Filename
stringlengths 22
64
| Paragraph
stringlengths 8
5.57k
|
|---|---|
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Abstract. The Polarized Electrons for Polarized Positrons (PEPPo) experiment has demonstrated the efficient transfer of polarization from electrons to positrons pro- duced by the bremsstrahlung radiation of a polarized electron beam in a high-Z target. Positron polarization up to 82% has been measured for an initial electron beam mo- mentum of 8.19 MeV/c, limited only by the electron beam polarization. Combined with the high intensity and high polarization performances of polarized electron sources, this technique extends efficient polarized positron capabilities from GeV to MeV elec- tron accelerators. This presentation reviews the PEPPo proof-of-principle experiment and addresses the perspectives for future applications. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | tering requires polarized and unpolarized elec- tron and positron beams for a model indepen- dent extraction [7]. In the several tens to hundreds GeV energy range, the lepton-hadron interaction is no longer re- stricted to neutral current exchange and reaches the charged currents domain of the electroweak sector. Charged W± currents interact with elec- tron and positron beams as essentially differ- ent experimental probes, able to uniquely iso- late positively or negatively charged quarks. In the deep inelastic regime, charged current in- teractions access combinations of quark flavors different from those measured with purely elec- tromagnetic neutral interaction, providing an al- ternative and novel source of information about parton distribution functions, particularly with respect to the strange and charm quarks. For instance, the availability of polarized electron and positron beams provide the necessary tools to measure the difference between the strange and anti-strange quark distributions free from any ambiguities related to the hadronization pro- cess [8]. Such polarized lepton beams also pro- vide the ability to test fundamental predictions of the Standard Model such as the absence of right-handed charged currents or some possible scenarios for the existence of a new physics be- yond the frontiers of the Standard Model. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | The production of high-quality polarized po- sitron beams relevant to these many applications remains however a highly difficult task that, un- til recently, was feasible only at large scale accel- erator facilities. Polarized positron beams based on radioactive sources are limited in polarization and intensity. The production of highly polarized and intense positron beams has been achieved at high energy storage facilities (> 10 GeV) taking advantage of the self-polarizing Sokolov-Ternov effect [9] for ultra-relativistic particles orbiting inside a magnetic field. Other scenarios [10, 11] have been proposed within the context of the International Linear Collider project, but still involving high energy electron beams and chal- lenging technologies that intrinsically limit their range of applications. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Positrons) technique [13] provides a novel and potentially widely accessible approach based on the production, within a tungsten target, of po- larized e+e− pairs from the circularly polarized bremsstrahlung radiation of a low energy (< 10 MeV) highly polarized electron beam. This presenta- tion reviews the principle of operation of this technique and its experimental demonstration performed at the injector of the Continuous Elec- tron Beam Accelerator Facility (CEBAF) of the Thomas Jefferson National Accelerator Facility (JLab) in Newport News, VA. Further develop- ments of the PEPPo method in the context of the JLab physics program and a possible low energy polarized positron beam facility are also sketched. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Polarization phenomena in electromagnetic pro- cesses have been investigated since the early thir- ties [14,15,16]. The originally more complete cal- culations of the polarization of the bremsstrah- lung radiation generated by an electron beam in the vicinity of a nuclear field [17] drove the de- velopment of polarized photon beams: an upolar- ized electron beam is predicted to generate a lin- early polarized photon beam, while a polarized electron beam would generate a circularly po- larized photon beam with polarization directly proportionnal to the initial electron beam po- larization. These features were used extensively at numerous accelerator facilities, and more re- cently in the experimental hall B of JLab [18] to operate a high energy polarized photon beam. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | shows the circular-to-longitudinal polarization trans- fer calculated for a high Z material and differ- ent initial beam conditions within the framework of Ref. [20]. Depending on the kinetic energy of the created positrons, the polarization transfer ranges from -1 to 1 with a remarkable kinemat- ically symmetric behaviour. This quite natural feature for a final system consisting of two par- ticles identical in mass and spin, is absent from older calculations. As a consequence, the low en- ergy half of the spectra shows strong differences between ultra-relativistic and finite mass calcu- lations, even when high initial photon energies are considered. These different approaches both predict that a circularly polarized photon beam creates a polarized e+e− pair whose longitudi- nal and transverse polarization components are both proportionnal to the initial photon beam polarization. Polarization transfer in the trans- verse plane is however much less efficient than in the longitudinal plane. The experimental demon- stration of the circular-to-longitudinal polariza- tion transfer is relatively recent and has been carried out at KEK [10], SLAC [11] and JLab [21] using completely different techniques to produce the initial polarized photon beam. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Fig. 1. Circular-to-longitudinal polarization trans- fer from photons to positrons emitted at a small an- gle for different initial photon energies and within a high Z material [20]. Full lines indicate calculations within a full screening approach, and the dashed line corresponds to a calculation neglecting screening ef- fects. Curves are represented as function of the full kinetic energy portion available to the positrons. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | intensity (∼mA), and high polarization (> 80%) electron beam driver, providing access to polar- ized positron beams to a much wider commu- nity because of the reduced cost of the driver accelerator. The PEPPo experiment [13] was de- signed to evaluate the performance of the PEPPo concept by measuring the polarization transfer from a primary electron beam to the produced positrons. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | The PEPPo experiment ran at the CEBAF in- jector [24] of JLab in june 2012 using a polar- ized electron beam up to pe=8.19±0.04 MeV/c to measure the momentum dependence of the polarization of the produced positrons in the mo- mentum range 3.07-6.25 MeV/c. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | where Ap is the positron analyzing power of the polarimeter and (cid:15)P is the electron-to-positron polarization transfer efficiency. Knowing PT , Pe, Ap and measuring Ap C provide a measurement of P(cid:107) and consequently (cid:15)P . |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | momentum of the PEPPo experiment and in- serting T1 in the beam path allowed us to es- tablish the S1 current by optimizing the trans- ported flux at FC for each momentum. Further optimization of that same flux while powering S2 set the current of that last magnet of the PEPPo beam transport line. Reversing the DD polar- ity is selecting and transporting to T2 positrons produced at T1. T2 is mounted on an electrically isolated aluminum target holder (TH) connected to a picoammeter and serving as a relative po- sitron flux monitor. The S1 and S2 current op- timization procedure was repeated with respect to the TH beam diagnostic for positive DD po- larity, and found to be in good agreement with previously optimized current values, supporting a negligible effect of the polarimeter magnet on incoming particle trajectories. The experimen- tally determined S1, DD, and S2 currents agree well with those determined by a GEANT4 [27] model of the experiment using magnetic fields modeled with OPERA-3D [28]. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | ton asymmetry for a known beam permits the calibration of the analyzing power of the polari- meter [30] assuming knowledge of the beam and target polarizations. The polarization of the electron beam, Pe, was measured to be 85.2 ± 0.6 ± 0.7% with a Mott polarimeter [29]. The first uncertainty is statis- tical and the second is the total systematic un- certainty associated with the theoretical and ex- perimental determination of the Mott analyzing power. The target polarization was experimentally de- duced from the measurement of the magnetic flux induced in a set of pick-up coils surround- ing the iron core [26], that was generated when powering the S3 magnet to the nominal 60 A op- erational current [31]. The comparison of the ex- perimental signal with the simulated signal mod- eled with OPERA-3D provides the average lon- gitudinal magnetic field polarizing the iron core Bz = 2.935±0.036 T [32]. The corresponding av- erage longitudinal polarization was determined to be PT = 7.06±0.09% [32], in very good agree- ment with the previously reported value [26]. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Electrons or positrons arriving at T2 convert into photons that eventually fire the crystal ar- ray. The signal delivered by each crystal is read with a R6236-100 Hamamatsu photomultiplier (PMT). The effective gain of the full electronics chain of each crystal was calibrated prior beam exposure with 137Cs and 22Na radioactive sour- ces, and monitored throughout data taking by controlling the position of the 511 keV peak pro- duced by the annihilation of positrons gener- ated inside the core. This method insures a ro- bust and stable energy measurement, intrinsi- cally corrected for possible radiation damages or PMT-aging effects. The PMT signals are fed into an analog-to-digital converter module that samples the signal into 500 successive time bins of 4 ns duration and enables three data taking modes. The sample mode, used for detector monitoring, allows reg- istration of the full sample set whenever an ex- ternal trigger is provided. The semi-integrated mode, used for positron data taking, collapses the 500 samples into 1 integrated value regis- tered for each positron trigger. The latter is built from the coincidence between the central crys- tal and a 1 mm thick scintillator (TS on Fig. 2) placed between the PEPPo line vacuum exit win- dow and T2; it provides an effective charged par- ticle trigger that considerably reduces the neu- tral backgrounds into the crystal array. The last or integrated data taking mode integrates the signal delivered by the crystal over a duration corresponding to a fixed beam polarization ori- entation (helicity gate). It is specific to high rate, background-free environments, and was used for the electron calibration measurements. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | to provide the actual Compton asymmetries Ae C for electrons, free from false asymmetries related to the beam or the analyzing magnet. Ae repre- sents here the electron analyzing power of the polarimeter. Electron experimental data feature high statistical accuracy (< 1%) and similar sys- tematic errors originating from the determina- tion of the pedestal signal. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Experimental asymmetries and uncertainties for each positron momentum are obtained by insur- ing a mimimum energy deposit Eij > 511 keV. Main sources of systematics originate from the energy calibration procedure, the random sub- traction method, and the selection of coincidence events. The relative contribution to asymmetry uncertainties does not exceed 2%. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | The PEPPo beam line, magnet fields, and detec- tion system was modeled using GEANT4, taking advantage of a previous implementation of polar- ized electromagnetic processes [33]. The determi- nation of the analyzing power of the polarimeter for positrons relies on the comparison between experimental and simulated electron asymme- tries. This benchmarks the GEANT4 physics pack- ages and resolves related systematic uncertain- ties within the limits of the measurement accu- racy. The measured electron analyzing power is shown in Fig. 3 and compared with the predic- tions of the GEANT4 model of the experiment. The reamarkable agreement between data and simulations indicates an accurate understanding of the full experimental setup. The analyzing power of the polarimeter for positrons (Fig. 3) is then directly simulated. The combination of the supplementary e+-to-γ annihilation conversion process together with the minimum energy de- posit requirement leads to larger analyzing power for positrons than electrons. The energy cut ef- fect is strong at low e+ momenta where it re- moves a significant part of the energy spectra acting as a dilution of the polarization sensitiv- ity. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Fig. 3. Electron and positron analyzing powers of the central crystal of the polarimeter. Statistical uncertainties were combined quadratically with sys- tematic uncertainties taken from Pe, PT , and Ae C to determine acutal error bars [21]. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | The positron longitudinal polarization P(cid:107) and the polarization transfer efficiency (cid:15)P are ob- tained from Eq. 1 knowing the electron beam polarization, the target polarization, and the po- sitron analyzing power. The PEPPo data demon- strate a remarkably highly efficient polarization transfer from photons to positrons over a large positron momentum range (Fig. 4). The high po- larization observed at energies significantly smaller than the kinematically achievable maximum is most likely a consequence of multiple scattering effects: higher energy particles with higher po- larization suffer loss of energy when interacting with atomic electrons but their polarization re- mains globally unchanged. The bremsstrahlung of longitudinally polarized electrons is therefore demonstrated to be an efficient process for gen- erating longitudinally polarized positrons. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Fig. 4. PEPPo measurements of the positron po- larization (top panel) and polarization transfer ef- ficiency (bottom panel); statistics and systematics are reported for each point, and the shaded area in- dicates the electron beam polarization [21]. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | to handle, making the PEPPo concept easier to realize. It is therefore more realistic to conceive of application of this technique in many different research fields. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | Fig. 5. GEANT4 simulations of the beam energy dependence of the positron production efficiency (top panel) and the circular-to-longitudinal polariza- tion transfer (bottom panel) at the positron energy corresponding to the optimum FoM [35]. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | wide community ranging from atomic and con- densed matter physics up to nuclear and particle physics. |
Processed_Polarized_positron_production_at_MeV_electron_acce.txt | We are deeply grateful to the SLAC E-166 Collabo- ration, particularly K. Laihem, K. McDonald, S. Rie- mann, A. Sch¨alicke, P. Sch¨uler, J. Sheppard and A. Stahl for loan of fundamental equipment parts and support in GEANT4 modeling. We also thank N. Smirnov for delivery of critical hardware. This work was supported in part by the U.S. Depart- ment of Energy, the French Centre National de la Recherche Scientifique, and the International Linear Collider project. Jefferson Science Associa-tes oper- ates the Thomas Jefferson National Accelerator Fa- cility under DOE contract DE-AC05-06OR23177. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | We present deep BVRI band images and H i line observations of the nearby, low surface brightness galaxy ESO215–G?009 which were obtained with the ANU 2.3-m Telescope and the Australia Telescope Compact Array, respectively. ESO215–G?009 was selected from the HIPASS Bright Galaxy Catalog because it has the second highest H i mass-to-light ratio of the galaxies with measured B band apparent magnitudes. We find that it is an isolated dwarf irregular galaxy with an old stellar population. We place an upper limit on the current star formation rate of ∼ 2.5 × 10−3 M⊙ yr−1. The extended H i disk shows regular rotation (vrot = 51 ± 8 km s−1), and at a column density of ∼ 5.0 × 1019 atoms cm−2 can be traced out to over six times the Holmberg radius of the stellar component (radius at µB = 26.6 mag arcsec−2). After foreground star subtraction, we measure a B band apparent magnitude of 16.13 ± 0.07 mag within a radius of 80′′. The H i flux density is 122 ± 4 Jy km s−1 within a radius of 370′′. Given a Galactic extinction of AB = 0.95 ± 0.15 mag, we derive an H i mass-to-light ratio of 22 ± 4 M⊙/L⊙,B for ESO215–G?009. To our knowledge this is the highest MHI/LB ratio for a galaxy to be confirmed by accurate measurement to date. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | One way to investigate aspects of star formation and galaxy evolution is to study galaxies which ap- pear to have done little of either, those with large quantities of unprocessed neutral hydrogen (H i) compared to their stellar content. The high quan- tities of H i gas in some low luminosity galaxies suggest that star formation within these galaxies has been impaired or halted, has lacked stimula- tion, or has only recently begun. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | parent magnitude, and AB is the B band Galactic extinction. Extinction correction from the host galaxy is not included (see Section 6.3). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | For most galaxies MHI/LB is typically less than 1 M⊙/L⊙,B, with late-type galaxies having a larger spread than early-types. Roberts & Haynes (1994) performed a statistical study of physical parameters along the Hubble sequence. For late- type galaxies they found a general trend of lower luminosity, surface brightness and H i mass, and higher MHI/LB and H i-to-total mass ratio than for early-type spirals. E.g., for the RC3 local Su- percluster sample they find a median MHI/LB of 0.78 M⊙/L⊙,B in type Sm/Im galaxies, with an inner quartile range of 0.44 M⊙/L⊙,B to 1.32 M⊙/L⊙,B, and a median MHI/Mtot of 0.15. However, some low surface brightness dwarf irreg- ular galaxies are know to have H i mass-to-light ratios that are significantly higher than this quar- tile range. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | One of the best known dwarf galaxies with a high MHI/LB is DDO 154, for which Hoffman et al. (1993) measured MHI/LB = 11 M⊙/L⊙,B. DDO 154, a dwarf galaxy of type Sm, has a very extended H i disk compared to its optical size, stretching to ∼6 times the Holmberg ra- dius (Carignan & Purton 1998). Another exam- ple is the Blue Compact Dwarf (BCD) galaxy NGC 2915, which has a moderately high MHI/LB of 2.6 M⊙/L⊙,B (Meurer, Mackie, & Carignan 1994; Meurer et al. 1996). The H i disk of this object extends to over five times the Holmberg radius and has distinct spiral arm structure (see Bureau et al. 1999). We do not consider MHI/LB ratios in tidal tails or bridges near galaxies, such as H i 1225+01, a tidally disturbed irregular galaxy described in detail by Chengalur, Giovanelli, & Haynes (1995). Likewise, massive gas-rich low sur- face brightness galaxies such as Malin 1 (Bothun et al. 1987; Impey & Bothun 1989) are not com- parable to low surface brightness dwarf systems with high MHI/LB ratios. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | al. (1997) found that the star formation process in the low surface brightness and normal galax- ies was inefficient, with the overall gas density well below the Toomre (1964) instability thresh- old. They measured some localized H i peaks which approached the threshold and found that these were associated with local star forming re- gions. Many of the galaxies in the later studies (van Zee 2000, 2001) are isolated, with no neigh- bors found within ∼200–400 kpc. van Zee (2001) concluded that unlike Blue Compact Dwarfs, star formation within these galaxies was occurring at a slow, constant rate across the galaxy disk with few major concentrations of activity, and that gas depletion timescales for these galaxies were of the order of 20 Gyr. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | While many galaxies with relatively high MHI/LB have been found, many more have gone unnoticed due in part to a bias towards opti- cally selected samples and the limitations of past blind H i surveys. The H i Parkes All-Sky Survey (HIPASS) has provided the first ever blind sur- vey of extragalactic neutral hydrogen in the local Universe (vsys < 12 700 km s−1) over the entire southern sky (Barnes et al. 2001). One of its first products is the HIPASS Bright Galaxy Catalog (hereafter BGC, Koribalski 2001; Koribalski et al. 2004), which lists the 1000 H i brightest sources in the southern hemisphere (selected by H i peak flux density). This catalog gives us the oppor- tunity to select interesting galaxies for detailed examination based on H i and optical properties rather than entirely on optical characteristics. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Fig. 1 shows a general trend for low luminosity galaxies in the BGC to have high MHI/LB. The low luminosity and low MHI/LB region of Fig. 1 is underpopulated as optically faint galaxies with lit- tle gas (dwarf elliptical galaxies) are not detected by HIPASS. We note that overall dwarf galaxies (those with the lowest luminosity) have a large spread in H i mass-to-light ratios. The lack of high MHI/LB bright galaxies, which should be easily detected, suggests that there is a physical limit to the H i mass-to-light ratio for a galaxy of a certain luminosity (stellar mass). Of the 148 galaxies in the BGC less luminous than MB,0 = −16.5 (here- after referred to as dwarfs), 50 have MHI/LB ≥ 3 M⊙/L⊙,B. Of these, seven galaxies have ini- tial ratios greater than 10 M⊙/L⊙,B and three have ratios in excess of 20 M⊙/L⊙,B. To un- derstand the nature of these galaxies, their abun- dance of neutral hydrogen gas and low stellar con- tent, we have started a multi-wavelength study of the galaxies with the highest H i mass to light ra- tios in the southern sky, selected from the BGC. The filled circle in Fig. 1 marks one of our sam- ple galaxies, ESO215–G?009, with MHI/LB = 24 ± 12 M⊙/L⊙,B, the second largest ratio in the BGC. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | In this paper, Section 2 looks at what was pre- viously known about ESO215–G?009. Section 3 summarizes our observations. Sections 4 and 5 present the new results and properties of ESO215– G?009 while Section 6 contains the discussion. Section 7 gives our conclusions. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | cal velocity is available to date. Lauberts & Valentijn (1989) measured a total blue magnitude of mB = 16.03 ± 0.09 mag from photographic plates. No previous H i observations are available other than those derived from HIPASS. Hucht- meier, Karachentsev, and Karachentseva (2001) published an MHI/LB for this galaxy (under the name KKs40) which was calculated using the H i flux density from an early public data release of HIPASS and the Lauberts & Valentijn (1989) pho- tographic blue magnitude. They found a high MHI/LB of ∼ 12 M⊙/L⊙,B, comparable with DDO 154. Unfortunately their H i flux density is an underestimate as the public data release spec- trum is off center and does not cover the entire galaxy. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Table 1 summarizes the properties of ESO215– G?009 from previous studies, values derived from those quantities, and some of the results of this study. Koribalski et al. (2004) measured a sys- temic velocity of vsys = 598 ± 1 km s−1, a velocity width of w20 = 83 ± 3 km s−1, and an integrated H i flux density of FHI = 104.4 ± 11.5 Jy km s−1. We adopt a distance to ESO215–G?009 of D = 4.2 Mpc from the Local Group velocity of 312 ± 1 km s−1 using D = vLG/H0 (see also Section 6.3). The proximity and isolation of ESO215–G?009 make it an excellent candidate for a study of the nature of high H i mass-to-light ratio galaxies, and of galaxy evolution in general. Using deep optical BVRI band images and H i data of ESO215–G?009 we will examine the stellar and H i distribution, stellar population, isolation and the H i kinematics to search for the physical reasons why ESO215– G?009 could maintain such a high fraction of H i gas, while other galaxies processed most of their gas into stars. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | The galaxy ESO215–G?009 was observed in two different wavelength regimes. Optical photometry was obtained with the Australian National Uni- versity (ANU) 2.3-meter Telescope at the Siding Springs Observatory. H i line and radio continuum data were obtained with the Australia Telescope Compact Array (ATCA). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Optical BVRI band (Cousins filters) images were taken as a series of 300 s exposures in April/June 2002 and May 2003 using the Nas- myth Imager (SITe 1124 × 1024 thinned CCD). The imager has a circular field of view with a di- ameter of 6.′62 and a pixel size of 0.′′59. The total exposure time in the four bands is 50, 40, 30, and 30 minutes, respectively. The typical seeing in the final images for the different bands was ∼1.′′9, 1.′′9, 1.′′8, and 2.′′0, respectively, and most observations were taken at low airmass. Twilight sky flat fields in all bands and bias images were obtained at the same time. Several Landolt (1992) standard stars were taken together with shallow 120 s BVRI im- ages of the galaxy field on a photometric night to perform the photometric calibration of the deeper images. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Data reduction and analysis was carried out with IRAF using standard procedures. After over- scan subtraction, bias subtraction, and flatfield- ing, individual 300s images were registered to each other and the sky level was subtracted. The im- ages for each band were then combined into a sin- gle image (to increase signal-to-noise, remove cos- mic rays, etc.) and the photometric calibration applied. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Foreground stars were removed by replacing them with the surrounding sky so that only the galaxy remained. Special care was taken to restore the light distribution under stars su- perimposed onto the galaxy, using the mirror image from across the galaxies center. Fig. 2 shows the resulting BVRI images before and after star subtraction. We measured the luminosity- weighted center of the galaxy (averaged be- tween the four bands) to be at α, δ(J2000) = 10h 57m 29.s9, −48◦ 10′ 43′′(±2′′), consistent with Lauberts (1982, see Table 1). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Data reduction and analysis was performed with the MIRIAD, AIPS, and KARMA pack- ages using standard procedures. Channels with Galactic emission were discarded. After con- tinuum subtraction, the H i data were Fourier- transformed using ‘natural’ weighting and a chan- nel width of 4 km s−1. The data were cleaned and restored with a synthesized beam of 38′′ × 35′′. We measure an r.m.s. noise of 1.6 mJy per chan- nel, close to the theoretical value. Primary beam correction was applied. H i moment maps were obtained using cutoffs of 4 mJy beam−1 (H i dis- tribution) and 6 mJy beam−1 (mean velocity field and dispersion). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | where D is the distance in Mpc and S20cm is the 20-cm radio continuum flux density in Jy. We cal- culate an upper limit to the star formation rate for ESO215–G?009 is ∼ 2.5 × 10−3 M⊙ yr−1. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | where the absolute Solar B magnitude, MB,⊙, is taken as 5.48 mag (Bessell, Castelli, & Plez 1998). We derive a B band luminosity of LB = (2.3 ± 0.4) × 107 L⊙,B, a factor of ∼80 below the median B luminosity of 1.9 × 109 L⊙,B for Sm/Im galaxies in Roberts & Haynes (1994). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | star in the field, which makes the star subtraction here more difficult and increased the uncertainty for any luminosity determination. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Table 2 gives a summary of the photometric results obtained from these profiles. No Galactic extinction correction was applied to the apparent magnitude and surface brightness. Column (1) gives the Cousins broad band filters used. Col- umn (2) gives the empirically derived total appar- ent magnitude. Column (3) gives the total ap- parent magnitude as extrapolated from the S´ersic profile fit. Column (4) gives the central surface brightness. Column (5) gives the effective surface brightness, the average surface brightness out to the half light radius, and Column (6) gives the half light (effective) radius. Column (7) gives the radii out to µ = 26.6 mag arcsec−2 (which is the Holmberg radius in the B band, extinction cor- rected). Column (8) gives the Galactic extinction correction from Schlegel et al. (1998). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | We obtained six colors from the photometry, which are shown at the end of Table 3 (which we will discuss in Section 6.6). From these colors it seems that ESO215–G?009 is very red for a late- type galaxy, suggesting that the stellar population is older than is typically seen in dwarf irregulars (see discussion in Section 6.4). Fig. 4 shows the color profiles for each of these six color combina- tions. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Fig. 5 shows the H i channel maps of ESO215– G?009. H i line emission was detected in the ve- locity range 547 to 647 km s−1. The emission has a resemblance to a slowly rotating gas disk, except that the pattern is not as sharply defined as might be seen in a larger disk (see example in Giovanelli & Haynes 1988). Also, the structure throughout the channels is not evenly distributed with sev- eral concentrations and voids. In particular, to the north of the galaxy center the gas has two strong peaks close to the systemic velocity, and to the south east of the center a depression is visible around 609 km s−1. We find no companion galax- ies within the ATCA primary beam (33′) and the velocity range of 200 – 1300 km s−1. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | ATCA. There is a resemblance to a double horn profile, characteristic of a rotating disk, although the central dip is not pronounced (which may be partially due to the low inclination of the source). Notably the ATCA spectrum is inconsistent with the single dish result, and measures a higher flux density. The HIPASS BGC flux density may be underestimated due to the large extent and high peak flux density of the source (see Barnes et al. 2001). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Fig. 7 shows the integrated H i distribution, the mean H i velocity field and the H i velocity dispersion of ESO215–G?009. From the H i dis- tribution we measure the dimensions of the H i disk as 730′′ × 640′′ (±40′′) at a position angle of P A ∼ 120◦. Fig. 8 shows the azimuthally averaged radial H i density profile of ESO215–G?009. The profile was measured with rings 10′′ wide assuming an inclination of 36◦ and PA = 119◦ (Section 5.2). Here 1 Jy beam−1 km s−1 corresponds to a col- umn density of 8.32 × 1020 atoms cm−2. H i can be traced out to a column density of ∼ 5.0 × 1019 atoms cm−2 at a radius of 370′′ ± 20′′. We obtain an H i flux density of FHI = 122±4 Jy km s−1. This is higher than the FHI = 104.4±11.5 Jy km s−1 ob- tained in the BGC (Koribalski et al. 2004). Using the adopted distance of 4.2 Mpc we derive an H i mass of MHI = (5.1 ± 0.2) × 108 M⊙. This is less than half the median H i mass of 1.27×109 M⊙for Sm/Im galaxies in Roberts & Haynes (1994), but is still well within the range of H i masses seen in this sample (see Table 3). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Through most of the galaxy the velocity disper- sion (see Fig. 7d) is between about 4 to 10 km s−1. In the central region around the stellar component of the galaxy, and extending to the north towards the strongest H i concentration, the velocity dis- persion is largest. This is reflected in Fig. 9, which shows the azimuthally averaged radial H i velocity dispersion profile of ESO215–G?009. The velocity dispersion is highest in the center of the galaxy where the profile is peaked. In the outer galaxy the dispersion is lower with a slightly declining profile. Fig. 10 shows an H i position velocity dia- gram along the major axis of ESO215–G?009, em- phasizing the symmetry between the approaching and receding sides of the galaxy. On the reced- ing side there are two gaps in the gas distribution accompanied by a large spread in velocity. These features could be expanding bubbles which have been recognized in other face-on galaxies (such as the large H i ‘superbubble’ in M101, Kamphuis, Sancisi, & van der Hulst 1991). They correspond to the low density regions in Fig. 7b to the south east. There is no evidence of a central H i bar in Fig. 10, however a weak bar on the scale of the optical component of the galaxy would be difficult to pick up from this diagram with our current res- olution. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | We obtained a rotation curve from the mean H i velocity field of ESO215–G?009 using the tilted ring algorithm in AIPS (Begeman 1989, rocur). Using all data within a radius of 6′, we derive the dynamical center of the galaxy at α, δ(J2000) = 10h 57m 30.s0, –48◦ 10′ 47′′ (±10′′), and its systemic heliocentric velocity, vsys = 597 ± 1 km s−1. This is consistent with Koribalski et al. (2004), and the position agrees with that derived from the center of the stellar component. The center and systemic velocity were fixed for the fitting of the remain- ing parameters for which we used 12 rings of 35′′ width. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | ceding sides of the galaxy revealed some differ- ences. The inclination for the best fit to the outer part of the galaxy on the receding side was i = 40◦±6◦ (beyond 150′′), while for the approach- ing side it was i = 31◦ ± 4◦. Fig. 11 includes the rotation curves for both sides fit with the same inclination angle as the overall fit, 36◦. There is a slight turn down on the approaching side, while the receding side continues rising slightly. Over- all the three rotation curves are reasonably flat beyond 150′′. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | With i = 36◦ ± 10◦ we find a rotational velocity of vmax = 51 ± 8 km s−1 out to a radius of 370′′ ± 20′′ (∼7.5 kpc). We derive a total dynamical mass of Mtot = (4.5 ± 1.6) × 109 M⊙. This gives an H i to dynamical mass ratio of MHI/Mtot = 0.11 ± 0.04, at the low end of what Roberts & Haynes (1994) found for Sm/Im galaxies (median value 0.15). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Using Eqn. 1, we can now recalculate MHI/LB using the ATCA integrated H i flux density of 122±4 Jy km s−1, the B band apparent magnitude of 16.13 ± 0.07 mag, and the Schlegel et al. (1998) Galactic extinction of 0.95 ± 0.15 mag. We de- rive an H i mass-to-light ratio for ESO215–G?009 of 22 ± 4 M⊙/L⊙,B. A large portion of the uncer- tainty is due to the uncertainty in the Galactic ex- tinction. This is in good agreement with the pre- liminary ratio from the BGC of 24±12 M⊙/L⊙,B. To our knowledge ESO215–G?009 has one of the highest (if not the highest) MHI/LB to be con- firmed by accurate measurement to date for any galaxy system, being approximately double the MHI/LB of DDO 154 (Hoffman et al. 1993). Us- ing Mtot = (4.5 ± 1.6) × 109 M⊙ and LB = (2.3 ± 0.4) × 107 L⊙,B, the dynamical mass-to- light ratio is Mtot/LB = 200 ± 110 M⊙/L⊙,B, making ESO215–G?009 an extremely dark matter dominated galaxy. If we assume a stellar mass-to- light ratio of ∼1 M⊙/L⊙,B, we estimate a bary- onic mass to total mass ratio of Mbary/Mtot = 0.12. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | dius is only 57.′′6 ± 0.′′6 (at the extinction corrected surface brightness of 26.6 mag arcsec−2). There- fore the gas extends to 6.4 ± 0.4 times the Holm- berg radius, comparable to the extent of DDO 154 (Carignan & Purton 1998). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | The brightest part of the central stellar region (see Fig. 2) is slightly elongated in a direction aligned much closer to the rotational minor axis (∼40◦). As the galaxy is seen at a low inclina- tion this feature is potentially a weak central bar. The optical morphology of ESO215–G?009 is oth- erwise rather smooth and featureless for a dwarf irregular, although its slightly mottled appearance in color images, the slowly rotating H i disk, and possible H ii regions (see Section 6.4) imply it is a late-type dwarf galaxy. The stellar distribution is located in the center of the H i disk at the position of a shallow depression in the H i distribution. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | AB = 3.17 ± 0.51 mag (Schlegel et al. 1998), and Burstein & Heiles (1982) give 2.84 mag. They may have mistaken AB for EB − V . Karachentsev & Musella (1996) performed the first CCD pho- tometry on this galaxy, getting a new mB of 15.7 mag, and using an extinction of 2.7 mag derived MHI/LB = 2.6 M⊙/L⊙,B. Barazza, Binggeli, & Prugniel (2001) found a similar B magnitude of 15.41 mag. Combined with the rise in the Galac- tic extinction, this galaxy appears approximately 5 magnitudes brighter than in van Zee et al. (1995), which means MHI/LB is overestimated by a fac- tor of around 100. Using the Barazza et al. (2001) B magnitude, the Schlegel et al. (1998) Galactic extinction, and the H i flux density from HIPASS (FHI = 73 ± 4 Jy km s−1), a more accurate H i mass-to-light ratio for the Orion dwarf (HIPASS J0545+05) is 0.9 ± 0.5 M⊙/L⊙,B. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Problems can also arise with the H i obser- vations. O’Neil, Bothun, & Schombert (2000) find four galaxies in the range 10 M⊙/L⊙,B < MHI/LB < 50 M⊙/L⊙,B as part of an Arecibo follow-up of a group of low surface brightness galaxies, as well as six galaxies which deviated from the Tully-Fisher relationship. These were the first extragalactic observations taken after an up- grade at Arecibo. Later Chung et al. (2000) found that many of the galaxies observed by O’Neil et al. (2000) had been contaminated by other neigh- boring objects within the sidelobes (see Heiles et al. 2001). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | We conclude that to obtain accurate H i mass- to-light ratios it is vitally important that the H i flux density, the B band apparent magnitude and the Galactic extinction are well determined; er- ror bars on all are essential and should always be quoted. Galaxies located near the Galactic Plane, where the extinction is high and uncertainty in- creases, are particularly difficult and should be treated with extra care. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | can calculate a distance from the Baryonic Tully- Fisher relationship (McGaugh et al. 2000) by first estimating a baryonic mass for ESO215–G?009 from the relationship given the galaxy’s rotation velocity of 51 ± 8 km s−1. We estimate Mbary = 2.1 × 108 M⊙. Given the H i flux density and the B band apparent magnitude then ESO215–G?009 would have to be at a distance of 2.6 Mpc to have this baryonic mass, almost half the Hubble flow distance. At the Hubble flow distance ESO215– G?009 would lie above the Baryonic Tully-Fisher relationship. Table 4 gives examples of the effect of this uncertainty on the main distance depen- dent properties in this paper. It shows the derived quantities at both distances, with the quoted un- certainty accounting for all other sources of error. Uncertainty in the estimated distance can have a large effect on properties such as B band luminos- ity and H i mass, both of which change by a factor of ∼2.5 between the two estimates. The Baryonic Tully-Fisher distance was not adopted in this pa- per as there is currently large uncertainty at the low mass end of the relation. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Galactic dust extinction can strongly affect the final MHI/LB result, as we have seen in Sec- tion 6.2. The relatively high Galactic extinction in the direction of ESO215–G?009 and the high number of foreground stars may mean that we could ultimately miss some very faint components of the galaxy, and so underestimate the total ap- parent magnitude. There is also the question of whether the value of Galactic extinction we are us- ing is completely wrong. The Schlegel et al. (1998) Galactic extinction is measured over large scales, and there may be smaller scale variation. The red- der color of the galaxy may suggest stronger ex- tinction than we are accounting for. To test this we looked at the extinction correction for galaxies in the surrounding field to see if there was a large extinction gradient in the area. We found that out to a radius of 1◦ the variation was between 0.8 and 1.1 mag, within the uncertainties of the extinction at ESO215–G?009’s position, and so we are confi- dent that AB is not too far from the Schlegel et al. (1998) value. However we note that close to the Galactic Plane the dust distribution can be patchy, and it is entirely possible that the extinction could be much higher at the position of ESO215–G?009 than current measurements indicate. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | must also be considered as that may reduce the to- tal light we receive and redden the galaxy. It is not possible currently to estimate what the dust ex- tinction could be in this galaxy as internal extinc- tion is poorly understood in dwarf irregular galax- ies, though it is often assumed to be low. The in- clination of the galaxy is quite low as seen from the rotation curve analysis, and assuming that most of the galaxy’s dust is in the plane of the H i, we find it unlikely that internal dust extinction has a large effect on the luminosity of ESO215–G?009. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | The photometric colors (see end of Table 3) are all fairly red for a dwarf irregular galaxy, sug- gesting that ESO215–G?009 contains a mainly old stellar population, which is uncommon in late- type galaxies. Parodi, Barazza, & Binggeli (2002) find the mean color for late-type galaxies in their sample is hB − RiLate = 0.90, while for the small number of early-types in their sample it is hB − RiEarly = 1.21. Their reddest late-type galaxy is ESO006–G001 with B − R = 1.24, a galaxy they claim has an intermediate morphol- ogy (both late- and early-type). ESO215–G?009’s value of (B − R)0 = 1.39 ± 0.22 is closer to the colors of the dwarf ellipticals and the more ex- treme late type dwarfs in the sample. The median B − V color for 210 Sm/Im galaxies in the Lo- cal Supercluster from Roberts & Haynes (1994) is 0.42. Again ESO215–G?009 has a much red- der color, (B − V )0 = 1.02 ± 0.21. It should be noted that the uncertainties in these colors are quite high due to the position of the galaxy close to the Galactic Plane. Unfortunately these higher uncertainties due to Galactic extinction prevent us from performing a detailed analysis of the age and metalicity of the underlying stellar population. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | stars in the inner regions and the older/redder population in the outer regions. This is the op- posite to larger spiral systems where spiral den- sity waves drive star formation in the outer disk and the central region is dominated by an older red bulge. Of the late-type galaxies in the Parodi et al. (2002) sample only eight show a positive gradient in B −R, while a further eight showed a flat profile and a few showed negative profiles. In this color ESO215–G?009 also shows a mostly flat trend. It appears that there may be a slightly higher pro- portion of younger stars in the center of ESO215– G?009, but as we don’t see it in all colors this trend is unlikely to be strong. van Zee (2001) find that color gradients in dwarf irregular galaxies in a quiescent phase (non-starbursting) are generally not as pronounced as those of starburst galaxies (such as BCDs, where there is strong star forma- tion confined to a small central region). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | During the star subtraction of the optical im- ages (Fig. 2) several bright slightly extended ob- jects close to the galaxy which we believe are H ii regions were not removed. The two most promi- nent can be seen best in the B image, on the outer north-eastern edge of the galaxy and close in on the north-western side. These regions appear to be rather blue compared to the rest of the galaxy. However they may be background galaxies or fore- ground features rather than H ii regions associated with ESO215–G?009. To try and confirm that these are indeed star forming regions we examined the SuperCOSMOS H-alpha Survey (SHS) image of the field around ESO215–G?009 (Hambly et al. 2001, and references therein). The image can be seen in Fig. 12 with the contours from our 20-cm radio continuum observation of the field overlaid. The two regions mentioned show up distinctly as diffuse objects, while the main galaxy is barely visible at all. Two further diffuse regions are visi- ble well beyond the optical galaxy, and almost to the edge of the H i disk on the south-eastern side. However, none of these diffuse regions in the Hα image correspond to any of the 20-cm continuum sources in the field. All of these objects could be H ii regions in ESO215–G?009, but this needs to be confirmed with follow up spectra and Hα imag- ing. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | where ΓB is the B-band stellar mass-to-light ratio in M⊙/L⊙,B, Tsf is the time for which the galaxy has been forming stars in years, and R is the re- cycling fraction (essentially the proportion of the stars formed within the galaxy that are no longer visible). |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | Kennicutt (1989) and Martin & Kennicutt (2001) (hereafter K89 and MK01, respectively) have found that gravitational instability may de- termine the star formation law for galactic disks. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | where σ(r) is the gas velocity dispersion, κ(r) is the epicyclic frequency, and αQ is a constant close to unity which is included to account for a more realistic disk (i.e. thicker disk, embedded stellar population). K89 derive αQ = 0.631, while for the MK01 sample (which was larger and had more de- tailed input parameters) they found αQ = 0.69. If µgas(r), the surface density of the gas in the disk, exceeds µcrit(r) then the disk will be unsta- ble to axisymmetric disturbances and large scale star formation can occur. MK01 also found that the simple Toomre criterion breaks down in some circumstances, such as low shear rates and stel- lar bars, and that there may be a weak trend for late-type galaxies to have higher αQ. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | where V is the rotation velocity in km s−1, and R is the radius in km. The derivative was deter- mined with a forward difference method. We have assumed αQ = 1. Only the azimuthally averaged H i surface density (µHI(r)) was available for the gas surface density. We have made no estimate of the molecular hydrogen content (as was done in K89) for ESO215–G?009 at this stage. |
Processed_ESO215-G?009:_An_Extreme_HI-Rich_Dwarf_Irregular_G.txt | can be seen in Fig. 13. Over all radii the ra- tio is less than the MK01 value of αQ = 0.69, strongly suggesting that the H i gas surface den- sity of ESO215–G?009 is not high enough to in- duce large scale star formation within the disk. The profile is similar to the galaxies in both K89 and MK01 with low levels of star formation. How- ever, the majority of those galaxies are early-type spiral galaxies (Sa to Sb), not dwarf galaxies. The exclusion of molecular gas from our estimate may have caused a significant underestimate of µgas(r)/µcrit(r), which may lead to part of the profile rising above the threshold. The estimate which K89 used (µgas(r) = 1.41µHI(r)) may not be appropriate for ESO215–G?009 or other dwarf irregular galaxies, so the only way to resolve this issue would be to obtain CO observations to esti- mate the H2 content (which may be problematic for ESO215–G?009 if the metalicity and excita- tion are low). It is important to note also that, as MK01 mention, using the azimuthally averaged H i surface density may not be appropriate for galax- ies like ESO215–G?009 where the variation in den- sity at a particular radius can be large. Several re- gions of the galaxy have H i densities much higher than the azimuthal average at their location (such as the region to the north of the galaxy center) and local densities may approach or exceed the threshold density. |
Processed_Evidential_Force_Aggregation.txt | Abstract - In this paper we develop an evidential force aggregation method intended for classification into recognized force of structures. We assume that the intelligence has already been partitioned into clusters and use the classification method individually in each cluster. The classification is based on a measure of fitness between template and fused intelligence that makes it possible to handle intelligence reports with multiple nonspecific and uncertain propositions. With this measure we can aggregate on a level-by-level basis, intelligence to achieve a starting from general complete force structure with recognized units on all hierarchical levels. |
Processed_Evidential_Force_Aggregation.txt | We define force aggregation as a combination of two processes. First, an association of intelligence reports, objects or units (depending on hierarchical level) by a clustering process [1−5] the left column in Figure 1. Secondly, a classification of cluster content through a comparison with templates, the right column in Figure 1. In Figure 1 such clustering and classification is performed on all hierarchical levels, force level-by-level aggregation of all units. |
Processed_Evidential_Force_Aggregation.txt | Evidential force aggregation is force aggregation from uncertain information. The classification in evidential force aggregation is the focus of this paper. The work described herein is an extension of previous work. In [6] we restricted each intelligence to carry only one proposition that could be report i.e., specific or nonspecific regarding object support any subset of all possible types, but was always certain. Here, in this paper, we allow for any number of nonspecific and uncertain propositions in each intelligence report. With this extension we may handle any general intelligence report. |
Processed_Evidential_Force_Aggregation.txt | The classification process deals with intelligence reports on a cluster-by-cluster basis. Looking at intelligence in one of the clusters, the classification from intelligence by templates take place in two phases. First, we combine all intelligence reports within the cluster, and secondly, we compare the combined intelligence with all available templates. |
Processed_Evidential_Force_Aggregation.txt | special concern is the representation used. As the reports in general are not reports about the same object or group of objects, we must not use a simple representation dealing only with object type. Instead, we must use a more advanced representation that allow us to keep track of different objects and their possible types. Intelligence reports that actually are referring to the same object or group of objects are precombined, and henceforth viewed as one intelligence report. When intelligence reports in the cluster this is done, all under investigation can be combined, giving us the possibility to investigate resulting hypothesis regarding force composition. |
Processed_Evidential_Force_Aggregation.txt | in Proc. Sixth Int. Conf. Information Fusion (Fusion 2003), pp. 1223−1229, Cairns, July 2003, Int. Soc. Inf. Fusion, Sunnyvale, CA, 2003. |
Processed_Evidential_Force_Aggregation.txt | absence of conflicts in numbers between what the intelligence propose and what each available template request for all subsets of types. With this measure we can select a template for intelligence with nonspecific propositions. |
Processed_Evidential_Force_Aggregation.txt | In Sect. 2 we describe the representation of intelligence and their combination. An example is given in Sect. 2.1. In Sect. 3 we describe the representation of templates and their evaluation and selection through a comparison with intelligence. A continuation of the example is presented in Sect. 3.1. An evidential force aggregation algorithm based on the result of the two previous sections is presented in Sect. 4. Finally, conclusions are drawn (Sect. 5). |
Processed_Evidential_Force_Aggregation.txt | object or set of objects they are precombined into Multiple nonalternative propositions objects than the •n j handled as reports. |
Processed_Evidential_Force_Aggregation.txt | We will here investigate the representation and combination of all intelligence referring to the same unit. We assume that a number of intelligence reports about different set of objects are available. These reports have already been partitioned into subsets where each subset corresponds to a unit on one hierarchical level higher [2, 3]. Let us hereafter focus on one such subset χ a and the aggregation of the intelligence in this subset. |
Processed_Evidential_Force_Aggregation.txt | In order to be able to handle reports about different objects that should not be combined on the object level, but should be viewed as fragments of a larger unit to be all combined, we need to refine our representation. Each report is now corresponding to a unique position in a unit structure. |
Processed_Evidential_Force_Aggregation.txt | stating that we have a 0.6 basic probability of one to two vehicles that are MBTs or APCs. |
Processed_Evidential_Force_Aggregation.txt | As we are also interested in the types of objects and their number regardless of their ordering, we sum up all contributions regarding the same type. |
Processed_Evidential_Force_Aggregation.txt | e.g., the direct sum {1, 2} ⊕ {2, 3} = {1+2, 1+3, 2+2, 2+3} = {3, 4, 4, 5} = {3, 4, 5}. This gives us information about different propositions in the initial representation with as frame of discernment. The result of Eq. (16) will not be used in the selection process of finding a template with maximal fitness for the intelligence in χ a. However, it is a result in itself that may be communicated for other purposes. |
Processed_Evidential_Force_Aggregation.txt | using Eq. (15). This result will be used in the next section to select a template with maximum fitness towards the intelligence. return previous the Temporarily, we representation using as the frame of discernment in order to obtain a basic probability assignment for each supported subset of all types of objects TY. |
Processed_Evidential_Force_Aggregation.txt | stating that we have a 0.5 basic probability in favor of two MBTs, a 0.3 basic probability in favor of two vehicles that are either MBTs or APCs. |
Processed_Evidential_Force_Aggregation.txt | This summarizes the support for each supported subset of all types TY. Note, that the first two sum to 1.0, and the four last sum to 1.0 as these are two different assignments. |
Processed_Evidential_Force_Aggregation.txt | Comparing templates having specific propositions that are certain in what they are requesting with intelligence propositions that are not only uncertain but may also be nonspecific in what they are supporting can be a difficult task. The idea we use to handle this problem is to compare a candidate template with intelligence from the perspective of each and every subset of all possible types of objects TY. |
Processed_Evidential_Force_Aggregation.txt | In doing this we investigate how much support a subset of TY receives both directly and indirectly from intelligence and template, respectively. The support for a subset of TY is summed up from all propositions that are equal to or itself a subset of this subset of TY. This is similar to the calculation of belief from basic probability numbers in Dempster- Shafer theory, except that we are not summing up basic probability numbers but natural numbers representing the number of objects of the proposed types. |
Processed_Evidential_Force_Aggregation.txt | as the measure of fitness of Ti towards all intelligence in χ a. This is the measure by which we rank all templates and make our selection of template. |
Processed_Evidential_Force_Aggregation.txt | For example, from the perspective of {MBT, APC} a template proposition of “four MBTs” lend indirect support to {MBT, APC} since {MBT} is a subset of {MBT, APC}, and intelligence proposing “two MBTs or APCs” lend direct support to the subset. With the summed up numbers being four and two, respectively, and we intelligence from the perspective of {MBT, APC}. We use this method to rank all templates based on a fitness measure of template to intelligence matching taking all subsets of TY into account. |
Processed_Evidential_Force_Aggregation.txt | with , is a measure of fitness looking for a worst matching between Ti and this proposition for all different subsets of all types TY. |
Processed_Evidential_Force_Aggregation.txt | By using the result obtained by Eq. (15) from the combination of all mass functions in Ja, we compare different templates in order to find a template with maximum fitness towards the set of intelligence reports. |
Processed_Evidential_Force_Aggregation.txt | the fitness of Ti on a type-by-type basis demanding a perfect fit for all types to give a full score, the second type entirely, and compare the measure ignores number of objects of all intelligence in χ a with the same in the template. While the first measure seems preferable it can be too extreme when considering of objects missing data for same small number in Ti. |
Processed_Evidential_Force_Aggregation.txt | also need the basic probability of the highest ranked template for any further aggregation from the next hierarchical level. |
Processed_Evidential_Force_Aggregation.txt | Each focal element in the resulting combination support a subset of all templates. Through a fitness weighted transformation, these templates will share this support in relation to their fitness towards the corresponding focal element in the intelligence. |
Processed_Evidential_Force_Aggregation.txt | In order to evaluate both templates T1 and T2 we ( calculate to find their measure J a of fitness towards the fused intelligence. To calculate these measures we must first calculate the number of objects requested by the templates for each subset of TY using Eq. (32). Using Eq. (33) with the fused intelligence give us the number of objects supported by different propositions in the fused intelligence. In Table 1 we have tabulated the result of ST1(.) and ST2(.) for all propositions in the fused intelligence. |
Processed_Evidential_Force_Aggregation.txt | the measure of fitness of each template towards every 〈 proposition in the intelligence given each subset of TY, in Table 2 using Eq. (30). In the .( ) same manner we find , i.e., the same measure of 〈 , , … x I a fitness towards all propositions in the intelligence, but given the set of all types TY (Table 3, using Eq. (31)). |
Processed_Evidential_Force_Aggregation.txt | From Table 2 we find the minimum fitness for both templates towards every proposition in the fused intelligence given all subsets of TY (tabulated in Table 4, using Eq. (29)). |
Processed_Evidential_Force_Aggregation.txt | as the first two focal elements intersects to {T1} and the remainder to {T1, T2}. |
Processed_Evidential_Force_Aggregation.txt | a as represented in Ja. Secondly, calculate the basic probability , , number for all propositions in x1 x2 the result from ⊕Ja, using Eq. (15). |
Processed_Evidential_Force_Aggregation.txt | At the same time, calculate the number of objects supported by intelligence SCa(.|〈.〉) for each subset of all each proposition in the intelligence using Eq. (33), and calculate the number of objects requested by each template STi(.) for each subset of all types TY and for each proposition in the intelligence, using Eq. (32). |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.