Investigation of snow depth determination potential of ICESat-2 point cloud data

Snow is one of the types of precipitation that is vital for ecosystems and the continuation of life cycles. However, in order to predict avalanche events that may occur as a result of heavy snowfall, it is important to evaluate the hazard occurrence and to predict the risk in order to reduce the loss of life and property in case of a possible disastrous event. For this reason, it is necessary to monitor the snow mass by measuring the snow depth at regular intervals; and in this context, it is applicable to benefit from satellite technologies that provide data with the help of active or passive sensor technologies on a local and/or global scale without the need for any field work. In the first part of this thesis, in line with the goal of determining snow depth using ICESat-2 satellite point cloud data, accuracy analyses were carried out primarily within the study areas of Erzurum and Van Ferit Melen airports. In this context, results in four different levels of reliability were obtained for point cloud data collected at different dates. The results revealed that the mean square error of point cloud data having high reliability level varies in a very narrow band (0.32 m – 0.49 m). Thereafter, in thesis, it was attempted to determine the potential for determining snow depth from the data collected by the ICESat-2 platform for the study areas in three different regions in Norway. In order to investigate the potential for determining snow depth from ICESat-2 data in the vicinity of 3 Norwegian airports (Bardufoss, Tromsø, Røros airports), it was calculated that the snow depth prediction errors were between -13 cm and +32 cm for the related study areas. When the results of snow depth estimation from the Norwegian study areas were analyzed, it was concluded that the results achieved were quite successful, and were highly compatible with the sensitivity level of the data having high level of confidence.