Artificial Intelligence for Snow Cover in Mountain Regions
Introduction
For the majority of the river catchments in the Northern Hemisphere, snow is the dominant water source. Therefore, especially mountain regions are considered as “water towers” for whole continents. Climate change leads to alterations in the mountain snow coverage, it affects not only its duration (onset of accumulation and snow melt) but also the snowpack itself. This is mainly the snow depth and the water stored within the snowpack – the snow water equivalent (SWE). Remote sensing-based snow cover products are available for several parameters, comprising fractional snow cover (FSC), albedo, wet/dry snow information, as well as SWE. Currently, operational sensors and algorithms cannot provide suitable information for mountain regions where the topography severely limits the capabilities: What would be required is high spatial (100 m pixel size at least) and temporal (daily availability) resolution. These requirements cannot be met based on the current algorithms and techniques. Artificial Intelligence (AI) has been used widely during the last decade to try and address many research challenges from different research fields, including remote sensing. Within the project “AI4Snow”, AI will be used to combine different remote-sensing products with meteorological in-situ observations to increase the spatial and temporal information on mountain snow cover. This information will be incorporated into hydrological models of test regions in Canada and Switzerland.
Innovation
“AI4Snow” will utilize the latest remote sensing products provided by the Copernicus Earth Observation (EO) program and innovative artificial intelligence (AI) methods to map mountain snow cover at unprecedented temporal and spatial coverage. The AI will be trained on a data cube, where remote sensing products from different sources and resolutions will be incorporated together with land cover information, elevation, and gridded meteorological data. The resulting dataset will be tested and validated in selected test sites which will be located in the Canadian Rockies and Suisse alps.
Contact
For further information, please contact:
Dr. Andreas Dietz (Project Leader)
German Aerospace Center (DLR)
Earth Observation Center (EOC)
German Remote Sensing Data Center (DFD)
Muenchener Strasse 20
82234 Wessling
Germany
Phone: +49 8153 28-1511