We present optical imaging and spectroscopy and HI imaging of the Virgo Cluster galaxy IC 3418, which is likely a smoking gun example of the transformation of a dwarf irregular into a dwarf elliptical galaxy by ram pressure stripping. IC 3418 has a s
pectacular 17 kpc length UV-bright tail comprised of knots, head-tail, and linear stellar features. The only H-alpha emission arises from a few HII regions in the tail, the brightest of which are at the heads of head-tail UV sources whose tails point toward the galaxy (fireballs). Several of the elongated tail sources have H-alpha peaks outwardly offset by 80-150 pc from the UV peaks, suggesting that gas clumps continue to accelerate through ram pressure, leaving behind streams of newly formed stars which have decoupled from the gas. Absorption line strengths, measured from Keck DEIMOS spectra, together with UV colors, show star formation stopped 300+/-100 Myr ago in the main body, and a strong starburst occurred prior to quenching. While neither H-alpha nor HI emission are detected in the main body of the galaxy, we have detected 4x10^7 M_sun of HI from the tail with the VLA. The gas consumption timescale in the tail is relatively long, implying that most of the stripped gas does not form stars but joins the ICM. The velocities of tail HII regions, measured from Keck LRIS spectra, extend only a small fraction of the way to the cluster velocity, suggesting that star formation does not happen in more distant parts of the tail. Stars in the outer tail have velocities exceeding the escape speed, but some in the inner tail should fall back into the galaxy, forming halo streams. One likely fallback stream is identified.
The Milky Way ultra-faint dwarf galaxies (UFDs) contain some of the oldest, most metal-poor stars in the Universe. We present [Mg/Fe], [Si/Fe], [Ca/Fe], [Ti/Fe], and mean [alpha/Fe], abundance ratios for 61 individual red giant branch stars across 8
UFDs. This is the largest sample of alpha abundances published to date in galaxies with absolute magnitudes M_V > -8, including the first measurements for Segue 1, Canes Venatici II, Ursa Major I, and Leo T. Abundances were determined via medium-resolution Keck/DEIMOS spectroscopy and spectral synthesis. The sample spans the metallicity range -3.4 < [Fe/H] < -1.1. With the possible exception of Segue 1 and Ursa Major II, the individual UFDs show on average lower [alpha/Fe] at higher metallicities, consistent with enrichment from Type Ia supernovae. Thus even the faintest galaxies have undergone at least a limited level of chemical self-enrichment. Together with recent photometric studies, this suggests that star formation in the UFDs was not a single burst, but instead lasted at least as much as the minimum time delay of the onset of Type Ia supernovae (~100 Myr) and less than ~2 Gyr. We further show that the combined population of UFDs has an [alpha/Fe] abundance pattern that is inconsistent with a flat, Galactic halo-like alpha abundance trend, and is also qualitatively different from that of the more luminous CVn I dSph, which does show a hint of a plateau at very low [Fe/H].
We present new constraints on the star formation histories of the ultra-faint dwarf (UFD) galaxies, using deep photometry obtained with the Hubble Space Telescope (HST). A galaxy class recently discovered in the Sloan Digital Sky Survey, the UFDs app
ear to be an extension of the classical dwarf spheroidals to low luminosities, offering a new front in efforts to understand the missing satellite problem. They are the least luminous, most dark-matter dominated, and least chemically-evolved galaxies known. Our HST survey of six UFDs seeks to determine if these galaxies are true fossils from the early universe. We present here the preliminary analysis of three UFD galaxies: Hercules, Leo IV, and Ursa Major I. Classical dwarf spheroidals of the Local Group exhibit extended star formation histories, but these three Milky Way satellites are at least as old as the ancient globular cluster M92, with no evidence for intermediate-age populations. Their ages also appear to be synchronized to within ~1 Gyr of each other, as might be expected if their star formation was truncated by a global event, such as reionization.
We investigate the kinematic and photometric properties of the Segue 3 Milky Way companion using Keck/DEIMOS spectroscopy and Magellan/IMACS g and r-band imaging. Using maximum likelihood methods to analyze the photometry, we study the structure and
stellar population of Segue 3. We find the half-light radius of Segue 3 is 26 +/- 5 (2.1 +/- 0.4 pc, for a distance of 17 kpc) and the absolute magnitude is a mere M_V = 0.0 +/- 0.8 mag, making Segue 3 the least luminous old stellar system known. We find Segue 3 to be consistent with a single stellar population, with an age of 12.0 +1.5/-0.4 Gyr and an [Fe/H] of -1.7 +0.07/-0.27. Line-of-sight velocities from the spectra are combined with the photometry to determine a sample of 32 stars which are likely associated with Segue 3. The member stars within three half-light radii have a velocity dispersion of 1.2 +/- 2.6 km/s. Photometry of the members indicates the stellar population has a spread in [Fe/H] of <0.3 dex. These facts, together with the small physical size of Segue 3, imply the object is likely an old, faint stellar cluster which contains no significant dark matter. We find tentative evidence for stellar mass loss in Segue 3 through the eleven candidate member stars outside of three half-light radii, as expected from dynamical arguments. Interpretation of the data outside of three half-light radii, is complicated by the objects spatial coincidence with a previously known halo substructure, which may enhance contamination of our member sample.
We present the results of a comprehensive Keck/DEIMOS spectroscopic survey of the ultra-faint Milky Way satellite galaxy Segue 1. We have obtained velocity measurements for 98.2% of the stars within 67 pc (10 arcmin, or 2.3 half-light radii) of the c
enter of Segue 1 that have colors and magnitudes consistent with membership, down to a magnitude limit of r=21.7. Based on photometric, kinematic, and metallicity information, we identify 71 stars as probable Segue 1 members, including some as far out as 87 pc. After correcting for the influence of binary stars using repeated velocity measurements, we determine a velocity dispersion of 3.7^{+1.4}_{-1.1} km/s, with a corresponding mass within the half-light radius of 5.8^{+8.2}_{-3.1} x 10^5 Msun. The stellar kinematics of Segue 1 require very high mass-to-light ratios unless the system is far from dynamical equilibrium, even if the period distribution of unresolved binary stars is skewed toward implausibly short periods. With a total luminosity less than that of a single bright red giant and a V-band mass-to-light ratio of 3400 Msun/Lsun, Segue 1 is the darkest galaxy currently known. We critically re-examine recent claims that Segue 1 is a tidally disrupting star cluster and that kinematic samples are contaminated by the Sagittarius stream. The extremely low metallicities ([Fe/H] < -3) of two Segue 1 stars and the large metallicity spread among the members demonstrate conclusively that Segue 1 is a dwarf galaxy, and we find no evidence in favor of tidal effects. We also show that contamination by the Sagittarius stream has been overestimated. Segue 1 has the highest measured dark matter density of any known galaxy and will therefore be a prime testing ground for dark matter physics and galaxy formation on small scales.
We investigate the kinematic properties and stellar population of the Galactic satellite Willman 1 (Wil 1) by combining Keck/DEIMOS spectroscopy with KPNO mosaic camera imaging. Wil 1 is an ultra-low luminosity Milky Way companion. This object lies i
n a region of size-luminosity space (M_V ~ -2 mag, d ~ 38 kpc, r_half ~ 20 pc) also occupied by the Galactic satellites Bootes II and Segue 1 and 2, but no other known old stellar system. We use kinematic and color-magnitude criteria to identify 45 stars as possible members of Wil 1. With a systemic velocity of v_helio = -12.8 +/- 1.0 km/s, Wil 1 stars have velocities similar to those of foreground Milky Way stars. Informed by Monte-Carlo simulations, we identify 5 of the 45 candidate member stars as likely foreground contaminants. We confirm a significant spread in the abundances of the likely Wil 1 red giant branch members ([Fe/H] = -1.73 +/- 0.12 and -2.65 +/- 0.12, [Ca/Fe] = -0.4 +/- 0.18 and +0.13 +/- 0.28). This spread supports the scenario that Wil 1 is an ultra-low luminosity dwarf galaxy rather than a star cluster. Wil 1s innermost stars move with radial velocities offset by 8 km/s from its outer stars and have a velocity dispersion consistent with 0 km/s, suggesting that Wil 1 may not be in dynamical equilibrium. The combination of the foreground contamination and unusual kinematic distribution make it difficult to robustly determine the dark matter mass of Wil 1. As a result, X-ray or gamma-ray observations of Wil 1 that attempt to constrain models of particle dark matter using an equilibrium mass model are strongly affected by the systematics in the observations presented here. We conclude that, despite the unusual features in the Wil 1 kinematic distribution, evidence indicates that this object is, or at least once was, a dwarf galaxy.
We present new metallicity measurements for 298 individual red giant branch stars in eight of the least luminous dwarf spheroidal galaxies (dSphs) in the Milky Way (MW) system. Our technique is based on medium resolution Keck/DEIMOS spectroscopy coup
led with spectral synthesis. We present the first spectroscopic metallicities at [Fe/H] < -3.0 of stars in a dwarf galaxy, with individual stellar metallicities as low as [Fe/H] = -3.3. Because our [Fe/H] measurements are not tied to empirical metallicity calibrators and are sensitive to arbitrarily low metallicities, we are able to probe this extremely metal-poor regime accurately. The metallicity distribution of stars in these dSphs is similar to the MW halo at the metal-poor end. We also demonstrate that the luminosity-metallicity relation previously seen in more luminous dSph galaxies (M_V = -13.4 to -8.8) extends smoothly down to an absolute magnitude of M_V = -3.7. The discovery of extremely metal-poor stars in dSphs lends support to the LCDM galaxy assembly paradigm wherein dwarf galaxies dissolve to form the stellar halo of the MW.
The galaxy circular velocity function at small masses is related to the matter power spectrum on small scales. Although this function is well-studied for Local Group dwarfs, theoretical predictions and observational measurements are difficult for sat
ellite galaxies, because of ram pressure and tidal stripping. By contrast, isolated dwarf galaxies are less affected by these processes, and almost always have enough 21cm emission to trace their dynamics robustly. Here, we test cold dark matter cosmology using isolated low mass dwarf galaxies from the SDSS with measured 21cm widths. We find consistency between the predicted and observed number density of isolated galaxies down to circular velocities of 50 km/s. Our technique yields a direct test of small-scale cosmology independent of the Lyman-alpha forest power spectrum, but our sample is currently statistically less powerful: warm dark matter particles heavier than 0.5 keV cannot be ruled out. Our major systematic uncertainty is the surface brightness limit of the SDSS. Blind HI surveys, such as the ALFALFA survey on Arecibo, will uncover a larger number of isolated low mass galaxies and increase the power of our constraints. With our sample, we also find that the Tully-Fisher relation for dwarf galaxies is a strong function of environment, and that the baryonic fraction is only a weak function of mass. These results suggest that for dwarf galaxies, gas loss is dominated by external, not internal, processes. [abridged]
