This website uses cookies, which are small text files that are used to make websites work more effectively. In order to continue using this website, you will need to accept the use of cookies.

Results show that subseasonal sea ice predictions are in an early stage, although skilful predictions 1.5 months ahead are already possible. The position of the sea ice edge is a key parameter for potential forecast users, such as Arctic mariners. However, until now little was known about the ability of current operational subseasonal forecast systems to predict the evolution of the ice edge. Sea ice forecasts are of critical importance to manage the opportunities and risks that come with increasing socioeconomic activities in the rapidly changing Arctic, which, despite the reduction of the sea ice cover, remains an extreme environment.

The demand of reliable forecasts that describe the evolution of the sea ice from days to months in advance has substantially grown in the last decades. Climate projections suggest that the retreat of the Arctic sea ice will lead to an ice-free Arctic Ocean in late summer by the second half of this century if the anthropogenic emissions of greenhouse gases will not be reduced. This scenario fosters the wide economic interests in the region of both private and public actors. Examples are the increasing investigations for new commercial navigation routes, the expansion of fossil fuel and mineral extraction and the development of tourism. Sea ice forecasts are fundamental tools to manage the risks associated with these activities in the Arctic, which, despite the observed sea ice decline, remains an extreme environment.

Operational forecast centers around the world recognize the need of a more accurate representation of the Earth’s weather and climate in their forecast products and are therefore moving towards using fully coupled forecast models. This advanced model configuration goes beyond that typically employed in the classical weather forecasting and is characterized by a global ocean, sea ice and land model that actively interacts with the atmospheric model. The forecast models are initialized with in-situ and satellite observations to obtain a realistic representation of the current state. These new forecasts products have the potential to target events beyond the classical weather timescale and to predict the sea ice evolution from weeks to months in advance.

A database of forecasts is collected in the frame of the Subseasonal to Seasonal (S2S) Prediction Project, a joint initiative of the World Climate Research Program (WCRP) and to the World Weather Research Program (WWRP) that aims to improve the forecast skill and the understanding on the S2S timescale. This database represents a unique occasion for a thorough assessment of state‐of‐the‐art operational predictions of the Arctic sea ice at the subseasonal time scale. Beside quasi-real time forecasts, reforecasts are available for the past two decades and cover the whole seasonal cycle.

“Our research represents the first overview of the subseasonal skill of state‐of‐the‐art coupled forecast systems in predicting the sea ice edge in the Arctic.” said Mr Lorenzo Zampieri, a doctoral student at the Alfred Wegener Institute who led the research. “The comparison of the S2S sea ice forecasts with the satellite observations has been carried out by focusing on the correctness of the sea ice edge position, which is a critical information for final users. The forecast systems contributing to the S2S database show a surprisingly large range of skill, with the best system producing skillful forecasts up to 45 days in advance.”

Spatial Probability Score

Image from Zampieri et al. (2018). Annual‐mean skill in terms of the SPS (Spatial Probability Score) of the different forecast systems (colored‐solid lines) and the climatological benchmark forecast (gray‐solid line) in predicting the Arctic sea ice edge as a function of lead time. Results have been averaged over the common reforecast period 1999–2010. Predictions with SPS values smaller than the climatological value ≈0.55·106 km2) can be considered skillful. The shading and dashed lines indicate ∼95% confidence intervals, based on standard errors obtained from the twelve individual annual means. Note that the CMA forecast system is not depicted given that its large errors lie outside of the range shown. ECMWF Pres. is based on the predecessor ECMWF system, the main difference being that sea ice was not simulated dynamically but prescribed based on a combination of persistence and climatology. SPS = Spatial Probability Score; NCEP = National Centers for Environmental Prediction; CMA = China Meteorological Administration; MF = Météo‐France; ECMWF = European Centre for Medium‐Range Weather Forecasts; UKMO = UK Met Office; KMA = Korea Meteorological Administration.

Results reveal that large errors are introduced already at the data assimilation stage, which is the procedure that adjusts the forecast initial state to the observations before each simulation. This suggests that the operational centers could improve their data assimilation approach for sea ice, atmospheric and ocean initial conditions. The study also point to the need for a further characterization of sea ice initial state beyond the simple sea ice concentration. In this context, new thickness measurements from satellite remote sensing will be fundamental to boost the prediction capabilities of subseasonal and seasonal forecast systems in the near future.


Zampieri, L., Goessling, H. F., & Jung, T. (2018). Bright prospects for Arctic sea ice prediction on subseasonal time scales. Geophysical Research Letters, 45, 9731–9738.