Lightning Fatalities in Africa From 2010-2017

Full text

1
Lightning Fatalities in Africa from 2010-2017
Ronald L. Holle
Executive Committee
African Centres for Lightning and
Electromagnetics Network
Oro Valley, Arizona, United States
Mary Ann Cooper, M.D.
Managing Director
African Centres for Lightning and
Electromagnetics Network
River Forest, Illinois, United States
Abstract—Lightning is a major hazard to people in Africa but
there is only a small amount of direct information about its
impacts. Since about 2010, there has been an improvement in
news reporting by use of the internet. The result is wider
dissemination of lightning incidents that had been occurring but
were not reaching newsgathering media. In the absence of
organized lightning casualty data collection in most of Africa, this
study will address the use of available internet-based reports. A
sample review of lightning fatalities for Zimbabwe is summarized
from English-language searches in order to obtain an overview of
scenarios involved in lightning casualties. Future data collection
will be stratified, when possible, by nation in Africa, month,
activity, location, and age, number and gender of individuals.
Keywords-Africa, lightning, lightning fatalities, monthly,
agriculture, hut, school
I. I
NTRODUCTION
The combination of frequent lightning occurrence and a
lightning-vulnerable population makes it apparent that large
numbers of lightning casualties are taking place in Africa. Yet
there is minimal organized information on the topic from
many locations in Africa. A summary of existing global
lightning fatality studies [1, 2] has been updated to include
several new countries, as shown in Fig. 1.
Table I lists the few national studies of lightning deaths
that have been published for multiple years since 1979 for
Africa. These are for Burundi [3], much of Malawi [4], South
Africa [5], Swaziland [6], Uganda [7], and Zimbabwe [8, 9]. It
is apparent in Fig. 1 that the fatality rate is typically large or
very large in Africa compared with developed nations of the
world. Australia and countries in Europe, North America, and
several other regions have rates below 0.5 deaths per million
people per year. Reasons for the differences are explored in
detail in [10]. This study will begin to address some gaps in
this knowledge despite the lack of national-scale studies in
Africa at this time.
II. D
ATA
S
OURCES
For such studies, data can be collected with some success
by online search engines. These daily or hourly updates
provide all links to reports related to the word “Lightning”.
However, this identification restricts the links to those in
English so that reports from other languages are not included.
A search on this word includes human casualties and damages
from lightning, but also has many other non-casualty entries
that comprise over 95% of the links. These discarded links
include sports teams, commercial products, colloquialisms and
other situations where “Lightning” appears in a news story.
Only by visually scanning each entry in the online search
engine can the event be determined if it involved people and
occurred in Africa. In most cases, the reports do not identify
the country, so it is necessary to search for the location where
the newspaper or online news reporting agency is located.
Figure 1. National lightning fatality rates per million people per year by
continent. Red shading indicates rates > 5.0 fatalities per million per year,
orange is 0.6 to 5.0, and yellow is 0.5 or less. White indicates no national
summaries have been published for periods ending in 1979 or later [updated
from 1, 2].
T
ABLE
I. P
UBLISHED
A
NNUAL
L
IGHTNING
F
ATALITY
R
ATES PER
M
ILLION
P
EOPLE AND
A
NNUAL
N
UMBER O F
F
ATALITIES BY
C
OUNTRY IN
A
FRICA
E
NDING IN
1979
OR
L
ATER
.
D
ATA
C
OLLECTION
T
YPES
:
A=M
EDICAL
R
ECORDS
;
B=P
ERSONAL
D
ATA
C
OLLECTION FROM
V
ARIETY OF
S
OURCES
;
C=P
RINT
M
EDIA
,
AND
D=M
IXTURE OF
S
OURCES
[U
PDATED FROM
1,
2].
Country Annual fatality rate Fatalities per year Data type
(deaths/million)
Burundi 2.5 26 B
Malawi 5.5 45 C
South Africa 6.3 264 A
Swaziland 15.5 15 D
Uganda 0.9 30 D
Zimbabwe 14 to 21 100 to 150 D

2
Many of the African reports are from remote areas where it
is not possible to independently check or verify them, since no
national data gathering of lightning fatalities is attempted by
any government agency in many developing nations. The
following is an estimate of the quality of the data obtained
from online search engines:
•
The location is likely to be accurate, beginning with
the country, then the local province or district, and
often the exact village or town.
•
The year and month are quite certain.
•
Numbers of deaths and injuries are usually accurate.
•
Ages and gender are often reported and likely to be
accurate when they are reported.
•
Location and activity are generally well known such as
in an agricultural field, under a tree, or in a hut or
school.
•
The day of the month and day of the week can be
ambiguous due to delays in the incidents reaching the
newsgathering agencies, and sometimes the time zones
are not well identified. Sometimes the day of the week
is specifically identified.
III. A
FRICAN
P
OPULATION AND
L
IGHTNING
A. Population
The population of Africa has increased by 19.7% from
1.049 billion in 2010 to 1.256 billion in 2017. Recent estimates
are that 60% of the population of sub-Saharan Africa is under
25 years old.
B. Lightning ocurrence
The global occurrence of lightning in Fig. 2 indicates that
lightning is very frequent in Africa south of the Sahara, but it
varies widely across the continent. By month, Fig. 3 shows
minimal variability through the year for all of Africa. For that
reason, Fig. 4 was prepared to indicate that lightning north of
the equator is primarily in the Northern Hemisphere warm
season from May through September. Conversely, areas south
of the equator have lightning in the Southern Hemisphere warm
season from November through March. The result is that for
some regions near the equator, there are two lightning seasons
during the course of the year.
IV. Z
IMBABWE
A. Events, fatalities and injuries
Zimbabwe was chosen as a first test of this approach since
it is a small lesser-developed nation that has regular English-
language reporting. Virtually nothing is known about lightning
casualties in Zimbabwe except for two estimates of 100 to 150
deaths per year [8, 9]. The web search engine can therefore be
of value in addressing general results for this developing
country from 2010 to 2017, as shown in Table II.
Figure 2. Lightning stroke density per square km per year from the Global
Lightning Dataset GLD360 network for the globe from 2013 through 2017.
The density map depicts 8,761,390,744 strokes. Scale is at lower left; grid size
is 20 by 20 km [11].
Figure 3. Lightning stroke percentages by month over all of Africa [11].
Figure 4. Lightning stroke percentages by month in Africa north and south
of the equator [11].
T
ABLE
II. S
UMMARY OF
L
IGHTNING
C
ASUALTIES IN
Z
IMBABWE
FROM
2010
TO
2017.
Number of events 57
Number of deaths 101
Number of injuries 186
Male 51
Female 63

3
This sample for Zimbabwe from the internet search engine
results in an average of 12.6 deaths per year during this period,
which is about 10% of the estimates of 100 to 150 deaths per
year in [8, 9]. The conclusion is that the internet search is
indicating a small portion of the total. Therefore, fatality rates
weighted by population cannot be determined from such data.
This dataset for Zimbabwe showing a majority of female
victims is unusual, since previous studies tend to find more
males [10]. For example, in Bangladesh it was found that 86%
of those who died due to lightning on 12 to 13 May 2016 were
males [12]. Similar research in Colombia [13] found that most
of those who died due to lightning were young males.
Furthermore, in Swaziland, 68% of those killed were males
[6]. In developed countries, the male death percentage is 65%
in the United Kingdom [14], and more males than females are
killed in the United States [15] and Australia [16].
B. Month
The month of Zimbabwe events involving one or more
deaths and/or injuries is shown in Fig. 5. The period from
November through February accounts for two-thirds of the
events. This cluster is similar to the Southern Hemisphere
curve for lightning occurrence in Fig. 4, and lends confidence
to this aspect of the data collection method.
C. Location/activity
The situation in which people were killed or injured by
lightning is shown in Fig. 6. This Zimbabwe dataset indicates
agriculture to be the most common activity [17]. Locating
under a tree to avoid rain is also common [18], followed by
being in a hut and on a school property [19]. These are useful
indications of the lightning exposure in Zimbabwe, despite the
limitations of the dataset.
D. Age
The age distribution in Fig. 7 shows a very wide range with
no specific grouping. Ages were not specified and therefore
not included in Fig. 7 for incidents where 1) many students
were killed and injured at schools and 2) a large group of
soldiers was injured.
E. Summary of Zimbabwe lightning casualties
The preceding information provides the first overview of
lightning deaths and injuries in Zimbabwe, as follows:
• The events are concentrated in the warm season
months of November through February, consistent
with the annual cycle of lightning occurrence that is
provided by data from a lightning detection network.
• Agriculture is the most frequent activity of those killed
or injured by lightning, followed by being under a tree,
in huts and on school property.
• More victims are female than male.
• The age distribution is more varied than in other
locations of the world.
Figure 5. Lightning casualty events by month in Zimbabwe from 2010 to
2017.
Figure 6. Lightning casualty events according to location and activity in
Zimbabwe from 2010 to 2017.
Figure 7. Lightning casualty events according to age in Zimbabwe from
2010 to 2017.
V. A
DDITIONAL
C
OUNTRIES
The Zimbabwe data based on web search-engine entries
have been able to provide a plausible general view of lightning
casualties in that country. Based on the success of this effort, it
is planned to expand the approach to additional nations. Those
with sufficient data at present are the following:

4
• Kenya and Zambia: No previous analyses have been
made to date of lightning casualties in these countries,
so the search-engine approach can be expected to
provide similar general information as in Section IV
for Zimbabwe.
• Burundi, South Africa, Swaziland, and Uganda:
Existing studies are mentioned in the Introduction, but
it may be possible to expand or verify those results
with the search-engine dataset.
• Malawi: A summary of recent lightning fatalities and
injuries has been completed for 16 of the 28 districts
in [4].
• Rest of Africa: None of the other countries in Africa
has had any prior lightning fatality studies published
for any length of time. The search-engine approach
has yielded relatively few reports from these nations
compared with those listed above. The reasons are 1)
news reports are not widely available in English,
and/or 2) remoteness makes any reports unlikely to
reach newsgathering agencies. Nevertheless, useful
data that may be obtained are deaths versus injuries,
gender and age, and location/activity. In addition, it
appears possible to group monthly event counts from
the nations north of the equator from those south of
the equator to compare with detected lightning.
VI. C
ONCLUSIONS
The improvement of internet search-engine data for
locating lightning casualties since about 2010 has made it
possible to gather such information from countries in Africa
where none currently exists. A sample of data from Zimbabwe
has proven to provide useful information that can be used to
address mitigation of the effects of lightning on people. This
approach can be expanded to other parts of Africa, although
the dataset is very limited in many regions due to language
and remoteness barriers.
R
EFERENCES
[1] Holle, R. L, “A summary of recent national-scale lightning fatality
studies,” Wea., Climate and Society, vol. 8, 2016, pp. 35-42.
[2] Holle R. L., “The number of documented global lightning fatalities,”
paper presented at the 33rd International Conference on Lightning
Protection, Estoril, Portugal, 2016, 4 pp.
[3] Nibigira, E., and C. Gomes, “ Lightning environment in Burundi,” paper
presented at the 32nd International Conference on Lightning Protection,
Shanghai, 2014, 4 pp.
[4] Kalindekafe, L., V. Katonda, T. K. Nthara, C. Chinsenga, P. Gomani, M.
Mkandawire, and R. L. Holle, “Lightning fatalities in Malawi: A
retrospective study from 2010 to 2017,” paper presented at the 34th
International Conference on Lightning Protection, Rzeszow, Poland,
2018, 7 pp.
[5] Blumenthal, R., “Lightning fatalities on the South African Highveld: A
retrospective descriptive study for the period 1997-2000,” Amer. J.
Forensic Med Pathology, vol. 26, 2005, pp. 66-59.
[6] Dlamini, W. M., “Lightning fatalities in Swaziland: 2000–2007,”
Natural Hazards, vol. 50, 2009, pp. 179–191.
[7] Mary, A. K., C. Gomes, A. Gomes, and W. F. W. Ahmad, “Lightning
accidents in Uganda,” paper presented at the 32nd International
Conference on Lightning Protection, Shanghai, 2014, 10 pp.
[8] Chitauro, J. J., “Welcoming speech. Discussion Section,” paper
presented at the First All-Africa International Symposium on Lightning,
Harare, Zimbabwe, 1990, 4 pp.
[9] Van Olst, M. D. A., “Minimising lightning fatalities: Lightning earth
currents in Zimbabwe,” paper presented at the First All-Africa
International Symposium on Lightning, Harare, Zimbabwe, 1990, 8 pp.
[10] Cooper, M. A., and R. L. Holle, “Reducing lightning injuries
worldwide,” Springer Natural Hazards, New York, 2018, 233 pp.
[11] Holle, R. L., R. K. Said, and W. A. Brooks, “Monthly GLD360 lightning
percentages by continent,” paper presented at the 7th International
Lightning Meteorology Conference, Fort Lauderdale, FL, Vaisala, 2018,
4 pp.
[12] Holle, R. L., and A. K. M. S. Islam, “Lightning fatalities in Bangladesh
in May 2016,” paper presented at the 8th Conference on the
Meteorological Applications of Lightning Data, Amer. Meteor. Soc.,
Seattle, WA, 2017, 4 pp.
[13] Navarrete-Aldana, N., M. A. Cooper, and R. L. Holle, “Lightning
fatalities in Colombia from 2000 to 2009,” Natural Hazards, vol. 74,
2014, pp. 1349-1362.
[14] Elsom, D. M., “Factors contributing to a long-term decrease in national
lightning fatality rates: case study of the United Kingdom with wider
implications,” Intl. J. Disaster Risk Reduction., vol. 31, 2018, pp. 341-
353.
[15] Holle, R. L., R. E. López, and B. C. Navarro, “Deaths, injuries, and
damages from lightning in the United States in the 1890s in comparison
with the 1990s,” J. Appl. Meteor., vol. 44, 2005, pp. 1563-1573.
[16] Coates L, R. Blong, and F. Siciliano, “Lightning fatalities in Australia,
1824–1991,” Natural Hazards, vol. 8, 1993, pp. 217-233.
[17] Holle, R. L., “Lightning-caused deaths and injuries related to
agriculture,” paper presented at the 33rd International Conference on
Lightning Protection, Estoril, Portugal, 2016, 6 pp.
[18] Holle, R. L., “Lightning-caused deaths and injuries in the vicinity of
trees,” paper presented at the International Conference on Lightning
Protection, Vienna, Austria, 2012, 8 pp.
[19] Holle, R. L., and M. A. Cooper, “Lightning-caused deaths and injuries at
schools,” paper presented at the 33rd International Conference on
Lightning Protection, Estoril, Portugal, 2016, 5 pp.