Title: Orbital distribution of infalling satellite halos across cosmic time

Abstract: The initial orbits of infalling subhalos largely determine the subsequent evolution of the subhalos and satellite galaxies therein and shed light on the assembly of their hosts. Using a large set of cosmological simulations of various resolutions, we quantify the orbital distribution of subhalos at infall time and its mass and redshift dependence in a large dynamic range. We further provide a unified and accurate model validated across cosmic time, which can serve as the initial condition for semi-analytic models. We find that the infall velocity $v$ follows a nearly universal distribution peaked near the host virial velocity $V_{\mathrm{h}}$ for any subhalo mass or redshift, while the infall orbit is most radially biased when $v\sim V_{\mathrm{h}}$. Moreover, subhalos that have a higher host mass or a higher sub-to-host ratio tend to move along a more radial direction with a relatively smaller angular momentum than their low host mass or low sub-to-host ratio counterparts, though they share the same normalized orbital energy. These relations are nearly independent of the redshift when using the density peak height as the proxy for host halo mass. The above trends are consistent with the scenario where the dynamical environment is relatively colder for more massive structures because their own gravity is more likely to dominate the local potentials. Based on this understanding, the more massive or isolated halos are expected to have higher velocity anisotropy.