JWST NIRCam+NIRSpec: Interstellar medium and stellar populations of young galaxies with rising star formation and evolving gas reservoirs

TL;DR

This work presents a joint spectro-photometric analysis of three spectroscopically confirmed $z\sim7.6-8.5$ galaxies in the SMACS J0723 field using JWST NIRCam imaging and NIRSpec spectroscopy. By fitting NIRCam photometry and emission lines with Prospector, including flexible SFHs, variable dust laws, and nebular emission, the authors infer rising SFHs, stellar masses around $M_{\star}\sim10^{8}\,M_{\odot}$, and mass-weighted ages $t_{50}\sim3-4$ Myr, with evidence for older components. Gas-phase metallicities from the SED fits agree with direct $T_e$ measurements and reveal low metallicity in the most compact galaxy, while more massive, multi-component systems show higher $Z_{\rm gas}$ and possible merger-driven star formation. Emission lines significantly improve metallicity and dust constraints, though their absolute fluxes require a nuisance scaling $f_{\rm scale}$ to account for slit losses and other effects; SFH priors substantially influence inferred masses and ages. Overall, the study highlights the importance of combining JWST imaging and spectroscopy to robustly characterize early galaxies and demonstrates the method's sensitivity to modelling assumptions and data treatment.

Abstract

We present an interstellar medium and stellar population analysis of three spectroscopically confirmed $z>7$ galaxies in the ERO JWST NIRCam and JWST NIRSpec data of the SMACS J0723.3-7327 cluster. We use the Bayesian spectral energy distribution (SED) fitting code \texttt{Prospector} with a flexible star-formation history (SFH), a variable dust attenuation law, and a self-consistent model of nebular emission (continuum and emission lines). Importantly, we self-consistently fit both the emission line fluxes from JWST NIRSpec and the broad-band photometry from JWST NIRCam, taking into account slit-loss effects. We find that these three $z=7.6-8.5$ galaxies ($M_{\star}\approx10^{8}~M_{\odot}$) are young with rising SFHs and mass-weighted ages of $3-4$ Myr, though we find indications for underlying older stellar populations. The inferred gas-phase metallicities broadly agree with the direct metallicity estimates from the auroral lines. The galaxy with the lowest gas-phase metallicity ($\mathrm{Z}_{\rm gas}=0.06~\mathrm{Z}_{\odot}$) has a steeply rising SFH, is very compact ($<0.2~\mathrm{kpc}$) and has a high star-formation rate surface density ($Σ_{\rm SFR}\approx22~\mathrm{M}_{\odot}~\mathrm{yr}^{-1}~\mathrm{kpc}^{-2}$), consistent with rapid gas accretion. The two other objects with higher gas-phase metallicity show more complex multi-component morphologies on kpc scales, indicating that their recent increase in star-formation rate is driven by mergers or internal, gravitational instabilities. We discuss effects of assuming different SFH priors or only fitting the photometric data. Our analysis highlights the strength and importance of combining JWST imaging and spectroscopy for fully assessing the nature of galaxies at the earliest epochs.

Figures (9)

• Figure 1: Composite imaging with NIRSpec slit and photometric data for Galaxy ID 04590 at redshift $z=8.50$. Galaxy ID 04590 is the centre of the F090W-F150W-F444W false-colour image on top of which the NIRSpec slit falls (upper left; scale bar and compass provided for reference). The source on the very left is a lower-redshift object given that it is visible in the F090W filter (see images on the upper right, middle row), while the source directly on the left might be at a similar redshift, but is not considered part of Galaxy ID 04590. The small residuals between the model and the data demonstrate the effectiveness of forcepho in capturing the multiband scene. forcepho may also be used to model the source photometry with high-accuracy (bottom right; black points correspond to the top component of the forcepho model. NIRCam filter transmissivity curves are shown as shaded regions) and is in rough agreement with 0.3" radius aperture photometry. The forcepho photometry is used in conjunction with the NIRSpec spectrum (lower left; red line. Shaded region shows the chip gap.) to supply the empirical constraints on our Prospector spectral energy distribution model (lower right; gray line). Both the photometry and emission line strengths are fit simultaneously by Prospector. The resulting SED model reveals a star-forming galaxy with strong emission line flux contributions to the rest-frame optical light.
• Figure 2: Same as Fig. \ref{['fig:data4590']}, but for Galaxy ID 06355 at redshift $z=7.67$. This galaxy is clearly more extended than Galaxy ID 04590 and consistent of multiple components.
• Figure 3: Same as Fig. \ref{['fig:data4590']}, but for Galaxy ID 10612 at redshift $z=7.66$. This galaxy is consists of two components.
• Figure 4: Comparison of inferred model with the observed data. Top panels: SED of the galaxies. Blue points are the observed photometry from NIRCam, while the orange boxes mark the posterior photometry from our model. The red solid line (and the shaded region) is the median SED of the posterior (and the 16th-84th percentile). The $\chi^2$ values indicate that our fits are reasonable. Bottom panels: NIRSpec observations versus model posteriors for the emission lines. The black circles mark the lines that are fitted, while the grey points indicate the masked emission lines, which have not been fitted. Overall, the emission lines are well reproduced with our model after rescaling the model-based fluxes by $f_{\rm scale}$, which can be motivated by slit-losses and physical effects such as LyC dust absorption or LyC escape.
• Figure 5: Star-formation histories (SFHs) over the past 100 Myr. The solid red line indicates the median of the posterior distribution inferred with our fiducial model, while the shaded region mark the 16th-84th percentiles. All of the galaxies are consistent with a recent burst of star formation: the SFRs in the past 5 Myr are higher by a factor of $\geq10$ than 20 Myr ago. Importantly, these are not the full SFHs, which extend back to $z=20$.