Only Nitrogen-Enhanced Galaxies Have Detectable UV Nitrogen Emission Lines at High Redshift

TL;DR

This work tackles whether nitrogen enhancement is universal among high-redshift galaxies or simply echoed by current detections. By computing JWST NIRSpec PRISM detection limits for UV nitrogen lines and comparing them to $N$-enhanced and $N$-normal photoionization models for both H II regions and AGN NLR, the authors assess the completeness of the nitrogen census at $z>5$. They find that present JWST data (CEERS) can only detect sources with $\ log(\rm N/O) \gtrsim -0.4$, while the deeper JADES survey can reach $\log(\rm N/O) \gtrsim -1.0$; crucially, UV nitrogen lines from $N$-normal galaxies at $z>5$ remain undetectable even in the deepest exposures, implying the current sample is biased toward highly nitrogen-enhanced systems. The study concludes that building a complete understanding of nitrogen enrichment in the early universe requires deep ($t_{\rm exp} \sim 40-500$ hours) spectroscopic surveys to uncover the true distribution of $\log(\rm N/O)$ at high redshift.

Abstract

The detections of bright UV nitrogen emission lines in some high-redshift galaxies suggest unexpectedly high nitrogen-to-oxygen ratios ($\log(\rm N/O)\gtrsim-1.0$) compared to local values ($\log(\rm N/O)\gtrsim-1.5$) at similar metallicities ($12+\log(\rm O/H)\lesssim8.0$). Although the presence of these `N-enhanced' galaxies indicates signatures of atypical chemical enrichment processes in the early universe, the prevalence of nitrogen enhancement in high-$z$ galaxies is unclear. So far, only $\sim$10 $z>5$ galaxies have nitrogen abundance measurements, and they all suggest elevated N/O ratios. Do all high-redshift galaxies exhibit elevated N/O ratios, or are we simply missing `N-normal' galaxies whose nitrogen abundances follow the local N/O scaling relation? To tackle these questions, we calculate the detection limits of UV NIII] or NIV] lines in current JWST surveys CEERS and JADES, and compare them to predictions from both `N-enhanced' and `N-normal' AGN narrow-line region and H II region photoionization models. We find that CEERS can only detect galaxies with significant nitrogen enhancement ($\log(\rm N/O)\gtrsim-0.4$), while JADES can only detect galaxies with moderately elevated N/O ratios compared to local values ($\log(\rm N/O)\gtrsim-1.0$). Even for the deepest exposure in JADES, UV nitrogen lines produced by `N-normal' galaxies at $z>5$ are too faint and thus not detectable, making their nitrogen abundance unmeasurable. Our results suggest that the existing sample of galaxies with measurable nitrogen abundances at $z\gtrsim5$ is incomplete and biased toward galaxies with significantly elevated N/O ratios. Deep ($t_{\rm exp}\sim40-500\,$hours) spectroscopic surveys will be crucial for building a complete sample to study nitrogen enrichment mechanisms in the early universe.