well understood. HB morphology is most affected by the metallicity of the cluster: more metal rich clusters exhibit redder HBs (``the first parameter'' of HB morphology). However, observations in some globular clusters show HBs which exhibit color distributions not expected for their metallicities. To explain this deviation, a ``second parameter'' is invoked. Age has been the most successful global candidate for the \"second parameter\" (Searle \\& Zinn 1978; Lee, Demarque, \\& Zinn 1994; for recent review, see Stetson, VandenBerg, \\& Bolte 1996; Sarajedini, Chaboyer, \\& Demarque 1997). However, recent observations show an increasing number of HB morphologies with anomalous features, e.g. bimodality in the color (temperature) distribution, extended blue tails, and gaps (e.g., Sosin et al. 1997; Rich et al. 1997; Catelan et al. 1998). Most of these anomalies are found near the blue end of the HB. Within the framework of the canonical theory of single star evolution, they appear to be inconsistent with a scenario where metallicity and age are the only two parameters that govern the HB morphology. NGC\\,6791 is an extreme example that exhibits deviations from the first and second parameter phenomena. This open cluster is believed to be extremely metal-rich, [Fe/H] $= 0.2$ -- 0.5 (Friel \\& Janes 1993; Kaluzny \\& Rucinski 1995; Peterson \\& Green 1998), which would suggest a predominantly red HB clump, following the first parameter phenomenon. Yet, nearly a third of its helium burning population (HB stars) are extremely blue (Liebert et al. 1994; Green et al. 1997) with a large temperature separation from the red clump. Dynamical effects do not seem to provide a promising explanation for this anomaly, because stellar densities are small in open clusters. Figure 1 shows the observed color-magnitude diagram (CMD) and a synthetic HB model based on canonical stellar models. The observed CMD of NGC\\,6791, which is from Kaluzny \\& Rucinski (1995), shows approximately 23 red clump stars and 9 hot stars. These hot stars have also been identified by ultraviolet (UV) observations using the {\\it Ultraviolet Imaging Telescope} (Landsman et al. 1998). The synthetic HB (diamonds) is based on the arbitrarily chosen chemical composition, i.e. [Fe/H]$ = 0.33$ and $Y=0.31$ (equivalent to ${\\Delta Y / \\Delta Z = 1.5}$). For illustrative purposes, a 9 Gyr isochrone from the main sequence (MS) to the red giant branch (RGB), is shown on the same plot. For $E(B-V) =0.125$ and $m-M=13.35$, the match is reasonable. It is beyond the scope of this paper to derive the age of this cluster or any other parameters by matching the CMD with isochrones. For the synthetic HB construction, we have assumed a Gaussian mass dispersion parameter $\\sigma_{HB} = 0.04 M_\\odot$, and a Reimers mass loss parameter $\\eta = 1.0$ (Lamers \\& Cassinelli 1999; Willson, Bowen, \\& Struck 1996). Readers are referred to Yi, Demarque, \\& Oemler (1997b) for details of the synthetic HB construction. We assume that there are approximately 32 HB stars in NGC\\,6791, and we have added some random scatter in both $M_{V}$ and $B-V$. The model reproduces, in accordance with the first parameter phenomenon, the prominent red clump seen in the data, but it lacks hot stars. NGC\\,6791 is the oldest Galactic open cluster known to-date at 6.5-9 Gyr (Demarque, Green, \\& Guenther 1992; Phelps et al. 1994; Garvanich et al. 1994; Chaboyer, Green, \\& Liebert 1999). The most recent estimate from Chaboyer et al. (1999) suggests $8.0 \\pm 0.5$ Gyr for the age of this cluster. However, even at the largest age estimate so far (9 Gyr), a synthetic HB based on the conventional input parameters would not reproduce anything even close to the HB morphology of NGC\\,6791. By increasing its age by a couple of Gyrs, we could produce some hot HB stars, but their numbers would still be much smaller than observed. Thus, the age effect alone seems unsuccessful in explaining the HB color distribution. A relatively large value of the dispersion ($\\sigma = 0.06 M_{\\odot}$, in comparison to the conventional value 0.03 -- 0.04) in the Gaussian mass distribution on the HB would result in a larger fraction of hot HB stars. This may match the blue-to-red HB star ratio found in NGC\\,6791 better, but the strong color-bimodality in NGC\\,6791 would not be reproduced. One or more local parameters must be at work in this cluster. As candidate mechanisms that produce hot HB stars, a few additional scenarios have been proposed lately. Sweigart (1997) has suggested that helium mixing on the red giant branch would cause further evolution beyond the tip, resulting in higher mass loss on the giant branch. As the color dispersion on the HB is the result of varying envelope masses, the enhanced mass loss results in a bluer HB. D'Cruz et al. (1997) assumed an episode of extreme mass loss at the tip of the giant branch to explain blue HB stars. \\placefigure{fig1} \\parbox{3.2in}{\\epsfxsize=3.2in \\epsfbox{yongetal_f1.eps}} \\centerline{\\parbox{3.2in}{\\small {\\sc Fig. 1} Observed CMD of NGC\\,6791 and a synthetic HB model. Note the presence of hot HB stars (enclosed in a box) only in the observed CMD. }} \\vspace{0.2in} \\addtocounter{figure}{1} In this letter, we explore yet another possible formation scenario for blue HB stars: mass loss during the core helium burning phase. Demarque \\& Eder (1985) first investigated this effect for low metallicity HB stars, and we extend on their work here. We will show that the inclusion of mass loss on the HB works in the direction of explaining the color distribution of the HB stars in NGC\\,6791. ", "conclusions": "In order to study the possibility of forming blue HB stars through mass loss during the HB phase of stellar evolution, we have constructed evolutionary tracks for HB stars of varying metallicity and envelope mass. Figure 2 shows HB evolutionary tracks and 6 hot stars in NGC\\,6791 whose temperatures and luminosities have been measured by Landsman et al. (1998) based on their UV fluxes. The track furthest to the right is the evolutionary track without mass loss and each successive track to the left has a larger mass loss rate; ${10^{-10}}$, ${10^{-9.5}}$, ${10^{-9} M_{\\odot} yr^{-1}}$, respectively. The effect of mass loss shown in Figure 2 is representative of the effects of mass loss in the other models calculated. Several points can be made by examining Figure 2. First, the greater the mass loss rate assumed, the hotter the star becomes. This is not a surprising result. It is generally believed that the HB is a sequence of stars which have the same core mass and varying envelope masses. The smaller the envelope mass, the hotter the star appears, as a deeper and hotter region is exposed with a smaller envelope. It seems natural then that mass loss during the HB phase will produce hotter stars with smaller envelopes. \\placefigure{fig2} \\parbox{3.2in}{\\epsfxsize=3.2in \\epsfbox{yongetal_f2.eps}} \\centerline{\\parbox{3.2in}{\\small {\\sc Fig. 2} Evolutionary tracks for stars starting on the zero-age HB for Z=0.04, ${\\Delta Y /\\Delta Z = 3}$ and envelope mass of ${0.07M_\\odot}$. The tracks of ${\\Delta Y /\\Delta Z = 1.5}$ show qualitatively the same phenomenon. The zero-age HB locus, which is insensitive to the adopted mass loss rate, is shown near the bottom of the plot. Tracks with increasing mass loss rates are shifted toward the hotter and fainter regions of the H-R diagram. The crosses indicate time intervals of ${10^7}$ years. Also shown are the hot stars in NGC\\,6791 whose temperatures and luminosities have been measured from their UV fluxes (Landsman et al. 1998). }} \\vspace{0.2in} \\addtocounter{figure}{2} At the same time, a larger mass loss rate also results in the stars becoming fainter than the stars without mass loss, particularly for the larger mass loss rates. As our mass loss scheme removes mass from the envelope, which is defined to be the hydrogen burning shell and outwards, the total amount of hydrogen burning is reduced. This results in the stars evolving with smaller luminosities. Figure 2 shows that the atmospheric properties of the hot stars in NGC\\,6791 (diamonds) are well matched by the theoretical models with mass loss if their envelope masses are approximately $0.07 M_\\odot$. If there is no mass loss on the HB, their properties can be matched only if their envelope masses are as small as $0.001 M_\\odot$ (the filled star symbol at log\\,$T_{\\rm eff} \\approx 4.4$ and log\\,L/L$_{\\odot} \\approx 1.1$). Whether HB stars can be born with such low envelope masses at all has been questioned by many groups: ``the minimum envelope mass hypothesis'' (see the discussion in \\S3.2 of Yi et al. 1997b). With mass loss on the HB, one can match the observed data without violating the minimum envelope mass hypothesis. \\placefigure{fig3} \\parbox{3.2in}{\\epsfxsize=3.2in \\epsfbox{yongetal_f3.eps}} \\centerline{\\parbox{3.2in}{\\small {\\sc Fig. 3} Comparison of evolutionary tracks for a constant metallicity Z=0.04 and varying envelope masses. Mass loss has a more profound effect on low envelope mass stars. }} \\vspace{0.2in} \\addtocounter{figure}{3} It should be noted that the adoption of mass loss on the HB does not change the zero-age HB locus at all in our scenario, because we are only comparing the tracks with and without mass loss when they begin their HB evolution at the same location in the CMD. The inclusion of mass loss on the HB also contributes to the wide color distribution in the HB of NGC\\,6791. Figure 3 shows evolutionary tracks for HB stars with varying envelope masses. Each panel is for a given envelope mass with four mass loss rates plotted for each. The composition for this particular set of models is $Z=0.04$ and ${\\Delta Y / \\Delta Z = 3}$. Mass loss has little effect to the evolution of the HB stars with large envelope masses but significantly affects that of less massive stars. This mass dependency would generate a wider color distribution in the HB. It may even create a color bimodality (or multimodality), as seen in the hot ends of the HBs of several globular clusters, if the mass dependency is not a smooth function of mass. This may explain the lack of intermediate HB stars in NGC\\,6791. A more detailed modeling will show if this is a plausible explanation to the case of NGC\\,6791. The effects of such mass loss in the HB phase have an impact on the study of ultraviolet spectral evolution of old stellar populations as well. Numerical studies by Demarque and Pinsonneault (1988) and by Horch, Demarque, \\& Pinsonneault (1992) suggested that in the advanced evolution of metal rich HB stars, stars with small envelopes evolve to become UV bright objects, the so-called slow blue phase (SBP) stars, rather than evolving into asymptotic giant branch stars (c.f. the AGB-manqu\\'{e} stars in Greggio \\& Renzini 1990). Yi, Demarque \\& Kim (1997a) elaborated on the theory and found a theoretical explanation to the SBP phenomenon and to its metallicity dependence, that is, the transitional mass (the maximum mass of a star that becomes a SBP star) increasing with metallicity. According to this theory, more metal-rich stars become UV bright SBP stars more easily. With the additional mass loss on the HB that we consider in this paper, stars would become SBP stars even more easily. Moreover, if mass loss on the HB increases with increasing metallicity and increasing effective temperature, as various studies suggest, it would cause metal-rich stellar populations to develop UV bright stars more quickly. Because SBP stars are considered important UV sources in elliptical galaxies (e.g., Tantalo et al. 1996; Yi et al. 1997a; O'Connell 1999) and the UV flux of elliptical galaxies may constrain the ages of galaxies (Yi et al. 1999), this study has an impact on extragalactic astronomy. Landsman et al. (1998) already pointed out that a population like NGC\\,6791 would exhibit a UV upturn with the magnitude shown in giant elliptical galaxies. If mass loss on the HB indeed contributes to the production of hot stars in the metal-rich environments, our current age estimates of giant elliptical galaxies are likely to be systematically overestimated. We should point out that there is no direct observational evidence for mass loss as substantial as we have assumed to occur on the HB. In fact, besides Koopmann et al.'s upper limit on the mass loss rate on the HB (${10^{-9} M_{\\odot}yr^{-1}}$), the diffusion study of Michaud et al. (1985) set a more strict upper limit of ${10^{-14} M_{\\odot}yr^{-1}}$. They suggest that if mass loss rate is larger than this it would be difficult to reproduce the silicon underabundance observed in sdB stars. If this is true, the mass loss rates investigated in our study may be too large to justify, unless mass loss rate increases substantially in the metal-rich regime. Once again, we do not advocate any value or mechanism for the mass loss on the HB. Instead, we wanted to see what level of mass loss rate on the HB would be needed if it is the only process that is responsible for the production of hot HB stars. Based on a cursory inspection of our new HB tracks with mass loss, adoption of such low mass rates ($ < {10^{-14} M_{\\odot}yr^{-1}}$) would not make any appreciable difference in the HB morphology. Some of the blue HB stars have been observed to be spectroscopic binaries (Green et al. 1997). This brings up the possibility that the blue HB stars are formed through binary evolution. Some of the formation scenarios such as helium white dwarf mergers (Bailyn \\& Iben 1989, Iben 1990) can perhaps be ruled out based on the fact that the luminosities are in a small range (see Figure 2). This small range of the luminosities of the hot stars is also in conflict with the binary scenario (Landsman et al. 1998). Mengel, Norris, \\& Gross (1976) demonstrated that blue HB stars can form through binary evolution with mass transfer via Roche Lobe overflow. However, the conditions under which this occurs are probably too infrequently met, although further study, applicable to metal rich stellar populations, is needed. Furthermore, it is unclear whether all blue HB stars in NGC\\,6791 are in mass-exchanging binary systems. Therefore, a mechanism to form blue HB stars through single star evolution is still needed. Mass loss on the HB may be such a mechanism." }, "0003/astro-ph0003080_arXiv.txt": { "abstract": "The results from an X-ray spectral analysis of a large sample of quasars, observed with ASCA, are presented. The sample was selected to include all ASCA observations of quasars, with z\\ $>0.05$ and M$_{V}<-23.0$, available up to January 1998. The data reduction leaves 62 quasars, 35 radio-loud and 27 radio-quiet, suitable for spectral analysis. Differences are found between the radio-quiet quasars (RQQs) and the radio-loud quasars (RLQs); the RLQs have flatter X-ray spectra ($\\Gamma\\sim1.6$), with little iron line emission or reflection and are more X-ray luminous than the softer ($\\Gamma\\sim1.9$) RQQs, in agreement with previous studies. A correlation between $\\Gamma$ and optical H$\\beta$ was also found for the {\\it radio-quiet} quasars in this sample, whereby the steepest X-ray spectra tend to be found in those objects with narrow H$\\beta$ widths. The correlation is significant at $>$99\\% confidence, confirming the well-known trend between $\\Gamma$ and H$\\beta$ FWHM in Seyfert 1s (Brandt \\et 1997), but at higher luminosities. Other spectral complexities are observed from this sample. A soft X-ray excess, with blackbody temperatures in the range 100 - 300 eV, is seen in many low z radio-quiet quasars. In most cases the temperatures are probably too hot to originate directly from the disk and could imply that some reprocessing is involved. Iron K line emission features are also found in the RQQs; but often from partially ionised material. Indeed in the highest luminosity RQQs there is neither evidence for iron line emission nor the reflection component expected from disk reflection models. These observations can be explained by an increase in the quasar accretion rate with luminosity, leading to an increase in the ionisation state of the surface layers of the disk. The occurrence of ionised or `warm' absorbers is rare in this sample, with only 5 detections in low z objects. However excess neutral X-ray absorption is found towards several of the high z, predominantly radio-loud, quasars. Although found to increase with quasar redshift, this `intrinsic' absorption may be associated with radio-loud AGN. ", "introduction": "Discovered in 1963 (Schmidt 1963), quasars are the most luminous continuously emitting objects in the Universe and represent the high luminosity end of the class of objects known as Active Galactic Nuclei (AGN). Like their lower luminosity cousins - Seyfert 1 galaxies - the bulk of the energy produced in quasars is thought to arise from accretion onto a compact object (the putative super-massive black hole). This central engine is also thought to be where the X-rays, that are observed from both quasars and Seyfert 1s, originate from. In one model, the UV photons produced by viscous dissipation in an accretion disk are Comptonised to X-ray energies by a hot corona above the surface of this disk (Haardt \\& Maraschi 1993). These hard X-rays in turn illuminate the accretion disk, being either `reflected' back towards the observer or thermalised into optical or UV photons. Evidence for these `reflection' features (in the form of an iron K$\\alpha$ emission line, Fe K absorption edge and Compton scattering hump) is commonly observed in the X-ray spectral band in Seyfert 1 galaxies (e.g. Pounds \\et 1990, Nandra \\& Pounds 1994). The detection of these reflection features however, in quasars, remains more ellusive (Reeves \\et 1997, Lawson \\& Turner 1997). In the {\\it radio-loud} quasars the situation is somewhat further complicated by the presence of a powerful relativistic radio-jet. In the X-ray band, these radio-loud quasars have flatter X-ray spectral emission (e.g. Wilkes \\& Elvis 1987, Lawson \\et 1992), and are generally more luminous than the radio-quiet quasars. The radio-loud quasars also have little or no X-ray (iron) line emission; this is often interpreted in terms of the Doppler boosting of the X-ray continuum, by the relativistic jet (see Reeves \\et 1997 and references therein). This paper presents the results of a detailed spectral analysis of 68 quasars obtained from the \\asca\\ public archive. The aims are to extend the results that were presented in Reeves \\et (1997), which contained a smaller sample of 24 objects. The objects considered in this sample contain a roughly equal mix of both radio-loud and radio-quiet quasars. The bigger sample, for instance, allows us to perform an investigation of the properties of iron K lines and reflection in quasars, which will be limited in some of the objects by signal-to-noise. It also permits us to investigate the properties of quasars over a large range of luminosity and redshift. A further aim of the paper is to make the results of this analysis available to the general community; the paper presents results from a considerable number of quasars that are currently unpublished. In the following section, the selection and properties of the sample are discussed. Section 3 then outlines the spectral fitting that was performed on the quasars in the sample. The following sections (4-7) then present and discuss the results in terms of the X-ray continuum emission, soft X-ray excesses from quasars, the properties of the iron line and reflection associated with the putative accretion disk and also the effects of absorbing material on the quasar spectra. Values of $ H_0 = 50 $~km~s$^{-1}$~Mpc$^{-1}$ and $ q_0 = 0.5 $ are assumed and all fit parameters are given in the quasar rest frame. ", "conclusions": "We now summarize the findings of this paper. In particular, comparison is drawn to our earlier paper (Reeves \\et 1997 or R97), which contained a smaller sample of quasars (24 objects compared with the current sample size of 62). Firstly we confirm the following main results from the R97 paper:- \\begin{itemize} \\item{A decrease in the X-ray photon index, with radio-loudness, for all quasars.} \\item{A strong confirmation of the correlation between the neutral X-ray absorption column (\\nh) and quasar redshift (z), in the sense that intrinsic \\nh\\ increases with z. Furthermore in this sample, the correlation does not depend on any calibration effects nor the rest-frame of the absorbing column.} \\item{A decrease of iron K line equivalent width with increasing radio-loudness. The interpretation is that the strength of the reflection disk component (and therefore any contribution from the iron K line) is diminished, due to Doppler boosting of the X-ray continuum by the relativistic jet, in the core-dominated radio-loud quasars.} \\end{itemize} However we have seen several new effects and correlations in this paper, that were not reported in our previous R97 sample:- \\begin{itemize} \\item{In the R97 paper, a correlation was found whereby the photon index for the radio-quiet objects increased with X-ray luminosity. No such correlation was found in this paper, over a large range of both redshift and luminosity, for the radio-quiet sources. The difference may be due to the increase in sample size (from 9 radio-quiet quasars in R97 to 27 in the present paper).} \\item{Two correlations were found involving the iron K emission line. Firstly the strength of the iron K emission was observed to decrease with luminosity (i.e. an `X-ray Baldwin' effect), regardless of whether the objects are radio-loud or radio-quiet. In addition the energy (or ionisation) of the iron line was found to increase with luminosity. Both of these trends confirm the result found in the Nandra \\et (1997b) paper, whereby the composite line profiles for AGN tend to be weaker at higher luminosities as well as narrower and shifted bluewards of 6.4 keV. The `X-ray Baldwin effect' was first proposed for AGN on the basis of Ginga data by Iwasawa \\& Taniguichi (1993).} \\item{A new effect is found whereby the strength of the Compton reflection `hump' is weaker in the most luminous quasars at high redshifts (z$>$1). The effect is observed not only in the jet-dominated radio-loud sources, but also in the {\\it radio-quiet} quasars. This finding is consistent with the `X-ray Baldwin effect' discussed above, and suggests that as a whole the {\\it neutral} disk reflection component in the high luminosity quasars is generally weaker than in the lower luminosity sources such as the Seyfert 1s.} \\item{A trend has also been found for the radio-quiet quasars in this sample, whereby the X-ray (2-10 keV) photon index increases with decreasing optical H$\\beta$ width. Thus the quasars with the steepest X-ray spectra tend to have the narrowest H$\\beta$ FWHM. This trend has previously been found in the lower luminosity Seyfert 1s (e.g. Brandt \\et 1997), but has not been reported before for the more luminous quasars.} \\item{Soft X-ray excesses are also found a significant proportion (9) of the low z quasars in this sample. Interestingly the majority of the quasars with strong soft excesses are those with the narrowest optical H$\\beta$ widths (where H$\\beta$ FWHM $<2000$~km/s).} \\item{The temperatures of the soft X-ray excesses in this paper vary in the range between kT = 100 - 300 eV, for simple blackbody fits. In a majority of the cases the temperatures are probably too hot to result by direct thermal emission from the putative quasar accretion disk. Instead, one possibility is that the soft excess originates via thermal Comptonisation of UV photons from the disk in a hot corona. Another possibility is that the `soft excess' results from reprocessing. In particular emission and/or reflection from the surface of a highly ionised inner accretion disk could reproduce the observed excess in soft X-ray flux.} \\item{A systematic search has been carried out for the presence of warm or ionised absorbers in this quasar sample. Only a smaller number were found. We confirm the presence of a warm absorber in three previously reported cases (PG 1114+445, IRAS 13349+243 and MR 2251-178). The only new warm absorbers reported here are in the radio-quiet quasar PDS 456 and a marginal detection in the radio-loud quasar PG 1425+267. Overall the apparent rarity of warm absorbers in quasars may be due to different (higher) ionisation, a smaller covering fraction, or is perhaps just due to the redshift effect.} \\end{itemize} So how can we place all these observations facts into a general scheme for quasars. Firstly the differences between radio-loud and radio-quiet quasars seem relatively straightforward. In the radio-loud quasars, a strong Doppler boosted emission component from the relativistic jet can account for the higher luminosities, the generally flatter X-ray spectra as well as the diminished iron K line and reflection component in these objects. One question of real interest for future study is whether the central engine is the same in the radio-loud quasars as it is in the radio-quiet quasars. For instance the structure of the accretion disk may be different in the radio-loud quasars; sensitive studies of the reflection component and iron line in the RLQs (i.e. with XMM) may help to determine this. The properties of the {\\it radio-quiet} quasars on the whole seem more complex. As has been seen in this paper, there is little or no dependence on the X-ray continua of quasars on luminosity and therefore presumably the black hole mass. However perhaps the one driving factor responsible for the individual properties of quasars may be the {\\it fractional accretion rate} $\\dot{m}$ of the central engine (i.e. the ratio of mass accretion to the Eddington rate - or the Eddington ratio). A high fractional accretion rate can result in the surface layers of the disk becoming highly photoionised, which subsequently can have several effects on the X-ray spectra. Depending on the degree of ionisation, ionised rather than neutral iron K emission lines can dominate the disk reflection spectrum, as observed. Furthermore at even higher ionisations, the neutral reflection component (and iron line emission) can appear to be weaker, particularly if the disk is fully ionised to several Thomson depths (e.g. Nayakshin \\et 1999, Ross \\et 1999), also in agreement with the apparent properties of the more luminous radio-quiet quasars. A further effect is that, for high Eddington ratios, stronger soft X-ray emission can be produced. This may partly arise as the intrinsic thermal emission from the disk can become stronger (Ross, Fabian \\& Mineshinge 1992). Another possibility is that as the disk is more highly ionised, it can become more reflective at soft X-ray energies, producing a steepening of the X-ray spectrum at low energies (i.e. as a result of the ionised disk reflection component). Thus a high accretion rate (relative to Eddington) may explain the strong soft excesses in some of the objects considered earlier. It is also interesting to return to the question of the dichotomy between the broad and narrow line quasars that was considered earlier. It has been postulated in the literature that the narrow optical H$\\beta$ lines may be an indicator of a high accretion rate (Pounds \\et 1995, Laor \\et 1997). If this is correct this could indeed account for the differences between the 2 types of objects. A high accretion rate may explain the strong soft excesses observed both in the narrow-line quasars (6 out of 8 objects in this sample) and also the lower luminosity narrow-line Seyfert 1s (NLS1s) in other samples (Vaughan \\et 1999, Leighly 1999). As explained this can be caused by increased intrinsic disk emission or an increase in the reflectiveness of the disk in the soft X-ray band. Also if the intrinsic disk emission is stronger, this can also account for the steeper 2-10 X-ray slopes, via increased Compton cooling (Pounds \\et 1995). Additionally some of the narrow-line quasars also show evidence for ionised iron K emission, also indicative of a high ionisation disk and thus a high accretion rate, although the evidence is tentative so far (also see Vaughan \\et 1999). So the narrow-line quasars (as well as the NLS1s) may radiate at a relatively high fraction of the Eddington rate, whereas in general the broad-lined quasars may be sub-Eddington, perhaps similar to the normal Seyfert 1s, but with more massive central black holes. An important question to ask is whether there are any narrow-line quasars at higher redshifts (z$>1$). Finally the amount of soft X-ray absorption towards quasars was found to increase with redshift (also see Fiore \\et 1998 for a similar analysis of \\rosat\\ quasars). This correlation is apparently robust, even when calibration effects and uncertainties in the amount of local absorption are taken into account. The main question that has arisen from this, is whether this absorption is intrinsic to the quasars or whether it originates from line-of-sight matter. Given the low-number density of high column systems (such as damped Ly-$\\alpha$ systems) that could cause appreciable X-ray absorption (O'Flaherty \\& Jakobsen 1997), the most likely scenario is that the bulk of this absorbing material is local to the quasars or the host galaxy environment. As there seems to be a comparative lack of absorption in some radio-quiet quasars (see Fiore \\et 1998), this excess \\nh\\ may be associated with radio-loud quasars (also see Sambruna \\et 1999). However further data (with XMM and Chandra) is required to determine the exact location and cause of this absorption." }, "0003/astro-ph0003049_arXiv.txt": { "abstract": "Many black hole X-ray transients are in a low state for several decades until an outburst occurs. We interpret this outburst behaviour as a marginal occurrence of a dwarf nova type disk instability in the cool outer accretion disk. We compute the disk evolution including evaporation of matter from the cool thin disk. This evaporation process causes the transition to a hot coronal flow. The efficiency depends on the black hole mass. The new results are the dependence of the outburst recurrence time on the mass transfer rate from the companion star and the fraction of the matter which is accumulated in the disk for the outburst. We determine a lower limit of the mass transfer rate, below which no disk instability can be triggered. We find that for rates slightly lower than those in the known black hole X-ray transients the disk would be stationary. We argue that many such optically faint black hole X-ray binaries with stationary cool accretion disks exist. ", "introduction": "Transient X-ray binaries containing a black hole form two different classes of objects, the high-mass and the low-mass binaries. The high-mass systems have an O or B star companion and the observations indicate mass transfer at a high rate onto the black hole primary. The low-mass systems, known as soft X-ray transients (SXT) or X-ray novae have a K or M dwarf companion. The Roche-lobe filling low-mass star transfers matter via an accretion disk onto the compact star. These binaries exhibit outbursts, usually detected in X-rays. The transient sources are interesting objects to study the accretion disk. Already a decade ago Huang \\& Wheeler (1989) and Mineshige \\& Wheeler (1989) argued that the rare outbursts are caused by a disk instability as in dwarf nova systems where the primary star is a white dwarf. Recently detailed modelling of the decline of the outburst lightcurve was done; for a comprehensive study and an overview see Cannizzo (1998,1999). Observational data for X-ray novae and system parameters deduced from the observations are summarized in reviews by Tanaka \\& Shibazaki (1996) and Tanaka (1999). Chen et al. (1997) carried out a statistical study of all long-term X-ray and optical lightcurves. Asai et al. (1998) investigated nine black hole X-ray transients in the quiescent state. If we compare SXTs with dwarf novae (Kuulkers 1999) we especially notice the long outburst recurrence times. The primary stars are black holes or neutron stars, and the orbital periods are longer. Only a few dwarf novae with very short orbital periods have such long recurrence times, connected with a high outburst amplitude. These systems are known as ``tremendous outburst amplitude dwarf novae'', TOADs (Howell et al. 1995). For the best observed system, WZ Sagittae, with outbursts every 31 years, the amount of matter accumulated in the disk for the outburst is about 1-2\\, $10^{24}$g (Smak 1993), a factor of only about 3 smaller than that estimated for the transient source A0620-00 (McClintock et al. 1983). But there is one important difference between WZ Sge stars and SXTs: the size of the disk in WZ Sge with an orbital period of only 81.6 minutes is much smaller than those in SXTs with typical periods of several days. The viscosity parameter in the disk in WZ Sge therefore has to be a factor of 10 to 100 lower if the same amount of matter is accommodated in the much smaller area (Meyer-Hofmeister et al. 1998). We want to point out another feature, different in dwarf novae and SXTs, but essential for the state of the disk. In transient sources and dwarf novae in quiescence the outer accretion disk is cool and matter is accumulated for the outburst. In dwarf novae such a disk is cool everywhere from the outer to the inner edge. In black hole transient sources, the disk can reach inward to the vicinity of the black hole. At such close distance a disk is hot even for very low mass flow rates. Since hot and cool disk regions cannot remain stationary side by side (Meyer 1981), transition fronts will sweep back and forth over the disk leading to a rapid sequence of hot and cool states preventing any long-term quiescent accumulation of mass in the disk. To circumvent this problem one either has to assume an extremely low viscosity so that matter cannot flow towards the inner disk (in contradiction to the amount of matter accumulated for a SXT outburst) or a hole in the inner thin disk due to evaporation. Our computation of disk evolution includes evaporation into a coronal flow and the co-existence of thin disk and corona consistently. Since in quiescence about half of the matter flows through the corona evaporation is an essential feature in the evolution of SXTs. The spectra of quiescent SXTs are not consistent with an accretion disk model of a thin disk reaching inward to the black hole, as pointed out by McClintock et al. (1995) and Narayan et al. (1996), but the problem can be resolved by accretion via an ADAF which is the inward continuation of the coronal evaporation flow (for a review see Narayan et al. 1999). The observed outburst recurrence time of transient sources ranges from around one year to several decades, for many systems only one outburst is recorded. We show in our investigation that the outbursts may be triggered only marginally and the recurrence time then can vary very significantly for a small difference in mass transfer rates. For slightly lower mass overflow rates from the companion star the systems remain in a stationary state with a cool disk. The question whether such faint non-transient black hole binaries exist was also approached, in a different way, in connection with the physics of an advection-dominated accretion flow (ADAF) by Menou et al. (1999b). The answer depends on the expected mass overflow rates in these binaries. But the predictions of the rates caused by magnetic braking are so uncertain that one can better draw conclusions from the outburst behavior of SXTs on the efficiency of magnetic braking than vice versa. In Table 1 we summarize properties of transient sources. Listed are binaries for which the observations document that the compact star is a black hole. In addition to these systems there exist a number of transients which are probably also black hole binaries, but also for those only in a few cases is more than one outburst known. \\begin{table*} \\pagestyle{empty} \\setlength{\\topmargin}{-2.5cm} \\setlength{\\textwidth}{26cm} \\setlength{\\textheight}{19cm} \\setlength{\\oddsidemargin}{-1.5cm} \\setlength{\\footskip}{-1.0cm} \\caption{Black--hole transient sources} \\begin{tabular}{llllllllr} \\hline \\hline Source & &BH mass & Companion & orbital &outburst&rec. time & acc. matter & Ref.\\\\ name & & ($M_\\odot$) & star & period (h)& year & $\\Delta t$\\, (ys) &log\\, $M_d$ & \\\\ \\hline \\\\ J0422+32 & XNova Per & $>$3.2 & M 2 V & 5.09 & '92 & $>$30 &23.9 & 1\\\\ 0620-003 &XNova Mon &$>$7.3 &K 5 V &7.75 &'17,'75 & 58&24.6 &2\\\\ 1124-684&XNova Mus &$\\sim 6$ &K 0-4 V &10.4 &'91 &$>$30 &25.3 &3\\\\ 1543-475&XN '71,'83,92 &2.7-7.5 &A 2 V & 26.95&'71,'83,'92 &$\\approx 10$ &25.1('92)\\ &4\\\\ J1655-40&XNova Sco &7.02$\\pm$0.22 &F 3-6 &62.7 &'94 &$>$30 &23.7 &5\\\\ 1705-250&XNova Oph'77 &$\\sim$ 6 &K$\\sim$ 3 V &12.51 &'77 &$>$30 &24.5 &6\\\\ 2000-251& XNova Vul& 6-7.5&early K &8.26 &'92 &$>$30 &25.1 &7\\\\ 2023-338&XNova Cyg &8-15.5 &K 0 IV &155.4 &'38,'56,'79,'89&10-20 &25.8('89) &8\\\\ & & & & & & & &\\\\ \\hline \\end{tabular} \\vspace {0.5cm} \\\\ Note: Systems established as black hole transients, data for black hole mass, spectral type of companion star, orbital period, outburst year and list of references from Tanaka \\& Shibazaki (1996), recently established also Nova Ophiuchi 1977 and 4U 1543-47. Amount of matter accumulated $log M_d$ derived from data collected by Chen et al. (1997). \\\\ References: (1) Filippenko et al. 1995a, (2) McClintock \\& Remillard 1986, (3) McClintock et al. 1992, (4) Orosz et al. 1998, (5) Orosz \\& Bailyn 1997, (6) Remillard et al. 1996, (7) Filippenko et al. 1995b, (8) Casares et al. 1992. \\end{table*} In our investigation we discuss the following points. We describe the computational code for evolution of the disk in quiescence including evaporation of the inner disk in Sect. 2. In Sect. 3 we show the results: the outburst recurrence time depends strongly on the black hole mass, the amount of matter accumulated in the disk during quiescence, and the fraction of mass accumulated to mass transferred from the companion star. We determine the lower limit for the overflow rate to trigger a disk instability. In Sect. 4 we discuss the regime of faint non-transient black hole low mass X-ray binaries. Conclusions follow in Sect. 5. ", "conclusions": "Our investigation gives new insight into the evolution of the disks in black hole X-ray transients. At the same time new questions also arise. \\subsection {The occurrence of outbursts} We follow the disk evolution including evaporation into a coronal flow. Conclusions on stability are only possible if one considers these truncated disks where at a certain radius $r_{\\rm {tr}}$ the thin accretion disk ends and the accretion changes to the form of a hot coronal flow. The outbursts are caused by the thermal-viscous instability as in dwarf novae, modelled with a viscosity value suitable for dwarf nova outbursts, which confirms the similarity. We found that the dependence of the evaporation process on the black hole mass essentially determines the outburst cycles. If the black hole mass is higher, a higher mass transfer rate is needed to trigger an outburst after a certain time interval of accumulation of matter. For example to get a recurrence time of 30 years for 4 or 8 $M_\\odot$ black holes about 2.5 or 6.5 $10^{-10}M_\\odot/ \\rm {yr}$ respectively are needed (compare Fig. 2). The outburst after long quiescence can be understood as marginal triggering of the disk instability. In such a case a small difference in the mass transfer rate results in an large change of the recurrence time. The location of the inner edge of the thin disk is important for the outburst cycles. In our modelling $r_{\\rm {tr}}$ follows from the evaporation model. The chosen viscosity parameter $\\alpha_{\\rm cool}$ also influences the result, but this not a free parameter because the total amount of accumulated matter has a constraint from the outburst energy. The systems listed in Table 1 are the best observed sources with the black hole mass established from observations. Assuming that no outburst was missed during the 30 years of X-ray observations (for a discussion see Chen et al. 1997) the recurrence times might be very long. In our view the instability is marginally triggered in these sources. For only a little lower rates these systems would be stationary, all matter transferred from the companion star steadily flows towards the black hole (wind loss excepted). Being so close to the marginal state in several black hole binaries, one expects a large number of similar systems in permanent quiescence. Such sources are very faint, with a spectrum like SXTs in quiescence. Menou et al. (1999b) discussed the observational signatures of such faint persistent black hole low-mass X-ray binaries. \\subsection {Mass transfer rates} Our computations of disk evolution to model the observations confirm that mass transfer rates of $10^{-10}$ to $10^{-9}M_\\odot/ \\rm {yr}$ cause the transient behaviour ( with a strong dependence on the black hole mass). These rates agree with the estimates for the amount of accumulated matter and with the rates derived from the spectral fits based on the ADAF model (Narayan et al. 1996, 1997, for a review see Narayan et al. 1999). For the accretion in disks around black holes we get a consistent description with the thin outer disk and the change to a coronal flow due to evaporation. \\subsection {The matter in the disk} The fact that the average rate of matter accumulation in the disk, derived from the observations (total outburst energy) and the mass flow rate towards the black hole (derived from ADAF spectral fitting) are just about the same seems surprising, as pointed out by Tanaka (1999). Menou et al.(1999b) estimated the relative importance of both rates and came to the conclusion that those are about equal (see also Menou et al. 1999a). Our computations naturally provide a value $\\langle\\dot M_{\\rm{acc}}\\rangle$/$\\dot M_T$ $\\approx$ 0.35-0.55 for outburst recurrence times of 30 to 50 years, characteristic for SXTs. \\subsection {The cause of the mass transfer} Assuming an evolved companion star the mass transfer rates caused by magnetic braking might be low enough to lead to outbursts. These rates of mass overflow from the secondary star according to the suggestions by Verbunt \\& Zwaan (1981) and Mestel \\& Spruit (1987) depend only weakly on the primary mass. The observed outbursts strongly point to the fact that the transfer rates are marginal to trigger an outburst. Then one would conclude that the rates have a spread such that no outbursts or only rare outbursts occur. But such an interpretation is not possible if the limiting rate is so different for different black hole mass, assuming that the observed systems do not all have the same black hole mass (for a discussion see Bailyn et al. 1998). Only if the transfer rates depend somehow on the primary mass could these rates be such that marginally triggered rare outbursts occur for different black hole masses. We do not know about any mechanism which could cause these transfer rates." }, "0003/nucl-th0003057_arXiv.txt": { "abstract": "The reaction rates for $\\alpha$ capture processes on self-conjugate nuclei in the mass range A=20-40 have been investigated. The rates were calculated using the statistical model code NON-SMOKER taking into account isospin suppression rules. These theoretical predictions are compared with rates derived from the available experimental data about the $\\alpha$ capture reactions but taking also into account additional experimental information from different reaction channels populating the $\\alpha$ unbound states of the self-conjugate compound nuclei. \\vspace{0.5cm} \\noindent {\\it PACS:} 21.10.Ma; 21.60.Gx; 24.30.-v; 24.60.Dr; 25.40.Ny; 26.50.+x; 27.30.+t\\\\ \\small {\\it Keywords:} nuclear reactions, nucleosynthesis, abundances --- supernovae: general\\\\ ", "introduction": "The astrophysical importance of $\\alpha$ capture on target nuclei with $N=Z$ is manifold. In the Ne- and O-burning phase of massive stars, alpha capture reaction sequences are initiated at $^{24}$Mg and $^{28}$Si, respectively, and determine the abundance distribution prior to the Si-burning phase~\\cite{CLS98}. Nucleosynthesis in explosive Ne and explosive O burning in type II supernovae depend on reaction rates for $\\alpha$ capture on $^{20}$Ne to $^{36}$Ar~\\cite{Arn95,Thie97}. While many of these processes are in quasistatistical equilibrium~\\cite{Thie97,ThH96,ThH99}, the reaction rates itself are important for a reliable description of nucleosynthesis in the subsequent cooling phase. An $\\alpha$ capture chain on such self-conjugate nuclei actually determines the production of $^{44}$Ti \\cite{hoff98}, which contributes to the supernova light curve by the energy release from its $\\beta$ decay to $^{44}$Ca via $^{44}$Sc. Charged-particle reaction networks have to consider $\\alpha$-capture rates in the conditions of an $\\alpha$-rich freeze-out \\cite{frei98}, or for an extended rp-process for proper treatment of the $\\gamma$ induced $\\alpha$ break-up of selfconjugate nuclei~\\cite{WGI94,sch98}. Such calculations involve also highly unstable nuclei, thus calling for a reliable prediction of the respective reaction rates. Due to isospin selection rules, $E1$ $\\gamma$ transitions with isospin $T=0 \\rightarrow T=0$ are forbidden. Likewise, $M1$ transitions will be strongly suppressed. Because of isospin conservation, only states with isospin $T=0$ can be populated by $\\alpha$ capture on $N=Z$ ($T=0$) targets. This leads to a strong suppression of the $\\gamma$ transitions in the compound nucleus and thus of the ($\\alpha$,$\\gamma$) cross section of self-conjugate nuclei~\\cite{WaW69}. Previous theoretical work used in astrophysical calculations either neglected this isospin effect \\cite{arn72,thi87} or accounted for it only in a phenomenological way with arbitrary suppression factors \\cite{hol76,woo78,cow91,sch98}. In this work we want to improve on the prediction of reaction rates on isospin-symmetric targets and compare the theoretical values to newly compiled available experimental information. In section \\ref{HF} the statistical model of nuclear reactions is introduced and the method for obtaining isospin suppression factors as well as the resulting theoretical cross sections and reaction rates are presented. Section \\ref{exp} reviews the available experimental information and presents reaction rates newly derived from all available experimental data covering different reaction channels. A summary and conclusion will be given in Section \\ref{conc}. ", "conclusions": "\\label{conc} In this paper we attempted to improve the statistical model description of $\\alpha$ capture rates on self-conjugate nuclei, which are important for many nucleosynthesis processes in stellar and explosive He-burning. The NON-SMOKER predictions presented here are typically lower than the results of previous Hauser Feshbach calculations~\\cite{hol76,woo78}, except for $^{28}$Si($\\alpha$,$\\gamma$)$^{32}$S for which the present theoretical rate is higher by a factor of 1.96 (see also Table 6 in~\\cite{hoff98}). Those previous calculations approximated the isospin effect by simply dividing the total $\\gamma$ width by a factor of 5 for isospin conjugated nuclei and employed equivalent square well potentials for the calculation of the $\\alpha$ transmission coefficients. As expected, differences are larger in comparison to calculations neglecting the isospin suppression. The statistical model rates are compared with reaction rates derived directly from experiment or calculated from resonance parameters which have been measured through different reaction channels. These rates agree reasonably well with the HF predictions in the astrophysically interesting temperature range $10.5$, and approximately solar enriched cluster gas has been detected at redshifts as large as $z\\sim1$ (Hattori et al. 1997). Hence it is natural to suppose that pre-enrichment and preheating are the consequence of very early and vigorous massive star-formation. Cluster gas is evidently a sink for enriched and heated material, and the most likely culprit for this enrichment is dwarf galaxies. Theoretical work has shown that supernovae and OB winds in these low-mass objects should lead to the production of energetic outflows with temperatures on the order of $10^6$ K (Larson 1974; Dekel \\& Silk 1986; Vader 1986). This behavior has been clearly identified in studies of both local starbursting galaxies (Axon \\& Taylor 1978; Marlowe et al.\\ 1995; Heckman 1997; Hunter et al.\\ 1998; Martin 1998) and spectroscopy of high-$z$ galaxies ( Franx et al.\\ 1997; Pettini et al.\\ 1998; Frye \\& Broadhurst 1998; Warren et al.\\ 1998). Whether these outflows lead to a catastrophic loss of the interstellar gas however, is likely to depend on a number of factors (De Young \\& Heckman 1994), and is a subject of current investigations (Murakami \\& Babul 1999; Mac Low \\& Ferrara 1999; Strickland \\& Stevens 1999). In the generally investigated hierarchical models of structure formation such as the Cold Dark Matter model (CDM), the existence of an era of widespread enrichment by outflows from dwarf galaxies is in fact quite natural, because low-mass galaxies are expected to form in large numbers and at early times (e.g., White \\& Frenk 1991). Since these early galaxies will be found preferentially in the large-scale overdense regions that later form clusters, dwarf outflows are the obvious candidates for pre-cluster heating. % The existence of large numbers of small galaxies at high redshifts is also favored by observations, to help understand the steep number counts and low luminosities of faint galaxies (Broadhurst, Ellis, \\& Glazebrook 1992), the sizes of faint galaxies in Hubble Deep Field images (Bouwens, Broadhurst, \\& Silk 1998a,b), and the small sizes of the distant Lyman-break galaxies (Steidel et al.\\ 1999). Locally, however, the space density of dwarf galaxies relative to massive galaxies is far less than predicted on the basis of the steep Press-Schechter slope for the faint end of the mass function (Ferguson \\& Binggeli 1994), prompting theoretical studies of the disruption of dwarves by tidal forces from neighboring objects (Moore et al. 1998), external UV radiation (Kepner, Babul, \\& Spergel 1997; Norman \\& Spaans 1997; Corbelli, Galli, \\& Palla 1997), and catastrophic mass loss during outflows (Larson 1974; Dekel \\& Silk 1986; Vader 1986). While many of these mechanisms may have had an impact on the formation of dwarf galaxies, relatively little attention has been directed towards the influence of outflows on neighboring galaxies. As the earliest galaxies to form were highly clustered (Kaiser 1984) and typical outflow temperatures and velocities were much larger than the virial temperatures and velocities of these galaxies, it is likely that dwarf galaxies were strongly influenced by their neighbors (Scannapieco, Ferrara, \\& Broadhurst 2000). Similarly, the sources responsible for pre-enrichment of the intracluster medium may well have enriched larger protogalaxies, with important consequences for their metallicity histories and cooling times. In this work, we conduct an idealized investigation of the impact of dwarf outflows on the history of galaxy formation. While the consequences of homogeneous heating on galaxy formation have been examined in the past (Blanchard, Valls-Gabaud, \\& Mamon 1992), the inhomogeneous nature of this process and the associated pre-enrichment of galaxies have not been addressed. Our aim is not, however, to construct a complete model of galaxy formation, and thus we do not track processes such as the formation of second-generation stars in galaxies, production of dust, transfer of angular momentum, or the structure of the interstellar medium. Rather we focus on the properties of the inhomogeneously heated and enriched intergalactic medium (IGM) out of which galaxies coalesced, the likely consequences of galaxy formation in this environment, and to what degree these issues must be accounted for within more detailed simulations of galaxy formation. The structure of this work is as follows. In \\S2 we review the observational evidence that leads us to consider a model in which widespread dwarf outflows shocked and enriched the medium out of which larger galaxies formed. In \\S3 we describe a simple numerical Monte Carlo code that we use to asses the overall features of such a model. The results of our simulations are given in \\S4 in which we examine which aspects of galaxy formation are most sensitive to the presence of outflows. In \\S5 we discuss the limitations of our modeling, and conclusions are listed in \\S6. ", "conclusions": "While the impact of preheating and enrichment on the observed properties of galaxy clusters has long been recognized, the impact on the observed properties of galaxies themselves has been little explored. By accounting for the observed outflows from dwarf galaxies, we have been able to show how many of the unexplained properties of galaxies and the IGM can be naturally understood. Firstly, the suppression of low-mass galaxy formation by outflows provides a natural explanation for the factor of $\\sim 4$ discrepancy between the number of observed Milky-Way satellites and predictions from standard CDM models that do not include outflows. Suppression also provides a natural mechanism for the formation of ``dark halos'' which may be associated with the High-Velocity clouds. Secondly, baryonic stripping results in a bell-shaped luminosity function of ellipticals. The lower the mass of a halo, the more likely it is to generate an outflow that strips material from a similar mass neighboring pre-virialized halo that would otherwise later form into a galaxy. This results in very few pairs of neighboring low-mass galaxies, and hence a relative deficit of major low mass-mergers. Finally, our models of enrichment of protogalactic gas predict a trend of increasing metallicity with galaxy mass in good agreement with inferences from observations. The initial metallicity predicted for a Milky-Way mass galaxy is $\\sim 0.1 Z_\\odot$ providing a natural initial floor at the level required to solve the G-dwarf problem and the more general lack of low metallicity stars in well studied massive elliptical galaxies relative to ``closed-box'' models of chemical enrichment. These results are persistent over a wide range of model parameters and cosmologies, and are not a result of fine-tuning parameters or invoking additional physics. While galaxy outflows are already incorporated into modern studies of galaxy formation, this is done only as an internal modification, ignoring pre-enrichment. Keeping track of the effect of outflows on neighboring halos is essential in understanding the properties of galaxy clusters, and hence it is not surprising that these effects would play a major role in the formation of galaxies as well. While the details await further investigation, it is clear that any complete picture of galaxy formation must account for heating and enrichment." }, "0003/astro-ph0003332_arXiv.txt": { "abstract": "When the logarithmic slope of the galaxy counts is lower than 0.4 (this is the case in all filters at large magnitude), the magnification bias due to the lens makes the number density of objects decrease, and consequently, the radial distribution shows a typical depletion curve. In this paper, we present simulations of depletion curves obtained in different filters for a variety of different lens models - e.g., a singular isothermal sphere, an isothermal sphere with a core, a power-law density profile, a singular isothermal ellipsoid and a NFW profile. The different model parameters give rise to different effects, and we show how the filters, model parameters and redshift distributions of the background populations affect the depletion curves. We then compare our simulations to deep VLT observations of the cluster MS1008--1224 and propose to constrain the mass profile of the cluster as well as the ellipticity and the orientation of the mass distribution. This application is based solely on deep photometry of the field and does not require the measurement of shape parameters of faint background galaxies. We finally give some possible applications of this method, useful for cluster lenses. ", "introduction": "Since a few years, a new application of gravitational lensing in clusters of galaxies has started to be explored, namely the depletion effect of number counts of background galaxies in cluster centers. This effect results from the competition between the gravitational magnification that increases the detection of individual objects (at least for flux limited data and marginally resolved objects) and the deviation of light beam that spatially magnifies the observed area and thus decreases the apparent number density of sources. This effect was pointed out as a possible application of the magnification bias by Broadhurst et al. \\cite*{broadhurst2} where they suggested a new method for measuring the projected mass distribution of galaxy clusters, based solely on gravitational magnification of background populations by the cluster gravitational potential. In addition, they suggested that the mass-sheet degeneracy, initially pointed out by Schneider and Seitz \\cite*{schneiderseitz} and observed in mass reconstruction with weak lensing measures, could be broken by using gravitational magnification information which is directly provided by the depletion curves. This method has been used by Fort et al. \\cite*{fort} and by Taylor et al. \\cite*{taylor} to reconstruct a two-dimensional mass map of Abell 1689 in the innermost 27 arcmin$^2$, taking into account the nonlinear clustering of the background population and shot noise. The results are consistent with those inferred from weak shear measurements and from strong lensing. However, the surface mass density cannot be obtained from magnification alone since magnification also depends on the shear caused by matter outside the data field \\cite{young,schneiderweaklensing}. But in practice, if the data field is sufficiently large and no mass concentration lies close to but outside the data field, the mass reconstruction obtained from magnification can be quite accurate \\cite{schneiderking}. This method is an attractive alternative to weak lensing because it is only based on galaxy counts and does not require the measure of shape parameters of very faint galaxies. In addition, it is still valid in the intermediate lensing regime, close to the cluster center, without any strong modifications of the formalism. Meanwhile, it is more sensitive to Poisson noise which increases when the number density decreases in the depletion area. Another weak point identified by Schneider et al. \\cite*{schneiderking} is that it may significantly depend on the galaxy clustering of background objects which can have large fluctuations from one cluster to another. A second application of the magnification bias that has been suggested is to use the shape and the width of depletion curves to reveal the redshift distribution of the background populations. This technique was first used by Fort et al. \\cite*{fort} in the cluster Cl0024+1654 to study the redshift distribution of background sources in the range $26<\\mathrm{B}<28$ and $24<\\mathrm{I}< 26.5$. They found that $60\\% \\pm 10\\%$ of the population in the B band is located between $z=0.9$ and $z=1.1$ while the remaining galaxies are broadly distributed around $z=3$. The population in the I band shows a similar redshift distribution but it extends to larger redshift $(\\simeq 20\\%$ of objects at $z>4)$. The last application of the magnification bias that has been explored is the search for constraints on cosmological parameters. This method is based on the fact that the ratio of the two extreme radii which delimit the depletion area depends on the ratio of the angular distances lens-source and observer-source. Consequently it depends on the cosmological parameters, as soon as all the redshifts are fixed, and mainly on the cosmological constant. This method was first used by Fort et al. \\cite*{fort} in Cl0024+1654 and in Abell 370. Their results favor a flat cosmology with $0.6<\\Omega_{\\Lambda}<0.85$ and are consistent with those obtained from other independent methods (see White \\cite*{white}). This technique requires a good modeling of the lens and could be improved by being extended to a large number of cluster lenses and by using an independent estimate of the redshift distribution (for example from photometric redshifts). However, this method was disputed by Asada \\cite*{asada} who claimed that it is difficult to determine $\\Omega$ by this method without the assumption of a spatially flat universe. He also mentioned the uncertainty of the lens model as one of the most serious problems in this test and concluded that this method cannot be taken as a clear cosmological test to determine $\\Lambda$. In order to better understand the formation of depletion curves and their dependence with the main characteristics of the lenses and the sources, we developed a detailed modeling of the curves in various conditions. In Sect. 2 we present the lens and counts models used in our simulations. In Sect. 3 we study the influence of the lens mass profile, with several sets of analytical mass distributions. In Sect. 4 we explore the influence of background sources distribution, selected through different filters. An application on real data is presented in Sect. 5, in the cluster MS 1008--1224, observed with the VLT. We compare these observations with the simulations in order to derive some constraints on the mass profile and the ellipticity of the cluster, as well as on the background sources distribution. Some prospects about the future of this method are given as a conclusion in Sect. 6.\\\\ Throughout the paper, we adopt a Hubble constant of H$_0=50$ km s$^{-1}$ Mpc$^{-1}$, with $\\Omega_\\Lambda=0$ and $\\Omega_0 = 1$. ", "conclusions": "The gravitational magnification by cluster lenses represents an effective tool to probe the distant universe and the mass distribution in the lenses. The first aim of this paper was to study the effects of lens mass profile model parameters on the typical features of depletion curves. We have also attempted to characterize some features associated with the background redshift distribution of the galaxies. Models were constructed in three bands covering a large spectral range and by using five different lens models. Our simulations agree well with very deep and high quality images of the cluster MS1008--1224 obtained with the VLT and FORS. The depletion effect is clearly seen in this cluster, and we have fitted its radial variation with several sets of mass profiles. The results are quite satisfying as we are able to constrain the mass profile up to a reasonable distance from the center (about 200\\arcsec, or equivalently 1.1 $h_{50}^{-1}$ Mpc). Our results marginally favor the NFW mass profile over the isothermal profile, and more significantly reject a power-law distribution, essentially because of the steep rise of the depletion curve just after its minimum. Note that this region corresponds to the ``intermediate'' lensing regime, where shear measurements are more difficult to relate to the mass distribution. We have also studied the shape of the depletion area, which is easy to relate to the ellipticity of the mass distribution. We were then able to constrain the ellipticity and orientation of the potential of MS1008--1224 with a good accuracy. This preliminary study highlights the need for additional exploration of several issues not fully explored in the present paper. For example, the question of clustering of the background sources still remains, although in the case of MS1008--1224, we have shown how it can be partly eliminated. Schneider et al. (2000) also mention this problem and insist on the fact that, at deep magnitudes, the two point correlation function of the sources is still quite uncertain, but a positive signal does not seem to extend much above $\\sim$ 1 \\arcsec. More quantitatively, we can use the most recent measures of the two point correlation function from the HDF-South \\cite{fynbo}. At the magnitude limits used in our paper, this correlation function does not exceed a few percent for an angular separation of 10\\arcsec , typical of our bin size. Previous measurements were less optimistic in the sense that their values attained 8 to 10 \\%\\ at 10\\arcsec , even at very faint magnitudes. But although there is still only poor knowledge of the amplitude of the correlation function at the faintest levels, we can hope that the effect will not be dramatic in our studies of the magnification bias, provided we retain a reasonable bin size at several arcseconds, to wash out most of the inhomogeneities. Following preliminary approaches from the observational point of view \\cite{taylor,athreya} or a more theoretical one \\cite{schneiderking}, one now clearly needs to extensively and quantitatively compare the weak lensing approach with the magnification bias. In particular, with the new facilities of deep wide-field imaging presently available, most of the difficulties related to a small field-of-view can be overcome: for weak lensing measurement, a complete mass reconstruction requires shear measurements up to the ``no-shear'' region in the outer parts of the cluster to integrate the mass inwards. The absolute normalization of the field number counts for the magnification bias can also be estimated outside the cluster, in exactly the same observing conditions (filter, magnitude limit, seeing, \\ldots ), giving an absolute calibration of the depletion effect. Moreover, the full 2D mass reconstruction of the cluster from the depletion signal alone should be tested. As the signal is directly related to the magnification $\\mu ( \\vec r )$, and then to $\\kappa ( \\vec r )$ and $\\gamma ( \\vec r )$, it should be in principle feasible to invert the depletion map to produce a non-parametric mass map. In practice, the reconstruction is simpler as soon as the shear $\\gamma$ becomes small, because in that case $\\mu$ and $\\kappa$ are simply linearly related. In any case, we have shown in this paper that it is rather easy to reach the outer parts of the cluster up to Mpc scales. This is quite similar to what can be done with deep X-ray maps for intermediate redshift clusters \\cite{soucail}. We want to insist on the fact that the magnification bias effect is easy to detect from the observational point of view because it is less sensitive to seeing conditions or to geometrical distortions of the instruments than are shear measurements \\cite{broadhurst2,fort}. It only requires deeper observations, not necessarily in photometric conditions, provided one is able to reach the outer parts of the cluster to normalise the number counts. In order to improve the mass reconstruction, one may progress with the help of photometric redshifts, quite useful for faint objects, to constrain the redshift distribution of the sources \\cite{pello,bolzonella}. This approach might also be useful in order to limit the influence of large scale over-density fluctuations when evaluating the magnification and the asymptotic limit of the depletion curves. This of course requires deep multi-color photometry, such as the one obtained on MS1008--1224. Extending this study up to large-scale structures (LSS) would probably be more difficult to implement as compared to the search for cosmic shear \\cite{vanwaerbeke}, and the depletion effect should be restricted to cluster scales, at least with the simple method used in this paper. Finally, one of our initial prospects was to try to constrain the background redshift distribution with multi-wavelength observations of the depletion effect. We have shown that this is quite a difficult task as the wavelength dependence of the depletion curves is a kind of second order effect. Nevertheless, this may be an interesting point to explore at other wavelengths, where the background redshift distribution is quite different than in the optical. For example, deep ISO observations in the mid-IR of a few cluster lenses \\cite{altieri} may represent an extension of our analysis, as well as submm observations with SCUBA, provided enough sources are detected behind the lenses for a statistical analysis. Another possibility would be to address the question of the nature of the X-ray background sources and their redshift distribution through deep and high resolution observations of clusters with the new X-ray satellites Chandra and XMM \\cite{refregier}." }, "0003/astro-ph0003042_arXiv.txt": { "abstract": "The soft gamma repeater SGR 1900+14 was observed in Pushchino observatory since 1988 December using BSA radio telescope operating at 111 MHz. We have detected the pulsed radio emission (Shitov 1999) with the same 5.16 s period that was reported earlier for this object (Hurley et al. 1998). The timing analysis has shown that this new radio pulsar PSR J1907+0919 associated with SGR 1900+14 has a superstrong magnetic field, which is $8.1 \\cdot 10^{14}$G, thereby confirming that it is a \"magnetar\" (Duncan \\& Thompson 1992; Kouveliotou et al. 1999). The dispersion measure of PSR J1907+0919 is $281.4(9) pc \\cdot cm^{-3}$ which gives an estimate of the pulsar's distance as about 5.8 kpc. ", "introduction": "Among known soft gamma- ray repeaters to date, only SGR 1900+14 and SGR 1806-20 are the objects for which a secular spin-down of the pulse periods (5.16 s and 7.47 s accordingly) with $\\dot{P}$ of order $10^{-10}$s/s was detected and thereby was established that these SGRs are neutron stars with a superstrong magnetic field of order $10^{15}$G (Kouveliotou et al. 1998; Kouveliotou et al. 1999), called as a \"magnetars\" (Duncan \\& Thompson 1992). Since the end of 1998 we carried out the observations of the SGR 1900+14 at low frequency (111 MHz) and have detected the periodic pulsed radio emission from this magnetar (Shitov 1999). In this paper we report the results of our observations obtained till August 1999. ", "conclusions": "New radio pulsar PSR J1907+0919 associated with the soft gamma repeater SGR 1900+14 is representative of a new class of pulsars with a superstrong magnetic field, slow down value of which for this pulsar is $8.1 \\cdot 10^{14}$ G. Presented results confirm that this object is a magnetar (Duncan \\& Thompson 1992; Kouveliotou et al. 1999). The pulsar distance of about 5.8 kpc determined from dispersion measure $DM = 281 pc \\cdot cm^{-3}$ supports the suggested earlier in a number of papers genetic connection of SGR 1900+14 with supernova remnants SNR G42.8+0.6. As timing analysis have shown there is no evidence for binary orbital motion of this pulsar, at least with $P_{orb} < 250$ days and with $a \\cdot \\sin{i} > 60$ ms." }, "0003/astro-ph0003274_arXiv.txt": { "abstract": "\\normalsize{ We use direct numerical simulations to study the evolution, or relaxation, of magnetic configurations to an equilibrium state. We use the full single-fluid equations of motion for a magnetized, non-resistive, but viscous fluid; and a Lagrangian approach is used to obtain exact solutions for the magnetic field. As a result, the topology of the magnetic field remains unchanged, which makes it possible to study the case of topological nonequilibrium. We find two cases for which such nonequilibrium appears, indicating that these configurations may develop singular current sheets. } ", "introduction": "\\label{I} Formation of singularities, or current sheets, is one of the striking features of astrophysical as well as tokamak plasmas \\cite{plasma}. Such singularities are key to understanding active phenomena related to fast magnetic field reconnection \\cite{flares}, \\cite{golub}. For example, fast dynamos rely on fast reconnection of magnetic field lines \\cite{priest}, \\cite{dynamo}. Despite their importance, key issues related to current sheet formation are still not well understood. Supposing, e.g., that they are formed due to instabilities, one has to assume that fluid dynamical processes are able to slowly deform equilibrium magnetic field configurations (and thereby build up regions of field gradients) without significant reconnection until a marginal state is reached. At this threshold, instability-driven reconnection would then lead to release of the stored free energy on the (observed) time scales thought to be too short to be consistent with, for example, Sweet-Parker reconnection \\cite{golub}, \\cite{priest}. However, it has been long recognized \\cite{galeev} that in the presence of reconnection, it is not obvious how one can attain (meta)stable configurations which store significant free energy. Furthermore, it is not clear why reconnection would not simply return the system to the marginal state, thus releasing only a small fraction of the available free energy. In this paper, we explore one possible solution to these puzzles: We consider specific magnetic field configurations which could arise from a slow evolution of (stable) quasi-equilibria, and then examine their subsequent (unforced) evolution. Our aim is to show that there exist configurations that evolve initially on the slow rate, but that can reach a point at which spontaneous current sheet formation occurs. These configurations have been referred to as ``topological nonequilibria\" (TN) \\cite{flares}, \\cite{parker}, and lead to situations in which the topology of the field is such that in a relaxed equilibrium state it inevitably contains discontinuities. TN results in spontaneous reconnection, because no external forces are involved; and in the cases we shall examine, the result is that extraction of all of the available free energy becomes possible. Finally, we note that an important aspect of this problem relates to the fact that there is a direct correspondence between magnetostatic equilibria and steady Euler flows, as pointed out by Moffatt \\cite{analogy}; this problem is therefore closely connected to the possible formation of singularities in hydrodynamics; see also \\cite{more}, \\cite{Bajer}. ", "conclusions": "The fundamental result emerging from our simulations is that the vanishing magnetic field on the outermost field line imposes strict constraints on the geometry of equilibrium: The type A topology can be at equilibrium only if it is axisymmetric; and therefore, if constrained by external walls, it is at TN. Similarly, the type B rosette structure develops discontinuities, but only in the presence of an external zero-line. The presence of zero-lines is thus an important aspect of topological nonequilibria. Finally, we comment briefly on the applicability of these results to astrophysical situations. Observations of the solar atmosphere \\cite{Shibata93} commonly show topologically unconnected magnetic flux systems which are seen to interact (viz., emerging flux loops). In such circumstances, in which one expects to encounter small but finite resistivity, these flux systems are initially unlinked, but as they are pushed together (and begin to reconnect), flux linkage is expected to occur and to lead to a field topology analogous to that depicted in Fig.\\ 2, or to the generic type B configuration, discussed here. The magnetic flux surrounding these two islands would be initially weak, and the current sheet which is formed is therefore expected to be weak. However, during the course of reconnection, more flux will be pushed outside the two islands, thus accelerating the process of reconnection. This process may therefore be self-accelerating, resulting in final (spontaneous) reconnection; preliminary numerical simulations of a resistive case of this sort suggest that the reconnection rate $v_d$ scales as $c_A/S^\\alpha$, where $\\alpha$ is a small power, $\\alpha\\sim O(0.1)$ \\cite{jon}. If confirmed, it would imply that the reconnection is fast enough to satisfy the observed (solar) constraints on reconnection times. (Recall that while the Sweet-Parker reconnection time for typical parameters corresponding to the solar corona is about three years, the time corresponding to $v_d=c_A/S^{0.1}$ is only 30 minutes, which is comparable to the energy release time scale for large solar flares, related to the so-called ``long-enduring\" events \\cite{golub}). Therefore the two topologies depicted in Figs.\\ 2 and 3 may be regarded as generic examples of ``fast\" reconnection and activity in magnetically active astrophysical systems." }, "0003/astro-ph0003038_arXiv.txt": { "abstract": "We explore the possibility of using the \\( B \\)-type polarization of the \\Comc to map the large-scale structures of the Universe taking advantage of the lens effects on the CMB polarization. The functional relation between the \\( B \\) component with the primordial CMB polarization and the line-of-sight mass distribution is explicited. Noting that a sizeable fraction (at least 40\\%) of the dark halo population which is responsible of this effect can also be detected in galaxy weak lensing survey, we present statistical quantities that should exhibit a strong sensitivity to this overlapping. We stress that it would be a sound test of the gravitational instability picture, independent on many systematic effects that may hamper lensing detection in CMB or galaxy survey alone. Moreover we estimate the intrinsic cosmic variance of the amplitude of this effect to be less than 8\\% for a \\( 100\\, \\textrm{deg}^{2} \\) survey with a \\( 10' \\) CMB beam. Its measurement would then provide us with an original mean for constraining the cosmological parameters, more particularly, as it turns out, the cosmological constant \\( \\Lambda \\). ", "introduction": "In the new era of precision cosmology we are entering in, the forthcoming experiments will provides us with accurate data on \\Comc anisotropies\\cite{CMBexperiments}. This should lead to accurate determinations of the cosmological parameters, provided the large-scale structures of the Universe indeed formed from gravitational instabilities of initial adiabatic scalar perturbations. It has been soon realized however that even with the most precise experiments, the cosmological parameter space is degenerate when the primary \\comc anisotropies alone are considered\\cite{BondEfst}. Complementary data, that may be subject to more uncontrollable systematics are thus required, such as supernovae surveys\\cite{SNLmbd} (but see \\cite{SNBof}) or constraints derived from the large-scale structure properties. Among the latter, weak lensing surveys are probably the safer\\cite{survlens}, but still have not yet proved to be accurate enough with the present day observations. Secondary \\comc anisotropies (i.e. induced by a subsequent interaction of the photons with the mass or matter fluctuations) offer opportunities for raising this degeneracy. Lens effects\\cite{TlensEffects} are particularly attractive since they are expected to be one of the most important.They also are entirely driven by the properties of the dark matter fluctuations, the physics of which involve only gravitational dynamics, and are therefore totally controlled by the cosmological parameters and not by details on galaxy or star formation rates. More importantly an unambiguous detection of the lens effects on \\comc maps would be a precious confirmation of the gravitational instability picture. Methods to detect the lens effects on \\comc maps have been proposed recently. High order correlation functions\\cite{T4pt}, peak ellipticities\\cite{peakEllip} or large scale lens induced correlators\\cite{SeljakZal} have been proposed for detecting such effects. All of them are however very sensitive to cosmic variance since lens effect is only a sub-dominant alteration of the \\comc temperature patterns. The situation is different when one considers the polarization properties. The reason is that in standard cosmological models temperature fluctuations at small scale are dominated by scalar perturbations. Therefore the pseudo-scalar part, the so called \\( B \\) component, of the polarization is negligible compared to its scalar part (the \\( E \\) component) and can only be significant when \\comc lens couplings are present. This mechanism has been recognized in earlier papers\\cite{B2E,BBcorde}. The aim of this paper is to study systematically the properties of the lens induced B field and uncover its properties. In section \\ref{LensEffectSec}, we perturbatively compute the lens effect on the \\comc polarization \\( E \\) and \\( B \\) field. This first order equation is illustrated by numerical experiments. Possibility of direct reconstruction of the projected mass distribution is also examined. As it has already been noted a significant fraction of the potential wells that deflect the \\comc photons can actually be mapped in local weak lensing surveys\\cite{SperZut,WBBellip}. This feature has been considered so far in relation to the \\comc temperature fluctuations. We extend in Section \\ref{CrossSec} these studies to the \\comc polarization exploiting the specificities of the field found in previous section. In particular we propose two quantities that can be built from weak lensing and \\Comc polarization surveys, the average value of which does not vanish in presence of \\comc lens effects. Compared to direct analysis of the \\comc polarization, such tools have the joint advantage of being less sensitive to systematics --systematic errors coming from \\comc mapping on one side and weak lensing measurement on the other side have no reason to correlate!-- and so emerge even in presence of noisy data, and of being an efficient probe of the cosmological constant. Indeed the expected amplitude of correlation is directly sensitive to the relative length of the optical bench, from the galaxy source plane to the \\comc plane, which is mainly sensitive to the cosmological constant. Filtering effects and cosmic variance estimation of such quantities are considered in this section as well. ", "conclusions": "We have computed a first order mapping that describes, in real space, the weak lensing effects on the \\comc polarization. In particular we derived the explicit mathematical relation between the primary \\comc polarization and the shear field at leading order in lens effect. It demonstrates that a \\( B \\)-component of the polarization field can be induced by lens couplings. We have shown however that the \\( B \\)-map alone cannot lead to a non-ambiguous reconstruction of the projected mass map. Nonetheless, the \\( B \\)-component can potentially exhibit a significant correlation signal with local weak lensing surveys. This opens a new window for detecting lens effects on \\comc maps. In particular, and contrary to previous studies involving the temperature maps alone, we found that such a correlation can be measured with a rather high signal to noise ratio even in surveys of rather modest size and resolution. Anticipating data sets that should be available in the near future, (\\( 100\\, \\textrm{deg}^{2} \\) survey, with \\( 5' \\) resolution for galaxy survey and \\( 10' \\) Gaussian beam size for \\comc polarization detection), we have obtained a cosmic variance around \\( 8\\% \\). Needless is to say that this estimation does not take into account systematics and possible foreground contaminations. It shows anyway that \\Comc polarization contains a precious window for studying the large scale mass distribution and consequently putting new constraints on the cosmological parameters. In this paper we have investigated specific quantities that would accessible to observations. They both would permit to put constraint on the cosmological constant. The simulated maps we presented here are only of illustrative interest. We plan to complement this study with extensive numerical experiments to validate our results (in particular on the cosmic variance), and explore the effect of realistic ingredients we did not include in our simple analytical framework, a shear non-gaussianity, lens-lens coupling and so forth." }, "0003/astro-ph0003454_arXiv.txt": { "abstract": "A long flux enhancement, with an exponential decay time of 86\\,min, is detected in 4U\\,1735$-$44 with the BeppoSAX Wide Field Cameras. We argue that this is a type\\,I X-ray burst, making it the longest such burst ever observed. Current theories for thermonuclear bursts predict shorter and more frequent bursts for the observed persistent accretion rate. ", "introduction": "Of the $\\simeq$150 low-mass X-ray binaries known in our galaxy, about 40\\%\\ show occasional bursts of X-rays, in which a rapid rise, lasting from less than a second to $\\simeq$10\\,s, is followed by a slower decay, lasting between $\\simeq$10\\,s to minutes. During the decay the characteristic temperature of the X-ray spectrum decreases. An X-ray burst is explained as energy release by rapid nuclear fusion of material on the surface of a neutron star and thus an X-ray burst is thought to identify the compact object emitting it unambiguously as a neutron star. If the burst is very luminous, reaching the Eddington limit $L_{\\rm Edd}$, the energy release may temporarily lift the neutron star atmosphere to radii of order 100\\,km. Reviews of observations of X-ray bursts are given by Lewin et al.\\ (1993, 1995). The properties of a burst depend, according to theory, on the mass and radius of the neutron star, on the rate with which material is accreted onto the neutron star, and on the composition of the accreted material. It is hoped that a detailed study of X-ray bursts can be used to determine the mass and radius of the neutron star, via the relation between luminosity, effective temperature and flux, and via the changes in the general relativistic correction to this relation when the atmosphere expands from the neutron star surface to a larger radius. However, the physics of the X-ray burst is complex. There is evidence that the emitting area does not cover the whole neutron star and changes with the accretion rate. Reviews of the theory of X-ray bursts are given by Bildsten (1998, 2000). \\nocite{lpt93}\\nocite{lpt95}\\nocite{bil00}\\nocite{bil98} In this paper we describe a long flux enhancement that we observed with the Wide Field Cameras of BeppoSAX in the X-ray burst source 4U\\,1735$-$44, and argue that this event is the longest type I X-ray burst ever observed. In Sect.\\,2 we describe the observations and data extraction, in Sect.\\,3 the properties of the flux enhancement. A discussion and comparison with earlier long bursts is given in Sect.\\,4. In the remaining part of this section we briefly describe earlier observations of 4U\\,1735$-$44. 4U\\,1735$-$44 is a relatively bright low-mass X-ray binary. Smale et al.\\ (1986) fit EXOSAT data in the 1.4-11\\,keV range with a power law of photon index 1.8 with an exponential cutoff above 7\\,keV, absorbed by an interstellar column $\\nh\\simeq 5\\times 10^{20}\\cmsq$. The flux in the 1.4-11\\,keV range is $\\simeq4\\times10^{-9}\\ergcms$. Van Paradijs et al.\\ (1988) show that a sum of thermal bremsstrahlung of $\\simeq10$\\,keV and black body radiation of $\\simeq2$\\,keV, absorbed by an interstellar column $\\nh<8\\times 10^{20}\\cmsq$, adequately describes EXOSAT data in the same energy range and at a similar flux level, obtained one year later. A similar spectrum, with a higher absorption column $\\nh\\simeq 3.4\\times 10^{21}\\cmsq$, fits the Einstein solid-state spectrometer and monitor proportional counter data (Christian \\&\\ Swank 1997). During GINGA observations, the source was somewhat brighter, at $\\simeq9\\times10^{-9}\\ergcms$ in the 1-37\\,keV range (Seon et al.\\ 1997). Bursts were detected at irregular time intervals during each of the five occasions in 1977 and 1978 that SAS-3 observed 4U\\,1735$-$44, leading to a total of 53 detected bursts (Lewin et al.\\ 1980). EXOSAT detected one burst in 1984 (Smale et al.\\ 1986) and five bursts during a continous 80\\,hr observation in 1985 (Van Paradijs et al.\\ 1988), one rather bright burst was detected with GINGA in 1991 (Seon et al.\\ 1997), and five X-ray bursts with RXTE in 1998 (Ford et al.\\ 1998). Burst intervals range from about 30 minutes to more than 50 hrs. Three of the bursts observed with EXOSAT and the single burst observed with GINGA were radius expansion bursts (Damen et al.\\ 1990, Seon et al.\\ 1997), and have been used to determine the distance to 4U\\,1735$-$44 as about 9.2\\,kpc (Van Paradijs and White 1995). 4U\\,1735$-$44 was the first X-ray burster for which an optical counterpart was found: V926 Sco (McClintock et al.\\ 1977). From optical photometry an orbital period of 4.65 hrs was derived (Corbet et al.\\ 1986). \\nocite{lpch80}\\nocite{scc+86}\\nocite{ppl+88}\\nocite{smy+97}\\nocite{dml+90} \\nocite{pw95}\\nocite{mbd+77}\\nocite{ctc+86}\\nocite{fkp+98} ", "conclusions": "In addition to the thermonuclear X-ray bursts, also called type I bursts, low-mass X-ray binaries show other sudden enhancements in X-ray flux. Type II bursts are different from type I bursts in that type II bursts do not show cooling of the characteristic temperature of the X-ray spectrum during the decline. X-ray flares have an irregular flux evolution. Type II bursts are thought to be accretion events; the nature of flares is unknown. The flux enhancement of 4U\\,1735$-$44 shows a smooth exponential decay of the countrate and of the characteristic temperature. Its rise must have been shorter than the decline. A black body gives a good fit to the observed spectrum, for a radius as expected from a neutron star, similar to earlier, ordinary bursts of 4U\\,1735$-$44. All these properties indicate a type\\,I burst. The only special property of the new burst is its duration, which when expressed as the ratio of fluence $E_{\\rm b}$ and peak flux $F_{\\rm max}$: $E_{\\rm b}/F_{\\rm max}>3400$\\,s, is more than 300 times longer than the longest burst observed previously from this source (see Lewin et al.\\ 1980). This duration also translates in a fluence which is several orders of magnitude larger than the previous record holder for 4U\\,1735$-$44, because the peak flux is similar to those of normal type\\,I bursts. The fluence of a type\\,I burst which burns all matter deposited onto a neutron star since the previous burst must be $\\simeq$1\\%\\ of the accretion energy released by deposition of this matter. We do not have a measurement to the previous burst, but in seven days following the burst no other burst was observed. Multiplying this time by the persistent luminosity we obtain $\\simeq2.7\\times10^{43}$\\,erg, or about 50 times the energy of the burst, well in the range of previously observed ratios for type\\,I bursts. The presence of clear cooling argues against a type\\,II burst; this and the smooth decay argues against a flare. If the flux enhancement were due to an accretion event, the amount of matter dropped extra onto the neutron star (assuming a mass of $1.4\\msun$ and a radius of 10\\,km) must have been $>3\\times10^{21}$\\,g, which may be compared to the average accretion rate of $2.3\\times10^{17}$\\,g\\,s$^{-1}$ derived for the persistent flux. If the inner part of the accretion disk would have depleted itself onto the neutron star during the flux enhancement, one would expect the accretion rate immediately after to be lower than before. The observations suggest the opposite. We conclude that a type\\,I X-ray burst is the best explanation for the enhanced flux event. We consider it significant that the occurrence of this burst is accompanied by the absence of any ordinary -- i.e.\\ short -- burst throughout our 9-day observation, whereas all previous observations of 4U\\,1735$-$44 did detect ordinary bursts (see Introduction). Searching the literature for long bursts we find that the longest type\\,I burst published previously is a radius expansion burst observed with SAS-3, probably in 4U\\,1708$-$23 (Hoffman et al.\\ 1978; see also Lewin et al.\\ 1995). The ratio of fluence and peak flux for that burst was $\\simeq 500$\\,s, so that the BeppoSAX WFC burst of 4U\\,1735$-$44 lasted at least six times longer. Other events published as long bursts from Aql\\,X-1 (Czerny et al.\\ 1987) and from X\\,1905+000 (Chevalier and Ilovaisky 1990) are in fact relatively short bursts followed by an enhanced constant flux level which persisted for several hours: in both cases the flux declined to 1/e of the peak level within 20\\,s. These events are clearly different from the long exponential bursts seen in 4U\\,1708$-$23 and 4U\\,1735$-$44. \\nocite{hld+78}\\nocite{ci90}\\nocite{ccg87} From the theoretical point of view, a long interval between bursts would allow hydrogen to burn completely before the onset of the burst, so that the energetics of the burst is dominated by pure helium burning. If matter accreted at a rate of $2.3\\times10^{17}$\\,g\\,s$^{-1}$ during one week, the energy released by helium burning is compatible with the energy of the observed burst. The problem with this model is that theory predicts for this accretion rate that the burst initiates well before hydrogen burning is completed, i.e.\\ that bursts are more frequent and less energetic, in accordance with those previously observed of 4U\\,1735$-$44. Indeed, Fujimoto et al.\\ (1987) find that a burst of $10^4$\\,s duration occurs only for accretion rates $\\dot M<0.01\\dot M_{\\rm Edd}$. The persistent flux during the BeppoSAX observation is a factor $\\simeq20$ higher than this limit; observations previous to ours have consistently found 4U\\,1735$-$44 at a similar luminosity. An alternative model for bursts with a duration of $10^4$\\,s is accretion of pure helium at an accretion rate in excess of the Eddington limit ($\\dot M>5\\times\\dot M_{\\rm Edd}$, Brown \\&\\ Bildsten 1998). The orbital period and optical spectrum indicate a main-sequence, i.e.\\ hydrogen-rich, donor star (Augusteijn et al.\\ 1998). \\nocite{fhir87}\\nocite{bb98} Perhaps the main challenge for any theoretical explanation is that the properties of the persistent flux during our nine day long observation, during which a single very long X-ray burst was observed, are not different from those during earlier observations with EXOSAT when more frequent ordinary bursts were found." }, "0003/astro-ph0003348_arXiv.txt": { "abstract": "We discuss the results expected from observing $\\gamma$ Dor stars with the MONS satellite. We first describe how MONS Star Trackers will consistently improve the present knowledge of this new class of variables. We then discuss how the $\\gamma$ Dor can be considered possible \"bridges\" between opacity driven pulsations and stochastically excited solar like oscillations. For this reason the possible inclusion of a $\\gamma$ Dor representative in the MONS main target list is suggested and discussed. Three possible candidates are presented: $\\gamma$ Dor, 9 Aur and BS 2740. ", "introduction": "The $\\gamma$ Dor stars have been recently defined (Kaye et al. 1999) as a class of variable stars in the lower-red part of the Cepheid instability strip, undergoing $g$-mode pulsation. Stars in this group show (multi)periodic photometric and spectroscopic variability with typical time scales between 0.3 and 3\\,d and typical amplitudes below $0^m.1$. The definition of the class was based on 12 objects contained in a so-called {\\it master list}. The principal physical properties of these stars are reported in Table 1. Several objects suspected to belong to the class have been recently added by Aerts, Eyer \\& Kestens (1998) and Handler (1999a) upon examination of HIPPARCOS light curves. At the present date the data-base of known $\\gamma$ Dor stars and related objects (Handler, private communication) contains 141 objects divided into {\\it Bona Fide} (12), {\\it Prime} (46) and {\\it Other} (83) candidates. Bona fide objects in the master list are stars for which extensive photometric and spectroscopic data prove that pulsations must be the cause for the variability. Prime and Other $\\gamma$ Dor candidates are stars for which either a moderate or a limited amount of photometric and/or spectroscopic data is available, allowing to suspect or speculate that nonradial pulsation (NRP) is the cause of the observed variations. \\begin{table}[h] \\footnotesize \\begin{center} \\begin{tabular}{cccccccccccc} \\tableline Star &$b-y$ & $$ & ST & $v \\sin i$ & $\\pi$ & [$Me/H$] & $M_V$ & $L$ & $T_{\\rm eff}$ & $R$ & $M$ \\\\ & (mag) & (mag) & & (kms$^{-1}$) & (mas) & & (mag) &$[L_\\odot]$ & (K) & $[R_\\odot]$ & $[M_\\odot]$ \\\\ \\tableline HD 224945 & 0.192 & 6.93 & F0 {\\sc v} & 55 & 16.92 & $-0.30$ & 3.07 & 5.1 & 7250 & 1.43 & 1.51 \\\\ $\\gamma$ Dor & 0.201 & 4.25 & F0 {\\sc v} & 62 & 49.26 & $-0.02$ & 2.72 & 7.0 & 7200 & 1.70 & 1.57 \\\\ 9 Aur & 0.217 & 5.00 & F0 {\\sc v} & 18 & 38.14 & $-0.19$ & 2.89 & 6.0 & 7100 & 1.62 & 1.52 \\\\ BS 2740 & 0.219 & 4.49 & F0 {\\sc v} & 40 & 47.22 & $-0.15$ & 2.86 & 6.2 & 7100 & 1.64 & 1.53 \\\\ HD 62454 & 0.214 & 7.15 & F1 {\\sc v} & 53 & 11.18 & ~~0.16 & 2.39 & 9.5 & 7125 & 2.02 & 1.66 \\\\ HD 68192 & 0.227 & 7.16 & F2 {\\sc v} & 85 & 10.67 & ~~0.05 & 2.30 & 10.5 & 7000 & 2.20 & 1.71 \\\\ HD 108100 & 0.234 & 7.14 & F2 {\\sc v} & 68 & 12.10 & $-0.03$ & 2.53 & 8.5 & 6950 & 2.01 & 1.62 \\\\ HD 164615 & 0.226 & 7.06 & F2 {\\sc iv} & 66 & 14.36 & ~~0.20 & 2.82 & 6.5 & 7000 & 1.73 & 1.53 \\\\ BS 6767 & 0.183 & 6.40 & F0 {\\sc v}n & 135 & 17.44 & $-0.10$ & 2.59 & 7.9 & 7300 & 1.76 & 1.61 \\\\ BS 8330 & 0.225 & 6.20 & F2 {\\sc iv} & 38 & 19.90 & $-0.01$ & 2.67 & 7.4 & 7000 & 1.85 & 1.57 \\\\ BS 8799 & 0.181 & 5.99 & kA5 & 45 & 25.04 & $-0.36$ & 2.96 & 5.7 & 7375 & 1.46 & 1.54 \\\\ HD 224638 & 0.198 & 7.49 & F1 {\\sc v}s & 24 & 12.56 & $-0.15$ & 2.98 & 5.5 & 7200 & 1.51 & 1.52 \\\\ \\tableline \\normalsize \\end{tabular} \\end{center} \\vskip-0.8truecm \\caption{Observational parameters of the confirmed $\\gamma$ Dor variables and calculated or inferred basic properties. Table taken from Kaye et al. (1999). \\label{deftab1} } \\end{table} The most relevant problems in the observation of $\\gamma$ Dor stars concern the amplitudes and time scales of their variability. Amplitudes are small and the clustering of the typical frequencies around 1\\,d$^{-1}$ creates conflicts with the 1\\,d$^{-1}$ alias typical of ground-based single-site observations and requires to organize multi-longitude campaigns. At the same time even in these campaigns problems of alignment, improper correction for atmospheric extinction and fluctuation in the instrumental setup often introduce spurious low-frequency terms that have the chance to be misinterpreted as signal. For this reason $\\gamma$ Dor stars will benefit as many other classes of variables of continuous out-of-atmospheric coverage as the one provided by a space mission like MONS. The interested community has largely discussed if $\\gamma$ Dor stars have any relevance for asteroseismology. These stars pulsate in $g$-modes that in some sense carry more information on the stellar interior than $p$-modes, but the low number of modes excited and the intrinsic uncertainties in the application of standard mode-identification techniques seem to make a seismic interpretation rather difficult. The $\\gamma$ Dor stars give us however two additional possibilities for seismological investigations which other classes of pulsating star cannot provide. Firstly, the domain of $\\gamma$ Dor stars in the HR diagram partly overlaps with that of the $\\delta$ Scuti stars (Handler 1999a). It can therefore be suspected that stars showing both types of pulsation could be present. This will allow to constrain the physical properties of a star tightly from the $\\delta$ Scuti-type $p$ modes, which helps in identifying the $\\gamma$ Dor-type $g$ modes which can then be used to probe the star's deep interior. Indeed, short-period light variations have already been detected for two $\\gamma$ Dor stars (see Handler 1999b for the first discovery). Secondly, Guzik et al. (2000) pointed out the major role played by convection in $\\gamma$ Dor pulsation theory. These authors calculated nonadiabatic pulsation properties of evolutionary models lying near the $\\gamma$ Dor instability region in the H-R diagram. The models considered have relatively deep envelope convection zones. By using the Pesnell (1990) non-adiabatic pulsation code these authors calculated the $l=0,1,2$ pulsation frequencies and found unstable high order $g$-modes with frequencies between 4 and 25 $\\mu$Hz (periods $\\sim$ 0.4 to 3 days). Furthermore the mode kinetic energy was found to reach a minimum at a frequency of about 11 $\\mu$Hz (period of $\\sim$ 1 d), i.e. near the most commonly observed $\\gamma$ Dor period. The frequency spacing between modes with same $l$ and subsequent $n$ was found to be $\\sim 0.1$ d$^{-1}$, exactly matching the typical frequency spacing observed in multiperiodic $\\gamma$ Dor stars. These results strongly suggest that asteroseismology of $\\gamma$ Dor stars might be easier as it seemed until very recently. Another interesting outcome of the model calculations originates from the \"frozen-in convection\" approximation used, in which fluctuations in the convective luminosity are set to zero during the pulsation cycle. Because convection does not adapt to transport the luminosity in this approximation, the radiation is periodically blocked by the high opacity at the convection zone base, resulting in pulsational driving. Therefore the overstable $g$-modes in $\\gamma$ Dor stars are first maintained by an opacity bump at the basis of the convective layer which has nothing to do with the classical $\\kappa/\\gamma$ mechanism and \"tunnel\" to the surface through a convective envelope which is \"lazy\" enough not to adjust and damp them. Second the turbulent motion in the convective envelope could trigger solar-like oscillations as in any other stars of this mass, temperature and luminosity. Indeed the temperature and luminosity range occupied by $\\gamma$ Dor stars completely overlaps with the one in which solar-like oscillations are predicted. In this sense $\\gamma$ Dor stars might represent a sort of \"bridge\" between overstable pulsation and stochastically generated solar-like oscillations. The MONS primary telescope represents an unique opportunity to investigate such a possible link. For this reason we encourage the MONS science team to consider the inclusion of an object of this kind in the primary target list. ", "conclusions": "The $\\gamma$ Dor stars are known to exhibit $g$-mode pulsations at time scales that make their observation difficult from the ground because of aliasing problems. The location of the class in the HR diagram and some theoretical arguments imply that solar-like oscillations could be excited in these stars as well. The MONS satellite is then the ideal tool to investigate both aspects taking profit of the clean spectral window of the STs and the precision of the CAM. We therefore strongly support the inclusion of a suitable candidate in the main target list and the recording of as many targets as possible with the STs." }, "0003/astro-ph0003313_arXiv.txt": { "abstract": "We have analyzed all the observations of BP Tauri taken by the International Ultraviolet Explorer in the low resolution ($\\bigtriangleup \\lambda \\sim 6 \\rm{\\AA}$), long wavelength (from $\\lambda=1850 \\rm{\\AA}$ to $\\lambda=3350 \\rm{\\AA}$) range. This dataset contains 61 spectra. We observe variability in the ultraviolet continuum of $\\bigtriangleup m_{cont.} \\sim 1 $ magnitude and variability in the \\ion{Mg}{2} line flux of $\\bigtriangleup m_{\\rm{Mg II}} \\sim 0.8 $ magnitudes. Moreover, these spectra do not show any correlation between the continuum flux and the \\ion{Mg}{2} line flux, thus resolving a standing controversy in the literature concerning the origin of the \\ion{Mg}{2} line flux. There is no correlation between the color temperature of the UV continuum and the average value of its flux. Using models of the accretion process developed by Calvet \\& Gullbring (1998), we obtain energy fluxes, accretion spot sizes, and accretion rates from the IUE observations of BP Tauri. We find average energy fluxes of $5.0\\ 10^{11} \\rm{ergs\\ cm^{-2}\\ s^{-1}}$, average spot sizes of $4.4\\ 10^{-3}$ times the stellar surface, and average accretion rates of $1.6\\ 10^{-8} {\\rm{\\msun/yr}}$. Our analysis shows that the particle energy flux and the UV flux in the stellar surface are proportional to each other. Most strikingly, we observe a correlation between accretion rate and spot size, with the spot size increasing as the square of the accretion rate. Based on the results of a simulation, we conclude that geometrical effects (i.e. the varying inclination of the spot with respect to the observer) are not enough to account for this effect. Current models of the accretion process fail to reproduce such an effect, suggesting the need of using more realistic descriptions of the stellar field when treating magnetospheric accretion. There may also be an unmodelled efficiency factor that determines how matter is loaded into the field lines. Non-dipole fields, geometry, oblique shocks and the possibility of ``limb brightening'' should be taken into account when creating models and explaining the results of observations of T-Tauri stars. ", "introduction": "The International Ultraviolet Explorer (IUE) was operational for almost 20 years and took over 107000 images of a wide array of objects. In particular, it observed about 130 T-Tauri stars (TTSs) in the low resolution ($\\bigtriangleup \\lambda \\sim 6 \\rm{\\AA}$), long wavelength range (from $\\lambda=1850 \\rm{\\AA}$ to $\\lambda=3350 \\rm{\\AA}$). This dataset constitutes the most complete set of observations of TTSs in the Balmer range. Now that the IUE Final Archive has been completed and all the data is in the public domain\\setcounter{footnote}{0}\\footnote{http://archive.stsci.edu/iue/} we have undertaken the project of exploring all this dataset. In this paper we will examine the observations of BP Tauri (also known as HBC 32 and HD 281934). A complete discussion of the full dataset will appear elsewhere. We have chosen this star in particular because of the large number of observations by IUE: there are 61 low resolution spectra. They will allow us to study correlations between different physical aspects of the T-Tauri phenomenon over a long period of time. BP Tauri is a single (Ghez, Neugebauer \\& Matthews 1993), Classical T-Tauri star (CTTS) located in L1495 (Lynds 1962) in the Taurus-Auriga complex of dark clouds at a distance of 140 pc (Elias 1978). The star has a constant radial velocity close to that of its molecular surroundings and a projected equatorial velocity of $v {\\rm sin}i < 10 $ km/s (Hartmann et al. 1986; Basri \\& Batalha 1990). The spectral type of BP Tauri corresponds to a K5-K7 dwarf with a strong absorption in the lithium resonance line (Duncan 1991; Hartigan et al. 1989). The mass is $M=0.5 - 0.8 \\msun$ (Gulbring 1994; Gullbring et al. 1998). Photometric observations (Vrba et al. 1986; Bouvier, Bertout \\& Bouchet 1988; Simon, Vrba \\& Herbst 1990 - SVH90 - and the observations summarized by Rydgren et al. 1984) have shown that BP Tauri has a large infrared emission and is variable in all optical bands, with a range in B of up to 3 magnitudes (Herbig \\& Bell 1988). Rapid fluctuations on time-scales of minutes have been observed by Schneeberger, Worden \\& Africano (1979). These fluctuations make it difficult to extract a possible photometric period. No period at all (Bouvier et al. 1988, Simon et al. 1990) and periods of 6.1 days (Simon et al. 1990), 7.6 days (Vrba et al. 1986), 7.7 days (Osterloh, Thommes \\& Kania 1996), 8.3 days (Richter et al. 1992), and $\\sim 10$ days (G\\'omez de Castro \\& Franqueira 1997 - GdCF97) have been proposed. The fact that BP Tauri grows redder as it becomes fainter has been used to explain this variability in the period in terms of a hot spot on the surface of the star (see for example SVH90). In a small fraction of CTTSs, spots cooler than the photosphere have also been observed, but no evidence of cool spots exists for BP Tauri (Gullbring et al. 1996). Variations in the measured period of BP Tauri might correspond to the hot spot (or spots) moving to different latitudes in a differentially rotating star. An inclination of $i\\sim 30^o-50^o$ for BP Tauri has been derived on the basis of the measured $v \\sin i$ and the longest measured periods (SVH90; Gullbring 1994). The variability in the photometric period can also be explained by considering a beat frequency between the Keplerian frequency of clumps of material circling the star and the rotation rate of the stellar surface (Bouvier et al. 1999; Smith, Bonnell \\& Lewis 1995; Smith, Lewis \\& Bonnell 1995). Measurements of spot sizes in this and other TTSs give values that vary between a fraction of a percent to almost 40\\% the visible stellar surface (Vrba et al. 1986; Vrba et al. 1993; Herbst, Herbst \\& Grossman 1994; Gullbring 1994; Bouvier et al. 1995; Fern\\'andez \\& Eiroa 1996). As has been observed in many CTTSs, the spectrum of BP Tauri appears veiled by variable continuum excess emission. This is the origin of the photometric variability. The optical veiling has been observed to be between 30 and 75\\% the value of the photospheric flux (Basri \\& Batalha 1990; Valenti, Basri \\& Johns 1993; Hartigan, Edwards \\& Ghandour 1995; Gullbring et al. 1998). Some spectral lines (notably, the Balmer series and the \\ion{Mg}{2} line) show evidence of blueshifted and redshifted absorption (see for example, Johns \\& Basri 1995ab; GdCF97). It has also been noted that line emission and the continuum level are sometimes correlated (GdCF97; SVH90). We will explore one of these correlations in Section 3. These observations can be framed in terms of the current paradigm for CTTS. As we understand it today, a CTTS consists of a late-type pre-main-sequence star with a magnetic field, surrounded by an accreting, dusty disk. At the distance from the star at which the torque due to the magnetic field equals the viscous torque in the disk, the disk is truncated. Some material from the disk will then be captured by the magnetic field (see for example K\\\"onigl 1991) and some will escape as a wind (Shu et al. 1994). Collisional coupling between ions and neutrals is such that even the neutral gas component follows the magnetic field lines (Martin, 1996). At the stellar surface, the material will have supersonic velocity. The simplest models assume that this velocity should be the free-fall velocity ($v_{ff}\\sim 300 \\ \\rm{km \\ s^{-1}}$, Calvet \\& Gullbring 1998) but more complicated models slow down the material to a fraction ot this velocity (Ostriker \\& Shu 1995). Nevertheless, a shock will occur some distance above the star. The shocked gas will heat up the surface of the star and a hot spot will be seen. Emission from this hot spot is thought to be responsible for the visible and ultraviolet excess observed in CTTS spectra (Valenti et al. 1993; Calvet \\& Gullbring 1998; Lamzin 1998). Furthermore, emission from the whole accretion flow is believed to be responsible for other features of the TTS spectra, in particular, the strong emission in the Balmer lines (Hartmann, Hewett \\& Calvet 1994; Johns-Krull \\& Basri 1997) and some atomic lines (Batalha et al. 1996; Beristain, Edwards \\& Kwan 1998). In this model redshifted absorption features would correspond to material falling into the accretion spot. Blueshifted absorption features would correspond to absorption by an optically thin wind. It is unclear at this point what fraction of the excess emission (if any) is due to the wind. In the context of the magnetospheric accretion model, the ultraviolet is a particularly interesting range. Emission from the underlying star is low compared to that of the veiling continuum. This means that, even though is not possible to measure the veiling directly (as it is possible in the visible range), conclusions derived from the ultraviolet analysis are not very sensitive to errors in the spectral type. In this paper we will explore the IUE dataset for BP Tauri. Relatively little work has been done on the individual IUE observations and their relation to the accretion paradigm. Here we assume the excess ultraviolet continuum emission to be caused by accretion. We start with a discussion of the raw data. A more complete discussion (including many more stars) will appear soon, in a paper dealing with the full low-resolution IUE dataset. In Section 3, we will use our dataset to study the correlation between the UV continuum and the flux in the \\ion{Mg}{2} blend. To understand the IUE data to the fullest we need a physical model. Therefore in Section 4 we will use the Calvet \\& Gullbring (1998) (CG98) description of the accretion shock to interpret our spectra in terms of accretion rates. Section 5 and 6 contain the analysis and the conclusions. ", "conclusions": "We have analyzed all the observations of BP Tauri taken by the International Ultraviolet Explorer in the low resolution ($\\bigtriangleup \\lambda \\sim 6 \\rm{\\AA}$), long wavelength range (from $\\lambda=1850 \\rm{\\AA}$ to $\\lambda=3350 \\rm{\\AA}$). We found 61 useful spectra. Compared with spectra of standard stars, all the BP Tauri spectra show strong ultraviolet excess and color temperatures around $8000$ K. Using 43 spectra, we observe variability in the ultraviolet continuum of $\\bigtriangleup m_{cont.} \\sim 1 $ magnitude and variability in the \\ion{Mg}{2} line flux of $\\bigtriangleup m_{\\rm{Mg II}} \\sim 0.8 $ magnitudes. We observe no correlation between continuum flux and \\ion{Mg}{2} line flux. This result resolves a controversy in the literature regarding the relationship of the continuum flux and \\ion{Mg}{2} line flux in BP Tauri. We conclude that the bulk of the \\ion{Mg}{2} in BP Tauri is not emitted below the accretion shock. Therefore, it may be emitted by the accretion column, the stellar wind and/or magnetospheric flaring. We do not observe any correlation between color temperature and the average value of the continuum. Within the context of the magnetospheric accretion paradigm, we have used models by CG98 to obtain the energy flux into the star due to the accretion process, and the spot size for each of the spectra of BP Tauri. We assume a constant inner truncation radius for the disk of $5\\ R_*$. With 45 spectra, we find an average energy flux of $5\\ 10^{11} {\\rm{ergs\\ cm^{-2}\\ sec^{-1}}}$ and an average projected spot size of $4.4\\ 10^{-3}$ times the area of one hemisphere of the star. The latter result coincides with previous observational determinations of spot sizes. From the energy flux and the spot size we obtain the accretion rate times $\\cos \\theta$, where $\\theta$ is the average inclination angle of the region that produces the emission. The average of this projected accretion rate is $1.6\\ 10^{-8} \\msun \\ \\rm{yr^{-1}}\\pm 30\\%$, consistent with values found by other researchers. This values are weakly sensitive to the value of the underlying star used obtain the continuum excess. No correlation is observed between the flux in \\ion{Mg}{2} line and the energy flux, the accretion rate or spot size, affirming the conclusion from the previous paragraph. The nature of the data and the models is such that we cannot say anything about intrinsic variations in the accretion rate or the real (i.e., deprojected) spot size, but we conclude that the dispersion in their ratio is $\\dot{M}/f\\sim 50\\%$, consistent with the errors. We observe a correlation between the ultraviolet flux and the accretion rate, similar to correlations found by other researchers between the U-band flux and the accretion rate: ${\\log (L_{acc}/ \\lsun)}={(1.1\\pm0.1)*\\log (L_{2900 \\rm{\\AA}}/ \\lsun)-(0.5\\pm0.05)}$. This correlation is implicit in the models by CG98. Also, we observe a correlation between the accretion rate ${\\dot M}\\cos\\theta$ and the projected spot size $f$: ${\\dot M}\\cos\\theta\\ \\alpha\\ f^{0.5\\pm0.05}$. Together, the two correlations imply that the UV flux in the surface of the star is proportional to the particle flux. These results are significant, in spite of the large errors involved in the determinations of each quantity. These correlations stand in contrast to results obtained by Calvet \\& Gullbring (1998). Using individual optical observations of 15 CTTSs, they concluded that the accretion rate increases linearly with spot size. A resolution of this discrepancy will have to wait until we have applied the magnetospheric accretion models to other stars. Using a simulation, we conclude that the geometry of the system determined by the observer, the rotation axis and the inclination of the star can not explain the correlation between accretion rates and spot sizes. One must look at other more fundamental explanations. Models of magnetospheric accretion by Ostriker \\& Shu (1995) and Mahdavi \\& Kenyon (1998) do predict that the spot size should increase with the accretion rate , but in both models, the predicted increase is much slower than what we found here ($f\\ \\alpha \\ {\\dot M}^2$). The correlation between accretion rate and spot size could be due to an unmodelled efficiency factor that determines how the spot is filled. Conversely, it indicates that the spot size is very sensitive to accretion rate. Future modeling should take into account the possibility of a non-dipole field for the star, oblique shocks, limb brightening and the effects of the geometry determined by the observer. The correlation between accretion rate and spot size puts a strong constraint in these models." }, "0003/astro-ph0003255_arXiv.txt": { "abstract": "A qualitative study of the surface trapping of low-frequency non-radial $g$-modes in rotating early-type stars is undertaken within the Cowling, adiabatic and traditional approximations. A dispersion relation describing the local character of waves in a rotating star is derived; this dispersion relation is then used to construct propagation diagrams for a $7\\,\\msun$ stellar model, which show the location and extent of wave trapping zones inside the star. It is demonstrated that, at frequencies below a cut-off, waves cannot be fully trapped within the star, and will leak through the surface. Expressions for the cut-off frequency are derived in both the non-rotating and rotating cases; it is found from these expressions that the cut-off frequency increases with the rotation rate for all but prograde sectoral modes. While waves below the cut-off cannot be reflected at the stellar surface, the presence of a sub-surface convective region in the stellar model, due to \\hetwo\\ ionization, means that they can become partially trapped within the star. The energy leakage associated with such waves, which are assigned the moniker {\\em virtual modes} due to their discrete eigenfrequencies, means that stability analyses which disregard their existence (by assuming perfect reflection at the stellar surface) may be in error. The results are of possible relevance to the 53 Per and SPB classes of variable star, which exhibit pulsation frequencies of the same order of magnitude as the cut-off frequencies found for the stellar model. It is suggested that observations either of an upper limit on variability periods (corresponding to the cut-off), or of line-profile variations due to virtual modes, may permit asteroseismological studies of the outer layers of these systems. ", "introduction": "\\label{sec:introduction} The self-excitation of global non-radial pulsation modes in a star is a prime example of positive feedback, whereby small oscillatory perturbations grow in amplitude via the efficient conversion of heat into vibrational energy by a suitable driving mechanism (see, \\eg, Unno \\etal\\ 1989 for a comprehensive review of the topic). A fundamental ingredient in the feedback loop is that the oscillations must be trapped in some part of the stellar interior, so that energy does not leak from the system faster than it can be generated. Such trapping can occur when a pair of evanescent regions, where traveling waves cannot be supported, enclose a propagative region; waves are repeatedly reflected at the two evanescent boundaries, and the resulting superposition leads to a standing wave of the normal-mode type. For waves excited in stellar envelopes, it is common for the surface layers to serve as one of the evanescent regions required for the formation of a trapping zone. Ando \\& Osaki \\shortcite{AndOsa1975} demonstrated that such a situation occurs in the Sun, where low-order $p$-modes are trapped beneath the photosphere, supporting a model first put forward by Ulrich \\shortcite{Ulr1970} to explain the five-minute solar oscillation \\cite{LeiNoySim1962}. However, the trapping is only effective for modes with frequencies below some cut-off; higher-frequency modes cannot be reflected at the photosphere, and will leak through the stellar surface. This issue was addressed in detail by Ando \\& Osaki \\shortcite{AndOsa1977}, who found that, although leakage does occur through the solar photosphere at frequencies above the cut-off, some waves can subsequently be reflected at the chromosphere-corona interface, and standing waves are able to form. More recently, Balmforth \\& Gough \\shortcite{BalGou1990} suggested that such coronal reflection can explain apparent observations of high-frequency chromospheric standing waves \\cite{FleDeu1989}, although debate concerning this interpretation still continues \\cite{Kum1994,DzhSta1995,Jef1998}. Pulsation in massive, early stars (types O and B) is qualitatively quite different from the solar case, due to the gross structural differences between the two stellar classes. However, it is still subject to the same wave trapping requirements, since the underlying physics remains the same. Shibahashi \\& Osaki \\shortcite{ShiOsa1976}, in their study of $g$-modes trapped within the hydrogen-burning shell of evolved massive stars, found that high-frequency (low-order) modes can tunnel through an evanescent region separating core and envelope, and thence escape from the star. A complementary situation was discussed by Osaki \\shortcite{Osa1977} when studying pulsation in Cepheid-type stars; non-radial $p$-modes trapped within the envelope were able to tunnel through an evanescent region into the core, where they were damped rapidly without reflection at the centre. In both cases, the appropriate region of the star was modeled as an isolating oscillating unit with the inclusion of wave leakage at one boundary. The leakage was found to stabilize some modes which would otherwise have been self-excited, due to the associated loss of vibrational energy from the star. Shibahashi \\shortcite{Shi1979} analyzed wave trapping in an idealized stellar model (corresponding to an evolved massive star) using an asymptotic method, and discussed in some depth these two cases; in addition, he considered the situation where low-frequency (high-order) $g$-modes are able to tunnel through an evanescent region in the envelope and thence escape through the stellar surface. More recently, however, relatively little attention has been shown regarding wave trapping issues at the surface of early-type stars; in particular, stability analyses \\cite{Cox1992,Kir1992,DziPam1993,GauSai1993}, based on the new opacity calculations of Rogers \\& Iglesias \\shortcite{RogIgl1992} and Seaton \\shortcite{Sea1992}, have assumed that the Lagrangian pressure perturbation $\\delta p$ tends to zero or some limiting value at the stellar surface. Such an assumption corresponds to the {\\em ab initio} condition that waves incident from the interior are totally reflected at the stellar surface; the possibility of leakage is thereby disregarded, and no consideration of trapping issues is undertaken. This is the first in a short series of papers studying the surface trapping of low-frequency $g$-modes in early-type stars, in an attempt to re-open discussion of, and investigation into, this important area. Much of the work is conceptually developed from that of Ando \\& Osaki \\shortcite{AndOsa1975}; however, in light of recent research into the influence of rotation on low-frequency modes \\cite{LeeSai1990,Bil1996,LeeSai1997}, and due to the fact that significant rotation appears to be commonplace in O- and B-star populations \\cite{How1997}, the theory is updated to include rotational effects. The current paper serves as a introduction, covering the more qualitative, general aspects of the study; subsequent papers will investigate various issues arising from this paper in greater depth. The following section reviews the pulsation equations appropriate for low-frequency $g$-modes in rotating stars, whilst section \\ref{sec:dispersion} derives the dispersion relation corresponding to these equations. The trapping of waves described by this dispersion relation is examined in section \\ref{sec:trapping} with the aid of propagation diagrams, and the effect of rotation on the eigenfrequencies of individual modes is discussed in section \\ref{sec:eigenfrequency}. The findings are discussed in section \\ref{sec:discussion}, and summarized in section \\ref{sec:conclusions}. ", "conclusions": "\\label{sec:conclusions} The prime conclusion to be drawn from the work presented herein is that the complete trapping of low-frequency $g$-modes beneath the surface of early-type stars is not guaranteed. This is especially the case in rotating stars, where the trapping cut-off frequency $\\omegat$ can be significantly increased by the action of the Coriolis force for all but the prograde sectoral modes. The fact that the latter are more effectively trapped in rapidly-rotating stars than other types of modes may explain anecdotal observational evidence which points to their favoured excitation. As a consequence of the dependence of $g$-mode eigenfrequencies on the rotation rate, the hypothesis has been put forward that the set of radial orders $\\{n\\}$ of trapped $g$-modes is invariant under the influence of rotation. Stability analyses which contain the {\\em ab initio} assumption of complete wave reflection at the stellar surface might be in error at frequencies below the cut-off $\\omegat$. More rigorous calculations can include the possibility of wave leakage, by adopting a more physically-realistic outer mechanical boundary condition. Such calculations will reveal to what extent advective damping associated with leaking virtual modes might suppress the self-excitation of these modes. These points may be of especial relevance to the 53 Per and SPB classes of variable stars." }, "0003/astro-ph0003063_arXiv.txt": { "abstract": "One of the most interesting results of the X-ray observations with the ASCA satellite of the Galactic center (GC) is the spatial distribution and the intensity of the Fe line at 6.4 keV. Up to now the morphology and the intensity of this line have been a puzzle. In this letter we present a map of the GC in the \\uc\\ line of SiO covering the same region than the ASCA observations. The SiO emission is restricted to molecular clouds with radial velocity between 10 and 60 \\kms in the Sgr A and Sgr B complexes. We find a correlation between the SiO morphology and the spatial distribution of the Fe 6.4 keV line, on the large scale and also within Sgr A and Sgr B. The SiO abundance increases by a factor of $\\gsim$20 in the regions with strong Fe 6.4 keV line. This indicates that the Fe 6.4 keV line mainly arises from molecular clouds with large gas phase abundance of refractory elements. We discuss the implications of the correlation on the origin of the hard X-rays, and the heating and the chemistry of the molecular clouds in the GC. ", "introduction": "The Galactic center (GC) is a strong source of diffuse X-ray emission in the 2-10 keV energy range and in lines from several ions \\cite{kawai88,sunyaev93,koyama96,koyama96,sidoli99}. Recently, the ASCA satellite mapped the X-ray emission from the GC \\cite{koyama89}. One of the most interesting results is the spatial distribution and the intensity of the iron K\\galpha\\ lines. The Fe K\\galpha\\ lines of highly ionized ions (He-like at $\\sim$6.70 keV and H-like at $\\sim$6.97 keV) arise from hot gas with temperatures of $\\sim$9 keV. This emission is concentrated towards Sgr A-West and symmetrically distributed along the galactic plane with a spatial distribution similar to that of the radio continuum emission and the molecular clouds \\cite{maeda96}. This is in sharp contrast with the K\\galpha\\ line from neutral or low ionized Fe atoms at $\\sim$6.40 keV (hereafter Fe\\deg\\ line) which shows emission only towards the Sgr A and Sgr B complexes \\cite{koyama96}. The Fe\\deg\\ line emission is caused by fluorescence and appears when neutral cold molecular clouds are exposed to a strong source of hard X-rays. In X-ray irradiated molecular clouds like those in the GC, it is expected that the X-rays will influence the heating, the ionization and the chemistry of these clouds (see \\eg Hollenbach et al., 1997). It is well known that the physical conditions and the chemistry of the molecular clouds in the GC differ substantially from those in the galactic disk (see \\eg Morris \\& Serabyn, 1996). High gas kinetic temperature \\cite{hutte93}, and large abundance of SiO are typical in the GC \\cite{Minh92,martin97,hutte98}. The origin of these unusual characteristics is unclear, but it is believed to be due to strong shocks in the GC \\cite{wilson82,martin97}. In this letter we present a correlation between the SiO radio emission and the Fe\\deg\\ line, suggesting that X-rays may play an important role in the heating and the chemistry of the GC molecular clouds. \\begin{figure*} \\rotatebox{-90}{\\includegraphics{fig1_xrays.ps}} \\caption{a-c) Spatial distribution of the integrated emission of the SiO \\uc\\ line, of the Fe\\deg\\ line (Koyama et al 1997) and of the CS \\du\\ line (Bally et al. 1987) in the GC. The velocity range used to calculate the integrated intensity for SiO and CS was \\menos100 to 100 \\kmsns. The contour levels are from 10 to 100 K\\kms in steps of 10 K\\kms for the SiO map, and 15 to 150 K\\kms in steps of 15 K\\kms for the CS map. For the Fe\\deg\\ map the contour levels (in 10$^{-6}$ counts/sec/0.106 min$^2$) are 0.25 and 0.4 to 1 by 0.1. The dots in Fig. 1a show the location where the SiO spectra were taken. d-e) Comparison of the averaged line profiles of the \\uc\\ line of SiO and of the \\du\\ line of CS for Sgr A and Sgr B respectively for the regions with and without Fe\\deg\\ emission. The areas used to obtain the averaged spectra towards Sgr A and Sgr B are shown in Fig. 1a by parallel lines at constant Declination. The region with Fe\\deg\\ emission has been defined using a threshold intensity for this line of 0.25$\\times$10$^{-6}$ counts/sec/0.106 min$^2$.} \\label{f:mapas} \\end{figure*} ", "conclusions": "" }, "0003/astro-ph0003019_arXiv.txt": { "abstract": "We present $Beppo$SAX observations of the $\\gamma $-ray emitting quasars 0836+710, 1510-089 and 2230+114. All the objects have been detected in the PDS up to 100 keV and have extremely flat power-law spectra above 2 keV ($\\alpha _x$=0.3--0.5). 0836+710 shows absorption higher than the galactic value and marginal evidence for the presence of the redshifted 6.4 keV Iron line. 1510-089 shows a spectral break around 1 keV, with the low energy spectrum steeper ($\\alpha _l$=1.6) than the high energy power-law ($\\alpha _h$=0.3). The data are discussed in the light of current Inverse Compton models for the high energy emission. ", "introduction": "Since the EGRET detection of about 60 blazars as strong $\\gamma $-ray emitters the study of these extreme objects has received a renewed interest. The overall Spectral Energy Distribution (SED) of Blazars shows two broad components, the first one peaking at IR-up to soft X-rays, the second one in the $\\gamma $-rays, from MeV up to TeV energies. The first peak is due to synchrotron radiation produced by relativistic electrons, while the high energy component is believed to be Inverse Compton scattered radiation. The seed photons for the IC scattering could be the synchrotron photons themselves (SSC model) or photons produced in the region external to the jet (EC model) which, especially for quasars with strong emission lines, is probably rich of optical-UV radiation. The subclass of quasar-like sources contains the most luminous sources, with apparent $\\gamma $-ray luminosity up to $10^{48}$ erg s$^{-1}$. In the following we present the $Beppo$SAX observations of three gamma-loud quasars (0836+710, 2230+114 and 1510-089), detected up to 100 keV with the high energy instrument PDS, and we discuss the external Compton scenario. A full paper is in preparation (Tavecchio et al. 2000). ", "conclusions": "\\noindent $\\bullet$ We constructed the SEDs of the observed sources using contemporaneous X-ray and optical observations and historical data taken from the literature (an example is reported in Fig.1) We have reproduced the observed spectrum using the homogeneous EC model discussed in detail in Ghisellini et al. (1998). In a spherical region with size $R$, a power-law electron distribution (with slope $n$ and limits $\\gamma _{min}$ and $\\gamma _{max}$) is continuously injected with luminosity $L_{inj}$. Electrons cool through the synchrotron and IC processes and are free to escape from the source at some velocity $v_{esc}$, forming a flat ($\\alpha < 0.5$) power-law below $\\gamma _{min}$. We assume that the external radiation field is described by a black body spectrum with luminosity $L_{ext}$, diluted in a spherical region with size $R_{ext}$. The parameter values for the models are reported in Tab. 2. \\noindent $\\bullet$ It is interesting to note that the derived spectral indices in the medium to hard X-ray band are flatter than 0.5 in two out of three sources. A population of electrons cooling through synchrotron and IC forms a distribution which produce a spectrum with $\\alpha =0.5$. Therefore some additional mechanism is required in order to produce the observed flatter spectrum, e.g. escape or injection of an intrinsically flat distribution (see e.g. Ghisellini 1996). \\noindent $\\bullet$ The excess of 1510-089 in the soft band is well understood as due to the SSC emission (see Fig.1), although another possible source is the tail of the strong UV bump. \\noindent $\\bullet$ We confirm the presence of absorption higher than the galactic one in 0836+710. The origin of this absorption is likely intrinsic to the source. The fluorescence iron line suggested by our data could be produced through reprocessing by the same material responsible for the absorption. On the other hand a broken power-law continuum with galactic absorption can reproduce the data equally well: in this case the break could be due to the incomplete comptonization of the soft external photons (see Ghisellini 1996). \\begin{figure} \\hskip -1.1cm \\vskip -1.9cm \\psfig{file=prova_ratio.ps, width=5.7cm, height=6cm} \\vskip -6.4cm \\hskip 5.35cm \\psfig{file=1510_sed.ps, width=7.cm} \\caption[]{Left panel: Data/Model ratio for 0836+710 with the power-law model ($N_H$ free). Right panel: Overall SED of 1510-089 with the spectrum calculated using the homogeneous EC model. Radio to UV (open squares) and gamma-ray data are taken from the literature.} \\vskip -.2cm \\end{figure}" }, "0003/astro-ph0003369_arXiv.txt": { "abstract": "Astrometric observations of microlensing events were originally proposed to determine the lens proper motion with which the physical parameters of lenses can be better constrained. In this proceeding, we demonstrate that besides this original usage astrometric microlensing observations can be additionally used in obtaining various important information about lenses. First, we demonstrate that the lens brightness can be determined with astrometric observations, enabling one to know whether the event is caused by a bright star or a dark lens. Second, we show that with additional information from astrometric observations one can resolve the ambiguity of the photometric binary lens fit and thus uniquely determine the binary lens parameters. Finally, we propose two astrometric methods that can resolve the degeneracy in the photometric lens parallax determination. Since one can measure both the proper motion and the parallax by these methods, the lens parameters of individual events can be uniquely determined. ", "introduction": "When a source is microlensed, it is split into two images. The flux sum of the individual images is greater than that of the unlensed source, and thus the source becomes brighter during the event. The sizes and brightnesses of the individual images change as the lens-source separation changes due to their transverse motion. Therefore, microlensing events can be detected either by photometrically monitoring the source brightness changes or by directly imaging the two separated images. However, with the current instrument direct imaging of the separate images is impossible due to the low precision of the instrument. As a result, current microlensing observations have been and are being carried out only by using the photometric method (Aubourg et al.\\ 1993; Alcock et al.\\ 1993; Udalski et al.\\ 1993; Alard \\& Guibert 1997). However, if an event is astrometrically observed by using the planned high precision interferometers from space-based platform, e.g.\\ the {\\it Space Interferometry Mission} (SIM), and ground-based interferometers soon available on 8-10 m class telescope, e.g.\\ the Keck and the Very Large Telescope, one can measure the shift of the source star image centroid caused by microlensing. The astrometric centroid shift vector as measured with respect to the position of the unlensed source is related to the lens parameters by \\begin{equation} \\bidelta\\bitheta_{c,0}={\\theta_{\\rm E}\\over u^2+2} \\left[ \\left( {t-t_0\\over t_{\\rm E}}\\right) \\hat{\\bf x} + \\beta \\hat{\\bf y}\\right], \\end{equation} where $\\theta_{\\rm E}$ is the angular Einstein ring radius, $t_{\\rm E}$ is the time required for the source to cross $\\theta_{\\rm E}$ (Einstein time scale), $t_0$ is the time of the closest lens-source approach (and thus the time of maximum amplification), and $\\beta$ is the separation at this moment (i.e.\\ impact parameter). The notation ${\\bf x}$ and ${\\bf y}$ represent the unit vectors with their directions that are parallel and normal to the lens-source proper motion. If one defines $x=\\delta\\theta_{c,x}$ and $y=\\delta\\theta_{c,y}-\\beta\\theta_{\\rm E}/2(\\beta^2+2)$, equation (1) becomes \\begin{equation} x^2 + {y^2\\over q^2} = a^2, \\end{equation} where \\begin{equation} a = {\\theta_{\\rm E}\\over 2(\\beta^2+2)^{1/2}}, \\end{equation} and \\begin{equation} q = {\\beta \\over (\\beta^2+2)^{1/2}}. \\end{equation} Therefore, during the event the image centroid traces out an elliptical trajectory (hereafter astrometric ellipse) with a semi-major axis $a$ and an axis ratio $q$. In Figure 1, we present astrometric ellipses for several example microlensing events with various lens-source impact parameters. \\begin{figure}[t] \\plotfiddle{ellipse.eps}{4.0cm}{0}{40}{40}{-120}{-130} \\vskip0.9cm \\caption{Trajectory of the source star image centroid shifts for several example microlensing events with various lens-source impact parameters. The directions of $x$- and $y$-axis are parallel and normal to the lens-source proper motion, respectively. } \\end{figure} The greatest importance of astrometric microlensing observation is that one can determine $\\theta_{\\rm E}$ from the observed astrometric ellipse (H\\o\\hskip-1pt g, Novikov \\& Polarev 1995; Walker 1995; Paczy\\'nski 1998; Boden, Shao, \\& Van Buren 1998). This is because the size (i.e.\\ semi-major axis) of the astrometric ellipse is directly proportional to $\\theta_{\\rm E}$ [see equation (3)]. Once $\\theta_{\\rm E}$ is determined, the lens proper motion is determined by $\\mu = \\theta_{\\rm E}/t_{\\rm E}$ with the independently determined $t_{\\rm E}$ from the light curve. While the photometrically determine $t_{\\rm E}$ depends on the three physical lens parameters of the lens mass ($M$), location ($D_{ol}$), and the transverse motion ($v$), the astrometrically determined $\\mu$ depends only on the two parameters of $M$ and $D_{ol}$. Therefore, by measuring $\\mu$ one can significantly better constrain the nature of lens matter. However, we note that to completely resolve the lens parameter degeneracy, it is still required to additionally determine the lens parallax (see more details in \\S\\ 4). In this proceeding, we demonstrate that besides this original usage astrometric microlensing observations can be additionally used in obtaining various important information about lenses. First, we show that the lens brightness can be determined with astrometric observations, enabling one to know whether the event is caused by a bright star or a dark lens (\\S\\ 2). Second, we demonstrate that additional astrometric microlensing observations allow one to uniquely determine the binary lens parameters (\\S\\ 3). Finally, we propose two astrometric methods that can uniquely determine the lens parallax, with which one can completely break the lens parameter degeneracy along with the measured proper motion (\\S\\ 4). \\begin{figure}[t] \\plotfiddle{omega.eps}{4.0cm}{0}{28.8}{28.8}{-170}{-70} \\plotfiddle{difference.eps}{4.0cm}{0}{30}{30}{0}{49} \\vskip-4.0cm \\caption{Left part: Astrometric behavior of bright lens events. The ellipses represents the trajectories of image centroid shift for events caused by bright lenses with various lens/source flux ratio $\\ell_L/\\ell_S$. The straight arrows represent the position vectors of the image centroid at different times during events. The curved arrow represents the direction of centroid motion with the progress of time. All example events have the same impact parameter of $\\beta=0.5$. Right part: Difference between the impact parameters determined from the centroid shift trajectory, $\\beta$, and the angular speed curve, $\\beta_0$, as a function of lens-source brightness difference.} \\end{figure} ", "conclusions": "In this proceeding, we demonstrate various additional usages of astrometric microlensing observations besides the original usage of the lens proper motion determination. These are summarized as follows. \\begin{enumerate} \\item By astrometrically observing a microlensing event caused by a bright lens, one can identify the bright lens and measure its flux. \\item With additional information from astrometric observations one can resolve the ambiguity of the photometric binary lens fit and thus uniquely determine the binary lens parameters. \\item With application of the two proposed astrometric methods, the degeneracy in the photometric lens parallax determination can be resolved, allowing one to completely break the lens parameter degeneracy along with simultaneously determined lens proper motion. \\end{enumerate}" }, "0003/astro-ph0003196_arXiv.txt": { "abstract": "Integral field optical spectroscopy with the INTEGRAL fiber-fed system and {\\it{HST}} optical imaging are used to map the complex stellar and warm ionized gas structure in the ultraluminous infrared galaxy IRAS 12112+0305. Images reconstructed from wavelength-delimited extractions of the integral field spectra reveal that the observed ionized gas distribution is decoupled from the stellar main body of the galaxy, with the dominant continuum and emission-line regions separated by projected distances of up to 7.5 kpc. The two optical nuclei are detected as apparently faint emission-line regions, and their optical properties are consistent with being dust-enshrouded weak-[O\\thinspace I] LINERs. The brightest emission-line region is associated with a faint ($m_{\\rm{I}}$= 20.4), giant H\\thinspace II region of 600 pc diameter, where a young ($\\sim$ 5 Myr) massive cluster of about 2 $\\times$ 10$^7$ $M_{\\odot}$ dominates the ionization. Internal reddening towards the line-emitting regions and the optical nuclei ranges from 1 to 8 magnitudes, in the visual. Taken into account the reddening, the overall star formation in IRAS 12112+0305 is dominated by starbursts associated with the two nuclei and corresponding to a star formation rate of 80 $M_{\\odot}$ yr$^{-1}$. ", "introduction": "Integral field spectroscopy of IRAS 12112+0305 was obtained with the INTEGRAL system (Arribas et al. 1998), a fiber-fed spectrograph mounted at the Naysmith No. 1 platform of the 4.2m William Herschel Telescope. The bundle of fibers consisted of 219 fibers, each 0.9$^{''}$ in diameter and covering a 16.5$^{''}$ $\\times$ 12.3$^{''}$ field-of-view. The spectra were taken using a 600 line/mm grating, covering the 5000$-$7900\\AA~ range, with an effective resolution of 4.8\\AA. The total integration time was 7500 sec, split into five separate integrations of 1500 sec each, with seeing $\\approx 1.0^{''}$. The reduction has been done following the standard procedures applied to spectra obtained with two-dimensional fiber spectrographs (Arribas et al. 1997 and references therein). The results are presented in Figure~1, together with the {\\it{HST}} $I$-band image (Borne et al. 2000a) for comparison. The derived astrophysical properties for the main continuum and line-emitting regions are presented in Table~1. Columns 2 and 3 give the relative positions. Column 4 presents the internal reddening values derived from the H$\\alpha$/H$\\beta$ ratio, assuming case B recombination. Column 5 gives the apparent magnitude within an aperture of 0.5$^{''}$ radius using the {\\it{HST}} $I$-band image, and column 6 gives the corresponding absolute magnitude after internal reddening correction assuming $A_{\\rm{I}}$= 1.494 $\\times$ E(B$-$V). Column 7 shows the observed H$\\alpha$ flux obtained using an aperture of 2.1$^{''}$ diameter, while column 8 gives the reddening-corrected H$\\alpha$ luminosity. The last four columns present the logarithm of the reddening-corrected emission line ratios and the corresponding activity classification. ", "conclusions": "\\subsection{Evidence for Decoupled Stellar and Ionized Gas Components} The stellar main body of IRAS 12112+0305 is concentrated in three dominant regions, each separated from each other by about 2$^{''}-3^{''}$ (see the H$\\alpha$ and H$\\beta$ continuum images in Figure~1). Two of these regions (called N$_s$ and N$_n$ hereinafter; see Table~1 for relative positions) located along position angle PA45, were already detected in the near-infrared, and associated with the nuclei of two galaxies involved in the final stages of a merger (Carico et al. 1990). The INTEGRAL-generated continuum images extend the wavelength coverage towards the blue, clearly showing the presence of a large differential extinction towards the southern nucleus (N$_s$). The third region (called R2 hereinafter) is located 3$^{''}$ north of N$_s$ at position angle PA20. The {\\it{HST}} $I$-band image (Figure~1) shows N$_s$ as a high-surface brightness compact region which coincides with the K-band point-like nucleus (Scoville et al. 2000). Regions N$_n$ and R2 are made of several fainter condensations distributed along an arc-like structure of about 5.4$^{''}$ (i.e., 8 kpc) extent and located 3$^{''}$ to 5$^{''}$ (i.e., 4.5 to 7.5 kpc) north$-$northeast of N$_s$. The brightest of these condensations, located 3$^{''}$ ($\\approx$ 4.5 kpc) northeast from N$_s$, is most likely associated with the northern nucleus detected in the INTEGRAL-generated continuum images (N$_n$) and in the near-infrared (Carico et al. 1990; Scoville et al. 2000). Although the overall structure of the ionized gas resembles that of the stellar light distribution, the dominant line-emitting regions do not coincide with the nuclei identified above but, on the contrary, are decoupled from them (see [O\\thinspace III] and H$\\alpha$ maps in Figure~1). The brightest line-emitting region (called R1 hereinafter) is located 5$^{''}$ ($\\approx$ 7.5 kpc) east of N$_s$ along position angle PA80. This region is associated with a faint $I$-band continuum source ($m_I$= 20.4) outside the main body of the galaxy (see {\\it{HST}} $I$-band image in Figure~1). The second-brightest H$\\alpha$ line-emitting region is associated with region R2, and it is therefore composed of a filament of faint continuum condensations that are also detected in the H$\\beta$ and [O\\thinspace III] maps (Figure~1). The optically dominant nucleus of the galaxy (N$_s$) is marginally detected in [O\\thinspace III] and appears as the faintest H$\\alpha$ emission source. The arc-like structure seen in the {\\it{HST}} $I$-band image corresponds to three well resolved H$\\alpha$ line-emitting peaks, the faintest of which is associated with the northern nucleus (N$_n$) detected in the optical and near-infrared (see discussion above). \\subsection{Massive Dust-Enshrouded Starbursts as Nuclear Ionizing Sources} The positions of the two apparently faint ionizing nuclei N$_n$ and N$_s$ coincide (within the {\\it{HST}} absolute astrometry uncertainties of $\\pm$0.7$^{''}-$1$^{''}$) with the two bright compact starburst nuclei detected at radio frequencies (Condon et al. 1991), thus favoring dust-enshrouded nuclear starbursts as the energy sources in these nuclei. The ionization sources in N$_s$ and N$_n$ are highly obscured, with visual extinctions ($A_{\\rm{V}}$) of 8 and 3.5 magnitudes, respectively (see spectra in Figure 2). The two nuclei appear almost equally bright with $I$-band absolute magnitudes of $-$18.9 (N$_s$) and $-$18.5 (N$_n$). However, when reddening is taken into account, N$_s$ is found to be ten times more luminous than N$_n$ (i.e. $M_I = -$22.8). These magnitudes are within the range of those measured for the nuclei of other luminous and ultraluminous infrared galaxies (Surace et al. 1998). The line ratios of the ionization sources associated with the nuclei N$_s$ and N$_n$ correspond to a mixture of weak-[O\\thinspace I] LINER and H\\thinspace II region spectra (see Table~1). The presence of weak-[O\\thinspace I] LINERs in the nucleus of galaxies has been taken as evidence for ionization by hot stars in a high-metallicity environment (Filippenko \\& Terlevich 1992; Shields 1992) or for ionization by a mixture of a low-luminosity AGN and hot stars (Ho et al. 1993). The integrated mid-infrared spectrum of IRAS 12112+0305 does not show any evidence for an AGN (Genzel et al 1998), further supporting the idea that the ionizing sources in the nuclei N$_s$ and N$_n$ are associated with dust-enshrouded starbursts. Although N$_s$ and N$_n$ are minor contributors to the observed H$\\alpha$ emission, the reddening-corrected flux emanating from these nuclei dominates the overall H$\\alpha$ luminosity with a value of 9.1 $\\times$ 10$^{42}$ erg s$^{-1}$ (Table~1). If the H$\\alpha$ flux emitted by the two nuclei were entirely due to stars, then the corresponding star formation rate would amount to about 80 $M_{\\odot}$ yr$^{-1}$ for a Salpeter initial mass function (IMF) with mass limits of 0.1 and 100 $M_{\\odot}$ (Leitherer et al. 1999). \\subsection{Brightest Line-Emitting Region: A Tidally-Induced Giant H\\thinspace II Region?} The {\\it{HST}} image (Figure~1) shows that the apparently more luminous line-emitting peak (R1) is associated with a faint ($m_I$= 20.35) region characterized by a physical size of about 600 parsecs ($\\approx$ 0.4$^{''}$), an $I$-band absolute magnitude of $-$17.7, an internal optical extinction of about one magnitude, and an H$\\alpha$ luminosity of 8.7 $\\times$ 10$^{40}$ erg s$^{-1}$ (see Table~1). The size, H$\\alpha$ luminosity and emission line ratios are typical of circumnuclear star-forming regions in nearby spirals (Gonzalez-Delgado \\& P\\'erez 1997; Planesas, Colina, \\& P\\'erez-Olea 1997), and of giant extragalactic H\\thinspace II regions like NGC 5471 in M101 (Shields 1990). The H$\\alpha$ luminosity and the equivalent widths of the H$\\beta$ (74 $\\pm$ 3\\AA) and H$\\alpha$ (450 $\\pm$ 10 \\AA) emission lines correspond to that of a 5-Myr old ionizing cluster of 2 $\\times$ 10$^{7}$ $M_{\\odot}$, assuming a Salpeter IMF with mass limits of 0.1 and 100 $M_{\\odot}$ (Leitherer et al. 1999). The mass of the ionizing cluster represents only a small fraction ($\\sim$ 3\\%) of the dynamical mass of this region (upper limit of 7.5 $\\times$ 10$^8$ $M_{\\odot}$), calculated assuming virialization and an instrument-corrected emission line width equivalent to a velocity dispersion of 60 km s$^{-1}$ (derived from the line-width measurements of the H$\\beta$ and [O\\thinspace III]5007\\AA~ lines). In summary, the derived properties of this region are characteristic of young massive H\\thinspace II regions and could represent a case for a tidally-induced giant extranuclear self-gravitating star-forming region, or even a dwarf galaxy, decoupled from the much older stellar body of the parent galaxies (Duc \\& Mirabel 1994, 1998)." }, "0003/astro-ph0003475_arXiv.txt": { "abstract": "We have obtained time-resolved ultraviolet spectroscopy for the pulsating DAV stars G226--29 and G185--32, and for the pulsating DBV star PG1351+489 with the Hubble Space Telescope Faint Object Spectrograph, to compare the ultraviolet to the optical pulsation amplitude and determine the pulsation indices. We find that for essentially all observed pulsation modes, the amplitude rises to the ultraviolet as the theoretical models predict for $\\ell=1$ non-radial g-modes. We do not find any pulsation mode visible only in the ultraviolet, nor any modes whose phase flips by $180^\\circ$ in the ultraviolet, as would be expected if high $\\ell$ pulsations were excited. We find one periodicity in the light curve of G185--32, at 141~s, which does not fit theoretical models for the change of amplitude with wavelength of g-mode pulsations. ", "introduction": "Observations of white dwarf stars are an important probe of stellar and galactic evolution. The properties of individual white dwarfs define the endpoints for models of stellar evolution, while the white dwarf luminosity function provides an observational record of star formation in our galaxy. For example, the coolest normal-mass white dwarfs are remnants of stars formed in the earliest epoch of star formation, so their cooling times can tell us the age of the galactic disk in the solar neighborhood (Winget et al. 1987, Wood 1992) and the effects of phase separation and crystallization at extreme densities (Chabrier et al. 1992, Segretain et al. 1994, Winget et al 1997). As the number of pulsation modes detected in the pulsating white dwarfs is insufficient for an inverse solution of the structure of the star, we must identify the pulsation modes to compare with the theoretical models and infer the structural parameters. A crucial step in determining the structure of a white dwarf from its pulsation periods is to identify the pulsation modes correctly. The pulsation modes in our models are indexed with three integers ($k$,$\\ell$,$m$) where $k$ represents the number of nodes in the pulsation eigenfunction along the radial direction, $\\ell$ is the total number of node lines on the stellar surface, and $m$ is the number of node lines passing through the pulsation poles. Pulsation modes with different indices generally have different pulsation periods. The usual procedure for identifying the mode indices is (1) calculate theoretical pulsation periods in models of white dwarfs; (2) compare the pattern of theoretical periods to the observed pattern of periods; (3) adjust the models to bring the theoretical and observed patterns into closer agreement. The problems with this procedure are clear: it does not work for white dwarfs with only a few excited pulsation modes, as it places too few constraints on the stellar structure; and, given the complexity and sophistication of the theoretical calculations and the large number of possible pulsation modes, there is ample opportunity to misidentify modes. Other methods of mode identification must be used to avoid these problems. ", "conclusions": "For a DAV, an $\\ell=1$ mode with 109~s period requires a $k=1$ radial index from pulsation calculations, and therefore the model for G226--29 has to have a thick hydrogen surface layer, around $10^{-4}\\mstr$ (Bradley 1998). The star may have a thick hydrogen layer as well. The effective temperature derived from the pulsation amplitudes, $T_{\\rm{eff}}=11750$, and surface gravity $\\log g=8.23$, indicate a mass of $(0.75 \\pm 0.04) \\msun$, according to the evolutionary models of Wood (1992). As all three modes fit $\\ell=1$, they must be a triplet from a rotationally split $\\ell=1$ mode. Even though the Bergeron et al. (1995) ML2/$\\alpha=0.6$ instability strip runs $12460~{\\rm K} \\geq T_{\\mathrm{eff}} \\geq 11160~{\\rm K}$, we know by comparison of its spectra with other ZZ Cetis that G226--29 is at the blue edge. As it is at the blue edge, such a low temperature indicates the instability strip is at much lower temperature than previously quoted. Also, the observed low amplitude of the pulsations, their short period and the small number of pulsations all indicate it is at the blue edge for its mass, and the higher than average mass suggests a higher temperature instability strip. According to Bradley \\& Winget (1994), the ML3 instability strip for a 0.75~$\\msun$ white dwarf is at 13\\,100~K, 330K hotter than for a 0.6~$\\msun$ star. Giovannini et al. (1998) show that the observed instability strip does depend on mass, as Bradley \\& Winget (1994) predicted. One of the problems with the determination of an effective temperature for a star is that it may vary with the wavelength used in the determination; even though the effective temperature is a bolometric parameter, none of our observations are. The problem is dramatic for stars with surface convection layers, because of the effects of turbulent pressure on the photosphere. None of the model atmospheres calculated with mixing length theory can reproduce the physical non-local processes involved (Canuto \\& Dubovikov 1998), requiring different parameterizations at different depths (Ludwig, Jordan \\& Steffen 1994), and a fine tuning of the convection mixing length coefficient for the wavelength region of interest (Bergeron et al. 1995, Koester \\& Vauclair 1997, Koester \\& Allard 2000). The model atmospheres used assume the atmosphere is in hydrostatic and thermodynamical equilibrium, an assumption that must be examined because the timescale for convection is of the same order of the timescale for pulsation. The periodicity at 141~s for G185--32 does not change its amplitude significantly with wavelength and therefore does not fit any theoretical model. As its period is twice that of the 70.92~s periodicity, one must consider if it is only a pulse shape effect on the 70.92~s periodicity, but normally a pulse shape effect occurs as an harmonic, not a sub-harmonic, because it normally affects the rise and fall of the pulse. We have no plausible explanation for this periodicity, showing that these stars still have much to teach us. The periodicity at 560~s rises slowly to the ultraviolet, but as the HST data sets are short, its amplitude has a large uncertainty due to aliasing, as shown by the phase change from ultraviolet ($-1.5 \\pm 8.2$~s) to zeroth order data ($44.8 \\pm 9.6$~s). Another surprise for G185--32 is the identification of the periodicities at 70.93~s and 72.56~s as $\\ell=1$ modes. The pulsation models are consistent with such short period $\\ell=1$, k=1 mode only for a total mass around $1~M_\\odot$. Higher k values require even larger mass. But the mass determination from our time average spectra, as well the mass determinations by Koester \\& Allard (2000) using the IUE spectra plus V mag and parallax, or the optical spectra of Bergeron et al. (1995) all derive a normal mass around 0.56~$M_\\odot$. One possibility to resolve such difference is if the observed modes were a rotationally split k=0 mode, but that would require that G185--32 be a binary star, to account for the center of mass changes during pulsation. As G185--32 is not a known binary star, such explanation requires the discovery of a companion, possibly with a high signal-to-noise red spectra. The models, while useful, clearly lack some of the pulsation physics present in the star. The splitting of the 70.93 to 72.56~s periodicities, assuming they are m-splitting of the same k and $\\ell$ mode, imply a rotation period around 26~min, but that of the 299.95~s to 301.46~s periodicities would imply a rotation period of 9.7~hr. In the estimate we used \\[P_{\\mathrm{rot}} = \\frac{1-C_{kl}^I}{\\Delta f}\\] where $\\Delta f$ is the frequency splitting, and used $C_{kl}^I\\simeq 0.47$ to 0.48 for k=1, $\\ell=1$. The asymptotic value for $C_{kl}^I$ is 0.5, for $k\\gg 1$. Either value for the rotation period is much shorter than for normal ZZ Ceti stars (around 1 day) so we must also consider the possibility that the 141~s periodicity, which does not follow the g-mode theoretical prediction, and is harmonically related to the 70.93~s periodicity, must arise from some other cause. The 141~s periodicity of G185--32 is a periodic brightness change that is not accompanied by a change in color, suggesting some kind of geometric effect." }, "0003/astro-ph0003461_arXiv.txt": { "abstract": "Low-frequency radio surveys are ideal for selecting orientation-independent samples of extragalactic sources because the sample members are selected by virtue of their isotropic steep-spectrum extended emission. We use the new 7C Redshift Survey along with the brighter 3CRR and 6C samples to investigate the fraction of objects with observed broad emission lines -- the `quasar fraction' -- as a function of redshift and of radio and narrow emission line luminosity. We find that the quasar fraction is more strongly dependent upon luminosity (both narrow line and radio) than it is on redshift. Above a narrow [OII] emission line luminosity of $\\log_{10}(L_{\\mathrm [OII]}/{\\mathrm W}) \\gtsimeq 35$ [or radio luminosity $\\log_{10} (L_{151} /$ W Hz$^{-1}$ sr$^{-1}) \\gtsimeq 26.5$], the quasar fraction is virtually independent of redshift and luminosity; this is consistent with a simple unified scheme with an obscuring torus with a half-opening angle $\\theta_{\\rm trans} \\approx 53^{\\circ}$. For objects with less luminous narrow lines, the quasar fraction is lower. We show that this is not due to the difficulty of detecting lower-luminosity broad emission lines in a less luminous, but otherwise similar, quasar population. We discuss evidence which supports at least two probable physical causes for the drop in quasar fraction at low luminosity: (i) a gradual decrease in $\\theta_{\\rm trans}$ and/or a gradual increase in the fraction of lightly-reddened ($0 \\ltsimeq A_{V} \\ltsimeq 5$) lines-of-sight with decreasing quasar luminosity; and (ii) the emergence of a distinct second population of low luminosity radio sources which, like M87, lack a well-fed quasar nucleus and may well lack a thick obscuring torus. ", "introduction": "Orientation-based unified schemes for radio-loud quasars and powerful radio galaxies play a key role in our understanding of these objects. Ever since the conception of the idea that powerful radio galaxies are simply quasars with their jet axes oriented away from our line of sight, such that the nuclear continuum and broad-line regions are obscured by a dusty torus or warped disc (Scheuer 1987; Barthel 1989), evidence has been sought to prove that these unification schemes are correct. Reviews of unification schemes for active galactic nuclei have been presented by Antonucci (1993) and Urry \\& Padovani (1995). Although the majority of observations are consistent with them, some observations have been used to cast doubt on their viability over all ranges of radio luminosities and redshifts. The narrow line emission in radio sources is observed to be emitted largely from beyond the obscuring material (e.g. McCarthy, Spinrad \\& van Breugel 1995; Hes, Barthel \\& Fosbury 1996) and therefore is independent of the jet axis orientation. Hence in this model one would expect radio galaxies and quasars to have similar narrow line luminosities. The strong positive correlation between the extended radio luminosities and narrow line luminosities of radio sources (Baum \\& Heckman 1989; Rawlings et al. 1989; Willott et al. 1999) means that the samples of radio galaxies and quasars to be compared must be matched in extended radio luminosity. At low redshift ($z<0.8$), there have been claims that quasars are observed to have [OIII] line luminosities a factor of 5-10 greater than radio galaxies of similar radio luminosities (Baum \\& Heckman 1989; Jackson \\& Browne 1990; Lawrence 1991), although the [OII] luminosities of radio galaxies and quasars at these redshifts are indistinguishable (Browne \\& Jackson 1992; Hes et al. 1996). These differences between [OII] and [OIII] have been interpreted as being due to partial obscuration of [OIII] since its higher ionization potential means it is likely to be emitted from closer to the nucleus than [OII]. However, Jackson \\& Rawlings (1997) have investigated the [OIII] luminosities of $z>1$ radio galaxies and quasars and find their distributions indistinguishable. Using a combined 7C/3CRR dataset Willott et al. (1999) find that quasars have more luminous narrow lines than radio galaxies at intermediate radio luminosities [$26<\\log_{10} (L_{151}$ / W~Hz$^{-1}$sr$^{-1}) <27$], but they are similar at higher radio luminosities; this result leads to different slopes for the narrow-line versus radio luminosity correlation for quasars and radio galaxies. Barthel (1989) showed that in the redshift range $0.544^{\\circ}$ from our line-of-sight and all the quasars having $\\theta_{\\rm trans}<44^{\\circ}$, where $\\theta_{\\rm trans}$ is the half-opening angle of the obscuring torus; a value of $\\theta_{\\rm trans}=45^{\\circ}$ has been adopted by many on the basis of this paper. However, at higher redshifts in the 3CRR sample, the fraction of quasars increases giving $\\theta_{\\rm trans} \\approx 60^{\\circ}$ (e.g. Singal 1993). Note that because of the tight luminosity--redshift correlation inherent in a single flux-limited sample, this apparent correlation with redshift may be due instead to a correlation with radio luminosity. Singal (1996) used several large samples with differing radio flux-density limits to find that the quasar fraction in radio samples declines with decreasing radio flux-densities. However, due to the correlation between the optical and radio luminosities of steep-spectrum quasars (Serjeant et al. 1998; Willott et al. 1998a), radio-fainter samples will contain optically-fainter quasars. Therefore the quasar fraction in faint samples might be underestimated if only quasars brighter than a certain optical magnitude limit are identified, which may well be the case in some of these fainter samples used by Singal. Using the new 7C Redshift Survey, a low-frequency radio sample selected at a flux-density limit $25 \\times$ lower than the 3CRR sample with $ \\approx 90\\%$ spectroscopic redshift completeness, we have previously shown that the quasar fraction does not depend strongly upon radio luminosity or redshift for $1$10$^{12}$L$_{\\odot}$, see \\cite{Sanders} for a review). Recent studies based on near-IR, mid-IR (\\cite{Murphy}, \\cite{Soifer}, \\cite{Genzelb}), and X-ray (\\cite{Risaliti}) observations indicate that those galaxies are probably dominated by starbursts over their global infrared luminosity, even though weak AGNs may still be present in most of them. The possible detection of weak AGNs in all ultra-luminous galaxies is still plausible and necessitates powerful diagnostics. In this paper, we examine the mid-IR spectral properties observed in the central region of the prototypical Seyfert 2 galaxy NGC 1068 (\\cite{Lefloch}). Due to the proximity of this galaxy (14.4\\,Mpc, 1$''$=72pc), we can disentangle the AGN from the star formation regions found at 20$''$ from the nucleus. A comparative study of their respective mid-IR emission will be presented in order to point out, in a more general way, different methods for distinguishing an AGN within a dominant starburst environment. ", "conclusions": "" }, "0003/astro-ph0003241_arXiv.txt": { "abstract": "We report the discovery of a 695-Hz quasi-periodic oscillation (QPO) in data taken with the {\\it Rossi X-ray Timing Explorer} of the low-mass X-ray binary (LMXB) EXO 0748--676. This makes EXO 0748--676 the second dipping LMXB, after 4U 1915--05, that shows kHz QPOs. Comparison with other sources suggests that the QPO corresponds to the lower frequency peak of the kHz QPO pair often observed in other LMXBs. The QPO was found in the only observation done during an outburst of the source in early 1996. This observation is also the only one in which the $\\sim$1 Hz QPO recently found in EXO 0748--676 is not present. ", "introduction": "High frequency (kHz) quasi-periodic oscillations (QPOs) have been found in many neutron-star low-mass X-ray binaries (see van der Klis \\markcite{va2000}2000 for a recent review). They are observed in the 300--1300 Hz range, and are often found in pairs with a nearly constant frequency separation of $\\sim$250--350 Hz. In addition to kHz QPOs, some sources have shown slightly drifting oscillations in the 330--590 Hz range, during type-I X-ray bursts (Strohmayer, Swank, \\& Zhang \\markcite{stswzh98}1998). In this paper we present our search for both kHz QPOs and burst oscillations in the low-mass X-ray binary EXO 0748--676. This source shows periodic (P=3.82 hr) eclipses, irregular intensity dips, and type-I X-ray bursts (Parmar et al. \\markcite{pawhgi1986}1986). From the eclipse duration a source inclination of 75$^\\circ$ to 82$^\\circ$ was derived (Parmar et al. \\markcite{pawhgi1986}1986). Based on its bursting behavior (e.g. burst rate and peak flux vs. persistent flux; see Gottwald et al. \\markcite{gohapa1986}1986) EXO 0748--676 may be a member of the atoll class (Hasinger \\& van der Klis \\markcite{hava1989}1989) of the neutron-star low-mass X-ray binaries. Recently, a variable 0.58--2.44 Hz QPO was found by Homan et al. (1999). This QPO was found in all observations, except in the only observation during an outburst of the source (early 1996) observed with the {\\it Rossi X-ray Timing Explorer} (RXTE), and is probably caused by an orbiting structure in the accretion disk, which modulates the radiation of the central source (Jonker et al. \\markcite{jowiva1999}1999; Homan et al. \\markcite{hojowi1999}1999). ", "conclusions": "\\begin{deluxetable}{ccccc} \\tablecolumns{5} \\tablehead{ \\colhead{Energy band (keV)} & \\colhead{rms amp. (\\%)} & \\colhead{rms amp. (\\%)} & \\colhead{rms amp. (\\%)} & \\colhead{rms amp. (\\%)}} \\tablecaption{95\\% confidence upper limits for kHz QPOs in non-outburst power spectra, in two energy bands and for four different fixed FWHM. Ranges represent the lowest and highest upper limits measured among all observations.\\label{upper_tab}} \\startdata & FWHM=10 Hz & FWHM=20 & FWHM=50 & FWHM=100 \\\\ 1--60 & 6.4--13.9 & 9.4--14.9 & 10.7--17.6 & 11.6--21.7 \\\\ 6.6--18.7 & 6.3--13.6 & 6.6--16.3 & 8.2--19.5 & 9.2--23.0 \\\\ \\enddata \\end{deluxetable} The properties of the 695 Hz QPO are similar to those of the kHz QPOs in atoll sources; the QPO is relatively narrow (5--18 Hz) and has an rms amplitude of $\\sim$6.5\\% (1--60 keV, outside the dip). Since only a single peak is observed, we can not tell whether it corresponds to the lower or the upper peak of a kHz QPO pair. However, comparison with kHz QPOs in atoll sources (see van der Klis \\markcite{va2000}2000) suggests that the observed QPO is the lower QPO of a kHz pair, for the following reasons: (1) of the 11 kHz QPO pairs found in atoll sources, 8 have ranges of lower peak frequencies that include 695 Hz, which is the case for only 3 of the upper peaks. (2) The upper peaks in atoll sources have widths in the 50--200 Hz range, although occasionally peaks with widths of only 10 to 20 Hz have been observed. On the other hand, the 4--18 Hz width we find is much more common for lower peaks. (3) When comparing the energy dependence of the QPO with that of the two kHz peaks in 4U 1608--52, which have rather different energy dependencies (Berger et al. \\markcite{bevava1996}1996; M\\'endez et al. \\markcite{mevava1998}1998; M\\'endez et al. \\markcite{me2000}2000), we find that it was very similar (i.e. steep) to that of the lower peak (see Figure \\ref{energy_fig}). Hence three of the QPO properties hint towards the QPO being the lower peak. The properties of the QPO varied on a time scale of a few $10^3$ s, as can be seen from Table \\ref{selection_tab}. Comparing the first time selection with the second, one can see that a relatively small frequency change is accompanied by a factor 3 (2$\\sigma$) increase in width, and an almost 50\\% (2$\\sigma$) increase in fractional rms amplitude. The other source in which only a single kHz QPO has been observed is XTE J1723--376 (Marshall \\& Markwardt \\markcite{mama1999}1999). However, most sources in which kHz QPO pairs have been found, have at times also shown single kHz QPOs. The fact that EXO 0748--676 and XTE J1723--376 have only shown single kHz QPOs is therefore most likely a matter of a small amount of data and coincidence. With $i=75^\\circ-82^\\circ$ EXO 0748--676 is probably the source with the highest inclination angle of the $\\sim$20 sources that have shown kHz QPOs. Twin kHz QPOs were already found in 4U 1915-05 (Barret et al. \\markcite{baolbo1997}1997, \\markcite{baolbo2000}2000; Boirin et al. \\markcite{bobaol2000}2000), a source that also shows dips (but no eclipses, which for a similar mass ratio would imply a lower inclination than EXO 0748--676). The fact that kHz QPOs are found over a large range of inclinations means that the radiation modulated by the kHz QPO mechanism should to a large extent be isotropic. The kHz QPO was detected during the dip, but at a significance of only 1.8$\\sigma$. This means that with $\\sim$90\\% confidence we can say that either the source producing the kHz QPO was not fully covered by the dipping material, or that a considerable amount of the modulated radiation went through the dipping material unperturbed, indicating that it has a scattering optical depth of at most a few. Also, the fact that the rms amplitude changes only a little in the dip suggests that the kHz QPO and the bulk of the flux are produced at the same site. The outburst of EXO 0748--676 in early 1996 (see Fig. \\ref{asm_fig}) may have been a transition from the island state to the banana state, and back, as is common for atoll sources. In addition to the increase in count rate, there are several power spectral properties that seem to confirm this idea: (1) The strength of 0.1--1.0 Hz noise during the outburst was lower than in the non-outburst observations (see Homan et al. 1999). Most atoll sources show a decrease of the noise strength when they move from the island to the banana state (Hasinger \\& van der Klis \\markcite{hava1989}1989). (2) The $\\sim$1 Hz QPO was not observed during the only outburst observation. In 4U 1746--37, one of the other two sources were a similar $\\sim$1 Hz QPO was found, the QPO was observed only in the island state, and not in the banana state (Jonker et al. \\markcite{jovaho}2000). (3) Although there are a few exceptions, in most atoll sources kHz QPOs are found only in the lower banana state (van der Klis \\markcite{va2000}2000). We find the kHz QPO in the only observation where the $\\sim$1 Hz QPO was absent. The $\\sim$1 Hz QPO is thought to be due to obscuration of the central source by an orbiting structure in the accretion disk at a distance of $\\sim$1000 km from the central source (Jonker et al. \\markcite{jowiva1999}1999; Homan et al. \\markcite{hojowi1999}1999). It is interesting to see that in two of the sources were the $\\sim$1 Hz QPOs are found, they are not observed in the banana state, indicating a change in the accretion disk geometry (at least in the area where the $\\sim$1 Hz QPO is formed). This, together with the fact that in most atoll sources kHz QPO are only found in the banana state, suggests that changes in the accretion disk geometry (at $\\sim$1000 km from the central source) may affect the production of kHz QPOs close to the central source." }, "0003/astro-ph0003131_arXiv.txt": { "abstract": "The classical [CO] index, i.e. the strength of the $\\Delta v$=2 CO absorption bands starting at 2.29 \\MIC, is sometimes used to constrain the maximum age of star formation events in galaxies. In this paper we critically analyze \\begin{description} \\item[{\\it i)\\hglue3pt}] theoretical models which could predict either a factor of $>$2 drop or a pronounced increase of [CO] at ages older than 100 Myr, depending on the evolutionary tracks one adopts (see Fig.~\\ref{evolco}). \\item [{\\it ii)}] observational data for young clusters in the LMC which do not show any strong relationship between the CO index and cluster age (see Fig.~\\ref{figobs}). \\end{description} The above scenario indicates that the value of [CO] does not provide a reliable tool for estimating the age of stellar populations older than $\\sim$10 Myr, i.e. after the first red supergiants have been formed. The contradictory results of theoretical models reflect problems in treating the evolution along the Asymptotic Giant Branch (AGB). In particular, those evolutionary synthesis models using stellar tracks which do not include the thermal pulsing AGB phase produce too weak CO features at 100--1000 Myr, i.e. in the range of ages when the near infrared emission is dominated by thermal pulsing AGB stars. ", "introduction": "An ideal instantaneous burst of star formation generates a so--called Simple Stellar Population (SSP), that is a stellar system which is coeval and initially chemically homogeneous (see Renzini \\& Buzzoni \\cite{buzzoni86}). The integrated near IR luminosity of a SSP is dominated by red stars since its very early stage of evolution ($\\simeq$10 Myr), when massive stars ($<$40~M$_{\\odot}$) evolve as red supergiants. When the stellar system gets older ($\\ga$100 Myr) intermediate mass giants evolving along the AGB and, after a few Gyr, low mass giants near the tip of the Red Giant Branch (RGB) dominate the integrated IR and bolometric luminosities (e.g. Renzini \\& Buzzoni \\cite{buzzoni86}, Chiosi et al. \\cite{chiosi86}). The time evolution of the observable parameters related to a SSP, such as e.g. photometric colours and spectral indices, provides the basic ingredient for constructing evolutionary models of star forming galaxies. Among the photometric and spectroscopic indices used to study the red stars of a SSP, the CO index has attracted quite some attention as a potential tool to trace red supergiants, i.e. young stellar systems. This idea primarily derives from the fact that field stars of similar spectral types show different CO indices depending on their spectral class, the strongest features being found in supergiants (see e.g. Fig. 4 of Kleinmann \\& Hall 1986, hereafter \\cite{KH86}). Several attempts of predicting the evolution of the CO index of a SSP appeared in the literature and were applied to the interpretation of IR spectral observations of starburst galaxies (e.g. Doyon et al. \\cite{doyon94}, Shier et al. \\cite{shier96}, Goldader et al. \\cite{goldader97}, Mayya \\cite{mayya97}, Leitherer et al. \\cite{leitherer99}). Most of the models are restricted to solar metallicities and predict a pronounced maximum at $\\simeq$10 Myr followed by a quite rapid and steady decline. The CO index drops by almost a factor of 3 at $\\simeq$100 Myr and, noticeably, reaches values much lower than those observed in old ($\\ga$10 Gyr) Galactic globular clusters and spheroidal galaxies of quasi--solar metallicities. The few models at sub--solar metallicities predict a similar time evolution with shallower CO features at all epochs. Taken at face value, these models would imply that star forming galaxies with prominent CO absorption features must be dominated by a young ($<$100 Myr) star formation event, while more mature, but still relatively young systems of a few $\\times\\;$100 Myr should be characterized by quite weak CO absorption features. In other words, the CO index could provide a powerful tool to constrain the age of the major star formation event of galaxies. This paper is a critical re--analysis of the time evolution of the CO index in simple SSPs and star forming galaxies. In Sect.~\\ref{CO1} we describe the, sometimes confusing, definition of CO index and discuss its relationship with stellar parameters. In Sect.~\\ref{model_evol} we present theoretical curves based on different stellar evolutionary models and briefly discuss the possible reasons for their very different behaviours. In Sect.~\\ref{data} we compare the predicted evolution of [CO] with measured parameters of template stellar clusters in the Magellanic Clouds, old globular clusters in the Galaxy, normal and starburst galaxies. In Sect.~\\ref{conclusions} we draw our conclusions. ", "conclusions": "" }, "0003/astro-ph0003307_arXiv.txt": { "abstract": "The extremely young cluster IC 348 has been monitored in the Cousins I band with a 0.6 m telescope at Wesleyan's Van Vleck Observatory. Photometry of 150 stars was obtained on 76 images taken on 27 separate nights during the period December, 1998, through March, 1999. As expected, spectral characteristics largely determine the nature of a star's variability in this cluster. None of the stars with H$\\alpha$ in absorption were found to be variables. On the other hand, all 16 stars identified as CTTS by their H$\\alpha$ emission equivalent widths and the majority of the 49 WTTS in the part of the cluster we monitored showed evidence of variability. Nineteen stars were found to be periodic, with periods ranging from 2.24 to 16.2 days and masses ranging from 0.35 to 1.1 M$_\\odot$. Seventeen of these are WTTS and the other 2 are of unknown spectral class. The period distribution is remarkably similar to what is found in the Orion Nebula Cluster for stars in the same mass range. Namely, it is bimodal with peaks at 2-3 days and 7-8 days, although there are not enough periods known to define these features significantly by the IC 348 data alone. The three fastest rotators are also the three most massive stars in the periodic sample. It is striking that none of the known CTTS were found to be periodic even though they are more highly variable than the WTTS in the cluster. This supports the canonical view that WTTS variability is primarily caused by the rotation of a surface with large, cool spots whose pattern is often stable for many rotation periods, while CTTS variability has an additional component caused by accretion hot spots which typically come and go on shorter timescales. Stars with significant infrared excess emission in this sample do tend to be CTTS, while the WTTS (including periodic ones, with one possible exception) show no infrared excess and, therefore, no evidence of disks. Among the CTTS, neither H$\\alpha$ emission equivalent width nor infrared excess emission shows any correlation with degree of variability. ", "introduction": "It has been known since the pioneering work of Joy (1945) that T Tauri stars are variable stars. They are now recognized to be pre-main sequence (PMS) stars (e.g. Bertout 1989) and their variations are thought to arise from a changing pattern of hot accretion zones and cool spots on their surfaces (Herbst et al. 1994). Classical T Tauri stars (CTTS), which have stronger emission lines and infrared excesses as well as a veiling continuum overlying their photospheric spectrum, typically vary in irregular fashion on timescales of hours to days with amplitudes of a few hudredths to several magnitudes in V. The principal source of the modulations is thought to be unsteady accretion, which causes the veiling continuum to wax and wane. Rotational modulation of the pattern of bright and dark spots on the CTTS surface is undoubtedly present, but often masked by changes in the spot pattern on timescales comparable to or shorter than a typical rotation period of 2 to 20 days. By contrast, the weak T Tauri stars (WTTS) are usually periodic variables with V magnitude amplitudes of less than 0.75 mag which can be attributed to the rotation of a star with very large, cool spots on its surface. These dark spots may be associated with the \"footprints'' in the photosphere of the strong dipole field, which is a central feature of the magnetospheric accretion model of TTS (e,g, Mahdavi and Kenyon 1998; Hartmann 1998). Earlier-type analogs of the CTTS exist, namely, the Herbig Ae/Be stars (Herbig 1960; Strom et al. 1972; Hillenbrand et al. 1992) and a group of early K through F stars which have no widely used moniker. Most of the Herbig Ae/Be stars and the F and G-type TTS are variables, with ranges and timescales similar to the CTTS. As a group, these variables have come to be known as UXors, after the proto-type UX Ori (Herbst, 1994; Herbst \\& Shevchenko 1998; Natta et al. 1999). No periodicity has been established with definiteness for any of these stars (Herbst \\& Shevchenko 1998). A commonly held view is that the variations result from occultations by proto-planets or proto-comets or other dust concentrations in circumstellar disks (e,g, Grinin 1994) but Herbst \\& Shevchenko (1998) propose that variable accretion might, instead, be the cause. The most extreme examples of PMS variability, the FUors (Herbig, 1977), can change brightness by more than 5 magnitudes on timescales as short as a few weeks. Unsteady accretion in a luminous disk is widely believed to account for their behavior (Hartmann \\& Kenyon 1996), although not universally (e.g. Petrov \\& Herbig 1992) Our ideas about the variations of pre-main sequence stars have largely been developed from studies of individual objects that are rarely considered within the context of a physically associated group. This is partly because the closest and brightest examples tend to be members of loose associations, particulary Taurus/Auriga, which are too spread out on the sky to be studied {\\it en masse}. Early cluster studies concentrated on the Orion region where Parenago (1954) discovered hundreds of variable stars and Haro, Chavira \\& Mendoza (1960) studied the flare stars. But photographic plates are poor detectors for work in Orion because their blue sensitivity restricts them to spectral regions where the nebular light is a serious pollutant and the late-type stars are intrinsically faint. Unlike the case for globular clusters, where photographic work on variable stars prospered during the era of photographic plates (e.g. Hogg 1972), variability studies of extremely young clusters languished during the 1960's - 1980's. The situation has now improved with the widespread availability of CCD detectors on small and medium-sized telescopes. By working in the far red spectral region, one can avoid the principal nebular emission lines and detect stars in Orion almost to the H-burning limit with a 0.6m telescope. Variability studies, especially those aimed at irregular variables, require lengthy observing runs and cannot usually be done on the larger telescopes, for which there is too much shared demand. Programs at Van Vleck Observatory on the campus of Wesleyan University directed at the Orion Nebula Cluster and, more recently, NGC 2264, have been underway since the early 1990's (Mandel \\& Herbst 1991; Attridge \\& Herbst 1992, Eaton, Herbst \\& Hillenbrand 1995; Choi \\& Herbst 1996; Herbst et al. 2000). Other recent examples of extremely young cluster monitoring programs include those by Adams, Walter \\& Wolk (1998) and Stassun et al. (1999), both directed at the Orion region, and Makidon et al. (1996) directed at NGC 2264. The principal focus of the extremely young cluster monitoring programs mentioned above has been the discovery of rotation periods for PMS stars. These data have been used to constrain models of the evolution of stellar angular momentum from the PMS to the main sequence phase. However, other interesting questions can be addressed by consideration of variability more broadly. It should be possible, for example, to test the notion that CTTS are more active and less regular variables than WTTS because of the role that accretion plays. Variability studies could also make a contribution to the study of the clusters themselves. For example, detection of irregular (or periodic) variations in a star within the cluster field is a good indicator of cluster membership which can be used, in conjunction with other such indicators as kinematics, location on the HR diagram, emission lines, etc. to help isolate true cluster members from field stars. Also, the optical variability of PMS stars is almost always neglected (for lack of knowledge) when HR diagrams are created. Magnitudes and colors are usually based on a small number of measures and could be significantly different for a particular star if they had been measured on different nights. To quantify the errors associated with this practice and to obtain the most applicable data for individual stars when possible, a large number of observations are required. For a variety of reasons, it appears that IC 348 is a young cluster uniquely well-suited to a general variability study of its PMS population. It is nearby, extremely young and relatively free of the nebulosity which complicates photometric studies in, for example, the ONC. The distance is somewhat in doubt since the most straightforward and precise method - averaging parallaxes of cluster members measured by Hippacrcos - leads to a result (260 $\\pm$ 25 pc; Scholz et al. 1999) which is marginally closer than traditional main sequence fitting methods (316 pc; Herbig 1998, hereinafter H98). The uncertainty in the distance translates to an uncertainty in the age of the PMS stars because it is determined by fitting their luminosities to models. At the larger, photometric distance, the median age of the PMS stars is about 1.3 million years, whereas at the Hipparcos distance, the stars are closer to 3 million years old, according to the models of D'Antona \\& Mazzitelli (1994). Since the typical age of a PMS star in the ONC is about 0.8 My according to the same models (Hillenbrand 1997), IC 348 is between 1.5 and 4 times older than the ONC. Of course, there is an age range in both clusters, so individual stellar ages determined by location on the HR diagram overlap with each other. The cluster has about 60 M$_{\\odot}$ in stars with an average mass of about 0.5 M$_{\\odot}$ within its \"core'' radius of 4\\arcmin (H98). This means that there are more than one hundred late-type members which can be monitored for variability within a 10$\\arcmin$\\ field. The earliest type star (BD+31$\\deg$ 632) is of spectral class B5, so it illuminates only a faint reflection nebula, not a bright emission nebula. The only disadvantage for photometry is that the cluster is rather heavily embedded in the local dust cloud out of which it presumably formed, since BD+31$\\deg$ 632 is insufficiently luminous to have cleared the region of grains. This means that extinction is rather high and variable from star to star. It restricts monitoring programs with small telescopes to a far red (e.g. Cousins I) band. A significant attraction of IC 348 for a variability study is that it has been the subject of several recent, detailed investigations by a variety of techniques, which have characterized the visible stellar population in important ways. The early history of this work is recounted in most of the papers below; here we mention only the contributions of the last five years, beginning with the near-infrared studies of Lada \\& Lada (1995) and Luhman et al. (1998). These have provided a deep stellar census and spectral types for most of the stars in the core region of the cluster. Information on the presence or absence of disks comes from the near-IR colors as well as veiling estimates in the IR spectroscopy and Br$\\alpha$ and H$\\alpha$ emission line strengths. A detailed optical spectroscopic and photometric study by Herbig (H98) provides the data to assess extinction and place the stars on an HR diagram, from which masses and ages can be inferred. This is supplemented by the photometric study of Trullols \\& Jordi (1997; TJ). An H$\\alpha$ survey reported in H98 provides information on whether stars are CTTS, WTTS or show absorption lines. A proper motion study by Scholz et al. (1999) has isolated the motion of the cluster from that of nearby and distant field stars and provided membership probabilities for some stars in the core. An X-ray study of the cluster by Preibisch et al. (1996) provides ROSAT data on many core members. Finally, Duchene, Bouvier \\& Simon (1999) have searched for optical binaries in the cluster using adaptive optics techniques. A variability study would seem to be a valuable addition to this outburst of activity on an important, nearby cluster. ", "conclusions": "" }, "0003/astro-ph0003021_arXiv.txt": { "abstract": " ", "introduction": "BL Lacertae objects are extreme extragalactic sources characterized by the emission of strong and rapidly variable nonthermal radiation over the entire electromagnetic spectrum. Synchrotron emission followed by inverse Compton scattering in a relativistic beaming scenario is generally thought to be the mechanism powering these objects (e.g. Kollgaard 1994., Urry \\& Padovani 1995). BL Lacs can be divided into different subclasses depending on their Spectral Energy Distribution (SED), namely LBL for objects with the synchrotron emission peaking at $\\nu_{peak}\\approx 10^{13-14} Hz $, intermediate objects ($\\nu_{peak}\\approx 10^{15-16} Hz $) and HBL or high energy peaked BL Lacs with $\\nu_{peak}\\approx 10^{17-18}Hz$ (Padovani \\& Giommi 1995). The wide X-ray band pass of the BeppoSAX satellite (Boella et al. 1997) is well suited for the detailed spectral study of all types of BL Lacs. In fact, direct measurements of the Compton part of the spectrum have been obtained for a number of LBLs (e.g. Padovani et al. 1999), and the very variable tail of the Synchrotron component has been studied in several HBLs (e.g. Pian et al. 1998, Wolter et al. 1998, Giommi, Padovani \\& Perlman 1999, Chiappetti et al. 1999). In the case of the two intermediate BL Lacs S5~0716+714 and ON~231 BeppoSAX for the first time was able to detect both spectral components within a single instrument (Giommi et al. 1999, Tagliaferri et al. 1999). The BeppoSAX archive at the Science Data Center (SDC, Giommi \\& Fiore 1998) presently includes over 100 observations of 56 distinct BL Lacs, about half of which are already publicly available. We have started a project to construct the SED of a large number of all types of BL Lacs by combining a) public BeppoSAX data (0.1-200 keV); b) simultaneous optical and radio data when these are available from monitoring campaigns, or from the University of Michigan Radio Astronomy Observatory (UMRAO) on-line data base (Aller et al. 1999); and c) non-simultaneous photometric data form NED. Here we present the first results of this project. \\begin{figure}[ht] \\centerline{\\psfig{file=fig1.ps, width=13.0cm}} \\caption[]{Spectral energy distribution of four LBL and intermediate BL Lacs.} \\end{figure} ", "conclusions": "" }, "0003/astro-ph0003217_arXiv.txt": { "abstract": "The core of the metal poor Galactic Globular Cluster M92 (NGC 6341) has been observed with WFPC2 on the {\\it Hubble Space Telescope} through visual, blue and mid-UV filters in a program devoted to study the evolved stellar population in a selected sample of Galactic Globular Clusters. In the UV $(m_{255}, m_{255}-U)$ color magnitude diagram we have discovered a faint `UV-dominant' object. This star lies within the error box of a Low Luminosity Globular Cluster X-ray source (LLGCX) recently found in the core of M92. The properties of the UV star discovered in M92 are very similar to those of other UV stars found in the core of some clusters (M13, 47 Tuc, M80, etc)--- all of them are brighter in the UV than in the visible and are located in the vicinity of a LLGCX. We suggest that these stars are a new sub-class of cataclysmic variables. ", "introduction": "\\label{sec:intro} Despite their rarity, it is the exotic creatures that attract the crowds at the zoo; similarly the exotic objects in the stellar zoo attract our attention. Unusual environments often lead to relatively large populations of the exotic. So it is with the cores of the Galactic Globular Cluster (GGCs), which have long been thought to harbor a variety of exotic objects---blue stragglers, low mass X-ray binaries, cataclysmic variables, millisecond pulsars, etc. Most of these objects are thought to result from various kinds of binary systems whose nature and even existence can be strongly affected by dynamics in the dense cluster cores. When a binary system contains a compact object (like a neutron star or white dwarf) and a close enough secondary, mass transfer can take place. The streaming gas, its impact on the compact object, or the presense of an accretion disk can give such systems observational signatures which make them stand out above ordinary cluster stars. These signatures might include X-ray emission, significant radiation in the ultraviolet (UV), emission lines, or rapid time variations. The first evidence for such objects in globular clusters was the discovery of X-ray sources. One population of X-ray sources with $ L_{X} > 10^{34.5} \\ers $ (the so-called Low Mass X-ray Binaries, LMXB) are thought to be binary systems with an accreting neutron star because of their X-ray bursts. LMXBs are very overabundant (a factor 100) in GGCs with respect to the field, presumably because the high stellar density has led to many capture binaries. Given the existence of neutron star systems in GGCs one might expect to find many more analogous systems involving white dwarfs (WDs). In the field, binary systems in which a WD is accreting material from a late type dwarf, i.e., a main sequence or subgiant star, are observed cataclysmic variables (CVs). CVs are well-studied objects in the field, where they are thought to form by the evolution of primordial binaries. They come in many varieties depending on stellar masses, mass transfer rates, magnetic field strength, etc. In GGCs one can expect even more variety because CVs located in dense clusters could have been created by dynamical processes (Hut \\& Verbunt, 1983, Bailyn 1995), while the CVs in low-density clusters result from primordial binary systems (Verbunt \\& Meylan 1988). Numerical simulations (e.g., DiStefano \\& Rappaport 1994) suggest that $> 100$ white dwarf binaries might be found in massive clusters like 47~Tuc and $\\omega$~Cen, and several 10's in more typical clusters. Despite the expectation of large numbers of CV like stars, searches have turned up only a relatively small number. Of course, part of the problem arises because of the difficulty of search for rather faint objects in crowded globular cluster fields. However, exploiting the high resolution of {\\it HST}, it has become possible to search GGC centers for several of the anticipated CV signatures. Still the number of candidates is small: \\begin{itemize} \\item More than 30 low luminosity X-ray sources with $L_x < 10^{34.5} \\ers$ (hereafter LLGCXs) have been discovered in 19 GGCs (Johnston \\& Verbunt 1994). Despite their relatively large numbers there is no consensus model for LLGCXs (see Verbunt et al. 1994) and Hasinger, Johnston \\& Verbunt 1996). The fainter LLGCXs ($L_x < 10^{32} \\ers$) might well be associated with CVs (van Paradijs (1983), Hertz \\& Grindlay, 1983). \\item There are three objects connected with conventionally detected CVs: a dwarf nova in M5 (Margon, Downes \\& Gunn 1981); HST UV detections of optical counterparts to a dwarf nova in 47 Tuc (Paresce \\& DeMarchi 1994) and possibly the historical nova in M80 (Shara \\& Drissen 1995) \\item Using \\HST, H$\\alpha$ emission has been observed from three objects in NGC 6397 (Cool et al. 1995), and two objects in NGC 6752 (Bailyn et al. 1996). \\item Also using \\HST, a number of candidate CVs have been selected on the basis of UV excess and variability: the Einstein dim source in 47 Tuc (Paresce, De Marchi \\& Ferraro 1992) and few CVs candidates in NGC 6624 (Sosin \\& Cool 1995). \\end{itemize} We are involved in two long-term HST projects to study in detail the evolved populations in a sample of GGCs, at different wavelengths ranging from the UV to the near IR. Although we were not specifically hunting for CVs, we have used the UV exposures to search for exotic objects in the core of GGCs. This search has been very fruitful: in our data-base (9 clusters) we have found in {\\it all GGCs properly observed (with exposures deep enough and in the right UV bands) there is at least one faint object with a strong UV excess with respect to the main stellar population of the cluster)}. These stars are brighter in the UV than in the visible, so we will refer to them as UV-dominant (UVD). We wish to carefully distinguish between these objects and objects which are called `UV-excess' objects on the basis of their colors in the visible or perhaps near UV. Previously we have reported on three UVD stars in the GGC M13 (Ferraro et al. 1997). Two of these objects have been found to lie within the error boxes of LLGCXs (Fox et al. 1996), and we argue that they are excellent CV candidates. Here we report on the discovery of another faint UVD star in the core of M92, and we suggest that this star is physically connected to the X-ray emission detected in the cluster. ", "conclusions": "We have now identified three UV dominant objects which appear to be associated with X-ray sources in the GGCs M13 and M92. We argue that these are generically CVs, i.e., white dwarf/main sequence binaries with some mass exchange. They share many properties with some field CVs---X-ray luminosity, UV colors, absolute visual magnitudes---although they tend to lie at the extremes of the distributions. The relatively high optical, UV, \\& X-ray luminosities are consistent with the notion that current surveys do not reach deep enough to detect most of the CVs in GGCs. On the other hand, cluster CVs are older and live in a dramatically different environment from their sisters in the field. Thus, it seems reasonable that cluster objects might be a new class of CVs with properties which slightly differ from those in the field. The CVs found in NGC~6397 and NGC~6752 differ significantly from our objects. Some of these have little or no UV excess. This should not be surprising. The searches in NGC~6397 and NGC~6752 relied on H$\\alpha$ and $R$ band, whereas we used the UV. There is considerable variety in the properties of field CVs and no reason to suspect less variety in GGC CVs. Different search techniques operating at the margin of detectability will certainly turn up different kinds of objects. The UVD objects in GGCs probably come in several varieties. (Our GGC projects do not seem to come up with simple answers.) Edmonds \\etal\\ (1999) have shown that one is a hot high gravity object, arguablly a He-WD. Still we suspect that most of these will turn out to be CVs, and that many UVD objects with no X-radiation detected to date will show up as LLGCXs in more sensitive X-ray surveys. This suspicion is fueled by the belief that a significant population of generic CVs must be present in GGCs. We have yet to make an observation directly showing the hot diffuse gas which would be the definitive evidence that our UVD objects are CVs. HST STIS spectra could give such evidence. It would be extremely valuable to develop a technique to identify GGC CVs using UV photometry. Cluster cores are very congested in the red and H$\\alpha$/$R$ band searches will obviously be incomplete because of the interference by bright red giants (see Figure~2 of Bailyn \\etal\\ 1996). On the other hand the core of even the densest clusters are relatively open in the UV (see Figure~1 of Ferraro \\etal 1999b)" }, "0003/astro-ph0003484_arXiv.txt": { "abstract": "We present a new calculation of the propagation of protons with energies above $10^{19}$ eV over distances of up to several hundred Mpc. The calculation is based on a Monte Carlo approach using the event generator SOPHIA for the simulation of hadronic nucleon-photon interactions and a realistic integration of the particle trajectories in a random extragalactic magnetic field. Accounting for the proton scattering in the magnetic field affects noticeably the nucleon energy as a function of the distance to their source and allows us to give realistic predictions on arrival energy, time delay, and arrival angle distributions and correlations as well as secondary particle production spectra. ", "introduction": "The world statistics of ultra high energy cosmic ray (UHECR) events of energy above 10$^{20}$ eV has now grown to 20 events~\\cite{SLC,AGASA}. It is very difficult to accelerate particles to such high energies in astrophysical shocks, the process thought to be responsible for the majority of the galactic cosmic rays~\\cite{Hillas84}. This has led to a large number of production models, many of them based on exotic particle physics scenarios~\\cite{BS00}. The gyroradii of 10$^{20}$ eV protons are significantly larger than our own Galaxy and this suggests an extragalactic origin~\\cite{Cocconi} for any astrophysical scenario ($r_g = 100 {\\rm kpc} \\times (E/10^{20} {\\rm eV}) \\times (1 \\mu{\\rm G}/B)$ with $E$ and $B$ being the proton energy and the magnetic field strength, respectively). The large distances between potential UHECR sources and Earth leads to another set of problems first pointed out independently by Greisen and by Zatsepin \\& Kuzmin, now widely known as the GZK effect~\\cite{GZK}. UHECR protons interact with photons of the microwave background radiation and lose their energy relatively rapidly during propagation over distances of tens of megaparsecs. This should result in a cutoff in the cosmic ray spectrum at an energy just below 10$^{20}$ eV. Many different calculations~\\cite{Hill85,BG88,YT93,RB93,AhCronin,PJ96,Lee98}, performed using various techniques, of the modification of the cosmic ray spectrum due to propagation have been published since the original suggestion. As a result, the general features of the cosmic ray spectrum after propagation are well established. Differences between the various approaches are, however, significant and the accuracy achieved is not sufficient for the interpretation of the existing experimental data, and more accurate calculations are needed for the expected significant increase of the experimental statistics~\\cite{HiRes,Auger,TelArr,OWL}. Previous calculations can be divided into two classes dealing mainly with: (a) the energy loss processes~\\cite{Hill85,BG88,YT93,RB93,AhCronin,PJ96,Lee98}, and (b) the deflection and scattering of protons in the extragalactic magnetic field~\\cite{Lampard97,Clay98a,Acht99}. The first group of calculations shows that small differences in the realization of the proton energy loss processes generate observable differences in the predicted spectra at Earth. Such calculations, however, cannot establish an accurate relation between the distance of a potential source and the modification of the proton spectrum emitted by this source because the influence of the extragalactic magnetic field is neglected. Among the calculations of the second kind, Refs.~\\cite{WME96,Lampard97,Clay98a} do not consider the proton energy losses in a satisfactory way, and Refs.~\\cite{Medina97,Sigl97a,Sigl99a} mostly discuss their results in a specific context. Only Achterberg {\\em et al.} \\cite{Acht99,Achterror} give a detailed discussion of the fundamental aspects of UHECR propagation in extragalactic magnetic fields, which we are interested in here. We present here calculations performed with the photoproduction event generator SOPHIA~\\cite{SOPHIA}, which is proven to reproduce well the cross section and final state composition in nucleon-photon interactions for energies from the particle production threshold up to hundreds of GeV in the center-of-mass system. We also account for all other energy loss processes of UHECR nucleons, and calculate the proton deflection in the extragalactic magnetic field in three dimensions. We restrict ourselves to proton injection energies up to 10$^{22}$ eV, and consider (with few exceptions) proton propagation for source distances less than 200 Mpc. The calculations are carried out using a Monte Carlo technique, and we propagate individual protons injected as either a mono-energetic beam, or with energies sampled from a fixed source energy spectrum. This approach has the advantage of representing fluctuations in the proton energy losses very well, thereby giving us a good handle on the correlations between energy loss, time of flight and angular deviation of the flight direction. As we will show, these important UHECR characteristics are deeply interconnected. For a given source distance, there is a strong correlation between the amount of energy lost, the time delay, and the scattering angle. Our calculations are thus mainly relevant to scenarios of UHECR acceleration at astrophysical shocks, for which 10$^{22}$ eV is a very generous upper energy limit. With this paper we wish to establish limits for the distance of potential UHECR proton sources as a function of proton energy and the average strength of the extragalactic magnetic field. We also study the angular distribution of UHECR with respect to the source direction (arrival angle) and the time delays after propagation over different distances. In addition, the neutrino fluxes produced during the propagation are presented. The article is organized as follows. We describe the propagation method, including the relevant features of the event generator SOPHIA, in Section 2. Section 3 gives some interesting results on the propagation of mono-energetic proton beams, and compares our results with other work. Section 4 analyzes the formation and development of the primary and secondary particle spectra for protons injected with a power law spectrum. In section 5 we discuss the results, present our conclusions, and make suggestions for future work. ", "conclusions": "The Monte Carlo propagation of ultra high energy protons in a random extragalactic magnetic field has obvious advantages over other approaches to calculations of proton propagation in the cosmologically nearby Universe. To start with, this approach takes fully into account fluctuations in the positions of proton interactions, and thus also in the proton energy losses and production of secondary particle fluxes. It also naturally generates the correlations between the proton's arrival energy, its time delay, and its angular deviation from the source direction. We have also shown that mathematical approaches which use a diffusion description of magnetic scattering, although superior in computational speed, can lead to significant systematic errors for propagation distances smaller then $\\sim 100$ Mpc. These features of the calculation become extremely valuable when applied to specific models of UHECR acceleration, especially models that involve a relatively short (compared to light travel time and proton time delay) active phase of the source. An extreme example for such a model is the GRB model for UHECR acceleration. However, other models involving interacting galaxies or radio galaxies of specific morphology could also be affected, especially if embedded in regions of high (random) magnetic field. At energies that allow protons to photoproduce, namely above 10$^{20}$ eV, the energy degradation is extremely rapid. This is not very surprising because of the very short photoproduction interaction length at energies corresponding to the maximum cross section -- i.e. $\\lambda_{\\rm ph}$ below 4 Mpc for energies between 4$\\times$10$^{20}$ eV and 10$^{21}$ eV. This energy range is very relevant, as it is just above the highest energy particles detected by the Fly's Eye and AGASA arrays~\\cite{FEHi,AGASA}. A large part of this rapid energy dissipation in our calculation is due to the correct implementation of the fluctuations in photoproduction interactions in SOPHIA. A good example for the size of the fluctuations is the proton energy distribution after propagation over 10 Mpc shown in Fig.~\\ref{power}, which covers more than one and a half orders of magnitude. This is an extreme case. However, every particle injected with an energy well above the photoproduction threshold would very rapidly result in a distribution extending down to the threshold, within the first 10 Mpc. This rapid energy dissipation creates additional problems for models of cosmic ray acceleration at astrophysical shocks. Apart from the difficult question of the maximum acceleration energy, such models require that a significant fraction (0.01 to 0.1) of their source luminosity contributes to the UHECR flux. The rapid energy dissipation increases the energy requirements in terms of total luminosity and severely limits the source distance. Because of magnetic scattering, such limits could also be set for particles injected with energy below the photoproduction threshold. \\begin{figure}[tb] % \\centerline{\\epsfig{file=p1_f9c.ps,width=85mm}} \\vspace*{10pt} \\caption{Proton 50\\% horizon as a function of injection energy for average random magnetic fields of 0.1 (dashed histogram), 1 (solid histogram), and 10 (dotted histogram) $n$G. See text for definition. The solid line is the total energy loss length from Fig.~\\protect\\ref{fig1}, shown here for comparison. } \\label{horiz} \\end{figure} Fig.~\\ref{horiz} shows the 50\\% horizon for UHECR sources as a function of source particle energy for $\\langle B \\rangle$ values of 0.1, 1 and 10 $n$G. The 50\\% horizon $R_{50}$ is defined here as the light propagation distance to the source at which 1/$e$ of all injected protons have retained 50\\% or more of their energy, i.e.\\ $R_{50}$ is achieved when \\begin{equation} \\int_{\\frac{E_0}{2}}^{E_0} {d N \\over d E} dE = N_0 \\exp(-1), \\end{equation} where $N_0$ is the number of particles injected with energy $E_0$. To start with, $R_{50}$ is small at any energy, and demonstrates the resonant nature of the photoproduction cross section. At E = 10$^{20}$ eV $R_{50}$ is about 100 Mpc, while at 2$\\times$10$^{20}$ eV it decreases to 20 Mpc and becomes smaller than 10 Mpc for energies above 3$\\times$10$^{20}$ eV. For injection energies above 10$^{20}$ eV the horizon energy dependence is similar to that of the energy loss distance shown in Fig.~\\ref{fig1}. These protons are not affected much by the magnetic field since their scattering angles are small, but suffer mainly from energy degradation due to $p\\gamma$ encounters. Below 10$^{20}$ eV the picture changes. The scattering in the magnetic field increases the propagation time and thus causes additional energy loss and an increase of the ratio $x_{\\rm loss}/R_{50}$. Stronger magnetic fields create delays, that could be longer than the light propagation time from the source and reverse the trend -- the horizon starts decreasing below $\\sim 6\\times 10^{19}$~eV and is restricted to 75 Mpc at 10$^{19}$ eV. Since the average time delay is inversely proportional to $E^2$, the decrease of $R_{50}$ is expected to become more drastic at lower energy. One consequence of the strong energy dependence of $R_{50}$ is, for example, that our attempts to correlate the arrival directions of UHECR with different types of astrophysical objects should use only objects within the particle horizon depending on the magnetic fields strength in different regions of the Universe. Independently of the magnetic field value, however, the horizon defined above is much smaller than the conventional numbers of 50 or 100~Mpc for the highest energy cosmic ray events. There are many relevant astrophysical problems which can be studied with the approach described in this paper. We plan to use the code for proton propagation in regular magnetic fields associated with large scale structures (local supercluster, supergalactic plane). The regular fields, especially if they reach the observationally allowed limits of 0.03~$\\mu$G and even 0.1~$\\mu$G, could change the propagation patterns for 10$^{19}$~eV cosmic ray protons and alter the horizon values shown in Fig.~\\ref{horiz}. We also plan to set limits on models of slow UHECR acceleration on shocks of very large dimensions and to look for possibilities of ultra-high energy $\\gamma$--ray halos around the sources and along the tracks of the UHECR protons." }, "0003/astro-ph0003167_arXiv.txt": { "abstract": "We present the results of new 3D TIGER spectroscopic observations and archived HST/WFPC2 and NICMOS images of the central region of M\\,104. The \\NII+\\Ha\\ images reveal the presence of a nuclear spiral structure, and the gaseous kinematics in the central arcsecond shows evidence for kinematical decoupling of the central peak. A straight nuclear dust lane, with a weak symmetric counterpart, is seen in the $V-I$ and $V-H$ colour maps. These results hint for the presence of a strong nuclear bar, that would be located inside the inner Linblad resonance of the large-scale bar discussed by Emsellem (\\cite{Ems95}). ", "introduction": "This is the third of a series of papers dealing with optical observations of the Sombrero galaxy (M~104, NGC~4594). We report here the results of 3D spectroscopy of the nuclear region of M\\,104 with the TIGER instrument, in the 6750/460~\\AA\\ spectral domain, which includes the \\NIIww, \\Ha\\ and \\SIIww\\ emission lines. This 3D dataset is used in combination with archived images acquired with the Wide Field and Planetary Camera 2 (WFPC2) and the Near Infrared Camera and Multi-Object Spectrometer (NICMOS) on board the Hubble Space Telescope (HST), to study the nuclear regions of this galaxy. This paper is organised as follows: Sect.~\\ref{SecObs} presents the observations and the data reduction; Sect.~\\ref{SecRes} shows the results; and Sect.~\\ref{SecDis} contains a brief discussion of the nature of the observed nuclear structures. Throughout this paper, we will use a distance to M~104 of 8.8~Mpc (Ciardullo \\etal\\ \\cite{Ciar93}), yielding an intrinsic scale of $\\sim$43~pc.arcsec$^{-1}$. We will also use a systemic (heliocentric) velocity $V_s$ of 1080~\\kms, infered from the two-dimensional stellar velocity field (Emsellem \\etal\\ \\cite{Ems96}, hereafter Paper~2), and a value of 84\\degr\\ ($\\pm 2$\\degr) for the inclination of the galaxy (Paper~2). \\begin{figure*} \\includegraphics[width=\\textwidth,clip=true,draft=false]{fig1.eps} \\caption{{\\bf Top row:} \\NIIwb\\ line flux (left), heliocentric centroid velocity $V$ (middle, minus $V_s = 1080$~\\kms) and dispersion $\\sigma$ (right, corrected from the instrumental dispersion) maps, as derived from a single component Gaussian fitting of the emission lines in each spectrum of the original (i.e. not deconvolved) TIGER data cube. Black or white contours correspond to selected isophotes (2.5, 5, 15, and 45 $\\times$ \\ten{-18} \\Wmarcsec) of the \\NIIwb\\ line flux map. Spectra with a \\NIIwb\\ line peak intensity $<$ 0.9 $\\times$~\\ten{-19} \\WmAarcsec\\ (i.e. $\\sim 3\\sigma_{\\mbox{\\tt noise}}$) have been discarded in the $V$ and $\\sigma$ maps. {\\bf Bottom row:} Same as top row, but derived from the deconvolved data cube (Richardson-Lucy deconvolution, 40 iterations, see Sect.~\\protect\\ref{SecObsTig}). Spectra with a fitted \\NIIwb\\ line peak intensity $<$ 0.24 $\\times$~\\ten{-19} \\WmAarcsec\\ (i.e. $\\sim 3\\sigma_{\\mbox{\\tt noise}}$) have been discarded in the $V$ and $\\sigma$ maps. North is up, East left.} \\label{FigNII} \\end{figure*} ", "conclusions": "\\label{SecDis} Recently, Regan \\& Mulchaey (\\cite{Regan99}) argued that nuclear bars may not be the primary agent for the fueling of active galactic nuclei (AGN) and suggested nuclear spirals as an alternative mechanism. Martini \\& Pogge (\\cite{MP99}) also analysed visible and near-infrared HST images of a sample of 24 Seyfert 2 galaxies and found that 20 of these exhibit nuclear spirals (with only 5 clear nuclear bars). Physical processes involved in the formation of such spirals are not clear yet, although acoustic instabilities have often been mentioned as a possible mechanism in non self-gravitating nuclear discs (Montenegro \\etal~\\cite{MYE99}, Elmegreen \\etal~ \\cite{Elm+98}, and references therein). In M~104 (classified as a liner), kiloparsec-scale spiral arms are indeed present but a straight dust lane is also found closer to the nucleus: it is a specific signature of strong bars with inner Lindblad resonance (Athanassoula \\cite{Lia92}). If it actually traces a nuclear bar, this dust lane should have a symmetric counterpart on the other side of the nucleus. There is indeed a hint for such a feature (see Sect.~\\ref{SecDust}) but it is rather weak. However, this weakness could be attributed to the fact that this second dust lane would be located on the far (north) side of the galaxy. To test this hypothesis, we computed the effect of the presence of two symmetric dust filaments using the luminosity density model of M~104 of Emsellem (\\cite{Ems95}, Paper~1 hereafter). We assumed the filaments to be in the equatorial plane of the galaxy (10\\degr\\ from end-on), and their characteristics were set to ensure that the mean extinction for the southern filament was consistent with the observations. The model predicts that the apparent extinction should rapidly decrease northward, with apparent $A_V < 0.03$~mag or $E(V-H) < 0.026$, $0\\farcs6$ north of the nucleus (for $R_V = 3.1$). These are upper limits since dust scattering and clumpiness would tend to significantly reduce these values. For these extinction levels, we indeed expect the northern filament to be barely detectable in our colour maps. The observed kinematics would also fit naturally into the strong nuclear bar picture. The fact that overall the velocities are small compared to the predicted circular velocities, obviously argue for the presence of strongly non-circular motions. The velocity profile in the spiral features could be explained by the combination of the streaming motions and projection effects as the spiral curves around the nucleus and becomes perpendicular to the line-of-sight. If the nuclear bar hypothesis is correct, the existence of offset straight dust lanes also requires the presence of an inner Linblad resonance (ILR), and an extended $x_2$ orbit family (e.g. Athanassoula \\& Bureau \\cite{Lia99}, hereafter AthB99). This would explain the observed kinematical decoupling between the gas in the nucleus and in the spiral arms, as clearly illustrated in the models of AthB99. The observed asymmetry in the central velocity gradients agrees also qualitatively with the predictions made by AthB99, when dust is included (see their Fig.~10 with e.g. $\\psi = 22.5$\\degr). We should however keep in mind that significant differences exist between the gas distribution, as observed in the core of M\\,104, and as idealised in the model of AthB99. The presence of a large-scale bar has already been suggested in Papers~1 and 2 (pattern speed $\\Omega_p$ of $\\sim$120~\\kms\\ kpc$^{-1}$ for a distance to M\\,104 of 8.8 Mpc). The mass model predicted a strong ILR for this primary bar, located roughly 20\\arcsec\\ from the nucleus. The secondary nuclear bar discussed in this letter would then be well inside this resonance. From the extension and orientation of the central dust lanes, we can roughly estimate a semi-major axis of $a = 425$~pc ($\\sim$10\\arcsec), and an orientation of approximately 10\\degr\\ from end-on\\footnote{See Gerhard \\cite{Ger89}} for this bar. The location of the corotation can be estimated from the relation $r_L \\simeq 1.2 \\times a$ (AthB99, $r_L$ is the Lagrangian radius), yielding a value of $\\sim$12\\arcsec. This is right where the transition region between the inner and outer disks occurs (Seifert \\& Scorza \\cite{Seifert96}). More detailed modelling is needed to accurately estimate its pattern speed. If there are clearly some hints of the presence of a nuclear bar in M\\,104, as discussed above, it is also clear that additional information are needed to confirm or infirm its existence. In particular, the main support for the nuclear bar scenario comes from the presence of the straight, southern dust filament, which could actually be much further from the nucleus than we assumed (e.g. if it is outside the equatorial plane of the galaxy). In the same way, the kinematical decoupling of the central regions could alternatively be due to, e.g. a nuclear keplerian disk. The answer to this should come soon, from HST/STIS observations of M\\,104. These high spatial resolution data will allow detailed comparison between the observed PVDs and the gaseous kinematics predicted by various models (nuclear bar, keplerian disk). Emission lines in the infrared, where dust is less problematic, could also be very valuable to understand the gas distribution and kinematics in more details." }, "0003/nucl-th0003045_arXiv.txt": { "abstract": "We calculate for the first time the surface tension and curvature coefficient of a first order phase transition between two possible phases of cold nuclear matter, a normal nuclear matter phase in equilibrium with a kaon condensed phase, at densities a few times the saturation density. We find the surface tension is proportional to the difference in energy density between the two phases squared. Furthermore, we show the consequences for the geometrical structures of the mixed phase region in a neutron star. ", "introduction": "The possibility of different phase transitions taking place in the superdense interior of neutron stars has been the target of considerable interest during the last few decades, where pion and kaon condensation as well as quark deconfinement have been investigated. But only less than a decade ago was it realized that if the phase transition is of first order, then a geometrically structured extended region will form in the superdense interior of the neutron star, where the two phases are in equilibrium \\cite{nkg91}. The reason for this richness in structure is that a neutron star has two {\\it globally} conserved charges, baryon number and electric charge, and two chemical potentials associated with these charges. Previous studies using the Maxwell construction could only ensure one chemical potential was common in the two phases, whereas the general phase equilibrium criteria by Gibbs \\cite{gibbs} ensure thermodynamical equilibrium for a system with any number of chemical potentials. The consequences are that the system is not locally charge neutral and a competition between Coulomb and surface energies are responsible for the geometrical structures. Moreover, the common pressure will vary with the proportion of the phases, and thus create an extended mixed phase region with structure in the neutron star. For a first order deconfinement transition, studies of the detailed crystalline structure of the mixed phase region have always been hindered by the lack of a single good model describing both phases. This is in contrast to the first order transition to a kaon condensed phase described in \\cite{glsb98,glsb99}, where both the normal nuclear matter phase and the kaon condensed phase are described by the same relativistic mean-field model, which allow us to calculate the profiles of all important quantities across the interface. From these profiles the Coulomb energy and the surface tension can be found, where, especially for the latter, only educated guesses were previously possible. A condensate consisting of negatively charged kaons is favored in neutron stars because they, contrary to other kaon types, can replace electrons as neutralizing agents \\cite{gl85,kaplan,brown,thor,fujii,li}. Detailed knowledge of the structure of a possible mixed phase region at densities above saturation is important, irrespective of which first order phase transition is responsible for it, as it may have important consequences for transport and superfluid properties and rotation in the form of r-mode instabilities, non-canonical values of the braking index, and glitch phenomena in pulsars \\cite{glbook,weber,reddy,madsen}. In the present paper we calculate the surface properties, i.e., the surface tension and the curvature coefficient, for a semi-infinite slab of normal nuclear matter in phase equilibrium with a semi-infinite slab of kaon condensed matter and show the resulting crystalline structure in the central part of the neutron star. Since we assure compliance with Gibbs phase equilibrium criteria, the two phases cannot be separately charge neutral, though overall the net charge vanishes. Thus for a infinite system we cannot explicitly take Coulomb interactions into account in the Lagrangian. However, it turns out that the typical radii of the geometrical structures are smaller than the Debye screening lengths of about 10 fm \\cite{pethick93}, and therefore it is a reasonable first approximation to ignore this effect in the calculation of the surface tension. Section II contains a description of the non-uniform relativistic mean-field model used to describe both phases. In Sec.\\ III the surface properties are described, whereas the consequences for the crystalline structure in a neutron star are illustrated in Sec.\\ IV. Finally our results are summarized in Sec.\\ V. ", "conclusions": "Little is known about the equation of state for nuclear matter above saturation density, but it is expected that at least a phase transition to deconfined quark matter happens at a few times the saturation density. However, this phase transition may not be the only one encountered in neutron stars. A transition to a phase with a $K^-$ condensate is also a possibility. If any of these phase transitions is of first order, a neutron star will have a mixed phase region in its dense interior, which is very likely to have some observable consequences. For the first time the surface properties in the interface between normal nuclear matter and kaon condensed matter have been calculated, which makes it possible to study this mixed phase region in greater detail. Our calculations are only a first approximation, there are a number of complicating aspects, which we have ignored or only treated approximately - e.g., explicit consideration of the Coulomb field which results in screening effects; the validity of the assumption that the surface tension is the same for semi-infinite slabs as it is for small slabs, rods, and drops; and the importance of the curvature and even higher order terms compared to the surface term. Concerning the first two points, we can generally say that screening effects will reduce the Coulomb energy, likewise the surface energy will decrease due to the decrease in the surface tension when the system is squeezed. The reason for the latter is that the surface region will dominate a small system, so that only $\\epsilon_s$ and not $d\\epsilon_s/dz$ is zero at the boundaries. For example a slab which is 10 fm thick and with a volume fraction of kaon condensed phase of about 0.4, the surface tension is reduced about 12\\% compared to the surface tension of the infinite system (30 fm thick), while the absolute value of the curvature coefficient drops about a factor of three. These two effects pull in opposite directions with regard to the size of the geometry which minimize the sum of Coulomb and surface energies, thus the overall effect on the size is expected to be only minor. We have taken another step towards a better understanding of the mixed phase region for a first order transition involving a kaon condensate. We do, however, realize there is still much room for improvement." }, "0003/astro-ph0003437_arXiv.txt": { "abstract": "The origin of the hard X--ray background (XRB) as a superposition of unabsorbed and absorbed Active Galactic Nuclei is now widely accepted as the standard model. The identification of faint X--ray sources in {\\sf ROSAT}, {\\sf ASCA}, and {\\sf BeppoSAX} medium--deep surveys and their average spectral properties are in broad agreement with the model predictions. However, AGN models, at least in their simplified version, seem to be at odds with some of the most recent findings calling for substantial revisions. I will review the recent XRB ``best fit\" models and discuss how the foreseen {\\sf XMM} and {\\sf Chandra} surveys will be able to constrain the allowed parameter space. ", "introduction": "It has been recognized, already a few years ago, that a self--consistent AGN model for the XRB requires the combined fit of several observational constraints in addition to the XRB spectral intensity such as the number counts, the redshift and absorption distribution in different energy ranges, the average spectra and so on (Setti \\& Woltjer 1989). First attempts towards a ``best fit\" solution relied on simplified assumptions for the AGN spectral properties and for the evolution of their luminosity function (Madau, Ghisellini \\& Fabian 1994 (MGF94), Comastri et al. 1995 (CSZH95), Celotti et al. 1995 (CFGM95)). A three step approach has been followed to build the so--called baseline model: the first step is to assume a single average spectrum for the type 1 objects which is usually parameterized as a power law plus a reflection component from a face--on disk and a high--energy cut--off at a few hundreds of keV. A distribution of absorbing column densities for type 2 objects is then added in the second step. Finally the template spectra are folded with an evolving XLF which, in the framework of unified models, does not depend on the amount of intrinsic obscuration. The number density and absorption distribution of obscured sources are then varied until a good fit is obtained. The baseline model led to a successful description of most of the observational data available before 1995 and to testable predictions for the average properties of the sources responsible for the bulk of the XRB. The increasing amount of data from soft and hard X--ray surveys combined with the study of nearby bright sources has been used to obtain a more detailed description of the AGN X--ray spectra and absorption distribution. In addition, the optical identification of sizeable samples of faint AGNs discovered in the {\\sf ROSAT}, {\\sf ASCA} and {\\sf BeppoSAX} surveys has shed new light on the evolution of the AGN luminosity function opening the possibility to test in more detail the AGN synthesis model predictions. As a consequence, the modelling of the XRB has attracted renewed attention and several variations/improvements with respect to the baseline model have been proposed. However, despite the increasing efforts, a coherent self--consistent picture of ``the\" XRB model has yet to be reached, as most of its ingredients have to be extrapolated well beyond the present limits. Besides the interest in a best--fit model it is by now clear that the problem of the origin of the XRB is closely related to the evolution of accretion and obscuration in AGN. As a consequence, the XRB spectrum should be considered as a useful tool towards a better understanding of the history of black hole formation and evolution in the Universe (Fabian \\& Iwasawa 1999) and the interplay between AGN activity and star--formation (Franceschini et al. 1999; Fabian this volume). ", "conclusions": "In order to achieve a major improvement in the exploration of XRB models parameter space, the resolved fraction of its energy density should be of the order of 50--60 \\% or higher. \\begin{table}[h] \\centerline{\\bf Table.~2 - Resolved fraction of the XRB} \\begin{center} \\begin{tabular}{|c|c|c|} \\hline \\multicolumn{1}{l}{\\bf 2-10 keV Flux interval} & \\multicolumn{1}{c}{\\bf Relative \\%} & \\multicolumn{1}{c}{\\bf Integral \\%} \\\\ \\hline $>$ 10$^{-11}$ & 0.5 & 0.5 \\\\ 10$^{-12}$-10$^{-11}$ & 2 & 2.5 \\\\ 10$^{-13}$-10$^{-12}$ & 8 & 10.5 \\\\ 10$^{-14}$-10$^{-13}$ & 32.5 & 43 \\\\ 10$^{-15}$-10$^{-14}$ & 39 & 82 \\\\ 10$^{-16}$-10$^{-15}$ & 16 & 98 \\\\ $<$ 10$^{-16}$ & 2 & 100 \\\\ \\hline \\end{tabular} \\end{center} \\end{table} The expected contribution of AGN to the 2--10 keV XRB is reported in Table 2 as a function of flux. The model parameters are such to account for an intensity of $\\sim$ 7 $\\times$ 10$^{-8}$ erg cm$^{-2}$ s$^{-1}$ sr$^{-1}$ (in between the {\\sf ASCA} and {\\sf BeppoSAX} measurements) at $\\sim$ 10$^{-17}$ cgs. The predictions are model dependent and should be considered as indicative. Nevertheless it is clear that at the fluxes sampled by the foreseen {\\sf Chandra} and {\\sf XMM} medium--deep surveys most of the XRB will be resolved allowing to unveil the nature of the sources making the bulk of its energy density. The most important challenge for XRB models will be the study of X--ray absorption and luminosity distribution for 2--10 keV fluxes $<$ 10$^{-13}$ cgs, the search for heavily obscured AGN which according to the predictions are expected to show up in a substantial fraction below $<$ 10$^{-14}$ cgs (cfr. Fig.~3), and the optical--infrared follow--up of X--ray obscured sources." }, "0003/astro-ph0003092_arXiv.txt": { "abstract": "We present optical photometry of the X-ray transient XTE J2123--058, obtained in July--October 1998. The light curves are strongly modulated on the 5.95hrs orbital period, and exhibit dramatic changes in amplitude and form during the decline. We used synthetic models which include the effect of partial eclipses and X-ray heating effects, to estimate the system parameters, and we constrain the binary inclination to be i=73$^{\\circ}\\pm4$. The model is successful in reproducing the light curves at different stages of the decay by requiring the accretion disc to become smaller and thinner by 30\\% as the system fades by 1.7 mags in the optical. From Aug 26 the system reaches quiescence with a mean magnitude of R=21.7$\\pm$0.1 and our data are consistent with the optical variability being dominated by the companion's ellipsoidal modulation. ", "introduction": "Soft X-ray transients (SXTs) are a subclass of low-mass X-ray binaries (LMXBs) that are characterized by episodic X-ray outbursts (usually lasting for several months), when the X-ray luminosities can increase by as much as a factor of 10$^7$ (van Paradijs \\& McClintock, 1995). The observed optical flux is generated by X-ray re-processing in the accretion disc and the companion star. These outbursts recur on a time scale of decades, but in the interim the SXTs are in a state of quiescence and the optical emission is dominated by the radiation of the faint companion star. This offers the best opportunity to analyze the properties of this star and obtain dynamical information which eventually enables us to constrain the nature of the compact object. There are currently 12 SXTs with identified optical counterparts, with 8 dynamical black-holes and 3 confirmed neutron stars: CenX-4, Aql X-1 and 1608-522 (van Paradijs \\& McClintock, 1995). In addition, there are a few neutron star binaries exhibiting X--ray on and off states ( EXO0748--676, 4U 2129+47 and SAX J1748.9--2021 ), although they are not classified as SXTs because they do not show the classic fast rise and slow exponential/linear decay. The X-ray transient J2123--058 was discovered on 29 June 1998 by the Rossi X-Ray Time Explorer ({\\it RXTE}) (Levine et al. 1989) reaching a peak X-ray flux of 100 mCrabs (2-12 keV). Its high Galactic latitude (b=-36$^{\\circ}$.2) is unusual among transients, an indication that J2123--058 might be a member of the galactic halo population. The optical counterpart was promptly identified with a variable star of R=17.2 (Tomsick et al. 1998a), which was only marginally visible on a digitized U.K. Schmidt plate, suggesting a preoutburst magnitude R$\\ge$20 (Zurita et al. 1998). Spectra obtained early in the outburst showed strong high-excitation lines of He\\,{\\sc ii} $\\lambda$4686, C\\,{\\sc iii}/N\\,{\\sc iii} $\\lambda$4640 and weak Balmer emission embedded in broad absorptions (Tomsick et al. 1998a, Hynes et al. 1999). These features are frequently observed in SXTs during outburst (e.g. Callanan et al 1995) and persistent LMXBs (e.g. Augusteijn et al. 1998). Type-I (thermonuclear) bursts have been detected both in X-rays (Takeshima and Strohmayer 1998) and optical (Tomsick et al. 1998b), a signature of a neutron star in the binary. The outburst light curve exhibited regular 0.7 mag deep triangular-shaped minima repeating every 6hrs (Casares et al. 1998, Tomsick et al. 1998b), a strong indication of high inclination. This provided the first evidence for the system orbital period (P= 5.957$\\pm$0.003) which was later confirmed by a radial velocity study of the He{\\sc ii} $\\lambda$4686 emission line (Hynes et al. 1998). In addition, Ilovaisky and Chevalier (1998) reported the presence of a 0.3 mag modulation with a period of 7.2 days, probably caused by the precessing disc. Since 26 Aug the system had settled down to its quiescent state at R$\\simeq$21.7 (Zurita \\& Casares 1998). This paper presents the results of a comprehensive set of observations that have led to detailed optical light curves from outburst through the decay into quiescence. Our spectroscopy will be the subject of a second paper (Hynes et al. 2000). \\vskip 10mm \\begin{tabbing} {\\bf Table 1} \\ Log of observations. \\end{tabbing} \\begin{tabular}{lcrc} \\hline \\hline {\\em Date} & {\\em HJD$^{(*)}$} & {\\em Exp/Filter} & {\\em Telescope}\\\\ \\hline {\\it Jul02 98} & -3 & 1xR & IAC80 0.8m$^{(1)}$\\\\ {\\it Jul04 ''} & -1 & 1xB,1xV,1xR & IAC80 0.8m\\\\ {\\it Jul06 ''} & 01 & 80xR & OGS 1m$^{(2)}$\\\\ {\\it Jul07 ''} & 02 & 79xR & OGS 1m \\\\ {\\it Jul08 ''} & 03 & 20xR & OGS 1m \\\\ {\\it Jul09 ''} & 04 & 33xR & OGS 1m \\\\ {\\it Jul10 ''} & 05 & 20xR & OGS 1m \\\\ {\\it Jul12 ''} & 07 & 42xR & OGS 1m \\\\ {\\it Jul13 ''} & 08 & 24xR & OGS 1m \\\\ {\\it Jul14 ''} & 09 & 2xV & M. Canopus 1m$^{(3)}$\\\\ {\\it Jul16 ''} & 11 & 2xV & M. Canopus 1m\\\\ {\\it Jul18 ''} & 13 & 1xB,1xV,1xR & M. Canopus 1m\\\\ {\\it Jul19 ''} & 14 & 40xBV & Crimean 0.5m$^{(4)}$\\\\ & & 36xR & Crimean 1.25 m$^{(5)}$\\\\ & & 2xB,2xV,2xR & M. Canopus 1m\\\\ {\\it Jul20 ''} & 15 & 42xBV & Crimean 0.5m\\\\ & & 74xR & Crimean 1.25m\\\\ & & 1xB,5xV,1xR & M. Canopus 1m\\\\ {\\it Jul21 ''} & 16 & 36xBV & Crimean 0.5m\\\\ & & 94xR & Crimean 1.25m\\\\ {\\it Jul22 ''} & 17 & 25xBV & Crimean 0.5m\\\\ & & 43xR & Crimean 1.25m\\\\ {\\it Jul23 ''} & 18 & 48xBV & Crimean 0.5m\\\\ & & 45xR & Crimean 1.25m\\\\ {\\it Jul24 ''} & 19 & 61xR & Kryonerion 1.2m$^{(6)}$\\\\ & & 55xBV & Crimean 0.5m\\\\ {\\it Jul26 ''} & 21 & 70xR & OGS 1m\\\\ & & 46xBV & Crimean 0.5m\\\\ {\\it Jul27 ''} & 22 & 77xR & OGS 1m\\\\ & & 33xBV & Crimean 0.5m\\\\ {\\it Jul28 ''} & 23 & 77xR & OGS 1m\\\\ & & 40xBV & Crimean 0.5m\\\\ {\\it Jul29 ''} & 24 & 74xR & OGS 1m\\\\ & & 49xBV & Crimean 0.5m\\\\ {\\it Jul30 ''} & 25 & 81xR & OGS 1m\\\\ & & 33xBV & Crimean 0.5m\\\\ {\\it Jul31 ''} & 26 & 40xBV & Crimean 0.5m\\\\ \\hline \\hline \\end{tabular} \\\\ \\begin{tabbing} \\end{tabbing} \\begin{tabular}{lcrc} \\hline \\hline {\\em Date} & {\\em HJD$^{(*)}$} & {\\em Exp/Filter} & {\\em Telescope}\\\\ \\hline {\\it Aug01 98} & 27 & 23xBV & Crimean 0.5m\\\\ & & 4xR & Crimean 0.38m\\\\ {\\it Aug02 ''} & 28 & 1xR & IAC80 0.8m\\\\ {\\it Aug03 ''} & 29 & 1xR & IAC80 0.8m\\\\ {\\it Aug04 ''} & 30 & 2xR & Crimean 0.38m\\\\ & & 1xR & IAC80 0.8m\\\\ {\\it Aug05 ''} & 31 & 2xR & Crimean 0.38m\\\\ {\\it Aug12 ''} & 38 & 1xR & IAC80 0.8m\\\\ {\\it Aug15 ''} & 41 & 12xR & IAC80 0.8m\\\\ {\\it Aug16 ''} & 42 & 26xR & IAC80 0.8m\\\\ {\\it Aug17 ''} & 43 & 8xR & IAC80 0.8m\\\\ {\\it Aug19 ''} & 45 & 3xV,1xR & M. Canopus 1m\\\\ {\\it Aug26 ''} & 52 & 18xR & OGS 1m\\\\ & & 1xV & IAC80 0.8m\\\\ {\\it Aug27 ''} & 53 & 14xR & OGS 1m\\\\ {\\it Sep02 ''} & 59 & 4xR & IAC80 0.8m\\\\ {\\it Sep23 ''} & 80 & 11xR & OGS 1m\\\\ {\\it Sep24 ''} & 81 & 8xR & OGS 1m\\\\ {\\it Sep25 ''} & 82 & 5xR & OGS 1m\\\\ {\\it Sep26 ''} & 83 & 5xR & OGS 1m\\\\ {\\it Jun21 99} &351 & 1xR,1xV,1xI & JKT 1m$^{(7)}$\\\\ & & & \\\\ \\hline \\hline \\end{tabular} \\vspace{5pt} \\footnotesize{$^*$HJD--2451000}\\\\ \\footnotesize{$^1$IAC80--80cm Telescope in the Observatorio del Teide (Tenerife).}\\\\ \\footnotesize{$^2$ 1 m Optical Ground Station in the Observatorio del Teide (Tenerife).}\\\\ \\footnotesize{$^3$ Mount Canopus 1m telescope at Tasmania.}\\\\ \\footnotesize{$^4$ 0.5m telescope at Crimea}\\\\ \\footnotesize{$^5$ 1.25m reflector of SAI Crimean Station.}\\\\ \\footnotesize{$^6$ 1.2m telescope at the Kryonerion Astronomical station of the National Observatory of Athens at Kryonerio of Korinthia.}\\\\ \\footnotesize{$^7$ 1m Jacobus Kapteyn Telescope in the Roque de los Muchachos Observatory (La Palma). \\normalsize \\noindent ", "conclusions": "XTE J2123--058 is a remarkable neutron star binary. Our optical light curve shows marked orbital modulations with dramatic variations as the outburst declines. These are very similar to the modulations observed in accretion disc corona (ADC) sources of comparable orbital periods (e.g. 4U 2129+47 and 2A 1822--371; see eg. McClintock et al. 1980 and Mason et al. 1982) although none of these are transient systems. The X-ray light curve displays the classic SXT properties, namely a FRED morphology with typical e-folding rise and decay times and secondary maximum. Furthermore, the ratio of X-ray to optical luminosity [$\\xi=B_{0} +2.5 \\log F_{\\rm x} (\\mu\\rm{Jy})$] is also in excellent agreement with the observed distribution of LMXBs. Taking B=17.28 (Tomsick et al. 1998a) and F$_{\\rm x} (2-12 keV) \\simeq 100$ mCrab (Levine, Swank \\& Smith 1998) at the outburst peak and assuming $A_{\\rm v}=0.37$ (Hynes et al 1998) we obtain $\\xi$=21.9, whereas the distribution peak of LMXBs gives $\\xi$=21.8 $\\pm$ 1.0 (see van Paradijs \\& McClintock 1995). This result implies that, despite its high inclination, the X-ray source in J2123-058 is not hidden by the accretion disc (i.e. $\\alpha\\geq$90$^{\\circ}$--i). This is consistent with the values our model favors for the binary inclination (i=73$^{\\circ}$) and disc flaring angle ($\\alpha$=5$^{\\circ}$.7 -- 7$^{\\circ}$.6). The longterm evolution of the optical light curve can be compared to those of other SXTs (e.g. GRO J0422+32, A0620-00, N. Muscae 91). They show a slow linear decay followed by a steeper fall. We find that J2123--058 also reproduces this behaviour, although the total amplitude and time scales are a factor of $\\sim$ 2 shorter (see e.g. Callanan et al 1995). We have modeled our R-band light curves of J2123--058 at different stages of the outburst including the obscuration effects and X-ray heating of the secondary star accretion disc. This led us to constrain the system inclination to i=73$^{\\circ}\\pm4$. We find encouraging the excellent agreement between the inclination values obtained for the two independent light curves (July 26-30 and 16 Aug) during which $L_{\\rm x}$ has dropped by one order of magnitude. The light curves at the plateau phase (July) are very similar to those of EXO 0748--676 and 2A 1822--371 with an extended depression of the luminosity from phase $\\sim$0.7 until the eclipse and a steeper rise to maximum (see Mason et al. 1980 and Schmidtke et al 1987). Our model fit indicates that the accretion disc is the dominant source of light and the triangular shaped minima can be interpreted as eclipses of the accretion disc by the secondary star together with the changing aspect of the heated polar caps of the companion star. The dramatic changes observed in the light curves during decline, are triggered by large changes in the disc size and geometry. Our fit to the decay data (16 Aug) demands a thinner and smaller accretion disc which implies a smaller fraction of the disc is X-ray heated. Conversely, the secondary star is more exposed to the X-ray radiation and therefore the total amplitude of the modulation increases by a factor of 2 (to $\\sim1.4$ mag). The resulting light curve has a sinusoidal-like shape and is reminiscent of the LMXBs 4U 2129+47(=V1727 Cygni) and HZ Her. Our model fits implies a change of $\\sim$30 percent in the disc size, as the system fades by 1.7 mags in the optical. The change in the radius of the disc size is what one expects. If angular momentum is transported outwards in the disc through viscous processes, then at outburst, since matter diffuses inwards, the angular momentum of that matter has to be transfered to the outer parts of the disc, and the radius of the disc is expected to expand. When the system is decaying, after the end of the mass transfer enhancement, the disc shrinks to its original radius (Livio \\& Verbunt 1988; Ichikawa \\& Osaki 1992). Observations of U Gem, Z Cha, OY Car, HT Car show that the accretion discs are indeed larger in outburst than in quiescence (Smak 1984b; O'Donohuge 1986; Harrop-Allin \\& Warner 1996). Comparing our results with disc radius variation in U Gem (Smak 1984a), we find approximately the same rate ofdecrease." }, "0003/astro-ph0003279_arXiv.txt": { "abstract": "Effects of young stellar clusters on their gas and dust environment are probed using mid-infrared (MIR) wavelengths. The strong MIR [NeIII]/[NeII] ratios ($\\sim$ 5 to 10) reveal the presence of current massive stars less than 5\\,Myr. Using MIR line ratios along with optical and NIR data from the literature, composite SEDs are constructed for NGC~1569, NGC~1140 and II\\,Zw40. The stellar SEDs are then used as input to a dust model to study the impact of the hard, penetrating radiation field on the dust components, particularly in low metallicity environments, where the destructive effects of the massive stellar clusters on the environments occur on global scales. For example, the smallest dust particles are destroyed over larger regions in the dwarf galaxies than in normal metallicity starbursts. ", "introduction": "The subject of stellar cluster formation history and environment has made great headway lately, with the high resolution and sensitivity currently available at optical and near infrared (NIR) wavelengths. In principle, the mid-infrared (MIR) wavelength regime should provide numerous advantages for such studies, since this wavelength range is relatively extinction free (A$_{15\\mu}$$_{m}$ $\\sim$ 5\\% A$_{J}$) and contains diagnostic ionic lines to probe HII regions. In addition, hot dust emission provides us with another link to the ultraviolet starlight that has been absorbed and reemitted by the nearby grains. Our knowledge of the MIR wavelength window has been limited by the low spatial and spectral resolution provided by the IRAS satellite, and has remained rather sketchy when it comes to detailed studies of the ISM of individual galaxies. The Infrared Space Observatory (ISO; Kessler et al. 1996) has been a recent turning point in this effort, providing high spectral and spatial resolution and unprecedented sensitivity in the MIR through the far infrared (FIR). We have incorporated these MIR and FIR observations in a study of the energy redistribution in starburst galaxies, with the aim of understanding the impact of the star formation on the surrounding gas and dust. The main limitation in MIR star cluster studies remains the spatial resolution, despite the great improvement over previous instrumentation provided by ISOCAM ($\\sim$ 6$\\arcsec$ at 15$\\mu$m; Cesarsky, C.J. et al. 1996). ISOCAM resolves about 600 pc at 20 Mpc, the distance of the closest massive merging system, the Antennae. However, even with this limitation, we are able to draw noteworthy conclusions from the MIR, from unique MIR diagnostics. Here, I concentrate primarily on results of the nearer dwarf galaxies, since impacts of the massive clusters on the global dust and gas environment are very pronounced in these relatively small objects. ", "conclusions": "MIR ISOCAM spectroscopy provides details of ionic lines, UIBs and the distribution of small hot grain emission in dwarf galaxies. The strong MIR [NeIII]/[NeII] ratios are signatures of the hard radiation fields and indicate the presence of clusters of young massive stars in dwarf galaxies. Because of the increase in T$_{eff}$ in low metallicity environments, this ratio is enhanced in dwarf galaxies to at least 5 to 10 times that observed in normal metallicity galaxies. The penetrating radiation field also affects the dust components, destroying the UIBs in some dwarf galaxies on global scales, as is evident in the MIR spectra and in the dust modeling. This dramatic global effect of the massive stellar population in dwarf galaxies, due to the decrease in attenuation of the UV flux, is not apparant in normal metallicity galaxies, where these effects are experienced much more locally." }, "0003/astro-ph0003086_arXiv.txt": { "abstract": "First results are reported of a new abundance study of neutron-capture elements in the ultra-metal-poor (UMP; [Fe/H]~= --3.1) halo field giant star \\cs22892. Using new high resolution, high signal-to-noise spectra, abundances of more than 30 neutron-capture elements (Z~$>$~30) have been determined. Six elements in the 40~$<$~Z~$<$~56 domain (Nb, Ru, Rh, Pd, Ag and Cd) have been detected for the first time in a UMP star. Abundances are also derived for three of the heaviest stable elements (Os, Ir, and Pb). A second transition of thorium, \\ion{Th}{2} \\wave{4086}, confirms the abundance deduced from the standard \\ion{Th}{2} \\wave{4019} line, and an upper limit to the abundance of uranium is established from the absence of the \\ion{U}{2} \\wave{3859} line. As found in previous studies, the abundances of the heavier (Z~$\\ge$~56) stable neutron-capture elements in \\cs22892\\ match well the scaled solar system r-process abundance distribution. From the observed Th abundance, an average age of $\\simeq$ 16~$\\pm$ 4~Gyr is derived for \\cs22892, consistent with the lower age limit of $\\simeq$~11~Gyr derived from the upper limit on the U abundance. The concordance of scaled solar r-process and \\cs22892\\ abundances breaks down for the lighter neutron-capture elements, supporting previous suggestions that different r-process production sites are responsible for lighter and heavier neutron-capture elements. ", "introduction": "Ultra-metal-poor (UMP) stars serve a critical role for understanding the initial epochs of our Galaxy; the observed abundances in these very old stars provide clues to the nucleosynthetic processes in the earliest Galactic stellar generations. The UMP halo giant \\cs22892\\ ([Fe/H] = --3.1\\footnote{ [A/B]~$\\equiv$~log$_{\\rm 10}$(N$_{\\rm A}$/N$_{\\rm B}$)$_{\\rm star}$~-- log$_{\\rm10}$(N$_{\\rm A}$/N$_{\\rm B}$)$_{\\odot}$, and \\eps{A}~$\\equiv$~log$_{\\rm 10}$(N$_{\\rm A}$/N$_{\\rm H}$)~+~12.0, for elements A and B. UMP stars are considered to be those with [Fe/H]~$<$~--2.5.}) merits special attention in nucleosynthesis studies. This star has extremely large overabundances of neutron-capture (n-capture) elements relative to iron, and the abundances of those elements with Z~$\\ge$~56 apparently are consistent only with a scaled solar system r[apid]-process abundance distribution (Sneden \\etal\\ 1996; Cowan \\etal\\ 1999; Norris, Ryan, \\& Beers 1997; Pfeiffer, Kratz \\& Thielemann 1997). In addition, this is the first star for which thorium and an extensive number of n-capture elements have been detected, allowing an estimation of its radioactive age (Sneden \\etal\\ 1996; Cowan \\etal\\ 1997, 1999; Pfeiffer \\etal\\ 1997). Thus far, two n-capture element domains in \\cs22892\\ and other UMP stars have been largely unexplored: the region 40~$<$~Z~$<$~56 (between Zr and Ba); and the region 75~$<$~Z~$<$~83 (the 3$^{\\rm rd}$ n-capture peak peak, Os$\\rightarrow$Pb). Using extensive new high resolution spectroscopic data, we have derived a new model-atmosphere for \\cs22892, and derived abundances for a set of n-capture elements never before seen in this (or any other UMP) star. These new abundances and their implications for early Galactic nucleosynthesis of n-capture elements are discussed in this {\\it Letter}. ", "conclusions": "In the top panel of Figure~2 we plot the \\cs22892\\ n-capture abundances from this study, and those of Sneden \\etal\\ (1996) for elements not analyzed by us. A scaled solar system r-process elemental abundance distribution is also shown. The solar distribution is obtained by a decomposition of the solar system elemental abundances (Anders \\& Grevesse 1989) into s[low]- and r-process contributions to individual isotopes, and is based upon the measured isotopic n-capture cross sections (K{\\\"a}ppeler \\etal\\ 1989, Wisshak \\etal\\ 1996). Summation of those contributions produces a solar system r-process {\\it elemental} abundance curve; see Burris \\etal\\ (2000) for details of this procedure. The solar system curve has been shifted to match the mean abundance level of the heavier n-capture elements (56~$\\leq$~Z~$\\leq$~72) in \\cs22892. The mean difference is $<$log~$\\epsilon_\\cs22892$~--~log~$\\epsilon_{\\rm s.s.}>$~= --1.41~$\\pm$~0.02 ($\\sigma$~=~0.08, 15 elements). The small scatter about the mean value confirms and extends all previous studies that have found consistency between \\cs22892\\ heavy n-capture element abundances and the solar system r-process distribution. In the bottom panel of Figure~2 this agreement is shown via a plot of the differences $\\delta$(log~$\\epsilon$) between the \\cs22892\\ abundances and the scaled solar curve. Abundances of the 3$^{\\it rd}$ n-capture peak elements Os, Ir, and Pb are also in good agreement with the scaled solar system r-process distribution. Abundances of 3$^{\\it rd}$ peak elements in two other UMP stars (\\eg, Sneden \\etal\\ 1998) have anticipated this result, but these are the first reliable 3$^{\\it rd}$ peak abundances in \\cs22892. Thus the solar r-process pattern extends throughout the 56~$\\leq$~Z~$\\leq$~82 element domain in this star.\\footnote{ The Pb abundance is based on extremely weak transitions, and remains poorly determined.} A nearly identical abundance distribution is observed in the UMP star HD 115444 (Westin \\etal\\ 2000). In addition, the [Ba/Eu] ratio in most UMP stars is in accord with the solar r-process value (\\eg, McWilliam 1998, Burris \\etal\\ 2000). The agreement of all of these abundance patterns with the solar system r-process distribution suggests a uniform site, and/or uniform conditions for synthesis of the heavier n-capture elements. The thorium abundance of \\cs22892\\ lies below the solar r-process curve in Figure~2, indicating that radioactive decay of this element has taken place over the time since it was created by the progenitor of this star. We computed a Th-based radioactive age for \\cs22892\\ with various input assumptions (such as using both theoretical r-process predictions and the observed solar system abundances). The calculations are sensitive to small parameter changes and give a range of results with a average of $\\simeq$~16~Gyr. The error bars on the derived age, including both observational and theoretical uncertainties, are $\\simeq$~4~Gyr. An age estimate for \\cs22892\\ from the upper limit on the uranium abundance can be done by comparing this limit with predictions (Cowan \\etal\\ 1999) for the long-lived uranium $^{238}$U isotope ({\\it i.e.}, assuming that we are not observing any of the relatively quickly decaying $^{235}$U isotope). This yields a lower age limit of $\\simeq$ 11~Gyr for \\cs22892. While this limit is even more uncertain than values based upon the detection of Th, it does provide a lower bound on the age of this star which, within the error uncertainty, is consistent with the age determination using the Th chronometer. In contrast to the heavy stable elements, our observations clearly demonstrate that the agreement between \\cs22892\\ and solar system r-process abundances fails for the lighter (Z~$<$~56) n-capture elements (see Figure~2). For these nine elements, $<$log~$\\epsilon_\\cs22892$~--~log~$\\epsilon_{\\rm s.s.}>$~= --1.72~$\\pm$~0.07 ($\\sigma$~=~0.20). Abundances of six of the lighter n-capture elements lie well below the solar system r-process curve that reproduces the heavier elements. The abundances of the odd-Z light n-capture elements are substantially less than those of the even-Z elements, a pattern typically seen in s-process nucleosynthesis. However, the scaled solar system s-process abundance distribution is a poor match to the \\cs22892\\ lighter n-capture abundances. Numerical experiments, analogous to those conducted by Cowan \\etal\\ (1995), suggest that a mix consisting of solar r-process abundances plus 10\\% of the solar s-process abundances can roughly fit the \\cs22892\\ data. But the Y and Ag abundances of \\cs22892\\ are 0.3-0.5~dex lower than this hybrid solar system distribution. In addition, the overall abundance level of these elements is still about 0.2~dex below the level of the heavier elements, when compared together to the solar distribution (Figure~2). There does not appear to be a simple way to mix solar n-capture abundances to match those of \\cs22892. The existence of two distinct r-process signatures in solar system meteoritic material, one for n-capture nuclei lighter than mass number 140 and one for heavier nuclei, has been previously suggested by Wasserburg, Busso, \\& Gallino (1996), Qian, Vogel, \\& Wasserburg (1998). The clear differences between the abundances of the heavier and the lighter r-process elemental abundances in CS 22892--052 are consistent with that suggestion. Thus lighter and heavier nuclei possibly could be produced on different Galactic timescales and come from supernovae of different mass ranges (Qian \\& Wasserburg 2000). Alternatively, neutron-star binaries also could be a source for one of the mass ranges of r-process nuclei (Rosswog \\etal\\ 1999). The total n-capture \\cs22892\\ abundance pattern is also consistent with a neutrino-heated supernova ejecta r-process in a single supernova event, albeit with two different epochs in the explosion/ejection process ({\\it cf.}, Woosley \\etal\\ 1994, and references therein). In neutrino-heated ejecta nucleosynthesis models the abundance yields are extremely sensitive to the electron fraction in the shock re-heating epoch when the lighter r-process species are synthesized (Hoffman \\etal\\ 1996). The electron fraction and the resulting lighter r-process abundances in this early epoch are expected to be different for each supernova event. By contrast, the later neutrino-driven wind epoch, where the heavier r-process nuclides originate in these models, should have similar conditions, therefore producing the same abundance pattern, in all supernovae. An additional clue about early Galactic n-capture nucleosynthesis lies in the now well-documented very large star-to-star scatter in the bulk [n-capture/Fe] ratios of UMP stars (\\eg, McWilliam \\etal\\ 1995, Burris \\etal\\ 2000). In \\cs22892\\ for example, [Eu/Fe]~$\\sim$~+1.6 (Sneden \\etal\\ 1996), while other UMP stars have [Eu/Fe]~$<$~0. The large scatter in overall n-capture element content is an indication of the chemical inhomogeneity of the Galactic halo; the Galaxy was not well-mixed at very early epochs. Possible explanations for the early scatter in the [n-capture/Fe] ratio involve separating the iron and r-process production into different types of supernovae with an initial iron production from very massive stars (Wasserburg \\& Qian 2000). Detailed abundance analyses of many very low metallicity stars, and more extensive theoretical r-process calculations will be needed to understand better the differences between the production of iron and the entire range of r-process elements in the earliest Galactic stellar populations." }, "0003/astro-ph0003423_arXiv.txt": { "abstract": "We measure the ratio of total to selective extinction, $R_{VI}\\equiv A_V/E(V-I)$, toward Baade's Window by comparing the $VIK$ colors of 132 Baade's Window G and K giants from Tiede, Frogel, \\& Terndrup with the solar-neighborhood $(V-I),(V-K)$ relation from Bessell \\& Brett. We find $R_{VI}=2.283 \\pm 0.016$, and show that our measurement has no significant dependence on stellar type from G0 to K4. Adjusting the Paczy\\'nski et al.\\ determination of the centroid of the dereddened Baade's Window clump for this revised value of $R_{VI}$, we find $I_{0,\\rm RC}=14.43$ and $(V-I)_{0,\\rm RC}=1.058$. This implies a distance to the Baade's Window clump of $d_{BW} = 8.63\\pm 0.16\\,\\kpc$, where the error bar takes account of statistical but not systematic uncertainties. ", "introduction": "Paczy\\'nski \\& Stanek (1998) found that the $VIK$ colors of clump giants in Baade's Window were anomalous in the sense that at fixed $(V-K)_0$, they were redder in $(V-I)_0$ than clump giants in the solar neighborhood by 0.2 mag. Stutz, Popowski, \\& Gould (1999) found a similar offset (0.17 mag) for RR Lyrae stars in Baade's Window compared to those in the solar neighborhood. Popowski (2000) showed that part of these offsets was simply due to errors in the original photometry used by both groups. When he incorporated the revised OGLE photometry of Paczy\\'nski et al.\\ (1999), he found that the offset in $(V-I)$ shrank to $\\sim 0.11$ mag in both cases. Popowski (2000) then reviewed the various attempts to explain such an offset in terms of a difference between the intrinsic properties of stars in the two populations, an idea advanced by Paczy\\'nski (1998) and by Stutz et al.\\ (1999). He argued that such an explanation was not impossible, but unlikely, and that a more plausible explanation is that the ratio of total to selective extinction $R_{VI}\\equiv A_V/E(V-I)=2.5$ adopted by both Paczy\\'nski \\& Stanek (1998) and Stutz et al.\\ (1999) from Stanek (1996) was incorrect. Both color anomalies could be solved, he noted, by adopting $R_{VI}= 2.1$. Popowski (2000) used this re-evaluation to draw various conclusions about the extragalactic distance scale. In the course of calibrating a $K$-band clump-giant distance indicator and applying it to measure the distance modulus of the Baade's Window bulge field, Alves (2000) measured $R_{VI}=2.26$. He did so by comparing the $VIK$ colors of his 20-star sample of Baade's Window clump giants taken from Tiede, Frogel, \\& Terndrup (1995) with the $VIK$ colors of local clump stars. Here we improve on the Alves (2000) measurement by incorporating a factor $\\sim 7$ more stars into the analysis, i.e., all 138 G0-K4 giants with $VIK$ photometry from Tiede et al.\\ (1995). We determine $R_{VI}$ by comparing these $VIK$ colors to the $VIK$ colors of nearby stars as determined by Bessell \\& Brett (1988). We find, \\begin{equation} R_{VI}\\equiv {A_V\\over E(V-I)}=2.283 \\pm 0.016. \\label{eqn:rvi} \\end{equation} This reduces the color anomalies to $\\sim 0.05\\,$mag and so qualitatively confirms Popowski's (2000) explanation of them. ", "conclusions": "As with essentially all methods for determining total and selective extinction, our measurement relies on the assumption that the colors of stars in Baade's Window are the same as those in the solar neighborhood. If the mean $(V-I)_0$ color at fixed $(V-K)_0$ differed between the two populations by $\\Delta(V-I)$, then our estimate of $\\alpha$ would likewise be in error by $\\Delta \\alpha = \\Delta(V-I)/\\langle A_V\\rangle$ where $\\langle A_V\\rangle=1.496$ is the mean value of $A_V$ over our final sample of 132 stars. However, from the fact that $\\beta$ is consistent with zero (eq.\\ [\\ref{eqn:alphabeta}]), such an offset would have to be independent of spectal type from G0 to K4 (and arguably to M5). This seems quite implausible. In addition, approximately the same offset would have to apply to RR Lyrae stars (Stutz et al.\\ 1999). Most probably, the fault lies not in the stars, but in the dust. Popowski (2000) reexamined the Wo\\'zniak \\& Stanek (1996) method by which Stanek (1996) determined $R_{VI}=2.5$. Making use of the original tests done by Wo\\'zniak \\& Stanek (1996), he found that this determination depends in part on an initial assumption about $R_{VI}$ so that values as low as $R_{VI}\\sim 2.3$ would be consistent with the data. Our measurement of $R_{VI}$ removes most, but not all, of the anomalous color problems found by Paczy\\'nski \\& Stanek (1998) and by Stutz et al.\\ (1999). In the latter case, Popowski's (2000) revised offset $0.11\\pm 0.02$ mag is now reduced to $0.04\\pm 0.02$, and so is only a $2\\,\\sigma$ discrepancy. However, for the clump giant anomaly found by Paczy\\'nski \\& Stanek (1998), the formal uncertainty is only $\\sim 0.003$ mag, so statistical fluctuations do not provide a plausible explanation. Nevertheless, the remaining discrepancy is small and may be due to a combination of small offsets between the photometric zero points of the various measurement systems that are used to make the comparison. Paczy\\'nski et al.\\ (1999) used $R_{VI}^{\\rm Stanek}=2.50$ from Stanek (1996) to deredden the observed centroid of the clump in their $VI$ color-magnitude diagram of Baade's Window. The zero point of their $V$-band extinctions is based on Gould et al.\\ (1998) and Alcock et al.\\ (1998) and is not affected by the present paper. However, the $I$-band photometry should be adjusted fainter by $\\Delta I_0 = (1/R_{VI}^{\\rm Stanek} - 1/R_{VI}^{\\rm GSF}) \\langle A_V\\rangle = 0.056$, where $R_{VI}^{\\rm GSF}=2.283$ is the value we determine here and $\\langle A_V\\rangle=1.48$ is the mean extinction of clump stars measured by Paczy\\'nski et al.\\ (1999). This adjustment yields a clump centroid of $I_{0,\\rm RC}=14.43$ and $(V-I)_{0,\\rm RC}=1.058$. The color is only 0.05 mag redward of the centroid of the Hipparcos clump (Paczy\\'nski et al.\\ 1999). Inserting the magnitude into Udalski's (2000) $I$-band calibration of the clump standard candle, $M_I= (-0.26 \\pm 0.02) + (0.13 \\pm 0.07)(\\rm [Fe/H] + 0.25)$, we obtain a distance modulus to Baade's Window \\begin{equation} \\mu_{BW} = 14.43 - M_I = 14.68\\pm 0.04, \\label{eqn:mubw} \\end{equation} where we have adopted [Fe/H]$_{BW}=-0.15\\pm 0.10$. The error bar takes account of all statistical uncertainties, 0.025 mag for $M_I$ (Udalski 2000), $0.02$ mag for the observed brightness of the Baade's Window clump (Paczy\\'nski et al.\\ 1999), $0.04\\times (1-R_{VI})=0.022$ mag for the zero-point uncertainty of the Stanek (1996) map (Alcock et al.\\ 1998), and $0.0033\\times \\langle A_V\\rangle=0.005$ mag for the uncertainty in $R_{VI}$ (this paper), but does not take account of systematic errors. Equation (\\ref{eqn:mubw}) is in good agreement with Alves' (2000) determination using the $K$-band clump distance indicator, $\\mu_{BW}=14.58\\pm 0.11$. {\\bf Acknowledgements}: We thank Piotr Popowski for his useful comments on the manuscript and David Alves for pointing out the error in the column headings of Tiede et al.\\ (1995). This work was supported in part by grant AST 97-27520 from the NSF." }, "0003/astro-ph0003109_arXiv.txt": { "abstract": "To investigate the link between weak tidal interactions in disk galaxies and the boosting of their recent star formation, we obtain images and spatially integrated spectra ($3615~{\\rm \\AA} \\leq \\lambda \\leq 5315~{\\rm \\AA}$) for 40 late-type spiral galaxies (Sab-Sbc) with varying degrees of lopsidedness (a dynamical indicator of weak interactions). We quantify lopsidedness as the amplitude $\\langle \\tilde{A}_1\\rangle$, of the $m=1$ Fourier component of the azimuthal surface brightness distribution, averaged over a range of radii. The median spectrum of the most lopsided galaxies shows strong evidence for a more prominent young stellar population (i.e. strong Balmer absorption, strong nebular emission, a weak $4000~{\\rm \\AA}$ break and a blue continuum) when compared to the median spectrum of the most symmetric galaxies. We compare the young stellar content, quantified by $EW(H\\delta_{abs})$ and the strength of the $4000~{\\rm \\AA}$ break ($D_{4000}$), with lopsidedness and find a $3-4\\sigma$ correlation between the two. We also find a $3.2\\sigma$ correlation between $EW(H\\beta_{emission})$ and lopsidedness. Using the evolutionary population synthesis code of Bruzual \\& Charlot we model the spectra as an ``underlying population'' and a superimposed ``boost population'' with the aim of constraining the fractional boost in the SFR averaged over the past $0.5$ Gyr (the characteristic lifetime of lopsidedness). From the difference in both $EW(H\\delta_{abs})$ and the strength of the $4000~{\\rm \\AA}$ break ($D_{4000}$) between the most and least symmetric thirds of our sample, we infer that $\\sim 1 \\times 10^9 M_{\\odot}$ of stars are formed over the duration of a lopsided event in addition to the ``underlying'' SFH (assuming a final galactic stellar mass of $10^{10} M_{\\odot}$). This corresponds to a factor of $ 8$ increase in the SFR over the past $5 \\times 10^8$ years. For the nuclear spectra, all of the above correlations except $D_{4000}$ vs. $\\langle \\tilde{A}_1\\rangle$ are weaker than for the disk, indicating that in lopsided galaxies, the SF boost is not dominated by the nucleus. ", "introduction": "Galaxies do not live isolated lives, but exist in the tidal fields of their environment. Arp (1966), in his Atlas of Peculiar Galaxies, lay the observational groundwork for the modern study of interacting galaxy systems by identifying many \"peculiar\" systems, later interpreted as various stages of major galaxy mergers. Strong galaxy-galaxy interactions may dramatically alter the stellar populations (e.g. Larson \\& Tinsley 1978; Kennicutt \\etal 1987; Turner 1998; Kennicutt 1998), morphology (e.g. Toomre \\& Toomre 1972; Hernquist, Heyl \\& Spergel 1993) and kinematics of galaxies (e.g. Toomre \\& Toomre 1972; Barnes \\& Hernquist 1992) driving evolution along the Hubble sequence. Massive mergers are also capable of funneling gas into the center of galaxies causing nuclear starbursts (Barnes \\& Hernquist 1991; Mihos, Richstone \\& Bothun 1992; Barnes \\& Hernquist 1996) and QSO activity (e.g. Sanders \\etal 1988). At the present epoch, however, major mergers are fairly rare events (e.g. Kennicutt \\etal 1987) and their broad evolutionary importance is unclear. Minor mergers and, in general, weak tidal interactions between galaxies occur with much higher frequency than major ones (e.g. Lacey \\& Cole 1993). By weak interactions we mean those which do not destroy the disk of the ``target'' spiral. Hierarchical structure formation models (e.g. cold dark matter) predict that the merging histories for high mass objects today contained multiple low mass accretion events in their past (e.g. Lacey \\& Cole 1993). The specific roles which weak interactions play in the evolution of galaxies, however, is uncertain. Weak interactions may cause disk heating (e.g. Toth \\& Ostriker 1992; Quinn, Hernquist \\& Fullagar 1993) and satellite remnants may build up the stellar halo (e.g. Searle \\& Zinn 1978; Johnston, Hernquist \\& Bolte 1996). Kennicutt \\etal (1987) studied the relation between interaction strength and star formation by making a comparison between isolated galaxies, close pairs, and galaxies from the Arp Atlas. They found that close pairs have larger values of $EW(H\\alpha_{em})$, i.e. higher star formation rates (SFR) than isolated galaxies. While pair spacing is weakly correlated with the SFR, they could not determine the specific role of interaction strength on the SFR. Hashimoto \\etal (1998) and Allam \\etal (1999) both studied the Hubble type specific effects of environment on the SFR in galaxies. They found that the SFR/{\\it mass of existing stars} was inversely proportional to the local galaxy density. They postulate that the anti-correlation is due partly to gas stripping and due partly to the anti-correlation of the merger cross-section with the galaxy-galaxy velocity dispersion. There is also evidence that interactions excite nuclear activity. In their close pair and strongly interacting sample Kennicutt {\\it et al.} (1987) found a strong correlation between $H\\alpha$ emission in the disk and that in the nucleus. Such a correlation between disk and nuclear emission is supported by theoretical work; Mihos \\& Hernquist (1994) and Hernquist \\& Mihos (1995) demonstrated that minor interactions form bar instabilities in the disk which in turn funnel large amounts of gas into the nucleus. The effectiveness of this process is suppressed by the presence of a dense bulge, which prevents bar formation. Due to the numerical expense in computing high resolution N-body/SPH (collisionless particle/smoothed particle hydrodynamics) models, the exact interaction parameters which result in such activity are uncertain. Weak interactions may also manifest themselves as kinematic or structural irregularities. Roughly $50\\%$ of all spiral galaxies have asymmetric HI profiles and rotation curves (Baldwin, Lynden-Bell \\& Sancisi 1980; Richter \\& Sancisi 1994; Haynes \\etal 1998). Baldwin \\etal (1980) postulated that these asymmetries are caused by weak interactions in the galaxy's past or by lopsided orbits. Barton \\etal (1999) examined the optical rotation curves of a set of observed and simulated interacting disk galaxies. They showed that interactions can cause large scale, time dependent asymmetries in the rotation curves of their sample galaxies. Swaters \\etal (1999) studied the kinematic asymmetries present in two galaxies lopsided in their optical and HI distributions. They qualitatively reproduced the kinematic asymmetries by placing closed orbits in mildly lopsided potential. A dynamical indicator of weak interactions may be ``lopsidedness.'' In the context of this paper (following Rudnick \\& Rix 1998; hereafter RR98), lopsidedness is defined as a bulk asymmetry in the {\\it mass} distribution of a galactic disk. Surveys for lopsidedness in the stellar light of galaxies were first carried out by Rix \\& Zaritsky (1995; hereafter RZ95) and Zaritsky \\& Rix (1997; hereafter ZR97). Using near-IR surface photometry of face-on spiral galaxies (spanning all Hubble types) they examined the magnitude of the $m=1$ azimuthal Fourier component of the I and K-band surface brightness, thus characterizing the global asymmetry of the stellar light. RZ95 and ZR97 found that a quarter of the galaxies in their sample were significantly lopsided. Using a larger, magnitude limited sample restricted to early type disks (S0 to Sab) and imaged in the R-band, RR98 found that the fraction of significantly lopsided early type disks is identical to that for late-type disks. RR98 convincingly demonstrated that lopsidedness is not an effect of dust, but is in fact the asymmetric distribution of the light from old stars and hence from the {\\it stellar mass} in the disk. Some theoretical work has been done in examining long lived $m=1$ modes (Syer \\& Tremaine 1996; Zang \\& Hohl 1978; Sellwood \\& Merritt 1994), little convincing evidence however has been put forth to show that isolated galaxies will form stable $m=1$ modes without external perturbations or significant counter-rotating populations. Without invoking the special cases above, long lived lopsidedness is possible if the disk resides in a lopsided potential. The question remains however: how is a lopsided potential created/maintained? Numerical simulations of hyperbolic encounters between disk galaxies fail to produce $m=1$ modes of amplitude $>10\\%$ without destroying the pre-existing stellar disk (Naab, T.; private communication). Minor mergers and possibly some weak interactions therefore remain as the most probable cause of lopsidedness (RR98). Recent work has shown that perturbations in the outer halo of a galaxy may be amplified and even transmitted down into the disk (Weinberg 1994). Work by Walker, Mihos \\& Hernquist (1996) and ZR97 showed that the type and magnitude of lopsidedness seen in RZ95, ZR97 and RR98 is comparable to the result of the accretion of a small satellite, if the mass ratio with the main galaxy is $\\approx 1/10$. In a preliminary study (i.e. a rigid halo with no dynamical friction) Levine and Sparke (1998) showed that lopsided galaxies may be formed by disks orbiting off center and retrograde in a flat-cored, dark matter dominated halo. They postulated that a galaxy may be pushed off center by a satellite accretion. Using phase mixing arguments (Baldwin \\etal 1980; RZ95) and analysis of N-body simulations (Walker \\etal 1996; ZR97) the lifetime of lopsided features has been estimated at $t_{lop}\\approx 1$ Gyr. That lopsidedness is transient ($t_{lop} \\ll t_{Hubble}$) yet common, requires that it must be recurring and therefore lopsidedness may have significant evolutionary consequences. The current paper focuses on the impact that minor mergers (observed as lopsidedness) may have on boosting the SFR and the recent star formation history (SFH) of disk galaxies. For the purpose of this discussion, we will assume that lopsidedness is caused by minor mergers. Regardless of what causes lopsidedness however, the perturbation in the gravitational potential manifestly exists and therefore may affect the gas in the galaxy to such a degree as to boost the SFR. Indeed, ZR97 find that lopsidedness is correlated (at $\\geq 96\\%$ confidence) with the ``excess'' of blue luminosity (over what is predicted by the Tully-Fisher relation). Modeling the integrated spectral evolution of starbursts using evolutionary population synthesis (EPS) codes has been been well studied (e.g. Couch and Sharples 1987; Barger \\etal 1996; Turner 1998) and despite its limitations, is a useful tool in determining the relative SFH over the past $1$ Gyr. The same techniques used to probe the SFH in massive starbursts should also work to probe the recent SFH in the putative mini-bursts which we seek to study. By comparing measured indicators of recent SF (e.g. $EW(H\\delta_{abs})$, $4000~{\\rm \\AA}$ break strength, A star content), to the same indicators derived from the EPS models, we will place limits on the mini-burst mass and duration. We have obtained spatially integrated spectra of a sample of 40 late type spiral galaxies (Sab-Sbc) of varying degrees of lopsidedness with the intent of using their relative stellar populations (as determined from stellar template fitting and EPS models) to determine their recent SF histories. Unlike the mass-normalized blue light excess, $\\Delta B$ used in ZR97, our method operates independently of assumptions about a galaxy's mass, inclination or luminosity. In addition to probing the recent ($\\leq 1$ Gyr) SFH with studies of the stellar continuum we probe the current SFR by measuring the integrated Balmer line emission strengths (e.g. Kennicutt \\etal 1994). The layout of the paper is as follows. In \\S2 we discuss the sample selection, observations, data reduction and determination of galaxy lopsidedness; In \\S3 we examine our methods for determining the current SFR and recent SFH via the measurement of emission and stellar continuum properties as a function of lopsidedness. The discussion of the significance of these results, including the correlation of the boost parameters with other galaxy characteristics and the impact of our results on previous works (i.e. RZ95,ZR97 \\& RR98) is contained in \\S4. In \\S5 we present a summary and possible directions for future work. ", "conclusions": "To quantify the correlation between the recent SF histories of present-day spiral galaxies and their global asymmetry, we compare the integrated spectral properties of late-type spirals of varying lopsidedness. We find that the recent ($\\leq 0.5$ Gyr) SFH and current ($\\leq 10^7$ years) SFR are both strongly correlated with $\\langle \\tilde{A}_1 \\rangle$ although there is appreciable scatter in the individual galaxy-to-galaxy properties. For $EW(H\\beta_{em})$, reflecting the current SFR, we find a $3.2\\sigma$ Spearman-rank correlation with $\\langle \\tilde{A}_1 \\rangle$. We fit a combination of A0V and G0III stellar spectra to our galaxy spectra to quantify the relative abundance of A-stars in the disk (which traces the SFR within 0.5 Gyrs). From these best fit model spectra, $I_{model}^{best} (\\lambda)$, we measure a number of spectral indices, and find that $EW_{mod}(H\\delta_{abs})$, $D_{4000}$, and $C_{A0V}$ are correlated with $\\langle \\tilde{A}_1 \\rangle$ at the $3.9\\sigma$, $3.0\\sigma$, and $4.2\\sigma$ levels, respectively. We measure the same spectral indices in the nucleus, and find them less correlated with $\\langle \\tilde{A}_1 \\rangle$ (except $D_{4000}$). Unless a nuclear starburst is obscured, the disk and not the nucleus is the primary site of the SF increase we see in lopsided galaxies. This is in contrast to numerical simulations where minor mergers funnel gas into the nucleus of galaxies, causing intense starbursts (Mihos \\& Hernquist 1994; Hernquist \\& Mihos 1995). Only by the presence of a dense bulge can the formation of a bar, and the subsequent funneling of gas, be prevented. To quantify the mass of additional stars formed in lopsided galaxies, we defined a boost vector in $EW_{mod}(H\\delta_{abs})$ vs. $D_{4000}$ space, by comparing the median values of these properties for the most symmetric third and the most lopsided third of our sample. We find $\\Delta EW_{mod}(H\\delta_{abs}) = 2.1\\pm 1.0~{\\rm \\AA}$ and $\\Delta D_{4000} = 0.024\\pm 0.01$. We fit this vector with an ``underlying population $+$ boost'' EPS model corresponding to a progenitor galaxy with $b=0.33$, $\\tau_{boost}=500$ Myr, and boost age of $0.5$ Gyr. Using this best fit EPS model, we find that $\\sim 1 \\times 10^9 M_{\\odot}$ is formed in the boost in addition to the ``underlying'' SFH (assuming a final stellar mass of $10^{10} M_{\\odot}$). This is a considerable fraction ($\\sim 10\\%$) of the final stellar mass of the galaxy and corresponds to a factor of 8 increase in the SFR over the past $5 \\times 10^8$ years. Given the increasing merger rates and increasing gas fractions towards higher redshifts, minor merger induced SF boosts of short duration played an important role in assembling the present day stellar content of galaxies. Finally, we address by how much the frequency of lopsidedness from a magnitude limited sample is increased by the corresponding luminosity boost. Our best fit EPS boost model corresponds to a $\\sim 1$ magnitude brightening when galaxies becomes lopsided, increasing their presence four-fold in magnitude limited samples. We lack the statistics however, to examine any Hubble type dependent differences in the luminosity boost. It is obvious that more work needs to be done to fully understand the cause of lopsidedness as well as the SFH of lopsided galaxies. To quantify the Hubble type specific boost in the recent SFH, a large sample should be obtained with significant numbers of galaxies in each Hubble type bin. Since imaging and spectroscopy will be needed for this project, a volume limited sample may be constructed which bypasses many of the problems encountered when selecting galaxies according to an apparent magnitude limit. Companion searches to sufficiently faint magnitudes will help to study the possible link between environment and lopsidedness (as caused by weak tidal interactions). With the recent commissioning of large area imaging and spectroscopy surveys such as Sloan Digital Sky Survey, constructing such a sample will become relatively straightforward. Numerical simulations have shown to be a useful tool in studying the evolution of the stellar and gas distributions in minor mergers. High resolution simulations with a live halo are crucial for studying the detailed response of the disk to the merger (Walker \\etal 1996). A thorough exploration of interaction parameter space is needed to quantify the structural and kinematic response in the stellar and gas components. High resolution N-body studies are also needed to explore the global stability of isolated galactic disks." }, "0003/astro-ph0003345_arXiv.txt": { "abstract": "The energy source of the giant HI supershells in nearby galaxies and in our own is still an unsettled issue. Proposed scenarios for production of near-complete ringlike supershells are multiple supernovae (SNe) and $\\gamma$-ray bursts (GRBs). In the late phase of evolution it is difficult to tell these models apart. We show that, if a supershell has been powered by multiple SNe, peculiar metal abundances should be found in the medium within the bubble. We present line diagnostics that could detect such unusual abundances, such as unusually high ratios of [O]/[Fe] and [Ne]/[Fe]. Among ions of the same element, a higher enhancement is expected in lines from a high-ionization state than in lines from a low-ionization state. Identification of the energy source of HI supershells would set strong constraints on the rates and energetics of GRBs, as well as on their location within a galaxy. ", "introduction": "For several decades, 21 cm surveys of spiral galaxies have revealed the puzzling existence of expanding giant HI supershells (see e.g. Tenorio-Tagle \\& Bodenheimer 1988 for a review). These nearly spherical structures have very low density in their interiors and high HI density at their boundaries, and they expand at velocities of several tens of ${\\rm km~{s}^{-1}}$. The radii of these shells are much larger than those of ordinary supernova remnants and often exceed $\\sim 1$ kpc; their ages are typically in the range of $10^6$--$10^8$ years. Heiles (1979) denominated as supershells the ones whose inferred kinetic energies are $\\ga 3\\times 10^{52}$ ergs. The Milky Way contains several tens of them (Heiles 1979; Heiles, Reach, \\& Koo 1996), and in one case the estimated kinetic energy is as high as $\\sim 10^{54}$ ergs. Similar supershells are also observed in other nearby galaxies. Whereas it is clear that these HI supershells result from deposition of an enormous amount of energy in the interstellar medium, the energy source is still a subject of debate. Collisions with high-velocity clouds (Tenorio-Tagle 1981) could account for those cases where only one hemisphere is present, and the required input energy is not too large. However, it is unclear how such collisions could produce the near-complete ringlike appearance observed in some cases (Rand \\& van der Hulst 1993). Small shells of radii $\\sim 200$--400 pc and energies $\\la 3\\times 10^{52}$ ergs are often explained as a consequence of the collective action of stellar winds and supernova explosions originating from OB star associations (McCray \\& Kafatos 1987; Shull \\& Saken 1995). The winds from the stars of the association create a bubble in the interstellar medium (ISM) that is filled with hot gas. The bubble further grows when the stars explode as supernovae, releasing their energy into the ISM. Multiple SN explosions are in principle a viable scenario even for the largest supershells, although this would require very large OB associations, not typically observed in nearby galaxies (Kennicutt, Edgar \\& Hodge 1989). Another possibility that has been put forward is that giant supershells could be the remnants of gamma-ray bursts (GRBs) (Efremov, Elmegreen \\& Hodge 1998; Loeb \\& Perna 1998). In fact, if GRBs occur in galaxies and can have energies $\\ga 10^{53}$ ergs, then remnants in the form of giant bubbles are unavoidable. Notice, however, that this conclusion relies on the assumption that the ratio of $\\gamma$-ray energy to kinetic energy of the ejecta is very small, as required by the popular 'internal shock' models for GRBs. If, on the other hand, this were not the case, as the analysis of GRB 970508 seems to imply, then the kinetic energy of GRBs would not be sufficient to produce a giant remnant (Paczy\\'nski 1999). The nature of the energy source can be more easily identified in young supershells. The ones due to multiple SNe would still show ongoing activity. Bubbles powered by a GRB explosion could instead be identified by signatures of the radiation emitted by the cooling gas, which had been heated and ionized by the GRB afterglow (Perna, Raymond \\& Loeb 2000). However, after a time $t\\ga 10^5$ yr, the imprints of this radiation have faded away. Old supershells remain, therefore, the most difficult to understand\\footnote{Among the observed supershells only about 10\\% of them seem to contain OB associations and could therefore be more naturally attributed to multiple SNe.}. However, given their ages, they are by far the most abundant in galaxies. An attempt to identify their energy source has been recently made by Rhode et al. (1999). Assuming that the HI holes are created by multiple SNe, and that the SNe represent the high-mass population (OB stars) of a cluster with a normal initial mass function, they observed that the upper main-sequence stars (late B, A and F) should still be present in the cluster. However, their observations showed that in several of the holes the observed upper limits for the remnant cluster brightness are inconsistent with the expected values. Therefore their test suggested problems with the multiple SNe scenario. On the other hand, no evidence that the holes could be due to GRBs was found either. More recently, Efremov, Ehlerova \\& Palous (1999) discussed possible differences between the structures produced by a GRB and by an OB association, based on their shapes, expansion velocities, and fragmentation times. Here we propose a new way of testing the multiple SNe versus GRB model to power supershells. This is based on the fact that SNe inject metals in the ISM in which they explode. As a result, if a supershell has been powered by multiple SNe, the abundances of some specific metals in its interior should be enhanced with respect to the typical values in the ISM surrounding the shell\\footnote{This is commonly observed in young supernova remnants (e.g. Canizares \\& Winkler 1981).}. As the high-mass stars which power the supershell explode as Type II SNe, the enhancement should be particularly pronounced in elements such as Oxygen, Silicon, Neon, Magnesium, but not in others (e.g. Nomoto et al. 1997). We present line diagnostics that could help detect such unusual abundances. If a supershell has been powered by a GRB, on the other hand, no peculiar metal enhancement is expected. The highly relativistic expansion of the ejecta requires that the baryonic load be very small \\footnote{Even if GRBs were associated with SNe (as it has been suggested in the case of SN 1998bw [Galama et al. 1998]), and there were some mass ejected at later times, it would be just that of a single SN, and therefore it would be highly diluted within the large volume of the supershell.}($M\\la 10^{-4}M_\\odot$). Therefore, detection of peculiar abundances in the medium within a supershell could provide a clue to was the energy source that powered it. Knowledge of the fraction of HI supershells that is likely to be associated to a GRB event would lead to important constraints on the energetics and rates of GRBs, as well as on their location within a galaxy. ", "conclusions": "The energy source which powers giant HI supershells is still a subject of debate. Its identification is particularly difficult in the late phases of evolution of the remnant. While hemispherical supershells could be perhaps attributed to collisions with high-velocity clouds, the near-complete ringlike ones could be more easily explained by either multiple SNe from an OB association or by a GRB. In this paper we have identified signatures that could help discriminate between the two models. Namely, we have shown that supershells powered by multiple SNe are likely to show enhanced abundances of the metals produced by the SNe themselves, and we have proposed some line diagnostics that could help reveal these unusual features. Being able to discriminate between the multiple SNe and the GRB scenario for the production of HI supershells would help constrain GRB rates and energetics, as well as their location within a galaxy." }, "0003/astro-ph0003173_arXiv.txt": { "abstract": "High resolution near-infrared adaptive optics imaging and polarimetry have been obtained of the Homunculus Nebula and its central illuminating star $\\eta$ Carinae (HD 93308). Images and maps of the linear polarization at a resolution of $<$0.2$''$ have been made in the J, H and K bands and in a narrow 2.15$\\mu$m continuum band (K$_c$). The overall appearance of the nebula in the near-IR is similar in all bands and to that at V, with specific features less sharp to longer wavelengths. A comparison between the published HST WFPC2 1.042$\\mu$m map and the AO J band image demonstrates that ground-based AO resolution can approach that of HST. The large-scale pattern of polarization vectors is centro-symmetric demonstrating that single scattering dominates everywhere except perhaps in the central bright core. The only difference between the near-IR and optical appearance is a narrow linear feature at position angle 320$^\\circ$ extending across the NW lobe of the Homunculus. Polarization maps at K$_c$ in the near vicinity of $\\eta$ Carinae were restored using a PSF derived from blind deconvolution. There is a definitely detected extension in the direction of the optically visible speckle knots and an estimate of 18\\% for the polarization of one of the knots was made. This level of polarization suggests that the knots arise in dust+gas clouds in the near environment of $\\eta$ Carinae, perhaps in an equatorial disc. The most remarkable result of the linear polarization mapping is the level of similarity in the spatial structure, and also in the degree of linear polarization, between the near-IR and optical data. Comparison of the polarization along the projected major axis of the Homunculus shows values in the SE lobe to within a few percent from V band to 2.2$\\mu$m. In the NW lobe the near-IR linear polarization values agree to within a few percent over the 1.2-2.2$\\mu$m range, but are upto 10\\% lower than at V. Such a polarization pattern cannot arise in Mie scattering from a single power law distribution of grain sizes unless the particles are very small compared with the wavelength, in disagreement with mid-IR observations. In addition the colour dependence of the extinction was found to be shallower than the typical ISM, indicating the presence of large grains. Several possibilities are explored to try to explain these contradictory results. Optical depth effects and a broad distribution in grain sizes are favoured, perhaps with a dependence of the grain size with depth into the small obscuring clouds in the lobes of the Homunculus. However the presence of aligned grains, previously inferred from mid-IR polarization, may also affect the scattered radiation from the dust. ", "introduction": "$\\eta$ Carinae, situated in the Carina complex at about 2.3kpc (Davidson \\& Humphreys \\cite{dahu97}), is one of the most massive stars known in the Galaxy and is going through the Luminous Blue Variable phase (Humphreys \\& Davidson \\cite{huda94}) of unsteady mass loss (Davidson et al. \\cite{da86}). During the 1840's it underwent an outburst and reached visual magnitude -1; since then is has been emerging from the dust which condensed after this ejection (Walborn \\& Liller, \\cite{wali}). Long term monitoring of optical, IR, radio and X-ray spectra has revealed evidence of periodicity perhaps related to a binary or multiple star at the core of the nebula (Daminelli et al. \\cite{dam97}). The compact nebula around $\\eta$ Carinae (HD 93308), called the Homunculus, was first shown by Thackeray (\\cite{tha56}) to be highly polarized. The initial polarimetry was confirmed by Wesselink (\\cite{wess}) who measured linear polarization of around 40\\%. Visvanathan (\\cite{vis}) observed that the polarization centred on $\\eta$ Carinae was almost constant with wavelength from U to R and increased with increasing aperture size. In a small aperture, higher polarization was observed on the NW side of the nebula than on the SE. The first systematic polarization maps were made by Warren-Smith et al. (\\cite{warr}) in the V band and demonstrated a centro-symmetric pattern of polarization vectors with a marked asymmetry in the polarization values along the major axis (position angle $\\sim$130$^\\circ$) with values upto 40\\% in the NW lobe. To produce such high values of polarization in a reflection nebula, Mie scattering by silicate particles with a size distribution weighted to smaller particles was invoked and modelled by Carty et al. (\\cite{cart}). In the near ($\\leq$0.5$''$) vicinity of $\\eta$ Car itself, speckle masking polarimetry in the H$\\alpha$ line and local continuum has revealed evidence for a compact equatorial disc aligned with the minor axis of the Homunculus (Falcke et al. \\cite{falc}). Within $<$1$''$ of $\\eta$ Car the polarization vector pattern does not remain centrosymmetric in the R band, suggesting that local structures and perhaps intrinsic emission may contribute to the morphology and scattered light (Falcke et al. \\cite{falc}). Polarimetry in the mid-infrared, where the dust emits rather than scatters radiation, shows an entirely different pattern of polarization vectors with a trend to be oriented radially, particularly near the boundary of the emission (Aitken et al. \\cite{ait95}). Such a pattern can be interpreted in terms of emission from aligned grains; Aitken et al. (\\cite{ait95}) suggest that the alignment mechanism may be gas-grain streaming, driven by the high outflow velocity, or a remnant magnetic field from a dense magnetized disc. There is a wealth of IR observations of $\\eta$ Car and the Homunculus on account of its intrinsic IR brightness, first observed by Westphal \\& Neugebauer (\\cite{wene}), and astrophysical interest. The IR spectrum is characterized by a peak around 10$\\mu$m, indicative of silicate grains (Mitchell \\& Robinson, \\cite{miro}). There is a central IR point source together with a second peak on the minor axis of the nebula, whose separation increases from 1.1 to 2.2$''$ from 3.6 to 11.2$\\mu$m (Hyland et al. \\cite{hyl}). The near-IR spectrum of $\\eta$ Car shows a steep increase with wavelength and prominent hydrogen lines of the Paschen and Brackett series as well as He~I lines (Whitelock et al. \\cite{whi}) and weaker Fe~II and [Fe~II] lines (Altamore et al. \\cite{alta}). Maps in the J, H and K bands show that the structure is dominated by scattering, but beyond about 3$\\mu$m dust emission dominates (Allen \\cite{all89}), with many clumps present. High spatial resolution observations have reported an unresolved central source (at L and M band, Bensammar et al. \\cite{bens}), with filaments and unresolved knots within 1$''$ detected in many IR bands (Gehring \\cite{gehr}). Maps in the mid-IR show a similar structure and the compact central source has a dust temperature $\\sim$650K and dust mass of 10$^{-4}$M$_\\odot$ with nearby dusty clouds associated into loop features (Smith et al. \\cite{smai}). This source has been so prodigiously studied at so many wavelengths that it possesses its own review article in Annual Reviews of Astronomy and Astrophysics (Davidson \\& Humphries \\cite{dahu97}). $\\eta$ Car can be considered an ideal source for adaptive optics on account of its very bright central, almost point, source (V$\\sim$7mag. - van Genderen et al. \\cite{vang}) and the limited radial extent ($\\pm\\leq$10$''$) of the Homunculus, which means that the source itself can be used as a reference star for the wavefront sensor. As a consequence, off-axis anisoplaniticity does not significantly affect the adaptive optics (AO) correction out to the edges of the nebula. Previous near-infrared AO imaging of $\\eta$ Car was obtained (Rigaut \\& Gehring \\cite{rige}), including some limited polarimetry (Gehring \\cite{gehr}) using the COME-ON AO instrument. Here we report on dedicated high resolution near-IR AO imaging polarimetry conducted at J, H, K, and in a continuum band at 2.15$\\mu$m, using the ADONIS system and SHARP~II camera with the aim of studying the small-scale polarization structure of the Homunculus. The observations are described in Sect. 2; the reductions and polarization data are presented in Sect. 3 and the relevance of the results for the structure and dust properties of this remarkable reflection nebula are discussed in Sect. 4. ", "conclusions": "The first high spatial resolution adaptive optics near-IR polarization maps of $\\eta$ Carinae and the Homunculus nebula have been presented. Since the Homunculus is dominated by scattering then the appearance in the near-IR is rather similar to that observed in the optical and a comparison of the AO results with an HST 1.04$\\mu$m WFPC2 image was presented showing essentially the same features. The most important single result from this work is the overall similarity of the linear polarization from the V band to 2.2$\\mu$m within a few percent for the SE lobe and the lower values at J and H compared with V for the NW lobe. Image restoration was applied to a set of 2.15$\\mu$m continuum images to determine the polarization distribution in the near vicinity of $\\eta$ Carinae. A tentative value of the polarization of the Weigelt et al. speckle knot D of 18\\% was determined suggesting that it is a dust cloud within the equatorial disc strongly illuminated by $\\eta$ Carinae. Various models are discussed in order to explain the flat distribution of polarization with wavelength over the Homunculus. A possible association of a narrow feature within the NW lobe of the Homunculus with one of the highly collimated emission line `whiskers' outside the nebula deserves further investigation." }, "0003/astro-ph0003203_arXiv.txt": { "abstract": "We have imaged a $45\\arcmin \\times 45\\arcmin$ area centered on the Hubble Deep Field \\mbox{(HDF)} in $UBVRI$ passbands, down to respective limiting magnitudes of approximately $21.5, 22.5, 22.2, 22.2, {\\rm and \\ } 21.2$. The principal goals of the survey are to identify QSOs and to map structure traced by luminous galaxies and QSO absorption line systems in a wide volume containing the HDF. The area surveyed is $400$ times that of the HDF, and $40$ times that of the HDF Flanking Fields. We have selected QSO candidates from color space, and identified 4 QSOs and 2 narrow emission-line galaxies (NELGs) which have not previously been discovered, bringing the total number of known QSOs in the area to $19$. The bright $z=1.305$ QSO only 12\\arcmin \\ away from the HDF raises the northern HDF to nearly the same status as the HDF-S, which was selected to be proximate to a bright QSO. About half of the QSO candidates remain for spectroscopic verification. Absorption line spectroscopy has been obtained for $3$ bright QSOs in the field, using the Keck 10m, ARC 3.5m, and MDM 2.4m telescopes. Five heavy-element absorption line systems have been identified, 4 of which overlap the well-explored redshift range covered by deep galaxy redshift surveys towards the HDF. The two absorbers at $z=0.5565$ and $z=0.5621$ occur at the same redshift as the second most populated redshift peak in the galaxy distribution, but each is more than $7h^{-1}$Mpc (comoving, $\\Omega_{m}=1$, $\\Omega_{\\Lambda}=0$) away from the HDF line of sight in the transverse dimension. This supports more indirect evidence that the galaxy redshift peaks are contained within large sheet-like structures which traverse the HDF, and may be precursors to large-scale ``pancake'' structures seen in the present-day galaxy distribution. ", "introduction": "} Deep galaxy redshift samples are permitting a new and often surprising view of the Universe at much younger epochs, and into which the role of gas, both hydrogen and processed, via QSO absorption line systems can be incorporated. Only recently, and with the help of the 10-m Keck telescopes, have deep galaxy redshift surveys been able to measure properties of galaxies at some of the redshifts ($2.5 \\la z \\la 4.5$) which have been easily accessible to absorption line studies for over three decades. Combining the study of QSO absorbers and galaxy surveys has the potential to greatly enhance our understanding of the formation and evolution galaxies as well as the large-scale structures which typically contain them. For example, even if galaxies and absorbers are closely related, biasing, which plays an important role in deciphering structure formation, is expected to be different for for galaxies, QSOs, absorbers, and the various classes of each \\citep[\\eg][]{demia1999,cen1998,fang1998, quashnock1998,bi1996}. A generic result of the deep galaxy pencil-beam surveys is that half or more of the galaxies measured tend to lie in very narrow redshift ``spikes'' which are present to redshifts of at least $z=1$ \\citep{cohen1996a,cohen1996b} and are often found at much higher redshifts ($z\\approx 3$) in the ``dropout'' surveys \\citep{steidel1998,adelberger1998}. The number density, redshift spacing, density enhancements, velocity dispersions, and morphological mixtures, all support the hypothesis that these structures in redshift space are parts of the precursors to present-day galaxy superclusters and walls \\citep{cohen1996a, cohen1996b}. This evidence is mostly circumstantial so far, since the deep pencil-beam surveys cover only very small (typically 50 sq.\\ arcmin.\\ or less) disjoint areas of the sky. Additional but shallower redshift surveys have been carried out in narrow fields adjacent to at least one deep pencil beam survey, which have supported the the idea that the redshift structures are coherent in the transverse spatial dimension on scales up to at least a degree, and for redshifts up to at least $z\\approx 0.4$ \\citep{cohen1999}. Extending this type of survey to deeper redshifts is difficult not only due to the faintness of the galaxies, but in the redshift range $1.2 \\la z \\la 2$ there is a lack of redshifted galaxy spectral features available at optical wavelengths. It is a highly desirable but currently difficult goal of future redshift surveys to cover both larger areas and a more complete redshift range. QSO absorption line systems offer a means of efficiently extending these studies to wider volumes and higher redshift, which is the aim of the program described here. The approach is to search for intervening absorption line systems in the spectra of QSOs at small angular separations. The selection function for heavy-element QSO absorption line systems, identified mainly by \\civ $\\lambda$1550{\\AA} and \\mgii $\\lambda$2799{\\AA} doublet transitions, is luminosity independent, and limited at high redshift only by the emission redshift of the backlighting QSOs. In optical spectra \\mgii lines can be detected from redshifts of $z \\approx 0.15 - 2.0$ and \\civ lines from $z\\approx 1$ to over 4. Absorption surveys towards groups of QSO sightlines have been successfully used to trace structure in three dimensions at high redshift \\citep{crotts1985,crotts1989, jakobsen1992,foltz1993,elowitz1995,dinshaw1996,williger1996,vandenberk1999, impey1999}. A few QSOs have also been observed directly within the areas covered by the galaxy surveys, and their spectra have revealed absorption line systems that very often lie within the redshift peaks defined by the galaxies \\citep{steidel1998}. These studies have demonstrated the utility of absorption line systems in probing large-scale structure both in radial and angular dimensions, and of using large-scale structure studies to decipher the relationship between galaxies and absorbing gas. QSOs bright enough to use for 3-dimensional absorption line studies generally have a high enough angular density so that suitable groups can be found in virtually any field of sufficiently high galactic latitude. For example, most UVX QSO surveys reveal a density of about 30 QSOs per sq.\\ degree to a limiting magnitude of $B\\le 21$ \\citep*[\\eg][]{zhan1989}, which is a practical limit for absorption line surveys with 4-m class telescopes. To take full advantage of this technique, one should select fields in which deep galaxy redshift surveys have also taken place. The galaxy and absorber surveys are then complementary: the galaxies provide the redshift locations and velocity dispersions of structures, while the absorbers can be used to quickly and efficiently widen the survey to larger areas and additional redshift ranges, and probe the otherwise invisible structure of the gas. The Hubble Deep Field \\citep[HDF;][]{williams1996} is the site of one of the most complete and comprehensive sets of deep redshift surveys, with over $300$ measured redshifts in an area of only $\\approx 50$ sq.\\ arcmin.\\ \\citep{cohen1996a,steidel1996,lowenthal1997,guzman1997, phillips1997,hogg1998}. The measured redshifts lie in the range $z \\la 1.3$ and $z \\ga 2.0$, with a gap between $1.3$ and $2.0$ due to restrictions of optical spectroscopy. We have chosen the area surrounding the Hubble Deep Field for our initial QSO/absorber study because of the large and continuing amount of research devoted to this sightline, and because it is easily accessible not only by northern-hemisphere telescopes, but also to the Hubble Space Telescope (HST) which can be used for follow-up observations of the low-redshift \\lya systems. Indeed, this latter approach is the primary justification for the construction of the Hubble Deep Field South, and a survey similar to ours for additional QSOs in that direction of the sky is currently taking place \\citep{teplitz1998}. \\citet[hereafter LPIF]{lpif} recently carried out a QSO survey in the one square degree surrounding the HDF, and found 30 QSOs brighter than $B=21$. While the LPIF survey and ours have similar goals and survey depths, ours uses 5-band photometry (LPIF used only $U, B, {\\rm and \\ } R$ bands) to search for high-redshift QSOs, and we have started QSO absorption line follow-up spectroscopy. Comparisons of the two surveys will be made when appropriate. In this paper we present our initial results on the QSO survey towards the HDF (there are no reasonably bright QSOs inside the HDF itself), and our preliminary absorption line study of 3 of the QSO lines-of-sight. The imaging observations and photometry, QSO candidate selection and verification, and QSO absorption spectroscopy, are presented in \\S\\,\\ref{imandph}, \\S\\,\\ref{candsec}, and \\S\\,\\ref{qsoals} respectively. We discuss the distributions of the QSOs and absorbers relative to the galaxy redshift sample in \\S\\,\\ref{discsec}. A summary is given in \\S\\,\\ref{summary}. ", "conclusions": "} We have begun a survey to identify QSOs and absorption line systems in a $45 \\times 45$ square arcmin area surrounding the Hubble Deep Field. So far $19$ QSOs have been identified within our survey area to a limiting magnitude of $B\\sim21$, and over 30 UVX and high-redshift QSO candidates remain. We have obtained absorption line spectra for three of the brighter QSOs in the field, which have revealed at least 5 heavy-element absorption line systems. Of the four systems that overlap the redshift range explored in deep galaxy redshift surveys of the HDF, two lie at or very near one of the strongest redshift peaks in the galaxy distribution. If the absorbers and galaxies in the peak are part of the same structure, it extends at least $7h^{-1}$Mpc ($\\Omega_{m}=1$, $\\Omega_{\\Lambda}=0$) in the transverse direction at a redshift of $z\\approx 0.56$. This supports earlier evidence from the galaxies alone that the peaks in the galaxy distribution are parts of larger structures, which may be the precursors to present-day superclusters or walls." }, "0003/gr-qc0003072_arXiv.txt": { "abstract": "{We present a spectral method for solving elliptic equations which arise in general relativity, namely three-dimensional scalar Poisson equations, as well as generalized vectorial Poisson equations of the type $\\Delta \\vec{N} + \\lambda \\vec{\\nabla}(\\vec{\\nabla}\\cdot \\vec{N}) = \\vec{S}$ with $\\lambda \\not= -1$. The source can extend in all the Euclidean space ${\\bf R}^3$, provided it decays at least as $r^{-3}$. A multi-domain approach is used, along with spherical coordinates $(r,\\theta,\\phi)$. In each domain, Chebyshev polynomials (in $r$ or $1/r$) and spherical harmonics (in $\\theta$ and $\\phi$) expansions are used. If the source decays as $r^{-k}$ the error of the numerical solution is shown to decrease at least as $N^{-2(k-2)}$, where $N$ is the number of Chebyshev coefficients. The error is even evanescent, i.e. decreases as $\\exp(-N)$, if the source does not contain any spherical harmonics of index $l\\geq k -3$ (scalar case) or $l\\geq k-5$ (vectorial case).} \\keywords {scalar and vectorial Poisson equation; spectral methods; Gibbs phenomenon; general relativity} \\begin{article} ", "introduction": " ", "conclusions": "We have presented a scalar Poisson equation solver based on spectral method. It enables us to solve the Poisson equation for a source extending to infinity and going to zero at least like $r^{-3}$. Our multi-domain approach enables to deal with a source which is $\\cont{\\infty}$ in each domain. Nevertheless some Gibbs phenomenon can appear due to the existence of particular solutions which contain logarithm functions in the external domain. Such functions are not well described in terms of Chebyshev polynomials, resulting in a Gibbs-like phenomenon. We exhibited the conditions for the appearance of such an effect and quantified it, leading to the conclusion that, for a source decaying as $r^{-k}$ ($k \\geq 3$), the error of the numerical solution is evanescent if the source does not contain any spherical harmonics with index $l\\geq k-3$. Otherwise, the error decreases at least as $N^{-2(k-2)}$, $N$ being the number of Chebyshev coefficients. We used this scalar Poisson equation solver to solve the generalized vectorial Poisson equation given by Eq. (\\ref{e:poisson_vect}) for a source going to zero at least like $r^{-4}$. Three different schemes have been discussed. We showed than the one proposed by Bowen \\& York \\cite{BowenY80} is not applicable to domains extending up to infinity, by means of our methods, because it gives rise to diverging auxiliary quantities. The scheme from Oohara \\& Nakamura \\cite{OoharN97} is applicable as long as the source is continuous and has been successfully implemented. The last scheme, proposed by Oohara, Nakamura and Shibata \\cite{OoharNS97}, is applicable even for discontinuous sources and has been successfully implemented too. The convergence properties of the two implemented schemes have been derived from the ones of the scalar Poisson equation solver and checked by comparison between calculated and analytical solutions." }, "0003/astro-ph0003035_arXiv.txt": { "abstract": "Near-infrared spectroscopic observations of a sample of very cool, low-mass objects are presented with higher spectral resolution than in any previous studies. Six of the objects are L-dwarfs, ranging in spectral class from L2 to L8/9, and the seventh is a methane or T-dwarf. These new observations were obtained during commissioning of NIRSPEC, the first high-resolution near-infrared cryogenic spectrograph for the Keck II 10-meter telescope on Mauna Kea, Hawaii. Spectra with a resolving power R$\\approx$2500 from 1.135 to 1.360\\mic~(approximately J-band) are presented for each source. At this resolution, a rich spectral structure is revealed, much of which is due to blending of unresolved molecular transitions. Strong lines due to neutral potassium (K I), and bands due to iron hydride (FeH) and steam (H$_{2}$O) change significantly throughout the L sequence. Iron hydride disappears between L5 and L8, the steam bands deepen and the K I lines gradually become weaker but wider due to pressure broadening. An unidentified feature occurs at 1.22\\mic~ which has a temperature dependence like FeH but has no counterpart in the available FeH opacity data. Because these objects are 3-6 magnitudes brighter in the near-infrared compared to the I-band, spectral classification is efficient. One of the objects studied (2MASSW J1523+3014) is the coolest L-dwarf discovered so far by the 2-Micron All-Sky Survey (2MASS), but its spectrum is still significantly different from the methane-dominated objects such as Gl229B or SDSS 1624+0029. ", "introduction": "After eluding undisputed detection for many years, numerous brown dwarfs -- objects with sub-stellar mass -- are now known. While some candidates were discovered in small-scale surveys of young nearby clusters, such as the Pleiades and Hyades, or as companions to low-mass stars, the biggest breakthrough has come as a result of large-scale surveys such as the Deep Near-Infrared Sky (DENIS) survey (Delfosse \\et 1997), 2MASS, the 2-Micron All-Sky Survey (Skrutskie \\et 1997, Kirkpatrick \\et 1999) and the Sloan Digital Sky Survey (SDSS) (Strauss \\et 1999). Recently, using optical (CCD) spectroscopy, Kirkpatrick \\et (1999) have defined a new spectral class, L-dwarfs, in which the metallic oxides (such as TiO and VO) found in M stars lose their dominance to metallic hydrides (such as FeH and CrH). The temperature range for the L class is given by Kirkpatrick \\et (2000) from about 2000 K for L0 to about 1250 K for L8, whereas Martin \\et (1999) suggest a range from 2200-1600 K. Depending on age and model calculations, Kirkpatrick \\et (1999) argue that at least one third of the L-dwarf objects must be brown dwarfs and perhaps all are. Spectral classification is based on spectroscopy between 6500 and 10000 \\AA~ at a resolution of 9\\AA~ (R $\\sim$1000). While use of this spectral region provides many important spectral diagnostics, it suffers from the fact that L- and T-dwarfs are extremely faint at these wavelengths and therefore long exposures on very large aperture telescopes are required to obtain spectra with good signal-to-noise ratios. Typical I-band magnitudes are about 19 or fainter (\\eg~ GD165B), but a gain of 3-6 magnitudes can be obtained in going to the near infrared. Using earlier generations of infrared instruments, previous observations of individual brown dwarf candidates have yielded a typical resolving power of about R=500-1000; see the observations of Geballe \\et (1996), Ruiz, Leggett and Allard (1997), Tinney, Delfosse and Forveille (1997), Kirkpatrick \\et (1999) and Strauss \\et (1999). These pioneering efforts were accomplished with instruments using an earlier generation of IR detector arrays, with at most 256 x 256 pixels. This resolution is sufficient to reveal the major differences that set apart the L-dwarfs and T-dwarfs from warmer stars, \\eg, the presence of deep steam bands and strong methane bands in the L- and T-dwarfs respectively. Kirkpatrick \\et (1993) modeled a spectral sequence of M-dwarfs using spectroscopy from 0.6 - 1.5 microns and identified the major bands and atomic features. Jones \\et (1996) performed a similar analysis from 1.16 - 1.22 microns with a sample of M dwarfs which also included GD165B. An excellent review of model atmospheres of very low mass stars and brown dwarfs is given by Allard \\et (1997). In this paper we report observations using NIRSPEC, a new cryogenic infrared spectrograph on the Keck II telescope employing a 1024 x 1024 InSb array. A consistent set of J-band spectra with R$\\sim$2500 is presented which, for the first time, allows a detailed comparison of the near-infrared features of the spectral sequence from early L-dwarfs to T-dwarfs. Our targets were selected from the list of L-dwarfs published by Kirkpatrick \\et (1999) and supplemented with new sources discovered more recently by the 2MASS (Kirkpatrick \\et, 2000). One of these objects is reported as being the closest known L-dwarf to date and another is likely the coolest L-dwarf discovered thus far. ", "conclusions": "The variation in the J-band spectra of this sample of objects is quite remarkable and it is relatively easy to place the objects in a temperature sequence. The water band strengthens as the temperature decreases. FeH weakens and then disappears, the K I lines weaken and broaden and the continuum around 1.15\\mic~ slowly drops relative to the continuum at 1.26\\mic. As expected, the DENIS L5 source and the 2MASS L5 object exhibit almost identical spectral characteristics. By ordering the spectra according to the classifications given by Kirkpatrick \\et (1999, 2000), with the earliest spectral type (L2) at the top, the following trends are apparent in the J-band spectra.\\\\ {\\it L2 (Kelu-1)}: strong K I and FeH lines are superimposed on a larger depression across the region, which is perhaps the result of residual oxide (either TiO and/or VO) absorption; VO is expected around 1.19\\mic.\\\\ {\\it L4 (GD165B)}: any residual oxide absorption has gone, effectively raising the continuum to produce a flatter spectrum, and making the K I and FeH features appear stronger although they are expected to decrease with decreasing temperature. The water absorption (steam) band at 1.30\\mic~ is increasing in strength. Numerous small features from 1.25--1.30\\mic~ closely match those in Kelu-1.\\\\ {\\it L5 (2MASSW J1507-1627)}: all features present in the L4 class remain. The K I and FeH features are very slightly weaker, while the water band at 1.30 microns is deeper than before and there is a slight slope of the continuum towards the blue end.\\\\ {\\it L5 (DENIS-P J1228-1547)}: this spectrum is almost the same as the previous one, confirming that they are indeed the same spectral class.\\\\ {\\it L8 (2MASSW J1623+1904)}: at L8, the FeH features have disappeared and the depth of the K I lines are significantly weaker but there is evidence of broadening in their wings. There is a slight downward slope of the continuum towards the blue. The steam band is relatively stronger.\\\\ {\\it L8/9 (2MASSW J1523+3014)}: Very similar to the previous L8, but the K I lines appear slightly broader and the water band is slightly deeper. The slope to the blue is a little stronger than in L8. Consequently, this object may be cooler than 2MASSW J1623+1904 as its designation suggests, but the difference is small.\\\\ {\\it T (SDSS 1624+0029)}: A dramatic slope towards the blue appears, due to the onset of methane absorption in this wavelength region, and there is also a slope or ``break\" towards the red from about 1.26--1.31\\mic~ before a deep water band sets in. The K I lines are still present but are now very broad. As an illustration of the density of molecular features and the problem of line blending, Figure 2 shows a model spectrum kindly provided by Peter Hauschildt (private communication). This sample spectrum is based on a model atmosphere code (AMES-Dusty, Allard \\& Hauschildt. in prep), with a self-consistent treatment for dust formation. The treatment of dust is complicated however, and theorists have yet to agree on the best approach. The parameters of this model are solar metallicity, log(g)=4.5 and T$_{eff}$=2000 K and the model spectrum was smoothed from an original resolution of R=50,000. Qualitatively, the agreement with the NIRSPEC spectra is very good. Another useful framework for understanding these spectra is the molecular equilibrium calculations by Burrows and Sharp (1999). As their analysis shows, the main absorbers characteristic of M stars (\\eg TiO and VO) decline rapidly in importance with decreasing effective temperature. These molecules are expected to condense onto dust grains; TiO for instance forms perovskite (CaTiO$_{3}$). The abundance of gaseous TiO begins to decrease around 2400 K and similarly, VO will become depleted near 1800 K. For iron, the first condensate to form is the metal itself, at about 2200 K, which can then form droplets and rain out of the atmosphere. We have carefully compared our spectra to the solar atlas and cannot make any conclusive identifications with iron lines, or any other metal lines (such as Mn and Al) among the dense forest of H$_{2}$O transitions. Interestingly, Jones \\et 1996 noted the presence of Fe in earlier spectral types, such as the M6 dwarf GL406, at a comparable resolution. A significant amount of iron may have rained out. Since they are less refractory and survive in monatomic form for a greater temperature range, the neutral alkali metals (Na, K, Rb, Cs) are expected to remain after the true metals become depleted. In effect, as the temperature falls the atmospheres of cool sub-stellar objects become more transparent. The column density of potassium and sodium, for instance, is expected to increase to the point where the wings of the absorption lines become damped. This result explains the strength, broadening and temperature dependence of the K I lines seen in our spectra. According to Burrows and Sharp (1999), sodium and potassium should become depleted around 1500 - 1200 K, with sodium disappearing first and potassium forming into KCl below about 1200 K. If there is settling of refractory species however, at higher, deeper temperatures, then both atomic sodium and potassium are expected to persist to lower temperatures, at which point they should form their sulfides, not chlorides (see Burrows, Marley and Sharp 1999, and Lodders 1999). Figure 1 shows that the very strong K I lines persist, albeit with broad wings, well into the T-dwarf temperature range. Some features apparent in the new data are not yet explained by the existing models. For example, a broad, relatively strong feature is seen in our spectra at 1.22 \\mic. This feature remains through L5, but is gone in the L8 spectra. Although this is the same pattern as followed by FeH, this broad feature does not appear in the opacity plot of FeH kindly supplied by Adam Burrows (private communication), nor in the model spectrum provided by Peter Hauschildt. Finally, our results imply that any L- or T- dwarf object meeting the discovery parameters of the 2MASS and/or the SDSS can be observed spectroscopically with NIRSPEC on Keck at medium to high spectral resolution. The near-infrared region from 1.13 - 1.36 \\mic~ is quite rich in spectral features, most of which appear to be unresolved blends of molecular species, namely H$_{2}$O and FeH in the L-dwarfs and CH$_{4}$ in the T-dwarfs. Evidently, even higher spectral resolution would help to constrain the models.\\\\" }, "0003/astro-ph0003258_arXiv.txt": { "abstract": "We discuss the prompt emission of the $\\gamma$-ray burst (GRB) 990510 and its subsequent X-ray afterglow from 8.0 to 44.3\\,hrs after the prompt emission, using observations with the {\\it Gamma-ray Burst Monitor} and {\\it Narrow Field Instruments} on {\\it BeppoSAX}. In the 40--700\\,keV band, GRB990510 had a fluence of $\\sim$1.9\\,$\\times 10^{-5}$\\,erg\\,cm$^{-2}$, whereas it reached a peak flux of $\\sim$2.4\\,$\\times$\\,10$^{-6}$\\,erg\\,cm$^{-2}$\\,s$^{-1}$. The X-ray afterglow decay light curve can be satisfactorily described by a single power law with index of $-1.42 \\pm 0.07$. Both the X-ray and optical behaviour of the afterglow can be explained by $\\gamma$-ray burst debris expanding as a jet; we find that the cooling frequency is (fixed) between the optical and X-ray wavelength bands. ", "introduction": "Among the about 25 $\\gamma$-ray bursts localized by the BeppoSAX {\\it Wide Field Cameras} (WFCs), most of those followed-up with the {\\em Narrow Field Instruments} (NFIs) onboard the same satellite have exhibited afterglows at X-ray energies (e.g., Costa et al.\\ 1997), whereas less than half of them have exhibited afterglows in the optical, IR, and/or radio (e.g., van Paradijs et al.\\ 1997; Frail et al.\\ 1997). Most X-ray afterglows show a smooth power-law decay (with indices between $-$1.1 to $-$1.9), the exceptions being GRB970508 (Piro et al.\\ 1998) and GRB970828 (Yoshida et al.\\ 1999), which exhibit re-bursting events on time scales of a few hours and a day, respectively, superimposed on a power-law trend. The brightest X-ray afterglow so far, i.e.\\ that of GRB990123, provided the first detection of hard X-ray (15--60\\,keV) afterglow emission (Heise et al.\\ 2000). Here we discuss BeppoSAX observations of the prompt $\\gamma$-ray emission and the X-ray afterglow of GRB990510. On 1999 May 10 the BATSE experiment onboard the {\\it Compton Gamma Ray Observatory} (CGRO) was triggered by GRB990510 at 8:49:06.29 UT (trigger 7560, see Kippen et al.\\ 1999). The GRB was also detected by the BeppoSAX {\\it Gamma-Ray Burst Monitor} (GRBM; Amati et al.\\ 1999a) and WFC unit 2 (Dadina et al.\\ 1999; Briggs et al.\\ 2000), as well as by {\\it Ulysses} (Hurley et al.\\ 2000) and the {\\it Near Earth Asteroid Rendezvous} (NEAR) spacecraft (Hurley 1999, private communication). In the WFC energy range (2--28\\,keV) the GRB had a duration of $\\sim$80\\,s and reached a peak intensity of 4.3~Crab. The WFC error box was followed up in X-rays by the Narrow Field Instruments (NFIs) onboard BeppoSAX $\\sim$8\\,hrs after the event and a strong decaying source was found (Piro et al.\\ 1999b, Kuulkers et al.\\ 1999). About 8.5\\,hr after the $\\gamma$-ray/X-ray event the optical counterpart was found (Vreeswijk et al.\\ 1999a) with a redshift of $z>1.62$ (Vreeswijk et al.\\ 1999b). A linear polarization of 2\\%\\ was measured (Covino et al.\\ 1999; Wijers et al.\\ 1999). Extended emission around the optical counterpart of GRB990510 has not been clearly detected, which indicates that a possible underlying host galaxy must be very faint (Israel et al.\\ 1999; Fruchter et al.\\ 1999b; Beuermann et al.\\ 1999). The light curve of the optical afterglow of GRB990510 does not follow a simple power-law decay, but showed smooth steepening after about one and a half day after the $\\gamma$-ray burst (Harrison et al.\\ 1999; Stanek et al.\\ 1999; Israel et al.\\ 1999). Traces of such a characteristic have also been found in the optical afterglow of GRB990123 (Kulkarni et al.\\ 1999; Fruchter et al.\\ 1999a) and the near-infrared afterglow of GRB990705 (Masetti et al.\\ 2000). It has been regarded as the signature of a decreasing collimation in a relativistic flow (Sari et al.\\ 1999; Rhoads 1999). Such behavior has never been observed in the X-ray afterglows of GRBs. The relatively large brightness of the GRB990510 X-ray afterglow allows an excellent opportunity to study the X-ray light curve in search of such a feature. ", "conclusions": "GRB990510 ranks among the top 25\\%\\ of the brightest GRB observed by the GRBM, while it ranks among the top 4\\%\\ (9\\%\\/) of the BATSE burst flux (fluence) distribution (Kippen et al.\\ 1999). The mean prompt $\\gamma$-ray spectrum is well described by a broken power-law, with a break energy of $\\sim$200\\,keV. The fluence, peak flux and spectrum as measured with the GRBM are comparable to those measured with BATSE. With the repeated pulses and \"hard-to-soft'' spectral evolution, the $\\gamma$-ray light curve and spectral behavior of GRB990510 are reminiscent of GRB970228 (Frontera et al.\\ 1998). The X-ray counterpart of GRB990510, 1SAX\\,J1338.1$-$8030, is also very bright if compared with other GRB X-ray afterglows (see e.g.\\ Piro 2000), and decays according to a typical power-law with index $-$1.42, which is consistent with that expected in relativistically expanding fireball models (e.g., Wijers, Rees \\&\\ M\\'esz\\'aros 1997). However, the optical light curve smoothly steepens $\\sim$1--2 days after the prompt $\\gamma$-ray emission (Stanek et al.\\ 1999; Harrison et al.\\ 1999; Israel et al.\\ 1999; see also Fig.~4). It was found that this steepening occurs at the same time in the different optical bands. To characterize its shape, the (V,R,I)-band data were simultaneously fitted by Harrison et al.\\ (1999) with the following four-parameter function\\footnote{As noted by Harrison et al.\\ (1999), the function which describes the optical (B,V,R,I) light curve by Stanek et al.\\ (1999) and Israel et al.\\ (1999) is different, leading to somewhat different values of the break time, i.e.\\ $\\sim$1.57~days.}: \\begin{equation} F_{\\nu}(t)=f_{\\star}(t/t_{\\star})^{\\alpha_1}[1-\\exp(-J)]/J;\\,\\,\\,\\,\\,J(t,t_{\\star},\\alpha_1,\\alpha_2)=(t/t_{\\star})^{(\\alpha_1-\\alpha_2)}, \\end{equation} with $t_{\\star}=1.20\\pm 0.08$\\,days, $\\alpha_1 = -0.82\\pm 0.02$, and $\\alpha_2 = -2.18\\pm 0.05$. In Fig.~4 we plot the optical R-band data taken in the same time span as the X-ray data, together with the above described function. It is clear that the optical data are not consistent with a power-law decay in that time span. We fitted the X-ray afterglow light curve with the same function as above, while fixing the decay indices to those derived in the optical. We find that the corresponding fits are bad, with $\\chi^2$ values of 24--27 for 14 d.o.f., depending on which parameter values we use among those reported by the different authors (Stanek et al.\\ 1999, Harrison et al.\\ 1999, Israel et al.\\ 1999). A steepening in the light curves can be expected in the fireball model when the cooling frequency moves towards lower frequencies in the observed frequency range. In that case the decay index $\\alpha$ changes by 0.25 (Sari et al.\\ 1998). However, the steepening of the optical decay is independent of wavelength (or achromatic) and the optical decay index $\\alpha$ changes by $\\sim$$-$1.36 (Harrison et al.\\ 1999). We provide additional evidence against a changing cooling frequency. In that case one would expect the optical decay index to be similar to that in the X-ray band, in contrast to what is observed. It has recently been realized that not all afterglow light curves are consistent with emission from expanding shells that are spherically symmetric, but that beaming may be important (i.e., jets, see e.g.\\ Sari et al.\\ 1999; Rhoads 1999). Such jets explain the presence of the steepening observed in the optical afterglow light curves of GRB990510 (e.g.\\ Harrison et al.\\ 1999). Sari et al.\\ (1999) presented general expressions for the expected spectral and decay index, appropriate for both spherical shell and jet evolutions shortly after the $\\gamma$-ray event. Our observed X-ray spectral index of $-$1.03$\\pm$0.08 implies a value of the index $p$ of the electron energy distribution in the expanding material of $p\\simeq 2.1$ in the case of fast cooling (i.e., when the cooling frequency is below the X-ray range). In the alternate case (i.e., the cooling frequency is above the X-ray range) we derive $p\\simeq 3.1$. Harrison et al.\\ (1999) found that the optical light curves can only imply $p\\simeq 2.1$, where the cooling frequency is above the optical wavelength range. Therefore, we conclude that the cooling frequency is between the optical and X-ray wavelengths. Note that the cooling frequency stays constant for a spreading jet (Sari et al.\\ 1999). At early times after the burst the decay light curve of a collimated source is identical to that of a spherical one, since then only a small portion of the emitting surface is visible due to relativistic beaming (the opening angle then is $1/\\gamma$, where $\\gamma$ is the Lorentz factor). In that case the decay index, $\\alpha$, is expected to be $-(3p-2)/4\\simeq -1.1$ in the case of fast cooling (i.e.\\ steeper than in the optical: $-(3p-1)/4\\simeq -1.3$; Sari et al.\\ 1999). As the fireball evolves, $\\gamma$ decreases, and the beaming angle will eventually exceed the jet opening angle. At that time one will see a break in the light curve, with $\\alpha = p \\simeq 2.1$, while the optical and X-ray decay index are similar after the break. Therefore, we fitted the X-ray afterglow light curve again, now fixing $\\alpha_1$ and $\\alpha_2$ to $-$1.1 and $-$2.1, respectively, and $t_{\\star}$ to that found in the optical. This leads to good fits with $\\chi^2$ values of about 12 for 14 dof. The corresponding fit is also shown in Fig.~4 (dotted line) with extrapolations to the boundaries of the plot. This shows that the observed X-ray afterglow of GRB990510 is consistent with the jet interpretation. As evident from Fig.~4, X-ray observations of the very early afterglow or a long time after the break time could have clearly discriminated whether the X-ray afterglow light curve is described by a single power law or consistent with the jet interpretation. We conclude that, even if we could not distinguish the presence of a clear break in the X-ray light curve, the only explanation within the fireball model consistent with the X-ray and optical data is a jet evolution, where the cooling frequency lies between the optical and X-ray wavelengths. Future observations of afterglows at late times with the X-ray observatories recently launched (Chandra and XMM-Newton) may provide a direct evidence of such a temporal X-ray feature." }, "0003/astro-ph0003291_arXiv.txt": { "abstract": "We report the detection of H{$\\alpha$} emission in the T dwarf (methane brown dwarf) 2MASSW J1237392+652615 over three days using the Keck Low Resolution Imaging Spectrograph. The measured line flux, log(L$_{H{\\alpha}}$/L$_{bol}$) = $-$4.3, is roughly consistent with early M dwarf activity levels and inconsistent with decreasing activity trends in late M and L dwarfs. Similar emission is not seen in two other T dwarfs. We speculate on several mechanisms that may be responsible for emission, including a strong magnetic field, continuous flaring, acoustic heat generation, and a close ($a$ $\\sim$ 4 - 20 R$_{Jup}$) interacting binary, with the cooler component overflowing its Roche lobe. We suggest that the M9.5Ve PC 0025+0447 could be a warm analogue to 2MASS J1237+65, and may be powered by the latter mechanism. ", "introduction": "Activity is an important parameter in the study of stellar populations. Numerous investigations of late-type (F-M) stars have shown correlations between emission (e.g., Ca II H \\& K, Mg II h \\& k, Balmer series, etc.) and fundamental parameters such as age, rotation, and metallicity \\citep{Ha00}. It is generally believed that the majority of this optical emission occurs in the chromosphere via collisional heating by ions and electrons along magnetic field lines. Indeed, this hypothesis is supported by the observed correlation of activity and rotation in late-type stars \\citep{Kr67,No84,Ba87}, which is expected if magnetic fields are generated by an internal dynamo (e.g.\\ $\\alpha$-$\\Omega$ dynamo; Parker 1955). Decrease in activity as stars age can be attributed to spin-down due to angular momentum loss in stellar winds \\citep{St86}. As we examine cooler M and L dwarfs, however, these relations begin to break down. As stars become fully convective ($\\sim$ 0.3 M$_{\\sun}$), the $\\alpha$-$\\Omega$ dynamo mechanism becomes ineffective, as it requires a low-buoyancy, radiative/convective boundary to anchor flux lines \\citep{Sp80}. However, the observed activity level remains roughly constant around this transition point \\citep{Ha96}, suggesting a turbulent dynamo as an alternate magnetic field source \\citep{Du93}. Indeed, flaring activity, which is magnetically driven, is seen in objects as late as M9.5 \\citep{Li99,Rd99}, supporting the existence of substantial magnetic fields beyond the convective cut-off. Alternately, acoustic heating could sufficiently heat the chromosphere \\citep{Sc87,Md92} to produce a ``basal flux'' of H$\\alpha$ emission. In either case, \\citet{Gi00} have shown that the fraction of objects with measurable emission rises to 100\\% at spectral type M7, then rapidly declines, so that no emission is seen in types L5 or later \\citep{Ki00}. The decrease in (steady) activity even encompasses objects with rapid rotation \\citep{Ba95,Ti98}, at odds with trends in hotter stars. Whether this drop in emission is due to ineffective chromospheric heating, decreased magnetic activity, or some other mechanism is unclear, but the end of the main sequence appears to mark a change in activity. Based on the results of \\citet{Gi00}, we would not expect significant activity in T dwarfs, brown dwarfs that show CH$_4$ absorption bands at 1.6 and 2.2 $\\micron$ \\citep{Ki99}. Nonetheless, we have observed H$\\alpha$ in emission in the T dwarf 2MASSW J1237392+652615 \\citep[herefter 2MASS J1237+65]{Bg99}, identified from the Two Micron All Sky Survey \\citep{Sk97}. We describe the optical observations of this and two other T dwarfs in $\\S$2; in $\\S$3 we discuss the H$\\alpha$ detection in 2MASS J1237+65 and possible emission mechanisms; in $\\S$4 we compare 2MASS J1237+65 to the unusual M9.5Ve PC 0025+0447; we summarize our results in $\\S$5. ", "conclusions": "We have reported the detection of H$\\alpha$ emission in the T dwarf 2MASS J1237+65, at a level of log(L$_{H{\\alpha}}$/L$_{bol}$) = $-$4.3. This emission is intriguing, as it is a salient exception to the cool dwarf temperature-activity relations identified thus far. We have proposed various activity mechanisms, including a strong magnetic field, continuous flaring, and an interacting brown dwarf binary, but these are speculative guesses at best. Comparison can be drawn with the M9.5Ve PC 0025+0447, which, if it is a brown dwarf, could be a warm analogue to 2MASS J1237+65. Both objects could also be close binary systems with lower-mass brown dwarf companions, which are steadily transferring mass to their primaries by Roche lobe overflow. Nonetheless, the mechanism for both objects remains unclear. Further investigation of the temporal stability of the H$\\alpha$ line in 2MASS J1237+65 and searches for emission in other T dwarfs is clearly warranted." }, "0003/astro-ph0003128_arXiv.txt": { "abstract": "\\small We present a new model to compute the effects of dust on the integrated spectral properties of galaxies, based on an idealized prescription of the main features of the interstellar medium (ISM). The model includes the ionization of H{\\sc ii} regions in the interiors of the dense clouds in which stars form and the influence of the finite lifetime of these clouds on the absorption of radiation. We compute the production of emission lines and the absorption of continuum radiation in the H{\\sc ii} regions and the subsequent transfer of line and continuum radiation in the surrounding H{\\sc i} regions and the ambient ISM. This enables us to interpret simultaneously all the observations of a homogeneous sample of nearby ultraviolet-selected starburst galaxies, including the ratio of far-infrared to ultraviolet luminosities, the ratio of H$\\alpha$ to H$\\beta$ luminosities, the H$\\alpha$ equivalent width, and the ultraviolet spectral slope. We show that the finite lifetime of stellar birth clouds is a key ingredient to resolve an apparent discrepancy between the attenuation of line and continuum photons in starburst galaxies. In addition, we find that an effective absorption curve proportional to $\\lambda^{ -0.7}$ reproduces the observed relation between the ratio of far-infrared to ultraviolet luminosities and the ultraviolet spectral slope. We interpret this relation most simply as a sequence in the overall dust content of the galaxies. The shallow wavelength dependence of the effective absorption curve is compatible with the steepness of known extinction curves if the dust has a patchy distribution. In particular, we find that a random distribution of discrete clouds with optical depths similar to those in the Milky Way provides a consistent interpretation of all the observations. A noteworthy outcome of our detailed analysis is that the observed mean relations for starburst galaxies can be closely approximated by the following simple recipe: use an effective absorption curve proportional to $\\lambda^{-0.7}$ to attenuate the line and continuum radiation from each stellar generation, and lower the normalization of the curve typically by a factor of 3 after $10^7\\,$yr to account for the dispersal of the birth clouds. This recipe or our full model for absorption can be incorporated easily into any population synthesis model. ", "introduction": "To interpret the observed spectral properties of galaxies, we require models for both the production of stellar radiation and its transfer through the interstellar medium (ISM). Currently, the accuracy of population synthesis models contrasts with the rudimentary way in which the absorption of starlight by dust is often treated. In many applications, dust is either ignored or assumed to be distributed in a uniform screen in front of the stars. The resulting uncertainties in the absorption of the ultraviolet radiation in galaxies can be as much as an order of magnitude. This problem has become especially acute in studies of galaxies at high redshifts. Nearby starburst galaxies may be suitable analogs of high-redshift galaxies and provide important clues to interpreting their spectral properties. The observations of nearby ultraviolet-selected starburst galaxies are numerous, including the ratio of far-infrared to ultraviolet luminosities, the ratio of H$\\alpha$ to H$\\beta$ luminosities, the H$\\alpha$ equivalent width, and the ultraviolet spectral slope. In particular, there is a remarkably tight correlation between far-infrared luminosity and ultraviolet spectral slope (Meurer et al. 1995; Meurer, Heckman, \\& Calzetti 1999). This wealth of observations can potentially help us quantify the effects of dust on various spectral properties of galaxies. There have been several analyses of the spatial distribution and optical properties of the dust in nearby starburst galaxies based on various subsets of the observations (e.g., Fanelli, O'Connell, \\& Thuan 1988; Calzetti, Kinney, \\& Storchi-Bergmann 1994, 1996; Puxley \\& Brand 1994; Meurer et al. 1995; Gordon, Calzetti, \\& Witt 1997). A generic result of these studies is that if the dust is distributed in a uniform foreground screen, it must have an unusually grey extinction curve. Otherwise, the distribution must be patchy. However, these analyses also raise questions of how to account self-consistently for all the observations. For example, the absorption inferred from the H$\\alpha$/H$\\beta$ ratio in starburst galaxies is typically twice as high as that inferred from the ultraviolet spectral slope. One interpretation of this result is that the ionized gas and ultraviolet-bright stars have different spatial distributions (Calzetti 1997). Issues such as this highlight the need for a simple yet versatile model to interpret simultaneously a wide range of phenomena related to the absorption of starlight by dust in galaxies. The purpose of this paper is to present such a model. We begin with the conventional view that young stars ionize H{\\sc ii} regions in the interiors of the dense clouds in which they are born. Line photons produced in the H{\\sc ii} regions and the non-ionizing continuum photons from young stars are absorbed in the same way by dust in the outer H{\\sc i} envelopes of the birth clouds and the ambient ISM. The birth clouds, however, have finite lifetimes. Thus, non-ionizing ultraviolet and optical photons from stars that live longer than the birth clouds are absorbed only by the ambient ISM. This allows the ultraviolet continuum to be less attenuated than the emission lines. Our model builds on several previous studies. For example, Silva et al. (1998) considered the effects of finite lifetimes of stellar birth clouds on the continuum but not the line emission from galaxies. Here, we treat the transfer of radiation (especially scattering) in an approximate way, which would preclude a detailed description of surface brightnesses but should be appropriate for angle-averaged quantities such as luminosities. Our model succeeds in accounting quantitatively for all the available observations of a homogeneous sample of nearby starburst galaxies. We present our model in \\S2, where we express the effective absorption curve describing the global transmission of radiation in terms of the different components of the ISM. In \\S3, we compare our model with observations and identify the specific influence of each parameter on the different integrated spectral properties of galaxies. One outcome of our detailed analysis is a remarkably simple recipe for absorption, which provides a good approximation to the observed mean relations and is easy to incorporate into any population synthesis model. In \\S4, we explore how the spatial distribution of dust can be constrained by the observations. Our conclusions are summarized in \\S5. ", "conclusions": "We have developed a simple model to compute the effects of dust on the integrated spectral properties of galaxies, based on an idealized prescription of the main features of the ISM. Our model includes the ionization of H{\\sc ii} regions in the interiors of the dense clouds in which stars form. Emission lines from H{\\sc ii} regions and the non-ionizing continuum from young stars are attenuated in the same way by dust in the outer H{\\sc i} envelopes of the birth clouds and the ambient ISM. However, since the model also includes the finite lifetimes of the birth clouds, the non-ionizing continuum radiation from stars that live longer than the birth clouds is attenuated only by the ambient ISM. We show that this can fully resolve the apparent discrepancy between the attenuation of line and continuum photons in starburst galaxies. This enables us, in turn, to interpret in a consistent way all the observations of a homogeneous sample of nearby ultraviolet-selected starburst galaxies, including the ratio of far-infrared to ultraviolet luminosities ($L_{\\rm dust}/L_{1600}$), the ratio of H$\\alpha$ to H$\\beta$ luminosities ($L_{{\\rm H}\\alpha}/L_{{\\rm H}\\beta}$), the H$\\alpha$ equivalent width ($W_{{\\rm H}\\alpha}$), and the ultraviolet spectral slope ($\\beta$). The different parameters in our model, including the effective age of the starburst, the lifetime and effective optical depth of the stellar birth clouds, the effective optical depth in the ambient ISM, and the fraction of dust in the ionized gas, each have a specific influence on the integrated spectral properties $L_{\\rm dust} /L_{1600}$, $L_{{\\rm H}\\alpha}/L_{{\\rm H}\\beta}$, $W_{{\\rm H} \\alpha}$, and $\\beta$. This provides new insights into the origin of the mean relations defined by the data and the scatter about these relations. In particular, the relation between the ratio of far-infrared to ultraviolet luminosities and the ultraviolet spectral slope in starburst galaxies reflects the wavelength dependence of the effective absorption in the ambient ISM. We find that a power law of the form $\\hat{\\tau }_\\lambda^{\\rm ISM} \\propto \\lambda^{ -0.7}$ accounts remarkably well for all the observations. The relation between $L_{\\rm dust} /L_{1600}$ and $\\beta$ can then be interpreted as a sequence in the overall dust content of the galaxies. Interestingly, this relation is accompanied by much weaker trends of the oxygen abundance and the optical luminosity with the ultraviolet spectral slope. The fact that our model reproduces the observed spectral properties of nearby starburst galaxies relatively easily leads us to suspect that, with suitable adjustment of the parameters, it could also reproduce those of more quiescent (but still star-forming) galaxies. The effective absorption curve required by the observations is much greyer than would be produced by a foreground screen of dust like that in the Milky Way, the LMC, or the SMC. We have explored whether this could be accounted for by a much steeper wavelength dependence of the optical depth, i.e. $\\tau_\\lambda \\propto \\lambda^{-m}$ with $1.0\\la m \\la 1.5$, combined with a more realistic spatial distribution of the dust. We find that a mixed slab model for the ambient ISM can produce the required effective absorption curve for low dust content but cannot explain the observations of starburst galaxies with very reddened ultraviolet spectra. In contrast, we show that a random distribution of discrete clouds provides a consistent interpretation of all the observed integrated spectral properties of starburst galaxies. While these results were anticipated in some previous studies, we have shown here for the first time how to reconcile them with the large H$\\alpha/$H$\\beta$ ratios and other observations. We also find that the optical depths of the clouds favored by our analysis are similar to those inferred from the statistics of stellar reddening in the Milky Way. The model we have developed for computing the absorption of starlight by dust in galaxies can be combined easily, by design, with any population synthesis model. The observed mean relations for starburst galaxies can also be reproduced by the following simple recipe: use an effective absorption curve proportional to $\\lambda^{-0.7}$ to attenuate the line and continuum radiation from each stellar generation, and lower the normalization of the curve typically by a factor of 3 after $10^7\\,$yr to account for the dispersal of the birth clouds. This recipe accounts at least as well as the one by Calzetti et al. (1994, and as modified by Calzetti 1997, 1999) for the effects of dust on the non-ionizing continuum radiation. In addition, it fully resolves the apparent discrepancy between the attenuation of line and continuum photons in starburst galaxies. We believe, therefore, that our model and the recipe derived from it provide simple yet versatile tools to interpret the integrated spectral properties of starburst and possibly other types of galaxies. In future work, we plan to apply them to the growing body of observations of high-redshift galaxies." }, "0003/astro-ph0003402_arXiv.txt": { "abstract": "The alignment of clusters of galaxies with their nearest neighbours and between clusters within a supercluster is investigated using simulations of $512^{3}$ dark matter particles for $\\Lambda$CDM and $\\tau$CDM cosmological models. Strongly significant alignments are found for separations of up to $15 h^{-1}$Mpc in both cosmologies, but for the $\\Lambda$CDM model the alignments extend up to separations of $30h^{-1}$Mpc. The effect is strongest for nearest neighbours, but is not significant enough to be useful as an observational discriminant between cosmologies. As a check of whether this difference in alignments is present in other cosmologies, smaller simulations with $256^{3}$ particles are investigated for 4 different cosmological models. Because of poor number statistics, only the standard CDM model shows indications of having different alignments from the other models. ", "introduction": "Differing claims have been made as to the scale and significance of alignments of clusters of galaxies. Binggeli (1982) was the first to point out this effect. He found that not only were clusters aligned with their nearest neighbours out to a separation of about $15h^{-1}$Mpc (where $h=H_{0}/100$ $\\mathrm{km}$ $\\mathrm{s}^{-1} \\mathrm{Mpc}^{-1}$) but that the orientation of a cluster was also related to the distribution of all surrounding clusters up to a separation of about $50h^{-1}$Mpc. Since then the effect has been studied by many authors (Struble \\& Peebles 1985; Flin 1987; Rhee \\& Katgert 1987; West 1989; Ulmer, McMillan \\& Kowalski 1989; West, Dekel \\& Oemler 1989; Fong, Stevenson \\& Shanks 1990; Plionis 1994; Martin et al. 1995; Splinter et al. 1997) but with conflicting results. A number of factors needed to be considered, particularly in the observational studies, before definite conclusions could be drawn as to the reality of the effect. One major problem is that of determining the position angles of the clusters. For this reason Flin (1987) used several independently determined position angles in his study, and Rhee and Katgert (1987) developed a more objective semi-automatic procedure to determine position angles. As a result both studies found some support for the alignment effects found by Binggeli (1982). However Martin et al. (1995) pointed out that a fundamental problem is that of the number of galaxies being too few to trace the cluster potential adequately. In addition, many of the studies have used clusters from catalogues (e.g. Abell and Lick) which are known to contain systematic biases (e.g. Lumsden et al. 1992). Using instead a statistical search of the Edinburgh-Milano cluster redshift survey, Martin et al. (1995) found no statistically significant evidence of cluster alignments. There would be considerable advantages in using X-ray data rather than optical data since hot gas traces the cluster potential directly and reduces the problem of confusion of cluster membership along the line of sight. Ulmer et al. (1989) used 45 X-ray clusters and found no significant effect. However, Rhee, van Haarlem \\& Katgert (1992) found that, using combined optical and X-ray data, clusters do tend to point towards neighbouring clusters if they are members of the same supercluster. Large scale simulations of clusters can provide large data sets of clusters for different cosmological models without observational biases and for which the orientation of the semi-major axis can be derived in a straightforward manner. A comparison between the simulations and observations can reveal whether the observations are consistent with any of the cosmologies. Whether the observations could be used to discriminate between models depends on the statistical significance of the difference in alignments between cosmological models, which can also be obtained from the simulations. If the alignments are sufficiently different in different models, then large observational data sets of cluster position angles could be used to test for the cosmology. The best data presently available is that of Plionis (1994) who estimated position angles for 637 clusters in contrast to the very much smaller numbers of clusters of about 50 or less used in other observational studies of alignments. Splinter et al. (1997) used simulations of $128^{3}$ particles to study the ellipticity and orientation of clusters of galaxies in N-body simulations for different cosmological models (ie. different values of density parameter, $\\Omega_{o}$, and initial power spectra). The box sizes varied from 110 to 300$h^{-1}$Mpc depending on the initial spectral index and the evolutionary stage. They found significant alignments for all spectra at separations less than $15h^{-1}$Mpc and that differences in $\\Omega$ had no measurable effect. In the present work cluster alignments in different cosmologies will be looked for using significantly larger simulations having $512^{3}$ and $256^{3}$ particles. Explanations of the alignment of galaxy clusters, if this effect exists, fall into 2 main categories. Tidal distortion due to interactions between clusters of galaxies has been proposed as the origin of the ellipticity of clusters (Binney \\& Silk 1979) and also of alignment (Salvador-Sol\\'{e} \\& Solanes 1993), although Dekel, West \\& Aarseth (1984) had earlier found that tidal interactions do not produce alignments and favoured instead an intrinsic origin for cluster anisotropies. Using N-body simulations they found that alignments only occurred in a 'top down' scenario in which superclusters collapsed from excess fluctuations on large scales rather than hierarchical clustering from fluctuations on smaller scales. However, other simulations of hierarchical clustering (e.g. White 1976; Cavaliere et al. 1986) have modelled the clumpy, non-spherical nature of clusters formed via subclustering. In line with this, van Haarlem \\& van de Weygaert (1993) found for cold dark matter (CDM) models that clusters are elongated in the direction from which the last subcluster fell into the cluster. West, Jones \\& Forman (1995) proposed that cluster formation proceeds by the merging of subclusters along large-scale filamentary features in the matter distribution. Thus the initial concern that the more accepted CDM models may be ruled out by the observed alignments in favour of a top down pancake scenario may be unfounded and cluster alignments may in fact be expected irrespective of the cluster formation model (Plionis 1994) In the present work large scale simulations of two CDM models are analysed for cluster alignment effects both among nearest neighbours and between a cluster and its neighbours within a supercluster. This is extended to a total of 4 different models using smaller simulations. The aims are to confirm the scale and significance of any alignments and to look for any significant differences between different cosmological models. The details of the simulations and cluster catalogues are given in Section 2, the method and results of searching for alignments are given in Section 3, and the results are discussed in Section 4. ", "conclusions": "The results from the very large simulations confirm without doubt that alignments between clusters of galaxies do exist. This effect is stronger between nearest neighbours than for all neighbours, even if they are restricted to cluster pairs within a supercluster. A difference in the alignments was found for the two cosmological models investigated in that the alignments extended to larger separations (about $30h^{-1}$Mpc) for the $\\Lambda$CDM model than for the $\\tau$CDM model ($<15h^{-1}$Mpc). However the significance of this difference is not large enough to be used as an observational test. Splinter et al. (1997) found that there appears to be a very weak trend that as $\\Omega$ is lowered more alignments are seen. They found a stronger trend that as the exponent of the primordial power spectrum (n) is made more negative there is increasing alignment between clusters. In the present simulations the power spectrum and shape parameter ($\\Gamma$) were the same for both the $\\Lambda$CDM and $\\tau$CDM models. However in our simulations the normalization of the power spectrum ($\\sigma_{8}$) was chosen to give the correct number of clusters at the present day for a particular cosmological model, giving more power on large scales for low $\\Omega$ models. It is thus not surprising that the alignments act over a longer range in the $\\Lambda$CDM model. Although Binggeli (1982) found no relation between the orientation effect and cluster richness, we found some evidence that alignments increased with cluster richness. However, in agreement with Fuller et al. (1999), we found that alignments may persist down to poorer clusters (mass in region $10^{13}-10^{14}\\mathrm{M}_{\\odot}$). In conclusion, it appears that cluster alignments are present for all CDM models up to separations of $15h^{-1}$Mpc. The alignments extend to greater separations for the low $\\Omega$ models at least, but the differences between models are not strong enough to be useful as a cosmological test. The alignments found may fit in with a general picture of cluster formation by hierarchical clustering in which material falls into the cluster along the large scale filamentary structure, possibly irrespective of cluster richness." }, "0003/hep-ph0003301_arXiv.txt": { "abstract": "\\baselineskip 16pt \\tightenlines Detailed predictions for the day-night (D-N) asymmetry in the energy-integrated one year signals in the SNO detector in the case of the MSW $\\nu_e \\rightarrow \\nu_{\\mu(\\tau)}$ and/or $\\nu_e \\rightarrow \\nu_{s}$ transition solutions of the solar neutrino problem are presented. The asymmetries in the charged current (CC) and $\\nu - e^{-}$ elastic scattering (ES) event rates are calculated for both MSW solutions; in the case of the $\\nu_e \\rightarrow \\nu_{s}$ transition solution the D-N asymmetry in the neutral current (NC) event rate are derived as well. The asymmetries are calculated for three night samples of events which are produced by the solar neutrinos crossing i) the Earth mantle only ({\\it Mantle}), ii) the Earth core ({\\it Core}) and iii) the Earth core and/or the mantle ({\\it Night}). The effects of the uncertainties i) in the values of the cross-sections of the CC and NC neutrino-induced reactions on deuterium, and ii) in the value of the bulk matter density and/or the chemical composition of the Earth core, on the corresponding D-N asymmetry predictions are analyzed. It is shown, in particular, that that due to the strong enhancement of the transitions of the solar neutrinos crossing the Earth core, at $\\sin^22\\theta_V \\leq 0.01$ the corresponding one year average D-N asymmetry in the {\\it Core} sample of CC events in the case of the $\\nu_e \\rightarrow \\nu_{\\mu(\\tau)}$ solution can be larger by a factor of up to $\\sim 8$ than the asymmetry in the {\\it Night} sample. In certain subregions of the MSW solution regions at small $\\sin^22\\theta_V$, the predicted magnitude of the {\\it Core} D-N asymmetry in the CC sample is very sensitive to the value of the electron fraction number in the Earth core. Iso - (D-N) asymmetry contours in the $\\dms - \\SdTvS$ plane for the SNO detector are derived in the region $\\SdTvS \\gsim 10^{-4}$ for the {\\it Core} and {\\it Night} samples of the CC, ES and NC events. The dependence of the D-N asymmetries considered on the final state e$^{-}$ threshold energy in the CC and ES reactions is also investigated. Our results show, in particular, that the SNO experiment will be able to probe substantial parts of the SMA and LMA MSW $\\nu_e \\rightarrow \\nu_{\\mu(\\tau)}$ solution regions be performing {\\it Night} and {\\it Core} D-N asymmetry measurements. ", "introduction": " ", "conclusions": "" }, "0003/gr-qc0003115_arXiv.txt": { "abstract": "A tensor description of perturbative Einsteinian gravity about an arbitrary background spacetime is developed. By analogy with the covariant laws of electromagnetism in spacetime, {\\em gravito-electro\\-magnetic} potentials and fields are defined to emulate electromagnetic gauge transformations under substitutions belonging to the gauge symmetry group of perturbative gravitation. These definitions have the advantage that on a flat background, with the aid of a covariantly constant timelike vector field, a subset of the linearised gravitational field equations can be written in a form that is fully analogous to Maxwell's equations (without awkward factors of 4 and extraneous tensor fields). It is shown how the remaining equations in the perturbed gravitational system restrict the time dependence of solutions to these equations and thereby prohibit the existence of propagating vector fields. The induced {\\em \\gemic} Lorentz force on a test particle is evaluated in terms of these fields together with the torque on a small gyroscope. It is concluded that the analogy of perturbative gravity to Maxwell's description of electromagnetism can be valuable for (quasi-)stationary gravitational phenomena but that the analogy has its limitations.\\newline \\noindent PACS numbers: 0420, 0450 ", "introduction": "Einstein's theory of gravitation remains a pinnacle in the evolution of theoretical physics. It offers an overarching description of phenomena ranging from the familiar behaviour of Newtonian gravitation to exotic astrophysical events at the extremes of space and time. Although its modern formulation is in terms of tensor fields on a manifold with a spacetime structure, its physical interpretation often benefits from a choice of observer and an appropriate reference frame. One of the traditional methods for extracting information from Einstein's gravitational field equations is to exploit the properties of observers in some fiducial background spacetime in which gravitational physics is either absent or familiar. This approach has led to various approximation in such backgrounds. Further reduction is afforded by a ``3+1'' decomposition in which spacetime tensors are expressed in terms of a field of frames adapted to some local foliation of spacetime by spacelike hypersurfaces. More generally frames are afforded by timelike vector fields, the integral curves of which model ideal observers. In such a manner it becomes possible to contemplate different limits in which matter moves slowly or the gravitational field is weak relative to such observers. It is known that Newtonian gravitation follows from such a limit. Within the framework of weak gravity non-Newtonian gravitational effects may arise and a number of experiments have been devised in order to detect such phenomena as the ``Lense-Thirring effect'' due to ``frame-dragging'' produced by the earth's rotation \\cite{Thirring}, \\cite{Thirring1}, \\cite{Thirring2}, \\cite{Everitt}. The nature of this effect may be detectable by a small orbiting gyroscope and is analogous to that produced by the torque on a small magnetic dipole in the presence of the magnetic field of a fixed magnetic dipole. Indeed the component of weak gravity (additional to the dominant Newtonian gravitational field) responsible for this effect is now referred to as the {\\em gravito-magnetic} field. The sensitivity of recently developed rotation sensors may also be increased to detect post-Newtonian effects in the future \\cite{Packard}. Several authors \\cite{bct}, \\cite{harris}, \\cite{thorneNZ}, \\cite{thorne}, \\cite{wald}, \\cite{forward}, \\cite{Zee}, \\cite{ehlers}, \\cite{jan}, \\cite{mashhoon}, \\cite{damour} have noticed that a subset of the Einstein equations when perturbed about flat spacetime can be written in a form that looks remarkably similar to Maxwell's equations with the Newtonian gravitational field corresponding to the {\\em \\gelec} field and mass-currents playing the role of electric currents. Since the laws of electromagnetism are well studied and understood this analogy has proved quite fruitful in the {\\em \\gemic} context particularly in astrophysical applications. Extended ``astrophysical jet-structures'' are now thought to have their origin in {\\em \\gemic} forces. In \\cite{NUT} the details of astrophysical lensing have been explored in terms of parameters in the NUT metric. It is also amusing to recall that one of the first theories of post-Newtonian gravitation was formulated by Heaviside in direct analogy with the then recently formulated theory of electromagnetism by Maxwell. In the language of the Poincar\\'{e} isometry group it predicted that gravitation like electromagnetism was mediated by an independent vector field rather than a second degree tensor field associated with the metric of spacetime. This difference of course must imply that any analogy between weak gravity and electromagnetism is incomplete and most derivations of the {\\em \\gemic} field equations take care to point this out. However in our view the caveats are themselves often incomplete and a close examination of various derivations of the {\\em \\gemic} equations display significant differences in detail. The difficulty in making objective comparisons often arises due to the implicit use of a particular coordinate system (usually adapted to a flat spacetime background) or a partial gauge fixing. Indeed the question of the gauge transformations induced on the {\\em \\gemic} fields from the underlying gauge covariance of the perturbative Einstein equations is usually dealt with rather cursorily. This leads one to contemplate the {\\it most useful way} to define the {\\em \\gemic} fields in terms of the perturbed components of the spacetime metric. Different choices are often responsible for the location of odd factors of 4 that permeate the {\\em \\gemic} equations compared with the Maxwell equations. Such choices also have implications for the form of the induced {\\em \\gemic} Lorentz force (and torque) in terms of the {\\em \\gemic} fields that enter into the equation for the motion of a massive point (spinning) particle. In \\cite{mashhoon} \\gemic gauge transformations are discussed from a perspective different from the one presented in this paper. Here such transformations are explicitly related to the gauge symmetry of perturbative gravitation and the definitions of \\gemic fields in turn induce the notions of gravito-magnetic and gravito-electric coupling strengths. In this article a tensor description of perturbative Einsteinian gravity about an arbitrary background spacetime is first constructed. By analogy with the covariant laws of electromagnetism in spacetime {\\em \\gemic} potentials and fields are then defined to emulate electromagnetic gauge transformations under substitutions belonging to the gauge symmetry group of perturbative gravitation. These definitions have the advantage that on a flat background, with the aid of a covariantly constant timelike vector field, a {\\it subset} of the linearised gravitational field equations can be written in a form that is {\\it fully analogous to Maxwell's equations} (without awkward factors of 4 and extraneous tensor fields. It is shown how the remaining equations in the perturbed gravitational system restrict the time dependence of solutions to these equations and thereby prohibit the existence of propagating vector fields. The induced {\\em \\gemic} Lorentz force on a test particle is evaluated by geodesic perturbation in terms of these fields together with the torque on a small gyroscope. It is concluded that the analogy of perturbative gravity to Maxwell's description of electromagnetism can be valuable for (quasi-)~stationary gravitational phenomena but that the analogy has its limitations. It has been argued that such limitations are absent in the approach to {\\em \\gemic} based on properties of the conformal tensor in a spacetime determined by Einstein's equations. Although this reformulation makes no reference to perturbative methods the analogy with the structure of Maxwell's equations is less direct. A tensorial description of this formulation is given in Appendix \\sect{conf}. Throughout this article the language of tensor fields as multi-linear maps on vector and co-vector fields is adopted. Co-vector fields are manipulated using the exterior calculus of differential forms and Hodge maps. Manifolds are assumed smooth and tensor fields sufficiently differentiable as required. Notations based on the tools used are summarised in Appendix \\sect{defs} and some technical computational details are relegated to Appendices \\sect{connex} and \\sect{trans}. In order to facilitate comparisons with other authors certain field redefinitions are discussed in section \\sect{altanalogy} together with the changes induced by them in the {\\em \\gemic} field equations. These alternatives are discussed in the concluding section where the salient features of this paper are summarised. ", "conclusions": "\\label{sec:discuss} In this article the analogy between Maxwell's equations for the electromagnetic field, \\eqn{maxcov1}, \\eqn{maxcov2} and the Einstein equations for weak gravitational fields in the transverse gauge, \\eqn{gemmaxcov1}, \\eqn{gemmaxcov2} has been made in terms of tensor fields. While the former are valid in an arbitrary Lorentzian spacetime the latter have been developed in terms of perturbations about a flat spacetime background. A comparison has been made between the general equation describing the motion of electrically charged point particles \\eqn{qeqnmotion} and the motion of massive point particles in a weak gravitational field \\eqn{eqnmotion}. Equations have also been developed \\eqn{gyroeqn} describing the motion of a freely falling small gyroscope in terms of \\gemic fields. In general it is asserted that any analogy between electromagnetism and weak gravity is closest in a restricted class of reference frames related by suitable non-relativistic transformations and for stationary physical field configurations in such frames. In addition to \\gemic fields the general equations of weak gravity involve a second degree symmetric tensor field $\\bsSigma_\\xi$ which has no electromagnetic analogue. The \\gemic fields defined in \\ref{sec:gemfieldeq} are coupled to $\\bsSigma_\\xi$ via \\eqn{secondgc} and this field produces non-Maxwellian terms in the weak gravitational force and torque equations, \\eqn{gemlorentz} and \\eqn{torque} respectively. In \\emism the Maxwell fields $A$, $F$ and $\\cJ$ are defined independent of any frame of reference. The latter is only required to define electric and magnetic fields and their sources in terms of electric charge and current density \\eqn{fsplit}, \\eqn{currentsplit}. In \\gemism the analogous fields $\\psi_\\xi$, $\\cF_\\xi$ and $\\cJ_\\xi$ are manifestly frame-dependent. The definition of these fields has been motivated by their behaviour under a class of gauge transformations belonging to the gauge symmetry of the weak Einstein equations. Unlike \\emism, $\\cF_\\xi$ is not gauge invariant under these transformations in general. However a subset of these transformations does exist for which $\\cF_\\xi$ remains invariant. These have been called \\gemic gauge transformations by analogy with the gauge symmetry of Maxwell's equations. Unlike electromagnetic interactions with electrically charged particles, weak gravitational interactions are not mediated by complex representations of these symmetries. In terms of the \\gemic fields a subset of the linearised Einstein equations take a remarkable form that is isomorphic to Maxwell's \\emic field equations. No limits or further approximations are required to establish this correspondence. The explicit appearance of the $\\bsSigma_\\xi$ tensor is confined to the remaining equations in the linearised system. This is a primary distinction of the approach adopted here compared with previous derivations of the \\gemic field equations. To exploit this reformulation and link physical field configurations with solutions to Maxwell's equations further conditions must be imposed on the linearised system. Conditions have been found that enable a useful analogy between weak gravitation and electromagnetism to be established. In the \\gemic limit $\\bsSigma_\\xi$ depends only on the a \\gemic potential \\eqn{gemlimiteqn}. Consequently, such weak gravitational fields can be described in terms of $\\psi_\\xi$ (or alternatively $\\bscE_\\xi$ and $\\bscB_\\xi$). However as stressed above, in order to obtain Maxwell-like equations the transverse gauge condition \\eqn{transgc} is imposed on $\\psi$. Condition \\eqn{transsplit} induces an equivalent condition \\eqn{firstgc} on $\\psi_\\xi$. (By contrast the electromagnetic gauge condition \\eqn{transgauge} is one of many that may be imposed on $A$.) The condition on $\\psi$ also imposes the restriction \\eqn{secondgc} which in the \\gemic limit implies that $\\psi_\\xi$ has a restricted time dependence. Physical \\gemism consequently shares more in common with electromagneto-statics than electromagnetism. Although both the Maxwell equations \\eqn{maxall} and the equivalent \\gemic equations \\eqn{gemmaxall} hold in any inertial frame (in Minkowski spacetime), in order to remain within the \\gemic limit only a class of \\gemic frames of reference is permitted (Appendix \\sect{trans}). By perturbing the equation of a physical timelike geodesic the relativistic equation of motion for a massive point particle in the weak gravitational field can be cast into a form containing a \\gemic Lorentz force \\eqn{gemlorentz} and an additional non-Maxwellian term. It is worth pointing out that in the context of the \\gemic fields defined in this paper the derivation of this equation of motion does not rely on the speed of the particle and the \\gemic Lorentz-like force takes its natural form. In the \\gemic and non-relativistic limit the particle acceleration is then determined by a non-relativisti Lorentz-like force \\eqn{pop} containing an additional factor of $\\frac{1}{4}$ multiplying the \\gelec field. When working in the \\gemic limit the field redefinitions presented in section \\sect{altanalogy} permit a comparison with the work of Thorne in \\cite{thorneNZ} with the notation $\\Phi_h = \\Phi$, $\\bfA_h = \\bsgamma$, $\\bfE_h = \\bfg$, and $\\bfB_h = \\bfH$, and with a current of the form $\\bfJ = \\rho \\bfv$. The work of \\cite{harris}, \\cite{mashhoon}, \\cite{bct}, \\cite{thorne} and others may be related to that of \\cite{thorneNZ} either by trivial field redefinitions or changes in metric signature. The analogy between electromagnetism and weak gravity developed by Wald \\cite{wald} is similar to the one presented here. However he does not discuss how the \\gemic 1-form-potential behaves under gauge transformations nor how the restricted time dependence arises from the transverse gauge condition. This article offers an alternative description of weak field gravitation in the language of \\gemic fields. We feel that it clarifies a number of issues concerning various other analogies between the equations of post-Newtonian gravity according to Einstein and Maxwell's description of electromagnetism. In the absence of a \\gemic limit formulations based on $\\Phi_h$ and $\\bfA_h$ give rise to gauge conditions \\eqn{hlorenz}, \\eqn{nonmaxgch} and field equations \\eqn{gemsigcoup1}, \\eqn{gemsigcoup2} thereby exposing couplings between the \\gemic potentials and fields and the tensor field $\\bsSigma_\\xi$. In the approach adopted here such couplings are relegated to the gauge condition \\eqn{nonmaxgcpsi}. Many analogies coalesce in the \\gemic limit modulo a re-shuffling of numerical factors that cannot be scaled away entirely. Although the mathematical analogy between weak gravity and the full system of Maxwells equations for electromagnetism in terms of covariant tensor fields on flat spacetime can be made close, the existence of gravitational gauge conditions limits the physical\\footnote{Fields derived from a potential $\\phi_{\\partial_t}$ that vary linearly with $t$ are deemed unphysical here.} analogy to stationary phenomena. Despite this limitation the interpretation of weak gravity in terms of \\gemic fields offers a fertile avenue of exploration for phenomena associated with the detection of the stationary gravito-magnetic field. The methods presented here are also applicable in principle to certain non-flat backgrounds and to weak field descriptions of non-Einsteinian gravitation (in which other geometrical fields may compete with metric-induced gravity) at the post-Newtonian level. These issues will be discussed elsewhere." }, "0003/astro-ph0003152_arXiv.txt": { "abstract": " ", "introduction": "Currently there are two general relativistic descriptions of spacetime in popular use. For planetary systems and other gravitationally bound structures which are small on the scale of the universe, there is a static description of the behavior of spacetime. On the other hand, for large-scale behavior, there is a time dependent description which is appropriate as a description of phenomena such as the observed red-shift of distant galaxies. The classic question is, ``How can these two disparate descriptions of spacetime possibly be reconciled with each other?'' The current standard answer is that these two are meshed together\\cite{Peb} on a spherical surface surrounding a mass concentration which grows with time in the static metric, but stays at a fixed coordinate radius in the non-static metric. This is the ``Schwarzschild solution in a cosmological model'' picture, or it is also called the ``Swiss cheese model.'' This later name refers to the fact the in this picture, the background material is removed inside the spherical boundary. The mass removed depends on the mass of the central concentration and the curvature of space. The replacement of the mass interior to the sphere by a concentration of mass at the center is based on Birkhoff's theorem\\cite{Br} which says that in a homogeneous, zero pressure cosmological model, as long as the material inside a sphere is spherically symmetric, we can replace it with a compact mass at the center with no change on its exterior effects. To make this matching work, the pressure of the exterior solution (Friedmann-% Lema\\^\\i tre metric class of the general Robertson-Walker line elements) must vanish, as does the pressure of the interior (the exterior Schwarzschild metric) solution. This condition places a restriction on the form of the universal expansion factor of the overall universe. Although the matching conditions can be met, as we shall see in the third section, an additional problem arises when the dynamics are considered. In the second section, for the convenience of the reader, I gather together the necessary classical equations from general relativity for the study at hand. One non-classical item in this section is, instead of the usual $3+1$ spacetime split into 3 space and 1 time dimensions, I split spacetime into one radial coordinate, and time plus the two angles of spherical coordinates. The change allows the direct computation of the relevant extrinsic curvatures. In the third section, I compute the stress-energy tensors and the extrinsic curvatures for both the Schwarzschild and the Friedman-Lema\\^\\i tre metrics used in the ``Swiss cheese model.'' By a reparameterization of the Schwarzschild metric, both the intrinsic and the extrinsic curvature can be made to be continuous. I remark that the stress-energy tensor, for certain parameter choices, displays no pressure discontinuity but only a cosmic fluid density discontinuity. That discontinuity is in line with the ``Swiss cheese model'' idea that there are holes in the cosmic fluid. In the fourth section, I show that it can happen, for trajectories which approach the metric interface an near grazing angles, that the subsequent trajectories are discontinuous functions of their initial conditions. Those which enter the inner or Schwarzschild metric region can be bound in a {\\it finite sized}, closed orbit, while those which do not, travel on a parabolic trajectory to infinity. To emphasis, this case is not the same as in Newtonian orbit theory where ellipses of progressively larger size blend into parabolas, but here the ellipse is just finite in size! In the fifth section, I introduce an alternative metric. This metric is basically an adaptation of the Schwarzschild metric in curved space. I compute the stress-energy tensor. It shows an isotropic pressure, and no mass-density flux. It is only second order in magnitude in both the Hubble constant and the inverse radius of curvature of the universe. The same statement is true of the Friedmann-Lema\\^\\i tre stress energy tensor. In addition I have computed the extrinsic curvature. Both the stress-energy tensor and the extrinsic curvature are continuous, outside the Schwarzschild radius of course, as they come from infinitely differentiable expressions. In the sixth section, I compute the equations of motion for a freely moving test particle in the alternative metric. I then transform them to a coordinate system at rest at the center of the mass concentration. The flat space, slowly moving particle, weak gravitational field limit of these equations of motion are also given. The only correction to Newton's equations of motion in this limit is a term proportional to the square of Hubble's constant, $H_0$. In the final section, I gives some examples of the dynamics found using my alternative metric. ", "conclusions": "" }, "0003/astro-ph0003478_arXiv.txt": { "abstract": "We observe G~117--B15A, the most precise optical clock known, to measure the rate of change of the main pulsation period of this blue-edge DAV white dwarf. Even though the obtained value is only within 1~$\\sigma$, $\\dot P = (2.3 \\pm 1.4) \\times 10^{-15} \\,\\rm s/s$, it is already constraining the evolutionary timescale of this cooling white dwarf star. ", "introduction": "We report our continuing study of the star G~117--B15A, also called RY LMi, and WD0921+354, one of the hottest of the pulsating white dwarfs with hydrogen atmospheres, the DAV or ZZ~Ceti stars (McGraw 1979). McGraw {\\&} Robinson (1976) found the star was variable, and Kepler \\etal\\ (1982) studied its light curve and found 6 pulsation modes. The dominant mode is at 215~s, has a fractional amplitude of 22 mma, and is stable in amplitude and phase. The other, smaller pulsation modes vary in amplitude from night to night (Kepler \\etal\\ 1995). Because the DAVs appear to be normal stars except for their variability (Robinson 1979, Bergeron {\\etal} 1995), it is likely that the DAV structural properties are representative of {\\it all} DA white dwarfs. The rate of change of a pulsation period is directly related to the evolutionary timescale of a white dwarf, allowing us to directly infer the age of a white dwarf since its formation. We have been working since 1975 to measure the rate of period change with time ($\\equiv \\dot P$) for the $P=215$~s periodicity of G117--B15A, and the Kepler et al. (1991) determination was $\\dot P= (12.0 \\pm 3.5) \\times 10^{-15}\\,{\\rm s/s}$, including all data obtained from 1975 through 1990. Kepler (1984) demonstrated that the observed variations in the light curve of G~117--B15A are due to non-radial {\\it g}-mode pulsations and therefore the timescale for period change is directly proportional to the cooling timescale. For comparison, the most stable atomic clocks have rates of period change of the order of $\\dot P \\simeq 2 \\times 10^{-14}$~s/s, while the most precise millisecond pulsars have $\\dot P \\simeq 10^{-20}$~s/s (Kaspi, Taylor \\& Ryba 1994 and references therein). Since the stability of a clock is measured by $P/\\dot P$, G117--B15A has the same order of stability as the most stable millisecond pulsar. G117--B15A is the first pulsating white dwarf to have its main pulsation mode index identified. The 215~s mode is an $\\ell=1$, as determined by comparing the ultraviolet pulsation amplitude (measured with the Hubble Space Telescope) to the optical amplitude (Robinson \\etal\\ 1995). Robinson \\etal\\ (1995), and Koester, Allard \\& Vauclair (1994) derive $T_{\\mathrm{eff}}$ near 12,400~K, while Bergeron \\etal\\ (1995), using a less efficient model for convection, derives $T_{\\mathrm{eff}}$=11,600~K. Bradley (1996) used the mode identification and the observed periods of the 3 largest known pulsation modes to derive a hydrogen layer mass lower limit of $10^{-6}\\,M_*$, and a best estimate of $1.5 \\times 10^{-4}\\,M_*$, assuming $k=2$ for the 215~s mode, and 20:80 C/O core mass. The core composition is constrained mainly by the presence of the small 304~s pulsation. ", "conclusions": "While it is true that the period change timescale can be proportional to the cooling timescale, other phenomena with shorter timescales can affect $\\dot P$. The cooling timescale is the longest possible one. As a corollary, if the observed $\\dot P$ is low enough to be consistent with evolution, then other processes (such as perhaps a magnetic field) are not present at a level sufficient to affect $\\dot P$. We compare the observed value of \\Pdot\\ with the range of theoretical values derived from realistic evolutionary models with $C/O$ cores subject to {\\it g}--mode pulsations in the temperature range of G117--B15A. The adiabatic pulsation calculations of Bradley (1996), and Brassard \\etal\\ (1992,1993), which allow for mode trapping, give \\Pdot $\\simeq (2-7) \\times 10^{-15} \\,\\rm s/s$ for the $\\ell=1$, low {\\it k} oscillation observed. The observed 3$\\sigma$ upper limit, $\\dot P \\leq 6.5 \\times 10^{-15} \\,\\rm s/s$, corresponding to a timescale for period change of $P/\\dot P \\leq 1.2 \\times 10^9$ yr, equivalent to 1~s in $6 \\times 10^6$~yr, is within the theoretical predictions and very close to it. Our upper limit to the rate of period change brings us to realms where reflex motion from the proper motion companion, if they form a physical binary, or an unseen orbiting planet is of the same order as the evolutionary timescale. The effect of proper motion of the star itself is only a few times smaller. These two effects must therefore be accurately measured. We are on the way to {\\it measure} the evolutionary time scale for this lukewarm white dwarf, but the observed phase scatter of the order of 1.8~s increased the baseline necessary for a measurement. This scatter is still present in our measurement." }, "0003/astro-ph0003364_arXiv.txt": { "abstract": "In this paper we propose a quintessence model with the potential $V(\\Phi )=V_{o}\\left[ \\sinh {(\\alpha \\,\\sqrt{\\kappa _{o}}\\Delta \\Phi })\\right] ^{\\beta }$, which asymptotic behavior corresponds to an inverse power-law potential at early times and to an exponential one at late times. We demonstrate that this is a tracker solution and that it could have driven the Universe into its current inflationary stage. The exact solutions and the description for a complete evolution of the Universe are also given. We compare such model with the current cosmological observations. ", "introduction": "This work was partly supported by CONACyT, M\\'{e}xico 119259 (L.A.U.) \\begin{references} \\bibitem{perlmutter} Perlmutter {\\it et al}. ApJ {\\bf 517}, 565 (1999). A. G. Riess {\\it et al.}, Astron.J. {\\bf 116}, 1009 (1998). \\bibitem{triangle} Neta A. Bahcall, Jeremiah P. Ostriker, Saul Perlmutter and Paul J. Steinhardt, Science {\\bf 284}, 1481-1488. Varun Sahni, Alexei Starobinsky, to appear in International Journal of Modern Physics D, also at astro-ph/9904398. \\bibitem{stein} R. R. Caldwell, Rahul Dave and Paul J. Steinhardt, Phys. Rev. Lett. {\\bf 80}, 1582 (1998). Liming Wang, R. R. Caldwell, J. P. Ostriker and Paul J. Steinhardt, astro-ph/9901388. \\bibitem{stein2} Ivaylo Zlatev, Limin Wang and Paul J. Steinhardt, Phys. Rev. Lett. {\\bf 82}, 896 (1999); P.J. Steinhardt, L. Wang and I. Zlatev, Phis. Rev. D {\\bf 59}, 123504 (1999). \\bibitem{ferr} Pedro G. Ferreira and Michael Joyce, Phys. Rev. D {\\bf 58}, 023503 (1998). \\bibitem{barr} T. Barreiro, E.J. Copeland and N.J. Nunes, Report-no: SUSX-TH-016. Available at astro-ph/9910214. \\bibitem{silviu} Silviu Podariu and Bharat Ratra, astro-ph/9910527. \\bibitem{peebles} P.J.E. Peebles and B. Ratra, Astrophys. L. Lett. {\\bf 325}, L17 (1988); B. Ratra and P.J.E. Peebles, Phys. Rev. D {\\bf 37}, 3406 (1988). \\bibitem{varun} Varun Sahni and Liming Wang, astro-ph/9910097. \\bibitem{urena} T. Matos, F. S. Guzm\\'an and L. A. Ure\\~na-L\\'opez, to be published in Classical and Quantum Gravity. Also available at astro-ph/9908152. \\bibitem{chimen} Luis P. Chimento, Alejandro S. Jakubi, Int. J. Mod. Phys. {\\bf D5}, (1996)71. \\bibitem{seljak} U. Seljak and M. Zaldarriaga, ApJ {\\bf 469}, 437 (1996). \\bibitem{belen} R. Bel\\'en Barreiro, to be published in New Astronomy Reviews. Also at astro-ph/9907094. \\bibitem{chung} Chung-Pei Ma {\\it et al}, Astrophys.J. {\\bf 521}, L1-L4(1999). \\bibitem{peacock} J. Peacock, Mon. Not. R. Astron. Soc. {\\bf 284}, 885 (1997). \\end{references} \\begin{figure}[h] \\centerline{ \\epsfysize=5cm \\epsfbox{omegas.eps}} \\label{fig:omegas} \\end{figure} \\begin{figure}[h] \\centerline{ \\epsfysize=5cm \\epsfbox{modelsnia.eps}} \\label{fig:snia} \\end{figure} \\begin{figure}[h] \\centerline{ \\epsfysize=5cm \\epsfbox{lamquint.eps}} \\label{fig:lamquint} \\end{figure} \\begin{figure}[h] \\centerline{ \\epsfysize=5cm \\epsfbox{power.eps}} \\label{fig:power} \\end{figure} \\begin{figure}[h] \\centerline{ \\epsfysize=5cm \\epsfbox{potential.eps}} \\label{fig:pot} \\end{figure} \\newpage ", "conclusions": "" }, "0003/astro-ph0003222_arXiv.txt": { "abstract": "We report on the complex structure of the red giant branch (RGB) of $\\omega$ Cen, based on a new wide field and wide color baseline $B$ and $I$ photometry. Our color magnitude diagram (CMD) shows the presence of multiple populations along this branch, in particular we discovered an anomalous branch (RGB-a), which appears to be well separated from the bulk of the RGB stars. On the basis of our CMD and from the previous literature we conclude that (1) these stars, clearly identified as a separate population in our CMD, represent the extreme metal rich extension ($[Ca/H]>-0.3$) of the stellar content of $\\omega$ Cen, and show anomalous abundances of s-process elements (as $Ba$ and $Zr$) as well; (2) they are physical members of the $\\omega$ Cen system; (3) they comprise $\\sim 5\\%$ of the stars of the whole system; (4) this component and the metal-intermediate one ($-0.4>[Ca/H]>-1$) have been found to share the same spatial distribution, both of them differing significantly from the most metal poor one ($[Ca/H]<-1$). This last evidence supports the hypothesis that metal rich components could belong to an independent (proto?) stellar system captured in the past by $\\omega$ Cen. ", "introduction": "\\label{sec:intro} The globular cluster $\\omega$ Centauri (NGC~5139) is the most luminous and massive object among the Galactic Globular Clusters (GGC), and surely the most peculiar one in terms of structure, kinematics and stellar content. It is the most flattened GGC, displaying also a decrease of ellipticity in the most internal region (Geyer, Nelles \\& Hopp 1983 - GNH83), and it has a significant rotation (Merrit, Meylan \\& Mayor 1997). The most interesting anomaly is its chemical inhomogeneity (first revealed by Dickens and Woolley 1967 and spectroscopically confirmed by Freeman \\& Rodgers 1975); since then a number of extensive spectroscopic surveys (Norris, Freeman \\& Mighell 1996 - NFM96, Suntzeff \\& Kraft 1996 - SK96) have shown that $\\omega$ Cen is the only GGC for which a multi-component heavy element distribution has been identified. Although SK96 found a single peaked distribution with an extended tail towards high metallicities, NFM96 showed that the distribution is at least bimodal with a main metal poor component at $[Fe/H]\\simeq -1.6$, a second smaller peak at $[Fe/H]\\simeq -1.2$, and a long tail extending up to $[Fe/H]\\simeq -0.5$. Furthermore, the most metal rich stars ($[Ca/H]>\\sim -1.0$) have been found to be more centrally concentrated than the bulk of the cluster population (Norris et al. 1997 - N97, and SK96). It has also been suggested that the kinematical properties (N97) and the spatial distribution of the two metallicity groups differ significantly (N97, Jurcsik 1998 - J98). This puzzling scenario has usually been explained either in terms of self-enrichment processes (Freeman 1993) and/or merging events (N97, J98). As part of a long term project specifically devoted to the study of the global stellar population in a sample of GGCs, we obtained wide field $B$, $I$ photometry in $\\omega$ Cen. The complete data set will be presented in a forthcoming paper, while in this letter we concentrate on the complex structure of the RGB. \\begin{figure*}[htb] \\vskip4truein \\special{psfile=fig1.ps voffset=-100 hoffset=40 vscale=60 hscale=65} \\caption[fig1.ps]{ $(B,B-I)$ CMDs for $\\omega$ Cen. Eight CMDs are plotted: one for each of the eight WFI mosaic chips. The cluster center is located in Chip \\#2 ({\\it panel (b)}). The arrows indicate the anomalous RGB. \\label{fig:map}} \\end{figure*} ", "conclusions": "" }, "0003/hep-th0003086_arXiv.txt": { "abstract": " ", "introduction": "The Pre-Big-Bang (PBB) model \\footnote{For an updated collection of papers on this model see http://www.to.infn.it/$\\sim$gasperin.} is based on the low energy effective action of string theory, which in $d$ spatial dimensions is given by \\be\\label{eq:lea} S = - \\frac{1}{2\\l_s^{d-1}} \\int d^{d+1}x \\, \\sqrt{-g} \\, e^{-\\varphi} \\, \\left[ R + (\\d_\\m \\varphi)^2 + \\cdots \\, \\right] \\,, \\ee where $\\varphi$ is the dilaton and $\\l_s$ is the string length scale. The qualitative differences between the PBB model, and the SBB model based on the Einstein-Hilbert action, \\be\\label{eq:eh} S = - \\frac{1}{2\\l_p^{d-1}} \\int d^{d+1}x \\, \\sqrt{-g} \\, R \\,, \\ee are most easily visualized by plotting the history of the curvature of the universe (see \\fig{pbb.eps}) according to each theory. In the SBB scenario the curvature increases as we go back in time, eventually reaching an infinite value at the Big-Bang singularity. In standard inflationary models the curvature reaches some fixed value as $t$ decreases at which point the universe enters a de Sitter phase. It has been shown however that such an inflationary phase cannot last forever, for reasons of geodesic completeness, and that the initial singularity problem still remains \\cite{ref:bordeet1993,ref:brandenberger1999}. The cosmology generated by (\\ref{eq:lea}) differs drastically from the standard scenarios. The action (\\ref{eq:lea}) without the ``$\\cdots$'' terms does not realize the PBB scenario, as we will discuss below. In the PBB model, as one travels back in time the curvature increases as in the previously mentioned models, but in the PBB a maximum curvature is reached at which point the curvature and temperature actually begin to {\\it decrease}. Although we will examine the details of how this occurs below, a few simple considerations make us feel more comfortable with this picture. For one, string theory predicts a natural cut-off length scale, \\be\\label{eq:lstring} \\l_s = \\sqrt{\\frac{\\hbar}{T}} \\sim 10 \\, l_{pl} \\sim 10^{-32} \\cm \\,, \\ee where $T$ is the string tension and $l_{pl}$ is the Planck length. So it is natural from the point of view of strings to expect a maximum possible curvature. Logically, as we travel back in time there are only two possibilities if we want to avoid the initial singularity. Either the curvature starts to grow again before the de Sitter phase, in which case we are still left with a singularity shifted earlier in time, or the curvature begins to decrease again, which is what happens in the PBB scenario (\\fig{pbb.eps}c). This behavior is a consequence of scale-factor duality. \\vspace{.75cm} \\hglue 1cm \\psfig{figure=pbb.eps,height=8.5cm,angle=-90}\\label{pbb.eps} \\begin{quote} \\scriptsize Figure \\ref{pbb.eps}: Curvature plotted versus time for, (a) the SBB model, (b) the standard inflationary model and (c) the PBB scenario. \\end{quote} \\subsection{More on Duality}\\label{mored} To demonstrate the enhanced symmetries present in the PBB model we will examine the consequences of scale-factor duality. The Einstein-Hilbert action (\\ref{eq:eh}) is invariant under time reversal. Hence, for every solution $a(t)$ there exists a solution $a(-t)$. Or in terms of the Hubble parameter $H(t) = \\dot a(t) / a(t)$, for every solution $H(t)$ there exists a solution $- H(-t)$. Thus, if there is a solution representing a universe with decelerated expansion and decreasing curvature ($H>0$, \\, $\\dot H < 0$) there is a ``mirror\" solution corresponding to a contracting universe ($H(-t)$, \\, $H < 0$). The action of string theory (\\ref{eq:lea}) is not only invariant under time reversal, but also under inversion of the scale factor $a(t)$, (with an appropriate transformation of the dilaton). For every cosmological solution $a(t)$ there is a solution $\\tilde a = 1/a(t)$, provided the dilaton is rescaled, $\\varphi \\rightarrow \\tilde \\varphi = \\varphi - 2d\\, \\ln a$. Hence, time reversal symmetry together with scale-factor duality imply that every cosmological solution has four branches, \\fig{branch.eps}. For the standard scenario of decelerated expansion and decreasing curvature ($H(t) > 0$, \\, $\\dot H(t) < 0$ ) there is a dual partner solution describing a universe with accelerated expansion parameter $\\tilde H(t)$ and growing curvature $\\dot {\\tilde H}(- t)$. \\vspace{.75cm} \\hglue 3cm \\psfig{figure=branch.eps,height=8cm,angle=-90}\\label{branch.eps} \\begin{quote} \\scriptsize Figure \\ref{branch.eps}: The four branches of a string cosmological solution resulting from scale-factor duality and time reversal. \\end{quote} We will now show how one can create a universe from the string theory perturbative vacuum, that today looks like the standard cosmology. This problem is analogous to finding a smooth way to connect the Pre-Big-Bang phase with a Post-Big-Bang phase, or how to successfully connect the upper-left side of \\fig{branch.eps} to the upper-right side. In general, the two branches are separated by a future/past singularity and it appears that in order to smoothly connect the branches of growing and decreasing curvature one requires the presence of higher order loop and/or derivative corrections to the effective action (\\ref{eq:lea}). This cancer of the PBB model is know as the Graceful Exit Problem (GEP) and is the subject of many research papers (see \\cite{ref:veneziano2000,ref:lidseyet1999} for a collection of references). One example of how the GEP can be solved is given in \\cite{ref:eassonet1999}. In this work we consider a theory obtained by adding to the usual string frame dilaton gravity action specially constructed higher derivative terms motivated by the limited curvature construction of \\cite{ref:mukhanovet1992}. The action is (\\ref{eq:lea}) with the ``$\\cdots$'' term being replaced by the constructed higher derivative terms. In this scenario all solutions of the resulting theory of gravity are nonsingular and for initial conditions inspired by the PBB scenario solutions exist which smoothly connect a ``superinflationary'' phase with $\\dot H > 0$, to an expanding FRW phase with $\\dot H<0$, solving the GEP in a natural way. \\subsection{PBB-Cosmology}\\label{pbbcos} Here we examine cosmological solutions of the PBB model. By adding matter in the form of a perfect fluid to the effective action (\\ref{eq:lea}) (without the ``$\\cdots$'' terms) and taking a Friedmann-Robertson-Walker background with $d=3$, we vary the action to get the equations of motion for string cosmology, \\begin{eqnarray}\\label{eq:eomstring} \\dot\\vp^2 - 6H\\dot\\vp + 6H^2 & = & e^\\vp \\r \\,, \\\\ \\dot H - H\\dot\\vp + 3H^2 & = & \\frac{1}{2} e^\\vp p \\,,\\nonumber \\\\ 2 \\ddot \\vp + 6H\\dot\\vp - \\dot\\vp^2 - 6\\dot H - 12H^2 & = & 0 \\nonumber \\,. \\end{eqnarray} As an example, for $p=\\r/3$ the equations with constant dilaton are exactly solved by \\be\\label{eq:rad} a \\propto t^{1/2}, \\qquad \\r \\propto a^{-4}, \\qquad \\vp = \\mbox{const.} \\,, \\ee which is the standard scenario for the radiation dominated epoch, having decreasing curvature and decelerated expansion: \\be\\label{eq:raddom} \\dot a > 0, \\qquad \\ddot a < 0, \\qquad \\dot H < 0 \\,. \\ee But there is also a solution obtained from the above via time translation and scale-factor duality, \\be\\label{eq:trans} t \\rightarrow -t, \\qquad a \\propto (-t)^{-1/2}, \\qquad \\vp \\propto -3 \\ln(-t), \\qquad \\r = -3p \\propto a^{-2} \\,. \\ee This solution corresponds to an accelerated, inflationary expansion, with growing dilaton and growing curvature: \\be\\label{eq:supinf} \\dot a > 0, \\qquad \\ddot a >0, \\qquad \\dot H > 0 \\,. \\ee Solutions with such behavior are called ``superinflationary\" and are located in the upper left quadrant of \\fig{branch.eps}. Let us briefly review the history of the universe as predicted by the PBB scenario. Recall, that in the SBB model the universe starts out in a hot, dense and highly curved regime. In contrast, the PBB universe has its origins in the simplest possible state we can think of, namely the string perturbative vacuum. Here the universe consists only of a sea of dilaton and gravitational waves. It is empty, cold and flat, which means that we can still trust calculations done with the classical, low-energy effective action of string theory. In \\cite{ref:kaloperet1999}, the authors showed that in a generic case of the PBB scenario, the universe at the onset of inflation must already be extremely large and homogeneous. In order for inflation to solve flatness problems the initial size of a homogeneous part of the universe before PBB inflation must be greater than $10^{19}l_s$. In response, it was proposed in \\cite{ref:buonannoet1998} that the initial state of the PBB model is a generic perturbative solution of the tree-level, low-energy effective action. Presumably, quantum fluctuations lead to the formation of many black holes (\\fig{bubbl.eps}) in the gravi-dilaton sector (in the Einstein frame). Each such singular space-like hypersurface of gravitational collapse becomes a superinflationary phase in the string frame \\cite{ref:feinsteinet2000, ref:ghoshet2000,ref:bozzaet2000,ref:buonannoet1998}. After the period of dilaton-driven inflation the universe evolves in accordance with the SBB model. \\vspace{.5cm} \\hglue 2cm \\psfig{figure=bubbl.eps,height=4cm,angle=-90}\\label{bubbl.eps} \\begin{quote} \\scriptsize Figure \\ref{bubbl.eps}: A $2+1$ dimensional slice of the string perturbative vacuum giving rise to black hole formation in the Einstein frame. \\end{quote} To conclude let us mention a few benefits of the PBB scenario. For one, there is no need to {\\it invent} inflation, or fine tune a potential for the inflaton. This model provides a ``stringy\" realization of inflation which sets in naturally and is dilaton driven. Pair creation (quantum instabilities) provides a mechanism to heat up an initially cold universe in order to produce a hot big-bang with homogeneity, isotropy and flatness. This scenario also has observable consequences. Problems with this scenario include the graceful exit problem, mentioned above. This is the problem of smoothly connecting the phases of growing and decreasing curvature, a process that is not well understood and requires further investigation. Most cosmological models require a potential for the dilaton to be introduced by hand in order to freeze the dilaton at late times. In general it is believed that the dilaton should be massive today, otherwise we would notice its effects on physical gauge couplings. Inclusion of a non-vanishing $B_{\\mu\\nu}$ into the action (\\ref{eq:lea}) greatly reduces the initial conditions which give rise to inflation \\cite{ref:lidseyet1999}. Also the initial collapsing region must be sufficiently large and weakly coupled. Lastly, the dimensionality problem is still present in this model. ", "conclusions": "In this review we have discussed a number of intriguing approaches to string and M-theory cosmology. While the past few years have shown a considerable increase in our understanding of M-theory, there is still plenty of room for further research. Perhaps the greatest advances have come from the discovery of duality symmetries in the M-theory moduli space, D-branes, the AdS/CFT correspondence and the development of Matrix theory. As demonstrated in this review we have taken the first steps to incorporate this new knowledge into cosmology. M-theory provides an innovative framework in which to study the early Universe and to search for alternatives to the Standard Big-Bang and Inflationary models. Conversely, cosmology is essential to our study of M-theory, since couplings and masses set by the vacuum state of string theory must agree with those observed in our Universe. The amalgamation of M-theory and cosmology may reveal the answers to a number of tantalizing questions and provide the tools to probe the earliest moments of creation." }, "0003/hep-ph0003190_arXiv.txt": { "abstract": " ", "introduction": "Until a few years ago cosmology with scalar fields was almost synonymous with cosmological inflation. Recently there has been an enormous upsurge in interest in the possibility that scalar fields can play an important role in the dynamics of the Universe at recent epochs, mainly due to the observations of the apparent magnitudes of distant supernovae \\cite{Supernovae} which may be explained by the presence of such a component \\cite{quintessence}. In this context it is certainly interesting to consider what the role of such fields can be at other epochs, and in particular how their behaviour between the end of inflation and their reappearance today might influence cosmology in the intervening period. This question is also related to the `fine-tuning' problem associated with such scenarios: how is it that such a field can give a significant contribution to the energy density today starting from a natural set of initial conditions after inflation? This apparent problem is in fact resolved in a wide class of potentials \\cite{jf,tracking} which generically have the property that in some part of the potential they may support modes which are dominated by the kinetic energy of the scalar field, so that their energy density scales away faster than that in the radiation, {\\it i.e.} $\\rho_\\phi\\propto a^{-n}$, with $42$, or equivalently an equation of state $p_\\phi=w_\\phi\\rho_\\phi$, with $w_\\phi>-1/3$) can be envisaged, with the case $n=3$ corresponding to the most standard reheating during the oscillation of the inflaton about the minimum of a quadratic potential. There is a continual release of entropy until the radiation dominated epoch, leading to a dilution of most relevant physical quantities sourced during the scalar dominated phase. In the latter case, which corresponds to domination by the kinetic energy of a homogeneous scalar field (or equivalently to an equation of state $p_\\phi=w_\\phi\\rho_\\phi$, with $w_\\phi>1/3$ ) the scalar field simply redshifts away until it becomes the sub-dominant component. There is no entropy release, and correspondingly a coherent energy remains in the scalar field which, given an appropriate potential (the `self-tuning' potentials of \\cite{jf}, or the `tracking' potentials of \\cite{tracking}) can become relevant again at late times \\cite{spokoiny,jp,peeblesvilenkin}. In \\cite{mj,jp} we have considered in a generic way the effect of a change in the expansion rate prior to nucleosynthesis on models of electroweak baryogenesis\\footnote{The effects on dark matter freeze-out can be inferred from the work of \\cite{Barrow, KamTurner}, who studied mainly modifications associated with anisotropy in the expansion.}, in particular on the effect on the sphaleron bound and the `no-go' theorem for electroweak baryogenesis in the case of a second order phase transition. As concrete realizations of such cosmologies we considered models which go through an epoch after inflation - which, following \\cite{mj} we termed `kination' - of domination by a kinetic mode of a scalar field. This occurs most naturally in a model in which the universe `reheats' not by the decay of the inflaton, but by gravitational particle creation at the end of the inflationary epoch \\cite{ford, spokoiny}. In a recent paper \\cite{dlr} it has been observed that, for low (sub-electroweak) reheat temperatures in more traditional models of reheating - in which the inflaton decays while oscillating in a mode with matter scaling after inflation - the effects discussed in \\cite{mj,jp} on electroweak cosmology also result. There is in this case an even larger relative boost to the expansion rate (see below), but a very large entropy release which tends to undo any of the enhancing effects of the greater expansion rate. In \\cite{tp} one of us (TP) has considered the general case of a decaying inflaton evolving in a mode scaling as $1/a^n$, and shown that, while the same larger boost to the expansion rate occurs as in the $n=3$ case of \\cite{dlr}, the entropy release problem is greatly reduced as the kinetic mode $n=6$ limit is attained. Here we concentrate on another aspect of such alternative cosmologies, which is a simple consequence of the observation which has been made in \\cite{js,jp,dlr}: Because of the enhanced expansion rate, the right-handed electrons of the standard model may remain out of equilibrium until a temperature below the electroweak phase transition. It is well known that asymmetry in right-handed electrons - because of their late equilibration time - may be important from at least two points of view: $\\bullet$ Since right-handed electrons couple to other particles in the standard model with only an extremely small Yukawa coupling, they remain out of equilibrium in an expanding Universe until relatively late - in the standard radiation dominated cosmology until $T \\gsim 20$TeV \\cite{cdeo, cko}. A pre-existing baryon asymmetry can survive the effect of standard model anomalous processes - which violate $B+L$ and are unsuppressed until the electroweak phase transition - only if there are non-zero CP-odd conserved global charges when they are operative. In the absence of such charges the equilibrium attained will be CP invariant with zero baryon number. As noted in \\cite{cko} above \\footnote{The scale quoted in \\cite{cko} is $10$GeV. The increase by a factor 2 is due to a tighter bound on the Higgs mass.} $20$TeV right electron number $e_R$ is in fact such an effective charge, and as a result other global charges like $B-L$ can be violated until close to this scale. This leads \\cite{cko} to a very significant reduction in the bounds on $B-L$ violating interactions in grand unified theories with the structure appropriate for them to generate baryon asymmetry. Here the consequences are much simpler and more dramatic: If the $e_R$ remain out of equilibrium all the way until the electroweak scale, a baryon number will result from this due to the $B+L$ violating processes. When the electroweak scale is reached this baryon number will simply be frozen when the $B+L$ violating processes abruptly switch off. This will be the case irrespective of whether there is primordial $B$ or $L$ (or $B-L$), and irrespective of whether these charges are violated or conserved. Just like in the case of electroweak baryogenesis all the non-trivial physics required is in principle present in the standard model. The problem of baryogenesis then becomes posed as what we will refer to as `electrogenesis', the generation of the source right handed electrons prior to the time at which the $B+L$ violating processes become suppressed. It is this process which we discuss below. $\\bullet$ The effective conservation of $e_R$ in the early Universe due to the fact that its perturbative decay channel is out of equilibrium is not exact, because the $e_R$ charge has an axial anomaly under the U(1) of hypercharge. There are no degenerate vacua as in the non-abelian case, but there are finite energy modes of the U(1) field with Chern-Simons number which can `eat' the charge. In fact, as discussed in~\\cite{js, giovshap} this leads to an instability at finite density to the formation of long wavelength modes of hypermagnetic field. When these modes come inside the horizon they can evolve during the time in which the right electron number is without its perturbative decay channel. Here this scenario will be modified as a result of the change in the expansion rate, since the perturbative channel does not come into play until the electroweak scale, at which time a first order phase transition may produce the turbulence needed to amplify the produced seed magnetic fields. ", "conclusions": "We have considered here one aspect of cosmologies in which a scalar field dominates the expansion rate prior to nucleosynthesis. Right-handed electrons may remain out of equilibrium until the electroweak scale, so that if they are generated the $B+L$ processes of the standard model will lead to a non-zero equilibrium density of baryons of the order of that in the $e_R$. We have discussed two kinds of post-inflationary cosmologies in which such a period of scalar field domination can occur: in the first the inflaton rolls away after inflation into a kinetic mode in a steep potential, and the Universe is `reheated' by the gravitational particle production at the end of the inflationary epoch, while in the second the inflaton rolls into a mode which can have a range of scalings and reheats the Universe itself by decaying sufficiently slowly to give a very low reheat temperature. We studied a specific model for the generation of the right handed electron asymmetry in which there are a set of scalars with CP-violating (and flavour changing) couplings to the leptons. We showed that in both scenarios such scalars can decay out of equilibrium at quite low temperatures and produce the desired asymmetry. While our models strongly favoured the case of kinetic mode domination, which have little or no entropy release, we note that in certain very special circumstances which may be satisfied in other models the generation of the observed baryon asymmetry may still be possible in the standard reheating scenario (with matter scaling during the reheating epoch). Finally we considered briefly the effect of the abelian anomaly which destabilizes such charges, and concluded that in the models with kinetic mode domination this effect is unimportant for the baryon number generation, while in the case with large entropy dilution it may be important and might allow the generation of magnetic field as in the case of standard radiation domination. Finally we return to the question of how this kind of mechanism might be implemented in other particle physics models, in particular in more popular (e.g. supersymmetric) extensions of the standard model. In general one need not consider necessarily the generation of right-handed electron number, but the generation of {\\it any CP odd charge which is effectively conserved after its creation on a timescale which is longer than the expansion rate of the Universe at the electroweak scale} (when the $B+L$ violating processes freeze-out). Given that the expansion rate at the electroweak scale can be enhanced in these models by many orders of magnitude -- up to a rate $\\sim 10^{-11}T_{\\rm ew}$ in models of type A, and $\\sim 10^{-6} T_{\\rm ew}$ in models of type B -- scenarios can be considered in which many of the lighter degrees of freedom will drop out of equilibrium (for example the lighter right-handed quarks). While in the standard model itself there would seem to be the obstacle of prohibitively small CP violation, in extensions there is generically new CP-violating structure in the added sectors ({\\it e.g.} in the chargino and squark mass matrices of the minimal supersymmetric standard model). The problem of baryogenesis then becomes the problem of the generation prior to the electroweak scale of CP-odd approximately conserved charge using this structure. Given our observation that for very modest masses (as low as a TeV for a particle with a coupling of order one) the decay of these heavier particles occurs out of equilibrium in these cosmologies, there is clearly the interesting possibility of sourcing CP-odd charges in this way, thus creating a baryon asymmetry. We will treat these issues in detail in forthcoming work \\cite{jp2}." }, "0003/astro-ph0003236_arXiv.txt": { "abstract": "{Using the exceptional long-term monitoring capabilities of the MACHO project, we present here the optical history of \\source\\ for a continuous 6-yr period. These data were used to investigate the previously claimed periodicities for this source of 8.15 h and 12.54 d : we find upper amplitude limits of 0.10 mag and 0.09 mag, respectively.} ", "introduction": "\\source\\ was discovered in the Large Magellanic Cloud (LMC) by early satellite flights (Leong et al.\\ 1971) which showed it to be a variable X-ray source with \\lx\\til(0.6--3)x10$^{38}$\\ergsec\\ (Markert \\& Clark 1975; Johnston, Bradt \\& Doxsey 1979; Long, Helfand \\& Grabelsky 1981). \\source\\ is the most luminous low mass X-ray binary (LMXB) known. After its precise position was determined (Johnston et al.\\ 1979), the optical counterpart was identified as a faint, {\\it V} \\til18.8, blue star similar to Sco~X--1 (Pakull 1978; Pakull \\& Swings 1979). Motch et al.\\ (1985) found variations of \\til0.4 mag in the optical light curve of \\source, which were modulated on a period of 6.4 h (see also Bonnet-Bidaud et al.\\ 1989). More extensive photometric observations obtained by Callanan et al.\\ (1990) showed flickering of \\til0.1 mag on timescales of \\leq0.5 h and \\til0.3 mag variations on \\til4 h timescales. Period searching revealed a period of 8.15$^{+}_{-}$0.02 h with a semi-amplitude of \\til0.08 mag which, being stable over their \\til14 d observations, they interpreted as being orbital in origin. However, Crampton et al.\\ (1990) suggested a much longer orbital period of 12.54 d with a semi-amplitude of \\til0.5 mag from their photometric observations that spanned nine nights. This substantial disagreement over LMC X-2's fundamental parameter, its orbital period, has still not been resolved and is, of course, crucial to any detailed interpretation of the source. \\subsection{Is the Orbital Period 8.15 h?} If an orbital period of 8.15 h is correct, we would expect a significant contribution from a heated secondary or outer disc bulge, and this should be evident from the shape of the folded light curve (Callanan et al.\\ 1990). However, as Crampton et al.\\ (1990) did not detect the 8.15-h modulation, it implies that the amplitude of this modulation is itself variable on timescales \\geq10 days. If it is owing to X-ray heating of the secondary, a semi-amplitude of 0.08 mag indicates that i\\leq70\\deg\\ (van Paradijs, van der Klis \\& Pederson 1988). \\subsection{Or is it 12.54 d?} If a 12.54-d period is correct, taking into account the structure and asymmetry of the light curve obtained, \\source\\ would be similar to Cyg X-2 (Cowley, Crampton \\& Hutchings 1979). This would suggest that the variations in the \\source\\ light curve might contain both a heated component and a partial eclipse of the accretion disc by the secondary. Irregularities in the disc cause some of the pre-eclipse modulation that gives rise to pre-eclipse dips and variability as observed in other LMXBs (see White 1989). In a long-period LMXB, only an evolved star can fill the Roche lobe, and the high X-ray luminosity of \\source\\ may be due to the high rate of mass transfer on to the neutron star as the companion evolves along the giant branch (Webbink, Rappaport \\& Savonije 1983). The period found could also be as a result of disc precession rather than orbital variations, or a beat between a photometric and orbital period (Crampton et al.\\ 1990). Short periodicities must also be present for this to be true, and although none were found by Crampton et al.\\ (1990), the periods found by Callanan et al.\\ (1990) and Bonnet-Bidaud et al.\\ (1989) may therefore be real. Evidence for the long period would be the direct detection of the secondary in the spectrum (cf. Cyg X--2; Casares, Charles \\& Kuulkers 1998 and references therein), but thus far the optical spectra show no absorption features that can be attributed to the companion star (Bonnet-Bidaud et al.\\ 1989). Lack of detection could be as a result of the extreme brightness of the accretion disc. This may be owing to the high luminosity fuelled by the higher rate of mass transfer allowed by the lower metal abundances in the LMC, leading to a higher Eddington luminosity (Russell \\& Dopita 1990). No correlation between X-ray and optical light curves has yet been detected that would help in the identification of the period. We present here the results of \\til6 years of optical monitoring of the source. These observations were acquired as a by-product of the MACHO project (Alcock et al.\\ 1995a), owing to the serendipitous location of \\source\\ in a surveyed field. Such extended monitoring is ideal for investigating modulations on timescales of tens of days. ", "conclusions": "The results from the MACHO data and the ESO light curve indicate that the orbital period of \\source\\ is not 12.54-d. There is however evidence of non-periodic variations on time-scales of \\til tens of days which would explain the results found by Crampton et al.\\ (1990). Long term variations occur in other LMXBs (cf Cyg X-2; Smale \\& Lochner 1992; Wijnands, Kuulkers \\& Smale 1996) which are not associated with the orbital period of the system. In high mass X-ray binaries (HMXB) {\\it superorbital} periods are known for several sources. Theories to explain these long term variations include precession of a tilted accretion disc, precession of the neutron star, mass transfer feedback and triple systems (see Priedhorsky \\& Holt 1987 and Schwarzenberg-Czerny 1992). In LMXBs such {\\it superorbital} periods are much less common. It is thought that they may be due to radiation-driven warped accretion discs (e.g. Wijers \\& Pringle 1999) or a disc instability in the system (Priedhorsky \\& Holt 1987). The light curves from the MACHO and ESO data do not confirm the 8.2-h period for \\source\\ found by Callanan et al.\\ (1990). However, had this periodicity been present at the Callanan et al.\\ (1990) amplitude we could not have detected it (as it is below our upper limit) and so this modulation still awaits confirmation." }, "0003/gr-qc0003101_arXiv.txt": { "abstract": " ", "introduction": "\\label{intro} The description of important areas of modern astronomy, such as high-energy astrophysics or gravitational wave astronomy, requires General Relativity. High energy radiation is often emitted by highly relativistic events in regions of strong gravitational fields near compact objects such as neutron stars or black holes. The production of relativistic radio jets in active galactic nuclei, explained by pure hydrodynamical effects as in the twin-exhaust model \\cite{blandford74}, by hydromagnetic centrifugal acceleration as in the Blandford-Payne mechanism~\\cite{blandford82}, or by electromagnetic extraction of energy as in the Blandford-Znajek mechanism~\\cite{blandford77}, involves an accretion disk around a rotating supermassive black hole. The discovery of kHz quasi-periodic oscillations in low-mass X-ray binaries extended the frequency range over which these oscillations occur into timescales associated with the relativistic, innermost regions of accretion disks (see, e.g.~\\cite{vanderklis98}). A relativistic description is also necessary in scenarios involving explosive collapse of very massive stars ($\\sim 30 M_{\\odot}$) to a black hole (in the so-called collapsar and hypernova models), or during the last phases of the coalescence of neutron star binaries. These catastrophic events are believed to exist at the central engine of highly energetic $\\gamma$-ray bursts (GRBs) \\cite{paczynski86,narayan92,woosley93,paczynski98}. In addition, non spherical gravitational collapse leading to black hole formation or to a supernova explosion, and neutron star binary coalescence are among the most promising sources of detectable gravitational radiation. Such astrophysical scenarios constitute one of the main targets for the new generation of ground-based laser interferometers, just starting their gravitational wave search (LIGO, VIRGO, GEO600, TAMA)~\\cite{thorne96,new02}. A powerful way to improve our understanding of the above scenarios is through accurate, large scale, three-dimensional numerical simulations. Nowadays, computational general relativistic astrophysics is an increasingly important field of research. In addition to the large amount of observational data by high-energy X- and $\\gamma$-ray satellites such as Chandra, XMM-Newton or INTEGRAL, and the new generation of gravitational wave detectors, the rapid increase in computing power through parallel supercomputers and the associated advance in software technologies is making possible large scale numerical simulations in the framework of general relativity. However, the computational astrophysicist and the numerical relativist face a daunting task. In the most general case, the equations governing the dynamics of relativistic astrophysical systems are an intricate, coupled system of time-dependent partial differential equations, comprising the (general) relativistic (magneto-) hydrodynamic (MHD) equations and the Einstein gravitational field equations. In many cases, the number of equations must be augmented to account for non-adiabatic processes, e.g., radiative transfer or sophisticated microphysics (realistic equations of state for nuclear matter, nuclear physics, magnetic fields, etc.). Nevertheless, in some astrophysical situations of interest, e.g., accretion of matter onto compact objects or oscillations of relativistic stars, the `test-fluid' approximation is enough to get an accurate description of the underlying dynamics. In this approximation the fluid self-gravity is neglected in comparison to the {\\it background} gravitational field. This is best exemplified in accretion problems where the mass of the accreting fluid is usually much smaller than the mass of the compact object. Additionally, a description employing ideal hydrodynamics (i.e., with the stress-energy tensor being that of a perfect fluid), is also a fairly standard choice in numerical astrophysics. The main purpose of this review is to summarize the existing efforts to solve numerically the equations of (ideal) general relativistic hydrodynamics. To this aim, the most important numerical schemes will be first presented in some detail. Prominence will be given to the so-called Godunov-type schemes written in conservation form. Since~\\cite{marti91} it has been gradually demonstrated~(\\cite{font94, eulderink95,romero96,falle96,banyuls97,wen97,pons98}) that conservative methods exploiting the hyperbolic character of the relativistic hydrodynamic equations are optimally suited for accurate numerical integrations, even well inside the ultrarelativistic regime. The explicit knowledge of the characteristic speeds (eigenvalues) of the equations, together with the corresponding eigenvectors, provides the mathematical (and physical) framework for such integrations, by means of either exact or approximate Riemann solvers. The article includes, furthermore, a comprehensive description of `relevant' numerical applications in relativistic astrophysics, including gravitational collapse, accretion onto compact objects and hydrodynamical evolution of neutron stars. Numerical simulations of strong-field scenarios employing Newtonian gravity and hydrodynamics, as well as possible post-Newtonian extensions, have received considerable attention in the literature and will not be covered in the review, which focuses in relativistic simulations. Nevertheless, we must emphasize that most of what is known about hydrodynamics near compact objects, in particular in black hole astrophysics, has been accurately described using Newtonian models. Probably the best known example is the use of a pseudo-Newtonian potential for non-rotating black holes which mimics the existence of an event horizon at the Schwarzschild gravitational radius~\\cite{paczynski80}, which has allowed accurate interpretations of observational phenomena. The organization of the article is as follows: Section 2 presents the equations of general relativistic hydrodynamics, summarizing the most relevant theoretical formulations which, to some extent, have helped to drive the development of numerical algorithms for their solution. Section 3 is mainly devoted to describing numerical schemes specifically designed to solve non-linear hyperbolic systems of conservation laws. Hence, particular emphasis will be paid on conservative high-resolution shock-capturing (HRSC) upwind methods based on linearized Riemann solvers. Alternative schemes such as Smoothed Particle Hydrodynamics (SPH), (pseudo-) spectral methods and others will be briefly discussed as well. Section 4 summarizes a comprehensive sample of hydrodynamical simulations in strong-field general relativistic astrophysics. Finally, in Section 5 we provide additional technical information needed to build up upwind HRSC schemes for the general relativistic hydrodynamics equations. Geometrized units ($G=c=1$) are used throughout the paper except where explicitly indicated, as well as the metric conventions of~\\cite{MTW}. Greek (Latin) indices run from 0 to 3 (1 to 3). ", "conclusions": "" }, "0003/astro-ph0003146_arXiv.txt": { "abstract": "We report the first use of Michelson interferometry on the Keck~I telescope for diffraction-limited imaging in the near infrared JHK and L bands. By using an aperture mask located close to the f/25 secondary, the 10\\,m Keck primary mirror was transformed into a separate-element, multiple aperture interferometer. This has allowed diffraction-limited imaging of a large number of bright astrophysical targets, including the geometrically complex dust envelopes around a number of evolved stars. The successful restoration of these images, with dynamic ranges in excess of 200:1, highlights the significant capabilities of sparse aperture imaging as compared with more conventional filled-pupil speckle imaging for the class of bright targets considered here. In particular the enhancement of the signal-to-noise ratio of the Fourier data, precipitated by the reduction in atmospheric noise, allows high fidelity imaging of complex sources with small numbers of short-exposure images relative to speckle. Multi-epoch measurements confirm the reliability of this imaging technique and our whole dataset provides a powerful demonstration of the capabilities of aperture masking methods when utilized with the current generation of large-aperture telescopes. The relationship between these new results and recent advances in interferometry and adaptive optics is briefly discussed. ", "introduction": "Developments in detector technology and opto-electronic hardware over the past decade have meant that real-time adaptive optical systems have now become a common feature of large ground-based optical and near-infrared telescopes (recent reviews may be found in Bonaccini \\& Tyson 1998, Hardy 1998 \\& Roddier 1999) However, while adaptive optics has enjoyed considerable recent success, other techniques that utilize post-detection data processing, rather than real-time compensation, have remained valuable for imaging at the very highest angular resolutions. The best known, and most straightforward of these to implement, is speckle imaging \\cite{Lab70,weig91,NR96} in which sequences of short-exposures of a target and an unresolved calibrator are used to recover high-resolution maps beyond the natural seeing limit. Although this method in principle allows the recovery of images of arbitrary complexity, the difficulty of attaining an adequate signal-to-noise ratio has meant that it has mainly been confined to studies of binary stars (see, for example, Patience et al.~1998) and other astronomical sources with similarly simple geometries (though Weigelt et al.~1998 is a recent counterexample). One solution to this signal-to-noise problem is to modify the pupil geometry of the telescope using a mask so as to mimic the operation of a separated-element interferometer array such as the VLBA. This process can be considered as finding an optimal balance between the level of atmospheric perturbations, the number of photons, and the amount of structural information measured about the source -- all of which increase as the pupil area rises. When an aperture mask is being used, the data collection and analysis methods are similar to those utilized for speckle imaging, but with a reduction in the number of independent spatial frequencies measured, which is balanced by an improved signal-to-noise ratio on the data which are obtained. This post-processing approach has been widely exploited at optical wavelengths where it has established itself as the only method by which reliable images of the surfaces of nearby stars at the diffraction limit have been recovered for ground-based telescopes (see, for example, Buscher et al.~1990; Wilson Dhillon \\& Haniff~1997; Tuthill Haniff \\& Baldwin~1999a). In this paper, we report the first aperture masking experiments to exploit the new generation of 10\\,m-class telescopes. We have used the the Keck~I telescope with a variety of sparse multi-aperture pupil masks both to verify the signal-to-noise and calibration advantages of these pupil geometries, and to demonstrate the ability of this method to provide diffraction-limited imaging of resolved targets in the near infrared with excellent dynamic range. We have used multi-epoch measurements to establish the reliability of our imaging, and present near infrared maps of the highly-structured dust shells of a number of evolved stars at resolutions exceeding 50\\,milli-arcseconds. ", "conclusions": "Results from the first aperture masking experiment performed on a 10\\,m class telescope are presented. A suitable choice of non- or low-redundancy pupil geometries has been found to dramatically improve the signal-to-noise on recovered bispectral data. Reliable images with complex and asymmetric structure at the diffraction limit have been routinely produced in the near-infrared JHK and L bands. With dynamic ranges in excess of 200:1, and demonstrated repeatability of map structure over multiple observing epochs, the expected advantages of sparse-aperture interferometry for bright targets have been confirmed. In a comparison of aperture masking, full-pupil speckle, and adaptive optics, the most reasonable conclusion appears that each technique has regions of the parameter space of source-brightness and spatial structure where it offers superior performance. The existence of such complementary observational techniques will certainly be beneficial in addressing a range of problems in high resolution astronomy, with masking being at its most effective for the brightest objects at the highest angular resolutions. The robust reconstruction of complex brightness distributions from sparsely sampled Fourier data augers well for the future of the next generation of separate-element ground-based imaging arrays with baselines in excess of an order of magnitude larger than those available here." }, "0003/astro-ph0003416_arXiv.txt": { "abstract": "\\noindent{\\it I discuss the chemical and spectrophotometric evolution of galaxies over cosmological timescales and present a first attempt to treat both aspects in a chemically consistent way. In our evolutionary synthesis approach, we account for the increasing metallicity of successive generations of stars and use sets of stellar evolutionary tracks, stellar yields, spectra, etc. for various metallicities. This gives a more realistic description of nearby galaxies, which are observed to have broad stellar metalllicty distributions, as well as of young galaxies at high redshift. Selected results are presented for the chemo-cosmological evolution of galaxies as compared to QSO absorption line observations and for the spectrophotometric evolution of galaxies to very high redshifts. On cosmological timescales, intercations are important drivers of galaxy evolution. Neglecting dynamical aspects we study the effects of interaction-induced starbursts on the spectrophotometric and chemical evolution of galaxies and briefly discuss the formation of star clusters and Tidal Dwarf Galaxies in this context. } ", "introduction": "The evolution of galaxies on cosmological timescales has 3 aspects: the chemical, spectral, and dynamical evolution. In reality, all of them are intimately coupled as shown by the existence of several observational relations involving quantities from any two of these 3 domains. Examples are the Faber -- Jackson, Fundamental Plane, or luminosity -- metallicity relations for ellipticals or the Tully -- Fisher relation and trends of characteristic HII region abundances, average colours, emission line strengths, luminosities, and mass - to - light ratios with spiral type, i.e. with the transition from bulge-dominated to disk-dominated systems. While over short timescales or lookback times and, at least, for giant galaxies, the 3 aspects of galaxy evolution might be treated independently, any modelling of galaxy evolution over cosmological timescales definitely requires a consistent treatment of all 3 aspects. Unfortunately, this is still too complex today, even for numerical simulations on powerful machines. Independently, all 3 aspects of galaxy evolution modelling have quite some tradition already. We are, however, still at the very beginning of a consistent understanding of galaxy formation and evolution over cosmological timescales including all 3 aspects. A consistent modelling of the chemical and dynamical evolution of galaxies is e.g. attempted by the group of G. Hensler in Kiel (`Chemodynamics'). With the availability of stellar input physics for various metallicities, we developed in G\\\"ottingen a consistent description of the chemical and spectral evolution over cosmological timescales, i.e. coupled to a cosmological model. This {\\bf chemically consistent evolutionary synthesis} method and selected applications for the interpretation of galaxy observations over a wide redshift range will be presented here. While still being far from a consistent coupling of our chemically consistent chemical, spectrophotometric and cosmological model with a dynamical galaxy evolution model, we at least started to study the starbursts accompanying strong interactions and mergers involving gas rich galaxies and their impact on the chemical and spectral evolution of galaxies. This brings us to some surprising features accompanying those strong bursts: the formation and evolution both of a new population of bright and massive young star clusters and of a new class of dwarf galaxies forming from recycled material along the tidal features of the merging giant galaxy pair (Tidal Dwarf Galaxies). ", "conclusions": "I presented a versatile model for the evolution of stellar populations and gas that offers a variety of applications from star clusters to nearby and distant galaxies. Only a few of them have been presented here. With a minimum number of parameters our combined chemical, spectrophotometric and cosmological evolution models describe a large number of observables and provide a long evolutionary baseline to compare with high-redshift galaxy abundances and spectra and understand the evolution of various galaxy types from the earliest phases to the present. The chemically consistent treatment is a first attempt to consistently combine 2 out of 3 aspects of galaxy evolution that nature, too, has coupled. Interactions play an important role for the evolution of galaxies over cosmological timescales. While in its present state, the model does not include any dynamical aspect nor spatial resolution, we tried and studied the effects of starbursts accompanying galaxy interactions if gas is involved. Several interesting phenomena were observed in this context, as e.g. the formation of large populations of bright star clusters and of ``recycled'' Tidal Dwarf Galaxies. Application of our models provided a first step to understand the nature and properties of these systems as well as their possible future evolution. The model has allowed for a series of precise observational predictions, part of which became verified already while others keep standing for a test. Our first attempt to also include the ${\\rm 3^{rd}}$ aspect of galaxy evolution, the formation and dynamical evolution of galaxies in their cosmological environments, could not be discussed here (cf. Contardo \\etal 1998). Over the years this model was developed, extended and refined, with its input physics continuously updated, observational extragalactic research has seen a tremendous progress. The amount of information from HST and large ground-based surveys is enormous and several quantum steps have been performed, e.g. concerning image resolution with HST, spectral resolution with KECK and WHT, and the number of high-redshift galaxies by the Lyman break technique. With 10 m telescopes, like VLT and HET, observational progress is challenging theory to keep path. The particularly close interplay between observations and the conceptually simple galaxy evolution modelling presented here, has proven very fruitful and stimulating for both sides. \\medskip\\noindent {\\bf Acknowledgements} {\\footnotesize \\noindent My thanks go to K. Fricke for his encouragement and to all the present and former members of our Galaxy Evolution Group, who -- over the years -- have contributed to various aspects of the work presented here, i.e. to Harald Kr\\\"uger, Christian Einsel, Johannes Loxen, Claudia M\\\"oller, Oliver Kurth, Ulrich Lindner, Peter Weilbacher, Jens Bicker, and Jochen Schulz. Partial financial support from the {\\sl Deutsche For\\-schungs\\-gemeinschaft} and the {\\sl Verbundforschung Astronomie} for various aspects of this work is gratefully acknowledged. } \\vspace{0.7cm} \\noindent {\\footnotesize" }, "0003/astro-ph0003285_arXiv.txt": { "abstract": "We have studied the UV extinction properties along 30 Galactic sightlines using data from the {\\it International Ultraviolet Explorer} (IUE) archive that have never been previously examined. These distant (d $>$ 1 kpc) sightlines were selected to investigate the distribution and physical conditions of gas located in low density regions of the Galactic disk and halo. The average densities along these sightlines are extremely low. It is likely that they are dominated by the warm intercloud medium and have little contribution from the cold cloud medium. We find that a subsample of these sightlines has extinction curves with weak bumps and very steep far-UV extinction reminiscent of the Magellanic clouds. These sightlines all lie in the region bounded by 325$^o \\leq l \\leq 0^o$ and -5$^o \\geq b \\geq -11^o$. The gas along these sightlines shows forbidden velocities which may indicate that the dust has been subject to shocks. This type of low density sightline may mimic the environments found in the Magellanic Clouds. Large values of N(Ca II)/N(Na I) indicating low depletion are associated with steep far-UV extinction. A possible correlation exists between decreasing bump strength and increasing far-UV steepness for extinction curves in the Galaxy and the Magellanic Clouds. ", "introduction": "There is an average Milky Way extinction relation, A($\\lambda$)/A(V), over the wavelength range 0.125 $\\mu$m to 3.5 $\\mu$m, which is applicable to a wide range of interstellar dust environments, including lines of sight through diffuse dust, and dark cloud dust, as well as dust associated with star formation (Cardelli, Clayton, \\& Mathis 1989 (CCM); Cardelli \\& Clayton 1991; Mathis \\& Cardelli 1992; Fitzpatrick 1999). The existence of this relation, valid over a large wavelength interval, suggests that the environmental processes which modify the grains are efficient and affect all grains. The CCM relation depends on only one parameter, the ratio of total-to-selective extinction, R$_V$ which is a crude measure of the size distribution of interstellar dust grains. However, the CCM relation does not appear to apply beyond the Milky Way. It does not always fit the observed extinction along sightlines observed in the Magellanic Clouds and M31 (e.g., Clayton \\& Martin 1985; Fitzpatrick 1985, 1986; Clayton et al. 1996; Bianchi et al. 1996; Gordon \\& Clayton 1998; Misselt, Clayton, \\& Gordon 1999). The 2175 \\AA~bump is weaker and the far-UV extinction is steeper in many of the Magellanic cloud sightlines but there are also sightlines in both the LMC and SMC where the dust extinction does follow CCM. The few lines of sight studied in M31 seem to show a CCM-like far-UV extinction and a weak 2175 \\AA~bump (Bianchi et al. 1996). On the other hand, the starburst nucleus of M33 appears to be associated with Milky-Way-type dust (Gordon et al. 1999). The variations in extinction properties seen in the Magellanic Clouds and M31 may be due to several factors. Different environments, such as star formation regions where large amounts of UV radiation and shocks are present, may play a large role in processing dust. Evidence for this can be seen in the LMC where two distinct wavelength dependences of UV extinction have been found for dust inside and outside the supergiant shell, LMC 2, which lies on the southeast side of 30 Dor. This structure was formed by the combined stellar winds and supernovae explosions from the stellar association at its center (Misselt et al. 1999). In the SMC, the dust properties are even more extreme, showing extinction curves for three of four sightlines which have virtually no bump and are very steep in the far-UV (Gordon \\& Clayton 1998). Although the dust responsible for these curves is located near regions of star formation in the SMC, the environment is likely to be less severe than for the LMC 2 dust. The 30 Dor region, where LMC 2 is located, is a much larger star forming region than any in the SMC. The dust environments in starburst galaxies and QSOs, which also show SMC-like extinction, are much more extreme than 30 Dor (e.g., Gordon, Calzetti, \\& Witt 1997; Pitman, Clayton \\& Gordon 2000; Gordon, Smith \\& Clayton 2000). The SMC has star formation occurring at only 1\\% the rate of a starburst galaxy so other factors such as the known differences in metallicity between galaxies may be important (Fitzpatrick 1986; Gordon \\& Clayton 1998; Misselt et al. 1999). Setting aside global metallicity differences, are there sightlines in the Galaxy where the dust environment is similar to those seen in the Magellanic clouds? Real deviations from CCM are seen in the Galaxy but deviations of the kind seen in the Magellanic clouds have been seen only rarely (Cardelli \\& Clayton 1991; Mathis \\& Cardelli 1992). A few sightlines (e.g., 62542, 204827, and 210121) show weak bumps and anomalously strong far-UV extinction for their measured values of $R_V$. Their extinction curves are plotted in Figure 1. These deviant sightlines represent a variety of dust environments. The Galactic sightline toward HD 62542 is somewhat similar to LMC 2. Its dust was swept up by bubbles blown by two nearby O stars (Cardelli \\& Savage 1988). HD 204827 is also in a star formation region where the dust has been subject to shocks (Clayton \\& Fitzpatrick 1987). HD 210121 lies behind a single cloud in the halo. There is no present activity near this cloud although it was ejected into the halo at some time in the past. There are some important differences between these Galactic extinction curves and those in the Magellanic clouds. The bump seen for HD 62542 is not just weak but it is very broad and shifted to the blue (Cardelli \\& Savage 1988). Mantles on the bump grains has been suggested as the reason for the weak, broad, and shifted Galactic bumps (Mathis \\& Cardelli 1992; Mathis 1994). These sightlines show that dust in a variety of environments with a range of $R_V$ values can have extinction curves similar to those in the LMC. However, none of the anomalous Galactic sightlines, seen in Figure 1, approach the SMC extinction properties. The SMC dust has weaker bumps and steeper far-UV extinction than any known Galactic or LMC sightline. \\vspace*{0.05in} \\begin{center} \\plotone{f1.eps} \\end{center} \\figurenum{1} \\figcaption{Sightlines in the Galaxy showing anomalous extinction. CCM curves for $R_V$ values of 2.5 and 3.1 are plotted for comparison.} \\vspace*{0.2in} Most of the Galactic sightlines, that have been studied previously, differ in one respect from the LMC and SMC sightlines. They are significantly more reddened than the Magellanic cloud sightlines. In particular, those sightlines showing the greatest deviations from CCM, those near the supershell LMC 2 and those in the SMC, all have E(B-V) $<$ 0.25. Of the twenty-nine CCM sightlines only two have E(B-V) $<$ 0.30. The others range up to E(B-V) = 1.2. Similarly, the Fitzpatrick \\& Massa sample of eighty stars includes only seven with E(B-V) $<$ 0.30 (Fitzpatrick \\& Massa 1990; Fitzpatrick 1999). Therefore, the dust along the Magellanic cloud sightlines is more diffuse and more representative of the warm intercloud medium than the cold cloud medium which is better represented in the Galactic samples. Kiszkurno-Koziej \\& Lequeux (1987) suggest from ANS extinction measurements of 1200 stars in the Galaxy that there may be a correlation between UV extinction parameters and distance from the Galactic plane. As $\\vert z \\vert$ increases, the bump becomes weaker and the far-UV extinction stronger. These sightlines have low reddenings and long sightlines so they are also more diffuse and therefore more like those in the Magellanic clouds. To investigate whether the extinction properties observed in the Magellanic clouds are related to the diffuse nature of the sightlines, a sample of long sightlines with low reddenings in the Galaxy was chosen and UV data were obtained with the {\\it International Ultraviolet Explorer} (IUE) so that extinction curves could be constructed. ", "conclusions": "\\noindent $\\bullet$ Magellanic-cloud-like extinction has now been found in the Milky Way.\\\\ $\\bullet$ Large values of N(Ca II)/N(Na I) indicating low depletion are associated with steep far-UV extinction as measured by $c_2$.\\\\ $\\bullet$ Global metallicity seems not to be a direct factor. Local environmental conditions seem to be the most important factor in determining dust properties.\\\\ $\\bullet$ Similar UV dust extinction properties have now been seen in the Milky Way, the Magellanic clouds, starburst galaxies and in high redshift star-forming galaxies.\\\\ $\\bullet$ There may be at least two ways to achieve similar extinction properties. A lack of dust coagulation has been suggested for HD 210121 to explain the observed extinction (Larson et al. 1996). The Galactic SD-region properties are closely tied to forbidden velocities indicating that processing of the grains in the diffuse ISM resulted in their observed properties. \\\\ $\\bullet$ There seems to be a correlation between decreasing bump strength and far-UV steepness that includes the Galaxy and the Magellanic Clouds. \\\\ $\\bullet$ All the sightlines contained in CCM lie within 1 kpc of the Sun. As this study shows, dust properties are not well mapped even in our own Galaxy. There are a larger range of UV extinction parameters seen in the Milky Way than implied by CCM.\\\\ Thanks to Ed Fitzpatrick for providing his IUE IDL procedure. This project was originally envisioned by the late Jason Cardelli." } }