A deep learning approach to test the small-scale galaxy morphology and its relationship with star formation activity in hydrodynamical simulations

Abstract

Hydrodynamical simulations of galaxy formation and evolution attempt to fully model the physics that shapes galaxies. The agreement between the morphology of simulated and real galaxies, and the way the morphological types are distributed across galaxy scaling relations are important probes of our knowledge of galaxy formation physics. Here, we propose an unsupervised deep learning approach to perform a stringent test of the fine morphological structure of galaxies coming from the Illustris and IllustrisTNG (TNG100 and TNG50) simulations against observations from a subsample of the Sloan Digital Sky Survey. Our framework is based on PixelCNN, an autoregressive model for image generation with an explicit likelihood. We adopt a strategy that combines the output of two PixelCNN networks in a metric that isolates the small-scale morphological details of galaxies from the sky background. We are able to quantitatively identify the improvements of IllustrisTNG, particularly in the high-resolution TNG50 run, over the original Illustris. However, we find that the fine details of galaxy structure are still different between observed and simulated galaxies. This difference is mostly driven by small, more spheroidal, and quenched galaxies that are globally less accurate regardless of resolution and which have experienced little improvement between the three simulations explored. We speculate that this disagreement, that is less severe for quenched discy galaxies, may stem from a still too coarse numerical resolution, which struggles to properly capture the inner, dense regions of quenched spheroidal galaxies.

Are all post-starbursts mergers? HST reveals hidden disturbances in the majority of PSBs

“…Finally, while galaxy mergers themselves are easy to simulate, the effect of the merger on the galaxy post-coalescence and the evolution of the post-starburst phase is notoriously difficult to reproduce due to the complex interplay of starburst-and AGNfeedback and shocks, which all require poorly-constrained sub-grid routines. In particular, Zanisi et al (2021) show that Illustris simulations fail to re-create small-scale morphological features, important in our sample, in compact and quenched galaxies. Isolated merger simulations, tailored to provide high enough resolution to reproduce post-starburst properties, can provide a better morphological comparison (e.g., Zheng et al 2020;Lotz et al 2020).…”

Euclidpreparation

“…Storey-Fisher et al (2020) and Margalef-Bentabol et al (2020) have used GANs to detect outliers in imaging surveys. Autoregressive flows can be used to compare simulations and observations (e.g., Zanisi et al 2021).…”

Using Artificial Intelligence and real galaxy images to constrain parameters in galaxy formation simulations

“…We note that small scale galactic holes dominate features number for n=0.1, but the numbers for n=80 are too small to make a definitive comparison. More notably, Figure 4 shows that it is indeed possible, using machine transfer learning, to detect image features and separate in a quantitative way the different n-values, going beyond a simple visual inspection of the maps, and hence be employed to quantitatively compare simulations and observations in a holistic way (see also Zanisi et al 2021) which enables to constrain otherwise free model parameters.…”

“…Relevant in the context of our paper, is the very recent work by Zanisi et al (2021) comparing small scale morpholigical features of simulated galaxies from the Illustris-TNG project (Pillepich et al 2018) to the SDSS dataset. Those authors find strong disagreement between simulated morphologies and observed galaxies in terms of galaxy morphological features.…”