Soil moisture retrieval errors using l-band radiometry induced by the soil type variability

The soil dielectric permittivity relates the soil water content to the brightness temperature. Thus, the precise estimation of the dielectric permittivity will lead to a better soil moisture retrieval. This paper presents a study which analyses the errors induced in the soil moisture retrieval due to the variability associated to soil type. Data sets over six different soil types were acquired during the MOUSE 2004 experiment, performed at the Joint Research Centre (JRC) outdoor test facility, Ispra, Italy. The radiometric data were processed using both laboratory measurements of soil dielectric permittivity and those computed using other semi-empirical models available in the literature. The soil texture was measured by the JRC, and soil height profiles were acquired by a laser profiler. Soil moisture retrievals using standard models lead to retrieval errors of 5.6% and 6.2%, while when the measured dielectric constant was used the errors were of the order of 5.1%. Errors associated to the modelling of the roughness effect and those due to interference signals are believed to be responsible of some scattering in the retrieved data.