ERGO-ML: Comparing IllustrisTNG and HSC galaxy images via contrastive learning

Abstract

Modern cosmological hydrodynamical galaxy simulations provide tens of thousands of reasonably realistic synthetic galaxies across cosmic time. However, quantitatively assessing the level of realism of simulated universes in comparison to the real one is difficult. In this paper of the ERGO-ML series (Extracting Reality from Galaxy Observables with Machine Learning), we utilize contrastive learning to directly compare a large sample of simulated and observed galaxies based on their stellar-light images. This eliminates the need to specify summary statistics and allows to exploit the whole information content of the observations. We produce survey-realistic galaxy mock datasets resembling real Hyper Suprime-Cam (HSC) observations using the cosmological simulations TNG50 and TNG100. Our focus is on galaxies with stellar masses between $10^9$ and $10^{12}{M}_{\odot }$ at $z=0.1-0.4$. This allows us to evaluate the realism of the simulated TNG galaxies in comparison to actual HSC observations. We apply the self-supervised contrastive learning method NNCLR to the images from both simulated and observed datasets (g, r, i - bands). This results in a 256-dimensional representation space, encoding all relevant observable galaxy properties. Firstly, this allows us to identify simulated galaxies that closely resemble real ones by seeking similar images in this multi-dimensional space. Even more powerful, we quantify the alignment between the representations of these two image sets, finding that the majority ($\gtrsim 70$ per cent) of the TNG galaxies align well with observed HSC images. However, a subset of simulated galaxies with larger sizes, steeper Sersic profiles, smaller Sersic ellipticities, and larger asymmetries appears unrealistic. We also demonstrate the utility of our derived image representations by inferring properties of real HSC galaxies using simulated TNG galaxies as the ground truth.

26 pages, 17 figures. Comments are welcome. Highlights in Fig. 3, 7 and 13

References in corpus (28)

• The EAGLE project: Simulating the evolution and assembly of galaxies and their environments
• Introducing the Illustris Project: Simulating the coevolution of dark and visible matter in the Universe
• The EAGLE simulations of galaxy formation: calibration of subgrid physics and model variations
• 3D-HST+CANDELS: The Evolution of the Galaxy Size-Mass Distribution since $z=3$
• Simba: Cosmological Simulations with Black Hole Growth and Feedback
• Introducing the Illustris Project: the evolution of galaxy populations across cosmic time
• The Evolution of Galaxy Structure over Cosmic Time
• Third Data Release of the Hyper Suprime-Cam Subaru Strategic Program
• The Size Evolution of Star-forming and Quenched Galaxies in the IllustrisTNG simulation
• SKIRT: an Advanced Dust Radiative Transfer Code with a User-Friendly Architecture
• Modelling the Pan-Spectral Energy Distribution of Starburst Galaxies: IV The Controlling Parameters of the Starburst SED
• Galactic Angular Momentum in the Illustris Simulation: Feedback and the Hubble Sequence
• Synthetic Galaxy Images and Spectra from the Illustris Simulation
• Deep learning predictions of galaxy merger stage and the importance of observational realism
• Galaxies in the Illustris simulation as seen by the Sloan Digital Sky Survey - II: Size-luminosity relations and the deficit of bulge-dominated galaxies in Illustris at low mass
• Identifying Galaxy Mergers in Observations and Simulations with Deep Learning
• Hyper Suprime-Cam Subaru Strategic Program: A Mass-Dependent Slope of the Galaxy Size-Mass Relation at $z<1$
• IllustrisTNG in the HSC-SSP: image data release and the major role of mini mergers as drivers of asymmetry and star formation
• A missing outskirts problem? Comparisons between stellar halos in the Dragonfly Nearby Galaxies Survey and the TNG100 simulation
• PetroFit: A Python Package for Computing Petrosian Radii and Fitting Galaxy Light Profiles
• ERGO-ML I: Inferring the assembly histories of IllustrisTNG galaxies from integral observable properties via invertible neural networks
• Morphological signatures of mergers in the TNG50 simulation and the Kilo-Degree Survey: the merger fraction from dwarfs to Milky Way-like galaxies
• MaNGIA: 10,000 mock galaxies for stellar population analysis
• ERGO-ML: Towards a robust machine learning model for inferring the fraction of accreted stars in galaxies from integral-field spectroscopic maps
• Morphological Parameters and Associated Uncertainties for 8 Million Galaxies in the Hyper Suprime-Cam Wide Survey
• A machine learning approach to infer the accreted stellar mass fractions of central galaxies in the TNG100 simulation
• Mock Galaxy Surveys for HST and JWST from the IllustrisTNG Simulations
• Dark against luminous matter around isolated central galaxies: a comparative study between modern surveys and Illustris-TNG

Cited by in corpus (6)

• Zooming in on the Circumgalactic Medium with GIBLE: Tracing the Origin and Evolution of Cold Clouds
• Cosmic reflections I: the structural diversity of simulated and observed low-mass galaxy analogues
• IllustrisTNG in the HSC-SSP: No Shortage of Thin Disk Galaxies in TNG50
• Gamma Analytical Modeling Evolution (GAME) I: The physical implications of deriving the stellar mass functions from z=0 to z=8
• Bridging Simulations and Observations: New Insights into Galaxy Formation Simulations via Out-of-Distribution Detection and Bayesian Model Comparison
• The detectability of bars at high redshift: a case study using Euclid-like mock observations of TNG50 simulated galaxies