Applications of Machine Learning (ML) in Earth and soil science are rapidly expanding to areas such as data mining, solution of nonlinear partial-differential equations, parameterization, data assimilation, scaling and guiding data-driven or process-based models. This unprecedented growth creates numerous new opportunities for advancing physical understanding, improving Earth system modeling, and enhancing predictive capabilities of Earth surface models across scales. This session provides a forum for scientists to exchange ideas on topics such as, but not limited to the following (1) innovative data analytics approach in soil modeling, (2) physics-informed ML for solving nonlinear partial-differential equations, (3) using ML and other innovative approaches in predicting soil hydraulic properties, transport parameters, thermal parameters, biogeochemical parameters, etc., (4) ML coupled with field data and numerical data, and (5) advances in theoretical and applied studies in soil modeling along with their predictability and uncertainty.

(This session is supported by the International Soil Modeling Consortium)

• SSSA: Soil Science Society of America -