NOAA Technical Memorandum NWS SR-193, Section 3

3. VARIATIONS BY STATE IN REPORTED FREQUENCIES

A. DEATHS AND INJURIES COMBINED (CASUALTIES)

Section 3 describes variations among states in actual frequencies of lightning reports by the use of maps and tables. This section shows results without population weighting; population is included in section 4. Data used to develop the following maps and tables are given at the end of section 3.

The sum of fatalities and injuries together is termed casualties. Figures 1 and 2 map the sum of lightning-caused deaths and injuries. Tables 4 and 5 give the top and bottom ten locations for deaths and injuries combined. All states are shown in descending order by the number of casualties in Figure 3. The largest numbers of casualties are in Florida, Michigan, Pennsylvania, North Carolina, and New York. Other high numbers are found in the southern and eastern regions of the US, and some states in the northeastern US where there is a large population. Also evident is a large number of casualties in the mountainous, dry states of Colorado, New Mexico, and Arizona. The smallest numbers of lightning casualties are in Alaska (none), Hawaii, the District of Columbia, northwest US states, Puerto Rico, and several small eastern states. Casualties by county for Colorado are in L�pez et al. (1995).

There are 13,057 Storm Data casualties from 1959 to 1994 (Table 3 and Figures 1 to 3) for an average of 363 per year. The National Lightning Detection Network identified an average of 21,746,000 cloud-to-ground flashes per year in the US from 1992 to 1995 (Orville and Silver, 1997). Assuming this four-year average to be representative, the result is a casualty every 60,000 flashes. Since around 70% of the flashes were detected, about 86,000 flashes occur for each casualty. A similar estimate can be made for Arizona from L�pez et al. (1997). Making corrections for season, analysis area, and detection efficiency, about 800,000 flashes per year occur in Arizona, while there were 4.5 casualties per year from 1959-1994 (Table 12). The resulting ratio in Arizona is one casualty for every 175,000 flashes; this is reasonable since there usually are fewer flashes in Arizona than in the rest of the US (Orville, 1991; Orville and Silver, 1997).

Figure 1: US map of number of lightning casualties (deaths and injuries combined) by state from 1959 to 1994.

Figure 2: US map of lightning casualties (deaths and injuries combined) ranked by state from 1959 to 1994.

Figure 3: Number of lightning casualties (deaths and injuries combined) ordered by state from 1959 to 1994.

B. DEATHS

Maps of lightning-caused fatalities are shown in Figures 4 and 5, and states with the most and least deaths in Tables 6 and 7. Florida has twice as many deaths compared to any other state in Storm Data.

A major change from previous results for casualties is the absence of Michigan in the highest totals of deaths (Table 6), indicating that injuries have been more commonly-reported than deaths in Michigan. Otherwise, the same states are in the first-ten list for both deaths and casualties, although in a different order. Also, Maryland and Arkansas for deaths replace Georgia and Colorado on the casualty list. The large Maryland entry is mainly due to the 81 deaths in a lightning-caused aircraft crash in 1963.

In terms of the least lightning-caused fatalities, there were no deaths in Hawaii and Alaska. The bottom-ten list in Table 7 includes the same states as casualties (Table 5), except New Hampshire and Rhode Island replace Delaware and Puerto Rico.

Lightning deaths from Storm Data at specific locations in the US were shown from 1959 to 1965 by Zegel (1967). Storm Data deaths were plotted by state in Mogil et al. (1977) for 1968 to 1976, and Duclos and Sanderson (1990) for 1968 to 1985.

Pakiam et al. (1981) plotted each fatality on a map of Singapore. Coates et al. (1993) for Australia and Gourbiere et al. (1997) for France showed maps of lightning deaths divided by political boundaries in formats similar to Figure 5. In the US, state maps of lightning deaths by county have been compiled for North Carolina (Langley et al., 1991), Michigan (Ferrett and Ojala, 1992), and Colorado (L�pez et al., 1995).

There were 3239 Storm Data deaths in the US from 1959 to 1994 (Table 3 and Figure 4 and 5) for an average of 90 per year. The same analysis can be made for deaths using the network-detected ground strikes as in the previous section for casualties. The result is one death for every 345,000 flashes in the US.

Figure 4: US map of number of lightning deaths by state from 1959 to 1994.

Figure 5: US map of lightning deaths ranked by state from 1959 to 1994.

C. INJURIES

Figures 6 and 7 show maps of injuries by state. Locations with the most and least injuries are in Tables 8 and 9. Florida had more injuries than any other state since 1959, as well as deaths (Table 6) and casualties (Table 4). For example, there were 105 injuries in Florida in 1996 alone (Paxton and Morales, 1997).

The same states appear on the first-ten list of fatalities in Table 8 except for two changes. Michigan is second in injuries but twelfth in deaths. Also, Georgia and Massachusetts on the injury list replace Arkansas and Maryland on the fatality list. The least injuries are in Alaska (0), Hawaii (4), Puerto Rico (6), and other western states and small eastern states.

The large number of Michigan injuries is due in part to two exceptional events (Ferrett and Ojala, 1992). During one Storm Data case in August 1975, 90 people were injured when lightning struck near the center of a campground at Leslie, Michigan. In June 1979, 45 National Guard soldiers suffered minor injuries when lightning struck their camp near Grayling, Michigan. A US map of injuries was shown for 1968 to 1976 by Mogil et al. (1977), and a state map of injuries by Colorado county was compiled by L�pez et al. (1995).

Most states have more injuries than deaths; the US average ratio is 2.54 injuries per death in Storm Data from 1959 to 1994. But Missouri had only slightly more injuries (93) than deaths (78), and Puerto Rico had 6 injuries but 30 deaths; a few other locations approach a 1:1 ratio. A map of the injury to death ratio by state (not shown) has no pattern. In Puerto Rico, the authors found that many lightning injuries occur there but are not widely known, while deaths are usually reported to the National Weather Service and included in Storm Data. A low ratio of injuries to fatalities may indicate underreporting of injuries, such that deaths are better reported. L�pez et al. (1993) found a greater underreporting of injuries requiring hospitalization (42%) than under-reporting of deaths (28%) in Colorado.

There were 9818 Storm Data injuries in the US from 1959 to 1994 (Table 3 and Figure 6 and 7). The same analysis for injuries using network-detected lightning as in previous sections results in one US injury for every 114,000 flashes.

Figure 6: US map of number of lightning injuries by state from 1959 to 1994.

Figure 7: US map of lightning injuries ranked by state from 1959 to 1994.

D. DAMAGE REPORTS

Maps of the damage reports are in Figures 8 and 9 by state. Locations with the most and least damage reports are in Tables 10 and 11.

It is apparent that damage reports in Storm Data are distributed very differently than are deaths and/or injuries. A high concentration of damage reports is evident over the plains from South Dakota to Texas. The highest number of damage reports is from Pennsylvania, where less than half as many casualties were reported as in Florida. In contrast, Florida is first on all casualty lists but is not high on the list of damages. Seven of the ten states with the highest damage counts are on the first-ten lists for casualties, deaths, or injuries. While Kansas, Oklahoma, and South Carolina rank in the first eight for damages, they are not in the top-ten list for any casualty category.

The least damage reports are from Alaska (3), Puerto Rico (4), and many of the same locations as in the casualty lists. An exception is the small number of damage reports from New Mexico, a state with a high number of deaths (Figure 4).

In the US, state maps of lightning damages by county have been compiled for Michigan (Ferrett and Ojala, 1992), Colorado (L�pez et al., 1995), and Colorado, Utah and Wyoming (Holle et al., 1996).

There is a weakly-defined geographical pattern (not shown) in the ratio of damage reports to casualties in Storm Data. The damage/casualty report ratio was near one, or less, in all southwestern states, and several southeastern states including Florida. The northwest half of the US tended to have nearly two damage reports for every casualty report. Whether the ratio is influenced by accurate reporting of every death or injury, or whether the reporting system in some states is not very complete in reporting damages is unknown.

It is also unknown why there should be any pattern since damage reports are so greatly underreported (Section 2). Damage reports in Storm Data are underreported by as much as 367:1, as described in Section 2 based on insurance claims in Holle et al. (1996). Therefore, rates of flashes per damage report cannot be made reliably from the available database as was done for casualties, deaths, and injuries in previous sections.

Figure 8: US map of number of lightning damage reports by state from 1959 to 1994.

Figure 9: US map of lightning damage reports ranked by state from 1959 to 1994.

E. SUMMARY

Table 12 provides the detailed list of reported frequencies and corresponding ranks of fatalities, injuries, casualties, and damage reports for all states and other locations in the dataset. Information in this table was used to develop the preceding maps and tables.

Florida led the nation in deaths, injuries, and therefore casualties, over all other locations by a wide margin. States with high numbers of casualties (Figures 1, 2) tended to be in the following categories:

• Florida--very frequent casualties and a sizable population.
• Southeast and southern plains--frequent casualties and medium to large populations (Alabama, Georgia, Mississippi, South Carolina, Tennessee, Texas).
• Midwest and northeast--densely populated (Michigan, New York, Ohio).
• Southern Rockies--less populated than many states (Arizona, Colorado, New Mexico).

The largest number of damage reports came from Pennsylvania, but it had less than half as many casualties as Florida. North Carolina had uniformly high frequencies in all categories - second in deaths, fourth in injuries, and fourth in damages. In contrast, Kansas was second in damage reports but twenty-fifth in deaths and twenty-second in injuries. There were a few locations, such as Alaska, Hawaii, Puerto Rico, the District of Columbia, and Nevada, with very few casualties and damage reports over the 36-year period.

Data from recent years collected by the US National Lightning Detection Network results in a rough estimate of one lightning casualty for every 86,000 flashes. A similar method gives an estimate of one death for about every 345,000 flashes, and an injury for about every 114,000 flashes.