Quantifying the intrinsic variability due to randomness of the Auriga galaxy formation model

TL;DR

This study quantifies the intrinsic variability of the Auriga galaxy formation model by re-simulating a Milky Way–like galaxy (Au-6) seven times with identical initial conditions and physics but different random seeds. It finds that global halo properties and the central stellar disc are robust to stochasticity (variations typically below ~10% for structural properties and up to ~30% for the star formation rate in the last Gyr), while some internal morphologies and satellite timings show stochastic differences. When comparing numerical resolution, the study shows that a factor of 8 in mass resolution induces larger, systematic changes than intrinsic variability, underscoring the importance of resolution in interpreting galaxy-scale results. Overall, intrinsic variability due to stochastic processes in Auriga is modest relative to resolution effects, enabling robust comparisons across fixed-resolution ensembles and informing future model development and convergence studies.

Abstract

Numerical simulations have become an indispensable tool in astrophysics. To interpret their results, it is critical to understand their intrinsic variability, that is, how much the results change with numerical noise or inherent stochasticity of the physics model. We present a set of seven realisations of high-resolution cosmological zoom-in simulations of a Milky Way-like galaxy with the Auriga galaxy formation model. All realisations share the same initial conditions and code parameters, but draw different random numbers for the inherently stochastic parts of the model. We show that global galaxy properties at $z=0$, including stellar mass, star formation history, masses of stellar bulge and stellar disc, the radius and height of the stellar disk change by less than $10\%$ between the different realisations, and that magnetic field structures in the disc and the halo are very similar. In contrast, the star formation rate today can vary by a factor of two and the internal morphological structure of the stellar disc can change. The time and orbit of satellite galaxies and their galaxy properties when falling into the main halo are again very similar, but their orbits start to deviate after first pericenter passage. Finally, we show that changing the mass resolution of all matter components by a factor of $8$ in the Auriga model changes galaxy properties significantly more than the intrinsic variability of the model, and that these changes are systematic. This limits detailed comparisons between simulations at different numerical resolutions.

Figures (12)

• Figure 1: Stellar light projections of all seven realisations of Au-6 at resolution L4. All projections cover a $60\,\mathrm{kpc}\times60\,\mathrm{kpc}$ box for the face-on projections and a $60\,\mathrm{kpc}\times30\,\mathrm{kpc}$ for the edge-on projections. The galaxies all look very similar in colour and size. However, there is variation in their internal structure, in particular in the strength of the non-axisymmetric bar-like structure in the centre and their spiral features.
• Figure 2: Projections of dark matter surface density and velocity dispersion with a two-dimensional colormap Springel2006. All projections cover a $500\,\mathrm{kpc}\times500\,\mathrm{kpc}$ box. The panels show all seven realisations of Au-6 at resolution L4. The properties of the main dark matter halo as well as the positions of the most massive and many smaller satellites are very similar.
• Figure 3: Circular velocity curves computed from the total enclosed mass of all seven realisations of Au-6 at resolution L4. The main difference is in the value of the peak or plateau around $20\,\mathrm{kpc}$. At larger radii circular velocity agrees on the per cent level.
• Figure 4: Star formation histories of all seven realisations of Au-6 at standard resolution L4 (top panel). They are essentially identical before $z=1$. As the star formation rates decline towards $z=0$, their relative differences (middle panel) increase, up to a deviation of $\sim~50\%$ relative to the mean. The total stellar mass deviates only on the level of a few percent over the whole evolution (bottom panel).
• Figure 5: Stellar mass surface density (left panel) and cumulative stellar mass (right panel) for all seven realisations of Au-6 at resolution L4 The surface density profiles are very similar in the inner part as well as in their slope in the disc. The total stellar mass in the full halo is very similar as well, but the distribution in the halo varies somewhat between realisations.