A re-examination of modeled and measured soil moisture spatial variability and its implications for land surface modeling

Using a spatially distributed water and energy balance model, we investigate the spatial structure of surface fluxes and states for the Washita '92 field experiment and the August campaign of the Washita '94 field experiments. For Washita '92, the model is validated against gravimetric and remotely sensed soil moisture, and for Washita '94, the model is validated against gravimetric soil moisture and measured energy fluxes. The model is shown to reasonably represent land-atmosphere interactions during the experimental periods. Scaling analysis of remotely sensed and modeled soil moisture and modeled latent heat flux is indicative of multiscaling behavior. The temporal behavior of the soil moisture scaling exponents for various moments suggests the existence of three distinct regimes during a dry-down. The multiscaling behavior inferred from simulated soil moisture and latent heat flux is hypothesized as a relationship which is a function of average soil moisture. Similar scaling analysis of important land surface properties indicates simple scaling for porosity, field capacity and wilting point, and multiscaling for residual soil moisture, leaf area index and the soils-topographic index. This is consistent with model results, which indicate a transition from simple scaling to multiscaling with dry-down. It is hypothesized that this transition is governed by the scaling properties which in wet conditions control infiltration (porosity, field capacity, leaf area index) to properties which in dry conditions control drainage (residual moisture content and soils-topographic index) and evaporation (wilting point, leaf area index). Land surface models which fail to incorporate these features will most likely be unable to capture the dynamic nature of soil moisture spatial variability. (C) 2001 Published by Elsevier Science Ltd.