Daily probabilities of severe thunderstorms in the United States

Development of a meaningful climatology of rare events, such as severe thunderstorms, is challenging in light of problems associated with collection of the reports, and the trade-off between acquisition of a sufficient sample size and a stable period of data collection. In support of the activities of the Storm Prediction Center and to provide information to emergency managers and the insurance industry, we have used kernel density estimation techniques in space and time to development daily climatological probabilities of a variety of severe thunderstorm hazards, including tornadoes (by damage classification), hail (both any hail and giant hail), and thunderstorm winds (any severe wind and extreme winds). We have developed display techniques to allow users to see animated fields of the probabilities throughout the annual cycle, and to see the annual cycle at any location in the contiguous 48 states.

The development of the daily probabilities has allowed us to see the annual cycle in ways not possible before, as well as to study the 'repeatability' of the annual cycle. It appears, for instance, that the relatively consistent peak in the seasonal cycle of tornadoes in the southern and central Great Plains helps increase public awareness of the threat, leading to fewer high-fatality tornadoes than would be expected otherwise.

See http://www.nssl.noaa.gov/~brooks/threatanim.html for the displays.