Summary
International Symposium on Antennas and Propagation
2009
Session Number:1F2
Session:
Number:1F2-4
Validity range of some approximations involved in vegetation modelling through numerical approaches
Pierre Borderies, Ludovic Villard,
pp.269-272
Publication Date:2009/10/21
Online ISSN:2188-5079
Summary:
Electromagnetic models of interaction with natural land media are extensively used to analyze and foresee the Synthetic Aperture Radar observables, and may be used to study the robustness of retrieval algorithms. However, usually, these models rely on approximations of various origins which may jeopardize the relevancy of some simulations, all the more so as retrieval algorithms rely more and more on complex sets of observables including polarimetry and interferometry, which makes the requirements toward these models still more demanding than for the mere radiometric outputs. In particular, the need for deriving phase properties has led to the development of coherent codes (e.g. [8][9][10]), which deliver greater range of outputs but may incorporate more approximations. Concerning these approximations, comparisons with experimental data are of course necessary but may be not sufficient. Indeed, since the land cover scenes are generally very complex, when a disagreement occur between simulated and experimental data it may be very difficult to attribute the discrepancy to a particular cause. Then the use of numerical techniques for studying a particular point inside a complex continental land scene may be relevant. The goal of the present paper is to illustrate that Finite Differences in Time Domain (FDTD) is well matched to treat such problems in their specificity and that it permits answering to lot of questions relative to these approximations. Section 1 deals with the numerical approach and section 2 illustrate its application to vegetation scattering.