Research project T4/DD/36 (Research action T4)
The proposed research project is a continuation of the currently ongoing project on ?Bio-physical characterisation of tropical ecosystems by space borne SAR remote sensing, nicknamed EcoSar (Telsat4/DD/005-006).
During the first phase of the project (1997-1999) a number of methodological and thematic issues were tackled. The main milestones to be mentioned here are:
- The incorporation of spatial heterogeneity and of grass blade structures into the UCL Radiative Transfer Model for vegetated targets;
- The implementation of a novel procedure for speckle filtering of coincident multi-temporal SAR data;
- The production of generalised land cover maps with high spatial consistency by multi-temporal SAR data segmentation and per segment classification;
- The implementation of a novel texture extraction procedure and its use in characterising forest canopy architecture.
The proposed second phase of the EcoSar project envisages to consolidate, validate and refine the results obtained so far by:
- producing operational versions of the laboratory developments (model and speckle filter) by incorporating validity range checks, restricting input to significant parameters, and providing it with a graphical user interface and full documentation;
- validating the thematic results obtained with the methods, developed based on historical data, with real-time reference measurements and outlining their range of validity;
- upgrading the conceptual water balance model to a physical model at the same time calibrating it with real time field observations;
- demonstrating the value of the methods developed at a scale representative for the TELSAT programme objectives through co-operation agreements established with the JRC on interpretation of large scale forest mapping exercises using ERS and JERS SAR data.
A new research line introduced for the second phase of the EcoSar project concerns the comparison and eventual synergetic use of space borne SAR data with optical data. The objective being the further improvement of the characterisation of vegetation seasonal features, SPOT Vegetation data are chosen for their high temporal resolution. They will be analysed with respect to the SAR data acquired during the filed campaign, and on a second occasion to Envisat ScanSar data.
