Snow and Cloud Detection
The availability of large amounts of satellite imagery data through the European Copernicus project and open source platforms like the OpenDataCube (ODC) greatly bolsters the opportunities to apply classical machine learning and deep learning algorithms. These models can be employed to meet many challenges. One challenge that is particularly relevant in our time is that of climate change and global warming. An important aspect of this challenge is monitoring the effects of climate change using the aforementioned satellite data. Thus change of snow cover in itself is already an interesting phenomenon to follow and report as a complement to other measurements which allow us to monitor the indicators of climate, such as surface temperature, ocean heat, or sea levels. More importantly, however, it has been widely researched how the change in snow cover influences surface energy level, water balance, thermal regimes, and vegetation and thus plays a major role in the ecological, and hydrological systems, as well as the climate of Arctic and other regions. Moreover, the extent of snow cover, has more direct economic effects too. For example in road cleaning costs, and in architecture. In order to accurately track the changes in snow cover, it is an important step to be able to identify snow in satellite images, and also to distinguish between a snow-covered area and an area occluded by clouds. This inspired the launch of this challenge.
👉Find the entire challenge here!
📷 Photo by Alberto Restifo on Unsplash