Kristen B. Wingfield McQuinn

Oxygen is forged by stars and redistributed through galaxies by feedback-driven outflows, leaving a record of star formation and the baryon cycle imprinted on its present-day abundance and distribution. The Galaxies Losing Oxygen via Winds project quantifies the production, distribution, and retention of oxygen in 37 low-mass, low–metallicity, gas-rich galaxies in the nearby Universe ( D < 6 Mpc) spanning a critical stellar mass range (10 ^6.5 < M _* < 10 ^9.5 ) where feedback is expected to strongly shape galaxy evolution. We incorporate literature results to extend the analysis over 10 ⁵ < M _* < 10 ^11.5 . We couple star formation histories derived from resolved stars with measurements of interstellar medium metallicity and gas mass, and compare the oxygen budgets with empirical constraints in the circumgalactic medium (CGM). We find low-mass galaxies retain between 4% and 25% (mean 11%) of their oxygen produced by stellar nucleosynthesis, and nearly all retained oxygen resides in the gas. Galaxies exhibit increased scatter to higher oxygen retention fractions when residing in denser environments (Θ ₅ > 0.5) and within 1 Mpc of a massive galaxy ( M _* > 10 ⁹ M _⊙ ), indicating the environment likely affects the amount of oxygen retained, recycled, or accreted to galaxies. Contrary to expectations, oxygen retention does not correlate with position on the mass–metallicity relation. Although ionized oxygen is detected in the CGM, it remains unclear whether most missing oxygen resides there or has been expelled entirely. Hydrodynamical simulations, despite successfully reproducing the mass–metallicity relation, predict much higher retention fractions than measured in low-mass galaxies. Simple modeling indicates that wind mass-loading and outflow metallicity govern oxygen retention, and that low-mass galaxies accrete less baryonic material relative to the cosmic baryon fraction.

We present radial stellar age gradients measured from star formation histories (SFHs) fit to resolved color–magnitude diagrams (CMDs) of 42 Local Volume dwarfs (6 ≲ Log M _⋆ / M _⊙ ≲ 9), spatially divided into elliptical annuli. Ages in each annulus are quantified using τ ₉₀ and τ ₅₀ , the lookback times to form 90% and 50% of the cumulative stellar mass. We find that radial age gradients are uncorrelated with environment, but gradients in τ ₉₀ are significantly correlated ( p -values ≲ 0.001) with lifetime galaxy-wide (“global”) SFHs, in agreement with two independent cosmological zoom-in simulations. For radial gradients of τ ₅₀ , simulation predictions differ. We demonstrate that given our large ( N = 42) and diverse observational sample, the strength of an observed correlation with global SFH is an actionable parameter to discriminate between different simulations with differing stellar feedback prescriptions. Overall, our results support predictions that dwarfs form inside-out like their more massive counterparts, with a combination of feedback-driven outward radial stellar migration and increasing birth radii for young stellar populations yielding present-day stellar age gradients ranging from outside-in to flat. In addition, the lack of a correlation between τ ₅₀ gradients and global SFH argues against recent star formation in radially outflowing gas. We also discuss the impact of differences between our observational sample and samples available in the latest simulations, highlighting areas for future investigation.

Gas-rich, star-forming, low-mass galaxies in the nearby Universe are powerful laboratories for studying baryonic physics in detail, including: stellar mass assembly, stellar feedback, chemical enrichment, and the interplay of the interstellar medium with star formation. Investigating these disparate yet interconnected processes requires data obtained by myriad observatories. Here, we present a comprehensive atlas of uniformly processed data on 37 low-mass galaxies within 6 Mpc. The atlas includes archival data on (i) the H i from the Very Large Array observatory; (ii) resolved stars from Hubble Space Telescope optical imaging; (iii) Spitzer Space Telescope 3.6 μ m imaging; and (iv) optical imaging from ground-based telescopes. We also compile measurements of (i) tip-of-the-red-giant-branch (TRGB) distances to the galaxies; (ii) direct method gas-phase oxygen abundances and nitrogen-to-oxygen abundance ratios; (iii) constraints on the local environment around each galaxy; and (iv) other measurements from the literature. We supplement the data with new observations where needed to complete the measurements for all galaxies in the sample. From these data, we find good agreement between stellar masses measured from color–magnitude diagrams and those estimated from 3.6 μ m imaging by assuming a mass-to-light ratio. We also provide the first mapping of the H i profiles as a function of structural parameters. These data sets and measurements are the foundation for the Galaxies Losing Oxygen via Winds (GLOW) project, whose main aim is to characterize the star formation−chemical enrichment cycle of low-mass galaxies by measuring the production, distribution, and retention of oxygen on a galaxy-by-galaxy basis.

The dependence of the stellar initial mass function (IMF) on star formation environment, particularly at low metallicities and high redshifts, remains poorly constrained. Ultrafaint dwarf galaxies (UFDs) are local fossils of high-redshift galaxies hosting old, metal-poor populations, and their resolved stellar populations provide unique pathways to constrain the subsolar IMF. We investigate the low-mass IMF in the Boötes I (Boo I) UFD with JWST/NIRCam, leveraging its capability to resolve over 10,000 stars reaching ≲0.15 M _⊙ , obtaining one of the largest, deepest resolved stellar samples for UFDs. We explore three different functional forms of the IMF with machine learning and statistical techniques, combining forward modeling of synthetic color–magnitude diagrams with simulation-based inference. We find that a single power-law IMF provides a poorer description of the observed luminosity function and yields a slope inconsistent with the canonical Salpeter IMF. Our best-fit broken power-law and log-normal IMF parameters are consistent with the Milky Way within 68% confidence level, providing evidence that star formation at metallicities as low as [Fe/H] ≈ −2.4 follows a similar IMF as in the Milky Way. By treating Boo I as a local relic analogous to a high-redshift galaxy with a stellar mass of ≲10 ⁵ M _⊙ at z ≳ 6, our results provide evidence for the universality of the IMF across both local and high-redshift environments.

The Treasury of Extremely Metal-Poor O Stars (TEMPOS) is a Hubble Space Telescope survey of hot and massive O-type stars in nearby, low-metallicity galaxies (≲20% of the solar metallicity, Z _⊙ ). Understanding massive-star physics in this regime is essential to interpret observations of metal-poor galaxies, including both low-mass dwarf galaxies and chemically unevolved galaxies in the early Universe. Yet, few far-ultraviolet (FUV) spectra of O stars of sufficient quality to characterize their fundamental properties and stellar winds exist below 20% Z _⊙ , and the heterogeneous observation design and incomplete coverage of parameter space pose significant barriers to progress. To remedy this, TEMPOS obtained new Cosmic Origins Spectrograph (COS) FUV spectra of 12 very metal-poor O stars, building upon archival data to assemble a spectroscopic atlas of 29 homogeneously observed stars that efficiently samples a wide range of spectral types and luminosity classes. Here, we describe the motivation, sample selection, and observation design for TEMPOS and present the first data release of reduced and coadded COS spectra. We then present initial results on the empirical properties of FUV O-star spectra below 20% Z _⊙ , including radial velocities, equivalent widths of photospheric lines, and terminal wind velocities ( v _∞ ). We show that v _∞ correlates with host galaxy metallicity across ∼5%–50% Z _⊙ and find tentative evidence of a steeper decline in wind strength below ∼10% Z _⊙ . The combined dataset of FUV spectra and planned releases of photometry and optical spectra from the TEMPOS Treasury program will advance our understanding of both stellar astrophysics and the interstellar medium in the extremely metal-poor regime.

We present deep Hubble Space Telescope (HST) imaging of one of the nearest ultra-diffuse galaxies (UDGs) outside of the Local Group: F8D1, a satellite of M81 known to be tidally disrupting. UDGs are an enigmatic and diverse population, with evolutionary pathways ranging from tidal processing to bursty feedback and high initial angular momentum. To determine F8D1’s evolutionary drivers, we resolve stars in F8D1’s central ∼1 kpc and in a parallel field ∼6 kpc along its major axis to deep photometric limits, reaching below the red clump. We also image eight shallower fields along F8D1’s major and minor axes. We calculate the star formation history (SFH) in the two deep fields, finding that while currently quiescent, both regions experienced a substantial burst ∼2 Gyr ago and a smaller burst ∼500 Myr ago, which likely formed F8D1’s nuclear star cluster. In the shallow fields, using the ratio of evolved asymptotic giant branch and red giant branch stars out to ∼13 kpc along F8D1’s known stellar stream, we confirm that F8D1 was globally star-forming until at least ∼2 Gyr ago. We estimate a total progenitor stellar mass, including the stream, of ∼1.3 × 10 ⁸ M _⊙ , with an average [M/H] ∼ −0.8. We compare F8D1’s properties to those of Local Group galaxies with similar initial stellar mass. We find that F8D1 is consistent with a progenitor star-forming galaxy similar to NGC 6822 that is in the midst of a transition to a Sagittarius-like system. Notably, this evolutionary sequence can be accomplished through tidal processing alone in galaxies that have experienced sufficiently bursty feedback to have created cored profiles.

The emissivity of dust is known to vary greatly with radiative environment, density, grain chemistry, and geometry. Discrepancies between dust-mass surface densities derived from far-infrared (FIR) emission and visible extinction persist across and within galaxies in the local Universe. Here, we use new extinction and emission measurements toward the LMC to show that this discrepancy is driven by the dust-mass opacity evolving with the intrinsic density of the interstellar medium, and that the ratio between FIR and optical dust-mass opacity varies with gas surface density. These new findings imply that the dust-mass opacity in the FIR could increase by nearly an order of magnitude (e.g., κ ₁₆₀ = 0.3–6 m ² kg ⁻² ) across over an order of magnitude of total hydrogen surface density (Σ _H = 4–100 M _⊙ pc ⁻² ), corroborating previous theoretical models for dust-mass opacity evolution in the FIR and providing new implications for emission-based dust-mass estimates.

We present a new framework for modeling the chemical enrichment histories of galaxies by integrating chemical evolution with resolved star formation histories (SFHs) derived from color–magnitude diagrams. This novel approach links the time evolution of the metallicity of the star-forming interstellar medium to the cumulative stellar mass formed in the galaxy, enabling a self-consistent description of chemical evolution. We apply this methodology to four isolated, gas-rich Local Group dwarf galaxies—Wolf–Lundmark–Melotte (WLM), Aquarius, Leo A, and Leo P—using deep Hubble Space Telescope and JWST imaging. For WLM, Aquarius, and Leo A, we independently validate our metallicity evolution results against ages and metallicities of individual red giant stars with spectroscopic measurements. We quantify systematic uncertainties by repeating our analysis with multiple stellar evolution and bolometric correction libraries. We compare the observed chemical enrichment histories to predictions from the TNG50 and FIREbox cosmological hydrodynamic simulations and the Galacticus semianalytic model. Of our four galaxies, only WLM is sufficiently massive to be reliably represented in these simulations; the remaining three fall below current resolution limits. We find that the enrichment history of WLM is best reproduced by FIREbox, while TNG50 and Galacticus predict higher metallicities at early times, suggesting that differences in stellar feedback and metal recycling prescriptions drive significant variation in predicted enrichment histories. This work demonstrates the power of combining resolved SFHs with physically motivated chemical evolution models to constrain galaxy formation physics and highlights the need for further observational and theoretical studies of metal retention and recycling in low-mass dwarf galaxies.

RR Lyrae stars are a common, dependable Population II distance indicator and provide an independent tracer of early star formation. Here, we utilize archival Hubble Space Telescope (HST) Advanced Camera for Surveys and JWST/NIRCam observations of the nearby dwarf star-forming galaxy Wolf–Lundmark–Melotte (WLM) to study RR Lyrae stars in JWST filters. We independently identify RR Lyrae stars in HST and JWST imaging in order to evaluate JWST’s efficacy at characterizing RR Lyrae stars in the near-IR. We use a Markov Chain Monte Carlo (MCMC) template-fitting technique to obtain periods, amplitudes, and mean magnitudes from the RR Lyrae time-series data. The spatially overlapping HST and JWST observations allow us to directly compare the same sources observed with the instruments and to calibrate the NIRCam F090W and F150W RR Lyrae period–Wesenheit–metallicity (PWZ) relation to the Gaia-consistent HST PWZ. We additionally assess the epoch-to-epoch consistency of the NIRCam photometry and find evidence of burn-in. We conclude that the zero-point offset is negligible compared to the uncertainties from the template fitting. We conduct an MCMC fit of the PWZ with both HST and JWST data. Our results are threefold. First, we find that we can reliably identify RR Lyrae stars in NIRCam data, but light-curve template fitting proves difficult on short-baseline observations. Second, the HST PWZ fit yields a distance modulus to WLM of μ = 24.85 ± 0.05 (0.93 ± 0.02 Mpc). This is closer than previous measurements, primarily attributed to consistency with the Gaia scale. Last, although the JWST PWZ fit has large uncertainties and a poorly constrained slope, it represents a first-of-its-kind PWZ calibration in NIRCam filters.

We use Hubble Space Telescope optical imaging from the Panchromatic Hubble Andromeda Southern Treasury (PHAST) to measure the spatially resolved recent star formation history (SFH) across the southern disk of M31. We fit color-magnitude diagrams (CMDs) of over 6500 individual 0.01 kpc ² regions to measure SFHs over the last ∼500 Myr. The resulting maps show coherent structure that traces the ringed morphology of the disk. We find a clear global decline in the recent star formation rate (SFR), with a pronounced drop in the last ∼40 Myr that is most evident in the region closest to M32. Combining PHAST and PHAT measurements, we now cover two-thirds of M31’s star-forming disk with homogeneous SFHs derived with the same technique, yielding the highest resolution spatially resolved SFHs of M31 measured with resolved stars. Inside the joint footprint, we measure a mean SFR of 0.445 ± 0.006 M _⊙ yr ⁻¹ over the last 100 Myr, and 0.285 ± 0.014 M _⊙ yr ⁻¹ over the last 20 Myr, implying a total disk SFR of ∼0.67 M _⊙ yr ⁻¹ and ∼0.43 M _⊙ yr ⁻¹ respectively, when extrapolated to the full star-forming disk. The observed decline is interpreted to be the late stages of a multi-Gyr wind-down from a previously more active state. Because recent star formation in M31 is concentrated primarily in the rings, the observed global decline is driven mainly by decreasing activity within those features. We also compare the CMD-based SFR surface densities to those inferred from far-ultraviolet (FUV)+24 μ m prescriptions and find that the FUV-based calibration underestimates the CMD-based 100 Myr average by a factor of ∼2.1. However, the PHAST SFHs produce a synthetic Galaxy Evolution Explorer FUV image that agrees well with observations, indicating that the CMD-derived SFHs provide an accurate description of recent star formation. The mismatch with the FUV+24 μ m estimates underscores that tracers implicitly averaged over ∼100 Myr are not reliable when the recent SFR is evolving.