The detection of snow avalanches using TerraSAR-X data

Inhalt:
Snow avalanches are natural hazards causing damage to nature, persons, and property in mountainous terrain. During periods with high avalanche activity, detailed avalanche information about date, location, and size is missing. In the presented study, two single-polarized TerraSAR-X StripMap image pairs acquired in January 2019 are used for avalanche detection. The study site is located in Tyrol and Vorarlberg, Austria, and South Tyrol, Italy. The relevant information of the Avalanche Warning Service Information System (LAWIS) has been used as reference data. The approach bases on the application of diverse texture filters with a subsequent change detection. All texture filters showed a classification correctness of around 70%, though the Entropy filter had the best result with 76%.