Simulating supernova driven winds in the Galactic center

We present here two videos showing the evolution of the gas column density in the central region of a simulation of the Milky Way ran with the PIERNIK code. The first video exhibits the face-on and edge-on view of the central 2 kpc for the entire simulation, displaying the formation and evolution of the Central Molecular Zone, and the outflows generated from it. The second video focuses on the outflow region, showing the galaxy edge-on up to 2 kpc heights both above and below the Galactic plane, from t~190 Myr (time at which we start analyzing the outflows).

The simulation models the entire Milky Way disk with an emphasis on the central region, reaching 3 pc resolution in the center and in the outflows generated from it. It includes gas cooling, gas heating (photo-electric, low energy cosmic rays, and supernovae), Nbody particles for star formation and feedback, and a fixed gravitational potential accurately describing our Galaxy (Hunter et al. 2024). The bar potential is introduced gradually over 150 Myr, time after which we increase the resolution in the center. Supernova feedback is implemented following the TIGRESS scheme (Kim & Ostriker 2017) in both kinetic and thermal forms, with a dependence on whether the Sedov-Taylor phase is resolved. This simulation does not include cosmic rays, as they will be included in a future study for comparison. The outflows generated in this simulation vary significantly over time due to cycles of star formation with episodic starbursts, and we observe the presence of cool HI gas clouds in the outflows up to 2 kpc distance from the Galactic mid-plane. These clouds are accelerated out of the disk by entrainment from the hot wind to velocities of order 100 km/s, consistently with observations of the Milky Way nuclear outflows. The results of this model suggest that the population of cold clouds observed in the outflows of the Galactic center could be generated from stellar feedback alone.

For more details and further analysis, please refer to the corresponding paper :

Peschken, N. ; Hanasz, M. ; Naab, T. ; Wóltański, D.; Gawryszczak, A. 2026 MNRAS, 522, 5529