Testing the inference of kinematics from mock JWST NIRSpec/MSA observations of TNG50 galaxies at $z\sim2-6$

TL;DR

This work builds an end-to-end framework to test how well JWST NIRSpec MSA observations can recover ionised gas kinematics in high-redshift galaxies by forward-modelling thousands of TNG50 galaxies into mock NIRCam images and NIRSpec spectra and fitting them with a thin-disc model. It finds that, on average, the inferred velocities $v(r_{\rm e})$ and dispersions $\sigma_{0}$ reproduce intrinsic values within factors ~2 and ~1.5 respectively, especially for disc-like systems, but individual galaxies can show large biases due to non-disc morphologies, clumpy emission, and limited resolution. Population trends such as the weak redshift evolution of $\sigma$ and of $v/\sigma$ are broadly recovered, and the $\sigma$–SFR relation aligns with theoretical models of disc turbulence driven by feedback and mass transport. The study provides a rigorous framework for comparing JWST observations to simulations and emphasizes the need for more flexible, 3D morpho-kinematic models to accurately interpret high-redshift gas kinematics across diverse galaxy populations.

Abstract

We use the TNG50 galaxy formation simulation to generate mock JWST NIRCam and NIRSpec microshutter array (MSA) observations of H$α$-emitting gas in $M_*=10^8-10^{11.5}\,M_\odot$ star-forming galaxies at $z=2-6$. We measure morphological properties from the mock imaging through Sersic profile fitting, and gas rotational velocities ($v$) and velocity dispersions ($σ$) by fitting the mock spectra as thin, rotating discs. To test the efficacy of such simple parametric models in describing complex ionised gas kinematics, we compare the best-fit quantities to intrinsic simulation measurements. At $z=3$, we find that $v$ and $σ$ for aligned and resolved sources generally agree well with intrinsic measurements, within a factor of $\sim$2 and $\sim$1.5, respectively. The recovery of kinematics is robust for smooth, disc-like systems, but $v$ and $σ$ can be over- or underestimated by more than a factor of 2, respectively, for intrinsically elongated systems. The scatter in the recovery accuracy is larger at higher redshift, as TNG50 galaxies at $z>3$ deviate more strongly from the thin rotating disc assumption. Despite uncertain measurements for individual galaxies, we find that key population trends, such as the weak redshift evolution of $σ$ and $v/σ$ as well as the dependence of $σ$ on the global star formation rate, are broadly recovered by our kinematic modelling. Our work provides the end-to-end framework needed to compare NIRSpec MSA observations to cosmological simulations and to quantify observational biases in measuring ionised gas kinematics, highlighting the need for the development of dedicated models for high-redshift galaxies.

Figures (18)

• Figure 1: Intrinsic $v$, $\sigma$ and $v/\sigma$ vs redshift for TNG50 galaxies with stellar masses of $10^8-10^{11.5}\,{\rm M}_{\odot}$. The markers and the errorbars correspond to the median and the 16th and 84th percentiles, respectively. Different colours correspond to various measurement choices. Measurements from Pillepich_2019 are shown in grey squares. The fiducial measurements of this work (i.e., within a radius of $r_{\rm e}$ of H$\alpha$, 0.5 ckpc pixels, unweighted) are shown as black diamonds. Light green points correspond to SFR-weighted measurements with the same binning scale as the fiducial. Orange points correspond to measurements with a pixel size of 5 ckpc (i.e., 10x worse spatial resolution than the fiducial). Purple diamonds correspond to measurements with both SFR-weighting and a pixel size of 5 ckpc. The black, grey and purple points are shown in larger sizes for emphasis (same as in Fig. \ref{['fig:kin_vs_z']}). Intrinsic measurements of TNG50 galaxy kinematics can vary by up to a factor of $\lesssim$2 at fixed redshift, depending on e.g. choices of aperture, spatial binning and SFR-weighting.
• Figure 2: Schematic of the mocking procedure (Sec. \ref{['subsec:mock']}), and inference of the morphological and kinematic properties (Sec. \ref{['subsec:inference']}) adopted throughout this paper and shown here for a single example object. The positions, velocities, and SFRs of the star-forming gas in each simulated galaxy from TNG50 (left) is used to create intrinsic images or cubes. The images (cubes) are then convolved with a NIRCam (NIRSpec) PSF and noise is added (see 12$\times$12 collages). The mock images are then fit with pysersic to obtain morphological parameters such as $q$, PA, $n_{\rm s}$ and $r_{\rm e}$. These parameters are then used in conjunction with kinematic forward models (assuming a thin, rotating disc) to infer the kinematics of the galaxies. We thus obtain the rotational velocity at $r_{\rm e}$$v(r_{\rm e})$ and the velocity dispersion $\sigma_0$ for each galaxy.
• Figure 3: Sample distribution of mock H$\alpha$ image sizes ($r_{\rm e}$) versus galaxy stellar mass ($M_*$) for TNG50 galaxies with NIRCam $(S/N)>10$, shown as orange points. The running averages for H$\alpha$ effective radii and 3D stellar half-mass radii are indicated in orange and black, respectively. The number of galaxies and the redshift of the snapshot are annotated in each panel. The 3D stellar half-mass radii and 2D H$\alpha$ effective radii agree well with each other on average.
• Figure 4: Probability distribution of axis ratios ($q$) inferred from the mock NIRCam imaging of TNG50 galaxies with stellar masses of $10^8-10^{11.5}\,{\rm M}_{\odot}$ and NIRCam $(S/N)>10$. The columns describe the redshift of the galaxy while the rows show the various bins in stellar mass (see annotations on the right). The gray line indicates the axis ratio distribution of the top left panel ($z=2$, $M_*=[10^8-10^9)\,{\rm M}_{\odot}$) for comparison. More massive galaxies at lower redshifts exhibit the smallest axis ratios, down to values of 0.1.
• Figure 5: Collage of fit H$\alpha$ photometry and kinematics for aligned and resolved $z=2$ TNG50 galaxies with good kinematic fits (i.e., low $\chi^2$ - see Sec. \ref{['sssec:kin_model']}). The five columns correspond to NIRCam mock H$\alpha$ images, NIRSpec mock H$\alpha$ spectra, thin disc kinematic model fit to the spectra, intrinsic and best-fit rotational velocity profiles, and intrinsic and best-fit velocity dispersion profiles, respectively. The stellar masses and scales are indicated in the first column of each TNG50 galaxy. The NIRSpec/MSA mock slit is placed vertically in each case as indicated by the white rectangles. The thin disc model successfully reproduces intrinsic quantities in a variety of systems.