Observing the stellar halo of Andromeda in cosmological simulations: the Auriga2PAndAS pipeline

TL;DR

This paper compares the stellar halo observations of Andromeda with simulated halos from Auriga, using a pipeline to create comparable mock observations and identify differences related to galaxy formation and simulation resolution.

Contribution

The authors develop and release a pipeline to generate mock observations of simulated stellar halos, enabling direct comparison with survey data like PAndAS.

Findings

Simulations are generally compatible with observations.

Some systematic differences, such as metal-rich star prevalence, are identified.

Differences may stem from star formation quenching or simulation resolution issues.

Abstract

We present a direct comparison of the Pan-Andromeda Archaeological Survey (PAndAS) observations of the stellar halo of M31 with the stellar halos of 6 galaxies from the Auriga simulations. We process the simulated halos through the Auriga2PAndAS pipeline and create PAndAS-like mocks that fold in all observational limitations of the survey data (foreground contamination from the Milky Way stars, incompleteness of the stellar catalogues, photometric uncertainties, etc). This allows us to study the survey data and the mocks in the same way and generate directly comparable density maps and radial density profiles. We show that the simulations are overall compatible with the observations. Nevertheless, some systematic differences exist, such as a preponderance for metal-rich stars in the mocks. While these differences could suggest that M31 had a different accretion history or has a different mass compared to the simulated systems, it is more likely a consequence of an under-quenching of the star formation history of galaxies, related to the resolution of the Auriga simulations. The direct comparison enabled by our approach offers avenues to improve our understanding of galaxy formation as they can help pinpoint the observable differences between observations and simulations. Ideally, this approach will be further developed through an application to other stellar halo simulations. To facilitate this step, we release the pipeline to generate the mocks, along with the six mocks presented and used in this contribution.