The objective of this work is to investigate the influence of forested areas on the ac- cumulation and ablation of snow which affect the water balance of forested drainage basins. It is important to know the main meteorological drivers of unloading and sublimation of intercepted snow, which alter the accumulation and ablation of under- canopy snow. Extensive fieldwork was conducted in the Seehornwald in Davos, Switzerland to obtain quantitative measurements of the amount of snow unloading of coniferous trees after snowfall events. Tarps were installed below trees and on non forested areas to quantify snowfall and subsequent unloading. The data was then compared to three different unloading models which are part of a physical snowpack model. In this thesis two specific snow fall events are investigated in detail. Mea- sured interception of both events is similar, but the measured amount of unloading is completely different. For both events the models can capture measured interception similarly, whereas the performance for unloading is weak. A difference between the performance of two models for small and big unloading events is found.
A method to evaluate webcam pictures taken daily from 2011 to 2015 is de- veloped. The increase/decrease of pixels representing snow is compared to three different unloading models. The best model has a correlation with decreasing pixels (unloading) of 0.58. To quantify the sublimation rate, time lapse cameras were built and installed in the winter season 2015/2016. Pictures of single branches were taken half-hourly. Branches, where no mechanical unloading or snow melt took place, were examined to calculate the sublimation rate. The results were then compared to the variables of the latent heat flux of sublimation and in the following referred to as ”meteorological drivers” (relative humidity, mean wind, temperature and incom- ing long/short -wave radiation). In general temperature shows the best correlation (-0.69) with the sublimation rate.
To sum up: Manual quantitative unloading measurements provide good data to see differences in unloading between single trees. The method has promising approaches, but to calibrate or evaluate different unloading models, the amount of data is not sufficient and the effort to provide the data is really high. However, the method with pixel analyses provides good data to compare and evaluate unloading and/or interception models.