The prediction of volcanic eruptions is difficult because, to be of practical use, they must be made before eruptions! Its a lot easier to see patterns in monitoring data after an eruption has occurred. But great progress has been made because of the lessons learned over many years at Kilauea volcano in Hawaii, and applied and modified at Mt. St. Helens before and during in eruption sequence in the 1980s.
Meaningful prediction requires careful monitoring of a volcano’s vital signs. Seismometers can be used to pinpoint earthquakes which track the rise of magma and its movement along fissures. Measurements of the tilt of the entire mountain provide additional information about the “breathing” of the volcano as magma moves inside it. Instruments that sniff SO2, CO2 and other gases also can signal changes in the volcano. At some volcanoes the seismic information seems most reliable, at others the tilt tells the story. But the best predictions come from the combination of all of these methods into a volcano monitoring and prediction system.
You must remember that each volcano is unique. The pattern of events that signifies an eruption at one volcano may not occur before an eruption at a different volcano. And the same volcano may change its eruptive behavior at any time! The good news is that general trends in precursor behavior our being observed at a variety of volcanoes around the world so that volcanologists are getting better at predicting eruptions.
Volcanologists use many different kinds of tools including instruments that detect and record earthquakes (seismometers and seimographs), instruments that measure ground deformation (EDM, Leveling, GPS, tilt), instruments that detect and measure volcanic gases (COSPEC), instruments that determine how much lava is moving underground (VLF, EM-31), video and still cameras, infrared cameras, satellite imagers, webcams, etc!
VW has a great lesson on Working on Volcanoes that goes into more detail.
In the past several decades, our short-term forecasting of large volcanic eruptions has improved by leaps and bounds. The success of scientists at Pinatubo was no fluke! But in reference to explosive volcanic behavior, short-term usually means hours to days, and rarely a week or more.
Volcanologists and seismologists who monitor active volcanoes have integrated several methods to track the state of an active volcano. The key ingredients in this integrated approach are: seismic monitoring, gas monitoring, and deformation studies.
Seismic monitoring consists of deploying a network of portable seismometers around the volcano. The seismometers are capable of detecting rock movement in the Earth’s crust. Some rock movements may be associated with the rise of magma beneath an awakening volcano. Volcanologists that monitor gases often use a correlation spectrometer (COSPEC) that measures sulfur dioxide (SO2) in plumes rising out of volcanic craters. An increase in SO2 may indicate an increase in magma near the Earth’s surface. Deformation studies monitor minute changes — on the order of several mm — in the shape of a volcano. When used together, these methods are capable of providing precise short-term information on volcanic behavior.
The science of monitoring volcanoes is still growing. Some volcanologists are using satellites to monitor active volcanoes. For example, the path of the volcanic plume erupted at Pinatubo was tracked using AVHRR weather satellites. Dr. Bill Rose and his students at Michigan Technological University are collecting and analyzing images of volcanic plumes from different eruptions to understand the evolution of plume in its journey through the atmosphere. Dr. Peter Francis (Open University) uses images from the LANDSAT multispectral satellite(s) to study the size and distribution of deposits from caldera-forming eruptions; this information provides important clues on the nature of future eruptions.
Other scientists are looking at the gases emitted from volcanoes. Carbon dioxide (CO2) is one of the gases to first leaves a magma as it creeps up into the Earth’s upper crust. If we could monitor CO2 accurately we might have a valuable tool for monitoring the state of unrest at volcanoes. The abundance of CO2 in the atmosphere has defeated us in the past, but volcanologist, Dr. Stanley Williams (Arizona State University), is working with the engineers who designed the COSPEC to design a similar tool that will detect CO2. In another few years we may be measuring both SO2 and CO2.
Still other scientists are using Geographical Positioning Systems satellites (GPS) to continuously monitor deformation at volcanoes. Dr. Tim Dixon (University of Miami) has monitored small changes (millimeters to centimeters) in the floor of the Long Valley caldera in California.
In short, the future is pretty good for forecasting volcanic eruptions over the short-term and in most cases (i.e., Rainier) loss of life can probably be minimized. Of course, an essential ingredient to mitigating volcanic hazards is international cooperation and rapid deployment of trained personnel and technology. The U.S. is a leader in international volcanic hazard mitigation, and right now, the Volcanic Action Crisis Team (VCAT), of the U. S. Geological Survey, is in Mexico assisting Mexican volcanologists monitor volcano Popocatepetl outside of Mexico City. Members of the VCAT crew cut their teeth on Mt. St. Helens in 1980, and later worked with the Philippine Institute of Volcanology and Seismology to forecast the powerful eruption at Pinatubo.
A caveat: eruption size is not always the controlling factor in the number of deaths incurred. The 1985 eruption of Nevado del Ruiz, Colombia, was, compared to Pinatubo, a rather small and insignificant eruption. Unfortunately, 25,000 people in the town of Armero were killed when a lahar, produced by melting of a summit glacier, swept through the town.
Here are a few references that may be more helpful:
- Volcanoes and Society by David Chester (Publisher: Edward Arnold, London, 1993) see chapter 7.
- Monitoring Active Volcanoes edited by Bill McGuire, Christopher RJ Kilburn and John Murray (UCL Press, London, 1995).
- UNESCO (1972). The Surveillance and Prediction of Volcanic Activity: A Review of Methods and Techniques. United Nations Educational, Scientific, and Cultural Organization (UNESCO), Paris, 166 pp.
- Tazieff J, Sabroux J-C (1983). Forecasting Volcanic Events, Developments in Volcanology I. Elsevier Publishers, Amsterdam, 635 pp.
- Tilling RI (1989). Volcanic Hazards. Short Course in Geology: Volume I, Presented at the 28th International Geological Congress, Washington, D.C. 123 pp.
Thanks to Chuck Wood, Scott Rowland, Mike Conway (Florida International University), the USGS, and the VW team!