Mapping change of functional forest traits and diversity using airborne laser scanning in the canton Aargau 2014-2019

Forests contribute substantially to ecosystem functions and services making their ecological quality valuable. Due to climate change, monitoring diversity is becoming increasingly important to record a possible decline. High functional diversity has been related to a decreasing vulnerability to disturbances like diseases, storms and insect attacks. Remote sensing and especially LiDAR are promising methods to assess functional traits and diversity in forests and have been linked to plant diversity and ecosystem functioning. However, large-scale and multitemporal analyses using LiDAR datasets are just at the beginning. This thesis aims to assess functional forest traits and diversity metrics out of ALS data and to compare them between the years 2014 and 2019. Three morphological traits, namely canopy height, foliage height diversity and plant area index were estimated for the entire forest area of the canton Aargau under defoliated conditions. Then, functional richness and divergence were computed out of the traits. For three subregions of the canton, occlusion in the lower canopy was computed to assess if traits and diversity metrics are influenced. More complex derivations of ALS point clouds, e.g. plant area index, richness or divergence, were found to be more sensitive to external influences like different sensor and flight settings and occluded fractions of the canopy volume. Various spatial patterns of the derived traits and diversity metrics were mapped, e.g. a decrease or smaller increase in steep and high altitude regions. Richness values showed a very large global increase of 123%, which cannot solely be attributed to biotic changes, but is rather caused by the sensitivity to sensor-related factors. The results demonstrate how the development of robust methods for trait and diversity estimations is important. The incorporation of sensor and flight parameters into the estimation methods is crucial for improved performance in multitemporal analyses using ALS point clouds