For data analysis optical satellite data are combined with radar data and in-situ data. The classification method for determining the types of surface uses the spectral segregation and ratios of the recording bands together to determine the spectral signal of the individual surface sections. In the Wadden Sea, we rarely have clearly distinguishable boundaries, but rather smooth transitions from one type of surface to another. However, in order to eventually arrive at a preferred class, the input parameters are divided into specific classes in a knowledge-based decision tree. 


By means of multi-temporal statistical parameters, the TerraSAR-X data provide information about different surfaces, which in turn are included in the classification via the decision tree. For example, so far the detection of mussel beds has been improved significantly through the synergistic use of optical and radar data.


Image of a mussel bed to the rear of the island of Norderney in optical satellite data

SPOT, false colour

TerraSAR-X, Spotlight-Modus



The field dat from the current and previous project provide important data for validating the classification results. The classification can be assessed using the field information and unresolved questions can be answered by the detailed photo documentation at each measurement point. The validation resulted in an overall accuracy of 69% for Lower Saxony, with the areas of sand achieving the best result with an accuracy of up to 83%. Area comparisons for seagrass meadows and mussel beds could be carried out via comparisons with monitoring data. For individual meadows and mussel beds, a surface correlation of 60 to 98% has so far been achieved.


Classification of the mussel bed im Rückseitenwatt der Insel Norderney

Classification by optical data alone

Classification by combining optical and radar data

Classified mussels are coloured in red, while the black line shows the results from mussel monitoring.



Numerous map products are and will be developed from different types of classification. Products with various spatial and temporal resolutions are developed, from classification of the intertidal flats, to the distribution and dynamics of different sediments, to probability maps:

The products are developed in close collaboration with the potential users. This ensures that the products fully meet their requirements.