Multiscale Soil-moisture Retrieval to Monitor the Global Change of the Water Cycle Perspectives from the Glowa-danube Project

The GLOWA-initiative, funded by the German Ministry of Research and Education (BMBF), has been established to address the consequences of Global Change, which is understood as affecting future development far beyond Climate Change and to include changes in land use, population and development, on regional water resources. The GLOWA-Danube project (www.GLOWA-Danube.de) is dealing with the Upper Danube watershed as a representative mesoscale test site (~ 75.000 km2) for mountain-foreland regions in the temperate mid-latitudes. The principle objective is to develop an integrated Decision Support System, called DANUBIA, to investigate the sustainability of future water use scenarios for interested stakeholders in the watershed. DANUBIA is object-oriented, spatially distributed and raster-based on a 1 km2 grid spacing. Remote Sensing and GIS form the basis of the integrative monitoring concept for GLOWA-Danube. Multitemporal and multisensoral (and hence multiscale) remote sensing information is used as a source of input and validation data for the SVAT model PROMET, which is part of DANUBIA. Hydrologic processes, such as runoff production or evapotranspiration, largely depend on the variation of soil moisture and its spatial pattern. Envisat’s ASAR-Wide Swath Mode (WSM) imagery for the first time synoptically covers mesoscale areas like the Upper Danube Basin in medium spatial resolution and high temporal repetition. At the same time Envisats Alternate Polarisation Mode (APS) data with its high spatial resolution offers the opportunity to distinguish between vegetation and soil effects on a scale where satellite measurements can directly be compared with field mesurements. The study presents the results of the application of a semi-empirical approach to retrieve near-surface soil moisture from ASAR-APS (30 m/pixel) and ASAR-WSM (150 m to 25 km/pixel) imagery. The procedure compensates for the different topographic and vegetation effects on the backscatter signal and also includes procedures to account for the look-angle effects in the WSM-data and the lack of soil-moisture information on different land uses like forest and settlements. Quantitative comparison of derived near-surface soil moisture maps from APS-data to field measurements shows good agreement. In order to derive surface soil moisture patterns for the mesoscale Upper Danube Basin WSM-data is processed and soil moisture is determined on a spatial resolution of 1 km and 25 km by only taking into account the backscatter of those land-uses, which contain soil-moisture information. For this purpose subscale land cover fraction data derived from AVHRR imagery is used . Retrieved mesoscale soil moisture statistics from WSM-data at a resolution of 1km are in good agreement with soil moisture retrieved from high spatial resultion APS data taken on the same day. Both the surface soil-moisture retrieved from WSMdata on the 1 and 25 km resolution are compared with model results of DANUBIA. For this purpose DANUBIA is run hourly for the dry year 2003 to account for memory effects of the soil system. The modelled and retrieved soil moisture patterns and magnitudes compare well with several distinct differences between model and measurements. They point at weaknesses of the current model parameterization, which can be overcome by using ASAR data in the future. The continuous model run allows the demonstration of the spatial heterogeneity and dynamics of mesoscale soil moisture pattern, which can be expected from SMOS-observations _____________________________________________________ Proc. of the 2004 Envisat & ERS Symposium, Salzburg, Austria 6-10 September 2004 (ESA SP-572, April 2005)