Research

My research interest is using combined physical (process-based) and statistical (machine learning) models to simulate complex systems, with a focus currently on earth and climate science. Especially, how complex physical models can be simplified for increased efficiency in estimation and prediction, how to quantify and improve the robustness and trustworthiness of inferred process understanding from model data fusion given the scarcity and uncertainty of observation data (a ubiquitous issue in earth and climate science), and how process-based and machine learning based models can be combined to reliably and efficiently simulate complex systems constrained by observation data of varying uncertainties.