Examining Tropospheric Precursors to Sudden Stratospheric Warming Events from an Ensemble Perspective

Sudden Stratospheric Warming (SSW) events are characterized by rapid warming of the high-latitude stratosphere, and are accompanied by a reversal of the 10-hPa zonal mean zonal wind at 60^˚N from westerly to easterly. Temperature and wind anomalies in the stratosphere can progress downward into the troposphere following SSW events. These anomalies can linger near the tropopause for up to 60 days and can have major impacts on tropospheric sensible weather, including increased likelihood of cold air outbreaks and active weather patterns in the mid-latitudes. Therefore, understanding the sources of uncertainty in our ability to predict SSW events can inform the uncertainty in sub-seasonal wintertime forecasts of the troposphere.

This analysis focuses on the 22 February 2008 SSW event, which serves as a predictability case study. Subseasonal forecast data from the European Center for Medium-range Weather Forecasting (ECMWF), available via the WCRP/WWRP Subseasonal-to-Seasonal (S2S) Prediction Project database, was used to examine ensemble forecasts at multiple lead times prior to the SSW. Preliminary results show two periods of anomalous zonal mean meridional eddy heat flux (hereinafter heat flux) in the tropopause regions in the two weeks prior to the SSW. Ensemble members that more accurately predicted the reversal of the zonal mean winds in the stratosphere generally had larger magnitudes of heat flux compared to those members that were not able to predict the SSW. Results readily show planetary-scale Rossby waves contributed a larger percentage of the overall heat flux compared to the synoptic scale. However, the vertical derivative of heat flux, which approximates the vertical convergence of wave activity flux, contributed by synoptic scale waves is on the same order of magnitude as for planetary scale waves in the lower stratosphere. These results suggest a mechanism for synoptic scale waves to directly contribute to modulating the lower stratospheric environment prior to an SSW event. Furthermore, the greatest variability in heat flux among the different ensemble members can be linked to the forecast evolution of a rapidly-deepening cyclone in the North Pacific and an upper-level ridge near Greenland. The importance of these two synoptic-scale features supports the hypothesis that the synoptic-scale played a role in forcing this particular SSW event.

A second case, from October–November 2016, analyzes the impact of two recurving tropical cyclones (TCs) (i.e., Nicole in the Atlantic and Haima in the West Pacific) on the stratospheric flow. The TCs reconfigured the midlatitude jet in late October 2016 and contributed to an increase in the tropopause level zonal mean heat flux. The increased heat flux preceded a deceleration of the stratospheric vortex, with several ECMWF ensemble members predicting an early season SSW event. Similarities and differences between these synoptic features and those seen in the 2008 case will be discussed in relation to our understanding of the dynamical mechanisms behind the (in)ability of numerical weather models to accurately predict the evolution of the stratospheric vortex in each case.