Managing and Monitoring Volcanic and Seismic Hazards

Managing and monitoring volcanic and seismic hazards means measuring the warning signs of volcanoes and earthquakes, mapping the areas at risk and preparing people, buildings and emergency services so that fewer lives are lost. In brief: volcanoes can often be forecast because magma gives warning signs (small earthquakes, ground swelling, gas release), while earthquakes cannot yet be predicted but can be forecast as long-term probabilities and detected seconds before strong shaking arrives. Protection relies on monitoring, hazard maps, evacuation plans, strong building design and education.

The aim of hazard management is to reduce risk, not to stop the hazard itself. This guide explains how volcanoes and earthquakes are monitored, what can and cannot be predicted, how early warning works, how buildings and land use reduce damage, includes a worked timing example, summarizes how the field developed and corrects common mistakes.

How are volcanoes monitored?

Before an eruption, magma rising through the crust changes the volcano in ways that instruments can detect. Volcano observatories combine several types of data, because no single signal is reliable on its own.

• Seismicity: seismometers record swarms of small earthquakes and a continuous shaking called volcanic tremor, both caused by moving magma and gas.
• Ground deformation: tiltmeters, GPS stations and satellite radar (InSAR) detect swelling of the volcano as magma fills a chamber, sometimes by only a few centimeters.
• Gas emissions: sensors measure sulfur dioxide and carbon dioxide; a rise in output or a change in the ratio of gases can signal fresh magma.
• Temperature and thermal imaging: infrared cameras and satellites show new hot spots, crater lakes that warm up and changes in lava dome growth.
• Other signs: changes in groundwater, springs and the behavior of animals are sometimes reported, but they are less reliable.

Observatories turn these data into an alert level. Many countries use a color or number scale, for example the USGS levels Normal, Advisory, Watch and Warning, each linked to actions for the public and authorities.

How are earthquakes monitored and forecast?

Earthquakes are recorded by networks of seismometers, which allow scientists to find the focus and epicenter and to calculate the magnitude within minutes. Plate movements are tracked with GPS, which shows where strain is building up along faults.

Scientists distinguish between prediction and forecasting. A prediction would state the exact time, place and size of a future earthquake, and no reliable method exists for this. A forecast gives a probability: for example, that a fault segment has a 30% chance of producing a large earthquake within 30 years. Forecasts are based on the earthquake history of a fault, the time since the last rupture and the rate of strain build-up. Because they are probabilities, they guide building codes and planning but cannot tell a city when to evacuate.

How does earthquake early warning work?

An earthquake produces several kinds of seismic waves. The fast P-waves (roughly 6 km/s in the crust) cause little damage, while the slower S-waves (roughly 3.5 km/s) and surface waves cause most of the shaking. An early warning system detects the P-waves with sensors near the epicenter, estimates the magnitude and sends alerts by radio, mobile phone and sirens at the speed of electromagnetic signals, which is far faster than the seismic waves.

The warning time depends on the distance from the epicenter: it can be tens of seconds for a city far from the fault and zero for places directly above it (the “blind zone”). Even a few seconds are enough to stop trains, open fire station doors, shut off gas and let people take cover. Such systems operate in Japan, Mexico and parts of the western United States, among others.

How is the risk reduced?

Monitoring only helps if people can act on it. Risk management therefore combines several measures.

• Hazard maps and zoning: maps show lava, lahar and pyroclastic flow paths, fault lines and areas of liquefaction, so that homes, schools and hospitals are not built in the most dangerous places.
• Earthquake-resistant design: flexible frames, cross-bracing, deep foundations and base isolators (bearings that let the building move separately from the ground) reduce collapse. Most earthquake deaths are caused by collapsing buildings.
• Evacuation plans: exclusion zones around volcanoes, escape routes, shelters and practiced evacuation, so that people leave in time.
• Education and drills: the advice to “drop, cover and hold on” during shaking and to move uphill and inland after a long earthquake near the coast (tsunami risk).
• Emergency supplies and insurance: stored food and water, and financial protection that speeds up recovery.

Wealthier countries usually have fewer deaths from the same magnitude because of stronger buildings and better organization, while poorer regions face higher losses and slower recovery.

Monitoring methods compared

Worked example: warning time and magnitude

1. Arrival times: a city lies 120 km from the epicenter. P-waves at 6 km/s arrive after 120 ÷ 6 = 20 s. S-waves at 3.5 km/s arrive after 120 ÷ 3.5 = about 34.3 s.
2. Warning time: a sensor 10 km from the epicenter detects the P-wave after 10 ÷ 6 = about 1.7 s. If the system needs 4 s to process the data, the alert is sent at about 5.7 s. Assuming the alert reaches the city almost instantly, the warning before the S-waves is about 34.3 − 5.7 = 28.6 s.
3. A city 20 km away: S-waves arrive after 20 ÷ 3.5 = about 5.7 s, so the alert at 5.7 s gives almost no warning. This is the blind zone.
4. Magnitude comparison: the magnitude scale is logarithmic. Each step of 1 corresponds to a tenfold increase in ground motion amplitude and about 31.6 times more energy. A magnitude 7 earthquake therefore has 10² = 100 times the amplitude of a magnitude 5 earthquake and about 31.6² = 1,000 times the energy.
5. Forecast probability: if a fault has a 2% chance of a large earthquake in any given year, the chance of no earthquake over 30 years is 0.98³⁰ = about 0.545, so the chance of at least one is about 1 − 0.545 = 45%.

How has hazard management developed?

Instrumental seismology began in the late 19th century, and the worldwide network of seismometers expanded in the 20th century. Volcano observatories were set up at Vesuvius in 1841 and in Hawaii in 1912. In 1975, authorities in Haicheng, China, evacuated the city after foreshocks and other signs shortly before a large earthquake, but the Tangshan earthquake of 1976 struck without warning and killed a very large number of people, showing that precursors are not reliable.

The eruption of Nevado del Ruiz in Colombia in 1985 produced lahars that buried the town of Armero and killed more than 20,000 people, even though hazard maps existed; the warnings were not turned into timely evacuation. In contrast, the eruption of Pinatubo in the Philippines in 1991 was forecast, and the evacuation of tens of thousands of people saved many lives. Japan introduced a nationwide early warning system and it was widely credited with helping during the 2011 Tōhoku earthquake.

Which mistakes are common?

Mistake 1: Believing that earthquakes can be predicted

Scientists can calculate long-term probabilities, but not the exact day. Claims of precise predictions have not stood up to testing.

Mistake 2: Thinking that early warning gives minutes

Warning times are usually a few seconds to tens of seconds, and none for places close to the epicenter.

Mistake 3: Assuming that volcanoes give no warning

Most eruptions are preceded by detectable changes in seismicity, ground deformation and gas. The difficulty is knowing exactly when, and how large, an eruption will be.

Mistake 4: Relying on monitoring alone

Good data saves lives only when combined with hazard maps, trusted communication and evacuation plans, as the case of Armero showed.