A Physics-Informed Neural Network Approach to the Richards Equation and Soil-Process Investigation
This paper presents a physics-informed neural network (PINN) approach for solving the Richards equation, which governs transient moisture transport in variably saturated soils.The model is trained on synthetic datasets generated using the porousMultiphaseFoam library based on OpenFOAM.The study investigates the ability of PINNs to accurately reproduce one-and two-dimensional infiltration dynamics and predict moisture content profiles over time.
Authors
- Ivan Maltsev (ORCID: https://orcid.org/0000-0002-0262-3389)
- Anna Varlamova
- Sergei Strijhak (ORCID: https://orcid.org/0000-0001-5525-5180)
- Andrei Kulinskii
Institutions
- Institute of Social and Economic Studies of Population (RU)
Publication Details
- Published
- 2025-09-21
- DOI
- https://doi.org/10.51408/csit2025_65
- Primary Topic
- Geophysical and Geoelectrical Methods
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Ministry of Science and Higher Education of the Russian Federation