Detection and modelling of biodiversity patterns in the Bavarian Forest National Park by means of hyperspectral and LIDAR remote sensing.
Benjamin Leutner (12/2010-06/2012)
Support: Björn Reineking
Spatial mapping, quantification and prediction of biodiversity patterns is one of the crucial prerequisites for resolving current issues in ecology. Spatially comprehensive biodiversity assessments require substantial efforts even on small scales but are almost impossible to conduct on larger scales such as on the landscape level. This is where the combination of comprehensive remote sensing imagery with point data, e.g. species diversity on plot level, comes into play and allows spatially explicit modelling of biodiversity patterns.
This thesis will apply plot based vegetation assessments from within the Bavarian Forest National Park, Germany. These will be linked with hyperspectral imagery from the airborne HyMap sensor as well as airborne laser scanning data on forest structure (LIDAR). The linking will be by means of spatial statistical modelling.
The main issues to be resolved are:
- Can the phyto-diversity within the forest be detected using hyperspectral imagery as proxy?
- Will the detection power increase significantly if spectral and structural (LIDAR) information are combined?
This thesis is being conducted in cooperation of the Chair of Biogeographie (University of Bayreuth), the junior professorship Biogeographical Modelling (University of Bayreuth) and the Chair of Remote Sensing (University of Würzburg).