资源环境科技发展态势分析平台

The North American continent experienced an extremely anomalous dipole climate in the 2014/2015 winter with record-breaking cold temperature anomalies in the east and warm anomalies in the west. The seasonable predictability of this extreme climate has not been understood. By using a large ensemble of model simulations with the Community Atmospheric Model (CAM4) and cluster analysis under observed sea surface temperature (SST) and sea ice conditions, we show that one of the simulation clusters can reproduce the observed dipole climate in this winter, but it is with low probability. This is in contrast to the two winters of 2013/2014 and 2015/2016 in which the leading clusters in surface temperature anomalies and the ensemble means are similar to observations in response to SST and sea ice conditions. Since the differences among the simulated clusters are entirely due to atmospheric initial conditions, we conclude that the 2014/2015 winter extreme dipole climate is a low-probability event that is primarily caused by internal atmospheric variability based on the single-model CAM4.