{ "0005/astro-ph0005277_arXiv.txt": { "abstract": "The Institute for Science Information (ISI) has generated two lists of citation information for astronomers that are restricted both as to the years surveyed for the {\\it cited} papers, and the years surveyed for the {\\it citing} papers. These databases are unique among the electronically-available citation data in their restrictions of both citing and cited years. The main list (P\\&A-100) gives citation data for 62,813 physicists and astronomers whose journal papers were cited 100 times or more from 1981.0 to 1997.5 by papers published during the same time interval. The second list (AST-top-papers) gives the 200 most-cited papers/year published in refereed astronomical journals from 1981--1996, as cited in papers in those same journals from 1981.0 to 1998.0. Astronomer names were selected from those given in the P\\&A-100 list using various sources, including the 1998 AAS Membership Directory, the 2000 list of the Astronomical Society of India, the names from the AST-top-papers list, the list of astronomers honored by the AAS, National Academy of Sciences and the Nobel Prize Committee, and the knowledge of this writer. From this work an Astronomy Citation Database has been constructed, containing citation data for 6458+ astronomers. Various problems, both substantial and subtle, of producing a reasonably fair citation database from either the data supplied by the ISI to this author, or from the Web, are detailed. Chief among these are whether to assign either parital or full credit for each author on a given paper. Whether one is honored with one of the top lifetime-awards given to astronomers is a strong function of how well your work stands out as your own. In particular, we can negatively impact our citation statistics in two ways. First, because the ISI does not keep track of meeting proceedings/books/catalogs, per se, we do not get citation credit for meeting papers/books/catalogs in the ISI lists. Second, if we fuzz our identities on the papers, such as publishing papers with two or more first initials or through confusion with the names of other astronomers/physicists. Name confusion affects this kind of analysis to the extent that that it would take an enormous effort to disentangle its effects and, even, then, not all name confusion would be settled. If sociologists, science historians and others (ourselves included?) feel that solving the name confusion problem is worthwhile, perhaps we, as a professional society, should take appropriate steps. A ``modest proposal'' is made that our professional field (and others) go to a system of uniquely associating an identification number to each author on each paper. ", "introduction": "For the past several decades, we in astronomy have relied on the Science Citation Index, as compiled and published by the Institute for Science Information (ISI), as our source for citation statistics of our papers. At the dawn of the 21st Century, we in astronomy are becoming more and more reliant on electronic databases for the papers we read, and for the citation statistics on those papers. The two main internet sites we access for listings of our published papers (as opposed to preprints) are the Astrophysics Data System (ADS) (adsabs.harvard.edu) and the Institute for Science Information's (ISI) Science Citation Index (www.webofscience.com). Indeed, the Science Library at our University no longer subscribes to the ISI's printed Science Citation Index; rather there is now complete reliance on the ISI's web-accessible Web of Science. Yet, as this author has discovered during the research conducted for this paper, many of us do not clearly understand the contraints and limitations of either the hard-copy Science Citation Index and the new Web of Science. However, what prompted the research done in this paper was the a new kind of citation analysis produced by the ISI research group over the past few years. At least two new citation lists were generated by the ISI research group before this author contacted them: A list of the chemists cited 500 times or more and a similar list for physicists/astronomers. Each list differs from what we can access either from the Web of Science or from the hard-copy Science Citation Index, in that time intevals for both {\\it citing} and {\\it cited} papers are specified. A French chemist, Dr. Armel Le Bail (Laboratoire des Flouresces, CNRS, ESA) purchased the ISI's most-cited chemist list (for \\$1000), and posted it on the Web (pcb4122.univ-lemans.fr/cgi-bin/physiciens.pl). Along with the most-cited chemist list, the ISI sent Dr. Le Bail the first 1120 names of physicists/astronomers on {\\it that} most-cited list, which Dr. Le Bail scanned into his computer and posted on the web. One of the reasons Dr. Le Bail posted these lists on the web was given by Dr. David Pendlebury of the ISI, who points out that, of the 50 most-cited chemists, 7 have been awarded the Nobel Prize (cf. Garfield \\& Welljam-Dorof 1992). The interest of the present author in this subject was piqued when one of his colleagues pointed out that his name was in the most-cited physicists/astronomers list on the web page of Dr. Le Bail. Thus began the journey of this author down the rabbit hole of web-accessible and web-generated paper and citation information. The present paper, with its lessons learned and data gathered, is the net result of that journey. Aside from the not-inconsiderable curiosity factor (e.g., most of us would like to know where we stand relative to others in the number of times our papers have been cited) and job-related factors (my own promotion to professor was aided by such a list compiled by a colleague for one year of citations for our department), why would one do such a study? This author can think of several questions one would like to answer. First, at the very least, if citation information is to be used in connection with job-related decisions, should not the available data be of highest possible quality? Such data should be treated like any other data, and investigated as to random and systematic errors. Should it not also be clear what assumptions go into the data being used? Given that a relationship exists between being most-cited and winning a Nobel prize among chemists, does the same relationship exist for honors received by astronomers? As a guide to those scientists just entering our field of study, what do these data tell us about how the way we put our names on our papers, and where we publish our papers, influence how we are honored by our peers? Previous papers which tried to assess citation information for astronomers (Abt 1981a,b, 1982, 1983, 1984a,b, 1985, 1987a,b, 1988a,b 1989, 1990a,b, 1992a,b, 1996, 1998a,b; Abt \\& Zhou 1996; Trimble 1985, 1986a,b, 1988, 1991, 1993a,b, 1996; White 1992; Girard \\& Davoust 1997; Davoust \\& Schmadel 1987, 1992) were limited by time and data access to asking statistical questions that are more restricted than those that now can be addressed electronically. The methodology employed by this paper are detailed in Section 2, where we address, in detail, what one can do, and what cannot do, with the present databases made available to this author by the ISI as well as those databases generally available on the web. The data we have generated from this study are discussed in Section 3. The statistical studies of citations for astronomers are discussed in Section 4, both among themselves and in comparison to the life-time honors bestowed to individuals. The main results of this paper are summarized in Section 5, where a ``modest proposal'' is made towards solving the ever-pervasive name confusion problem. ", "conclusions": "An Astronomy Citation Database (ACD) has been assembled, which gives the number of citations, papers cited and the cite/pap ratio for 6331+ astronomers for a 16.5 year shapshot of time. These data correspond to citation information assembled by the Institute for Science Information (ISI) for astronomers and physicists for papers published in the years 1981-1997.0, as cited in papers published in 1981-1997.5. The data for the astronomy list was assembled from two databases given to this writer by the ISI. One list (P\\&A-100) contains citation data for 62,813 physics and astronomy ISI ``names'' cited 100 or more times during the specific intervals. The other list (AST-top-papers) contains both citation data for the 200 most-cited papers published each year in astronomy during 1981-1996, as well as for the astronomy-related citations for 5,035 astronomer names from those papers. The databases given to this author by the ISI give full citation credit to each author for each of her/his papers, whether first author or not, whether multi-authored or not. This is also the methodology used for ISI's Web of Science. While the methodology of the ISI in this regard has not changed over the years, how it presents its data has. In particular, users familiar with the hard-copy Science Citation Index will note that only papers on which you are a first author are listed in the Author section, while all papers on which you were an author (first or not) are given the Source section. As such, readers should be aware that the ISI Web of Science now gives full credit for the citations for all papers on which each of us is an author, but only for refereed journals listed from their list of sampled journals (cf. Table~1). In order to make sense of these citation data, this writer had to engage in a series of laborious, time-consuming tasks. These tasks also involved discovering and correcting for a number of biases in the original data lists, some of which are inherent to any electronically-assembled database. The main file for ACD is divided into three subsets: 4617 ``unique-one'' astronomers (those whose names are uniquely identified with individuals); 357 ``unique-two'' astronomers (those whose last names are cited with two or more sets of first initials); and 1484 ``confused-named'' astronomers (those with names and initials that are confused with those of other astronomers and/or physicists). The use of the word ``unique'' to name two of the subsets refers to the fact that these are names singly assigned to an individual astronomer. Such is not the case for the ``confused'' ISI names, which are associated with two or more individuals in astronomy and/or physics. Due to name confusion, it is likely than many astronomers are listed twice in the ACD, either two times in the confused-named list, or once in the unique-one list and once in the confused-named list. Two other files are provided with the ACD. One is an ``honorable mention'' list of 173 names for those astronomers whose names are in the AST-top-papers list but not in the P\\&A-100 list, and which have less than 100 citations from the AST-top-papers list. The other is the ``honors'' list, which correlates the citation data information from the main database with the honors 278 individuals have received for their astronomical work. A list of the 10 most-cited papers per year, from 1981 to 1996 is provided in this paper (but not in the electronic database). Comparison of the authors' names on the 10 most-cited papers to those names most cited overall shows a good correspondence. Moreover, the papers in the 10 most-cited-per-year list are of a wide range of paper type (e.g., review, data, theory, observation) and are all known to be of high quality. Hence, the old shibboleth that one can get many citations from publising papers with wrong results is shown to be the myth that it is. The errors for citations for individuals in the ACD are both random and systematic. Random errors exist in proportion (estimated to be 4\\%) to the number of citations, owing to the way in which the ISI compiles citations. Systematic errors include errors of omission and comission. Errors in citations owing to variance in self-citations among astronomers can also affect the statistics at the $\\sim 6\\%$ level, if one chooses to apply such a criterion to these data. The likely most-egregious error in this database in the eyes of some of the readers of this paper is the use of full citations for each author of a paper. Tests using the papers of top-cited authors in the AST-top-papers database shows that no way we can think of to calculate citations for authors will give the same results. The plain fact is, however, whether one agrees with using full citations or not for astronomers, this is what the ISI gives us in the Web of Science, as well as in the P\\&A-100 and AST-top-papers lists. We show that, due to the manner in which citations are assembled by the Institute for Science Information, which maintains the science citations for our field of study, we hurt the impact of our papers if we publish many papers in meeting proceedings, or if our name is confused with those of others. Two more of our findings are of specific sociological interest. The first is that astronomers who divide their papers among two or more first sets of initials on average publish more papers, have more citations, and have a significantly higher citations/paper ratio than astronomers who publish under just one set of initials. The second is that there is a very good, but not perfect, correspondence of a person being near the top of the citation list and the chance that person has been honored by her/his peers in U.S. astronomy for life-time achievement. Closer examination of this good correlation reveals that assigning honors to our peers is as human an enterprise as any that we do. The evidence asuggests that anything we do, intended or not, to blur the focus of our papers to our peers has a measurable effect on how we are honored by them. Such blurring effects can come in several forms: authoring papers using two or more sets of initials; having a name confused with those of others; publishing your papers predominantly in non--U.S.--based journals or meeting proceedings. Other means of blurring the focus of our papers (having high citations, but low cite/pap ratio; mostly publishing in large groups or with more well-known authors; publishing in several different scientific fields) also likely exist. In a paper devoted to analyzing the citation data for astronomers, it is relevent to note that citation statistics give us but one view of the impact of individuals on our science. This is best evidenced by the most recent Nobel Prize awarded to astronomers, that to JH Taylor and RA Hulse in 1993. Where these two individuals stand in the citation list has little relationship to their scientific impact on our field. The two ISI databases used in this paper were generated with ISI software and given to this author expressly for the analysis done in this paper. These databases are different from those we can access via the Web in three ways: ability to specify specific ranges of years for cited and citing papers; kinds of papers cited, and number of citations attributed to each author. The idiosyncracies discovered in the course of this survey of the use of current web-based database for the bibliography of astronomers are detailed. The full ACD (all three files) will be made available via anonymous ftp from samuri.la.asu.edu, as well as through the Astrophysics Data Center. The top ten papers cited for each year between 1981-1996 are given in this paper. The full top-200 cited papers database produced by the ISI is proprietary (termed ``High-Impact Papers in Astronomy'') and is available for a fee separately from the ISI. Since the ACD is a database about the accomplishments of people, any error in the ACD is a serious error. As such, readers are encouraged to contact this writer if errors of omission or comission are found in the ACD. Those appropriate modifications to the ACD that should be made, will be made. It is hoped that with the help of readers, a full {\\it unconfused}, ACD can be eventually made of cited papers during 1981--1997.5. Such a list can then act as a baseline against which future investigations of this kind may be made. It also hoped that the lessons learned in this paper about the idiosyncracies of the various electronic databases will aid others in their own searches. Towards this end, I end this paper by unashamedly borrowing from Jonathan Swift in putting forward a ``modest proposal'' for eliminating name confusion in our field. We are used to having social security numbers, university ID numbers, shopper ID numbers; each of us is now various numbers in various databases. If this is so, why do we not assign a ``publishing ID number'' (PID for short) for each person who publishes a paper in our journals? I suggest that this be done, and we start by assigning PIDs to all astronomers who {\\it have} published papers in the journals in the past. The ACD could be a start, but only a start, as to solve name confusion the process must work iteratively among the ISI, the individuals involved and the ACD. The PID number would then be carried by each journal in the author list (but not necessarily printed out for each paper). If we, and other scientists in other fields of study, are interested in having a true, honest assessment of citation data for astronomers as a function of time, then the problem of name confusion should be, and can be, solved. \\vspace*{3mm} This work could not have been done without the active participation and cooperation of Dr. David Pendlebury of the ISI, who supplied the two ISI datasets used in this study, and who also did a very careful reading of the first draft of this paper. This author owes Dr. Pendlebury much thanks. Conversations with Helmut Abt, Sandra Faber and Anne Cowley were also helpful in writing this paper. My thanks to Arnab Choudhuri for pointing me in the direction of the online Astronomy Society of India directory. Insightful comments from the editor, Bob Milkey, also greatly helped the presentation of this paper." }, "0005/astro-ph0005088_arXiv.txt": { "abstract": "We have obtained low resolution ($\\sim$20~\\AA\\ FWHM) slit spectra of 10 quasar candidates located within one degree of the $z=2.181$ quasar Q0122$-$380 with the objective of searching for signs of large scale structure matching the intervening rich absorption complexes seen over the redshift range $1.81 \\la z \\la 1.97$ toward this object. Of the 8 confirmed quasars, 4 turn out to have redshifts $z<1.8$, placing them well in front of the redshift range of interest. Two of the three confirmed quasars at redshift $z>1.8$ show no obvious absorption matching that of Q0122$-$380 at our spectral resolution and signal-to-noise ratio. The third object at a redshift of $z=1.868$ displays strong $z_{\\rm abs} \\sim z_{\\rm em}$ absorption systems at $z \\simeq 1.84$ and $z\\simeq 1.86$ and a possibly BAL-like trough at $z\\simeq 1.76$. If not intrinsic in nature, the former two systems could potentially be related to the absorption seen in Q0122$-$380, albeit over a distance of $50\\farcm2$ ($D_\\perp \\simeq 15\\,h^{-1}$ Mpc at $z\\simeq 1.9$). ", "introduction": "The existence of large scale structure at high redshift ($z \\ga 1$) provides an important constraint on theories for the formation of structure and evolution of the Universe. One approach to probing for such structure is through the study of intervening metal line absorption systems in quasar spectra. Such systems are believed to trace galaxies through their extended gaseous halos. Statistical analysis of the redshift distribution of available samples of quasar absorption systems suggest that large scale clustering on comoving scales up to $\\sim$100 Mpc may have been in place already at $z$$\\sim$2--3 (e.g.\\ Quashnock et al.\\ \\cite{QVY96}). The complementary technique of probing for large scale structure in the plane of the sky by searching for correlated absorption in adjacent lines of sight is hampered by the relatively low density of high redshift quasars bright enough for detailed absorption line work. Nonetheless, several potential high redshift 'absorption superclusters' spanning tens of Mpc on the sky have been identified in this manner. These include the $z\\simeq1.65$ absorption systems seen toward PKS~0273$-$233 (Foltz et al.\\ \\cite{Fetal93}); the two pairs of damped Ly$\\alpha $\\ systems seen at $z\\simeq 2.38$ and $z\\simeq 2.85$ toward Q2138$-$4427 and Q2139$-$4434 (Francis \\& Hewitt \\cite{FH93}; Francis et al.\\ \\cite{Fetal96}); the apparent structures at $z\\simeq 2.3$ and $z\\simeq2.5$ detected in a dense quasar field near the south Galactic pole by Williger et al.\\ (\\cite{Wetal96}); and the well-studied case of the strong absorption spanning $1.8 \\la z \\la 2.2$ in the field of the quasar pair Tol~1037$-$2703/1038$-$2712 (Jakobsen et al.\\ \\cite{Jetal86}; Dinshaw \\& Impey \\cite{DI96}; Lespine \\& Petitjean \\cite{LP97}; and references therein). With the aim of searching for further such cases of intervening high-redshift superclusters Romani et al.\\ (\\cite{RFS91}) searched the quasar catalogs for suitable background objects near quasars known to display rich metal line absorption systems. One of the most promising fields identified by Romani et al.\\ is that of the $z=2.181$ quasar Q0122$-$380, an object whose absorption spectrum contains at least seven C\\,{\\sc iv}\\ systems between $1.81 \\la z \\la 1.98$ (Carswell et al.\\ \\cite{Cetal82}) and happens to lie within a field in which Savage et al.\\ (\\cite{Setal84}) have carried out a deep objective prism quasar search. Q0122$-$380 is therefore surrounded by 11 quasar candidates of brightness $V\\simeq$19--20 within a $1^\\circ$ radius, corresponding to comoving distances $D_\\perp \\la 44\\,h^{-1}$ Mpc at $z\\simeq1.95$. In this paper we present exploratory slit spectra and redshifts of these quasar candidates. As it turns out, several of the objects of interest are either not confirmed as quasars, or lie at significantly lower redshift than indicated by their preliminary catalog entries, thereby rendering the field toward Q0122$-$380 rather less promising for the purpose of searching for high redshift superclusters than originally thought. \\begin{table*}[ht!] \\setlength{\\tabcolsep}{3pt} \\begin{center} \\parbox[b]{0.725\\linewidth}{ \\caption[]{Objects observed in the field of Q0122$-$380} \\label{Q-Tab.coordsz} } \\begin{tabular*}{0.725\\linewidth}[h]{ p{0.15\\linewidth} lll lll r r r r } \\hline\\noalign{\\smallskip} \\multicolumn{1}{c}{Object$\\ \\qquad\\ $} & \\multicolumn{3}{c}{$\\alpha$(J2000)} & \\multicolumn{3}{c}{$\\delta$(J2000)} & \\multicolumn{1}{c}{$\\quad m_{\\rm V}$} & \\multicolumn{1}{c}{$\\quad z^{\\ast}{}^{\\dag}$} & \\multicolumn{1}{c}{$z_{\\rm em}$} & \\multicolumn{1}{c}{$\\quad\\theta^\\ddag$} \\\\ \\noalign{\\smallskip} \\hline\\noalign{\\smallskip} \\ Q0122$-$380\\ & 01$^h\\!\\!$&24$^m\\!\\!$&17\\fs52 & $-$37\\degr$\\!\\!$&44\\arcmin$\\!\\!$&27\\farcs0\\ \\ & $\\quad$16.5 & $\\quad$2.181 & 2.189$\\pm$0.006 & $\\quad$ \\phantom{0.00} \\\\ \\noalign{\\medskip} \\ Q0117$-$379\\ & 01 & 19 & 56.47 & $-$37 & 38 & 39.\\,3\\ \\ & $\\quad$20.0 & $\\quad$1.48~ & 1.484$\\pm$0.007 & $\\quad$52.0 \\\\ \\ Q0117$-$380\\ & 01 & 19 & 45.65 & $-$37 & 48 & 27.\\,1\\ \\ & $\\quad$18.9 & $\\quad$2.02~ & 2.020$\\pm$0.007 & $\\quad$53.9 \\\\ \\ Q0118$-$377\\ & 01 & 20 & 43.15 & $-$37 & 28 & 38.\\,5\\ \\ & $\\quad$19.0 & $\\quad$0.34~ & 1.728$\\pm$0.007 & $\\quad$45.3 \\\\ \\ Q0120$-$3781\\ & 01 & 22 & 35.50 & $-$37 & 32 & 56.\\,9\\ \\ & $\\quad$19.5 & $\\quad$2.15~ & 2.124$\\pm$0.007 & $\\quad$23.2 \\\\ \\ Q0120$-$3785\\ & 01 & 22 & 45.05 & $-$37 & 35 & 31.\\,6\\ \\ & $\\quad$19.5 & $\\quad$2.17~ & 1.516$\\pm$0.004 & $\\quad$20.3 \\\\ \\ Q0121$-$373\\ & 01 & 24 & 15.06 & $-$37 & 05 & 13.\\,9\\ \\ & $\\quad$19.4 & $\\quad$1.49~ & 1.055$\\pm$0.009 & $\\quad$39.2 \\\\ \\ Q0124$-$373\\ & 01 & 26 & 15.52 & $-$37 & 07 & 46.\\,1\\ \\ & $\\quad$19.8 & $\\quad$0.88~ & 0.920$\\pm$0.010 & $\\quad$43.5 \\\\ \\ Q0125$-$376\\ & 01 & 28 & 05.78 & $-$37 & 22 & 37.\\,6\\ \\ & $\\quad$19.0 & $\\quad$1.84~ & 1.868$\\pm$0.008 & $\\quad$50.2 \\\\ \\noalign{\\medskip} \\ \\ ~0121$-$379$^a$ & 01 & 23 & 25.89 & $-$37 & 42 & 23.\\,6\\ \\ & $\\quad$19.7 & $\\quad$2.21~ & \\multicolumn{1}{c}{$\\cdots$} & $\\quad$10.4 \\\\ \\ \\ ~0123$-$372$^a$ & 01 & 25 & 27.10 & $-$36 & 58 & 06.\\,1\\ \\ & $\\quad$20.3 & $\\quad$2.13~ & \\multicolumn{1}{c}{$\\cdots$} & $\\quad$48.4 \\\\ \\ \\ ~0117$-$378$^b$ & 01 & 19 & 56.82 & $-$37 & 37 & 28.\\,4\\ \\ & $\\quad$20.0 & $\\quad$2.25~ & \\multicolumn{1}{c}{$\\cdots$} & \\multicolumn{1}{c}{$\\quad\\cdots$} \\\\ \\noalign{\\smallskip}\\hline\\noalign{\\smallskip} \\end{tabular*} \\parbox[b]{0.725\\linewidth}{ \\noindent $^\\dag$ Preliminary redshift listed in Hewitt \\& Burbidge (1993)\\\\ $^\\ddag$ Angular distance from Q0122$-$380 in arcmin.\\\\ $^a$ Object is a star, not a quasar.\\\\ $^b$ No object was found at the catalog coordinates.\\\\ } \\end{center} \\end{table*} ", "conclusions": "Romani et al.\\ (1991) originally drew attention to the rich absorption line quasar Q0122$-$380 on the basis that it was surrounded by a number of quasar candidates closer than $1^\\circ$. Of the 11 objects listed in Table~1, seven had preliminary redshifts $z>1.8$, placing them near or behind the absorption seen between $1.81 \\la z \\la 1.98$ toward Q0122$-$380. Of these seven high redshift candidates, only three (Q0117$-$380, Q0120$-$3781 and 0125$-$376) are confirmed as $z>1.8$ quasars. Of the remainder, one (0117$-$378) could not be located, two (0121$-$379 and 0123$-$372) are identified as stars, and another (Q0120$-$3785) turns out to be at a lower redshift of $z=1.52$. The four quasar candidates with lower preliminary redshifts $z<1.8$ are all confirmed as such. \\begin{figure}[ht!] \\centerline{ \\hfill\\includegraphics[width=0.375\\textwidth]{qsofig2.eps}\\hfill }\\par \\parbox[t]{0.485\\textwidth}{ \\caption[]{Spectra of the two stars misclassified as quasars in Hewitt \\& Burbidge (1993). The stellar absorption features are indicated. } } \\end{figure} As is evident from Fig.~1, our low resolution spectra of two of the three confirmed $z>1.8$ quasars (Q0120$-$3781 and Q0117$-$380) reveal no obvious C\\,{\\sc iv}\\ absorption features in the wavelength range 4350-4600~\\AA\\ that could potentially be associated with the absorption spanning $1.81 \\la z \\la 1.97$ toward Q0122$-$380. However, the line detection limit of our spectra is only $W_\\lambda \\ga 8$~\\AA\\ at these wavelengths. A more promising case is that of the final object, Q0125$-$376, whose redshift of $z=1.868$ lies close to that of the absorption complex at $z \\simeq 1.91$ in Q0122$-$380. Moreover, our spectrum of Q0125$-$376 (which has a better S/N ratio than those of Q0120$-$3781 and Q0117$-$380) shows three strong ($W_\\lambda \\ga 9$~\\AA) absorption features at $\\lambda\\simeq 4395$~\\AA, $\\lambda\\simeq 4435$~\\AA\\ and $\\lambda\\simeq 4275$~\\AA, respectively. The former two features are almost certainly due to C\\,{\\sc iv}\\ absorption from two $z_{\\rm abs} \\sim z_{\\rm em}$ systems at $z\\simeq 1.837$ and $z\\simeq 1.864$, an interpretation that is further strengthened by plausible detections of matching Si\\,{\\sc iv}\\ lines at shorter wavelengths. Based on a matching weaker feature seen at the anticipated position of Al\\,{\\sc iii}, we tentatively identify the third feature as the C\\,{\\sc iv}\\ trough of a weak and possibly detached BAL-like complex at $z\\simeq 1.76$\\/. \\newpage The two $z_{\\rm abs} \\sim z_{\\rm em}$ systems seen in Q0125$-$376 could conceivably be associated with the absorption seen toward Q0122$-$380, falling squarely between the $z \\simeq 1.91$ complex and the weaker $z \\simeq 1.814$ system detected in that object by Carswell et al.\\ (1982). In a standard cosmological model with $q_0=0.5$ the angular distance between Q0125$-$376 and Q0122$-$380 of $50\\farcm2$ at $z=1.9$, corresponds to a projected comoving separation of D$_\\perp = 36\\,h^{-1}$ Mpc $(\\mbox{H$_0$}=h\\,100\\,\\mbox{km s$^{-1}$ Mpc$^{-1}$})$, which is comparable to the extent of local superclusters of galaxies. On the other hand, the presence of possibly BAL-like absorption at lower redshift would argue that the two $z_{\\rm abs} \\sim z_{\\rm em}$ systems seen in Q0125$-$376 are intrinsic in nature. While higher resolution observations would be required to further delineate these possibilities and properly map the absorption toward the three confirmed $z>1.8$ quasars above, the exploratory observations presented here already make it clear that the field surrounding Q0122$-$380 is not as promising for searching for large scale structure at high redshift as had initially been hoped." }, "0005/astro-ph0005041_arXiv.txt": { "abstract": "High-speed spectroscopy of two pulsating subdwarf B stars, KPD\\,2109+4401 and PB\\,8783, is presented. Radial motions are detected with the same frequencies as reported from photometric observations and with amplitudes of $\\sim 2\\kmsec$ in two or more independent modes. These represent the first direct observations of surface motion due to multimode non-radial oscillations in subdwarf B stars. In the case of the sdB+F binary PB\\,8783, the velocities of both components are resolved; high-frequency oscillations are found only in the sdB star and not the F star. There also appears to be evidence for mutual motion of the binary components. If confirmed, it implies that the F-type companion is $\\simge1.2$ times more massive than the sdB star, while the amplitude of the F star acceleration over 4 hours would constrain the orbital period to lie between 0.5 and 3.2d. ", "introduction": "The discovery of small-amplitude non-radial pulsations in a number of subdwarf B stars (sdBVs) has introduced a powerful new tool for the study of stellar remnants (Kilkenny et al. 1997=SDBV\\,I and subsequent papers). Long time-series photometric campaigns have detected rich spectra of oscillations in over a dozen targets, with frequencies and amplitudes generally indicative of low-order ($\\ell=0-2$) and low-degree modes. The discovery has revolutionized the study of subdwarf B stars for the simple reason that pulsations in these stars were not expected. From a theoretical point of view, it appears that the pulsation mechanism is only effective when diffusion processes modify the outer layers of the star. Metal-enrichment in a specific layer must conspire with the local temperature to drive pulsations through the opacity mechanism (Charpinet et al. 1997). In terms of their effective temperature and surface gravity, sdBVs are indistinguishable from their non-variable counterparts -- there is no instability strip in which all sdBs pulsate. The origin of all sdBs remains a puzzle, although the increasing detection of sdB binarity, including several sdBVs, may point to a previous phase of common-envelope evolution in many, if not all. The structure of sdBs is partially hidden as a consequence of atmospheric diffusion, which disguises the true composition and the mass of the hydrogen-rich envelope, both of which are key diagnostics of previous evolution. By enabling an exploration of the composition and mass of these outer layers, asteroseismology may be of a pivotal importance. Immediately after the discovery of sdBV pulsations, we recognized the potential for spectroscopy to provide additional diagnostics. Mode identification from photometry alone is challenging, and could be assisted by the identification of line-profile variations, whilst the comparison of radial and light amplitudes could be used to determine stellar radii directly. Spectroscopy might also demonstrate the presence of higher-order modes not detected photometrically. The limitations are that the periods are short (100-500s) compared with conventional CCD readout times, the stars are faint (12--15 mag.), and the photometric amplitudes are typically only a few tenths of one per cent. High-resolution high-S/N multi-line studies such as those obtained for non-radial oscillations in rapidly rotating bright O and B stars (e.g. Reid et al. 1993, Telting, Aerts \\& Mathias 1997) would not appear to be feasible. However, the development of new techniques offered the possibility to acquire high-speed spectroscopy of sdBVs and here we report our first successful observations obtained in 1998 October. Subsequently, O'Toole et al. (2000) announced preliminary results of independent radial velocity observations. They report a 9\\kmsec\\ amplitude periodic variation in the large amplitude sdBV PG1605+072 at the principal frequency of 2.10 mHz found photometrically by Koen et al. (1998, SDBV\\,VII). ", "conclusions": "High-speed high-resolution spectroscopy of two non-radially pulsating subdwarf B stars KPD\\,2109+4401 and PB\\,8783 has been acquired with the William Herschel telescope. These data show radial velocity variations at both high- and low-frequencies. High-frequency peaks in the velocity amplitude spectrum correspond to frequencies identified in the light curves of both stars, and have allowed the velocity amplitudes associated with these stellar oscillations to be estimated. Typically $\\sim2\\kmsec$, these translate into radial variations of some $100-250$km within 60--90s. The prospect for measuring the velocity amplitudes of non-radial pulsations in sdBVs with much higher amplitudes is promising. Line profile variations were not detected in the data although there are good reasons to suppose they should be present. The possibility that frequencies not present in the photometry may contribute to the velocity amplitude spectrum should also be pursued. In the case of the binary sdB PB\\,8783, low-frequency velocity variations corresponding to the mutual motion of both components have been resolved. These provide an upper limit to the orbital period of between 0.8 and 3d, depending principally on the orbital inclination. Establishing the orbital period for this star should therefore be a priority for future observations. The deconvolution of component star velocities also demonstrated that high-frequency velocity variations occur only in the sdB star and not in the F star." }, "0005/astro-ph0005331_arXiv.txt": { "abstract": "Similarity solutions are found for the adiabatic collapse of density perturbations $\\delta M/M \\propto r^{-\\sp}$ $(\\sp>0)$ in a flat universe containing collisional gas only. The solutions are obtained for planar, cylindrical, and spherical perturbations with zero initial pressure. For adiabatic index $\\gamma\\ge 4/3$, a shock develops at a fixed fraction of the current turnaround distance. Near the center of a spherical perturbations with $\\gamma>4/3$ and $\\sp > 1/2$, the gas is in quasi-hydrostatic equilibrium (pressure supported) and has an asymptotic power law density profile, $\\rho\\sim r^{-3\\sp/(\\sp+1)}$, independent of $\\gamma$. For $\\sp\\le 1/2 $, the profile depends on $\\gamma$, the pressure is finite, the temperature decreases inward, and gravity dominates pressure causing a continuous inward flow. Although for $1/2<\\sp<2$ the temperature decreases at the center, the gas is pressure supported. The pressure is finite in cylindrical perturbations for $\\sp\\le 2(\\gamma-1)/(3\\gamma-4)$, and in planar perturbations for any $\\sp>0$. We also derive the asymptotic behaviour of the gas variables near the center in a universe dominated by collisionless matter. In such a universe, the gas in a spherical perturbation with $\\sp<2$ cannot be pressure supported and the temperature approaches a constant near the center. The solutions and the asymptotic behaviour are relevant for modelling the gas distribution in galaxy clusters and pancake-like superclusters, and determining the structure of haloes of self-interacting dark matter with large interaction cross section. ", "introduction": "\\label{equations} We write the Newtonian equations of motion governing the adiabatic collapse of symmetric perturbations in a collisional fluid (gas) of adiabatic index $\\gamma$ and zero initial pressure. Except section \\ref{asymptotic:dm}, we restrict the analysis here and throughout to the collapse in a flat universe containing collisional gas only. The initial gas pressure is zero, so the expansion scale factor of the universe is $a(t)\\propto t^{2/3}$, the Hubble function is $H(t)=2/(3t)$, and the background density is $\\rho_c=3H^2/(8\\pi G)=1/(6\\pi G t^2)$. Denote by $r$ and $\\upsilon\\equiv \\dd r/ \\dd t$ the physical position and velocity of a gas shell, where $ r=0$ is the symmetry center of the perturbation. Further, let $\\rho(r,t)$ and $p(r,t)$ be the gas density and pressure at $r$. As in Fillmore \\& Goldreich (1984) define the mass within a distance $r$ from the symmetry center by $m(r,t)=\\int_0^r x^{\\nn-1}\\rho(x,t)\\dd x$, where $\\nn=1,2$, and 3 refer, respectively, to planar, cylindrical, and spherical perturbations. The mass within a fixed shell varies with time like $m \\sim t^{-2(3-\\nn)/3}$, because of the Hubble expansion along $3-\\nn$ of the axes. In this notation, the equations of motion are, the continuity equation, \\def\\ff{\\frac{2(3-\\nn)}{3}} \\begin{equation} \\frac{\\dd(\\rho t^{\\ff})}{\\dd t}=-t^\\ff \\rho r^{1-\\nn}\\partial_r(r^{\\nn-1}\\upsilon) \\; , \\label{eom1} \\end{equation} Euler, \\begin{equation} \\frac{\\dd \\upsilon}{\\dd t}-\\frac{2}{9}\\frac{3-\\nn}{\\nn}\\frac{r}{t^2} =-\\frac{\\partial_r p}{\\rho}- \\frac{4\\pi G m}{r^{n-1}} \\; , \\label{eom2} \\end{equation} adiabatic condition, \\begin{equation} \\frac{\\dd}{\\dd t}(p\\rho^{-\\gamma})=0 \\; , \\label{eom3} \\end{equation} and the relation, \\begin{equation} \\partial_r m= r^{\\nn -1}\\rho \\; . \\label{eom4} \\end{equation} These equations are satisfied everywhere except at the shock where the fluid variables are described by jump conditions obtained from mass, momentum, and energy conservation. The initial conditions leading to self-similar collapse are specified at an early time close to zero, $t_i$, as \\begin{eqnarray} \\label{inid} \\frac{\\delta M}{M}&=&\\left(\\frac{r}{r_0}\\right)^{-\\sp} \\; , \\\\ \\label{inivel} \\upsilon(r,t_i)&=&\\frac{2}{3t_i}r \\; , \\\\ \\label{inip} p(r,t_i)&=&0 \\; , \\label{inic} \\end{eqnarray} where $\\delta M/M$ is the mean density contrast interior to $r$, and $\\sp>0$. For cosmological initial conditions the initial density contrast must be tiny, so we will be interested in the solution in the region $r\\gg r_0$. A perturbation with $\\sp > \\nn$ can be realized by placing a high narrow positive density peak at the center ($\\ll r_0$) of a symmetric void with local density contrast $\\sim (-r^{-\\sp})$. The condition (\\ref{inivel}) means that a gas shell at $r$ moves initially with the general universal expansion. This condition can be relaxed to allow for a non vanishing initial zero peculiar velocity according to late time linear theory (e.g., Peebles 1980). However, this does not affect the details of the collapse (Peebles 1980, Bertschinger 1985), so we use (\\ref{inivel}) which is commonly adopted in the literature. Bertschinger (1985) and White \\& Forcada (1997), respectively, considered the collapse of spherical perturbations with $\\sp=3$, and $\\sp=2$. The equations of motion (\\ref{eom1}--\\ref{eom4}) together with the initial conditions (\\ref{inid}--\\ref{inip}) are insufficient to completely determine the evolution of the perturbation. Still missing is an inner boundary condition specifying the velocity and mass at $r=0$, for $t\\ge t_i$. For a shock to develop without the accumulation of a central mass (a black hole for $n=3$) we must have $v(r=0,t\\ge t_i)=0$ and $m(r=0,t\\ge t_i)=0$. Relaxing the condition $V(0)=0$ leads to a non-vanishing central mass with or without the presence of a shock. In a critical density universe ($\\Omega=1$) the only length scale relevant to the collapse is the scale of non-linearity. At any time, $t$, this scale can be defined as the distance of the shell at the maximum expansion, i.e., the shell with $\\upsilon=0$ (Gunn 1977, Fillmore \\& Goldreich 1984, Bertschinger 1985). This radius is termed the current turnaround radius, $r_{ta}(t)$. Starting from tiny initial density contrast, the mean overdensity (density in units of $\\rho_c$) interior to $r_{ta}(t)$ is a fixed number independent of time. For time $t\\gg t_i$, when shells with $r\\gg r_0$ reach their turnaround, the collapse develops a self-similar behaviour that depends on $r$ and $t$ through the combination $\\lambda=r/r_{ta}$. The turnaround radius $r_{ta}(t)$ is given by (e.g., Fillmore \\& Goldreich 1984), \\begin{equation} r_{ta}=r_0\\frac{C_x}{C_t^{3\\alpha/2}} \\left(\\frac{t}{t_i}\\right)^{\\alpha} \\quad ; \\quad \\alpha=\\frac{2}{3}\\frac{\\sp+1}{\\sp} \\end{equation} where, \\begin{equation} C_x=\\frac{5}{12}, 0.741, 1 \\quad ;\\quad C_t=\\frac{5}{6}, 1.386, \\left(\\frac{3\\pi}{4}\\right)^{2/3} \\; , \\end{equation} for $\\nn=1,2,$ and 3, respectively. The turnaround radius grows faster than the scale factor $a\\sim t^{2/3}$. This is because the mass, $\\sim \\rho_c r_{ta}(t)^3$, interior to $r_{ta}$ must grow with time while the mass, $\\sim \\rho_c a^3(t)$, inside a fixed shell in a homogeneous universe is constant. For $\\sp <2$ the turnaround radius grows faster than $t$ reaching the horizon scale in finite time. When this happens relativistic description must be used and $r_{ta}$ ceases to be the only scale in the problem (Fillmore \\& Goldreich 1984). The equations can be cast into a non-dimensional form using the scaled variables $V(\\lambda)$, $D(\\lambda)$, $P(\\lambda)$, and $M(\\lambda)$ defined by (Bertschinger 1985), \\begin{eqnarray} \\label{scalev}\\upsilon(r,t)&=&\\frac{r_{ta}}{t}V(\\lambda)\\\\ \\label{scaled}\\rho(r,t)&=&\\rho_c D(\\lambda)\\\\ \\label{scalep}p(r,t)&=&\\rho_c\\left(\\frac{r_{ta}}{t}\\right)^2 P(\\lambda)\\\\ \\label{scalem}m(r,t)&=&\\frac{1}{3}\\rho_c r_{ta}^\\nn M(\\lambda) \\; . \\end{eqnarray} Expressed in terms of these variables, the equations (\\ref{eom1}-\\ref{eom4}) become, respectively, \\begin{equation} \\label{a1} \\left(V-\\alpha\\lambda\\right)D'+\\left(\\frac{\\nn-1}{\\lambda}V+V'- \\frac {2\\nn}{3}\\right)D=0 \\; , \\end{equation} \\begin{equation} \\label{a2}\\left(\\alpha-1\\right)V+\\left(V-\\alpha\\lambda\\right)V' -\\frac{2}{9}\\frac{3-\\nn}{\\nn}\\lambda = -\\frac{P'}{D}-\\frac{2}{9}\\frac{M}{\\lambda^{\\nn-1}} \\; , \\end{equation} \\begin{equation} \\label{a3}\\left(\\gamma\\frac{D'}{D}- \\frac{P'}{P}\\right)\\left(V-\\alpha\\lambda\\right)=2\\left(\\alpha-2+\\gamma\\right) \\; , \\end{equation} \\begin{equation} \\label{a4}M'=3\\lambda^{\\nn-1}D \\; , \\end{equation} where the prime symbol denotes derivatives with respect to $\\lambda$. We will mainly be concerned with solutions for shocked collapse with vanishing mass at the center. The inner boundary condition appropriate for this collapse are vanishing mass and velocity at $\\lambda=0$, i.e., \\begin{equation} V(0)=0 \\quad {\\rm and} \\quad M(0)=0\\; , \\label{innerb} \\end{equation} Self-similarity implies that the shock appears at fixed $\\lambda=\\lambda_s=r_s/r_{ta}$, so the physical radius of the shock $r_s\\propto t^\\alpha$ and its non-dimensional speed is $(r_{ta}/t)^{-1}(\\dd r_s/\\dd t)= \\alpha\\lambda_s$. At the surface of the shock the fluid variables satisfy the jump conditions obtained from mass, momentum, and energy conservation. In terms of the non-dimensional fluid variables, the jump conditions appropriate for an adiabatic shock are, \\begin{eqnarray} \\label{jump1} V^+&=&\\alpha\\lambda_s+\\frac{\\gamma-1}{\\gamma+1}(V^- -\\alpha\\lambda_s)\\; ,\\\\ \\label{jump2}D^+&=&\\frac{\\gamma+1}{\\gamma-1}D^- \\; ,\\\\ \\label{jump3}P^+&=&\\frac{2}{\\gamma+1}D^-(V^--\\alpha\\lambda_s)^2\\; ,\\\\ \\label{jump4}M^+&=&M^- \\; , \\end{eqnarray} where the superscripts of the minus and plus signs refer to pre- and post-shock quantities. In employing energy conservation we have taken $\\frac{P}{D(\\gamma-1)}$ as the non-dimensional internal energy per unit mass. In section \\ref{numerical} we will find numerical solutions satisfying the requirements for shocked collapse without a central mass. Except spherical perturbations with $\\gamma=4/3$ only one value $\\lambda_s$ can yield solutions satisfying these requirements. Spherical perturbations with $\\gamma=4/3$ allow a range of values for $\\lambda_s$. Before presenting the numerical solutions we derive in the next section the asymptotic behaviour of the fluid variables near the center, and two integrals of motion which will be used as a check on the numerical solutions. ", "conclusions": "\\label{discussion} The similarity solutions are found for collapse in a flat universe with matter density parameter $\\Omega = 1$. Because of Birchhoff's theorem, the solutions for spherical collapse are valid in an open universe if the current turnaround radius is well inside the spherical region interior to which the perturbation is bound. The statement is incorrect for planar and cylindrical perturbations because of the explicit appearance of cosmology dependent terms in the equations of motion (\\ref{eom1}--\\ref{eom4}), like $t^{2(3-n)/3}$ in the continuity equation (\\ref{eom1}). The solutions are appropriate for the adiabatic collapse of perturbations with deep gravitational potential so that the initial thermal energy of the gas can be ignored. Such perturbations are probably the seeds for massive galaxies, galaxy groups and clusters. In the intergalactic medium (IGM) most of the gas is continuously photo-heated and is of moderate density. There is considerable interest in analytic modelling of the IGM in current methods for extracting cosmological information from the Lyman forest (Croft et. al. 1998, Nusser \\& Haehnelt 1999, 2000). So far these methods have heavily relied on linear analysis (e.g., Bi, B\\\"orner, \\& Chu 1992, Gnedin \\& Hui 1998, Nusser 2000) and hydrodynamical simulations (e.g., Petitjean et. al. 1995, Theuns et. al. 1999). Analytic treatment of the IGM beyond the linear regime is exceedingly complicated. Consider a situation in which photo-heating establishes the relation $p=k \\rho^\\gamma$ in the IGM, where $k$ and $\\gamma$ depend non-trivially on time (e.g., Theuns et. al. 1999). The pressure in this case introduces a length scale $k^{1/2} G^{(1-\\gamma)/2} t^{2-\\gamma}$ (e.g., Sedov 1959). If we take constant $k$ and $\\gamma$, this length scale varies with time like $r_{ta}$ only in the special case of $\\gamma=4/3$ and infinite $\\sp$. So physically interesting situations including initial pressure in which the collapse is self-similar do not exist. Spherical Perturbations with $\\sp>2$ when $\\gamma>4/3$, and with any $\\sp>0 $ when $\\gamma=4/3$ deserve special attention. In the corresponding solutions for shocked collapse without a black hole at the center, the quantity $V_c^2=2Gm/r$ diverges towards the center, and so there is a point $r=r_g$ at which $V_c^2=c^2$, where $c$ is the speed of light. According to general relativity this implies the presence of a black hole at the center, invalidating the assumption of no central mass, made in deriving the solutions. In particular, the condition $\\upsilon(r=0)=0$, necessary for shocked collapse, is incompatible with the presence of a central black hole. Near $r_g$, however, radiation pressure and angular momentum can prevent the formation of a black hole. Should this occur, we expect our solution for shocked accretion to be valid away from the central region. The evolved gas variables in the symmetric self-similar collapse contain full information on the initial perturbation. So the system retains memory of the initial conditions, even in the highly nonlinear regime. On the other hand, a collapsing system of collisionless matter can develop density profiles which do not depend on the initial shape of the perturbation. For example, according to the solutions of Fillmore \\& Goldreich (1984), a spherical density perturbation develops into $r^{-2}$ for $\\sp<2$, and a cylindrical perturbation into $r^{-1}$ for all $\\sp$. Haloes identified in cosmological simulations of collisionless particles with generic initial conditions, also tend to have density profiles independent of the spectrum of the initial fluctuations (Navarro, Frenk \\& White 1997). Our results are relevant for describing the gas distribution in various physical systems such as the cores of clusters or pancake-like superclusters. Over a limited range of scales, the index $\\sp$ can be related to the index, $l$, of the three dimensional power spectrum, $p(k)\\sim k^l$, of the linear density fluctuations. If the initial density field is gaussian with a scale free power spectrum then the properties of the nonlinear field depend only on one scale. This is the nonlinear scale, $R_{nl}$, defined as the scale on which the rms value of density fluctuations is unity. This scale grows with time like\\footnote{$R_{nl}$ does not involve the dimension $\\nn$ because $l$ refers to the three dimensional $p(k)$ so the rms value on a scale $R$ is $R^{(l+3)/2}$ independent of $\\nn$. } $R_{nl}\\sim t^{\\frac{2(l+5)}{3(l+3)}}$. By matching the time dependence of $R_{nl}$ and $r_{ta}\\sim t^{\\frac{2(\\sp+1)}{3\\sp}}$ we identify $\\sp=(l+3)/2$. So the collapse of gas into galaxies and clusters can be, respectively, modeled by our solutions for $ \\sp\\!\\sim\\! 0.4$ and $0.65$, where we have taken $l\\!\\sim\\! -2.2$ and $-1.7$ assuming a Cold Dark Matter power spectrum. Taking $l\\!\\sim\\! -1$ for collapse on a pancake-like large scale superclusters gives $\\sp\\sim 1$. Another way to relate $\\sp$ and $l$ is to identify symmetric perturbations with local maxima in the linear density field (Hoffman \\& Shaham 1985). The shape of high density peaks in a gaussian field varies with $r$ like the two-point correlation function, $\\sim r^{-(l+3)}$ (e.g., Bardeen et. al. 1986). So, at least in the limit of high peaks, $\\sp=l+3$. On galaxy and cluster scales the relation $\\sp=l+3$, respectively, gives $\\sp=0.8$ and $1.3$ in contrast to $\\sp=0.4 $ and $0.65$ obtained from $\\sp=(l+3)/2$. Note however that gas in galaxies tends to settle into disks and so our solutions are less relevant than they are for cluster size objects. In spherical geometry the asymptotic behaviour shows that the gas cannot be pressure supported if $\\sp<1/2$, and $\\sp<2$ for collapse with, and without collisionless matter, respectively. Estimates of the masses of rich galaxy clusters from X-ray observations of the intracluster gas rely on hydrostatic equilibrium (e.g., Fabian 1994). If on cluster scales $\\sp\\sim 0.7$--$\\sim 1.3$, then the asymptotic behaviour implies that the cluster gas may not be in hydrostatic equilibrium. How large is the error introduced in the mass estimates by assuming hydrostatic equilibrium? The following argument shows that this error is negligible. Hydrostatic equilibrium calculations neglect the term $ G^{-1}r^2\\dd \\upsilon /\\dd t$ in the mass estimate. Using the asymptotic expansion one finds that neglecting this term amounts to a relative mass error of $\\sim (2/\\pi^2) (t/t_{ta})^{3V_0-2}$ where $t_{ta}$ is the turnaround time of the shell present at $r$ at the current time $t$. Shells in the inner regions have passed their maximum expansion a few dynamical times ago. Therefore $t\\gg t_{ta}$ and since $V_0<0$ we conclude that the error is negligible. The solutions are related to modelling the structure of haloes made of self interacting dark matter (SIDM) (Spergel \\& Steinhardt 1999) with large interaction cross section. On scales of massive galaxies and clusters, our results predict final density profile $\\rho\\sim r^{-1.2}$ to $\\sim r^{-1.7}$. These profiles are consistent with the results obtained by Moore et. al. in their simulations of SIDM with large cross section." }, "0005/astro-ph0005107_arXiv.txt": { "abstract": "A new semi-analytical model of a star evolving in a tidal field is proposed. The model is a generalization of the so-called 'affine' stellar model. In our model the star is composed of elliptical shells with different parameters and different orientations, depending on time and on the radial Lagrangian coordinate of the shell. The evolution equations of this model are derived from the virial relations under certain assumptions, and the integrals of motion are identified. It is shown that the evolution equations can be deduced from a variational principle. The evolution equations are solved numerically and compared quantitatively with the results of 3D numerical computations of the tidal interaction of a star with a supermassive black hole. The comparison shows very good agreement between the main ``integral'' characteristics describing the tidal interaction event in our model and in the 3D computations. Our model is effectively a one-dimensional Lagrangian model from the point of view of numerical computations, and therefore it can be evolved numerically $10^{2}-10^{3}$ times faster than the 3D approach allows. This makes our model well suited for intensive calculations covering the whole parameter space of the problem. ", "introduction": "Starting from the seminal paper by Roche the problem of the tidal influence of a gravitating source on a satellite has been addressed by numerous researchers. More recently, interest in this problem has been raised by a paper of Hills (Hills 1975), who proposed tidal disruption processes as the main processes of fueling of QSO's and AGN's. From the point of view of the astrophysics of QSO's and AGN's there are several approaches to that problem. Firstly one can consider the tidal interaction event as an elementary process in the complicated astrophysical environment of a supermassive black hole, presumably situated in the cores of QSO's and AGN's. Then one could find the main evolutionary characteristics of such a system and its average luminosity, taking into account additional gas dynamical and stellar dynamical processes occurring in the cores (e.g. Hills, 1975, Frank $\\&$ Rees 1976, Young et al 1977; Young 1977; Hills 1978; Frank 1979; Gurzadian $\\&$ Ozernoi 1980; Lacy et al 1982; Illarionov $\\&$ Romanova 1986a; Illarionov $\\&$ Romanova 1986b; Dokuchaev 1991; Beloborodov et al 1992; Roos 1992; Syer $\\&$ Ulmer 1999; Magorrian $\\&$ Tremaine 1999). One can also consider the evolution of remnants of a single tidal stripping or tidal disruption event and find the characteristic luminosity change of the object due to the accretion of the remnants through an accretion disk or quasi-spherical configuration onto the central black hole (e.g. Lacy et all 1982; Rees 1988; Evans $\\&$ Kochanek 1989; Cannizzo et all 1990; Roos 1992; Kochanek 1994; Ulmer et al 1998; Kim et al 1999; Syer $\\&$ Ulmer 1999; Ulmer 1999). On the other hand it is very important to understand quantitatively the main characteristics of the tidal encounter itself, and a lot of of work has been devoted to the physical processes occurring in a star during its fly-by around a black hole. The papers on that subject could be classified by the different stellar models used in the calculations. The simplest possible approach to the problem uses an incompressible model of the star. Thus one can reduce the complicated hydrodynamical nonlinear partial differential equation governing the evolution of the stellar gas to a set of ordinary differential equations, which are easy to analyze by analytical and numerical means. The study of incompressible models has been performed for Newtonian and relativistic tidal fields and different kinds of orbits of the star (e.g. Nduka, 1971; Fishbone, 1973; Mashhoon, 1975; Luminet $\\&$ Carter 1986; Kosovichev $\\&$ Novikov, 1992). However, this approach is highly unrealistic, since effects determined by the compressibility of the star can play a major role during the tidal disruption event (e.g. Carter $\\&$ Luminet 1982). A significant step forward was made by Lattimer and Schramm (Lattimer $\\&$ Schramm 1976) and by Carter and Luminet (Carter $\\&$ Luminet 1983, 1985) who proposed the so-called affine model of the tidally disrupted star, which allows for the compressibility of the stellar gas. In this model the law of time evolution of different elements of the star is defined in terms of some spatially uniform $3\\times 3$ matrix ${\\bf Q}(t)$: $$x^{i}=Q^{i}_{j}(t)x^{j}_{0}, $$ where $x^{i}$ are the components of the position vector of a gas element, $x^{j}_{0}$ are the components of the position vector in some reference state (say, before the tidal field ``is switched on''), and summation over repeated indices is assumed. Then one can find the evolution equations for the matrix elements from the so-called virial relations written for the whole star. The affine model has successfully been applied to the problem of tidal interaction and tidal disruption of a star by a supermassive black hole during close encounters (e.g. Carter $\\&$ Luminet 1983, 1985; Luminet $\\&$ Mark 1985; Luminet $\\&$ Carter 1986; Luminet $\\&$ Pichon 1986; Novikov et al 1992; Diener et al 1995). Lai, Rasio and Shapiro used the same model for an approximate treatment of an isolated rotating star, as well as for a star in a binary system (e.g. Lai et al 1994; Lai $\\&$ Shapiro 1995). Recent progress in numerical simulations has allowed researchers to perform direct 3D simulations of the tidal interaction and tidal disruption events. The first SPH simulations were run in the beginning of eighties by Nolthenius and Katz, and by Bicknell and Gingold, although the number of particles in these simulations was too small to be representative (Nolthenius $\\&$ Katz 1982, 1983; Bicknell $\\&$ Gingold 1983). In the following decades, SPH simulations were improved both by increasing of the number of particles, and by using of more complicated stellar models (e.g. Evans $\\&$ Kochanek 1989; Laguna et al 1993; Laguna 1994; Fulbright et al 1995; Ayal et al 2000). Three dimensional finite difference simulations were done by Khokhlov, Novikov and Pethick for a polytropic star in a Newtonian tidal field (Khokhlov et al 1993a,b, hereafter Kh a,b), by Frolov, Khokhlov, Novikov and Pethick for a white dwarf (Frolov et al 1994), and by Diener, Frolov, Khokhlov, Novikov and Pethick for a polytropic star in the tidal field of a Kerr black hole (Diener et al 1997). An interesting attempt to combine the affine model and a simple version of 3D finite difference hydrodynamics has been made by Mark, Lioure and Bonazzola (Mark et al 1996). Although the 3D simulations promise the most direct and thoughtful approach to the problem, they are still very time consuming. All in all, less than one hundred different sets of values of the problem parameters have been tested with numerical experiments, and due to very poor statistics these experiments cannot be used to characterize the general properties of the tidal encounters for a broad range of available parameters. There is another, more fundamental difficulty connected with the 3D simulations. The complexity of 3D hydrodynamical flows makes the interpretation of the results of numerical work increasingly difficult. The situation is reminiscent of a real physical experiment, and a simple 'reference' model of the tidally disrupted star would be very welcome in order to interpret the results of the numerical simulations. On the other hand, the astrophysics of AGN's and QSO's requires a rather rough description of a single tidal encounter, and only a few 'averaged' quantities such as e.g. the amount of mass lost by the star during the tidal interaction, or the amount of energy deposited in the star by the tidal forces are of interest from the astrophysical viewpoint. In this paper we propose a new, semi-analytical model of the tidally interacting or tidally disrupted star which could be used for intensive calculations covering the whole parameter space of the problem, and also as a 'reference' model for 3D simulations. Our model is a straightforward generalization of the affine model. However, in contrast to the affine model, the different layers of the star evolve differently in our model, and are connected to each other by a force determined by pressure. This allows us to employ our model for calculation of quantities such as the loss of mass from the star after a fly-by over a black hole without complete disruption, which cannot be calculated in the affine approximation. Instead of the position matrix ${\\bf Q}(t)$ of the affine model, we use the position matrix ${\\bf T}(t, r_{0})$, which depends not only on time, but in addition on the value $r_{0}$ of the 'reference' vector $x_{0}^{i}$ (obviously the radius $r_{0}$ plays the role of a Lagrangian coordinate, so we will later call it the Lagrangian radius). Thus, in our model the star consists of elliptical shells which are composed of all elements of the star with a given Lagrangian radius $r_{0}$. The evolution of the shell depends on the Lagrangian radius, and therefore the shells have different ratios between their major axes and different rotation angles with respect to a (locally inertial) coordinate frame centered on the star's center of mass, for the different values of the Lagrangian radius. The evolution equations of our model follow from the virial relations written for each shell (see e.g. Chandrasekhar 1969, hereafter Ch). Unlike the affine model the virial relations written for a shell inside the star must contain surface terms, and these surface terms lead to interactions between shells with different Lagrangian radii, and therefore to the propagation of a disturbance through the star. In fact, the evolution equations are of hyperbolic type, and the disturbance induced by a tidal field propagates over the star as a non-linear sound wave. We derive the evolution equations of our model in the next Section using certain approximations for the pressure terms, and the terms describing the self-gravity of the star. In the simplest formulation of our model the interaction between the shells depends only on their relative volumes, and therefore the shells are allowed to intersect each other. Therefore the position matrix ${\\bf T}$ has no direct physical meaning in such a case, and one should only use quantities averaged over many shells in order to infer physical information (such as e.g. the energy and the angular momentum contained inside some part of the star, the components of the quadrupole moment tensor for that part, the amount of mass lost by the star, and so on). We show that the energy and the angular momentum are well defined in our model, and derive the law of evolution of these quantities due to the presence of a tidal field. We also show that the circulation of velocity of a gas element over the shell is exactly conserved even in the presence of the tidal field. Then we apply our model to the simplest problem, the parabolic fly-by of a polytropic star around a source of Newtonian gravity, and numerically calculate the evolution of the quantities characterizing the star during the fly-by. We compare our results with 3D finite difference simulations of Kh a,b for the same problem and the same parameters, and find very good agreement. Additionally we compare our model with results of SPH simulations, calculations based on the affine model and results from the linear theory of tidal perturbations (Press $\\&$ Teukolsky 1977; Lee $\\&$ Ostriker 1986). Then we calculate the energy deposited in the star, its angular momentum and the amount of mass lost by the star as a function of the pericentric separation between the star and the center of gravity. As it will be clear for the results in Section 3, our model gives a better agreement with the results of the 3D simulations than the affine model. We use a rather unusual summation convention assuming that summation is performed over all indices appearing in our expressions more than once, but summation is not performed if indices are enclosed in brackets. Bold letters represent matrices in abstract form. All indices can be raised or lowered with help of the Kronecker delta symbol, but nevertheless we distinguish between the upper and lower indices in order to enumerate the rows and columns of matrices, respectively. Therefore, the expression $A^{k}_{i}B^{l}_{i}=C^{kl}$ means ${\\bf AB}^{T}={\\bf C}$, and $A^{i}_{k}B^{i}_{l}=C_{kl}$ means ${\\bf A}^{T}{\\bf B}={\\bf C}$ (here $T$ stands for the transpose of a matrix). Finally, we would like to list the main approximations made in the derivation of the dynamical equations of our model. 1) We assume that the star is composed of elliptical shells, and the shells are not deformed during the evolution of the star in a tidal field. 2) We calculate the self-gravity of the star in a simplified manner. Namely, in order to calculate the force of gravity acting on some particular shell, we neglect the contribution of the star's mass concentrated in the outer (with respect to that shell) layers of the star. It is also assumed that the gravitational force determined by the inner layers is equivalent to the gravitational force of a uniform density ellipsoid inserted in this shell. 3) We assume a polytropic equation of state of the stellar gas. 4) We use an ``averaged'' density and an ``averaged'' pressure instead of the exact quantities. These averaged quantities depend only on time and the Lagrangian radius of the star. The approximations 1,2 are essential for our model. The approximations 3,4 can be relaxed in a more advanced variant of the model. ", "conclusions": "We have constructed a new model of a star perturbed by a tidal field. In this model the star consists of a set of elliptical shells which in general have different principal axes and different orientations with respect to a fixed locally inertial reference frame comoving with the star's center of mass. The model obeys certain evolution equations. The results of calculations of a simple problem of tidal encounter of a polytropic star moving on a parabolic orbit around a source of Newtonian gravity have been compared with the results of three dimensional finite difference simulations. We found that the main characteristics of the tidal encounter agree with the results of the finite difference approach with a typical accuracy $10-20$ percent. Taking into account that the astrophysical applications do not demand very high accuracy in description of a single tidal encounter, we think that our model could be used in order to investigate all possible variants of the problem of tidal interaction between a supermassive black hole and a star interesting from an astrophysical point of view. The main advantage of our model is its effectively one dimensional character, which allows us to calculate all interesting variables much faster than the 3D approach, and over a much longer time of evolution. Also, all characteristics of the tidally perturbed star could be inferred from our model in a straightforward and unambiguous way. Our model could also be used in a study of a rotating single star, or a binary star. In principal, the agreement between our model and 3D computations could be improved if one uses more advanced variants of our model (see below). However we would like to note that in order to make a comparison between an advanced variant of our model and 3D computations, one should also increase both the number of the numerical experiments and their resolution. For example, recently it has been claimed (Ayal et al, 2000) that the difference between the different numerical experiments in SPH models is about 10 percent depending on number of particles used in the calculations. We are not aware of similar convergence studies for the 3D finite difference models. We think that such studies should be undertaken parallel to work on improvement of our model. Now we would like to discuss possible extensions of our formalism. One obvious way for such extension consists in using a more complicated stellar model for the unperturbed state. Then the realistic stellar models could be generalized to our problem by using the energy evolution equation and the virial relations similar to the eq. (8,9), but written for a realistic stellar gas, with possible inclusion of e.g. non-adiabatic effects, viscosity, effects of radiative transfer, and generation of heat due to nuclear reactions in the stellar core. The evolution equations for such a model could be derived from the energy equation and the virial relations in the way described above. One could also use a more refined numerical scheme, say an implicit scheme with a nonuniform grid. We suppose that certain powerful methods developed in numerical investigations of pulsating stars could be directly applied to our problem. Another interesting extension consists in using the real distribution of the pressure and the density over the volume of the star in our approximation. Thus it is possible to construct a model with no intersection between the shells, which could provide more information about displacements of particular elements of the star during the tidal encounter. Assuming that the star consists of an ideal gas with constant ratio of specific heats $\\gamma$, and neglecting the possible presence of shocks in the system, this problem is reduced to the evaluation of the following integrals: $$I_{1}=\\int_{S}R^{2(\\gamma -1)}d\\Omega, \\eqno 35$$ and $$I_{2i}=\\int_{S}l^{2}_{(i)}R^{2\\gamma}d\\Omega, \\eqno 36$$ where the integration is performed over the surface of an ellipsoid $\\sum \\lambda_{(i)}y^{2}_{(i)}=1$. $\\lambda_{i} > 0$ are the eigenvalues of the matrix $R^{ln}$ (see. eq. 4), $R$ is the value of the radius vector joining the center of the ellipsoid to a point on its surface, $l_{i}$ are the direction cosines of the radius vector, and $d\\Omega$ is the elementary solid angle. The Cartesian coordinates $y_{i}$ are associated with the frame of eigenvectors of the matrix $R^{ln}$. If $\\gamma=1$ the integral (35) is trivial, and the integrals (36) are related to the integrals $D_{i}$ defined above (eq. 18), but with $\\lambda_{j}^{-1}$ playing the role of $a_{j}^{2}$. One can see that the pressure tensor (defined as the left hand side of the eq. 13) can be evaluated with help of the integrals (35, 36). The volume part of the pressure tensor (the first term on l.h.s. of the eq. (13)) is obtained by integration over the mass of a quantity proportional to the integral (35) with the upper limit of integration determined by some given Lagrangian coordinate $x_{*}$. The surface term ( the second term on l.h.s. of the eq. (13)) is expressed with help of the integrals (36) with $\\lambda_{i}=\\lambda_{i}(x_{*})$. Note that now the surface term is not symmetric, and therefore transfer of angular momentum between the neighboring shells due to pressure is allowed. If some shells are close to intersection, the density at some particular value of $x$ tends to infinity. That means that some eigenvalues $\\lambda_{i}$ go to zero, and as a consequence the integrals (35), (36) tend to infinity causing an increase of pressure. In turn the increase of pressure could prevent the shells from intersecting. The integrals (35), (36) can be evaluated e.g. by numerical means, and serve as main building blocks of a model without intersections between the shells corresponding to different Lagrangian radii. Finally one can generalize our model to the case of a relativistic tidal field. In a separate paper we apply our model to the problem of tidal interaction of a star with a supermassive Kerr black hole." }, "0005/astro-ph0005382_arXiv.txt": { "abstract": "We report a high resolution (0\\farcs2), 15~GHz survey of a sample of 48 low-luminosity active galactic nuclei with the Very Large Array{\\footnote{The VLA is operated by the National Radio Astronomy Observatory, a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.}}. Compact radio emission has been detected above a flux density of 1.1~mJy in 57\\% (17 of 30) of low-ionization nuclear emission-line region (LINER) nuclei and low-luminosity Seyferts. The 2~cm radio power is significantly correlated with the emission-line ([O~I]~$\\lambda$6300) luminosity. Using radio fluxes at other frequencies from the literature, we find that at least 15 of the 18 detected radio cores have a flat to inverted spectrum ($\\alpha~\\geq$ $-$0.3, S$_{\\nu}~\\propto~\\nu^{\\alpha}$). While the present observations are consistent with the radio emission originating in star forming regions (the brightness temperatures are $\\geq$~10$^{2.5-4.5}$~K), higher resolution radio observations of 10 of the detected sources, reported in a companion paper \\citep{falet00}, show that the cores are very compact ($\\lesssim$~pc), of high brightness temperature (T$_b~\\gtrsim$ 10$^8$~K) and probably synchrotron self-absorbed, ruling out a starburst origin. Thus, our results suggest that at least 50\\% of low-luminosity Seyferts and LINERs in the sample are accretion powered, with the radio emission presumably coming from jets or advection-dominated accretion flows. We have detected only 1 of 18 ``transition'' (i.e. LINER + H~II) nuclei observed, indicating their radio cores are significantly weaker than those of ``pure'' LINERs. Compact 2~cm radio cores are found in both type 1 (i.e. with broad H$\\alpha$) and type 2 (without broad H$\\alpha$) nuclei. There is weak evidence, limited in significance by small numbers, that low-luminosity active galactic nuclei with compact radio cores exhibit radio ejecta preferentially aligned along the rotation axis of the galaxy disk. If this result is confirmed by a larger sample, it would lend support to the idea that the misalignment of accretion disks with the galaxy stellar disk in more luminous Seyfert galaxies is a result of radiation-pressure induced warping of their accretion disks. ", "introduction": "There is increasing evidence that a large fraction of the nuclei of nearby galaxies show many similarities with powerful active galactic nuclei (AGN); these objects are termed low-luminosity active galactic nuclei (LLAGN; active galaxies with nuclear L$_{H\\alpha}~\\leq$ 10$^{40}$ erg s$^{-1}$; Ho, Filippenko \\& Sargent 1997a, hereafter H97a). These similarities include broad H$\\alpha$ lines (Ho et al. 1997b), broader H$\\alpha$ lines in polarized emission than in total emission \\citep{baret99}, nuclear UV sources \\citep{maoet95,baret98} and water vapor megamasers \\citep{braet97}. However, the emission-line spectra of LLAGNs (i.e. low-luminosity Seyferts, LINERs, and ``transition'' nuclei [nuclei with spectra intermediate between those of LINERs and H~II regions]), can also be modeled in terms of photoionization by hot, young stars \\citep{termel85,filter92,shi92}, by collisional ionization in shocks \\citep{kosost76,foset78,hec80,teret92,dopsut95}, or by aging starbursts \\citep{aloet99}. Thus, it is not clear that accretion onto a black hole powers all LLAGNs. How does one distinguish accretion-powered LLAGNs from LLAGNs powered by hot stars or supernova shocks? Broad H$\\alpha$ lines and bright unresolved optical or UV sources are ambiguous indicators because they can all be produced in starburst models (see Terlevich et al. 1992), and a search for broader polarized H$\\alpha$ emission is currently feasible in only a few of the brightest LLAGNs \\citep{baret99}. Further, all these indicators are highly dependent on viewing geometry and obscuration and on the signal to noise of the observations, a problem exacerbated by the low optical and UV luminosities of LLAGNs. Their observation demands both high signal-to-noise spectra and high spatial resolution to separate weak nuclear emission lines from the starlight of the host galaxy. One well-known property of some powerful AGNs is a compact, flat-spectrum nuclear radio source, usually interpreted as the synchrotron self-absorbed base of the jet which fuels larger-scale radio emission. Astrophysical jets are known to be produced in systems undergoing accretion onto a compact object \\citep[e.g.][]{pri93,bla93}, so such compact radio sources in galactic nuclei may reasonably be considered a signature of an AGN. Much theoretical work \\citep[e.g.][]{beget84,lovrom96,falbie99} has been devoted to this disk-jet relationship in the case of galactic nuclei and it has been suggested that scaled-down versions of AGN jets can produce flat-spectrum radio cores in LLAGNs \\citep{fal96,falbie96}. Nuclear radio emission with an inverted spectrum is also expected from an advection-dominated accretion flow \\citep[ADAF; e.g.][]{naret98}, a possible form of accretion onto a black hole at low accretion rates \\citep{reeet82}, or from bremsstrahlung, cyclotron and synchrotron emission from plasma accreting quasi-spherically onto a black hole \\citep{mel94}. From the observational perspective, \\citet{hec80} showed that LINER nuclei tend to be associated with a compact radio source, and compact, flat-spectrum radio cores are known to be present in many `normal' E/S0 galaxies \\citep{sadet89, wrohee91, sleet94}. Flat-spectrum radio cores are, however, uncommon in normal spirals or Seyfert galaxies \\citep{ulvwil89,vilet90,sadet95}. Flat-spectrum radio sources can also result through thermal emission from optically-thin ionized gas or through free-free absorption of non-thermal radio emission, a process which probably occurs in compact nuclear starbursts \\citep{conet91}. The brightness temperature, T$_b$, in such starbursts is limited to log~[T$_b~$(K)]~$\\lesssim$ 5 \\citep{conet91}. Thus it is necessary to show that T$_b$ exceeds this limit before accretion onto a black hole can be claimed as the power source. In this paper, we present a high-frequency (15~GHz or 2~cm), high-resolution ($\\sim$~0{\\farcs}15) survey of LLAGNs with the Very Large Array \\citep[VLA,][]{thoet80}. This resolution is high enough to isolate nuclear emission from that of the host galaxy, and the radiation is unaffected by the obscuration present at UV or optical wavelengths, and less affected by free-free absorption than observations at longer cm wavelengths. Further, large samples can be quickly studied, as deep radio maps are achievable in as little as 15--20 min per object. Higher resolution follow-up observations are then required in order to eliminate the possibility that the radio cores are thermal in origin, as discussed above. We have therefore embarked on a program to observe a large number of LLAGNs at high resolution with the VLA and the Very Long Baseline Array \\cite[VLBA,][]{napet94} in order to identify accretion-powered nuclei, to test the predictions of ADAF and jet models, and to characterize the presence and structure of radio jets on sub-pc to hundred-pc scales. In this paper, we report on the results of the first stage of this program --- VLA observations of 48 LLAGNs which have been extensively observed at other wavebands. Preliminary results of this project have been published in \\citet{falet97,falet98,falet99} and \\citet{naget99b}. ", "conclusions": "Our detection of compact, flat-spectrum radio cores in about 50\\% of low-luminosity Seyferts and LINERs in a sample of 48 low-luminosity AGNs, when combined with VLBA observations \\citep{falet00}, suggests that at least half of all low-luminosity Seyferts and LINERs are accretion powered. Given the sensitivity limit of our survey, the true incidence of radio cores is likely to be higher. The 2~cm radio power is significantly correlated with the [O~I]~$\\lambda$6300 luminosity for LLAGNs in our sample. The lower detection rate (at $\\geq$92\\% significance) of compact radio cores in ``transition'' nuclei suggests that either these nuclei are intrinsically different from ``pure'' LINERs, or their ``pure'' LINER component is of lower luminosity, with correspondingly lower luminosity radio cores. The latter interpretation is favored by the correlation between the radio and [O~I]~$\\lambda$6300 luminosities coupled with the lower [O~I]~$\\lambda$6300 luminosities of the ``transition'' nuclei as compared to the LINERs. The presence or absence of a detected broad H$\\alpha$ line is not a good indicator of the presence or absence of a compact, flat-spectrum 2~cm radio core. A significant number of the LINERs and low-luminosity Seyferts which contain 2~cm compact radio cores show evidence, at low resolution (1{\\arcsec}--5{\\arcsec}) and frequency (1.4~GHz or 20~cm), for extended radio emission along the galaxy disk minor axis. This radio emission may trace a wide-angle outflow or a weak, highly-collimated jet along the disk rotation axis. If the latter is true, it lends support to the idea that it is the ``radiative instability'' which causes warps in the nuclear accretion disks of more luminous Seyfert galaxies. Finally, we note that the data presented here are the initial results of a larger program to study a well-defined sample of LLAGNs at high resolution with the VLA and VLBA; results of these will appear in future papers in this series." }, "0005/astro-ph0005457_arXiv.txt": { "abstract": "We have carried out numerical evolutionary calculations of binary systems to investigate the formation of binary millisecond pulsars (pulsars with white dwarf companions). We apply the ``standard scenario'' in which the binary pulsars are formed from low-mass and intermediate-mass X-ray binaries as well the alternative scenario in which the neutron stars are formed by accretion-induced collapse (AIC) of white dwarfs. The mass transfer processes are carefully followed by taking into account a number of binary interactions. Assuming that the magnetic fields of the neutron stars decay due to the accretion, we calculate the pulsar surface dipole magnetic field strength at the end of the mass transfer as a function of the final orbital period. We find that while the observed data of the majority of pulsars are compatible with the derived relations, we fail to produce binary pulsars with relatively high magnetic fields and short orbital periods (such as PSR B0655+64). We conclude that those systems are most likely formed through common-envelope phase. ", "introduction": "It has been suggested that binary pulsars with white dwarf companions (binary millisecond pulsars or BMSPs) are the descendants of low-mass X-ray binaries (LMXBs; Joss \\& Rappaport 1983, Savonije 1983, and Paczynski 1983) or intermediate-mass X-ray binaries (IMXBs; Podsiadlowski \\& Rappaport 2000; Tauris, et al. 2000). The white dwarf companions of the pulsars are the remnants of the donor stars which have dumped matter to the neutron stars during the X-ray phase. The nearly circular orbits observed in the systems support the view that extensive mass transfer has happened in the systems (before mass transfer the orbits must have been eccentric due to the supernova explosions which have formed the neutron stars, tidal effects become effective to circularize the orbits when the donor stars swell up to fill up their Roche lobes). Hence, pulsars in those systems will have accreted some amount of mass and have been recycled (spun up) by the accretion. The millisecond spin periods observed in many of the pulsars (despite their old ages, as the nondegenerate companions need some $10^8$ -- $10^9$ yr to become white dwarfs) indicate that they are indeed recycled pulsars. Magnetic fields of isolated pulsars do not decay significantly during their lifetime (Bhattacharya et al. 1992). However, many pulsars which are, or have been members of binary systems have relatively weak surface magnetic fields ($10^8$ -- $10^9$ G). This leads to the suggestion that the magnetic field of neutron stars decays due to the accretion (Taam \\& van den Heuvel 1986, Shibazaki et al. 1989) or related effects, such as spin up and spin down (Srinivasan et al. 1990). However, a number of BMSPs are observed to have relatively strong fields ($\\sim 10^{10}$ -- $10^{11}$ G; see Table 1) which implies that those neutron stars have accreted only a small amount of matter. We investigate a number of evolutionary models of binary systems which lead to the formation of BMSPs and examine whether such high-magnetic field BMSPs can be formed. In Sect. 2 we discuss the ``standard scenario'' in which the systems are formed from LMXBs or IMXBs. In Sect. 3 we discuss an alternative scenario in which the pulsars are formed by accretion-induced collapse (AIC) of white dwarfs. The discussion and conclusion are given in Sect. 4. \\begin{table} \\caption[ ]{Binary millisecond pulsars with relatively strong magnetic fields. f(M) is the mass function and $M_{\\rm wd}$ is the mass of the white dwarf companion with the assumption $i = 60 \\degr$ (Tauris \\& Savonije 1999, Tauris et al. 2000)} \\begin{flushleft} \\begin{tabular}{llllll} \\hline\\noalign{\\smallskip} PSR & $P_{\\rm orb}$ & $P_{\\rm spin}$ & $\\log(B)$ &f(M) & $M_{\\rm wd}$ \\\\ & (days) & (ms) & (G) &(\\Msun)& (\\Msun) \\\\ \\hline\\noalign{\\smallskip} B1831-00 & 1.811 & 520.95 & 10.94 &0.000124 &0.075 \\\\ B0655+64 & 1.029 & 195.67 & 10.07 &0.0714 &0.814 \\\\ J1232-6501 & 1.863 & 88.3 & 9.97 &0.0014 &0.175 \\\\ J1157-5112 & 3.507 & 43.6 & $< 9.80$ &0.2546 &$>1.2$ \\\\ \\hline\\noalign{\\smallskip} J1803-2712 & 406.781 & 334.42 & 10.89&0.0013 &0.170\\\\ B0820+02 & 1232.47 & 864.87 & 11.48&0.003 &0.231\\\\ \\noalign{\\smallskip} \\hline \\end{tabular} \\end{flushleft} \\end{table} ", "conclusions": "We have derived the relations between the surface magnetic field strengths and the final orbital periods of BMSPs. The observed data of BMSPs mostly fit well with the derived relations. This is compatible with the view that those systems are indeed the descendants of LMXBs or IMXBs. The existence of high-magnetic field pulsars with long orbital periods (PSR B0820+02 and J1803-2712) can naturally be explained that they have originated from long period LMXBs. In such long period systems the red giant donors have a deep convective envelope at the onset of the mass transfer, hence the mass transfer happens violently (super Eddington) and occurs in short timescale (van den Heuvel 1994) such that the neutron stars can only accrete a small fraction of the transferred mass. As a result, the pulsars only experience a mild decay of the magnetic field. We, however, fail to produce high-magnetic field pulsars with short orbital periods such as PSR B0655+64 and PSR B1831-00. We conclude that those systems cannot be formed from either ``standard'' or AIC scenario. Tauris et al. (2000) also reach similar conclusion that binary pulsars with high-mass CO/ONeMg white dwarf companions (PSR B0655+64 belongs to this category as it has a $\\sim 0.8 \\Msun$ companion) and short orbital periods cannot be formed from LMXBs or IMXBs. Then, the most plausible scenario for forming those systems is that they are formed from the common-envelope (CE) evolution as is originally suggested by van den Heuvel \\& Taam (1984) for PSR B0655+64. In this scenario, the initial mass of the companion star is considerably larger than the mass of the neutron star (say $5 \\Msun$) and the initial orbital period is very long ($\\sim 1000$ days). The companion will fill up its Roche lobe when it ascends the asymptotic giant branch (AGB), i.e. after helium exhaustion in its core. Mass transfer from such an evolved and massive companion tends to be unstable. Such a system undergoes a common envelope (CE) phase in which the neutron star spirals into the envelope of the primary and both stars are embedded in a common envelope. During the spiral-in of the neutron star, the envelope is blown off at the expense of the orbital energy (and probably also in combination with some other energy sources; i.e., accretion energy, recombination energy etc.). At the end of the CE phase, the system consists of a $\\sim 1 \\Msun$ CO white dwarf (initially the CO core of the primary) and a neutron star. As a large fraction of the orbital energy is used to expel the envelope, the orbital separation of the system is very small ($P_{\\rm orb} \\la 1 {\\rm d}$). The system resembles closely PSR B0655+64. Since the duration of the CE phase is short ($\\sim 10^3$ -- $10^5$ yr), the neutron star will not accrete much matter, so it will maintain its high magnetic field. However, Chevalier (1993) and Brown (1995) have suggested that a spiral-in neutron star may experience a hypercritical accretion to become a black hole. Despite this suggestion, we argue that the existence of short orbital period BMSPs with relatively strong dipole fields indicates that CE evolution is still an attractive way to form binary pulsars. The remaining problem is that, the small mass function of PSR B1831-00 may imply a small mass of the white dwarf ($\\sim 0.075 \\Msun$ if we assume $i = 60 \\degr$). This mass is much too small to be reconciled with the CE evolution. In this respect, the small mass function may indicate, instead, a small inclination angle. However, if we require $M_{\\rm wd} \\ga 0.6 \\Msun$ then $i \\la 7.5 \\degr$. Assuming a random orientation of the orbit, the probability of observing a binary with such a small inclination angle is only 0.8\\%. Another possibility is that the system may result from AIC of a 1+1.2 $\\Msun$ main-sequence and white dwarf system with unstable mass transfer (Li \\& Wang 1998). But the outcome is highly speculative. We conclude therefore that the history of PSR B1831-00 is indeed very special and it needs further studies. We have used a specific relation between magnetic field strength B and amount of matter accreted by the neutron star $\\Delta M$, as given in Eq. (1). This relation well reproduces the observed relation between B and orbital period of BMSPs originating from LMXBs and IMXBs. This is certainly a strong argument in favour of a relation of the type of Eq. (1), however it is not necessarily a confirmation for this type of relation. In the literature, there are at least two other relations have been suggested (see e.g. Shibazaki et al. 1989 and Urpin \\& Geppert, 1995). The field decay described by these relations are more slow or rapid compared to that by Eq. (1). But, evidently, they will give similar ``well fitting''for the vast majority of the BMSPs (Li \\& Wang, 1998) in which PSR B0655+64 and PSR B1831-00 will always remain the exceptions." }, "0005/astro-ph0005511_arXiv.txt": { "abstract": "We have determined the mass profiles of dark halos in 83 objects observed by ASCA. The point spread function of X-ray telescope was deconvoled by the Richardson-Lucy algorithm and the temperature profiles were calculated to obtain the mass profiles. The derived mass profiles are consistent with the NFW model in $ 0.01-1.0 r_{virial}$. We found a good correlation between the scale radius $ r_{s}$ and the characteristic mass density $ \\delta_{c} $ , which indicates the self-similarity of dark halos. The spectrum index of primordial density fluctuation, $ P(k) \\propto k^{n}$, was determined from the slope of $ r_{s} - \\delta_{c} $ relation. For $M_{200}=10^{12}-10^{15} M_{\\odot}$, our analysis gives $n=-1.2 \\pm 0.3$ with a confidence level of $90\\%$. The mass density of dark halos is a good indicator of the mean mass density of the universe at the time when the halos were assembled, $z=z_{f}$. Assuming $\\delta _{c} \\propto (1+z_{f} )^{3}$, we have determined the epoch when each dark halo was assembled. Our analysis indicates that the field elliptical galaxies and groups of galaxies formed approximately at $1+z_{f} \\simeq 15$ and at $1+z_{f} \\simeq 7-10$ respectively. ", "introduction": "Early on, the density fluctuation grows linearly as the universe expands, $d\\rho / \\rho \\propto (1+z)^{-1}$ . Once a density enhancement of a spherical region approaches $ d\\rho /\\rho \\sim 1$, the sphere begins to turn around from the Hubble flow and collapse rapidly to form a virialized halo. The over density of virialized halo reaches $178\\rho_{0}(1+z_{c})^{3}$, where $z_{c}$ is the collapse redshift and $\\rho_{0}$ is the critical mass density in the present \\citep{kai86}. As the universe expands, the virial radius expands gradually by steady mass accretion. N-body simulations show that the accreting dark matter forms an extended envelope, while the density profile of the original halo remains unchanged. The mass profile of a dark halo is rearranged only by a major merger through violent relaxation. A halo formed by a major merger at $z_{m}$ is characterized by the mass density of $178\\rho_{0}(1+z_{m})^{3}$, and it is preserved until the next major merger takes place \\citep{sal98, rai98, hen99}. By this means, the density profile of a dark halo can be a good indicator of the redshift when the last merger took place. \\citet{nav96} (hereafter NFW) suggested that all of the dark halos have the self-similar mass profile described as \\begin{equation} \\rho_{(r)} = \\frac{\\rho_{0} \\delta_{c}}{(r/r_{s}) (1+r/r_{s})^{2} }, \\end{equation} where $\\rho_{0}, r_{s}$ and $\\delta_{c}$ are the critical mass density of universe, the scale radius and the characteristic density respectively. In this case, the $\\delta_{c}$ would be a direct indicator of the formation epochs of dark halos. In this paper, we will present the mass profiles of dark halos over the mass range of $10^{12}-10^{15} M_{\\odot}$ using a sample of 83 objects observed with ASCA. This furnishes us with a possible way to determine the formation epochs of galaxies and clusters of galaxies. Throughout the paper, we assume $ \\Omega=1$, $\\Lambda =0$ and $H_{0}=75km s^{-1} Mpc^{-1}$. ", "conclusions": "" }, "0005/astro-ph0005505_arXiv.txt": { "abstract": "Recent measurements of young accreting binary neutron stars are determining more precise magnetic field and accretion parameters. A low magnetic field accreting, millisecond pulsar has finally been found in a binary burster. At least 20 low-mass binaries have exhibited high frequency oscillations, 300--1200 Hertz. The majority have, for some range of luminosity, the pair of quasiperiodic oscillations that have never been seen in a black hole candidate. Recent evidence from burst oscillations strengthens the case that the difference frequency of this pair is close to the spin frequency of the neutron star. These oscillations are correlated with the spectra, luminosity, and low frequency oscillations. Quasiperiodic oscillations are also seen sometimes in strong magnetic field pulsars, where their origin can be closely examined. ", "introduction": "It is now believed that we have galactic binary neutron stars with magnetic fields that are as low as $10^8$ Gauss and as high as $5 \\times 10^{14}$ Gauss. There are single neutron stars with magnetic fields as low, and probably single neutron stars with a field as high. Because of recent measurements it is now possible to plot the magnetic fields and rotation periods of representative members of the varied populations. These are displayed in Figure 1, along with the fields and periods of some related objects. The evolutions of systems from birth of the neutron stars to their place in this diagram are subjects in themselves. Only now are the X-ray measurements establishing the places of the objects with reasonable accuracy rather than as an order of magnitude hypothesis. There are many aspects of the accretion onto the binary neutron stars that are not yet understood. I will touch upon some of these. But observations by RXTE and BeppoSAX have established some facts that vindicate former years of assumptions, enable calculations to be done that depend on the parameters that were determined, and lead to a new round of questions. ", "conclusions": "The new results on accreting binary neutron stars found by RXTE and BeppoSAX provide new tools to gain physical insight both into the compact objects that are the last equilibrium point before collapse to a black hole and into the binary accretion flows onto them. In the case of the strong field accreting pulsars, cyclotron features allow direct measurement of their magnetic fields and the timing signatures of the flow then illuminate apparent irregularities in the disks. The conditions of their occurrence are not clear and they are far from being well studied. There are hints of vibration signatures that might come from the accretion mound above the polar cap and are possible evidence for photon bubbles, but they need confirmation and more study. Magnetic fields down near $10^8$ Gauss in low-mass X-ray binaries are confirmed by the discovery of one irrefutable example. The uncertainty of important parameters of the system leave it still ambiguous whether we should find more such systems. A burster itself, it could resolve many technical questions about the QPO in LMXB if it ever manifests them. Evidence continues to support the conclusion that bursters are fast rotating neutron stars with low fields. The burst frequency and the difference between two high frequencies in the persistent flux seem rather clearly related. But the phenomena are still complex in detail. The persistent flux QPO around 1 kHz and around 100 Hz are correlated and correlations appear very general, holding for atoll sources, Z sources, and possibly black hole candidates as well. Competitive models are still working out detailed implications for comparison. It is notable that they all involve effects within a few kilometers of the neutron star, where effects of General Relativity are involved." }, "0005/astro-ph0005219_arXiv.txt": { "abstract": "Results from BeppoSAX observations of Compton--thick Seyfert 2 galaxies are summarized and reviewed, and their general properties derived and discussed. In five out of the seven observed sources, the nucleus is directly visible at high X-ray energies, where the photons penetrate absorbers with column densities in the range 1.1--4.3$\\times10^{24}$ cm$^{-2}$ (in the other two sources, NGC~1068 and NGC~7674, the nucleus is instead totally obscured at all energies, implying even larger column densities). In most sources there is unambiguous evidence of a reflection component from optically thick, cold matter, while in two (or maybe four) cases there is also evidence of reflection from ionized matter. For the sources with a measured X--ray luminosity, a comparison with the infrared luminosity is made; while in two cases (the Circinus galaxy and NGC~4945) the IR emission appears to be dominated by starburst activity, in the other three sources (NGC~6240, Mrk~3 and TOL~0109-383) it is likely to be dominated by reprocessing of the UV and X--ray photons emitted by an AGN. ", "introduction": "Compton--thick Seyfert 2 galaxies are by definition those AGN in which the X--ray obscuring matter has a column density equal to or larger than the inverse Thomson cross section, i.e. $N_H \\geq \\sigma_T^{-1} = 1.5\\times10^{24}$ cm$^{-2}$. The Thomson cross section is equal to the photoelectric cross section at around 10 keV (assuming cosmic abundances), and this energy may be assumed as the boundary between photoelectric-- and Compton--dominated regimes. By chance, this is also the upper energy of the working band of many past X--ray satellites, which therefore could observe Compton--thick sources only in the photoelectric regime, where the X--ray emission is dominated by scattered components. BeppoSAX (Boella et al. 1997), thanks to the unprecedented sensitivity of its high energy collimated detector, the PDS (Frontera et al. 1997), has now extended the sensitive observing range well into the Compton--dominated regime. There are several reasons why Compton--thick sources deserve to be studied. Firstly, most AGN, in the local universe at least, are obscured by Compton--thick matter (Maiolino et al. 1998). Therefore, they are an important ingredient not only of the Cosmic X--ray Background, but also of the IR background, where most of the absorbed radiation is re--emitted (Fabian \\& Iwasawa 1999). Secondly, the heavy absorption means that spectral components, which would otherwise have been completely dominated by the nuclear emission, can be observed. In particular, in the $\\sim$1--10 keV band the emission is dominated by reflection from both cold and ionized matter of the nuclear radiation, and the geometrical and physical properties of the circumnuclear matter can then be studied (e.g. Matt, Brandt \\& Fabian 1996). In this paper we summarize and discuss the results from the BeppoSAX Core Program on bright Compton--thick Seyfert 2 galaxies, as well as sources observed in other programs which were found to be Compton--thick, and explore the consequences. \\begin{table*} \\caption{Exposure times and count rates for the BeppoSAX observations of Compton--thick sources.} \\begin{tabular}{||l|c|c|c|c|c||} \\hline & & & & & \\cr Source & Date~of~obs & Exp.~time~(MECS) & LECS~CR & MECS~CR & PDS~CR \\cr & & (ks) & (s$^{-1}$) & (s$^{-1}$) & (s$^{-1}$) \\cr & & & & & \\cr NGC~1068$^{1,2}$ & 1996-Dec/1998-Jan-11 & 101.6/37.3 & 0.110 & 0.096$^{a}$/0.073$^{b}$ & 0.21 \\cr Circinus Galaxy$^{2,3}$ & 1998-Mar-13 & 137.7 & 0.065 & 0.132$^{b}$ & 2.01 \\cr NGC~6240$^4$ & 1998-Aug-14 & 119.4 & 0.012 & 0.024$^{b}$ & 0.38 \\cr Mrk~3$^{5}$ & 1997-Apr-16 & 112.8 & 0.021 & 0.069$^{a}$ & 1.16 \\cr NGC~7674$^{6}$ & 1996-Nov-25 & 116.0 & 0.003 & 0.005$^{a}$ & 0.13 \\cr NGC~4945$^7$ & 1999-Jul-01 & 93.8 & 0.031 & 0.057$^{b}$ & 2.77 \\cr TOL~0109-383$^8$ & 1999-Jul-26 & 64.3 & 0.005 & 0.010$^{b}$ & 0.16 \\cr & & & & & \\cr \\hline \\end{tabular} ~\\par 1) Matt et al. 1997; 2) Guainazzi et al. 1999; 3) Matt et al. 1999; 4) Vignati et al. 1999; 5) Cappi et al. 1999; 6) Malaguti et al. 1998; 7) Guainazzi et al. 2000a; 8) Iwasawa et al. 2000;\\par $^{a}$3 MECS units; $^{b}$2 MECS units (see Sec.2) \\end{table*} $H_0$=50 km s$^{-1}$ Mpc$^{-1}$ is adopted throughout the paper. ", "conclusions": "Compton--thick Seyfert 2 galaxies are very likely the most common subclass of AGN in the local Universe (Maiolino et al. 1998), and possibly also at high redshifts (Fabian 1999). The hard X--ray band is certainly the best with which to study these sources, because part of the nuclear radiation can penetrate the obscuring matter, if the column density does not exceed a few times 10$^{24}$ cm$^{-2}$. This is the reason why BeppoSAX has permitted a great advance in this limited but important field. Unfortunately, even this instrument does not allow the exploration of hard X--rays beyond the local Universe, and the cosmological evolution of the column density and covering factor of the absorber, which are important in order to understand the growth of the black holes and its relation with the star formation rate (e.g. Fabian 1999), is still unknown. Moreover, only a small fraction of the extragalactic sky has been covered so far at these energies with sufficient sensitivity, which implies that many sources like NGC~4945 and Circinus are still awaiting discovery. To make significant progresses in this field, a large improvement in sensitivity (like that will be provided by Constellation--X\\footnote{\\sc http://constellation.gsfc.nasa.gov/}), and large area, deep surveys (like that provided by Swift\\footnote{\\sc http://swift.gsfc.nasa.gov/} and, even better, that proposed with the EXIST\\footnote{\\sc http://hea-www.harvard.edu/EXIST/EXIST.html} project) are needed." }, "0005/astro-ph0005443_arXiv.txt": { "abstract": "The amount and properties of high-redshift galactic and intergalactic (IG) dust are largely unknown, but could be investigated using multi-wavelength photometry of high-$z$ objects that have a known intrinsic spectrum. Observations of $\\gamma$-ray burst (GRB) afterglows appear to support the theoretical model of an adiabatic blast wave expanding into an external medium. In this model, the synchrotron peak flux is independent of frequency, providing a flat spectrum when observed over time, and therefore allowing straightforward measurement of the relative attenuation of afterglow flux in widely separated bands. Applying this method to dust extinction, we show that for a sample of afterglows which have been corrected by galactic extinction, comparison between the number counts of peak fluxes in $X$-ray versus optical can provide constraints on an intergalactic component of dust. A similar technique can probe the redshift-dependence of extinction in GRB-forming regions without requiring an assumed relation between extinction and reddening by the dust. Probing systematic changes in extinction with redshift -- particularly in IG and/or non-reddening dust -- is crucial to a proper interpretation of the Type Ia Supernova Hubble diagram and similar observations, and useful in understanding GRB progenitor environments. ", "introduction": "There is now substantial evidence that $\\gamma$-ray bursts (GRBs) originate at cosmological distances (e.g. Metzger et al. 1997) from very powerful explosions (e.g. Kulkarni et al. 1998, 1999). They have been detected at very high redshifts, making them useful probes of the universe out to early cosmological epochs. In addition, the $\\gamma$-ray emission is followed by delayed emission at longer wavelengths, from the $X$-ray to the radio band (Costa et al. 1997; van Paradijs et al. 1997; Frail et al. 1997). This afterglow is described reasonably well as synchrotron radiation, emitted when a relativistic shell collides with an external medium (Paczy\\'nski \\& Rhoads 1993; Katz 1994; Waxman 1997a,b; Wijers, Rees \\& M\\'esz\\'aros 1997; Sari, Piran \\& Narayan 1998). Afterglows, like their high energy counterpart, are also expected to be detected out to very high redshift. The absorption-line systems and the Ly$\\alpha$ forest visible in their spectra can therefore be used to trace the evolution of metallicity in the universe, and to constrain, or possibly measure, the epoch at which re-ionization of the universe occurred (e.g. Lamb \\& Reichart 2000). In this {\\em Letter} we point out other uses that multi-wavelength observations of afterglows might have in cosmology. Whereas it is by now clear that GRBs (and consequently their afterglows) are far from being standard candles, the afterglows do have a very interesting property: the flux at the peak of the synchrotron spectrum is independent of time for adiabatic hydrodynamic evolution (Katz 1994; M\\'esz\\'aros \\& Rees 1997; Sari, Piran \\& Narayan 1998), and the adiabatic shock model has received robust support from observations (Waxman 1997a,b). Because the frequency of the peak flux smoothly decreases with time, this provides a strong theoretical connection between observations at various widely-separated frequencies which is lacking in other high-$z$ objects such as quasars or galaxies. This property makes GRB afterglows well suited to absorption studies of both the GRB immediate environment and host galaxy, and of the intervening intergalactic medium (IGM). One such possibility is the study of intergalactic dust, which may play a role in observations of high-$z$ Supernovae (SNe). In the past few years, observations of Type Ia SNe by two separate groups have revealed a progressive dimming of SNe at high redshift with respect to the predictions of a matter-dominated universe or even an open universe with a zero cosmological constant (Riess et al. 1998; Perlmutter et al. 1998). This dimming has been interpreted as evidence for acceleration in the cosmic expansion, probably caused by a positive cosmological constant. Two important systematic effects which must be accounted for in drawing this conclusion are evolution in the supernova intrinsic brightness and dust obscuration; both groups have discussed these at length. However, their dust corrections, based on reddening, rely on the assumption that dust has everywhere the same characteristics as in the Milky Way, which might not be the case. As Aguirre (1999) has shown, a scenario in which galaxies expel a significant fraction of their metals in winds or by radiation pressure ejection of dust, and in which very small grains are selectively destroyed or retained by galaxies, can self-consistently provide a viable candidate for an intergalactic (IG) dust component which is more 'grey' (i.e. less reddening) than dust in the Milky Way. Dust reddening properties also vary significantly within and between galaxies. GRB afterglows are observed in widely-separated bands at which even grey dust will attenuate radiation very differently, and, as explained above, a known relation exist between the expected fluxes in those bands in the absence of extinction. This {\\em Letter} shows that, given a sample of GRB afterglows for which a reliable correction for the host galaxy absorption has been made, a comparison between the number count distribution of peak fluxes in various bands can sensitively test the presence of a cosmological distribution of grey dust. Whereas this test is insensitive to the choice of model parameters made, such as type of cosmology, shape of the GRB luminosity function, redshift evolution of the GRB rate, etc. (because these parameters influence in the {\\em same} way the number counts in all bands, and we are {\\em comparing} data in different bands), it does rely on the assumption of the constancy of the peak flux in the various bands. In reality, several effects can contribute to a departure from the simple, ideal behavior (see e.g. Meszaros, Rees \\& Wijers 1998), and we discuss how and to what extent these can be corrected for, so that our test would still remain possible. Besides probing the existence of IG dust, we show that the same type of multi-wavelength study of GRB afterglows can provide a useful probe of the evolution of metallicity (or dust-to-gas ratio) in the GRB-forming regions of galaxies. Independent probes of metallicity evolution (i.e. not related to GRB sites) can then yield information on the specific environments of GRBs in relation to galaxies, helping to elucidate the nature GRB progenitors. Conversely, assuming a GRB progenitor type, one could obtain information about dust in the environment of that type of object, at very high $z$. ", "conclusions": "One of the outstanding problems in observational cosmology is to understand the type and amount of dust existing at high-$z$, whether outside galaxies or within them. Traditional methods of estimating extinction require a known dust reddening curve and an accurate model for the intrinsic UV/optical spectrum of the observed object, both of which may be rather uncertain; very few existing techniques allow the measurement of absolute extinction. In this {\\em Letter}, we have proposed GRB afterglows as probes of dust in cosmology. Like QSOs and galaxies they can be observed to very high $z$. In addition, there is a well-based theoretical spectrum which is flat between $X$-ray and near infra-red frequencies. This allows very straightforward estimates of dust extinction {\\em independent of its reddening properties} (though the reddening would be observable and useful) in GRB environments or in the intergalactic medium; the former can help help determine the properties of GRB progenitor environments (including the dust properties), while the latter is crucial for careful interpretation of the Type Ia Supernova Hubble diagram. While relying on the assumption that afterglow evolution is well described by the simplest adiabatic blast model, with a large number of afterglows (such as {\\em Swift} will provide), the technique should be applicable even if deviations from the theory occur." }, "0005/astro-ph0005169_arXiv.txt": { "abstract": "We investigate the origin of the broad absorption features detected near 1-1.4 keV in several Narrow Line Seyfert 1 galaxies, by modelling the absorbing medium with various physical parameters, using the ionisation code PEGAS. The observed properties of the X-ray absorption features can be reproduced by taking into account the peculiar soft X-ray excess which is well fitted by a blackbody plus an underlying power law. We equally stress that the emission coming from the absorbing medium (related to the covering factor) has a strong influence on the resulting X-ray spectrum, in particular on the apparent position and depth of the absorption features. A non-solar iron abundance may be required to explain the observed deep absorption. We also investigate the influence of an additional collisional ionization process (``hybrid case'') on the predicted absorption features.\\\\ ", "introduction": "A systematic analysis of the X-ray spectral properties of a sample of 22 so-called Narrow Line Seyfert\\,1s (NLS1s) based on ASCA observations has shown evidence for a broad absorption feature centred in the energy range 1.1-1.4\\,keV, in 6 of the 22 objects \\cite{Vaughan99}. The absorption feature strength is typically about 100\\,eV with an intrinsic width ranging from 0.1 to 0.3\\,keV. Until now, this type of absorption has never been detected in Broad Line Seyfert\\,1s (BLS1s) spectra. Only 3 NLS1s exhibit absorption edges in the 0.7-0.9\\,keV range consistent with that seen in at least 50\\% of the BLS1s. Several explanations of the 1.1-1.4\\,keV absorption features have been proposed: a blueshift of \\ion{O}{vii}--\\ion{O}{viii} edges or lines (outflow: z$\\sim$0.2-0.6\\,c) \\cite{Leighly97,Ulrich99}; resonance absorption lines from Mg, Si, S and Fe\\,L \\cite{Nicastro99b,Turner99}; an enhancement of Fe or Ne \\cite{Leighly97,Ulrich99}. ", "conclusions": "Several properties of NLS1s eg. UV-X-ray energy distribution, iron overabundance, hybrid plasma with high covering factor, could account for the peculiarities observed in some soft X-ray spectra of these objects. The spectral resolution attainable by ASCA was insufficient to disentangle between these different possibilities. The new X-ray satellites (Chandra, XMM-Newton) offer the prospect of detailed spectra that will certainly allow us to determine the nature of the 1\\,keV feature (emission or absorption). Moreover X-ray spectroscopic diagnostics such as those based on the ratios of He-like ion lines will enable us to determine the ionizing process (either pure photoionization, or photoionization plus an additional ionization process), as well as the gas density \\cite{PorquetDubau2000b,Kaastra2000}. The determination of the physical parameters for the Warm Absorber media in NLS1s and in BLS1s will provide constraints on unified schemes.\\\\" }, "0005/astro-ph0005439_arXiv.txt": { "abstract": "$\\tau$ Air-showers are the best trace of rarest Ultra High Energy neutrinos UHE $\\nu_{\\tau}$, $ \\bar\\nu_{\\tau} $ and $ \\bar{\\nu}_e $ at PeV and higher energy. $\\tau$ Air-showers may generate billion times amplified signals by their secondaries . Horizontal amplified $\\tau$ air-showers by $\\nu_{\\tau} N$ and UHE $\\bar\\nu_{e} e$ at PeV emerging from mountain chain might be the most power-full imprint. Upward UHE $ \\nu_{\\tau} N $ interaction on Earth crust at horizontal edge and from below, their consequent UHE $\\tau$ air-showers beaming toward high mountains should flash $\\gamma$,$\\mu$,X and optical detectors on the top. Upward $\\tau$ air-shower may hit nearby satellite flashing them by short, hard, diluted $\\gamma-$burst at the edge of Gamma Ray Observatory BATSE threshold. We identify these events with recent (1994) discovered upward Terrestrial Gamma Flashes (TGF) and we probed their UHE $\\tau$ - UHE $\\nu_\\tau$ origin. From these TGF data approximated UHE $\\nu_\\tau$ flux and $\\Delta m_{\\nu_\\mu \\nu_\\tau}$ sever lower bound are derived. Partial TGF Galactic signature is also manifest within known 47 TGF events at ${ \\simeq 2\\cdot 10^{-3}}$probability. Well known $X-\\gamma-\\mathrm{TeV}$ active galactic and extragalactic sources have found probable counterpart in TGF arrival directions. Detection of elusive UHE $\\nu_\\tau$ seem finally achieved. ", "introduction": "_e$, $\\nu_\\tau$, $\\bar{\\nu}_\\tau$ interactions and their UHE $\\tau$ secondary} Ultra high energy astrophysical neutrino (UHE$\\nu$) from PeVs ($\\gtrsim 10^{15}$ eV) up to ($10^{18}$ eV) EeV and GZK cut off energies ($ \\gtrsim 10^{19}$ eV) might be traced by $\\tau$ induced air showers and by their millions to hundred billions multiplicity in secondaries particles, ($N_{\\mu} \\sim 10^6 \\left( \\frac{E_{\\nu}}{PeV} \\right)$, $N_{X} \\sim 10^{10} \\left( \\frac{E_{\\nu}}{PeV} \\right)$, $N_{opt} \\gtrsim 10^{11} \\left( \\frac{E_{\\nu}}{PeV} \\right)$).\\\\ Indeed astrophysical PeVs UHE anti-neutrino electrons, $\\bar{\\nu}_e$, near the Glashow W resonance peak, $E_{\\bar{\\nu_e}} = M^2_W / 2m_e \\simeq 6.3 \\cdot 10^{15}\\, eV$, (dominant over expected UHE PeV atmospheric neutrino signals), may be observable by their secondary horizontal $\\tau$ air showers originated by UHE chain reaction $\\bar{\\nu}_e + e \\rightarrow W^- \\rightarrow \\bar{\\nu}_{\\tau} + \\tau^-$ inside the concrete rock of a high mountain. Also UHE $\\nu_\\tau$, $\\bar\\nu_{\\tau} $ at ($10^{16}$ - $10^{17}$eV) interacting with nuclear matter ($\\nu_\\tau$ $N$) must be observable if flavor mixing $\\nu_{\\mu}\\leftrightarrow \\nu_{\\tau}$ take place as shown by Superkamiokande data, because of huge astrophysical distances respect to oscillation ones.\\\\ Therefore UHE $\\nu_{\\tau}$ and $\\bar{\\nu_{\\tau}}$ may be converted and they may reach us from high energy galactic sources, as pulsars, SNRs, black holes or galactic SGRs microquasars, as well as from powerful extragalactic AGNs, quasars or GRBs, even for any small mass mixing, ($\\Delta m_{ij}^2 \\sim 10^{-4}$ eV $^2$) or any high (GZK) energy because of the large galactic (Kpcs) and extreme cosmic (Mpcs) distances: \\begin{equation} L_{\\nu_{\\mu} - \\nu_{\\tau}} = 4 \\cdot 10^{-3} \\,pc \\left( \\frac{E_{\\nu}}{10^{16}\\,eV} \\right) \\cdot \\left( \\frac{\\Delta m_{ij}^2 }{(10^{-2} \\,eV)^2} \\right)^{-1} \\end{equation} Rare upward UHE $\\tau$, born by $\\nu_{\\tau}$ and $\\bar{\\nu_{\\tau}}$ nuclear (or rare $\\bar{\\nu_e} - e$ interactions near the upward earth surface), may escape outside on air where they may spontaneously decay triggering upward vertical, oblique or near horizontal $\\tau$ air showers. The vertical ones (by small nadir angle) occur preferentially at low energies nearly transparent to the Earth ($E_{\\nu} \\sim 10^{15} - 10^{16} $ eV). The oblique $\\tau$ air showers whose arrival directions have large nadir angle, are related mainly to higher energy $\\nu_{\\tau}$, or $\\bar{\\nu_{\\tau}}$ nuclear interactions ($E_{\\bar{\\nu_{\\tau}}} \\geq 10^{17} - 10^{20}$ eV). Indeed these horizontal - upward UHE $\\nu_{\\tau}$ cross a smaller fraction of the Earth volume and consequently they suffer less absorption toward the horizon. \\\\ These huge horizontal or upward air-shower signals being at least million to billion times more abundant than the original and unique UHE $\\tau$ or UHE $\\mu$ track in underground Km cube detectors are much easier to be discovered with no ambiguity. We remind that long tracks in km$^3$ detectors are mostly noisy signals by TeVs to tens of TeVs muons secondaries generated by atmospheric neutrinos born by common cosmic ray interactions in upper atmosphere. The $\\tau$ air shower is analogous to the Learned and Pakwasa $(1995)$ \"double bang\" in underground neutrino detectors. The novelty of the present \"one bang in\" (the rock) - \"one bang out\" (the air) lays in the self-triggered explosive nature of $\\tau$ decay in flight and its consequent huge amplified air shower signal at a characteristic few Kms distance. Detectable horizontal gamma bursts (mainly bremsstrahlung photons) are among the most abundant signal. The source of UHE $\\nu_{\\tau}$, since Super Kamiokande evidence of neutrino flavour mixing, must be as abundant as muon $ \\nu_{\\mu}$ ones. \\\\ Moreover the expected $ \\nu_{\\tau} $ signals, by their secondary tau tracks at highest cosmic ray energy window $1.7\\cdot 10^{21} \\,eV > E_{\\tau} > 1.6\\cdot 10^{17} \\, eV $, must exceed the corresponding $ \\nu_{\\mu} $ (or muonic) ones, making UHE $ \\nu_{\\tau} $ above $0.1$ EeV the most probable UHE signal. Indeed, the Lorentz-boosted tau range length grows (linearly) above muon range, for $ E_{\\tau} \\geq 1.6 \\cdot 10^8 GeV $; (see Fig (1) eq.3): the tau track reaches its maxima extension, bounded not by pair production (eq. 2), but by growing nuclear electro-weak interactions (eq. 4), $ R_{\\tau_{\\max}} \\simeq 191\\;Km$, at energy $ E_{\\tau} \\simeq 3.8\\cdot 10^9\\;GeV$ . \\begin{equation}\\label{6} R_{R_{\\tau}} \\cong 1033 \\; Km \\left(\\frac{\\rho_r}{5}\\right)^{-1} \\cdot \\left\\{\\, 1 \\,+\\, \\frac{\\ln\\left[\\left(\\frac{E_{\\tau}}{10^8 \\, \\mathrm{GeV}}\\right)\\left(\\frac{E_{\\tau}^{\\min}}{10^4 \\, \\mathrm{GeV}}\\right)^{-1}\\right]}{(\\ln \\, 10^4 )}\\right\\} . \\end{equation} \\begin{equation}\\label{7} R_{\\tau_o} = c \\tau_{\\tau} \\gamma_{\\tau} = 5 \\, \\mathrm{Km} \\, \\left(\\frac{E_{\\tau}}{10^8 \\, \\mathrm{GeV}}\\right) \\; . \\end{equation} \\begin{equation}\\label{9} R_{W_{\\tau}} = \\frac{1}{\\sigma N_A \\rho_r} \\simeq \\frac{2.6\\cdot 10^3 \\, \\mathrm{Km}}{\\rho_r} \\, \\left(\\frac{E_{\\tau}}{10^8\\, \\mathrm{GeV}}\\right)^{-0.363} \\; . \\end{equation} At this peak the tau range is nearly $20$ times longer than the corresponding muon range (at the same energy) implying, for comparable fluxes, a ratio 20 times larger in $ \\nu_{\\tau} $ over $ \\nu_{\\mu} $ detection probability. This dominance, may lead to a few rare spectacular event a year (if flavor mixing occurs) preferentially in horizontal plane in underground $Km^3$ detectors. The Earth opacity at those UHE regimes at large nadir angles (nearly horizontal, few degree upward direction) is exponentially different for UHE muons respect to tau at GZK energies (corresponding to 500 Kms UHE Tau lenghts), making the muon/tau flux ratio of such lenghts severely (half billion time) suppressed. UHE $\\nu$ above GZK are transparent to BBR cosmic photons and they may easily reach us from far cosmological distances. Therefore the puzzle of UHECR above GZK cut off may be solved assuming that neutrinos (possibly of heaviest Muon/Tau nature)share a light mass of few eV , in the frame-work of Hot Dark Matter halos clustered around galaxies. Such light neutrinos may form a huge hidden dark calorimeter able to beam dump UHE $\\nu$ via $Z$ (s-channel), via virtual $W$ ($t$ channel) or W pair productions. The corresponding cross sections for such $\\nu$ $\\nu$ interactions are shown in Fig. 2; their secondaries may be final UHE anti-protons (or anti-neutrons) or UHE protons (or neutrons) (Fargion,Mele,Salis 1997-1999) responsible of final observed UHECR above GZK cut off. The interaction efficency by relic light neutrinos via UHE $\\nu$ at GZK cut off is thousands times larger than UHE $\\nu$ interactions on Earth atmosphere and/or direct UHECR (nucleons,nuclei) propagations above GZK distances. Therefore light neutrino mass may explain both hot dark matter and UHECR above GZK (as well as their recent clustering in triplets or doublets). Just to emphasize the $\\nu$ mass roles in high energy astrophysics, we remind the important case of a SN MeVs neutrino burst arriving slowed by its mass relativistic flight and its delayed arrival from far SN (galactic or better extragalactic) events respect to the massless (prompt coeval) gravitational waves. The expected time delay between the massless graviton wave burst (by supernova quadrupole emission at distance L and the $\\nu_e$ neutronization neutrino burst), will be an additional tests test to the elusive mass detection: $ \\Delta t \\sim 50$ sec $ \\ffrac{E_{\\nu}}{5\\,MeV}^{-2} \\ffrac{m_{\\nu}}{5\\,eV}^{2} \\ffrac{L}{Mpc}$. (Fargion 1981). Let us remind that massive neutrino imply new right handed interactions in early Universe (Antonelli, R.Konoplich, D.Fargion 1981) and multifluid gravitational clustering during galaxy formation epochs (D.Fargion 1981, 1983). UHE Tau $ E_{\\tau} \\geq \\ 10^5 GeV - 5 \\cdot 10^7 GeV $ air-shower in front of high mountains chains will be easily induce peculiar horizontal UHE $\\tau$ decay beyond a thick mountain (Fargion, Aiello, Conversano 1999). The high mountain act as a clever filter: \\\\a) as a wide angle screen of undesirable horizontal Ultra High Energy Cosmic Rays (UHECR) (electro-magnetic shower, secondary Cherenkov photons and muons),b) as a calorimeter for UHE $\\nu_\\tau$,$ \\bar\\nu_{\\tau} $ and $ \\bar{\\nu}_e $, c) as a distance meter correlating tau relativistic track and birth and its air-shower opening distance from the mountain with UHE tau original energy. d) as an unique source, by tau electromagnetic showering, of horizontal rich, sharp, $\\mu$second burst $\\gamma$, X flash, electron pairs and Cherenkov showers source. e) as an unique source, by tau hadronic showering, of additional horizontal dense muon pairs sharp bundle burst. \\\\ An hybrid detector (gamma/optical)would get precise signal and arrival direction. Because of the different neutrino interactions with energy and flavors it will be possible to estimate, by stereoscopic, directional and time structure signature, the spatial air-shower origination in air, the primary tau distance decay from the mountain (tens of meter for PeVs UHE $ \\bar{\\nu}_e $ and nearly hundred meters up to Kms for UHE $\\nu_\\tau$,$ \\bar\\nu_{\\tau} $) $ E_{\\tau} \\geq \\ 10^5 GeV , 5 \\cdot 10^7 GeV $ , the consequent most probable original UHE tau range and energy. Additional energy calibration may be derived sampling shower intensities.\\\\ Hundreds of detectors in deep wide valley would be necessary to get tens taus of event a year. \\\\ Screening by undesirable lateral or downward noisy cosmic rays or natural radiation is possible by directional and time clustering filter; therefore the induced $\\bar{\\nu}_e e \\rightarrow \\tau$ air shower even in absence of $\\nu_{\\mu} \\leftrightarrow \\nu_{\\tau}$ oscillation should be identified and detectable soon. Its unique $\\bar{\\nu}_e$ origin is marked by the peaked W resonance, and by the small mountain $\\bar{\\nu}_e$ opacity and its high neutrino cross-section. Its identity is marked by the expected fine tuned PeV energy at W peak and the tau air-shower birth place near (a hundred meter) the mountain wall.\\\\ More copious ($> 5$ times more) events by PeV up to tens PeV $\\;$ $\\nu_{\\tau} N$ interaction occur if, as most of us believe, $\\nu_{\\mu}$ oscillate in $\\nu_{\\tau}$.\\\\ It will be also possible to discover UHE $\\tau$, by observing the upward tau air-shower arriving from hundred Kilometers away (near horizontal edges) from high mountains, high balloon and satellites; such UHE tau created within a wide (tens thousands to millions square km$^2$ wide and hundred meter UHE Tau depth in Earth crust) target would discover easely UHE $\\nu_\\tau$,$ \\bar\\nu_{\\tau} $ neutrinos at PeV up to EeV energies and above, just within the mysterious GZK frontiers. The discover will need capable gamma, optical and mainly muon bundle detectors within present technology as studied elsewhere.\\\\ From the same highest mountains, balloons and near orbit satellite, looking more downward toward the Earth it is possible to discover more frequent but lower energetic astrophysical $\\simeq$ PeV - tens PeV neutrinos still nearly transparent to the Earth volumes and (Gandhi et al. 1998),(see Fig.3). \\\\ The UHE neutrinos $\\bar{\\nu_e}$,${\\nu}_{\\mu}$ $\\bar{\\nu}_{\\mu}$ are default and expected UHECR ( $\\gtrsim 10^{16}$ eV) secondary products near AGN or microquasars by common photo-pion decay relics by optical photons nearby the source (PSRs, AGNs) ($p + \\gamma \\rightarrow n + \\pi^+, \\pi^+ \\rightarrow \\mu^+ \\nu_{\\mu}, \\mu^+ \\rightarrow e^+ \\nu_e \\bar{\\nu}_{\\mu} $), or by proton proton scattering in galactic interstellar matter. The maximal observational distances from mountains, baloons or satellites, may reach $\\sim$ 110 Km $(h/Km)^{\\frac{1}{2}}$ toward the horizon, corresponding to a UHE $\\tau$ energy $\\sim 2 \\cdot 10^{18}$ eV $(h/ Km)^\\frac{1}{2}$. Therefore we propose to consider such upward shower nearly horizontal detection from high mountains to test this highest $\\nu_{\\tau} \\bar{\\nu_{\\tau}}$ energy window almost opaque to Glashow UHE $\\bar{\\nu_e}$ fluxes. ", "conclusions": "" }, "0005/astro-ph0005113_arXiv.txt": { "abstract": "We report on detailed abundances of giants in the Galactic bulge, measured with the HIRES echelle spectrograph on the 10-m Keck telescope. We also review other work on the bulge field population and globular clusters using Keck/HIRES. Our new spectra have 3 times the resolution and higher S/N than previous spectra obtained with 4m telescopes. We are able to derive $\\log g$ from Fe II lines and excitation temperature from Fe I lines, and do not rely on photometric estimates for these parameters. We confirm that the iron abundance range extends from $-1.6$ to $+0.55$ dex. The improved resolution and S/N of the Keck spectra give [Fe/H] typically 0.1 to 0.2 dex higher than previous studies,\\cite{mr94} for bulge stars more metal rich than the Sun. Alpha elements are enhanced even for stars at the Solar metallicity (as is the case for bulge globular clusters). We confirm our earlier abundance analysis of bulge giants\\cite{mr94} and find that Mg and Ti are enhanced relative to Ca and Si even up to [Fe/H]=+0.55. We also report the first reliable estimates of the bulge oxygen abundance. Our element ratios confirm that bulge giants have a clearly identifiable chemical signature, and suggest a rapid formation timescale for the bulge. ", "introduction": "Because of their faintness, reddening, severe crowding, and high metallicity, the stars of the Galactic bulge remained among the last Galactic population to be studied with high resolution spectroscopy. In the scientific cases for large telescopes, the goal of successfully defining the abundances and chemistry of bulge stars has often figured prominently. Of course, the real driver for studying these stars is not the technical challenge, rather it is their potential to yield insights into the formation of bulges and ellipticals. Within the last five years, the combination of spectroscopy with the Keck telescopes and imaging with the Hubble Space Telescope has revolutionized the study of galaxies at high redshift. A population of plausible progenitors \\cite{steidel96} to present-day $L^*$ galaxies has been discovered at $z>3$ and a proposed star formation history \\cite{madau96} of the Universe has been sketched out. However, these observations cannot trace the evolution of the $z>3$ galaxies into their present-day counterparts. In many respects, such as luminosity and clustering, they strongly resemble the progenitors of present-day luminous galaxies. It is also possible to constrain the formation time of bulges from observations of galaxies at $z\\leq 1$. Recent pixel-by-pixel analysis \\cite{ellis2000} of resolved images of high redshift galaxies with clearly visible bulges apparently shows that at any given redshift, bulges are bluer than the reddest galaxies of elliptical morphology. Unfortunately, this imagery cannot easily distinguish between a late starburst on top of an old population versus a mostly intermediate-age population. So it is valuable to seek other available evidence, such as the ages and abundances of stars in the Galactic bulge. The exact agreement between HST luminosity functions of old metal rich globular clusters, and NTT luminosity functions of the Galactic bulge field\\cite{sergio95} strongly suggests that the bulge formed early and rapidly. HST photometry in a number of different bulge fields also shows that the stars brighter than the oldest turnoff point are foreground stars associated with the disk, not the bulge\\cite{feltz00}. Age constraints from luminosity functions or the luminosity of the main sequence turnoff point, while powerful, are only accurate to (at best) $\\approx 1-2 Gyr$. The detailed composition of stars in the bulge does not constrain the absolute age of the bulge. However, it does constrain the timescale for chemical enrichment, and it helps to relate the bulge (or not) to elliptical galaxies. As a larger sample of stars is accumulated, more detailed theoretical inferences about the enrichment history will be possible. The bulge of the Milky Way is a clearly distinct population, as defined by the classical characteristics of a stellar population, age, abundance, kinematics, and structure. The central 1000 pc of our galaxy is dominated by old, metal rich \\cite{rich88}$^,$\\cite{mr94} stars with very high phase space density. The stellar mass of the bulge is $2\\times 10^{10}M_\\odot$, roughly 1/3 that of the disk, but it still accounts for a large fraction of the baryonic mass of the Galaxy. The image of the bulge\\cite{hauser90} obtained using the DIRBE instrument on board the COBE satellite dramatically illustrates its distinct nature and its similarity to more distant ellipticals. It is possible to develop a model\\cite{zhao96} that both fits the surface brightness in the COBE image, solves Poisson's equation, and gives stellar orbits that reproduce the observed kinematics of the bulge. Presently, there is no clear consensus on the ages and formation timescales of bulges in general. The colors of bulges imaged in detail in the optical and IR by HST are consistent with very large ages\\cite{pel99}, a result first found in 1969 for the bulge of M31\\cite{sandage69}. On the other hand, the integrated Mg line strengths of bulges are less than those of ellipticals at the same iron line strengths\\cite{proc00}, which would argue that bulges might have experienced a less intense and more extended period of star formation than the ellipticals. \\subsection{How Element Ratios May Constrain the Formation of the Bulge} The motivation for measuring abundance ratios in old stars is that they preserve the fossil record of the early star formation process. Potentially, the initial mass function, star formation rate, and importance of infall or extended star formation at late times can all be recovered from abundance ratios. The material treated briefly below is discussed in more detail elsewhere. \\cite{andy97}$^,$\\cite{mr99} Scenarios for forming the bulge predict a wide range of timescales, from $\\sim 10^8$ yr for a violent starburst, to a few Gyr for a massive disk that thickens into a bar. The modeling of observed abundance trends can distinguish among these models. \\begin{figure}[ht] \\begin{center}\\leavevmode \\psfig{figure={fig1.eps},width=100mm,angle=0} \\vspace*{-5.0cm} \\caption{A schematic plot of $\\rm [\\alpha/Fe]$ vs [Fe/H] in disk/halo stars. A shallower IMF slope (increasing number of massive stars) will enhance the alpha elements. A higher star formation (faster enrichment) will keep the alpha elements high, even above the Solar iron abundance.} \\label{fig:sfr} \\end{center} \\end{figure} {\\it Metallicity:} The fundamental notion of chemical evolution is that other than those light elements produced in the Big Bang, metals are made in supernovae. Because SNe explode in $\\sim 10^6$ yr and distribute their metals widely, it is possible to model the process as a simple differential equation (the Simple Model\\cite{ss72} of chemical evolution). In the case of the bulge, the deep potential well and likely violence of the early starburst satisfy the model assumptions, and the abundance distribution fits the Simple Model\\cite{rich90}. The yield is the ratio of the mass of metals produced to the total mass locked up in long-lived stars. In the Simple Model, the yield is the mean metal abundance of the population. The shallower the initial mass function slope (more massive stars) the higher the yield. {\\it Alpha Elements:} When the first 200-inch echelle spectra of metal poor stars and globular clusters were obtained\\cite{wall62} it was noted that some even-Z elements (O, Mg, Si, Ca, and Ti) were overabundant by $\\sim +0.3$ dex relative to the Solar Neighborhood. These are the so-called $\\alpha$-elements, although their actual synthesis is far more complicated than transmutation by successive capture of helium nuclei in massive stars. The widely accepted explanation for these over abundances\\cite{tinsley79}$^,$\\cite{wheeler89}is that massive star (Type II supernovae) dominated the enrichment at early times; models\\cite{ww95} indicate that the ejecta of these SNe are very rich in alpha elements. Although type I SNe produce the iron peak elements, their contribution to the iron abundance becomes important only after $\\sim 1$Gyr, as time is required for the formation of a prior generation of white dwarfs. The diagnostic value of trends of $\\rm [\\alpha/Fe]$ vs [Fe/H] extend beyond their use as a crude clock, as has been suggested for the bulge\\cite{matt90}. If the IMF is dominated by massive stars, the alpha elements can be enhanced by more than +0.3 dex\\cite{ww95}, while a high star formation rate will result in stars of Solar iron abundance having an alpha-enhanced composition, as appears to be the case for the bulge. Finally, although Ti is observed to be elevated with the alpha elements, the nucleosynthesis calculations\\cite{ww95} predict a low yield of Ti in massive stars; this remains a problem. {\\it Neutron-Capture Elements:} The two dominant modes of neutron capture also offer the potential to serve as clocks, and as a fossil record of early star formation. Supernovae (probably Type II) are believed to be the site of the r-process\\cite{wh92}, while the helium burning shells of AGB stars are suspected as the site of s-process production, as was shown in early calculations\\cite{iben75}. The Ba/Eu ratio is is especially useful because it is sensitive to the r-process fraction of heavy elements. However, practical use of this diagnostic in the bulge is somewhat complicated by the lack of weak Ba lines, although La and Nd offer excellent possibilities as s-process indicators. Depending on whether [Ba/Eu] as a function of [Fe/H] approaches the s-process or the lower r-process value, one can infer either a disk-like or halo-like (Type II SN ejecta dominated composition) star formation history.\\cite{andy98} In principle, evidence for r-process nucleosynthesis indicates the presence of enrichment due to Type II SNe which could be due either to a rapid burst of star formation or a shallow IMF. In the bulge, we hope to use [Ba/Eu] and other heavy element diagnostics to test the hypothesis\\cite{wyse92} that the bulge formed from gas initially enriched by the astration of the halo. The heavy elements have tremendous potential to constrain the enrichment timescale (and stellar masses responsible) in great detail, especially in the difficult 1-5 Gyr regime\\cite{busso99}. The production of stable isotopes of some s-process elements such as Rb turns out to be very sensitive to the temperature of the helium burning shells of AGB stars. The derivation of abundances from the equivalent widths of the lines of heavy elements is done with caution. Each absorption line is split into multiple sub-components by nuclear hyperfine splitting; failure to account for this effect can lead to serious errors in the abundances. Before turning to a discussion of our results, we point out that our program would not have been possible without the HIRES echelle spectrograph \\cite{vogt94} as well as the Keck telescopes. Just now, in the year 2000, we are seeing the successful first light of UVES at the VLT, and HDS at Subaru. HIRES paved the way for these successful instruments, at a time when the operational success of such an instrument on a 10m telescope was far from guaranteed. ", "conclusions": "In contrast to the well known achievements in the high redshift universe, the impact of Keck on stellar abundances is less widely known, yet significant. Keck/HIRES spectroscopy has placed the abundance scale of the bulge on a secure footing. We have just begun to tap the potential information in these spectra. Prior efforts at measuring the oxygen abundance in the bulge from data obtained on 4m class telescopes were ineffective. For the first time, we are beginning to see emerging some clear trends in oxygen as a function of iron abundance. The abundance range, and puzzling element trends found by McWilliam \\& Rich (1994)\\cite{mr94} are confirmed. Two metal rich globular clusters toward the bulge have also been the subject of a major campaign with HIRES.\\cite{cohen99}$^,$\\cite{carretta00} NGC 6553 and 6528 have been found to have Solar metallicity with the alpha elements of O and Ca enhanced. The compositions of their stars are precisely those of bulge field giants at the same metallicity. The formation of the proto-bulge probably proceeded much as is observed in starburst galaxies today, with the production of numerous star clusters, a few of the more luminous of which are observed to survive to the present day. As spectroscopy of fainter stars becomes feasible, enrichment trends are now available for new stellar populations, such as dwarf spheroidal galaxies. As more high resolution spectra from large telescopes are analyzed, these trends may become valuable in distinguishing the formation histories of stellar populations. The Sagittarius dwarf spheroidal galaxy (a tidally disrupted dwarf galaxy lying in the direction of the bulge) is the only dwarf companion of the Milky Way that contains stars as metal rich as the Sun. One might speculate that the bulge could have been built from the shards of a few such disrupted systems, and the presence of Solar metallicity stars in the Sgr dwarf strengthens this idea. However, the Galactic bulge and disk populations are dramatically different from the Sgr dwarf stars, which have subsolar Ca and Si abundances\\cite{smecker99} at [Fe/H]=0. The trends of Mn with [Fe/H] and [Ba/Y] with [Fe/H] are even more different between the bulge and the Sag dwarf\\cite{msh00}, and it is possible to explain these differences as being caused be early, rapid enrichment in the bulge. The origin of the metal rich population in the Sgr dwarf is an interesting problem in chemical evolution, given the low mass of that galaxy and its encounter with the Milky Way. We can pretty much rule out, however, that the metal rich population in the Sgr dwarf was somehow captured from the bulge, or that Sgr was once a much larger galaxy that enriched as quickly as the bulge did. One may also compare the bulge composition to metal rich dwarf stars in the Solar neighborhood, which are $\\approx 10$ Gyr old and reach the same high metallicities ([Fe/H]=+0.55). High resolution spectroscopy\\cite{castro97}$^,$\\cite{feltz98} of these stars shows them to clearly have disk-like compositions: Mg, Ti, and O abundances are at approximately Solar values with no clear trends. In contrast, the old open cluster NGC 6791 has [Fe/H]=+0.4 and enhanced Ca\\cite{peterson98}. Chemical enrichment reaching high iron abundance evidently does not proceed the same way in all environments. Based on the compositions of stars, one clearly cannot produce the bulge out of the disintegrated remnants of systems like the Sagittarius dwarf spheroidal. Qualitatively, the abundance pattern in the bulge strongly suggests rapid, early enrichment, consistent with the predictions of chemical evolution models\\cite{matt99}. The notion of rapid enrichment agrees with other studies of the age of the stellar population\\cite{sergio95} in the bulge. The distinct nature of the bulge composition gives us confidence that abundance ratios offer a powerful diagnostic tool that may help to decipher the fossil record of galaxy formation. Many open questions remain. The bulge has a bar-like morphology, and the most successful scenario\\cite{merritt94} for forming a bar-like bulge requires dynamical instabilities occurring in a pre-existing disk. However, N-body simulations of bars indicate that they are unlikely to survive for a Hubble time, yet the Galactic bulge is extremely old.\\cite{sergio95} Further, the extreme stellar density near the nucleus is evidence for strong dissipation being a factor in the formation of the Galactic bulge. If the bulge abundance ratios favor a top-heavy IMF and very rapid formation, one must infer that ellipticals enrich more rapidly (and perhaps with a heavier IMF) because their $\\rm Mg_2$ indices at a give $<\\rm Fe>$ line strength are so much higher compared to the bulges; in fact spiral bulges lie near the lower range in Mg index in these diagrams \\cite{worthey92}$^,$\\cite{proc00} Before addressing these questions, and the challenge of relating the local data to high redshift observations, we plan to increase our sample size and explore the behavior of different atomic species. However, the study of the Milky Way bulge stars (and eventually, perhaps, individual stars in the bulge of M31) does have promise in illuminating the chemical evolution of ellipticals. \\subsection{Looking Towards the Future} This year, two new powerful high dispersion spectrographs come on line. At the VLT UVES has already passed science verification and has produced beautiful data. The HDS spectrograph at Subaru is just about to see first light. Fiber feeds to UVES will enable the acquisition of as many as 8 stars in a single exposure covering all orders, or spectroscopy of over 100 stars in a single echelle order. The latter capability will be enjoyed by the new echelle spectrograph that will be commissioned next year on the Magellan I (Baade) telescope. On Keck, the NIRSPEC infrared spectrograph can reach $R=30,000$ in the near-IR, and places old giants in the Galactic center within reach. We plan to extend our abundance studies to the field and cluster stars of the Galactic center in the next few years. The hard reality remains that analysis of the data will still be time consuming. For metal rich stars, it is clear that even at $R=60,000$ we require a full spectrum synthesis before we can feel completely secure in our results. It will be a challenge to keep up with the flood of new data in the coming years. This situation should be an inspiration to observers and theorists alike, as we enter these unprecedented times." }, "0005/astro-ph0005325_arXiv.txt": { "abstract": "We present a method for constraining the evolution of the galaxy luminosity-velocity (LV) relation in hierarchical scenarios of structure formation. The comoving number density of dark-matter halos with circular velocity of $200\\kms$ is predicted in favored CDM cosmologies to be nearly constant over the redshift range $0\\lsim z\\lsim 5$. Any observed evolution in the density of bright galaxies implies in turn a corresponding evolution in the LV relation. We consider several possible forms of evolution for the zero-point of the LV relation and predict the corresponding evolution in galaxy number density. The Hubble Deep Field suggests a large deficit of bright ($M_V<-19$) galaxies at $1.4\\lsim z\\lsim 2$. If taken at face value, this implies a dimming of the LV zero-point by roughly 2 magnitudes. Deep, wide-field, near-IR selected surveys will provide more secure measurements to compare with our predictions. ", "introduction": "\\label{sec:intro} Within the framework of hierarchical structure formation, galaxies form within virialized dark-matter (DM) halos. The halo properties and their time evolution can be predicted robustly within a given Cold Dark Matter (CDM) model using N-body simulations or analytic approximations. A fundamental property of a DM halo is its maximum circular velocity, which is likely to be related to the observable internal velocity of the luminous galaxy that resides in the halo. In the local universe, it is well known that galaxies obey fundamental scaling relations, including the relation between total luminosity and internal velocity (commonly known as Tully-Fisher or Faber-Jackson relations), which we will focus on here. This LV relationship surely reflects some fundamental aspect of galaxy formation, although its origin is poorly understood. An observational determination of the \\emph{redshift evolution} of this relationship would provide crucial clues as to its physical origin and would constitute an important constraint on models of galaxy formation. However, a direct observational determination of the internal velocities of galaxies at high redshift is extremely difficult. Rotation curves have been obtained for a small number of galaxies up to redshift $z\\sim1$ (Vogt et al. 1996, 1997, 2000). These results suggest that there is little evolution in the Tully-Fisher relation for objects similar to local large spirals out to $z\\sim1$ (but see Mallen-Ornelas et al. 1999). At higher redshifts ($z\\ga 1$), no direct observations are yet available. While galaxies of different morphological types obey different LV relations, one can define a mean LV relation for any given morphological mix. This can be used to make a direct connection between the distribution functions of galaxies in terms of their \\emph{internal velocity} and \\emph{luminosity}. Clearly, if the evolution of the velocity function is predicted by theory, and the evolution of the luminosity function is determined observationally, this connection can provide a constraint on the evolution of the LV relation. Determining the number density of galaxies as a function of their rest-frame luminosity is considerably easier than measuring their rotation velocities directly at high redshift. Observed luminosities can be k-corrected to the rest visual frame up to $z\\sim2$ using multi-band photometry extending into the near-IR. It has been very difficult to obtain spectroscopic redshifts in the `desert' regime of $1.4 \\lsim z \\lsim 2$, but photometric redshifts offer a means of filling in this gap. These advances, combined with theoretical progress in predicting the number density and structural properties of dark matter halos, suggest the approach for constraining the redshift evolution of the LV relation that we exploit here. The basis of our analysis is the evolution of the velocity function of DM halos as determined via a high-resolution N-body simulation of the \\lcdm cosmology (Sigad et al. 2000, hereafter S00). We extend these results to other cosmologies using analytic models. We consider a wide range of possibilities for the redshift evolution of the LV relation, including some motivated by the results of semi-analytic models (cf. Somerville \\& Primack 1999). We then present predictions of the corresponding redshift evolution of the comoving number density of bright galaxies, which may be compared directly with observations to provide quantitative constraints on how the LV relation evolves in time. ", "conclusions": "\\label{sec:conc} We have argued that observed estimates of the number density of galaxies above a given luminosity as a function of redshift can be used to constrain the redshift evolution of the LV relation. Lilly et al. (1995) found only little evolution in the number density of \\emph{bright} galaxies in the rest-frame B-band out to $z\\sim1$, consistent with the direct constraints on LV evolution out to this redshift (Vogt et al. 1996, 1997, 2000). Redshift surveys of the past decade were unable to reach the higher redshift range $1.0 \\lsim z \\lsim 2.0$ associated with the spectroscopic 'desert' or `high place of sacrifice'. The results presented by Dickinson (2000), based on NICMOS (near-IR) imaging of the Hubble Deep Field (HDF) suggest that the assembly of present-day luminous galaxies may be taking place precisely in this redshift interval. Using photometric redshifts, Dickinson finds that there are only one-third as many galaxies with rest-frame $M_V < -19$ in the redshift interval $1.4 \\lsim z \\lsim 2$ as in the interval $0 \\lsim z \\lsim 1.4$. In our fiducial \\lcdm\\ cosmology, the comoving volumes of the HDF for these two intervals are equal, implying an overall drop in the comoving number density at $z \\ga 1.4$ of about a factor of three. Taken at face value, this would correspond to dramatic evolution in the LV relation at $z\\ga 1$, rather similar to the most extreme `dimming' model shown in Figure~2. Probably the weakest link in our calculation is our assumption that the halo maximum velocity is the same as the measured internal velocity for a galaxy. The actual measured velocity will depend on the baryon fraction in the galaxy, the spatial distribution of the baryons, and the degree to which angular momentum is preserved in the collapse. However, note that because we expressed the space density in Figure~2 \\emph{relative} to that at $z=0$, our plotted results will not change by much as long as the ratio of observed velocity to the halo maximum velocity, $C \\equiv v_{\\rm obs}/\\vm$, is of order unity and roughly constant with redshift. We have calculated the value of $C(z)$ using a simple analytic model of disk formation, which assumes that a constant fraction of the halo mass condenses into a disk, and angular momentum is conserved (Mo, Mao,\\& White 1998). We find that $C(z) \\simeq 1.3$ for the redshift range $0 \\lsim z \\lsim 2$, and falls to $C(z) \\simeq 1.0$ by $z\\sim5$. This suggests that our predictions should be reasonably robust at least out to $z\\sim2$. How significant is the effect seen in the HDF? One possibility is that galaxies are preferentially missed at $z\\ga 1.4$ because of the $(1+z)^4$ cosmological surface-brightness dimming. Dickinson (2000) claims that this is unlikely to be a large effect; when bright HDF galaxies from $z\\lsim1$ are artificially redshifted to $z\\lsim 2$ with no intrinsic luminosity evolution, most of them are still easily detectable. Another possibility is that the photometric redshifts are inaccurate; Dickinson points out that there are \\emph{no} spectroscopic redshifts in the range $1.4\\lsim z\\lsim 2$ where the deficit is seen. However, this would require quite a dramatic failure where \\emph{all} of the galaxies (with photometric redshifts) assigned to $0.5\\lsim z_{\\rm phot}\\lsim 1.4$ would have to be reassigned to the higher redshift range. Most importantly, recall that the HDF covers a relatively small volume, and therefore random fluctuations along the line of sight are likely to be large. The observed deficit of galaxies might simply reflect large-scale ``spikes'' in the redshift distribution. Although there is some evidence for a similar deficit of bright galaxies at $z \\gsim 1$ in other deep optical/IR surveys (Fontana et al. 1999), secure confirmation of these results will require deep wide-field surveys with multi-band optical-IR photometry. Several such surveys are now in progress (e.g. the Cambridge-Carnegie Las Campanas IR Survey). Moreover, the DEEP survey (Davis \\& Faber 1998) will obtain measured line-widths for a large number of galaxies at $z\\ga 0.7$, which will provide direct constraints on the LV relation at high redshift. Finally, our results may be relevant to a recent proposal by Newman \\& Davis (2000) to use velocity-selected galaxies as a tracer population in a classical $dN/dz$ measure of cosmological parameters. This analysis would be subject to similar uncertainties in relating the halo velocity to observable galaxy velocity, e.g., the dependence on whether the halos are selected by $\\vm$ or $\\vvir$. Moreover, because the proposed spectroscopic surveys are magnitude limited, the indicated evolution of the LV relation could introduce severe selection biases into the analysis." }, "0005/astro-ph0005055_arXiv.txt": { "abstract": "The abundance of Li in stars formed within the past 5 Gyr is logN(Li) = 3.2($\\pm$0.2),while the corresponding value for the oldest stars in the Galaxy is logN(Li) = 2.2($\\pm$0.2).The global evidence suggests that the latter represents the full, or the major part of the primordial abundance, so that the difference of an order of magnitude is due to Li produced in the Galaxy. It is well known that spallation of insterstellar CNO by $^{4}$He and protons in galactic cosmic rays (GCR) can produce Li,but models yield a shortfall of almost an order of magnitude compared with the current observed abundance range.Another GCR reaction, $\\alpha$+$\\alpha$ fusion has been invoked to explain some Li production in the early Galaxy,but application of this to the disk yielded too much early Li or too little current Li.These failures led to a search for alternative mechanisms,essentially stellar, at particular phases of evolution:the helium flash phase in AGB stars,in novae,and during supernova. Here we stress the importance of the observed upper envelope in the plot of Li v. Fe in stars as a constraint on any mechanism in any model aiming to account for disk Li. We show that a good can be found assuming that low energy GCRs produce the Li,with the $\\alpha$+$\\alpha$ reaction as the key mechanism although production in supernovae cannot at this stage be excluded.There is an apparent time delay in the Li production, relative to O and Fe, which if confirmed could be explained by the origin of a low energy $\\alpha$-particle component in processes associated with stars of intermediate and low mass.The $\\alpha$ flux at a given epoch would then be proportional to the amount of gas expelled by low and intermediate mass stars in the Galaxy, though the acceleration of these alphas could still be linked to more energetic events as SN explosions. The present scenario appears to account coherently for the closely related observations of the temporal evolution in the Galaxy (Halo+Disk) of abundances of $^{12}$C,$^{13}$C,$^{14}$N,$^{16}$O,$^{26}$Fe, the two main peaks (one in the Halo and one in the Disk) in the G-dwarf stellar frequency distribution, and the evolution of $^{9}$Be and $^{10}$B+$^{11}$B via GCR spallation reactions without requiring the very high local cosmic-ray fluxes implied by the spallation close to SN (Casuso $\\&$ Beckman 1997).Adding a natural mechanism of differential depletion in red supergiant envelopes, we can explain the observed time evolution of the abundance of D and that of the isotopic ratios $^{7}$Li/$^{6}$Li and $^{11}$B/$^{10}$B (Casuso $\\&$ Beckman 1999) starting from an SBBN model with baryon density $\\sim$0.05. Our model also predicts the second Li-\"plateau\" found for [Fe/H] between -0.2 and +0.2, due to the \"loop back\" implied for Li (also for $^{9}$Be and B) because of the required infall of low metallicity gas to the disk. Without ruling out other mechanisms for the main production of Li in the Galactic Disk, the low-energy $\\alpha$+$\\alpha$ fusion reaction in the ISM offers a promising contribution. ", "introduction": "A few seconds after the Big Bang, four light isotopes were produced: D,$^{3}$He,$^{4}$He and $^{7}$Li (see eg. Walker et al. 1991,Copi et al. 1995,Shramm and Turner 1998); all of these warrant careful study, and here we are focusing on $^{7}$Li.The importance of understanding the evolution of the Galactic abundance of Li was highlighted in the key discovery by Spite $\\&$ Spite (SS) (1982) that the observed abundance of Li in Galactic stars does not continue to fall uniformily with decreasing iron abundance below [Fe/H]$\\simeq$-1, but levels off to a \"plateau\" at a level of logN($^{7}$Li)$\\simeq$2, which SS interpreted as corresponding to the abundance produced by big bang nucleosynthesis (SBBN).Spite and Spite measured the $^{7}$Li abundance as a function of metallicity (iron abundance) and surface temperature.They found that the $^{7}$Li abundance is flat for surface temperatures greater than about 5600K, and further, it is also flat for the stars with the lowest iron abundance.The first plateau suggests that the stars with the highest surface temperatures are not destroying their $^{7}$Li by convection (the depth of the convective zone depends on surface temperature and is shallowest for stars with the highest surface temperatures).The second plateau indicates that any post-big-bang production must be insignificant for the most metal-poor stars because the $^{7}$Li abundance does not increase with iron abundance.The case against major depletion (and hence for a plateau abundance that reflects the primeval abundance) was strengthened by the observation of $^{6}$Li in certain population II stars (Smith et al. 1993, Hobbs and Thorburn 1994).Big-bang production of $^{6}$Li is negligible; the $^{6}$Li seen was probably produced by cosmic-ray processes (along with beryllium and boron).Because $^{6}$Li is much more fragile than $^{7}$Li and yet still survived with the abundance relative to Be and B expected from cosmic-ray production, depletion of of $^{7}$Li cannot have been very significant (Steigman et al. 1993). Using this interpretation the primordial abundance is given by logN$_{P}$($^{7}$Li)=2.2($\\pm$0.2), a value confirmed in detailed work by a succession of authors (Rebolo,Beckman $\\&$ Molaro 1987,Hobbs $\\&$ Thorburn 1991,Spite 1991,Thorburn 1994) which can be combined with the SBBN produced abundance of $^{4}$He (see e.g. Pagel et al. 1992), to infer basic cosmological parameters: the universal baryon density $\\Omega$$_{b}$, and the number of massless two-component neutrino types N$_{\\nu}$.To be sure that the population II abundance of $^{7}$Li is a largely undepleted SBBN abundance, entails two essential steps: showing that population II stars (with T$_{eff}$$\\geq$5500 K) have not depleted or barely depleted their $^{7}$Li, and showing that most of the $^{7}$Li in population I stars is of Galactic origin.The first step has already been accomplished via the theoretical work of the Yale group, who showed (Pinsonneault, Deliyannis $\\&$ Demarque 1992) that sub-surface convective transport, and hence $^{7}$Li depletion is strongly suppressed at low metallicities.As a result of this, and of steadily accumulating observations, opinion (see Spite $\\&$ Spite 1993) has swung strongly behind the view that logN($^{7}$Li)$\\simeq$2.2 is the SBBN value.Thorburn's (1994) refined work on the \"plateau\", has brought out a scatter in the $^{7}$Li v. [Fe/H] plot below [Fe/H]=-1.5, which is incompatible with zero production of $^{7}$Li in the halo, but in practice strongly supports a primordial value for $^{7}$Li not far above logN$_{P}$($^{7}$Li)$\\simeq$2. The second step is quite complicated, because during the disk lifetime there may have been a number of significant production processes for Li, and also a number of destruction, or depletion processes.The initial primordial abundance masks any Li evolution in the halo, so we concentrate our attention in the present paper on production in the disk and its interpretation. The evolution of the lithium abundance in the Galactic disk can be followed via observations which define the upper envelope of the lithium abundance in stars over the range of iron metallicity, -1.5${\\leq}$[Fe/H]${\\leq}$0.1, which characterizes the disk population.The underlying assumption is that while lithium is in general depleted within stars, for a given value of metallicity the highest observed abundance value for a set of stars will correspond to minimum depletion, and hence to an optimum approximation to the Galactic interstellar lithium abundance at the epoch when the stars were formed.Following the evolution of lithium should give similar insight into the processes which form it to that which we can obtain by following the evolution of any other element.The rise in the ratio O/Fe with decreasing Fe, for example, gives the key to understanding the origin of a major fraction of Galactic oxygen in supernovae of type II, whereas iron is formed in all stars with masses greater than or equal to 1 solar mass. Although the general trend in Galactic Li evolution can be followed via the Li-Fe envelope, the effect of depletion imposes the need for the greatest care when interpreting the observed abundance in any single object.Depletion is well-known to occur in cool stars (see e.g. Pinsonneault et al. 1992,Deliyannis et al. 1990): those with T$_{eff}$ less than the solar value, and occurs also in the \"{\\it lithium gap}\" (Cayrel et al. (1984), Boesgaard (1987)) in the middle-F range of spectral classes.Stellar depletion renders lithium particularly interesting as a probe of stellar structure (Steigman et al. 1993),but makes it more difficult to interpret measured abundances in terms of production processes. However, if we are not able to account adequately for the present-day and post-solar system abundance values: logN(Li)$\\geq$3, there must remain some room for doubt about the BBN value.For this reason, as well as for its intrinsic importance as a test of Galactic evolution models, the source(s) of Galactic lithium continue to be of considerable research interest.A number of production mechanisms have been proposed: cosmic ray spallation of CNO (Reeves,Fowler \\& Hoyle (1970), Meneguzzi,Audouze \\& Reeves (1971), Walther,Mathews \\& Viola (1989)), nucleosynthesis in novae (Arnould \\& Norgaard (1979), Starrfield et al. (1978)), in the atmospheres of red giants (Cameron \\& Fowler (1971)), in supernovae (Dearborn et al. 1989,Woosley et al. 1990), and in AGB stars and carbon stars (D'Antona and Matteucci (1991)) and in black hole binaries (Martin et al. (1994)).It is well accepted that processes in the interior of normal stars not only fail to yield lithium, but tend to deplete it.In spite of the detection of individual lithium rich objects which might be characteristic sources, models which incorporate such sources into a Galactic evolution scheme (Audouze et al. (1983), Abia \\& Canal (1988), D'Antona \\& Matteucci (1991)) do not give good agreement with the observed lithium-iron envelope.Further, the spatial homogeneity of the Fe-Li curve points against sparse sets of point sources, even distributed sources, and in favour of a more diffuse origin for the lithium. Recent detailed models of Li production assumed in carbon stars, massive AGB stars,SNII and novae (Romano 1999), do not give fair fits to the very high slope of the Li abundance vs. [Fe/H] near [Fe/H]$\\simeq$-0.3 (see Fig. 1 of Romano 1999). The light nuclide $^{6}$Li is not produced dignificantly in SBBN and is expected to be produced over the lifetime of the Galaxy in Galactic cosmic ray spallation as well as $\\alpha$+$\\alpha$ fusion reactions.Its high fragility to stellar processing makes it a less useful tool than $^{7}$Li to constrain big bang nucleosynthesis, but many authors have modelled $^{6}$Li time evolution due to the assumed conexion with the $^{9}$Be and B abundances (Yoshii et al. 1997,Lemoine et al. 1997,Vangioni-Flam et al. 1999,Fields and Olive 1999, Ryan et al. 1999). It has been suggested (see,e.g.,Steigman et al. (1993)) that since (due to dust grain depletion, and ionization equilibrium uncertainly) isotope ratios can be determined more reliably in the interstellar medium than absolute abundances or ratios of different elements, the interstellar isotope ratio $^{6}$Li/$^{7}$Li might offer a better parameter to test source models than the absolute lithium abundance estimated directly in the ISM.However, measurements of the local interstellar $^{7}$Li/$^{6}$Li ratio (e.g. Lemoine et al. (1993),Meyer,Hawkins \\& Wright (1993)) show major variations, with differences of up to an order of magnitude from one IS cloud to another.Further, given the extreme difficulty of the ratio measurement in a stellar atmosphere and the consequent extreme paucity of such data as a function of metallicity together with the difficult interpretation of these data in terms of differential stellar depletion as a function of stellar surface temperature, it would be especially risky to attempt to draw conclusions at this stage by using a chemical evolution model to predict the evolution of the isotope ratio against, say, iron abundance.Because of the apparent spatial inhomogeneity it is not even safe to place too much emphasis on the well measured solar system $^{7}$Li/$^{6}$Li ratio of 12.5 (Mason (1971)).These considerations have led us to the present approach of concentrating on the overall Li abundance envelope as a key model constraint. In this paper we adopt the technique we were the first to use in Rebolo et al. (1988) of assuming that the upper envelope of lithium vs. iron abundance plot is at least a close approximation to the undepleted curve.We do in fact examine the alternative hypothesis: that this envelope represents a depletion curve, and show that this interpretation is quantitatively improbable, so leaving the way clear for the use of the lithium vs. iron envelope as a test of lithium production processes.The purpose of the paper is to show, using this envelope, which types of production processes are excluded, and which permitted.Without going into any numerical detail it is evident from inspection (see Fig. 1) that the rise in the lithium abundance towards the values found in objects close to solar (iron) metallicity occurs relatively late in the Galactic disk evolution time scale; the lithium rise lags the rise in iron, precisely the opposite case to that of oxygen (see Fig. 3).A direct implication is that processes associated with type II supernovae could, but with difficulty, yield the observed lithium production.This consideration not only covers hypothetical processes within the supernovae, but interactions of the energetic particles which they produce in processes occurring in the interstellar medium (ISM).This is just an example of how we can hope to constrain the Galactic lithium production process using the available observational data.Below we will use quantitative modelling (both analytical and numerical) with the aim of reproducing the Li-Fe envelope, thereby eliminating processes which predict significantly different envelopes.What remains will be candidate material for the process (or processes) which gave rise to some 90$\\%$ of the lithium we can observe today. In section 2 we show, using simplified analytical models, how the overall shape of the lithium-iron curve for the Galactic disk can be reproduced on the assumptions of delayed production of lithium and increasing infall of gas to the disk.In section 3 we describe briefly a numerical chemical evolution model used to handle the detailed evolution of lithium.In section 4 we examine quantitatively the problem of the galactic cosmic ray (GCR) flux as a candidate source for lithium.In section 5 we compare some of the suggested production mechanisms for lithium.Finally we draw some conclusions about the primordial abundance of lithium. ", "conclusions": "We have surveyed mechanisms of Li production in the disk, and confronted them with the upper envelope of the Li-Fe observations, which we have taken to represent the Li-Fe evolution curves in the absence of stellar depletion for the individual objects observed.As a result we can conclude that: 1)Mechanisms relying on SNIIe to produce Li cannot be excluded in at least an approximate explanation of the observations.This is true for production within the SNe themselves, but also holds for GCR fluxes originating in SNIIe. 2)Mechanisms whose time dependence is that of the SFR give either too much Li in the early disk or too little in the later disk. 3)Mechanisms which rely on SNIe to produce the Li (again either in the immediate surroundings of the SNa or via a more generally dissipated GCR flux originating in SNIe) predict that the disk Li should grow proportionally to Fe, which does not appear to fit the observations. 4)An attempt to reproduce the results on the assumption that the contemporary maximum abundance, logN(Li)$\\simeq$3.4 is the true primordial abundance, and that the Li-Fe envelope is a pure depletion curve also fails by a wide margin. 5)Mechanisms which produce an increase in disk Li significantly delayed with respect to that of Fe can explain the observations very well. We have in this article explored one such mechanism: the production of Li via $\\alpha$+$\\alpha$ fusion reaction in the ISM due to low energy cosmic rays whose source of origin is the atmospheres of low and intermediate mass stars.This mechanism has the virtue that these stars have lifetimes comparable with that of the disk, so that their collective gas expulsion rate has accumulated progressively throughout the disk lifetime, leading automatically to a delay with respect to Fe in the Li production curve.We have explained that even if the acceleration of the GCR is due to SNe envelopes, the product of injection rates and acceleration rates retains the delay implied by the observations (further work on acceleration mechanisms such as that due to stellar wind termination shocks is, however, well worth exploring in this context).Support for the possibility of this mechanism is provided by the observed similarity between the GCRS composition and that of the solar corona which is biased according to the first ionization potential, and we note in this context the statement of Ellison et al. (1997) that \"in the outer solar atmosphere the solar coronal gas, the solar wind, and the $\\sim$MeV solar energetic particles have undoubtedly a composition biased according to FIP\", together with the fact that the hydrogen and precisely helium are not well fitted by the alternative model of Meyer et al. (1997) and Ellison et al. (1997) based on volatility and mass to charge to explain the GCRS.These considerations permit an origin in an environment close to thermal equilibrium, i.e. typical of stars of moderate mass.We have incorporated the mechanism in an evolutionary model of the disk previously demonstrated to be capable of accounting well for the Be and B vs. Fe observations (Casuso \\& Beckman (1997)), and which gives a particularly good account of the G-dwarf metallicity distribution in the solar neighborhood.The resulting Li-Fe plots include very fair fits to the observed Li-Fe envelope. We have included in this scenario a natural mechanism of differential depletion (Casuso \\& Beckman 1999) operating within red supergiant envelopes, which can account for the observed D/H v. time and isotopic ratios of $^{7}$Li/$^{6}$Li and $^{11}$B/$^{10}$B v. time. However we would not at this stage wish to rule out the possibility of other mechanism or mechanisms for disk lithium production.The observational weight of the stellar Li abundances, as we have shown, does place some strong constraints on Li-production models.One of the clearest conclusions we can draw is that the \"high\" value log N(Li)$\\simeq$3.4 for the primordial Li abundance can be quantitatively rejected using the Li-Fe observational constraint.The assignation of a value close to the \"Spite plateau\" (Spite \\& Spite (1982)) value: log N(Li)$\\simeq$2.2 as primordial is thereby strengthened.In this context the comprehensive study by Thorburn (1994) of Li in halo stars, in which a contribution to the plateau produced by the $\\alpha$+$\\alpha$ reaction due to the halo GCR flux is shown to account well for the observed scatter and slight rise in the Li abundance below [Fe/H]=-1.5, makes a suggestive link with the disk model tested in the present paper.The importance of the $\\alpha$+$\\alpha$ process has almost certainly been previously underestimated in the disk, and the powerful constraint on evolutionary processes and models implied by the Li vs. Fe observations has not been adequately taken into account; it is these aspects of the lithium puzzle which the present paper has been designed to expose. {\\large Note added in Proof:} Newly observations of Li and $^{7}$Li/$^{6}$Li in ISM (toward o Per and $\\zeta$ Per) by Knauth, Federman, Lambert, Crane (Nature in press), give a variation in $^{7}$Li/$^{6}$Li ratio (from near 2 which is the expected for Li production from spallation or alpha-alpha fusion reactions purely, to near 11 which is very similar than that of solar value), together with very similar reported values for Li/H abundance (near 11x10$^{-10}$) for the two clouds in contrast with the solar value of 20x10$^{-10}$).Also, the two clouds are near the star forming region IC 348. All of these data agree very well with our picture of production of light elements in the ISM via GCRs (Be,B) (Casuso and Beckman 1997) and via alphas of low-energy (Li).We explained this variation (in fact a fall off) via a model in which the envelopes of red-supergiant stars (so, star forming region) deplete differentially $^{6}$Li and $^{7}$Li , and the increasing infall of non-depleted gas with time (Casuso and Beckman 1999).And also, we explained in the present article the decay on Li/H abundance from solar to actual ISM due precisely to the depletion in star forming regions in addition with the infall of non-enriched gas (see Fig. 4).So, we can explain these data without the problem inh erent to the explanation by Knauth et al., which point to the differential production of Li in the o Per direction and in the $\\zeta$ Per direction because of the higher flux of cosmic rays in the o Per direction, while observations point to almost the same total Li/H abundance." }, "0005/astro-ph0005263_arXiv.txt": { "abstract": "The next generation of gamma-ray telescopes may be able to observe gamma-ray blazars at high redshift, possibly out to the epoch of reionization. The spectrum of such sources should exhibit an absorption edge due to pair-production against UV photons along the line of sight. One expects a sharp drop in the number density of UV photons at the Lyman edge $\\epsilon_{L}$. This implies that the universe becomes transparent after gamma-ray photons redshift below $E \\sim (m_{e} c^{2})^{2}/\\epsilon_{L} \\sim 18 {\\rm GeV}$. Thus, there is only a limited redshift interval over which GeV photons can pair produce. This implies that any observed absorption will probe radiation fields in the very early universe, regardless of the subsequent star formation history of the universe. Furthermore, measurements of differential absorption between blazars at different redshifts can cleanly isolate the opacity due to UV emissivity at high redshift. An observable absorption edge should be present for most reasonable radiation fields with sufficient energy to reionize the universe. Ly$\\alpha$ photons may provide an important component of the pair-production opacity. Observations of a number of blazars at different redshifts will thus allow us to probe the rise in comoving UV emissivity with time. ", "introduction": "Our knowledge of UV radiation fields and energy injection into the IGM at $z>5$ is fairly tenuous. There are two main constraints: observations of the integrated background light (Madau \\& Pozzetti 2000, Bernstein et al 1999), and the fact that no Gunn-Peterson trough is observed in the spectra of the highest-redshift quasar to date (Fan et al 2000), implying that the universe must be reionized by $ z = 5.8$. The upcoming Next Generation Space Telescope (NGST) will be able to image high-redshift star clusters or AGNs in rest frame UV continuum emission (Haiman \\& Loeb 1997,1998), and their redshifts may be obtained via H$\\alpha$ observations (Oh 1999). Nonetheless, the redshift-binned number counts will be fairly sparse, and one is unlikely to probe sufficiently far down the luminosity function to get a good measure of the comoving emissivity as a function of redshift. Observations of gamma-ray blazars (``grazars'') probe extragalactic IR and UV radiation fields, by observing the pair production opacity to $\\gamma$ rays at the high energy end (Gould \\& Schreder 1967, Stecker, De Jager \\& Salamon 1992, Madau \\& Phinney 1996, Primack et al 1999). All theoretical models have confined their predictions to low redshift grazars, with the exception of Salamon \\& Stecker (1998), who computed the $\\gamma$-ray opacity up to z=3. They concluded that because the stellar emissivity peaks between z=1 and z=2, the $\\gamma$-ray opacity shows little increase at high redshift, and thus is not dependent on the initial epoch of galaxy formation. To date, EGRET has detected 66 gamma-ray loud blazars (Hartman et al 1999), out to redshifts $z>2$. The next generation of gamma-ray telescopes (GLAST, CELESTE, STACEE, MAGIC, HESS, VERITAS, and Milagro) should greatly enlarge this sample. If the low redshift correlation between black hole mass and bulge mass (Magorrian et al 1998) continues to high redshift, then it is possible that high-redshift halos could host mini-quasars (Haiman \\& Loeb 1998, Haehnelt, Natarajan \\& Rees 1998), which should be detectable in rest frame UV emission by NGST and X-ray emission by Chandra (Haiman \\& Loeb 1998, 1999) in the redshift range $z \\sim 5-15$. This raises the exciting possibility that grazars could be detected at similarly high redshifts. It is worth noting that EGRET has detected $\\sim 56$ sources at high Galactic latitudes $b > 10^{\\circ}$ (Mukherjee, Grenier \\& Thompson 1997), with no known counterparts at other wavelengths. Their spatial distribution and log N- log S plot can be well fit by a Galactic component plus an isotropic, extragalactic contribution. Some of these may well be unidentified high-redshift blazars. In this paper, I point out that if grazars are detected at high redshifts $z>3$, the pair production opacity to gamma ray photons can be used to probe the comoving emissivity longward of the Lyman break at these extremely high redshifts, independent of the star formation rate at lower redshifts. Due to the small escape fraction of ionizing photons $f_{esc} < 5 \\%$ from host galaxies, as well as the high photoelectric opacity of the IGM at these wavelengths, the comoving number density of UV photons exhibits a sharp drop at the Lyman edge at all redshifts. Thus, there is only a limited pathlength over which a gamma-ray photon can pair produce against UV photons, before it redshifts to energies which require UV photons above the Lyman edge for pair production to take place. For $z < z_{break}$, the universe becomes optically thin to the gamma-ray photon. Thus, the detection of an absorption edge in a high-redshift grazar places an immediate constraint on the mean radiation field over the redshifts $z_{break} < z < z_{s}$. Furthermore, measurement of the different absorption at a given observed energy between blazars at redshifts $z_{1},z_{2}$ places an immediate constraint on the radiation field in the redshift range $z_{1} < z < z_{2}$. Detection of grazars at a number of redshifts would then enable one to probe the UV emissivity history of the universe. In all numerical estimates, we assume a background cosmology given by the 'concordance' values of Ostriker \\& Steinhardt (1995): $(\\Omega_{m},\\Omega_{\\Lambda},\\Omega_{b},h,\\sigma_{8 h^{-1}},n)=(0.35,0.65,0.04,0.65,0.87,0.96)$. ", "conclusions": "In this paper, I have suggested that if blazars can be detected at high redshift, detection of gamma-ray absorption due to pair production against high-redshift UV photons will provide a valuable probe of high-redshift UV radiation fields. This is because the sharp Lyman edge in the intergalactic radiation fields implies that gamma-ray photons have only a limited redshift interval in which to pair-produce. As they redshift to lower energies, they require photons with $\\epsilon > 13.6$eV to pair-produce, so the universe becomes optically thin. The shape of the attenuation curve is primarily sensitive to the overall number density of photons longward of the Lyman edge at high redshifts: the higher this number density, the lower the gamma-ray photon energy at which pair-production opacity sets in. This makes it a useful test of the overall level of star formation and ambient UV radiation fields present at high redshift. Ly$\\alpha$ photons provide an important contribution to this pair production opacity, and indeed may be the dominant source of opacity if sources with a relatively lower fluxes longward of the Lyman edge (such as quasars or low metallicity stars) are abundant. Finally, measurements of differential absorption between blazars at the same observed energies will allow us to cleanly isolate the increase in opacity due to radiation fields at high redshift. There are two large uncertainties. The first is whether GLAST will be able to see high-redshift blazars at all. However, in the unified model of AGN, the scarcity of ultra-luminous blazars is a geometrical effect (due to relativistic beaming) rather than a requirement of extremely large black hole masses. In fact, the luminosity boost provided by beaming reduces the black hole mass by several orders of magnitude below that demanded by the Eddington limit. So it is at least plausible that high redshift blazars will be detectable. The second uncertainty is whether absorption seen in a blazar will be internal, rather than due to pair production against photons in the IGM. However, at GeV energies we have some physical understanding of the observed EGRET spectra (e.g., Ghisellini et al 1998); opacity to gamma-ray photons due to internal radiation fields can be constrained by time variability arguments and other constraints. Furthermore, GLAST should assemble an extremely large catalog ($>$ few thousand) of low redshift blazars, whose spectra can be studied in detail (and the contribution to opacity due to low redshift star formation can be quantified by other means); provided blazar properties do not evolve too strongly with redshift, we should have a firm handle on the intrinsic unabsorbed blazar spectrum." }, "0005/astro-ph0005238_arXiv.txt": { "abstract": "This catalog comprises an up-to-date (December 1999) list of luminous ($>$10$^{36}$ erg/s), binary supersoft X-ray sources. This electronic version (including the accompannying Web-pages) supersedes the printed version of Greiner (1996). ", "introduction": "After the discovery of supersoft X-ray sources with Einstein Observatory observations, the ROSAT satellite with its PSPC detector has discovered about four dozen new supersoft sources and has thus established luminous supersoft X-ray sources (SSS) as a new class of objects. Though many different classes of objects emit supersoft X-ray radiation (defined here as emission dominantly below 0.5 keV which corresponds to effective temperatures of the emitting objects of $<$50 eV), we consider here sources with bolometric luminosities in the range $10^{36}-10^{38}$ erg/s. Optical observations have revealed the binary nature of several of these objects. A white dwarf (WD) model, the so-called close-binary supersoft source (CBSS) model, is perhaps the most promising (van den Heuvel et al. 1992; Rappaport, DiStefano, Smith 1994, Kahabka \\& van den Heuvel 1997). It invokes steady-nuclear burning on the surface of an accreting WD as the generator of these systems' prodigious flux. Indeed, SSS temperatures and luminosities as derived from the X-ray data suggest an effective radius comparable to that of WDs. Eight SSSs have orbital periods between approximately 4 hrs and 3.5 days. These are the candidates for the CBSS model. Mass transfer rates derived from the CBSS model are in the right range for steady nuclear burning of the accreted matter. This catalog comprises an up-to-date (December 1999) list of luminous ($>10^{36}$ erg/s) supersoft X-ray sources. We include in this catalog accreting binary sources of high luminosity which are thought to be in a state of (steady or recurrent) hydrogen burning. Since CAL 83, the prototype, is known to have an ionisation nebula (Pakull and Motch 1989), and further supersoft binaries are expected to also have one, we include also sources associated with very luminous planetary nebulae. Not included are the low-luminosity objects like single (i.e. non-interacting) white dwarfs and magnetic cataclysmic variables, and PG 1159 stars which reach similar luminosities but form a rather distinct class (e.g. Dreizler et al. 1995). Excluded are also supersoft active galactic nuclei which reach luminosities up to 10$^{45}$ erg/s, and the recently found examples of large-amplitude outbursts of supersoft X-ray emission which have been interpreted as tidal disruption events (e.g. Komossa \\& Greiner 1999). Since most of the new sources are X-ray discoveries, the final inclusion in the group of luminous close binary supersoft sources has to await the optical identification. Only then a distinction is possible among the various and quite different types of objects which show a supersoft X-ray spectrum (i.e. emission only below 0.5 keV) but have different luminosities. Due to this fact of necessary follow-up optical observations, it can well happen that a source is included in an early version of the catalog but later turns out to be of a different type. An example is \\astrobj{RX J0122.9--7521} which has long been thought to be a SMC supersoft source (Kahabka et al. 1994), but has been identified as a galactic PG 1159 star (Cowley et al. 1995, Werner et al. 1996), and therefore has been removed from this catalog. ", "conclusions": "" }, "0005/astro-ph0005524_arXiv.txt": { "abstract": "We present new deep multi-frequency radio-polarimetric images of a sample of high redshift radio galaxies (HzRGs), having redshift between 1.7 and 4.1. The radio data at 4.7 and 8.2 GHz were taken with the Very Large Array in the A configuration and provide a highest angular resolution of 0.2$''$. Maps of total intensity, radio spectral index, radio polarization and internal magnetic field are presented for each source. \\\\ The morphology of most objects is that of standard FRII double radio sources, but several contain multiple hot-spots in one or both lobes. Compared to similar samples of HzRGs previously imaged, there is a higher fraction (29 \\%) of compact steep spectrum sources (i.e. sources with a projected linear size less than 20 kpc). Radio cores are identified in about half of the sample and tend to have relatively steep spectra ($\\alpha \\le -1$). \\\\ Polarization is detected in all but 4 sources, with typical polarization at 8.2 GHz of around 10-20\\%. The Faraday rotation can be measured in most of the radio galaxies: the observed rotation measure (RM) of 8 radio sources exceeds 100 rad m$^{-2}$ in at least one of the lobes, with large gradients between the two lobes. We find no dependence of Faraday rotation with other properties of the radio sources. If the origin of the Faraday rotation is local to the sources, as we believe, then the intrinsic RM is more than a 1000 rad m$^{-2}$. Because low redshift radio galaxies residing at the center of clusters usually show extreme RMs, we suggest that the high-z large RM sources also lie in very dense environments. Finally, we find that the fraction of powerful radio galaxies with extreme Faraday rotation increases with redshift, as would be expected if their average environment tends to become denser with decreasing cosmic epoch. However this result has to be taken with caution, given the limitations of our analysis. ", "introduction": "High redshift radio galaxies play an important role in the study of the early universe: thanks to their extreme luminosity at different wavelengths it has been possible to us them as cosmological probes already for several decades. Currently there are more than 150 radio galaxies known with redshift greater than 2, and recently a powerful radio source at a redshift of 5.19 has been discovered by \\cite{bre99}, becoming the most distant known AGN. \\\\ High redshift radio galaxies (HzRGs) comprise a different population to high redshift radio-quiet galaxies, e.g. Ly-dropouts: there is evidence that they are older and more massive, and will evolve into brightest cluster galaxies rather than L$_{*}$ ellipticals (Best et al. 1997; van Breugel et al. 1998). \\nocite{bes97,bre98} \\\\ In the past few years a number of studies have concentrated on the properties of HzRG host galaxies, at visual and near infrared wavelengths (e.g. Pentericci et al. 1999, van Breugel et al. 1998, Best et al. 1997, Eales et al. 1997) \\nocite{pen99,bre98,bes97,eal97} with the main goal of studying the morphological evolution of these host galaxies, understanding the nature of the radio-optical alignment effect and discerning the various components (stellar light, scattered light, etc.) that contribute to the optical and infrared continuum emission. \\\\ However, one of the potentially most important results from recent studies on powerful radio galaxies came from radio observations and with the discovery that a significant fraction ($\\sim 20\\%$) of HzRGs have extremely large Faraday rotation, of the order of several thousands rad m$^{-2}$ (Carilli et al. 1997; Athreya et al. 1998), \\nocite{car97,atr98} similar to low redshift powerful radio galaxies residing at the center of X-ray clusters with extreme cooling flows (Taylor et al. 1994).\\nocite{tay94} This makes HzRGs potential excellent targets for finding and studying high redshift (proto) clusters. \\\\ Therefore the main purpose of these new high resolution radio polarimetric imaging observations was to enlarge the number of known HzRGs with high intrinsic Faraday rotation, and to provide cluster targets for future observations with facilities such as the new X-ray telescopes, Chandra and XMM. \\\\ High resolution radio imaging is important not only for finding high Faraday rotation radio galaxies, but also for a number of other important issues, such as the identification of the location of the active nucleus, the study of the correlation between the optical morphology and the radio jets, or the line emission gas and the radio jets, and the study of the evolution of radio size structure. \\\\ Throughout the paper we adopt a cosmology with $H_0=50$ km s$^{-1}$ Mpc$^{-1}$ and $q_0=0.5$. ", "conclusions": "We have presented high resolution multi-frequency radio polarimetric observations of a sample of 27 high redshift radio galaxies. Maps of the sources and the fundamental parameters of the observations were presented. This, together with previous samples makes now an extended data base from which the relation between basic properties can be studied. The main results are the following: \\begin{itemize} \\item We detect radio cores in about half of the sample. The cores often have steep spectra ($\\alpha < -1$). The core fractions depend only weakly on radio sources size, contrary to the predictions of radio source evolutionary models. The median core fraction is larger than that of matched-luminosity 3CR radio galaxies at redshift $\\sim 1$. \\item We have shown that high redshift radio galaxies tend to be more distorted than at low redshift. This implies a larger density of the external medium in which they reside. \\item We have discovered 8 new radio galaxies with very high Faraday rotation and large gradients between the different components. Given that the Faraday rotation properties do not depend on radio sources parameters such as power, total size, distortion etc, our interpretation is that these sources reside in very high density environments, possibly proto-clusters. We also find that the fraction of powerful radio galaxies with extreme Faraday rotation increases with redshift, in agreement with the change of their average environment with cosmic epoch. \\end{itemize} \\begin{figure} \\centerline{ \\psfig{figure=f5.ps,width=9.cm}} \\caption{The fraction of powerful radio galaxies with Faraday rotation in excess of 1000 rad m$^{-2}$ as a function of redshift; the horizontal error bars indicate the redshift range for each bin.} \\label{f:frac} \\end{figure}" }, "0005/gr-qc0005080_arXiv.txt": { "abstract": "The origin of cosmic gamma-ray bursts remains one of the most intriguing puzzles in astronomy. We suggest that purely general relativistic effects in the collapse of massive stars could account for these bursts. The late formation of closed trapped surfaces can occur naturally, allowing the escape of huge energy from curvature-generated fireballs, before these are hidden within a black hole. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005148_arXiv.txt": { "abstract": "The ROSAT Ultradeep HRI survey in the Lockman Hole contains a complete sample of 91 X-ray sources with fluxes in the 0.5-2 keV band larger than 1.2 $\\cdot$ 10$^{-15}$ erg cm$^{-2}$ s$^{-1}$, where over ~75 \\% of the sources are quasars or Seyfert galaxies. During the course of our optical identification work, we have obtained optical spectra of 67 narrow emission line galaxies (NELG), which are physically not associated with the X--ray sources. We have derived the equivalent width (EW) and the full width at half maximum (FWHM) for the most prominent emission lines of 41 quasars and Seyfert galaxies taken from the ROSAT Deep Survey (RDS), which has a flux limit of 5.5~$\\cdot$~10$^{-15}$ erg cm$^{-2}$ s$^{-1}$ in the 0.5-2.0 keV band. Furthermore we have obtained the EW and FWHM values of the field NELGs. Here we present the spectroscopic discrimination between RDS Seyfert galaxies and field galaxies (NELG). The analysis of the emission lines has revealed that a single object out of 69 spectros\\-copically identified AGN fits the optical criteria of Narrow-Line Seyfert 1 galaxies (NLS1). This may indicate that NLS1 contribute only marginally to the soft X-ray background, but we can not exclude a possible larger contribution. ", "introduction": "The most sensitive ROSAT surveys consist of a 207 ksec ROSAT PSPC exposure, a 205 ksec HRI raster scan and a total 1112 ksec HRI expsoure of a $0.3\\ deg^2$ area in the Lockman Hole region. The HRI images are the basis for the Ultradeep HRI Survey (Hasinger at al. 1999). At a flux limit of 10$^{-15}$~erg~s$^{-1}$ in the 0.5-2.0 keV energy band, the HRI survey has resolved about 70-80\\% of the soft X-ray background into discrete sources. The ROSAT Deep Survey (RDS), based on the PSPC image, includes a statistically complete sample of 50 X-ray sources with fluxes in the 0.5-2.0 keV band greater than 5.5$\\cdot$10$^{-15}$~erg~s$^{-1}$ (Hasinger et al. 1998). The spectroscopic identification of the RDS using the Keck and Palomar telescopes have shown that about 75\\% of the sources are quasars and Seyfert galaxies (Schmidt et al. 1998, Lehmann et al. 2000). Both surveys contain 91 X-ray sources, where the faintest sources reach a flux of 1.2$\\cdot$10$^{-15}$ erg s$^{-1}$ in the 0.5-2.0 keV band. Recent optical/infrared work has led to an identification of 88 of the 91 X-ray sources, confirming a high fraction of AGNs (72 objects). This is the largest fraction of AGNs found in any previous X-ray survey (Boyle et al. 1995, Georgantopolous et al. 1996, Bower et a. 1996 and McHardy et al. 1998). Among our AGNs is the most distant X-ray selected quasar at a redshift of 4.45 (Schneider et al. 1998). Groups and clusters of galaxies ($\\sim$10\\%) form the second most abundant class of objects. One X-ray source has been classified as a narrow emission line galaxy (NELG), whereas some deep ROSAT PSPC surveys found a significantly larger fraction of NELGs (Boyle et al. 1995, McHardy et al. 1998). We see no evidence that NELGs or other classes of objects dominate the soft X-ray counts at faint fluxes. ", "conclusions": "" }, "0005/astro-ph0005181_arXiv.txt": { "abstract": "We study the evolution of ionization fronts around the first proto-galaxies by using high resolution numerical cosmological ($\\Lambda$+CDM model) simulations and Monte Carlo radiative transfer methods. We present the numerical scheme in detail and show the results of test runs from which we conclude that the scheme is both fast and accurate. As an example of interesting cosmological application, we study the reionization produced by a stellar source of total mass $M = 2\\times 10^8 M_\\odot$ turning on at $z\\approx 12$, located at a node of the cosmic web. The study includes a Spectral Energy Distribution of a zero-metallicity stellar population, and two Initial Mass Functions (Salpeter/Larson). The expansion of the I-front is followed as it breaks out from the galaxy and it is channeled by the filaments into the voids, assuming, in a 2D representation, a characteristic butterfly shape. The ionization evolution is very well tracked by our scheme, as realized by the correct treatment of the channeling and shadowing effects due to overdensities. We confirm previous claims that both the shape of the IMF and the ionizing power metallicity dependence are important to correctly determine the reionization of the universe. ", "introduction": "An increasing number of works has been recently dedicated to the study of the reionization of the universe (Gnedin \\& Ostriker 1997; Haiman \\& Loeb 1998; Valageas \\& Silk 1999; Ciardi \\etal 2000, CFGJ; Miralda-Escud\\'e, Haehnelt \\& Rees 2000; Chiu \\& Ostriker 2000; Bruscoli \\etal 2000; Gnedin 2000; Benson \\etal 2000) which use both analytical and numerical approaches. An important refinement has been introduced by the proper treatment of a number of feedback effects ranging from the mechanical energy injection to the H$_2$ photodissociating radiation produced by massive stars (CFGJ). However, probably the major ingredient still lacking for a physically complete description of the reionization process is the correct treatment of the transfer of ionizing photons from their production site into the intergalactic medium (IGM). Attempts based on different approximated techniques have been proposed (Razoumov \\& Scott 1999; Abel, Norman \\& Madau 1999; Norman, Paschos \\& Abel 1998; Gnedin 2000; Umemura, Nakamoto \\& Susa 1999), which sometimes are not readily implemented in cosmological simulations. Therefore, it is crucial to develop exact and fast methods that can eventually yield a better treatment of the propagation of ionization fronts in the early universe. The first three papers are based on the ray-tracing method, whereas Gnedin (2000) uses the so called local optical depth approximation. Finally Umemura, Nakamoto \\& Susa (1999) implemented a time-independent ray-tracing method. Monte Carlo (MC) methods have been widely used in several physical/astrophysical areas to tackle radiative transfer problems (for a reference book see Cashwell \\& Everett 1959) and they have been shown to result in fast and accurate schemes. Here we build up on previous experience of our group (Bianchi, Ferrara \\& Giovanardi 1996; Ferrara \\etal 1996; Ferrara \\etal 1999; Bianchi \\etal 2000) in dealing with MC problems to present a case study of cosmological H$_{\\rm II}$ regions produced by the first stellar sources. The study is intended as a test of the applicability of the adopted techniques in conjunction with cosmological simulations in terms of convergency, accuracy, and speed of the scheme. As an example of interesting cosmological application, we study the reionization produced by a stellar source of total mass $M = 2\\times 10^8 M_\\odot$ turning on at $z\\approx 12$, located at a node of the cosmic web. The study includes a Spectral Energy Distribution of a zero-metallicity stellar population, and two Initial Mass Functions (Salpeter/Larson); the IGM spatial density distribution is deduced from high resolution cosmological simulations described below. In a forthcoming paper we will then improve the results of CFGJ by exploiting the strength of the MC method to release the approximations relative to the radiative transfer made in that paper. ", "conclusions": "\\begin{figure} \\hskip 1.truecm \\psfig{figure=clump_fig08.ps,height=13cm} \\caption{\\label{fig8}\\footnotesize{Comparison between the final stages ($t=100$~Myr) of the I-front evolution for a Larson (upper panel) and a Salpeter (bottom) IMF for the same source properties as in Fig. \\ref{fig7}.}} \\end{figure} The combined cosmological and radiative transfer simulations described above allow us to determine the evolution of the ionized region around the selected zero-metallicity stellar object. Illustrative slices extracted from our simulation box through the source location are shown in Fig. \\ref{fig7} at different times (10, 20, 40, 60, 100 Myr) after the source has been turned on and a for Larson IMF, together with the initial density field. The above time interval corresponds to less than two redshift units at $z\\approx 12$, and it is therefore not unreasonable for the purposes of this paper to neglect the IGM density evolution As seen from Fig. \\ref{fig7}, the ionization front (I-front) breaks out from the galaxy very rapidly, leaving behind a very clumpy ionization structure in the immediate surroundings where the IGM is overdense and, consequently, recombination times are shorter. As an aside, it is interesting to note that, as the circular velocity of the parent galaxy is a few km s$^{-1}$, the ionized gas (whose sound speed is of order of 10 km s$^{-1}$) will very likely be able to leave the galaxy, thus quenching further star formation. If this is a widespread phenomenon, it might have strong implications for the evolution of dwarf galaxies as already outlined by some authors (Barkana \\& Loeb 1999; Ferrara \\& Tolstoy 2000). Once the I-front expands in the IGM, it is channeled -- for similar reasons -- by the large filaments into the underdense volumes (voids) and its structure becomes very complex and jagged, assuming, in a 2D representation, a characteristic butterfly shape. The typical overdensity of the clumps encountered by the expanding I-front is $\\approx 30$. Qualitatively, the typical final extent of the ionized region is of about 1 comoving Mpc; this size is reached already after 60 Myr. After that time, the rapid decrease of the source ionization power (Fig. \\ref{fig2}) slows down the expansion. The final stage of the evolution is constituted by a relic H$_{\\rm II}$ region which slowly starts to recombine on a time scale which in the voids can be as long as 0.8 Gyr. The MC technique shows here all his power in following the details of the I-front evolution. For example, it tracks remarkably well the channeling induced by the large scale structure: the ionization front cannot propagate inside the densest filaments due to their large optical depth and short recombination time. In addition, also the effects of shadowing produced by isolated clumps are clearly recognized from the fingers protruding from the H$_{\\rm II}$ region to the left of the source in the upper panel of Fig. \\ref{fig8}. The ionization cone visible below the source, caused by a large overdensity located close to the source in that direction, is also a result of shadowing. Thus it seems that our scheme is highly suitable at least for this type of cosmological radiative transfer calculations. We now turn to the differences induced by our two assumptions concerning the IMF. Fig. \\ref{fig8} shows a comparison between the final stages (100 Myr after the source turn on) of the I-front evolution for a Larson (upper panel) and a Salpeter (bottom) IMF. The source stellar mass and other properties of the simulations are the same as those discussed above. For a Salpeter IMF, the volume of the ionized region is smaller by a factor 8, although the shape is very similar to the one for a Larson IMF case at an earlier stage, roughly corresponding to 10 Myr (see Fig. \\ref{fig7}). This was expected from the two adopted SEDs. In fact, the total number of ionizing photons (see Fig. \\ref{fig2}) per stellar mass formed integrated over the entire source lifetime and spectral extent is $5\\times 10^{61}$ ($10^{61}$) for the Larson (Salpeter) IMF. Thus, the IMF might play an important role for the reionization of the universe; in addition, zero-metallicity stars have larger ionizing power as already stressed previously and recently addressed by other authors (Cojazzi \\etal 2000; Tumlinson \\& Shull 2000). To assess the differences between the evolution of an I-front propagating in an inhomogeneous medium, as in the simulations discussed here, and in an homogeneous gas at the mean IGM density, we have calculated the ratio ${\\cal V} = V/V_h$ between the ionized volumes in our simulation and in the homogeneous case. This is found to be roughly independent of time and equal to ${\\cal V}=0.45$ for the Larson IMF. In the Salpeter IMF case ${\\cal V}$ varies slightly with time from 0.125 to 0.2. Hence, this result confirms that the ionized volume tends to be smaller in the inhomogeneous case, but suggests that the effect is not dramatic." }, "0005/astro-ph0005132_arXiv.txt": { "abstract": "X-ray spectral features which are unusually strong in many Narrow Line Seyfert galaxies are found to be consistent with reflection from strongly ionized matter, providing further evidence of a high accretion rate in these objects and offering a unique signature of that key parameter in future observations. ", "introduction": "The abstracts of papers for this meeting showed a strengthening consensus that the defining parameter of Narrow Line Seyfert 1 galaxies (NLS1s) is indeed a high accretion rate. That proposal was first made$^{1}$ by direct analogy with the X-ray properties of GBHC in their high state, where it was also noted that the resulting increase in ionising flux would lead to the `broad line' clouds forming at larger radii, providing a natural explanation for the narrow permitted optical lines characteristic of NLS1. A prediction that the hard X-ray spectrum of NLS1 would be unusually steep, due to increased cooling of the Comptonising electrons, was borne out by subsequent \\asca\\ spectra$^{2}$. In the present paper we discuss the interpretation of additional spectral features, apparently characteristic of NLS1, which lends further support to the high accretion rate thesis and offers the exciting potential of a direct observational signature of one of the fundamental parameters of the AGN phenomenon. ", "conclusions": "There is now good evidence that another distinguishing property of NLS1 is the presence of strong spectral features supersimposed on an intrinsically steep power law. At least for Ark 564, we find that ALL the observed spectral features, from the ionized Fe K-absorption edge to the `soft excess' below $\\sim1$~keV, can be interpreted as arising from an ionized disc. We note encouraging agreement with the predictions of the model of Nayakshin \\et, where the unusual intensity and steepness of the intrinsic X-ray emission of a NLS1 is critical for a significant optical depth of intermediate ionisation to occur. We note, in passing, that this interpretation offers the exciting future prospect of using X-ray spectra as a unique diagnostic of the accretion rate in AGN. Finally, a comment on the question of NLS1 being `low mass' or `high accretion rate' objects, a debate continued in this meeting. That difference may be merely semantic, given that the NLS1 and `normal' Seyfert 1 galaxies accessible to current observation lie in a limited range of luminosity. Correspondingly, NLS1 tend to be both high accretion rate and low mass (for their luminosity). More sensitive future X-ray observations should allow the expected differences in appearance of high mass AGN with different accretion rates to be studied." }, "0005/astro-ph0005242_arXiv.txt": { "abstract": "We present observations of a sample of optically-faint, hard X-ray sources of the kind likely to be responsible for much of the hard X-ray background. We confirm that such sources are easily detected in the near-infrared, and find that they have a featureless continuum suggesting that the active nucleus is heavily obscured. The infrared colours of the majority of the targets observed are consistent with absorbed elliptical host galaxies at $z=1-2$. It is likely that we are observing some of the brighter members of the important new class of X-ray Type II quasars. ", "introduction": "The X-ray Background (XRB) above 2\\keV\\ has at last been mostly resolved into point sources by the {\\sl Chandra} X-ray Observatory (Mushotzky et al 2000; Brandt et al 2000). {\\sl Chandra}'s superb sub-arcsecond imaging provided the high-sensitivity confusion-less images for this breakthrough. The generally accepted model for the XRB is that it is dominated by absorbed active galactic nuclei (AGN: Setti \\& Woltjer 1989; Madau, Ghisellini \\& Fabian 1994; Comastri et al 1995), a significant fraction of which are Compton-thick (i.e. have an absorbing column $>10^{24}\\pcmsq$): collectively such objects are Type~II AGN. So far, only the 2-7 keV XRB has been (mostly) resolved with {\\sl Chandra}. This regime is not expected to be sensitive to large numbers of Compton-thick objects, where the X-ray emission emerges only above 5\\keV\\ (Wilman \\& Fabian 1999; Wilman, Fabian \\& Nulsen 2000); it should reveal sources with column densities more typically of $10^{22-23}$\\pcmsq. The Compton-thick objects provide much of the power where the XRB spectrum peaks in $\\nu I_{\\nu}$. Correction of the XRB for absorption (Fabian \\& Iwasawa 1999) shows that accretion at the standard 10 per cent efficiency onto massive black holes can approximately account for the local mass density in black holes (Magorrian et al 1998) and that about 85 per cent of that accretion power is absorbed and re-radiated in the mid- to far-IR. Despite the recent progress resolving the hard XRB into discrete sources with the correct collective spectrum, the actual identification of many of these objects is, however, not straightforward. Roughly one-third are blue broad-line quasars, another third are identified with faint, optically-normal galaxies and the final third have only extremely faint optical counterparts or no detectable counterpart at all (Mushotzky et al 2000; Brandt et al 2000; Maiolino et al 2000). Mushotzky et al (2000) find infrared HK$'$ band counterparts for most of their X-ray detected sources. We have used SCUBA maps of the cores of the lensing clusters A2390 and A1835 to place deep submillimetre limits on the three serendipitous {\\sl Chandra} sources which lie in the field (Fabian et al 2000). Only one (marginal) source is detected in both the X-ray and submillimetre bands. Of three X-ray sources in the HST field of A2390 we find that one plausibly has a photometric redshift of 0.9 and conclude from its hard X-ray spectrum that it is a Type~II quasar; the other two have $V>26$ (Fabian et al 2000). Such objects are therefore difficult to follow up in the optical band alone. Although two SCUBA sources in the A370 field that optically resemble AGN are detected in X-rays (Bautz et al, in preparation), none of the 10 SCUBA sources in fields flanking the Hubble Deep Field is detected in a deep Chandra observation (Hornschemeier et al 2000). In this paper we present observations which are part of a programme to determine the origin of the optically-faint, hard X-ray sources that are most likely responsible for the hard X-ray background. Given the possibility that the sources are highly redshifted and/or obscured, we have sought infrared counterparts at the X-ray target position. Our work bridges other surveys in this field which are either very deep, or shallow with a wider area coverage. By targeting the serendipitous sources from several 10-20~ksec {\\sl Chandra} observations, we are able to select the very brightest absorbed sources in each field for follow-up. ", "conclusions": "We have carried out follow-up observations of optically-faint, X-ray hard, serendipitous {\\sl Chandra} sources, and find that they are readily detected in the near-infrared. Spectra in the infrared of some of them appear flat and featureless, suggesting that strong emission-line activity is either absent or heavily obscured. Only one source -- which is the brightest optically -- shows a strong emission line in an optical spectrum, which we cannot identify unambiguously. The 0.5-7\\keV\\ fluxes of our sources are in the range $0.3-4.6\\times10^{-14}$\\ergpcmsqps which, if due to an unabsorbed non-thermal quasar spectrum, imply bolometric luminosities of a few times $10^{44}-10^{46}$\\ergps, for redshifts between $0.1$ 25~$\\mu$m) than in YSOs, indicative of much cooler dust temperatures. A particularly intriguing case is the A-type star $\\beta$~Pictoris, exhibiting both an infrared excess and a circumstellar disk seen in scattered light, which may form the evolutionary link between YSOs and Vega-type stars. Thirteen years after the IRAS mission its successor saw first light: the {\\it Infrared Space Observatory}\\footnote{Based on observations with ISO, an ESA project with instruments funded by ESA Member States (especially the PI countries: France, Germany, the Netherlands and the United Kingdom) and with the participation of ISAS and NASA.} (ISO; Kessler et al. 1996). Unlike the all-sky IRAS mission, ISO could do pointed observations and offered multiple modes of operation. It gave us our first glimpse of the sky at wavelengths longer than 100~$\\mu$m and revealed the richness of the infrared spectrum. In this review, I will discuss the results obtained by ISO on pre-main sequence and Vega-type stars, focussing on the the evolution of circumstellar dust. In section~2 I will briefly introduce the capacities of the instruments on board ISO. The next section is devoted to a qualitative discussion of the infrared spectra of young stars, and outlines a possible evolutionary scenario for circumstellar dust. Section~4 deals with the originating region of the infrared radiation from pre-main sequence stars, whereas section~5 reviews the new picture of Vega-type stars that ISO had given us and their relation to young stars. In the last section I will summarize the main conclusions of this review and will look forward to the new questions that may be answered by future infrared missions. ", "conclusions": "The {\\it Infrared Space Observatory} has yielded great new insights in the the evolution of disks into planetesimals and planets. For the first time we are able to analyze the chemical composition and evolution of circumstellar dust and only after the ISO mission has it been realized how important solid-state resonances are for the interpretation of infrared energy distributions. In fact the flux in all four IRAS bands, centered at 12, 25, 60 and 100~$\\mu$m, can be dominated by solid-state emission features, demonstrating the necessity of infrared spectroscopy to interpret the energy distributions of pre-main sequence objects. One of the most exciting ISO results has been the discovery of crystalline silicates, such as found in our own solar system, in some Herbig Ae/Be stars. The fact that in embedded YSOs the silicates are invariably amorphous shows that for intermediate-mass stars, the transition from amorphous to crystalline must occur in the HAeBe phase of evolution. The mechanism responsible for this transition is yet unclear. Laboratory experiments show that crystallization of silicates may be achieved by either a slow annealing of warm grains, or by heating the dust grains to temperatures above $\\sim$ 1000~K. Since the crystalline dust in Herbig Ae/Be stars is observed to have much lower temperatures, substantial mixing in the disk would be required in this scenario. However, a study of ISO spectra of post-AGB stars by Molster et al. (1999) shows that in these objects crystallization occurs at much lower temperatures over long time-scales. They suggest that the same low-temperature annealing process at work in these stars may also be responsible for the crystallization of silicates in young stars. If proven, this same process will be responsible for the creation of crystalline silicates in solar-system comets, a conclusion which ultimately may alter our view of the origin of our own solar system. ISO has also left us with a feeling for the richness and variety of the infrared spectra of young stars, resulting in the qualitative picture of spectral evolution sketched in section~3. However, a quantitative conformation of this picture awaits bigger samples and in particular its extension towards solar-mass young stars. Indeed, van den Ancker et al. (2000c) remark upon the fact that stellar ages as derived from the position in the Hertzsprung-Russell diagram do not show a one-to-one correlation with degree of crystallinity. The fact that in a typical young cluster only half of the pre-main sequence objects shows an infrared excess (e.g. Beckwith et al. 1990), suggests that this may be due to greatly differing time-scales for disk evolution from object to object. The lack of correlation between submm spectral index, an indicator for grain growth, and degree of crystallinity noted by these authors is more puzzling. Apparently these two processes are not necessarily coupled, suggesting that processes other than stellar mass and age, perhaps the environment of the star, are important. The discovery that the Vega-type phenomenon is common in relatively young main-sequence stars, whereas it is rare in stars older than 400~Myrs has considerably strengthened the connection between pre-main sequence and Vega-type stars. However, the final proof of an evolutionary link between these two categories of objects still remains to be made and most likely will be made soon: the last of NASA's great observatories, SIRTF, is scheduled to be launched in the near future and, with its greatly enhanced sensitivity, promises to continue the path of discoveries that has been explored by ISO." }, "0005/astro-ph0005583_arXiv.txt": { "abstract": "\\normalsize A young, nearby compact aggregate of X-ray emitting pre-main sequence stars was recently discovered in the vicinity of $\\eta$ Chamaeleontis (Mamajek, Lawson \\& Feigelson 1999, ApJ, 516, L77). In this paper, we further investigate this cluster: its membership, its environs and origins. \\rosat\\/ High-Resolution Imager X-ray data for the cluster's T Tauri stars show high levels of magnetic activity and variability. The cluster has an anomalous X-ray luminosity function compared to other young clusters, deficient in stars with low, but detectable X-ray luminosities. This suggests that many low-mass members have escaped the surveyed core region. Photographic photometry from the USNO-A2.0 catalog indicates that additional, X-ray-quiet members exist in the cluster core region. The components of the eclipsing binary RS Cha, previously modeled in the literature as post-main sequence with discordant ages, are shown to be consistent with being coeval \\prems\\, stars. We compute the Galactic motion of the cluster from {\\it Hipparcos\\,} data, and compare it to other young stars and associations in the fourth Galactic quadrant. The kinematic study shows that the $\\eta$ Cha cluster, as well as members of the TW Hya association and a new group near $\\epsilon$ Cha, probably originated near the giant molecular cloud complex that formed the two oldest subgroups of the Sco-Cen OB association roughly 10-15 Myr ago. Their dispersal is consistent with the velocity dispersions seen in giant molecular clouds. A large H {\\small I} filament and dust lane located near $\\eta$ Cha has been identified as part of a superbubble formed by Sco-Cen OB winds and supernova remnants. The passage of the superbubble may have terminated star-formation in the $\\eta$ Cha cluster and dispersed its natal molecular gas. \\\\ ", "introduction": "} Intermediate-age pre-main sequence (\\prems) stars (ages of $\\sim5-30$ Myr) with established distances and ages are rare in the astronomical literature (\\cite{Herbig78}). Low-mass stars ($0.1-3$ \\msol) in this age range are predominantly past their active classical T Tauri phase, and are usually called weak-lined T Tauri stars (WTTs) or post-T Tauri stars (\\cite{Martin97}). Nearby stars ($d < 150$ pc) in this age range are especially important for studies of stellar angular momentum evolution, stellar multiplicity, the evolution of young and luminous brown dwarfs and planets, and the evolution and longevity of circumstellar disks (e.g.\\, \\cite{Beckwith96}, \\cite{Bouvier97}, \\cite{Brandner98}, \\cite{Low99}, \\cite{RayJay99}, \\cite{Bejar99}). It is particularly valuable to studies of these issues to have samples of nearby, coeval, codistant stars in this age range. Coronal X-ray emission (or more precisely, the ratio $L_x/L_{bol}$) is elevated $1-3$ orders of magnitude above main sequence levels in low-mass stars throughout the \\prems\\, phase (e.g., \\cite{Briceno97}). Pointed X-ray observations of nearby active star-forming molecular clouds, where the bulk of the stars have modeled ages of $<$2 Myr, have identified many new weak-lined T Tauri stars missed by previous surveys (in Chamaeleon I (\\cite{Feigelson93}), Taurus-Auriga (e.g. \\cite{Strom94}), and other active star-forming regions). Copious numbers of older \\prems\\, stars have been found in star clusters such as the $30-50$ Myr-old clusters IC 2602 and IC 2391 (\\cite{Randich95}, \\cite{Stauffer97}), around the Orion molecular clouds ($<$7 Myr, \\cite{Alcala98}) and associated with Gould's Belt ($<$30 Myr, \\cite{Guillout98}). Hundreds of isolated \\prems\\, and ZAMS stars have been discovered with X-ray telescopes, particularly the $ROSAT$ All-Sky Survey (RASS, see reviews by \\cite{Neuhauser99} and \\cite{Feigelson99}), and thus provides an excellent means for locating older WTT stars which may no longer be proximate to their parent molecular cloud. Two stellar associations with X-ray-selected samples of WTTs with ages around $5-20$ Myr lie nearby: the Sco OB2 (Sco-Cen) association at $d \\simeq 110-150$ pc (e.g., \\cite{deZeeuw99}, \\cite{Preibisch99}) and the TW Hya T Association at $d \\simeq 50$ pc (\\cite{Webb99}). Last year, a new, nearby stellar aggregate was added to the list: the $\\eta$ Chamaeleontis cluster (\\cite{Mamajek99}, hereafter Paper I) with age $t \\approx 8$ Myr. The new cluster has $d$ = 97\\,pc, and contains 13 known members with masses $0.1-3$ \\msol\\, within a very small region (0.2 square degree) of sky. The cluster was discovered with a \\rosat\\/ High Resolution Imager (HRI) pointing of a tight group of 4 RASS X-ray sources previously established to be WTTs (Alcala et al. 1995, Covino et al. 1997). The low-mass X-ray-discovered stars have the lithium and H$\\alpha$ spectral signatures of WTT stars and are clustered around several intermediate-mass stars including the $V$ = 5.5 B8V star $\\eta$ Cha, and the $V$ = 6, A8V+A8V double-lined, eclipsing binary RS Cha. Paper I announced the group as the fourth nearest open cluster to the Sun and the second nearest grouping of \\prems\\, stars after the TW Hya Association. Since then, three additional candidate groups in the Sun's neighborhood have been announced: the Carina-Vela group (a possible extension of Sco OB2; \\cite{Makarov00}) the ``Tucanae Association'' (\\cite{Zuckerman00}), a new association of $\\sim$30-Myr-old post-T Tauri stars in Horologium (\\cite{Torres00}), and a small group of T Tauri stars associated with the isolated MBM 12 cloud (\\cite{Hearty00}). As with the $\\eta$ Cha cluster, these new groups will require further study to piece together the recent star-formation history of the solar neighborhood, as well as investigate the dynamics of disintegrating star clusters.. The present paper discusses the properties and origins of the $\\eta$ Cha cluster in detail. The X-ray data for cluster members is presented in Section 2. Section 3 gives preliminary evidence for the existence of additional cluster members in the core region. Section 4 compares the $\\eta$ Cha cluster to other open clusters. In Section 5, we investigate the kinematics and origins of the $\\eta$ Cha cluster with respect to other young stars in the 4th Galactic quadrant. We argue that the $\\eta$ Cha cluster, the Sco-Cen OB association, the TW Hya Association, and a group of young stars near $\\epsilon$ Cha, likely formed in or near the same giant molecular cloud complex $5-15$ Myr ago. Section 6 discusses the interstellar medium (ISM) in the direction of the $\\eta$ Cha cluster and strengthens the claim for a kinematic origin in Sco-Cen. Our findings are summarized in Section 7. Appendix A details the kinematics of the $\\eta$ Cha cluster, the TW Hya association, the $\\epsilon$ Cha group, and the Sco-Cen association. Appendix B presents notes on individual $\\eta$ Cha cluster members. Particular attention is paid to the eclipsing binary RS Cha, which we find to be consistent with being two \\prems\\, A stars, but which has been modelled in previous literature as a non-coeval post-main sequence system. ", "conclusions": "\\subsection{Summary} The principal results of this study are: (1) We have convincingly established that the group of X-ray selected stars around $\\eta$ Cha constitutes a physical open cluster, and not a chance superposition of unrelated stars. In addition to the arguments in Paper I (spatial coincidence of high- and low-mass stars; identical distances and motions of the brighter stars from {\\it Hipparcos\\,} measurements, and a self-consistent HR diagram), we find a compact cluster of photometric candidates coincident with the X-ray-discovered population (\\S 3, Figure \\ref{Fig_SDE}). The $\\eta$ Cha cluster is smaller and more poorly populated than most classical open clusters (\\S 4, Table \\ref{Table_LYNGA}). Its size and population is comparable to the small clusters seen in nearby stellar nursuries (e.g. Taurus, Lupus, etc.), but the model ages are older ($\\approx$8 Myr) and there is no associated molecular gas or dust. (2) The X-ray selected stars exhibit very high levels of magnetic activity, with powerful and high-amplitude X-ray variability (Table \\ref{Table_XRAY}, Figure \\ref{Fig_XRAY}). This is supported by photometric evidence for large star\\-spots on the stellar surfaces (Lawson et al., in preparation). (3) The kinematics of the $\\eta$ Cha and other young stellar groupings over a large region of the southern sky indicate that many have a common origin about 10-15 Myr ago during the star-formation epochs of the Upper Centaurus Lupus and Lower Centaurus Crux subgroups of the Sco-Cen OB Association (\\S 5, Table \\ref{Table_KINE}, Figure \\ref{Fig_XYZ}). In particular, both the $\\eta$ Cha cluster and TW Hya association appear to be outliers of the $\\approx$11-Myr-old LCC subgroup of the Sco-Cen association. We conclude that Sco-Cen is far larger than usually assumed (e.g. \\cite{deg89}, Z99), and that more young stars or groups of stars will be found with motions consistent with an origin in or near Sco-Cen. The presence of stellar groups like $\\eta$ Cha and TW Hya lying today $\\sim 50$ pc from the core of the Sco-Cen association can be readily understood as the consequence of their velocities inherited from the parent giant molecular cloud (\\cite{Feigelson96}). It is well-established that, on large $50-100$ pc scales, giant molecular clouds exhibit velocity gradients and dispersions around 5 \\kms (\\cite{Larson81}, \\cite{Myers83}, \\cite{Efremov98}). This high velocity dispersion is usually interpreted as the consequence of turbulence in cloud complexes. If the outlying stellar groups were unbound from the main OB association, they would have dispersed at a rate around 5 \\kms for 10 Myr, which corresponds to their observed $\\sim 50$ pc separations from the association today. The fate of the molecular material from which the $\\eta$ Cha stars formed is less certain. One possibility, mentioned by Jones \\& Herbig (1979) in another context, is that moving clouds experience resistance from the ambient medium and become separated from their newly formed stars on timescales of $10^7$ years. But the findings reviewed in \\S 6 suggest that the $\\eta$ Cha cluster resided in a changing interstellar environment. Initially, the cloud from which the cluster formed may have shared the stellar motion. At some later time the expanding stellar winds and/or supernova bubbles from the most massive Sco-Cen stars caught up to the cluster and evacuated or evaporated the $\\eta$ Cha cloud material. As the H {\\small I} and dust filament is just a few degrees southwest of the cluster today (Figure \\ref{Fig_IRAS}), this cloud-stripping may have occurred quite recently. If the star-formation efficiency of the $\\eta$ Cha cluster progenitor cloud was around $5-20$\\%, then the cloudlet was several hundred solar masses comparable to the Cha I and II clouds which are active today. It remains to be investigated astrophysically whether the Sco-Cen supernova remnants, now seen as the H {\\small I} filament with $M \\sim 10^4$ \\msol\\, could disperse a molecular cloud of this mass in the requisite time period. It is possible that the tiny molecular clouds found by Mizuno et al.\\ (1999) throughout the Chamaeleon region are left over from this dispersal process. The kinematic tie between the $\\eta$ Cha, $\\epsilon$ Cha, and TW Hya aggregates and the enormous Sco-Cen OB complex suggests that many of the young stars in the fourth Galactic quadrant may have their origins in the Sco-Cen giant molecular cloud. The three young stellar groups discussed here --- $\\eta$ Cha, $\\epsilon$ Cha, and TW Hya --- may be only a small fraction of a large population of $\\sim 10$ Myr \\prems\\, stars stars distributed in a ``halo'' around the main concentration of the Sco-Cen association. Some of these \\prems\\, stars will lie very close to the Sun. \\subsection{Future Work} While study of the dynamical state and history of $\\eta$ Cha is beyond the scope of this paper, we can make some preliminary comments on this topic. The escape velocity from the cluster is likely to be about $\\simeq 0.5-0.7$ km s$^{-1}$ near the edge of the field that {\\it ROSAT} HRI surveyed. Molecular clouds with several hundred masses of gas, a likely progenitor of the $\\eta$ Cha cluster, have internal three-dimensional velocity dispersions around $1-2$ km s$^{-1}$ (\\cite{Larson81}). If stars inherit the velocity dispersions of their progenitor molecular clouds (\\cite{Feigelson96}), then many stars may have escaped from the cluster, and the cluster core we study here may be evanescent. The dynamical history of the cluster is probably more complex than a simple loss of high velocity stars, as it depends on the rate of gas dissipation (e.g., slow thermal evaporation or sudden removal by supernova remnants). The role of binaries and possibility of mass segregation also should be studied (e.g. \\cite{Mathieu85}, \\cite{Bonnell98}). Radial velocity measurements of more cluster members and N-body simulations are needed to address these questions with confidence. It would also be interesting to compare the dynamical evolution of $\\eta$ Cha with that of the TW Hya association, which probably had a similar origin in the Sco-Cen giant molecular cloud but is now clearly unbound and dispersed. Future kinematic simulations should also include the effects of galactic differential rotation, and the gravitational field of the disk in the Z direction. The relative proximity of this cluster of stars, unobscured by molecular material, makes the $\\eta$ Cha cluster an ideal laboratory for studying aspects of the evolution of intermediate-aged \\prems\\, stars and substellar objects, such as angular momentum evolution along the Hayashi tracks, stellar multiplicity, the luminous phases of brown dwarfs, and the evolution and longevity of circumstellar disks. At a distance of 97 pc and an age of $\\sim$10 Myr, the low-mass stellar limit is defined by $I \\approx 14$ and spectral type M5 -- M6. Brown dwarfs with masses $20-70$ $M_{Jupiter}$ are relatively bright and hot objects with $I \\approx 14-17$, $T_{\\rm eff} = 2700-3100$ K, and spectral types M6 -- M8. These are more easily located and studied than the under-luminous $T_{\\rm eff} = 1500-2000$ K, L-type and T-type objects that characterize older brown dwarf populations (\\cite{Kirkpatrick99}). Identification of additional cluster members can be provided by a combination of optical spectroscopy (spectral typing and detection of key T Tauri-star diagnostic lines such as H$\\alpha$ and Li I $\\lambda$6707), optical photometry and proper motion studies. The $\\eta$ Cha cluster may even be a suitable location for searching for the presence of early planet formation; current $8-10$-m-class instruments already allow the resolution of sub-solar system-sized structures at $d \\approx$ 100 pc." }, "0005/astro-ph0005256_arXiv.txt": { "abstract": "\\rightskip 0pt \\pretolerance=100 \\noindent The EGRET source \\source\\ is the brightest and most accurately positioned of the as-yet unidentified high-energy $\\gamma$-ray sources at high Galactic latitude ($\\ell,b=89^{\\circ},25^{\\circ}$). We present a multiwavelength study of the region around it, including X-ray, radio, and optical imaging surveys, as well as optical spectroscopic classification of most of the active objects in this area. Identifications are made of all but one of the {\\it ROSAT} and {\\it ASCA} sources in this region to a flux limit of approximately $5 \\times 10^{-14}$~erg~cm$^{-2}$~s$^{-1}$, which is $10^{-4}$ of the $\\gamma$-ray flux. The identified X-ray sources in or near the EGRET error ellipse are radio-quiet QSOs, a galaxy cluster, and coronal emitting stars. We also find eight quasars using purely optical color selection, and we have monitored the entire field for variable optical objects on short and long time scales without any notable discoveries. The radio sources inside the error ellipse are all fainter than 4~mJy at 1.4~GHz. There are no flat-spectrum radio sources in the vicinity; the brightest neighboring radio sources are steep-spectrum radio galaxies or quasars. Since no blazar-like or pulsar-like candidate has been found as a result of these searches, \\source\\ must be lacking one or more of the physically essential attributes of these known classes of $\\gamma$-ray emitters. If it is an AGN it lacks the beamed radio emission of blazars by at least a factor of 100 relative to identified EGRET blazars. If it is an isolated neutron star, it lacks the steady thermal X-rays from a cooling surface and the magnetospheric non-thermal X-ray emission that is characteristic of all EGRET pulsars. If a pulsar, \\source\\ must be either older or more distant than Geminga, and probably an even more efficient or beamed $\\gamma$-ray engine. One intermittent \\ro\\ source falls on a blank optical field to a limit of $B > 23.4, V > 23.3$, and $R > 22.5$. In view of this conspicuous absence, RX~J1836.2+5925 should be examined further as a candidate for identification with \\source\\ and possibly the prototype of a new class of high-energy $\\gamma$-ray source. ", "introduction": "One of the most important advances in high-energy astrophysics in recent years is the discovery of 271 persistent high energy $\\gamma$-ray sources by the EGRET instrument aboard the Compton Gamma-ray Observatory ({\\it CGRO\\/}, Hartman \\etal~1999). While the detection of these sources is a major success, identification of their nature and origin has turned out to be a more challenging task. The principal method of identification, which relies on statistical evidence that blazars are the dominant population, is to find positional coincidences between EGRET sources and flat-spectrum radio/millimeter sources (Thompson \\etal~1995, 1996; Mattox \\etal~1997; Bloom et al. 1997). By definition blazars are flat-spectrum, radio-loud AGNs with polarized and variable optical emission. Although numerous efforts have been made at various wavelengths, only about one third of all EGRET sources have been identified with any degree of confidence. On the latest count these identifications include 66 blazars, \\ie, flat-spectrum radio quasars or BL~Lac objects (Hartman \\etal~1999), seven rotation-powered pulsars (Hartman \\etal~1999, Kaspi \\etal~2000, Ramanamurthy \\etal~1995), the nearby radio galaxy Cen A, and the Large Magellanic Cloud. Therefore approximately 196 EGRET sources remain unidentified with roughly half of these located at high Galactic latitude, $b > 10^\\circ$. Many difficulties attend the identification of EGRET sources close to the Galactic plane, but even at high Galactic latitude, the size of the typical error circle and the lack of a tight relation between gamma-ray flux and other properties such as X-ray flux and core radio flux prevent all but the brightest counterparts from being identified securely on the basis of position alone. The absence of obvious counterparts also admits the possibility that there is another population with characteristics unlike the identified EGRET sources. We have decided to explore the latter possibility by means of detailed work at other wavelengths, while in the long term the situation should improve considerably with the next generation high-energy $\\gamma$-ray mission {\\it GLAST\\/}, which will produce more precise source locations. We have chosen for a case study the unidentified EGRET source \\source. This object may be the best candidate for the prototype of a new population different from blazars or pulsars. It is the brightest of the as-yet unidentified EGRET sources at high Galactic latitude ($\\ell,b=89^{\\circ},25^{\\circ}$), and the one with the smallest error circle. Because it is strongly detected and well away from the confusing diffuse emission in the Galactic plane, \\source\\ is localized to within a radius of only $12^{\\prime}$ at 99\\% confidence, which makes a deep multiwavelength search for a counterpart feasible. The latest analysis of the EGRET observations of \\source\\ leads to the conclusion that it shows no strong evidence for variability (Reimer \\etal~2000). Its spectrum can be fitted by a power law of photon index --1.7 from 70~MeV to 4~GeV, with a turndown above 4~GeV. Such temporal and spectral behavior is more consistent with a rotation-powered pulsar than a blazar. Unlike \\source, blazars are highly variable, and exhibit steeper spectra. Prior to the observations reported herein, there were no known active galactic nuclei (AGNs) or pulsars in the error circle of \\source. Examination of existing catalogs finds no flat-spectrum radio source (Mattox \\etal~1997), no 1.4~GHz radio source of any type brighter than 4~mJy in the NRAO-VLA Sky Survey catalog (NVSS, Condon \\etal~1998), and no 4.85~GHz source brighter than 20~mJy (Becker, White, \\& Edwards~1991). Observations by Nice \\& Sayer (1997) find and no radio pulsar to an upper limit of 1~mJy at 770~MHz. Furthermore, all of the known gamma-ray blazars and pulsars appear brighter in X-rays than the upper limit that we shall present for \\source. In light of these facts, \\source\\ cannot be a blazar unless it is a radio-quiet one (requiring a redefinition of this concept), nor a pulsar unless, as we shall show, it is one with unprecedented characteristics. In this paper we present the results of radio, X-ray, and optical observations of the location of \\source. The outline of the paper is as follows: \\S 2 describes our multiwavelength data acquisition and selection techniques. \\S 3 describes the optical spectroscopy of candidates and the overall results. \\S 4 details notable properties of individual objects and assesses their prospects as the identification of \\source. Multiwavelength comparisons with known $\\gamma$-ray sources are addressed in \\S 5, and the implications and conclusions of our work are discussed in \\S\\S 6 and 7. ", "conclusions": "We identified all but one of the X-ray sources in the field of \\source\\ to a flux limit of approximately $5 \\times 10^{-14}$~erg~cm$^{-2}$~s$^{-1}$. These are radio-quiet QSOs [$F(1.4\\ {\\rm GHz}) < 0.5$~mJy], coronal emitting stars, and a cluster of galaxies. There are no flat-spectrum radio sources in the vicinity to a flux limit of $\\approx 20$~mJy, and no radio sources in the EGRET error ellipse brighter than 4~mJy at 1.4~GHz. In addition, we find no evidence of a BL~Lac object hosted in any low-redshift galaxy. We also found several QSOs, as one would expect, using purely optical color selection. Multiple-epoch optical imaging of the entire EGRET error ellipse has not revealed any notable variability. The discovery of only radio-quiet quasars in the error circle of \\source\\ is a sobering development in the search for its identification. Although the $\\gamma$-ray properties of \\source\\ are more similar to those of Geminga and other EGRET pulsars, no other indirect evidence for a pulsar, apart from one unidentified X-ray source (RX~J1836.2+5925) whose optical counterpart is probably fainter than $B=23.4, V=23.3$, and $R=22.5$, has been found. Yet, the fact that this X-ray source is variable by at least a factor of 2 would make it unique among rotation-powered pulsars. Taken together, these findings point to the possibility of a truly remarkable object, one that cannot be matched by any known class of $\\gamma$-ray source. Even in the absence of a definite identification, it is clear that \\source\\ is lacking in one or more of the physically essential attributes of any known class of $\\gamma$-ray emitter. Its radio flux is at least two orders of magnitude fainter than any of the securely identified EGRET blazars, and its soft X-ray flux is at least 50 times fainter than that of Geminga and similar EGRET pulsars. If it is an AGN it lacks the beamed radio emission of blazars. If it is an isolated neutron star, it lacks the steady thermal X-rays from a cooling surface and the magnetospheric non-thermal X-ray emission that is characteristic of all EGRET pulsars. If a pulsar, \\source\\ must be either older or more distant than Geminga, and probably an even more efficient or highly beamed $\\gamma$-ray engine. We have plans to complete the optical spectroscopy of fainter candidates in this field to $B \\approx 21.5$ and we will also study fundamental properties such as polarization and optical variability of the newly discovered AGNs. Perhaps the most important technique which we have not yet applied is polarimetry. Polarimetry provides a definitive test for synchrotron emission in an ordered magnetic field, and polarization is one of the essential properties of blazars. Perhaps the blazar nature of a radio-quiet beam in an AGN can only be demonstrated in this way. A deeper radio pulsar search would also be warranted. Finally, we will pursue the optical identification of the \\ro\\ source RX~J1836.2+5925 to the faintest magnitudes that are necessary in order to find our whether or not it is a neutron star. In combination, these observations may result in the identification of an important EGRET source, and possibly the prototype of a new class of $\\gamma$-ray emitter." }, "0005/astro-ph0005126_arXiv.txt": { "abstract": "The main results of broad band (0.1--10 keV) BeppoSAX observations of a selected sample of NLS1s are presented and discussed. It is shown that all the available data are consistent with a scenario in which NLS1s are running at a high accretion rate. ", "introduction": "An observational program of a small sample of Narrow Line Seyfert 1 Galaxies (NLS1s) has been carried out with BeppoSAX with the aim to investigate the broad band X--ray spectral and variability properties of these objects. The capabilities of the BeppoSAX detectors, and especially the relatively large MECS effective area at high energy ($>$ 5 keV), have been fully exploited to further investigate, with respect to previous ROSAT and ASCA observations, several of the distinctive properties of NLS1s. More specifically, we want to test whether the 2--10 keV spectral index distribution and the properties of the iron K--shell features in the 6--10 keV region are different from those of normal, broad--line Seyfert 1s (BLS1s). In addition, the broad energy range covered by the BeppoSAX LECS (0.1--4 keV), MECS (2--10 keV), and PDS (13--100 keV) detectors will be used to constrain the overall shape of the X--ray continuum and in particular the strength of the soft excess component and the nature of the 1 keV absorption/emission features reported in several ASCA observations \\cite{L97}, \\cite{F98}, \\cite {TGN98}, \\cite{L992}, \\cite{TGN99}, \\cite{V991}, \\cite{V992}. In this paper, a summary of the most important results obtained by BeppoSAX are presented and discussed. A more detailed analysis of the X--ray observations complemented by optical and UV data for some sources can be found elsewhere : \\cite{C98}, \\cite{C00}, \\cite{P00}. ", "conclusions": "The present results fit fairly well with the hypothesis of a higher accretion rate relative to the Eddington rate in NLS1s with respect to BLS1s. If the energy conversion efficiency is the same, NLS1s should have smaller black hole masses and a correspondingly higher accretion disc temperature. The steep soft excess and the good fit obtained with thermal components are explained by a shift of the accretion disc spectrum in the soft X--ray band, while the rapid variability is naturally accounted for by the small black hole mass. The strong soft excess could lead to a strong Compton cooling of the hot corona electrons and thus to a steep hard tail. In some models the disc surface layers become strongly ionized when the accretion rate approaches the Eddington limit, which fits nicely with the detection of ionized iron K$\\alpha$ lines in a few objects. Finally the optical line width is inversely proportional to $L/L_{Edd}$ if the broad line region is virialized and its radius is a function of luminosity alone \\cite{LF97}, \\cite{BBF96}. The observation of ionized lines implies that reprocessing is occurring at some level; however, the strong soft component cannot be due to disc reprocessing alone unless the primary emission is highly anisotropic or the high energy spectrum extends up to the MeV region without any cut--off. Unfortunately the BeppoSAX sensitivity at high energies is not good enough to measure the shape of the high energy spectrum especially for steep spectrum sources, and sensitive X--ray observations of NLS1s at E$>$ 10 keV are not foreseen in the near future. A reliable estimate of the amount of reprocessed radiation in NLS1s would be also extremely important to better understand the nature of the soft X--ray features detected by ASCA. Broad band observations in the 0.1--10 keV range with better sensitivity and higher energy resolution will be/are being carried out by {\\it Chandra} and {\\it XMM--Newton}. The X--ray continuum shape and the intensity of any emission/absorption features will be measured with unprecedented detail. In addition, variability studies of the various spectral components will be valuable to test the leading hypothesis of an extreme accretion rate in NLS1s. \\begin{ack} I thank all the people who, at all levels, have made possible the SAX mission. This research has made use of SAXDAS linearized and cleaned event files (Rev.2.0) produced at the BeppoSAX Science Data Center. It is a pleasure to thank all the scientists involved in the BeppoSAX Core Program NLS1 observations for the fruitful collaboration. Partial support from the Italian Space Agency under the contract ASI--ARS--98--119, and the Italian Ministry for University and Research (MURST) under grant Cofin--98--02--32 are acknowledged. \\end{ack}" }, "0005/astro-ph0005476_arXiv.txt": { "abstract": "The claim that large-scale structure data independently prefers the \\uLambda~Cold Dark Matter model is a myth. However, an updated compilation of large-scale structure observations cannot rule out \\uLambda CDM at 95\\% confidence. We explore the possibility of improving the model by adding Hot Dark Matter but the fit becomes worse; this allows us to set limits on the neutrino mass. ", "introduction": "It has been known for several years that the Standard Cold Dark Matter model, (SCDM, \\uOmega$_m=1, h=0.5, n=1$)\\footnote{\\uOmega$_x$ refers to the fraction of critical energy density present in component $x$; $m$ represents all forms of matter, $\\nu$ signifies massive neutrinos, and \\uLambda~stands for the cosmological constant. Hubble's constant is given by $H_0 = 100 h$km/s/Mpc. $n$ refers to the power-law index of the primordial power spectrum.} could not simultaneously agree with COBE and large-scale structure observations. Viable alternatives included Tilted Cold Dark Matter (TCDM, \\uOmega$_m=1, h=0.5, n=0.8$), Cold + Hot Dark Matter (CHDM, \\uOmega$_m=1$, \\uOmega$_\\nu=0.2, h=0.5, n=1$), Open Cold Dark Matter (OCDM, \\uOmega$_m=0.3, h=0.7, n=1$), and \\uLambda CDM (\\uOmega$_m=0.3$, \\uOmega$_\\Lambda=0.7, h=0.7, n=1$). Gawiser \\& Silk\\cite{gawisers98} performed a quantitative comparison of these models with a compilation of CMB anisotropy and large-scale structure data and found that CHDM was the most successful model. \\uLambda CDM, OCDM, and TCDM were inconsistent with the data at 99\\% confidence. The discrimination between models came primarily from large-scale structure data, particularly the APM galaxy power spectrum, and different values of the shape parameter ($\\Gamma$ = \\uOmega$_m h$ in $n=1$ CDM models) were preferred on large and small scales. ", "conclusions": "\\uLambda CDM is now the standard cosmological model but large-scale structure data does not prefer it over CHDM or OCDM. Current large-scale structure data cannot rule out any of those models at 95\\% confidence. \\uLambda CDM does not prefer the addition of a Hot Dark Matter component in the form of massive neutrinos. This leads to an upper limit on the mass of the most massive neutrino of 4~eV if a power-law primordial power spectrum is assumed. Forthcoming observations of large-scale structure from 2dF and SDSS will probe the horizon size at matter-radiation equality, the primordial power spectrum on small scales, and neutrino masses. \\nopagebreak" }, "0005/astro-ph0005195_arXiv.txt": { "abstract": "We present visibility measurements of the nearby Mira-like star R~Doradus taken over a wide range of wavelengths (650--990\\,nm). The observations were made using MAPPIT (Masked APerture-Plane Interference Telescope), an interferometer operating at the 3.9-m Anglo-Australian Telescope. We used a slit to mask the telescope aperture and prism to disperse the interference pattern in wavelength. We observed in R~Dor strong decreases in visibility within the TiO absorption bands. The results are in general agreement with theory but differ in detail, suggesting that further work is needed to refine the theoretical models. ", "introduction": "\\label{sec.mappit} MAPPIT (Masked APerture-Plane Interference Telescope) is an interferometer operating at the 3.9-m Anglo-Australian Telescope\\cite{Bed92,BRM94b,Mar97}. Its main features are a one-dimensional aperture mask (either a slit or an array of holes) and a prism to disperse the fringes in wavelength. This allows us to record interference fringes simultaneously over a wide wavelength range. Here we report multi-wavelength visibility measurement of the Mira-like red giant star R~Doradus, which we compare with our recent theoretical work on the wavelength dependence of angular diameters of M~giants\\cite{Jac2000,JBR2000}. MAPPIT is located in the west coud\\'{e} room of the AAT at Siding Spring Observatory, Australia. It consists of a series of optical components mounted on two parallel optical rails (see Fig.~\\ref{mappit}). For this experiment, the aperture mask was a slit placed diametrically across the pupil (right panel of Fig.~\\ref{mappit}). The wavelength region was selected by adjusting the second flat mirror, to centre the desired part of the spectrum on the detector. The parameters of the configuration are given in Table~\\ref{dec96params}. \\begin{figure} \\centering \\includegraphics[bb=41 156 513 297,width=13.5cm]{mappit2d.ps} \\hfill \\includegraphics[bb=2 20 550 600,clip=true,scale=0.18]{2daperture.eps} \\caption[]{\\label{mappit} A schematic view of the main components of MAPPIT, as used in this experiment (not to scale). Note that the field and collimating lenses are actually achromats. The detector was a Thomson 1024$\\times$1024 CCD. The aperture mask for the observations reported here was a slit, whose position relative to the primary mirror of the telescope is shown on the right. } \\end{figure} \\begin{table} \\caption[]{\\label{dec96params} MAPPIT parameters for December 1996 observations} \\vspace{0.5cm} \\begin{center} \\begin{tabular}{lc} \\hline Pupil diameter at aperture mask & 25.4\\,mm \\\\ Projected slit width & 97\\,mm \\\\ Projected maximum baseline & 3.89\\,m \\\\ Position angles & 10.4$^\\circ$ and 135$^\\circ$ \\\\ Wavelength range & 652 to 987\\,nm \\\\ Wavelength resolution & 5 to 12\\,nm \\\\ Detector pixel size & 19\\,$\\mu$m \\\\ Detector angular scale & 10.5\\,mas/pixel \\\\ CCD window ---\\\\ ~~spatial direction: & 400 pixels (not binned) \\\\ ~~wavelength direction: & 220 pixels, binned by a factor of 5 to 44 pixels \\\\ Exposure time\t& 10\\,ms (at 650\\,ms intervals) \\\\ \\hline \\end{tabular} \\end {center} \\label{obsparams} \\end {table} Most previous aperture-masking observations have used a CCD with full on-chip binning and no shutter, to collapse the interference pattern along the fringes and produce a high-speed one-dimensional readout\\cite{BHB90,WBB92,THB97,BZvdL97}. However, the curved fringes in a broadband wavelength-dispersed system require full two-dimensional detection. For previous observations with MAPPIT in wavelength-dispersed mode, we therefore used the IPCS (Image Photon Counting System)\\cite{Bed92,BRM94b}. The IPCS had fast two-dimensional readout with no readout noise but, like other photon counting systems, it had poor red sensitivity and suffered from non-linear effects at high count rates. The observations reported here used a CCD operating as a conventional two-dimensional detector. Only a portion of the CCD was read out and some on-chip binning was used (see Table~\\ref{dec96params}), to give faster readout and reduce the effect of readout noise. A shutter was used to set the exposure time to 10\\,ms, with one frame being recorded every 650\\,ms. The wavelength resolution ranged from about 5\\,nm per binned pixel at 650\\,nm to about 11\\,nm per binned pixel at 950\\,nm. ", "conclusions": "We have presented multi-wavelength visibility observations of R~Dor and compared them with theoretical models. We find that fundamental-mode 1M$_{\\odot}$ and 1.2M$_{\\odot}$ Mira models provide the best fits to our observations. First overtone 1M$_{\\odot}$ Mira models and static models do not fit our observations well. Therefore, if R~Dor does pulsate in the first overtone mode (as Bedding et al.\\cite{BZJ98} suggest) then the first-overtone models will need to be extended to other parameters and phases, and/or to be refined to accommodate our results. Finally, we note that there is still uncertainty as to whether the pulsation mode of Miras is the fundamental or the first overtone. Observed line velocities are too large for first-overtone pulsation, but angular diameters are too large for fundamental-mode pulsation. To resolve this, it has been suggested that scattering, e.g., by molecules or dust, may cause angular diameter measurements to over-estimate the true stellar size\\cite{PCDFR99}. Multi-wavelength observations like those presented here should be able to place constraints on the amount of scattering that can be present." }, "0005/astro-ph0005530_arXiv.txt": { "abstract": "We have obtained optical long-slit spectroscopy of the nucleus of M32 using the Space Telescope Imaging Spectrograph aboard the {\\it Hubble Space Telescope}. The stellar rotation velocity and velocity dispersion, as well as the full line-of-sight velocity distribution (LOSVD), were determined as a function of position along the slit using two independent spectral deconvolution algorithms. We see three clear kinematical signatures of the nuclear black hole: a sudden upturn, at $\\sim 0.3''$ from the center, in the stellar velocity dispersions; a flat or rising rotation curve into the center; and strong, non-Gaussian wings on the central LOSVD. The central velocity dispersion is $\\sim 130$ \\kms\\ (Gaussian fit) or $\\gap 175$ \\kms\\ (corrected for the wings). Both the velocity dispersion spike and the shape of the central LOSVD are consistent with the presence of a supermassive compact object in M32 with a mass in the range $2-5\\times 10^6\\Msolar$. These data are a significant improvement on previous stellar kinematical data, making M32 the first galaxy for which the imprint of the black hole's gravitation on the stellar velocities has been observed with a resolution comparable to that of gas-dynamical studies. ", "introduction": "\\subsection{Method} An observed spectrum $I(\\lambda)$ is the convolution of the line-of-sight velocity distribution $N(V)$ of the stars within the aperture with the spectrum of a single star $T(\\lambda)$: \\begin{equation} I(\\ln\\lambda) = \\int N(V)T(\\ln\\lambda-V/c)dV. \\end{equation} The goal is to extract an estimate $\\hat{N}(V)$ of the true stellar broadening function $N(V)$ given $I(\\lambda)$ and $T(\\lambda)$, both observed with the same instrument. For $T(\\lambda)$ we adopt the template spectrum of Figure \\ref{fig_spectra}. Two independent deconvolution routines were used. The first algorithm, the ``Fourier Correlation Quotient'' (FCQ) method (\\cite{ben90}; \\cite{bsg94}), constructs an estimate of the broadening function using Fourier techniques. The FCQ routine differs from earlier Fourier algorithms (e.g. \\cite{sar77}) in that the deconvolution is based on the template-galaxy correlation function rather than on the spectra themselves. This approach is less sensitive to template mismatch (\\cite{ben90}). The second algorithm, ``Maximum Penalized Likelihood'' (MPL), finds $\\hat{N}(V)$ as the solution to a penalized likelihood problem. The MPL estimate of $N(V)$ is computed on a grid in $V$ in such a way as to optimize the fit of the convolved template to the galaxy spectrum, subject to a ``penalty'' that measures the lack of smoothness of $\\hat{N}(V)$ (\\cite{mer97}). Both algorithms are nonparametric in the sense that no explicit constraints are placed on the functional form of $\\hat{N}(V)$. However they differ in two ways that are important for the current study. The FCQ algorithm requires that the absorption lines in the template spectrum be narrow compared to the broadened lines of the galaxy spectrum, i.e. that the galaxy velocity dispersion be large compared to the instrumental resolution. The MPL routine works well even when the galaxy velocity dispersion is small, as long as both template and galaxy spectra are observed at the same spectral resolution, at least in the case that the template star and galaxy have the same intrinsic absorption line properties (an assumption that will not be tested here). The two algorithms differ also in the way they deal with the amplification of noise that accompanies the deconvolution. The FCQ routine uses a Wiener filter to suppress high-frequency components of the template-galaxy correlation function $\\tilde K_{T,G}$. The degree of smoothing is determined by a factor, called here $W$, which fixes the width of the Gaussian function used to model the low-frequency, or signal, component of $\\tilde K_{T,G}$. The choice $W=1$ corresponds to ``optimal'' filtering and larger values produce less smoothing; the FCQ algorithm adopts a default value of $W=1$ but automatically increases $W$ (to a maximum of 1.3) if the recovered LOSVD shows evidence of significantly non-Gaussian wings. In the MPL algorithm, the level of smoothing is determined by a factor $\\alpha$ that multiplies the smoothness penalty function. This penalty function defines any $N(V)$ that is Gaussian as ``smooth,'' regardless of its mean or dispersion, via Silverman's (1982) prescription; in the limit $\\alpha\\rightarrow\\infty$, the MPL estimate of $N(V)$ is the Gaussian function which is most consistent, in a maximum-likelihood sense, with the galaxy spectrum. There is no a priori way of computing the optimum value of $\\alpha$ in the MPL algorithm, a point that we return to below. The different effects of smoothing on the form of $\\hat{N}(V)$ are illustrated in Figure \\ref{fig_compare}, which shows estimates of the LOSVD in the central resolution element of M32 as computed by the two routines. Both algorithms produce rapidly-fluctuating solutions when undersmoothed, a consequence of the amplification of noise that always accompanies deconvolution. The only significant difference in this regime is the non-negativity of the MPL estimates, a consequence of the logarithmic form of the penalty function (\\cite{sil82}). As the smoothing is increased, systematic differences begin to appear which are related to the different smoothing algorithms in the two codes. Solutions obtained via MPL tend to be more robust with respect to the degree of smoothing, producing in the limit of large $\\alpha$ a Gaussian fit. However Figure \\ref{fig_compare} suggests that estimates of certain quantities, e.g. the wings of the LOSVD, might depend sensitively on the choice of smoothing level in either algorithm. Once an estimate of $N(V)$ has been obtained, various quantites related to the line-of-sight velocity distribution can be derived. The simplest of these are the mean and rms velocities, which we denote by $\\overline{V}$ and $\\sigma$ respectively. As is well known, both quantities are difficult to estimate for numerically-recovered LOSVD's since they are sensitively dependent on the form of $\\hat{N}(V)$ at large velocities where this function is most poorly determined. A standard alternative is to describe $\\hat{N}(V)$ by a Gram-Charlier or Gauss-Hermite (GH) series, the product of a normalizing Gaussian with a sum of Hermite polynomials $H_i$, both expressed in terms of $(V-V_0)/\\sigma_0$ (\\cite{tht90}). The parameters $V_0$ and $\\sigma_0$ take the place of $\\overline{V}$ and $\\sigma$; while their definitions are to an extent arbitrary, these parameters are typically determined by requiring the coefficients of $H_1$ and $H_2$, called $h_1$ and $h_2$, to be zero (\\cite{ger93}; \\cite{vdf93}). Because $V_0$ and $\\sigma_0$ describe the Gaussian core of the LOSVD, they are relatively insensitive to deviations of $\\hat{N}(V)$ from Gaussianity at high velocities. Information about these deviations is contained in the higher-order coefficients $h_3$, $h_4$ etc.; $h_3$ measures asymmetries in $N(V)$ and $h_4$ measures the strength of symmetric, non-Gaussian wings. The FCQ and MPL algorithms derive the GH parameters from $\\hat{N}(V)$ in slightly different ways; details are given in Appendix A. When applied to the STIS M32 spectra, the two algorithms were found to give consistent results for the lowest moments of $N(V)$, i.e. $V_0$, $\\sigma_0$ and $h_3$. However the estimates of $h_4$ differed significantly at positions outside of the central $\\sim 0.1''$. The FCQ algorithm gave $-0.15\\lap\\hat{h}_4\\lap 0$ at almost all positions; negative values of $h_4$ imply an $N(V)$ that falls off more sharply than a Gaussian at large velocities. The MPL algorithm gave values for $h_4$ in the range $0\\lap h_4\\lap 0.1$, almost all positive, corresponding to LOSVD's with super-Gaussian wings. Positive values of $h_4$ are expected near a black hole (\\cite{baw76}; \\cite{vdm94}) and are also characteristic of models with radially-anisotropic velocity distributions. We discuss the origin of this discrepancy in Appendix B. We believe that the primary reason for the systematic difference in $\\hat{h}_4$ values is the low velocity dispersion of M32. When a galaxy's velocity dispersion is comparable to the dispersion of the template star spectrum ($\\sim 50$ \\kms\\ in the case of HR7615), the FCQ algorithm has difficulty recovering the true LOSVD (Figures B2,3; \\cite{bpn91}). The $\\hat{N}(V)$'s recovered by FCQ in this regime are more sharply truncated than the true $N(V)$'s, leading to systematically low estimates of $\\hat{h}_4$. For values of $\\sigma_0$ and S/N comparable to those of M32 at $\\sim 1''$, Figure B2 shows that the estimates of $h_4$ generated by FCQ depend only weakly on the true $h_4$, with a bias that approaches $-0.1$ for a true $h_4$ of 0.1. The MPL algorithm suffers much less from this bias (Figures B3-5). Bias in nonparametric function estimates can always be reduced by reducing the degree of smoothing (e.g. \\cite{sil86}), which in the case of the FCQ algorithm means increasing $W$. Figure B5c suggests that increasing $W$ from its default value of $1$ to values of $\\sim 2$ can reduce the bias in FCQ estimates of $h_4$ by factors of $2$ or greater, even when $\\sigma_0$ is as large as $100$ \\kms . We carried out this experiment with the STIS data; the results are shown in Figure \\ref{fig_h4}. The average $\\hat{h}_4$ values recovered by FCQ in M32 are indeed dependent on $W$; a change in $W$ from $1$ to $1.5$ has the effect of increasing $\\langle\\hat{h}_4\\rangle$ from $\\sim -0.08$ to $\\sim +0.08$. The latter value is essentially identical to the mean value of $\\hat{h}_4$ recovered via MPL. We note that the dependence of $\\hat{h}_4$ on $W$ could be due either to the suggested explanation, i.e. the need to include more frequency channels for dispersions close to the instrumental resolution, or alternatively to a mismatch between galaxy and template spectral properties in the wings of the lines which results in incorrect continuum subtraction. We will not explore the second possibility here but note again that the Monte Carlo experiments in Appendix B suggest that values of $W$ of order 2 are appropriate even when there is no template mismatch. The values of $\\hat{h}_4$ recovered by MPL are also dependent on the value of the smoothing parameter $\\alpha$ but much less so (cf. Figure \\ref{fig_compare}), until $\\alpha$ is made so large that the LOSVD is forced into a Gaussian shape. The Monte Carlo tests summarized in Figure B5b suggest that the bias in $\\hat{h}_4$ as recovered by MPL is likely to be only of order $\\sim -0.02$, several times smaller than with FCQ. We conclude that the values of $\\hat{h}_4$ recovered by the two algorithms are consistent once their different biases are taken into account and that the values returned by MPL are likely to be more accurate. Henceforth we adopt the MPL estimates. The full set of GH parameters derived from the STIS spectra and their $1\\sigma$ confidence intervals are given in Table 2 (FCQ) and Table 3 (MPL). For radii $\\lap 0.7''$ from the center the data were sampled at full resolution($\\sim 0.05''$) while at larger radii they were binned spatially. The sampling at small radii is fine enough that the data points are somewhat correlated; this was done to ensure that no information concerning the steep radial gradients of the profiles was lost. \\subsection{Results for M32} Figure \\ref{fig_broad} presents LOSVDs computed via the MPL algorithm at positions separated by about $0.1''$ along the slit. One expects these broadening functions to obey $N(V;R) = N(-V;-R)$, since for a point-symmetric galaxy, the velocity distributions should reverse after passing from one side of the galaxy to the other. The LOSVDs of Fig. \\ref{fig_broad} show approximately the expected symmetry. The right-hand column of Figure \\ref{fig_broad} plots mean broadening functions averaged over the two sides of the galaxy, $\\overline{N}(V)={1\\over 2}[N(V,R)+N(-V,-R)]$; the central LOSVD has been symmetrized about $V=0$. These broadening functions show clear and consistent deviations from Gaussian form, in two respects. First, the central LOSVD exhibits strong super-Gaussian ``wings'' at high velocities. These wings are possibly present also in some of the off-center LOSVDs although with lower amplitude. Second, the off-center LOSVDs are asymmetric, with tails extending toward velocities opposite in sign to the mean velocity at each radius. These asymmetric tails are similar to those exhibited by a rotating system superimposed on a slowly-rotating bulge. Systematic problems in the spectral deconvolution, e.g. template mismatch or incorrect continuum subtraction, can easily produce features like the wings and tails seen in the broadening functions of Figure \\ref{fig_broad}. Such errors in most cases would be expected to produce features located at the {\\it same} velocity on both sides of the galaxy (e.g. \\cite{bsg94}) and are therefore an unlikely explanation for the asymmetric tails seen in the off-center LOSVDs. The strong wings seen in the central LOSVD might more plausibly be attributed to systematic errors. However we found that the wings in the central LOSVD were robust; they appeared in both the MPL and FCQ estimates of $N(V)$ (though less clearly in the latter -- see Figure \\ref{fig_compare}) and were relatively unaffected by changes in the assumed continuum level or slope. We carried out MPL deconvolutions where the fit to the galaxy spectrum was restricted to the region around only one, or two, of the three calcium-triplet lines; these LOSVDs also exhibited strong wings. We also tried using one of the other available STIS stellar templates; again the wings were only slightly affected. (The adopted template, Figure 1, produced the best overall fit to the galaxy spectrum.) Finally, we show in \\S4.2 that the wings are consistent with those predicted by stars in the gravitational field of a supermassive black hole. Figure \\ref{fig_rotate} shows the Gauss-Hermite parameter $V_0$, a measure of the stellar rotation, in the inner arc second of M32. Also plotted are $h_3$, the lowest, odd GH moment of the LOSVD, and the ``corrected'' rotation velocity, $V_{0,c}=V_0 + \\sqrt{3}\\sigma_0h_3$. $V_{0,c}$ is a closer approximation than $V_0$ to the true mean line-of-sight velocity $\\overline{V}$ (\\cite{vdf93}). The corrected rotation velocity is lower in absolute magnitude than $|V_0|$ due to the asymmetric wings of the LOSVD noted above. The STIS rotation curve is consistent with earlier ground-based measurements (Figure \\ref{fig_gb}) at radii $\\gap 1''$ but with a larger peak value, $\\sim 60$ \\kms. Furthermore the rotation curve remains flat or slightly rising into smaller radii than seen heretofore, before falling at $R\\lap 0.25''$ due to the blending of light from the two sides of the galaxy. There is a suggestion of an east-west asymmetry in the rotation curve though the effect is probably not significant. The $h_3$ profile is approximately antisymmetric about the center of M32, as expected in a relaxed galaxy. $|h_3|$ reaches a maximum value of $\\sim 0.05$ at $|R|\\approx 0.3''$ and appears to gradually decline at larger radii. This behavior is similar to that predicted in axisymmetric models (e.g. Figure 8 of \\cite{deh95}) where $h_3$ remains essentially constant at radii outside the seeing disk. The Gauss-Hermite parameter $\\sigma_0$ is shown in Figure \\ref{fig_dispersion}. Also plotted is $h_4$, the lowest, even moment of the LOSVD, and the ``corrected'' velocity dispersion, $\\sigma_{0,c}=\\sigma_0(1+\\sqrt{6}h_4)$; $\\sigma_{0,c}$ is a closer approximation than $\\sigma_0$ to the true rms velocity $\\sigma$. The velocity dispersion rises suddenly inside of $\\sim 0.3''$, approximately the same radius at which the rotation curve begins to fall. This coincidence suggests that at least part of the rise in $\\sigma_0$ is due to averaging of the rotation velocity over the two sides of the galaxy near the center, which has the effect of converting a rotation into an apparent dispersion (\\cite{ton87}). The corrected velocity dispersion $\\sigma_{0,c}$ rises well above $\\sigma_0$ near the center due to the strong non-Gaussian wings of the LOSVD. The central value of $\\sigma_{0,c}$ is $\\sim 175$ km s$^{-1}$; this should probably be interpreted as a lower limit since $h_4$ is only sensitive to the inner parts of the wings. (We argue below, based on model fits, that the rms velocity in the central resolution element may be as high as $\\sim 200 $ \\kms.) The ground-based data (Figure \\ref{fig_gb}) are consistent with the STIS dispersions at radii $\\gap 1''$ but fail to resolve the continued rise in $\\sigma_0$ inside of $\\sim 0.5''$. Dynamical models (e.g. \\cite{deh95}; \\cite{qia95}) predict $h_4(R)$ profiles similar to that in Figure \\ref{fig_dispersion} when observed with HST resolution: a central maximum; a rapid drop, to small or negative values, at $R\\approx 0.1''$; and a nearly constant value at larger radii. The predicted drop at $\\sim 0.1''$ is due to blending of the light from the two sides of the galaxy, which broadens the low-velocity part of the LOSVD and lowers the observed $h_4$. The predicted central value of $h_4$ depends strongly on the black hole mass and on the PSF; our value, $\\hat{h}_4\\approx 0.14$, is larger than in the two studies just cited, but these studies were based on rather low assumed black hole masses, $M_h=1-2\\times 10^6\\Msolar$. The true black hole mass is probably greater (\\cite{vdm98}). The behavior of $\\hat{h}_4$ at larger radii is surprising. Previous observational studies (e.g. \\cite{vdm94a}; \\cite{bkd96}) have returned smaller estimates for $h_4$ in M32. However we believe that these earlier results are not inconsistent with ours giving the difficulties involved with estimating this parameter. The van der Marel (1994a) study was based on WHT observations with a much lower spatial resolution than the STIS data. At radii $\\lap 1''$, the value of $\\hat{h}_4$ recovered by those authors was strongly affected by the PSF blending of the rotation curve discussed above, yielding negative values in the central aperture. Outside of $\\sim 2''$, van der Marel et al. found $\\hat{h}_4$ to increase sharply to $\\sim 0.03$ on both sides of the galaxy (their Figure 12). Van der Marel (private communication) notes that the values of $\\hat{h}_4$ derived from the WHT data depended sensitively on the choice of template spectrum and on the algorithm for continuum subtraction. Using a single, best-fit template, $\\hat{h}_4$ was found to lie between $\\sim0.3$ and $\\sim0.5$ throughout the inner $2''$; the lower values of $\\hat{h}_4$ in the published paper were derived using a spectral deconvolution routine that constructs an ``optimal'' template by linear superposition of a set of stellar spectra. In another ground-based study, Bender, Kormendy \\& Dehnen (1996) applied the FCQ algorithm to CFHT data of higher spatial resolution and found $\\hat{h}_4\\approx 0.05$ inside of $0.2''$, gradually falling to $\\sim 0$ at $\\sim 1.0''$. However the spectral resolution in this study was only $80$ \\kms and the derivation of $h_3$ and $h_4$ correspondingly difficult; as noted above, we also found $\\hat{h}_4\\approx 0$ from the STIS data using the FCQ algorithm and argued that these values were significantly negatively biased. Although we believe that all of these studies are consistent with our conclusion that $h_4$ is significantly positive throughout the nucleus of M32, we are less willing to strongly endorse the precise $\\hat{h}_4$ values shown in Figure \\ref{fig_dispersion}, due to the sensitive dependence of this parameter on the details of the spectral deconvolution algorithm, continuum subtraction, smoothing level, etc. In Paper II we will construct dynamical models based on a range of assumed $h_4$ profiles in order to test the dependence of the inferred black hole mass on this parameter. We may also compare our results to the van der Marel et al. (1997, 1998) HST/FOS measurements of $V_0$ and $\\sigma_0$ (Figure \\ref{fig_fos}). The FOS measurements were taken through square apertures as small as $\\sim 0.1''$ on a side, hence their spatial resolution is comparable to that of the STIS data. However the FOS is a low spectral resolution instrument and not well suited to objects like M32 with a relatively low velocity dispersion; furthermore there are difficulties in positioning the FOS and these were probably the cause of the large point-to-point variations seen by van der Marel et al. One advantage of STIS over FOS is the continuous spatial sampling which avoids potential errors in aperture placement. We find a hint in the STIS data of the asymmetry seen in the FOS $\\sigma_0(R)$ profile (a more rapid falloff on the west side). The central FOS value of $\\sigma_0$ seems significantly bigger than found here, and the FOS rotation velocities are systematically lower. \\clearpage ", "conclusions": "We used HST and STIS to obtain stellar absorption line spectra near the center of M32 in a wavelength region centered on the Calcium triplet. The spectra were analyzed using two independent spectral deconvolution routines; these gave fully consistent results except in the case of the Gauss-Hermite $h_4$ parameter, but we argued that the differences could be reconciled after taking into account the different biases of the two algorithms. The stellar rotation velocities in M32 are slightly higher than observed from the ground and remain constant into $\\sim 0.25''$ from the center. The velocity dispersions exhibit a clear spike beginning at approximately the same radius. These two kinematical profiles are consistent with those predicted by simple axisymmetric models containing central black holes with masses in the range $2-5\\times 10^6\\Msolar$. The stellar LOSVDs show significant deviations from Gaussian form as measured by the Gauss-Hermite parameters $h_3$ and $h_4$. The central LOSVD is particularly non-Gaussian, exhibiting strong, high-velocity wings. We showed that the amplitude of these wings is consistent with that predicted by simple models containing black holes with masses of order $3\\times 10^6\\Msolar$. Detailed dynamical modelling of M32 based on these data and estimates of the black hole mass will be presented in Paper II. \\bigskip\\bigskip We thank W. Dehnen and R. van der Marel for helpful discussions. This work was supported by NASA grants NAG 5-3158 and NAG 5-6037, by NSF grant AST 96-17088, and by STIS GTO funding. Data presented here were based on observations with the NASA/ESA {\\it Hubble Space Telescope}, obtained at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc. (AURA), under NASA contract NAS5-26555. \\clearpage \\centerline{\\bf Appendix A} \\centerline{\\bf Gauss-Hermite Moments} The two spectral deconvolution algorithms described above yield nonparametric estimates $\\hat{N}(V)$ of the stellar LOSVD. Here we describe the methods used by the two algorithms to derive the GH moments from $\\hat{N}(V)$. Let $N(X,Y;V)$ be the distribution of line-of-sight stellar velocities in the aperture centered at $(X,Y)$. Define the GH moments of $N$ as \\beq h_i(X,Y) = 2\\sqrt{\\pi}\\int_{-\\infty}^{\\infty} N(X,Y;V)g(w)H_i(w)dV, \\eqnum{A1} \\eeq where $H_i$ are the Hermite polynomials (as defined by \\cite{ger93}) and the weight function \\beq g(w) = {1\\over\\sqrt{2\\pi}\\gamma_0} e^{-w^2/2}, \\ \\ \\ \\ w=(V-V_0)/\\sigma_0 \\eqnum{A2} \\eeq has three free parameters $(\\gamma_0,V_0,\\sigma_0)$. Following van der Marel \\& Franx (1993), we choose these parameters at every point $(X,Y)$ such that \\beq h_0(X,Y)=1,\\ \\ \\ \\ h_1(X,Y)=h_2(X,Y)=0. \\eqnum{A3} \\eeq These definitions impose the following implicit conditions on $(\\gamma_0,V_0,\\sigma_0)$: \\beq \\gamma_0 =\\sqrt{2}\\sigma_0\\int_{-\\infty}^{\\infty} N(V)e^{-w^2/2}dw, \\eqnum{A4a} \\eeq \\beq 0 = \\int_{-\\infty}^{\\infty} N(V)e^{-w^2/2}\\ w\\ dw,\\eqnum{A4b} \\eeq \\beq 0 = \\int_{-\\infty}^{\\infty} N(V)e^{-w^2/2}(2w^2-1)\\ dw.\\eqnum{A4c} \\eeq The relations (A4) define a nonlinear minimization problem with solutions $(\\gamma_0,V_0,\\sigma_0)$ given $N(V)$. The MPL algorithm (\\cite{mer97}) derives the three parameters in just this way, using the NAG routine {\\tt E04FDF} to minimize the sum $(h_0-1)^2 + h_1^2 + h_2^2$ as a function of $(\\gamma_0,V_0,\\sigma_0)$. The higher-order GH moments are then derived using equation (A1), by numerical integration over $\\hat{N}(V)$. Most spectral deconvolution algorithms of which we are aware derive the parameters $(\\gamma_0,V_0,\\sigma_0)$ in a different way. The LOSVD is compared to the trial function \\beq {\\cal N}(V) = {\\gamma_0\\over\\sqrt{2\\pi}\\sigma_0}e^{-w^2/2}\\left[1 + \\sum_{j=3}^{j_{max}} h_jH_j(w)\\right] \\eqnum{A5} \\eeq where $j_{max}$ is the index of the highest GH moment fitted to $\\hat{N}(V)$; typically $j_{max}=4$. The integrated square deviation between $\\hat{N}(V)$ and $\\cal{ N(V)}$ is then minimized by varying the $(j_{max}+1)$ free parameters $(\\gamma_0,V_0,\\sigma_0,h_3,h_4,...,h_{j_{max}})$. This is the technique used by the FCQ algorithm. A theorem (\\cite{myl08}) guarantees the equivalence of the two approaches if $j_{max}=\\infty$ in equation (A5) (\\cite{vdf93}). However if $j_{max}\\ne\\infty$, and if the input $N(V)$ can not be precisely represented by a finite GH series with $j\\le j_{max}$, the results given by the two algorithms will differ. For instance, in attempting to represent an $N(V)$ having $h_6\\ne 0$ using $j_{max}=4$, the FCQ algorithm will adjust $\\sigma_0$ and $h_4$ to incorrect values in order to better fit the high-velocity wings of the profile with the limited number of terms allowed to it. This is illustrated in Figure A1, which shows the values of $\\sigma_0$ and $h_4$ generated by the second algorithm, $\\hat\\sigma_0$ and $\\hat h_4$, compared to the true values for an input $N(V)$ with $h_4=0.15$ and nonzero $h_6$: \\beq N(V) = {1\\over\\sqrt{2\\pi}}e^{-V^2/2}\\left\\{1 + 0.15H_4(V) + h_6H_6(V)\\right\\}. \\eqnum{A6} \\eeq For $|h_6|\\gap 0.1$, the errors in $\\sigma_0$ and $h_4$ as derived from the second algorithm are $\\gap 15\\%$ and $\\gap 20\\%$ respectively. \\clearpage \\centerline{\\bf Appendix B} \\centerline{\\bf Performance Evaluation of the FCQ and MPL Algorithms} Here we compare the performance of the FCQ and MPL algorithms given simulated data. Our primary goal is to understand the source of the systematic offset of $h_4$ values as derived from the M32 spectra by the two algorithms (\\S3.1). Two independent sets of tests were carried out, the first by R. Bender and the second by D. Merritt. All tests were based on synthesized galaxy spectra generated from the STIS template spectrum (Figure \\ref{fig_spectra}) by convolving it with an assumed $N(V)$ and adding noise. The first set of tests addressed the accuracy of FCQ estimates when the galaxy velocity dispersion is low. It is well known that the accuracy of FCQ begins to fall off when the galaxy velocity dispersion becomes comparable to the dispersion of the template spectrum (e.g. \\cite{bpn91}). Figure B1 shows values of $\\hat{\\sigma}_0$ recovered by FCQ given a Gaussian-broadened template spectrum and thirty random noise realizations. The default value ($W\\approx 1$) of the smoothing parameter was used. There is a positive bias in the estimated values beginning at $\\sigma_0\\approx 100$ \\kms; the bias increases with decreasing $\\sigma_0$ becoming significant for $\\sigma_0\\approx 50$ \\kms. The bias is only weakly dependent on S/N. This bias in the estimation of $\\sigma_0$ is unlikely to be important for the nucleus of M32 where $\\sigma_0\\gap 100$\\kms. Figure B2 shows the performance of FCQ at recovering $h_4$. The template spectrum was broadened using an $N(V)$ of the form \\beq N_1(V) = {1\\over\\sqrt{2\\pi}}e^{-V^2/2\\sigma_0^2}\\left\\{1 + h_4H_4(V/\\sigma_0)\\right\\} \\eqnum{B1} \\eeq with various values of $\\sigma_0$ and $h_4$. Figure B2 reveals significant biases in $\\hat{h}_4$ for $\\sigma_0\\lap 100$ \\kms, even when S/N is as great as 100. When $\\sigma_0\\approx 50$ \\kms\\ and S/N $\\approx 30$, characteristic of M32 at $\\sim 1''$, the bias in $h_4$ is $\\sim -0.1$ for an input $h_4$ of $\\sim 0.1$. The second set of tests compared the performance of the FCQ and MPL algorithms on galaxy spectra generated from the broadening function \\beq N_2(V) = {1\\over\\pi\\sigma}{1\\over 1+(V/\\sigma)^2}, \\eqnum{B2} \\eeq a Lorentzian function, with $\\sigma=100$ \\kms. This LOSVD is qualitatively similar to what is expected in a black-hole cusp, with $N\\sim V^{-2}$ high-velocity wings. The non-trivial GH parameters are \\beq \\gamma_0=0.76986\\ \\ \\ \\ \\sigma_0=108.07\\ {\\rm km\\ s}^{-1} \\ \\ \\ \\ h_4=0.14546\\ \\ \\ \\ h_6=0.01850. \\eqnum{B3} \\eeq Figure B3 shows mean estimates of $N(V)$ obtained using the two algorithms for 100 random realizations of the noise. The smoothing parameter in both algorithms was adjusted to minimize the mean square error of $\\hat{N}(V)$ (as defined below) for each value of S/N. There is a greater bias in the FCQ estimates, as well as a persistent ``ringing'' at high velocities. Figure B4 plots the mean integrated square error (MISE) and integrated square bias (ISB) of the recovered broadening functions as functions of S/N; in the case of the FCQ algorithm, the integrated errors are shown both for the optimal choice of smoothing parameter $W_{opt}$ that minimizes the MISE, as well as for the value chosen by the algorithm ($\\sim 1.3$). The MISE of an estimate $\\hat{f}(x)$ is defined as \\beq {\\rm MISE}\\left[\\hat{f}(x)\\right] = E\\int\\left\\{\\hat{f}(x) - f(x)\\right\\}^2dx \\ \\ \\ \\ \\ \\ \\ \\ \\eqnum{B4a} \\eeq \\beq \\ \\ \\ \\ \\ \\ \\ \\ \\ \\ = \\int\\left\\{E\\hat{f}(x) - f(x)\\right\\}^2dx + \\int \\left(E\\left\\{\\hat{f}^2(x)\\right\\} - E\\left\\{\\hat{f}(x)\\right\\}^2\\right) dx \\eqnum{B4b} \\eeq \\beq = {\\rm ISB}\\left[\\hat{f(x)}\\right] + {\\rm IV}\\left[\\hat{f(x)}\\right], \\ \\ \\ \\ \\ \\ \\eqnum{B4c} \\eeq the sum of the integrated square bias ISB and the integrated variance IV (\\cite{sil86}); here $E$ denotes the expectation value, i.e. the average over many random realizations of the noise. The MISE and ISB displayed in Fig. B4 were divided by the normalizing factor $\\int \\left[N_2(V)\\right]^2 dV$. The MISE of the MPL estimates falls roughly as a power law, ${\\rm MISE}[\\hat{N}(V)]\\sim (S/N)^{-1.3}$, close to the asymptotic $(S/N)^{-1}$ of parametric estimators. Approximately $1/2$ of the total square error comes from the bias and $1/2$ from the variance. In the case of the FCQ algorithm, the MISE behaves in a more complicated way with S/N, at first falling with S/N then appearing to level off for S/N $\\gap 50$. This levelling off is a consequence of the low-velocity-dispersion bias of FCQ discussed above. For the FCQ estimates, the bulk of the MISE is due to the variance; adjusting the smoothing parameter primarily affects the bias and has little effect on the MISE. For S/N $\\approx 20$, the mean square error of the optimal FCQ estimate is a factor $\\sim 3$ greater than that of the MPL estimates. The bias in $\\hat{N}(V)$ is in the direction of wider and more steeply truncated functions, particularly in the case of the FCQ estimates (Fig. B3). This bias in $\\hat{N}(V)$ is consistent with the negative bias found above in estimates of $h_4$. Figure B5 compares the ability of the two algorithms to recover $h_4$ from the Lorentzian $N_2(V)$. Plotted there are the mean square error (MSE) and bias in estimates of $h_4$ from 100 random noise realizations; the MSE is defined, for any estimated parameter $\\hat{P}$, as \\beq {\\rm MSE}(\\hat{P}) = E\\left\\{\\hat{P} - P\\right\\}^2, \\eqnum{B5} \\eeq which can also be decomposed into contributions from the squared bias SB and the variance V: \\beq \\ \\ \\ \\ {\\rm MSE}(\\hat{P}) = \\left(E\\hat{P} - P\\right)^2 + \\left(E\\left\\{\\hat{P}^2\\right\\} - E\\left\\{\\hat{P}\\right\\}^2\\right) \\eqnum{B6a} \\eeq \\beq = {\\rm SB}(\\hat{P}) + {\\rm V}(\\hat{P}). \\ \\ \\ \\ \\ \\ \\ \\ \\eqnum{B6b} \\eeq The MSE of estimates obtained with the MPL algorithm again varies roughly as a power law, ${\\rm MSE}(\\hat{h_4})\\sim (S/N)^{-1.5}$. The bias in the MPL estimates is always negative, i.e. in the direction of more Gaussian $N(V)$'s; for $S/N\\sim 20$, this bias is a modest $\\sim -0.03$, dropping to $\\lap -0.01$ for S/N $= 100$. The FCQ estimates of $h_4$ show a considerably greater error, both in the bias and the variance. Two sets of FCQ estimates were made: first using the default value of the smoothing parameter returned by the code, $W\\approx 1.3$; and second using the optimum value $W_{opt}$ that minimized the MSE of the $h_4$ estimates at each S/N. For the default value of $W$, the algorithm returns mean estimates of $h_4$ that lie in the range $0.03-0.05$ for all values of S/N $\\ge 10$, an average error of $\\sim 70\\%$. However the optimum smoothing parameter for the recovery of $h_4$ was found to vary strongly with S/N, from $W_{opt}\\sim 0.5$ for S/N $=5$ to $W_{opt}\\sim 2$ for S/N $=100$ (Fig. B5c). Nevertheless a substantial bias remains when $W_{opt}$ is used, of order $\\sim -0.05$ even for S/N $=50-100$. These biases are larger than found above using a more Gaussian $N(V)$ with smaller $h_4$ and suggest that FCQ estimates of $h_4$ may be substantially biased even for $\\sigma_0$ as large as $\\sim 100$ \\kms\\ when the true $N(V)$ is sufficiently non-Gaussian. \\clearpage" }, "0005/astro-ph0005420_arXiv.txt": { "abstract": "We present multi-frequency maps of a portion of the Galactic Plane centered on a declination of $-60^{\\circ}$ with resolutions ranging from $0^{\\circ}.2$ to $0^{\\circ}.9$. The maps are optimized to detect unresolved source emission and are cross-calibrated against the planet Jupiter. We discuss six millimeter-bright regions, three of which are visible in all bands, and list measured flux densities. Variability is limited to less than 3.8\\% for all sources seen at 31 and 42~GHz and less than 10\\% at 144~GHz. Fractional polarization limits smaller than 0.08 are measured for all sources seen at 31 and 42~GHz. No fractional polarization limits are reported at 144~GHz. ", "introduction": "Galactic radiation at centimeter and millimeter wavelengths is attributed to a combination of the following sources of emission: synchrotron, free-free, thermal interstellar dust and variations in the electric and magnetic dipoles of dust ({\\it i.e.} spinning dust). Large-area, multi-frequency surveys help us to understand the objects and environments which emit in the far-infrared and assist in determining the mechanisms driving microwave emission. The atmosphere is relatively opaque at these frequencies making such surveys difficult to perform from low altitudes. At present, the only comprehensive microwave survey data available is from the Cosmic Background Explorer ({\\it COBE}) \\cite{ben96}. It will not be until the launch of the Microwave Anisotropy Probe ({\\it MAP}) satellite in late 2000 that a full sky survey with pixel resolution smaller than $1^{\\circ}$ will be completed. Bright millimeter-wavelength sources are important for cosmology as calibrators and a possible Cosmic Microwave Background (CMB) foreground contaminate \\cite{toff99,sok98}. There are several open questions about this class of object that include: what are their spectra; are they variable; are they polarized? We address these questions for six Galactic sources using observations made with the Mobile Anisotropy Telescope ({\\it MAT}). We focus on characterizing emission from unresolved and localized millimeter-bright regions. These data will also help assess daytime atmospheric conditions relevant to observations from similar Chilean sites. We present maps of the Galactic Plane centered on a declination of $-60^{\\circ}$ (approximately $280^{\\circ}3$ in order to detect the primary bispectrum. We estimated the secondary bispectra from the coupling between the Sunyaev--Zel'dovich (SZ) and the weak lensing effects, and from the extragalactic radio and infrared sources. Only Planck will detect the SZ--lensing bispectrum, while both MAP and Planck will detect the bispectrum from extragalactic point sources (table~\\ref{tab:sn}). We also studied how well we can discriminate among the primary, the SZ--lensing coupling, and the extragalactic point sources bispectra. We found that MAP and Planck will separate the primary from other secondary sources at 1\\% or better accuracies. This conclusion is due to the presence of acoustic oscillation in the primary bispectrum that does not appear in the secondary bispectra. The SZ--lensing coupling and the extragalactic sources are well separately measured by Planck experiment, although COBE and MAP cannot discriminate between them (table~\\ref{tab:corr}). Our arguments about the ability to discriminate among various bispectra were fully based upon the shape difference, and thus did not take into account the spectral difference in the frequency space. As pointed out by \\cite{TE96,CHT00}, the multi-band observation is so efficient to discriminate among the primary signal and the other foreground contaminants for measuring the CMB anisotropy power spectrum. Their scheme should be effective on the bispectrum as well, and the accuracy of the foreground removal will be improved further. Thus, we expect that MAP and Planck will measure the primary bispectrum separately from the foregrounds. The simplest inflationary scenario usually predicts small $f_{NL}$ $(\\sim 10^{-2})$\\cite{SB9091,GLMM94}, and the second order perturbation theory yields $f_{NL}\\sim 1$\\cite{PC96}. Thus, the significant detection of the primary bispectrum or the skewness with any experiments means that the simplest inflationary scenario needs to be modified. According to our results, if the reported detections\\cite{FMG98,Magueijo00} of the bispectrum in the COBE map were the cosmological origin, then MAP and Planck would detect the primary bispectrum much more significantly. Although Banday, Zaroubi and G\\'orski\\cite{BZG00} pointed out the one of those detections\\cite{FMG98} could be accounted for by the experimental systematic effects of COBE, the other\\cite{Magueijo00} is claimed to be significant even after removing such the systematics. Although we have not discussed so far, the spatial distribution of emissions from interstellar dust is a potential source of the microwave non-Gaussianity. Since it is very hard to estimate the bispectrum analytically, the dust map compiled by Schlegel, Finkbeiner and Davis\\cite{SFD98} could be used to estimate the dust bispectrum. For example, we found that the dimensionless skewness parameter defined by $\\left<(\\Delta T)^3\\right>/\\left<(\\Delta T)^2\\right>^{3/2}$ is as large as 51. We used the publicly available HEALPix-formatted\\cite{GHW98} $100~\\mu{\\rm m}$ map which contains 12,582,912 pixels without sky cut. The mean intensity in the map was $14.8~{\\rm MJy~sr^{-1}}$. Of course, this skewness is largely an overestimate for the CMB measurement in reality; we need to cut a fraction of sky which contains the Galactic plane, and then this will greatly reduce the non-Gaussianity. Nevertheless, residual non-Gaussianity is still a source of the microwave bispectrum, and has to be taken into account. Moreover, the form of the bispectrum measured in the dust map would reflect the physics of interstellar dust, which is highly uncertain at present, and thus studying the interstellar dust bispectrum would be challenging field." }, "0005/astro-ph0005493_arXiv.txt": { "abstract": "This paper considers scattering screens that have arbitrary spatial variations of scattering strength transverse to the line of sight, including screens that are spatially well confined, such as disks and filaments. We calculate the scattered image of a point source and the observed pulse shape of a scattered impulse. The consequences of screen confinement include: (1)~Source image shapes that are determined by the physical extent of the screen rather than by the shapes of much-smaller diffracting microirregularities. These include image elongations and orientations that are frequency dependent. (2) variation with frequency of angular broadening that is much weaker than the trademark $\\nu^{-2}$ scaling law (for a cold, unmagnetized plasma), including frequency-independent cases; and (3) similar departure of the pulse broadening time from the usually expected $\\nu^{-4}$ scaling law. We briefly discuss applications that include scattering of pulses from the Crab pulsar by filaments in the Crab Nebula; image asymmetries from Galactic scattering of the sources Cyg~X-3, Sgr~A*, and NGC~6334B; and scattering of background active galactic nuclei by intervening galaxies. We also address the consequences for inferences about the shape of the wavenumber spectrum of electron density irregularities, which depend on scaling laws for the image size and the pulse broadening. Future low-frequency ($<100$~MHz) array observations will also be strongly affected by the Galactic structure of scattering material. Our formalism is derived in the context of radio scattering by plasma density fluctuations. It is also applicable to optical, UV and X-ray scattering by grains in the interstellar medium. ", "introduction": "\\label{sec:intro} Images of scattered radio sources and distorted pulses from pulsars provide some of the most-used observables for probing microstructure in the electron density of interstellar gas. Over the last decade, interstellar scattering measurements have revealed asymmetries in the scattered images of radio sources. These are most often interpreted in terms of underlying anisotropy of the very small irregularities that diffract the radiation. That anisotropy, in turn, most likely reflects the orientation of magnetic fields in the \\ion{H}{2} gas that contains the microstructure. Angular broadening of compact sources and pulse distortions due to multi-path propagation are used to probe the amplitude of scattering and also, through the frequency scaling, to constrain the shape of the wavenumber spectrum for the microstructure. Inversion of scattering observables into information about the microstructure almost invariably relies on the assumption that the scattering strength is uniform in directions transverse to the line of sight. We reconsider the assumptions used to analyze angular and temporal broadening, in particular the assumption of uniformity of the scattering medium transverse to the line of sight. One reason is that the interstellar medium (ISM) shows structures on a wide variety of scales and so it is reasonable to expect manifestations of nonuniformities, at least in some directions. Secondly, the physics that underlies asymmetric images is quite different if the asymmetry occurs on scales much larger than diffractive scales, as they would if the asymmetry is caused by the large-scale distribution of diffracting irregularities. Thirdly, observations of the Crab pulsar show anomalous scalings of pulse broadening with frequency. These are interpreted by some as indicating that scattering occurs within the pulsar magnetosphere rather than in a cold plasma (J.~Eilek~1997, private communication; Hankins \\& Moffett~1998; Lyutikov \\& Parikh~2000). As we show, anomalous scalings occur quite naturally from cold plasma extrinsic to the pulsar if the scattering region is bounded in the transverse direction. Additionally, future observations at low radio frequencies of a variety of sources -- including high redshift sources -- are expected to reveal further anomalous scattering that most likely will be the result of confined scattering structures. Finally, the scaling with frequency of angular and pulse broadening is often used to constrain the shape of the wavenumber spectrum of scattering irregularities (e.g., Cordes, Weisberg, \\& Boriakoff~1985; Fey et al.~1991). Weakening of the frequency dependence by confined plasma structures would be interpreted as a steeper wavenumber spectrum. Thus it is important to assess the role of confined structures in the observations of scattered radio sources. In \\S\\ref{sec:summary1} we discuss previous treatments of angular and pulse broadening. In \\S\\ref{sec:theory} we derive a general formalism for scattering that takes into account arbitrary variations of scattering strength transverse to the line of sight. Examples are given in \\S\\ref{sec:examples}. Applications to the Crab pulsar and other Galactic sources are given in \\S\\ref{sec:galactic}. Future observations of extragalactic sources and at low frequencies are considered in \\S\\ref{sec:future}. In \\S\\ref{sec:ism} we discuss possible implications for the interstellar medium. Finally, in \\S\\ref{sec:summary}, we summarize our results. ", "conclusions": "\\label{sec:summary} In this paper we have shown that radio scattering observables such as image shapes and pulse broadening functions can be strongly influenced by structure in the scattering medium on length scales substantially larger than those that cause the scattering. As such, careful multi-frequency observations can be used to constrain properties of the interstellar medium on scales that are typically $\\sim 1$--10~\\hbox{AU}. Intergalactic scattering has not been identified but is certainly expected from intervening spiral galaxies, probably expected for some Lyman-$\\alpha$ clouds, and may occur from distributed ionized gas. For intergalactic scattering, relevant length scales can be comparable to the sizes of galaxies. A low-frequency VLBI survey of extragalactic sources may thus probe the level of scattering in other galaxies and in the general intergalactic medium. It is also expected that scattering of radiation from gamma-ray burst afterglows will be influenced in some cases by intervening ionized gas in the IGM as well as in the Milky Way's \\hbox{ISM}. These issues will be explored in separate articles. We thank B.~Rickett for helpful discussions. This work was supported by NSF Grant AST 9819931 to Cornell University. Basic research in radio astronomy at the NRL is supported by the Office of Naval Research." }, "0005/astro-ph0005346_arXiv.txt": { "abstract": "Since the discovery of low-ionization nuclear emission-line regions in many galaxies (LINERs), it has been recognized that they constitute a class of active galactic nuclei (AGNs) which are thought to be powered by gas accretion onto a central, supermassive black hole. LINERs are observed in approximately one third of galaxies in the local universe and it has been often thought that they harbor an AGN-like central engine with moderate activity. However, some LINERs show no direct evidence for AGNs such as broad emission lines, radio jets, hard X-ray emission, spectral energy distributions which are inconsistent with starlight, and so on. For such LINERs (a subset of type 2 LINERs), we present new poststarburst models which explain some of their most important optical narrow emission-line ratios. In these models, the ionization sources are planetary nebula nuclei (PNNs) with temperature of $\\sim 10^5$ K which appear in the late-phase evolution of intermediate-mass stars with mass between $\\approx 3 M_\\odot$ and $\\approx 6 M_\\odot$. Such PNNs left in a typical starburst nucleus can produce an H$\\alpha$ luminosity of $L({\\rm H}\\alpha) \\sim 10^{38}$ ergs s$^{-1}$ for typical poststarburst LINERs and $\\sim 10^{39}$ ergs s$^{-1}$ only in exceptionally bright cases. The PNN phase lasts until the death of the lowest-mass stars formed in the starburst, which is $\\sim 5 \\times 10^8$ yr for an assumed lower limit of the initial mass function of 3 $M_\\odot$. This long duration appears consistent with the observed higher frequency of occurrence of LINERs if every galaxy could experience the starburst activity several times in its life. We therefore propose that some LINERs which show no direct evidence for AGNs may be poststarburst nuclei powered by a cluster of PNNs. ", "introduction": "In order to investigate the evolution of a star cluster formed in a starburst, it is necessary to know both the initial mass function (IMF) and the star formation rate (SFR). The SFR for nuclear starbursts can be estimated using observed H$\\alpha$ luminosities and so on (Kennicutt 1998). Although it is generally difficult to derive the IMF accurately, it seems reasonable to adopt a power-law form of IMF; $\\phi(m) \\propto m^{-\\mu}$ (e.g., Scalo 1986). In this formulation, there are three free parameters; the power index ($\\mu$), and the upper and lower mass limits of the IMF ($m_u$ and $m_l$). The following values are often adopted for the evolution of the solar neighborhood; $\\mu$ = 1.35, $m_l = 0.1 M_\\odot$, and $m_u = 60 M_\\odot$. On the other hand, summarizing various kinds of observational constraints, Joseph (1991) suggests that $m_u \\approx$ 30 -- 60 $M_\\odot$ and $m_l \\approx$ 3 -- 6 $M_\\odot$ for typical nuclear starbursts. Although the top-heavy IMF is suggested for the nuclear starbursts (i.e., $\\mu$ is smaller than 1.35; e.g., Scalo 1990), we adopt $\\mu$ =1.35, $m_u = 60 M_\\odot$, and $m_l = 3 M_\\odot$ as a modest combination. Another parameter is the duration of the starburst, $\\tau_{\\rm SB}$; we adopt $\\tau_{\\rm SB} = 10^7$ yr because the negative feedback from supernova explosions to star-forming gas clouds is expected to occur $\\sim 10^7$ yr (i.e., the lifetime of B stars) after the onset of the starburst (e.g., Larson 1987). The lifetime of a starburst can also be estimated as a gas consumption timescale, $\\tau_{\\rm SB} \\sim M_{\\rm gas} \\eta_{\\rm SF} / SFR$ where $M_{\\rm gas}$ is the gas mass available for the starburst, $\\eta_{\\rm SF}$ is the star formation efficiency, and $SFR$ is the star formation rate. If we adopt $M_{\\rm gas} \\sim 10^9 M_\\odot$, $\\eta_{\\rm SF} \\sim 0.1$, and $SFR \\sim 10 M_\\odot$ yr$^{-1}$, we obtain $\\tau_{\\rm SB} \\sim 10^7$ y. We will discuss the SFR later. \\begin{figure*}[b] \\plotone{f2.ps} \\caption{% The H$\\alpha$ luminosity evolution of the star cluster for the starburst models. Note that the H$\\alpha$ luminosity is normalized by a unit gas mass, $1 M_\\odot$. The number of Lyman continuum photons which are capable of ionizing He$^0$ (i.e., $h\\nu \\geq$ 24.6 eV) is also plotted as a function of time by dotted lines. This is also normalized by a unit gas mass, $1 M_\\odot$. \\label{fig2}} \\end{figure*} Now let us consider the evolution of the star cluster with the above parameters. Phase I ($0 \\leq t \\leq 10^7$ yr): We assume that the starburst lasts for $10^7$ yr. The star formation rate is assumed to be constant. In this phase, main photoionization sources are most massive stars. Dynamical effect of supernova explosions may be weak in this phase. Phase II ($10^7 \\leq t \\lesssim 2 \\times 10^7$ yr): Since the star formation ceases at a time $t = 10^7$ yr in our model, main photoionization sources change from O stars to B stars as time goes. Wolf-Rayet stars also work in the photoionization (Vacca \\& Conti 1992; Conti 1999). Phase III ($2 \\times 10^7 \\lesssim t \\lesssim 1 \\times 10^8$ yr): Continuous supernova explosions develop a superwind and then shock heating also works in this phase (Heckman, Armus, \\& Miley 1990; Ohyama, Taniguchi, \\& Terlevich 1997; Alonso-Herrero et al.\\ 2000). Phase IV [$t \\simeq (1 {\\em -} 5) \\times 10^8$ yr]: Intermediate-mass stars with mass of several $M_\\odot$ play a role as the photoionization sources. The bolometric luminosity of each star in the main sequence phase is $L_* \\approx 1 \\times 10^3 L_\\odot$. Each star evolves from the main sequence to the asymptotic giant branch (AGB). Following this AGB phase, each star will lose its gaseous envelope and then a hot stellar core appears and ionizes the surrounding gas, making a so-called planetary nebula (Kaler 1985; Vassiliadis, \\& Wood 1994). Since this core (i.e., a planetary-nebula nucleus; hereafter PNN) is so hot (i.e., $T_{\\rm eff} \\gtrsim 10^5$ K), they become to be major photoionization sources. Finally, stars with mass of $\\approx 3 M_\\odot$ die at $t \\simeq 5 \\times 10^8$ yr\\footnote{Note that the ages of intermediate-mass stars are estimated as $5.7 \\times 10^8$ yr, $2.5 \\times 10^8$ yr, and $1.4 \\times 10^8$ yr for stellar masses of $3 M_\\odot$, $4 M_\\odot$, and $5 M_\\odot$, respectively using a relation of $\\tau(m) = 1.2 \\times 10^{10} m^{-2.78}$ yr for $m < 10 M_\\odot$ (Theis, Burkert, \\& Hensler 1992).} and then the effect of the nuclear starburst disappears unless stars less massive than 3 $M_\\odot$ were formed in the starburst. The starburst evolution is summarized schematically in Figure 1. In many previous studies, Phase IV has not been taken into account seriously. However, the temperature of PNNs is hot (i.e., $T_{\\rm eff} \\gtrsim 10^5$ K) enough to ionize the surrounding nebula. More importantly, this higher temperature can be responsible for the formation of partly-ionized regions in which [O {\\sc i}] emission is thought to arise. If the ionization parameter is as low as that for typical LINERs, the optical spectrum of such a nebula is expected to be quite similar to those of LINERs. Note that each PNN has a bolometric luminosity of $L_{\\rm PNN} \\approx 10^4 L_\\odot$ (e.g., Vassiliadis, \\& Wood 1994), being comparable to that of a hot O star (Filippenko, \\& Terlevich 1992; Shields 1992). \\subsection{Poststarburst evolution model} In order to demonstrate the importance of the PNN cluster, we investigate the poststarburst evolution for the star cluster discussed in section 2.1 using luminosity evolution models of Kodama \\& Arimoto (1997). In Figure 2, we show the time evolution of H$\\alpha$ luminosity which is normalized by a unit gas mass, $1 M_\\odot$. Here we assume that stars are formed from the gas with the solar metallicity for simplicity. The H$\\alpha$ luminosity is estimated from the number of Lyman continuum photons; $L({\\rm H}\\alpha) \\simeq 1.36 \\times 10^{-12} N({\\rm Lyc})$ ergs s$^{-1}$ (Leitherer \\& Heckman 1995). Note that the lower mass cutoff is fixed at 0.1 $M_\\odot$ in Kodama \\& Arimoto's models. Therefore, the H$\\alpha$ luminosity shown in Figure 2 is corrected for the case of $m_l = 3 M_\\odot$ by us. According to Kennicutt (1998), the SFR is related to $L({\\rm H}\\alpha)$ as $SFR = 7.9 \\times 10^{-42} L({\\rm H}\\alpha) ~ M_\\odot {\\rm yr}^{-1}$ for the IMF with $\\mu = 1.35$, $m_l = 0.1 M_\\odot$, and $m_u = 100 M_\\odot$. This relation can be replaced by $SFR = 4.0 \\times 10^{-42} L({\\rm H}\\alpha) ~ M_\\odot {\\rm yr}^{-1}$ for the Salpeter IMF with $m_l = 3 M_\\odot$ and $m_u = 60 M_\\odot$. Balzano (1983) found $L({\\rm H}\\alpha) \\simeq 10^{40}$ -- $6 \\times 10^{42}$ ergs s$^{-1}$ for her sample of Markarian starburst nuclei with a Hubble constant of $H_0$ = 75 km s$^{-1}$ Mpc$^{-1}$ (see also Kennicutt, Keel, \\& Blaha 1989; Ho, Filippenko, \\& Sargent 1997b). Since the duration of H$\\alpha$-bright phase is very short (e.g., $\\lesssim 10^7$ yr; see Figure 2) in the starbursts, it seems rarer to detect starbursts with such a bright phase from a statistical point of view. For example, the age of the starburst occurring in Mrk 1259, which is one of the brightest starburst nuclei studied by Balzano (1983), is estimated to be $\\sim 5 \\times 10^6$ yr (Ohyama et al. 1997). Although the observed H$\\alpha$ luminosity of this galaxy is $\\sim 10^{41}$ ergs s$^{-1}$, this galaxy would be observed as a less luminous starburst if it will be observed $10^{7-8}$ yr after now. Typical H$\\alpha$ luminosities listed in Balzano's catalog are $10^{40}$ - $10^{41}$ ergs s$^{-1}$. If the above statistical effect is taken into account, their initial H$\\alpha$ luminosities may be of the order of $10^{41}$ - $10^{42}$ ergs s$^{-1}$. Furthermore, since the mean reddening correction is about a factor of 4-5 for H$\\alpha$ (e.g., Balzano 1983), the reddening-corrected, initial H$\\alpha$ luminosities seem to be as luminous as $\\sim 10^{42}$ ergs s$^{-1}$ on average and some brightest nuclear starbursts may have H$\\alpha$ luminosities of $\\sim 10^{43}$ ergs s$^{-1}$. These values lead to $SFR \\sim$ 10 -- 100 $M_\\odot$ yr$^{-1}$. Then, we estimate $M_{\\rm gas} \\sim 10^{8 - 9} M_\\odot$ for the nuclear starbursts. Indeed, the molecular gas mass in circumnuclear regions of typical starburst nuclei is $M_{\\rm gas} \\sim 10^9 M_\\odot$ (Devereux et al.\\ 1994). As shown in Figure 2, the H$\\alpha$ luminosity density per unit mass at $t \\sim 10^8$ yr is $L({\\rm H}\\alpha)/M_{\\rm gas} \\sim 10^{30}$ ergs s$^{-1}$ $M_\\odot^{-1}$. Therefore, we estimate typical H$\\alpha$ luminosities of poststarburst nuclei with $t \\sim 10^8$ yr as $L({\\rm H}\\alpha) \\sim 10^{38}$ ergs s$^{-1}$ for a typical LINER powered by PNNs, and $\\sim 10^{39}$ ergs s$^{-1}$ in exceptionally bright cases. It should be mentioned that our poststarburst models cannot be applied to on-going starburst galaxies because most starburst nuclei have $L({\\rm H}\\alpha) \\gtrsim 10^{39}$ ergs s$^{-1}$ (Balzano 1983; Kennicutt et al. 1989; Ho et al. 1997b). Namely, if a nuclear starburst occurs, the photoionization should be dominated by massive stars in the starburst even if there is a cluster of PNNs left from the recent past nuclear starburst. \\begin{figure*}[t] \\plotone{f3.ps} \\caption{% The SED evolution of the star cluster adopted in our study for $t$ = 0.01, 0.05, 0.1, 0.2, and 0.5 Gyr. \\label{fig3}} \\end{figure*} \\subsection{Optical emission-line properties of the ionized nebula} Next we investigate optical emission-line properties of the gaseous nebula photoionized by the PNN cluster. We use the photoionization code CLOUDY version 90.05 (Ferland 1996), which solves the equations of statistical and thermal equilibrium and produces a self-consistent model of the run of temperature as a function of depth into the nebula. Here we assume that a uniform-density gas cloud with plane-parallel geometry is irradiated by the star cluster formed in the starburst. The parameters for the calculations are 1) the hydrogen density of the cloud ($n_{\\rm H}$), 2) the ionization parameter (Osterbrock 1989); $U = Q({\\rm H^0}) (4 \\pi r^2 n_{\\rm H} c)^{-1}$ where $Q({\\rm H^0})$ is the number of ionizing photons, $r$ is the distance from the ionizing source, and $c$ is the light velocity, 3) the spectral energy distribution (SED) of the ionizing radiation, and 4) the chemical compositions. The SED of the ionizing radiation is calculated during the course of the starburst evolution using Kodama \\& Arimoto's models. The SED evolution is shown in Figure 3. Note that the ordinate is the flux in units of $L_\\odot$ \\AA$^{-1}$, corresponding to the case that the total gas mass is $1 M_\\odot$. We perform our photoionization calculations as a function of the ionization parameter between log $U = -4$ and log $U = -3$ with a logarithmic interval of 0.2. Following Filippenko \\& Terlevich (1992), we adopt $n_{\\rm H} = 10^3$ cm$^{-3}$. In order to examine the effect of dust grains on emission-line intensity ratios, we perform photoionization calculations for the following two cases; a) the dust-free gas with the solar abundances, and b) the gas with the Orion Nebula abundances where effects of dust grains are taken into account. For both the cases we simply adopted the abundance sets available in CLOUDY. The solar composition is taken from Grevesse \\& Anders (1989) and Grevesse \\& Noels (1993). The Orion Nebula abundances are a subjective mean of the abundances determined by Baldwin et al. (1991), Rubin et al. (1991), and Osterbrock, Tran, \\& Veilleux (1992). The grains are the large-R grains described by Baldwin et al. (1991). The calculations were stopped when the temperature fell to 3000 K, below which little optical emission is expected. The results are shown in Figures 4 and 5 for the above two cases, respectively. Here we use the same emission-line diagnostic diagrams as those used in Filippenko \\& Terlevich (1992). It is shown that our poststarburst models with age of $t \\simeq (1$ -- $5)\\times 10^8$ yr appear consistent with the observed emission-line properties of LINERs; the depletion of metals into dust grains leads to higher relative intensities of low-ionization lines such as [O {\\sc i}] and [S {\\sc ii}], being more consistent with the observations. However, it should be noted that the [O {\\sc i}]/H$\\alpha$ ratio barely touches the LINER region of the diagnostic diagram in both Figure 4 and 5. In this respect, our models are similar to the previous O-star models for weak [O {\\sc i}] LINERs proposed by Filippenko \\& Terlevich (1992) and Shields (1992); i.e., the weak [O {\\sc i}] LINERs are defined as LINERs with $I$([O {\\sc i}]$\\lambda$6300)/$I$(H$\\alpha$) $\\leq 1/6$. However, since PNNs are hot enough to supply such high-energy photons, the other emission-line ratios appear more consistent with the observations (see Figures 4 and 5) than those of the O-star models. The weakness of [O {\\sc i}] emission in our models may be attributed to the fact that the production rate of high-energy photons is lower than that of the non-thermal continuum radiation from an AGN. \\begin{figure*} \\epsscale{1.2} \\plotone{f4.ps} \\caption{% Optical emission-line diagnostics used to classify different types of emission-line galaxies adopted in Filippenko \\& Terlevich (1992). Typical loci of ratio combinations used in Filippenko \\& Terlevich (1992) are reproduced for normal H {\\sc ii} regions, NLRs of typical AGNs such as Seyfert 2 galaxies, and classical LINERs. Results of poststarburst photoionization models for $t$ = 0.01, 0.05, 0.1, 0.2, and 0.5 Gyr are marked as line segments joining the points between log $U = -4$ and log $U = -3$. \\label{fig4}} \\end{figure*} \\begin{figure*} \\epsscale{1.2} \\plotone{f5.ps} \\caption{% The same diagnostic diagrams as those in Figure 4, but for the Orion Nebula abundances. \\label{fig5}} \\end{figure*} \\begin{figure*}[b] \\epsscale{0.8} \\plotone{f6.ps} \\caption{% The evolutions of equivalent width of both H$\\beta$ emission and absorption lines are shown for the three cases; 100 percent, 50 percent, or 10 percent of the blue continuum emission arises from stars formed in the starburst. Gaseous H$\\beta$ emission, stellar H$\\beta$ absorption, and their sum total are shown by dashed lines, dotted lines, and thick lines, respectively, for the three cases. Here in order to estimate the equivalent width of H$\\beta$ absorption, we have used another spectral synthesis code GISSEL96 (Bruzual 1993; Leitherer et al.\\ 1996) because Kodama \\& Arimoto's models do not give it. Since the two codes use both the same stellar evolution models and stellar atmosphere models, there is no problem in the estimate. The equivalent width is defined as $EW = \\int {{F_{\\rm line}(\\lambda) - F_{\\rm cont}(\\lambda)} \\over F_{\\rm cont}(\\lambda)} d\\lambda$ \\AA. \\label{fig6}} \\end{figure*} In order to demonstrate this quantitatively, we show the time evolution of the number of ionizing photons which are capable of ionizing He$^0$ (i.e., $h\\nu \\geq$ 24.6 eV) in Figure 2 (see the dotted line). Although high-energy photons are supplied by O stars in Phase I, they decrease in number significantly in Phase III. However, they increase in number again in Phase IV because of the supply from high-temperature PNNs. This makes our models more consistent with the observations than the previous O-star photoionization models. It is known that intermediate-mass stars produce strong Balmer absorption lines in their stellar atmospheres and thus the Balmer emission lines arising from the ionized nebulae are sometimes quenched by the Balmer absorption. In order to investigate this effect quantitatively along the starburst evolution, we show time variations of the equivalent width ($EW$) of both Balmer emission and absorption for the case of H$\\beta$ in Figure 6. The H$\\beta$ emission dominates in the first a few times $10^7$ yr while the H$\\beta$ absorption does after this age. This comparison suggests that LINERs whose Balmer lines are observed as absorption are more popular by about one order of magnitude than those with the Balmer emission. On the other hand, observations show that the fraction of the former LINERs is significantly smaller than that of the latter ones (Ho et al.\\ 1997a). However, if the Balmer emission has a narrower line width on average than the Balmer absorption, the Balmer emission can be seen even if the $EW_{\\rm absorption} > EW_{\\rm emission}$. This trend is actually seen in the observed spectra of many LINERs (Ho et al.\\ 1997a). ", "conclusions": "We have presented a new poststarburst model of LINERs. In this model, the ionization sources are planetary nebula nuclei (PNNs). Our main point is that the ionization sources are planetary nebula nuclei (PNNs) with temperature of $\\sim 10^5$ K which appear in the late-phase evolution of intermediate-mass stars with mass between $\\approx 3 M_\\odot$ and $\\approx 6 M_\\odot$. Our models are able to reproduce the observed optical narrow emission-line ratios of LINERs although the [O {\\sc i}] emission is underpredicted to some extent. We give a summary of the limitations of our models. 1) Our models cannot be applied to LINERs associated with nuclei of elliptical galaxies and most of S0 galaxies in which nuclear starbursts occur seldom. 2) Our models cannot be applied to LINERs with direct evidence for AGN such as broad-line emission, radio jets, hard X-ray emission, and so on (mostly type 1 LINERs). 3) Our models cannot be applied to on-going starburst nuclei because the photoionization is dominated by massive stars in the starburst rather than a cluster of PNNs left from a recent past starburst. 4) Accordingly, our models are applied to a subset of type 2 LINERs located in nuclear regions of spiral galaxies. Since the H$\\alpha$ luminosity in the poststarburst LINER phase is less luminous by $\\sim$ 4 orders of magnitude than the initial H$\\alpha$ luminosity at the onset of the starburst, our models are preferentially applied to low-luminosity LINERs; e.g., $L$(H$\\alpha$) $\\sim 10^{38}$ ergs s$^{-1}$. Our models are constructed to explain the observed optical narrow emission-line ratios and we have not examined properties at other wavelengths (e.g., radio continuum, hard X rays, and so on). However, since the ionization sources are PNNs, poststarburst LINERs will not show any evidence for the presence of AGNs such as radio jets, hard X-ray emission and so on. As shown in Figure 3, our models predict that the X-ray emission is much weaker than that of typical AGNs, being consistent with some type 2 LINERs (Terashima et al. 2000 and references therein; see also for AGN-like SEDs of LINERs, Ho 1999). One interesting prediction of our models is that the ionization sources may be spatially extended because massive stars in nuclear starbursts are often distributed within central 100-pc regions (Meurer et al. 1995; see also Sugai \\& Taniguchi 1992). Recently, Pogge et al. (2000) have shown that some LINERs have spatially-extended emission-line regions with sizes of tens to hundreds parsecs. However, since they have either a compact UV source or no bright UV source, they may be not poststarburst LINERs but genuine low-ionization AGNs. In conclusion, in order to examine how many poststarburst LINERs are really present, we need systematic investigations of radio-continuum and X-ray properties of low-luminosity LINERs as well as UV continuum imaging in future. \\vspace{0.5cm} We would like to thank an anonymous referee for many useful comments and suggestions. YS and TM are supported by JSPS. This work was financially supported in part by Grant-in-Aids for the Scientific Research (Nos. 10044052, and 10304013) of the Japanese Ministry of Education, Culture, Sports, and Science." }, "0005/astro-ph0005164_arXiv.txt": { "abstract": "A complete flat-spectrum radio-loud sample of AGN includes a significant fraction of Seyfert-like AGN including a NLS1. Analysis of their optical spectra suggests that the reddest continuum colours are either associated with AGN in nearby resolved galaxies, or distant quasars showing relatively narrow permitted emission lines. ", "introduction": "The relationship between Seyfert galaxies and radio-loud galaxies has not been explored in detail. However, a complete sample of flat-spectrum radio-loud AGN contains a significant fraction of Seyfert-like sources, including one which would be classified as a narrow-line Seyfert 1 (NLS1). To understand the physical mechanisms responsible for producing the different characteristics of AGN, a multiwavelength approach is needed. Correlations between properties at different wavelengths can be used to reject and refine physical models of the central regions of AGN. The Parkes Half-Jansky Flat-Spectrum Sample (PHFS) \\cite{D1} is interesting in this context because it is a radio-selected sample. The selection criteria are as follows: \\begin{itemize} \\item Radio-loud: 2.7GHz flux $>$ 0.5 Jy. \\item Flat-spectrum: $\\alpha_{2.7/5.0} > -0.5$, where $S_{\\nu}=\\nu^{\\alpha}$ \\item Galactic latitude: $|$b$| >20^{\\circ}$ \\item $-45^{\\circ} <$ Dec(B1950) $< +10^{\\circ}$ \\end{itemize} This sample contains 323 sources with a wide range of properties which can be quite different to AGN selected by optical colours. The PHFS AGN have a large range in optical luminosities, with more than 50 objects having absolute magnitudes in the Seyfert luminosity range (M$_{B}>$-23). Another interesting characteristic is the large dispersion in optical colours of this sample compared with optically selected samples such as the Large Bright Quasar Survey \\cite{W1,F1}. ", "conclusions": "" }, "0005/astro-ph0005487_arXiv.txt": { "abstract": "\\xone\\ and \\xthree\\ are the only known persistent stellar-mass black hole candidates that have almost always shown spectra that are dominated by a soft, thermal component. We present here results from 170\\,ksec long Rossi \\mbox{X-ray} Timing Explorer (\\rxte) observations of these objects, taken in 1996~December, where their spectra can be described by a disc black body plus an additional soft ($\\Gamma\\sim 2.8$) high-energy power-law (detected up to energies of 50\\,keV in \\xthree). These observations, as well as archival Advanced Satellite for Cosmology and Astrophysics (\\asca) observations, constrain any narrow Fe line present in the spectra to have an equivalent width $\\aproxlt 90$\\,eV. Stronger, broad lines ($\\approx 150$\\,eV EW, $\\sigma \\approx 1$\\,keV) are permitted. We also study the variability of \\xone. Its \\mbox{X-ray} power spectral density (PSD) is approximately $\\propto f^{-1}$ between $10^{-3}$ and $0.3$\\,Hz with a root mean square (rms) variability of $\\approx 7\\%$. At energies $>5$\\,keV the PSD shows evidence of a break at $f > 0.2$\\,Hz, possibly indicating an outer disc radius of $\\aproxlt 1000~GM/c^2$ in this likely wind-fed system. Furthermore, the coherence function $\\gamma^2(f)$, a measure of the degree of linear correlation, between variability in the $> 5$\\,keV band and variablity in the lower energy bands is extremely low ($\\aproxlt 50\\%$). We discuss the implications of these observations for the mechanisms that might be producing the soft and hard X-rays in these systems. ", "introduction": "\\label{sec:intro} Since the discovery of Cygnus~X-1 in 1964 \\cite{bowyer:65a}, the study of galactic black hole candidates (BHCs) has shown that these objects display a large variety of states which are characterized by their distinct spectral shapes and temporal behaviours. The most important states which have been identified are the ``low/hard state'', which is characterized by a hard X-ray spectrum with a photon index $\\Gamma=1.7$ and large root mean square (rms) variability $\\aproxgt 30\\%$ (Tanaka \\& Lewin 1995; Nowak 1995\\nocite{tanaka:95a,nowak:95a}; and references therein), and the ``high/soft state'', which is spectrally softer ($\\Gamma \\sim 2.5$) and exhibits less variability. The fractional Eddington luminosity of sources in the soft state tends to be higher than that for sources in the hard state (Nowak 1995\\nocite{nowak:95a}, and references therein). The soft state has been observed in steady sources such as \\xone\\ and \\xthree\\ \\cite{treves:88a,ebisawa:89a,treves:90a,ebisawa:93a,schmidtke:99a}, in recurring transients such as GX~339$-$4 \\cite{grebenev:93a}, and in a number of transients such as Nova Muscae \\cite{miyamoto:94a}. The persistent BHC Cygnus~X-1 has been observed to switch between the hard and the soft state (albeit with $kT\\sim 0.3$\\,keV for the soft state), with the total X-ray luminosity staying roughly constant \\cite{cui:97b,zhang:97b}. A great deal of observational attention has been focused on the more commonly observed hard state, since most of the brighter galactic BHCs only occasionally transit to the soft state. Only two of the persistent nearby BHCs, \\xone\\ and \\xthree, until recently have always been observed in the soft state. Wilms et al. \\shortcite{wilms:99b}, hereafter Paper II, present evidence that \\xthree\\ periodically transits into the low/hard state. In this work we present 170\\,ksec long Rossi \\mbox{X-ray} Timing Explorer (\\rxte) observations of both \\xone\\ and \\xthree\\ during high/soft X-ray states. \\xthree\\ is a highly variable BHC with a probable 9\\,${\\rm M}_{\\odot}$ compact object mass \\cite{cowley:83a}. Its luminosity has been observed to be as high as $4\\times 10^{38}\\,{\\rm erg~s^{-1}}$, which is $\\approx 30\\%$ of its Eddington luminosity, and its soft X-ray flux ($\\sim 1$--$9$\\,keV) is variable by a factor of more than four on long time scales (see Paper II). \\xthree\\ also has exhibited a strong 99- or 198-day periodicity in its soft X-ray flux \\cite{cowley:91a,cowley:94a}, which is also apparent in the \\asm\\ monitoring, albeit with a period that varies over time (Paper II). Previously we had suggested that this periodicity might be associated with a warped, precessing accretion disc \\cite{wilms:99a}; however, with the recently observed ``state changes'' to a low/hard flux it now seems likely that a systematic variation of the accretion rate is an important part of this long-term variability. \\xone\\ is also a good candidate for a black hole. Using a large number of \\textsl{ROSAT} HRI observations, Cowley et al. \\shortcite{cowley:95a} were able to identify the counterpart with ``star number 32'' of Cowley et al. \\shortcite{cowley:78a}. This object has a mass function of only $f=0.144\\,\\Msun$, but including other evidence the mass of the compact object appears to be $M>4\\,\\Msun$ \\cite{hutchings:87a}, and probably $\\aproxgt 6\\,\\Msun$ \\cite{cowley:95a}. The luminosity of the object is typically about $2\\times 10^{38}\\,{\\rm erg~s^{-1}}$ \\cite{long:81a}. Although small differences between \\xone\\ and \\xthree\\ are apparent, their spectra and short-time temporal behaviours have historically been quite similar. \\xone, however, does not exhibit any obvious periodic behaviour in its long term X-ray lightcurve. We have monitored \\xone\\ and \\xthree\\ with \\rxte\\ in three to four weekly intervals since the end of 1996 in order to enable a systematic study of the soft state. The campaign started with 170\\,ksec long observations of both sources, and in this paper we present results from the spectral and temporal analysis of these long observations. In addition, we consider spectral results of archival Advanced Satellite for Cosmology and Astrophysics (\\asca) observations. Preliminary results of our analyses have already appeared elsewhere \\cite{wilms:99a,wilms:99c}. Results from the monitoring observations are presented in Paper II. The remainder of this paper is structured as follows. We start with a description of our \\rxte\\ data analysis methodology (\\S\\ref{sec:rxte}). We then present the results from the spectral analysis from \\rxte\\ (\\S\\ref{sec:spectra}) and \\asca\\ (\\S\\ref{sec:asca}). The \\rxte\\ timing analysis is discussed in section~\\ref{sec:timing}. We discuss our results in the context of current physical models for the soft state (\\S\\ref{sec:discuss}) and then summarize the paper (\\S\\ref{sec:summary}). ", "conclusions": "" }, "0005/astro-ph0005214_arXiv.txt": { "abstract": "We report on the discovery of \\psr, a radio pulsar in an eccentric, relativistic 5-hr binary orbit. The pulsar shows no evidence for being recycled, having pulse period $P = 394$~ms, characteristic age $\\tau_c = 1.4 \\times 10^6$~yr, and inferred surface magnetic dipole field strength $B = 1.3 \\times 10^{12}$~G. From the mass function and measured rate of periastron advance, we determine the total mass in the system to be (2.300 $\\pm$ 0.012)~\\sm, assuming that the periastron advance is purely relativistic. Under the same assumption, we constrain the pulsar's mass to be $M_p \\leq 1.348$~\\sm\\ and the companion's mass to be $M_c \\geq 0.968$~\\sm\\ (both 99\\% confidence). Given the total system mass and the distribution of measured neutron star masses, the companion is probably a massive white dwarf which formed prior to the birth of the pulsar. Optical observations can test this hypothesis. ", "introduction": "Relativistic binary pulsars have been celebrated as being unique laboratories for high-precision tests of general relativity and accurate determinations of neutron star masses (\\cite{tw89,twdw92,sac+98}). To date, post-Keplerian general relativistic parameters have been measured for nine binary pulsar systems (see reviews by \\cite{twdw92,tc99,kas99}). Five are double neutron star binaries (PSRs B1913+16, B1534+12, B2127+11C, J1518+4904, J1811$-$1736) and four (PSRs J1713+0747, B1802$-$07, B1855+09, B2303+46) have white dwarf companions. Through radio timing observations of PSR~B1913+16, general relativity has been confirmed at the $\\sim$0.3\\% level (\\cite{tay92}). Timing observations of these pulsars show that the neutron star masses fall in a narrow range centered on 1.35~\\sm\\ (\\cite{tc99}). The lack of variation of these masses is surprising given the variety of binary evolution mechanisms by which they were formed. Of the above relativistic binary systems, all but one are likely to share a similar evolutionary history. The observed pulsar in most of these systems has a large characteristic age $\\tau_c$ and low surface magnetic field $B$ relative to the bulk of the isolated pulsar population. This is because the observed neutron star, the first-born in the binary, was ``recycled'' by a phase of mass transfer from its companion as the latter ascended the giant branch (see Bhattacharya \\& van den Heuvel 1991 \\nocite{bv91} for details of neutron-star binary evolution). This results in spin up of the neutron star and the reduction of its magnetic field. The binary observed today has survived the supernova explosion of the secondary (although in the cases of PSRs B2127+11C and B1802$-$07, which are in globular clusters, the possibility of a more complicated evolutionary history cannot be discounted). The exception among the above binaries is PSR~B2303+46 which shows no evidence for having been recycled, having $B$ and $\\tau_c$ comparable to the bulk of the isolated pulsar population. PSR~B2303+46, long thought to be part of a double neutron-star binary, has recently been shown to have a white-dwarf companion (\\cite{vk99}) which must have formed {\\it prior} to the pulsar's birth. Formation mechanisms for such a system have been suggested by Portegies Zwart \\& Yungelson (1999) and Tauris \\& Sennels (2000) and involve an initial primary, the white dwarf progenitor, which transferred sufficient matter onto the initial secondary that the latter underwent core collapse. We report here on \\psr, a relativistic binary pulsar recently discovered as part of the Parkes Multibeam Pulsar survey. A preliminary report on this system was presented by Manchester et al. (2000).\\nocite{mlc+00} ", "conclusions": "\\label{sec:disc} \\psr\\ is different from most of the other known relativistic binary pulsars in several important respects. The pulsar's characteristic age, $\\tau_c = 1.4 \\times 10^{6}$~yr, and inferred surface magnetic field strength, $B = 1.3 \\times 10^{12}$~G, are similar to those of the bulk of the isolated pulsar population. Thus, \\psr\\ is unlikely to have ever been recycled. Like the other short-$P_b$, eccentric relativistic binaries, it may be a double neutron star binary, but if so, we must be observing the second-formed neutron star in its short-lived radio pulsar phase. In this case, its companion could still be an observable radio pulsar that was not detected in our Parkes observations, either because its short spin period varies too rapidly due to the binary orbit (a possibility we are checking, given our knowledge of the binary orbit) or because its radio beam does not intersect our line of sight. However, we show here that the companion is unlikely to be a second neutron star, given the system's mass function, its rate of precession $\\dot{\\omega}$, and what is known about the neutron star mass distribution. The mass function is given by \\begin{equation} f(M_p) = \\frac{4 \\pi^2 (a \\sin i)^3}{G P_b^2} = \\frac{M_c^3 \\sin^3 i}{(M_p + M_c)^2}, \\end{equation} where $M_p$ and $M_c$ are the pulsar and companion masses, respectively, $P_b$ is the orbital period, and $a \\sin i $ is the projected semi-major axis of the pulsar orbit. This provides a constraint on the minimum $M_c$ for any assumed $M_p$, by setting $i$, the inclination of the orbital angular momentum with respect to the line-of-sight, to $90^{\\circ}$. In addition, $\\dot{\\omega}$, under the assumption that it is due to general relativistic periastron precession (but see \\S\\ref{sec:classom}), is given by \\begin{equation} \\dot{\\omega}_{GR} = 3 \\left( \\frac{P_b}{2 \\pi} \\right)^{-5/3} (T_{\\odot} M)^{2/3} (1 - e^2)^{-1}, \\end{equation} where $T_{\\odot} = 4.925490947 \\times 10^{-6}$~s. This then determines the total system mass $M = M_p + M_c$. For $\\dot{\\omega} = (5.32 \\pm 0.02)^{\\circ}$yr$^{-1}$ (Table~\\ref{ta:parms}), $M = (2.300 \\pm 0.012)$~\\sm. Figure~\\ref{fig:mass} shows these constraints in $M_p - M_c$ phase space. The shaded region is ruled out by the mass function. The intersection of the $\\dot{\\omega}$ straight line with the boundary of this region determines the maximum allowed pulsar mass: $M_p < 1.331$~\\sm\\ ($1\\sigma$) or $M_p < 1.348$~\\sm\\ ($3\\sigma$). Similarly, we set a lower limit $M_c > 0.974$~\\sm\\ ($1\\sigma$) or $M_c > 0.968$~\\sm\\ ($3\\sigma$). Thorsett \\& Chakrabarty (1999) \\nocite{tc99} showed that measurements of neutron star masses are consistent with all being drawn from a Gaussian distribution having mean and standard deviation 1.35 and 0.04~\\sm, respectively. The parameters in the \\psr\\ system are thus interesting: if $M_p$ is close to its maximum allowed value corresponding to $i \\simeq 90^{\\circ}$, as would be consistent with other neutron star masses, then its companion has mass much lower than those of all known neutron stars, and therefore is unlikely to be one. Of course on {\\it a posteriori} statistical grounds, $i \\simeq 90^{\\circ}$ is improbable. On the other hand, a smaller $i$ implies a less massive pulsar. For the median value $i = 60^{\\circ}$, $M_p = 1.17$~\\sm, and $M_c = 1.13$~\\sm, both significantly lower than for any other known neutron stars. Note that none of these values can be ruled out based on neutron star stability arguments, since masses of as little as $\\sim$0.1~\\sm\\ are allowed (e.g. \\cite{cst93}). However the formation mechanism for such low-mass neutron stars is unclear. \\subsection{A Neutron Star/White Dwarf Binary?} A more likely possibility is that the \\psr\\ system is a neutron star/CO white-dwarf binary seen edge-on. (ONeMg white dwarfs have minimum mass $\\sim$1.1~\\sm [Wanajo, Hashimoto \\& Nomoto 1999] \\nocite{whn99} so this possibility cannot be ruled out.) The evolutionary history of the \\psr\\ binary, if the companion is a massive white dwarf, is then clear (see e.g. \\cite{dc87}, \\cite{py99}, \\cite{ts00}): the system originated as a binary consisting of two main sequence stars having mass ratio near unity, but with neither sufficiently massive to independently form a neutron star. For example, the primary may have had mass 7~\\sm\\ and the secondary 5~\\sm. The primary evolved first to form the white dwarf, in the process transferring sufficient matter onto the secondary for its mass to exceed that necessary to form a neutron star. After the white dwarf formed, it spiraled into the envelope of the now more massive secondary as the latter ascended the giant branch, greatly decreasing the orbital period, and ejecting the common envelope. The secondary, a helium star after the ejection of the envelope, then exploded in a supernova, which fortuitously did not disrupt the binary. The radio pulsar we see is therefore the second evolved star. The evolutionary scenario in which a massive white dwarf forms prior to the neutron star has been considered in population synthesis studies. Dewey \\& Cordes (1987) \\nocite{dc87} and Portegies Zwart \\& Yungelson (1999) \\nocite{py99} showed that the birth rate of white dwarf/young pulsar binaries is comparable to that of double neutron star binaries. However, Portegies Zwart \\& Yungelson (1999) argued that the white dwarf should have mass $\\gapp 1.1$~\\sm, a result of their assumption that a neutron-star forming binary had to have an initial primary mass of $>7$~\\sm. To form a $\\sim$1~\\sm\\ white dwarf, an initial minimum primary mass of $\\sim$6~\\sm\\ is required (S. Portegies Zwart, private communication). The work of Tauris \\& Sennels (2000) \\nocite{ts00} confirms that white dwarf/young neutron star systems should be observable, but they suggest that the birth rate of such systems is much higher than that of double neutron star systems. A lower allowed initial primary mass for forming white dwarfs in the Portegies Zwart \\& Yungelson (1999) analysis could reduce the disagreement. One prediction of the binary evolution theory is that the space velocity of \\psr\\ will be greater than $\\sim$150~km~s$^{-1}$ (\\cite{ts00}). At a distance of 3.2~kpc, this implies a proper motion of $> 10$~mas~yr$^{-1}$, which could be measurable by timing on a time scale of a few years. VLBI observations may also be able to detect it. In fact, given the pulsar's timing age, Galactic latitude, distance estimate, and assuming that the pulsar was born in the Galactic plane, the component of the pulsar's space velocity perpendicular to the plane must be $\\sim150(d/3.2 \\; {\\rm kpc})$~km~s$^{-1}$, already roughly consistent with the prediction. If the companion is indeed a white dwarf, allowing for interstellar reddening, its B magnitude is likely to be in the range 25 -- 26.5 and R in the range 24.5 -- 26. From a Digital Sky Survey image of the field, it is clear that the field is relatively crowded, but not impossibly so. Excellent seeing conditions on the VLT would allow detection of the companion. If detected, then a follow-up temperature measurement using HST may provide a useful age constraint. \\subsection{Classical Contribution to $\\dot{\\omega}$?} \\label{sec:classom} The observed $\\dot{\\omega}$ may not be purely relativistic, in which case the above conclusions would have to be modified. Classical contributions to $\\dot{\\omega}$ can come either from tidal deformation of the companion by the neutron star (significant only if the companion is non-degenerate), or from a rotation-induced quadrupole moment in the companion, possibly relevant if it is a rapidly rotating white dwarf. We consider these possibilities in turn. A tide raised on the companion star gives it a quadrupole moment that results in a classical apsidal advance, $\\dot{\\omega}_{tide}$, given by \\begin{equation} \\dot{\\omega}_{tide} = 3.44 \\times 10^6 k_2 \\left(\\frac{M_p}{M_c}\\right) \\left( \\frac{M}{M_{\\odot}}\\right)^{-5/3} \\left( \\frac{R_c}{R_{\\odot}} \\right)^5 \\;\\; ^{\\circ}{\\rm yr}^{-1}, \\end{equation} where $R_c$ is the companion radius and $k_2$ its apsidal constant, a measure of its internal density distribution (\\cite{rma76,sb76}). A hydrogen main sequence star companion could in principle fit in the orbit, but would result in a classical $\\dot{\\omega}$ $\\sim$1000 times the relativistic value (see \\cite{mr75}) so is certainly ruled out. However, a helium main sequence star has much smaller radius, and would result in a classical $\\dot{\\omega}$ that is comparable to the expected relativistic value. We have explored this possibility by considering a range of values for $M_p$ and $M_c$, and finding the expected $\\dot{\\omega} = \\dot{\\omega}_{tide} + \\dot{\\omega}_{GR}$ that match the observed value. For these calculations, we assumed $k_2(R_c/R_{\\odot})^5 = 4 \\times 10^{-6}(M_c/{\\rm M}_{\\odot})^{4.59}$, as derived by Roberts et al. (1976). In Figure~\\ref{fig:mass}, the dashed line shows the allowed locus in $M_p - M_c$ space for a non- or slowly rotating helium star companion. Clearly this is possible only if $M_p \\leq 0.75$~\\sm. This seems unlikely given observed masses of other neutron stars. A rapidly rotating white dwarf would have a quadrupole moment that could result in a significant classical $\\dot{\\omega}$ (Roberts et al. 1976, \\cite{sb76}). However unlike the tidal quadrupole, the rotationally-induced quadrupole lies in a plane perpendicular to the white dwarf spin axis. Since the pulsar formed after the white dwarf, it is unlikely that the white dwarf spin axis is aligned with the orbital angular momentum, as this would demand a fortuitously symmetric supernova explosion. A misalignment of the angular momenta results in classical spin-orbit coupling, that is, a precession of the orbital plane, $di/dt$ as in the pulsar/B-star binary PSR~J0045$-$7319 (\\cite{lbk95,kbm+96}). This precession would manifest itself as a time-variable projected semi-major axis $x \\equiv a \\sin i$. In general, the magnitude of $di/dt$ is comparable to that of the periastron precession $\\dot{\\omega}$ (\\cite{wex98}). For \\psr, we find an upper limit on $\\dot{x}$ that is much smaller than $\\dot{\\omega}$ (Table~\\ref{ta:parms}). This might suggest that spin-orbit coupling is unlikely to be occurring. However, $\\dot{x}$ varies as $\\cos i$; since $i \\simeq 90^{\\circ}$ (especially if $\\dot{\\omega}_{GR}$ is smaller than our observed value), we cannot rule out orbital plane precession, i.e. a large $di/dt$. The above reasoning also holds for a helium-star companion that is rotating rapidly. In this case, the dashed line in Figure~\\ref{fig:mass} is inappropriate as it assumes no contribution from a rotation-induced quadrupole moment. Hence, we cannot presently rule out this possibility. However, a 1-\\sm\\ helium main sequence star would be considerably brighter than a white dwarf: assuming it is on the helium main sequence, it would have bolometric luminosity $\\log (L/L_{\\odot}) = 2.4$ and $T_{eff} = 50,000$~K (\\cite{kw90a}). Using bolometric corrections from Bessell, Castelli \\& Plez (1998) \\nocite{bcp98} and given the distance and expected reddening, a helium star should have $B \\simeq 17$, and should be easily distinguishable from a white dwarf. \\subsection{Future Relativistic Observables} \\label{sec:relobs} The prospects for a high-precision determination of both component masses in the \\psr\\ binary system are excellent. Our upper limit on the combined time dilation and gravitational redshift parameter $\\gamma$ is shown in Table~\\ref{ta:parms}. Note that this is largely independent of which model for the system is correct. Given that the fractional uncertainty in $\\gamma$ scales as $T^{-3/2}$, where $T$ is the observing span (\\cite{dt92}), we expect a 3$\\sigma$ detection by 2001, assuming $\\dot{\\omega}$ is purely relativistic (for which $\\gamma \\simeq 0.7$~ms for $M_p = 1.32$~\\sm). Furthermore, a general relativistic orbital period derivative, $\\dot{P_b}$, should be measurable with interesting precision by 2004, given that its fractional uncertainty scales as $T^{-5/2}$ and its expected value is $-3.5 \\times 10^{-13}$. Thus, continued high-precision timing observations of \\psr\\ will eventually test our assumptions about a purely relativistic $\\dot{\\omega}$. Although the orbit is likely to be highly inclined, measurement of Shapiro delay will be difficult: the maximum expected delay is only $\\sim$5~$\\mu$s, well below our current timing precision. Although \\psr\\ will coalesce in $\\sim$1.5~Gyr due to gravitational-radiation-induced orbit decay, it, and coalescing neutron-star/white-dwarf binaries in general, will not be detected by LIGO. The detectability depends on the frequency of the emitted radiation near the time of coalescence. This depends on the size of the orbit at coalescence, which, to order of magnitude, is the white dwarf radius, $\\sim 10^9$~cm. This implies an orbital frequency of $\\sim$0.25~Hz at coalescence; the implied gravitational wave frequency, twice the orbital frequency, is well outside the LIGO band of 10--10,000~Hz. However, such signals would be within the 0.00001--1~Hz band of the proposed Laser Interferometric Space Antenna (LISA; \\cite{ben98}). General relativistic geodetic precession (\\cite{bo75}) is predicted to result in the precession of the pulsar's spin axis at a rate of $\\sim 1.7^{\\circ}$yr$^{-1}$ (assuming a purely relativistic $\\dot{\\omega}$). This is greater than that predicted for the two pulsars for which it has been observed, PSR~B1913+16 (\\cite{wrt89}; \\cite{kra98}) and PSR~B1534+12 (\\cite{arz95,stta00}), although the measured amplitude will be suppressed by a possibly large factor that depends on the unknown line-of-sight geometry and spin-orbit misalignment. Given the narrow pulse profile seen in \\psr\\ (Fig.~\\ref{fig:grand}, FWHM of 0.011$P$), this effect could be detected in a few years, barring unfortunate geometries. Thus far, we have detected no strong evidence for pulse profile changes, although a detailed analysis is beyond the scope of this paper. That the pulsar was not detected in past pulsar radio surveys of the region is intriguing. The Johnston et al. (1992) \\nocite{jlm+92} 20-cm survey of the Galactic plane had minimum detectable flux density of $\\sim$1~mJy, below the pulsar's 20-cm flux density of 3.3~mJy. At this frequency, the pulsar displays no evidence for scintillation. The Parkes all-sky survey at 70~cm may also have been able to detect it. That survey's sensitivity reached 20~mJy for Galactic plane sources; extrapolating the flux densities in Table~\\ref{ta:parms}, the estimated flux density at 70~cm is $\\sim$40~mJy. Scattering for the pulsar at that frequency is negligible. Thus, it is plausible that the pulsar's beam has only recently precessed into our line-of-sight, a result of geodetic precession. If so, the pulse profile should be evolving rapidly. \\bigskip We thank M. Kramer for assistance at Parkes, B. Gaensler for help with the ATCA data, and V. Kalogera, M. van Kerkwijk, S. Portegies Zwart, F. Rasio and S. Thorsett for helpful discussions. VMK is supported by a National Science Foundation CAREER award (AST-9875897). FC is supported by NASA grant NAG~5-3229." }, "0005/astro-ph0005022_arXiv.txt": { "abstract": "The recently released BOOMERanG data was taken as ``contradicting topological defect predictions''. We show that such a statement is partly misleading. Indeed, the presence of a series of acoustic peaks is perfectly compatible with a non-negligible topological defects contribution. In such a mixed perturbation model (inflation and topological defects) for the source of primordial fluctuations, the natural prediction is a slightly lower amplitude for the Doppler peaks, a feature shared by many other purely inflationary models. Thus, for the moment, it seems difficult to rule out these models with the current data. ", "introduction": " ", "conclusions": "" }, "0005/hep-ph0005271_arXiv.txt": { "abstract": "{ The collapse of sufficiently large closed domain wall produced during second order phase transition in the vacuum state of a scalar field can lead to the formation of black hole. The origin of domain walls with appropriate size and energy density could be a result of evolution of an effectively massless scalar field at the inflational epoch. We demonstrate that in this case the situation is valued when there are compact domains of less favorable vacuum surrounded by a sea of another vacuum. Each domain has a surface composed of vacuum wall that stores a significant amount of energy, and can collapse into the black hole. This offers the way of massive primordial black holes formation in the early Universe.} ", "introduction": "It is well known that any object put within its gravitational radius forms a black hole (BH). At present time BHs can be naturally created only in the result of gravitational collapse of stars with the mass exceeding three Solar masses \\cite{1} in the end of their evolution. From the other hand, it has long been known \\cite{kp,3,4,mi,ck,klopmop} that primordial black hole (PBH) formation is possible in the early Universe. PBH can form when the density fluctuations become larger then unity on a scale intermediate between Jeans length and horizon size. Other possibilities are related to the dynamics of various topological defects such as the collapse of cosmic strings \\cite{3} from the thermal second order phase transition or to the collisions of the bubble walls \\cite{4,mi} created at the first order phase transitions. Formally, there is no limit on the mass of PBH that forms after the collapse of highly overdense region, it is only needed to form appropriate spectrum of initial density fluctuations at the inflation~\\cite{lk}. We cannot expect {\\it {a priori}} the same \"no--mass--limit\" condition in the case of PBHs formed by topological defects, because the mass of such PBHs is defined by the correlation length of the respective phase transition. In this paper we concern with the possibility to form PBHs after the self--collapse of closed domain walls created during a second order phase transition. Such PBHs should have small masses in the case of thermal second order phase transition with usual equilibrium initial conditions. It takes place because the characteristic size of walls coincides with the correlation length of phase transition, which initiate the formation of that domain walls. Usually we deal with high temperature phase transition in the early Universe, that makes the respective correlation length quite small and does not allow to store a significant amount of energy within the individual closed vacuum wall. Moreover the phase transition can have quite complicated dynamics, namely it can be accompanied by another phase transition that creates another type of topological defects, namely, vacuum strings~\\cite{kim,sik} (see discussion in Section~2). As a result we get the mesh of vacuum strings connected by walls where the probability of the existence of closed walls is strongly suppressed~\\cite{sik}. We discuss here the non--equilibrium scenario of closed vacuum walls formation that opens a new possibility for the massive PBHs production in the early Universe. The starting point is rather general and evident. If a potential of a system possesses at least two different vacuum states there are two possibilities to populate that states in the early Universe. The first one is that the Universe contains both states populated with equal probability, that takes place under the usual circumstances of thermal phase transition. The other possibility corresponds to the case when two vacuum states are populated with different probability and there are islands of the less probable vacuum, surrounded by the sea of another, more preferable, vacuum. The last possibility can take place when we go beyond the equilibrium conditions, established by pure thermal dynamics. More definitely, it is necessary to redefine effectively the correlation length of the scalar field that drives a phase transition and consequently the formation of topological defects. We will show that the only necessary ingredient for it is the existence of an effectively flat direction(s), along which the scalar potential vanishes during inflation. Then the background de-- Sitter fluctuations of such effectively massless scalar field could provide non-- equilibrium redefinition of correlation length and give rise to the islands of one vacuum in the sea of another one. In spite of such redefinition the phase transition itself takes place deeply in the Friedman-- Robertson-Walker (FRW) epoch. After the phase transition two vacua are separated by a wall, and such a wall can be very big. The motion of closed vacuum walls has been first driven analytically in~\\cite{tk}. At some moment after crossing horizon they start shrinking due to surface tension. As a result, if the wall does not release the significant fraction of its energy in the form of outward scalar waves, almost the whole energy of such closed wall can be concentrated in a small volume within its gravitational radius what is the necessary condition for PBH formation. The mass spectrum of the PBHs which can be created by such a way depends on the scalar field potential which parametrizes the flat direction during inflation and triggers the phase transition at the FRW stage. Through the paper we will deal with so called pseudo Nambu--Goldstone (PNG) potential, that is quite common for the particle physics models. The plan of this paper is as following. In Section~2 the origin of vacuum walls at the PNG potential is discussed. The possibility to form the non--equilibrium conditions for the second order phase transition at the inflantional stage is considered in Section~3. In Section~4 the minimal conditions of PBHs formation from collapsing closed vacuum walls are discussed. The mass spectrum of PBHs is evaluated numericaly for the certain choises of parameters. ", "conclusions": "As we have seen in the preceding section all regions with phase $\\theta >\\pi$ are converted into islands with vacuum $\\theta_{\\min}=2\\pi$ surrounding by the closed walls. The size distribution of closed walls imprints the size distribution of domains filled with phase coming from fluctuations that have crossed the point $\\pi$ during the one dimensional brownian motion. The physical size that leaves the horizon during e-- fold number N ($N\\le N_{max}$) reads \\beq \\label{ps} l=H_i^{-1}e^N \\eeq This scale becomes comparable to the FRW particle horizon at the moment \\beq \\label{ph} t_h=H^{-1}e^{2N} \\eeq It is clear that we will start to observe the selfcollapse of a closed domain wall when its size is causally connected. Approximately at the same time the wall is acquiring spherical form due to the surface tension. Thus, if the amount of energy stored in a such vacuum configuration \\beq \\label{mass} E\\approx\\sigma t_h^2 \\eeq is large enough, the BH can be formed in the result of its selfcollapse. More definitely, the gravitation radius of configuration should exceed the minimal size up to which it can collapse. In our case the collapse of closed domain wall, coming from potential (\\ref{potential}), changes on repulsion at the size comparable with wall's width~\\cite{we}. This establish cut off for the PBH's mass spectrum at the small masses range. To evaluate numerically mass spectrum of PBHs we have to calculate the size distribution of domains that contains phases, which are at the range $\\theta >\\pi$. Suppose that at e--fold $N=\\ln{(lH_i)}$ before the end of inflation the volume $V(\\bar{\\theta},N)$ has been filled with phase value $\\bar{\\theta}$. Then at the e--fold $N-1$ the volume filled with average phase $\\bar{\\theta}$ obeys following iterative expression \\cite{anti} \\begin{equation} \\label{iteration}V(\\bar{\\theta},N- 1)=e^{3}V(\\bar{\\theta},N)+(V_{U}(N)- e^{3}V(\\bar{\\theta},N))P(\\bar{\\theta},N-1)\\delta\\theta, \\end{equation} here the $V_{U}(N)\\approx e^{3N}H^{-3}_i$ is the volume of the Universe at $N$ e--fold. We applied here distribution (\\ref{gaus}). Now one can easily calculate the size distribution of domains filled with appropriate value of phase corresponding to $N$, with the use of expression~(\\ref{iteration}). For our numerical calculations we have chosen the following reasonable values for inflational Hubble constant $H=10^{13}$~GeV and for the radius of PNG potential $f=10^{14}$~GeV (see for example \\cite{frees}). Also we suppose that $N_{max}=60$ and that the total energy of the wall is smaller than the total energy of the medium inside it. The simulation has been performed for two cases that depend on the $\\Lambda$ and $\\theta_{N_{max}}$. The results are following. \\vspace{0.3cm} {\\centering \\begin{tabular}{c|c|c|c|c|c|c|c|c|c} \\hline \\( \\log_{10} \\frac{M_{PBH}}{1g.} \\)& 12& 13& 14& 15& 16& 17& 20& 21& 22\\\\ \\hline \\hline \\( \\log_{10} n_{PBH} \\)& 20.4& 18.4& 16.3& 14.2& 12.0& 9.79& 5.18& 2.77& 0.30\\\\ \\hline \\multicolumn{10}{c}{ }\\\\ \\multicolumn{10}{c}{ \\( \\Lambda =10^{8} \\)GeV (see for example\\cite{frees}) , \\( \\theta _{N_{max}}=0.65 \\)}\\\\ \\multicolumn{10}{c}{}\\\\ \\multicolumn{10}{c}{Table 1. The total number \\( n_{PBH} \\) of PBHs against their masses \\( M_{PBH} \\).}\\\\ \\end{tabular}\\par} \\vspace{0.3cm} \\noindent The results of Table 1. are fitted in such a manner to satisfy the most stronger astrophysical constraints. It is known \\cite{21} that only PBHs with masses larger then $\\approx 10^{15}$~g. can survive in respect to Hawking evaporation. The observations of diffused gamma ray background establish strong limit~\\cite{mg} on the density fraction $\\Omega_{PBH}<10^{-9}$ of PBHs with masses $10^{14}\\div 10^{15}$~g.. The next limit, which has to be checked is the limit on the abundance of PBHs with masses $10^{12}\\div 10^{13}$~g.. Although such light PBHs disappeared already due to Hawking radiation it could produce a large amount of entropy at the epoch of nucleosinthesis \\cite{nas}. Actually the constraints \\cite{nas} is valued for the PBH's mass range $10^9\\div 10^{13}$~g. and reads $\\beta <10^{- 15}(10^9g./M_{PBH})$, here $\\beta$ is the density fraction of PBHs at the moment of their formation. It means that at the moment of full evaporation of such PBHs $\\tau =M^3_{PBH}/(g_*m^4_p)$ their contribution into the total density of the Universe should satisfy the following restriction $\\alpha <10^{-4}$. To check our spectrum we express the density fraction of PBHs at the moment of their full evaporation in the terms of $n_{PBH}$, it gives $\\alpha\\simeq 10^{-8}n_{PBH}(1g./M_{PBH})^2$. Thus in the case of Table 1. the entropy production limit~\\cite{nas} is also satisfied. The more careful consideration of particle content of $10^{10}\\div 10^{13}$g. PBH decay products shows that such PBHs are very effective sources of antiprotons at the period of their evaporation~\\cite{ck}. According to the theory on non--equilibrium nucleosynthesis (see~\\cite{klopmop} and references therein) it can change dramatically the abundance of such elements as $^3He$, $^6Li$ and $^7Li$. This fact establishes the most stringent constraints on the density fraction of PBHs in the mass rage $10^{10}\\div 10^{13}$g.~\\cite{klopmop}~\\footnote{Here we take into account only the standard model set of particle spices. The invoking of the some supersymmetric extensions of the standard model with a large number of unstable moduli fields which could come from evaporating PBHs also would influence on the primordial chemical content of the Universe (see for example~\\cite{moduli}).}. Thus, taking into account the effects of non--equilibrium nucleosynthesis of $^3He$ and lithium we have got following limits on the density fraction of PBHs with mass range indicated above: $\\alpha_{^3He}<10^{-15}\\Omega_b(M_{PBH}/1g.)k^{-1}$, $\\alpha_{Li}<10^{-19}\\Omega_b(M_{PBH}/1g.)^{5/4}k^{-1}$, where the $\\Omega_b$ is the fraction of baryon density and $k$ varies from $1/4$ to $1/6$ for PBHs of different masses~\\cite{klopmop}. These limits are also satisfied our model in Table 1. As well we can see from the Table 1. that the number of BH with masses more than $10^{22}$~g. is negligible, whereas BHs with masses smaller than $10^{12}$~g. were not produced at all because of their small gravitational radius. Another interesting case is small $\\Lambda$ limit that comes from QCD~\\cite{kim}. It gives rise to massive BH. \\vspace{0.3cm} {\\centering \\begin{tabular}{c|c|c|c|c|c|c|c|c} \\hline \\( \\log_{10} \\frac{M_{PBH}}{1g.} \\)& 28.3& 29.5& 30.6& 31.8& 33.0& 34.1& 35.3& 36.4\\\\ \\hline \\hline \\( \\log_{10} n_{PBH} \\)& 17.4& 13.0& 11.1& 9.2& 7.2& 5.1& 3.0& 0.66\\\\ \\hline \\multicolumn{9}{c}{ }\\\\ \\multicolumn{9}{c}{ \\( \\Lambda =1 \\)GeV (see for exapmle~\\cite{kim}) , \\( \\theta _{N_{max}}=0.8 \\)}\\\\ \\multicolumn{9}{c}{}\\\\ \\multicolumn{9}{c}{Table 2. The total number \\( n_{PBH} \\) of PBHs against their masses \\( M_{PBH} \\).}\\\\ \\end{tabular}\\par} \\vspace{0.3cm} \\noindent In this case the wall width is very large and so that only rather massive BHs could be formed. This fact can be applied for explanation of giant BHs origin in the centres of galaxies (see for example~\\cite{agn}) The maximal PBH masses in the Table 2 are at least by 3--4 orders of the magnitude smaller than the masses of BHs, assumed to be present in the centres of galaxies. However the account for possible strong concentration of small mass PBHs around close-to-maximal mass PBHs can provide the evolution and collapse of PBH systems into black holes with the mass, exceeding by several orders of the magnitude the maximal PBH mass. It provides the possibility to evolve the approach to the origin of AGNs on the base of our model. To complete our discussion let us mention following. We assumed that the coherent field oscillations around the true vacua, which start after the triggering of phase transition with the amplitude proportional to the residual between the local value of phase and the local vacuum phase, are damped due to relatively strong dissipation, so that no energy density is stored in the form of such oscillations. If the dissipation is not effective, as it is the case for invisible axion~\\cite{kim}, the inhomogeneity of the phase distribution before the phase transition results in the inhomogeneity of the energy density distribution of coherent field oscillations. Then such energy density contributes into the total cosmological density and its large scale inhomogeneity induces effects of anisotropy of relic radiation. In this case one can strongly constrain the model parameters from the observational upper limits on the total cosmological density and on the possible anisotropy of thermal background radiation induced by isocurvature perturbations~\\cite{iso}. Leaving a more completely study of the last two issues to future publications, we can conclude that the non--equilibrium dynamics of scalar field with effectively flat direction can give a new spin to the PBHs formation mechanisms applying collapse of domain walls. \\\\ \\\\ {\\it Acknowledgements}. The work of SGR and MYuK was partially performed in the framework of Section \"Cosmoparticle physics\" of Russian State Scientific Technological Program \"Astronomy. Fundamental Space Research\", with the support of Cosmion-ETHZ and Epcos-AMS collaborations. ASS and MYuK acknowledge supporte from Khalatnikov--Starobinsky school (grant 00--15--96699). \\small" }, "0005/astro-ph0005058_arXiv.txt": { "abstract": "Radio, optical and X-ray observations of the powerful radio galaxy PKS 1138--262 at $z=2.156$ have suggested that this galaxy is a massive galaxy in the center of a forming cluster. We have imaged 1138--262 and the surrounding 38 square arcminute field with the Very Large Telescope\\footnote{Based on observations carried out at the European Southern Observatory, Paranal, Chile, programme P63.O-0477(A).} in a broad band and a narrow band encompassing the redshifted \\lya\\ emission. We detect 50 objects with rest equivalent width larger than 20 \\AA\\ and a luminous, highly extended \\lya\\ halo around 1138--262. If the radio galaxy is at the center of a forming cluster, as observations at other wavelengths suggest, these objects are candidate \\lya\\ emitting cluster galaxies. ", "introduction": "Observations of clusters at high redshift ($z > 2$) can directly constrain cosmological models (e.g. Bahcall \\& Fan \\cite{bah98}), but searches based on colours or narrow band emission have not established the presence of massive clusters (Le F\\`evre et al.\\ \\cite{fev96}; Pascarelle et al.\\ \\cite{pas96}; Keel et al.\\ \\cite{kee99}). There are several indications (e.g. Pentericci et al.\\ \\cite{pen99}) that powerful radio galaxies at high redshift (HzRGs) tend to be in the center of forming clusters. The powerful radio galaxy PKS 1138--262 at redshift 2.156 is a prime example of a forming brightest cluster galaxy and has extensively been studied (e.g.\\ Pentericci et al.\\ \\cite{pen97}). The arguments for 1138--262 being at the center of a cluster include (a) the very clumpy morphology as observed by the HST (Pentericci et al.\\ \\cite{pen98}), reminiscent of a massive merging system; (b) the extremely distorted radio morphology and the detection of the largest radio rotation measures (6200 rad m$^{-2}$) in a sample of more than 70 HzRGs, indicating that 1138--262 is surrounded by a hot, clumpy and dense magnetized medium (Carilli et al.\\ \\cite{car97}; Pentericci et al. \\cite{pen00}); (c) the detection of X-ray emission around 1138--262 (Carilli et al.\\ 1998), indicating the presence of hot cluster gas, although a contribution to the X-ray luminosity by the AGN cannot be precluded. For this reason, we chose 1138--262 to carry out a pilot study with the VLT, to search for direct evidence of clusters at high redshift. There are various techniques for detecting high redshift companion galaxies. The colour selection technique used to find Lyman Break Galaxies (LBGs) (Steidel \\& Hamilton \\cite{ste92}) is not feasible at the redshift of 1138--262, since the Lyman limit falls at 2878 \\AA, which is well below the atmospheric cutoff. Therefore we have adopted the strategy of narrow band imaging at the wavelength of the redshifted \\lya\\ line. This technique is capable of detecting galaxies at redshifts similar to the radio galaxy redshift having strong \\lya\\ emission. ", "conclusions": "We have detected 50 candidate \\lya\\ emitters close to radio galaxy 1138--262 at redshift 2.2. These \\lya\\ emitters are candidate starburst galaxies in the cluster, which is presumed to exist or to be forming around 1138--262. We do not find a significant overdensity of candidates compared to luminosity functions of blank fields nor do we detect a strong concentration gradient in our 8 Mpc$^2$ field. The next step in our search for clusters at high redshift is to confirm the existence of the \\lya\\ emitters by multi object spectroscopy at the VLT and determine the spatial correlation function and velocity dispersion, which together with the size of the cluster will give a direct estimate of the total mass. Additionally, we will carry out X-ray observations of 1138--262 with the Chandra telescope." }, "0005/astro-ph0005091_arXiv.txt": { "abstract": "This paper presents photometric redshifts for 431 Abell clusters imaged as part of the Palomar Abell Cluster Optical Survey (PACOS), of which 236 are new redshifts. We have obtained moderately deep, 3--band (Gunn $gri$) imaging for this sample at the Palomar Observatory $60''$ telescope, as part of the photometric calibration of DPOSS. Our data acquisition, reduction, and photometric calibration techniques are described, and photometric accuracy and consistency is demonstrated. An empirical redshift estimator is presented, utilizing background-corrected median $g-r$ colors and mean $g$ magnitudes for the ensemble of galaxies in each field. We present photometric redshift estimates for the clusters in our sample with an accuracy of $\\sigma_z=0.038$. These redshift estimates provide checks on single-galaxy cluster redshifts, as well as distance information for studies of the Butcher-Oemler effect, luminosity functions, $M/L$ ratios, and many other projects. ", "introduction": "Clusters of galaxies are the largest bound systems in the Universe providing useful constraints for theories of large-scale structure formation and evolution. They are the samples of choice for studying galaxy evolution in dense environments, with many tens or even hundreds of galaxies in a small, physically associated volume. Multicolor optical photometry of galaxy clusters is commonly used to study the Butcher-Oemler effect (Butcher \\& Oemler 1978), the morphology-density relation (Dressler 1980), and other correlations between overall cluster properties, galaxy properties, and redshift. Comparisons between optical and X--ray properties of galaxy clusters are also of considerable scientific interest. For instance, mass-to-light ratios of clusters are also useful for constraining cosmological parameters, including $\\Omega$, the mass density of the universe. Properly understood catalogs in the optical and X--ray can help us better understand the various selection effects present in both types of cluster samples. To obtain the maximal scientific return from such studies, it is necessary to know the redshifts of the clusters. Unfortunately, the majority of known galaxy clusters do not have measured redshifts. Nearly fifty years has passed since the publication of Abell's (1958) optically selected cluster catalog, and only $\\sim 1/3$ of the Northern clusters have had spectroscopically measured redshifts, with many of these based on only one or two galaxies. Even at low redshift, obtaining accurate cluster redshifts requires a four-meter class telescope with multi-object spectroscopic capability; performing a survey of hundreds or even thousands of clusters is prohibitively time consuming. In recent years, there has been an increasing recognition that redshifts of individual objects or clusters can be estimated quite accurately from photometric data (Frei \\& Gunn 1994, Brunner \\etal 1997, {\\it etc.}) These estimators have traditionally relied on either empirical correlations between individual galaxy colors and spectroscopically measured redshifts (Connolly \\etal 1995 ), or a template method wherein model spectra are created from evolutionary synthesis codes or spectroscopic data (Gwyn \\& Hartwick 1996). However, both techniques aim to measure the redshifts of single objects using many ($n\\ge4$) colors; increasing the number of colors results in more accurate redshifts over a larger redshift range. An extensive discussion and comparison of existing techniques can be found in Hogg \\etal (1998). Unlike these methods, we are instead relying on only two filters (Gunn $g$ and $r$) to derive photometric redshifts for an ensemble of objects (a galaxy cluster) over a relatively small redshift range ($00.3$, compared with the largest eccentricities in our Solar System, of about~0.2 for Mercury and Pluto, and 0.05 for Jupiter. Even more significantly, about two thirds are orbiting their host star much closer than Mercury orbits the Sun (0.39~AU). While the Doppler technique preferentially selects systems in tight orbits, giant planets so close to the parent star were generally unexpected. Theoretical progress in understanding their formation and their properties has been rapid, but it remains far from complete. \\footnote{The text employs a number of standard astronomical terms and units. Planetary systems are conveniently characterized in terms of corresponding Solar System quantities ($\\odot$ = Sun; $\\oplus$ = Earth; \\Jupiter\\ = Jupiter): $M_\\odot\\simeq2.0\\times10^{30}$~kg; $M_\\Jupiter\\simeq9.5\\times10^{-4}\\,M_\\odot$; $M_{\\rm Saturn}\\simeq2.9\\times10^{-4}\\,M_\\odot$; $M_{\\rm Uranus}\\simeq4.4\\times10^{-5}\\,M_\\odot$; $M_\\oplus\\simeq3.0\\times10^{-6}\\,M_\\odot\\simeq 3\\times10^{-3}\\,M_\\Jupiter$. 1~AU~=~1~astronomical unit (mean Sun-Earth distance) $\\simeq1.5\\times10^{11}$~m. For Jupiter, $a=5.2$~AU and $P=11.9$~yr. Stellar distances are conveniently given in parsec (pc), defined as the distance at which 1~AU subtends an angle of 1~second of arc (or arcsec); 1~pc~$\\simeq3.1\\times10^{16}$~m $\\simeq$~3.26 light-years. For reference, distances to the nearest stars are of order 1~pc; there are about 2000 known stars within a radius of 25~pc of our Sun, and the distance to the Galactic centre is about 8.5~kpc. Stellar masses range from around $0.1-30\\,M_\\odot$, with spectral types providing convenient spectral classification related to the primary stellar properties of temperature and luminosity. Our Sun is of spectral type G2V: cooler stars (types K, M) are of lower mass and have longer lifetimes; hotter stars (types F, A, etc.) are of higher mass and have shorter lifetimes. Object names such as 70~Vir (for 70~Virginis) reflect standard astronomical (constellation-based) nomenclature, while other designations reflect discovery catalogues or techniques variously labelled with catalogue running numbers (e.g.\\ HD~114762) or according to celestial coordinates (e.g.\\ PSR~1257+12). The International Astronomical Union is in the process of formulating recommendations for the nomenclature of extra-solar planets (cf.\\ \\cite{wd98}), meanwhile the {\\it de facto\\/} custom denotes (multiple) planets around star~X as X~b, c,... according to discovery sequence. } Based on present knowledge from the radial velocity surveys, about 5\\% of solar-type stars may harbour massive planets, and an even higher percentage may have planets of lower mass or with larger orbital radii. If these numbers can be extrapolated, the number of planets in our Galaxy alone would be of order 1~billion. Although only one main-sequence extra-solar planetary system is known which contains more than one planet, present understanding of planet formation could imply that many of these massive planets are accompanied by other objects in the same orbiting system. For the future, experiments capable of detecting tens of thousands of extra-solar planetary systems, lower mass planets down to around $1\\,M_\\oplus$, and spectral signatures which may indicate the presence of life, are now underway or are planned. The most substantial advances may come from space observatories over the next 10--20~years. This review covers published literature to March 2000. It provides a summary of the theoretical understanding of planet formation, a review of detection methods, reporting on major relevant experiments both ongoing and planned, and outlines the properties of the extra-solar planetary systems detected to date. Amongst others it addresses the following questions: What defines a planet? What prospects are there for the various search programmes underway? How common are planetary systems? What is their formation process? What information can be deduced from them? What fraction are likely to lie in the `habitable zone'? What are the prospects for detecting the presence of life? Section~\\ref{sec:stars-planets} provides a background to the processes believed to underpin star formation and planetary formation. Section~\\ref{sec:methods} reviews detection methods, including instruments and programmes under development, and their prospects for planet detection in the future. The detection of protoplanetary disks, from which planets are formed, is also covered. Section~\\ref{sec:properties} presents the observed and inferred properties of the known systems and their host stars, explaining how these characteristics may fit into a revised picture of the formation and evolutionary of extra-solar planets. Section~\\ref{sec:life} touches upon some of the issues relevant for the development of the field from planet discovery to the detection of life. ", "conclusions": "\\label{sec:summary} Precise radial velocity measurements have discovered 34 extrasolar giant planets within about 50~pc over the last 5~years. With many radial velocity monitoring programmes underway, and an occurrence rate of some 5\\% in systems measured to date, we may expect that some 100~such planets will be discovered over the next 5~years, some 10--20 of which may be close-in systems for which transit measurements may be possible. A number of other experimental techniques capable of detecting extra-solar planets are under development. Global space astrometric measurements at the 10~microarcsec level may furnish a list of 10--20\\,000 giant planets out to 100--200~pc by the year 2020, while accurate photometric monitoring from space should lead to the detection and characterisation of significantly lower mass planets. Together, such data sets will provide a vast statistical description of planetary masses, orbits, and eccentricities, allowing important constraints to be placed on the complex processes believed to be involved in planetary formation. Improved knowledge of individual systems, in particular from transit measurements, combined with improved atmospheric modelling and theories of habitability, will narrow down the range of identified planets on which life may have developed. Nearby, Earth-mass planets should exist, and would be the natural targets for infrared space interferometers which, by 2020, may succeed in imaging and providing evidence for life on them. We now know that other worlds -- large ones at least -- are common. Developments have been so rapid over the last few years that many significant developments, and many new surprises, can be predicted with confidence. \\ack I am grateful to the European Southern Observatory for hospitality while writing the major part of this review. Its preparation has benefitted from up-to-date information on systems and www links of the Extra-Solar Planets Encyclopaedia maintained by Dr~Jean Schneider (http://www.obspm.fr/planets) and updated orbital parameters maintained by Dr~Geoff Marcy (http://exoplanets.org/). On-line journals, the NASA Astrophysics Data System (ADS), and Latex/Bibtex have facilitated its preparation. I~am grateful to all colleagues who readily furnished preprints and authorised the use of their figures for this review. Figures originally published in the Astrophysical Journal are reproduced by permission of the AAS. It is a pleasure to thank G.~Marcy (U.C.~Berkeley and San Francisco State University), P.D.~Sackett (University of Groningen) and F.P.~Israel and P.T.~de Zeeuw (Sterrewacht Leiden) for comments on the manuscript. I am particularly grateful to J.~Schneider (Observatoire de Paris-Meudon) for numerous valuable and perceptive suggestions for improvements and additions. \\vskip 10pt Note added in proof: a further 8~planets were reported, from the Coralie spectrometer observations, in a press release from the European Southern Observatory, 2000 May~4. \\newpage {\\footnotesize\\parskip 0pt" }, "0005/astro-ph0005434_arXiv.txt": { "abstract": "We present Far Ultraviolet Explorer ({\\it FUSE}) spectra for three Magellanic Cloud O stars (Sk 80, Sk $-67^\\circ$05 and Sk $-67^\\circ$111) with repeated observations. The data demonstrate the capabilities of {\\it FUSE}\\/ to perform time-resolved spectroscopy on extragalactic stars. The wavelength coverage of {\\it FUSE}\\/ provides access to resonance lines due to less abundant species, such as sulfur, which are {\\em unsaturated} in O supergiants. This allows us to examine wind variability at all velocities in resonance lines for stars with higher mass loss rates than can be studied at longer ($\\lambda \\geq 1150$\\AA) wavelengths. The {\\it FUSE}\\/ wavelength range also includes resonance lines from ions which bracket the expected dominant ionization stage of the wind. Our observations span 1-4 months with several densely sampled intervals of 10 hours or more. These observations reveal wind variability in all of the program stars and distinctive differences in the ionization structure and time scales of the variability. Sk $-67^\\circ$111 demonstrates significant wind variability on a time scale less than 10 hours and the coolest O star (Sk $-67^\\circ$05) exhibits the largest variations in O {\\sc vi}. ", "introduction": "Not long after the first observations of UV wind lines by Morton (1967), repeated observations with the {\\it Copernicus}\\/ satellite discovered that these lines were time variable (York et al.\\ 1977, Snow 1977). Subsequent {\\it IUE}\\/ observations established that discrete absorption components (DACs) are ubiquitous in hot stars with well-developed, but unsaturated wind lines (e.g., Howarth \\& Prinja 1989) and that the presence of DACs is indicative of wind variaility. Later work by Kaper et al.\\ (1996) demonstrated the universality UV wind line variability directly, through time series observations of selected O stars. Massa et al.\\ (1995) and Kaper et al.\\ determined that the wind lines in many stars vary on their stellar roation time scale, verifying Prinja's (1988) suggestion that wind line activity is linked to the stellar rotation period. This has lead researchers to consider physical explanations which rely on phenomena present on the stellar surface, such as magnetic fields (Henrichs et al.\\ 1998) or non-radial pulsations (Gies et al.\\ 1999). Whatever the origin, Owocki et al.\\ (1995) have shown that winds with large scale spiral patterns are consistent with some of the observations (Fullerton et al.\\ 1997, Kaper et al.\\ 1999), although other aspects of the variability remain unexplained. All of the previous results were deduced from observations of Galactic stars. Although there is no reason to doubt similar variability in normal extragalactic stars (Prinja \\& Crowther 1998 have detected DACs in several Magellanic Cloud stars), direct observations of wind variability have not been obtained. Beside demonstrating its universality, observing wind variability in Large and Small Magellanic Cloud (LMC, SMC) stars with {\\it FUSE}\\/ is important in other respects. First, the lower metallicity of the clouds provides an opportunity to study how instabilities in line driven winds are affected by abundances. Second, the ions available to {\\it FUSE}\\/ complement those normally observed at longer wavelengths. This is especially true for the S {\\sc iv} and S {\\sc vi} lines which can be used to determine the ionization structure of wind variability. Third, because sulfur is less abundant than ions with comparable ionization potentials accessible at longer wavelengths, its resonance lines are less optically thick and can be used to probe activity at deeper levels in the winds. \\medskip \\begin{tabular}[*t]{|l|l|c|c|} \\multicolumn{4}{c} {{\\bf Table 1}} \\\\ \\hline \\multicolumn{4}{|c|} {\\bf {Program Stars}} \\\\ \\hline Name\t & Sp Ty$^a$\t & $v_{\\infty}$ (km s$^{-1}$) & $V$ (mag) \\\\ \\hline Sk 80\t\t & O7 Iaf+\t & 1400$^b$ & 12.36 \\\\ Sk $-67^\\circ$05 & O9.7 Ib\t & 1665$^c$ & 11.34 \\\\ Sk $-67^\\circ$111 & O7 Ib(f)\t & 1800$^b$ & 12.57 \\\\ \\hline \\end{tabular} \\\\ \\small \\noindent $^a$ Spectral types from Fitzpatrick (1988) and Walborn (1977) \\\\ \\noindent $^b$ Bianchi et al.\\ (2000) \\\\ \\noindent $^c$ Patriarchi \\& Perinotto (1992) \\normalsize \\medskip ", "conclusions": "Although we lack well-sampled temporal coverage and repeated observations for S {\\sc vi} at this time, the current data still provide a glimpse of the capabilities of {\\it FUSE}\\/ to perform wind variability studies. The combination of {\\it FUSE}\\/ wavelength coverage and the reduced metallicity of the Magellanic Cloud stars enabled us to examine wind variability in new density regimes and new ionization states. In many respects, S {\\sc iv} and Si {\\sc iv} $\\lambda \\lambda$ 1400 are sensitive to similar plasmas. However, the reduced abundance of sulfur allows us to probe more massive flows. Both O {\\sc vi} and N {\\sc v} sample very hot plasma. However, the wider separation of the O {\\sc vi} $\\lambda \\lambda 1032, 1038$ doublet removes the complications introduced by the overlap of the N {\\sc v} $\\lambda \\lambda 1240$ components in many stars (however, interstellar Ly $\\beta$ contaminates O {\\sc vi} $\\lambda 1032$). These features allowed us to detect wind variability at intermediate velocities in O7 supergiants. In Galactic O7 supergiants, all of the wind lines in the {\\it IUE}\\/ or {\\it HST}\\/ range are saturated, and it is not until O6 and earlier supergiants that Si {\\sc iv} desaturates enough for variability to be detected (see $\\lambda$ Cep in Kaper et al., 1996). Even then, N {\\sc v} remains completely saturated, so ionization information cannot be obtained. In contrast, the P~Cygni profiles of S {\\sc iv} are well developed and unsaturated in the {\\it FUSE}\\/ spectra of both O7 supergiants. O {\\sc vi} is similarly well-suited studying variations in spectra of Sk $-67^\\circ$111, but is too weak to be useful in Sk 80. Several aspects of the current data are noteworthy. First, to provide an indication of the magnitude of the physical changes implied by the morphological variability, consider the results from Bianchi et al.\\ (2000) for S {\\sc iv} in Sk 80 and Sk $-67^\\circ$111. Analyzing the same observations displayed here, they determined that the line of sight wind column density varied by 68\\% in Sk 80 and 45\\% in Sk $-67^\\circ$111. We see, therefore, that the variability is not a small perturbation on top of an otherwise steady flow. Instead, {\\em variability is a fundamental property of the winds}. Second, the {\\em wind lines vary in every Magellanic Cloud O star with repeated observations} separated by more than a few days, a result which is similar to that observed in Galactic O stars (Kaper et al.\\ 1996). Third, although the S {\\sc iv} variability in Sk 80 is stronger than in Sk $-67^\\circ$111, its O {\\sc vi} variability is weaker, if present at all -- an effect that may be due to the relative abundances of the two stars. Fourth, although Sk $-67^\\circ$05 is cooler than either Sk 80 or Sk $-67^\\circ$111, its O {\\sc vi} variability is much stronger. This implies that the variability we happened to observe is of a higher ionization state than in the two O7 stars. Fifth, while there is no evidence of short term ($\\leq 1$ day) variability in two of the program stars, it is clearly present in Sk $-67^\\circ$111 on time scales similar to those observed for the Galactic O6 supergiant $\\lambda$ Cep (Kaper et al., 1996). From our very limited data, it appears that the overall wind variability of the Magellanic Cloud O stars is similar to Galactic stars. However, longer and better sampled time series will be needed to determine whether the time scale of the variability is related to the rotation period of the stars, as it is in many Galactic O stars and B supergiants." }, "0005/astro-ph0005145_arXiv.txt": { "abstract": "Ultraviolet through optical spectroscopy of four NLS1s shows strong absorption features in the high-ionization UV resonance lines such as C~IV, N~V, and Si~IV. Mg~II is not absorbed. The absorption could originate in the warm absorber. ", "introduction": "We observed four NLS1s with the HST Faint Object Spectrograph (FOS). Spectra were obtained on consecutive orbits to provide near-simultaneous coverage from the UV through the optical ($\\lambda\\lambda1150-6800$). The effects of variability should be minimized in these data. Three of the targets, Mrk 493, WPVS007, and RX J$0134-42$, show strong optical Fe II emission and weak [O III] $\\lambda5007$ narrow-line emission. Two objects, WPVS007 and RX J$0134-42$, show steep X-ray spectra and large X-ray variability. Two, Akn 564 and RX J$0134-42$, have been reported to have warm X-ray absorbers. The original goals were: (a) to calculate physical parameters of the gas from the diagnostic UV emission lines, (b) to study the UV iron-line complexes, and (c) to obtain a good piece of the spectral energy distribution (SED). ", "conclusions": "" }, "0005/astro-ph0005373_arXiv.txt": { "abstract": "A study of the absorption systems toward the gravitationally lensed quasar APM~08279+5255 is presented. \\par\\noindent Most of the Mg~{\\sc ii} systems in the redshift range $z$~$\\sim$~1.2--2.07, although saturated, show large residuals at the bottom of the lines. The most likely interpretation is that individual clouds within Mg~{\\sc ii} halos do cover only one of the two brightest QSO images. The separation between the two lines of sight decreases from 1.7 to 0.7~$h^{-1}_{75}$~kpc ($q_{\\rm o}$~=~0.5, $z_{\\rm lens}$~=~1) between $z$~=~1.22 and $z$~=~2.07. This reveals that Mg~{\\sc ii} halos are made of a collection of clouds of radius smaller than about 1~$h^{-1}_{75}$~kpc. \\par\\noindent Two strong Mg~{\\sc ii} absorbers at $z_{\\rm abs}$~=~1.062 and 1.181 are studied in detail. This is the first time that the Na~{\\sc i}$\\lambda$3303 doublet is detected in such high redshift systems. Together with the detection of the Mg~{\\sc i}$\\lambda$2852 transition, this strongly constrains the physical characteristics of the gas. The $N$(Na~{\\sc i})/$N$(Mg~{\\sc i}) ratio is found to be larger than unity, implying that the gas is cool and neutral. The Doppler parameters measured in individual and well detached components is probably as small as 1~km~s$^{-1}$. The column densities of Na~{\\sc i}, Ca~{\\sc ii}, Mg~{\\sc i}, Ti~{\\sc ii}, Mn~{\\sc ii} and Fe~{\\sc ii} observed at $z_{\\rm abs}$~=~1.1801 are very close to that observed along the line of sight towards 23~Ori in our Galaxy. The shape of the QSO continuum is consistent with attenuation by dust at $z$~$\\sim$~1 ($A_{\\rm V}$~$\\sim$~0.5~mag). Altogether it is found that the H~{\\sc i} column density at $z$~=~1 is of the order of 1 to 5~10$^{21}$~cm$^{-2}$, the corresponding metallicity is in the range 1--0.3~$Z_{\\odot}$, the overall dust-to-metal ratio is about half that in our Galaxy and the relative depletion of iron, titanium, manganese and calcium is similar to what is observed in cool gas in the disk of our Galaxy. The objects associated with these two systems could both contribute to the lens together with another possible strong system at $z_{\\rm abs}$~=~1.1727 and the strong Lyman-$\\alpha$ system at $z_{\\rm abs}$~=~2.974. \\par\\noindent The probable damped Lyman-$\\alpha$ system at $z_{\\rm abs}$~=~2.974 has 19.8~$<$~log~$N$(H~{\\sc i})~$<$~20.3. The transverse dimension of the absorber is larger than 200~$h^{-1}_{75}$~pc. Column densities of Al~{\\sc ii}, Fe~{\\sc ii}, Si~{\\sc ii}, C~{\\sc ii} and O~{\\sc i} indicate abundances relative to solar of $-$2.31, $-$2.26, $-$2.10, $-$2.35 and $-$2.37 for, respectively, Fe, Al, Si, C and O (for log~$N$(H~{\\sc i})~=~20.3). These surprizingly similar values indicate that the amount of dust in the cloud is very small as are any deviations from relative solar abundances. It seems likely that the upper limits found for the zinc metallicity of several damped Lyman-$\\alpha$ systems at $z$~$>$~3 in previous surveys is indicative of a true cosmological evolution of the metallicity in individual systems. ", "introduction": "$~=~2.974} \\begin{figure} \\centerline{\\vbox{ \\psfig{figure=9595.f12,height=10.cm,width=9.3cm,angle=0} }} \\caption[]{Absorptions in a few transitions on a velocity scale with origin at $z_{\\rm abs}$~=~2.974. Vertical dashed lines mark the position of the strongest components.} \\label{s297} \\end{figure} There is a strong absorption feature with $W_{\\rm obs}$~$>$~19~\\AA~ at $\\lambda$~=~4831~\\AA. It corresponds to H~{\\sc i} Lyman-$\\alpha$ at $z_{\\rm abs}$~=~2.974. The only possibility of coincidence with a BAL transition could be Lyman$\\beta$ at $z_{\\rm abs}$~=~3.71 but there is no corresponding Lyman$\\alpha$ transition (see Srianand \\& Petitjean 2000). Although the red-wing of the absorption has the characteristic shape of a damped transition, uncertainties in the continuum determination prevent an accurate determination of the column density. Associated absorptions from Al~{\\sc ii}, Fe~{\\sc ii}, Si~{\\sc ii}, C~{\\sc ii} and O~{\\sc i} are detected in four components spanning $\\sim$100~km~s$^{-1}$ (see Fig.~{\\ref{s297}). By fitting the Lyman-$\\alpha$ line, we estimate that the total H~{\\sc i} column density in the four components is in the range 19.8~$<$~log~$N$(H~{\\sc i})~$<$~20.3. It is important to note that there is no evidence that the cloud does not cover the three lines of sight which are separated by $\\sim$~200$h^{-1}_{75}$~pc at the redshift of the absorber. Indeed, the core of the H~{\\sc i} absorption is black over $\\sim$~15~\\AA. Column densities integrated over the absorption profiles have been obtained for all species and summarized in Table~3. Column \\#3 of Table~3 gives the abundance of the element assuming that log~$N$(H~{\\sc i})~=~20.3 in the cloud and that the observed ion is the dominant species. Given the uncertainty in the neutral hydrogen column density, the absolute values are unreliable and could be 0.5~dex higher. Column \\#4 and \\#5 of Table~3 give the solar metallicity and the metallicity relative to solar respectively. It is remarkable how consistent the measurements are, which point toward metallicities less than 10$^{-1.5}$~$Z_{\\odot}$. The low metallicities are not due to depletion into dust-grains. Indeed, the relatively small neutral hydrogen column density implies that column densities of relatively abundant elements can be measured. It is apparent that iron is not depleted compared to carbon or oxygen and that the amount of dust in this cloud must be very small. Metallicity in damped Lyman-$\\alpha$ systems is usually measured using zinc, an element that is not very much depleted into dust-grains in our Galaxy and, because it has relatively low metallicity compared to other elements, induces non-saturated absorptions even for clouds of high hydrogen column density. There is barely no evolution in the measured Zinc metallicity from $z$~$\\sim$~1 to $z$~$\\sim$~3 (Pettini et al. 1997, 1999). At $z$~$>$~3, in most damped Lyman-$\\alpha$ studied up to now, zinc is not detected. This is most probably a consequence of limited S/N ratio of the data however. Indeed, the detection limit of most of the spectra is log~$N$(Zn~{\\sc ii})~$\\sim$~11.5 which means [Zn/H]~$<$~$-$1.4 (see e.g. Prochaska \\& Wolfe 1997). It can be noted that in the system at $z_{\\rm abs}$~=~2.974 toward APM~08279+5255, the limit on zinc is of this order. However, as we can measure metallicities for more abundant elements, we know that metallicities are less than $-$1.5. For such abundances, zinc would have been detectable in the spectrum of APM~08279+5255 only for H~{\\sc i} column densities larger than 10$^{21}$~cm$^{-2}$. It is therefore possible that the upper limit found for the zinc metallicity at $z$~$>$~3 in previous surveys is indicative of a true evolution of the metallicity in individual systems (see also Prochaska \\& Wolfe 2000; Savaglio et al. 1999). This should be checked by measuring in the same systems abundances of species like carbon, aluminium, silicon and iron. \\begin{table} \\begin{tabular}{lllll} \\multicolumn{5}{l}{{\\bf Table 3.} Column densities$^a$ in the $z_{\\rm abs}$~=~2.974 system}\\\\ \\hline \\multicolumn{1}{c}{Species}& \\multicolumn{1}{c}{log~$N$}&\\multicolumn{1}{c}{$Z^b$}& \\multicolumn{1}{c}{$Z_{\\odot}$}& \\multicolumn{1}{c}{[X/H]$^{b,c}$}\\\\ \\multicolumn{1}{c}{} &\\multicolumn{1}{c}{(cm$^{-2}$)}&\\multicolumn{1}{c}{}& \\multicolumn{1}{c}{}\\\\ \\hline H~I & 19.8--20.3 & & & \\\\ C~{\\sc ii} & 14.5 & $-$5.80 & $-$3.45 & $-$2.35 \\\\ O~{\\sc i} & 14.8: & $-$5.50 & $-$3.13 & $-$2.37 \\\\ Al~{\\sc ii} & 12.5 & $-$7.78 & $-$5.52 & $-$2.26 \\\\ Si~{\\sc ii} & 13.8 & $-$6.55 & $-$4.45 & $-$2.10 \\\\ Fe~{\\sc ii} & 13.5 & $-$6.80 & $-$4.49 & $-$2.31 \\\\ Zn~{\\sc ii} & $<$11.8 & $<-$8.50 & $-$7.35 & $<-$1.1 \\\\ Ni~{\\sc ii} & $<$12.5 & $<-$7.80 & $-$5.75 & $<-$2.0 \\\\ \\hline \\multicolumn{5}{l}{$^a$ logarithm of, in cm$^{-2}$; $^b$ log~$N$~(H~{\\sc i})~=~20.3 is assumed; }\\\\ \\multicolumn{5}{l}{$^c$ typical errors are $\\pm$0.3~dex.}\\\\ \\label{tab297} \\end{tabular} \\end{table} ", "conclusions": "The doublet ratio of several intervening Mg~{\\sc ii}$\\lambda\\lambda$2796,2803 systems along the line of sight to APM~08279+5255 is observed close to unity, indicating saturation of the lines, whereas the depth of the lines is close to 0.5 in the normalized spectrum (see Fig.~\\ref{cf}). This can be understood if the absorption profile is made of components with arbitrarily small Doppler parameters ($b$~$\\sim$~1--1.5~km~s$^{-1}$). This would imply however a surprisingly low temperature (1500--3000~K) when the gas is expected to be heated by photo-ionization to temperatures larger than 10$^{4}$~K (e.g. Petitjean et al. 1992). A more likely explanation of these observations is that Mg~{\\sc ii} galactic halos are composed of a collection of clouds each of them having dimensions less than $\\sim$1~kpc. Individual clouds, regularly spread over the velocity profile by kinematics, cover only one of the two brighest image of the lensed quasar. The number density of clouds is not large enough for the absorption material to cover the two lines of sight at all velocity positions. The total covering factor of the halo however is close to one, consistent with observations of associated galaxies at intermediate redshift. In contrast, the two strong Mg~{\\sc ii} systems at $z_{\\rm abs}$~=~1.06 and 1.18 have covering factor equal to one (the lines are saturated and go to the zero level) over more than 200~km~s$^{-1}$. These latter systems are likely to arise due to absorption through the central part of galaxies where the number of clouds is so large that saturated absorption occurs along both lines of sight whatever the radius of the individual clouds might be. Models by Mo \\& Miralda-Escud\\'e (1996) have shown that halos with small rotation velocity ($<$~100~km~s$^{-1}$) should contribute little to the total cross-section of Mg~{\\sc ii} systems as they have dimensions as small as 5~kpc. This conclusion is probably true at $z$~$<$~1 (see also Churchill et al. 1996). It is however interesting to note that most of the $z$~$>$~1 Mg~{\\sc ii} systems studied here (see Fig.~\\ref{cf}) are spread over much less than 100~km~s$^{-1}$. Moreover, they have equivalent widths $W_{\\rm r}$~$\\sim$~0.43, 0.37, 0.12, 0.99, 0.28 and 0.17~\\AA~ at $z_{\\rm abs}$~=~1.211, 1.5497, 1.5523, 1.813, 2.0418 and 2.0668 respectively. Therefore the systems we see have similar strengths as the systems which, at lower redshift, are associated with large halos of galaxies. In particular, they are generally stronger than the weak Mg~{\\sc ii} systems studied by Churchill et al. (1999). This means that, contrary to what is seen at lower redshift, these systems could be associated with halos of low rotation velocity ($<$~100~km~s$^{-1}$) and thus small radii. The two strong Mg~{\\sc ii} systems at $z_{\\rm abs}$~=~1.06 and 1.18 are studied in detail. Absorption from Ca~{\\sc ii}, Mg~{\\sc i}, Ti~{\\sc ii}, Mn~{\\sc ii} and Fe~{\\sc ii} have been observed in several damped Lyman-$\\alpha$ systems over a large range of redshift (Meyer et al. 1995, Lu et al. 1996, Vladilo et al. 1997, Proschaska \\& Wolfe 1997, Churchill et al. 2000). This is, however, the first time that Na~{\\sc i} is also detected at such redshift, thanks to the combination of high S/N ratio and high spectral resolution. This additional strong constraint shows that the gas in these systems is cool and neutral. Indeed, similar column densities are observed in our Galaxy for Ca~{\\sc ii}, Mg~{\\sc i}, Ti~{\\sc ii}, Mn~{\\sc ii} and Fe~{\\sc ii} in warm and cool gas clouds toward, respectively, $\\mu$Col and 23~Ori. Only the Na~{\\sc i} column density differs; it is more than an order of magnitude larger through the cool cloud. Doppler parameters as low as $b$~$\\sim$~1~km~s$^{-1}$ are derived from Voigt-profile fitting of isolated subcomponents. We find that the H~{\\sc i} column density at $z$~=~1 is of the order of $1\\times 10^{21}$~cm$^{-2}$ to $5\\times 10^{21}$~cm$^{-2}$, the corresponding metallicity is in the range 1--0.3~$Z_{\\odot}$, the dust-to-metal ratio is about a third that in our Galaxy and the relative depletions of iron, titanium, manganese and calcium are similar to those in cool gas in the disk of our Galaxy. The dust-to-metal ratio measured here is similar to what is derived in most of the damped Lyman-$\\alpha$ systems (Vladilo 1998, Savaglio et al. 1999). The presence of dust is supported by the reddening of the QSO spectrum over the $R$-band. These are probably amongst the most metal and dust-rich damped Lyman-$\\alpha$ systems at $z$~$\\sim$~1. The dust depletion pattern is similar to that observed in cool gas in the Galaxy. All this is consistent with the finding by Petitjean et al. (1992) that although most of the damped Lyman-$\\alpha$ systems arise in warm gas, the highest column densities are due to a collection of clumps that condense out of the warm phase due to thermal instability (see also Lane et al. 2000). Another damped Lyman-$\\alpha$ system is seen at $z_{\\rm abs}$~=~2.974 with 19.8~$<$~log~$N$(H~{\\sc i})~$<$~20.3. As the Lyman-$\\alpha$ line is black over about 15~\\AA, the cloud must cover the three QSO images. The transverse dimension of the absorber is therefore larger than 200~$h^{-1}_{75}$~pc. Column densities of Al~{\\sc ii}, Fe~{\\sc ii}, Si~{\\sc ii}, C~{\\sc ii} and O~{\\sc i} indicate abundances relative to solar of $-$2.31, $-$2.26, $-$2.10, $-$2.35 and $-$2.37 for, respectively, Fe, Al, Si, C and O (for log~$N$(H~{\\sc i})~=~20.3). Metallicities are therefore less than 10$^{-1.5}$~$Z_{\\odot}$ and, if any, the amount of dust in the cloud is very small, as are any deviations from relative solar abundances. It seems likely that the difficulty to detect Zinc in several damped Lyman-$\\alpha$ systems at $z$~$>$~3 in previous surveys is indicative of a true cosmological evolution of the metallicity in individual systems (see also Prochaska \\& Wolfe 2000, Savaglio et al. 1999). This should be checked by measuring in the same systems abundances of species like carbon, aluminium, silicon and iron." }, "0005/astro-ph0005529_arXiv.txt": { "abstract": "The \\emph{Far Ultraviolet Spectroscopic Explorer} satellite observes light in the far-ultraviolet spectral region, 905 -- 1187 \\AA\\ with high spectral resolution. The instrument consists of four coaligned prime-focus telescopes and Rowland spectrographs with microchannel plate detectors. Two of the telescope channels use Al:LiF coatings for optimum reflectivity from approximately 1000 to 1187 \\AA\\ and the other two use SiC coatings for optimized throughput between 905 and 1105 \\AA. The gratings are holographically ruled to largely correct for astigmatism and to minimize scattered light. The microchannel plate detectors have KBr photocathodes and use photon counting to achieve good quantum efficiency with low background signal. The sensitivity is sufficient to examine reddened lines of sight within the Milky Way as well as active galactic nuclei and QSOs for absorption line studies of both Milky Way and extra-galactic gas clouds. This spectral region contains a number of key scientific diagnostics, including \\ion{O}{6}, \\ion{H}{1}, \\ion{D}{1} and the strong electronic transitions of H$_2$ and HD. ", "introduction": "The \\emph{Far Ultraviolet Spectroscopic Explorer} (\\emph{FUSE}) is a NASA astronomy mission, developed in cooperation with the Canadian Space Agency and the Centre National d'Etudes Spatiales of France, that is exploring the far-ultraviolet (FUV) universe from 905 to 1187 \\AA\\ with high spectral resolution. \\emph{FUSE} was launched 1999 June 24 on a Delta II rocket. Early Release Observations, which are the basis of the papers in this issue of the Astrophysical Journal Letters, began in 1999 October and regular science operations commenced in 1999 December. After the decommissioning of the \\emph{Copernicus} Mission \\citep{spi75} in 1981, it was clear that a follow-on mission with much higher sensitivity and a velocity resolution comparable to or better than that of \\emph{Copernicus} ($\\sim15$ \\kms\\ FWHM \\footnote{Full-width-half-maximum of a spectral profile. All resolution widths are FWHM unless specifically stated otherwise.}) was highly desirable. The utilization of modern detectors and mirror technology opens the spectral bandpass from the short-wavelength cutoff of the \\emph{Hubble Space Telescope} (\\emph{HST}) down to the \\ion{H}{1} photoionization limit at 912 \\AA\\ for observations at distances far beyond the $\\sim1\\kpc$ limit for routine \\emph{Copernicus} measurements. \\emph{FUSE} received strong support from two decadal survey committees \\citep{fie82,bah91}. Although cost considerations restricted the telescope for such a mission to roughly the one meter class, modern detectors that simultaneously cover most of the bandpass made high spectral resolution measurements with excellent sensitivity possible. In addition, by using modern mirror technology, the bandpass could be extended below $\\sim1000$ \\AA\\ where the \\emph{Copernicus} sensitivity was limited by the drop in reflectivity of mirrors overcoated with Al and a thin covering of LiF (Al:LiF). Finally, the installation of the Space Telescope Imaging Spectrograph (STIS) \\citep{woo98} on the \\emph{HST} with a lower limit of $\\sim1150$ \\AA\\ meant that the FUSE mission could be designed primarily for the wavelengths below the STIS limit. Exploration of this spectral region since the \\emph{Copernicus} mission has been limited. The Voyager ultraviolet spectrometers \\citep{bro77} obtained low resolution ($\\Delta\\lambda\\ge18$ \\AA) spectra of a number of sources, e.g. \\citep{hol91}. Several missions covering this wavelength range have been flown as space shuttle sortie missions with operational lifetimes of up to two weeks. These included the Hopkins Ultraviolet Telescope (HUT) \\citep{dav92, kru95}, the Orbiting and Retrievable Extreme Ultraviolet Spectrometers (ORFEUS) \\citep{gre91,hur98} and the Interstellar Medium Absorption Profile Spectrograph (IMAPS) \\citep{jen96}. These missions performed a wide range of important studies. However, they were precluded from carrying out more detailed and comprehensive studies by the limited length of the missions as well as limits to the velocity resolution or sensitivity. This article presents an overview of the \\emph{FUSE} mission. A discussion of the scientific background is followed by a description of the mission. The following article \\citep{sah00} discusses the on-orbit performance of the \\emph{FUSE} satellite in more detail. ", "conclusions": "" }, "0005/astro-ph0005003_arXiv.txt": { "abstract": "This paper is intended to offer a pedagogical treatment of inflationary cosmology, which is accessible to undergraduates. In recent years, inflation has become accepted as a standard scenario making predictions that are testable by observations of the cosmic background. It is therefore manifest that anyone wishing to pursue the study of cosmology and large-scale structure should have this scenario at their disposal. The author hopes this paper will serve to `bridge the gap' between technical and popular accounts of the subject. ", "introduction": "The Standard Model of Cosmology has successfully predicted the nucleosynthesis of the light elements, the temperature and blackbody spectrum of the cosmic background radiation, and the observed redshift of light from galaxies which suggests an expanding universe. However, this model can not account for a number of initial value problems, such as the flatness and monopole problems. Inflationary cosmology resolves these concerns, while preserving the successes of the Big-Bang model. Inflation was originally introduced for this reason and its motivation relied on predictions from particle theory. In more recent times, inflation has been abstracted to a much more general theory. It continues to resolve the initial value problems, but also offers an explanation of the observed large-scale structure of the universe. In this paper, the fundamentals of modern cosmology for an isotropic and homogeneous space-time, which is naturally motivated by observation, will be reviewed. The Friedmann equations are derived and the consequences for the dynamics of the universe are discussed. A brief introduction to the thermal properties of the universe is presented as motivation for a discussion of the horizon problem. Moreover, other issues suggesting a more general theory are presented and inflation is introduced as a resolution to this conundrum. Inflation is shown to actually exist as a scenario, rather than a specific model. In the most general case one speaks of the inflaton field and its corresponding energy density. Models of inflation differ in their predictions and the corresponding evolution of an associated inflaton field can be explored in a cosmological context. The equations of motion are cast in a form that makes observational consequences manifest. The slow-roll approximation (SRA) is discussed as a more tractable and plausible evolution for the inflaton field and the slow-roll parameters are defined. Using the SRA, inflation predicts a near-Gaussian adiabatic perturbation spectrum resulting from quantum fluctuations in the inflaton field and the DeSitter space-time metric. These result in a predicted power spectrum of gravity waves and temperature anisotropies in the cosmic background, both of which will be detectable in future experiments. Inflation is shown to be a rigorous theory that makes concise predictions in regards to a needed inflaton potential at the immediate Post-Planck or perhaps even the Planck epoch ($\\sim 10^{-43}$s). This offers the exciting possibility that inflation can be used to predict new particle physics or serve as a constraint for phenomenology from theories such as Superstring theory. \\newpage ", "conclusions": "" }, "0005/astro-ph0005553_arXiv.txt": { "abstract": "The evolution of the temperature in the intergalactic medium is related to the reionization of hydrogen and helium, and has important consequences for our understanding of the \\lya forest and of galaxy formation in gravitational models of large-scale structure. We measure the temperature-density relation of intergalactic gas from \\lya forest observations of eight quasar spectra with high resolution and signal-to-noise ratio, using a new line fitting technique to obtain a lower cutoff of the distribution of line widths from which the temperature is derived. We carefully test the accuracy of this technique to recover the gas temperature with a hydrodynamic simulation. The temperature at redshift $\\bar{z}=$(3.9, 3.0, 2.4) is best determined at densities slightly above the mean: $T_\\star=$($20200\\pm 2700$, $20200\\pm 1300$, $22600\\pm 1900$)K (statistical error bars) for gas density (in units of the mean density) $\\Delta_\\star=$($1.42\\pm 0.08$, $1.37\\pm 0.11$, $1.66\\pm 0.11$). The power-law index of the temperature-density relation, defined by $T=T_\\star (\\Db/\\Delta_\\star)^{\\gmo}$, is $\\gmo=$($0.43\\pm 0.45$, $0.29\\pm 0.30$, $0.52\\pm 0.14$) for the same three redshifts. The temperature at the fixed over-density $\\Delta=1.4$ is $T_{1.4}$=($20100\\pm 2800$, $20300\\pm 1400$, $20700\\pm 1900$)K. These temperatures are higher than expected for photoionized gas in ionization equilibrium with a cosmic background, and can be explained by a gradual additional heating due to on-going \\heii reionization. The measurement of the temperature reduces one source of uncertainty in the lower limit to the baryon density implied by the observed mean flux decrement. We find that the temperature cannot be reliably measured for under-dense gas, because the velocities due to expansion always dominate the widths of the corresponding weak lines. ", "introduction": "The Lyman-$\\alpha$ forest absorption in the spectra of quasars provides a wealth of information about the properties of the intergalactic medium (hereafter, IGM). There has recently been a lot of interest in using the distribution of Doppler parameters of fitted absorption lines, measuring the total velocity dispersion of the gas, to constrain the temperature of the IGM \\citep*{stl99,rgs99,bm99}. In this paper we develop a method to identify and to fit absorption lines, and to obtain the gas temperature from the distribution of the Doppler parameters of the lines. Our algorithm is intended to be simple to implement and be applied identically on simulations and observations. The temperature of the IGM as a function of density is primarily determined by the balance between adiabatic cooling and photoionization heating, once ionization equilibrium with the background radiation has been established. However, during the epoch of reionization, the heating rate is higher because every atom needs to be ionized once (and the ionization can occur on a short time-scale compared to the recombination rate), and the high opacity of the low-density IGM implies that high-frequency photons are absorbed, delivering a much greater amount of heat for each ionization \\citep[e.g.,][]{mr94,hg97,hs98,ah99,g99}. Other sources of heating may also contribute, such as Compton heating by the X-ray background \\citep{me99}, or photoelectric heating by dust grains \\citep*{nss99}. Constraining these sources of heating is one of the two primary reasons why we are interested in measuring the temperature. The other reason is the need to make accurate predictions for the statistics of the \\lya forest flux in order to constrain cosmological parameters \\citep*[e.g.,][]{rms97,wmh97,cwk98,h99,mm99,hsb99,cwp99,wch99,chd99, nh99,mmr99}. The temperature-density relation affects the predicted relationship between the power spectrum of the transmitted flux and the power spectrum of the initial mass density perturbations \\citep[]{nh99}, as well as the predicted mean transmitted flux \\citep{rms97,mmr99}, which can be used to constrain the baryon density of the universe. Recent \\lya forest simulations have shown that, when the structure of the absorption systems is adequately resolved, the predicted absorption line widths are smaller than observed if the temperature of the IGM is determined from photoionization equilibrium, well after reionization has ended \\citep{tle98,tls99,bma99}. To solve the discrepancy the temperature apparently needs to be higher. Several authors have presented measurements of the IGM temperature using different methods, generally finding values moderately higher than expected from photoionization equilibrium \\citep{tsh99,rgs99,bm99,str00}. Our aim in this paper is to provide a new unambiguous measurement of the temperature, making a more exhaustive analysis than in previous work of the model uncertainties that result from comparing the observational results with a simulation. We develop a new line-fitting method as an alternative to the standard Voigt-profile fitting with line deblending, which is much faster, unambiguous, and easy to implement. Our method works by essentially assigning one line to each sufficiently deep minimum in the transmitted flux, and measuring the line width and central optical depth for each line. The gas temperature at each density is then derived from the distribution of line widths at each central optical depth. The systematic uncertainties and model dependence of the method used to derive the temperature are carefully analyzed, in a more extensive way than it was done in previous work. The new method is applied to observational data and to a simulation in exactly the same way, computing error bars due to the variance in our observed sample. The main idea of the method to measure the temperature of the IGM was suggested by \\citet{stl99}, \\citet{rgs99}, and \\citet{bm99}. The probability distribution of Doppler parameters, $P(B)$, is characterized by a lower cutoff, $B_C$, where $P(B)$ rises sharply, with very few lines having narrower Doppler parameters than this cutoff. The idea is that this cutoff is a measure of the gas temperature. In general, absorption lines have both a thermal and a hydrodynamical contribution to their breadth; however, for any set of lines with similar gas temperature, the narrowest ones will be those where the velocity field along the line of sight through the absorber is close to a caustic, so that the variation in the fluid velocity is minimized and thermal broadening dominates the observed line width. In fact, it was found by \\cite{tsh99} that the narrowest absorption lines are primarily thermally broadened. A tight relationship between density and temperature in the IGM for gas at low densities, where shock heating is not very important, is expected theoretically and is found in numerical simulations \\citep{hg97,tle98}. This implies that the narrowest lines at a given gas density (corresponding approximately to the optical depth at the line center) are not selected to have low gas temperature, but low fluid velocity dispersion. This justifies estimating the temperature from the lower cutoff of the Doppler parameter distribution. In \\S 2 we briefly describe the observational data and the simulation that we use. In \\S 3 we describe our line fitting algorithm. In \\S 4 we demonstrate how the line fitter works by running it on spectra from the numerical simulation. In \\S 5 we describe our method for estimating the temperature from Doppler parameter distribution, testing the conditions under which the temperature can be recovered in a model-independent way. In \\S 6 we use the line fitter on the observational data and give results for the measured temperatures. The results are discussed in \\S 7. The Appendix describes further details of our line-fitting method. ", "conclusions": "\\\\ 1. The temperature of the IGM is $\\sim 20000\\pm 2000$K at density 1.4 times the mean, independent of redshift, although an increase of $\\sim 3500$K from $z=3.9$ to $z=3.0$ cannot be ruled out. \\\\ 2. The high temperature cannot be explained by heating in ionization equilibrium, and probably indicates on-going \\heii reionization. \\\\ 3. The contribution of temperature uncertainty to the uncertainty in the baryon density required by the observed mean flux decrement in the \\lya forest is now well constrained. We thank Adam Steed and David Weinberg for helpful comments on the manuscript. \\appendix" }, "0005/astro-ph0005079_arXiv.txt": { "abstract": "Since the chemical evolution of galaxies seems to differ between morphological types and deviates in many details from the standard scenario the question has to be addressed when, how and to what amount metal-enriched ejecta from Supernovae and Planetary Nebulae polute their environment. Since recent observations of dwarf galaxies show no significant metal abundance gradients throughout the galaxies while enhancement of metals happens in isolated HII regions, an effective mixing process has to be assumed. Chemodynamical evolution models can provide a possible explanation by demonstrating that strong evaporation of gas clouds by hot gas and following condensation leads to an almost perfect mixing of the gas. We focus on the different phases of chemodynamical evolution that are experienced by a representative dwarf irregular galaxy model and present a quantitative analysis of the chemodynamical gas flow cycles. ", "introduction": "\\vspace{-2mm} Observations of dwarf galaxies show no significant metal abundance gradients throughout the galaxies (Kobulnicky, Kennicutt, \\& Pizagno 1999; van Zee et al.\\ 1998). A possible mechanism suggested by Tenorio-Tagle (1996) is the mixing of metal-enriched blow-out gas with fresh matter from the environment of the galaxy. After Supernovae (SNe) have caused an outflow of matter (depending on the mass of the galaxy; see Mac Low \\& Ferrara, 1998), the hot gas cools, condenses and falls back into the galactic body. The moderate-to-low stellar metallicities in dwarf elliptical galaxies (dEs) and the related dwarf spheroidals (dSphs) suggest that extensive gas loss has occured during their evolution by means of SNe-driven galactic winds (Larson 1974; Dekel \\& Silk 1986). Current starburst dwarf galaxies (SBDGs) are characterized by superwinds (Marlowe et al.\\ 1995) or by large expanding X-ray plumes which are often confined by swept-up H$\\alpha$ shells. Yet many dSph galaxies show not only a significant intermediate-age stellar population (Hodge 1989; Grebel 1997), but also more recent star formation (SF) events (Smecker at al.\\ 1994; Han et al.\\ 1997) indicating that gas was partly kept in the systems. On the other hand, gas infall might also cause a new SF episode as in NGC 205 (Welch, Sage, \\& Mitchell 1998). In several SBDGs large HI reservoirs envelope the luminous galactic body (e.g. NGC 4449: Hunter et al.\\ 1998; I Zw 18: van Zee et al.\\ 1998; NGC 1705: Meurer, Staveley-Smith, \\& Killeen 1998) and obtrude that the starburst is fueled by enhanced infall. Since the infall rate cannot yet be evaluated observationally the task for evolutionary models is to investigate the efficiency of SF and its according gas consumption.\\\\[-5mm] ", "conclusions": "\\vspace{-2mm} In the previous section we have shown that a constant infall of metal-poor matter from the enveloping \"HI reservoir\" drives a permanent gas mixing cycle keeping SNe-produced metals effectively in the gravitational field of the galaxy. Since recent deep HI observations of dIrrs and SBDGs have discovered an increasing number of large gas envelopes, i.e., of enormous gas reservoirs circling or accumulating around the visible body of the BCDGs, the relevance of infall episodes for the DG evolution is obvious and serves as the most promissing explanation for their observed abundances (see Hensler, Rieschick, \\& K\\\"oppen, 1999). Additionally we demonstrate by chemodynamical models that the SF cycle (Fig. 3) is triggered by only a small fraction of infalling matter, but produces sufficient energy to cause major evaporation and to drive more than $3/4$ of the infalling gas, incorporated into the ICM, as a galactic outflow back into an outer long-term cycle. Chemodynamical models can provide a fundamental insight into strong interactions between dynamical and energetical processes that happen in these sensitively balanced systems of low gravitational energy.\\\\[-1mm]" }, "0005/astro-ph0005286_arXiv.txt": { "abstract": "Recent observations of the size-frequency distribution of zodiacal cloud particles obtained from the cratering record on the LDEF satellite (Love and Brownlee 1993) reveal a significant large particle population (100 micron diameter or greater) near 1 AU. Our previous modeling of the Solar System dust bands (Grogan et al 1997), features of the zodiacal cloud associated with the comminution of Hirayama family asteroids, has been limited by the fact that only small particles (25 micron diameter or smaller) have been considered. This was due to the prohibitively large amount of computing power required to numerically analyze the dynamics of larger particles. The recent availability of cheap, fast processors has finally made this work possible. Models of the dust bands are created, built from individual dust particle orbits, taking into account a size-frequency distribution of the material and the dynamical history of the constituent particles. These models are able to match both the shapes and amplitudes of the dust band structures observed by IRAS in multiple wavebands. The size-frequency index, $q$, that best matches the observations is approximately 1.4, consistent with the LDEF results in that large particles are shown to dominate. However, in order to successfully model the `ten degree' band, which is usually associated with collisional activity within the Eos family, we find that the mean proper inclination of the dust particle orbits has to be approximately 9.35$^{\\circ}$, significantly different to the mean proper inclination of the Eos family (10.08$^{\\circ}$). This suggests that either the ten degree band is produced from collisional activity near the inner edge of the family or that the inclinations of dust particle orbits from the Eos family as a whole no longer trace the inclinations of their parent bodies but have been degraded since their production. ", "introduction": "A little over fifteen years ago, the phenomenon of the zodiacal light was attributed to a smooth, lenticular distribution of cometary debris, centered on the Sun, lying in the plane of the ecliptic (see Giese et. al 1986 for a review). However, the launch of IRAS in 1983 revolutionized our knowledge of the interplanetary medium. For the first time, brightness profiles of the zodiacal cloud became available which clearly showed a level of structure, particularly near the ecliptic, which could not be explained by the previous paradigm. Figure 1 shows such a brightness profile of the zodiacal cloud, along with the results of passing the profile through a fast Fourier filter to isolate the near-ecliptic features. These features appear as `shoulders' superimposed on the background emission at roughly $\\pm 10^{\\circ}$, and a `cap' near the ecliptic plane. In the discovery paper, Low et. al (1984) suggest that these dust bands are traces of collisional debris within the main asteroid belt, based on a determination of their color temperature. This is an important point: the traditional source of the interplanetary dust complex was assumed to be the debris of short period comets (Whipple 1967; Dohnanyi 1976). Although asteroid collisions should inject at least some material into the cloud, the lack of observational constraints had otherwise made the contribution of asteroidal material next to impossible to estimate. A dust band is a toroidal distribution of dust particles with common inclinations. The dust particles themselves are asteroidal collisional debris. Particles in cometary type orbits have high orbital eccentricities; planetary gravitational perturbations produce large variations in these eccentricities and these variations are coupled to those in the inclinations (Liou et al. 1995). Therefore even if a group of cometary type orbits initally had identical inclinations, planetary perturbations would disperse those inclinations over a wide range on a timescale of a few precession periods, showing that it is impossible for a comet to produce a well defined dust band. A given asteroid undergoing a collision will break up producing debris according to some size-frequency distribution. This distribution can be defined by the equation, \\begin{equation} n(D) \\sim D^{2-3q}, \\end{equation} where $D$ is the diameter of the particle. For a system in collisional equilibrium, q=11/6 (Dohnanyi 1969) and the distribution is dominated by small particles. Assuming the excess velocities after escape are small compared with the mean orbital speed of an asteroid (15-20 km/s), the orbits of individual fragments will be similar as their orbital elements will be only slightly perturbed from those of the parent asteroid (Davis et al. 1979). Even a small initial distribution in relative velocity (10-100 m/s), corresponding to a minor dispersion in semimajor axis $\\Delta a/a$ (0.1-1\\%) rapidly produces a ring of material over the parent asteroid's orbit (10$^{2}$-10$^{3}$ years). Secular precession acts upon the particles' longitude of ascending node due to the effect of Jovian perturbations. To first order, \\begin{equation} \\langle \\dot{\\Omega} \\rangle = - \\frac{3GM_{J}a^{3/2}}{4R_{J}^{3} \\sqrt{GM_{\\odot}}} \\left[ 1+\\frac{15}{8}\\frac{a^{2}}{R_{J}^{2}} \\right] \\end{equation} where $\\Omega$ is the longitude of node, $M_{J}$ is Jupiter's mass, $M_{\\odot}$ is the Sun's mass, $R_{J}$ is the mean orbital distance of Jupiter (5.2 AU), $a$ is the semimajor axis of a given particle and $G$ is the gravitational constant (Sykes and Greenberg 1986). The rate of nodal regression is found from the derivative of the above equation, \\begin{equation} \\Delta \\langle \\dot{\\Omega} \\rangle = \\frac{9GM_{J}a^{3/2}}{8R_{J}^{3} \\sqrt{GM_{\\odot}}} \\left[ 1+\\frac{35}{8}\\frac{a^{2}}{R_{J}^{2}} \\right] \\frac{\\Delta a}{a}. \\end{equation} The time taken to distribute the nodes around the ecliptic to form a dust band is then given by \\begin{equation} \\Delta t = \\frac{2\\pi}{\\Delta\\langle \\dot{\\Omega} \\rangle}. \\end{equation} For a collisional event at 2.2 AU with ejection velocities of 100 m/s, a dust band would form after approximately 2 x 10$^{6}$ years. Now since particles in inclined orbits spend a disproportionate amount of time at the extremes of their vertical harmonic oscillations, a set of such orbits with randomly distributed nodes will give rise to two apparent bands of particles symmetrically placed above and below the mean plane of the system (Neugebauer et al. 1984). This gives a natural explanation for the `shoulders' on the IRAS profiles at approximately $\\pm 10^{\\circ}$. Similarly, the central `cap' may be simply explained as a low inclination dust band. Any dispersion in the proper inclinations of the dust particles will lead to the dust band profile appearing broader, with the peak intensity shifted to a lower latitude (Dermott et al. 1990, Grogan et al. 1997). A point of debate in the literature rests on whether the dust bands are equilibrium or non-equilibrium features. In other words, are the dust bands produced by a gradual grinding down of asteroid family members, or do they represent regions of random, catastrophic disruptions in the asteroid belt? The equilibrium model, first discussed by Dermott et al. (1984) and most recently by Grogan et al. (1997), observes that the positions of the dust bands follow the locations of the major Hirayama asteroid families. This would be the natural consequence if the local volume density of dust, produced from continual asteroid erosion, followed the local volume density of asteroids. The catastrophic model follows from a discussion of dust band production rates (assuming the random disruption of a small single asteroid of approximately 15km diameter) and dust band lifetimes (material will be removed by Poynting-Robertson (P-R) drag). Following this logic Sykes and Greenberg (1986) conclude that several dust bands should be visible at any given time. This is in agreement with the IRAS observations and represents the main argument for the non-equilibrium model. The question is an important one to answer, and has implications for the investigation of the long-term evolution of the asteroid belt. If the equilibrium model proves correct, then the dust bands can be used as probes of collisional activity within their corresponding families and ultimately employed to estimate the percentage contribution of asteroidal material to the zodiacal dust complex. If the catastrophic paradigm is correct, then individual dust band features cannot be related to given asteroids in the belt with any confidence, and the question of the asteroidal contribution to the cloud will be much more difficult to unravel. ", "conclusions": "The results presented in this paper improve upon those reported in a previous paper (Grogan et al. 1997), particularly in regard to the ten degree band associated with the Eos family. In order for a dust band model to match the observations, it needs to fit both the latitude of peak flux (driven by the mean proper inclination of the particles) and the width of the dust band feature (a function of the dispersion in proper inclinations). Previously, the dispersion in proper inclinations of the Eos dust particles was reported at a relatively high 2.5$^{\\circ}$, which minimized the residuals while the mean proper inclination of the particles was fixed at the mean proper inclination of the Eos asteroid family. In this paper, smaller residuals are found when the mean proper inclination of the particles is allowed to float as a free parameter; the best fit then corresponds to a mean proper inclination of 9.35$^{\\circ}$ and a dispersion of only 1.5$^{\\circ}$. Figure 23 shows the members of the Eos asteroid family in (e,i) space as determined by the hierarchical clustering method (Zappala et al. 1995). Shown on this diagram is the position of the mean proper inclination of the ten degree band model. The consequence is that the ten degree dust band material is not tracing the orbital element space of the Eos family as a whole, as would perhaps be expected from the equilibrium model. Either the collisional activity is occurring near the inner edge of the Eos family, or the inclinations of dust particle orbits originating from the Eos family as a whole no longer trace the inclinations of their parent bodies but have been degraded since their production. If some mechanism was degrading the dust particle orbits it would presumably apply to particles from all sources, but may be more easily observed within the Eos family owing to its high inclination. Trulsen and Wikan (1980) have suggested based on their numerical simulations that the combined influence of P-R drag and collisions acts to decrease both the mean eccentricity and inclination of dust particle orbits. This subject is however open to debate; the nature of collisions between interplanetary dust particles is still poorly understood. Figure 24 shows the cumulative surface area as a function of different size-frequency distrubtion indices for the Eos, Themis and Koronis families and also a single 15km diameter asteroid. At first this appears to contradict our result that a low $q$ of around 1.43 is needed to model the dust bands. However, the diagram is set up such that size-frequency distribution is constant from the source point all the way down to the smallest IDPs, which we know is not the case since P-R drag will act to preferentially remove the small particles. In reality, the size-frequency distribution will change from the large to the small end of the distribution, and will also be a function of heliocentric distance. The diagram does suggest that for a single asteroid to be responsible for the ten degree dust band, the size-frequency index of the collisional debris would initially have needed to be extremely high to produce the surface area required to match the observations. The justification of cutting off the distribution of dust band material at 2 AU is essentially given by Figure 12. As the particles move out of the asteroid belt the action of the secular resonance disperses them into the background cloud, an effect which is more marked as the particle size increases. For this reason the Fourier filter is particularly sensitive to material located in the asteroid belt, and models that confine the material to the asteroid belt match the observations very well. In the future, our models will populate the inner Solar System as well as the main-belt region, but to do this properly we will have to: \\begin{enumerate} \\item Investigate the dynamical history of a much greater number of particle sizes than the five sizes we have considered so far in order to properly account for their behavior at the 2 AU secular resonance; \\item Take into account collisional processes: larger particles will have shorter collisional lifetimes compared to their P-R drag lifetimes and will therefore not penetrate as far into the inner Solar System. Each distribution of orbits of a given particle size will therefore have a natural inner edge defined by the lifetime of the particles in the cloud. \\end{enumerate} However, we can obtain an estimate for the dust band contribution to the zodaical cloud as a whole by simply extending our best fit dust band models to populate the inner Solar System. The distribution of orbits obtained in this manner will not be exactly correct, due to our insufficient treatment of the secular resonance, but will still be reasonably accurate in terms of the total surface area associated with the dust bands. Figure 25 compares the thermal emission obtained from this raw dust band model to the corresponding IRAS profile in the 25 $\\mu m$ waveband. The result is shown for inner Solar System distributions of material corresponding to $\\gamma$=1.0, as expected for a system evolved by P-R drag, and $\\gamma$=1.3 as predicted in parametric models of the zodiacal cloud, most recently Kelsall et al. (1998). The dust bands appear to contribute approximately 30\\% to the total thermal emission. Also shown is the amplitude of the dust band material confined to the main belt (exterior to 2 AU), which represents the component of the dust band material isolated by the fast Fourier filter. This clearly shows the extent to which the dust band contribution is underestimated if it is assumed that the filtered dust band observations represent the entirety of the dust band component of the cloud. Figure 26 shows the ratio of areas of material associated with the entire main belt asteroid population and all families, for asteroid diameters greater than 1 km. The best fit lines have a slope corresponding to a size-frequency index $q=1.795$. This diagram can be used to estimate the total contribution of main belt asteroid collisions to the dust in the zodiacal cloud, by extrapolating the observed size distributions of larger asteroids in both populations assuming a collisional equilibrium power law size distribution. The result is that the main belt asteroid population contributes approximately three times the dust area of the Hirayama families alone, and the total asteroidal contribution to the zodiacal cloud could account for almost the entireity of the interplanetary dust complex. In reality, evolved size distributions are more complex than simple power laws (Durda et al. 1998) and the size distribution of individual asteroid families likely preserve some signatures of the original fragmentation events from which they were formed. However, small dust-size particles and their immediate parent bodies have collisional lifetimes in the main belt that are considerably shorter than the age of the Solar System or the major asteroid families. Thus the dust size distributions associated with both the background main belt and family asteroids may well be considered to have achieved an equilibrium state, with total areas related to the equivalent volumes of the original source bodies in each population. An alternative, and perhaps more satisfactory, approach to obtaining the total asteroidal contribution to the zodiacal cloud will be to apply our methods to the main-belt asteroid population in the same way we have investigated the dust bands. This is the subject of a future paper. The origin of the large dispersion in proper inclination (1.5$^{\\circ}$) required to successfully model the ten degree band, in rough agreement with the 1.4$^{\\circ}$ found by Sykes (1990) and the 2$^{\\circ}$ found by Reach et al. (1997), remains unclear. Dispersion in inclination due to the Lorentz force is expected to behave such that the root mean square of the dispersion will increase with the square root of the distance traveled, and will be inversely proportional to the cube of the radius of the particle (Leinert and Gr\\\"{u}n 1990). Morfill and Gr\\\"{u}n (1979) report a value of only 0.3$^{\\circ}$ for a particle of 1 $\\mu m$ radius by the time it has spiraled in to 1 AU from the asteroid belt after 3000 years, with that expected for a 100 micron particle to be significantly less. Subsequent treatments by Consomagno (1979), Barge et al. (1982) and Wallis and Hassan (1985) differ by more than an order of magnitude due to the lack of detailed knowledge of the magnetic field structure. A more likely source of the dispersion is simply the action of the secular resonance at 2 AU. However, this leaves open the question of why a large dispersion is required to model the ten degree band, and only the small dispersion of the Themis and Koronis families is required to successfully reproduce the central band observations. One answer may be that the emission associated with the central band is due to relatively recent collisions within these families. Figure 27 shows the variation with time of the total cross-sectional area associated with the main belt and describes the stochastic breakup of asteroidal fragments. This numerical approach to describing the collisional evolution of the asteroid belt is detailed by Durda and Dermott (1997). The initial main belt mass is taken to be approximately three times greater than the present mass (Durda et al. 1998); this population evolves after 4.5 Gyr to resemble the current main belt. The calculation is performed for particles from 100 $\\mu m$ through the largest asteroidal sizes, with a fragmentation index $q=1.90$. The dust production rate in the main asteroid belt becomes more stochastic with time following a relatively smooth decrease in area as the small particles are created directly from the breakup of the parent body are destroyed. The spikes in the dust production are due to the breakup of small to intermediate size asteroids. Therefore while the observable volume of a family may decay at a fairly constant and well-defined rate, the total area of dust associated with the family during that time may fluctuate by an order of magnitude or more. We have shown in this paper how the Solar System dust bands can be investigated and used as a tool for addressing fundamental questions about the nature of the zodiacal cloud and the origin of the material from which it is composed. A key component of this process has been the realization that large particles play a dominating role in the structure of the cloud and that their dynamical histories need to be included in any physically motivated model. In the future we will extend our knowledge of the dust dynamics to a wider range of particle sizes, and address the main-belt contribution as well as the dust band component on the way to our ultimate goal of providing a global model for the zodiacal emission. \\clearpage \\vspace{-0.5cm}" }, "0005/astro-ph0005309_arXiv.txt": { "abstract": "It has been suggested that ultrahigh energy neutrinos can acquire cross-sections approaching hadronic size if the string scale is as low as $1-10$~TeV. In this case, the vertical air showers observed with energies above the Greisen-Zatsepin-Kuzmin cutoff at $E\\ap 6\\cdot 10^{19}$~eV could be initiated by neutrinos which are the only known primaries able to travel long distances unimpeded. We have calculated the neutrino-nucleon cross-section $\\skk$ due to the exchange of Kaluza-Klein excitations of the graviton in a field theoretical framework. We have found that $\\skk$ and the transferred energy per interaction are too small to explain vertical showers even in the most optimistic scenario. ", "introduction": "Several experiments using different techniques have observed ultrahigh energy cosmic rays (UHE CR) with energies up to $3\\cdot 10^{20}$eV~\\cite{rev,exp}. The isotropy of the UHE CR arrival directions argues for their extragalactic origin, since galactic and extragalactic magnetic fields cannot isotropize charged particles of such energies. However, all known extragalactic sources of UHE CR, such as AGN~\\cite{bi}, topological defects~\\cite{td} or the Local Supercluster~\\cite{bg}, result in a well pronounced Greisen-Zatsepin-Kuzmin (GZK) cutoff~\\cite{GZK} at $E_{\\rm GZK}\\ap 6\\cdot 10^{19}$~eV, although in some cases the cutoff energy is shifted closer to $1\\cdot 10^{20}$~eV \\cite{bg}. As possible solution to this puzzle it has been proposed that the UHE primaries initiating the observed air showers are not protons, nuclei or photons but neutrinos~\\cite{alt,do98,ja00}. Neutrinos are the only known stable particles which can traverse extragalactic space even at energies $E\\gsim E_{\\rm GZK}$ without attenuation, thus avoiding the GZK cutoff. Although only two dozens of UHE events have been observed, we can pin down rather precisely the required interactions of UHE neutrinos. Since the shower profile of Fly's Eyes highest energy event with $E\\ap 3\\cdot 10^{20}$~eV is well fitted by a proton \\cite{p} and also the lateral electron and muon distributions observed by AGASA are consistent with this hypothesis, neutrino-nucleon interactions should mimic nucleon-nucleon interactions at cms energies $\\sqrt{s}\\ap 500$~TeV. In particular, the neutrino-nucleon cross-section should reach $\\sigma=100-200$~mbarn, while the average energy fraction $y$ transferred per interaction to the shower should be large, $y\\ap 0.6$. Most models introducing new physics at a scale $M$ to produce large cross-sections for UHE neutrinos fail because experiments generally constrain $M$ to be larger than the weak scale, $M\\gsim m_Z$, and unitarity limits cross-sections to be ${\\cal O}(\\stot)\\lsim 1/M^2 \\lsim 1/m_Z^2$~\\cite{bu98}. String theories with large extra dimensions \\cite{ex} are different in this respect: if the Standard Model (SM) particles are confined to the usual $3+1$-dimensional space and only gravity propagates in the higher-dimensional space, the compactification radius $R$ of the large extra dimensions can be large, corresponding to a small scale $1/R$ of new physics. The weakness of gravitational interactions is a consequence of the large compactification radius, since Newton's constant is then given by $G_N^{-1}=8\\pi R^{\\delta} M_D^{\\delta+2}$, where $\\delta$ is the number of extra dimensions and $M_D\\sim\\;$TeV is the fundamental mass scale. Such a scenario is naturally realized in theories of open strings \\cite{type1}, where SM particles correspond to open strings beginning and ending on D-branes, whereas gravitons correspond to closed strings which can propagate in the higher-dimensional space. From a four-dimensional point of view the higher dimensional graviton in these theories appears as an infinite tower of Kaluza-Klein (KK) excitations with mass squared $m_{\\vec{n}}^2=\\vec{n}^2/R^2$. Since the weakness of the gravitational interaction is partially compensated by the large number of KK states and cross-sections of reactions mediated by spin 2 particles are increasing rapidly with energy, it has been argued in Refs.~\\cite{do98,ja00} that neutrinos could initiate the observed vertical showers at the highest energies. In the calculations of Refs.~\\cite{ja00,nu99} it was assumed that the massless four-dimensional graviton and its massive KK excitations couple with the usual gravitational strength $\\overline{M}_{\\mbox{\\small Pl}}^{-1}= \\sqrt{8\\pi}/M_{\\mbox{\\small Pl}}$. Then the sum over all KK contributions to a given scattering amplitude only converges in the case of one extra dimension, and for two or more extra dimensions a cutoff has to be introduced by hand. However, it has recently been pointed out \\cite{fluct} that due to brane fluctuations the effective coupling $g_{\\vec{n}}$ of the level $\\vec{n}$ KK mode to four-dimensional fields is suppressed exponentially, \\begin{equation} g_{\\vec{n}}={1\\over\\overline{M}_{\\mbox{\\small Pl}}} \\exp\\left(-{c\\,m_{\\vec{n}}^2\\over \\Ms^2}\\right)\\;, \\label{coupling} \\end{equation} where $c$ is a constant of order $1$ or larger, which parametrizes the effects of a finite brane tension \\cite{fluct}, and $\\Ms$ is the string scale. This exponential suppression thereby provides a dynamical cutoff in the sum over all KK modes. We have recalculated the neutrino-nucleon cross-section $\\skk$ due to the exchange of KK gravitons in a four-dimensional, effective field theory valid for $s\\lsim\\Ms^2$, taking this dynamical cutoff into account. As a consequence, our result for $\\skk$ is considerably smaller than in previous calculations~\\cite{nu99}. Since we have found moreover that the energy transfer per interaction is small at the energies of interest, $y\\ap 0.1$, neutrinos behave like deeply penetrating particles and cannot explain the observed vertical air showers. This conclusion holds even if one extrapolates the unitarity violating cross-section valid for $s\\lsim\\Ms^2$ to the region $s\\gsim\\Ms^2$. We have also derived an upper bound for $\\skk$ consistent with unitarity for $s\\gsim\\Ms^2$ using the eikonal method. In this case, the resulting cross-section respects the Froissart bound and is numerically too small to lead to observable consequences in UHE CR experiments. ", "conclusions": "We have calculated the neutrino-nucleon cross-section $\\skk$ due to the exchange of KK excitations of the graviton taking into account the exponential suppression of modes with $m_{\\vec n}^2\\gsim \\Ms^2$. Because of the smallness of the resulting cross-section and energy transfer per interaction, the neutrino behaves also in these theories as a deeply penetrating particle. In the case that the cross-section $\\skk\\propto s^2$ continues to grow for $s\\gsim\\Ms^2$, thereby violating four-dimensional unitarity, future UHE CR experiments like AUGER or OWL could be more sensitive to large extra dimensions than LHC. However, an accurate determination of the sensitivity of these experiments would require a string-theoretical calculation of the neutrino-nucleon cross section." }, "0005/astro-ph0005401_arXiv.txt": { "abstract": "The interstellar medium of Centaurus A (NGC 5128) has been studied extensively in recent years, using mostly molecular lines tracing low to medium density gas (500 to several $10^3$ cm$^{-3}$). The amount and distribution of the dense molecular gas was largely unknown. Here we present new millimeter data of the HCN(1--0), CS(2--1), and CS(3--2) rotational transitions towards the nearby radio galaxy Centaurus~A observed with the SEST on La Silla, Chile. We obtained spectra of the HCN(1--0) emission which traces dense 10$^4$ cm$^{-3}$ molecular gas at the center and along the prominent dust lane at offset positions $\\pm$60$^{\\prime \\prime}$ and $\\pm$100$^{\\prime \\prime}$. We also obtained a few spectra of CS(2--1) and (3--2) tracing densities of $\\sim$10$^5$ cm$^{-3}$. The emission in these lines is weak and reaches a few mK at the available angular resolutions of 54$^{\\prime \\prime}$ - 36$^{\\prime \\prime}$. At the central position, the integrated intensity ratio I(HCN)/I(CO) peaks at 0.064, and decreases to $\\sim$0.02 to 0.04 in the dust lane. Using the new high density tracer data, we estimate the amount, distribution and physical conditions of the dense molecular gas in the dust lane of Centaurus A. We find that Cen~A and the Milky Way are comparable in their HCN(1-0) line luminosity. However, towards the nucleus the fraction of dense molecular gas measured via the line luminosity ratio L(HCN)/L(CO) as well as the star formation efficiency $L_{\\rm FIR}/L_{\\rm CO}$ is comparable to ultra-luminous infrared galaxies (ULIRGs). Within the off-nuclear dust lane and for Cen~A as a whole these quantities are between those of ULIRGs and normal and infrared luminous galaxies. This suggests that most of the FIR luminosity of Centaurus~A originates in regions of very dense molecular gas and high star formation efficiency. ", "introduction": "Centaurus~A (NGC 5128) is the closest radio galaxy (distance 3.5 Mpc, 1$^{\\prime \\prime}=$ 17 pc, Soria et al. 1996, Hui et al. 1993, Israel 1998, Ebneter \\& Balick 1983, de Vaucouleurs 1979) and exhibits a very prominent dust lane. Centaurus A is a strong radio galaxy with a milliarcsecond nuclear continuum source (Kellermann et al. 1997; Shaffer \\& Schilizzi 1975; Kellermann 1974) and two giant radio lobes. Absorption against the nuclear source has been found in H{\\sc I} (van der Hulst et al. 1983) and many molecular species and transitions (Gardner \\& Whiteoak 1976; Whiteoak \\& Gardner 1971; Bell \\& Seaquist 1988; Seaquist \\& Bell 1986, 1990; Phillips et al. 1987; Eckart et al. 1990a; Israel et al. 1990, 1991; Wild et al. 1997; Wiklind \\& Combes 1997). Several studies of Centaurus~A in the millimeter wavelength range have been carried out. Although other elliptical galaxies with dust lanes have recently been detected in CO (Sage \\& Galleta 1993), Centaurus is the best object for a detailed study due to its proximity and corresponding large angular size. Phillips et al. (1987) and Quillen et al. (1992) observed several positions in the CO(2-1) line along the dust lane at a resolution of 30$^{\\prime \\prime}$. In previous papers we presented a fully sampled map of the $^{12}$CO(1-0) emission together with {\\it IRAS} observations of the FIR continuum (Eckart et al. 1990b), measurements of the millimeter absorptions lines towards the nucleus (Eckart et al. 1990a; see also Wiklind \\& Combes 1997), and a $^{12}$CO(2-1) map along the dust lane at a resolution of 22$^{\\prime \\prime}$ (Rydbeck et al. 1993). In Wild et al. (1997) we presented the first fully sampled $^{13}$CO(1-0) map along the dust lane of Centaurus~A, as well as single spectra of the $^{13}$CO(2-1) emission in the disk and C$^{18}$O(1-0) emission at the central position. These new data allowed us, in combination with the $^{12}$CO(1-0) and $^{12}$CO(2-1) maps obtained earlier, to study the excitation conditions of the molecular gas in detail throughout the dust lane. Using different CO line ratios and their variation across the disk of Centaurus A, we inferred the physical parameters of the molecular ISM and their spatial variations. Here we extend our investigation to the very dense (10$^4$ to 10$^5$cm$^{-3}$) phase of the molecular gas by observing line emission of the density tracers HCN and CS towards the nucleus and the off-nuclear dust lane in Centaurus~A. \\begin{table}[b] \\caption{Observational parameters} \\label{obstab} \\begin{tabular}{lcrrr}\\hline \\hline Transition & offset & integration & $T^{*}_{\\rm A,max}$ & $\\Delta v$ \\\\ & RA, Dec & time \\\\ & (arcsec) & (min) & (mK) & (km/s) \\\\ \\hline HCN(1-0) & +105,-73 & 178 & $<$2.0 & \\\\ & +79,-55 & 460 & 4.0 & 230 \\\\ & +50,-25 & 396 & 5.0& 190 \\\\ & 0,0 & 252 & 8.5 & 390 \\\\ & -52,+37 & 836 & 3.8 & 170 \\\\ & -79,+55 & 360 & 4.0 & 170 \\\\ \\hline CS(3-2) & +50,-25 & 1644 & 0.9 & 250 \\\\ & 0,0 & 712 & 3.8 & 105 \\\\ CS(2-1) & -52,+37 & 288 & 3.0 & 135 \\\\ \\hline \\hline \\end{tabular} \\end{table} \\begin{figure} \\resizebox{\\hsize}{!}{\\includegraphics{h2024f1.eps}} \\hfill \\caption{ Beam positions at which the dense molecular gas in the dust lane of Centaurus~A has been investigated. Shown is the beam for the HCN J=1--0 measurements.} \\label{figbeams} \\end{figure} \\begin{figure} \\resizebox{\\hsize}{!}{\\includegraphics{h2024f2.eps}} \\hfill \\caption{ HCN J=1--0 spectra along the dust lane of Centaurus~A. Offsets are given relative to the center position $\\alpha(1950)=13^{h}22^{m}31.8^{s}$, $\\delta(1950)=-42^{\\circ}45^{\\prime}30^{\\prime \\prime}$ } \\label{fighcn} \\end{figure} \\begin{figure} \\resizebox{\\hsize}{!}{\\includegraphics{h2024f3.eps}} \\hfill \\caption{ CS J=3--2 and J=2--1 spectra along the dust lane of Centaurus~A. Offsets are given relative to the center position $\\alpha(1950)=13^{h}22^{m}31.8^{s}$, $\\delta(1950)=-42^{\\circ}45^{\\prime}30^{\\prime \\prime}$ } \\label{figcs} \\end{figure} ", "conclusions": "While the line luminosities of the HCN(1-0) and CS(3-2) as well as the FIR luminosity of the molecular gas in the dust lane of Cen~A are quite comparable to each other there are definite differences in the overall fraction of dense molecular gas and the efficiency with which stars are formed from it. This star formation activity is also the source of the FIR luminosity of Cen~A. About 40\\% or even more of the total molecular line luminosity in Cen~A originates in dense gas. This suggests that star formation as well as the bulk of the dense molecular gas is mostly concentrated in GMC complexes rather than in a more diffuse molecular gas component. This is already indicated by the ring-like distribution of HII region found by Graham (1979) as well as the MEM deconvolved $^{12}$CO(2-1) line emission mass by Rydbeck et al. (1993). We also note that with respect to other positions in the dust lane the I(HCN)/I(CO) ratio is larger at separations of about 100$^{\\prime \\prime}$ from the nucleus rather than at separations of about 60$^{\\prime \\prime}$. The larger distance is close to the inner edge of the ring of HII regions and corresponds well with the position of the folds in the warped molecular gas disk of Centaurus~A (Quillen et al. 1992, 1993; Sparke 1996) and an increased intensity in the 15$\\mu$m continuum dust emission (Block \\& Sauvage 2000; Mirabel et al. 1999). Therefore a higher I(HCN)/I(CO) ratio may be due to a combination of enhanced star formation efficiency at these positions and an increase in column density due to the folds. Due to the low intrinsic velocity dispersion of the thin molecular disk (Quillen et al. 1992) and due to the fact that the $^{12}$CO line emission is originating in optically thick molecular gas (Wild et al. 1997) this may lead to shadowing of molecular clouds along the line of sight toward the folds. This effect will be stronger for $^{12}$CO than for the small, dense cloud cores seen in the less abundant HCN line emission. This effect may therefore lead to an intensity decrease in the $^{12}$CO line and an increase in the HCN(1-0) line, resulting in the observed variation of the I(HCN)/I(CO) line ratio. Future interferometric measurements will allow us to study the distribution of molecular gas in the dust lane of Centaurus~A in much greater detail. Line ratios, luminosities and star formation can then be investigated for individual GMC complexes." }, "0005/astro-ph0005292_arXiv.txt": { "abstract": "Numerical solutions of the isothermal Lane-Emden equation are presented, corresponding to self-gravitating gaseous cores embedded within a finite density envelope of overall cylindrical symmetry. These structures may be members of a fragmentation hierarchy proceeding from sheets, to filaments, to elongated, prolate clumps. The embedded solutions are the first of their kind, and as such represent a significant improvement upon the isolated cloud paradigm used almost exclusively by previous authors. The properties of the equilibria are in reasonable agreement with observations of dense molecular cores in star-forming clouds, despite the fact that there is only one free parameter in the models. We show that this parameter may be identified with the critical wavelength for instability in the parent filament. The implications of further fragmentation and the possible influence of magnetic fields are briefly discussed. ", "introduction": "\\label{sec-intro} The present observational picture of star-forming regions conveys cloud structures considerably more varied than any one theoretical scenario can explain. Nevertheless, there are certain points of correspondence, when one restricts attention to the molecular, self-gravitating component of the interstellar medium. On scales of tens to hundreds of parsecs, both atomic and molecular gas clouds appear shell-like and filamentary (Scalo 1985; Kulkarni \\& Heiles 1988). Shells may be formed by several distinct dynamical processes: cloud-cloud collisions (Smith 1980), compressional shock waves from supernovae or OB stars (McCray \\& Kafatos 1987), and large-scale shocks associated with spiral density waves (Roberts 1969) are but a few possibilities. Filamentary structure persists down to scales of several parsecs, to the regime of individual molecular clouds [Loren 1989, Nozawa et al 1991 ($\\rho$ Oph); Heyer et al 1987, Onishi et al 1996 (Taurus); Bally et al 1987, Tatematsu et al 1993 (Orion)]. Each filament typically contains several distinct subcondensations in close proximity to each other, some of which harbor infrared continuum sources (Onishi et al 1998). In some cases, the embedded clumps are spaced quite regularly along the filament (Schneider \\& Elmegreen 1979; Dutrey et al 1991). Thus, several authors have speculated that the formation of star-forming clumps proceeds via a hierarchical fragmentation process, in which filaments are formed out of larger structures, and then clumps out of the filaments (Schneider \\& Elmegreen 1979; Gaida, Ungerechts, \\& Winnewisser 1984; Hanawa et al 1993; Fiege \\& Pudritz 2000$a$). Focusing now on the cores themselves, maps in dense tracers such as NH$_3$ and CS display roughly elliptical intensity contours, with a mean apparent major-to-minor axis ratio of around 2 (Jijina, Myers, \\& Adams 1999). Statistical arguments applied to the distribution of measured axial ratios for several surveys have prompted some authors to conclude that the cores are more likely to be intrinsically prolate than oblate (David \\& Verschueren 1987; Myers et al 1991; Ryden 1996). Other statistical arguments imply that the observed elongation is unlikely to be a result of star formation or outflows (Myers et al 1991). Coupled with the additional result that the majority of cores are near virial equilibrium (Jijina et al 1999), this argues against a wholly dynamical origin for prolateness, since the implied lifetimes are then so short that observations of cores without embedded stars would be exceedingly rare (the latter represent about one-half of the ammonia cores detected in the Taurus region, and a somewhat smaller fraction in Ophiuchus and Orion). Rather, it appears more likely that cores are at least quasi-equilibrium structures, and that their shapes therefore offer some clue to the forces responsible for their formation. If this reasoning is correct, then it leads to a formidable crisis in our current theoretical picture of cloud equilibria, which typically envisions star forming cores as self-gravitating clumps bounded by a zero density, constant pressure---and so high temperature---medium (e.g., McKee et al 1993). That is, it is difficult to conceive of prolate, quasi-equilibrium, gaseous cores as purely {\\it isolated} structures. Non-gravitational forces, such as rotation or magnetic fields, are not likely to aid in maintaining prolate equilibria, although their role in oblate structures has been made abundantly clear.\\fn{It is in fact possible to construct isolated, prolate magnetic clouds, but these possess magnetic field structures which are highly unusual and, in any case, lack direct observational justification. See Fiege \\& Pudritz (2000$b$) and Curry \\& Stahler (2000) for examples.} The absence in the literature of even a single, physically acceptable, gaseous, prolate equilibrium solution suggests that a fresh theoretical approach is necessary. We examine a scenario in this paper whereby core morphology is directly attributable to a fragmentation process. As we show in \\S \\ref{sec-plane} and \\ref{sec-fil}, the existence of prolate structures within cylindrical filaments may be understood in exactly the same way as the existence of the filaments themselves: as a result of the fragmentation of the parent cloud and nonlinear growth of the fragments. What has apparently escaped attention until now, and what we demonstrate explicitly, is that {\\it accessible, long-lived states exist wherein cores and their extensive filamentary envelopes occupy the same hydrostatic structure}. We now proceed to outline a particular fragmentation hierarchy that might be responsible for such structures. ", "conclusions": "We have presented a new family of 2D, numerical solutions of the isothermal Lane-Emden equation in cylindrical symmetry. The equilibria have an embedded, periodic structure, and in this sense are direct counterparts to the 2D solution of Schmid-Burgk (1967) in planar symmetry. Together, the existence of the two equilibria suggest a fragmentation scheme which seems consistent with the observed hierarchical structure in several well known star-forming regions. Moreover, our results constitute a remarkably simple yet robust explanation for the origin and maintenance of the prolate, gaseous cores that represent the lower rung of this hierarchy. Although we have ignored many physical effects that should be included in subsequent studies (particularly nonthermal motions and magnetic fields), we expect that the shift in emphasis from isolated to embedded structures will prove fruitful in future theoretical work on the origin and evolution of dense cores. \\\\\\\\ \\noi It is a pleasure to thank Steve Stahler for discussions that prompted a search for these solutions. I am grateful to Richard Larson and Dean McLaughlin for useful comments on the manuscript, and to an anonymous referee for suggestions which helped to clarify certain properties of the equilibria. \\begin{appendix}" }, "0005/astro-ph0005547_arXiv.txt": { "abstract": "We present a new catalogue of symbiotic stars. In our list we include 188 symbiotic stars as well as 28 objects suspected of being symbiotic. For each star, we present basic observational material: coordinates, $V$ and $K$ magnitudes, ultraviolet (UV), infrared (IR), X-ray and radio observations. We also list the spectral type of the cool component, the maximum ionization potential observed, references to finding charts, spectra, classifications and recent papers discussing the physical parameters and nature of each object. Moreover, we present the orbital photometric ephemerides and orbital elements of known symbiotic binaries, pulsational periods for symbiotic Miras, Hipparcos parallaxes and information about outbursts and flickering. ", "introduction": "Symbiotic stars are interacting binaries, in which an evolved giant transfers material to much hotter, compact companion. In a typical configuration, a symbiotic binary comprises a red giant transferring material to a white dwarf via a stellar wind. Amongst the evidence for this predominant mass-transfer mechanism is the fact that ellipsoidal light variations, characteristic of tidally distorted stars, are rarely observed for symbiotic stars. Thus far, only two systems, T~CrB (\\citetb{46}) and CI~Cyg (\\citetb{m229}), are known to have the ellipsoidal light variations of a distorted giant. In some symbiotic systems, the red giant is replaced by a yellow giant or a carbon star, and the white dwarf by a main-sequence or neutron star. Most symbiotic stars ($\\sim 80\\%$) contain a normal giant star and these, based on their near-IR colours (showing the presence of stellar photospheres, $T_{\\rm eff} \\sim$\\ 3000 -- 4000\\,{\\sc k}), are classified as S-type systems ({\\it stellar}). The remainder contain Mira variables and their near-IR colours indicate temperatures of $\\sim 1000$\\,{\\sc k}, giving away the presence of warm dust shells; these are classified as D-type systems ({\\it dusty}). The IR type seems to be dependent on the orbital separation of the components. For large separations (long periods), the cool star seems able to evolve to the Mira stage and produce a dust shell that enshrouds the system; for smaller separations (shorter periods), we deal with normal giants. For a detailed review of symbiotic stars, we refer the reader to \\citetb{358}. Two catalogues of symbiotic stars have been published. The first was by David Allen in 1984 (\\citetb{30}); it included 129 symbiotic stars and 15 possible symbiotic objects with a concise summary of available observational data, finding charts and optical spectra for the most of listed objects. The second catalogue was by Scott Kenyon in 1986 (\\citetb{31}); it included 133 symbiotic stars and 20 possible symbiotic objects, as well as tables describing selected observational properties of all the objects and a spectroscopic summary of a selected sample. Kenyon's work also provides the reader with an excellent overview and bibliography of selected symbiotic stars. Since 1986, a number of papers have presented surveys of large samples of symbiotic stars (e.g.\\ \\citetb{7,vwds93,ibm94,sih95,79,3,313}) and in-depth investigations (\\citetb{ii88,37,ibesm93,sk94,go96} for AX~Per alone). New stars have been included in the family of symbiotic stars each year and, at the same time, better data have been collected and better data analysis has been performed for a number of well-known symbiotic stars. The aim of this work is to present the symbiotic star research community with a comprehensive compilation of existing data collected from a number of astronomical journals, electronic databases and unpublished data resources. For many objects a new classification has been necessary: some have been confirmed as symbiotic stars; some have been rejected; some new objects have been added. Our catalogue lists 188 symbiotic stars and 28 objects suspected of being symbiotic stars. ", "conclusions": "" }, "0005/hep-ph0005244_arXiv.txt": { "abstract": "We study the effect of random matter density fluctuations in the sun on resonant neutrino conversion in matter. We assume no specific mechanism for generation of the fluctuation and we keep the amplitude and correlation length as independent parameters. We do not work under the approximation that fluctuations have spatial correlations only over distances small compared to the neutrino oscillation lengths. Instead we solve numerically the evolution equation for the neutrino system including the full effect of the random matter density fluctuations of given amplitude and correlation length. In order to establish the possible effect on the MSW solutions to the solar neutrino problem we perform a global analysis of all the existing observables including the measured total rates as well as the Super--Kamiokande measurement on the time dependence of the event rates during the day and night and the recoil electron energy spectrum. We find the effects of random noise to be larger for small mixing angles and they are mostly important for correlation lengths in the range few 100 km $\\lesssim L_0\\lesssim$ few 1000 km. They can be understood as due to a parametric resonance occuring when the phase acquired by the oscillating neutrino state on one fluctuation length $L_0$ is a multiple of 2$\\pi$. We find that this resonant parametric condition is mainly achieved for low energy neutrinos such as the pp-neutrinos and therefore its effect is mostly seen on the total event rates while the other Super--Kamiokande observables are very marginally sensitive to the presence of noise due to the higher energy threshold. ", "introduction": "Solar neutrinos were first detected already three decades ago in the Homestake experiment \\cite{homestake0} and from the very beginning it was pointed out the puzzling issue of the deficit in the observed rate as compared to the theoretical expectation based on the standard solar model \\cite{SSMold} with the implicit assumption that neutrinos created in the solar interior reach the Earth unchanged, i.e. they are massless and have only standard properties and interactions. This discrepancy led to a change in the original goal of using solar neutrinos to probe the properties of the solar interior towards the study of the properties of the neutrino itself and it triggered an intense activity both theoretical as well as experimental, with new measurements being proposed in order to address the origin of the deficit. On the theoretical side, enormous progress has been done in the improvement of solar modelling and calculation of nuclear cross sections. For example, helioseismological observations have now established that diffusion is occurring and by now most solar models incorporate the effects of helium and heavy element diffusion~\\cite{Bahcall:1997qw,Bahcall:1995bt}. From the experimental point of view the situation is now much richer. Four additional experiments to the original Chlorine experiment at Homestake \\cite{homestake} have also detected solar neutrinos: the radiochemical Gallium experiments on $pp$ neutrinos, GALLEX \\cite{gallex} and SAGE~\\cite{sage}, and the water Cerenkov detectors Kamiokande~\\cite{kamioka} and Super--Kamiokande \\cite{sk1,sk99}. The latter have been able not only to confirm the original detection of solar neutrinos at lower rates than predicted by standard solar models, but also to demonstrate directly that the neutrinos come from the Sun by showing that recoil electrons are scattered in the direction along the Sun--Earth axis. Moreover, they have also provided us with useful information on the time dependence of the event rates during the day and night, as well as a measurement of the recoil electron energy spectrum. After 825 days of operation, Super--Kamiokande has also presented preliminary results on the seasonal variation of the neutrino event rates, an issue which will become important in discriminating the MSW scenario from the possibility of neutrino oscillations in vacuum~\\cite{ourseasonal,v99}. At the present stage, the quality of the experiments themselves and the robustness of the theory give us confidence that in order to describe the data one must depart from the Standard Model (SM) of particle physics interactions by endowing neutrinos with new properties. In theories beyond the SM, neutrinos may naturally have new properties, the most generic of which is the existence of mass. It is undeniable that the most popular explanation of the solar neutrino anomaly is in terms of neutrino masses and mixing leading to neutrino oscillations either in {\\sl vacuum}~\\cite{Glashow:1987jj} or via the matter-enhanced {\\sl MSW mechanism}~\\cite{msw}. The standard MSW analysis is based on a mean--field treatment of the solar background through which the neutrinos propagate. In this approximation the global analysis of the full neutrino data sample described above \\cite{oursolar} leads to the existence of three allowed regions in the $\\Delta m^2$ $\\sin^2 2\\theta$ parameter space for neutrino oscillations \\begin{itemize} \\item[$\\bullet$] non-adiabatic-matter-enhanced oscillations or small mixing angle (SMA) region with $ \\Delta m^2=(0.4$--$1)\\times 10^{-5}$ eV$^2$ and $\\sin^2(2\\theta)=(1$--$10)\\times 10^{-3} $, and \\item[$\\bullet$] large mixing (LMA) region $\\Delta m^2=(0.2$--$5)\\times 10^{-4}$ eV$^2$ and $\\sin^2(2\\theta)=0.6$--$1$. \\item[$\\bullet$] low mass solution (LOW) $\\Delta m^2=(0.3$--$2)\\times 10^{-7}$ eV$^2$ and $\\sin^2(2\\theta)=0.8$--$1$. \\end{itemize} There are several works in the literature \\cite{Krastev,otherper,other,Burgess0,Nunokawa,Burgess} where corrections to such mean--field picture have been studied. The influence of periodic matter density fluctuations of given amplitude and fixed frequency above the average density on resonant neutrino conversion was investigated in Refs.\\cite{Krastev,otherper}. In Ref.~\\cite{Krastev} a parametric resonance is found when the fixed frequency of the perturbation is close to the neutrino oscillation frequency. This approach however gives not answer to the physical origin of such fixed frequency perturbation. More recently, the main approach to fluctuations that has been pursued \\cite{other,Burgess0,Nunokawa} is to model the matter density as a Gaussian random variable (white noise). In this approach the number of free parameters remains the same as in the case of fixed frequency perturbations -- two, the perturbation amplitude and the correlation length-- but the presence of white noise in the sun is doubtless since there are many mechanism to generate random density perturbations. For technical reasons these analysis where performed under the assumption that fluctuations have spatial correlations only over distances small compared to the neutrino oscillation lengths. Within this approximation the conclusions obtained were that such fluctuations in the solar electron density can significantly modify the MSW solutions to the solar neutrino problem (SNP) provided that their relative amplitude near the MSW resonance point can be as large as few percent. In Ref.~\\cite{Burgess} criticisms to these results were raised based on two facts: i) the unexistence of a plausible source for such $\\delta$-correlated fluctuations in the vicinity of the MSW resonance point and ii) whatever its origin, the effect of the density perturbation is maximum in the regime where the short correlation--length approximation fails. In particular in Ref.~\\cite{Burgess} they concentrate on helioseismological waves as origin of the perturbation and in particular on g-waves whose amplitude increases with the solar depth and can, in principle, reach the interesting values to affect neutrino propagation. However they conclude that such g-waves do not affect the MSW neutrino conversions since the wavelength for the lower modes, for which the largest amplitudes are possible, is much longer than the characteristic MSW neutrino oscillation length. It has been recently argued \\cite{Dzhalilov}, however, that in the magnetohydrodynamical (MHD) generalization of the Helioseismology the objection in Ref.~\\cite{Burgess} does not hold. Assuming modest central large-scale magnetic fields ($B_0=$1--10--100~Gauss) one can find magneto-gravity eigenmodes with much shorter wave lengths for density perturbations $\\lambda_{MHD}\\sim 200-2000~$km \\cite{Dzhalilov} that is comparable with the neutrino oscillation length at the MSW resonance for large and small mixing angles correspondingly. In this paper we revisit the problem of the effect of matter of density fluctuations in the sun on the MSW solutions to the SNP. In our approach we assume no specific mechanism for generation of the fluctuation and we keep the amplitude and correlation length as independent parameters. There are two main differences in our analysis as compared to those in Refs.~ \\cite{other,Burgess0,Nunokawa,Burgess}. First we do not work under the approximation that fluctuations have spatial correlations only over distances small compared to the neutrino oscillation lengths. Instead we solve numerically the evolution equation for the neutrino system including the full effect of the random matter density fluctuations of given amplitude and correlation length. Second, in order to establish the possible effect on the MSW solutions to the SNP we perform a global analysis of all the existing observables including not only the measured total rates but also the Super--Kamiokande measurement on the time dependence of the event rates during the day and night, as well as the recoil electron energy spectrum. In particular we include the regeneration effects when neutrinos cross the Earth \\cite{daynight} which were neglected in Ref. \\cite{Nunokawa}. The outline of the paper is as follows. In Sec.~\\ref{solutions} we discuss our approach to the solution of the neutrino evolution in the presence of density fluctuations. In Sec.~\\ref{parke} we briefly summarize the standard analytical approach based on the short correlation length approximation and in Sec.~\\ref{numerical} we discuss our numerical treatment and present our results for the survival probabilities as a function of the relevant oscillation and noise parameters. In Sec.~\\ref{resonance} we interpret our results for the enhancement of the survival probability in the language of parametric resonance in MSW conversions. Section~\\ref{analysis} is devoted to the statistical analysis of the solar neutrino observables in the framework of the MSW solutions of the SNP in presence of the noisy density fluctuations. We study the variation of the allowed regions of the SNP for different combinations of observables when noise fluctuations with different correlation lengths are included. Our results are summarized in Figs.~\\ref{rates}--\\ref{global} and Table~\\ref{chimin}. We show that even for noise levels as large as 4\\% the relative quality of the three allowed regions of the MSW solutions to the SNP depends on the value of the correlation length studied in the wide range $L_0=70$--$10^4$ Km and that the three allowed regions of the MSW solutions of the SNP remain valid at the present level of solar neutrino experiments. Finally in Sec.~\\ref{conclu} we discuss our results and summarize our conclusions. ", "conclusions": "\\label{conclu} In this paper we have studied the effect of density matter fluctuations in the sun on the MSW solutions to the SNP. Assuming no specific mechanism for generation of the fluctuations we have kept the amplitude and correlation length as independent parameters. Our analysis is performed under no assumption on the relative size of the correlation length of fluctuations as compared to the neutrino oscillation length. To perform such a study we have solved numerically the evolution equation for the neutrino system including the full effect of the random matter density fluctuations of given amplitude and correlation length. Our procedure is to generate a realization of the density profile for given values of the perturbation amplitude and correlation length and then to solve numerically the evolution equation for the neutrino states for that given realization of the density profile and different neutrino production points and finally to average the obtained survival probability over different density realizations (with the same amplitude and correlation length) and neutrino production points. This numerical approach of averaging the solutions $P_{ee}(r_0)$ over noise realizations is different from any previous approaches with the averaging of the Schr\\\"{o}dinger equation itself before obtaining the solution. Our results for the survival probabilities are presented in Figs.~\\ref{probdm}--\\ref{probl}. We find that the effects are larger for small mixing angles. The larger the mixing angle the shorter the correlation length needed to observe an effect. For the SMA the larger effects occur for correlation lengths in the range few 100 km $\\lesssim L_0\\lesssim$ few 1000 km. They can be understood as due to a parametric resonance occuring when the phase acquired by the oscillating neutrino state on one fluctuation length $L_0$ is a multiple of 2$\\pi$. This resonance is maximal when this condition is verified close to the MSW resonance. We find that this resonant parametric condition is mainly achieved for {\\it low neutrino energies} such as the pp-neutrinos seen in GALLEX and SAGE. Next, in order to establish the possible effect of the presence of noise on the MSW solutions to the SNP we have performed a global analysis of all the existing observables including not only the measured total rates but also the Super--Kamiokande measurement on the time dependence of the event rates during the day and night, as well as the recoil electron energy spectrum. The result of such analysis is presented in Figs.~\\ref{rates}--\\ref{global} where we plot the allowed regions for MSW neutrino oscillations in the framework of two--neutrino mixing with the Sun density profile generated from the BP98, after including random noise with amplitude $\\xi=4\\%$ and different correlation lengths $L_{0}$ (70, 200, 700, 2000 y 10000 km). The main conclusions are that the total rates are the most sensitive observables to the presence of noise. On the other hand when the many degrees of freedom corresponding to the Super--Kamiokande spectrum are included the dependence of the allowed mixing parameters on the matter noise is smoothed. This is caused by the larger sensitivity of the mean $\\langle P_{ee}\\rangle$ to the noise for low energy neutrinos. Due to its higher energy threshold, the Super--Kamiokande experiment is mostly insensitive to these effects. For the same reason one expects that the Borexino experiment would be more suitable to place bounds both on the level of neutrino noise $\\sqrt{<\\xi>^2}$ and on the correlation length $L_0$." }, "0005/astro-ph0005221_arXiv.txt": { "abstract": "The anomalously slow rotation of the solar core is just one from a remarkable lists of fundamental indications showing that the solar core is somehow coupled to the surface activity cycle. On the other hand, the atmospheric, LSND and solar neutrino problems are not consistent, therefore one or more of the neutrino experiments must be attributed - at least in part - to some phenomena other than neutrino oscillations, or a fourth neutrino is required. We use the luminosity constraint in a general case not considered yet, taking into account nuclear reactions occurring in small hot bubbles indicated to be present in the solar core (Grandpierre, 2000). The dynamic solar model fit the observed neutrino fluxes within one-and-half sigma even without oscillations. This fit is shown to be better than the present day MSW and VAC fits. An attracting perspective is obtained for a simultaneous solution of the neutrino problems and the solar core-related astrophysical problems. Predictions of the dynamic solar model are presented for the SNO measurements. ", "introduction": "Two different approaches seem to be promising in exploring the roots of the solar neutrino problems. One is the traditional particle physics approach, suggesting that neutrino oscillations are responsible for the missing solar neutrinos. Unfortunately, the small mixing angle, large mixing angle and vacuum oscillation solutions all suffer from being poor fits to the observations. Eleven years ago Bahcall (1989) wrote in his book, that the MSW neutrino oscillation solution of the solar neutrino problem (SNP) is attractive since the mixing angles and mass differences can each vary by orders of magnitude, and so it does not need fine tuning as the vacuum solution does. But not so much later Paterno and Scalia (1994) noticed that the allowed region shrinkened to a point-like area, and so \"hypotheses on non-conventional neutrino properties are strongly disfavoured, except for the matter neutrino oscillations, the latter surviving within very narrow limits\". Now even this remained point-like area (and the surrounding small 95 $\\%$ C.L. region, which extends only due to the theoretical and experimental errors) do not have a high allowance. The probability of these solutions decreased from 100 $\\%$ to cca. 10$ \\%$ for the widely regarded best fit of small mixing angle (SMA) solution (and similarly for the LMA and VAC solutions), when the SuperKamiokande (SK) rates, spectral and day-night effect data are also taken into account. One can evaluate the acceptability of the presented solutions by the probability $P(\\chi^2)$ belonging to the calculated $\\chi^2_{min}$. The rule of thumb telling that \"the value of $ \\chi^2$ for a \"moderately\" good fit is $ \\chi^2 \\approx d.o.f.$\" (Press et al., 1992) may be useful. Bahcall, Krastev and Smirnov (1998) obtained for the SMA global fit $\\chi^2_{min} = 26.5/17 d.o.f.$ which is acceptable at the probability $P(\\chi^2)$ = 7 $ \\%$ C.L. Suzuki (1998) obtained $\\chi^2_{min} = 50.2/31 d.o.f.$ for the case when oscillations are not allowed. The corresponding probability is $P(\\chi^2) = 1.6 \\%$. The best fit he obtained is for the vacuum oscillation (VAC), with $\\chi^2_{min} = 38.7/31$ d.o.f., with $P(\\chi^2) \\approx 20 \\% $. This latter case we recognize also as a solution worse than moderately good, since $\\chi^2_{min} > d.o.f.$ At the same time, Bahcall, Krastev and Smirnov (1998) obtained a global fit (when the constraints from the rates, the spectrum shape and the Day-Night asymmetry are all included) to for the best fit vacuum oscillation solution $\\chi_{min}^2=28.4$ for 18 d.o.f., which is acceptable only at 6$ \\%$ C.L. It may be disturbing that two different and contrasting confidences are introduced to characterise the acceptance of a fit. One is $P(\\chi^2)$, the probability of a fit. The other one, $P( \\Delta \\chi ^2)$, characterises how far the (allegedly normal) distribution of the experimental and theoretical errors extends. Since the larger errors leads to larger allowed regions, and the farther iso-probability contours belong to increasing $n$ characterising the distance as $n \\sigma$, therefore the higher is this second probability, the worse is the fit. Therefore, the higher is the former probability $P(\\chi^2)$, the better is the fit; while, on the contrary, the higher is the latter probability, $P( \\Delta \\chi ^2)$, the worse is the fit. In Gonzalez-Garcia and Pena-Garay (2000) Table 3, the larger C.L. belongs to a higher $\\chi^2/d.o.f$ and so it shows a poorer fit. They found for the SMA a C.L. $83 \\%$, which tells us that it is somewhere outside the 1.5 $\\sigma$ region around the best fit which has a low (cca. 15 $\\%$) probability itself. Langacker (1999) noted that the LMA solution and the no oscillation hypothesis has the same rate of goodness, i.e. both of them is disfavored at the 95-99 $ \\%$ CL level, while the LMA solution is also a very poor fit, although it is allowed at 95$\\%$ CL. In other terms, this means that SMA is refuted with more than 1.5 $\\sigma$. Recently Maris and Petcov (2000) found that \"the conservative SMA around the point $ \\Delta m^2 = 4 \\times 10^{-6}$, $sin^2 2 \\Theta = 0.0085$ is ruled out at 1.5 sigma\" by the degree of the day-night effect observed by the SK. Now Haxton (2000) noted that \"One puzzling aspect of atmospheric, solar, and LSND neutrino results is that they require three independent $ \\delta m^2$s. That is, they do not respect the relation $ \\delta m^2_{21} + \\delta m^2_{31} + \\delta m^2_{32} = 0 $, thus either one or more of the neutrino experiments must be attributed to some phenomena other than neutrino oscillations, or a fourth neutrino is required.\" We learned that the neutrino oscillation solutions of the SNP actually are poor fits. Moreover, we recognise that even if they would be moderately good fits, it would be still an urgent need to find new physics outside from the neutrino oscillations to explain the results of the neutrino experiments. In this Letter we suggest to look for the new physics in a conservative region, in the field of astrophysics. The reason is the (not yet recognised) significance of the facts corresponding to the coupling of the solar core to the surface activity phenomena (one can find the presentation of the astrophysical core-related problems in Grandpierre, 1996, 1999, 2000). It is known, that the solar core rotates so slowly that its rotation rate is the surface rate $\\pm 30 \\%$ (Elsworth et al., 1995). At the same time, the so-called 'best solar model' of Pinsonneault et al. (1989) predicted a rotation rate of the core which is 4-15 times the surface rate. The discrepancy may be surmounted only when allowing a coupling between the solar core and the surface active regions. But if the core may participate in the surface activity cycle, the energy production has to be touched, and so the neutrino production may be influenced as well. In this way we found a way how the new physics arises in the field of astrophysics. The objections raised against a possible astrophysical solution to the solar neutrino problems (see e.g. Bludman, Hata and Langacker 1994) are valid only for non-standard models without new physics (without changing the spectral shape of the individual neutrino fluxes). But it is known that the solution of the SNP needs new physics (Bludman, Hata and Langacker, 1994). The often used term \"model-independence\" refers to general models in which the individual pp, $Be^7$, CNO and $B^8$ neutrino fluxes are allowed to vary as free parameters. But if new physics is present, another kind of nuclear reactions may also contribute to the energy and neutrino production of the Sun. In this paper we attempt to show how the presence of a high-temperature energy source (Grandpierre, 1996, 2000) could contribute to the neutrino detector data. It is a general view that the principal neutrino sources are pp, pep, $^7Be$, $^8B$, $^{13}N$, and $^{15}O$. As Bahcall and Krastev (1996) remarked, this picture has become so accepted that it is sometimes referred to as \"model-independent\". Nevertheless, if other neutrino sources do exist, then these previously thought \"model-independent\" models may all fail at a certain rate. Therefore, the related luminosity constraint also may prove insufficient and over-constrained when compared to the actual Sun, since the hot bubbles may produce a smaller or larger part of the solar luminosity. The luminosity constraint expresses the fact that the energy productions is related to the neutrino production since both is produced by the nuclear reactions: \\begin{eqnarray} L_{\\odot}/4\\pi R^2 = \\sum_{\\alpha} (Q/2 - _{\\alpha} ) \\phi(\\alpha), \\end{eqnarray} where R=1 A. U. (1.469$\\times 10^{13}$ cm), $_{\\alpha}$ is the average neutrino energy and $\\phi(\\alpha)$ is the $\\alpha$th flux ($\\alpha$=pp, pep, $^7Be,^8B,$...), and $Q$ is the energy released in the fusion reaction $4p+2e \\rightarrow \\alpha + 2\\nu$. The most general luminosity constraint we found in the literature is presented in Minakata and Nunokawa, 1999 in the following form: \\begin{eqnarray} L_{\\odot}/4\\pi R^2 & = 13.1 \\phi(pp) + 11.92 \\phi(pep) + 12.5 \\phi(^7{Be})+ \\nonumber\\\\ & 6.66 \\phi(^8B) + 3.46 \\phi(^{13}N)+21.57 \\phi(^{15}O)\\nonumber\\\\ & + 2.36 \\phi(^{17}F) + 10.17 \\phi(hep). \\end{eqnarray} We found it more suitable to convert these neutrino fluxes to fractional fluxes $\\Phi$, normalised to the BP98 SSM fluxes: \\begin{eqnarray} 1 & = 0.9119 \\Phi(pp) + 0.001966 \\Phi (pep) + 0.0758 \\Phi(Be) \\nonumber\\\\ & + 5.193 \\times 10^{-5} \\Phi(B)+ 3 \\times 10^{-3} \\Phi(CNO). \\end{eqnarray} Neglecting the small terms, and including the pep into the pp, the CNO into the Be term, \\begin{eqnarray} 0.99= 0.914 \\Phi_1 + 0.076 \\Phi_7, \\end{eqnarray} where $ \\Phi_1 $ refers to the pp+pep fluxes, and $\\Phi_7$ to the Be+CNO fluxes (where we modified the left side with the condition that for $\\Phi_1$=$ \\Phi_7$=1 the equation must be valid). We used the fractional neutrino fluxes following Minakata and Nunokawa (1998). The chlorine equation is: \\begin{eqnarray} 2.56 = 1.8 \\Phi_7 + 5.9 \\Phi_8. \\end{eqnarray} The gallium-equation is: \\begin{eqnarray} 72.4 = 69.6 \\Phi_1 + 46.9 \\Phi_7 + 12.4 \\Phi_8. \\end{eqnarray} \\begin{figure}[t] \\includegraphics[angle=0,width=7.5cm]{1.ps} \\caption[ ]{The $^8B$ and $^7Be$+CNO neutrino fluxes, consistent with the luminosity constraint and experimental results for standard neutrinos. The SSM flux region is around the (1,1) point. The dashed (solid) lines correspond to the central ($\\pm 1 \\sigma$) experimental values for Cl, Ga and $\\nu -e$ scattering experiments. The crosses indicate the behaviour of non standard solar models with low central temperature. } \\label{F1} \\end{figure} \\begin{figure} \\includegraphics[angle=0,width=7.5cm]{3.ps} \\caption[ ]{The same as in Fig. 1, except that the luminosity constraint is now written for the quiet solar core of the dynamical solar model. The hot bubbles are allowed to produce mu and tau neutrinos, therefore the Kam line shifts leftwards. The triangle LQ'99 refers to the calculations of Lavagno and Quarati, 1999. The figure shows that LQ'99 does not fit exactly to the low-temperature parabola.} \\label{F2} \\end{figure} It is easy to derive from these equations the lines of the Fig. 1 of Fiorentini and Ricci (1998), shown here as Fig. 1. It is confirmed that the non standard cool Sun models do not give an acceptable fit, as they are farther than $3 \\sigma$ from the intersection of the acceptance zones. Now we are prepared to consider the case when hot bubbles are present. If the energy production of the hot bubbles is not negligible when compared to the total solar luminosity, we should involve it into the luminosity constraint. Regarding the energy budget of the hot bubbles, it depends on their temperature, which is indicated to be in the range of $10^8$ to $10^{11}$ K (Grandpierre, 1996, 2000). One can think that the bubbles may produce energy through the hot CNO cycle, triple alpha cycle (this cycle does not produce neutrinos) and other nova-type nuclear reactions (Audouze, Truran, Zimmerman, 1973). From these reactions the one which may produce the largest number of events in the neutrino detectors may be the hot CNO cycle. Writing the luminosity constraint for the hot bubble separately, for the case when the CNO cycle gives ten percent of the total solar luminosity as an upper limit, from (1) we obtain: \\begin{eqnarray} 0.1 L_{\\odot}= 12.525 \\Phi^b(CNO). \\end{eqnarray} From this constraint, we can derive an upper limit for the bubble CNO neutrino flux, \\begin{eqnarray} \\Phi^b(CNO) < 6 \\times 10^9 cm^{-2} s^{-1}, \\end{eqnarray} which is close to the value of the SSM beryllium-neutrino flux. This value is compatible with the present-day global constraints $0.0 < \\Phi(^7 Be) < 6.35$ (Bahcall and Krastev, 1996). Now we can calculate how Fig. 1 is modified when the hot bubbles are producing 22$ \\%$ of the total solar luminosity. In this case the luminosity constraint for the quiet solar core without the bubbles should be formulated as: \\begin{eqnarray} 0.77= 0.914 \\Phi_1 + 0.076 \\Phi_7. \\end{eqnarray} Using this constraint in the Ga-equation (6), we observe that the effect of the hot bubble energy generation, or, more precisely, the constraint that the quiet solar core should produce a less than total solar luminosity, is to shift the Ga-zone upwards. This effect is helpful in obtaining better and physical (i.e. $\\Phi(Be) > 0$) fluxes, and so to resolve one of the solar neutrino problems, the problem of beryllium flux. Now one may consider what happens with the Kamiokande zone if hot bubbles are present. Since the bubbles may reach very high temperatures, they may produce mu and tau neutrinos, which can be observed by the Kamiokande and not by the other detectors. Therefore, since Fig. 2 describes the quiet solar core only, the Kam-zone may be shifted to the left. Actually, the rate of the necessary shift is determined by the intersection of the Cl and Ga zones. We observe from Fig. 2 that the shift occurs around $\\Phi_8$=0.26, which means that the bubbles has to produce a $\\nu_{\\mu, \\tau}$ neutral current with a contribution to the SK \\begin{eqnarray} \\Phi^b_{\\mu, \\tau} = 0.21. \\end{eqnarray} We note here, that since the hot bubbles may contribute also to the high-energy excess observed at the SK, therefore the amount of $\\Phi^b_{\\mu, \\tau}$ should not be so large. The Cl-zone of Fig. 1 will not be modified when going to Fig. 2. ", "conclusions": "It is interesting that the dynamic solar model (DSM) obtained with the inclusion of the hot bubbles into the standard solar model (SSM) modifies the Ga, Cl, Kam and non-standard zones in a way the create shifting an overlap region in a physical range ($\\Phi_7 > 0$. Remarkably, the cool Sun model also overlaps with this overlap region. For a bubble-luminosity around 0.22 (of the total solar luminosity), they all fit with cca. 1.2 $\\sigma$. Fig. 2 shows that the three kinds of neutrino detectors actually do not contradict to the standard neutrino picture. The situation is that the combined neutrino results actually indicate a lower than standard central temperature of the Sun with a remarkable confidence. Fig. 2 gives a better fit (cca. 1.2 $\\sigma$) than that of the presently favoured SMA solution (higher than 1.5$\\sigma$). The best fit may be reached when both the bubbles and the oscillations are taken into account. One can observe from Fig. 2 that the new astrophysics may offer powerful perspectives to influence the solution of the SNP. It seems to be not true, that we know the basic physics of the Sun enough. Even if one can consider the presence of the hot bubbles in the solar core as yet not established, the frequent statements that astrophysical solutions are ruled out, prove to be unfounded in the here presented more general basis. In Fig. 2 the non standard low-temperature solutions are also presented. Although in Fig. 1 they are found far from the 1 $\\sigma$ zones, in the case of Fig. 2 they are around 1.2 $\\sigma$ from the overlap region when $T_c = 0.942$. This value of central temperature is quite consistent with the 0.78 solar luminosity produced by the quiet solar core. We do not attempt to suggest that the hot bubbles permanently produce a significant part of the solar luminosity. Our attempt is more narrowly confined: it is to point out that the bubbles in principle may create a situation in which the astrophysical solution may be {\\it alone} enough to solve the problem of the missing solar neutrinos. It is quite plausible, that the electron neutrinos do oscillate, and this phenomenon is responsible for a large part of the SNP and other apparent \"neutrino anomalies\". Nevertheless, the point is that if we ignore the role played by the dynamics of the solar core in the production of solar neutrinos, we may found ourselves in the uncomfortable situation of poor fits of the MS parameters and inconsistent consequences of atmospheric, LSND and solar neutrino problems. To resolve the astrophysical problems of the solar core it is not necessary that the hot bubbles produce a significant part of the solar luminosity. Actually, they may produce only a negligible part of the solar luminosity and they could prove still be able to trigger and influence the surface solar activity. Gorbatzky (1964) already calculated that hot bubbles arising from point explosions with an initial surplus energy around $10^{35}$ ergs may be able to reach the stellar surfaces from 0.1 stellar radius. He ignored completely the energy production of these hot bubbles with temperatures $T > 10^8 $ K. But even a slight influence of the solar core dynamics might be able to improve significantly the fit and consistency of the neutrino problems. The result red from Fig. 2 has predictive value for the future neutrino detectors. It is easy to estimate the consequences of this dynamic solar model (for standard neutrinos) regarding the SNO observations. {\\it Our picture modifies the conclusion that the [NC]/[CC] is larger than unity} ([i]=observed rate/standard solar model rate, see Bahcall, Krastev, Smirnov, 2000) {\\it will definitely indicate the presence of neutrino observations}. Since at high temperatures like $ 10^{10} - 10^{11} K$ the hot bubbles may produce mu and tau neutrinos independently of the presence of neutrino oscillations, they may contribute to [NC] increasing it. For example, if we take $T=0.942$, than we will have $R^{qc}_{SK}=0.23$, and so [NC]/[CC]=2.04. If we allow a value of T closer to the SSM value 1, the ratio $([NC]/[CC])_{DSM}$ decreases towards unity. This conjecture of the DSM should be taken into account at the interpretations of the SNO and SK observations. Regarding the helioseismic context, we note that sound speed is not sensitive to the nuclear reactions. Bahcall and Ulrich (1988) and Basu, Pinsonneault and Bahcall (2000) remarked that even when switching out the $He^3$ + $He^4$ reaction (producing 14 $\\%$ of the total solar luminosity), the sound speed differ only by 0.1 $\\%$ from the sound speed obtained from the standard solar model. On the other hand, the calculations of the dynamic solar model show that when bubbles and neutrino oscillations are both present the SMA and LMA regions (of the purely MSW solutions) shift significantly even for such a small change as $T=0.995$. Moreover, the many-body effects of the particles in the nuclear reactions led to solar models which may compensate the cool solar models towards an SSM central temperature (Lavagno, Quarati 1999)." }, "0005/astro-ph0005367_arXiv.txt": { "abstract": "We have obtained echelle spectra with the Hubble Space Telescope (HST) of the MgII 2800\\,\\AA\\ region of ten stars whose metallicities range from 1/300 to 1/3 that of the Sun, and whose space velocities suggest a halo or old thick-disk origin. Spectra of all ten show double-peaked emission in the MgII core, very much like the quiet Sun. A half-dozen apparently non-rotating stars were observed more than once, and show at most a low level of variability in the emission flux, comparable to that of quiet stars of solar metallicity. For four stars, we have obtained Lyman\\,$\\alpha$ spectra at 0.2\\,\\AA\\ resolution; all four show emission. The data thus strongly suggest that chromospheric activity at a minimum level is present in all stars of near-solar temperature, regardless of age or metallicity. While this points to non-magnetic sources such as acoustic waves, a contribution from globally-organized magnetic fields is possible. A longer series of MgII and Lyman\\,$\\alpha$ observations is needed to constrain this. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005198_arXiv.txt": { "abstract": " ", "introduction": "A normal galaxy is one that derives its luminosity primarily from nuclear burning in stars, and is at neither the high nor the low extreme of the luminosity distribution. In such galaxies, the visible and ultraviolet is the only direct window onto the photospheres of stars, whereas the rest of the spectrum reflects reprocessed light from stars (Figure 1). Dust in the interstellar medium reprocesses part of the stellar luminosity into infrared emission from a few \\um to 1~mm or longer wavelengths. The radio emission at millimeter wavelengths derives from thermal emission from ionized plasmas in HII regions, whereas the synchrotron emission from cosmic ray electrons ($\\rm CR\\ e^-$) trapped in the magnetic field of the galaxy fill in the cm-wavelength spectrum. In the X-rays, the luminosity is dominated by very hot plasmas at $\\rm T\\wig>10^6~K$ created by shocks from supernova explosions. The global spectrum of normal galaxies has some very stable signatures, and other quite variable aspects. Much of this article is an attempt to understand what makes for stability in the spectrum, and what significance to attach to the variable parameters. Since stars form in the interstellar medium (ISM), it is no surprise that galaxies forming stars actively have substantial ISM luminosities, conveyed overwhelmingly in the infrared, and often exceeding by large factors the optical and ultraviolet luminosity. Studying galaxies in the infrared is thus tantamount to studying the ISM and its properties, and star formation (SF) activity on large scales. Astration, the cycle of star formation, nuclear burning and ISM enrichment in heavy elements, is the primary secular change and main evolutionary process affecting the chemical make-up and energy balance of the Universe. Because the luminosity function ({\\it e.g.} Kim \\& Sanders 1998) falls fast enough that the integral over the population is dominated by the low end of the luminosity interval, over 90\\% of star formation in the local Universe takes place in normal disk galaxies, rather than the spectacular extremes. Thus, normal galaxies essentially cause and host Cosmic nuclear evolution. Normal galaxies form a unique bridge between our detailed understanding of the Milky Way and our global understanding of the Universe and its history. Empirically, the integrated emission from galaxies must be compatible with the properties of Galactic objects. Our physical models of the local energy balance and of large-scale processes in the ISM must be consistent with the observed global behavior. On the other hand, galaxy luminosity and spectral properties are a crucial ingredient to modeling the observed faint extragalactic source counts and infrared background. Moreover, modeling the counts and background requires more than a census of the passive population ({\\it cf.} Puget's article in this volume), and our models of galaxy behavior must be consistent with population evolution required to explain deep counts. The enrichment in individual galaxies must also add up to the observed current-day abundances. Thus achieving an accurate representation of normal galaxies would verify directly the validity of underlying models of the Milky Way ISM, and is a necessary step to the interpretation of cosmological results. The InfraRed Astronomy Satellite (IRAS; Beichman \\etal 1986) ushered in the era of space-based infrared astronomy in a dramatic fashion, revealing a stunningly rich infrared sky, unanticipated from the bits of infrared data previously and heroically collected from the ground. IRAS conducted the first unbiased all-sky survey at 12, 25, 60 and 100\\um, as well as a spectroscopic survey with a resolving power of 20 between 7.5 and 23\\um, for objects brighter than 10 Jy or so. IRAS had a profound influence on astronomy in general, not just the infrared, because it represented such a very large gain in sensitivity and spatial coverage, comparable perhaps to going from attempting visual astronomy in daylight to observing in a dark night (Beichman 1987; Soifer, Houcj \\& Neugebauer 1987). Another significant factor in this influence was the dissemination of data products from IRAS, including source catalogs, image atlases and sky brightness estimates, generated with great care and well characterized and documented as to reliability, completeness, flux accuracy, and other statistical attributes. The Infrared Space Observatory (ISO, Kessler \\etal 1996) took infrared astronomy in space to new levels of sensitivity and into a new realm of spectroscopy across the whole range of 3 to 200\\mic. Its versatile instrumentation took to the rich and fascinating extragalactic infrared sky first revealed by IRAS, and moved our knowledge of that sky an order of magnitude further and deeper. In what follows, methods traditionally used in the study of normal galaxies are described, and outstanding questions currently pursued in the field are stated. The most prominent results from the IRAS survey are reviewed in \\S3. Contributions by ISO in the field of broad-band photometry are then presented (\\S4), followed by results in spectrospcopy (\\S5). Normal galaxy studies not directly concerned with the ISM are reviewed in \\S6. The outlook and challenges in pursuing the interpretation of infrared data on the ISM are discussed in \\S7. ", "conclusions": "" }, "0005/astro-ph0005244_arXiv.txt": { "abstract": "Current estimates of number counts of radio sources in the frequency range where the most sensitive Cosmic Microwave Background (CMB) experiments are carried out significantly under-represent sources with strongly inverted spectra. Hot accretion flows around supermassive black holes in the nuclei of nearby galaxies are expected to produce inverted radio spectra by thermal synchrotron emission. We calculate the temperature fluctuations and power spectra of these sources in the Planck Surveyor 30 GHz energy channel, where their emission is expected to peak. We find that their potential contribution is generally comparable to the instrumental noise, and approaches the CMB anisotropy level at small angular scales. Forthcoming CMB missions, which will provide a large statistical sample of inverted-spectra sources, will be crucial for determining the distribution of hot accretion flows in nearby quiescent galactic nuclei. Detection of these sources in different frequency channels will help constrain their spectral characteristics, hence their physical properties. ", "introduction": "The upcoming cosmic microwave background (CMB) experiments, e.g.~MAP and the Planck Surveyor, will be able determine the primordial anisotropies to an unprecedented level of accuracy. Because of its high sensitivity, excellent angular resolution and wide range of frequencies, Planck in particular, will be extremely sensitive to extragalactic foreground point sources, which provide the major source of uncertainty in the measurement of the intrinsic fluctuations. Several studies have therefore been carried out to calculate the contribution of point sources to the CMB anisotropies. Much of this work (see Toffolatti et al. 1999a,b; De Zotti et al. 1999; Gawiser \\& Smoot 1997; Sokasian, Gawiser \\& Smoot 1998) has dealt with the contribution from radio sources, the number counts of which are determined down to $\\mu$Jy but only up to frequencies $\\la$ 8 GHz. These counts are usually extrapolated to the higher frequencies relevant for the CMB experiments. This implies that the available counts are sensitive enough to include the most significant contribution from the ``steep'' and ``flat'' spectrum sources (with $F_\\nu\\propto \\nu^{-\\alpha}$, and $\\alpha\\ge 0$, such as compact radio galaxies and radio loud quasars), but are missing, or are strongly under-representing, an important contribution from a class of sources with inverted spectra ($\\alpha\\la 0$; e.g. De Zotti et al. 1999). This is further emphasized by recent observations at 28.5 GHz, which find up to a factor of 7 more sources than predicted from low-frequency surveys (Cooray et al.~1998). Inverted-spectrum sources, such as those discussed here, may peak in the frequency range of a few tens to a few hundreds GHz, and could therefore provide a considerable contribution in the region where the most sensitive CMB experiments are carried out. GHz Peaked Spectrum (GPS) sources (see O'Dea et al.~1998, Guerra, Haarsma \\& Partridge 1998) have been recognized to be an important class of inverted-spectrum sources. Their emission is attributed to synchrotron radiation from compact and high density regions often associated with the early stages of the formation of more classical double radio sources (the so called ``young source'' scenario; Philips \\& Mutel 1982). However, as pointed out by Toffolatti et al. (1999), there may be another, distinct, class of strongly inverted spectra due to thermal synchrotron emission in hot or advection dominated accretion flows (ADAFs). Unlike the relatively rare and bright GPS sources (peak fluxes of $\\sim 1-10$ Jy), usually associated with bright active galaxies or quasars at high redshifts, ADAF sources should be common in nearby galaxies and provide the most significant contribution to the emission in the high radio frequencies of the faint ($\\sim$ a few mJy) radio cores observed in such galaxies. The reason why we consider hot accretion flows to be common in nearby galactic nuclei is that, in recent years, it has become apparent (e.g. Fabian \\& Rees 1995; Narayan \\& Yi 1995; Di Matteo et al. 2000 and references therein) that the nuclei of such galaxies, which host the largest black holes known with masses of $10^8-10^{10} M_{\\odot}$ (e.g., Magorrian et al. 1998), are remarkably underluminous for the typically expected accretion rates (determined from measuraments of densities and sound speeds of their hot interstellar medium). In particular, it has been shown (e.g. Di Matteo et al. 2000) that the relative quiescence and spectral characteristics of the early-type galactic nuclei can be well-explained if the central black holes accrete via low radiative-efficiency accretion flows or ADAFs (Rees et al. 1982; for a review see, e.g., Narayan, Mahadevan \\& Quataert 1998). Moreover, it has been proposed (Di Matteo and Allen 1999) that such flows, which also produce significant emission in the $X$-ray band, could provide a significant contribution to the cosmic $X$-ray background (XRB). Within the context of these models, a significant fraction of the hard number counts in the X-ray energies should arise from sources at low redshift ($z \\la 1$). This picture is supported by recent deep {\\em Chandra} observations, which have resolved about 40 per cent of the hard XRB in point sources in bright early-type galaxies (Mushotzky et al. 2000). The potential contribution of GPS sources to fluctuations in the CMB anisotropy has been discussed by De Zotti et al. (1999). In this {\\em Paper}, we examine the specific contribution of inverted spectra ADAF sources in the nuclei of early-type galaxies to the CMB anisotropy. We evaluate their foreground contribution to the small-scale cosmic microwave fluctuations in the low-energy channels foreseen for the Plank surveyor mission. These sources, if indeed common in elliptical galaxies, should be much more numerous albeit fainter than the GPS population, and may therefore provide a stronger noise contribution at the small angular scales. While it is important to assess the potential contribution of advection-dominated sources to the CMB fluctuations, the forthcoming CMB experiments themselves will, for the first time, provide a large statistical sample of objects with inverted radio spectra. Because most of the ADAF emission occurs in the high radio and in the $X$-ray band, Planck observations will possibly provide the most powerful test for the presence of ADAFs around supermassive black holes. In particular, such studies will provide strong constraints on the spectral properties of this class of objects, and will help determine how common they are in the nearby Universe. Confirming the presence of these sources would also support the conjecture that they provide a significant contribution to the hard XRB. ", "conclusions": "We have computed the temperature fluctuations and power spectrum produced by inverted radio spectra from hot accretion flows in the nuclei of nearby elliptical galaxies in the Planck 30 GHz channel, where their emission is expected to peak. We have shown that the contribution from this class of sources approaches the intrinsic CMB fluctuation level only at small angular scales. However, because of the different nature of its power spectrum, the source contribution should not affect the most important goal of the Planck mission, that is the accurate measurement of the primary CMB anisotropy. On the other hand, Planck will provide a large statistical sample of sources characterized by inverted spectra. Therefore, it should be possible to use this study to determine how common this mode of accretion is in the nearby supermassive black holes. In particular, as most of the contribution from this population is expected to peak at high radio frequencies, Planck should allow us to study their spectral characteristics. In turn, because different spectral distributions and luminosities reflect the shape of the density profiles, CMB experiments could allow us to gain important information on the physical conditions in these accretion flows. As already noted by Toffolatti et al. (1999b), the implications of such a study could, more generally, be significant as a way of testing the physical processes in the medium surrounding massive black holes, and the evolution of the interstellar medium in galaxies up to moderate redshifts. Even more, it would provide a test for current ideas according to which a significant fraction of the $X$-ray background may due to accretion in this regime in early-type galaxies in the local universe (Di Matteo \\& Allen 1999). Note that such a significant statistical study would be more difficult to carry out with surveys at other wavelengths because of the rapid decline of the ADAF flux, which makes the emission from this type of accretion flows extremely weak in the far infrared and optical bands. We need to stress that, in principle, the contribution from ADAF sources should be easily disentangled not only from that due to sources with a flat and steep spectrum, but also from that due to GPS sources which also have strongly inverted spectra. GPS sources are typically much brighter (with fluxes typically ranging from a few to 10 Jy) but rarer (usually associated with QSOs) than the expected ADAF sources. The number of GPS sources rapidly decreases with decreasing flux, whereas ADAFs are expected to be much more numerous at faint flux levels. As a result, GPS are only minor contributors to the fluctuations at small angular scales, whereas ADAFs would be mostly significant at these scales. Therefore it should be possible to study these two populations independently. Note that we have shown the expected temperature fluctuations due to ADAF sources only for the lowest energy channel of Planck. If most of the sources are indeed in the range of luminosities consistent with those observed so far, then this channel is expected to have the largest (possibly major) contribution, due to the high-frequency cutoff in the spectrum of these sources. However, if a substantial population of high-luminosity sources is present, then some contribution should also be present in the other channels of the Planck LFI. The availability of multifrequency data should allow an efficient identification of pixels contaminated by discrete sources. In order to carry out a substraction of the contaminating flux one should therefore take into account that strongly inverted spectra such as those considered here may not be present in most frequency channels but give rise to a strong contamination up to a certain frequency, and then abruptly drop. It should also be pointed out that, contrary to some of the GPS sources for which variability has been observed (e.g. Stanghellini et al. 1998), the radio sources in the hot accretion flows are usually not very variable. A lack of variability is particularly important for a proper removal of sources from the spectral fitting. We note that ADAFs around massive black holes could also be found in spiral galaxies such as the Galactic nucleus Sgr A$^*$. However, even if ADAFs were indeed common in spiral nuclei, their potential contribution to the CMB anisotropy would still be dominated by that from ellipticals. Inferred black hole masses are found to be proportional to the mass of the bulge component of their host galaxy, implying $M_{BH} \\sim 10^6-10^7 M_\\odot$ for spiral galaxies. As a consequence, the contribution from spirals should be much lower, as the radio flux scales as $M_{\\rm BH}^{2.5-3}$ (Franceschini et al. 1998), and (see \\S3) their spectrum would peak at frequencies higher than those of elliptical cores and affect higher energy (e.g. sub-mm, mm) channels of CMB experiments. Because of this, given enough sensitivity, the relevance of ADAFs in quiescent spiral nuclei may also be assesed separately by the forthcoming experiments. Finally, we note that in our analysis we do not take into account the effects of source clustering. Clustering decreases the effective number of objects in randomly placed cells and, consequently, enhances the cell to cell fluctuations. There is indeed evidence that the positions in the sky of a wide variety of extragalactic sources are correlated (Shaver 1988). However, the specific correlation function of our radio--submm sources in early-type galaxies is not well-constrained. The analyses of Toffolatti et al. (1998) have shown that the contribution due to clustering (using the two-point correlation function from sources selected at 5 GHz; Loan, Wall \\& Lahav 1997) is generally small in comparison with the Poisson term; however, the relative importance of clustering increases if sources are substracted out from the Planck maps down to faint flux levels." }, "0005/astro-ph0005558_arXiv.txt": { "abstract": "We estimate the {\\it mass} density fluctuations power spectrum (PS) on large scales by applying a maximum likelihood technique to the peculiar velocity data of the recently completed redshift-distance survey of early-type galaxies (hereafter ENEAR). Parametric CDM-like models for the PS are assumed, and the best fit parameters are determined by maximizing the probability of the model given the measured peculiar velocities of the galaxies, their distances and estimated errors. The method has been applied to CDM models with and without COBE normalization. The general results are in agreement with the high amplitude power spectra found from similar analysis of other independent all-sky catalogs of peculiar velocity data such as MARK III (Willick \\etal 1997)and SFI (Giovanelli \\etal 1998; da Costa \\etal 1996), in spite of the differences in the way these samples were selected, the fact that they probe different regions of space and galaxy distances are computed using different distance relations. For example, at $k=0.1 \\ihmpc$ the power spectrum value is $P(k) \\Omega^{1.2} =(6.5 \\pm 3) \\times 10^3 (\\hmpc)^3$ and $\\eta_8\\equiv\\sigma_8 \\Omega^{0.6} = 1.1_{-0.35}^{+0.2}$; the quoted uncertainties refer to $3\\sigma$ error level. We also find that, for \\lcdm\\ and \\ocdm\\ COBE-normalized models, the best-fit parameters are confined by a contour approximately defined by $\\Omega h^{1.3}=0.377\\pm0.08$ and $\\Omega h^{0.88}=0.517\\pm0.083$ respectively. $\\Gamma$-shape models, free of COBE normalization, results in the weak constraint of $\\Gamma \\geq 0.17$ and in the rather stringent constraint of $\\eta_8=1.0\\pm 0.25$. All quoted uncertainties refer to $3\\sigma$ confidence-level (\\cl). The calculated PS has been used as a prior for Wiener reconstruction of the density field at different resolutions and the three-dimensional velocity field within a volume of radius $\\approx 80 \\hmpc$. All major structures in the nearby universe are recovered and are well matched to those predicted from all-sky redshift surveys. The robustness of these features has been tested with Constrained Realizations (CR). Analysis of reconstructed three-dimensional velocity field yields a small bulk flow amplitude ($\\sim 160\\pm 60~\\kms$ at $60~\\hmpc$) and a very small rms value of the tidal field ($\\sim 60~\\kms$). The results give further support to the picture that most of the motion of the Local Group arises from mass fluctuations within the volume considered. ", "introduction": "\\label{sec:intro} The canonical model of cosmology assumes that large-scale structure has grown out of small density perturbations via the process of gravitational instability. These initial fluctuations are usually assumed to satisfy the statistics of a Gaussian random field, solely characterized by its power spectrum. In the linear regime, the fluctuations grow self-similarly and retain their initial distribution and power spectrum shape. Therefore, mapping the underlying cosmological velocity field and its power spectrum on large scales, provides a direct probe to the origin of structure in the universe. The PS, the three-dimensional distribution of luminous matter and the predicted peculiar velocity field have been derived from a variety of data sets, especially from all-sky redshift surveys (\\eg\\ Strauss \\& Willick 1995 for a review of earlier work; Sutherland \\etal 1999; Branchini \\etal 1999). Unfortunately, however, the distribution of galaxies in these catalogs is not necessarily an unbiased tracer of the underlying mass distribution, and suffer from the infamous ``galaxy biasing'' problem. Furthermore, in estimates from redshift surveys, uncertainties arise from the complicated relation between the real space and the redshift space distributions, known as redshift distortions (\\eg , Kaiser 1987, Zaroubi and Hoffman 1996). In order to avoid these problems altogether it is advantageous to appeal to dynamical data, in particular catalogs of galaxy peculiar velocities on large scales. Peculiar velocities enable a direct and reliable determination of the mass PS and distribution, under the natural assumption that the galaxies are unbiased tracers of the large-scale, gravitationally-induced, velocity field. Furthermore, since peculiar velocities are non-local and have contributions from different scales, analysis of the peculiar velocity field provides information on scales somewhat larger than the sampled region (\\eg\\, Hoffman \\etal 2000). For the same reason peculiar velocities are adequately described by linear theory even when densities become quasi-linear (\\eg , Freudling \\etal 1999). Consequently, the dynamics and the distribution of peculiar velocities are well described by the linear regime of gravitational instability and by a Gaussian probability distribution function (PDF), respectively. Assuming that both the underlying velocity field and the errors are drawn from independent random Gaussian fields, the observed peculiar velocities constitute a multi-variant Gaussian data set, albeit the sparse and inhomogeneous sampling. The corresponding {\\it posterior} PDF is a multivariate Gaussian that is completely determined by the assumed PS and covariance matrix of errors. Under these conditions one can write the joint PDF of the model PS and the underlying velocity or density field. The purpose of the present study is to calculate, from the joint PDF, the PS and 3D mass distribution, as well as the 3D peculiar velocity field, as derived from the newly completed ENEAR galaxy peculiar velocity catalog (da Costa \\etal 2000a, Paper~I). First, the PS model parameters are estimated by maximizing the likelihood function given the model (Zaroubi \\etal 1997). An identical likelihood estimation of the power spectrum has been previously applied to the {\\rm Mark III} (Zaroubi \\etal 1997) and the SFI (Freudling \\etal 1999) data sets. In both cases the analysis yielded a high amplitude power spectrum. Although the results from those two catalogs are consistent with each other, they are marginally inconsistent with the power spectra measured from redshift catalogs (\\eg , da Costa \\etal 1996; Sutherland \\etal 1999), inferred from the analysis of the velocity correlation function (\\eg\\, Borgani \\etal 2000a, 2000b), and from velocity-velocity comparisons (\\eg , Davis \\etal 1996, da Costa \\etal 1998). One of our goals is to use the same methodology employed before for the Mark~III and SFI to the new ENEAR catalog to directly test the reproducibility of the results with an independent sample based on a different distance indicator but probing a comparable volume. Second, the Wiener filter (WF) solution of the field is recovered by finding the most probable field given the PS and the data (Zaroubi \\etal 1995, 1999). Constrained realizations (CR) are then used to sample the statistical scatter around the WF field (Hoffman and Ribak 1991). The mass density PS is used to calculate the smoothed Wiener filtered density and 3D velocity fields given the measured radial velocities (Zaroubi \\etal 1995, 1999). The WF provides an optimal estimator of the underlying field in the sense of a minimum-variance solution given the data and an assumed \\prior\\ model (Wiener 1949; Press \\etal\\ 1992). The \\prior\\ defines the data auto-correlation and the data-field cross-correlation matrices. In the case where the data is drawn from a random Gaussian field, the WF estimator coincides with the conditional mean field and with the most probable configuration given the data (see Zaroubi \\etal 1995). It should be noted that Kaiser \\& Stebbins (1991) were the first to propose a Bayesian solution to the problem of reconstruction from peculiar velocity data sets. Finally, the recovered three-dimensional velocity field is used to compute the amplitude of the bulk flow and to decompose the velocity field in terms of a divergent and tidal components which enables one to separate the contribution to the measured peculiar velocity field from mass fluctuations within and outside the volume probed by the data (Hoffman \\etal 2000). The methods adopted in this study do not involve any explicit window function, weighting or smoothing the data. In addition, they automatically underweight noisy, unreliable data. However, a few simplifying assumptions are required: 1) peculiar velocities are drawn from a Gaussian random field; 2) peculiar velocities are related to the densities through linear theory; 3) errors in the $D_n-\\sigma$ inferred distances constitute a Gaussian random field with two components, the first scales linearly with distance while the second models the nonlinear evolution of the velocities as a constant scatter. The need to assume a parametric functional form for the PS is also a limitation. The outline of this paper is as follows. In \\S~\\ref{sec:data} we briefly describe the peculiar velocity data used in the present analysis. The PS analysis is carried out in \\S~\\ref{sec:PS}. The Wiener filtering is applied to the ENEAR data in \\S~\\ref{sec:wiener}, where maps of the density field are presented and compared to those predicted from redshift surveys. Also shown in this section are the recovered three-dimensional velocity field and the results of its analysis. Our results are summarized and discussed in \\S~\\ref{sec:conclusion}. ", "conclusions": "\\label{sec:conclusion} In the first part of this paper the maximum-likelihood method (Zaroubi \\etal 1997) has been used to measure the mass-density power spectrum from the newly completed ENEAR early--type redshift-distance survey. The method assumes that the galaxy peculiar velocities satisfy Gaussian random statistics and that they are linearly related to the mass-density field. The initial fluctuation power spectrum is assumed to be CDM-like, with or without COBE normalizations. In addition the measured peculiar velocities error are assumed to be proportional to the distance with some thermal component to account for the nonlinear evolution of high-density environment in which the early--type galaxies reside. General results that are valid for all the models used in the analysis, and are independent of the detailed parameterization and normalization used in each model, can be summarized as follows. The amplitude of the power spectrum at $k=0.1 \\ihmpc$ is $P(k)\\Omega^{1.2}= (6.5\\pm 3) \\times 10^3 (\\ihmpc)^3$ yielding $\\eta_8=1.1^{+0.2}_{-0.35}$. For the family of COBE-normalized CDM models the following range of parameters was considered: $\\Omega \\le 1$; $0.4 < h <1$; and $n=1$. Within this range we have obtained a constraint on a combination of the parameters $\\Omega$ and $h$ which can be approximated by $\\Omega\\approx (0.38\\pm 0.08) h^{-1.3}$ for \\lcdm, and $\\Omega\\approx (0.52\\pm 0.083) h^{-0.88}$ for \\ocdm. For $h=0.65$, \\lcdm\\ yields $\\Omega= 0.5-0.8$. Similar constraints are obtained from the analysis of the generic $\\Gamma$-models, independent of the COBE normalization. We find that the power spectrum amplitude and shape parameter are constrained to be $\\eta_8=1.0^{+0.3}_{-0.28}$ and $\\Gamma \\ge 0.18$, with larger values of $\\Gamma$ ($>0.4$) being more probable. We point out that these constraints are consistent with the results obtained from a similar analysis of the {\\rm Mark III} and the SFI peculiar velocity catalogs. This agreement is encouraging since it shows that the results are robust and independent of the sample used. Examination of the $\\chi^2/d.o.f.$ for the most likely COBE-normalized models shows that their values are of the order of 0.93. These values are about $2\\sigma$ away from the preferred value of 1. However, this should not be too alarming as many of the models within the errors have $\\chi^2/d.o.f \\sim 1$. The $\\chi^2/d.o.f$ for the best-fit $\\Gamma$-model is 0.99. As pointed out by previous papers that have analyzed the PS derived from peculiar velocity data (Zaroubi \\etal 1997, Freudling \\etal 1999), the constraints on $\\eta_8$ and $\\Gamma$ are considerably higher than those obtained from other types of analyzes including peculiar velocity data (Borgani \\etal 1997, 2000), cluster abundances, and the galaxy power-spectrum (Efstathiou \\etal 1992; Sutherland \\etal 1999). They are also not consitent with those obtained by combining the results from high redshift supernovae type Ia (Perlmutter \\etal 1999) and the CMB data (Efstathiou \\etal 1999) which yields values of $\\Omega \\approx 0.25\\pm 0.15$ and $\\Lambda\\approx 0.65 \\pm 0.2$. Furthermore, assuming a linear galaxy-mass relation the value of $\\eta_8$ obtained from the present analysis would imply $\\beta=1.0$ or a $\\beta_I \\sim 1.4$ (\\eg Willmer, da Costa \\& Pellegrini 1999; Sutherland \\etal 1999), where the subscript refers to \\iras\\ galaxies, at least a factor of 2 larger than those derived from a velocity-velocity comparison of the \\iras\\ 1.2 Jy gravity field and the {\\rm Mark III} (Davis \\etal 1996), SFI ( da Costa \\etal 1998) and ENEAR (Nusser \\etal 2000) all leading to $\\beta_I \\sim 0.5$. These values are also consistent with those derived from small-scale velocities (Fisher \\etal 1995). There are many possible explanations for the above discrepancies. One possibility is that all other analyses have somehow conspired to produce consistent results but yet incorrect interpretation. Even though at first glance this seems unlikely, the recent results from the CMB ballon experiments Boomerang (Bernardis \\etal 2000; Lange \\etal 2000) and MAXIMA (Hanany \\etal 2000; Balbi \\etal 2000) have shown that the height of the second peak in the CMB angular power-spectrum is consistent with higher values of $\\Omega$. From their most likely models these authors derive $\\Omega= 0.4-0.8$ It is important to point out that the method is very sensitive to the assumed error model which can add or supress power. It also implicitly gives a high weight to nearby galaxies, likely to be slow rotators or low velocity dispersion systems, for which the measurements and the distance relations are the least reliable. However, tests have shown that these effects are not very important for the present data set. Another potential problem arises due to the rapid decrease of the weight with distance, the effective volume of the currently available catalogs is small and the shape of the power spectrum is poorly constrained, as illustrated by the case of the $\\Gamma$-model. All these factors may impact on the reliability of the constraints obtained from the PS analysis. Finally, one or more of the theoretical model ingredients could be inaccurate, \\eg, power spectrum assumed shapes, Gaussianity of the distribution; or even some inherent bias in the method itself that has eluded the extensive numerical tests carried out with the data and mock samples (\\eg Freduling \\etal 1999). In fact, through an eigenmode expansion of the {\\rm Mark III} and SFI galaxy catalogs, Hoffman and Zaroubi (2000) have conducted a mode--by--mode goodness--of--fit analysis. They found that when the surveys are analyzed with their corresponding CDM most likely models, there is a systematic inconsistency between the data and the `best-fit' models suggesting either a generic problem in the peculiar velocity data sets or the inadequacy of the theoretical or error models. Unfortunately, however, the analysis has not been able to point to the exact source of inconsistency. Finally, in this study we have also performed, given the most probable power spectrum, a Wiener reconstruction of the density and velocity fields. The maps shown here have $1200 \\kms$ and $900 \\kms$ Gaussian resolution and they are limited to the Supergalactic plane. The main features shown are similar to the features in the \\iras\\ reconstruction, corrected for perculiar velocities. The constrained realizations allow us to estimate the point-by-point uncertainties in the recovered maps. In terms of their recovered density fields ENEAR, SFI and {\\rm Mark III} mostly agree. However, they do differ in the velocity fields. ENEAR shows no significant tidal component which contributes about half of the {\\rm Mark III} local bulk velocity. This tidal field accounts for the very different bulk velocities obtained from ENEAR and {\\rm Mark III}, with SFI situated in between these surveys. The results suggest that volumes of $60-80\\hmpc$ are essentially at rest relative to the CMB and that the Local Group motion is primarily due to mass fluctuations within the volume sampled by the existing catalogs of peculiar velocity data." }, "0005/astro-ph0005302_arXiv.txt": { "abstract": "We study the statistics of gamma-ray bursts, assuming that gamma-ray bursts are cosmological and they are beamed in the form of a conical jet with a large bulk Lorentz factor $\\sim 100$. In such a conic beam, the relativistic ejecta may have a spatial variation in the bulk Lorentz factor and the density distribution of gamma-ray emitting jet material. An apparent luminosity function arises because the axis of the cone is randomly oriented with respect to the observer's line of sight. The width and the shape of the luminosity function are determined by the ratio of the beam opening angle of the conical jet to the inverse of the bulk Lorentz factor, when the bulk Lorentz factor and the jet material density is uniform on the photon emitting jet surface. We calculate effects of spatial variation of the Lorentz factor and the spatial density fluctuations within the cone on the luminosity function and the statistics of gamma-ray bursts. In particular, we focus on the redshift distribution of the observed gamma-ray bursts. The maximum distance to and the average redshift of the gamma-ray bursts are strongly affected by the beaming-induced luminosity function. The bursts with the angle-dependent Lorentz factor which peaks at the center of the cone have substantially higher average gamma-ray burst redshifts. When both the jet material density and the Lorentz factor are inhomogeneous in the conical beam, the average redshift of the bursts could be 5 times higher than that of the case in which relativistic jet is completely homogeneous and structureless. Even the simplest models for the gamma-ray burst jets and their apparent luminosity distributions have a significant effect on the redshift distribution of the gamma-ray bursts. ", "introduction": "The BATSE experiment on the {\\it Compton Gamma Ray Observatory} and the study of the afterglows (e.g., Piran 1999 and references therein). have established that the gamma-ray bursts (GRBs) are cosmological \\citep{mao92, meegan92,pi92}. Even though the distance scale seems settled \\citep{met97}, it appears that uncertainties remain in the total energy and the burst rate of GRBs \\citep{ku99, kupi99}. These two important issues depend on the level of beaming in GRB emission. That is, the issues critically depend on whether the geometry of the gamma-ray emitting ejecta is spherical or jet-like \\citep{har99, kul99, mes99, sari99}. A number of authors have studied energetics and geometry of the ejecta \\citep{maoyi94, rho97, pan98, rho99, mod00}. It is also important to put constraints on the width of the luminosity function by comparing the observed intensity distribution with those predicted by a physical model \\citep{maoyi94, yi94}. In essence, the rate, the energy, and the luminosity function of GRBs are all closely related to whether or not the geometry of the ejecta is spherical. Two most frequently quoted statistics in GRB observations are $$ and $\\log N (>F) - \\log F$, where $F$ refers to the peak flux (or peak count rate) and $N$ denotes the number of GRBs with fluxes higher than $F$ (e.g., Yi 1994). These two quantities contain information on the lumonosity function of GRBs and the spatial number density of the sources. A value of $$ consistent with that of an observed sample is a necessary condition but not a sufficient condition for a luminosity function $\\Phi(L)$ which is neither directly observable nor theoretically well undertood. The luminosity function of GRBs can be obtained for an assumed source distribution $n(z)$ such that the calculated $\\log N(>F) - \\log F$ fits the observed distribution, and vice versa. The density $n(z)$ refers to the rate of GRBs per unit time per unit comoving cosmological volume. However, due to the very nature of $N(>F)$, which is the convolution of $n(z)$ and $\\Phi(L)$, one almost always obtains $n(z)$ for a given $\\Phi(L)$ such that the theoretical $\\log N(>F) - \\log F$ curve fits the observed intensity distribution. Therefore, in order to extract information concerning $n(z)$ or $\\Phi(L)$, one has to assume one of these two functions or to develop a techinique to separate the effects of these two unknown functions (Horack and Emslie 1994; Horack et al. 1994; Ulmer et al. 1995; Ulmer and Wijers 1995). It is therefore of great interest to construct $\\Phi(L)$ on the basis of a physical model, which is one of our major goals in this {\\it Letter}. Since there remain uncertainties in GRB engine models, we focus on the consequences of the conical beaming without specifying how a beam is formed in a physical engine model. Using the first BATSE catalog of gamma-ray bursts \\citep{fishman94}, \\citet{maoyi94} studied the effects of the relativistic bulk motion in a conical beam on the statistics of gamma-ray bursts. They found that the luminosity function is naturally introduced by the random distribution of the space orientation of the cone axis and that the case of the standard candle is not easily distinguished from that of the beaming-induced luminosity function with a sharp peak. This is especially the case for large beam opening angle and the large Lorentz factor $\\gamma$, as one may expect. Different Lorentz factors and opening angles however result in non-trivial changes for the distances to GRBs and especially the highest redshift of or the maximum distance to the most distant GRB for a given sample. For instance, the maximum redshift $z_{max}$ increases as the ratio of the opening angle to $\\gamma^{-1}$ decreases. We modify the conical beam model by allowing a spatial variation of $\\gamma$ and the density profile of gamma-ray emitting electrons on the photon-emitting surface of the cone. From numerical simulations of relativistic jets \\citep{marti97, renaud98, rosen99} and observations of the astrophysical jets \\citep{zen97, spruit00}, it is clear that jets do have some significant structure in them and the bulk Lorentz factors evolve as the jets propagate. Therefore, it is plausible to extend the simplest jet model such as that of \\citet{maoyi94}. In a more realistic jet model, the bulk Lorentz factor has a spatial profile at the surface where the observed gamma-ray emission occurs and the spatial electron density distribution is significantly inhomogeneous. In $\\S$ 2 we begin with a brief presention of data we use, which are parts of the BATSE 4B catalog \\citep{paciesas99}, and we describe our conical jet geometry, following \\citet{maoyi94}, in $\\S$ 3 we present results. Finally, we conclude with summary of our results and discussions in $\\S$ 4. ", "conclusions": "The theoretical models for GRBs are abundant. Despite remarkable progresses in understanding physical mechanisms involved in these models, the GRB prompt emission mechanisms and engine models have so far been unable to constrain the extent of beaming and the luminosity distribution of GRBs. This in turn has been a major uncertainty in interpreting the observed flux data in terms of the cosmological spatial distribution of the bursts. In this regard, the present work has shown that the simple beaming models and their resulting apparent luminosity functions have significant effects in interpreting the observed data. If the GRBs are indeed standard candles with a single well-defined luminosity, the spatial distribution of GRBs in connection with the cosmic star formation rate could be translated into the cosmological source evolution. However, the luminosity distributions we have considered affect the maximum redshift and the average redshift significantly. It is therefore important to derive a theoretical luminosity function for a given GRB model. The jet models we adopted are obviously over-simplified. Despite this major drawback, the models capture the essential ingredients of the beamed relativistic jets concerning the apparent luminosity function. One of the major uncertainties is that the jets and GRB sources differ greatly and GRB luminosities and jets' physical conditions are intrinsically different in each source. Given the wide range of burst durations and the diverse burst types (Fishman and Meegan 1995, and references therein), such a possibility cannot be ruled out. If this is indeed the case, our standard source approach is not applicable." }, "0005/astro-ph0005134_arXiv.txt": { "abstract": "We propose a model of the accretion flow onto a black hole consisting of the accretion disk with an accreting two-temperature corona. The model is based on assumptions about the radiative and conductive energy exchange between the two phases and the pressure equilibrium. The complete model is determined by the mass, the accretion rate, and the viscosity parameter. We present the radial dependencies of parameters of such a two-phase flow, with advection in the corona and the disk/corona mass exchange due to evaporation/condensation included, and we determine the transition radius from a two-phase disk/corona accretion to a single-phase optically thin flow (ADAF) in the innermost part of the disk as a function of accretion rate. We identify the NLS1 galaxies with objects accreting at a rate close to the Eddington accretion rate. The strong variability of these objects may be related to the limit cycle behaviour expected in this luminosity range, as the disk, unstable due to the dominance by the radiation pressure, oscillates between the two stable branches: the advection-dominated optically thick branch and the evaporation branch. ", "introduction": "Broad band spectra of Seyfert galaxies and X-ray binaries in their hard states are well described by models which consist of a standard optically thick disk, disrupted and replaced by an optically thin hot flow in the innermost part (e.g. Loska \\& Czerny 1997, Esin, McClintock \\& Narayan 1997). The question remains: which physical mechanism is responsible for such a transition? Meyer \\& Meyer-Hoffmeister (1994) suggested in the context of CVs that the accretion flow is basically a two-phase disk/corona flow, with the heat generation in the corona leading to gradual disk evaporation due to the conduction flux, and finally, to the disappearance of the disk in the innermost part of the flow for low accretion rates. Here we follow this basic idea but we adjust it to the parameter range appropriate for accreting black holes. ", "conclusions": "" }, "0005/astro-ph0005378_arXiv.txt": { "abstract": "We study the dynamical stability against bar-mode deformation of rapidly spinning neutron stars with differential rotation. We perform fully relativistic 3D simulations of compact stars with $M/R \\geq 0.1$, where $M$ is the total gravitational mass and $R$ the equatorial circumferential radius. We adopt an adiabatic equation of state with adiabatic index $\\Gamma=2$. As in Newtonian theory, we find that stars above a critical value of $\\beta \\equiv T/W$ (where $T$ is the rotational kinetic energy and $W$ the gravitational binding energy) are dynamically unstable to bar formation. For our adopted choices of stellar compaction and rotation profile, the critical value of $\\beta = \\beta_{dGR}$ is $\\sim 0.24-0.25$, only slightly smaller than the well-known Newtonian value $\\sim 0.27$ for incompressible Maclaurin spheroids. The critical value depends only very weakly on the degree of differential rotation for the moderate range we surveyed. All unstable stars form bars on a dynamical timescale. Models with sufficiently large $\\beta$ subsequently form spiral arms and eject mass, driving the remnant to a dynamically stable state. Models with moderately large $\\beta \\gtrsim \\beta_{dGR}$ do not develop spiral arms or eject mass but adjust to form dynamically stable ellipsoidal-like configurations. If the bar-mode instability is triggered in supernovae collapse or binary neutron star mergers, it could be a strong and observable source of gravitational waves. We determine characteristic wave amplitudes and frequencies. ", "introduction": "Neutron stars in nature are rotating and subject to nonaxisymmetric rotational instabilities. An exact treatment of these instabilities exists only for incompressible equilibrium fluids in Newtonian gravity (see, e.g., Chandrasekhar 1969; Tassoul 1978; Shapiro \\& Teukolsky 1983). For these configurations, global rotational instabilities arise from nonradial toroidal modes $e^{im\\varphi}$ ($m=\\pm 1,\\pm 2, \\dots$) when $\\beta\\equiv T/W$ exceeds a certain critical value. Here $\\varphi$ is the azimuthal coordinate and $T$ and $W$ are the rotational kinetic and gravitational potential binding energies. In the following we will focus on the $m=\\pm 2$ bar mode, since it is the fastest growing mode when the rotation is sufficiently rapid. There exist two different mechanisms and corresponding timescales for bar mode instabilities. Uniformly rotating, incompressible stars in Newtonian theory are {\\em secularly} unstable to bar mode formation when $\\beta \\geq \\beta_s \\simeq 0.14$. However, this instability can only grow in the presence of some dissipative mechanism, like viscosity or gravitational radiation, and the growth time is determined by the dissipative timescale, which is usually much longer than the dynamical timescale of the system. By contrast, a {\\em dynamical} instability to bar mode formation sets in when $\\beta \\geq \\beta_d \\simeq 0.27$. This instability is independent of any dissipative mechanisms, and the growth time is determined by the hydrodynamical timescale of the system. The secular instability in compressible stars, both uniformly and differentially rotating, has been analyzed numerically within linear perturbation theory, by means of a variational principle and trial functions, by solving the eigenvalue problem, or by other approximate means. This technique has been applied not only in Newtonian theory (Lynden-Bell \\& Ostriker 1967; Ostriker \\& Bodenheimer 1973; Ipser and Lindblom 1989; Friedman and Schutz 1977) but also in post-Newtonian theory (Cutler and Lindblom 1992; Shapiro and Zane 1998 for incompressible stars) and full general relativity (Yoshida and Eriguchi 1995; Bonazzola, Frieben and Gourgoulhon 1996; Stergioulas and Friedman 1998). For relativistic stars, the critical value of $\\beta_s$ depends on the compaction $M/R$ of the star (where $M$ is the gravitational mass and $R$ the circumferential radius at the equator), on the rotation law and on the dissipative mechanism. The gravitational-radiation driven instability occurs for smaller rotation rates, i.e. for values $\\beta_s < 0.14$, in general relativity. For extremely compact stars (Stergioulas and Friedman 1998) or strongly differentially rotating stars (Imamura et al. 1995), the critical value can be as small as $\\beta_s < 0.1$. By contrast, viscosity drives the instability to higher rotation rates $\\beta_s > 0.14$ as the configurations become more compact (Bonazzola, Frieben and Gourgoulhon 1996; Shapiro and Zane 1998). Determining the onset of the dynamical bar-mode instability, as well as the subsequent evolution of an unstable star, requires a numerical simulation of the fully nonlinear hydrodynamical equations. Simulations performed in Newtonian theory (e.g.~Tohline, Durisen \\& McCollough 1985; Durisen, Gingold \\& Tohline 1986; Williams \\& Tohline 1988; Houser, Centrella \\& Smith 1994; Smith, Houser \\& Centrella 1995; Houser \\& Centrella 1996; Pickett, Durisen \\& Davis 1996; New, Centrella \\& Tohline 1999) have shown that $\\beta_d$ depends only very weakly on the stiffness of the equation of state. Once a bar has developed, the formation of spiral arms plays an important role in redistributing the angular momentum and forming a core-halo structure. Recently, it has been shown that, similar to the onset of secular instability, $\\beta_d$ can be smaller for stars with a higher degree of differential rotation (Tohline \\& Hachisu 1990; Pickett, Durisen \\& Davis 1996) To date, the dynamical bar-mode instability has been analyzed only in Newtonian theory, hence almost nothing is known about the role of relativistic gravitation. The reason is that until quite recently a stable numerical code capable of performing reliable hydrodynamic simulations in three dimensions plus time in full general relativity has not existed. Some recent developments, however, have advanced the field significantly. New formulations of the Einstein equation based on modifications of the standard $3+1$ ADM system of equations (Arnowitt, Deser \\& Misner 1962) have resulted in codes which have proven to be remarkably stable over many dynamical timescales (e.g., Shibata \\& Nakamura 1995; Baumgarte \\& Shapiro 1999; Oohara \\& Nakamura 1999). In addition, gauge conditions which warrant long-time stable evolution for rotating and self-gravitating systems and are manageable computationally have been developed (e.g., Shibata 1999b). In this paper we adopt the relativistic hydrodynamic implementation of Shibata (1999a) to study the onset and growth of the dynamical bar-mode instability in relativistic stars. Although this study is carried out only for a simple equation of state and rotational law, it demonstrates how, as numerical relativity in full $3+1$ matures, it is becoming more useful as a tool to solve long-standing problems in relativistic astrophysics characterized by strong gravitational fields and little or no spatial symmetry. There are numerous evolutionary paths which may lead to the formation of rapidly rotating neutron stars with $\\beta \\sim 0.3$. The parameter $\\beta$ increases approximately as $R^{-1}$ during stellar collapse. During supernova collapse, the core contracts from $\\sim 1000$ km to $\\sim 10$ km, and hence $\\beta$ increases by about two orders of magnitude. Thus, even moderately rapidly rotating progenitor stars may yield rapidly rotating neutron stars which may reach the onset of dynamical instability (Bonazzola \\& Marck 1993; Rampp, M\\\"uller \\& Ruffert 1998). Similar arguments hold for accretion induced collapse of white dwarfs to neutron stars and for the merger of binary white dwarfs to neutron stars. In fact, recent X-ray and radio observations of supernova remnants have identified several young, isolated, rapidly rotating pulsars, suggesting that these stars may have been born with periods of several milliseconds (Marshall et al.~1998; Kaspi et al.~1998; Torii et at.~1999). These neutron stars could be the collapsed remnants of rapidly rotating progenitors. Rapidly rotating neutron stars may naturally arise in the merger of binary neutron stars. Baumgarte, Shapiro and Shibata (2000) have studied equilibrium configurations of differentially rotating neutron stars and found examples where the maximum allowed mass increases by a factor of about 2 due to differential rotation. This suggests that the merger of binary neutron stars could result in a ``hypermassive'' neutron star which has rest mass exceeding the maximum value for uniformly rotating stars. Recent hydrodynamic simulations in full general relativity indicate that such hypermassive neutron stars can indeed be produced in the merger of moderately compact neutron stars (Shibata \\& Uryu 2000). They show that the remnant is unlikely to exceed the onset point of dynamical instability initially. Subsequent neutrino emission and cooling, however, will make the star shrink in size, leading to an increase in $\\beta$, possibly beyond the onset of nonradial dynamical instability, $\\beta_d$. Rapidly rotating neutron stars experiencing the bar-mode instability could have significant observable consequences. According to Newtonian simulations (Tohline, Durisen \\& McCollough 1985; Durisen Gingold \\& Tohline 1986; Williams \\& Tohline 1988; Houser, Centrella \\& Smith 1994; Smith, Houser \\& Centrella 1995; Houser \\& Centrella 1996), a dynamically unstable star may evolve into a two-component system containing a central star and circumstellar accretion disk. Such a system may be observable in a supernova remnant. In the case of merged binaries, the differentially rotating remnant may be more massive, hot and bloated than a typical rapidly rotating, old pulsar. Consequently, the frequency of gravitational waves excited by the bar-mode instability could be low, {\\it i.e.}, less than 1kHz (see Eq.~(\\ref{wavef}) below), and hence detectable by kilometer-size laser interferometers such as LIGO (Lai \\& Shapiro 1995; Thorne 1995). In this paper we summarize the results of our fully relativistic simulations of bar-mode instabilities in neutron stars. We determine $\\beta_d$ for highly relativistic stars, follow the growth of the bar-mode, and find the frequency and amplitude of the emitted gravitational waves. We implement the numerical scheme described in Shibata (1999a), using differentially rotating neutron stars of high $\\beta$ for initial data. We focus on differentially rotating stars, since uniformly rotating stars do not reach $\\beta \\gtrsim 0.2$ except for extremely stiff equations of state, and hence do not become dynamically unstable to bar-modes (Tassoul 1978). We adopt an adiabatic equation of state with $\\Gamma=2$ as a reasonable qualitative approximation to moderately stiff nuclear equation of state. The adiabatic assumption is justified even for hot neutron stars, since energy dissipation is small over the dynamical timescales of interest. In Sec.~2, we briefly summarize our formulation of the fully relativistic system of equations and our numerical scheme. In Sec.~3, initial models of differentially rotating, equilibrium neutron stars are presented. Following Shibata, Baumgarte \\& Shapiro (2000) we adopt the so-called conformal flatness approximation to prepare differentially rotating neutron stars in (approximate) equilibrium states for computational convenience. To confirm the reliability of this approximation, we also compute numerically exact equilibrium states and demonstrate that this approximation is accurate (cf.~Cook, Shapiro \\& Teukolsky 1996). In Sec.~4, we present our numerical results, focusing on the onset of the bar-mode instability, its early growth and corresponding waveforms and frequencies. We briefly summarize our results in Sec.~5. Throughout this paper, we adopt geometrized units with $G=1=c$ where $G$ and $c$ denote the gravitational constant and speed of light. In numerical simulation, we use Cartesian coordinates $x^k=(x, y, z)$ with $r=\\sqrt{x^2+y^2+z^2}$, $\\varpi=\\sqrt{x^2+y^2}$ and $\\varphi=\\tan^{-1}(y/x)$; $t$ denotes coordinate time. Greek indices $\\mu, \\nu, \\dots$ denote $x, y, z$ and $t$, and Latin indices $i,j,k, \\dots$ denote $x,y$ and $z$. ", "conclusions": "We perform hydrodynamic simulations in full $3+1$ general relativity (GR). We use the same formulation and gauge conditions as in Shibata (1999a), to which the reader may refer for details and basic equations. The fundamental variables used in this paper are: \\beqn \\rho &&:{\\rm rest~ mass~ density},\\nonumber \\\\ \\varep &&: {\\rm specific~ internal~ energy}, \\nonumber \\\\ P &&:{\\rm pressure}, \\nonumber \\\\ u^{\\mu} &&: {\\rm four~ velocity}, \\nonumber \\\\ v^{k}&& ={u^k \\over u^0}; ~~\\Omega =v^{\\varphi}, \\nonumber \\\\ \\alpha &&: {\\rm lapse~ function}, \\nonumber \\\\ \\beta^k &&: {\\rm shift~ vector}, \\nonumber \\\\ \\gamma_{ij} &&:{\\rm metric~ in~ 3D~ spatial~ hypersurface},\\nonumber \\\\ \\gamma &&=e^{12\\phi}={\\rm det}(\\gamma_{ij}), \\nonumber \\\\ \\tilde \\gamma_{ij}&&=e^{-4\\phi}\\gamma_{ij}, \\nonumber \\\\ K_{ij} &&:{\\rm extrinsic~curvature}.\\nonumber \\eeqn Geometric variables, $\\phi$, $\\tilde \\gamma_{ij}$, the trace of the extrinsic curvature $K\\equiv K_{ij}\\gamma^{ij}$, $\\tilde A_{ij}\\equiv e^{-4\\phi}(K_{ij}-\\gamma_{ij}K/3)$, as well as three auxiliary functions $F_i \\equiv \\pa_j \\tilde \\gamma_{ij}$, where $\\pa_j$ is the partial derivative, are evolved with an unconstrained evolution code in a modified form of the ADM formalism (Shibata \\& Nakamura 1995). GR hydrodynamic equations are evolved using a van Leer scheme for the advection terms (van Leer 1977; Hawley, Smarr \\& Wilson 1984). Numerical simulation is performed using Cartesian coordinates. Violations of the Hamiltonian constraint and conservation of mass and angular momentum are monitored as code checks. Several test calculations, including spherical collapse of dust, stability of spherical neutron stars, and the stable evolutions of rigidly and rapidly rotating neutron stars have been described in Shibata (1999a). Simulations using this code and exploring the dynamical (quasi-radial) stability against gravitational collapse of rigidly rotating ``supramassive'' neutron stars, which have rest masses exceeding the TOV limit for a nonrotating spherical star, have been presented in Shibata, Baumgarte \\& Shapiro (2000). A simulation using this code and demonstrating the existence of dynamically stable differentially rotating ``hypermassive'' stars, which have rest masses exceeding the maximum value for uniformly rotating stars, was presented in Baumgarte, Shapiro and Shibata (2000). The stress energy tensor for an ideal fluid is given by \\beq T_{\\mu\\nu}=(\\rho + \\rho \\varep + P)u_{\\mu} u_{\\nu} + P g_{\\mu\\nu}, \\eeq where $g_{\\mu\\nu}$ is the spacetime metric. We adopt a $\\Gamma$-law equation of state \\beq P=(\\Gamma-1)\\rho \\varep,\\label{EOS} \\eeq where $\\Gamma$ is the adiabatic constant. For isentropic configurations the $\\Gamma$-law equation of state can be rewritten in the polytropic form \\beq P = \\kappa \\rho^{\\Gamma}, \\mbox{~~~~~} \\Gamma = 1 + \\frac{1}{n} \\label{eos}, \\eeq where $\\kappa$ is the polytropic constant and $n$ the polytropic index. This is the form that we use for constructing initial data. Throughout this paper, we adopt $n = 1$ as a reasonable qualitative approximation to a moderately stiff, nuclear equation of state for simplicity. Instead of $\\rho$ and $\\varep$ we numerically evolve the densities $\\rho_* \\equiv \\rho \\alpha u^0 e^{6\\phi}$ and $e_* \\equiv (\\rho\\varepsilon)^{1/\\Gamma}\\alpha u^0 e^{6\\phi}$ as the hydrodynamic variables (Shibata, Oohara \\& Nakamura 1997; Shibata 1999a). Since these variables satisfy evolution equations in conservation form, the total rest mass of the system \\beq M_0=\\int d^3 x \\rho_*. \\eeq is automatically conserved, as is the the volume integral of the energy density $e_*$ in the absence of shocks. The time slicing and spatial gauge conditions we use in this paper for the lapse and shift are the same as those adopted in our series of papers (Shibata 1999a, 1999b; Shibata, Baumgarte \\& Shapiro 2000); i.e.~we impose an ``approximate'' maximal slice condition ($K \\simeq 0$) and an ``approximate'' minimum distortion gauge condition ($\\tilde D_i (\\pa_t \\tilde \\gamma^{ij}) \\simeq 0$ where $\\tilde D_i$ is the covariant derivative with respect to $\\tilde \\gamma_{ij}$, see Shibata 1999b). We have performed numerical simulations of rapidly and differentially rotating neutron stars in full $3+1$ general relativity. We treated compact stars of $10 \\gtrsim R/M \\gtrsim 6$ and focussed on their dynamical stability against bar-mode formation. We found that when plotted in a $\\bar M_0$ versus $\\bar \\rho_{\\rm max}$ diagram, a region of stable stars can be clearly distinguished from a region of unstable stars, with the onset of instability almost independent of the degree of differential rotation. We showed that the parameter $\\beta=T/W$ remains a good diagnostic of the onset point of instability in the relativistic domain as it did for Newtonian stars. The critical value for the instability onset depends only weakly on the degree of differential rotation for the models surveyed to date. For those cases we find that $\\beta_{dGR} \\sim 0.24-0.25$, and that $\\beta_{dGR}$ decreases slightly for stars with a higher degree of differential rotation. We also have preliminary evidence that $\\beta_{dGR}$ decreases with compaction as well. We will systematically study this hypothesis with a post-Newtonian numerical analysis in a forthcoming paper (Saijo, Shibata, Baumgarte \\& Shapiro 2000). For selected models, we followed the growth and saturation of bar-mode perturbations up to late times. Stars with sufficiently large $\\beta > \\beta_{dGR}$ develop bars first and then form spiral arms, leading to mass ejection. Stars with smaller values of $\\beta \\sim \\beta_{dGR}$ also develop bars, but do not form spiral arms and eject only very little mass. In both cases, unstable stars appear to form differentially rotating, triaxial ellipsoids once the bar-mode perturbation saturates. Typically, these flattened ellipsoids appear to have $\\beta \\gtrsim 0.2$, so that they would be secularly unstable due to gravitational waves and viscosity. We expect that this secular instability will allow the stars to maintain a bar-like shape for many dynamical timescales, leading to quasi-periodic emission of gravitational waves. We estimate the initial frequency and amplitude of gravitational waves to be $f \\sim (1-1.4) $kHz and $h \\sim 5\\times 10^{-22}$ for stars of mass $\\sim 2.5M_{\\odot}$ and radius $R \\sim 7M$ at a distance of $10$Mpc. The effective amplitude of gravitational waves could be much larger during the subsequent evolution because of the accumulation of quasi-periodic wave cycles (Lai \\& Shapiro 1995). In order to accurately determine the secular evolution of the ellipsoidal star together with emitted gravitational wave signal, a more detailed calculation is necessary. Since the secular timescale is larger than the dynamical timescale by several orders of magnitude, it may be impossible to follow the evolution with a fully dynamical code, even with implicit differencing to avoid the Courant criterion for stability. This suggests that in full GR, the secular evolution problem may best be solved within an appropriate, quasi-stationary scheme similar in spirit to the approach used in stellar evolution calculations." }, "0005/astro-ph0005187_arXiv.txt": { "abstract": "A sample of 59 sight lines to reddened Galactic OB stars was examined for correlations of the strength of the CO Fourth Positive ($A^1\\Pi - X^1\\Sigma^+$) absorption band system with the ultraviolet interstellar extinction curve parameters. We used archival high-dispersion NEWSIPS IUE spectra to measure the CO absorption for comparison to parametric fits of the extinction curves from the literature. A strong correlation with the non-linear far-UV curvature term was found with greater absorption, normalized to \\ebv, being associated with more curvature. A weaker trend with the linear extinction term was also found. Mechanisms for enhancing CO in dust environments exhibiting high non-linear curvature are discussed. ", "introduction": "The extinction of starlight in the ultraviolet (UV; $\\lambda\\lesssim3200$ \\AA) can be characterized by three main components: a linear rise, a Lorentzian-like bump centered on $\\lambda\\sim2175$ \\AA, and a far-ultraviolet ($\\lambda\\lesssim1700$ \\AA) non-linear rise. The general features of the extinction curve can be understood in terms of dust grain populations \\citep[e.g.]{Mathis77,Draine84}. \\citet{CCM88,CCM89} showed that the overall shape of the UV extinction curve can be estimated by the single parameter \\rv [$\\equiv A(V)/E(B-V)$]. However, individual sight lines exhibit large deviations from the average extinction curve, particularly in the far-UV rise \\citep{Mathis92}. A detailed understanding of these deviations can provide information about the nature of the particles responsible for the extinction, putting constraints on interstellar grain models. A number of studies have investigated the possible correlations between the various components of the UV extinction curve and gas phase abundances \\citep{Joseph89,Jenniskens92}, infrared emissions \\citep{Cox87,Hackwell91,Boulanger94}, very broadband structure \\citep{Jenniskens94}, and diffuse interstellar band (DIB) absorptions \\citep{Desert95}. The \\citet{Jenniskens92} study was the first to positively identify a correlation of the strength of the non-linear rise with the abundance of a gas phase atom or molecule, finding that a larger column of CH implies a stronger far-UV non-linear rise, and the inverse relationship for CH$^+$. No such study has to date been performed for CO, a molecule of great astrophysical importance. Being the second most abundant molecule after \\htwo, it is used as a tracer of \\htwo\\ gas and an indicator of the total mass of molecular clouds \\citep{Scoville87,Bloemen89,Hunter94}. Having allowed rotational transitions that are readily observed in emission at radio wavelengths, the CO molecule is commonly used to probe dark molecular clouds. CO also has an extensive band system in the ultraviolet, which can be used to investigate the molecular content of diffuse and translucent clouds when observed in absorption toward background stars. The wavelength coverage of the short-wavelength prime (SWP) camera of the \\textit{International Ultraviolet Explorer} (IUE) satellite, 1150 -- 2000 \\AA, spans as many as 19 bands of the Fourth Positive ($A^1\\Pi - X^1\\Sigma^+$) ($v'$--0) band system beginning with the (0--0) band at 1544.5 \\AA\\ and continuing to shorter wavelengths. For this study, we searched the IUE archival data for ultraviolet absorption of CO along 59 lines of sight toward Galactic O and B stars for which UV extinction curves have already been published, and parameterized as described below, to investigate the possible correlations with the various curve features. \\citet[FM hereafter]{FM90} have proposed a parameterization that accurately fits the extinction curves with a single analytical expression using six parameters. Using this method, the extinction curve, normalized to unit \\ebv\\ and with $x\\equiv\\lambda^{-1}$, can be expressed as \\begin{equation} k(x)\\equiv\\frac{E(\\lambda-V)}{E(B-V)}=c_1+c_2x+c_3D(x;\\gamma,x_0)+c_4F(x) \\label{eq1} \\end{equation} where \\begin{equation} D(x;\\gamma,x_0)=\\frac{x^2}{(x^2-x_0^2)^2+x^2\\gamma^2} \\end{equation} is the Lorentzian-like ``Drude'' profile, representing the 2175 \\AA\\ bump with $x_0$ the bump peak and $\\gamma$ its width, and \\begin{equation} F(x)= \\left\\{ \\begin{array}{ll}0.53(x-5.9)^2+0.05446(x-5.9)^3 & x > 5.9 \\mu m^{-1} \\\\ 0 & x \\leq 5.9 \\mu m^{-1} \\end{array} \\right. \\end{equation} is the far-UV curvature term, whose strength is determined by a single parameter, $c_4$. ", "conclusions": "The abundance of CO, as measured by the equivalent width of the UV absorption of the $A-X$ (2--0) band normalized to unit \\ebv, is correlated with the strength of the non-linear far-UV rise in the UV extinction curves toward Galactic O and B stars. This correlation is indicative of either a decrease in the photodestruction rate of CO with increased extinction or of a dust environment that is conducive to the enhanced formation of CO or its precursor molecules. The linear rise does not correlate with CO abundance, although for sight lines with large \\rv\\ ($\\gtrsim 3.6$) we observe mostly small columns ($\\lesssim10^{15}$ cm$^{-2}$ per \\ebv). These results, together with those of \\citet{Cardelli88} reinforces the notion that the use of the visual extinction, $A_V$, as the independent variable in global studies of interstellar chemistry has to be viewed with some caution. Future study of these relationships could benefit from a quantitative determination of the CO column density and an exploration of the \\htwo/CO ratio in these environments. For \\htwo, the ideal instrument is the recently launched \\textit{Far-Ultraviolet Spectroscopic Explorer} (FUSE) satellite. FUSE will be able to measure the \\htwo\\ column from the individual lines of the Werner and Lyman band systems, as well as observe the absorptions from the CO Hopfield-Birge bands. However, more accurate CO observations could be made by the Space Telescope Imaging Spectrograph (STIS) for two reasons. The wavelength separation of the rotational lines within the $A-X$ bands is greater than that of the $B-X$ and $C-X$ and the STIS Echelle modes have higher resolution than FUSE, allowing for the measurement of the ground state rotational temperature and doppler velocity, and therefore a precise derivation of column density. Finally, we note that the ratio of true absorption to scattering by dust in the far-UV is critical to the interpretation of these results. Studies aimed at investigating this relation in conjunction with the FM extinction parameterization would be useful. Such measurements would facilitate a more detailed analysis of the conditions in these cloud environments." }, "0005/astro-ph0005464_arXiv.txt": { "abstract": "We investigate where in protoplanetary disks magnetorotational instability operates, which can cause angular momentum transport in the disks. We investigate the spatial distribution of various charged particles and the unstable regions for a variety of models for protoplanetary disks taking into account the recombination of ions and electrons at grain surfaces, which is an important process in most parts of the disks. We find that for all the models there is an inner region which is magnetorotationally stable due to ohmic dissipation. This must make the accretion onto the central star non-steady. For the model of the minimum-mass solar nebula, the critical radius, inside of which the disk is stable, is about 20 AU, and the mass accretion rate just outside the critical radius is $10^{-7}$ -- $10^{-6} ~M_{\\odot}~{\\rm yr}^{-1}$. The stable region is smaller in a disk of lower column density. Dust grains in protoplanetary disks may grow by mutual sticking and may sediment toward the midplane of the disks. We find that the stable region shrinks as the grain size increases or the sedimentation proceeds. Therefore in the late evolutionary stages, protoplanetary disks can be magnetorotationally unstable even in the inner regions. ", "introduction": "Various infrared and radio observations have revealed the existence of accretion disks around T Tauri stars, so-called protoplanetary disks, where planet formation may be going on (e.g., Rucinski 1985; Adams, Lada, \\& Shu 1987). Photometric observations at optically thin millimeter wavelengths give disk masses of $\\sim 0.1$ -- $0.01 ~M_{\\odot}$ (Beckwith et al. 1990), and fitting of spectral energy distributions (SEDs) at infrared wavelengths gives disk sizes of tens of AU or 100 AU (Adams, Lada, \\& Shu 1988). Some authors tried to reproduce the SEDs of T Tauri stars with models of viscous accretion disks (e.g., Bertout, Basri, \\& Bouvier 1988), and succeeded in deriving the slopes of SEDs at infrared wavelengths for some T Tauri stars. Optical and ultraviolet veiling of absorption lines has been discovered for some T Tauri stars, which indicates the existence of boundary layers at the interface between a Keplerian accretion disk and a slowly spinning star (Lynden-Bell \\& Pringle 1974). All these suggest the existence of accreting matter in protoplanetary disks at least in their early active stages. However, the mechanisms of angular momentum transport, necessary for accretion, have not yet been well understood, any more than in other astrophysical accretion disks. The magnetorotational instability must be the most promising source of anomalous viscosity in accretion disks (Balbus \\& Hawley 1998, and references therein). Balbus \\& Hawley (1991) showed that astrophysical accretion disks are unstable to axisymmetric disturbances in the presence of a weak magnetic field. The instability sets in under a broad range of conditions and is capable of initiating and sustaining MHD turbulence in accretion disks, as long as the approximation of the ideal MHD holds. However, protoplanetary disks are so cold and dense that the ionization fraction is very low (Umebayashi \\& Nakano 1988). Because the coupling of some charged particles with magnetic fields is not strong enough, both the processes of ambipolar diffusion and ohmic dissipation have to be considered (Nakano 1984). These processes have an effect of suppressing the magnetorotational instability. It is necessary to clarify which parts of the disks are magnetorotationally unstable in relation to the evolution of protoplanetary disks. Blaes \\& Balbus (1994) examined the effect of the ambipolar diffusion on the instability, and showed that differentially rotating disks are unstable when the collision frequency of neutral particles with ions exceeds the local epicyclic frequency. Sano \\& Miyama (1999, hereafter Paper I) investigated the stability of weakly ionized disks including the effect of ohmic dissipation. This process dominates the ambipolar diffusion in the regions of higher density, or in the inner parts (within about 100 AU) of the disks. In paper I, we derived the conditions for the existence of unstable modes by the global linear analysis taking into account the effect of the vertical structure of the disks. The wavelength of the most unstable mode expected from the local linear analysis is approximately given by \\begin{equation} \\lambda_{\\rm local} \\approx \\max \\left( 2 \\pi \\frac{v_{\\rm A}}{\\Omega}, 2 \\pi \\frac{\\eta}{v_{\\rm A}} \\right) \\;, \\label{eqn:lambda} \\end{equation} where $v_{\\rm A} = B / ( 4 \\pi \\rho )^{1/2}$ is the Alfv{\\'e}n speed, $\\Omega$ is the angular rotational velocity of the disk, and $\\eta$ is the magnetic diffusivity (Jin 1996; Paper I). If the magnetic Reynolds number $R_{\\rm m} \\equiv v_{\\rm A}^2 / \\eta \\Omega$ is larger than unity, the ohmic dissipation is not efficient and the wavelength of the most unstable mode is approximately given by $2 \\pi v_{\\rm A} / \\Omega$, which is the result of the ideal MHD. When $R_{\\rm m} \\lesssim 1$, on the other hand, perturbations with shorter wavelengths are stabilized due to the dissipation. Thus, as the magnetic diffusivity increases, $2 \\pi \\eta / v_{\\rm A}$ increases and finally becomes the wavelength of the most unstable mode. The local analysis gives the wavelength $\\lambda_{\\rm local}$ of the most unstable mode as a function of the local values of $v_{\\rm A}$ and $\\eta$. Thus $\\lambda_{\\rm local}$ takes different values at different points in the disk. Our results of the global analysis show that the vertical distribution of the wavelength $\\lambda_{\\rm local} (z)$ is quite important for the understanding of the global stability. In paper I, we have found that the layer at a height $z$ is globally unstable when $\\lambda_{\\rm local} (z)$ obtained by the local analysis is shorter than the scale height of the disk $H$, or \\begin{equation} \\lambda_{\\rm local} (z) \\approx \\max \\left[ 2 \\pi \\frac{v_{{\\rm A}}(z)}{\\Omega}, 2 \\pi \\frac{\\eta(z)}{v_{{\\rm A}}(z)} \\right] \\lesssim H \\;. \\label{eqn:rzcrit} \\end{equation} The condition given by equation~(\\ref{eqn:rzcrit}) agrees with the idea that the stability criterion is approximately given by the requirement that the shortest wavelength for local instability is smaller than the scale height of the disk. Thus the ratio $\\lambda_{\\rm local} / H$ is a good indicator for the existence of global unstable modes. This ratio is sensitive to the magnetic diffusivity, which is determined by the physical quantities in protoplanetary disks such as the density, the temperature, and the ionization fraction. Umebayashi \\& Nakano (1988) examined the radial distribution of magnetic diffusivity on the midplane for the so-called minimum-mass solar nebula proposed by Hayashi, Nakazawa, \\& Nakagawa (1985). However, to make use of equation~(\\ref{eqn:rzcrit}) as the stability criterion of protoplanetary disks, we have to know the vertical distribution of magnetic diffusivity. In this paper we investigate the two-dimensional distribution of magnetic diffusivity for various disk models, and reveal where in protoplanetary disks the magnetorotational instability is operative. Because the recombination processes are more effective in denser regions, the ionization fraction is lower at higher density. Thus the inner dense region of the disks would be decoupled from magnetic fields except for the innermost region where the thermal ionization is efficient. However, even in the inner dense region, the gas near the surface layer of the disks may be well ionized because the density is low. Gammie (1996) proposed a layered accretion model which has the decoupled region (dead zone) sandwiched by the ionized active layers. The ionization by X-rays from the central object would also make a similar structure (Glassgold, Najita, \\& Igea 1997). In such situations, the column densities of the active layers are one of the most important quantities for the dynamical evolution of protoplanetary disks. We shall calculate the column densities of these layers and compare them with the previous works. The recombination of ions and electrons on grain surface is the dominant process in the dense region. Therefore the characters of dust grains affect significantly the ionization fraction and the column density of the active layers. In the previous works grain surface reactions were not included because they were concerned mainly in the late phases of the disks when dust grains of the disks have already settled toward the midplane. In this paper we investigate the effect of dust grains on the abundances of charged particles and the contribution of charged grains to the interaction of the gas with magnetic fields in protoplanetary disks, and show where the magnetorotational instability can be operative. The plan of this paper is as follows. In \\S 2 we describe the disk models adopted in this paper and the magnetic diffusivity of the gas composed of various kinds of charged particles. We also describe the reaction scheme which determines the abundance of charged particles in ionization-recombination equilibrium. In \\S 3 we show numerical results on the spatial distribution of charged particles, the time scale of magnetic field dissipation, and the unstable regions in the disks. Because charged grains can be the dominant charged particles at least in some parts of the protoplanetary disks, their size and abundance are highly important in determining the magnetic diffusivity. Dust grains may grow by mutual sticking. They will sediment onto the midplane if the disk is quiescent. The coupling between the gas and magnetic field varies as the disk evolves in such ways. In \\S 3 we also examine the unstable regions in some model disks which must correspond to different evolutionary stages. In \\S4 we investigate the column densities of active layers and discuss the evolutionary scenario of protoplanetary disks taking account of the magnetorotational instability. We also discuss the validity of our assumptions and some related problems in \\S 4. Finally we summarize our results in \\S 5. ", "conclusions": "\\subsection{Layered Accretion} In paper I we have found that, in some disk conditions, the unstable modes have large amplitudes localized at the upper layer of the disk. This implies that the magnetorotational instability in the nonlinear regime enables the angular momentum transport only in this layer, and thus the layered accretion occurs as proposed by Gammie (1996). This layer corresponds to the unstable region at $r \\lesssim r_{\\rm crit}$ shown by the stripes in Figure~\\ref{fig:d100}. Since this unstable layer locates at a height a few times the scale height of the disk, the column density of this layer is very small. Therefore, the layered accretion might hardly affect dynamical evolution of the disks. However, the thickness of this layer is determined by the distribution of the magnetic diffusivity, which depends not only on the density distribution but also on the size and abundance of grains. First, we examine the dependence of the column density of the unstable layer, $\\Sigma_{\\rm uns}$, on the depletion factor of dust grains, $f_{\\rm g}$. Because the column density of the layer above the critical curve $\\lambda_{\\rm ideal} / H = 1$ is quite small, we approximate $\\Sigma_{\\rm uns}$ as the total column density of the disk subtracted by the column density of the dead zone. Figure~\\ref{fig:col2}a shows $\\Sigma_{\\rm uns}$ as a function of $r$ for $\\beta = 100$ for three cases of grain depletion, $f_{\\rm g} = 1$, $10^{-2}$, and $10^{-4}$. The grain size is taken to be $a = 0.1$ $\\mu$m and the other parameters are the same as in the fiducial model. The column density $\\Sigma_{\\rm uns}$ increases as the grain fraction decreases. Gammie (1996) assumed that the column density of the unstable region was comparable to the attenuation length $\\chi_{\\rm CR} \\approx 96$ g cm$^{-2}$. We have found that the column density $\\Sigma_{\\rm uns}$ is much smaller than $\\chi_{\\rm CR}$ unless $f_{\\rm g} \\lesssim 10^{-4}$. Magnetic fields can be decoupled from the gas even when cosmic rays are not attenuated. Next we examine the dependence of $\\Sigma_{\\rm uns}$ on the grain size. Figure~\\ref{fig:col2}b shows the column density $\\Sigma_{\\rm uns}$ for the models with $a = 0.1$, 0.3, and 1 $\\mu$m. Here we have assumed the grain fraction $f_{\\rm g} = 1$ and $\\beta_{\\rm c} = 100$. The column density $\\Sigma_{\\rm uns}$ increases as the grain radius increases. However, $\\Sigma_{\\rm uns}$ is at most a few g cm$^{-2}$ within $r \\sim 1$ AU, even when the grain is as large as $a = 1$ $\\mu$m. Therefore, unless the abundance of dust grains is reduced to $10^{-4}$ times the interstellar abundance, mass accretion through the unstable layer above the dead zone is very faint and the matter must accumulate around the outer boundary of the dead zone. \\subsection{Mass Accretion Rate} We estimate the mass accumulation rate into the dead zone. Sano, Inutsuka, \\& Miyama (1998) investigated the nonlinear evolution of magnetorotational instability including the effect of ohmic dissipation using two-dimensional MHD simulations. Their results show that, at the saturated turbulent state, the efficiency of angular momentum transport depends on the strength of the initial magnetic field perpendicular to the disk, and that when the magnetic Reynolds number $R_{\\rm m} \\equiv v_{\\rm A}^2 / \\eta \\Omega$ is around unity, the $\\alpha$ parameter of the viscosity (Shakura \\& Sunyaev 1973) is given by \\begin{equation} \\alpha \\approx 1.8 \\times 10^{-2} \\left( \\frac{\\beta}{1000} \\right)^{-1} \\;, \\label{eqn:alpha} \\end{equation} where $\\beta$ is the plasma beta for the initial field. Because the magnetic Reynolds number $R_{\\rm m}$ is close to unity around the critical radius, the mass accretion rate can be estimated with equation~(\\ref{eqn:alpha}). The radial accretion velocity $v_r$ is given by $v_r \\sim \\nu / r$, where $\\nu = \\alpha c_{\\rm s} H$ is the kinematic viscosity. Then the mass accretion rate at $r \\approx r_{\\rm crit}$ is given by \\begin{equation} \\dot{M} \\sim 2 \\pi r_{\\rm crit} \\Sigma v_r \\approx 4.9 \\times 10^{-7} f_{\\Sigma} \\left( \\frac{r_{\\rm crit}}{19 ~{\\rm AU}} \\right)^{-1/2} \\left( \\frac{\\beta_{\\rm c}}{100} \\right)^{-1} \\quad M_{\\odot}~{\\rm yr}^{-1} \\;, \\label{eqn:mdot} \\end{equation} for the disks whose structure is given by equations~(\\ref{eqn:sigr}), (\\ref{eqn:csr}), and (\\ref{eqn:hr}) with $q = 3 / 2$ and $M_{\\ast} = 1 ~M_{\\odot}$. For the case of $\\beta_{\\rm c} = 1000$, the mass accretion rate at $r \\approx r_{\\rm crit} \\approx 22$ AU is an order of magnitude smaller. The typical age of T Tauri stars is estimated to be $10^6$ yr (Kenyon \\& Hartmann 1995), and the lifetime of protoplanetary disks must be comparable to or longer than the age of T Tauri stars. If mass accretion with a rate $\\dot{M} \\sim 10^{-6}$ $M_{\\odot}$~yr$^{-1}$ continues and the accreted matter is distributed relatively widely inside $r \\approx r_{\\rm crit}$, the disk of this part becomes gravitationally unstable at some stage, and the resultant gravitational torque causes mass accretion in this region. This makes the accretion onto the central star non-steady. If the accreting matter accumulates near the critical radius $r \\approx r_{\\rm crit}$ and a dense ring forms, a planet or a brown dwarf would be formed due to fragmentation of the ring (Nakano 1991). In this case, the critical radius gives the position where the first small companion forms. Higher mass disks have higher accretion rate as seen from equation~(\\ref{eqn:mdot}). The mass accretion rate for a disk of $f_{\\Sigma} = 10$ at $r \\approx r_{\\rm crit} \\approx 39$ AU is 7 times larger than that of the fiducial model at $r_{\\rm crit} \\approx 19$ AU. For the model of $f_{\\Sigma} = 0.3$, we have the mass accretion rate $\\dot{M} \\sim 1.8 \\times 10^{-7} ~M_{\\odot}~{\\rm yr}^{-1}$ at $r \\approx r_{\\rm crit} \\approx 13$ AU. Therefore, higher mass disks become gravitationally unstable in shorter time scale. \\subsection{Evolution of Dust Grains in Protoplanetary Disks} Investigating the growth and sedimentation of dust grains in protoplanetary disks, Nakagawa et al. (1981) showed that dust grains can grow by sticking through mutual collisions in thermal Brownian motion, e.g., when the grain size is smaller than microns at $r \\approx 1$ AU. The collision time in Brownian motion, $t_{\\rm col}$, is given by \\begin{equation} t_{\\rm col} = 2 \\pi \\frac{(\\rho_{\\rm g}/3)^{3/2} \\, \\bar{a}^{5/2}}{\\zeta_{\\rm g} \\, \\rho \\, (kT)^{1/2}} = 21 \\left( \\frac{\\rho_{\\rm g}}{3 ~{\\rm g}~{\\rm cm}^{-3}} \\right)^{3/2} \\left( \\frac{\\bar{a}}{1 ~\\mu {\\rm m}} \\right)^{5/2} \\left( \\frac{\\zeta_{\\rm g}}{0.0034} \\right)^{-1} \\left( \\frac{r}{1 ~{\\rm AU}} \\right)^{3/2} t_{\\rm K} \\;, \\label{eqn:growth} \\end{equation} where $\\rho_{\\rm g}$ and $\\bar{a}$ are the internal density and the mean radius, respectively, of dust grains, and $\\zeta_{\\rm g}$ is the mass fraction of dust grains, and the last expression is at the midplane of the fiducial model. As seen from equation~(\\ref{eqn:growth}), dust grains in dense regions grow to micron size in a time comparable to the dynamical time scale, $t_{\\rm K}$, and much shorter than the time scale of sedimentation, \\begin{equation} t_{\\rm sed} = \\frac1{2 \\pi} \\frac{\\rho H}{\\rho_{\\rm g} \\bar{a}} t_{\\rm K} = 5.2 \\times 10^{5} \\left( \\frac{\\rho_{\\rm g}}{3 ~{\\rm g}~{\\rm cm}^{-3}} \\right)^{-1} \\left( \\frac{\\bar{a}}{1 ~\\mu {\\rm m}} \\right)^{-1} {\\rm yr} \\;, \\end{equation} where the last expression is at the midplane of the fiducial model and is independent of $r$. Numerical calculations of Nakagawa et al. (1981) show that the mass fraction of floating grains, $f_{\\rm g}$, at $r = 1$ AU is $\\sim 10^{-1}$, $10^{-2}$, and $10^{-4}$ at the time $t \\sim 2 \\times 10^3$, $10^4$, and $10^5$ yr, respectively, and that the size of floating grains, $a$, is 1 -- 10 $\\mu$m in the late stages. Although laminar flow was assumed in their analysis, we discuss the evolution of the unstable regions using their results. Figure~\\ref{fig:a}d shows the unstable region for the model with $a = 1$ $\\mu$m, an order of magnitude larger than in the fiducial model. The critical radius on the midplane $r_{\\rm crit}$ is 4.6 and 5.8 AU for the cases of $\\beta_{\\rm c} = 100$ and $1000$, respectively. The dead zone shrinks as the sedimentation of dust grains proceeds; in the cases of $f_{\\rm g} = 10^{-1}$, $f_{\\rm g} = 10^{-2}$, and $f_{\\rm g} = 10^{-4}$, the critical radius of the models with $a = 1$ $\\mu$m is 3.2 (3.8), 2.3 (2.7), and 1.4 (1.6) AU, respectively, for $\\beta_{\\rm c} = 100$ (1000). In all these cases the critical radius $r_{\\rm crit}$ is less than 6 AU. If dust grains in protoplanetary disks could grow to micron size, most parts of the disk become magnetorotationally unstable. Therefore, sedimentation of dust grains must be prevented by turbulence induced by this instability contrary to the assumption of Nakagawa et al. (1981). According to the standard scenario of solar system formation (Hayashi et al. 1985), planets form by accretion of planetesimals, which are produced through gravitational instability of a thin dust layer (Goldreich \\& Ward 1973). The thin layer is formed by sedimentation of dust, which must occur only in a quiescent disk (Mizuno, Markiewicz, \\& V{\\\"o}lk 1988). By sedimentation the disk becomes magnetorotationally unstable and turbulent, preventing sedimentation as shown in \\S 3.3. Therefore we have to investigate the gas motion and the evolution of dust simultaneously. \\subsection{Ionization by X-Rays} So far we have considered only cosmic rays and radioactive elements as the ionization sources of protoplanetary disks. In this subsection we consider the effect of photo-processes. The protoplanetary disks are exposed to the UV radiation from the central star and from the interstellar space. Because of the very small attenuation length, a few $\\times$ 10$^{-3}$ g cm$^{-2}$, the ionization by UV radiation is inefficient except for very thin surface layers. Young stellar objects are strong X-ray sources (e.g., Glassgold, Feigelson, \\& Montmerle 2000). The X-ray luminosity of low-mass young stellar objects is typically in the range $L_{\\rm X} \\sim 10^{28}$ -- $10^{30}$ erg s$^{-1}$, or a factor of $10^2$ -- $10^3$ above the contemporary solar levels. Some X-ray sources exhibit high-amplitude rapid flares with peak luminosity $L_{\\rm X} \\sim 10^{30}$ -- $10^{32}$ erg s$^{-1}$. These emissions require the presence of a large volume of high-density plasma at a temperature of about $10^{7}$ K. The plasma must be magnetically confined, probably in large loops on a scale comparable to or larger than the X-ray emitting star itself. Field lines linking the star with the disk at the corotating radius may explain the strongest of these powerful flares (Shu et al. 1994; Hayashi, Shibata, \\& Matsumoto 1996). The X-rays can induce a wide variety of changes in chemical and physical properties of protoplanetary disks. Soft X-rays can be the most promising ionization sources for protoplanetary disks as well as cosmic rays. While the underlying ionization mechanism is due to electronic collisions in both cases, X-rays are absorbed in a smaller column of matter than cosmic rays. Thus cosmic rays provide a more global, low-level ionization of the disks, whereas X-rays can produce a localized high-level ionization. Igea \\& Glassgold (1999) investigated the X-ray transfer and ionization in the disks around young stellar objects using a Monte Carlo method. The X-ray emitting region was modeled as a ring of radius $r = 10 ~R_{\\odot}$ at height $z = 10 ~R_{\\odot}$ with $L_{\\rm X} = 10^{29}$ erg s$^{-1}$ based on the x-wind model of Shu et al. (1994). They obtained the ionization rates at 1 -- 10 AU of the minimum-mass solar nebula as functions of the vertical column density into the disk by hydrogen number, $N_{\\perp}$. Their results show that the ionization rate by X-rays is larger than that by cosmic rays, $\\zeta_{\\rm CR} \\approx 10^{-17}$ s$^{-1}$, only at $N_{\\perp}$ smaller than $10^{25}$ to $2 \\times 10^{23}$ cm$^{-2}$ depending on $r$, which is much smaller than the total column density of the disk at each $r$. Therefore, most of our results are preserved even if X-ray ionization is taken into account. For instance, the thickness of the dead zone is hardly affected by the X-ray ionization. Even if cosmic rays are excluded from the inner regions of protoplanetary disks by, e.g., stellar winds (Parker 1960), the active surface layer can be maintained by the X-ray ionization (Igea \\& Glassgold 1999). Igea \\& Glassgold (1999) neglected the effects of dust grains on ion densities and on interaction with magnetic fields. However, the column densities of unstable regions $\\Sigma_{\\rm uns}$ must be affected by the characters of dust grains in the disk in the same way as in the cases with the cosmic ray ionization investigated in \\S 3. \\subsection{Uncertainties in the Disk Temperature} \\label{sec:temp} The temperature distribution given by equation~(\\ref{eqn:temp}) with $p = 1 / 2$ holds for the disk transparent to the radiation from the central star. Actually, the inner dense regions of protoplanetary disks would be optically thick and there would be some other heating sources such as viscous heating and dissipation of magnetic fields. However, the results are not very sensitive to the temperature as long as it is below several hundred Kelvin where the thermal ionization is inefficient. Figure~\\ref{fig:temp} shows the column densities of the unstable layers for the models with temperature ten times higher (dashed curve) and ten times lower (dot-dashed curve) than the fiducial model (solid curve). The other parameters are the same as those of the fiducial model. The difference in $\\Sigma_{\\rm uns}$ is surprisingly small for such large changes in the temperature. We also investigated for $p = 0$ and $3 / 4$, and obtained quite similar results assuming as above that the thermal ionization is inefficient. When the temperature is higher than several hundred Kelvin, the thermal ionization becomes efficient and there would be significant effects on the instability. This would happen only in the innermost region $r \\lesssim 0.1$ AU for the minimum-mass model, which is much smaller than the dead zones obtained in this paper. Notice that the temperature of steady accretion disks ($p = 3 / 4$) with typical mass accretion rate is within the range we examined here. \\subsection{Validity of Ionization-Recombination Equilibrium} We have used the densities of charged particles obtained by assuming that the gas is in ionization-recombination equilibrium. This assumption is valid when the relaxation time to an equilibrium state for each kind of charged particles is much shorter than the characteristic time scale of protoplanetary disks. The relaxation time $t_{\\rm r}({\\rm X}_i)$ to the ionization-recombination equilibrium for a species X$_{i}$ is approximately given by the minimum of \\begin{equation} \\left| \\frac{n({\\rm X}_i)}{\\beta_{ijk} n({\\rm X}_j)n({\\rm X}_k)} \\right| \\;. \\end{equation} We estimate $t_{\\rm r}({\\rm X}_i)$ by using the equilibrium abundances of particles adopted in the previous section. Figure~\\ref{fig:ti}a shows the time scales near the midplane as functions of $r$, and Figure~\\ref{fig:ti}b shows those in the Earth's region ($r = 1$ AU) as functions of $|z|$, for the fiducial model. As seen from Figure~\\ref{fig:ti}a, the gas near the midplane is almost in ionization-recombination equilibrium at $r \\gtrsim 1.3$ AU because $t_{\\rm r}$ is smaller than $t_{\\rm K}$ for all the particles. At $r \\lesssim 1$ AU, however, $t_{\\rm r}$ for neutral grains is much longer than $t_{\\rm K}$. Because $t_{\\rm r}$ for ions and electrons is much shorter than $t_{\\rm r}$ for grains, ions and electrons always take quasi-equilibrium abundances determined for the temporary charge-state distribution of grains. When the charge-state distribution deviates significantly from the equilibrium, the relaxation time to its true equilibrium is about 250 yr, much longer than $t_{\\rm K}$ (see Fig~\\ref{fig:ti}a). However, because this is much shorter than the sedimentation time of grains in this region of the solar nebula (Nakagawa et al. 1981), the ionization-recombination equilibrium can be attained even in this region. In Earth's region the time scale $t_{\\rm r}$ for metal ions and electrons is longer than $t_{\\rm K}$ only at $|z| \\gtrsim 3.7 ~H$ as seen from Figure~\\ref{fig:ti}b. Because most of the disk matter is inside this layer, a possible deviation from the ionization-recombination equilibrium hardly affects the results obtained in the previous sections and subsections. As the depletion factor $f_{\\rm g}$ of dust grains decreases, the time scale $t_{\\rm r}$ for dust grains decreases considerably in regions where ions and electrons are the dominant charged particles. For ions and electrons $t_{\\rm r}$ remains at least 10 times smaller than $t_{\\rm K}$ even at such late evolutionary stages as $f_{\\rm g} = 10^{-4}$. In the inner regions where dust grains are the dominant charged particles, $t_{\\rm r}$ for dust grains increases extensively as $r$ decreases, and it finally exceeds $t_{\\rm K}$ at a certain position. For the case of $f_{\\rm g} = 10^{-4}$ the time scale $t_{\\rm r}$ for charged grains exceeds $t_{\\rm K}$ only at $r \\lesssim 0.66$ AU, and its value is at most 300 times longer than $t_{\\rm K}$ even in the innermost region $r \\lesssim 0.5$ AU. Because the evolutionary time scale of the protoplanetary disk is much longer than $t_{\\rm K}$, we can regard that the solar nebula is in ionization-recombination equilibrium even at stages of $f_{\\rm g} \\approx 10^{-4}$ except at high $|z|$ where little matter exists. The situation is qualitatively the same for the other disks investigated in the previous section." }, "0005/astro-ph0005008_arXiv.txt": { "abstract": "The properties of globular clusters in dwarf galaxies are key to understanding the formation of globular cluster systems, and in particular in verifying scenarios in which globular cluster systems of larger galaxies formed (at least partly) from the accretion of dwarf galaxies. Here, we revisit the globular cluster system of the dE,N galaxy NGC 3115 DW1 -- a companion of the nearby S0 galaxy NGC 3115 -- adding Keck/LRIS spectroscopy and HST/WFPC2 imaging to previous ground-based photometry. Spectra for seven globular clusters reveal normal abundance ratios with respect to the Milky Way and M31 clusters, as well as a relatively high mean metallicity ([Fe/H]$\\approx -1.0\\pm0.1$ dex). Crude kinematics indicate a high velocity dispersion within 10 kpc which could either be caused by dark matter dominated outer regions, or by the stripping of outer globular clusters by the nearby giant galaxy NGC 3115. The total galaxy mass out to 3 and 10 kpc lies between $1\\cdot10^{10}$ and $1\\cdot10^{11}$M$_\\odot$ and $2\\cdot10^{10}$ and $4\\cdot10^{11}$ M$_\\odot$, respectively, depending on the mass estimator used and the assumptions on cluster orbits and systemic velocity. The HST imaging allows measurement of sizes for two clusters, returning core radii around 2.0 pc, similar to the sizes observed in other galaxies. Spectroscopy allows an estimate of the degree of contamination by foreground stars or background galaxies for the previous ground-based photometry, but does not require a revision of previous results: NGC 3115 DW1 hosts around $N_{\\rm GC}=60\\pm20$ clusters which corresponds to a specific frequency of $S_{\\rm N}=4.9\\pm1.9$, on the high end for massive dEs. Given its absolute magnitude ($M_V=-17.7$ mag) and the properties of its cluster system, NGC 3115 DW1 appears to be a transition between a luminous dE and low-luminosity E galaxy. ", "introduction": "The study of globular cluster systems of dwarf galaxies complements the numerous studies of such systems in larger elliptical and spiral galaxies. Few globular cluster systems (GCSs) around dwarf galaxies beyond the Local group have been studied to date with respect to their cluster system (see Ashman \\& Zepf 1998). This is mostly due to the low numbers of globular clusters present in such galaxies. However, their properties are relevant for a number of globular cluster system formation scenarios. Dwarf galaxies are expected to provide insight into how the smallest galaxies build up a system of globular clusters. Further, their properties must be known in order to verify scenarios in which larger globular cluster systems are predicted to build up by the accretion of proto-galactic fragments or dwarf galaxies (\\citeauthor{kisslerpatig98} 1998a, \\citeauthor{cote98} 1998, \\citeauthor{hilker99} 1999). These scenarios relate to the older idea that galaxy halos might have formed through the assembly of such small stellar systems \\citep[e.g.][]{searlezinn78}. Photometric studies of several globular cluster systems in dwarf galaxies were carried out by \\citet{durrell96b} and \\citet{miller98}. \\citet{durrell96b} studied the systems of 11 dwarf galaxies in the Virgo cluster. All were found to host globular cluster candidates and have specific frequencies ranging from 3 to 8, similar to Local Group dwarfs and giant elliptical galaxies. \\citet{miller98} studied 24 dwarf ellipticals in the Virgo and Fornax clusters as well as in the Leo group. They found that dE,N galaxies had higher specific frequencies than dE galaxies, with values around $S_{\\rm N}=6.5\\pm1.2$, increasing with increasing $M_V$ (decreasing luminosity). Not much is known yet about the metallicities of globular clusters in dwarf galaxies. \\citet{minniti96} constructed a metallicity distribution for all Local Group dwarf galaxies and noticed that the distribution was peaked around [Fe/H]$\\approx -1.7$ dex with no clusters more metal-rich than [Fe/H]$=-1.0$ dex. \\citet{durrell96b} derived metallicities from Washington colors for two of their Virgo dwarf ellipticals and obtained a mean metallicity of [Fe/H]$=-1.45\\pm0.2$ dex. \\citet{durrell96} studied the GCS of the dE NGC 3115 DW1 in more detail and found it to be relatively rich (see below for a more detailed description of their results). This motivated us to carry out spectroscopy for some of the globular cluster candidates in this galaxy to get a more detailed picture of their chemical and kinematical properties. Further, HST/WFPC2 data were available from the archive, allowing us to study the sizes of some of the clusters. NGC 3115 DW1 is a dE1,N galaxy in the vicinity of the giant S0 galaxy NGC 3115. It is located RA: 10h 05m 41.6s; Dec: $-07^{\\rm o}$ 58\\arcmin\\ 53.5\\arcsec\\ ($l=248.12^{\\rm o}$; $b=36.69^{\\rm o}$). We will assume a distance of $11^{+5.0}_{-2.3}$ Mpc throughout this paper following \\citet{durrell96}. Additional properties will be given in the text where they are relevant. In Section 2 we describe the new data. In Section 3 we analyze the spectroscopic data giving first a brief kinematical study of the globular cluster system before discussing the abundances and the overall metallicity of the system. In Section 4 we revisit the previous photometry (number of clusters, colors, specific frequency), compare the photometric metallicities with the spectroscopic ones and add the HST/WFPC2 imaging to derive sizes for two clusters. In Section 5 we discuss whether NGC 3115 DW1 could have suffered stripping by its giant companion. We summarize our results in Section 6. ", "conclusions": "\\label{ln:discussion} In section \\ref{ln:rotdisp}, we derived a high globular cluster velocity dispersion, and thus a high galaxy mass, when we included the two outermost globular clusters. The high mass is not unexpected given the bright absolute magnitude of NGC 3115 DW1. Based on its $M_B=-16.8$ mag \\citep{durrell96}, NGC 3115 DW1 falls in the transition region between dwarfs and ellipticals in the mass-luminosity relation of \\citet{dekel86} (see Fig.~3 therein). Its high mass ($M_{\\rm PME}=(4.8\\pm2.3)\\cdot10^{10}$M$_\\odot$) and the high velocity dispersion ($\\sigma=130\\pm15$ km s$^{-1}$, see below) are more consistent with an elliptical galaxy. We therefore discuss whether the two outermost clusters could be in the process of being stripped by the nearby giant S0 galaxy NGC 3115. \\subsection{Possible Stripping?} Figure \\ref{ps:fov} and Table \\ref{tab:specphot} show that 2 (L1 and L63) of the 7 globular clusters have significantly larger projected radii, i.e. 161.4\\arcsec\\ ($8.6$ kpc, L1) and 189.4\\arcsec\\ ($10.1$ kpc, L63), than the ``inner'' ($r\\leq56.3$\\arcsec $=3$ kpc) globular clusters. These large projected distances from NGC 3115 DW1 could be due to stripping by the nearby S0 galaxy NGC 3115. Figure \\ref{ps:localgroup} shows the relative positions of NGC 3115 DW1 and NGC 3115. The projected distance between the two galaxies is 17.3\\arcmin\\ which corresponds to 55 kpc at the distance of $d\\approx$11 Mpc. The mean radial velocities of L1 and L63 are $v_{\\rm rad}=420\\pm29$ km s$^{-1}$ and $v_{\\rm rad}=605\\pm74$ km s$^{-1}$, respectively. Only L1 shows a significant deviation from the systemic velocity of NGC 3115 DW1 \\citep[$v_{\\rm rad}=698\\pm74$ km s$^{-1}$,][]{capaccioli93} and NGC 3115 \\citep[$v_{\\rm rad}=663\\pm6$ km s$^{-1}$,][]{capaccioli93}. We expect no contamination from globular clusters of the nearby galaxy NGC 3115. \\citet{kavelaars97} found the surface density over-abundance of globular clusters around NGC 3115 (power-law index of radial distribution $\\alpha=-1.8\\pm0.5$) disappearing at 6\\arcmin\\ radius from the center of NGC 3115 (at a photometric limit of $V=23.5$ mag). The globular clusters L1 and L63 have a radial distance to NGC 3115 of $\\approx14$\\arcmin . The projected GC surface density of the GCS of NGC 3115 at the position of these two clusters is $<0.01$ arcmin$^{-2}$. The extrapolated GC surface density of NGC 3115 DW1 at this position lies between 0.2 and 6.9 GCs arcmin$^{-2}$, given the large uncertainties on the density profile. As the numbers are too small (we only found 2 clusters to the north and 0 to the south) it cannot statistically be concluded whether the two globular clusters found in the northern field are chance detections or a statistically significant overabundance. Assuming that both galaxies are roughly at the same distance, we can estimate the dwarf galaxy's gravitational potential and the ratio of potentials of NGC 3115 DW1 and NGC 3115. Both globular clusters are at about 1/5 of the distance separating NGC 3115 DW1 and NGC 3115. As a rough estimate, we assume that NGC 3115 DW1 and NGC 3115 have similar M/L$_V$. In this case, the ratio of their $M_V$'s would imply that NGC 3115 has a mass 10 times larger than NGC 3115 DW1. Hence, the gravitational potentials are comparable at the projected position of the distant globular clusters (L1 and L63). Since the mass of NGC 3115 is likely to be higher than the adopted value (assuming an extended dark matter halo) the motion of both globular clusters is no longer dominated by the gravitational potential of NGC 3115 DW1 alone. Both clusters could then be considered as intergalactic globular clusters. Note that stripping of globular clusters appears to be common among interacting galaxies. \\citet{dacosta95} show that four globular clusters of the Sagittarius dSph are in the process of being stripped by the Milky Way and are being added to its globular cluster system. Other studies have indicated that stripping may be important in galaxy clusters (e.g. in the Fornax cluster \\citeauthor{kisslerpatig99} 1999, \\citeauthor{hilker99} 1999). However, there are no other (optical) hints of interaction from NGC 3115 DW1's stellar light. \\citet{durrell96} found the isophotes to be consistent with little or no tidal disruption out to a projected radius of 60\\arcsec\\ (corresponding to 3.2 kpc) where their photometric errors start to dominate. A simple test for the stripping hypothesis would be a wide-field study of the system in order to rule out (spectroscopically) the presence of any similar clusters around NGC 3115 DW1. \\subsection{The Expected Velocity Dispersion} A look at the fundamental plane of dwarf elliptical galaxies \\citep[e.g.][]{peterson93} shows that NGC 3115 DW1 fits reasonably well into the relation for dwarf and giant elliptical galaxies, under their assumption of $M_V=-16.7$ mag. Adopting the absolute magnitude of $M_V=-17.7$ mag \\citep{durrell96b} the galaxy falls slightly off the relation and would imply a higher velocity dispersion than measured in the central 3\\arcsec . With the measured velocity dispersion of $\\sigma=74\\pm36$ km s$^{-1}$ for the 5 globular clusters inside $r<56.3$\\arcsec\\ ($R<3$ kpc) we obtain from the fundamental-plane relation of \\citet{peterson93} an absolute magnitude of $M_V=-18.0\\pm0.5$ mag. The measured velocity dispersion for the total sample of 7 clusters inside $r<189.4$\\arcsec\\ ($R<10.1$ kpc) of $\\sigma=130\\pm15$ km s$^{-1}$, would correspond to far brighter absolute magnitude ($M_V=-19.5\\pm0.5$ mag) than the measured $M_V=-17.7$ mag. This discrepancy can be explained by a close encounter and subsequent stripping of the dwarf galaxy's halo by the nearby S0 galaxy NGC 3115. Stripping of outer halo regions might well have introduced violent perturbations and led to an enhanced velocity dispersion of the halo region (which is traced by the globular clusters). The relaxation time of such a system far exceeds the Hubble time \\citep{binney94} and therefore it is not possible to reject this scenario just from considerations of dynamical timescales. Alternatively, a high velocity dispersion in the outskirts of a galaxy could be due to a dark-matter dominated massive halo. The outer parts of a number of lower-luminosity Local Group galaxies are known to be dominated by dark matter \\citep[e.g.][]{mateo98}. This picture could explain the fact that we measure an uncommonly high mass for a dwarf elliptical (see sec.~\\ref{ln:massestimate}) at a projected radius of 189.4\\arcsec\\ ($10.1$ kpc). We cannot discriminate between the above possibilities at this point." }, "0005/astro-ph0005522_arXiv.txt": { "abstract": "We present high resolution interferometric and single dish observations of molecular gas in the Serpens cluster-forming core. Star formation does not appear to be homogeneous throughout the core, but is localised in spatially- and kinematically-separated sub-clusters. The stellar (or proto-stellar) density in each of the sub-clusters is much higher than the mean for the entire Serpens cluster. This is the first observational evidence for the hierarchical fragmentation of proto-cluster cores suggested by cluster formation models. ", "introduction": "\\label{sintro} It is generally accepted that most stars are born in clusters (cf. Clarke et al.~\\cite{CBH00}). The way in which clusters form and evolve is therefore likely to influence the distribution of masses for stars in the field, the initial mass function, IMF (Salpeter~\\cite{S55}; Scalo~\\cite{S86}). In very young, embedded clusters the distribution of stellar masses is often similar to the IMF (Palla \\& Stahler~\\cite{PS99}; Meyer et al.~\\cite{Mea00}), and the mass spectra of prestellar and protostellar condensations in the Serpens and $\\rho$--Ophiuchi cluster-forming cores are also consistent with the IMF (Testi \\& Sargent~\\cite{TS98}, hereafter TS98; Motte et al.~\\cite{MAN98}). It has been suggested that within stellar clusters star formation occurs preferentially in sub-clusters where the stellar density is much enhanced (Clarke et al.~\\cite{CBH00}). This has important implications for cluster evolution. For example, more massive stars could be produced by coalescence (Stahler et al.~\\cite{SPH00}). To date, there is little evidence for sub-clustering in the Orion Nebula Cluster (Bate et al.~\\cite{BCM98}) or in the smaller clusters around intermediate mass pre-main sequence stars (Testi et al.~\\cite{TPN99}). However, models advocate sub-clusters with a much higher stellar density at the time of formation (Bonnell et al.~\\cite{Bea98}). It is therefore important to establish if sub-clustering is present in the very youngest clusters. The mean stellar densities and the stellar to gas mass ratio in such sub-clusters can provide critical observational constraints on coalescence, competitive accretion and binary evolution models (Bonnell et al.~\\cite{Bea97}; \\cite{Bea98}; Kroupa~\\cite{Kea99}). The Serpens molecular cloud, at $\\sim310$~pc (de~Lara et al.~\\cite{dLea91}), is one of the most active nearby cluster-forming cores. Inside the 500--1500~M$_\\odot$, $\\sim$0.6~pc diameter cloud of molecular gas (White et al.~\\cite{WCE95}) is a young protocluster comprising about one hundred embedded young stellar objects (YSOs), protostars and prestellar clumps (Strom et al.~\\cite{SVS76}; Eiroa \\& Casali~\\cite{EC92}; Giovannetti et al.~\\cite{Gea98}; Kaas~\\cite{K99}; Casali et al.~\\cite{CED93}; Hurt \\& Barsony~\\cite{HB96}; TS98). Numerous jets and molecular outflows have also been detected (Rodr\\'{\\i}guez et al.~\\cite{Rea89}; White et al~\\cite{WCE95}; Eiroa et al.~\\cite{Eea97}; Herbst et al.~\\cite{HBR97}; Wolf-Chase et al.~\\cite{WCea98}; Davis et al.~\\cite{Dea99}; Hodapp~\\cite{H99}; Hogerheijde et al.~\\cite{H99}). The total estimated mass of the YSOs, protostars, and prestellar clumps is in the range 40--80~M$_\\odot$ (Giovannetti et al.~\\cite{Gea98}; TS98), implying an overall star formation efficiency of 2--5\\%, similar to most nearby molecular clouds. The proto-cluster radius, $\\sim 0.2$~pc, and mean stellar density, $\\sim 400$-$800$~stars/pc$^3$, are typical of very young embedded clusters (Testi et al.~\\cite{TPN99}), making this an ideal laboratory for studying early cluster formation processes. Here, we present wide field, high resolution, aperture synthesis and single dish millimeter-wave molecular line observations of the Serpens core which support the concept of at-birth sub-clustering. ", "conclusions": "Our new interferometer and single-dish maps of the Serpens core indicate sub-clustering at an early epoch of cluster formation. Three separate properties argue for sub-clustering, spatial segregation, outflow orientations, and circumcluster gas kinematics. Approximately one third of the near infrared cluster members are concentrated in a $\\sim$0.1~pc radius region surrounding SVS-2 and SVS-20 (Giovannetti et al.~\\cite{Gea98}; Kaas~\\cite{K99}), while the millimeter and sub-millimeter sources are largely concentrated in the SE (A, B) and NW (C, D) fragments (TS98; Davis et al.~\\cite{Dea99}). The orientations of the outflows observed in the SE and NW are quite different. In the NW, all three flows are oriented along p.a.$\\sim$140$^\\circ$, as is the near infrared reflection nebula centered on SMM5 (Kaas~\\cite{K99}; see also Figure~\\ref{fcsmap}b). In the SE, the two outflows are aligned approximately north-south, with mean p.a.$\\sim$175$^\\circ$. In Figure~3, the NW and SE sub-clusters are embedded in discrete \\nduehp\\ clumps, separated in velocity by $\\sim$1~km/s (Table~1). Each clump comprises two cores: peak velocities of cores A and B differ by only 0.5~km/s ($\\sim 1/2$~$\\Delta$V), C and D by 0.1~km/s ($\\sim 1/10$~$\\Delta$V). Thus the spatially distinct clumps are also kinematically separated, while there is reasonable internal velocity coherence. It appears that the Serpens core encompasses at least three sub-structures -- the NW and SE sub-clusters and the NIR cluster. Star formation is currently occurring simultaneously in the NW and SE sub-clusters, both of which contain roughly equal fractions of prestellar, protostellar and infrared sources. The NIR cluster is probably more evolved but we see no evidence of the progressive pattern of star formation proposed by Casali et al.~(\\cite{CED93}). The kinematics and outflow orientations indicate that each subcluster originated in a separate fragment of the cloud which subsequently fragmented into the smaller cores seen in the \\nduehp\\ maps. Within these are the 3~mm continuum cores that are likely progenitors of single stellar systems (TS98). Taken together these observations are consistent with either the hierarchical fragmentation picture advocated by Elmegreen~(\\cite{E97}; \\cite{E99}) or the spontaneous fragmentation suggested by Myers~(\\cite{M98}) for the formation of stellar clusters, and provide the first observational support for hierarchical fragmentation within a cluster-forming core. We note that while the mean stellar/protostellar density of the entire Serpens core is $\\sim$400-800~stars/pc$^3$, most of the proto-cluster members are within the three subclusters where densities reach 2000-4000~stars/pc$^3$. It is very likely that, within a few million years, the cluster will evolve to a size and mean density very similar to those of embedded clusters around intermediate mass stars. \\smallskip \\noindent {\\bf Acknowledgements:} We thank Cathie Clarke and the referee, Paul Ho, for comments which much improved this paper. The Owens Valley millimeter-wave array is supported by NSF grant AST-96-13717. Research on young star and disk systems is also supported by the {\\it Norris Planetary Origins Project} and NASA's {\\it Origins of Solar Systems} program (through grant NAGW--4030). The FCRAO observations were supported by NSF grant AST-97-25951. JSO thanks the Cornell University Department of Astronomy for continuing support and warm hospitality." }, "0005/astro-ph0005536_arXiv.txt": { "abstract": "We measure redshifts for 67 galaxies in the field of six cluster candidates from the ESO Imaging Survey (EIS). The cluster candidates are selected in the EIS patches C and D among those with estimated mean redshifts $0.5 \\leq z \\leq 0.7$. The observations were made with EFOSC2 at the 3.6m ESO telescope. In the six candidate cluster fields, we identify 19 possible sets of 2 to 7 galaxies in redshift space. In order to establish which of the 19 sets are likely to correspond to real dense systems we compare our counts with those expected from a uniform distribution of galaxies with given luminosity function. In order to take into account the effect of the Large Scale Structure, we modulate the probability computed from the luminosity function with random samplings of the Canada-France Redshift Survey. We find that four out of six candidate EIS clusters are likely to correspond to real systems in redshift space ($> 95$~\\% confidence level). Two of these systems have mean redshift in agreement with the redshift estimate given by the matched filter algorithm ($\\Delta z = \\pm 0.1$). The other two systems have significantly lower redshifts. We discuss the implications of our results in the context of our ongoing research projects aimed at defining high-redshift optically-selected cluster samples. ", "introduction": "\\label{s-int} Clusters of galaxies are the largest virialized structures observed in the Universe. Since they arise from exceptionally high peaks of the primordial fluctuation density field, their properties are highly sensitive to the nature of such cosmic fluctuations. Therefore, the mass function of both local (e.g. White et al. \\cite{whi93}; Girardi et al. \\cite{gir98}) and distant clusters (e.g. Oukbir \\& Blanchard \\cite{ouk92}; Carlberg et al. \\cite{car97}; Eke et al. \\cite{eke98}; Borgani et al. \\cite{bor99}) is a powerful tool to constrain cosmological models for the formation and evolution of cosmic structures. Moreover, clusters are useful laboratories for testing models of galaxy evolution. While early-type galaxies only show evidence for passive evolution (e.g. Stanford et al. \\cite{sta98}), the fraction of blue galaxies increases significantly with redshift (Butcher \\& Oemler \\cite{but78}), at least up to $z \\sim 0.5$, and the fraction of S0's decreases (Dressler et al. \\cite{dre99}). It is therefore essential to have reliable cluster catalogues over the largest possible redshift range. Most distant clusters, at $z \\geq 0.5$, have so far been identified through optical follow-ups of X-ray selected clusters (see, e.g. Gioia et al. \\cite{gio90} and Rosati et al. \\cite{ros00} for a recent review), or by looking at the environment of high-redshift radio galaxies (e.g. Smail \\& Dickinson \\cite{sma95}; Deltorn et al. \\cite{del97}). In the optical, clusters at $z \\simeq 0.5$ and beyond started to be classified in the 80's (Gunn et al. \\cite{gun86}). In the 90's a large catalogue of objectively selected distant clusters, identified in the optical, became available (Postman et al. \\cite{pos96}). These last clusters are identified using a matched-filter algorithm using both positional and photometric data. In brief, this algorithm filters a galaxy catalogue to remove fluctuations in the projected distribution of galaxies that are not likely to be galaxy clusters. For this purpose, the filter is built around parametrizations of the spatial distribution and luminosity function of cluster galaxies. This algorithm also provides an estimate of the redshift for each candidate cluster (hereafter we refer to the matched-filter estimated redshift as $z_{mf}$). Currently, $\\simeq 30$ PDCS clusters have been confirmed spectroscopically, most of them at $z < 0.5$ (Holden et al. \\cite{hol99a}, \\cite{hol99b}; Oke et al. \\cite{oke98}). Recently, Olsen et al. (\\cite{ols99a}, \\cite{ols99b}) and Scodeggio et al. (\\cite{sco99}) have presented a catalogue of 302 cluster candidates from the $I$-band images of the ESO Imaging Survey (EIS, see Renzini \\& da Costa \\cite{ren97}). Clusters are identified in two dimensions (hereafter, 2-d) using the matched filter algorithm of Postman et al. (\\cite{pos96}; see Olsen et al. \\cite{ols99a}). The estimated redshifts for EIS clusters span the range $0.2 \\leq z_{mf} \\leq 1.3$, with a median redshift $z_{mf}=0.5$. Several EIS cluster candidates have been confirmed so far, most at $z < 0.5$, either from the existence of the red sequence of cluster ellipticals/S0's in colour-magnitude diagrams (Olsen et al. \\cite{ols99b}), or from a combination of photometric and spectroscopic data (da Costa et al. \\cite{dac99}). The EIS cluster catalogue is the largest optically selected cluster sample currently available in the Southern Hemisphere to this depth. This catalogue constitutes an obvious reference for follow-up observations at the ESO VLT aimed at determining the structure and dynamics of distant clusters, as well as the spectroscopic properties of their member galaxies. Unfortunately, little is currently known on the performance of the matched filter algorithm in detecting real clusters at $z \\geq 0.5$. As we already pointed out, most confirmed PDCS and EIS clusters have redshifts $z < 0.5$. Therefore, to point blindly at EIS cluster candidates would make for an inefficient use of VLT time, because we expect several of these candidate clusters not to be real, in particular at $z_{mf} \\geq 0.5$. The aim of our investigation is twofold: we want to confirm as many EIS clusters as possible, in order to build a reliable sample of distant clusters with well determined redshift, and, at the same time, evaluate the performance of the matched filter algorithm in the detection of high-redshift clusters. In order to achieve this purpose, we use two independent methods: (1) multi-object spectroscopic observations of EIS cluster candidates in the redshift range $0.5 \\leq z_{mf} \\leq 0.7$, and (2) the detection of the colour-magnitude sequences traced by early-type galaxies through multi-colour optical and near-IR photometry of the most distant EIS cluster candidates (Scodeggio et al., in preparation). In this paper we report the first results of the spectroscopic investigations of 6 EIS clusters. We are able to confirm the existence of significant concentrations in redshift space in correspondence of four of the six EIS fields targeted. For two of these confirmed clusters, the spectroscopic mean redshift agrees with the matched-filter estimate to within $\\Delta z = \\pm 0.1$. In Sect.~\\ref{s-obs} we describe our spectroscopic observations, data reduction, and give the new galaxy redshifts. In Sect.~\\ref{s-zsys} we analyse the data, and define sets of galaxies in redshift space. We also discuss the concordance of the mean redshifts of these sets with the matched-filter estimates of the cluster mean redshifts. We then make a likelihood analysis of the reality of the galaxy sets, and flag four of them as reliable at $> 95$~\\% confidence level (Sect.~\\ref{s-likely}). Finally, we discuss our results and give our conclusions in Sect.~\\ref{s-con}. We use H$_0=$ h$_{75}$~75~km~s$^{-1}$~Mpc$^{-1}$, $\\Omega_0=0.2$ and $\\Omega_{\\Lambda}=0$ throughout this paper, unless otherwise stated. ", "conclusions": "\\label{s-con} We obtain 67 new redshifts for galaxies in six EIS candidate cluster fields. Based on these data, we establish the existence of real systems in redshift space in the direction of four of these candidate clusters. The reality of the systems is established at $> 95$~\\% confidence level, and in two cases, at $> 99$~\\%. The redshift overdensities, coupled with the 2-d overdensities detected by the use of the matched-filter algorithm, strongly supports the reality of four of the six examined EIS clusters. These 4 clusters add to the other two spectroscopically confirmed EIS clusters (da~Costa et al. \\cite{dac99}). Two of the four $z$-systems have a median redshift in good agreement with the matched-filter estimate for the EIS cluster redshift ($\\mid$ median$(z)-z_{mf} \\mid < 0.1$). The other two have significantly lower redshifts (median$(z) = 0.129$, 0.236 vs. $z_{mf}=0.5$). Taken at face value, these results suggest that, in several cases, the matched-filter algorithm over-estimates the mean cluster redshift by a large amount. However, it is quite possible that in some cases we have not detected the EIS cluster, but a foreground galaxy system projected along the same line-of-sight of the cluster. Similarly, it is difficult to conclude about the reality of the EIS clusters where we do not detect any real redshift system. In particular, we note that in the field of EIS0540-2418 we have a marginal detection (92~\\% probability) of the galaxy set 2d at median$(z)=0.698$ (see Table~\\ref{t-cs}), in fair agreement with the matched-filter estimate of the cluster redshift, $z_{mf}=0.6$. We also note that the cluster EIS0533-2353 has $z_{mf}=0.7$, larger than any of our galaxy sets. This suggests that it could have escaped detection because our observations were not deep enough. In fact, da~Costa et al. (\\cite{dac00}) suggest that all our six EIS cluster candidates could be real, based on the analysis of the colour-magnitude diagrams for galaxies in the cluster fields. We conclude that our spectroscopic confirmation rate must be considered as a lower limit. If at least one third of the EIS clusters in the redshift range sampled by our observations are real, there are more than 25 EIS clusters with $z_{mf}$ in the range 0.5--0.7. This sample is large enough for the derivation of the properties of clusters at intermediate to high redshifts. Optical selection of clusters of galaxies at high redshifts is a necessary complementary approach to X-ray selection. While X-ray selection tends to detect only rich Abell-like clusters, optically selected cluster samples contain a large number of poor clusters. In fact, the space density of PDCS clusters is five times higher than that of rich Abell clusters (Holden et al. \\cite{hol99a}), and very few PDCS clusters are X-ray bright (Holden et al. \\cite{hol97}). Consistently, the velocity dispersions of our two systems with $\\geq 5$ galaxy redshifts (system 4c, at $\\overline{z}=0.236$ and system 6c, at $\\overline{z}=0.445$, see Table~\\ref{t-cs}) are $\\sim 600$~km~s$^{-1}$, typical of low-richness clusters ($R \\leq 1$, see Girardi et al. \\cite{gir93}). With the current and near-future ground-based facilities for wide-field optical and near-infrared imaging, we can expect a rapid increase in the samples of optically selected clusters. Currently, our spectroscopic sample only comprises $\\sim 10$~\\% of all the clusters in the two patches C and D, and in the (estimated) redshift range 0.5--0.7. We plan to extend our sample in forthcoming observing runs. Confirmed EIS clusters at high redshift will be the natural targets of VLT observations aimed at determining their dynamical properties." }, "0005/astro-ph0005470_arXiv.txt": { "abstract": "We show with analytic models that the assumption of uncorrelated intrinsic ellipticities of target sources that is usually made in searches for weak gravitational lensing due to large-scale mass inhomogeneities (``field lensing'') is unwarranted. If the orientation of the galaxy image is determined either by the angular momentum or the shape of the halo in which it forms, then the image should be aligned preferentially with the component of the tidal gravitational field perpendicular to the line of sight. Long-range correlations in the tidal field will thus lead to long-range ellipticity-ellipticity correlations that mimic the shear correlations due to weak gravitational lensing. We calculate the ellipticity-ellipticity correlation expected if halo shapes determine the observed galaxy shape, and we discuss uncertainties (which are still considerable) in the predicted amplitude of this correlation. The ellipticity-ellipticity correlation induced by angular momenta should be smaller. We consider several methods for discriminating between the weak-lensing (extrinsic) and intrinsic correlations, including the use of redshift information. An ellipticity--tidal-field correlation also implies the existence of an alignment of images of galaxies near clusters. Although the intrinsic alignment may complicate the interpretation of field-lensing results, it is inherently interesting as it may shed light on galaxy formation as well as on structure formation. ", "introduction": "Searches for weak gravitational lensing due to large-scale mass inhomogeneities are coming of age. Ellipticities of high-redshift sources are taken to be indicators of the shear field induced by weak gravitational lensing by mass inhomogeneities along the line of sight, and shear-shear correlations can be used as a probe of the lensing-mass distribution \\cite{gun67,mir91,bla91,kai92,bar92,bar99}. The advantage of weak lensing is that it determines the power spectrum (as well as higher-order statistics; e.g., Bernardeau, van Waerbeke \\& Mellier 1997; Munshi \\& Jain 2000; Cooray \\& Hu 2000) for the {\\it mass} rather than the light. In just the past few months, four groups have reported the first detections of such ``field lensing'' \\cite{BacRefEll00,KaiWilLup00,Witetal00,Waeetal00}. Noise for the weak-lensing signal is provided by the intrinsic ellipticities of the sources. With a sufficiently large sample of sources, the random orientation of these sources can be overcome. One thus looks for an ellipticity correlation in excess of the Poisson noise provided by randomly oriented intrinsic ellipticities. The analysis always assumes that the intrinsic orientations of the sources are entirely random and isotropically distributed. The point of this paper will be to demonstrate that this should not be the case. To do so, we consider two {\\it ansatzen} for the origin of the ellipticity of the high-redshift sources. We first suppose that the ellipticity of the galaxy image may be determined primarily by the shape of the halo in which it forms; this might be expected if the sources are isolated ellipticals. In this case, a modification of the spherical-top-hat model for gravitational collapse in a tidal field suggests a preferential elongation of the galaxies along the direction of the tidal field. We show that in this case, long-range correlations in the ellipticities of widely-separated sources are proportional to long-range correlations in the tidal field, and thus to correlations in the mass distribution. We then consider what happens if the orientation of the image is determined by the angular momentum of the halo in which it forms; this should be a good description if the sources are disk galaxies. The simplest hypothesis---adopted in nearly all disk-formation models (e.g., Dalcanton, Spergel \\& Summers 1997; Mo, Mao \\& White 1998; Buchalter et al. 2000)---is that the plane of the disk is perpendicular to the angular-momentum vector of the galactic halo in which the disk forms. According to linear perturbation theory, a galactic halo acquires its angular momentum via torquing of the aspherical protogalaxy in the tidal gravitational field that arises from the large-scale mass distribution \\cite{Hoy49,Pee69,Dor70,Whi84,HeaPea88,CatThe96}. Averaging over all possible orientations of the protogalaxy, the disk orientations are correlated with the tidal field. In this case, long-range correlations in the ellipticities are expected to be smaller, as they will be proportional at lowest order to the square of the correlations in the tidal-field and/or mass distribution. In the next Section, we review briefly the statistics used to describe the weak-lensing signal. In Section 3, we explan how ellipticals should be preferentially elongated along the direction of the tidal field, and we present the calculation of the shear power spectrum for this case. Section 4 presents numerical results. In Section 5, we show how tidal torquing can align galaxies preferentially along the tidal gravitational field, and we explain why this should lead to smaller ellipticity correlations that are of higher order in the mass correlation. In Section 6 we put forth some ideas for disentangling the intrinsic and weak-lensing signals, including the use of redshift information, and we predict a corresponding alignment in the images of galaxies near clusters. We close with some concluding remarks in Section 7. During preparation of this paper, we learned of related work (using numerical simulations) by Heavens, Refregier \\& Heymans (2000) and Croft \\& Metzler (2000). Our analytic approach should complement their numerical work and perhaps help shed some light on the origin of their observed correlations. The analytic calculation should also be useful in determining the correlation at large angular separations, where it becomes increasingly difficult to measure in simulations. Our analytic approach also suggests some possible intrinsic/weak-lensing discriminators. ", "conclusions": "We have shown that the shapes and/or angular momenta of galactic halos should depend to some extent on the tidal gravitational field in which they are produced. Long-range correlations in the gravitational field should thus lead to long-range correlations in the shear inferred from images of distant galaxies. Although uncertainties in the relation between the luminous-galaxy shape and the halo shape prohibit us from carrying out a ``first principles'' calculation of the correlation, we can estimate the magnitude of a the correlation that arises if ellipticities are determined by halo shapes by calibrating to the observed distribution of ellipticities. The amplitude of the intrinsic power spectrum is increased (decreased) with a larger (smaller) smoothing length $R$. Realistically, there will be factors in addition to those that we have considered that contribute to the observed orientation, and these could decrease the correlation. One example is the halo spin. As another example, major mergers could affect the halo shape as well as the orientation of the disk relative to that of the halo. All of these effects will tend to diminish the correlations predicted by our model. However, results from numerical simulations (Heavens, Refregier \\& Heymans 2000; Croft \\& Metzler 2000) seem to indicate that the correlations in the halos have not been much diluted by these effects. Although these simulations quantify the correlations of the parent halos, there is still a considerable amount of physics relating the halo shape to the shape of the luminous galaxy that cannot yet be described properly with simulations. Heuristically, these effects should tend to diminish the intrinsic correlations even further. If future theoretical work determines that the degradation is considerable, then the effects we are discussing will be unimportant for interpretation of recent field-lensing detections. However, even if the correlation is small, it should not be zero---we have indeed identified realistic physical effects that should play at least some role in aligning galaxy images. Thus, the physical effects we have discussed here will be important for interpretation of future more precise weak-lensing maps, as well as for understanding the implications of measurements of higher-order weak-lensing statistics. At first, this intrinsic correlation may be seen as a nuisance for field-lensing searches. However, the intrinsic correlation arises from the same long-range correlations in the density field that give rise to the weak-lensing correlation. Moreover, the galaxy-formation physics that produces spins and shapes of galaxies is itself inherently interesting. Thus, measurement of these intrinsic correlations would be of fundamental significance for structure formation and galaxy formation." }, "0005/astro-ph0005193_arXiv.txt": { "abstract": "The internal shock model for gamma-ray bursts involves shocks taking place in a relativistic wind with a very inhomogeneous initial distribution of the Lorentz factor. We have developed a 1D lagrangian hydrocode to follow the evolution of such a wind and the results we have obtained are compared to those of a simpler model presented in a recent paper (Daigne \\& Mochkovitch \\cite{Daigne2}) where all pressure waves are suppressed in the wind so that shells with different velocities only interact by direct collisions. The detailed hydrodynamical calculation essentially confirms the conclusion of the simple model: the main temporal and spectral properties of gamma-ray bursts can be reproduced by internal shocks in a relativistic wind.\\\\ ", "introduction": "Since the discovery of the optical counterpart of GRB 970228 (van Paradijs et al. \\cite{vanParadijs1}) the accurate localizations provided by the \\textit{Beppo--SAX} satellite have led to the detection of the optical afterglow for more than ten gamma--ray bursts (hereafter GRBs). The most spectacular result of these observations is to have provided a direct proof of the cosmological origin of GRBs. The detection of absorption lines at $z=0.835$ in the spectrum of GRB 970508 (Metzger et al. \\cite{Metzger1}) followed by other redshift determinations (between $z=0.43$ and $z=3.41$) confirmed the indications which were already available from the \\textit{BATSE} data showing a GRB distribution perfectly isotropic but non homogeneous in distance (Fishman and Meegan \\cite{Fishman1} and references therein). The energy release of GRBs with known redshifts extends from $E_{\\gamma}= 2\\ 10^{51} \\frac{\\Omega}{4 \\pi}$ to $E_{\\gamma} = 2\\ 10^{54} \\frac{\\Omega}{4 \\pi}\\ \\rm erg$. The solid angle $\\Omega$ in which the emission is beamed is quite uncertain. A small $\\Omega$ should reveal itself by a break after a few days in the afterglow light curve. A break is indeed observed in a few cases such as GRB 990510 (Harrison et al. \\cite{Harrison1}) where $\\frac{\\Omega}{4 \\pi}$ could be as small as 0.01. However, most afterglows do not show any break which means that $\\Omega$ is usually not very small ($\\frac{\\Omega}{4 \\pi}\\sim 0.1$\\,?).\\\\ The source of cosmic GRBs must therefore be able to release a huge energy in a very short time. Possible candidates include the coalescence of two neutron stars (Eichler et al. \\cite{Eichler1}; Paczy\\'nski \\cite{Paczynski1}), the disruption of the neutron star in a neutron star -- black hole binary (Narayan et al. \\cite{Narayan1}; Mochkovitch et al. \\cite{Mochkovitch1}) or the collapse of a massive star (Woosley \\cite{Woosley1}, Paczy\\'nski \\cite{Paczynski2}). In all these cases the resulting configuration is expected to be a stellar mass black hole surrounded by a thick disc. Since the power emitted by GRBs is orders of magnitude larger than the Eddington limit it cannot be radiated by a static photosphere. The released energy generates a fireball which then leads to the formation of a wind. Moreover, this wind has to become highly relativistic in order to avoid the compactness problem and produce gamma--rays (Baring \\cite{Baring1}; Sari \\& Piran \\cite{Sari1}). Values of the Lorentz factor as high as $\\Gamma=100$--$1000$ are required, which limits the allowed amount of baryonic pollution to a remarkably low level. Only a few mechanisms have been proposed to produce a wind under such severe constraints : ({\\it i}) magnetically driven outflow originating from the disc or powered by the Blandford--Znajek (\\cite{Blandford1}) process (Thomson \\cite{Thomson1}; M\\'esz\\'aros \\& Rees \\cite{Meszaros2}; Daigne \\& Mochkovitch \\cite{Daigne3}; Lee et al. \\cite{Lee1}) ; ({\\it ii}) reconnection of magnetic field lines in the disc corona (Narayan et al. \\cite{Narayan1}) ; ({\\it iii}) neutrino--antineutrino annihilation in a funnel along the rotation axis of the system (M\\'esz\\'aros \\& Rees \\cite{Meszaros3}; Mochkovitch et al. \\cite{Mochkovitch1}, \\cite{Mochkovitch2}). Mechanisms ({\\it i}) and ({\\it ii}) require that the magnetic field in the disc reaches very high values $B \\gsim 10^{15}\\ \\rm G$. Our preliminary study (Daigne \\& Mochkovitch \\cite{Daigne3}) of the wind emitted from the disc shows that it can avoid baryonic pollution only if some very severe constraints on the dissipation in the disc and the field geometry are satisfied. Some recent works (Ruffert et al. \\cite{Ruffert1}) have also shown that mechanism ({\\it iii}) is probably not efficient enough to power a gamma--ray burst, except may be for the shortest events. When the wind has reached its terminal Lorentz factor, the energy is mainly stored in kinetic form and has to be converted back into gamma--rays. Two main ideas have been proposed to realize this conversion. The first one is the so-called external shock model (Rees \\& M\\'esz\\'aros \\cite{Rees1}; M\\'esz\\'aros \\& Rees \\cite{Meszaros4}). The wind is decelerated by the external medium, leading to a shock. Gamma--rays are emitted by the accelerated electrons in the shocked material through the synchrotron and/or inverse Compton mechanisms. This model has been studied in details (Fenimore et al. \\cite{Fenimore1}; Panaitescu et al. \\cite{Panaitescu1}; Panaitescu \\& M\\'esz\\'aros \\cite{Panaitescu2}) and seems unable to reproduce some important features of GRBs such as their strong temporal variability (see however Dermer \\& Mitman \\cite{Dermer1}). Conversely, the external shock model reproduces very well the delayed emission at lower energy from the afterglows (M\\'esz\\'aros \\& Rees \\cite{Meszaros5}; Wijers et al. \\cite{Wijers1}). The second proposal is the internal shock model (Rees \\& M\\'esz\\'aros \\cite{Rees2}) where the wind is supposed to be formed initially with a very inhomogeneous distribution of the Lorentz factor. Rapid parts of the wind then catch up with slower ones leading to internal shocks where gamma--rays are again produced by synchrotron or inverse Compton radiation. We have started a study of this model in a previous paper (Daigne \\& Mochkovitch \\cite{Daigne2}, hereafter DM98) where the wind was simply made of a collection of ``solid'' shells interacting by direct collisions only (all pressure waves were suppressed). The very encouraging results we obtained had to be confirmed by a more detailed study. We have therefore developed a relativistic hydrocode to follow the evolution of the wind. We present the code and the main results in this paper. We write in Sect. 2 the lagrangian equations of hydrodynamics in special relativity. In Sect. 3 we describe the numerical method we use to solve them and we present the tests we performed to validate the method. We display our results in Sect. 4 and Sect. 5 is the conclusion. ", "conclusions": "This paper is the continuation of our study of the internal shock model started in DM98. We developed a 1D lagrangian relativistic hydrocode (in spherical symmetry) to validate our previous simpler approach where all pressure waves were neglected in the wind. Our code is an extension of the classical \\textit{PPM} method of Colella and Woodwards (\\cite{Colella1}) in the spirit of the work by Mart\\'{\\i} \\& M\\\"uller (\\cite{Marti2}) for the eulerian case in planar symmetry. A detailed comparison has been made between the hydrocode and the simple model in the case of a single pulse burst. It appears that the dynamical evolution of the wind is well reproduced by the simple model, which is not too surprising because the wind energy is largely dominated by the kinetic part so that the effect of pressure waves is small. Only one physical quantity -- the density of the shocked material -- is strongly underestimated in the simple model. In order to make valuable comparisons between the two calculations we have therefore adjusted the equipartition parameters so that the mean value of the synchrotron energy is the same in the two cases. The synthetic bursts which are then obtained are very similar which proves that our first approach was essentially correct and confirm our previous results. A similar conclusion was reached by Panaitescu and M\\'esz\\'aros (\\cite{Panaitescu3}) who performed a comparable study. The internal shock model can easily explain the great temporal variability observed in GRBs. The main characteristic features of individual pulses are well reproduced: (1) pulses have typical asymmetric ``FRED'' profiles; (2) the pulse width decreases with energy following a power--law $W(E)\\propto E^{-p}$ with $p \\sim 0.4$; (3) short pulses show a tendency to become more symmetric. Our model still gives very short pulses which decay faster than they rise but the hydrodynamical simulation improves the situation compared to the simple model. Spectral properties of GRBs are also well reproduced. We obtain synthetic spectra which can be nicely fitted with Band's function with parameters comparable to those observed in real GRBs. The spectral hardness and the count rate are correlated during the evolution of a burst with the hardness usually preceeding the count rate. As also pointed in DM98, the duration--hardness relation is a natural consequence of the internal shock model. These results are very encouraging and the main difficulty which remains is the low efficiency (about $10 \\%$) of the internal shock model. As long as the energetics of GRBs and the mechanism initially operating in the central engine are not precisely identified, we cannot say if this is a critical problem or not. We still believe that the internal shock model is at present the most convicing candidate to explain the gamma--ray emission from GRBs. Next steps in this work will address the following questions. We first want to extend our hydrodynamical code to a non--adiabatic version in order to include the radiative losses in the dynamical calculation. We have already developed an ``isothermal Rieman Solver'' for that purpose (Daigne \\& Mochkovitch \\cite{Daigne1}). We would also like to study the effects of the external medium, with a special attention to the reverse shock which propagates into the wind and possibly interacts with the internal shocks. Preliminary results with the simple method using ``solid layers'' have already been obtained (Daigne \\& Mochkovitch \\cite{Daigne4}) but they have to be confirmed by a hydrodynamical calculation. Finally, we would like to investigate the details of the emission process during internal shocks to solve some of the problems encountered by the synchrotron model." }, "0005/astro-ph0005120_arXiv.txt": { "abstract": "We discuss a new method for inferring the stellar mass of a distant galaxy of known redshift based on the combination of a near-infrared luminosity and multi-band optical photometry. The typical uncertainty for field galaxies with $I<22$ in the redshift range $0L_p$ implied in Beckenstein's framework. Thus, the present available data set, considered within Bekenstein's framework, is capable to rule out $\\alpha$ variability, while the original paper had to recourse to E\\\"otv\\\"os-like experiments to achieve the same result. \\citet{LyS} have also analyzed $\\alpha$ variation in the context of Bekenstein's theory. Our results are in agreement with their analysis, even though they didn't allow both free parameters of the theory: $\\frac{l}{L_p}$ and $\\gamma$ to vary independently. However, it should be noted that Beckenstein's framework is very similar to the dilatonic sector of string theory, and it has been pointed out that in the context of string theories \\citep{strings1,strings2} there is no need for an universal relation between the Planck and the string scale. Finally, our results are consistent with no time variation of fundamental constants over cosmological time in agreement most of the experimental results. Indeed, excluding the Webb et al. data points from our fits does not change significantly the values of the adjusted constants. Thus, this rather large class of theories cannot explain this discrepant result." }, "0005/astro-ph0005066_arXiv.txt": { "abstract": "Mkn~421 was repeatedly observed with \\textit{Beppo}SAX in 1997--1998. This is the first of two papers where we present the results of a thorough temporal and spectral analysis of all the data available to us, focusing in particular on the flare of April 1998, which was simultaneously observed also at TeV energies. Here we focus on the time analysis, while the spectral analysis and physical interpretation are presented in the companion paper. The detailed study of the flare in different energy bands reveals very important new results: i) hard photons lag the soft ones by 2--3 ks -- a behavior opposite to what is normally found in high energy peak BL Lacs X--ray spectra; ii) the flare light curve is symmetric in the softest X--ray band, while it becomes increasingly asymmetric at higher energies, with the decay being progressively slower than the rise; iii) the flux decay of the flare can be intrinsically achromatic if a stationary underlying emission component is present. The temporal and spectral information obtained challenge the simplest models currently adopted for the (synchrotron) emission and most importantly provide clues on the particle acceleration process. ", "introduction": "\\label{sec:introduction} Blazars are radio--loud AGNs characterized by strong variability, large and variable polarization, and high luminosity. Radio spectra smoothly join the infrared-optical-UV ones. These properties are successfully interpreted in terms of synchrotron radiation produced in relativistic jets and beamed into our direction due to plasma moving relativistically close to the line of sight (e.g. Urry \\& Padovani \\citealp{up95}). Many blazars are also strong and variable sources of GeV $\\gamma$--rays, and in a few objects the spectrum extends up to TeV energies. The hard X-- to $\\gamma$--ray radiation forms a separate spectral component, with the luminosity peak located in the MeV--TeV range. The emission up to X--rays is thought to be due to synchrotron radiation from high energy electrons in the jet, while it is likely that $\\gamma$-rays derive from the same electrons via inverse Compton (IC) scattering of soft (IR--UV) photons --synchrotron or ambient soft photons (e.g. Sikora, Begelman \\& Rees~\\citealp{sbr94}, Ghisellini et al. \\citealp{gg_sed98}). The contributions of these two mechanisms characterize the average blazar spectral energy distribution (SED), which typically shows two broad peaks in a $\\nu F_\\nu$ representation (e.g. von Montigny et al. \\citealp{vmon95}; Sambruna, Maraschi \\& Urry \\citealp{smu96}; Fossati et al. \\citealp{fg_sed98}): the energies at which the peaks occur and their relative intensity provide a powerful diagnostic tool to investigate the properties of the emitting plasma, such as electron energies and magnetic field (e.g. Ghisellini et al. \\citealp{gg_sed98}). Moreover variability studies, both of single band and of simultaneous multifrequencies data, constitute the most effective means to constrain the emission mechanisms at work in these sources as well as the geometry and modality of the energy dissipation. The quality and amount of X--ray data on the brightest sources start to allow us to perform a thorough temporal analysis as function of energy and determine the spectral evolution with good temporal resolution. In X--ray bright BL Lacs (HBL, from High-energy-peak-BL Lacs, Padovani \\& Giommi \\citealp{pg95}) the synchrotron maximum (usually) occurs in the soft-X--ray band, and the inverse Compton emission extends in some cases to the TeV band where -- thanks to ground based Cherenkov telescopes -- four sources have been detected up to now: Mkn~421 (Punch et al. \\citealp{punch92}), Mkn~501 (Quinn et al. \\citealp{quinn96}), 1ES~2344+514 (Catanese et al. \\citealp{catanese_2344_98}), and PKS~2155--304 (Chadwick et al. \\citealp{chadwick98}). If the interpretation of the SED properties in terms of synchrotron and IC radiation is correct, a correlation between the X--ray and TeV emission is expected. \\smallskip Mkn~421 ($z$ = 0.031) is the brightest BL Lac object at X--ray and UV wavelengths and the first extragalactic source discovered at TeV energies, where dramatic variability has been observed with doubling times as short as 15 minutes (Gaidos et al. \\citealp{gaidos96}). As such it was repeatedly observed with X--ray satellites, including \\textit{Beppo}SAX. Remarkable X--ray variability correlated with strong activity at TeV energies has been found on different occasions (Macomb et al.~\\citealp{macomb95,macomb96}, Takahashi et al.~\\citealp{takahashi96}, Fossati et al.~\\citealp{fg_lincei98}, Maraschi et al.~\\citealp{maraschi_letter}). In particular, the 1998 \\textit{Beppo}SAX data presented here were simultaneous with a large TeV flare detected by the Whipple Observatory (Maraschi et al.~\\citealp{maraschi_letter}). This paper is the first of two, which present the results of a uniform, detailed spectral and temporal analysis of \\textit{Beppo}SAX observations of Mkn~421 performed during 1997 and 1998. Here we focus on the data reduction and the timing analysis, and also discuss the results on the spectral variability derived from the different properties of the flux variations in different energy bands. The paper is organized as follows. We briefly summarize the characteristics of \\textit{Beppo}SAX (\\S\\ref{sec:bepposax_overview}), and introduce the observations studied (\\S\\ref{sec:observations}). We then address the temporal analysis of the variability, considering several energy bands and comparing the light curve features by means of a few simple estimators for the 1997 and 1998 observations (\\S\\ref{sec:light_curves}). The remarkable flare observed in 1998 is the object of a further deeper analysis, reported in Section \\S\\ref{sec:timing98}, focused on timescales and time lags. Section~\\ref{sec:disc:variability} contains a summary of the results of the temporal analysis, preparing the ground for the comprehensive discussion presented in Paper~II (Fossati et al. \\citealp{fossati_II}). There they are considered together with the results of the spectral analysis and thus used to constrain a scenario able to interpret the complex spectral and temporal findings. ", "conclusions": "" }, "0005/astro-ph0005585_arXiv.txt": { "abstract": "We present extensive metal--line absorption properties for 45 absorption systems that were selected by their {\\MgII} absorption at redshifts between 0.4 and 1.4. For each system the properties of several chemical species are determined, including a wide range of ionization conditions. In the optical, the absorption systems have been observed at $\\sim 6$~{\\kms} resolution with HIRES/Keck, which covered {\\MgII}, several {\\FeII} transitions, {\\MgI}, and in some cases (depending upon redshift), {\\CaII}, {\\TiII}, {\\MnII}, and {\\AlIII}. Ultraviolet, lower resolution ($\\sim 230$~{\\kms}) Faint Object Spectrograph data (1600~{\\AA}--3275~{\\AA}) were obtained from the {\\it Hubble Space Telescope\\/} archive. These spectra covered {\\AlII}, {\\AlIII}, {\\SiII}, {\\SiIII}, {\\SiIV}, {\\CII}, {\\CIII}, {\\CIV}, {\\NV}, {\\OVI}, and several Lyman series transitions, with coverage dependent upon the absorption system redshift. From these data, we infer that {\\MgII} absorbing galaxies at intermediate redshifts have multiphase gaseous structures. ", "introduction": "\\label{sec:intro} During the last decade a great deal of progress has been made toward understanding the physical properties of intervening metal--line absorption systems measured in the spectra of high redshift quasars. This is particularly true at intermediate redshifts, $0.5\\leq z \\leq 1.5$, for absorbers selected by the presence of the resonant {\\MgIIdblt} doublet (e.g.\\ \\cite{ltw87}; \\cite{tytler87}; \\cite{sbs88}; \\cite{pb90}, \\cite{ss92}). One of the most notable achievements was the demonstration that {\\MgII} absorbers with $W_{r}(2796)\\geq0.3$~{\\AA} are almost always associated with galaxies (\\cite{bb91}; \\cite{steidel95}). Those works substantiated the $\\sim 30$--year standing hypothesis by Bahcall \\& Spitzer (1969\\nocite{bs69}) that metal--line absorption in quasar spectra arises in extended gaseous envelopes surrounding intervening galaxies. The general picture today is that a wide variety of morphological types (from ellipticals to irregulars) have gaseous ``halos'' that extend to roughly 40~kpc, with the most common being Sbc--Scd types (Steidel, Dickinson, \\& Persson 1994\\nocite{sdp94}; \\cite{guillemin}). The line--of--sight gas kinematics of the absorbers is consistent with that expected for material bound in galactic potential wells (\\cite{pb90}; Churchill, Steidel, \\& Vogt 1996\\nocite{csv96}; \\cite{kinematicpaper}). This picture, however, is not without its counter examples or ambiguities. In some cases there is evidence that compact star forming objects spread out over $\\sim 200$ kpc are seen at the {\\MgII} absorption redshift and there is no directly associable bright galaxy (\\cite{yanny92}; \\cite{yannyyork92}). It is also not yet established whether the more numerous ``weak'' {\\MgII} absorbers, those with $W_{r}(2796)<0.3$~{\\AA}, are related to galaxies similar in type to those associated with ``strong'' {\\MgII} absorption. There is mounting evidence that a fair number of the weak systems do not arise within $\\sim 40$~kpc of normal, bright galaxies (\\cite{cwc-lb98}; \\cite{weak}). What are the typical low to high ionization absorption conditions in intermediate redshift {\\MgII} absorption--selected galaxies? Do the majority of {\\MgII} systems have an associated high ionization phase as seen in {\\SiIV}, {\\CIV}, {\\NV}, and {\\OVI} absorption? Are there any trends between the high ionization and low ionization absorption strengths? Are there other relationships (or lack of relationships!) that provide clues to the physical nature of galactic gas at intermediate redshifts? Motivated by these and similar questions, we have undertaken a program to measure the absorption properties of a wide variety of chemical and ionization species associated with {\\MgII} absorbers. Unique to our study is that the {\\MgII} systems have been observed at high resolution ($\\sim 6$~{\\kms}) with HIRES/Keck~I (\\cite{thesis}). These spectra also provide a population of weak systems, which are significantly more numerous in their redshift path density (\\cite{weak}). The HIRES spectra cover {\\MgII}, several {\\FeII} transitions, {\\MgI}, and depending upon redshift coverage, {\\CaII}, {\\TiII}, {\\MnII}, and {\\AlIII}. The remaining absorption properties, including neutral hydrogen and higher ionization species, have been measured in lower resolution ($\\sim 230$~{\\kms}) spectra obtained from the {\\it Hubble Space Telescope\\/} archive of the Faint Object Spectrograph. In this paper, we present the measurements of the absorption lines found in the FOS spectra. Additional analysis focused on the above motivational questions is presented in a parallel companion paper (\\cite{paper2}, hereafter Paper II). In \\S~\\ref{sec:sample}, we outline our sample selection. Details of the data analysis are presented in \\S~\\ref{sec:data}. In \\S~\\ref{sec:systems}, we provide a brief description of each system. The general absorption properties are presented in \\S~\\ref{sec:discussion} and a brief synopsis is given in \\S~\\ref{sec:conclusions}. ", "conclusions": "\\label{sec:conclusions} We have measured the absorption properties of 45 intermediate redshift {\\MgII} absorbers in FOS spectra from the {\\it HST\\/} archive and from the database of the {\\it HST\\/} QSO Absorption Line Key Project. The sample was selected from the 51 {\\MgII} systems observed with HIRES/Keck for which FOS spectra of the same quasars existed. The {\\MgII} profiles, and other transitions observed in the optical, have been resolved at $\\sim 6$~{\\kms} resolution. The UV FOS spectra have resolution $\\sim 230$~{\\kms}. In this paper, we presented the data, the data analysis, and a brief description of the properties of each system. We present additional analysis of the data in a parallel companion paper (Paper II\\nocite{paper2}). We have found evidence for a high ionization gaseous phase in intermediate redshift {\\MgII} absorbing galaxies. Mostly, the high ionization species detected are {\\CIV} and {\\SiIV}, which are commonly seen in absorption in the Galaxy (e.g.\\ \\cite{savagearaa}; \\cite{savage97}). These data lead us to suggest that these galaxies have multiphase interstellar media and halos similar to those observed locally (\\cite{dahlem}; also see \\cite{letter}). What is the kinematic spread of the {\\CIV} and what is its line of sight velocity structure? How is this structure related to that seen in {\\MgII}? High resolution spectroscopic observation are sorely needed for sorting out the physical nature of this high ionization material and its relation to the kinematically complex low ionization gas. Only if the high ionization profiles are resolved at resolutions comparable to the HIRES/Keck data, can the {\\it relative\\/} kinematics of the low and high ionization gas can be quantified." }, "0005/astro-ph0005250_arXiv.txt": { "abstract": "The development of germanium Compton telescopes for nuclear $\\gamma$-ray astrophysics ($\\sim$0.2-20~MeV) requires new event reconstruction techniques to accurately determine the initial direction and energy of photon events, as well as to consistently reject background events. This paper describes techniques for event reconstruction, accounting for realistic instrument/detector performance and uncertainties. An especially important technique is Compton Kinematic Discrimination, which allows proper interaction ordering and background rejection with high probabilities. The use of these techniques are crucial for the realistic evaluation of the performance and sensitivity of any germanium Compton telescope configuration. ", "introduction": "Looking beyond the INTErnational Gamma-Ray Astrophysics Laboratory (INTEGRAL), the next generation soft $\\gamma$-ray ($\\sim$0.2-20~MeV) observatory will require high angular and spectral resolution imaging to significantly improve sensitivity to astrophysical sources of nuclear line emission. Building upon the success of COMPTEL/CGRO (\\cite{schon93}), and the high spectral resolution of the upcoming SPI/INTEGRAL (\\cite{vedre98,licht96}), a number of researchers (\\cite{johns96, jean96, boggs98}) have discussed the merits of a high spectral/angular resolution germanium Compton telescope (GCT); the ability to achieve high sensitivity to point sources while maintaining a large field-of-view make a high resolution Compton telescope an attractive option for the next soft $\\gamma$-ray observatory. The development of Compton telescopes began in the 1970's, with work done at the Max Planck Institut (\\cite{schon73}), University of California, Riverside (\\cite{herzo75}), and the University of New Hampshire (\\cite{lockw79}), culminating in the design and flight of COMPTEL/CGRO. These historical Compton telescopes consist of two scintillation detector planes -- a low atomic number `converter' and a high atomic number `absorber.' The model interaction of a Compton telescope is a single Compton scatter in the converter plane, followed by photoelectric absorption of the scattered photon in the absorber. By measuring the position and energy of the interactions, the event can be \\textit{reconstructed} to determine the initial photon direction to within an annulus on the sky. A handful of groups are actively developing imaging germanium detectors (GeDs) partly in anticipation of a GCT (\\cite{luke94, kroeg96}). The goal of these researchers is to develop large area detectors with (sub)millimeter spatial resolution, while maintaining the high spectral resolution ($E/\\delta E \\sim 500$ at 1~MeV) characteristic of GeDs. The use of high spectral/spatial resolution GeDs as converter and absorber planes would significantly improve the performance of a Compton telescope, but will add a number of complications to the event reconstruction. Most significantly, with the moderate atomic number $(Z = 32)$ of germanium, photons will predominantly undergo multiple Compton scatters before being photoabsorbed in the instrument. Furthermore, with interaction timing capabilities of $\\sim$10~ns, the interaction order will not be determined unambiguously by timing alone. Compton Kinematic Discrimination (CKD) is proposed here to overcome these complications, an extension of a method first discussed in context of liquid xenon time projection chambers (\\cite{april93}). The ability of this technique to allow proper event reconstruction is investigated in detail. Due to their relatively low efficiency (typically $\\sim 1\\%$), Compton telescopes rely on efficient background suppression to maintain their sensitivity. In addition to interaction ordering, techniques are presented using CKD, in combination with other tests and restrictions, to suppress the dominant background components. The goal of this work is to outline a complete set of event reconstruction techniques for GCTs, taking into account realistic detector/instrument performance and uncertainties. Examples of the techniques are presented for a GCT configuration outlined in Appendix~A; however, full analysis of this configuration will be presented in a second paper dedicated to the optimization and performance of several GCT configurations. The full analysis of a GCT configuration is complicated, requiring a detailed study of the tradeoffs between efficiency, angular and spectral resolution; therefore, this paper focuses only on the detailed discussion of the event reconstruction techniques which will be used in future work dedicated to analyzing GCT performance. \\begin{figure} \\resizebox{\\hsize}{!}{\\includegraphics{figure1.eps}} \\caption{Example Compton telescope. If a photon undergoes one or more Compton scatters in the instrument and then is photoelectrically absorbed, then by using the positions $(\\vec{r}_{1}, \\ldots,\\vec{r}_{N})$ and energy deposits $(E_{1}, \\ldots, E_{N})$ of the interactions, the initial direction of the photon can be determined by the Compton scatter formula to within an annulus on the sky, $\\phi_{1}$. The width of this annulus is determined by the uncertainties in both the interaction locations and energy deposits.} \\label{Figure 1} \\end{figure} ", "conclusions": "Event reconstruction in future high resolution Compton telescopes will present a number of complications compared to historical configurations. The initial complication of multiple-scattering of photons in GCTs, however, turns out to be an advantage: the application of CKD to 3+ site events, combined with the high spectral and spatial resolution of GeDs, allows extremely efficient background suppression, crucial for Compton telescope performance. This paper has outlined a set of tests and restrictions, accounting for realistic instrument/detector performance, to reconstruct photopeak events in GCTs while rejecting a large fraction of the background events. Table 2 presents the fraction of events, photon and background, that remain after each rejection technique is subsequently applied. (The numbers in Table 2 assume only $W_{1}$ is tested for $\\beta^{-}$ decay energies.) Development of these event reconstruction techniques allows realistic evaluation of the performace and sensitivity of GCT designs. Our next goal is to simulate the efficiency, resolution, background and sensitivity of several Compton telescope configurations, utilizing the event reconstruction techniques developed here to realistically determine the performace of these instruments. CKD rejection has been shown to be the most efficient background rejection technique; however, the addition of effective TOF, backscatter, nonlocalized $\\beta^{-}$ decay, and positron signature tests dramatically improve background rejection capabilities. We anticipate that use of these techniques will achieve overall sensitivity improvements in GCTs by factors of $\\sim 5-10$." }, "0005/astro-ph0005517_arXiv.txt": { "abstract": "We report millimetre-wave continuum observations of the X-ray binaries Cygnus~X-3, SS~433, LSI+61$^{\\circ}$303, Cygnus~X-1 and GRS~1915+105. The observations were carried out with the IRAM 30 m-antenna at 250 GHz (1.25 mm) from 1998 March 14 to March 20. These millimetre measurements are complemented with centimetre observations from the Ryle Telescope, at 15 GHz (2.0 cm), and from the Green Bank Interferometer at 2.25 and 8.3 GHz (13 and 3.6 cm). Both Cygnus X-3 and SS 433 underwent moderate flaring events during our observations, whose main spectral evolution properties are described and interpreted. A significant spectral steepening was observed in both sources during the flare decay, that is likely to be caused by adiabatic expansion, inverse Compton and synchrotron losses. Finally, we also report 250 GHz upper limits for three additional undetected X-ray binary stars: \\object{LSI+65$^{\\circ}$010}, \\object{LSI+61$^{\\circ}$235} and \\object{X Per}. ", "introduction": "\\label{intro} An interesting group of X-ray binary systems in our Galaxy are known to be powerful and efficient sources of radio waves. The number of radio emitting X-ray binaries (REXRBs) detected so far is about 10 $\\%$ of the total $\\sim$ 200 systems catalogued (Hjellming \\& Han 1995). Although far from representing a numerous population, their remarkable properties and scaled down similarity with extragalactic AGNs and quasars makes them to deserve a careful study based on multi-wavelength monitoring programs. Radio emission in REXRBs is normally highly variable and of non-thermal synchrotron origin. Radio outbursts with different amplitude are frequently detected and interpreted as synchrotron radiation due to the ejection and expansion of ionized plasma clouds (plasmons), usually following a super Eddington accretion event. Recent multi-wavelength monitoring of radio outbursts from the microquasar system GRS 1915+105 (Fender et al. 1997a; Mirabel et al. 1998) have revealed that the flaring synchrotron emission extends well beyond the centimetric domain, reaching up to infrared wavelengths. The energetic implications of this fact are considerable (Mirabel et al. 1998), and it would be important to investigate if similar behaviour is observed in other REXRBs. In an attempt to better assess this issue, we undertook a daily monitoring campaign in the mm domain for some well known objects in the REXRB class. Our main goal here was to study the variability and spectral index properties of selected sources in a wide frequency range. Whenever possible, we have taken advantatge from the availability of daily monitorings in the cm domain, thanks to the Ryle Telescope and to the Green Bank Interferometer. This allowed us to estimate the source spectral indices between radio frequencies separated by two orders of magnitude. The targets for the observing program were chosen among the brightest REXRBs with luminous massive companions and declination $\\delta > -30^{\\circ}$. They include: Cygnus X-3, SS 433, LSI+61$^{\\circ}$303, Cygnus X-1 and GRS~1915+105. A summary of their main physical properties is condensed in Table~\\ref{properties}. Previous cm observations for all of them are abundant in the literature, and a few mm detections have been reported as well. Nevertheless, no extended and truly simultaneous cm/mm monitoring has been systematically carried out to our knowledge. The present work is an exploratory step in this direction. \\begin{table*} \\caption[]{\\label{properties} REXRBs properties} \\begin{tabular}{lccccc} \\hline Source & Classification & cm-Flux density (Jy) & mm-Flux density (Jy) & Periods & Remarks \\\\ & & Quiescent~~~Outburst & Quiescent~~~Outburst & \\\\ \\hline Cygnus X-3 & W-R+? & $\\sim$0.05~~~$>$20 & $\\sim$0.05~~~$>$3 & 4.8 h (orbital) & jets \\\\ Cygnus X-1 & O9.7Iab+BH & $\\sim$0.015~~~$\\sim$0.040 & 0.010 & 5.6 d (orbital) & \\\\ GRS~1915+105& Be?+BH & $\\sim$0.01~~~~$>$1 & 0.015 & & superluminal jets \\\\ LSI+61$^{\\circ}$303& Be+NS? &$\\sim$0.03~~~~$>$0.5 & $\\sim$0.01 &26.5 d (orb.), 4 yr ? & radio and X-ray period \\\\ SS~433 & OB?+? & $\\sim$0.5~~~~$>$10 & $\\sim$0.12 & 13 d (orb.), 164 d & precessing jets \\\\ \\hline \\end{tabular} ~\\\\ \\end{table*} \\begin{table} \\caption[]{\\label{obsres} IRAM results at 250 GHz (1.25 mm)} \\begin{tabular}{lccc} \\hline Source & Date & Number of & Flux density \\\\ & (1998 March) & subscans & (mJy) \\\\ \\hline Cygnus X-3 & 14.532 & 60 & $44\\pm 4$ \\\\ & 14.555 & 19 & $47\\pm 9$ \\\\ & 15.531 & 124 & $28\\pm 4$ \\\\ & 16.555 & 84 & $32\\pm 4$ \\\\ & 18.569 & 60 & $91\\pm 6$ \\\\ & 19.571 & 93 & $31\\pm 5$ \\\\ & 20.558 & 60 & $45\\pm 7$ \\\\ \\\\ Cygnus X-1 & 14.530 & 120 & $<$ 9 \\\\ & 15.555 & 100 & $<$ 17 \\\\ & 16.537 & 100 & $<$ 12 \\\\ & 18.583 & 100 & $<$ 18 \\\\ & 19.587 & 50 & $<$ 24 \\\\ & 20.575 & 102 & $<$ 21 \\\\ \\\\ GRS 1915+105 & 15.510 & 29 & $<$ 27 \\\\ & 16.508 & 126 & $<$ 12 \\\\ & 18.543 & 100 & $<$ 37 \\\\ & 19.543 & 32 & $<$ 47 \\\\ & 20.519 & 61 & $<$ 23 \\\\ \\\\ LSI+61$^{\\circ}$303 & 14.500 & 60 & $<$ 12 \\\\ & 14.715 & 224 & $9\\pm3$ \\\\ & 15.591 & 140 & $14\\pm4$ \\\\ & 16.588 & 107 & $<$ 12 \\\\ & 17.705 & 300 & $<$ 6 \\\\ & 18.610 & 100 & $<$ 15 \\\\ & 19.612 & 90 & $<$ 12 \\\\ & 20.610 & 160 & $<$ 11 \\\\ \\\\ SS 433 & 15.497 & 24 & $70\\pm 7$ \\\\ & 16.491 & 50 & $68\\pm 7$ \\\\ & 18.525 & 100 & $51\\pm10$ \\\\ & 19.525 & 100 & $48\\pm 6$ \\\\ & 20.508 & 28 & $72\\pm10$ \\\\ \\hline \\end{tabular} ~\\\\ \\end{table} ", "conclusions": "\\begin{enumerate} \\item We have conducted a series of millimetre and centimetre observations of a sample of REXRBs during a week long interval. Our target list included both Cygnus X-3 and SS 433, that were persistently detected at 250 GHz (1.25 mm) throughout the whole run. The REXRB LSI+61$^{\\circ}$303 was also detected at 250 GHz near the peak of one of its periodic radio outbursts. Several 250 GHz upper limits for other REXRBs are also reported. \\item For the detected sources, our results are in agreement with the synchrotron spectrum in REXRBs extending commonly up to millimetre wavelengths and possibly beyond. This observed fact reinforces the idea that these systems are able to accelerate relativistic electrons to very high energies, at least $\\gamma \\geq 10^3$. \\item The high frequency radio spectrum of Cygnus X-3 and SS 433 was also observed to steepen noticeably during the decay of flaring events. This behavior is interpreted in terms of energetic losses of the synchrotron emitting electrons. \\end{enumerate}" }, "0005/astro-ph0005384_arXiv.txt": { "abstract": "We explore the possibility that an intrinsic correlation between galaxy ellipticities arising during the galaxy formation process may account for part of the shear signal recently reported by several groups engaged in weak gravitational lensing surveys. Using high resolution N-body simulations we measure the projected ellipticities of dark matter halos and their correlations as a function of pair separation. With this simplifying, but not necessarily realistic assumption (halo shapes as a proxy for galaxy shapes), we find a positive detection of correlations up to scales of at least $20 \\hmpc$ (limited by the box size). The signal is not strongly affected by variations in the halo finding technique, or by the resolution of the simulations (over the range tested). We translate our three dimensional results into angular measurements of ellipticity correlation functions and shear variance which can be directly compared to observational results. We also make simulated angular surveys by projecting our simulation boxes onto the plane of the sky and applying a radial selection function. Measurements from these catalogs are consistent with the analytic projection of the statistics. Interestingly, the shear variance we measure is a small, but not entirely negligible fraction (from $\\sim10-20 \\%$, depending on the angular scale) of that seen by the observational groups, and the ellipticity correlation functions approximately mimic the functional form expected to be caused by weak lensing. The amplitude of these projected quantities depends strongly on the width in redshift of the galaxy distribution. If in the future photometric redshifts are used to pick out a screen of background galaxies with a small redshift width, then the intrinsic correlation may become comparable to the weak lensing signal. Although we are dealing with simulated dark matter halos, we might expect there to be a similar sort of signal when real galaxies are used. This could be checked fruitfully using a nearby sample with known redshifts. ", "introduction": "The large--scale mass distribution in the Universe is expected, through gravitational lensing, to imprint itself on the pattern of ellipticities measured from background galaxies (e.g., Blandford \\etal 1991; Miralda-Escud\\'{e} 1991). The angular correlations of such ellipticities, or the variance in ellipticities averaged in angular cells (amongst other statistics) can be compared to expectations for different cosmological models, and in principle can discriminate between them (e.g., Blandford \\etal 1991; Miralda-Escud\\'{e} 1991; Kaiser 1992; Jain \\& Seljak 1997). Since lensing is induced by the foreground mass distribution only, it provides the most direct method of studying the structure of mass in the Universe on large scales. Accordingly, detecting the shear signal induced by large scale structure has been the subject of a great deal of observational effort. Recently, four separate groups have reported detections of this cosmic shear, at levels comparable to that expected from currently popular models of structure formation (Van Waerbeke \\etal 2000 [hereafter VW]; Bacon, Refregier \\& Ellis 2000; Wittman \\etal 2000; Kaiser, Wilson \\& Luppino 2000 [hereafter KWL]). In attributing the observed correlations of ellipticities, or the shear variance, to large--scale structure, an important assumption is that the sample of background galaxies used contains no intrinsic correlation of ellipticities. If such a correlation were present in the sample of lensed sources, it could be attributed to lensing, and may enhance the detected signal. The argument for discounting this possibility is that a pair of galaxies separated by a small distance on the sky are nonetheless on average separated by a large distance along the line of sight. If a particular pair of galaxies are separated by a large distance, there is no good theoretical reason to expect their ellipticites to be intrinsicly correlated. However, the angular correlation of ellipticites predicted to be due to lensing is quite small --- on the order of $10^{-4}$ out to scales of several arcminutes. To detect this correlation against the random variations of galaxy ellipticities, observers take deep images and use large samples of background galaxies, perhaps $10^{5}$ per square degree. Some of these galaxies can be expected to be close not only in projection, but in real space as well. To what degree should we expect correlations in the actual ellipticities of nearby galaxies, and how much would the projection of such correlations add to any observed lensing signal? That such intrinsic correlations in galaxy ellipticities may exist is not implausible. If elongation by local tidal fields contributed significantly to galaxy ellipticities, then nearby galaxies could be expected to sample the tidal field in the same fashion, producing similar elongations. Alternately, if some elongation originating from a galaxy's last merger were to survive for a time comparable to the characteristic merger timescale for its environment, then one might expect galaxies to be preferentially aligned along the local large--scale structure, and thus similarly to each other. On larger scales, such elongation in cosmic structures appears to be present. For instance, the effects of large scale structure on the shapes of galaxy clusters have been the subject of much study. Cosmological N--body simulations have suggested that clusters tend to be oriented towards neighboring clusters or in directions defined by adjoining filaments and the merging subclusters which drain along them. (Dekel, West \\& Aarseth 1984; West, Dekel \\& Oemler 1989; West, Villumsen \\& Dekel 1991; van Haarlem \\& van de Weygaert 1993; Splinter \\etal 1997; de Theije, van Kampen \\& Slijkhuis 1998). Observations have typically indicated the presence of such alignments, either towards nearby clusters (Binggeli 1982; Flin 1987; West 1989a,b; Rhee, van Haarlem \\& Katgert 1992; Plionis 1994) or towards nearby large--scale structure in the galaxy distribution (Argyres \\etal 1986; Lambas, Groth \\& Peebles 1988); although not all studies support the presence of such alignments (Struble \\& Peebles 1985; Ulmer, McMillan \\& Kowalski 1989; Fong, Stevenson \\& Shanks 1990). The existence of correlations in the alignment of large--scale structures appears quite possible; perhaps similar intrinsic correlations in alignment exist on galactic scales. Theoretical expectations for the degree of correlation of intrinsic ellipticities can in principle be derived. The local gravitational shear can be expected to either align the intrinsic angular momentum of nearby galaxies (Lee \\& Pen 2000), or to similarly deform neighboring, non--rotating galaxies through tidal distortion (Ciotti \\& Dutta 1994). Thus, the statistics of the local tidal field can be related to the statistics of galaxy angular momenta (Catelan \\& Theuns 1996a,b; Catelan \\& Theuns 1997; Sugerman, Summers \\& Kamionkowski 1999); and therefore to the intrinsic correlations in galaxy ellipticities (Coutts 1996; Lee \\& Pen 2000; Catelan \\etal 2000, in preparation; Mackey and White 2000, in preparation). There have been numerous attempts to detect intrinsic correlations in galaxy alignments using low redshift samples; the picture painted by this work is unclear, as we can see from the following sample. Flin (1988) considered a sample of 118 galaxies in the Perseus supercluster and found that the spin axes of these galaxies were aligned with the supercluster plane. Muriel \\& Lambas (1992) reported a correlation of alignments seen with spirals taken from the ESO catalog and analyzed in three dimensions; when only projected data was considered, the correlation was no longer present. Garrido \\etal (1993) analyzed a sample covering a large area of sky in the northern hemisphere and claimed to find no evidence for correlations in alignment except within the Coma supercluster. Han, Gould \\& Sackett (1995) examined the spins of 60 galaxies in the Ursa Major filament and found no evidence for any alignment of spins. Cabanela \\& Aldering (1998) considered galaxy shapes extracted from a survey of Perseus--Pisces conducted using an automated plate scanner; statistically significant and color dependent correlations of galaxy ellipticities were found. On the other hand, Cabanela \\& Dickey (1999) used HI observations to determine the spins of 54 galaxies in the Perseus--Pisces supercluster; and found no evidence for preferential alignments of spin vectors. At this time evidence favors an orientation alignment between cD galaxies and the major axis of their parent cluster; but the presence or absence of any other galaxy shape correlations remains undetermined. In this paper, we use Nbody simulations to make theoretical predictions for the correlation of intrinsic galaxy ellipticities. For simplicity, we work with directly with the projected ellipticities of the simulated dark matter halos, without making assuming any model for the way galaxies form within them. The significance of our results will therefore be entirely dependent on whether galaxy ellipticities behave significantly differently from their halos, a problem we leave to future gasdynamical simulations. The layout of the paper is as follows. In \\S2, we describe the N--body dataset used, our halo catalog, and our measurement of projected ellipticities from the halos. We measure the three dimensional correlation functions of projected ellipticities in \\S3, and in \\S4 we decribe the construction of simulated surveys from our halo catalogs, with a geometry designed to mimic weak lensing observations. In \\S5, we project the three dimensional correlation functions into angular statistics, including the shear variance. We also compute these angular measures directly from our simulated surveys (as a consistency check). We compare our results to current observational data in \\S6, before discussing and summarizing our results in \\S7. ", "conclusions": "We have used Nbody simulations to make predictions for intrinsic correlations of galaxy ellipticities, under the assumption that galaxy shapes follow the shapes of their dark matter halos. Measurements of the ellipticity correlation functions in three-dimensions give a distinctive signal, which we measure with relatively small uncertainties on scales from $\\sim 0.5-30 \\hmpc$. These correlations vary by less than a factor of $\\sim 2$ for different halo finding techniques and different simulation resolutions. We project these three dimensional correlations into angular statistics, including the shear variance. We have done this both analytically, using a modified Limber's equation, and by making direct measurements from simulated surveys constructed by projecting the simulation boxes. We find that the amplitude of the angular statistics depends strongly on the redshift width of the galaxy distribution. With widths appropriate to present day surveys, we find that the intrinsic correlations we predict are around $10-20\\%$ of the currently measured signal, somewhat smaller than the $1 \\sigma$ errors on the measurements. Since the area of the sky surveyed for weak lensing is increasing rapidly, the intrinisic correlation may become detectable from these deep and wide surveys in the future. In any case, it seems to be worth bearing in mind that there could be this sort of contamination. In particular, one possible way of extracting more information from lensing which has received attention is the use of photometric redshift information, to break down the background galaxy distribution into a number of ``screens''. This would enable tomography to be carried out (e.g., Hu 1999). We have seen however that the intrinsic correlation may be quite large for these narrow redshift bins, so that it might become comparable to the weak lensing signal (Fig. \\ref{shearobs}). Of course, the extra information available in the form of photometric redshifts is likely to be very useful for deciding whether there is an intrinsic component, and if it exists, to separate it from the lensing signal. For example the cross-correlation of ellipticities (or co-variance of the shear) between different redshift bins can be compared to the correlation within bins, with only the later responding to intrinisic correlations. Something along these lines has already been carried out by KWL, albeit with two colour bands which both give wide redshift distributions (but with one deeper than the other). These authors find a higher shear signal for the deeper redshift sample, which is consistent with lensing, but in the wrong direction for intrinsic correlations. For the cross-correlation between samples, they do find slightly anomalous results, however. Another way of trying to measure any intrinsic ellipticity correlations would be to stick to the local universe, and to measure the three dimensional correlation functions (\\S3) from a redshift survey. If there really is a signal like that plotted in Fig. \\ref{xie}, then this could be measurable from a relatively small survey (by todays standards), with a few thousand galaxies. Even without redshifts, one might expect to find a measurable intrinsic signal from a relatively nearby angular sample of galaxies, like the APM survey (Maddox \\etal 1990), or Sloan Digital Sky Survey (Gunn \\& Weinberg 1995). If there are in fact some measurable correlations between real galaxy ellipticities, then this can be understood in the framework of structure formation by gravitational instability, with the ellipticities being linked to the angular momenta of galaxies, which are in turn set up by tidal torques from the shear in the initial density field (e.g., Peebles 1969, Barnes \\& Efstathiou 1987, Catelan and Theuns 1996a,b). This may explain why the ellipticity correlation functions we measure have similar functional forms to those caused by weak-lensing: both are responding to a cosmic shear field. Detection of correlated ellipticities, if they exist, may be useful for the study of galaxy formation (e.g., Sugerman \\etal 2000), or even cosmology (Lee \\& Pen 2000). It is also likely that the signal due to the intrinsic correlation will give qualitatively and measurably different results from weak lensing for some statistics we have not considered here. For example, the probability distribution of the lensing convergence is predicted to have a measurable skewness, something which can be used to determine $\\Omega$ (Bernardeau \\etal 1997). Measurements of this parameter from our simulated surveys by L. Van Waerbeke (private communication) yield a null result, the convergence pdf being consistent with a Gaussian distribution. The intrinisic correlations do not therefore appear to interfere with our ability to do cosmology in this way, and should not act as more than an additional source of noise (albeit correlated) when reconstructed mass maps are made. On the simulation side, one important issue is the fact that our results have apparently not converged with resolution. Although we find that the higher resolution of two simulations gives more intrinsic correlations, it is possible that given even higher resolution, things will begin to go the other way. Clearly this needs to be tested somehow in the future. Also, perhaps most important of all, we have assumed a very simple relationship between projected halo ellipticities and projected galaxy ellipticities. It is possible that adding gas dynamics and star formation to simulations will result in their being no significant correlation between the two. The tests which we have carried out which have most bearing on this are the use of two sets of different friends of friends groups, which respond to ellipticities either of the whole halo, or just the dense central region. As we find results for the two which are very similar, this is at least some evidence that the intrinsic correlation may be fundamental. \\bigskip As this paper was being completed, we became aware of similar work by Heavens \\etal (2000). These authors use the angular momentum of Nbody halos (also from Virgo simulations, but only at the lower of the two resolutions) to predict the intrinsic correlation of spiral galaxy ellipticities. They reach final results which are broadly similar (although they find much more noise), and also conclude that while these effects are likely to be minor for present surveys, they may become important with small redshift widths (in their case for shallower surveys). We also became aware of analytic work on a similar theme by Catelan \\etal (in preparation), and Mackey and White (in preparation). \\bigskip" }, "0005/astro-ph0005451_arXiv.txt": { "abstract": "We show that planet formation via both gravitational collapse and core accretion is unlikely to occur in equal mass binary systems with moderate ($\\sim 50$~AU) semi-major axes. Internal thermal energy generation in the disks is sufficient to heat the gas everywhere so that spiral structures quickly decay rather than grow or fragment. This same heating will inhibit dust coagulation because the temperatures rise above the vaporization temperatures of many volatile materials. We consider other processes not included in the model and conclude that our temperatures are conservatively estimated (low), i.e. planet formation is less likely in real systems than in the model. ", "introduction": "Both indirect evidence \\citep{ALS88,BSCG} and later direct imaging \\citep{Close97,MO96}, have shown that disks are quite common in young stellar systems. These disks are commonly thought \\citep{BS96} to be sites for planet or brown dwarf formation. A large fraction of stars are formed in binary systems \\citep{MGJS_PP4} and in the same star formation regions as single stars. Theory suggests that the most likely mechanisms responsible for forming Jovian mass planets or low mass brown dwarfs are either gravitational collapse of large scale spiral structure or coagulation of small solid grains followed by later accretion of additional gas (`core accretion') in the disks of forming stellar systems. Evaluating the effectiveness of these mechanisms is important for understanding the origin of our own solar system as well as planetary systems in other mature single or multiple systems. The \\irsfive\\ system serves as a useful observational testbed for comparison to theoretical modeling because of its relative youth \\citep[$\\sim10^5$~yr,][]{BTC94} and many previous detailed observations \\citep[see e.g.][]{Men-Hen-97}. This system consists of an extended nebulosity some 2400$\\times$1100~AU in size with an inner core of 220$\\times$76~AU \\citep{Mom98}. Two bipolar jets flow outward in each direction from the core in the plane perpendicular to its long axis. The core has been resolved into two sources with projected separation of about 50~AU and inferred disk masses of $\\sim.05$\\msun, each $\\sim$20-25~AU in diameter \\citep{Rodriguez98}. The total mass in the core has been estimated to contain 0.5--1.0\\msun\\ of material \\citep{ALS88,Mom98}, which produces $\\sim$30\\lsun\\ in luminous output \\citep{KeeMas90}. We present a numerical simulation of a binary star/disk$+$star/disk system using a two dimensional ($x,y$) Smoothed Particle Hydrodynamic (SPH) code. The dimensions of the disks and semi-major axis of the binary are chosen to be similar to the inner core region of \\irsfive. In the absence of strong constraints on the constituents of the binary (e.g. the masses of the two stars), we choose to set up a binary system consisting of identical components, obtained by setting up a single system in isolation, then duplicating it exactly. We assume each star and disk have mass $M_*=0.5$\\msun\\ and $M_D=0.05$\\msun, respectively. The disk radius is set to $R_D=15$~AU which, for a semi-major axis of $a=50$~AU, is comparable to the largest stable streamline \\citep{Pac77}. The mass and temperature of the disk are distributed according to $r^{-3/2}$ and $r^{-1/2}$ power laws respectively. The absolute scale of each power law is determined from the disk mass, the radial dimensions of the disk and the condition that the Toomre stability parameter, $Q$, is no smaller than \\qmin=1.5 over the entire disk. This value ensures that the simulation begins in a state marginally stable against the growth of spiral structure, so that we do not accidently `discover' a collapsed object early in the evolution which in reality is an artifact of our initial condition. Both density and temperature are free to vary in time and space, so the initial condition will not prevent spiral structure growth or fragmentation, if the evolution leads to such. The gas is set up on circular orbits around the star so that pressure and gravitational forces exactly balance centrifugal forces. Radial motion is zero. The magnitudes of the pressure and self-gravitational forces are small compared to the stellar gravity, so the disk is nearly Keplerian in character. Approximately 60000 equal mass particles are set on a series of concentric rings around the star in a single, star/disk system, then duplicated, bringing the total number of particles to $\\sim$120000. The two stars and disks are offset equal distances in the $+x$ and $-x$ directions. We define the binary semi-major axis to be $a=50$~AU, similar to \\irsfive. Only weak constraints on eccentricity exist in \\irsfive, primarily consisting of the sizes of the observed disks: eccentricities larger than $e=0.3$ would lead to rapid Roche lobe overflow. We set $e=0.3$ to be the initial value in this simulation. The system is at apoapse at time $t=0$ with the orbital velocities defined by approximating each star$+$disk system as a point mass, so that the orbit determination reduces to the solution of the two body problem. The disks are self gravitating and each star is modeled as a point mass free to move in response to gravitational forces from the rest of the system. The stellar gravitational forces are calculated using a Plummer potential with a softening radius of 0.2~AU, which also serves as an accretion radius, $r_{acc}$. SPH particles with trajectories that pass closer than $r_{acc}$ to a star are absorbed, and the star's mass and momentum increase accordingly. The thermodynamic evolution is identical to that described in \\citet{DynamII}. Thermal energy is added to the gas due to active hydrodynamic processes using an artificial viscosity scheme, which approximately models shocks and turbulence. This heating is roughly equivalent in magnitude to an alpha model with $\\alpha\\sim 2-5 \\times10^{-3}$. Thermal energy is removed from the disk gas by radiative cooling due to passive blackbody emission from the disk's photosphere surfaces. The blackbody temperature is calculated at each time step and for each SPH particle. This treatment remedies a major shortcoming of previous models \\citep{DynamI,Pick98,Boss97} which used a `locally isothermal' or `locally adiabatic' approximation to show that relatively low mass disks can undergo fragmentation and/or collapse, despite earlier claims \\citep{PodPP3} that a very massive disk is required. ", "conclusions": "" }, "0005/astro-ph0005498_arXiv.txt": { "abstract": "Observations of Type I X-ray bursts have long been taken as evidence that the sources are neutron stars. Black body models approximate the spectral data and imply a suddenly heated neutron star cooling over characteristic times of seconds to minutes. The phenomena are convincingly explained in terms of nuclear burning of accreted gas on neutron stars with low mass companion stars. Prospects are promising that detailed theory and data from RXTE and future missions will lead to better determinations of important physical parameters (neutron star mass and radius, composition of the accreting gas, distance of the source). Among the variety of bursts observed, there are probably representatives of different kinds of explosive burning. RXTE's discovery of a 2.5 ms persistent coherent period from one Type I burster has now linked bursters indisputably to the epitome of a neutron star, a fast spinning magnetic compact object. Oscillations in some bursts had already been thought to arise from the neutron stars' rotations. Detailed observations of these oscillations are touchstones of how the explosive bursts originate and progress, as well as independent measures of the neutron star parameters. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005337_arXiv.txt": { "abstract": "The BeppoSAX Wide Field Cameras detected a transient in NGC\\,6440 on 1998 Aug 22. ROSAT did not detect this source on 1998 Sep 8, indicating that the X-ray flux had decreased by a factor 400 at least, and/or that the X-ray spectrum had become appreciably softer. Analysis of archival ROSAT HRI data reveals two sources associated with NGC\\,6440; one of these may be the transient in quiescence. We have also obtained B and R images of NGC\\,6440 on 1998 Aug 26 and again on 1999 July 15, when the transient had returned to quiescence. Subtraction of these images reveals one object in the core which was brighter in B, but not in R, during the X-ray outburst. We estimate $B\\simeq22.7$ and $(B-R)_0\\ltap0$ on 1998 Aug 26, which makes it a viable candidate counterpart. ", "introduction": "NGC\\,6440 is a globular cluster near the center of the Galaxy, at a distance of 8.5\\,kpc and reddened by $E(B-V)=1.0$ (Ortolani et al.\\ 1994). A bright X-ray source was detected near this cluster with OSO-7 and with UHURU from 1971 December 17 to 1972 January 21 (Markert et al.\\ 1975, Forman et al.\\ 1976). UHURU observations obtained before 1971 Oct 23 and after 1972 Mar 1 did not detect the source (Forman et al.\\ 1976). During the outburst the transient X-ray source had a virtually constant luminosity of about $3\\times 10^{37}\\ergs$, in the 2-11\\,keV band. Before and after the outburst the flux was less than 5\\%\\ of this. (We use the conversion of UHURU $\\cts$ to flux given by Bradt \\&\\ McClintock 1983; and the absorption column $N_H=6.9\\times10^{21}\\cmsq$ determined by in 't Zand et al.\\ 1999.) \\nocite{obb94}\\nocite{mbc+75}\\nocite{fjt76}\\nocite{bm83}\\nocite{zvs+99} A dim source was detected in the core of NGC\\,6440 with the Einstein satellite, and again with ROSAT, at a luminosity of $\\sim10^{33}\\ergs$, in the 0.5-2.5\\,keV band (Hertz \\&\\ Grindlay 1983; Johnston et al.\\ 1995); after conversion to the 2-11\\,keV band, this corresponds to $\\ltap 10^{-4}$ of the outburst flux. \\nocite{hg83}\\nocite{jvh95} On 1998 August 22 a bright transient source appeared again in NGC\\,6440, observed with BeppoSAX. The position coincides with the globular cluster within the accuracy of 1$'$. This time the outburst lasted rather shorter: the source had a luminosity of $6.0\\times 10^{36}\\ergs$ on Aug 22, $3.6\\times 10^{36}\\ergs$ on Aug 26, and $< 10^{36}\\ergs$ on Sep 1, in the 2-10\\,keV band (in 't Zand et al.\\ 1999, also see Fig.\\,\\ref{xcur}). Like persistent bright X-ray sources, transients occur more often in globular clusters per unit of stellar mass than in the galactic disk. To understand this overabundance one would like to study these sources at optical and ultraviolet wavelengths. So far, no transient X-ray source in a globular cluster has been optically identified. Such identification is difficult because the relatively large error circle of the X-ray position contains a large number of stars. We therefore obtained a ROSAT HRI observation as soon as possible after the detection of the transient with BeppoSAX, in the hope of improving the X-ray position. The optical brightness of soft X-ray transients is known to vary in tandem with the X-ray luminosity (for a review, see e.g.\\ Chen et al.\\ 1997). We therefore obtained optical images of NGC\\,6440 to look for objects that vary in tandem with the X-ray flux, in the hope of identifying the optical counterpart of the transient. \\nocite{csl97} In Sect.\\,2 we describe the results of the new ROSAT HRI observation, and also analyse archival ROSAT data of NGC\\,6440. In Sect.\\,3 we describe the optical observations and the search for an optical counterpart to the X-ray source. Our results and their implications are discussed in Sect.\\,4. ", "conclusions": "We have tried to reconstruct the approximate X-ray light\\-curve of the August 1998 outburst by combining data from various measurements. Detections were made by the Wide Field Cameras and the Narrow Field Instruments on board of BeppoSAX on Aug 22 and 26, respectively; and by the XTE All Sky Monitor in a seven-day period starting on Aug 19. We revise the flux detected with the Wide Field Cameras to $35\\pm4$\\,mCrab, slightly upwards from the values given in In 't Zand et al.\\ (1999), on the basis of a better calibration. Upper limits were obtained with the BeppoSAX Wide Field Cameras on Sep 1 and with the XTE All Sky Monitor in the seven days periods preceding and following the detection. Fig.\\,\\ref{xcur} shows the resulting lightcurve. \\begin{figure} \\centerline{\\psfig{figure=xcur.ps,width=\\columnwidth,clip=t} {\\hfil}} \\caption{The August 1998 outburst of the X-ray transient in NGC\\,6440 as observed with BeppoSAX ($\\bullet$), and with the XTE All Sky Monitor (horizontal lines, solid for the detection, dotted for upper limits). The dashed line indicates exponential decay with e-folding time 5 days, passing through the BeppoSAX detections. The V indicates the time of the optical observations. \\label{xcur}} \\end{figure} Extrapolating the 5-day exponential decay from the BeppoSAX detections to the time of the ROSAT HRI observation, we predict an X-ray luminosity of $\\sim 10^{35}\\ergs$ in the ROSAT band; this is about an order of magnitude above the observed upper limit. This may imply that the decay accelerated; or alternatively that the spectrum softened, since low-energy photons are much more affected by the heavy absorption towards NGC\\,6440. There is indeed evidence for other X-ray transients that the spectrum in the low state is much softer than during outburst, e.g.\\ for Aql X-1 (Verbunt et al.\\ 1994).\\nocite{vbj+94} The X-ray lightcurve shown in Fig.\\,\\ref{xcur} implies that the optical observations were made at an X-ray flux level of about 10\\,mCrab, corresponding to a source luminosity $\\simeq2\\times10^{36}\\ergs$. Van Paradijs \\&\\ McClintock (1994) give a semi-empirical relation between the orbital period, X-ray luminosity and absolute visual magnitude of a low-mass X-ray binary. Applying this relation to the transient in NGC\\,6440 with the estimate of the X-ray luminosity for the time of the optical observation, we obtain $M_{\\rm V}\\simeq4.0$ for an assumed 1\\,hr period. At the distance and reddening of NGC\\,6640 this corresponds to $V\\simeq21.7$. \\nocite{vpm94} The intrinsic $B-V$ colour of low-mass X-ray binaries is close to zero; with the reddening to NGC\\,6440 we thus predict $B\\simeq22.7$ for a one hour period. For a period of 5\\,hr (0.2\\,hr) the predicted magnitude is about 1 magnitude brighter (fainter). We conclude that the candidate in the error circle of X2 is viable; the proximity of the predicted $B$ magnitude to the observed one is fortuitous, considering that the spread in the relation given by Van Paradijs \\&\\ McClintock is about a magnitude, and that our estimate of the X-ray luminosity is uncertain. We therefore dare not estimate an orbital period on the basis of the magnitude of our candidate. The core of NGC\\,6440 contains PSR\\,B\\,1745$-$20 (Lyne et al.\\ 1996). The total energy loss $\\dot E\\equiv I\\Omega\\dot\\Omega$ for the pulsar is about $6.6\\times 10^{32}\\ergs$, where $I$ is the moment of inertia of the neutron star, $\\Omega$ its rotation frequency and $\\dot\\Omega$ the time derivative of $\\Omega$. Typical X-ray luminosities for radio pulsars are of order $L_{\\rm x}\\sim 10^{-3}I\\Omega\\dot\\Omega$ (e.g.\\ Fig.\\,4 in Verbunt et al.\\ 1996).\\nocite{vkb+96} We conclude that it is very unlikely that the pulsar is responsible for the observed X-ray flux of X1 or X2." }, "0005/astro-ph0005101_arXiv.txt": { "abstract": "The expected data rate produced by the Low Frequency Instrument (LFI) planned to fly on the ESA Planck mission in 2007, is over a factor 8 larger than the bandwidth allowed by the spacecraft transmission system to download the LFI data. We discuss the application of lossless compression to Planck/LFI data streams in order to reduce the overall data flow. We perform both theoretical analysis and experimental tests using realistically simulated data streams in order to fix the statistical properties of the signal and the maximal compression rate allowed by several lossless compression algorithms. We studied the influence of signal composition and of acquisition parameters on the compression rate \\Cr\\ and develop a semiempirical formalism to account for it. The best performing compressor tested up to now is the arithmetic compression of order 1, designed for optimizing the compression of white noise like signals, which allows an overall compression rate $\\overline{\\Cr} = 2.65 \\pm 0.02$. We find that such result is not improved by other lossless compressors, being the signal almost white noise dominated. Lossless compression algorithms alone will not solve the bandwidth problem but needs to be combined with other techniques. ", "introduction": "\\label{sec:introduction} The {\\sc Planck} satellite (formerly COBRAS/SAMBA, \\cite{Bersanelli:etal:1996}), which is planned to be launched in 2007, will produce full sky CMB maps with high accuracy and resolution over a wide range of frequencies (\\cite{Mandolesi:etal:1998a,Puget:etal:1998}). Table~\\ref{tab:PlanckTable}\\ summarizes the basic properties of LFI aboard Planck. The reported sensitivities per resolution element -- i.e. a squared pixel with side equal to the Full Width at Half Maximum (FWHM) extent of the beam --, in terms of antenna temperature, represents the goals of LFI for 14 months of routine scientific operations) as recently revised by the LFI Consortium (\\cite{Mandolesi:etal:1999}). The limited bandwidth reserved to the downlink of scientific data calls for huge lossless compression, theoretical upper limit being about four (\\cite{Maris:etal:SAIT}). Careful simulations are demanded to quantify the capability of true compressors for ``realistic'' synthetic data and improve the theoretical analysis, including CMB signal (monopole, dipole and anisotropies), foregrounds and instrumental noise. During the data acquisition phase the Planck satellite will rotate at a rate of one circle per minute around a given spin axis that changes its direction every hour (of 2.5$'$ on the ecliptic plane in the case of simple scanning strategy), thus observing the same circle on the sky for 60 consecutive times (\\cite{Mandolesi:etal:1998a,Mandolesi:etal:1998b}). LFI will produce continuous data streams of temperature differences between the microwave sky and a set of on-board reference sources; both differential measurements and reference source temperatures must be recorded. The LFI Proposal assumes a sampling time $\\tau_{\\mathrm{s}} \\sim 7$~msec for each detector (\\cite{Mandolesi:etal:1998a}), thus calling for a typical data rate of $\\sim 260$~Kb/sec, while the allocated bandwidth to download Planck data to ground is in total $\\sim 60$~Kb/sec. Assuming the total bandwidth to be equally split between instruments, $\\approx 30$\\ Kb/sec on the average would be assigned to LFI asking for a compression of about a factor $8.4$. Data have to be downloaded without information losses and by minimizing scientific processing on board. A possible solution would be to adapt the sampling rate to the angular resolution specific for each frequency. This should allow to save about up to a factor $\\approx 9$\\ for the 30 GHz channel, but since only $\\approx 7\\%$\\ of the samples come from such channel (see table \\ref{tab:PlanckTable}) the overall reduction in the final data rate would be $\\approx 17\\%$. On the other hand, it is unlikely that the bandwidth for the downlink channel may be enhanced to solve the bandwidth problem, since the ground facilities are shared between different missions and there is the need to minimize possible cross-talks between the instrument and the communication system. With the aim of optimizing of the transmission bandwidth dedicated to the downlink of LFI data from the Planck spacecraft to the FIRST/Planck Ground Segment, we analyze in detail the role that can be played by lossless compression of LFI data before they are sent to Earth. We apply different compression algorithms to suitable sets of Planck LFI simulated data streams generated by considering different combinations of astrophysical and instrumental signals and for different instrumental characteristics and detection electronics. The first considered contribution is that introduced by receiver noise: we consider here the case of pure white noise and of white noise coupled to $1/f$ noise with different knee frequencies. The reference load temperature is assumed to be 20~K for present tests; because of the strong dependence of the $1/f$ noise on the load temperature, this can be considered a worst case, since the actual baseline reference load is of 4~K. Different sky signal sources are subsequently added to the receiver noise: CMB fluctuations, CMB dipole, Galaxy emission and extragalactic point sources. The signal from the different sky components are convolved with the corresponding antenna pattern shapes, assumed to be symmetric and gaussian with the FWHM reported in Table~1. We generate simulated data streams at the two extreme frequency channels, 30 GHz and 100 GHz and consider data streams with different time lengths. Regarding the detection electronics, we explore different signal offset and scaling. \\noindent The large number of above combinations was systematically explored using an automated program generator as described by \\cite{Maris:etal:1998}. \\noindent In Section \\ref{sec:components} we characterize quantitatively the LFI signal component by component. Section \\ref{sec:acquisition} we discuss how the acquisition chain is modeled to perform compression simulations. A theoretical analysis of the compression efficiency is presented in section \\ref{sec:theorethical}. While section \\ref{sec:statistical} is devoted to the analysis of the signal statistics. The subject of quantization error is illustrated in section \\ref{sec:quantum}. The experimental protocol and results about compression are reported in section \\ref{sec:evaluation:and:results}. Further constraints on the on-board data compression are reported in section \\ref{sec:packeting}. A proposal for an alternative coding method is made in section \\ref{sec:coding:scheme}. The overall compression rate is estimated in section \\ref{sec:overall:cr}. Conclusions are in section \\ref{sec:conclusions}. Appendix \\ref{appendix:a} is included to further illustrate the estimation of the overall compression rate. ", "conclusions": "\\label{sec:conclusions} The expected data rate from the Planck Low Frequency Instrument is $\\approx 260$\\ kbits/sec. The bandwidth for the scientific data download currently allocated is just $\\approx 60$\\ kbit/sec. Assuming an equal subdivision of the bandwidth between the two instruments on-board Planck, an overall compression rate of a factor 8.7 is required to download all the data. In this work we perform a full analysis on realistically simulated data streams for the 30 GHz and 100 GHz channels in order to fix the maximum compression rate achievable by loss-less compression methods, without considering explicitly other constrains such as: the power of the on-board Data Processing Unit, or the requirements about packet length limits and independence, but taking in account all the instrumental features relevant to data acquisition, i.e.: the quantization process, the temperature / voltage conversion, number of quantization bits and signal composition. As a complement to the experimental analysis we perform in parallel a theoretical analysis of the maximum compression rate. Such analysis is based on the statistical properties of the simulated signal and is able to explain quantitatively most of the experimental results. Our conclusions about the statistical analysis of the quantized signal are: I) the nominally quantized signal has an entropy $h \\approx 5.5$\\ bits at 30GHz and $h \\approx 5.9$\\ bits at 100GHz, which allows a theoretical upper limit for the compression rate $\\approx 2.9$\\ at 30 Ghz and $\\approx 2.7$\\ at 100 GHz. II) Quantization may introduce some distortion in the signal statistics but the subject requires a deepest analysis. Our conclusions about the compression rate are summarized as follows: I) the compression rate \\Cr\\ is affected by the quantization step, since greater is the quantization step higher is \\Cr\\ (but worse is the measure accuracy). II) \\Cr\\ is affected also by the stream length $L_u$, i.e. more circles are compressed better then few circles. III) the dependencies on the quantization step and $L_u$\\ for each compressor may be summarized by the empirical formula (\\ref{eq:cr:fit}). A reduced compression rate $\\CrOne$\\ is correspondingly defined. IV) the \\Cr\\ is affected by the signal composition, in particular, by the white noise r.m.s. and by the dipole contribution, the former being the dominant parameter and the latter influencing \\Cr\\ for less than $\\approx 6\\%$. The inclusion of the dipole contribution reduces the overall compression rate. The other components (1/f noise, CMB fluctuations, the galaxy, extragalactic sources) have little or no effect on \\Cr. In conclusion, for the sake of compression rate estimation, the signal may be safely represented by a sinusoidal signal plus white noise. V) since the noise r.m.s. increases with the frequency, the compression rate \\Cr\\ decreases with the frequency, for the LFI $\\Delta\\Cr/\\Cr \\lsim 10\\%$. VI) the expected random r.m.s. in the overall compression rate is less than $1\\%$. VII) we tested a large number of off-the-shelf compressors, with many combinations of control parameters so to cover every conceivable compression method. The best performing compressor is the arithmetic compression scheme of order 1: {\\tt arith-n1}, the final \\CrOne\\ being 2.83 at 30 GHz and 2.61 at 100 GHz. This is significantly less than the bare theoretical compression rate (\\ref{eq:cr:theory}) but when the quantization process is taken properly into account in the theoretical analysis, this discrepancy is largely reduced. VIII) taking into account the data flow distribution among different compressors the overall compression rate for {\\tt arith-n1} is: $$ \\overline{\\Cr}_{,\\mathrm{arith-n1}} \\approx 2.65 \\pm 0.02 $$ \\noindent This result is due to the nature of the signal which is noise dominated and clearly excludes the possibility to reach the required data flow reduction through loss-less compression only. Possible solutions deal with the application of lossy compression methods such as: on-board averaging, data rebinning, or averaging of signals from duplicated detectors, in order to reach an overall lossy compression of about a factor 3.4, which coupled with the overall loss-less compression rate of about 2.65 should allow to reach the required final compression rate $\\approx 8.7$. However each of these solutions will introduce heavy constraints and important reduction of performances in the final mission design, so that careful and deep studies will be required in order to choose the best one. Another solution to the bandwidth problem would be to apply a coarser quantization step. This has however the drawback of reducing the signal resolution in terms of $\\Delta T/T$. Lastly the choice of a given compressor cannot be based only on its efficiency obtained from simulated data, but also on the on-board available CPU and on the official ESA space qualification: tests with this hardware platform and other compressors will be made during the project development. Moreover, in the near future long duration flight balloon experiments and ground experiments (see \\cite{Lasenby:etal:1998,Debe:masi:98}) will provide a solid base to test and improve compression algorithms. In addition the final compression scheme will have to cope with requirements about packet length and packet independence. We discuss briefly this problems recalling two proposals (\\cite{Fast:Compression}, \\cite{Coding:Compression}) which suggest solutions to cope with these constrains. \\appendix" }, "0005/astro-ph0005271_arXiv.txt": { "abstract": "We present spectra of 6 type 1 Seyfert galaxies, 2 Seyfert 2, a starburst galaxy and a compact narrow line radiogalaxy, taken in two spectral ranges centered around the near--IR \\hbox{Ca\\,{\\sc ii}} triplet ($\\sim$ 8600\\AA\\ ), and the Mgb stellar feature at 5180\\AA . We measured the equivalent width (EWs) of these features and the Fe$_{52}$ and Fe$_{53}$ spectral indices. We found that the strength of the IR \\hbox{Ca\\,{\\sc ii}} triplet (CaT) in type 1 Seyfert galaxies with prominent central point sources, is larger than what would be expected from the observed strength of the blue indices. This could be explained by the presence of red supergiants in the nuclei of Seyfert 1 galaxies. On the other hand, the blue indices of these galaxies could also be diluted by the strong \\hbox{Fe\\,{\\sc ii}} multiplets that can be seen in their spectra. We have also measured the stellar and gas velocity dispersions of the galaxies in the sample. The stellar velocity dispersions were measured using both, the Mgb and CaT stellar features. The velocity dispersion of the gas in the narrow line region (NLR) was measured using the strong emission lines \\hbox{[O\\,{\\sc iii}]} $\\lambda \\lambda$5007, 4959 and \\hbox{[S\\,{\\sc iii}]} $\\lambda$9069. We compare the gas and star velocity dispersions and find that both magnitudes are correlated in Seyfert galaxies. Most of the Seyfert 1 we observe have stellar velocity dispersion somehow greater than that of the gas in the NLR. ", "introduction": "The presence of young massive stars in the nuclear regions of Seyfert galaxies was strongly suggested by the detection of the near--IR absorption \\hbox{Ca\\,{\\sc ii}} triplet ($\\sim$ 8600\\AA ) (CaT) in a sample of active galactic nuclei by Terlevich, D\\'{\\i}az \\& Terlevich \\shortcite{TDT90} (hereafter TDT90). This stellar feature depends strongly on gravity and only weakly on metallicity, and is known to be specially strong in young red supergiants (Jones, Alloin \\& Jones 1984; D\\'{\\i}az, Terlevich \\& Terlevich 1989). The analysis of this feature led TDT90 to conclude that strong starbursts should be present in the nuclei of Seyferts 2, since, despite the weakness or dilution observed in the blue stellar absorption lines, the CaT was found to be very strong in the nuclear spectra of the twelve galaxies of this type that were observed. Moreover, the only three Seyfert 1 galaxies included in their sample also show CaT in absorption, suggesting at least some contribution by a young stellar population. Subsequent IR spectroscopy (1.5 to 2.3 $\\mu$m) of normal and active galaxies performed by Oliva et al. (1995) allowed them to conclude that the Seyfert 2 nuclei of their sample were compatible with evolution from a pre-cursor starburst. They also find that the 1.6 to 2.3 $\\mu$m stellar continuum of Seyfert nuclei is too red to be accounted for by a non-thermal continuum but is compatible with reprocessed radiation from hot dust. Cid Fernandes \\& Terlevich (1992;1993;1995), in a critical analysis of the simple unified scenario \\cite{Ant93}, proposed that the observed strong CaT in Seyfert type 2 galaxies with strong and blue optical continuum plus the absence of broad lines combined with the low continuum polarization, were the result of the presence in the nuclear region of unpolarized starlight from very young stars, i.e.~a nuclear/circumnuclear starburst or star forming toroid. This simple suggestion seems able to overcome most of the difficulties faced by the basic unified model for Seyfert 2, while preserving its attractive features. Schmitt, Storchi--Bergmann \\& Cid Fernandes \\shortcite{SSC98} have recently found that the spectra of many Seyfert 2 can be modelled by the sum of the spectra of a young stellar cluster (age $\\sim$ 100 Myr) and an old one (age $\\sim$ 10 Gyr), and that these models reproduce the observations better than the traditional ones consisting of a blue featureless continuum (BFC) and an old stellar population, thus confirming Cid Fernandes \\& Terlevich suggestions. The question about the origin of the nuclear continuum in Seyfert 2, has given rise to searches of young stars in their nuclei. Heckman et~al. \\shortcite{Heck97} and G\\'onzalez Delgado et~al. \\shortcite{Gon98} have presented high resolution UV images, taken with the HST, of 4 Seyfert 2 galaxies, finding compact nuclear starbursts in all of them. They have also found spectral features from young hot stars in the UV spectra of these 4 galaxies. Their main conclusion is that, in all the galaxies they have studied, the observed continuum is exclusively due to a nuclear/circumnuclear young starburst and the energy emitted by the nuclear starburst is, at least, of the same order as the one produced by the buried AGN. Powerful starbursts have also been found in the nuclei of several LINERS by Colina et~al. \\shortcite{Col98} and Maoz et~al. \\shortcite{Maoz98} and very recently in a QSO by Brotherton et~al. \\shortcite{BBS99}. If the results presented above can be generalized to all type 2 Seyferts then, we could conclude that nuclear starbursts ought to play an important role in the total energy emitted in their nuclear regions. Furthermore, since in unified models Seyferts 1 and 2 are not physically different kinds of object, but the consequence of a different viewing angle, one can conclude that, if nuclear starbursts are found to be energetically important in Seyfert 2, so they should be in Seyfert 1. However, the starbursts in Seyfert 1 nuclei, if present, may be somewhat different from those in Seyfert 2: the narrow H$\\alpha$ + \\hbox{[N\\,{\\sc ii}]} emission in Seyfert 2 galaxies is more extended than in Seyfert 1 hosts (Pogge 1989; Gonz\\'alez Delgado \\& P\\'erez 1993). Also, the galaxies with type 2 Seyfert nuclei have enhanced infrared emission from their disks compared to those with type 1 nuclei or with normal spiral galaxies \\cite{Mai95}. From their near--IR observations Oliva et~al. \\shortcite{Oli95} found that the $M/L$ ratio in Seyfert 1 is similar to the one shown by normal early type spiral galaxies, whereas this ratio is found to be lower in Seyfert 2, pointing to younger stellar populations. Also, Gonz\\'alez Delgado et~al. \\shortcite{Gon97} found that circumnuclear star-forming rings are more common in Seyfert 2 than in Seyfert 1. Moreover the few UV HST images that exist of type 1 Seyferts do not show compact starbursts in their nuclei but point sources. In this paper we use the fact that the nuclear light output in Seyfert 1 galaxies has little contamination from the surrounding bulge to search for signatures of a nuclear starburst. If, for example, strong CaT absorptions were detected in the unresolved nuclear component, because of the limited surface brightness, they could not be due to bulge contamination. A starburst signature should be possible to detect, even in the presence of a dominant nuclear component. In other words, if the result from Heckman et~al. \\shortcite{Heck97} and Gonz\\'alez Delgado et~al. \\shortcite{Gon98} can be generalized -- i. e. that the luminosity of the starburst is equal or larger than that of the AGN -- then the surface brightness of the stellar population giving rise to the CaT feature should be much higher than that from the old bulge component in order to be detected on top of the dominant AGN spectrum and therefore detectable in the presence of the nuclear continuum. We have looked for signatures of young stars in the optical/NIR spectra of 6 type 1 Seyfert galaxies. Although, in principle, the UV would be the best spectral band to detect absorptions from young stars, the strong broad line region (BLR) contamination at the wavelengths of the stronger stellar features (\\hbox{[C\\,{\\sc iv}]} $\\lambda$1550 \\AA , \\hbox{[Si\\,{\\sc iv}]} $\\lambda$1400 \\AA ) makes this method not viable. On the other hand, as discussed by TDT90, the region around the near--IR CaT is relatively free of strong emissions. \\begin{footnote}{The \\hbox{O\\,{\\sc i}} $\\lambda$8400 \\AA\\ and the \\hbox{[Fe\\,{\\sc II}]} $\\lambda$8617 \\AA\\ lines are exceptions to this. The first line is broad and prominent in luminous Seyfert 1 galaxies, while the second one is narrow and seems to be present in some starburst and Seyfert 2 galaxies.} \\end{footnote} It is known that the stellar kinematics in the nuclei of Seyfert galaxies is similar to that of normal spiral galaxies, since both kinds follow the same Faber--Jackson relation (TDT90; Nelson \\& Whittle 1996). It is also known that the gas motions in the narrow line region (NLR) of Seyfert galaxies seem to be dominated by the gravitational field of the bulge, since the width of the \\hbox{[O\\,{\\sc iii}]} $\\lambda$ 5007 emission line is correlated with the nuclear stellar velocity dispersion (Wilson \\& Heckman 1985; TDT90; Nelson \\& Whittle 1996). TDT90 found a small population of type 2 Seyfert galaxies with gas velocities much larger than the stellar ones. The existence of this population was confirmed by Nelson \\& Whittle \\shortcite{NW96}. The kinematics of the stars and gas in the nuclei of our sample galaxies was also studied in this paper, improving previous results by adding 6 new Seyfert 1 galaxies to the above refered samples. \\begin{figure} \\centerline{ \\psfig{figure=comp.eps,width=8cm,clip=}} \\caption{Elimination of atmospheric absorption bands. In both frames, the top spectrum is the atmospheric absorption corrected one. In Mk~1098 (bottom), the correction method caused broad ``emission bands'' to appear. On the contrary in Mk~423 (top), only weak and narrow ``emission lines'' (marked with arrows) appear after the correction procedure.} \\label{comp} \\end{figure} \\begin{table} \\caption{\\label{obs}Observational parameters.} \\begin{tabular}{lcc} \\hline $\\lambda_{c}$ & 5100 \\AA & 8700 \\AA \\\\ \\hline Dates & 3--4 May 1991 & 3--4 May 1991 \\\\ Telescope & WHT & WHT \\\\ Spectrograph & ISIS & ISIS \\\\ Grating & 600B & 316R \\\\ Detector & CCD TEK 1 & CCD TEK 2 \\\\ Filter & --- & CG495 \\\\ Spectral range & 4706--5607\\ \\AA & 7983--9573 \\AA \\\\ Dispersion & 0.73 \\AA\\ pixel$^{-1}$ & 1.39 \\AA\\ pixel$^{-1}$ \\\\ Spatial scale & 0.3 arcsec pixel$^{-1}$ & 0.3 arcsec pixel$^{-1}$ \\\\ Slit width & 1.04 arcsec & 1.04 arcsec \\\\ \\hline \\end{tabular} \\end{table} \\begin{table*} \\begin{minipage}{13cm} \\caption{\\label{obssample}Observations and galaxy sample} \\begin{tabular}{lccccccc} \\hline Galaxy & Alternate & R.A. & Decl. & P.A. & Night & $\\lambda_{c}$ & Exposure\\\\ & Designation & (1950) & (1950) &($^{\\circ}$) && (\\AA) & (s) \\\\ \\hline NGC 4235 & IC 3098 & 12 14 36 & + 07 28 09 & 70 & 2 & 5100 & 3600 \\\\ &&&&&& 8700 & 3600 \\\\ NGC 5940 & UGC 9876 & 15 28 51 & + 07 37 38 & 70 & 2 & 5100 & 3600 \\\\ &&&&&& 8700 & 3600 \\\\ NGC 6104 & UGC 10309 & 16 14 40 & + 35 49 50 & 83 & 1 & 5100 & 3600 \\\\ &&&&&& 8700 & 3600 \\\\ Mk 270 & NGC 5283 & 13 39 41 & + 67 55 28 & 70 & 2 & 5100 & 3600 \\\\ &&&&&& 8700 & 3600 \\\\ Mk 423 & MCG 6-25-72 & 11 24 07 & + 35 31 34 & 170 & 1 & 5100 & 5400 \\\\ &&&&&& 8700 & 5400 \\\\ Mk 759 & NGC 4152 & 12 08 05 & + 16 18 41 & 35 & 2 & 5100 & 3600 \\\\ &&&&&& 8700 & 3600 \\\\ Mk 766 & NGC 4253 & 12 15 55 & + 30 05 27 & 108 & 1 & 5100 & 5400 \\\\ &&&&&& 8700 & 5400 \\\\ Mk 885 & --- & 16 29 43 & + 67 29 06 & 90 & 1 & 5100 & 3600 \\\\ &&&&&& 8700 & 3600 \\\\ Mk 1098 & --- & 15 27 37 & + 30 39 24 & 70 & 2 & 5100 & 3600 \\\\ &&&&&& 8700 & 3600 \\\\ 3C305 & IC 1065 & 14 48 18 & + 63 28 36 & 57 & 1,2 & 5100 & 3600 \\\\ &&&&&& 8700 & 3600 \\\\ \\hline \\end{tabular} \\end{minipage} \\end{table*} ", "conclusions": "We have studied the CaT and the Mgb stellar absorption indices and kinematics of the nuclei of a sample of ten active galaxies, centering our analysis on the study of the properties of Seyfert 1 galaxies. In spite of problems related to contamination by emission lines, we find that the IR CaII triplet stellar indices of Seyfert 1 nuclei with EW(H$\\beta_{broad}$) $\\la$ 45 \\AA, is stronger than what the standard model would predict from the observed strength of the Mgb index. This result is naturally explained by the presence, in the nuclei of these type 1 Seyferts of young stellar clusters whose luminosity is somehow related to the luminosity of the active nucleus itself. This conclusion is weaker for those nuclei with strong \\hbox{Fe\\,{\\sc ii}} emission affecting the measurement of the Mgb index. Our measurements of the velocity dispersions in Seyfert galaxies, support previous conclusions by TDT90 and Nelson \\& Whittle \\shortcite{NW96} that the main factor that controls the gas motions in the NLR of Seyfert galaxies is the mass of the bulge. We also find that other factors may also be important, since the correlation between the gas and stellar velocity dispersions shows a large scatter. Among the factors that may broaden the emission lines in the NLR, induced motions by shocks are the most likely. Some differences between the kinematics of Seyfert 1 and Seyfert 2 have been found. Seyfert galaxies with gas velocity dispersions much larger than that of the stars are preferentially found among type 2. Also we found some Seyfert 1 nuclei with emission lines narrower than the stellar absorption features. Although we have outlined some hypotheses, like orientation an evolutionary effects, a satisfactory explanation for this second aspect still needs to be found." }, "0005/astro-ph0005047_arXiv.txt": { "abstract": "We present the discovery of a candidate substellar companion from a survey of nearby, young stars with the NICMOS coronagraph on the Hubble Space Telescope. The H$\\approx$12 mag object was discovered approximately 4$\\arcsec$ from the young A0V star HR 7329. Using follow-up spectroscopy from STIS, we derive a spectral type between M7V and M8V with an effective temperature of $\\sim$ 2600 K. We estimate that the probability of a chance alignment with a foreground dwarf star of this nature is $\\sim 10^{-8}$ and therefore suggest the object (HR 7329B) is physically associated with HR 7329 with a projected separation of 200 AU. Current brown dwarf cooling models indicate a mass of less than 50 Jupiter masses for HR 7329B based on age estimates of $\\leq$ 30 Myr for HR7329A. ", "introduction": "The discovery of substellar objects in stellar systems is a key goal in contemporary astronomy, and an essential element in furthering our knowledge of the mass function of binary star and planetary system formation. The substellar mass range from 10 to 80 Jupiter masses (0.01 $-$ 0.08 M$_{\\odot}$) is crucial to our understanding of the bridge between the lowest mass stars and the giant planets. To this end, the Near Infrared Camera and Multi-Object Spectrometer (NICMOS) Instrument Definition Team (IDT) has conducted an infrared coronagraphic survey of young, main-sequence stars to search for substellar companions. Substellar objects cool with age because they do not sustain hydrogen fusion and are more difficult to detect with time as they become fainter (e.g. Burrows et al. 1997). Using independently determined ages and distances for the target stars, the masses of newly-detected secondaries can be ascertained from infrared fluxes and theoretical evolutionary tracks on the H$-$R diagram. Follow-up spectroscopy further constrains the effective temperature and probability of companionship. Here we present a spectrum obtained with the Space Telescope Imaging Spectrograph (STIS) of a substellar companion candidate, HR 7329B, from our NICMOS imaging survey. Previously, this survey revealed TWA5B, a $\\sim$ 20 Jupiter mass brown dwarf companion to TWA 5A (Lowrance et al. 1999). ", "conclusions": "\\subsection{Likelihood of Companionship} From its H-magnitude and M7.5V spectral type, HR 7329B can be either a background object, a foreground main-sequence M star, or a companion to HR 7329A. A main-sequence M7.5V star has M$_{H}$ $=$ 10.3 (Kirkpatrick \\& McCarthy 1994), so HR 7329B is too bright to be a background main-sequence star. If it were on the main-sequence, its photometric distance would be 19 parsecs. Henry (1991), in a volume limited infrared survey, finds six objects with M$_{H}$ $>$ 9.5 within five pc from the sun. If we assume a spherical distribution of low mass stars in the solar neighborhood, we can extrapolate the results within 5 pc to expect 1000 such objects out at 25 pc, so the a priori probability of finding one in projection within a 4$\\arcsec$ radius circle is $\\sim 10^{-7}$. Proper motion measurements of the companion and primary in the time between the NICMOS and STIS observations could be used to further constrain the probability of companionship. Unfortunately, the positional errors are too large. However, we can further constrain the probability that the object is not a foreground M dwarf. Searching the Tycho cataloque, we find that the mean proper motion of 1000 stars between 16 and 25 parsecs is 0.373$\\arcsec \\pm$ 0.277. Therefore, if we assume a gaussian distribution of proper motions about this mean, almost 80\\% of foreground stars have moved more than the half-width of the STIS slit (0.1$\\arcsec$)(taking into account angles along the slit), and would not be visible in the second epoch. Given these arguments, it is unlikely ($\\sim 10^{-8}$) HR 7329B is a foreground object and for the remainder of the paper, we assume it is physically associated with HR 7329A. \\subsection{Age of System} It is difficult to determine an age for A-type stars, but HR 7329 appears to be young ($<$ 40 Myr) based on its rotation, and more importantly, location on an H$-$R diagram. For massive stars, rotational velocities decline with age; HR 7329 has an especially large v$sini$ ($=$ 330 km/s) (Abt \\& Morrel 1995) which is considerably above the majority of A-type stars ($\\sim$ 100 km/s). Figure 3 reproduces the H$-$R diagram from Jura et al (1998) for A stars from the Yale Bright Star Catalog and overplots nearby, young clusters. There seem to be common areas of similar age stars; the 50-90 Myr IC2391 and Alpha Per clusters lie below the older (600 Myr) Hyades and Preasepe. There is a large scatter in the Pleaides (70-125 Myr), which could be due to a range of distances and ages as well as unresolved binaries. HR 7329 lies on a line located below the Alpha Per and IC 2391 cluster which intersects $\\beta$ Pic, HR 4796 and HD 141569. The latter stars have recently been assigned ages from their late-type companions of 20, 8, and 4 Myr respectively (Barrado Y Navascues et al. 1999; Stauffer, Hartmann, \\& Barrado Y Navascues 1995; and Weinberger et al. 2000). This suggests that HR 7329 is between 10 and 30 Myrs old. Finally, it has recently been suggested that HR 7329 is found within a young co-moving cluster much like the TW Hydrae Association with an age of $\\sim$40 Myr (Zuckerman \\& Webb 2000; Webb et al. 2000). \\placefigure{fig3} \\subsection{Effective Temperature and Bolometric Luminosity} An effective temperature of HR 7329B is required to position it on an H$-$R diagram, but the temperature scale for late, young M-dwarfs is uncertain (Allard et al. 1997). Luhman \\& Rieke (1998) extrapolate from Leggett et al.'s (1996) model fits to derive 2670 K and 2505 K for M7 V and M8 V respectively, which agrees with the newer models used by Leggett, Allard, \\& Hauschildt (1998) with an uncertainty of about 100 K. With this uncertainty for late M dwarf stars and the added uncertainty due to the spectral type, we plot the derived temperatures for each spectral class (Figure 4) and their associated uncertainty which overlaps and gives a possible range from 2405 K to 2770 K. The parallactic distance measured to HR 7329A by the Hipparcos mission is 47.67 $\\pm$ 1.6 pc. With a derived H magnitude of 11.93 for HR7329B, and a distance modulus of 3.39, we calculate an absolute H magnitude of 8.54 mag. There exists a number of bolometric corrections in the literature for M7 V and M8 V stars (Tinney et al. 1993; Kirkpatrick et al. 1993; Bessel, Castelli, \\& Plez 1998) based on I and K magnitudes. However, none give the BC in the H band. We have used the BC at the other bandpasses and the colors of late-type stars as a function of spectral type from Kirkpatrick \\& McCarthy (1994) to find a relationship between BC(H) and spectral type. For M7 and M8 we find a range of BC$_{H}$ from 2.54 to 2.78. Using a solar M$_{bol}$ of 4.75, we derive a luminosity for HR 7329B of 0.0026 $\\pm$ 0.0003 L${\\odot}$, with an uncertainty that includes the 0.5 spectral type range, bolometric correction, and distance errors. \\subsection{Derived Mass} We place HR 7329B on pre-main sequence evolutionary tracks (Baraffe et al. 1998) to infer a mass (Figure 4). Assuming only companionship, and therefore distance, indicates a mass of less than 50 \\mj\\ (less than 35\\mj\\ is not covered in Baraffe's models) and an age of less than 30 Myr. This supports the young age attributed to HR 7329A from its position on the H$-$R diagram, other youth indicators and possible membership in a young moving group. Evolutionary tracks from different authors do differ somewhat due to the different model atmospheres used. The tracks of D$'$Antona \\& Mazitelli (1997) indicate a mass range of 40 \\mj\\ or less for this luminosity and temperature. Burrows et al.'s (1997) models predict a 40 \\mj\\ brown dwarf will have an effective temperature of 2800K and a luminosity of 0.0023 L${\\odot}$ at an age of 22 Myr. \\placefigure{fig4} We present high signal-to-noise near-infrared photometry and optical spectroscopy of a probable companion (HR 7329B) at a projected distance of 200 AU from HR 7329 (A). We suggest the mass of B is less than 40 \\mj\\ . The derived age of less than 30 Myr for this companion supports the very young age of the primary A0V star indicated by its placement on the H$-$R diagram of nearby A-type stars. We do not detect any 1.6$\\micron$ scattered light from the far-infrared emitting dust seen by IRAS around HR 7329A. The HR 7329 system stands out from other binaries in that it has a very high mass ratio, q$\\sim$0.01. Zuckerman and Becklin (1992) found that around $\\sim$ 200 white dwarf stars whose progenitors are F and A main sequence stars, the percentage of systems with low-mass M star companions (M$\\sim$0.1M$_{\\odot}$, q$\\sim$0.06) was 5 to 10\\%, and the number of detected brown dwarfs was one (GD 165B), or $<$1\\%. The small percentage of white dwarfs with detectable brown dwarf companions is probably the result of the decline in brown dwarf luminosity with age. The discovery of the brown dwarf, HR 7329B, among a small sample of young A and F stars ($\\sim$10) observed by NICMOS suggests that the number of companion brown dwarfs and low mass stars may not be too different. In the field (Reid 1999), and the Pleiades cluster (Zapatero Osorio et al 1997), the relative number of low mass stars and brown dwarfs per log mass interval is also about equal, suggesting a flat inital-mass-function (IMF) for single stars. Clearly, greater statistics are needed before firm conclusions can be reached about the IMF of secondaries. We would like to thank M. Jura, C. Chen and J. Patience for their invaluable help and assistance. We thank the anonymous referee for comments which help to clarify the presentation. This work is supported in part by NASA grants NAG 5-4688 to UCLA and NAG 5-3042 to the University of Arizona NICMOS Instrument Definition Team. This paper is based on observations obtained under grant GO-8176.01-97A with the NASA/ESA Hubble Space Telescope at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc. under NASA contract NAS 5-26555." }, "0005/astro-ph0005579_arXiv.txt": { "abstract": "Observing ROSAT sources in an area \\mbox{$20\\degr \\times 25\\degr$} centered at the high latitude (\\mbox{$b = -59\\degr$}) active star \\objectname[]{ER\\,Eri}, we found evidences for a nearby association, that we call the Horologium Association (HorA), formed by at least 10 very young stars, some of them being bona fide Post-T\\,Tauri stars. We suggest other six stars as possible members of this proposed association. We examine several requirements that characterize a young stellar association. Although no one of them, isolated, gives an undisputed prove of the existence of the HorA, all together practically create a strong evidence for it. In fact, the Li line intensities are between those of the older classical T\\,Tauri stars and the ones of the Local Association stars. The space velocity components, of the HorA relative to the Sun (\\mbox{U= $-9.5 \\pm 1.0$}, \\mbox{V = $-20.9 \\pm 1.1$}, \\mbox{W = $-2.1 \\pm 1.9$}) are not far from those of the Local Association, so that it could be one of its last episodes of star formation. In this region of the sky there are some hotter and non-X-ray active stars, with similar space velocities, that could be the massive members of the HorA, among them, the nearby Be star \\objectname[]{Achernar}. The maximum of the mass distribution function of its probable members is around \\mbox{0.7 -- 0.9 $M_\\sun$}. We estimate its distance as $\\sim$60\\,pc and its size as $\\sim$50\\,pc. If spherical, this size would be larger than the surveyed area and many other members could have been missed. We also observed 3 control regions, two at northern and southern galactic latitudes and a third one in the known \\objectname[]{TW\\,Hya} Association (TWA), and the properties and distribution of their young stars strengthen the reality of the HorA. Contrary to the TWA, the only known binaries in the HorA are 2 very wide systems. The HorA is much more isolated from clouds and older ($\\sim$30\\,Myr) than the TWA and could give some clues about the lifetime of the disks around T\\,Tauri stars. Actually, none of the proposed members is an IRAS source indicating an advanced stage of the evolution of their primitive accreting disks. \\objectname[]{ER\\,Eri} itself was found to be a RS\\,CVn-like system. ", "introduction": "Surveys of young pre-main-sequence (PMS) stars based on IRAS colors detect mainly the classical T\\,Tauri stars (CTT), with ages less than 10\\,Myr, due to their important dusty accretion disks. Among them, one of the most comprehensive was the ``Pico dos Dias Survey'' (PDS) \\citep{12, 38, 45}. The weak T\\,Tauri stars (WTT) are mainly detected by the X-rays produced in their active coronae, since their disks in general contain no more sufficient dust. Finally, there should exist even more evolved young stars, with masses smaller or comparable to the Sun and ages between about 10 to 70\\,Myr - the so-called Post-T\\,Tauri stars (PTT). If star formation were constant during the last 100\\,Myr, a large number of uniformly distributed PTT should exist in the solar neighborhood. Nevertheless, the first searches for PTT resulted in very few objects \\citep{14}. This lack of discoveries may indicate that star formation was not constant in time. However, as the disks around PTT are even more evolved than those around WTT, they should have been dissipated or agglutinated into planetesimals and they hardly would be detected in surveys based on the far infrared or even on H$_\\alpha$ emission, as it can be very weak in PTT. The detection of X-ray sources by the ROSAT All-Sky Survey (RASS) \\citep{30} outside important star formation regions gave some hope that these are the expected population of PTT. Nevertheless, several considerations cast doubts about the PTT nature of those dispersed objects \\citep{3, 32}, indicating that most of them could be active main sequence (MS) stars. \\citet{20} suggests that most of them are connected with the Local Association. On the other hand, \\citet{10} found that some of those objects could be PMS stars ejected from forming clouds. She also notes that the WTT around the Chamaleon region appear to be formed in small short-lived cloudlets \\citep{11}. In any case, this large population of X-ray sources around star forming clouds seems to be represented by a mixture of PMS and young MS objects. Do genuine PTT exist and where are they? The main point of this work is that PTT can be found in physically characterized dispersed groups, with ages larger than the mean lifetime of the original clouds. In such case they may be located far from any cloud. In fact, we discovered the first of this kind of association when we were searching for new T\\,Tauri stars among IRAS point sources in a five degree radius around \\objectname[]{TW\\,Hya}, an already known isolated CTT at high galactic latitude \\mbox{($b$ = $23\\degr$)} \\citep{8, 12}. Hipparcos parallaxes show that this is a nearby association, the distance of \\objectname[]{TW\\,Hya} being \\mbox{$56.4 \\pm\\,7$ pc} \\citep{40}. Several research groups became interested in this association and other members were found [see. e.g., \\citet{15, 19, 39, 18}]. Today nearly 13 young systems totalling 21 stars and a brown dwarf are known to belong to the \\objectname[]{TW\\,Hya} association (TWA) \\citep{39, 36} and this constitutes a sufficiently large density of PMS stars in a localized region of the sky to consider it as a real cluster. Futhermore, the stars appear to have a common origin \\citep{24, 39} and the best age estimated for the association is $\\sim$10\\,Myr \\citep{35, 39}. Other groups of WTT, apparently isolated from clouds, have been detected at larger distances, as the \\objectname[]{$\\eta$\\,Cha} cluster at 100\\,pc \\citep{27} or the one in front of the translucent clouds MBM\\,7 and MBM\\,55 \\citep{13}. The former appears to be more clearly defined, beeing nearly 40 times more compact than the TWA and seems to be related to the Sco\\,-\\,Cen\\,OB association. Studies of this kind of clusters are important to understand the beginning of dispersion of regions containing also massive stars. We propose in this paper the existence of a new nearby association around \\objectname[]{ER\\,Eri} (\\objectname[]{PDS\\,1} = \\objectname[]{Hen\\,1}), older and more isolated from clouds than the TWA. ", "conclusions": "Exploring a region of about \\mbox{$20\\degr \\times 25\\degr$} around the high galactic latitude \\mbox{($b = -59\\degr$)} active star \\objectname[]{ER\\,Eri}, previously classified as a WTT \\citep{12, 45}, we found evidences of a new association, the ``Horologium Association'', formed mainly by bona fide low mass \\mbox{($\\sim$0.8 {\\it M}$_\\sun$)} PTT. Its probable and possible members are marked with asterisks in Table\\,1 and are discriminated in Table\\,5. Possible hot members are presented in Table\\,7. \\objectname[]{ER\\,Eri} itself was found to be a background RS\\,CVn-like system. Since we found no low mass PTT in the two control areas, as can be seen in Figure\\,5, we believe that HorA-like stellar groups are not numerous in the solar vicinity, indicating a non constant rate of star formation in the last 100\\,Myr. This new association is presently represented by at least 10 members (Table\\,5), having an age of $\\sim$30\\,Myr and is older than the isolated TWA, with an age of approximately 10\\,Myr. Until now there are no detected binaries in the HorA (stars ERX\\,22N and ERX\\,22S have a separation of 2380\\,AU and should be considered as common proper motion stars). If confirmed, this marked difference with the TWA must be due to different intrinsic conditions of star formation of these two associations. The HorA is even more isolated from clouds than the TWA and, in any case, we can expect that its original cloud could have dissipated in less than 30\\,Myr. The distances of the stars in the HorA cover an interval from nearly 40 to about 90\\,pc (at a mean distance of $\\sim$60\\,pc) giving a diameter of $\\sim$50\\,pc, compatible with the size produced after 30\\,Myr by an initial velocity dispersion of \\mbox{$\\sim$1.8 km s$^{-1}$}. If this size is the same in angular extent, it surpasses the surveyed region and could contain many other members, such as the interesting young star \\objectname[]{AB\\,Dor}. In the surveyed region there may be hotter members, that are not X-ray sources, (for example, the nearby Be star \\objectname[]{Achernar}) having similar ages, distances and space motions. If this is the case, the HorA could be the remnant of an old OB association, but, due to the kinematical errors, there is little hope to localize its birth place. The space velocity components of the HorA are near but distinct from those of the Local Association. In fact, in all surveyed regions we found young field stars, possible members of the Local Association, with a compatible Li depletion. But, for its well defined kinematics, physical properties, Li abundances and restricted location in the sky, the HorA can be distinguished from these stars, resembling more the general behavior of the TWA. Considering the observational limitations, as that of the magnitude limit of the Hipparcos measurements, we are aware of the challenge that represents the detection of a coeval moving group of stars with an age around 30\\,Myrs. To arrive at this, we examine several requirements that characterize a young stellar association. If no one of them, isolated, is completely sufficient, none of them are mutually contradictory and all together practically create the necessary condition for the establishment of the HorA as a real nearby association. Following \\citet{5} and considering that large moving groups, more commonly known as Eggen's superclusters, are not real clusters but are formed by a chance coincidence of smaller coeval streams, we can conceive the HorA as one small, coeval structure in the Pleiades supercluster. But its overall characteristics are distinct from those of the young Pleiades subgroups found by \\citet{48}. In any case, the HorA may be one of the last episodes of star formation of the Local Association and could be useful to understand better its fine structure." }, "0005/astro-ph0005115_arXiv.txt": { "abstract": "We show how two simple modifications to the standard cosmic string scenario for structure formation compare to the recently released Boomerang data set. Namely we consider pure string closed models and mixed models where both inflation and strings are responsible for the perturbations. In the closed models we find that pure string models would require a universe with roughly $\\Omega_M=0.8$, $\\Omega_{\\Lambda}=1.6$ to agree with the peak position revealed by the data and in agreement with the SNIa data. In the hybrid scenario with local strings we find that we require even more tilt and baryon content ($n_s\\sim 0.8$, $\\Omega_b=0.08$, $H_0=70$) to match the data than with pure inflation models. The case with global strings fares better with a standard period of $\\Lambda$CDM inflation and a $\\sim30\\%$ contribution from strings being in good agreement with the data. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005609_arXiv.txt": { "abstract": "{ We present Ulysses and NEAR data from the detection of the short or intermediate duration (2 s) gamma-ray burst \\object{GRB~000301C} (2000 March 1.41 UT). The gamma-ray burst (GRB) was localised by the Inter Planetary Network (IPN) and RXTE to an area of $\\sim$50 arcmin$^2$. A fading optical counterpart was subsequently discovered with the Nordic Optical Telescope (NOT) about 42~h after the burst. The GRB lies at the border between the long-soft and the short-hard classes of GRBs. If \\object{GRB~000301C} belongs to the latter class, this would be the first detection of an afterglow to a short-hard burst. We present UBRI photometry from the time of the discovery until 11 days after the burst. We also present JHK photometry obtained with UKIRT on 2000 March 4.5 UT (3.1 days after the burst). Finally, we present spectroscopic observations of the optical afterglow obtained with the ESO VLT Antu telescope 4 and 5 days after the burst. The optical light curve is consistent with being achromatic from 2 to 11 days after the burst and exhibits a break. A broken power-law fit yields a shallow pre-break decay power-law slope of $\\alpha_1=-0.72\\pm$0.06, a break time of $t_{\\rm break}=4.39\\pm$0.26 days after the burst, and a post-break slope of $\\alpha_2=-2.29\\pm$0.17. These properties of the light curve are best explained by a sideways expanding jet in an ambient medium of constant mean density. In the optical spectrum we find absorption features that are consistent with \\ion{Fe}{ii}, \\ion{C}{iv}, \\ion{C}{ii}, \\ion{Si}{ii} and Ly$\\alpha$ at a redshift of 2.0404$\\pm$0.0008. We find evidence for a curved shape of the spectral energy distribution of the observed afterglow. It is best fitted with a power-law spectral distribution with index $\\beta\\sim -0.7$ reddened by an \\object{SMC}-like extinction law with A$_{V}\\sim 0.1$ mag. Based on the Ly$\\alpha$ absorption line we estimate the \\ion{H}{i} column density to be $\\log$(N(\\ion{H}{i}))$=21.2\\pm0.5$. This is the first direct indication of a connection between GRB host galaxies and Damped Ly$\\alpha$ Absorbers. ", "introduction": "The discovery of the first $X$-ray afterglow (Costa et~al.~\\cite{C1997}) and optical counterpart (van~Paradijs et~al.~\\cite{PGG1997}) to a long-duration gamma-ray burst (GRB) have led to a revolution in GRB research. The determination of a redshift of 0.835 for \\object{GRB~970508} (Metzger et~al.~\\cite{MDK1997}), and the subsequent determination of redshifts of 13 bursts with a median redshift of $\\sim$1.0, have firmly established their cosmological origin (Kulkarni et al.~\\cite{KBB2000}; This work; Bloom et al.~\\cite{BDD2000}). The intriguing case of an association of the peculiar supernova \\object{SN1998bw} with \\object{GRB~980425} (Galama et~al.~\\cite{GVP1998}) was the first indication of a possible connection with supernovae. Evidence for supernova signatures in the late light curves of \\object{GRB~970228} (Reichart~\\cite{Rei1999}; Galama~et~al.~\\cite{GTV1999}) and \\object{GRB~980326} (Castro-Tirado \\& Gorosabel~\\cite{CTG1999}; Bloom~et~al.~\\cite{BKD1999}) suggests that at least some long-duration GRBs may be related to the collapse of massive ($>25$ M$_\\odot$) stars. Breaks in the power-law declines of \\object{GRB~990123} (Kulkarni~et~al.~\\cite{KDO1999}) and \\object{GRB~990510} (Harrison~et~al.~\\cite{HBF1999}) are interpreted as evidence for collimated outflows (`jets') (see also Holland et al.~\\cite{HBH2000}). Further evidence for this collapsar + jet model (e.g., MacFadyen \\& Woosley~\\cite{MW1999}) comes from the light curve of \\object{GRB~980519} which is best interpreted as a jet expanding into a preexisting circumburst stellar wind (Jaunsen et~al.~\\cite{J2001}). The high-energy properties of GRBs show a bi-modal distribution of burst durations (Kouveliotou et~al.~\\cite{K1995}) which, in the simplest scenario, may indicate the existence of binary compact mergers as the progenitors of the short-duration bursts (T$_{90}<2$ s). From an analysis of the Third BATSE Catalog, Mukherjee et al.~(\\cite{MFB1998}) have shown that, in addition to the short (T$_{90}<2$ s) and long (T$_{90}>5$ s) classes, there may exist a third, intermediate soft-spectrum class of GRBs with duration 2 $\\mathrm{s} < \\mathrm{T}_{90} < 5$ s. In this paper we report the discovery and subsequent observations and analysis of the afterglow of the short-to-intermediate duration \\object{GRB~000301C} (Fynbo et~al.~\\cite{FJH2000a}). Sect.~\\ref{SECTION:detection} reports the detection, IPN localisation and the high-energy data of the GRB obtained from Ulysses and NEAR\\@. Sect.~\\ref{SECTION:discovery} describes the discovery of the optical counterpart and our subsequent optical and infrared observations. Sect.~\\ref{SECTION:photometry} details the optical and infrared photometry and Sect.~\\ref{SECTION:Spectroscopy} describes the VLT spectroscopy. Sect.~\\ref{SECTION:Results} describes the results obtained on the spectroscopy and spectral energy distribution and Sect.~\\ref{SECTION:discussion} is devoted to the discussion and interpretation, with Sect.~\\ref{SECTION:conclusion} presenting our conclusions. Throughout this paper, we adopt a Hubble constant of H$_0$ = 65 km s$^{-1}$ Mpc$^{-1}$ and assume $\\Omega_{\\mathrm{m}}=0.3$ and $\\Omega_{\\Lambda}=0.7$. ", "conclusions": "\\label{SECTION:conclusion} \\object{GRB~000301C} is so far the GRB of shortest duration, for which a counterpart has been detected. The high-energy properties of the burst are consistent with membership of the short-duration class of GRBs, though \\object{GRB~000301C} could belong to the proposed intermediate class of GRBs or the extreme short end of the distribution of long-duration GRBs. Our VLT-spectra show that \\object{GRB~000301C} occurred at a redshift of 2.0404$\\pm$0.0008. The light curve of the optical transient is well-fitted by a broken power-law and it is consistent with being achromatic. From the light-curve properties we find that the best model for \\object{GRB~000301C} is that of a sideways expanding jet in an ambient medium of constant density. This interpretation is further supported by the achromatic light-curve evolution, and by the agreement between the theoretically predicted and observationally derived spectral indices. The spectral energy distribution at March 4.5 reveals \\object{SMC}-like extinction in the host galaxy at a level of $\\mathrm{A}_V < 0.10$, which is significantly lower than for the strongly star-forming LBGs. Hence, the extreme faintness of the host galaxy indicates a low overall star-formation rate in the host galaxy, raising the possibility that the host may be a chemically less evolved, relatively low-luminosity galaxy containing SMC-type dust. We argue that there may be a connection between the host galaxy of \\object{GRB~000301C} and DLAs, suggesting that substantial star-forming activity at high redshift takes place in relatively faint galaxies. Future studies of high redshift GRBs will further help explore this connection." }, "0005/astro-ph0005323_arXiv.txt": { "abstract": "We study the distribution functions of mass and circular velocity for dark matter halos in N-body simulations of the $\\Lambda$CDM cosmology, addressing redshift and environmental dependence. The dynamical range enables us to resolve subhalos and distinguish them from ``distinct\" halos. The mass function is compared to analytic models, and is used to derive the more observationally relevant circular velocity function. The distribution functions in the velocity range 100--500$\\kms$ are well fit by a power-law with two parameters, slope and amplitude. We present the parameter dependence on redshift and provide useful fitting formulae. The amplitudes of the mass functions decrease with $z$, but, contrary to naive expectation, the comoving density of halos of a fixed velocity $\\sim 200\\kms$ actually increases out to $z\\sim 5$. This is because high-$z$ halos are denser, so a fixed velocity corresponds to a smaller mass. The slope of the velocity function at $z=0$ is as steep as $\\sim -4$, and the mass and velocity functions of distinct halos steepen with increasing $z$, while the functions of subhalos do not steepen with $z$, and become even flatter at $z>2$. A simple observable prediction is that the slope of the velocity function of isolated galaxies is steeper than that of galaxies in groups by as much as unity, reflecting the density biasing of high-velocity halos. We confirm that the Press-Schechter approximation typically overestimates the halo mass function by a factor of $\\sim 2$, while modified approximations provide improved predictions. ", "introduction": "\\label{sec:intro} Understanding the evolution of structure in the universe and its relation to initial conditions is a fundamental issue in cosmology. Galaxy formation involves physical processes such as star formation, supernovae feedback, and dust extinction which we still are far from understanding in full. On the other hand, the clustering of the collisionless dark-matter (DM) component is well understood. DM particles cluster into halos, which subsequently accrete more mass either by gradual infall or by merging with other halos. An accurate description of the distribution of DM halo mass (the mass function) and its temporal evolution are crucial for our understanding of the evolution of the luminous galaxies that reside in these halos. A popular approximation to the mass function is provided by the Press-Schechter formalism (Press \\& Schechter 1974, hereafter PS). The initial density fluctuation field, smoothed on some comoving scale $R$, is assumed to evolve according to linear theory until it reaches a critical value, $\\delta_c=1.69$, at which time it is assumed to have collapsed and virialized. The critical value is determined by a linear extrapolation to the collapse time as set by the spherical collapse model. Many of the simplifying assumptions behind the PS formalism do not seem to be physically motivated, yet it turns out to predict with reasonable accuracy the mass function as seen in simulations; the neglected effects tend to cancel each other (Monaco 1998). While being very useful for qualitative and semi-quantitative analyses, the accuracy of the PS approximation may not be enough when quantitative details are concerned. For example, it has been found based on simulations that for halo masses below $M=10^{13} - 10^{14}\\hmsun$ (depending on the specific cosmological model) the PS approximation overestimates the mass function by as much as a factor of $\\sim 2$ (Gross \\etal 1998; Lee \\& Shandarin 1999; Somerville \\etal 2000). At the high-mass end, $M \\gsim 10^{15} \\hmsun$, the PS formalism underestimates the abundance of DM halos, especially at high redshift (\\eg, Governato \\etal 1999, Somerville \\etal 2000). Improved approximations have been proposed in order to correct these inaccuracies, \\eg, a modification of PS by Sheth and Tormen (1999, hereafter ST), practically replacing the spherical collapse model with an ellipsoidal collapse model (Sheth, Mo, \\& Tormen 2000). Lee \\& Shandarin (1998, hereafter LS) also use a non-spherical approach, based on the Zeldovich approximation (Zeldovich 1970), to improve the PS approximation. The need for a more accurate description of the mass function is further highlighted by the development of a useful approach for studying galaxy formation and evolution based on semi-analytical models (Kauffmann \\etal 1999; Somerville \\& Primack 1999; Baugh \\etal 1999). Semi-analytical models try to circumvent the complications associated with baryonic processes by simplified prescriptions and artificially embedding galaxies within DM halos, which allows them to subsequently predict observable statistical quantities such as the galaxy luminosity function or the Tully-Fisher relation. Some of the important processes governing the evolution of galaxies in semi-analytical models are directly related to the DM clustering properties; for example, the mass function and merger histories of the halo populations are key ingredients in semi-analytical models (Kauffmann \\etal 1999; Somerville \\& Kolatt 1999). These need to be followed with a better accuracy than provided by the PS approximation. Furthermore, the PS formalism does not address substructure within halos. When two halos merge, the PS approximation immediately labels them as one virialized halo. However, high-resolution simulations (e.g., Klypin \\etal 1999a, Ghigna et al. 1998) show that substructure does maintain its identity after merging into larger halos. This means that galactic halos are expected to survive tidal stripping for some time, with possible important consequences for collisions and starbursts (Kolatt \\etal 1999, 2000). An alternative to modeling the complicated physics needed to predict galaxy luminosities in individual halos is to calculate statistical properties such as the distribution function of the circular velocity of halos --- the velocity function. By using observed luminosity-velocity relations, one can then relate the predicted velocity function to observational luminosity functions (Gonzales \\etal 2000; Bullock \\etal 2000b) and address halo number counts directly. In addition, because modeling luminosities of the high-redshift galaxies is even more uncertain than modeling the local population, the redshift evolution of the halo velocity function provides a much-needed, and more direct handle on the galaxy population than the corresponding mass function. In order to obtain the velocity function we use a high-resolution simulation which has the force resolution and mass resolution necessary to determine the maximum circular velocities of halos, and is capable of resolving substructure within halos (Kravtsov, Klypin \\& Khokhlov 1997). The simulation is complemented by a halo finding algorithm which also classifies halos into nesting levels of halos within halos (Bullock \\etal 2000a; Bullock 1999; see \\S\\ref{sec:method}). Our halo finder/classifier (HFC) fits each halo by an NFW density profile (Navarro, Frenk \\& White 1996) and automatically assigns a circular velocity profile to each halo. For the purpose of comparing to observations, some of the halos are assumed to be galactic halos based on a simple prescription following basic observational constraints. In \\S\\ref{sec:method} we present our method for extracting the mass (\\S\\ref{sec:massf}) and velocity (\\S\\ref{sec:velf}) functions for the different populations of halos (\\eg, subhalos that reside in massive hosts, halos of galaxies in groups, etc.), while correcting for incompleteness in the halo-finding algorithm by an iterative procedure. In \\S\\ref{sec:gal_id} we present a simple scheme for identifying ``galactic'' halos. In \\S\\ref{sec:results} we compare results for distinct halos to the PS, ST, and LS predictions, and extend our investigation to the mass and velocity functions of substructure as well. The results for halos and subhalos at $z=0$ are discussed in \\S\\ref{sec:z0}, and their redshift evolution is presented in \\S\\ref{sec:evolution}. Results for galactic halos in different environments are outlined in \\S\\ref{sec:gal_res}. We discuss our results and compare them to other studies in \\S\\ref{sec:conc}. ", "conclusions": "\\label{sec:conc} We have studied the distributions of mass and maximum circular velocity of DM halos in a cosmological simulation of the $\\Lambda$CDM model. The mass function is useful for semi-analytic modeling of galaxy formation, while the velocity function can be confronted with observations once a relation between halo velocity and disk rotation velocity is assumed. The high resolution allows us to address distinct halos as well as subhalos, at a range of redshifts, and to distinguish between field and grouped galactic halos. We find that the Schechter function, and in particular its power-law behavior over a wide range below the characteristic mass $M_*$, is a good fit to the different halo distributions of mass and velocity, over the whole range studied in this simulation, and at all times monitored. This is once the incompleteness of the halo finder at small halos and the associated errors are properly accounted for. Each of these distribution functions is therefore characterized by a slope and an amplitude. The characteristic large scale where the function bends is not constrained properly because we do not sample enough large halos. At the low end, we do not attempt to recover the velocity function below $100\\kms$, and therefore cannot address possible deviations from a power-law at low velocities (Klypin \\etal 1999b; Moore \\etal 1999). At $z=0$, for distinct halos, we find mass and velocity function slopes of $\\alpha=-1.85\\pm0.4$ and $\\beta=-3.82\\pm0.03$ in \\equ{schechter_m} and \\equ{vfitbeta} respectively. We measure the time evolution of the mass and velocity functions via the redshift dependence of the slope and amplitude parameters. The $z$ dependences are described for convenience by simple functional fits out to $z=5$ [Figures~\\ref{fig:massf_redshift} and \\ref{fig:velf_redshift}], to be used in semi-analytic models of galaxy formation. The slope of the velocity function for distinct halos steepens with redshift like $\\beta=-3.9 -2.0\\tz +1.4\\tz^2 \\pm0.2$, where $\\tz\\equiv \\log(1+z)$. A particular prediction that should be tested observationally is that the amplitude of the velocity function at fixed $\\vm =300\\kms$ hardly varies with redshift, for all types of halos. This implies that the number density of halos with $\\vm \\sim 200\\kms$ and below is actually predicted to {\\it increase} slightly with increasing redshift, as opposed to the naive expectation based on the way the mass function evolves. This prediction for the weak evolution of the velocity function can also be tested indirectly. For example, by combining this result with the observed evolution of the luminosity function of spiral galaxies, one can deduce the way the Tully-Fisher relation might evolve with redshift (cf. Gonzalez et al. 2000; Bullock \\etal 2000b). To make this connection between the luminosity function and the velocity function using the Tully-Fisher relation, it will also be necessary to take into account the effects of baryonic infall. But this is important, since an inconsistency between this prediction and direct observations of Tully-Fisher evolution may question the validity of the $\\Lambda$CDM model. Most earlier cosmological simulations, and analytic approximations such as Press - Schechter, overlook substructure within halos, which is bound to play an important role in galaxy formation. Our simulations and halo finder enable a study of the hierarchical halo population within a cosmological volume. We find that the subhalo distributions can also be fitted by power laws, though the errors are larger because there is an order of magnitude fewer subhalos than distinct halos. At redshifts of order 2 and beyond, we find that the subhalo population evolves differently from the distinct halo population. While the slope of the distinct-halo functions becomes steeper with increasing redshift, the slope of the subhalo becomes flatter at higher redshifts. On the other hand, the amplitudes of the subhalo and distinct-halo functions decrease with increasing redshift in a similar way. Several other simulations were used to study the evolution of substructure within individual clusters of galaxies (\\eg, Ghigna \\etal 1998, van den Bosch \\etal 1999, Sensui \\etal 2000, Okamoto \\& Habe 1999). The results of Okamoto \\& Habe (1999), in particular, are quite similar to what we find. They see very little evolution of the subhalo mass function in the range $03$, but the discrepancies between the simulation and the ST and LS predictions are much smaller than the deviations of the PS formalism. All three approximations predict a higher rate of evolution for the mass function than seen in the simulations." }, "0005/astro-ph0005445_arXiv.txt": { "abstract": "We investigate the dynamical importance of a newly recognized possible source of significant feedback generated during structure formation; namely cosmic ray (CR) pressure. We present evidence for the existence of numerous shocks in the hot gas of galaxy clusters (GCs). We employ for the first time an explicit numerical treatment of CR acceleration and transport in hydro simulations of structure formation. According to our results, CRs provide an important fraction of the total pressure inside GCs, up to several tenths. This was true even at high redshift (z=2), meaning that such non-thermal component could affect the evolution of structure formation. ", "introduction": "During the hierarchical process of structure formation, supersonic gas infall and merging events invariably generate powerful, large and long-lived shock waves (Miniati et al. 1999). These should produce copious amounts of CRs, by way of diffusive shock acceleration (e.g. Blandford \\& Ostriker 1978), including both electrons and ions. In addition, the post-shock gas and diffusively trapped CRs are mostly advected into non-expanding regions, such as filaments and clusters. It turns out that the energy of most of the CR-protons is only marginally affected by radiative losses during a Hubble time. The important possibility, then, is that the latter might accumulate inside forming structures, storing up a substantial fraction of the total pressure there. In addition to cosmic shocks other sources of CRs are also possible. These include AGNs, SNe and stellar winds all of which are candidates for important contributions to the total population of CRs in cosmic structures, although they are not discussed here. There is growing observational evidence that significant non-thermal activity takes place in GCs. This evidence is provided by extended sources of polarized radio emission, interpreted as synchrotron radiation from relativistic electrons (e.g. Hanisch 1984; Deiss et al. 1997); and by the detection of radiation in excess to what is expected from the hot, thermal X-ray emitting Intra Cluster Medium (ICM), both in the extreme ultra-violet (e.g. Lieu et al. 1996; Kaastra 1998) and in the hard X-ray band above $\\sim 10$ KeV (e.g. Fusco-Femiano et al. 1999; Valinia et al. 1999). Although a coherent picture of the non-thermal status of the ICM is still lacking, a very plausible origin for these radiation excesses is inverse-Compton (IC) due to relativistic electrons (e.g. Sarazin \\& Lieu 1998). Based on this assumption and on the measured EUV excess in Coma cluster, Lieu et al. (1999) have estimated the existence of a CR proton component in approximate {\\it equipartition} of energy with the thermal gas. ", "conclusions": "We have shown that the ICM of GCs is commonly populated by numerous internal flow shocks with similar characteristic to, but not necessarily associated with major merger events. These along with accretion shocks and merger shocks are likely to play an important role for the non-thermal activity of the ICM. We have also shown that CR pressure could provide a substantial fraction of the total pressure in GCs today, thus affecting GC mass estimates based on the hydrostatic equilibrium assumption and in turn, the baryonic fraction estimates (which end up being biased high). We have also shown that CR pressure was significant already at high $z$, therefore possibly affecting the evolution of structure formation. Since this is often used as a tools for discriminating among different cosmological models (e.g. Carlberg et al. 1997; Bahcall \\& Fan 1998), the role of CR pressure should be well understood in order to apply evolutionary arguments with confidence." }, "0005/astro-ph0005390_arXiv.txt": { "abstract": "We extend the formalism for the calculation of the relativistic corrections to the Sunyaev-Zel'dovich effect for clusters of galaxies and include the multiple scattering effects. We present a systematic method for the inclusion of the multiple scattering effects. The multiple scattering contribution is found to be very small compared with the single scattering contribution. For high-temperature galaxy clusters of $k_{B} T_{e} \\approx 15$keV, the ratio of the both contributions is $-0.3\\%$ in the Wien region. In the Rayleigh--Jeans region the ratio is $-0.03\\%$. Therefore the multiple scattering contribution is safely neglected for the observed galaxy clusters. ", "introduction": "Compton scattering of the cosmic microwave background (CMB) radiation by hot intracluster gas --- the Sunyaev-Zel'dovich effect (Zel'dovich \\& Sunyaev 1969; Sunyaev \\& Zel'dovich 1972, 1980a, 1980b, 1981) --- provides a useful method to measure the Hubble constant $H_{0}$ (Gunn 1978; Silk \\& White 1978; Birkinshaw 1979; Cavaliere, Danese, \\& De Zotti 1979; Birkinshaw, Hughes, \\& Arnaud 1991; Birkinshaw \\& Hughes 1994; Myers et al. 1995; Herbig et al. 1995; Jones 1995; Markevitch et al. 1996; Holzapfel et al. 1997; Furuzawa et al. 1998; Komatsu et al. 1999). The original Sunyaev-Zel'dovich formula has been derived from a kinetic equation for the photon distribution function taking into account the Compton scattering by electrons: the Kompaneets equation (Kompaneets 1957; Weymann 1965). The original Kompaneets equation has been derived with a nonrelativistic approximation for the electron. However, recent X-ray observations have revealed the existence of many high-temperature galaxy clusters (David et al. 1993; Arnaud et al. 1994; Markevitch et al. 1994; Markevitch et al. 1996; Holzapfel et al. 1997; Mushotzky \\& Scharf 1997; Markevitch 1998). In particular, Tucker et al. (1998) reported the discovery of a galaxy cluster with the electron temperature $k_{B} T_{e} = 17.4 \\pm 2.5$ keV. Rephaeli and his collaborator (Rephaeli 1995; Rephaeli \\& Yankovitch 1997) have emphasized the need to take into account the relativistic corrections to the Sunyaev-Zel'dovich effect for clusters of galaxies. In recent years remarkable progress has been achieved in the theoretical studies of the relativistic corrections to the Sunyaev-Zel'dovich effects for clusters of galaxies. Stebbins (1997) generalized the Kompaneets equation. Itoh, Kohyama, \\& Nozawa (1998) have adopted a relativistically covariant formalism to describe the Compton scattering process (Berestetskii, Lifshitz, \\& Pitaevskii 1982; Buchler \\& Yueh 1976), thereby obtaining higher-order relativistic corrections to the thermal Sunyaev-Zel'dovich effect in the form of the Fokker-Planck expansion. In their derivation, the scheme to conserve the photon number at every stage of the expansion which has been proposed by Challinor \\& Lasenby (1998) played an essential role. The results of Challinor \\& Lasenby (1998) are in agreement with those of Itoh, Kohyama, \\& Nozawa (1998). The latter results include higher-order expansions. Itoh, Kohyama, \\& Nozawa (1998) have also calculated the collision integral of the Boltzmann equation numerically and have compared the results with those obtained by the Fokker-Planck expansion method. They have confirmed that the Fokker-Planck expansion method gives an excellent result for $k_{B}T_{e} \\leq 15$keV, where $T_{e}$ is the electron temperature. For $k_{B}T_{e} \\geq 15$keV, however, the Fokker-Planck expansion results show nonnegligible deviations from the results obtained by the numerical integration of the collision term of the Boltzmann equation. Nozawa, Itoh, \\& Kohyama (1998b) have extended their method to the case where the galaxy cluster is moving with a peculiar velocity with respect to CMB. They have thereby obtained the relativistic corrections to the kinematical Sunyaev-Zel'dovich effect. Challinor \\& Lasenby (1999) have confirmed the correctness of the result obtained by Nozawa, Itoh, \\& Kohyama (1998b). Sazonov \\& Sunyaev (1998a, b) have calculated the kinematical Sunyaev-Zel'dovich effect by a different method. Their results are in agreement with those of Nozawa, Itoh, \\& Kohyama (1998b). The latter authors have given the results of the higher-order expansions. Itoh, Nozawa, \\& Kohyama (2000) have also applied their method to the calculation of the relativistic corrections to the polarization Sunyaev-Zel'dovich effect (Sunyaev \\& Zel'dovich 1980b, 1981). They have thereby confirmed the result of Challinor, Ford, \\& Lasenby (1999) which has been obtained with a completely different method. Recent works on the polarization Sunyaev-Zel'dovich effect include Audit \\& Simons (1999), Hansen \\& Lilje (1999), and Sazonov \\& Sunyaev (1999). As stated above, Itoh, Kohyama, \\& Nozawa (1998) have carried out the numerical integration of the collision term of the Boltzmann equation. This method produces the exact results without the power series expansion approximation. Sazonov \\& Sunyaev (1998a, b) have reported the results of the Monte Carlo calculations on the relativistic corrections to the Sunyaev-Zel'dovich effect. In Sazonov \\& Sunyaev (1998b), a numerical table which summarizes the results of the Monte Carlo calculations has been presented. This table is of great value when one wishes to calculate the relativistic corrections to the Sunyaev-Zel'dovich effect for galaxy clusters of extremely high temperatures. Accurate analytic fitting formulae would be still more convenient to use for the observers who wish to analyze the galaxy clusters with extremely high temperatures. For this purpose, Nozawa et al. (2000) have presented an accurate analytical fitting formula of 0.1\\% accuracy for the numerical results for the relativistic corrections to the thermal Sunyaev-Zel'dovich effect for clusters of galaxies. For the analyses of the galaxy clusters with extremely high temperatures, the results of the calculation of the relativistic thermal bremsstrahlung Gaunt factor (Nozawa, Itoh, \\& Kohyama 1998a) and their accurate analytic fitting formulae (Itoh et al. 2000) will be useful. In this series of papers devoted to the study of the relativisitc corrections to the Sunyaev-Zel'dovich effect for clusters of galaxies, we have so far restricted ourselves to the case of single Compton scattering. This is justified because the optical depth for the Compton scattering of the CMB photon inside the galaxy clusters is generally about $10^{-2}$ or smaller (Birkinshaw 1999). Nevertheless, it would be desirable to evaluate the effects of the multiple Compton scattering of the CMB photon inside the galaxy clusters accurately, as we have already developed the method to calculate the relativistic corrections to the Sunyaev-Zel'dovich effect for the galaxy clusters with high accuracy. The multiple scattering effects have been already considered by many authors (see Birkinshaw 1999 for references). In this paper we wish to evaluate the multiple scattering effects in the same theoretical framework of this series of papers. The present paper is organized as follows. In $\\S$ 2 we give the method of the calculation and the results. In $\\S$ 3 we give discussion of the results and concluding remarks. ", "conclusions": "From the results presented in the previous section it is clear that the multiple scattering contribution $\\Delta I_{2}$ is very small compared with the single scattering contribution $\\Delta I_{1}$. For high-temperature galaxy clusters of $k_{B} T_{e} \\approx 15$keV, we obtain the ratio $\\Delta I_{2}/\\Delta I_{1} \\approx -0.3\\%$ at $X=5$. In the Rayleigh--Jeans region we have $\\Delta I_{2}/\\Delta I_{1} \\approx -0.03\\%$. Therefore it is concluded that the multiple scattering contribution to the thermal Sunyaev-Zel'dovich effect for galaxy clusters can be safely neglected. The reader is therefore referred to the previous four papers in this series of papers which deal with the single scattering contribution in detail." }, "0005/astro-ph0005359_arXiv.txt": { "abstract": "Radio-astronomical observations are increasingly corrupted by RF interference, and online detection and filtering algorithms are becoming essential. To facilitate the introduction of such techniques into radio astronomy, we formulate the astronomical problem in an array signal processing language, and give an introduction to some elementary algorithms from that field. We consider two topics in detail: interference detection by rank estimation of short-term covariance matrices, and spatial filtering by subspace estimation and projection. We discuss experimental data collected at the Westerbork radio telescope, and illustrate the effectiveness of the space-time detection and blanking process on the recovery of a 3C48 absorption line in the presence of GSM mobile telephony interference. ", "introduction": "Radio-astronomical observations are increasingly corrupted by RF interferers such as wireless communication and satellite navigation signals. Online detection and filtering algorithms are essential to reduce the effect of interference to an acceptable level. However, existing methods have a limited scope. Until now, the most widely implemented algorithm is a single-channel total power change detector, followed by a blanking of the correlator output. \\citet{friedman96a} has implemented an improved power detector at the RATAN600, based on detection of change in the power. \\citet{weber97} proposed the use of the quantized correlation at all lags to test the presence of interference. Another detector based on wavelet decomposition has been proposed by \\citet{maslakovic96}. These are all single channel detectors which do not exploit the spatial properties of the interference. The only detector which considered combining multiple telescopes for improved detection and blanking was proposed by \\citet{kasper82} for low frequency interferometry, where a robust data censoring method based on the temporal behavior of the cross spectrum was proposed. This requires a large number of estimated spectra ($10^5$) to obtain reliable robust estimates, and only two channels are used. Finally, adaptive filtering techniques have recently been considered by \\citet{bradley98} who propose to excise interference from the Green-Bank radio telescope using a reference antenna and an LMS type algorithm. Our aim in this paper is to introduce modern array signal processing techniques to the context of radio astronomy, and to investigate the merits of {\\em multichannel} detection and filtering algorithms at the Westerbork Synthesis Radio Telescope (WSRT). By combining cross-correlation information of a large number of sensor pairs, we can increase the detection performance significantly, and also estimate the spatial signature of interferers. In essence, our approach is to compute (on-line) short-term spatial correlation matrices in narrow sub-bands, and then to compute the eigenvalue decomposition of each of these matrices \\citep{leshem99spawc}. A rank estimate based on the eigenvalues allows to detect the number of interfering signals in each time-frequency slot, and the dominant eigenvectors give information on the ``spatial signature'' of the interferers. After detection, we can follow two directions. We can reduce the interference by rejecting corrupted time-frequency slots (blanking). This approach is suitable for time-slotted communication signals such as the European mobile telephony standard GSM, or the TDMA (time-division multiple access)-based mobile telephony standards IS-54/136 in the US. A more challenging approach is to also use the eigenvector information. Indeed, we can project out those dimensions in the spatial correlation matrices that correspond to the spatial signature vectors of the interference. Such spatial filtering techniques will greatly enhance the performance of observations with continuously-present interference. The effectiveness of the space-time detection and blanking process is demonstrated by applying the algorithms to data measured at the WSRT using an on-line 8-channel recording system. As will be shown in section \\ref{sec:experiments}, we were able to recover an absorption line of 3C48 which was completely masked by a superimposed GSM interference, and could not be recovered by single channel techniques. The paper is written in a tutorial style, to appeal to both the radio astronomy and the signal processing communities. The structure of the paper is as follows. After posing the astronomical measurement equations in section \\ref{sec:astron}, we reformulate the model in terms of array processing matrix language in section \\ref{sec:array}. We then introduce RF interference and describe its effect on the received data. In section \\ref{sec:detection} we discuss various detection algorithms. We compare the single and multichannel detectors, for the case of a narrow-band interferer with known spatial signature vector, and then present two multichannel detectors that do not assume this knowledge. We then move to spatial filtering techniques in section \\ref{sec:spatfilt}, where we formulate the basic ideas and describe a projections based approach. Finally, experimental results on multichannel blanking are shown in section \\ref{sec:experiments}. ", "conclusions": "\\label{sec:concl} In this paper, we considered various aspects of multichannel interference suppression for radio-astronomy. It was shown that by sub-band processing we have access to the many narrow-band techniques available in array signal processing. We have demonstrated the benefits of multichannel spatio-spectral blanking, both theoretically and on measured data. The results are very pleasing. We have also discussed spatial filtering techniques and demonstrated how they can be incorporated into the radio-astronomical measurement equation." }, "0005/astro-ph0005503_arXiv.txt": { "abstract": "We give a summary of recent results on spatial and velocity biases in cosmological models. Progress in numerical techniques made it possible to simulate halos in large volumes with a such accuracy that halos survive in dense environments of groups and clusters of galaxies. Halos in simulations look like real galaxies, and, thus, can be used to study the biases -- differences between galaxies and the dark matter. The biases depend on scale, redshift, and circular velocities of selected halos. Two processes seem to define the evolution of the spatial bias: (1) statistical bias and (2) merger bias (merging of galaxies, which happens preferentially in groups, reduces the number of galaxies, but does not affect the clustering of the dark matter). There are two kinds of velocity bias. The pair-wise velocity bias is $b_{12}=0.6-0.8$ at $r< 5h^{-1}$Mpc, $z=0$. This bias mostly reflects the spatial bias and provides almost no information on the relative velocities of the galaxies and the dark matter. One-point velocity bias is a better measure of the velocities. Inside clusters the galaxies should move slightly faster ($b_v =1.1-1.3$) than the dark matter. Qualitatively this result can be understood using the Jeans equations of the stellar dynamics. For the standard LCDM model we find that the correlation function and the power spectrum of galaxy-size halos at $z=0$ are antibiased on scales $r<5h^{-1}$Mpc and $k\\approx(0.15-30)h$Mpc$^{-1}$. ", "introduction": "The distribution of galaxies is likely biased with respect to the dark matter. Therefore, the galaxies can be used to probe the matter distribution only if we understand the bias. Although the problem of bias has been studied extensively in the past (e.g., Kaiser 1984; Davis et al., 1985; Dekel \\& Silk 1986), new data on high redshift clustering and the anticipation of coming measurements have recently generated substantial theoretical progress in the field. The breakthrough in analytical treatment of the bias was the paper by Mo \\& White (1996), who showed how bias can be predicted in the framework of the extended Press-Schechter approximation. More elaborate analytical treatment has been developed by Catelan et al. (1998ab), Porciani et al.(1998), and Sheth \\& Lemson (1998). Effects of nonlinearity and stochasticity were considered in Dekel \\& Lahav (1998) (see also \\citet{TaruyaSuto}). Valuable results are produced by ``hybrid'' numerical methods in which low-resolution N-body simulations (typical resolution $\\sim 20$kpc) are combined with semi-analytical models of galaxy formation \\citep[e.g.][]{ Diaferio, Benson, Somerville}. Typically, results of these studies are very close to those obtained with brute-force approach of high-resolution ($\\lesssim 2$kpc) N-body simulations \\citep[e.g.,][]{Colina, Ghignab}. This agreement is quite remarkable because the methods are very different. It may indicate that the biases of galaxy-size objects are controlled by the random nature of clustering and merging of galaxies and by dynamical effects, which cause the merging, because those are the only common effects in those two approaches. Direct N-body simulations can be used for studies of the biases only if they have very high mass and force resolution. Because of numerous numerical effects, halos in low-resolution simulations do not survive in dense environments of clusters and groups (e.g., Moore, Katz \\& Lake 1996; Tormen, Diaferio \\& Syer, 1998; Klypin et al., 1999). Estimates of the needed resolution are given in Klypin et al. (1999). Indeed, recent simulations, which have sufficient resolution have found hundreds of galaxy-size halos moving inside clusters (Ghigna et al., 1998; Col\\'in et al., 1999a; Moore et al., 1999; Okamoto \\& Habe, 1999). It is very difficult to make accurate and trustful predictions of luminosities for galaxies, which should be hosted by dark matter halos. Instead of luminosities or virial masses we suggest to use circular velocities $V_c$ for both numerical and observational data. For a real galaxy its luminosity tightly correlate with the the circular velocity. So, one has a good idea what is the circular velocity of the galaxy. Nevertheless, direct measurements of circular velocities of a large complete sample of galaxies are extremely important because it will provide a direct way of comparing theory and observations. This lecture is mostly based on results presented in Col\\'in et al. (1999ab) and Kravtsov \\& Klypin (1999). ", "conclusions": "There is a number of physical processes, which can contribute to the biases. In our papers we explore dynamical effects in the dark matter itself, which result in differences of the spatial and velocity distribution of the halos and the dark matter. Other effects related to the formation of luminous parts of galaxies also can produce or change biases. At this stage it is not clear how strong are those biases. Because there is a tight correlation between the luminosity and circular velocity of galaxies, any additional biases are limited by the fact that galaxies ``know'' how much dark matter they have. Biases in the halos are reasonably well understood and can be approximated on a few Megaparsec scales by analytical models. We find that the biases in the distribution of the halos are sufficient to explain within the framework of standard cosmological models the clustering properties of galaxies on a vast ranges of scales from 100~kpc to dozens Megaparsecs. Thus, there is neither need nor much room for additional biases in the standard cosmological model. In any case, biases in the halos should be treated as benchmarks for more complicated models, which include non-gravitational physics. If a model can not reproduce biases of halos or it does not have enough halos, it should be rejected, because it fails to have correct dynamics of the main component of the Universe -- the dark matter." }, "0005/astro-ph0005029_arXiv.txt": { "abstract": "New photometric solutions have been carried out on the important eccentric eclipsing system V380 Cygni (B1.5~II-III + B2~V) from $UBV$ differential photoelectric photometry obtained by us. The photometric elements obtained from the analysis of the light curves have been combined with the spectroscopic solution recently published by Popper \\& Guinan and have led to the physical properties of the system components. The effective temperature of the stars has been determined by fitting IUE UV spectrophotometry to Kurucz model atmospheres and compared with other determinations from broad-band and intermediate-band standard photometry. The values of mass, absolute radius, and effective temperature, for the primary and secondary stars are: $11.1\\pm 0.5$~M$_{\\odot}$, $14.7\\pm0.2$~R$_{\\odot}$, $21\\,350\\pm400$~K, and $6.95\\pm 0.25$~M$_{\\odot}$, $3.74\\pm0.07$~R$_{\\odot}$, $20\\,500\\pm500$~K, respectively. In addition, a re-determination of the system's apsidal motion rate has been done from the analysis of 12 eclipse timings obtained from 1923 to 1995. The apsidal motion study yields the internal mass distribution of the more luminous component. Using stellar structure and evolutionary models with modern input physics, tests on the extent of convection in the core of the more massive B1.5 II-III star of the system have been carried out. Both the analysis of the $\\log g - \\log T_{\\rm eff}$ diagram and the apsidal motion study indicate a star with a larger convective core, and thus more centrally condensed, than currently assumed. This has been quantified in form of an overshooting parameter with a value of $\\alpha_{\\rm ov}\\approx 0.6\\pm0.1$. Finally, the tidal evolution of the system (synchronization and circularization times) has also been studied. ", "introduction": "The bright eclipsing binary V380 Cyg (HR~7567; HD~187879; HIP~97634; $V_{max}$ = 5.68; $P=12.426$~days; B1.5~II-III + B2~V; Hill \\& Batten 1984) has properties that make it an important ``astrophysical laboratory'' for studying the structure and evolution of massive stars. In particular, the extent of convection in the stellar core and the internal mass distribution of the primary component can be probed because the system has an eccentric orbit with a well-established apsidal motion rate. Also, V380~Cyg can provide independent measures of the initial fractional helium abundance of the system ($Y$), which is an important and fundamental quantity but empirically difficult to measure. Accurate fundamental physical properties of the components (e.g., mass, radius, effective temperature, etc.) are however required to carry out such analyses. V380~Cyg consists of an evolved, more massive, and more luminous primary component and a main sequence secondary star. It has an eccentric orbit ($e=0.23$) and an orbital period of 12.426~days. Several spectroscopic studies have been carried out to determine its orbital properties and the masses, temperatures, and luminosities of the component stars (see Batten 1962; Popper 1981; Hill \\& Batten 1984; Lyubimkov et al. 1996). The first light curve of the system by Kron (1935), as well as the more modern photometry by Semeniuk (1968) and Battistini, Bonifazi, \\& Guarnieri (1974), revealed shallow eclipses with depths of $\\sim0.12$~mag and $\\sim0.09$~mag for primary and secondary minima, respectively. Moreover, these photometric measurements show that the secondary eclipse is displaced from $0\\fp5$, indicating an eccentric orbit, while changes in the displacement between primary and secondary minima indicate the presence of apsidal motion with a period of about 1500 yrs. As discussed by Gim\\'enez (1984) and Gim\\'enez, Claret, \\& Guinan (1994), V380~Cyg is an ideal binary system for the study of convective overshooting in the cores of massive stars because of the evolutionary stages of its component stars. Because of the eccentric orbit and the eclipsing nature of the binary, it is possible to determine the apsidal motion rate. From this, additional constraints on the internal mass distribution and the evolutionary state of an evolved massive star also can be established. Because V380~Cyg provides potentially important tests of stellar structure and evolution, we have carried out new high signal-to-noise spectroscopic and photometric observations of the system and performed a new, detailed investigation of its properties. In \\S \\ref{sec:new} we present the new observations. The effective temperature determination, based on both standard photometry and UV/optical spectrophotometry, is discussed in \\S \\ref{sec:teff}. In \\S \\ref{sec:modLC} we concentrate on the analysis of the light curves. \\S \\ref{sec:aps} is devoted to the study of eclipse timings, leading to an accurate determination of the apsidal motion rate. The properties of the system components are compared with the predictions of theoretical models in \\S \\ref{sec:mod}. The tidal evolution of the system (circularization and synchronization times) is analyzed in \\S \\ref{sec:tidal}. Finally, the main conclusions of our study are presented in \\S \\ref{sec:conc}. ", "conclusions": "\\label{sec:conc} This study complements the spectroscopic investigation of V380~Cyg published in Paper I. New differential photometry has been collected in two different epochs, resulting in about 300 observations in each of the $UBV$ bandpasses. The photometry has been modeled using an improved version of the Wilson-Devinney program that includes recent model atmospheres. The best fit to the light curves was achieved for the stellar and orbital properties listed in Table \\ref{tab:prop}. The effective temperatures of the stars were determined using UV spectrophotometry and optical photometry. Also, four new eclipse timings were added to the published ones and a complete re-evaluation of the apsidal motion rate of V380~Cyg was carried out. The determined physical properties (Table \\ref{tab:com}) of the V380~Cyg components were compared with evolutionary models (with the same input physics as in Claret 1995) computed for the observed masses, and with different values of convective overshoot ($\\alpha_{\\rm ov}=0.2$, $0.4$, and $0.6$). The fractional abundances of metals and helium were estimated to be $Z=0.012\\pm0.003$ (from UV/optical spectrophotometry) and $Y=0.26\\pm0.02$ (from evolutionary model fit to the secondary component) respectively. The best agreement was found for an overshooting parameter of $\\alpha_{\\rm ov} \\approx0.6$, for which the primary component is predicted to be located near the blue point of the MS hook. The excellent agreement obtained for the secondary component excludes the possibility of resorting to a different $Y$ value or an error in the temperature scale zero point for explaining the location of the primary component, because the temperature ratio of the components is well-constrained from the light curve analysis. The apsidal motion study of V380~Cyg also indicates that a value of $\\alpha_{\\rm ov}\\approx0.6$ is necessary to fit the observed apsidal motion rate of the system. The study of the locations of the components in the $\\log g - \\log T_{\\rm eff}$ diagram and the internal structure (from apsidal motion) lead to mutually consistent results. However, we caution that the relatively large value of overshooting should be restricted to the mass, evolutionary stage, and chemical composition of the primary component of V380~Cyg. Our convective overshooting result has a more profound implication than merely providing a better fit to the V380~Cyg data. We have shown that the observed physical properties of V380~Cyg cannot be described by evolutionary models with the most commonly used physical ingredients (including mild overshooting). The physical difference between these standard models and our better-fitting models (with enhanced convective overshooting) is that the latter have larger convective cores. Thus, our more general conclusion is that massive stars have larger convective cores and, therefore, are more centrally condensed than predicted by standard theory. Note that convective overshooting is not the only physical process that might produce such a result. Further increases in the opacity (as it was demonstrated when substituting the LAOL by the newer OPAL opacity tables), diffusive mixing, turbulence or other mechanisms may certainly lead to stellar models with larger and denser convective cores, relaxing the need for our large overshooting parameter. In the meantime, with the current physical ingredients, our results indicate that models should consider a moderate amount of convective overshooting for matching the observed properties of real massive stars. An investigation of the tidal evolution of V380~Cyg was also carried out. The numerical integrations of the relevant differential equations indicate a binary system with components in an eccentric orbit that have reached pseudo-synchronism. From the theory of Tassoul (1987, 1988), it appears that the orbit circularization should occur on a timescale of $10^4$ years. This is on the same timescale as it takes the primary component to fill its Roche lobe and start mass transfer. This paper clearly demonstrates the importance of specific eclipsing binaries as ``astrophysical laboratories''. In the present case of V380~Cyg, fundamental problems of stellar structure and evolution are addressed. Early-type massive systems with at least one evolved component constitute a very important source of observational data for testing stellar structure and core convection. Such systems are scarce (due to strong observational selection effects) and gathering the required photometric and spectroscopic data is not easy (because of the typically long orbital periods -- 10 days or more). The relative faintness of the secondary component also presents difficulties in the spectroscopic aspects of the problem. In spite of the observational challenges, these systems (although rare) have proved to be crucial tools and their study should be a priority in stellar astrophysics. Some examples of early-type eclipsing binaries suitable for testing core convection are: V1756~Cyg, V453~Cyg and V346~Cen (see Ribas et al. 1999). V1765~Cyg has a very massive and evolved component, already in the supergiant stage. V453~Cyg and V346~Cen have primary components of similar mass to that of V380~Cyg but appear to be at a less evolved stage ($\\log g \\approx 3.8$). Other suitable binaries for testing models are those belonging to the $\\zeta$-Aur class of eclipsing binaries (Schr\\\"oder, Pols, \\& Eggleton 1997). These are systems with at least one evolved member in the core helium burning or in the supergiant stage. The long periods (hundreds or even thousands of days) make the observations challenging. Hopefully, there will be many more potentially interesting systems to study. The long term monitoring programs associated with (future) space missions and the ground-based systematic surveys (e.g., the Sloan Digital Survey and the OGLE experiment towards the Galactic Bulge) are expected to yield valuable results, both from the detection point of view and from the densely-covered light curves that they will provide. Finally, extensive photometry of extragalactic eclipsing binaries (in LMC, SMC, and M31) are becoming available as a result of several microlensing (Grison et al. 1995; Alcock et al. 1997; Udalski et al. 1998) and variable star (Kaluzny et al. 1998, 1999; Stanek et al. 1998, 1999) surveys. Such a wealth of data guarantees a nearly inexhaustible source of eclipsing systems meeting the demanding requirements for critical model analysis, not only for solar-type abundances but also for other chemical compositions." }, "0005/hep-th0005212_arXiv.txt": { "abstract": "Over the past decade it has become clear that fundamental strings are not the only fundamental degrees of freedom in string theory. D-branes are also part of the spectrum of fundamental states. In this paper we explore some possible effects of D-branes on early Universe string cosmology, starting with two key assumptions: firstly that the initial state of the Universe corresponded to a dense, hot gas in which all degrees of freedom were in thermal equilibrium, and secondly that the topology of the background space admits one-cycles. We argue by t-duality that in this context the cosmological singularities are not present. We derive the equation of state of the brane gases and apply the results to suggest that, in an expanding background, the winding modes of fundamental strings will play the most important role at late times. In particular, we argue that the string winding modes will only allow four space-time dimensions to become large. The presence of brane winding modes with $p > 1$ may lead to a hierarchy in the sizes of the extra dimensions. ", "introduction": "In this paper we consider an approach to string cosmology in close analogy to the usual starting point of standard big-bang cosmology. We assume that the Universe started out small, dense, hot, and with all fundamental degrees of freedom in thermal equilibrium. We also assume that the background space is toroidal in all spatial dimensions \\footnote{In fact, we do not need to be this specific. The crucial assumption is the existence of one-cycles in all spatial directions.}. Given these assumptions, the initial state will consist of a gas of all fundamental branes which the theory admits. We will study the equation of state of the individual types of branes, neglecting for simplicity brane interactions, and will use the results to determine the source terms in the equations of motion for the background. In particular, we will study obstructions to spatial dimensions becoming large. We will find that fundamental string winding modes dominate the evolution at late times, and prevent more than three spatial dimensions from becoming large, in agreement with the scenario proposed in \\cite{BV} in the context of perturbative string theory. We also argue that because of t-duality the usual singularities of a the homogeneous and isotropic big bang and inflationary cosmologies are not present. The main goal of this paper is to generalize the considerations of \\cite{BV} to the context of our present understanding of string theory. In the 1980's, it was believed that the only fundamental degrees of freedom of string theory were the fundamental strings. In this context, it was shown that t-duality and string winding modes could have a very important effect on early Universe cosmology. Assuming that the background space is toroidal and thus admits string winding modes, it was shown that t-duality could explain the absence of the initial big-bang singularity. In addition, it was speculated that string winding modes would only allow three spatial dimensions to become large. The second point was put on a firmer basis by the work of Tseytlin and Vafa \\cite{TV}, who discussed the effects of gases of strings on the background equations of motion which were taken to be those of dilaton gravity. The first main point of the BV scenario \\cite{BV} is that t-duality will lead to an equivalence of the physics if the radius of the background torus changes (in string units) from $R$ to $1/R$. This corresponds to an interchange of momentum and winding modes. Thus, $R$ becoming small is equivalent to $R$ tending to infinity. Neither limit corresponds to a singularity for string matter. For example, the temperature $T$ obeys \\begin{equation} \\label{tempdual} T({1 \\over R}) \\, = \\, T(R) \\, . \\end{equation} Thus, in string cosmology the big bang singularity can be avoided. The second point suggested in \\cite{BV} was that string winding modes would prevent more than three spatial dimensions from becoming large. The point is that string winding modes cannot annihilate in more than three spatial dimensions (by a simple classical dimension counting argument). In the context of dilaton cosmology, a gas of string winding modes (which has an equation of state $\\tilde{p} = - (1/d) \\rho$, where $\\tilde{p}$ and $\\rho$ denote pressure and energy density, respectively, and $d$ is the number of spatial dimensions) will lead to a confining potential in the equation of motion for $\\lambda = log(a)$, where $a(t)$ is the scale factor of the Universe \\cite{TV}. Note that this is not the result which would be obtained in a pure metric background obeying the Einstein equations. The dynamics of classical strings in higher dimensional expanding backgrounds was studied numerically in \\cite{MS}, confirming the conclusions of \\cite{BV}. However, it is now clear that string theory has a much richer set of fundamental degrees of freedom, consisting - in addition to fundamental strings - of D-branes \\cite{Pol} of various dimensionalities. The five previously known consistent perturbative string theories are now known to be connected by a web of dualities \\cite{Witten1}, and are believed to represent different corners of moduli space of a yet unknown theory called M-theory. Which branes arise in the effective string theory description depends on the particular point in moduli space. We will be making a specific assumption below. The question we would like to address is whether the inclusion of the new fundamental degrees of freedom will change the main cosmological implications of string theory suggested in \\cite{BV}, namely the avoidance of the initial cosmological singularity, and the singling out of 3 as the maximal number of large spatial dimensions, in the context of an initial state which is assumed to be hot, dense and small, and of a background geometry which admits string winding modes. Our concrete starting point is 11-dimensional M-theory compactified on $S^1$ to yield 10-dimensional Type II-A string theory. The resulting low energy effective theory is supersymmetrized dilaton gravity. As fundamental states, M-theory admits the graviton, 2-branes and 5-branes. After compactification, this leads to 0-branes, 1-branes, 2-branes, 4-branes, 5-branes, 6-branes and 8-branes as the fundamental extended objects of the 10-dimensional theory. The dilaton represents the radius of the compactified $S^1$. We are in a region of moduli space in which the string coupling constant $g_s$ is smaller than 1. We assume that all spatial dimensions are toroidal (radius $R$), and that the Universe starts out small, dense, hot, and in thermal equilibrium. Thus, the Universe will contain a gas of all branes appearing in the spectrum of the theory. Note that this starting point is in close analogy with the hot big bang picture in standard cosmology, but very different from brane-world scenarios in which the existence of a particular set of branes is postulated from the outset without much justification from the point of view of cosmology. There have been several interesting previous studies of the cosmology of brane gases. Maggiore and Riotto \\cite{MR99} (see also \\cite{R99}) studied the phase diagram of brane gases motivated by M-theory as a function of the string coupling constant and of the Hubble expansion rate (as a measure of space-time curvature) and discovered regions of the phase diagram in which brane gases determine the dynamics, and regions in which the effective action is no longer well described by a ten-dimensional supergravity action. Given our assumptions, we are in a region in moduli space in which the ten-dimensional effective description of the physics remains true to curvature scales larger than that given by the string scale. In this paper, we consider the time evolution of the system through phase space starting from some well-defined initial conditions. We will argue that as a consequence of t-duality, curvature scales where the ten-dimensional description breaks down are never reached. In another interesting paper, Park et al. \\cite{PSL00} take a starting point very close to our own, a hot dense gas of branes. However, they did not consider the winding and oscillatory modes of the branes. In the following section we will study the equation of state of the brane gases for all values of their spatial dimension p. We will separately analyze the contributions of winding and non-winding modes (the latter treated perturbatively). The results will be used as source terms for the equations of motion of the background dilaton gravity fields, following the approach of \\cite{TV}. We find that the winding modes of any p-brane lead to a confining force which prevents the expansion of the spatial dimensions, and that the branes with the largest value of p give the largest contribution to the energy of the gas in the phase in which the scale factor is increasing. In Section 3 we use the results of the previous section to argue that the main conclusions of the scenario proposed in \\cite{BV} are unchanged: t-duality eliminates the cosmological singularity, and winding modes only allow three dimensions of space to become large. We point out a potential problem (the {\\it brane problem}) of cosmologies based on theories which admit branes in their spectrum of fundamental states. This problem is similar to the well-known domain wall problem \\cite{problem} in cosmological models based on quantum field theory. It is pointed out that a phase of loitering (see e.g. \\cite{SFS}) yields a natural solution of this problem, and it is shown that the background equations of motion may well yield a loitering stage during the early evolution of the Universe. Some limitations of our considerations and avenues for future research are discussed in the final section. \\vskip 0.4cm ", "conclusions": "In this paper we have generalized the approach to superstring cosmology pioneered in \\cite{BV} to include the contribution of branes. Our starting point is the assumption that the Universe starts out small, hot and in thermal equilibrium, with all of the spatial dimensions being equivalent and compact (string scale). We also assume that the background admits one cycles. We work in the corner of moduli space resulting from compactification of 11-d M-theory on $S^1$, and in which the string coupling constant is small. We argue that, as a consequence of t-duality, the usual big bang singularity is absent in the resulting cosmology. Furthermore, since winding modes of all of the branes are excited in thermal abundance in the initial state, and since the energy in winding modes increases as the background spatial scale expands, the thermodynamics of the winding modes coupled to the background equations of motion, which in the corner of moduli space of M-theory which we consider are the dilaton-gravity equations, dominates the initial evolution of the background. Our thermodynamic considerations suggest that the mechanism first pointed out in \\cite{BV}, by which the string winding modes will prevent all but 3 spatial dimensions from becoming large, persists. In addition, the presence of winding modes of 2-branes may lead to a hierarchy in the sizes of the extra dimensions, with exactly internal dimensions being larger than the others. We pointed out some further interesting characteristics of the resulting cosmology. In particular, the horizon problem is absent. However, in order to produce a Universe in which the large spatial dimensions exceed the present Hubble radius, it seems necessary to have a background evolution of the large spatial dimensions resembling inflation. We also pointed out the existence of a {\\it brane problem}, a problem for cosmology in theories with stable branes which is analogous to the domain wall problem in cosmological scenarios based on quantum field theories which admit stable domain walls. A phase of loitering in the background cosmological evolution will naturally solve this problem. Based on the background equations of motion, it appears that as long as the winding modes do not disappear, the background solutions approach a point of loitering. It would be interesting to extend our considerations to other regions in moduli space, in particular to regions of strong string coupling. In those regions it appears \\cite{MR99}that the effective ten-dimensional background description breaks down before the Hagedorn temperature is reached. It is also important to point out that on Calabi-Yau threefold backgrounds one cycles are absent, and thus our cosmological scenario does not apply. Calabi-Yau three-folds are required if the four-dimensional low energy effective theory is to have $N=1$ supersymmetry. In the context of early Universe cosmology, however, it is not reasonable to demand $N=1$ supersymmetry. In particular, one could have maximal supersymmetry which is consistent with the toroidal background we are using. It would be interesting to explore whether consistent four-dimensional low-energy effective theories can be constructed from compact backgrounds which admit one-cycles. Given that the tension of the branes exceeds that set by the string scale, the applicability of the homogeneous background field equations to brane gases might be questionable, in particular at late times when the branes are fairly widely separated \\footnote{We thank D. Lowe for stressing this concern to us.}. This issue deserves further study. \\vspace{0.4cm} \\centerline{\\bf Acknowledgements} The research was supported in part by the U.S. Department of Energy under Contract DE-FG02-91ER40688, TASK A. Two of us (S.A. and D.E.) are supported by fellowships by the U.S. Department of Education under the GAANN program. We are grateful to E. Akhmedov, R. Easther, B. Greene, A. Jevicki, D. Lowe and S. Ramgoolam for many discussions about this project." }, "0005/hep-ph0005256_arXiv.txt": { "abstract": "s{The possibility that the Fermi scale is the only fundamental energy scale of Nature is under serious consideration at present, yet cosmic rays may already have provided direct evidence of new physics at a much higher scale. The recent detection of very high energy particles with no plausible astrophysical sources suggests that these originate from the slow decays of massive particles clustered in the halo of our Galaxy. Such particles had in fact been predicted to exist beforehand with mass and lifetime in the range required to explain the observations. I discuss recent work focussing on experimental tests of this speculative but exciting idea.} ", "introduction": "\\label{sec:Intro} The only massive particles in the Standard Model to have survived from the Big Bang are nucleons --- protons and (bound) neutrons --- along with a commensurate number of electrons to yield the observed charge neutrality of the universe.\\footnote{We know now that massive relic neutrinos contribute at least as much as the luminous component of nucleons to the present energy density. However they are unlikely to be the dominant component of the dark matter, based on arguments concerning structure formation.} Considerations of primordial nucleosynthesis restrict the nucleonic contribution to the density parameter to $\\Omega_{\\rm N}\\lessim0.1$ and it is widely accepted that the dark matter in galaxies and clusters which contributes $\\Omega_{\\rm DM}\\gtrsim0.3$ is non-nucleonic and probably composed of a new stable relic particle. There are many candidates for the identity of this particle but the most popular notion is that it is associated with the new physics beyond the Standard Model necessary to stabilize the hierarchy between the Fermi scale, $G_{F}^{-1/2}\\simeq300$~GeV, and the Planck scale, $G_{N}^{-1/2}\\simeq10^{19}$~GeV. In particular theories of (softly broken) low energy supersymmetry (SUSY) typically imply that the lightest SUSY partner is a neutralino with mass of order the Fermi scale, which is absolutely stable if the discrete symmetry termed $R$-parity is exactly conserved. Interestingly enough the relic abundance of such a weakly interacting particle which was in thermal equilibrium in the early universe can account for the dark matter. In supergravity theories, there is a new energy scale of ${\\cal O}(10^{11})$~GeV --- the geometric mean of the Fermi and Planck scales. This is the scale of the `hidden sector' in which SUSY is broken through gaugino condensation induced by a new strong interaction, and communicated to the visible sector through gravitational interactions. Following the emergence of superstrings (for which $N=1$ supergravity is the effective field theory) it was realised \\cite{confine} that the hidden sector can also serve to confine fractionally charged states which are a generic prediction \\cite{charge} of string theory. This avoids a serious conflict with the unsuccessful experimental searches for fractional charges but necessarily implies the existence of (integrally charged) bound states with mass of ${\\cal O}(10^{11})$~GeV. In a specific construction with $SU(5)\\otimes\\,U(1)$ unification, it was noted \\cite{crypton} that most such states would be short-lived but that the lightest such state would only decay through non-renormalizable operators of dimension $\\geq8$ and thus have a lifetime exceeding the age of the universe. This introduces a new candidate for the constituent of the dark matter --- named ``cryptons'' --- interestingly similar to nucleons which too are bound states of fractional charges and can only decay through non-renormalizable operators. However, just as with nucleons, their cosmological origin is a puzzle. If such particles were ever in thermal equilibrium their relic abundance would have been excessive since their self-annihilations are rather inefficient. For nucleons the problem is just the opposite and their very existence today requires an out-of-equilibrium origin. If the same were true of cryptons, their relic abundance may well have a cosmologically interesting value.\\footnote{It has recently been noted \\cite{wimpzilla} that particles with mass of ${\\cal O}(H_{\\rm inf})\\sim10^{13}$~GeV --- also dubbed ``wimpzillas'' --- can be created with a cosmologically interesting abundance through quantum vacuum fluctuations during inflation or during the subsequent (re)heating process.} It is then interesting to ask what the observational signatures of such particles might be. Reviving an old suggestion \\cite{fg80}, we recognised \\cite{cryptondm} that the most sensitive probe would be in extremely high energy cosmic rays (EHECR), specifically in the flux of high energy neutrinos which would necessarily be created by crypton decays. The best constraint we obtained followed from the upper limit on deeply penetrating air showers set by the Fly's Eye atmospheric fluorescence experiment; this implied that such particles must have a lifetime exceeding $\\sim10^{18}$~yr if they are an important constituent of the dark matter. As this was close to the theoretically expected lifetime in the ``flipped'' $SU(5)$ model, I was optimistic enough to suggest in a conference talk \\cite{taup91} that `` \\ldots some improvement of these experimental sensitivities can rule out (or detect!) such particles''. Just a few months later the Fly's Eye array detected \\cite{flyseye} an event, consistent with a proton primary, but with an energy of $(3.0\\pm0.9)\\times10^{11}$~GeV. This was well above the Greisen-Zatsepin-Kuzmin (GZK) cutoff \\cite{gzk} energy of $\\sim5\\times10^{10}$~GeV, beyond which resonant photopion production losses on the cosmic microwave background should limit the propagation distance of any such strongly interacting particle to less than about a hundred Mpc. Over a dozen such events have been detected subsequently by the Akeno airshower array (AGASA) as well as HiRes, the successor to Fly's Eye, so the absence of the GZK cutoff \\cite{spectrum} is now well established. However contrary to the expectation that such high energy particles, being essentially undeflected by the weak intergalactic magnetic fields, should point back to their sources, the observed distribution on the sky \\cite{isotropy} is consistent with isotropy. This is quite baffling given that that only a few astrophysical sites (active galactic nuclei or the extended lobes of radio galaxies) are capable of accelerating such particles, even in principle, and there are none \\cite{sources} along the arrival directions within the propagation range. Hence it is generally acknowledged \\cite{astro} that there is no ``conventional'' astrophysical explanation for the observed EHECR. ", "conclusions": "Although some progress has been made in sharpening the spectral predictions of the decaying halo particle model for EHECR, much work still needs to be done. The calculations so far have assumed the simplest decay channel --- into two partons. However non-renormalizable operators are in fact likely to induce many-body decays. The effects of supersymmetry also need to be investigated more carefully, e.g. the effects of varying the SUSY parameters and inclusion of sparticle decay channels. Nevertheless it is already clear that the general trend in the EHECR data can be accounted for by this hypothesis, if the particle mass is $m_X\\sim10^{12-13}$~GeV and its lifetime is $\\tau_X\\sim10^{16}~{\\rm yr}(\\xi_X/3\\times10^{-4})$, so that even with a very long lifetime such particles need constitute only a tiny fraction $\\xi_X$ of the halo CDM. It is also clear that TD models \\cite{td}, in which $m_X$ corresponds to the GUT-scale, are already {\\em ruled out} by the spectral data. The next generation of large area cosmic ray, gamma-ray and neutrino observatories (Auger, Amanda, Antares, \\ldots) is now under construction so it is important to refine these calculations in order to make specific predictions for the expected fluxes. We emphasize that previous estimates of high energy gamma-ray and neutrino fluxes from TD \\cite{tdgamnu} are based on the Hill fragmentation functions (\\ref{hill1},\\ref{hill2}), while other work \\cite{tdgamnu2} use the (M)LLA spectrum (\\ref{lla}) or its SUSY variant. Blasi \\cite{b99} has calculated in detail the flux of $\\gamma$-rays in the decaying halo particle model but he too uses the Hill and the MLLA spectra. All these approximations are {\\em inapplicable} at the high energies of interest as explained earlier, and moreover the spectra of pions are not simply proportional to that of nucleons as assumed. Hence it is clear that all these estimates are unreliable. It is essential that further work use the physically more realistic approach to calculating fragmentation spectra outlined above in order to devise definitive experimental tests \\cite{us} of the decaying particle hypothesis." }, "0005/hep-ph0005310_arXiv.txt": { "abstract": "\\widetext The appearance of high energy tau neutrinos due to $\\nu_\\mu \\rightarrow \\nu_\\tau$ oscillations of extragalactic neutrinos can be observed by measuring the neutrino induced upward hadronic and electromagnetic showers and upward muons. We evaluate quantitatively the tau neutrino regeneration in the Earth for a variety of extragalactic neutrino fluxes. Charged-current interactions of the upward tau neutrinos below and in the detector, and the subsequent tau decay create muons or hadronic and electromagnetic showers. The background for these events are muon neutrino and electron neutrino charged-current and neutral-current interactions, where in addition to extragalactic neutrinos, we consider atmospheric neutrinos. We find significant signal to background ratios for the hadronic/electromagnetic showers with energies above 10 TeV to 100 TeV initiated by the extragalactic neutrinos. We show that the tau neutrinos from point sources also have the potential for discovery above a 1 TeV threshold. A kilometer-size neutrino telescope has a very good chance of detecting the appearance of tau neutrinos when both muon and hadronic/electromagnetic showers are detected. ", "introduction": "A recent breakthrough in the study of neutrino oscillations came from the observation by the Super-Kamiokande experiment of a deficit of upward-going atmospheric muon neutrinos \\cite{superk}. The observed electron neutrino flux was found to be consistent with the theoretical expectation from models of cosmic ray production of neutrinos. Furthermore, SuperK measurements are consistent with earlier experiments \\cite{kamioka,kamR,imb,soudan} which detected anomalous ratios of the $\\nu_\\mu$ to $\\nu_e$ flux. The new high-statistics data disfavor scenarios in which $\\nu_\\mu$'s oscillate into sterile neutrinos ($\\nu_s$) \\cite{fgv}, and the data are consistent with $\\nu_\\mu$ to $\\nu_\\tau$ oscillation (99$\\%$ CL) with a large mixing angle, $\\sin^2 \\theta > 0.84$ and a neutrino mass squared difference of $2 \\times 10^{-3}$ eV$^2 < \\Delta m^2 < 6 \\times 10^{-3}$ eV$^2$. Direct detection of $\\nu_\\tau$ appearance is extremely difficult because at low energies, the charged-current cross section for producing a tau is small and the tau has a very short lifetime. Several long-baseline experiments with accelerator sources of $\\nu_\\mu$ \\cite{MINOS,K2K,ICARUS,NOE,OPERA} have been proposed with the goal of detecting tau neutrinos from oscillations, thus confirming the SuperK results. The only convincing evidence of neutrino oscillations to date involves astrophysical sources, neutrinos from the sun and atmospheric neutrinos. These observations involve indirect measurements, namely the disappearance of the expected neutrino fluxes. We have recently discussed the possibility of using a kilometer-size neutrino telescope to detect tau neutrinos from extragalactic sources of high-energy neutrinos such as Active Galactic Nuclei (AGN) and Gamma Ray Bursts (GRB), assuming $\\nu_\\mu\\leftrightarrow \\nu_\\tau$ with the oscillation parameters of the SuperK experiment \\cite{irs}. The probability for $\\nu_\\mu \\rightarrow \\nu_\\tau$ is given by \\cite{kayser} \\begin{equation} P(\\nu_\\mu \\rightarrow \\nu_\\tau; L) = \\rm{sin^2}2\\theta \\,\\,\\,\\rm{sin{^2}} \\Biggl(\\frac{1.27\\Delta m^2({\\rm eV}^2)L({\\rm km})}{E({\\rm GeV})}\\Biggr)\\, . \\end{equation} Assuming two flavor oscillations, muon neutrinos produced in AGN or GRB would oscillate to tau neutrinos as they travel to the Earth. Over astronomical distances in the range of a megaparsec to thousands of megaparsecs, by measuring tau neutrino fluxes, one could, in principle, probe oscillations down to $\\Delta m^2 \\sim 10^{-17}$ eV$^2$, nine orders of magnitude below current neutrino experiments \\cite{halzen,wspl}. On the other hand, for the SuperK parameter range, with $\\Delta m^2$ on the order of $10^{-3}$ eV$^2$ and $\\sin^22\\theta\\simeq 1$, the oscillation probability is 0.5. It is this latter possibility that we explore in this paper. We use the simplest assumption for the flavor content of extragalactic sources of neutrinos, in the absence of oscillations, for the ratio $\\nu_e:\\nu_\\mu:\\nu_\\tau$ to be $1:2:0$. This is based on a counting argument applied to $\\pi\\rightarrow \\mu\\nu_\\mu$ and $\\mu\\rightarrow \\nu_\\mu e\\nu_e$ processes. With the two-flavor oscillations suggested by the SuperK experiment, the flavor ratio becomes $1:1:1$ after the neutrinos have traveled over astronomical distances. Even in the three-flavor oscillation scenario, the ratio is still $1:1:1$, because the path length of high energy extragalactic neutrinos is much larger than any neutrino oscillation length supported by the solar, atmospheric or accelerator data \\cite{yasuda}. The ratio for $\\nu_e:\\nu_\\mu$ might get modified at high energies due to muon cooling \\cite{rm}. In addition, $\\nu_e$ from neutron decay might give significant contribution, resulting in neutrino fluxes dominated by electron neutrinos as in the case of diffuse neutrino fluxes from propagating cosmic rays \\cite{stanev0003}. We comment qualitatively in the discussion section on how our results are altered with more realistic, flavor-dependent neutrino energy cutoffs. Regardless of the flavor content of the source, the maximal mixing suggested by the SuperK experimental results mean that there will be an appreciable tau neutrino component at the Earth, so one is interested in tau neutrino detection in high energy neutrino telescopes such as ANTARES \\cite{antares}, NESTOR \\cite{nestor}, AMANDA \\cite{amanda} and the next generation of large underground detectors \\cite{km3}. Tau neutrino detection requires an understanding of the effect of propagating through the Earth on the tau neutrino flux. The propagation of ultra-high energy tau neutrinos through the Earth is quite different from muon and electron neutrinos. The Earth never becomes opaque to tau neutrinos, while muon and electron neutrinos are absorbed via charged-current interactions before reaching the opposite surface \\cite{halzen}. Ultrahigh-energy tau neutrinos interact in the Earth producing taus which, due to the short lifetime, decay back into tau neutrinos with lower energy. This cascade continues until the tau neutrinos reach the detector on the opposite side of the Earth or until the energy of the neutrinos is small enough that the interaction length of the neutrino is longer than the path length through the Earth. The energy and nadir angle dependence of the extragalactic tau neutrinos fluxes have been examined quantitatively in Refs. \\cite{irs,bottai}. For certain fluxes, those that do not decrease too steeply with energy, there are significant enhancements of the tau neutrino flux relative to the muon neutrino flux at energies below $\\sim 10^6$ GeV. An enhancement of the tau neutrino flux does not necessarily translate to dramatic modifications of the standard model (no-oscillation) rates for upward-going muons, especially in view of uncertainties in the normalization of the extragalactic fluxes. However, by comparing rates for upward-going muons with rates for upward hadronic/electromagnetic (EM) showers, the signature of tau neutrino interactions is unambiguous for a large range of neutrino fluxes. In the next section, we briefly introduce the extragalactic and generic $\\nu_\\tau$ fluxes $F_{\\nu_\\tau}^{o}\\sim E^{-n}$ for $n=1,2$ that are used here. After reviewing neutrino propagation through the Earth, we describe $\\nu_\\tau$ signatures. The fluxes considered here have a range of energy behaviors. Even if the normalizations of the neutrino fluxes are uncertain, and in some cases optimistic, it is useful to make quantitative comparisons of the event rates for upward muons and upward hadronic/EM showers, with and without neutrino oscillations, which we do in Section IV. The quantitative results for specific models lead to model independent conclusions, which we summarize graphically. Tau neutrino appearance would provide an independent confirmation of the SuperK results and would point towards the better understanding of physics beyond the Standard Model. ", "conclusions": "We have studied signals for $\\nu_\\mu \\rightarrow \\nu_\\tau$ oscillations with extragalactic high energy muon neutrinos. Assuming SuperK oscillation parameters, muon neutrinos convert into tau neutrinos as they travel megaparsec distances, with both fluxes being equal at the surface of the Earth. High energy muon neutrinos get absorbed as they pass through the Earth, while tau neutrinos cascade down to lower energies. We find this enhancement of the $\\nu_\\tau$ flux in the low energy region to be prominent for flat initial spectrum, such as $E^{-1}$, the AGN model of Stecker and Salamon, and the topological model of Sigl {\\it et al}. For steeper spectra, the enhancement is small because the number of higher energy neutrinos that contributes to the lower energy flux via tau decay is relatively small compared to the low energy flux of neutrinos. Upward tau neutrinos, once they reach the detector, interact producing tau leptons which decay with very short lifetimes. We have considered muons from tau decay as well as its hadronic decay mode. Since the planned detectors are unable to distinguish between hadronic and electromagnetic showers, we have included all the processes that give both hadronic and electromagnetic showers. We find that upward muons alone would not be sufficient to separate the tau neutrinos contribution, due to the large background from $\\nu_\\mu$ charged-current interactions, the small branching fraction for $\\tau\\rightarrow \\mu$ decay mode and the model uncertainty for the incident neutrino flux. In the case of upward hadronic/EM showers, we find that tau neutrinos give significant contributions, signaling the $\\nu_\\tau$ appearance. Given the uncertainties in the normalizations of the extragalactic neutrino fluxes, combining muon rates and hadronic/EM rates offer the best chance to test the $\\nu_\\mu\\rightarrow \\nu_\\tau$ oscillation hypothesis. As concluded in earlier work \\cite{gqrs96,gqrs98}, in general, an energy threshold of between 10 TeV and 100 TeV for upward muons and showers is needed in order to reduce the background from atmospheric neutrinos. We find that diffuse AGN neutrino fluxes, as described by the Stecker-Salamon and Mannheim models, as well as neutrinos from GRBs can be used to detect tau appearance. By measuring upward showers with energy threshold of 10 TeV, and upward muons, the event rates exceed the atmospheric background and are about a factor of 1.5-2 larger than in the no-oscillation scenario. Here we also comment on the effect of muon and pion cooling to the flavor ratio. Athar {\\it et al.} in Ref. \\cite{yasuda} have shown that with a negligible electron neutrino content at the source, the electron neutrino content at the Earth (in the three-flavor model) is reduced if not negligible compared to the nearly equal muon and tau neutrino fluxes. Keeping the energy spectrum unchanged, this means that the hadronic/electromagnetic shower background, which has significant contributions from $\\nu_e N\\rightarrow e X$ with $\\nu_e>E_{\\rm shr}^{\\rm min}$ would be reduced. Electron (anti-) neutrinos from processes in the propagation of cosmic rays may dominate at some energies \\cite{stanev0003}. We have not considered that possibility here because of the low rates below 1 PeV. Steepening of the energy spectra displayed in Fig. 1 due to a neutrino energy cutoff from pion and muon cooling will have implications for the tau neutrino `pileup', especially for the flatter spectra where the pileup is more pronounced. As an estimate of the lower bound on the relative enhancement of the hadronic/EM signal compared to the muon signal, one can compare the rates for horizontal events, where tau neutrino pileup is small. For example, Figures 22 (a-f) show clear distinction between oscillation and no-oscillation scenarios, even in directions near horizontal, where there is no pile up. Furthermore, for $E^{-2}$ flux, where the pileup is very small \\cite{irs}, the ratio of ratios $R$ discussed above ranges from 2.5 to 2.8. Thus, even without the tau neutrino pileup, the oscillation scenario can be distinguished from the no-oscillation scenario. The detection of $\\nu_\\mu \\rightarrow \\nu_\\tau$ oscillations with a point source might also be possible. With the resolution for the planned neutrino telescopes of $2^{\\circ}$, the atmospheric background is reduced by $3.8 \\times 10^{-3}$. For upward showers, this gives less than 1 event per year for $E_{\\rm shr}^{\\rm min}=1$ TeV, and even less for higher energy thresholds. Thus, if the point source has a flat spectrum, $F_{\\nu+\\bar{\\nu}} = 10^{-16} E^{-1}$, then one would be able to detect tau neutrinos by measuring upward showers with $E_{\\rm shr}^{\\rm min}=1$ TeV. In the more realistic case, when the point source has a steeper spectrum ($E^{-2}$), such as Sgr A* \\cite{markoff98}, a normalization of $10^{-7}$/cm$^2$/s/sr/GeV would be sufficient for the detection of tau neutrinos with threshold of 1 TeV. Time correlations with variable point sources would further enhance the signal relative to the background. We have demonstrated that extragalactic sources of neutrinos can be used as a very-long baseline experiment, providing a source of tau neutrinos and opening up a new frontier in studying neutrinos oscillations. \\vskip 0.1true in \\leftline{Acknowledgements} The work of S.I.D. and I.S. has been supported in part by the DOE under Contracts DE-FG02-95ER40906 and DE-FG03-93ER40792. The work of M.H.R. has been supported in part by National Science Foundation Grant No. PHY-9802403. \\appendix" }, "0005/astro-ph0005139_arXiv.txt": { "abstract": "The number of publications considering Narrow-Line Seyfert 1 galaxies has increased dramatically in recent years. Especially after the launch of the X-ray missions ROSAT and ASCA, Narrow-Line Seyfert 1s have become very popular. In these proceedings I will give an overview of how they are distributed over the electromagnetic spectrum. I will describe what we know about them at radio, infrared, optical, and X-ray bands, and how they differ and how they are similar to Broad-Line Seyfert 1s. Finally I will introduce a method to find them with high probability. ", "introduction": "Looking back in time, Narrow-Line Seyfert 1 galaxies (NLS1) have become one of the most popular issues in astronomy in recent years. Between the definition of NLS1 by Osterbrock \\& Pogge in 1985 and the launch of ROSAT in 1990, only a couple of papers on NLS1 had been published. The launches of ROSAT and ASCA seem to be the break points. After these launches in 1990 and 1993, respectively, the number of publications that deal with NLS1 has increased dramatically. In Figure \\ref{distr} the number of publications is shown that contain the term `Narrow Line Seyfert 1' in their title, derived from the CDS/ADS abstract server. However, this is only the tip of the iceberg. The real number of publications is much higher. NLS1 have become an important class of objects in astrophysics. \\begin{figure} \\psfig{file=dgrupe1_fig_1.ps,width=12.0cm,bbllx=3.0cm,bblly=1.0cm,bburx=28.0cm,bbury=19.5cm,clip=} \\caption{\\label{distr} Development of the number of publications per year that contain the word 'Narrow Line Seyfert 1' in their title. The arrows mark the launches of ROSAT and ASCA. Important publications about NLS1 are marked as well. } \\end{figure} ", "conclusions": "NLS1 are AGN with extreme properties. They show the steepest X-ray spectra, the strongest FeII emission and the lowest emission from NLR lines. They are more variable in X-rays than BLS1 and their soft X-ray excess is stronger. Soft X-ray selected NLS1 do not show significant cold absorption and polarization. However, optically selected NLS1 do (see Goodrich 1989). One question still remains: Are NLS1 their own class or are they just a subclass of Seyfert 1s? Well, their properties are extreme and NLS1 are the objects with the steepest X-ray spectra and the most pronounced soft X-ray excess. However, there is no distinct physical boundary between NLS1 and BLS1. My personal point of view is, that they are just the continuation of Seyfert 1s towards extreme properties. They are at the end of the Boroson \\& Green (1992) Eigenvector 1 relation. NLS1 are probably AGN that accrete at Eddington accretion rates, or if Eigenvector 1 represents the age of AGN, they can be considered to be very young objects (Grupe 1996, see also S. Mathur's contribution in these proceedings, and her recent article (2000)). One final remark: The number of NLS1 that have been found has indeed increased dramatically in the last decade. Nevertheless, NLS1 existed of course before Osterbrock \\& Pogge (1985) defined them. It took years to make the term `Narrow-Line Seyfert 1' popular. Many objects which we would consider today to be NLS1 were not named so in those days. On the other hand today, objects are offen called NLS1 which do not belong in this category. {\\em Acknowledgments:} I would like to thank Drs. Bev Wills and Thomas Boller for carefully reading the manuscript and their suggestions and comments to this article." }, "0005/astro-ph0005555_arXiv.txt": { "abstract": "Zinc is a good indicator of metallicity in Damped Lyman $\\alpha$ (DLA) systems because it is almost unaffected by dust depletion. However, the use of zinc as a tracer of metallicity evolution has been hampered by the difficulty of detecting the Zn II resonance lines at high redshift. The measurement of zinc abundance in a DLA system at $z_{\\rm abs} > 3$ obtained by means of the UVES spectrograph at the VLT prompted us to re-analyse the full sample of zinc abundances present in the literature to search for a metallicity-redshift relation in DLA systems. The study of the metallicities of individual systems shows evidence for an anti-correlation between [Zn/H] and redshift supported by different types of statistical tools. The zinc metallicity decreases by $ -0.3 \\pm 0.1 $ dex per unit redshift interval in the range $0.5 \\lsim z_{\\rm abs} \\lsim 3.5$. This rate is in good agreement with that found by Savaglio, Panagia \\& Stiavelli (2000) in their recent study of DLA abundances corrected for dust depletion. The present result does not require a knowledge of the dust depletion pattern(s) in DLA systems. On the other hand, the analysis of the column-density weighted metallicity of the sample, $$, does not show a clear evidence for redshift evolution, consistent with previous studies of zinc abundances. We propose that the apparent lack of evolution of $$ is due to the combination of selection bias effects together with the extreme sensitivity of $$ to low-number statistics. ", "introduction": "The redshift-metallicity relation of QSO absorption systems is a fundamental probe of the chemical evolution of the universe. The study of such relation in Damped Lyman $\\alpha$ (DLA) systems\\footnote{The QSO absorption systems with neutral hydrogen column density $N$(HI) $> 2 \\times 10^{20}$ atoms cm$^{-2}$ are called Damped Lyman $\\alpha$ systems owing to the presence of radiative damping wings in their Lyman $\\alpha$ absorption profiles.} probes, in particular, the metal enrichment of the associated galaxies located at cosmological distances along the QSO line of sight (Lu et al. 1996). Abundance studies of DLA systems can be used to trace the chemical evolution of galaxies in the early universe, starting from the redshifts of the most distant QSOs. Abundance determinations for different elements are presently available for about 60 DLA systems, but the study of their redshift evolution is hampered by two main difficulties. One is the uncertainty of the abundance measurements owing to the possible effects of dust depletion. The other is the limited redshift coverage of the sample, which is not always adequate for probing the presence of evolution. One approach to tackle the first difficulty is to correct the observed abundances for dust depletion effects (Vladilo 1998, Savaglio et al. 2000). Another approach is to use an element unaffected by dust depletion as a tracer of metallicity evolution (Pettini et al. 1997, 1999). Zinc is known to have little affinity with dust since it is essentially undepleted in the interstellar medium (Roth \\& Blades 1995). Abundance measurements in Galactic metal-poor stars yield [Zn/Fe]\\footnote{We adopt the usual convention [X/Y] = log $N$(X)/$N$(Y) $-$ log (X/Y)$_{\\sun}$} $\\approx 0$ (Sneden, Gratton \\& Crocker 1991), suggesting that zinc is a good tracer of iron. The quality of available stellar data leaves open the possibility that new, more precise measurements may reveal small deviations of [Zn/Fe] from the solar ratio such as those found for the iron-peak elements Cr, Mn and Co (Ryan, Norris \\& Beers 1996). However, even in this case, zinc would still be a good indicator of metallicity, even if not a perfect tracer of Fe. Studies of zinc abundances in DLA systems have not revealed evidence for evolution of the column-density weighted [Zn/H] metallicity (Pettini et al. 1997, 1999). One difficulty in detecting evolution is the lack of measurements at $z_{\\rm abs} > 3$, when the redshifted Zn\\,II resonant doublet $\\lambda_{\\rm rest} 2025, 2062$ \\AA\\ falls in the reddest part of the visible spectrum. In addition, the column-density weighted metallicity is more prone to be affected by low number statistics than the unweighted metallicity, as we discuss in Section 3. Here we present the results of a search for redshift evolution performed by considering both the unweighted and the weighted metallicities (Sections 2 and 3, respectively) and based on the sample of [Zn/H] literature data, which includes now our recent measurement at $z_{\\rm abs} \\simeq 3.4$ (Molaro et al. 2000). ", "conclusions": "We find evidence for an anti-correlation between the absolute zinc abundance [Zn/H] and the absorption redshift $z_{\\rm abs}$ of DLA systems, with a slope $\\simeq -0.3 \\pm 0.1$ in the range $0.5 \\lsim z_{\\rm abs} \\lsim 3.5$. The zinc metallicity increases from $\\approx 3\\%$ up to $\\approx 25\\%$ of the solar value from $z_{\\rm abs} \\simeq 3.5$ to $z_{\\rm abs} \\simeq 0.5$. Should DLA absorbers continue the same trend also from $z_{\\rm abs} \\simeq 0.5$ to $z_{\\rm abs} \\simeq 0$, the typical present-day metallicity would be $\\approx 35\\%$ solar, even though a value as high as $60\\%$ is still within the errors of the intercept. Correcting for dust depletion effects would slightly steepen the anti-correlation, but well within the statistical error of the slope; the characteristic present-day metallicity would rise up to $\\approx 50\\%$ solar, with values as high as $\\approx 100\\%$ solar still within the errors. The slope of the metallicity redshift relation that we derive is in good agreement with the value recently derived by Savaglio et al. (2000). The literature data base considered by these authors is larger than ours and includes different elements in addition to zinc. However, most of these elements are known to be severely depleted into dust in the ISM and the results by Savaglio et al. are based on an algorithm that corrects the abundances for depletion effects. The present results do not require a modeling of the elemental depletion patterns nor assumptions on the intrinsic abundance patterns in DLA systems. In spite of the correlation with redshift, the zinc metallicities show evidence for intrinsic scatter. Models of galactic chemical evolution have already been able to explain such scatter by considering the surface brightness and the formation redshift of the galaxies, as well as the galactocentric distance of the gas intercepted (Jim\\'enez, Bowen \\& Matteucci 1999). While such analysis has shown a general consistency between the zinc observations and the predicted evolutionary tracks, the present results demonstrate for the first time the evolution of zinc metallicity on pure observational grounds. While we find evolution of the zinc metallicity of individual systems, we do not find evolution of the column-density weighted metallicity $$. This is consistent with the results of previous studies by Pettini et al. (1997, 1999). The lack of evolution of the zinc mean cosmic metallicity might be due, at least in part, to the lack of a sufficiently large data base, since the measurement of $$ is extremely sensitive to low number statistics. However, there are also reasons to believe that $$ is affected by some selection bias. In Fig. \\ref{Fig2} we use different symbols for the systems with $N$(HI) $> 10^{21}$ cm$^{-2}$ (empty squares) and $N$(HI) $\\leq 10^{21}$ cm$^{-2}$ (empty circles). One can see from the figure that the systems with high column density have, in general, low metallicity --- an effect originally pointed out by Boiss\\'e et al. (1998). High column density systems are the main contributors to $$ and the lack of such absorbers with [Zn/H] $> -1$ at low $z$ tends to hide the global rise of metallicity with cosmic time. The lack of DLA systems of high column density and metallicity at low redshift is somewhat surprising because (i) clouds with $N$(HI) $> 10^{21}$ atoms cm$^{-2}$ and high metallicity do exist in the disk of our Galaxy and in low redshift spirals; (ii) study of the HI content of the local universe suggest that spirals should be the main contributors to the DLA population at $z\\approx 0$ (Rao \\& Briggs 1993). Nevertheless, spirals are a small fraction of the intervening DLA galaxies observed in low-$z$ imaging studies (Le Brun et al. 1997; Rao \\& Turnshek 1998; see also refs. in Table 1 by Vladilo 1999). This deficiency of spirals suggests the presence of some selection effect. Selection effects that can bias the observed population of DLA absorbers include QSO obscuration by DLA dust (Fall \\& Pei 1993) and gravitational lensing (Smette, Claeskens \\& Surdej 1997). However, also the surface brightness of the intervening galaxies and the galactocentric distances of the clouds intercepted can play a role in affecting the observed population. As discussed in Vladilo (1999), these effects generally conspire to decrease the fraction of chemical enriched regions in the sample population, dust obscuration alone yielding a QSO visual extinction of $\\approx 1$ magnitude when $N$(HI) $> 10^{20.7}$ cm$^{-2}$ at solar metallicity. Considering the likely presence of this bias and the severe dependence of $$ on low number statistics, the lack of evolution of $$ should not be used to conclude that the mean cosmic metallicity of DLA absorbers does not evolve. Comparison between empirical $$ determinations and model predictions of global enrichment of the universe should await a better understanding of the role played by any selection bias and a significant enlargement of the observational data base." }, "0005/astro-ph0005280_arXiv.txt": { "abstract": "We present a critical analysis of the nature of the so-called Li-rich RGB stars. For a majority of the stars, we have used Hipparcos parallaxes to determine masses and evolutionary states by comparing their position on the Hertzsprung-Russell diagram with theoretical evolutionary tracks. Among the twenty Li-rich giants whose location on the HR diagram we were able to determine precisely, five appear to be Li-rich because they have not completed the standard first dredge-up dilution, and three have abundances compatible with the maximum allowed by standard dilution. Thus, these should be re-classified as Li-normal. For the remaining stars, the high Li abundance must be a result of fresh synthesis of this fragile element. We identify two distinct episodes of Li production which occur in advanced evolutionary phases depending upon the mass of the star. Low-mass RGB stars, which later undergo the helium flash, produce Li at the phase referred to as the bump in the luminosity function. At this evolutionary phase, the outwardly-moving hydrogen shell burns through the mean molecular weight discontinuity created by the first dredge-up. Any extra-mixing process can now easily connect the $^3$He-rich envelope material to the outer regions of the hydrogen-burning shell, enabling Li production by the Cameron \\& Fowler (1971) process. While very high Li abundances are then reached, this Li-rich phase is extremely short lived because once the mixing extends deep enough to lower the carbon isotopic ratio below the standard dilution value, the freshly synthesized Li is quickly destroyed. In intermediate-mass stars, the mean molecular weight gradient due to the first dredge-up is not erased until after the star has begun to burn helium in its core. The Li-rich phase in these stars occurs when the convective envelope deepens at the base of the AGB, permitting extra-mixing to play an effective role. Li production ceases when a strong mean molecular weight gradient is built up between the deepening convective envelope and the shell of nuclear burning that surrounds the inert CO core. This episode is also very short lived. Low-mass stars may undergo additional mixing at this phase. The compiled data provide constraints on the time scales for extra mixing and some insight on processes suggested in the literature. However, our results do not suggest any specific trigger mechanism. Since the Li-rich phases are extremely short, enrichment of the Li content of the ISM as a result of these episodes is negligible. ", "introduction": "During the first dredge-up, the lithium abundance at the surface of a red giant star decreases due to dilution of the external convective stellar layers with the lithium-free region in the interior. Depending on the stellar mass and metallicity, the surface lithium abundance decreases with respect to its value at the end of the main sequence by a factor that varies between $\\sim$ 30 and 60. The post dredge-up lithium abundance also depends on the surface depletion of this fragile element during the pre-main sequence and main sequence phases; in Population I stars this is known to be important at masses lower than $\\sim$ 1.2 M$_{\\odot}$ and in stars originating from the Li dip (see volume edited by \\cite{Crane94} for reviews). Starting from the present interstellar medium abundance of log N(Li)$\\simeq$3.3 (where log N(Li) = log[n(Li)/n(H)]+12), one thus expects a post-dilution value lower than about 1.8 to 1.5 for Pop I stars and, indeed, most G-K giants fall below this upper limit (\\cite{Lambertetal80}, \\cite{Brownetal89}, \\cite{Mallik99}). However, about 1$\\%$ of G-K giants show unexpectedly strong lithium lines (\\cite{WS82}, \\cite{Brownetal89}, \\cite{GrattonDAntona89}, \\cite{Pilachowskietal90}, \\cite{Pallavicinietal90}, \\cite{FekelBalachandran93}). Some of these Li-rich giants have abundances that are even higher than the present interstellar medium value (\\cite{DeLaRezaDaSilva95}, \\cite{Balachandranetal00}). Various suggestions have been made to explain the Li-rich giant phenomenon. Some are related to external processes, like the contamination of the external layers of the giant by the debris of nova ejecta or by the engulfing of a planet (\\cite{Alexander67}, \\cite{Brownetal89}, \\cite{GrattonDAntona89}, \\cite{SiessLivio99}). Other explanations explore internal processes, like the preservation of the initial lithium content or fresh lithium production (\\cite{FekelBalachandran93}, \\cite{DeLaRezaetal96}, \\cite{Castilhoetal99}, \\cite{SackmannBoothroyd99}). The other notable disagreement betwen the prediction of abundances in first ascent giants and observations is the carbon isotopic ratio. It has been observed that the carbon isotopic ratio in evolved stars of open clusters with turnoff masses lower than about 2.2M$_{\\odot}$ (\\cite{Gilroy89}, \\cite{GilroyBrown91}) and in field giants at various metallicities (\\cite{Snedenetal86}, \\cite{Shetroneetal93}, \\cite{Pilachowskietal97}, \\cite{cbw:98}, \\cite{Carrettaetal98}, \\cite{Grattonetal2000}) is lower than the value predicted by the standard theory. Observations reveal that the carbon isotopic ratio does not decrease below the standard model predictions until the mean molecular weight gradient produced by the first dredge up is erased by the outwardly-burning hydrogen shell. This evolutionary phase is referred to as the bump in the luminosity function on the HR diagram and corresponds to a temporary decrease in the luminosity and a small increase in the effective temperature of the star when the chemical discontinuity is removed. It has therefore been surmised that a non-standard mixing process, previously inhibited by the mean molecular weight barrier, begins to act at this phase and results in ``extra mixing\" of the convective zone material with regions hot enough to convert $^{12}$C to $^{13}$C (Sweigart \\& Mengel 1979; Charbonnel 1994, 1995; Charbonnel et al. 1998). The nature of the mechanism which produces the drop of the $^{12}$C/$^{13}$C ratio remains uncertain, though rotation-induced mixing is a probable candidate. We do know that (i) the extra-mixing is inhibited by molecular weight gradients because, as our previous discussion showed, it has not been observed to occur before the star enters the bump, (ii) it occurs in $\\sim 96 \\%$ of the low-mass stars (\\cite{CharbonnelDias98}) and (iii) it destroys part of the $^3$He produced on the main sequence (as suggested first by \\cite{rbw84}; see also \\cite{cchar:95}, \\cite{hog:95} and \\cite{SackmannBoothroyd99}). It may be responsible for other chemical anomalies. For instance, a significant decrease in the surface lithium abundance is seen in Population II giants which are at the red-giant bump (\\cite{Pilachowskietal93}, \\cite{cchar:95}, \\cite{Grattonetal2000}). The continuous decline of the carbon abundance along the RGB, and the presence in the atmosphere of material processed by the ONeNa-cycle (see \\cite{kra:94} and \\cite{daco:98} for reviews) seen in globular cluster giants may also result from a manifestation of the extra-mixing process (see \\cite{weissDC00} and references therein). In this paper we draw a connection between the Li enhancement and the carbon isotopic ratio decline in red-giants and provide evidence for the hypothesis outlined in \\S2. For the first time it is shown that two phases of mixing occur in Population I stars depending upon the mass of the star; mixing occurs at the luminosity bump for low-mass stars and in the early-AGB phase before the completion of the second dredge-up in both low and intermediate-mass stars. Both lead to short-lived Li-rich phases. We discuss the similarities between the two events. ", "conclusions": "On the basis of their position on the HR diagram, we are able to separate the so-called Li-rich giants into three different groups. The first group consists of normal stars which have only recently started lithium dilution, or have normal post-dilution lithium abundances, and are thus mis-labelled as Li-rich. The second group contains low-mass stars at the luminosity bump on the RGB. The third group contains low and intermediate-mass stars in the early-AGB phase. The second and third groups support our hypothesis that Li-rich stars are formed by an extra-mixing process which is effective when the convective zone is in close proximity to the hydrogen-burning shell and when the two regions are not separated by a strong gradient of molecular weight. Li production will be followed by a decrease in the $^{12}$C/$^{13}$C ratio as the material mixes to deeper layers. When this occurs the freshly synthesized Li will be steadily destroyed. Other elements processed in the hydrogen-burning shell may be affected and this remains to be studied. As the Li production phase is short and these stars have only a moderate mass loss rate, they are not expected to contribute significantly to the Li enrichment of the ISM." }, "0005/astro-ph0005249_arXiv.txt": { "abstract": "A numerical double burst model of the chemical evolution of gas-rich dwarf galaxies has been developed. The model is fitted to a sample of N/O, O/H, Y and gas fraction observations, where N/O and O/H are the relative abundances by number of nitrogen to oxygen and oxygen to hydrogen, respectively. Y is the abundance by mass of helium. Closed models as well as models including enriched outflow, ordinary outflow and ordinary outflow combined with inflow are considered.The bursts are assumed to be instantaneous but ordered in pairs to explain the scatter in N/O-O/H. The method of gas fraction fitting is revised, and it is found that it is very important to specify whether dwarf irregulars (dIrrs) or blue compact galaxies (BCGs) are considered. Effective enriched winds fail when fitting N/O, whereas closed models, models with ordinary winds or a combination of ordinary winds and inflow are all viable. ", "introduction": "The chemical evolution of dwarf irregular (dIrr) \\footnote{In this paper a dwarf galaxy is defined to have absolute magnitude M$_{\\rm B}\\geq -17$.} and blue compact emission-line galaxies (BCGs) is of particular interest because a substantial body of observational data is available and some degree of simplicity exists because of the low level of `metal' enrichment and absence of large abundance gradients. Furthermore, their wide range of intrinsic properties makes them suitable objects for testing certain expectations from stellar nucleosynthesis theory and the `Simple' or other models of galactic chemical evolution, although at the same time there are complications associated with inflow of unprocessed material, outflow in homogeneous or selective galactic winds and bursting (or `gasping') modes of star formation. Chemical evolution models attempt to apply all these concepts to account for the distribution of different elements, notably helium, oxygen and nitrogen, in relation to star formation rates and gas fractions. Because many parameters such as these last two are generally very poorly determined, the most convincing tests come from the comparison of different elements with one another. Back in the 1970s, Smith \\shortcite{smith:1975}, Peimbert \\shortcite{peimbert:1978} and Edmunds \\& Pagel \\shortcite{edmunds:1978} noticed a contribution of primary nitrogen to the N/O ratio in Galactic and extragalactic H II regions with low oxygen abundance and Edmunds \\& Pagel attributed the existence of scatter in N/O at a given O/H to the existence of a time delay in primary nitrogen production by intermediate-mass stars, combined with differing effective ages of the underlying stellar populations. Alloin et al. \\shortcite{alloin:1979} also noted the primary nitrogen and attributed scatter in N/O to variations in the initial mass function (IMF), whereas Lequeux et al. \\shortcite{leq:1979} in their classic study of helium, nitrogen and oxygen in irregular galaxies and BCGs confirmed the primary nitrogen likewise, but were not convinced that there was any real scatter in their data. The models of Alloin et al. and Lequeux et al. assumed evolution to take place smoothly as a function of time; Matteucci \\& Chiosi \\shortcite{mc:1983} were the first to incorporate into chemical evolution models for these systems the idea of bursting modes of star formation as prevously inferred by Searle \\& Sargent \\shortcite{searle:1972} and Searle, Sargent \\& Bagnuolo \\shortcite{bagnuolo:1973}, and interpreted on the basis of the SSPSF hypothesis by Gerola, Seiden \\& Schulman \\shortcite{gerola:1980}. The basic pattern of a primary (constant N/O) pattern at low metallicities in H~II regions changing over to a secondary pattern (N/O $\\propto$ O/H) at higher ones has been confirmed in many more recent investigations (e.g. Vila-Costas \\& Edmunds 1993; van Zee, Salzer \\& Haynes 1998). The chemical evolution of dIrrs and BCGs has been studied in many more recent investigations. Matteucci \\& Tosi \\shortcite{mt:1985} found good fits to the data with a Salpeter IMF, inflow, homogeneous outflow, bursting star formation and a choice of third dredge-up parameters from Renzini \\& Voli \\shortcite{RV:1981}, attributing scatter in the N/O ratio to variations in $M_{\\rm up}$, the upper limit to the masses of stars undergoing the third dredge-up with hot-bottom burning. Garnett \\shortcite{garnett:1990} indicated schematically how the occurrence of bursts could in itself lead to variations in the N/O ratio just as a result of observing systems at different stages in the burst cycle. Pilyugin \\shortcite{pilyugin:1992,pilyugin:1993} developed similar ideas in quantitative numerical models involving self-enrichment of H II regions and selective galactic winds as well as bursting modes of star formation, and Marconi, Matteucci \\& Tosi \\shortcite{marconi:1994} also developed models with bursts and selective winds, while Carigi, Col\\' in \\& Peimbert \\shortcite{carigi:1998} have investigated similar models, but prefer a `bottom-heavy' IMF similar to one claimed in some globular clusters and giving rise to low true yields. However, part of the motivation for invoking selective winds was the apparent existence of a large $dY/dZ$ ratio suggested by Pagel et al. \\shortcite{pagel:1992}, which no longer seems valid \\cite{izotov:1998}, and the scatter in N/O also seems to have been overestimated in those investigations. In this paper, therefore, we investigate the problem again, making use of more recent data and models of stellar nucleosynthesis and exploring in particular the role of mixing processes and of differing burst phases in leading to scatter in the N/O, O/H relation and the relationship between oxygen abundance and gas fraction. The structure of the article is as follows: Section\\ \\ref{abundance} presents the adopted sample of abundance observations. Section\\ \\ref{mix} discusses the evolution of \\hbox{H\\,{\\sc ii}}-regions, wind-driven bubbles and supernova-driven supershells to investigate possible mixing scenarios. This leads to the description of our double-bursting models in section\\ \\ref{model}. The results of fitting the models to the sample of observations are presented in section\\ \\ref{result}, and discussed in section\\ \\ref{conc}. ", "conclusions": "\\label{conc} Numerical models calculating the chemical evolution of gas-rich dwarf galaxies have been presented. The models have been fitted to a sample of abundance and gas fraction observations. A chemical evolutionary model has to fit these observations simultaneously. It is clear that a realistic model is not just a couple of equations including a lot of parameters that one can change until all data are fitted. One has to remember that the equations are applied to a physical system, obeying physical laws. Hence, before doing any calculations, some considerations were made about ejecta dispersal and mixing processes. It was found that the processes involved are complicated and no complete theory exists yet. However, the four most important features in the enrichment process are found to be stellar ejecta, giant \\hbox{H\\,{\\sc ii}}-regions, wind-driven superbubbles and SN-driven supershells. Wind-driven superbubbles arise shortly after a burst due to the strong stellar winds of massive stars. The radii of these superbubbles are in general smaller than the radii of the \\hbox{H\\,{\\sc ii}}-regions. As the massive stars explode as SNe, a supershell is swept-up, soon catching up on the superbubble. However, a simple calculation showed that their radii become comparable to the radii of \\hbox{H\\,{\\sc ii}}-regions only in the late stages of \\hbox{H\\,{\\sc ii}}-region existence. Hence, it is suggested that observed emission lines, used for abundance determinations, arise in the ionized medium outside of the superbubble/supershell. Numerical hydrodynamical models indicate that SN-ejecta always stay within the supershell. If this is true, the observed abundances are not affected by the SN-ejecta from the stars producing the \\hbox{H\\,{\\sc ii}}-region, hence being typical for the ISM. The details of supershell evolution are still not known, but both theory and observations support star formation in expanding shells. From theory, it is expected to start not earlier than about 20 Myr after the starburst, resulting in star formation involving a mass comparable to the mass of the burst that initiated the supershell, hence appearing as a double burst. The double bursting mode of our numerical model ensures the appearance of scatter in the N/O-O/H plane, according to the time-delay idea, namely that N is released some time after O. The time interval between the two bursts of a pair is tuned to give maximum scatter. It is found that this timescale is comparable to the timescale of star formation in an expanding supershell. The requirement for the 'shell burst' to be the second pair-burst is that the O-rich ejecta mix into the supershell. Because of the poorly understood physics of supershells and mixing processes, this assumption should be seen so far as a working hypothesis. All bursts are assumed to be instantaneous, hence representing short but intense star formation events. The closed model is able to explain the observations of N/O-O/H in both scatter and level. If assuming the upper limit of hot bottom burning to be 5$M_{\\odot }$, $\\alpha $=1.1 is used to explain the observations, where $\\alpha $ is the mixing length parameter. This is a rather low value compared to the results of recent works however \\cite{van_den_hoek:1997,marigo:1998}, favouring a value close to 2. Using yield set 3 (the Padova set), it is necessary to extrapolate the primary N yields below Z=0.008 to obtain a higher N yield. Otherwise, the level of N/O becomes too low compared to the observations. The primary N production has to be increasing with decreasing metallicity. Calculations of stellar yields for metallicities lower than Z=0.008 are definitely desired to quantify this.. A very important conclusion is that no primary N production in massive stars is needed to explain the observations. Intermediate mass stars are in position to produce a sufficient amount of primary N. For the Y-O/H observations, a linear trend is visible, and a linear fit gives $dY/dZ=2.63\\pm 2.21$ and $Y_p=0.238\\pm 0.004$. These values are consistent with those of Izotov et al. \\shortcite{izotov:1997} within the uncertainties. The closed model is found to explain the Y-O/H observations perfectly, only with different slopes, depending on the yield set in use. In all cases the slopes are within the uncertainties. As seen by inspection of the true He yields for the three sets in table\\ \\ref{totalyield}, one finds the explanation for the slope difference to be that the He yields of set 3 are 2-3 times higher than those of set 2a or 2b. It is noteworthy that the Y-O/H relation was fitted using exactly the same parameters as for the N/O-O/H fitting. The problem arises when fitting O/H-$\\mu $ data. It is argued that the observed gas fractions are actually lower limits, because dark matter is implicitly included in dynamical mass estimates and molecular hydrogen ignored. Hence, it is found useful to calculate upper limits using M/L estimates from starburst evolutionary models. For most objects, very extreme upper limits have to be used, assuming the galaxies to experience maximum luminosity of their first burst, except for three galaxies where the known star formation histories allow us to adopt more moderate and realistic upper limits. If star formation histories are found for a larger sample of dIrrs and even BCGs constraining chemical evolution models with observed gas fraction intervals may eventually turn out to be extremely useful. The closed models do not reproduce the gas fractions of the three well-known objects, even if a lower IMF-cutoff, equal to 0.01 $M_{\\odot }$ instead of 0.1 $M_{\\odot }$ is adopted. Hence, open models are considered, allowing gas to escape or to accrete on to the galaxy. Two kind of winds, enriched and ordinary, have been used. The results when incorporating enriched winds are not in accordance with the observed level of N/O and the Y-O/H fitting is not satisfactory, when the gas fractions of the three well-known systems are fitted. Hence, it is concluded that our models employing enriched winds is in conflict with the observations. The next step is to include ordinary instead of enriched winds. Both N/O-O/H and Y-O/H are fitted, with results resembling those of the closed model. Unfortunately, only the gas fraction of one of the three selected galaxies is fitted, not differing much from the results of the closed model. To check the inclusion of ordinary winds, the outcome of the model is compared to the results of a simple analytical model, employing continuous star formation. A close resemblance is found between the two models at low metallicities, but at higher metallicities, the numerical model seems to have problems in getting the yield down. The difference may be caused by the behavior of starbursts at a low absolute gas mass. This is confirmed when comparing the numerical model, changed slightly to employ continuous star formation, to the analytical model, displaying almost identical outputs. Thus, it is concluded that it is important to specify clearly whether instantaneous bursts or continuous star formation is used, when including ordinary winds. Instantaneous bursts resemble the intense bursts of BCGs, whereas the more moderate bursts of dIrrs are better explained using a continuous SFR. Finally, inflow and ordinary winds were included. The results are in accordance with all observations, except that it is difficult to obtain the right gas fractions and N/O scatter simultaneously. Only for quite extreme parameter choices as in fig.\\ \\ref{gaslargescatter} and \\ref{NOlargescatter}, one may be succesful. It may be important to note that the results, when using the combined inflow/ordinary wind model, show the upturn in N/O to be more pronounced, than it was for the closed model. One question is unavoidable: which model is preferred? It is impossible to give an unambiguous answer. Dwarf galaxies are different, both in mass and appearance. Some are explained well by a closed model, others need a combination of ordinary winds and inflow. This is true for both dIrrs an BCGs. However, for dIrrs one should prefer to adopt a more continuous SFR before fitting the observations. In all cases, the model including enriched winds seems to be ruled out, since it is in direct conflict with the observations, as also found by e.g. Carigi et al. \\shortcite{carigi:1998}." }, "0005/astro-ph0005005_arXiv.txt": { "abstract": " ", "introduction": "Since the discovery of the events of gamma-ray bursts (GRBs) about thirty years ago (Klebesadel et al. 1973), many achievements have been obtained, but the full comprehension of the objects seems still to be a longstanding problem. Among the many efforts, investigating statistical properties of the events is as necessary as poking into the details of the bursts. Since more and more data of GRBs have been available (e.g., Fishman et al. 1994; Meegan et al. 1994; Meegan et al. 1996; Meegan et al. 1998; Paciesas et al. 1999), statistical results become more and more reliable. Possible correlations among various parameters of GRBs were studied previously (e.g., Golenetskii et al. 1983; Barat et al. 1984; Belli 1993). Investigations of the issues were continued recently with more sizable sets of data (e.g., Mallozzi et al. 1995; Dezalay et al. 1997; Belli 1999). With a large number of bursts observed with BATSE, Fishman (1999) found that the hardness-duration correlation, which had been described previously, was confirmed. In the following, we will make a further investigation on this issue. ", "conclusions": "" }, "0005/astro-ph0005233_arXiv.txt": { "abstract": "An analytical understanding of the strongly nonlinear regime of gravitational collapse has been difficult to achieve. The only insight has been the stable clustering hypothesis, which assumes that the number of neighbors for objects averaged over small length scales is constant in time. Our recently proposed analytic halo model for $N$-point correlation functions now provides a tool for calculating gravitational clustering properties in the strongly nonlinear regime. This model also provides a new physical framework for an independent evaluation of the validity of the stable clustering hypothesis. We derive the asymptotic nonlinear behavior of the $N$-point correlation functions and pairwise peculiar velocities in terms of dark matter halo properties. We show that these statistics exhibit stable clustering only when the halo mass function and halo density profile obey specific relations. The long-cherished stable clustering hypothesis therefore is not necessarily realized in practice. ", "introduction": "Gravitational clustering is the fundamental physical process responsible for the formation and evolution of structure in the universe. Galaxy-hosting dark matter halos are the products of the strongly nonlinear stage of this process, but a detailed understanding of this important regime has mostly eluded us because few numerical simulations have the dynamic range to explore comfortably such a small length scale (e.g., Moore et al. 1999; Bullock et al. 1999). The one analytic handle has been the stable clustering assumption (Peebles 1974; Davis \\& Peebles 1977; Peebles 1980), which has allowed us to extrapolate into nonlinear regimes beyond the reach of numerical simulations. Predictions of the stable clustering hypothesis are: (1) the two-point correlation function of the density field $\\xi(r)$ and the power spectrum $P(k)$ are power laws, with $\\xi(r) \\propto r^{-\\gamma}$ and $\\Delta(k)\\equiv 4\\pi k^3 P(k) \\propto k^\\gamma$, where $\\gamma = {(9+3n)/(5+n)}$ for a primordial spectral index $n$; (2) the higher order $N$-point correlation functions $\\xi_N(r)$ scale as $\\xi_N(r) \\propto r^{-\\gamma_N}$, with $\\gamma_N = (N-1)\\gamma$, or $ \\xi_N \\propto \\xi^{N-1} $; and (3) the pairwise peculiar velocity exactly cancels the Hubble expansion on small scales, $v/Hr=-1$, so that bound, high-density halos maintain a fixed physical size. Some of these predictions have agreed well enough with numerical simulation results (Jain 1997 and references therein) to have obtained general acceptance. In two recent papers we presented evidence from high-resolution $N$-body simulations that gravitational clustering does not necessarily follow the scaling $\\xi_3\\propto \\xi^2$ required by the stability condition for the two- and three-point functions $\\xi$ and $\\xi_3$ (Ma \\& Fry 2000a). For a deeper understanding of the strongly nonlinear regime beyond simulations, we proceeded to construct an analytic halo model (Ma \\& Fry 2000b) in which mass is distributed in spherical dark matter halos with phenomenological mass distribution functions, density profiles, and halo-halo correlations. We showed that this halo model can reproduce analytically the two- and three-point correlation functions measured in numerical simulations, and that it also makes useful predictions beyond the limited range of validity of simulations. In this {\\it Letter} we investigate the extent to which the analytic halo model is consistent with stable gravitational clustering in the strongly nonlinear regime. In particular, we derive the halo model predictions for the asymptotic nonlinear behavior of the $N$-point correlation functions (\\S 2) and the pairwise peculiar velocities (\\S 3) in terms of dark matter halo properties. We then obtain the conditions that must be satisfied by the halo mass function and density profile in order to reproduce results of the stable clustering hypothesis. ", "conclusions": "In this {\\it Letter}, we have derived the asymptotic nonlinear behavior of the $N$-point correlation functions $\\xi_N$ (eqs.~[\\ref{gamma_model}], [\\ref{gamma3_model}], [\\ref{gammaN_model}]) and the pairwise peculiar velocities $-v/Hr$ (eq.~[\\ref{h_model}]) in the framework of the recently proposed analytic halo model of Ma \\& Fry (2000b). We have shown that their small scale behavior is consistent with the stable clustering hypothesis only if dark matter halos satisfy certain criteria. The two halo parameters whose variation we have explored are the logarithmic slope $\\alpha$ of the halo mass distribution $dn/dM$ at the low mass end in eq.~(\\ref{alpha}) and the slope $\\beta_0$ for the mass dependence of the halo concentration parameter $c(M)$ in eq.~(\\ref{beta}) (recall $\\beta\\approx 0.8\\beta_0$). These two parameters are highlighted because results to date from numerical simulations have indicated a significant uncertainty, with $0.4\\la \\alpha \\le 1$ and $0\\la \\beta \\la 1/2$ being plausible values. From the derived asymptotic nonlinear behavior, we have obtained analytically the relations that must be satisfied by $\\alpha$ and $\\beta$ in order for stable clustering to occur. Equations~(\\ref{stable2}), (\\ref{stable3}), and (\\ref{stableN}) summarize the results for the $N$-point correlation function $\\xi_N$. These equations and Figure~1 indicate that although the two-point stable clustering condition alone is satisfied by an infinite set of $(\\alpha,\\beta)$ for a given $n$, the three- and higher-point stable clustering conditions, $\\xi_N\\propto \\xi^{N-1}\\,, N\\ge 3$, imply $\\alpha$ must be zero. Achieving stable clustering to all orders therefore requires stringent conditions: $\\alpha=0$ and $\\beta=(3+n)/6$. For non-scale-free models (e.g., cold dark matter), $n$ is the index at the high-$k$ end of the power spectrum. Compared with the hierarchical model of Davis \\& Peebles (1977), results from this paper and Ma \\& Fry (2000b) show that the analytic halo model makes more general and physical predictions for the behavior of the $N$-point correlation functions and the pairwise velocities. Imposing the stability condition $\\alpha=0$ and $\\beta=(3+n)/6$ in equations~(\\ref{gamma_model}), (\\ref{gamma3_model}), and (\\ref{gammaN_model}), we indeed recover their results of $\\xi\\propto r^{\\gamma}$ with $\\gamma=(9+3n)/(5+n)$ at small $r$ and $\\xi_N \\propto \\xi^{N-1}$. The condition $\\beta=(3+n)/6$ can be understood further by examining individual halos. Using $c\\propto M_*^{\\beta_0}$, $M_*\\propto a^{6/(n+3)}$, and $\\beta_0\\approx \\beta$ up to a logarithmic factor, we see that $\\beta=(3+n)/6$ gives approximately $c\\propto a(t)$, which is consistent with the simulation results in Bullock et al. (2000). The halo scale radius $R_s$ is then approximately constant in time in physical coordinates, as expected for a stable system, and the physical density $\\rho$ of individual halos is indeed very close to constant in time, more so for very small halos that form early ($\\nu \\gg 1$, $ c \\gg 1$) than for those that form recently ($\\nu \\approx 1$, $c\\approx 1$). Yano \\& Gouda (1999) have found that for halo density $\\rho\\propto r^{-\\epsilon}$ with $\\epsilon\\le 3/2$ (such as Navarro et al. and Moore et al.), the velocity parameter $v/Hr$ approaches 0 and the two-point correlation function $\\xi$ approaches a constant (i.e. $\\gamma=0$) in the nonlinear limit. Neither behavior is seen in our Figures because their derivation ignores the mass distribution function $dn/dM$ and is therefore valid only for equal mass halos. Our earlier paper (Ma \\& Fry 2000a) has already shown signs of departure from the stable clustering hypothesis in high resolution $N$-body simulations. In this paper, the specific values $\\alpha=0$ and $\\beta=(3+n)/6$ required for stable clustering provide additional evidence that this long-cherished hypothesis may not be applicable in all situations. In such cases, one consequence is that the frequently used fitting formulas for the nonlinear power spectrum (Hamilton et al 1991; Jain et al. 1995; Peacock \\& Dodds 1996; Ma 1998) will need modifications at high $k$, e.g., $k/k_{nl}\\ga 50$ for the $n=-2$ model and $k\\ga 20\\,h$ Mpc$^{-1}$ for cold dark matter models with a cosmological constant (see Figs.~3 and 4 of Ma \\& Fry 2000b). We believe the analytic halo model is a powerful tool that is providing new insight into the nonlinear regime of gravitational clustering, the most fundamental process in cosmology. C.-P. M thanks Jim Peebles and Uros Seljak for useful discussions. She acknowledges support of an Alfred P. Sloan Foundation Fellowship, a Cottrell Scholars Award from the Research Corporation, a Penn Research Foundation Award, and NSF grant AST 9973461. \\clearpage" }, "0005/astro-ph0005469_arXiv.txt": { "abstract": "% We have developed a technique to map the three-dimensional structure of the local interstellar medium using a maximum entropy reconstruction technique. A set of column densities ${\\bf N}$ to stars of known distance can in principle be used to recover a three-dimensional density field ${\\bf n}$, since the two quantities are related by simple geometry through the equation ${\\bf N} = {\\sf C} \\cdot {\\bf n}$, where ${\\sf C}$ is a matrix characterizing the stellar spatial distribution. In practice, however, there is an infinte number of solutions to this equation. We use a maximum entropy reconstruction algorithm to find the density field containing the least information which is consistent with the observations. The solution obtained with this technique is, in some sense, the model containing the minimum structure. We apply the algorithm to several simulated data sets to demonstrate its feasibility and success at recovering ``real'' density contrasts. This technique can be applied to any set of column densities whose end points are specified. In a subsequent paper we shall describe the application of this method to a set of stellar color excesses to derive a map of the dust distribution, and to soft X-ray absorption columns to hot stars to derive a map of the {\\it total} density of the interstellar medium. ", "introduction": "% Given a set of projected density measurements of columns with known end points, one can, at least in principle, recover the three dimensional distribution of absorbing material. One way to accomplish this deprojection is to simply draw isocolumn contours \\citep{sfeir,welsh}. Here we propose an alternative technique -- maximum entropy reconstruction (\\MER) -- and apply a computer implementation of the algorithm to simulated data sets. We test its ability to recover structure in the interstellar medium (\\ISM) by varying several parameters and the functional form of the entropy. We then briefly sketch its application to a number of archived data sets, to be presented in a subsequent paper. ", "conclusions": "% Maximum entropy reconstruction is a powerful technique for recovering the structure of the \\ISM\\ from column density data sets. By modulating the cell size, one can find a solution which recovers the underlying structure in the \\ISM. The lower limit to the size scale of detected structures is set by the stellar sampling density. The reconstruction algorithm is insensitive to the particular form of the entropy employed, so it is most prudent to use computationally simple and rapidly converging forms such as the quadratic entropy (equation \\ref{eq14}). The next step is to apply this technique to existing astronomical data sets. Different components of the \\ISM\\ can be traced by absorption and emission measurments made at different wavelengths. Using a variety of absorptive and emissive tracers, we should be able to map regions of varying ionization and molecular state and elemental composition, and compare their spatial distributions. We are currently working to map the distribution of dust in the \\ISM\\ using column densities derived from stellar color excess measurements. In addition, we are using archived \\ROSAT\\ observations of hot stars to determine X-ray absorption columns. Since the soft X-ray cross section is insensitive to the molecular and (modest) ionization state of the absorber \\citep{arabadjis}, reconstructing the density field using X-ray columns will map the {\\it total} density of material the local \\ISM. Additionally, by exploiting the spectral resolution of instruments aboard XMM-Newton and Chandra, one may be able to separate this material by ionization state. For example, $\\Delta E/E \\sim 50$ for the non-dispersive ACIS detector aboard Chandra, resulting in a spectral resolution of about 10 eV at the prominent absorption edge of oxygen at 0.5 keV. Since the neutral and first three ionization states of oxygen are separated by about 20 eV each (the potentials are 13.6, 35.1, and 54.9 eV), it may be possible to distill the oxygen distribution into its dominant ionization states. Several \\ISM\\ tracers also contain velocity information. Optical absorption lines in stellar spectra, originating from the \\ISM, provide a measure of the distribution of metals not bound up in dust grains, and maps constructed from their columns may provide insight into the gas phase chemistry of \\ISM. Additionally, the velocity structure of a resolved absorption line can be used in conjuction with a rotation curve and a second application of the maximum entropy principle to further constrain the distribution of material along the line of sight. Similarly, velocity-resolved 21 cm emission maps can be used to construct a distribution of neutral hydrogen along each line of sight. The applicability of this technique to absorption measurements at almost any wavelength comes at a particularly fortuitous time. Most astronomical satellite programs support an on-line archive containing their public domain observations, many of which can be mined for projected density measurements. With the number number of of these archives growing on an almost daily basis, this technique should find wide application in the near future. The authors would like to acknowledge support from NASA grant NAG5-3247. JSA would like to thank Doug Richstone for useful discussions. \\clearpage" }, "0005/astro-ph0005143_arXiv.txt": { "abstract": "We report on a three-year program of coordinated X-ray and optical monitoring of the narrow-line Seyfert 1 galaxy NGC~4051. The rapid continuum variations observed in the X-ray spectra are not detected in the optical, although the X-ray and optical continuum fluxes are correlated on time scales of many weeks and longer. Variations in the flux of the broad \\Hbeta\\ line are found to lag behind the optical continuum variations by approximately 6 days (with an uncertainty of 2--3 days), and combining this with the line width yields a virial mass estimate of $\\sim1.1 \\times 10^6$\\,\\Msun, at the very low end of the distribution of AGN masses measured by line reverberation. Strong variability of \\4686\\ is also detected, and the response time measured is similar to that of \\Hbeta, but with a much larger uncertainty. The \\4686\\ line is almost five times broader than \\Hbeta, and it is strongly blueward asymmetric, as are the high-ionization UV lines recorded in archive spectra of NGC~4051. The data are consistent with the Balmer lines arising in a low-inclination (nearly face-on) disk-like configuration, and the high-ionization lines arising in an outflowing wind, of which we observe preferentially the near side. During the third year of monitoring, both the X-ray continuum and the \\4686\\ line went into extremely low states, although the optical continuum and the \\Hbeta\\ broad line were both still present and variable. We suggest that the inner part of the accretion disk may have gone into an advection-dominated state, yielding little radiation from the hotter inner disk. ", "introduction": "Beginning in early 1996, the International AGN Watch\\footnote{All International AGN Watch papers and data are available at the AGN Watch website at URL {\\sf http://www.astronomy.ohio-state.edu/$\\sim$agnwatch/}.} undertook a program of contemporaneous X-ray and optical spectroscopic monitoring of the galaxy NGC~4051, the only NLS1 galaxy in Seyfert's \\cite{Sey43} original list of high surface-brightness galaxies with strong emission lines. The X-ray variability characteristics of NGC~4051 are typical of the NLS1 class \\cite{Law87,McH95}. X-ray observations were made with the {\\em Rossi X-Ray Timing Explorer (RXTE)}, and optical spectra were obtained with the Perkins 1.8-m telescope at Lowell Observatory and the 1.6-m Tillinghast Reflector on Mt.\\ Hopkins. The purpose of this program has been twofold: \\begin{enumerate} \\item To determine the nature of the relationship between the X-ray and UV--optical continuum variations. This is a particularly interesting question in the case of NGC~4051 since the X-ray flux dropped to an extremely low level towards the end of this campaign \\cite{Utt99}. \\item To determine the size of the broad-line region (BLR) and virial mass of the central object via reverberation techniques \\cite{Bla82,Pet93}. \\end{enumerate} The emission-line cross-correlation lag $\\tau$ can be taken to be the light-travel time across the BLR, so the BLR size is given by $r=c\\tau$. By combining this with the emission-line width $\\vFWHM$, the mass of the central source can be inferred to be \\begin{equation} M = \\frac{f \\vFWHM^2 c\\tau}{G}, \\end{equation} where $f$ is a factor of order unity that depends on the still unknown geometry and kinematics of the BLR. There is an implicit assumption that the gravitational force of the central object dominates the kinematics of the BLR; this is formally unproven, but at least in the case of the well-studied Seyfert 1 galaxy NGC~5548, the reverberation-mapping data are consistent with the required $\\vFWHM \\propto r^{-1/2}$ relationship \\cite{Pet99}. Virial mass estimates based on reverberation-mapping data are now available for nearly 40 AGNs \\cite{Wan99,Kas00}. ", "conclusions": "\\subsection{The Continuum} It has already been pointed out based on these same {\\em RXTE} data \\cite{Utt99} that the X-ray continuum of NGC~4051 virtually ``turned off'' in early 1998. However, the optical spectroscopic data show that the optical continuum and emission lines (and therefore, by inference, the ionizing UV continuum) did not disappear at the same time. Figure 2a shows the {\\em RXTE} light curve. The two lower panels of Fig.\\ 2 show the optical rms spectra based on data obtained during the intervals marked on Fig.\\ 2a. Figure 2b shows the rms spectrum obtained during a period of relatively active X-ray and optical continuum variability. The rms spectrum in Fig.\\ 2c is based on spectra obtained during the period when the X-rays were about an order of magnitude fainter than normal. Figure 2b shows that when the X-ray flux was high and variable, the optical continuum and the \\Hbeta\\ and \\4686\\ emission lines were also variable. Figure 2c, however, shows that when the X-rays were in a faint state, both the optical continuum and the \\Hbeta\\ line were still varying, but the \\4686\\ emission line was no longer varying. A possible interpretation of the behavior of NGC~4051 is that the inner X-ray producing part of the accretion disk has entered an advection-dominated accretion-flow (ADAF) state, in which radiation is emitted with very low efficiency \\cite{Nar94,Nar98}. The outer part of the disk, which produces the UV--optical continuum and drives the emission lines, remains relatively unaffected, however. The implication is that there is a transition radius inside of which the disk is an ADAF and outside of which it radiates efficiently, perhaps like a classical thin disk \\cite{Sha73}, and the persistence of the optical continuum and the \\Hbeta\\ emission line suggests that this transition radius is somewhere between the regions that are most responsible for the soft X-rays and the H-ionizing continuum. \\begin{figure} \\psfig{figure=peterson_fig_2.ps,width=12cm,angle=0} \\caption{Top panel shows the X-ray light curve as in Fig.\\ 1. Bottom left panel shows the optical rms spectrum based on data obtained between JD2450183 and JD2450262. Bottom right panel shows the optical rms spectrum based on data obtained between JD2450810 and JD2451022.} \\end{figure} \\subsection{The Virial Mass and Implications for NLS1s} Reverberation-based size estimates for the broad emission lines and resulting virial mass estimates provide a potential means of distinguishing among the various NLS1 models. In Fig.\\ 3a, we show the relationship between the BLR radius as measured from the Balmer-line lags as a function of the optical continuum luminosity for all AGNs with Balmer-line lags known to reasonable accuracy \\cite{Wan99, Kas00}. This compilation contains six additional AGNs that could be classified as NLS1s as they meet the criterion $\\vFWHM \\ltsim 2000$\\,\\kms. The best-fit regression line ($R_{\\rm BLR} \\propto L^{0.62 \\pm 0.02}$), based on all objects (excluding NGC~4051), is shown as a dotted line. NGC~4051 lies approximately 2.8$\\sigma$ above this regression line, although all the other narrow-line objects clearly fall in the locus defined by the AGNs with broader lines. Given the large dispersion in this relationship, reflected in the high $\\chi^2_{\\nu} (= 15.7)$ of the fit, the offset of NGC~4051 is not unusual. \\begin{figure} \\psfig{figure=peterson_fig_3.ps,width=12cm,angle=0} \\caption{(a). The relationship between size of the Balmer-line emitting region and optical luminosity. The dotted line is a least-squares fit to all the data, excluding NGC~4051. (b). The relationship between virial mass and optical luminosity. The dashed line is a least-squares fit to the narrow-line objects only, and the dotted line is a least-squares fit to all of the other points. In both panels, the filled circles are Seyfert 1 galaxies and the open circles are QSOs. The large triangles are those AGNs in which \\Hbeta\\ has a half width less than 2000\\,\\kms. NGC~4051 is shown as a filled diamond.} \\end{figure} Figure 3b shows the mass--luminosity relationship for these AGNs. We show (a) the best-fit regression line based on all objects {\\em except} the seven narrow-line objects and (b) that based on the narrow-line objects alone. These two fits are separated by typically an order of magnitude in black-hole mass; the black holes in the narrow-line objects are about a factor of 10 lower than those of other AGNs of comparable luminosity. How well do these results allow us to distinguish among the various explanations for the NLS1 phenomenon? We consider the possibilities: \\begin{enumerate} \\item {\\em Do the BLRs of NLS1s have anomalously large radii?} The position of NGC~4051 in Fig.\\ 3a might suggest that this is possible, but the distribution of other narrow-line objects does not support this. Furthermore, as noted above, the scatter in the BLR-radius luminosity relationship is very large, and NGC~4051 is in a statistical sense not the largest outlier in this relationship (simply because other sources have smaller uncertainties in their measured lags). \\item {\\em Are NLS1s simply low-inclination systems?} If the BLR is a flattened system, at low inclination (i.e., nearly face-on) the line widths will be decreased by a factor $\\sin i$, but the measured emission-line lags will be relatively unaffected. On the other hand, assuming that the UV--optical continuum arises in an accretion disk at the same inclination, the apparent UV--optical luminosity is higher at lower inclination. Thus, relative to similar sources at intermediate inclination, the masses of low-inclination sources will be underestimated, and their luminosities will be overestimated, displacing the narrow-line objects in Fig.\\ 3b towards the lower right. This is generally consistent with the location of all seven of the narrow-line objects, including NGC~4051. The line transfer function for \\Hbeta\\ would provide a more definitive test of this hypothesis since it would allow determination of the inclination of the system. This would require more and better data than we have obtained in this experiment. \\item {\\em Are NLS1s undermassive systems with relatively high accretion rates?} Again, the distribution of the narrow-line objects, including NGC~4051, in Fig.\\ 3b is consistent with this hypothesis. The narrow-line sources on this plot lie below the mass-luminosity relationship for other AGNs, at the lower end of the envelope around this relationship. \\end{enumerate} In summary, the hypothesis that NLS1s have unusually distant BLRs for their luminosity is probably not viable in general, although it could apply to the specific case of NGC~4051. At the present time, however, we cannot distinguish between the low-inclination and low-mass, high accretion-rate hypotheses on the basis of the reverberation results alone, although considerations based on X-ray data favor the latter. Indeed, it is entirely possible that both effects (i.e., low inclination and low black-hole mass) contribute. An improvement in the quality and quantity of the optical spectroscopic data could allow determination of the \\Hbeta\\ transfer function, which could allow discrimination between these competing models. The differences between the characteristics of the \\Hbeta\\ emission line on one hand and of the high-ionization lines such as \\4686\\ on the other suggests a two-component BLR, which has been proposed on numerous occasions on other grounds \\cite{Col88}. In this particular case, an interpretation that is at least qualitatively consistent with all the data and relatively simple is that the Balmer lines arise primarily in material that is in a flattened disk-like configuration at a low inclination (to account for the narrow width of the \\Hbeta\\ line), and the high-ionization lines arise in an outflowing polar wind, of which we see preferentially the component on the near side of the disk (to account for high velocity and blueward asymmetry) \\cite{Col88}. If indeed the \\Hbeta\\ emission arises primarily in a low-inclination disk, our virial mass estimate of $M=1.1\\times10^6$\\,\\Msun\\ might seriously underestimate the black-hole mass. It could then be inferred that the NLS1 class might be best explained as low-inclination rather than low-mass, high accretion-rate systems. The strong rapid X-ray variability of NLS1s seem to favor the latter explanation, but we note that at least some low-inclination accretion-disk models predict relatively strong, variable EUV/soft X-ray fluxes \\cite{Net87, Mad88}, consistent with observations of NLS1s." }, "0005/gr-qc0005104_arXiv.txt": { "abstract": "We have investigated cosmological models with a self-interacting scalar field and a dissipative matter fluid as the sources of matter. Different variables are expressed in terms of a {\\it generating function}. Exact solutions are obtained for one particular choice of the {\\it generating function} The potential corresponding to this generating function is a standard tree-level potential arising in the perturbative regime in quantum field theory. With suitable choice of parameters, the scale factor in our model exhibits both inflationary behaviour in the early universe as well as an accelerating phase at late times with a decelerating period in between. It also satisfies the constraints for primeval nucleosynthesis and structure formation and seems to solve the cosmic coincidence problem. The solution exhibits a attractor nature towards a asymptotic de-sitter universe. ", "introduction": "A number of recent observations \\cite{flat} suggest that the $\\Omega_m$, the ratio of the matter density(baryonic+dark) to the critical density, is significantly less than unity suggesting that either the universe is open or that there is some other sources of this missing energy which makes $\\Omega_{total}\\sim 1$. The recent findings of BOOMERANG experiments \\cite{BOOM} strongly suggests the second possibility of a flat universe. At the same time, the measurements of the luminosity-redshift relations observed for the 50 newly discovered type Ia supernova with redshift $z>0.35$ \\cite{super} indicate that at present the universe is expanding in an accelerated fashion suggesting a net negative pressure for the universe. Initial suggestions were to identify this missing energy density to a cosmological constant $\\Lambda$\\cite{Rbah,Rsah}. For a flat matter dominated universe with $\\Lambda$ in Einstein gravity observations strongly suggest $\\Omega_\\Lambda\\sim 0.72$. However, this possibility that $\\Lambda$ could be the dominant energy density has the drawback that the energy scale involved is lower than normal energy scale predicted by the most particle physics models by a factor of $\\sim 10^{-123}$. An alternative source of energy density that may be admissible for this acceleration could be a dynamical $\\Lambda$\\cite{Rcal} in the form of a scalar field with some self interacting potential \\cite{Rpeebetc}. If the energy density of this kind of source varies slowly with time, it mimics an effective cosmological constant. The idea of this candidate, called {\\it quintessence} \\cite{Rcal}, is borrowed from the inflationary paradigm of the early universe. The difference, however, is that this new field evolves at a much lower energy scale. The energy density of this field, though dominant at present epoch, must remain sub-dominant at very early stages and should have evolved in such a way that it becomes comparable to the matter density $\\rho_m$ now. When quintessence is modeled using a minimally coupled scalar field, in general, parameters need to be fine-tuned so as to ensure that $\\rho_m$ and $\\rho_{\\phi}$ are of the same order today. This fine tuning problem has been termed as the cosmic coincidence problem. A new form of quintessence field called ``{\\it tracker field}''\\cite{Rzla} has been proposed to solve the cosmic coincidence problem. It has an equation of motion with an attractor like solution in the sense that for a wide range of initial conditions the equation of motion converges to the same solution. There are a number of quintessence models which have been suggested and most of these involve scalar fields with minimal coupling with potentials dominating over the kinetic energy of the field. A purely exponential potential is one of the widely studied cases \\cite{Rfer}. Inspite of the several advantages the energy density is not enough to make up for the missing part. Inverse power law is another form of the potential (\\cite{Rpeebetc}-\\cite{Rzla}) that has been considered extensively for quintessence models, in particular, for solving the cosmic coincidence problem. Though many of the problems are resolved successfully with this potential, the predicted value for the equation of state for the quintessence field, $\\gamma_Q$, is not in good agreement with the observed results. In search of suitable models that would eliminate the problems, new types of potentials, like $V_0[\\cos h \\lambda\\phi-1]^p$\\cite{Rsah2} and $V_0\\sin h (\\alpha\\sqrt k_0\\Delta\\phi)^\\beta$\\cite{Rsah,Rlop} have been considered, which have asymptotic forms like the inverse power law or exponential ones. Different physical considerations have lead to the study of other types of the potentials also\\cite{Ruzan}. Recently Saini {\\it et al} \\cite{Rsai} have reconstructed the potential in context of general relativity and minimally coupled quintessence field from the expression of the luminosity distance $d_L(z)$ as function of redshift obtained from the observational data. However, none of these potentials are entirely free of problems. Hence, there is still a need to identify appropriate potentials to explain current observations \\cite{Rfer}. Also it has been recently shown by Pietro and Demaret\\cite{Rpietro} that for constant scalar field equation of state, which is a good approximation for a tracker field solutions, the field equations and the conservation equations strongly constrain the scalar field potential. Most of the widely used potential for quintessence, such as inverse power law one, exponential or the cosine form, are incompatible with these constraints. The CDM is in general considered to be a perfect fluid. However, in some scenarios, certain physical processes can make the CDM fluid effectively a dissipative one. In such a situation the fluid has an effective pressure that is negetive. Recently it has been proposed that the CDM must self interact in order to explain the detailed structure of the galactic halos \\cite{CDM}. This self interaction will create a viscous pressure whose magnitude will depend on the mean free path of the CDM particles. In a recent work Chimento et.al have shown that a mixture of minimally coupled self interacting scalar field and a perfect fluid is unable to drive the accelerated expansion as well as solve the cosmic coincidence problem at the same time \\cite{chimento}. However, a mixture of a dissipative CDM with bulk viscosity and a minimally coupled self interacting scalar field can successfully achieve both features simultaneously. Also, as demonstrated in a recent paper by Zimdahl \\etal ~\\cite{zim} one can also have a negative $\\pi$ if there exists an interaction which does not conserve particle numbers. This may be due to the particle production out of gravitational field. In this case, the CDM is not a conventional dissipative fluid, but a perfect fluid with varying particle number. Substantial particle production is an event that occurs in the early universe. But Zimdahl \\etal have shown that even extremely small particle production rate can also cause the sufficiently negative $\\pi$ to violate the strong energy condition. In this paper we have used a minimally coupled scalar field with a self interacting potential together with a matter fluid having a dissipative pressure over and above its positive equilibrium pressure. We have not assumed any particular model for this negative pressure. Instead, we have investigated what kind effects it has in the expansion of the universe. Unlike other works in scalar field cosmology with a dissipative pressure ~\\cite{dis} we have neither assumed the behaviour of the scale factor nor have we assumed any specific form of the potential. Rather we have expressed all the variables in terms of what we call the `{\\it generating function}'. For this we have followed the method described by Chimento et.al. \\cite{chimento1} with some additional assumptions. We have proceeded with a particular choice of the generating function for which the potential is constructed using a combination of different power-law functions of $\\phi$. From the behaviour of decelerating parameter it has been shown that one can indeed generate both inflationary era in the early time and also an late time accelerating phase with a decelerating period in between. We have also investigated the stability and attractor structure of the general solutions of the field equations with this kind of potential and have found that for certain choices of the constants the solutions indeed exhibit attractor behaviour in the late times. ", "conclusions": "The most important conclusion is that models with a self-interacting scalar field and a matter fluid having a negative pressure in addition to its positive equilibrium pressure can produce a scenario for the cosmological evolution in which one can have an inflationary phase to begin with, an accelerated phase at late times (like the present era) and a decelerating phase in-between. Recently Lopez and Matos~\\cite{matos} have shown that this kind of complete history for the scale factor can be described by a hyperbolic potential. But the physical origin of such potential is still not well known. But here we have shown that such kind of behaviour for the scale factor can be generated with a potential given in (22),which has been widely used by many authors for inflationary models. \\par The behaviour of $\\rho$ and $\\rho_{\\phi}$ in our model shows that although in early universe, $\\rho$ is greater than $\\rho_{\\phi}$ which is necessary for different physical phenomena like nucleosynthesis and structure formation etc, in the late times, $\\rho_{\\phi}$ starts dominating. This feature explains the missing energy density and also the ration of two energy densities becomes a constant in late time showing the ``{\\it tracking nature}''. One should note that both the assumptions (8) and (16) play crucial role in our model. Given a barotropic equation of state between $\\rho$ and $p$ one can not assume (8) and (16) at the same time if the dissipative pressure $\\pi$ is zero in our model as t hat will lead to an over-determined problem ( the number of unknowns will be less than the number of independent equations). Even if one assumes these two condition one can check that will lead to an negative equilibrium pressure which is not desirable. H ence the existence of dissipative pressure also plays an important role in our model. We have also studied the general equation of motion for the scalar field (equation (24)) for the potential (22) and have shown that for the choices of constant for which the potential is minimum at $\\phi=0$, the phase space diagram exhibit a attractor behaviour towards the asymptotic de-sitter solution. We want to mention that previously tracker and attractor solutions have been studied for scalar fields having inverse power law, exponential, cosine potential. But in all of these cases the equation of state $w$ is a constant in radiation era as well as in a matter dominated era. It was later shown by Pietro and Demaret ~\\cite{Rpietro} that these kind of potential with a constant $w$ is not consistent with the field equations. In our case, the equation of state for the scalar field $w$ is not a constant but it varies with the cosmic evolution and approaches towards -1 asymptotically showing the existence of a cosmological constant in late times." }, "0005/astro-ph0005375_arXiv.txt": { "abstract": "The recently formulated Two Oscillator (TO) model interprets the lowest of the kilohertz frequencies of the twin peak quasi-periodic oscillations in X-ray binaries as the Keplerian frequency $\\nu_{\\rm K}$. The high twin frequency $\\nu_h$ in this model holds the upper hybrid frequency relation to the rotational frequency of the neutron star's magnetosphere ${\\bf\\Omega}$: $\\nu_h^2=\\nu_{\\rm K}^2+4(\\Omega/2\\pi)^2$. The vector ${\\bf\\Omega}$ is assumed to have an angle $\\delta$ with the normal to the disk. The first oscillator in the TO model allows one to interpret the horizontal branch observed below 100 Hz as the lower mode of the Keplerian oscillator under the influence of the Coriolis force, with frequency $\\nu_L$ being dependent on $\\nu_h$, $\\nu_{\\rm K}$ and $\\delta$. For some stars such as 4U 0614+09, Sco X-1 and 4U 1702-42, $\\nu_h$, $\\nu_{\\rm K}$ and $\\nu_L$ have been observed simultaneously providing the opportunity to check the central prediction of the TO model: the constancy of $\\delta$ for a particular source. Given the considerable variation of each of these three frequencies, the existence of an observational invariant with a clear physical interpretation as a global parameter of the neutron star magnetosphere is an important test of the TO model. Using the results of recent observations of 4U 0614+09 we verify the existence of this invariant and determine the angle $\\delta=15^o.6\\pm 0.^o5$ for this star. The second oscillator in the model deals with a radial (presumably sound) oscillation and diffuse process in the viscous layer surrounding the neutron star. Our analysis of the viscous oscillation frequency $\\nu_V$ and the break frequency $\\nu_b$ of the diffusion shows that the spin value of the inner boundary of the transition layer for 4U 0614+09 is at least two times more than values for 4U 1728-34 and Sco X-1. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005361_arXiv.txt": { "abstract": "More than 1300 variables classified provisionally as first overtone RR~Lyrae pulsators in the {\\sc macho} variable star database of the Large Magellanic Cloud (LMC) have been subjected to standard frequency analysis. Based on the remnant power in the prewhitened spectra, we found 70\\% of the total population to be monoperiodic. The remaining 30\\% (411 stars) are classified as one of 9 types according to their frequency spectra. Several types of RR~Lyrae pulsational behavior are clearly identified here for the first time. Together with the earlier discovered double-mode (fundamental \\& first overtone) variables this study increased the number of the known double-mode stars in the LMC to 181. During the total 6.5~yr time span of the data, 10\\% of the stars show strong period changes. The size, and in general also the patterns of the period changes exclude simple evolutionary explanation. We also discovered two additional types of multifrequency pulsators with low occurrence rates of 2\\% for each. In the first type there remains one closely spaced component after prewhitening by the main pulsation frequency. In the second type the number of remnant components is two, they are also closely spaced, and, in addition, they are symmetric in their frequency spacing relative to the central component. This latter type of variables is associated with their relatives among the fundamental pulsators, known as Blazhko variables. Their high frequency ($\\approx 20\\%$) among the fundamental mode variables versus the low occurrence rate of their first overtone counterparts makes it more difficult to explain Blazhko phenomenon by any theory depending mainly on the role of aspect angle or magnetic field. None of the current theoretical models are able to explain the observed close frequency components without invoking nonradial pulsation components in these stars. ", "introduction": "In this continuing series of papers dealing with the variable star data of the {\\sc macho} project, we examine here the temporal behavior of a large sample of {\\it first overtone RR~Lyrae} stars. Until very recently, short-periodic RR~Lyrae stars have been known to appear in two different flavors: (a) singly-periodic; (b) doubly-periodic (or double-mode). In this latter case the two periods are associated with the first two radial normal modes of the star. Because of their dominant first overtone content, these variables were hidden for a long period of time among the monoperiodic first overtone stars. In 1977 AQ~Leo was discovered as the first double-mode RR~Lyrae star by Jerzykiewicz \\& Wenzel (1977). The first variable of this type in a globular cluster was discovered by Goranskij (1981). However, it was only in 1983 when the first systematic studies started in globular clusters (Cox, Hodson \\& Clancy 1983), although there were suggestions that some of the first overtone variables in the appropriate period range with 'excessive scatter' might be actually double-mode stars (Sandage, Katem \\& Sandage 1981). Less than a year ago Olech et al. (1999a, b, hereafter O99a,b) found that some first overtone stars in the globular clusters M5 and M55 exhibit two frequencies, very closely spaced. They argued that the high period ratio strongly indicated the presence of nonradial modes in those stars. In the course of another recent study of the pulsation behavior of the RR~Lyrae stars in the Galactic bulge sample of the {\\sc ogle} project, Moskalik (2000, hereafter M00) also found several variables with closely spaced frequencies. Furthermore, in a selectively chosen sample of the {\\sc macho} RR~Lyrae inventory, Kurtz et al. (2000) also picked a few amplitude- and phase-modulated (Blazhko-type) variables among the first overtone stars. In the present paper we carry out a {\\it systematic} study of the frequency spectra of 1350 first overtone stars of the {\\sc macho} project for the Large Magellanic Cloud. The results of the investigation of a shorter segment of the same data set have already been briefly summarized by Kov\\'acs et al. (2000). This is the first large-scale survey of the finer details of the temporal behavior of the variable star data of a microlensing survey. (We note however, that in a recent paper, Udalski et al. 1999, without going into the details, mention a similar massive analysis in a search for double-mode Cepheids in the Small Magellanic Cloud.) Due to the large size of the sample, this study also yields valuable statistics in respect of the occurrence rate of the various modal behaviors among first overtone stars. This, together with other information on the fundamental mode RR~Lyrae stars, supplies crucial observational data for understanding RR~Lyrae pulsation, and, in particular, the Blazhko phenomenon. Because of the various pulsation behaviors discovered among RR~Lyrae stars, we found it necessary to introduce new notation for the already known classical types. Throughout this paper we use RR0 for fundamental, RR1 for first overtone and RR01 for double-mode (fundamental \\& first overtone), instead of the traditional notation of RRab, RRc and RRd. We will see that the new notation can be more successfully adapted to label new types and subtypes of variables. ", "conclusions": "A massive frequency analysis has been carried out on a large sample of the first overtone RR~Lyrae (RR1) population of the {\\sc macho} variable star base of the Large Magellanic Cloud. The study was aimed primarily at finding multiperiodic and non-stationary variables among RR1 stars. Special effort was made to search for amplitude- and phase-modulated (Blazhko) variables. Occurrence of this phenomenon among RR1 stars was not known previously. Because widely acceptable theoretical interpretation of the Blazhko effect is still missing, it is important to accumulate more observational data in order to constrain further the possible models. Because of the high occurrence rate of the Blazhko effect among fundamental mode RR~Lyrae stars and the lack of theoretical understanding of this phenomenon, known for almost a century, further study of these stars has a general significance. \\begin{deluxetable}{llrrc} \\tablecaption{Variable types in the {\\sc macho} database for the first overtone RR~Lyrae stars in the LMC } \\tablenum{7} \\tablehead{ {\\it Type} & {\\it Short description} & {\\it Number} & \\% & $\\sigma$(\\%) } \\startdata RR1 & Single-mode & 916 & 69.0 & 1.3 \\\\ RR1$-\\nu1$ & 1 close component & 24 & 1.8 & 0.4 \\\\ RR1$-\\nu2$ & 2 asymmetric close components & 3 & 0.2 & 0.1 \\\\ RR1$-\\nu M$ & Multifrequency close components & 5 & 0.4 & 0.2 \\\\ RR1$-BL$ & 2 symmetric close components & 28 & 2.1 & 0.4 \\\\ RR1$-PC$ & Period change & 141 & 10.6 & 0.9 \\\\ RR01 & 1st \\& 0th overtone double-mode & 181 & 13.6 & 0.9 \\\\ RR12 & 2nd \\& 1st overtone double-mode & 3 & 0.2 & 0.1 \\\\ RR1$-D1$ & Integer d$^{-1}$ frequencies & 20 & 1.5 & 0.3 \\\\ RR1$-NC$ & Non-classified & 6 & 0.5 & 0.2 \\\\ \\enddata \\tablecomments{The symbol $\\sigma$(\\%) denotes the standard deviation of the population ratio, assuming Poisson distribution in the different populations.} \\end{deluxetable} The basic technique of analysis we applied was the standard Fourier frequency analysis with the important sequential steps of prewhitening. All 1350 variables passed through a basic analysis which resulted in the selection of more than 400 multimode candidates for further studies. In Table 7 we summarize the statistics of the various types of variables we identified in the sample. This table does not contain the more than 30 doubly identified variables which were among the 1350 stars in the basic analysis of the data set. However, to make the statistics more complete, we added the 73 double-mode (RR01) variables of Alcock et al. (1997) to the newly discovered 108 stars of this study. As we have already warned in the paper, the classification of some of the variables may not be definitive. Although this results in some ambiguity in the derived statistics, it does not change our basic conclusions. Furthermore, because of the finite noise level, there is a limit for the lowest detectable secondary signal components. Considering specifically the Blazhko-type stars, from the statistical tests presented in Sect. 5, we estimate this limit to be 10--15\\% in the units of the main signal component (10\\% being the level of marginal, whereas 15\\% is the level of close to secure detectability). The 108 newly discovered RR01 stars follow the pattern on the $P_0\\rightarrow P_1/P_0$ diagram of the already known 73 double-mode variables (Alcock et al. 1997). This pattern implies metallicity and/or mass spread among the RR01 population of the LMC. Direct metallicity measurements through the $\\Delta S$ method by Clementini et al. (2000) suggest the existence of sufficient [Fe/H] spread. The preliminary theoretical work of Popielski \\& Dziembowski (2000) comes to a similar conclusion. Accurate photometry and metallicity measurements of these stars would be very valuable also for constraining further the LMC distance scale (Kov\\'acs \\& Walker 1999). We found three variables suspected for double-mode pulsations in the first and second overtones. Further study of these stars with additional photometry would be necessary to confirm their suggested pulsational status. A considerable fraction of the RR1 population shows period change. Although some of them can be approximated by linear period changes, their high values exclude any evolution-related explanation. Understanding period changes in these and in other RR~Lyrae stars remains an important unsolved problem in stellar pulsation theory. We regard the most exciting finding of this study the discovery of the variables with closely spaced frequencies (for preliminary works on these and some fundamental mode Blazhko stars we refer to Kov\\'acs et al. 2000 and Kurtz et al. 2000). About half of these stars show only two components, similarly to the recent discoveries of the same type of stars by Olech et al. (1999a,b) and Moskalik (2000). Although in some of these RR1$-\\nu1$-type stars in our sample a weak third component with symmetric frequency spacing can be suspected, neither the large majority of our RR1$-\\nu1$ stars, nor the variables reported by other authors show such components. However, in 28 variables we observe symmetric frequency spacing, very similar to the frequency spectra of the fundamental mode (RR0) Blazhko-type variables (e.g., Kov\\'acs 1995). Therefore, we associate these variables with the Blazhko-type stars. At this moment it is only the LMC in which first overtone Blazhko stars have been identified. Compared with the commonly used incidence rate of 20--30\\% of the RR0 Blazhko variables (Szeidl 1988), our 2\\% frequency among RR1 stars seems to indicate that whatever is the underlying cause of the Blazhko phenomenon, it should work quite {\\it differently} in RR1 than in RR0 stars. Although in a milder sense, this difference still survives even if we consider the recent statistics obtained by Moskalik (2000) on the {\\sc ogle} sample of Galactic bulge RR~Lyrae stars. He found 11 Blazhko variables from the 149 RR0 stars but none from the 66 RR1 stars analyzed (however, he found RR1$-\\nu1$-type stars: 11 among the RR0, and 2 among the RR1 stars). According to our tests, if the lowest level of modulation among RR1 stars were 15\\%, assuming 20\\% incidence rate for the Blazhko phenomenon, we should have detected at least 30 more variables of this type. Although with a 10\\% modulation level under the same circumstance this number decreases to about 10, it is still large enough to conclude that we probably have not missed any Blazhko RR1 variables above the 10\\% modulation level. Adding to these that the high frequency of the Blazhko RR0 stars refers to relatively older data with marginal probability of the detection of low-amplitude modulations, and that most of them have 30--50\\% total modulation level, we think that the large difference in the occurrence rates of the two types of Blazhko stars is significant. It follows from this result that {\\it magnetic oblique rotator} model (Shibahashi 1994; 2000) of the Blazhko phenomenon must face with the difficulty in explaining why the role of magnetic field becomes so much less important as a star pulsating in the fundamental mode and showing Blazhko phenomenon moves to the RR1 regime. Most of these stars must either stop being Blazhko-type variables, or their modulation level must decrease from 30--50\\% down to below 10\\% (in the units of the total pulsation amplitude). In addition, there is also the problem of asymmetric modulation amplitudes. This difficulty becomes extreme for the above theory to explain both Blazhko- and RR1$-\\nu1$-type variables within the same framework. In order to deal with these basic observational facts, the magnetic model needs to be more complex than it is now. This, without the precise measurement of magnetic field, would lead to introducing more free parameters, and thereby to decreasing the predictive power of the theory. The other theory, the model of {\\it direct ($\\omega_0\\approx \\omega_l$) resonant mode coupling between a radial and nonradial modes} stays within the framework of nonlinear stellar pulsation, without invoking the presence of magnetic field. According to Van Hoolst, Dziembowski \\& Kawaler (1998), with significant probability this resonance is capable of destabilizing the fundamental limit cycle and leading either to steady {\\it phase locked} pulsation, or to an amplitude- and phase-modulated one. In both cases we would see Blazhko variability but in the steady amplitude and phase solution observability would depend on the aspect angle of the observer relative to the rotation axis. Concerning the resonance theory, it is important to remark that the recent linear pulsation survey of Dziembowski \\& Cassisi (1999) shows that the excitation rates for the resonant $l=1$ modes are comparable to those of the radial fundamental and first overtone modes. Although these results are encouraging, many questions remain to be answered, such as the relation between RR1$-\\nu1$- and Blazhko-type stars, and, of course, the low number of the Blazhko RR1 stars. We look forward to completing a similar survey of multimode RR~Lyrae stars in our six fields in the Small Magellanic Cloud. While we do not anticipate significant differences in the fractions of the types of multimode RR~Lyrae variables or sample metallicity, such a survey will provide a test of the degree to which such samples are homogeneous. In many cases the signal-to-noise of the SMC RR~Lyrae photometry is expected to be better than that of the LMC photometry because our SMC exposure times were twice as long and this more than compensates for the greater distance of the SMC." }, "0005/astro-ph0005157_arXiv.txt": { "abstract": "We have analyzed optical spectra of 473 X-ray and 235 optically selected AGNs, to study their emission line properties. We present results of an analysis of the \\hbeta\\ linewidths. We find that the linewidth distribution of quasars is shifted towards higher velocities ($<$v$>$=4300 \\kms) compared to the distribution of Sy\\,1s ($<$v$>$=3000 \\kms). There are no {\\em Narrow Line Quasars}, i.e. there are no AGNs with quasar luminosities and FWHM(\\hbeta)$<$2000 \\kms. NLSy1s comprise 20--30\\% of the AGN population at faint absolute magnitudes (\\MB $>-22$), irrespective of the selection method. In the RASS sample we find $\\Gamma$\\,[0.1-2.4 keV] $<$3.3. The $\\Gamma$ vs. FWHM(\\hbeta) distribution for Sy\\,1 galaxies is consistent with previous work. For QSOs the spectral index also flattens with increasing FWHM(\\hbeta), but they have larger linewidths than Seyfert 1s. ", "introduction": "\\vspace{-0.5cm} Seyfert\\,1 galaxies with small widths ($<$2000 \\kms) of the permitted emission lines coming from the broad line region (BLR) have been given a separate name {\\em Narrow Line Seyfert\\,1 galaxies (NLSy1)}, although they probably do not form a class physically separated from classical Sy\\,1s. Close attention was drawn to them resulting from the discovery that steep soft X-ray spectra of AGNs are almost always associated with NLSy1s \\cite{Boller96}. It was also claimed that X-ray selected samples contain considerably more NLSy1s than optically selected ones \\cite{Stephens89}. Since Sy\\,1 galaxies are not the most luminous examples of AGNs, we searched for similar effects among quasars, their high luminosity counterparts. We analyzed the Hamburg database of optical AGN spectra to determine the fraction of NLSy1s in optically and X-ray selected samples, to search for {\\em Narrow Line QSOs}, and to study the relation between soft X-ray spectral index and linewidth, as a function of luminosity. ", "conclusions": "" }, "0005/astro-ph0005227_arXiv.txt": { "abstract": "The core-dominated radio-loud quasar PKS 0637-752 ($z = 0.654$) was the first celestial object observed with the Chandra X-ray Observatory, offering the early surprise of the detection of a remarkable X-ray jet. Several observations with a variety of detector configurations contribute to a total exposure time with the Chandra Advanced CCD Imaging Spectrometer (ACIS; Garmire et al. 2000, in preparation) of about 100~ks. A spatial analysis of all the available X-ray data, making use of Chandra's spatial resolving power of about 0.4 arcsec, reveals a jet that extends about 10 arcsec to the west of the nucleus. At least four X-ray knots are resolved along the jet, which contains about 5\\% of the overall X-ray luminosity of the source. Previous observations of PKS~0637-752 in the radio band (Tingay et al. 1998) had identified a kpc-scale radio jet extending to the West of the quasar. The X-ray and radio jets are similar in shape, intensity distribution, and angular structure out to about 9 arcsec, after which the X-ray brightness decreases more rapidly and the radio jet turns abruptly to the north. The X-ray luminosity of the total source is $\\log L_X \\approx 45.8 \\ \\rm erg \\, s^{-1} \\ (2 - 10 \\ keV)$, \\footnote{We use $H_{0}$ = 50 km s$^{-1}$ Mpc$^{-1}$ and $q_{0}$ = 0 throughout} and appears not to have changed since it was observed with ASCA in November 1996. We present the results of fitting a variety of emission models to the observed spectral distribution, comment on the non-existence of emission lines recently reported in the ASCA observations of PKS~0637-752, and briefly discuss plausible X-ray emission mechanisms. ", "introduction": "Until recently most of our knowledge regarding the spatial structure and spectral shape of extragalactic jets has relied on observations performed in the radio and optical bands. X-ray detections and upper limits have provided insights into the physical conditions responsible for the observed radiation from knots and hotspots in extragalactic radio jets. However, in most cases the poor spectral and spatial resolution available has made the interpretation of the X-ray data difficult. The spectral energy distributions (SED) seen from jets have been ascribed to combinations of synchrotron radiation, synchrotron self-Compton (SSC) radiation, and thermal bremsstrahlung from shock-heated gas near the jets. In particular, the synchrotron and SSC models have been successfully employed to explain the X-ray emission from hotspots and jets in M87 (Biretta, Stern, $\\&$ Harris 1991), Cygnus A (Harris, Carilli, \\& Perley 1994) and 3C~295 (Harris et al., 2000). In the cases of 3C~273 (Harris \\& Stern 1987) and Pictor A (Meisenheimer et al. 1989), none of the standard processes yield satisfactory results. The Chandra X-ray Observatory (CXO), launched on 1999 July 23, provides a significant improvement over previous missions in combined spatial and spectral resolution (see Weisskopf and O'Dell 1997, van Speybroeck et al.~1997) which we expect will result in a significant increase in the number of detected and resolved X-ray jets. PKS~0637-752 is the first X-ray jet to have been discovered by {\\it Chandra}. The quasar was originally detected in X-rays with the Einstein Observatory (Elvis \\& Fabbiano 1984). Since then a peculiar emission-line feature at $\\sim 0.97$~keV was claimed in an ASCA SIS observation of the source (Yaqoob et al. 1998). One of two Ginga observations of PKS 0637-752 has been reported as showing a marginal detection of an Fe $\\rm K\\alpha$ line with an equivalent width of $103 \\pm 85 \\ \\rm eV$ ($\\pm 1\\sigma$ errors; Lawson \\& Turner 1997). In $\\S$ 2 we describe the spectral and spatial analysis of the core and jet, and our observations of the radio jet. The properties of sources in the vicinity of PKS~0637-752 are also briefly presented. Section 3 contains a brief description of our attempt to apply standard jet models to the observed SED of PKS~0637-752. We provide a thorough investigation of the underlying emission processes in a companion paper (Schwartz et al. 2000, in preparation) where more complex models are considered. ", "conclusions": "In figure 8 we present the SED of the WK7.8 knot of PKS~0637-752. The radio observations of PKS~0637-752 were performed at ATCA at 4.8 and 8.6~GHz. The 4.8 and 8.6~GHz beam width is $\\sim$ 2 and $\\sim$ 1 arcsec FWHM, respectively. The spectral indices and flux densities of the resolved components of the core, jet, and knots are presented in Table 7. Values for the optical flux density were obtained from the recent Hubble Space Telescope WFPC2 observations (Schwartz et al. 2000, in preparation and references therein). The SED shows that a single-component power-law synchrotron model cannot explain the combined radio, optical and X-ray flux densities, since the optical lies far below a power-law interpolation between the radio and X-ray measurements. We also tested whether SSC emission or inverse Compton scattering of cosmic microwave background (CMB) photons could explain the observed X-ray emission. The model components in Figure 8 are for the case of equipartition between the magnetic-field and electron energy densities, and assume a sphere of radius 0.15 arcsec and a power-law electron number spectrum of slope 2.4 between 100~MeV and 230~GeV, steepening by unity at 30~GeV due to energy losses. A detailed description of the SSC and inverse Compton calculations and the assumptions made for the model parameters are presented in a companion paper (Schwartz et al. 2000, in preparation). We estimate an equipartition field, B$_{eq}$, of about 2 $\\times$ 10$^{-4}$ Gauss. Based on this B$_{eq}$ value both SSC and IC on the CMB under-predict the X-ray flux by several orders of magnitude. Thermal models were also considered. Assuming a plasma temperature of 10~keV, an emission volume of 4 $\\times$ 10$^{-3}$ arcsec$^{3}$ and a 2-10~keV luminosity of 1 $\\times$ 10$^{42}$ erg s$^{-1}$, we estimate a plasma density of about 1 cm$^{-3}$. The derived rotation measures (RM) for radio waves propagating through such a dense plasma are quite large, inconsistent with the recent radio ATCA observations at 4.8 and 8.6~GHz (Schwartz et al. 2000, in preparation; Lovell et al. 2000). These observations show a RM $\\sim$ 80 rad m$^{-2}$ in the core, but no Faraday rotation in the jet, with an upper limit of $\\pm$ 10 rad m$^{-2}$. A contrived geometry where the jet collides with a giant molecular cloud in a companion galaxy producing thermal X-rays but, from our line-of-sight, the cloud is located behind the jet, may explain the non-detection of Faraday rotation in the jet. An examination of the radio and X-ray brightnesses of different parts of the main emission region in the jet suggests that the X-ray brightness to radio brightness ratio is remarkably constant out to the last knot (WK9.7), where the X-rays are relatively fainter. The interpretation of the change of X-ray brightness between WK8.9 and WK9.7 depends, however, on the emission process: if the X-radiation has a synchrotron origin, then the emitting electrons must be locally accelerated, and the change in X-ray brightness of WK9.7 would be telling us about changes in particle acceleration at different points in the jet. If the X-rays have an inverse Compton origin, then it is possible that the brightness change is entirely due to aging of an electron population accelerated in WK8.9, but then other difficulties in understanding the energetics of the source must be faced (see Schwartz 2000 for further discussion)." }, "0005/astro-ph0005011_arXiv.txt": { "abstract": "We describe a moderate-resolution (20--25 \\kms) FUSE study of the low-redshift intergalactic medium. We report on studies of 7 extragalactic sightlines and 12 \\lyb\\ absorbers that correspond to \\lya\\ lines detected by HST/GHRS and STIS. These absorbers appear to contain a significant fraction of the low-$z$ baryons and were a major discovery of the HST spectrographs. Using FUSE data, with 40 m\\AA\\ ($4\\sigma$) \\lyb\\ detection limits, we have employed the equivalent width ratio of \\lyb/\\lya, and occasionally higher Lyman lines, to determine the doppler parameter, $b$, and accurate column densities, N$_{\\rm HI}$, for moderately saturated lines. We detect \\lyb\\ absorption corresponding to all \\lya\\ lines with $W_{\\lambda} \\geq 200$ m\\AA. The \\lyb/\\lya\\ ratios yield a preliminary distribution function of doppler parameters, with mean $\\langle b \\rangle = 31.4 \\pm 7.4$ \\kms\\ and median 28 \\kms, comparable to values at redshifts $z = 2.0-2.5$. If thermal, these $b$-values correspond to $T_{\\rm HI} \\approx 50,000$~K, although the inferred doppler parameters are considerably less than the widths derived from \\lya\\ profile fitting, $\\langle b/b_{\\rm width} \\rangle = 0.52$. The typical increase in column density over that derived from profile fitting is $\\Delta$log~N$_{\\rm HI} = 0.3$ but ranges up to 1.0 dex. Our data suggest that the low-$z$ \\lya\\ absorbers contain sizable non-thermal motions or velocity components in the line profile, perhaps arising from cosmological expansion and infall. ", "introduction": "Since the discovery of the high-redshift \\lya\\ forest over 25 years ago, these abundant absorption features in the spectra of QSOs have been used as evolutionary probes of the intergalactic medium (IGM), galactic halos, large-scale structure, and chemical evolution. The {\\it Hubble Space Telescope} (HST/FOS) Key Project on QSO absorption systems found that \\lya\\ absorbers persist to low redshift in surprisingly large numbers (Bahcall et al. 1991; Morris et al.\\ 1991; Jannuzi et al. 1998; Weymann et al. 1998). In this paper, we assume that the \\lya\\ (and \\lyb) lines are intergalactic. Richards et al.\\ (1999) discuss the possibility that some C~IV absorption systems could be intrinsic to the AGN or ejected at relativistic velocities. The Colorado group has used HST to conduct a major survey of \\lya\\ absorbers at low redshift ($z \\leq 0.07$) along 15 AGN sightlines, using the moderate-resolution (19 \\kms) GHRS spectrograph (Stocke et al.\\ 1995; Shull, Stocke, \\& Penton 1996; Penton et al.\\ 2000a,b). Additional moderate-resolution HST/STIS data along 13 sightlines were taken during HST cycle 7. These observations measured \\lya\\ absorbers down to equivalent widths of 10--20 m\\AA\\ and determined distributions of the low-$z$ \\lya\\ absorbers in H~I column density for $12.3 < \\log N_{\\rm HI} < 14.0$ and in line width for $15 < b_{\\rm width} < 100$ \\kms\\ (Penton et al.\\ 2000b). The distribution function, $d {\\cal N}/dN_{\\rm HI} \\propto N_{\\rm HI}^{-1.80 \\pm 0.05}$, together with photoionization corrections suggest (Shull et al.\\ 1999a) that the low-$z$ \\lya\\ forest may contain a significant fraction of the baryons predicted by nucleosynthesis models of D/H (Burles \\& Tytler 1998). However, a precise baryon census in the low-$z$ IGM requires measurements of the true doppler parameter, $b$, to obtain accurate column densities in saturated \\lya\\ lines with log~N$_{\\rm HI} \\geq 13.5$. Measurements of \\lyb\\ or higher Lyman lines are needed to constrain the degree of saturation through a curve-of-growth (COG) analysis. Indeed, recent ORFEUS studies (Hurwitz et al.\\ 1998) of \\lyb/\\lya\\ ratios in two absorbers toward 3C~373 suggest that $b$ is less than the line width determined from \\lya\\ profile fitting. With the goal of characterizing the distribution of $b$-values and measuring accurate H~I columns, we conducted a FUSE mini-survey of \\lyb\\ absorbers toward 7 AGN with well-known \\lya\\ lines. The FUSE mission and its capabilities are described by Moos et al.\\ (2000) and Sahnow et al.\\ (2000). In \\S~2 we describe our FUSE \\lyb\\ observations. We also compare simple COG (\\lyb/\\lya) estimates of $b$ with single-component fits to (HST) \\lya\\ lines to understand the kinematic structure of the low-$z$ absorbers. In \\S~3 we present our conclusions and give directions for future work on the IGM baryon content, line kinematics, and temperature. ", "conclusions": "First and foremost, FUSE has identified the low-redshift \\lyb\\ forest. Lines between 1216~\\AA\\ and (1216~\\AA)$(1+z_{\\rm em}$) with no clear identification are often labeled as \\lya. The FUSE detection of their \\lyb\\ counterparts makes these identifications conclusive. We have used the COG concordance of Ly$\\beta$/Ly$\\alpha$ equivalent widths, and occasionally higher Lyman lines, to derive reliable values of $b$ and N$_{\\rm HI}$ for the stronger, saturated lines ($W_{\\lambda} > 200$ m\\AA). Our major findings are: (1) The doppler parameters from single-component COG fits to \\lyb/\\lya\\ equivalent widths are considerably less than those derived from line profile fitting, with $\\langle b/b_{\\rm width} \\rangle = 0.52$; (2) We find $\\langle b \\rangle = 31.4 \\pm 7.4$ \\kms\\ (median 28 \\kms), similar to values at redshifts $z =$ 2.0--2.5; (3) Although these $b$-values correspond to $T_{\\rm HI} \\approx 50,000$~K, the low-$z$ absorbers may contain non-thermal motions or line broadening from cosmological expansion and infall. Over its lifetime, FUSE will observe many AGN sightlines for the O~VI, D/H, and AGN projects. This will produce a large survey of Ly$\\beta$ absorbers at $z < 0.155$ that we can use to characterize the distributions in $b$ and N$_{\\rm HI}$ in the low-redshift \\lyb\\ forest. The distribution function, $f(b)$, can be used to derive the IGM temperature distribution and infer its equation of state (Schaye et al.\\ 1999; Ricotti, Gnedin, \\& Shull 2000). With high-S/N data and a flux-limited sample, we can search for the hot baryons predicted by cosmological simulations (Cen \\& Ostriker 1999a). These absorbers should appear in the high-$b$ tail of the distribution as broad, shallow absorbers. Because \\lya\\ lines with $W_{\\lambda} > 130$ m\\AA\\ (log N$_{\\rm HI} > 13.5$) appear to be saturated (Penton et al.\\ 2000b), HST alone cannot provide accurate column densities for the strong \\lya\\ absorbers that probably dominate the baryon content and opacity of the low-$z$ IGM. For lines in which $b$ is well determined, log~$N_{\\rm HI}$ typically increases by 0.3 dex, and up to 1 dex, compared to \\lya\\ profile fitting. This increase, which is greatest in the most saturated \\lya\\ lines, means that an even larger fraction of baryons may be found in the low-$z$ \\lya\\ forest. A full \\lyb\\ survey will also provide a sample of high-$N_{\\rm HI}$ absorbers which can be used to estimate the level of metallicity in the low-$z$ IGM (Shull et al.\\ 1998; Cen \\& Ostriker 1999b). As an illustration, we constructed simple photoionization models using CLOUDY (Ferland 1996) that assume a specific ionizing intensity at 13.6 eV of $J_\\nu = 10^{-23}$ ergs cm$^{-2}$ s$^{-1}$ sr$^{-1}$ Hz$^{-1}$ (Shull et al.\\ 1999b) with $J_\\nu \\propto \\nu^{-1.8}$ and column density $\\log N_{\\rm HI} = 15$. For total hydrogen densities (cm$^{-3}$) of $\\log n_H = (-3,-4,-5)$, a 20 m\\AA\\ measurement of C~III $\\lambda$977 implies $Z/Z_{\\odot} = (0.3, 0.15, 0.01)$. For $\\log n_H = (-5, -6)$, a 20 m\\AA\\ measurement of O~VI $\\lambda$1032 implies $Z/Z_{\\odot} = (0.015, 0.006)$. Observations of C~III $\\lambda977$ will be especially useful, because they can be compared with C~II and/or C~IV to obtain reliable ionization corrections." }, "0005/hep-ph0005238_arXiv.txt": { "abstract": "The Mirror Matter or Exact Parity Model sees every standard particle, including the physical neutral Higgs boson, paired with a parity partner. The unbroken parity symmetry forces the mass eigenstate Higgs bosons to be maximal mixtures of the ordinary and mirror Higgs bosons. Each of these mass eigenstates will therefore decay $50\\%$ of the time into invisible mirror particles, providing a clear and interesting signature for the Large Hadron Collider (LHC) which could thus establish the existence of the mirror world. However, for this effect to be observable the mass difference between the two eigenstates must be sufficiently large. In this paper, we study cosmological constraints from Big Bang Nucleosynthesis on the mass difference parameter. We find that the temperature of the radiation dominated (RD) phase of the universe should never have exceeded a few 10's of GeV if the mass difference is to be observable at the LHC. Chaotic inflation with very inefficient reheating provides an example of how such a cosmology could arise. We conclude that the LHC could thus discover the mirror world and simultaneously establish an upper bound on the temperature of the RD phase of the universe. ", "introduction": "The Mirror Matter or Exact Parity Model (EPM) sees every standard particle paired with a parity partner. This idea was first mentioned by Lee and Yang in their seminal paper on parity violation \\cite{ly} as a way to retain the full Poincar\\'e Group as a symmetry of nature despite the $V-A$ character of weak interactions. Some follow up work was performed in the ensuing decades on some aspects of the mirror matter hypothesis \\cite{kob}. In 1991, the idea was independently rediscovered and the full gauge theory constructed for the first time \\cite{flv1}. Shortly thereafter the EPM was extended to include nonzero neutrino masses and mixings and applied to the solar, atmospheric and LSND anomalies \\cite{flv2,flv3}. The EPM can also alter standard Big Bang cosmology in interesting ways, through the possible identification of some dark matter with mirror matter, and through modifications of Big Bang Nucleosynthesis (BBN) \\cite{fv}. The ordinary and mirror particle sectors can interact in a number of ways. The first is through gravitation, with immediate consequences for the dark matter problem and astrophysics. Non-gravitational interactions can be induced through the mixing of colourless and neutral particles with their mirror counterparts. Neutrinos, the photon, the $Z$ boson and the physical neutral Higgs boson can mix with the corresponding mirror states. Coloured and/or electrically charged particles are prevented from mixing with their mirror analogues by colour and electric charge conservation laws. The purpose of this paper is to study the Higgs boson sector of the EPM. It has been previously noted that the mass eigenstate physical Higgs bosons must be maximal mixtures of the underlying ordinary and mirror states because of the unbroken parity symmetry \\cite{flv1,flv2}. Each mass eigenstate will therefore decay $50\\%$ of the time into invisible mirror particles, providing a striking experimental signature in principle. The production cross-section for such Higgs bosons would be 1/2 of that of the standard Higgs boson of the same mass. This is a very simple and important observation, because it provides a clear way to experimentally establish the existence of the mirror world \\footnote{From the recent results of the L3 Collaboration \\cite{L3} we can establish a lower bound of about 65 GeV for a Higgs boson with these properties.}. In recent years there has been a strong focus on using the neutrino anomalies as a way to discover mirror matter \\cite{flv2,flv3,fv}. Neutrino oscillation physics certainly does provide a very interesting way to garner experimental evidence for mirror matter, or to at least constrain the model (if one is being pessimistic). However, the terrestrial neutrino phenomenology of the EPM is similar to that of pseudo-Dirac neutrinos \\cite{pd}, so complementary information would be useful. The Higgs boson sector is one potentially important way to obtain this information. (The mixing of ortho-positronium with mirror-ortho-positronium is another \\cite{fg}.) The strength of Higgs boson mixing with its mirror partner is controlled by an {\\it a priori} independent dimensionless parameter $\\lambda_{HH'}$. The mass splitting between the mass eigenstate Higgs bosons is proportional to this same parameter. Standard cosmology, through BBN, can be used to constrain $\\lambda_{HH'}$ and hence the Higgs boson mass splitting also \\cite{cf}. In this paper, we will demonstrate that the temperature of the radiation dominated (RD) phase of the universe should never have exceeded a few tens of GeV if the mass splitting is to be substantial (of order 1 GeV). Chaotic inflation with very inefficient reheating is {\\it an example} of how such a cold cosmology could arise. Remarkably, the Large Hadron Collider (LHC) could thus discover the mirror world as a byproduct of its Higgs boson search programme, and simultaneously establish an upper bound on the temperature of the RD phase of the universe. ", "conclusions": "The Exact Parity or Mirror Model predicts some simple and interesting Higgs physics. There will be two physical neutral Higgs boson mass eigenstates, each with a $50\\%$ invisible width. This would be a remarkable way to discover mirror particles. A detectable mass splitting between the two eigenstates would strongly suggest that the radiation dominated phase of the universe was never hotter than, say, a few tens of GeV. This would in turn be interesting information for cosmological model builders." }, "0005/astro-ph0005294_arXiv.txt": { "abstract": "We have constructed a family of simple models for spiral galaxy evolution to allow us to investigate observational trends in star formation history with galaxy parameters. The models are used to generate broad band colours from which ages and metallicities are derived in the same way as the data. We generate a grid of model galaxies and select only those which lie in regions of parameter space covered by the sample. The data are consistent with the proposition that the star formation history of a region within a galaxy depends primarily on the local surface density of the gas but that one or two additional ingredients are required to fully explain the observational data. The observed age gradients appear steeper than those produced by the density dependent star formation law, indicating that the star formation law or infall history must vary with galactocentric radius. Furthermore, the metallicity--magnitude and age--magnitude correlations are not reproduced by a local density dependence alone. These correlations require one or both of the following: (i) a combination of mass dependent infall {\\it and} metal enriched outflow, or (ii) a mass dependent galaxy formation epoch. Distinguishing these possibilities on the basis of current data is extremely difficult. ", "introduction": "\\label{c4intro} The chemical evolutionary histories of spiral galaxies provide considerable insight into many of the important processes involved in galaxy formation and evolution. For example, we can study star formation laws (SFLs; e.g.\\ Wyse \\& Silk 1989; Phillipps \\& Edmunds 1991), the interactions between newly-formed stars and the interstellar medium (e.g.\\ Dekel \\& Silk 1986; MacLow \\& Ferrera 1999), the importance and effects of gas flows (e.g.\\ Lacey \\& Fall 1985; Edmunds 1990; Edmunds \\& Greenhow 1995) and the infall that must accompany disc formation (e.g.\\ Tinsley \\& Larson 1978; Lacey \\& Fall 1983; Steinmetz \\& M\\\"{u}ller 1994). The main challenge is obtaining {\\it unambiguous} insight into particular physical processes. Some of the ambiguity can be circumvented by studying both the ages and the metallicities of galaxies: in this paper we use the ages and metallicities of a sample of face-on spiral galaxies to constrain which processes are the most important in affecting their observational properties. Despite these difficulties, considerable progress has been made in understanding some important aspects of galaxy formation and evolution. A local density dependence in the SFL is strongly favoured, although other factors may affect the star formation rate (SFR) over galactic scales (e.g.\\ Schmidt 1959; Dopita 1985; Kennicutt 1989; Wyse \\& Silk 1989; Dopita \\& Ryder 1994; Prantzos \\& Aubert 1995; Kennicutt 1998). Infall may be important in determining the metallicity distribution of stars in the solar neighbourhood (e.g.\\ Tinsley 1980; Prantzos \\& Aubert 1995; Pagel 1998). Other processes are more controversial: e.g.\\ metal-enriched outflows (e.g. MacLow \\& Ferrera 1999) or radial gas flows (e.g.\\ Edmunds \\& Greenhow 1995; Lacey \\& Fall 1985). However, recent observational advances, coupled with the development of multiple metallicity stellar population synthesis codes has allowed comparison of galaxy evolution models with both the gas metallicities and colours of spiral galaxies (Contardo, Steinmetz \\& Fritze-von Alvensleben 1998; Jimenez et al.\\ 1998; Boissier \\& Prantzos 2000; Prantzos \\& Boissier 2000; Cole et al.\\ 2000). The colours of spiral galaxies depend on both their ages and metallicities, therefore study of their colours offers fresh insight into galaxy formation and evolution, although inevitably degeneracies remain. In Bell \\& de Jong (2000; BdJ hereafter), we analysed the optical--near-infrared (near-IR) colours of a sample of 121 low-inclination spiral galaxies in conjunction with up-to-date stellar population synthesis models to explore trends in age and metallicity with galaxy parameters, such as magnitude or surface brightness. In particular, we found that there are significant trends between the age and $K$ band surface brightness of a galaxy, and between the metallicity and both the $K$ band magnitude and surface brightness of a galaxy. In that paper, we argued that these correlations could be the result of a surface density-dependent SFL, coupled with galaxy mass-dependent chemically-enriched gas outflows. In this paper, we investigate these ideas in more detail. We use a family of simple models to explore the effects of infall, outflows, age differences and SFLs on the colour-based ages and metallicities of spiral galaxies. Our aim is not to construct a self-consistent, realistic model of galaxy formation and evolution. This work is intended to guide future, more detailed explorations of the star formation histories (SFHs) of spiral galaxies: this simple modelling isolates which physical processes affect which observables, to allow more realistic models to concentrate on formulating self-consistent prescriptions for the most important physical phenomena. The plan of this paper is as follows. In section \\ref{c4model}, we outline the data and its main limitations. We describe the chemical evolution model, the basic assumptions and equations and outline how we translate the model output into observables which we can readily compare with the data. In section \\ref{c4evo}, we describe the properties of the closed box model. In section \\ref{c4inf} we explore the effects of infall, outflow and systematic trends in galaxy formation epoch. In section \\ref{c4sfl}, we investigate the effects of changing the SFL on our results. In section \\ref{c4disc} we discuss the results further, checking the plausibility of these models with other observational constraints. There, we also compare our models to a comparable, but more detailed model by Boissier \\& Prantzos. Finally, in section \\ref{c4conc}, we summarise our results. ", "conclusions": "\\label{c4conc} We have constructed a simple family of chemical evolution models with the aim of gaining some insight into the origins of many of the trends in SFH with galaxy parameters presented in BdJ. The model is used to generate colour-based ages and metallicities, which are directly comparable with those derived in BdJ. We generated a grid of model galaxies and selected only those which lie in a pre-defined region of the $K$ band absolute magnitude--central surface brightness plane. Using this model, we have found the following: \\begin{itemize} \\item A local gas surface density-dependent Schmidt SFL describes many of the colour-based age and metallicity trends from BdJ surprisingly well. A model of this type does not explain the mass dependence in SFH required by BdJ and significantly underpredicts the age gradient in spiral galaxies and the slope of the age--central surface brightness correlation. \\item The global properties of the fiducial model can be improved in either of two ways. \\begin{itemize} \\item[(i)] A combination of mass-dependent infall and metal enriched outflow imprint independent mass correlations on the galaxy colour-based ages and metallicities. Smaller galaxies have a more extended period of inflow in this model, and lose a greater fraction of their freshly-synthesized metals. \\item[(ii)] Galaxy formation epoch varies systematically with galaxy mass. If less massive galaxies are younger (i.e.\\ if they assembled the bulk of their gas content at very late times) we explain the mass--metallicity and mass--age correlation without resorting to outflow. \\end{itemize} \\item Regarding the radial variations within galaxies: \\begin{itemize} \\item[(i)] If the infall timescale varies with radius, an age gradient can be generated. This has little or no effect on the metallicity gradient. \\item[(ii)] Alternatively, Kennicutt's (1998) SFL (which involves both gas density and the dynamical time) produces both strong age and gas metallicity gradients and a reasonable scatter in the local age/metallicity--surface brightness correlation. The main shortcoming of this model is a `backwards' age/metallicity--magnitude correlation. \\end{itemize} \\end{itemize} One deliberate limitation of our empirical approach is that we do not incorporate our model galaxies into a detailed cosmological context. For example, cold dark matter cosmologies make well-defined predictions about the formation mechanisms and infall histories of galactic discs that we ignore (e.g.\\ Steinmetz \\& M\\\"{u}ller 1994; Mo, Mao \\& White 1998; Somerville \\& Primack 1999; Cole et al.\\ 2000). A more realistic treatment of the formation of our initial discs would be desirable, but is sensitive to the poorly understood details of angular momentum evolution in forming disc galaxies (e.g.\\ Navarro \\& Steinmetz 1997; 2000). In this paper our philosophy has been to determine which of our conclusions seem the most robust to model details, and which may change if our initial disc formation was made more realistic. Our study gives an indication of which of our conclusions (and those of Boissier \\& Prantzos) are robust: i.e.\\ which physical processes must operate in any galaxy formation scenario to reproduce the observations. From the above discussion, we find the following. \\begin{itemize} \\item Gas surface density determines the SFR, although other factors (such as dynamical time or a critical density for star formation) may also influence the SFR. \\item Radial dependence in the SFL and/or the infall history is favoured. \\item The infall of gas {\\it either} varies strongly with galaxy mass, peaking at late times in low-mass galaxies (as parameterised by our formation epoch model, E, or the model of Boissier \\& Prantzos), {\\it or} infall varies more weakly with galaxy mass but operates in conjunction with a higher efficiency of metal-enriched outflows in low-mass systems (Models I and O). The data (as it stands) marginally favours the latter option (Fig.\\ \\ref{fig:c4outeff}). \\end{itemize} However, the last conclusion is strongly dependent on the treatment of very low metallicity galaxies for which our colour-based ages and metallicities are highly uncertain. A combination of models I and O receives further support from studies of resolved stellar populations: there is ample evidence for older stellar populations in local faint, gas-poor and metal-poor dwarf Spheroidal galaxies. These cannot fit into the mass-dependent age scheme postulated above (e.g.\\ Grebel 1998; Hurley-Keller, Mateo \\& Nemec 1998). Of course, it is quite possible that very low-mass dwarf galaxies have a metallicity--magnitude correlation driven primarily by outflows, and more massive galaxies (such as spirals, which have managed to keep the bulk of their gas content) have a metallicity--mass correlation driven primarily by differences in galaxy formation epoch. The key observable is the effective yield of galaxies: the effective yield gives insight into whether a galaxy is simply under-evolved and metal-poor (like, perhaps, low surface brightness disc galaxies; de Blok et al.\\ 1996; Bell et al.\\ 2000) or has had the bulk of its metals removed in gas outflows (as appears likely for dwarf Spheroidals: e.g.\\ Dekel \\& Silk 1986). What is clear is that more work, both on observational and theoretical fronts, is required to fully elucidate the origin of the mass--metallicity correlation." }, "0005/astro-ph0005588_arXiv.txt": { "abstract": "At high gluon or string densities, gluons' saturation or the strong interaction among strings, either forming colour ropes or giving rise to string's percolation, induces a strong suppression in the particle multiplicities produced at high energy. This suppression implies important modifications on cosmic ray shower development. In particular, it is shown that it affects the depth of maximum, the elongation rate, and the behaviour of the number of muons at energies around $10^{17}$--$10^{18}$ eV. The existing cosmic ray data point out in the same direction. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005407_arXiv.txt": { "abstract": "A new investigation of the supernova remnant (SNR) \\objectname[]{N157B} was carried out with the Australia Telescope Compact Array. Radio continuum images of the entire \\objectname[]{30 Doradus} region have been made at 3.5 and 6\\,cm wavelength with a resolution of $\\sim$ 2\\arcsec. These data allow a high resolution study of the spectral index distribution and polarization properties of both N157B and the nearby \\objectname[]{30 Doradus} nebula (the latter will be reported in a subsequent paper). N157B is an extended Crab--type SNR which may be beginning the transition to a composite remnant. There is little apparent fine structure and the brightest radio region is several parsecs from the probable position of the X--ray pulsar. The SNR has a radio spectral index of $-$0.19 and is significantly polarized at 3.5\\,cm but not at longer wavelengths. ", "introduction": "\\objectname[]{N157B} \\citep{hen56} (also 30 Dor B or SNR 0538--691) is a supernova remnant located near the edge of the \\objectname[]{30 Doradus} nebula \\citep{bod01} in the \\objectname[]{Large Magellanic Cloud} (LMC). The \\objectname[]{30 Doradus} complex is the nearest extragalactic giant \\ion{H}{2} region and an active star--forming region \\citep{chu92}. The nebula and the SNR are in the same interstellar region of the LMC as \\objectname[]{SN1987A} and other SNRs, among which is \\objectname[]{SNR 0540--693}, a composite remnant with a pulsar and a plerion component \\citep{man93}. The SNR nature of \\objectname[]{N157B} was first determined by its excess of radio over optical emission \\citep{lem68}, and confirmed by the detection of [\\ion{S}{2}] lines in the optical spectrum \\citep{dan81}. The Crab nebula represents the class of pleronic supernova remnants; it contains a pulsar as a central energy source and a well--organised magnetic field indicated by strong linear polarization at radio wavelengths and relatively flat radio spectrum \\citep{wei83}. A number of detected SNRs have radio characteristics similar to the \\objectname[]{Crab nebula}, yet the pulsar stimulation of the synchrotron nebula is confirmed for just a few of them \\citep{sew83,sew89,wei88}. \\objectname[]{N157B} is classified as a Crab--type remnant because of its centre--filled morphology, flat radio spectrum \\citep{mil78,mil80}, and strong soft X--ray emission \\citep{lon79}. Recently a 16\\,ms pulsar detected only in X--rays \\citep{mar98,got96,wan98} has been associated with \\objectname[]{N157B}. Previous observations of \\objectname[]{N157B} at radio wavelength have not detected any polarized emission \\citep{dic94}. However, our current investigation at a shorter wavelength shows a significant level of polarization. All of the data together confirm the classification of \\objectname[]{N157B} as a Crab--type SNR. The presence of \\objectname[]{N157B} in a relatively dense and complex interstellar environment together with the younger SN1987A and the presumably older SNR 0540--693, offers a unique opportunity to study the evolution of Crab--type SNRs in such areas. The best Galactic examples, the Crab and 3C58, both appear to be in exceptionally low--density environments \\citep{fes97}. ", "conclusions": "\\objectname[]{N157B} is very close to \\objectname[]{30 Doradus} and it is also embedded in an \\ion{H}{2} region around the OB association \\objectname[]{LH99} \\citep{luc70}. \\citet{chu92} concluded that the southern part of the SNR is obscured and possibly interacts with the dark cloud within the \\ion{H}{2} region. With our high radio sensitivity we can see that the SNR does extend well to the south of the bright northern component. With a size of 30\\,pc$\\times$18\\,pc, it is the largest Crab--type SNR known. The feature at $05^h37^m52^s$ and $-69\\arcdeg10\\arcmin22\\arcsec$ and the elongated structure centered at about $05^h37^m50^s$ and $-69\\arcdeg11\\arcmin10\\arcsec $ might indicate the beginning of a transition into a composite remnant like SNR 0540--693. The dense interstellar environment of the \\objectname[]{30 Doradus} region may be conducive to the formation of a shell and the irregular structure of \\objectname[]{N157B} may be accounted for by a complex density distribution. \\citet{chu92} found that \\objectname[]{N157B} contributes up to 80\\% of the flux density in this region at 0.843 GHz , the rest coming from the surrounding \\ion{H}{2} region. They suggested that the flat radio spectral index ($\\alpha$) of $-$0.1 ($S_{\\nu} \\approx \\nu^{\\alpha}$) indicated from earlier measurements \\citep{mil80} should be characteristic of the SNR itself. However, they also found that the optical emission lines from \\objectname[]{N157B} originate in conditions associated with blast waves and they therefore concluded that the corresponding non--thermal radio emission should not have the flat spectrum typically associated with pulsar stimulation. Our results below differ somewhat from the previously published values derived from low resolution data but still show that N157B has a flat radio spectrum. Data used for determining the radio spectrum of \\objectname[]{N157B} are listed in Table\\, \\ref{tbl-spind}. To obtain a consistent background and source extent, the images were all convolved to the $43\\arcsec\\times45\\arcsec$ resolution of the MOST image and the integrated flux density calculated over the same area in each convolved map. A background intensity around the source was determined from the average of several areas around the SNR but far enough from 30 Doradus to avoid contamination. \\begin{deluxetable}{lll} \\footnotesize \\tablecaption{Flux Densities of N157B. \\label{tbl-spind}} \\tablewidth{0pt} \\tablehead{ \\colhead{Frequency (GHz)} & \\colhead{Flux Density (Jy)} & \\colhead{References}} \\startdata 0.843 & $2.86 \\pm 0.24$ & Anne Green (1997, private comunication) \\\\ 1.400 & $2.64 \\pm 0.16$ & Data of \\citet{meb97} \\\\ 2.378 & $2.36 \\pm 0.14$ & Data of \\citet{dic94} \\\\ 4.740 & $2.20 \\pm 0.10$ & Data of this paper \\\\ 8.600 & $1.82 \\pm 0.12$ & Data of this paper \\\\ \\enddata \\end{deluxetable} The resultant integrated flux densities listed in Table \\ref{tbl-spind} give a mean spectral index for \\objectname[]{N157B} of $-$0.19 (see Fig.\\ref{fig-spind}). Brightness--brightness plots also show that the spectrum does not vary significantly across the whole extent of the SNR. This argues that the entire SNR is pulsar--powered. \\begin{figure} \\plotone{figure4.ps} \\caption{Radio spectrum of N157B. The slope of the fitted line is $-$0.19.} \\label{fig-spind} \\end{figure} In the X--ray band, however, the spectrum of \\objectname[]{N157B} differs slightly relative to other such remnants. It has a power--law energy slope of $-$1.5 which is steeper than those of other Crab-like remnants with values of about $-$1.0 \\citep{wan98}. The X--ray and entire spectrum of \\objectname[]{N157B} also strikingly resemble that of SNR 0540--693, leaving no doubt of the pleronic origin of \\objectname[]{N157B}. The SNR 0540--693 contains a pulsar and is a composite remnant, sharing the characteristics of pure shell remnants and pure plerions \\citep{sew89,man93}. This latter composite SNR is often called a Crab--type remnant, because the core is Crab--like and the faint shell component was only investigated later. This further similarity between \\objectname[]{N157B} and 0540--693 is more evidence that \\objectname[]{N157B} is beginning the transition to a composite remnant. The polarization properties of the \\objectname[]{30 Doradus} region were previously investigated by \\citet{dic94} at longer wavelengths. They used ATCA observations at 13\\,cm but concluded that no polarized emission was detected in any region of \\objectname[]{N157B} above the level of instrumental polarization (less than 2\\%). We report a significant level of polarized emission at 3.5\\,cm from \\objectname[]{N157B} for the first time. This relatively high level of 10\\% is typical for a Crab--type remnant. The data at 6\\,cm give a marginal detection which we consider an upper limit of 3\\% fractional polarization at the peak of \\objectname[]{N157B}. This low level of polarized emission at this wavelength implies very strong internal depolarization. The observed polarization could also be reduced by the thermal emission from the associated \\ion{H}{2} region around \\objectname[]{LH99} which may be intertwined with the SNR so that its thermal electrons may cause some of the Faraday depolarization as well. To estimate the internal rotation necessary to cause the observed depolarization, we use a slightly modified version of the relation given by \\citet{bur66}: \\begin{equation} P(\\lambda_2)/P(\\lambda_1) = sin(\\delta)/\\delta \\end{equation} \\noindent where $P$ is the fractional polarization at $\\lambda_{2}$, the longer wavelength, and $\\lambda_{1}$, the shorter wavelength; $\\delta$ is the rotation in position angle of the measured electric vector between the two wavelengths. We will adopt the possible detection of 3\\% polarization at 6\\,cm and so use a rough mean value for $P(\\lambda_{2})/P(\\lambda_{1})$ of 0.3 which requires an average rotation of 135$\\arcdeg$. Then using the standard formula for Faraday rotation we can estimate the line--of--sight magnetic field strength: \\begin{equation} \\delta = 7.9 \\times 10^5 \\lambda^2 \\int N_e \\vec{B} \\dot{\\times} \\vec{dl} \\end{equation} \\noindent where $\\delta$ is in radians, $\\lambda$ is in meters, $N_{e}$ is the electron density in cm$^{-3}$, $B$ is the magnetic field strength in Gauss, and $l$ is the path length through the SNR in pc. Because the Faraday rotation increases progressively along the path, we must actually find the adopted rotation along half the depth along the line of sight. We do not, of course, know this depth directly but choose its value to be approximately equal to the short observed axis of the SNR or a half depth of 10\\,pc. From an analysis of ROSAT and ASCA X--ray spectra, \\citet{wan98} derived an electron density of $0.6f^{-0.5}$\\,cm$^{-3}$, where $f$ is the filling factor of the object. Because Crab--type SNRs should be reasonably uniform inside, we adopt a value for $f$ of about 0.5 to get a mean density of about 2\\,cm$^{-3}$. These numbers give a line--of--sight magnetic field strength of 30\\,$\\mu$Gauss. Most Crab--type SNRs have rather uniform magnetic fields; over the limited region where it can be observed, \\objectname[]{N157B} shows more variation than most. It may be that the field is directed mainly toward us, which would reduce the observed polarization at all wavelengths and also give a large Faraday rotation. No correction has been made for an unknown field orientation which would tend to increase the field strength in the remnant. We should emphasize that this result is only an order of magnitude estimate. Neglected factors include the uncertainties and probably variations in the depolarization, magnetic field reversals, density uncertainties and variations. For example, significant clumping can change the apparent integrated electron density because the X--ray emission depends upon the density squared but the Faraday rotation depends linearly on the density and the sum of squares does not equal the square of sums. If we know the magnetic field strength, we can then determine the energy in relativistic electrons from the intensity of the synchrotron radiation. \\citet{gin65} give the approximate formula: \\begin{equation} E_e = \\Sigma_\\nu \\times R^2 \\times A \\times area \\times pathlength \\times B^{-3/2} \\end{equation} \\noindent where $E_{e}$ is the total energy in relativistic electrons, $\\Sigma_{\\nu}$ is the surface brightness of the remnant at a given observing frequncy $\\nu$, $R$ is its distance, and $A$ is a parameter that depends on the frequency, the spectral index and the adopted cutoff frequencies for the synchrotron radiation. There appears to be a break in the spectrum of the non--thermal emission between the radio and X--ray \\citep{wan98} but we do not know specifically where it occurs. We shall choose an upper frequency of $10^{12}$\\,Hz and a lower one of $10^{7}$\\,Hz, with a spectral index of $-$0.19, $A = 3.7 \\times 10^{19}$ at 8.6\\,GHz where the flux density is 1.82\\,Jy. With the sizes and magnetic field strength given above, we find that the mean relativistic electron energy density is $4\\times10^{-10}$\\,ergs\\,cm$^{-3}$. This is about 10 times the magnetic energy density. While within the errors, we cannot definitely state that the values are different. It is interesting to note that in the three objects for which we have now measured the relative energies by this method, each has between 3 and 10 times greater relativistic electron energy than magnetic energy (Kepler's SNR \\citet{mat84} and N23 SNR \\citet{dic98}). \\citet{wan98} give an approximate temperature for the thermal gas in \\objectname[]{N157B} of 0.4 -- 0.7\\,keV. Choosing 0.6\\,keV and the approximate density of 2\\,cm$^{-3}$ used above, the thermal energy density is about $3\\times10^{-9}$\\,erg\\,cm$^{-3}$. Thus the thermal energy is about 10 times that in relativistic electrons and 100 times that in the magnetic field. Even in this Crab--type SNR the thermal energy dominates. In summary, the filled structure and uniform spectral index of \\objectname[]{N157B} are Crab--like but it also has some characteristics similar to those of the composite SNR 0540--693. Although the dense, irregular surroundings may be partly responsible for the non--uniform brightness and outline of this SNR particularly to the south, the overall morphology and spectral index suggest that this SNR is just beginning the transition from a pure Crab--like remnant to a composite remnant. It will be interesting to see how it changes at radio, optical and X--ray wavelengths in the future. We have detected for the first time a significant level of linearly polarized emission in this object at 3.5\\,cm wavelength. The encompassing H II region and clumpy internal structure within the SNR may cause sufficient Faraday rotation to reduce the polarization to below detectable levels at 6\\,cm and make the large variations in the polarization observed at 3.5\\,cm at different positions around the remnant. Observations at higher frequencies, e.g. 23\\,GHz, are needed to enable more accurate determination of the Faraday rotation." }, "0005/astro-ph0005176_arXiv.txt": { "abstract": "I briefly review the X/UV absorber models and show that the observations of NLS1s are generally consistent with the models. The covering factor of absorbers in NLS1s is likely to be high and there is some evidence of super-solar metallicities. I argue that NLS1s may be active galaxies in the early stage of their evolution and as such, may be low luminosity, low redshift analogues of the high redshift quasars. NLS1s may reside in rejuvenated gas-rich galaxies. I also propose that the high Fe~II emission in NLS1s may be a direct consequence of their large accretion rate and so a collisional ionization origin of Fe~II is favored. ", "introduction": "The organizers of this workshop have asked me to talk about the X-ray/UV absorbers in AGN and discuss what they can tell us about the Narrow Line Seyfert 1 phenomenon. However, I am going to go well beyond my expertise and speculate. I think that this meeting is an ideal platform for discussing new ideas, so let me start with my conclusions first and get them out of the", "conclusions": "That the AGN phenomenon was so much stronger at z$\\sim$2--3 than today has long elicited the suspicion that there is a connection between the youth of a galaxy and the likelihood that an AGN forms inside it. The question then naturally arises, ``what are the local counterparts to the young galaxies in the early universe in which local AGN may live?'' (Krolik 1999). A standard answer to this question is ``Starburst galaxies''. Heckman (1999) has argued that starburst galaxies are the low redshift analogues of Lyman break galaxies at high redshift. Similarly, we ask, what are the low redshift analogues of high redshift (z\\gax 4) quasars? I propose that they might be NLS1s. It is my pleasure to thank Th. Boller and the organizing committee for inviting me to this wonderfully stimulating workshop. I thank the Wilhelm and Else Heraeus foundation for travel support and the delightful stay at the Physikzentrum, Bad Honnef. This work is supported in part through NASA grant NAG 5-3249 (LTSA). The figures were created using the archives at the Space Telescope Science Institute, operated by the Association of Universities for Research in Astronomy, Inc., from NASA contract NAS5-26555." }, "0005/astro-ph0005340_arXiv.txt": { "abstract": "We use hydrodynamic cosmological simulations to predict the star formation properties of high-redshift galaxies ($z=2-6$) in five variants of the inflationary cold dark matter scenario, paying particular attention to $z=3$, the redshift of the largest ``Lyman-break galaxy'' (LBG) samples. Because we link the star formation timescale to the local gas density, the rate at which a galaxy forms stars is governed mainly by the rate at which it accretes cooled gas from the surrounding medium. At $z=3$, star formation in most of the simulated galaxies is steady on $\\sim 200$ Myr timescales, and the instantaneous star formation rate (SFR) is correlated with total stellar mass. However, there is enough scatter in this correlation that a sample selected above a given SFR threshold may contain galaxies with a fairly wide range of masses. The redshift history and global density of star formation in the simulations depend mainly on the amplitude of mass fluctuations in the underlying cosmological model. The three models whose mass fluctuation amplitudes agree with recent analyses of the \\lya forest also reproduce the observed luminosity function of LBGs reasonably well, though the dynamic range of the comparison is small and the theoretical and observational uncertainties are large. The models with higher and lower amplitudes appear to predict too much and too little star formation, respectively, though they are not clearly ruled out. The intermediate amplitude models predict SFR$\\;\\sim 30-40\\msunyr$ for galaxies with a surface density $\\sim 1\\;{\\rm arcmin}^{-2}$ per unit redshift at $z=3$. They predict much higher surface densities at lower SFR, and significant numbers of galaxies with SFR$\\;>10\\msunyr$ at $z \\geq 5$. ", "introduction": "\\label{sec:intro} The discovery and characterization of ``Lyman-break'' galaxies (LBGs) has opened a new window on the high-redshift universe, revealing a population of star-forming galaxies at $z>3$ whose comoving space density exceeds that of $L_*$ galaxies today \\citep{steidel96,lowenthal97}. These galaxies can be identified by their unusual colors in deep imaging surveys because the intrinsic continuum break at $\\lambda \\sim 912$\\AA\\ and the intergalactic absorption by the \\lya forest at $\\lambda < 1216$\\AA\\ redshift into optical bands. Spectroscopic follow-up shows that photometry of Lyman-break objects yields robust approximate redshifts. {}From an optical imaging survey, one can therefore construct a sample of high-$z$ galaxies limited primarily by rest-frame ultraviolet (UV) luminosity, which, in the absence of dust extinction, is itself determined mainly by the instantaneous formation rate of massive stars. Application of this approach to the Hubble Deep Field \\citep[HDF; ][]{williams96} and other deep imaging surveys has yielded first attempts at one of the long-standing goals of observational cosmology, determination of the star formation history of the universe \\citep[e.g.,][]{madau96,madau97,connolly97,steidel99}. In this paper, we examine the ability of models based on inflation and cold dark matter (CDM) to account for the observed population of LBGs, using cosmological simulations that incorporate gravity, gas dynamics, and star formation. We consider five variants of the CDM scenario: three $\\Om=1$ models, a spatially flat low density model with a cosmological constant, and an open universe low density model with $\\Ol=0$. The spatial clustering of the high-redshift galaxies in these simulations was discussed by \\citet{katz99}; here we focus on the masses and star formation properties of these galaxies. Numerical simulations play two overlapping but distinct roles in cosmological studies. First, they provide quantitative predictions that can be compared to observations in order to test the underlying cosmological models. Second, they provide greater understanding of the observational phenomena themselves, by showing how observable structures might arise and evolve in a given cosmological scenario. In this paper we will emphasize the second of these roles, mainly because the numerical limitations of the simulations and our limited knowledge of the physics of star formation contribute uncertainties that are comparable to the differences between cosmological models. The examination of different cosmologies is still a useful exercise, however, because it shows how cosmological parameters and properties of primordial mass fluctuations affect the properties of the high-redshift galaxy population when other physical and numerical parameters are held fixed. Hydrodynamic simulations complement the main alternative approach to the theoretical study of high-redshift galaxies, based on semi-analytic models of galaxy formation \\citep[e.g.,][]{baugh98,kauffmann99,somerville00}. Semi-analytic models have the advantages of simplicity, flexibility, and speed. The price is a substantial number of approximations and tunable parameters; the values of some parameters are fixed by matching selected observations, leaving other observables as predictions of the model. Semi-analytic models incorporate simplified descriptions of gravitational collapse, mergers, and cooling of gas within dark halos. The strength of numerical simulations is their more realistic treatment of these processes. The only free parameters (apart from the physical parameters of the cosmological model being studied) are those related to the treatment of star formation and feedback. Given these parameters, simulations provide straightforward, untunable predictions. However, the simulation approach must contend with the numerical uncertainties caused by finite volume and finite resolution, and computational expense makes it a slow way to explore parameter space. Over the next few years, interactions between the numerical and semi-analytic approaches should strengthen both. Here we mainly present the numerical results on their own terms, with a brief comparison to interpretations based on semi-analytic models in \\S\\ref{sec:disc}. We describe our numerical methods, treatment of star formation, and choice of cosmological model in \\S\\ref{sec:sims}. In \\S\\ref{sec:gals} we present results for the LCDM model (CDM with a cosmological constant) at $z=3$, the redshift best probed by recent Lyman-break galaxy surveys. In \\S\\ref{sec:csf} we broaden our scope, examining predictions of five different CDM models for the population of star-forming galaxies from $z=6$ to $z=2$. We discuss our results and prospects for future progress in \\S\\ref{sec:disc}. ", "conclusions": "\\label{sec:disc} Our main result is that inflationary CDM models, combined with straightforward assumptions about galactic scale star formation, predict a substantial population of star-forming galaxies at $z \\geq 2$. The stellar masses and star formation rates of these high-redshift systems are sensitive to the amplitude of the underlying mass power spectrum (compare, e.g., Figure~\\ref{fig:pk} and Figure~\\ref{fig:madau}.) The results of the LCDM, OCDM, and SCDM simulations appear roughly consistent with the observed properties of Lyman-break galaxies, given the theoretical and observational uncertainties. The low-amplitude, TCDM model predicts an anemic LBG population that is probably inconsistent with current observations, though this conclusion may be sensitive to our finite numerical resolution and our adopted value of $\\Omb$. The high-amplitude, CCDM model appears to predict too much high-redshift star formation, by a significant factor. The \\lya forest offers a more direct probe of the amplitude of mass fluctuations at high redshift \\citep{croft98}. Recent observational analyses (\\citealt{croft99,mcdonald00}; Croft et al., in preparation) imply that the matter power spectrum at $z \\approx 2-3$ is similar to that in our LCDM, OCDM, and SCDM models but incompatible with the CCDM or TCDM models. It is reassuring that the models supported by the \\lya forest data appear to be the most compatible with the star formation properties of observed LBGs, though an increase in $\\Omb$ to the values supported by recent D/H studies (\\citealt{burles97}, \\citeyear{burles98}) and other \\lya forest analyses \\citep{rauch97,weinberg97b} might spoil this agreement to some extent. We will examine the influence of $\\Omega_b$ on the high-redshift galaxy population elsewhere (Gardner et al., in preparation); our initial results imply that galaxy star formation rates in the SCDM model scale roughly as $\\Omega_b^{1.5}$. In KHW, we showed that the clustering of high-redshift galaxies in these simulations is consistent with observed LBG clustering \\citep{adelberger98,giavalisco98}, and that the clustering is insensitive to the cosmological model because galaxies form at the same ``biased'' locations of the dark matter distribution in all five simulations. In \\citet{gardner00}, we examine the predictions of these simulations for damped \\lya absorption, which is the other main observational probe of the high-redshift galaxy population. The galaxies resolved in these simulations account for only a fraction of the observed damped \\lya absorption at $z \\approx 3$, ranging from $\\sim 3\\%$ in TCDM to $\\sim 30\\%$ in LCDM, SCDM, and CCDM to $\\sim 50\\%$ in OCDM. Since the simulations already go to higher space densities than existing spectroscopic samples of LBGs, our results imply that these LBG samples are not yet deep enough to include the galaxies responsible for most damped \\lya absorption. \\citet{haehnelt00} reach a similar conclusion using analytic arguments. By extrapolating the simulation results with the aid of the Press-Schechter (\\citeyear{press74}) mass function, \\cite{gardner00} conclude that absorption in lower mass systems is sufficient to account for observed damped \\lya absorption in any of these cosmological models, with the possible exception of TCDM. However, the uncertainties in the extrapolation are large, and definitive examination of the compatibility between LBG and damped \\lya constraints will require higher resolution simulations. Semi-analytic methods, sometimes combined with N-body simulations of the dark matter distribution, are the main alternative to hydrodynamic simulations for theoretical investigation of high-redshift galaxy formation. Using these methods, several groups have found that CDM models like the ones studied here can reproduce the numbers, luminosities, colors, and clustering properties of observed LBGs \\citep[e.g., ][]{baugh98,governato98,kauffmann99,somerville00}. The semi-analytic papers have led to three rather different suggestions about the nature of Lyman-break galaxies. In the first view, observed LBGs are the most massive galactic systems present at high redshift, forming stars at a fairly steady rate \\citep{baugh98}. In the second view, interactions play a crucial role in triggering bursts of star formation, and many LBGs are low mass systems boosted temporarily, and briefly, to prominence \\citep{kolatt99,somerville00}. A third, intermediate case is one in which LBGs are massive galaxies experiencing bursts of star formation stimulated by minor or major mergers \\citep{somerville00}. This variety of views is mirrored to some extent in the observational literature on LBGs (compare, for example, \\citealt{steidel96} to \\citealt{lowenthal97} or \\citealt{trager97}). Our simulations suggest a picture that is intermediate between the extremes of this debate, but closest to the first point of view. Star formation in the simulated galaxies is steady on timescales of 200 Myr (Figure~\\ref{fig:sfrComp}), and the instantaneous star formation rate is fairly well correlated with stellar mass (Figure~\\ref{fig:sfrVsMstar}). However, there is enough scatter in galaxy star formation rates that a sample of galaxies selected above a SFR threshold includes objects with a substantial range of stellar masses (Figure~\\ref{fig:sfrVsMstar}), and these may reside in halos with a wide range of circular velocities (Figure~\\ref{fig:vc40}). The simulations automatically include interactions and mergers, but they do not resolve the existence of the low mass systems envisioned to play an important role in the extreme version of the collision-induced starburst model. The properties of the simulated LBG population depend on the cosmological initial conditions and on the basic physics of gravity and gas dynamics, but they also depend on our adopted prescription for galactic scale star formation. The crucial features of this prescription are (1) that the local star formation timescale decreases with increasing gas density, as implied by studies of local galaxies \\citep{kennicutt98}, and (2) that supernova feedback energy is deposited mainly in the dense interstellar medium, where it is usually radiated away before it has a large dynamical impact. Since we do not require any external triggers for star formation, an isolated galaxy that steadily accretes cold gas will convert that gas into stars, albeit over many orbital times. Interactions and mergers can enhance star formation by driving gas to higher density, but galaxies do not build up large reservoirs of dense gas that wait to be ignited. Limited resolution may reduce the influence of interactions and mergers in these simulations, since they do not resolve low mass satellites and do not resolve the nuclear star formation that is prominent in high resolution simulations of starbursts induced by minor (\\citealt{mihos94a}; \\citeyear{mihos96}) or major \\citep{mihos94b,hernquist95} mergers. A more efficient feedback mechanism could also lead to more episodic star formation histories, by producing cycles of starbursts followed by suppressed accretion and cooling. Since we have not investigated scenarios in which interactions or feedback play a larger role, we cannot draw conclusions about their viability. However, our results suggest that the straightforward treatment of star formation described in \\S\\ref{sec:numerics} is sufficient to explain at least the basic properties of the observed LBG population within the CDM cosmological framework. The clearest prediction of the simulations is that the Lyman-break galaxies studied by Steidel et al.\\ (\\citeyear{steidel96}, \\citeyear{steidel99}) and \\citet{lowenthal97} represent the tip of an iceberg. The cumulative distribution curves in Figure~\\ref{fig:cumSfr} should be taken as lower limits to the predicted galaxy number densities, especially at high redshifts, since limited resolution causes these simulations to underestimate the star formation rates in low mass systems, or to miss the systems entirely. Nonetheless, the curves for, e.g., the LCDM model show that there should be large numbers of $z=3$ galaxies below the magnitude limits of current LBG samples, and significant numbers of galaxies with SFR$\\;\\geq 10 \\msunyr$ even at $z \\geq 5$. Recent searches have already yielded a number of spectroscopically confirmed galaxies at $z=5-7$ \\citep{spinrad98,weymann98,chen99,hu99}, and analyses of deep HST/NICMOS images show candidate objects to redshifts $z \\geq 10$ \\citep{yahata00,dickinson00}. Systematic characterization of this faint galaxy population will be challenging, so it will be some time before simulations and data can be compared quantitatively in the very high redshift regime. But the existence of a significant population of star-forming galaxies at $z > 5$ is a natural prediction of the CDM scenario. Figures~\\ref{fig:sfrVsMstar} and~\\ref{fig:starFrac} also imply some correlations between observable properties of $z=3$ galaxies. While less massive galaxies tend to have lower star formation rates, they usually have higher ratios of instantaneous SFR to time-averaged SFR, and they should therefore have bluer spectral energy distributions unless they are more heavily reddened by dust. Less massive galaxies also tend to be more gas rich. Unfortunately, both of these predicted trends could be exaggerated by numerical resolution effects, so we do not regard them as very robust. This paper presents a first attempt to characterize the star formation properties of high-redshift galaxies using hydrodynamic simulations, but there is much room for progress with future simulations. The emerging consensus on cosmological parameters, if it survives the tightening of observational constraints, makes the task easier by focusing effort on a preferred background model. Within such a framework, simulations with different box sizes and resolutions can be combined to model the galaxy population over a wider dynamic range of mass and redshift, improving the comparison to the observed luminosity function (as in Figure~\\ref{fig:diffSfr}) and global star formation history (as in Figure~\\ref{fig:madau}). \\citet{weinberg99} take a first step along this path, using multiple simulations of the LCDM model (with higher $\\Omega_b$) to predict cumulative SFR distributions from $z=0.5$ to $z=10$. Since the present simulations resolve galaxies far below the limits of current LBG spectroscopic samples, simulations of larger volumes at lower resolution will improve the comparison between predicted and observed LBG clustering. Higher resolution simulations, on the other hand, can probe the connection between LBGs and damped \\lya systems and test the robustness of some of the trends found in \\S\\ref{sec:gals}. They can also provide predictions of the structural properties of high-redshift galaxies, such as size and morphology, though these may be best investigated with simulations that zoom in to follow the formation of individual objects \\citep[e.g., ][]{haehnelt98,contardo98}. In the long run, we also hope to examine different formulations of star formation and feedback, to determine what descriptions of these physical processes match the observed properties of galaxies and the intergalactic medium over the full range of accessible redshifts. Where are the Lyman-break galaxies today? Because our present simulations stop at $z=2$, we will save a detailed examination of this question for a future paper on the assembly history of galaxies, using simulations (like those of \\citealt{dave99}) that continue to $z=0$. A first look at these simulations suggests that there is no simple, one-sentence answer. Between $z=3$ and $z=0$, galaxies accrete fresh material and merge with each other, and many new galaxies form that had no $z=3$ progenitors at all (at least above the numerical resolution limits). The particles that lie in galaxies at $z=3$ are distributed at $z=0$ among galaxies with a wide range of environments and masses, though the most massive $z=0$ galaxies always contain some of these particles and the least massive often do not. Any link between LBGs and present-day ellipticals, or bulges, or halos, or clusters, can at best capture a general trend, one that is likely to be violated nearly as often as it is obeyed. Fortunately, cosmological simulations are an ideal tool for characterizing the full range of possible histories, providing the theoretical thread that can tie snapshots of the galaxy population at different redshifts into a coherent picture of galaxy formation and evolution." }, "0005/astro-ph0005389_arXiv.txt": { "abstract": "\\noindent We have extensively searched for periodic signals from the soft-gamma repeater SGR~1900+14, at 430 and 1410 MHz with the Arecibo telescope. Our observations did not reveal the 5.16-s periodicity discovered at X-ray wavelengths by Hurley et al.~(1998). We place pulsed flux-density upper limits of 150 and 30 $\\mu$Jy at 430 and 1410 MHz respectively. In the course of the 1410-MHz search we discovered a 226-ms radio pulsar, PSR~J1907+0918. Its period derivative implies that the age of J1907+0918 is only 38 kyr, making it one of the youngest members of the known pulsar population. Independent lines of evidence in support of this apparent youth are the unusually high degree of circular polarization and a relatively flat radio spectrum. The close proximity of this young radio pulsar to the supernova remnant G42.8+0.6 poses a problem for the proposed association between the G42.8+0.6 and SGR~1900+14. ", "introduction": "\\label{sec:intro} Since the discovery of the first soft-gamma repeater (SGR), the source of the famous ``March 5'' burst from the supernova remnant N49 in the Large Magellanic Cloud \\nocite{mgi+79,cdt+82} (Mazets et al.~1979; Cline et al.~1982), three other SGRs have been found close to the Galactic plane \\nocite{mgg79,lff+86,kkw+98} (1900+14: Mazets, Golenetskii \\& Gur'yan 1979; 1806--20: Laros et al.~1986; 1627--41: Kouveliotou et al.~1998b). Recently, Cline et al.~(2000) \\nocite{cfg+00} report a possible fifth SGR, 1801--23. Several lines of evidence support the notion that SGRs are magnetars --- young, slowly-rotating neutron stars with intense ($\\sim 10^{14}$ G) magnetic fields \\nocite{dt92a,td95,td96b} (Duncan \\& Thompson 1992; Thompson \\& Duncan 1995,1996). Firstly, all SGRs presently known are located close to or within a supernova remnant. SGR~1900+14 lies just outside the remnant G42.8+0.6 \\nocite{vkfg94,hkp+99} (Vasisht et al.~1994; Hurley et al.~1999a). Secondly, coherent pulsations in the range 5--8 s have been observed in three SGRs \\nocite{kds+98,hkms98} (Mazets et al.~1979; Kouveliotou et al.~1998a; Hurley et al.~1998). Finally, the measurement of high spin-down rates in two SGRs \\nocite{ksh+99,mrl99} (Kouveliotou et al.~1998a,1999; Marsden, Rothschild \\& Lingenfelter 1999) are suggestive of a neutron star spinning down via relativistic wind emission. Following the precise localization and discovery of coherent 5.16-s X-ray pulsations from SGR~1900+14 during an intense outburst in 1998 \\nocite{hkms98} (Hurley et al.~1998), we undertook a high-sensitivity search for radio pulsations from SGR~1900+14 using the upgraded Arecibo telescope. In this {\\it Letter}, we report on these observations and discuss their implications. The pulsar search is described in \\S \\ref{sec:search}. Although we did not detect the 5.16-s periodicity, we did find a young radio pulsar, J1907+0918, located 2 arcmin from SGR~1900+14. In \\S \\ref{sec:1907} we summarize 15 months of Arecibo timing observations of PSR~J1907+0918. As we discuss in \\S \\ref{sec:disc}, our discovery of PSR~J1907+0918 is the latest in a number of independent pieces of evidence which challenges the proposed association between SGR~1900+14 and G42.8+0.6. ", "conclusions": "\\label{sec:disc} Although our radio search for the 5.16-s pulsations from SGR~1900+14 was unsuccessful, the discovery of PSR~J1907+0918 has important implications for the origins of SGR~1900+14 and its proposed association with the supernova remnant G42.8+0.6. \\subsection{PSR~J1907+0918 and SGR~1900+14} The angular separation between PSR~J1907+0918 and SGR~1900+14 is $\\sim 2$ arcmin. Apart from globular cluster pulsars and double neutron star binaries, this is the closest pair of neutron stars on the plane of the sky. Either the two stars are physically close to each other (at the dispersion-measure distance of J1907+0918, 7.7 kpc, the projected separation is only 5 pc), or they are at different distances and only appear close when seen in projection. To investigate the case of a simple geometric projection, we require an estimate of the expected number of pulsars per unit area that would be detectable in our deep search. Cordes \\& Chernoff (1997) provide a number of useful analytic expressions for these purposes. Starting from their equation (8) we find the mean search volume $\\overline{V}$ to be proportional to $S_{\\rm min}^{-3/2}$. Here for simplicity we have neglected any strong period dependence upon the sensitivity (reasonable for long-period pulsars) and weighted the mean over a pulsar luminosity function of the form $d\\log N/d\\log L=-1$ \\nocite{lbdh93} (see e.g.~Lorimer et al.~1993). Under the simplest assumption that the number density of pulsars is approximately constant, the expected number of pulsars $N \\propto \\overline{V} \\propto S_{\\rm min}^{-3/2}$. To estimate the pulsar detection rate in our search we extrapolate the results of the current Parkes multibeam survey of the Galactic plane which is detecting about 1.5 pulsars deg$^{-2}$ within $|b|<1^{\\circ}$ around the location of PSR J1907+0918 on the Galactic plane (F.~Camilo, private communication). Under the above assumptions, we would expect a deep Arecibo search of this part of the plane to detect around $1.5\\times2.5^{3/2}\\simeq6$ pulsars deg$^{-2}$. Since the width of the 1410-MHz beam is around 3.5 arcmin, this corresponds to about one pulsar detection every 60 telescope pointings. It is therefore not unreasonable to appeal to simple chance detection in this case where we have only made one pointing at 1410 MHz. Aside from simple probability calculations, there is one other reason to believe that the apparent close angular proximity between PSR~J1907+0918 and SGR~1900+14 is simply a projection effect. If we were to assume that the two neutron stars had a common binary origin, this would require a binary system in which both stars are sufficiently massive to undergo a supernova explosion. Although theoretical arguments can be made to explain the close proximity of PSR~B1853+01 and PSR~B1854+00 \\nocite{wcd91} (Wolzczan Cordes \\& Dewey 1991) i.e.~that both these neutron stars were formed from a massive binary system, the ages of the radio pulsars in this case differ by over 100 million years. The similar ages of SGR~1900+14 and PSR~J1907+0918 (see below) would require the progenitor stars to have essentially identical main sequence lifetimes and hence masses. For example, to produce two neutron stars that differ in age by only $10^4$ yr, we estimate from eq.~1.3.9 of Shapiro \\& Teukolsky (1983) \\nocite{st83} that main sequence stars of the order of 6 M$_{\\odot}$ would differ in their absolute initial mass by only 0.05 $M_{\\odot}$, somewhat unlikely given the mass ratio distributions of binary stars (see e.g.~Dewey \\& Cordes 1987 \\nocite{dc87} and references therein). Both on statistical and evolutionary grounds, we conclude that SGR~1900+14 and PSR~J1907+0918 did not share a common origin. \\subsection{Which neutron star is associated with G42.8+0.6?} As demonstrated in \\S \\ref{sec:1907}, J1907+0918 is clearly a young pulsar. Given that plausible cases for an association with a supernova remnant can be made for several young radio pulsars with similar characteristic ages to J1907+0918, it is appropriate to revisit the case for the association between SGR~1900+14 and the supernova remnant G42.8+0.6 following our discovery of PSR~J1907+0918. There are three possibilities to be considered: (1) SGR~1900+14 is the neutron star produced in the supernova explosion that produced G42.8+0.6; (2) PSR~J1907+0918 is associated with G42.8+0.6; (3) neither of these neutron stars are associated with G42.8+0.6. In order to make a good case for {\\it any} neutron star-supernova remnant association, the following criteria should be satisfied (see \\nocite{kas96} Kaspi 1996): (a) the distances to both objects should agree; (b) the ages of both objects should agree; (c) the implied transverse velocity, based on the neutron star offset from the remnant center and the age, should be reasonable. We now review the current evidence for both SGR~1900+14 and PSR~J1907+0918 in connection with what is known about G42.8+0.6. (a) Distance estimates: PSR~J1907+0918 is estimated to lie at 7.7 kpc based on its dispersion measure and assuming the Taylor \\& Cordes (1993) Galactic electron density model. The statistical uncertainty in this model is at least 25\\%. Hurley et al.~(1999b) \\nocite{hlk+99} estimate the distance to SGR~1900+14 to be 5.7 kpc based on a spectral analysis of {\\it ASCA} data, although no estimate of the uncertainty in this measurement is quoted. The distance to G42.8+0.6 is commonly taken in the literature to be 5 kpc (see e.g.~Vasisht et al.~1994). Whilst this appears to be in agreement with SGR~1900+14, it should be stated that the latter distance was derived using the notoriously unreliable $\\Sigma$--$D$ relationship. A more recent $\\Sigma$--$D$ study places G42.8+0.6 at $10\\pm3$ kpc \\nocite{cb98} (Case \\& Bhattacharya 1998). Given the considerable uncertainties associated with this technique, it would obviously be premature to rule out an association between G42.8+0.6 and either of the two neutron stars in question. In this regard, we note that the recent discovery by Vrba et al.~(2000) \\nocite{vhl+00} of a massive star cluster only 12 arcsec from SGR~1900+14 suggests that it may have been formed in this cluster rather than G42.8+0.6. Vrba et al.~estimate the star cluster to lie at 14.5 kpc. (b) Age estimates: The ages of any non-historical supernova remnants are strongly coupled with their distances since absolute remnant sizes, along with assumptions about the expansion velocities are required to constrain the ages. Hence the age of G42.8+0.6 is also subject to considerable uncertainty. Vasisht et al.~(1994) quote an age of $10^4$ yr but, given the above range of distance estimates, this could easily be uncertain by factors of a few. For SGR~1900+14, the traditional assumptions about dipolar spin-down are thought not to apply and the age is quite uncertain with current estimates of $\\sim10^4$ yr (see e.g.~Kouveliotou et al.~1999). For PSR~J1907+0918, the 38-kyr characteristic age is probably indicative of its true age. This is somewhat model dependent since the age would be reduced if e.g.~the birth spin period of PSR~J1907+0918 was close to its current value or even increased if the neutron star braking is less than that expected from pure magnetic dipole braking (see e.g.~Manchester \\& Taylor 1977). In summary, based on currently-available evidence we conclude that both neutron stars appear to be young enough to be considered as plausible candidates for an association with G42.8+0.6. (c) Transverse speed estimates: Both neutron stars lie about 20 arcmin from the center of G42.8+0.6 which implies a transverse velocity of $4000\\,\\,D_7/t_4$ km s$^{-1}$ to carry either of them to their present position. Here $D_7$ is the distance in units of 7 kpc and $t_4$ is the age in units of $10^4$ yr. Although the exact values of $D_7$ and $t_4$ are highly uncertain, it is unlikely that they are such that the required velocity estimate is below 1000 km s$^{-1}$. For either of the neutron stars, then, the implied transverse velocities would place them at the far extremes of the presently-observed distribution \\nocite{hla93} (Harrison, Lyne \\& Anderson 1993). To ultimately test for an association between PSR~J1907+0918 and G42.8+0.6, and constrain the age of the pulsar, a proper motion measurement is required. The predicted pulsar proper motion is $120/t_4$ mas yr$^{-1}$. Future VLBI proper motion measurements of PSR~J1907+0918, perhaps using Arecibo-Effelsberg-GBT are highly desirable. To summarize, based on the currently-available information, we conclude that the proposed association between G42.8+0.6 and SGR~1900+14 is, contrary to frequent claims in the literature, far from secure since there is no reason against arguing equally strongly in favor of PSR~J1907+0918 as being the neutron star produced in the supernova explosion rather than SGR~1900+14. Indeed, the additional possibility that neither of these young neutron stars is associated with G42.8+0.6 remains attractive at this stage! As noted by Gaensler \\nocite{gae00} (2000), large positional offsets between neutron stars and supernova remnants are more likely a result of random line-of-sight alignments rather than genuinely associated high-velocity neutron stars. This may well be the case here and further observations (e.g.~deeper multi-wavelength maps of the region and proper-motion measurements) are clearly desirable to help resolve this most perplexing situation." }, "0005/astro-ph0005030_arXiv.txt": { "abstract": "We have obtained the full 1-200~${\\rm \\mu m}$ spectrum of the low luminosity (36~$\\rm L_{\\odot}$) Class~I protostar \\object{Elias~29} in the $\\rho$~Ophiuchi molecular cloud. It provides a unique opportunity to study the origin and evolution of interstellar ice and the interrelationship of interstellar ice and hot core gases around low mass protostars. We see abundant hot CO and ${\\rm H_2O}$ gas, as well as the absorption bands of CO, ${\\rm CO_2}$, ${\\rm H_2O}$ and ``6.85~${\\rm \\mu m}$'' ices. We compare the abundances and physical conditions of the gas and ices toward \\object{Elias~29} with the conditions around several well studied luminous, high mass protostars. The high gas temperature and gas/solid ratios resemble those of relatively evolved high mass objects (e.g. GL~2591). However, none of the ice band profiles shows evidence for significant thermal processing, and in this respect \\object{Elias~29} resembles the least evolved luminous protostars, such as \\object{NGC~7538~:~IRS9}. Thus we conclude that the heating of the envelope of the low mass object \\object{Elias~29} is qualitatively different from that of high mass protostars. This is possibly related to a different density gradient of the envelope or shielding of the ices in a circumstellar disk. This result is important for our understanding of the evolution of interstellar ices, and their relation to cometary ices. \\keywords {Stars: formation -- individual: Elias~29 -- ISM: dust, extinction -- molecules -- abundances -- Infrared: ISM: lines and bands} ", "introduction": "~\\label{se29:intro} The general picture of low mass star formation has been formed since the 1980's with the availability of infrared and millimeter wavelength broad band photometry from the ground, and with the IRAS satellite and KAO observatory (e.g., Lada \\& Wilking \\cite{lad84}; Adams et al. \\cite{ada87}; Hillenbrand et al. \\cite{hil92}; Andr\\'e et al. \\cite{and93}). A classification scheme was made, where the continuum emission of Class~0 and~I objects peaks in the submillimeter and far-infrared. These objects are still deeply embedded in their accreting envelopes. In the Class~II phase, the wind of the protostar has cleared its surrounding environment, such that it becomes optically visible, and shows \\ion{H}{i} emission lines. The continuum emission of these objects peaks in the near-infrared, but there is still significant excess emission above the stellar continuum. They are believed to be surrounded by optically thick dusty disks. Finally, little dust emission remains for Class~III objects, when the disk is optically thin, and planetary companions may have been formed. Our knowledge of the physical and chemical state and evolution of the material surrounding protostars, has progressed with the availability of medium and high resolution spectroscopic instrumentation at near and mid-infrared wavelengths ($\\sim2-20$~${\\rm \\mu m}$). The progress made, is best illustrated by the observations of high mass protostars, which are bright and easy to observe. The (ro-)vibrational bands of various molecules (CO, ${\\rm H_2O}$, ${\\rm CH_3OH}$, silicates) were observed from the ground, revealing profile variations of the 3.07~${\\rm \\mu m}$, and 4.67~${\\rm \\mu m}$ ${\\rm H_2O}$ and CO ice bands (e.g. Smith et al. \\cite{smi89}; Tielens et al. \\cite{tie91}; Chiar et al. \\cite{chi98}). This was interpreted as evaporation of the volatile CO ice, and crystallization of ${\\rm H_2O}$ ice in the molecular envelopes. This heating effect is strengthened by the detection of hot gas in 4.6~${\\rm \\mu m}$ CO observations (Mitchell et al. \\cite{mit91}). With the launch of the {\\it Infrared Space Observatory} in 1995 (ISO; Kessler et al. \\cite{kes96}), it became possible to observe all other molecular bands in the infrared (Whittet et al. \\cite{whi96}). It was shown that high mass protostellar evolution can be traced in the gas-to-solid abundance ratios (van Dishoeck et al. \\cite{dis96}; van Dishoeck \\& Blake \\cite{disb98}), and the profiles of the ice bands, in particular solid ${\\rm CO_2}$ (Gerakines et al. \\cite{ger99}; Boogert et al. \\cite{boo00}). Thus, there is overwhelming evidence that thermal processing, i.e. evaporation and crystallization of ices in and around hot molecular cores, plays an important role in the evolution of high mass molecular envelopes. The composition and evolution of the molecular material around low mass protostars are not as well studied. It seems unlikely that the molecular material evolves similar to that around high mass protostars. Low mass protostars evolve much slower, release less radiative energy, drive less energetic winds, and form disks. It is not established whether low mass objects possess hot cores as well, and whether the ices survive the process of star formation. If (some of) the ices survive, are they included into comets, and if so, are the ice structure and composition still the same compared to interstellar ices? How important are energetic processes, such as cosmic ray bombardment, in altering the ice composition on the long time scale of the formation of low mass stars? To investigate the influence of low mass protostars on their molecular envelope, we make an infrared spectroscopic study \\object{Elias~29}, also called WL~15 and YLW~7 (Elias \\cite{eli78}, Wilking et al. \\cite{wil83}, Young et al. \\cite{you86}). On a large scale, \\object{Elias~29} lies in core E, which is in the south-east corner of the 1$\\times$2~pc extended compact CO ridge L~1688 (Loren et al. \\cite{lor90}) in the densest part of the $\\rho$~Ophiuchi cloud, at a distance of $\\sim$160~pc from the earth (Wilking \\& Lada \\cite{wil83}; Whittet \\cite{whi74}). It is the reddest object found in the near-infrared survey of this cloud by Elias (\\cite{eli78}), without a counterpart at optical wavelengths. For our observations, we used Elias' coordinates (J2000): \\begin{center} $\\rm \\alpha~=~16^h27^m09^s.3$ \\hspace{20pt} $\\rm \\delta~=~-24^o37'21''$. \\end{center} The overall spectrum of \\object{Elias~29} is typical for a heavily embedded Class~I source, probably in a late accretion phase (Wilking et al. \\cite{wil89}; Andr\\'e \\& Montmerle \\cite{and94}; Greene \\& Lada \\cite{gre96}; Saraceno et al. \\cite{sar96}). The embedded nature is also revealed by its high extinction, and by the cold compact envelope observed at millimeter wavelengths (Andr\\'e \\& Montmerle \\cite{and94}; Motte et al. \\cite{mot98}). \\object{Elias~29} is associated with a molecular outflow (Bontemps et al. \\cite{bon96}; Sekimoto et al. \\cite{sek97}). With a bolometric luminosity of $\\sim 36~L_{\\odot}$ (Chen et al. \\cite{che95}), \\object{Elias~29} is the most luminous protostar in the $\\rho$~Oph cloud, which makes this source very suitable for spectroscopic studies. The relatively high luminosity, and high bolometric temperature ($T_{\\rm bol}\\sim 410$~K) imply an age in the range 0.5--4 10$^5$~yr (Chen et al. \\cite{che95}). In the pre-main-sequence evolutionary tracks of Palla \\& Stahler (\\cite{pal93}), this corresponds to a star with end mass 3.0--3.5 $M_{\\odot}$. \\object{Elias~29} might thus be a precursor Herbig AeBe star. This classification is however uncertain. For example, it has been argued from the SED, and the absence of mid-infrared emission features, that \\object{Elias~29} is a 1 $M_{\\odot}$ protostar with a large accretion luminosity, and a spectral type of K3-4 at the birth line (Greene \\& Lada \\cite{gre00}). This Paper is structured as follows. Technical details on the ISO infrared observations are given in Sect.~\\ref{se29:obs}. All the observed emission and absorption features are discussed in detail in Sect.~\\ref{se29:res}. Section~\\ref{se29:sed} gives a description of the continuum shape, and a comparison to other lines of sight. The ice composition and thermal history, and the silicate band depth with inferred extinction and column densities toward \\object{Elias~29} are discussed in Sect.~\\ref{se29:ice}. Then, numerous lines of gaseous CO and ${\\rm H_2O}$ are detected, and modeled to derive gas temperatures and column densities (Sect.~\\ref{se29:gas}). The molecular abundances and gas-to-solid ratios of \\object{Elias~29} are compared to a sample of sight-lines, ranging from dark cloud cores to evolved protostars. A comparison with high mass protostars is made (Sect.~\\ref{se29:abun}). Section~\\ref{se29:geom} discusses the origin of the wealth of observed emission and absorption features and puts them in a geometrical picture, where we review the evidence for an extended envelope and an accretion disk. We conclude in Sect.~\\ref{se29:summary} with a summary and suggestions for future observations. ", "conclusions": "~\\label{se29:summary} The 1.2--195~${\\rm \\mu m}$ spectrum of the low mass protostellar object \\object{Elias~29} in the $\\rho$~Ophiuchi molecular cloud shows a wealth of absorption lines of gas and solid state molecules. Hot CO and ${\\rm H_2O}$ gas are detected ($T_{\\rm ex}>$300~K) at rather high abundances, on scales of not more than a few hundred AU. The ice abundances are relatively low. In this respect, \\object{Elias~29} resembles luminous protostars with significantly heated cores, such as GL~2591. However, {\\it none} of the many ice bands that are detected, i.e. from ${\\rm H_2O}$, CO, ${\\rm CO_2}$, and the 6.85~${\\rm \\mu m}$ band, shows outspoken signs of thermal processing. Again in comparison with luminous protostars, \\object{Elias~29} now resembles less evolved objects, such as \\object{NGC~7538~:~IRS9}. Our combined gas and solid state analysis thus shows that high and low mass protostars heat their molecular envelopes in different ways. This may be related to their different structure, such as the presence of a circumstellar disk in low mass protostars. The hot gas of \\object{Elias~29} could be present on the surface of a flaring disk, which is efficiently heated by the central star. The ices toward \\object{Elias~29} must be well shielded in a circumstellar disk seen close to edge-on, or far away in the envelope. Does this imply that in general the ices in the disks or outer envelopes of low mass protostars remain unaltered, both in composition and structure, during the process of star formation? Are these ices the building blocks of the early solar system and are they preserved in present day observed cometary nuclei? To date, no Class~I protostar has been found with strong signs of crystalline ices (Boogert et al. \\cite{boo00}). On the other hand, the presence of crystalline ices and silicates has been reported in several isolated, less embedded Herbig AeBe objects (Malfait et al. \\cite{mal99}). This research needs to be extended to a larger sample of low mass protostars, in a range of evolutionary stages and luminosities. Furthermore, it is essential for the interpretation of the gas and solid state characteristics toward Elias 29 that the presence of a circumstellar disk, and its inclination are determined by future high spatial resolution infrared or millimeter continuum observations." }, "0005/astro-ph0005206_arXiv.txt": { "abstract": "The canonical cosmological constant dominated cold dark matter model (\\lcdm) may possess too much power on small scales at $z=0$, manifested as central over-concentration of dark matter and over-abundance of dwarf galaxies. We suggest an alternative model, $\\Lambda$DCDM, where one half of the cold dark matter particles decay into relativistic particles by $z=0$. The model successfully lowers the concentration of dark matter in dwarf galaxies as well as in large galaxies like our own {\\it at low redshift}, while simultaneously retaining the virtues of the \\lcdm model. The model solves the problem of over-production of small dwarf galaxies in the \\lcdm {\\it not by removing them but by identifying them with failed, ``dark\" galaxies}, where star-formation is quenched due to dark matter evaporation and consequent halo expansion. A dramatic difference between the \\ldcdm model and other proposed variants of the \\lcdm model is that the small-scale power {\\it at high redshift} ($z>2$) in the \\ldcdm model is enhanced compared to the \\lcdm model. A COBE-and-cluster normalized \\ldcdm model can be constructed with the following parameters: $H_0=60$km/sec/Mpc, $\\lambda_0=0.60$, $\\Omega_{0,CDM}=0.234$, $\\Omega_{0,b}=0.044$, $n=1.0$, and $\\sigma_8=1.06$. A clean test of this model can be made by measuring the evolution of gas fraction in clusters. The prediction is that the gas fraction should decrease with redshift and is smaller by $31\\%$ at $z=1$ than at $z=0$. X-ray and Sunyaev-Zel'dovich effect observations should provide such a test. ", "introduction": "While the canonical \\lcdm model is remarkably successful in many ways (Ostriker \\& Steinhardt 1995; Bahcall \\etal 1999), there is now some tentative evidence that it may have too much power on small scales ($l\\sim 1-100$\\kpc) today. Evidence includes a large excess % of dwarf galaxies (Klypin \\etal 1999; Moore \\etal 1999), the over-concentration of dark matter in % dwarf galaxies % (Moore 1994; Flores \\& Primack 1994; Burkert 1995; McGaugh \\& de Blok 1998; Moore \\etal 1999) as well as in % large galaxies % (Navarro \\& Steinmetz 2000). Without impairing the many notable virtues of the \\lcdm model, in this {\\it Letter} we suggest a \\lcdm model in which one half of the CDM particles decay into relativistic particles by $z=0$. Decaying CDM was suggested before in attempts to save the CDM model in an Einstein-de Sitter universe % (Turner, Steigman, \\& Krauss 1984; Doroshkevich \\& Khlopov 1984; Olive, Seckel, \\& Vishniac 1985). Suggestions were also made in the context of neutrino models for similar rescue missions (Davis \\etal 1981; Hut \\& White 1984). ", "conclusions": "We have shown that, by allowing for one half of the CDM particles to decay by $z=0$ % into relativistic particles, the problem of excess small-scale power in the \\lcdm model is remedied. In essence, the decay of CDM particles results in reduction of CDM mass and expansion of the halo, and lowers the concentration of CDM in the inner region to brings the model into agreement with observations of dwarf galaxies. The problem of excess number of dwarf galaxies in the \\lcdm is solved by the same mechanism but manifested in a quite different way: instead of suppressing the number of predicted dwarf halos, % we argue that these dwarf halos failed to form a sufficient amount of stars to be identified as dwarf galaxies even in our local neighborhood. It is important to search for these gas-rich (not necessarilly neutral gas-rich) dark galaxies in the Local Group. The model is consistent with COBE, the local abundance of rich clusters of galaxies, the age constraint and $q_0$ from high-z SNe. A test of the model will be provided by measuring the evolution of gas fraction in clusters. The prediction is that the gas fraction should decrease with redshift and is smaller by $31\\%$ at $z=1$ than at $z=0$. X-ray and SZ effect observations should provide such a test." }, "0005/astro-ph0005083_arXiv.txt": { "abstract": "We present mid-infrared maps and preliminary analysis for 61 galaxies observed with the ISOCAM instrument aboard the {\\it Infrared Space Observatory}. Many of the general features of galaxies observed at optical wavelengths---spiral arms, disks, rings, and bright knots of emission---are also seen in the mid-infrared, except the prominent optical bulges are absent at 6.75 and 15 \\m. In addition, the maps are quite similar at 6.75 and 15 \\m, except for a few cases where a central starburst leads to lower \\ISOcolor\\ ratios in the inner region. We also present infrared flux densities and mid-infrared sizes for these galaxies. The mid-infrared color \\ISOcolor\\ shows a distinct trend with the far-infrared color \\IRAScolor. The quiescent galaxies in our sample (\\IRAScolor\\ $\\lesssim 0.6$) show \\ISOcolor\\ near unity, whereas this ratio drops significantly for galaxies with higher global heating intensity levels. Azimuthally-averaged surface brightness profiles indicate the extent to which the mid-infrared flux is centrally concentrated, and provide information on the radial dependence of mid-infrared colors. The galaxies are mostly well resolved in these maps: almost half of them have $<$ 10\\% of their flux in the central resolution element. A comparison of optical and mid-infrared isophotal profiles indicates that the optical flux at 4400 \\AA\\ near the optical outskirts of the galaxies is approximately eight (seven) times that at 6.75 \\m\\ (15 \\m), comparable with observations of the diffuse quiescent regions of the Milky Way. ", "introduction": "Normal galaxies, defined as those with on-going star formation, account for most of the luminous mass in the local Universe. Their luminosity is derived from stars and they span a broad range of observed morphologies, luminosities, and infrared-to-blue ratios. Results from \\IRAS\\ have shown that the infrared colors from the four \\IRAS\\ bands are sensitive indices of the radiation intensity in the interstellar medium. The mid-infrared (5-20 \\m) emission is dominated by very small grains fluctuating to high temperatures and polycyclic aromatic hydrocarbons (PAHs) (Helou et al. 2000). These grains are not in thermal equilibrium but they still convert heating photons and thereby trace star formation. Larger grains in thermal equilibrium dominate the emission from normal galaxies at longer wavelengths. By comparing the mid-infrared emission to other components of the galaxy such as \\HI, H$_{2}$, ionized gas, and starlight, one can derive the physical properties of the interstellar gas, dust and radiation field in galaxies (e.g. Vigroux et al. 1999). The {\\it Infrared Space Observatory} (\\ISO) U.S. Key Project on Normal Galaxies (PI: G. Helou, proposal id: SF\\_GLX\\_*, hereafter the ``Key Project'') was proposed to study the interstellar medium of a broad range of normal galaxies using three of the four instruments aboard \\ISO: ISOCAM, ISOLWS, and ISOPHOT. Under NASA guaranteed time, the U.S. Key Project obtained \\ISO\\ observations of 69 galaxies. Nine relatively nearby and extended galaxies were chosen to provide spatially resolved cases so that various phases of the interstellar medium could be studied independently. The remaining 60 galaxies in the Key Project sample cover the full range of observed morphologies, luminosities, and \\IRAS\\ colors seen in normal galaxies. Using this diverse sample, we hope to gain new insight into the star formation process on the scale of galaxies, especially its drivers and inhibitors (Helou et al. 1996). ", "conclusions": "We present mid-infrared maps at 6.75 and 15 \\m\\ for 61 normal star-forming galaxies; for a subset of 13 of these galaxies we also show maps at 4.5 \\m. All galaxies for which observations were attempted at these wavelengths were successfully detected. Qualitatively, the mid-infrared morphology is not a strong function of wavelength, and many of the optical features of the galaxies are also observed in the mid-infrared, except for the bulges of spiral galaxies, consistent with the findings of Helou et al. (1996) in NGC 6946. Moreover, the data support non-negligible photospheric contributions at 4.5 \\m\\ for galaxies exhibiting low FIR/B ratios, consistent with the conclusion drawn from mid-infrared ISOPHOT spectroscopy for normal galaxies (Helou et al. 2000; Lu et al. 2000). However, the evidence presented here is tenuous, as we only have broad band data at 4.5 \\m\\ for a small number of galaxies. Mid-infrared curve of growth profiles indicate that the mid-infrared emission is very well resolved by the ISOCAM maps for most of these galaxies. Moreover, the profiles are generally exponential in nature, and the distribution of 6.75 and 15 \\m\\ emission is quite similar for most galaxies. However, four of the six galaxies that show an enhanced \\ISOcolorb\\ color also show signs of active central star formation, with yet greater enhancement of \\ISOcolorb\\ in the central regions. Quiescent galaxies, those showing low global interstellar heating intensities (i.e. \\IRAScolor\\ $\\geq 0.6$), have an almost constant \\ISOcolor\\ color near unity. For galaxies with higher global heating intensities, the mid-infrared color drops rapidly with increasing far-infrared color. We interpret this as evidence for global mid-infrared spectral energy distributions that are increasingly dominated by \\HII\\ region emission, characterized by a relatively steep slope in the mid-infrared continuum and a depressed contribution from PAHs. It is interesting to note that galaxies of all morphological types appear to follow the same color-color trend. We have estimated the average mid-infrared surface brightness at which the mid-infrared semi-major axis matches that in the optical (at the $B$-band 25 mag arcsec$^{-2}$ level). We find \\RB\\ $\\approx$ \\Rmira\\ at 0.04 MJy sr$^{-1}$ and \\RB\\ $\\approx$ \\Rmirb\\ at 0.09 MJy sr$^{-1}$ on average for this sample. These mid-infrared surface brightness levels are consistent with observations of Galactic cirrus and the Solar Neighborhood, implying a reasonable similarity in interstellar heating intensity for the outskirts of normal galaxies and our Galaxy. A final interesting finding from the mid-infrared size analysis centers on the ratio of total heating output to the total heating input for quiescent regions. We find that ratio to be of order 0.7." }, "0005/astro-ph0005560_arXiv.txt": { "abstract": "We present X-ray emission line equivalent width images of the bright Galactic supernova remnant Cassiopeia A for the elements Si, S, Ar, Ca, and Fe using a 50,000 s observation with the Advanced CCD Imaging Spectrometer on the Chandra X-ray Observatory. The images essentially identify the bulk of detectable ejecta of these elements over a wide range of surface brightness, and show morphologies distinctly different from that of the broadband X-ray emission and of the 4-6 keV continuum emission. The Si, S, Ar, and Ca maps, while different in turn from those of Fe, show that these X-ray emitting ejecta are distributed similarly to the fast optical ejecta knots, and clearly delineate the X-ray counterpart of the northeast optical jet. Low surface brightness regions just outside the bright shell in the north and west are also shown to have strong line emission. The strong Fe emission is exterior to that of other elements in the east, as previously noted, but is generally coincident elsewhere. The projected interior has relatively little emission traced by high line equivalent widths. ", "introduction": "A record of a supernova explosion is imprinted in the spatial distribution of ejecta in the remnant, but then is gradually erased by dynamic interaction with the surrounding medium. Cassiopeia A, the brightest and youngest known Galactic remnant, is one of the best targets for study of the ejecta distribution. Detailed observations and modelling at optical wavelengths (e.g., Reed et al. 1995, Lawrence et al. 1995, Fesen \\& Gunderson 1996) show a complex system of chemically enriched knots with high velocities that form a shell with a linear extension to the northeast called the jet. Although these ejecta are nearly undecelerated, they show asymmetries in their red- and blue-shifted velocities and in their line intensity ratios that arise from their interaction with an inhomogeneous medium. Optical emission, however, accounts for only a small fraction of the total mass, as typical shocks in young supernova remnants heat the bulk of the mass to X-ray emitting temperatures. X-ray spectral imaging, which is required to identify and map these ejecta, has been carried out most extensively with the ASCA Observatory (e.g., Holt et al. 1994, Fujimoto et al. 1995, Hwang \\& Gotthelf 1997; also Vink et al. 1999 with SAX). The moderate spectral resolution of the ASCA CCD detectors allows strong line blends of individual elements to be imaged with the 3$'$ half-power diameter point spread function of the ASCA mirrors. With the launch of the Chandra X-ray Observatory comes a tremendous increase in the capability to make these observations. The Chandra mirrors provide $<0.5''$ imaging capability, which is on par with excellent ground-based optical observations, while the Advanced CCD Imaging Spectrometer (ACIS) on Chandra gives spectral resolution comparable to that of the ASCA CCDs. Cas A was the Chandra first light target in 1999 August, and Hughes et al. (2000) use this early observation to note the existence of Fe-dominated zones in the east that are exterior to Si-dominated zones. Because Fe is synthesized in the innermost layers of the star, this is interpreted as an overturning of ejecta layers in this region during the explosion. In this paper, we present the global distribution of the X-ray line emission in Cas A using a new Chandra observation that is ten times longer than any previous one with ACIS. We account for the locally variable underlying continuum to produce maps of the line-to-continuum ratio that reveal the distribution of strong line emission throughout the remnant over a wide range of surface brightness. These are the first such images that have been produced for X-ray emission from all these elements in any young supernova remnant. ", "conclusions": "It can be verified that the maps in Fig. 3 trace gas that is enriched with supernova ejecta by comparing the observed line intensity ratios to theoretical line/continuum emissivity ratios. Using the approximate Si energy range shown in Fig. 1 for a solar abundance plasma with temperatures between $10^6-10^8$ K and ionization ages from $10^9-10^{12.4}$ cm$^{-3}$ s, the calculated ratio of the Si line emissivity relative to the continuum does not exceed 2 (and this only at extreme temperatures). A ratio of 2 traces the bright regions of the Si EQW image in Fig. 3, which is consistent with the consensus that the X-ray Si emission in Cas A comes predominantly from ejecta. A similar calculation for Fe K gives maximum ratios of just over 2 for the same range of parameters, with the highest computed ratios occurring at temperatures of a few keV near collisional ionization equilibrium. A ratio of 2 again outlines the Fe K EQW image in Fig. 3, suggesting that much of this emission may be from ejecta. Another possibility has been suggested by Borkowski et al. (1996), who modelled the Fe K emission in the integrated ASCA spectrum as arising from shocked circumstellar wind material. The images do not rule out the possibility that some of the Fe K emission does come from such a contribution, and to some extent, this could also be true of the emission from Si and other elements. The resolution of this issue requires detailed examination of the spectra, which we defer to future work. The images in Fig. 3 give a good indication of the distribution of X-ray emitting ejecta, but we emphasize that they do not give detailed abundance information. Theoretical equivalent widths vary with temperature and ionization age, so that a given equivalent width can correspond to significantly different element abundances depending on the local plasma conditions. The EQW maps by themselves should therefore be taken only as an indication of where the abundances are definitely enhanced above the solar value. The details of the maps, particularly in faint emission regions, are moreover dependent on the rather simple assumptions that the local continuum is linear in shape and free of contamination from weak lines, and that the line emission in each image is due entirely to the element identified. Though this last assumption is likely to be quite valid in most cases, it may not be valid for Fe L, nor would it be for Mg. Finally, the EQW maps by themselves are not actual maps of the ejecta mass, as the mass depends not only on the abundance and emitting volume but also on the local density. The resemblance between our Si and S line maps and optical maps explicitly demonstrates that the X-ray and optically emitting ejecta are spatially coexistent over much of the remnant. This was anticipated in that the X-ray Si and S line emission, inferred to include a significant contribution from ejecta, show asymmetric Doppler shifts corresponding to bulk velocities of $\\sim$2000 km/s that are comparable to those of the optical ejecta knots (Markert et al. 1983, Holt et al. 1994). The ejecta in a given region therefore have a wide range of densities and temperatures, and the shocks propagating into the ejecta are highly variable on small scales; the slow shocks required for the optical emission (Reed et al. 1995) cannot explain the X-ray emission. The faint X-ray jet extends more than 3.5$'$ beyond the expansion center used by Kamper \\& van den Bergh (1976, a position nearly coincident with that of the newly discovered X-ray point source) making it of comparable extent to its optical counterpart. The qualitative resemblance between the X-ray image of the jet and the optical images of Fesen \\& Gunderson (1996) is striking. In particular, the three prominent filaments of the outer optical jet and the triangular shaped base are all seen to have X-ray counterparts. Our maps also show that the abundances of Si and S are enhanced in faint regions just outside the bright shell in the north and west, but they do not reveal any other extended region of enhanced abundance comparable to the northeast jet. The Si image, containing the highest number of photons, does show faint streamers of emission extending radially outward directly opposite the jet to the west, but with a larger opening angle, and with equivalent widths that are not high enough to indicate unambiguously that these are ejecta. We have noted that Si, S, Ar, and Ca have similar general morphologies, but examination of the individual panels of Fig. 3 does show differences between them. A detailed survey of such differences is beyond the scope of this paper, but two notable examples are the brightest knots in the Si image of Fig. 2, at the inner boundary of the shell to the northeast and southeast: these are much weaker in the Ar and Ca EQW maps than in Si and S. The southeast knot was analyzed qualitatively by Hughes et al. (2000), and it appears that the Ar and Ca abundances here are indeed lower, and that these ejecta originate in an outer ejecta zone compared to knots with strong emission for all four elements. The overall difference between the Si and Fe EQW maps in the east appears to support the suggestion that the inner Fe ejecta layers have been overturned and propelled beyond the Si ejecta layers in this part of the remnant, and may be related to the asymmetries that created the jet nearby. Such overturning of ejecta layers was first suggested by Chevalier \\& Kirshner (1979 and references therein) to explain the properties of the optical knots, and also occurs in recent models (e.g., Kifonidis et al. 2000, Khokhlov et al. 1999). In summary, the X-ray line maps presented in this paper reveal the spatial striking correspondence between the optically and X-ray emitting ejecta in Cas A, explicitly showing that the different physical conditions giving rise to each coexist on small spatial scales." }, "0005/astro-ph0005426_arXiv.txt": { "abstract": "Using an Extreme Ultraviolet Explorer (EUVE) spectrum, we confirm the identification of a white dwarf companion to the B9.5V star 16 Dra (HD150100), and constrain its surface temperature to lie between 29,000K and 35,000K. This is the third B star $+$ white dwarf non-interacting Sirius-type binary to be confirmed, after y Pup (HR2875, HD59635) and $\\theta$ Hya (HR3665, HD79469). 16 Dra and its white dwarf companion are members of a larger resolved proper motion system including the B9V star 17 Dra A (HD150117). The white dwarf must have evolved from a progenitor more massive than this star, $M_{\\rm MS}\\approx3.7M_\\odot$. White dwarf companions to B stars are important since they set an observational limit on the maximum mass for white dwarf progenitors, and can potentially be used to investigate the high mass ends of the initial-final mass relation and the white dwarf mass-radius relation. ", "introduction": "Unresolved Sirius-type binary systems consisting of a white dwarf and a main sequence star (spectral type B$-$K) are difficult to identify optically, since the bright main sequence companion completely swamps the degenerate star's flux. However, through the ROSAT Wide Field Camera (WFC, Pounds et~al. 1993) and Extreme Ultraviolet Explorer (EUVE, Bowyer et~al. 1994) surveys, EUV radiation with the spectral signature of a hot white dwarf has been detected originating from apparently inactive main sequence stars, giving a clue to the existence of a previously unidentified population of Sirius-type binaries. Over 20 new systems have now been identified (e.g.~Barstow et~al. 1994, Burleigh et~al. 1997, Vennes et~al. 1998). For companions of spectral type $\\sim$A5 or later, far-ultraviolet spectra obtained with the International Ultraviolet Explorer (IUE) have been used to confirm the identifications, since the white dwarf is actually the brighter component at these wavelengths. Unfortunately, stars of spectral types O, B and early A will still dominate any emission from a white dwarf in the far-UV regime, and IUE or HST cannot be used to identify any putative degenerate companions to these objects. \\begin{figure} \\vspace{9cm} \\special{psfile=h1904.f1 hscale=50 vscale=50 hoffset=-80 voffset=0} \\caption{Optical image of the 16 Dra system from the Digitized Sky Survey. The field is 8$\\times$8 arcmin. The lower of the bright stars is 16 Dra. The upper bright star is the 17 Dra A/B pair, unresolved here but in fact separated by 3.2 arcsec. The circles denote, in decreasing order of size, the ROSAT WFC, EUVE and ROSAT PSPC source error boxes.} \\end{figure} \\begin{figure*} \\vspace{8cm} \\special{psfile=h1904.f2 hscale=55 vscale=45 hoffset=50 voffset=15} \\caption{EUVE short wavelength spectrum of 16 Dra, shown together with a pure-H white dwarf $+$ ISM model (solid line) for log $g=8.5$, $T_{\\rm eff}=31,900$K, $N_{\\rm HI}=8.2\\times10^{18}$ atoms cm$^{-2}$, $N_{\\rm HeI}=9\\times10^{17}$ atoms cm$^{-2}$, and $N_{\\rm HeII}=3\\times10^{17}$ atoms cm$^{-2}$. Inset (upper): the same data, displayed at the instrument resolution of $\\approx0.5${\\AA}. Inset (lower): EUVE short wavelength spectrum of Sirius~B, for comparison. } \\end{figure*} Two bright B stars, $\\theta$ Hya (HR3665) and y Pup (HR2875), were unexpectedly detected in the ROSAT and EUVE surveys. Since their soft X-ray and EUV colours were similar to many known hot white dwarfs, it was suspected that they too were hiding hot white dwarf companions. Fortunately, both EUV sources were bright enough to be observed by EUVE's spectrometers. y Pup was observed by EUVE in 1996, $\\theta$ Hya in 1998, and the formal discovery of these Sirius-type systems was subsequently reported by Vennes et~al. (1997, y Pup), Burleigh \\& Barstow (1998, y Pup), and Burleigh \\& Barstow (1999, $\\theta$ Hya). White dwarf companions to B stars are of significant importance since they must have evolved from massive progenitors, perhaps close to the maximum mass for white dwarf progenitor stars. They are also likely to be significantly more massive than the mean for white dwarf stars in general ($\\approx0.57M_\\odot$, Bergeron et~al. 1992). The value of the maximum mass for a white dwarf progenitor star, and hence the minimum mass for producing a Type II supernova through core collapse in a single star, is a long-standing astrophysical problem. Weidemann (1987) gives the limit as $\\sim8M_\\odot$ in his semi-empirical initial-final mass relation. Observationally, this limit is best set by the white dwarf companion to y Pup (HR2875). Echelle spectroscopy of this object by Vennes (2000) has recently revealed that this system comprises two main sequence B stars (B3.5V$+$B6V) in an eccentric $\\approx15$ day orbit, with the white dwarf forming a third, wider component. The white dwarf must then have evolved from a star more massive than B3.5V, $\\sim5.5M_\\odot$ (Vennes 2000). We also note that Bergh\\\"ofer et~al. (2000) have recently suggested that the spectroscopic companion to the B1.5IV star $\\lambda$ Sco might be a hot ultramassive white dwarf ($1.25M_\\odot$0.91 & Burleigh \\& Barstow (1998) \\\\ & & & & & & & Vennes et~al. (2000) \\\\ 16 Dra$^\\dagger$ & HD150100 & B9V & 3.7 & & 115$-$131 & $>$0.69 & this paper \\\\ $\\theta$ Hya$^\\dagger$$^\\dagger$ & HR3665 & B9.5V & 3.4 & $\\buildrel > \\over {_\\sim}$10 yrs & 38$-$41 & $>$0.68 & Burleigh \\& Barstow (1999) \\\\ Sirius B$^\\ddagger$ & & A0V & 3.25 & $\\approx$50 yrs & 2.637$\\pm$0.011 & 1.034$\\pm$0.026 & Holberg et~al. (1998) \\\\ Beta Crt.$^\\dagger$$^\\dagger$ & & A1III & 2.9 & $\\buildrel > \\over {_\\sim}$10 yrs$$ & 77$-$87 & 0.44 & Vennes et~al. (1998) \\\\ \\end {tabular} \\end{center} Main sequence masses from Allen (1973). \\\\ $\\star$ Suggested, not confirmed. Note that the Hipparcos distance estimate is inconsistent with the photometric distance ($\\sim140$pc). \\\\ An alternative explanation is that the system consists of two B stars. \\\\ $^\\dagger$ The spectral type of 16 Dra is in fact B9.5V. Its proper motion companion 17 Dra A is B9V, and thus any white dwarf in the system must have evolved from a progenitor more massive than this. See text for futher discussion.\\\\ $^\\dagger$$^\\dagger$ Micro-variability in the proper motions of these stars as measured by Hipparcos indicate binary periods $\\buildrel > \\over {_\\sim}$10 yrs. \\\\ $^\\ddagger$ Holberg et~al. (1998) use an initial-final mass relation to estimate the progenitor mass of Sirius B as 6$-$7M$_\\odot$. \\\\ \\end{table*} We have analysed the weak EUVE spectrum of the B9.5V star 16 Dra, and confirm that there is an unresolved hot white dwarf in the field. Fig.1 clearly shows that the white dwarf is not resolved from 16 Dra in the Digitized Sky Survey image, and their angular separation can be no more than $\\approx$30 arcsec. However, if the white dwarf lies at the same distance as 16 Dra and its proper motion companions 17 Dra A \\& B, and is related to them, then it must have evolved from a progenitor more massive than the earliest extant star in the system, 17 Dra A (B9V). Thus this degenerate has the second most massive progenitor among known white dwarfs. Table 4 lists the earliest type stars known to have white dwarf companions, including all three B star $+$ white dwarf binaries and Sirius. EUVE spectra provide us with little information with which to constrain a white dwarf's surface gravity, and hence its mass, but we can use a theoretical initial-final mass relation between main sequence stars and white dwarfs, e.g.~that of Wood (1992), to estimate the mass of the white dwarf if the progenitor was slightly more massive than a B9V star: $M_{\\rm WD} \\, = \\, A \\, exp \\, (B \\times M_{\\rm MS})$, where $A \\, = \\, 0.49M_\\odot$ and $B \\, = \\, 0.094M_\\odot^{-1}$. For $M_{\\rm MS} \\, = \\, 3.7M_\\odot$, we find $M_{\\rm WD} \\, = \\, 0.69M_\\odot$. This would suggest the surface gravity of the white dwarf log $g>8.0$ and, therefore, its surface temperature most likely lies between $\\approx29,000$K and $\\approx35,000$K. Finally, we note that if this white dwarf can be resolved from 16 Dra, then an optical spectrum may potentially be obtained (e.g.~with HST/STIS) from which its temperature and gravity can be tightly constrained. The mass can then be estimated, and this binary could be used to investigate the initial-final mass relation and to test the high mass end of the mass-radius relation. \\begin{figure*} \\vspace{8cm} \\special{psfile=h1904.f5 hscale=55 vscale=45 hoffset=50 voffset=15} \\caption{Left to right: EUV, UV and optical spectra of 16 Dra, shown together with a white dwarf model for $T_{\\rm eff}=31,900$K and log $g=8.5$. The B9.5V star clearly dominates the UV and optical flux, and the white dwarf is only detectable in the EUV. The UV spectrum was extracted from the IUE final archive (LWP18244). The optical spectrum was obtained on 1997 March 17 at the Russain Academy of Sciences' Special Astrophysical Observatory's 6m telescope, located in the Karachaevo-Cherkesia region of southern Russia. } \\end{figure*}" }, "0005/hep-ph0005263_arXiv.txt": { "abstract": "Relic neutrinos with mass in the range indicated by Super-Kamiokande results if neutrino masses are hierarchial (about 0.07 eV) are many times deemed too light to be cosmologically relevant. Here we remark that these neutrinos may significantly contribute to the dark matter of the Universe (with a large lepton asymmetry $L$) and that their existence might be revealed by the spectrum of ultra high energy cosmic rays (maybe even in the absence of a large $L$). ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005604_arXiv.txt": { "abstract": "We present details of the Behind The Plane survey of IRAS galaxies, which extends the PSCz survey to cover the 93\\% of the sky with complete and reliable IRAS data from the Point Source Catalog. At low latitudes, our catalogue is not complete to $0.6\\Jy$, but the incompleteness is physically understood and can be corrected for. IRAS galaxies at low latitudes are heavily or completely obscured optically, and are heavily outnumbered by Galactic sources with similar IRAS properties. We have used radio, optical, mm, near and far-infrared data to identify the galaxies, and the 2D catalogue is now complete. We have used optical and HI spectroscopy to obtain redshifts for the galaxies; the southern spectroscopy is completed and the north nearly so. ", "introduction": "\\vspace{-5pt} The 84\\% sky coverage of the PSCz survey is effectively limited by the need to get, for every galaxy, an optical identification from sky survey plates. The IRAS Point Source Catalog data itself is reliable to much lower latitudes, although genuine galaxies are outnumbered by Galactic sources with similar IRAS properties. Previous attempts to go further into the Plane have either been restricted to the Arecibo declination range, or have relied on optical identifications from Sky Survey Plates. Because the extinction may be several magnitudes, they have inevitably suffered from progressive and unquantifiable incompleteness as a function of latitude. In 1994 we embarked on a program, parallel with the PSCz survey, to systematically identify low latitude IRAS galaxies wherever the PSC data allowed, using new near-infrared observations where necessary. \\vspace{-5pt} ", "conclusions": "" }, "0005/astro-ph0005118_arXiv.txt": { "abstract": "Low mass X--ray transients hosting black hole candidates display on average a factor of $\\sim 100$ larger swing in the minimum (quiescent) to maximum (outburst) X--ray luminosity than neutron star systems, despite the fact that the swing in the mass inflow rate is likely in the same range. Advection dominated accretion flows, ADAFs, were proposed to interpret such a difference, because the advected energy disappears beyond the event horizon in black hole candidates, but must be radiated away in neutron star systems. The residual optical/UV emission of quiescent low mass X--ray transients, after subtraction of the companion star spectrum, was originally ascribed to optically thick emission from the outer accretion disk regions, where matter accumulates. Difficulties with this interpretation, led to a revised ADAF model where the bulk of the residual optical/UV emission in quiescence does not originate in the outermost disk regions but is instead produced by synchrotron radiation in the ADAF, and therefore is part of the ADAF's luminosity budget. We demonstrate that, once the residual optical/UV emission is taken into account, the bolometric luminosity swing of black hole candidates is consistent with that of neutron star systems. Therefore ascribing the bulk of the residual optical/UV flux to the ADAF removes much of the evidence on which ADAF models for low mass X--ray transients were originally developed, namely the higher luminosity swing in black holes than in neutron stars. We also find that, for the neutron star spin periods (a few ms) and magnetic fields ($\\sim 10^8-10^9$~G) inferred from some low mass X--ray transients, the mass to radiation conversion efficiency of recently proposed ADAF/propeller models is considerably higher than required to match the observations, once the contribution from accretion onto the magnetospheric boundary is taken into account. Motivated by these findings, we explore here an alternative scenario to ADAFs in which very little mass accretion onto the collapsed star (if at all) takes place in the quiescence intervals, whereas a sizeable fraction of the mass being transferred from the companion star (if not all) accumulates in an outer disk region. As in some pre-ADAF models, the residual optical/UV emission of black hole candidate systems are expected to derive from the gravitational energy released by the matter transferred from the companion star at radii comparable to the circularisation radius. The quiescent X--ray luminosity originates either from accretion onto the black hole candidates at very low rates and/or from coronal activity in the companion star or in the outer disk. For comparably small mass inflow rates, it can be concluded that the neutron stars in these systems are likely in the radio pulsar regime. In the interaction of the radio pulsar relativistic wind with matter transferred from the companion star, a shock forms, the power law-like emission of which powers both the harder X--ray emission component and most of the residual optical/UV observed in quiescence. The soft, thermal-like X--ray component may arise from the cooling of the neutron star surface in between outbursts or, perhaps, heating of the magnetic polar caps by relativistic particles in the radio pulsar magnetosphere. This scenario matches well both the X--ray and bolometric luminosity swing of black hole candidate as well as neutron star systems, for comparable swings of mass inflow rates toward the collapsed object. ", "introduction": "We estimate here the quiescent optical/UV luminosity of LMXRTs once the contribution from the mass donor star is subtracted (see also Menou, Narayan \\& Lasota 1999). We adopt the source sample and X--ray luminosities (0.5--10 keV range) of G98, complemented with recent results on the NS systems SAX J1808.4--3658 (Stella et al. 2000) and X~1732--304 (Guainazzi et al. 1999). Our results remain unchanged if the X--ray luminosities derived by C97) are used in place of those in G98 (see also Tab. 1 and Fig. 1). Being many orders of magnitude higher than the optical luminosity, the maximum X--ray luminosity provides a reliable estimate of the bolometric luminosity at the outburst peak (i.e. $L_{\\rm max}^X=L_{\\rm max}^{\\rm bol}=L_{\\rm max}$). Optical V magnitudes and absorptions are from C97, unless otherwise specified. \\subsection{Black hole candidates} The BHC transient A~0620--00 has a quiescent 0.5--10 keV luminosity of $L^X_{\\rm min}\\sim10^{31}\\ergs$ (for a distance of 1.2 kpc). This value is obtained by extrapolating the ROSAT data and using a fixed column density of $N_H=1.2\\times10^{21}\\cmdue$ (McClintock et al. 1995). Due to the small number of collected photons in the ROSAT observation ($\\sim 40$) the spectrum is very poorly determined and can be well fit by a variety of single component models. A short wavelength HST/FOS spectrum of the quiescent optical counterpart yielded a 1350--2200\\,\\AA\\ luminosity of $0.6-4\\times 10^{31}\\ergs$. These results have been confirmed by higher quality HST/STIS spectra (McClintock \\& Remillard 2000). At optical wavelengths (2200--4750\\,\\AA) the spectrum can be fit by a 9000 K black body (luminosity of $\\sim 10^{32}\\ergs$), after subtraction of the $58\\pm4\\%$ contribution from the K5V star companion, which affects mainly the spectrum at wavelengths $\\gsim 4000$\\,\\AA. At longer wavelengths (2.0--2.5\\,$\\mu$m) the K dwarf flux dominates, making up $75\\pm17\\%$ of the infrared luminosity (Shahbaz et al. 1999). Therefore, the quiescent optical/UV luminosity of A~0620--00 outshines the X--ray luminosity by a factor of $\\sim 10$ (see also McClintock et al. 1995). We estimate a rough bolometric quiescent luminosity of $L^{\\rm bol}_{\\rm min} \\sim 10^{32}\\ergs$. The lowest quiescent X--ray luminosity detected from GS 2023+338 (V~404 Cyg) is $L^X_{\\rm min}\\sim 2\\times 10^{33}\\ergs$ (Narayan, Barret \\& McClintock 1997; Campana 2000). The spectrum is well fit by either a power law (photon index $\\Gamma\\sim 1.5-2$) or a bremsstrahlung ($k\\,T_{\\rm br}\\sim 5-10$ keV). GS 2023+338 is significantly reddened ($A_{\\rm V} \\sim 4$ mag) and no UV data are available. Casares et al. (1993) estimate that the contribution of the accretion disk to the optical flux relative to the G9V--K0III companion is 72\\%, 36\\% and 19\\% in the B, V and R bands, respectively. These fractions convert to dereddened luminosities of $L_{\\rm B}\\sim 8\\times 10^{33}\\ergs$, $L_{\\rm V}\\sim 8\\times 10^{33}\\ergs$ and $L_{\\rm R}\\sim 6\\times 10^{33}\\ergs$. For GS 2023+338 we therefore adopt a value of $L^{\\rm bol}_{\\rm min} \\sim 2\\times 10^{34}\\ergs$, which is a factor of $\\sim 10$ higher than $L_{\\rm min}^X$. GRO J1655--40 has been detected in quiescence at a level of $L^X_{\\rm min} \\sim 2\\times 10^{32}\\ergs$ (Hameury et al. 1997). The spectrum can be described by a power law model ($\\Gamma\\sim 1.5$). Soft X--ray and UV flux measurement are severely hampered by a large absorption ($A_{\\rm V}\\sim 4$ mag). The F5IV companion star outshines the disk in the optical ($95\\pm2\\%$ at 5500\\,\\AA; Orosz \\& Bailyn 1997). An estimate of the V-band luminosity of the accretion disk based on the model by Orosz \\& Bailyn (1997) yields $L_{\\rm V} \\sim 4\\times10^{33}\\ergs$. The optical quiescent luminosity therefore dominates the X--ray emission by a factor of $\\sim 10$. For all other known BHC transients there are only upper limits to their quiescent X--ray flux (see Tab. 1). Yet, their optical counterparts are relatively well studied and the residual optical flux can be estimated. In particular, Keck spectra were used to estimate the fraction of the 6600--6800\\,\\AA\\ luminosity that originates from the companion star. This is 85\\%, 95\\%, 30\\% and 60\\% in GS 1124--684 (Nova Mus 91), GS 2000+25 (Nova Vul 88), H~1705--250 (Nova Oph 77) and GRO J0422+32 (Nova Per 92), respectively (Casares et al. 1997; Harlaftis et al. 1996, 1997, 1999). All these BHCs have K--M dwarf companions. V-magnitudes are ${\\rm V}=20.5,\\ 21.2,\\ 21.3$ and 22.2 mag, respectively. By assuming that the same fractions above hold for the V band, a very conservative assumption for K--M dwarf stars, we infer that the dereddened V luminosities of the accretion disk are $L_{\\rm V}=2\\times 10^{32},\\ 4\\times 10^{32},\\ 2\\times 10^{34}$ and $5\\times 10^{31}\\ergs$, respectively. The optical counterpart of 4U~1543--47, an A2V star, has V=16.7. Orosz et al. (1998) estimate a disk contribution of 10\\%, 21\\% , 32\\% and 39\\%, in the B, V, R and I bands, respectively. The increasing disk contribution for longer optical wavelengths is due to the relatively hot companion. The dereddened disk V and I luminosities are $L_{\\rm V}\\sim 3\\times 10^{34}\\ergs$ and $L_{\\rm I}\\sim 10^{34}\\ergs$, respectively. Consequently we estimate $L^{\\rm bol}_{\\rm min} \\sim 4\\times 10^{34}\\ergs$. Therefore, we conclude that the bolometric luminosity of quiescent BHC LMXRTs is dominated by optical/UV disk emission. When both the X--ray and optical quiescent luminosities are available, the latter are systematically higher by about an order of magnitude. \\subsection{Neutron stars} The two best studied NS LMXRTs, Aql X-1 and Cen X-4, have $L^X_{\\rm min} \\sim 6\\times 10^{32}\\ergs$ (Campana et al. 1998b) and $L^X_{\\rm min}\\sim 5\\times 10^{32}\\ergs$ (Asai et al. 1998; Campana et al. 2000), respectively. Recent optical studies of the field of Aql X-1 indicate that the true optical counterpart is located $0.5''$ from the previously known star (e.g. Shahbaz, Casares \\& Charles 1997). The magnitude of the counterpart is ${\\rm V}=21.6$ mag (Chevalier et al. 1999). The dereddened ($A_{\\rm V}=1.2$ mag) V luminosity can therefore be $L_{\\rm V}\\sim 10^{32}\\ergs$ at the most (i.e. if the entire V luminosity came from the disk). In any case this luminosity represents a small fraction of the quiescent X--ray luminosity. In the case of Cen X-4 the residual optical flux is estimated to contribute 80\\%, 30\\%, 25\\% and 10\\% in the B, V (18.7 mag), R and I bands, respectively (Shahbaz, Naylor \\& Charles 1993). The corresponding reddening-corrected luminosities ($A_{\\rm V}=0.3$ mag) are $L_{\\rm B}\\sim 9\\times10^{31}\\ergs$, $L_{\\rm V}\\sim 5\\times 10^{31}\\ergs$, $L_{\\rm R}\\sim 6\\times10^{31}\\ergs$ and $L_{\\rm I}\\sim 2\\times 10^{31}\\ergs$. Recently, an UV spectrum has been obtained with the HST/STIS (McClintock \\& Remillard 2000). The main result is that in a $\\nu\\,F_{\\nu}$ vs. $\\nu$ representation there is only factor of $\\sim 1-2$ increase from the X--rays to the optical. Therefore, the quiescent X--ray and optical luminosities are comparable in Cen X-4. The quiescent X--ray state of these two sources has been studied in some detail with BeppoSAX and ASCA pointed observations. Their 0.1--10 keV spectrum comprises a soft component, modeled by a black body with $k\\,T_{\\rm bb}\\sim 0.1-0.3$ keV and equivalent radius of $\\sim 1-3$ km, plus a hard power law component with photon index $\\sim 1-2$ (Campana et al. 1998b, 2000; Asai et al. 1996, 1998). The contribution of the two spectral components to the 0.5--10 keV luminosity is comparable. The quiescent X--ray flux of 4U~1608--522, 4U 2129+47 and EXO 0748--676 has also been detected (e.g. Campana et al. 1998a, see Tab. 1). 4U~1608--522 has been revealed in quiescence at a level of $L^X_{\\rm min} \\sim 2\\times 10^{33}\\ergs$ (0.5--10 keV for $d=3.3$~kpc; Asai et al. 1998). The highly absorbed ($A_{\\rm V}=5.2$ mag) optical counterpart of 4U~1608--522 has ${\\rm J}=18.0$ mag (${\\rm R}>22$ mag) and a luminosity of $\\sim 2\\times 10^{32}\\ergs$ at the most (including the companion star). 4U 2129+47 has $L^X_{\\rm min}\\sim 6\\times 10^{32}\\ergs$ (0.5--10 keV for $d=6.3$~kpc). The F9 subgiant companion dominates the optical flux (V=18.5 mag). Garcia \\& Callanan (1999) estimate ${\\rm V}=24.5$ mag for the disk of 4U 2129+47, implying a dereddened ($A_{\\rm V}\\sim 1.5$ mag) V luminosity of only $\\lsim 10^{30}\\ergs$. EXO 0748--676 has a relatively high quiescent luminosity of $L^X_{\\rm min}\\sim 10^{34}\\ergs$ (0.5--10 keV for $d=3.8$~kpc). However, being a high inclination system, EXO 0748--676 should be treated with caution since its X--ray flux variations might be driven by geometrical effects (e.g. obscuration by a variable height of the disk rim) rather than genuine mass inflow rate variations. Optical observations provided an upper limit on the V-magnitude of the quiescent optical counterpart of EXO 0748--676 (${\\rm V}>23$ mag for $A_{\\rm V}=1.2$ mag); this translates to a V luminosity of $\\sim 10^{30}\\ergs$. In addition to the sources in the G98 sample, SAX J1808.4--3658 (Stella et al. 2000) and X 1732--304 (Guainazzi et al. 1999) have also been detected in quiescence. SAX J1808.4--3658 has a quiescent X--ray luminosity of $2-3\\times 10^{32}\\ergs$. The optical counterpart was detected only during the outburst decay. An upper limit on the quiescent V magnitude of $>20.5$ mag has been derived (Giles et al. 1999). By using the galactic column density to estimate $A_{\\rm V}$, we derive an upper limit to the V luminosity of $4\\times 10^{32}\\ergs$. X 1732--304 in the globular cluster Terzan 1 was previously considered a persistent (though highly variable) source. In April 1999 it was observed in a quiescent state at a 2--10 keV luminosity of $1.4\\times 10^{33}\\ergs$. The quiescent X--ray spectrum was compatible with the two-component spectrum inferred for Aql~X-1 and Cen~X-4. The optical counterpart is not known. All the data above indicate that in quiescent NS LMXRTs the X--ray luminosity exceeds (or, at the most, is comparable to) the optical luminosity. \\subsection{Comparison of luminosity ratios and quiescent luminosities} The ratio of minimum to maximum luminosity of LMXRTs, as estimated by N97 and G98 on the basis of the X--ray data alone, is plotted in the left panel of Fig.~1, versus the maximum luminosity. BHCs are clearly separated from NSs both in terms of maximum luminosity and X--ray luminosity ratio. The middle panel shows instead $L^{\\rm bol}_{\\rm min}/L_{\\rm max}$ as estimated above based on both X--ray and optical measurements (see also Tab. 1): the distinction between BHC and NS systems is no longer apparent. Using a Kolmogorov-Smirnov, KS, test we estimate that the values of $L^{\\rm bol}_{\\rm min}/L_{\\rm max}$ for the two classes of transients have a 9\\% probability of being drawn by chance from the same parent distribution (note that using the X--ray data alone the KS probability is $<0.2\\%$). To check that this conclusion is robust and independent of the method for estimating $L_{\\rm max}$ adopted by G98, we calculated also the minimum to maximum luminosity ratios by using the $L_{\\rm max}$ values derived by C97 (see also Tab. 1). The results are shown in the right panel of Fig.~1. Also in this case, it is not possible to distinguish BHC from NS systems by using $L^{\\rm bol}_{\\rm min}/L_{\\rm max}$ (KS probability of 22\\%). We note that the conclusions above are even strengthened if EXO~0748-676, and/or SAX J1808.4--3658 and X 1732--304 (i.e. the sources not included in the G98 sample) are excluded from the source sample. Considering minimum bolometric luminosities only, one has that the BHC and NS transients population have a 56\\% probability of being drawn by chance from the same parent distribution. This probability is 22\\% if minimum bolometric luminosities in Eddington units are used instead. This is at variance with the results obtained by M99, who used minimum X--ray luminosities. These results show that once the contribution from the optical luminosity (after subtraction of the mass donor's spectrum) is included in the evaluation of the quiescent luminosity of LMXRTs, there is no evidence that the luminosity swing of BHCs is larger than that of NSs, neither that the minimum (quiescent) luminosity of BHCs is lower. In fact the optical luminosity, while usually negligible in NS LMXRTs, dominates the quiescent luminosity of BHC transients. ", "conclusions": "The scenario we explored in this paper for the quiescent emission of LMXRTs assumes only very little matter (if at all) proceeds toward the collapsed object while most of the mass transferred from the companion star accumulates in an outer disk region or is lost in a wind. This idea is in line with models designed to explain the delays between the optical and UV light curves in the outbursts of cataclysmic variables. A suppression of the innermost disk regions (or nearly so) during quiescence is envisaged in those models as well (e.g. Livio \\& Pringle 1992; Meyer \\& Meyer-Hofmeister 1994; King 1997). Our suggested scenario is also reminiscent of earlier (pre-ADAF) disk instability models that were proposed to explain the long recurrence time of A 0620--00 within the context of standard accretion theory (Huang \\& Wheeler 1989; Mineshige \\& Wheeler 1989). At present we can only speculate on the reason why the quiescent outer disk would remain stable despite its effective temperature in $\\sim 5000-10000$~K range. Perhaps the {\\it ad hoc} variation of the $\\alpha$ viscosity parameter that is required to make current disk instability models reproduce the outburst recurrence times (see Lasota \\& Hameury 1998) does not take place at the temperature at which the hydrogen ionisation and opacity change (as in current models), but rather at somewhat higher temperatures. In this case the outer disk region of quiescent LMXRTs could be in the lower branch of the corresponding surface density vs. viscosity relationship. In our model the residual quiescent optical/UV emission (after subtraction of the contribution from the mass donor star) in BHC systems derives entirely from the gravitational energy released by the matter transferred from the companion star at radii comparable to the circularisation radius. The low quiescent X--ray luminosity originates either from standard accretion into the BHC at very low rates (some $\\sim 10^{11}-10^{12}\\gs$), and/or from coronal activity of the outer disk where matter accumulates or, limited to short orbital period systems, the companion star which is forced to corotate (see also Bildsten \\& Rutledge 2000). For comparably small mass inflow rates ``leaking\" from the outer disk regions, it can be safely concluded that the NSs of LMXRTs are in the radio pulsar regime, if their spin periods are a few millisecond and magnetic field some $\\sim 10^8-10^9$~G, as indicated by the observations. The interaction of the radio pulsar relativistic wind with the matter transferred from the companion star gives rise to a shock, the power law like emission of which powers the harder X--ray emission component and optical/UV excess which are observed at a level of $\\sim 10^{32}-10^{33}\\ergs$ in the quiescent state of Cen~X-4. The soft, thermal-like component which contributes about half of the quiescent X--ray luminosity of several NS LMXRTs arises from the cooling of the NS surface in between outbursts or, perhaps, heating of the magnetic polar caps by relativistic particles in the radio pulsar magnetosphere. Both the X--ray and bolometric luminosity swing of of BHC as well as NS systems are well matched by the model, for comparable ratios of minimum to maximum mass inflow rates toward the collapsed object across the two classes of LMXRTs. Moreover, different predictions are made about the quiescent emission of LMXRTs, which could be tested through higher spectral resolution and throughput observations to be obtained in the near future (e.g. with Newton-XMM). For example, if the quiescent X--ray emission of BHC systems resulted from coronal activity, emission lines from heavy elements and an optically thin thermal spectrum would be expected. Moreover residual optical/UV and X--ray flux variations should be correlated. On the contrary if a hot accretion disk with standard (as opposed to ADAF) efficiency gives rise to the X--ray flux, uncorrelated X--ray and residual optical/UV variations might take place. Variations in the optical/UV excess and the X--ray power law component of quiescent NS systems should be correlated, if they both arose from radio pulsar shock emission. The detection of pulsations at the NSs spin in the quiescent soft X--ray component would rule out emission from the whole NS surface and argue in favor of heated magnetic polar caps. The ultimate test of the radio pulsar regime in quiescent NS LMXRTs would be the detection of a pulsed radio signal. Yet, the matter in the outer disk and/or the shock might enshroud the pulsar making any radio signal very difficult to detect (Kochanek 1993; Stella et al. 1994; Campana et al. 1998a). The geometry of the radio pulsar shock and matter accumulating during the quiescent intervals of NS LMXRTs is highly uncertain; detailed Balmer line Doppler mapping (possibly at different times after the end of an outburst) could provide important clues on this issue." }, "0005/astro-ph0005432_arXiv.txt": { "abstract": "We calculate the spectrum of density fluctuations in models of inflation based on a weakly self-coupled scalar matter field minimally coupled to gravity, and specifically investigate the dependence of the predictions on modifications of the physics on length scales smaller than the Planck length. These modifications are encoded in terms of modified dispersion relations. Whereas for some classes of dispersion relations the predictions are unchanged compared to the usual ones which are based on a linear dispersion relation, for other classes important differences are obtained, involving tilted spectra, spectra with exponential factors and with oscillations. This is the case when the dispersion relation becomes complex. We conclude that the predictions of inflationary cosmology in these models are not robust against changes in the super-Planck-scale physics. ", "introduction": "Most current models of inflation \\cite{Guth} are based on weakly self-coupled scalar matter fields minimally coupled to gravity. In most of these models, the period of inflation lasts for a number of e-foldings much larger than the number needed to solve the problems of standard cosmology \\cite{Linde}. In these cases, the physical length of perturbations of cosmological interest today (those which today correspond to the observed CMB anisotropies and to the large-scale structure) was much smaller than the Planck length at the beginning of inflation. Hence, the approximations which go into the calculation of the spectrum of cosmological perturbations \\cite{flucts} break down. It is then of interest to investigate whether the predictions are sensitive to the unknown super-Planck-scale physics, or whether the resulting spectrum of perturbations is determined only by infrared physics. An analogous problem arises for black hole evaporation. The original computations of the thermal spectrum from black holes \\cite{Hawking} appear to involve mode matching at super-Planck scales. However, in the case of black holes it can be shown \\cite{Unruh,Brout,Hambli,Corley} that the predictions are in fact insensitive to modifications of the physics at the ultraviolet end. Our goal is to explore whether and in which cases the spectrum of fluctuations resulting from inflationary cosmology depends on the unknown ultraviolet physics. We will adapt the method of \\cite{Unruh,Corley} and consider theories obtained by replacing the linear dispersion relation for the linearized fluctuation equations by classes of nonlinear dispersion relations. We find that for the class of dispersion relations introduced by Unruh \\cite{Unruh} one recovers a scale-invariant spectrum of fluctuations in the case of exponential inflation. In contrast, for the class of dispersion relations modelled after the one introduced in \\cite{Corley}, the resulting spectrum may be tilted and may include exponential and oscillatory factors if the dispersion relation becomes complex. Such spectra are inconsistent with observations. We thus conclude that the predictions for observables in weakly coupled scalar field models of inflation depend sensitively on hidden assumptions about super-Planck-scale physics. ", "conclusions": "We have studied the robustness of the predictions for the spectrum of cosmological perturbations of weakly coupled inflationary models. The method used was to replace the usual linear dispersion relation by special classes of nonlinear ones, where the nonlinearity is confined to physical wavelengths $\\lambda$ smaller than some critical length $l_{\\rm C}$. We found that for the class of dispersion relations first introduced by Unruh \\cite{Unruh}, the predictions are unchanged. This is connected with the fact that the initial vacuum state evolves adiabatically up to the time $\\eta_1$ when $\\lambda = l_{\\rm C}$ \\footnote{We thank Bill Unruh for pointing this connection out to us.}. However, in the case of the dispersion relation modelled after the one used by Corley and Jacobson \\cite{Corley} in the situation where it becomes complex, the resulting spectrum can have oscillations, non-standard tilts and exponential factors which render the resulting theory in conflict with observations. The specific predictions depend on the sign of $b_m$, on the value of $m$, and on the initial state chosen. The results are summarized in Table 1. We thus conclude that the predictions in weakly coupled scalar field-driven inflationary models are not robust to changes in the unknown fundamental physics on sub-Planck lengths. This opens up another potentially very interesting link between fundamental physics and observations. Note, however, that in strongly coupled scalar field models of inflation such as the model discussed in \\cite{BZ}, the spectrum of fluctuations is robust to changes in the underlying sub-Planck-length physics. \\vspace{0.5cm} \\centerline{\\bf Acknowledgements} \\vspace{0.2cm} We are grateful to Lev Kofman, Dominik Schwarz, Carsten Van de Bruck and in particular Bill Unruh for stimulating discussions and useful comments. We acknowledge support from the BROWN-CNRS University Accord which made possible the visit of J.~M. to Brown during which most of the work on this project was done, and we are grateful to Herb Fried for his efforts to secure this Accord. One of us (R.~B.) wishes to thank Bill Unruh for hospitality at the University of British Columbia during the time when this work was completed. J.~M. thanks the High Energy Group of Brown University for warm hospitality. The research was supported in part by the U.S. Department of Energy under Contract DE-FG02-91ER40688, TASK A." }, "0005/astro-ph0005268_arXiv.txt": { "abstract": "Two giants of the Sagittarius dwarf spheroidal have been observed with the UVES spectrograph on the ESO 8.2m Kueyen telescope, during the commissioning of the instrument. Sgr 139 has [Fe/H]$=-0.28$ and Sgr 143 [Fe/H]$=-0.21$, these values are considerably higher than photometric estimates of the metallicity of the main population of Sgr. We derived abundances for O, Na, Mg, Al, Si, Ca, Sc, Ti, V, Cr, Mn, Co, Ni, Cu, Y, Ba, La, Ce, Nd and Eu; the abundance ratios found are essentially solar with a few exceptions: Na shows a strong overdeficiency, the heavy elements Ba to Eu, are overabundant, while Y is underabundat. The high metallicity derived implies that the Sgr galaxy has experienced a high level of chemical processing. The stars had been selected to be representative of the two main stellar populations of Sagittarius, however, contrary to what expected from the photometry, the two stars show a very similar chemical composition. We argue that the most likely explanation for the difference in the photometry of the two stars is a different distance, Sgr 143 being about 2Kpc nearer than Sgr 139. This result suggests that the interpretation of colour -- magnitude diagrams of Sgr is more complex than previously thought and the effect of the line of sight depth should not be neglected. It also shows that spectroscopic abundances are required for a correct interpretation of Sgr populations. ", "introduction": "In the recent years our ideas on galaxy formation and evolution have considerably developed, and it is generally acknowledged that it is a complex process which may well take different paths in different galaxies. Much attention is being devoted to dwarf spheroidal galaxies, essentially for two reasons: 1) they seem to be relatively simple systems, typically characterized by a single stellar population; in such a system we hope to be able to isolate some of the key ingredients of the phenomenon; 2) Interaction of these dwarf galaxies with large galaxies (such as our own or the Andromeda galaxy) could, in principle, play an important role in shaping the morphology of the large galaxies. The nearest members of the class, the dwarf spheroidals of the Local Group, are close enough that their stars are amenable to detailed analysis with the same techniques employed to study Galactic stars, with the advent of the new 8m class telescopes. In this paper we report on such an observation: the first detailed chemical analysis of two stars in the Sgr dwarf spheroidal based on high resolution spectra obtained with the UVES spectrograph on the ESO 8.2m Kueyen telescope. Ever since the discovery of Sgr (Ibata et al 1994) photometric studies have shown the red giant branch (RGB) of Sgr to be wider than expected for a population with a single age and metallicity. This has been generally interpreted as evidence that Sgr displays a spread in metallicity which is likely due to different bursts of star formation. Ibata et al (1995) found a mean metallicity of [Fe/H]=-1.25 and their metallicity distribution displays a spread of over 1 dex. Sarajedini \\& Layden (1995) found a main population with [Fe/H]= -0.52 and suggested the possible existence of a population of [Fe/H]$\\approx -1.3$. Mateo et al (1995) provide a mean metallicity of $-1.1 \\pm 0.3$, Ibata et al (1997) estimate metallicities in the range $-1.0 -0.8 $, Marconi et al (1998) $ -1.58 -0.7 $ , Bellazzini et al (1999) $-2.1, -0.7$. The age of Sgr may not be disentangled from its metallicity, from Main Sequence fitting, Fahlman et al (1996) found acceptable solutions for an age of 10 Gyr and metallicity $-0.8$ or an age of 14 Gyr and a metallicity $-1.3$. Clearly if Sgr may not be described as a single population the concepts of age and metallicity loose some of their significance; Bellazzini et al (1999) proposed an extreme scenario in which star formation began rather early and continued for a period longer than 4 Gyr. Foreseeing the potentiality of UVES to perform detailed abundance analysis of these stars, to confirm or refute the photometrically inferred spread in metallicity, we undertook already in 1995 observations of low resolution spectra of photometrically identified (Marconi et al 1998) Sgr candidates. The main purpose was to obtain confirmed radial velocity members of Sgr for subsequent high resolution follow-up with UVES. From the low resolution spectra we also devised a method to obtain crude metallicity estimates from spectral indices defined in the Mg I b triplet region. The two stars were selected from this low-resolution study of Sgr with photometry and abundance estimates which suggested these stars to differ by at least 0.5 dex in metallicity. \\begin{table}[b] \\caption{Basic Data} \\begin{center} \\renewcommand{\\tabcolsep}{0.05cm} \\begin{tabular}{lrrrrrrr} \\hline \\# & $\\alpha_{2000}$ & $\\delta_{2000} $ & $V$ \\hfill& $(V-I)_0$ &$T_{eff}$ & log g & $\\xi$\\\\ & & & & & K & & kms$^{-1}$\\\\ \\hline 139 & 18 53 50 & $-30$ 30 45 & 18.33 & 0.965 &4902&2.5 &1.4\\\\ 143 & 18 53 49 & $-30$ 31 60 & 18.15 & 0.947 &4932&2.5 & 1.5\\\\ \\hline \\end{tabular} \\end{center} \\end{table} \\begin{table}[t] \\caption{Mean abundances} \\begin{center} \\begin{tabular}{lrrrrrr} \\hline & [X/H] & $\\sigma$ & n &[X/H] & $\\sigma$ & n \\\\ & 139 & & & 143 & & \\\\ \\hline O I & $<-0.53 $ & -- & 1 & $-0.37$ & & 1 \\\\ Na I & $ -0.61 $ & 0.12 & 2 & $ -0.78 $ & 0.17 & 3 \\\\ Mg I & $ -0.33 $ & 0.11 & 2 & $ -0.44 $ & 0.12 & 3 \\\\ Al I & $ -0.52 $ & 0.08 & 3 & $ -0.40 $ & 0.26 & 3 \\\\ Si I & $ -0.35 $ & 0.26 & 5 & $ -0.28 $ & 0.09 & 5 \\\\ Ca I & $ -0.49 $ & 0.12 & 5 & $ -0.47 $ & 0.21 & 4 \\\\ Sc II & $ -0.67 $ & 0.05 & 2 & $ -0.67 $ & 0.20 & 2 \\\\ Ti I & $ -0.19 $ & 0.13 & 7 & $ -0.37 $ & 0.14 & 7 \\\\ Ti II & $ -0.31 $ & 0.07 & 2 & $ -0.46 $ & 0.06 & 2 \\\\ V I & $ -0.21 $ & 0.16 & 2 & $ -0.44 $ & 0.05 & 2 \\\\ Cr II & $ -0.32 $ & & 1 & $ -0.35 $ & & 1 \\\\ Mn I & $ -0.41 $ & & 1 & $ -0.30 $ & & 1 \\\\ Fe I & $ -0.28 $ & 0.16 & 15& $ -0.21 $ & 0.19 & 15 \\\\ Fe II & $ -0.33 $ & 0.14 & 3 & $ -0.23 $ & 0.04 & 4 \\\\ Co I & $ -0.34 $ & 0.21 & 2 & -- & -- & -- \\\\ Ni I & $ -0.56 $ & 0.16 & 6 & $ -0.44 $ & 0.21 & 6 \\\\ Cu I & $ -0.45 $ & & 1 & $-0.22 $ & & 1 \\\\ Y II & $ -0.61 $ & 0.08 & 3 & $-0.66 $ & 0.13 & 4 \\\\ Ba II & $ -0.08 $ & & 1 & $-0.10 $ & & 1 \\\\ La II & $ +0.30 $ & 0.30 & 3 & $+0.33 $ & 0.23 & 3 \\\\ Ce II & $ +0.15 $ & 0.06 & 3 & $+0.00 $ & 0.13 & 3 \\\\ Nd II & $ +0.12 $ & 0.27 & 5 & $+0.09 $ & 0.19 & 8 \\\\ Eu II & $ +0.05 $ & & 1 & $+0.05 $ & & 1 \\\\ \\hline \\end{tabular} \\end{center} \\end{table} \\begin{table}[] \\caption{Line data and abundances for Fe} \\begin{center} \\renewcommand{\\tabcolsep}{0.1cm} \\begin{tabular}{lrrrrrr} \\hline Ion & $\\lambda $ & log gf & EW(pm) & $\\epsilon$ & EW(pm) & $\\epsilon$ \\\\ & nm & & 139& & 143 & \\\\ \\hline Fe I & 585.5091 & -1.76 & 4.12 & 7.56 & 4.24 & 7.61 \\\\ Fe I & 585.6083 & -1.64 & 3.85 & 7.04 & 6.24 & 7.51 \\\\ Fe I & 585.8779 & -2.26 & 1.97 & 7.14 & 2.37 & 7.27 \\\\ Fe I & 586.1107 & -2.45 & 1.81 & 7.35 & 1.22 & 7.18 \\\\ Fe I & 506.7151 & -0.97 & 7.46 & 7.06 & 8.34 & 7.24 \\\\ Fe I & 510.4436 & -1.69 & 5.89 & 7.27 & 6.58 & 7.65 \\\\ Fe I & 510.9650 & -0.98 & -- & -- & 7.90 & 7.24 \\\\ Fe I & 489.2871 & -1.29 & 7.09 & 7.32 & 5.91 & 7.06 \\\\ Fe I & 552.5539 & -1.33 & 6.25 & 7.13 & - & -- \\\\ Fe I & 587.7794 & -2.23 & 2.74 & 7.25 & 3.34 & 7.25 \\\\ Fe I & 588.3813 & -1.36 & 7.88 & 7.18 & 9.15 & 7.43 \\\\ Fe I & 615.1617 & -3.30 & 9.59 & 7.28 & 8.02 & 6.98 \\\\ Fe I & 616.5361 & -1.55 & 5.98 & 7.17 & 6.46 & 7.26 \\\\ Fe I & 618.7987 & -1.72 & 7.36 & 7.37 & 6.34 & 7.17 \\\\ Fe I & 649.6469 & -0.57 & 6.26 & 6.97 & 8.45 & 7.38 \\\\ Fe I & 670.3568 & -3.16 & 7.72 & 7.42 & 7.16 & 7.33 \\\\ Fe II & 483.3197 & -4.78 & 2.72 & 7.34 & 1.94 & 7.27 \\\\ Fe II & 499.3358 & -3.65 & 5.97 & 7.10 & 6.37 & 7.30 \\\\ Fe II & 513.2669 & -4.18 & -- & -- & 3.98 & 7.33 \\\\ Fe II & 526.4812 & -3.19 & 6.00 & 7.11 & 6.09 & 7.24 \\\\ \\hline \\end{tabular} \\end{center} \\end{table} ", "conclusions": "The metallicity of the two stars examined here is higher than all previous photometric estimates. Although it is possible that we happened to select two members of the high--metallicity tail of Sgr, this position is hardly tenable, the event of finding two such stars in a 9 square arcmin field must be quite rare. It is more likely that Sgr actually possesses a population, perhaps the main population, this metal--rich. For our two stars the Schlegel et al (1998) maps provide $E(B-V)= 0.14$. By comparison Marconi et al (1998) used E(B-V)=0.18. The fact that the actual reddening could be 0.04 less than this could explain why the metallicity we find is 0.3 dex higher than the highest metallicity estimated by Marconi et al. \\begin{table}[] \\caption{Line data and abundances} \\begin{center} \\renewcommand{\\tabcolsep}{0.1cm} \\begin{tabular}{lrrrrrr} \\hline Ion & $\\lambda $ & log gf & EW(pm) & $\\epsilon$ & EW(pm) & $\\epsilon$ \\\\ & nm & & 139& & 143 & \\\\ \\hline O I & 630.0304 & -9.82 & $<1.5 $ & $<8.33$ & 2.08 & 8.49 \\\\ Na I & 498.2814 & -0.95 & -- & -- & syn & 5.44 \\\\ Na I & 615.4226 & -1.56 & 3.73 & 5.80 & 3.32 & 5.74 \\\\ Na I & 616.0747 & -1.26 & 4.48 & 5.63 & 3.42 & 5.46 \\\\ Mg I & 571.1088 & -1.83 & 11.48 & 7.32 & 10.16 & 7.12 \\\\ Mg I & 631.8717 & -1.98 & 4.00 & 7.17 & 3.19 & 7.03 \\\\ Mg I & 631.9237 & -2.20 & -- & -- & 3.24 & 7.26 \\\\ Al I & 669.8673 & -1.65 & 2.88 & 5.95 & syn & 6.27 \\\\ Al I & 669.6788 & -1.42 & syn & 6.04 & syn & 6.17 \\\\ Al I & 669.6023 & -1.35 & syn & 5.87 & syn & 5.77 \\\\ Si I & 577.2146 & -1.75 & 6.67 & 7.48 & 4.82 & 7.16 \\\\ Si I & 594.8541 & -1.23 & 7.12 & 7.00 & 8.84 & 7.28 \\\\ Si I & 614.2483 & -1.48 & 2.36 & 7.05 & 4.34 & 7.40 \\\\ Si I & 614.5016 & -1.43 & 4.99 & 7.47 & 3.79 & 7.24 \\\\ Si I & 615.5134 & -0.77 & 5.80 & 6.95 & 7.87 & 7.27 \\\\ Ca I & 615.6023 & -2.20 & 1.81 & 5.77 & -- & -- \\\\ Ca I & 616.1297 & -1.02 & 7.99 & 5.89 & 9.20 & 6.00 \\\\ Ca I & 616.6439 & -0.90 & -- & -- & 8.01 & 5.66 \\\\ Ca I & 645.5598 & -1.35 & 6.37 & 5.81 & 7.85 & 6.07 \\\\ Ca I & 649.9650 & -0.59 & 9.95 & 5.70 & 9.95 & 5.69 \\\\ Ca I & 650.8850 & -2.11 & 3.28 & 6.01 & --- & -- \\\\ Sc II & 552.6790 & 0.13 & 9.94 & 2.53 & 10.91 & 2.65 \\\\ Sc II & 632.0851 & -1.77 & 1.92 & 2.46 & 1.62 & 2.36 \\\\ Ti I & 488.5082 & 0.36 & 9.92 & 4.88 & 8.97 & 4.63 \\\\ Ti I & 497.7719 & -0.92 & syn & 4.66 & syn & 4.80 \\\\ Ti I & 497.8222 & -0.39 & syn & 4.76 & syn & 4.60 \\\\ Ti I & 498.9131 & -0.22 & syn & 4.76 & syn & 4.44 \\\\ Ti I & 499.7098 & -2.12 & syn & 5.05 & syn & 4.70 \\\\ Ti I & 586.6452 & -0.84 & 10.14 & 4.91 & 8.41 & 4.57 \\\\ Ti I & 612.6217 & -1.42 & 6.40 & 4.76 & 6.72 & 4.83 \\\\ Ti II & 498.1355 & -3.16 & syn & 4.76 & syn & 4.60 \\\\ Ti II & 499.6367 & -2.92 & syn & 4.66 & syn & 4.52 \\\\ V II & 573.7059 & -0.74 & -- & -- & 3.28 & 3.57 \\\\ V II & 613.5361 & -0.75 & syn & 3.68 & syn & 3.50 \\\\ V II & 615.0157 & -1.78 & syn & 3.91 & syn & 3.60 \\\\ Cr II & 488.4607 & -2.08 & 3.76 & 5.35 & 3.76 & 5.32 \\\\ Mn I & 511.7934 & -1.14 & 3.72 & 4.98 & 4.24 & 5.09 \\\\ Co I & 553.0774 & -2.06 & 5.76 & 4.73 & -- & -- \\\\ Co I & 533.1452 & -1.96 & 4.10 & 4.43 & -- & \\\\ Ni I & 585.7746 & -0.64 & 2.74 & 5.41 & 4.38 & 5.75 \\\\ Ni I & 612.8963 & -3.33 & 5.60 & 5.75 & 5.86 & 5.81 \\\\ Ni I & 613.0130 & -0.96 & 1.89 & 5.61 & 1.88 & 5.62 \\\\ Ni I & 617.5360 & -0.53 & 5.74 & 5.81 & 4.93 & 5.64 \\\\ Ni I & 617.6807 & -0.53 & 5.73 & 5.81 & 7.86 & 6.21 \\\\ Ni I & 617.7236 & -3.50 & 3.69 & 5.75 & 3.95 & 5.82 \\\\ Cu I & 510.5537 & -1.52 & 13.10 & 3.99 & 11.80 & 3.76 \\\\ Y II & 488.3684 & 0.07 & 5.84 & 0.95 & 9.37 & 1.67 \\\\ Y II & 498.2129 & -1.29 & 3.35 & 1.71 & 2.78 & 1.57 \\\\ Y II & 508.7416 & -0.17 & 7.58 & 1.55 & 7.13 & 1.40 \\\\ Y II & 511.9112 & -1.36 & 2.91 & 1.63 & 3.15 & 1.67 \\\\ Ba II & 649.6897 & -0.38 & 18.80 & 2.06 & 18.92 & 2.03 \\\\ La II & 480.4039 & -1.50 & 3.76 & 1.40 & 3.53 & 1.33 \\\\ La II & 511.4559 & -1.06 & 7.74 & 1.81 & 7.78 & 1.76 \\\\ La II & 632.0376 & -1.61 & 3.20 & 1.21 & 4.31 & 1.42 \\\\ Ce II & 518.7458 & 0.13 & 4.30 & 1.74 & 3.00 & 1.44 \\\\ Ce II & 533.0556 & -0.36 & 3.53 & 1.66 & 3.26 & 1.59 \\\\ Ce II & 546.8371 & 0.14 & 3.60 & 1.78 & -- & -- \\\\ Ce II & 604.3373 & -0.43 & -- & -- & 2.04 & 1.71 \\\\ Nd II & 491.4382 & -1.00 & 5.89 & 2.01 & 3.82 & 1.53 \\\\ Nd II & 495.9119 & -0.98 & 11.66 & $^n$2.87 & 6.73 & 1.75 \\\\ Nd II & 496.1387 & -0.71 & 2.80 & 1.32 & 3.01 & 1.36 \\\\ Nd II & 498.9950 & -0.50 & -- & -- & syn & 1.79 \\\\ Nd II & 499.8541 & -1.10 & -- & -- & syn & 1.79 \\\\ Nd II & 508.9832 & -1.16 & 3.79 & 1.49 & 2.94 & 1.30 \\\\ Nd II & 529.3163 & -0.10 & 5.70 & 1.54 & 6.48 & 1.67 \\\\ Nd II & 543.1516 & -0.57 & -- & -- & 2.84 & 1.71 \\\\ Nd II & 548.5696 & -0.30 & 3.53 & 1.77 & 2.32 & 1.48 \\\\ \\hline \\end{tabular} \\noindent{ $^n$ not used to compute the mean abundance} \\end{center} \\end{table} \\begin{table}[] \\caption{HFS data for Eu II} \\begin{center} \\begin{tabular}{lrr} \\hline & $\\lambda$ (nm) & log gf \\\\ \\hline Eu II & 664.516 &$ -0.800$\\\\ Eu II & 664.513 &$ -0.886$ \\\\ Eu II & 664.511 &$ -0.521$ \\\\ Eu II & 664.510 &$ -0.856$ \\\\ Eu II & 664.509 &$ -0.506$ \\\\ Eu II & 664.508 &$ -0.771$ \\\\ Eu II & 664.507 &$ -0.427$ \\\\ \\hline \\end{tabular} \\end{center} \\end{table} Quite obviously our results do not rule out the existence of a more metal--poor population. Preliminary results of abundance analysis in Sgr are given by Smecker-Hane \\& McWilliam (1999), who find two metal--poor Sgr member stars, with [Fe/H]$=-1.41$ and [Fe/H]$=-1.14$. It is interesting that out of 11 stars analyzed by them 7 have metallicities in the range $-0.6 < \\rm [Fe/H] < -0.2$, two are metal-poor and two are metal rich ([Fe/H]$\\sim 0.$). The 7 stars of intermediate metallicity, which should be analogous to the two under study here, show solar abundance ratios and no enhancement of $\\alpha$ elements, in agreement with our findings. Also the Na abundance displays a similar pattern: for all their stars, except the two metal--poor ones Na is over-deficient with respect to iron by 0.3 -- 0.5 dex. Unfortunately, these results have not been published in a more detailed form and we lack information on the temperatures and luminosities of the the stars considered by Smecker-Hane \\& McWilliam so we do not know if we are comparing similar giants. It is also interesting to compare the present results with the abundances of the two Sgr planetary nebulae He 2-436 and Wray 16-423, studied by Walsh et al (1998). The only element in common in the two analysis is O, for which Walsh et al find [O/H]$= -0.64 \\pm 0.08$ and [O/H]$=-0.62 \\pm 0.07$ for He 2-436 and Wray 16-423, respectively. Our result for Sgr 143 is about 0.2 dex higher, but it is also more uncertain, because it is based on a single weak line, which is also very sensitive to gravity. An increase of gravity of 0.5 dex results in an increase of O abundance of 0.25 dex. O should be only marginally affected during AGB evolution, so that the O abundance in the PN ought to be quite close to that in the progenitor star. Walsh et al (1998) argued that their abundances suggested a mild enhancement of O over Fe, because they assumed $-0.8$ to be the mean [Fe/H] of Sgr. Another scenario appears more likely, in view of our results: a solar O/Fe ratio, which suggests that the PNe have [Fe/H]$\\sim -0.5$. Having established that the two stars are quite similar in atmospheric parameters and abundances we must explain why their photometry is different and why the metallicity estimated from the low resolution spectra for star 143 is far lower than the one derived here. We consider 5 possibilities: 1) errors in $V$; 2) errors in $V-I$; 3) different reddening; 4) different age; 5) different distance. Let us examine all of these cases. That a difference of 0.18 mag in $V$ may be due to the photometric error may be discarded since this is a factor of ten larger than the photometric error of Marconi et al (1998). An error in $V-I$ is more likely; a 0.03 -- 0.04 mag error in $V-I$ would allow to slide sideways one of the two stars in the colour-magnitude diagram in such a way that both stars lie on the same isochrone, since the RGB, in this range of $V-I$ is very steep. The implied difference in T$_{\\rm eff}$ is of $\\sim 100$ K, the errors of our analysis. Differential reddening seems unlikely for three reasons. The dust maps of Schlegel et al (1998) give $E(B-V)=0.14$ for both stars, suggesting that the reddening of the two stars is the same within 0.01 mag. The absence of detectable amounts of HI in Sgr (Burton \\& Lockman, 2000) also argues against a differential reddening. If the 0.18 mag difference in $V$ were due to reddening it would imply a difference of almost 0.08 mag in $V-I$, i.e. $\\sim 200$ K in T$\\rm _{eff}$. Although such a difference is within the errors of the present analysis and cannot be ruled out, it does seem somewhat unlikely, given the similarity of the two spectra. An age difference of $\\sim 1 $ Gyr could be enough to explain the difference in the photometry of the two stars. A larger age spread would be necessary to explain the width of the RGB, like in the scenario proposed by Bellazzini et al (1999). Although such a possibility is attractive, it appears somewhat contrived and it is not so clear that star formation may continue for several Gyrs without resulting in a spread in metallicity, as well as ages. A distance difference of about 2Kpc would be enough to explain the difference in $V$. This value is not unreasonable, Ibata et al (1997), estimate the half-brightness depth of Sgr to be about 1.2 kpc . It is interesting that recent N--body simulations by Helmi \\& White (2000) support a considerable depth of Sgr: inspection of their figure 2 shows that the bulk of their model for Sgr has a depth of about 2 Kpc, however considering the debris shed during previous orbits, one has a sizeable population over a depth of 10 Kpc. Further support to the possibility that the two stars have a different distance comes from inspection of the Na I D lines (Fig. 2), three interstellar components belonging to our Galaxy are evident in both the spectra of Sgr 143 and of Sgr 139 at radial velocity +16.5 kms$^{-1}$, $+28.0$ kms$^{-1}$ and $+47.3$ kms$^{-1}$; while the Na I D lines of star 143 appear symmetric and there is no hint of an interstellar component at the radial velocity of Sgr, the lines of star 139 show a weak but definite asymmetry, which we interpret as a weak interstellar line associated with Sgr. Star 139 is in fact the fainter of the two and hence the most distant, according to this interpretation, this would explain why the interstellar Na I D lines appear in its spectrum but not in the spectrum of star 143, which would then be in the side of Sgr nearer to us. So of the five possibilities considered only the photometric error in $V$ and the differential reddening are discarded. We may not decide which is the correct one with the present data, new accurate photometric measurements will allow to settle at least the issue of errors in $V-I$. However we consider that the distance difference is the most likely explanation, because it is the simplest and is supported by several arguments. This suggests that the non-negligible line of sight depth of Sgr could explain at least a part of the width of the RGB of Sgr. Up to now all investigators have adopted a unique distance modulus for Sgr, in order to compare their photometry to fiducial ridge lines of Galactic clusters or to theoretical isochrones. This assumption may prove to be bit too naive. A full discussion of the metallicity estimates from low resolution spectra shall be given elsewhere. Suffice to say here that the method of estimating abundances from low resolution needs a relatively high S/N ratio. In the case of star 139 the metallicity derived from high resolution analysis coincides with that estimated from low resolution to within the errors of the latter. We verified that the degraded UVES spectrum is very similar to the low resolution EMMI spectrum. The indices measured on this degraded spectrum yield in fact almost the same abundance provided by those measured on the low resolution spectrum. In the case of star 143 instead the method has been applied to a spectrum of too low signal to noise ratio, in this case the degraded UVES spectrum bears almost no resemblance to the low resolution spectrum, except for the strongest feature of the Mg I b triplet, which was enough to provide the correct radial velocity for this star. The ratios of all elements are essentially solar, noticeable exceptions are: Na which is overdeficient with respect to iron, and the heaviest elements Ba, La, Ce, Nd, Eu, which appear over-abundant while Y appears underabundant. Such anomalies are not readily interpretable, deep mixing would produce an enhanced Na and low O and Mg, at variance to what is observed. While it would be tempting to interpret the overabundance of heavy elements as due to s-process enrichment, the stars do not appear luminous enough to be on the thermally pulsating asymptotic giant branch, where this mechanism is operative. Moreover, s-process enrichement would produce also a high Y abundance and no Eu (which is thought to be a ``pure'' r-process element), at variance to the low Y and high Eu abundances observed here. On the other hand, these stars could have been born in r-process enhanced material (sugested by the Eu enhancement). Howevever this seems also quite implausible since the r-process is thought to take place in Type II supernovae which also produce large amounts of O and other $\\alpha$-elements, which are not observed to be enhanced in our stars. This surprising pattern is reminiscent to what is observed in the young supergiants in both Magellanic Clouds, where the ratios of the moderate-mass s-process elements Y and Zr to iron are essentially solar, whereas the heavier species Ba to Eu are overabundant by ratios [X/Fe] of the order of 0.3 and 0.5~dex respectively in the LMC and SMC (Hill et al. 1995, Hill 1997, Luck et al 1998). In the Magellanic Clouds also, we are at loss of an explanation for this behaviour (see discussion in Hill 1997). Note that our two Sgr giants have the same overall metallicity as the LMC young population, and that the heavy elements overabundances are also of the same order as in the LMC." }, "0005/astro-ph0005097_arXiv.txt": { "abstract": "We report on a ROSAT PSPC observation of the highly-luminous $z = 4.72$ radio-loud quasar GB~1428+4217 obtained between 1998 December 11 and 17, the final days of the ROSAT satellite. The low-energy sensitivity of the PSPC detector was employed to constrain the intrinsic X-ray absorption of the currently most distant X-ray detected object. Here we present the detection of significant soft X-ray absorption towards GB~1428+4217, making the absorbing material the most distant matter yet probed with X-ray spectroscopy. X-ray variability by $25\\pm 8$ per cent is detected on a timescale of 6500 s in the rest frame. The X-ray variation requires an unusually high radiative efficiency $\\eta$ of at least 4.2, further supporting the blazar nature of the source. ", "introduction": "Observations of high redshift quasars are of wide cosmological importance since these objects are thought to be associated with the earliest collapsed structures. Many high-redshift radio-loud quasars have recently been found to show low-energy X-ray cutoffs, and these are believed to be associated with intrinsic X-ray absorbers of column densities several times $\\rm 10^{22}\\ cm^{-2}$ (e.g. Elvis et al. 1994a; Elvis et al. 1998; Fiore et al. 1998). The radio-loud quasar GB~1428+4217 is currently the most distant X-ray detected object. It has been studied with the ROSAT HRI and ASCA (Fabian et al. 1997, 1998). These observations revealed an extreme isotropic X-ray luminosity of about $\\rm 1.3 \\times 10^{47}\\ erg\\ s^{-1}$ as well as variability by a factor of 2 on a timescale of about 2.4 days in the rest frame. The ASCA observations did not allow tight constraints to be placed on X-ray absorption due to the lack of low-energy response in the 0.1--0.5 keV band and the limited low-energy calibration. The ROSAT HRI did not give the needed spectral information. Therefore, we proposed GB~1428+4217 for observation during the last ROSAT observations in December 1998, employing the excellent low-energy sensitivity of the PSPC detector to constrain the intrinsic X-ray absorption in GB~1428+4217 as well as search for flux and associated spectral variability. A value of the Hubble constant of $H_0$=$\\rm 70\\ km\\ s^{-1}\\ Mpc^{-1}$ and a cosmological deceleration parameter of $q_0 = \\rm \\frac{1}{2}$ have been adopted throughout. ", "conclusions": "In this paper we report on the detection of significant soft X-ray absorption in the most distant X-ray detected object, the highly X-ray luminous quasar GB~1428+4217, obtained during the last ROSAT observations in December 1998. The low-energy sensitivity of the PSPC detector was employed to constrain the intrinsic X-ray absorption. The soft X-ray absorption of $N_{\\rm H, fit}^{\\rm z=0} \\rm = (3.14 \\pm 0.35) \\times 10^{20}\\ cm^{-2}$ is larger than the Galactic column of $N_{\\rm H, Gal} \\rm = (1.4 \\pm 0.4) \\times 10^{20}\\ cm^{-2}$ towards GB~1428+4217 at the 5 sigma level. The inferred intrinsic column density is $N_{\\rm H, fit}^{\\rm z=4.72} \\rm = (1.52 \\pm 0.28) \\times 10^{22}\\ cm^{-2}$. The most probable explanation for the soft X-ray absorption is intrinsic absorption in GB~1428+4217. An interesting possibility is that this gas is connected with the youth and possible formation of the host galaxy. Absorption by intergalactic singly ionized helium (He~{\\sc ii}), ionized oxygen (O~{\\sc vii}) or an intrinsic break in the spectrum of GB~1428+4217 are found to be unlikely explanations for the soft X-ray absorption. X-ray flux variability by $25\\pm8$ per cent is detected on a time scale of about 6500 s in the rest frame. The derived efficiency limit of $\\eta\\ge 4.2$ is remarkably large and further supports the blazar nature of the object." }, "0005/astro-ph0005212_arXiv.txt": { "abstract": "We found a strong QPO feature at $0.085\\pm0.002$ Hz in the power spectrum of X-ray transient XTE J1118+480. The QPO was detected in PCA/RXTE data with an amplitude close to 10\\% rms, and the width $0.034\\pm0.006$ Hz. The shape of the power spectrum is typical for black hole candidates: almost flat at frequencies lower than 0.03 Hz, roughly power law with slope $\\sim$1.2 from 0.03 to 1 Hz, with a following steepening to $\\sim$1.6 at higher frequencies. The hard energy spectrum detected up to $\\sim$150 keV and the absence of significant X-ray variability at the high frequencies above 100 Hz strongly support the identification of XTE J1118+480 as black hole transient. ", "introduction": "The transient X-ray source XTE J1118+480 was discovered with the RXTE All-Sky Monitor on March 29th, 2000. Subsequent RXTE pointed observations revealed a power law energy spectrum with a photon index of about 1.8 up to at least 30 keV. No X-Ray pulsations were detected (Remillard et al., 2000) In hard X-rays the source was observed by BATSE up to 120 keV (Wilson\\&McCollough 2000). Uemura, Kato \\& Yamaoka(2000) reported the optical counterpart of 12.9 magnitude in unfiltered CCD. The optical spectrum was found typical for the spectrum of an X-Ray Nova in outburst (Garcia et al. 2000). Pooley \\& Waldram (2000) using Ryle Telescope detected a noisy radiosource with flux density of 6.2 mJy at 15 GHz. All existing observations show that XTE J1118+480 is similar to the black hole transients in close binaries with a low mass companion. In this Letter we report on the detection of quasi-periodical oscillations (QPO) in the power spectrum of this source. ", "conclusions": "Low frequency QPO peaks were reported earlier in the power spectra of several black hole candidates in their low/hard state -- at $\\sim$0.03--0.07 Hz with $Q\\sim1$ for Cyg X-1 (Vikhlinin et al. 1992, 1994, Kouveliotou et al. 1992a), at $\\sim$0.3 Hz for GRO J0422+32 (Kouveliotou et al. 1992b, Vikhlinin et al. 1995), $\\sim$0.8 Hz for GX 339-4 (e.g. Grebenev et al, 1991) and in the high/soft state of LMC X-1 (\\cite{eb_lmcx1}) and XTE J1748--288 (\\cite{we_1748}). Impressive QPOs with harmonics were observed in the power spectra of Nova Muscae 1991 (e.g. \\cite{jap_muscae}, \\cite{bel_qpo}), GRS 1915+105 (e.g. Greiner et al. 1996, Trudolyubov et al. 1999b). The detection of low frequency QPO in the power spectrum of XTE J1118+480 allows us to add another black hole candidate to this sample. In all these cases the QPO peak lies close to the first (low frequency) break in the power spectrum (see also \\cite{wvdk}). The optical counterpart of XTE J1118+480 is sufficiently bright to check for the presence of corresponding low frequency optical variability with $f\\sim0.085$ Hz. The power spectra of black hole candidates are drastically different from those of neutron stars in LMXBs in similar low/hard spectral state. Sunyaev and Revnivtsev (2000) presented a comparison of power spectra for 9 black hole candidates and 9 neutron stars. None of the black hole candidates from this sample show a significant variability above $\\sim$100 Hz, while all 9 neutron stars were noisy well above 500 Hz, with the significant contribution of high-frequency noise $f>150$ Hz to the total variability of the source. The power spectrum of the newly discovered X-ray transient XTE J1118+480 (see Fig 2) looks very similar to other black hole PDSs (see Fig.1 of Sunyaev and Revnivtsev, 2000). The detection of low frequency QPO, lack of high-frequency noise and a hard energy spectrum detected up to $\\sim$150 keV in X-rays are supportive arguments for the earlier identification of XTE J1118+480 as a black hole candidate." }, "0005/astro-ph0005024_arXiv.txt": { "abstract": "The proposed global astrometry mission {\\it GAIA}, recently recommended within the context of ESA's Horizon 2000 Plus long-term scientific program, appears capable of surveying the solar neighborhood within $\\sim$ 200 pc for the astrometric signatures of planets around stars down to the magnitude limit of $V$=17 mag, which includes late M dwarfs at 100 pc. Realistic end-to-end simulations of the GAIA global astrometric measurements have yielded first quantitative estimates of the sensitivity to planetary perturbations and of the ability to measure their orbital parameters. Single Jupiter-mass planets around normal solar-type stars appear detectable up to 150 pc ($V\\le $12 mag) with probabilities $\\ge$ 50 per cent for orbital periods between $\\sim$2.5 and $\\sim$8 years, and their orbital parameters measured with better than 30 per cent accuracy to about 100 pc. Jupiter-like objects (same mass and period as our giant planet) are found with similar probabilities up to 100 pc. These first experiments indicate that the {\\it GAIA} results would constitute an important addition to those which will come from the other ongoing and planned planet-search programs. These data combined would provide a formidable testing ground on which to confront theories of planetary formation and evolution. ", "introduction": "At the very end of 1995, the discovery~\\cite{mayo} of the first Jupiter-mass ($M_J$) planet orbiting a normal star other than the Sun heralded the beginning of a new era of extraordinary discoveries in the realm of extra-solar planets, bringing with them the hope for a better understanding of the formation and frequency of planetary systems, and perhaps of bringing us closer to the ultimate goal of discovering extraterrestrial life. After four years since that discovery, spectroscopic programs have been able to reveal some twenty extra-solar planets, i.e., objects with a lower mass limit below the 13-$M_J$ cut-off which has been adopted by Oppenheimer and Kulkarni~\\shortcite{oppen99} to differentiate giant planets from brown dwarfs. However, these discoveries have raised new and troubling questions in our understanding of the properties of planetary systems. The fundamental tenets upon which present theories are based include nearly circular orbits and giant planets formed several AU from the central star, in contrast with the very short orbital periods~\\cite{mayo} and high eccentricities~\\cite{lath89,coch,mazeh} found for several of the new discoveries. Their interpretation as {\\it bona-fide} planets rests on our understanding of correlations shown by their orbital and physical parameters, as recently discussed by Black~\\shortcite{black} and earlier by Duquennoy \\& Mayor~\\shortcite{duque} in their work on solar-type binary stars. New models, which employ specific physical and dynamical mechanisms like {\\it in-situ} formation~\\cite{bode2} or orbital migration~\\cite{lin1,trill,murray}, have been proposed to justify the presence of {\\it hot jupiters} around normal stars, demonstrating that the interplay between additional theoretical work and more observational data will be necessary for a continued improvement in our theoretical understanding of how planets form and evolve, and where Earth-like planets could eventually be found. However, simply adding a few tens of additional detections of giant extra-solar planets is not enough. A better understanding of the conditions under which planetary systems form and of their general properties requires large, {\\it complete} samples of planets, with useful upper limits on Jupiter-mass planets at several AU from the central star.\\\\ Ongoing and planned radial velocity surveys~\\cite{mayo,coch,noyes,marcy,marcy2} have started filling significant portions of the relevant parameter space. Searches based on relative astrometry from the ground and in space~\\cite{map,pti,keck,vlti,pravdo} will be an important complement to the spectroscopic work and, probably, the preferred means for establishing the existence of planets around young stars and that of low mass planets down to a few Earth masses, as will be the case for SIM~\\cite{boden,unwin}. A {\\it HIPPARCOS}-like, space-borne global astrometric mission, which can survey the whole sky to faint magnitudes and with high astrometric accuracy, will enable the monitoring of large ($> 10^5$) samples of stars, with well understood completeness properties. This, depending on actual values of planetary frequencies~\\cite{ppiv}, could yield the possibility of making firm measurements of statistical properties of planetary systems. For, different correlations among orbital parameters (eccentricity, period or semi-major axis) and measurable differences in planetary frequency are likely to be generated by diverse planetary formation scenarios (core accretion and disk instability are the two known to date) and evolution mechanisms, as well as different formation and evolution processes of the parent star (binarity, spectral type, metallicity, age). An astrometric mission such as {\\it GAIA} appears well poised for such a systematic census of planetary systems within $\\sim$ 200 pc from the Sun. The {\\it GAIA} concept was originally proposed by Lindegren and Perryman~\\shortcite{linde} as a possible Cornerstone--class mission within the Horizon 2000+ program of scientific satellites of the European Space Agency. This satellite is designed to chart more than one billion objects (stars, extra-galactic objects, and solar system objects) on the sky down to the limiting magnitude of $I=20$. The targeted final accuracy is $\\sim$ 10 $\\mu$as on positions and parallaxes, and $\\sim$ 10 \\muas/year on proper motions at the reference magnitude of $V=15$ for a G2V star~\\cite{gilm}, and for a mission life time of 5 years. In the following sections we show and discuss some relevant results derived from detailed end-to-end simulations of the data acquisition and analysis process for {\\it GAIA}, which, as we will see, appears capable of discovering Jupiter-mass planets around $\\sim$ 3$\\times 10^5$ candidate stars (including dwarfs earlier than K5). ", "conclusions": "In this work we have given the first quantitative evaluation of the detectability horizon of single extra-solar giant planets around single normal stars in the neighborhood of our solar system for the global astrometry mission {\\it GAIA}. Complete simulations, comprehensive of observations of star-planet systems and successive statistical analysis of the simulated data, have yielded the following results: \\begin {itemize} \\item [a)] it will be possible to detect more than 50 per cent of all {\\em Jupiter-like} planets (orbital period $P = 11.8$ years) orbiting solar-type stars within 100 pc; {\\em Jupiter-size} planets, with shorter orbital periods, will be detectable up to 200 pc, with similar probabilities; \\item [b)] for {\\em true Sun-Jupiter systems} it will be possible to determine the full set of orbital parameters and to derive accurate estimates of the masses up to distances of order of 50 pc, value which doubles if we consider the range of periods in the vicinity of the mission lifetime; \\item [c)] simulated observations of a selection of the actually known extra-solar planets, discovered by means of spectroscopic measurements, provide a meaningful estimate of the uncertainty with which masses and orbital elements can be determined for the known star-planet systems and for a substantial fraction of those that will be found within the context of such a global astrometry mission. Although preliminary, our results indicate that these systems will be easy to discover and their orbital parameters will be accurately determined with {\\it GAIA}, except very short-period systems such as 51 Peg. \\end {itemize} Hence, our results indicate that: $1)$ {\\it GAIA} would monitor all of the hundreds of thousand F-G-K stars (i.e., whose masses are within a factor $\\sim$ 1.5 that of the Sun) up to a distance of $\\sim$ 200 pc from the Sun, in search for astrometric signatures due to the presence of giant planets ($M\\simeq M_J$) with orbital periods up to Jupiter's; $2)$ a significant fraction of the detected planets would have the main orbital parameters (semi-major axis, period, eccentricity, inclination) measured to better than 30 per cent accuracy. Therefore, the {\\it GAIA} survey would uniquely complement the expectations from other ongoing and planned spectroscopic and astrometric planet searches, both from ground and in space, thus helping with the creation of the fundamental testing ground on which to measure the validity of actual theoretical models of formation and evolution. {\\it GAIA}'s discovery potential might have significant impact on our knowledge of the distribution laws of the most relevant orbital parameters, and it would contribute to determine the frequency of planetary systems themselves in the solar neighborhood and, by extrapolation, in the whole Galaxy. A vast all-sky astrometric survey would help understand peculiar characteristics of these systems, e.g., whether giant planets lying in the outer regions are common: such planetary scenarios may be worth further investigation, as, according to present theoretical models, this could indicate presence of low mass planets in the inner regions, possibly in the parent stars' habitable zones. The monitoring of hundreds of thousands stars directly implies the chance to investigate objects belonging to a wide range of spectral types, thus providing the important observational material~\\cite{boss} to decide whether giant planets are more likely to form by means of gravitational instability in disks (once they are found more often around young stars~\\cite{kui,came,bode}), or by means of accretion of planetesimals (once they are found more often around old stars~\\cite{poll,lissa,polla}). The high precision global astrometric measurements will estimate the inclination $i$ of the orbital planes for the majority of the presently known planetary systems and for a large fraction of those that will be eventually discovered: it will then be possible to provide unambiguous mass estimations of such dark companions, reducing significantly the uncertainty on the mass range in the transition region from brown dwarfs to giant planets." }, "0005/gr-qc0005125_arXiv.txt": { "abstract": "In this paper we study the perturbations of the charged, dilaton black hole, described by the solution of the low energy limit of the superstring action found by Garfinkle, Horowitz and Strominger. We compute the complex frequencies of the quasi-normal modes of this black hole, and compare the results with those obtained for a Reissner-Nordstr\\\"{o}m and a Schwarzschild black hole. The most remarkable feature which emerges from this study is that the presence of the dilaton breaks the \\emph{isospectrality} of axial and polar perturbations, which characterizes both Schwarzschild and Reissner-Nordstr\\\"{o}m black holes. ", "introduction": "\\label{intro} The study of the effects produced by the coupling between gravity and scalar fields is a rather hot issue, since a definite prediction of superstring theory is just the existence of a scalar field, namely the \\emph{dilaton}. For instance, this coupling could have played an important role during the early phases of the life of the Universe. Indeed, it has been shown that a possible cosmological implication of the dilaton is the so-called ``pre-big-bang'' scenario\\ \\cite{PBB}, characterized by a dilaton-driven inflationary period, that should have left its peculiar fingerprints on the cosmic gravitational wave background\\ \\cite{maggiore}. On the experimental side, recent investigations on the interaction of scalar waves with gravitational detectors have shown that a scalar component of the gravitational radiation should excite the monopole mode of a resonant spherical antenna\\ \\cite{maura} and also give rise to specific correlations between the signals revealed by a resonant sphere and an interferometer\\ \\cite{nicolis}. Among all possible astrophysical sources of gravitational waves, black holes should have the most typical and in principle recognizable frequency spectrum. According to the Einstein theory of gravity, a black hole which dynamically interacts with its surroundings emits wave bursts having shape and intensity which initially depend on the features of the external perturbations. As a late-time effect, however, it generates gravitational wave trains having characteristic frequencies which are \\emph{independent} of the initial perturbations, the so-called \\emph{quasi normal modes}\\ \\cite{qnm}. And we know from the theory that these frequencies can only depend on the three parameters characterizing a black hole, namely the mass, the charge and the angular momentum. The question whether a black hole can support an additional scalar degree of freedom and still generate a regular, asymptotically flat, spacetime free of naked singularities or any other kind of pathological behaviour has been widely investigated in the literature. Having in mind the ``no-hair theorem'' program, Bekenstein\\ \\cite{bek1} in 1972 showed that black holes cannot support a \\emph{minimally coupled} scalar field, if the solution is required to be asymptotically flat. Subsequently, he developed a procedure to generate black hole solutions surrounded by a \\emph{conformal} scalar field\\ \\cite{bek3}, but they proved to be \\emph{unstable} against radial perturbations\\ \\cite{bron}. In 1990 Ferrari and Xanthopoulos \\ \\cite{valexan}, using a Kaluza-Klein approach, derived equations describing a gravitational field coupled with a massive scalar field, obtained the conformal structure of the metric, and discussed several kind of possible couplings. Yet, no definite answer was given to the question whether a satisfactory black hole can exist surrounded by a massive scalar field. The search for regular exact solutions describing black holes endowed with a scalar field got a formidable boost when it became clear that, as long as the curvature is small, all vacuum solutions of general relativity are also approximate solutions of string theory. As a matter of fact, charged black hole solutions coupled to a scalar field, viz a {\\it dilaton} in this context, can be obtained by the action (we use geometric units throughout) \\be\\label{theaction} S = \\int d^{4}x \\sqrt{-g}\\ [R - 2(\\nabla \\phi )^{2} + e^{-2a\\phi} F^{2}], \\ee which, by variations, gives the equations: \\begin{equation}\\label{maxxeq} \\nabla_{\\mu}( e^{-2a\\phi} F^{\\mu\\nu})=0, \\end{equation} \\begin{equation}\\label{dilaeq} \\nabla^{2} \\phi - \\frac{1}{2} e^{-2 a \\phi} F^{2}=0, \\end{equation} \\begin{equation}\\label{einseq} R_{\\mu\\nu} =2\\nabla_{\\mu}\\phi\\nabla_{\\nu}\\phi - 2e^{-2a\\phi} F_{\\mu\\rho}F_{\\nu}^{\\ \\rho} + \\frac{1}{2} g_{\\mu\\nu} e^{-2a\\phi} F^{2}, \\end{equation} where $F^{2} = F_{\\mu\\nu}F^{\\mu\\nu}$ is the first Maxwell invariant and $a$ is a non-negative real constant regulating the strength of the coupling between the dilaton and the Maxwell field. This theory has a number of interesting limiting cases. The limit $a=0$ corresponds to the ordinary Einstein-Maxwell theory plus a Klein-Gordon scalar field with zero mass. The case $a=\\sqrt{3}$ is the four-dimensional reduction of Kaluza-Klein theory. Finally, for $a=1$ the action (\\ref{theaction}) describes the tree-level low energy limit of superstring theory in the so-called \\emph{Einstein frame}. A large class of black hole solutions of this theory in an arbitrary number of dimensions has been found by Gibbons and Maeda\\ \\cite{gibb1}. Their results were next specialized to four dimensions by Garfinkle, Horowitz and Strominger\\ \\cite{ghs} (GHS) who studied a solution describing a spherically symmetric, charged, dilaton black hole. In the GHS solution the electric charge and the dilaton are not independent parameters: when the charge is set equal to zero the dilaton also disappears, and the solution reduces to that of a Schwarzschild black hole. In a sense, this is a consequence of the ``no-hair theorems'' which limits the number of free parameters of a black hole to three. The equations satisfied by small perturbations of this solution have been derived and studied by Holzhey and Wilczek\\ \\cite{HW} in the line of the general treatement given by Chandrasekar\\ \\cite{MT}. They first showed that it is possible to reduce the perturbed equations to five decoupled wave equations with potential barriers, and, as a by-product of their analysis, they argued that the GHS solution is stable under external small perturbations. In this paper we take a further step in the study the properties of the coupled emission of electromagnetic, scalar and gravitational radiation by such black holes, computing the \\emph{quasi-normal mode frequencies} of the GHS solution. Comparing the spectrum of the latter with those of Schwarzschild and Reissner-Nordstr\\\"om, we show that the presence of the scalar field breaks the relevant feature of the \\emph{isospectrality} of the axial and polar perturbations. The paper is organized as follows. In Section 2 we summarize the features of the GHS solution. In Sections 3 and 4 the equations governing the axial and polar perturbations are discussed separately. We shall not derive the axial equations, since there is nothing to add to the derivation made in Ref. \\ \\cite{HW}. On the other hand, the polar equations will be discussed in greater detail. Some misprints appearing in Ref. \\ \\cite{HW} are corrected and the explicit expression of the matrix whose eigenvalues are the potentials governing polar perturbations is given in the Appendix. In Section 5 the quasi-normal frequencies of the GHS black hole obtained with an extended WKB approch\\ \\cite{iyerwill} are calculated for different values of the parameters. Finally, in Section 6 the results obtained are discussed and the main {qualitative} differences between the spectra of the dilaton and the Reissner Nordstr\\\"{o}m black holes are pointed out. ", "conclusions": "\\label{sec:6} The spectrum of the quasi-normal modes of a charged, dilaton black hole is different from that of a Schwarzschild or a Reissner-Nordstr\\\"om black hole. For a Schwarzschild black hole the perturbations are completely described by the Regge-Wheeler and the Zerilli equations for the axial and the polar perturbations. Although the analytic form of the two potential barriers is different, they are related by a very simple equation (MT, ch. 5, \\& 43) which allows them to have the same reflexion and trasmission coefficients. Therefore, since the quasi-normal mode frequencies are the singularities of the scattering amplitude, it follows that the two potentials are \\emph{isospectral}. Thus, a perturbed Scharzschild black hole emits axial and polar gravitational waves at exactly the same frequencies. For a Reissner-Nordstr\\\"om black hole the perturbed equations can be reduced to four wave equations, two for the axial and two for the polar perturbations, respectively. The four wave-functions \\op Z^\\pm_1\\cl and \\op Z^\\pm_2,\\cl where $+$ stands for polar and $-$ for axial, are a linear combination of the gravitational and the electromagnetic functions belonging to the corresponding parity. It turns out that the two potentials\\op V^+_1\\cl and\\op V^-_1\\cl are again related in such a way that they have the same reflection and trasmission coefficients, and the same is true for \\op V^\\pm_2.\\cl Thus the coupling GW-EM is such that it preseves the isospectrality of the axial and polar perturbations. However there is an important difference with respect to the Schwarzschild case: no quasi-normal mode exists that is purely electromagnetic or gravitational, which means that the excitation of a mode will be accompanied by the simultaneous emission of both gravitational and electromagnetic waves. For a charged black hole in a theory described by action (\\ref{theaction}) the situation is different. Let us consider the axial perturbations first. As shown in Section 3, the perturbed equations can be reduced to two wave equations, but the perturbed dilaton does not couple to the electromagnetic and gravitational fields. This is due to the fact that the dilaton is a scalar, and its axial perturbation vanishes. Consequently, the excitation of an axial mode will be accompanied only by the emission of gravitational and electromagnetic waves. However, the dilaton appears in the unperturbed metric functions that determine the shape of the potentials of the axial wave equations. Thus it affects the scattering properties of the axial potentials. It is interesting to note that the real part of the quasi-normal mode frequencies of the axial potential \\op V_2^a\\cl are very similar to those of the Reissner-Nordstr\\\"om black hole (see Fig. \\ref{fig:4}), even though the dilaton solution does not reduce to the Reisnner-Nordstr\\\"om solution in any limiting case. On the other hand, the two wave equations which describe the polar perturbations of a Reissner-Nordstr\\\"om black hole, are replaced by three wave equations in the case of a dilaton black hole, and they couple the gravitational, electromagnetic and scalar perturbations. This occurrence breaks the symmetry between axial and polar perturbations, and makes the scattering properties of the two parities different (see Fig. \\ref{fig:5}). We conclude that for a dilaton black hole the excitation of an axial mode induces the simultaneous emission of gravitational and electromagnetic waves, whereas the excitation of a polar mode is accompanied by the further emission of scalar radiation. In addition, gravitational and electromagnetic polar waves are emitted with frequencies and damping times different from the axial ones." }, "0005/astro-ph0005162_arXiv.txt": { "abstract": "The responses of perturbations added into the optically thick, advection-dominated accretion disk (ADAD), what we call the slim disk (SD), are investigated through numerical simulations. Although it is proposed that the SD is thermally stable, I find that a perturbation added into the disk is not rapidly damped and moves through the disk in its free-fall time. After the perturbation moves, the global structure of the disk does not vary very much. These facts may account for the substantial variability of the X-ray luminosities of stellar super-luminal jet sources (SLJSs) and Narrow-Line Seyfert~1s (NLS1s). ", "introduction": "Recent X-ray observations report that not only stellar black hole candidates (SBHCs) in their low states (=faint sources), but also NLS1s and stellar SLJSs (=bright sources) exhibit X-ray fluctuations (variability). The fluctuations of bright sources are made in the optically thick ADAD, what we call SD (=bright disk). However, the time evolution of the SD has not been investigated so far. The numerical simulation by Manmoto et~al. (1996) is well known as a time-evolution calculation of the optically thin ADAD (=faint disk). The disturbance added into the optically thin ADAD falls into the central star. The disk luminosity increases when the disturbance falls, and the light curve of this process is in good agreement with the X-ray shot configuration of the SBHC Cyg X-1 in its low state. I investigate how the luminosities of disturbed SDs vary. I add a similar disturbance as Manmoto et~al. (1996) into the SD, and as a result, a similar light curve to that of the optically thin ADAD is obtained. The basic equations and numerical procedures are described in \\S 2. The resultant time evolution and discussion will be presented in \\S 3. ", "conclusions": "" }, "0005/astro-ph0005448_arXiv.txt": { "abstract": "We have obtained an exceptionally high S/N, high resolution spectrum of the gravitationally lensed quasar Q1422+231. A total of 34 \\CIV\\ systems are identified, several of which had not been seen in previous spectra. Voigt profiles are fitted to these \\CIV\\ systems and to the entire \\lya\\ forest in order to determine column densities, $b$-values and redshifts for each absorption component. The column density distribution for \\CIV\\ is found to be a power law with index $\\alpha = 1.44 \\pm 0.05$, down to at least log $N$(\\CIV) = 12.3. We use simulations to estimate the incompleteness correction and find that there is in fact no evidence for flattening of the power law down to log $N$(\\CIV) = 11.7 -- a factor of ten lower than previous measurements. In order to determine whether the \\CIV\\ enrichment extends to even lower column density \\HI\\ clouds, we utilize two analysis techniques to probe the low column density regime in the \\lya\\ forest. Firstly, a composite stacked spectrum is produced by combining the data for Q1422+231 and another bright QSO, APM 08279+5255. The S/N of the stacked spectrum is 1250 and yet no resultant \\CIV\\ absorption is detected. We discuss the various problems that affect the stacking technique and focus in particular on a random velocity offset between \\HI\\ and its associated \\CIV\\ which we measure to have a dispersion of $\\sigma_v$ = 17 \\kms. It is concluded that, in our data, this offset results in an underestimate of the amount of \\CIV\\ present by a factor of about two and this technique is therefore not sufficiently sensitive to probe the low column density \\lya\\ clouds to meaningful metallicities. Secondly, we use measurements of individual pixel optical depths of \\lya\\ and corresponding \\CIV\\ lines. We compare the results obtained from this optical depth method with analyses of simulated spectra enriched with varying \\CIV\\ enrichment recipes. From these simulations, we conclude that more \\CIV\\ than is currently directly detected in Q1422+231 is required to reproduce the optical depths determined from the data, consistent with the conclusions drawn from consideration of the power law distribution. ", "introduction": "Our understanding of the intergalactic medium (IGM) has undergone a paradigm shift in recent years, largely as a result of powerful hydrodynamical simulations. When the \\lya\\ forest was first observed in the late 1960s, the rich field of \\HI\\ absorption blueward of the QSO's \\lya\\ emission was interpreted as the detection of discrete intergalactic clouds (e.g.\\ Lynds \\& Stockton 1966; Lynds 1971; Sargent et al.\\ 1980). In order to understand the existence of such isolated absorbers, various theories of cloud confinement, including self-gravity (Melott 1980), cold dark matter minihaloes (Rees 1986) and the presence of an inter-cloud medium (e.g.\\ Sargent et al.\\ 1980; Ostriker \\& Ikeuchi 1983) were proposed. However, the advent of hydrodynamical simulations, which model the growth of structure in the high redshift universe, provided a significant revision to our picture of the IGM (see the recent review by Efstathiou, Schaye and Theuns 2000). It has been found that in the presence of a UV ionizing background, the `bottom-up' hierarchy of structure formation knitted a complex, but smoothly fluctuating `cosmic web' in the IGM (e.g.\\ Cen et al.\\ 1994; Hernquist et al.\\ 1996; Bi \\& Davidsen 1997). The absorption in the \\lya\\ forest is caused not by individual, confined clouds, but by a gradually varying density field characterized by overdense sheets and filaments and extensive, underdense voids. The advance in theoretical simulations has been matched by increasingly high quality data, as comprehensively reviewed by Rauch (1998). One of the major discoveries concerning the IGM has been the identification of metal absorption lines associated with many of the \\lya\\ forest clouds (Cowie et al.\\ 1995; Tytler et al.\\ 1995). Thus, whilst the \\lya\\ forest was once thought to be chemically pristine, it has now been well-established that a large fraction of the high column density \\lya\\ clouds ($N$(\\HI) $>$ 14.5), associated with collapsing, over-dense structures, contain metals (most notably \\CIV) -- the signature of enrichment by the products of stellar nucleosynthesis (Songaila \\& Cowie 1996). The presence of metals in the \\lya\\ forest may be reasonably explained either by in-situ enrichment (local star formation in the \\HI\\ cloud itself or in a nearby galaxy) or by early pre-enrichment by a high redshift episode of Population III stars. Whilst the effects of supernova feedback are still not fully understood and therefore the spatial extent of wind-driven ejecta is poorly constrained, enrichment by galactic winds and superbubbles is unlikely to be an efficient way to distribute metals over distances large in comparison with the mean separation between galaxies (MacLow \\& Ferrara 1999). Instead, models have appealed to larger scale processes such as merging and turbulent diffusion as the dominant mixing mechanisms (Gnedin \\& Ostriker 1997; Gnedin 1998; Ferrara, Pettini \\& Shchekinov 2000). Such processes would take time to smooth out the metallicity of the IGM so that at $z \\sim 3$ the metal enrichment is still expected to be very patchy. Whilst the deep potential wells of galaxies inhibit efficient, widespread distribution of metals far from their sites of formations, small regions of star formation at high redshift may be able to eject their nucleosynthetic products for more homogeneous mixing (e.g.\\ Nath \\& Trentham 1997 and references therein). An episode of Population III star formation may therefore have spread metals far from their sites of formation, seeding `sterile' regions of the IGM with metals (Ostriker \\& Gnedin 1996). Clearly, distinguishing between in-situ and Pop III scenarios has important implications for understanding not only the first generation of stars, but also the mechanisms by which metals are mixed and distributed from their stellar birthplaces. Several papers (Cowie \\& Songaila 1998; Lu et al.\\ 1998; Ellison et al.\\ 1999a, hereafter Paper I) have addressed this question by attempting to probe the low density regions of the IGM where the difference in metallicity predicted by in-situ and Pop III enrichment may be most marked. The detection of \\CIV\\ $\\lambda\\lambda$1548,1550 associated with low column density \\lya\\ clouds (log $N$(\\HI) $<$ 14.0) is observationally challenging, even with the capabilities of Keck, due to the extreme weakness of the metal lines. Analysis techniques have therefore been developed to effectively enhance the sensitivity of the data beyond the normal equivalent width limits of the spectra. Two in particular have received recent attention, namely the production of a stacked \\CIV\\ spectrum (Lu et al.\\ 1998) and the use of individual pixel optical depths of Ly~$\\alpha$\\ and \\CIV\\ (Cowie \\& Songaila 1998). In Paper I we attempted to reconcile the apparently conflicting results obtained from these two techniques with an analysis of a very high signal-to-noise ratio (S/N) spectrum of the ultra-luminous BAL quasar APM 08279+5255. Rigorous testing of the analysis procedures revealed that both methods suffered from hitherto unrecognised limitations and it was concluded that the question of whether or not the low density regions of the IGM have been enriched remains unanswered. In order to probe deeper into the low density IGM, we have obtained an exceptionally high S/N spectrum of the well-known lensed QSO, Q1422+231, using the HIRES instrument (Vogt 1994) on the KeckI telescope. Although this QSO has been well studied in the past (e.g.\\ Songaila \\& Cowie 1996; Songaila 1998), we have roughly doubled the exposure time of earlier spectra, allowing us to probe the metallicity of the \\lya\\ forest to more sensitive levels than has previously been achieved. The spectrum is much better suited to the present work than the the much more complex BAL quasar APM 08279+5255 (Ellison et al.\\ 1999a). We present a careful and extensive analysis of the \\CIV\\ systems in order to determine the extent of metal enrichment in the IGM. This paper is organised as follows. In \\S2 we describe the observations, the data reduction procedures and the Voigt profile fitting process used to determine column densities. We briefly discuss in \\S3 the suitability of Q1422+231 for this work and define the redshift interval over which we will perform the analysis. In \\S4, we determine the column densities of the 34 detected \\CIV\\ absorption systems in the spectrum and investigate the column density distribution of these absorbers. Finally, we critically assess the two methods of analysis, described in sections \\ref{stacksec} and \\ref{tausec}, which have been developed to probe low column density \\lya\\ clouds to very sensitive levels. We utilize a suite of simulation techniques to fully test these methods and quantify the potential inaccuracies in our analysis. We adopt $\\Omega_M$ = 1.0 throughout. ", "conclusions": "In this paper, we have addressed the enrichment history of the IGM by studying the \\lya\\ forest and its associated \\CIV\\ systems in a very high S/N ($\\sim 200)$, high resolution ($\\sim$ 8 \\kms) spectrum of the well-known lensed quasar, Q1422+231 obtained with Keck/HIRES. The numerous \\CIV\\ systems associated with high column density \\lya\\ absorbers are fitted with Voigt profiles defined by a redshift, $b$-value and column density for each component line. We investigate the \\CIV\\ column density distribution, $f(N)$, to very sensitive levels and detect several weak \\CIV\\ systems which had not been previously identified in lower quality spectra of the same object. We determine a power law index $\\alpha = 1.44 \\pm 0.05$ which continues down to log $N$(\\CIV) = 12.2 before starting to turnover. By simulating synthetic absorption lines with $b$-values taken at random from the observed distribution, we estimate a correction factor to account for incompleteness and find that the corrected data points now indicate that the power law continues down to at least log $N$(\\CIV) = 11.75, a factor of ten more sensitive than previous measurements (e.g.\\ Songaila 1997). This shows that even at these low column densities there is no evidence for a flattening of the power law and therefore there are probably many more \\CIV\\ systems that lie below the current detection limit. We investigate two methods with which it may be possible to recover these weak \\CIV\\ systems. Firstly, we select 67 \\lya\\ lines with 13.5 $<$ $N$(\\HI) $<$ 14.0 in Q1422+231 and APM 08279+5255 and produce a stacked spectrum centered on the predicted position of \\CIV\\ $\\lambda$1548. The composite stack has a S/N = 1250 and shows no residual absorption; we use synthetic stacked spectra to determine a 4$\\sigma$ upper limit of log \\CIV/\\HI\\ = $-3.1$. We critically assess the accuracy of this method by performing the stacking procedure on a suite of simulated, synthetic spectra and identify several associated problems. We investigate, in particular, the effect of a redshift offset between the position of the \\lya\\ line and its associated \\CIV. With improved statistics, we refine the redshift offset determined in Paper I and find that a $\\sigma_v$=17 \\kms\\ is present in the \\CIV\\ systems which we detect directly. By including a random redshift offset drawn from a Gaussian distribution with $\\sigma_v$=17 \\kms\\ in our stacking simulations, we find that log \\CIV/\\HI\\ = $-2.8$ is now required to achieve a 4$\\sigma$ detection. This limit is still consistent with current measured metallicities in higher column density \\lya\\ clouds and is therefore not sufficiently sensitive to determine whether the \\CIV/\\HI\\ ratio drops in low $N$(\\HI) lines. A feasibility study is performed to assess the effectiveness of a `re-center' on the maximum optical depth pixel prior to stacking for removing the effect of an unknown offset. We find that this technique can not improve the quality of the stack result in the \\CIV/\\HI\\ regime that we are targeting. It is not yet clear whether the observed redshift offset persists in the low column density \\lya\\ clouds, but it must be considered a factor. Moreover, the effects of contamination, continuum fitting errors and anomolous pixels also pose problems for this technique, although in simulations the redshift offset appears to be a major effect. Therefore, if this technique is to be pursued further to reach a meaningful detection limit, an improvement in S/N by at least a factor of two is required. The optical depth technique introduced by Cowie \\& Songaila (1998) is considered as an alternative approach. This technique is shown to exhibit several advantages over the stacking method such as its insensitivity to redshift offsets and its ability to exclude contamination from other absorption features. We develop this technique as a method that can be used to test whether the detected \\CIV\\ systems represent the full tally of absorbers. The data have optical depths consistent with an almost constant log \\CIV/\\HI\\ $\\sim -3$ down to $\\tau$(\\lya)$\\sim$ 2--3, below which $\\tau$(\\CIV) flattens off to an approximately constant value. It is unlikely that this flattening is real and is most probably caused by the effect of noise and/or continuum errors, even at such high S/N ratios as have been achieved in this spectrum. Given the many effects that may alter the measured $\\tau$ values, such as blending and noise, we do not attempt to fit the observed distribution of optical depths. Instead, our strategy is to test whether the detected \\CIV\\ systems are sufficient to reproduce the measured $\\tau$(\\CIV) and if not, determine how much additonal \\CIV\\ may be present below our current detection limit. By simulating synthetic spectra with different enrichment recipes, we have shown that the \\CIV\\ systems detected directly in the spectrum are not sufficient to reproduce the results of the optical depth analysis of Q1422+231. This is in agreement with the conclusions drawn from the column density distribution of \\CIV, i.e.\\ that the data are consistent with a continuous power law $f(N)$ down to at least $N$(\\CIV) = 11.75 and that there is therefore likely to be a large number of weak metal lines not yet directly detected. This agrees with the conclusions of Cowie \\& Songaila (1998). In order to interpret the results from the optical depth method, we have simulated synthetic spectra with a range of input \\CIV/\\HI\\ ratios. We find that including \\CIV\\ associated with strong \\lya\\ lines ($N$(\\HI) $>$14.5) but below the current detection limit, in addition to the 34 identified \\CIV\\ systems, can reproduce the optical depths measured in the observed spectrum for $\\tau$(\\lya) $>$ 3. For smaller values of $\\tau$(\\lya), some additional metals are required and we find that including \\CIV\\ in low column density \\HI\\ lines ($N$(\\HI) $<$ 14.5) with log \\CIV/\\HI\\ $= -2.6$ produces optical depth results consistent with those measured in the data. However, determining the precise \\CIV/\\HI\\ in the low $N$(\\HI) \\lya\\ clouds and the density to which the enrichment persists is still uncertain due to effects such as noise and continuum fluctuations and it is therefore not possible to say whether the low column density forest is pristine. Nevertheless, we find that even in the high optical depth \\HI\\ pixels (which will not be seriously affected by noise or small continuum errors) the detected \\CIV\\ systems are not sufficient to cause all the measured absorption and clearly there are more metals in the IGM than we can currently detect." }, "0005/astro-ph0005168_arXiv.txt": { "abstract": "We have correlated a sample of 99 Seyfert 1 \\& 98 Seyfert 2 (Rafanelli et al., 1995) interacting galaxies with the survey and pointed ROSAT data and have detected 91 Seyferts 1 \\& 47 Seyfert 2s in the X-ray band. We have performed spectral and timing analysis of the X-ray detected sources, and we have examined the relation of the X-ray luminosity $L_X$ with the interaction strength $Q$. ", "introduction": "Interacting galaxies are considered to have a wide importance on triggering of starburst and AGN activity. Therefore, we have investigated two subsamples of spectroscopically-selected, X-ray detected interacting Seyfert 1s, including NLS1, and Seyfert 2s galaxies (Rafanelli et al.,1995) with $ z<0.11, \\; m \\le 15.5 \\; and \\; \\delta \\ge -23^\\circ $.\\\\ The pairs have been selected on POSS plates using the following criteria given by Rafanelli et al. (1995): separation between the components $S < 3D_p$, where $D_p$ is the apparent major axis of the\\ Seyfert galaxy, and a magnitude-difference between Seyfert and the companion\\ $\\Delta m = m_{comp} - m_{seyfert} < 3$ . The identification is based on X-ray contour lines overlaid on optical images taken from DSS. In Fig. 1 we show as an example the overlay from the NLS1 galaxy Mkn 896. Fig. 2 shows the spectrum of this object. The deviation from the fit at energy 0.7 keV in Fig. 2 indicates the presence of a warm absorber. The pointing and survey lightcurves in Fig. 3 show the variability of this source on short time scales. \\begin{figure}[htb] \\centerline{\\psfig{figure=Pfefferkorn_fig_1.ps,height=3.1truein,width=3.1truein,angle=0}} \\caption{The NLS1 galaxy Mkn 896 with a companion galaxy. The X-ray contours are overlaid on the optical image.} \\end{figure} \\begin{figure}[htb] \\centerline{\\psfig{figure=Pfefferkorn_fig_2.ps,height=3.3truein,width=5.6truein,bbllx=25mm,bblly=10mm,bburx=200mm,bbury=270mm,angle=-90,clip=}} \\caption{Simple power-law fit results to the NLS1 Mrk 896.} \\end{figure} \\begin{center} \\begin{figure}[htb] \\hbox{ \\psfig{figure=Pfefferkorn_fig_3.ps,height=2.5truein,width=2.8truein,bbllx=10mm,bblly=20mm,bburx=190mm,bbury=240mm,angle=-90,clip=} \\psfig{figure=Pfefferkorn_fig_4.ps,height=2.5truein,width=2.6truein,bbllx=10mm,bblly=35mm,bburx=190mm,bbury=240mm,angle=-90,clip=} } \\caption{The lightcurves from the pointing and survey observation shows the short-time variability of Mkn896.} \\end{figure} \\end{center} ", "conclusions": "" }, "0005/astro-ph0005397_arXiv.txt": { "abstract": "Our study begins by revisiting the {\\it photometric} Butcher-Oemler effect using data from 7 CNOC clusters at 0.23$\\lsim$$z$$<$0.43. We construct the foreground/background corrected colour-magnitude diagrams for these clusters. Our analysis shows that the CNOC clusters reproduce the trend of increasing blue galaxy fraction with redshift as seen by Butcher \\& Oemler. We use these data to investigate the history of star formation in clusters by connecting these clusters as snapshots at different redshifts. We address two key issues. Firstly, we ask whether the simple fading and passive evolution of the blue galaxies is consistent with the properties of galaxies in nearby clusters, such as the Coma cluster. We find that the evolution of star forming field galaxies towards redder colour (once the star formation ceases on entry into the cluster environment) can successfully reconstruct colours and magnitudes of galaxies in the local cluster. There is no requirement for widespread disruption of these galaxies. Since the blue galaxies fade as they age, the fainter galaxies on the present-day colour-magnitude relation tend to have more extended star formation histories than their bright counterparts. However, this effect is {\\it not} sufficient to cause a sizable age variation for the galaxies along the colour-magnitude relation, implying that the slope is dominated by variations in metal abundance. Secondly, we address the nature of the Butcher-Oemler effect itself. We compare the distribution of galaxies in the colour-magnitude diagrams and hence infer the evolution of the rate at which galaxies have arrived in the cluster. Models in which star formation is abruptly truncated as galaxies are accreted by the cluster have difficulty in reproducing the observed colour distribution. In contrast, if star formation declines on a 1~Gyr timescale after accretion, the galaxy accretion history we infer is consistent from cluster to cluster and matches well the distribution expected in simple theoretical models. The Butcher-Oemler effect is thus driven both by the declining star formation rates of field galaxies and by a decline in the rate at which fresh galaxies are accreted by the cluster. Our study naturally leads to a comparison of the global star formation histories of galaxies in clusters and the field. We show that the star formation rate per galaxy mass for galaxies in cluster cores is significantly smaller than that of the field environment below $z$$<$1 due to the truncation of star formation. However, the factor by which star formation is suppressed is dependent on the cluster accretion history. High quality observations of clusters at higher redshifts are needed to better define this relation. ", "introduction": "It is 20 years since Butcher \\& Oemler (1978; 1984, hereafter BO84) first reported their startling results on the blue galaxy fractions of distant clusters. This and subsequent photometric work suggested that the cores of distant clusters at redshift $z$$>$0.2 contain large numbers of blue galaxies (Couch \\& Newell 1984; Lubin et al. 1996; Rakos \\& Schombert 1995), in stark contrast to the homogeneous red populations of local galaxy clusters (eg., Bower, Lucey \\& Ellis 1992; Terlevich, Bower \\& Caldwell 2000). There is an apparent contradiction between the narrow colour-magnitude (CM) relation of evolved galaxies in local clusters (which indicates a long period of passive evolution) and the blue colours of intermediate redshift galaxies (which indicates considerable, relatively recent star formation activity). Bower, Kodama \\& Terlevich (1998, BKT98), Smail et al.\\ (1998) and Balogh et al.\\ (2000) have begun the process of modelling the combined dynamical and photometric evolution of clusters in order to reconstruct the galaxies' formation histories. We continue these themes in this paper. BO84's original discovery of strong cluster evolution involved only the photometric properties of galaxies. We will refer to this aspect as the `photometric Butcher-Oemler effect'. Dressler \\& Gunn (1992) and Couch \\& Sharples (1987) undertook spectroscopy initially with the motivation of confirming spectroscopically that the blue galaxies were cluster members. They discovered that many of the blue cluster galaxies, and some red ones, had spectra of a type that are rare in local galaxy samples. These galaxies (referred to as E+A, PSG, a+k/k+a) have anomalously strong Balmer lines indicating an excess of A-stars superimposed on an K-star type spectrum. The galaxies are produced when star formation is abruptly truncated (Couch \\& Sharples 1987). The strongest lined examples require a burst of star formation just before the truncation (eg., Barger et al. 1996; Poggianti \\& Barbaro 1996; Poggianti et al. 1999; Couch et al. 1998). We refer to the increase in the incidence of such spectra with redshift as the `spectroscopic Butcher-Oemler effect' With the advent of HST, it has become possible to study the morphologies of galaxies in the distant clusters (Smail et al. 1997; Dressler et al. 1994; 1997; Couch et al. 1994; 1998). They found that the morphological mix in these clusters also evolved with redshift, creating a third variant of the Butcher-Oemler effect, which we will refer to as the `morphological BO effect'. In particular, the fraction of S0 galaxies in the cluster cores decreases rapidly with redshift (Dressler et al. 1997). Several attempts have been made to connect photometric/spectroscopic features and the morphology. Dressler et al. (1994) and Couch et al. (1994) showed that most of the blue BO84 galaxies are actually normal spirals with the rest being interacting galaxies or mergers. Poggianti et al. (1999) compared the spectroscopic and the morphological effects, suggesting that the time scale of the morphological transition is longer than the A-type star evolution since the post star burst galaxies (k+a/a+k) are observed as normal spirals rather than S0 galaxies. In this paper, we aim to clearly separate these three different (but related) evolutionary trends. We will focus on the {\\it photometric} effect, returning to the questions raised by Butcher \\& Oemler's original papers. The advantage of the photometric approach is that it allows us to study large, complete samples of galaxies rather than a smaller subset of the population that is complicated by target selection and the strength of spectral features and often limited to brighter galaxies. Recent studies have tended to concentrate on particular spectroscopic features to learn details about the nature of the star-burst/truncation phenomena seen in some cluster members (Barger et al. 1996; Poggianti et al. 1999; Balogh et al. 1999). On the other hand, we have made little progress on the evolution of the cluster population as a {\\it whole}, as galaxies accrete from field, truncate their star formation, and fade and become red towards the CM relation ridge-line. A conspicuous contradiction exits between Poggianti et al. (1999) and Balogh et al. (1999) over the way star formation is suppressed in the cluster environment. While Poggianti et al. (1999) require an abrupt end to star formation activity to explain the incidence of a+k/k+a spectra, Balogh et al. (1999) favor a slow decline in star formation in order to explain the radial dependence of the [\\oii] emission line strength. To remedy this situation, we will look at the CM diagrams of the distant clusters in more detail, examining how the whole cluster population evolves with time. {\\it The essential idea is that we determine the statistical star formation histories of galaxies in clusters by connecting a series of cluster snapshots at different redshifts.} The process has been extensively applied to global galaxy populations (eg., cosmic star formation history by Madau, Pozzeti \\& Dickinson 1998), but this is the first application specifically targeting cluster galaxies. We also address how the blue fraction varies with galaxy magnitude and with radius from the cluster centre. The variation with magnitude is of interest given Cowie's claim of cosmic `down-sizing': the progression of star formation into smaller and smaller units as the universe ages (Cowie et al. 1996). Clearly, the history of galaxy accretion will vary from cluster to cluster, and this will limit the degree to which we can reconstruct cluster galaxy evolution from a set of observations of different clusters. However, it is important to distinguish between the growth of the total cluster mass, which will be dominated by mergers between similar mass units, and the accretion of field galaxies. Monte-Carlo simulation of cluster growth (Lacey \\& Cole, 1993) shows that while the sequence of mergers varies greatly between clusters, the role of accretion is more stable, and the fraction of galaxies accreted varies $\\sim$ 2 per cent per Gyr. This is particularly the case if we base our study on X-ray selected clusters since the existence of a dense intracluster medium biases against the inclusion of clusters which are not yet dynamically relaxed (Bower et al. 1997). Nevertheless, differences in cluster accretion histories can only be eliminated using a much larger sample of clusters than we have available at present. It is also important that we try to match the masses of the clusters in order to form an evolutionary sequence. The X-ray luminosities of the CNOC clusters used in this study bracket that of the Coma cluster (\\S~2.1). Since the X-ray luminosity function evolves little over the look-back times involved (eg., Ebelling et al. 1996), the clusters should be reasonably well matched in terms of their space density. Possibly, we would have liked to compare with an even richer local cluster (\\S~2.4). However, although there is no such massive local cluster, current data on the richness dependence of the CM relation suggests that it would have even an tighter CM ridge-line (and smaller blue fraction) than that of Coma (L\\'opez-Cruz 1996). Our goals in this paper are two-fold. Initially, we will make an accurate comparison of the distant and local clusters. In particular, we will test whether the galaxies in the distant clusters simply fade and become red after their star formation is truncated or whether an additional mechanism such as tidal stripping or harassment (Moore et al. 1996) is required to explain how the distant clusters evolve into clusters like those at the present-day. We then go on to explore the distribution of galaxies in the colour-magnitude plane, and to address the difference between clusters at different redshifts. The distribution of galaxies can be used to derive the history of galaxy accretion from the field and the way the star formation is truncated. We will show that a consistent picture emerges in which the accretion rate of cluster galaxies from the field declines smoothly with time. Combined with the decline of star formation activity of field galaxies, this naturally accounts for the Butcher-Oemler effect. The layout of the paper is as follows. In \\S~2, we present the sources of data used to compose the distant and local cluster CM diagrams, and outline how we correct for background and foreground galaxies. In \\S~3, we calculate the blue galaxy fraction from the field corrected CM diagrams and discuss its dependence on magnitude and the distance from cluster centre. In \\S~4, we apply our stellar population models to calculate the colour change (red-wards) and fading of the distant cluster CM diagrams and compare with a local cluster (Coma). In \\S~5, we investigate the colour distribution of galaxies in more detail, examining how the galaxy accretion rate must evolve in time. Our conclusions are summarised in \\S~6. We use H$_0$=50 km s$^{-1}$ Mpc$^{-1}$ and q$_0$=0.1 throughout this paper, which are the same parameters adopted in the original BO84 work. The age of the universe is 16.5~Gyr for this parameter set. ", "conclusions": "We have created the field corrected CM diagrams for the 7 CNOC distant clusters (0.23$\\lsim$$z$$<$0.43) in order to study their galaxy populations in the same manner as in BO84. We make a comparison with the redshift using data for the Coma cluster from Terlevich et al. (2000). The form for the redshift evolution of the blue galaxy fraction agrees well with the trend seen in BO84. In addition, the blue galaxy fraction is generally a strong increasing function of distance from the cluster centre, and depends only weakly on magnitude. We have then applied these data-sets to build up a sequence of snap-shots of cluster evolution. We follow the evolution of the whole population of cluster galaxies in these CM diagrams accounting for the flow of galaxies across the diagram as they are accreted by the cluster. \\begin{itemize} \\item Firstly, we have shown that the blue galaxies are incorporated into the present-day tight CM relation as they fade and become redder rapidly after the truncation of their star formation, giving the possible link to the faint ($>$$M_*$+1) S0 galaxies in present-day clusters. The brighter S0 galaxies, however, must have ceased star formation earlier than $z\\sim0.4$. In clusters, as in the universe as a whole, fainter galaxies have a more extended star formation history than their bright counterparts. Nevertheless, this effect does not lead to a sizable age variation along the CM relation at the present-day. \\item Secondly, we have used the distribution of galaxies in the colour space to infer the evolution of the rate at which galaxies have been accreted by the cluster. We interpret the observed distribution in colour by following the declining star formation rate in the field and statistically assigning a star formation truncation epoch. If star formation is truncated abruptly, we find that too many recent accretion events are required relative to the past rate. In order to be consistent with a smoothly declining accretion rate, such as suggested by the extended P-S theory, we need to allow for some level of residual star formation after galaxy accretion. This suggests that ram-pressure and/or mergers/harassment cannot be so effective as to remove all the remaining gas from the system in a single orbital time-scale. The Butcher-Oemler effect is thus shown to result from a combination of the three effects, namely, declining accretion rate, declining field star formation rate, and the truncation of star formation. \\item Finally, we have presented the global star formation history for the galaxies in rich cluster cores for the first time. It is shown that the star formation in cluster galaxies is much less than that in the field galaxies below $z$=1. However, the factor by which star formation is suppressed depends on the effectiveness of the cluster environment in suppressing star formation. \\end{itemize} It is important to ask how robust these conclusions are to the assumptions that must be made in order to model the galaxy properties. When propagating the observed properties of galaxies forward in time, the past star formation rate has only a relatively weak effect (as shown by Fig.~\\ref{fig:vector}) and thus the results derived from this analysis are robust. Determining the distribution of truncation times, and hence comparing the past and present accretion rates in clusters is significantly more model dependent. Firstly, the ratio of the present to past star formation rate must be estimated for the field galaxy population. This must be parameterised by a simple model and matched to the observed colour distribution of field galaxies. The properties of the accreting galaxies can, however, be constrained by a wide range of observations in addition to the observed field galaxy colours. The need to match the cosmic star formation history suggested by Madau et al. (1998) has motivated our choice of the gas infall model. Secondly, we must parameterise the effect of the cluster environment on the galaxies that are accreted. This is the more critical component of the model, and we illustrate this by considering two scenarios. If star formation is cut-off abruptly by a star burst and gas stripping upon accretion into the cluster, the model struggles to explain the high fraction of blue objects seen in the distant clusters. In contrast an extended period of star formation decay naturally reproduces a smooth decay in the accretion rate that is similar in all of the clusters. More complex scenarios are easily possible: for example galaxies might either undergo a star burst or might decay slowly, depending on their orbit within the cluster. An important avenue for future work is to integrate spectral star formation indicators (such as the [\\oii] or H$\\alpha$ line strength), or even radio continuum flux (Smail et al. 1999), and H$\\delta$ absorption line strength into this analysis so that the degeneracy between colour and past/present star formation rate can be broken. This will improve our understanding of the truncation process. So far we have not considered the effects of dust obscuration. This is a potentially important complication that is extremely difficult to model in a straightforward manner. Adding a uniform dust extinction that depends on the present star formation rate reddens the observed colours, causing us to underestimate star formation rates. However, both the field and some of the cluster galaxies will be affected so that the reddening cancels out to a first approximation. On the other hand, if localised dust extinction accompanies a star-burst on entry to the cluster environment as suggested by Poggianti et al. (1999) and Smail et al. (1999), we would recover an uneven accretion history (similar to the behaviour shown by Model~(a)) since the `intrinsic' (ie., dust corrected) colour distribution would include many more `blue' galaxies. Clearly the potential of these techniques are best exploited by observing more and more distant clusters. To resolve the star formation history in clusters at higher redshifts ($z$$>$1), we need a good homogeneous sample of clusters at higher redshifts ($z$$>$0.5--1.0). The {\\it Rosat} deep X-ray cluster survey (Rosati et al. 1998) and the coming new surveys by {\\it Chandra} and {\\it Newton} with high sensitivity cameras will provide an ideal sample of high redshift clusters for this purpose. The analysis presented in this paper is limited by the accuracy of the the distant cluster photometry, as well as by the number of clusters in our sample. A larger sample of clusters at each redshift would allow projection effects to be averaged over and the variation in evolution history to be investigated. Wider spatial coverage is also important to more accurately subtract field population and also to investigate the environmental effect beyond the cluster core out to the virial radius. This is essential if we are to improve our knowledge of the star formation truncation mechanism itself. At present, our approach is also compromised because we do not allow for the dynamical evolution of the cluster populations. Because of this, we cannot compare the properties of galaxies at different radii within the clusters. Recent rapid progress in the high resolution N-body simulation of clusters (eg., Moore et al. 1999b; Balogh et al. 2000; Diaferio et al. 2000) will allow this aspect to be taken into account." }, "0005/astro-ph0005504_arXiv.txt": { "abstract": "We give a review of different properties of dark matter halos. Taken from different publications, we present results on (1) the mass and velocity functions, (2) density and velocity profiles, and (3) concentration of halos. The results are not sensitive to parameters of cosmological models, but formally most of them were derived for popular flat \\LCDM~ model. In the range of radii $r=(0.005-1)\\rvir$ the density profile for a quiet isolated halo is very accurately approximated by a fit suggested by Moore \\etal (1997): $\\rho\\propto 1/x^{1.5}(1+x^{1.5})$, where $x=r/\\rs$ and \\rs~ is a characteristic radius. The fit suggested by Navarro \\etal (1995) $\\rho\\propto 1/x(1+x)^2$, also gives a very satisfactory approximation with relative errors of about 10\\% for radii not smaller than 1\\% of the virial radius. The mass function of $z=0$ halos with mass below $\\approx 10^{13}\\Msunh$ is approximated by a power law with slope $\\alpha =-1.85$. The slope increases with the redshift. The velocity function of halos with $\\Vmax< 500$~km/s is also a power law with the slope $\\beta= -3.8-4$. The power-law extends to halos at least down to 10~km/s. It is also valid for halos inside larger virialized halos. The concentration of halos depends on mass (more massive halos are less concentrated) and environment, with isolated halos being less concentrated than halos of the same mass inside clusters. Halos have intrinsic scatter of concentration: at $1\\sigma$ level halos with the same mass have $\\Delta(\\log{c_{\\rm vir}})=0.18$ or, equivalently, $\\Delta\\Vmax/\\Vmax =0.12$. Velocity anisotropyfor both subhalos and the dark matter is approximated by $\\beta(r) =0.15 + 2x/[x^2+4]$, where $x$ is radius in units of the virial radius. ", "introduction": "During the last decade there was an increasingly growing interest in testing predictions of variants of the cold dark matter (CDM) models at subgalactic ($\\lesssim 100{\\rm\\ kpc}$) scales. This interest was first induced by indications that observed rotation curves in the central regions of dark matter dominated dwarf galaxies are at odds with predictions of hierarchical models. Specifically, it was argued (Moore 1994; Flores \\& Primack 1994) that circular velocities, $v_c(r)\\equiv[GM( 0.5$ in the neutrino-heated matter, which suggests that the overproduction problem of neutron-rich nuclei with mass numbers $A\\approx 90$ in exploding models may be absent when a Boltzmann solver is used for the $\\nu_e$ and $\\bar\\nu_e$ transport. ", "introduction": "The mechanism of supernova explosions of massive stars is still not satisfactorily understood. Detailed numerical models showed that the hydrodynamic shock, which is launched when the collapsing stellar core bounces abruptly by the stiffening of the equation of state (EoS) at nuclear densities, cannot propagate out promptly but stalls because of energy losses due to photodisintegration of iron-group nuclei and neutrino emission from the shock-heated matter (e.g., Bruenn 1985, 1989a,b; Myra et al.~1987). Early suggestions that energy deposition by neutrinos might cause an explosion reach back to Colgate \\& White (1966). The modern version of the neutrino-driven ``delayed'' explosion mechanism is due to Wilson (1985), who found that neutrino energy deposition can revive the stalled shock on a time scale of several hundred milliseconds after bounce (Bethe \\& Wilson 1985). Because of the complexity of the involved physics and the low efficiency of the neutrino energy transfer it remained unclear for years whether the explosions are sufficiently energetic and whether the delayed mechanism works for a larger range of stellar masses (Wilson et al.\\ 1986; Bruenn 1993). Recognizing that neutron-finger convection in the newly formed neutron star could increase the neutrino luminosities, Wilson \\& Mayle (1988, 1993) managed to obtain healthy explosions. However, the question of neutron star convection is not finally settled and currently it is not clear whether neutron-finger instabilities or Ledoux convection (Burrows 1987, Pons et al.~1999) or quasi-Ledoux convection (Keil, Janka, \\& M\\\"uller 1996; Janka \\& Keil 1998) or none (Bruenn et al.~1995; Bruenn \\& Dineva 1996; Mezzacappa et al.~1998a) occur and how they affect the explosion. Multi-dimensional hydrodynamic models (Herant et al.\\ 1994; Miller, Wilson, \\& Mayle 1993; Burrows, Hayes, \\& Fryxell 1995; Janka \\& M\\\"uller 1996; Mezzacappa et al.\\ 1998b) have demonstrated the existence and the importance of convective overturn in the neutrino-heating layer behind the supernova shock. Driven by a negative entropy gradient which emerges behind the weakening prompt shock and is enhanced by the neutrino energy deposition, the convective motions transport energy from the region of strongest heating to the shock, thus raising the postshock pressure and pushing the shock farther out. At the same time, cold, low-entropy matter is advected downward where it can readily absorb energy from the upstreaming neutrinos. These hydrodynamic instabilities have a bearing on the measured kick velocities of pulsars (Lyne \\& Lorimer 1994, Cordes \\& Chernoff 1998) and the anisotropies observed in many supernovae. They are essential to understand the production of radioactive elements in the vicinity of the nascent neutron star and their large-scale mixing into the hydrogen and helium layers of the exploding star (Kifonidis et al.\\ 2000). All multi-dimensional simulations have so far been carried out with serious simplifications of the neutrino transport. Even the most advanced spherically symmetric post-bounce models have only employed MGFLD (Bruenn 1993, Bruenn et al.\\ 1995) until recently. The significance of an accurate neutrino transport for the delayed explosion mechanism, however, has long been recognized (Janka 1991, Messer et al.\\ 1998, Yamada, Janka, \\& Suzuki 1999, Burrows et al.\\ 2000). It is therefore a natural step that a new generation of supernova models will employ schemes based on a solution of the Boltzmann equation. In fact, Mezzacappa et al.~(2000) have published results for a 13 M$_{\\odot}$ star which show that a better transport can make a qualitative change to the outcome of the simulations. However, they considered a model with an exceptionally small iron core of 1.17 M$_{\\odot}$ (Nomoto \\& Hashimoto 1988) and the explosion energy was only $0.41 \\times 10^{51}\\,$erg at a post-bounce time of $\\sim 550\\,$ms. The growth rate of this energy of $0.05\\times 10^{51}\\,$erg per 100$\\,$ms cannot easily be extrapolated in time and will probably not increase the explosion energy significantly, because the density around the mass cut drops rapidly and the heating region is evacuated by the developing bifurcation between neutron star and ejecta. In this {\\em Letter} we present results for a Newtonian simulation of a 15 M$_{\\odot}$ star with a 1.28 M$_{\\odot}$ iron core (Woosley \\& Weaver 1995) which show that an accurate neutrino transport does not produce an explosion for this star in spherical symmetry. ", "conclusions": "Our spherically symmetric, Newtonian simulation of a 15 M$_{\\odot}$ star with a 1.28 M$_{\\odot}$ iron core, using a new Boltzmann solver for the neutrino transport, did not give an explosion until 350$\\,$ms after core bounce, although the shock reached a larger maximum radius than in a comparable MGFLD simulation of Bruenn et al.\\ (1995). This is probably explained by stronger neutrino heating of the postshock medium with the more accurate Boltzmann transport. Since both simulations were done with the same progenitor, EoS, and neutrino opacities, and excellent agreement during the core-collapse phase was found, uncertainties due to the different numerics seem to be minimized. Although we have only included $\\nu_e$ and $\\bar\\nu_e$ in our simulation, we consider our conclusions as solid, because muon and tau neutrinos would drain energy from the $\\nu_e$ and $\\bar\\nu_e$ luminosities but contribute to the postshock heating only at an insignificant level due to the lack of charged-current interactions. The main effect of adding pair processes would be a weakening of the early shock propagation by additional energy losses. Also general relativity would probably hamper an explosion (Fryer 1999), but the situation is still ambiguous (De Nisco, Bruenn, \\& Mezzacappa 1998; Baron 1988). The importance of an accurate $\\nu_e$ and $\\bar\\nu_e$ transport is emphasized by the finding that $Y_e > 0.5$ in the region of net neutrino energy deposition. This is interesting because $Y_e \\la 0.48$ was obtained in the neutrino-heated ejecta in supernova models, e.g., by Herant et al.\\ (1994), Burrows et al.\\ (1995) and Janka \\& M\\\"uller (1996), causing a large overproduction of neutron-rich nuclei around $N = 50$ and $A\\approx 90$ ($^{88}$Sr, $^{89}$Y, $^{90}$Zr). This is in conflict with measured galactic abundances. With values $Y_e > 0.5$ this problem disappears (Hoffman et al.\\ 1996). Using their Boltzmann solver for the neutrino transport, Mezzacappa et al.\\ (2000) obtained a successful, but weak explosion in case of a 13 M$_{\\odot}$ progenitor with an extraordinarily small iron core of 1.17 M$_{\\odot}$. For a 15 M$_{\\odot}$ star with a larger core (and therefore most likely also for more massive progenitors), we cannot confirm a qualitative difference from spherically symmetric simulations with MGFLD transport, although we find important quantitative differences with our more accurate neutrino transport. In order to obtain explosions via the neutrino-heating mechanism, multi-dimensional simulations seem indispensable for stars with typical iron core masses. Convection inside the neutron star (Keil et al.\\ 1996) or lower neutrino opacities --- due to suppression relative to the standard description by nucleon correlation effects (e.g., Janka et al.\\ 1996, Burrows \\& Sawyer 1998, Reddy et al.\\ 1999) --- could raise the neutrino emission significantly on the relevant time scale of a few 100$\\,$ms after bounce, and convective overturn in the postshock region has been shown by several groups to support the explosion." }, "0005/astro-ph0005112_arXiv.txt": { "abstract": "We present a measurement of the Hubble Constant based on Cepheid distances to 27 galaxies within 20 Mpc. We take the Cepheid data from published measurements by the Hubble Telescope Key Project on the Distance Scale ($H_0$KP). We calibrate the Cepheid Period-Luminosity (PL) relation with data from over 700 Cepheids in the LMC obtained by the OGLE collaboration; we assume an LMC distance modulus of 18.50 mag ($d_{{\\rm LMC}}=50.1$ kpc). Using this PL calibration we obtain new distances to the $H_0$KP galaxies. We correct the redshifts of these galaxies for peculiar velocities using two distinct velocity field models: the phenomenological model of Tonry \\etal\\ and a model based on the IRAS density field and linear gravitational instability theory. We combine the Cepheid distances with the corrected redshifts for the 27 galaxies to derive $H_0$, the Hubble constant. The results are $H_0 = 85 \\pm 5\\,\\kmsmpc$ (random error) at 95\\% confidence when the IRAS model is used, and $92 \\pm 5\\,\\kmsmpc$ when the phenomenological model is used. The IRAS model is a better fit to the data and the Hubble constant it returns is more reliable. Systematic error stems mainly from LMC distance uncertainty which is not directly addressed by this paper. Our value of $H_0$ is significantly larger than that quoted by the $H_0$KP, $H_0 = 71 \\pm 6\\,\\kmsmpc$. Cepheid recalibration explains $\\sim 30\\%$ of this difference, velocity field analysis accounts for $\\sim 70\\%$. We discuss in detail possible reasons for this discrepancy and future study needed to resolve it. ", "introduction": "A long-standing goal of observational cosmology is the measurement of the expansion rate of the universe, parameterized by the Hubble constant, $H_0.$ Knowledge of $H_0$ enables us to assign galaxies absolute distances $d$ from their redshifts $cz$,using $d=cz/H_0,$ for $z \\ll 1.$ More fundamentally, the Hubble constant measures the time since the Big Bang, or the ``expansion age'' of the universe: $t_0=f(\\omegam,\\omegal) H_0^{-1},$ where $\\omegam$ and $\\omegal$ are the density parameters for mass and the cosmological constant (or ``dark energy''), respectively. The function $f(\\omegam,\\omegal)$ has well known limiting values $f=2/3$ for $\\omegam=1$ and $f=1$ for $\\omegam=0,$ if $\\omegal=0.$ In the flat ($\\omegam+\\omegal=1$) models now favored by CMB anisotropy measurements (e.g., Tegmark \\& Zaldarriaga 2000; Lange \\etal\\ 2000), $f$ is larger, at given $\\omegam,$ than in the $\\omegal=0$ case. However, unless $\\omegam \\simlt 0.25,$ which is disfavored by a variety of data (Primack 2000), $f \\le 1$, even in a flat universe. It follows that for currently acceptable values of the density parameters, the expansion age of the universe is $\\simlt H_0^{-1}.$ By convention, extragalactic distances are measured in Mpc, redshifts in $\\kms,$ and $H_0$ in $\\kmsmpc.$ From the mid-1960s through the mid-1980s, two groups dominated the debate over $H_0.$ One, associated mainly with Sandage and collaborators, argued that $H_0=50\\,\\kmsmpc$ with relatively small ($\\sim 10\\%$) uncertainty. A second, led by de Vaucouleurs, advocated $H_0=100\\,\\kmsmpc$ with similarly small error. The corresponding values of the expansion timescale are $H_0^{-1}=9.8\\,h^{-1}$ Gyr, where $h\\equiv H_0/100\\kmsmpc.$ Thus, the large Hubble constant favored by de Vaucouleurs leads to a ``young'' ($t_0 \\simlt 10$ Gyr) universe, while the small $H_0$ favored by Sandage corresponds to an ``old'' ($t_0 \\simgt 15$ Gyr) universe. In recent years, the debate has shifted, with many groups finding $H_0$ to be intermediate between the Sandage and de Vaucouleurs values. Especially important in this regard has been the work of the Hubble Space Telescope (HST) Key Project on the Extragalactic Distance Scale ($H_0$KP), which finds $H_0=71 \\pm 6\\,\\kmsmpc$ (Mould \\etal\\ 2000). We discuss their methods and results further below. An independent constraint on the age of the universe comes from the age of the oldest stars $t_*$. This can be measured from the turnoff point in the Hertzsprung-Russell diagrams of old globular clusters. The best current estimates (Krauss 1999; see also Caretta \\etal\\ 1999) suggest that $t_* = 12.8 \\pm 1.0$ Gyr ($1\\,\\sigma$ error), and that $10 \\le t_* \\le 17$ Gyr at 95\\% confidence. If one furthermore assumes that the globular clusters did not form until about $\\Delta t \\sim 1$ Gyr after the Big Bang, then the age of the universe as indicated by the oldest stars is $t_* + \\Delta t \\approx 14 \\pm 1$ Gyr. With the above estimates, we thus require that $t_0$ be strictly larger than 10 Gyr, and prefer that it be $\\simgt 13$ Gyr, to ensure consistency of the Big Bang model with stellar ages. From this perspective, the de Vaucouleurs value of $H_0$ yields far too small an expansion time, while the Sandage value produces one that is comfortably large. The current modern value ($H_0$KP) gives a marginally consistent $t_0 = 13.3 \\pm 1.3$ Gyr if we assume an $\\omegam=0.3,$ $\\omegal=0.7$ universe as preferred by a variety of present data. A Hubble constant only 20\\% larger, however, would give an expansion age of 10.6 Gyr, and thus conflict with the best estimates of the globular cluster ages. The determination of the Hubble constant clearly remains a crucial part of the cosmological puzzle. Recent efforts by the $H_0$KP and other groups have greatly reduced the allowed range for $H_0,$ but have not unequivocally demonstrated consistency between the timescales of Big Bang cosmology and stellar evolution. The main purpose of this paper is to underscore the importance of ongoing work on the problem, by presenting an alternative approach, using existing data, to measuring $H_0.$ The outline of this paper is as follows: In \\S 2 we discuss the effects of peculiar velocities on the determination of $H_0,$ and strategies for overcoming these effects. In \\S 3 we present a derivation of the Cepheid PL relation using the Optical Gravitational Lensing Experiment (OGLE; Udalski \\etal\\ 1999a,b) database of LMC Cepheids, and in \\S 4 we apply the new PL relation to the $H_0$KP Cepheid database to obtain distances for the $H_0$KP galaxies. In \\S 5 we constrain the local peculiar velocity field by applying the maximum likelihood \\velmod\\ method to a sample of galaxies with accurate relative distances from surface brightness fluctuations. In \\S 6 we apply the resulting velocity models to the $H_0$KP Cepheid galaxies, and thereby obtain a value of $H_0.$ In \\S 7 we further discuss and summarize our results. ", "conclusions": "We have argued in this paper that $H_0 = 85 \\pm 5\\,\\kmsmpc$ at 95\\% confidence, considering random error only. This result, if correct, leads to an expansion timescale $H_0^{-1} = 10.9$--$12.2$ Gyr, and thus an expansion age $t_0 = f(\\omegam,\\omegal)H_0^{-1}$ that is shorter still (see the discussion in \\S 1), unless $\\omegam = 1-\\omegal \\simlt 0.25.$ For example, for % $H_0=85\\,\\kmsmpc$ and an $\\omegam=0.3,$ $\\omegal=0.7$ cosmology, $t_0 = 11.1$ Gyr. This expansion age may be compared with the estimated age of the oldest globular clusters, $t_* = 12.8 \\pm 1$ Gyr ($1\\,\\sigma$ uncertainty; Krauss 1999). At first blush, then, our estimated Hubble constant leads to a universe younger than its oldest stars. Given this logical contradiction, our result obviously requires further scrutiny. We discuss a number of salient issues in this final section. \\subsection{Why do we disagree with the $H_0$KP?} The $H_0$KP team reported $H_0=71 \\pm 6 \\, \\kmsmpc$ (Mould \\etal\\ 2000). However, of their reported 9\\% ($1\\,\\sigma$) error, approximately 6.5\\% is systematic error due mainly to uncertainty in the distance to the LMC. This systematic error affects our value in precisely the same way as theirs, and thus should not be considered in comparing our $H_0$ estimates. The $1\\,\\sigma$ {\\em random\\/} error in the $H_0$KP Hubble constant is $\\sim 4.4\\,\\kmsmpc,$ corresponding to a $2\\,\\sigma$ error of $\\sim 9\\,\\kmsmpc.$ Thus, the $H_0$KP estimate overlaps with ours only at the very edges of our respective $2\\,\\sigma$ error bars (i.e., at $80\\,\\kmsmpc$). Since we have used the same Cepheid data set to arrive at our estimates, and therefore share much of their random error, this constitutes a significant disagreement. There are two principal causes of this disagreement. The first is the difference in Cepheid calibration as discussed in \\S 3. Our OGLE-based PL calibration produces distances that are smaller by $\\sim 5\\%$ on average, primarily because the OGLE calibration yields larger reddening and thus larger extinction estimates. When applied to the $H_0$KP distances and their procedure for estimating the Hubble constant, the OGLE calibration should bring their value up to $\\sim 75\\,\\kmsmpc,$ closer to the value we derive, though still inconsistent, given that we use the same data. The second source of disagreement is more fundamental: the different strategies we have adopted for determining $H_0.$ The $H_0$KP approach (referred to as ``Method I'' in \\S 2) has been to use the Cepheid galaxies as calibrators for secondary distance indicators (DIs), especially Type Ia Supernovae (SN Ia), the Tully-Fisher and Fundamental Plane relations, and the SBF relation. The secondary DIs are then applied to galaxies at much larger distances than the Cepheid galaxies themselves, typically in the $3000$--$10,000\\,\\kms$ range, where peculiar velocities can be largely neglected. In contrast, we have derived $H_0$ from the Cepheid galaxies themselves, correcting the effects of non-Hubble motions with velocity field models (``Method II'' of \\S 2). The two strategies are subject to different pitfalls. Method I can go awry with the propagation of Cepheid errors into the calibration of the secondary DIs. Such errors might occur for a variety of reasons, the foremost being the difficult nature of the measurements. SN Ia are often historical, i.e., occurred many years or decades ago, and the data on their brightnesses may not be consistent with modern methods. And yet, such historical SN Ia must be used in the calibration procedure, SN Ia being rare events and Cepheid galaxies being few in number. Four out of six SN Ia calibrated by Gibson \\etal\\ (2000) occurred prior to 1990, and two of these occurred prior to 1975. This small sample of calibrators introduces the biggest uncertainty in the SN-based $H_0$ (Suntzeff \\etal\\ 1999). The calibration of the SBF and Fundamental Plane (FP) methods suffer from another problem: Cepheids are found in late-type spiral galaxies, whereas SBF applies best, and FP applies only, to early type galaxies. Consequently, the absolute calibration of the FP relation using Cepheid distances (Kelson \\etal\\ 2000) must be obtained indirectly, by assuming that the Cepheid calibrators and the FP ellipticals are members of a group lying at a common distance. The SBF relation has been calibrated using a small number of spirals with prominent bulges (Ferrarese \\etal\\ 2000a), but possible stellar population differences between spiral bulges and ellipticals make the validity of this calibration uncertain (Tonry \\etal\\ 1997; TBAD00). Neither of the above problems applies to the $H_0$KP calibration of the Tully-Fisher relation by Cepheid galaxies (Sakai \\etal\\ 2000). However, Sakai \\etal\\ obtained their value of $H_0$ from a single $I$ band Tully-Fisher data set, that of Giovanelli and collaborators (e.g., Giovanelli \\etal\\ 1997). Although this data set is of high quality, there are a number of other large Tully-Fisher data sets of recent vintage that were not considered by Sakai \\etal, such as those collected in the the Mark III Catalog (Willick \\etal\\ 1997a). Tully-Fisher measurements are prone to systematic differences in velocity width and photometric measurement conventions, and application of the Cepheid-calibrated Tully-Fisher relation to a wider range of Tully-Fisher data sets is needed; an important first step in this direction has been taken by Tully \\& Pierce (1999). A coincidence of multiple secondary DI miscalibrations is very unlikely, at best. Nevertheless, our Method II analysis does not involve secondary DI's and cannot suffer from the problem of propagated Cepheid calibration errors.\\footnote{It does, of course, potentially suffer from calibration error in the Cepheid PL relation, but this is essentially the problem of the LMC distance.} However, our approach is vulnerable to inaccurately modeled peculiar velocities because we measure $H_0$ locally (see the discussion in \\S 2). Our peculiar velocity models are ``state-of-the-art,'' especially the IRAS models, and we have optimized them with respect to the SBF data set, which is the best current sample for constraining the local velocity field. It is evident from our Hubble diagrams, however, that our velocity models are not perfect. A clear indication of this is the ridge of $\\sim 5$ galaxies that lie well below our best-fit Hubble line in Figures~\\ref{fig:f6a} and~\\ref{fig:f7a}. These are objects that lie within, and in the background of, the Virgo-Ursa Major region, and that are falling in toward Virgo or Ursa Major at higher velocity than predicted by the model. None of these objects deviates from our model at more than the $2\\,\\sigma$ level---indeed, the only $3\\,\\sigma$ deviant point is NGC 1326A, which is above the $H_0=100\\,\\kmsmpc$ line at a distance of 16.4 Mpc---but it is still disquieting to see the large scatter at distances near and beyond Virgo. This scatter does not invalidate our approach, but reminds us that peculiar velocities are not fully accounted for in our model, and that some caution is needed in interpretation. \\subsection{Considerations for future work} The discussion above shows that the debate on the Hubble constant will continue. Its ultimate resolution will require that several key needs are met: \\begin{enumerate} \\item {\\it More nearby galaxies with accurate Cepheid distances}.--- Our local measurement of $H_0$ could be greatly improved with a larger, and more uniformly distributed, sample of Cepheid galaxies. It is to be hoped that further observations by the HST, and later by NGST, will yield such samples. Indeed, such measurements would enable a better grasp of the systematic error in Method II values of $H_0$, along the lines of an angular variance analysis suggested by Turner, Cen \\& Ostriker (1992). \\item {\\it Improved models for the local velocity field}.--- The IRAS model presented here is a reasonable but imperfect fit to both the SBF and the Cepheid distances. Work is currently under way by WNSB to test enhancements of the model. In particular, we intend to further investigate the effects of nonlinear dynamics and nonlinear bias. Another dynamical variable that needs to be better understood is the small-scale velocity noise and its dependence on galaxy density. If these efforts succeed, the uncertainty in $H_0$ due to peculiar velocities will be reduced. \\item {\\it A re-calibration of secondary DIs}.--- As noted in \\S 2, both the ``distant'' and ``local'' strategies for measuring $H_0$ should be pursued, and eventually they should agree. We have pointed to a few issues where the $H_0$KP calibrations of secondary DIs could contain subtle errors. For further investigation, one could use the Cepheid distances to calibrate the Tully-Fisher relations for samples not considered by Sakai \\etal\\ 2000, in particular, those tabulated in the Mark III Catalog (Willick \\etal\\ 1997a), especially after possible calibration errors in that and other catalogs have been corrected via comparison with the uniform, all-sky Shellflow survey recently presented by Courteau \\etal\\ (2000). \\item {\\it Further exploration of the ``Hubble bubble''}.--- The approach of this paper, and, to a lesser extent, that of the $H_0$KP, could overestimate the Hubble constant if the local universe ($d \\simlt 30\\hmpc$) is expanding more rapidly than the global average, as has been suggested by Zehavi \\etal\\ (1998) on the basis of SN Ia data within $\\sim 10,000\\,\\kms.$ Such a situation is certainly possible on theoretical grounds; a fractional mass fluctuation $\\delta_M$ within a sphere of radius $R$ produces a deviation of the Hubble constant within that sphere of $\\delta H_0/H_0 \\approx - \\omegam^{0.6} \\delta_M / 3$ (see Turner, Cen \\& Ostriker 1992 and Tomita 2000 for more detailed theoretical analyses). Typical mass fluctuations on scales $R \\simgt 10\\hmpc$ are $\\simlt 1$ in most cosmological scenarios, so that one would expect that the local value of $H_0$ on a $\\sim 10\\,\\hmpc$ scale to deviate by $\\sim 10$--20\\% from the global value if $\\omegam \\approx 0.3,$ with smaller deviations on larger scales. However, for our local value of $H_0$ to {\\em exceed\\/} the global value, it would be necessary for our local neighborhood to be {\\em underdense\\/} relative to the mean. And yet, the 1.2 Jy IRAS density field suggests that the opposite is true---our local region within 20 Mpc is overdense relative to the volume within 100 Mpc well that is well-sampled by IRAS. Moreover, recent Tully-Fisher data (Dale \\& Giovanelli 2000) do not support the claim of Zehavi \\etal\\ (1998) for a local Hubble bubble. It is prudent to view this issue as an open one for now, and to continue to test the relationship between the local and distant Hubble flow. Such tests do not require DIs that are absolutely calibrated, and thus are independent of the question of $H_0$ itself. \\item {\\it More accurate absolute calibration of the PL relation}.--- As discussed in \\S 2, the largest systematic error in the analysis of this paper is due to uncertainty in the zero point of the Cepheid PL relation, which is itself due to uncertainty in the distance to the LMC. We have adopted the ``canonical'' value $\\mu_{{\\rm LMC}}=18.50$ in this paper, the same as that adopted by the $H_0$KP. Our $H_0$ estimate could be range between $78$ and $98\\,\\kmsmpc$ depending on the distance to the LMC chosen. \\end{enumerate} It is, finally, worth taking a moment to consider the question, What if $H_0$ really is as large as, say, $90\\,\\kmsmpc?$ Would that constitute a ``crisis for Big Bang cosmology,'' a claim that has been heard in some quarters? The answer, for the moment, is clearly ``No,'' for two reasons. First, the globular clusters could still be as young as $t_*=10$ Gyr. Such a young age is unlikely but possible at the few percent level (Krauss 1999). If this were the case, then $t_0 > t_*$ for $H_0=90\\,\\kmsmpc$ and $\\omegam=0.3,$ $\\omegal=0.7.$ One would then require only that the globular clusters formed very shortly ($\\simlt 10^8$ yr) after the Big Bang, which is not impossible. Second, even if $t_*=13$ Gyr is correct, one can obtain $t_0 \\ge t_*$ for $H_0=90\\,\\kmsmpc$ if $\\omegam = 1-\\omegal \\le 0.13.$ A universe of such low density has not been ruled out. In short, a Hubble constant $\\approx 90\\,\\kmsmpc$ does not pose an insurmountable problem for Big Bang cosmology so long as the ages of the oldest stars and the values of the parameters $\\omegam$ and $\\omegal$ remain poorly determined. \\subsection{Summary} We have presented a new determination of the Hubble constant using Cepheid PL data published by the $H_0$KP. Rather than use the nearby ($d \\simlt 20$ Mpc) Cepheid galaxies as calibrators for secondary distance indicators, which are then applied to more distant ($d \\simgt 50$ Mpc) galaxies for which peculiar motions are fractionally small (the $H_0$KP strategy), we use Cepheid galaxies directly to measure $H_0.$ We first redetermined the Cepheid galaxy distances using a calibration of the PL relation derived from a large sample of LMC Cepheids presented by the OGLE group. Our absolute PL calibration assumed $\\mu_{{\\rm LMC}}=18.5.$ (We reemphasize that the $H_0$KP group will shortly present their own revision of Cepheid distances in light of the OGLE LMC Cepheid data [Madore \\& Freedman 2000, in preparation; Freedman \\etal\\ 2000, in preparation].) We then presented two models of the local peculiar velocity field. The first was obtained from the IRAS galaxy density field using the linear relation between large-scale velocity and density fields and the assumption that IRAS galaxies trace the mass density field up to a linear biasing factor $b.$ The IRAS model applies in the Local Group reference frame. The second was the phenomenological model of TBAD00, which applies in the CMB reference frame. The Tonry model assumes the local velocity field is dominated by infall to the Virgo and Great Attractors, along with a dipole and quadrupole term. We optimized each model by fitting it, using the maximum likelihood \\velmod\\ algorithm of WSDK and WS, to a calibration-free 281-galaxy subset of the SBF sample of Tonry \\etal\\ (2001), currently the most accurate set of relative-distances for galaxies in the nearby ($cz \\simlt 3000\\,\\kms$) universe. In the case of the IRAS model, this optimization consisted mainly of determining the value of $\\beta=\\omegam^{0.6}/b,$ which was found to be $0.38 \\pm 0.06$ ($1\\,\\sigma$ error), with $0.2 \\le \\beta \\le 0.5$ allowed at the $3\\,\\sigma$ level. For the Tonry model the optimization involved constraining several parameters that determine the influence of the Virgo and Great Attractors. The velocity model fits used only relative distances for the SBF galaxies and thus in no way prejudiced our subsequent determination of $H_0.$ We then applied the IRAS and Tonry velocity models to 27 Cepheid galaxies, again using the \\velmod\\ algorithm, with the one remaining free parameter now being the Hubble constant. This yielded $H_0=85 \\pm 5\\,\\kmsmpc,$ essentially independent of the value of $\\beta,$ when the IRAS velocity field was used. When the Tonry model was used we obtained $H_0 = 92 \\pm 5\\,\\kmsmpc.$ The quoted random errors are at the 95\\% confidence level. The IRAS model produced a better fit likelihood than the Tonry model, and a Hubble diagram with markedly less scatter. We thus favor the result from IRAS, and adopt the IRAS value of $H_0$ for our final conclusion. This value is significantly larger than the $H_0$KP result, $H_0=71 \\pm 6\\,\\kmsmpc$ (Mould \\etal\\ 2000). Until Method I and Method II analyses give consistent results, we find it untenable to state that $H_0$ is known to within $10 \\%$. We discussed at length in \\S 7.1 several possible reasons for the difference, as well as (\\S 7.2) a number of lines of further investigation needed to clarify the issue. Should the larger value we quote here turn out to be correct, it would be difficult to reconcile the expansion age of the universe, $t_0 = 11.1 $ Gyr for $\\omegam=0.3,$ $\\omegal=0.7,$ with the estimated ages of oldest globular clusters, $12.8 \\pm 1$ Gyr (Krauss 1999). However, the remaining uncertainty in these stellar ages, as well as in the cosmological density parameters, is sufficiently large that a Hubble constant as large or even somewhat larger than what we have argued for here does not yet pose a logical inconsistency for Big Bang cosmology." }, "0005/astro-ph0005324_arXiv.txt": { "abstract": "A new technique has been developed to remove the ionosphere's distorting effects from low frequency VLBI data. By fitting dispersive and non-dispersive components to the phases of multi-frequency data, the ionosphere can be effectively removed from the data without the use of {\\em a priori} calibration information. This technique, along with the new gating capability of the VLBA correlator, was used to perform accurate astrometry on pulsar B0950+08, resulting in a much improved measurement of this pulsar's proper motion ($\\mu_{\\alpha} = -1.6 \\pm 0.4$ mas/yr, $\\mu_{\\delta} = 29.5 \\pm 0.5$ mas/yr) and parallax ($\\pi = 3.6 \\pm 0.3$ mas). This puts the pulsar at a distance of $280 \\pm 25$ parsecs, about twice as far as previous estimates, but in good agreement with models of the electron density in the local bubble. ", "introduction": "Accurate radio pulsar distances are important for population and birthrate modeling and, combined with measurements of dispersion and Faraday rotation, for studies of the ionized interstellar medium and Galactic magnetic fields. Parallax measurements can firmly fix the bottom rungs of the pulsar distance ladder, but parallaxes of even the closest pulsars are only a few milliarcseconds in amplitude, and such measurements are available for only nine pulsars (see Toscano et al.\\ 1999 for a recent list).\\footnote{An optical parallax is also available for the pulsar B0633+17 (Geminga), which has no known radio counterpart.} Four of these parallax measurements were made using timing studies of millisecond pulsars with particularly stable spin periods. The others were made with various VLBI techniques. The use of calibration sources within the primary antenna beam shows promise, as demonstrated by \\citet{fom99}. Observing at L~band with 25 meter telescopes provides a primary beam size of about 30 minutes of arc. Only a small fraction of pulsars have calibrators within this small distance. The number of accessible target pulsars is greatly increased by the use of calibrators at greater offset angles, up to several degrees, but differential ionospheric phase delays make high precision comparison measurements very difficult. Most pulsars are weak radio sources at high frequencies, where array resolution is high and ionospheric effects are minimized. Observations are often made at L~band (1.4 to 1.7 GHz) as a compromise, but even at L~band the ionosphere remains a significant obstacle. We have developed a new calibration technique to remove the effects of the ionosphere from the data and used it to measure a revised parallax of PSR B0950+08. ", "conclusions": "" }, "0005/hep-th0005239_arXiv.txt": { "abstract": " ", "introduction": "\\setcounter{equation}0 Much attention has been paid for the possibility that we are living on a 3-brane in higher dimensional spacetime \\cite{BW1,BW2}. This brane world picture alters the conventional notion of the extra dimensions. Particularly if the bulk is Anti de Sitter (AdS) spacetime, the extra dimensions could be large or even infinite. The action describing the brane world picture is given by \\begin{equation} S= \\frac{1}{2 \\kappa^2}\\int d^5 x \\sqrt{-g} \\left( {\\cal R}^5 + \\frac{12}{l^2} \\right) - \\sigma \\int d^4 x \\sqrt{-g_{brane}} + \\int d^4 x \\sqrt{-g_{brane}} {\\cal L}_{matter}, \\label{0-1} \\end{equation} where ${\\cal R}^5$ is the 5D Ricci scalar, $l$ is the curvature radius of the AdS spacetime and $\\kappa^2=8 \\pi G$ where $G$ is the Newton constant in the 5D spacetime. The brane has tension $\\sigma$ and the induced metric on the brane is denoted as $g_{brane}$. Matter is confined to the 4D brane world and is described by the Lagrangian ${\\cal L}_{matter}$. We will assume $Z_2$ symmetry across the brane. Recently, Randall and Sundrum (RS) constructed a simple model for a brane world \\cite{RS}. They assumed the effect of the matter confined to the brane is negligible compared with that of the surface tension. Their solution is described by the metric; \\begin{equation} ds^2= \\left( \\frac{l}{z} \\right)^2 \\left( dz^2 -d \\tau^2 + \\delta_{ij} dx^i dx^j \\right). \\label{0-2} \\end{equation} It has been shown that the usual 4D gravitational interactions are recovered on the 3-brane. One of the fascinating feature of their model is that the 5D spacetime is not necessarily compactified. In RS model, the 3-brane is Minkowski spacetime. Solutions for homogeneous expanding brane world are obtained by many people [4-14]. It has been shown that the evolution of the universe is identical with that of the conventional 4D cosmology at sufficiently low energies. However in the real world, the universe has inhomogeneity which leads to our structure of the universe [15-17]. This inhomogeneity can be observed today, for example, in the Cosmic Microwave Background Radiation (CMB). Then the cosmological perturbations in the brane world give direct tests for a viability of the brane world idea. In addition, the inhomogeneous fluctuations on the brane could be a powerful observable to probe the existence of the extra dimensions. This is because the inhomogeneous fluctuations on the brane inevitably produce the perturbations of the bulk geometry \\cite{SMS}. The perturbations in the bulk affect the motion of the brane in turn. Then, in general, the dynamics on the brane cannot be separated from the dynamics in the bulk. This could add a new property to the evolution of the cosmological perturbations and could reject the brane world idea. To study the evolution of cosmological perturbations, we should treat the coupled system of brane-bulk dynamics. The problem has the similarity with the dynamics of the domain wall interacting with the gravitational wave, which has been investigated in 4D spacetime \\cite{II}. In our case, the matter on the brane is dynamical. This makes the problem very difficult. We should find a solution for the brane with the cosmological expansion and inhomogeneous fluctuations. The most straightforward way is to solve the 5D Einstein equation, however, it would be difficult to carry out in general. In this paper we propose a new method to deal with the problem. We observe that the brane world cosmology can be constructed by cutting the perturbed AdS spacetime along the suitable slicing and gluing two copies of remaining spacetime. The point is as follows. If we choose a slicing to cut the 5D AdS spacetime, the jump of the extrinsic curvature along the slicing is determined. Because the jump of the extrinsic curvature should be equated with the matter localized on the brane, the matter on the brane is also determined. In other words, a solution for a brane with the given matter can be obtained by finding a suitable slicing. To find the suitable slicing for the given matter, we need two kinds of coordinate transformations. One is a large coordinate transformation which leads to the slicing which determines the background matter. Another is an infinitesimal coordinate transformation which leads to the slicing which determines matter perturbations. The coordinate transformations will be determined by imposing the conditions on the matter such as equation of state. More detailed procedures will be described in the next section. Our main result is \\begin{equation} \\frac{\\delta \\rho}{\\rho}=-2 \\Phi_0 =const. \\label{ee-0} \\end{equation} at superhorizon scales and for late times when the Hubble scale $H$ is sufficiently low ($H \\ll l^{-1}$). Here $\\delta \\rho$ is the density fluctuations and $\\Phi_0$ is the metric perturbations in the longitudinal gauge in the brane world and we assumed the barotropic index of the matter is constant. The point to observe is that the solution (\\ref{ee-0}) is identical with the one obtained in the 4D cosmology. We can also show that the late time evolution of the perturbations agrees with the one obtained in the 4D cosmology at subhorizon scales larger than the AdS curvature scale $l$. The assumption to obtain the above results is that the effect of the massive graviton with mass $m > m_{eff}=l^{-1} e^{\\alpha_0}$ in the brane world is negligible where $e^{\\alpha_0}$ is the scale factor of the brane world. We can understand the fact that the massive graviton with $m > m_{eff}$ can modify the evolution from the following arguments. For late times, the cosmological brane approaches to the RS brane. For the RS brane, the 4D gravity is recovered by the zero-mode of 5D graviton [3,20-23]. The Kaluza-Klein modes give the correction to the 4D gravity. However, in the Anti-de Sitter spacetime, the brane is protected from the Kaluza-Klein modes by the potential barrier which arises from the curvature of the AdS spacetiem. For earlier times, the cosmological brane is located at larger $z$ in the coordinate (\\ref{0-2}). The point is that, for larger $z$, the potential barrier becomes lower. Then the relatively light graviton can interact with the brane. Thus for early times, the 4D cosmology will be susceptible to the Kaluza-Klein modes of 5D graviton. If the brane interacts with the 5D gravitational perturbations, the gravitational waves are inevitably emitted to the 5D bulk. It will cause the modification in the evolution of the perturbations. This picture is consistent with the result that the modes with large $m > m_{eff} =l^{-1} e^{\\alpha_0}$ can modify the evolution because $m_{eff}$ becomes smaller for earlier times. The paper is organized as follows. In section II, we describe our formalism in details and derive the background solution using it. In section III, we calculate the perturbations at superhorizon scales for late times $H < l^{-1}$ using the formalism. In section VI, we calculate the perturbations at subhorizon scales for late times $H < l^{-1}$. It will be shown that the evolution of the perturbations is identical with the one obtained in the 4D cosmology for any scales larger than the AdS curvature scale, if the effect of the massive graviton $m > m_{eff}$ is negligible. In section V we study the effect of the massive graviton $m > m_{eff}$ on the evolution of the perturbations. Finally we discuss the implication of our results on the brane world cosmology. In the Appendix, we listed useful formulas for calculations. ", "conclusions": "\\setcounter{equation}0 The cosmological perturbations in the brane world provide useful tests for the brane world idea. This is because the perturbations in the brane world interact with the perturbations in the bulk which is inherent nature of perturbations in the brane world. The dynamics of the brane can not be separated from the dynamics of the bulk. This is because the inhomogeneous fluctuations on the brane inevitably produces the gravitational waves in the bulk, which in turn affect the evolution of perturbations on the brane. Thus, naively, we think the evolution of the cosmological perturbations is modified significantly. We showed that this is not a case for late time $H < l^{-1}$ and at scales larger than the AdS curvature scale. The metric perturbations become frozen once the perturbations exit the horizon as in the conventionoal 4D cosmology. This result is important for the inflationary scenario in the brane world. If the scale $l^{-1}$ is sufficiently higher than the scales of the inflation, and if heavy graviton modes may be neglected the constancy of the curvature perturbations can be used to estimate the scalar temperature anisotropies of the CMB at large scales. Our results are consistent with those of Ref \\cite{MWBH}, where curvature perturbations on large scales is shown to be conserved, and the density perturbations generated during high-energy inflation on the brane are calculated. The assumption in obtaining the above results is that the effect of massive gravitons with $m >m_{eff}= e^{\\alpha_0}l^{-1}$ can be neglected. The contribution of these modes depends on the initial spectrum of the fluctuations. If the primordial fluctuations are generated during inflation at low energies $H 10^{6-7}$ K). Position-velocity maps show that kpc-scale shells/arcs formed through the nonlinear evolution in a model without stellar energy feedback has similar kinematics to explosional phenomena, such as supernovae. We also find that dense clumps and filamentary structure are formed due to a natural consequence of the non-linear evolution of the multi-phage ISM. Although the ISM in a small scale looks turbulent-like and transient, the global structure of the ISM is quasi-stable. In the quasi-stable phase, the volume filling factor of the hot, warm, cold components are $\\sim 0.2, \\sim 0.6$, and $\\sim 0.2$, respectively. We compare the observations of \\ion{H}{1} and molecular gas of the LMC with the numerically obtained \\ion{H}{1} and CO brightness temperature distribution. The morphology and statistical properties of the numerical \\ion{H}{1} and CO maps are discussed. We find that the cloud mass spectrum of our models represent a power-law shape, but their slopes change between models with and without the stellar energy injection, and also the slope depends on the threshold brightness temperature of CO. ", "introduction": "The topology of the neutral interstellar medium (ISM) can be studied in great detail by the spatial and velocity structures in the neutral \\ion{H}{1} gas. A recent high-resolution \\ion{H}{1} survey of the Large Magellanic Cloud (LMC) reveals that the structure of the neutral atomic inter stellar gas is dominated by numerous holes and shells as well as complex filamentary structure (\\cite{KIM98}). These features are commonly seen in recent high-resolution \\ion{H}{1} images of nearby galaxies obtained with radio synthesis interferometers (\\cite{DD90}; \\cite{P92}; \\cite{SS97}; \\cite{WB99}; \\cite{ST99}). In general, the shell-like and hole structure seen in \\ion{H}{1} has been understood as the cumulative effect of stellar winds from massive stars and supernova explosions evacuating the cool ISM (\\cite{TT88}; \\cite{VDH96}; \\cite{OEY96} ; \\cite{OC97}). However, the extensive study of \\ion{H}{1} shells in the LMC shows that there is a relatively weak correlation between the \\ion{H}{1} shells and the ionized gas traced out by the \\ion{H}{2} regions and \\ion{H}{2} filaments (\\cite{KI99}). Furthermore, the correlation between the \\ion{H}{1} shells and 122 OB stellar associations in the LMC is not very tight (\\cite{KI00}). Rhode et al. (1999) claimed that there is no remnant star clusters at the center of the \\ion{H}{1} holes in Holmberg II, and it is inconsistent with the SNe hypothesis. Moreover an energy source generating the kpc-scale supergiant \\ion{H}{1} holes is a puzzle. These issues raise an interesting question about whether \\ion{H}{1} shells/holes have been formed by the interaction between stars and the ISM or not. Recent hydrodynamical simulations by Wada \\& Norman (1999) demonstrate that a gravitationally and thermally unstable disk, which models an ISM in galaxies, can generate the cold, dense clumps and filaments surrounded by hot, diffuse medium. They show that porous structure is a natural consequence of the non-linear evolution of the ISM. This result strongly suggests that some fraction of \\ion{H}{1} shells, supershells or holes seen in galaxies does not relate to the interaction between stellar activities and the ISM. The other important component of the ISM is the molecular gas. Molecular clouds are potential sites of star formation, but their formation mechanism, and the relationship between their evolution and star formation have been poorly understood. Dense molecular hydrogen is traced by the rotational transition of CO, and a recent high resolution survey of $^{12}$CO (J$=$1-0) in the LMC with NANTEN, which is 4m millimeter-wave telescope at Las Campanas Observatory, established a comprehensive view of the giant molecular clouds in the LMC (\\cite{FU98}). Since the gas clumps, whose size is typically $10-100$ pc, in the simulations by Wada \\& Norman (1999) are dense ($n > 1000$ cm$^{-3}$) and cold ($T=10-100 $ K), they are counterparts of the observed giant molecular clouds in the LMC or other galaxies. Using numerical simulations, Feitzinger et al. (1981) and Gardiner, Turfus, \\& Putman (1998) have studied global dynamics of the ISM and star formation in the LMC. Although the numerical methods they used are different, the stochastic self-propagating star formation model (e.g. Seiden \\& Gerola 1984) in Feitzinger et al. (1981) and sticky particle method in Gardiner, Turfus, \\& Putman (1998), are both phenomenological concerning the structure and dynamics of the ISM and star formation processes. Unfortunately the spatial resolution (100 pc in Feitzinger et al. 1981) and mass resolution (6$\\times 10^4 M_\\odot$, which corresponds to $\\sim 200$ pc for $n\\sim 1$ cm$^{-3}$ and $H \\sim 100$ pc, in Gardiner, Turfus, \\& Putman 1998) in these simulations are not good enough for comparison between models and the recent high resolution observations ($\\sim$ 15 pc for HI (\\cite{KIM98}) and $\\sim $ 40 pc for CO (J$=$1-0) (\\cite{FU98})). In this paper, we apply the numerical scheme used in Wada \\& Norman (1999) to a LMC-type model galaxy, and we conduct two-dimensional hydrodynamical simulations of the multi-phase ISM in a LMC-like model galaxy, taking into account self-gravity of the gas, radiative cooling and various heating processes, such as supernova explosions. High spatial resolution (7.8 pc) and a modern hydrodynamical scheme allow us to model the star formation and its feedback less phenomenologically, and therefore it needs less assumptions than the previous semi-analytic and numerical approaches (e.g. a review by \\cite{SF95}). In contrast to the model used in Wada \\& Norman (1999), the rotation curve assumed here is nearly rigid as suggested in the LMC or other LMC-type dwarf galaxies. We derive the \\ion{H}{1} and CO brightness map from the simulations. Then we compare these simulation results with the \\ion{H}{1} and CO observations of the LMC. In \\S 2, we describe our numerical method and models. In \\S 3, the numerical results with and without star formation are discussed on morphology and statistical structure of the ISM, then they are compared with the observations. Position-Velocity diagrams derived from the numerical results are also discussed. Comparison with past numerical simulations and other implications are discussed in \\S 4, and conclusions are presented in \\S 5. ", "conclusions": "Using high resolution hydrodynamical simulations, we have computed that the global dynamics and structure of the multi-phase ISM in a LMC-type galaxy. Due to gravitational and thermal instability in the gas disk, clumpy fluctuations evolve, and then the clumps merge and form filamentary structure in the non-linear phase. Higher density clumps are formed in the filaments due to the gravitational instability, or collisions between the filaments. Our numerical model with the star formation allows us to provide the evolution of the blastwaves due to supernovae explosions in the rotating, inhomogeneous, multi-phase, and turbulent-like media. We find that the supernova rate in the model with stellar energy feedback is typically of the order of 0.001 yr$^{-1}$ during several hundred Myrs, but fluctuates rapidly (time scale $\\sim$ 10 Myr) by a factor of three or four. The model also shows that kpc-scale low density cavities seen in the observed \\ion{H}{1} map (\\cite{KIM98}) are difficult to be formed under frequent supernovae, but SNe are necessary to form hot bubbles where the gaseous temperature is greater than $10^{6-7}$ K. Our result suggests that there are two possible causes of low density regions in the ISM. The kpc-scale inhomogeneity and arcs can be formed as a natural consequence of non-linear evolution of the multi-phase interstellar medium in a LMC-type galaxy. We find in the PV-diagram of the numerical models that filamentary structure in the non-star forming model, which are caused by the gravitational instability, has similar kinematics to the structure formed by SN explosions. The dense clouds are rotating, and about half of of them show retrograde rotation against the sense of galactic rotation. Using the Monte Carlo radiative transfer code, we have computed \\ion{H}{1} and CO brightness temperature distributions, and compared them with those from the recent observations. The CO cloud mass spectrum in the model with stellar energy feedback is similar to the observed one (\\cite{FU98}), but \\ion{H}{1} distribution function is well fitted by the model without SNe. Therefore we conclude that the small scale structure and dynamics of the ISM in the LMC is mainly affected by the stellar activities, but the gravitational instability significantly contributes to the global morphology and dynamics of the interstellar matter in a kpc-scale." }, "0005/astro-ph0005040_arXiv.txt": { "abstract": "We have applied the MCS image deconvolution algorithm (Magain, Courbin \\& Sohy 1998) to HST/WFPC2 $V$, $I$ data of three M31 bulge globular clusters (G170, G177, and G198) and control fields near each cluster. All three clusters are clearly detected, with an increase in stellar density with decreasing radius from the cluster centers; this is the first time that stars have been resolved in bulge clusters in the inner regions of another galaxy. From the RGB slopes of the clusters and the difference in $I$ magnitude between the HB and the top of the RGB, we conclude that these three clusters all have roughly solar metallicity, in agreement with earlier integrated-light spectroscopic measurements. Our data support a picture whereby the M31 bulge clusters and field stars were born from the same metal-rich gas, early in the galaxy formation. ", "introduction": "Stellar populations are used as tools to probe star formation histories in galaxies. The bulge of our Galaxy, together with that of our Local Group galaxy companion M31, have colors, metallicities, and kinematics which are typical of early type spiral bulges. Although our knowledge about bulges has improved during the last decade, little is known about their formation and evolution, especially in connection with the other components of galaxies (Freeman 1993). There is, however, recent evidence that bulges originate, on rather short time scales, during the very first phases of galaxy formation. For example, in our Galaxy, the analysis of Color-Magnitude Diagrams (CMDs) of star clusters located within five degrees of the Galactic center has revealed a metallicity distribution similar to that of the surrounding Galactic field stars (Barbuy et~al. 1998). In addition, the metal-rich bulge clusters have been demonstrated to have the same age as the inner-halo metal-rich old globular cluster 47~Tucanae (Ortolani et~al. 1995). The globular cluster system of M31, which is about twice as rich as the Galactic one, is among the most studied cluster systems in external galaxies (Harris 1991). However, our knowledge comes mainly from the photometric and/or spectroscopic integrated properties of these clusters. A few pioneering attempts have been made to obtain cluster CMDs from the ground, but even with good seeing at the CFHT, Heasley et~al. (1988) and Christian \\& Heasley (1991) were only able to reach the upper part of the red giant branch, without reaching the horizontal branches of G1 and G219, two of the brightest M31 clusters. With the advent of the Hubble Space Telescope (HST), the situation has changed. The CMDs of 10 globular clusters in M31 have been published, reaching about one magnitude below the horizontal branch (Fusi Pecci et~al. 1996, Ajhar et~al. 1996, Rich et~al. 1996, Holland et~al. 1997). These clusters are mainly located in the halo of M31, where contamination from foreground stars and the M31 stellar disk is minimized. The closest to the center are G108 (with [Fe/H]=$-$0.94) and G280 (with [Fe/H]=$-$0.4), located at 19.2\\arcmin\\ = 4.81 kpc and 18.4\\arcmin\\ = 4.59 kpc from the M31 center, respectively. Recently, Jablonka et al. (1992, 1998) obtained integrated spectrophotometric observations for a sample of globular clusters in the bulge of M31, looking for possible extreme cases of metal enrichment, and consequently naturally investigated the inner regions of M31. The two clusters G170 and G177 belong to their sample. G177 exhibits metallic absorption features which are as strong as those characterizing the central regions of elliptical galaxies; with its higher-than-solar metallicity, this cluster clearly challenged the conventional view of old globular clusters as metal-poor objects, although such metal-rich clusters seem to be rare. Another cluster, G198, as super-metal rich as G177 and at about the same distance from the M31 galactic center, had been identified by Huchra et al. (1991) in an independent spectroscopic analysis. G170, located slightly further away has about a solar metallicity. As a first attempt to learn more about the outstanding properties of extremely metal-rich globular clusters, we applied for deep imaging observations with the HST, taking advantage of the high spatial resolution and red sensitivity of the Planetary Camera of WFPC2, in order to build the CMDs of these three M31 bulge clusters, all of them located within about six arcmin from the center of this galaxy. These cluster data, so far the closest to the center of M31, allow insight into the questions of metallicities, age and link between bulge field and cluster stellar populations. Given their central location and consequently their very high stellar density, they challenge the best observational capabilities. G170, being slightly further away from the center, is surrounded by a lower density of field stars than the other two clusters, and was meant to be a relative calibrator. A fourth cluster, G1, has been observed during the same program; its properties, viz. CMD, structural parameters, dynamical mass estimate, and M/L ratio, will be addressed in another paper (Meylan et al. 2000). The structure of this paper is as follows: Section~2 presents the observations and data reduction; Section~3 describes the image deconvolution technique used; Section~4 analyses and discusses the results, while Section~5 summarizes the highlights of this study. ", "conclusions": "We have applied the MCS deconvolution algorithm to three M31 bulge globular clusters. This deconvolution appears as a very efficient method in such extremely dense regions, but still rather demanding in terms of computing time, so we have been limited to small regions. However, our initial results are already very promising. Our radial star counts for the cluster and control fields clearly show that we have resolved the clusters: this is the first time that cluster stars have been resolved in the bulge of another galaxy so close to the center. From the deconvolved photometry, we clearly see the cluster red giant branches, though we do not detect their horizontal branches. Taking into account systematic biases introduced by crowding, there are no significant differences between the CMDs for the three clusters. Moreover, there is no apparent difference between the cluster and their respective control field CMDs. From the RGB slopes and the difference in $I$ magnitude between the HB and RGB, all three clusters have roughly solar metallicity, making them similar to NGC~6528 or NGC~6553 in our Galaxy; by inference, the same is true of the mean field population. We thus conclude that the M31 bulge clusters and field stars originate from the same material. They are representative of old, metal-rich populations. Similar results have been found in our Galaxy, where it becomes more and more apparent that the metallicity distributions of field stars and globular clusters in the bulge are identical (Ortolani et al. 1995; Barbuy et al. 1998). From a detailed study of the element ratios for two stars in NGC 6553, Barbuy et al. (1999) conclude that the Galactic bulge underwent rapid star formation and chemical enrichment. Certainly our work corroborates this view. Bulges appear to be old, metal-rich systems, similar in many respects to elliptical galaxies. There is mounting evidence that most stars in ellipticals and bulges formed at high redshift z$_f$ $>$ 3 (see review by Renzini \\& Cimmati 1999). Unfortunately, we have pushed HST to its limits, and M31 seems to be the furthest galaxy for which we can use HST in this way. Higher spatial resolution studies of bulge populations in M31 and more distant galaxies awaits adaptive optics on 8--10m ground-based telescopes, and eventually the NGST." }, "0005/astro-ph0005510_arXiv.txt": { "abstract": "As a part of the ISOPHOT CIRB (Cosmic Infrared Background Radiation) project we have searched for point-like sources in eight fields mapped at two or three wavelengths between 90$\\mu$m and 180$\\mu$m. Most of the 55 sources detected are suspected to be extragalactic and cannot be associated with previously known objects. It is probable, also from the far-infrared (FIR) spectral energy distributions, that dust-enshrouded, distant galaxies form a significant fraction of the sources. We present a tentative list of new extragalactic FIR-sources and discuss the uncertainties involved in the process of extracting point sources from the ISOPHOT maps. Based on the analyzed data we estimate the number density of extragalactic sources at wavelengths 90$\\mu$m, 150$\\mu$m and 180$\\mu$m and at flux density levels down to 100\\,mJy to be 1$\\cdot$10$^5$\\,sr$^{-1}$, 2$\\cdot$10$^5$\\,sr$^{-1}$, and 3$\\cdot$10$^5$\\,sr$^{-1}$, respectively. Strong galaxy evolution models are in best agreement with our results, although the number of detections exceeds most model predictions. No-evolution models can be rejected at a high confidence level. Comparison with COBE results indicates that the detected sources correspond to $\\ga$20\\% of the extragalactic background light at 90$\\mu$m. At longer wavelengths the corresponding fraction is $\\sim$10\\%. ", "introduction": "The cosmic infrared background (CIRB) consists in the far-infrared of the integrated light of all galaxies along the line of sight plus any contributions by intergalactic gas and dust, photon-photon interactions ($\\gamma$-ray vs. CMB) and by hypothetical decaying relic particles. A large fraction of the energy released in the universe since the recombination epoch is expected to be contained in the CIRB. An important aspect is the balance between the UV-optical and the infrared backgrounds: what is lost by dust obscuration will re-appear through dust emission in the CIRB. Some central, but still largely open, astrophysical problems to be addressed through CIRB measurements include the formation and early evolution of galaxies, and the star formation history of the universe. The primary goal of the ISOPHOT CIRB project is the determination of the flux level of the FIR CIRB. The other goals are the measurement of its spatial fluctuations and the detection of the bright end of FIR point source population contributing to the CIRB. The full analysis of the data from the DIRBE (Hauser et al. \\cite{hauser}; Schlegel et al.\\cite{schlegel98}) and FIRAS (Fixsen et al. \\cite{fixsen98}) experiments indicated a CIRB at a surprisingly high level of $\\sim$1 MJy\\,sr$^{-1}$ between 100 and 240 $\\mu$m. Preliminary results had been obtained already by Puget et al. (\\cite{puget}). Lagache et al. (\\cite{lagache99a}) detected a component of Galactic dust emission associated with warm ionized medium and the removal of this component lead to a CIRB level of 0.7\\,MJy\\,sr$^{-1}$ at 140$\\mu$m. Because of the great importance of the FIR CIRB for cosmology these results definitely require confirmation by independent measurements. ISOPHOT observation technique is different from COBE: (1) with relatively small f.o.v. ISOPHOT is capable of looking at the darkest spots between the cirrus clouds; (2) ISOPHOT has good sensitivity in the important FIR window at 120 -- 200 $\\mu$m; (3) with the good spatial and spectral sampling ISOPHOT gives the possibility of recognizing and eliminating the emission of galactic cirrus. In the ISOPHOT CIRB project we have mapped four low-cirrus regions at high galactic latitude at the wavelengths of 90, 150, and 180 $\\mu$m (see Fig.~\\ref{fig:allsky}). Through this multi-wavelength mapping we will try to separate the cirrus component and confirm the detection of sources at neighboring wavelengths. In addition, we have performed absolute photometry in several filters between 3.6 -- 200 $\\mu$m at the darkest spots of the fields. This photometry will be used (1) to secure the zero point for the maps at 90, 150, and 180 $\\mu$m, and (2) to determine the contribution by the zodiacal emission using measurements of its SED at mid-IR wavelengths where it dominates the sky brightness. This paper presents the first step in the analysis of the ISOPHOT CIRB observations. Here we will concentrate on the data reduction and the study of the point sources (galaxies) found in the FIR maps. The source counts determined in the FIR are important for the study of the star formation history of the universe and for the testing of the current models of galaxy evolution. With recent observations at infrared and sub-mm wavelengths it has become obvious that star formation efficiencies derived from optical and UV observations only (e.g. Madau et al. \\cite{madau}; Steidel et al. \\cite{steidel}, \\cite{steidel99}; Cowie et al. \\cite{cowie96}, \\cite{cowie97}, \\cite{cowie98}; Hu et al. \\cite{hu}) underestimate the true star formation activity at high redshifts because the correction for dust extinction is unknown (e.g. Heckman et al. \\cite{heckman}). IRAS has shown that in the local universe about one third of the luminosity is emitted at infrared wavelengths. In starburst galaxies the fraction can be much higher as most of the starlight is absorbed by dust and re-radiated in the infrared. In extreme sources like the hyperluminous galaxy $IRAS$\\,10214+4724 the energy spectrum peaks around 100$\\mu$m in the rest frame and more than 90\\% of the energy is emitted in the infrared and sub-mm regions. The emission maximum moves further towards sub-mm with increasing redshift, causing optical studies to seriously underestimate the true star formation activity. If the dust content is high enough the objects can remain completely undetected at optical wavelengths. With ISO and new sub-mm instruments like the SCUBA bolometer array (Holland et al. \\cite{holland99}) it has become possible to study the star formation history of the universe at infrared to sub-mm wavelengths (for reviews see e.g. Hughes et al. \\cite{hughes3}, \\cite{hughes1}). Due to the negative $K$-correction the observed flux densities will not depend strongly on the redshift and it is possible to detect more distant galaxies (e.g. van der Werf \\cite{werf}; Guiderdoni et al. \\cite{guiderdoni97}). Recent studies (e.g. Dunlop et al. \\cite{dunlop}; Omont et al. \\cite{omont}; Hughes et al. \\cite{hughes97}, \\cite{hughes3}; Stiavelli et al. \\cite{stiavelli99}; Abraham et al. \\cite{abraham99}; Lilly et al. \\cite{lilly99}; Blain et al. \\cite{blain99}) have shown that star formation activity remains high at $z>$1. In observations with the SCUBA instrument at 450$\\mu$m and 850$\\mu$m (e.g. Hughes et al. \\cite{hughes3}; Barger et al. \\cite{barger1}; Smail et al. \\cite{smail97}; Blain et al. \\cite{blain99}; Eales et al. \\cite{eales99}; Lilly et al. \\cite{lilly99}) galaxies have been detected up to redshifts $z$$\\sim$5. Compared with galaxies seen in optical surveys the objects have higher dust content and the star formation rates are an order of magnitude higher. The surface density of the detected sources exceeds predictions of no-evolution models by at least one order of magnitude (Smail et al. \\cite{smail97}; Eales et al. \\cite{eales99}; Barger et al. \\cite{barger99}). The number of sources detected by Eales et al. (\\cite{eales99}) at 850$\\mu$m above $\\sim$3\\,mJy accounts for $\\sim$20\\% of the CIRB detected by FIRAS (Fixsen et al. \\cite{fixsen98}). Similar results were obtained by Barger et al. (\\cite{barger99}). At the level of 0.5\\,mJy the sources contain most of the sub-mm CIRB ( Smail et al. \\cite{smail97}, \\cite{smail98}; Blain et al. \\cite{blain99b}). Kawara et al. (\\cite{kawara}) observerved the Lockman Hole at 95$\\mu$m and 175$\\mu$m using ISOPHOT. The number of sources found was at least three times higher than predicted by no-evolution models. The conclusions of the FIRBACK (Puget et al. \\cite{puget99}) and ELAIS (Oliver et al. \\cite{ringberg}) projects are similar and at 175$\\mu$m sources with $S_{\\nu}$$>$120\\,mJy account for $\\sim$10\\% of the CIRB detected by FIRAS. In this article we study the number density of extragalactic sources and their contribution to the FIR background radiation using observations made with ISOPHOT. The data consist of maps made at wavelengths 90$\\mu$m, 150$\\mu$m and 180$\\mu$m, and for some smaller areas at 120$\\mu$m. The total area is close to 1.5 square degrees. Most of the regions have been observed at three wavelengths (90$\\mu$m, 150$\\mu$m, 180$\\mu$m) some at two wavelengths (120$\\mu$m and 180$\\mu$m). Both the galactic foreground cirrus emission and the emission from typical extragalactic objects will reach their maxima within or near the observed wavelength range. In particular, we will be able to determine the cirrus spectrum for each region separately. We have developed a point source extraction routine based on the fitting of the detector footprint to spatial data. The method is different from those used in most previous studies where the source detection algorithms have concentrated on the variations (off-on-off) of the detector signal as function of time. Our analysis will therefore be independent of and complementary to the previous results. ", "conclusions": "\\label{sect:conclusions} We have searched for FIR point sources in raster maps observed with the ISOPHOT C100 and C200 detectors at wavelengths between 90$\\mu$m and 180$\\mu$m. The total area covered is $\\sim$1.5\\,square degrees. Most of the FIR sources detected are presumably IR galaxies which, due to the negative $K$-correction, can be observed at redshifts $z>1$. A comparison of the SEDs of sources detected at 90, 150, and 180$\\mu$m with cirrus spectra shows that for most sources an explanation in terms of cirrus knots can be excluded. Based on the number counts of the sources we can conclude: \\begin{itemize} \\item We have found 55 FIR sources that, due to the multi-wavelength confirmation, correspond to detections with high confidence level ($\\ga$4$\\sigma$) \\item We have derived a FIR source density of $\\sim$60 sources per $\\sq\\degr$ at 100\\,mJy level \\item The source density is much higher than predicted by no-evolution galaxy models; at 180$\\mu$m the excess is close to a factor of five \\item The source counts are in agreement with models where the star formation rate and the relative number of ULIGs increases strongly with $z$, e.g. the counts are slightly higher than predicted by model E of Guiderdoni et al. (\\cite{guiderdoni98}) \\item At 150$\\mu$m and 180$\\mu$m the combined flux of detected sources accounts for $\\sim$10\\% of the CIRB intensity as derived from the COBE FIRAS data; at 90$\\mu$m the fraction is over 20\\% \\end{itemize}" }, "0005/astro-ph0005383_arXiv.txt": { "abstract": "We have used the VLBA at 5 GHz to observe all galaxies with nuclear radio flux densities above 3.5 mJy found in a VLA survey at 15 GHz of a sample of nearby LINER galaxies. All galaxies were detected revealing high brightness temperature ($T_{b} \\ga 10^8$ K) radio sources. Free-free emission is unlikely since it greatly overpredicts the soft X-ray luminosities. We infer the presence of AGN-like, non-thermal radio emission most likely powered by under-fed black holes. Together with our VLA sample we estimate from our observations that at least half of LINER galaxies host genuine AGN. We find no evidence for highly inverted radio cores as predicted in the ADAF model: the (non-simultaneous) spectral indices are on average around $\\alpha=0.0$. In the two brightest sources we detect some extended emission, which appears to originate in jets in at least one of these galaxies. Together with the spectral indices this suggests that the nuclear emission at centimeter radio waves is largely dominated by emission from radio jets, very similar to the situation in more luminous AGN. The energy released in these jets could be a significant fraction of the energy budget in the accretion flow. ", "introduction": "The evidence for supermassive black holes in the nuclei of most galaxies has become much stronger recently. Some of the best cases are the Milky Way (Eckart \\& Genzel 1997), NGC 4258 (Miyoshi et al. 1995), and a number of other nearby galaxies (Richstone et al. 1998) where convincing dynamical evidence for black holes exists. In quasars and radio galaxies their existence is commonly inferred from the huge energy output of the active galactic nucleus (AGN) which is probably powered by accretion onto the black hole. However, despite the alleged presence of black holes in both cases, there is a huge span in luminosity between weakly active galaxies like the Milky Way and AGN. The question of how these central engines are related to each other and why they appear so different despite being powered by the same type of object is therefore of major interest. For many nearby galaxies with low luminosity nuclear emission lines, it is not even clear whether they are powered by an AGN or by star formation. This is especially true for Low Ionization Nuclear Emission Region (LINER) galaxies (Heckman 1980), some of which can be explained in terms of aging starbursts (Alonso-Herrero et al.~2000). One of the best ways to probe the very inner parts of these engines is to study the compact radio sources found in many AGN. Indeed, despite their low optical luminosity, quite a few nearby galaxies have such radio sources in their nuclei (e.g. Jones et al.~1981), prominent cases in spiral galaxies being the Milky Way (Sgr A*) and M~81 (see Bietenholz et al. 1996). In addition some relatively nearby elliptical galaxies such as M87 and NGC~1275 appear as low-power FR\\,I radio galaxies and also contain well-known compact radio cores (Cohen et al.~1969; Schilizzi et al.~1975). These radio sources resemble the cores of radio-loud quasars, showing a very high brightness temperature and a flat to inverted radio spectrum that extends up to sub-mm wavelengths. Models proposed for these low luminosity radio nuclei are either a scaled AGN model, in which the core is the synchrotron self-absorbed base of a radio jet coupled to an underluminous accretion disk (Falcke, Mannheim, \\& Biermann 1993; Falcke 1996; Falcke \\& Biermann 1999) or an advection-dominated accretion flow (ADAF; Narayan et al.~1998; see also Melia 1992; Fabian \\& Rees 1995). Earlier surveys have shown that E and S0 galaxies often have compact, flat-spectrum radio sources in their nuclei (Wrobel \\& Heeschen 1984, 1991; Sadler et al.~1989; Slee et al.~1994). Some of the most prominent flat-spectrum nuclear radio sources in nearby galaxies are found in galaxies with LINER nuclear spectra (O'Connell \\& Dressel 1978), but so far there has been no comprehensive study of radio nuclei in a significant sample of LINER galaxies, which make up the majority of galaxies with low-level nuclear activity. We have, therefore, recently conducted a survey of LINER galaxies with the Very Large Array (VLA; Thompson et al. 1980) in its A configuration at 15 GHz (resolution $\\sim$0\\farcs15) to search for compact radio emission (Nagar et al.~2000, Paper I). The sources were drawn from the extensive and sensitive spectroscopic study of a complete, magnitude-limited sample of 486 nearby galaxies (Ho, Filippenko, \\& Sargent 1995), one third of which showed LINER-like activity (Ho, Filippenko, \\& Sargent 1997). From those active galaxies with a LINER spectrum a subsample of 48 bright sources was drawn with no well-defined selection criterion other than that they had been observed with other telescopes as well, e.g. ROSAT, the HST (UV imaging, Maoz et al. 1996, Barth et al. 1998), and the VLA at 15 GHz in A configuration (Nagar et al. 2000) and at 1.4 and 8.4 GHz in A and B configuration (van Dyk \\& Ho 1997). The sample also included so called transition objects which have spectra intermediate between LINER and \\ion{H}{2} region galaxies. While the project was being conducted a few sources in the original sample were re-classified as low-luminosity Seyfert galaxies. However, only one out of the ten sources discussed here has a ``pure'' Seyfert spectrum. The 15~GHz VLA survey found a surprisingly large number (15 out of 48) of galaxies with compact radio cores and flat spectral indices. Here we present Very Long Baseline Array (VLBA; Napier et al. 1994) observations of the eleven brightest of these galaxies to investigate the central region of LINER galaxies at the sub-parsec scale and clarify the nature of their radio cores. ", "conclusions": "Our result has a number of interesting implications. The presence of high brightness temperature radio cores in our LINER sample confirms the presence of AGN-like activity in these galaxies. It is unlikely that the radio sources represent free-free emission, as has been claimed for example in NGC~1068 (Gallimore, Baum, O'Dea 1997), since a much higher soft X-ray luminosity than is typically observed in low-luminosity AGN would result. The emission coefficient for thermal bremsstrahlung from a gas at temperature $T$ is (e.g. Longair 1992, eq.~3.43) \\begin{eqnarray} \\epsilon_\\nu~&=& 6.8 \\times 10^{-51} Z^2 T^{{\\small{-\\onehalf}}} N_p N_e\\nonumber\\\\ &&\\cdot\t g(\\nu, T) exp(- h\\nu/kT)\\; {\\rm W m^{-3} Hz}^{-1} \\end{eqnarray} where $g(\\nu, T)$ is the Gaunt factor. If we consider a plasma at temperature $T\\simeq10^8$ K, then at 5~GHz, the exponential factor in eqn. (2) is 1, while the Gaunt factor is $\\sim$12. In the soft X-ray regime, taken as 0.4~keV to 2~keV, the Gaunt factor varies between 1.7 and 0.8, respectively, while the exponential factor in eq. (2) varies between 0.95 and 0.8, respectively. Therefore, the luminosity, per Hertz, at 0.4~keV and 2~keV is $\\sim$0.25 times and $\\sim$0.05 times that at 5~GHz, respectively. The geometric mean monochromatic luminosities of the nuclei observed by us is 10$^{27.5\\pm0.6}$ erg sec$^{-1}$ Hz$^{-1}$ at 5~GHz. If this emission traces thermal bremsstrahlung we would expect the total 0.4--2~keV luminosity of these nuclei to be 10$^{43.9}$ erg s$^{-1}$. However, the observed 0.4-2~keV luminosities for low-luminosity AGN tend to be of the order of 10$^{39-40}$ erg s$^{-1}$ (e.g. Ptak et al. 1999)---many orders of magnitude lower and thus rendering a thermal origin of the radio emission very unlikely. Of course photoelectric absorption could attenuate some of the soft X-ray emission, however, since seven of our galaxies show broad H$\\alpha$ emission (spectral type 1.9) the absorption should only be moderate. To make this argument more watertight one will need to investigate multiwavelength data for our galaxies on a case-by-case basis. On the other hand, the compact, flat-spectrum cores we have found are similar to those typically produced in many AGN. Hence we can take the presence of compact, non-thermal radio emission as good evidence for the presence of an AGN in our galaxies. The 100\\% detection rate with the VLBA, based on our selection of flat-spectrum cores found in a 15 GHz VLA survey, shows that for statistical purposes we could have relied on the VLA alone for identification of these compact, high brightness radio sources. Hence, with 15 GHz VLA surveys of nearby galaxies one has an efficient tool for identifying low-luminosity AGN. This complements other methods for identifying AGN, such as searching for broad emission-lines or hard X-rays, and has the advantage of not being affected by obscuration. If we only consider galaxies with a LINER spectrum, we found at least eleven flat-spectrum radio cores at 15 GHz in a sub-sample of 24 LINERs observed by Nagar et al.~(2000). Eight of these eleven LINERs are included in our sample here, yielding a lower limit to the AGN fraction for LINERs of at least $33\\pm12$\\% (8/24). Based on our 100\\% detection rate of these flat-spectrum cores with the VLBA, we can, however, argue that all eleven flat spectrum sources found in the VLA study are likely to be AGN, raising the AGN fraction of LINERs to at least $46\\pm14$\\% (11/24). These ratios do not change significantly if we include the galaxies classified as Seyferts. Since the selection of our parent sample is not very well defined, we could still be subject to an unquantifiable bias. This can be minimized by studying the radio emission of a distance-limited sample, which we plan in a future paper. First results (Falcke et al.~1999) seem to indicate that the bias is not large. The two brightest radio sources in our sample show extended structure suggestive of jet-like outflows, and the other seven sources are unresolved or slightly resolved. Our very limited dynamic range is not good enough to prove or exclude the presence of jets for the latter. Moreover, VLBA observations of M81 (Bietenholz et al.~2000) have shown that jets in low-luminosity AGN can be very compact and difficult to detect. The only clue we therefore have is the spectrum which is flat or slightly inverted. Such a spectrum is obtained in jet models (Blandford \\& K\\\"onigl 1979; Falcke 1996; Falcke \\& Biermann 1999), where the spectral index ranges from $\\alpha=0.0$ to 0.23 as a function of inclination angle to the line-of-sight. In no case do we find a spectral index as high as $\\alpha=0.4$ as predicted in the ADAF model (Yi \\& Boughn 1998). This does not necessarily exclude the ADAF model, but argues for the parsec scale radio emission at centimeter radio waves being dominated by another component, such as a radio jet or a wind. A combination of an underluminous disk or an ADAF and a radio jet is one possibility (e.g.~Donea, Falcke, \\& Biermann 1999). Assuming the cores are produced by randomly oriented, maximally efficient jets from supermassive black holes (of order $10^8 M_\\odot$) we can use eq. (20) of Falcke \\& Biermann (1999) to calculate that for an average monochromatic luminosity of $10^{27.5}$ erg sec$^{-1}$ Hz$^{-1}$ at 5 GHz the jets would require a minimum {\\em total} jet power of order $Q_{\\rm jet}\\ga10^{42.5}$ erg sec$^{-1}$. Compared to quasars this is a rather low value and supports the conclusion, based on their low UV and emission line luminosities, that the cores are powered by under-fed black holes. On the other hand this jet power is well within the range of the bolometric luminosity of typical low-luminosity AGN ($10^{41-43}$ erg sec$^{-1}$; Ho 1999) and, compared to radiation, jets could be a significant energy loss channel for the accretion flow." }, "0005/astro-ph0005127_arXiv.txt": { "abstract": "The excellent soft X-ray sensitivity of the PSPC detector onboard the \\rosat satellite provided the first chance to study precisely the spectral and timing properties of Narrow-Line Seyfert~1 galaxies. \\rosat observations of Narrow-Line Seyfert~1 galaxies have revealed (1) the existence of a giant soft X-ray excess, (2) a striking, clear correlation between the strength of the soft X-ray excess emission and the FWHM of the \\Hb line, (3) the general absence of significant soft X-ray absorption by neutral hydrogen above the Galactic column, (4) short doubling time scales down to about 1000 seconds, (5) the existence of persistent giant (above a factor of 10), and rapid (less than 1 day) X-ray variability in extragalactic sources. The soft X-ray results on Narrow-Line Seyfert~1 galaxies indicate that their black hole regions are directly visible, further supporting the Seyfert~1 nature of these objects. The extreme X-ray properties of Narrow-Line Seyfert~1 galaxies make them ideal objects for understanding many of the problems raised generally by the Seyfert phenomenon. ", "introduction": "This paper reviews the important \\rosat contributions to the field of Narrow-Line Seyfert~1 (NLS1) research. The excellent soft X-ray sensitivity of the PSPC detector (Pfeffermann et al. 1987) onboard the \\rosat satellite (Tr\\\"umper 1983), the \\rosat All-Sky Survey data (Voges et al. 1999), and deep \\rosat PSPC and HRI pointed observations provided the best opportunities to study NLS1s before the launches of \\xmm and {\\it Chandra}. \\rosat observations have triggered the rapid growth in the definition of the phenomenological parameters of NLS1s throughout the electromagnetic spectrum, as well as the theoretical modeling of their exciting properties. This is clearly demonstrated with the papers presented in these proceedings. Fifteen years after the definition of the peculiar optical properties of NLS1s (see the review article of Pogge), followed by a period in which their importance had been suggested (e.g. Halpern \\& Oke 1987; Stephens 1989; Puchnarewicz et al. 1992), NLS1s now represent an important class of the AGN family, holding many keys to our understanding of the problems posed by the Seyfert phenomenon. ", "conclusions": "Great progress in defining the observational properties of NLS1s has been achieved based on \\rosat and \\asca observations. The observations have shown many NLS1s to have characteristic, unique and extreme X-ray properties. These include the strongest soft X-ray excess emission seen in Seyfert~1 galaxies, steep 2--10~keV power-law continua, and extremely rapid and large-amplitude X-ray variability. The extreme properties of NLS1s have additionally stimulated the theoretical modeling of many aspects of Seyfert activity." }, "0005/astro-ph0005311_arXiv.txt": { "abstract": "Blue compact dwarf galaxies (BCDGs) are currently undergoing strong bursts of star formation. Nevertheless, only a few of them have been clearly detected in CO, which is thought to trace the \"fuel\" of star formation: H$_2$. In this paper we present a deep search for CO $J=1\\rightarrow0$ and $J=2\\rightarrow1$ emission lines in a sample of 8 BCDGs and two companions. Only 2 of them (Haro\\,2 and UM\\,465) are detected. For the other galaxies we have obtained more stringent upper limits on the CO luminosity than published values. We could not confirm the previously reported ``detection'' of CO for the galaxies UM\\,456 and UM\\,462. We analyze a possible relation between metallicity, CO luminosity, and absolute blue magnitude of the galaxies. We use previously determined relations between $X\\equiv N$(H$_2)/I_{\\rm CO}$ and the metallicity to derive molecular cloud masses or upper limits for them. With these ``global'' $X_{\\rm CO}$-values we find that for those galaxies which we detect in CO, the molecular gas mass is similar to the \\hi\\ mass, whereas for the non-detections, the upper limits on the molecular gas masses are significantly lower than the \\hi\\ mass. Using an LVG (Large Velocity Gradient) model we show that $X_{\\rm CO}$ depends not only on metallicity but also on other physical parameters such as, volume density and kinetic temperature, which rises the question on the validity of ``global'' $X_{\\rm CO}$-factors. ", "introduction": "} A particular class of dwarf galaxies named Blue Compact Dwarf Galaxies (BCDGs, Sargent \\& Searle, \\shortcite{sargent:searle70}) has seen increasing interest among astrophysicists because of their extreme current star forming activity which is in contrast to their apparent ``youth'' in terms of chemical evolution. BCDGs represent about 5\\% of all dwarfs \\cite{salzer89}, \\cite{sage:etal92} and are among the smallest star forming galactic systems known. \\par One of their outstanding properties is that their optical spectra are dominated by lines characteristic of \\hii\\ regions, which is the reason why they are frequently termed ``\\hii\\ galaxies''. From optical spectroscopy we know that many \\hii\\ galaxies have low heavy element abundances, typically down by a factor of three up to more than twenty compared to the solar neighbourhood % \\cite{kunth:oestlin2000}. It quickly became clear that these objects must form stars in what is called a burst, otherwise the observed star formation rate would be in conflict with their total gas masses as derived from \\hi\\ observations \\cite{thuan:martin81}. Such a burst may last some $10^8$\\,yrs, with a time span between bursts of the order of 10$^9$ yrs. It has been suggested that interaction with companions might trigger their star formation \\cite{brinks90}, but Taylor et al. \\shortcite{taylor:etal95} found that only about 60\\% of \\hii\\ galaxies have companions, often with masses about $\\frac{1}{10}$ of the main galaxy. One of the most interesting and important issues which has not been settled so far is the molecular gas content of these galaxies. Molecular hydrogen is believed to be the preponderant seed for star formation, so it is a natural assumption that large amounts of H$_2$ should be present in BCDGs. Yet the results have been anything but conclusive so far. Following early attempts to detect the CO line in BCDGs \\cite{tacconi:young84}, there have been a number of observations to confirm or reject those inconclusive measurements (e.g. Sage et al. \\shortcite{sage:etal92}, hereafter SSLH; Gondhalekar et al. \\shortcite{gondhalekar:etal98}; Taylor, Kobulnicky \\& Skillman \\shortcite{taylor:etal98}, hereafter TKS). Surprisingly, the results remained partially contradictory, as for instance in the case of II~Zw~40 (Arnault et al. \\shortcite {arnault:etal88}; SSLH), although improved instrumentation had been involved. Prompted by the difficulty to detect the CO line -- relied upon as a good tracer of molecular hydrogen content -- in BCDGs, some of the pertinent publications prematurely concluded that molecular gas is deficient in these systems. However, part of the difficulties to detect CO might have been due to beam filling and sensivitity problems. Taking e.g. 30\\,Dor in the LMC as a template giant star-forming complex, it is clear that, if placed at some larger distance and covered by the beam, it could have escaped detection, as CO brightness is rather low there, due to strong photo-dissociation in the high interstellar radiation field \\cite{cohen:etal88}. The same might be true for BCDGs. In this case, high-sensitivity mapping might reveal previously undetected CO emission. We have therefore conducted a search for CO in gas-rich (based on \\hi) \\hii\\ galaxies, using the IRAM 30\\,m telescope. In contrast to previous projects (e.g. SSLH, Gondhalekar et al. \\shortcite{gondhalekar:etal98}), our observations not only consisted of single pointings towards the brightest position in the galaxies, but involved mapping a number of positions in them, to detect possible gas concentrations away from photodissociation regions. Obtaining sensitive upper limits to the CO luminosity in these systems is as much a goal of this study as detecting emission. In Section 2 % we present details of our observations. In Section 3 % we present our data, and compare it with previous results. In the subsequent section (Sec. 4) % we discuss possible causes for the detections and non-detections. This section is divided into three subsections: Section \\ref{gas} deals with the physical conditions of the gas derived from a LVG model; in Section \\ref{metlum} we analyze the relationship between metallicity and CO luminosity; finally, in Section \\ref{factor} we discuss the $X_{\\rm CO}$ factor problem, which has been heatedly debated in the past years and has not yet found a clear resolution. Our conclusions are presented in the last section (Sec. 5). ", "conclusions": "} We have searched for emission from the $^{12}$CO ($J=1\\rightarrow0$ and $J=2\\rightarrow1$) transitions in 10 dwarf galaxies, 8 of which are BCDGs and 2 are the companions of one of these. We detected CO in 2 of them (Haro\\,2 and UM\\,465) and found it to be extended in both galaxies. Although we mapped part of the other galaxies, we were unable to detected CO. We obtained very stringent upper limits. We could not confirm the ``marginal detection'' of CO in UM\\,456 and UM\\,462 previously reported by SSLH. The observed line ratios of the $2\\rightarrow1$ to $1\\rightarrow0$ transitions are not very sensitive to changes in the kinetic temperature. Modelling the ratio with a simple LVG code helps only to exclude low densities. Higher CO transitions and/or observations of CO isotopomers will help to get more stringent limits on these physical parameters. We could not find any simple relation between metallicity and CO luminosity. Molecular gas masses for the galaxies are derived assuming the relation between $X_{\\rm CO}$ and metallicity given by \\cite{arimoto:etal96}. We find that for those galaxies detected in the CO lines the molecular gas mass is comparable to the \\hi\\ mass, whereas for those galaxies undetected in CO the \\hi\\ mass is significantly larger than the limits on the molecular gas mass. Even in the sources where CO has not been detected, we do not argue against the presence of H$_2$. While it is certainly possible that in the extreme environment of a BCDG not just CO but also H$_2$ is destroyed, at least in regions close to young massive stellar clusters, a picture in which a large amount of H$_2$ exists without CO is attractive. Sensitive observations of \\ci\\ and \\cii\\ in these galaxies would thus be desirable in the future to shed light on this issue." }, "0005/astro-ph0005582_arXiv.txt": { "abstract": "The classical Blandford \\& Payne (1982) model for the magnetocentrifugal acceleration and collimation of a disk-wind is revisited and refined. In the original model, the gas is cold and the solution is everywhere subfast magnetosonic. In the present model the plasma has a finite temperature and the self-consistent solution of the MHD equations starts with a subslow magnetosonic speed which subsequently crosses all critical points, at the slow magnetosonic, Alfv\\'en and fast magnetosonic separatrix surfaces. The superfast magnetosonic solution thus satisfies MHD causality. Downstream of the fast magnetosonic critical point the poloidal streamlines overfocus towards the axis and the solution is terminated. The validity of the model to disk winds associated with young stellar objects is briefly discussed. \\vskip 0.5 true cm ", "introduction": "Astrophysical jets are systematically associated with the presence of an underlying accretion disk, both observationally and theoretically (see K\\\"{o}nigl \\& Pudritz 2000 for a recent review). In the case of protostellar objects, accretion disks are resolved by means of infrared and millimeter surveys and interferometric mappings down to scales of a few tens of AU. In the optical and the near infrared, HST high resolution images of disks in several jet sources have also been obtained (Padgett et al. 1999). With an apparent relation found between accretion and ejection in the form of a strong correlation between outflow signatures and accretion diagnostics (see e.g. Cabrit et al. 1990, Cabrit \\& Andr\\'e 1991, Hartigan et al. 1995), stellar jets seem to be powered by the gravitational energy released in the accretion process. These facts and considerations have led several authors to develop models of disk winds. The pioneering work of Bardeen \\& Berger (1978) on a hydrodynamic radially self-similar model of a hot galactic wind was generalized in the seminal paper of Blandford \\& Payne (1982, henceforth BP82) by including a rotating magnetic field. In particular, in BP82 it was shown that a cold plasma can be launched magneto-centrifugally from a Keplerian disk, similarly to a bead on a wire, provided that the magnetic field lines are sufficiently inclined from the axis. Since then, steady and axisymmetric MHD models, self-similar in the radial direction, have been successfully analyzed and generalized in the literature (see e.g. Contopoulos \\& Lovelace 1994, henceforth CL94, Li 1995, 1996, Ferreira 1997, Ostriker 1997, Vlahakis \\& Tsinganos 1998, henceforth VT98, Lery et al. 1999). A major problem is however still open on the validity of the various classes of radially self-similar solutions analyzed so far. Because, as it is well known since the original work of Weber \\& Davis (1967) on the rotating magnetized solar wind in the equatorial region, acceptable outflowing solutions must cross smoothly all singularities related to the characteristic speeds of the MHD perturbations, i.e., the poloidal Alfv\\'en velocity and the slow/fast magnetosonic velocities. However, in radially self-similar equations the critical points are not found where the poloidal speed of the flow is equal to the characteristic velocities of these magnetosonic waves. In the cold model of BP82 the ``modified'' fast magnetosonic critical point (where $t=1$ in the BP82 notation) is found downstream of the position where the poloidal velocity of the wind is equal to the fast magnetosonic velocity. Subsequently it has been shown that this is a general property of the axisymmetric steady MHD equations: the singularities of the equations coincide with the positions of the limiting characteristics, or separatrices, within the hyperbolic domain of the governing equations (Bogovalov 1994, Tsinganos et al. 1996). In particular, Bogovalov (1994, 1996) pointed out the key role played by the singularity occurring at the fast magnetosonic separatrix surface (FMSS). Namely, the asymptotic region of the jet is causally disconnected from the base of the flow, only for solutions that cross the critical point at the FMSS. This means that every terminal perturbation or shock does not affect the outflow structure upstream of the position of this critical point. And, Tsinganos et al. (1996) have given several analytical examples where the true singularities of the equations do not coincide with the positions where the governing partial differential MHD equations change character from elliptic to hyperbolic and vice versa. For the sake of simplicity from now on we shall indicate by `fast/slow magnetosonic singularity', or in short 'modified fast/slow', the critical points at the FMSS/SMSS. It turns out that in none of the previous models of disk-winds a solution has been found to cross the FMSS. For example, Li (1995, 1996) and Ferreira (1997), starting from the accretion disk, succeeded to cross the slow magnetosonic and the Alfv\\'en ones, but downwind turning points were found where the solutions terminate. Such solutions can be connected to infinity only through a shock, as suggested by Gomez de Castro \\& Pudritz (1993). However in this case, as the wind velocity is subfast magnetosonic, a temporal evolution of the outflow is expected (Ouyed \\& Pudritz 1997). Cylindrically collimated solutions were found by Ostriker (1997) for a cold plasma, integrating the MHD system upstream from infinity and crossing the Alfv\\'en singularity, but always in the subfast magnetosonic regime. On the other hand, it has been shown that in collimated winds oscillations of streamlines are a common feature (Vlahakis \\& Tsinganos 1997). It thus seems that cylindrically collimated solutions without oscillations correspond to a rather particular choice of parameters that completely suppresses such oscillations. A slight change in these parameters induces the onset of oscillations which increase in amplitude until the configuration is destroyed (Vlahakis 1998). Since the Ostriker (1997) solutions are asymptotically subfast magnetosonic they are likely to be sensitive to perturbations from the external medium, unlike solutions that really satisfy all the criticality conditions. Therefore, such solutions are likely to be structurally and topologically unstable (Vlahakis 1998). However, it has been shown by Contopoulos (1995) that, in the restricted case of a purely toroidal magnetic field, a smooth crossing of the FMSS is possible. On the other hand in such a case an asymptotically cylindrically collimated configuration is not found; in fact, a new transition to subfast magnetosonic velocities must occur anyway for radially self-similar winds. The only way out is then to match the superfast magnetosonic solution with a shock which is in this case in the physically disconnected domain. In the present study we extend the analysis of BP82, CL94 and Contopoulos (1995) showing that an exact and simultaneous smooth crossing of \\underbar{all} three MHD critical surfaces is possible. In Sec. 2 we define the equations of the hot wind in the framework of a radially self-similar approach and outline the numerical technique. In Sec. 3 we explore the solution topologies in the region around and particularly downstream of the FMSS, where the solution terminates, while in Sec. 4 are shown the features of a few solutions crossing all three critical points with conditions similar to those of BP82. Finally, in Sec. 5 we discuss the possible astrophysical applications of these solutions to stellar jets, and summarize the main implications of our results in comparison with previous ones obtained by other authors. ", "conclusions": "Before discussing the main physical implications of our results, also in connection with those obtained by other authors, we show that the present solutions are suitable to describe the physical properties of astrophysical outflows. \\subsection{Astrophysical applications} The modeling of a particular astrophysical outflow requires first the calculation of all physical quantities from the non dimensional parameters characterizing the particular model. We will address here this question of calculating some observable quantities of disk-winds associated with protostellar objects from the parameters of our model. We deduce first the ratios of some characteristic speeds at the disk level, keeping in mind that from the numerical results we have obtained $M_o \\sim 0.01$ and $G_o \\sim 0.1$. We will refer in the following mainly to the solutions with $x=0.75$. {\\it First}, the ratio of the poloidal Alfv\\'en and Keplerian speeds at the disk level is: \\begin{equation}\\label{rho_o} \\left( \\frac{V_{Ap}}{V_{\\rm Kep}} \\right)_o = 0.316 \\times \\left( \\frac{M_o}{0.01} \\right) \\left(\\frac{G_o}{0.1}\\right)^{-3/2} \\frac{1}{\\kappa \\sin\\psi_o} \\approx 0.178 \\,. \\end{equation} The poloidal magnetic field which is essential in the launching of the outflow is anchored in the disk and its energy density is less than the rotational kinetic energy density of the disk. Thus, the field is rather weak to brake the rotation of the plasma at the disk and it is carried passively around by azimuthal rotation. {\\it Second}, the ratio of the sound and initial speeds at the disk level is: \\begin{equation}\\label{CsVo} \\left(\\frac{C_s}{V_o}\\right)_o = 70.7 \\times \\left({\\gamma \\mu}\\right)^{1/2} \\left(\\frac{G_o}{0.1}\\right)^{2}\\frac{10^{-4}}{M_o^{(\\gamma+1)}} \\approx 94 \\,, \\end{equation} where $V_o=V_z (z=0)$. The initial ejection speed is negligible in comparison to the thermal speed at the disk, a situation similar to a thermally driven wind. {\\it Next}, the ratio of the sound and Keplerian speeds at the disk level is: \\begin{equation}\\label{CsVk} \\left(\\frac{C_s}{V_{\\rm Kep}}\\right)_o = 0.22 \\times (\\gamma \\mu)^{1/2}\\left(\\frac{G_o}{0.1}\\right)^{1/2}\\frac{1}{ \\kappa M_o^{(\\gamma -1)}} \\approx 0.314 \\,. \\end{equation} We notice that the Keplerian speed is about 3 times higher than the thermal speed at the disk. Thus, thermal effects cannot inhibit the rotation of the disk. {\\it Finally}, the ratio of the Keplerian and initial speeds at the disk level is: \\begin{equation}\\label{VkVo} \\left(\\frac{V_{\\rm Kep}}{V_o}\\right)_o = 316 \\times \\kappa \\left(\\frac{G_o}{0.1}\\right)^{3/2}\\left(\\frac{10^{-2}}{M_o}\\right)^2 \\approx 300 \\,, \\end{equation} i.e., the initial speed is negligible in comparison to the Keplerian speed. In our case the flow speed at the fast critical point is about $10^3$ the initial speed $V_o$. In agreement with the observations we can reasonably assume the terminal speed of the outflow to be $\\sim 400$ km s$^{-1}$, such that its velocity at the base is $V_o = $ 0.4 km s$^{-1}$. In principle, radially self-similar models do not have an intrinsic scale length; however from the previous estimate of the initial speed one allows to calculate the footpoint of the reference fieldline $\\alpha=1$ on the disk. In units of 10 solar radii this cylindrical distance $\\varpi_o$ is: \\begin{equation}\\label{varpi_o} \\frac{\\varpi_o}{10 R_{\\odot}} = 0.19 \\times \\left(\\frac{M_o}{0.01}\\right)^4\\left(\\frac{0.1}{G_{o}}\\right)^3 \\frac{1}{\\kappa^2}\\frac{\\cal M}{\\cal M_{\\odot}} \\left(\\frac{V_o}{{\\rm km} \\, {\\rm s}^{-1}}\\right)^{-2}. \\end{equation} Hence, for a one solar mass star we get $\\varpi_{o} \\approx 12.5 R_{\\odot}$. It is also interesting to calculate the mass-loss rate $\\dot {\\cal M}_w$ in units of $10^{-8} {\\cal M}_{\\odot} \\, {\\rm yr}^{-1}$ : \\begin{eqnarray}\\label{mdot} \\begin{array}{l} \\displaystyle \\frac{\\dot {\\cal M}_w}{10^{-8}{\\cal M_{\\odot}} \\,{\\rm yr}^{-1}} = 0.0386 \\times \\left ( \\displaystyle \\frac{M_o}{0.01}\\right)^2 \\left(\\frac{B_{z,o}}{10 G}\\right)^2 \\times \\\\ \\\\ \\left( \\displaystyle \\frac{\\varpi_o}{10 R_{\\odot}}\\right)^{2} \\left ( \\displaystyle \\frac{V_o}{\\rm km \\,{\\rm s}^{-1} }\\right)^{-1} f(\\alpha_{out}, \\alpha_{in} ) \\,, \\end{array} \\end{eqnarray} where \\begin{equation}\\label{mdot1} f(\\alpha_{out}, \\alpha_{in} ) = \\frac{\\alpha_{out}^{x-3/4}-\\alpha_{in}^{x-3/4}}{x-3/4} \\qquad {\\rm if} \\quad x\\neq 0.75 \\,, \\end{equation} and \\begin{equation}\\label{mdot2} f(\\alpha_{out}, \\alpha_{in} ) = \\ln \\frac{\\alpha_{out}}{\\alpha_{in}} \\qquad {\\rm if} \\quad x= 0.75 \\,. \\end{equation} By assuming $\\varpi_{in} = \\varpi_{o}$, $\\varpi_{out} \\approx 10 \\varpi_{o}$ and $B_{z,o}= 8$ G we have ${\\dot {\\cal M}_w}/(10^{-8} {\\cal M_{\\odot}} \\, {\\rm yr}^{-1}) \\approx 1$, with a temperature at the base of the flow of: \\begin{equation}\\label{T_o} T_{o,in} = 3 \\times 10^5 \\mu\\left(\\frac{G_o}{0.1}\\right)^4\\frac{10^{-8}} {M_o^{2(\\gamma+1)}} \\left( \\frac{V_o}{{\\rm km}\\, {\\rm s}^{-1}} \\right)^2 \\approx 8\\times 10^4 \\,{\\rm K} \\,, \\end{equation} \\begin{equation}\\label{T_o,out} T_{o,out} = T_{o,in} \\frac{\\varpi_{in}}{\\varpi_{out}} \\approx 8\\times 10^3 \\,{\\rm K} \\,. \\end{equation} We remind that $T_o$ is not the temperature of the disk as we have assumed a transition layer between the disk surface and the base of the flow (see Sec. 2). This region could be reasonably related to a corona heated by dissipative processes in the plasma (e.g. magnetic reconnection, ohmic heating, etc.; see e.g. K\\\"onigl \\& Pudritz 2000). As the flow corotates roughly up to the Alfv\\'en point (Fig. 5) the specific angular momentum carried by the wind is $\\dot J_w = \\dot {\\cal M}_w \\Omega \\varpi_\\alpha^2$ while the angular momentum that has to be extracted locally at the foot point $\\varpi_o$ of the fieldline in order that the disk accretes at a rate $\\dot {\\cal M}_a$ is $\\dot J_a= (1/2) \\Omega \\varpi_o^2 \\dot {\\cal M}_a$ (Spruit 1996). If the angular momentum carried by the wind is a fraction $f$ of $\\dot J_a $ while $1-f$ is the fraction carried away by viscous stresses, then the ratio of the mass fluxes in the wind and in the accretion flow is $$ {\\dot {\\cal M}_w \\over \\dot {\\cal M}_a} = {f\\over 2} {\\varpi_o^2 \\over \\varpi_\\alpha^2} \\lesssim 0.015 \\,, $$ taking into account that $ \\varpi_\\alpha = 5.8 \\varpi_o$ for model I. It follows that the rate of the outflowing mass is at most of the order of 1$\\%$ of the rate of the accreted mass; and this is achieved when the wind carries all the angular momentum of the accreted mass. When the outflow carries a smaller fraction of the angular momentum of the disk, the mass loss rate in the wind is an even smaller fraction of the mass loss rate in the wind. In other words, the mass loss rate in the wind is a negligible fraction of the accreted mass, despite that the jet may carry most of the angular momentum of the accreted mass. Similar results are obtained for the case $x=0.7525$. Therefore, from the above arguments we may conclude that from our solutions we deduce for the physical parameters values in reasonable agreement with those observed in this class of objects. Our solution terminates at $z/ \\varpi_o \\approx 2 \\times 10^4$, i.e., at $\\approx 400$ Astronomical Units (AU) from the central star. At this position we could argue that there exists a shock matching the solution with the outermost region of the outflow (Gomez de Castro \\& Pudritz 1993). It is well known that bright knots are observed on scales of thousands AU along most protostellar jets. These configurations are shocks that are interpreted as originated either by fluid instabilities on the jet surface or by temporal variations in the velocity of the outflow (Burrows et al. 1996, Ray 1996, 1998, Micono et al. 1998, K\\\"onigl \\& Pudritz 2000). It could be reasonable to associate the terminal shock of our solutions with the inner knots, found at distances down to $\\approx 100$ AU from the star. However we cannot ignore that these knots are non steady configurations and move outwards with velocities $\\sim 100 \\div 200$ km s$^{-1}$ (Ray 1996). We could assume that the shock is well upstream of the optical knots: polarimetric radio data on the T Tauri object are consistent with the presence of a shock at $\\approx 20 \\div 40$ AU from the star (Ray et al. 1997). Alternatively the terminal shock could indeed be located approximately at the positions of the inner knots, but there the flow looses both self-similarity and steadiness. However as we are in the superfast magnetosonic regime, the upwind configuration is not affected. Only a much more detailed parametric study will be able to test these two possibilities. \\subsection{Physical properties of the critical solutions} The solutions of this model, in particular Fig. 7, illustrate nicely the physical process of transferring electromagnetic Poynting energy flux and enthalpy to directed kinetic energy flux of the flow in order to accelerate a disk wind and then form a jet along the symmetry axis of the system. Thus, the analysis of the previous section is interesting in the sense that it reveals the driving mechanisms of the outflow. The poloidal kinetic energy is negligible at the disk level. It then increases rather sharply up to the region of the SMSS and Alfv\\'en surfaces. This increase is at the expense of both, the enthalpy and the electromagnetic Poynting energy flux (see, Fig. 7). The poloidal velocity is directed basically in the radial direction (Figs. 5 and 6), i.e., here part of the random thermal energy together with a part of the electromagnetic energy are mostly transformed to directed wind expansion. Downstream of the Alfv\\'en surface it is mainly the Poynting energy flux that is effectively transformed into kinetic energy directed along the rotational axis, till the FMSS is encountered. After the FMSS, the flow has already reached the maximum speed available from the total energy E, which is also approximately equal to the initial electromagnetic Poynting energy flux. Then, the acceleration asymptotically stops. Despite the fact that most of the acceleration to high speeds is apparently of magnetic origin, the role of the polytropic index and thus of the initial thermal acceleration may not be negligible, in particular in the region before the SMSS. For example, in the case of Fig. 4 where the flow is exactly adiabatic and $\\gamma=5/3$, the critical solution achieves only a very small axial component of the velocity which is twice the axial velocity on the equatorial plane. In the quasi-isothermal case of models I and II where $\\gamma=1.05$, the maximum velocity is 1000, higher than the equatorial one (Fig. 5). As a matter of fact, this last case is closer to the one analyzed in Li (1995) and Ferreira (1997) where the gas is isothermal up to the first critical surface and then it is taken to be cold afterwards, wherein the pressure has sufficiently dropped. However, another possibility is that the low terminal speed obtained in the adiabatic case of Fig. 4 could be due to the lower value of the rotation parameter $\\lambda^2$ which is $\\sim 2.8$ in the adiabatic case of Fig. 4, as opposed to values $\\sim 137$ and 136 in models I and II and similarly for the case examined in Ferreira (1997). When the gas has reached a high speed along the $z$-axis, its inertia causes it to lag behind the rotation of the field line and the field is wound up, as shown in Fig. 6, resulting to a highly twisted magnetic field. Consequently, the strong curvature force of this predominantly azimuthal magnetic field towards the $z$-axis, causes the poloidal field to collimate. Initially the field is flaring away from the rotation axis but the curvature force bends the poloidal field lines toward the rotation axis. The azimuthal velocity peaks around the Alfv\\'en point which is at a height $z= 3.5$ and a cylindrical distance $\\varpi=5.8$ times the starting distance $\\varpi_o$ in model I. Beyond the Alfv\\'en point the rotation drops in accordance to angular momentum conservation and thus the centrifugal force becomes negligible. Then, the strong inwards curvature force of the twisted field, wins, over the weak outwards centrifugal force and gas pressure gradient with the result that the lines are bent and eventually collapse towards the rotation axis. It is interesting that this feature of the collapse of the outflow towards the rotation axis which appears in cold models (BP82) and models that do not cross the FMSS (Li 1995, Ferreira 1997), is also preserved in the present hot model where also all critical points are crossed. This result seems to indicate the rather dominant role of the magnetic hoop stress in radially self-similar models, contrary to what happens in meridionally self-similar models wherein the structure becomes asymptotically cylindrical (Trussoni et al. 1997, Sauty et al. 1999, Vlahakis \\& Tsinganos 1999). It is worth to clarify for a moment the term ``disk-wind'' that we used in this study. By that term we simply intend to indicate that we describe an outflow from a disk-like structure accreting onto a central gravitational object. Thus, the flow starts at some angle $\\theta_o$ above or at the equatorial plane of the disk, as opposed to a ``stellar'' wind flow that starts radially above or at a spherical or quasi-spherical source. Needless to say that a consistent solution of the accreting part of the flow would be required for a consistent solution of the inflow-outflow structure in the case of a disk-wind. However, such a complete undertaking is beyond the scope of the present paper which only intends to emphasize the possibility to construct complete steady self-similar solutions for the wind crossing all critical points. To make such a connection between the disk and the outflow, in the spirit of BP82, Li (1995) and Ferreira (1997), the first step would be to see how our parameters may fall into the range of parameters considered by those models. For that purpose, in Eqs. (\\ref{BPk}) - (\\ref{BPxii}) we have made a correspondance between our parameters and those used by BP82. Thus, in the ``standard\" solution of BP82 the parameters are: $\\kappa_{\\rm BP}= 0.03$, $\\lambda_{\\rm BP} = 30$ and $\\xi'_o = 1.58$ corresponding to a launching angle of the jet at the disk $\\psi_o \\approx 32^{\\circ}< 60^{\\circ}$. In our case, we find $\\kappa_{\\rm BP}\\approx 0.13$, $\\lambda_{\\rm BP}\\approx 14.57$, for both, model I and model II. We also have ${\\xi'_o}_{\\rm BP}=\\cot \\psi_o=0.425$ ($\\psi_o = 67^{\\circ}$) for model I and ${\\xi'_o}_{\\rm BP}=\\cot \\psi_o=0.675$ ($\\psi_o = 56^{\\circ}$) for model II, in the BP82 notation. We note that the values of $\\kappa_{\\rm BP}, \\lambda_{\\rm BP}$ are close in BP82 and the present model. However, the value of the launching angle $\\psi_o$ is $> 60^{\\circ}$ in our model I because of the additional thermal driving of the outflow at the disk level, contrary to the cold model of BP82 where $\\psi_o \\approx 32^{\\circ} < 60^{\\circ}$. In summary, our models I and II occupy in the space of $\\kappa_{\\rm BP}$ and $\\lambda_{\\rm BP}$, roughly the same domain as in BP82 (cf. Fig. 2 in BP82). The only difference is in the value of the launching angle $\\psi_o$ which can be larger in the present hot model, as expected. These values are within the range of the allowed parameters in the ($\\kappa_{\\rm BP}$, $\\lambda_{\\rm BP}$) space also in the analysis of Li (1995, cf. Fig. 3) provided that the magnetic diffusivity is of order one. Note also that model II with $x=0.7525$ corresponds to an ejection index in the notation of Ferreira (1997) $\\xi=2x-3/2=0.005$. \\subsection{Summary} In this paper we have extended the classical work of Blandford and Payne (1982), mainly by showing via examples for the first time that a solution passing through all MHD critical points can indeed be constructed. As is well known, the FMSS plays the role of the MHD signal horizon such that in an outflow crossing this MHD horizon all perturbations which the outflow may encounter are convected downstream by the superfast outflow and so the steady state solution is maintained. In other words, the outflow interior to the FMSS is causally disconnected and protected against any conditions it may encounter in the interstellar or intergalactic medium towards which the jet propagates after it is launched by magnetocentrifugal forces from the surface of an accretion disk. Unlike other analytical models which produce asymptotically cylindrically collimated outflows (Sauty \\& Tsinganos 1994, Trussoni et al. 1997, VT98, Sauty et al. 1999, Vlahakis \\& Tsinganos 1999), this class of radially self-similar models cannot continue to infinity but it has to be stopped downstream of the FMSS and matched via a MHD shock to a subfast outflow that mixes with the interstellar medium (Gomez de Castro \\& Pudritz 1993). This shock can connect the present solutions to some breeze, subAlfv\\'en or subslow magnetosonic branch perhaps also preserving the self-similarity. Thus, the main difference here with previous results presented in the literature is that the asymptotic part of the present solutions is causally disconnected from the source and hence any perturbation downstream of the superfast transition cannot affect the whole structure of the steady outflow. This task of matching the present solutions with a downstream shock however remains a challenge for future studies, together with a (time-consuming) more extended parametric analysis and also a correct matching of the ideal MHD outflow solutions with an inflow in a non-ideal accretion disk (Ferreira 1997)." }, "0005/astro-ph0005257_arXiv.txt": { "abstract": "We examine the observational consequences of the two possible origins for irregular galaxies: formation from collapse of a primordial cloud of gas early in the age of the Universe, and formation from tidal tails in an interaction that could have occured any time in the history of the Universe. Because the formation from tidal tails could have occurred a long time ago, proximity to larger galaxies is not sufficient to distinguish tidal dwarfs from traditional dwarfs. We consider the effects of little or no dark matter on rotation speeds and the Tully-Fisher relationship, the metallicity-luminosity relationship, structure, and stellar populations. From these selection criteria, we identify a small list of dwarf irregular galaxies that are candidates for having formed as tidal dwarfs. ", "introduction": "Models of interacting galaxies have shown that tidal forces in the interaction can produce long tails of stars and gas that have been pulled out of the interacting galaxies (Toomre \\& Toomre 1972). Zwicky (1956) pointed out the possibility that self-gravitating objects might develop in these tidal tails that could then evolve into small galaxies. Since then, concentrations of stars and gas that are probable ``tidal dwarfs'' in the making have been observed at the tips of tidal tails in several interacting systems (for example, the Antennae system; Mirabel \\et\\ 1992). Numerical modelling confirms that bound, galaxy-sized clumps can form in tidal tails (Barnes \\& Hernquist 1992; Elmegreen, Kaufman, \\& Thomasson 1993). These tidal dwarfs, once the tidal tail itself has dispersed and the parent galaxies have moved off, could bear a striking resemblance to small, independent, Im-type galaxies (Schweizer 1974). The tidal dwarfs are small, gas-rich, morphologically disorganized, and already have on-going star formation (see also Mirabel, Lutz, \\& Maza 1991). Furthermore, the properties measured for tidal dwarfs are well within the range of properties seen for normal, relatively isolated irregular galaxies (Mirabel, Dottori, \\& Lutz 1992; Duc \\& Mirabel 1994; Hibbard \\et\\ 1994; Hunter 1997). Gravitational interactions are an on-going process in the Universe that began when galaxies themselves first formed. Therefore, this mechanism for forming irregular galaxies has been taking place for the age of the Universe. Hunsberger, Charlton, \\& Zaritsky (1996), for example, estimate that as many as one-half of the current dwarf galaxy population of compact groups may have been formed from the interactions of giant spiral galaxies. The formation of dwarf irregulars in compact groups is accelerated because of the increased crowdedness and potential for interactions there. However, interactions can and do occur outside of compact groups of galaxies as well. One must then consider that any given dwarf irregular galaxy, including field galaxies, could have been formed in one of two ways: traditional formation from collapse of a primordial cloud of gas early in the age of the Universe, and tidal dwarf formation from an interaction of larger galaxies at any time during the history of the Universe. Because the time scale since the formation of a tidal dwarf can be large, a tidal dwarf could appear to be relatively isolated if the formation took place many Gyrs ago. Because the formation mechanism of traditional dwarfs and of tidal dwarfs are different, some key characteristics of these two groups of galaxies could also be different, as outlined by Barnes \\& Hernquist (1992) and Elmegreen \\et\\ (1993). In this paper we examine a sample of irregular galaxies with these observational differences in mind and ask whether any seemingly normal Im galaxy might be a candidate for an old tidal dwarf. The observational characteristics of Im galaxies and distinctions with tidal dwarfs are too imprecise at this time to do more than point out candidate tidal dwarf systems, but it is a way to begin thinking about the issue that not all irregulars may have had the same initial conditions. Dwarf galaxies, galaxies that are intrinsically not luminous, come in a variety of types, including spirals, irregulars, ellipticals, and spheroidals, and Gerola, Carnevali, \\& Salpeter (1983) suggested galaxy interactions as a means for forming dwarf ellipticals. However, the most common type of dwarf is a gas-rich Im galaxy, and we will primarily concentrate on irregulars in this discussion. ", "conclusions": "We have examined properties of a sample of irregular galaxies from the perspective of features that might distinguish galaxies formed in tidal interactions at some time shorter than a Hubble time from those formed from collapse of a primordial gas cloud a Hubble time ago. We have considered the lack of dark matter predicted by models as manifested in rotation speeds and the Tully-Fisher relationship, the fact that tidal dwarfs may have formed from enriched material, structure, and peculiar stellar populations. However, using these features to identify old tidal dwarfs is currently imprecise. Abundances and star formation histories are entangled in other evolutionary and observational effects, and not enough is known about the amount and location of dark matter and the true structure of irregulars. Nevertheless, we have identified candidates for old tidal dwarfs, and they are listed in Table \\ref{tablist}. We have also listed the distance to the nearest large galaxy. A little over one-quarter of the galaxies in this list are in the Local Group. Eighty-five percent of the galaxies are within 0.5 Mpc of a large galaxy; and one lies near supergiant gas streamers wrapped around a nearby Im galaxy. Because of the difficulties in identifying old tidal dwarfs, these galaxies can only be considered candidates at this point. In addition this is not an exhaustive list, and we have not included representative samples of other groups of dwarfs including dwarf ellipticals and dwarf spheroidals. The peculiar galaxy IZw18, for example, has the peculiar stellar populations and complex kinematics that might make it a candidate. Clearly, it is important to understand the formation and evolutionary processes of the most common galaxy in the universe: irregular and other dwarf galaxies. The fact that irregulars could potentially be formed in more than one way complicates our ability to interpret the properties of the galaxies that we see today. How can we improve our understanding of irregulars so that differences due to different origins might be more apparent? We need to better understand the kinematics and structures of irregular and dwarf galaxies. This includes the gas and stellar kinematics and velocity dispersions from which we can infer the distribution and amount of dark matter and the stellar structure of the galaxy. We also need more very deep studies of stellar populations of irregulars, particularly probing the extremes of galaxy characteristics. Only once there is a statistically significant sample of star formation histories can we begin to see trends. Finally, we need numerical simulations that can show whether interactions are feasible, perhaps between two unequal mass partners, that can produce a tidal dwarf and still leave the parent spiral intact. This is particularly important for the Local Group system in which we identify 6 candidate old tidal dwarfs, but the obvious parents are relatively normal looking spirals." }, "0005/astro-ph0005531_arXiv.txt": { "abstract": "Launch of the {\\em Far Ultraviolet Spectroscopic Explorer} ({\\em FUSE}) has been followed by an extensive period of calibration and characterization as part of the preparation for normal satellite operations. Major tasks carried out during this period include initial coalignment, focusing and characterization of the four instrument channels, and a preliminary measurement of the resolution and throughput performance of the instrument. We describe the results from this test program, and present preliminary estimates of the on-orbit performance of the {\\em FUSE} satellite based on a combination of this data and prelaunch laboratory measurements. ", "introduction": "The {\\em Far Ultraviolet Spectroscopic Explorer} ({\\em FUSE}) is obtaining high resolution, far ultraviolet spectra of faint astronomical objects in the 905 -1187 \\AA\\ wavelength range \\citep{Moos00}. Details of the {\\em FUSE} design and the predicted performance based on preflight measurements have been given previously \\citep{Friedman99, Sahnow96}. After an initial period of spacecraft checkout and on-orbit guidance tests, the two far ultraviolet detectors were powered on in 1999 August, and several months of checkout and science verification activities began. This included evaluation of the overall satellite performance, and preliminary instrument characterization. Science operations began in 1999 October, but characterization activity will continue throughout the three year mission with a decreasing frequency. Results from these early investigations show that the satellite is, with a few exceptions, performing quite well; most measures show that the performance is at or near preflight predictions, and adequate to meet the goals of the mission. The FUSE design consists of four coaligned optical channels, two of which have optics coated with SiC (SiC1 and SiC2), and two coated with LiF over Al (LiF1 and LiF2). Each channel is made up of a telescope primary mirror, a focal plane assembly containing the spectrograph entrance apertures, a holographically-ruled diffraction grating, and a portion of a detector. The previous paper \\citep{Moos00} presents an overview of the {\\em FUSE} mission, including the scientific background and an overall description of the instrument. This paper discusses the measured performance of the {\\em FUSE} satellite, as of 2000 February, with an emphasis on the properties of the instrument which affect the scientific data. ", "conclusions": "The {\\em Far Ultraviolet Spectroscopic Explorer} is performing well on orbit; early characterization activity is now complete, and routine scientific observations have begun. We have presented the first results from the characterization program, which will continue throughout the mission. Despite several as yet unexplained anomalies, high quality data addressing a wide variety of scientific problems is being collected, as seen in the accompanying papers." }, "0005/astro-ph0005477_arXiv.txt": { "abstract": "We present in this paper analyses of the $ROSAT$ PSPC and $ASCA$ SIS and GIS observations of the Crab-like supernova remnant (SNR) G54.1+0.3. Its spectrum obtained by $ROSAT$ PSPC favors a power law model with a photon index of -0.8$^{+0.8}_{-2.0}$, absorbed energy flux in 0.1-2.4 keV of 1.0$\\times 10^{-12}$ erg cm$^{-2}$ s$^{-1}$, and absorption column density of 12.3$^{+8.0}_{-3.2}\\times 10^{21}$ cm$^{-2}$. $ASCA$ SIS observation shows that its spectrum can also be best fitted with power law model. The fitted parameters are, photon index -1.9$^{+0.1}_{-0.2}$, absorbed energy flux in 0.7-2.1 keV 6.5$\\times$10$^{-13}$ erg cm$^{-2}$ s$^{-1}$, and column density 17.9$^{+2.8}_{-2.5}\\times 10^{21}$ cm$^{-2}$. The high absorption column density indicates a distance similar to the radius of the galaxy. The 0.1-2.4 keV X-ray luminosity of G54.1+0.3 is 3.2$\\times$10$^{33}$$d_{10}^2$ erg s$^{-1}$, where $d_{10}$ is the distance in 10 kpc. With an image restoration method we have obtained high spatial resolution X-ray image of the remnant, which clearly shows an X-ray jet pointing to the northeast with a length about 40$\\arcsec$ from the center of the nebula. Its X-ray luminosity in 0.1-2.4 keV is about 5.1$\\times$10$^{32}$$d_{10}^2$ erg s$^{-1}$. The X-ray jet is consistent with the radio extension to the northeast in both direction and position. We propose that the X-ray jet is connected with the pulsar assumed to exist in the remnant. ", "introduction": "Radio source G54.1+0.3 was first suggested to be a Crab-like SNR by Reich et al. (1985) for its flat spectral index of $\\alpha\\sim-0.1\\pm0.1$, filled-center morphology and significant polarization. This identification to G54.1+0.3 was confirmed by Velusamy $\\&$ Becker (1988) with high resolution multifrequency observations with the VLA and OSRT. In the high resolution VLA maps, G54.1+0.3 has a filled-center brightness distribution peaks around R.A.(2000) =19:30:30, DEC(2000)=18:52:11 and extends to the northeast and north (Velusamy \\& Becker 1988). They pointed out that these extensions are reminiscent of the radio jets seen in the Crab (Velusamy 1984), CTB80 (Angerhofer et al. 1981) and G332.4+0.1 (Roger et al. 1985). X-rays from G54.1+0.3 was detected by $EINSTEIN$ IPC (resolution $\\sim$ 1$\\arcmin$) with a source strength of 0.016$\\pm$0.004 counts s$^{-1}$ in the energy band 0.5-4.0 keV (Seward 1989). No extent to the X-ray emission was found, due to both its small angular size (2.0$^{\\prime}\\times1.2^{\\prime}$) (Velusamy \\& Becker 1988) and its low flux. A power law spectral fitting with energy index of 1.0 gives column density N$_H$ between 5$\\times$10$^{21}$ and 1$\\times$10$^{23}$ cm$^{-2}$, with the best fit value of 3$\\times$10$^{22}$ cm$^{-2}$, indicating a large distance of this source. In the paper we present the analyses of $ROSAT$ PSPC and $ASCA$ GIS and SIS observations of G54.1+0.3. We obtain its spectral information, and, with the aid of an image restoration method, we obtain a high spatial resolution X-ray map of the remnant which clearly shows an X-ray jet pointing to the northeast. ", "conclusions": "\\subsection{Distance and X-ray luminosity of G54.1+0.3} Velusamy \\& Becker (1988) suggested that G54.1+0.3 may have a distance of about 3.2 kpc, if its progenitor is in the star-forming region G53.9+0.3. The galactic HI column density in this direction is about 14.5$\\times$10$^{21}$ cm$^{-2}$ (Dickey $\\&$ Lockman 1990). The best fit column density we get from $ROSAT$ PSPC observation is a little lower and the best fit column density of the $ASCA$ SIS observation is a little higher than that value. These column densities imply a distance comparable with the radius of the galaxy, similar to the result of $EINSTEIN$ IPC (Seward 1989), 3.2 kpc might be then too close and 10 kpc should be a reasonable estimation. The X-ray luminosity in 0.1-2.4 keV is $L_X$ = 3.2$\\times$10$^{33} d_{10}^{2}$ erg s$^{-1}$, where $d_{10}$ is the distance to G54.1+0.3 in unit of 10 kpc. If the distance does not deviate from 10 kpc very much, its X-ray luminosity is three or four orders' lower than Crab Nebula (Helfand \\& Becker 1987), lies in the lower end of Crab-like SNRs, similar to that of SNR 3C58 (Helfand \\& Becker 1987; Helfand et al. 1995). The radio luminosity of G54.1+0.3 is about 5$\\times$10$^{33}$ $d_{10}^{2}$ erg s$^{-1}$, derived from the radio observations of Velusamy \\& Becker (1988). The ratio $L_x/L_r$ = 0.6, also similar to that of 3C58 (Helfand \\& Becker 1987). Seward \\& Wang (1988) found that a relation between the X-ray luminosity ($L_X$) of a plerionic SNR and the spin-down luminosity ($\\dot{E}$) of the central pulsar. Using that relation we can derive $\\dot{E}\\sim$8$\\times$10$^{35}$ erg s$^{-1}$ for the central pulsar in G54.1+0.3. \\subsection{Electron energy distribution} The $ROSAT$ PSPC observation of G54.1+0.3 shows that the X-ray flux at 1 keV is about 6.4$\\times$10$^{-4}$ mJy. As the radio flux at 1.4 GHz is 478 mJy, the flux index between radio and X-ray is about -0.7, a little flatter than the $ASCA$ SIS obtained X-ray energy index (-0.9) and much steeper than that of radio spectral index (-0.13) (Velusamy $\\&$ Becker 1998), indicating that the spectrum contains a break between radio and X-ray. Comparing the radio to X-ray flux index, radio flux index and the X-ray flux index, we find that the break is around 10$^{11}$ Hz. If the relativistic electrons have a power law energy distribution $n_e$ = $E^{\\gamma}$, the spectral index $\\alpha$ is $\\frac{\\gamma+1}{2}$. The spectral break in the spectral means a similar break in the electron energy distribution. The critical radiation frequency of a relativestic electron with energy $E$ in a magnetic field with strength $B$ is $\\nu_c = 16.1BE^{2}sin\\psi$ MHz, where $B$ is in $\\mu$G, $E$ is in GeV and $\\psi$ the incident angle of electron (Lang 1998). The electrons whose maximum radiations are at 1 GHz have typical energies of 9.4$B^{-0.5}$ GeV with $B$ the magnetic field strength in $\\mu$G, assuming that the incident angle is 45$\\degr$. Similarly the X-ray (around 1 keV) emitting relativestic electrons will have typical energies of 94$B^{-0.5}$ TeV. If the magnetic field is about 10$\\mu$G, the above estimations show that the electron energy distribution is $\\sim$$E^{-1.3}$ around 30 GeV and $\\sim$$E^{-2.8}$ around 300 TeV. An index break exists between 30 GeV and 300 TeV, probably around 300 GeV. The life time of a relativestic electron can be represented by $t_{1/2}$, the time of the electron loses half of its initial energy $E_0$, $t_{1/2}=\\frac{8.35\\times10^{9}}{(B sin\\psi)^{2}E_0}$ years, where $B$ is in $\\mu$G, $E_0$ is in GeV and $\\psi$ the incident angle of the electron (Lang 1998). The lifetimes for the 30 GeV, 300 GeV and 300 TeV photons in 10 $\\mu$G magnetic field are about 5.6$\\times$10$^{6}$, 5.6$\\times$10$^{5}$, 5.6$\\times$10$^{2}$ years, respectively. These three typical lifetimes will be used in the discussions of the origin of the electron energy distribution index break in the next paragraph. If the electrons from the center pulsar have a continuous power law energy distribution initially, the observed break should be due to the short lifetime of the high energy electrons. Because the low energy electrons which radiate radio emission have a long lifetime, their energy distribution represents the initial electron energy distribution well. The initial energy flux ratio $\\frac{f_{1.4GHz}}{f_{1keV}}$ is then 11.75. The currently observed energy flux ratio $\\frac{f_{1.4GHz}}{f_{1keV}}$ is 7.8$\\times10^{5}$, indicates that the age of G54.1+0.3 would be at least $\\frac{7.8\\times 10^{5}}{11.75}$$\\times$5.6$\\times$10$^{2}$=3.7$\\times$10$^{7}$ years. This large age value shows that the observed electron energy distribution break is quite probably an intrinsic property of the electrons from the central pulsar. \\subsection{X-ray jet} For the first time an well resolved X-ray image of G54.1+0.3 has been obtained. It shows a JLF pointing to the northeast. The simulation shows that such a structure can be clearly resolved by $ROSAT$ PSPC with the aid of an image restoration technique. The simulation also shows that this feature can not be attributed as the fluctuations of the bright source, it is an intrinsic structure of the object. We have studied the possibility that the JLF is indeed a separate object lies in a similar direction with G54.1+0.3. We find that there is no identified object in the 30$\\arcsec$ vicinity of the JLF except G54.1+0.3. The optical plate obtained by Palomar Observatory Sky Survey and electronically reproduced by Skyview of NASA/GSFC shows no source in the JLF region too. The JLF shown in figure 5 shows some enhancements in the head. But it might be a false phenomenon caused by the low quality of the original data and the restoration process, as can be found in the simulation, although some similar structures exist in the 4.8 GHz radio map. More simulations show that the length of the JLF is quite reliable, the width of the JLF might have an uncertainty up to $\\sim$ 50$\\%$. We have compared figure 4 with the 4.8 GHz VLA map obtained by Velusamy \\& Becker (1988) in details. The brightest point of the extended X-ray source locates at R.A.(2000)=19:30:30.0, DEC(2000)=18:52:07, which coincides with the brightest region of the radio source. The head of the JLF has a coordinate of R.A.(2000)=19:30:32.2, DEC(2000)=18:52:31, which also coincides with the northeastern enhancement in the radio map. The nice position coincidence of the X-ray and radio sources strongly favor their same origin. However, the X-ray source has a smaller extent than the radio source and no significant X-ray emission has been detected along the northward feature, which was suggested to be the most probable radio JLF by Velusamy \\& Becker (1988). It might be due to the intrinsic deficiency or the limited sensitivity of the present observation. There are two possible ways to explain the origin of the X-ray JLF. One is that it is a fragment produced in the supernova explosion, like the fragments detected around the Vela SNR, especially its `bullet'-like fragment A. (Aschenbach et al. 1995; Strom et al. 1995). However, significant radio emission has only been detected around the head of the fragments, implies that most of the relativistic electrons are in the leading edge of the fragments, close to the shock front (Strom et al. 1995). But in the case of G54.1+0.3 the radio emission has a similar distribution with the X-ray JLF, indicating a similar distribution of relativistic electrons with the X-ray brightness. It makes the fragment origin of the X-ray JLF implausible. The second is that the X-ray JLF is due to the relativistic electrons produced by the central pulsar, like X-ray jets detected in PSR 1929+10 (Wang et al. 1993), Crab SNR (Hester et al. 1995), Vela pulsar (Markwardt \\& \\\"Ogelman 1995), SNR MSH 15-52 (Tamura et al. 1996), SNR CTB80 (Wang \\& Seward 1984; Safi-Harb et al. 1995) in the galaxy and SNR N157B in the Large Magellanic Cloud (Wang \\& Gotthelf 1998). The coincidence of radio and X-ray emission in the case of Vela pulsar jet (Frail et al. 1997) and that of SNR N157B (Wang \\& Gotthelf 1998) strongly support this scenario. We conclude that the JLF we discoveried is quite probably an X-ray jet connected with the pulsar in G54.1+0.3. The X-ray emission of Vela pulsar jet can be fitted with both power law and thermal plasma model (Markwardt \\& \\\"Ogelman 1995), and the X-ray pulsar jet in MSH15-52 appears to be nonthermal. It is difficult to get the spectral properties of the X-ray jet in G54.1+0.3 with the present data. We assume that it share the same power law model with the whole remnant, and is due to the synchrotron radiation of relativistic electrons from the pulsar. The X-ray luminosity of the jet in 0.1-2.4 keV is then about 5.1$\\times$10$^{32}$$d_{10}^2$ erg s$^{-1}$. From the radio map of Velusamy $\\&$ Becker (1988) we estimate that the flux of the jet at 4.8 GHz is about 40 mJy. Its X-ray flux at 1 keV is about 9.4$\\times10^{-5}$ mJy. The two fluxes give a spectral index from radio to X-ray of about -0.73, quite similar to that of the whole remnant. As no significant radio spectral variation across the source has been detected (Velusamy $\\&$ Becker 1988), the jet electrons have a break with the energy distribution too, similar to the whole remnant. The distance of the jet head to the nebula center is about 40${\\arcsec}$. It corresponds to 2 pc if the SNR is 10 kpc away. Reccent distance measurements to Vela SNR obtained a distance of 250$\\pm30$ pc (Cha et al. 1999). If so the Vela pulsar jet is about 3 pc long (Cha et al. 1999, Markwardt \\& \\\"Ogelman 1995). The lengths of the two jets are quite similar." }, "0005/astro-ph0005194_arXiv.txt": { "abstract": "In the modified dynamics (MOND) the inner workings of dwarf satellites can be greatly affected by their mother galaxy--over and beyond its tidal effects. Because of MOND's nonlinearity a system's internal dynamics can be altered by an external field in which it is immersed (even when this field, by itself, is constant in space). As a result, the size and velocity dispersion of the satellite vary as the external field varies along its orbit. A notable outcome of this is a substantial increase in the dwarf's vulnerability to eventual tidal disruption--rather higher than Newtonian dynamics (with a dark-matter halo) would lead us to expect for a satellite with given observed parameters. The space of system parameters of the dwarf may be divided according to three main criteria: It can be either external- or internal-field dominated; it can be in the adiabatic or in the impulsive regime; and it can be in the tidal or non-tidal regime. The Milky Way's dwarf satellites populate all these regions of parameter space, and a single dwarf on an eccentric orbit can traverse several of them. The situation is particularly transparent in the external-field-dominated, adiabatic regime where the dynamics can be described analytically: due to the variation in the external-field strength with the galactocentric distance of the dwarf, $R$, its size changes as $R^{-1}$, and the internal velocities change as $R$. As the dwarf approaches the center it puffs up, becoming susceptible to tidal disruption. Adiabaticity is lost roughly at the same $R$ were tidal effects become important. The behavior near and in the impulsive regime is studied numerically. ", "introduction": "The dynamical behavior of dwarf spheroidals and other satellites of the Milky Way holds much information pertinent to the dark-matter problem. Attempts to elicit such knowledge include, on the one hand, measurements of the satellites' intrinsic properties such as the size, luminosity, and velocity dispersion, which evince mass discrepancies in the satellites (\\cite{ao,pryor91,matar}). This discrepancy is removed in the modified dynamics--MOND (\\cite{dwarf,mdeb,matar}). On the other hand, the satellites can be used to probe the gravitational field of their mother galaxy (specifically, the Milky Way) by using them as test particles to probe the galaxy's potential field (e.g., \\cite{lt}), or by studying tidal effects of the galaxy on the structure of the satellite taken as a finite body (e.g., \\cite{fl}). In Newtonian dynamics, the history of the center-of-mass motion may influence the internal workings of the satellite via tidal effects. Tidal disruption may have culled from the satellite population those that are internally weakly bound and/or move on elongated orbits, thus affecting the distribution of galactic orbits seen today (see, e.g., \\cite{lcg}). \\par In MOND, the interaction between the internal and center-of-mass motions, brought about by the theory's nonlinearity, goes beyond the Newtonian effects. For small systems (smaller than the scale over which the external field varies) the effect goes in one direction: while the center-of-mass motion is not affected by the internal motions, it may strongly affect them as explained in \\cite{bm} and \\cite{mgfe}. This occurs when the accelerations inside the satellite are of the order of or smaller than its center-of-mass acceleration; it is also required that the internal accelerations be small compared with the acceleration constant of MOND, $a_0$, as is always the case for the Milky Way's dwarf satellites. Due to this external-field effect (EFE) a satellite that plunges into the galaxy on an eccentric orbit increases in size, making itself an easier victim for tidal disruption. An additional destructive effect results when the changes in the external field become resonant with the internal motions. \\par The purpose of the paper is to describe, and demonstrate the pertinence, of these processes, which are peculiar to MOND. \\par In the next section we briefly recapitulate the external field effect. Next, in section 3, we consider a dwarf on an elongated orbit, delineating the different regimes of application of the MOND effects, and give an analytic description of dynamics in the adiabatic regime. In section 4 we describe the MOND N-body simulations, the results of which are described in section 5. Section 6 lists our conclusions and briefly comments on the Milky Way's dwarf satellites. ", "conclusions": "We have studied the existence, the nature, and the influence on dwarf satellites of the external field effect in MOND. For dwarf parameters in the EFE regime two situation are grossly distinguished: a) the adiabatic regime, in which tidal effects are not so important and b) the impulsive region, which also roughly coincides with the region where tidal forces become important. Due to the EFE the radius of a dwarf in the adiabatic regime increases as it approaches the mother galaxy. If the whole orbit is in the adiabatic regime, the structure of the dwarf simply changes periodically with the orbital period. If, however, some segment of the orbit is in the impulsive-tidal regime near pericenter, then the dwarf might lose much of its mass there. Even if it does not, it can emerge from this region having a larger radius and smaller velocity dispersion (hence, a longer intrinsic dynamical time). In its next approach to perigalacticon it will thus enter the impulsive-tidal regime at a larger distance from the center. \\par Clearly, all the above is highly germane to the dwarf system of the MW. The distribution of intrinsic and orbital parameters of presently observed dwarfs must have been greatly affected by interaction with the MW. And, one expects, MOND would give a different answer than Newtonian dynamics with dark matter. To actually deduce the present-day properties of the dwarfs would, however, require knowledge of the initial distribution of the orbital and intrinsic parameters of the dwarf-satellite population. Nothing is really known about this, so we refrain from speculating on the subject. We only estimate where our dwarf satellites stand as regards external-field dominance, adiabaticity, and the importance of tidal effects. \\par We consider the 10 dwarf spheroidal satellites with known parameters (\\cite{matar}): Sculptor, LSG 3, Fornax, Carina, Leo I, Sextans, Leo II, Ursa Minor, Draco, and Sagittarius. We take for the MW $V_{\\infty}= 220$ kms$^{-1}$. Since only core radii, $r_c$, are given we write for the mean radius $r=\\eta r_c$ to get for the adiabaticity parameter of those dwarfs $\\gamma \\sim \\eta^{-1}(22,150,14,15,47,8,39,14,20,2)$, respectively. So, with the exception of Sagittarius--which is known to be in the throes of disruption--and perhaps Sextans, these dwarfs are in the adiabatic regime within reasonable margins for $\\eta$, and even considering the approximate nature of the $\\gamma$ criterion. According to our analysis they are also only weakly affected by tidal forces at their present positions. As has been pointed out (\\cite{dwarf}, \\cite{mdeb}), most of the above dwarfs (with the exception of LSG 3, Leo I, and Leo II) are materially affected by the EFE: with the above choices of system parameters we get $\\beta\\sim \\eta^{-1}(0.7, 4.4, 0.7, 0.5, 1.9, 0.2, 1.2, 0.6, 0.9, 0.1)$. \\par If we apply the MOND scaling $\\alpha\\propto R^2$, which is valid in the $\\beta\\ll 1$ regime, to the dwarfs with $\\eta\\beta < 1$ (except for Sagittarius) we can estimate the minimum galactocentric distance above which the bulk of the dwarf is immune to tidal effects. This is given by $R_t^M\\approx R_0 \\alpha_0^{-1/2}=R_0 \\gamma_0^{-1/3}$, where here a subscript 0 marks present-day values. (If a dwarf in now on an outgoing section of its orbit it will return to the same $R$, as it goes in, in the same state.) For Sculptor, Fornax, Carina, Sextans, UMi, and Draco we get, respectively $R_t^M\\sim \\eta^{1/3}(28, 57, 41, 43, 27, 32)$ kpc. The corresponding Newtonian values ($R_t^N\\approx R_0\\gamma_0^{-1}$) are $R_t^N\\sim \\eta(4, 10, 7, 11, 5, 4)$ kpc. They are smaller than the corresponding MOND values if $\\eta$ is not so large that $\\alpha_0<1$. (For some dwarfs these Newtonian radii may fall within the stellar MW where our approximation of a spherical, logarithmic potential is not valid.) \\par Our results imply that for a given dwarf in the adiabatic regime on an elongated orbit under a strong EFE the size and velocity dispersion would be strongly dependent on the distance from the mother galaxy. One might then try to look for such correlations in the time-frozen population as seen today. This seems to us quite hopeless at present because the effects will be swamped by other factors of which we know very little; in particular, the unknown distribution of initial (intrinsic and orbital) parameters for the dwarfs. This is aggravated by the small sample size. \\par We leave for a future publication some other interesting effects predicted by MOND that result from the EFE. For example, in a dwarf in the EFE regime the total angular momentum is not conserved. We alluded to the fact that the direction of the external field if felt by the ``internal'' dynamics of the dwarf. In a static or adiabatic situation only the angular momentum along the external-field direction is conserved. \\par We thank the referee, Tad Pryor, for many useful comments and suggestions" }, "0005/astro-ph0005149_arXiv.txt": { "abstract": "While NLS1s have been studied intensively at X-ray and optical wavelengths, comparatively little is known about their characteristics in the radio band. Therefore, we have carried out an investigation of the radio luminosities, source sizes, spectral index distribution, and variability of a large, uniformly selected sample of NLS1s. Our results indicate that, in some respects, the radio properties of NLS1s differ significantly from those of classical Seyfert galaxies. Radio observations of NLS1s may thus provide important clues regarding the nature of their nuclear activity. ", "introduction": "It is well established that the optical and X-ray characteristics of narrow-line Seyfert~1 galaxies distinguish them from all other types of active galactic nuclei (AGNs). Unfortunately, the radio properties of NLS1s have been less well explored. In the only study dedicated to the subject, Ulvestad, Antonucci, \\& Goodrich (1995; hereafter UAG) found that NLS1s do not differ noticeably from nearby classical type~1 and type~2 Seyfert galaxies at centimeter wavelengths, in contrast to the results obtained in the optical and X-ray bands. This conclusion was based on the modest radio powers ($10^{20}$--$10^{23}$ W~Hz$^{-1}$) and small radio source sizes ($\\ltwid 300$~pc) of the objects they examined. But as UAG candidly noted, their sample of NLS1s was not defined using a uniform set of criteria, and only a fraction of the galaxies in it (9/15) was detected. We have investigated the radio emission of a larger, uniformly selected sample of NLS1s in order to gain further insight into the radio nature of these objects and their relation to other classes of AGNs. Our sample of 24 NLS1s is drawn from the catalog of {\\sl IRAS\\/} sources detected in the {\\sl ROSAT\\/} All-Sky Survey (Boller et al.\\ 1992; Moran et al.\\ 1996). Full details regarding the sample definition are provided in Moran et al.\\ (2000). We have obtained simultaneous high-resolution A-array VLA observations at 20~cm and 3.6~cm of most of the {\\sl IRAS-} and {\\sl ROSAT-}Observed NLS1 (``IRON'') galaxies. In addition, nearly all of the objects have been imaged at 20~cm in the moderate-resolution B and C arrays by Condon et al.\\ (1998a) and in the low-resolution D array as part of the NRAO VLA Sky Survey (Condon et al.\\ 1998b). All but one of the IRON galaxies are detected at 20~cm; 22 have three or more flux density measurements at that wavelength. ", "conclusions": "" }, "0005/astro-ph0005463_arXiv.txt": { "abstract": "We have carried out a spatio-kinematic study of the outflow from the classical T~Tauri star DG~Tau using the {\\em Space Telescope Imaging Spectrograph} (STIS) on board the Hubble Space Telescope (HST). A series of seven spatially offset long-slit spectra spaced by 0.07$''$ were obtained along the axis of the outflow to build up a 3-D intensity-velocity ``cube'' in various forbidden emission lines (FELs) and H$\\alpha$. Here we present high spatial resolution synthetic line images close to the star in distinct radial velocity intervals (from $\\sim $ +50\\,km\\,s$^{-1}$ to $\\sim$ -450\\,km\\,s$^{-1}$ in four bins, each $\\sim$ 125\\,km\\,s$^{-1}$ wide). The lowest velocity emission is also examined in finer detail (from +60\\,km\\,s$^{-1}$ to -70\\,km\\,s$^{-1}$ in five bins $\\sim$ 25\\,km\\,s$^{-1}$ wide). We have found that the highest velocity and most highly collimated component, i.e.\\ the jet, can be traced from DG~Tau to a distance D$\\sim$0.7$''$. The jet is on the axis of a pear-shaped limb-brightened bubble which extends between 0.4$''$ and 1.5$''$ from the source and which we interpret as a bow shock. Other condensations are seen close to the star indicating ongoing temporal variations in the flow. The low-velocity component of the outflow is found to be spatially wide close to the source ($\\sim$0.2$''$ at D=0.2$''$), in contrast to the high velocity jet (width $\\laeq$0.1$''$). We have also found evidence to suggest that not only does the density in the outflow increase longitudinally with proximity to the source but that it also increases laterally towards the flow axis. Thus, at least in the case of DG~Tau, the flow becomes gradually denser as it increases in velocity and becomes more collimated. Our observations show a continous bracketing of the higher speed central flow within the lower speed, less collimated, broader flow, down to the lowest velocity scales. This suggests that the low and high velocity FELs in the highly active T~Tauri star DG~Tau are intimately related. Implications of these observations for FEL models will be considered in a future paper \\citep{brmesc00}. ", "introduction": "\\label{intro} One of the most interesting questions in young stellar object (YSO) research is how their jets, e.g.\\ \\citet{camen97}, \\citet{crete96} or \\citet{eismundt97}, are collimated and accelerated. To address this problem one must obtain not only high spatial resolution but in addition kinematic information as close as possible to their source. Ground-based long-slit spectroscopic studies of optically visible jet sources have shown that the structure and kinematics of the outflow region on scales $\\laeq$1$''$ is rather complex (e.g. \\cite{solfbohm93}; \\cite{hirthms97}). In a significant fraction of classical T~Tauri stars (CTTSs) the forbidden emission lines (FELs) show two (or more) blueshifted velocity components which have very different properties (note that in many CTTSs the corresponding redshifted part of the flow is occulted by a circumstellar disk, at least close to the source). In the case of double-peaked FEL profiles, the emission consists of a so-called high-velocity component (HVC), having typical radial velocities of -60 to -200 km\\,s$^{\\rm -1}$ and a low-velocity component (LVC), having typical radial velocities of -5 to -20 km\\,s$^{\\rm -1}$ with respect to the systemic velocity of the star (see \\citet{hart95} and \\citet{solf97} for comprehensive discussions). While the HVC can be spatially very extended (at least in the outflow direction) and is often identified with a jet, the LVC is much more compact ($\\laeq$1$''$). Other differences between the two components have also been noted, for example the LVC appears to be of higher density but lower excitation than the HVC \\citep{hirthms97} and the LVC shows a clear inverse correlation of velocity with increasing critical density. The latter effect has been interpreted as evidence of acceleration in the LVC with increasing distance from the star. A number of theories have been put forward to explain the origin of the various FEL components. For example Kwan and Tademaru (1988, 1995) have suggested that the LVC and HVC are separate flows. According to their model the LVC is a poorly collimated wind coming from the outer periphery of the YSO disk while the HVC is a separate jet launched from closer to the star. Others have sought to explain the observations in terms of a one component, non-isotropic, wind model in which the appearence of separate FEL components is due to projection effects \\citep{hartray89, saf93, ouypud94}. In any event, \\citet{calvet97} has shown that the luminosity of the two components is tightly correlated, which seems to imply that they are not independent flows, and that the apparent dominance of one component over the other may be a density effect. In order to examine in more detail the nature of the compact FEL region, we have observed the CTTS DG~Tau with STIS on-board the HST. Multiple overlapping slit positions parallel to the outflow from this star were chosen so as to build up a 3-D spatial intensity-velocity ``cube''. Our target was picked not only because it is one of the closest CTTSs, but more significantly because its FEL region has a broad range of velocities, probably also a result of the relatively small angle between the jet axis and the line of sight ($\\approx 38^{\\circ}$, see \\cite{eismundt98}). This is important given the moderate wavelength resolution of STIS. Historically, DG~Tau was amongst the first T~Tauri stars from which a jet-like outflow (HH~158) was discovered \\citep{mundtfried83} and it has been imaged by the HST prior to the installation of the telescope's correcting optics \\citep{kep93}. On large scales ($\\approx$~10$''$) the jet seems to terminate in a bow shock \\citep{eismundt98} while high resolution spectro-imaging ground-based studies \\citep{lavalley97} have shown that the flow close to the star contains at least two resolved knots (at 2.7$''$ and 4$''$, epoch 1994.8). The outermost of these has a morphology and velocity gradient which is also consistent with it being a bow shock. In \\S 2 we describe our observational technique and give details of our STIS data. Our primary results are described in \\S 3 and discussed in \\S 4. A more detailed analysis will be presented in \\citet{brmesc00}. ", "conclusions": "\\label{disc} The STIS observations presented here clearly show that the HVC emission in DG~Tau comes from the most highly focused part of the outflow i.e.\\ the jet. The jet at high velocities can be traced back to at least 0.1$''$ (15\\,AU) from the star where quenching effects become important even in the case of the [OI]$\\lambda\\lambda$6300,6363 lines, which have the highest critical density ($ \\sim 10^6$ cm$^{-3}$) among the studied FELs. Obviously it is likely that the jet is collimated on even smaller scales. The shifts of the emission centroids in the different lines and in the different velocity bins can be explained quite naturally if there is an increase in jet density not only with proximity to the star (in the longitudinal direction) but also to the central outflow axis. That is to say the high velocity ``core'' of the jet is denser than its periphery. This is confirmed by an inspection of the [SII]$\\lambda\\lambda$6716,6731 doublet ratio (see Fig.\\ 2 and Bacciotti et al.\\ 2000). Note also that structures such as A1 and A2 are reminiscent of the ``bubbles'' recently seen by \\citet{krist99} in the case of XZ~Tauri. These are almost certainly internal working surfaces caused by temporal variations in the outflow from DG~Tau. Certainly there is plenty of evidence for strong jet velocity variations in DG~Tau (and in many other CTTS stars) on timescales of years. For example, the data of \\citet{mundtetal87} and \\citet{solfbohm93} show an increase in the radial velocity of the jet at D $\\sim$ 0.5$''$ by a factor of 2 within 8 years. Also the proper motion data of \\citet{eismundt98} indicate large velocity variations. Finally, several of the knots in the DG~Tau jet show bow shock-like structures (see also the HST Archive data presented in \\citet{brmesc00}) and these provide indirect evidence for strong velocity variations.\\\\ Turning now to the LVC, its nature still remains somewhat enigmatic. It was already clear from groundbased observations that the LVC and HVC differ in many properties such as density and excitation. A new difference is reported here i.e.\\ the rather large spatial width W of the LVC perpendicular to the jet at distances D of about 0.1$''$--0.3$''$ from the source. A comparison of W (at FWHM) between the high and low-velocity emission in [SII] and [OI] shows that the average W(LVC) $\\sim$ 0.18$''$ while the HVC is hardly resolved at a distance D of 0.2$''$. We note that this comparison of spatial widths can only be done using our [SII] and [OI] ``images'' as the H$\\alpha$ data is heavily contaminated by stellar H$\\alpha$ emission and in [NII] the LVC is very weak. Another interesting result of our study is the smaller velocity of the LVC at the edges of the flow (for full details see \\citet{brmesc00}). For example at D\\,=\\,0.2$''$ the LVC peaks at $\\sim$ -95\\,km\\,s$^{-1}$ in [OI] for the central slit position (S4) while in the two outermost slit positions (S1,S7) it peaks at about -18\\,km\\,s$^{-1}$. The corresponding values for the [SII] line are -60\\,km\\,s$^{-1}$ and -40\\,km\\,s$^{-1}$, respectively. Since these lines are optically thin, such observations clearly point to a rise in the average LVC velocity as the central outflow axis is approached.\\\\ To what degree the observations of DG~Tau presented here, particularly of the LVC, are representative of other CTTSs is an open question. DG~Tau is one of the most active CTTSs known and we caution that the LVC of DG~Tau is unusual in that it has the highest absolute velocity of all the CTTSs listed by \\citet{hart95}, as well as one of the highest accretion rates. That said it shares the typical properties of other LVCs and, in particular, the ratio of its luminosity (L) to that of the DG~Tau HVC is in perfect agreement with the L(HVC) v.\\ L(LVC) relationship noted by \\citet{calvet97}. Thus DG~Tau may simply be displaying the higher activity tail of the distribution of outflow properties amongst CTTSs. In conclusion our STIS observations show, for the first time, a quasi-continuous variation in the outflow velocity close to a YSO in the transverse direction to the flow. Detailed studies are required, however, to test whether these observations can constrain models for the generation of the LVC and HVC." }, "0005/astro-ph0005239_arXiv.txt": { "abstract": "We report the detection of 1.3\\,mm continuum and near-infrared $K-$band ($2.2\\micron$) emission from the submillimeter galaxy SMM\\,J00266+1708. Although this galaxy is among the brightest sub-mm sources detected in the blank-sky surveys ($L\\sim 10^{13} L_{\\sun}$), SMM\\,J00266+1708 had no reliable optical/near-infrared counter-part. We used sensitive interferometric 1.3\\,mm observations with the Owens Valley Millimeter Array to accurately determine the position of the sub-mm galaxy. Follow-up near-infrared imaging with the Keck~I telescope uncovered a new faint red galaxy at $K=22.5$ mag which is spatially coincident with the 1.3\\,mm emission. This is currently the faintest confirmed counter-part of a sub-mm galaxy. Although the redshift of SMM\\,J00266+1708 is still unknown, its high sub-mm/radio spectral index suggests that the system is at high redshift ($z\\ga2$). Approximately 50\\% or more of the sub-mm galaxies are faint/red galaxies similar to that of SMM\\,J00266+1708. These ultraluminous obscured galaxies account for a significant fraction of the total amount of star-formation at high redshift despite being missed by optical/ultraviolet surveys. ", "introduction": "Deep surveys of the submillimeter sky using the Submillimeter Common User Bolometer Array (SCUBA) camera (Holland et al. 1999) on the James Clerk Maxwell Telescope have uncovered a population of ultraluminous dusty galaxies at high-redshift (Smail, Ivison, \\& Blain 1997; Hughes et al.\\ 1998; Barger et al.\\ 1998; Eales et al.\\ 1999; Blain et al.\\ 1999a). This population accounts for a large fraction of the extragalactic background at mm/sub-mm wavelengths (Blain et al.\\ 1999b) and hence is important to our understanding of the distant universe. The sub-mm population is thought to contribute significantly to both the total amount of star-formation (Blain et al.\\ 1999b) and AGN activity (Almaini et al.\\ 1999) at high-redshift. The sub-mm population will likely show a mixture of AGN and starburst properties given their apparent similarities to the local population of ultraluminous ($L>10^{12}L_{\\sun}$) infrared galaxies (ULIGs, Sanders \\& Mirabel 1996). However, we could expect the majority ($\\sim 70$--80\\%) of the sub-mm galaxies to be predominantly powered by starbursts since this has been found for the local ULIGs (Genzel et al.\\ 1998). The early CO and X-ray data on the sub-mm population support the starburst nature of the population by showing the presence of sufficient molecular gas to fuel the star-formation activity (Frayer et al.\\ 1998, 1999) and the lack of expected X-ray emission if mostly dominated by AGN (Fabian et al.\\ 2000; Hornschemeier et al.\\ 2000). Observations of the dust-rich sub-mm galaxies complement the studies of the ultraviolet-bright Lyman-break sources (Steidel et al.\\ 1996, 1999) which tend to be much less luminous at infrared wavelengths (Chapman et al.\\ 2000). Only by studying both the Lyman-break and the sub-mm populations of galaxies will a complete picture for the star-formation history of the universe emerge. In order to understand the nature of the sub-mm population, we have been carrying out multi-wavelength observations of individual systems in the SCUBA Cluster Lens Survey (Smail et al.\\ 1998). This survey represents sensitive sub-mm mapping of seven massive, lensing clusters which uncovered 15 background sub-mm sources. The advantage of this sample is that the amplification of the background sources allows for deeper source frame observations. Also, lensing by cluster potentials does not suffer from differential lensing so that the observed flux ratios will represent intrinsic values, despite the possible variation of source size at different wavelengths. The most challenging aspect for follow-up observational studies of the sub-mm population is determining the proper counter-parts to the sub-mm emission and obtaining their redshifts (Ivison et al.\\ 1998, 2000a). The large 15$''$ SCUBA beam leaves ambiguity in identifying the galaxy associated with the sub-mm emission. The early results based on optical imaging and spectroscopy were overly optimistic in the identification of the sub-mm counter-parts (Smail et al.\\ 1998; Barger et al.\\ 1999; Lilly et al.\\ 1999). Radio data (Smail et al.\\ 2000a) and initial near-infrared (NIR) imaging (Smail et al. 1999) suggest that several of the original candidate optical counter-parts (e.g., Barger et al.\\ 1999) are incorrect. Despite their ultra-high luminosities, many sub-mm galaxies are nearly completely obscured by dust at ultraviolet/optical wavelengths. For these highly obscured galaxies, follow-up radio (Smail et al.\\ 2000a) and/or mm interferometry (Downes et al.\\ 1999; Bertoldi et al.\\ 2000) as well as near-infrared observations are required in order to uncover the proper counter-part. The galaxy SMM\\,J00266+1708 is an excellent example of such a source. ", "conclusions": "We report the identification of the sub-mm source SMM\\,J00266+1708 with a faint red galaxy ($K=22.5$\\,mag) which is undetected at optical wavelengths, despite very deep observations. This source has an extremely high luminosity of approximately $10^{13}L_{\\sun}$ even after correcting for lensing. The current data for the sub-mm Cluster Lens Survey suggest that 40\\%--70\\% of the sub-mm population as a whole are faint/red galaxies which are undetected at optical wavelengths. These faint/red sub-mm galaxies are thought to contribute significantly to the total amount of star formation at high redshift and are hence important to our understanding of the early evolution of galaxies. The redshift of SMM\\,J00266+1708 is currently unknown, but the galaxy is expected to be at a redshift $z>2$. Obtaining a redshift will be extremely challenging with current instrumentation. At $K=22.5$ mag, the galaxy pushes the capabilities of even the largest ground based telescopes. Since the galaxy is relatively red ($I-K > 3.6$), we expect H$\\alpha$ to be the brightest optical emission line, and perhaps the only optical line currently detectable based on comparisons with the ERO HR10 (Dey et al.\\ 1999). If the redshift is similar to that estimated from the SED of the galaxy ($z\\sim3.5$), H$\\alpha$ would be shifted redward of $K$-band, making ground based observations extremely challenging. We could expect much fainter $K-$band magnitudes of 23--26 for similar ULIG/sub-mm galaxies ($\\ga 10^{12} L_{\\sun}$) which are unlensed. Therefore, obtaining optical/NIR redshifts for many of the sub-mm galaxies may have to wait for the {\\it Next Generation Space Telescope}. Alternatively, redshifts could be directly measured from the CO lines themselves with future ground-based instruments, such as the Atacama Large Millimeter Array (ALMA), operating at mm-wavelengths (Blain et al.\\ 2000). Sensitive interferometric observations at sub-mm/mm-wavelengths with the next generation of instruments will be crucial for our understanding of these dust-obscured systems." }, "0005/astro-ph0005525_arXiv.txt": { "abstract": "We present line observations of different molecular species toward the dense core in \\object{L483} and its bipolar outflow powered by the Class 0 object \\object{IRAS 18148-0440}. $^{12}$CO(2--1) maps show that the outflow is well collimated and asymmetric, and that its gas is warmer than the surrounding cloud by at least a factor of 2. In contrast with the outflows from other Class 0 objects, the CH$_3$OH($2_k$--$1_k$) lines in L483 do not show strong high velocity wings, although there is a small ($\\sim 0.3$ km s$^{-1}$) velocity shift approximately along the outflow direction. We do not find evidence for a CH$_3$OH abundance enhancement in the flow, and the CH$_3$OH lines trace a centrally concentrated core which we model, assuming optically thin emission, as having a density gradient between $r^{-1}$ and $r^{-1.5}$ for radii between $15''$ and $100''$. H$_2$CO(2$_{12}$--1$_{11}$) lines show strong high-velocity wings with the same distribution as the outflow, and evidence for a H$_2$CO abundance enhancement of a factor of 20 with respect to the ambient cloud. At ambient velocities and over the central $40''$, this line presents a strong self absorption and a brighter blue peak, a characteristic signature of inward motions. A simple analysis of the H$_2$CO line profiles suggests an infall rate of $2\\times 10^{-6}$~M$_\\odot$ yr$^{-1}$. Combining the results from our observations with previous work, we discuss the evolutionary status of IRAS 18148-0440 and its outflow. The bipolar outflow presents some characteristics common to other outflows from Class 0 sources, like high degree of collimation, gas heating, and H$_2$CO abundance enhancement. However, other characteristics, like its low velocity, the lack of bright SiO or CH$_3$OH outflow wings, and the association with a NIR scattering nebula (optically invisible) seem more common to outflows from the more evolved Class I sources. As IRAS 18148-0440 is a Class 0 object based on its spectral energy distribution, we propose that it is more evolved than other objects in its class, probably in transition from Class 0 to Class I. ", "introduction": "Class 0 objects are the youngest stellar objects known (Andr\\'e et al. \\cite{And93}, \\cite{And99}). They commonly power bipolar outflows with extreme characteristics like a very high degree of collimation and evidence for shock processing of molecular gas even in cases of very low stellar luminosity (see Bachiller \\& Tafalla \\cite{Bac99} for a recent review). In order for these outflows to evolve into the more quiescent (``standard'') outflows associated with Class I sources, rapid changes in outflow morphology and kinematics have to occur in the few $10^4$ yr that Class 0 lasts (Andr\\'e et al. \\cite{And93}). These changes are most likely associated with changes in the source itself, which is undergoing its major phase of assembling via gravitational infall (e.g., Bontemps et al. \\cite{Bon96}, Mardones et al. \\cite{Mar97}). Understanding how these first evolutionary changes of the stellar and outflow life occur is a major challenge to star formation studies, and it requires the simultaneous analysis of Class 0 objects, their outflows, and their dense gas environments. Here we present a molecular line study of the L483 core and its outflow powered by IRAS 18148-0440 (IRAS 18148 hereafter), a system that we find at the end of its Class 0 stage, starting its transition to become a Class I object. The source IRAS 18148 in L483, first identified as an embedded object by Parker (\\cite{Par88a}), is one of the reddest low-mass sources known (Ladd et al. \\cite{Lad91a}, \\cite{Lad91b}), and is located toward the Aquila Rift, at a most likely distance of 200 pc (Dame \\& Thaddeus \\cite{Dam85}). Ladd et al. (\\cite{Lad91a}) and Fuller et al. (\\cite{Ful95}) estimate a source bolometric temperature (in the sense of Myers \\& Ladd \\cite{Mye93}) of 50-60 K, and using the flux compilation by Fuller et al. (\\cite{Ful95}) (their Fig. 4), we estimate a $L_\\mathrm{smm}$/$L_\\mathrm{bol} \\gtrsim 0.9$~\\% (also, Fuller et al. \\cite{Ful95} fit the spectral energy distribution with a single-temperature dust model at 40~K, and from their 1.1mm flux, we estimate $L_\\mathrm{bol}$/$L_\\mathrm{1.1mm} < 2.5 \\times 10^4$). These numbers suggest that IRAS 18148 is a Class 0 object (Andr\\'e et al. \\cite{And93}, \\cite{And99}, Chen et al. \\cite{Che95}), as already proposed by Fuller et al. (\\cite{Ful95}) and Fuller \\& Wooten (\\cite{Ful00}), although it is less extreme than the prototype Class 0 source \\object{VLA1623} (Andr\\'e et al. \\cite{And90}, Andr\\'e et al. \\cite{And93}). IRAS 18148 has a luminosity of about 10~L$_\\odot$ and drives a well-collimated bipolar CO outflow (Parker et al. \\cite{Par88b}, \\cite{Par91}, Fuller et al. \\cite{Ful95}, Bontemps et al. \\cite{Bon96}, Hatchell et al. \\cite{Hat99}), and is associated with a variable H$_2$O maser (Xiang \\& Turner \\cite{Xia95}) and shocked H$_2$ emission (Fuller et al. \\cite{Ful95}, Buckle et al. \\cite{Buc99}). NIR imaging of the source vicinity shows a well-defined, parabolic reflection nebula, which is optically invisible and coincides with the blue lobe of the CO outflow (Hodapp \\cite{Hod94}, Fuller et al. \\cite{Ful95}). Ammonia observations by Goodman et al. (\\cite{Goo93}), Fuller \\& Myers (\\cite{Ful93}), Anglada et al. (\\cite{Ang97}), and Fuller \\& Wootten (\\cite{Ful00}) reveal that the L483 core is centrally concentrated, has a strong velocity gradient across it, and a gas kinetic temperature of about 10~K. H$_2$CO and CS spectra toward the central source present strong self absorption with lines having brighter blue peak, a signature of infall motions (Myers et al. \\cite{Mye95}, Mardones et al. \\cite{Mar97}). The combination in L483 of Class 0 characteristics, like a low $T_\\mathrm{bol}$ and infall asymmetry, together with the presence of a bright NIR nebula, indicative of partial core disruption, makes this source an interesting object to study the early evolution of a very young stellar object. To carry out such a study, we have observed L483 in tracers sensitive to different aspects early stellar life, like the outflow (CO, section 3.1), the dense core and possible chemical outflow anomalies (CH$_3$OH, section 3.2), and infall and shock chemistry (H$_2$CO, 3.3). From the combination of these observations, we propose that the central source in L483 has already started its transition toward Class I, and that the outflow has lost part of the chemical richness characteristic of Class 0 flows (section 4). \\smallskip ", "conclusions": "We have observed the L483 core and outflow in different mm molecular transitions and made full maps in CO(2--1), CH$_3$OH(2$_k$--1$_k$), and H$_2$CO(2$_{21}$--1$_{11}$). With these data, we have studied the outflow, the core, and their relation with the IR cometary nebula around IRAS 18148. The main conclusions of our work are as follows: 1. The $^{12}$CO emission at ambient velocities is extremely thick with the brightness temperature expected for gas at 9~K, the temperature previously estimated for the core gas. Outside the ambient regime the $^{12}$CO lines present bright wings indicating outflow material warmer than the ambient gas by at least a factor of 2. A simple model for the C$^{17}$O(1--0) emission towards the core center shows that this line is optically thin and non Gaussian due to the presence of two velocity components. From the integrated C$^{17}$O(1--0) emission we estimate a central H$_2$ column density of $3 \\times 10^{22}$ cm$^{-3}$ in the inner $20''$. 2. The CO outflow emission is compact and slow, with a total length of 0.15 pc and a kinematical age of $10^4$~yr. Lower limits to the outflow mass, momentum, and energy are 0.01~M$_\\odot$, 0.03~M$_\\odot$ km s$^{-1}$, and $2\\times 10^{42}$ erg, respectively. The CO outflow is asymmetric, with a blue lobe having a bright spot coinciding with the NIR nebula and the red lobe being weaker near IRAS 18148 and having a relative maximum $45''$ from the source. This asymmetry suggests that the reflection nebula around the IRAS source may also be asymmetric and have a more prominent blue side. 3. The CH$_3$OH emission traces a dense core with no appreciable outflow wing contribution, although there is a shift in the line velocity along the direction of the outflow. The CH$_3$OH emission is centrally concentrated on the IRAS position, and in the central $200''$ (0.1 pc) it decreases radially in a manner intermediate between what would be expected for optically thin emission with density power laws of $r^{-1}$ and $r^{-1.5}$. The estimated mass in this region is 5-10 M$_\\odot$. No evidence for CH$_3$OH or SiO abundance enhancement is found in the outflow. 4. The H$_2$CO(2$_{12}$--1$_{11}$) emission is self absorbed at ambient velocities, and presents spectra with brighter blue peak, characteristic of inward motions, toward the central $40''$. With a simple model, we estimate an average infall speed of 0.02 km s$^{-1}$ and an infall rate of $2\\times 10^{-6}$~M$_\\odot$ yr$^{-1}$. At high velocities, the H$_2$CO(2$_{12}$--1$_{11}$) line presents bright wings in the same sense as the CO outflow wings, indicative of outflow acceleration. Comparing the H$_2$CO and CO wing intensities we find that the H$_2$CO abundance in the outflow regime is enhanced with respect to the ambient regime by a factor of 20. 5. The combination of CO, CH$_3$OH, and H$_2$CO data shows that the L483 outflow is less extreme than other outflows from Class 0 objects, like L1448-C and L1157, although it has some of their characteristics, such as gas heating and some abundance enhancement. We therefore suggest that the Class 0 source at the center of the L483 outflow is more evolved than other Class 0 sources, and it is in its transition to become a Class I object." }, "0005/astro-ph0005243_arXiv.txt": { "abstract": "The giant elliptical galaxy M86 in Virgo has a $\\sim28$ kpc long dust trail inside its optical halo that points toward the nucleated dwarf elliptical galaxy, VCC 882. The trail seems to be stripped material from the dwarf. Extinction measurements suggest that the ratio of the total gas mass in the trail to the blue luminosity of the dwarf is about unity, which is comparable to such ratios in dwarf irregular galaxies. The ram pressure experienced by the dwarf galaxy in the hot gaseous halo of M86 was comparable to the internal gravitational binding energy density of the presumed former gas disk in VCC 882. Published numerical models of this case are consistent with the overall trail-like morphology observed here. Three concentrations in the trail may be evidence for the predicted periodicity of the mass loss. The evaporation time of the trail is comparable to the trail age obtained from the relative speed of the galaxies and the trail length. Thus the trail could be continuously formed from stripped replenished gas if the VCC 882 orbit is bound. However, the high gas mass and the low expected replenishment rate suggest that this is only the first stripping event. Implications for the origin of nucleated dwarf ellipticals are briefly discussed. ", "introduction": "M86 (NGC 4406) is a bright elliptical (E3/S0) galaxy located in the Virgo Cluster at a distance of $\\sim18.3$ Mpc (Capaccioli et al. 1990). It has a redshift of $-227$ km s$^{-1}$ (Binggeli et al. 1985), while the mean heliocentric velocity of the cluster as a whole is 1050 km s$^{-1}$ (Binggeli et al. 1993). M86 is thought to be on an orbit passing through the core of the cluster approximately every 5 billion years (Forman et al. 1979). It is an X-ray object and appears to be a weak radio source (Laing et al. 1983; Fabbiano et al. 1992; Rangarajan et al. 1995). A plume of X-ray, HI, and infrared emission from M86 suggests that its interstellar medium was swept back by the ram pressure from its motion through the intracluster medium (Forman et al. 1979; Fabian, Schwarz, \\& Forman 1980; Takeda, Nulsen, \\& Fabian 1984; Bregman \\& Roberts 1990; Knapp et al. 1989; White et al. 1991). There is also an optical asymmetry to M86 that gives it a slightly enhanced emission along the plume (Nulsen \\& Carter 1987). Here we discuss a nucleated dwarf elliptical galaxy, VCC 882 (NGC 4406B; Binggeli et al. 1985), that lies just to the northeast of M86, inside its projected stellar halo. Deep CCD images show a 28 kpc long dust trail inside M86 that appears to follow VCC 882 in its orbit. This trail is possibly the result of ram pressure stripping of gas originally inside VCC 882 that was removed by the high pressure of its motion through the hot gaseous halo of M86. The gas mass obtained from the extinction in the trail is consistent with this former connection to VCC 882. Other evidence for an interaction between M86 and VCC 882 could be an isophotal twist in the central 8 to 80 arcsec of M86 (Bender \\& Mollenhoff 1987), and the asymmetric outer isophotes of M86 (Nulsen \\& Carter 1987). Gas stripping is pervasive within this core region of the Virgo cluster. Three spirals are close to the M86/VCC 882 pair in projection: NGC 4438, NGC 4388, and NGC 4402. There is also a nearby dwarf galaxy, IC 3355. The spirals have been severely stripped of their outer HI disks (Warmels 1986; Hoffman, Helou, \\& Salpeter, 1988; Cayatte et al. 1990), but the dwarf looks normal. The peculiar negative velocity of M86 is also shared by NGC 4402, NGC 4438, and IC 3355, as if at least some of these galaxies are comoving like a group through the Virgo cluster (Kotanyi \\& Ekers 1983). Detailed studies of one of these galaxies, NGC 4438, suggest that ram pressure from its motion through the Virgo intracluster medium has visibly distorted its outer disk (Arp 1966), pushing the gas and star formation off the normal plane, and producing a short diffuse trail of radio continuum and X-ray emission (Kotanyi \\& Ekers 1983; Kotanyi, van Gorkom \\& Ekers 1983). A second galaxy in this group, NGC 4388, has been studied by Pogge (1988), Petitjean \\& Durret (1993) and Veilleux et al. (1999), with mixed results on a stripped origin for extraplanar material. The spiral galaxy NGC 4569, at an equal distance from M87 but on the eastern side, has a negative velocity too, along with an anemic classification (van den Bergh 1976) and depletion in HI (Cayatte et al. 1990). Another dwarf galaxy, IC 3475, is closer in projection to M87 than IC 3355 and is highly stripped of HI (Vigroux et al. 1986), while another anemic spiral, NGC 4548, is about twice as far from M87 as these others and has a distorted outer HI disk (Vollmer et al. 1999). Ram pressure seems to have stripped the Virgo spiral NGC 4694 also. This galaxy is far to the east of the core region in Virgo. It has a 36 kpc long trail of HI streaming off to the west, with a linear velocity gradient along the trail that smoothly connects it with the galaxy (van Driel \\& van Woerden 1989). A very faint dwarf galaxy is also in the trail, where the HI column density peaks. All of these cases suggest stripping from the motion of a galaxy through the hot, low-density, intracluster medium inside Virgo. The case discussed here differs because VCC 882 was apparently stripped by its motion through the much denser hot gas associated with the elliptical galaxy M86. Its relative speed is just as large as in the other cases, exceeding 1000 km s$^{-1}$, but the ram pressure it felt must have been much larger because of the higher ambient density. A similar case is the Virgo dwarf galaxy UGC 7636, which was apparently stripped as it moved through the giant elliptical, NGC 4472. The evidence for this is: a gas cloud to the side of UGC 7636 (Sancisi, Thonnard \\& Ekers 1987) that has the right mass to have been formerly part of UGC 7636 (Patterson \\& Thuan 1992; Irwin \\& Sarazin 1996); a 30 kpc long optical trail of luminous debris adjacent to the cloud and the dwarf (McNamara et al. 1994); absorption of the elliptical galaxy X-ray radiation by the HI cloud (Irwin \\& Sarazin 1996), and an oxygen abundance in an HII region of the cloud that is consistent with the abundance expected for UGC 7636 (Lee et al. 2000). Previous photometric studies of Virgo cluster galaxies were made by Binggeli et al. (1985, 1993), Bender \\& Mollenhoff (1987), and Caon et al. (1994). Katsiyannis et al. (1998) co-added thirteen Schmidt exposures to produce a deep R-band image of the Virgo southeast region, from which they studied extended and overlapping halos. The dust trail discussed here did not appear in these other images. ", "conclusions": "Deep B, V, and I images of the southeast core region of the Virgo cluster reveal a dust trail in the central region of M86, spanning 28 kpc with a thickness of $\\sim500$ pc. The dust appears to connect to the nucleated dwarf galaxy VCC 882 located just north of M86. We estimate that the dust has a V-band extinction of 0.1 mag, with a total mass of $10^8$ M$_\\odot$. This dust and the associated gas may have been ram-pressure stripped from VCC 882 over the last 10--20 My as the dwarf galaxy passed through the hot X-ray emitting gas of M86. The mass of gas is somewhat consistent with its former presence inside VCC 882, and the morphology of a trail is consistent with theoretical predictions for relatively weak ram pressures and periodic outflows. The similarity between the evaporation time of the trail and the dwarf orbit time suggest that evaporation is important. The length of the trail could even be determined by steady evaporation at the back end. A problem with this interpretation is that VCC 882 is a classical nucleated dwarf. It seems to have no recent star formation as if it just had a gas disk, and the gas mass estimate is much too high for such a morphology. If the trail really did come from VCC 882, then it would have had a mixed morphology before the stripping, like several other rare cases that have been studied elsewhere. More observations will be required to determine the origin and history of this gas. We gratefully acknowledge summer student support from the Wm. Keck Foundation to the Keck Northeast Astronomy Consortium, and thank Vassar College for a research publication grant." }, "0005/astro-ph0005596_arXiv.txt": { "abstract": "Measurements of the intensity of the cosmic X-ray background (XRB) carried out over small solid angles are subject to spatial variations due to the discrete nature of the XRB. This cosmic variance can account for the dispersion of XRB intensity values found within the \\ASCA, \\SAX\\ and \\ROSAT\\ missions separately. However there are differences among the values obtained in the different missions which are not due to spatial fluctuations but, more likely, to systematic cross-calibration errors. Prompted by recent work which shows that \\ROSAT\\ PSPC has calibration differences with all the other missions, we compute a bayesian estimate for the XRB intensity at 1 keV of $10.0^{+0.6}_{-0.9}\\, \\uint$ (90 per cent confidence errors) using the \\ASCA\\ and \\SAX\\ data points. However, this value is still significantly larger than the $HEAO-1$ intensity measured over many thousands of square degrees ($8\\, \\uint$). ", "introduction": "Most of the X-ray background (XRB) above photon energies of a few keV is known to be extragalactic in origin (see, e.g, reviews by Boldt 1987 and Fabian \\& Barcons 1992). Its spectrum was well measured by the $HEAO-1$ A2 experiment in the 3-50 keV energy range (Marshall et al 1980), where it fits a thermal bremsstrahlung model with a temperature $kT\\approx 40\\, \\keV$. An overall fit to the XRB spectrum from 3 keV to 10 MeV was presented by Gruber (1992) based on data from the A2 and A4 experiments on board $HEAO-1$, which showed an extrapolated intensity at 1 keV of $\\approx 8 \\uint$. More recently, measurements of the XRB spectrum and intensity at photon energies $< 10\\, \\keV$ have been obtained with the use of imaging instruments. Using \\ASCA\\ data, Gendreau et al (1995) confirmed that the thermal bremsstrahlung shape (which at photon energies below $\\sim 20\\, \\keV$ can be approximated by a power law with energy spectral index $\\alpha\\sim 0.37$) provides a good fit to the XRB spectrum down to about 1 keV. Vecchi et al (1999) used the LECS and MECS instruments on board \\SAX\\ in the 1-8 keV band to confirm that the XRB spectrum is consistent with a power law with $\\alpha\\sim 0.4$. $ROSAT$ PSPC observations have also provided data on the spectrum of the extragalactic XRB. Hasinger (1992) found a very steep spectrum for the XRB below 2 keV ($\\alpha\\sim 1$). However, shadowing experiments with $ROSAT$ have provided the more stringent upper limits to the slope of the extragalactic XRB at soft X-ray energies ($\\alpha<0.7$, Barber \\& Warwick 1994). The relative steepness of the \\ROSAT\\ spectra with respect to many other instruments has been studied in detail by Iwasawa, Fabian \\& Nandra (1999). The fact that the spectral shape of the same sources is usually consistent in observations of many instruments with the exception of \\ROSAT\\ (which usually finds steeper spectra) is highly suggestive of a calibration mismatch which will certainly affect the XRB spectrum as well. The normalisation of the extragalactic XRB (that we parametrise as the XRB intensity $I_{XRB}$ at 1 keV in units of $\\uint$) remains uncertain. The Marshall et al (1980) measurement is the most robust result as it was performed over a very large solid angle ($\\sim 10^4\\, \\deg^2$) with instrumentation especially designed to subtract efficiently and accurately the particle background. The XRB intensity measured by the imaging instruments on board \\ASCA, \\SAX\\ and \\ROSAT\\ invariably yield higher values which are often statistically discrepant among them. \\begin{table*} \\begin{center} \\begin{tabular}{l l l c c c l} \\hline Instrument & Field & $\\Omega_i$ & $I_i$ & $\\sigma_i$ & $\\Sigma_i$ & Ref\\\\ & Name & ($\\deg^2$) & ($\\uint$)\\\\ \\hline ROSAT PSPC & QSF3 & 0.223 & 11.4 & 0.34 & 1.00 & Chen et al (1996)\\\\ ASCA SIS & QSF3 & 0.134 & 10.0 & 0.37 & 0.76 & Chen et al (1996)\\\\ ROSAT PSPC & GSGP4 & 0.283 & 11.8 & 0.44 & 4.49 & Georgantopoulos et al (1996)\\\\ ROSAT PSPC & SGP2 & 0.283 & 12.0 & 0.61 & 4.53 & Georgantopoulos et al (1996)\\\\ ROSAT PSPC & SGP3 & 0.283 & 12.0 & 0.68 & 4.52 & Georgantopoulos et al (1996)\\\\ ROSAT PSPC & QSF1 & 0.283 & 9.9 & 0.65 & 4.49 & Georgantopoulos et al (1996)\\\\ ASCA GIS & Lockman & 0.165 & 10.9 & 0.61 & 1.31 & Miyaji et al (1998)\\\\ ASCA GIS & Lynx & 0.144 & 9.3 & 0.61 & 1.38 & Miyaji et al (1998)\\\\ ROSAT PSPC & Lockman & 0.165 & 10.0 & 0.30 & 2.10 & Miyaji et al (1998)\\\\ ROSAT PSPC & Lynx & 0.144 & 11.5 & 0.43 & 2.28 & Miyaji et al (1998)\\\\ ASCA SIS & Various& 0.538 & 8.9 & 0.50 & 0.96 & Gendreau et al (1995)\\\\ ROSAT PSPC & Various& 5.9 & 13.4 & 0.18 & 0.66 & Hasinger (1992)\\\\ SAX LECS+MECS & Various& 0.726 & 11.0 & 0.30 & 0.77 & Vecchi et al (1999)\\\\ \\hline \\end{tabular} \\caption{Measurements of the XRB intensity at 1 keV carried out with imaging instruments.$\\Omega_i$ is the solid angle covered, $I_i$ the measured intensity, $\\sigma_i$ the 1-sigma statistical uncertainty and $\\Sigma_i$ the 1-sigma uncertainty derived when statistical and cosmic variances are included.} \\end{center} \\end{table*} In this paper we point out that the discrete nature of the XRB introduces a cosmic variance in its intensity which is large for observations carried out over small solid angles. This cosmic variance is just the confusion noise caused by unresolved or non removed sources in the images (Scheuer 1974, Barcons 1992). We find variations of the order of 10 per cent for solid angles under 1$\\deg^2$ and therefore cosmic variance often dominates over the statistical uncertainties quoted in the various estimates of the XRB intensity. Once this is included, we combine individual measurements of the XRB to compute bayesian estimates of the overall XRB intensity. We find that measurements carried out within the same mission are brought to consistency by cosmic variance, but that systematic differences between different missions remain. ", "conclusions": "Cosmic variance is able to account for the dispersion of the measured values of the extragalactic XRB intensity within the same mission. However, systematic differences remain among different missions, which cannot be understood in terms of spatial fluctuations. A cross-calibration mismatch between \\ROSAT\\ observations and those form other missions has been reported (Iwasawa et al 1999). But even using the \\ASCA\\ and \\SAX\\ data only the result ($10.0^{+0.6}_{-0.9}\\, \\uint$) is higher than the XRB intensity at 1 keV extrapolated from the HEAO-1 data at photon energies above 3 keV. A steepening of the XRB spectrum at energies below 3 keV does not really help as both \\ASCA\\ and \\SAX\\ are sensitive to energies $\\sim 3\\, \\keV$ and above where the mismatch persists. The only conclusion we can reach is that although internal calibration uncertainties in each mission amount to probably less than 10 per cent, the cross-calibration among them has still large residual errors, preventing a more precise determination of the XRB intensity. XMM is and will be for many years the most sensitive X-ray imaging facility over the 0.1-10 keV energy band, providing also moderate spectral resolution. Over 1 year, XMM will carry out observations of 500-1000 fields of $\\sim 0.2\\, \\deg^2$, so we expect $\\sim 100\\, \\deg^2$ to be covered each year. The cosmic variance (roughly scaling as $\\Omega^{-1/2}$) will then be small enough for unambiguous determinations of the XRB spectrum and intensity. If the EPIC instruments can be calibrated to significantly better than 10 per cent, this will certainly solve the issue of the intensity of the XRB." }, "0005/astro-ph0005075_arXiv.txt": { "abstract": " ", "introduction": "Ever since the discovery of radiation which comes from cosmos by Hess in 1912 and christened by Millikan in 1925 as `cosmic rays', physicists and astronomers have speculated upon their origin. Fermi (1949) made the first serious attempt at explaining the power law nature of the cosmic ray spectrum. He noted that a particle could increase its energy at collisions against magnetic field irregularities. In his model cosmic rays interact with galactic molecular clouds that move randomly. Particles increase their energy in head-on collisions which are more frequent than overtaking collisions when they loose energy. The process is known as the second-order Fermi acceleration because the mean particle momentum gain $\\Delta p/p$ in one interaction is proportional to $(V/v)^{2}$, $V$ is the root-mean-square velocity of a cloud and $v$ is the particle velocity, considered below to be comparable to the speed of light -- c. Presently the second-order Fermi acceleration is considered in plasma where the magnetic field fluctuations play a role of the Fermi `clouds'. \\noindent {\\bf Nonrelativistic shocks.} The concept that shock waves accelerate particles in a mechanism similar to the one described by Fermi (1949) appeared in four seminal papers: Krymski (1977), Axford et al. (1977), Bell (1978a,b) and Blandford \\& Ostriker (1978). The idea was foreshadowed by Hoyle (1960) who postulated that shocks could efficiently accelerate particles but without specifying a mechanism. Parker (1958) and Hudson (1965, 1967) attempted to obtain such mechanism based on pairs of converging shocks and, most notably, Schatzman (1963) constructed a theory based on perpendicular shocks. Contrary to the original mechanism in the convergent shock flow pattern particles interact with the flowing plasma only like in head-on collisions. Mean momentum gain in such interaction is proportional to $U_{1}/c$ ($U_{1}$ is the shock velocity) and hence the process is known as the first-order Fermi acceleration. Efficiency of the first-order relative to the second-order Fermi acceleration equals roughly $U_{1}/V_{A}$, where $V_{A}$ is the Alfv\\'en speed in plasma (cf. Ostrowski \\& Schlickeiser 1996). A shock wave, or briefly a shock, can be described as a sharp transition layer which propagates through plasma with a velocity exceeding the speed of sound and changes its state through the compression. The thickness of the layer is determined by the physical process responsible for thermodynamic parameters transfer from incoming plasma, upstream of the shock, to flowing away plasma, downstream of the shock. In tenuous plasma the transfer proceeds through collective electromagnetic effects and the shock width is of the order of the gyroradius of a thermal ion. In the acceleration process we will consider relativistic particles which move with speeds close to the speed of light and have a gyroradii much larger than thermal ions and consequently they see the shock as a discontinuity. The acceleration processes in nonrelativistic -- $U_{1}\\ll c$ -- shocks yield power-law particles momentum spectra, $f(p)\\propto p^{-\\alpha}$, with a very simple formula for the spectral index of accelerated particles $$ \\alpha = \\frac{3r}{r-1} \\qquad ,\\eqno(1.1) $$ where $$ r = \\frac{\\gamma_{a}+1}{\\gamma_{a}-1+2 M^{-2}} \\eqno(1.2) $$ is a shock compression ratio, $M$ is the shock Mach number and $\\gamma_{a}$ is the plasma adiabatic index. For a strong shock, $M\\rightarrow\\infty$, propagating in a nonrelativistic plasma with $\\gamma_{a} = 5/3$ we have $r\\rightarrow 4-$ and $\\alpha\\rightarrow 4+$. This is encouragingly close to the index of 4.3 inferred for the source of the galactic cosmic rays. Similarly, the acceleration time expressed by a simple diffusive formula is discussed in Section 3. \\noindent {\\bf Relativistic shocks.} A consistent method to tackle the problem of first-order Fermi acceleration in relativistic shock waves was conceived by Kirk \\& Schneider (1987a; see also Kirk 1988). They assumed a parallel shock geometry and that particles are subject to pitch-angle scattering on each side of the shock. By extending the diffusion approximation to higher order terms in the anisotropy of the particle distribution, they obtained solutions to a kinetic equation of the Fokker--Planck type with the isotropic form of pitch angle diffusion coefficient. Since pitch-angle scattering conserves the particle momentum in the fluid frame, the energy spectrum is obtained by matching the solutions at the shock. Their $Q_{L}$ method yielded a particle energy spectral index for strong nonrelativistic shocks as $\\sigma\\simeq 2.0$ -- where $\\sigma \\equiv \\alpha - 2$ -- in agreement with previous results. For relativistic shocks with realistic compression of Heavens \\& Drury (1988), the method produced particle spectra with $\\sigma$ slightly smaller than 2 provided the Lorentz factor of the shock $\\gamma\\leq 5$, and slightly larger at higher $\\gamma$. The authors derived also an angular distribution function at the shock as measured in the upstream and the downstream fluid frame. In the upstream fluid frame the distribution is strongly peaked even for a mildly relativistic case of $U_{1}=0.3c$. Next, Kirk \\& Schneider (1988) extended the analysis by involving both diffusion and large-angle scattering in particle pitch angle. They discovered that -- in relativistic shock waves -- the presence of large angle scattering can substantially modify the spectrum of accelerated particles. An extension of the Kirk \\& Schneider's (1987a) approach to more general conditions in the shock was given by Heavens \\& Drury (1988) who took into consideration the fluid dynamics of relativistic shock waves. They also noted that the resulting particle spectral indices depend on the perturbations spectrum near the shock in contrast to the nonrelativistic case. Kirk \\& Heavens (1989) considered the acceleration process in shocks with magnetic fields oblique to the shock normal (see also Ballard \\& Heavens 1991) by extending the method of Kirk \\& Schneider (1987a). Oblique shock fronts may be conveniently classified into two categories: subluminal and superluminal. In the former ones it is possible to find a Lorentz transformation to a frame of reference in which the electric field is zero in both the upstream and the downstream regions, and the shock front is stationary. In this frame, called the de Hoffman-Teller frame (de Hoffman \\& Teller 1950), the energy of a particle remains constant provided it does not suffer scattering. Superluminal shocks, however, do not admit a transformation to such a frame of reference. They correspond to shock fronts in which the point of intersection of the front with a magnetic field line moves at a speed greater than c. Kirk \\& Heavens used the de Hoffman-Teller frame to consider the subluminal shocks. They showed, contrary to nonrelativistic results again, that such shocks led to flatter spectra than parallel ones approaching the value $\\sigma\\simeq 1.0$ when the shock velocity along the magnetic field $U_{B}\\simeq c$. Their work relied on the assumption of adiabatic invariant $p_\\perp^2 /B$ conservation for particles interacting with the shock, which restricted considerations to the case of a nearly uniform magnetic field upstream and downstream of the shock. A different approach to particle acceleration was presented by Begelman \\& Kirk (1990) who noted that in relativistic shocks most field configurations lead to super-luminal conditions for the acceleration process. In such conditions, particles are accelerated in a single shock transmission by drifting parallel to the electric field present in the shock. Begelman \\& Kirk showed that there is more efficient acceleration in relativistic conditions than that predicted by a simple adiabatic theory. The acceleration process in the presence of finite amplitude magnetic field perturbations was considered by Ostrowski (1991; 1993) and Ballard \\& Heavens (1992). Ostrowski considered a particle acceleration process in the relativistic shocks with oblique magnetic fields in the presence of field perturbations, where the assumption $p_\\perp^2 /B = {\\rm const}$ was no longer valid. To derive particle spectral indices he used a method of particle Monte Carlo simulations and noted that the spectral index was not a monotonic function of the perturbation amplitude enabling the occurrence of steeper spectra than those for the limits of small and large perturbations. It was also revealed that conditions leading to very flat spectra involve an energetic particle density jump at the shock. The acceleration process in the case of a perpendicular shock shows a transition between the compressive acceleration described by Begelman \\& Kirk (1990) and, for larger perturbations, the regime allowing for formation of the wide range power-law spectrum. The Ostrowski (1991) method was based on the `mean field + perturbation' decomposition of magnetic field, i.e. a particle is considered to propagate in the mean field along its undisturbed `adiabatic' trajectory, while the magnetic field inhomogeneities are allowed for by perturbing the trajectory parameters in finite time-steps. The simulations of Ostrowski (1993) were based on the numerical integration of the particle equations of motion in a perturbed magnetic field. Finite-amplitude field perturbations were described with analytic formulae as a superposition of static sinusoidal waves. The analogous simulations by Ballard \\& Heavens (1992) for highly disordered background magnetic fields show systematically steeper spectra in comparison to the above results, as discussed by Ostrowski (1993). In terms of the Lorentz factor of the shock Ballard \\& Heavens found a rough relation $\\alpha \\simeq (3\\gamma +1)/8$ that is valid up to $\\gamma \\simeq 5$. They checked their results considering different power-law fluctuations spectra for the magnetic field and stated that differences between the resulting particle spectra were quite small. The particle spectrum formation in the presence of non-linear coupling of accelerated particles to the plasma flow has been commented by Ostrowski (1994). The shock waves propagating with relativistic velocities rise also interesting questions concerning the cosmic ray acceleration time scale, $T_{acc}$. Until our results published in 1996 (Bednarz \\& Ostrowski 1996 - see chapter 3) there was only somewhat superficial information available about that problem. A simple comparison to the nonrelativistic formula based on numerical simulations shows that $T_{acc}$ relatively decreases with increasing shock velocity for parallel (Quenby \\& Lieu 1989; Ellison et al. 1990) and oblique (Takahara \\& Terasawa 1990; Newman et al. 1992; Lieu et al. 1994; Quenby \\& Drolias 1995; Naito \\& Takahara 1995) shocks. However, the numerical approaches used there, based on assuming the particle isotropization at each scattering, neglect or underestimate a significant factor controlling the acceleration process -- the particle anisotropy. Ellison et al. (1990) and Naito \\& Takahara (1995) included also derivations applying the pitch angle diffusion approach. The calculations of Ellison et al. for parallel shocks show similar results to the ones they obtained with large amplitude scattering. In their computations for the shock with velocity $0.98\\,c$ the acceleration time scale is reduced on a factor $\\sim 3$ with respect to the nonrelativistic formula. Naito \\& Takahara considered shocks with oblique magnetic fields. They confirmed the reduction of the acceleration time scale with increasing inclination of the magnetic field derived earlier for nonrelativistic shocks (Ostrowski 1988). However, their approach neglected the effects of particle cross field diffusion and assumed the adiabatic invariant conservation at particle interactions with the shock. These two simplifications limit their results to the cases with small amplitude turbulence near the shock\\footnote{One should note that the spatial distributions near the shock derived by these authors (their figures 1 and 2) do not show a particle density jump proved to exist in oblique relativistic shocks by Ostrowski (1991). It is also implicitly present in analytic derivations of Kirk \\& Heavens (1989).}. One should also note that comparing some of the mentioned papers the derived time scales to the nonrelativistic expression does not have any clear physical meaning when dealing with relativistic shocks. In the present paper we use pitch angle diffusion approximation for particle transport in the acceleration process. Let us note that some earlier derivations of the acceleration time scale were based on the numerical simulations involving particle scattering at point like scattering centers isotropizing the particle momentum at each scattering, the so called large angle scattering model. This approach does not provide a proper description for the acceleration processes in shock waves moving with velocities comparable to the particle velocity because it removes particle anisotropy and changes the factors related to it. Moreover, against arguments presented in some papers such scattering patterns can not be realized in turbulent magnetic fields near relativistic shocks, where most particles active in the acceleration process are able to diffuse only a short distance below a few particle gyroradii off the shock\\footnote{ However, for the nonrelativistic shock velocity and particles much above the injection energy such approximations can be safely used (cf. Jones \\& Ellison 1991).}. Such distances are most often insufficient to allow for big particle pitch-angle changes. In shocks with oblique magnetic fields such large angle scattering patterns can substantially change the shape of the accelerated particle spectrum with respect to the pitch angle diffusion model. Additionally, as an individual particle interaction with the shock can involve a few revolutions along the magnetic field, the usually assumed adiabatic invariant conservation, $p_\\perp^2/B = {\\rm const}$, cannot be valid for short inter-scattering intervals. \\noindent {\\bf Ultrarelativistic shocks.} The acceleration mechanism described in section 4.1 is quite different from that in the nonrelativistic and mildly relativistic regime so that we distinguish a class of ultrarelativistic shocks if their Lorentz factors $\\gamma\\gg1$. The condition $\\gamma\\gg1$ implies also some simplifications that allow to consider ultrarelativistic shocks as a separate class. First, the magnetic field inclination downstream of the shock is, in practice, always perpendicular to the shock normal as one can derive from Eq. 2.14. Similarly, we can approximate in Eq. 2.13 the ratio of the value the magnetic field downstream of the shock to upstream as $B_{2}/B_{1}\\simeq \\sqrt{8} \\gamma\\sin\\psi_{1}$. A turbulence downstream of the shock could amplify this value and for example assuming equipartition with the thermal pressure downstream, one obtains $B_{2}/B_{1}\\sim (c/V_{A})\\gamma$. Moreover, independently of the plasma composition (proton-electron or electron-positron) the shock velocity relative to the downstream medium is $U_{2}=c/3$ in the limit of large $\\gamma$. The ultrarelativistic shocks are characterized by large anisotropy of particle momentum distribution near the shock that was presented in Bednarz \\& Ostrowski (1998, see Figs.~4.4~-~4.7 below). The values of two main parameters describing the acceleration process, namely the energy spectral index and the acceleration time, are independent of shock conditions if fluctuations upstream of the shock ensure the acceleration process to be effective. They tend to 2.2 (spectral index, Bednarz \\& Ostrowski 1998; also Bednarz \\& Ostrowski 1997a,b) and 1.0 $r_{g}/c$ (acceleration time, Bednarz 1998, 1999). The rough analytical calculations of Gallant \\& Achterberg (1999) are consistent with the Bednarz \\& Ostrowski (1998) paper and Gallant et al. (1998) confirm the value of spectral index for the specific condition of the extremely disordered magnetic field downstream of the shock. Ultrarelativistic shocks are considered as sources of cosmic rays with energies exceeding $10^{20}$ eV and several papers suggested that gamma ray bursts (GRBs) could be sources of these particles (cf. Waxman 1995, Vietri 1995). Vietri (1995) argued that in the Fermi-type acceleration at an ultrarelativistic shock, a particle could have an relative energy gain $\\sim\\gamma^{2}$ per shock crossing cycle. Gallant \\& Achterberg (1999) showed that particles with initial momenta isotropically distributed upstream of the shock gain $\\sim\\gamma^{2}$ energy, but only at the first interaction of the shock. They also showed that for parameters typical of the millisecond pulsars in the neutron star binaries observed in our Galaxy, the gamma ray burst blast wave would decelerate within the pulsar wind bubble, yielding an energy spectrum $\\sigma\\simeq 2$ for the boosted particles. Moreover, this spectrum would typically extend over the energy region $10^{18.5}-10^{20}$ eV, i.e. precisely where the ultra-high-energy cosmic rays (UHECR) component is observed. Bednarz (1999) suggested that such extremely energetic particles could be produced by reflections of the shock directly in GRBs. \\noindent {\\bf Relativistic shocks in astrophysical objects.} Results presented further in the theses could be applied in models of some galactic and extragalactic objects. One of those are active galactic nuclei where a central black hole ejects plasma in the form of relativistic jets. A few tens of blazars has been detected in GeV $\\gamma$-rays by the EGRET detector (von Montigny et. al. 1995). It is widely believed that the $\\gamma$-ray production in blazars is strictly related to the existence of relativistic jets because many of them show superluminal motions (Vermeulen \\& Cohen 1994). Jiang et al. (1998) applied the K\\\"onigl inhomogeneous jet model (Blandford \\& K\\\"onigl 1979; K\\\"onigl 1981) to a sample of quasars and BL Lacs objects and found the Lorentz factors of jets to be a significant part the ultrarelativistic ones. In unified schemes for active galactic nuclei the Fanaroff-Riley type II (FR II) radio sources are formed by AGNs, similarly to blazars, but jets are ejected at higher angles to the line of sight. Evidence that they are relativistic even on tens or hundreds kiloparsec scales suggest that the hotspots in these sources are the downstream regions of relativistic shocks. The recent finding of microquasars in our Galaxy, a class of objects that mimic -- on scales million of times smaller -- the properties of quasars opened new possibilities to study physical processes in accretion disks of black holes. The observations of Mirabell \\& Rodriguez (1994), Tingay et al. (1995), and Hjellming \\& Rupen (1995) confirm the existence of relativistic flows related to these objects, and it is expected that they form relativistic shocks in the interstellar medium. A relativistic wind of magnetized electron-positron plasma blowing from a pulsar with the flow Lorentz factor of $\\sim 10^{6}$ is expected to form a termination shock (e.g. Kennel \\& Coroniti 1984; Gallant \\& Arons 1994 and Chiueh et al. 1998). Non-thermal radiation apparently seen in the class of such objects -- plerions -- suggests the existence of acceleration processes inside the nebula. The Crab Nebula as the young and energetic source is the best plerion to study it. Recent optical observations of Crab Nebula using the Hubble Space Telescope and also the X-ray observations of ROSAT (cf. Hester et al. 1995) show a fascinating structure of jets, a torus of X-ray emission and complexes of sharp wisps. $\\gamma$ ray observations of the Crab Nebula exhibit the existence of extremely energetic electrons near the pulsar (cf. de Jager et al. 1996). The electron energy is a few magnitudes larger than that in the blowing wind so an acceleration mechanism has to take place near the pulsar. Gallant \\& Arons (1994) proposed a mechanism where electrons gain their energy from electromagnetic waves generated by gyrating ions. The mechanism tries to explain wisps at the distance of 10\" from the pulsar but a knot found at 0.7\" (cf. Hester et al. 1995) is not explained in the model. We expect that acceleration mechanism presented by Bednarz \\& Ostrowski (1998) and Bednarz (1999) is able to account for the generation of such energetic electrons at if the ultrarelativistic shock formed near the pulsar. Observations carried out by the Burst and Transient Source Experiment show that GRBs originate from cosmological sources (Meegan et al. 1992 and Dermer 1992). Identification of the host galaxy for the GRB 971214 (Kulkarni et al. 1998) and several other bursts causes there is little doubt now that some, and most likely all GRBs are cosmological. These phenomena are surely related to ultrarelativistic shocks with $\\gamma>10^{2}$ (cf. Woods \\& Loeb 1995). The power-law form of the spectrum often observed at high photon energies suggests the existence of nonthermal populations of energetic particles. Bednarz \\& Ostrowski (1998, see chapter 4 below) showed that such shocks are able to accelerate charged particles and values of their energy spectral indices converge to $\\sigma=2.2$ when $\\gamma\\rightarrow\\infty$ and/or the magnetic turbulence amplitude grows. Below, we will present our results on relativistic shock acceleration published in a series of papers Bednarz \\& Ostrowski (1996, 1998, 1999) and Bednarz (1999). In the next chapter we discuss our numerical simulations and problems with their application to relativistic shock conditions. Then, in chapter 3, the acceleration time scales in mildly relativistic shocks are derived for a number of magnetic field configurations. Chapter 4 is devoted to ultrarelativistic shocks. We show convergence of the particle energy spectral index to the asymptotic value $\\sigma_{\\infty} \\simeq 2.2$ for $\\gamma \\rightarrow \\infty$. We also discuss particle reflections from large $\\gamma$ shocks providing a limit for models involving GRBs as sources of UHECR. The acceleration time scale is also derived. In the last chapter 5 a short summary is presented. \\chapter{Numerical simulations} In order to consider the role of particle anisotropic distributions and different configurations of the magnetic field in shocks the present work is based on the small angle particle momentum scattering approach described by Ostrowski (1991). It enables us to model effects of cross-field diffusion, important in shocks with oblique magnetic fields. Let us note (cf. Ostrowski 1993) that this code allows for a reasonable description of particle transport in the presence of large amplitude magnetic field perturbations also. Some earlier derivations of the acceleration time scale were based on the numerical simulations involving particle scattering at point like scattering centers isotropizing the particle momentum at each scattering. This approach does not provide a proper description for the acceleration processes in shock waves moving with velocities comparable to the particle velocity because it removes particle anisotropy and changes factors related to it. Moreover, against arguments presented in some papers, such scattering pattern can not be realized in turbulent magnetic fields near relativistic shocks, where most particles active in the acceleration process are able to diffuse only a short distance below a few particle gyroradii off the shock\\footnote{ However, for the nonrelativistic shock velocity and particles much above the injection energy such approximation can be safely used (cf. Jones \\& Ellison 1991).}. Such distances are often insufficient to allow for big particle pitch-angle changes occurring with the {\\em point-like} scattering centers which isotropize particle momentum at each scattering. In shocks with oblique magnetic fields such scattering pattern can substantially change the shape of the accelerated particle spectrum with respect to the pitch angle diffusion model. Additionally, as an individual particle interaction with the shock can involve a few revolutions along the magnetic field, the usually assumed adiabatic invariant conservation, $p_\\perp^2/B = {\\rm const}$, cannot be valid for short inter-scattering intervals. Below, the light velocity is used as the velocity unit, $c=1$. As the considered particles are ultrarelativistic ones, $p = E$, we often put the particle momentum for its energy. In the shock we label all upstream (downstream) quantities with the subscript `1' (`2'). The quantities are given in their respective plasma rest frames but subscripts U or D mean that a parameter is measured in upstream plasma rest frame or downstream plasma rest frame, respectively. The shock normal rest frame is the one with the plasma velocity normal to the shock, both upstream and downstream the shock (cf. Begelman \\& Kirk 1990). The acceleration time scales in relativistic shocks (chapter 3), $T_{acc}$, are always given in this particular frame in units of the upstream gyroradius divided by c but the downstream plasma rest frame quantities are used (chapter 4) for the case of ultrarelativistic ones, $t_{acc}$. Here we affix a gyroradius with the index `$g$' when it is a value given for the local {\\em uniform} (tantamount to {\\em mean} or {\\em homogeneous}) magnetic field component. Index `$e$' means the {\\em effective} field including the field perturbations (see Eq. 2.15). Let us denote parallel diffusion coefficient as $\\kappa_{\\parallel}$ and perpendicular diffusion coefficient as $\\kappa_{\\perp}$. Moreover, we will sometimes use shortcuts $\\tau \\equiv \\kappa_\\perp / \\kappa_ {\\parallel}$ and $\\lambda \\equiv \\log_{10} (\\kappa_\\perp / \\kappa_\\| )$. If it will not cause ambiguity we will use symbol $\\psi$ to designate the magnetic field inclination to the shock normal upstream of the shock, instead of $\\psi_{1}$, and the Lorentz factor of the shock as seen upstream of the shock as $\\gamma$, instead of $\\gamma_{1}$. For the same magnetic field fluctuation patterns upstream and downstream of the shock we will use symbols without indices for these patterns. ", "conclusions": "" }, "0005/astro-ph0005305_arXiv.txt": { "abstract": "We present the first infrared (IR) color-magnitude diagram (CMD) for the halo of a giant elliptical galaxy. The CMD for the stars in the halo of NGC~5128 (Centaurus~A) was constructed from HST NICMOS observations of the WFPC2 CHIP-3 field of Soria et al. (1996) to a 50\\% completeness magnitude limit of [F160W]=23.8. This field is located at a distance of 08\\myarcmin50\\myarcsec ($\\sim9$~kpc) south of the center of the galaxy. The luminosity function (LF) shows a marked discontinuity at [F160W]$\\approx$20.0. This is $1-2$ mag above the tip of the red giant branch (TRGB) expected for an old population ($\\sim12$~Gyr) at the distance modulus of NGC~5128. We propose that the majority of stars above the TRGB have intermediate ages ($\\sim2$~Gyr), in agreement with the WFPC2 observations of Soria et al. (1996). Five stars with magnitudes brighter than the LF discontinuity are most probably due to Galactic contamination. The weighted average of the mean giant branch color above our 50\\% completeness limit is [F110W]$-$[F160W]=1.22$\\pm0.08$ with a dispersion of 0.19 mag. From our artificial-star experiments we determine that the observed spread in color is real, suggesting a real spread in metallicity. We estimate the lower and upper bounds of the stellar metallicity range by comparisons with observations of Galactic star clusters and theoretical isochrones. Assuming an old population, we find that, in the halo field of NGC~5128 we surveyed, stars have metallicities ranging from roughly 1\\% of solar at the blue end of the color spread to roughly solar at the red end, with a mean of [Fe/H]=$-0.76$ and a dispersion of 0.44 dex. ", "introduction": "Individual stars are the visible building blocks of galaxies and direct tracers of the galaxy formation process. Massive elliptical galaxies are believed to contain the majority of the oldest stars in the Universe \\citep{tinsley76}. Two main scenarios have been proposed for the formation of these galaxies. In the traditional scenario of single ``monolithic'' collapse, the most massive early-type (E and S0) galaxies form at early times ($z\\gtrsim2$) and evolve with little or no star formation thereafter \\citep{tinsley80, bruzual80,koo81,shanks84, king85, yoshii88, guiderdoni90}. In hierarchical models of galaxy formation, massive galaxies assemble later ($z\\lesssim2$) from mergers of smaller subunits \\citep{white78,white91,lacey93,kauffmann93, cole94, somerville97}. At least some elliptical galaxies show signatures of intermediate-age stars in addition to an old stellar population \\citep{worthey92}. Dynamical studies also suggest that the pressure support of stellar populations in elliptical galaxies could result from mergers \\citep{toomre72,hibbard94}. Moreover, detailed observations of some elliptical galaxies reveal morphological and kinematic signs of a past merging event. These range from the observations of proto-elliptical merger remnants like NGC~7252 to evidence of counter-rotating or otherwise decoupled cores for nearby ellipticals (star-star or star-gas, Bertola 1997). The dominance of old stars and evidence of merging in elliptical galaxies can be understood simultaneously if the youngest stars contribute only a small fraction of the observed integrated light. In fact, \\citet{silva98} show that nearby elliptical galaxies with morphological signatures of recent merger activity have near-IR colors similar to those of galaxies not showing signs of mergers. They conclude that intermediate-age ($1-3$ Gyr) stars contribute at most 10\\%$-$15\\% of the total stellar mass in galaxies with recent merger activity in their sample. Resolving individual stars in elliptical galaxies has become feasible only recently with the advent of the Hubble Space Telescope (HST), since no suitable examples are near enough to be observed from the ground (this will change with the application of adaptive optics systems to large ground-based telescopes). Such detailed information about the stellar content of elliptical galaxies can help us reconstruct their star formation history, and hence, constrain their process of formation. With HST, it is possible to resolve the population of the massive elliptical galaxy NGC~5128 (Centaurus~A) due to its proximity to us. NGC~5128 is part of a group of 25 galaxies composed mainly of dwarf galaxies extending over about 25 degree on the sky (see review by Israel 1998). Its distance estimate relies upon different measurement methods. A distance modulus of $(m-M)_0=27.53 \\pm 0.25$ \\citep{tonry90} is derived from the globular cluster luminosity function (LF) of \\citet{harris86}. The planetary nebula luminosity function yields $(m-M)_0=27.73\\pm0.14$ \\citep{hui93}. The distance modulus of $(m-M)_0=27.48\\pm0.06$ measured by \\citet{tonry90} from surface brightness fluctuations is revised to $(m-M)_0=27.71\\pm0.10$ by \\citet{israel98} using the results of \\citet{tonry91} (the more recent results of \\citet{tonry97} yield a higher revised value of $(m-M)_0=28.18\\pm0.07$). An estimate of $(m-M)_0=27.72\\pm0.20$ is derived from the magnitude of the tip of the red giant branch (TRGB) for stars in the halo observed with HST WFPC2 by Soria et al. (1996; hereafter SMW96). A more recent measurement by Harris et al. (1999; hereafter HHP99) comes from the magnitude of the TRGB for stars in another halo field observed with HST WFPC2 (these data reach $\\sim3$ magnitudes down the RGB). They find a distance modulus of $(m-M)_0=27.98\\pm0.15$, or $D=3.9\\pm0.3$ Mpc. The weighted average of distance moduli of $(m-M)_0=27.75\\pm0.06$ is adopted throughout this paper and corresponds to a distance of $D=3.5\\pm0.1$ Mpc. At this distance, 1 arcmin corresponds to 1018~pc. NGC~5128 is a clear case of an elliptical galaxy showing signs of past merger activity. It is one of the largest known radio galaxies (500 $\\times$ 250 kpc wide) and a massive disk of gas, young stars, and dust is embedded in its center. Within a radius of about 18\\myarcmin from the nucleus, the galaxy shows optical shell structures made up of old disk stars and associated H\\,{\\sc i} shells \\citep{schiminovich94}. This suggests that NGC~5128 might have experienced more than just one merger in its past \\citep{weil96}. The H\\,{\\sc i} shells detected in the outer part of the galaxy seem to show signs of recent star formation \\citep{graham98}. The halo globular clusters in NGC~5128 (region in radius $R<$~24\\arcmin) have a mean metallicity of [Fe/H]=$-0.8\\pm0.2$ (0.5 dex higher than their Milky Way counterparts; Harris et al. 1992) and show a bimodal distribution in color, with peaks at [Fe/H]$\\simeq-1.1$ and $-0.3$ ($R>$ 4\\arcmin; Harris et al. 1992), an effect commonly associated with a merging event. \\citet{jablonka96} find no object with a metallicity higher than solar in a similar sample. The globular clusters in the inner 3~kpc are more metal rich with $-0.6\\leq$[Fe/H]$\\leq+0.1$ than in the outer regions and show signs of a metallicity gradient \\citep{jablonka96, minniti96, alonso97}. Recent HST WFPC2 observations yield a mean value of [Fe/H]$>-0.9$ (SMW96) and [Fe/H]=$-0.41$ (HHP99) for red giant branch stars in the halo of NGC~5128. The age of the current burst of star formation in the disk of NGC~5128 is typically a few times $10^7$ years \\citep{vandenbergh76,dufour79}. In the halo, the presence of $\\sim200$ stars found to be brighter than the TRGB prompted SMW96 to suggest the presence of an intermediate-age population of $\\sim5$~Gyr, making up at most 10\\% in number of the total halo population. Dynamical estimates based on the model of a merger of a late-type galaxy of mass a few times $10^{10} M_{\\odot}$ with NGC~5128 suggest a more recent merging timescale of a few hundred million years \\citep{tubbs80,malin83}. The total mass of the galaxy estimated from velocity dispersion measurements of the planetary nebula system is $M=4\\pm1 \\times 10^{11} M_{\\odot}$, with half of it estimated to be due to dark matter \\citep{mathieu96}. We present in this paper the first IR color-magnitude diagram (CMD) for the halo of a giant elliptical galaxy. Our NICMOS observations of the halo of NGC~5128 probe a range of $\\sim4$ magnitudes down the luminosity function to our 50\\% completeness limit. Section~2 presents the details of the NICMOS observations we obtained in August 1998. The data analysis using the stellar photometry package DAOPHOT is described in Section~3. The importance of doing artificial-stars experiments is emphasized in Section~4, where the completeness functions are presented. The luminosity functions and discontinuities are derived in Section~5. Section~6 presents the first IR CMD of the halo of NGC~5128. This section is divided into four sub-sections discussing the total uncertainties, deriving an estimate of the metallicity spread of the halo stars, and discussing the nature of the bright stars above the TRGB of an old population. A summary of our results and conclusions appears in Section~7. The more technical details of the magnitude system transformations are given in Appendix~A. ", "conclusions": "We have presented the first IR CMD for the halo of a giant elliptical galaxy. Assuming a distance to NGC~5128 of 3.5 Mpc, we have detected a discontinuity in the luminosity function at [F160W]$\\approx$20.0 and have measured IR magnitudes and colors for stars in the halo of NGC~5128 to [F160W]=23.8 (50\\% completeness limit). We are confident that we are not confused by crowding to [F160W]$\\simeq23.5$ based on careful analysis of artificial-stars tests. The weighted average of the mean color of our giant branch above our 50\\% completeness limit is [F110W]$-$[F160W]=1.22$\\pm0.08$ ($(J-H)_{CIT}=0.78$) with a dispersion of 0.19 mag. From our artificial-star experiments we have determined that there is a real spread in color in our CMD. By comparing our data with star cluster giant branches and theoretical isochrones, we were able to constrain the metallicity spread associated with this real color spread. Assuming an old population, we find that, in the halo field of NGC~5128 we surveyed, stars have metallicities ranging from roughly 1\\% of solar at the blue end of the color spread to roughly solar at the red end, with a mean of [Fe/H]=$-0.76$ and a dispersion of 0.44 dex. We assert that the five brightest stars above the SMW96 determination of the TRGB are most probably due to Galactic contamination. We found that the majority of stars above the TRGB of an old population belong to an intermediate-age population ($\\sim2$~Gyr). The presence of an intermediate-age population in the halo of NGC~5128 is consistent with the findings of SMW96. We conclude from our analysis that the IR data are consistent with the halo of NGC~5128 being composed of at least two age populations, a population with ages $\\sim2$~Gyr and an old population. Assuming an old population, we find that the stars have a wide range of metallicities. Future work will combine the WFPC2 CHIP-3 observations of SMW96 and our NICMOS data to examine the multi-color (F555W, F814W, F110W, and F160W) properties of stars in the halo of NGC~5128 and try to reconstruct the galaxy's formation history." }, "0005/astro-ph0005133_arXiv.txt": { "abstract": "We argue that both the extreme soft X-ray excess and the large-amplitude variability of narrow-line Seyfert 1 galaxies (NLS1s) can be explained in the framework of the slim disk model. When the disk luminosity approaches the Eddington luminosity, the disk becomes a slim disk, exhibiting a multi-color blackbody spectrum with a maximum temperature, $T_{\\rm bb}$, of $\\sim 0.2 (M/10^5 M_\\odot)^{-1/4}$keV, and size of the X-ray emitting region, $r_{\\rm bb}$, of $\\sim r_{\\rm S}$ (the Schwarzschild radius). Furthermore, magnetic energy can be amplified up to a level exceeding radiation energy emitted from the disk, causing substantial variability in X-rays by consecutive magnetic flares. ", "introduction": "It has been recently established that NLS1s are characterized by extreme soft excesses and extreme variability \\cite{Lei}, although the origin still remain a puzzle. Since these features are quite reminiscent of those of Galactic black hole candidates during the very high state, i.e.\\ the state in which the luminosity is comparable to the Eddington luminosity, $L_{\\rm E}$, it is natural to assume that NLS1s have a systematically large disk luminosity, $L$ (e.g. \\cite{Bol}). In such a case, the disk is known to become a slim disk \\cite{Abra}. What, then, is the observational signature of the slim disk? ", "conclusions": "" }, "0005/astro-ph0005419_arXiv.txt": { "abstract": "% The existence of antimatter domains in baryon asymmetrical Universe can appear as the cosmological consequence of particle theory in inflationary models with non-homogeneous baryosynthesis. Such a domain can survive in the early Universe and form globular cluster of antimatter stars in our Galaxy. The model of antimatter pollution of Galaxy and annihilation with matter gas is developed. The proton-antiproton annihilation gamma flux is shown to reproduce the observed galactic gamma background measured by EGRET. From comparison with observational data the estimation on the maximally allowed amount of antimatter stars, possibly present in our Galaxy, is found. ", "introduction": "The generally accepted motivation for baryon asymmetric Universe is the observed absence of the macroscopic amounts of antimatter up to the scales of clusters of galaxies, which probably extends on all the part of the Universe within the modern cosmological horizon \\cite{Cohen98}. The modern cosmology relates this baryon asymmetry of the Universe to the process of baryosynthesis., i.e. to the creation of baryon excess in very early Universe \\cite{Sakharov67,Kuzmin70}. In the homogeneous baryon asymmetric Universe the Big Bang theory predicts exponentially small fraction of primordial antimatter. Therefore, any non exponentially small amount of antimatter in the modern Universe is the profound signature for new phenomena, related to the existence of antimatter domains and leads to the respective predictions for antinuclear component of galactic cosmic rays. The most recent analysis finds that the size of possible antimatter domains in baryon symmetrical Universe should be only few times smaller than the modern cosmological horizon to escape the contradictions with the observed gamma ray background \\cite{Cohen98}. The distribution of antibaryon excess, corresponding to relatively small ($<\\,10^{-5}$) volume occupied by it, can arise in inflational models with baryosynthesis and is compatible with all the observational constraints on the annihilation of antimatter in the baryon dominated Universe \\cite{Khlopov00}. The size and amount of antimatter domains is related to the parameters of models of inhomogeneous baryosynthesis (see for review \\cite{Chechetkin82,Khlopov96}). With the account for all possible mechanisms for inhomogeneous baryosynthesis, predicted on the base of various and generally independent extensions of the standard model, the general analysis of possible domain distributions is rather complicated. But the main point of the existing mechanisms of baryosynthesis, important for our aims, is that all of them can lead to inhomogeneity of baryon excess generation and even to generation of antibaryon excess in some regions of space, when the baryon excess, averaged over the whole space, being positive (see reviews in \\cite{Chechetkin82,Khlopov99,Chechetkin87}). On the other hand, EGRET data \\cite{EGRET97} on diffuse gamma background show visible peak around $E_{\\gamma}\\,\\approx\\,70$ MeV in gamma spectrum, which fact can be naturally explained by the decays of $\\pi^0$-mesons, produced in nuclear reactions. Interactions of the protons with gaseous matter in the Galaxy shift the position of such a peak to higher values of gamma energy due to $4$-momentum conservation. At the same time the secondary antiprotons, produced in the cosmic ray interactions with interstellar gas, are too energetic \\cite{Chardonnet96} and their annihilation also cannot explain the observational data. The above consideration draws attention to the model with antimatter globular cluster existing in our Galaxy, which cluster can serve as a permanent source of antimatter due to (anti)stellar wind or (anti)Supernova explosions. The isolated antimatter domain can not form astronomical object smaller than globular cluster \\cite{Khlopov98}. The isolated anti-star can not be formed in the surrounding matter since its formation implies the development of thermal instability, during which cold clouds are pressed by hot gas. Pressure of the hot matter gas on the antimatter cloud is accompanied by the annihilation of antimatter. Thus anti-stars can be formed in the surrounding antimatter only, what may take place when such surrounding has at least the scale of globular cluster. One can expect to find antimatter objects among the oldest population of the Galaxy \\cite{Khlopov98}, in the halo, since owing to strong annihilation of antimatter and matter gas the formation of secondary antimatter objects in the disk component of our Galaxy is impossible. So in the estimation of antimatter effects we can use the data on the spherical component of our Galaxy as well as the analogy with the properties of the old population stars in globular clusters and elliptical galaxies. The total mass of such cluster(s) is constrained from below by the condition of antimatter domain survival in the surrounding baryonic matter because small antimatter domains completely annihilate in the early Universe before the stage of galaxy formation. The upper limit on the total mass of antimatter can be estimated from the condition, that the gamma radiation from annihilation of antimatter with galactic matter gas does not exceed the observed galactic gamma background. The expected upper limit on cosmic antihelium flux from antimatter stars in our Galaxy was found \\cite{Khlopov98,Golubkov00} only factor of two below the modern level of sensitivity in experimental cosmic antihelium searches \\cite{AMS}. In the first approximation the integral effect we study depends on the total mass of the antimatter stars and does not depend on the amount of globular clusters. The only constraint is that this amount does not exceed the observed number of galactic globular clusters (about 200). Assume that antimatter globular cluster, moves along elliptical orbit in the halo. The observed dispersion of velocity of globular clusters is $\\,\\sim\\,300$ km/s and of the long axis of their orbits is $\\,\\sim\\,20$ kpc. This gives $T\\,\\sim\\,2\\cdot 10^{15}$ s as the order of the magnitude for the period of orbital motion of the cluster in the Galaxy. The period the cluster moves along the dense region of the disk with the mean half--width $D\\,\\sim\\,100$ pc depends on the angle at which the orbit crosses the plane of the disk and is of the order \\bma t_d\\ \\sim\\ \\frac{D}{}\\ \\sim\\ 10^{13}\\ s. \\ema This means that the cluster spends not more than 1\\% of the time in the dense region of galactic disk, where the density of gas is of the order of $n_H^{disk}\\,\\sim\\,1$ cm$^{-3}$, moving the most time in the halo with much lower density of the matter gas $n_H^{halo}\\,\\sim\\,5\\cdot 10^{-4}$ cm$^{-3}$. Therefore, we can neglect the probability to find the cluster in the disk region and consider the case when the source of the antimatter is in the halo. One could expect two sources of the annihilation gamma emission from the antimatter globular cluster. The first one is the annihilation of the matter gas captured by the antimatter stars. Another source is the annihilation of the antimatter, lost by the antimatter stars, with interstellar matter gas. It is clear that the gamma flux originating from the annihilation of the matter gas on the antimatter stars surface is negligible. Really, an antimatter star of the Solar radius $R\\,=\\,R_{\\odot}$ and the Solar mass $M\\,=\\,M_{\\odot}$ captures matter gas with the cross section \\bma \\sigma\\ \\sim\\ \\pi\\,R\\,\\left (R\\,+\\,\\frac{2GM}{v^2}\\right ) \\ \\sim\\ 4\\cdot10^{22}\\ cm^2\\,, \\ema \\noindent so that the gamma luminosity of cluster of $10^5$ stars does not exceed $L_{\\gamma}\\,\\le\\,M_{5}\\cdot 10^{29}\\,erg/s$, where $M_5$ is the relative mass of the cluster in units $10^5\\,M_{\\odot}$, $M_{cl}\\,=\\,M_5\\cdot 10^5\\,M_{\\odot}$. Such a low gamma luminosity being in the halo at the distance of about $10$ kpc results in the flux $F_{\\gamma}\\,\\le\\,10^{-13}$ (ster$\\cdot$ cm$^2\\cdot$ s)$^{-1}$ of $1000$ MeV gamma rays near the Earth, what is far below the observed background. This explains why the antimatter star itself can be rather faint gamma source elusive for gamma astronomy and shows that the main contribution into galactic gamma radiation may come only from the annihilation of the antimatter lost by the antistars with the galactic interstellar gas. There are two sources of an antimatter pollution from the (anti-)cluster: the (anti-)stellar wind and the antimatter Supernova explosions. In both cases the antimatter is expected to be spread out over the Galaxy in the form of positrons and antinuclei. The first source provides the stationary in-flow of antimatter particles with the velocities in the range from few hundreds to few thousands km/s to the Galaxy. The (anti)Supernova explosions give antimatter flows with velocities order of 10$^4$ km/s. The relative contributions of both these sources will be estimated further on the base of comparison with the observational data assuming that all the contribution into diffuse gamma background comes from the antimatter annihilation with the interstellar matter gas. We assume in present paper that the chemical content to be dominated by anti-hydrogen and consider the contribution from the annihilation of the antiprotons only. We consider the quasi-stationary case, provided by the presence of a permanent source of the antimatter. The assumption about stationarity strongly depends on the distribution of magnetic fields in the Galaxy, trapping charged antiparticles, annihilation cross section and on the distribution of the matter gas. We shall see that the assumption about stationarity is well justified by existing experimental data and theoretical models. We carried out a careful consideration of the possibility to reproduce the observed spectrum of diffuse gamma background, suggesting the existence of maximal possible amount of the antimatter in our Galaxy. We showed that the predicted gamma spectrum is consistent with the observations. In this case the integral amount of galactic antimatter can be estimated, which estimation leads to definite predictions for cosmic antinuclear fluxes \\cite{Khlopov98,Golubkov00}, accessible for cosmic ray experiments in the nearest future \\cite{AMS}. ", "conclusions": "Let us estimate the intensity of the antiproton source and, as result, the total mass of the hypothetical globular cluster of antistars for three values of the minimal antiproton velocity: $v_{disp}$, $v_{SW}$ and $v_{SN}$. The first case assumes that antiprotons have been decelerated and travel in the halo with velocities equal to the velocity dispersion defined by the galactic gravitational field. The second value of $v_{min}$ is the order of the speed of the fast stellar wind and the third case is the velocity of the particles blown off by the Supernova explosion without possible deceleration. If we integrate over the volume of the whole halo and take into account the antiproton storaging in the halo during the confinement time, we obtain for the integral intensity of the antiproton source $\\dot{M}\\,\\sim\\,\\left (n_0\\,m_p\\,V_{halo}\\right )/t_{conf}$. For above three variants of the minimal velocity of the antiprotons and $t_{conf}\\,\\sim\\,5\\cdot 10^8$ years from \\Eq{aprsw} and \\Eq{aprsn} we obtain the following values of the necessary antiproton source intensity: \\beq{aprate} \\begin{array}{lll} \\dot{M}^{disp} & \\approx & 3.0\\cdot 10^{-9}\\ M_{\\odot}/yr\\\\ \\\\ \\dot{M}^{SW} & \\approx & 8.5\\cdot 10^{-9}\\ M_{\\odot}/yr\\\\ \\\\ \\dot{M}^{SN} & \\approx & 1.0\\cdot 10^{-7}\\ M_{\\odot}/yr\\\\ \\end{array} \\eeq From the analogy with elliptical galaxies in the case of constant mass loss due to stellar wind one has the mass loss $10^{-12}M_{\\odot }$ per Solar mass per year. In the case of stellar wind we find for the mass of the anticluster: \\beq{mclusw} M_{clu}^{SW}\\ \\approx\\ 2\\cdot 10^4\\,M_{\\odot}\\,. \\eeq To estimate the frequency of Supernova explosions in the antimatter globular cluster the data on such explosions in the elliptical galaxies were used \\cite{Khlopov98}, what gives the mean time interval between Supernova explosions in the antimatter globular cluster $\\Delta T_{SN}\\, \\sim\\,1.5\\cdot 10^{15}\\,M_{5}^{-1}$ s. For $M_{5}\\,>\\,1$ this interval is smaller than the period of the orbital motion of the cluster, and one can use the stationary picture considered above with the change of the stellar wind mass loss by the $\\dot{M}\\,\\sim\\,f_{SN}\\cdot M_{SN}$, where $f_{SN}\\,=\\,6\\cdot 10^{-16} \\,M_{5}$ s$^{-1}$ is the frequency of Supernova explosions and $M_{SN}\\,=\\, 1.4\\,M_{\\odot}$ is the antimatter mass blown off in the explosion. Following the theory of Supernova explosions in old star populations only the supernovae of the type I (SNI) take place, in which no hydrogen is observed in the expanding shells. In strict analogy with the matter SNI the chemical composition of the antimatter Supernova shells should include roughly half of the total ejected mass in the internal anti-iron shell with the velocity dispersion $v_i\\,\\le\\,8\\cdot 10^8$ cm/s and more rapidly expanding $v_e\\,\\sim\\,2\\cdot 10^9$ cm/s anti-silicon and anti-calcium external shell. The averaged effective mass loss due to Supernova explosions gives the antinucleon flux $\\dot{N}\\,\\sim\\,10^{42}\\,M_{5}\\,s^{-1}$, but this flux contains initially antinuclei with the atomic number $A\\,\\approx\\,30\\,-\\,60$, so that the initial flux of antinuclei is equal to $\\dot{A}\\,\\sim\\,(2\\,-\\,3)\\cdot 10^{40}\\,M_{5}\\,s^{-1}$. Due to the factor $\\sim\\,Z^2 A^{2/3}$ in the cross section the annihilation life-time of such nuclei is smaller than the cosmic ray life-time, and in the stationary picture the products of their annihilation with $Z\\,<\\,10$ should be considered. With the account for the mean multiplicity $\\,\\sim\\,8$ of annihilation products one obtains the effective flux $\\dot{A}_{eff}\\,\\sim\\,(1.5\\,-\\,2.5)\\cdot 10^{41}\\,M_{5}$ s$^{-1}$, being an order of magnitude smaller than the antiproton flux from the stellar wind. If to take the antimatter stellar wind as small as the Solar wind $(\\dot{M_{\\odot}}\\,=\\,10^{-14}\\,M_{\\odot}$ yr$^{-1})$ this corresponds to the antiproton flux by two orders of magnitude smaller than one chosen above in \\Eq{aprate}, and the antimatter from Supernova should play the dominant role in the formation of galactic gamma background. For the Supernova case we have for the mass of the anticluster the value \\bma M_{clu}^{SN}\\,\\approx\\,4.0\\cdot 10^5\\,M_{\\odot}\\,, \\ema \\noindent which value agrees with the estimation \\cite{Khlopov98}. If we assume that significant fraction of the antiprotons from stellar wind is decelerated up to $v_{disp}$ the respective mass of the globular cluster of antistars can be reduced up to \\bma M_{clu}^{disp}\\,\\approx\\,7\\cdot 10^{3}\\,M_{\\odot}. \\ema It is necessary to make small remark. Namely, in principle, one cannot exclude that the secondary antiprotons produced in $pp$ collisions can be decelerated in the halo magnetic fields up to velocities order of few hundreds km/s. In this case they will also give contribution in the diffuse gamma flux annihilating with the matter gas and the calculations performed in present paper are valid in this case also. In conclusion we can say that the hypothesis on the existence of antimatter globular cluster in the halo of our Galaxy does not contradict to either modern particle physics models or observational data. Moreover, the Galactic gamma background measured by EGRET can be explained by antimatter annihilation mechanism in the framework of this hypothesis. If the mass of such a globular cluster is of order of $10^4\\div 10^5\\ M_{\\odot}$, we can hope that other signatures of its existence like fluxes of antinuclei can be reachable for the experiments in the nearest future. The analysis of antinuclear annihilation cascade is important in the realistic estimation of antinuclear cosmic ray composition but seems to be much less important in its contribution into the gamma background as compared with the effect of antimatter stellar wind. This means that the gamma background and the cosmic antinuclei signatures for galactic antimatter are complementary and the detailed test of the galactic antimatter hypothesis is possible in the combination of gamma ray and cosmic ray studies. \\bigskip {\\it Acknowledgements}. The authors acknowledge the COSMION Seminar participants for useful discussions. The work was partially carried out in framework of State Scientific Technical Programme ''Astronomy. Fundamental Space Research'', Section ''Cosmoparticle Physics''. One of the authors (M.Kh.) expresses his gratitude also to COSMION-ETHZ and AMS-EPICOS collaborations for permanent support. \\vfill\\eject \\newcommand{\\jour}[4]{ #1, {\\bf #2}, #3 (#4)}" }, "0005/astro-ph0005215_arXiv.txt": { "abstract": "The Pyxis globular cluster is a recently discovered globular cluster that lies in the outer halo ($R_{gc} \\sim 40$ kpc) of the Milky Way. Pyxis lies along one of the proposed orbital planes of the Large Magellanic Cloud (LMC), and it has been proposed to be a detached LMC globular cluster captured by the Milky Way. We present the first measurement of the radial velocity of the Pyxis globular cluster based on spectra of six Pyxis giant stars. The mean heliocentric radial velocity is $\\sim\\,$36 km/sec, and the corresponding velocity of Pyxis with respect to a stationary observer at the position of the Sun is $\\sim-191$ km/sec. This radial velocity is a large enough fraction of the cluster's expected total space velocity, assuming that it is bound to the Milky Way, that it allows strict limits to be placed on the range of permissible transverse velocities that Pyxis could have in the case that it still shares or nearly shares an orbital pole with the LMC. We can rule out that Pyxis is on a near circular orbit if it is Magellanic debris, but we cannot rule out an eccentric orbit associated with the LMC. We have calculated the range of allowed proper motions for the Pyxis globular cluster that result in the cluster having an orbital pole within $15^{\\circ}$ of the present orbital pole of the LMC and that are consistent with our measured radial velocity, but verification of the tidal capture hypothesis must await proper motion measurement from the Space Interferometry Mission or HST. A spectroscopic metallicity estimate of $[$Fe/H$] = -1.4\\pm0.1$ is determined for Pyxis from several spectra of its brightest giant; this is consistent with photometric determinations of the cluster metallicity from isochrone fitting. ", "introduction": "Evidence continues to accumulate that the outermost Milky Way globular clusters may not have originated in the same process that formed the inner globular clusters. Based on the recognition that the second parameter effect of horizontal branch morphology in globular clusters is found predominantly among outer halo ($R_{gc} > 8$ kpc) clusters, \\cite{SZ} proposed that the outermost globular clusters may have formed in chemically distinct ``fragments'' that later fell into the Milky Way halo. Building on the suggestion by \\cite{kd76} that several red horizontal branch (second parameter) globular clusters were potentially associated with the Magellanic Plane group of dwarf galaxies, \\cite{srm94} showed that there is a planar alignment between a particular sample of second parameter globular clusters and the Milky Way dwarfs. Recently, \\cite{palma99} reaffirmed that there may be a dynamical relationship between the second parameter globular clusters and the Milky Way dwarf satellites. The Pyxis globular cluster (\\cite{idk95}; \\cite{d95}) at $R_{gc} = 41$ kpc (Sarajedini \\& Geisler 1996) defines the inner edge of the prominent gap in the globular cluster radial distribution between $40 \\lesssim R_{gc} \\lesssim 60$ kpc. The presence of this gap has been used to argue that the primordial Galactic globular cluster system ends at $\\sim\\,$40 kpc while the distant, $R_{gc} > 60$ kpc clusters originated in Galactic satellite dwarf galaxies (e.g., \\cite{zinn85}). Although Pyxis lies among the ``inner group'' of globular clusters (i.e., inside the gap in $R_{gc}$), Irwin et al.\\ (1995) propose that Pyxis may be a captured LMC globular cluster based on the young age they infer for the globular cluster and on its proximity to the plane of the LMC orbit derived from the \\cite{jones} proper motion. Further support for the tidal capture hypothesis comes from \\cite{palma99}, where a statistical analysis of the likely orbital poles of the Galactic satellite galaxies and the globular clusters identifies Pyxis, NGC 6229, NGC 7006, and Pal 4 as the globular clusters most likely to share a common orbital pole with either the Magellanic Plane galaxies (the LMC, the SMC, Draco, and Ursa Minor) or the Fornax--Leo--Sculptor Stream galaxies. However, these postulations on the origin of Pyxis have been made without the benefit of any kinematical data on the cluster. Although deep photometry of Pyxis exists (Sarajedini \\& Geisler 1996), no spectroscopic observations have been published. Indeed, Pyxis is one of the last few known globular clusters lacking a radial velocity (cf.\\ Harris 1996). We report here on du~Pont 2.5-m Telescope spectroscopic observations of Pyxis stars (\\S 2). With our derived radial velocity for the cluster, we re-address the stripped LMC hypothesis for Pyxis' origin (\\S 3), but point out that, in the end, we can only make predictions on the proper motions expected under this scenario. Unfortunately, the proper motion is required for a definitive solution to the question of the cluster's origin. ", "conclusions": "It has been proposed since its discovery that the Pyxis globular cluster may have been captured by the Milky Way from the Magellanic Clouds. If the space motion for Pyxis were known, a comparison of the position of its orbital pole with respect to the LMC as well as a comparison of its angular momentum and orbital energy to that of the LMC would allow one to determine if the two objects share similar orbits. Although only one component of the space motion of Pyxis is now measured, some constraints can be placed on its possible orbit in the tidal capture scenario. A circular orbit with an orbital pole at $(l,b)=(163,-22)^{\\circ}$ is completely ruled out by the measured radial velocity. However, we have shown here that the large radial velocity of Pyxis with respect to a stationary observer at the position of the Sun does not rule out the possibility that the cluster was captured from the LMC since a reasonable range of viable orbits with $e \\sim\\,$0.8 exist for Pyxis that are also similar in energy and angular momentum to that of the LMC. No suitable first epoch plate material is known to exist for Pyxis, so an attempt to measure its proper motion to better determine the likelihood that Pyxis may be a captured LMC globular cluster will require precise observations with the HST or the Space Interferometry Mission (SIM). Although proper motions are not available for the majority of the outer halo globular clusters, their spatial distribution has been used to argue that they are likely to have been accreted into the halo (e.g., Majewski 1994, Palma et al.\\ 2000). Recently, Dinescu et al.\\ (2000) have measured a proper motion for the young globular cluster Pal 12 and they find that its orbit is what one would expect if it had been captured from the {\\em Sagittarius} dwarf galaxy. An accretion origin of the outer halo, second parameter horizontal branch globular clusters is often invoked to explain the possible younger age of some of these objects (where youth is inferred either from the second parameter effect itself or from relative age estimates determined from the cluster CMDs). The physical mechanism that causes the second parameter effect in globular clusters is still unknown: Although it is now generally agreed that there are indeed some globular clusters with anomalously young ages, age differences alone may not be enough to explain the second parameter effect. Whether or not the physical mechanism that causes the effect is age, the possibility that conditions somehow favor the formation of second parameter globular clusters preferentially in Milky Way satellite galaxies (which later get accreted by the Galaxy) may explain the source of the differences between second parameter and non-second parameter globular clusters. The age measurement for the Pyxis globular cluster by \\cite{sg96}, $13.3\\pm1.3$ Gyr, suggests that it is younger by $\\sim\\,$3 Gyr than the oldest Milky Way globular clusters when measured on the same age scale. Recently, age measurements for the oldest LMC globular clusters have been made (\\cite{olsen98}) using a different technique than that used for Pyxis, but their average age of $15.3\\pm1.5$ Gyr places them similar in age to the oldest Milky Way globular clusters, when calibrated onto the same absolute age scale. Another study of a different sample of LMC clusters (\\cite{johnson99}) also finds the oldest LMC clusters to be as old as the old Milky Way clusters. Thus, we may conclude that typical LMC clusters are older than Pyxis. However, at least one of the clusters in the Olsen et al.\\ (1998) sample is $\\sim\\,$2 Gyr younger than the others (NGC 1898), which makes it similar in age to Pyxis. Therefore, it is not impossible to place Pyxis in the ``LMC family'' of clusters from age arguments, though it does appear that Pyxis would be at the young end of the age range for old LMC clusters. It may be noted, however, that the current orbital pole of the Small Magellanic Cloud (SMC) is also very near the intersection point of the poles of the LMC and Pyxis (see Figure 2). Since the SMC is more fragile due to its weaker gravitational potential, perhaps a more attractive origin for Pyxis is from stripping of the SMC rather than the LMC. Recent studies of SMC globular clusters have found that the SMC clusters show a range in ages (e.g., \\cite{sfrz98}, \\cite{msf98}) including at least one cluster with an age similar to Pyxis (NGC 121). The orbital energy of the SMC has a larger magnitude and a smaller error bar than that of the LMC, so not all of the orbits produced from a proper motion in the shaded region in Figure 4 have orbital energies similar to expectations of SMC debris. Only the orbits having proper motions found in the inner part of the shaded region, with a total magnitude of the proper motion of $\\sim\\,$0.75 mas/yr, have orbital energies consistent with an SMC capture origin. Since the same orbital energy and age arguments applied to support the LMC capture origin also apply to the SMC, we consider it a possibility that Pyxis may have been captured from either the LMC or the SMC." }, "0005/astro-ph0005023_arXiv.txt": { "abstract": "Recent 21 cm radio observations of H$_I$ regions in the Small Magellanic Cloud, have revealed spatial power spectra of the intensity, which are quite similar in shape to those previously deduced for the Galaxy. The similarity, in spite the differences in the physical parameters between the Galaxy and the SMC, suggests that the shape of the power spectra reflects some underlying mechanism which is not too sensitive to the environmental specifics. In this paper we present an interpretation for the observational power spectra in terms of a large scale turbulence in the interstellar medium, in which the emitting H$_I$ regions are embedded. The turbulence gives rise to density fluctuations which lead to the observed intensity fluctuations, in the H$_I$ regions. The observational power spectra are used to deduce the turbulence spectral function. In the SMC, the turbulence largest eddies are comparable in scale to the SMC itself. This implies that turbulent mixing should have smoothed out any large scale abundance gradients. Indeed, this seems to be the case, observationally. The turbulence is also expected to amplify and shape up the large scale magnetic field. Indeed, the observational data indicate the existence of a large scale disordered field of the strength expected from energy equilibrium with the turbulent velocity field. The large scale turbulence is most probably generated by instabilities in the large scale flows induced by the tidal close encounter with the LMC $ \\sim 2\\times 10^8{\\rm yr}$ ago. The life-time of the largest eddies is $\\sim 4\\times 10^8{\\rm yr}$ so the turbulence had not yet enough time to decay and persists even though the energy source is no longer there. ", "introduction": "Crovisier \\& Dickey (1983), and Green (1993) have used 21 cm radio observations to derive the intensity spatial power spectra for our Galaxy. Recently, Stanimirovi$\\acute{{\\rm c}}$ \\etal (1999) obtained H$_I$ spatial power spectra for the Small Magellanic Cloud (SMC). Some of the observations were interferometric and yielded the power spectrum directly, and some single dish observations from which the power spectrum was calculated by Fourier transform. The resulting power spectra exhibit a power-law behavior, $p(q)\\propto q^{\\gamma}$, over a wide range of spatial scales: $\\sim 10$--$200{\\rm pc}$ in the Galaxy, and $\\sim 30{\\rm pc}$--$4{\\rm kpc}$ for the SMC. Here $q$ is the absolute value of the 2-dimensional wavenumber in the plane of the sky, and the observational power-law indices are $\\gamma \\sim -(2.8\\div 3) $ for the Galaxy and $\\gamma \\sim -3$ for the SMC. Given the differences in the spatial scales for which the spectra were obtained, and the differences in the physical parameters between the Galaxy and the SMC, the similarity of the indices is remarkable. It suggests the existence of some underlying mechanism that is not sensitive to the specific physical parameters of the environment. Stanimirovi$\\acute{{\\rm c}}$ \\etal (1999) attributed the power-law shape to a fractal structure of H$_I$ clouds hierarchy. The observational power spectra reveal the existence of intensity correlations over a wide range of spatial scales, and this naturally suggests that turbulence in the interstellar medium (ISM) is the sought for underlying mechanism. Such an approach was suggested by Lazarian (1995) and followed up by Lazarian \\& Pogosyan (2000). In this paper we present an interpretation of the observational power spectra in terms of density fluctuations which are generated by a large scale turbulence in the ISM, in which the emitting neutral hydrogen regions are embedded. The model determines the power spectrum for any given spectral function of the underlying large scale turbulence. Therefore, the observational power spectra can be used to study the above large scale turbulence. From the observational power spectra, we deduce the spectral function of the underlying turbulence: $F(k)\\propto k ^{-2}$, with $k$ denoting the absolute value of the 3-dimensional wavenumber. As detailed in the Discussion section, this spectral function is quite common in various components of the ISM and characterizes turbulence in a compressible medium. ", "conclusions": "To interpret the observational spatial power spectra we have assumed that the density fluctuations that produce them are a response to a large scale 3-dimensional turbulence in the ISM, in which the emitting neutral hydrogen regions are embedded. We derived an expression for the intensity power spectrum in terms of the underlying 3-dimensional large scale turbulence. The power spectra were seen to be the sum of contributions from thin slabs of neutral hydrogen, parallel to the plane of the sky, with depth {\\it much smaller} than their extent in the plane of the sky. The depth is also {\\it much smaller} than the distances, along the line of sight, between slabs. In this case, we found that a power-law 3-dimensional turbulence spectrum, $F(k)\\propto k^{-m}$, leads to a power-law 2-dimensional intensity power spectrum, $p(q)\\propto q^{-m-1}$. From the observational values of the power spectrum indices, it follows that $m\\sim 2$ for the SMC and $m=1.8 -2$ for the Galaxy. \\subsection{Turbulence spectrum} This 3-dimensional turbulence spectrum resembles the Kolmogorov spectrum characterized by $m=5/3$ but differs in the value of the index which is closer to $m=2$. For the Kolmogorov spectrum the turbulent velocity on scale $l$ satisfies $v_t(l)\\propto l^{1/3}$, while in the present case $v_t(l)\\propto l^{1/2}$. A velocity-size relation with index $\\sim 0.5$, has been deduced for molecular clouds of a wide range of sizes (Larson 1981; Leung, Kutner \\& Mead 1982; Myers 1983; Dame \\etal 1986; Myers \\& Goodman 1988; Falgarone, Puget \\& Perault 1992; Miesch \\& Bally 1994). A turbulence spectrum characterized by $m=2$ was also found in a H$_{II}$ region by Roy \\& Joncas (1985). It is interesting that this spectrum seems to prevail in different components of the ISM. The common factor in these cases is that the medium is compressible. Indeed, numerical simulations, with and without a magnetic field (Passot, Pouquet, \\& Woodward 1988, V$\\acute{{\\rm a}}$zquez-Samedani, Ballesteros-Paredes, \\& Rodrigu$\\acute{{\\rm e}}$z 1997), indicate that such a spectrum is characteristic to turbulence in a compressible medium (whether or not magnetic field fluctuations are important). The Kolmogorov spectrum strictly applies to the inertial wavenumber range for homogeneous and isotropic incompressible turbulence. In this range there is no energy input from the source generating the turbulence (input is at larger scales) nor energy losses (dissipation is at smaller scales). Thus, turbulent kinetic energy is cascaded from larger to smaller scales due to the nonlinear eddy interactions, at a {\\it constant rate} that is {\\it independent of wavenumber}. The present steeper spectrum, with $m\\sim 2$, indicates that at each wavenumber a part of the cascaded turbulent kinetic energy is transformed to other energy forms. As a result, the turbulent kinetic energy rate is smaller the larger the wavenumber. In a compressible medium, part of the kinetic turbulent energy is indeed transformed to energy associated with density fluctuations. Finally, we note that the emitting neutral regions themselves, modeled here as thin slabs, are probably transient features created and disrupted by the large scale turbulence (Ballesteros-Paredes, V$\\acute{{\\rm a}}$zquez-Samedani, \\& Scalo 1999). Their life-times are of the order of the timescales of the largest eddies. The spatial scales for which the power spectra were obtained are smaller than the above, and therefore the corresponding turbulence timescales are shorter than those of the very largest scales. With respect to the observed scales, the slab can be considered as a quasi-stationary structure in which the observed smaller scale eddies generate the density fluctuations that give rise to the intensity fluctuations. \\subsection{Turbulence Scales and Energy Sources} The scale of the large-scale turbulence in the Galaxy should exceed 200 pc -- the largest scale probed by the power spectra. The power-law shape of the spectra suggests that energy input to the turbulence is on yet larger scales. A probable energy source on these larger scales is the galactic differential rotation, as also suggested by numerical simulations of Wada \\& Norman (1999). In this case, a natural scale will be a fraction of the galactocentric distance. However, for a 3- dimensional turbulence, the largest scale is likely to be comparable to the width of the galactic H$_{II}$ layer of $\\sim$~2kpc (Reynolds 1989; Reynolds 1991). In the SMC, the turbulence can be generated by a host of instabilities related to the bulk flows that result from the tidal interaction with the Galaxy and with the LMC, as well as to the expanding shells and super shells. The existence of large scale correlated density fluctuations in the SMC indicates the existence of turbulence on a scale $\\ges$~4 kpc which is comparable to the size of the SMC . This observational evidence by itself is a very interesting result as the turbulence scale is an order of magnitude larger than that of the turbulence in the LMC, studied by Spicker \\& Feitzinger (1988) on the basis of radial velocities of 21-cm emission. The most plausible source feeding energy to such an extremely large scale turbulence are bulk flows induced by the tidal interactions with the Galaxy and the LMC, notably the very close encounter with the LMC that peaked about $2\\times 10^8{\\rm yr}$ ago and lasted for about $2\\times 10^8{\\rm yr}$ (Gardiner \\& Noguchi 1996). Such large scale bulk flows generate turbulence due to shear instabilities e. g. the Kelvin-Helmholtz instability (KH) or shock induced Richtmier- Meshkov instability(RM). These two mechanisms were shown to induce large scale turbulence in the intracluster gas of merging galaxy clusters by Roettiger \\etal (1998) and by Goldman (1997), respectively. The simulations of Gardiner \\& Noguchi (1996) exhibit large scale (few kpc) velocity gradients with shear values $S\\simeq 5-10{\\rm km/s/kpc}$, consistent with the (long period, population I) Cepheid radial velocity data of Mathewson, Ford,\\& Visvanathan (1986, 1988). One expects that the large scale shear will produce turbulence with the largest eddies size $l_t$ comparable to these spatial scales, and root mean squared turbulent velocities $v_t\\les S l_t=10\\div 15{\\rm km/s}$. This value is comparable to the sound speed of the embedding ionized medium, and supersonic with respect to the H$_I$ regions, thus enabling the density fluctuations in the latter. On smaller scales of $\\les$~1~kpc, also the giant super-shells observed in the SMC, contribute, to the turbulence. The giant supersehlls themselves are probably the result of a star formation burst that followed the interaction with the LMC (Zaritsky \\etal 2000). The time scale for decay of the turbulence $t_d\\sim l_t/v_t$ with a rms turbulent velocity of $v_t\\approx 10 {\\rm km/s}$ and scale $l_t=1 {\\rm kpc}$ is $\\sim 4 \\times 10^8{\\rm yr}$. Thus, we are witnessing a turbulence that didn't have enough time to decay, since the close encounter with the LMC of $\\sim 2\\times 10^8{\\rm yr}$ ago. Moreover, the simulations of Roettiger \\etal (1998) and simulations of RM instability (Rotman, 1991) indicate that continued excitation (even at a lower level than the original one) can keep such a turbulence going for times considerably longer than the above estimate. Such a lower level excitation can be provided by ongoing tidal interactions of smaller' strength. Another question of interest is the near constancy of the slope of the spectra. We know that small scale (from 1kpc and down) ) energy sources exist (supergiant shells, smaller scale supernovae remnants, smaller scale hydrodynamic instabilities) as well as smaller scale energy sinks (e.g. magnetic excitations). Consider a source /sink a at wavenumber $k$ in an interval $\\Delta k$ such that $k\\ges \\Delta k>> k_0$ with $k_0$ corresponding to the largest scale in the turbulence. Denote by $\\epsilon$ and $\\Delta \\epsilon$ the energy rate due to the large scale turbulence and the change due to the small scale source, respectively. Since $F(k)\\propto \\epsilon(k)^{2/3}$ the jump in the turbulence spectral function $$\\frac{\\Delta F(k)}{F(k)}\\sim \\frac{2}{3}\\frac{\\Delta \\epsilon}{\\epsilon}\\ . \\eqno(18)$$ Even for a relative change of $\\sim 20\\%$, the jump will be $\\les 10\\%$, consistent with the scatter of the observational spectra. \\subsection{Implications of the Turbulence on the SMC Magnetic Field and Abundances Gradients} The consequences of this turbulence may be quite important regarding magnification of large scale magnetic fields, and enhancing diffusion coefficients. In the SMC, the turbulence largest eddies are comparable in scale to the size of the SMC. This implies that turbulent mixing should smooth any large scale abundance gradients. This seems indeed to be the case observationally (Dufour 1975; Kobulnicky 1998) . The turbulence is also expected to amplify and shape up the large scale magnetic field. Indeed, the SMC observational data indicate the existence of a large scale disordered field (Haynes \\etal 1991; Ye \\& Turtle 1991). The observed magnetic field is $\\sim 6.5\\mu{\\rm G}$. a value close to that expected from a dynamo amplification to an equilibrium value $B=5\\mu{\\rm G} (v_t/10{\\rm km s}^{-1})(n/1 {\\rm cm}^{-3})^{1/2}$ with $v_t$ the turbulent root mean squared velocity and $n$ the number density in the ionized embedding region. \\subsection{Concluding Remarks} The interpretation of the $H_I$ power spectrum of the SMC in terms of an underlying large scale turbulence seems natural and as noted above is consistent with other observational data. The existence of a large scale turbulence in the SMC gains also support from the recent finding of Stanimirovi$\\acute{{\\rm c}}$ \\etal (2000), that the autocorrelation of the dust column density is also a power law with an index $\\sim -3$. In our interpretation, the dust will be just another passive marker waved by the turbulence and thus should have the same power-law spectral index as the $H_I$ intensity. Moreover, observations of stars of various ages were used by Zaritsky \\etal (2000) and by Kunkel, Demers, \\& Irwinl (2000), to conclude that hydrodynamic interactions, following the recent close passage of the LMC, rather than just gravity are responsible for the morphology and dynamics of the ISM of the SMC. \\vskip 0.5 truecm Comments by the referee and by Dr. Steven Shore have contributed to the improvement of the presentation. This work has been supported by the US-Israel BSF grant 94-314 and Israel Science Foundation grant 561-9911.2." }, "0005/astro-ph0005353_arXiv.txt": { "abstract": "We present a grid of theoretical models where the calculation of absorption line spectral indices in both the blue and red wavelength ranges is done with the same evolutionary synthesis code. We have computed some of these indices: CaT, Na~{\\sc i}, Mg~{\\sc i} in the near infrared and Mg$b$, Mg$_{2}$, Fe52, Fe53, NaD and H$\\beta$, in the blue-visible range, for Single Stellar Population (SSP) of 6 different metallicities, (Z=0.0004, 0.001, 0.004, 0.008, 0.02 and 0.05), and ages from 4 Myr to 20 Gyr. From the comparison of these evolutionary synthesis models with a compilation of elliptical galaxy data from the literature, we find that the observed CaT index follows the blue $\\langle{\\rm Fe}\\rangle$ index rather than Mg$_{2}$ as the models predict. If this implies an {\\sl over-abundance} [Mg/Ca] and we take into account the masses of stars which produce Mg and Ca, these stars could form in a time scale shorter than 5 Myr from the beginning of the star formation process. Alternatively, an IMF biased towards very massive stars (M$> 40 {\\rm M_{\\odot}}$) at the early epoch of star formation in elliptical nuclei has to be assumed. We also suggest to revise the calculation of the nucleosynthesis yield of Magnesium. By using the diagnostic diagram CaT-H${\\beta}$ to disentangle age and metallicity in such populations, we obtain around solar abundances and a sequence of ages between 4 and 16 Gyr for the galaxy sample. ", "introduction": "The study of chemical abundances in elliptical galaxies has traditionally been performed through the analysis of absorption features usually present in their spectra (see the recent review by Henry \\& Worthey \\cite{HW}). The observation of such indices in the blue spectral range ---in particular the so called {\\sl Lick} indices--- has been a very fruitful tool to interpret the physical properties of elliptical galaxies and globular clusters, both assumed to consist of old stellar populations. There are many articles compiling observational data for some of these indices (e.g. Trager et al. \\cite{trager} and references therein), specially in Mg$_{2}$ and $\\langle {\\rm Fe} \\rangle =({\\rm Fe5270}+{\\rm Fe5335})/2$. Some works have also measured other indices in the same spectral region such as Mgb, NaD and H$\\beta$. Evolutionary synthesis models are the tool most frequently used to interpret observed spectra. There are a large number of different models (see Leitherer et al. \\cite{leitherer}, and references therein ) which have become available thanks to the development of theoretical isochrones, computed for a wide range of ages and metallicities. An additional basic input for these models is an atlas of stellar spectra (empirical or theoretical) which provides the spectral energy distribution of each elemental area of the Hertzsprung-Russell Diagram (HRD). If the spectral resolution of the available stellar atlas is good enough, line-strength indices can be measured directly in the final spectrum resulting from the calculation (see Vazdekis \\cite{V99}). When this is not the case, or when the stellar atlas consists of atmosphere models, empirical calibrations of line-strength indices (also known as {\\sl fitting functions}) must be incorporated into the models. These fitting functions are obtained by observing a large sample of stars covering the widest available range of the basic atmospheric stellar parameters (effective temperature T$_{\\rm eff}$, surface gravity $\\log g$, and metallicity ---usually parameterized by ${\\rm [Fe/H]}$---; some authors also include relative abundances ${\\rm [X/Fe]}$ parameters to introduce elemental ratios different from solar). Among the most employed sets of fitting functions are those provided by the Lick group (Gorgas et al. \\cite{Gorgas93}; Worthey et al. \\cite{WFGB94} ---hereafter WFGB94), and those of the Marseille group (Idiart \\& Freitas-Pacheco \\cite{idiart95}, Borges et al. \\cite{BIFT95}; hereafter BIFT95). Examples of evolutionary synthesis models, in which blue spectral indices for single stellar populations (SSP) of different ages and metallicities are computed, are those of Worthey (\\cite{W94}, hereafter W94), Casuso et al. (\\cite{casuso}), Bressan et al. (\\cite{BCT96}, hereafter BCT96), Vazdekis et al. (\\cite{vazdekis96} hereafter VCPB96), and Kurth et al. (\\cite{KFF99}, hereafter KFF99). All these works have employed the polynomial functions of WFGB94. On the other hand, BIFT95 have made use of their own set of fitting functions, which have also been employed in the models of Tantalo et al. (\\cite{tantalo98}). Most of these synthesis models give estimates for the blue-yellow line-strength indices, such as Mgb, Mg$_{2}$, Fe5270 and Fe5335 (sometimes only $\\langle {\\rm Fe} \\rangle$), NaD and H${\\beta}$. An important result is obtained from the study of the locus of data in the plane Mg$_{2}$--Fe (where Fe means an iron index such as Fe5270, Fe5335 or $\\langle {\\rm Fe} \\rangle$). The correlation followed by globular cluster data is adequately reproduced by synthetic models of spectral indices applied to old stellar populations of low metallicities, a result which is not unexpected, since most of the poor-metal stars used to calibrate the spectral index dependence on metallicity are members of these globular clusters. This correlation is steeper than that found for elliptical galaxy nuclei which cannot be fitted by the models even by using the oldest and more metal-rich stellar populations (Burstein et al. \\cite{B84}; Gorgas et al. \\cite{GEA90}; Worthey et al. \\cite{W92}; Davies et al. \\cite{davies93}; Carollo \\& Danzinger \\cite{CD94a},\\cite{CD94b}; Fisher et al. \\cite{fisher}; Vazdekis et al. \\cite{vazdekis97}). In fact, elliptical galaxies are located below the lines in the mentioned diagrams. The usual explanation states that old elliptical galaxies formed stars very quickly in the past, after the production of large quantities of magnesium and other elements by massive stars (through the ejection of metals by Type~II supernovae, SNe), but before the bulk of iron production, which is mainly synthesized by Type~I supernovae resulting from the evolution of low-mass stars. The iron-peak elements appear at least 1~Gyr later than the $\\alpha$-elements in the interstellar medium. This result limits the star formation duration to less than 1 Gyr, after the start of the process Therefore, the so-called {\\sl over-abundance} of Magnesium over Iron is actually an {\\sl under-abundance} of Iron, in terms of the absolute values of total abundances, and since Calcium is also an $\\alpha$-element, it should be expected that Ca indices follow the Mg$_2$ behavior: if CaT and Mg$_{2}$ indices were directly related to the abundances of Calcium and Magnesium, and both elements were mostly produced by Type~II~SNe, one should expect that a large Mg enrichment would also imply a large proportion of Ca in comparison with the Iron abundance, implying [Ca/Fe] $>0$, too. On the contrary, if models are not able to reproduce the observational data, a new explanation should be proposed. However, elliptical galaxies data seem to follow the model predictions in the Ca4455--$\\langle {\\rm Fe} \\rangle$ plane (Worthey \\cite{W98}) thus implying a $\\rm [Ca/Fe]=0$. Since this result is not well understood, here we propose the use of the Calcium Triplet index at $\\sim \\lambda8600$~\\AA, CaT ($\\lambda8542+8662$~\\AA), to test the predictions of theoretical models against observational data in the plane CaT-$\\langle {\\rm Fe}\\rangle$. This point will be discussed in detail in the following sections. It is important to stress that model predictions used to compare the variation of the CaT with other spectral features in the blue spectral region should be obtained with the same computational technique and inputs, i.e., the same models with identical stellar tracks and atlases of Stellar Energy Distributions, SEDs. For this reason, in this paper we will present index predictions obtained with a revised version of the evolutionary synthesis models already presented by Garc\\'{\\i}a-Vargas et al. (\\cite{Paper I}, hereinafter Paper I). There we modeled the equivalent width of the two main lines of the CaII Triplet ($\\lambda\\lambda 8542, 8662$~\\AA), following the index definition given by D\\'{\\i}az et al. (\\cite{DTT}, hereafter DTT), for SSPs with ages ranging from 1 Myr to 17 Gyr, and for 4 different metallicities Z=0.004, 0.008, 0.02 and 0.05. An important conclusion derived from Paper I is that the CaT index is almost constant with age, and only dependent on metallicity for ages older than 1 Gyr, when the IR flux is dominated essentially by giants. This result indicates that the CaT is a potential tool, in conjunction with other age-sensitive indices such as H$\\beta$, to confront the well-known age-metallicity degeneracy problem in old populations. In fact, both indices produce a quasi-orthogonal grid of constant age and metallicity lines (see Fig.~7 in Paper I). In the revised version of the models employed in this work, we have included the computation of the most common indices in the blue spectral region, following the same strategy as that employed in Paper I for the near-IR indices. In order to check our model results about CaT in that work, we compared the predicted indices with the globular cluster data (see paper I), and obtained a dependence of this CaT for the oldest stellar populations on the metallicity similar to that estimated from those data. Unfortunately, there were just a few CaT observations in elliptical galaxies to compare with the model results. In this new piece of work we have compiled data for a sample of elliptical galaxies, for which both the CaT and Lick indices are available in the literature. This will allow us to compare the predictions of the new models with the indices measured in both blue and near-IR spectral regions. This paper is organized in the following way: in Section~2 we give a description of the evolutionary synthesis model, with special attention to changes introduced with respect to Paper~I, and we discuss the criteria followed to select the fitting functions. The comparison of models with data is shown in Section~3. A discussion is performed in Section~4 and finally, our conclusions are presented in Section~5. ", "conclusions": "We have developed an evolutionary synthesis model with which we have produced a grid of models for SSP at 6 metallicities and a wide age range. This code is able to predict indices in the blue-visible spectral range, the classical {\\sl Lick} indices and indices in the near-IR such CaT, Mg~{\\sc i} and Na~{\\sc i} at the same time. We have carefully analyzed the behavior of this index for the coolest stars (Teff $<$ 4000), given by some samples of data available in the literature, obtaining the generic trend of CaT decreasing with effective temperature, in agreement with the extrapolated JCJ92's theoretical functions. Therefore, although systematic effects may still be present in the theoretical predictions, mainly due to potential errors in the extrapolation of these JCJ92 equations, we use the most adequate solution until more reliable fitting functions become available. We have compiled a set of data from the literature for galaxies for which both kind of indices, blue-visible and near-IR, had been observed, and we have compared their predictions with data. We have used our results to study the relationship between different indices for old stellar populations by producing diagnostic diagrams in which observed data and models can be plotted to determine the basic physical properties of the dominant stellar population in these galaxies. We find that most of the galaxies with known data for Mg2, CaT and $\\langle {\\rm Fe} \\rangle$ remain in the CaT -- Mg$_{2}$ plane at the place of Z=0.02, while they seem to have overabundances of Mg. This conclusion is in agreement with other data of Ca4300 in the blue region found by Worthey (\\cite{W98}), but raises the question of how it is possible to have solar metallicity ratio for the calcium element and over-solar ratio for Mg abundance, while both are $\\alpha$-element produced by the same type of massive stars. If we accept the assumption of a relative abundance [Mg/X]$>0$, adopted to explain this kind of diagram and that it is due to a short time scale for the star formation, and we apply the same argument for the [Mg/Ca], this would imply that the star formation time scale in elliptical galaxy nuclei must be shorter than $\\sim 5-10$ ~Myr. Otherwise, we should not find a discrepancy between the data and the models in the later diagnostic, where both indices proceed from the same kind of alpha-elements, which is not the case. We suggest that an update of the nucleosynthesis yields of Mg, increasing the production of Mg for the more massive stars, may solve this problem, by extending the elapsed time between the production of Mg and of Ca and, in consequence, the time scale for the star formation. An alternative explanation might be an IMF biased towards the massive stars ($\\rm M> 40 M_{\\odot}$) in the early phases of star formation in elliptical galaxies. We must keep in mind that the emission over these spectra also may affect Mg$_{2}$ data: they may be reduced by 5 \\% if a careful analysis is done before obtaining the spectral indices Mgb and Mg2. We use the orthogonal diagram CaT -- H${\\beta}$ to date elliptical galaxies and to determinate their abundances, reaching the conclusion that elliptical galaxies are nearby solar in their abundances of calcium, in the same way as for iron, and that their ages range between 8 -16 Gyr. A large campaign of observations in the near-infrared to estimate the CaT index, {\\sl e.g.} in the same set of galaxies given in Davies et al. (1987), would be very useful to date them and to determinate their metallicities/abundances in a clear way." }, "0005/astro-ph0005165_arXiv.txt": { "abstract": "We have discovered a Narrow-Line Seyfert 1 (NLS1) galaxy and identified it with an ultrasoft X-ray source {\\bf (RX~215319-1514)} detected with $ROSAT$. We present its X-ray and optical spectral characteristics. Its redshift is found to be 0.0778$\\pm$0.0002. Its optical spectrum shows fairly strong Fe II emission lines, and its X-ray spectrum has an extremely steep power-law index of $\\Gamma$ $\\geq$4, making it an extremely interesting example of this class of AGNs. No low-energy absorption or variability is observed. The soft X-ray (0.1--2.0 keV) luminosity is estimated to be 3.4--12.5$\\times$10$^{43}$ ergs s$^{-1}$, depending on the spectral model. ", "introduction": "Narrow-line Seyfert 1 (NLS1) galaxies with narrow (FWHM= 500--2000 km s$^{-1}$) emission lines of hydrogen in their optical/UV spectrum tend to have the steepest soft X-ray spectra in the $ROSAT$ band of 0.1--2.0 keV, and form a distinct class of active galactic nuclei (see Boller et al. 1996 and references therein). Catalogues of ultra-soft X-ray sources are, therefore, very useful to find such objects. While carrying out optical spectroscopy of candidate objects in the positional error circles of such sources from the catalogue given by Singh et al. (1995), we have found a new narrow-line Seyfert 1 (NLS1) galaxy. The galaxy is a strong candidate for being a counterpart to the ultra-soft X-ray source -- WGA~J2153.3-1514 (RXJ~2153.3-1514). A brief summary of its optical and X-ray properties is presented below. ", "conclusions": "" }, "0005/astro-ph0005486_arXiv.txt": { "abstract": "We present an algorithm to optimally process uniformly sampled array image data obtained with a nondestructive readout. The algorithm discards full wells, removes cosmic ray (particle) hits and other glitches, and makes a nearly optimum estimate of the signal on each pixel. The algorithm also compresses the data. The computer requirements are modest, and the results are robust. The results are shown and compared to results of Fowler sampled and processed data. Non-ideal detector performance may require some additional code, but this is not expected to cost much processing time. Known types of detector faults are addressed. ", "introduction": "The optimal rejection of cosmic ray glitches from astronomical images is of critical importance for large-area pixelized detectors in space. The detectors (CCDs, etc) are generally stable and repeatable, so they can be carefully calibrated. They are often sensitive to cosmic rays and other radiation. The signals from cosmic rays can be the largest contamination. But the contamination is far from Gaussian; it tends to be dominated by ``glitches\" which have a large effect on one or a few pixels for a short duration. Finding, limiting, and rejecting affected data is a problem common to many observation and data reduction strategies. Cosmic rays affect ground-based detectors as well, but to a much smaller extent. This is partly because the atmosphere and magnetic field act as a shield, eliminating most of the cosmic rays, and partly because atmospheric emission and scattering are variable and limit the extent to which it is possible to uncover and understand other systematic errors. This study was focused on the Next Generation Space Telescope (Stockman \\etal\\ 1998); however, many of the results are applicable over a wide range of observatories. Although we specifically consider IR detectors, the results can be applied to visible light detectors or any detectors where both readout noise and Poisson statistics are present. ", "conclusions": "The algorithm presented here is robust as it excises points far from the line and uses a linear fit on the points near the line. It preserves the maximum dynamic range allowed by the hardware readout and rejects almost all cosmic ray hits. Its adjustable weighting efficiently uses uniformly sampled data, and yet it uses a minimum of computer resources. These studies are supported by the NASA Remote Exploration and Experimentation Project (REE), which is administered at the Jet Propulsion Laboratory under Dr. Robert Ferraro, Project Manager." }, "0005/astro-ph0005329_arXiv.txt": { "abstract": "We present measurements of the angular correlation function of galaxies selected from a $B_J\\sim 23.5$ multicolour survey of two $5^\\circ \\times 5^\\circ$ fields located at high galactic latitudes. The galaxy catalogue of $\\sim 4\\times 10^5$ galaxies is comparable in size to catalogues used to determine the galaxy correlation function at low-redshift. Measurements of the $z\\sim 0.4$ correlation function at large angular scales show no evidence for a break from a power law though our results are not inconsistent with a break at $\\ga 15 h^{-1} {\\rmn Mpc}$. Despite the large fields-of-view, there are large discrepancies between the measurements of the correlation function in each field, possibly due to dwarf galaxies within $z\\sim 0.11$ clusters near the South Galactic Pole. Colour selection is used to study the clustering of galaxies $z\\sim 0$ to $z\\sim 0.4$. The galaxy correlation function is found to strongly depend on colour with red galaxies more strongly clustered than blue galaxies by a factor of $\\ga 5$ at small scales. The slope of the correlation function is also found to vary with colour with $\\gamma \\sim 1.8$ for red galaxies while $\\gamma \\sim 1.5$ for blue galaxies. The clustering of red galaxies is consistently strong over the entire magnitude range studied though there are large variations between the two fields. The clustering of blue galaxies is extremely weak over the observed magnitude range with clustering consistent with $r_0\\sim 2 h^{-1} {\\rmn Mpc}$. This is weaker than the clustering of late-type galaxies in the local Universe and suggests galaxy clustering is more strongly correlated with colour than morphology. This may also be the first detection of a substantial low redshift galaxy population with clustering properties similar to faint blue galaxies. ", "introduction": "The galaxy two-point correlation function is commonly used to measure the structure of the galaxy environment from high redshift until the present epoch. The clustering properties of galaxies in the local Universe are well measured by large representative surveys of the galaxy population (Maddox, Efstathiou \\& Sutherland 1996). Catalogues of galaxies selected by morphology show large variations of the galaxy correlation function with late type galaxies having considerably weaker clustering than early type galaxies (Davis \\& Geller 1976, Loveday {\\it et al.} 1995). The results from studies of galaxies with fainter apparent magnitudes and higher redshifts are less conclusive. Pencil-beam surveys with CCDs and photographic plates from $4{\\rmn m}$ telescopes have measured the amplitude of the $B>22$ correlation function; however, estimates vary by $\\ga 100\\%$ (Infante \\& Pritchet 1995). Also, while $B>22$ surveys show evidence for a rapid decline of the amplitude of the correlation function (Efstathiou {\\it et al.} 1991, Infante \\& Pritchet 1995, Roche {\\it et al.} 1996), $I$ band imaging surveys to similar depths show no evidence for a rapid decrease of the correlation function amplitude (Postman {\\it et al.} 1998). The small areas of previous studies of the faint galaxy correlation function are a possible source of the discrepancy. Large individual structures and voids in the Universe could bias estimates of the correlation function if the field-of-view of the survey is small. The use of single band data to select catalogues of galaxies could suffer from biases as the morphological mix of galaxies will change as a function of limiting magnitude. It is probable that the differing amplitudes of the $B$ and $I$ band correlation functions are due to faint $B$ band data being dominated by weakly clustered blue galaxies (Efstathiou {\\it et al.} 1991) while the $I$ band data has a larger fraction of early type galaxies. In this paper, we use a $B_J\\sim 23.5$ multicolour catalogue of galaxies derived from two $5^\\circ \\times 5^\\circ$ fields to measure the clustering properties of faint galaxies. Section 2 discusses the observations and data reduction used to produce the galaxy catalogue. The method used to determine the angular and spatial correlation functions is described in Sections 3 and 4. Estimates of the correlation function at large angular scales are presented in Section 5. Section 6 discusses the angular correlation function as a function of limiting magnitude for the $U$, $B_J$, $R_F$ and $I$ bands. We show that it is impossible to use a single model to describe the observed clustering of galaxies as different populations of galaxies are measured by each band as a function of limiting magnitude. In Section 7, we use colour selection to measure galaxies with similar stellar populations over a range of redshifts. Our estimates of the spatial correlation function indicate clustering is more strongly correlated with colour than morphology. Our main conclusions are summarised in Section 8. ", "conclusions": "We have used deep multicolour galaxy catalogues of two $5^\\circ \\times 5^\\circ$ fields to study clustering from $z\\sim 0$ to $z\\sim 0.4$. The key conclusions of this paper are: (i) The galaxy spatial correlation function is a power law on comoving scales less than $15 h^{-1} {\\rmn Mpc}$. At larger scales, the correlation function is consistent with a power law though a break in the correlation function is not inconsistent with the data. (ii) Despite the large fields-of-view, there are significant differences in the measured amplitude of the clustering; with the possible exception of blue galaxies. It is clear that fields larger than $100 \\Box ^\\circ$ are required to accurately measure the clustering of $B_J\\sim 22$ galaxies. (iii) Dwarf galaxies in relatively nearby clusters ($z\\sim 0.11$) may effect estimates of faint galaxy correlation function. The effect is colour dependent with the clustering of red galaxies varying significantly between the 2 fields observed. (iv) The clustering properties of galaxies strongly depend on the band used to select the catalogue. Bluer bands show weaker clustering than red bands and there is a rapid decline of the amplitude of the $B_J$ correlation function at faint magnitudes. It is probably inappropriate to fit a simple clustering model to correlation functions derived from single band imaging due to the changing morphological mix with magnitude. (v) The clustering properties of galaxies strongly depend on colour. Such behaviour is consistent with colour being correlated with morphological type. Red galaxies (early types) exhibit stronger clustering with larger values of $\\gamma$ than blue galaxies (late and irregular types). (vi) Blue galaxies have extremely weak clustering with $r_0\\la 3 h^{-1} {\\rmn Mpc}$. This is considerably weaker than the clustering of late type galaxies and is consistent with the clustering of galaxies being more strongly correlated with colour and stellar population than morphology. (vii) The clustering of $B_J<21.5$ blue galaxies is comparable to $B_J>23$ blue galaxies. This is strong evidence for star forming galaxies being weakly clustered from $z\\sim 0.4$ until the present epoch." }, "0005/astro-ph0005603_arXiv.txt": { "abstract": "We present full disk X-ray reflection spectra for two currently popular accretion flow geometries for AGN -- the lamppost model frequently used to discuss the iron line reverberation in AGN, and the model where the X-rays are produced in magnetic flares above a cold accretion disk (AD). The lamppost spectra contain several spectroscopic features characteristic of highly ionized material that are not seen in the X-ray spectra of most AGN. The magnetic flare model, on the other hand, produces reflected spectra that are roughly a super-position of a power-law and a {\\em neutral-like} reflection and iron K$\\alpha$ line, and are thus more in line with typical AGN X-ray spectra. Furthermore, because of the difference in the ionization structure of the illuminated material in the two models, the line equivalent width increases with the X-ray luminosity, $L_x$, for the lamppost, and decreases with $L_x$ for the flare model. In light of these theoretical insights, recent iron line reverberation studies of AGN, the X-ray Baldwin effect, and the general lack of X-ray reflection features in distant quasars all suggest that, for high accretion rates, the cold accretion disk is covered by a Thomson thick, {\\em completely ionized} skin. Because the latter is only possible when the X-rays are concentrated to small emitting regions, we believe that this presents a strong evidence for the magnetic flare origin of X-rays in AGN. ", "introduction": "\\label{sect:intro} Iron K$\\alpha$ emission and the so-called reflection hump centered around $\\sim 30$ keV are perhaps the only significant observational signatures of the presence of cold matter close to the event horizon around accreting black holes in Active Galactic Nuclei (AGN) and Galactic Black Hole Candidates (GBHC). This is why many theory papers investigated X-ray reflection spectra from AGN and GBHCs in great detail under the assumption that the matter is non-ionized or that the density of the illuminated layer is constant (e.g., Lightman \\& White 1988; George \\& Fabian 1991, Ross \\& Fabian 1993; Matt, Fabian \\& Ross 1993, 1996; \\zycki et al. 1994, and additional references in NK). Basko, Sunyaev \\& Titarchuk (1974); Kallman \\& White (1989); Raymond (1993); Ko \\& Kallman (1994); \\rozanska \\& Czerny (1996) relaxed the constant density assumption and all found that the thermal ionization instability (Krolik et al. 1981) plays a central role in establishing the equilibrium temperature and density profiles of the X-ray illuminated gas. Nayakshin, Kazanas \\& Kallman (2000; hereafter NKK) extended results of these authors by providing accurate radiation transfer for illuminating spectra appropriate for the inner part of ADs in AGN and GBHCs. The results of NKK show that a self-consistent gas density determination may provide valuable physical insights into the problem that allow one to put tight constraints on AD theories. As an example, Nayakshin \\& Kallman (2000; NK hereafter) considered the X-ray illumination problem in the three different AD geometries: (1) the ``lamppost'' geometry, where the X-ray source is located above the black hole at some height $h_x$; (2) full corona geometry (e.g., Liang \\& Price 1979) and (3) the two-phase patchy corona model (e.g., Galeev, Rosner \\& Vaiana 1979; Haardt, Maraschi \\& Ghisellini 1994; and Svensson 1996). They pointed out that the reflected spectra and correlations between the X-ray continuum and the atomic features, such as the Fe K$\\alpha$ line and the associated edge are very different for these three geometries. Here we present the full disk spectra for the lamppost and the flare models and broadly compare our theoretical predictions to current observations of AGN. We find that a number of observational facts rules out the lamppost model geometry and, at the same time, supports the magnetic flare origin for the X-rays. (Note that we do not discuss here the Advection Dominated Accretion Flows (e.g., Ichimaru 1977; Rees et al. 1982; Narayan \\& Yi 1994) or modifications of this model due to winds (e.g., Blandford \\& Begelman 1999; Quataert \\& Gruzinov 2000) since these models are not expected to work for many luminous AGN that have broad iron lines -- see the recent review by Fabian et al. [2000]). ", "conclusions": "\\label{sect:discussion} A search for the iron line reverberation in response to continuum variations was the goal of recent observations of MCG--6--30--15 (Lee et al. 2000; Reynolds 2000) and NGC 5548 (Chiang et al. 2000). These authors showed that whereas the continuum X-ray flux was strongly variable on short time scales, the iron K$\\alpha$ line flux stayed roughly constant on these time scales and formally did not vary during the whole observation, which is far longer than the light crossing time of the innermost region. This seems to argue that the reflection takes place very far from the black hole -- e.g., in the putative molecular torus. However, the line profile is {\\em broad}, indicating that most of its flux does come from the region rather close to the black hole. According to Figure \\ref{fig:ew}b, the lamppost geometry and the full corona case cannot explain the lack of iron line reverberation, because that requires the EW of the line to drop with increase in the X-ray flux approximately as EW$\\propto \\fx^{-1}$, whereas it increases with $\\fx$ for the lamppost model and stays constant for the full corona geometry. On the other hand, the magnetic flare model can potentially explain these interesting observations because the skin is completely ionized in this model and thus the EW of the line decreases with the X-ray flux. Nandra et al. (1997) have shown that EW of the iron line monotonically decreases with the increasing $L_x$ for a sample of AGN (the X-ray ``Baldwin'' effect). This study is corroborated by observations of Vignali et al. (1999) who found that the available data for luminous (high $z$) quasars indicate that the latter lack the reflection component and also have no or weak iron lines. This suggests that either there is no cold matter near the X-ray source, or it is hidden by a Thomson thick and completely ionized skin (as suggested by Nayakshin 2000). The former possibility would be somewhat unexpected because the currently popular AD theories either have the cold disk going all the way to the last stable orbit black hole (e.g., magnetic flares, lamppost) for all accretion rates, or suggest that the hot part of the accretion flow diminishes with increase in $L$ and eventually disappears (e.g., Esin, McClintok \\& Narayan 1997; see also \\S 6.1 in Fabian et al. 2000). In other words, in all these theories the luminous quasars are expected to have cold disk persisting down to the last stable orbit, and hence the absence of the line and reflection have to be explained by the ionization physics effects (unless X-ray sources relativistically move away from the disk -- see Beloborodov 1999 -- the more so the higher $L_x$ is). Out of the two models, the lamppost is clearly ruled out while the magnetic flare model seems quite viable. Further, the full disk spectra (Fig. \\ref{fig:spectra}) show that the lamppost model predicts many strong features due to Oxygen, Mg, Si and other elements in soft X-rays, practically as prominent as those in the iron recombination band, and an enormous EW for the iron line, and a notably large iron absorption edge at $\\sim 9$ keV. Such spectra are uncommon for real AGN, which again argues against the lamppost model. On the other hand, crudely speaking, the reflected spectra of the magnetic flare model is a combination of the mirror-like reflection from the completely ionized skin and the cold-like reflection hump and Fe K$\\alpha$ line formed in the cold layers below the skin. These spectra do appear to be neutral (for $\\Gamma \\simlt 2$, see NKK) and are reminiscent of spectra of real AGN and hard state GBHCs (see Done \\& Nayakshin 2000). In addition, while some Seyfert 1 AGN do have extremely broadened iron lines like MCG--6--30--15 (e.g. NGC 3516; Nandra et al. 1999), others do not have the extreme skewed lines expected from the very inner disk (e.g. IC4329a; Done, Madejski \\& \\zycki\\ 2000). As is clear from Figure 2b, the magnetic flare model can explain both of these facts if we suggest that broad iron line AGN are those with low $\\dm \\simlt 0.01$ or so, whereas narrow iron line AGN accrete at larger $\\dm$ such that the skin in their inner disk completely destroys the broad line component. Note that the lamppost model again fails here, because its skin increases the line EW (see Fig. 2a). Therefore, it emerges that the current observations indicate that the ionized skin is indeed present on the top of accretion disks in AGN and that it is {\\em completely} ionized. The latter is only possible if the illuminating X-ray flux is much larger than the disk flux, which is most natural for accretion disks with magnetic flares occurring above the disk. The condition $\\fx\\gg \\fdisk$ is in the fact the basic assumption of the two-patchy phase model (same as the magnetic flare model), because this requirement is central in producing the X-ray continuum spectra as hard as those observed and for a broad range of the parameter space (e.g., Haardt et al. 1994; Stern et al. 1995; Poutanen \\& Svensson 1996; Nayakshin \\& Dove 2000)." }, "0005/astro-ph0005435_arXiv.txt": { "abstract": "We critically examine the basic paradigm for the origin of the 2-3 hr period gap in cataclysmic variables (CVs), i.e., binary systems in which a white dwarf accretes from a relatively unevolved, low-mass donor star. The observed orbital period distribution for $\\sim300$ CVs shows that these systems typically have orbital periods, $P_{orb}$, in the range of $\\sim80$ min to $\\sim8$ hr, but a distinct dearth of systems with 2 $\\la P_{orb}$(hr) $\\la 3$. This latter feature of the period distribution is often referred to as the ``period gap\". The conventional explanation for the period gap involves a thermal bloating of the donor star for $P_{orb} \\ga 3$ hr due to mass transfer rates which are enhanced over those which could be driven by gravitational radiation (GR) losses alone (e.g., magnetic braking). If for some reason the supplemental angular momentum losses become substantially reduced when $P_{orb}$ decreases below $\\sim3$ hr, the donor star will relax thermally and shrink inside of its Roche lobe. This leads to a cessation of mass transfer until GR losses can bring the system into Roche-lobe contact again at $P_{orb} \\sim 2$ hr. We carry out an extensive population synthesis study of CVs starting from $\\sim 3 \\times 10^6$ primordial binaries, and evolving some $\\sim 2 \\times 10^4$ surviving systems through their CV phase. In particular we study current-epoch distributions of CVs in the $\\dot M-P_{orb}$, $R_{2}-P_{orb}$, $M_{2}-P_{orb}$, $q-P_{orb}$, $T_{eff}-P_{orb}$, and $L_{2}-P_{orb}$ planes, where $\\dot M$ is the mass transfer rate, $q$ is the mass ratio $M_2$/$M_1$, and $M_2$, $R_2$, $T_{eff}$, and $L_2$ are the donor star mass, radius, effective temperature, and luminosity, respectively. This work presents a new perspective on theoretical studies of the long-term evolution of CVs. In particular, we show that if the current paradigm is correct, the secondary masses in CVs just above the period gap should be as much as $\\sim 50\\%$ lower than would be inferred if one assumes a main-sequence radius-mass relation for the donor star. We quantify the $M_{2}-P_{orb}$ relations expected from models wherein the donor stars are thermally bloated. Finally, we propose specific observations, involving the determination of secondary masses in CVs, that would allow for a definitive test of the currently accepted model (i.e., interrupted thermal bloating) for the period gap in CVs. ", "introduction": "Cataclysmic variables (CV) are short period binary systems consisting of a white dwarf that accretes matter via Roche-lobe overflow from a low-mass companion star. These objects exhibit a wide range of phenomenology including optical flickering in nova-like systems, dwarf nova eruptions which are thought to be caused by thermal instabilities in the accretion disks, and classical nova explosions which are thermonuclear runaways of the accreted matter on the white dwarf (see, e.g., Warner 1995). The range of observed phenomena depends on the mass transfer rate, the mass ratio of the stellar components, and the magnetic field strength of the accreting white dwarf. The orbital periods of the majority of CVs range from 8 hours down to about 78 minutes, but both longer and shorter period systems are known. In the former case, the donor stars are typically somewhat evolved, while in the latter case, the donor stars are hydrogen exhausted. In this paper we focus on the gap that exists in the orbital period distribution of CVs in the range of $\\sim 2-3$ hr (see, e.g., Warner 1976; Rappaport, Verbunt, \\& Joss 1983, hereafter RVJ; Spruit \\& Ritter 1983; Hameury, et al. 1988a; Warner 1995). The overall evolution of CV binaries is thought to be fairly well understood. The widely accepted explanation for the period gap rests on a mechanism for extracting angular momentum from the binary orbit (e.g., via magnetic braking of the secondary) for periods down to $\\sim3$ hours, followed by a relatively substantial decrease in the angular momentum loss rate\\footnote{ We note that this scenario does not require the angular momentum loss rate to drop suddenly. Instead, it requires only that the timescale over which the angular momentum loss rate decreases must be shorter than the thermal timescale of the donor.}. The donor star, which had been thermally ``bloated'' in response to the mass loss driven by the systemic angular momentum losses, is then able to relax inside of its Roche lobe and mass transfer ceases. The donor star is then thought to reestablish Roche-lobe contact by the time the orbital period has decreased to about 2 hr, after which mass transfer resumes. In this paper we critically examine this paradigm for the creation of the period gap. While most workers believe in the existence of the so-called ``2-3 hr period gap\", a few (e.g., Wickramasinghe \\& Wu 1994; Verbunt 1997, but see also Warner 1995; Wheatley 1995) have questioned its reality, especially when all types of CV are considered; however, we adopt the view that the period gap is a real feature of the CV population as a whole and, as such, requires a theoretical explanation (with observational tests) within the context of their binary evolution. Finally in this regard we note a suggestion by Clemens et al. (1998) that the period gap results from a ``kink'' in the radius-mass relation for main-sequence stars at a mass of about $\\sim 0.25~M_\\odot$ (but see the rebuttal by Kolb, King, \\& Ritter 1998). In \\S 2 we describe the conventional picture of the evolution of a typical CV, including the period gap, and show some illustrative examples of binary evolution calculations for individual systems. In \\S 3 we explore how the binary evolution alters the relations among mass, radius, and orbital period of the secondary star. Specifically we discuss how the main-sequence radius-mass relationship must be modified to include the addition of a ``bloating factor'' that accounts for the changes caused by departures from thermal equilibrium of the mass-losing secondary star. We derive semi-analytic mass-period and radius-period relationships for CV secondaries. In \\S 4 we describe our population synthesis and binary evolution codes, while in \\S 5 we present results from our population synthesis study of CVs in which the binary parameters of the CVs at all phases of their evolution are explored. In \\S 6 we show how assumptions that the donor star has a main-sequence radius-mass relation can lead to large errors in the assignment of the constituent stellar masses, most notably within the orbital period range of 3-5 hr. This period range should encompass the maximum bloating exhibited by a CV secondary compared to a main-sequence star of the same mass. Also in \\S 6 we discuss some specific observational implications resulting from our theoretical work. In particular, a specific test for CVs just above the period gap which will enable us, in principle, to distinguish unambiguously among different possible explanations for the period gap is presented. Finally, we present our summary and conclusions in \\S 7. ", "conclusions": "In this paper we briefly reviewed our current understanding of the secular evolution of CVs through their mass transfer phase, including the currently accepted model for the 2-3 hr ``period gap\" in the orbital period distribution. The results of evolution calculations for a representative sample of individual systems are presented, both as functions of time and of orbital period. A population synthesis code, that starts with some $3 \\times 10^6$ primordial binaries, was then used to generate $\\sim2 \\times 10^4$ systems which evolve successfully through the CV phase of mass transfer. This allows for a more complete exploration of parameter space. The results are displayed as probability densities in the $\\dot M-P_{orb}$, $M_{2}-P_{orb}$, $R_{2}-P_{orb}$ $q-P_{orb}$, and $T_{eff}/L_{2}-P_{orb}$ planes, for CVs at the current epoch. This method of displaying the results can lead to considerable insight into the relationships among the various system parameters. We find that for CVs with orbital periods above 5.5 hr and below the period gap (but above the period minimum) the secondary stars closely follow the main-sequence R-M relation (cf. Beuermann et al. 1998). However, for those with $P_{orb}$ between 3-5.5 hr, the effect of bloating causes them to deviate substantially from this same relation. Among our more interesting results, we have shown that the donor star masses in CVs with orbital periods just above the period gap should be as much as $30-50\\%$ lower than would be inferred on the assumption that the donor stars obey a main-sequence radius-mass relation. This conclusion is only valid if the basic underlying cause of the period gap is thermal bloating of the donor star for systems above the period gap (see \\S \\S 1-6). On the basis of our results, we have proposed a direct observational test of, in particular, the basic paradigm of the period gap and, more generally, our overall understanding of the evolution of CVs. This test involves the challenging, but realistic, task of making relatively accurate (e.g., $10\\%$) determinations of the secondary masses in about a half dozen CVs in the period range of 3-4 hr. If the masses are consistent with the assumption of a main-sequence radius-mass relation for the donor stars, then the currently accepted explanation of the period gap cannot be correct, and the very existence of the gap would pose a major conundrum. If, on the other hand, the masses are mostly consistent with the lower values predicted in this work, then a substantial part of our basic understanding of the secular evolution of CVs will be validated. Previously, much observational attention in CV studies has been focused on determinations of the white dwarf masses. While this is clearly of great interest, we hope with this work to stimulate more interest in the important issue of determining the secondary masses. This research was supported in part by NASA under ATP grants GSFC-070 and NAG5-8500 (to S.B.H.), and NAG5-7479 and NAG5-4057 (to S.A.R.). L.A.N. acknowledges the financial support of NSERC (Canada) and thanks CITA and the University of Toronto for a Reinhardt Fellowship and for their hospitality. We thank M. Politano for a number of useful discussions relating to this work. We are grateful to an anonymous referee who made numerous helpful and insightful comments that led to significant improvements in the paper. We also thank D. MacCannell and G. Esquerdo for their technical assistance. \\begin{deluxetable}{lcccc} \\tablenum{1} \\tablecaption{Summary of Model Parameters} \\tablehead{ \\colhead{Model} &\\colhead{$\\beta$\\tablenotemark{a}} &\\colhead{$\\alpha$\\tablenotemark{b}} &\\colhead{$\\gamma$\\tablenotemark{c}} &\\colhead{$C_{MB}$\\tablenotemark{d}} } \\startdata A - Standard Model & 0 & 1 & 3 & 1 \\\\ B - Reduced Magnetic Braking & 0 & 1 & 3 & 1/2 \\\\ C - High Angular Momentum Losses & 0 & 2 & 3 & 1 \\\\ D - Conservative Mass Transfer & 1 & -- & 3 & 1 \\\\ \\enddata \\tablenotetext{a} {Fraction of mass lost by the donor star that is transferred to, and ultimately retained by, the white dwarf.} \\tablenotetext{b} {Specific angular momentum carried away in nova explosions in units of the specific angular momentum of the white dwarf.} \\tablenotetext{c} {Magnetic braking parameter ``$\\gamma$\" as defined in RVJ.} \\tablenotetext{d} {Proportionality constant in the magnetic braking expression used by RVJ, in units of their ``standard\" value.} \\end{deluxetable}{} \\begin{deluxetable}{lccccc} \\tablenum{2} \\tablecaption{Summary of Polynomial Fits to $M_{2}-P_{orb}$ Relations\\tablenotemark{a}} \\tablehead{ \\colhead{Model\\tablenotemark{b}} &\\colhead{$c_0$} &\\colhead{$c_1$} &\\colhead{$c_2$} &\\colhead{$c_3$} &\\colhead{$M_{2}(P_{orb} = 3$ hr)\\tablenotemark{c}} } \\startdata A\t & 0.005863 & --0.001251 & 0.02353 & 0.0 & 0.214 \\\\ B & --0.4323 & 0.3294 & --0.04942 & 0.005028 & 0.247 \\\\ C & --0.1829 & 0.1031 & 0.01041 & 0.0 & 0.220 \\\\ D & --0.5280 & 0.3856 & --0.06261 & 0.006076 & 0.230 \\\\ Main-Sequence Donor & -- -- & -- -- & -- -- & -- -- & 0.35 \\\\ \\enddata \\tablenotetext{a}{For systems with $P_{orb} >$ 3 hr. Fits are of the form: $M_2 = c_0 + c_{1}P_{orb} + c_{2}P^2_{orb} + c_{3}P^3_{orb}$}. \\tablenotetext{b} {Models are defined in Table 1.} \\tablenotetext{c} {In units of M$_\\odot$.} \\end{deluxetable}{} \\newpage Fig. 1 -- Evolution with time of the mass transfer rate, $\\dot M$, and orbital period, $P_{orb}$, for several model cataclysmic variable systems. Left panel - the evolution of a single CV with initial masses ($M_2 = 0.9 ~M_\\odot$; $M_{WD} = 1.1~M_\\odot$). This system first comes into Roche-lobe contact at $P_{orb} = 6$ hr, evolves through the period gap, to the minimum in $P_{orb}$, and back up to longer periods by $10^{10}$ yr. Right panel - the evolutions of a selection of four other illustrative initial binary constituent masses, $M_2$,$M_{WD}$ = 0.2,0.4 (solid), 0.35,0.35 (dotted), 0.3,0.6 (dashed), and 0.65,0.7 (long dashed), all in units of $M_\\odot$. Fig. 2 -- Evolution with orbital period, $P_{orb}$, of the mass transfer rate, $\\dot M$, secondary mass, $M_2$, and secondary radius, $R_2$, for several illustrative model cataclysmic variable systems. The initial masses for the systems whose evolutions are displayed in the left and right sets of panels are the same as described in Fig. 1. Fig. 3 -- Computed population of cataclysmic variables at the current epoch in the $\\dot M-P_{orb}$ plane for our Standard Model (see Table 1). Here $\\dot M$ is the mass transfer rate, and $P_{orb}$ is the orbital period. The color represents the logarithm of the number of systems in a particular $\\dot M-P_{orb}$ cell, of which there are 100 per hour interval in $P_{orb}$ and 100 per decade in $\\dot M$. The color scale is given on the right side of the figure. We note that the scattered, isolated (red) points in the image {\\it below} the main tracks are minor numerical artifacts of the evolution code that occasionally appear when the Roche lobe makes initial contact with the atmosphere of the donor star. One of these dots corresponds to only $\\sim0.1$ CVs in the entire Galaxy at the current epoch, and so is of no significance. Figs. 4 -- Computed population of cataclysmic variables at the current epoch in the $R_{2}-P_{orb}$ plane for our Standard Model (see Table 1). Here $R_2$ is the radius of the donor star. The color represents the logarithm of the number of systems in a particular $R_{2}-P_{orb}$ cell, of which there are 100 per 0.1 $R_\\odot$ and 100 per hour interval in $P_{orb}$. The color scale is given on the right side of the figure. Figs. 5 -- Computed population of cataclysmic variables at the current epoch in the $M_{2}-P_{orb}$ plane for our Standard Model (see Table 1). Here $M_2$ is the mass of the donor star. The color represents the logarithm of the number of systems in a particular $M_{2}-P_{orb}$ cell, of which there are 100 per 0.1 $M_\\odot$ and 100 per hour interval in $P_{orb}$. The color scale is given on the right side of the figure. Fig. 6 -- Computed population of cataclysmic variables at the current epoch in the $q-P_{orb}$ plane for our Standard Model (see Table 1); $q \\equiv M_{2}/M_{WD}$. The color represents the logarithm of the number of systems in a particular $q-P_{orb}$ cell, of which there are 100 per $\\Delta q = 0.1$ and 100 per hour interval in $P_{orb}$. The color scale is given on the right side of the figure. Fig. 7 -- Computed population of cataclysmic variables at the current epoch in the {$T_{eff}$ -- $P_{orb}$} plane (left panel), and the {$Luminosity - P_{orb}$} plane (right panel) for our Standard Model (see Table 1). We show both the stellar luminosity (top curve) and the core nuclear luminosity (lower distributions). The color represents the logarithm of the number of systems in a particular {$L - P_{orb}$} or {$T_{eff} - P_{orb}$} cell of which there are 100 per decade in $L$, 100 per 500K in $T_{eff}$, and 100 per hour interval in $P_{orb}$. The color scale for both plots is given on the right. Fig. 8 -- Same as Figure 3, except that the population has been scaled by $\\dot M^{1}$ to crudely take into account observational selection effects. Fig. 9 -- Computed orbital period distributions for cataclysmic variables at the current epoch. Left panel - solid curve is the distribution for all systems that appear in Figure 3; the dashed curve was produced by scaling the contributions of each system evolved by $\\dot M^{3/2}$ while the dotted curve is for an $\\dot M^{1}$ scaling (see text). The $\\dot M^{3/2}$- and $\\dot M^{1}$-scaled curves have been shifted vertically by arbitrary amounts for ease in comparison. Right panel - solid curve is for all systems in Fig. 3 which have not yet reached orbital period minimum; dashed curve is for systems that have evolved past the orbital period minimum. Fig. 10 -- Computed distributions of the secondary (right panels) and white dwarf masses (left panels) in cataclysmic variables at the current epoch. The mass distributions are ordered according to the range of orbital period. The dotted histogram (upper right) is for post-period minimum CVs and has been arbitrarily divided by 1.5 for presentation purposes. Fig. 11 -- Computed distribution of mass ratios in cataclysmic variables at the current epoch. The top panel is for systems with orbital periods in the range of 1-3 hr (which includes all post period-gap systems), while the bottom panel is for systems above the period gap. Fig. 12 -- Same as Figure (5), except that in addition to the Standard Model (A), the results for three other models are shown (see Tables 1 and 2): (B) reduced magnetic braking constant; (C) specific angular momentum lost with the ejected matter is twice that of the white dwarf; and (D) conservative mass transfer and retention by the white dwarf. Fig. 13 -- Secondary (donor) mass, $M_2$ as a function of orbital period. The solid curve is based on the assumption that the donor star fills its Roche lobe and has a radius-mass relation appropriate to stars on the main sequence (i.e., eq. [2]) The main-sequence models were generated with the same bipolytrope code that was used to carry out the binary stellar evolution calculations and are discussed in the text. The dashed curves are polynomial fits to the $M_{2}-P_{orb}$ relations derived from the population synthesis study shown in Figure 12. The labels, A through D, correspond to the four different panels in Figure 12. \\newpage" }, "0005/astro-ph0005573_arXiv.txt": { "abstract": "We calculate the exact kinetic evolution of cosmic neutrinos until complete decoupling, in the case when a large neutrino asymmetry exists. While not excluded by present observations, this large asymmetry can have relevant cosmological consequences and in particular may be helpful in reconciling Primordial Nucleosynthesis with a high baryon density as suggested by the most recent observations of the Cosmic Microwave Background Radiation. By solving numerically the Boltzmann kinetic equations for the neutrino distribution functions, we find the momentum-dependent corrections to the equilibrium spectra and briefly discuss their phenomenological implications. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005090_arXiv.txt": { "abstract": "We report the results of initial {\\it FUSE} observations of molecular hydrogen (H$_2$) in translucent clouds. These clouds have greater optical depth than any of the diffuse clouds previously observed for far-UV H$_2$ absorption, and provide new insights into the physics and chemistry of such regions. Our initial results involve observations of HD 73882, a well-studied southern hemisphere star lying behind substantial interstellar material ($E_{B-V}$ = 0.72; $A_V$ = 2.44). We find a total H$_2$ column density, N(H$_2$) = 1.2 $\\times$ 10$^{21}$ cm$^{-2}$, about three times larger than the values for diffuse clouds previously measured in the far-UV. The gas kinetic temperature indicated by the ratio N($J$=1)/N($J$=0) is 58 $\\pm$ 10 K. With the aid of ground-based data to calculate an appropriate multi-component curve of growth, we have determined column densities for all rotational levels up to $J$ = 7. The $J$ $\\geq$ 2 states can be reasonably fitted with a rotational excitation temperature of 307 $\\pm$ 23 K. Both the kinetic and rotational temperatures are similar to those found in previous investigations of diffuse clouds. The ratios of carbonaceous molecules to hydrogen molecules are also similar to ratios in diffuse clouds, suggesting a similar chemistry for this line of sight. ", "introduction": "Molecular hydrogen is the most abundant molecular species in the Galactic interstellar medium. While H$_2$ is an important constituent of diffuse interstellar clouds with visual extinctions $A_V$ less than about 1 magnitude, it becomes the dominant form of hydrogen in the so-called ``translucent'' clouds, which are characterized by $A_V$ in the range 1--5 mag. H$_2$ dominates the dynamics, chemistry, and physics of denser clouds, and is the basic raw material for star formation. Due to the homonuclear structure of H$_2$ and its lack of a dipole moment, ro-vibrational transitions within the electronic ground state are quadrupolar with low spontaneous emission coefficients. They are therefore very difficult to observe except where special excitation conditions produce detectable ro-vibrational emission in the near-IR (e.g. Gautier et al.\\ 1976) or from a space-borne observatory such as ISO, when a long enough line of sight ($A_V$ $\\approx$ 20) allows detection of the mid-IR pure rotational emission from warm diffuse gas (Falgarone et al.\\ 2000). Cold H$_2$, the dominant form, can be widely observed through electronic transitions in the far UV ($\\lambda$ $<$ 1150 \\AA ), or with difficulty in the IR, where the weak quadrupole transition can be observed in absorption only for very high column densities (Lacy et al. 1994). Previous instruments capable of resolving the far-UV absorption bands of H$_2$ include the {\\it Copernicus} mission (Savage et al.\\ 1977; Spitzer \\& Jenkins 1975), the {\\it IMAPS} experiment (Jenkins \\& Peimbert 1997), and the {\\it ORFEUS} spectrometer (Richter et al.\\ 1998; de Boer et al.\\ 1998). All of those, however, were limited to relatively bright stars with $A_V$ $\\la$ 1. The {\\it FUSE} observatory (Moos et al.\\ 2000) is well suited for observations of cold H$_2$ in the diffuse and translucent interstellar medium, owing to its high throughput in the wavelength region from 905 \\AA\\ to 1187 \\AA, encompassing the strong Werner and Lyman bands of H$_2$. A survey of H$_2$ in these regions was therefore identified by the {\\it FUSE} PI team as a project of high priority. We have assembled a list of 35 lines of sight, which sample a wide variety of interstellar environments. These lines of sight, with 0.3 mag $\\la$ $E(B-V)$ $\\la$ 1.1 mag, are also characterized by a range in extinction properties (determined both from $R_V$ and from the shape of the far-UV extinction curve). Most of the lines of sight have information on molecular abundances from previous optical and/or mm-wave data. In conjunction with the {\\it FUSE} observing program, we and several collaborators are obtaining additional data using various ground-based telescopes. We have obtained very high-resolution (R $\\sim$ 150,000--250,000) spectra of interstellar \\ion{K}{1}, \\ion{Na}{1}, \\ion{Ca}{2}, and CH absorption along most of the lines of sight to understand the cloud velocity structure (Welty, Morton, \\& Snow 2000). We (led by DGY) are obtaining moderately high-resolution (R $\\sim$ 50,000), very high S/N, nearly complete optical spectra to measure many of the diffuse interstellar bands, and the molecules CH, CH$^+$, CN, C$_2$, and C$_3$. T. Oka and B. J. McCall are pursuing near-IR measurements of H$_3$$^+$, and we and F. Chaffee are conducting near-IR observations to compare grain mantle features such as water ice, the 3.4-$\\mu$m hydrocarbon band, and the silicate feature at 9.6 $\\mu$m, with the UV data on dust extinction and gas-phase depletions. The ultimate goals of this program are to fully understand the masses of dense interstellar clouds, to probe the chemistry and physics of translucent clouds, and to probe the transition region between diffuse and dense interstellar clouds. In this {\\it Letter}, we describe the first observations of H$_2$ in a translucent cloud line of sight, toward the star HD 73882. The following sections describe the properties of the line of sight deduced from previous observations (\\S 2), the {\\it FUSE} observations and our analysis of the data (\\S 3), and the insights gained from these new data (\\S 4). ", "conclusions": "The total H$_2$ column density of 1.2 $\\times$ 10$^{21}$ cm$^{-2}$ we have measured toward HD 73882 is statistically identical to the largest previously observed via far-UV absorption spectroscopy, 1.1 $\\times$ 10$^{21}$ cm$^{-2}$ toward HD 24534 (Mason, et al.\\ 1976). The molecular fraction $f$ = 2N(H$_2$)/[2N(H$_2$) + N(H I)] is $f$ = 0.65, nearly identical to that found toward Zeta Oph ($f$ = 0.63; Bohlin, et al.\\ 1978). The only known larger value, $f$ = 0.80, toward HD 24534 (Snow et al.\\ 1998), will be redetermined via a planned {\\it FUSE} observation. The abundances of CO, CN, CH, and C$_2$, relative to the total hydrogen abundance, are more similar to those found in diffuse clouds (i.e. toward Zeta Oph) than in dark clouds. Because these molecular abundances represent only a small fraction of the presumed total carbon abundance, we conclude that the clouds toward HD 73882 have not reached the predicted transition point where carbon becomes primarily molecular (e.g. van Dishoeck \\& Black 1986). The limited data presently available also suggest that the depletions toward HD 73882 are similar to those found in cold diffuse clouds, an issue we will explore in more detail in a future paper (Rachford et al. 2000). In principle, the CO abundance can be derived from our {\\it FUSE} spectrum. However, the best candidate lines, from the C--X (0,0) band at 1088 \\AA, lie on the flat portion of the curve of growth where the derived abundance is highly sensitive to the chosen $b$-value. Our derived H$_2$ column density, combined with the line-of-sight extinction properties cited above, show that the ratio of hydrogen, both molecular and total, to dust extinction, is also similar to the value found for diffuse clouds (Bohlin et al.\\ 1978). The ratio of molecules in $J$ = 1 to $J$ = 0 (ortho- to para-hydrogen) is usually interpreted as a measure of the gas kinetic temperature, because in these clouds the collisional timescale for depopulating these levels is shorter than the radiative timescale. Our N(1)/N(0) ratio yields $T_{\\rm kin} = 58 \\pm 10$ K, consistent with the average value $T_{\\rm kin} = 77 \\pm 17$ K found for diffuse clouds with {\\it Copernicus} data (Savage et al.\\ 1977). Shull et al.\\ (2000) find a somewhat larger value in a small sample of {\\it FUSE} data. The excitation diagram in Figure 3 shows that the levels $J \\geq 2$ in this line of sight are not in thermal equilibrium at $T_{\\rm kin}$, but instead follow a different distribution characterized by an excitation temperature, $T_{\\rm ex} = 307 \\pm 23$ K. Non-thermal excitation of the high-$J$ levels has been explained in terms of UV pumping, in which the molecules cascade down through upper rotational and vibrational lines following the absorption of a UV photon (Black \\& Delgarno 1973). In this case, the excited H$_2$ may arise in the outer, optically thin regions of the cloud, rather than in the self-shielded interior. If so, the high-$J$ lines may exhibit small velocity shifts or enhanced $b$-values. However, the limited spectral resolution and uncertain wavelength scale of our {\\it FUSE} data precludes a sensitive search for these effects. It is noteworthy that even in a line of sight thought to be dominated by molecular cloud material within a single cloud complex, the velocity structure is very complicated. This structure, along with the similarities between the current results and previous results for diffuse clouds, suggests the possibility that we looking through several Zeta Oph-type clouds rather than a single very dense cloud. Our analysis of the excited rotational levels of H$_2$ depends heavily on high-resolution ground-based optical absorption and mm-wave emission line data to which we had access. It is thus imperative for future {\\it FUSE} interstellar-line observations to obtain appropriate high-resolution spectra in order to have any hope of unambiguous interpretation of the lower resolution data." }, "0005/astro-ph0005567_arXiv.txt": { "abstract": "We present results from BeppoSAX observations of 10 low redshift quasars. The quasars are part of the Laor et al. (1997) sample of 23 optically selected PG quasars with redshift $<0.4$ and low Galactic absorption along the line of sight. Significant spectral curvature is detected for the 6 quasars with the highest signal to noise ratio in their low energy spectra. The average curvature can be parameterized as a flattening of the underlying power law by $\\Delta\\alpha\\sim0.5$ above $\\sim$ 1 keV. We find that quasars with a steeper soft X-ray (0.1--2 keV) spectrum tend to be steeper also at higher (2--10 keV) energies. The distribution of the best fit 2--10 keV slopes is similar to that found for other radio-quiet AGN and characterized by a large dispersion around the mean: $\\alpha_E \\simeq$ 1.0 $\\pm$ 0.3. Iron K$_{\\alpha}$ lines are detected in 4 quasars. In the narrow--line quasar PG~1115$+$407, the rest frame line energy (6.69$\\pm$0.11 keV) and equivalent width (580$\\pm$280 eV) are consistent with those found in a few low redshift narrow--line Seyfert 1 galaxies (NLSy1). This, together with the similarity of the 0.1--10~keV X-ray continuum, suggests that this quasar is the higher redshift and luminosity analog of a NLSy1. In the broad line quasar PG~0947$+$396, the rest-frame line energy suggests fluorescence from cold iron. The line equivalent width ($>$ 400 eV) is however about 2--3 times higher than that usually found in Seyfert 1 galaxies. The high energy power-law slopes and the iron line emission properties seem to be unrelated to the X--ray luminosity. ", "introduction": "Observations of quasars above 2 keV have in the past been largely devoted to objects selected via X-rays. Their 2--10 keV spectral indices cluster tightly around a ``canonical'' value of 0.9 ($f_E \\propto E^{-\\alpha}$; Williams et al. 1992, Comastri et al. 1992, Reeves et al. 1997, Lawson \\& Turner 1997). This uniformity may be the result of a selection effect (Elvis 1992). Seyfert galaxies, in fact, do indeed show a significantly wider range of 2--10 keV spectral indices (Brandt, Mathur \\& Elvis 1997) and QSOs might show the same effect as reported by recent studies of ASCA observations (George et al. 2000). We have therefore started a program to observe, with BeppoSAX (Boella et al. 1997a), a reasonably large, well defined, and representative sample of optically selected quasars. Our goals are to perform a systematic study of their hard ($E>2$ keV) emission spectra and, taking advantage of the broad BeppoSAX band, to compare them with the soft 0.1--2 keV spectra acquired simultaneously. BeppoSAX has so far observed 10 quasars extracted from the Laor et al. (1997) sample, nine as part of a Core Program and one as part of the Science Verification Phase (SVP). The Laor et al. sample consists of 23 PG quasars selected to have $z<0.4$, Galactic $N_{\\rm H}<1.9 \\times10^{20}$ cm$^{-2}$ and $M_B<-23$. These quasars all have good ROSAT PSPC X-ray spectra, radio fluxes, IR photometry, high S/N optical spectro photometry, IUE and HST spectra (Laor et al. 1997 and references therein). The 0.2--2 keV PSPC spectra of most of these 23 quasars are well fitted by a single power law plus Galactic absorption. The average spectral index is $\\alpha_E=1.62\\pm0.45$, and the $\\alpha_E$ range is 0.9--2.8. The large spread in $\\alpha_E$ permitted the discovery of correlations between the soft X-ray spectral shape and optical emission line properties (Laor et al. 1994, 1997, Boller, Brandt \\& Fink 1996, Ulrich-Demoulin \\& Molendi 1996): steeper soft X-ray quasars tend to have narrower $H_\\beta$ lines, fainter [OIII] emission and stronger FeII emission with respect to $H_\\beta$. Laor et al. (1997) interpret the $\\alpha_E-H_\\beta$ FWHM anticorrelation in terms of a dependence of $\\alpha_E$ on $L/L_{\\rm Edd}$. The line width is inversely proportional to $\\sqrt {L/L_{\\rm Edd}}$ if the broad line region is virialized and if its size is determined by the central source luminosity (see Laor et al. 1997, \\S 4.7). So narrow-line, steep (0.1--2 keV) spectrum AGNs emit close to the Eddington luminosity (Nicastro 2000). A proposed scenario for these sources, as described by Pounds et al. (1995), is that the hard X-ray power law is produced by Comptonization in a hot corona: as the object becomes more luminous in the optical-UV, Compton cooling of the corona increases. The corona becomes colder, thus producing a steeper X-ray power law. If this mechanism is operating, then steep $\\alpha_E$(PSPC) quasars should also have a steep hard X-ray power law that would have been missed by hard X-ray surveys. BeppoSAX 0.1--10 keV observations should help in verifying this picture; test if the correlations between the soft X-ray spectrum and the optical lines properties hold also at high X-ray energies; and study the Iron K emission lines in high luminosity objects, to test the claim of an X-ray Baldwin effect (Iwasawa \\& Taniguchi 1993, Nandra et al. 1997). In this paper we report the analysis of BeppoSAX observations of 10 PG quasars. In \\S 2 we describe the observations and the data analysis; in \\S 3 we present the broad band 0.1--100 keV spectra; in \\S 4 we discuss in detail the 0.1--10 keV spectral shape and compare it with the PSPC results; in \\S 5 we present results on the iron line; in \\S 6 we present our conclusions. A correlation between the 0.1--10 keV spectral shape and the optical line properties is deferred to a future publication. In the paper $H_{0}$=50 km s$^{-1}$ Mpc$^{-1}$ and $q_{0}$=0.5 are assumed. ", "conclusions": "We presented the spectral analysis of the BeppoSAX observations of 10 PG quasars selected from the Laor et al. (1997) sample. The main results can be summarized as follows: \\begin{itemize} \\item[$\\bullet$] Together with PG~1226$+$023 a positive detection of the continuum in the 15--100 keV energy range has been found in two more quasars: PG~0947$+$396 and PG~1115$+$407. However, a possible contamination from hard serendipitous sources in the PDS field of view cannot be ruled out with the present data. \\item[$\\bullet$] The distribution of the 2--10 keV power law energy indices is similar to that observed in other quasars samples. The dispersion around the average value of $\\alpha_E = 1.0\\pm 0.3$ reflects the large spread of the best--fit values over the range 0.5 $< \\alpha_E <$ 1.5. \\item[$\\bullet$] No intrinsic absorption has been detected in any of the objects since the absorbing columns are always compatible with the Galactic values. \\item[$\\bullet$] Significant spectral curvature is present in the BeppoSAX spectra of most of the quasars and is related to the statistics of the observed spectrum: sources with the same energy index over the 0.1--10 keV band are fainter. The average curvature can be parameterized as a spectral flattening by $\\Delta\\alpha\\sim0.5\\pm0.2$ towards high energies requiring two component models. The exact spectral shape and intensity of these components vary from object to object. In a few cases the curvature is due to a strong ``soft excess\" below about 1 keV, while for PG 1626$+$554, independent LECS--PSPC fits suggest the presence of an \"ultra soft excess\" (below ~0.3 keV). The curvature in PG~1402$+$26 is mainly due to a hard tail above $\\sim$ 5 keV rather than a ``soft excess\"; also PG~1626$+$554 shows evidence for a hard tail. A spectral hardening at even higher energies ($>$ 10 keV) could be present in the two quasars detected in the PDS band. \\item[$\\bullet$] The origin of the spectral curvature is likely due to the combined effect of thermal emission from an accretion disk peaking in the far UV and the onset of a Compton reflection component at high energies. The addition of these components does provide a better description of the observed spectra. We also note that quasars with a steeper 0.1--2 keV spectrum tend to be steeper also in the 2--10 keV band, although the effect is detected only at the 68 \\% confidence level. \\item[$\\bullet$] Iron K$_{\\alpha}$ lines are detected in 4 quasars. The rest frame line energy (6.7 keV) and equivalent width (580 eV) of PG~1115$+$407 are consistent with those found in a few low redshift narrow--line Seyfert 1 galaxies (Comastri et al. 1998, Turner et al. 1998, Vaughan et al. 1999, Leighly 1999). The contemporaneous presence of a steep 0.1--10 keV continuum and narrow H$\\beta$ line allows us to classify this object as a relatively high redshift NLSy1. For the optically broad lined quasar PG~0947$+$396, the rest frame line energy of 6.4 keV is similar to that found in many local Seyfert 1 galaxies, suggesting fluorescence from cold iron. The line equivalent width (670 eV) is however higher (at the $2-3 \\sigma$ level) than that usually found in Seyfert 1 galaxies. We investigate the possibility that the high EW of the line, as well as an excess of counts detected in the PDS, are the result of fitting a complex continuum with a simple power-law model. We tried several other models (the inclusion of a Compton reflection component or of a thick and partial covering absorber) but in all cases the line is still required and its EW is never smaller than 400 eV. A line has also been detected in PG~1352$+$183 and PG~1226$+$023. In the first case the line energy and intensity are compatible with the values detected in many Seyfert galaxies. In the second case it could be witnessing a Seyfert-like spectrum diluted in the jet emission (Grandi et al. 1997, Haardt et al. 1998). The detection of significant Fe K--shell emission in 3 radio--quiet quasars with 2--10 keV luminosities in the range 1--5 $\\times$ 10$^{44}$ erg s$^{-1}$ seems to be inconsistent with the trend seen in other radio--quiet AGN (Iwasawa \\& Taniguchi 1993; Nandra et al. 1997, George et al. 2000), where iron lines are more frequently found in lower luminosity objects. \\item[$\\bullet$] A detailed comparison with the recent ASCA observations of a larger sample of radio--quiet quasars (George et al. 2000) is not possible owing to the different responses, sensitivities and energy ranges covered by ASCA and BeppoSAX. We note however that there is a good agreement between the two samples especially for what concerns the average 2--10 keV slope and intrinsic dispersion and the presence of curved convex spectra in most of the objects. The detection with BeppoSAX of a strong iron line in a few relatively high luminosity objects deserves further investigation. The foreseen XMM-Newton quasar surveys will most likely settle several open issues. \\end{itemize}" }, "0005/astro-ph0005421_arXiv.txt": { "abstract": "We study the gamma-ray emissions from an outer-magnetospheric potential gap around a rotating neutron star. Migratory electrons and positrons are accelerated by the electric field in the gap to radiate copious gamma-rays via curvature process. Some of these gamma-rays materialize as pairs by colliding with the X-rays in the gap, leading to a pair production cascade. Imposing the closure condition that a single pair produces one pair in the gap on average, we explicitly solve the strength of the acceleration field and demonstrate how the peak energy and the luminosity of the curvature-radiated, GeV photons depend on the strength of the surface blackbody and the power-law emissions. Some predictions on the GeV emission from twelve rotation-powered pulsars are presented. We further demonstrate that the expected pulsed TeV fluxes are consistent with their observational upper limits. An implication of high-energy pulse phase width versus pulsar age, spin, and magnetic moment is discussed. ", "introduction": "The EGRET experiment on the Compton Gamma Ray Observatory has detected pulsed signals from seven rotation-powered pulsars (e.g., Nolan et al. 1996, and references therein; Kapsi et al. 2000): Crab, Vela, Geminga, PSR B1706-44, PSR B1951+32, PSR B1046-58, and PSR B1055-52, with PSR B0656+14 being a possible detection (Ramanamurthy et al. 1996). The modulation of the $\\gamma$-ray light curves at GeV energies testifies to the production of $\\gamma$-ray radiation in the pulsar magnetospheres either at the polar cap (Harding, Tademaru, \\& Esposito 1978; Daugherty \\& Harding 1982, 1996; Dermer \\& Sturner 1994; Sturner, Dermer, \\& Michel 1995; Shibata, Miyazaki, \\& Takahara 1998; Miyazaki \\& Takahara 1997; also see Scharlemann, Arons, \\& Fawley 1978 for the slot gap model), or at the vacuum gaps in the outer magnetosphere (Chen, Ho, \\& Ruderman 1986a,b, hereafter CHR; Chiang \\& Romani 1992, 1994; Romani and Yadigaroglu 1995; Romani 1996; Zhang \\& Cheng 1998; Cheng \\& Zhang 1999). Effective $\\gamma$-ray production in a pulsar magnetosphere may be extended to the very high energy (VHE) region above 100 GeV as well; however, the predictions of fluxes by the current models of $\\gamma$-ray pulsars are not sufficiently conclusive (e.g., Cheng 1994). Whether or not the spectra of $\\gamma$-ray pulsars continue up to the VHE region is a question which remains one of the interesting issues of high-energy astrophysics. In the VHE region, positive detections of radiation at a high confidence level have been reported from the direction of the Crab, B1706-44, and Vela pulsars (Bowden et al. 1993; Nel et al. 1993; Edwards et al. 1994; Yoshikoshi et al. 1997; see also Kifune 1996 for a review), by virtue of the technique of imaging Cerenkov light from extensive air showers. However, with respect to {\\it pulsed} TeV emissions, only the upper limits have been, as a rule, obtained from these pulsars (see the references cited just above). If the VHE emission originates the pulsar magnetosphere, rather than the extended nebula, significant fraction of them can be expected to show a pulsation. Therefore, the lack of {\\it pulsed} TeV emissions provides a severe constraint on the modeling of particle acceleration zones in a pulsar magnetosphere. In fact, in CHR picture, the magnetosphere should be optically thick for pair production in order to reduce the TeV flux to an unobserved level by absorption. This in turn requires very high luminosities of tertiary photons in the infrared energy range. However, the required fluxes are generally orders of magnitude larger than the observed values (Usov 1994). We are therefore motivated by the need to contrive an outer gap model which produces less TeV emission with a moderate infrared luminosity. High-energy emission from a pulsar magnetosphere, in fact, crucially depends on the acceleration electric field, $\\Ell$, along the magnetic field lines. It was Hirotani \\& Shibata (1999a,b; hereafter Paper I, II) who first solved the spatial distribution of $\\Ell$ together with particle and $\\gamma$-ray distribution functions. They explicitly demonstrated that there is a stationary solution for an outer gap which is formed around the null surface at which the local Goldreich-Julian charge density \\begin{equation} \\rho_{\\rm GJ}= \\frac{\\Omega B_z}{2\\pi c [1-(\\Omega \\varpi/c)^2]} \\label{def_rhoGJ} \\end{equation} vanishes, where $B_z$ is the component of the magnetic field along the rotation axis, $\\Omega$ refers to the angular frequency of the neutron star, $\\varpi$ indicates the distance of the point from the rotation axis, and $c$ is the speed of light. Subsequently, Hirotani (2000a, hereafter Paper IV) investigated the $\\gamma$-ray emission from an outer gap, by imposing a gap closure condition that a single pair produces one pair in the gap on average. He demonstrated that $\\Ell$ becomes typically less than $10\\%$ of the value assumed in CHR and that the resultant TeV flux is sufficiently less than the observational upper limit of the pulsed flux, if the outer gap is immersed in a X-ray field supplied by the blackbody radiation from the whole neutron star surface and/or from the heated polar caps. In this paper, we develop his method to the case when a magnetospheric power-law component contributes in addition to the blackbody components. In the next section, we formulate the gap closure condition. Solving the condition in \\S 3, we investigate the acceleration field and the resultant $\\gamma$-ray emissions as a function of the X-ray field. In \\S 4, we further apply the theory to twelve rotation-powered pulsars and predict the absolute fluxes of TeV emission from their outer gaps. In the final section, we discuss the validity of assumptions and give some implications on pulse profiles of GeV emissions. ", "conclusions": "\\label{sec:discussion} \\subsection{Summary} \\label{sec:summary} To sum up, we have considered the electrodynamic structure of an outer gap accelerator in which relativistic particles emit $\\gamma$-rays via curvature process. Imposing the gap closure condition that a single pair produces one pair in the gap on average, we solve selfconsistently the gap width as a function of the X-ray fields and the pulsar parameters. Once the gap width is known, we can further compute the acceleration field and the resultant $\\gamma$-ray emissions. It was demonstrated that the luminosities of GeV and TeV emissions are a decreasing function of the X-ray energy and number density. We also showed that the expected $\\nu F_\\nu$ fluxes ($<10^{11.5}$Jy~Hz) of IC-scattered, TeV $\\gamma$-rays from the outer gaps of rotation-powered pulsars are less than the observational upper limits, except for Crab and B0656+14. For Crab, energy-dependent particle distribution function should be considered, whereas for B0656+14, more accurate X-ray observations are required. It is concluded that the difficulty of excessive TeV emission, which appears in the CHR picture, does not arise in the present outer gap model. \\subsection{Stability of the Gap} \\label{sec:stability} The outer gap in the present model is stable, regardless of whether the X-ray field is dominated by a surface blackbody or a magnetospheric power-law component. Consider the case when the gap width $H$ slightly increases as an initial perturbation. It increases both $\\Ell$ and $V_{\\rm gap}$, which in turn increases both $E_{\\rm c}$ and $N_\\gamma$. The increase of $E_{\\rm c}$ results in the decrease of $E_{\\rm th}$. \\\\ {\\bf (1)} \\quad When the surface blackbody dominates the X-ray field, the X-ray spectrum and luminosity are unchanged by the perturbation. Therefore, the decrease of $E_{\\rm th}$ implies the decrease of $\\lambda_1$ or $\\lambda_2$ and hence $\\lambda_{\\rm p}$. \\\\ {\\bf (2)} \\quad When the magnetospheric emission dominates, the secondary and tertiary emissions will increase with $E_{\\rm c}$ and $N_\\gamma$; therefore, $N_{\\rm pl}$ increases as well. Accordingly, the decrease of $E_{\\rm th}$ and the increase of $N_{\\rm pl}$ imply a significant decrease of $\\lambda_3$ and hence $\\lambda_{\\rm p}$. \\\\ In either case, it follows that $\\lambda_{\\rm p}$ decreases owing to the initial increase of $H$. Reminding the gap closure condition $H=\\lambda_{\\rm p}/N_\\gamma$, we find a negative feedback which cancels the initial perturbation of $H$. \\subsection{Pulse Sharpness} \\label{sec:sharpness} Let us discuss the expected sharpness of GeV pulses. It seems unlikely that the azimuthal width of the gap increases with decreasing $H$. Therefore, it would be possible to argue that the solid angle in which the primary $\\gamma$-rays are emitted decreases as the arc of the gap along the last open field line (i.e., $2H$) decreases. On these grounds, we can expect a sharp pulse when $h \\ll 1$ holds, such as for Crab and J0822-4300. Qualitatively speaking, the same conclusion can be expected for millisecond pulsars and magnetars. In the case of a millisecond pulsar, its fast rotation shrinks the light cylinder. In such a small-volume magnetosphere, the outer gap is immersed in a dense magnetospheric, power-law X-ray emission. As a result, $\\lambda_{\\rm p}$ decreases to reduce $h$. In the case of a magnetar, its strong magnetic field makes the expansion coefficient $A$ in equation (\\ref{eq:Poisson-2}) be large. Therefore, a very thin ($h \\ll 1$) gap with a strong $\\Ell$ would be expected. In short, for young pulsars, millisecond pulsars, and magnetars, their high-energy pulsations are expected to show sharp peaks. \\subsection{Validity of Assumptions} \\label{sec:valid_assump} First, we reduced the Poisson equation into the one-dimensional form (eq.[\\ref{eq:Poisson-2}]), by assuming $D_\\perp \\gg H$. Let us briefly consider the two-dimensional effect due to the transfield derivative in the Poisson equation. When $D_\\perp$ becomes small, the gap shifts outwards, $\\Ell$ is partially screened, and $H$ enlarges (fig.~12 in Paper I; see also Cheng, Ho, \\& Rudermann 1986a for a screened, or spatially constant $\\Ell$ in a thin gap). Owing to the screened acceleration field, the GeV and TeV fluxes becomes small compared with those obtained in $D_\\perp \\gg H$ case. On these grounds, we can constrain the upper limit of the TeV fluxes in the transversely thick limit, $D_\\perp \\gg H$.'' Secondly, let us discuss the case when the assumption of the vacuum gap breaks down. In this case, the charges in the gap partially cancel the original $\\Ell$ obtained in the vacuum gap (eq. [\\ref{eq:Poisson-2}]). The partially screened $\\Ell$ results in the decrease of the TeV fluxes. On these grounds, we can regard the TeV fluxes presented in the present paper as the firm upper limits. Thirdly, we consider the influence of cyclotron resonance scatterings. For one thing, the {\\it soft} blackbody emission from the whole surface may be scattered to be anisotropic (Daugherty \\& Harding 1989). Such effects are important for polar cap models, because the collision angles ($\\cos^{-1}\\mu_{\\rm c}$) suffer significant corrections. Nevertheless, in an outer gap, such corrections are negligibly small. Moreover, the cyclotron resonance increases the effective emitting area and decreases the temperature. For simplicity, we neglect these two effects in this paper, because they cancel each other. For example, the decreased temperature results in a decrease of the target photons above a certain threshold energy for pair production. On the other hand, the increased emitting area increases the number of target photons above the threshold, thereby cancel the effect of the decreased temperature. What is more, the {\\it hard} blackbody emission from the heated polar caps may be scattered to be smeared out. That is, most of the hard X-rays may be scattered back to the stellar surface owing to cyclotron resonance scatterings and reemitted as soft X-rays (Halpern \\& Ruderman 1993). In this case, the hard component will be indistinguishable with the original soft component due to the neutron-star cooling. Nevertheless, for older pulsars such as B0656+14 and B1055-52, these effect seems to be ineffective probably due to their less dense electrons around the polar cap near the neutron star surface. \\subsection{Gamma-ray Luminosity vs. Spin-down Luminosity} \\label{sec:Lgev-Lspin} Curvature-radiated luminosity, $L_{\\rm GeV}$, has a weak dependence on the spin-down luminosity, $L_{\\rm spin}$, if we fix the transfield thickness of the gap, $D_\\perp/\\rlc$. In another word, the evolution of $D_\\perp/\\rlc$ is crucial to discuss the $L_{\\rm GeV} \\propto L_{\\rm spin}{}^{0.5}$ relation (Thompton et al. 1994; Nel et al. 1996). To solve $D_\\perp$, we must analyze the two-dimensional Poisson equation on the poloidal plane; however, it is out of the scope of the present paper. \\subsection{Synchrotron Radiation Below 10 MeV Energies} \\label{sec:sycnrotron} As we have seen, the accelerated particles reach curvature-radiation reaction limit to become roughly monoenergetic. The curvature spectrum in lower energies then becomes a power law with a spectral index $1/3$, which is much harder than the observed $\\gamma$-ray pulsar spectra. In this subsection, we demonstrate that the $\\gamma$-ray spectrum below a certain energy (say $10$ MeV) is dominated by a synchrotron radiation from freshly born particles and that the expected $\\gamma$-ray spectra further softens. As an example exhibiting a soft power-law $\\gamma$-ray spectrum from eV to GeV energies, we consider the Crab pulsar. To discriminate whether curvature or synchrotron process dominates, we separately consider each process and take the ratio of the radiation-reaction forces. That is, we ignore much complicated synchro-curvature process, because such details are not important for the present purpose. Let us first consider the case of $\\inc=45^\\circ$, which gives $B_5=6.5 \\times 10^2$ at the gap center. Since the curvature-radiated $\\gamma$-ray energy is $E_{\\rm c}=5.2$ GeV, the freshly born particles have the Lorentz factors of $\\Gamma_0 \\sim 5 \\times 10^3$. A particle with this Lorentz factor emit synchrotron radiation around the energy \\begin{equation} h \\nu_{\\rm sync}= \\frac{3 h \\Gamma_0{}^2 eB\\sin\\chi_{\\rm p}} {4\\pi m_{\\rm e} c}, \\label{eq:synchro_Ec} \\end{equation} where $\\chi_{\\rm p}$ denotes the pitch angle of the particles. We can solve the evolution of the Lorentz factor and $\\chi_{\\rm p}$ simultaneously by the method described in \\S~5.3 of Paper~I. We present the evolution of $\\sin\\chi_{\\rm p}$ due to synchrotron radiation shortly after the pair production in figure~\\ref{fig:pitch}, and the evolution of the longitudinal momenta in figure~\\ref{fig:synch}. In both figures, the abscissa designates the distance along the fieldlines in $\\rlc$ unit with respect to the birth place (distance$=0$). The particles are supposed to be created with positive momenta; therefore, electrons turn back to have negative longitudinal momenta. Positrons lose longitudinal momenta on the initial stage of acceleration, because the relativistic beaming effect causes the synchrotron-radiation-reaction force not only in the transverse but also in the longitudinal directions. It follows from figure~\\ref{fig:pitch} that we can approximate $\\sin\\chi_0 \\sim 0.3$ when the particles have not run $\\Delta l \\sim 1 \\times 10^{-6}\\rlc$ for $\\inc=45^\\circ$. When $\\chi_{\\rm p}$ is kept around $0.3$, Lorentz factors are also in the same order of $\\Gamma_0$. Substituting $\\sin\\chi_{\\rm p}=0.3$ and $\\Gamma_0 \\sim 5 \\times 10^3$ into equation (\\ref{eq:synchro_Ec}), we obtain $h \\nu_{\\rm sync} \\sim 8.5$ MeV as the central energy of the synchrotron spectrum for $\\inc=45^\\circ$. The fraction of the particles having $\\sin\\chi_{\\rm p} \\sim 0.3$ and Lorentz factor $\\sim \\Gamma_0$ to the saturated particles then becomes \\begin{equation} \\frac{\\Delta N_{\\rm e}}{N_{\\rm e}} \\sim \\frac{\\Delta l}{H} \\sim 6 \\times 10^{-5}. \\label{eq:ratio_population} \\end{equation} We can estimate the ratio between the synchrotron radiation from the freshly born particles and the curvature radiation from the saturated particles as follows: \\begin{equation} R_{\\rm sc}(\\nu)= \\frac{\\Delta N_{\\rm e}}{N_{\\rm e}} \\frac{dP_{\\rm sync}/d\\nu}{dP_{\\rm curv}/d\\nu}, \\label{ratio_forces} \\end{equation} where \\begin{equation} \\frac{dP_{\\rm sync}}{d\\nu} \\equiv \\frac{\\sqrt{3}e^3 B \\sin\\chi_0}{m_{\\rm e}c^2} F \\left(\\frac{h\\nu}{h\\nu_{\\rm sync}}\\right), \\label{eq:def_Psync} \\end{equation} \\begin{equation} \\frac{dP_{\\rm curv}}{d\\nu} \\equiv \\frac{\\sqrt{3}e^2 \\Gamma}{R_{\\rm c}} F \\left(\\frac{h\\nu}{E_{\\rm c}}\\right), \\label{eq:def_Pcurv} \\end{equation} \\begin{equation} F(x) \\equiv x \\int^\\infty_x K_{5/3}(y)dy; \\label{eq:def_F} \\end{equation} $K_{5/3}$ refers to the modified Bessel function of 5/3 order; $\\Gamma$ in equation~(\\ref{eq:def_Pcurv}) denotes the saturated Lorentz factor and becomes $2.4 \\times 10^7$ for $45^\\circ$ for Crab. At the synchrotron peak energy, $0.29 h\\nu_{\\rm sync}=2.5$ MeV, the ratio becomes $R_{\\rm sc}=12$. In the same manner we can consider the case of $\\inc=30^\\circ$. In this case, we have $h\\nu_{\\rm sync} \\sim 1.7$ MeV and $\\Delta N_{\\rm e}/N_{\\rm e} \\sim 1 \\times 10^{-4}$. As a result, we obtain $R_{\\rm sc}=11$ at $0.5$ MeV. We can therefore conlude that the $\\gamma$-ray spectrum below certain energy ($\\sim 10$ MeV) is dominated by the synchrotron radiation from freshly born particles. In addition, in the case of Crab, the unsaturated motion of particles ($l_{\\rm acc} \\sim H$) implies that the synchro-curvature radiation from unsaturated particles are also important. Therefore, the spectrum below GeV will become much softer compared with the simple curvature spectrum with central energy $E_{\\rm c}=5.2$ (or $4.2$) GeV for $\\inc=45^\\circ$ (or $30^\\circ$). \\begin{figure} \\centerline{ \\epsfxsize=9cm \\epsfbox[70 50 400 460]{pitch.eps} } \\caption{\\label{fig:pitch} Pitch angle evolution due to synchrotron radiation as a function of the position. The abscissa denotes the distance (in $\\rlc$ unit) along the field lines from the birth place (distance$=0$). The thick and thin lines correspond to $\\inc=30^\\circ$ and $\\inc=45^\\circ$, respectively. } \\end{figure} \\begin{figure} \\centerline{ \\epsfxsize=9cm \\epsfbox[70 50 400 460]{synch.eps} } \\caption{\\label{fig:synch} Longitudinal momentum evolution due to synchrotron radiation as a function of the position. The ordinate are nomalized in $m_{\\rm e}c^2$ unit. The abscissa and the lines are the same as figure~7. } \\end{figure} \\subsection{Comparison with Zhang and Cheng model} \\label{sec:cf_ZC} Finally, we point out the difference between the present work and Zhang and Cheng (1997); they considered a gap closure condition so that the curvature-radiated $\\gamma$-ray energy may be adjusted just above the threshold of pair production. That is, they considered the $\\gamma$-ray energy to be about $E_{\\gamma, {\\rm ZC}} \\equiv (m_{\\rm e}c^2)^2 / E_{\\rm x}$, where $E_{\\rm x}$ refers to the characteristic X-ray energy. By equating $E_{\\gamma, {\\rm ZC}}$ with the central energy of curvature radiation (eq. [\\ref{eq:Ec}] in our notation), they closed the equations. When the soft (or hard) blackbody emission dominates, $E_{\\rm x}$ can be approximated by $3 kT_{\\rm s}$ (or $3 kT_{\\rm h}$). The model of Zhang and Cheng (1997) is, in fact, qualitatively consistent with our gap closure condition, provided that the X-ray are supplied by the soft or hard blackbody emission. More specifically, our model gives about 2 times larger characteristic $\\gamma$-ray energy compared with their model. To see this, we present in figure~\\ref{fig:cfZC} the ratio between $E_{\\rm c}$ computed from equation (\\ref{eq:Ec}) and $E_{\\gamma, {\\rm ZC}}$; the hard blackbody or the power-law components are not considered in this calculation. The abscissa indicates the soft blackbody temperature, $kT_{\\rm s}$. For the three thick curves, $\\Omega_2$ is fixed at $0.5$; the solid, dashed, and dotted lines corresponds to $\\mu_{30}= 1.0$, $3.0$, and $0.3$, respectively. For the two thin curves, on the other hand, $\\mu_{30}$ is fixed at $1.0$; the dashed and dotted curves corresponds to $\\Omega_2= 1.0$ and $0.25$, respectively. At small $kT_{\\rm s}$, our model gives more than twice greater $\\gamma$-ray energy compared with Zhang and Cheng (1997); nevertheless, the difference is not very prominent. It should be noted, however, that the spectra of the X-ray radiation are explicitly considered in our present model in the sense we perform the integration over X-ray energies in equations (\\ref{eq:def_lambda_4}), (\\ref{eq:def_lambda_5}), and (\\ref{eq:def_lambda_3}) and that the additional, power-law component is considered in our present model. It would be interesting to investigate the back reaction of the accelerated particles on the polar cap heating, which was deeply investigated by Zhang and Cheng (1997). Consider the case when the voltage drop in the gap approaches the surface EMF, $V_*$. Such an active gap will supply copious relativistic primary particles to heat up the polar cap due to bombardment. The resultant hard blackbody emission supplies target photons for pair production to make the realistic solution deviate from the thick solid curve and approach thin curves at small $kT_{\\rm s}$ (fig.~\\ref{fig:width}). Therefore, this sort of back reaction on the X-ray field due to the relativistic particles needs further consideration. \\begin{figure} \\centerline{ \\epsfxsize=8.5cm \\epsfbox[200 2 700 350]{cfZC.ps} } \\caption{\\label{fig:cfZC} The ratio between $E_{\\rm c}$ computed from equation (13) and $E_{\\gamma, {\\rm ZC}}$ (see text). Neither the hard blackbody nor the power-law components are considered. X-rays are supplied by the whole surface blackbody emission; the abscissa refers to the temperature, $kT_{\\rm s}$. } \\end{figure}" }, "0005/astro-ph0005492_arXiv.txt": { "abstract": "We consider the accretion model for anomalous X-ray pulsars proposed recently by Chatterjee, Hernquist and Narayan, in which the emission is powered by accretion from a fossil disk formed by the fallback of material from a supernova explosion. We demonstrate that this model is able to account for the spin period, luminosity and age distributions of the observed population of AXPs for reasonable and broad distributions of the free parameters of the model, namely, the surface magnetic field of the neutron star, the mass of its accretion disk and its initial spin period. In particular, this model is able statistically to account for the puzzlingly narrow observed spin distribution of the AXPs. We show also that if the establishment of fallback accretion disks around isolated neutron stars is a universal phenomenon, then a fairly large minority ($\\sim 20\\%$) of these objects become X-ray bright AXPs or X-ray faint systems spinning down by propeller action, while the rest become radio pulsars. ", "introduction": "Anomalous X-ray pulsars (AXPs), about half a dozen of which are known, have properties significantly different from those of binary X-ray pulsars (see Mereghetti 1999 for a recent review). AXPs are sources of pulsed X-ray emission with relatively low persistent X-ray luminosities, $L_x \\sim 10^{35} - 10^{36}$ erg/sec, and soft spectra which are well-fitted by a combination of blackbody and power-law contributions, with effective temperatures and photon indices in the range $T_e \\sim 0.3 - 0.4$ keV and $\\Gamma \\sim 3 - 4$, respectively. They have relatively long spin periods of about $P\\sim 6 - 12$ seconds, which increase steadily with time. Their characteristic ages are about $P/2\\dot{P} \\sim 10^{3} - 10^{5}$ years. No binary companions have been detected for these objects, and observations have placed strong constraints on companion masses (e.g. Mereghetti, Israel \\& Stella 1998; Wilson et al. 1998). At least three AXPs have been associated with young supernova remnants which limit their ages to $\\sim 10 - 20$ kyr. There are at present two theories seeking to explain the properties of AXPs. In one, they are modeled as isolated, ultramagnetized neutron stars or ``magnetars'' spinning down by the emission of magnetic dipole radiation, which is powered either by residual thermal energy (Heyl \\& Hernquist 1997a,b) or by magnetic field decay (Thompson \\& Duncan 1996). The required surface magnetic field strength is $B \\sim 10^{14} - 10^{15}$ G, which is similar to the values inferred from timing data for soft gamma repeaters or SGRs (e.g. Kouveliotou et al. 1998, 1999; see, however, Marsden, Rothschild \\& Lingenfelter 1999). The other class of theories proposes that the X-ray emission from AXPs is powered by accretion from binary companions of very low mass (Mereghetti \\& Stella 1995), from the interstellar medium (see Wang 1997 in reference to the AXP candidate RX J0720.4-3125; however, see Heyl \\& Hernquist 1998 for a magnetar model of this object), or from the debris of a disrupted high-mass X-ray binary system, after a stage of common-envelope evolution (van Paradijs et al. 1995; Ghosh, Angelini \\& White 1997). The inferred magnetic field strengths of the neutron stars then are similar to those of ordinary radio pulsars and luminous X-ray pulsars ($B \\simlt 10^{12}$ G). More recently, Chatterjee, Hernquist and Narayan (1999; hereafter CHN) proposed another accretion model in which AXPs are neutron stars with standard magnetic fields accreting from a disk formed after fallback of material from a supernova explosion. The star is rapidly spun down close to the observed AXP periods on a timescale of $\\sim 10^{4}$ years by the action of a propeller effect (for other accretion scenarios for AXPs and SGRs based on the propeller effect, see Alpar [1999] and Marsden et al. [1999]). However, it has been observed that the spin-down of certain AXPs is very stable over time (Kaspi, Chakrabarty \\& Steinberger 1999). This could pose a problem for all accretion models, in which a higher level of timing noise might be expected. Moreover, it is not clear whether optical and infra-red emission from such accretion disks would be low enough to evade the limits set by observations (Perna, Hernquist \\& Narayan, 1999; Hulleman et al. 2000). It remains to be seen whether accretion models can circumvent these difficulties. A question that has hitherto largely been unaddressed is the reason for the puzzlingly strong clustering of spin periods of the observed AXPs. Colpi, Geppert \\& Page (2000) contend that this phenomenon implies magnetic field decay in the magnetar model. Marsden et al. (1999) seek to explain this by limiting the magnetic fields of the neutron stars and the strength of the propeller wind emission by which they spin down to narrow ranges of values. In this paper, we show that the observed narrow period range of AXPs is consistent with the accretion model described in CHN if the neutron stars are drawn from an underlying population characterized by broad and reasonable distributions of magnetic field strength, initial spin and accretion disk mass. It has been suggested that AXPs and SGRs are drawn from the same underlying population of objects. This is motivated by the observation that SGRs (see Hurley 1999 for a review) have properties that are extremely similar to those of AXPs, except for the fact that they occasionally undergo energetic outbursts. We show in this paper that the narrow period range of AXPs and SGRs taken together is also consistent with the above model. ", "conclusions": "We have shown in this paper that according to our model for fallback accretion disks around solitary neutron stars, not all isolated neutron stars which are the product of supernova explosions become radio pulsars; a substantial fraction of them ($\\sim 5-10\\%$, for the models considered here) end up as X-ray bright AXPs, a possibly larger fraction ($\\sim 8-14\\%$) as X-ray faint propeller systems, while the remainder end up as radio pulsars. If the 6 observed AXPs were produced by supernova explosions $\\sim 10^4$ years ago, then the ratio of birthrates of AXPs and radio pulsars is expected to be of the order of 10\\%. Whether a neutron star becomes a radio pulsar, an AXP or a propeller system depends on the values of its surface magnetic field $B$, initial spin period $P_0$, and initial circumstellar disk mass $M_d$. Using the observationally constrained distribution function of $B$ for radio pulsars, and assuming simple distribution functions of initial spin period and disk mass covering a broad range in $P_0$ and Log$(M_d)$ respectively, it has been shown that our accretion model produces a population of AXPs with a spin period distribution peaking at $\\sim 6$ seconds, X-ray luminosity $\\sim 10^{35}$ ergs/s, age $\\sim 10^4$ years, and with approximately the above AXP-radio pulsar birthrate ratio. Systems with relatively high values of $B$ and $M_d$ end up as AXPs, while those with low values end up as radio pulsars; systems with intermediate values of these parameters become propeller systems. It is difficult to make definitive statements about possible relations between the properties of the underlying neutron star, such as its magnetic field strength or its initial accretion disk mass, and its observable properties such as spin period, age or luminosity, since the wide ranges of values the initial parameters can take give rise to a lot of scatter in the final AXP properties that can be observed. Nevertheless, some rough general trends may be discerned. For example, if a neutron star has a low disk mass (``low'' and ``high'' here are to be understood relatively in the context of the corresponding distributions considered in the previous section and in Figure 1), then it would need a high surface magnetic field strength in order to become a visible AXP; this would in turn tend to produce a probable observed spin period which lies slightly rightwards of the peak of the spin period distribution in Figure 1; the probable observed age is low, and the observed luminosity could take a range of values. On the other hand, a star with a high disk mass could reach the AXP phase with a range of surface magnetic field strengths, and the probable observed spin period, luminosity and age could also span a range of values. In this paper, we wished also to address ourselves to the narrow spin period range observed in AXPs. One might suspect that this would in turn constrain the parameters of any model of AXPs to unphysically slender ranges of values. However, we have shown in this paper that even under the assumption of quite reasonable and broad distributions of the parameters, our model leads to a spin period distribution that is statistically consistent with the narrowly clustered distribution observed in the case of the 6 known AXPs, and also in the case of the 6 AXPs and 4 SGRs taken together (if indeed they can be grouped together as fundamentally the same kind of objects). We note in conclusion that the above model would not work as a viable explanation for AXPs if the underlying neutron stars had magnetar-like field strengths ($B \\sim 10^{14.5}$ G); in this case, spin-down would be extremely rapid and would produce spin periods of hundreds of seconds, rather than close to 10 seconds. Thus, it is unnecessary to invoke neutron stars with non-standard magnetic fields in the context of this model for AXPs. \\bigskip \\bigskip We are particularly grateful to Ramesh Narayan for help and advice during the conception of this paper. We would also like to thank Josh Grindlay and Vicki Kaspi for useful discussion." }, "0005/astro-ph0005171_arXiv.txt": { "abstract": "The morphological types of galaxies in nine clusters in the redshift range $0.1\\ls z\\ls 0.25$ are derived from very good seeing images taken at the NOT and the La~Silla--Danish telescopes, with all galaxies at $M_V < -20$ and within the central $\\sim$1 $\\rm Mpc^2$ area being classified. With the purpose of investigating the evolution of the fraction of different morphological types with redshift, we compare our results with the morphological content of nine distant clusters studied by the MORPHS group (Dressler \\et 1997), five clusters observed with {\\it HST-WFPC2} at redshift $z = 0.2-0.3$ (Couch \\et 1998), and Dressler's (1980) large sample of nearby clusters. After having checked the reliability of our morphological classification both in an absolute sense and relative to the MORPHS scheme (Smail \\et 1997), we analyze the relative occurrence of elliptical, S0 and spiral galaxies as a function of the cluster properties and redshift. We find a large intrinsic scatter in the S0/E ratio, mostly related to the cluster morphology. In particular, in our cluster sample, clusters with a high concentration of ellipticals display a low S0/E ratio and, vice-versa, low concentration clusters have a high S0/E. At the same time, the trend of the morphological fractions (\\%Es, \\%S0s, \\%Sp) and of the S0/E and S0/Sp ratios with redshift clearly points to a morphological evolution: as the redshift decreases, the S0 population tends to grow at the expense of the spiral population, whereas the frequency of Es remains almost constant. We also analyze the morphology-density (MD) relation in our clusters and find that -- similarly to higher redshift clusters -- a good MD relation exists in the high-concentration clusters, while it is absent in the less concentrated clusters. Finally, the comparison of the MD relation in our clusters with that of the D97 sample suggests that the transformation of spirals into S0 galaxies becomes more efficient with decreasing local density. ", "introduction": "When Butcher \\& Oemler (1978, 1984) discovered an excess of galaxies bluer than the elliptical sequence in clusters at $z\\ge0.2$, nothing was known about the galactic morphologies at such large distances. The first evidence of the disky/spiral nature of the Butcher-Oemler galaxies came from high-resolution ground-based imagery which also found several cases of galaxies with disturbed morphologies and/or close neighbors (Thompson 1986, 1988, Lavery \\& Henry 1988, 1994, Lavery, Pierce \\& McClure 1992). Over the past five years, thanks to the high spatial resolution imaging achieved with the {\\it Hubble Space Telescope} {(\\it HST)}, it has been established that the morphological properties of galaxies in rich clusters at intermediate redshift differ dramatically from those in nearby clusters. The most obvious difference is the overabundance of spirals in the cluster cores at $z=0.3-0.5$ (Couch et al.\\ 1994, 1998, Dressler et al.\\ 1994, Wirth et al.\\ 1994, Dressler et al.\\ 1997 [D97, MORPHS collaboration], Oemler et al.\\ 1997, Smail et al.\\ 1997 [S97, MORPHS]). The spiral population in the distant clusters consist of the great majority of the blue galaxies responsible for the Butcher-Oemler effect, as well as a sizeable fraction of the \\sl red \\rm population (Dressler et al.\\ 1999 [MORPHS], Poggianti et al.\\ 1999 [MORPHS]). A considerable proportion of these spirals have disturbed morphologies, in some cases quite clearly as the result of an ongoing merger/interaction while in others possibly connected to some other dynamical mechanism (e.g. interaction with the hot intracluster medium or the cluster tidal field, Moore et al.\\ 1996, 1998, Abadi et al. 1999). The second major piece of evidence for morphological evolution in clusters was uncovered only from post-refurbishment data: Coupled to the increase in the spiral fraction, the S0 galaxies at intermediate redshifts are proportionately (x2--3) \\sl less \\rm abundant than in nearby clusters, while the fraction of ellipticals is already as large or larger (D97, S97). This result strongly suggests that a large number of the cluster spirals observed at $z \\sim 0.4$ have evolved into the S0's that dominate the cores of rich clusters today (D97, Couch et al.\\ 1998, van Dokkum et al.\\ 1998, but see Andreon 1998 for a different view). Thus the disk galaxy populations appear to be greatly affected by the cluster environment, while the ellipticals in dense regions seem to have changed little since $z\\sim 0.5$ as far as both their abundance and their stellar populations are concerned (van Dokkum \\& Franx 1996, Andreon, Davoust \\& Heim 1997, D97, Ellis et al.\\ 1997 (MORPHS), Kelson et al. 1997, S97, Barger et al.\\ 1998 (MORPHS), van Dokkum et al.\\ 1998, Kelson et al.\\ 1999). Morphological studies at redshift greater than 0.6 have been limited to three clusters so far, pointing to a low fraction of early-type galaxies in two clusters at $z\\sim 0.8$ (Lubin et al. 1998, van Dokkum et al. 2000), a high early-type galaxy fraction in a cluster at z=0.9 (Lubin et al. 1998) and a surprisingly high rate of mergers in a cluster at z=0.83 (van Dokkum et al. 1999). Further proof of the changes occurring in clusters is the observed evolution of the morphology-density (MD) relation -- the correlation between galaxy morphology and local projected density of galaxies that Dressler (1980a, D80a) found {\\it in all types of clusters} at low redshift, whereby the elliptical fraction increases and the spiral fraction decreases with increasing local galaxy density. An MD relation qualitatively similar to that found by D80a was discovered by D97 to be present {\\it in regular clusters and absent in irregular ones at $z\\sim 0.5$}. Interestingly, the incidence of ellipticals is already very high in all the distant clusters regardless of their dynamical status, therefore the formation of the ellipticals must occur independently of and before cluster virialization (D97). Overall, the available data seem to require a strong morphological evolution in clusters between $z=0.4$ and $z=0$. Still, it is worth keeping in mind that these conclusions, although grounded on high-quality data obtained with a monumental observational effort, are based on a ``small'' sample of distant clusters and on the comparison of a limited redshift range around $z\\sim 0.4$ with the present-day cluster populations (Dressler 1980b, D80b). Clearly the $z\\sim 0.1-0.2$ regime - which up until now has remained largely unexplored - is crucial for a better understanding of the progression of galaxy evolution in dense environments. At these moderate redshifts, performing an analysis comparable to that of the MORPHS requires either ground-based CCD imaging taken over quite a large field under excellent seeing conditions, or a time-consuming mosaic coverage with {\\it HST}. The goal of this paper is to begin to fill in the observational gap between the distant clusters observed with {\\it HST} and the nearby clusters, and hence trace, for the first time, the evolution of the morphological mix at a look-back time of $2-4$ Gyr. In addition, by enlarging the sample of clusters whose galactic morphologies have been studied in detail, we hope to shed some light on the dependence of the observed evolutionary trends on the cluster properties. We present ground-based, good-seeing images of the central regions of 9 clusters at $z=0.09-0.25$ (\\S2) and we perform a detailed morphological analysis of the galaxies in these clusters (\\S3). We study the relative occurrence of ellipticals, S0's and spirals as a function of the cluster properties and we compare them with similar studies at lower and higher redshift (\\S4). Finally, we examine the morphology-density relation of the total sample and of the high- and low-concentration clusters separately (\\S5) and we present our conclusions in \\S6. Throughout this paper we use $H_0=50 \\rm \\, km \\, sec^{-1} \\, Mpc^{-1}$ and $q_0=0.5$. ", "conclusions": "1) The morphological properties of the galaxy populations in nine clusters at $z=0.1-0.25$ are found to be intermediate between those at $z\\sim 0.4-0.5$ and those at low-z, with a moderate spiral content and a moderate ``deficiency'' (as compared to lower redshifts) of S0 galaxies. Our results support the evolutionary scenario, inferred from higher redshift studies (D97, S97, Couch et al. 1998), involving the disk galaxy populations in which there is a progressive morphological conversion in clusters, from spirals into S0's. 2) At $z\\sim 0.2$, we find a dichotomy in the relative occurrence of S0 and elliptical galaxies: four of our clusters display a low S0/E ratio ($\\sim 0.8$) while two of our clusters have a significantly higher ratio ($\\sim 1.9$) similar to the other clusters in our sample at $z=0.1$. The most likely interpretation of this dichotomy and of the large scatter in the S0/E ratio at $z\\sim 0.2$ is that such a ratio is both a function of the redshift and of the cluster ``type'', being significantly lower in clusters with a strong central concentration of elliptical galaxies. 3) At $z\\sim 0.1-0.2$ a morphology-density relation exists only for the high-concentration clusters and is absent in the low-concentration ones. The same result was found at $z \\sim 0.5$ (D97), while at low redshift the correlation between galaxy morphology and local density is present in all types of clusters of the D80a sample. Although the number of galaxies is too small to draw definite conclusions, these results seem to suggest that the morphology-density relation in low-concentration clusters was established only in the last 1-2 Gyr, but only additional data and a homogeneous systematic analysis both at low and moderate redshifts will clarify this matter. Moreover, comparing our MD relations with the corresponding ones at $z \\sim 0.5$ (D97), we suggest that the efficiency of the Sp$\\rightarrow$S0 transformation process anticorrelates with the local density. The relation between the S0/E ratio and the spatial concentration of the ellipticals is not surprising in the light of the well-known correlations between the galaxy content and the cluster type in low-redshift clusters. As mentioned in \\S4, Oemler (1974) grouped clusters in three classes: spiral-rich (SR), elliptical-rich (ER, the most spherical in shape and concentrated) and S0-rich (S0R, named ``spiral-poor'' by O74, ``not quite as centrally concentrated as the ER class, but more regular than the SR class''). Interestingly, O74 suggested that S0-rich clusters are dynamically evolved clusters (they already show segregation by mass and morphological type) representing a later evolutionary stage of spiral-rich clusters, following the evolution of a significant fraction of the spiral galaxies into S0's. In contrast, in the scenario proposed by O74, E-rich clusters are well evolved but \\sl intrinsically different \\rm from the S0R clusters: although possibly the dynamically oldest type of clusters, their high elliptical content implies that they did not evolve from the spiral-rich clusters and is likely due to an enhanced formation rate of ellipticals in regions that began as the densest fluctuations in the early universe. We speculate that in our sample at $z \\sim 0.2$, the four clusters with a strong central concentration of ellipticals (and the lowest S0/E ratios) are presumably the analogues (and progenitors) of the low-z E-rich clusters (see the extrapolation at low redshift of dotted line in Figure~10a), while the low-concentration clusters (with the highest S0/E ratio) seem to be the analogues of the ``S0-rich'' clusters. The effects of redshift (evolution) and cluster type are expected to mingle in various proportions at the different epochs. Following O74, we suggest that the relative occurrence of S0's and spirals is mostly linked with the ``maturity'' of the cluster with spirals progressively evolving into S0's, while the ellipticals are well in place at redshifts greater than those considered here and their abundance and concentration reflect an ``original imprinting'' (see also D97). Then, the S0/Sp ratio should be related to the evolutionary epoch of the cluster and the S0/E value should be determined both by the epoch and the cluster type (nurture and nature, in a way), with the redshift being the dominant effect at early epochs. \\footnote{If the ER and S0R clusters are the endpoints of the evolution of two originally-different types of clusters and both types experienced the accretion of large numbers of spirals that with time turned into S0's, then when looking further back in time, the difference in the S0/E ratio between the precursors of the ER and of the S0R clusters should become smaller and smaller: at $z\\sim 0.4-0.5$ all types of clusters are expected to display a low S0/E ratio (with smaller fluctuations among the different types of clusters), as indeed is observed in the MORPHS dataset, simply because many of the S0's have not yet formed.}" }, "0005/astro-ph0005347_arXiv.txt": { "abstract": "We have applied a powerful numerical approach to compute, with a high resolution, the structure of dusty circumstellar disks with embedded planets. We emphasize some testable implications of our simulations which would verify the presence of a planet via thermal emission of one or more dusty clumps which are in mean motion resonances with the planet. In particular, our simulations indicate that Vega may have a massive planet of $m\\sim 2~m_J$ ($m_J$ being Jupiter's mass) at a distance of 50--60~AU, and Epsilon Eri may have a less massive planet of $m\\sim 0.2~m_J$ at a similar distance of 55--65 AU. This conclusion is testable: Each resonant feature is stationary in the reference frame co-rotating with the planet, but it is not so for the observer at Earth. Therefore, if our interpretation of asymmetric clumps in circumstellar disks as dynamical resonant structures is correct, the above pattern revolves around the star with an angular velocity of $(1.2-1.6)^\\circ$/yr (Vega) and $(0.6-0.8)^\\circ$/yr ($\\epsilon$~Eri) -- a prediction that can be tested on a timescale of several years. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005037_arXiv.txt": { "abstract": "{ Evolutionary effects with redshift of core collapse supernovae and their application to cosmology have been studied based on an extensive grid of stellar models between 13 and 25 $M_\\odot$, and their light curves after the explosion. With decreasing metallicity Z and increasing mass, progenitors tend to explode as compact Blue Supergiants (BSG) and produce subluminous supernovae which are about $1.5^m$ dimmer compared to 'normal' SNeII with Red Supergiant (RSG) progenitors. Progenitors with small masses tend to explode as RSGs even with low Z. The consequences are obvious for probing the chemical evolution. We identify {\\it extreme SNe~IIp} supernovae as a rather homogeneous class which may allow their use as standard candles for distance determination accurate within $30 \\%$. Due to their unique light curves, no spectra need to be taken for their identification, and follow-up observations can be limited to a very small dynamical range in brightness. This means that distance determination by 'extreme SNe~IIp\" are possible up to redshifts of $\\approx 3 $ using 8-meter class telescopes. SIRTIF may push the limit by another magnitude. } \\resumen{ Hemos estudiado los efectos evolutivos en las supernovas gravitacionales debidos al {\\it redshift} y su aplicaci\\'on en cosmolog\\'\\i a bas\\'andonos en una extensa red de modelos con masas comprendidas entre 13 y 25 M$_\\odot$ y sus curvas de luz. Al disminuir la metalicidad, Z, e incrementar la masa, las explosiones tienden a producirse cuando el progenitor es una supergigante azul (BSG) y consecuentemente se obtienen supernovas subluminosas, $\\approx 1.5^m$ m\\'as d\\'ebiles que una supernova {\\it normal} producida en la explosi\\'on de una supergigante roja (RSG). Los progenitores de baja masa tienden a explotar como supergigantes rojas incluso cuando Z es peque\\~na. Analizamos las consecuencias que de ello se derivan en la evoluci\\'on qu\\'\\i mica. Obtenemos que las {\\it SNe IIP extremas} constituyen una clase muy homog\\'enea, por lo que podr\\'\\i an emplearse como indicadores de distancia con una fiabilidad de un 30$\\%$. Debido a que presentan una curva de luz muy caracter\\'\\i stica, pueden identificarse sin recurrir al espectro, y su seguimiento observacional se limitar\\'\\i a un reducido intervalo de brillo. Esto hace posible el empleo de las {\\it SNe IIP extremas} como indicadores de distancia hasta un {\\it redshift} $\\approx$ 3 usando telescopios de clase 8 metros. SIRTIF podr\\'\\i a aumentar el l\\'\\i mite observacional en 1 magnitud. } \\begin{document} ", "introduction": "Supernovae are among the brightest single objects which may reach the same brightness as the entire host galaxy. This allows to measure distances on cosmological scales if the intrinsic brightness of the object is known. During the last few years, the main attention has been drawn by Type Ia Supernovae because the homogenity in their properties allows accurate distance determinations based on either statistical correlations in combination with exact calibrations by $\\delta-Ceph.$ stars (Phillips 1993, Saha et al. 1997), or detailed models. Both the empirical and the theoretical approach provided consistent values for $H_o$ (e.g. M\\\"uller \\& H\\\"oflich 1994, Ries et al. 1995, Hamuy et al. 1996, H\\\"oflich \\& Khokhlov 1996, Nugent et al. 1996). The routine detections of SNe~Ia at redshifts of 0.5 to 1.2 provided strong evidence for a positive cosmological constant (e.g. Perlmutter et al. 1999, Riess et al. 1999). For the latter results, an absolute accuracy of about 10 \\% is required. This leaves potential systematic effects of SNe~Ia properties with redshift as major concern (H\\\"oflich et al. 1998), and it explains the goal to extend the current efforts to large redshifts. In the current scenario, SNe~Ia are thermonuclear explosions of white dwarfs which have grown to the Chandrasekhar mass $M_{Ch}$ by accretion of H or He rich material from a companion, and by burning of the accreted material to C/O. The strength of He shell flashes and the wind around accreting WD is very metal dependent. For Z $\\lesssim 0.1 \\times $solar, the WD may not reach $M_{Ch}$ (Nomoto et al. 2000, fully consistent with our findings). SNe~Ia may not (or rarely) occur at large redshifts. From the observational point of view, another restriction for the use of SNe~Ia is that both spectra and light curves have to be taken to identify the objects. The other class of SN, core collapse supernovae, are thought to be the final results of massive stellar evolution for stars with main sequence masses $\\gtrsim 10 M_\\odot $. The light curves and spectra depend sensitively on the initial stellar mass, metallicity, mass loss and explosion energy. Therefore, these objects show a wide variety of brightness and properties of the light curves which prevents their use as quasi standard candles. On the other hand, these objects will occur soon after the initial star formation period and, therefore, can be used to probe the structure of the universe at high z. In this work, we present a study focussed on core collapse supernovae to answer the following questions: How do the light curves of core collapse supernovae depend on the metallicity which must be expected to decrease with redshift? Can we identify a subclass among the core collapse supernovae which can be used as quasi-standard candles, and what accuracy do we expect? Can this subclass be identified purely by their light curves, without a follow-up which requires to \"go\" much fainter than maximum light? \\begin{figure} \\begin{center} \\leavevmode \\psfig{figure=p1.eps,width=12cm,rwidth=13.3cm,clip=,angle=270} \\vskip -0.3cm \\caption{Stellar evolution from the main sequence to the onset of the core collapse for masses between 13 and 25 $M_\\odot$ with metallicities Z=0.02 (black), 0.066 (light grey) and 0.001(grey).} \\label{fig:cartoon} \\end{center} \\vskip -0.4cm \\end{figure} ", "conclusions": "Light curves for plateau supernovae have been studied. A set of detailed calculations for stellar evolution and light curves have been computed for a variety of initial masses, explosion energies, mixing during the stellar evolution or during the explosion, and metal abundances. Based on our models, we suggest the use of a subclass of Type II Supernovae, the {\\it extreme SNeIIp}, as quasi standard candles. These objects are characterized by a plateau phase in excess of 50 to 60 days (e.g. SN1999em). They can be understood as explosions of Red Supergiants which have undergone rather moderate mass loss during the stellar evolution. The V brightness during the plateau phase changes/declines by about 0.2 to 0.7 $^m$. The mean absolute brightness in V ($\\approx 17.4 - 17.8^m$) during the plateau phase is rather insensitive to the mass of the progenitor and the explosion energy (within $\\approx 0.6 ^m$). Note that line blocking in B and, in particular, in the UV depends on the metallicity causing a somewhat larger spread. The overall similarity of the LCs is caused by the similarity of the density structures of red giants, the resulting {\\sl flat} density slopes, the expanding H-rich envelope and the 'self-regulating' propagation of the recombination front which determines the brightness during the plateau phase. In contrast, if the progenitor explodes as a blue supergiant, the resulting steep density profile results in a long lasting phase of increasing photospheric radius and brightness. The maximum brightness is lower by about $1.5^m$ compared to the explosion of a RSG because of the increased expansion work for BSGs. It is well known from SN1987A that low metallicity stars may explode as blue supergiant. Qualitatively, this tendency is reproduced by our models. Note, however, that we showed above that, at the lower end of the mass scale, the star may explode as a RSG even for Z as low as a 1.E-3. The mass dependence of the final outcome has two main consequences. Firstly, the discovery probability for SNe~II at high z will decrease with the progenitor mass. The supernovae statistics will be systematically biased, starting at $z \\approx 1$. The consequences for the study of the chemical evolution and the element production at high red-shifts (e.g. by NGST) shall be noted. Secondly, even at high redshifts, some {\\it extreme SNe~IIp} should be visible. Taking their unique properties, they may prove to be the key for the use of SN for cosmology at high z before SNe~Ia occur. Although the use of {\\it extreme SNe~IIp} will not achieve the same accuracy as Type Ia Supernovae, there are some distinct advantages: 1) due to their unique light curves and colors, no spectrum is required for identification. 2) The requirements on the time coverage of the light curves are very moderate: Three or four deep images with a sample rate of 50 to 60 days in the rest frame will allow their discovery, identification and their use for cosmology. At some time, two color images should be taken to deselect flare star and to get at handle on the reddening. 3) Finally, there is no need to follow the light curves after the plateau towards dimmer magnitudes. For the use of SNe~Ia, the requirement to obtain a spectrum limits the current use of SNIa of $\\approx 24^m$ if 8m-class telescopes are employed. For the {\\it extreme SN~IIp}, 1) to 3) implies that the largest ground based telescopes with IR detectors can be used as search instruments which pushes the limit to about $27 $ to $28^m$. Therefore, {\\it extreme SNe~IIp} may be used up to $z \\approx 3$ using 8-meter class telescopes. SIRTIF may push the limit by another magnitude by long time exposures. One potential pitfall is the unisotropic luminosity caused by aspherical explosions of core collapse SN. In general, the light of core collapse supernovae is polarized by $\\approx 0.5... 1 \\% $ (e.g. Wang et al. 2000). Polarization of this size corresponds to asymmetries in the envelope which produce directional dependence in the observed L of $\\approx 0.3 ~...~0.6^m$ (H\\\"oflich, 1991). However, extended H rich envelopes tend to spherize the H-rich layers of the envelopes even if the explosions are assumed jet-like (Khokhlov, H\\\"oflich \\& Wang, 2000 in preparation). This tendency is consistent with recent observations for SN1999em (Wang 2000, private communication). The statistical data base for {\\it extreme SNe~IIp} is very incomplete. For the years 1998 and 1999, about 5 to 10 \\% of all nearby SNII fall into this category making the rate for this type about a factor of 3 to 5 less abundant than SNe~Ia. However, the star formation rate at redshifts between 2 and 3 was higher by a factor of $\\approx 3 ... 5 $ (Kravtsov \\& Yepes 2000) compared to the current rate, making the expected rates comparable to those of SNe~Ia. For more details, see Chieffi et al. 2000." }, "0005/astro-ph0005201_arXiv.txt": { "abstract": "A remarkable correlation between the \\Ha\\ emission line and the radio behaviour of \\lsi\\ over its \\4 modulation is discovered. The radio outburst peak is shifted by a quarter of the \\4\\ modulat ion period (about 400 days) with respect to the equivalent width of the \\Ha\\ emission line variability. The onset of the \\lsi\\ radio outbursts varies in phase with the changes of the \\Ha\\ emission line, at least during the increase of \\Ha\\ equivalent width. This is the first clear correlation between the emission associated to the compact object and the Be circumstellar disk in a Be/X-ray binary system. ", "introduction": "The Be/X-ray binaries are the major subclass of massive X-ray binary systems in which a neutron star accretes material from the wind of an early type Be star. The Be stars are known to exhibit emission in the Balmer lines and infrared excess, which are attributed to the presence of cool circumstellar disk. Correlation between the changes of the Be circumstellar envelope and the emission of the compact object can be expected, as a result of the compact object interaction with the surrounding matter. However, no clear correlation has been detected till now -- only loose correlations between the optical/infrared properties of the Be circumstellar disks and the X-ray emission of the neutron star have been reported to exist (e.g. Corbet et al. 1985; Coe et al. 1994; Negueruela et al. 1998). \\object{\\lsi} (\\object{V615~Cas}, \\object{GT~0236+610}) is a radio emitting X-ray binary which exhibits radio outbursts every $26.5\\:$d. The radio outburst peak and the outburst phase are known to vary over a time scale of \\4 (Gregory et al. 1989; Gregory, 1999). Hereafter, we will use the latest values reported and we will refer to these radio periods as $P_{1}=26.4917$~d and $P_{2}=1584$~d. Phase zero for both has been set at JD2443366.775 (Gregory, Peracaula \\& Taylor, 1999). The $26.5$~d period is believed to be the orbital period. The \\4\\ modulation has been discovered on the basis of continued radio monitoring. Both relativistic jet precession or cyclic variability in the Be star envelope have been proposed as a possible origin of the long term modulation (Paredes, 1987; Gregory et al. 1989), with the second interpretation being the most likely one. This suggestion is supported by the fact that the \\Ha\\ emission line varies on the same (4 yr) time scale (Zamanov et al. 1999). However, these authors were not able to derive what is the connection between the radio and \\Ha\\ parameters. In this letter we report an intriguing correlation between the synchrotron non-thermal radio emission, associated with the compact star, and the Be circumstellar disk visible in the $H\\alpha$ emission line. This is the first clear connection between the Be circumstellar disk variability and the emission from the neutron star in the Be/X-ray binaries. ", "conclusions": "\\begin{table} \\caption[]{ Fitted parameters, $\\;\\; y=A+B\\:cos(2\\pi(\\phi +\\phi_0))$.} \\begin{tabular}{c@{\\hspace{1ex}}l@{\\hspace{1ex}}cccc} \\hline parameter & A & B & $\\phi_{0} $ \\\\ \\hline & & & & \\\\ Radio peak [Jy] & 0.187$\\pm$0.003 & 0.059$\\pm$0.003 & 0.125$\\pm$0.010 & \\\\ Onset phase$^*$ & 0.548$\\pm$0.005 & 0.12$\\pm$0.01 & 0.62$\\pm$0.01 & \\\\ EW($H\\alpha$) [\\AA] & 12.8$\\pm$0.01 & 2.3$\\pm$0.1 & 0.60$\\pm$0.01 & \\\\ $\\Delta V_{\\rm peak}$ [km s$^{-1}$] & 323$\\pm$1 & $-20\\pm$2 & 0.59$\\pm$0.01 & \\\\ & & & & \\\\ \\hline \\end{tabular} $\\;(^*)$ the phase of the beginning of the radio outburst calculated relative to the orbital period \\vskip 0.4cm \\end{table} A successful modeling of the radio outburst of \\lsi\\ is based on the synchrotron radiation from relativistic particles injected into an expanding plasmon (Paredes et al. 1991). The genesis of the plasmon can be a result of the transition of the neutron star from propeller to ejector state (Zamanov, 1995), or in other words from accretion onto the magnetosphere to \"young radio pulsar\" every orbital period. In such a picture, the start of the outburst will correspond to the moment when the neutron star emerges from the denser parts of the circumstellar disk. Therefore, the bigger the disk the later the outburst can be expected. This can be seen on Fig.\\ref{P1584}. At phases 0--0.25 we observe increase of the EW(\\Ha), decrease of the $\\Delta\\:V_{\\rm peak}$ and slow shift of the onset of the outburst. The behaviour of the beginning of the outburst on the increasing branch (phases 0--0.25) is stable, but on the decreasing branch (phases 0.25--0.5) the scatter of the points is considerably bigger. The stable behaviour at phases 0.25--0.5 is observed twice at about JD~2449800 and JD~2451400 (Fig.\\ref{JDDD}) so it is unlikely to be a data artifact. Probably this is a result that, during the disk build-up, the increase of the material of the Be disk is feeded only from one source - the B star equatorial region. In contrast the disk-decline can be in two directions - accretion onto the B star or slow dissipation outwards. The behaviour of the start of the outburst points that the disk build up is a stable process and the disk-decline is a more complicated and probably unstable process, or may be it suggests formation of structures like the double disk observed in \\object{X Persei} (Tarasov \\& Roche, 1995). In context of the propeller-ejector transition the surrounding matter will basically influence the expansion velocity of the plasmon, affecting in this way the intensity of the radio outbursts. The remaining plasmon physical parameters (initial magnetic field, injection rate of relativistic particles, etc.) are not expected to vary significantly from one to another outburst. The expanding plasmon calculations predict that there will be weaker outbursts for higher expansion velocities. By expanding faster, the energy losses of the electrons due to the adiabatic expansion are more important and less electron energy is available to be radiated. In addition, the faster decrease of the magnetic field will also contribute to less synchrotron radiation being produced. Supposing that the plasmon is a result of the propeller-ejector transition, the expanding plasmon will appear when the neutron star is receding from the periastron and the plasmon will expand outside of the \\Ha\\ emitting disk. The size of this disk is about 40-65$\\,R_{\\odot}$ (Zamanov \\& Mart\\'{\\i} 2000) and the apastron separation between components is about 150$\\,R_{\\odot}$ (Hutchings \\& Crampton 1981). The enigmatic behaviour of the outburst peak flux density (its phase shift with 0.25 or $\\sim$400 days) indicates that the conditions outside the \\Ha\\ disk vary in a different way compared to the changes inside the \\Ha\\ emitting disk. The X-ray emission of \\lsi\\ is observed to exhibit maximum every orbital period and the X-ray outburst is shifted relatively to the radio outburst (Taylor et al. 1996, Harrison et al. 2000). In terms of the ejector-propeller model, the X-ray maximum is due to the propeller action and higher mass accretion rate onto the magnetosphere at the periastron passage (Zamanov \\& Zamanova, 1997). In this sense it will be very interesting to see what is the behaviour of the X-ray maximum observed in the high-energy emission of \\lsi\\ over the \\4 modulation. Another possible origin of the \\4\\ modulation may be the precession of the Be star. Lipunov \\& Nazin (1994) have demonstrated that this value is in rough agreement with the expected period ($\\sim 10^3$d) for precession of the B star. The precession of the B star can be expected, because after the supernova kick the neutron star orbital plane may be different from the Be disk plane (e.g. Bradt \\& Podsiadlowski, 1995). Our attempts to model the behaviour of the outburst phase as a result of the precession are unsuccessful till now but it can be due to of insufficient data sample, because the systematic radio observations cover about 6 yr (1.5 periods) with considerable gaps and scatter inside the data set. In this context, it deserves to be noted that if the \\Ha\\ variability is a result of a precessing disk seen at different inclination angles, this will imply an inclination angle $i>60^\\circ$ and a precession angle $\\Delta i > 6^\\circ$ (this estimate is obtained using the values from Table~1 and assuming everywhere an optically thick in \\Ha\\ disk). To conclude, the behaviour of \\lsi\\ radio and \\Ha\\ emission is evidence that the picture of the interaction between the neutron star and the circumstellar disk in the Be/X-ray binaries is not as simple as generally expected. We need long series of observations over different wavelengths to better understand the behaviour of the Be stars and the Be/X-ray binaries." }, "0005/nucl-th0005029_arXiv.txt": { "abstract": "We study the neutrino emissivity of strongly magnetized neutron stars due to the charged and neutral current couplings of neutrinos to baryons in strong magnetic fields. The leading order neutral current process is the one-body neutrino-pair bremsstrahlung, which does not have an analogue in the zero field limit. The leading order charged current reaction is the known generalization of the direct Urca processes to strong magnetic fields. While for superstrong magnetic fields in excess of $10^{18}$ G the direct Urca process dominates the one-body bremsstrahlung, we find that for fields on the order $10^{16}-10^{17}$ G and temperatures a few times $10^9$ K the one-body bremsstrahlung is the dominant process. Numerical computation of the resulting emissivity, based on a simple parametrization of the equation of state of the $npe$-matter in a strong magnetic field, shows that the emissivity of this reaction is of the same order of magnitude as that of the modified Urca process in the zero field limit. ", "introduction": "It is now well established that the neutron stars which are observed as radio-pulsars posses $B$-fields of the order of $10^{12}$-$10^{13}$ G at the surface. The interior fields are unknown, but can be by several orders of magnitude larger than the ones inferred for the surface. The scalar virial theorem sets an upper limit on the magnetic field strength of a neutron star of the order of $10^{18} \\rm{G}$ (Lai \\& Shapiro \\cite{Lai}). Similar conclusion is reached through more sophisticated numerical studies (Bocquet et al. \\cite{Boc}). Recent measurements of the spin-down timescales of several soft gamma-ray repeaters, such as SGR 0526-66 (Mazets et al. \\cite{Maz}), SGR 1806-20 (Murakami et al. \\cite{Mur}), and SGR 1900+14 (Kouveliotou et al. \\cite{Kou1}) with RXTE, ASCA and BeppoSAX have made a strong case for SGRs as being newly born neutron stars that have very large surface magnetic fields (up to $10^{15}$ G). The subsequent discoveries of the SGR 1627-41 by BATSE (Woods et al. \\cite{Woods}) and SGR 1801-23 by Ulysses, BATSE, and KONUS-Wind (Cline et al. \\cite{Cline}) lent further support to the identification of SGRs with highly magnetized neutron stars. These objects were naturally related to the magnetars, which are thought to be remnants of a supernova explosion which develop high magnetic fields via a dynamo mechanism (Duncan \\& Thompson \\cite{Dun}, Thompson \\& Duncan \\cite{Thom}). The magnetars also serve as a model for the anomalous X-ray pulsars (AXP) (van Paradijs, Taam \\& van den Heuvel \\cite{Para}) such as 1E 1841-045 (Kes 73) (Gotthelf, Vasisht, \\& Dotani \\cite{Gott}), RX J0720.4-3125 (Haberl et al. \\cite{Hab}), and 1E 2259+586 (Rho \\& Petre \\cite{Rho}). Neutrino-nucleon interactions in the strong magnetic fields have been studied recently both in the supernova and neutron star contexts. It has been pointed out that the neutrino emission from proto-neutron stars, which is anisotropic in strong $B$-fields, could produce the ``pulsar kicks\" if the fields are in excess of $10^{16}$ G (Horowitz \\& Li \\cite{Horo}; Arras \\& Lai \\cite{Arras} and references therein). The strong magnetic fields relax the kinematical constrains on the direct Urca process and hence give rise to finite neutrino emissivity even when the proton fraction is small (Leinson \\& Perez \\cite{Leinson}, Bandyopadhyay et al. \\cite{Band}). The effect is most pronounced in the ultra-high magnetic fields when protons and the electrons occupy the lowest Landau levels. The direct Urca process for arbitrary magnetic fields, when the protons and electrons are allowed to occupy many Landau levels, has also been studied (Baiko \\& Yakovlev \\cite{Baiko}). The neutrino emissivities via the one-body processes sensitively depend on the abundances of the various species of baryons and leptons which are controlled by the equation of state (EoS) of the dense matter in strong magnetic fields. The strong magnetic fields lead to an increase of the proton fraction (Broderick, Prakash \\& Lattimer \\cite{Bro}). The muon production and pion condensation in strong magnetic softens the EoS of the dense matter (Suh \\& Mathews \\cite{Suh}). For the purpose of estimating the magnitude of the neutrino emissivities, we employ in this paper a simple parametrization of the EoS for the $npe$-matter in strong magnetic fields. The main objective of this paper is to show that the strong magnetic fields open a new channel of neutrino emission via one-body neutral current bremsstrahlung, which does not have an analogue in the zero field limit. We also briefly discuss the direct Urca process, which is forbidden in the low-density zero-field limit (as long as the triangular condition $p_{Fp}+p_{Fe}\\ge p_{Fn}$ is not satisfied), but is allowed in strong magnetic fields because of the relaxation of the kinematical constrains. We compare the emissivities of various reactions using a simple parametrization of the EoS for the $npe$ matter in a strong $B$-field. As a standard reference for our comparison we use the modified Urca process. The presence of a magnetic field has two different effects on neutrino emissivities: (i) in pure neutron matter it allows for spin-flip transitions where the finite difference of the momenta of neutrons at two different Fermi surfaces enables one to satisfy energy-momentum conservation, (ii) the charged particles occupy Landau levels leading to a smearing of the transverse momenta over an amount $\\sqrt{eB}$ \\footnote{We use the natural units, $\\hbar=c=k_B=1$.}. As a consequence there are two typical scales of the magnetic field, where effects on the emissivity are expected; first for $|\\mu_B| B \\sim T$ which is relevant in neutrino-pair bremsstrahlung from neutrons, and second $e B \\sim p_F^2$ relevant for the Urca process. As well know, the emissivity of any particular reaction can be related to the imaginary part of the polarization function of the medium (Voskresensky \\& Senatorov \\cite{VS}, Raffelt \\& Seckel \\cite{Raffelt}, Sedrakian \\& Dieperink \\cite{Se1}). We compute the polarization function of neutrons and protons in strong magnetic fields employing the finite temperature Matsubara Green's functions technique. For the case of the bremsstrahlung the time-like properties of the polarization function are relevant. The space-like properties of the polarization function, relevant for the neutrino-nucleon scattering, have been studied by Arras \\& Lai (\\cite{Arras}) in an equivalent response function formalism. The plan of this paper is as follows. The bremsstrahlung emissivity is related to the polarization function of the medium in Sect. 2. The neutrino emissivity via neutrino-pair bremsstrahlung from neutrons is discussed in Sect. 3 and that from protons in Sect. 4. The Urca process in strong magnetic fields is briefly discussed Sect. 5. The EoS of $npe$ matter in a magnetic field is discussed in Sect. 6. Our numerical results are presented in Sect. 7. Sect. 8 contains our conclusions. ", "conclusions": "We have studied the neutrino emissivity of strongly magnetized neutron stars due to the one-body processes driven by the charged and neutral current couplings between the neutrinos and baryons. We have shown that, in addition to the well-known charged current process (the direct Urca reaction), there is a new channel of energy loss - the one-body neutrino pair-bremsstrahlung in a magnetic field. The process does not have an analogue in the zero field limit and competes with the modified (two-body) bremsstrahlung process as the dominant neutral current reaction for fields on the order $10^{16}-10^{17}$ G and temperatures a few times $10^9$ K. For superstrong magnetic fields in excess of $10^{18}$ G the direct Urca process takes over. \\\\ Our numerical evaluation of the emissivities of several competing reactions, which is based on a simple parametrization of the EOS of $npe$-matter in a strong magnetic field, shows that under certain conditions the emissivities of the one-nucleon processes, such as the direct Urca and the one-body bremsstrahlung, are of the same order of magnitude or dominate the standard processes commonly included in the cooling simulations in the zero-field limit. \\appendix" }, "0005/hep-ph0005073_arXiv.txt": { "abstract": "We investigate a topological inflation model in supergravity. By means of numerical simulations, it is confirmed that topological inflation can take place in supergravity. We also show that the condition for successful inflation depends not only on the vacuum-expectation value (VEV) of inflaton field but also on the form of its K\\\"ahler potential. In fact, it is found that the required VEV of the inflaton $\\varphi$ can be as small as $\\langle \\varphi \\rangle \\simeq 1 \\times M_G$, where $M_{G}$ is the gravitational scale. ", "introduction": "Superstring theories compactified on $(3+1)$-dimensional space-time have many discrete symmetries in the low-energy effective Lagrangian \\cite{discrete}. A spontaneous breakdown of such discrete symmetries creates topological defects, i.e. domain walls, in the early universe \\cite{KIB}. If the vacuum-expectation value (VEV) of a scalar field $\\varphi$ is larger than the gravitational scale $M_{G} \\simeq 2 \\times 10^{18}$~GeV, the region inside the wall undergoes inflationary expansion and eventually becomes the present whole universe \\cite{Linde,Vilenkin}. If the universe is open at the beginning, it expands and the spontaneous breakdown of the symmetries always takes place at some epoch in the early universe. It has been recently argued that the quantum creation of the open universe may take place with appropriate continuation from the Euclidean instanton \\cite{open}. Thus, topological inflation is a natural consequence of the dynamics of the system, and it does not require any fine-tuning of initial conditions for the beginning universe. Furthermore, it does not cause the ``graceful exit\" problem and the universe becomes homogeneously radiation dominated after reheating. A simple and interesting model for topological inflation was proposed in the framework of supergravity \\cite{Izawa}.\\footnote{Other topological inflation models were studied in the superstring inspired models \\cite{BBN,EKOY}.} However, it was not explicitly shown whether topological inflation really takes place. In this paper, we perform a numerical analysis on the above model and show that topological inflation indeed occurs in a wide range of parameter space. We also show that the condition for successful inflation depends not only on the superpotential, which determines the vacuum expectation value (VEV) of inflaton $\\varphi$, but also on the form of its K\\\"ahler potential. We in fact find that the required VEV can be as small as $\\langle \\varphi \\rangle \\simeq 1 \\times M_G$, which is far below the lower bound of $\\la \\varphi \\ra = \\eta_{cr} \\simeq 1.7M_{G}$ derived in Ref. \\cite{sakai}. ", "conclusions": "We have studied a topological inflation in supergravity. First, we have shown that topological inflation really takes place in supergravity. Also, the criterion of successful topological inflation depends not only on the breaking scale of the discrete symmetry but also on the mass of the inflaton near the origin. This is because the inflaton rolls down rapidly from the origin if its mass is large. For a very flat case favored by the observation of the spectral index, $n_{s} \\simeq 1 - 0.8$ (i.e., $0 < \\kappa < 0.1$), we have found that the critical breaking scale $\\eta_{cr}$ becomes as small as $M_{G}$, which is smaller than the critical value, $\\eta_{cr}\\cong1.7M_{G}$ observed in Ref.~\\cite{sakai}. Finally we have discussed the primordial spectrum produced by the topological inflation. In general, the topological inflation predicts the tilted spectrum $n_{s} < 1$ depending on $\\kappa$.\\footnote{It is possible to produce more exotic spectrum including blue one. This is due to the exponential blow of the potential, which is significant for the region $\\varphi \\gtilde M_{G}$ and makes the potential more complex than the simple double-well potential.} The present topological inflation model is free from the thermal and nonthermal overproduction of gravitinos since the reheating temperature can be as low as $10^{8}$~GeV. Furthermore, as pointed out in Ref.~\\cite{Asaka}, this model is consistent with a leptogenesis scenario in which heavy Majorana neutrinos are produced in the inflaton decay and successive decays of the Majorana neutrinos result in lepton asymmetry enough to explain the observed baryon asymmetry in the present universe. \\subsection*" }, "0005/astro-ph0005366_arXiv.txt": { "abstract": "% With exact three-flavor Boltzmann neutrino transport, we simulate the stellar core collapse, bounce, and postbounce evolution of a 13 \\msolar\\, star in spherical symmetry, the Newtonian limit, without invoking convection. In the absence of convection, prior spherically symmetric models, which implemented approximations to Boltzmann transport, failed to produce explosions. We are motivated to consider exact transport to determine if these failures were due to the transport approximations made and to answer remaining fundamental questions in supernova theory. The model presented here is the first in a sequence of models beginning with different progenitors. In this model, a supernova explosion is not obtained. We discuss the ramifications of our results for the supernova mechanism. ", "introduction": "Core collapse supernovae are among the most important phenomena in astrophysics because of their energetics and nucleosynthesis. Beginning with the first numerical simulations conducted by Colgate and White\\cite{cw66}, three decades of supernova modeling have established a basic supernova paradigm. The supernova shock wave---formed when the iron core of a massive star collapses gravitationally and rebounds as the core matter exceeds nuclear densities---stalls in the iron core as a result of enervating losses to nuclear dissociation and neutrinos. The failure of this ``prompt'' supernova mechanism sets the stage for a ``delayed'' mechanism, whereby the shock is reenergized by the intense neutrino flux emerging from the neutrinospheres carrying off the binding energy of the proto-neutron star\\cite{w85,bw85}. The heating is mediated primarily by the absorption of electron neutrinos and antineutrinos on the dissociation-liberated nucleons behind the shock. This past decade has also seen the emergence of multidimensional supernova models, which have investigated the role convection, rotation, and magnetic fields may play in the explosion \\cite{hbhfc94,bhf95,jm96,mcbbgsu98a,mcbbgsu98b,fh99,khowc99}. Although a plausible framework is now in place, fundamental questions about the explosion mechanism remain: Is the neutrino heating sufficient, or are multidimensional effects such as convection and rotation necessary? Can the basic supernova observable, explosion, be reproduced by detailed spherically symmetric models, or are multidimensional models required? Without a doubt, core collapse supernovae are not spherically symmetric. For example, neutron star kicks\\cite{fbb98} and the polarization of supernova emitted light\\cite{w99} cannot arise in spherical symmetry. Nonetheless, ascertaining the explosion mechanism and understanding every explosion observable are two different goals. To achieve both, simulations in one, two, and three dimensions must be coordinated. The neutrino energy deposition behind the shock depends sensitively not only on the neutrino luminosities but also on the neutrino spectra and angular distributions in the postshock region, necessitating exact multigroup (multi-neutrino energy) Boltzmann neutrino transport. Ten percent variations in any of these quantities can make the difference between explosion and failure in supernova models\\cite{jm96,bg93}. Past simulations have implemented increasingly sophisticated approximations to Boltzmann transport, the most sophisticated of which is multigroup flux-limited diffusion\\cite{br93,wm93}. A generic feature of this approximation is that it underestimates the isotropy of the neutrino angular distributions in the heating region and, thus, the heating rate\\cite{ja92,mmbg98}. It is important to note that, without invoking proto-neutron star (e.g., neutron finger) convection, simulations that implement multigroup flux-limited diffusion do not produce explosions\\cite{br93,wm93}. Moreover, the existence and vigor of proto-neutron star convection is currently a matter of debate\\cite{mcbbgsu98a,bd96,kjm96}. Wilson\\cite{w71} implemented an approximation to Boltzmann neutrino transport by using order-of-magnitude parameterizations of the neutrino--matter weak interactions. His models failed to produce explosions. Core collapse simulations that implemented exact Boltzmann neutrino transport were completed by Mezzacappa and Bruenn\\cite{mb93a,mb93c}. Following this work, we now present the findings of a core collapse, bounce, and postbounce simulation. Recognizing the need for more accurate time-dependent neutrino transport in supernova models, other groups have now developed Boltzmann solvers\\cite{yjs99,bypet99,rj00}. \\section {\\bf Foundations} We model the explosion of a 13 M$_{\\odot}$ star, beginning with the precollapse model of Nomoto and Hashimoto\\cite{nh88}. The core collapse, bounce, and explosion were simulated with a new neutrino radiation hydrodynamics code for both Newtonian and general relativistic spherically symmetric flows: AGILE--BOLTZTRAN. BOLTZTRAN is a three-flavor Boltzmann neutrino transport solver\\cite{mb93b,mm99}, now extended to fully general relativistic flows\\cite{l00}. In this simulation, it is employed in the $O(v/c)$ limit with 6-point Gaussian quadrature to discretize the neutrino angular distributions and 12 energy groups spanning the range from 5 to 300 MeV to discretize the neutrino spectra. AGILE is a conservative general relativistic hydrodynamics code\\cite{l00,lt98}. Its adaptivity enables us to resolve and seamlessly follow the shock through the iron core into the outer stellar layers. The equation of state of Lattimer and Swesty\\cite{ls91} (LS EOS) is employed to calculate the local thermodynamic state of the matter in nuclear statistical equilibrium (NSE). For matter initially in the silicon layer, the temperatures are insufficient to achieve NSE. In this region, the radiation and electron components of the LS EOS are used, while an ideal gas of \\nuc{Si}{28} is assumed for the nuclear component. For typical hydrodynamic timesteps ($\\sim .1$ millisecond), silicon burning occurs within a single timestep for T $\\sim 5 \\gk$\\cite{ht99}; therefore, when a fluid element exceeds a temperature of 5 \\gk\\ in our simulation, the silicon is instantaneously burned, achieving NSE and releasing thermal energy equal to the difference in nuclear binding energy between \\nuc{Si}{28} and the composition determined by the LS EOS. We investigated the convergence of the net neutrino heating rate as the number of Gaussian quadrature points and the number of neutrino energy groups in our Boltzmann simulations were varied, as in Messer et al\\cite{mmbg98}. In the heating region, the 4- and 6-point rates, the 6- and 8-point rates, and the 12- and 20-group rates differed by at most 5 percent, 3 percent, and 3 percent, respectively. Moreover, during the course of the important first 300 ms of our simulation, the maximum variation in the total energy is $\\sim 3\\times 10^{49}$ erg, which is a few percent of the total energy, and the total lepton number is conserved to within a fraction of a percent. Note that the numerical uncertainty in the net heating rate (which is at most 3 percent in our model) is no larger than the uncertainty in the total energy. Therefore, any further numerical convergence in the computation of this rate would be meaningless. \\begin{figure} \\begin{center} \\epsfig{file=traces.ps} \\end{center} \\caption{Radial trajectories of equal mass shells in the iron core and silicon layer. We also trace the shock, nuclear burning, and dissociation fronts, which carve out three regions in the $(r,t)$ plane. A: Silicon. B: Iron produced by infall compression and heating. C: Free nucleons and alpha particles. } \\label{fig1} \\end{figure} ", "conclusions": "" }, "0005/astro-ph0005150_arXiv.txt": { "abstract": "We discuss the optical coronal line spectra observed for a sample of 19 Narrow Line Seyfert 1 galaxies. We find no correlation between the coronal line strength and the soft X-ray power-law index derived from {\\it ROSAT PSPC} data. There is a trend for broader coronal lines to have larger equivalent widths. In addition, a strong trend is found between line width and velocity relative to the NLR. This trend is interpreted in terms of a decelerating outflow, originating close to the nucleus. ", "introduction": "As a class Narrow Line Seyfert 1 galaxies (NLS1s) are defined in terms of their optical emission-line properties \\cite{Osterbrock}. In recent years, however, NLS1s have been studied mainly in relation to their X-ray emission, which displays some peculiar properties. In particular, the soft X-ray emission from NLS1s seems to be extreme in its shape and variability, possibly due to an extreme accretion rate \\cite{Pounds}. To study this emission directly is difficult due to Galactic absorption. However, an indirect probe is provided by the high-ionization coronal lines. These forbidden lines arise from species with high ionization potentials ($>100$ eV), and are thought due to photoionization by the hard AGN continuum. As part of a programme to study their multi-wavelength properties, we obtained optical spectra of 19 NLS1s with the {\\it IDS\\/} mounted on the {\\it INT\\/} at La Palma. These objects were not chosen to have strong coronal line emission, but rather to represent the range in observed X-ray spectral indices as observed using the {\\it ROSAT PSPC}. The optical data were reduced using {\\sc STARLINK} software. ", "conclusions": "The lack of correlation between the coronal-line strengths and the {\\it ROSAT} power-law indices is somewhat surprising given the predictions of photoionization models. It may be that the range in gas conditions or covering factors for the coronal-line regions between different NLS1s is quite large. The continuum shape may also be aspect-dependent, such that the steep soft X-ray continuum is not always seen by the coronal-line emitting gas. Finally, a single power law may provide a poor parameterization of the soft X-ray spectral shape. {\\it ASCA} spectra of NLS1s do suggest a complicated spectral shape in some objects \\cite{Vaughan}. The shape of the soft X-ray continuum in NLS1s will be accurately determined by forthcoming {\\it Chandra\\/} and {\\it XMM-Newton\\/} observations. The kinematical results suggest a connection between the coronal-line emitting gas and the central region of NLS1s. In the context of a decelerating outflow model, the gas velocity is $\\approx 500$ km s$^{-1}$ at a distance from the centre $\\approx 1/40$ that of the NLR. Given an NLR size $\\approx 100$~pc, this implies that the coronal-line gas originates at $\\approx 2$~pc from the centre. Such a small size could be associated with the outer region of the BLR and/or the inner edge of the proposed dusty torus." }, "0005/astro-ph0005016_arXiv.txt": { "abstract": "Radio galaxies are uniquely useful as probes of large-scale structure as their uniform identification with giant elliptical galaxies out to high redshift means that the evolution of their bias factor can be predicted. As the initial stage in a project to study large-scale structure with radio galaxies we have performed a small redshift survey, selecting 29 radio galaxies in the range $0.19}{_{\\sim}} 200\\,$mJy. In particular, at faint flux levels the radio source population is a mix of nearby star-forming galaxies and AGN-powered radio sources with a range in redshifts from 0 to $>$4 (Condon et al.\\ 1998), which have quite different clustering properties. It is therefore important to test the results of the angular correlation function studies with direct measurements of clustering from radio galaxy redshift surveys. Studies of the clustering of radio-quiet AGN seem to show a generally similar correlation length, but there is a wide range in estimates of $r_0$ from different samples. This can probably be explained if the correlation function depends both on redshift and AGN luminosity [e.g.\\ Sabbey et al.\\ (2000), La Franca, Andreani \\& Christiani (1998)]. Magliocchetti et al.\\ (1999) discuss theoretical predictions for the evolution of the two-point correlation function of radio sources. Perhaps the most appropriate case to take is that where radio galaxies form at high redshift ($z\\gg 1$). We can trace the evolution of the host population out to $z\\sim 3$, and find that the hosts vary little with redshift, apart from some passive evolution. The host magnitudes are also only weakly dependent on radio luminosity (Lacy, Bunker \\& Ridgway 2000). Hence uncertainties in the evolution in the bias factor are unlikely to be as important an issue for radio galaxies as they are for normal galaxies or radio-quiet quasars. Fry (1996) shows that in this ``galaxy conservation'' scenario, the bias factor increases with redshift according to $b(z)=1+(b_0-1)(1+z)$, where $b_0$ is the bias factor at the present epoch. This is because fluctuations in the galaxy density field are fixed at the epoch of formation, but the fluctuations in the matter density field grow with time. The decrease in the bias factor with time is mostly compensated for by the clustering of matter under gravity, for which the growth factor $D(z)=(1+z)^{-1}$ for an $\\Omega_{M}=1$, $\\Omega_{\\Lambda}=0$ cosmology. The two-point correlation function, $\\propto D^2(z)b^2(z)$, should therefore show little evolution. To test this model, and to examine the nature of intermediate redshift superclusters, we therefore decided to begin a survey of large-scale structure at moderate redshifts ($z\\sim 0.2-0.65$). This paper describes the initial result from this survey and also the prospects for future surveys. ", "conclusions": "We have succeeded in developing an effective method for studying the clustering of moderate redshift radio galaxies directly, and have detected clustering of radio galaxies at $z\\approx 0.3$. The amplitude of the cross-correlation function we measure is consistent with that for radio galaxies locally. This is as expected in the simple model discussed in the introduction, in which radio source hosts evolve little with redshift, and is higher than that for normal galaxies at $z\\sim 0.3$, for which Small et al.\\ (1999) measure $r_0=3.7\\, h^{-1} {\\rm Mpc}$. At present, however, the small size of our survey prevents us ruling out all but the most extreme evolution in the correlation function. Expansion of this survey to $\\stackrel{>}{_{\\sim}} 100$ redshifts will allow a measurement of the two-point correlation function to be made which has comparable accuracy to that for normal galaxies at these redshifts. The relatively large volume probed by a larger survey will allow us to define a sample of superclusters and to examine their structures and the evolution of those structures to the present day. The discovery of redshift clustering in deep pencil beam galaxy surveys, e.g. that of the HDF (Cohen et al.\\ 2000) has raised the possibility that large-scale structures continue to be present in the Universe at least to $z\\sim 1$. Therefore the evolution of these structures should place interesting constraints on cosmology. Crucially, however, because we can trace the evolution of radio galaxy hosts to $z\\sim 3$, we can, in principle, predict how the bias should evolve with redshift, removing an important uncertainty from the interpretation of the results of correlation function studies." }, "0005/astro-ph0005234_arXiv.txt": { "abstract": "Biased galaxy-formation theories predict that massive galaxies at high redshifts should act as signposts to high-density environments in the early universe, which subsequently evolve into the cores of the richest clusters seen at the present day. These regions are characterised by over-densities of young galaxies, perhaps including a population of dusty, interaction-driven starbursts --- the progenitors of massive cluster ellipticals. By searching for this population at submillimeter (submm) wavelengths we can therefore test both galaxy- and structure-formation models. We have undertaken such a search in the field of a $z=3.8$ radio galaxy, 4C\\,41.17, with the SCUBA submm camera. Our extremely deep 450- and 850-$\\mu$m maps reveal an order-of-magnitude over-density of luminous submm galaxies compared to typical fields (the likelihood of finding such an over-density in a random field is $<2 \\times 10^{-3}$). The SCUBA galaxies have bolometric luminosities, $>10^{13}\\,$L$_\\odot$, which imply star-formation rates (SFRs) consistent with those required to form a massive galaxy in only a few $10^8$ years. We also note that this field exhibits an over-density of extremely red objects (EROs), some of which may be associated with the submm sources, and Lyman-break galaxies. We propose that the over-densities of both submm and ERO sources in this field represent young dusty, starburst galaxies forming within a proto-cluster centered on the radio galaxy at $z=3.8$, which is also traced by a less-obscured population of Lyman-break galaxies. ", "introduction": "Galaxy-formation theories are developing rapidly and have claimed some success at reproducing the properties of galaxies in the local universe (Cole et al.\\ 1994). However, at the moment their predictions of the early evolution of galaxies are comparatively untested. Moreover, the areas where the theoretical predictions are most reliable tend to be those where the observational tests are most difficult at high redshift, e.g., the evolution in the mass function of galaxies (White \\& Frenk 1991). One testable prediction of hierarchical galaxy-formation models at high redshifts deals with the clustering behaviour of massive galaxies at early epochs: such galaxies are expected to cluster strongly in regions of highest density, areas which should subsequently form the cores of massive clusters of galaxies in the local universe (Kaiser 1984; Baron \\& White 1987; Efstathiou \\& Rees 1988; Kauffmann et al.\\ 1999). Thus by searching for over-densities of massive galaxies in the distant universe we can test the predictions of hierarchical models {\\it as well as} investigating the formation and evolution of the luminous galaxies seen in present-day clusters. However, to achieve this we must find a method of selecting the densest regions in the early universe. One possible technique is to use massive galaxies as signposts for high-density environments at high redshifts. Observations of distant, luminous radio galaxies have led to the suggestion that they are massive ellipticals (Matthews, Morgan \\& Schmidt 1964; Lilly \\& Longair 1984), a belief confirmed recently by {\\it Hubble Space Telescope} ({\\it HST}) studies which show they possess $r^{1/4}$-law profiles characteristic of elliptical galaxies (McLure et al.\\ 1999; Zirm et al.\\ 2000). Locally these massive elliptical galaxies typical reside in galaxy clusters and hence luminous radio galaxies should make good markers to search for the progenitors of rich galaxy clusters at high redshifts. The picture of hierarchical formation of massive galaxies and their environments is supported on small scales by recent rest-frame optical imaging of radio galaxies (van Breugel et al.\\ 1998; Pentericci et al.\\ 1998): at the earliest epochs ($z\\ge 4$) there is evidence of diffuse emission on large scales (5--10$\\,$$''$) and sub-clumps similar in scale ($\\sim10$\\,kpc) to radio-quiet star-forming galaxies; these then appear to evolve into more compact structures by $z\\sim 2$. Interestingly, on somewhat larger scales there is also an apparent 10--100-fold increase in the surface density of EROs --- some of which are dusty, star-forming galaxies akin to local ULIRGs (Dey et al.\\ 1999; Smail et al.\\ 1999) --- around high-redshift radio galaxies (HzRGs) and quasars as compared with the field (Hu \\& Ridgway 1994; Elston, Rieke \\& Rieke 1988; Arag\\'on-Salamanca et al.\\ 1994; Dey, Spinrad \\& Dickinson 1995; Yamada et al.\\ 1997). On similar scales, the distorted radio morphologies and high rotation measures of some HzRGs suggest they reside in high-density environments (Carilli et al.\\ 1997) with extended X-ray emission detected around one example, 1138$-$262 at $z=2.2$ (Carilli et al.\\ 1998). \\begin{figure*} \\centerline{\\psfig{file=f1.eps,width=7.3in,angle=0}} \\noindent{\\scriptsize {\\sc Fig.~1.}---a) Full 850-$\\mu$m image of the $z=3.8$ radio galaxy, 4C\\,41.17, at the original $\\sim$14$''$ resolution. Negative features are due to the chopped/nodded observing procedure. Contours are plotted at $3, 4, 5, 6, 7 \\times 1.5\\,$mJy\\,beam$^{-1}$. b) 450-$\\mu$m image of 4C\\,41.17, smoothed to a resolution of 10$''$ FWHM. Contours are plotted at $3, 4 \\times 8.6\\,$mJy per 10$''$ beam$^{-1}$. c) Central, cleaned portion of the 850-$\\mu$m image. Contours are plotted at $3, 4 \\ldots 15 \\times 0.8\\,$mJy\\,beam$^{-1}$, where the beam has been smoothed to 20$''$ in the background regions. \\vspace*{-1mm} } \\end{figure*} Confirming that HzRGs reside in high-density environments would also provide substantial insight into galaxy evolution in these regions, especially the formation of giant ellipticals in the hierarchical models already discussed. Optical/IR studies have targeted the evolution of ellipticals in clusters at $z < 1$ (Arag\\'on-Salamanca et al.\\ 1993; Stanford, Eisenhardt \\& Dickinson 1998; Bower, Lucey \\& Ellis 1992; Ellis et al.\\ 1997); theoretical work has investigated their evolution in the context of hierarchical formation in high-density regions (Kauffmann \\& Charlot 1998; Baugh et al.\\ 1999). Pinpointing the progenitors of ellipticals at very high redshifts would allow straightforward tests of the latter models and would enable the time-line of the optical/IR cluster studies to be extended beyond $z\\sim 1.3$ (Stanford et al.\\ 1997, 1998; Rosati et al.\\ 1999; Liu et al.\\ 2000). In the hierarchical picture, lower-mass structures will accrete onto the high-mass peak of the proto-cluster identified with the radio galaxy, the final assembly of this structure achieved through a series of mergers which will strongly affect the galaxies residing within the dark matter halo, triggering intense starbursts analogous to mergers in the local universe (Sanders \\& Mirabel 1996). The dust created in these starbursts will absorb a considerable fraction of the UV/optical light emitted by young stars, re-emitting it in the rest-frame far-IR. An extreme and highly obscured starburst with a SFR of $\\sim 1000\\,$M$_\\odot$yr$^{-1}$ would have a bolometric luminosity, $L_{\\rm bol}$, of $\\sim 10^{13}\\,$L$_\\odot$, and the majority of this would appear at rest-frame far-IR wavelengths (Ivison et al.\\ 1998). The negative $K$-correction for dust emission in the submm passband (Blain \\& Longair 1993) means that a luminous, dusty starburst with $L_{\\rm bol} \\sim 10^{13}\\,$L$_\\odot$ would have an 850-$\\mu$m flux of $\\sim 10\\,$mJy if observed at {\\it any} redshift between 1 and 10. The advent of sensitive arrays working at submm wavelengths, in particular the SCUBA camera (Holland et al.\\ 1999) on the James Clerk Maxwell Telescope (JCMT) have enabled efficient and sensitive surveys of dusty starbursts, probing out to high redshifts (Smail, Ivison \\& Blain 1997; Barger et al.\\ 1998; Hughes et al.\\ 1998; Eales et al.\\ 1999). Thus by using SCUBA to undertake targeted observations of the environments of known HzRGs, searching for over-densities of bright submm sources, we can test whether HzRGs are located in the cores of rich proto-clusters {\\it and} constrain the formation epoch of massive cluster galaxies. For targets at $z\\gs 2$ the resolution and field of view of SCUBA at 850$\\,\\mu$m means we are sensitive to star-forming galaxies distributed on scales from 100 to 1000\\,kpc --- well-matched to the predicted virial radii of the most massive clusters at these epochs (Jenkins et al.\\ 2000). In this paper we present deep 450- and 850-$\\mu$m maps of one of the most distant and powerful known radio galaxies, 4C\\,41.17 at $z=3.8$ (for which 1$''$ = 6.6\\,$h_{50}^{-1}$\\,kpc, with $h_{50} = H_0/50$\\,km\\,s$^{-1}$\\,Mpc$^{-1}$ and $q_0=0.5$). There is already some evidence for a massive structure associated with 4C\\,41.17. Based upon a Lyman-break search using $UVR$ imaging, Lacy \\& Rawlings (1996) have suggested that there is a modest excess of high-redshift galaxies in this region, having found six candidate $z\\gs 3.4$ galaxies in a 1.5-arcmin$^2$ field, roughly centered on 4C\\,41.17. They show that this density is slightly higher than that expected from the field density found by Steidel, Pettini \\& Hamilton (1995), but the complications introduced by comparing their $UVR$ study with Steidel et al.'s $\\mathcal{UGR}$ survey, as well as issues of cosmic variance (Steidel et al.\\ 1999), make it difficult to assess the over-density. Other evidence for a deep potential centered on the radio galaxy includes the presence of an extended Ly$\\alpha$ halo reaching to $>$100\\,kpc (Chambers, Miley \\& van Breugel 1990; Dey 1999). In the next section we present our submm, near-IR and optical observations of this field and their reduction. We discuss our analysis and results, in the context of supporting archival data at other wavelengths, in \\S3 and give our conclusions in \\S4. ", "conclusions": "We report the discovery of five new luminous submm sources within a $2.5'$-diameter (1\\,Mpc at $z=3.8$) area centered on the submm-bright $z=3.8$ radio galaxy, 4C\\,41.17. Three of these sources are as bright or brighter than any other known blank-field submm galaxies. This surface density of submm galaxies is an order of magnitude greater than that expected from blank-field counts: we would need to map at least 600 blank fields to find a chance configuration of such bright sources. Using the available constraints, we suggest that this over-density lies at $z>2.8$ and is therefore consistent with a structure associated with 4C\\,41.17. We note that this field exhibits an over-density above that expected in blank fields, not only at submm wavelengths, but also in samples of galaxies selected via the Lyman-break technique (Lacy \\& Rawlings 1996) and for their extremely red colors (EROs). As many as five of the latter class of galaxies may be directly associated with the luminous submm sources seen in this field. We introduce a classification scheme for the counterparts of submm sources, based upon observations of well-studied SCUBA galaxies. Within the framework of this scheme we propose that submm galaxies and ERO starbursts represent different aspects of a single evolutionary cycle. Following on from this, we suggest that the over-densities of both the submm and ERO populations in this field represent young, dusty starbursts forming within a proto-cluster centered on the radio galaxy at $z=3.8$, which also hosts a population of less-obscured Lyman-break galaxies. More observations are clearly needed to confirm the nature of this over-density, to test our suggestion that it is associated with the radio source at $z=3.8$ and, most interestingly, to investigate if it represents a virialised structure. The most feasible test of the latter issue will be to image this field using the {\\it Chandra} or {\\it Newton} X-ray observatories to identify emission from the hot intracluster medium confined within the potential well of the proposed proto-cluster. We are currently expanding our survey to cover 15 fields centered on HzRGs to determine more reliable limits on the prevalence of similar over-densities of submm sources around massive galaxies at high redshifts, $z=3$--5." }, "0005/astro-ph0005144_arXiv.txt": { "abstract": "AGNs with narrow Balmer lines show various extreme properties. In particular, rapid X-ray variability, steep X-ray spectra, peculiar optical and UV line ratios, and possibly peculiar line profiles. Since all these phenomena occur together they are likely to be related to one specific underlying physical parameter. I review recent evidence, based on HST imaging of low $z$ quasars, which suggests that the H$\\beta$ line width and continuum luminosity of quasars provide a reasonably accurate estimate of the black hole mass. This implies that narrow-line AGN have relatively low black hole masses, and thus high $L/L_{\\rm Edd}$, as independently suggested based on their steep X-ray spectra. I present additional evidence suggesting that the X-ray variability and the radio loudness are primarily driven by the black hole mass. The high mass inflow rate into the core of narrow-line AGNs may produce a denser and more enriched BLR, a high column radiation pressure driven outflow, and a smaller illumination angle for the NLR, as suggested by the observed emission line properties. Narrow-line AGNs may thus provide important clues for understanding the rich overall phenomenology of AGNs. ", "introduction": "The term `Narrow-Line Seyfert 1 Galaxies' (NLS1s) was coined by Osterbrock \\& Pogge (1985) who noted the overall peculiar optical emission-line spectra of Seyfert galaxies with narrow Balmer lines (see Pogge, these proceedings). Follow-up studies of their radio emission and optical polarization properties did not reveal anything outstanding. The first hint for their remarkable X-ray properties was found by Stephens (1989) who noted, based on {\\em Einstein} data, that ``X-ray selection may be an efficient way to find NLS1s.'' This conclusion was much strengthened by Puchnarewicz et al. (1992) who found that $\\sim 50$\\% of their {\\em Einstein} ultrasoft survey AGNs were NLS1s, thus establishing that NLS1s have steeper than usual soft X-ray spectra. This result was further refined by Laor et al. (1994) who noted a remarkably strong correlation between the H$\\beta$ FWHM and the {\\em ROSAT} $\\alpha_x$ in a sample of 10 PG quasars ($r_s$ = 0.842, Pr=$2\\times 10^{-3}$ ). This result was further strengthened when the complete sample of all 23 $M_B<-23$, $z<0.4$, $N_{\\rm H~I}<1.9\\times 10^{20}$~cm$^{-2}$, PG quasars was analyzed (Laor et al. 1997a, hereafter L97; $r_s$= 0.79, Pr= $7\\times 10^{-6}$). Boller, Brandt \\& Fink (1996) studied a large sample of NLS1s with {\\em ROSAT} and noted the clear absence of broad line AGNs with a steep $\\alpha_x$. However, they found a much larger scatter in the H$\\beta$ FWHM vs. $\\alpha_x$ relation. In particular, some of their NLS1s show normal $\\alpha_x$ values, unlike the sample of L97 where all narrow-line quasars display steep $\\alpha_x$ values. This difference most likely results from the large luminosity range of the AGNs in the Boller et al. sample. In particular, a plot of $\\alpha_x$ vs. $L_X$ for the Boller et al. sample (using the data in their Table 1) reveals a clear trend of flattening of $\\alpha_x$ with decreasing $L_X$. All the flat ($\\alpha_x>-2$) NLS1s in the Boller et al. sample have low luminosity ($L_X<1.3\\times 10^{44}$~erg~s$^{-1}$), and all their luminous AGNs ($L_X>1.3\\times 10^{44}$~erg~s$^{-1}$) are steep ($\\alpha_x<-2$). Thus, the Boller et al. sample indicates that a significant H$\\beta$ FWHM vs. $\\alpha_x$ correlation appears in bright AGNs, and not when lower luminosity Seyferts are included, consistent with the strong correlation in the Laor et al. sample which includes only $M_B<-23$ PG quasars. The H$\\beta$ FWHM vs. $\\alpha_x$ correlation thus involves luminosity as well. The luminosity dependence can be understood if the primary driver of this correlation is $L/L_{\\rm Edd}$, rather than just the H$\\beta$ FWHM (see \\S 2). It will be interesting to explore if lower luminosity AGNs do follow an H$\\beta$ FWHM vs. $\\alpha_x$ correlation, but offset towards flatter $\\alpha_x$. ", "conclusions": "The highly simplified $M_{\\rm BH}(\\Delta v, L)$ estimate in quasars appears to be accurate to within a factor of about 2--3. NLS1s thus most likely have a relatively low $M_{\\rm BH}$, and if they are not very faint, a relatively high $L/L_{\\rm Edd}$. The rapid X-ray variability of NLS1s is mostly due to their low $M_{\\rm BH}$, but there may also be some enhancement of variability for NLS1s with the highest $L/L_{\\rm Edd}$. NLS1s, and Seyferts in general, are radio quiet most likely because of their relatively low $M_{\\rm BH}$. The UV lines suggest the BLR is denser and possibly more enriched. A possible scenario is that the high $L/L_{\\rm Edd}$ results from large amounts of gas being dumped into the center of the galaxy. This increased accretion rate brings in denser gas (implied by the UV line ratios), enhances star formation rate and therefore metalicity (implied by the strong N~V), and possibly blocks most of the NLR illumination (implied by the weak narrow lines). The increased $L/L_{\\rm Edd}$ may enhance the column of the radiation-pressure ablated surface layer in the BLR (implied by the UV line profiles). Clearly, there are many open questions which need to be addressed to reliably establish or disprove the above scenario. Specifically: 1. Do NLS1s follow the $M_{\\rm BH}$ vs. $L_{\\rm bulge}$ relation? The best way to proceed here is to extend the analysis of Bahcall et al. and Kirhakos et al. to lower luminosity PG Quasars which qualify as NLS1s. 2. How tight is the X-ray variability vs. $M_{\\rm BH}$ relation? One needs a well defined sample, high quality optical spectroscopy, and long X-ray integrations (if the PSD is non-stationary). 3. Do NLS1s generally show the UV emission-line profile trends seen in I~Zw~1? One needs high-quality UV spectroscopy for a well defined sample of NLS1s, and the PG sample can again be a very useful parent sample. 4. Are AGN absorption outflows related to $L/L_{\\rm Edd}$? The ideal route here is to survey the UV absorption properties of AGNs spanning a large range in $L$ and in $L/L_{\\rm Edd}$ (the PG sample again!). 5. What controls the strength of [O III]? Is it the NLR gas covering factor? Ionization state? Density? A careful study of various line ratios in the NLR is required for a conclusive answer. 6. What controls the other intriguing correlations noted by Boroson et al (host colors, extended [O III], and radio morphology)? Follow-up studies of these correlations with higher quality data is the first step required here. I thank the organizing committee for their kind invitation, generous support, and for a very interesting meeting." }, "0005/astro-ph0005372_arXiv.txt": { "abstract": "Recent observations of the diffuse Galactic gamma-ray glow at 1.809 MeV, attributed to the radioactive decay of \\element[][26]{Al}, point towards a massive star origin of this radioactive isotope. Wolf Rayet stars and core-collapse supernovae appear to dominate the production of this isotope. Massive stars are commonly located in clusters and OB associations, regions of recent star formation. We thus discuss the temporal evolution of \\element[][26]{Al}, and \\element[][60]{Fe} within evolving OB associations. The goal of this study is to utilize the associated gamma-ray lines as a diagnostic tool for the study of correlated star formation, and also to provide more detailed models for the interpretation of data obtained with COMPTEL on the Compton Observatory. We investigate the effects of possible aluminum yield enhancements, predicted for some massive close binary systems. In addition to the ejection of chemically processed matter, massive stars also drive strong stellar winds and emit large fluxes of ionizing radiation. This energy and radiation input into the interstellar medium (ISM) is crucial for the dynamical evolution of the gas and subsequent star forming activity in the galactic disk. We discuss population synthesis models for a variety of star formation histories, and compare the predicted gamma-ray line lightcurves to COMPTEL measurements in the Cygnus region. Radioactive tracers such as \\element[][26]{Al} and \\element[][60]{Fe} provide a unique gamma-ray tracer of Galactic star formation activity, complementary to other methods using spectral information in the radio, IR or optical bands. ", "introduction": "Introduction} For almost a decade the COMPTEL instrument aboard the Compton Observatory (Schoenfelder et al. \\cite{Schoen93}) has mapped the diffuse Galactic emission in the 1.809 MeV line (Diehl et al. \\cite{diehl95}, Oberlack et al. \\cite{ugo96}, and Pl\\\"uschke et al. \\cite{pl99a}). This gamma-ray line is believed to be due to the decay of radioactive \\element[][26]{Al}, which can be produced in many different environments. One of the primary issues in the interpretation of this map is the identification of the dominant source of this isotope. Several studies of correlations between the 1.8 MeV map and diffuse maps taken in low energy bands suggest that the gamma- ray patterns follow closely those of the massive star population (Prantzos \\& Diehl \\cite{p_d96}, Diehl \\& Timmes \\cite{dt98}, Kn\\\"odlseder et al. \\cite{kn99}). In particular, Kn\\\"odlseder et al. (\\cite{kn99}) recently demonstrated that the 1.809 MeV emission profile correlates best with electron bremsstrahlung emission in the microwave regime, which traces the free electrons which in turn are produced by the ionizing radiation from massive stars. The gamma-ray line from \\element[][26]{Al} thus appears to directly trace the presence of massive stars, so that Wolf Rayet stars and core-collapse supernovae emerge as the most promising source candidates (Prantzos \\& Diehl \\cite{p_d96}). With a typical stellar yield of $10^{-4}\\mathrm{M}_{\\odot}$ of \\element[][26]{Al} injected into the ISM per massive star (during the WR phase and in the supernova event) and a million year lifetime, one expects a steady state amount of $\\sim$ 1 M$_\\odot$ of radioactive aluminum in the ISM at any time. Spread throughout the disk of the Galaxy this amount can explain the observed diffuse 1.8 MeV line flux. In addition to the nucleosynthesis of radioactive matter, which traces the evolution over the past few million years, massive stars dynamically shape the interstellar medium on a similar time-scale (e.g., Lozinskaya \\cite{loz92}). Due to their high mass loss rates (\\cite{bar81,dj88}) and large wind velocities (Cassinelli \\& Lamers \\cite{cl87}), massive stars impart a large amount of momentum and kinetic energy to the surrounding medium (e.g., van der Hucht et al. \\cite{vdh87}; Leitherer et al. \\cite{lei92}). Due to their large surface temperatures, massive stars also emit a large fraction of their radiative luminosity in the wavelength regime below 91.2 nm, causing photoionization of the surrounding medium (e.g., Panagia \\cite{pan73}, Vacca et al. \\cite{vac96}). The subsequent supernova explosions contribute additional energy (typically 10$^{51}$ erg (Jones et al. \\cite{jo98}, and references therein)), although much of this energy might be radiatively lost (Thornton et al. \\cite{th98}), as well as fresh products of stellar nucleosynthesis.\\\\ In a population synthesis approach we compute the light-curves of \\element[][26]{Al} (and also \\element[][60]{Fe}) together with the mechanical and extreme ultra-violet luminosities as a function of time for different star formation histories. \\element[][26]{Al} and \\element[][60]{Fe} yields are taken from recent WR models (Meynet et al. \\cite{m97}) and supernova simulations (Woosley \\& Weaver \\cite{ww95}; Woosley et al. \\cite{wlw95}). We also study the potentially important role of yield enhancements for stars that are members of particular binary systems (Langer et al. \\cite{lan98}). For an assumed star formation history the corresponding gamma-ray line flux then provides a unique diagnostic tool to study star forming regions, complementing other tracers such as IR, UV, or H$_\\alpha$ emission. We use the gamma-ray light curves to constrain the star formation history of the the Cygnus region, which is one of the brightest isolated features in COMPTEL's 1.809 MeV map.\\\\ The paper is organized as follows. In section \\ref{MEASURE} we review the COMPTEL 1.8 MeV results. In section \\ref{MODEL} we construct an OB association model based on three major aspects; nucleosynthesis of radioactive matter, injection of kinetic energy, and emission of extreme ultra violet radiation. We contrast results from a starburst with those from a continuous star formation history. A comparison with recent COMPTEL observations of the Cygnus region is presented in section \\ref{COMPARE}. We present our concludions in section \\ref{SUMMARY}. ", "conclusions": "" }, "0005/astro-ph0005414_arXiv.txt": { "abstract": "We present observations of SDSSp J104433.04--012502.2, a luminous quasar at $z=5.80$ discovered from Sloan Digital Sky Survey (SDSS) multicolor imaging data. This object was selected as an $i'$-band dropout object, with $i^*=21.8 \\pm 0.2$, $z^*=19.2 \\pm 0.1$. It has an absolute magnitude $M_{1450} = -27.2$ ($H_{0} =50$ km s$^{-1}$ Mpc$^{-1}$, $q_{0} = 0.5$). The spectrum shows a strong and broad Ly$\\alpha$ emission line, strong Ly$\\alpha$ forest absorption lines with a mean continuum decrement $D_{A} = 0.91$, and a Lyman Limit System at $z=5.72$. The spectrum also shows strong OI and SiIV emission lines similar to those of quasars at $z\\lesssim 5$, suggesting that these metals were produced at redshift beyond six. The lack of a Gunn-Peterson trough in the spectrum indicates that the universe is already highly ionized at $z \\sim 5.8$. Using a high-resolution spectrum in the Ly$\\alpha$ forest region, we place a conservative upper limit of the optical depth due to the Gunn-Peterson effect of $\\tau < 0.5$ in regions of minimum absorption. The Ly$\\alpha$ forest absorption in this object is much stronger than that in quasars at $z\\lesssim 5$. The object is unresolved in a deep image with excellent seeing, implying that it is unlensed. The black hole mass of this quasar is $\\sim 3 \\times 10^9 M_{\\odot}$ if we assume that it is radiating at the Eddington luminosity and no lensing amplification, implying that it resides in a very massive dark matter halo. The discovery of one quasar at $M_{1450} < -27$ in a survey area of 600 deg$^2$ is consistent with an extrapolation of the observed luminosity function at lower redshift. The abundance and evolution of such quasars can provide sensitive tests of models of quasar and galaxy formation. ", "introduction": "At what epoch did the first generation of galaxies and quasars form? How was the universe re-ionized, ending the ``dark ages'' (\\cite{Rees98})? These fundamental questions can only be answered with studies of high-redshift objects. The last few years have witnessed the first direct observations of galaxies at redshift higher than five (\\cite{Dey98}, \\cite{Weymann98}, \\cite{Spinrad98}, \\cite{Chen99}, \\cite{Breugel99}, \\cite{Hu99}, see also the review by \\cite{Stern99}), while detailed studies of the ensemble properties and large scale distribution of galaxies at $z\\sim4$ have begun (Steidel et al.~1998, 1999). Several quasars have been found at $z\\gtrsim 5$ (Fan et al.~1999, 2000a, \\cite{Zheng00}), including a low-luminosity quasar at $z=5.50$ (\\cite{Stern00}). Studies of high-redshift quasars provide important probes of this critical epoch in cosmic evolution. The lack of the Gunn-Peterson (1965) effect in the absorption spectrum of a $z\\sim 5.0$ quasar (\\cite{S99}) indicates that the universe is already highly ionized at that redshift. The exact epoch of re-ionization could be determined from the absorption spectra of quasars at even higher redshift (\\cite{Jordi97}, \\cite{HL99}). The study of the luminosity function of high-redshift quasars will constrain models of quasar and galaxy evolution (\\cite{HL98}, \\cite{Haehnelt98}), and determine whether it was UV radiation from AGNs or from young massive stars that re-ionized the universe, ending the ``dark ages'' (\\cite{HL98}). Measurements of the chemical abundance in the quasar environment will reveal the metal production process at the very early stage of galaxy evolution (\\cite{HF99}). Finally, luminous high-redshift quasars represent high peaks of density fields, and may be the markers of large scale structure at these early epochs (\\cite{George99}, \\cite{HH00}, \\cite{MW00}). The Sloan Digital Sky Survey (SDSS; \\cite{York00}) is using a dedicated 2.5m telescope and a large format CCD camera (\\cite{Gunnetal}) at the Apache Point Observatory in New Mexico to obtain images in five broad bands ($u'$, $g'$, $r'$, $i'$ and $z'$, centered at 3540, 4770, 6230, 7630 and 9130 \\AA, respectively; \\cite{F96}) over 10,000 deg$^2$ of high Galactic latitude sky. The multicolor data from SDSS have proven to be very effective in selecting high-redshift quasars: more than 50 quasars at $z>3.5$ have been discovered to date from about 600 deg$^2$ of imaging data (Fan et al.~1999, 2000a, \\cite{HET1}, \\cite{Zheng00}). The inclusion of the reddest band, $z'$, in principle enables the detection of quasars up to $ z \\sim 6.5$ in SDSS data. In this paper, we report the discovery of SDSSp J104433.04--012502.2 (the name reflecting its J2000 coordinates from the preliminary SDSS astrometry, accurate to $\\sim 0.1''$ in each coordinate), a very luminous, ``$i'$-dropout'' quasar at $z=5.80$, selected by its very red $i^*-z^*$ color. In a $\\Lambda$-dominated flat universe ($H_{0}$ = 65 km s$^{-1}$ Mpc$^{-1}$, $\\Lambda=0.65$ and $\\Omega=0.35$, referred to as the $\\Lambda$-model in this paper, \\cite{OS95}, \\cite{KT95}), $z=5.80$ corresponds to an age of 0.9 Gyr in an universe 13.9 Gyr old, or a look-back time of 93.2\\% of the age of the universe. Similarly, the universe was 0.7 Gyr old at $z=5.80$ in a universe 13.0 Gyr old at present for a model with $\\Omega=1$ and $H_{0}$ = 50 km s$^{-1}$ Mpc$^{-1}$, which we refer to as the $\\Omega=1$ model in this paper. We present the photometric observations and target selection in \\S 2, and the spectroscopic observations in \\S 3. In \\S 4, we discuss the cosmological implications, including the constraints on the Gunn-Peterson effect, quasar evolution models, and black hole formation. ", "conclusions": "\\subsection{Absorption Properties and Gunn-Peterson Effect} The high luminosity of SDSS 1044--0125 makes it an ideal object for high signal-to-noise ratio observations to study the intergalactic medium at high redshift. In order to detect continuum break caused by the Lyman Limit System, an edge filter with width of 40 \\AA\\ was convolved with the low resolution spectrum in Figure 2. A strong peak at 6131 \\AA\\ is detected in the convolved spectrum, indicating the existence of a Lyman Limit System at $z_{LLS} = 5.72$. No flux is detected blueward of this break. A Lyman Limit System is usually detected within 0.1 of the emission redshift in essentially all quasars at $z>4$ (\\cite{SSG91}, \\cite{APM}, \\cite{Fan99}). The most striking feature of the spectrum of SDSS 1044--0125 is the very strong absorption caused by Ly$\\alpha$ forest lines. However, the flux level in the Ly$\\alpha$ forest region never reaches zero. It lacks the Gunn-Peterson (1965) trough that would exist in the spectrum of a quasar at redshift higher than the re-ionization redshift (\\cite{HL99}), indicating that the intergalactic medium is already highly ionized at $z \\sim 5.8$. We estimate the average continuum decrements as: $D_{A,B} \\equiv \\left\\langle 1 - f_\\nu^{obs}/f_\\nu^{con} \\right\\rangle $, where $f_\\nu^{obs}$ and $f_\\nu^{con}$ are the observed and the unabsorbed continuum fluxes of the quasar, and $D_{A}$ and $D_{B}$ measure the decrements in the region between rest-frame Ly$\\alpha$ and Ly$\\beta$ ($\\lambda = 1050 - 1170$ \\AA) and between Ly$\\beta$ and the Lyman Limit ($\\lambda = 920 - 1050$ \\AA), respectively (\\cite{OK82}). The measurements of $D_{A}$ and $D_{B}$ require knowledge of the continuum shape redward of Ly$\\alpha$. However, with $D_{A}$ approaching unity, the effect of different slopes is quite small. Assuming a power law continuum $\\nu^{\\alpha}$ with $\\alpha = -0.5$, as indicated in Figure 3, we obtain $D_{A} = 0.91$ and $D_{B}$ = 0.94. Using a slope of --1.0 only changes the $D_{A}$ and $D_{B}$ values to 0.92 and 0.96, respectively. We therefore adopt $D_{A} = 0.91 \\pm 0.02$ and $D_{B} = 0.95 \\pm 0.02$. These values are in close accordance with those from quasar RD J030117+002025 ($z=5.50$, \\cite{Stern00}) and from distant galaxies in the Hubble Deep Field (\\cite{Weymann98}), and are much higher than that of the $z=5.00$ quasar SDSSp J033829.31+002156.3 ($D_{A} = 0.75$, \\cite{S99}), suggesting that the strong evolution of the strength of the Ly$\\alpha$ forest at $z>5$, $N(z) \\propto (1+z)^{2.3-2.75}$, measured at redshifts below five continues to a redshift of nearly six. Assuming this number density evolution, \\cite{Zuo93} and \\cite{F99} show that at $z\\sim 5.8$, the expected average $D_{A}$ ranges from 0.8 to 0.9. We further derive an upper limit on the Gunn-Peterson optical depth following Songaila et al.~(1999). Figure 3 shows the high-resolution echellette spectrum of SDSS 1044--0125 in the Ly$\\alpha$ forest region (binned to 2 \\AA/pixel). The continuum level is approximated by a $\\nu^{-0.5}$ power law as above. Even the most transparent part of the forest does not return close to the continuum level at this resolution. It is evident that the Ly$\\alpha$ forest is much stronger in SDSS 1044--0125 than in SDSSp J033829.31+002156.3 (Figure 2 of \\cite{S99}), where a fraction of the forest has flux comparable to the extrapolated continuum. In the region between 7926 \\AA\\ and 7929 \\AA, SDSS 1044--0125 has an optical depth $\\tau = 0.35 \\pm 0.07$, where the error bar only reflects the statistical noise in the spectrum (note the noise level indicated in the figure). This value changes to 0.40 and 0.31 for power law slopes of 0.0 and $-1.0$, respectively. Therefore, we adopt a conservative limit of $\\tau < 0.5$ at this redshift of 5.52. With higher resolution and signal-to-noise ratio, we might be able to select regions even less affected by the Ly$\\alpha$ forest lines. Therefore, it is only an upper limit. For comparison, Songaila et al.~derived $\\tau < 0.1$ for $z=4.72$ with similar resolution. The Gunn-Peterson effect analysis above is based on an attempt to measure the amount of flux between Ly$\\alpha$ forest lines (e.g., \\cite{G94}). At redshift higher than five, even with a moderately high resolution spectrum, these forest lines overlap, making it impossible to find a truly ``line-free'' region. Modern hydrodynamic simulations and semi-analytic models show that under the influence of gravity, the intergalactic medium becomes clumpy, and the Gunn-Peterson optical depth should vary even in the lowest column density regions (e.g. \\cite{Bi92}, \\cite{Jordi93}, \\cite{Cen94}, \\cite{Hernquist96}). The minimum absorption regions in the forest merely represent regions that are most underdense in this fluctuating Gunn-Peterson effect. An accurate measurement of the Gunn-Peterson effect and the ionizing background from the high resolution spectrum of SDSS 1044--0125 requires detailed comparison with cosmological simulations; this is beyond the scope of the current paper. Figure 2 also shows the detection of an intervening MgII absorption system. The MgII doublet $\\lambda 2796.4+2803.5$ is detected at wavelengths 9166.7 \\AA\\ and 9190.3 \\AA\\ in the high-resolution spectrum; the redshift of this system is $z_{abs}=2.278$. The rest frame equivalent widths of the doublet lines are 2.43 and 1.90 \\AA, respectively. This system is very similar to the one detected in SDSSp J033829.31+002156.3 ($z_{abs}=2.304$, \\cite{S99}). It is possible that SDSS 1044--0125 is amplified by lensing from this intervening absorber. However, we saw in \\S~2 that this object is {\\em unresolved} under $0.4''$ seeing in the K band. \\subsection{Number Density of Very High Redshift Quasars} The total area of SDSS imaging data that we have searched for high-redshift quasars thus far is of order 600 deg$^2$. All that satisfy $z^*<19.3$ and $i^* - z^* > 2.2$ in this 600 deg$^2$ region have been observed spectroscopically. Only SDSS 1044--0125 is identified as a high-redshift quasar; the remaining objects are L and T dwarfs. Using the luminosity function and redshift dependence of \\cite{SSG95} (for the $\\Omega=1$ model), extrapolating it to higher redshifts and assuming $f_{\\nu} \\propto \\nu^{-0.5}$, we predict that in a total area of 600 deg$^2$, for $z>5.65$ ($i'$-dropout objects ), there should be 1.5 quasars with $M_{1450} < -27.0$ and 1.1 quasars with $M_{1450} < -27.2$. For $z>5.8$, the extrapolation predicts 1.1 and 0.8 quasars for $M_{1450} < -27.0$ and $-27.2$, respectively. This assumes that our selection efficiency is 100\\%. The Schmidt-Schneider-Gunn luminosity function is derived using objects with $2.7 < z < 4.7$ and $-27.5 < M_{B} < -25.5$. Although it is difficult to draw any reliable conclusion from the observation of a single high-redshift quasar, the discovery of SDSS 1044--0025 is consistent with the expectations from this rather large extrapolation from lower redshift results. Assuming that the same luminosity function holds at even higher redshift, the SDSS will be able to discover one quasar at $z\\gtrsim 6$, $z^* \\lesssim 19$ in every 1500 deg$^2$ of the survey. SDSS 1044-0125 is a very luminous quasar. Assuming that (1) its bolometric luminosity equals the Eddington luminosity, $L_{\\rm bol}=L_{\\rm Edd}=1.5\\times 10^{38}(M_{\\rm BH}/{\\rm M_\\odot})~{\\rm erg~s^{-1}}$; (2) its intrinsic continuum spectrum is the same as the mean spectral template of \\cite{Elvis94}, and (3) neither beaming nor lensing affects the observed flux, we find a black hole mass of $M_{\\rm BH}=3.4\\times 10^{9}{\\rm M_\\odot}$ in the $\\Lambda$--model, or $M_{\\rm BH}=2.7\\times 10^{9}{\\rm M_\\odot}$ in the $\\Omega=1$ model. In either case, the implied black hole mass is quite large, similar to that of the black hole at the center of the nearby giant elliptical galaxy M87 (Harms et al.~1994, Macchetto et al.~1997). If the quasar were radiating below the Eddington limit, the inferred black hole mass would be even higher. In the \\cite{Elvis94} template, $\\approx 1\\%$ of the bolometric luminosity is emitted in the observed $z^\\prime$ band. If this fraction is larger for SDSS 1044-0125, the implied black hole mass would be reduced. Note that the universe was less than 1 Gyr old at this redshift, while the Eddington time scale, the $e$-folding time for the growing of a black hole shining at the Eddington luminosity, is $4 \\times 10^7 (\\epsilon/0.1)$ yr, where $\\epsilon$ is the radiative efficiency for the accretion. If the black hole started accreting with an initial mass of $\\sim 10^3 \\rm M_\\odot$ with 10\\% efficiency, the seed black hole would have to form and begin accreting at redshift well beyond 10 in order to grow to $3\\times 10^{9} \\rm M_\\odot$ at $z=5.80$ (see also \\cite{Turner91}). Forming such a massive black hole in such a short time is a remarkable feat. The observations of high-redshift quasars can be used to constrain the formation epoch of the first star clusters and the fueling process of early black holes. How likely is it to find a quasar like SDSS 1044-0125 in popular cold dark matter cosmological models? We use the following simple model to estimate the abundance of high-redshift quasars (see also \\cite{HH00}): \\cite{Magoo98} have found a correlation between central black hole mass and bulge mass, $M_{\\rm BH}/M_{\\rm bulge}=6\\times10^{-3}$ in nearby galaxies. If this correlation holds at high redshift, this would imply a bulge mass of $5.7\\times10^{11}~{\\rm M_\\odot}$ ($\\Lambda$-model) or $4.5\\times10^{11}~{\\rm M_\\odot}$ ($\\Omega=1$) for the host galaxy of SDSS 1044-0125, and a lower limit of $5.7\\times10^{12}~{\\rm M_\\odot}$ ($\\Lambda$-model) or $4.5\\times10^{12}~{\\rm M_\\odot}$ ($\\Omega=1$) for its dark halo, assuming $M_{\\rm halo}/M_{\\rm bulge}\\geq \\Omega_{\\rm DM}/\\Omega_{\\rm b}\\approx 10$. The comoving abundance of dark matter halos at this epoch is very sensitive to this halo mass, and can be estimated by means of the \\cite{PS74} formalism. In a $\\Lambda$-CDM model, assuming $\\sigma_{8} = 0.87$ and an untilted primordial power spectrum, for parent halo masses of $10^{12}$ M$_\\odot$, $6 \\times 10^{12}$ M$_\\odot$, and $10^{13}$ M$_\\odot$, we expect 50,000, 60, and 4 candidate halos respectively within the survey volume in a redshift window $\\Delta z = 1$. The duty cycle of quasar activity is poorly known, but is certainly much less than unity (\\cite{HH00}). Given these uncertainties, this model is not contradictory to our discovery of a single quasar like SDSS 1044-0125 in the 600 deg$^2$ survey area. Note that this model does not address the physical process by which the massive black hole formed. The assumptions we made about $M_{\\rm BH}/M_{\\rm bulge}$ and the lifetime of the quasars are completely untested at high redshift. Indeed, Rix et al.~(1999) argue that the host galaxies of $z \\sim 2$ quasars are appreciably less luminous than the universal $M_{\\rm BH}/M_{\\rm bulge}$ hypothesis would imply. Because of its high luminosity, SDSS 1044-0125 likely probes the exponential, high--mass tail of the underlying dark halo distribution, making predictions of the expected number counts also sensitive to cosmological parameters, especially the normalization of the power spectrum (i.e. $\\sigma_8$). In principle, the SDSS survey will be able to probe quasars $\\approx 1$ mag fainter than SDSS 1044-0125, and could reveal tens of additional sources at $z\\approx 6$. The detection of these objects will yield strong constraints on cosmological models for the formation and evolution of quasars at very high redshifts. \\bigskip The Sloan Digital Sky Survey (SDSS) is a joint project of the University of Chicago, Fermilab, the Institute for Advanced Study, the Japan Participation Group, The Johns Hopkins University, the Max-Planck-Institute for Astronomy, Princeton University, the United States Naval Observatory, and the University of Washington. Apache Point Observatory, site of the SDSS, is operated by the Astrophysical Research Consortium. Funding for the project has been provided by the Alfred P. Sloan Foundation, the SDSS member institutions, the National Aeronautics and Space Administration, the National Science Foundation, the U.S. Department of Energy, and Monbusho, Japan. The SDSS Web site is {\\tt http://www.sdss.org/}. This publication makes use of data products from the Two Micron All Sky Survey, which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by NASA and NSF. XF and MAS acknowledge additional support from Research Corporation, NSF grant AST96-16901, the Princeton University Research Board, and a Porter O. Jacobus Fellowship. RHB acknowledges support from the Institute of Geophysics and Planetary Physics (operated under the auspices of the U.S. Department of Energy by the University of California Lawrence Livermore National Laboratory under contract No.~W-7405-Eng-48). ZH acknowledges support from Hubble Fellowship grant HF-01119.01-99A. DPS acknowledges support from NSF grant AST99-00703. We thank Wolfgang Voges, Hans-Walter Rix, David Weinberg, and Peng Oh for helpful comments, and the expert assistance of Bob Goodrich and Terry McDonald during the Keck observations." }, "0005/astro-ph0005552_arXiv.txt": { "abstract": "The Infrared Space Observatory observed the field of the $\\gamma$--ray burst GRB\\,970508 with the CAM and PHT instruments on May 21 and 24, 1997 and with PHT in three filters in November 1997. A source at 60\\,$\\mu$m (flux in May of $66\\pm 10$\\,mJy) was detected near the position of the host galaxy of this $\\gamma$--ray burst. The source was detected again in November 1997, at a marginally lower flux ($43\\pm 13$\\,mJy). A Galactic cirrus origin and a stellar origin for the emission can be ruled out on the basis of the infrared colours. The marginal evidence for variability in the 60\\,$\\mu$m flux between May and November is not sufficient to warrant interpretation of the source as transient fireball emission. However, the infrared colours are physically reasonable if attributed to conventional dust emission from a single blackbody source. The probability of detecting a 60\\,$\\mu$m by chance in a PHT beam down to a detection limit of 50\\,mJy is $\\sim 5\\times 10^{-3}$. If the source is at the redshift of the host galaxy of the $\\gamma$--ray burst the fluxes and upper limits at wavelengths from 12\\,$\\mu$m to 170\\,$\\mu$m indicate it is an ultraluminous infrared galaxy (L$_{\\rm ir} \\sim 2\\times 10^{12}$\\,L$_{\\sun}$). The star formation rate is estimated to be several hundred solar masses per year, depending significantly on model-dependent parameters. If this source is associated with the host galaxy of GRB\\,970508, progenitor models which associate GRBs with star-forming regions are favoured. ", "introduction": "The recent discoveries of fading afterglows to a number of $\\gamma$--ray bursts (GRBs) have been precipitated by the accurate and prompt localisation capability of the two Wide Field Cameras (WFC) aboard the `BeppoSAX' X-ray satellite (Boella et al. 1997; Piro et al. 1998a). \\nocite{bb97,pcf+98} The typical $\\sim 3'$ radius error circles, obtained in a matter of hours after the occurrence of the GRBs, are easily covered by ground-based optical and radio telescopes, allowing rapid and deep follow-up observations. The reduction of GRB error circles by BeppoSAX also made it feasible to study the content of these regions at far-infrared wavelengths with the European Space Agency's Infrared Space Observatory\\footnote{Based on observations with ISO, an ESA project with instruments funded by ESA Member States (especially the PI countries: France, Germany, the Netherlands and the United Kingdom) with the participation of ISAS and NASA.}, ISO \\cite{ksa+96}. The error region of GRB\\,970402 was the first to be rapidly surveyed at far-infrared wavelengths \\cite{ajct98}. ISO observed this GRB error circle 55 hours, and again 8 days, after the burst event and detected no new transient sources down to a 5\\,$\\sigma$ limit of 0.14\\,mJy at 12\\,$\\mu$m and 350\\,mJy at 170\\,$\\mu$m. Details of the ISO observations of GRBs may be found in Delaney et al. (1999) \\nocite{dhm+99}. ", "conclusions": "The ISO observations of GRB\\,970508 place unique limits on the level of far-infrared emission in the weeks and months following the burst event. Non-transient emission observed at 60\\,$\\mu$m indicates the presence of a ULIG which may be the parent of the host galaxy of the GRB. This result may have important implications for GRB progenitor models, favouring those which place GRBs in or near star-forming regions." }, "0005/astro-ph0005287_arXiv.txt": { "abstract": "The Sunyaev-Zel'dovich (SZ) effect associated with pressure fluctuations of the large scale structure gas distribution will be probed with current and upcoming wide-field small angular scale cosmic microwave background experiments. We study the generation of pressure fluctuations by baryons which are present in virialized dark matter halos, with overdensities $\\gtrsim 200$ and by baryons present in overdensities $\\lesssim 10$. For collapsed halos, assuming the gas distribution is in hydrostatic equilibrium with matter density distribution, we predict the pressure power spectrum and bispectrum associated with the large scale structure gas distribution by extending the dark matter halo approach which describes the density field in terms of correlations between and within halos. The projected pressure power spectrum allows a determination of the resulting SZ power spectrum due to virialized structures. The unshocked photoionized baryons present in smaller overdensities trace the Jeans-scale smoothed dark matter distribution. They provide a lower limit to the SZ effect due to large scale structure in the absence of massive collapsed halos. We extend our calculations to discuss higher order statistics, such as bispectrum and skewness in SZ data. The SZ-weak lensing cross-correlation is suggested as a probe of correlations between dark matter and baryon density fields, while the probability distribution functions of peak statistics of SZ halos in wide field CMB data can be used as a probe of cosmology and non-Gaussian evolution of large scale pressure fluctuations. ", "introduction": "In recent years, increasing attention has been given to the physical properties of the intergalactic warm and hot plasma gas distribution associated with large scale structure and the possibility of its detection (e.g., \\cite{CenOst99}). It is now widely believed that at least $\\sim$ 50\\% of the present day baryons, when compared to the total baryon density through big bang nucleosynthesis, are present in this warm gas distribution and have remained undetected given its nature (e.g., \\cite{Fuketal98}). Currently proposed methods for the detection of this gas with include observations of the thermal diffuse X-ray emission (e.g., \\cite{Pieetal00}), associated X-ray and UV absorption and emission lines (e.g., \\cite{Trietal00}) and resulting Sunyaev-Zel'dovich (SZ; \\cite{SunZel80}) effect (e.g., \\cite{Cooetal00a}). The SZ effect arises from the inverse-Compton scattering of CMB photons by hot electrons along the line of sight. This effect has now been directly imaged towards massive galaxy clusters (e.g., \\cite{Caretal96,Jonetal93}), where temperature of the scattering medium can reach as high as 10 keV producing temperature changes in the CMB of order 1 mK at Rayleigh-Jeans wavelengths. Previous analytical predictions of the resulting SZ effect due to large scale structure have been based on either through a Press-Schechter (PS; \\cite{PreSch74}) description of the contributing galaxy clusters (e.g., \\cite{ColKai88,KomKit99}) or using a biased description of the pressure power spectrum with respect to the dark matter density field (e.g., \\cite{Cooetal00a}). Numerical simulations (e.g., \\cite{daS99,Refetal99,Seletal00}) are beginning to improve some of these analytical predictions, but are still limited to handful of simulations with limited dynamical range and resolution. Therefore, it is important that one consider improving analytical models of the large scale structure SZ effect, and provide predictions which can be easily tested through simulations. Our present study on the large scale baryon pressure and the resulting SZ effect is timely for several reasons, including the fact that improving numerical simulations have recently begun to make detailed predictions for the pressure power spectrum and SZ effect such that those predictions can be extended and improved with analytical models \\cite{daS99,Refetal99,Seletal00}. Also, several studied have considered the possibility that large scale baryon distribution can be probed with upcoming CMB missions using SZ effect (e.g., \\cite{Cooetal00a}). Our calculations can be used to further refine these predictions and to investigate the possibility how such analytical model as the one presented here can be tested with observations. As part of this study, we extend previous studies by considering the full power spectrum and bispectrum, the Fourier space analog of the three-point function, of pressure fluctuations. The pressure power spectrum and bispectrum contains all necessary information on the large scale distribution of temperature weighted baryons, whereas, the SZ power spectrum is only a projected measurement of the pressure power spectrum. This can be compared to weak gravitational lensing, where lensing is a direct probe of the projected dark matter density distribution. The bispectrum of pressure fluctuations, and SZ bispectrum, contains all the information present at the three-point level, whereas conventional statistics, such as skewness, do not. An useful advantage of using the 3d statistics, such as the pressure power spectrum, is that they can directly compared to numerical simulations, while only 2d statistics, such as the projected pressure power spectrum along the line of sight, basically the SZ power spectrum, can be observed. Our approach here is to consider both such that our calculations can eventually be compared to both simulations and observations. The calculation of pressure power spectrum and bispectrum requires detailed knowledge on the baryon distribution, which can eventually be obtained numerically through hydrodynamical simulations. Here, we provide an analytical technique to obtain the pressure power spectrum and bispectrum by describing the baryon distribution in the universe as (1) present in virialized halos with overdensities $\\gtrsim 200$ with respect to background densities (2) unshocked diffuse baryons in overdensities $\\lesssim 10$ that trace a Jeans-smoothed dark matter density field (3) the intermediate overdensity region, which is likely to be currently undergoing in shock heating and falling on to structures such as filaments. In the present paper we discuss the first two regimes, while a useful approach to include the latter, through simulations, is discussed. Our description of baryons present in virialized halos follow recent studies on the dark matter density field through halo contributions \\cite{Sel00,MaFry00,Scoetal00} following \\cite{SchBer91} and applied to lensing statistics in \\cite{Cooetal00b} and \\cite{CooHu00b}. For the description of baryons, the critical ingredients are: the PS formalism \\cite{PreSch74} for the mass function; the NFW profile of \\cite{Navetal96}, and the halo bias model of \\cite{Moetal97}. The baryons are assumed to be in hydrostatic equilibrium with respect to dark matter distribution, which is a valid assumption, at least for the high mass halos that have been observed with X-ray instruments, given the existence of regularity relations between cluster baryon and dark matter physical properties (e.g., \\cite{MohEvr97}). We take two descriptions of the temperature structure: (1) virial temperature and (2) virial temperature plus an additional source of minimum energy. The latter consideration allows the possibility for a secondary source of energy for baryons, such as due to preheating through stellar formation and feedback processes. Numerical simulations (e.g., \\cite{CenOst99,Pen99}), as well observations (e.g., \\cite{Davetal95,Ren97}), suggest the existence of such an energy source. The low photoionized overdensity baryons are described following the analytical description of \\cite{GneHui98}. The fraction of baryons present in such low overdensities are assumed to follow what has been measured in numerical simulations of \\cite{CenOst99}. We suggest that such baryons provide a lower limit to the SZ effect in the absence of any contribution from baryons present in virialized halos. 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}) and consistent with COBE \\cite{BunWhi97}. For the linear power spectrum, we take the fitting formula for the transfer function given in \\cite{EisHu99}. The paper is organized as following: In \\S \\ref{sec:density}, we review the dark matter halo approach to modeling the density field and extend it to model properties associated with large scale baryon distribution, mainly the pressure fluctuations that contributes to the observable SZ effect. We suggest recent papers by Seljak \\cite{Sel00}, Ma \\& Fry \\cite{MaFry00}, Cooray \\& Hu \\cite{CooHu00b}, and Scoccimarro et al \\cite{Scoetal00} for details on the dark matter halo approach and applications to other observable statistics such as galaxy properties and weak gravitational lensing. As necessary, we use techniques developped in these papers for our current calculation. In \\S \\ref{sec:sz} we apply the formalism to the convergence power spectrum, skewness, and bispectrum. We conclude in \\S \\ref{sec:conclusions} with a summary of our main results. \\begin{figure*}[t] \\centerline{\\psfig{file=fig1a.eps,width=3.8in,angle=-90} \\psfig{file=fig1b.eps,width=3.8in,angle=-90}} \\caption{Pressure power spectra with single halo (PP; dotted line), double halo (hh; dashed line) and total (t; solid line) contributions. (a) Here, we use virial temperature to describe electrons and show pressure bias, $b_\\pi(k)$ (long dashed line), filtered non-linear dark matter density power spectrum (\\S~\\ref{sec:gnehui}) and the measured pressure power spectrum in simulations (thick solid line). (b) The variations in pressure power with maximum mass. The dot-dashed lines show the total power with a minimum electron temperature of $0.75$ keV, as an attempt to reproduce power spectra under a possible preheating scenario.} \\label{fig:pressurepower} \\end{figure*} ", "conclusions": "\\label{sec:conclusions} Using an extension of the dark matter halo approach, we have presented an efficient method to calculate the large scale structure pressure power spectrum and its high order moments, such as bispectrum. We have divided the contribution to large scale pressure power spectrum based on the overdensities in which contributing baryons are present with (1) baryons present in virialized halos with overdensities greater than $\\sim$ 200 and in hydrostatic equilibrium with the density field of such halos, (2) photoionized baryons in overdensities less than $\\sim$ 10 and which trace the Jeans-scale smoothed dark matter density field, and (3) baryons present in the mid overdensity regime which are likely to be undergoing collapse and shock heating. Our approach allows us to calculate not only 2d statistics such as the projected pressure power spectrum, or the SZ effect, which will be observed, but also the 3d statistics that can be directly compared to predictions based on numerical simulations. We have performed such a comparison to recently published numerical simulations by \\cite{Refetal99} and found good agreement between our analytical calculations and their simulations. The current simulations are limited to a handful of realizations and limited dynamical range and resoltion. With improving resolution and accuracy, analytical models such as the one presented here will be tested in detail against numerical calculations. Analytical calculations, aided by numerical simulations, will eventually allow detailed studies of large scale baryon distribution using observations such as the wide-field SZ effect. The projected pressure power spectrum along the line of sight, provides a direct calculation of the large scale structure SZ effect and its higher order correlations. In the absence of massive and rare halos, we have suggested that baryons present in small overdensities provide a lower limit to any contribution to SZ effect. The extent to which baryons present in overdensities between 10 and 200 contribute to the correlations in large scale pressure and, from it, the SZ effect, requires additional studies, preferably with numerical simulations. Presently, our Understanding the role of preheating and its effect of baryons will also be another challenge as the SZ observations will clearly depend on such additional energy contributions to large scale baryon distribution. We have suggested the possibility of using SZ power spectrum and higher order correlations, such as the SZ skewness, as a probe of preaheating. Such a study will require a wide-field SZ map and this task will be completed with Planck observations. The unique frequency dependance of the SZ effect, together with its non-Gaussian behavior, will allow the construction of a reliable SZ cluster catalog with will aid in cosmologicalo studies of structure formation. Our approach allows one to study possible systematic effects that may be present in upcoming SZ observations of small area fields due to the presence or absence of rare massive halos in such fields that will be observed. We have shown that the SZ effect as well as its non-Gaussian properties are mainly due to the most massive and rarest virialized halos in the universe. The lack of massive halos in observed SZ fields can introduce a systematic bias in the power spectrum, but the sample variance introduced by the lack of such masses, can be easily corrected based on the prior knowledge on mass distribution of observed fields. Due to additional mass dependence through temperature, the effect of mass is such that the SZ effect is more dependent on the rare halos than weak gravitational lensing convergence. The same SZ halos also contribute to lensing convergence and the cross-correlation between SZ and lensing can be used a probe of clustering properties between density and temperature weighted baryon fields. Given the great potential to study baryon distribution using SZ, various issues suggested here involving such correlations merit further study." }, "0005/astro-ph0005471_arXiv.txt": { "abstract": "We investigate the impact of hyperfine splitting on stellar abundance analyses of Mn and Sc, and find that incorrect hfs treatment can lead to spurious abundance trends with metallicity. We estimate corrections to a recent study by Nissen et al. (2000), and find: (1) $\\lbrack$Mn/Fe$\\rbrack$ is described by a bimodal distribution, with $\\lbrack$Mn/Fe$\\rbrack$~$\\sim -$0.3 for stars $\\lbrack$Fe/H$\\rbrack$~$< -0.7$, and $\\lbrack$Mn/Fe$\\rbrack$~$\\sim -$0.05 for stars at higher metallicity, suggestive of a transition between halo/thick disk and thin disk populations. (2) The large majority of stars show nearly solar $\\lbrack$Sc/Fe$\\rbrack$ ratios; although important deviations cannot be ruled out. ", "introduction": "A fundamental goal of stellar abundance studies is to investigate and identify the nucleosynthetic origin and evolution of individual elements in various stellar populations. Observed abundance ratios may be used to constrain theoretical nucleosynthesis yields \\citep[e.g.][]{ww95}, and gain insight into the formation and chemical enrichment history of galactic systems. \\cite{tinsley79} developed a paradigm of chemical evolution to explain early observations \\citep{waller62,conti67} which showed that oxygen and other $\\alpha$-elements (e.g. Mg, Si, Ca) are enhanced in metal-poor halo stars relative to iron, but not in solar-metallicity disk stars (for more discussion see McWilliam 1997). Based on this simple idea one expects certain trends of [$\\alpha$/Fe] with [Fe/H] for systems with different formation time scales (i.e. SFR, IMF, and binary fractions); thus the measurement of [$\\alpha$/Fe] can be used to probe these parameters, and our understanding of chemical evolution. This also suggests that additional diagnostics of the Galactic star formation history may be obtained by careful abundance studies of other elements. In the following, we consider the abundance trends of Sc and Mn, two so-called iron-peak elements. \\cite{waller62} first noted that Mn is deficient relative to Fe in low metallicity stars, consistent with the odd-even effect for iron-peak elements suggested by \\cite{helfer59} and \\cite{arnett71}. These early observations were later confirmed by \\cite{gratton89}, who found that Mn behaves in an opposite sense to the $\\alpha$-elements: a steady decline to [Mn/Fe]~$\\sim -0.34$ from solar metallicity to [Fe/H]~$\\sim -1.0$~dex, followed by a plateau at [Mn/Fe]~$\\sim -0.34$ between [Fe/H] = $-$1.0 to $-$2.4~dex. Gratton concluded that Mn is under-produced in early massive stars, and over-produced in type~Ia SN. Below [Fe/H]~$\\sim -2.5$, \\cite{mpss2} showed that the [Mn/Fe] ratio declines rapidly, and suggests the presence of a distinct early population~II, or population~III in the Galactic halo. More recently, Nissen et al.\\ (2000; N00) published a comprehensive study of $\\approx$120 [Mn/Fe] measurements from F and G dwarf stars in the metallicity interval $-1.4<[{\\rm Fe/H}]<+0.1$ . Their results found the same gradual decline in [Mn/Fe] as \\cite{gratton89}, but to lower [Mn/Fe] values and with the suggestion of a drop in [Mn/Fe] at [Fe/H]~$\\sim -0.7$ dex. If Mn is a type~Ia SN product, we expect sub-solar [Mn/Fe] in the Galactic bulge and giant elliptical galaxies, which are enhanced in type~II SN products \\citep[e.g.][]{worthey92,mr94}. For the LMC and dwarf galaxies excesses of elements produced by type~Ia SN are anticipated \\citep{wg91}. The behavior and origin of Mn is also relevant to interpretation of the abundance patterns in damped \\lya systems \\citep{lu96,pro97,ptt00}. In nearly every damped system where Mn is observed, one finds an underabundance of Mn relative to Fe which suggests an underlying type~II SN enrichment pattern independent of dust depletion. This interpretation, however, hinges on the fact that Mn is primarily produced in type~Ia SN. Identification of the formation processes for Mn, therefore, has direct bearing on the nucleosynthetic history of these protogalaxies. Nucleosynthesis calculations \\citep[e.g.][]{tww95} predict deficient [Sc/Fe] ratios in metal-poor stars. This contrasts starkly with the observations of Zhao \\& Magain (1990, 1991) who found $+$0.3 dex Sc enhancements. However, \\cite{gs91}, \\cite{peterson90}, and \\cite{mpss2} found no evidence for a deviation from [Sc/Fe]=0.0, and \\cite{gilroy88} found Sc deficient by $\\approx$0.2 dex. The recent analysis of N00 supports the enhancements found by Zhao \\& Magain, with the surprising conclusion that Sc behaves like an $\\alpha$-element; particularly convincing is the fact that N00 find relatively deficient Sc in their sub-sample of $\\alpha$-element poor stars. \\begin{table*}[ht] \\tabletypesize{\\scriptsize} \\caption{\\centerline {\\sc HYPERFINE SPLITTING OF Mn~I AND Sc~II LINES\\tablenotemark{a}}} \\begin{center} \\begin{tabular}{cc|cc|cc|cc|cc|cc|cc} \\tableline \\tableline \\tskip \\multicolumn{2}{c|}{Mn~I $\\lambda$6013} & \\multicolumn{2}{|c|}{Mn~I $\\lambda$6016} & \\multicolumn{2}{|c|}{Mn~I $\\lambda$6021} & \\multicolumn{2}{|c|}{Sc~II $\\lambda$5526} & \\multicolumn{2}{|c|}{Sc~II $\\lambda$5657} & \\multicolumn{2}{|c|}{Sc~II $\\lambda$6245} & \\multicolumn{2}{|c}{Sc~II $\\lambda$6604} \\\\ \\multicolumn{1}{c}{$\\Delta \\lambda$\\tablenotemark{b}} & $\\log gf$\\tablenotemark{c} & \\multicolumn{1}{c}{$\\Delta \\lambda$} & $\\log gf$ & \\multicolumn{1}{c}{$\\Delta \\lambda$} & $\\log gf$ & \\multicolumn{1}{c}{$\\Delta \\lambda$} & $\\log gf$ & \\multicolumn{1}{c}{$\\Delta \\lambda$} & $\\log gf$ & \\multicolumn{1}{c}{$\\Delta \\lambda$} & $\\log gf$ & \\multicolumn{1}{c}{$\\Delta \\lambda$} & $\\log gf$ \\\\ \\tableline \\tskip .478 & $-$0.766 & .619 & $-$1.460 & .746 & $-$2.668 & .770 & $-$3.051& .886 & $-$1.229 & .621 & $-$1.736 & .582 & $-$2.505 \\\\ .499 & $-$0.978 & .645 & $-$1.297 & .772 & $-$1.451 & .775 & $-$2.648& .888 & $-$1.799 & .629 & $-$2.476 & .590 & $-$2.347 \\\\ .518 & $-$1.251 & .647 & $-$0.682 & .774 & $-$2.316 & .779 & $-$2.436& .893 & $-$1.799 & .631 & $-$1.907 & .594 & $-$1.936 \\\\ .527 & $-$1.455 & .667 & $-$1.292 & .795 & $-$1.275 & .779 & $-$1.838& .894 & $-$1.627 & .636 & $-$3.476 & .596 & $-$2.358 \\\\ .533 & $-$1.661 & .668 & $-$0.945 & .798 & $-$2.191 & .783 & $-$2.326& .895 & $-$1.641 & .638 & $-$2.293 & .599 & $-$2.333 \\\\ .538 & $-$1.309 & .684 & $-$1.276 & .804 & $-$0.533 & .783 & $-$1.633& .901 & $-$2.323 & .640 & $-$2.114 & .602 & $-$2.531 \\\\ .547 & $-$1.330 & .685 & $-$1.394 & .813 & $-$1.249 & .786 & $-$1.567& .902 & $-$1.652 & .644 & $-$3.058 & .604 & $-$3.029 \\\\ .552 & $-$1.485 & .696 & $-$1.723 & .817 & $-$2.271 & .787 & $-$2.305& .904 & $-$1.652 & .646 & $-$2.260 & .607 & $-$4.455 \\\\ .562 & $-$1.807 & .696 & $-$1.460 & .821 & $-$0.673 & .788 & $-$1.571& .906 & $-$1.825 & .647 & $-$2.385 & .609 & $-$2.707 \\\\ .566 & $-$1.853 & .698 & $-$1.656 & .827 & $-$1.316 & .789 & $-$0.909& .906 & $-$3.749 & .650 & $-$2.824 & .611 & $-$2.505 \\\\ .566 & $-$2.409 & .704 & $-$2.422 & .834 & $-$0.831 & .790 & $-$2.414& .908 & $-$1.825 & .651 & $-$2.319 & .613 & $-$2.347 \\\\ .567 & $-$2.029 & .707 & $-$1.297 & .837 & $-$1.492 & .790 & $-$1.632& .909 & $-$2.001 & .652 & $-$2.803 & .615 & $-$2.531 \\\\ & & .713 & $-$1.292 & .843 & $-$1.015 & .791 & $-$1.030& & & .655 & $-$2.678 & .615 & $-$2.358 \\\\ & & .714 & $-$1.656 & .846 & $-$1.522 & .792 & $-$1.754& & & .656 & $-$2.502 & & \\\\ & & .716 & $-$1.394 & .847 & $-$1.237 & .793 & $-$1.937& & & .657 & $-$2.581 & & \\\\ & & & & & & .793 & $-$1.166& & & & & & \\\\ & & & & & & .794 & $-$1.322& & & & & & \\\\ & & & & & & .795 & $-$2.083& & & & & & \\\\ & & & & & & .795 & $-$1.741& & & & & & \\\\ & & & & & & .795 & $-$1.507& & & & & & \\\\ \\tableline \\end{tabular} \\end{center} \\tablenotetext{a}{Line data taken from http://cfaku5.harvard.edu} \\tablenotetext{b}{The correct wavelength equals $\\lambda + \\Delta \\lambda$} \\tablenotetext{c}{The absolute $gf$ values have no bearing on our analysis} \\end{table*} While one of us (JXP) was pursuing an analysis of Mn for a sample of thick disk stars \\citep{pro00}, we became concerned about the N00 Mn analysis due to their use of the incorrect hyperfine splittings of Steffen (1985; S85). In particular, we were concerned that an incorrect hfs treatment and the change in line-strength with metallicity could lead to an artificial trend of [Mn/Fe] versus [Fe/H]. In the course of this work we also worried about the N00 treatment of Sc~II lines; the large S85 hfs splittings seem to have caused N00 to derive large hfs abundance corrections in contrast to previous work \\citep{mr94,mpss2} which found only small hfs corrections for scandium. In this paper, we estimate the error in derived Mn and Sc abundances due to use of the S85 hfs lists and consider the effects of the corrections on the abundance trends found by N00. The focus of this paper is to warn the reader of the uncertainties associated with hyperfine splitting. While we make an estimate of the true Mn and Sc abundances from the N00 sample, our analysis is limited by the fact that neither the microturbulent velocities or the iron equivalent widths were available to us. ", "conclusions": "\\label{sec-discuss} We now apply the corrections derived in the previous section in a line-by-line fashion to the N00 compilation and thereby estimate the true [Mn/Fe] and [Sc/Fe] trends. Although our results should qualitatively describe the metallicity dependence of Mn and Sc, a quantitative discussion will await a full reanalysis by the authors of N00. Figure~\\ref{fig-new}a compares the corrected [Mn/Fe] values based on our hfs calculations versus the N00 results. The corrected analysis offer a considerably different picture from that of N00. The most robust change is the systematic increase of $\\approx 0.15$~dex in [Mn/Fe] for the most metal-poor stars. This [Mn/Fe] shift arises from the substantial hfs corrections to the solar Mn abundance. One also notes that the trend toward lower [Mn/Fe] values from [Fe/H]~=~$0 \\to -0.5$ is no longer evident in the reanalysis. In short the corrected [Mn/Fe] values are bimodal; those stars with [Fe/H]~$< -0.7$ have [Mn/Fe]~$\\sim -0.3$ in agreement with \\cite{gratton89} and the higher metallicity stars show [Mn/Fe]~$\\sim -0.05$. The most straightforward explanation for these results is that [Fe/H]~$\\sim -0.7$ marks the passing from the thin disk stellar population to the thick disk and halo populations. The Mn/Fe trends are consistent with Mn being produced primarily by type~Ia SN and do not require a metallicity dependent yield. We note in passing that the plateau of Mn/Fe for the thin disk stars lies rather uncomfortably below solar: [Mn/Fe]~$\\sim -0.05$~dex. While the Sun might have a high Mn abundance relative to the bulk of the thin disk dwarf stars, the offset could be due to a zero-point error in the N00 solar analysis. Now consider the results for Sc presented in Figure~\\ref{fig-new}b. In short, the corrected values no longer show the $\\alpha$-enhanced trend discussed in the N00 analysis. The solar Sc abundance has been revised upward by $\\approx 0.2$~dex which, in turn, has eliminated the relative enhancement in the metal-poor stars. Our results agree well with the Sc abundance analysis performed by \\cite{gs91} indicating that essentially all stars exhibit solar Sc/Fe. There are two sets of stars, however, with reasonably significant departures from [Sc/Fe]~=~0: (i) the $\\alpha$-poor stars (marked with a +) identified by \\cite{nissen97} exhibit sub-solar Sc/Fe, consistent with an origin related to the $\\alpha$-elements; (ii) nearly every star with [Fe/H]~$\\sim -0.6$ shows enhanced Sc/Fe. This trend is difficult to explain and we wonder if it is simply a statistical anomaly or perhaps a small error in the original analysis of N00. Finally, we note the two weakest Sc~II lines ($\\lambda 6245,6604$) give systematically higher $\\e{Sc}$ values than the two stronger lines, particularly in the metal-poor stars. This difference may be explained by blends between the strong lines of the Solar spectrum and unidentified neutral metal lines which would lead to erroneously high $gf$ values. To account for this systematic offset, we recommend that future studies on Sc include many more Sc~II lines. In this Letter, we have investigated the effects of miscalculating the hfs correction on elemental abundance trends. We have performed a simple exercise to correct for the erroneous hfs compilation used by N00 in their analysis of Mn and Sc. In turn, we have demonstrated that hfs errors can mimic metallicity-dependent abundance trends. Our results highlight the importance of accurately addressing hyperfine splitting and we encourage the reader to utilize the large database of hfs lines being compiled by R. Kurucz. Having estimated corrections to the Mn and Sc abundances from N00, we find the following: (1) the [Mn/Fe] values are best described by a bimodal distribution with stars at [Fe/H]~$< -0.7$ exhibiting [Mn/Fe]~$\\sim -0.3$ and stars at higher metallicity showing [Mn/Fe]~$\\sim -0.05$. These trends are consistent with type~Ia SN being the principal site for Mn nucleosynthesis; (2) the large majority of stars have nearly solar Sc/Fe ratios consistent with the results of \\cite{gs91} and \\cite{mpss2}. There are exceptions, however, namely the $\\alpha$-poor stars of the N00 sample which show systematically low Sc/Fe values and the bulk of stars at [Fe/H]~$\\sim -0.6$ which exhibit enhanced Sc." }, "0005/astro-ph0005192_arXiv.txt": { "abstract": "This paper investigates the discrepancy between recent measurements of the density profile of the globular cluster 47~Tucanae that have used {\\it Hubble Space Telescope\\/} data sets. A large core radius would support the long-held view that 47~Tuc is a relaxed cluster, while a small core radius may indicate that it is in a post--core-collapse phase or possibly even on the verge of core collapse, as suggested by a variety of unusual objects --- millisecond pulsars, X-ray sources, high velocity stars --- observed in the core of the cluster. Guhathakurta et al. (1992) used pre-refurbishment Wide Field/Planetary Camera~1 (WFPC1) $V$-band images to derive $r_{\\rm core}=23''\\pm2''$. \\cite{calzetti} suggested that the density profile is instead a superposition of two King profiles, one with a small, $8''$ core radius and the other with a $25''$ core radius, based on $U$-band Faint Object Camera (FOC) images. More recently, \\cite{guido} have used deep WFPC1 $U$-band images to derive $r_{\\rm core}=12''\\pm2''$. The cluster centers used in these studies are in agreement with one another; differences in the adopted centers are not the cause of the discrepancy. Our independent analysis of the data used by De~Marchi et~al.\\ reaches the following", "introduction": "Globular clusters are excellent laboratories for studying the dynamics of a stellar system. The high density of stars near their centers results in frequent interactions---e.g.,~single star-single star, single star-binary, and binary-binary. Such interactions redistribute energy throughout the cluster and drive its global evolution on the so-called ``two-body relaxation'' timescale. This can sometimes be comparable to the typical orbital period of stars near the cluster center, and significantly shorter than the cluster age, implying that dynamical evolution is important. The orbital period or crossing/dynamical time is related to the core radius (the characteristic length scale associated with the inner density profile) and the velocity dispersion of the cluster. It is customary to characterize the radial distribution of various stellar populations in terms of the core radius (cf.~\\cite{rasio}). The inner stellar density profile of a globular cluster is suggestive of its evolutionary state (\\cite{hut96}). Most clusters are characterized by constant density cores, well fit by models based on a relaxed, Maxwellian distribution function of stars out to the limiting (tidal) radius (\\cite{king66a}). On the other hand, about twenty percent of all globular clusters appear to have undergone catastrophic gravothermal collapse as a result of runaway energy loss from the core due to two-body relaxation (\\cite{dk86}; see the recent review in \\cite{meyheg}). Even these post--core-collapse (PCC) clusters, however, are {\\it not\\/} expected to develop central singularities as heating by binaries will drive a ``quasi-steady post-collapse'' stage followed by gravothermal oscillations. Theoretical modeling suggests that clusters in the post-collapse phase should have core radii in the range $r_{\\rm core}\\sim0.01\\>$--$\\>0.04r_h$ (\\cite{good87}; \\cite{gao91}) or even as large as $r_{\\rm core}\\sim0.09r_h$ (\\cite{vc94}), where $r_h$ is the half-mass radius. These calculations are based on a single-mass stellar population; models with a realistic stellar mass function could yield a larger core radius. The globular cluster 47~Tucanae has been studied extensively, and has long been considered a prototypical relaxed globular cluster (\\cite{king85}) with a large core radius: $r_{\\rm core}\\sim25''$ (\\cite{dk84}). Recent observations of a variety of exotic objects, such as millisecond pulsars (\\cite{rob95}), X-ray sources (\\cite{hg83}; \\cite{hjv94}), and nine~stars whose velocities differ from the cluster mean by $\\gtrsim30$~km~${\\rm s}^{-1}$ (\\cite{karl95}), have raised the possibility that 47~Tuc may be approaching core collapse. The half-light (or half-mass) radius of 47~Tuc is $r_h=174''$ (\\cite{tdk93}), so if it has undergone core collapse it should have a core radius in the range $1\\farcs7$ to $16''$ (following \\cite{good87}; \\cite{gao91}; \\cite{vc94}). Two recent determinations of the core radius of 47~Tuc fall within the upper end of this range, as discussed in \\S\\,2 below. Moreover, \\cite{fisgeb95} have constructed nonparametric dynamical models based on the surface brightness and velocity dispersion profiles, and conclude that the mass profile of 47~Tuc is as centrally concentrated as that of M15. The density profile slope of 47~Tuc beyond $2'$ is found to be similar to that seen in Fokker-Planck simulations of PCC clusters (\\cite{cohn80}). Ground-based studies typically use integrated surface brightness measurements to determine a cluster's density profile (cf.~\\cite{dk84}). A substantial fraction of the optical light of a cluster comes from a handful of the brightest red giant branch (RGB) stars so that the effective Poisson error associated with integrated light measurements is large. This kind of ``sampling'' error makes the measured density profile noisy, and any error in the center determination tends to bias the measured core radius towards large values. {\\it Hubble Space Telescope\\/} ({\\it HST\\/}) images, even with the aberrated pre-refurbishment point spread function (PSF), offer the advantage of resolving individual stars down to the main sequence turnoff even in the cluster core (Guhathakurta et~al. 1992, hereafter GYSB). Star counts are more representative of the stellar mass density than the integrated light, and the effective Poisson errors are smaller (\\cite{king66b}). It is preferable to work at short wavelengths (e.g.,~the $U$ band) where the brightest RGB stars are suppressed relative to the more numerous, bluer faint subgiants: this reduces sampling effects in integrated light measurements and increases the degree of faint star completeness (most notably in the vicinity of bright giants) for star count measurements (Calzetti et~al.\\ 1993; De Marchi et~al.\\ 1996, hereafter DPSGB). This paper focuses on the question: What is the core radius of 47~Tuc as defined by the radial distribution of evolved stars? It examines recent measurements of the density profile of 47~Tuc that appear to be in disagreement with one another. In particular, the {\\it HST\\/} Wide Field/Planetary Camera~1 (WFPC1) data set analyzed by DPSGB is reanalyzed here using somewhat different photometric techniques; our results and those of DPSGB are compared to archival Wide Field Planetary Camera~2 (WFPC2) data. The background of the core radius controversy is given in \\S\\,2. In \\S\\,3, the available 47~Tuc data sets and the methodologies of this paper and DPSGB are discussed. In \\S\\,4 the photometric errors associated with each method are examined, demonstrating that the aperture photometry technique of DPSGB produces a radially-varying bias in the star counts used to measure $r_{\\rm core}$. Core radius calculations are presented in \\S\\,5 along with tests which show that the core radius discrepancy can be explained in terms of the star count bias in the DPSGB study. The conclusions of this paper are presented in \\S\\,6. ", "conclusions": "This paper presents estimates of the density profile of the globular cluster 47~Tuc based on three samples of stars (star list and photometry) derived independently from {\\it Hubble Space Telescope\\/} WFPC1 and WFPC2 images. Apparent discrepancies amongst the core radius measurements published by \\cite{gysb}, \\cite{guido}, and \\cite{calzetti} are investigated. Our conclusion is that there is severe, radially-varying bias in the magnitude-limited star counts used by De~Marchi et~al.\\ and Calzetti et~al., and this causes their core radius estimates to be spuriously low ($r_{\\rm core}\\sim14''$) relative to other determinations ($r_{\\rm core}\\sim23''$). This ``limiting magnitude bias'' is a result of large photometric scatter/bias associated with the application of their aperture photometry method to the crowded central regions of the cluster, coupled with a choice of limiting magnitude near the steep part of the stellar luminosity function. In general, such a choice of limiting magnitude is dangerous; even with symmetric errors the resulting sample will be contaminated by large numbers of stars just fainter than the cutoff. Any radial variation in the magnitude of the errors will cause the degree of this contamination to vary, resulting in an incorrect determination of the radial density profile. Combining De~Marchi et~al.'s photometry with a limiting magnitude near the main sequence turnoff at the peak of the luminosity function reduces, but does not eliminate, the discrepancy ($r_{\\rm core}\\sim18''$); the radial variations in DPSGB's photometry (larger photometric scatter/bias at small radii) have a significant effect on the derived core radius even with an optimal choice of limiting magnitude. A more accurate PSF-fitting method is used in this paper to indepedently derive two sets of stellar photometry, one from the deep WFPC1 data analyzed by De~Marchi et~al.\\ and the other from an archival WFPC2 data set. The core radii derived using these two photometry sets are independent of the choice of limiting magnitude and star list, and are consistent with each other and with previous ground-based and {\\it HST\\/} work: $r_{\\rm core}\\sim23''$ (cf.~\\cite{hr79}; \\cite{dk84}; Guhathakurta et~al.\\ 1992). The best fit core radius for the surface density distribution of evolved stars in 47~Tuc is about 15\\% of the cluster half-mass radius ($r_{\\rm h}=174''$). This is significantly larger than the range of $(r_{\\rm core}/r_{\\rm h})$ values found in numerical simulations of post--core-collapse clusters: 0.01--0.04 (\\cite{cohn80}; \\cite{good87}; \\cite{gao91}). It should be noted however that the surface brightness profile is not a perfect discriminant between a relaxed and post--core-collapse cluster, and it is advisable to combine it with velocity dispersion data (Gebhardt \\& Fischer 1995). \\bigskip \\bigskip" }, "0005/astro-ph0005537_arXiv.txt": { "abstract": "The propagation of $\\gamma$-rays through metagalactic space is associated with pair creation and subsequent inverse-Compton scattering off low-energy background radiation. As a consequence, $\\gamma$-rays of very high energy emitted by remote sources are reprocessed into the window from 10~MeV to 30~GeV conserving the injected energy. Any cosmologically distributed population of $\\gamma$-ray sources therefore contributes to the diffuse $\\gamma$-ray background in this energy band which is well-determined from recent observations with the Compton Gamma Ray Observatory (CGRO). Since the $\\gamma$-rays trace accelerated particles, the observed flux of diffuse $\\gamma$-rays also constrains the global efficiency for particle acceleration. Radio galaxies can account for the $\\gamma$-ray background if their particle acceleration efficiency considerably exceeds $\\sim 18$\\% implying that particle acceleration is an essential part of the thermodynamics in these sources. ", "introduction": "The microphysics of particle acceleration in the solar system is a fascinating realm and provides useful paradigms for particle acceleration in other places of the Universe. From the observed flux of cosmic rays it is clear that much more powerful particle accelerators than main sequence stars and their winds must be at work in our Galaxy and in other galaxies. Supernova remnants \\cite{ellison97} and radio galaxies \\cite{biermann95} are among the most interesting putative sources of cosmic rays. The cosmic ray pressure is rather large and compares with that of the interstellar magnetic fields implying that cosmic rays are an important ingredient in the dynamics of the interstellar medium, and possibly also in the dynamics of the intergalactic medium (in superclusters). This imposes strong constraints on the energetics of possible cosmic ray sources from which one can infer the particle acceleration efficiency. For instance, supernova blast waves can supply the flux of cosmic rays up to the so-called knee in their spectrum at an energy of $10^{15}$~eV only if their particle acceleration efficiency is as large as $\\sim 13\\%$ \\cite{drury90} which basically invalidates simple test-particle approaches for the description of the acceleration mechanism. This has enforced the two-fluid theory for shock acceleration in which the momentum flux due to accelerated particles is self-consistently included into the dynamics of the shock wave \\cite{achterberg84}. In this contribution it is argued that if radio galaxies are the most powerful particle accelerators in the Universe responsible for the diffuse isotropic gamma-ray background, and possibly also for the ultrahigh-energy cosmic rays up to at least $5\\times 10^{19}$~eV (the so-called Greisen-Zatsepin-Kuz\\'min cutoff), they require a particle acceleration efficiency even larger than that of supernova remnants. ", "conclusions": "The particle acceleration efficiency of the jets in radio galaxies must be larger than $\\sim 18\\%$ if the $\\gamma$-rays from these jets are to explain the DIGB." }, "0005/astro-ph0005067_arXiv.txt": { "abstract": "Mkn~421 was repeatedly observed with \\textit{Beppo}SAX in 1997--1998. The source showed a very rich phenomenology, with remarkable spectral variability. This is the second of two papers presenting the results of a thorough temporal and spectral analysis of all the data available to us, focusing in particular on the flare of April 1998, which was simultaneously observed also at TeV energies. The spectral analysis and correlations are presented in this paper, while the data reduction and timing analysis are the content of the companion paper. The spectral evolution during the flare has been followed over few ks intervals, allowing us to detect for the first time the peak of the synchrotron component shifting to higher energies during the rising phase, and then receding. This spectral analysis nicely confirms the delay of the flare at the higher energies, which in Paper~I we quantified as a hard lag of a few ks. Furthermore, at the highest energies, evidence is found of variations of the inverse Compton component. The spectral and temporal information obtained challenge the simplest models currently adopted for the (synchrotron) emission and most importantly provide clues on the particle acceleration process. A scenario accounting for all the observational constraints is discussed, where electrons are injected at progressively higher energies during the development of the flare, and the achromatic decay is ascribed to the source light crossing time exceeding the particle cooling timescales. ", "introduction": "\\label{sec:introduction} Blazars are radio--loud AGNs characterized by strong variability, large and variable polarization, and high luminosity. Radio spectra smoothly join the infrared-optical-UV ones. These properties are successfully interpreted in terms of synchrotron radiation produced in relativistic jets and beamed into our direction due to plasma moving relativistically close to the line of sight (e.g. Urry \\& Padovani \\citealp{up95}). Many blazars are also strong and variable sources of GeV $\\gamma$--rays, and in a few objects, the spectrum extends up to TeV energies. The hard X to $\\gamma$--ray radiation forms a separate spectral component, with the luminosity peak located in the MeV--TeV range. The emission up to X--rays is thought to be due to synchrotron radiation from high energy electrons in the jet, while it is likely that $\\gamma$-rays derive from the same electrons via inverse Compton (IC) scattering of soft (IR--UV) photons --synchrotron or ambient soft photons (e.g. Sikora, Begelman \\& Rees~\\citealp{sbr94}; Ghisellini \\& Madau \\citealp{gg_madau_96}; Ghisellini et al. \\citealp{gg_sed98}). The contributions of these two mechanisms characterize the average blazar spectral energy distribution (SED), which typically shows two broad peaks in a $\\nu F_\\nu$ representation (e.g. Fossati et al. \\citealp{fg_sed98}). The energies at which the peaks occur and their relative intensity provide a powerful diagnostic tool to investigate the properties of the emitting plasma, such as electron energies and magnetic field (e.g. Ghisellini et al. \\citealp{gg_sed98}). In X--ray bright BL Lacs (HBL, from High-energy-peak-BL Lacs, Padovani \\& Giommi \\citealp{pg95}) the synchrotron maximum occurs in the soft-X--ray band. Variability studies constitute the most effective means to constrain the emission mechanisms taking place in these sources as well as the geometry and modality of the energy dissipation. The quality and amount of X--ray data on the brightest sources start to allow thorough temporal analysis as function of energy and the characterization of the spectral evolution with good temporal resolution. Mkn~421 is the brightest HBL at X--ray and UV wavelengths and thus it is the best available target to study in detail the properties of the variability of the highest frequency portion of the synchrotron component, which traces the changes in the energy range of the electron distribution which is most critically affected by the details of the acceleration and cooling processes. This paper is the second of two, which present the uniform analysis of the X--ray variability and spectral properties from \\textit{Beppo}SAX observations of Mkn~421 performed in 1997 and 1998. In Paper~I (Fossati et al. \\citealp{fossatiI}) we presented the data reduction and the timing analysis of the data, which revealed a remarkably complex phenomenology. The study of the characteristics of the flux variability in different energy bands shows that significant spectral variability is accompanying the pronounced changes in brightness. In particular, a more detailed analysis of the remarkable flare observed in 1998 revealed that: i) the medium energy X--rays lag the soft ones, ii) the post--flare evolution is achromatic, and iii) the light curve is symmetric in the softest X--ray band, and it becomes increasingly asymmetric at higher energies, with the decay being progressively slower that the rise. The general guidelines which we followed for the data reduction and filtering are described in Paper~I (in particular in \\S2 and \\S3.2). Here we will only report on details of the treatment of the data specific to the spectral fitting. The paper is organized as follows. In Sections~\\S\\ref{sec:bepposax_overview} and \\S\\ref{sec:observations} we briefly summarize the basic information on \\textit{Beppo}SAX and the 1997 and 1998 observations. The results and discussion relative to the spectral analysis are the content of \\S\\ref{sec:spectral_analysis} and \\S\\ref{sec:discussion}. In particular, the observed variability behavior strongly constrains any possible time dependent particle acceleration prescription. We will therefore consider these results together with those of the temporal analysis, discuss which constraints are provided to current models and present a possible scenario to interpret the complex spectral and temporal findings (\\S\\ref{sec:interpretation}). Finally, we draw our conclusions in \\S\\ref{sec:conclusions}. ", "conclusions": "" }, "0005/astro-ph0005584_arXiv.txt": { "abstract": "Type Ia supernovae (SN Ia) are generally believed to be the result of the thermonuclear disruption of Chandrasekhar-mass carbon-oxygen white dwarfs, mainly because such thermonuclear explosions can account for the right amount of $^{56}$Ni, which is needed to explain the light curves and the late-time spectra, and the abundances of intermediate-mass nuclei which dominate the spectra near maximum light. Because of their enormous brightness and apparent homogeneity SN Ia have become an important tool to measure cosmological parameters. In this article the present understanding of the physics of thermonuclear explosions is reviewed. In particular, we focus our attention on subsonic (``deflagration'') fronts, i.e. we investigate fronts propagating by heat diffusion and convection rather than by compression. Models based upon this mode of nuclear burning have been applied very successfully to the SN Ia problem, and are able to reproduce many of their observed features remarkably well. However, the models also indicate that SN Ia may differ considerably from each other, which is of importance if they are to be used as standard candles. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005251_arXiv.txt": { "abstract": " ", "introduction": "Optical observations jumped out to a rapid lead in studies of high-redshift (z$\\ge$1) galaxies, because of the sensitivity of large-format detectors, and the accessibility of very strong spectral features in the far-UV rest frame at z$\\ge$ 2.5 (\\lya, the trinity of deep interstellar absorptions due to Si~II, O~I, and Si~II, at 1260, 1303, and 1334\\AA, respectively, and of course the dramatic Lyman continuum ``break\" which is robustly produced by both intrinsic and extrinsic absorption in (presumably) all normal galaxies at z$\\ge$2.5--3. (Steidel,\\et\\ 1996) \\footnote {As long as the intrinsic depth of the Lyman limit break is ``deep enough\" (e.g. two magnitudes), it can be used as an excellent indicator of high-redshift galaxies in broad-band photometry which extends to short enough wavelengths. However, the full depth of the Lyman break has still not been actually measured in galaxies lacking active nuclei. HUT provided a few strong upper limits to the amount of Lyman continuum which manages to leak out of nearby galaxies, and a deep FUV imaging program with the STIS MAMA detector in HST's Cycle 9 should be able to {\\it detect} this emission in a sample of moderate-redshift ($1.1 \\ge z \\ge 1.4$) starburst galaxies (GO program 8561).} In surprisingly short order, this lead to an impressive list of results e.g. luminosity functions and integrated star formation rates, morphologies, and even spatial clustering (Steidel, \\et\\ 1998). It is certainly exciting to be able to see much data on an evolutionary phase in normal galaxies that was hardly observed at all less than ten years ago. \\footnote {The last time I attended a meeting at this lovely conference center, in 1988, there were virtually {\\it no} observational data on normal galaxies above redshifts of 1.} However, as the initial euphoria of discovery has subsided, researchers have started to face the practical reality: all of this new information, as invaluable as it is, rests fundamentally on the strong stellar continuum escaping from these galaxies in the far-UV. It is well known that hardly any other region of the electromagnetic spectrum is more sensitive to absorption by interstellar dust grains. And this effect is especially strong when we are observing newly-formed stars, since they are too short-lived to escape from the interstellar matter out of which they are born. This leads immediately to the outstanding questions about these ``Lyman Break\" galaxies: 1) how can we estimate their properties, such as mass and metallicity, 2) which of their measured properties need to be corrected, 3) by how much, and 4) what galaxies are missing altogether from these searches? Thus the next task is to use observations at longer wavelengths to obtain some answers. This review is an attempt to see how long-wavelength observations are starting to catch up with the optical ones, and fill in key gaps. ", "conclusions": "" }, "0005/astro-ph0005298_arXiv.txt": { "abstract": "Low resolution spectra of SN 1999dn at early times are presented and compared with synthetic spectra generated with the parameterized supernova synthetic-spectrum code SYNOW. We find that the spectra of SN 1999dn strongly resemble those of SN 1997X and SN 1984L, and hence we classify it as a Type Ib event. Line-identifications are established through spectrum synthesis. Strong evidence of both ${\\rm H}{\\alpha}$ and C II $\\lambda 6580$ is found. We infer that ${\\rm H}{\\alpha}$ appears first, before the time of maximum brightness, and then is blended with and finally overwhelmed by the C II line after maximum; this favors a thin high-velocity hydrogen skin in this Type Ib supernova. ", "introduction": "Type Ib and Ic supernovae (SNe Ib and SNe Ic) are distinguished from SNe Ia primarily by their lack of a strong $6150{\\rm~\\AA}$ Si II absorption feature in early-time spectra \\citep{fil97}. Their nature is still in debate due to their relative rareness \\citep{cap99} and faintness \\citep{pie97}. The most widely accepted scenario is the core-collapse of a massive star that has lost its hydrogen (SN Ib) and possibly helium envelope (SN Ic) either through a stellar wind or the effects of a close companion, although other models like the thermonuclear explosions of white dwarfs have not been completely excluded. Currently, there is intense interest in SNe Ib/Ic because of their possible connection with $\\gamma$-ray bursts \\citep{whe99}. SN 1999dn, located in NGC 7714, was discovered by \\citet{qiu199} on Aug. 19.76UT with the BAO 0.6-m telescope. It was originally identified as a Type Ic \\citep{aya99,tur99} or Ib/c event \\citep{pas99}. The derived velocity of the weak ``Si II $\\lambda 6355$'' absorption line is much smaller than that of other lines, which encouraged us to reconsider the line-identification and spectroscopic classification through spectrum synthesis. We used the fast parameterized LTE code SYNOW and early-time spectra observed with the BAO 2.16-m telescope. The observations, spectrum synthesis procedure, and results are presented and discussed here. ", "conclusions": "\\citet{pas99} indicated that the spectra of SN 1999dn were very similar to that of SN 1997X and identified it as a Type Ib/c event in accordance with the previous classification of the latter \\citep{gar97,mun98}, while others noted both as Type Ic supernovae \\citep{sun97,ben97,aya99,tur99}. But by comparing the spectrum of SN 1999dn in Aug. 30 and of SN 1997X, presented by \\citet{mun98}, with that of the prototypical Type Ib SN 1984L of Aug. 30, plotted by \\citet{har87} in their Figure 3, one can see that those three spectra, all near maximum light, strikingly resemble each other almost in all features and differ with the early-time spectrum of Type Ic supernovae mainly in two aspects: (1) the He I optical P Cygni lines of SN 1999dn, SN 1997X and SN 1984L are prominent, while in the prototypical Type Ic SN 1994I He I line are hardly discernible, although \\citet{clo96} claimed evidence of very weak He I lines in spectra with good signal-to-noise ratio; and (2) unlike the $6150{\\rm~\\AA}$ absorption feature usually seen in the early-time spectra of SNe Ic, the minimum at $\\sim 6200{\\rm~\\AA}-6400{\\rm~\\AA}$ in SN 1999dn, SN 1997X and SN 1984L is hard to attribute to Si II $\\lambda 6355$, as demonstrated in the lower panels of Figure 1 --- Figure 3 . It has been supposed that C II $\\lambda 6580$ is responsible for this feature in SNe Ib \\citep{har87}; this is supported by our spectrum synthesis here and in particular by the identification of C II $\\lambda\\lambda 4738,4745$ in the Sep. 14 spectrum of SN 1999dn. We regard both SN 1999dn and SN 1997X as typical Type Ib supernovae. The relative strengths of the He I lines $\\lambda\\lambda 5876$, $6678$ and $7065$ can not be completely fitted by SYNOW. In Figure 1 and Figure 2, we choose to fit $\\lambda 5876$ best, and then $\\lambda\\lambda 6678$, $7065$ are somewhat weaker than observed. In Figure 3, the calculated $\\lambda 6678$ is the right strength, but $\\lambda 5876$ is too strong and $\\lambda 7065$ is too weak. Such NLTE effects may be caused by nonthermal excitation from Comptonized gamma rays released in the decay of $^{56}{\\rm Ni}$ that has been mixed into the helium envelope \\citep{shi90,luc91}. Note the Ni II line which we introduced above to fit the ${3930 \\rm~\\AA}$ absorption in the spectrum of Sep. 14. This could be direct early-time spectral evidence of $^{56}{\\rm Ni}$ mixing in the explosion of a Type Ib supernova. As for the Aug. 31 spectrum, despite its poor signal-to-noise ratio, we are sure that the introduction of Ni II lines to the synthetic spectrum can not improve the fit. The velocity upper limit of the mixed $^{56}{\\rm Ni}$ may be $\\sim 10,000{\\rm~km~s^{-1}}$, which is larger than the value predicted by explosion simulations \\citep{har94,kif99}. One question remains of the minimum velocity of C II that we used to explain the absorption near $6300{\\rm~\\AA}$. Why should this ion be detached around $20,000{\\rm~km~s^{-1}}$ at the earliest times, but then detached at a much lower velocity of $\\sim 10,000{\\rm~km~s^{-1}}$ on Sep. 14? Does it imply two different C II velocity components in the helium layer? An alternative scenario that seems more attractive and probable, is that at the earliest times this feature is produced by ${\\rm H}{\\alpha}$, which always fits as well as C II $\\lambda 6580$ (see the dotted lines in the lower panels of Figure 1 --- Figure 3). In the Aug. 21 and Aug. 31 spectra, H I would be detached by $19,000{\\rm~km~s^{-1}}$ and $18,000{\\rm~km~s^{-1}}$, respectively. In the Sep. 14 spectrum, on the other hand, the identification of C II $\\lambda 6580$ is unambiguous because H I can not explain the observed absorption at $4580{\\rm~\\AA}$. Actually, on Aug. 31 a contribution from C II can not be excluded because the observed feature is stronger than in Aug. 21 and Sep. 14. The other synthesized H I lines like ${\\rm H}{\\beta}$ are swamped by Fe II lines and are too weak to be identified. To summarize, we suppose that a thin high-velocity hydrogen skin exits outside of the helium layer in SN 1999dn, with ${\\rm H}{\\alpha}$ appearing only before and near maximum light. As the envelope expands and the photosphere recedes deeper into the helium layer, the optical depth of ${\\rm H}{\\alpha}$ diminishes naturally and eventually is overwhelmed by C II $\\lambda 6580$. We note that other authors \\citep{fil90,jef91} have presented some evidence, though inconclusive, of ${\\rm H}{\\alpha}$ in some Type Ic supernovae, especially SN 1987M, and argued that SNe II and SNe Ic have similar physical origins. Similarly, our identification of ${\\rm H}{\\alpha}$ in SN 1999dn provides a link between the progenitors and explosion mechanisms of Type Ib supernovae and Type IIb events such as 1987K and 1993J." }, "0005/astro-ph0005121_arXiv.txt": { "abstract": "It was recently pointed out that the distribution of times between solar flares (the flare waiting-time distribution) follows a power law, for long waiting times. Based on 25 years of soft X-ray flares observed by Geostationary Operational Environmental Satellite (GOES) instruments it is shown that 1.\\ the waiting-time distribution of flares is consistent with a time-dependent Poisson process, and 2.\\ the fraction of time the Sun spends with different flaring rates approximately follows an exponential distribution. The second result is a new phenomenological law for flares. It is shown analytically how the observed power-law behavior of the waiting times originates in the exponential distribution of flaring rates. These results are argued to be consistent with a non-stationary avalanche model for flares. ", "introduction": "\\noindent The distribution of times between flares (``waiting times'') gives information about whether flares occur as independent events, and also provides a test for models for flare statistics. For example, the avalanche model for flares (\\cite{lu&ham91}, \\cite{lu&93}) is a model designed to reproduce the observed power-law distributions of flare energy and duration. Flares are described as redistribution events in a cellular automaton (CA) that is driven to a self-organized critical state. Because the system is driven at a constant (mean) rate and flares occur as independent events, the model makes the specific prediction that the flare waiting-time distribution (WTD) is a simple exponential, consistent with a Poisson process. Observational determinations of the flare WTD have given varying results. Determinations based on hard X-ray observations have focused on the distribution of short waiting times (seconds -- hours). Biesecker (1994) found the WTD for hard X-ray bursts observed by the Burst and Transient Source Experiment on the Compton Gamma Ray Observatory to be consistent with a time-dependent Poisson process, i.e.\\ one in which the mean flaring rate is time-varying. This result is consistent with a non-stationary avalanche model for flares (an avalanche model driven with a non-constant rate). However, Wheatland et al.\\ (1998) found an overabundance of short waiting times (by comparison with a time-dependent Poisson process) in hard X-ray bursts observed by the Interplanetary Cometary Observer (ICE) spacecraft. (For another determination of the WTD based on hard X-ray, see \\cite{pea&93}.) Recently, the distribution of times between soft X-ray flares observed by the Geostationary Operational Environmental Satellite sensors (GOES) between 1976 and 1996 was examined by Boffeta et al.\\ (1999). The advantage of the GOES data is that it provides a long sequence of data with few gaps, and so the flare WTD can be examined for long waiting times. Boffeta et al.\\ found that the distribution follows a power law for waiting times greater than a few hours. They argued that this result is inconsistent with the avalanche model, and that the appearance of a power law suggests a turbulence model for the origin of flares. In this paper the GOES data is re-examined. It is shown that the observed, power-law like WTD is consistent with a piecewise-constant Poisson process, and hence with the non-stationary avalanche model. Further, it is shown that the time distribution of rates of the GOES flares averaged over several solar cycles is approximately exponential. This is a new phenomenological law for flaring. Finally, it is shown analytically how a piecewise-constant Poisson process with an exponential distribution of rates has a WTD that is power-law distributed for long waiting times, consistent with the observations. ", "conclusions": "\\noindent In this paper the waiting-time distribution for 25 years of GOES soft X-ray flares (of greater than C1.0 class) has been investigated. The observed WTD is found to be qualitatively consistent with a piecewise-constant Poisson process, with a time history of rates determined from the data using a Bayesian procedure. This result indicates that the GOES flares are independent, random events. There does not appear to be good evidence in the GOES events for flare sympathy, or for long-term correlations in the times of flare occurrence. The GOES WTD displays a power-law tail for long waiting times, as pointed out by Boffeta et al.\\ (1999), and confirmed here. In this paper the power-law behavior is demonstrated to originate from two basic assumptions, that are well supported by the data: 1.\\ that flare process is Poisson, and 2.\\ that the distribution of flaring rates follows an approximate exponential. Subject only to these assumptions, the theoretical WTD is equation~(\\ref{eq:result}), which reproduces the qualitative features of the observed WTD, including the power-law tail. There is some discrepancy between equation~(\\ref{eq:exp}) and the observationally determined WTD. For example, the observational determination of the power-law index of the tail of the distribution is around $-2.2$ (Boffeta et al.\\ found $-2.4\\pm 0.1$), whereas equation~(\\ref{eq:result}) predicts an index of $-3$. This difference is most likely due to the departure of the observed distribution of flaring rates from a simple exponential form, particularly for low flaring rates (which influence the behavior of the WTD for long waiting times). This paper presents the new result that the probability of flare occurrence per unit time, when averaged over the solar cycle, follows an approximate exponential distribution (see Figure~3). This is a new phenomenological law for flaring, that must be explained by any theory for the origin of flare energy. The rate of flare occurrence reflects the total rate of energy release in flaring, which must match the rate of energy supply to the system. Hence it follows that the rate at which energy is supplied to the corona for flaring also follows an exponential distribution. It is clear from Figure~2 that this new law does not hold instantaneously -- e.g.\\ at times of maxima of the cycle, there are few low flaring rates. For certain periods of time during each solar cycle the flaring rate is approximately constant. From these points it also follows that the observed flare WTD is time-dependent, and may have a different form depending upon the interval of observation. If the WTD is constructed for a short period of observation, during which time the rate of flaring is approximately constant, then the distribution will resemble an exponential. The power-law tail of the WTD appears in the GOES data taken over several solar cycles, during which time there is wide variation in the flaring rate. For shorter periods of observation the power-law form might not appear, depending on whether there is sufficient variation in the flaring rate. The time-dependence and cycle-dependence of the rate and waiting-time distributions will be investigated in more detail in future work. In this paper, waiting times between flares from all active regions present on the Sun have been considered, so that the Sun is treated as a single flaring system. Boffeta et al.\\ (1999) also considered flares in individual active regions, as identified (in the GOES catalog) from H$\\alpha$ events. The distribution for waiting times in individual active regions was found to be similar to that from all active regions. In future work the WTD in individual active regions will be considered in more detail. The results presented in this paper are consistent with the avalanche model for flares. Although avalanche cellular automata produce an exponential WTD when driven with a constant rate, if the rate of driving is varied so that the distribution of rates is exponential, then the resulting model (referred to here as a non-stationary avalanche model) should reproduce the qualitative features of the observed WTD. There is no need to consider models that produce a power-law WTD through long-term correlations between events (e.g.\\ models of MHD turbulence, cf.\\ Boffeta et al.\\ 1999), because the WTD is seen to be a simple consequence of the statistics of independent flare events together with an exponential distribution of flaring rates. The author acknowledges the support of a U2000 Post-doctoral Fellowship at the University of Sydney." }, "0005/astro-ph0005317_arXiv.txt": { "abstract": "We present sensitive (T$_R^*\\ \\approx\\ $0.1K), large-scale (47$^{\\prime}$ $\\times$ 7$^{\\prime}$--corresponding to 4 pc $\\times$ 0.6 pc at the source) maps of the CO J=1$\\rightarrow$0 emission of the L1448 dark cloud at 55$^{\\prime\\prime}$ resolution. These maps were acquired using the On-The-Fly (OTF) capability of the NRAO 12-meter telescope atop Kitt Peak in Arizona. CO outflow activity is seen in L1448 on parsec-scales for the first time. Careful comparison of the spatial and velocity distribution of our high-velocity CO maps with previously published optical and near-infrared images and spectra has led to the identification of six distinct CO outflows. Three of these are powered by the Class 0 protostars, L1448C, L1448N(A), and L1448N(B). L1448 IRS 2 is the source of two more outflows, one of which is newly identified from our maps. The sixth newly discovered outflow is powered by an as yet unidentified source outside of our map boundaries. We show the direct link between the heretofore unknown, giant, highly-collimated, protostellar molecular outflows and their previously discovered, distant optical manifestations. The outflows traced by our CO mapping generally reach the projected cloud boundaries. Integrated intensity maps over narrow velocity intervals indicate there is significant overlap of blue- and redshifted gas, suggesting the outflows are highly inclined with respect to the line-of-sight, although the individual outflow position angles are significantly different. The velocity channel maps also show that the outflows dominate the CO line cores as well as the high-velocity wings. The magnitude of the combined flow momenta, as well as the combined kinetic energy of the flows, are sufficient to disperse the 50 M$_{\\odot}$ NH$_3$ cores in which the protostars are currently forming, although some question remains as to the exact processes involved in redirecting the directionality of the outflow momenta to effect the complete dispersal of the parent cloud. ", "introduction": "It has long been an open question whether young stars could be the agents of dispersal of their parent molecular clouds through the combined effects of their outflows \\citep{nor79,ber96}. The answer to this question depends on whether the outflows have the requisite kinetic energy to overcome the gravitational binding energy of the cloud, as well as the efficiency with which outflows can transfer momentum, in both magnitude and direction, to the surrounding cloud. For considerations of molecular cloud dispersal, addressing the question of the adequacy of outflow momenta has historically lagged behind determinations of outflow energetics. This is because evaluation of the available energy sources needed to account for the observed spectral linewidths in a cloud is adequate for quantitative estimates of outflow energies. However, in order to address whether the requisite momentum for cloud dispersal exists in a given case requires well-sampled, sensitive, large-scale mapping of sufficiently large areas to encompass entire molecular clouds. Such observing capability has been beyond reach until the last few years, with the implementation of ``rapid'' or ``On-The-Fly'' mapping capabilities at large-aperture millimeter telescopes. The fact that many outflows powered by young stellar objects actually extend well beyond their parent molecular cloud boundaries has been recognized only recently, with the advent of large-scale, narrowband optical imaging surveys that have revealed shock-excited Herbig-Haro objects at parsec-scale separations from their exciting sources \\citep{ba96a,ba96b,bal97,dev97,eis97,wil97,gom97,gom98,rei98} and from equally large-area, sensitive, millimeter line maps that show parsec-scale molecular outflows \\citep{den95,lad96,ben96,ben98,oli99}. The millimeter line maps of parsec-scale flows have been almost exclusively confined to instances of single, well-isolated cases, due to the tremendous confusion of multiple outflows in regions of clustered star formation, such as are found in NGC 1333 (Sandell \\& Knee 1998; Knee \\& Sandell 2000), $\\rho$ Oph, Serpens \\citep{whi95}, or Circinus \\citep{bal99}. The L1448 dark cloud, with a mass of 100 M$_{\\odot}$ over its $\\sim$ 1.3 pc $\\times$ 0.7 pc extent as traced by C$^{18}$O emission, \\citep{ba86a}, is part of the much more extensive (10 pc $\\times$ 33 pc) Perseus molecular cloud complex, which contains $\\approx$ 1.7 $\\times$ 10$^4$ M$_{\\odot}$, at a distance of 300 pc \\citep{ba86b}. The two dense ammonia cores within L1448 contain 50 M$_{\\odot}$ distributed over a 1 pc $\\times$ 0.5 pc area \\citep{ba86a,ang89}. The core at V$_{LSR}$ = 4.2 km s$^{-1}$ contains the Class 0 protostar L1448 IRS 2, while the other core, at V$_{LSR}$ = 4.7 km s$^{-1}$, harbors four Class 0 protostars: L1448C, L1448N(A), L1448N(B), and L1448NW. \\citep{bar98,oli99,eis00}. The Class I source, L1448 IRS 1, lies close to the western boundary of the cloud, just outside the lowest NH$_3$ contours in the maps of \\citep{ba86b}. High-velocity molecular gas in L1448 was discovered a decade ago via CO J$=$2$\\rightarrow$1 and CO J$=$1$\\rightarrow$0 mapping of a $\\sim$ 2$^{\\prime}$ $\\times$ 6$^{\\prime}$ area centered on L1448C, acquired with 12$^{\\prime\\prime}$ and 20$^{\\prime\\prime}$ angular resolutions, respectively \\citep{bac90}. Due to its brightness, high-velocity extent ($\\pm$ 70 km s$^{-1}$), and symmetrically spaced CO bullets, the L1448C molecular outflow has been the object of much study, unlike the flows from its neighbors, the 7$^{\\prime\\prime}$ (in projected separation) protobinary, L1448N(A) \\& (B), just 1.2$^{\\prime}$ to the north, or L1448 IRS 2, 3.7$^{\\prime}$ to the northwest (e.g., \\citep{cur90,gui92,bal93,bac94,dav94,bac95,dut97}). Although outflow activity in the vicinity of the protobinary had been reported previously, the H$_2$ and CO flows, driven by L1448N(A) and L1448N(B), respectively \\citep{bac90,dav95}, were not recognized as distinct until recently \\citep{bar98}. Identification of these flows was aided by noting the position angle of the low-excitation H$_2$ flow, centered on L1448N(A), to be distinct from the position angle of the CO flow from L1448N(B), defined by the direction of the line joining L1448N(B) with the newly discovered Herbig-Haro object, HH 196 \\citep{bal97}. Recent, wide-angle ($\\sim$ 70$^{\\prime}$ field-of-view), narrowband optical imaging of the entire extent of the L1448 cloud has resulted in the discovery of several systems of Herbig-Haro objects, some displaced several parsecs from any exciting source \\citep{bal97}. In order to investigate the link between high-velocity molecular gas and the newly discovered Herbig-Haro objects, as well as to study the possibility of cloud dispersal via outflows, we acquired new, sensitive, large-scale CO J$=$1$\\rightarrow$0 maps of a substantial portion of the L1448 cloud. These new molecular line maps were acquired with the On-The-Fly (OTF) mapping technique as implemented at NRAO's 12-meter millimeter telescope atop Kitt Peak, Arizona. \\section {Observations and Data Reduction} The CO J$=$1$\\rightarrow$0 maps of L1448 presented in this paper were acquired using the spectral-line On-The-Fly (OTF) mapping mode of the NRAO's\\footnote{The National Radio Astronomy Observatory is a facility of the National Science Foundation, operated under cooperative agreement by Associated Universities, Inc.} 12-meter telescope on 23 June 1997, UT 13$^h$53$^m$ $-$ UT 19$^h$25$^m$. We stress that the OTF technique allows the acquisition of large-area, high-sensitivity, spectral line maps with unprecedented speed and pointing accuracy. For comparison, it would have taken eight times the amount of telescope time, or nearly a week in practice, to acquire this same map using conventional, point-by-point mapping. Although OTF mapping is not a new concept, given the rigor of the position encoding that allows precise and accurate gridding of the data, the fast data recording rates that allow rapid scanning without beam smearing, and the analysis tools that are available, the 12-meter implementation is the most ambitious effort at OTF imaging yet. To produce our CO maps of L1448, we observed a 47$^{\\prime}$ $\\times$ 7$^{\\prime}$ field along a position angle P.A. $=$ 135$^{\\circ}$, (measured East from North), centered on the coordinates of L1448 IRS 2 ($\\alpha_{1950}=$ 3$^h$ 22$^m$ 17.9$^s$, $\\delta_{1950}=$ 30$^{\\circ}$ 34$^{\\prime}$ 41$^{\\prime\\prime}$). The 115 GHz beamwidth was $\\approx$ 55$^{\\prime\\prime}$. We scanned a total of 33 rows at a rate of 50$^{\\prime\\prime}$/s, along P.A. $=$ 135$^{\\circ}$, with a row spacing of 12.7$^{\\prime\\prime}$. (The row spacing is determined by the optimum spatial sampling and by the scanning position angle.) We calibrated and integrated on an absolute off position ($\\alpha_{1950}=$ 3$^h$ 20$^m$ 00.0$^s$, $\\delta_{1950}=$ 31$^{\\circ}$ 00$^{\\prime}$ 00$^{\\prime\\prime}$) at the start of every row. Each row took approximately one minute to scan. Each map coverage took 41 $-$ 56 minutes to complete. We performed six map coverages to attain an RMS in each OTF spectrum of T$_R^*$ $\\approx$ 0.11 K. A dual-channel, single-sideband SIS receiver was used for all observations. The backend consisted of 250 kHz and 500 kHz resolution filterbanks, yielding velocity resolutions of 0.65 km s$^{-1}$ and 1.3 km s$^{-1}$, respectively. The filterbanks were used in parallel mode, each of the two receiver polarization channels using 256 filterbank channels. The polarization channels were subsequently averaged together to improve signal-to-noise. Only the 250 kHz resolution data were used to produce the maps presented here. Line temperatures at the 12-meter are on the T$_R^*$ scale, and must be divided by the corrected main beam efficiency, $\\eta_m^*$, to convert to the main-beam brightness temperature scale. For our very extended source, $\\eta_m^*$ is approximately 1.0. Since the corrected main-beam efficiency is the fraction of the forward power in the main diffraction beam relative to the total forward power in the main beam plus error beam, contribution from the error beam can make $\\eta_m^*$ $>$ 1.0. At 115 GHz, the theoretical error beam width is $\\approx$ 17$^{\\prime}$, but the ratio of the error beam amplitude to the main beam amplitude is only 6 $\\times$ 10$^{-4}$, suggesting contribution from the error beam can be ignored. We used the NRAO standard source, B5 ($\\alpha_{1950}=$ 3$^h$ 44$^m$ 52.4$^s$, $\\delta_{1950}=$ 32$^{\\circ}$ 44$^{\\prime}$ 28$^{\\prime\\prime}$), to check absolute line temperatures. The OTF data were reduced with the Astronomical Image Processing Software (AIPS), Version 15JUL95. AIPS tasks specific to OTF data are `OTFUV', which converts a single 12-meter OTF map (in UniPOPS SDD format) to UV (single-dish) format, and `SDGRD', which selects random position single-dish data in AIPS UV format in a specified field of view about a specified position and projects the coordinates onto the specified coordinate system. The data are then convolved onto a grid. OTF data maps were first combined, then gridded into a data cube and baseline-subtracted. Channel maps as well as individual spectra were inspected to ensure good baseline removal and to check for scanning artifacts. Only the first and last rows scanned contained corrupted spectra and were rejected. ", "conclusions": "\\subsection{Interpretation of Outflow Structure} Figures 7$-$9 present the various outflow extents and position angles. Figure 7 is relevant for the discussion of alternative interpretations of the outflow emission centered on the L1448 IRS 2 ammonia core. Figure 8 is used for the discussion of the outflows originating from sources embedded in the ammonia core associated with L1448C and L1448N. For both Figures 7 and 8, the outflow axes and extents are superposed on our integrated high-velocity CO linewing map of the L1448 cloud. Figure 9 shows the same outflow axes superposed on a much higher spatial-resolution ($\\sim$ 1$^{\\prime\\prime}$ vs. $\\sim$ 55$^{\\prime\\prime}$) H$_2$ image of the L1448 region from Eisl\\\"offel (2000). (This is the only figure using J2000 coordinates.) In all cases (except for the two newly-identified outflow features seen in our CO maps associated with L1448 IRS 2 and the unidentified source outside our map boundaries), outflow position angles were determined from previously-published, arcsecond-scale outflow data. For all of the outflows, we find good agreement between large-scale CO features seen in our maps and flow axes that have been determined from the previous, higher-resolution observations. In Figures 7$-$9, {\\it solid}, colored lines denote well-established flow position angles and extents, derived from our own CO data and the published literature, whereas {\\it dashed} lines denote outflow position angles and extents that are consistent with our new CO data. \\subsubsection{Outflow Emission from the L1448 IRS 2 Ammonia Core} High-velocity CO outflow activity centered on L1448 IRS 2 was discovered from low-spatial resolution ($\\approx$ 55$^{\\prime\\prime}$) mapping (O'Linger et al. 1999). These authors suggested that L1448 IRS 2 was the source of a single outflow, with a constant opening angle, as depicted in Figure 7a. The previous outflow symmetry axis along P.A. $\\sim$ 133$^{\\circ}$ and opening angle, $\\phi \\sim 27^{\\circ}$, were derived from the high spatial resolution K$^{\\prime}$ images of Hodapp (1994). We derive a new outflow symmetry axis along P.A. $\\sim$ 138$^{\\circ}$ from the more recent H$_2$ images of Eisl\\\"offel (2000). Therefore, we have been able to determine more accurately the position angles for the proposed outflow cavity walls, which should lie along P.A. $\\sim$ 152$^{\\circ}$ and P.A. $\\sim$ 125$^{\\circ}$, if the IRS 2 outflow retains its initial opening angle out to large distances. This model explains the presence of high-velocity blue- and redshifted CO emission seen along these position angles at large distances from IRS 2, notably the V-shaped morphology of the blueshifted gas and the presence of several CO ``bullets'' located along the proposed outflow axis, well beyond HH 195A$-$D. In a constant opening angle outflow scenario, HH 193 lies along one arm of this V, at P.A. $\\sim$ 152$^{\\circ}$. The blueshifted gas along the other arm of the V (P.A. $\\sim$ 125$^{\\circ}$) would be confused with emission from the E-W outflow in this vicinity, but since the blueshifted emission extends {\\it past} IRS 1 towards the {\\it redshifted} HH 194 knots, at least part of this emission could be due to IRS 2. Although there is redshifted CO emission along P.A. $\\sim$ 138$^{\\circ}$ and along P.A. $\\sim$ 152$^{\\circ}$, there is little evidence for redshifted emission along P.A. $\\sim$ 125$^{\\circ}$. However, this may be due to confusion with the three redshifted lobes associated with L1448C, L1448N(A), \\& L1448N(B). Figure 9 clearly shows an outflow associated with IRS 2 along a P.A. $\\sim$ 138$^{\\circ}$. The redshifted gas along the outflow axis is prominent in the H$_2$ emission, but is not clearly apparent in our CO maps beyond the redshifted peak about 1$^{\\prime}$ southeast of IRS 2. Hints of more extended emission along P.A. $\\sim$ 138$^{\\circ}$ may be seen, however, in Figures 6b \\& c, and curiously, in {\\it blueshifted} emission extended along this axis to the southeast of IRS 2 in Figure 5d. Such overlap of blueshifted gas along the redshifted outflow axis is expected for outflows oriented nearly in the plane of the sky. Indications of extended blueshifted emission along P.A. $\\sim$ 138$^{\\circ}$, at least a few arcminutes downstream of HH 195A$-$D, are seen in Figures 5b \\& c. However, there are a few problems with the single outflow, constant opening angle model: (1) Figure 9 indicates that although the initial opening angle of the IRS 2 outflow is $\\phi \\sim 27^{\\circ}$, the two strands of H$_2$ defining this opening angle join in a bow shock structure in the vicinity of HH 195A-D; (2) The CO data show little evidence for emission along the cavity wall at P.A. $\\sim$ 125$^{\\circ}$, although we note that there is much confusion from high-velocity gas associated with other outflows along this position angle to both sides of IRS 2; (3) HH 193 lies precisely at the end of the outflow wall in this model, an unlikely location for a shock; (4) The velocity dispersion of the blueshifted feature along P.A. $\\sim$ 152$^{\\circ}$ is high, since it is prominent at both ambient cloud velocities and at highly blueshifted velocities (Figures 5, 6d\\&e); (5) The highest-velocity outflow emission should converge towards the outflow axis (P.A. $\\sim$ 138$^{\\circ}$) due to projection effects, but the highest-velocity outflow emission lies along P.A. $\\sim$ 152$^{\\circ}$, the proposed outflow wall. Our maps are not sensitive enough to have picked up the highest-velocity outflow emission, however, which is severely diluted in our large ($\\sim$1$^{\\prime}$) beam. Nevertheless, the absence of CO emission along P.A. $\\sim$ 138$^{\\circ}$ in the highest velocities suggests that the feature along P.A. $\\sim$ 152$^{\\circ}$ defines a separate outflow axis. This has led to an alternate interpretation of the high-velocity CO associated with L1448 IRS 2 in which the presence of {\\it two} outflows is required, as depicted in Figure 7b, with one outflow along P.A. $\\sim$ 138$^{\\circ}$, and a new, second outflow along P.A. $\\sim$ 152$^{\\circ}$, so prominent in the CO data. In this scenario, the new IRS 2 outflow would be responsible for exciting HH 193. Two outflows, along distinctly different position angles, would also suggest that IRS 2 is a binary system. Although there is currently no evidence to indicate this source to be binary from available continuum data obtained with the Submillimetre Common User Bolometer Array (SCUBA) at the JCMT on Mauna Kea, Hawaii (O'Linger et al. 1999), it is possible that IRS 2 is a compact binary on a scale smaller than 7$^{\\prime\\prime}$ (the resolution of the SCUBA 450 $\\mu$m data). Recent work indicates a high incidence of binarity among young stellar systems (eg., Ghez, Neugebauer, \\& Matthews 1993; Looney, Mundy, \\& Welch 2000). Only arcsec/sub-arcsec imaging at either millimeter or centimeter wavelengths could test the binary hypothesis further. Evidence of other outflow activity near L1448 IRS 2 is found by noting that the very confined outflow, whose lobes peak only $\\sim$1$^{\\prime}$ on either side of IRS 2, has its redshifted peak well-aligned along the P.A. $=$ 138$^{\\circ}$ outflow that excites HH 195A-D, but the corresponding blue peak closest to IRS 2 is skewed at a somewhat shallower position angle closer to P.A. $\\sim$ 125$^{\\circ}$. No CO emission (Figure 2, Figures 6a$-$c) is seen along P.A. $\\sim$ 125$^{\\circ}$ on the opposite side from IRS 2, which would be expected if there were an outflow along this direction. The blue peak is spatially coincident with HH 195E, however, the only HH 195 knot which is off the P.A. $=$ 138$^{\\circ}$ axis of the IRS 2 outflow. It has been argued that IRS 1 drives an east-west oriented outflow and is the most probable driving source of HH 195E (Bally et al. 1997; Eisl\\\"offel 2000). Although the presence of such an E-W oriented outflow in this region is undisputed, it is possible that an as yet undiscovered source, other than IRS 1, may be the responsible agent. Thus, the positioning of the blue peak so close to IRS 2 may be coincidental, and due primarily to local heating associated with HH 195 E, and overlapping outflows associated with blueshifted emission from IRS 2 and IRS 1. This picture is supported by the L1448 H$_2$ mosaic of Eisl\\\"offel (2000), shown in Figure 9, which shows the H$_2$ emission in the vicinity of HH 195E pointing toward IRS 1, not IRS 2. \\subsubsection{Outflow Emission from the L1448C/L1448N Ammonia Core } The position angles of the blue- and red-shifted outflowing gas powered by L1448C are indicated by the green lines in both Figures 8 \\& 9. Solid green lines indicate the L1448C outflow's direction and extent as determined by previous workers (see the caption of Figure 8 for references). The dashed green line on the blue-shifted side represents the continuation of the L1448C outflow proposed by Barsony et al. (1998). The blueshifted L1448C outflow suffers a large deflection from its original direction, as can be seen clearly in Figure 9, which shows the blueshifted L1448C outflow axis passing directly through H$_2$ emission knots I \\& S, and about 20$^{\\prime\\prime}$ north of emission knot R. Emission knot R lies within more extended H$_2$ emission that appears to form a U-shaped or bow-shock structure which opens toward the southeast, bisected by the L1448C outflow axis. The sides of the U are separated by $\\sim$ 40$^{\\prime\\prime}$, with the brighter side (including knot R) lying to the south of the L1448C outflow axis. This final, deflected, blue-shifted outflow axis lies along P.A. $\\sim$ 127$^{\\circ}$, where a long finger of high-velocity blueshifted CO emission is found. Blueshifted CO emission surrounds the HH 267 complex in a horseshoe shape about the proposed extension of the L1448C outflow, suggesting the outflow may be responsible for this emission. The dashed green line on the redshifted side in Figure 8 indicates the possible continuation of the L1448C outflow to the south. The L1448N(A) molecular outflow axis and extent are depicted by the purple line in Figures 8 \\& 9. Only redshifted molecular gas associated with the L1448N(A) outflow, along P.A. $\\sim$ 150$^{\\circ}$, is detected in our CO maps, with a total length of at least 0.7 pc, as seen in Figure 8. The U-shaped molecular hydrogen emission that traces part of the redshifted outflow cavity wall from the L1448N(A) outflow, as seen in Figure 9, is unresolved in our CO maps. Knots of H$_2$ emission which trace the outflow wall are $\\le$30$^{\\prime\\prime}$ apart as far as 3$^{\\prime}$ to the southeast of L1448N(A) (Barsony et al. 1998; Eisl\\\"offel 2000). Taken together with the observed length of the CO redshifted lobe, this suggests a collimation factor of at least 16:1. The opening half-angle of the L1448N(A) outflow was estimated to be $\\phi/2\\ \\approx\\ 25^{\\circ}$, from the morphology of the near-infrared reflection nebulosity to the north of L1448N(A), associated with what would be the blueshifted outflow lobe (Bally et al. 1993). The blueshifted flow powered by L1448N(A) is not apparent in our CO maps, judging by the drop in the blueshifted, high-velocity CO contour levels along the symmetry axis of its NIR reflection nebula. The lack of blueshifted CO emission from the L1448N(A) outflow is most likely accounted for by the likelihood that the cloud boundary has been reached in this direction, and that the flow has broken out of the molecular cloud. The L1448N(B) molecular outflow axis and extent are depicted by the mustard-colored lines in Figures 8 \\& 9. The position angle, P.A. $\\sim$ 129$^{\\circ}$, of the L1448N(B) outflow was determined by the orientation of the redshifted CO outflow driven by L1448N(B) from Figure 1b and noting that this CO flow symmetry axis intersects HH 196 (Barsony et al. 1998). The true spatial extent of the L1448N(B) CO outflow, however, is demonstrated here for the first time. On the scale of our map, the L1448N(B) outflow remains unresolved along its width. The optical emission knots of HH 196 are, indeed, found to lie right along the L1448N(B) CO outflow axis, confirming the identification of L1448N(B) as their exciting source (Figure 8). Approximately 2$^{\\prime}$ downstream of HH 196, the L1448N(B) flow becomes confused in projection with the P.A. $\\sim$ 152$^{\\circ}$ outflow from IRS 2. If the outflow continues along P.A. $\\sim$ 129$^{\\circ}$, it could account for the bulges in the blueshifted emission to the south and southwest of HH 193 (Figure 2, Figure 5c, Figure 8), and might even be responsible for the ``C''--shaped blueshifted emission structure east of the HH 267 system, $\\sim$2.5 parsecs from L1448N(B). The redshifted lobe associated with this source appears to terminate $\\sim$2$^{\\prime}$ northwest of HH 277, although this could be due to confusion with the high-velocity CO emission surrounding HH 277, which seems to be part of an outflow driven by an unidentified source outside of the boundaries of our map. Using the most conservative length for the L1448N(B) outflow, the major axis taken from the HH 196 knots through the end of the redshifted lobe ($\\sim$12$^{\\prime}$), the derived lower limit for the collimation factor is $\\ge$12:1. Estimates of the initial opening angle and width of the L1448N(B) outflow await higher spatial resolution, interferometric imaging. A dark blue dashed line in Figure 8 indicates a possible outflow axis passing through an otherwise unexplained, high-velocity redshifted feature associated with HH 277, in the southeast quadrant of our CO map. This redshifted CO velocity feature is most prominently seen in Figure 6c. The orientation of this structure is almost perpendicular to the general orientation of our map. Thus, HH 277 is probably driven by a source off the edge of our map. Finally, although the origin of the HH 267 knots cannot definitively be resolved based on the CO data we present here, our observations do constrain the driving source. L1448N(A) \\& (B) can be ruled out as possible driving sources of HH 267, since the P.A.'s of their associated outflows are along completely different directions than the lines linking them with HH 267. Furthermore, blue-shifted molecular gas has yet to be detected from L1448N(A). Terminal velocities for the L1448 IRS 2 outflows can not be determined from our data, but the P.A.'s of both of these outflows also miss the HH 267 complex completely. However, the P.A. of the deflected blueshifted lobe of the L1448C flow goes right through HH 267, and the reported terminal velocity for this outflow (70 km s$^{-1}$: Bachiller et al. 1990, 1995) is in good agreement with the measured HH 267 velocities (Bally et al. 1997), as suggested by Barsony et al. (1998). \\subsection{Cloud Dispersal by Giant Protostellar Flows?} The most dramatic evidence for the direct effects of the outflows on the L1448 molecular cloud is seen in the distortions of the cloud contours at all velocities in Figures 5 \\& 6. Although we cannot estimate the masses and energetics of each individual outflow in our maps due to confusion in space and velocity, we can, nevertheless, estimate the {\\it total} contribution of the outflows to the cloud's energetics. The Local Thermodynamic Equilibrium (LTE) analysis used to estimate the combined mass of the outflows (M$_{tot}\\approx 0.7$ M$_{\\odot}$) is discussed in O'Linger et al. (1999). Optically thin high-velocity CO emission was assumed, given the lack of observed high-velocity $^{13}$CO emission in the velocity intervals outside the cloud core velocities (see Figure 3). Therefore, the resultant derived mass is a strict lower limit, since no attempt was made to correct for the considerable mass expected to be masked by the line core emission. For highly inclined outflows ($i>70^{\\circ}$), the characteristic velocity which is used to calculate outflow energetic parameters is best chosen as the geometrical mean between the highest observed velocities, $V_{CO}$, and the inclination-corrected velocity, $V_{CO}/cos(i-{{\\phi}\\over {2}})$, where ${\\phi}\\over {2}$ is the half-opening angle of the outflow (Cabrit \\& Bertout 1992). Assuming the outflow inclinations, 70$^{\\circ}$ $\\le$ $i$ $\\le$ 90$^{\\circ}$, V$_{char} = 22 - 34$ km s$^{-1}$ for the L1448 outflows, the total momentum in all the flows is computed to be 16 M$_{\\odot}$ km s$^{-1}$ $\\le$ M$_{tot}$V$_{char}$ $\\le$ 24 M$_{\\odot}$ km s$^{-1}$. This range of values is nearly equivalent to the momentum content of the quiescent NH$_3$ cores, assuming 50 M$_{\\odot}$ total cores and $v_{turb}$ $\\sim$ 0.5 km s$^{-1}$ \\citep{ba86a}. Even more striking, the total kinetic energy in all the flows, 2$\\times 10^{45}$ ergs $\\le$ $1\\over {2}$M$_{tot}$V$_{char}^2$ $\\le$ 8$\\times 10^{45}$ ergs, exceeds the gravitational binding energy ($\\sim\\ GM^2/R\\ \\approx$ 5 $\\times$ 10$^{44}$ ergs) of the NH$_3$ cores by an order of magnitude, and the gravitational binding energy of the 100 M$_{\\odot}$ C$^{18}$O cloud (9 $\\times$ 10$^{44}$ ergs), contained within a 1.3 pc $\\times$ 0.7 pc region \\citep{ba86b}, by a factor of five. The total outflow momentum quoted above is, in fact, a lower limit, since these outflows are still gaining momentum from the force provided by the central driving engines, and the total outflow mass may be grossly underestimated. The magnitude of both the total energy and momenta of the outflows suggests these outflows are capable of dispersing the NH$_3$ cores, with the caveat that it is unclear, both from the outflow and ambient cloud morphology, how the outflow momenta can be adequately transferred to the surrounding core. Possibly, this can be accomplished as the individual outflow opening angles increase with time." }, "0005/astro-ph0005303_arXiv.txt": { "abstract": "We present 1 micron \\textit{Hubble Space Telescope\\/}/NICMOS resolved imaging polarimetry of the \\hbox{GG Tau} circumbinary ring. We find that the ring displays east-west asymmetries in surface brightness as well as several pronounced irregularities, but is smoother than suggested by ground-based adaptive optics observations. The data are consistent with a $37\\arcdeg$ system inclination and a projected rotational axis at a position angle of $7\\arcdeg$ east of north, determined from millimeter imaging. The ring is strongly polarized, up to $\\sim50$\\%, which is indicative of Rayleigh-like scattering from sub-micron dust grains. Although the polarization pattern is broadly centrosymmetric and clearly results from illumination of the ring by the central stars, departures from true centrosymmetry and the irregular flux suggest that binary illumination, scattering through unresolved circum\\-\\textit{stellar\\/} disks, and shading by these disks, may all be factors influencing the observed morphology. We confirm a $\\sim0\\farcs25$ shift between the inner edges of the NIR and millimeter images and find that the global morphology of the ring and the polarimetry provide strong evidence for a geometrically thick ring. A simple Monte Carlo scattering simulation is presented which reproduces these features and supports the thick ring hypothesis. We cannot confirm filamentary streaming from the binary to the ring, also observed in the ground-based images, although it is possible that there is material inside the dynamically cleared region which might contribute to filamentary deconvolution artifacts. Finally, we find a faint 5th point source in the \\hbox{GG Tau} field which, if it is associated with the system, is almost certainly a brown dwarf. ", "introduction": "\\label{sec:intro} GG Tau is one of the most studied systems known to possess a tidally truncated circumbinary disk. It is a double binary, in which all four components display evidence of infrared excesses \\citep{white99} indicating the presence of unresolved circum\\-\\textit{stellar\\/} disks. However, the most striking object in the system is the large disk of gas and dust surrounding the northern \\hbox{($\\sim0\\farcs25$)} binary. Since the circumbinary structure was first detected \\citep{beckwith90,sg92,kawabe93}, many of its physical parameters have been constrained, particularly from the millimeter interferometry of \\citet{dgs94} and \\citet[hereafter GDS]{gds99}. The disk is tidally truncated by the dynamical action of the binary at a radius of \\hbox{$\\sim180$AU} from the binary center of mass and extends out to at least 800AU. From their most recent millimeter continuum and CO line flux measurements, GDS propose that 90\\% of the dust component of the circumbinary disk lies in a well-defined \\textit{ring\\/} (inner radius \\hbox{$\\sim180$AU;} outer radius \\hbox{$\\sim260$AU)} within the extended (800AU) disk. Their observations are consistent with an almost circular Keplerian disk, rotating about the binary's center of mass with a projected rotational axis at a position angle (PA) of $7\\arcdeg$ east of north, and viewed at an inclination of $37\\arcdeg$ to face-on. This implies that the northern part of the ring is closest to us (the `front') and that the southern part of the ring corresponds to the `back'. The first near-infrared (NIR) observations of the ring were obtained with ground-based adaptive optics (AO) by \\citet[hereafter RRNGJ]{roddier96} who published deconvolved $J$\\/, $H$\\/, and $K$\\/ band images of the ring. Their data show a very clumpy ring and suggest a large degree of anisotropy in its illumination. The images also show radial filaments \\textit{inside\\/} the dynamically cleared cavity, extending from the binary to ring. At the $J$\\/ and $K$\\/ bands, the ring appears incomplete (or at least \\textit{very\\/} faint) at the back. GDS found that when they registered their 1.4mm and RRNGJ's $J\\/$ band image, there is a $\\sim$0\\farcs25 shift of the ring's NIR inner edge towards the binary. This could be explained if the dust ring is \\textit{geometrically thick\\/}, extending to a height of $\\sim$120AU at its tidally truncated inner radius. In this paper, we present the first resolved imaging polarimetry of the \\hbox{GG Tau} system and the first space-based images of the circumbinary ring. We refer to the northern binary as \\hbox{`GG Tau'} and follow GDS in assuming a distance of 140AU to the Taurus star-forming region \\citep{elias78}. ", "conclusions": "" }, "0005/astro-ph0005135_arXiv.txt": { "abstract": "Narrow-Line Seyfert 1s (NLS1s) are generally considered to be ``strange'' Active Galactic Nuclei (AGNs). Surprisingly, this makes them very useful for constraining models. I discuss what happens when one attempts to qualitatively fit the NLS1 phenomenon using the stellar wind model for AGN line emission (e.g., Kazanas 1989). The simplest way of narrowing profile bases of this model to the widths observed in NLS1s is probably to lower the mass of the supermassive black hole. In a flux-limited and redshift-limited data set, this is indeed similar to increasing \\( L/L_{\\rm Edd} \\). Because the broad line region (BLR) of the stellar line emission model scales with the tidal radius of the stars, this model predicts maximal BLR velocities of \\( {\\rm FWZI}\\propto (L/L_{\\rm Edd})^{-1/3} \\). This implies that the black holes of NLS1s are approximately \\( 3^{3}=27 \\) times less massive than those in other Seyfert 1s if the stellar line emission model is correct. Another consequence of increasing \\( L/L_{\\rm Edd} \\) in this model is that it results in an increase in the wind edge densities. NLS1 spectra appear to support this result as well. Even the collateral features of NLS1s, such as the line asymmetries and continuum properties, appear to be easily explained within the context of this model. For better or worse, if the stellar wind line emission is correct, NLS1s are not much stranger than other AGNs. ", "introduction": " ", "conclusions": "" }, "0005/astro-ph0005245_arXiv.txt": { "abstract": "The recent finding by Chevalier \\& Ilovaisky (1998) from {\\it Hipparcos} observations that OB-supergiant X-ray binaries have relatively large runaway velocities (mean peculiar tangential velocity\\footnote{The values given here are not identical (though similar) to those listed in Chevalier \\& Ilovaisky (1998). The corrections we applied are outlined below.} $\\avvt = 42 \\pm 14$~\\kms), whereas Be/X-ray binaries have low runaway velocities ($\\avvt = 15 \\pm 6$\\,\\kms), provides confirmation of the current models for the formation of these two types of systems. These predict a difference in runaway velocity of this order of magnitude. This difference basically results from the variation of the fractional helium core mass as a function of stellar mass, in combination with the conservation of orbital angular momentum during the mass transfer phase that preceded the formation of the compact object in the system. This combination results into: (i) Systematically narrower pre-supernova orbits in the OB-supergiant systems than in the Be-systems, and (ii) A larger fractional amount of mass ejected in the supernovae in high-mass systems relative to systems of lower mass. Regardless of possible kick velocities imparted to neutron stars at birth, this combination leads to a considerable difference in average runaway velocity between these two groups. If one includes the possibility for non-conservative mass transfer the predicted difference between the runaway velocity of the two groups becomes even more pronounced. The observed low runaway velocities of the Be/X-ray binaries confirm that in most cases not more than 1 to 2\\,\\msun\\ was ejected in the supernovae that produced their neutron stars. This, in combination with the --on average-- large orbital eccentricities of these systems, indicates that their neutron stars must have received a velocity kick in the range 60 - 250 \\,\\kms\\ at birth. The considerable runaway velocity of Cygnus X-1 ($\\vt = 50\\pm15$\\, \\kms) shows that also with the formation of a black hole considerable mass ejection takes place. ", "introduction": "A high-mass X-ray binary (HMXB) consists of a massive OB-type star and a compact X-ray source, a neutron star or a black hole. The X-ray source is powered by accretion of wind material, though in some systems mass transfer takes place through Roche-lobe overflow; the compact stars in the latter systems are surrounded by an accretion disk. Since wind accretion plays an important role, in practice only an OB supergiant or a Be-star companion have a strong enough stellar wind to result in observable X-ray emission. In a Be/X-ray binary the X-ray source is only observed when the neutron star moves through the dense Be-star disk at periatron passage. About 75\\% of the known HMXBs are Be/X-ray binaries, although this is a lower limit given their transient character. Chevalier \\& Ilovaisky (1998) derived the proper motions for a sample of HMXBs from {\\it Hipparcos} measurements. The four OB-supergiant HMXBs for which proper motions are available (0114+65, 0900-40 [Vela X-1], 1700-37 and Cyg X-1) have relatively large peculiar tangential velocities. Some corrections to the values given by these authors are needed (cf.\\ Steele et al.\\, 1998, Kaper et al. 1999). Taking these into account (Table 1) the mean peculiar velocity of these systems is $42 \\pm 14$\\,\\kms. It was already known that the OB-supergiant system of 1538-52 (QV Nor) has a peculiar radial velocity of about 90\\, \\kms\\ with respect to its local standard of rest (Crampton et al. 1978; Gies \\& Bolton 1986; van Oijen 1989). For the 13 Be/X-ray binaries with measured proper motions Chevalier \\& Ilovaisky found peculiar tangential velocities ranging from $\\vt = 3.3 \\pm 0.7$ to $21 \\pm 7.4$\\, \\kms, with an average of $\\avvt = 11.4 \\pm 6.6$\\,\\kms. Again, after corrections (see Sect.\\,2) and excluding the Oe systems X~Per (0352+309) and V725~Tau (0535+262), one finds for the genuine Be/X-ray binary a slightly higher value of $\\vt = 15 \\pm 6$\\,\\kms. We would like to point out here that these mean values are in good agreement with the runaway velocities of these two types of systems predicted on the basis of simple ``conservative'' evolutionary models (van den Heuvel 1983, 1985, 1994; Habets 1985; van den Heuvel \\& Rappaport 1987) and even better agreement is obtained when mass is not conserved in the transfer process (Portegies Zwart 2000). The effect of sudden mass loss during the supernova explosion is taken into account and in a massive binary this is the dominant contribution to the runaway velocity; a random kick velocity of a few hundred \\kms\\ imparted to the neutron star at birth (see e.g.\\ Hartman 1997) has only a small effect, as the kick's impulse has to be distributed over the entire massive ($\\apgt 15$\\,\\msun) system. (See Portegies Zwart \\& van den Heuvel 1999, for arguments in favor of kicks). Therefore, in first-order approximation, these kicks can be neglected in calculating the runaway velocities of HMXBs, but {\\em not} in calculating their orbital eccentricities (see Sects.\\, 3.4 and 3.5). The aim of the present paper is to give a quantitative assessment of the above-mentioned conjectures. It should be noted here that five Be-star systems in the Be/X-ray binary sample studied by Chevalier \\& Ilovaisky (1998) are of spectral type B4~Ve or later (masses $\\leq 6 M{_\\odot}$). The companions of these stars might be white dwarfs instead of neutron stars. Therefore, a supernova explosion is not necessarily the reason for their (excess) space velocity, which, in any case, is relatively small. It may be due to the typical random velocities observed in young stellar systems. Leaving these late-type Be/X-ray binaries out does not result in a significant change in the observed mean peculiar velocity of the Be-systems. Furthermore, there is some doubt concerning the use of the distances based on {\\it Hipparcos} parallaxes of several of the other Be-systems, as these distances differ very much from the distances determined in other ways, e.g. by using reddening etc. (Steele et al. 1998). In Sect.\\, 2 we therefore critically examine the distances and proper motions of all the systems with Be companions. In Sects.\\, 3.1 and 3.2 we present an analytical calculation of the expected runaway velocities and orbital eccentricities of typical OB-supergiant and Be HMXBs, on the basis of the standard evolutionary models for these systems, adopting conservative mass transfer during phases of mass exchange, and including the effects of stellar-wind mass loss for the OB-supergiant systems. In Sect.\\ 4 we discuss the effect of non-conservative mass transfer on the runaway velocity and in Sect.\\, 5.1 for the Be/X-ray binaries with known orbital eccentricities. We calculate which kick velocities should be imparted to the neutron stars of Be/X-ray binaries in order to produce their, on average, large orbital eccentricities (since the mass-loss effects alone cannot produce these). In Sect.\\, 5.2, as an alternative, we compare the observed runaway velocities and orbital eccentricities of the Be/X-ray binaries with those expected on the basis of symmetric mass ejection and show that without kicks their combination of high orbital eccentricities and low space velocities cannot be explained. Our conclusions are summarized in Sect.\\, 6. ", "conclusions": "The measured tangential velocities of the Be/X-ray binaries and OB-supergiant X-ray binaries by the {\\it Hipparcos} satellite confirm the expectations from the evolution of massive close binaries in which little mass is lost from the binary systems during the first mass transfer phase. The much higher tangential velocities of supergiant X-ray binaries than those of the Be-systems follow from a combination of (1) the much larger fractional helium core masses in the progenitors of the OB-supergiant systems which cause their pre-supernova orbital periods to be shorter, and thus their pre-supernova orbital velocities to be much larger than those of the less massive Be-systems, and (2) the much lower amounts of mass ejected during the supernova explosion in the lower-mass Be-systems compared to the OB-supergiant systems. The combination of a high orbital eccentricity with a low space velocity observed for the Be type X-ray binaries can only be understood if a kick with appreciable velocity --in the range 60 to 250~\\kms-- is imparted to the newly born neutron star. Such a kick tends to only slightly affect the space velocity of the binary system since the neutron star has to drag along its massive companion. The orbital eccentricity, however, is strongly affected by such a asymmetric velocity kick. If the supernova explosions in these systems had been symmetric, the high orbital eccentricities observed in the class of Be X-ray binaries are impossible to reconcile with their on average low runaway velocities." }, "0005/astro-ph0005590_arXiv.txt": { "abstract": "We investigate a large number of cosmological evolution models for QSOs and Active Galactic Nuclei (AGN). We introduce a spectrum-luminosity correlation as a new input parameter and adopt the estimated mass function (MF) of massive black holes in centers of nearby galactic nuclei as a constraint to distinguish among different QSO/AGN models. We explore three basic types of phenomenological scenarios; (i) Models with multiple short-lived ($\\sim$ a few $10^{6-8}$ yrs) populations, (ii) Models with a single long-lived ($\\sim 10^9$ yr) QSO population, and (iii) Models with recurrent QSO/AGN activities which are driven by long-term variabilities of the disk instability type. In each model, we derive the expected theoretical luminosity function (LF) and the MF of black holes which grow through mass accretion. We assess the plausibility of each model based on whether each model's LF and MF are compatible with the observed data. We find that the best fits to the observed LFs are obtained in the model with multiple short-lived populations and without any significant spectral evolution. This finding suggests that the QSO populations may be composed of many short-lived generations ($\\sim$ a few 10$^8$ yrs) and that there is no significant spectral evolution within each generation. On the other hand, we also show that there is no satisfactory model which can simultaneously account for the observed LF and the estimated MF. We speculate that some of the present-day black holes (BHs) found in galactic nuclei may have formed without undergoing the QSO/AGN phase. ", "introduction": "It has been well established that the observed comoving space density of optically bright QSOs reaches a peak at a critical redshift $z\\sim 2$ (e.g. Peterson 1997, Hartwick \\& Shade 1990, Weedman 1986). A similar trend is also seen in the X-ray evolution (Miyaji et al. 2000 and references therein). In both cases, the evolution of the luminosity function (LF) is roughly accounted for by the number-conserving luminosity evolution (e.g. Mathez et al. 1976) in which the luminosities of QSOs first gradually decrease at $z>2$ from their births near $z>4$ and rapidly decline at $z<2$. It has recently been questioned whether the X-ray evolution can be adequately described by the pure luminosity evolution (Hasinger 1998, Miyaji et al. 2000). Although the critical redshift at which the QSO activities show a suddden transition is firmly established, it is unclear what determines such a redshift. There have been numerous attempts and models leading to debates as to what physical processes in QSO/AGN engines and/or their surroundings determine the basic characteristics of the cosmological evolution of QSO/AGN (e.g. Caditz et al. 1991, Small \\& Blandford 1992, Fukugita \\& Turner 1995, Yi 1996, Haehnelt et al. 1998, Nulsen \\& Fabian 2000). We intend to explore most of the existing classes of QSO/AGN evolution in their broad categories. The phenomenological scenarios we explore in this work can be roughly classified as follows. First, we consider a class of models in which a single long-lived ($\\geq 10^9$ yr) QSO population evolves throughout the cosmological time after birth at high redshifts $z>4$ (Mathez 1976, Yi 1996, Peterson 1997 and references therein). Second, we study the models in which many short-lived (a few $10^8$ yr) QSO populations form and evolve. In these models, the overall observed evolutionary trend is a result of the collective evolution of many different generations of QSO/AGN (Haehnelt et al. 1998). Third, a model of recurrent QSO/AGN activities in which some long-term variabilities of QSO/AGN emission dominates their recurrent activities (Siemiginowska \\& Elvis 1997). As a specific model of the third type, we assume that the variabilities are mainly caused by the accretion disk instability of the dwarf novae type (Frank et al. 1992) as is occasionally discussed in connection with the origin of high luminosity QSO activities. We consider two fundamentally different assumptions which are motivated by the recent works in the general area of acrretion disk physics. These assumptions have direct implications on the luminosity evolution of the QSO/AGN and as a consequence they are indirectly testable. The first one of the two is that the change of accretion flow around a central black hole, which powers the QSO/AGN activities, is mainly driven by the change in mass accretion rate. Such a change then results in the correlated spectral and luminosity evolution of QSO. This assumption is based on the observed behavior of black hole X-ray binaries (BHXBs). Some BHXBs show prominent spectral changes which are strongly correlated with the luminosity changes (e.g. Rutledge et al. 1999, Choi et al. 1999a, 1999b, 2000). Although the details of the models accounting for this behavior vary significantly, it is widely believed that the changes in the mass accretion rate is the underlying cause of this phenomenon. For instance, the ADAF models (Narayan et al. 1998, Yi 1996 and references therein) have enjoyed a relative success in explaining the hard X-ray emitting states of BHXBs while the thin disk models (Frank et al. 1992) have been widely applied to the soft X-ray emitting states. In these models, the continuous changes in the spectral state is interpreted as changes in the accretion rate and the accretion disk's physical conditions (e.g. Esin, McClintock, \\& Narayan 1997, and reference therein, Rutledge et al. 1999). We assume that the QSOs' luminosity evolution is accompanied by a strong spectral evolution caused by the accretion flow transition (see Choi et al. 1999a for details). The spectral state is determined by the BH mass and the physical accretion rate. The dominating effect is obviously that the resulting luminosity evolution shows distinct evolutionary behavior in different energy bands. The nearly mass scale-invariant nature of the accretion flow properties further support this assumption (Narayan \\& Yi 1995). The second one is that the QSOs luminosities are interpreted as fixed fractions of the bolometric luminosity and each band's luminosity is simply given as a fixed fraction of the bolometric luminosity regardless of changes in the accretion rate. The latter approach is closer to those of the conventional studies carried out so far (Peterson 1997). Although this assumption is not clearly supported by any physical models of accretion flows, the lack of detailed QSO spectral information in different bands makes it hard to rule out this simple but convenient approach. In our previous work (see Choi, Yang, \\& Yi 1999b, 2000 for details), we looked into the pure luminosity evolution model (Mathez 1976, Peterson 1997) with the explicit inclusion of the spectrum-luminosity correlation. In this model, all QSOs are long-lived and they become gradually dimmer with their comoving space number density conserved throughout the evolution, i.e. from their roughly synchronous births to the present epoch. We arrived at a conclusion that the first assumption, i.e. the accretion flow transition, can be accommodated in the QSO evolution with relatively good fits to the observed luminosity evolution. According to this model, however, the smaller mass black holes with the masses $< 10^8 M_{\\odot}$ in galactic nuclei (Magorrian et al. 1998) cannot be direct remnants of the past QSO activities as the QSOs' black holes grow much more massive than these black holes (cf. Wandel 1999). In this model, the QSO remnants have to exist in the nuclei of rare massive galaxies. Then, the often discussed massive dark objects in ordinary galaxies (Magorrian et al. 1998) should have formed and grown without experiencing the QSO phase, which does not appear to be a popular proposition. In the present work, we explore a number of QSO/AGN evolution models and make both qualitative and quantitative comparisons among them. The three types of the models we consider are (i) the single long-lived population, pure luminosity evolution model (Yi 1996, Choi et al. 1999b), (ii) the density evolution model in which the QSOs' evolution is the superposition of multiple generations of short-lived ($\\sim$ a few $10^6-10^8$ yr) QSOs (Haehnelt et al. 1998), and (iii) the model in which the QSO luminosities undergo recurrent variabilities (Siemiginowska \\& Elvis 1997). The second type of the models have been studied recently (e.g. Haehnelt \\& Rees 1993, Haehnelt, Natarajan, \\& Rees 1998) and they are specifically based on the idea that the hierarchical build-up of normal galaxies and the evolution of the AGNs/QSOs are closely connected to each other. It has been claimed that such models are supported by the recent observational evidences such as the existence of massive black holes (BHs) with $ 10^{6-10} M_{\\odot}$ (e.g. Franceschini et al. 1998), and the strong correlation in their masses between the supermassive BH and the spheroidal components of nearby galaxies (e.g. Magorrian et al. 1998, cf. Wandel 1999). In addition, we also explore a specific physical model in which the QSOs' long-term variabilities caused by accretion disk instabilities contribute considerably to the observed QSO LFs (Siemiginowska \\& Elvis 1997). We use these classes of models and make comparisons among them using the derived LFs and MFs along with the observational data. In sections 2 and 3, we summarize the evolution models and describe how we determine model parameters in each model. We derive the resulting analytical LFs of QSO/AGN and MFs of BH remnants. In section 4, we draw our conclusions and discuss their implications on how to interpret the observational data. ", "conclusions": "We have explored diverse types of possible phenomenological scenarios for the cosmological evolution of QSOs. Based on the basic paradigm in which QSO/AGN are powered by accreting supermassive black holes (Frank et al. 1992), we have considered various spectral states correlated with the luminosity level (Choi et al. 1999b) and constructed the analytical LFs in the various energy bands. Using the MF of remnant BHs in nearby galactic nuclei as an additional constraint, we have examined whether there exists a model simultaneously satisfying both LF and MF. The best and hence most promising scenario along with the best fit parameters points to the following formation and accretion history for supermassive BHs and QSOs. \\begin{enumerate} \\item In the multiple population models, due to the rapid decrease of $\\dot M$ caused by the short evolutionary time scale and to the redshift dependence of the evolutionary time scale, the number density of bright QSOs has to evolve strongly over the cosmological time. According to the derived soft X-ray LFs in a good agreement with observed LFs (Miyaji et al. 2000), the QSOs with less massive seed BHs must have been born more abundantly and/or refueled more efficiently at lower redshifts. \\item While the space number density obtained from the soft X-ray LF of AGN by (Miyaji et al. 2000) is about only 3 times higher than that observed by Boyle et al. (1991), the resulting comoving space densities in the BH remnants inferred from the derived best-fit soft X-ray LFs in all QSO evolution models, which we have tested, are smaller than those estimated for the putative BHs in nearby galaxies by a factor of about 10 to 100. No models we have explored can derive the LF and MF satisfying both of the two observational constraints well enough. This could imply that it is highly unlikely that all galaxies contain MDOs which are actually supermassive BH remnants of QSOs. Alternatively, there could be a population of highly obscured AGN which harbor growing BHs (Fabian 1999, Hasinger 2000). \\item There is little difference in evolving shapes of soft X-ray LFs among all the models we considered. This is inevitable given the fact that our models have at least been required to fit the observed LFs of QSO/AGN in the soft X-ray bands (Miyaji et al. 2000). We, however, find substantial differences in optical ($4400 \\AA$) LFs and hard X-ray LFs, which is a natural consequence of the assumed spectrum-luminosity correlation (Choi et al. 1999a, 1999b, 2000). In the case of the spectral evolution, the optical LF evolves more rapidly than the X-ray LF especially at lower redshifts or when QSOs become faint. There is a great difference among the evolving features of the hard X-ray LFs in various models. This is one of the strongest outcomes in the spectral evolution models. Expected observational data by hard X-ray missions such as the {\\it Chandra} X-ray observatory could play a role in providing a much needed, additional observational constraint. Such a constraint will shed light on which QSO models are more plausible and provide a statistical test on whether QSOs have experienced the spectral transition over the cosmological time scale. Such a transition is obviously untestable in individual galaxies due to the prohibitively long evolution and transition time scales. \\item The multiple population model with no spectral evolution gives the evolving shape of LFs which fit the observed B band LF of QSO derived by Boyle et al. (1991) in the optical ($4400 \\AA$) band as well as the soft X-ray LF. It is by far the best model among the various evolution models we have tested. This result supports the possibility that the QSO population is composed of many short-lived (a few 10$^8$ yr) generations while in each short generation QSOs do not experience any significant spectral evolution. \\end{enumerate} This work was supported by a KRF grant No. 1999-001-D00365. JY was also supported by the MOST through the National R \\& D program (99-N6-01-01-A-06) for women's universities. IY wishes to thank Anna I. Yi and Fojyik Yi for numerous discussions and helpful suggestions and Ethan Vishniac for hospitality while this work was in progress." }, "0005/astro-ph0005073_arXiv.txt": { "abstract": "I argue that it is very unlikely that X-rays from quiescent black-hole low-mass X-ray binary transients are emitted by coronae of companion stars. I show that in a simple model in which these X-rays are emitted by an ADAF filling the inner part of an unsteady, dwarf-nova type disc, the X-ray luminosity is correlated with the orbital period. I predict what values of X-ray luminosities from black-hole transient systems should be observed by {\\sl Chandra} and {\\sl XMM-Newton}. ", "introduction": "Low-mass X-ray binary transient systems (LMXBTs; these systems are also called `X-ray novae' or `Soft X-ray transients') are low-mass X-ray binaries (LMXBs) which sometimes (rarely for the most part) undergo outbursts during which the X-ray luminosity increases by more than 5 to 6 orders of magnitude. In LMXBs, a black hole or neutron star primary accretes matter lost by a Roche-lobe filling, low-mass secondary star. All known low-mass X-ray binaries (LMXBs) containing black holes are transient (Tanaka \\& Shibazaki 1996), whereas many neutron star LMXBs are steady, in the sense that unlike black-hole LMXBs they do not show high amplitude outbursts, but only low amplitude X-ray flux variation. Matter transferred from the secondary, forms an accretion disc, which far away from the accreting object is quasi-Keplerian. Accretion discs in LMXBTs appear to be truncated in their inner regions (Esin et al.1997; \\.Zycki et al. 1998,1999). Since here magnetic fields can play no role (because the magnetic moments of neutron stars are too low, and because of the absence of black holes magnetic fields) the inner disc `hole' can be due only to some kind of evaporation (Narayan \\& Yi 1995; Honma 1996). In such a case the inner accretion flow onto the compact object may form an advection-dominated accretion flow (ADAF; Abramowicz et al. 1995; Narayan \\& Yi 1995). Truncated discs are also required by the disc instability model (DIM), which is supposed to describe LMXBT outbursts (Lasota 1996; Menou et al. 2000; Dubus, Lasota \\& Hameury 2000). This model was devised to describe dwarf nova (DN) outbursts (see Cannizzo 1993 and Lasota 2000b for reviews). Also in these systems truncated discs are required to reconcile models with observations (e.g. Lasota, Kuulkers \\& Charles 1999; Meyer \\& Meyer-Hofmeister 1994; Shaviv, Wickramasinghe \\& Wehrse 1999) According to models of such truncated discs the inner accretion flow is an optically thin, very hot plasma, in which temperature may be close to the virial temperature. It is therefore expected to emit a considerable part of its energy in X-rays. This is indeed observed in quiescent DN and LMXBTs where such inner hot flows should be present (Eracleus, Helfand \\& Patterson 1991; Richards 1996; van Teeseling, Beuermann \\& Verbunt 1996; McClintock, Horne \\& Remillard 1995; Verbunt et al. 1994; Wagner et al. 1994; Asai et al. 1998; Barret, McClintock \\& Grindlay 1996). The properties of this X-ray emission provide an important test of accretion flow models (see e.g. Quataert \\& Narayan 1999; Meyer, Meyer-Hofmeister \\& Liu 1996). However, since observed X-ray luminosities are often rather low, one should be sure that the X-rays are not emitted by other sources, in principle less powerful than accretion. For DN it was shown that quiescent X-rays are emitted by the accretion flow and not by the secondaries coronae (Eracleus et al. 1991 van Teeseling et al. 1996; Richards 1996). For LMXBTs Verbunt (1996) concluded that (``except maybe for A0620-00\") X-rays cannot be emitted by coronae of secondary stars. In the case of neutron-star LMXBTs Brown, Bildsten \\& Rutledge (1998) attribute the quiescent X-rays to thermal emission from the neutron-star surface. This emission would be due to repeated deposition during the outbursts of nuclear energy deep in the crust. This could be a viable alternative to the accretion model (Rutledge et al. 1999, see however Menou et al. 1999c). Recently Bildsten \\& Rutledge (1999) concluded that in the case of black-hole LMXTBs the quiescent X-rays may be due to coronal emission from stellar companions. They argue that in these systems the ratio of the X-ray flux to the stellar, bolometric flux is $\\lta 10^{-3}$ as in RS CVn's, which are active, close, detached binaries of late-type stars (a G of K type giant or subgiant in orbit with a late-type main-sequence or subgiant) in which, for orbital periods $\\lta 30$ days, the rotation of both components is synchronous with the orbit. Their coronal X-ray emission may be as large as $10^{31}$ erg s$^{-1}$ (Dempsey et al. 1993). Unfortunately, only three quiescent black-hole LMXBT systems were detected in X-rays. In A0620-00 the quiescent ($\\sim$ 2 - 10 keV) luminosity is $L_X=10^{31}$ erg s$^{-1}$, in the other two systems (GRO J1655-40 and V404 Cyg) $L_X>10^{32}$ erg s$^{-1}$ (see Garcia et al. 1997 and references therein). This sample is not only small but also very eclectic as far as companion stars are concerned. The secondary in A0620-00 is a late type dwarf (K5V, McClintock \\& Remillard 1986), in GRO J1655-40 the F3-6 (Orosz \\& Bailyn 1997) secondary is either near the end of its main-sequence life (Reg\\\"os, Tout \\& Wickramasinghe 1998) or is crossing the Hertzsprung gap on its way to the giant branch (Kolb et al. 1997, Kolb 1998), and finally in V404 Cyg the K0 (Casares, Charles \\& Naylor 1992) secondary is a `stripped' giant (King 1993). In the case of the last system Bildsten \\& Rutledge (1999) admitted that its $L_X= 1.6\\times 10^{33}$ erg s$^{-1}$ cannot be emitted by the companion's corona (they find $L_X/L_{\\rm bol}=8\\times 10^{-2}$). However, except for this system and for 4U1543-47 (see below), Bildsten \\& Rutledge (1999) expect quiescent X-ray luminosity of black-hole LMXTBs to originate in the coronae of secondaries. I discuss this hypothesis in Sect. 2. and conclude that it cannot be correct. Black-hole LMXBT's secondary stars cannot be the source of quiescent X-rays because they are not different from their dwarf nova counterparts. In the (two known) cases where these secondaries are different, their coronal X-ray luminosity should be {\\sl lower} than in the corresponding active star binaries, so that also in this case quiescent X-ray luminosity can only result from accretion. In Section 3 I discuss what the disc instability model of dwarf novae and LMXBTs has to say about quiescent X-ray emission and in Section 4 I show that, on simple assumptions, this model combined with an ADAF model (as first proposed by Narayan, McClintock \\& Yi 1996; see also Lasota, Narayan \\& Yi 1996) predicts a correlation between the quiescent X-ray luminosity and the orbital period. This correlation is satisfied by the three observed systems, which allows one to make predictions about future observations by {\\sl Chandra} and {\\sl XMM-Newton} of systems for which up to now only upper limits are known. Section 5 ends the article with discussion and conclusions. \\begin{figure} \\resizebox{\\hsize}{!}{\\includegraphics{fig1_lum.eps}} \\vskip -.3cm \\caption{X-ray luminosities of quiescent dwarf-novae and soft X-ray transients and X-ray luminosities of RS CVn stars. The continuous line marked `FGM' is the limit given by Eq. (\\ref{fgm}) with $M_2=1M_{\\odot}$. The dashed line corresponds to the relation Eq. (\\ref{lx}). Down-pointing triangles correspond to upper limits.} \\label{luminosities} \\end{figure} ", "conclusions": "The quiescent X-ray luminosity model is rather unsophisticated, but it is based on several natural properties of the ADAF + DIM-disc model of LMXTs and has the advantage of making definite predictions that, one might hope, will be tested by observations. Any detection of GRO J0422+32, H 1705-25, GS1124-68 would eliminate the coronal model, as pointed out by Bildsten \\& Rutledge (1999). The detection of all five systems close to the predicted fluxes would count as a success for the ADAF + DIM-disc model and in particular for the MAH (first suggested by Narayan \\& Yi 1995). One should keep in mind, however, that if the MAH were to apply in these systems this would not prove its general validity. In binaries the mass-transfer stream provides a a `barrier' which prevents the spreading of evaporation. What would form such a barrier in galactic nuclei is not clear, but in NGC 4258 the ADAF extends only to few hundred $R_G$ (Gammie, Narayan \\& Blandford 1999, whereas it could in principle extend up to $\\sim 10^4R_G $). The confirmation of correlation between X-ray luminosity and orbital period given by Eq. (\\ref{lx}) would not necessarily mean that the inner, hot, accretion flow is an ADAF in the sense of Abramowicz et al. (1995) and Narayan \\& Yi (1995). This correlation means only that the rate at which accretion enters the hot flow is correlated with the size of the inner `hole', hence with the orbital period. The `standard' ADAF model is a good representation of the inner flow. It is not clear that it is only one. More detailed observational diagnostics is required to decide what is the real solution. Quiescent dwarf novae are not expected to satisfy Eqs. (\\ref{slope}) or (\\ref{lx}). First, most of them have recurrence times much shorter than black-hole LMXBs, so that the quiescent accretion rate should vary on a short time-scales. In fact X-ray quiescent luminosity is highly variable in these systems (Verbunt, Wheatley \\& Mattei 1999). Second, in many cases the inner disc truncation might be due to the action of the white dwarf's magnetic field rather than to evaporation (e.g. Lasota et al. 1999). In such a case the transition radius is just the magnetospheric radius and is independent of the orbital period. In the case of neutron-star LMXTBs there are two possibilities. First, as proposed by Brown et al. (1998), the quiescent X-ray luminosity could be due to the neutron star surface being heated by thermonuclear reactions in the matter accreted during outbursts. In this case X-ray observations would allow the quiescent disc to extend down to the neutron star (or the last stable orbit). Second, as described in Menou et al. (1999c), the inner part of the quiescent accretion flow would form an ADAF as in black-hole LMXTBs. In this case the advected thermal energy must be emitted from the stellar surface so, for the same mass-transfer rate, quiescent neutron-star LMXTBs should be more luminous than those containing black holes, as observed. The problem, however, is that the model predicts quiescent neutron-star LMXTB luminosities much higher than observed. In order to reduce the accretion rate onto the neutron star (by three orders of magnitude) an ADAF model has, therefore, to involve the `propeller effect' and perhaps the presence of winds (Menou et al. 1999c), so it is not as simple as the black-hole LMXBT models (see however Quataert \\& Narayan 1999). In this case, however, as in DN, one would not expect a correlation of the X-ray luminosity with the orbital period. Observations of variability of quiescent neutron-star LMXTBs could help in choosing the correct model. It is worth noting that simulations of outburst cycles of LMXBTs show that disc truncation is necessary if one wishes to reproduce observed properties of outburst cycles (Menou et al. 2000; Dubus et al. 2000). For black-hole LMXTBs things are simpler: all available evidence suggests that quiescent X-rays are emitted by an ADAF filling the inner part of a truncated DIM-type accretion disc. A particular version of the model of such a configuration will be tested by {\\sl Chandra} and {\\sl XMM-Newton}." }, "0005/astro-ph0005559_arXiv.txt": { "abstract": "We have detected a bright X-ray flare from the nearby ($d=$5.0 pc) brown dwarf \\lp\\ with the \\chandra/ACIS-S. This is an old (500 Myr), rapidly rotating, lithium-bearing M9 object, with a bolometric luminosity of $\\approx 6$\\tee{29} \\cgslum. It was only detected by {\\it Chandra} during an X-ray flare of duration 1-2 hours near the end of a 12.1 hour observation. The peak X-ray luminosity was 1.2\\ud{0.5}{0.3}\\tee{26} \\cgslum\\ in the brightest $\\approx 550$ seconds, corresponding to $L_X/L_{\\rm bol}\\approx 2\\times 10^{-4}$. A total of 2\\tee{29} ergs was released during the 43,773 sec observation, giving a time-averaged $L_X/L_{\\rm bol}\\approx 7\\times 10^{-6}$. \\lp\\ was not detected before the flare, with a 3$\\sigma$ upper limit on the emission at $L_X/L_{\\rm bol}< 2\\times 10^{-6}$ ($L_X<$1\\tee{24} \\cgslum). This is faint for a rapidly rotating late-type star, and establishes a record lower limit to the quiescent flux about an order of magnitude below the flux limit (and a factor of 5 below the \\lxlbol\\ limit) placed on quiescent X-ray emission from the M8 dwarf VB 10. The inferred flaring duty cycle is comparable to that measured via variable H$\\alpha$ emission for other late M-type, fully convective stars. ", "introduction": "\\lp\\ (=BRI 0337$-$3535) is an isolated, non-accreting, brown dwarf identified through its Li abundance and low luminosity \\cite{tinney98discovery}. Its parallactic distance and bolometric luminosity are 5.0\\ppm0.1 pc and 6\\tee{29}\\cgslum \\cite{tinney96}. Tinney ~\\cite*{tinney98discovery} infers an age of about $\\tau=$500 Myr, which implies that it is a fully collapsed object (see \\citenp{feigelson99} for a recent review of protostellar evolution). This requires that any observed coronal activity not be due to accretion, as may power younger brown dwarfs ($\\tau<$10 Myr) in analogy with T-Tauri stars. We report here on a bright 1-2 hour X-ray flare detected with \\chandra \\ during a 12 hour observation, and a strong upper limit on persistent X-ray emission. We attribute the energy release in the flare to transient magnetic activity on this fully convective, rapidly rotating star. Stars with masses $M>0.3 M_\\odot$ have an outer convective zone and an interior radiative region that need not be rotating at the same rate. A poloidal magnetic field in the convective layers will be stretched and amplified into strong toroidal fields when it is dragged by convective overshoot (see \\citenp{weiss94}) into the radial shear in rotation that resides at the boundary (in and near the so-called \"tachocline\"; \\citenp{spiegel92}). In these cases, rapid rotation is associated with enhanced coronal activity. The activity level correlates with the Rossby number -- the ratio of the stellar rotation period to the convective turnover time \\cite{Noyes84}. For Rossby numbers between 10 and 0.1, coronal activity (as measured by \\lxlbol) increases with decreasing Rossby number (that is, with more rapid rotation), ``saturating'' at about \\lxlbol $\\sim 10^{-3}$ for Rossby numbers of 0.1 to 0.01 \\cite{randich98}. For less massive stars and young brown dwarfs, the energy is transported throughout the star by convection; no radiative core is present. For this reason, it has been supposed that the activity and its dependence on rotation might change near the spectral type where the radiative layer disappears (about M5.5; see \\citenp{giampapa96} for a physical overview). However, a search for this effect found no evidence for a change in the saturated value of \\lxlbol $\\sim 10^{-3}$, down to spectral types as late as M7, well into the region of fully convective stars \\cite{fleming93}. This implied no dramatic change in the rotation-activity dependence in fully convective stars. The first hint that activity might be decreasing in the very late M dwarfs was observed from the rapidly rotating ($v \\sin i = 40$ \\km\\ \\persec) M9.5 star BRI 0021-0214, for which a strong upper-limit on \\hal\\ emission indicated a substantially lower persistent coronal activity than expected from a rapidly rotating, fully convective star \\cite{basri95}. On the other hand, Reid \\etal\\ \\cite*{reid99dec} observed a strong \\hal\\ flare from BRI 0021-0214; this indicated that activity was indeed present, and that the star has outbursts no more than 7\\% of the time. Liebert \\etal\\ ~\\cite*{liebert99} reported a bright \\hal\\ flare from 2MASSW J0149090+295613, an M9.5 V star which is otherwise quiescent. Intensity variability from brown dwarfs has been searched for, with limited success. Bailer-Jones \\& Mundt \\cite*{bailerjones99} found no infra-red variability in three Pleiades brown dwarfs, with a limit of $\\delta I<$0.05 mag on timescales between 25 min-27 hrs. A search for ``weather'' in two brown dwarfs (\\lp\\ and DENIS-P J1228-1547) produced a claim of evidence of variability in one, at the 2.3$\\sigma$ level \\cite{tinney99}. Previous X-ray observations of brown dwarfs and brown dwarf candidates have produced detections of persistent emission from several \\cite{neuhauser98,neuhauser99a}, but these are all young objects (\\approxlt 10 My), still in the process of proto-stellar collapse, and so are actively accreting; neither are they as cool or faint as older brown dwarfs such as \\lp. The activity observed from these collapsing brown dwarfs is analogous to that in the also-young T-Tauri stars -- powered by accretion and collapse -- rather than that in fully collapsed, older M dwarfs. There are only a few observations of X-ray emission in older late M dwarfs and brown dwarfs. Giampapa \\etal~\\cite*{giampapa96} reported on a ROSAT detection from VB 8 (spectral type M7) with a time-averaged \\lxlbol =1.6\\tee{-3}. Fleming \\etal~\\cite*{fleming00} detected the M8 star VB 10 during a flare at \\lxlbol =5\\tee{-4}. No quiescent X-ray emission was detected, limiting it to \\lxlbol $<$\\ee{-5}. In the only previous observations of \\lp , an upper limit of \\lxlbol $<$7\\tee{-5} was found \\cite{neuhauser99a}. Hence, as with \\hal\\ observations, the X-ray behavior of late M stars tends toward flaring activity and an absence of persistent activity at the X-ray detection limits of present instrumentation. Our work confirms this tendency. We report here on our \\chandra \\ observation of an X-ray flare from the brown dwarf \\lp. In \\S~\\ref{sec:obs}, we describe the observation, analysis and detected flux level. In \\S 3, we interpret the light curve of the flare. We conclude in \\S~\\ref{sec:con} with a brief summary of our results and a comparison to other work. ", "conclusions": "\\label{sec:con} We have detected an X-ray source which we identify with the brown dwarf \\lp, with 99.98\\% confidence, based on positional coincidence. The detection was during a flare, with a peak X-ray luminosity of $L_X=$1\\ud{0.5}{0.3}\\tee{26} \\cgslum, or $L_X/L_{\\rm bol}\\approx 2\\times 10^{-4}$. This flare's peak luminosity is a factor of ten below that of an X-ray flare detected from the comparably luminous M8 star VB~10 ($L_{\\rm Bol}\\approx 1.7 \\times 10^{30} \\ {\\rm erg \\ s^{-1}}$; \\citenp{fleming00}). In VB~10, the peak X-ray luminosity exceeded $10^{-3} L_{\\rm Bol}$. We have analysed the X-ray light curve with a simple sharp rise plus exponential decay model; this indicates a decay time $5400\\pm700 \\ {\\rm s}$ when we specify the flare start-time, and in the range of 500-7100 s when we do not specify the flare start time. This implies \\lp\\ is flaring $\\sim$10\\% of the time, comparable to the amount of \\hal\\ flaring among the coolest M dwarfs \\cite{gizis00}, although the uncertainty on our value is large. The light-curve is consistent with rotational modulation, but the counting statistics are too poor to conclude that it is required, and discovery of such modulation must wait for higher signal-to-noise data. A previous attempt to detect this source in X-rays \\cite{neuhauser99a} placed an upper-limit on the time-averaged luminosity ($L_X< 4\\times 10^{25} \\ {\\rm erg \\ s^{-1}}$) during a 66 ksec \\rosat/PSPC observation and a 220 ksec \\rosat/HRI observation. These limits are above the time-averaged luminosity we detect here ($L_X\\approx 4\\times 10^{24} \\ {\\rm erg \\ s^{-1}}$), and are therefore consistent with our results. X-ray variability was not discussed by these previous works; however, a cursory review of 220~ksec of ROSAT/HRI data revealed no variability on timescales of 500 sec; the detection sensitivity (5 counts in a 500 sec bin, energy conversion factor of 3.2\\tee{-11} \\cgsflux\\ \\perval{count}{-1}), is about a factor of 10 above the present peak-flare detection. The time-averaged luminosity ($L_X$=4\\tee{24}\\cgslum) is a factor of $10^4$ below the previous detections of brown dwarfs and brown dwarf candidates; the time-averaged ratio \\lxlbol(=7\\tee{-6}) is factor of $10^2$ below this ratio \\cite{neuhauser98,neuhauser99a}. This might be due to different energy production mechanisms; the previously detected brown dwarfs are young compared with \\lp\\ ($\\sim$1 My vs. 500 My) and are still forming, while \\lp\\ is isolated and undergoing slow gravitational contraction. The persistent X-ray emission of the very young brown-dwarfs in open clusters is more analogous to the pre-main sequence T-Tauri stars -- where convection is stronger, the atmospheres are less neutral, and accretion may play a role -- while the flaring X-ray emission of \\lp\\ is more analogous to the flaring emission of late-type main-sequence stars, such as VB~8 and VB~10. We estimate (order of magnitude) an eddy turnover time of $\\sim$1 year for \\lp, yielding a Rossby number of $R_0 \\sim$5\\tee{-4}. As discussed earlier, this is below the Rossby number limit at which the X-rays are observed to be \"saturated\" (at \\lxlbol$\\sim$\\ee{-3}). That saturation level is much greater than our quiescent X-ray flux upper limit. Either the \"supersaturation\" suggested by Randich has strongly set in, or the connection between Rossby number and observed activity is no longer relevant. It is possible that the rapid rotation suppresses persistent coronal activity, either by forcing the magnetic field into a more organized form or by suppressing the turbulent dynamo. It is also possible that the neutrality of a cool photosphere quenches coupling between atmospheric motions and the magnetic field, which forces the field into dissipative configurations (as pointed out by \\citenp{fleming00}). In any case, the detected flare requires that a magnetic field be present on LP 944-20, and that at least occasionally it is forced into a dissipative configuration high in the atmosphere. Our results confirm the impression from previous studies of \\hal\\ that stellar activity is dying at the bottom of the main sequence, at least in the form that it has in more massive late-type stars. We have pushed the limits on quiescent coronal emission levels to new lows for fully convective objects. We have also helped confirm that such objects do apparently still have magnetic fields and an ability to flare about 10\\% of the time." }, "0005/hep-th0005028_arXiv.txt": { "abstract": "{We examine the cosmological effects of the Hagedorn phase in models where the observable universe is pictured as a D-brane. It is shown that, even in the absence of a cosmological constant, winding modes cause a negative `pressure' that can drive brane inflation of various types including both power law and exponential. We also find regimes in which the cosmology is stable but oscillating (a bouncing universe) with the Hagedorn phase softening the singular behavior associated with the collapse.} \\begin{document} ", "introduction": "\\label{sec:intro} Inflation \\cite{guth} is a beautiful solution to several difficult problems in cosmology; the horizon problem, the flatness problem, and the monopole problem. In field theory however, the standard way to obtain inflation is to add a positive cosmological constant (which has a negative pressure $p=-\\rho)$. This ingredient is, without doubt, the least attractive feature of standard inflation and it is generally extremely difficult to control its adverse effects (\\eg the graceful exit and moduli problems). It is worth asking therefore if there are other forms of matter that can have a negative pressure and hence give an accelerating scale factor. In this paper we introduce a candidate that actually has this property -- open strings on D-branes at temperatures close to the string scale. At sufficiently high temperatures and densities fundamental strings enter a curious `long string' Hagedorn phase~\\cite{carlitz,general,deo,thorl,abkr,after}. To date applications of this phase have been quite limited in string cosmology~\\cite{general} because the thermodynamics is governed by the finite temperature partition function. A rigorous analysis therefore requires nothing less than solving the string system in a cosmological setting, a difficult problem that might at best be tractable only in a few special cases. Moreover, in order to understand the effect of macroscopic phenomena such as winding modes we need the microcanonical ensemble (as we shall discuss) -- an ensemble that does not particularly lend itself to cosmological applications. In this paper we show that, for certain systems, it is possible to bypass these technical difficulties by using a classical random walk picture to model the behaviour of the strings in cosmological backgrounds. The particular systems we will focus on are D-branes in the weak coupling limit~\\cite{polch}. In this limit one can separate the energy momentum tensor into two components; a localized component corresponding to the D-brane tension, and a diffuse component that spreads into the bulk corresponding to open string excitations of the brane. (At the risk of causing confusion we will often refer to the latter as a `bulk' component.) We will in addition allow a bulk cosmological constant, although our focus in this paper will be on the cosmology when the combined effect of the brane and bulk cosmological constants is subdominant. The random walk picture allows us a first glimpse at the cosmological effects of a primordial Hagedorn phase of open strings on branes and we find two interesting types of behaviour. \\begin{itemize} \\item The first we call {\\em Hagedorn inflation}. We will show that the transverse `bulk' components of a D-brane's energy-momentum tensor can be negative. If all of the transverse dimensions have winding modes, this negative `pressure' causes the brane to power law inflate along its length with a scale factor that varies as $a\\sim t^{4/3}$ even in the absence of a nett cosmological constant (as shown in eq.(\\ref{61})). If there are transverse dimensions that are large (in the sense that the string modes are not space-filling in these directions), then we can find exponential inflation (as shown in eq.(\\ref{expinf})). \\item If there is a small but negative cosmological constant, the Universe can enter a stable but oscillating phase; \\ie a `bouncing' universe. The nett effect of the Hagedorn phase is to soften the singular behaviour associated with the collapse. Such singularity smoothing is a familiar aspect of strings, but the nice feature here is that we find it in a purely perturbative regime. \\end{itemize} We should at this point also emphasize a general observation that we make, namely that the diffuse stringy component can have a dominant effect on the cosmology even in the weak coupling limit. At first sight this may be somewhat surprising given that the intrinsic tension energy of a D-brane goes like $\\rho_{br}\\sim 1/g_s \\sim 1/\\hat{\\kappa} $ where $g_s$ is the string coupling and $\\hat{\\kappa}$ is the effective gravitational coupling. However, we will see that the {\\em cosmological} effects of the two components are proportional to $\\hat{\\kappa}^4 \\rho_{br}^2 $ and $\\hat{\\kappa}^2 \\rho$ for the brane and diffuse `bulk' components respectively. Then the contribution of the brane component is proportional to $\\hat{\\kappa}^4 \\rho_{br}^2 \\sim \\hat{\\kappa}^2$. Since $\\frac{1}{g_s}\\geqsim \\rho \\geqsim 1$ to be in the Hagedorn phase, if \\eg the transverse volumes are of order unity in string units, then the contribution of the bulk component is $\\hat{\\kappa} \\geq \\hat{\\kappa}^2 \\rho \\geq \\hat{\\kappa}^2$. Hence the cosmological effect of the diffuse bulk component can be dominant when $g_s$ becomes small. We begin in sections 2 and 3 by deriving the energy-momentum tensor, the principal ingredient of Einstein's equations. Since the results can be understood rather intuitively, this part of the discussion is organized so that cosmologists (and indeed anybody else) can skip the bulk of sections 2 and 3 concerning string thermodynamics and proceed directly to the energy-momentum tensor which is summarized at the end of section 3. The thermodynamic discussion of section 2 gives a detailed introduction to the behaviour of both type I and type IIA/B open strings on D-branes as calculated from the microcanonical ensemble in a flat background. In particular we discuss the importance of macroscopic modes such as winding modes. Much of this section is a collation of results from ref.\\cite{abkr}. We then reintroduce the classical random walk picture paying special attention to the meaning of quantities such as average string length. In section 3 we use the thermodynamic results to calculate the energy momentum tensor $T_{\\mu\\nu}$ of the Hagedorn phase. $T_{\\mu\\nu}$ enters into the higher dimensional Einstein's equations and determines the cosmology, and in particular we show that open string winding modes gives negative transverse components. For convenience the results for $T_{\\mu\\nu}$ are summarized at the end of section 3 where we also discuss the heuristic interpretation of this negative pressure. Armed with the energy-momentum tensor, we examine the resultant cosmology. In sections 4 and 5, we solve the equations of motion with various ans\\\"atze, and find the advertised inflationary behaviour as well as bouncing solutions with singularity smoothing behaviour. We will, purely for definiteness, consider adiabatic systems in solving the evolution equations of the universe. Under the assumption of adiabaticity, the inflationary growth period drives down the temperature of the system; eventually the temperature drop causes the universe to leave the Hagedorn regime, and consequently inflation ends automatically. However adiabatic systems are probably unable to provide a realistic scenario with sufficient inflation. In section 6 we therefore discuss how, in non-adiabatic systems, inflation can be sustained. We conclude in section 7. ", "conclusions": "In this paper we have studied the possible cosmological implications of the Hagedorn regime of open strings on D-branes in the weak coupling limit. Our main result in sections 2-5 is that, due to the non-extensive dependence of the free energy on the volumes, a gas of open strings can exhibit negative pressure leading naturally to a period of power law or even exponential inflation -- Hagedorn inflation. We also find that the open string gas can dominate the cosmological evolution at weak coupling even though the D-brane tension becomes large in this limit. Hagedorn inflation also has a natural exit since any significant cooling can cause a change in the thermodynamics if winding modes become quenched or if the density drops below the critical density, $\\rho_c\\sim 1$, needed for the entropy of the Hagedorn phase to be dominant. Such a cooling can be caused by a sudden adiabatic increase in the transverse radius or by the inflation itself. We find this `easy-exit' feature of open-string Hagedorn inflation to be of its most appealing features. In addition, we found that a small but negative cosmological constant, can cause the universe to enter a stable but oscillating phase. The effect of the Hagedorn phase is to soften the singular behaviour associated with the bounce. The most striking aspect of our discussion is probably the existence of negative pressure. One might therefore ask how general a feature this is expected to be. By T-dualizing we argued that we can put the negative pressure down to the fact that in any particular direction, the gravitational degrees of freedom have both Kaluza-Klein modes and winding modes whereas the open strings (which are dominant in the entropy) have only one or the other. Hence we expect negative pressure to be possible whenever there are large space-filling modes that dominate the entropy. In section 6 we speculated on how the inflation might be sustained through the well known phenomenon of string instability in a de Sitter background. Hagedorn inflation may be thought of as a first example in the search for alternatives to the cosmological constant within the framework of string/brane systems. One possible direction for further study in this area is connected with the fact that we have throughout been taking the string coupling to be weak enough so that the brane tension does not play an important role in the cosmology. It is therefore interesting to ask if new cosmological effects might arise from large scale fluctuations in the branes themselves~\\cite{abkr,riotto}. The arguments of section 6 indicate that if it does then the brane driven inflation may be qualitatively different to the string driven inflation discussed here. This is because open strings are one dimensional objects with $S\\sim E + const \\sqrt{E} $ whereas fluctuating $p$-branes have an entropy~\\cite{penalba} \\be S\\sim E^{{2p} \\over{p+1}}. \\ee On calculating $\\beta=\\partial S/\\partial E$ we see that $p\\neq 1$ branes do not have the divergent behaviour which is the defining feature of strings. The thermodynamics of these objects has been the subject of much study \\cite{penalba} and there is probably more to be learnt here. Finally, an interesting connected issue which we did not discuss is related to the effect of brane melting discussed in ref.\\cite{abkr}, in which the non-perturbative aspect of D-brane thermal production must be taken into account. At the present time it is hard to make any quantitative estimate of these effects on the cosmological solutions we have been discussing here, but we hope to be able to address this question in future work. \\subsection*{Acknowledgements} \\noindent We thank Dan Chung, Cedric Deffayet, Emilian Dudas, Keith Olive, Geraldine Servant, Carlos Savoy and Richard Woodard for discussions. S.A.A. and I.I.K. thank Jose Barb\\'on and Eliezer Rabinovici for a previous collaboration and discussions concerning this work. S.A.A. thanks the C.E.A. Saclay for support. K.F. acknowledges support from the Department of Energy through a grant to the University of Michigan. K.F. thanks CERN in Geneva, Switzerland and the Max Planck Institut fuer Physik in Munich, Germany for hospitality during her stay. I.I.K. is supported in part by PPARC rolling grant PPA/G/O/1998/00567, the EC TMR grant FMRX-CT-96-0090 and by the INTAS grant RFBR - 950567." }, "0005/astro-ph0005009_arXiv.txt": { "abstract": "The STARE instrument is a small aperture, wide-field, CCD-based telescope that delivers high cadence time series photometry on roughly 40~000 stars in a typical field centered on the galactic plane. In a two-month observing run on a field, we obtain sufficient precision on roughly 4~000 stars to detect a close-in Jupiter-sized companion in an edge-on orbit. We also used this instrument to detect the planetary transits across the Sun-like star HD~209458. The project is now in its third season, and we have acquired a large dataset on several fields. Given the frequency of close-in extrasolar planets found by the radial velocity surveys, and the recent confirmation that at least some of these are indeed gas giants, the STARE project should be able to detect roughly a dozen Jupiter-sized planets in its existing dataset. ", "introduction": "Radial velocity surveys of nearby F, G, K and M dwarf stars have revealed a class of close-in extrasolar massive planets that orbit their stars with an orbital separation of $a \\la 0.1$ AU. Prior to the transit results for HD~209458, the radial velocity method has been the only method by which we have learned anything about these planets. The radial velocity technique measures the period, semi-amplitude, and eccentricity of the orbit, and by inference the semi-major axis. It also yields a value for the minimum mass, dependent upon the assumed value for the stellar mass, but aside from this it gives no direct information on the structure of the planet itself. The search to measure the transit photometrically is motivated by fact that, for a star for which both the radial velocity and transits are observed, one can estimate both the mass (with negligible error due to $\\sin i$) and radius of the planet. These can then be combined to calculate such critically interesting quantities as the surface gravity and average density of the planet, and thus provide constraints on structural models for these low-mass companions. Assuming random orbital alignment for systems with $a = 0.05$ A.U. and Sun-like primaries, the chance of a transiting configuration is roughly 10\\%. ", "conclusions": "" }, "0005/astro-ph0005379_arXiv.txt": { "abstract": "We report on the discovery of the radio afterglow from the gamma-ray burst (GRB) of 1999 May 6 (GRB 990506) using the Very Large Array (VLA). The radio afterglow was detected at early times ($\\Delta t=1.5$ days), but began to fade rapidly sometime between 1 and 5 days after the burst. If we attribute the radio emission to the forward shock from an expanding fireball, then this rapid onset of the decay in the radio predicts that the corresponding optical transient began to decay between 1 and 5 minutes after the burst. This could explain why no optical transient for GRB 990506 was detected in spite of numerous searches. The cause of the unusually rapid onset of the decay for the afterglow is probably the result of an isotropically energetic fireball expanding into a low density circumburst environment. At the location of the radio afterglow we find a faint ($R \\sim 24$ mag) host galaxy with a double morphology. ", "introduction": "\\grb\\ was detected by the BATSE (trigger No.~7549) on board the Compton Gamma-Ray Observatory (CGRO) on 1999 May 6.47 UT (\\cite{kip99a}). It was a bright burst, lasting approximately 150 s, with a fluence in the 50-300 keV band of 2.23$\\times10^{-4}$ erg cm$^{-2}$ -- placing it in the top 2\\% of BATSE bursts as ranked by fluence. The PCA instrument on board the Rossi X-Ray Timing Explorer (RXTE) began scanning a 8$^\\circ\\times6^\\circ$ region centered on the BATSE position some three hours after the burst (\\cite{mt99}) and discovered a previously uncataloged X-ray source, which faded significantly over the observing interval (\\cite{tm99}). The initial 12\\arcmin\\ error circle for the RXTE-PCA position of \\grb\\ was further refined to a 30 arcmin$^2$ region through the addition of timing annuli from the Interplanetary Network (IPN), involving the NEAR, Ulysses and CGRO satellites (\\cite{hkkc99}). Despite the fact that at least nine different optical telescopes imaged the IPN/RXTE-PCA region containing the X-ray transient (some as early as 90 minutes after the burst), no corresponding optical transient was detected. In this paper we report the discovery of a radio transient (\\vla) in this field, which we propose is the radio afterglow of GRB 990506. We also present upper limits on the brightness of the corresponding optical afterglow. We discuss the radio light curve of this transient, and its implications for the lack of detection of the optical afterglow. Finally, we present the discovery of the probable host galaxy of the radio transient, and by implication the GRB itself. \\grb\\ joins a small but growing class of bursts which evidently lack a bright optical transient but are seen at both X-ray and radio wavelengths. ", "conclusions": "We have identified the radio afterglow from GRB 990506. This GRB was unusual in that it produced a radio afterglow that began to fade at very early times (between 1 and 5 days after the burst) and in that no optical afterglow was detected in spite of the numerous deep images obtained. Both these observations may be explained by a high energy spherical fireball expanding into a low density environment. In this simple picture there is no need to invoke dust extinction to account for the lack of detection of an optical afterglow. We cannot rule out, however, that the radio emission originated in a reverse shock. If the reverse shock produced the radio afterglow, then the emission from the forward shock (in both optical and radio) was presumably too faint to be seen. To find additional GRBs of this type will require more rapid followup to precise burst localizations than has typically been achievable. Upcoming satellite missions such as HETE II and SWIFT should improve upon this situation." }, "0005/astro-ph0005465_arXiv.txt": { "abstract": "We present broad-band radio observations of the afterglow of \\grb{}, spanning from 1.4 to 350 GHz for the period of 3 to 83 days after the burst. This radio data, in addition to measurements at the optical bands, suggest that the afterglow arises from a collimated outflow, i.e. a jet. To test this hypothesis in a self-consistent manner, we employ a global fit and find that a model of a jet, expanding into a constant density medium (ISM+jet), provides the best fit to the data. A model of the burst occurring in a wind-shaped circumburst medium (wind-only model) can be ruled out, and a wind+jet model provides a much poorer fit of the optical/IR data than the ISM+jet model. In addition, we present the first clear indication that the reported fluctuations in the optical/IR are achromatic with similar amplitudes in all bands, and possibly extend into the radio regime. Using the parameters derived from the global fit, in particular a jet break time, $t_{jet} \\approx 7.5$ days, we infer a jet opening angle of $\\theta_0 \\approx 0.2$, and consequently the estimate of the emitted energy in the GRB itself is reduced by a factor $\\sim$50 relative to the isotropic value, giving $E\\approx 1.1 \\times 10^{51}$ ergs. ", "introduction": "\\label{sec:intro} \\grb\\ is the latest afterglow to exhibit a break in its optical/IR light curves. An achromatic steepening of the light curves has been interpreted in previous events (e.g. \\cite{kdo+99}; \\cite{H+99}) as the signature of a jet-like outflow (\\cite{rho99}; \\cite{SPH99}), produced when relativistic beaming no longer ``hides'' the non-spherical surface, and when the ejecta undergo rapid lateral expansion. The question of whether the relativistic outflows from gamma-ray bursts (GRBs) emerge isotropically or are collimated in jets is an important one. The answer has an impact both on estimates of the GRB event rate and the total emitted energy --- issues that have a direct bearing on GRB progenitor models. An attempt by Rhoads and Fruchter (2000) to model this break using only the early time ($\\Delta t \\lesssim 14$ days) optical/IR data has led to a jet interpretation of the afterglow evolution, but with certain peculiar aspects, such as a different jet break time at R band than at K' band. However, subsequent papers by Masetti et al.~(2000), and Sagar et al. (2000), with larger optical data sets, pointed out that there are large flux density variations ($\\sim 30\\%$) on timescales as short as a few hours, superposed on the overall steepening of the optical/IR light curves. While the origin of these peculiar fluctuations remains unknown, it is clear that they complicate the fitting of the optical/IR data, rendering some of the Rhoads and Fruchter results questionable. In this paper we take a different approach. We begin by presenting radio measurements of this burst from 1.4 GHz to 350 GHz, spanning a time range from 3 to 83 days after the burst. These radio measurements, together with the published optical/IR data, present a much more comprehensive data set, which is less susceptible to the effects of the short-timescale optical fluctuations. We then use the entire data set to fit a global, self-consistent jet model, and derive certain parameters of the GRB from this model. Finally, we explore the possibility of a wind, and wind+jet global fit to the data, and compare our results with the conclusions drawn in the previous papers. ", "conclusions": "\\label{sec:conc} The afterglow emission from \\grb\\ can be well described in the framework of the jet model of GRBs. Global fitting of the radio and optical data, allows us to calculate the values of $p$, $t_{jet}$, and the time evolution of $\\nu_a$, $\\nu_m$, and $F_{\\nu,max}$ in a self-consistent manner. Within this approach the proposed discrepancy between the behaviors of the R band and K' band light curves, suggested by Rhoads and Fruchter, is explained as the result of the lack of data for $t>7.18$ days $(\\lesssim t_{jet})$ at K', and the existence of achromatic substructure from fluctuations in the radio and optical/IR regimes. The value for the break time from the global, self-consistent approach we have used is $t_{jet}=7.5$ days at all frequencies. The long-lived radio emission from the burst, spanning a large range in frequency and time, plays a significant role in our ability to extract the time evolution of $\\nu_a$, $\\nu_m$, and $F_{\\nu,max}$ from the data. In the case of this GRB in particular, the large range in frequency and time is crucial, since it serves to reduce the effects of unexplained deviations from the simple theory, such as the short-timescale fluctuations in the optical bands, on the overall evolution of the fireball. We end with some words of caution. In our analysis we assumed that $\\nu_c$ lies above the optical band throughout the evolution of the fireball, and successfully got a reasonable fit to the data. However, it is possible that another set of parameters, with $\\nu_c$ below the optical band, can fit the data equally well. Preliminary work in this direction indicates that the gross features of the fireball evolution (e.g. a break time of $\\sim$7 days) remain unaltered." }, "0005/astro-ph0005114_arXiv.txt": { "abstract": "The low second acoustic peak in the recent Boomerang data may indicate a cosmological baryon density which is larger than allowed by standard big bang nucleosynthesis. We show that the decay of the tau-neutrino: $\\nu_\\tau \\rightarrow \\nu_e + \\phi$, where $\\nu_e$ is the electron neutrino and $\\phi$ is a scalar, essentially can assure agreement between BBN calculations and light element observations for a large baryon density. ", "introduction": "The Boomerang experiment recently measured the angular power spectrum of the cosmic microwave background up to $l=600$~\\cite{boom}. Accordingly to what is expected for a flat Universe, the data show a peak in the power spectrum at $l=197\\pm6$. At the same time, however, the data seem to indicate that the second acoustic peak is rather low~\\cite{first}, which may be an indication of a high baryon number~\\cite{tegmark} (see also~\\cite{Hu:1995uz,white}) (say e.g. $\\Omega_b h^2 \\sim 0.03$, where $h$ is the Hubble constant in units $100 \\, {\\rm km \\, s}^{-1} \\, {\\rm Mpc}^{-1}$). This simple conclusion, however, immediately leads to disagreement with the well established Big Bang Nucleosynthesis (BBN), which predicts $ \\Omega_b h^2 \\approx 0.019 \\pm 0.0024$~\\cite{tytler}. Although this discrepancy is still very preliminary, it is interesting to investigate specific models which can reconcile BBN with a high baryon density. The problem is the following. In the standard BBN scenario there is only one free parameter, namely the baryon to photon number ratio, $\\eta = n_B/n_\\gamma$, which is related to the baryon density according to $\\eta\\simeq 2.68 \\cdot 10^{-8} \\Omega_b h^2 $. Observationally $\\eta_{10} = \\eta \\cdot 10^{10}$ has long been known to be in the interval $1$ {\\small $\\lapprox$} $\\eta _{10}$ {\\small $\\lapprox$} $10$, and the recent deuterium measurements favor $\\eta_{10} \\approx 5$. When one increases $\\eta_{10}$ the helium abundance increases slightly while the deuterium abundances decreases rapidly: \\be \\eta_{10} \\nearrow \\, \\, \\, \\, {\\rm implies} \\, \\, \\, \\, Y_{he} \\nearrow \\, \\, \\, \\, {\\rm and} \\, \\, \\, \\, D/H \\searrow \\ee It is then clear that a high baryon number leads to a too low deuterium prediction and a too high helium-4 value. Now, with a lower deuterium abundance one must seek a method of increasing the deuterium, to achieve agreement with observations. This is easily found; one could simply increase the effective number of massless degrees of freedom. For two different value of $\\eta$ one can achieve the same deuterium abundance by varying the energy density (expressed through $\\Delta N_{eff} = N_{eff} - 3$). According to \\cite{franc}, the needed $\\Delta N_{eff}$ is roughly: \\be \\Delta N_{eff} \\approx \\frac{1}{0.03} \\cdot {\\rm log_{10}} \\left( \\frac{\\eta_2}{\\eta_1} \\right) \\label{EF} \\ee and we thus see, that if $\\eta_{10}$ is 7 (9) instead of 5, then we need additional energy density corresponding to 5 (8.5) extra neutrinos. Increasing the energy density (which effectively corresponds to adding new particles), however, also affects the helium abundance. An increase in $N_{eff}$ correspond to a variation $\\Delta Y_{he} \\approx 0.013\\Delta N_{eff}$ in the helium abundance and the observational data leave room for not more than one extra neutrino. We thus conclude that even if one can make deuterium calculations agree with observations by increasing $N_{eff}$, then the helium predictions will be in strong disagreement with observations. The goal of this Letter is to point out a specific model, which may solve this apparent problem. ", "conclusions": "We have shown, how a specific model with a decaying massive tau-neutrino can make BBN calculations for the light element abundances agree with the observations in a Universe with as high baryon number as $\\eta_{10}=7$. Such a high baryon number may be needed to explain a lower second peak in the CMB power spectrum as seen in the recent Boomerang data. The scenario with a high $\\Omega_b$ naturally predicts a high $3^{\\rm rd}$ peak, and can hence easily be excluded by future CMB observations. On the other hand, should the future CMB experiments find a high $3^{\\rm rd}$ peak, then one must distinguish between the various models. One can distinguish a high $\\Omega_b$ scenario from the delayed recombination picture suggested in ref.~\\cite{peebles}, since high $\\Omega_b$ will lower the diffusion damping, and hence the $3^{\\rm rd}$ peak should be higher in this case than in the models proposed in~\\cite{peebles}. Distinguishing between a decaying neutrino and a chemical potential~\\cite{julien} is more difficult, and one would probably need refined observations of other light elements like helium-3 and lithium." }, "0005/astro-ph0005322_arXiv.txt": { "abstract": "Using realistic hydrodynamical simulations of the solar surface convection as 3D, time-dependent, inhomogeneous model atmospheres, the solar photospheric Si abundance has been determined to be log\\,$\\epsilon_{\\rm Si} = 7.51\\pm0.04$. This constitutes a difference of 0.04\\,dex compared with previous estimates based on the 1D Holweger-M\\\"uller (1974) model, of which half is attributable to the adopted model atmosphere and the remaining part to the improved quantum mechanical broadening treatment. As a consequence, all meteoritic abundances should be adjusted downwards by the same amount. In particular the meteoritic Fe abundance will be log\\,$\\epsilon_{\\rm Fe} = 7.46\\pm0.01$, in good agreement with the recently determined photospheric Fe abundance (Asplund et al. 2000b). The existing uncertainties unfortunately prevent an observational confirmation of the postulated effects of elemental migration of metals in the Sun. ", "introduction": "} To compare the meteoritic and solar photospheric chemical compositions it is necessary to have a common reference element. The honour is normally given to Si, which is an abundant element and a natural choice to measure other elemental abundances against in meteorites considering the volatility of hydrogen. In order to place all other meteoritic abundances on an absolute scale to confront with the photospheric abundances, one must therefore accurately know the photospheric Si/H ratio. As a consequence all meteoritic absolute abundances will depend on the measured solar photospheric Si abundance. In practice, one often utilizes additional elements besides Si to anchor the two scales more firmly to each other (e.g. Anders \\& Grevesse 1989). Several determinations of the solar photospheric Si abundance exist in the literature (e.g. Holweger 1973; Lambert \\& Luck 1978; Becker et al. 1980) using the Holweger-M\\\"uller semi-empirical model atmosphere (Holweger \\& M\\\"uller 1974) and the use of equivalent widths and the microturbulence concept. Given the recent progress in the construction of self-consistent, 3D, hydrodynamical simulations of the solar surface convection (e.g. Stein \\& Nordlund 1989, 1998; Asplund et al. 2000a,b, hereafter Paper I and II) and the significance of the solar photospheric Si abundance, a re-analysis seemed warranted, which is presented here. These convection simulations have now reached a very high degree of realism. Without relying on any free parameters, such as the mixing length parameters, the simulations successfully reproduce the granulation topology and statistics (Stein \\& Nordlund 1998), constraints from helioseismology (Rosenthal et al. 1999) and detailed spectral line shapes, shifts and asymmetries (Paper I and II). In particular, line profiles are accurately described without invoking any micro- and macroturbulence which are necessary in classical 1D spectral analyses. As a consequence several possible sources of uncertainties (e.g. model atmosphere limitations, equivalent widths and microturbulence) can be removed when deriving elemental abundances, which should result in more secure determinations. ", "conclusions": "" }, "0005/astro-ph0005052_arXiv.txt": { "abstract": "We present broadband and narrowband {\\it Hubble Space Telescope} images of the central region of the heavily disturbed Virgo cluster galaxy NGC~4438 (Arp120), whose nucleus has been described as a type 1 LINER or dwarf Seyfert. Narrowband H$\\alpha$ and [NII] HST images reveal striking bipolar shell features, 1 kpc in projected length from end-to-end, which are likely the result of an outflow from the nuclear region experiencing a strong interaction with the ISM. While these outflow shells share similarities with those in some other starburst or AGN galaxies, these in NGC~4438 are notable because NGC~4438 harbors neither a luminous circumnuclear starburst nor a luminous AGN. The shells appear to be closed at their outer ends, suggesting that the outflow in NGC~4438 is dynamically younger than those in some other galaxies. The radio continuum emission is strongly enhanced near the outer ends of the shells, suggesting working surfaces arising from collimated nuclear outflows which have impacted and shocked the surrounding ISM. The 2 shells are quite different, as the northwestern (NW) shell is luminous and compact, while the southeastern (SE) shell is 2.5 times longer and much fainter, in both optical emission lines and the non-thermal radio continuum. The differences between the 2 shells may be attributed to a difference in ISM density on the 2 sides of the nuclear disk. Such an ISM asymmetry exists on larger scales in this heavily disturbed galaxy. At the base of the outflow is a nuclear source, which is the highest surface brightness source in the galaxy at optical wavelengths, This source is resolved with a FWHM=0.3$''$=25 pc and has modest luminosities, uncorrected for extinction, of 5$\\times$10$^{38}$ erg s$^{-1}$ in H$\\alpha$, and M$_{\\rm B}$=-13. We discuss whether the outflow is powered by a low luminosity AGN or a compact nuclear starburst. The kinetic energy associated with the ionized gas in the shells is $\\sim$10$^{53}$ ergs, which could be produced either by massive star formation or an AGN. While the NW shell, which contributes most of the flux in most ground-based apertures centered on the nucleus, exhibits LINER-type line ratios, the nucleus has an H$\\alpha$/[NII] ratio consistent with an HII region. Although there appears to be very little massive star formation occuring in the central kpc, the nucleus may contain a partially obscured, young nuclear star cluster. On the other hand, the collimation of the southeastern shell, and the strongly enhanced radio continuum emission at the outer ends of the shells are more easily explained by jets associated with a nuclear black hole than a compact nuclear starburst, although there is no direct evidence of a jet. The 2000 km s$^{-1}$ broad line component could be due to an AGN broad line region but might also be due to high velocity gas in the outflow. Since NGC~4438 has a large bulge, a large nuclear black hole might be expected. ", "introduction": "Ultimately we wish to know nature of the nuclear source and whether the outflow is driven by AGN or starburst. The present study has not given definite answers for these questions, and future studies will help provide the answers. High spatial resolution optical or infrared spectroscopy of the nucleus could show the spatial distribution of the 2000 km s$^{-1}$ broad line component, revealing whether it originates in the outflowing gas or is a true AGN broad line component. It could also show whether there is a very broad line component on sub-arcsecond scales possibly arising from an accretion disk, as in the Virgo LINER NGC~4579 (Barth \\etal 2001). Studies of the stellar kinematics from absorption lines can constrain the mass of a central black hole and show whether the outflow originates from the dynamical center of the galaxy. IR imaging and spectroscopy can provide better extinction estimates, further revealing the true nature of the nucleus, and yielding new information on the spatial extent and intensity of the central star formation activity. High resolution x-ray observations (Chandra) can show how much of the central 10$''$ Einstein and ROSAT x-ray source is associated with the nucleus, and how much with the outflow. The x-ray spectrum of the nuclear source can reveal the true nature of the nucleus, whereas that of the outflow can reveal whether the hot gas has sufficient energy to be dynamically important for the evolution of the bubble. In M82 the high temperature (4$\\times$10$^7$ K) thermal x-rays may contain enough pressure to drive the outflow (Griffiths \\etal 2000). High resolution molecular gas (e.g. CO) observations may be important for learning the energy budget of bubbles, since molecular gas may be participating in the outflow, and could be carrying a significant amount of kinetic energy. Most of the kinetic energy associated with a starburst superbubble in M82 appears to be in the form of molecular rather than ionized gas (Matsushita \\etal 2000), and it will be interesting to learn whether the bubbles in NGC~4438 are similar. We find it remarkable that there is a such a striking outflow from a nucleus without a very luminous starburst or AGN, and think followup studies of NGC~4438 and similar galaxies are important because of the possible implications for the deposition of matter and energy into galactic haloes and the intergalactic medium. ", "conclusions": "" }, "0005/astro-ph0005264_arXiv.txt": { "abstract": "This paper presents an interpretation based on gamma-ray burst source kinematics for the relationship found by Norris et al.~between peak luminosity and energy-dependent pulse lag. I argue that the correlation should instead be between {\\it number} luminosity and pulse lag. This interpretation improves the least-squares fit of this correlation for the known bursts by 25 percent or more. It also suggests a distance estimation scheme. I propose that this relationship is due to the variation in line-of-sight velocity among bursts. This interpretation allows one to speculate on the range of gamma-ray burst expansion velocities or the size of their jet opening angles. ", "introduction": "It has recently been reported by \\citet{nmb00,nmb00b} that there is a relationship between the peak luminosity of gamma-ray bursts (GRB) and the pulse time lag between BATSE energy channels. Specifically, they find that the peak luminosity $L_{pk}$ is anticorrelated with the duration of the lag, $\\Delta t$, according to $L_{pk} \\approx 1.3 \\times (\\Delta t/0.01 \\text{ sec})^{-1.15} \\times 10^{53}$ ergs/sec . This intriguing relationship is useful because it probes the properties of an ensemble of bursts and thereby can provide clues of the global dynamics of GRBs. It also has the potential for distance estimation to GRBs. Several workers have attempted to explain the origin of the pulse lag, or so called pulse paradigm \\citep{nnb+96}. \\citet{dm98} were able to produce a pulse lag by modeling the burst as a series of internal shocks, but obtained timescales an order of magnitude too large. However, as \\citet{wf00} point out, the lag timescale is much longer than the synchrotron cooling time. \\citet{pm98} modeled the kinematics of internal shocks and found no evidence of pulse peak lag. As of yet, the source of the pulse paradigm remains a mystery. This paper does not attempt to explain the origin of the pulse lag, but assumes that it derives from a process common to all bursts and has some proper decay timescale $\\Delta t'$ in the reference frame of the emitter. In this paper I will present an interpretation of the luminosity-lag relationship as being due to relativistic motion of the emitting region toward the observer. Thus it is a purely kinematic effect. ", "conclusions": "\\label{sectionbeaming} In this letter I have argued that the correlation of \\citet{nmb00} is between number luminosity and spectral lag and I interpret this as being due to the variety, among bursts, of relativistic velocities at which emitting regions move toward the observer. Suppositions as to the physical emission process have been deliberately avoided in order to highlight this purely kinematic effect. There are two possible scenarios by which this might happen. The first is simply that the variation is due to the variation of expansion Lorentz factor among bursts. From Eqn (\\ref{dtobseqn}), this implies that the ratio of maximum to minimum spectral lags is the same as the ratio of maximum to minumum Lorentz factors \\begin{equation} \\frac{\\gamma_{max}}{\\gamma_{min}} = {\\frac{\\Delta t_{max}}{\\Delta t_{min}}} \\gtrsim 100 \\end{equation} where observed lags are seen to range over roughly two orders of magnitude. Thus if the fastest bursts have $\\gamma_{max} \\sim 100$, then the minimum bursts would have $\\gamma_{min} \\sim 1$ and thus may be only mildly relativistic. The second possibility is that GRB ejecta is directed in a jet such that the Lorentz factor $\\gamma(\\theta)$ has some maximum $\\gamma(\\theta = 0) = \\gamma_{max}$ and monotonically decreases with increasing angle. Thus the function $\\gamma(\\theta)$ will determine the relative numbers of bursts observed with given spectral lags. \\citet{band97} has noted that the distribution of bursts is strongly peaked at small lags indicating that the Lorentz factor of emitting ejecta, $\\gamma(\\theta < \\theta_0/2) \\sim \\gamma_{max}$, is basically constant over the jet opening angle $\\theta_0$, with a fairly narrow edge region $\\theta > \\theta_0$ of decreasing $\\gamma$. Thus let us estimate a characteristic opening angle $\\theta_0$ by assuming that $\\gamma = \\gamma_{max}$ is constant for $\\theta < \\theta_0/2$ and decreases to order unity within a narrow boundary or ``edge'' region $0 < \\theta - \\theta_0/2 < 1/\\gamma_{max}$. Thus the solid angle subtended by ``face'' bursts (i.e. observed at inclinations $\\theta < \\theta_0/2$) is $\\pi (\\theta_0/2)^2$, and that for ``edge'' bursts is $\\pi \\theta_0/\\gamma_{max}$ and the ratio is: \\begin{equation} \\frac{\\text{``edge'' bursts}}{\\text{``face'' bursts}} = \\frac{4}{\\gamma_{max} \\theta_0} ~. \\end{equation} From Fig. 3 of \\citet{nmb00}, one can estimate this ratio, by defining short lags $< 0.1$ sec as ``face'' and the longer lags as ``edge'' bursts, to be $\\sim 2/5$. Assuming $\\gamma_{max} = 100$, then this gives $5^o < \\theta_0 < 10^o$. With more statistics of the distribution of bursts along the curve of Eqn. (\\ref{nteqn}) one can fit the function $\\gamma(\\theta)$ much more precisely. This work was performed under the auspices of the U.S. Department of Energy by University of California Lawrence Livermore National Laboratory under contract W-7405-ENG-48.\\\\" }, "0005/astro-ph0005523.txt": { "abstract": "Using a non-linear Monte-Carlo code we investigate the radiative response of an accretion disk corona system to static homogeneous flares. We model a flare by a rapid (comparable to the light crossing time) energy dissipation in the corona or the disk. If the flares originate from the disk, the coronal response to the soft photon shots produces a strongly non-linear Comptonised radiation output, with complex correlation/anti-correlations between energy bands. This behavior strongly differs from those found with usual linear calculations. Thus any model for the rapid aperiodic variability of X-ray binaries invoking a varying soft photon input as a source for the hard X-ray variability has to take into account the coronal temperature response. On the other hand, if the flare is due to a violent heating of the corona, when the perturbation time scale is of the order of a few corona light crossing times, the shot spectrum evolves from hard to soft. This general trend is independent of the shot profile and geometry. We show that for short dissipation time, the time averaged spectra are generally harder than in steady state situation. In addition, annihilation line and high energy tails can be produced without need for non-thermal processes. ", "introduction": "The hard X/$\\gamma$-ray spectra of galactic black hole candidates (GBHC) in their low hard state as well as Seyfert galaxies can be generally represented as a sum of a hard power-law continuum with a cutoff at a few hundred keV and a Compton reflection bump (with a Fe K$\\alpha$ line at $\\sim 6.4$ keV) produced when high energy photons interact with cold material (Zdziarski et al. 1996\\nocite{1996A&AS..120C.553Z}). The presence of the Compton reflection component (Guilbert \\& Rees 1988\\nocite{1988MNRAS.233..475G}; George \\& Fabian 1991\\nocite{1991MNRAS.249..352G}) implies that cold material could be present in the direct vicinity of the X/$\\gamma$-ray producing region. A soft excess present below $\\sim 1$ keV is usually associated with the thermal emission from the cold accretion disk and can be powered by the viscous dissipation in the disk itself as well as by reprocessing hard photons. These three components are generally interpreted in the framework of accretion disk corona models. These models assume that the soft thermal radiation is emitted by the disk and then Comptonised in a very hot plasma, the ``corona''. The reflection features arise naturally from the disk illumination. The nature and geometry of the corona are unclear. Several geometries have been proposed: slab sandwich-like corona (Haardt \\& Maraschi, 1991, 1993)\\nocite{1993ApJ...413..507H}\\nocite{1991ApJ...380L..51H}, localized active regions on the disk surface powered by magnetic reconnections (Liang et al. 1977; \\nocite{1977ApJ...218..247L}Galeev et al. 1979; \\nocite{1979ApJ...229..318G} Haardt et al. 1994\\nocite{1994ApJ...432L..95H}) or a hot accretion disk in the center surrounded by a cold standard disk (Shapiro et al. 1976 \\nocite{1976ApJ...204..187S}; Ichimaru 1977\\nocite{1977ApJ...214..840I}; Narayan \\& Yi 1994\\nocite{1994ApJ...428L..13N}). Unfortunately, spectroscopy alone does not allow one to firmly probe either the geometry or the way the corona is powered. X-ray variability studies seem to be a key for the understanding of accretion processes around compact objects. They should at least bring indications on the nature and structure of the corona. The short term variability of black hole X-ray binaries is now well known thanks to an impressive amount of observational data accumulated since many years by several space experiments (Cui 1999)\\nocite{1999hepa.conf...97C}. There is however no accepted model that accounts for most of the observational data. The inner disk dynamic predicts important variability at kHz frequencies while a stronger variability is observed around 1~Hz. The power density spectrum is roughly a power-law of index between -1.0 and -2.0. Another important feature of the temporal behavior is the time lags between hard and soft photons that depend on Fourier frequency $\\propto f^{-1}$. Whether the intrinsic source of the variability is the disk (Payne 1980\\nocite{1980ApJ...237..951P}; Miyamoto et al. 1988\\nocite{1988Natur.336..450M}) or the corona (Haardt et al.~1997\\nocite{1997ApJ...476..620H}, Poutanen \\& Fabian 1999) is still a matter of debate. As the characteristic time scales are expected to grow linearly with the black hole mass, the time scales are far longer in Seyfert galaxies. The long observations required prevent the acquisition of as many data as for X-ray binaries. Their Power Density Spectra (PDS), at least, are similar to those of stellar black holes, modulo the mass scale factor (Edelson \\& Nandra 1999). To understand these temporal characteristics it seems important to introduce the temporal dimension in spectral models. The most important difficulty when dealing with time dependent problems is that the system is not necessarily in radiative equilibrium. The corona is strongly coupled with radiation, its physical parameters such as temperature and optical depth can fluctuate in response to changes in the photon field. These variations in turn influence the radiative field, making the system strongly non-linear. The dynamics of compact plasma has been extensively studied during the eighties in the context of the models for the high energy emission of active galactic nuclei; first using analytical arguments (Guilbert et al. 1983\\nocite{1983MNRAS.205..593G}), then using more and more accurate numerical methods (Guilbert \\& Stepney 1985;\\nocite{1985MNRAS.212..523G} Fabian et al. 1986\\nocite{1986MNRAS.221..931F}; Kusunose 1987;\\nocite{1987ApJ...321..186K} Done \\& Fabian 1989\\nocite{1989MNRAS.240...81D}). The most detailed treatment of micro-physics was achieved by Coppi (1992) \\nocite{1992MNRAS.258..657C} using a method based on the solution of the kinetic equations. These studies gave an understanding of the behaviour of the plasma and the spectral evolution of the emitted radiation, when the input parameters (heating, external soft-photons injection...) vary on time scales of the order of the light crossing time. However, at the time of those studies the importance of the coupling with cold matter did not appear as crucial as it does now. Indeed, in the accretion disk corona framework, another complication appears: the hard X-ray radiation is produced by Comptonisation of soft photons that, in turn, can be mostly produced by reprocessing the same hard radiation in the cold accretion disk (the ``feedback'' mechanism). Until now, in most studies only steady-state situations have been considered. The physical characteristics of the emitting region (such as temperature and optical depth of the Comptonising cloud) that determine the observed X/$\\gamma$-ray spectrum are assumed not to vary in time (e.g. Sunayev \\& Titarchuk 1980)\\nocite{1980A&A....86..121S}. The most detailed calculations considered only steady states, where the temperature and the optical depth are defined by the energy balance and electron-positron pair balance, assuming a constant heating (Haardt \\& Maraschi 1993; Stern et al. 1995a,b\\nocite{1995ApJ...449L..13S}\\nocite{1995MNRAS.272..291S}; Poutanen \\& Svensson 1996\\nocite{1996ApJ...470..249P}). The observed rapid spectral changes imply the presence of rapid changes in the physical conditions of the source. When taken into account in radiative transfer modeling, these changes have been considered as a succession of steady state equilibria (Haardt et al. 1997;\\nocite{1994ApJ...432L..95H} Poutanen \\& Fabian 1999). This approximation is acceptable as long as the underlying perturbation evolves on time scales, $t_{\\mathrm{c}}$, far larger than the light crossing time. Actually, we do not know if this assumption is valid. Here we aim at giving a first look at the rapid spectral and temporal evolution of a hot plasma coupled with a reprocessor. In this first attempt to model the non-linear behavior of a time dependent accretion disk corona system, we try to point out the main properties of the accretion disk corona when the dissipation parameters change on time scales of the order of the light crossing time of the corona. We show that the non-linear Monte Carlo method (Stern et al. 1995a) can be an efficient tool to perform the task of computing the time evolution of the plasma parameters together with a detailed radiative transfer treatment enabling the production of light curves. \\nocite{1999ApJ...514..682E} Our model assumptions are presented in Sect.~\\ref{assumptions}. Section ~\\ref{montecarlo} gives a description of our computational method. The different consistency tests that we performed in order to check the validity of the code for steady state situations are presented in Sect.~\\ref{comparisons}. Then we present some applications to time dependent situations. We first investigate the case of an equilibrium modified by a variation in the soft photon input, in Sect.~ \\ref{variabilitydrivenbythedisk}. In Sect.~\\ref{variabilitydrivenbythecorona}, we then consider situations where the energy dissipation in the corona occurs during short flares. ", "conclusions": "The NLMC method has significantly contributed to the study of the spectral properties of a steady corona radiatively coupled with an accretion disk (e.g. Stern et al. 1995b\\nocite{1995ApJ...449L..13S}). However, in order to understand the short term X-ray variability of accreting black hole sources, it seems necessary to take into account the dynamical aspect of the coupling. We have shown that, here again, the NLMC method can be an efficient tool. Our code is able to deal with different situations in which the disk-corona equilibrium is perturbed by a violent energy dissipation in the disk or the corona. The few examples given here are far from being a definitive study of these problems. However, they enable us to outline some important general properties. On the one hand, we showed that models invoking a variability in the injection of soft seed photons as the origin of hard X-ray variability have to take into account the response of the corona to such fluctuations. Indeed, the corona is quickly Compton cooled when the soft photons flux increases. As long as the coronal heating is kept constant, the whole luminosity of the Comptonised radiation is constant, even if the thermal emission is strongly variable. There is however an important spectral evolution, leading to complex correlations between different energy bands. Details of these correlations depend on the intrinsic soft flux variability. On the other hand, if the variability arises from dissipation in the corona, the quasi-static approximation is valid as long as the dissipation time scale is far larger than the corona light crossing time. If this is not the case the feedback from the disk leads to a hard-to-soft spectral evolution. Such models thus predict soft lags at high fourier frequencies ($> \\sim$ 150 Hz). We also showed that a short dissipation time scale produces harder spectra than steady state dissipation. Moreover, in such conditions, a pure thermal model can produce spectral features such as high-energy tails or annihilation lines which are generally considered as the signature of non-thermal processes. The present work could be developed in numerous ways. For example, as the disk and corona may be coupled by some physical mechanism, it is likely that the observed variability originates from nearly simultaneous perturbations of the disk and the corona. It is also not very realistic to consider that the dissipation occurs both homogeneously and instantaneously in the corona or the disk: effects of propagation should be introduced. Other complications may arise, such as bulk motions or modifications of the geometry of the emitting region during a flare. The studies of the individual flare evolution should provide predictions for the lag inversion frequencies that can be used to put constraints on the geometry of the sources. Then, it would be interesting to investigate different stochastic models describing the interaction between flares. Taken together with the evolution of individual flares it would make it possible to generate the light curves in different energy bands, and compute the time-averaged energy spectra and various temporal characteristics such as power spectral density, time/phase lags between different energy bands, cross-correlation functions, coherence function, and compare them with the observations." }, "0005/astro-ph0005502_arXiv.txt": { "abstract": "We give a short description of different methods used in cosmology. The focus is on major features of $N$-body simulations: equations, main numerical techniques, effects of resolution, and methods of halo identification. ", "introduction": "Numerical simulations in cosmology have a long history and numerous important applications. Different aspects of the simulations including history of the subject were reviewed recently by \\citet{Bert}; see also \\citet{Sellwood}. More detailed aspects of simulations were discussed by \\citet{Gelb}, \\citet{Gross}, and \\citet{Kravtsov}. Numerical simulations play a very significant role in cosmology. It all started in 60s \\citep{AarsethA} and 70s \\citep{Peebles, PressSchecter} with simple N-body problems solved using N-body codes with few hundred particles. Later the Particle-Particle code (direct summation of all two-body forces) was polished and brought to the state-of-art \\citep{AarsethB}. Already those early efforts brought some very valuable fruits. \\citet{Peebles} studied collapse of a cloud of particles as a model of cluster formation. The model had 300 points initially distributed within a sphere with no initial velocities. After the collapse and virialization the system looked like a cluster of galaxies. Those early simulations of cluster formation, though producing cluster-like objects, signaled the first problem -- simple model of initially isolated cloud (top-hat model) results in the density profile of the cluster which is way too steep (power-law slope -4) as compared with real clusters (slope -3). The problem was addressed by \\citet{GunnGott}, who introduced a notion of secondary infall in an effort to solve the problem. Another keystone work of those times is the paper by \\citet{WhiteA}, who studied collapse of 700 particles with different masses. It was shown that if one distributes the mass of a cluster to individual galaxies, two-body scattering will result in mass segregation not compatible with observed clusters. This was another manifestation of the dark matter in clusters. This time it was shown that inside a cluster the dark matter can not reside inside individual galaxies. Survival of substructures in galaxy clusters was another problem addressed in the paper. It was found that lumps of dark matter, which in real life may represent galaxies, do not survive in dense environment of galaxy clusters. \\citet{WhiteRees} argued that the real galaxies survive inside clusters because of energy dissipation by the baryonic component. That point of view was accepted for almost 20 years. Only recently it was shown the energy dissipation probably does not play a dominant role in survival of galaxies and the dark matter halos are not destroyed by tidal stripping and galaxy-galaxy collisions inside clusters \\citep{KGKK, Ghigna99}. The reason why early simulations came to a wrong result was pure numerical: they did not have enough resolution. But 20 years ago it was physically impossible to make a simulation with sufficient resolution. Even if at that time we had present-day codes, it would have taken about 600 years to make one run. Generation of initial condition with given amplitude and spectrum of fluctuations was a problem for some time. The only correctly simulated spectrum was the flat spectrum which was generated by randomly distributing particles. In order to generate fluctuations with power spectrum, say $P(k) \\propto k^{-1}$, \\citet{AarsethGT} placed particles along rods. Formally, it generates the spectrum, but the distribution has nothing to do with cosmological fluctuations. \\citet{Doroshkevich} and \\citet{KlypinShandarin} were the first to use the \\citet{Zeldovich} approximation to set initial conditions. Since then this method is used to generate initial conditions for arbitrary initial spectrum of perturbations. Starting mid 80s the field of numerical simulations is blooming: new numerical techniques are invented, old ones are perfected. The number of publications based on numerical modeling skyrocketed. To large extend, this have changed our way of doing cosmology. Instead of questionable assumptions and waving-hands arguments, we have tools of testing our hypothesis and models. As an example, I mention two analytical approximations which were validated by numerical simulations. The importance of both approximations is difficult to overestimate. The first is the Zeldovich approximation, which paved the way for understanding the large-scale structure of the galaxy distribution. The second is the \\citet{PressSchecter} approximation, which gives the number of objects formed at different scales at different epochs. Both approximations cannot be formally proved. The Zeldovich approximation formally is not applicable for hierarchical clustering. It must start with smooth perturbations (truncated spectrum). Nevertheless, numerical simulations have shown that even for the hierarchical clustering the approximation can be used with appropriate filtering of initial spectrum \\citep[see][and references therein]{SahniColes}. The Press-Schechter approximation is also difficult to justify without numerical simulations. It operates with the initial spectrum and the linear theory, but then (a very long jump) it predicts the number of objects at very nonlinear stage. Because it is not based on any realistic theory of nonlinear evolution, it was an ingenious, but a wild guess. If anything, the approximation is based on a simple spherical top-hat model. But simulations show that objects do not form in this way -- they are formed in a complicated fashion through multiple mergers and accretion along filaments. Still this a very simple and a very useful prescription gives quite accurate predictions. This lecture is organized in the following way. Section 2 gives the equations which we solve to follow the evolution of initially small fluctuations. Initial conditions are discussed in section 3. A brief discussion of different methods is given in section 4. Effects of the resolution and some other technical details are also discussed in Section 5. Identification of halos (``galaxies'') is discussed in Section 6. ", "conclusions": "" }, "0005/astro-ph0005028_arXiv.txt": { "abstract": "The discoveries that GRBs have X-ray, optical and radio afterglows have connected the study of GRBs to the rest of astronomy, and revolutionized the field. In this review, I discuss the implications that the observation of these afterglows have for burst energies and luminosities, and for models of the bursts and their afterglows. I describe recent evidence linking the long, softer, smoother GRBs detected by BeppoSAX and core collapse supernovae. Finally, I summarize recent work showing that, if these GRBs are due to the collapse of massive stars, they may provide a powerful probe of the very high redshift universe. ", "introduction": "Gamma-ray bursts (GRBs) were discovered serendipitously more than a quarter century ago~\\cite{kleb73}. The bursts consist of short, intense episodes of gamma-ray emission, lasting anywhere from $\\sim 10^{-2}$ seconds to $\\sim 10^3$ seconds. The time histories of GRBs are diverse, as Figure 1 illustrates. However, studies of the time histories have shown that the bursts can be separated into two classes: short, harder, more variable bursts; and long, softer, smoother bursts~\\cite{lamb93,kouveliotou93} (see Figure 2). Burst spectra are nonthermal; the photon number spectrum is a broken power law, with average slopes $\\sim -1.5$ and $\\sim -2.5$ at low and high energies and a shoulder at 100 keV - 1 Mev~\\cite{band93,schaefer94}. The data gathered by the Burst and Transient Source Experiment (BATSE) on the {\\it Compton} Gamma-Ray Observatory confirmed earlier evidence of a rollover in the cumulative brightness distribution of GRBs, showing that the burst sources are inhomogeneously distributed in space~\\cite{meegan92}. The data also showed that the sky distribution of even faint bursts is consistent with isotropy~\\cite{meegan92} (see Figure 3). This combination of results implies that we are at, or near, the center of the spatial distribution of burst sources and that the intrinsic brightness and/or spatial density of the sources decreases with increasing distance from us. \\begin{figure}[t] \\hfil \\begin{minipage}[t]{2.5truein} \\mbox{}\\\\ \\psfig{file=grb_980920.ps,width=2.5truein,clip=} \\end{minipage} \\hfill \\begin{minipage}[t]{2.5truein} \\mbox{}\\\\ \\hfil \\psfig{file=grb_980923.ps,width=2.5truein,clip=} \\hfil \\vfil \\end{minipage} \\\\ \\begin{minipage}[t]{2.5truein} \\mbox{}\\\\ \\psfig{file=grb_981021.ps,width=2.5truein,clip=} \\end{minipage} \\hfill \\begin{minipage}[t]{2.5truein} \\mbox{}\\\\ \\psfig{file=grb_981203.ps,width=2.5truein,clip=} \\vfil \\end{minipage} \\caption{Four GRB time histories of GRBs 980920, 980923, 981021, and 981203. These four bursts occurred within about two months, illustrating the diversity of GRB time histories.~\\cite{paciesas99}} \\end{figure} \\begin{figure}[t] \\hfil \\psfig{file=two_classes.eps,width=5.12truein,clip=} \\hfil \\caption{Distribution of duration (as measured by the time interval containing 90\\% of the photon counts) versus spectral hardness (as measured by the ratio of fluence in the 50 - 100 keV and 100 - 300 keV energy bands) for bursts in the BATSE 4B catalog~\\cite{paciesas99} (diamonds), showing clear evidence for two classes of bursts: short, harder, more variable bursts; and long, softer, smoother bursts. Events detected by BeppoSAX (solid squares) belong to the latter class. From~\\cite{kulkarni00}.} \\end{figure} \\begin{figure}[b] \\hfil \\psfig{file=grb_sky.eps,width=5.12truein,clip=} \\hfil \\caption{The positions in Galactic coordinates of the GRBs in the BATSE 4B catalog~\\cite{paciesas99}, showing the isotropy of the burst sky distribution~\\cite{meegan92}.} \\end{figure} The BATSE results showed that the bursts cannot come from a population of neutron stars in a thick Galactic disk (as was previously thought) and spurred interest in the possibility that the sources of the bursts lie at cosmological distances. Yet the evidence remained circumstantial. Consequently, the GRB distance scale -- and even more, the nature of the burst sources -- was debatable~\\cite{lamb95,pac95}. The principal reason for the continuing uncertainty in the distance scale of the bursts was that no definite counterpart to any burst could be found at other wavelengths, despite intense efforts spanning more than two decades. Consequently, the study of GRBs was isolated from the rest of astronomy. Scientists studying them had only the laws of physics and the properties of the bursts themselves to guide them in attempting to solve the GRB mystery. \\begin{figure}[t] \\hfil \\psfig{file=bepposax_970228.ps,width=5truein,clip=} \\caption{BeppoSAX observations of the fading X-ray afterglow of GRB 970228. Left panel: MECS image on February 28. Right panel: a deeper MECS image on March 3~\\cite{costa97b}} \\hfil \\end{figure} \\begin{figure}[b] \\hfil \\psfig{file=wht_970228.ps,angle=270,width=5truein,clip=} \\hfil \\caption{William Herschel Telescope images taken on February 28 and March 8 of the fading optical afterglow of GRB 970228.~\\cite{paradijs97}} \\end{figure} \\begin{figure}[t] \\hfill \\begin{minipage}[t]{2.16truein} \\mbox{}\\\\ \\psfig{file=hst_stis_cls_970228.ps,width=2.16truein,clip=} \\end{minipage} \\hfill \\begin{minipage}[t]{2.66truein} \\mbox{}\\\\ \\caption{HST STIS image taken on September 4 of the optical afterglow of GRB 970228, which revealed the presence of a faint ($R = 25.5$) host galaxy. Both are clearly visible near the center of the WFPC2 image~\\cite{sahu97}.} \\end{minipage} \\end{figure} The relatively accurate (3') gamma-ray burst positions found using BeppoSAX, and disseminated within a day or so, revolutionized the field. They led to the remarkable discoveries that GRBs have X-ray~\\cite{costa97}, optical~\\cite{paradijs97} and radio~\\cite{frail97} afterglows, finally connecting the study of GRBs with the rest of astronomy. The breakthroughs in our understanding of GRBs made possible by these discoveries cannot be overstated. In this review, dedicated to the memory of my friend and colleague David Schramm, I describe these recent breakthroughs. I first relate the discovery of GRB X-ray, optical and radio afterglows. I then discuss the implications that the observations of these afterglows have for burst energies and luminosities, and for models of the bursts and their afterglows. I describe recent evidence linking GRBs and core collapse supernovae. Finally, I summarize recent work showing that, if GRBs are due to the collapse of massive stars, GRBs may provide a powerful probe of the very high redshift universe. ", "conclusions": "The discoveries that GRBs have X-ray, optical and radio afterglows have connected the study of GRBs to the rest of astronomy, and revolutionized the field. We now know that the sources of most (and perhaps all) GRBs lie at cosmological distances, and that the bursts are among the most energetic and luminous events in the universe. There is increasing indirect evidence that the long, softer, smoother GRBs detected by BeppoSAX are associated with the star-forming regions of galaxies, and tantalizing direct evidence that these bursts are connected to core collapse supernovae. If these GRBs are due to the collapse of massive stars, they may be a powerful probe of the very high redshift universe." }, "0005/astro-ph0005444_arXiv.txt": { "abstract": "We have examined the properties of shock waves in simulations of large scale structure formation. Two cosmological scenarios have been considered: a standard cold dark matter model with $\\Omega_M=1$ (SCDM) and a cold dark matter model with cosmological constant and $\\Omega_M+\\Omega_{\\Lambda}=1$ ($\\Lambda$CDM) having $\\Omega_\\Lambda=0.55$. Large-scale shocks result from accretion onto sheets, filaments and knots of mass distribution on a scale of order of $\\sim 5$\\hinv Mpc in both scenarios. Energetic motions, partly residual of past accretion processes and partly caused by current asymmetric inflow along filaments, end up generating additional shocks. These extend on a scale of order of $\\sim1$\\hinv Mpc and envelop and penetrate deep inside the clusters. Also collisions between substructures inside clusters form merger shocks. Consequently, the topology of the shocks is very complex and highly connected. During cosmic evolution the comoving shock surface density decreases, reflecting the ongoing structure merger process in both scenarios. Accretion shocks have very high Mach numbers, typically between 10 and a few $\\times 10^3$, when photo-heating of the pre-shock gas is not included. The characteristic shock velocity is of order $v_{sh}(z)=H(z)\\lambda_{NL}(z)$, where $\\lambda_{NL}(z)$ is the wavelength scale of the nonlinear perturbation at the given epoch. However, the Mach number for merger and flow shocks (which occur within clusters) is usually smaller, in the range $\\sim 3 - 10$, corresponding to the fact that the intracluster gas is hot (i.e., already shock heated). Statistical fits of shock velocities around clusters as a function of cluster temperature give power-law functions in accord with those predicted by one-dimensional solutions. On the other hand, a very different result is obtained for the shock radius, reflecting extremely complex shock structures surrounding clusters of galaxies in three-dimensional simulations. The amount of in-flowing kinetic energy across the shocks around clusters, which represents the power available for cosmic-ray acceleration, is comparable to the cluster X-ray luminosity emitted from a central region of radius 0.5 \\hinv Mpc. Considering their large size and long lifetimes, those shocks are potentially interesting sites for cosmic-ray acceleration, if modest magnetic fields exist within them. ", "introduction": "``Cosmic shock waves'', formed in the course of large-scale structure formation, can contribute important roles in cosmology. They include external {\\it accretion shocks} as well as {\\it merger} and {\\it flow shocks} internal to galaxy clusters. The pristine cosmic plasma accreting onto the large scale structure is deflected from the Hubble flow and first processed by {\\it accretion shocks} (see \\eg Ryu \\& Kang 1997b). Evidence for their existence might be inferred from the observation of hot gas in the intracluster medium (ICM). In the commonly accepted paradigm for structure formation in the universe, gas accreting onto cosmic filaments and clusters of galaxies (GCs) has a typical bulk velocity up to $\\sim {\\rm a~few}~10^3 \\kms$. This gas is then shock-heated to temperatures ranging from $10^5-10^7$ K in filaments and up to $10^7- 10^8$ K in GCs (\\eg Kang \\etal 1994a, KCOR94 hereafter; Cen \\& Ostriker 1994, CO94 hereafter; Cen \\& Ostriker 1999a). {\\it Merger shocks} are produced during the mergers of sub-structures within a galaxy cluster and propagate through the hot ICM. In addition, during such a process, accretion shocks associated with the merging units also propagate through the ICM. Together with merging shocks they form a complex structure that can survive for long times inside the ICM after the end of the merger, because of the continuous gas inflow through filaments and sheets. We refer to these as {\\it flow shocks} (see \\S \\ref{morp.s}). There is now substantial observational evidence for temperature structure in clusters due to internal shock waves; these are appear to be mostly produced by merger events (\\eg Markevitch \\etal 1998; Donnelly \\etal 1999 and references therein) although recently some evidence might have appeared for flow shocks, too (\\cite{e98}, \\S \\ref{dc.s}). Among other reasons for interest in cosmic shocks is their ability to efficiently accelerate particles to relativistic energies (\\cf Blandford \\& Ostriker 1978, 1980; and Jones \\& Ellison 1991 for a recent review of this subject). In fact, relativistic cosmic-ray (CR) electrons have been observed in GCs through their synchrotron emission (\\eg Kim \\etal 1989; Giovannini \\etal 1993; Deiss \\etal 1997). Extended sources of synchrotron radiation are commonly observed with spatial distribution similar to that of the thermal X-ray emission (see \\eg Liang 1999). Although the cooling time for such CR electrons is much shorter than the cluster ages, explicit signatures of particle aging are rare in the spectra of the observed sources. Since individual cluster galaxies are unlikely to replenish the ICM adequately with populations of relativistic particles, an efficient mechanism for extended particle acceleration is probably required to understand the CR electron replenishment. Moreover, recently the EUVE satellite has revealed that many clusters possess an excess of extreme ultra-violet (EUV) radiation compared to what is expected from the hot, thermal X-ray emitting ICM (\\eg Lieu \\etal 1996; Fabian 1996; Mittaz \\etal 1997; Kaastra 1998). Further evidence for nonthermal activity in the ICM comes from detection of radiation in excess to thermal emission in the hard X-ray band above $\\sim 10$ KeV (\\eg Henriksen 1998; Fusco-Femiano \\etal 1999; Valinia \\etal 1999; Sarazin 1999). The mechanism proposed for the origin of these components is the inverse-Compton (IC) scattering of cosmic microwave background photons by CR electrons, although it is not clear if the same electron population is responsible for producing both the EUV excess and the hard X-ray excess (Ensslin \\etal 1999). Such detections suggest the possibility that nonthermal activities in the ICM are much higher than previously expected (Sarazin \\& Lieu 1998; Lieu \\etal 1999). Cosmic shock waves should be capable of accelerating CRs electrons responsible for the above emissions. However, at the same time, CR protons are produced. It is possible then, although not established yet, that if the above interpretation for the nonthermal radiation is correct, the CR protons produced at these shocks and accumulated throughout the cosmological evolution could provide a substantial fraction of the total pressure in GCs (Sarazin \\& Lieu 1998; Lieu \\etal 1999). It is clear that if the CR pressure was ever comparable to the thermal pressure during the evolution of the universe, that would have a profound impact on cosmology. For instance, structure formation is heavily used as a probe for discriminating among cosmological models (\\eg Carlberg \\etal 1997; Bahcall \\& Fan 1998) and hydrostatic equilibrium of the thermal ICM gas in the potential well of the total cluster mass is commonly assumed in order to derive GC masses (\\eg White \\etal 1993; Evrard 1997). The presence of a nonthermal component obviously would alter the results in proportion to its relative importance. Furthermore some additional source of pressure is clearly required in GCs over that produced by adiabatic hydrodynamics both to produce the correct density profiles (\\eg Evrard 1990; Navarro, Frenk \\& White 1995) and to prevent catastrophic cooling flows (Suginohara \\& Ostriker 1999). Cosmic ray pressure in the inner parts of GCs may thus play a vital role in the hydrodynamic equilibrium of these systems. Accretion shocks were also proposed as sites for acceleration of high energy CRs, protons and heavy nuclei, up to $10^{18} - 10^{19}$eV (Kang \\etal 1996, 1997). In fact, given the large velocity of the accretion flows and the large size and long lifetimes of the associated shocks, such energies would be achievable through ``cross-field'' diffusion in perpendicular magnetohydrodynamic shocks. Magnetic fields in the ICM of GCs have been observed with strengths of the order of a few $\\times 0.1\\mu$G (\\eg Kim \\etal 1989; Fusco-Femiano \\etal 1999; Molendi \\etal 1999). Outside GCs constraints on rotational measure from quasars impose an upper limit of $\\sim$nG, based on the assumption of regularly alternating magnetic field (see \\eg Kronberg 1994). However, this limit can be shifted to higher values if a realistic distribution of magnetic field associated with cosmic structures is assumed. On such basis, Ryu \\etal (1998) and Blasi \\etal (1999) claimed a new {\\it upper limit} \\ltsima 1$\\mu$G, at least along cosmic structures. Cosmic shock waves could serve also as sites for the generation of weak seeds of magnetic field by the Biermann battery mechanism. It was proposed that these seeds could be amplified to strong magnetic field of up to $\\sim\\mu{\\rm G}$ in clusters if flows there can be described as Kolmogoroff turbulence (Kulsrud \\etal 1997). However, further development into coherent magnetic field is unclear, since there is as yet no detailed theory capable of describing this process (see, \\eg Chandran 1997). Additional roles which shock waves may play in cosmology have been explored by a number of authors (\\eg Ryu \\& Kang 1997a, RK97 hereafter; Quilis \\etal 1998). RK97 compared analytical self-similar solutions for cluster formation in the Einstein-de Sitter universe (Bertschinger 1985) with one-dimensional numerical simulations in low density universes with/without a cosmological constant ($0.1 < \\Omega_M < 1$), where the properties of the accretion flow are related with the cluster's mass, radius and temperature. The major conclusion was a possibly testable prediction about the difference in the accretion flow in different cosmological models. In particular, the accretion velocity onto clusters of a given mass or radius in low density universes is smaller by up to 45 \\% and 65 \\% respectively compared to that in the Einstein-de Sitter universe. In the present paper we focus on the quantitative properties of large scale shocks produced by gas during the formation of cosmic structures. For this, the simulation data described in KCOR94 and CO94 have been used. The roles played by those shocks, especially with regard to CR acceleration and magnetic field generation and their consequences on cosmology, will be studied in future work. Details of our data analysis are described in \\S \\ref{num2.s}. In particular, we computed the velocity, Mach number, radius of shocks and kinetic energy flow across them, which were not studied in previous works, and calculated their correlation with the cluster core temperature. The results are reported in \\S \\ref{res.s}. Finally, \\S \\ref{dc.s} concludes with a discussion. ", "conclusions": "\\label{dc.s} We have studied the properties of ``cosmic shock waves'' associated with the large scale structure of the universe in two different cosmological scenarios, namely SCDM and $\\Lambda$CDM. Such shocks reveal remarkable properties. In fact, hierarchical formation histories of GCs produce highly complex flows and shock structures, which extend over scales of several Mpc. In addition to accretion shocks (responsible for heating infalling gas) and merger shocks, {\\it flow} shocks also appear and propagate through the thermalized ICM, providing extra gas heating. It turns out that the morphology of shocks associated with a large scale structure is complex and irreducibly three-dimensional and spherical shapes are inadequate to their description. Only for the external accretion shocks, located far away from the cluster core, some form regularity is recovered. This is an important issue especially in perspective of those missions with the next generation of high resolution X-ray telescopes (Chandra and XMM) which are planning to detect shocks in the ICM. It is worth mentioning that Ensslin \\etal work (Ensslin \\etal 1998) might already provide observational evidence for the presence of flow shocks in cosmic structure. Their conclusions are based on the assumption that the observed radio emission is due to particles currently accelerated at shocks there and injected from a ``radio relic'', a remnant previously associated with some radio galaxies. For example, for 1253+275, they find a pre-shock gas temperature $T\\sim 0.5-1$ KeV which shows that this is not the case of an accretion shock but that of either a merger or a flow shock (propagating through the ICM). Since there is no evidence for a merging process in 1253+275, it must be a flow shock. In addition, from their reported data on the pressure jump we infer $M\\sim 3 - 4.2$, well in the expected range for the internal shocks shown in Fig. \\ref{mh.f} (bottom panel). Cosmic shocks are also ideal sites for particle acceleration. We have shown already in \\S \\ref{rel.p} that cosmic shocks provide enough power to produce copious CRs. The details of the produced populations will depend on the injection mechanism and scattering agent, \\ie the magnetic field and the diffusion properties. Here, we further stress the important role of merger and flow shocks. These shocks may be responsible, not only for acceleration of CRs out of the thermal pool of the ICM, but also for the re-acceleration of CRs produced at accretion shocks and/or ejecta from radio-galaxies, AGNs or normal galaxies. In addition, they could be crucial in terms of acceleration and transport of ultra high energy CRs, because the scattering mean free path of these particles is of the same order as the typical separation between accretion/merger/flow shocks in the ICM. As already pointed out at the beginning of this section, there seems to be solid foundation for the existence of such shocks. As described in \\S 1, the presence of relativistic CR electrons in GCs has been inferred through observations of diffuse synchrotron radiation from radio halos. In addition, there is evidence for excess of radiation in both EUV and hard X-ray bands, with respect to thermal emission. Although still of controversial interpretation, such excesses are probably due to IC emission of CR electrons scattering off cosmic microwave background photons. Published studies, however, reveal that for an accurate interpretation of the constraints from the combined non-thermal emission components, it is crucial to have a detailed depiction of the relative distribution of particles and magnetic fields (\\cite{elb99}). The proton component of CRs has not been directly observed. Nonetheless, given the estimates for the CR electron component, Lieu \\etal (1999) concluded that their contribution in terms of dynamical pressure in GCs could be comparable to the thermal gas. This is consistent with the estimate inferred in the previous section and, as already pointed out in the introduction, has important consequences for cosmology. We point out that cosmic shock waves have existed ever since nonlinear structure formation was initiated at high redshift. This was shown through Fig. \\ref{nmz.f} in \\S \\ref{nmz.s}. Therefore, the importance of nonthermal activities in the cosmic plasma traces back to early epochs. Collisions of CR proton in the ICM generate, however, a flux of gamma ray photons through the production and subsequent decay of neutral pions; as pointed out by Blasi (1999), such gamma ray flux seems to be only marginally compatible with the upper limits measured by EGRET for Coma and Virgo clusters. But, again, the spatial and spectral distribution of CR, both depending on the overall cosmological history of these particles, play a crucial role in the determination of the expected gamma ray flux. In any case, the advent of the new generation of $\\gamma$-ray facilities (GLASS, VERITAS) characterized by a much higher sensitivity (\\cf Blasi 1999 for more details) will definitely settle the issue. From this depiction it emerges the importance and the necessity to understand the role of CRs in cosmology. For this purpose we are developing numerical tools in order to treat consistently magnetic fields and CRs in numerical simulations of structure formation. Such tools, in fact, will allow us to follow explicitly the evolution of the magnetic field as well as the acceleration and transport of CRs. With this information we will be able to carry out very useful comparisons between numerical and observational results in various bands of the electromagnetic spectrum." }, "0005/astro-ph0005391_arXiv.txt": { "abstract": "We have obtained two deep exposures of the \\ro Oph cloud core region with the \\ROSAT {\\it High Resolution Imager}. The improved position accuracy (1$\\arcsec$--6$\\arcsec$) with respect to previous recent \\xr observations (\\ROSAT {\\it PSPC}, and {\\it ASCA}\\,) allows us to remove positional ambiguities for the detected sources. We also cross-correlate the X-ray positions with IR sources found in the {\\it ISOCAM} survey of the same region at 6.7 and 14.3\\,$\\mu$m, in addition to sources (optical and IR) known from ground-based observations, which are young stars (T~Tauri stars, with and without circumstellar disks, and protostars). We thus obtain the best-studied sample of X-ray emitting stars in a star-forming region (63 \\xr sources detected, and 55 identified). We find that there is no statistically significant difference between the X-ray luminosity functions of {\\it HRI}-detected Class~II and Class~III sources, i.e., T~Tauri stars with and without disks, confirming that the contribution of these disks to X-ray emission (for instance by magnetic reconnection between the star and the disk), or to X-ray absorption, must be small. X-ray variability of T~Tauri stars can be studied by comparing the \\HRI data with the previously obtained \\PSPC data, but also using the fact that some \\HRI observations were done at different epochs. The resulting statistics show that most of the sources are variable, and that their variability is consistent with a solar-like (hence magnetic) flare origin. We use the information given both by the \\ISOCAM survey and by our \\HRI deep exposure to study the T~Tauri star population of the \\ro Oph dense cores. We confirm that essentially all Class~II and Class~III sources (embedded T~Tauri stars) are X-ray emitters, and that a strong correlation exists between their X-ray luminosity, $L_{\\mathrm X}$, and their stellar luminosity, $L_\\star$, with $L_{\\mathrm X}/L_\\star \\sim 10^{-4}$. Most of the new \\ISOCAM Class~II sources are not detected, however, which we explain by the fact that their X-ray luminosities ``predicted'' on the basis of this correlation are too faint to be detected by the {\\it HRI}. We predict that $\\sim$40 unknown faint or embedded Class~III sources remain to be discovered in X-rays in the {\\it HRI/ISOCAM} overlapping area, down to a limit of L$_{\\mathrm{X}} \\sim 3 \\times 10^{28}$\\,erg\\,s$^{-1}$. We show that the bulk of these unknown Class~III sources should be made of low- to very low-mass stars (M$_\\star <0.1$--0.6\\,M$_\\odot$). Prospects for future detections with {\\it XMM-Newton} and {\\it Chandra} are discussed. ", "introduction": "The \\ro Ophiuchi dark cloud complex is one of the nearest active site of low-mass star formation (see Wilking \\cite{wilking92} for a review). It is composed of two main dark clouds, L1688 and L1689, from which filamentary dark clouds, called streamers, extend to the north-east over tens of parsecs (e.g., Loren \\cite{loren89}; de Geus et~al. \\cite{degeus90}). The main star formation activity is observed in the westernmost dark cloud, L1688, which shows a rich cluster of low mass young stellar objects (YSO) around two dense molecular cores, ``core~A'' and ``core~F'' in the terminology of Loren (\\cite{loren89}) and Loren et~al. (\\cite{loren90}). The distance to the molecular complex remains somewhat controversial (see Wilking \\cite{wilking92}), with a usually adopted distance $d \\sim 160$ pc from the Sun. From {\\it Hipparcos} parallaxes and Tycho B--V colors of classes V and III stars, Knude \\& H$\\o$g (\\cite{knude98}) have detected at $d = 120$\\,pc an abrupt rise of the reddening as expected from a molecular cloud. Based on the {\\it Hipparcos} positions, proper motions, and parallaxes, de Zeeuw et~al. (\\cite{dezeeuw99}) gives $d = 145 \\pm 2$\\,pc for the mean distance of the Upper Scorpius OB association. We adopt $d =140$\\,pc in this article, instead of 160\\,pc used in our previous work. From infrared (IR) observations of star-forming regions, Lada and collaborators (e.g. Lada \\cite{lada87}; Wilking, Lada, \\& Wilking \\cite{wilking89}, hereafter WLY) introduced an IR classification and distinguished different stages of evolution of young stellar objects (YSO). This classification was subsequently revisited by Andr{\\'e} \\& Montmerle (\\cite{AM}, \\ha AM) to incorporate results of millimeter continuum studies on circumstellar dust. The IR sources are classified in three classes, according to their spectral energy distributions (SEDs). This classification, initially defined empirically, is now well understood in terms of evolution of low-mass stars at their earliest stages. Submillimeter observations led to the discovery of cold objects, younger than the IR sources, and thus to the introduction of a fourth class named ``Class 0'' (Andr{\\'e} et~al. \\cite{andre93}, \\cite{andre00}). Class 0 sources are very young protostars, peaking in the submillimeter range, at the beginning of the main accretion phase. \\co sources are evolved IR protostars, optically invisible, in the late accretion phase. \\cd sources are YSO surrounded by optically thick circumstellar disks. \\ct sources are YSO with an optically thin circumstellar disk or no circumstellar disk. Studies of optically visible YSO, T~Tauri stars, led to another classification based on the H$_{\\alpha}$ line, which separates ``classical'' T~Tauri stars (CTTS) from ``weak-line'' T~Tauri stars (WTTS) according to their equivalent width in emission, with a boundary at EW[H$_{\\alpha}$]$\\sim 5-20$\\,\\AA, depending on the spectral type (Mart{\\'\\i}n 1997). CTTS and WTTS are usually taken to be identical to Class~II and Class~III sources respectively, on the basis of their IR SED (see AM for a discussion about these two classifications). We will associate in this article Class~II (Class~III) sources with CTTS (WTTS). Several ground-based near-IR surveys (e.g.Wilking et~al. \\cite{wilking89}; Greene et~al. \\cite{GWAYL}, \\ha GWAYL; Barsony et~al. \\cite{barsony97}, \\ha BKLT; and references therein) discovered in a $\\approx$1 square degree area around the densest regions (with survey completeness limit down to $K \\sim 14$), $\\sim$100 low-luminosity embedded sources. More recently, the \\ISOCAM camera on-board the {\\it Infrared Space Observatory} satellite imaged a half square degree centered on L1688 in the mid-IR (LW2 and LW3 filters, respectively centered at 6.7\\,$\\mu$m and 14.3\\,$\\mu$m -- \\ISOCAM central programme surveys by Nordh et~al.; see Abergel et~al. \\cite{abergel96}), and recognized 68 new faint young stars with infrared excess (Bontemps et~al. \\cite{bontemps00}). Near-IR spectroscopy has been used to determine spectral types of an increasingly large number of \\ro Oph YSO (see the pioneering works of Greene \\& Meyer \\cite{greene95}, and Greene \\& Lada \\cite{greene96}). Recently, Luhman \\& Rieke (\\cite{luhman99}) obtained $K$-band spectroscopy for $\\sim$100 sources, combining a magnitude-limited sample in the cloud core ($K \\le 12$) with a representative population from the outer region of the cluster ($K \\le 11$). The \\ro Oph dark cloud YSO have also been extensively studied in X-rays. Early observations with the \\Einstein satellite showed that at the \\TT star stage YSO are bright and variable \\xr emitters in the 0.2--4\\,keV energy band (Montmerle et~al. \\cite{montmerle83}). When the S/N ratio is sufficient large, their \\xr spectra can be fitted by a thin thermal model, with temperatures $\\approx 1$\\,keV and absorption column densities $N_H \\sim 10^{20}$--$10^{22}$\\,cm$^{-2}$. Variability studies and modeling led to explain the \\xr emission in terms of bremsstrahlung from a hot ($T_{\\mathrm X} \\sim 10^7$\\,K) plasma trapped in very large magnetic loops, in other words in terms of an enhanced solar-like flare activity (see reviews by Montmerle et~al. \\cite{montmerle93}; and Feigelson \\& Montmerle \\cite{FM}, \\ha FM). Casanova et~al. (\\cite{CMFA}) -- \\ha CMFA -- reported deep \\ROSAT {\\it Position Sensitive Proportional Counter} ({\\it PSPC}\\,) imaging of the \\ro Oph cloud dense cores A and F. They detected in the $35\\arcmin \\times 35\\arcmin$ central portion of the field (the inner ring of the \\ROSAT detector entrance window support structure) 55 X-ray sources in the 1.0--2.4\\,keV energy band. For three \\xr sources, one or several Class~I sources lie within the error boxes of \\xr peaks, but other counterparts are possible (unclassified IR sources, \\TT stars). \\xr emission from one of these Class~I sources, \\object{YLW15} (=\\object{IRS43} in WLY), was unambiguously confirmed with a follow-up \\ROSAT {\\it High Resolution Imager} ({\\it HRI}\\,) observation by Grosso et~al. (\\cite{grosso97}). The outer portion of the CMFA \\PSPC field, analyzed by Casanova (\\cite{casanova94}), contains 36 \\xr sources. The optical spectroscopic classification of these \\xr sources and other \\xr selected stars in the \\ro Oph dark cloud vicinity, based on H$_{\\alpha}$ and \\ion{Li}{I} (670.8\\,nm) spectroscopy, was made by Mart{\\'\\i}n et~al. (\\cite{martin98}), doubling the number of PMS stars spectroscopically classified in the \\ro Ophiuchi area. Observations of harder \\xr ($>$4\\,keV) from the \\ro Oph dark cloud were initially only possible with non-imaging instruments. {\\it Tenma} and {\\it Ginga} revealed unresolved emission from the cloud core region, with a hard \\xr spectrum with $kT_{\\mathrm X}\\sim4$\\,keV and $N_H\\sim10^{22}$\\,cm$^{-2}$ (Koyama \\cite{koyama87}; Koyama et~al. \\cite{koyama92}). Wide-energy band imaging observations became possible with \\ASCA in the range 0.5--10\\,keV. In the \\ro Oph dark cloud, Koyama et~al. (\\cite{koyama94}) detected hard \\xrs from T~Tauri stars, with $kT_{\\mathrm X}$ up to $\\sim$8\\,keV in the case of the WTTS DoAr21. There is also some evidence for unresolved hard \\xr emission from embedded young stars below the point source detection limit. From this \\ASCA observation, Kamata et~al. (\\cite{kamata97}), found additional \\TT stars and detected three \\xr sources associated with Class~I sources, but with large \\xr error boxes (15$\\arcsec$--30$\\arcsec$). There is a deep connection between IR and \\xr observations of star-forming regions. Sensitive ground-based near-IR surveys penetrate dark clouds (except for dense cores) so that their source populations are frequently dominated by ordinary stars in the Galactic disk. Space-based mid-IR isolates YSO with significant circumstellar material and effectively eliminates the background star population, but they will miss the recognition of YSO with less massive or absent disks. X-ray emission, in contrast, is elevated by 1--4 orders of magnitude in YSO of all ages, irrespective of a disk presence. It thus provides a unique tool for improving the census of young star clusters. In this article, we present the results from the \\HRI follow-up of the CMFA \\PSPC observation. The high angular resolution of these observations allows us to find counterparts to all \\xr sources without ambiguity. The comparison with the sensitive \\ISOCAM survey of the \\ro Oph dark cloud significantly improves the existing classification of these counterparts and allows us to do statistical studies on a well defined sample. We first present the \\ROSAT \\HRI observations: image analysis, source detection and identification ($\\S$\\ref{X-ray}). We incorporate the \\ISOCAM survey results from Bontemps et~al. (\\cite{bontemps00}) and we present the resulting IR classification for the \\HRI sources ($\\S$\\ref{Nature}). The next sections discuss the \\xr luminosity of the \\HRI detected TTS ($\\S$\\ref{Luminosity}), and the X-ray detectability of the embedded TTS population ($\\S$\\ref{properties}). Next (\\S\\ref{origins}), we show that the \\HRI census of Class~III sources cannot be complete, and that numerous unknown low-luminosity Class~III sources, perhaps including brown dwarfs, must exist. Summary of the main results and conclusions are presented in $\\S$\\ref{Summary}, where prospects for improvements with {\\it XMM-Newton} and {\\it Chandra}, are also discussed. Appendix~A gives details about the \\HRI \\xr source detection, and lists the \\xr detections. Optical finding charts, and identification list of the \\HRI \\xr sources can be found in Appendix~B. Appendix~C compares these \\HRI observations with previous \\PSPC ones. Appendix~D discusses the status of optical/IR counterparts without IR classification. \\begin{table*}[!htb] \\small \\caption{Log of \\ROSAT \\HRI Observations.} \\label{tab:log} \\begin{tabular}{lcclrlrrrr} \\hline \\hline \\vspace{-0.3cm}\\\\ & & & \\multicolumn{2}{c}{Begin} & \\multicolumn{2}{c}{End} & \\multicolumn{1}{c}{Elapsed} \\\\ \\vspace{-0.4cm}\\\\ \\multicolumn{1}{c}{Pointing} & \\multicolumn{1}{c}{Core} & \\multicolumn{1}{c}{Obs.} &\\multicolumn{2}{c}{\\hrulefill}&\\multicolumn{2}{c}{\\hrulefill}& \\multicolumn{1}{c}{Time} & \\multicolumn{1}{c}{Exposure}\\\\ \\multicolumn{1}{c}{ID} & & \\multicolumn{1}{c}{\\#} & \\multicolumn{1}{c}{Date} & \\multicolumn{1}{c}{Hour (UT)} & \\multicolumn{1}{c}{Date} & \\multicolumn{1}{c}{Hour (UT)} & \\multicolumn{1}{c}{[ks]} & \\multicolumn{1}{c}{[ks]} \\\\ \\hline 201835h & A & 1 & 1995 Aug 29 & 18$^{\\mathrm h}$31$^{\\mathrm m}$30$^{\\mathrm s}$ & 1995 Sep ~12 & 22$^{\\mathrm h}$35$^{\\mathrm m}$47$^{\\mathrm s}$ & 1\\,224.3 & 51.3\\\\ \\hline 201834h & F & 1 \t& 1995 Mar 09 & 00$^{\\mathrm h}$51$^{\\mathrm m}$03$^{\\mathrm s}$ & 1995 Mar 14 & 05$^{\\mathrm h}$16$^{\\mathrm m}$43$^{\\mathrm s}$ & 447.9 & 12.5 \\\\ 201834h-1 & F & 2 \t& 1995 Aug 18 & 19$^{\\mathrm h}$29$^{\\mathrm m}$21$^{\\mathrm s}$ & 1995 Aug 20 & 14$^{\\mathrm h}$52$^{\\mathrm m}$05$^{\\mathrm s}$ & 27.8 & 27.5 \\\\ 201834h-2 & F & 3 \t& 1996 Sep ~07 & 23$^{\\mathrm h}$12$^{\\mathrm m}$46$^{\\mathrm s}$ & 1996 Sep ~11& 13$^{\\mathrm h}$46$^{\\mathrm m}$23$^{\\mathrm s}$ & 311.6 & 37.2 \\\\ & F & 1+2+3 & 1995 Mar 09 & 00$^{\\mathrm h}$51$^{\\mathrm m}$03$^{\\mathrm s}$ & 1996 Sep ~11& 13$^{\\mathrm h}$46$^{\\mathrm m}$23$^{\\mathrm s}$ & 47\\,739.3 & 77.2 \\\\ \\hline \\hline \\end{tabular} \\end{table*} ", "conclusions": "\\label{Summary} \\subsection{Main observational results} We have obtained two deep exposures of the \\ro Oph cloud core region (d=140\\,pc) with the \\ROSAT {\\it High Resolution Imager} (core~A: 51\\,ks, core~F: 77\\,ks, in three partial exposures). The improved position accuracy (1$\\arcsec$--6$\\arcsec$) with respect to previous recent \\xr observations (\\ROSAT {\\it PSPC}, Casanova et~al. \\cite{CMFA}; and {\\it ASCA}, Koyama et~al. \\cite{koyama94} and Kamata et~al. \\cite{kamata97}) have allowed us to remove a number of positional ambiguities for the detected sources. We have cross-correlated the X-ray positions with IR sources found in the {\\it ISOCAM} survey of the same region at 6.7 and 14.3\\,$\\mu$m, in addition to sources known in the optical, IR, and radio from ground-based observations. We thus have now at our disposal the best-studied sample of X-ray emitting YSO in a star-forming region. We first summarize the main observational results of this article. \\begin{list}{}{\\itemsep 0pt} \\item[(1)] We detect 63 \\HRI \\xr sources, and 55 are identified. Of the 55 identified \\xr sources 40 are \\PSPC sources, and 9 are \\ASCA sources. \\item[(2)] The IR classification (ground-based and \\ISOCAM survey) for the 55 identified \\xr sources yields: one Class~I protostar (\\object{YLW15}=\\object{IRS43}); 23 Class~II sources, including 4 new \\ISOCAM Class~II sources; 21 Class~III sources, including 13 new Class~III sources; 8 new Class~II or Class~III source candidates; one early-type Class~III source (the young magnetic B3 star \\object{S1}), and one field star (the F2V star \\object{HD148352}). The contamination of the sample of new X-ray sources by field stars is negligible. \\item[(3)] There is no statistically significant difference between the X-ray luminosity functions of {\\it HRI}-detected Class~II and Class~III sources, i.e. T~Tauri stars with and without disks, confirming that the contribution of these disks to X-ray absorption, or emission (for instance by magnetic reconnection between the star and the disk), must be small. \\item[(4)] X-ray variability of {\\it HRI}-detected T~Tauri stars has been studied by comparing the \\HRI data with the previously obtained \\PSPC data, and using \\HRI observations done at three different epochs. The resulting statistics show that most of the sources are variable, and that their \\xr variability is consistent with a solar-like (hence magnetic) flare origin. \\item[(5)] We use the information given both by the \\ISOCAM survey and by our \\HRI deep exposure to study the T~Tauri star population of the \\ro Oph dense cores. We confirm that essentially all Class~II and Class~III YSO are X-ray emitters, and that a strong correlation ($\\log(L_{\\mathrm X}/erg\\,s^{-1}) = (1.0\\pm0.2) \\times \\log(L_{\\star}/L_{\\odot}) + 30.1$) exists between the X-ray luminosity and the stellar luminosity of T~Tauri stars, likely down to low luminosities (L$_\\star \\sim 0.1$\\,L$_\\odot$). We confirm that the characteristic $L_{\\mathrm X}/L_{\\star}$ for T~Tauri stars is $\\sim$10$^{-4}$ in the $\\rho$~Oph cloud, albeit with a large dispersion. There is no evidence for a magnetic ``saturation'' seen at a level of 10$^{-3}$ in late-type main sequence stars. \\item[(6)] However, most of the new \\ISOCAM Class~II sources are not detected by the \\HRI. We show that this is consistent with their intrinsic X-ray luminosities being too faint if ``predicted'' using the above $L_{\\mathrm X}$--$L_{\\star}$ correlation. \\end{list} \\subsection{What have we learned ?} \\begin{list}{}{\\itemsep 0pt} \\item[(1)] The first general conclusion we can draw from the \\HRI results presented above is a complete confirmation of the \\PSPC results obtained by CMFA. This was not {\\it a priori} obvious, since the CMFA population (\\PSPC and near-IR) overlaps, but is different from, the {\\it HRI}/near-IR/\\ISOCAM population presented in this paper: many \\PSPC sources are not detected by the \\HRI (see Appendix C), and some \\HRI sources are Class~II and Class~III newly classified thanks to a combined identification with {\\it ISOCAM}. This shows that the $L_\\mathrm{X}$--$L_\\star$ correlation is robust for the $\\rho$~Oph TTS. \\item[(2)] The second, and perhaps most important, conclusion is the probable existence of $\\sim 40$ unknown X-ray YSO down to a limit of L$_\\mathrm{X} \\sim 3 \\times 10^{28}$\\,erg\\,s$^{-1}$ in the {\\it HRI/ISOCAM} overlapping area, which should be mainly low- to very low-mass ($<0.1$--0.6\\,M$_\\odot$) diskless, ``Class~III TTS''. This prediction is based both on the use of the $L_\\mathrm{X}$--$L_\\star$ correlation, legitimated by its robustness, and on the discovery of a large number of faint new IR sources by {\\it ISOCAM}. As shown below, it may be soon verified by the next generation of X-ray satellites, namely {\\it XMM-Newton} and {\\it Chandra}. In this respect, the present paper can be taken as a ``transition'' paper between two generations of X-ray satellites. \\end{list} Why is the detection of these ``unknown TTS'' important ? Because they are diskless, they are unlikely to be recognized as YSO by IR observations alone; and because they are likely to be as numerous as the YSO with IR excess, they have to be included in any reliable census of YSO, with an impact on such basic quantities as the initial mass function, or the star formation efficiency, especially if considered from an evolutionary point of view. For instance, from the results in this paper it is impossible to study the real connection between the distributions of the Class~II and Class~III sources in the densest regions, in particular to see whether the distribution of the Class~III sources is also centered on the same DCO$^+$ cores as the Class~II sources. The number of Class~III sources embedded in the densest regions may, or may not, be comparable to that of the Class~II sources, depending on the timescale for disk dispersal, especially among low-mass YSO. An X-ray improved census of Class~III sources may also be crucial in determining whether a burst of star formation is presently going on in the $\\rho$~Oph cores, as some recent indications suggest (see Mart{\\'\\i}n et al. 1998). It will also allow to study the $L_\\mathrm{X}$--$L_\\star$ correlation for Class~III and Class~II sources seperately, which was not possible in this paper (\\S\\label{Lstar-Lx}) due to insufficient statistics. \\subsection{The potential of {\\it XMM-Newton} and {\\it Chandra}} To quantify the prospects for improvement in the \\xr domain, we have computed the detection threshold for the \\xr camera {\\it EPIC} aboard {\\it XMM-Newton}, which was successfully launched in December 1999 (see Fig.~\\ref{detections}). The improved sensitivity and enlarged energy range (0.5--12\\,keV) of {\\it EPIC} will allow to detect the weak \\ISOCAM Class~II sources, and also to discover numerous unknown faint or embedded Class~III sources, in particular if they have high plasma temperatures (several keV) reached during flares, and extend the census of this population towards the low-mass end. In the best case, the {\\it XMM-Newton} sensitivity will reach L$_\\mathrm{X} \\sim 10^{28}$\\,erg\\,s$^{-1}$ for A$_\\mathrm{V} \\simgt 20$, for long exposures ($>$75\\,ksec). This is nearly two orders of magnitude more sensitive than {\\it ROSAT}. In case the faint Class~III sources turn out to be so crowed that confusion problems arise, the excellent angular resolution of {\\it Chandra} will be critical. In \\ro Oph, there are already several identified {\\it bona fide} and candidate brown dwarfs (see review in Neuh{\\\"a}user et~al. \\cite{neuhaeuser99}, and references therein), and four of them have been recently detected in \\xr using the \\ROSAT \\PSPC archive (Neuh{\\\"a}user et~al. \\cite{neuhaeuser99}). Neuh{\\\"a}user et~al. have also shown that brown dwarfs could be \\xr emitters with the same ratio $\\log(L_{\\mathrm X}/L_\\star) \\sim -4$ than for T~Tauri stars. Thus {\\it Chandra} and {\\it XMM-Newton} should be able to detect many more of these objects with low stellar luminosity and masses, shedding a new light on their nature and early evolution. {" }, "0005/astro-ph0005358_arXiv.txt": { "abstract": "We evaluate the dependence of the cluster correlation length $r_c$ on the mean intercluster separation $D_c$, for three models with critical matter density, vanishing vacuum energy ($\\Lambda = 0$) and COBE normalized: a tilted CDM (tCDM) model ($n=0.8$) and two blue mixed models with two light massive neutrinos yielding $\\Omega_h = 0.26$ and 0.14 (MDM1 and MDM2, respectively). All models approach the observational value of $\\sigma_8$ (and, henceforth, the observed cluster abundance) and are consistent with the observed abundance of Damped Lyman$\\alpha$ systems. Mixed models have a motivation in recent results of neutrino physics; they also agree with the observed value of the ratio $\\sigma_8/\\sigma_{25}$, yielding the spectral slope parameter $\\Gamma$, and nicely fit LCRS reconstructed spectra. We use parallel AP3M simulations, performed in a wide box (side 360$\\, h^{-1}$Mpc) and with high mass and distance resolution, enabling us to build artificial samples of clusters, whose total number and mass range allow to cover the same $D_c$ interval inspected through APM and Abell cluster clustering data. We find that the tCDM model performs substantially better than $n=1$ critical density CDM models. Our main finding, however, is that mixed models provide a surprisingly good fit of cluster clustering data. \\vskip 0.2 truecm PACS: 95.35; 98.80; 98.65.Cw ", "introduction": "The study of the clustering of galaxy clusters, in the early eighties, allowed a basic advancement in our understanding of Large Scale Structure (LSS). The discrepancy between the galaxy correlation length $r_g$ and the cluster correlation length $r_c$ (Bahcall \\& Soneira 1983, Klypin \\& Kopylov 1983, but see also Hauser \\& Peebles 1973) led to the introduction of the concept of bias (Kaiser 1984). Data on $r_c$ were then worked out, in further detail, for Abell clusters by Peacock \\& West (1992) and Postman, Huchra \\& Geller (1992), as well as for clusters in APM and in Edinburgh--Durham Southern Galaxy catalogs, by Dalton et al (1992), Nichol et al (1992) and Croft et al (1997). These analyses show that the value of $r_c$ depends on the mass threshold ($M_{th}$) of the cluster sample, through its mean intercluster separation $D_c = n^{-1/3} (>M_{th})$, and that $r_c$ increases with $D_c$. However, $r_c$ values obtained from Abell and APM data seem only partially consistent; this is to be partially ascribed to different cluster definitions; Bahcall \\& Burgett (1986), Bahcall \\& Cen (1992) and Bahcall \\& West (1992) suggested that observational ambiguities are wide enough to allow to conjecture that the scaling relation $r_c \\simeq 0.4\\, D_c$ holds for $20 < D_c h/{\\rm Mpc} < 100$. Herebelow, we shall refer to this relation as BW conjecture. It ought to be born in mind that, above $\\sim 50\\, h^{-1}$Mpc, such conjecture hinges on the estimates of $r_c$ for 55 and 94$\\, h^{-1}$Mpc mean separations, for richness $R \\geq 1$ and $R \\geq 2$ Abell clusters, while APM data, for the same $D_c$ range, give smaller $r_c$. Dekel et al (1989) and Sutherland \\& Efstathiou (1991) suggested that the projection effects and peculiar inhomogeneities in the Abell sample might have biased upward $r_c$ at large $D_c$. Peacock \\& West (1992), instead, confirmed such points (see also Jing, Plionis \\& Valdarnini, 1992). Altogether, it may be fair to say that the controversy on the observational behaviour of $r_c$ for high $D_c$ values has not been solved yet, although, as we shall see, there may be good reasons to assess that different cluster definitions play a key role. This paper is devoted to a comparison of cluster clustering, as it emerges from such data, with simulations of three cosmological models: a tilted CDM (tCDM) model and two mixed models (MDM1 and MDM2) with cold+hot DM. All models have critical matter density, vanishing vacuum energy, and are COBE normalized. During the last few years, much attention has been devoted to models with a positive cosmological constant $\\Lambda$, also because of the remarkable data sets concerning SN Ia (see,e.g., Riess et al 1998, Perlmutter et al 1998, and references therein). In this work, we shall not debate whether mixed models can still offer a fair fit to all cosmological data; they certainly do not fit SN Ia data, unless their current interpretation was misled by some systematic bias. In a number of cases, however, mixed models were just not tested and the success of $\\Lambda$--models to fit some data set was directly taken as further evidence in their favour. In the case of cluster clustering, we shall show that mixed models perform quite well and are surely better than any other model with matter density parameter $\\Omega_m = 1$ considered until now. In order to fit cluster data with a model, a large simulation volume is required; in fact, we need a fair sample of galaxy clusters for large $D_c$, as well as adequate mass and force resolutions, to identify clusters in a reliable way, for small $D_c$. Simulation parameters are therefore set so to allow a sample of 90 clusters, at least, for large $D_c$ and $\\sim 60$ baryon--CDM particles per cluster, at least, for small $D_c$ (as we shall see, 60 particles correspond to $\\sim 10^{14} h^{-1} M_\\odot$). Altogether, at redshift $z=0$, we shall therefore span a $D_c$ interval ranging from $\\sim 20$ to $80\\, h^{-1}$Mpc. Cluster clustering has been studied by various authors in simulations. In particular, the behaviour of $r_c~vs.~D_c$, for standard CDM and open CDM, was studied by Bahcall \\& Cen (1992), Watanabe et al (1994), Croft \\& Efstathiou (1994), Eke et al (1996), Croft et al (1997), Governato et al (1999). Their results allow to conclude that CDM models with $n=1$ may approach the observed behaviour of $r_c~vs.~D_c$, only for $\\Omega_m < 1$. The behaviour of $r_c~vs.~D_c$ in a mixed model was also studied, using PM simulations, by Klypin \\& Rhee (1994) and Walter and Klypin (1996). Their work treated a different mix from those considered here, using smaller box and resolution. Accordingly, they could inspect only the $D_c$ interval running from $\\sim 20$ to 45$\\, h^{-1}$Mpc. The behaviour they found is only marginally consistent with a constant $r_c/D_c$ ratio, but their model does not exhibit much improvement in respect to pure CDM. The mixed models we consider here were selected on the basis of recent tests on $\\nu$ flavour mixing, which seem to support a non--vanishing $\\nu$--mass. Mixing data come from the solar $\\nu$ deficit (see, e.g., Hampel et al 1996, for GALLEX, and Abdurashitov et al 1996, for SAGE), the atmospheric $\\nu$ anomaly (Fukuda et al 1994) and the LSND experiment (Athanassopoulos et al 1995) on $\\nu$'s arising from $\\mu^+$ and $\\pi^+$ decay. Barger, Weiler \\& Whisnant (1998) and Sarkar (1999) show that all above results can agree if a fourth sterile $\\nu$ exists, which can be however added without harming BBNS or LEP standard results. Diagonalizing the mass matrix, they eventually obtain the four $\\nu$--mass eigenvalues, which split into two nearly degenerate pairs, corresponding to $m_\\nu \\simeq 0$ and $m_\\nu \\sim 1.4$--1.5 eV. It must be outlined that, within this picture, there remains no contradiction among different experimental results, at variance with earlier analyses which seemed to find contradictions between LSND and other $\\nu$--mixing results. In a cosmological context, however, mixed models have been considered since long. The transfer function for several mixed models was first computed by Bonometto \\& Valdarnini (1984). Results on mixed models were then found by a number of authors (see, e.g., Achilli, Occhionero \\& Scaramella 1985, Valdarnini \\& Bonometto 1985, Holtzmann 1989, for results obtainable from the linear theory, and Davis et al 1992, Klypin et al 1993, Ghigna et al 1994, for early simulations). After the release of LSND data, Primack et al (1995) performed simulations of models with 2 massive $\\nu$'s and yielding $\\Omega_h = 0.20$ and found that such mixture eased some problems met by greater $\\Omega_h$ models. The possibility of considering mixed models together with blue spectra (primeval spectral index $n > 1$) was first considered by Liddle et al (1996) and Lucchin et al (1996). In the former paper, blue mixed models able to fit all linear and analytical constraints were shown to exist. In the latter paper, inflationary models leading to blue spectra were discussed and results of an N--body simulation of blue mixed models were reported. Unfortunately, the model considered violated some observational constraints. A systematic study of blue mixed models was recently performed by Bonometto \\& Pierpaoli (1998) and Pierpaoli \\& Bonometto (1999), selecting those consistent with CMB data and data predictable from the linear theory. In the next section we show that the models considered here, on the basis of $\\nu$--physics motivation, are also suitable to fulfill the main observational constraints. In {\\S} ~3 we review the technique used to simulate their non--linear evolution. In {\\S} ~4 we describe how clusters are selected in simulations. Then, in {\\S} ~5 we describe how the 2--point correlation function and its error estimates were worked out. In {\\S} 6, we will show the main results of the $r_c ~vs.~D_c$ behaviour derived from fits to the 2--point functions. {\\S} 7 is devoted to discussion of the results and the main conclusions we derived from this work. ", "conclusions": "" }, "0005/astro-ph0005385_arXiv.txt": { "abstract": "We present supportive evidence that the Boroson \\& Green eigenvector~1 is not driven by source orientation and further that both [OIII]\\,$\\lambda$5007 and [OII]\\,$\\lambda3727$ are isotropically emitted in the radio-quiet sample of BQS (Bright Quasar Survey) quasars, contrary to results found for radio-loud AGN. Studies of optical emission lines in quasars have revealed a striking set of correlations between various emission line properties, known as the Boroson \\& Green eigenvector~1. Until recently it was generally accepted that eigenvector~1 does not depend on orientation as it strongly correlates with [OIII]\\,$\\lambda$5007 emission, thought to be an isotropic property. However, recent studies of radio-loud AGN have questioned the isotropy of [OIII] emission and concluded that [OII]\\,$\\lambda3727$ emission is isotropic. In this paper we investigate the relation between eigenvector~1 and [OII] emission in radio-quiet BQS quasars, and readdress the issue of orientation as the driver of eigenvector~1. We account for the small blue bump present at [OII] wavelengths and subtract FeII emission that contaminates [OIII] emission. We find significant correlations between eigenvector~1 and orientation independent [OII] emission, which implies that orientation does not drive eigenvector~1. The luminosities and equivalent widths of [OIII] and [OII] correlate with one another, and the range in luminosities and equivalent widths is similar. This suggests that our radio-quiet quasar sample of the Bright Quasar Survey is largely free of orientation dependent obscuration and/or ionization effects. We conclude that neither the [OIII] emission nor the [OII]/[OIII] ratio are dependent on orientation in radio-quiet quasars, contrary to recent results found for radio-loud quasars. ", "introduction": "Studies of optical emission lines in quasars have revealed some striking correlations that may well be related to the fundamental properties of the accreting black hole system. Boroson \\& Green (1992, hereafter BG92) performed a principal component analysis (PCA) on the BQS quasar sample (Schmidt \\& Green 1983) and showed that the primary eigenvector (hereafter eigenvector~1 or EV1), which was responsible for $\\sim$ 30\\% of the variance in the data, was anticorrelated with various measures of FeII\\,$\\lambda 4570$ strength (equivalent width and FeII/H$\\beta$ ratio), correlated with [OIII]\\,$\\lambda 5007$ strength (luminosity and peak) and H$\\beta$ FWHM, and anticorrelated with the blue asymmetry of the H$\\beta$ line. It was later found that these optical line properties correlate with UV properties: CIII] width, SiIII]/CIII] ratio, CIV and NV strength (Wills et al. 1999; Kuraszkiewicz et al. 2000) and with soft X-ray properties: luminosity and spectral index (Boroson \\& Green 1992; Corbin 1993; Laor et al. 1994, 1997). Recently Brandt \\& Boller (1998) showed that the correlations between EV1 and the X-ray properties are stronger than those with the individual line parameters, suggesting that the EV1 has a more fundamental physical meaning. A number of physical parameters have been suggested to drive EV1 including accretion rate, orientation, and black hole spin. BG92 and Boroson (1992) argued that EV1 is not driven by an orientation effect (i.e. some anisotropic property), despite the strong dependence on H$\\beta$ line width, as it is strongly correlated with the [OIII]\\,$\\lambda 5007$ (hereafter [OIII]) luminosity, which was assumed to be isotropic. However, the isotropy of the [OIII] emission in other AGN has since been called into question. Jackson \\& Browne (1990) studied a sample of powerful narrow-line radio galaxies and radio-loud quasars, which, in the context of Unified Models (e.g. Antonucci 1993), are considered to be the same type of object viewed from different angles to the radio axis. The [OIII] line luminosity of the narrow-line radio galaxies (viewed edge-on) is lower by 5--10 times than that of the quasars, matched in redshift and extended radio luminosity. This result was surprising. It was expected that the [OIII] emission would be the same in both samples, as it was thought to originate from distances large enough to be unaffected by obscuring material from the dusty torus. Hes, Barthel \\& Fosbury (1996) found that radio-loud quasars and powerful narrow-line radio galaxies show no difference in [OII]\\,$\\lambda3727$ (hereafter [OII]) emission, suggesting that [OII] emission, and not [OIII] emission, is isotropic. As [OIII] has a higher critical density and higher ionization potential, and hence lies nearer to the central engine, this difference can be explained if the [OIII] emission region extends to sufficiently small radii to be obscured by the dusty torus when the active nucleus is viewed ``edge-on''. Support for this scenario was provided by the detection of [OIII] emission in polarized light in 4 out of 7 radio galaxies (one also showing [OII] polarization) while a sample of radio-loud quasars showed none (Di Serego Alighieri et al. 1997). Polarized [OIII] emission has also been observed in NGC~4258 (Wilkes et al. 1995; Barth et al. 1999) a Seyfert galaxy with an edge-on molecular disk surrounding the nucleus. Baker (1997), studying a complete sample of low frequency radio selected quasars from the Molonglo Quasar Sample, found that the [OII] to [OIII] ratio is anticorrelated with the radio-core to lobe flux density ratio $R$, which is generally used as an orientation indicator. This again implies that [OIII] is affected by dust absorption as the orientation becomes more edge-on. Similarly, the [OIII] luminosity versus radio luminosity correlation shows a larger scatter than the similar [OII] versus radio correlation (Tadhunter et al. 1998) in the 2 Jy extended radio selected sample (Wall \\& Peacock 1985). This additional scatter could again be explained by dust obscuration of the [OIII] emission although the authors prefer an interpretation in terms of the higher sensitivity of [OIII] to the ionization parameter (Tadhunter et al. 1998). If the central regions of radio-loud quasars and powerful radio galaxies are basically similar to the central regions of radio-quiet quasars (with the exception of the existence of the radio jets) then by analogy we would expect the behavior of the [OII] and [OIII] lines to be similar in both classes. Indeed, Seyfert~1 galaxies, which in the Unified Model scenario correspond to the face-on Seyfert~2 galaxies, have higher [OIII] luminosities than Seyfert~2s with comparable radio luminosity (Lawrence 1987; but see Keel et al. 1994 who find no difference in a sample of IRAS selected Seyfert galaxies). This suggests that L([OIII])/L([OII]) could be an orientation indicator not only in radio-loud but also in radio-quiet quasars. Similarly FeII emission strength and the broad line widths are strongly dependent on orientation in radio-loud QSOs (e.g. Miley \\& Miller 1979; Wills \\& Browne 1986; Vestergaard, Wilkes \\& Barthel 2000) with stronger FeII and narrower lines in face-on sources. Again, by analogy, this suggests that the extreme EV1 objects, Narrow-Line Seyfert 1s (NLS1), which have stronger FeII emission and narrow lines, could also be face-on. Given the strong evidence for anisotropic [OIII] emission in radio-loud quasars, we present an investigation of the behavior of [OII] emission in a radio-quiet subset of the optically-selected Palomar BQS sample to study the [OII] relation to EV1 in comparison with that of [OIII]. This allows us to revisit the question of orientation as a driver of EV1 and will lead to a better understanding of the underlying physics driving the strongest set of emission line correlations found to date for quasars, and so provide information on their central regions. We also compare the [OII] and [OIII] emission in our radio-quiet, optically selected sample with radio-loud samples to investigate whether the behavior of [OII] and [OIII] emission is similar in the two classes. Finally, we address the question of whether the [OII]/[OIII] ratio is an orientation indicator in radio-quiet quasars. ", "conclusions": "Until recently it was generally accepted that eigenvector~1 does not depend on orientation as it is strongly correlated with [OIII] emission, originally thought to be an isotropic property in quasars. As recent studies of radio selected AGN samples have questioned the isotropy of [OIII] emission, we have investigated the relation between [OII] emission, which appears to be more isotropic, and eigenvector~1 and once again addressed the question of orientation as a driver of eigenvector~1. We chose radio-quiet quasars from the optically selected Bright Quasar Survey which showed either high or low [OIII] luminosity, spanning a wide range of EV1 values in BG92. We subtracted FeII emission, which contaminates the [OIII] emission, from our spectra (using the BG92 iron template). We also demonstrated the significant effect of the presence of the small blue bump (Balmer continuum and FeII emission) on accurate measurements of the [OII] emission line, emphasizing the need for spectra covering a wide ($\\geq$ 1000\\AA) wavelength range in order to determine the underlying continuum. We found: \\begin{enumerate} \\item strong correlations between L([OII]), L([OIII]) and EV1 implying that EV1 does not depend on orientation, confirming earlier conclusions of BG92 and Boroson (1992), based on [OIII] alone. EV1 is likely driven by an intrinsic property (e.g. accretion rate or black hole spin). \\item significant EW([OIII])-- EW([OII]) and L([OIII])-- L([OII]) correlations \\item similar ranges in EW([OIII]) and EW([OII]) and in L([OIII]) and L([OII]) respectively. \\end{enumerate} These results lead us to conclude that the optically selected BQS sample (at least our radio-quiet sample) is free from orientation dependent dust effects and ionization dependent effects in the narrow-line region. Assuming our sample is representative of bright, optically selected radio-quiet quasars, this implies that their [OIII] emission is isotropic and the [OII]/[OIII] ratio is not an orientation indicator. This is in contrast with earlier results for the radio selected AGN (Baker 1997; Jackson \\& Browne 1990). We suggest that this discrepancy may be due to, contamination of the [OIII] emission by orientation dependent FeII emission in the latter samples. Acknowledgements - We are grateful to Perry Berlind for observing the spectra of our sample quasars, Martin Elvis and Joanne Baker for helpful discussions, and Todd Boroson for providing the FeII optical template and the eigenvector~1 values. We gratefully acknowledges the support: of the Smithsonian pre-doctoral fellowship at the Harvard-Smithsonian Center for Astrophysics and grant no. 2P03D00410 of the Polish State Committee for Scientific Research (JK), NASA contract NAS8-39073(CXC) (BJW), NASA LTSA grant NAG5-8107 and the Alfred P. Sloan Foundation (WNB), and a Research Assistantship at SAO made possible through NASA grants: NAGW-4266, NAGW-3134, NAG5-4089 to BJW and the Columbus Fellowship at The Ohio State University (MV)." }, "0005/astro-ph0005450_arXiv.txt": { "abstract": "We present an analysis of the ENEAR sample of peculiar velocities of field and cluster elliptical galaxies, obtained with $D_n$--$\\sigma$ distances. We use the velocity correlation function, $\\psi_1(r)$, to analyze the statistics of the field--object's velocities, while the analysis of the cluster data is based on the estimate of their r.m.s. peculiar velocity, $V_{\\rm rms}$. The results are compared with predictions from cosmological models using linear theory. The statistics of the model velocity field is parameterized by the amplitude, $\\eta_8=\\sigma_8 \\Omega_m^{0.6}$, and by the shape parameter, $\\Gamma$, of the CDM--like power spectrum. This analysis is performed in redshift space, so as to circumvent the need to address corrections due to inhomogeneous Malmquist bias and to the redshift cutoff adopted in the sample selection. From the velocity correlation statistics we obtain $\\eta_8=0.51^{+0.24}_{-0.09}$ for $\\Gamma=0.25$ at the $2\\sigma$ level for one interesting fitting parameter. This result agrees with that obtained from a similar analysis of the SFI I-band Tully--Fisher (TF) survey of field Sc galaxies. Even though less constraining, a consistent result is obtained by comparing the measured $V_{\\rm rms}$ of clusters to linear theory predictions. For $\\Gamma=0.25$ we find $\\eta_8=0.63_{-0.19}^{+0.22}$ at $1\\sigma$. Again, this result agrees, within the uncertainties, with that obtained from the SCI cluster sample based on TF distances. Overall, our results point toward a statistical concordance of the cosmic flows traced by spirals and early-type galaxies, with galaxy distances estimated using TF and $D_n$--$\\sigma$ distance indicators, respectively. \\vspace*{6pt} \\noindent {\\em Subject headings: } Cosmology: observations -- cosmology: theory -- galaxies: distances and redshifts -- large-scale structure of universe. ", "introduction": "\\label{intro} The analysis of the peculiar velocities of galaxies and clusters is one of the most promising ways to investigate the amplitude of cosmic density perturbations on $\\sim$ 100\\h1Mpc scales (e.g., Strauss \\& Willick 1995). The importance of cosmic flows for cosmology has motivated a two-decade long effort of building large and homogeneous redshift-distance samples of galaxies and clusters. Analyses of early redshift-distance surveys of spirals (Aaronson et al. 1982) and of early-types (e.g. Lynden-Bell et al. 1988), even though leading to the development of several statistical methods of analyzing peculiar velocity data, left many issues unresolved, primarily because they were based on relatively small and shallow data sets. Recently, a second-generation of redshift-distance surveys has become available involving high-quality data and significantly larger samples of both spirals (Mathewson, Ford \\& Buchhorn 1992; Haynes et al. 1999a,b) and early-types (da Costa et al. 2000a). The existence of these new samples has raised the hope that some of the discrepancies found in earlier analyses may soon be settled. Indeed, the analyses of the different all-sky catalogs of peculiar velocity data currently available such as Mark~III (Willick et al. 1997) and SFI (e.g. da Costa et al. 1996; Giovanelli et al. 1998), lead to a roughly consistent picture of the peculiar velocity field and the local mass distribution (Dekel et al. 1999). However, some quantitative disagreements still remain ranging from the amplitude of the bulk velocity (da Costa et al. 1996; Giovanelli et al. 1998; Dekel et al. 1999), to estimates of the parameter $\\beta=\\Omega_m^{0.6}/b$ (e.g., Davis, Nusser \\& Willick 1996; Zaroubi et al. 1997; da Costa et al. 1998; Willick \\& Strauss 1998; Freudling et al. 1999; Borgani et al. 2000), where $\\Omega_m$ is the cosmological matter density parameter and $b$ is the linear galaxy biasing factor. It is important to emphasize that the two most important catalogs currently in use, Mark~III and SFI, consist of combinations of distinct data sets covering different parts of the sky and therefore could be susceptible to subtle systematic effects. Both catalogs also rely predominantly on Tully-Fisher distances of spiral galaxies and we should note that earlier statistical comparisons of the velocity fields derived from $\\dnsig$ and TF distances found significant differences between them (e.g., G\\'orski et al. 1989; Tormen et al. 1993). There have also been claims of significant differences, larger than expected from the estimated errors, between cluster distances estimated using galaxies of different morphological types (e.g. Mould et al. 1991; Scodeggio, Giovanelli \\& Haynes 1998). In this context, the recently completed all-sky redshift--distance survey of early-type galaxies (da Costa et al. 2000a, hereafter ENEAR), probing a volume comparable to that of the existing catalogs of peculiar velocity data, is a welcome addition. The ENEAR galaxies sample different regions of space and density regimes; the peculiar velocities are measured using an independent distance indicator; and the distances are based on separate types of observations, reduction techniques and corrections. Finally, the ENEAR sample has well defined selection criteria, the completeness of the observations is uniform across the sky and the data, mostly new measurements by the same group, are in a homogeneous system. The present Letter has the twofold aim of comparing global statistical quantities, which describe the velocity fields traced by the TF and $\\dnsig$ distance indicators, and of placing constraints on the nature of the fluctuation power--spectrum. Our analysis is based on the velocity correlation statistics and the r.m.s. one-dimensional peculiar velocity of clusters. These statistics were used by Borgani et al. (2000, B00 hereafter) and Borgani et al. (1997, B97 hereafter) to analyze the SFI sample of field spirals and the SCI sample of cluster spirals (Giovanelli et al. 1997), respectively. In this paper, the same analysis is carried out for the ENEAR sample of field galaxies and groups and for ENEAR clusters (hereafter ENEARc; Bernardi et al. 2000, in preparation). ", "conclusions": "\\label{conclusions} We presented statistical analyses of the peculiar velocity field within $cz=6000\\vel$ traced by field objects and clusters in the ENEAR sample based on $D_n$--$\\sigma$ distances. We use the velocity correlation statistics $\\psi_1(r)$ to characterize the velocity field traced by field ellipticals and loose groups ad find results which are consistent with those obtained from the SFI sample of spirals with TF distances. Contrary to past claims, we find no statistically significant differences between the peculiar velocity fields mapped by spirals and ellipticals. This result is in general agreement with and generalizes the findings of da Costa et al. (2000b) using the bulk-velocity statistics. Constraints on the power spectrum of density fluctuations were derived by resorting to linear theory. Assuming the shape of the power spectrum to be consistent with results from galaxy galaxy clustering analyses, $\\Gamma=0.25$, we find $\\eta_8=0.51^{+0.24}_{-0.09}$ at $2\\sigma$ level for one interesting fitting parameter. A consistent constraint is also obtained from the analysis of the r.m.s. velocity of ENEAR clusters; for the same value of the shape parameter $\\Gamma$, it implies $\\eta_8=0.63_{-0.19}^{+0.22}$ at $1\\sigma$, thus consistent with results from the SCI cluster TF velocities (B97, Watkins 1997). Our results confirm the conclusion by B00 that the amplitude of cosmic flows can be reconciled with independent constraints on the amplitude of density perturbations as that required by the number density of nearby rich clusters. They also show that consistent results are obtainable from independent distance indicators, once they are applied to homogeneously selected galaxy samples." }, "0005/astro-ph0005499_arXiv.txt": { "abstract": "$^{12}$CO $J$ = 1--0, 2--1, 4--3, 7--6, and $^{13}$CO 1--0, 2--1, and 3--2 line emission was mapped with angular resolutions of 13$''$ -- 22$''$ toward the nuclear region of the archetypical starburst galaxy M\\,82. There are two hotspots on either side of the dynamical center, with the south-western lobe being slightly more prominent. Lobe spacings are not identical for all transitions: For the submillimeter CO lines, the spacing is $\\sim$15$''$; for the millimeter lines (CO $J$ = 2--1 and 1--0) the spacing is $\\sim$26$''$, indicating the presence of a `low' and a `high' CO excitation component. A Large Velocity Gradient (LVG) excitation analysis of the submillimeter lines leads to inconsistencies, since area and volume filling factors are almost the same, resulting in cloud sizes along the lines-of-sight that match the entire size of the M\\,82 starburst region. Nevertheless, LVG column densities agree with estimates derived from the dust emission in the far infrared and at submillimeter wavelengths. 22$''$ beam averaged total column densities are $N$(CO) $\\sim$ 5\\,10$^{18}$ and $N$(H$_2$) $\\sim$ 10$^{23}$\\,\\cmsq; the total molecular mass is a few 10$^{8}$\\,\\solmass. Accounting for high UV fluxes and variations in kinetic temperature and assuming that the observed emission arises from photon dominated regions (PDRs) resolves the problems related to an LVG treatment of the radiative transfer. Spatial densities are as in the LVG case (\\numd\\ $\\sim$ 10$^{3.7}$\\,\\percc\\ and $\\sim$10$^{3}$\\percc\\ for the high and low excitation component, respectively), but $^{12}$CO/$^{13}$CO intensity ratios $\\ga$10 indicate that the bulk of the CO emission arises in UV-illuminated diffuse cloud fragments of small column density ($N$(H$_2)$ $\\sim$ 5\\,10$^{20}$\\,\\cmsq/\\kms) and sub-parsec cloud sizes with area filling factors $\\gg$1. Thus CO arises from quite a different gas component than the classical high density tracers (e.g. CS, HCN) that trace star formation rates more accurately. The dominance of such a diffuse molecular interclump medium also explains observed high [C\\,{\\sc i}]/CO line intensity ratios. PDR models do not allow a determination of the relative abundances of $^{12}$CO to $^{13}$CO. Ignoring magnetic fields, the CO emitting gas appears to be close to the density limit for tidal disruption. Neither changes in the $^{12}$C/$^{13}$C abundance ratio nor variations of the incident far-UV flux provide good fits to the data for simulations of larger clouds. A warm diffuse ISM not only dominates the CO emission in the starburst region of M\\,82 but is also ubiquitous in the central region of our Galaxy, where tidal stress, cloud-cloud collisions, shocks, high gas pressure, and high stellar densities may all contribute to the formation of a highly fragmented molecular debris. $^{12}$CO, $^{12}$CO/$^{13}$CO, and [C\\,{\\sc i}]/CO line intensity ratios in NGC\\,253 (and NGC\\,4945) suggest that the CO emission from the centers of these galaxies arises in a physical environment that is similar to that in M\\,82. Starburst galaxies at large distances ($z$ $\\sim$ 2.2--4.7) show $^{12}$CO line intensity ratios that are consistent with those observed in M\\,82. PDR models should be applicable to all these sources. $^{12}$CO/$^{13}$CO line intensity ratios $\\gg$10, sometimes observed in nearby ultraluminous mergers, require the presence of a particularly diffuse, extended molecular medium. Here [C\\,{\\sc i}]/CO abundance ratios should be as large or even larger than in M\\,82 and NGC\\,253. \\ \\ \\ \\ \\\\ ", "introduction": "Low lying rotational transitions of CO are widely used as tracers of molecular hydrogen and are essential to determine dynamical properties and total molecular masses of galaxies. The widespread use of CO $J$ = 1--0 and 2--1 spectroscopy is however not sufficiently complemented by systematic surveys in higher rotational CO transitions to confine the excitation conditions of the dense interstellar medium (ISM). While the $J$ = 1 and 2 states of CO are only 5.5 and 17\\,K above the ground level, the $J$ = 3 to 7 states are at 33, 55, 83, 116, and 155\\,K and trace a component of higher excitation. `Critical densities', at which collisional deexcitation matches spontaneous decay in the optically thin limit, are $\\sim$ 10$^{5-6}$\\,\\percc\\ for CO $J$ = 3--2 to 7--6 in contrast to 10$^{3.5}$ and 10$^{4.3}$\\,\\percc\\ for the ground rotational CO transitions. Starburst galaxies are known to contain large amounts of molecular gas that may be heated to $T_{\\rm kin}$ $\\sim$ 100\\,K by young massive stars, cosmic rays or turbulent heating. Therefore highly excited CO transitions, observed at submm-wavelengths, are {\\it the} appropriate tool to study this interstellar gas component. Among the three nearest ($D$ $\\sim$ 3\\,Mpc) nuclear starburst galaxies, NGC\\,253, NGC\\,4945, and M\\,82 (NGC\\,3034) M\\,82 is most readily accessible from telescopes of the northern hemisphere. Containing one of the brightest IRAS point sources beyond the Magellanic Clouds (\\SHUN\\,\\,$\\sim$ 1000\\,Jy), M\\,82 has been observed at a variety of wavelengths, ranging from the radio to the $\\gamma$-ray domain of the electromagnetic spectrum. The starburst in M\\,82 is likely triggered by a tidal interaction with M\\,81, causing a high infrared luminosity (\\irlum\\,\\,$\\sim$ 4\\,10$^{10}$\\,\\solum), a high density of supernova remnants, and copious amounts of dense gas with strong OH and \\WAT\\ masers and a large number of molecular high density tracers (for CO maps, see Sutton et al. 1983; Olofsson \\& Rydbeck 1984; Young \\& Scoville 1984; Nakai et al. 1986, 1987; Lo et al. 1987; Loiseau et al. 1988, 1990; Phillips \\& Mampaso 1989; Turner et al. 1991; Tilanus et al. 1991; Sofue et al. 1992; White et al. 1994; Shen \\& Lo 1995; Kikumoto et al. 1998; Neininger et al. 1998). So far, few CO 4--3 maps of external galaxies were published (for M\\,51, M\\,82, M\\,83, and NGC6946 see White et al. 1994; Petitpas \\& Wilson 1998; Nieten et al. 1999). Among these M\\,82 is the only true starburst galaxy but its CO 4--3 map (White et al. 1994) is confined to the very central region. With respect to higher rotational CO transitions, only a few CO 6--5 spectra were presented from nearby galaxies (Harris et al. 1991; Wild et al. 1992). We have used the Heinrich-Hertz-Telescope (HHT) on Mt. Graham (Baars \\& Martin 1996) to map M\\,82 in the CO $J$= 7--6, 4--3, and $^{13}$CO 3--2 transitions. These data are complemented by new $J$ = 2--1 and 1--0 spectra taken with the IRAM 30-m telescope. ", "conclusions": "We have studied millimeter and submillimeter CO line emission up to the $J$ = 7--6 rotational transition toward the central region of the starburst galaxy M\\,82 and obtain the following main results: \\begin{enumerate} \\renewcommand{\\labelenumi} {(\\arabic{enumi})} \\item The spatial structure of the millimeter and submillimeter CO emission is distinct. While integrated intensity maps suggest that the lobe separation of the low-$J$ transitions is $\\sim$26$''$, it is $\\sim$15$''$ for the mid-$J$ transitions. Major-axis position-velocity maps in the CO $J$ = 2--1 and 4--3 lines show however agreement in the lobe positions. This indicates that, at the inner edges of the low-$J$ CO lobes, line profiles are wider in the higher excited CO transitions. We thus distinguish between a `low' and a `high' CO excitation component, the latter coinciding with the main source of millimeter and submillimeter dust emission. \\item An LVG excitation analysis of CO submillimeter lines leads to internal inconsistencies. While measured line intensities are reproduced with $T_{\\rm kin}$ $\\sim$ 60 -- 130\\,K, \\numd\\ $\\sim$ 10$^{3.3-3.9}$\\,\\percc, cloud averaged column densities $N$(CO)$_{\\rm cloud}$ $\\sim$ 10$^{20}$ and $N$(H$_2$)$_{\\rm cloud}$ $\\sim$ 10$^{24-25}$\\,\\cmsq, [$^{12}$CO]/[$^{13}$CO] abundance ratios $\\ga$50, and a total molecular mass of a few 10$^{8}$\\,\\solmass, area filling factors ($f_{\\rm a}$ $\\sim$ 0.05--0.10) and volume filling factors ($f_{\\rm v}$ $\\sim$ 0.05) are similar. This results in cloud sizes that do not match their angular scale. On the other hand, the resulting H$_2$ column density is consistent with that derived from the dust continuum at millimeter and submillimeter wavelengths. For the low excitation component, densities are $\\sim$ 10$^{3}$\\percc. \\item An application of PDR models resolves the inconsistencies of the LVG calculations. LVG densities, column densities, and total mass are confirmed. The bulk of the CO emission arises, however, from a diffuse, low column density ($N$(H$_2$) $\\sim$ 5\\,10$^{20}$\\,\\cmsq/\\kms) interclump medium with small $X$ = $N$(H$_2$)/$I_{\\rm CO}$ conversion factors, area filling factors $\\gg$1, and sub-parsec cloud sizes. The relations defined by Larson (1981) are not fulfilled and the gas may not be virialized. Such a scenario explains why CS or HCN are better tracers of global star formation rate and infrared luminosity than CO. Our scenario also explains observed high [C\\,{\\sc i}]/CO line intensity ratios, while relative abundances of $^{12}$CO versus $^{13}$CO cannot be accurately determined. Higher column density clouds, even accounting for variations in far-UV flux and $^{12}$C/$^{13}$C isotope ratios, do not reproduce observed $^{12}$CO/$^{13}$CO line intensity ratios $\\ga$10. Densities are close to the minimum values required for tidal stability in the absence of magnetic fields. \\item In regard to $^{12}$CO line intensity ratios, the central region of M\\,82 appears to be representative for the entire family of starburst galaxies, both at small and at high redshifts. A comparison of the starburst regions in M\\,82 and NGC\\,253 demonstrates that this similarity extends to $^{12}$CO/$^{13}$CO and [C\\,{\\sc i}]/CO line intensity ratios. The large $^{12}$CO/$^{13}$CO line intensity ratios ($\\gg$10) observed toward `nearby' mergers prove, however, that differences exist at least w.r.t. rare CO isotopomers. Galaxy pairs with such high $^{12}$CO/$^{13}$CO line ratios require the presence of a particularly diffuse highly fragmented low column density ISM. \\end{enumerate} Apparently, dropping the assumption of constant temperature in the CO excitation model is a necessary step to provide a self-consistent approach to the physical properties of molecular clouds in the nuclear starburst region of M\\,82. While the use of PDR models is crucial for a better understanding of the molecular gas phase in a starburst environment, important information is still missing. Interferometic observations of high density tracers (e.g. CN, CS, HCN, HNC, N$_2$H$^{+}$), coupled with PDR model calculations including chemical aspects, are needed to fully understand the spatial morphology, density distribution, and molecular excitation of this archetypical starburst complex. An interesting aspect is provided by HCO$^{+}$ $J$ = 1--0 line emission (Table\\ \\ref{lobes}). The bulk of this emission might arise from regions intermediate between those of the low and high CO excitation component. Since this molecule (as well as N$_2$H$^{+}$) is a sensitive tracer of ionization conditions in the dense gas, a detailed knowledge of its spatial distribution would be crucial for a better understanding of structure and excitation. So far, models were calculated for the cosmic ray flux of the solar neighbourhood. A flux enhancement by two to three orders of magnitude (with all necessary chemical implications) has still to be incorporated into PDR codes (but see Schilke et al. 1993). Another important quantity is the spatial distribution of the UV flux. For M\\,82, we do not know the variation of the UV flux as a function of galactocentric radius. \\appendix" }, "0005/astro-ph0005516_arXiv.txt": { "abstract": "The radio emission from supermassive black holes in nearby early-type galaxies can be used to test possible explanations for their low luminosities. We calculate the expected contribution from thermal synchrotron emission from hot accretion flows to the high radio frequency observations of NGC 2300, NGC 1399, NGC 4278 and NGC 4594. We find that, in all cases, and in accordance with our previous findings, hot flows accreting close to their Bondi rates overestimate significantly the observed fluxes. This implies that simply assuming a low radiative efficiency for the accreting gas is not enough to explain their low luminosities. Smaller central densities and accretion rates, as expected in the presence of strong mass loss or convection in the flows, can help reconcile the models with observations. We also show that a significant contribution to the high-frequency radio spectra can be due to non-thermal synchrotron emission from the small scale radio jets observed in these systems, allowing for even lower accretion rates in the inflows. We suggest that these outflows or jets may dump significant energy into the surronding medium close to the accretion radius and so reduce the accretion rates onto these systems. We discuss the relationship between the radio flux and black hole mass for the observed sample and its potential importance for probing accretion models. ", "introduction": "Most nearby galaxies exhibit little or no nuclear activity. However dynamical arguments based on the observed stellar and gas distributions firmly imply the presence of supermassive compact objects in their cores (Magorrian et al.~1998; Richstone et al.~1999; van der Marel 1999). These studies show that virtually all early-type galaxies host black holes with masses in the range $10^8-$ to a few $10^9$ \\Msun. The central black holes in nearby early--type galaxies are probably the remnants of QSO phenomena (McLure et al.~1999). Unlike the giant ellipticals at high redshifts which host radio galaxies and radio-loud quasars, they only display low-luminosity radio cores (Sadler, Jenkins \\& Kotanji 1989; Wrobel \\& Heershen 1991). However, the black holes in the centers of nearby early--type galaxies have enough fuel so that they should still exhibit quasar--like activity. X-ray studies show that they possess extensive hot-gaseous halos which should accrete onto the central black holes and give rise to far more activity than is observed. The Bondi accretion rates for the typical temperatures and densities of their interstellar medium (ISM) are typically estimated to be $\\sim 0.01-0.1$ \\Msunpyr, implying luminosities of $10^{45-46} \\ergps$, for standard accretion disks radiative efficiencies of $\\sim 10\\%$ (see e.g. Di Matteo et al. 2000 for details). Because of the lack of any such activity it has been suggested (Fabian \\& Rees 1995; Reynolds et al. 1996; Mahadevan 1997; Di Matteo \\& Fabian 1997, Di Matteo et al. 1999, 2000 and references therein) that accretion in the nuclei of ellipticals occurs at low radiative efficiency as predicted in advection dominated accretion flow (ADAF) models (e.g. Rees et al. 1982; see Narayan, Mahadevan \\& Quataert 1998 for a review). In these models, thermal synchrotron emission is predicted to give a strong contribution to the radio emission. Because the only obvious sign of activity from these supermassive black holes is their radio emission, studies in this band provide a useful tool for constraining accretion models for these systems. In order to discriminate between a potential accretion flow component and emission from the more extended, scaled-down jets, also common around the supermassive black holes in early-type galaxies, it is crucial to examine their core radio emission component at high resolution and high radio frequencies. In previous work we have shown that the high-frequency radio emission from the cores of three elliptical galaxies is strongly suppressed with respect to the standard ADAF model predictions for black holes accreting at close to the Bondi rates (Di Matteo et al. 1999, hereafter DM99). We also showed that radio emission can provide powerful constraints on ADAF models for the cores of ellipticals, implying that the low-radiative efficiency of an ADAF in not enough to explain their low luminosities. and that the accretion rates onto the black holes ought to be smaller than the expected Bondi rates. We have examined how the required suppression can be obtained if strong mass loss is present in hot accretion flows and/or if matter is fed at much lower rates than those expected (DM99). This is in agreement with the proposal that a direct consequence of the dynamics of hot quasi-spherical flows is the development of strong outflows (as emphasized by Begelman \\& Blandford 1999; Igumenshchev \\& Abramowicz 1999; Stone, Pringle \\& Begelman 2000) leading to suppressed density profiles $\\rho \\propto r^{-3/2+p}$ for $0