Evolution of gas-phase metallicity across cosmic time

• Main Author: Acharyya, Ayan
• Format: Article in Journal/Newspaper
• Language: English
• Published: The Australian National University 2022
• Subjects: Milky Way; sami
• Online Access: https://dx.doi.org/10.25911/zpaw-gp13; https://openresearch-repository.anu.edu.au/handle/1885/258981

Chemical enrichment of the interstellar medium (ISM) in galaxies by generations of star-formation is a crucial ingredient to tracing galaxy evolution. Local galaxies have been the rigorously studied owing to their proximity, whereas distant and relatively faint high-redshift (1 < z < 4) galaxies remain poorly understood. In order to trace chemical evolution in high-z galaxies, it is necessary to develop and test new UV nebular diagnostics because at high-z we primarily observe the rest-frame UV spectra. We compare new rest-frame UV and existing optical nebular emission line diagnostics, by applying both sets, for the first time, to the brightest known lensed galaxy at cosmic noon (z~2) RCS0327-E. We infer the metallicity (12+log(O/H)), ionisation parameter (log (q)), electron temperature (T_e), electron density (n_e) and ISM pressure (log(P/k)) via UV and optical emission line diagnostics. Moreover, we extend an existing Bayesian inference code (IZI) to 3D (IZIP), enabling it to infer 12+log(O/H), log (q) and log (P/k) simultaneously. We find that 12+log(O/H) is harder to infer based on UV emission lines alone while inferred values log (q) and T_e are broadly similar for UV and optical diagnostics. UV diagnostics yield a higher (~1.5 dex) pressure and density than optical, because the UV lines probe the inner, denser parts of nebula. We employ the new UV diagnostics to obtain metallicities for the full MEGaSaURA sample -- consisting of rest-frame UV spectra of 22 bright, star-forming, gravitationally lensed galaxies in redshift 1.6 < z < 3.6. We focus on the N3O3 ([N III] 1750/[O III] 1660,6) diagnostic because this ratio has the highest number of detections (N = 11) in our sample. Our sample expands the literature of lensed galaxy metallicities in the above redshift range by ~30%. We compare the N3O3 metallicities with those inferred from IZIP, as well as stellar and optical gas-phase metallicities from the literature. We find it difficult to draw strong conclusions regarding the redshift evolution of metallicity in the MEGaSaURA sample because of large scatter in metallicities -- both intrinsic to the sample and across different diagnostics. Probing distant galaxies not only requires us to study the rest-frame UV regime, but also leads to worsening spatial resolution of the observations. This impacts studies of the spatial distribution of metals, particularly the metallicity gradient { which is crucial to understand gas flow history of galaxies. We quantify the impact of spatial resolution and signal-to-noise ratio (SNR) on inferred metallicity gradients in integral field unit (IFU) datacubes by modeling a suite of synthetic IFU data cubes from a simulation of an isolated, Milky Way-type disk galaxy. We find that coarse resolution leads to artificially shallow inferred metallicity gradients, and that recovery of the true gradient with an accuracy of ~10% requires resolving the galaxy scale length by at least 4-5 resolution elements. Based on these numerical experiments, we propose a method to correct observed metallicity gradients for spatial resolution effects. Thereafter, we correct the observed metallicity gradients of the MaNGA, CALIFA and SAMI samples and study the mass-metallicity gradient relation. We find that while the correction has little effect on the overall average mass-metallicity gradient relation, individual galaxies can undergo significant correction depending on the resolution. Our results are qualitatively consistent across the three different IFU samples -- more massive galaxies exhibit steeper gradients, up to log (M_*/M_sun) < 10.5, above which the the gradients are shallower, likely due to an increased prevalence of major mergers leading to gas mixing. Our models and method of correcting for resolution effects, along with our cross-survey comparisons, will facilitate planning, analysing and comparing current and future large surveys.