A Comparison of Tornado Statistics from Tornado Alley and Dixie Alley

1
A Comparison of Tornado Statistics from Tornado
Alley and Dixie Alley

• Alan Gerard and John Gagan NWS WFO Jackson, MS
• And
• John Gordon
  NWS WFO Louisville, KY

2
Tornado Alley?
www.cdli.ca/CITE/tornadoes_alley.htm
3
Tornado Alley Definitions

• Region in the middle of the United States where
  more tornadoes occur than anywhere else in the
  world.
• Most tornadoes in the United States form in an
  area called "Tornado Alley". This area includes
  parts of Texas, Oklahoma, Kansas, and Nebraska.
  www.windows.ucar.edu/tour/link/earth/Atmosphere/t
  ornado/alley.html
• Where is tornado alley? geography.about.com/librar
  y/faq/blqztornadoalley.htm
• The United States Tornado Alley, stretches
  from northwest TX across OK and through northeast
  KS.

4
Fawbush and Miller 1948
The weather situation is discussed by Capt.
Robert C. Miller (left) and Lieut. Col. E. J.
Fawbush
5
Perception

• Historically, the term tornado alley has been
  used loosely, and the locations indicated are
  often based on personal perception rather than
  scientific data. (The Tornado Nature's Ultimate
  Windstorm pg 278 Grazulis)
• Wizard of Oz

6
Origins of Tornado Alleys.

• Personally spoke to Doswell, Grazulis, Dan
  McCarthy, Joe Schaeffer, Howie Bluestein, Fred
  Ostby, and Allen Pearson about the origins of
  Tornado Alley.
• Searched many books (including Flora and internet
  sources)
• No one knew the origins of Tornado Alley.
• Found papers from Kelly 78 et al., Concannon 2000
  et al., Grazulis books, and no definitive answer.
  

7
Fawbush and Miller

• On Feb 15, 1952, Fawbush and Miller set up a new
  project called, "Tornado Alley," in which a
  concentrated study of severe weather activity was
  made over an area extending from Lubbock, Texas
  to Eastern Colorado and Nebraska.
• This was done in stages, the 1st being from
  LUB-END, 2nd from END north to the Nebraska
  line.

8
Many Different Alleys

• Evidence of smaller tornado alleys across the
  United States based on a long track F3 to F5
  tornado climatology study from 1880 to 2003"
  (Broyles and Crosbie 2004)
• There is more than one tornado alley. Any area
  that seems to experience above average tornado
  frequency is eventually labeled a tornado
  alley. There are dozens of such regions.
  (Grazulis The Tornado Nature's Ultimate Windstorm
  pg 278).

9
Dixie Alley Origins

• AR/TN OUTBREAK March 21-22, 1952 28 tornadoes
  204 deaths
• First tornado watch ever issued by NWS
• MS Delta Outbreak Feb 21, 1971 10 tornadoes 121
  deaths, 1524 injuries

10
Dixie Alley

• HoweverPearson

Named Dixie Alley after 71 Outbreak in MS Delta
(no formal documentation that we can locate).
11
Jeopardy Question?

• Which alley is under the greatest threat of
  tornadoes for the entire year?
• Plains Tornado Alley
• Dixie Alley

12
Depiction of Plains and Dixie Tornado Alleys
13
Plains Alley and Dixie Alley Breakdown

• Plains Alley
• Size 407,734 sq mi
• Population 17,824,023
• Dixie Alley
• Size 276,890 sq mi
• Population 24,948,345

Population Data from the 2000 Census
14
Methodology

• Our own depictions of Tornado Alley and Dixie
  Alley are personal
• Data is from 1950 to 2003 and is derived from the
  SPC ONETOR database
• A close approximation to actual tornadoes.
• Focused on strong/violent tornadoes
• Data not normalized for population or spatial
  extent

15
Tornadoes by F-Scale
16
Strong Violent Tornadoes by Month
17
Killer Strong Violent Tornadoes by Month
18
Strong Violent Tornadoes by Time of Day
19
Killer Strong Violent Tornadoes by
Time of Day
20
Strong Violent Tornado Path Length
21
Tornado Deaths
22
Strong Violent Tornado Fatalities
23
Post NWS Modernization

• In addition to the longer term statistics, felt
  it would be revealing to compare to post-NWS
  modernization.
• Increased number of tornadoes since deployment of
  88D.
• Increased ability to see potentially tornadic
  storms.
• More emphasis on verification/storm data.
• Look at statistics from 1998-2004, focusing
  particularly on strong tornadoes.

24
Methodology

• Used same definitions of Tornado Alley and
  Dixie Alley.
• Looked at total tornadoes, strong tornadoes,
  strong tornado days, and numbers of tornadoes, by
  month.
• Data taken from the NWS storm data/verification
  homepage.
• Subject to potential biases
• County based versus track based
• Population density and county size potential
  biases same as longer term dataset.

25
Summary of Stats

• Dixie Alley had about 1.5 times as many strong
  tornadoes as Tornado Alley (338 vs 206) while
  Tornado Alley had more tornadoes overall (2278
  vs. 1703)
• Killer tornadoes much more prevalent in Dixie
  Alley area than Tornado Alley
• 66 vs 24
• Dixie Alley had twice the number of outbreak
  days
• 9 days with 10 strong tornadoes versus 5 in
  Tornado Alley

26
Seasonal Trends
27
Percentage of Annual Strong Tornado Days
28
Annual Trend

• 76 of strong tornado days in Tornado Alley occur
  in a three month period (April, May, June)
• Adding March in takes the value to nearly 90
• The highest percentage over a three month period
  in Dixie Alley is 46 (March, April, May)
• The next two highest three month periods are
  February, March, April (41) and November,
  December, January (37)
• Eight out of 12 months (Oct-May) have at least 5
  of the annual strong tornado days for Dixie Alley
• Only 5 out of 12 for Tornado Alley (Mar-June,
  Oct)

29

• Over the last several years in Dixie Alley
• March has been significantly below long-term
  normal with regard to strong/violent tornadoes
• November has been significantly above normal with
  regard to strong/violent tornadoes

30
Conclusions

• Confirms validity of Plains Tornado Alley
• Not meant to diminish magnitude of risk in this
  area
• Highlights magnitude of tornado risk in Dixie
  Alley area
• Strong/violent tornadoes as prevalent
• Killer tornadoes much more prevalent
• Temporal aspect of threat much different in Dixie
  Alley versus Plains Tornado Alley
• Threat most of year killer strong/violent
  tornadoes not uncommon in late fall/winter
  months!!
• While a peak in strong/killer tornadoes occurs in
  late afternoon and evening, a much higher risk
  exists in the 9 pm to 5 am timeframe than in the
  Plains Tornado Alley.

31
Tornado fatalities declined by a similar
magnitude between 1916 and 1965 as between 1966
and 2005
More rapid dropoff in latter period suggests
improved warnings/technology played large role
BUT outbreaks of violent tornadoes will probably
still be big killers!