Dynamic Ice Map”: Combining High-Resolution Sea Ice Type Classification With Sea Ice Drift Forecast

Kurzfassung

To assist navigation through ice-covered waters, Synthetic Aperture Radar (SAR) Earth observation data is often used due to its independence of daylight and weather. Currently, SAR images are acquired and automatically delivered onboard the research vessel Polarstern during campaigns in the Arctic. However, the surrounding sea ice has drifted by the time of delivery and the SAR scenes have to be shifted manually to collocate them to the current ice situation around the ship. This hinders the direct use of SAR scenes for route planning. We aim to enable automatic route planning through drifting sea ice. The sea ice drift forecast data provided on Copernicus Marine Environmental Monitor Service [1, 2] is applied in a value-adding process for this operational application. We use SAR-based sea ice type classification, derived with machine learning from Sentinel-1 or TerraSAR-X SAR data, as static ice type maps [3, 4]. Then we polygonise the ice map and obtain the corner points of the polygons for data reduction. To enable dynamic route planning in a dynamic environment, the static ice map is advected along Lagrangian-Trajectories, using the sea ice drift forecast model data, e.g. TOPAZ [6] or neXtSIM [5]. We provide hourly slices of advected corner points. The result is a spatial-temporal ice map ready for routing (dynamic ice map) that bridges scales from the 160 metres to 6 kilometres resolution. In further processing, speed-optimized route suggestions considering the obtained dynamic ice map are calculated. The routes are then sent to ship bridges. At the current stage of development, routes are calculated optimized for speed, where reduced sailing velocities are assumed in thicker ice situations. Bathymetry is also considered, i.e. navigating through shallow waters is prohibited. We tested our routing support on RV Polarstern with this new data product in a two weeks period during the ArcWatch-II expedition in early autumn 2024 to the Central Arctic. The calculated routes were used to derive decisions for navigation, together with the ice-radar of the ship and spaceborne SAR images. The test case has shown that especially longer routes can benefit from the dynamic ice map. We will present the new algorithm used for the creation of dynamic ice maps. Moreover, an evaluation of the used sea ice drift forecast models together with a detailed quality assessment of the dynamic sea ice maps will be provided. References [1] European Union - Copernicus Marine Environmental Monitor Service: neXtSIM-F: Arctic Ocean Sea Ice Analysis and Forecast, https://doi.org/10.48670/moi-00004, 2020. [2] European Union - Copernicus Marine Environmental Monitor Service: TOPAZ5: Arctic Ocean Physics Analysis and Forecast, https://doi.org/10.48670/moi-00001, 2015. [3] Kortum, K., S. Singha, and G. Spreen, Robust Multiseasonal Ice Classification From High-Resolution X-Band SAR: IEEE Transactions on Geoscience and Remote Sensing, 60, 1–12, http://dx.doi.org/10.1109/TGRS.2022.3144731, 2022. [4] Murashkin, D., and A. Frost, Arctic Sea ICE Mapping Using Sentinel-1 SAR Scenes with a Convolutional Neural Network, 5660–63, http://dx.doi.org/10.1109/IGARSS47720.2021.9553206, 2021. [5] Rampal, P., Bouillon, S., Ólason, E. and Morlighem, M.: neXtSIM: a new Lagrangian sea ice model, The Cryosphere, 10, 1055–1073, https://doi.org/10.5194/tc-10-1055-2016. [6] Sakov, P., Counillon, F., Bertino, L., Lisæter, K. A., Oke, P. R. and Korablev, A.: TOPAZ4: an ocean-sea ice data assimilation system for the North Atlantic and Arctic, Ocean Science., 8, 633–656, https://doi.org/10.5194/os-8-633-2012.