GLIMPSE-D: Metallicity Decline in Faint Galaxies: Implications for [O III]+Hb Luminosity Function and Reionisation Budget

TL;DR

This study leverages JWST/NIRSpec observations of strongly lensed galaxies in Abell S1063 to measure the emission-line ratio $R_3=[O\,III]\lambda5008/H\beta$ for 54 very faint galaxies at $z\sim6-9$, enabling metallicity inferences down to roughly 2% solar and extending the $R_3$–$M_{\mathrm{UV}}$ relation to $M_{\mathrm{UV}}\lesssim -14$. By calibrating $R_3$ on the low-metallicity branch, the authors derive metallicities and identify a turnover in $R_3$ at $M_{\mathrm{UV}}\sim-18$, with metallicities decreasing toward fainter UV magnitudes. They construct four prescription models for $R_3(M_{\mathrm{UV}})$ and propagate them through the GLIMPSE [O III]+H$\beta$ LF to estimate the ionising photon production rate $\dot{N}_{\rm ion}$ under a fixed escape fraction $f_{\rm esc}=14\%$, comparing against reionisation constraints. The results show that intermediate models can reproduce the ionisation budget and support a galaxy-dominated reionisation, while extremely flat or sharply truncated scenarios are disfavoured; the findings emphasize that faint, metal-poor galaxies play a nuanced role and that metallicity evolution must be incorporated when interpreting emission-line LFs as proxies for the reionisation budget.

Abstract

We report the measurement of the R3=[O III]5008/Hb ratios for 54 galaxies in the GLIMPSE-D survey. Thanks to gravitational lensing, our sample includes galaxies with -20 < Muv < -14 at z=6-9. We derive oxygen abundances using calibrated relationships. We observe a significant decline in R3 values below Muv > -18, which we interpret as evidence of decreasing metallicities in fainter regimes. We explore four prescription models of the evolution of R3 with UV emission based on the new measurements and results from previous surveys. Applying these models to the GLIMPSE [O III]+Hb luminosity functions, we measure and extrapolate the ionising photon production rate $\dot{N}_{ion}$ of galaxies down to very faint limits SFR(Ha) > 5e-3 Msun/yr. Our results support the dominant contribution of star-forming galaxies to reionisation, and are consistent with the recent discovery of ultra-faint metal-poor galaxies. Our measurements of the relative contribution of each luminosity bin show that galaxies with L(Ha)~1e41 to 1e42 erg/s dominate at 8<z<9, but the relative contributions become more uniform at $7<z<8$. Extreme models either under- or over-estimate the ionising photon budget, while intermediate models align with recent observational constraints.

Figures (5)

• Figure 1: Evolution of the R3 ratio as function of _UV $M_\text{UV}$ for the GLIMPSE-D galaxies, as well as the public JADES DR4 scholtz_jades_2025 and UNCOVER DR4 price_uncover_2025 and chemerynska_extreme_2024. We further included the few very metal poor galaxies detected between redshift $6<z<9$ (V23,V24: vanzella_extremely_2023vanzella_extreme_2024; W25: willott_search_2025). We also show the evolution as assumed in korber_glimpse_2025. Left:$6<z<7$ redshift range that includes many new objects from GLIMPSE-D in the very faint end. Right:$7<z<9$ redshift range corresponding to the ranges with [O iii] [O iii]+Hβ H$\beta$ in the F444W NIRCam filter, as discussed in korber_glimpse_2025.
• Figure 2: Oxygen abundance of the selected GLIMPSE-D galaxies using the R3 calibration of nakajima_empress_2022. Measurements and uncertainties for individual sources from R3 and the calibration are shown in black, and the oxygen abundance of the R3 running median is shown in blue.
• Figure 3: Prescription models of R3(Muv) with the UNCOVER, JADES and GLIMPSE-D datapoints in the background.
• Figure 4: Integrated ionising photon production rate N_ ion$\dot{N}_{\rm ion}$ down to given $L_{\text{\alpha} H$α$}$, measured from the GLIMPSE [O iii] [O iii]+Hβ H$\beta$ LF with the four R3 models (top panels), and the relative contribution of each luminosity bin to the ionising photon production rate (bottom panels). We use the average escape fraction measured for GLIMPSE $f\textrm{esc}$f_esc$=14\%$ (Jecmen et al. Submitted). The left and right panels differentiate the LF from the two redshift bins from korber_glimpse_2025. More details in App. \ref{['app:nion']}.
• Figure 5: Four galaxies with their observed spectrum and their associated best fit model. The galaxies are ordered by their R3 (from highest to lowest).