The Durability of Subhalos and the Fragility of Subhalo Finders - Phil Mansfield (KIPAC)

Abstract: Every dark matter halo is surrounded by an orbiting swarm of rapidly disintegrating, small "subhalos." Our ability to make theoretical predictions for subhalos impacts a vast swath of science, ranging from core cosmological probes like large-scale clustering and satellite abundances, to bread-and-butter astrophysics like the galaxy-halo connection. However, despite their scientific importance, there are substantial short-comings in our ability to predict the properties of subhalos. There are strong first-principles to suspect that even our best simulations have trouble simulating subhalos, and different tools disagree on some of the most important summary statistics of subhalo populations. In this talk, I present a new, quantitative system for evaluating how biased our subhalo predictions are, and use it to assess the reliability of different simulations and subhalo-finding tools. I show that simulations do exhibit substantial numerical biases, as do some of the most popular subhalo-finding tools. I present a new tool which outperforms these tools by orders of magnitude and show that it is able to track subhalos to the point of galaxy disruption without requiring orphan modeling.