1.4 - Hazards, vulnerability and disaster risk

Hazards become disasters through exposure and vulnerability

A large earthquake is a physical event; a disaster is a serious disruption of a community's functioning when that event interacts with people, assets and their capacity to cope. Event magnitude alone therefore cannot tell us the eventual death toll or economic loss.

A volcano on an uninhabited island can be hazardous without immediately creating a human disaster. A moderate earthquake beneath a crowded area of weak buildings may create a disaster. Exposure is about presence; vulnerability is about susceptibility. A wealthy coastal city can have high exposure because valuable buildings lie in an inundation zone, while an individual low-income household may be especially vulnerable because it lacks safe housing, insurance or transport.

Resilience is the ability to resist, absorb, adapt to and recover from disruption while maintaining or restoring essential functions. It includes strong infrastructure, functioning institutions, mutual aid, financial reserves and access to services. A threshold for resilience is exceeded when event demands overwhelm available coping systems: for example, hospital demand exceeds staffed capacity while damaged roads prevent outside help. It is context dependent rather than a fixed earthquake magnitude that applies to all communities. Resilience can differ between neighbourhoods and can be reduced by a preceding crisis.

These expressions organise reasoning; they are not universally calibrated equations for predicting deaths. Definitions and units must be specified before numerical use, and vulnerability and capacity may overlap. An original illustrative index gives Community A $H=4,V=3,C=2$, producing 6, and B $H=4,V=1,C=4$, producing 1. The comparison shows how lower vulnerability and greater capacity reduce this chosen index for the same hazard. It does not mean A will have six deaths, six times the annual probability of a disaster, or a 600% risk.

Pressure and Release: how vulnerability develops

The Pressure and Release model, often shortened to PAR, explains a disaster as the meeting of a physical hazard and a socially produced progression of vulnerability. The model helps move an explanation beyond blaming victims or treating poverty as an unexplained label.

Root causes operate at national or international scales as well as locally. Dynamic pressures turn them into place-specific unsafe conditions; the stages are linked, not three independent lists. Rapid urban growth is not intrinsically unsafe: it becomes a pressure where safe housing, services and governance fail to keep pace. Wealth alone is not a guarantee of safety either if regulations are poorly enforced or benefits are distributed unequally.

A disaster can intensify the same vulnerabilities. Loss of a wage earner and productive assets can reduce household income; debt can delay rebuilding, leaving people in unsafe shelter before another event. Physical and human conditions also interact: steep land may be occupied because affordable flatter land is unavailable, while shaking triggers landslides on those slopes.

“Release” means reducing the pressures that create vulnerability, such as extending safe housing, enforcing appropriate standards, maintaining drainage and securing equitable access to services. It does not mean stopping plate motion. PAR is a conceptual explanatory model; it does not precisely forecast an event date or death toll, and it needs specific evidence about households, institutions and hazards. It should be combined with exposure and capacity information rather than used as a universal checklist.

Contrasting earthquake and tsunami impacts

Contrasting events show how source processes, exposure and coping systems interact, but country development categories are broad descriptions rather than complete causal explanations. Use the events and reporting dates below, not a claim that every household in a country has the same vulnerability.

Haiti, 12 January 2010, is a developing-country example. The earthquake struck near the densely settled capital region. The 2010 government-led Post-Disaster Needs Assessment estimated damage and losses of about US$7.8 billion, close to the country's entire 2009 GDP, with housing heavily affected. Destruction of homes, hospitals and transport disrupted shelter, care and livelihoods. Lost public infrastructure constrained response, while rubble and damaged water/sanitation created environmental-health problems. A World Bank retrospective, 12 October 2021 gives an estimated 220,000 deaths and 1.5 million displaced; these are reported estimates, not an exact final census.

Chile, 27 February 2010, is an emerging-country case at the event date. The magnitude 8.8 Maule earthquake generated a tsunami that damaged coastal settlements. A USGS response report published in 2011 reported approximately US$30 billion in economic losses, equivalent to 17% of GDP, and 521 confirmed fatalities with 56 missing at that reporting stage. Damage to transport, homes, coastal businesses and services affected employment and recovery. Building standards and earthquake experience helped limit some shaking losses, but tsunami exposure and failures in warning/response created residual risk. Coastal debris and damage to water systems added environmental consequences.

Japan, 11 March 2011, is a developed-country case. The magnitude 9.1 Tōhoku earthquake generated a destructive tsunami. Preparedness and engineered buildings reduced some shaking vulnerability, yet inundation overwhelmed parts of the coastal protection and evacuation system. Deaths, displacement, destroyed settlements and damaged ports affected people and production. The Fukushima nuclear accident added contamination concerns, exclusions and long-term disruption to livelihoods. Japan's Reconstruction Agency records the June 2011 estimate of direct damage as 16.9 trillion yen. This is a dated direct-damage estimate, not a directly comparable measure of Haiti's damage plus economic losses.

The contrast supports two conclusions. First, the larger absolute asset loss in a wealthy economy does not necessarily mean a larger proportional burden or greater household vulnerability. Second, earthquake-resistant buildings cannot remove a tsunami's different physical pathway or every cascading infrastructure failure. Resilience must fit the hazard and the exposed groups, and even substantial national capacity can be overwhelmed locally.

Volcanic impacts from local damage to distant disruption

The impact footprint of a volcanic eruption depends on what it emits, where material travels and which systems are exposed. Economic effects can extend far beyond the country containing the volcano.

In developed Iceland in April–May 2010, the Eyjafjallajökull eruption supplied ash to the atmosphere. Wind carried it across European flight routes; safety restrictions then disrupted travel and air-dependent trade. EUROCONTROL's November 2021 retrospective records about 104,000 cancelled flights in 2010. Stranded passengers illustrate social disruption, while airlines, airports and time-sensitive businesses illustrate economic disruption. The Icelandic Meteorological Office’s 2010 eruption record also records ash production and sediment-bearing glacial floods near the volcano. This case shows high exposure in interconnected systems despite substantial response capacity; a distant traveller is exposed through a transport network rather than through residence on a volcanic slope.

In developing DR Congo on 22 May 2021, Nyiragongo lava reached the Goma area. According to UNICEF's report of 7 June 2021, lava damaged principal water pipes and a reservoir, cutting safe-water access for about 200,000 people. Destroyed homes and displacement created immediate social disruption. Water-system damage increased disease risk and forced spending and time into finding supplies, reducing capacity for ordinary work and schooling. Replacement pipes, water trucking and chlorination were needed. The physical lava footprint therefore generated a larger service and environmental-health footprint; it is incorrect to count only buildings covered by lava.

These cases are not a controlled test of development level: they differ in eruption style, duration, settlement proximity and measured outcome. They do show distinct pathways. Ash interacting with aviation produces a transnational economic cascade, while lava interacting with essential water infrastructure produces acute local vulnerability and dependence on repair. A useful comparison explains those pathways instead of treating a flight cancellation as equivalent to losing safe drinking water.

Comparing impacts without confusing measures

A defensible comparison separates the physical event from the outcome measures and asks whether the figures are genuinely comparable. Similar magnitudes do not control for depth, rupture location, duration, tsunami generation or the population exposed.

Social measures include deaths, injuries, displacement and interrupted schooling or healthcare. Economic measures include direct damage to assets, lost income/output, reconstruction spending and losses relative to GDP. Environmental measures include landslide sediment, polluted water, ash deposition and damaged ecosystems. One measure cannot represent all three, and indirect effects can develop long after the event.

An original illustrative comparison reports direct asset damage of US$2 billion in Economy X, whose annual GDP is US$10 billion, and US$20 billion in Economy Y, whose GDP is US$1,000 billion. X's damage equals 20% of GDP, compared with 2% in Y. Y loses ten times more in absolute money; X faces the larger proportional burden. GDP is an annual flow, while destroyed assets are a stock, so this percentage is a scale comparison, not literally the percentage of that year's output destroyed. It also cannot show the unequal household distribution of loss or the affordability of rebuilding.

Reporting dates matter. An early death toll may exclude missing people; later estimates may revise classifications or boundaries. Monetary values depend on exchange rates, insurance coverage and whether informal assets are recorded. Comparing the Haiti, Chile and Japan figures above requires particular caution because they use different damage/loss definitions and currencies. They establish the importance of proportional burden and cascading effects; they do not isolate national wealth as the only explanation.