Galactic Accelerations from the GD-1 Stream Suggest a Tilted Dark Matter Halo

Abstract: Cold dark matter halos are expected to be triaxial and often tilted relative to the stellar disk. Stellar streams provide a sensitive tracer of the Milky Way's halo shape, though models for the Galactic potential are typically limited to simple, symmetric functional forms. Here, we measure the Galactic acceleration field along the GD-1 stellar stream using a direct differentiation of the stream's track in phase-space. Using a fully data-driven catalog of stream members from Gaia, SDSS, LAMOST, and DESI, we map the stream in 6D phase-space. We fit splines to the stream track, and infer cylindrical acceleration components $a_R = -2.5 \pm_{0.1}^{0.2}, \ a_z = -1.8\pm 0.1, \ a_\phi = 0.2\pm 0.1~\rm{km \ s^{-1} \ Myr^{-1}}$ at $(R,z,\phi) = (11.9~\rm{kpc}, 7.3~\rm{kpc}, 171.1~\rm{deg})$. We measure mass enclosed within $14~\rm{kpc}$ of $1.4\pm 0.1 \times 10^{11} M_\odot$ and z-axis density flattening of $q_{\rho, z} = 0.81\pm^{0.06}_{0.03}$, both consistent with previous estimates. However, we find a 2$\sigma$ deviation from an axisymmetric acceleration field, which can be explained by a triaxial dark matter halo with axis ratios 1:0.75:0.70. The major axis of the halo is consistent with a tilt of $18~\rm{deg}$ above the Galactic plane in the direction of the Sun. The magnitude and direction of the tilt are consistent with measurements of the Milky Way's stellar halo from Gaia and the H3 survey. A tilted triaxial halo has important consequences for orbit-integration-based studies of the Galaxy, and can be further tested by deriving acceleration constraints from multiple streams.