Effective versus measured correlation length for radar-based surface soil moisture retrieval

At present, radar-based surface soil moisture (SM) retrieval is hampered by the influence of surface roughness on the backscattering coefficient (0). Surface roughness is typically represented by two parameters, namely the standard deviation of surface heights (s) and the surface correlation length (l). The latter is a very problematic parameter, since it is extremely variable and very difficult to measure adequately. Therefore, several authors proposed calibrating it using backscattering models yielding optimum or effective l values. Baghdadi et al. found that those effective l were related to the parameter s and the configuration of the sensor, and proposed an approach to calculate it. The objective of this study is to evaluate the validity of that approach using data acquired on a complementary test site. RADARSAT-1 scenes acquired over an experimental watershed are used. Soil moisture and surface roughness parameters were measured in detail, coinciding with satellite overpasses. The effective l values calculated from the equations of Baghdadi et al. (2006) are used to perform forward and inverse simulations using the Integral Equation Model that are compared with radar observations and ground measurements of SM. The results obtained highlight the potential of the evaluated approach towards an operational radar based soil moisture estimation.