Forecasting eruptions

Forecasting volcanic eruptions is one of the most important goals of volcanology, as people and property are at risk in active areas. Although the phenomena are very different, volcanic forecasting is similar in principle to weather forecasting. It is based on statistics on previous events and information about current activity.

General and Specific Forecasts

All forecasting of volcanic eruptions is probabilistic and based on long-term patterns and present-day observations. Forecasts can be subdivided into two categories—general and specific depending on how constrained the forecast is. General forecasts are based mainly on statistical averages. For example, if it rains an average of 182 days per year in a given city, the chance of rain is 50 percent on any day. A general forecast for any other day would report a 50 percent chance of rain, which is hardly a useful prediction but is statistically valid. Specific forecasts are based on historical statistics and current information. Thus, the 50 percent general forecast could be greatly improved by using satellite photographs and information on the day's movements of atmospheric pressure systems. In that case, rain might be forecast with a 90 percent chance, and the specific forecast would be much more useful.

A similar rationale can be applied to volcanoes. To forecast eruptions, it is necessary to understand the volcano’s past activity. The problem is that volcanoes have widely differing eruptive patterns. Some erupt frequently, and historical records are reliable indicators of eruption patterns. Other volcanoes erupt so infrequently that the historical record is not a statistically valid sampling of their activity. Some potentially dangerous volcanoes have never erupted in recorded history. Although prehistoric eruptive patterns give important long-term information in general forecasting, it is also necessary to have current information about volcanoes that show signs of awakening. Three types of modern observations help forecast volcanic eruptions—geophysical, geochemical, and geological.

Geophysical Observations

Indirect physical changes usually accompany the underground movement of magma, such as changes in the volcano's magnetic, thermal, or gravitational properties, changes in the configuration of the ground surface, seismic activity, and fluctuations in the electrical properties of rocks. These geophysical phenomena are related to the fact that magma is introduced to the volcano or its plumbing system before an eruption.

Magma commonly produces only minor temperature changes in the overlying near-surface rocks and soil because rocks are generally poor heat conductors. Furthermore, any temperature effects are usually masked by heating and cooling during the daily solar cycle. Most rocks, and especially iron-rich volcanic rocks, have a natural magnetization acquired during cooling within the Earth's magnetic field. Just as an iron magnet can be destroyed by heating, few rocks can remain magnetized at temperatures greater than about 550 degrees Celsius. The temperatures of magma range from about 800 to 1,200 degrees Celsius, and molten rock is, therefore, nonmagnetic. Underground rocks are slowly heated by magma, decreasing local magnetic field strength.

As magma approaches within a few kilometers of the Earth's surface, it makes room for itself by shouldering aside the surrounding rocks. As a result, the overlying rocks and soil may bulge upward in a phenomenon known as tumescence. Similarly, deflation of the ground surface commonly occurs when magma withdraws, sometimes signaling the end of an eruption. The shape and extent of the uplifted area are controlled by the configuration of the underground magma body, and the magnitude of tumescence can be substantial. For several months before the catastrophic eruption of May 18, 1980, the northern summit region of Mount St. Helens bulged several meters per day, achieving a total of about 100 meters of northward bulging. Cracks and faults also opened on the summit of St. Helens, some a few meters wide and extending for more than one kilometer. At other volcanoes where the volume of magma is small, or the magma is relatively deep, tumescence may be imperceptible to the eye, amounting to only a few millimeters of uplift over several kilometers.

The occurrence of earthquake swarms also marks the movement of underground magma. Thousands of small earthquakes may occur daily as the magma forces itself upward, fracturing the surrounding rocks at successively shallower depths and sending out seismic waves. Volcanic earthquakes differ in several ways from earthquakes associated with large faults in the Earth's crust. First, volcanic earthquakes are localized beneath active or potentially active volcanoes and occur due to local magma injection rather than large-scale shifting along major faults. Second, volcanic earthquakes are relatively shallow, usually at depths less than five or ten kilometers. Third, volcanic earthquakes have low magnitudes (usually less than 4 on the Richter scale) because the brittle rocks around local magma bodies are easily fractured.

In contrast to a single-shock volcanic earthquake, a volcanic tremor is a continuous vibration that can last for minutes, hours, or days and has been detected only on active volcanoes. Also called harmonic tremors, the vibrations are thought to be produced by the subsurface flow of magma or by the formation of gas bubbles within shallow magma bodies. At basalt volcanoes such as Hawaii, harmonic tremor commonly begins a few minutes before the eruption of fluid lava and continues during the surface emission of lava. At volcanoes with more viscous types of magma, such as Mount St. Helens, harmonic tremor does not always occur, and its origins are poorly understood. A 1996 survey of earthquake swarms between 1979 and 1989 showed that 58 percent were followed by eruptions. Those not followed by eruptions had a shorter average duration than those that were.

Other geophysical studies include monitoring the electrical properties of rocks and soil near volcanic vents and measuring any changes in the strength of the Earth's local gravitational field. Small but measurable changes in gravity may accompany the underground injection of magma if the magma is lighter or heavier than the rocks it displaces or if the ground is uplifted.

Geochemical Observations

Geochemical measurements of gas and lava compositions are also useful during volcano monitoring. The types and volumes of gases emitted from lava, or fumaroles (vents that emit only gases), commonly change when fresh magma is injected underground. For example, at the Kīlauea Volcano in Hawaii, the ratio of carbon dioxide to sulfur dioxide gases is greater in fresh magma that arrives from the Earth's mantle than is the ratio of those gases emitted from older magma that has been stored beneath the volcano for some time. Thus, changes in the gas chemistry and temperature of lava lakes and fumaroles can signal the injection of fresh, gas-rich magma that might be more likely to erupt than would older, gas-poor magma.

Studies of the gas plume from the Mount St. Helens lava dome have shown that increased emissions of sulfur dioxide and carbon dioxide occurred when fresh batches of magma were injected beneath the summit crater. Increased rates of sulfur dioxide emission have been measured before several non-explosive eruptions at Mount St. Helens. The emissions have been interpreted as resulting from rapid degassing of small bodies of magma as they moved toward the surface from a larger, deeper magma chamber. Changes in the percentages of hydrogen, helium, and radon in volcanic gases may also prove useful in signaling eruptions.

Measurements of lava compositions can also be useful for recognizing fresh batches of magma that have moved into the volcano. At Mount St. Helens, each non-explosive eruption since October 1980 has added a new flow of viscous lava to the summit crater. The chemistry and mineralogy of lava samples from Mount St. Helens have remained nearly constant through many successive eruptions. This consistency suggests that the magma feeding the eruptions has come from a single, shallow magma body that was injected in 1980 and has not undergone any significant changes in composition since that time. A different scheme of magma supply seems to operate beneath the Kīlauea volcano in Hawaii, where repeated injections of fresh magma into shallow magma chambers have occurred many times during historical eruptions of that volcano. The arrival of fresh magma causes variations in the temperatures of lava flows and lava lakes and variations in the chemical composition and mineralogy of the lavas.

Geological Observations

Events that are too complex to be numerically measured in any simple way can nevertheless be observed and recorded, such as the opening of new fractures or vents, the advent of new fumaroles, lava flows, or pyroclastic fountains, and the fluctuation or cessation of lava output. Together with geophysical and geochemical data, such geological observations are an important aspect of eruption forecasting.

Geologists also study prehistoric volcanic deposits, which can give important information about long-term eruption patterns on a time scale of thousands or millions of years. Prehistoric volcanic deposits are sometimes difficult to study because of poor outcroppings of the deposits, the destruction of ancient deposits by erosion, and difficulties with accurately dating the ancient materials. Despite these limitations, geological field studies are an important aspect of volcano forecasting because volcanism can be examined in a much longer time frame than is possible with historical records alone. The periods of repose at different volcanoes vary from months to hundreds of thousands of years; some volcanoes erupt so infrequently that the historical record does not accurately represent their long-term behavior. This is true for many Cascade volcanoes, including Mount St. Helens, which was active between 1831 and 1857, then lay dormant for more than a century before explosively awakening in May 1980.

Past Eruptions and Probabilities

The statistics of past eruptions, based on historical records and the study of prehistoric deposits, are very important both in finding the average probability of future eruptions and in recognizing any patterns of eruption characteristics. At least three patterns have been recognized from a few well-studied, active volcanoes. The first pattern is characterized by completely random behavior. No matter how long it has been since the last eruption, the average chance of a future eruption remains the same. This is analogous to cutting a deck of cards to get the ace of spades; no matter how many cuts are made, the chance of getting the ace in the next cut remains 1 in 52. Mauna Loa volcano in Hawaii seems to behave this way, and the average chance for a new eruption is about 2 percent each month, regardless of how long it has been since the previous eruption.

The second pattern is like cutting a deck of cards for the ace of spades and throwing away each unwanted card. The chance of getting the ace increases with each new cut because the deck grows smaller. At Hekla volcano in Iceland, the probability of an eruption increases with time since its last eruption. A third pattern has no analogy in card cutting. At volcanoes such as the Kīlauea in Hawaii, eruptions tend to cluster together in time, and the probability of a new eruption decreases the longer the time has passed since the last eruption. This is the opposite of the second pattern. Recognition of one of these time patterns is important in forecasting future eruptive activity, but such recognition depends on having observed a statistically significant number of eruptions at a given volcano—about twenty or so. Only a few of the world's volcanoes are sufficiently active or have been studied long enough for scientists to recognize patterns in their activity. Ironically, eruptions must occur for reliable forecasts to be made.

Volcano Monitoring

No single approach to volcano forecasting can yield trustworthy results, but the combined data of geophysics, geology, and geochemistry, interpreted within the context of long-term eruptive behavior of a given volcano, can often be compiled into reliable forecasts. As with any forecasting, the reliability of volcanic forecasts increases with the quantity and diversity of information available to volcanologists. The 1980 awakening of Mount St. Helens after more than a century of dormancy was predicted by U.S. Geological Survey (USGS) scientists in a very general way, leading to an evacuation several weeks before the eruption. After the major eruption, continuous monitoring of smaller, more numerous eruptions led to more specific forecasts, generally reliable to within a few hours or days. Specific, reliable forecasts are also routinely issued by the U.S. Geological Survey at the Kīlauea volcano, which has erupted dozens of times since the USGS Hawaiian Volcano Observatory (HVO) was established there in 1912. Other US observatory monitor programs include the Yellowstone Volcano Observatory (YVO), Alaska Volcano Observatory (AVO), USGS California Volcano Observatory (CalVO), and USGS Cascades Volcano Observatory (CVO). These observatories belong to the World Organization of Volcano Observatories (WOVO).

The geophysical equipment used in volcano monitoring is diverse, and some instruments are also used for industrial applications. The temperatures of lava and gases can be measured with a thermocouple—a pair of wires of different metals that are welded together at both ends. When one end of the circuit is immersed in hot material, an electrical current is generated, its strength depending on the temperature difference between both ends of the circuit. An ammeter near the cold end of the circuit reads the electrical current, from which the temperature at the hot end of the circuit can be calculated. Another temperature-measuring instrument is the optical pyrometer, which measures the wavelengths of visible and near-infrared radiation emitted from the surfaces of incandescent objects.

Magnetometers measure the strength and sometimes the direction of the Earth's magnetic field. Over large regions, the instruments can be used from airplanes, but volcano monitoring usually involves ground measurements at fixed stations that are reoccupied to get comparative readings. Tumescence and faulting of the ground are measured with surveying instruments such as steel tapes, transits, levels, or theodolites. A network of survey stations and reflective targets is established on a volcano, and measurements are repeated to recognize any deformation of the ground that might be associated with the movement of underground magma. Tiltmeters are used to measure changes in the slope or inclination of the ground surface. Using the same principle as a carpenter's level, electronic tiltmeters employ sensitive bubbles to measure tilt in two directions at right angles. Thus, the direction and magnitude of tilting can be monitored. Anchored on the ground, the instruments are extremely sensitive—the amount of tilt is expressed in microradians, the equivalent of lifting one end of a one-kilometer-long rod by one millimeter.

Volcanic earthquakes are monitored with a network of seismometers encircling the volcanic area. The instruments can be tuned to detect very small, unfelt earthquakes with Richter magnitudes less than 1. With each shock, several types of seismic waves are emitted from the site of an earthquake. Because different waves travel at different speeds, their arrival times can be used to calculate the distance between the seismometer and the earthquake center in much the same way that the time lag between lightning and thunder can be used to estimate the distance to the lightning.

Gravimeter readings can be repeatedly taken at fixed stations to measure any changes in the Earth's gravitational force. In one form, the instrument consists of a weight suspended from a sensitive spring; variations in gravity cause variations in the extension of the spring. Several mechanical and optical schemes have been developed to measure this very small deflection, and the most sensitive instruments can detect gravitational changes of one part in ten million. Measurable changes sometimes occur when magma is either lighter or heavier than the underground rocks that it displaces or if the ground surface has been uplifted by the magma. In the latter case, the gravitational force is diminished because the gravimeter is farther away from the Earth's center of mass.

Collections of volcanic gases are made directly at vents with evacuated and flow-through sampling tubes or indirectly by using remote devices. One such device is the correlation spectrometer (COSPEC), an instrument that measures the output of sulfur dioxide gas from volcanic plumes at a safe distance. The COSPEC can be used from the ground or an airplane. It measures the amount of solar ultraviolet light absorbed by sulfur dioxide gas in the plume, compares it to a standard, and thus measures the amount of sulfur dioxide. In 1983, sulfur dioxide emissions at the Kīlauea volcano averaged about 250 tons per day, while Mount St. Helens emitted about 100 tons per day, down from its 1980 peak of about 1,500 tons per day. By the late 2010s, Kīlauea emitted between 500 and 14,000 metric tons of sulfur dioxide each day. During the Lower East Rift Zone eruption in 2018, Kīlauea emissions exceeded 100,000 metric tons per day.

The practice of forecasting eruptions is shifting from description-based models to physics-based models with real-time data monitoring to improve prediction accuracy.

Principal Terms

correlation spectrometer (COSPEC): an instrument that measures the output of sulfur dioxide from volcanic gas plumes

earthquake swarm: a number of earthquakes that occur close together and closely spaced in time

tiltmeter: an instrument that precisely measures tilting of the ground surface

tumescence: a local swelling of the ground that commonly occurs when magma rises toward the surface

volcanic earthquakes: small-magnitude earthquakes that occur at relatively shallow depths beneath active or potentially active volcanoes

volcanic tremor: a continuous vibration of long duration, detected only at active volcanoes

Bibliography

Acocella, Valerio. "How Scientists Forecast Volcanic Eruptions." Frontiers, 28 Mar. 2022, kids.frontiersin.org/articles/10.3389/frym.2022.813910. Accessed 20 July 2024.

Acocella, V., et al. "Towards Scientific Forecasting of Magmatic Eruptions." Nature Reserve Earth Environment, vol. 5, 2024, pp. 5–22. doi.org/10.1038/s43017-023-00492-z.

Chester, Roy. Furnace of Creation, Cradle of Destruction. AMACOM Books, 2008.

Decker, Robert, and Barbara Decker. Volcanoes. 4th ed. W. H. Freeman, 2005.

Francis, Peter, and Clive Oppenheimer. Volcanoes. 2nd ed. Oxford UP, 2004.

Grattan, John, and Robin Torrence, eds. Living Under the Shadow: Cultural Impacts of Volcanic Eruptions. Left Coast Press Inc., 2007.

Heliker, C., et alt. Earthquakes and Volcanoes, vol. 18, 1986.

Mason, Betsy. “Season of Fire.” Discover, vol. 24, 2003, pp. 32-39.

Paone, Angelo, and Sung-Hyo Yun. Forecasting Volcanic Eruptions. IntechOpen, 2020.

Papale, Paolo. Forecasting and Planning for Volcanic Hazards, Risks, and Disasters. Elsevier, 2021.

Sigurdsson, Haraldur, ed. Encyclopedia of Volcanoes. 2nd ed., Academic Press, 2015.

Trombley, R. B. The Forecasting of Volcanic Eruptions. Lincoln, Neb. iUniverse, 2006.