Kernel Density Estimates of Tornado Occurrence in the United States
Stephen Weinbeck, SUNY, Brockport, NY; and S. Trebes and J. Maliekal
Issuing probabilistic forecasts of the threat of severe weather events within 25 n mi of any location in the United States is one of the objectives of the National Weather Service. Among severe weather events, none is more numerous and dangerous than severe thunderstorms. Many severe thunderstorms produce dangerous tornadoes. Typically 800 tornadoes form in association with severe thunderstorms in the United States in a typical year. Forecasting the threat of tornado occurrence, therefore, is an integral part of this initiative. A thorough understanding of the tornado climatology, especially the density of their occurrence, has the potential to enhance the quality of such forecasts.
In this study, we used data compiled by the Storm Prediction Center (SPC) and estimated the density of tornado occurrence in the eastern two thirds of the United States for the period 1950 to 1995. ArcGIS, the standard desktop Geographical Information System (GIS), and the kernel density approach, an interpolation technique used to generalize locations of specific incidents to an entire area, were used for this purpose. Kernel density was estimated by placing a circle of radius 75 km over each point where a tornado occurred, evaluating the distance from that point to the center of a nearby 80 km x 80 km square, and integrating all such distances associated with that square. By repeating this procedure for all such square within the study regions, density surfaces were created for spring and summer months. These surfaces were then superimposed on maps. Similarly, maps of tornadoes densities were generated for El Niņo, La Niņa, and neutral years. Density maps show that conditions over the equatorial eastern and central Pacific have an impact on the occurrence of tornadoes in the United States.
Poster Session 1, IIPS Poster Session I
Monday, 30 January 2006, 2:30 PM-4:00 PM, Exhibit Hall A2
Previous paper Next paper
Browse or search entire meeting
AMS Home Page