ES-03-0008

Analysis of JWST brown dwarf data using self-consistent models enabled by machine learning

Francisco Ardevol Martinez, Inga Kamp, Michiel Min, Paul I. Palmer

Interpreting observations of exoplanet and brown dwarf atmospheres is a complex and computationally expensive problem, typically requiring hundreds of thousands to millions of forward model evaluations. Here we present a machine learning approach based on normalising flows that is able to significantly reduce the number of forward models needed. We apply this novel method to the recent JWST observations of the brown dwarf WISE-J1828.