3B.7
Climatology and Case Studies of Northeast Severe Convection with Low-Predictive Skill

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 29 June 2015: 3:00 PM
Salon A-5 (Hilton Chicago)
Matthew Vaughan, University at Albany, SUNY, Albany, NY; and B. Tang and L. F. Bosart

This study objectively identifies severe weather events with low-predictive skill from 1980 to 2013 over the northeast U.S. The database consists of 0600 UTC Storm Prediction Center (SPC) outlooks for the 24-hour period beginning at 1200 UTC to 1200 UTC the next day. Slight risk areas are projected on a 40 x 40 km grid over the northeast U.S. Storm reports of hail, wind, and tornados aggregated over each 24-hour period are used to calculate probability of detection (POD), false alarm ratio (FAR), and critical success index (CSI) skill scores. Events scoring in the lowest 25th percentile of POD and FAR are separated into two categories to construct a climatology consisting of high-impact severe weather cases with low POD (type one) and low-impact cases with high FARs (type two). Type one events have no discernable trend through the 33-year period, whereas type two events decline substantially with time. Both types of events are then subdivided according to background midlevel flow and convective parameters and compared to cases with high-predictive skill, defined as having a CSI in the highest 25th percentile of high-impact cases.

We find that type one events occur more often under low shear conditions than cases with high predictive skill. Furthermore, high-impact events under low-shear conditions have a median CSI of 0.187, while high-impact events under high-shear conditions have a median CSI of 0.137. Additionally, high-impact events occurring under 500-hPa northerly flow have lower predictive skill than high-impact events occurring under other flow regimes. Case studies, segregated by flow regime and convective mode, are conducted to further analyze mesoscale features impacting predictability. Particular emphasis is placed on identifying mesoscale phenomena, such as terrain-channeled flows and surface boundaries. These features may locally enhance convective parameters such as low-level lapse rates, vertical wind shear, and low-level theta-e in the near-storm environment within an otherwise marginal synoptic pattern.