Airborne laser scanning has already become an essential part of forest inventories in the Nordic countries and in Canada. However, its wider practical forestry application in the Central European countries is awaiting. In the first part of this thesis, a procedure for modelling of the basic stand variables, especially forest stand volume, was designed using an area-based approach. Not only linear regression was used for modelling but also machine learning (k-nearest neighbor algorithm, Random Forest, Support Vector Machine and neural networks). In the second part, the thesis deals with biomass estimation using the area-based approach and with comparison of the empirical and semi-empirical approaches to modelling. The third part deals with leaf area index (LAI) estimation using penetration indices and LiDAR metrics. The eLAI estimated by optical method is commonly used for model fitting in overwhelming majority of scientific papers. The benefit of this study is usage of LAI derived by destruction method (directly). The thesis as a whole has shown that the airborne laser scanning is also usable in a variety of forestry applications in Central European countries.
| Identifer | oai:union.ndltd.org:nusl.cz/oai:invenio.nusl.cz:425015 |
| Date | January 2017 |
| Creators | Patočka, Zdeněk |
| Source Sets | Czech ETDs |
| Language | Czech |
| Detected Language | English |
| Type | info:eu-repo/semantics/doctoralThesis |
| Rights | info:eu-repo/semantics/restrictedAccess |
Page generated in 0.0019 seconds