Huan Lei

My research draws inspiration from a variety of scientific and engineering questions relevant to fluid dynamics, chemistry, material sciences, and climate physics. In particular I am interested in developing high-fidelity computational models for multiscale multiphysics systems that no longer fit into traditional hypothesis-driven based descriptions. Current research focuses on the development of numerical methods for integrating the model reduction and projection theory with data-learning algorithms to probe low-dimensional evolution dynamics relevant to diffusion, transport, transition, etc. The major object is to construct data-driven models encoded with nonlocal correlations and fluctuations arising from interactions on smaller scales juxtapose with proper physical constraints. The particular areas of application includes nanoscale and mesoscale hydrodynamics, coarse-grained molecular dynamics, material synthesis,  biomolecule modeling and climate modeling.