Physical soil moisture diffusion modeling through sensors and numerical simulation.

Abstract

This work presents a methodology for the identification of soil hydraulic parameters based on the comparison between numerical simulations of the Richards equation and experimental soil moisture measurements obtained using sensors installed at depths of 10 cm, 20 cm, 40 cm, and 60 cm. Soil hydraulic behavior is modeled using the Van Genuchten model, whose parameters are commonly estimated from soil textural compositions through the Rosetta database. To improve model accuracy, a one-dimensional numerical model with spatial and temporal discretization is implemented to simulate the evolution of soil moisture along the soil profile. The proposed methodology evaluates a multimodel set corresponding to different combinations of sand, silt, and clay contents, selecting the model that minimizes the RMSE error with respect to the experimental measurements. Results obtained using real data from a citrus crop field demonstrate the capability of the proposed method to identify hydraulic parameters able to adequately reproduce the observed soil moisture dynamics.