Diffuse coronae in cosmological simulations of Milky Way-sized galaxies

We investigate the properties of halo gas using three cosmological `zoom-in' simulations of realistic Milky Way-galaxy analogs with varying sub-grid physics. In all three cases, the mass of hot ($T > 10^6$ K) halo gas is $\sim 1\%$ of the host's virial mass. The X-ray luminosity of two of the runs is consistent with observations of the Milky Way, while the third simulation is X-ray bright and resembles more closely a very massive, star-forming spiral. Hot halos extend to 140 kpc from the galactic center and are surrounded by a bubble of warm-hot ($T = 10^5 - 10^6$K) gas that extends to the virial radius. Simulated halos agree well outside 20-30 kpc with the $\beta$-model of Miller & Bregman (2013) based on OVII absorption and OVIII emission measurements. Warm--hot and hot gas contribute up to $80\%$ of the total gas reservoir, and contain nearly the same amount of baryons as the stellar component. The mass of warm-hot and hot components falls into the range estimated for $L^*$ galaxies. With key observational constraints on the density of the Milky Way corona being satisfied, we show that concealing of the ubiquitous warm-hot baryons, along with the ejection of just $20-30 \%$ of the diffuse gas out of the potential wells by supernova-driven outflows, can solve the "missing baryon problem". The recovered baryon fraction within 3 virial radii is $90\%$ of the universal value. With a characteristic density of $\sim 10^{-4}$ cm$^{-3}$ at $50-80$ kpc, diffuse coronae meet the requirement for fast and complete ram--pressure stripping of the gas reservoirs in dwarf galaxy satellites.