Hazard event prediction

Why is it important to predict hazards?
  • Reduce the risk of death and injury.
  • Reduce damage to property and land
  • To protect agriculture e.g. move cattle to higher land during a flood or cover vulnerable crops during frosts
  • To give warning to citizens in order to give them time to evacuate and/or prepare
  • To prepare emergency services e.g. cancel non-emergency medical procedures, call up rescue teams, etc.
  • Cancel all non-essential travel e.g. flights and shipping to reduce the chance of accidents
  • Seek international assistance
  • Give warnings to vulnerable industries e.g. recall fishing boats during hurricanes or close mines during flooding.

When making predictions, it is necessary to look at the probability of an event happening, the probable magnitude of the event, the location of the hazard and its likely impact. Because all hazards contain so many variables, it is impossible to make 100% accurate predictions of hazard outcomes, but scientists are getting better.


Of all the world's major natural hazards, earthquakes are probably the hardest to predict. Unlike most hazards there is no build up of magnitude e.g. droughts and floods (they tend to get progressively worse, whereas the strongest part of the earthquake is normally the first quake), there are no warning signs e.g. volcanoes (changes in shape, heat, etc.) and the hazard does not move towards land e.g. hurricanes and even tsunamis (so you can prepare for its arrival).

Even though it is still impossible to predict the exact location, strength and time of an earthquake, seismologists have made some advances in predicting earthquakes. Below are some of the advances seismologists have used and some of the methods that they use to try and aid prediction.

Foreshocks: An increase in the frequency of small earthquakes (foreshocks) have been used to predict large earthquakes. Probably the most famous example was the prediction of the 1975 Haicheng earthquake which measured 7.2 on the Richter scale. However, only about 5% of foreshocks lead to bigger earthquakes so there can be a lot of false alarms. (USGS - Haicheng earthquake prediction)

Seismic History: Seismologists can study the seismic history of earthquakes and try and make predictions of when future earthquakes are likely to happen. For example El Salvador has a major earthquake roughly once every ten to twenty years. However, at best this can only give a rough time frame and can certainly not pinpoint the time or location of an earthquake.

Animal behaviour: Some scientists believe that small animals e.g. cats, toads and dogs are able to detect pre-seismic activity and alert people to an imminent earthquake. Some scientists believe that it is low frequency electromagnetic signals that they are responding to. It is believed that toads en mass hopping across the road in Taizhou, China two days before a major earthquake that killed 10,000 people was actually a warning sign that local authorities should have acted upon.

How Animals Predict Earthquakes - BBC article

Did toads predict the earthquake - France 24 article

Plate boundaries: Most earthquakes are found along plate boundaries so scientists can alert countries and populations to the risk of earthquakes. However, even knowing the potential location certainly does not help predicting the time of a quake. Also some earthquakes happen along old and unknown plate boundaries and faults or actually happen with plate boundaries e.g. intraplate earthquakes. Because these earthquakes are almost impossible to predict that they can cause a lot of damage because populations are not prepared.

Virginia earthquake felt in Washington and New York - BBC article

Radon: The release of radon has been studied as a precursor to a major earthquake. However, all studies have proved inconclusive, but what scientist claimed that he did predict the recent L'Aquila earthquake in Italy using this technique.

Did a technician accurately predict the L'Aqulia earthquake - Scientific American Article

Geological Changes: Scientists believe that small-scale uplift, tilt or subsidence of the ground can be precursor to major earthquakes. However, it would be almost impossible to try and monitor all geological changes around the world to try and predict earthquakes.

Rock stress: Scientists also believe that changes in the stress of rocks can also be a sign of imminent earthquakes. Some research done along the San Andres fault suggested changes in rock stress 2 hours before an earthquake. Again though it would be almost impossible to monitor all plate boundaries look for changes in stress.

Pre-quake changes seen in rocks - BBC article

Earthquake prediction using accelerometers in laptops - BBC article

The Parkfield Earthquake Prediction Experiment - Decoded Science Article

Long range earthquake prediction - really? - BBC article

Can we predict when and where quakes will strike - BBC article

Italy scientists on trial over L'Aquila earthquake - BBC article

L'Aquila quake: Italy scientists guilty of manslaughter - BBC article


Compared to earthquakes, volcanoes are much easier to predict, because unlike earthquakes they normal release warning signs before they erupt. Below are some examples of how scientists predict volcanoes and also a case study of how the Japanese volcano Sakurajima is monitored in an attempt to predict future eruptions.

Gas emissions: Scientists often measure the release of gases around volcanoes. An increase in the release of some gases e.g. sulphur dioxide can indicate that an eruption is likely.

Ground deformation: On known active volcanoes, scientists will often study the tilt (slope) of the volcano or the development of any bulges. Vulcanologist knew that the Mount St. Helens volcano was about to erupt because of the development of large bulge on its side.

Thermal tracking: The build up of magma can often cause an increase in the temperature of a volcano. Scientists can monitor temperatures changes through underground probes, infrared or even satellite.

Mass movements: Prior to volcanoes increases in seismic activity, changes in the shape of volcano e.g. slope angle or changes in temperature can trigger a variety of mass movements e.g. rockfall, avalanches and lahars which can be studied by scientists as warning signs of an imminent eruption.

Hydrology: Scientists can monitor changes in water in a number of ways. They can study changes in temperature and chemical composition. They can also look for the presence of volcanic gases. Also scientists study rivers flowing from volcanoes to look for volcanic related sediment, but also increases in snow melt and possible the presence of lahars caused by increased temperatures.

Remote sensing: Remote sensing equipment like satellites, thermal image cameras and gas monitors detect changes in shape, temperature, gas and chemical composition, etc to try and predict likely eruptions.

Seismic History: By studying previous volcanic history, scientists can look for patterns of eruptions and also establish if volcanoes are active, dormant or even extinct. However, volcanoes don't always follow patterns, so apart from giving rough estimates this system is not particularly useful.

Seismic activity: An increase in earthquakes can signify that a volcanic eruption is likely. Therefore scientists carefully monitor seismic activity around known volcanoes.

Expert predicted volcano eruption - BBC article

Iceland volcano's uncertain timescale - BBC article

Mega Volcanoes may be Predicted - BBC article

Big Tokyo Earthquake Likely in the Next 30 Years - BBC article

Supervolcanoes 'can grow in just hundreds of years' - BBC article

Sakurajima Volcano

Sakurajima volcano lies near the city of Kagoshima on the Japanese island of Kyushu. The city of Kagoshima has a population of about 500,000 people so scientists monitor the volcano very careful in an attempt to protect the local population. Two organisations monitor the volcano; the Japanese Meteorological Agency and Kyoto University's Sakurajima Volocanological Observatory. They monitor the volcano in a number of ways including:

Monitoring seabed: The seabed in Kagoshima Bay is monitored to look for changes in the height of the seabed. An increase in the height of the seabed indicates a build up of magma.

Groundwater levels: Scientists look for changes in the temperate of groundwater which maybe caused by volcanic activity as well as changes in the chemical composition and gases released. The rising seabed can also cause tides to rise that can also be monitored.

Seismic activity: Seismometers constantly monitor areas around the volcano look for increases in earthquakes which may signal eruptions.

Volcano shape: Tiltmeters carefully monitor the shape of the volcano to look for changes in its shape. If the volcano grows or begins to bulge it can signal that a volcanic eruption is likely.

Rock structure: An underground tunnel monitors the structure and temperature of rock. If splitting or melting rock is detected it normally signifies an increase in volcanic activity.

Remote sensing technology like satellites and aerial images are also used to monitor changes in the volcano.
external image 250px-Kagoshima_and_Sakurajima.jpg
external image ven3volcano.jpg


Although hurricanes are very hard to predict and track, scientists gaining a better understanding so can never give more accurate warnings. To begin with scientists know the conditions necessary for hurricanes to form, namely warm water. Therefore they know that hurricanes are likely to form in tropical areas towards the end of summer e.g. in the Caribbean between June and November. They also know that hurricanes normally move westwards (because of easterly winds) and slightly towards the poles (see map to the right). Even though scientists can not say exactly when a hurricane will form, because they form at sea they can make attempts to predict the path of hurricanes so that they can predict landfall and therefore warn communities. Hurricane tracks are predicted in a number of ways, by attempting to measure; temperature (air and sea), pressure, wind speed/direction and moisture:

Satellites: Satellites are now much more sophisticated and can measure the size of hurricanes, the direction they are travelling, but also cloud structure and ocean temperatures.

Weather radar: The US has a total of 155 radars constantly scanning the skies over the US and its surrounds. They are capable of recording wind and precipitation data.

Aircraft: The US Air Force uses aircraft to drop sensors into hurricanes to measure wind speed, wind direction, pressure, etc.

Buoys and floats: These can measure water and air temperature, wave height and the direction and speed of wind.

Computer models: All the information that is collected from the four methods above is then fed into computers to try and predict future movements and changes of hurricanes. However, even with improved data, models can often pick very varied courses and changes of hurricanes. To the right is a map showing the different computer predictions for Hurricane Katrina - you can see the predicted tracks are very varied.

Hurricane Katrina

Hurricane Katrina was a large category 5 hurricane that caused devastation when it made landfall on New Orleans on 29th August 2005. The winds of over 175mph killed over 1,800 and caused damaged totalling over $100 billion. The destruction caused by this storm was despite some of the most intensive tracking and predicting ever made about a hurricane. Once Hurricane Katrina had passed Florida, forecasters new that the storm would intensify in the warm waters of the Gulf of Mexico. However, there predictions of where it may make landfall covered much of the US's south coast. Because of this the NHC could only forecast the probability of Hurricane Katrina hitting certain areas. Despite this this time their predictions were reasonable accurate, from 56 hours out, they predicted that Hurricane Katrina would hit somewhere is SE Louisiana. Even with the reasonably accurate predictions many people choose to ignore evacuation warnings and stayed in their homes. Evacuation warning may have been ignored for many reasons including:
  • Can't afford to leave (no transport)
  • Fear of leaving home (possible water/wind damage and robbery)
  • To old, young or ill to leave/travel
  • Don't trust predictions
  • Confidence in defences
  • Warnings too late
  • Lack of knowledge, no communication, no warning
  • Underestimate of risk

NOAA: National Oceanic and Atmospheric Administration (US government agency that monitors conditions in the oceans and atmosphere) - NOAA Website

NHC: National Hurricane Center (part of NOAA) - National Hurricane Center Website
external image Stages_of_Hurric_Dev_fig01.jpg
external image at200512_verify.gif

New Orleans pounded by hurricane - BBC article

I should have evacuated New Orleans earlier - BBC article

New Orleans evacuation underway - BBC article


Tsunamis are a secondary hazard that can be caused by one of the following primary hazards:
  • Earthquake
  • Submarine volcano
  • Mass movement
  • Meteor

Therefore a lot of tsunami predictions can be made by monitoring the primary hazards that cause tsunamis. Astronomers study the sky for potential meteors and can make predictions about their impact, seismologists can record earthquakes and vulcanologist study potential volcanoes. However, monitoring primary hazards is not the only method used of predicting tsunamis. The Pacific Ocean and now the Indian Ocean have very sensitive and sophisticated tsunami warning systems. It must be noted that the method below is not a prediction, but a warning after a tsunami has been triggered.

Tsunami warning systems consist of two important components; one a network of sensors and two a communications infrastructure. A tsunami warning is often triggered by seismic gauges monitoring for earthquakes (most common cause of tsunamis), but then onshore tidal or offshore buoys measure for changes in sea height/tide etc. to confirm the presence of a tsunami. At sea tsunamis maybe not noticeable to humans because it is not until they reach shallower water near the coast so they develop into their devastating size. Once the presence of a tsunami has been confirmed, the second component of the system is essential - communications. All vulnerable populations must be alerted via television, internet, radio, texts, twitter, sirens, etc. Without proper communication to vulnerable populations, knowledge of a tsunami is useless. It is also important that vulnerable communications have plans to how to evacuate and protect populations.

However, it must be remembered that if an earthquake happens very close to land and a tsunami is triggered any warning can be too late to evacuate communities. It was estimated that Bandah Aceh (Northern Sumatra, Indonesia) was hit within 15 minutes of the 26th December 2004 earthquake. Therefore even if there was a warning system in place, evacuation would not have been possible.

external image 473879963_f64813b442.jpg

Pacific Tsunami Warning Center Website

Asia tsunami warning system is ready - BBC article

Action Urged Over Giant Wave Threat - BBC article

Evacuations Based on Predictions

Despite improvements in predictions, evacuation warnings are not always made or made very late. Some of the reasons for the delay in evacuation notices include:

  • Governments don't want to cause panic. A panic may cause more deaths than the hazard.
  • The cost of ordering an evacuation (both in physically moving people and closed businesses)
  • Unpredictable warnings. People may actually be evacuated into a more dangerous area.
  • Late predictions. If a warning comes late it maybe safer for people to stay in their homes, rather than attempting to move away.
  • Some old, young and sick people are unable to travel
  • Some people don't hear evacuation warnings or can't afford to evacuate (no car, no where to go, etc.)
  • Some people don't want to leave their home, business, pets, etc. unattended
  • People believe they are safe in their homes and can defend against the hazard.

Related articles

Amazon Fire Season Linked to Ocean Temperatures - BBC article

Predicting droughts with greater certainty - Science Daily Article

Mexico raises Popocatepetl volcano alert - BBC article