Science Service System

Summary of Proposal COA0118

TitleThe Use of TerraSAR-X and Optical Multispectral Data for Improved Classification of Wadden Sea Surface Types (TUTOR).
Investigator Brockmann, Carsten - Brockmann Consult, Geo-Information Services
Team Members
Dr. Gade, Martin - University of Hamburg, Institut fuer Meereskunde (IfM)
SummaryMotivation
The Wadden Sea is of major commercial and ecological significance, and has for centuries been exploited by man for communication, transportation, waste disposal, power generation, fisheries and amenity development. For this reason the governments of Germany, Holland and Denmark have proposed that areas of the Wadden Sea be added to the global list of Particularly Sensitive Sea Areas. The long term sustainability of this area requires proper management and this in turn requires a) the appropriate policy directives and b) the necessary tools to carry out these directives. Monitoring is an integral part of this process and having the appropriate tools for carrying out reliable and accurate measurements is essential for informed decision making. By comparing and combining the two remote sensing technologies (SAR and optical) it is hoped that new insights will be obtained into the potential of classifying tidal flat surface structures for monitoring purposes.

Team
The present team is comprised of a small but efficient group of experts in both earth observation techniques, including IfM for SAR and BC for optical techniques, as well as the Wadden Sea Nationalpark Agency responsible for implementing EU Directives and expert in present day monitoring requirements.

Objectives
The present study will assess the value of TerraSAR-X data in combination with optical multispectral scanner data to improve the classification and mapping of important Wadden Sea tidal flat structures (sediment type and mussel beds).

Data requirements
The anticipated data requirements are:
50 StripMap (SM) images
20 SpotLight (SL) images
20 High-Resolution SpotLight (HS) images
and covering seasonal changes on the tidal flats.

Methods
TerraSAR-X data will be used together with data from other sensors. The data will be geo-referenced and collocated using state-of-the-art data processing tools. Supervised classification techniques will be applied to the data groups to highlight and locate peculiarities. In parallel, the inversion method using the IEM will then be applied to derive surface roughness parameters. This new method is based on the inversion of the single-scattering Integral Equation Model (IEM) and will be applied to provide estimates of the surface roughness parameters, rms height and correlation length. The roughness parameters will be derived from pairs of co-polarized SAR images acquired quasi-simultaneously at different electro-magnetic frequencies (i.e. differnt platforms) during low water in the German Wadden Sea.
Analysis of optical RS data will involve classification methods which include linear spectral unmixing and supervised classification techniques. This is part of ongoing studies at BC. Validation will be carried out using in-situ measurements obtained by monitoring agencies.
Finally, the information gained (from the classification and inversion procedures) will be used to provide improved maps of the distribution of sediment types and mussels beds.

Deliverables
  • Improved methodology for the classification of sediment types and mussel banks through the synergetic use of TerraSAR-X and optical multispectral data (Report) .
  • Improved understanding of the specific capabilities and advantages of each of the two technologies (Report).
  • Possibility of new and additional parameters being ‘recognised’ and better classified using both technologies in combination (Report).
  • Maps of distribution of sediment types and mussel banks (Maps)


Schedule
The project will run for one year, beginning in July, 2006 and ending in June, 2007.

Funding
The proposed project is viewed as a pilot study and funding will be required and sought for any follow-up project which may result from its findings.

Back to list of proposals

© DLR 2004-2016