sign in sign up
<<
Home , Dixie Alley, Tornado Alley, Tornado climatology, Tornado records

A HISTORICAL AND STATISTICAL COMPARISON OF " ALLEY" TO " ALLEY"

John P. Gagan

National Weather Service Springfield,Alan Gerard

Jackson, John Gordon

Louisville, National Weather Service

Abstract describe various parts of the that are perceived to have a high incidence of tornado The media and public frequently use the term “,” an unofficial term which is used to occurrence. This paper examined the history of the term “Tornado Alley,” as well as the lesser known term “.” A basic statistical analysis of the temporal and spatial occurrence of tornadoes and tornado-related casualties was done in each of these areas, both long term and in the recent years following the modernization of the National Weather Service.

Corresponding Author: Alan Gerard

NationalJackson, Weather MS 39232 Service email:234 [email protected] Service Drive Gagan, et al.

1. History and Terminology • “States included in ‘Tornado Alley’ are , a. Tornado , , and Missouri.” • , , , , , phenomena and certainly one of the most destructive. southern“An L-shape Oklahoma , whichand then stretches over Arkansas from western and Tornadoes are one of thetornado most talked about weather LouisianaIowa down to through southeastern Nebraska Mississippi.” and Kansas to

The history of the word took several twists Some publications that address “tornado alley” included: ,before it landed tronada with its, and current then turned meaning into in ternado the United by EnglishStates. sailors Ultimately, in the it comes1500s fromto mean the “a Spanish violent word tropical for • tornadoes occur most often is in an L-shaped “The primary area of the U.S. in which significant storm.” In pop-culture, movies such as “The Wizard of Oz” in 1939, “Mr. and Mrs. Bridge” in 1990, and “Twister” (Concannonregion from Iowaet al. to2000) Oklahoma to Mississippi, in 1996 have certainly increased the public’s awareness. • with the highest threat in Oklahoma.” KansasPublic awareness turnpike bridge has been on fueled26 April with 1991, impressive the 3 May tornado 1999 videovideos, from including the Moore, a television Oklahoma crew F5 hiding tornado, underneath and the 24 a separately.“Texas, Oklahoma, This region Kansas, essentially Nebraska, contains , the Missouri, Arkansas, were analyzed most prevalent.” (Schaefer and Tatom1999) June 2003 National Geographic video from inside an F3 proverbial ‘tornado alley’ where tornadoes are b.tornado “Tornado near Alley” Manchester, . Historically, the term “Tornado Alley” has been used loosely, and the locations indicated are often based on term “Tornado Alley.” Generally speaking, this is an area Many Americans are familiar with the widely used personal perception rather than scientific data (Grazulis thought of as the location for peak tornado occurrence in 1993). The authors personally spoke to Dr. Charles of the and Midwest which is commonly Doswell III and Dr. Howard Bluestein from the University of Oklahoma, Tom Grazulis from The Tornado Project, the United States. However, a less commonly known term retired (SPC) Director Dr. Joe is that of “Dixie Alley.” This generally refers to that part of Schaefer, retired SPC Science and Operations Officer the southeastern United States which is prone to tornado Bob Johns, retired National Severe Storms Forecast development. This manuscript will take an in depth look Center (NSSFC) Director Fred Ostby, and former NSSFC at theIf you similarities use any of and the popular differences internet between search “Tornado engines aforementionedDirector Allen Pearson. referenced Nobody Snoden could Flora’s provide 1954 a definitive tornado Alley” and “Dixie Alley.” answer to the origin of “Tornado Alley.” Most of the Flora’s manuscript. The authors conducted a thorough orand so, type the in location “tornado for alley tornado location”, alley variesyou will greatly. come Someup with of climatology, although we could not locate a reference in theseover 400,000 include: hits. Of the first 100 internet search results every decade from the 1950s through today. This included literary search which entailed numerous papers from • several papers from Snoden Flora, Dr. Theodore Fujita, “Region in the middle of the United States where and The Dr. Howardonly “tornado Bluestein alley” which reference offered that no the insights authors to • “Mostmore tornadoes tornadoes occur in the than United anywhere States form else in anthe definitive location or origin. Take-off, the Tinker, areaworld.” called ‘Tornado Alley.’ This area includes Januarycould obtain 16, 1953 was from[Available an article online in at: • OK Air Force Base (AFB) newspaper, which]. appeared The article on fromparts north-centralof Texas, Oklahoma, Texas through Kansas, centraland Nebraska.” http://www.nssl. “By most definitions, ‘Tornado Alley’ extends noaa.gov/GoldenAnniversary/awards.html stated: “On Feb 15, 1952, Fawbush and Miller set up a new dryOklahoma, air collide Kansas, time andafter Nebraska time.” into the extendingproject called, from ‘TornadoLubbock, Alley,’Texas into whichEastern a Colorado concentrated and • “TheDakotas, United the Statesnorth-south ‘Tornado line Alley’ where stretches moist and from study of severe weather activity was made over an area Lubbock, Texas to Enid, Oklahoma, and the second from northeast Kansas.” Nebraska. This was done in stages, the first being from northwest Texas across Oklahoma and through 146 National Weather Digest Enid, Oklahoma to the Nebraska line.” A Historical and Statistical Comparison of "Tornado Alley" to "Dixie Alley" c. “Dixie Alley” perceptions. Tornado Alley, hereafter referred to as the

Kansas and Oklahoma, and a large portion of northern the authors once again consulted those tornado experts andPlains central Tornado Texas Alley (Fig. 1). (PTA), includes all of Nebraska, previouslyRegarding cited, the and origin researched and location numerous of “Dixie previous Alley,” been an idea cultivated through the observations of The existence of a Dixie Tornado Alley (DTA) has publications. Most experts referred the authors to Dr. Harold Brooks, but Dr. Brooks told us explicitly that he meteorological professionals who have lived in the Pearson.did not coin Mr. Pearson the phrase. explained The only that source he personally that knew coined the southern states and those with severe weather forecasting origin of “Dixie Alley” was former NSSFC Director Allen severeexperience. thunderstorm The first climatology. statistical analysis A higher suggesting frequency the of of 21 February 1971. This event had 10 long-track validity of a DTA was provided by Brooks et al. (2003) tornadoesthe phrase thatafter resulted working inthe 121 Mississippi deaths and Delta over outbreak 1,500 tornadoes was located from Louisiana through much of Mississippi and , when accounting for the mean injuries. Pearson compared that event with the Arkansas- thenumber frequency of days of withlong-track tornadoes (path from length October greater through than outbreak of 21-22 March 1952, where 28 25December. miles) violent Broyles tornadoes and Crosbie (rated F3 (2004) to F5) investigated across the tornadoes struck, killing over 200 people. Both events were in the south, and PearsonUSA Todaycoined Weather the phrase [Available “Dixie Alley”online inat: 1971. One of the few websites defining Dixie inUnited central States Mississippi from 1880 andto 2003. an area They extending discovered from that Alleystorms/1999/ndstue09.htm was from an article in two of the most prominent tornado alleys were located http://www.usatoday.com/weather/tornado/ from northeast Texas to northern] which Alabama, stated: then “A skips third proposedwestern Tennessee by this investigation into northern includes portions all of of Mississippi Arkansas, tornado focus is the so-called ‘Dixie Alley’, which runs and Alabama. With these discoveries in mind, the DTA Carolina.” Tennessee, and northern and central (Fig. 1). over the southern Appalachians and reforms in North Louisiana, Mississippi and Alabama, western and central b. Methodology Since there was such limitedSignificant formal Tornadoes, documentation, 1680- 1991it was compendium, difficult to compare there is more the two than tornado one tornado alleys. alley. As “AnyGrazulis area (1993)that seems stated to experience in above average tornado frequency is eventually labeled a ‘tornado alley.’ There are databaseTornado [Available data for online both at: the PTA and DTA were gathered from]. This the database Storm Predictionprovides detail Center’s ranging ONETOR from tornado alleys across the United States based on a long the physical attributes of eachhttp://www.spc.noaa.gov/ tornado (e.g., location, dozens of such .” There is also evidence of smaller wcm/#data track F3 to F5 study from 1880 to damage rating, path length and width), to the effects of statistical2003 (Broyles analysis and Crosbieof the frequency 2004). This of papertornadoes will bridgein the the gap between perception and reality by offering a basic

“Tornado“Dixie Alley.” Alley.” In an effort to support the validity of a “Dixie Alley,” these statistics will be compared to those from 2. Tornado Alley and Dixie Alley – A Historical Study a. Background accepted by the meteorological community for decades, thoughThe as existence discussed of “Tornadoin the introduction Alley” has beensection, widely its exact spatial extent has had numerous interpretations.

Alley referred to in this text should be considered the authors’Without interpretation a definitive areabased to on reference,a collective theof published Tornado Fig. 1. A map outlining the Plains Tornado Alley, in red, and

Volume 34 Number 2 ~ December 2010 147 the Dixie Tornado Alley, in green. Gagan, et al.

3). In the PTA, 72% of the observed strong and violent individualthe tornado tornadoes (e.g., injuries [Available and fatalities). online at: The ONETOR tornadoesreports by monthoccur (Fig.during 2) asthe well three-month as by the time period of day from(Fig. database was developed by counts of , not pdf http://www. strong and violent tornadoes have been observed from spc.noaa.gov/wcm/SPC_severe_database_description. April through June. In the DTA, only 38% of the observed ] . A vast majority of the records within the database aare tornado counts crosses of individual one or tornadoes; more state however, lines. Therefore, duplicate LikeApril the through PTA, June.there is In a the maximum DTA the in risk the for frequency strong and of inrecords order forto develop a tornado a data exist set within of individual the database tornadoes, when violent tornadoes is spread out over two distinct . a prominent, secondary peak in tornado frequency in to exclude duplicate counts of tornadoes. In addition, tornadoes during the spring ; however, there is tornadorecords withinsegments the outsideONETOR of databasethe boundaries were pared of the down PTA violent tornadoes are most pronounced in the months November. The differences in the frequency of strong and strong and violent tornadoes have occurred during this eventsand DTA (one were for excluded each alley). and tornadoes For this study,that moved reports from of from October through February. In the DTA, 40% of one alley to another were segmented into two separate time frame, while only 10% have occurred in the PTA. tornadoes from 1950 through 2007 were interrogated. occurWhen during considering the afternoon the time and of early day (Fig. evening 3), it hours is evident (12 frequencyFrom this of subset all tornadoes, of the ONETOR the frequency database, of raw strong tornado and PMthat to the 9 PM, vast Central majority Standard of strong Time). and Seventy-six violent tornadoes percent statistics for each alley were extracted, including the and 59% of all strong and violent tornadoes occur during the number of casualties, and the frequency of killer violent tornadoes (F2/EF2 or greater on the ), the overnight hours (9 PM to 7 AM CST), the frequency ofthis strong time frameand violentin the PTA tornadoes and DTA, in respectively. the PTA decreases During strong and violent tornadoes. In order to draw a more andaccurate population. one-to-one The comparison PTA spans of 407,734 the DTA mi and2 the PTA, in frequency during the overnight hours, 29% of strong statistics have been normalized for2 . bothAccording spatial to areathe andsubstantially violent tornadoes to 19%. While occur the during DTA alsothis timeshows frame. a decrease 2000 United States Census [Available online at:, dwarfing http:// Casualties sustained by each tornado alley also vary thequickfacts.census.gov/qfd/index.html DTA, which covers 276,890 mi , the population of is smaller in geographic coverage as compared to the considerably in both amount and timing. While the DTA the DTA was 24,948,345, while the population of the PTA c.was Comparison 17,824,023. of the Dixie Tornado Alley and the Plains http://quickfacts.census.gov/qfd/index.htmlPTA, over 7 million more persons reside in the]. DTAFigures than Tornado Alley the PTA, per the U.S. Census Bureau [Available online at

4-5 highlight the distribution of fatalities and injuries for comes to overall tornado frequency. Over 13,500 tornadoes the DTA and PTA by month. A higher incidence of fatalities Without question, the PTA is second to none when it and injuries have been reported in the DTA as compared to the PTA (in the DTA 1705 fatalities and 26,026 injuries andwere violent reported tornadoes in the (rated PTA, with F2/EF2 over to 7500 F5/EF5 tornadoes on the were reported, while in the PTA nine hundred ninety-one reported in the DTA. However, when considering strong fatalities and 14,709 injuries were reported.) One would expect that the DTA would incur more casualties given that Fujita scale), the annual frequency of tornadoes is much a larger population lives within this alley. Normalizing more comparable between alleys. Over 2850 strong the amount of casualties by 100,000 people confirms that frequenciesand violent bytornadoes area (10,000 were mireported2) produces in the similar PTA, resultswith a the DTA has been slightly more dangerous than the PTA. little over 2450 reported in the DTA. Normalizing these The DTA has observed 6.8 fatalities and 104 injuries per frequency of all tornadoes for the PTA outpaces that of 100,000 people, while the PTA has observed 5.6 fatalities for the overall frequency of tornadoes. The normalized2, and 83 injuries per 100,000 people. As expected, the majority of the reported casualties in the PTA, 88% of the DTA, with 332 and 276 tornadoes per 10,000 mi fatalities and 84% of injuries, occur between April and 70respectively. tornadoes Forper 10,000strong andmi2, violentrespectively. tornadoes, however, June. During this same time frame, only 35% of fatalities the DTA has a higher frequency than the PTA, with 89 and and 38% of injuries occur in the DTA. Far more fatalities frequencies of strong and violent tornadoes, each alley (64%)Killer and stronginjuries and (61%) violent occur tornadoes from the (Fig. period 6) are October most has Whileunique bothdistributions the PTA of andstrong DTA and observeviolent tornado similar through March in the DTA.

148 National Weather Digest common in the DTA. Three hundred seventy-one killer A Historical and Statistical Comparison of "Tornado Alley" to "Dixie Alley"

Fig. 2. The frequency of strong and violent tornadoes by month in the Plains Tornado Alley

(red) and the Dixie Tornado Alley (green).

Fig. 3. The frequency of strong and violent tornadoes in the Plains Tornado Alley (red) and the

Dixie Tornado Alley (green) by hour (CST).

Volume 34 Number 2 ~ December 2010 149 Gagan, et al.

Fig. 4. The frequency of all tornado-related fatalities by month in the Plains Tornado Alley (red)

and the Dixie Tornado Alley (green).

Fig. 5.

The frequency of all tornado-related injuries by month in the Plains Tornado Alley (red) and the Dixie Tornado Alley (green).

150 National Weather Digest A Historical and Statistical Comparison of "Tornado Alley" to "Dixie Alley"

violenttornadoes tornadoes have been by area observed (Fig. 7) in thefurther DTA, demonstrates with 205 in the timeframe when most people are asleep and more the PTA. Normalizing the frequency of killer strong and vulnerable3. Post-NWS to tornado Modernization related injury Statistics or death. the disparity between the DTA and PTA. In the months In addition to the long-range statistics discussed in mispanning2 October through March, 8.4 and 1.0 killer monthsstrong and spanning violent April tornadoes through were June, observed 4.7 and per 3.9 10,000strong in the DTA and PTA, respectively. Meanwhile, in 2the in the previous section, a separate analysis was also made for the time period after the modernization of the NWS andPTA violentexhibit tornadoesa maximum were risk observedof killer strongper 10,000 and violent mi (post-modernization era, or PME). The rationale for the DTA and PTA, respectively. While both the DTA and looking at this time period separately is that with the advent of operational Doppler radar, NWS personnel are months.tornadoes around April, the DTA demonstrates a much more aware of over sparsely populated more substantial risk during the late fall and winter areas, which prior to Doppler radar might have gone goneundetected. unreported These during mesocyclones the conventional would radar be capable era. This of Figure 8 highlights the distribution of killer strong and producing tornadoes which were more likely to have strongviolent andtornadoes violent bytornadoes time of day.occur As during is the the case afternoon with all andstrong early and evening violent hours tornadoes, (12 PM the to 9vast PM majorityCentral Standard of killer is particularly the case for the DTA where storm spotting Time). Seventy-six percent and 62% of killer strong and andis extremely more frequent difficult nightdue to time forestation, events. hilly Therefore, terrain, andthe violent tornadoes occur during this time frame in the the diffculty in storm spotting due to lower cloud bases

PM to 7 AM CST), there is a noticeable difference in the Doppler time period might yield different results than PTA and DTA, respectively. During the overnight hours (9 the relatively longer time frame when only conventional 21% of killer strong and violent tornadoes occur during radar was generally used. Of course, the much shorter thefrequency overnight of killerin the strongPTA, an and alarming violent 34% tornadoes. of killer strongWhile Doppler period of analysis introduces an inherent error and violent tornadoes occur during the overnight in the source that would most likely be smoothed in a longer termThe analysis. methodology Thus, this used drawback in calculating must be kept the in mindpost- when drawing any conclusions from these results. timeDTA. period The DTA than has the nearly PTA, twice and thethis number obviously of incidence includes of strong/violent tornadoes during the Midnight to Noon modernization statistics differs from that previously described in a few key ways. First, the period evaluated

Fig. 6. The frequency of killer strong and violent tornadoes by month in the Plains Tornado

Alley (red) and the Dixie Tornado Alley (green). Volume 34 Number 2 ~ December 2010 151 Gagan, et al.

Fig. 7.

The frequency of killer strong and violent tornadoes by month, normalized by area (per10,000 mi2) in the Plains Tornado Alley (red) and the Dixie Tornado Alley (green).

Fig. 8. The frequency of killer strong and violent tornadoes in the Plains Tornado Alley (red)

and the Dixie Tornado Alley (green) by hour (CST).

152 National Weather Digest A Historical and Statistical Comparison of "Tornado Alley" to "Dixie Alley"

actually increases again in January before decreasing into February and March before reaching its spring peak in is a nine-year period from 1998 to 2006. Selection of April. This trend is in contrast to the longer term statistics this time period is of course somewhat subjective, but frombased a on severe the authors’ convection experience perspective it was due felt to that the 1998 full before increasing up to the peak in April. best demarcated the full modernization of the NWS which show a relative minimum in December and January implementation of the Weather Surveillance Radar (WSR- Comparing the two alleys’ traces reveals the magnitude 88D ) system. Secondly, because of the desire to look at of the PTA spring peak was greater than that of DTA, veryverification end of thestatistics study inperiod, this part NWS of Storm the study Data as[Available well as while the fall peak is clearly much larger (by nearly an onlinepotential at concerns about the ONETOR database at the order of magnitude) in the DTA than the PTA. While the February),PTA trace reachesthe incidence zero or of near strong zero tornadoes for several during months the http://www4.ncdc.noaa.gov/cgi-win/wwcgi. of the year (July through September, November through dll?wwEvent~Storms] was used to calculatecounty-based the statistics summer months. increased the tornadorather than events. ONETOR. This This introduces means thata potential instead biasof looking since incidencePME only ofapproached tornadoes or during reached the zerolate summer,in DTA during and cool- the at actual tornadoes, we are looking at season tornado events led to strong tornado occurrence at a rate of at or above .05 per 10,000 mi2 areas with more numerous, smaller counties will likely arehave relatively somewhat short more (< 20 reports km), it is than expected those that areas this withbias incidence of strong tornado events per 10,000 for all mi winter2 shouldfewer but be relativelylarger counties. minimal, Because but could most result tornado in numbers tracks and spring months other than December. The total annual was counties are generally smaller than in the PTA area. 1.74 for DTA versus 0.81 for PTA. thatThe are overall slightly trend tilted of tornado toward thereports DTA during region, the where PME While there exists potential biases or errors in the (greaterdata which population might leaddensity, to possibility an artificial of more inflation active of time the number of strong tornado events in DTA versus PTA was not unexpected given the results found from the incidence nearly three times as great as that of PTA seems longer term study. Overall, there were about 50% more toperiod, clearly and support smaller the counties), premise thatthe fact strong that tornadoes the DTA had are an at reports of strong tornadoes from the DTA region (384) killerthan PTAtornadoes (265). occurred When looking in this region at killer during tornadoes, the period the importantly, the much larger number of killer tornadoes ofdominance study, compared of the DTA to only region 24 inis veryPTA. clear.The number A total inof the79 least as common in the DTA region as PTA. Perhaps more

factorsin the DTA (e.g., region both would meteorological indicate that and a combinationsocioeconomic of the as DTA comprised more than one-third (36.9%) of all killer rate of strong tornado occurrence combined with other tornadoes in thePTA United comprised States only during 12% the 1998 to 2006 Strong Tornadoes 98-06 ofperiod, the national while correspondingly total. the killer strong tornadoes in both alleys during the Figure PME 9 period, shows theusing annual the trendsimple of of tornadoes by the area of each alley (expressednormalization as oftornadoes dividing theper number10,000 mi2 distinct maximum in the spring months, ). As can be seen, both alleys show a and PTA’s decreasing after May, and with the DTA’s decreasing after April

DTA’s larger in magnitude. However, DTA shows a pronounced fall maximum in November, much larger than PTA’s spring maximum, which the PTA does (there is a small increase in frequency Fig. 9. The per annum frequency of strong and violent tornadoes by month, duringnot show October). in any After significant a decrease manner from 2

normalized by area (per 10,000 mi ), for the Plains (red) and Dixie Alley (green) November into December, the DTA trace areas, for the PME (1998-2006). Volume 34 Number 2 ~ December 2010 153 Gagan, et al.

residents there also do not discount the risk of nighttime thandiscussed the PTA in theregion. next section) means that the DTA region tornadoes.is sufficient Another risk of such concern events is that in the cool PTA season to ensure tornadoes that was much more susceptible to tornado-induced loss of life are often moving at a greater speed than storms in the 4. Conclusion to take shelter. warmThe season, challenge potentially facing leadingoperational to less meteorologists time for people in arguments over the “champion” area as far as tornado The purpose of this work is not to provoke scholarly onlinethe DTA storm can bereport summed program, up looking available at at:statistics http://spc.nssl. from the incidence. Rather, it is to show that public and media noaa.gov/climo/online/sp3/plot.phplast 20 years. Between 1986 and 2005, SVRPlot [SPC perception is just that, perception, and that reality is often themore complexity complicated. of the Recent United outstandingStates’ tornado work vulnerability, by Ashley ] shows that 504 and(2007) the andpotential Ashley for et theal. (2008)mid-South has and helped Southeast to illustrate to be people have been killed by tornadoes in the two “alleys” “overlooked” by the focus on “Tornado Alley,” including combined. Seventy-five percent of these fatalities (377) discussions of the potential socioeconomic factors (e.g., occurred in the DTA. On average, a strong tornado in the DTA is associated with 0.44 fatalities, more than twice the rate as in the PTA (0.18). Also, NWS verification statistics tornadohigher incidence reporting of mobilemake anyhome one-to-one use) which comparison will not be during this time period show that the probability of restated here. While the various biases associated with LT:detection 13.6 min (POD) versus and 10.2 lead min). time (LT) for strong tornadoes best, the statistics outlined in this paper make several is greater in the PTA than the DTA (POD: .80 versus .68; clearbetween points “Tornado Alley” and “Dixie Alley” difficult at part due to the inability to effectively spot storms in the First, both the monthly and diurnal distributions of southeastThe disparity United States in verification due to heavy skill forestation, is likely in and large it should be noted that this disparity has shrunk markedly ontornado the day incidence of peak arerisk much is substantially different in less the than DTA thethan risk in the PTA. In the DTA, the risk of a strong or killer tornado since NWS modernization introduced Doppler radar into warning operations [POD: .75; (1996-2005) versus .45 on the peak day in the PTA. However, this peak is much (1986-1995); LT: 12.6 min versus 6.1 min]. Still, tornado riskless throughpronounced much in of the the DTA, year and rather the than annual an extremeincidence risk in events associated with large numbers of fatalities continue the DTA can be viewed as more of a persistent, moderate to be a serious risk in the DTA even with improved warnings, as shown by the 2008 Super Tuesday outbreak. during a more confined “season” as is observed in the PTA. tornadoes.This outbreak Clearly, killed if the 57 people,fatality rate nearly is to all be in decreased the DTA, When the more sustained risk of the DTA is added up over even with advance warnings issued for all of the killer greatan entire a risk year, of strong/violentit seems clear thattornadoes even when as the taking PTA, andany certainlypotential datahas a biases greater into risk account, of killer the tornadoes DTA has at than least the as substantially in the DTA, improved meteorological science PTA. operationaland warnings meteorology will not be sufficient,community at to least better not understand in the near publicterm. Atresponse least as importantthrough an will increased be a commitment focus on bysocial the

Additionally, the DTA has about a 50% greater risk of 7strong am. Thetornadoes annual duringand diurnal the overnight distribution hours, of thewith tornado over a fromsciences, tornadoes as well faced as increased by the region. understanding among the third of its killer tornadoes occurring between 9 pm and public and government officials of the significant risk risk in the DTA has a substantial impact on preparedness maintainand warning a level dissemination of tornado efforts preparedness in this area. throughout Clearly, Acknowledgments residents and emergency personnel in the DTA must preparednessmost of the year, activities which contrastsaround the to the“tornado PTA in season” which the of The authors would like to thank the reviewers for their lateNWS, March emergency through management June (although and media occasional can focus atpyical their thoughtful comments which improved the quality of this paper. We also would like to thank Tom Grazulis events do occur outside of the “season” here as well). from The Tornado Project for his assistance with the Warning dissemination and preparedness activities in thishistorical study. part of this work. Finally, we would like to the DTA must also account for the greater frequency of thank Jon Davies for his encouragement to formalize overnight154 National tornadic Weather activity Digest as well, although clearly there A Historical and Statistical Comparison of "Tornado Alley" to "Dixie Alley"

Authors References

John P. Gagan is currently a Senior Forecaster at the

Ashley,Wea. W.S., Forecasting 2007: Spatial and temporal analysis of National Weather Service (NWS) Forecast Office in tornado fatalities in the United States: 1880–2005. researchSpringfield, focused MO. He on receivedheavy rainfall his Master and the of Science role of systemdegree , 22, 1214-1228. in Meteorology from Saint Louis University in 2001, with Wea. Ashley,Forecasting, W.S., A.J. Krmenec and R. Schwantes, 2008: propagation. John began his weather career as a medium- Vulnerability due to nocturnal tornadoes. asrange a General forecaster Forecaster and weather in Jackson, derivative MS trader from forlate Aquila 2002 23, 795-807. throughEnergy, based early in2006. Kansas John’s City, researchMO. He began interests his NWS have career been climatology. [Available online at: Brooks, H.E., cited 2003: Severe thunderstorm] an emphasis in the early detection and risk assessment of http://www.nssl. focused on the improvement of weather forecasts, with noaa.gov/hazard/svrwxmisc.html Climatological estimates of local daily tornado Alanhazardous, Gerard high is impact the weatherMeteorologist-In-Charge events. of the Brooks,probability H.E., Doswell, for the United C.A. III States. and Wea. Kay, Forecasting,M.P., 2003:

MS, a position he has held since 2002. Prior to that, he National Weather Service Forecast Office in Jackson, 18, 626-640. tornado alleys across the United States based on a was the Science and Operations Officer and a Senior Broyles,long J.C., track and F3-F5K.C. Crosbie, tornado 2004: climatology Evidence study of smaller from Forecaster in the same office. He has also held positions . Preprints, 22nd Conf. Severe Local at the NWS forecast office in Cleveland, OH and the NWS Storms, Amer. Meteor. Soc., Hyannis MA. [Available Commissioneroffice in Columbus, of Committees. OH. He is He a has past-president published articles of the online1880-2003 at: inNational numerous Weather publications, Association, including and is currentlyNational theWeather NWA broyles/longtrak.pdf]. Digest, Electronic Journal of Operational Meteorology, and http://www.spc.noaa.gov/publications/ Weather and Forecasting. He is a graduate of St. Louis University. Climatological risk of strong and violent tornadoes Concannon,in the P.R.,United Brooks, States. H.E. andPreprints, Doswell, 2nd C.A. III,Conf. 2000: on John Gordon is currently the Meteorologist-In-Charge Environmental Applications, Amer. Meteor. Soc., 212-219. Long Beach, CA, counties(MIC) of of the Kentucky National and Weather southern Service Indiana. (NWS) Previously, Forecast Tornadoes of the United States. University Office in Louisville, KY, which serves citizens across 59 of Oklahoma Press, 206 pp. Flora, S.D., 1954: MeteorologyJohn was the firstfrom MIC Parks at the College new NWS of St. office Louis in Huntsville,University Significant Tornadoes, 1680-1991. Alabama from 2002-2004. Mr. Gordon received a BS in Environmental Films, 1326 pp. Mississippi State in 2002. John served as a Lead Forecaster Grazulis,T.P., 1993: in 1986 and a Masters in Teaching in Geoscience from The Tornado: Nature’s Ultimate Windstorm. University of Oklahoma Press, 352 pp. at WFO Nashville, TN from 1998-2002, Journeyman Grazulis, T.P., 2001: 1994.Forecaster Mr. atGordon WFO Springfield,retired from MO the from U.S. Air 1994 Force to 1998, (AF) Reservesand Meteorologist after serving Intern nearly at WFO 21 Jackson, years. MostMS from of his 1993- AF Schaefer,activity. J. T, Preprints, and F. B. 19th Tatom, Conf. 1999: Severe The Local relationship Storms, between El Nino, La Nina and United States tornado career was spent as a flight meteorologist for the US Air CategoryForce Reserve 4 and 5 53rd hurricanes Weather over Reconnaissance 3 years from 2003-2005 Squadron , MN, Amer. Meteor. Soc. (Isabel,Hurricane Ivan, Hunters and Katrina). based in Biloxi, MS. He flew three

Volume 34 Number 2 ~ December 2010 155