I am an associate research scholar in Department of Civil and Environmental Engineering at Princeton University. I am in transition to School of Atmospheric Sciences at Sun Yat-sen University as an associate professor.

Current Research

Multi-GPU, CUDA-Aware acceleration of particle tracking on distributed platforms

Large-scale hydrologic modeling: National hydrologic modeling platform of China

Data-driven approach in hydrology: Neural particle tracking using Neural ODE

Coupling between ParFlow and Land Surface Models

• Coupling with NASA LIS, ongoing project in US
  • LIS is a flexible land surface modeling and data assimilation framework
• Coupling with Common Land Model (CoLM), future collaboration in China

Applications of integrated hydrologic models for scientific discoveries

• Topography-controls on spatiotemporal variations of ET age
  • Particle tracking in the North China Plain
• Connections between groundwater and streamflow (in revision)
  • Backward partilce tracking in the continental US

Performance of our modeling platform

Our modeling platform showed great performance not only in the ParFlow hydrologic modeling, but also in the EcoSLIM particle tracking.

Simulated groundwater ages compare favorably to that reported in literatures and expand these previous understandings by a more complete age distribution of all aquifers and wider simulated ranges of individual aquifers. (a) shows locations of the 21 principal aquifers studied in Jurgens et al. (2022). Simulated groundwater age distribution (b) agrees with that based on means of 1279 samples from Jurgens et al. (2022) (measured) at peak ages. The latter cannot capture the distribution of groundwater age at large values due to the limited number of samples. The average age of 14 macroscale basins summarized in Stewart et al. (2010) is also indicated by the shaded grey line in (b). Peak ages of three different data sources match each other well. (c) presents the age range of each aquifer bracketed by minimum age and maximum age with the mean value indicated. Lines and bars are for simulated and Jurgens et al. (2022), respectively. Simulated means compare well with that reported in Jurgens et al. (2022) and simulated ranges bracket the reported ranges showing the expanded understandings.