To interpret DESI scheduled for the summer of 2023, this team will produce a set of high-resolution hydrodynamical simulations which are needed to accurately compute the large-scale clustering of the intergalactic medium within the cosmic web while capturing small-scale effects due to pressure broadening of gas or a finite mass of dark matter particles leading to free-streaming, at sufficient precision.
DOE-supported sky surveys such as Dark Energy Spectroscopic Instrument (DESI) and Rubin Observatory (a.k.a. LSST) require sophisticated hydrodynamical simulations in order to interpret the data and produce realistic mock skies on which working groups can test different methodologies. Combining in this proposal two exascale-ready cosmology codes – Nyx and HACC – allows us to model a wide range of redshifts and cosmological observables, from the epoch of reionization and vast regions of the intergalactic medium, to small-scale physics relevant for modern-day galaxies.
The ongoing DESI survey will observe 0.7 million quasars at z>2, more than three times as much as collected with previous surveys. This exquisite dataset will strongly improve several cosmological constraints, e.g., neutrino mass, the nature of dark matter, and the epoch of inflation. To interpret DESI, especially focusing on the DESI Year 1 Data Release of the Lyman alpha forest, scheduled for the summer of 2023, we will produce a set of high-resolution hydrodynamical simulations which are needed to accurately compute the large-scale clustering of the intergalactic medium within the cosmic web while capturing small-scale effects due to pressure broadening of gas or a finite mass of dark matter particles leading to free-streaming, at sufficient precision. At lower redshifts, the interpretation of the galaxy clustering and weak lensing signal from data collected by the LSST will be limited if we do not understand the effects of baryonic physics on small scales. Sophisticated hydrodynamical simulations which will be run under this project will include different subgrid/feedback models and will provide urgently needed predictions to fully exploit new information present in the LSST data.