Selecting ultra-faint dwarf candidate progenitors in cosmological N-body simulations at high redshifts

The smallest satellites of the Milky Way ceased forming stars during the epoch of reionization and thus provide archaeological access to galaxy formation at $z>6$. Numerical studies of these ultra-faint dwarf galaxies (UFDs) require expensive cosmological simulations with high mass resolution that are carried out down to $z=0$. However, if we are able to statistically identify UFD host progenitors at high redshifts \emph{with relatively high probabilities}, we can avoid this high computational cost. To find such candidates, we analyze the merger trees of Milky Way type halos from the high-resolution ${\it Caterpillar}$ suite of dark matter only simulations. Satellite UFD hosts at $z=0 $ are identified based on four different abundance matching (AM) techniques. All the halos at high redshifts are traced forward in time in order to compute the probability of surviving as satellite UFDs today. Our results show that selecting potential UFD progenitors based solely on their mass at z=12 (8) results in a 10\% (20\%) chance of obtaining a surviving UFD at $z=0$ in three of the AM techniques we adopted. We find that the progenitors of surviving satellite UFDs have lower virial ratios ($\eta$), and are preferentially located at large distances from the main MW progenitor, while they show no correlation with concentration parameter. Halos with favorable locations and virial ratios are $\approx 3$ times more likely to survive as satellite UFD candidates at $z=0.$