Modeling preferential flow in unsaturated soil using active region model and its sensitivity analysis

A total of 4 dye infiltration experiments, using iodine-starch staining method and Brilliant Blue dye staining method, were conducted in loam and sand with various initial and boundary conditions. The measured soil water content and solute concentration distributions resulted from preferential flow were simulated and predicted using active region model (ARM) and mobile-immobile region model (MIM), and were evaluated and compared using the relative root mean square error (RRMSE) analysis. Furthermore, sensitivity analysis was conducted to study the impacts of ARM parameter (γ) on the predictions using ARM with various infiltration conditions. The model evaluation and comparison revealed that, ARM produced more accurate infiltration depth, and soil water content and solute concentration distributions predictions; the ARM captured the macroscopic behavior of preferential flow and transport. Sensitivity analysis revealed that, the predicted infiltration depth using ARM was more sensitive to ARM parameter (γ) when the ARM parameter (γ) was higher; while the predicted infiltration depth using ARM was less sensitive to the ARM parameter (γ) under higher infiltration amount and higher initial soil water content conditions.