The Stratospheric Sudden Warming Event in February 2018 and its Prediction by a Climate System Model

Abstract

A major stratospheric sudden warming (SSW) event was observed in February 2018 after a 4-year absence since the winter of 2013/2014. Based on the reanalysis data, the polar night jet changed from a very strong state to a moderate state during 12-19 January, and the moderate westerlies directly reversed to easterlies during 5-15 February. The intensified East Asian trough, Alaskan blocking, and East U.S. trough amplified the extratropical climatological wave 2, which propagated upward into the stratosphere, leading to a vortex-splitting SSW event. Predictions of the February 2018 SSW event are explored in hindcasts initialized 0-4 weeks in advance by the Beijing Climate Center Climate System Model. Less than 20% of the 28 ensemble members predict the reversal of [U] 60°N, 10hPa in hindcasts initialized 3 or 4 weeks in advance if a 5-day error is allowed, while this ratio increases to 43% in hindcasts initialized 1 week in advance. Based on the climatological occurrence of SSW events in the forecast system, the maximum deterministic predictable limit of this event is 1-2 weeks in this forecast system. The eddy heat flux and its domination by wave 2 can only be predicted within the predictable time limit. A comparison between hindcast members initialized 2 weeks in advance suggests that the extratropical troughs and blockings are responsible for the upward propagation of waves from the troposphere to the stratosphere. The predictable limit of the stratospheric circulation pattern for the February 2018 SSW, 1-2 weeks, also generalizes to other vortex split SSW events such as the

Predictability of Stratospheric Sudden Warmings in the Beijing Climate Center Forecast System with Statistical Error Corrections

“…To efficiently use and explain the hindcast outputs, statistical error corrections (ECRs) methods based on mathematical modeling are usually necessitated. Rao et al () showed that the predictability of the February 2018 SSW in the BCC_CSM hindcasts was 1 to 2 weeks, which is consistent with the average predictability of SSW events in other models (Karpechko, ; Rao et al, ; Taguchi, , ; Tripathi et al, ). The present paper extends that work by comparing the predictability of 12 SSW events with/without ECRs.…”

“…Every hindcast has four ensemble members, which are produced by a lagged average forecasting strategy with a 6‐hr interval of atmospheric initial conditions ahead. For more details on the model, the reader is directed to Liu et al () and Rao et al (, ).…”

“…The selected SSW events are generally consistent with Butler et al () except that the marginal SSW events in 20 March 2000, 2 January 2002, and 18 January 2003 are excluded in this study (e.g., Hu et al, ). Based on Charlton and Polvani (), Hu et al (), and Rao et al (), half of the SSW events are vortex displacement events (marked by a “D”), and half are vortex split events (marked by an “S”). Table also lists the actual change in the zonal mean zonal wind at 60°N and 10 hPa from day –5 to the day when the easterly reaches its maximum after the SSW onset (d[ U ]).…”

“…Ample evidence has shown that different SSWs are predictable on different timescales in many models. For example, the prediction limit has been stated as around 10 days to 2 weeks for a few SSW events [e.g., February 1989 (Taguchi, ); December 2001 (Hirooka et al, ; Mukougawa et al, ); December 2003 (Hirooka et al, ); January 2013 (Tripathi et al, ); February 2018 (Karpechko et al, ; Rao et al, ); multiple cases (Jung & Leutbecher, ; Karpechko, ; Rao et al, ; Taguchi, , )]. Previous studies found that the predictable limit decreases for several extremely strong SSW events with a large decrease in the zonal mean zonal winds [e.g., February 1979 (Mechoso et al, ); the Southern Hemisphere event in September 2002 (Allen et al, ; Simmons et al, ); January 2009 (Kim et al, ; Kim & Flatau, )].…”

Winter 2018 major sudden stratospheric warming impact on midlatitude mesosphere from microwave radiometer measurements

“…The primary definition of a SSW event provided by the World Meteorological Organization requires a stratosphere temperature increase and an accompanying zonal wind reversal to easterlies at the 10 hPa pressure level (approximately 30 km altitude) and 60 • latitude (WMO, 1978). This definition was broadened and detailed in recent papers (Butler et al, 2015;Butler and Gerber, 2018;Rao et al, 2019). The summarizing paper, where a SSW database is described, was published in Butler et al (2017).…”