Extended U.S. Tornado Outbreak During Late May 2019: A Forecast of Opportunity

Abstract

The second half of May 2019 was an unusually active period for tornadic thunderstorms across the U.S. Great Plains, Midwest, and lower Great Lakes. While this period typically coincides with the peak climatological frequency of tornadoes, preliminary reports of tornadoes were over triple the expected 30‐year average. Multiple‐day outbreaks of tornadoes are not unprecedented in the United States; however, this event was perhaps the first to be forecast at subseasonal lead times (3–4 weeks) by the Extended Range Tornado Activity Forecast team. This forecast of opportunity was driven, in part, by anomalous convective forcing in portions of the tropical Indian and Pacific Oceans, causing subsequent changes in Northern Hemisphere atmospheric angular momentum. This manuscript analyzes the evolution of hemispheric‐scale circulation features leading up to the event, examines teleconnection processes known to influence U.S. tornadoes, and provides insights into the forecast process at subseasonal lead times.

Hybrid Prediction of Weekly Tornado Activity Out to Week 3: Utilizing Weather Regimes

“…Additionally, tornadoes are more likely to occur in Phases 1–4 of the global wind oscillation (GWO) when atmospheric angular momentum is low (Gensini & Marinaro, 2016; Moore, 2018). Statistical models have been developed based on these climate modes to predict severe storm activity on the S2S time scale (Baggett et al, 2018; Gensini et al, 2019; Lepore et al, 2017). However, the ENSO, MJO, and GWO explain limited variability of tornado activity.…”

Madden–Julian Oscillation–Induced Suppression of Northeast Pacific Convection Increases U.S. Tornadogenesis

“…This severe weather forecast is based on synoptic-scale atmospheric instability [e.g., convective available potential energy (CAPE) and low-level wind shear (LLWS)] from numerical weather forecast models and observations. To extend the current forecast lead time for tornadogenesis to subseasonal time scales (i.e., 14-30 days of lead time), several studies have explored a potential link between U.S. tornado activity and the Madden-Julian oscillation (MJO; Madden and Julian 1972), which is defined by the convective activity over the Maritime Continent propagating eastward around the equatorial tropics with a 30-90-day period (e.g., Thompson and Roundy 2013; Barrett and Gensini 2013; Barrett and Henley 2015;Gensini and Marinaro 2016;Tippett 2018;Baggett et al 2018;Moore and McGuire 2020;Gensini et al 2019). Thompson and Roundy (2013) found that in March-May, the chances of violent tornado outbreaks in the United States are maximized during MJO phase 2, when the convection over the Indian Ocean is enhanced (Fig.…”

“…Recent studies also explored the impact of interactive relationship between MJO and global wind oscillation (GWO), which is defined by the global atmospheric angular momentum, on the U.S. tornado activity (Gensini et al 2019;Moore and McGuire 2020). Moore and McGuire (2020), for instance, explored the interactive impact of GWO and MJO on U.S. tornado days during different seasons.…”

“…Moore and McGuire (2020), for instance, explored the interactive impact of GWO and MJO on U.S. tornado days during different seasons. Additionally, Gensini et al (2019) suggested that MJO and GWO events can produce favorable atmospheric conditions for U.S. tornadic storms in the United States. Baggett et al (2018) further showed useful forecast skills of the MJO-induced tornadic environmental parameters (i.e., CAPE and low-level storm relative helicity) within 45-day lead time using the MJO index.…”

Introduction to Special Collection: “Bridging Weather and Climate: Subseasonal‐to‐Seasonal (S2S) Prediction”

“…For example, the occurrence and frequency of tornado and hail activity in the central United States can be skillfully predicted weeks in advance using the MJO (Baggett et al, ). Similarly, a case study of the May 2019 tornado outbreak over the central United States shows that the favorable supercell conditions were probabilistically predicted nearly 4 weeks in advance (Gensini et al, ). Beyond severe weather, the occurrence of increased periods of cloud‐to‐ground lightning strikes was shown to show skill for lead times of up to 15 days (Tippett and Koshak, ).…”