|Affiliation||Institute of Agriculture|
【FOOD】 Saito Unit
Many of the parameters required in physical models used to analyze soil water dynamics are related to physical properties of soil and can be measured directly in the laboratory. However, the physical model usually cannot deal with the lack of parameters and requires a set of appropriate initial and boundary conditions. On the other hand, data-driven models that do not require parameters such as physical properties of soil have also been widely used to predict soil water dynamics. The predictive ability of data-driven models is limited by the quality and quantity of the data available. In this research, when the input parameters of the physical model are predicted from the data-driven model, the effect of the data-driven model on the result of the physical model is evaluated. Furthermore, we will consider using the results of the physical model as training data for the data-driven model, and examine how we can integrate the physical model and the data-driven model in soil water dynamics analysis.
|Affiliation||University of California, Riverside (U.S.A.)|
|Division / Department||Department of Environmental Sciences|