Research Problem and Data

We intend to investigate tornadoes, their frequency, and magnitude from 1950-1999.

Data was taken from the entire United States, which we hope will give us a large picture and convey any overall themes, as well as several states. These states include (1) Texas, which historically experiences the most tornadoes, (2) Nebraska, a state consistently ranking in the top 10 for tornado frequency, and (3) Ohio, our state and one in which tornadoes are known to happen, but is not considered a 'hotspot.' It is our intent that these states will give a complete picture of tornado occurrence, and show what may happen to hotspots and areas of lower activity when climate is changed.

Tornado Data for the United States from 1950 - 1999.

  1950's 1960's 1970's 1980's 1990's
Total # Tornadoes 4796 6813 8579 8141 10696
Yearly Avg 533 757 953 905 188
Monthly Avg 44 63 79 75 99

 

El Nino La Nina
1958 1951
1964 1952
1966 1956
1969 1965
1973 1971
1983 1974
1987 1989
1988  
1992  
1995  
1998  

FREQUENCY DATA

Tornado Frequency By Decade

  1950's 1960's 1970's 1980's 1990's
Texas 647 1195 1484 1492 1836
Nebraska 319 302 405 377 701
Ohio 71 148 202 165 251

There is a significant difference in the rank orders of data for Texas and Nebraska. There is no significant difference in the Texas-Ohio and Nebraska-Ohio tests.

Spearman Rank Correlation p-values:
Texas-Nebraska p=.1096
Texas-Ohio p=.0719
Nebraska-Ohio p=.0719

Tornado Frequency in El Nino Years

Years
Texas Nebraska Ohio United States
1958 75 55 12 608
1964 78 46 9 760
1966 77 10 3 606
1969 131 20 21 650
1973 152 19 55 1199
1983 190 16 10 995
1987 135 26 6 695
1988 92 23 0 773
1992 203 77 63 1404
1995 204 64 10 1479
1998 134 93 35 2149

The only two tests which revealed no significant difference between rank orders were Texas-USA and Nebraska-USA. All others revealed a significant difference.

Spearman Rank Correlation p-values:
Texas-Nebraska p=.4902
Texas-Ohio p=.1837
Texas-USA p=.0170
Nebraska-Ohio p=1983
Nebraska-USA p=.0785
Ohio-USA p=.1223

Tornado Frequency Non-El Nino Years

  1950's 1960's 1970's 1980's 1990's
Texas 572 909 1332 1075 1295
Nebraska 264 226 386 312 467
Ohio 59 115 147 149 143
United States 4188 4797 7380 5678 5664

Only the Texas-USA and Ohio-USA tests revealed no significant difference in rank orders.

Spearman Rank Correlation p-values:
Texas-Nebraska p=.1096
Texas-Ohio p=.1615
Texas-USA p=.0719
Nebraska-Ohio p=.3173
Nebraska-USA p=.2301
Ohio-USA p=.0719

Tornado Frequency La Nina Years

  Texas Nebraska Ohio United States
1950 15 9 3 269
1952 13 11 2 272
1956 56 36 10 567
1965 110 48 39 995
1971 193 52 15 963
1974 118 36 25 1123
1985 166 41 19 921
1997 237 43 18 1735

Texas-Ohio and Nebraska-Ohio were the only two rank orders which show a significant difference. The majority of tests showed a conformity in rank orders, or no significant difference.

Spearman Rank Correlation p-values:
Texas-Nebraska p=.0422
Texas-Ohio p=.1474
Texas-USA p=.0376
Nebraska-Ohio p=.1117
Nebraska-USA p=.0751
Ohio-USA p=.0438

Tornado Frequency Non La Nina Years

  Texas Nebraska Ohio United States
1950's 563 263 56 3688
1960's 1085 250 109 5818
1970's 1173 317 162 6493
1980's 1326 336 143 7220
1990's 1599 658 233 8961

Only Nebraska-Ohio revealed a significant difference in rank order correlations. All other pairs were not significantly different.

Spearman Rank Correlation p-values:
Texas-Nebraska p=.0719
Texas-Ohio p=.0719
Texas-USA p=.0455
Nebraska-Ohio p=.1096
Nebraska-USA p=.0719
Ohio-USA p=.0719

MAGNIGTUDE DATA

Magnitude counts total 1950 - 1999

Magnitude Texas Nebraska Ohio
F0 2823 758 185
F1 1976 622 335
F2 1112 267 177
F3 330 87 55
F4 78 46 31
F5 6 4 9

Our Spearman rank correlation tests reveal p-values of less than .10: we can reject the null hypothesis, which for Spearman rank tests says there is a significant difference in the two sets of data. In lay terms: the rank orders of the sets of data tested here are statistically similar.

Spearman Rank Correlation p-values:
Texas-Nebraska p=.0253
Texas-Ohio p=.0350
Nebraska-Ohio p=.0350

El nino magnitude counts 1950 - 1999

Magnitude Texas Nebraska Ohio
F0 668 120 48
F1 407 105 67
F2 195 46 38
F3 64 5 25
F4 15 2 1
F5 2 4 0


Here we also see p-values of less than .10, so we reject the null hypothesis. We say there is no significant difference between any of the sets of data for total magnitude counts.

Spearman Rank Correlation p-values:
Texas-Nebraska p=.0350
Texas-Ohio p=.0350
Nebraska-Ohio p=.0476

La nina magnitude counts 1950 - 1999

Magnitude Texas Nebraska Ohio
F0 168 100 11
F1 261 74 44
F2 185 20 34
F3 40 16 11
F4 8 9 10
F5 0 0 0

Again, we see p-values of less than .10, which means we reject the null hypothesis: there is no significant difference between any sets of data in these tests. The rank orders of magnitude in la nina years for Texas, Nebraska, and Ohio are statistically similar.

Spearman Rank Correlation p-values:
Texas-Nebraska p=.0639
Texas-Ohio p=.0275
Nebraska-Ohio p=.0905

 

Intro
background
relevance
methods
data
conclusion
bibliography