Advisor: Gottfried Mandlburger

The increasing availability of high-resolution airborne laser bathymetry data enables detailed three-dimensional mapping of river systems, including structures located beneath the water surface. One such structure is deadwood, which plays an important role in river ecology but can also pose potential risks to infrastructure. Reliable monitoring therefore requires a reproducible method capable of detecting and geometrically describing deadwood directly from point clouds. Conventional approaches are often limited in their applicability to submerged environments.The aim of this master thesis is the development of an reproducible workflow for the detection and quantification of submerged deadwood from airborne laser bathymetry point clouds. The study area is located in the Natura 2000 protected area Neubacher Au along the river Pielach in Lower Austria. Data processing is performed using the OPALS software framework as well as Python-based scripts and includes several processing steps such as outlier filtering, derivation of a water surface model, refraction correction based on Snell’s law, terrain classification, and the computation of eigenvalue-based geometric metrics for point cloud classification. Based on these features, a tree-based classification is applied to separate deadwood points from terrain, water surface, and noise points. Subsequently, the detected deadwood points are segmented and geometrically modelled. Using principal component analysis and robust curve fitting, geometric parameters such as length, diameter, and volume of individual deadwood elements are derived.The results demonstrate that submerged deadwood can be detected and geometrically characterized using airborne laser bathymetry data. The developed workflow enables detection and quantitative description of deadwood structures and provides a basis for river monitoring. This work contributes to improved environmental monitoring and supports applications in river ecology, hazard assessment, and river management.