Multitemporal and dual-polarization observations of agricultural vegetation covers by X-band SAR images

Abstract

This paper presents a study of SAR images which aims to demonstrate the capability of radar to identify vegetation types and to determine the vegetation canopy parameters. The investigations, per­ formed on bi-temporal and dual-polarizatio n (HH and VV) X-band SAR images of an agricultural scene, included all the steps involved, from SAR images to applications: 1) The retrieval of radar backscatter val­ ues from SAR images has been possible with the use of external ref­ erence targets; 2) the analysis of the polarization and temporal re­ sponses of the vegetation covers has highlighted the particular behavior of flooded rice fields; 3) the observations of ricefields have been inter­ preted by a theoretical model. During an extended period of rice plant growth, the model, in agreement with the observations, suggests a pos­ sible use of both HH and VV images for ricefield mapping and moni­ toring; and 4) a result of ricefield mapping with an algorithm based on the above results is presented.