Low-z Circum-Galactic Medium: Insights from Local Group Simulations

Gas flows around galaxies (aka the Circum-Galactic Medium or CGM) closely shape many important galactic observables such as the star formation history, baryon and metal budget and distribution, energy budget, etc. Our neighbourhood, the Local Group (LG) of galaxies, possesses the richest CGM observational data, both qualitatively and quantitatively. Constrained LG simulations are extremely effective in such specific use cases because they provide a decent representation of the real universe LG and its associated CGM, unlike conventional cosmological simulations.

I use one such suite of high-resolution simulations, Hestia, for comparing the properties of the simulated LG CGM with the observed one. Hestia simulations use the AREPO-based hybrid Eulerian-Lagrangian approach in order to solve hydrodynamical equations as well as incorporate state-of-the art Auriga galaxy formation model to include key astrophysical processes. I model five tracer ions (chosen to represent various CGM phases) using Cloudy post-processing tool. I confirm certain key observed CGM properties - the cold/cool CGM is clumpy and distributed close to the galactic centers while the hot CGM is diffuse and volume-filling.

I find that the Andromeda analogues in Hestia tend to produce less gas in the galactic outskirts, unlike that seen in the case of real observations. Among various likely reasons causing this discrepancy, I probe the possibility of Andromeda CGM observations being overestimated due to a bias from intervening Milky Way CGM gas clouds.

By generating mock Andromeda observations within Hestia, I am able to demonstrate, for two out of three Hestia realizations, that it is possible that a few gas clouds in the Milky Way’s CGM can be mistakenly associated with Andromeda’s CGM in observations. This could, at least partly, result in Andromeda CGM observations being overestimated and hence, at tension with simulation results. I propose that future simulation studies could provide an even more accurate and statistically even more robust bias estimation to observers such that LG CGM observations can be better corrected for.