Demonstrating the Utility of Conditional Probabilities of Tornado Damage Rating in the Impact-Based Warning Era

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Wednesday, 5 November 2014
Capitol Ballroom AB (Madison Concourse Hotel)
Bryan T. Smith, NOAA/NWS/NCEP/SPC, Norman, OK; and R. L. Thompson, A. R. Dean, P. T. Marsh, R. Wagenmaker, M. Hudson, and J. T. Ferree
Manuscript (2.1 MB)

Handout (1.8 MB)

The Storm Prediction Center developed conditional probabilities of tornado damage rating from near-storm environment data and radar-based storm-scale characteristics from a 5-year sample of tornadoes (>4,700) reported in the contiguous United States (CONUS) during 2009-2013. The probabilities are derived from filtering tornado segment data by the maximum EF-scale tornado event per hour on a 40-km horizontal grid. Near-storm environment data, consisting primarily of supercell-related convective parameters from hourly objective mesoscale analysis calculated at the Storm Prediction Center (SPC), accompanied each tornado grid-hour event. Convective mode was assigned to each tornado event by manually examining full volumetric WSR-88D data at the beginning time of each event. Peak 0.5 rotational velocity was also manually identified immediately prior to and during the duration of each tornado event.

A best-practices approach for short-term tornado hazard diagnosis considers the use of information from 1) near-storm environment, 2) convective mode, 3) peak 0.5 rotational velocity and 4) supporting or confirming evidence of a tornado as a unifying frame-of-reference. A few examples highlighting the strengths and weaknesses of applying this conditional probability technique will be discussed.