Comparison of three modelling approaches of potential natural forest habitats in Bavaria, Germany

In the context of the EU Habitats Directive, which contains the obligation of environmental monitoring, nature conservation authorities face a growing demand for effective and competitive methods to survey protected habitats. Therefore the presented research study compared three modelling approaches (rule-based method with applied Bavarian woodland types, multivariate technique of cluster analysis, and a fuzzy logic approach) for the purpose of detecting potential habitat types. The results can be combined with earth observation data of different geometric resolution (ASTER, SPOT5, aerial photographs or very high resolution satellite data) in order to determine actual forest habitat types. This was carried out at two test sites, situated in the pre-alpine area in Bavaria (southern Germany). The results were subsequently compared to the terrestrial mapped habitat areas of the NATURA 2000 management plans. First results show that these techniques are a valuable support in mapping and monitoring NATURA 2000 forest habitats