Active polarimetric microwave remote sensing of corn and multiparametric retrieval of corn parameters with an artificial neural network

Abstract

AbstractPolarimetric signatures of a corn canopy are studied with the vector radiative transfer theory. Comparisons are made with experimental data at L and C bands. Multiparametric inversions of corn canopy parameters are performed with an artificial neural network (ANN) trained with vector radiative transfer theory. We have performed simultaneous retrieval of three parameters: corn height, volumetric moisture of corn stalks, and volumetric soil moisture from a total of eight averaged Mueller matrix elements at L and C bands and at one incidence angle. It is shown that the performance of the neural network is good, with errors of less than 10%. It is also shown that the performance of the ANN is better if measurements at both frequencies are used and also if polarimetric measurements including information about the polarization phase difference are used. © 1993 John Wiley & Sons, Inc.