Effect of Land-Cover Type on the SMAP Active/Passive Soil Moisture Downscaling Algorithm Performance

A brightness temperature (Tb) downscaling algorithm based on the synergy between active and passive microwave observations is tested using airborne data that simulate the Soil Moisture Active Passive (SMAP) mission of the National Aeronautics and Space Administration scheduled for launch in January 2015. While this algorithm has been adopted as the baseline for SMAP, it has only been tested on a limited variety of land uses and vegetation types. Consequently, this study evaluates the SMAP active/passive downscaling algorithm using data with varied conditions. The SMAP experiment conducted in Australia has been used for this purpose. The algorithm was applied over several 9 km × 9 km pixels with different land covers, so as to evaluate the accuracy of this algorithm under different heterogeneity levels. Brightness temperatures were downscaled from 9 to 3 km (approximating the resolution ratio of SMAP downscaling approach) across nine days of data. Results show that the root-mean-square error of Tb in grassland could meet the 2.4-K target accuracy of SMAP, while in cropping, it was 2 K higher than the target. The influence from water bodies was also assessed and confirmed to have a significant impact if not removed prior to downscaling.