Articles | Volume 19, issue 10
https://doi.org/10.5194/tc-19-4149-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/tc-19-4149-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
02 Oct 2025
Research article |
| 02 Oct 2025
| 02 Oct 2025
Developing a deep learning forecasting system for short-term and high-resolution prediction of sea ice concentration
Development Centre for Weather Forecasting, Norwegian Meteorological Institute, Oslo, Norway
Department of Geosciences, University of Oslo, Oslo, Norway
Cyril Palerme
Development Centre for Weather Forecasting, Norwegian Meteorological Institute, Oslo, Norway
Malte Müller
Development Centre for Weather Forecasting, Norwegian Meteorological Institute, Oslo, Norway
Department of Geosciences, University of Oslo, Oslo, Norway
Jean Rabault
IT Department, Norwegian Meteorological Institute, Oslo, Norway
Nick Hughes
Ice Service, Norwegian Meteorological Institute, Tromsø, Norway
Related authors
Cyril Palerme, Johannes Röhrs, Thomas Lavergne, Jozef Rusin, Are Frode Kvanum, Atle Macdonald Sørensen, Arne Melsom, Julien Brajard, Martina Idžanović, Marina Durán Moro, and Malte Müller
EGUsphere, https://doi.org/10.5194/egusphere-2025-2001, https://doi.org/10.5194/egusphere-2025-2001, 2025
Short summary
Short summary
We present MET-AICE, a sea ice prediction system based on artificial intelligence techniques that has been running operationally since March 2024. The forecasts are produced daily and provide sea ice concentration predictions for the next 10 days. We evaluate the MET-AICE forecasts from the first year of operation, and we compare them to forecasts produced by a physically-based model (Barents-2.5km). We show that MET-AICE is skillful and provides more accurate forecasts than Barents-2.5km.
Cyril Palerme, Thomas Lavergne, Jozef Rusin, Arne Melsom, Julien Brajard, Are Frode Kvanum, Atle Macdonald Sørensen, Laurent Bertino, and Malte Müller
The Cryosphere, 18, 2161–2176, https://doi.org/10.5194/tc-18-2161-2024, https://doi.org/10.5194/tc-18-2161-2024, 2024
Short summary
Short summary
Sea ice forecasts are operationally produced using physically based models, but these forecasts are often not accurate enough for maritime operations. In this study, we developed a statistical correction technique using machine learning in order to improve the skill of short-term (up to 10 d) sea ice concentration forecasts produced by the TOPAZ4 model. This technique allows for the reduction of errors from the TOPAZ4 sea ice concentration forecasts by 41 % on average.
Cyril Palerme, Johannes Röhrs, Thomas Lavergne, Jozef Rusin, Are Frode Kvanum, Atle Macdonald Sørensen, Arne Melsom, Julien Brajard, Martina Idžanović, Marina Durán Moro, and Malte Müller
EGUsphere, https://doi.org/10.5194/egusphere-2025-2001, https://doi.org/10.5194/egusphere-2025-2001, 2025
Short summary
Short summary
We present MET-AICE, a sea ice prediction system based on artificial intelligence techniques that has been running operationally since March 2024. The forecasts are produced daily and provide sea ice concentration predictions for the next 10 days. We evaluate the MET-AICE forecasts from the first year of operation, and we compare them to forecasts produced by a physically-based model (Barents-2.5km). We show that MET-AICE is skillful and provides more accurate forecasts than Barents-2.5km.
Åsmund Bakketun, Jostein Blyverket, and Malte Müller
EGUsphere, https://doi.org/10.5194/egusphere-2025-1693, https://doi.org/10.5194/egusphere-2025-1693, 2025
Short summary
Short summary
Obtaining accurate estimates of seasonal snow conditions requires a combination of observations and numerical models. We use a model accounting for the vertical structure of the snow, and a data assimilation method representing varying uncertainty of the model in time and space. Compared to existing products, neglecting these considerations, our system produced improved estimates of seasonal snow conditions. Snow mass estimates suggest a potential impact on derived hydrological applications.
Jean Rabault, Trygve Halsne, Ana Carrasco, Anton Korosov, Joey Voermans, Patrik Bohlinger, Jens Boldingh Debernard, Malte Müller, Øyvind Breivik, Takehiko Nose, Gaute Hope, Fabrice Collard, Sylvain Herlédan, Tsubasa Kodaira, Nick Hughes, Qin Zhang, Kai Haakon Christensen, Alexander Babanin, Lars Willas Dreyer, Cyril Palerme, Lotfi Aouf, Konstantinos Christakos, Atle Jensen, Johannes Röhrs, Aleksey Marchenko, Graig Sutherland, Trygve Kvåle Løken, and Takuji Waseda
EGUsphere, https://doi.org/10.48550/arXiv.2401.07619, https://doi.org/10.48550/arXiv.2401.07619, 2024
Short summary
Short summary
We observe strongly modulated waves-in-ice significant wave height using buoys deployed East of Svalbard. We show that these observations likely cannot be explained by wave-current interaction or tide-induced modulation alone. We also demonstrate a strong correlation between the waves height modulation, and the rate of sea ice convergence. Therefore, our data suggest that the rate of sea ice convergence and divergence may modulate wave in ice energy dissipation.
Cyril Palerme, Thomas Lavergne, Jozef Rusin, Arne Melsom, Julien Brajard, Are Frode Kvanum, Atle Macdonald Sørensen, Laurent Bertino, and Malte Müller
The Cryosphere, 18, 2161–2176, https://doi.org/10.5194/tc-18-2161-2024, https://doi.org/10.5194/tc-18-2161-2024, 2024
Short summary
Short summary
Sea ice forecasts are operationally produced using physically based models, but these forecasts are often not accurate enough for maritime operations. In this study, we developed a statistical correction technique using machine learning in order to improve the skill of short-term (up to 10 d) sea ice concentration forecasts produced by the TOPAZ4 model. This technique allows for the reduction of errors from the TOPAZ4 sea ice concentration forecasts by 41 % on average.
Qin Zhang and Nick Hughes
The Cryosphere, 17, 5519–5537, https://doi.org/10.5194/tc-17-5519-2023, https://doi.org/10.5194/tc-17-5519-2023, 2023
Short summary
Short summary
To alleviate tedious manual image annotations for training deep learning (DL) models in floe instance segmentation, we employ a classical image processing technique to automatically label floes in images. We then apply a DL semantic method for fast and adaptive floe instance segmentation from high-resolution airborne and satellite images. A post-processing algorithm is also proposed to refine the segmentation and further to derive acceptable floe size distributions at local and global scales.
Laust Færch, Wolfgang Dierking, Nick Hughes, and Anthony P. Doulgeris
The Cryosphere, 17, 5335–5355, https://doi.org/10.5194/tc-17-5335-2023, https://doi.org/10.5194/tc-17-5335-2023, 2023
Short summary
Short summary
Icebergs in open water are a risk to maritime traffic. We have compared six different constant false alarm rate (CFAR) detectors on overlapping C- and L-band synthetic aperture radar (SAR) images for the detection of icebergs in open water, with a Sentinel-2 image used for validation. The results revealed that L-band gives a slight advantage over C-band, depending on which detector is used. Additionally, the accuracy of all detectors decreased rapidly as the iceberg size decreased.
Graig Sutherland, Victor de Aguiar, Lars-Robert Hole, Jean Rabault, Mohammed Dabboor, and Øyvind Breivik
The Cryosphere, 16, 2103–2114, https://doi.org/10.5194/tc-16-2103-2022, https://doi.org/10.5194/tc-16-2103-2022, 2022
Short summary
Short summary
The marginal ice zone (MIZ), which is the transition region between the open ocean and the dense pack ice, is a very dynamic region comprising a mixture of ice and ocean conditions. Using novel drifters deployed in various ice conditions in the MIZ, several material transport models are tested with two operational ice–ocean prediction systems. A new general transport equation, which uses both the ice and ocean solutions, is developed that reduces the error in drift prediction for our case study.
Joey J. Voermans, Qingxiang Liu, Aleksey Marchenko, Jean Rabault, Kirill Filchuk, Ivan Ryzhov, Petra Heil, Takuji Waseda, Takehiko Nose, Tsubasa Kodaira, Jingkai Li, and Alexander V. Babanin
The Cryosphere, 15, 5557–5575, https://doi.org/10.5194/tc-15-5557-2021, https://doi.org/10.5194/tc-15-5557-2021, 2021
Short summary
Short summary
We have shown through field experiments that the amount of wave energy dissipated in landfast ice, sea ice attached to land, is much larger than in broken ice. By comparing our measurements against predictions of contemporary wave–ice interaction models, we determined which models can explain our observations and which cannot. Our results will improve our understanding of how waves and ice interact and how we can model such interactions to better forecast waves and ice in the polar regions.
Cyril Palerme and Malte Müller
The Cryosphere, 15, 3989–4004, https://doi.org/10.5194/tc-15-3989-2021, https://doi.org/10.5194/tc-15-3989-2021, 2021
Short summary
Short summary
Methods have been developed for calibrating sea ice drift forecasts from an operational prediction system using machine learning algorithms. These algorithms use predictors from sea ice concentration observations during the initialization of the forecasts, sea ice and wind forecasts, and some geographical information. Depending on the calibration method, the mean absolute error is reduced between 3.3 % and 8.0 % for the direction and between 2.5 % and 7.1 % for the speed of sea ice drift.
Yu Zhang, Tingting Zhu, Gunnar Spreen, Christian Melsheimer, Marcus Huntemann, Nick Hughes, Shengkai Zhang, and Fei Li
The Cryosphere Discuss., https://doi.org/10.5194/tc-2021-85, https://doi.org/10.5194/tc-2021-85, 2021
Revised manuscript not accepted
Short summary
Short summary
We developed an algorithm for ice-water classification using Sentinel-1 data during melting seasons in the Fram Strait. The proposed algorithm has the OA of nearly 90 % with STD less than 10 %. The comparison of sea ice concentration demonstrate that it can provide detailed information of sea ice with the spatial resolution of 1km. The time series shows the average June to September sea ice area does not change so much in 2015–2017 and 2019–2020, but it has a significant decrease in 2018.
Joey J. Voermans, Jean Rabault, Kirill Filchuk, Ivan Ryzhov, Petra Heil, Aleksey Marchenko, Clarence O. Collins III, Mohammed Dabboor, Graig Sutherland, and Alexander V. Babanin
The Cryosphere, 14, 4265–4278, https://doi.org/10.5194/tc-14-4265-2020, https://doi.org/10.5194/tc-14-4265-2020, 2020
Short summary
Short summary
In this work we demonstrate the existence of an observational threshold which identifies when waves are most likely to break sea ice. This threshold is based on information from two recent field campaigns, supplemented with existing observations of sea ice break-up. We show that both field and laboratory observations tend to converge to a single quantitative threshold at which the wave-induced sea ice break-up takes place, which opens a promising avenue for operational forecasting models.
Cited articles
Andersson, T. R., Hosking, J. S., Pérez-Ortiz, M., Paige, B., Elliott, A., Russell, C., Law, S., Jones, D. C., Wilkinson, J., Phillips, T., Byrne, J., Tietsche, S., Sarojini, B. B., Blanchard-Wrigglesworth, E., Aksenov, Y., Downie, R., and Shuckburgh, E.: Seasonal Arctic sea ice forecasting with probabilistic deep learning, Nat. Commun., 12, https://doi.org/10.1038/s41467-021-25257-4, 2021. a, b, c, d
Batrak, Y. and Müller, M.: On the warm bias in atmospheric reanalyses induced by the missing snow over Arctic sea-ice, Nat. Commun., 10, https://doi.org/10.1038/s41467-019-11975-3, 2019. a
Blair, B., Müller, M., Palerme, C., Blair, R., Crookall, D., Knol-Kauffman, M., and Lamers, M.: Coproducing sea ice predictions with stakeholders using simulation, Weather Clim. Soc., 14, 399–413, https://doi.org/10.1175/wcas-d-21-0048.1, 2022. a
Bommer, P., Kretschmer, M., Hedström, A., Bareeva, D., and Höhne, M. M. C.: Finding the right XAI method – A Guide for the Evaluation and Ranking of Explainable AI Methods in Climate Science, arXiv [preprint], https://doi.org/10.48550/arXiv.2303.00652, 2023. a
Cavalieri, D., Parkinson, C., Gloersen, P., and Zwally, H. J.: Sea Ice Concentrations from Nimbus-7 SMMR and DMSP SSM/I-SSMIS Passive Microwave Data. (NSIDC-0051, Version 1), NASA National Snow and Ice Data Center Distributed Active Archive Center [data set], https://doi.org/10.5067/8GQ8LZQVL0VL, 1996. a
Comiso, J. C., Meier, W. N., and Gersten, R.: Variability and trends in the Arctic Sea ice cover: results from different techniques, J. Geophys. Res.-Oceans, 122, 6883–6900, https://doi.org/10.1002/2017jc012768, 2017. a
Dinessen, F., Hackett, B., and Kreiner, M. B.: Product User Manual For Regional High Resolution Sea Ice Charts Svalbard and Greenland Region, Tech. rep., Norwegian Meteorological Institute, https://doi.org/10.48670/moi-00128, 2020. a
Ebert-Uphoff, I. and Hilburn, K.: Evaluation, tuning, and interpretation of neural networks for working with images in meteorological applications, B. Am. Meteorol. Soc., 101, E2149–E2170, https://doi.org/10.1175/bams-d-20-0097.1, 2020. a
Eguíluz, V. M., Fernández-Gracia, J., Irigoien, X., and Duarte, C. M.: A quantitative assessment of Arctic shipping in 2010–2014, Sci. Rep.-UK, 6, https://doi.org/10.1038/srep30682, 2016. a
Fritzner, S., Graversen, R., and Christensen, K. H.: Assessment of high-resolution dynamical and machine learning models for prediction of sea ice concentration in a regional application, J. Geophys. Res.-Oceans, 125, https://doi.org/10.1029/2020jc016277, 2020. a, b, c
Germann, U. and Isztar, Z.: Scale-dependence of the predictability of precipitation from continental radar images. Part I: Description of the methodology, Mon. Weather Rev., 130, 2859–2873, https://doi.org/10.1175/1520-0493(2002)130<2859:sdotpo>2.0.co;2, 2002. a
Goessling, H. F., Tietsche, S., Day, J. J., Hawkins, E., and Jung, T.: Predictability of the Arctic sea ice edge, Geophys. Res. Lett., 43, 1642–1650, https://doi.org/10.1002/2015gl067232, 2016. a, b
Grigoryev, T., Verezemskaya, P., Krinitskiy, M., Anikin, N., Gavrikov, A., Trofimov, I., Balabin, N., Shpilman, A., Eremchenko, A., Gulev, S., Burnaev, E., and Vanovskiy, V.: Data-driven short-term daily operational sea ice regional forecasting, Remote Sens.-Basel, 14, https://doi.org/10.3390/rs14225837, 2022. a, b, c, d, e, f
Gunnarsson, B.: Recent ship traffic and developing shipping trends on the Northern Sea Route–Policy implications for future arctic shipping, Mar. Policy, 124, 104369, https://doi.org/10.1016/j.marpol.2020.104369, 2021. a
He, K., Zhang, X., Ren, S., and Sun, J.: Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification, arXiv [preprint], https://doi.org/10.48550/arXiv.1502.01852, 2015. a
Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D., Simmons, A., Soci, C., Abdalla, S., Abellan, X., Balsamo, G., Bechtold, P., Biavati, G., Bidlot, J., Bonavita, M., De Chiara, G., Dahlgren, P., Dee, D., Diamantakis, M., Dragani, R., Flemming, J., Forbes, R., Fuentes, M., Geer, A., Haimberger, L., Healy, S., Hogan, R. J., Hólm, E., Janisková, M., Keeley, S., Laloyaux, P., Lopez, P., Lupu, C., Radnoti, G., de Rosnay, P., Rozum, I., Vamborg, F., Villaume, S., and Thépaut, J.-N.: The ERA5 global reanalysis, Q. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803, 2020. a
Hunke, E. C., Lipscomb, W. H., Turner, A. K., Jeffery, N., and Elliott, S.: CICE: the Los Alamos Sea Ice Model Documentation and Software User's Manual Version 5.1 LA-CC-06-012, techreport, Los Alamos National Laboratory, Los Alamos NM 87545, https://www.osti.gov/biblio/1364126 (last access: 15 September 2023), 2015. a
Johnson, S. J., Stockdale, T. N., Ferranti, L., Balmaseda, M. A., Molteni, F., Magnusson, L., Tietsche, S., Decremer, D., Weisheimer, A., Balsamo, G., Keeley, S. P. E., Mogensen, K., Zuo, H., and Monge-Sanz, B. M.: SEAS5: the new ECMWF seasonal forecast system, Geosci. Model Dev., 12, 1087–1117, https://doi.org/10.5194/gmd-12-1087-2019, 2019. a
Kingma, D. P. and Ba, J.: Adam: A Method for Stochastic Optimization, arXiv [preprint], https://doi.org/10.48550/arXiv.1412.6980, 2014. a
Kucik, A. and Stokholm, A.: AI4SeaIce: selecting loss functions for automated SAR sea ice concentration charting, Sci. Rep.-UK, 13, https://doi.org/10.1038/s41598-023-32467-x, 2023. a, b
Kvanum, Are Frode: AreFrode/Developing_ice_chart_deep_learning_predictions: Developing a deep learning forecasting system for short-term and high-resolution prediction of sea ice concentration (Latest), Zenodo [data set], https://doi.org/10.5281/zenodo.17121457, 2025. a
Kwok, R.: Arctic sea ice thickness, volume, and multiyear ice coverage: losses and coupled variability (1958–2018), Environ. Res. Lett., 13, 105005, https://doi.org/10.1088/1748-9326/aae3ec, 2018. a
Lavergne, T., Sørensen, A. M., Kern, S., Tonboe, R., Notz, D., Aaboe, S., Bell, L., Dybkjær, G., Eastwood, S., Gabarro, C., Heygster, G., Killie, M. A., Brandt Kreiner, M., Lavelle, J., Saldo, R., Sandven, S., and Pedersen, L. T.: Version 2 of the EUMETSAT OSI SAF and ESA CCI sea-ice concentration climate data records, The Cryosphere, 13, 49–78, https://doi.org/10.5194/tc-13-49-2019, 2019. a
Liu, Q., Zhang, R., Wang, Y., Yan, H., and Hong, M.: Short-term daily prediction of sea ice concentration based on deep learning of gradient loss function, Frontiers in Marine Science, 8, https://doi.org/10.3389/fmars.2021.736429, 2021a. a, b
Liu, Y., Bogaardt, L., Attema, J., and Hazeleger, W.: Extended range Arctic sea ice forecast with convolutional long-short term memory networks, Mon. Weather Rev., 149, 1673–1693, https://doi.org/10.1175/mwr-d-20-0113.1, 2021b. a, b, c
Long, J., Shelhamer, E., and Darrell, T.: Fully convolutional networks for semantic segmentation, in: 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), IEEE, Boston, MA, USA, 7–12 June 2015, 7-12, 3431–3440, https://doi.org/10.1109/cvpr.2015.7298965, 2015. a
Metzger, E. J., Smedstad, O. M., Thoppil, P., Hurlburt, H., Cummings, J., Walcraft, A., Zamudio, L., Franklin, D., Posey, P., Phelps, M., Hogan, P., Bub, F., and DeHaan, C.: US Navy operational global ocean and Arctic ice prediction systems, Oceanography, 27, 32–43, https://doi.org/10.5670/oceanog.2014.66, 2014. a
Müller, M., Batrak, Y., Kristiansen, J., Køltzow, M. A. Ø., Noer, G., and Korosov, A.: Characteristics of a convective-scale weather forecasting system for the European Arctic, Mon. Weather Rev., 145, 4771–4787, https://doi.org/10.1175/mwr-d-17-0194.1, 2017. a, b, c
Müller, M., Knol-Kauffman, M., Jeuring, J., and Palerme, C.: Arctic shipping trends during hazardous weather and sea-ice conditions and the Polar Code's effectiveness, npj Ocean Sustainability, 2, https://doi.org/10.1038/s44183-023-00021-x, 2023. a, b
Nair, V. and Hinton, G.: Rectified Linear Units Improve Restricted Boltzmann Machines Vinod Nair, in: ICML'10: Proceedings of the 27th International Conference on International Conference on Machine Learning, Haifa, Israel, 21–24 June 2010, 27, 807–814, https://dl.acm.org/doi/10.5555/3104322.3104425 (last access: 15 September 2023), 2010. a
Norwegian Meteorological Institute: met.no Barents-2.5km Files, Norwegian Meteorological Institute [data set], https://thredds.met.no/thredds/catalog/barents25km_files/catalog.html, last access: 13 May 2024, 2024a. a
Norwegian Meteorological Institute: AROME Arctic archive, Norwegian Meteorological Institute [data set], https://thredds.met.no/thredds/catalog/aromearcticarchive/catalog.html, last access: 13 May 2024, 2024b. a
Norwegian Meteorological Institute: Sea Ice Concentration (OSI-401-b) Files, Norwegian Meteorological Institute [data set], https://thredds.met.no/thredds/catalog/osisaf/met.no/ice/conc/catalog.html, last access: 13 May 2024, 2024c. a
Notz, D., D”orr, J., Bailey, D. A., Blockley, E., Bushuk, M., Debernard, J. B., Dekker, E., DeRepentigny, P., Docquier, D., Fučkar, N. S., Fyfe, J. C., Jahn, A., Holland, M., Hunke, E., Iovino, D., Khosravi, N., Madec, G., Massonnet, F., O'Farrell, S., Petty, A., Rana, A., Roach, L., Rosenblum, E., Rousset, C., Semmler, T., Stroeve, J., Toyoda, T., Tremblay, B., Tsujino, H., Vancoppenolle, M. and SIMIP Community: Arctic sea ice in CMIP6, Geophys. Res. Lett., 47, https://doi.org/10.1029/2019gl086749, 2020. a
Notz, D. and Marotzke, J.: Observations reveal external driver for Arctic sea-ice retreat, Geophys. Res. Lett., 39, https://doi.org/10.1029/2012gl051094, 2012. a
Ólason, E., Boutin, G., Korosov, A., Rampal, P., Williams, T., Kimmritz, M., Dansereau, V., and Samaké, A.: A new brittle rheology and numerical framework for large-scale sea-ice models, J. Adv. Model. Earth Sy., 14, https://doi.org/10.1029/2021ms002685, 2022. a
OSI SAF: Global Sea Ice Concentration (netCDF) – DMSP, EUMETSAT SAF on Ocean and Sea Ice [data set], https://doi.org/10.15770/EUM_SAF_OSI_NRT_2004, 2017. a
Owens, R. and Hewson, T.: ECMWF Forecast User Guide, ECMWF, https://doi.org/10.21957/M1CS7H, 2018. a
Palerme, C., Müller, M., and Melsom, A.: An intercomparison of verification scores for evaluating the sea ice edge position in seasonal forecasts, Geophys. Res. Lett., 46, 4757–4763, https://doi.org/10.1029/2019gl082482, 2019. a, b
Palerme, C., Lavergne, T., Rusin, J., Melsom, A., Brajard, J., Kvanum, A. F., Macdonald Sørensen, A., Bertino, L., and Müller, M.: Improving short-term sea ice concentration forecasts using deep learning, The Cryosphere, 18, 2161–2176, https://doi.org/10.5194/tc-18-2161-2024, 2024. a, b, c, d
Rainville, L., Wilkinson, J., Durley, M. E. J., Harper, S., DiLeo, J., Doble, M. J., Fleming, A., Forcucci, D., Graber, H., Hargrove, J. T., Haverlack, J., Hughes, N., Hembrough, B., Jeffries, M. O., Lee, C. M., Mendenhall, B., McCormmick, D., Montalvo, S., Stenseth, A., Shilling, G. B., Simmons, H. L., Toomey, J. E., and Woods, J.: Improving situational awareness in the Arctic ocean, Frontiers in Marine Science, 7, https://doi.org/10.3389/fmars.2020.581139, 2020. a
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, 2016. a
Ren, Y., Li, X., and Zhang, W.: A data-driven deep learning model for weekly sea ice concentration prediction of the pan-Arctic during the melting season, IEEE T. Geosci. Remote, 60, 1–19, https://doi.org/10.1109/tgrs.2022.3177600, 2022. a, b
Ronneberger, O., Fischer, P., and Brox, T.: U-Net: convolutional networks for biomedical image segmentation, in: Lecture Notes in Computer Science, edited by: Nassir, N., Joachim, H., William, W. M., and Alejandro, F. F., Springer International Publishing, https://doi.org/10.1007/978-3-319-24574-4_28, 234–241, 2015. a, b, c, d
Röhrs, J., Gusdal, Y., Rikardsen, E. S. U., Durán Moro, M., Brændshøi, J., Kristensen, N. M., Fritzner, S., Wang, K., Sperrevik, A. K., Idžanović, M., Lavergne, T., Debernard, J. B., and Christensen, K. H.: Barents-2.5km v2.0: an operational data-assimilative coupled ocean and sea ice ensemble prediction model for the Barents Sea and Svalbard, Geosci. Model Dev., 16, 5401–5426, https://doi.org/10.5194/gmd-16-5401-2023, 2023. a, b, c, d
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 Sci., 8, 633–656, https://doi.org/10.5194/os-8-633-2012, 2012. a, b
Serreze, M. C. and Barry, R. G.: Processes and impacts of Arctic amplification: a research synthesis, Global Planet. Change, 77, 85–96, https://doi.org/10.1016/j.gloplacha.2011.03.004, 2011. a
Serreze, M. C. and Meier, W. N.: The Arctic's sea ice cover: trends, variability, predictability, and comparisons to the Antarctic, Ann. NY Acad. Sci., 1436, 36–53, https://doi.org/10.1111/nyas.13856, 2019. a
Stocker, A. N., Renner, A. H. H., and Knol-Kauffman, M.: Sea ice variability and maritime activity around Svalbard in the period 2012–2019, Sci. Rep.-UK, 10, https://doi.org/10.1038/s41598-020-74064-2, 2020. a, b
Toms, B. A., Barnes, E. A., and Ebert-Uphoff, I.: Physically interpretable neural networks for the geosciences: applications to Earth system variability, J. Adv. Model. Earth Sy., 12, https://doi.org/10.1029/2019ms002002, 2020. a
Veland, S., Wagner, P., Bailey, D., Everett, A., Goldstein, M., Hermann, R., Hjort-Larsen, T., Hovelsrud, G., Hughes, N., Kjøl, A., Li, X., Lynch, A., Müller, M., Olsen, J., Palerme, C., Pedersen, J. L., Rinaldo, Ø., Stephenson, S., and Storelvmo, T.: Knowledge needs in sea ice forecasting for navigation in Svalbard and the High Arctic, Tech. Rep. NF-rapport 4/2021, Svalbard Strategic Grant, Svalbard Science Forum, https://doi.org/10.13140/RG.2.2.11169.33129, 2021. a
Wang, L., Scott, K., and Clausi, D.: Sea ice concentration estimation during freeze-up from SAR imagery using a convolutional neural network, Remote Sens.-Basel, 9, 408, https://doi.org/10.3390/rs9050408, 2017. a
Williams, T. D., Bennetts, L. G., Squire, V. A., Dumont, D., and Bertino, L.: Wave–ice interactions in the marginal ice zone. Part 1: Theoretical foundations, Ocean Model., 71, 81–91, https://doi.org/10.1016/j.ocemod.2013.05.010, 2013. a
Williams, T., Korosov, A., Rampal, P., and Ólason, E.: Presentation and evaluation of the Arctic sea ice forecasting system neXtSIM-F, The Cryosphere, 15, 3207–3227, https://doi.org/10.5194/tc-15-3207-2021, 2021. a, b
Wu, Y. and He, K.: Group Normalization, arXiv [preprint], https://doi.org/10.48550/arXiv.1803.08494, 2018. a
Yu, X., Rinke, A., Dorn, W., Spreen, G., Lüpkes, C., Sumata, H., and Gryanik, V. M.: Evaluation of Arctic sea ice drift and its dependency on near-surface wind and sea ice conditions in the coupled regional climate model HIRHAM–NAOSIM, The Cryosphere, 14, 1727–1746, https://doi.org/10.5194/tc-14-1727-2020, 2020. a
Zampieri, L., Goessling, H. F., and Jung, T.: Predictability of Antarctic sea ice edge on subseasonal time scales, Geophys. Res. Lett., 46, 9719–9727, https://doi.org/10.1029/2019gl084096, 2019. a
Zhang, L., Shi, Q., Leppäranta, M., Liu, J., and Yang, Q.: Estimating winter Arctic sea ice motion based on random forest models, Remote Sens.-Basel, 16, 581, https://doi.org/10.3390/rs16030581, 2024. a
Zhang, Z., Luo, P., Loy, C. C., and Tang, X.: Facial landmark detection by deep multi-task learning, in: Computer Vision – ECCV 2014, edited by: Fleet, D., Pajdla, T., Schiele, B., and Tuytelaars, T., Springer International Publishing, https://doi.org/10.1007/978-3-319-10599-4_7, 94–108, 2014. a
Short summary
Recent studies have shown that machine learning models are effective at predicting sea ice concentration, yet few have explored the development of such models in an operational context. In this study, we present the development of a machine learning forecasting system which can predict sea ice concentration at 1 km resolution up to 3 d ahead using real-time operational data. The developed forecasts predict the sea ice edge position with better accuracy than physical and baseline forecasts.
Recent studies have shown that machine learning models are effective at predicting sea ice...
