N-body Cosmology with Abacus
Abstract
Interpreting galaxy surveys in a cosmological context requires an accurate forward model of large-scale structure. N-body simulations are the standard tool for this but are not without their drawbacks. For one, they are computationally expensive, requiring 10s of thousands of GPU node-hours for large simulations; for another, they are only as accurate as their discrete ``macroparticle'' representation of dark matter allows. In my dissertation, I address both of these challenges. Abacus is a code for cosmological N-body simulations based on an exact decomposition of the near-field and far-field force, making it exceptionally accurate and fast. Using one dual-GPU node, Abacus can solve a supercomputer-sized N-body problem in a fraction of the node-hours of other codes while retaining significantly higher force accuracy. This accuracy has allowed us to investigate and correct discreteness effects that arise from the macroparticle representation of dark matter. Using modified initial conditions, we can suppress discreteness errors in the small-scale late-time matter power spectrum by an order of magnitude. Additionally, a suite of halo catalogs and particle data from 165 Abacus N-body simulations spanning 40 wCDM cosmologies is publicly available at https://lgarrison.github.io/AbacusCosmos/ .
- Publication:
-
American Astronomical Society Meeting Abstracts #233
- Pub Date:
- January 2019
- Bibcode:
- 2019AAS...23342803G