Derivation of Soil Moisture Patterns from a simple Soil Moisture Index

Soil moisture and its spatio-temporal pattern is one of the main drivers in complex soil-vegetation-atmosphere exchange processes. In order to observe long-term patterns of surface soil moisture, we analyzed a historical data set of ERS SAR (synthetic aperture radar) data using 85 ERS scenes from 1995-2003 for the Rur catchment (2364 km2) in Western Germany. The ERS satellites operated in C-band and single-channel VV polarization. To derive surface soil moisture from the microwave backscatter intensity, the influence of surface roughness and vegetation biomass on the backscatter must be taken into account. Thus, a simple soil moisture index was developed to retrieve semi-quantitative information about spatial soil moisture patterns with a simple yet robust approach. By using data from all available scenes for each month of the year, histograms of σ0-values for each agricultural land use class (cereals, sugar beet, pasture) were generated. Within each of these histograms, the influence of biomass and surface roughness on backscatter is assumed to be constant. Thus, changes in backscatter intensity are due to changes in surface soil moisture. Since the histograms are based on data from 8 years, we assume that each histogram contains pixels representing the wet and the dry soil moisture state. An index was spanned between high and low backscatter values, identifying wet and dry areas. By using soil texture information of the given location, the qualitative index can be translated into volumetric soil moisture. The resulting soil moisture maps were compared to precipitation data from nearby meteorological stations.