The high latitude regions have warmed three times as fast as the global average in recent decades, leading to a dramatic decline in sea-ice extent and thickness. These sea-ice changes vastly impact the polar environment, ecosystem, and local communities - They bring about both risks and opportunities for human activities in polar regions, e.g., scientific missions, tourism, fisheries, shipping, natural resource exploitation and wildlife management. Sea-ice changes may also have an influence extending beyond the polar regions, e.g., in parts of northern mid-latitude continents. Sea-ice prediction on subseasonal-to-seasonal (S2S) timescales benefits society and is a hot topic being researched internationally. Thanks to the recent advances in High-performance computing, model development and initialisation methods, and the emergence of novel observations, the gap between potential and realistic sea-ice predictability is closing in some regions and seasons but still remains significant. The 4SICE project is managed by the Nansen Environmental and Remote Sensing Center in Norway and the Sun Yat-sen University in China. 4SICE aims to improve the capability of S2S sea-ice prediction by improving the model, improving the initialisation method, and using yet untapped observations. The research in the project mostly focuses on the Arctic region, but will also assess the Antarctic which receives less attention. 4SICE sea-ice predictions will support scientific and economic activities in the polar regions.
Sea ice prediction on subseasonal-to-seasonal (S2S) timescales is extremely beneficial for society and is a hot topic of international research. Thanks to the recent advances in model development and initialisation, and the emergence of novel observations, the gap between potential and realistic sea ice predictability is closing in some regions and seasons but overall, it remains very significant. As of today, dynamical sea ice prediction is skilful only up to a few months, while potential sea ice predictability extends well beyond 1-year lead time. 4SICE will address the challenges in improving model initialisation and reducing model bias to significantly enhance dynamical sea ice prediction on S2S timescales. 4SICE will investigate the mechanism, skill and limit of dynamical sea ice prediction in the Arctic and Antarctic, through observations and model data outputs. 4SICE will also investigate the impact of small-scale ocean and sea ice processes on sea ice prediction. 4SICE will implement the assimilation of sea ice drift data, estimate sea ice dynamic parameters with data assimilation (DA), and implement advanced approaches for the non-Gaussian distribution issue of sea ice states in DA in the Norwegian Climate Prediction Model (NorCPM). The COIN DA method will be implemented in the high-resolution regional prediction system SHAPS to improve sea ice initialisation. 4SICE will downscale the global predictions from NorCPM with regional prediction systems SHAPS and SISPS to high-resolution predictions in the Arctic and Antarctic. Furthermore, 4SICE will investigate the impact of enhanced sea ice prediction on seasonal predictions in lower latitudes and study the trans-Arctic maritime accessibility predictability on S2S timescales. 4SICE will also contribute to the Sea Ice Prediction Network (SIPN) and SIPN South, and apply sea ice predictions to support scientific missions and economic activities in the polar regions for end users.