We present uniformly measured star formation histories (SFHs) of 40 Local Group (LG) dwarf galaxies based on color-magnitude diagram (CMD) analysis from archival Hubble Space Telescope imaging. We demonstrate that accurate SFHs can be recovered from CMDs that do not reach the oldest main sequence turn-off (MSTO), but emphasize that the oldest MSTO is critical for precisely constraining the earliest epochs of star formation. We find that: (1) the average lifetime SFHs of dwarf spheroidals (dSphs) can be approximated by an exponentially declining SFH with τ ∼ 5 Gyr; (2) lower luminosity dSphs are less likely to have extended SFHs than more luminous dSphs; (3) the average SFHs of dwarf irregulars (dIrrs), transition dwarfs, and dwarf ellipticals can be approximated by the combination of an exponentially declining SFH (τ ∼ 3-4 Gyr) for lookback ages >10-12 Gyr ago and a constant SFH thereafter; (4) the observed fraction of stellar mass formed prior to z = 2 ranges considerably (80% for galaxies with M < 10 5 M to 30% for galaxies with M > 10 7 M ) and is largely explained by environment; (5) the distinction between "ultra-faint" and "classical" dSphs is arbitrary; (6) LG dIrrs formed a significantly higher fraction of stellar mass prior to z = 2 than the Sloan Digital Sky Survey galaxies from Leitner and the SFHs from the abundance matching models of Behroozi et al. This may indicate higher than expected star formation efficiencies at early times in low mass galaxies. Finally, we provide all the SFHs in tabulated electronic format for use by the community.

We search for signatures of reionization in the star formation histories (SFHs) of 38 Local Group dwarf galaxies (10 4 < M < 10 9 M ). The SFHs are derived from color-magnitude diagrams using archival Hubble Space Telescope/Wide Field Planetary Camera 2 imaging. Only five quenched galaxies (And V, And VI, And XIII, Leo IV, Hercules) are consistent with forming the bulk of their stars before reionization, when full uncertainties are considered. Observations of 13 of the predicted 'true fossils' identified by Bovill & Ricotti show that only two (Hercules and Leo IV) indicate star formation quenched by reionization. However, both are within the virial radius of the Milky Way and evidence of tidal disturbance complicates this interpretation. We argue that the late-time gas capture scenario posited by Ricotti for the low mass, gas-rich, and star-forming fossil candidate Leo T is observationally indistinguishable from simple gas retention. Given the ambiguity between environmental effects and reionization, the best reionization fossil candidates are quenched low mass field galaxies (e.g., KKR 25).

We use archival Hubble Space Telescope observations of resolved stellar populations to derive the star formation histories (SFHs) of eighteen nearby starburst dwarf galaxies. In this first paper we present the observations, color-magnitude diagrams, and the SFHs of the eighteen starburst galaxies, based on a homogeneous approach to the data reduction, differential extinction, and treatment of photometric completeness. We adopt a star formation rate (SFR) threshold normalized to the average SFR of the individual system as a metric for classifying starbursts in SFHs derived from resolved stellar populations. This choice facilitates finding not only currently bursting galaxies but also "fossil" bursts increasing the sample size of starburst galaxies in the nearby (D< 8 Mpc) universe. Thirteen of the eighteen galaxies are experiencing ongoing bursts and five galaxies show fossil bursts. From our reconstructed SFHs, it is evident that the elevated SFRs of a burst are sustained for hundreds of Myr with variations on small timescales. A long > 100 Myr temporal baseline is thus fundamental to any starburst definition or identification method. The longer lived bursts rule out rapid "self-quenching" of starbursts on global scales. The bursting galaxies' gas consumption timescales are shorter than the Hubble time for all but one galaxy confirming the shortlived nature of starbursts based on fuel limitations. Additionally, we find the strength of the Hα emission usually correlates with the CMD based SFR during the last 4−10 Myr. However, in four cases, the Hα emission is significantly less than what is expected for models of starbursts; the discrepancy is due to the SFR changing on timescales of a few Myr. The inherently short timescale of the Hα emission limits identifying galaxies as

We measure the mass function for a sample of 840 young star clusters with ages between 10-300 Myr observed by the Panchromatic Hubble Andromeda Treasury (PHAT) survey in M31. The data show clear evidence of a high-mass truncation: only 15 clusters more massive than > 10 4 M are observed, compared to ∼100 expected for a canonical M −2 pure power-law mass function with the same total number of clusters above the catalog completeness limit. Adopting a Schechter function parameterization, we fit a characteristic truncation mass of M c = 8.53 M . While previous studies have measured cluster mass function truncations, the characteristic truncation mass we measure is the lowest ever reported. Combining this M31 measurement with previous results, we find that the cluster mass function truncation correlates strongly with the characteristic star formation rate surface density of the host galaxy, where M c ∝ Σ SFR ∼1.1 . We also find evidence that suggests the observed M c -Σ SFR relation also applies to globular clusters, linking the two populations via a common formation pathway. If so, globular cluster mass functions could be useful tools for constraining the star formation properties of their progenitor host galaxies in the early Universe. Subject headings: galaxies: star clusters: general -galaxies: star formation -galaxies: individual (M31) -globular clusters: general 1. INTRODUCTION Star cluster populations are observational tracers of star formation activity in galaxies out to ∼100 Mpc distances. By comparing the properties of star cluster populations to the properties of overall star formation activity, studies of nearby galaxies have established that there is a correlation between the star formation rate (SFR) surface density, Σ SFR , and the fraction of stars that form in long-lived star clusters (e.g., Adamo et al. 2015, Johnson et al. 2016. This correlation demonstrates a close connection between star clusters and their formation environment, where the rate of cluster formation is linked to the total SFR, but also to local galactic properties such as gas surface density and interstellar pressure (Kruijssen 2012). One implication of this result is that star clusters can reveal the characteristics of past star formation episodes long after they have ended. While cluster destruction through evaporation due to two-body relaxation, tidal shocks, and other processes will erode lowmass star cluster populations over time, globular clusters and other massive clusters provide long-lived records of star formation activity.The mass function of star clusters is another observable property that we can exploit to study episodes of past star formation. Numerous studies have characterized the mass function of young star clusters using a power-law distribution (dN/dM ∝ M α ) with an index of α=−2.0 ± 0.3 that holds over a wide range of cluster

We map the distribution of dust in M31 at 25 pc resolution, using stellar photometry from the Panchromatic Hubble Andromeda Treasury survey. The map is derived with a new technique that models the near-infrared color-magnitude diagram (CMD) of red giant branch (RGB) stars. The model CMDs combine an unreddened foreground of RGB stars with a reddened background population viewed through a log-normal column density distribution of dust. Fits to the model constrain the median extinction, the width of the extinction distribution, and the fraction of reddened stars in each 25 pc cell. The resulting extinction map has a factor of 4 times better resolution than maps of dust emission, while providing a more direct measurement of the dust column. There is superb morphological agreement between the new map and maps of the extinction inferred from dust emission by Draine et al. (2014). However, the widely-used Draine & Li (2007) dust models overpredict the observed extinction by a factor of ∼ 2.5, suggesting that M31's true dust mass is lower and that dust grains are significantly more emissive than assumed in Draine et al. (2014). The observed factor of ∼ 2.5 discrepancy is consistent with similar findings in the Milky Way by Plank Collaboration et al. (2014), but we find a more complex dependence on parameters from the Draine & Li (2007) dust models. We also show that the discrepancy with the Draine et al. (2014) map is lowest where the current interstellar radiation field has a harder spectrum than average. We discuss possible improvements to the CMD dust mapping technique, and explore further applications in both M31 and other galaxies.