Forecast Skill of the Arctic Sea Ice Outlook 2008–2022

• Published in: Geophysical research letters Vol. 50; no. 6
• Main Authors: Blanchard‐Wrigglesworth, Edward, Bushuk, Mitchell, Massonnet, François, Hamilton, Lawrence C., Bitz, Cecilia M., Meier, Walter N., Bhatt, Uma S.
• Format: Journal Article
• Language: English
• Published: Washington John Wiley & Sons, Inc 28-03-2023; Wiley
• Subjects: Arctic sea ice; Benchmarks; Circulation patterns; Ensemble forecasting; Ice volume; Initial conditions; Mathematical models; Modelling; Sea ice; Sea ice concentrations; Seasonal forecasting; Statistics; Summer; Summer circulation; Summer weather; Weather forecasting; Arctic region
• ISSN: 0094-8276; 1944-8007
• DOI: 10.1029/2022GL102531

We assess the skill of forecasts of Arctic September sea ice in the Sea Ice Outlook over 2008–2022. The multi‐model median June initialized forecast of September sea ice extent (SIE) is slightly more skilled (RMSE = 0.48 million km2) than a damped anomaly forecast, but July and August initialized forecasts (RMSE = 0.52 and 0.36 million km2 respectively) do not beat this benchmark. The skill of individual dynamical and statistical SIE forecasts is lower than the multi‐model median forecast skill. Overall skill is lower than expected from retrospective forecasts. Several forecasts initialized in early September 2021 and 2022 imply physically improbable values. Spatial forecasts of sea ice concentration show multi‐model forecast skill and an improvement in individual forecast skill in recent years. Initial conditions show large spread in sea ice volume and a positive correlation between initialized sea ice volume and September SIE forecast. Summer weather has an impact on forecast error.
Plain Language Summary
We have collected, analyzed, and disseminated seasonal forecasts of September sea ice over 2008–2022. Here, we analyze the skill of these forecasts. We show that individual forecasts of September sea ice extent (SIE) have limited skill, but the median forecast shows skill that is at least as good as a statistical benchmark. Overall the skill is lower than expected from existing retrospective forecasts. Shorter‐term forecasts produced in early September show that many forecasts imply physically unprecedented rates of SIE change. The skill of spatial forecasts from individual models is also limited, but the multi‐model forecast is twice as skilled, showing the value of a multi‐model forecast ensemble. To explore the spread across forecasts we have assessed sea ice initial conditions and found large differences in the sea ice volume used to initialize forecasts, which has an impact on the forecast. We also find that summer weather has a modest impact on forecast error, which tends to underpredict September SIE during summers with circulation patterns that favor high September SIE and vice versa.
Key Points
The skill of September sea ice extent (SIE) forecasts over 2008–2022 is modest. The multi‐model forecast has the highest skill
Sea ice forecast initial conditions are variable. Forecasts initialized with more sea ice volume forecast greater September SIE
Summer atmospheric circulation has an impact on forecast error