Title: Advancing Forecasts of Weather Extremes into Subseasonal-to-Seasonal (S2S) Timescales
Abstract: The intrinsic chaos of the atmosphere makes the prediction of extreme weather at subseasonal-to-seasonal (S2S) timescales (~2 to 5 weeks) an inherently difficult task. We demonstrate here that the potential exists to advance forecast lead times of extreme weather events into S2S timescales through knowledge of two of the atmosphere’s most prominent oscillations; the Madden-Julian oscillation (MJO) and the Quasi-biennial oscillation (QBO). We present evidence of a dynamical relationship linking atmospheric rivers (narrow plumes of intense atmospheric moisture that lead to floods and droughts over the west coast) to the MJO and the QBO through modulation of North Pacific blocking using reanalyses and retrospective forecasts of S2S forecast systems, and demonstrate differences in predictability across MJO phases. We conclude by presenting an empirical prediction scheme for anomalous atmospheric river activity based solely on the MJO and QBO and demonstrate skillful subseasonal “forecasts of opportunity” 4+ weeks ahead. This methodology is also shown to produce skillful forecasts of hail and tornado activity over the Plains and Southeast U.S.. With the wide-ranging impacts associated with extreme weather events, even modest gains in the subseasonal prediction of anomalous activity may support early action decision making and benefit numerous sectors of society.