Sea Ice Outlook: Post-Season Highlights

19 December 2014

The Sea Ice Outlook Post-Season Highlights were developed by ARCUS, based on the full report.

  • The Sea Ice Outlook (SIO), a contribution to the Study of Environmental Arctic Change provides an open forum for researchers and others to develop, share, and discuss Arctic sea ice predictions.

  • The SIO is now part of the Sea Ice Prediction Network(SIPN) project. SIPN provides the SIO with a larger group of contributors—including the SIPN leadership team the Sea Ice Outlook Action Team, and the broader SIPN network—as well as more activities (workshops, etc.) that feed into the development of the SIO reports.

  • New content in this year's SIO reports, including this post-season report, includes analysis of the relative skill of various Arctic sea ice prediction models and methods, and an expanded focus on the spatial pattern, probability, and ice-free dates for specific regions—this information is often of more use to stakeholders and decision-makers than a single value for Arctic sea ice extent.

  • This year, we had 28 groups contribute pan-arctic Outlooks and 5 groups contribute regional outlooks, with a total of 88 submissions from June to August. We appreciate all this energy and work from the Arctic sea ice prediction community!

  • Observed Arctic September average ice extent in 2014 was 5.3 million square kilometers, according to National Snow and ice Data Center (NSIDC) estimates

  • This is the 6th lowest extent since satellite observations began in 1979, and close to the downward linear trend line of decreasing Arctic sea ice extent over time. The eight lowest minimum ice extents have occurred in the last eight years, 2007-2014.

  • The observed extent for September 2014 suggests that in the absence of anomalous patterns of weather and wind that results in large ice loss (such as occurred in 2012), sea ice extent will tend to stay near the downward linear trend line.

  • The median Outlook estimates were 4.7 million square kilometers for the June report, 4.8 million square kilometers for the July report, and 5.0 million square kilometers for the August report. Contributions are based on a range of methods: statistical; modeling; subjective information (or "heuristic"); and a mix of methods (e.g., a mix of statistical and modeling methods).

  • Through the season, the Outlook contributions tended to converge—as we would expect with a shorter forecast period—and were closer to the observed value. This convergence is most visible among the statistical and modeling contributions.

  • For the modeling contributions specifically, the later the prediction date, the more confident the predictions. In addition, the inter-model spread is also reduced as the prediction start dates get closer to the month of September. The median value of all the models combined each month was remarkably close to the observed extent of 5.3 million square kilometers.

  • When evaluating the SIO over the past several years (2008-2014) we can compare the SIO predictions to three reference forecasts, all based from 1979 up to but not including the predicted year: 1. the climatological mean (1981-2010), 2. the linear trend (following the downward trend since 1979), and 3. an anomaly persistence forecast (e.g., if the weather is colder than normal today, the forecast is that it will be colder than normal next week). SIO median predictions do considerably better than the climatological mean and slightly better than the linear trend. The anomaly persistence predictions do better than SIO at lead times of 1 to 2 months, but the SIO median predictions do better at a 3-month lead time, which would be considered a "true" seasonal forecast lead time.

  • For modeling methods, SIPN has the following recommendations to improve Arctic sea ice predictions:

    • Better initialization of sea ice conditions, including incorporating more observational data in the sea ice and the ocean models;
    • More advanced model physics (e.g., melt ponds);
    • Better account for initial condition and physics uncertainties;
    • Encourage groups to provide information on what they consider as the greatest source of uncertainty in their simulations.
  • The highlights above are discussed in more depth in the full report.

Citation: Sea Ice Prediction Network. 2014. Arctic Sea Ice Outlook 2014: Post Season Report Highlights. http://www.arcus.org/sipn/sea-ice-outlook/2014/post-season-highlights