The [O iii]+H β equivalent width distribution at z ≃ 7: implications for the contribution of galaxies to reionization

Abstract

We quantify the distribution of [OIII]+Hβ line strengths at z≃7 using a sample of 20 bright ($\mathrm{M}_{\mathrm{UV}}^{ }$ ≲ -21) galaxies. We select these systems over wide-area fields (2.3 deg2 total) using a new colour-selection which precisely selects galaxies at z≃6.63–6.83, a redshift range where blue Spitzer/IRAC [3.6]−[4.5] colours unambiguously indicate strong [OIII]+Hβ emission. These 20 galaxies suggest a log-normal [OIII]+Hβ EW distribution with median EW = 759$^{{+112}{4.5pt}}_{{-113}{4.5pt}}$ Å and standard deviation = 0.26$^{{+0.06}{4.5pt}}_{{-0.05}{4.5pt}}$ dex. We find no evidence for strong variation in this EW distribution with UV luminosity. The typical [OIII]+Hβ EW at z≃7 implied by our sample is considerably larger than that in massive star forming galaxies at z≃2, consistent with a shift toward larger average sSFR (4.4 Gyr−1) and lower metallicities (0.16 Z⊙). We also find evidence for the emergence of a population with yet more extreme nebular emission ([OIII]+Hβ EW>1200 Å) that is rarely seen at lower redshifts. These objects have extremely large sSFR (>30 Gyr−1), as would be expected for systems undergoing a burst or upturn in star formation. While this may be a short-lived phase, our results suggest that 20% of the z≃7 population has such extreme nebular emission, implying that galaxies likely undergo intense star formation episodes regularly at z>6. We argue that this population may be among the most effective ionizing agents in the reionization era, both in terms of photon production efficiency and escape fraction. We furthermore suggest that galaxies passing through this large sSFR phase are likely to be very efficient in forming bound star clusters.

Ultraviolet spectra of extreme nearby star-forming regions: Evidence for an overabundance of very massive stars

“…The very high equivalent width of the optical lines these systems power implies a dominant young 10 Myr stellar population, comparable to the extremely high sSFR systems glimpsed in the reionization era (e.g. Roberts-Borsani et al 2016;Endsley et al 2020) and probing conditions where the uncertain treatment of very massive stars in population synthesis will have an outsized impact on our ability to correctly interpret and model observations. As discussed in Section 4, in this paper we will focus on comparisons to the latest predictions of the C&B stellar population synthesis framework.…”

“…Evidence from both photometric bands contaminated by rest-optical line emission and the first rest-UV spectra in the reionization era suggests that a significant fraction of z > 6 galaxies are dominated by populations of massive stars formed in bursts initiated within < 20 Myr of observation (e.g. Stark et al 2015b;Endsley et al 2020). Thus, our understanding of the most distant galaxies JWST will observe will depend sensitively on models for the evolution of populations of massive stars on much shorter timescales than typical star-forming galaxies even at z ∼ 2 (e.g.…”

“…Leveraging the fact that the strong rest-optical lines contaminate the Spitzer /IRAC [3.6] and [4.5] micron bands at z ∼ 6-9, several authors have robustly demonstrated that the specific star formation rates (sSFRs) of typical systems at these redshifts are well inexcess of galaxies at lower redshift (Labbé et al 2013;Smit et al 2014). Indeed, Endsley et al (2020) have recently shown that 20% of the most luminous systems at z ∼ 7 reveal evidence for extremely high equivalent width optical nebular emission corresponding to sSFRs 50 Gyr −1 , indicative of a substantial population of galaxies dominated entirely by massive stars formed within the last <20 Myr. Accurately interpreting the nebular line emission detected in such extreme systems will rely sensitively on the accuracy of models for the ionizing radiation fields powered by very young populations of massive stars.…”

How robustly can we constrain the low-mass end of the z ∼ 6−7 stellar mass function? The limits of lensing models and stellar population assumptions in the Hubble Frontier Fields

“…We use 𝑟 = 0.5 kpc, a typical half-light radius of a galaxy at 𝑧 ∼ 6 − 7 (Kawamata et al 2015), and 𝑀 ∼ 10 8 M as a typical galaxy mass, resulting in a dynamical time 𝑡 dyn ∼ 20 Myr. We set the lower boundary of 𝑡 age to 20 Myr in order to avoid solutions around 𝑡 age ∼ 10 Myr with extremely high nebular equivalent widths (EWs) ∼ 7000 Å for [O ] 5007 Å, which remain unlikely given evidence from observations of high-redshift galaxies (Atek et al 2011(Atek et al , 2014Smit et al 2014;Reddy et al 2018b) and stellar population models (but see Endsley et al 2021). We refer the reader to appendix A for further discussion of this issue.…”

“…While constant or exponentially decreasing SFHs were previously assumed (e.g., Eyles et al 2007;Stark et al 2009;González et al 2011;Grazian et al 2015;Kikuchihara et al 2020), results from hydrodynamical simulations predict exponentially increasing SFRs for high-redshift galaxies (e.g., Finlator et al 2011). Observations of very high EW optical emission lines of [O ] and H𝛽, up to rest-frame EW ∼ 2000 Å, (e.g., Atek et al 2011Atek et al , 2014Reddy et al 2018b) and of IRAC excesses in high-redshift galaxies (which have been attributed to strong contributions from rest-frame optical emission lines; Smit et al 2014Smit et al , 2015De Barros et al 2019;Endsley et al 2021) further support young galaxies with relatively high rates of recent star formation, which could be due to a SFH with an increasing SFR as a function of time. Alternatively, given their compact sizes, the star formation in young high-redshift galaxies might also be episodic with several relatively strong bursts occurring over relatively short duty cycles (e.g., Stark et al 2009).…”