An examination of the relationship between the Southern Annular Mode and Antarctic surface air temperatures in the CMIP5 historical runs

Strong relationships exist between the Southern Annular Mode (SAM) and surface air temperature (SAT) across much of Antarctica. Changes in the SAM will have a profound influence on future Antarctic climate so it is important that the models used to predict climate change can accurately reproduce current SAM–SAT relationships. We analyse data from 50 Climate Model Intercomparison Project (CMIP) 5 models to assess how well they reproduce the observed mean and variability of annual and seasonal SAM–SAT relationships at six Antarctic stations. Overall, the models do better at reproducing these relationships when meridional flow has its largest influence on SAT, doing best (worst) in winter (autumn and summer). They are generally unable to replicate existing seasonal cycles in the strength of the SAM–SAT relationship and show much less spatial and especially temporal variability in the strength of these relationships than is observed. Using an estimate of intrinsic variability to quantify the skill of the CMIP5 models, their average ability to successfully replicate a seasonal SAM–SAT relationship at the six locations studied ranges from 16 % in autumn to 32 % in winter. The mean success rate of a single model across all four seasons is 24 %, ranging from 8 to 38 % (compared to a ‘perfect model’ with 46 %). Analysing the different atmospheric circulation patterns associated with extreme SAM–SAT correlations in the models demonstrates the importance of correctly reproducing both the climatological mean and variability of the planetary longwaves at Southern Hemisphere high-latitudes (particularly wave-number 3), in order to accurately reproduce observed SAM–SAT relationships across Antarctica.


Publication status:
Authors: Marshall, Gareth J., Bracegirdle, Thomas J.

On this site: Gareth Marshall, Thomas Bracegirdle
1 September, 2015
Climate Dynamics / 45
Digital Object Identifier (DOI):