OTI on FIRE: Testing the Efficacy of Orbital Torus Imaging to Recover the Galactic Potential

Abstract Orbital Torus Imaging (OTI) is a dynamical inference method for determining the Milky Way’s gravitational potential using stellar survey data. OTI uses gradients in stellar astrophysical quantities, such as element abundances, as functions of dynamical quantities, like orbital actions or energy, to estimate the Galactic mass distribution, assuming axisymmetry and a steady state of the system. While preliminary applications of OTI have shown promising outcomes, its sensitivity to disequilibrium effects is unknown. Here, we apply OTI to a benchmark Feedback in Realistic Environments (FIRE-2) cosmological hydrodynamic simulation, m12i, which enables a comparative analysis between known FIRE-2 vertical acceleration profiles and total surface mass densities to the analogous OTI-inferred results. We quantify OTI’s accuracy within 16 solar-analog volumes embedded in the simulated galactic disk. Despite a dynamically evolving system, we find that OTI recovers the known vertical acceleration profiles within 3 σ /1 σ errors for 94%/75% of the volumes considered. We discuss the method’s sensitivity to the local, instantaneous structure of the disk, reporting a loss in accuracy for volumes that have large (>1.5 kpc) scale heights and low total density at z = 1.1 kpc. We present realistic OTI error bars from both Markov Chain Monte Carlo sampling and bootstrapping the FIRE-2 simulated data, which provides a touchstone for interpreting results obtained from current and forthcoming surveys such as SDSS-V, Gaia, WEAVE, and 4MOST.