THE SOUTHERN EXTENT of TORNADO ALLEY Philip W

INTRODUCTION PURPOSE OF THIS STUDY RESULTS “Tornado Alley” is a term that is often used to refer to a region in the central This study attempts to address the discrepancies in the geographic depiction of The kernel density estimation was applied to our 1° latitude by 1° longitude United States where tornado frequency is especially high (Bluestein, 1999). For the Tornado Alley, specifically south of the Oklahoma heart of tornado peak occurrence. quadrant tornado frequency data for Oklahoma and Texas. Figure 8 provides a United States, it is well-known that tornado frequency is greatest within the state of In other words, what is an appropriate representation of Tornado Alley within the monochromatic depiction of the results while a color version is presented in Figure 9. Oklahoma. The National Climatic Data Center’s informational website, in an article state of Texas? For further illustration, Figure 10 shows the dot map tornado locations (i.e. as shown titled Tornado Climatology, shows tornado frequency based on long-term data from in Figure 7) overlain onto the KDE map results (i.e. as presented in Figure 9). the Storm Prediction Center of the National Severe Storms Laboratory (Figure 1). The tornado density results clearly illustrate that the southern portion of A depiction of Tornado Alley is also presented on this map running from northern Tornado Alley is located distinctly further to the east than the governmental agency Texas is a north-north-eastward direction towards Iowa and southern Wisconsin. depictions and most popular website depictions presented in Figures 1-5. The NCDC article also shows another depiction of Tornado Alley based on the DATA AND METHODS Interestingly, the Weather for Dummies depiction of Tornado Alley (shown in Figure work of Concannon et al. (2000) (see Figure 2). The National Severe Storm In recent decades, the capability for detection of tornadoes has increased 6) is a closer representation to that suggested for the geographic location of Tornado Laboratory’s severe weather primer website presents yet another depiction (see considerably due to technological advances (Grazulis et al., 1993). However, the Alley by our study. Figure 3). In both of these cases, Tornado Alley is shown to stretch from north- trend in the frequency of significant (i.e. moderate-strong) F2-F5 tornadoes within central and northwest Texas due northward to South Dakota. These latter depictions the United States has not increased appreciably in recent decades (Grazulis, are located further to the west than that illustrated in Figure 1. The map in Figure 3 2001). Therefore, for our study, we focused on significant tornadoes (F2-F5) that also attempts to illustrate the causation for high tornado frequency in this area by occurred in Oklahoma and Texas. showing locations of typical warm and cold air masses. Several public informational Figure 7 presents a simple dot map showing the location (of initial websites depict Tornado Alley using the NSSL geographic depiction shown in Figure touchdown) for each of the F2-F5 tornadoes that occurred in Oklahoma and Texas 3. during the period 1950-2005. Considerable spatial variability is evident. Other popular informational web sources depict Tornado Alley in various In order to facilitate map illustration, total tornado frequency was calculated manners. For example, the Tornado Chasers website simply highlights states in the within 1° latitude by 1° longitude quadrants for the study area. An attempt to central part of the country (Figure 4). Chuck Doswell’s tornado chasers website provide an isoline map of tornado frequency based on these quadrant data failed illustrates Tornado Alley (Figure 5) similar to that depicted by Concannon et al. due to the evident considerable spatial heterogeneity. (2000) (as shown in Figure 2). By contrast, Tornado Alley is illustrated in the book In order to create a smooth surface that represents the density of events Weather for Dummies (Cox, 2000) to be further eastward (Figure 6). across an area, a popular technique is the “kernel density estimation” (KDE) The concept of Tornado Alley is indeed very subjective and there may be method. The goal of this technique is to model the distribution of the underlying many smaller tornado alleys across the country as argued by Broyles and Crosbie population from the sample data (Jiang et al., 2007). In our case, this would entail (2004). Nevertheless, the depictions presented in these maps (Figures 1-6) using the 1° latitude by 1° longitude quadrant data. KDE has been widely used for generally correspond to the popular perception of Tornado Alley, albeit with notable geospatial analyses in the social sciences, and a computer algorithm for the variations in its specific geography. method is available within ArcGIS. Recently, Edwards et al. (2004) and Weinbeck et al. (2006) have successfully used the KDE technique to illustrate spatial density

of severe weather events including severe thunderstorms and tornadoes. Figure 8 Figure 9

Figure 1 Figure 2

Figure 7 Figure 10

Figure 3 Figure 4

References Cited Bluestein, H.B. (1999) Tornado Alley: Monster Storms of the Great Plains. New York: Oxford University Press, 180 pp. CONCLUSION Broyles, C. and C. Crosbie (2004) “Evidence of smaller tornado alleys across the United States based on a long track F3 to F5 tornado climatology study from 1880 to 2003.” Preprints, 22nd Conference on Severe Local Storms. Hyannis, MA: American Meteorological Society. This study has shown that the southern area of Tornado Alley is located well Concannon, P.R., H.E. Brooks and C.A. Doswell III (2000) “Climatological risk of strong and violent tornadoes in the United States.” Second Conference on Environmental Applications. Long Beach, CA: American Meteorological Society. within northern Texas, but especially in the eastern part of the state. This geographic Cox, J.D. (2000) Weather for Dummies. Hoboken, NJ: John Wiley and Sons, 384 pp. Edwards, R., R.L. Thompson, K.C. Crosbie, J.A. Hart and C.A. Doswell III (2004) “Proposals for modernizing the definitions of tornado and severe thunderstorm outbreaks.” location is distinctly further to the east (within Texas) than depictions found on most Preprints, 22nd Conference on Severe Local Storms. Hyannis, MA: American Meteorological Society. Grazulis, T.P. (2001) The Tornado: Nature’s Ultimate Windstorm. Norman, OK: University of Oklahoma Press, 324 pp. governmental and popular informational websites. Grazulis, T. P., J. T. Schaefer, and R. F. Abbey, Jr. (1993) “Advances in tornado climatology, hazards, and risk assessment since tornado symposium II.” In: The Tornado: Its Structure, Dynamics, Prediction, and Hazards, ed. C. Church, D. Burgess, C. Doswell, and R. Davies-Jones, Geophysical Monograph 79, 408-426. Washington DC: American Geophysical Union. Jiang, D., C.F. Eick and C. Chen (2007) “On supervised density estimation techniques and their application to spatial data mining.” Proceedings of the 15th International Symposium on Advances in Geographic Information Systems. Seattle, WA: Association for Computing Machinery. Weinbeck, S., S. Trebes and J. Maliekal (2006) “Kernel density estimates of tornado occurrence in the United States.” 22nd International Conference on Interactive Acknowledgements Processing Systems for Meteorology, Oceanography, and Hydrology. Atlanta, GA: American Meteorological Society. The technical assistance of Roberto Campos and Emariana Taylor is gratefully acknowledged.

Figure 5 Figure 6