Sensitivity analysis of land and water productivities predicted with an empirical and a process-based root water uptake function

Abstract

Rootzone hydraulic conditions govern root water uptake and transpiration under drought stress. Process-based approaches to predict the soil water status are advocated for an improved simulation of soil hydrology and crop yield. We investigated the sensitivity to system parameters in root water uptake simulation using a process- based function (MFlux) and an empirical function (Feddes) embedded in the SWAP hydrological model, applied to irrigated pasture scenarios in New Zealand. Data from two locations and three soils were used to simulate 42 growing seasons. The sensitivity analysis of both Feddes and MFlux parameters was performed for a rainfed and two irrigated scenarios, one triggering irrigation based on relative evapotranspiration (I–ETr), the other based on common practice using total available water (I–TAW). Results confirm that some parameters of the MFlux function are more sensitive than those from the Feddes function and both functions support the I–ETr criterion to optimize the water use in grazed pastures in New Zealand.