Spatial variability in the water balance of an experimental catchment

Abstract

Data from a 5 ha catchment were used to assess systematic spatial variability in the soil water content. The importance of incorporating this variability into hydrologic models was evaluated. The catchment was subdivided into three zones on the basis of slope and topographic position to reflect variations in the gain, loss and transfer of soil water. Temporal patterns of soil moisture behaviour of these zones were shown to be statistically different using a detailed network of sampling sites by the neutron moisture meter. This subdivision was justified because computed values of catchment vapour loss which incorporated the variable influence of soil water were better than point measurements using a lysimeter for the catchment scale. The respective performances of two variants of a process model, lumped and cascaded, were compared with measured catchment discharge. Both methods were comparable during wet periods for monthly yields of runoff. For dry periods and for weekly time scales, the cascaded variant was superior to the lumped version of the process model.