POLINSAR Workshop 2003
Session: Theoretical Modelling
Chair/s: Thuy Le Toan and Paul Saich
1. Session highlights
Although the scheduled theoretical modelling session had only three
oral papers, the use of physical models to interpret data or to derive
inversion algorithms was present in several presentations and poster papers.
Most current theoretical models are not adapted to polarimetric and
interferometric radar data, for which phase information needs to be estimated
accurately. The difficulties in such modelling originate both in the detailed
description of the medium and in calculations of higher order scattering.
For natural media such as vegetation, ice or soil, appropriate models are
needed to understand the (few) experimental results and to assess the generality
and robustness of inversion algorithms through model simulations.
2. Seed questions
Q1: Is theoretical modelling useful?
Yes. Despite their complexity, physical models are essential for the
development of inversion algorithms and the definition of optimum future
Q2: Reasons for relatively little modelling work?
Lack of funding: In the remote sensing community, theoretical modelling
is usually considered as upstream basic research, for which less funding
is available compared to application-oriented studies.
Gap between electromagnetism and remote sensing communities: There
are insufficient links between the two communities. For the electromagnetics
community, studies of natural media may appear more complex and less rewarding
than studies of man made material and targets.
Lack of appropriate datasets: The available datasets are often insufficient
for physical modelling work. There are few polarimetric and interferometric
datasets with relevant ground data. The datasets that are available are
often limited to few frequencies, and / or to particular times of the year.
Q3: Actions to be undertaken?
1. Identification of specific topics needed in modelling work.
Based on experimental observations and the state of the art in modelling,
progress depends on several topics which remain to be solved. A non-exhaustive
list might include, for example:
Vertical extinction in tree and forest canopies,
Higher order scattering for soil surfaces,
Polarisation behaviour of microwave measurements of crops,
Phase difference in sea ice
2. Conduct of dedicated experiments
To solve a given topic, dedicated experiments should be conducted making
use of one or more of the following: indoor, ground based in-situ, airborne,
spaceborne systems, with relevant ground data. Past datasets, often oriented
to general modelling purposes, usually did not contain sufficient information
to investigate particular crucial points.
3. Provision of radar and ground datasets
Funding should be used to set up databases containing relevant radar
and ground data for use by the community. Existing initiatives such
as ERA-ORA (EU Network, 1998-2000) should be extended. In particular, there
is a need to have ENVISAT datasets adapted to modelling and to the development
of retrieval algorithms. Such datasets, with incidence angle and
polarisation diversity, need to be assembled from different test
sites by teams working with the same ground data collection procedures.
Model Generation from SAR Interferometry
Mr Mohammad Halmi Kamaruddin
Modelling of the
scattering by a smooth dielectric cylinder: study of the complex scattering
matrix using two different models
On the physical
modelling of polarimetric parameters of forests
Dr Thuy Le Toan
of the sensitivity of polarisation coherence to soil surface roughness
Dr Francesco Mattia