Extended Abstract (224K)P5.1
Tornado Outbreak Days: an updated and expanded climatology (1875-2003)
Russell Schneider, NOAA/NCEP/SPC, Norman, OK; and J. T. Schaefer and H. E. Brooks
Tornado Outbreaks are a key focus of the severe weather forecast community. A Tornado Outbreak Day climatology was constructed for the period from 1875 through 2003 using both NOAA and Grazulis (1990) databases. Daily summaries (midnight to midnight) of a variety of tornado and severe weather activity measures were constructed from each database to aid the classification of each day in the 128 year record. The resulting database included both traditional measures such as counts of tornadoes as a function of estimated intensity (F scale) as well as alternative measures such as total path length or integrated Destructive Potential Index. Particular care was taken to track the nuances of each dataset that influences the interpretation of the daily summary statistics. Examples include indications of a “skipping path” within long track tornadoes or combination of a family of tornadoes into a single event or events with unknown characteristics contributing to uncertainty in combined parameter statistics.
The comprehensive database was used to categorize whether each day qualified as a Tornado Outbreak Day and to further categorize outbreak days according to their severity. Initial (objective) criteria for each Outbreak Day severity category (1-6) were established but each day was subsequently analyzed subjectively both individually and within its severity category to reduce biases due to collection era, availability of event statistics, and to resolve disagreements between the two datasets.
The resulting Tornado Outbreak Day database is analyzed to illuminate key characteristics of the events. Preliminary results document the primary Outbreak Day climatological maximum in April, May and June with a shift of the maximum to March, April and May for outbreaks of historic severity (category 4-6). A secondary Outbreak Day maximum is also indicated in late Fall. This seasonal evolution is illustrated in greater detail through the use of “daily” time series plots of Outbreak Day activity.
The data are also used to gage the rarity and importance of Outbreak Days within the overall tornado forecasting problem. Preliminary data suggest that approximately 3.5 percent of all days can be classified as Tornado Outbreak Days, but they account for over 80 percent of all recorded tornado fatalities. Major Outbreaks are observed on less than 0.5 percent of days (approximately 1.6 per year) but accounted for nearly 50 percent of all fatalities.
Poster Session 5, CLIMATOLOGICAL STUDIES OF SEVERE STORMS
Tuesday, 5 October 2004, 3:00 PM-4:30 PM
Next paper