2020/10/27 by Manu Tom, Tom, Manu, Rajanie Prabha +9
Earth and Planetary Sciences · #Arctic and Antarctic ice dynamics #Artificial Intelligence (cs.AI) #Climate change and permafrost #Computer Vision and Pattern Recognition (cs.CV) #Cryospheric studies and observations #FOS: Computer and information sciences
paper · pdf · doi:10.48550/arxiv.2010.14300
openalex publication_date 2020/10/27 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Continuous observation of climate indicators, such as trends in lake\nfreezing, is important to understand the dynamics of the local and global\nclimate system. Consequently, lake ice has been included among the Essential\nClimate Variables (ECVs) of the Global Climate Observing System (GCOS), and\nthere is a need to set up operational monitoring capabilities. Multi-temporal\nsatellite images and publicly available webcam streams are among the viable\ndata sources to monitor lake ice. In this work we investigate machine\nlearning-based image analysis as a tool to determine the spatio-temporal extent\nof ice on Swiss Alpine lakes as well as the ice-on and ice-off dates, from both\nmultispectral optical satellite images (VIIRS and MODIS) and RGB webcam images.\nWe model lake ice monitoring as a pixel-wise semantic segmentation problem,\ni.e., each pixel on the lake surface is classified to obtain a spatially\nexplicit map of ice cover. We show experimentally that the proposed system\nproduces consistently good results when tested on data from multiple winters\nand lakes. Our satellite-based method obtains mean Intersection-over-Union\n(mIoU) scores >93%, for both sensors. It also generalises well across lakes and\nwinters with mIoU scores >78% and >80% respectively. On average, our webcam\napproach achieves mIoU values of 87% (approx.) and generalisation scores of 71%\n(approx.) and 69% (approx.) across different cameras and winters respectively.\nAdditionally, we put forward a new benchmark dataset of webcam images\n(Photi-LakeIce) which includes data from two winters and three cameras.\n