Measuring Surface Roughness on Base of the Circular Polarization Coherence as an Input for a Simple Inversion of the IEM Model
In this paper a simple inversion approach is applied to the integral equation method model. The inversion of the radar backscatter into soil moisture was accomplished with an iterative method. To date most inversion approaches with the IEM try to determine both the surface roughness and the soil moisture on base of two copolarized radar channels. Because there are a number of possible combinations of roughness and moisture parameters to reach the same values for the radar backscatter, these approaches always contain a fragment of uncertainty and inaccurateness. Therefore this work follows the idea of arising the number of input variables to reduce the inaccurateness and ambiguity of the inversion results. For this reason the surface roughness parameters are declared as input variables and were calculated external of the IEM. The base for computing the RMS values is the circular polarization coherence of the copolarized channels. The real part of this coherence was found to be a good measure for this surface roughness parameter. For agricultural surfaces the correlation length of the autocorrelation function was found to correlate with the RMS heights. Because of the fact, that the autocorrelation length is not the most deciding parameter, this contiguity was used to calculate the autocorrelation length from the RMS heights. The final result in the form of a soil moisture map seems quite promising. Good correlations between measured and modelled values were realised. The arrangement of the soil moisture outside the test areas are rather real with higher values in depressed areas and lower values on hilltops.