Numerous synonyms exist for ‘disaster’, that is, when there are casualties, damage, and disruption: calamity, catastrophe, crisis, debacle, emergency, and tragedy, among others. Some have been defined and differentiated, rarely consistently, in academic books, glossaries, dictionaries, and laws. Even then, we struggle to translate and interpret the vocabulary and concepts across languages and cultures.

A ‘disaster’ is perhaps best described as a situation requiring outside help. Something happens that cannot be dealt with, so assistance is needed. It applies to the individual and to the family – think house fires or vehicle crashes. It applies to countries and to continents, such as the North American power outages of 1965 and 2003; and it happens on a worldwide level, with pandemics such as 1918-1920 flu, and the ongoing COVID-19 outbreak. It could happen again, through space weather knocking out power grids and communication systems, for example, or via a major meteorite strike.

While ‘a situation requiring outside help’ helps to define and describe a ‘disaster,’ it does not explain why disasters arise. The fundamental causes of disasters, important for exploring disaster trends, can be distant from the observed impacts.

Some disasters involve environmental phenomena, with prominent examples being volcanic eruptions, rainfall, heat and humidity, landslides, blizzards, and tsunamis. The scale of a disaster is not necessarily correlated with an environmental phenomenon’s size or impact.

Some people are easily able to evacuate from a hurricane or wildfire, for example, and then rebuild their home with insurance money. Others cannot afford to buy insurance or to evacuate; they might lose their job if they leave for a few days; or perhaps they do not have the physical or mental health to start again after losing everything. When a virus sweeps the world, some people are comfortable working from home and stockpiling food from occasional supermarket visits. Others must continue to commute, lack the cash to purchase large quantities of supplies, or experience violence at home. Some people are happy heading to a communal evacuation shelter for a few days after a tornado or chemical explosion, while others legitimately fear discrimination or assault.

These differences in people and places, not the environmental phenomenon itself, explain disaster causes. In December 2003, similar earthquakes rattled central California and southern Iran killing, respectively, two people and around 25,000 people. The disparity did not come from the tremors, but from the number of people and types of infrastructure shaken. Similarly, Bangladesh has long been a landfall zone for cyclones blowing off the Bay of Bengal. Perhaps half a million people died in a 1970 cyclone disaster, while storm disasters in 2020 and 2021 resulted in just dozens of fatalities. In the interim, extensive effort to prepare for cyclones was made, resulting in vastly improved knowledge, awareness, preparedness, evacuation, sheltering, and recovery.

Thus, describing a disaster must focus on society, not nature. The phrase ‘natural disaster’ is best avoided on the premise that disasters are not natural. They are caused by society’s actions and choices, typically giving some people opportunities and resources to avoid problems, while forcing many others to be vulnerable through poverty, marginalisation, and violence.

In other words, as is often said, earthquakes don’t kill people, collapsing infrastructure does. Infrastructure collapses because codes and regulations do not exist or are not enforced, while the people most affected retain few options to examine their seismic vulnerability or to improve their situation.

Hybrid phrases are similarly misnomers. The idea of a ‘natech’ (natural-technological or ‘natural hazards triggering technological’) disaster suggests that technological or industrial failures follow an environmental phenomenon in a cascade, or as an extra impact, perhaps releasing chemicals, severing transportation links, or producing explosions. Yet buildings and vehicles are technology. A school flattened in a tornado, or an aeroplane crashing after a lightning strike, for example, happened not because of the tornado or lightning, but because of society’s choices about technology, notably the strength of the school, or the sturdiness of the aircraft. All disasters are societal, including technology, industry, and economies, so neither natural disasters nor ‘natech’ disasters exist.

The seeds of destruction

How does this definitional discussion help track disaster trends? Although nature does not cause disasters, tackling misapprehensions about increasing environmental dangers is part of examining the environmental phenomena – in tandem with society’s actions – that most influence disaster numbers.

The Earth is dynamic. From hourly temperature variations to millennial climate change as the planet varies in its orbit around the Sun, to tectonic plates drifting through the aeons, nothing is constant. Space weather leading to geomagnetic storms displays a decadal variance, as do over a dozen regional climate indices. The numbers, intensities, and durations of specific environmental phenomena are always changing for different reasons in different places over different lengths of times.

At the same time, human activity modifies them. A road improperly cut through a mountainside can increase landslides or rockslides. Paving over green areas augments runoff, worsening floods. We then separate floodwater from our properties with walls or levees that confine the flow and increase its speed — not helped by dredging to improve waterway navigability. As our buildings soar upwards, so do wind speeds through city wind tunnels and air temperatures through urban heat islands. We manage forests and grasslands, suppressing small fires, which in turn changes the ecosystem and prefaces huge conflagrations often started by arson, discarded cigarettes, or sparking electricity lines.

Our changes are global as well as local. Releasing greenhouse gases into the atmosphere and wrecking ecosystems which absorb them has noticeably increased the amount of heat retained by the planet, changing the climate. This human-caused climate change is affecting the weather now, with one of the most dangerous impacts being increased frequency, intensity, and duration of heat waves. They are moving into realms that make human survival difficult, especially when it does not cool down at night, portending major heat-related disasters and disruption to food supplies, if, for example, agricultural labourers can no longer work. Simultaneously, the oceans are warming, acidifying, and rising with major implications for marine food supplies and coastal freshwater sources.

Human-caused climate change affects the environment in other ways. Many storms are expected to increase in intensity while decreasing in numbers, from North Atlantic polar lows to hurricanes, cyclones, and typhoons. Floods, avalanches, landslides, wildfires/bushfires, and many more are being altered, but specific changes typically depend as much on local actions as on the global climate. For instance, far too many droughts are caused by water (over)use and (mis)management, irrespective of precipitation or snowmelt.

Another major factor confusing an analysis of environmental and human changes is observational bias. Complete and comparable records of many tempests, including hurricanes and cyclones, do not extend before the 1970s, requiring assumptions and proxies to calculate trends. As the world’s population has increased, with a concomitant expansion of available media and hunger for information, more reporting of everything appears alongside more misinformation. The advent of the internet and data crowd-sourcing have vastly changed the amount, form, and quality of material. Even the presence of other news affects how we view the environment. Reported volcanic eruptions declined during World War II, not because of environmental changes, but because the media was preoccupied with stories deemed to be more important – although 1944 was an exception, since Mount Vesuvius erupted in Italy, a major combat area.

Determining increases or decreases of such environmental numbers, and reasons for the changes, is complicated. It can be achieved, but is fraught with difficulties and debates. We must remain grounded in the query ‘are disasters increasing?’ which means focusing on human factors – the key to disasters.

Population numbers, infrastructure, owned assets, and the world’s connectivity have long been increasing. More people exist to be harmed, more infrastructure and assets exist to be damaged, and more expectations of movement of people, goods, services, and information, exist to be disrupted. More potential for disasters is evident, as is the potential for large disasters. At the time of writing, a million-death earthquake has not yet been recorded in human history, but possibilities are frighteningly apparent from Manila to Mexico City. Enormous meteorite strikes, tsunamis, and volcanic eruptions have previously blasted swathes of land that now host large populations.

Meanwhile, our creativity produces more vulnerabilities. Building land into water without proper seismic resistance might create problems if the land liquifies during a tremor and buildings topple, as happened during the 1999 Turkey earthquake. This remains a major concern worldwide, from Wellington in New Zealand to Vancouver in Canada. River levees and sea dikes are constructed to stop regular flooding, but often, subsequent development takes over the floodplain on the assumption it will always remain dry. This assumption is poor, since no shoreline or bank is static. Water eventually reclaims what it covered before, shocking those who thought they were ‘protected’ – as residents in the USA’s Louisiana can attest to. All these activities create disasters.

On the other hand, we reduce disasters by recognising vulnerabilities and attempting to redress them. Hygiene and drinking water treatment have stopped cholera outbreaks in London since 1866. In the wake of Hurricane Hazel’s devastation across Toronto in 1954, planning has so far averted similar disasters in the region, despite further hurricanes. Irrespective of poignant criticisms of the international humanitarian system, aid organisations save tens of thousands of lives, helping people who would otherwise be destitute. For humanity as a whole, life expectancy and many health and social indicators continue to improve while inequities and inequalities continue to widen.

This concept is key. We have the resources and knowledge to reduce disaster potential, and we have had some admirable successes. But as a whole, humanity has squandered the chance to do so much more for so many more, especially for the worst off. Rather than doing what we know we can and should do to avoid disasters, we choose instead to increase disaster potential.

What does the data tell us?

Based on the diverse definitions of ‘disaster’ and its synonyms, databases try to compile statistics. One of the most widely used is EM-DAT (Emergency Events Database) which includes disasters from 1900 until the present, conforming to at least one of the following criteria:

· 10 or more people dead;

· 100 or more people affected;

· The declaration of a state of emergency;

· A call for international assistance

‘People affected’ is defined as ‘requiring immediate assistance during a period of emergency.’ Aside from the technicalities of excluding disasters involving intangible heritage losses or people requiring assistance after the immediate period of emergency, not to mention the database using the term ‘natural disaster’, more fundamental problems prevail.

First, the ability and legal scope to declare a formal emergency and to call for international assistance have improved substantially since 1900. Second, neither is always based on actual disaster impacts. Formally calling for help and making declarations are mainly political decisions. Leaders can show reluctance, despite palpable need, because they might lose face and look incompetent. Conversely, they may call for help in order to seem decisive or to take credit for incoming aid.

Third, the numbers ‘10’ and ‘100’ are arbitrary. Eleven and ninety-nine respectively could serve just well. More insidious is how these thresholds deny some disaster experiences. A village of ninety people lacking the power to declare a state of emergency could lose nine people in a lightning strike and manage with nearby aid. It would not count as a disaster, despite the obvious calamity. Or the ‘nearby aid’ might be a neighbour over an international border, suddenly satisfying the criterion ‘a call for international assistance.’ Theoretical and practical analyses show how the overall cumulative effect of ‘small,’ more frequent disasters can outweigh the impacts of ‘big’ ones matching EM-DAT’s criteria.

Another disaster database, DesInventar, was set up in 1994 to improve understanding of local impacts and disaggregated data. Due to its short operational time period, any computations from this database do not soundly reflect disaster patterns.

Irrespective of the importance of local and indirect experiences, large-scale disasters cannot be sidelined. Neither Hurricane Katrina in 2005 nor 9/11 in 2001 should ever be dismissed as mere data ‘blips’ in US deaths related to hurricanes and terrorism. Despite Toronto’s and Bangladesh’s successes regarding hurricanes and cyclones respectively, they are far from immune to further catastrophe. A probably underestimated 2.06 million deaths are ‘attributed to weather, climate and water hazards’ from 1970-2019, yet COVID-19 killed more than double that figure in eighteen months. Plenty of scenarios, perhaps a Tehran or Tokyo earthquake, or a meteorite strike, could overturn disaster data trajectories to date.

These points show how important the question ‘what is a disaster?’ is for ‘are disasters increasing?’ Even deciding which deaths and injuries to list is perniciously subjective.

Should traffic fatalities occurring during evacuation be tallied, for example? Equally, post-disaster suicides are witnessed long after a disaster’s immediate aftermath, while suicides that may have been prevented when disaster-affected people are assisted are hard to identify. When an earthquake leads to a terrible tsunami, as with Papua New Guinea in 1998 and Japan in 2011, is it one disaster per location or should earthquake-related deaths be differentiated from tsunami-related deaths? The toll of 250,000 killed in the 2004 Indian Ocean earthquake and tsunami is annually exceeded by the number of children under five said to be killed by malaria worldwide. People perishing in Texas’ cold snap of February 2021—including in easy-to-avoid traffic crashes—are disaster deaths, so why aren’t those succumbing to long-term illness because they cannot afford to heat their homes every winter?

Disaster-related data are needed, so collection and analysis ought to continue. The data’s current status presents challenges in arguing a defensible, accurate, precise, or robust picture of disaster trends.

No country for black swans

To reiterate, the answer to ‘are disasters increasing?’ depends. We can, at least, debunk myths such as: climate change increases disasters; disasters are becoming more frequent; and today’s disasters are worse than the past. We can also reject so-called ‘black swans’: allegedly high-impact, low-predictability events which people claim afterwards were readily foreseeable. They are readily foreseeable because, even while nature likely has plenty of surprises in store for us, we know that disasters result from vulnerabilities, and we know how and why vulnerability arises and is maintained.

Although narratives of increasing disasters are, for now, neither convincingly provable nor disprovable, we should never be complacent. Some specifics are clear, such as the strong likelihood of lethal heat-humidity combinations from human-caused climate change. In the meantime, high vulnerabilities persist for those with the fewest options and resources, ensuring that the potential for disasters continues to increase.