Identification of van Genuchten Model Fitting Parameters for Soil Water Characteristic Curve

Abstract

When performing unsaturated seepage flow analysis for unsaturated layers in the ground, a soil water characteristic curve is required. The van Genuchten model is often used to generate this curve, and its parameters are generally identified using gradient methods such as the Gauss–Newton method (G–N method). However, the G–N method is prone to local suboptimal solutions, so in recent years, attempts have been made to apply a random search method using a genetic algorithm (GA method). Here, the van Genuchten model fitting parameters were identified by the G–N and GA methods, based on the results of water retention tests for four soil types, and the identification accuracy of the two methods was compared. As a result, it was confirmed that by setting reasonable initial values, the van Genuchten model constants can be identified with sufficient accuracy using the G–N method, which has a lighter computational load.