Relating satellite gravimetry data to global soil moisture products via data harmonization and correlation analysis

Due to limited in-situ data global soil moisture products should also be validated with respect to independent global data sets. Our study investigates possibilities and benefits of relating soil moisture products from remote sensing and hydrological modeling to information on total water storage change from satellite gravimetry. We use soil moisture data from the active satellite sensor ASCAT and the hydrological model WGHM as well as satellite gravity field observations from the GRACE mission. First we apply a data harmonization procedure to equalize the distinct data representations and formats of those data sets. Then we perform a correlation analysis. The results show correlations close to one between GRACE and soil moisture data specifically for humid and temperate regions. A comparison of correlation coefficients from different data pairs highlights that in arid environments total water storage from GRACE corresponds better to surface soil moisture captured by ASCAT than to total soil moisture from WGHM. In humid and temperate regimes the observation is reversed. Furthermore regions could be identified where the input data of the WGHM might be of low quality, producing higher correlations between ASCAT and GRACE than between ASCAT and WGHM. We therefore conclude that GRACE data can deliver valuable information for the quality assessment of soil moisture products and provide a link to their contribution to continental water storage. (C) 2013 Elsevier Inc. All rights reserved.