Under Pressure: Decoding the Effect of High Densities on Derived Nebular Properties

Abstract

Recent JWST observations have uncovered a population of compact, high-redshift ($z>6$) galaxies exhibiting extreme nebular conditions and enhanced nitrogen abundances that challenge standard chemical evolution paradigms. We present a joint UV and optical abundance analysis using a new suite of $\texttt{Cloudy}$ photoionization models covering a wide density range ($n_e=10^2-10^9$ cm$^{-3}$), combined with HST and JWST spectroscopy for a sample of star-forming galaxies across $0.0\lesssim z \lesssim10.6$. We find that assuming uniform, low-density conditions ($n_e\sim10^2$ cm$^{-3}$) in high-density environments ($n_e\sim10^5$ cm$^{-3}$) can bias nebular diagnostics by overestimating $T_e$ (up to 1800 K), overpredicting $\log U$ (by $>1$ dex), and underestimating O/H (up to 0.67 dex), while only modestly inflating N/O. Therefore, robust abundance determinations at high-$z$ require a multi-phase density model. Using this model, we recalculate O/H and N/O abundances for our sample and present the first $\log U$ diagnostics and ICFs for high-ionization UV N lines. We find that the UV tracers systematically overestimate N/O by $\sim0.3-0.4$ dex relative to the optical benchmark. We find that N/O increases with redshift, correlating with both $n_e$ and star formation rate surface density ($\rmΣ_{SFR}$), suggesting that N/O is temporarily enhanced in compact, high-pressure environments. However, the $n_e$ evolution with $z$ is more gradual than the $(1+z)^3$ scaling of virial halo densities, suggesting that $n_e$evolution is shaped by both cosmological structure growth and baryonic processes. These trends point to prompt N/O enrichment potentially driven by very massive stars, with key implications for interpreting UV emission and determining reliable chemical abundances from JWST observations of the early universe.

Figures (21)

• Figure 1: Rest-UV spectra for the Low-$z$ Sample (see Section \ref{['sec:lzsamp']} and Table \ref{['tab:samp_prop']}). The detected N4] $\lambda$$\lambda$1483,1487 (left) and N3] $\lambda$1750 (right) emission profiles are shown; four galaxies have N4] $\lambda$$\lambda$1483,1487 emission and five have N3] $\lambda$1750 emission, but none have both. We plot the emission line fits (blue line) for the eight of these galaxies without reported N4] and N3] line fluxes in the literature.
• Figure 2: UV N-Emitting Sample Properties
• Figure 3: Electron Temperature and Density Diagnostics
• Figure 4: The emissivities of N4] $\lambda$$\lambda$1483,1487, N3] $\lambda$1750, and O3] $\lambda$$\lambda$1661,1666, determined with PyNeb, are shown in the left column, while the right column shows the same for the emissivities of [O2] $\lambda$$\lambda$3727,3730, [N2] $\lambda$6584, and [O2] $\lambda$$\lambda$7321,7332. The upper panels show emissivities as a function of $T_e$ (5,000 to 25,000 K), and the lower panels describe these emissivities as a function of $n_e$ (10$^2$ - 10$^9$ cm$^{-3}$). Note that we show the full range of densities used in photoionization models to show the different ranges each line is sensitive over. The top panels illustrate that $n_e$ has a larger impact on optical lines that have lower critical $n_e$ , while the bottom panels show that UV lines are more impacted by $T_e$ than the optical lines given that they have higher excitation energies.
• Figure 5: Emission Line Characteristics