57A029
How well can we model the observed decline in the summer Arctic sea-ice extent?
Ann Keen, Helene Hewitt, Jeff Ridley, Alex West
Corresponding author: Ann Keen – ann.keen@metoffice.gov.uk
The September 2009 Arctic sea-ice extent was the third lowest on record, returning to the long-term trend after the exceptionally low ice extent in 2007 and 2008. The HadGEM1 global coupled climate model, which has a relatively sophisticated treatment of sea ice compared with many global climate models, can capture the observed rate of decline of sea-ice extent since 1979, and the model can also reproduce an abrupt reduction in extent of similar magnitude to that observed in summer 2007. After such an event, in HadGEM1 the ice extent recovers to the long-term trend within a few years. However, some coupled climate models show a step change in extent after such abrupt reductions. In order to predict when the Arctic will become ice-free in the summer we need to understand the different model behaviours and the feedbacks driving the recovery or otherwise of the ice. Key factors causing the record low sea-ice extent in 2007 were the unusual and persistent atmospheric circulation pattern over the Arctic during that summer, together with a longer-term ‘preconditioning’ of the ice and ocean due to gradual warming, making regions of the Arctic more prone to melting completely during the summer. In HadGEM1, a low summer sea-ice extent is often associated with a negative MSLP anomaly in the Arctic basin in the early summer, and the resulting reduction in ice extent and thickness can recover over a few years if more ‘normal’ weather conditions return. In this study, results from a number of climate simulations of the HadGEM1 model will be compared with observed changes to the sea ice, focusing on years with especially little sea ice, to identify key mechanisms controlling the ice extent.
Go Back