Sea ice concentration forecasts for 1-day lead time have been produced using a deep learning model. The deep learning model used for producing these forecasts was developed from the work of Palerme et al., 2023. Though the list of predictors and the model architecture is almost the same as in Palerme et al., 2023 (only the number of layers was modified), the deep learning model developed in SEAFARING is regional and at a higher spatial resolution (5 km). Palerme C., Lavergne T., Rusin J., Melsom A., Brajard J., Kvanum A.F., MacDonald Sørensen A., Bertino L., Müller M. (20223) Calibration of short-term sea ice concentration forecasts using deep learning. The Cryosphere Discussions. link to the paper
Sea-ice charts are provided by the Norwegian Ice Service. They are analysed from high-resolutions satellite information and available on weekdays.
Model forecasts for the following seven days are from the ocean and sea-ice forecasting system TOPAZ4. It is part of the Copernicus Arctic Marine Forecasting Center (ARCMFC). Details on the model quality can found on MET Norway's information page or in the documents provided here.
The sea-ice edge is defined as the 15% sea-ice concentration contour.