SENIOR COMPUTATIONAL GEOSCIENTIST
Dunyu Liu is a Senior Computational Geoscientist at the University of Texas Institute for Geophysics, working at the intersection of artificial intelligence and physics-based Earth system modeling. He earned his PhD from Texas A&M University and his earlier degrees from Peking University. Liu and his colleagues recently developed a Graph Neural Network-based simulator for earthquake rupture dynamics that achieves significant computational speedup with strong generalization capability, as well as AI models to synthesize disparate oceanic drilling data. These efforts build on his core expertise in physics-based modeling of earthquake rupture and cycle dynamics, ground motion, and seismic hazard assessment using parallel finite element methods integrated with paleoseismic and geodetic observations. His research extends across climate modeling with CESM, geodynamic inversions and deformation imaging, analytical frameworks for viscous anisotropy, InSAR analysis through GMTSAR development, and planetary modeling. A strong advocate for clean, reproducible scientific software, Liu emphasizes that code development should be user-friendly, reliable, and driven by scientific questions. His multidisciplinary approach leverages high-performance computing and emerging AI techniques to advance understanding of earthquakes, climate, tectonics, and planetary processes.
To access Dunyu’s software and learn more about his research, software philosophy, book recommendations and general comments, visit him on GitHub or Substack.
INTERESTS
Earthquake dynamic ruptures and cycles, numerical modeling, parallel computing, fault geometric complexity, and ground motion simulation.
ACADEMICS
Ph.D., Geophysics, Texas A&M University
M.S., Geophysics, Peking University
B.S., Geophysics, Peking University
CONTACT AND LINKS
dliu@ig.utexas.edu
Curriculum Vitae
Publications
GitHub
EQdyna Substack
Twitter
LinkedIn