| Description |
The downward influence of stratospheric sudden warmings (SSWs) can create significant tropospheric circulation anomalies that last for several weeks. It is therefore of interest to understand the month of the year during which SSWs are most likely to occur and the controlling factors of their temporal distribution. Conceivably, the distribution is controlled by the interplay between decreasing stratospheric wave driving and weakening stratospheric vortex strength. General circulation models (GCMs) tend to produce their SSW maximum later in winter than observations, which is considered a model deficiency. However, the observed record is short, suggesting that under-sampling of SSWs may contribute to this discrepancy. Here, we study the distribution of SSWs and related events in a long control simulation with a stratosphere resolving GCM. Further, we create a simple statistical model to determine the primary factors controlling the SSW distribution. The model is based on the daily climatological mean, standard deviation, and autocorrelation of stratospheric winds and assumes that the winds follow a normal distribution. Results indicate that we cannot reject the null hypothesis that model and observations stem from the same distribution suggesting that the mid-winter SSW maximum seen in the observations is due to sampling uncertainty. We conclude that the late SSW distribution seen in models is not unrealistic and that it is likely that future observations will show more late winter SSWs. We further find that the statistical model reproduces the seasonal evolution of SSWs well and that the decreasing climatological strength of the vortex is the primary factor in controlling the SSW distribution. This supports the idea that stratospheric winds are approximately normally distributed and that SSWs simply form the tail of this monotonic and unimodal distribution. |
