Interpreting nebular emission lines in the high-redshift Universe

Abstract: One of the most remarkable outcomes from JWST has been the exquisite UV-optical spectroscopic data for galaxies in the high-redshift Universe ($z \geq 5$), enabling the use of various nebular emission lines to infer conditions of the interstellar medium. In this work, we evaluate the reliability of these diagnostics, specifically those used to recover the star formation rate (SFR), the ionising photon production efficiency, and the gas-phase oxygen abundance. We use forward-modelled galaxy spectra from idealised toy models and the FLARES cosmological hydrodynamical simulations to assess how these diagnostic lines respond to variations in the underlying stellar populations and star-dust geometry. We find that due to the clumpy nature of FLARES galaxies, they exhibit strong internal variation in age, metallicity and dust attenuation, biasing the interpretation using these diagnostics. In FLARES, the SFRD at the bright-end of the SFR function can be underestimated by as much as $30 \%$ compared to the true values. Similarly, the true ionising photon production efficiency in FLARES remains almost constant with stellar mass, while estimates using H$\alpha$ or H$\beta$ can underestimate it by more than 0.5 dex at the massive end ($>10^{9.5}$ M$_{\odot}$), introducing an artificial declining trend. The FLARES mass-metallicity relation derived from line ratios (dust corrected) exhibits a flat trend, whereas the intrinsic mass-weighted relation shows a strong positive correlation. These findings highlight the critical importance of understanding the biases introduced due to the coupling of star-dust geometry and heterogeneous stellar population when interpreting nebular emission lines at high redshift. Hence the accurate interpretation of high-redshift galaxy spectra requires careful forward modelling that accounts for realistic star-dust geometry and population variations.