|The graph at the right shows the number of tornadoes reported each year since 1950. The numbers shown are likely an underestimate of the true tornado count since some events, especially in rural areas, go unreported. In the early 1990s, a major modernization program within NWS took place, which included the deployment of Doppler radar and a more rigorous spotter training program. In addition, the 1990s saw an increasing number of storm chasers using video cameras and cell phones to document and report severe weather. These facts are likely responsible for the dramatic increase in annual tornadoes reported since 1990. Okay, and maybe the movie "Twister"!|
|Another way to illustrate the increase in reported tornadoes in recent years is to take a look at the number of tornado reports in each decade since 1950. From the 1950s through the 1980s, the number of tornadoes reported in the Tri-State area was fairly consistent with 65 to 85 events logged in each ten year period. Since 1990 however, the number of tornado events logged is about three times that of prior decades.|
|Still another way to look at this trend is to again divide the data into decades, but now calculate the average number of tornadoes each year within that decade. For example, in the 1950s, 87 tornadoes were reported over the ten years, giving an average of 9 tornadoes each year from 1950-1959. Again notice that from the 1950s through the 1980s, the average annual tornado count was consistent between 7 and 9 tornadoes. In the past two decades however, over 200 tornadoes have been logged during each 10-year period, giving a more recent annual average of 23 or 24 tornadoes per year. The last bar on the right is a bit misleading in that only three years of data (2010-2012 is included).|
|This chart shows the number of tornado days per year. A tornado day is defined as a day on which a tornado occurs. This is not the same as the number of tornadoes per year. For example, in 2008 there were 39 tornadoes reported in the Goodland County Warning Area. The chart to the right shows there were 12 tornado days in 2008. So those 39 tornadoes occurred on 12 days. Some days obviously had multiple tornadoes, while other may not have.|
|This chart is similar to the chart above, but shows the average number of tornado days per year in each decade. For example, during the 1950s, there was an average of about five tornado days each year. Compare that to the 1990s decade when there was an average of over 10 tornado days each year. This does not necessarily imply anything about the number of tornadoes during the 1990s compared to the 1950s. To illustrate, if one year has a single tornado day when 25 tornadoes occur, and the following year has ten tornado days but only one tornado occurs each day, the second year would have fewer tornadoes even though it had many more days when a tornado was observed.|
|The next two charts illustrate the month-to-month variation in tornado reports. Although tornadoes can occur in any month given the right atmospheric conditions, over the past 63 years, no tornadoes have occurred between November 1 and February 28. The peak months for tornadoes in the Tri-State area historically are May and June, and to a lesser extent, July.|
|This chart takes the monthly information above and breaks each month's tornado count down into a percentage of the total. Taking June as an example, 35 percent of all tornado reports received since 1950 have occurred during that month, while 30 percent have occurred in May.|
The Annual Tornado Season chart to the right shows a bar for each year of the 63-year tornado record. You'll definitely need to click on the chart for readability. The years 1950 to 2012 are shown along the x axis, and dates are shown on the y axis. The length of each vertical bar is the length of that year's tornado season, which begins on the day of the first observed tornado and ends on the day of the last observed tornado. A tall bar shows a year with a long tornado season while a short bar shows a year with a short tornado season.
The dates along the left axis give an idea of when the first and last tornado occurred each year. Look at 1973. The first tornado was observed in early March, and the last tornado that year was reported in mid-May. Meanwhile, in 1992 the first tornado also occurred in early March but the season lasted into mid-September. Two years (1974 and 1976) do not have a bar. Those years each had only one tornado day, which didn't graph well. In 1974, one tornado occurred on April 20th, while in 1976, two tornadoes were reported on just a single day...September 12th, which by-the-way, is the latest day on record for the first tornado.
On average, the date of the first tornado in the Tri-State area is May 10, and the date of the last tornado is August 9. But not all years are average! 1992 saw the first tornado develop on March 8, while the year 2000 saw the last tornado on October 31, the "Halloween Outbreak".
Similar to the Annual Tornado Season chart immediately above, this chart shows another way to display the length of the tornado season each year. Rather than displaying the dates of the first and last tornado, this chart shows the number of days between the first and last tornado, so the height of each bar (and the data label above) is related to the length of the bars in the previous chart. Once again, look at 1974 and 1976. As stated above, these two years had the shortest tornado season on record, with tornadoes occurring on only one day.
|The Date of First Tornado shows the number of years in the 63-year record in which the first tornado of the year fell into each 10-day period shown along the bottom axis. Looking at Mar 21-31, records show that in only four years did the first tornado occur during that period. However, 13 of the 63 years saw the first tornado occur between May 1 and May 10. This chart gives a quick view of when the first tornado of the season is likely to occur, which is from mid-April into May. The year 1976 shows up again as the outlier bar on the right when the first tornado was reported between September 11-20.|
|Dr. Ted Fujita developed the F-Scale in 1971 to provide a method to rate the intensity of tornadoes. The intent of the scale was to distinguish between weak and strong tornadoes. Dr. Fujita's scale was quickly accepted by the NWS and came into wide use by the mid-1970s. Values on the scale range from F0 for tornadoes producing minimal damage, to F5 for tornadoes causing devastating damage. A rough estimate of the wind speeds responsible for the damage could then be inferred using the F scale. In February 2006, this scale was modified and is now referred to as the "EF" scale, or Enhanced Fujita Scale. Sixty four tornadoes in the historical records were not assigned a rating, perhaps due to lack of information about the tornado. These unrated tornadoes are shown in the first column and labeled EF?.|
This chart expresses the EF ratings in the chart above as a percentage. Tornadoes rated EF0 and EF1 are considered weak, tornadoes rated EF2 and EF3 are considered strong, and those with ratings of EF4 and EF5 are considered violent.
It would appear that most tornadoes in the Tri-State area are weak since 83% are either EF0 or EF1. This is likely not the case however. Tornado ratings are based on observed damage. In rural areas with vast expanses of open fields there are simply less things to damage compared to urban areas. Thus, a tornado which in reality might be violent, will receive an EF0 rating if it remains over open country producing no damage.
|The number of tornadoes reported in each of the 19 counties in the Goodland CWA varies quite a bit over the years. Sherman county Kansas takes the dubious honor of having the most tornadoes reported since 1950, with Kit Carson and Yuma counties in Colorado coming in 2nd and 3rd. Logan County Kansas and Dundy County Nebraska have had the fewest number of tornadoes reported at 25. Also note there are two Cheyenne counties. The one at the far left is Cheyenne county Kansas, while the one toward the right is Cheyenne county Colorado.|
The graph in the previous slide is somewhat misleading in that the size of the counties is not uniform. It would seem logical that if county A is twice as large as county B that there would be more tornadoes reported in county A, all other things being equal.
To remove some of that bias, each county's tornado count in the slide above was divided by the area of the county and normalized to 1000 square miles. The graph to the right shows a more uniform tornado tally among the counties when this factor is taken into account. Sherman county Kansas still comes in with a high number of tornado reports. And no, the presence of the Doppler radar in Goodland is not to blame!