Estimation of Border-Strip Soil Hydraulic Parameters

Abstract

The inverse problem of determining soil hydraulic parameters (saturated hydraulic conductivity and water retention parameters) of border-strip irrigation from irrigation event data is analyzed. The inverse problem is solved using sequential unconstrained minimization technique. The forward problem involves the solution of coupled Saint-Venant’s equation governing overland flow and Richard’s equation governing subsurface flow. Saint-Venant’s equations are solved using the MacCormack scheme–based finite-difference method while Richard’s equation is solved using a mass conservative fully implicit finite-difference method. Field experiments are conducted on two border strips to obtain surface and subsurface irrigation data such as irrigation advance, recession, flow depth, and soil moisture content. The soil hydraulic parameters, i.e., saturated hydraulic conductivity and soil retention parameters, are estimated by minimizing the deviations between the model-predicted and field-observed irrigation data. The results indicate that defining the objective function in terms of flow depths results in the optimization algorithm converging to the true values as compared to the use of irrigation advance data. Further, it is observed that underestimating the initial guess results in the least number of iterations for the optimization algorithm to converge to the true values. It is also observed that simultaneous estimation of all three soil hydraulic parameters is not possible even with the inclusion of subsurface moisture content data in the objective function.