5.3 Seasonal sea ice predictability in a rapidly changing Arctic

Tuesday, 19 May 2009: 9:00 AM
Capitol Ballroom AB (Madison Concourse Hotel)
Marika M. Holland, NCAR, Boulder, CO; and D. Bailey and S. J. Vavrus

Seasonal predictions of Arctic sea ice conditions have typically been based on statistical regression models using a number of different sea ice predictors or on results from ensemble ice model forecasts. However, in the rapidly changing Arctic environment, the predictability characteristics of summer ice cover could undergo important transformations. Here we use coupled climate model simulations to assess the seasonal predictability of the September Arctic ice extent over the 20th-21st century. We examine the role of preconditioning of the ice cover versus intrinsic atmospheric variations in determining the end-of-summer ice extent through both perfect initialization experiments and statistical relationships from standard 20th-21st century coupled climate model integrations. These are complemented by results from ice-only model integrations with specified atmospheric forcing. We find important changes in the relationship between preceding ice and atmospheric conditions and the September ice extent as the Arctic transitions from a perennial to seasonal ice covered state. This has implications for the design of forecasting systems and sea ice observational networks.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner