Mapping of small-scale intertidal macrophyte communities at the island of Helgoland (North Sea) using hyperspectral remote sensing images

Edit Item Edit Item

General Information:

Bartsch, I. , Thiemann, S. , Hennig, B. D. and Cogan, C. (2005): Mapping of small-scale intertidal macrophyte communities at the island of Helgoland (North Sea) using hyperspectral remote sensing images , 2nd EARSeL Workshop Remote Sensing of the Coastal Zone, 9-11 June, Poro, Portugal. .
Cite this page as:
Contact Email:

Supplementary Information:


AbstractHard bottom marine biocoenoses constitute one of the most important and productive ecosystems in the world. The detection of stability or change in different hierarchical levels of these complex systems is difficult but of major international relevance for basic and applied research. World-wide traditional biological field methods created only sparse information regarding the recognition of spatial changes of intertidal biotopes. Remote sensing techniques may help to assess biodiversity change on a high hierarchical level and with a synoptic view of the system, complementing the detailed biological biodiversity studies with their complex spatial information.The marine nature reserve of the island of Helgoland (North Sea, Germany) comprises a large amount of the representative species for northern European rocky coastlines, but the major communities are small-scaled (patches of 10-1000 m2 size) and interwoven. The mapping of these biotopes by airborne remote sensing was the aim of this study in order to develop a method for long-term spatial monitoring of intertidal communities. In July 2002 and September 2003, two data sets of Helgolands coast were acquired with the hyperspectral sensor ROSIS operating at approximately 1m pixel size. The data were radiometrically, atmospherically, and geometrically corrected. Based on ground truth data and detailed spectral analysis, a supervised hierarchical classification scheme was developed to classify the major intertidal communities. Comparisons between the 2002 and 2003 data and between different classification protocols are presented. Simple, cost-friendly machine classification is not possible, but the advanced analysis of the hyperspectral data show a good congruence with a traditionally derived biotope map. The spectral classes are not an exact copy of the mapped biotopes, but represent an even higher hierarchical level of biodiversity and accurately record the patchiness of classes in contrast to the biotope map. The potential and limitations of this spectral approach will be discussed and suggestions for improvement are presented. The method has potential to monitor the spatial change of intertidal communities in the context of the EU-water frame work directive.

Further Details:

read more
OAI 2.0:
ePIC is powered by:
EPrints 3