Auriga Galaxy Simulations

• Auriga simulations are high-resolution cosmological MHD simulations that model galaxy evolution including dark matter, gas, stars, and magnetic fields.
• They employ the AREPO moving-mesh code with advanced subgrid physics to mimic star formation, feedback, and galaxy dynamics with precise resolution.
• The suite offers insights into secular evolution, bar formation, CGM diversity, and subhalo suppression, providing robust benchmarks for theoretical models.

Auriga simulations are a comprehensive suite of cosmological magneto-hydrodynamical (MHD) zoom-in simulations of galaxies spanning dwarf to Milky Way mass scales. They utilize the moving-mesh code AREPO and embed a modern galaxy formation physics model within the $\Lambda$CDM framework. Their primary goal is to track the coupled evolution of dark matter, gas, stars, black holes, and magnetic fields, providing a predictive platform for galaxy dynamics, structure, and scaling relations across cosmic time.

1. Numerical Framework and Physical Ingredients

The Auriga simulations employ AREPO, which solves the Euler equations for MHD on a moving unstructured Voronoi mesh and includes key subgrid physics:

• Cosmology: Planck2013 parameters ($\Omega_m=0.307$, $\Omega_\Lambda=0.693$, $\Omega_b=0.048$, $h=0.6777$, $\sigma_8=0.829$).
• Resolution: Milky Way-mass halos use $m_{\rm DM} \simeq 3\!\times\!10^5\,M_\odot$, $m_{\rm bary} \simeq 5\!\times\!10^4\,M_\odot$, and softening lengths $\epsilon_* = 369$ pc (physical; gas adaptive from $\approx500$ pc to 1.85 kpc).
• Hydro and Gravity: Gravity via TreePM; ideal MHD with divergence cleaning; cooling and heating (primordial + metal lines, UVB); two-phase pressurized ISM ($n_{\rm th}=0.13$ cm$^{-3}$, Springel & Hernquist 2003); stochastic star formation and stellar feedback (winds with mass loading $\eta \propto v_{\rm vir}^{-1}$).
• Stellar/AGN Feedback: SN II/Ia/AGB metal enrichment, black hole seeding and Bondi accretion, dual-mode AGN feedback (thermal quasar, kinetic radio).
• Magnetic Fields: Seeded at high-$z$; evolved self-consistently through ideal MHD.

The augmented suite now comprises 66 high-resolution runs: 40 MW-mass and 26 dwarf-mass halos (Grand et al., 16 Jan 2024). Data products include raw snapshots, group catalogs, merger trees, accreted/in-situ star tags, mock Gaia and PAndAS catalogs, and high-level analysis libraries.

2. Secular and Bar-Driven Evolution

Auriga galaxies spontaneously develop robust bars, pseudobulges, and realistic disc breaks (Blázquez-Calero et al., 2019, Fragkoudi et al., 12 Jun 2024). Bar identification employs Fourier decomposition of disc surface density (global $A_2\!\geq\!0.25$), with bar half-length from the drop in $A_2(R)$. Photometric decompositions model bulges (Sérsic $n\!<\!2$, always classified as pseudobulges), discs (exponential), and bars (modified Ferrer profile).

Key bar properties:

• Lengths $r_{\rm out}=3\!-\!9$ kpc (median $5.3$ kpc); strengths $f_{\rm bar}=0.4\!-\!0.8$.
• Boxy/peanut bulges identified in $\sim 30\%$ of bars ($a_X=2.03\!-\!2.45$ kpc, $b_X=1.12\!-\!1.65$ kpc).
• Barred galaxies are more baryon-dominated ($f_{\rm bd}\!\geq\!0.8$ within $5$ kpc), assemble stellar mass earlier, and show lower Toomre $Q$ at bar formation ($Q=1.0\!-\!1.2$ versus $Q\!\gtrsim\!1.3$ for unbarred) (Fragkoudi et al., 12 Jun 2024).
• The barred fraction decreases with redshift, plateauing near $20\%$ at $z\sim3$; bar lengths grow post-formation except for those formed via high-$z$ mergers.

Bar-driven secular evolution builds pseudobulges and disc breaks in full agreement with observed scaling distributions.

3. Gas and Stellar Disc Morphology and Kinematics

Star-forming and HI gas discs are resolved with high fidelity (Marinacci et al., 2016, Grand et al., 16 Jan 2024):

• HI disc radii $R_{\rm HI}=5.6\!-\!55$ kpc (median $\sim35$ kpc), systematically larger and more gas-rich than nearby observed discs (mass-diameter relation $\alpha=1.96$, $\beta=6.52$).
• HI thickness ($h_{\rm HI}$) correlates with SFR: $\log h_{\rm HI} = a \log{\rm SFR} + b$, with $a=0.30$–$0.35$, $b=0.69$–$0.78$.
• From $z\!\gtrsim\!3$, discs build $V/\sigma$ from turbulent ($V/\sigma\sim2\!-\!3$) to settled ($V/\sigma\!\gtrsim\!10$ at $z=0$), tracking observed trends from H$\alpha$ kinematics.

Stellar migration, including both churning and blurring, mixes stars in the outer cold disc; the mean migration is $\langle|\Delta R|\rangle\sim1$–$3$ kpc, with diffusion-like age and radius dependence. Bars increase migration and flatten metallicity gradients for older populations (Okalidis et al., 2022).

4. Bulge, Thick Disk, and Halo Formation Pathways

Bulges in Auriga are predominantly pseudobulges, shaped by bar-driven secular inflows and in-situ star formation (Gargiulo et al., 2019). Their Sérsic indices ($n=0.6$–$1.9$), B/T ratios, and rapid rotation place them above the classical bulge locus; accreted fractions in bulges are typically low ($f_{\rm acc}<0.1$ for $21\%$, $<0.2$ for $60\%$).

Thick disks form early (mean ages $6$–$9$ Gyr), are $\sim3$ Gyr older and $0.25$ dex more metal poor than the thin disk, and are enhanced in $\rm[Mg/Fe]$ by $\sim0.06$ dex (Pinna et al., 2023). Growth follows three channels: in-situ star formation from a turbulent epoch, dynamical heating, and accretion of ex-situ stars ($\langle f_{\rm accr,thick}\rangle\sim22\%$), with mergers playing a key role. Chemical bimodality in $\rm[Mg/Fe]$–$\rm[Fe/H]$ robustly separates geometric thick and thin components in mock IFS projections, supporting future spectroscopic diagnostics (Pinna et al., 11 Sep 2024).

Stellar halos exhibit mass, shape, metallicity, and gradient diversity set by stochastic accretion histories (Monachesi et al., 2018). Halo masses span $2\!\times\!10^9$–$2\!\times\!10^{10}$ M$_\odot$, with median metallicities $-1.3$ to $-0.3$ dex at 30 kpc, matching empirical mass–metallicity relations. Inner halo shapes are oblate, becoming prolate at large radii.

5. Dwarfs, Substructure, and Environmental Effects

Auriga reproduces field and satellite ultra-diffuse galaxies (UDGs) as high-spin tail and tidal transformation products, respectively (Liao et al., 2019). Field UDGs correlate linearly between size and halo spin parameter ($r_e\propto\lambda_{\rm halo}$), inhabit dwarf-mass halos ($M_{200}\!\lesssim\!10^{10}$ M$_\odot$), and show no evidence for a failed $L^\star$ origin.

Baryonic physics dramatically suppresses subhalo abundance near galaxy centers: $S(r)\approx0.2$ at 0.1 $r_{200}$ (80% subhalo destruction in Auriga) versus $0.5$ (50%) in APOSTLE, scaling directly with central galaxy mass (Richings et al., 2018). Velocity distributions of surviving subhalos peak higher and are narrower in Hydro runs due to steeper potential wells.

CGM diversity at $z=0$ is extreme: column densities span $3$–$4$ dex, covering fractions range $5$–$90\%$; covering fraction of H I, C IV, and Si II increases with disc fraction and anticorrelates with AGN luminosity, while parent-element covering fractions correlate with stellar mass. Neither recent SFR nor long-term mergers regulate CGM properties in the absence of major recent mergers (Hani et al., 2019).

6. Dark Matter Halo Shape, Morphology, and Global Robustness

Baryons render Auriga dark matter halos round at all radii: $(b/a,c/a)_{\rm MHD}=(0.96,0.70)$ at $14$ kpc versus $(0.55,0.43)$ in DM-only simulations (Prada et al., 2019). Triaxiality reverses from prolate (DMO) to mildly oblate with baryon inclusion. Halo–disc alignment is strong (median disc-to-minor axis misalignment $\theta_{\rm min}\sim19^\circ$), but inner-to-outer orientation twisting occurs in $\sim$20% of cases—correlating with bulge age.

Intrinsic stochasticity of Auriga simulations at fixed resolution results in $<10\%$ scatter in most global properties, $<30\%$ in local metrics, and factor of $\sim2$ in current SFR; morphology (bar presence, spiral arm pattern) and satellite disruption times are most variable (Pakmor et al., 17 Jul 2025). Resolution changes introduce systematic offsets—stellar mass, SFR, bulge/disc mass, scale radii—well beyond stochastic variability, necessitating careful ensemble averaging and recalibration for cross-resolution comparisons.

7. Data Products and Mock Surveys

Aurigaa provides mock Gaia DR2 catalogs replicating selection, astrometric/photometric errors, phase-space and stellar parameters. Stellar disc flaring and halo spin can be correctly recovered using Gaia’s observables. These catalogues enable the validation of analysis pipelines and the benchmarking of substructure and dynamical studies (Grand et al., 2018).

Summary Table: Select Auriga Simulation Properties (MW-mass runs)

Property	Value/Range	Reference
DM particle mass	$3\times10^5\,M_\odot$	(Grand et al., 16 Jan 2024)
Baryonic mass resolution	$5\times10^4\,M_\odot$	(Grand et al., 16 Jan 2024)
HI disc radii (z=0)	$5.6$–$55$ kpc (median $\sim35$ kpc)	(Marinacci et al., 2016)
SFR at z=0	$1.9$–$12.4\,M_\odot\,{\rm yr}^{-1}$	(Grand et al., 16 Jan 2024)
Bar strength (A2), z=0	$0.4$–$0.8$	(Blázquez-Calero et al., 2019)
Pseudobulge Sérsic index	$0.6$–$1.9$ (<2 for all MW analogues)	(Gargiulo et al., 2019)
CGM covering fraction (HI, C IV)	$5$–$90\%$	(Hani et al., 2019)
Subhalo suppression $S(0.1r_{200})$	$0.2$ ($80\%$ destroyed)	(Richings et al., 2018)
Halo roundness (c/a, b/a, MHD)	$0.70$, $0.96$ at $14$ kpc	(Prada et al., 2019)
Thick disk accreted fraction	mean $\sim22\%$, up to $55\%$	(Pinna et al., 2023)

Concluding Remarks

Auriga delivers a validated, high-resolution, MHD-enabled, cosmologically consistent suite of galaxy formation simulations, supporting detailed studies of secular evolution, baryonic/AGN feedback, gas dynamics, substructure suppression, stellar migration, bulge and thick disk assembly, CGM diversity, halo morphology, and providing a rich public data resource for observational and theoretical benchmarking. All major scaling relations, morphological features, and kinematic properties in MW-mass regimes are quantitatively consistent with contemporary observations, while systematics of model robustness and resolution dependence are now fully characterized (Pakmor et al., 17 Jul 2025).