A Five-Year Evaluation of SMOS Level 2 Soil Moisture in the Corn Belt of the United States

Abstract

We compare Soil Moisture Ocean Salinity Level 2 Soil Moisture (L2SM) retrievals to the five-year in situ soil moisture dataset available for the predominately agricultural South Fork Iowa River (SFIR) watershed in the U. S. Corn Belt. SMOS L2SM is 0.039 m $^3$ m $^{-3}$ drier than the SFIR network soil moisture and has an unbiased RMSE (ubRMSE) of 0.062 m $^3$ m $^{-3}$ for the period of April 2013 to November 2017 (excluding DJF). The bias is 11% of the range of in situ soil moisture. The largest monthly dry biases occur in April (0.052 m $^3$ m $^{-3}$ ), July (0.072 m $^3$ m $^{-3}$ ), and November (0.061 m $^3$ m $^{-3}$ ). Potential sources of the dry bias we discuss are: bias in auxiliary modeled temperatures; errors in soil texture maps; and nonrepresentative parameterizations of single scattering albedo and soil surface roughness. Auxiliary skin temperature was colder than expected and may explain why SMOS L2SM has a slightly drier bias for evening overpasses. Increasing the parameterized soil surface roughness produces wetter SMOS L2SM retrievals but also decreases the range of SMOS L2SM. Random error in auxiliary surface temperature, edge-of-field locations of in situ sensors, and differences in sensing volume between SMOS and in situ sensors contribute to the soil moisture ubRMSE. ubRMSE can be decreased by using a nonzero single scattering albedo more representative of a corn and soybean canopy at the cost of increasing the dry bias.