Photometric identification and MMT spectroscopy of new extremely metal-poor galaxies: towards a better understanding of young stellar populations at low metallicity

TL;DR

This study introduces a new photometric method to identify extremely metal-poor galaxies with young stellar populations, confirmed through spectroscopy, to better understand low-metallicity star formation akin to early universe conditions.

Contribution

The paper presents a novel photometric selection technique for discovering nearby XMPs with young stellar populations, validated by spectroscopic follow-up, expanding the sample of known extremely metal-poor galaxies.

Findings

32 of 53 targets confirmed as extremely metal-poor with $12+\log\mathrm{O/H}<7.7$

Detection of two galaxies with the lowest known metallicities, $Z/Z_\odot<0.05$

He II emission does not scale with H$\beta$, indicating complex ionization sources

Abstract

Extremely metal-poor star-forming galaxies (XMPs) represent one of our only laboratories for study of the low-metallicity stars we expect to encounter at early epochs. But as our understanding of the $z>6$ universe has improved, it has become clear that the majority of known XMPs within 100 Mpc host significantly less prominent massive star populations than their reionization-era counterparts, severely limiting their utility as testbeds for interpreting spectral features found at the highest redshifts. Here we present a new photometric selection technique designed to identify nearby XMPs dominated by young stellar populations comparable to those expected in the reionization era. We apply our technique to uncover candidate XMPs in SDSS imaging at magnitudes $16<i'<23$, extending significantly below the completeness limits of the SDSS spectroscopic survey. Spectroscopic observations with the MMT confirm that 32 of the 53 uniformly metal-poor and high specific star formation rate targets we observed have gas-phase oxygen abundances $12+\log\mathrm{O/H}<7.7$ ($Z/Z_\odot<0.1$), including two in the range of the lowest-metallicity galaxies known, $Z/Z_\odot<0.05$. Our observations shed new light onto the longstanding mystery of He II emission in star-forming galaxies: we find that the equivalent width of the He II $\lambda 4686$ high-ionization emission line does not scale with that of H$\beta$ in our sample, suggesting that binary evolution or other processes on $>10$ Myr timescales contribute substantially to the $\mathrm{He^+}$-ionizing photon budget in this metallicity regime. Applying such selection techniques coupled with deep spectroscopy to next-generation photometric surveys like LSST may eventually provide a basis for an empirical understanding of metal-poor massive stars.