1.5 - Comparing tectonic hazard profiles
Learn to compare the physical characteristics of earthquakes, eruptions and tsunamis, then explain why an event's measured size does not determine its social or economic impact.
Measuring size and effects
The size of a tectonic event and the effects experienced at a particular place are different measurements.
Magnitude and intensity
Magnitude describes an event's size at its source. Intensity describes the strength of its effects at a location, so one event can produce many local intensities.
Earthquake moment magnitude, written , is calculated from seismic moment: the rigidity of the rocks multiplied by the fault area that slipped and the average displacement. It describes the overall rupture, rather than counting collapsed buildings. A larger magnitude normally means greater potential for extensive shaking, but depth, distance, rupture direction and local geology affect the shaking at each settlement.
The Modified Mercalli intensity scale uses observed shaking and effects, from I (not felt) to XII (extreme). A report from a town on soft sediments may indicate greater intensity than one from a town on firm rock at a similar distance. Building vulnerability also affects observed damage. Mercalli values are ordered categories: XII is not mathematically twice VI. A magnitude has no unit such as metres or joules; it is an index.
An approximate comparison of radiated seismic energy is
Here is the difference in magnitudes and the ratio is dimensionless. A one-unit increase gives times the energy. This empirical comparison does not predict deaths or local intensity.
For example, a magnitude 7.0 event releases approximately times the energy of a magnitude 6.0 event. It does not necessarily cause 31.6 times the damage.
The Volcanic Explosivity Index (VEI) compares explosive eruptions using erupted tephra volume, plume height and observations. Its familiar categories run from 0 to 8; from VEI 2 upwards, successive volume bands generally increase tenfold. The smallest bands differ, and observations also matter. VEI measures explosivity: a low-VEI effusive eruption can still bury buildings in lava. A tsunami is often described using wave height, inundation distance or run-up height at a specified coast; the magnitude of its generating earthquake is not a local tsunami-height measurement.
Sources: USGS magnitude and intensity, energy comparison and VEI explanation, 2022.
Comparing hazard characteristics
A hazard profile compares several characteristics using stated categories, units and an identified event or setting.
| Characteristic | Earthquake | Volcanic eruption | Tsunami |
|---|---|---|---|
| Magnitude | describes rupture size; intensity varies locally | VEI describes explosivity; lava volume or discharge may also matter | Coastal height/run-up varies with seabed and coastal form |
| Speed of onset | Strong shaking can start with little usable warning | Unrest may develop before eruption; explosive phases can begin rapidly | Near-source waves may arrive within minutes; distant coasts have longer lead times |
| Areal extent | Strong shaking zone plus secondary hazards | Lava/pyroclastic flows relatively local; ash can travel far downwind | Many coasts across an ocean may be exposed, unevenly |
| Duration | Main shaking usually seconds to minutes; aftershocks can last much longer | Individual phases may last hours; activity can persist for months or years | A series of waves and dangerous currents may persist for hours |
| Frequency | Small events generally more frequent than very large events | Depends on the volcano and eruption size | Depends on generating events and the coast studied |
| Spatial predictability | Broad risk zones known; exact next rupture location uncertain | Known active centres help identify zones; new vents and flow paths remain uncertain | Travel paths can be modelled after generation; local inundation requires detailed modelling |
These are characteristic ranges, not fixed scores for every event. Separate onset from duration, and spatial predictability from predicting the date. A known plate boundary gives a useful risk zone without an exact earthquake prediction. A long volcanic eruption may give time to evacuate one area while a fast pyroclastic flow allows little escape once it starts.
Magnitude and frequency also depend on the observation window. Suppose an original illustrative 100-year catalogue contains 900 events of magnitude 4–4.9 and nine of magnitude 7–7.9. Their observed average rates are nine and 0.09 events per year. The inverse rate for the larger class is about 11.1 years per event across that whole region. It is neither a countdown nor the recurrence interval of one particular fault. Short catalogues can miss rare extremes; a quiet recent period does not establish safety.
When using a radar-style profile, define which end of each axis means more hazardous. An arbitrary score of 5 for frequency cannot be added meaningfully to 5 for duration. A profile helps identify management needs; it is not automatically a numerical risk calculation.
A fast onset reduces reaction time; a wide extent increases coordination demands; a long duration can exhaust shelters and livelihoods. The combination matters more than an unlabelled total profile score.
Event profiles and impacts
Contrasting events show how a physical profile interacts with exposed people, assets and their capacity to cope.
The following historical examples represent different development contexts; the labels describe broad circumstances at the time, rather than a permanent ranking of every community.
| Event | Physical profile and location | Social and economic implications |
|---|---|---|
| Haiti, 12 January 2010, earthquake | Rapid onset near the densely settled Port-au-Prince region; severe local shaking | Extensive building failure caused catastrophic mortality and displacement. Damage to institutions, roads and services weakened rescue and recovery in a low-income context. Death estimates differ substantially between sources. |
| Pinatubo, Philippines, June 1991, VEI 6 eruption | Explosive eruption with local pyroclastic flows and widespread ash; warning signs preceded the climax | Evacuation reduced exposure to the flows, but wet ash collapsed roofs and rain generated lahars. Disrupted homes, farming and infrastructure extended loss beyond the main eruption in an emerging-economy context. |
| North-east Japan, 11 March 2011, earthquake and tsunami | Large offshore rupture; a rapidly arriving tsunami flooded exposed coastal settlements | Strong earthquake preparation did not prevent devastating coastal mortality and infrastructure loss. Flooding at Fukushima Daiichi created a technological emergency, displacement and prolonged costs in a high-income country. |
The Haiti interpretation is supported by USGS reconnaissance, 2010; the eruption profile by USGS Pinatubo evidence; the Japanese cascade by the IAEA accident report, 2015. The Japan magnitude follows the revised USGS event record, rather than early 9.0 estimates.
Social severity includes deaths, injury, displacement, interrupted schooling and health effects. Economic severity includes destroyed assets and interrupted production. A richer country may record a larger monetary bill because expensive assets are exposed, yet have more resources for rebuilding. A smaller absolute loss may be a much greater share of a poorer country's economy. Damage, indirect loss and reconstruction needs are related but different quantities; do not combine them without checking definitions.
The examples do not isolate development as the cause. They differ in hazard type, location, secondary hazards and exposed population. Comparing their raw death totals alone would confuse these influences. The useful claim is that physical magnitude sets potential severity while exposure, vulnerability and management influence how that potential becomes loss.
Testing magnitude–impact links
A correlation investigates whether two variables vary together; it cannot by itself establish why they do so.
To study magnitude and deaths, begin with comparable earthquake records and common rules for counting deaths. Plot magnitude against deaths, inspect outliers, and consider a logarithmic deaths axis if values span orders of magnitude. Repeat with economic damage using a common currency, price year and damage definition. A high-magnitude sparsely populated event and a lower-magnitude urban event can break a simple relationship. Tsunami fatalities must be consistently included or excluded.
Spearman's rank correlation tests the direction and strength of a monotonic association. Rank both variables in the same direction, calculate the rank difference for each paired event, square the differences, then add them. For untied ranks:
is the number of paired observations. Values near +1 indicate a strong positive rank association; values near −1 a strong negative association; values near 0 little monotonic association. Tied ranks require the taught tie-aware method rather than uncritical use of this shortcut.
This original illustrative dataset isolates the method; these are not real disaster records.
| Event | Magnitude | Deaths | Magnitude rank | Deaths rank | |
|---|---|---|---|---|---|
| A | 6.0 | 20 | 1 | 1 | 0 |
| B | 6.2 | 90 | 2 | 3 | 1 |
| C | 6.4 | 40 | 3 | 2 | 1 |
| D | 6.6 | 500 | 4 | 6 | 4 |
| E | 6.8 | 150 | 5 | 4 | 1 |
| F | 7.0 | 300 | 6 | 5 | 1 |
Here , so to three decimal places. Larger magnitudes tend to have more deaths in this small sample, but D and F show that the order is imperfect. A coefficient is not a significance test: deciding statistical significance also needs the correct critical value or probability procedure, sample size and stated significance level. Six selected illustrative events cannot prove a global relationship.
Even a significant real correlation would not prove magnitude alone caused the deaths. Compare ground motion, depth, exposed population and construction, and check whether missing low-loss events biased the sample. A sound interpretation combines the statistical pattern with a geographical explanation.