Resolving stellar populations, star formation, and ISM conditions with JWST in a large spiral galaxy at z $\sim$ 2

TL;DR

The paper investigates how stellar populations, star formation, and ISM conditions organize in a massive spiral galaxy during the cosmic noon epoch ($z\sim2$). It exploits a rich JWST dataset (NIRSpec MSA, NIRCam imaging and WFSS) together with ground-based AO IFU observations and ALMA data to perform resolved ISM diagnostics and pixel-by-pixel SED fitting, achieving ~1 kpc spatial resolution. Key findings include a massive, heavily obscured bulge coexisting with bright, moderately metal-poor star-forming clumps (metallicity $12+\log(O/H)\approx8.5$) and an apparent enhancement of N/O due to dilution from inflowing metal-poor gas; sulfur appears subsolar with $\log(S/O)\approx-1.9$, hinting at complex nucleosynthetic and depletion histories. The study supports an inside-out growth scenario for K20-ID7 and demonstrates the power of JWST to deliver a holistic, spatially resolved view of galaxies at cosmic noon, with implications for gas accretion and disk evolution in the early universe.

Abstract

Cosmic noon represents the prime epoch of galaxy assembly, and a sweet spot for observations with the James Webb Telescope (JWST) and ground-based near-IR integral-field unit (IFU) spectrographs. This work analyses JWST NIRSpec Micro Shutter Array (MSA), NIRCam Wide Field Slitless Spectroscopy (WFSS) of K20-ID7, a large spiral, star-forming (SF) galaxy at z=2.2, with evidence for radial gas inflows. By exploiting the synergy with ground-based IFU ERIS observations, we conduct a comprehensive and resolved study of the interstellar medium (ISM) and stellar properties, from rest optical to near-IR, via emission-line diagnostics, resolved spectral energy distribution (SED) fitting of high-resolution imaging, and Pa$β$ line detection in NIRCam WFSS data. Our analysis reveals massive ($M_{\star}\simeq$(0.67-3.5)$\times$10$^{9}$ $M_{\odot}$) SF clumps with star formation rates (SFRs) ~3-24 $M_{\odot}$/yr, and quite low dust attenuation ($A_V\simeq$0.4), electron density ($n_{e}$<300 cm$^{-3}$), and ionisation (log(U)$\simeq -3.0$). The central bulge turns out to be modestly massive ($M_{\star}$=(7$\pm$3)$\times$10$^{9}$ M$_{\odot}$), heavily obscured ($A_V$=6.43$\pm$0.55), and likely to have formed most of its stellar mass in the past (SFR=82$\pm$42 $M_{\odot}$/yr over the last 100 Myr), yet still forming stars at a lower rate (SFR=12$\pm$8 M$_{\odot}$/yr over the last 10 Myr). We infer a metallicity 12+log(O/H)~8.54 and an apparent enhancement of the N/O abundance (log(N/O)$\simeq -1.0$) in all distinct galaxy regions, a likely consequence of dilution effects due to radial inflows of metal-poor gas. We measure a sub-solar sulfur abundance (log(S/O)$\simeq$-1.9). Finally, the radial stellar age profile reveals older stellar populations in the inner galaxy regions compared to the outskirts, pointing to an inside-out growth of K20-ID7.

Figures (16)

• Figure 1: Images of K20-ID7 in four representative HST/ACS (F435W, F814W) and JWST/NIRCam (F150W, F444W) filters, tracing galaxy emission from rest-frame UV to near-IR wavelengths. Redder NIRCam imaging unambiguously reveals a regular, unperturbed galaxy morphology, with well-defined spiral arms, and a compact bulge which appears brighter in the reddest F444W filter. All images are in units of nJy.
• Figure 2: NIRSpec MSA spectroscopy of K20-ID7. Left: Location of the two distinct MSA mask arrays, with open shutters drawn on the top of a NIRCam RGB image of K20-ID7, composed of F356W+F444W (red), F182M+F200W+F277W (green), and F090W+F115W+F150W (blue) filters. Right: PRISM/CLEAR spectra extracted from the seven coloured MSA shutters (those exhibiting some galaxy emission, same colors as in the left panel), using our custom-made reduction (described in Sect. \ref{['sec:data_msa']}). White shutters do not contain any galaxy emission. Spectra from different regions clearly differ in continuum shape and line intensities, over the total covered rest-frame near-UV to near-IR wavelength range. For the purpose of a better visualization, each spectrum has been vertically shifted by an arbitrary constant.
• Figure 3: Continuum-subtracted NIRCam F444W WFSS data of K20-ID7, zoomed-in over 3.95 -- 4.40 $\mu$m, encompassing Pa$\beta$ line emission. In the 2D slitless spectra (upper panel), Pa$\beta$ line emission is clearly spatially resolved perpendicularly to the dispersion axis, and also extended along the parallel direction, which leads to a double-peaked Pa$\beta$ line profile in the total 1D spectrum (lower panel). This is consequence of resolved spatial offsets combined with velocity gradients along the grism dispersion axis (discussed in Sect. \ref{['sec:pab']}).
• Figure 4: 2D maps at 25mas pixel scale of the main galaxy properties, as resulting from our resolved SED fitting with CIGALE of HST/ACS and JWST/NIRCam images. From left to right, the top panels show maps of stellar mass, and SFRs averaged over the last 10 Myr and 100 Myr, while the bottom ones $A_V,{\rm star}$, stellar age, and $\chi^2_{\rm red}$. The $\chi^2_{\rm red}$ map demonstrates the goodness of our pixel-by-pixel SED modeling, displaying values $\sim$ 1 across the entire galaxy.
• Figure 5: 2D spatial distribution of Pa$\beta$ in K20-ID7 from NIRCam grism F444W observations. Left. Pa$\beta$ line map, as reconstructed by grizligrizli, with white contours tracing F444W continuum emission. Pa$\beta$ emission appears offset with respect to continuum emission, along the x-axis (i.e., grism dispersion direction), due to velocity gradients. This is particularly evident for the northernmost clump (circled in red). Middle. Pa$\beta$ line map after correcting for the H$\alpha$ velocity-field model inferred from VLT/ERIS AO data (Pulsoni et al., in prep.). Both Pa$\beta$ and continuum emission now trace star-forming clumps and spiral arms well. Right. Pa$\beta$-traced SFR map, obtained by assuming negligible dust attenuation. Cyan circular apertures (diameters of 0.8$"$, 0.5$"$, 0.3$"$) and the central blue one (0.25$"$) are used to infer Pa$\beta$-based SFRs for the three brightest clumps and the bulge, respectively. The two Pa$\beta$ maps have a 0.05$"$ pixel scale, while the SFR map has been rebinned to a 0.025$"$ pixel scale, and smoothed with a Gaussian kernel of $\sigma = 1$ pix.