Dr. Pin-Ching Li possesses extensive expertise in hydrologic and hydraulic modeling, machine learning (ML), geotechnical engineering, and remote sensing. He has developed data-driven models from available operational and monitoring data using Transfer Function-Noise (TFN) and ML techniques, such as Long Short-term Memory (LSTM) models, for simulating and predicting groundwater level response to groundwater extraction. In addition, he is engaged in developing conceptual and numerical models using MODFLOW USG and TOUGH families of codes to simulate groundwater flow as well as subsurface fate and transport of chemicals. He is working on tomographic analysis using MODFLOW USG with matrix diffusion to delineate the spatial distribution of flow and transport parameters.
Dr. Li is proficient in analyzing and simulating land subsidence induced by decline and fluctuations in groundwater levels. He developed numerical models to predict future land subsidence by considering geomechanics and geohydrology. In addition, he has analyzed the carbon sequestration using the ToughReact model and determined the potential AoR (Area of Review) of the CO2 plume. Furthermore, he applied ML techniques with Principal Component Analysis (PCA) for classifying subsurface materials and analyzed physical parameters from remote sensing data using ML models, such as Convolution Neural Network (CNN) and Generative Adversarial Networks (GANs).