1. Introduction
Where, when, and at what frequency lightning occurs across the globe is steadily becoming better known by the use of a variety of sensors (Holle and Cooper 2016b; Nag et al. 2015). Lightning is due to upward vertical motions accompanied by microphysical processes at elevations with typical ambient temperatures between −5° and −15°C. The upward motion resulting from vertical instability in the atmosphere can then become strong enough to produce lightning in the relevant subfreezing layer. There is a variety of meteorological systems capable of producing vertical motion in this layer that is also modulated by large elevation changes and major land–water contrasts. About two-thirds of lightning occurs during the afternoon due to the strongest heating of the land surface during this time. At higher latitudes in both hemispheres, about two-thirds of lightning occurs during the three summer months. In tropical regions lightning tends to occur more often when the equatorial trough passes over a region (Holle and Cooper 2016b). In the Asian monsoon affecting Bangladesh, tropical moisture is brought from the south into the Indian subcontinent as the Northern Hemisphere summer arrives. Tropical regions of the world, including both land and water, are estimated to account for 78% of global lightning (Christian et al. 2003), especially where there are marked elevation changes and land–water boundaries (Albrecht et al. 2016; Holle and Murphy 2016).
There are several paths connecting cloud-to-ground lightning with people. Five mechanisms have been proposed by Cooper and Holle (2010). In the order of importance, the ground current is the largest (50%–55%) followed by side splash (30%–35%), upward streamer (10%–15%), contact injury, and direct strike (3%–5%).
In contrast to the greatly improved knowledge of the location and time of lightning occurrence, the distribution of lightning-related human fatalities and injuries is not well characterized in many regions, especially in lesser-developed nations. Holle (2016a) summarized 23 published national-scale lightning fatality statistics. That study was extended in Holle (2016c) to indicate a known total of 4176 annual lightning fatalities in 26 nations where fatalities have been documented in the last quarter century. Previous estimates are that global annual lightning fatalities range from 6000 to 24 000 with larger tolls reported in developing countries (Cardoso et al. 2014; Gomes and Ab Kadir 2011; Holle and López 2003). Such estimates are constrained by the lack of accurate statistics since there is no proper recording system of lightning death and injury in many countries (Holle 2016a). Particular attention is needed for collecting data in regions of the world where many or most people are dependent on labor-intensive agriculture (Holle 2016b) and gather in many structures such as schools that are not safe from lightning (Holle and Cooper 2016a).
In developed countries lightning fatalities have been greatly reduced during the last century as the population shifted away from labor-intensive agricultural labor; this change can be indicated in part by the movement of the population from rural to urban locations. Additional factors reducing lightning casualties in developed countries are economic advancements that provide lightning-safe structures and dwellings and readily available, fully enclosed, metal-topped vehicles (Holle 2016c). In developing nations, however, lightning is an underestimated natural hazard despite the fact that it poses considerable risk to life and property (Dlamini 2009). Here, the majority of the population continues to be engaged in subsistence agriculture for long periods (Holle 2016b), live in lightning-unsafe dwellings, and work in lightning-unsafe structures (Holle 2009).
With a current estimated population of 164 million and density of 1237 persons per square kilometer [Bangladesh Bureau of Statistics (BBS) 2016; Fig. 1], Bangladesh is impacted frequently by various natural hazards including lightning. The rural population of Bangladesh has decreased during the period of this study from 80% in 1990 to 70% in 2015 (Biswas et al. 2016). Floods and tropical cyclones are responsible for the most widespread losses of life and damages in Bangladesh and are widely reported. In contrast, the meteorological hazards of thunderstorms, hail, and tornadoes result in frequent casualties and damages on more localized time and space scales and are therefore less likely to be reported outside the specific locations where they occur. Since their effects have been overshadowed by the scale of damages caused by floods and cyclones (Ono and Schmidlin 2011), none of the relevant government agencies in Bangladesh possess information related to lightning casualty distributions. However, following the deaths of 89 people on 12 and 13 May 2016 (Holle and Islam 2017), the Bangladesh government declared lightning as a natural disaster.
Spatial distribution of average population density (persons per square kilometer). Color shading indicates population density by district, calculated as the average of values of 1991, 2001, and 2011 censuses.
Citation: Weather, Climate, and Society 9, 3; 10.1175/WCAS-D-16-0128.1
Precise information with regard to lightning casualties for all of Bangladesh has been unavailable, although three previous studies provide some insight. Gomes et al. (2006) reported 133 deaths plus 137 injuries from 73 incidents in 2005. Ono and Schmidlin (2011) estimated an annual death toll of 500 to 1000 for Bangladesh. An additional study demonstrated that lightning constitutes more than 25% of the total number of electricity-related casualties in Bangladesh (Mashreky et al. 2012). A mixture of personal interviews over parts of the country in 2003 and a media review during a 3-day period in 2016 is reported for mainly injuries by Biswas et al. (2016). A better understanding of lightning-related deaths and injuries can be expected to allow the development of public policy, greater public awareness, and safety measures (Curran et al. 2000; Roeder et al. 2015).
This work explores a variety of features of lightning-related deaths and injuries in Bangladesh from 1990 to mid-2016. Three research questions are addressed: (i) what is the geographic variation of lightning-related death and injury in Bangladesh, (ii) at what time during the day and year do these casualties occur and what are the changes in recent years, and (iii) how do the circumstances of the victims, such as age and gender, differ from those in other countries? The order of this presentation is as follows: First, there will be a description of the development of this unique lightning casualty dataset for Bangladesh. Then will follow maps by district of fatalities and injuries and associated rates. Subsequent sections will describe the time of these events followed by summaries of the demographic characteristics of the casualties.
2. Data and methods
Lightning casualties in this study refer to the sum of fatalities and nonfatal injuries to people. It was necessary to consider a number of sources to develop a national-scale database. Data employed in the development of this substantial labor-intensive project included information from the following:
two leading national daily newspapers, one in Bengali and the other in English;
one regional/local daily newspaper in Bengali;
district civil surgeon offices;
yearly/monthly disaster reports published by the Disaster Forum and Network for Information Response and Preparedness Activities on Disaster (NIRAPAD);
district police headquarters; and
historical documents and newsletters of agencies such as the Space Research and Remote Sensing Organization (SPARRSO).
Local and national newspapers were the dominant sources of information, which involved physically scanning each daily newspaper from 1990 through June 2016. As observed in prior lightning fatality studies, underreporting is a major issue due to various reasons (Holle et al. 2005; Dlamini 2009; Trengove and Jandrell 2015). As an example of the difficulty in collecting lightning casualty information, Selvi and Rajapandian (2016) found an average of 2234 deaths per year in neighboring India, Illiyas et al. (2014) found 1755 fatalities per year, while Singh and Singh (2015) found only 159 fatalities from another database for India.
Media-based information is often limited in geographic coverage (Mills et al. 2008). Because the likelihood of a report being disseminated about a single lightning event is small, especially an injury, the risk associated with lightning is not perceived to be as large a natural disaster as is the actual case (Duclos and Sanderson 1990). The result is underreporting, especially of injuries (Cherington et al. 1999; Elsom 2001; Holle et al. 2005).
A standardized form based on an extensive literature review was developed to record the location, age and gender, date and time, and activity of the casualty at the time of the incident. To avoid duplication, data were first matched using all of the demographic variables and then cross checked against the corresponding day, year, and location of other entries. Cases were excluded if any duplication was encountered, which amounted to 2% of the total casualties. Two issues with the reported cases were observed: (i) not each fatality record included all relevant attributes of interest and (ii) much more emphasis was placed on detailing fatality attributes rather than injury. Therefore, attributes related to injury were largely missing. After scrutiny the final database resulted in a total number of casualties of 5468 from January 1990 through June 2016.
A geographic information system (GIS) was used to encode lightning casualty data according to a district boundary shape file, obtained from the Bangladesh Water Resources Planning Organization (WARPO). In this study, GIS managed phenomena with reference to locations on Earth’s surface by integrating spatial and nonspatial data (e.g., attribute data) across a diverse range of applications and helped link connections between activities based on geographic proximity. Initially, it was intended to encode data according to the subdistrict (upazila) level; however, it was not possible because of missing such information in many cases. Apart from the fatality data, a number of sociodemographic variables such as population, literacy rate, and percent of agricultural population were collected from the BBS and joined to the district boundary shape file. The total population of each district was acquired from the three census years of 1991, 2001, and 2011, so that the average populations for the respective districts plus the country can be computed, while other variables represent the 2011 population and housing census values (BBS 2012).
The fatality and injury rates per million people per year, and fatality rate per area per year, were calculated within a GIS package by following the methodology of Singh and Singh (2015). To obtain death and injury rates, the number of deaths and injuries from 1990 to 2015 (the first 6 months of 2016 are not included in annual totals and rates) were first totaled and then divided by the average population of each district and number of years. A similar procedure was followed at the national scale so that a comparison with other global records can be made. Likewise, fatality data were normalized by the area of each district to determine the fatality density rate.
The temporal variations of diurnal, monthly, seasonal, and yearly casualties were derived for the entire country. The seasonal variation was computed by dividing the year into the four periods of the premonsoon (March–May), monsoon (June–September), postmonsoon (October–November), and winter (December–February).
3. Results and discussion
a. Spatial distributions of lightning-related casualties
1) Spatial distribution of population
The average population and area of each district in Bangladesh are listed in Table 1. The population density by district in Fig. 1 indicates a wide range in the distribution of population. The highest population densities are in the capital Dhaka in the center of the country and nearby districts, while the smallest population densities are scattered elsewhere.
District name, district average population in millions, district area in square kilometers, number and rank of lightning-related deaths and injuries, fatality rate per million people per year and its rank, and fatality density per million per square kilometers per year and its rank by district for Bangladesh from 1990 through June 2016.
2) Spatial distribution of fatalities and injuries
Across Bangladesh, the annual average number of fatalities and injuries for the entire period are 114 and 89, respectively. For comparison, Holle (2016c) summarized 26 recently published studies that found an average number of fatalities per year to range from only a few in many European countries to as many as 2234 in India.
The fatality and injury rates per million per year according to each district are plotted in Figs. 2 and 3. The countrywide fatality rate for the entire study period is 0.92 per million people per year (Table 1), while the injury rate is 0.71. The rates for the last 6 years are 1.6 and 1.4, respectively. This recent fatality rate approaches the 2.0 reported for nearby India (Illiyas et al. 2014; Selvi and Rajapandian 2016). Rates from additional countries are in Table 1 of Holle (2016a).
Lightning fatality rate per million people per year by district. Color shading in scale ranges from highest rates in red to lowest rates in blue. Fatality ranks by district are numbered.
Citation: Weather, Climate, and Society 9, 3; 10.1175/WCAS-D-16-0128.1
Lightning injury rate per million people per year by district. Color shading in scale ranges from highest rates in red to lowest rates in blue. Injury ranks by district are numbered.
Citation: Weather, Climate, and Society 9, 3; 10.1175/WCAS-D-16-0128.1
District fatality rates range widely from 0.20 to 4.47 per million people per year (Fig. 2). The largest rates are in the Moulvibazar district (4.47) followed by Chapai Nawabganj (3.67), Cox’s Bazar (2.78), Lalmonirhat (2.38), and Sunamganj (2.37). The smallest rates are in Shariatpur (0.20), Pirojpur (0.20), Dhaka (0.21), Jhalokati (0.22), and Sirajganj (0.29). It has been observed that weighting the number of fatalities by population sometimes shifts maxima from higher- to lesser-populated regions (Curran et al. 2000; Holle 2016a; Navarrete-Aldana et al. 2014; Zhang et al. 2011). Our results partially support this finding. While the fatality rate per million people per year in Moulvibazar ranks the highest, the population of 0.6 million is relatively small, yet there are 70 fatalities. In comparison, Chapai Nawabganj has the largest number of deaths of any district (140) and also ranks second in population-weighted fatality rate. The coastal Cox’s Bazar district also exhibits a similar pattern to Chapai Nawabganj (Table 1).
With respect to injuries, the largest rates are in Moulviabazar and Chapai Nawabganj followed by the northern district of Thakurgaon (Fig. 3). The number of injuries is less than fatalities because of a bias in the underreporting of injuries compared with fatalities. In the United States, a comprehensive review of medical records found 10 injuries per death (Cherington et al. 1999). Salerno et al. (2012) found a ratio of four injuries per death in Malawi. The ratio of less than one injury per fatality in this Bangladesh database is an indication that the sources used in developing the dataset are not complete with respect to injuries and therefore may also be incomplete with respect to fatalities. As the quality of data improves, this ratio could approach the 4:1 found in Malawi that was determined from personal on-site interviews rather than from retrospective documentation, as used in this Bangladesh work.
To address these regional differences of casualties that cannot be explained with the present dataset, a subsequent Bangladesh study is being considered by combining the following available monthly datasets: 1) population, 2) lightning frequency, and 3) crop-by-crop agricultural activity. Such a study may be able to compare lightning fatalities and injuries during labor-intensive crop cycle activities through the year as related to lightning occurrence according to districts to understand variations shown by these multiyear maps.
3) Area weighting of fatalities
The fatality data were normalized by district area in Fig. 4. Chapai Nawabganj ranks the highest followed by Cox’s Bazar. Since fatality density per square kilometer takes into account both the area and number of deaths of each district, the result generally contradicts the Indian study by Singh and Singh (2015), who observed that the maximum fatality density shifts from larger- to smaller-sized states.
Lightning fatality rate density per million per square kilometer per year by district. Color shading in scale ranges from highest rates in red to lowest rates in green. Fatality density ranks by district are numbered.
Citation: Weather, Climate, and Society 9, 3; 10.1175/WCAS-D-16-0128.1
4) Urban versus rural
During the period of this study, the national rural population decreased somewhat from 80% in 1990 to 70% in 2015 (Biswas et al. 2016). However, because of the growth in population, the actual number of residents in rural scenarios was relatively unchanged. Nevertheless out of 1605 reported fatality events, 1497 (93%) are in rural areas, whereas only 108 events (7%) are in urban Bangladesh. Biswas et al. (2016) found a similar rural percentage of 91%. This result is consistent with earlier findings (Cardoso et al. 2014; Gomes and Ab Kadir 2011; Holle 2016a; Zhang et al. 2011), which reported that the majority of fatalities in developing countries are located in rural areas. Contributing factors include extensive labor-intensive agriculture (Holle 2016b); minimal or no lightning protection provided by unsubstantial dwellings and other structures, such as schools (Holle and Cooper 2016a); a lack of readily available, fully enclosed, metal-topped vehicles; and a low level of lightning understanding (Gomes and Ab Kadir 2011; Raga et al. 2014).
b. Temporal distributions of lightning-related casualties
1) Diurnal variations of casualties
It is found that 1436 out of the 3086 fatalities occurred between early morning (0600 LST) and early evening (2000 LST; Fig. 5). Two maxima are present within this period, one in late morning (1000 LST) and the other in the midafternoon (1400 LST). A daytime maximum in lightning casualties has been observed in all prior studies of lightning casualties around the world. This maximum occurs when the heating of the land surface results in vertical instability leading to lightning-producing convection (Yamane et al. 2010; Holle and Cooper 2016b).
Number of lightning-related fatalities in Bangladesh according to local time.
Citation: Weather, Climate, and Society 9, 3; 10.1175/WCAS-D-16-0128.1
A weak morning minimum has been shown in the northeastern plains region of the Indian subcontinent for cloud-to-ground (CG) lightning by the Global Lightning Dataset GLD360 network (Nag et al. 2017). Lightning typically reaches a distinct minimum around 1000 LST (Holle 2014); however, the northeastern plains region in Nag et al. (2017) shows a much weaker morning minimum than elsewhere. More details about the seasonal and spatial extent of morning lightning in this region need to be pursued in subsequent GLD360 studies.
2) Seasonal distribution
The seasonal variation of lightning fatalities and injuries (Table 2) indicates the premonsoon season to have more casualties than others. During the premonsoon, 1916 fatalities occurred, while there were 998 during the monsoon season. Winter had the smallest number of fatalities. The premonsoon is characterized by strong, incoming solar radiation, such that thunderstorms are very frequent in this time period (Chowdhury and De 1995). Nag et al. (2017) found somewhat more CG lightning in the monsoon than the premonsoon season. Additional spatial and temporal lightning climatologies inferred from the Lightning Imaging Sensor (LIS) and Global Lightning Dataset GLD360 datasets are expected to clarify the seasonal variations.
Seasonal distribution of lightning-related fatalities and injuries for Bangladesh from 1990 through June 2016. The seasons are premonsoon (March–May), monsoon (June–September), postmonsoon (October–November) and winter (December–February).
3) Interannual distribution
The interannual variation of deaths and injuries for Bangladesh is shown in Fig. 6. Both are much more frequent between 2010 and 2015, while the smallest totals were recorded in the 1990s. The database has 30 deaths and 22 injuries per year from 1900 to 1999, 106 deaths and 72 injuries from 2000 to 2009, and 251 deaths and 220 injuries per year from 2010 to 2015. This trend is the opposite of that found in developed nations that have a dramatically decreasing trend of lightning casualties (Holle 2016a).
Interannual variation of lightning-related fatalities and injuries in Bangladesh.
Citation: Weather, Climate, and Society 9, 3; 10.1175/WCAS-D-16-0128.1
Several factors were considered to account for the recent major increase in lightning casualties. One consideration is that casualties actually changed, while another approach is to examine major changes in how casualties are being reported. Considering the former category, the population of Bangladesh increased during the period of this study from 107 million in 1990 to 163 million in 2016. This 52% change, a growth of 2% yr−1, resulted in many more people being exposed to the threat of lightning. Regarding the latter factor, the reporting of the news of lightning casualties can be expected to have been impacted by a widely recognized increase in cellular phone subscribers in the country by 47%, from 87 million in January 2012 to 128 million in July 2016, a growth of 10% yr−1 (http://www.btrc.gov.bd/content/mobile-phone-subscribers-bangladesh-july-2016). The impacts of these two changes on lightning fatalities appear to be quite important. The annual average is 144 deaths from 1990 through 2015. However, it is more representative to identify the latest 6 years with 251 deaths per year as an indication of the present fatality total. For example, Table 3 lists the top 10 1-day totals of 17 or more fatalities; many are since 2011. More specifically, exposure to agricultural activity accounted for a majority of the deaths during the latest 2 days in Table 3 of 12 and 13 May 2016 (Holle and Islam 2017). Such collections of scattered reports throughout Bangladesh, primarily in agricultural scenarios, were not reported a decade or two ago, although they likely took place. A similar recent increase was found in Turkey due to a large population growth and other issues, such that the recommendation was to use only the most recent 3 years to identify the present lightning fatality status (Tilev-Tanriover et al. 2015).
Top 10 lightning days in chronological order that caused multiple lightning-related fatalities since 1990 (figures in parentheses represent deaths in each district).
Three periods of fatalities can be identified in Fig. 6. While Romps et al. (2014) estimate a relatively modest 1% increase in lightning every 2 years due to warming temperatures, the much larger changes in Fig. 6 are likely dominated by vastly improved news reporting and a much more populous nation.
c. Demographic distributions of lightning-related casualties
1) Fatalities per incident
Table 4 indicates that a single fatality is present in 50.1% of the events (798 of the 1593 people), whereas 30.6% of the events involve two fatalities (488 victims). The remaining 19.3% of the incidents have more than two victims per incident (Table 4). Examples of multiple deaths in a single day at one or multiple locations are in Table 3. Note that Table 3 entries are not from a single flash at the same time and location; instead, a grouping was made by the media of separate events that came to their attention on the same day. For example, the 12–13 May 2016 event in Table 3 was examined by Holle and Islam (2017) and Biswas et al. (2016), where it was found to be widely dispersed across central Bangladesh during widespread occurrence of frequent lightning. For comparison, single fatalities occurred in 78% of Swaziland cases (Dlamini 2009) and 91% in the United States (Curran et al. 2000).
Number and percent of fatalities per lightning event.
2) Gender and age variations
When the gender of lightning fatalities in the database is known, Fig. 7 reveals that 73.4% are males (2265 persons), 17.2% are children (530), and 9.4% are females (291). Children are defined in Bangladesh as ≤18 years old. Injury statistics (not shown) indicate that 56.7% (1351 persons) are male, 33.7% (802) are female, and 9.6% (229 persons) are children. A similar high ratio of males was determined by Biswas et al. (2016).
Gender variation of lightning-related fatalities. Males accounted for 73.4%, females for 9.4%, and children ≤ 18 years old for 17.2%.
Citation: Weather, Climate, and Society 9, 3; 10.1175/WCAS-D-16-0128.1
Both ages and gender are known for 2870 fatalities (Fig. 8a). Of these 79.6% (2286 persons) are male and 20.4% (584) are female. The age distribution indicates that 68.9% (1575) of the male fatalities were in the age group of 10–39 years, and the same female age group had 73.1% (427). A contiguous secondary peak is in the age group of 40–49 for both males and females (Fig. 8a). Biswas et al. (2016) indicate a high incidence of injuries in the age group of 50 and above based on personal interviews. Note that the number of fatalities is more frequent for men than women in all age groups.
Age distribution by gender of (a) lightning-related fatalities and (b) population-weighted fatalities per year.
Citation: Weather, Climate, and Society 9, 3; 10.1175/WCAS-D-16-0128.1
The male population has consistently been found elsewhere to have more deaths from lightning. Elsom (2001) reported 73% male deaths in the United Kingdom, Raga et al. (2014) found 79% in Mexico, Holle et al. (2005) found 80% in the 1990s and Jensenius (2016) found 79% from 2006 to 2015 in the United States, and Mills et al. (2008) found 84% in Canada. The reason for more male deaths has been attributed to higher risk taking by young males (Badoux et al. 2016) and specifically in rural Bangladesh to outdoor work commitments of males (Dlamini 2009; Raga et al. 2014; Singh and Singh 2015).
Jaim and Hossain (2011) found that women’s participation in agriculture in Bangladesh has changed during the period of this lightning casualty study. Mechanization has greatly reduced the involvement in crop production by all workers, especially males. However, during this time there has been an increase in livestock and poultry production by females as well as homestead gardening. As mentioned elsewhere, it is beyond the scope of this paper to examine the crop-by-crop exposure to lightning, but such a study is planned by region and time of year, and gender variation will be included.
Fatalities in each age group are also weighted by population and shown in Fig. 8b. Population data are from the 2011 census since many lightning deaths occurred near that year. For males, the addition of population weighting shifts the maximum to the 30–39 age range and indicates a larger relative frequency at ages over 60 than the number of fatalities in Fig. 8a. For females, there is also an indication of more deaths per population over age 50 than indicated by fatalities in Fig. 8a. These are the first known representations of age- and gender-weighted lightning deaths by population in a developing country.
3) Socioeconomic conditions affecting activities of lightning fatalities
Examination of the activity of 3086 fatalities reveals that farming constitutes the largest portion (40%) of lightning-related deaths in the country (Table 5). Farming includes plowing, irrigation, harvesting crops, and planting crops. Another 24% of fatalities occur inside houses (Table 5). The amount of lightning protection provided by Bangladesh housing is not directly quantifiable; however, it is reasonable to conclude that many dwellings are not lightning safe. Globally about a quarter of the world’s population lives in substandard conditions that may not have lightning protection (Jacob et al. 2016). Hafiz (2000) indicates that houses in rural Bangladesh, where lightning related deaths are observed to be high, are generally made of bamboo, thatch, and mud. Similar weakly constructed housing is documented in detail in the Cox’s Bazar region of Bangladesh by Mohiuddin and Latif (2013). It can be concluded that an important but unknown portion of the rural Bangladeshi housing is vulnerable to lightning.
Activities and locations of Bangladesh fatalities due to lightning.
Another 11% of fatalities in Table 5 are attributed to returning home or walking around their homesteads or courtyards. Other factors, including repairing vehicles, clearing forests, working in construction sites, and collecting sand from rivers account for 9% of the deaths. An additional 8% are related to fishing; bathing in rivers, lakes, or ponds; and boating.
These activities are aligned with other studies in developing countries. Labor-intensive agriculture where people spend a substantial amount of time outdoors was identified in Gomes and Ab Kadir (2011) and Holle (2016b). Incidents inside dwellings have been identified elsewhere by Dlamini (2009) in Swaziland, Salerno et al. (2012) in Malawi, Cardoso et al. (2014) in Brazil, Navarrete-Aldana et al. (2014) in Colombia, and Holle (2016c) for multiple nations. Walking to and around homesteads was also observed in Swaziland (Dlamini 2009). Water activities are emphasized in China by Zhang et al. (2011), and water-related leisure activities are a frequent activity of recent lightning fatalities in the United States (Jensenius 2016). In contrast, recent United States activities of lightning fatalities are dominated by the leisure category (Jensenius 2016).
4. Conclusions
To the best of our knowledge, this is the first study to document lightning-related deaths and injuries in Bangladesh over a long period. Using a variety of sources, a national-scale database was constructed from 1990 through mid-2016. A GIS was used to estimate death and injury rates per million per year and fatality density rates per area according to district. A similar technique was used to estimate population-weighted fatality and injury rates at the national scale. It is recognized that data inhomogeneity across Bangladesh through the period of this summary is an important shortcoming of this first comprehensive survey of Bangladesh lightning casualties.
The analysis revealed annual averages of 144 fatalities and 88 injuries for the entire period, although the totals increased sharply in recent years. The full-period fatality rate is 0.92 and the injury rate is 0.71 per million people per year, but the latest 6 years have rates of 1.6 and 1.4, respectively. The urban–rural variation indicated that 93% of lightning fatalities occurred in rural Bangladesh as opposed to 7% in urban areas.
The majority of lightning-related Bangladesh fatalities occurred between early morning (0600 LST) and early evening (2000). Such an incidence is likely to be due to an overlap of outdoor agricultural activity involving a portion of the Bangladesh population with the daytime occurrence of CG lightning as the ground is heated by daytime incoming solar radiation. The premonsoon season had the largest number of casualties followed by the monsoon season, although the premonsoon has somewhat less lightning than the monsoon. The interannual variations of lightning deaths showed the smallest total fatalities to have occurred in 1992. However, fatalities have greatly increased to a current average of 251 per year since 2010 as media coverage has improved, due in part to a major increase in cellular telephone usage.
More males were killed than females. Investigation of activities showed that agriculture is the major cause of lightning deaths followed by inside a dwelling, returning home or walking around homesteads, and water-related situations. Further studies are planned to match the annual crop cycle in Bangladesh with the locations of fatalities and CG lightning.
Such information can be beneficial to agencies accountable for the management of this severe meteorological hazard. Furthermore the results of this study have the potential to develop specific public policies and lightning safety education in Bangladesh and elsewhere.
Acknowledgments
The authors very much appreciate the careful and constructive comments of four anonymous reviewers.
REFERENCES
Albrecht, R. I., S. J. Goodman, D. E. Buechler, R. J. Blakeslee, and H. J. Christian, 2016: Where are the lightning hotspots on Earth? Bull. Amer. Meteor. Soc., 97, 2051–2068, doi:10.1175/BAMS-D-14-00193.1.
Badoux, A., N. Andres, F. Techel, and C. Hegg, 2016: Natural hazard fatalities in Switzerland from 1946 to 2015. Nat. Hazards Earth Syst. Sci., 16, 2747–2768, doi:10.5194/nhess-16-2747-2016.
BBS, 2012: Housing and population census 2011. BBS, Statistics and Informatics Division, Ministry of Planning Rep., 363 pp.
BBS, 2016: Bangladesh disaster-related statistics 2015: Climate change and natural disaster perspectives. BBS, Statistics and Informatics Division, Ministry of Planning Rep., 333 pp.
Biswas, A., K. Dalal, J. Hossain, K. U. Baset, F. Rahman, and S. R. Mashreky, 2016: Lightning injury is a disaster in Bangladesh?—Exploring its magnitude and public health needs. F1000Research, 5, 2931, doi:10.12688/f1000research.9537.1.
Cardoso, I., O. Pinto Jr., I. R. C. A. Pinto, and R. Holle, 2014: Lightning casualty demographics in Brazil and their implications for safety rules. Atmos. Res., 135–136, 374–379, doi:10.1016/j.atmosres.2012.12.006.
Cherington, M., J. Walker, M. Boyson, R. Glancy, H. Hedegaard, and S. Clark, 1999: Closing the gap on the actual numbers of lightning casualties and deaths. Preprints, 11th Conf. on Applied Climatology, Dallas, TX, Amer. Meteor. Soc., 379–380.
Chowdhury, M. A. M., and U. K. De, 1995: Pre-monsoon thunderstorm activity over Bangladesh from 1983 to 1992. Terr. Atmos. Oceanic Sci. J., 6, 591–606, doi:10.3319/TAO.1995.6.4.591(A).
Christian, H. J., and Coauthors, 2003: Global frequency and distribution of lightning as observed from space by the Optical Transient Detector. J. Geophys. Res., 108, 4005, doi:10.1029/2002JD002347.
Cooper, M. A., and R. L. Holle, 2010: Mechanisms of lightning injury should affect lightning safety messages. Preprints, Third Int. Lightning Meteorology Conf., Orlando, FL, Vaisala, 5 pp. [Available online at http://www.vaisala.com/Vaisala%20Documents/Scientific%20papers/11.Cooper,%20Holle.pdf.]
Curran, E. B., R. L. Holle, and R. E. López, 2000: Lightning casualties and damages in the United States from 1959 to 1994. J. Climate, 13, 3448–3464, doi:10.1175/1520-0442(2000)013<3448:LCADIT>2.0.CO;2.
Dlamini, W. M., 2009: Lightning fatalities in Swaziland: 2000–2007. Nat. Hazards, 50, 179–191, doi:10.1007/s11069-008-9331-6.
Duclos, P. J., and L. M. Sanderson, 1990: An epidemiological description of lightning-related deaths in the United States. Int. J. Epidemiol., 19, 673–679, doi:10.1093/ije/19.3.673.
Elsom, D. M., 2001: Deaths and injuries caused by lightning in the United Kingdom: Analyses of two databases. Atmos. Res., 56, 325–334, doi:10.1016/S0169-8095(00)00083-1.
Gomes, C., and M. Z. Ab Kadir, 2011: A theoretical approach to estimate the annual lightning hazards on human beings. Atmos. Res., 101, 719–725, doi:10.1016/j.atmosres.2011.04.020.
Gomes, C., M. Z. Ab Kadir, F. Hussain, and K. R. Abeysinghe, 2006: Lightning accidents and awareness in South Asia: Experience in Sri Lanka and Bangladesh. Proc. 28th Int. Conf. on Lightning Protection, Kanasawa, Japan, ICLP, 1240–1243.
Hafiz, R., 2000: Financing housing and increasing the affordability of the rural people in Bangladesh. Proc. Third Housing and Hazards Int. Conf., Dhaka, Bangladesh, and Exeter, United Kingdom, 61–65.
Holle, R. L., 2009: Lightning-caused deaths and injuries in and near dwellings and other buildings. Fourth Conf. on the Meteorological Applications of Lightning Data, Phoenix, AZ, Amer. Meteor. Soc., JP1.3. [Available online at https://ams.confex.com/ams/89annual/techprogram/paper_144220.htm.]
Holle, R. L., 2014: Diurnal variations of NLDN-reported cloud-to-ground lightning in the United States. Mon. Wea. Rev., 142, 1037–1052, doi:10.1175/MWR-D-13-00121.1.
Holle, R. L., 2016a: A summary of recent national-scale lightning fatality studies. Wea. Climate Soc., 8, 35–42, doi:10.1175/WCAS-D-15-0032.1.
Holle, R. L., 2016b: Lightning-caused deaths and injuries related to agriculture. Preprints, 24th Int. Lightning Detection Conf. and Sixth Int. Lightning Meteorology Conf., San Diego, CA, Vaisala, 1–5. [Available online at http://www.vaisala.com/en/events/ildcilmc/archive/Pages/ILDCILMC-2016-Archive.aspx.]
Holle, R. L., 2016c: The number of documented global lightning fatalities. Preprints, 24th Int. Lightning Detection Conf. and Sixth Int. Lightning Meteorology Conf., San Diego, CA, Vaisala, 1–4. [Available online at http://www.vaisala.com/en/events/ildcilmc/archive/Pages/ILDCILMC-2016-Archive.aspx.]
Holle, R. L., and R. E. López, 2003: A comparison of current lightning death rates in the U.S. with other locations and times. Int. Conf. on Lightning and Static Electricity, Blackpool, United Kingdom, Royal Aeronautical Society, Paper 103-34 KMS, 7 pp.
Holle, R. L., and M. A. Cooper, 2016a: Lightning-caused deaths and injuries at schools. Preprints, 33rd Int. Conf. on Lightning Protection, Estoril, Portugal, ICLP, 5 pp.
Holle, R. L., and M. A. Cooper, 2016b: Lightning occurrence and social vulnerability. Atmospheric Hazards—Case Studies in Modeling, Communication, and Society Impacts, J. S. M. Coleman, Ed., InTech, 18 pp., doi:10.5772/63001.
Holle, R. L., and M. J. Murphy, 2016: Lightning over three large tropical lakes and the Strait of Malacca: Exploratory analyses. Preprints, 24th Int. Lightning Detection Conf. and Sixth Int. Lightning Meteorology Conf., San Diego, CA, Vaisala, 1–16. [Available online at http://www.vaisala.com/en/events/ildcilmc/archive/Pages/ILDCILMC-2016-Archive.aspx.]
Holle, R. L., and A. K. M. S. Islam, 2017: Lightning fatalities in Bangladesh in May 2016. Eighth Conf. on the Meteorological Application of Lightning Data, Seattle, WA, Amer. Meteor. Soc., 853. [Available online at https://ams.confex.com/ams/97Annual/webprogram/Paper304061.html.]
Holle, R. L., R. E. López, and B. C. Navarro, 2005: Deaths, injuries, and damages from lightning in the United States in the 1890s in comparison with the 1990s. J. Appl. Meteor., 44, 1563–1573, doi:10.1175/JAM2287.1.
Illiyas, F. T., K. Mohan, S. K. Mani, and A. P. Pradeepkumar, 2014: Lightning risk in India: Challenges in disaster compensation. Econ. Polit. Wkly., 49, 23–27.
Jacob, S., W. H. Siew, N. Basharat, and S. Choi, 2016: Lightning protection of a temporary structure in open area. Preprints, 33rd Int. Conf. on Lightning Protection, Estoril, Portugal, ICLP, 5 pp. [Available online at https://pure.strath.ac.uk/portal/files/53503380/Jacob_etal_2016_lightning_protection_of_a_temporary_structure_in_open_area.pdf.]
Jaim, W. M. H., and M. Hossain, 2011: Women’s participation in agriculture in Bangladesh 1988–2008: Changes and determinants. Seventh Asian Society of Agricultural Economists Int. Conf., Hanoi, Vietnam, ASAE, 16 pp.
Jensenius, J. S., 2016: NOAA’s lightning safety awareness efforts—What we’ve accomplished in 15 years. Preprints, 24th Int. Lightning Detection Conf. and Sixth Int. Lightning Meteorology Conf., San Diego, CA, Vaisala, 1–5. [Available online at http://www.vaisala.com/en/events/ildcilmc/archive/Pages/ILDCILMC-2016-Archive.aspx.]
Mashreky, S. R., M. J. Hossain, A. Rahman, A. Biswas, T. F. Khan, and F. Rahman, 2012: Epidemiology of electrical injury: Findings from a community based national survey in Bangladesh. Injury, 43, 113–116, doi:10.1016/j.injury.2011.04.013.
Mills, B., D. Unrau, C. Parkinson, B. Jones, J. Yessis, K. Spring, and L. Pentelow, 2008: Assessment of lightning-related fatality and injury risk in Canada. Nat. Hazards, 47, 157–183, doi:10.1007/s11069-007-9204-4.
Mohiuddin, M., and M. B. Latif, 2013: Housing condition of coastal area in Bangladesh: A case study of Kutubdia, Cox’s Bazaar. J. Environ. Sci. Nat. Res., 6, 15–20, doi:10.3329/jesnr.v6i1.22034.
Nag, A., M. J. Murphy, W. Schulz, and K. L. Cummins, 2015: Lightning locating systems: Insights on characteristics and validation techniques. Earth Space Sci., 2, 65–93, doi:10.1002/2014EA000051.
Nag, A., R. L. Holle, and M. J. Murphy, 2017: Cloud-to-ground lightning over the Indian subcontinent. Eighth Conf. on the Meteorological Applications of Lightning Data, Seattle, WA, Amer. Meteor. Soc., 861. [Available online at https://ams.confex.com/ams/97Annual/webprogram/Paper313608.html.]
Navarrete-Aldana, N., M. A. Cooper, and R. L. Holle, 2014: Lightning fatalities in Colombia from 2000 to 2009. Nat. Hazards, 74, 1349–1362, doi:10.1007/s11069-014-1254-9.
Ono, Y., and T. W. Schmidlin, 2011: Design and adoption of household tornado shelters for Bangladesh. Nat. Hazards, 56, 321–330, doi:10.1007/s11069-010-9571-0.
Raga, G. B., M. G. de la Parra, and B. Kucienska, 2014: Deaths by lightning in Mexico (1979–2011): Threat or vulnerability? Wea. Climate Soc., 6, 434–444, doi:10.1175/WCAS-D-13-00049.1.
Roeder, W. P., B. H. Cummins, K. L. Cummins, R. L. Holle, and W. S. Ashley, 2015: Lightning fatality risk map of the contiguous United States. Nat. Hazards, 79, 1681–1692, doi:10.1007/s11069-015-1920-6.
Romps, D. M., J. T. Seeley, D. Vollaro, and J. Molinari, 2014: Projected increase in lightning strikes in the United States due to global warming. Science, 346, 851–854, doi:10.1126/science.1259100.
Salerno, J., L. Msalu, T. Caro, and M. B. Mulder, 2012: Risk of injury and death from lightning in northern Malawi. Nat. Hazards, 62, 853–862, doi:10.1007/s11069-012-0113-9.
Selvi, S., and S. Rajapandian, 2016: Analysis of lightning hazards in India. Int. J. Disaster Risk Reduct., 19, 22–24, doi:10.1016/j.ijdrr.2016.08.021.
Singh, O., and J. Singh, 2015: Lightning fatalities over India: 1979–2011. Meteor. Appl., 22, 770–778, doi:10.1002/met.1520.
Tilev-Tanriover, S., A. Kahraman, M. Kadioglu, and D. M. Schultz, 2015: Lightning fatalities and injuries in Turkey. Nat. Hazards Earth Syst. Sci., 15, 1881–1888, doi:10.5194/nhess-15-1881-2015.
Trengove, E., and I. Jandrell, 2015: Lightning myths in southern Africa. Nat. Hazards, 77, 101–110, doi:10.1007/s11069-014-1579-4.
Yamane, Y., T. Hayashi, A. M. Dewan, and F. Akter, 2010: Severe local convective storms in Bangladesh: Part I. Climatology. Atmos. Res., 95, 400–406, doi:10.1016/j.atmosres.2009.11.004.
Zhang, W., Q. Meng, M. Ma, and Y. Zhang, 2011: Lightning casualties and damages in China from 1997 to 2009. Nat. Hazards, 57, 465–476, doi:10.1007/s11069-010-9628-0.
