Research Division Seminar
Pushing the frontiers of galaxy evolution modeling with multi-scale simulations and machine learning

Abstract

Supermassive black holes (SMBHs) in Active Galactic Nuclei (AGN) play a key role in the formation of galaxies and large-scale structure, but the triggering and impact of AGN feedback across scales and the origin of the observed SMBH–galaxy connection remain major open questions owing to the multi-scale and multi-physics nature of the problem. AGN feedback can also profoundly affect the properties and spatial distribution of baryons on scales that contain a large amount of cosmological information.  Current and upcoming cosmological surveys will provide unprecedented data to constrain the fundamental cosmological parameters, but uncertainties in galaxy formation physics remain a major theoretical obstacle to extract information from cosmological experiments.  In this talk, I will present new simulation techniques that are pushing the frontiers of galaxy formation modeling towards (1) the smallest scales, developing physically predictive models of SMBH accretion and feedback explicitly at sub-pc resolution in a full cosmological context and (2) the largest scales, using thousands of large-volume simulations exploring a wide range of sub-grid feedback implementations to train machine learning algorithms that can maximize the extraction of information from cosmological surveys while marginalizing over uncertainties in galaxy formation physics.  I will demonstrate the feasibility of these orthogonal approaches to address fundamental problems and discuss their potential to advance the fields of galaxy evolution and cosmology.

Zoom link: https://rediris.zoom.us/j/95949230133?pwd=xxXArEDCwNg4iXt4f5vUiCGvUFC9ph.1