An Improved UV-Optical Diagnostic for Rejuvenating Galaxies in the Local Universe and Implications for Galaxy Evolution

TL;DR

The paper addresses the challenge of identifying rejuvenating galaxies by decoupling short- and intermediate-timescale star formation signals in the local universe. It improves the CM21 UV–optical diagnostic by applying consistent dust corrections and by subtracting the O-star–driven NUV contribution from H$\alpha$–traced ionizing populations using a FSPS-based calibration, yielding a reliable $10$–$100\,\mathrm{Myr}$ signal. Applying this to a large SDSS sample, the authors identify roughly $10^4$ rejuvenating galaxies (~$4.5\%$) with intermediate stellar masses, finding that rejuvenation is more common in group outskirts and correlates with lower gas-phase metallicities by about $0.15$–$0.25$ dex, consistent with accretion of metal-poor gas. The method is scalable and minimizes reliance on full SED fitting, with future prospects to extend to IFU surveys and higher redshifts to map rejuvenation in the context of gas supply and quenching cycles.

Abstract

Rejuvenating galaxies are important probes of galaxy evolution, yet identifying them observationally is challenging as constraining recent star formation histories requires both photometric and spectroscopic data. We present a method for identifying rejuvenating galaxies in the local Universe using ultraviolet (UV) imaging and optical spectroscopy, building on a recent selection that identifies a system as rejuvenating if it is quenched in the near-UV (NUV; tracing $\sim\!100\,\mathrm{Myr}$ timescales) but star-forming in H$α$ (tracing $\sim\!10\,\mathrm{Myr}$ timescales). Shortly after a star formation episode, however, the NUV is dominated by the same massive stars that power H$α$, so these indicators do not always trace distinct timescales. To address this, we derive a relation that predicts the NUV emission associated with the ionizing O-star population traced by H$α$, enabling us to isolate the NUV contribution from longer-lived stars (primarily B/A stars with $M\lesssim\!20\,M_\odot$). Subtracting the predicted O-star NUV from the dust-corrected NUV yields a more reliable rejuvenation diagnostic. Using this method, we identify $\sim\!10^{4}$ rejuvenating galaxies in a sample of Sloan Digital Sky Survey (SDSS) galaxies ($\sim\!4.5\%$). These galaxies have intermediate stellar masses and are found primarily in lower-density environments, becoming increasingly rare toward the centers of groups and clusters. Rejuvenating galaxies also exhibit systematically lower gas-phase metallicities, consistent with fueling by the accretion of metal-poor gas.

Figures (8)

• Figure 1: Distribution of the GSWLC parent sample in $\mathrm{EW}_{H\alpha}$ vs. $\mathrm{sSFR}_{\mathrm{NUV}}$ space, illustrating four subpopulations---star-forming (blue), rapidly quenching (yellow), quenched (red), and rejuvenating (green)---as defined by CM21. The boundaries separating active star formation from quiescence are set at $\log_{10}\!(\mathrm{EW}_{H\alpha}/\text{\AA})=0.5$ and $\log_{10}\!(\mathrm{sSFR}_{\mathrm{NUV}}/\mathrm{yr}^{-1})=-11.1$, corresponding to minima in the respective relations with stellar mass.
• Figure 2: Inclination $i$ from axis ratio following Blanton2011 vs. $V$-band optical depth $\tau_V$ from MPA--JHU for the CM21 rejuvenating sample (red) and our dust-corrected rejuvenating sample (green). The faint blue points show the distribution of star-forming GSWLC--A galaxies. Kernel density contours are shown for both rejuvenating samples. Marginal kernel density estimates (KDEs) along the axes summarize $i$ and $\tau_V$. Dashed lines mark medians.
• Figure 3: Kernel density distributions comparing H$\alpha$ and NUV luminosities before (thin black) and after (thick blue) dust corrections for the MPA--JHU star-forming sample Brinchmann2004. Best-fit linear trends are overplotted and illustrate improved correlation after correction. Marginal KDEs display one-dimensional distributions. Dashed lines indicate medians.
• Figure 4: FSPSConroy2009Conroy2010 tracks of $\log_{10}(L_{\mathrm{NUV,ion}}/L_{\mathrm{H}\alpha})$ using the O-star NUV component and nebular H$\alpha$. The black curve shows a constant-SFR model; the colored dotted curves show exponentially declining SFHs with varying $e$-folding times. All curves converge to the equilibrium plateau of $\simeq-13.8$ after $\sim\!3$--$6\,\mathrm{Myr}$. Shading marks the EPR (pre-equilibrium) regime.
• Figure 5: Distribution of galaxy subpopulations in $\mathrm{EW}_{H\alpha}$ vs. stellar mass. Kernel-density contours for rejuvenating (green) and quenched (orange) systems are shown over the full parent sample (black). Green points mark RGs classified as starbursts in MPA--JHU. Marginal KDEs along the axes display the one-dimensional distributions (mass on the $x$-axis; $\mathrm{EW}_{H\alpha}$ on the $y$-axis). Points and shaded bars denote medians and 16th--84th percentiles.