You must have a javascript-enabled browser and javacript and stylesheets must be enabled to use some of the functions on this site.
 
   ESA       
   
Introduction

 

Polarimetric Options for P-Band SAR Tomography of Forested Areas

Stefano Tebaldini(1) and Fabio Rocca(1)

(1) Politecnico di Milano, P.za Leonardo da Vinci, 20133, Milan, Italy

Abstract

The aim of this paper is to provide a discussion about the potentialities of SAR Tomography (T-SAR) to infer information about the vertical structure of forested areas at P-Band, basing on the availability of either a fully polarimetric or a single polarimetric data. The core of the analysis is represented by a model of the second order statistics of the multi-baseline, multi-polarimetric data. Such a model will serve as the basis to pose the problem of the retrieval of the vertical structure of a forest as the problem of detecting multiple targets within the SAR resolution cell and estimating the spatial structure and backscattered power of each detected target. In this framework, the well posing of the Tomographic problem can be discussed in a simple fashion, simply be considering the number of equations that arise from the data-set with respect to the number of the unknowns required to solve the model. Our conclusion is that the availability of multiple baselines can be successfully exploited in order to estimate the vertical structure even in the presence of a single polarimetric data, provided that two conditions are fulfilled: 1) temporal decorrelation is not such to completely destroy the information relative to the targets; 2) phase disturbances, such as those due to residual platform motion and propagation through the atmosphere, can be successfully compensated for through a calibration step. As a result, it turns out that at least 3 fully polarimetric acquisitions or 4 single polarimetric acquisitions are required to solve a model that accounts for the presence of ground and canopy contributions. If fewer acquisitions are available, then further hypotheses are required, in order to reduce the number of unknowns within the model. These conclusions are fully supported by the results obtained in the framework of the ESA project BioSAR, aimed at the investigation of the Radar signature of boreal forests at P-Band. Results relative to single and multi polarimetric Tomographic analyses are presented.

 

Workshop presentation

Full paper

 

  Higher level                 Last modified: 07.05.06