Impact of filtering soil roughness low frequencies on the radar backscattering coefficient simulated by the IEM model

摘要

The low frequency component of the soil surface roughness profile (LF) has little impact on microwave backscattering from soils. The presence of LF makes the estimation of the correlation length difficult and so numerous and long profiles are required leading to tedious sampling efforts. In this study we proposed to filter the LF by limiting the computation of the autocorrelation function (ACF) parameters to the beginning of the relationship i.e. the short distances. It is demonstrated that the obtained correlation length by the proposed method is smaller and less variable than considering the first 1/e value. When introducing this value in the IEM model to compute the microwave backscattering coefficient we obtained better agreement with experimental values.