Characterizing subpixel variability of low resolution radiometer derived soil moisture using high resolution radar data

[1] Soil moisture estimates obtained using passive remote sensing from satellite platforms often suffer from the drawback of coarse spatial resolution. In this current work, low resolution soil moisture estimates from passive remote sensing are fused with high resolution radar backscatter data to produce soil moisture change estimates at the spatial resolution of radar. More specifically, soil moisture estimated from AMSR-E and TMI (separate cases) for a single 50 km × 50 km pixel has been fused with TRMM-PR backscatter data at 5 km resolution to produce soil moisture change estimates at 5 km resolution. A brief sensitivity analysis has been presented as a baseline study for soil moisture sensitivity of TRMM-PR backscatter. Soil moisture change estimates have been computed using a simple methodology and validated using in situ measurements from the Little Washita Micronet. It is seen that fusing radar data with radiometer soil moisture estimates leads to a better representation of the soil moisture variability within the radiometer pixel as compared to the baseline (radiometer estimate only) case where uniform subpixel distribution of soil moisture is assumed. The TMI/PR case performs better than the AMSR-E/PR case indicating the need for temporally coincident radar radiometer observations for producing high resolution soil moisture change estimates.