Sensitivity Study of X-Band Multiparametric SAR Data From Cabbage Fields

Abstract

To make polarimetric synthetic aperture radar (SAR) accessible to mainstream applications, the true composition ratio of scattering mechanisms within a radar backscatter must be accurately identified. To validate polarimetric SAR decomposition techniques, a novel multiparametric SAR observation combined with a theoretical model simulation has been applied to rice paddies having simple vegetation structure on smooth and flat surfaces, where the surface scattering was small. In this article, cabbage fields in Miura city in Japan were selected as outdoor-grown vegetables, so the effect of exposed ground surfaces must be taken into account. In combination with the multiparametric SAR observation, a dominant scattering mechanism is reliably isolated through the theoretical characterization of the data by a discrete scatterer model (DSM). Suppressing unnecessary effects by row structures of the crop field, the volume scattering from cabbage leaves is identified as a dominant scattering mechanism over most incidence angles. On the other hand, the surface scattering is dominant for copolarizations only at small incidence angles. The analysis also indicates the possibility that greater soil moisture could be kept by roots of mature cabbages. To demonstrate the validation of polarimetric decomposition by the results, a popular three-component decomposition was applied to the observation. Differences between the decomposition and the simulation results were analytically explained by features of the backscatter from cabbage fields. Model-based polarimetric decompositions shall be improved through this approach.