THE GALAXY-DARK MATTER CONNECTION WITH HST, JWST AND EUCLID – Part II

Groundbreaking optical/IR observations of the first galaxies born shortly after the Big Bang are providing new information about the amount of star light in early-Universe galaxies. New radio observations of neutral hydrogen gas surrounding the galaxies reveal how star light ionizes the gas. To convert this into knowledge about the physical processes involved, we must fit physical models to the data.

We will develop, test and apply new physics-informed machine learning methods to construct highly efficient and flexible emulator-like computational models of early-Universe galaxy formation and hydrogen gas ionization (reionization).

This will realize a new regime of predictive  and physical model testing capability needed for tractably handling large and precise near-future early-Universe data sets, improving a factor 1000+ in computational efficiency.This will allow us to produce some of the strongest, most robust constraints on dark matter physics and early-Universe galaxy formation to date using data from telescopes like the Hubble and James Webb Space Telescopes, EUCLID, the Low-Frequency Array (LOFAR) and the Hydrogen Epoch of Reionization Array (HERA).

We will also forecast model testing prospects for the Square Kilometre Array in synergy with EUCLID and the Roman Space Telescope, and establish a valuable set of useful tools for the fast-emerging fields of early-Universe galaxy formation, structure formation cosmology beyond the standard model, and reionization.