One cloud is not enough: Extreme conditions biasing chemical abundances in high-redshift galaxies
Abstract
Since its launch, JWST has enabled the characterisation of the ionised interstellar medium (ISM) of high-redshift galaxies using a range of rest-frame ultraviolet (UV) and optical diagnostics originally developed in the Local Universe. These observations challenge the validity of such classical methods in extreme high-redshift environments. We present an analysis of the ISM in three case studies at z=2-6 (MARTA-4327, the Sunburst Arc, and RXCJ2248-ID) using the multi-cloud photoionisation modelling framework HOMERUN, which models integrated galaxy spectra as linear combinations of photoionised regions spanning a wide range of ionisation parameter and density conditions. We show that emission from unresolved high-density (n_ = 10^4–10^7 e cm^ -3 ) regions boosts temperature-sensitive line ratios, but it shows a weak impact on standard optical density tracers. If this density-ionisation structure is not properly constrained, gas temperatures are biased high and the derived T_ e -method metallicities are underestimated by ∼0.15–0.3 dex, as we found for MARTA-4327. We modelled the rest-frame UV and optical data simultaneously to interpret the observed discrepancies between N/O estimates derived from UV and optical emission lines. In RXCJ2248-ID, the disagreement can arise naturally from ionisation and density effects alone. In contrast, the Sunburst Arc shows evidence of chemical stratification, with a nitrogen-enriched matter-bounded component coexisting with a chemically normal ionisation-bounded one. Finally, high-ionisation lines such as He II łambda1640 and C IV łambdałambda1548,1551 could be explained within a pure star-formation scenario that invokes matter-bounded regions, consistent with non-negligible Lyman-continuum leakage. However, in the case of RXCJ2248-ID, we cannot rule out a minor contribution from an active galactic nucleus (AGN) based solely on the observed emission-line fluxes. These results show that classical diagnostics can be significantly biased in high-redshift galaxies and highlight the need for self-consistent approaches to interpret the complex ISM conditions and chemical enrichment in the early Universe.