Assessment of soil moisture retrieval with numerical weather prediction model temperatures

摘要

The effect of using a Numerical Weather Prediction (NWP) soil temperature product instead of estimates provided by concurrent 37 GHz data on satellite-based passive microwave retrieval of soil moisture was evaluated. This was prompted by the change in system configuration of preceding multi-frequency satellites to new single frequency L-band missions. In situ soil moisture data from four watershed sites in the USA were used to assess this change with one soil moisture retrieval algorithm. The temperature product substitution resulted in a large decrease in sensitivity to in situ soil moisture changes, and illustrates the complications of moving from a coincident source to interpolation of modelled temperature.