Dr. Christoph Keller’s research is focused on improving the representation of reactive trace gases in atmospheric models. He uses the GEOS-Chem chemistry module within the GEOS Earth System model to simulate and better understand the evolution of air pollutants such as ozone and nitrogen dioxide. He uses atmospheric observations to further improve model constraints, in particular emissions. He also develops new methods to emulate processes relevant to atmospheric chemistry based on machine learning. Keller is the principal developer of the GEOS composition forecast (GEOS-CF) model.