BlastBerries: How Supernovae Affect Lyman Continuum Escape Fractions and Ionizing Photon Production in Local Analogs of High-Redshift Galaxies

TL;DR

This study uses low-redshift Blueberry galaxies as local analogs of high-redshift star-forming systems to test whether SN feedback facilitates LyC photon escape. By cross-matching 1242 BB galaxies with transient surveys, the authors identify 11 SN hosts (primarily core-collapse) and compare their properties to the broader BB and field populations using SED fitting, UV slopes, and emission-line diagnostics. They find SN hosts possess higher SFR and stellar mass with burstier SFHs, and they exhibit bluer UV continua suggestive of increased LyC escape, but they simultaneously show significantly lower ionizing photon production efficiency ξion, implying a potential decline in photon production over ~50 Myr due to SN feedback. The results hint at episodic SN-driven channels that could enhance escape in comparatively massive compact galaxies, offering important constraints for reionization models and highlighting the need for direct LyC/Lyα measurements and larger samples in the future.

Abstract

While compact, star-forming galaxies are believed to play a key role in cosmic reionization, the physical mechanisms enabling the escape of ionizing photons through the galactic interstellar medium remain unclear. Supernova (SN) feedback is one possible mechanism for clearing neutral gas channels to allow the escape of Lyman continuum photons. Here, we use SN discoveries in low-redshift analogs of high-redshift star-forming galaxies -- Green Pea galaxies and their even lower-redshift counterparts, Blueberry (BB) galaxies -- to understand how SNe shape the properties of their host galaxies at high redshifts. We cross-match 1242 BB galaxies with transient discovery reports and identify 11 SNe, ten of which are likely core-collapse SNe, and compare their hosts to the larger BB population. We find that SN-hosting BBs exhibit elevated star formation rates, burstier star formation histories within the last $\sim$50 Myr, and higher stellar masses. We estimate the occurrence rates of SNe in BB galaxies, finding that the SN rate may be slightly suppressed in BBs compared to field galaxies of similar mass, but we are unable to fully control for observational uncertainties. Finally, SN hosts show bluer UV slopes than non-host BB galaxies at 2.1$σ$ significance and lower ionizing photon production efficiency at 7.9$σ$ significance; the former result offers modest support for the hypothesis that SN-driven feedback plays a role in facilitating the escape of ionizing photons, while the latter may imply that SN-driven quenching decreases the rate of ionizing photon production in compact star-forming galaxies during the epoch of reionization.

Figures (9)

• Figure 1: Three-color stamp images of the BB SN hosts in our sample, using grz filters from the DESI Legacy Imaging Survey dey_overview_2019levi_dark_2019. The images are centered on the host galaxies, while the red crosses mark the coordinates of each SN. Redshifts are derived from SDSS catalogs, which include spectroscopically derived parameters for each host galaxy york_sloan_2000. The scale bar is given in angular size in the top left panel, where 3" corresponds to $\sim3$ kpc at $z=0.05$ and $\sim5.5$ kpc at $z=0.1$. Classification information is from the Transient Name Server (TNS) when available; some SNe in our sample were never spectroscopically classified.
• Figure 2: ZTF, ATLAS, and Pan-STARRS light curves for all SNe in our matched sample, with SN 2010ay from sanders_sn_2012 including synthetic photometry from Gemini observations. SN candidates 2024ruy, 2020adtl, and 2023oou are excluded from our analysis. We show the best-fit STARDUST2 model for each bona fide SN rodney_type_2014.
• Figure 3: SN event frequency ($N_{\rm SN}/N_{\rm total}$) as a function of mass for the SDSS field galaxy sample (green), the color-filtered SFG sample (blue), and the BB sample (pink). Error bars are derived from Poisson statistics for galaxies in the given stellar mass bin. The BB sample shows a lower frequency of SN events.
• Figure 4: SN frequency for the SDSS field galaxy sample (green) and the BB sample (pink). Each sample is binned by FUV-calculated SFR, and frequency is calculated by $N_{{\rm SN}}/N_{{\rm total}}$ in each bin. See the Figure \ref{['fig:rates']} caption for additional details.
• Figure 5: Corner plot showing galaxy properties of BB galaxies and BB SN hosts, measured from the SDSS spectroscopic pipeline. Properties include the H$\alpha$-inferred SFR, electron density, metallicity, and stellar mass. The contours displayed correspond to the $0.5, 1, 1.5, 2 \sigma$ levels. The unclassified SNe are marked with brown "x" markers, while the classified Type Ia and Type II SNe are marked with orange and pink stars, respectively. Dotted lines represent the population means. Top labels give the medians of each property for the BB galaxies, with error bars corresponding to the 16th and 84th percentiles.