IllustrisTNG in the HSC-SSP: No Shortage of Thin Disk Galaxies in TNG50

TL;DR

This study compares the shapes of galaxies in the TNG50 simulation with real observations from HSC-SSP, finding excellent agreement for certain mass ranges and highlighting the importance of mock observations for accurate comparisons.

Contribution

It introduces a forward-modeling approach for shape comparison, demonstrating that TNG50 accurately reproduces the abundance of thin disk galaxies in the observed universe.

Findings

Excellent agreement between TNG50 and observations for galaxies with $10 \,\leq\, \log (M_{\star}/M_{\odot}) \leq 11.5$.

Low-mass TNG50 galaxies are thicker than observed, likely due to dynamical heating effects.

TNG50 galaxies are more concentrated at low and high masses but less at intermediate masses.

Abstract

We perform a thorough analysis of the projected shapes of nearby galaxies in both observations and cosmological simulations. We implement a forward-modeling approach to overcome the limitations in previous studies, which hinder accurate comparisons between observations and simulations. We measure axis ratios of $z=0$ (snapshot 99) TNG50 galaxies from their synthetic Hyper Suprime-Cam Subaru Strategic Program (HSC-SSP) images and compare them with those obtained from real HSC-SSP images of a matched galaxy sample. Remarkably, the comparison shows excellent agreement between the observations and the TNG50 simulation, challenging previous claims that $\Lambda$CDM models underproduced the abundance of thin galaxies. Specifically, for galaxies with stellar masses $10\leq \log (M_{\star}/M_{\odot}) \leq 11.5$, we find $\lesssim 0.1\sigma$ tensions between the observations and the simulation, a stark contrast to the previously reported $\gtrsim 10\sigma$ tensions. We reveal that low-mass galaxies ($M_{\star}\lesssim 10^{9.5}\,M_{\odot}$) in TNG50 are thicker than their observed counterparts in HSC-SSP and attribute this to the spurious dynamical heating effects that artificially puff up galaxies. We also find that, despite the overall broad agreement, TNG50 galaxies are more concentrated than the HSC-SSP ones at the low- and high-mass end of the stellar mass range of $9.0\leq \log (M_{\star}/M_{\odot}) \leq 11.2$ and are less concentrated at intermediate stellar masses. But we argue that the higher concentrations of the low-mass TNG50 galaxies are not likely the cause of their thicker/rounder appearances. Our study underscores the critical importance of conducting mock observations of simulations and applying consistent measurement methodologies to facilitate proper comparison with observations.