The Pandora project. II: how non-thermal physics drives bursty star formation and temperate mass-loaded outflows in dwarf galaxies

Abstract

Dwarf galaxies provide powerful laboratories for studying galaxy formation physics. Their early assembly, shallow gravitational potentials, and bursty, clustered star formation histories make them especially sensitive to the processes that regulate baryons through multi-phase outflows. Using high-resolution, cosmological zoom-in simulations of a dwarf galaxy from \textit{the Pandora suite}, we explore the impact of stellar radiation, magnetic fields, and cosmic ray feedback on star formation, outflows, and metal retention. We find that our purely hydrodynamical model without non-thermal physics - in which supernova feedback is boosted to reproduce realistic stellar mass assembly - drives violent, overly enriched outflows that suppress the metal content of the host galaxy. Including radiation reduces the clustering of star formation and weakens feedback. However, the additional incorporation of cosmic rays produces fast, mass-loaded, multi-phase outflows consisting of both ionized and neutral gas components, in better agreement with observations. These outflows, which entrain a denser, more temperate ISM, exhibit broad metallicity distributions while preserving metals within the galaxy. Furthermore, the star formation history becomes more bursty, in agreement with recent JWST findings. These results highlight the essential role of non-thermal physics in galaxy evolution and the need to incorporate it in future galaxy formation models.

Figures (13)

• Figure 1: Projected views centred on the Pandora dwarf illustrating outflow events across our four studied models. The area displayed in each panel approximately encompasses the central galactic region ($0.2\,r_\text{halo}$), 5 $\,\mathrm{kpc}$ across. Each of the columns, from left to right, corresponds to the following models: standard hydrodynamics (HD), calibrated SN feedback hydrodynamics (HD+Boost), radiative transfer and MHD (RTiMHD), and our 'full-physics' simulation with radiation, CRs and MHD (RTCRiMHD). (Top row panels) Mass-weighted densities for the gas (blue) and stellar (gold) components. (Second row panels) Gas temperature maps. (Third row panels) Gas density flows separated according to radial velocity ($v_\text{r}$) into inflows (blue; $v_\text{r} < -v_\text{vir} = 12.5\,\,\,\mathrm{km}\,\,\mathrm{s}^{-1}$) and outflows (golden; $v_\text{r} > v_\text{vir}$, and red for faster outflows $v_\text{r} > 40\,\,\,\mathrm{km}\,\,\mathrm{s}^{-1}$). For further visual guidance, in this row we display the full gas density field using a gray scale. (Bottom row panels) Mass-weighted stellar metallicity. To guide the eye, the overlaid contours display the $[10^{5}, 10^{6}, 10^{7}, 10^{8}]\,\,\mathrm{M_\odot}\,\,\mathrm{kpc}^{-2}$ stellar surface density isocontours. Galactic outflows in HD and HD+Boost appear hotter and with rugged and defined shock structures. Including radiation reduces the outflows, whereas CRs lead to temperate and more homogeneous outflows, as well as complex metallicity topology.
• Figure 2: (Top row) Star formation rate (SFR) for our hydrodynamical models HD and HD+Boost (panel a1), and for our more complex models including stellar radiation and magnetic fields RTiMHD, as well as CRs, RTCRiMHD (panel a2). (Central row) Outflowing gas mass-loading factor ($\eta = \dot{M}_\text{gas} /$ SFR; see text for further details) radially ejected through a spherical shell positioned at $0.5\,r_\text{halo}$. Including radiation leads to a more continuous SFR, whereas its combination with CRs leads to a more bursty star formation history. The calibrated feedback HD+Boost and full physics RTCRiMHD simulations attain the highest outflow mass-loading factors. The RTiMHD has the lowest $\eta$ values, albeit those remain comparable to the HD scenario. (Bottom row) Fourier transform of the star formation history for the same models, normalised to the zero-frequency mode (left panel). The average power is separated into the three frequency regimes corresponding to long, intermediate and short timescales (right panel).
• Figure 3: PDF of the small-scale environmental thermodynamical properties where SN events take place in each of the models: gas density (top panel) and gas temperature (bottom panel). Across all simulations, most SNe take place at densities $\rho \sim 10^{-26} \,\,\mathrm{g}\,\,\mathrm{cm}^{-3}$. Different galaxy formation physics leads to notably different fractions of SN events taking place at high densities ($\rho > 10^{-24} \,\,\mathrm{g}\,\,\mathrm{cm}^{-3}$). Early feedback in the form of stellar radiation significantly increases the number of SNe at high densities by reducing star formation (and consequently SN) clustering, whereas 'boosted' SN feedback leads to their drastic reduction, due to effectiveness of few events in dispersing dense gas. The RTCRiMHD simulation shows the strongest temporal SN clustering, while maintaining the proportion of events taking place in dense, photo-heated regions.
• Figure 4: Radial profiles for various outflow-related quantities, centred on the Pandora dwarf. Profiles correspond to star formation episodes at $z = 3.5$, and are representative of times with non-negligible star formation (Appendix \ref{['ap:profile_variability']}). The RTCRiMHD model in particular exhibits larger variations due to its burstier star formation. (Top row) Pressure profiles for the thermal (solid), magnetic (dotted), radiative (dot-dashed), and cosmic ray (dashed) components. (Central row) Fractional pressures for the same components as shown in the top row. (Bottom row) Density profiles, separated into total (solid line) and outflowing components (dotted line). We overlay the radial profile of the net outflowing momentum using golden solid lines, displayed as dashed for radii when the net momentum is inflowing. We include additional vertical lines in each panel corresponding to twice the half-mass radii of the stellar component (thick black dashed line, $2\, r_{*}$), and the 0.2, 0.5 and 1.0 $r_\text{halo}$ (gray dashed lines). In the models without CRs, outflowing gas (e.g., $\sim 1\,\,\mathrm{kpc}$ in panel b2) is thermally supported (panels a1 and a2). Conversely, outflows in the RTCRiMHD model ($\sim 3\,\,\mathrm{kpc}$ in panel b3) are dominated by CR pressure (panel a3). Overall, in the RTCRiMHD simulation, thermal pressure dominates across most radii - although with important contributions from the other pressure components. CRs dominate in the outflow bubbles and provide significant support at $r \lesssim 3\,r_{*}$. Magnetic pressure is only globally important inside the galaxy, whereas the radiative photo-heating pressure becomes more important at large radii, driving the thermal pressure through photo-heating.
• Figure 5: Gas mass fraction distribution functions as a function of gas density (left column) and gas temperature (right column) for the central galactic region ($r < 0.2\,r_\text{halo}$). Distributions are shown for snapshots of non-negligible star formation activity in the galaxy, at $z = 3.5$ (except for HD+Boost, see main text). Total gas fraction (gray histograms), inflowing gas fraction (blue histograms, $v_\text{gas,radial} < v_\text{vir} \sim 12.5\,\,\mathrm{km}\,\,\mathrm{s}^{-1}$), outflowing gas fraction (orange histograms, $v_\text{gas,radial} > v_\text{vir}$) and escaping gas fraction (red histogram, $v_\text{gas,radial} > v_\text{esc} (r = 0.2\,r_\text{halo})$) are shown for our four simulation models (top to bottom rows). Each temperature panel lists the SFR of each galaxy during the past 100 Myr, and the total mass fraction of the inflowing, outflowing and escaping gas component. The RTCRiMHD simulation has lower inflow rates as well as higher outflow and escaping rates per given SFR. The outflowing gas is also colder and denser, particularly notable in the escaping gas regime.