NOAA Technical Memorandum NWS SR-182

NOAA Technical Memorandum NWS SR-182


Nelson A. de Villiers
National Weather Service Forecast Office
Little Rock, Arkansas

Scientific Services Division
Southern Region
Fort Worth, TX

February 1997

1. Introduction

The National Weather Service is undergoing major restructuring. Part of this process has resulted in a different area of warning responsibility for the Little Rock office. The purpose of this paper is to examine the occurrence of severe weather in the new County Warning Area (CWA). Those counties are shown on Fig. 1. There are 75 counties in Arkansas, and the Weather Service Forecast Office (WSFO) at Little Rock warns 46 of them. Adjacent offices of the National Weather Service provide warnings for the remaining counties.

Statistics will be used to quantify and describe severe weather occurrences within the area of interest. In particular, figures will refer to tornado segments, hail, and damaging winds. The database used was provided by the Verification Section of the Storm Prediction Center (SPC) in Norman, Oklahoma, formerly known as the National Severe Storms Forecast Center (NSSFC) in Kansas City. The CLIMO program (Vescio 1995) produced data for the period from 1950 to 1993 for tornado segments, and from 1955 to 1993, except 1972, for hail and damaging winds.

Tornado segments are used instead of actual tornadoes because the SPC verifies weather warnings on a county by county basis. For example, if a tornado travels across three counties, the data base will show three tornado segments, although they were produced by a single tornado. Keep in mind as you look through the data that the term tornado will refer to segments rather than the full path of the tornadoes.

Large hail is defined as hailstones with a diameter greater than or equal to 0.75 in. Hail reports are subdivided into three categories--hail with diameter from 0.75 in to less than 1.75 in, from 1.75 in to less than 2.75 in, and with diameter equal to 2.75 in or larger. The base values of each of the three sizes are described as penny or dime, golf ball, and baseball.

Damaging wind reports refer to instances when wind damage occurred (whether speeds were known or not) or where wind speeds were greater than or equal to 50 kt (58 mph).

Data contained in this paper could serve several interests. At the WSFO data can be used to acquaint new forecasters with what they might expect throughout the year. Insurance companies can use the numbers to assess risk from severe weather. Emergency managers can evaluate the data to prepare for weather contingencies. Newcomers to the area can become familiar with meteorological disaster threats to which they may be exposed, and the curious may have their interest satisfied. Many more uses are possible.

2. Tornadoes in the NWSFO Little Rock CWA

2.1. Tornado Classification

Tornadoes have been classified (Table 1) by the strength of their winds using the Fujita Scale (Fujita 1987), ranging from F0 to F5. The CLIMO software tracked the total number of tornadoes, the number of strong tornadoes (F2 or greater), and the number of killer tornadoes (any intensity tornado as long as it caused at least one fatality).

Table 1. Fujita Scale of Tornado Intensity

Fujita Scale Wind speed range in mph Damage description
F0 Under 72 Light
F1 73-112 Moderate
F2 113-157 Considerable
F3 158-206 Severe
F4 207-260 Devastating
F5 261-318 Incredible

Grazulis (1993) augmented the F-scale to include median and average path lengths and widths, as shown in Table 2 below.

Table 2. Average and Median Path Length and Width per F-scale

F-scale Average length (mi) Median length (mi) Average width (yds) Median width (yds)F-sc
F0 1.11 0.30 46 17
F1 2.59 0.98 93 47
F2 5.66 2.19 167 99
F3 12.08 6.76 290 180
F4 22.42 13.80 432 297
F5 34.17 23.44 616 496
all 4.40 0.98 128 48

2.2. Tornado Frequencies

During the period in question (1950-1993), there were a total of 616 tornadoes, with 335 of them being classified as strong (F2 or greater intensity), and 45 of them were killer tornadoes causing 174 deaths (Fig. 2). Weak tornadoes (F0 and F1) numbered 271, leaving 10 tornadoes without an F-scale value assigned to them. It is interesting to note that 1987 was the only year which registered no tornadoes at all in the CWA, while 1982 had the highest number at 71.

Tornadoes occur primarily in the spring months, with a marked decline in the summer, and a secondary peak during the late fall (Table 3). March, April, and May receive 56 percent of all tornadoes, 58 percent of strong tornadoes, 69 percent of killer tornadoes, and 89 percent of all fatalities. Table 3 shows the monthly percentage of tornadoes, strong tornadoes, and killer tornadoes. Table 4 shows the monthly distribution of tornadoes. Figure 3 shows that even with most of the tornadoes occurring during two peaks, all months are susceptible to tornado occurrences.

Table 3. Monthly Percentage of Tornadoes

All 4 7 16 23 17 6 3 1 2 2 9 10
Strong 3 8 19 21 18 3 0 1 1 2 12 12
Killer 2 5 31 16 22 0 0 0 0 2 13 9

Table 4. Monthly Tornado Distribution

All 26 41 98 139 107 34 21 8 12 10 58 62
Strong 11 26 65 70 59 10 0 4 3 7 40 40
Killer 1 2 14 7 10 0 0 0 0 1 6 4

As Fig. 4 shows, tornadoes occur mostly during the late afternoon and early evening. Table 5 shows the percentage distribution by hours of all tornadoes, strong tornadoes, and killer tornadoes. The hours between 4 p.m. and 9 p.m. encompass 58 percent of all tornadoes, 58 percent of strong tornadoes, and 65 percent of all killer tornadoes. Although there is a frequency minimum around 6 a.m., tornadoes can occur at all hours of the day.

Table 5. Hourly Percentage Distribution of Tornadoes

1A 2A 3A 4A 5A 6A 7A 8A 9A 10A 11A NN
All 1 2 1 1 1 1 1 1 1 2 2 2
Strong 2 2 2 1 1 1 0 1 0 2 0 2
Killer 0 7 0 0 2 2 0 0 0 0 2 0

1P 2P 3P 4P 5P 6P 7P 8P 9P 10P 11P MN
All 4 3 4 9 10 10 11 9 9 7 4 4
Strong 4 2 4 9 8 10 12 8 11 10 4 4
Killer 5 0 2 11 7 16 11 7 13 7 4 4

The intensity distribution of tornadoes is shown in Fig. 5. Of the 616 tornadoes, only 606 were given an F-scale classification. Table 6 shows the percentage distribution by F-scale for tornadoes within the CWA for WSFO Little Rock. Table 7 shows the distribution of observed tornadoes by the same categories.

Table 6. Percentage Distribution by F-scale

F0 F1 F2 F3 F4 F5
17 28 35 17 3 0

Table 7. Distribution by F-scale

F0 F1 F2 F3 F4 F5
101 170 216 102 17 0

2.3. Tornado Fatalities

The yearly distribution of tornado deaths is shown in Fig. 8, with the monthly and hourly distributions following in Figs. 9 and 10, respectively. The reader can see that tornado-related fatalities occur mostly during the spring months and during the late afternoon and early evening hours, as was the case with tornado occurrences. Table 8 shows the monthly death totals, while Table 9 shows the hourly totals.

Table 8. Monthly Distribution of Tornado Fatalities

No. 1 2 115 18 22 0 0 0 0 1 11 4

Table 9. Hourly Distribution of Tornado Fatalities

1A 2A 3A 4A 5A 6A 7A 8A 9A 10A 11A NN
No. 4 1 0 0 1 1 0 0 0 0 1 0

1P 2P 3P 4P 5P 6P 7P 8P 9P 10P 11P MN
No. 2 0 2 8 63 38 19 9 11 7 3 4

2.4 Tornado Injuries

It is interesting to note in Table 10 that in the monthly distribution of tornado injuries only the month of July was free from injuries during the period under review (1950-1993). Similarly, there was only one injury-free hour during the same period, as shown in Table 11. The numbers of tornado injuries are shown in yearly, monthly, and hourly distribution, respectively, in Figs. 11-13.

Table 10. Monthly Distribution of Tornado Injuries

No. 29 85 841 460 196 11 0 9 5 8 153 133

Table 11. Hourly Distribution of Tornado Injuries

1A 2A 3A 4A 5A 6A 7A 8A 9A 10A 11A NN
No. 41 21 5 7 1 1 0 6 7 6 21 44

1P 2P 3P 4P 5P 6P 7P 8P 9P 10P 11P MN
No. 21 22 61 100 429 278 355 165 158 113 32 37

3. Hail

Since hail is a product of severe thunderstorms, as are tornadoes, it is not surprising that hail occurs most frequently in the spring months and during the late afternoon and early evening hours (Figs. 14-16). Table 12 shows the monthly distribution of large hail occurrences (0.75 in or larger), and Table 13 shows the hourly distribution of large hail.

Table 12. Monthly Distribution of Large Hail

No. 35 29 167 364 180 121 93 55 24 37 68 21

Table 13. Hourly Distribution of Large Hail

1A 2A 3A 4A 5A 6A 7A 8A 9A 10A 11A NN
No. 23 21 10 14 3 9 15 16 13 10 16 15

1P 2P 3P 4P 5P 6P 7P 8P 9P 10P 11P MN
No. 37 67 73 108 133 148 129 101 93 67 43 30

4. Damaging Wind

Wind damage has a tendency to occur during the warm months of the year, not just in the spring as with tornadoes and hail. The yearly, monthly, and hourly distribution of wind damage occurrences are depicted in Figs. 17-19. Table 14 shows the monthly distribution of damaging wind occurrences, while Table 15 shows the hourly distribution of damaging winds.

Table 14. Monthly Damaging Wind Occurrences

No. 22 37 122 254 310 264 258 268 72 83 149 35

Table 15. Hourly Damaging Wind Occurrences

1A 2A 3A 4A 5A 6A 7A 8A 9A 10A 11A NN
No. 69 55 37 41 30 31 16 20 25 11 26 32

1P 2P 3P 4P 5P 6P 7P 8P 9P 10P 11P MN
No. 62 100 168 160 184 164 121 124 133 122 81 65

5. Note Regarding Interpretation of Results

No attempt has been made to address population density, the increased efforts to solicit public reports of severe weather, nor the increased emphasis on severe weather warning verification, all of which may have changed over the period of time covered by this study. Another factor which seems to have increased the number of severe weather occurrences is the concerted effort by the NWS to train severe weather spotters and to deploy them during severe weather episodes. Some of these points are addressed by Grazulis, et al. (1993).

6. Conclusions and Recommendations

Severe weather occurs all year long within the CWA assigned to the Little Rock NWS Office. It does have a tendency to show its effects in the spring and during the fall, but no month should be considered completely safe. By the same token, one can see that no hour during the day is safe. It is true that most severe weather tends to happen during the late afternoon and early evening hours, but severe weather also occurs even at dawn, when heating is at a minimum and cooling at a maximum.

Forecasters should take a vigilant posture during their shifts to anticipate the possibility of severe weather, according to the current weather conditions, without discounting automatically the probability of severe weather simply because it is the wrong time of the year or the wrong hour of the day. Emergency managers should be prepared to respond to watches and warnings issued by the National Weather Service at any time, as well. The general public and severe weather spotters need to know and review severe weather safety rules, and they should be prepared to respond appropriately whenever a warning is issued or severe weather is observed.


The author would like to thank George R. Wilken (Science and Operations Officer at the Little Rock NWS Forecast Office) for his suggestions, proofreading, and patient prodding to finish this project.


Fujita, T. F., 1987: U. S. Tornadoes, Part 1, 70-Year Statistics, Univ. Of Chicago Press, Chicago.

Grazulis, T. P., J. T. Schaefer, and R. F. Abbey, Jr., 1993: Advances in tornado climatology, hazards, and risk assessment since Tornado Symposium II. Geophysical Monograph 79, C. Church, et al, ed. Amer. Geophysical Union, Washington, DC, pp. 409-426.

Vescio, M. D., 1995: CLIMO--Software to generate severe weather statistics for NWS County Warning Areas. NWS/NCEP Storm Prediction Center, Norman, OK.