The KLEVER survey: Nitrogen abundances at $z\sim$2 and probing the existence of a fundamental nitrogen relation

TL;DR

This work investigates nitrogen abundances in z~2 galaxies from the KLEVER survey and compares them with a large local SDSS sample to test the N/O–O/H relation, the N/O–M* relation, and the existence of a fundamental nitrogen relation (FNR). By deriving metallicities and N/O using consistent, Te-based calibrations (N2O2 and N2S2) and cross-checking against the FMR framework, the study finds only a mild N/O enhancement at fixed O/H but a strong evolution of N/O with stellar mass, mirroring the evolution of the mass-metallicity relation. The KLEVER data are largely consistent with both the FMR and the FNR, supporting a picture in which nitrogen enrichment timescales and star formation histories shape chemical evolution in a way that remains coherent from z~2 to today. The results imply that gas inflows alone do not fully drive the FMR and that factors such as galaxy age and star formation efficiency play important roles, with the FNR-FMR symbiosis revealing a unified, redshift-invariant framework for understanding galaxy chemical evolution.

Abstract

We present a comparison of the nitrogen-to-oxygen ratio (N/O) in 37 high-redshift galaxies at $z\sim$2 taken from the KMOS Lensed Emission Lines and VElocity Review (KLEVER) Survey with a comparison sample of local galaxies, taken from the Sloan Digital Sky Survey (SDSS). The KLEVER sample shows only a mild enrichment in N/O of $+$0.1 dex when compared to local galaxies at a given gas-phase metallicity (O/H), but shows a depletion in N/O of $-$0.36 dex when compared at a fixed stellar mass (M$_*$). We find a strong anti-correlation in local galaxies between N/O and SFR in the M$_*$-N/O plane, similar to the anti-correlation between O/H and SFR found in the mass-metallicity relation (MZR). We use this anti-correlation to construct a fundamental nitrogen relation (FNR), analogous to the fundamental metallicity relation (FMR). We find that KLEVER galaxies are consistent with both the FMR and the FNR. This suggests that the depletion of N/O in high-$z$ galaxies when considered at a fixed M$_*$ is driven by the redshift-evolution of the mass-metallicity relation in combination with a near redshift-invariant N/O-O/H relation. Furthermore, the existence of an fundamental nitrogen relation suggests that the mechanisms governing the fundamental metallicity relation must be probed by not only O/H, but also N/O, suggesting pure-pristine gas inflows are not the primary driver of the FMR, and other properties such as variations in galaxy age and star formation efficiency must be important.

Figures (14)

• Figure 1: The N/O abundance as a function of N2O2, as measured via detection of auroral lines in stacked spectra of SDSS galaxies (Curti_2017). The red line represents the linear fit to the data for the new N/O calibration provided by Equation \ref{['N2O2_eq']}. Each point is colour-coded according to the number of galaxies within each stack, whilst the dashed black line shows, for reference, the log(N2O2) vs log(N/O) calibration presented by PMC09.
• Figure 2: The N/O abundance as a function of N2S2, as measured via detection of auroral lines in stacked spectra of SDSS galaxies (Curti_2017). The red line represents the linear fit to the data for the new N/O calibration provided by Equation \ref{['N2S2_eq']}. Each point is colour-coded according to the number of galaxies within each stack, whilst the dashed black line shows, for reference, the log(N2S2) vs log(N/O) calibration presented by PMC09.
• Figure 3: N/O derived from both N2O2 and N2S2 for galaxies in our SDSS sample, with contours showing the regions encompassing 30%, 60% and 90% of the total sample. The coloured lines show how the relationship between N2S2 and N2O2 varies as a function of star formation rate. At a fixed N2O2, more star forming galaxies have systematically larger values of N2S2, causing a discrepancy in the estimation of N/O.
• Figure 4: The nitrogen abundance as a function of the gas-phase metallicity for the SDSS and KLEVER galaxy samples. The grey contours show the regions encompassing 30%, 60%, 90% and 99% of the local SDSS sample. The white points show the median binned SDSS data with corresponding standard deviations, with the solid red curve representing the best fit to the data using Eq. \ref{['NOOH_eq']}. The black dash-dotted line represents the transition metallicity at 12+log(O/H) = 8.534. The KLEVER galaxies are shown in magenta with corresponding errorbars, with the average position of the KLEVER galaxies shown as a yellow star. The histogram in the top left shows the deviations of the SDSS sample and the KLEVER sample from the best-fit line to the local data. KLEVER galaxies have a large scatter, yet on average show a mild enrichment in N/O of 0.1 dex when compared to SDSS galaxies.
• Figure 5: The nitrogen abundance as a function of the stellar mass for the SDSS and KLEVER galaxy samples. The grey contours show the regions encompassing 30%, 60%, 90% and 99% of the total SDSS sample. The white points show the median binned SDSS data with corresponding standard deviations, with the solid red curve representing the best fit to the data using Equation \ref{['NOmass_eq']}. The magenta points are the KLEVER galaxies with corresponding errorbars, with the average position of the KLEVER galaxies shown as a yellow star. The histogram in the top left shows the deviations from the best fit in log(N/O) for both the KLEVER and SDSS samples. KLEVER galaxies have an average depletion in N/O of 0.35 dex compared to SDSS galaxies, suggesting significant redshift evolution in the stellar mass-N/O relation.