Modelling of Forest Structure with Remote Sensing Data
Potential and Limits of Hyperspectral Information
Versandkostenfrei!
Versandfertig in 6-10 Tagen
33,99 €
inkl. MwSt.
PAYBACK Punkte
17 °P sammeln!
Forest structure, e.g. the composition and distribution of tree species, is a key element for characterizing ecological functions and ecological state of forest ecosystems. The aim of this master thesis was to model 15 forest structure measures of a mixed temperate forest in the Bavarian Forest National Park (Germany) with hyperspectral remote sensing data (HyMap). The findings indicated that hyperspectral data has high potential to identify forest structure even in heterogeneous mixed forests like the Bavarian Forest National Park. Forest structure measures were derived from vegetation survey...
Forest structure, e.g. the composition and distribution of tree species, is a key element for characterizing ecological functions and ecological state of forest ecosystems. The aim of this master thesis was to model 15 forest structure measures of a mixed temperate forest in the Bavarian Forest National Park (Germany) with hyperspectral remote sensing data (HyMap). The findings indicated that hyperspectral data has high potential to identify forest structure even in heterogeneous mixed forests like the Bavarian Forest National Park. Forest structure measures were derived from vegetation surveys on 102 ground plots and served as dependent variables in a decision tree based random forest model. The independent variables were obtained from hyperspectral data in 7 m resolution, which was transformed by a minimum noise fraction rotation (MNF). With these two datasets random forest model performance on each forest structure measure was compared to model performance derived from literature. Furthermore, descriptive statistics, correlation analysis and ordination methods were used to discuss the results.
Andere Kunden interessierten sich für
