Retrieval of bare surface soil moisture from AMSR-E data

Passive microwave plays an important role in soil moisture monitoring. In this study, a layered soil emission model has been built based on theory of radiative transfer, combined with the random rough surface, soil dielectric constant model and soil moisture algorithm. The soil moisture in different frequencies was simulated and the look-up table set was established between soil moisture and soil brightness temperature in the taklimakan desert study area. The look-up table set was used to retrieve the soil moisture from the study area´s brightness temperature observed by AMSR-E. The results of soil moisture retrieval were compared to the AMSR-E soil moisture data product and their rms errors were calculated for validation. The validation results suggest that there is a good agreement between retrieved and AMSR-E soil moisture in study area at certain frequencies, and the soil moisture has been retrieved to the highest accuracy at 6.9 GHz, with H poarization. In conclusion, the soil moisture in this study area can be well reflected by the retrieval results.