Advancing Convective Severe Weather Outlooks into Subseasonal-to-Seasonal Timescales

Every year, convective severe weather poses a threat to society - endangering lives and devastating properties. Flash floods, hail, lightning, tornadoes, and severe winds can wreak havoc on an area in a matter of seconds. For example, on April 27^th, 2011 the southeast United States experienced a destructive severe weather event that consisted of large hail and almost 200 tornadoes that killed more than 300 people and cost more than 4 billion dollars in damages. With such extreme impacts, it would be beneficial to have as much lead time as possible to determine when and where convective severe weather regimes are likely to occur. This study aims to identify sources of predictability that foretell a greater likelihood of severe weather at subseasonal-to-seasonal timescales (~3 to 6 weeks). Recent studies have shown that knowledge of the current states of the Madden-Julian Oscillation (MJO) and the Quasi-Biennial Oscillation (QBO) can be used as a potential source of predictability. Using ERA-Interim reanalysis data (1979-2015), we perform a lagged-composite analysis of several severe weather parameters: convective available potential energy, the magnitude of the 0 to 6 km vertical wind shear, and their product (CS06). This analysis is performed during February-June according to the 8 phases of the MJO, parsed by easterly and westerly QBO periods. In our composites, we observe a clear modulation of these parameters out to lags of 6 weeks across all 8 phases of the MJO. Moreover, when we parse the MJO phases into easterly and westerly QBO periods, there is further modulation of when and where severe weather regimes occur. As an example, we examine these parameters for Tuscaloosa, AL which was struck by a devastating EF4 tornado during the aforementioned severe weather event of 2011. The event occurred during a westerly QBO period, 3 to 4 weeks following a strong propagation of the MJO through phases 5 and 6. Consistent with our composites, there was a greater likelihood of anomalously high CS06 (and, by extension, severe weather) for this particular combination of lag weeks, QBO period, and MJO phases. This result alludes to the potential usefulness of the MJO and QBO as predictors for severe weather regimes at subseasonal-to-seasonal time scales, offering an opportunity to improve severe weather outlooks with lead times out to 6 weeks.