Spatial–Temporal snapshots of global natural disaster impacts Revealed from EM-DAT for 1900-2015

Abstract

Natural disasters causing human, economic, or environmental losses include all types of geophysical, meteorological, hydrological, climatological, or biological events that disturb human and natural environments. Emergency preparedness for, response to, mitigation of, and recovery from natural disasters that are bases for sustainable development requires sound impact assessments at proper spatial and temporal scales. This research provides spatial–temporal views of world natural disasters recorded in the EM-DAT database for the period of 1900-2015. Views of natural disaster impacts in terms of human fatalities, injuries, affected, and property damages are summarized and ranked at the world, continent, and country levels and by decade and the whole period. Top 10, 20, and 30 out of 221 countries are highlighted and referenced with the world totals. Correlates of country disaster impacts with social-economic attributes are explored. While countries with significant disaster impacts are found in all continents, large developing countries (e.g. China, India) or developed countries (e.g. United States, Germany) lead the global natural disaster hotspots, which are small in numbers but large in total disaster losses. Such global inter-country views of natural disaster impacts provide useful insights for sustainability policy-making relevant to international and national humanitarian efforts towards natural disasters.

Keywords:

• natural disaster
• hazard
• natural hazard
• risk
• risk analysis
• emergency management

Introduction

In recent decades, we have seen a wide range of natural (e.g. floods and hurricanes) and technological disasters (e.g. hazardous material releases) across the countries. In particular, every year natural hazards have negative impacts on millions of people all over the world. Some natural hazards, which include a wide range of geophysical, meteorological, hydrological, climatological, or biological events that disturb human and natural environments, turn into disasters, causing physical impacts, such as injuries, casualties, property damages, and/or non-physical impacts, such as psychological, mental, and political wounds (Lindell 2013). More specifically, research has shown that human injuries, causalities, and economic losses due to natural disasters have been increasing over the past six decades (O’Keefe et al. 1976; Berke 1995; Mileti 1999; Dewan 2013). Therefore, a good understanding of natural hazard impacts at various spatial and temporal scales is essential for effective and efficient emergency/disaster management; preparedness, response, mitigation, and recovery for sustainable and resilient development (Shen et al. 2018).

This research examines country-level impacts caused by world natural disasters recorded in the international Emergency Events Database (EM-DAT) for the period of 1900–2015. EM-DAT, compiled by the Centre for Research on the Epidemiology of Disaster (CRED), documents all the natural disasters as a group, which contains five subgroups – geophysical, meteorological, hydrological, climatological, and biological – each of which further contains one or more types of specific natural disasters (CRED 2015). As defined in CRED (2015), those five natural disasters and their physical impacts or risks – Fatality, Injured, Affected, and Damage are described in Table 1.

Table 1. Definition and classification of natural hazards and impacts of natural disasters.

Hazard	Classification		Definition	Disaster type
Natural Hazard Types	Geophysical		A hazard originating from solid earth. This term is used interchangeably with the term geological hazard	Earthquake, Mass Movement, Volcanic activity
	Meteorological		A hazard caused by short-lived, micro- to meso-scale extreme weather and atmospheric conditions that last from minutes to days	Extreme, Temperature, Fog, Storm
	Hydrological		A hazard caused by the occurrence, movement, and distribution of surface and subsurface freshwater and saltwater.	Flood, Landslide, Wave action
	Climatological		A hazard caused by long-lived, meso- to macro-scale atmospheric processes ranging from intra-seasonal to multi-decadal climate variability	Drought, Extreme Temperature, Glacial Lake Outburst, Wildfire
	Biological		A hazard caused by the exposure to living organisms and their toxic substances (e.g., venom, mold) or vector-borne diseases that they may carry. Examples are venomous wildlife and insects, poisonous plants, and mosquitoes carrying disease-causing agents such as parasites, bacteria, or viruses (e.g. malaria)	Epidemic, Insect infestation, Animal Accident
	Extraterrestrial		A hazard caused by asteroids, meteoroids, and comets as they pass near-earth, enter the Earth’s atmosphere, and/or strike the Earth, and by changes in interplanetary conditions that effect the Earth’s magnetosphere, ionosphere, and thermosphere	Impact, Space weather
Terms			Definitions
Fatality			Number of people who lost their life because of natural hazards
Injured			People suffering from physical injuries, trauma or an illness requiring medical treatment as a direct consequence of a disaster
Affected			Sum of injured, homeless (number of people whose house is destroyed or heavily damaged and therefore need shelter after an event), and affected (people requiring immediate assistance during a period of emergency)
Damage			The amount of damage to property, crops, and livestock. In EM-DAT estimated damages are given in US$ (‘000). For individual disaster, the registered figure corresponds to the damage value at the moment of the event

Source: Adopted and modified based on CRED (2015), http://www.emdat.be/new-classification, http://www.emdat.be/explanatory-notes, http://www.emdat.be/database).

The EM-DAT database contains information from various sources, including UN agencies, national governments, the International Federation of Red Cross, and Red Crescent Societies. A natural disaster is recorded into EM-DAT if at least one of the following criteria is fulfilled: 10 or more people dead; 100 or more people affected; the declaration of a state of emergency, and a call for international assistance (CRED 2015).

The database also shows an upward trend not only in natural disaster occurrences but also in disaster impacts (e.g., causalities, property damages). For example, from 1950 to 1979, 1,779 natural disasters occurred, causing 4,860,449 casualties, 1,372,606 injuries, and $78 billion in property damages. However, from 1980 to 2015 the reported number of natural disasters increased to 11,494 or about 6.5 times as many as that in the period of 1950–1979, causing 2,599,237 fatalities, 6,354,195 injuries and $2.71 trillion in property damages (CRED 2015).

This research is based upon EM-DAT’s country specific natural disasters classifiers for the period of 1900–2015, including disaster occurrence, group, subgroup, type, year, and country and disaster impacts, including human fatalities, injuries, people affected, and damages. Disaster impacts are summarized through snapshots at spatial scales of world, continent, and country and temporal scales of decade and the total period. In addition, representative percentages, rankings, and correlations are computed and plotted to better reveal the global hotspots of natural disasters.

Specifically, human fatalities, injuries, affected, and property damages were accumulated and ranked at the world, continent, and country levels. Top 10, 20, and 30 out of 221 countries were highlighted and referenced by time (e.g. decade) and type (e.g. climatological). Additionally, this study explored whether country disaster impacts are correlated with social-economic attributes such as population, area, and gross domestic product (GDP). These global spatial and temporal views of natural disaster impacts will provide useful insights for national and international sustainability and resiliency policies.

This article is organized into six sections. Following the introduction in Section one, Section two concisely reviews the relevant literature. Section three introduces databases and methods. Section four shows the results. Section five provides conclusions and remarks.

Literature review

A hazard is regarded as an extreme and severe event that poses a potential risk to human beings and their settlements (Deyle et al. 1998; Cutter et al. 2006). According to CRED (2009a, 2009b), as shown in Table 1, natural hazards are classified into subgroups as (1) Geophysical hazards, originating from solid earth, take the forms of earthquakes, tsunamis, mass movement, and volcanic eruptions; (2) Meteorological hazards, due to short-lived micro- to meso-scale extreme weather and atmospheric conditions, occur as extreme temperature, storm, and fog; (3) Hydrological hazards, caused by the occurrence, movement, and distribution of surface and sub-surface fresh-water and salt-water, appear as flood, landslide, or wave action; (4) Climatological hazards, affected by long-lived, meso- to macro-scale atmospheric processes ranging from intra-seasonal to multi-decadal climate variability, show as drought, glacial lake outbursts, and wildfires; and (5) Biological hazards, due to the exposure to living organisms and their toxic substances or vector-borne diseases, appear as epidemic, insect infestation, or animal accident.

CRED (2015) also defines disaster impacts as (1) Fatality – number of people who lost their life because of natural hazards; (2) Injured – people suffering from physical injuries, trauma or an illness requiring medical treatment as direct consequence of a disaster; (3) Affected – sum of homeless (people whose house is destroyed or heavily damaged and therefore need shelter) and affected (people requiring immediate assistance during a disaster); and (4) Damage – the amount of damage to property, crops, and livestock. These impacts are serious threats to sustainable development.

Various methods to study natural disaster impacts exist in the literature (Cutter et al. 2006), including risk/hazards approach (White and Hass 1975); political ecology approach (O’Keefe et al. 1976); pressure and release approach (Blaikie et al. 1994; Wisner et al. 2004); hazard-of-place approach (Cutter 1996); and vulnerability/sustainability approach (Turner et al. 2003). These models have been applied to studies to enhance our understanding of impacts of various natural hazards. For example, Mustafa (1998) used the pressure and release model to investigate flood impacts, so did Gaillard et al. (2007) to analyse the impacts of typhoons. Additionally, the hazards-of-place approach suggested by Cutter (1996) was applied to various hazard studies (e.g. Cutter et al. 2000; Wu et al. 2002). The vulnerability/sustainability approach was applied for quantitative and vulnerability analysis (e.g. Tyler et al. 2007). Also, a spatial impact model for tornadoes was developed by Shen and Hwang (2015), a spatial–temporal method for assessing multi-hazard damages by Orencio and Fujii (2014), and a mapping approach for risk analysis by Beck and Kropp (2011). Moreover, extensive quantitative impact assessment models were reviewed by Cox (2001), Greenberg and Cox (2014), and Aven and Cox (2015).

It should be noted that most of the previous impact studies focus on single subgroup or type of disasters within a local area. Some are on multiple disasters for a country or a region. Such a study often starts with the area’s historical hazards, disasters, and impacts and involves risk determination, which is to identify the types of hazards and measure their historical or potential disaster impacts at various spatial and temporal levels (Murphy and Gardoni 2006; Greenberg et al. 2012). Global natural disaster impact studies are relatively few, but can be found in Dilley et al. (2005) and Cvetković and Dragicević (2014) on global natural disaster risk assessments, Peduzzi et al. (2009) for global risk indexes, Giuliani and Peduzzi (2011) on natural hazard geoportals, Peduzzi et al. (2005), Kron et al. (2012), Wirtz et al. (2014), Shi and Karsperson 2015, and Shi et al. (2016) for mapping and management of various global hazard datasets, and Shen and Hwang (2016) for country-level expected risks of technological disasters using the well-known EM-DAT data set. In addition, studies at the global level conducted in the most recent years are found in Shen and Hwang (2018) who proposed a risk analysis model of global technological disasters; Shen et al. (2018) created a risk analysis model of world natural disasters; Lesk et al. (2016) who explored how extreme weather events influence global crop production; Winsemius et al. (2016) who analysed how global future river flood risk can be calculated into climate change impacts; Carrivick and Tweed (2016) who examined the effects of the human systems (such as population pressure and land use) on the glacier outburst flood; Kirschbaum et al. (2015) who evaluated a global landslide catalogue created based on a variety of sources including media report and online database to reveal spatial and temporal trends in landslide on a global level; Dottori et al. (2018) who investigated the impacts of anthropogenic warming on human and economic losses from river flooding at the regional levels such as the Asian continent; and Bolin and Kurtz (2018) who studied how human factors such as race, class, and ethnicity are related to disaster vulnerability.

Our literature review shows that, even though various types of natural disasters have been explored, most of the existing studies focused only on one country, several countries, or world regions of Asia, America, or Europe. Moreover, in spite of these well-done studies, impact analyses of natural disasters at various spatial-temporal scales at the international level, including almost all the countries of the world, are few and far between. This research aims to fill this gap by examining global natural disaster impacts in terms of fatalities, injuries, people affected, and damages at the country level, which is ranked globally and by continent to highlight hotspot countries or regions by time and social-economic correlates. This country-level impact analysis may be coarse, especially for large countries in which intra-country variability may be quite high, but global research on inter-country disparities in impacts from country-reported natural disasters is still useful, especially for international and national humanitarian efforts towards natural disasters.

One caveat is warranted regarding the EM-DAT database. Records of natural-disaster impacts are usually affected by data availability, accuracy, and other reasons, especially for earlier years of 1900–2015 used in this research. Therefore, data collection, classification, interpretation, and recording used in EM-DAT could induce biases (Berz, 2000; WHO, 2002; Shen and Hwang, 2016). In addition, Gall et al. (2009) delineate six major biases that alone or in combination slant the interpretation of disaster loss information and eventually lead to a number of common misperceptions or fallacies about hazard events.

Data and method

First of all, straightforward data analyses with ranking, percentage, total, and classification are summarized in tables, charts, and maps. These summary views focus on top 10, 20, and 30 hotspot countries by disaster impacts in fatality, injured, affected, and damage. Similar views are also drawn at the continent level and for top countries from each continent. Finally, representative linear or nonlinear correlates between country disaster impacts and social-economic attributes, such as population, land area, and GDP are explored and reported.

Three types of databases were used as inputs. First, the natural disaster database was obtained from EM-DAT published by CRED (2015). The database was compiled from various sources, including UN agencies, non-governmental organizations, insurance companies, research institutes and press agencies and organized in a hierarchical manner from high levels to specific types of natural disasters in the world from 1900 to 2015. Each natural disaster contains important attribute such as disaster groups, sub-groups, types, sub-types, and disaster impacts – human injuries, fatalities, people affected, and damages – by space (e.g., continent, region, country) and time (e.g., year). Second, the world country boundary GIS database was drawn from Global Administrative Unit Layers (GAUL) for the world developed by the Food and Agriculture Organization (FAO) of the United Nations (FAO-UN 2014). Third, the world social-economic attributes of historical country population, area, and gross domestic product (GDP) data were extracted from the World Development Indicators database by the World Bank (2015).

Figure 1 can be summarized into the following three steps. The first step is to process tables of disaster impacts and world development indicators spatially by country/continent and temporally by year/decade/total period for natural disasters using the EM-DAT database for the period of 1900–2015. The second step is to calculate ranking, classification, percentage, and correlate for human fatalities, injuries, people affected, and damages by spatial and temporal scales. The third step is to perform GIS functions using ‘Summarize’, ‘Join’, and ‘Field Calculator’ on the tables produced in the first and second steps, and the world boundary GIS layer. This step also produces visualizations of natural disaster impact by country.

Figure 1. Data processing flowchart.

Results and analysis

A temporal view of natural disaster impacts at the world level

As shown in Figure 2(a), the reported annual human fatalities from natural disasters in the world were relatively higher in the first half of the 20 century than in the second half, and fluctuated dramatically (e.g. about hundreds to 3.7 million) until after the early 1960. Since the late 1960s, the annual average deaths decreased dramatically, though there were some fluctuations (e.g. from thousands to hundreds of thousands). It is very noticeable that the dynamics of the annual casualties, the injured in Figure 2(b), the affected in Figure 2(c), property damage in Figure 2(d), and the annual disaster occurrence (e) were clearly different. A decrease in loss of life from natural disasters might be due to improved systems for forecasting, warning, and evacuation in hazard-prone areas in a country; meanwhile, an increase in the injured, the affected, and property damage might be due to increasing interactions between human beings and natural systems under rapid urbanization with more people and buildings, especially in natural hazard prone areas in both developing and developed countries in the world (Mileti 1999; Burby 1998). It is interesting to note that the sharp increasing trend from 1960 to 2000 and the quick drop from 2000 to 2015 of occurrences show a sharp self-contradictory since the hazard and disaster monitoring and reporting technologies have been improving tremendously since 1960.

Figure 2. A temporal view of world natural disasters, 1900–2015.

Table 2 shows natural disaster impacts worldwide by decade during 1900 to 2015. The five different types of natural hazards showed a gradual increase in the number of global disaster occurrence for the entire period, and so did the annual number of the injured and affected, and damage accordingly. Conversely, the annual deaths worldwide decreased. What is noticeable is that the reported number of disaster occurrence was generally higher for both hydrological and meteorological hazards than for biological, climatological and geophysical ones; while the total number of fatalities from geophysical hazard stayed almost the same during the entire period. A dramatic decrease was seen in the total number of fatalities for biological and climatological hazards as times went by (due perhaps to improved public health, weather prediction and warning systems), even though an increase was seen in the number for climatological hazards due to a couple of low-probability and high-impact events (e.g., Bangladesh and India droughts of 1943). In general, people affected were relatively higher for climatological and hydrological hazards than for the other three while the property damage was relatively higher for geophysical, hydrological, and meteorological hazards than for the remaining two.

Table 2. Natural disaster impacts by decade, 1900–2015.

Decade	Occurrence					Fatality					Injured					Affected ($1,000)					Damage ($1,000,000)
	Β	C	G	H	M	Β	C	G	H	M	Β	C	G	H	M	Β	C	G	H	M	Β	C	G	H	M
1900–1909	5	4	42	7	14	3,000,070	1,261,000	218,082	681	18,014	0	0	41	0	0	0	0	240	0	0	0	0	777	480	50
1910–1919	7	12	30	4	18	3,009,700	86,073	69,392	101,379	59,948	0	0	1,816	0	200	2,000	32	2	3,635	98	0	100	110	0	390
1920–1929	10	6	33	13	34	3,323,000	4,724,103	554,492	4,682	119,988	0	0	105,763	0	10	18,000	25,011	217	8	1,106	0	8	688	70	253
1930–1939	3	3	38	16	41	2,016	71	240,911	4,362,428	95,528	0	0	61,782	0	6,500	0	29	98	13,630	165	0	0	1,044	1,838	466
1940–1949	4	17	52	15	53	11,776	3,450,250	163,992	118,557	127,120	0	0	48,675	0	300	15	2	307	2	2,534	0	30	1,445	915	619
1950–1959	2	1	61	101	128	481	11	25,975	2,063,848	36,753	0	0	12,459	0	1,563	80	0	247	13,184	6,167	0	0	823	1,799	3,436
1960–1969	37	56	89	179	221	9,467	1,510,724	57,776	37,415	135,058	0	100	60,664	123,577	624,333	478	117,905	4,440	46,376	30246	0	1,077	4,885	4,995	7,488
1970–1979	64	91	126	316	306	9,913	119,086	440,821	58,155	358,892	0	200	482,187	15,741	51,682	2,239	263,749	17,108	207,974	52,540	1	6236	16,942	8,888	21,780
1980–1989	171	186	245	625	597	43,895	557,664	85,426	57,770	52,068	0	584	131,624	110,583	72,190	1,686	593,862	33,648	471,927	141,446	0	20,908	64,564	46,149	53,860
1990–1999	395	240	335	1,010	991	91,702	3,972	105,426	103,803	220,488	79,775	694	341,892	825,756	301,037	13,537	313,290	26,675	1,438,640	231,206	109	59,113	196,109	214,877	229,331
2000–2009	616	318	358	1,918	1,277	58,573	1,791	454,149	61,580	263,281	1,870	2,340	954,683	229,399	2,093,362	5,445	757,757	84,066	952,057	482,605	120	45,749	174,370	167,421	503,704
2010–2015	117	92	131	697	473	16,370	20,232	249,900	32,450	77,180	404,609	1,465	401,024	29,213	139,922	4,781	107,340	18,912	424,388	121,683	0	45,873	305,324	201,563	219,421

B = Biological, C = Climatological, G = Geological, H = Hydrological, M = Meteorological.

Table 2 contains some zero entries or extremely small numbers, though the zero entries may not be true zeros according to CRED (2015) due to various data collection and processing issues. More zeros and small values occurred more for injury and damage, more for early years (e.g. prior to 1960), and more for biological, climatological, and hydrological disasters. In fact, the EM-DAT has few zeros for all losses and all disaster types since 1960s, especially since 1990. The zeros and extremely small values in EM-DAT raised some quality concerns about and clarifications for the database. Nevertheless, EM-DAT is still one of the most comprehensive and widely used public databases on global natural disasters (Dilley et al. 2005; Peduzzi et al. 2009; Wirtz et al. 2014; CRED 2015).

A spatial view of natural disaster impacts at the country level

Figure 3 maps 221 countries of the world with the top (in darker brown colors) and bottom (in lighter yellow colors) countries to represent the frequency of natural disasters and their impacts while Table 3 shows top 30 countries ranked by the frequency and impacts (fatality, injured, affected, and damage) and lists cumulative shares of the top 10, 20, and 30 countries with respect to the world total for the period of 1900–2015. There are several interesting features worthy of some discussions. First, the natural disaster occurrence in top countries include the world’s large or most populated countries such as the United States, China, India, Philippines, Indonesia, Bangladesh, and Japan. These top countries spread all over continents. On the other hand, the bottom countries where only a single natural disaster was reported during 1900–2015 include small or less-populated countries such as St. Helena, Palau, French Guiana, Equatorial Guinea, and Brunei Darussalam. Second, the top four countries with over 2 million people killed from different natural disasters are found in Asian countries (e.g. China, India, and Bangladesh), while the bottom countries where no person was killed include small or less-populated countries such as Macau, Finland, Virgin Islands (UK), Brunei Darussalam, French Guiana, Palau, and St. Helena. Third, the spatial patterns of the number of the injured and the affected from natural disasters roughly echo those of fatalities, except for several countries. Fourth, the top six countries in economic damage include the developed countries such as the United States, Japan, Italy, Germany and developing China and India, whereas Africa was a less naturally disastrous continent except for a few countries (e.g. Algeria, Egypt, Sudan, Nigeria, South Africa, Mozambique). Fifth, China was among the top two in disaster occurrence and impacts.

Figure 3. A spatial view of natural disaster impacts by country, 1900–2015. Source: Author

Table 3. Top 30 countries ranked by total disaster impact.

Rank	: Country	Occurrence	Country	Fatality	Country	Injured	Country	Affected	Country	Damage(S1K)
1	United States	934	China	12,722,367	Peru	2,064,155	China	3,166,731,184	United States	789,206,310
2	China	818	India	9,128,538	China	1,681,429	India	2,044,391,121	China	461,306,381
3	India	662	Soviet Union	3,868,439	Bangladesh	1,044,957	Bangladesh	431,608,158	Japan	436,887,500
4	Philippines	593	Bangladesh	2,992,779	Haiti	582,593	Philippines	202,647,266	Italy	87,472,263
5	Indonesia	464	Ethiopia	416,201	Indonesia	431,668	Brazil	105,332,925	India	84,580,506
6	Bangladesh	330	Japan	244,706	Japan	274,649	Pakistan	91,14,780	Germany	54,402,130
7	Japan	328	Haiti	242,396	India	246,089	Thailand	90,920,886	Australia	48,578,873
8	Mexico	247	Indonesia	241,544	Philippines	211,666	Viet Nam	84,865,580	Thailand	47,787,747
9	Australia	226	Uganda	204,697	Iran	173,710	Ethiopia	79,797,859	Mexico	44,039,210
10	Brazil	220	Niger	195,198	Pakistan	162,929	Kenya	58,898,815	France	41,121,700
11	Iran	208	Pakistan	176,464	Turkey	98,676	Iran	44,468,180	Chile	40,726,970
12	Viet Nam	197	Sudan	162,617	Guatemala	78,781	Sudan	38,935,498	England	38,587,180
13	Pakistan	190	Iran	156,329	Chile	77,474	Korea	34,078,030	Indonesia	29,260,308
14	Afghanistan	170	Myanmar	146,392	Afghanistan	70,704	Mozambique	31,300,097	Pakistan	28,295,969
15	Colombia	170	Italy	140,538	Tanzania	55,368	Indonesia	29,268,710	Canada	27,740,600
16	Peru	169	Mozambique	106,114	El Salvador	47,304	Niger	28,217,750	Spain	27,525,297
17	Turkey	161	Peru	96,568	Mexico	37,105	Sri Lanka	28,212,225	Turkey	26,910,100
18	Russia	157	Turkey	92,125	Argentina	36,096	United States	27,756,578	New Zealand	26,714,035
19	France	151	Cabo Verde	85,295	Ghana	29,927	Malawi	27,306,458	Philippines	25,930,746
20	Italy	136	Guatemala	84,689	Sri Lanka	29,354	Soviet Union	25,428,794	Korea (North)	23,653,310
21	Hong Kong	134	Philippines	69,620	United States	28,215	South Africa	21,512,482	Iran	23,432,696
22	Thailand	131	Russia	61,624	Nepal	27257	Cambodia	20,421,449	Brazil	22,175,381
23	Canada	127	Chile	61,413	Morocco	25,985	Zimbabwe	19,479,915	Taiwan	21,329,712
24	Nepal	120	Canada	51,871	Colombia	24,174	Japan	19,010,675	Soviet Union	20,712,700
25	Congo	114	Somalia	49,054	Algeria	21,992	Peru	18,970,450	Bangladesh	18,350,780
26	Chile	112	United States	44,008	Nicaragua	21,467	Mexico	18,954,319	Korea (South)	16,067,257
27	Nigeria	112	Sri Lanka	39,079	Taiwan	21,166	Colombia	18,011,357	Greece	13,076,659
28	Ethiopia	111	Colombia	33,756	Myanmar	20,762	Somalia	17,372,462	Algeria	11,814,846
29	Haiti	110	Nepal	32,786	Yugoslavia	20,059	Ghana	16,505,767	Cuba	11,641,697
30	Korea (South)	110	Venezuela	31,284	Sudan	19,420	Australia	16,196,062	Russia	11,258,996
	221 Countries	13,757		32,583,037		7,951,888		7,283,897,190		2,799,344,224
	Country Group	%	Country Group	%	Country Group	%	Country Group	%	Country Group	%
	Top 10	35.05%	Top 10	92.86%	Top 10	86.44%	Top 10	87.27%	Top 10	74.85%
	Τop 20	47.47%	Top 20	96.69%	Top 20	93.50%	Τop 20	91.59%	Top 20	85.40%
	Top 30	56.06%	Top 30	98. 14%	Top 30	96.39%	Top 30	94.15%	Top 30	91.47%
	Middle 161	42.62%	Middle 161	1.86%	Middle 161	3.61%	Middle 161	5.85%	Middle 161	8.53%
	Bottom 30	0.81%	Bottom 30	0.00%	Bottom 30	0.00%	Bottom 30	0.00%	Bottom 30	0.00%

Table 3 clearly shows that the share of the top 10, 20, and 30 countries ranged from about 35% to 98%: Occurrence (35.02%, 47.47%, and 56.06%), Fatality (92.86%, 96.69%, and 98.14%), Injured (86.44%, 93.50%, and 96.39%), Affected (87.27%, 91.59%, and 94.15%), and Damage (74.85%, 85.40%, and 91.47%). The middle 161 countries (except for the top 30 and the bottom 30 countries) had 42.62% in disaster occurrence, 1.86% for fatality, 3.61% for injured, 5.85% for affected, and 8.53% for damage. The bottom 30 countries had almost zero disaster impacts. These percentages suggest that the top 30 disaster countries share over 90% disaster impacts and many of them have a large population (e.g. China, India), fast growing pace (e.g. China, India, and Brazil), quick industrialization (e.g. China, India, and South Korea), or an advanced economy (e.g. the United States, Japan, and Germany). Also, they are all over the continents and more than 1/3 countries are in Asia, especially China, India, Bangladesh, Japan, Indonesia, and Philippines.

Table 4 shows the top 10 countries ranked by occurrence and impact by natural disaster subgroup. Top countries by each disaster impact were found in all continents, with some leading countries such as China, India, and Bangladesh in Asia; the United States, Canada, and Mexico in North America; Columbia, Brazil, Chile, and Peru in South America; Sudan, Ethiopia, and South Africa in Africa; Iran and Turkey in Middle East or Eurasia; and Germany, Russia, France, and Italy in Europe, and Australia in Oceania. Some countries are considerably riskier than others when multiple disaster impacts are considered together, such as China, India, Bangladesh, Indonesia, the United States. Again, large developing and large developed countries were hot-spots for natural disaster impacts.

Table 4. Top 10 countries of disaster impacts by natural disaster subgroup.

Biological		Climatological		Geophysical		Hydrological		Meteorological
Country	Occurrence	Country	Occurrence	Country	Occurrence	Country	Occurrence	Country	Occurrence
Congo	69	United States	87	Indonesia	172	China	326	United States	620
India	69	Australia	47	China	162	India	315	Philippines	333
Nigeria	58	China	42	lran	106	Indonesia	225	China	277
Niger	47	Hong Kong	39	Turkey	78	United States	179	India	229
Sudan	42	Russia	29	Japan	75	Philippines	173	Bangladesh	193
Uganda	39	Canada	26	Philippines	57	Brazil	154	Japan	176
Indonesia	35	Brazil	21	Peru	48	Colombia	113	Mexico	116
Tanzania	34	Indonesia	20	United States	43	Pakistan	105	Australia	109
Kenya	32	South Africa	18	Mexico	42	Afghanistan	95	Viet Nam	98
Bangladesh	30	Spain	18	Chile	40	Bansladesh	93	Taiwan	77
Country	Fatality	Country	Fatality	Country	Fatality	Country	Fatality	Country	Fatality
India	4,543,874	India	4250,326	China	876,978	China	6,605,379	Bangladesh	637,103
Soviet Union	2,500,000	China	3,503,799	Haiti	222576	India	74,061	India	181,918
China	1,561,498	Bangladesh	1,900,018	Indonesia	216,928	Bangladesh	52,434	China	174,713
Bangladesh	403,188	Soviet Union	1200,000	Japan	194,948	Guatemala	41,844	Myanmar	144,666
Uganda	203,664	Ethiopia	402,367	Soviet Union	155,545	Venezuela	30,560	Russia	58,126
Niger	109,859	Sudan	150,047	Iran	147,475	Pakistan	17,539	Philippines	48,987
Canada	50,562	Mozambique	100,117	Pakistan	143,790	Japan	14212	United States	35,851
Nigeria	24,S99	Niger	85,000	Italy	116,380	Soviet Union	12,430	Japan	35,545
Burkina Faso	17,101	Cabo Verde	85,000	Turkey	89,497	Peru	10,485	France	24,660
Ethiopia	11,726	Somalia	39,673	India	78,359	Brazil	9,467	Honduras	24,621
Country	Injured	Country	Injured	Country	Injured	Country	Injured	Country	Injured
Haiti	277,451	Russia	2,319	China	667,547	China	828,617	Peru	1,824,992
Philippines	124,139	Australia	1,134	Haiti	300,000	Indonesia	255,737	Bangladesh	941,340
Peru	79,725	United States	639	India	220,376	Bangladesh	102,992	China	185,041
Tanzania	54,953	Chile	532	Indonesia	175210	Sudan	19375	Japan	116,393
Ghana	29,861	South Africa	530	Iran	172,879	El Salvador	18,000	Philippines	70,809
Mozambique	5,118	Indonesia	478	Japan	157274	Yugoslavia	15,000	Myanmar	20,280
Niger	5,009	China	224	Peru	156,047	Brail	12,830	India	17,915
Kenya	3,459	Benin	200	Pakistan	150212	Pakistan	10517	United States	13,857
Cameroon	2,314	Portugal	186	Turkey	97,682	India	7,798	Ukraine	13,353
Iraq	2,217	France	161	Guatemala	77,711	Russia	7,571	Viet Nam	12227
Country	Affected	Country	Affected	Country	Affected	Country	Affected	Country	Affected
Soviet Union	18,000,000	India	1,061,841,000	China	75,461,140	China	1,999,673,896	China	573,529,927
Kenya	6,881,995	China	518,056,392	India	28,333,879	Inda	845,901,204	Philippines	159,251,382
Bangladesh	3,042,429	Brazil	78,824,000	Indonesia	10,276,188	Bangladesh	322475,704	India	107,893,565
Madagascar	2,341,425	Ethiopia	76,941,879	Chile	9,932,528	Pakistan	79,290,116	Bangladesh	81,369,525
Canada	2,008,917	Kenya	48,800,000	Philippines	7,578,787	Thailand	55,581,872	Viet Nam	47,299,781
Japan	2,000,534	Iran	37,625,000	Pakistan	7,125,009	Viet Nam	31,427,045	United States	14,112,697
Brazil	1,984,376	Sudan	31,510,000	Turkey	6,827,716	Philippines	29238,107	Cuba	11,885,064
Burundi	1,377,966	Thailand	29,982,602	Guatemala	6,400204	Brail	24276,029	Madagascar	9,734,556
Congo	715,614	Niger	26,243,186	Nepal	6,344,880	Colombia	16,273,837	Mexico	8,719,703
Indonesia	689,029	Bangladesh	25,002,000	Peru	6,102122	Sri Lanka	14,900,934	Australia	8,700,154
Country	Damage*	Country	Damage*	Country	Damage*	Country	Damage*	Country	Damage*
Australia	120,000	United States	62,680,600	Japan	359,794,400	China	219242,796	United States	616,127,780
Colombia	104,000	China	29,870,420	China	110,288,857	United States	67,796260	China	101,904,308
Madagascar	3,500	Australia	14,059,194	Italy	4928 7,952	India	56,861,688	Japan	63,472,800
Cameroon	1,700	Spain	13,414,108	United States	42601,670	Thailand	45390,408	Mexico	31,510,610
Philippines	925	Canada	11,272,500	Chile	36,327,070	Germany	26359,830	France	30,895,060
Nicaragua	7	Brazil	11219,100	New Zealand	24,787,669	Italy	26331,810	Germany	27,780,300
Soviet Union	0	Indonesia	10,489,200	Turkey	24,685,400	England	23,771230	Philippines	21208258
Kenya	0	Portugal	4,918,136	Soviet Union	16,869,800	Pakistan	20586,178	India	20,052,996
Bangladesh	0	Russia	4,860,156	Taiwan	15,127,900	Korea (North)	17,542,800	Australia	19271,304
Canada	0	Italy	3,690,000	Indonesia	12,227,316	Australia	14,118,700	England	14,755,950

* $1K.

Table 5 shows the top 10–30, bottom 10–30, and middle 65–132 for the total 221 countries by natural disaster occurrence, impact, and subgroup. Hydrological disasters (5,228) were ranked first in occurrence, followed by meteorological disasters (4,376), geophysical disasters (1,599), biological disasters (1,468), and climatological disasters (1,086). Even though climatological and biological disasters happened less, they caused more fatalities (e.g. climatological – 11.74 million; biological – 9.59 million) than any other disaster type did (e.g. hydrological – 7.01 million; geophysical – 2.68 million; meteorological – 1.57 million). Meanwhile, meteorological disasters produced the most injuries – 3.33 million, followed by geophysical – 2.68 million, hydrological −1.34 million, and biological −0.59 M. Hydrological disasters ranked first for the number of people affected – 3.64 billion. This was followed by climatological – 2.30 billion, and meteorological – 1.11 billion. Finally, meteorological disasters generated the largest damages – $11.09 trillion, followed by geophysical – $7.80 trillion, hydrological – $7.10 trillion, and climatological – $2.01 trillion. These figures show that different kinds of natural hazards could have quite different disaster impacts, for example, while biological or climatological hazards were more deadly with most casualties, meteorological and geophysical hazards were more costly with highest economic losses.

Table 5. Disaster impacts by disaster subgroup and country group.

Subgroup	Country Group	Occurrence	Fatality	Injured	Affected	Damage
	Top 10	21.73%	98.32%	99.16%	81.31%	100.00%
	Top 20	38.90%	99.19%	99.94%	89.76%	100.00%
	Top 30	54.90%	99.63%	100.00%	94.36%	100.00%
	Middle 95	42.98%	0.37%	0.00%	5.63%	0.00%
Biological	155 Countries	1,468	9,587,151	589,168	48,018,597	230,132
	Top 10	13.44%	99.84%	3.62%	77.32%	16.15%
	Top 20	33.15%	99.98%	35.97%	82.18%	59.58%
	Top 30	45.49%	99.99%	80.67%	83.80%	72.83%
	Middle 96	47.70%	0.01%	19.33%	16.20%	27.17%
Climatological	156 Countries	1,086	11,735,092	6,959	2,296,896,283	201,025,928
	Top 10	45.53%	83.77%	73.33%	71.18%	77.39%
	Top 20	60.79%	95.92%	89.31%	91.20%	85.66%
	Top 30	75.92%	99.12%	97.96%	96.84%	94.20%
	Middle 65	22.20%	0.88%	2.03%	3.15%	5.80%
Geophysical	125 Countries	1,599	2,676,924	2,687,269	194,775,974	779,891,095
	Top 10	23.66%	97.95%	72.40%	90.30%	48.11%
	Top 20	43.00%	98.87%	92.81%	93.21%	53.23%
	Top 30	54.40%	99.30%	93.43%	96.04%	75.85%
	Middle 130	44.19%	0.70%	6.57%	3.96%	24.15%
Hydrological	190 Countries	5,228	7,012,063	1,341,931	3,638,662,441	709,535,497
	Top 10	45.27%	86.92%	41.53%	86.46%	78.45%
	Top 20	58.55%	95.58%	42.40%	93.63%	87.44%
	Top 30	64.69%	97.56%	97.81%	95.53%	89.77%
	Middle 132	33.48%	2.44%	2.19%	4.47%	10.23%
Meteorological	192 Countries	4,376	1,571,807	3,326,561	1,105,543,895	1,108,661,572
All Subgroup	221 Countries	13,757	32,583,037	7,951,888	7,283,897,190	2,799,344,224

* Damage = percentage or $1K.

Meanwhile, the occurrence percentages in Table 5 are based on ranking of fatality. Other percentages are based on disaster impacts. For fatality occurrence, the top 10, 20, and 30 countries accounted for 21.73%-54.90% for biological, 13.44%-45.49% for climatological, 45.54%-75.92% for geophysical, 23.66%-54.44% for hydrological, and 45.27%-64.69% for meteorological, while the middle 65-132 countries shared from low 22.20% for geophysical to high 47.70% for climatological, indicating disasters occurred mostly in the top 30 countries, especially geophysical and meteorological disasters with nearly a half happened in top 10 countries.

Further, among the five disaster subgroups, the top 10 countries had 71.18% (affected by geophysical) -100% (damage by biological) shares of their four impacts, except 3.62% injured and 16.15% damage for climatological, 41.53% injured for meteorological, and 48.11% damage for hydrological disasters. The top 20 countries had more than 53.23% (damage by hydrological) except 35.95% and 42.40% injured by climatological and meteorological disasters. Also, the top 30 countries experienced more than 80% impacts except 72.83% and 75.85% damage by climatological and hydrological hazards. These figures again indicate the concentration of disaster impacts in a relatively small number of countries.

A continental view of natural disaster impacts

Table 6 compares natural disaster impacts by continent. Asia is clearly the top continent for disaster happenings and leads in the world in damage (46.13% or $1.29 trillion), injured (57.27% or 4.55 million), fatality (80.36% or 26.18 million), and affected (87.65% or 6.39 billion). The next is Americas continent, which was more impacted in injury (38.49% or 3.06 million) and damage (36.53% or $1.02 trillion). Europe came in third and was more impacted in fatality (12.75% or 4.15 million) and damage (13.43% or $376 billion). Oceania continent, which consists primarily Australia, New Zealand, had the least impacts, especially in fatality (0.06% or 19,304), injured (0.14% or 10,963), and affected (0.33% or 24.14 million). Interestingly, Africa, the relatively least developed continent, shared 18.28% or 2,515 of world natural disaster occurrence, but had the second lowest disaster impacts, especially in injured (2.68% or 212,979) and damage (1.08% or $30.19 billion).

Table 6. Natural disaster impact by continent.

Continent	Occurrence	Fatality	Injured	Affected	Damage ($1K)	Occurrence (%)	Fatality (%)	Injured (%)	Affected (%)	Damage (%)
Africa	2,515	1,407,287	212,979	501,335,099	30,194,142	18.28	4.32	2.68	6.88	1.08
Americas	3,426	819,587	3,060,899	306,522,733	1,022,616,622	24.90	2.52	38.49	4.21	36.53
Asia	5,469	26,182,130			1291250305	39.75	80.36	57.27	87.65	46.13
Europe	1,717	4,154,729	114221	67,633,665	375,940,647	12.48	12.75	1.44	0.93	13.43
Oceania	631	19,304	10,963	24,137,374	79,375,308	4.59	0.06	0.14	0.33	2.84
221 Countries	13,757	32,583,037	7,951,888	7283,897,190	2,799344224	100.00	100.00	100.00	100.00	100.00

A view of top countries in disaster impacts

Table 7 illustrates the natural disaster occurrences and losses for the top country in each continent by disaster subgroup and type. First of all, China as a single country had large shares of people affected (43.48% or 3.17 billion), fatality (39.05% or 12.72 million), and injured (21.15% or 1.68 million) in the world over the period of 1900–2015. These shares came with 5.95% or 818 of the world total disaster occurrence. While with the most disaster happenings at 6.79% or 934, the United States had relatively low fatality (0.14% or 44,008), injured (0.35% or 28,215), and people affected (0.38% or 27.76 million). However, the United States had the highest economic loss in total damage (28.19% or $7.89 trillion). Following China and the US were Iran, Russia and Ethiopia in disaster impacts.

Table 7. Natural disaster impact by disaster type for top countries in each continent.

Subgroup	Type	Country	Occu	Fatality	Injured	Affected	Damage*	Occurrence (%)	Fatality (%)	Injured (%)	Affected (%)	Damage (%)
Β	Epidemic	China	10	1,561,498	0	9,829	0	1.22	12.27	0.00	0.00	0.00
Β	Insect infestation	China	1	0	0	0	0	0.12	0.00	0.00	0.00	0.00
C	Drought	China	36	3,503,534	0	518,000,000	29,760,420	4.40	27.54	0.00	16.36	6.45
C	Wildfire	China	6	265	224	56,392	110,000	0.73	0.00	0.01	0.00	0.02
G	Earthquake	China	155	876,478	667,412	75,455,500	110,280,857	18.95	6.89	39.69	238	23.91
G	Mass movement (dry)	China	7	500	135	5340	8,000	0.86	0.00	0.01	0.00	0.00
H	Rood	China	261	6,599,551	826,906	1997,433,161	217,392,396	31.91	51.88	49.18	63.08	47.13
H	Landslide	China	65	5,528	1,711	2240,735	1,850,400	7.95	0.04	0.10	0.07	0.40
M	Extreme temperature	China	13	379	4,822	81,215,180	21,430,200	1.59	0.00	0.29	2.56	4.65
M	Storm	China	264	174,334	180,219	492314,747	80,474,108	32.27	1.37	10.72	15.55	17.44
Β	Epidemic	Ethiopia	25	11,726	0	192,447	0	22.52	2.82	0.00	0.24	0.00
Β	Insect infestation	Ethiopia	4	0	0	0	0	3.60	0.00	0.00	0.00	0.00
C	Drought	Ethiopia	16	402,367	0	76,941,879	92,600	14.41	96.68	0.00	96.42	77.89
C	Wildfire	Ethiopia	1	0	5	0	0	0.90	0.00	1.44	0.00	0.00
G	Earthquake	Ethiopia	7	24	165	420	7,070	6.31	0.01	47.55	0.00	5.95
G	Mass movement (dry)	Ethiopia	1	13	0	0	0	0.90	0.00	0.00	0.00	0.00
G	Volcanic activity	Ethiopia	3	69	0	11,000	0	2.70	0.02	0.00	0.01	0.00
H	Rood	Ethiopia	52	1,976	167	2,651,929	19,220	46.85	0.47	48.13	332	16.17
H	Landslide	Ethiopia	2	26	10	184	0	1.80	0.01	2.88	0.00	0.00
Β	Epidemic	Iran	3	372	0	2,500	0	1.44	0.24	0.00	0.01	0.00
C	Drought	Iran	2	0	0	37,625,000	3,300,000	0.96	0.00	0.00	84.61	14.08
C	Wildfire	Iran	1	0	0	0	0	0.48	0.00	0.00	0.00	0.00
G	Earthquake	Iran	106	147,475	172,879	2,518,609	11,826,628	50.96	94.34	99.52	5.66	50.47
H	Rood	Iran	79	7,860	693	4,117,271	8,277,528	37.98	5.03	0.40	926	35 32
H	Landslide	Iran	4	116	44	100	0	1.92	0.07	0.03	0.00	0.00
M	Extreme temperature	Iran	1	158	0	0	0	0.48	0.10	0.00	0.00	0.00
M	Storm	Iran	12	348	94	204,700	28,540	5.77	0.22	0.05	0.46	0.12
Β	Epidemic	Russia	10	33	0	158,246	0	6.33	0.05	0.00	3.38	0.00
Β	Insect infestation	Russia	1	0	0	0	0	0.63	0.00	0.00	0.00	0.00
C	Drought	Russia	5	0	0	1,000,000	2,540,000	3.16	0.00	0.00	2134	22.49
C	Wildfire	Russia	24	169	2319	113,855	2320,156	15.19	0.27	14.65	2.43	20.55
Ε	Impact	Russia	1	0	1,491	300,000	33,000	0.63	0.00	9.42	6.40	029
G	Earthquake	Russia	11	2,066	1,177	59,013	551,520	6.96	3.35	7.43	1.26	4.88
G	Mass movement (dry)	Russia	2	60	0	1,750	2,600	1.27	0.10	0.00	0.04	0.02
H	Rood	Russia	56	756	7,563	2276,047	4,145,570	35.44	1.23	47.77	48.56	36.71
H	Landslide	Russia	7	414	8	800	0	4.43	0.67	0.05	0.02	0.00
M	Extreme temperature	Russia	21	57,914	2,998	756200	1,400,100	13.29	93.98	18.94	16.13	12.40
M	Storm	Russia	20	212	276	21,012	299,050	12.66	0.34	1.74	0.45	2.65
Β	Epidemic	United States		317	0	406,703	0	0.54	0.72	0.00	1.47	0.00
C	Drought	United States	14	0	0	0	41,835,000	1.50	0.00	0.00	0.00	5.30
C	Wildfire	United States	73	1,281	639	859,854	20,845,600	7.82	2.91	226	3.10	2.64
G	Earthquake	United States	41	2,587	13,294	61,953	41,741,670	4.39	6.56	47.12	022	529
G	Volcanic activity	United States	2	90	0	2,500	860,000	0.21	0.20	0.00	0.01	0.11
H	Rood	United States	174	2,924	410	12,312,596	67,776260	18.63	6.64	1.45	44.36	8.59
H	Landslide	United States	5	658	15	275	20,000	0.54	1.50	0.05	0.00	0.00
M	Extreme temperature	United States	36	5,201	31	0	16,085,000	3.85	11.82	0.11	0.00	2.04
M	Storm	United States	584	30,650	13,826	14,112,697	600,042,780	62.53	69.65	49.00	5054	76.03
All Sub	All Type	China	818	12,722367	1,681,429	3,166,731,184	461,306,381	5,95	5.95	39.05	21.15	43.48
All Sub	AU Type	Ethiopia	111	416,201	347	79,797,859	118,890	0.81	1.28	0.00	1.10	0.00
Αll Sub	All Type	Iran	208	156329	173,710	44,468,180	23,432,696	1.51	0.48	2.18	0.61	0.84
All Sub	All Type	Russia	158	61,624	15,832	4,686,923	11291,996	1.15	0.19	0.20	0.06	0.40
All Sub	All Type	United States	934	44,008	28,215	27,756578	789,206,310	6.79	0.14	0.35	0.38	28.19
All Sub	All Type	221 Countries	13,757	32,583,037	7,951,888	7,283,897,190	2,799,344,224	100.00	100.00	100.00	100.00	0.00

B = Biological, C = Climatological, G = Geological, H = Hydrological, M = Meteorological.

*$1K

As for disaster types in each top country, China had the flood was the most severe hazards, leading in disaster occurrence (31.91%), fatality (51.88%), injured (49.18%), people affected (63.08%), and damage (47.13%). Following flood is earthquake, which accounted 18.95% of occurrences, 39.69% of injuries, and 23.91% of damages. Storm and drought were also deadly and disastrous, with storms causing 27.54% fatality and drought leading 17.44% economic losses. Epidemic was catastrophic as well, accounting for 12.27% deaths. In the United States, storms led the way in all categories of disaster impact, with 62.53% of the occurrences causing 76.03%, 69.65%, 50.84%, and 49.00% in damage, fatality, affected, and injured, respectively. Following storm type were flood, earthquake, and wildfire in disaster impacts while other types were minor.

In Russia, flood (e.g. 48.56% affected, 47.77% injured, and 36.71% damage) and extreme temperature (e.g. 93.98% fatality and 18.94% injured) were most severe, followed by drought (e.g. 22.49% damage) and wildfire (e.g. 20.55% damage). In Iran, the top natural disaster was earthquake, causing 94.34% fatality, 99.52% injured, and 50.47% damage. Drought (e.g. 84.61% affected) and flood (e.g. 35.32%) came to second or third. In Ethiopia, Drought took the top spot with 96.48% for fatality, 96.42% for affected, and 77.89% for damage. Following drought were flood, causing 48.13% injured and 16.17% damage and earthquake, causing 47.55% injured and 5.95% damage.

Correlates between disaster impacts and social-economic attributes

The above spatial and temporal views indicate that disaster occurrences and impacts varied by country and year. However, natural hazards are mostly caused by natural forces below, on, and above the earth. Although knowledge about certain patterns of natural hazards has been accumulated, for example, earthquakes along geological faults, full understanding of the exact disaster occurrences and impacts in space and time is few and far in between. Complicating this are features of human settlements and errors of disaster impact recordings. Despite all these, it would be intriguing to see what plausible correlates exist between disaster impacts and social-economic attributes.

Scatter box plots in Figure 4 show overall correlates between country-level population, area, and GDP and disaster occurrence and disaster impact by total and subgroup. These plots include all data values including outliers and influential points, which are included because they are more likely to be the top disaster hotspots and, hence, important to this research. The boxes labelled with R² > 0.500 show the correlation R > 0.70. In general, population, area, and GDP show positive correlations with occurrence and the disaster impacts except for a few cases, as the lines indicate.

Figure 4. Correlates with outliers and influential points.

Specifically, population correlates well with fatality or people affected with R² > 0.500 (or R > 0.70) for 4 of the 5 boxes. The two highest are R² = 0.904 (or R = 0.951) for peopled affected and R² = 0.876 (R = 0.936) for fatality in total for all disaster subgroups. Population is also good in correlation with occurrence (R² = 0.569 for total, R² = 0.824 for Hydrological), injured (R² = 0.616 or R = 0.785 for Geographical, R² = 0.735 or R = 0.857 or for Hydrological), and damage (R² = 0.835 or R = 0.914 for Hydrological). GDP correlates well with damage (R² = 0.897 or R = 0.947 for total, R² = 0.824 or R = 0.908 for Climatological, R² = 0.500 or R = 0.707 for Hydrological, and R² = 0.733 or R = 0.856 for Meteorological). Area has no single R² > 0.500 or R = 0.707, but does have some weak to moderate correlations, especially with occurrence, fatality, and damage.

Figure 5 shows sample log-linear and non-linear correlations with outliers and influential points removed for the population vs. fatality in linear (R² = 0.532 or R = 0.729), GDP vs. damage in polynomial (R² = 0.641 or R = 0.801), and area vs. occurrence in exponential (R² = 0.382 or R = 0.618).

Figure 5. Sample correlates without outliers and influential points.

Conclusions and discussions

Natural hazards (such as flood, earthquake, and hurricane) happen in mother-nature. However, natural disasters causing deaths, injuries, and property losses occur through interactions between natural and man-made environment. No country can be completely immune from natural hazards, but some countries suffered more from natural disasters than others. The spatial–temporal views in this research show that many natural disasters occurred during 1900–2015 and generated severe impacts to countries all over the world, especially large, populated, developed, or rapidly developing nations. However, in terms of sustainability and residency, both developed and developing countries have a long way to go.

The United States ranked highest in natural disaster occurrence, followed by China, India, Philippines, Indonesia, Bangladesh, and Japan which are all located in Asia. The nations with the most fatalities were China, India, Russia, Bangladesh, Haiti, and Ethiopia while the highest numbers of injuries were recorded in Peru, China, Bangladesh, Haiti, Indonesia, Japan, and India. Also, the United States was ranked first in terms of damage of properties, crops, and live stocks, followed by Japan, China, India, Italy, Germany, and Chile. All of these nations were among the ones with top GDP outputs. These results also show that natural disasters happened in nations across every continent (e.g. Asia, Europe, Africa, and Americas), even though some countries in one continent were riskier than others in another continent.

Correlations between social-economic attributes and disaster occurrences or impacts indicate that only some of them are strong and significant. In general, country population is more correlated with occurrence, fatality, injured, and affected, GDP is more correlated with damage, while country land area has only weak correlations with disaster occurrence and impacts. These correlations are stronger when all natural disaster subgroups are combined, especially with outliers or influential points included. Sample log-linear or non-linear correlations were explored after excluding outliers or influential points, but they may not fit well to the disaster hotspot countries, which tend to have large social-economic values and disaster impacts. Perhaps stratified correlations corresponding to country and impact classifications make more senses.

Along with population increase, economic development, and technology adoption, recent decades have seen an increase of natural disasters and their disastrous impacts across the countries. Thus, this research is important in revealing spatial and temporal risk disparities by natural disasters, especially to the wellbeing of disaster hotspot countries. The spatial–temporal snapshots of global natural disasters taken in this research can further facilitate disaster-related policies, resilient measures, and sustainable development at the national and international level. More specifically, this research can be used as references for country-wide disaster impact assessments or analyses, especially for country governments to make sound emergency preparedness, disaster response and recovery, and hazard mitigation decisions, policies, or plans regarding natural disasters. If country-level views are too aggregated and, hence, cannot reveal intra-country variabilities of disaster impacts, local governments at the state, provincial, or municipal levels may want to use finer spatial units, such as sub-country units (e.g. state, city), to conduct jurisdiction-specific impact assessments for more sustainable and resilient policies, initiatives, or plans to minimize losses from natural disasters. In addition, the results of this research can be useful to the insurance plans or policies against losses from natural disasters. This aspect is essential for both public and private sectors of the insurance business (e.g. Gizzi et al., 2016). Towards these ends and for better quality of live for all countries in the world, more research on global natural disaster impacts and their assessment, prevention, mitigation, and recovery is much warranted.

Disclosure statement

No potential conflict of interest was reported by the authors.