Snow Height Classification Dataset

Snow Height Classification dataset provides manually annotated snow height data that can be used for development and evaluation of automatic snow height classification approaches.

A subset of 20 IMIS stations which span different locations and elevations and vary in underlying surface (e.g., vegetation, bare ground, glacier, etc.) were selected and manually annotated with binary two-class ground truth information regarding snow height data:

• Class 0 - Snow - the surface is covered by snow
• Class 1 - No Snow - the surface is snow-free (e.g., vegetation, soil, rocks, etc.)

Data has been annotated with the help of domain experts. It should be mentioned that annotating historical data is problematic, as there is no way of checking whether there really was snow at the station or not. This means that assessing the presence of snow with the help of information from other sensors should be considered a best effort approach.

Besides the annotated snow height data, the dataset also contains additional data which are relevant to reproduce results in the following repository: https://gitlabext.wsl.ch/jan.svoboda/snow-height-classification

Any further use of the data has to comply with the CC BY-NC license: https://creativecommons.org/licenses/by-nc/4.0/

• Swiss Data Science Center (link) (Grant/Award: C21-15L)
• WSL Institute for Snow and Avalanche Research SLF