Modelling the spatial variability of soil moisture in a micro-scale catchment and comparison with field data using geostatistics

The temporal and spatial variability of soil moisture in the micro-scale Berrensiefen catchment is simulated with SWMS_3D (three-dimensional variably saturated flow based on finite elements) and a loosely coupled model for the atmospheric boundary conditions. An additional model component for macropore and surface runoff and their concentration time is introduced. Time step of precipitation, atmospheric parameters and the simulated hydrologic processes is one hour. Soil parameters are discretized with a horizontal resolution of 10 m and five nodal layers in the vertical direction. A high resolution digital elevation model is used to reconstruct catchment surface morphology. For a nearly 2-month period hydrologic processes are simulated. The process model results are compared with measurements and the influence of the topography, the distributed soil and land use parameters on the soil water fluxes is examined. The spatial structure of soil moisture during a drying period is analysed using geostatistics. The application of standard geostatistical techniques (variogram modelling) captures the main features of spatial variability of the modelled soil moisture distributions. The integral scale or correlation length of soil moisture is constant while the semivariance shows clear correlation with soil moisture status during that period. A geostatistical comparison of the modelled soil moisture patterns of the whole catchment with a measured data set of a single hillslope of the catchment addresses the problem of differing scales of modelling and measurements. It is shown that due to the different spatial scales the results of point measurements of soil moisture can hardly be used to validate the modelled spatial structure of soil moisture. (C) 2003 Elsevier Science Ltd. All rights reserved.