Evaluation of Sudden Stratospheric Warming Event in the Korean Integrated Model Ensemble Simulations

Predictability of the sudden stratospheric warming (SSW) event occurred in January 2013 has been evaluated in the Korean Integrated Model (KIM). Ensemble simulations with 50 members were initialized 15 days before the SSW onset and integrated for 40 days. 7 ensemble members predict the zonal-mean zonal wind reversal from westerly to easterly at 60^oN and 10 hPa between -1 and +5 days of the SSW onset, and those members were classified to SSW group in this study. Compared to the analysis data, the easterly is weaker and maintained shorter in the simulations. In 31 ensemble members, the zonal-mean zonal wind at 60oN reverses but it was not within the specified time window or does not extend down to 10 hPa. 12 ensemble members, which fail to predict zonal-mean zonal wind reversal, were classified to non-SSW group. Ensemble members were averaged for the SSW group and non-SSW group, respectively, and two groups were compared in order to figure out which factors and processes are important and need to be improved for the model to predict SSW better. It is found that planetary wave amplitudes and forcings with zonal wavenumbers of 1 and 2 are larger in the SSW group as shown in previous studies. This signal appears ~10 day before the SSW onset and expands from the troposphere to the stratosphere. In the northern hemisphere mid and high latitudes, biases of the mid tropospheric geopotential height perturbations against the analysis data in the SSW group are similar to those in the non-SSW group. The biases are even greater in the SSW group than in the non-SSW group depending on the forecast time. However, for the geopotential height perturbations with zonal wavenumbers of 1 and 2, the SSW group is more successful to simulate them than the non-SSW group. For the present model to simulate SSW, proper representation of the planetary-scale waves is more likely to be important than the accurate representation of the mean state.