Using backscatter models to define the limits of crop information recovery from SAR data

Abstract

We develop here a methodology for the comparison of ERS-2 multi-temporal backscatter values from agricultural crops with the predictions of a physically-based radiative transfer model, RT2, driven by field observations contemporaneous with the satellite overpasses. The approach makes clear the determining scatter mechanisms in the crop canopies as represented by the model, emphasising the roles of each of the generic plant components. We give a detailed description of the intensive field campaign, explaining how estimates of measurement uncertainties were derived. Using these, quantitative comparisons of the model with the ERS-2 values were possible. We draw attention to discrepancies between the two, and where possible suggest reasons for these; again these are facilitated by using the simplified core model for RT2 that is explained here.