ABSTRACT
The planning of diverse agricultural crop management practices requires accurate spatio-temporal information. In this context, synthetic aperture radar (SAR) data is well suited for crop morphological monitoring within the growing season. Many radar-derived characterization parameters are utilized to capture the changes in the scattering mechanisms from crop fields concerning their phenological stages. This chapter elaborates the derivation of recently proposed target descriptors and by Dey et al. and also the corresponding model-free four-component scattering power components (MF4CF). The temporal variations of these parameters for different incidence angles are shown for rice crops over test sites in Vijayawada, India, and Colinas, Spain. RADARSAT-2 images acquired over these two test sites are used for complete analysis. It is known that polarimetric information provides excellent sensitivity for identifying crop phenology stages. Hence, an assessment of the rice phenology classification using model-free scattering power components is performed. The overall classification accuracy achieved using the MF4CF components is 93.42%. The temporal analysis of the scattering-type parameters and the scattering power components helps investigate the detailed scattering characteristics from rice across its phenological stages.
