A forecast for a volcanic eruption is a probabilistic statement concerning the time, place, and character of an eruption before it occurs. It is analogous to forecasting the weather and is not as precise a statement as a prediction. Forecasting volcanic eruptions is a major component of the goal to reduce volcanic hazards. It is unlikely that we will be able to forecast the majority of volcanic activity accurately in the near future, but valuable information is being gathered about phenomena that occur before eruptions. One problem is that most forecasting techniques require experience with actual eruptions before the mechanism is understood. Thus, we are better able to predict eruptions in the Hawaiian Islands then elsewhere because we have had so much experience there.
The methods of forecasting volcanic eruptions include:
- Monitoring of seismic activity.
- Monitoring of thermal, magnetic, and hydrologic conditions.
- Topographic monitoring of tilting or swelling of the volcano.
- Monitoring of volcanic gas emissions.
- Studying the geologic history of a particular volcano or volcanic centre.
Seismic Activity
Our experience with volcanoes, such as Mount St. Helens and those on the big island of Hawaii, suggests that earthquakes often provide the earliest warning of an impending volcanic eruption. In the case of Mount St. Helens, earthquake activity started in mid-March before the eruption in May. Activity began suddenly, with near-continuous shallow seismicity. Unfortunately, there was no increase in earthquakes immediately before the May 18 event. In Hawaii earthquakes have been used to monitor the movement of magma as it approaches the surface. Several months before the 1991 Mt. Pinatubo eruptions, small steam explosions and earthquakes began. Mt. Pinatubo (present elevation 1700 m, or 5578 ft) was an eroded ridge, and, as a result, did not have the classic shape of a volcano. Furthermore, it had not erupted in 500 years; most of the people living near it did not even know it was a volcano! Scientists began monitoring earthquake activity and studying past volcanic activity, which was determined to be explosive. Earthquakes increased in number and magnitude before the catastrophic eruption, migrating from deep beneath the volcano to shallow depths beneath the summit.
Geophysicists have proposed a generalized model for seismic activity that may help in predicting eruptions. The model is for explosive composite volcanoes, such as those in the Cascade Mountains, which may awaken after an extended period of inactivity. As a dormant volcano reawakens, the magma must fracture and break previously solidified igneous rock above the magma chamber in order to work its way to the surface. Several weeks before an eruption, increasing pressure creates numerous fractures in the plugged volcano conduit above the chamber. At first, the increase in seismic events will be very gradual, and a seismologist may need 10 days or so to confidently recognize an accelerating trend toward an eruption. Once the trend has been recognized, there will still be several days before the eruption occurs. Unfortunately, this short warning time may be insufficient for a large-scale evacuation. Thus to forecast eruptions, it may be best to use seismic activity in concert with other eruption precursors discussed below. It is fortunate that, in contrast to earthquakes, volcanoes provide warning signs prior to eruption.
Thermal, Magnetic, and Hydrologic Monitoring
Monitoring of volcanoes is based on the fact that, before an eruption, a large volume of magma moves up into some sort of holding reservoir beneath the volcano. The hot material changes the local magnetic, thermal, hydrologic, and geochemical conditions. As the surrounding rocks heat, the rise in temperature of the surficial rock may be detected by remote sensing or infra-red aerial photography. Increased heat may melt snowfields or glaciers; thus, periodic remote sensing of a volcanic chain may detect new hot points that could indicate potential volcanic activity. This method was used with some success at Mount St. Helens before the main eruption on May 18, 1980. When older volcanic rocks are heated by new magma, magnetic properties, originally imprinted when the rocks cooled and crystallized, may change. These changes can be detailed by ground or aerial monitoring of the magnetic properties of the rocks that the volcano is composed of.
Topographic Monitoring
Monitoring topographic changes and seismic behaviour of volcanoes has been useful in forecasting some volcanic eruptions. The Hawaiian volcanoes, especially Kilauea, have supplied most of the data. The summit of Kilauea tilts and swells before an eruption and subsides during the actual outbreak. Kilauea also undergoes earthquake swarms that reflect moving subsurface magma and an imminent eruption. The tilting of the summit in conjunction with the earthquake swarms was used to predict a volcanic eruption in the vicinity of the farming community of Kapoho on the flank of the volcano, 45 km (28 mi) from the summit. As a result, the inhabitants were evacuated before the event, in which lava overran and eventually destroyed most of the village. Because of the characteristic swelling and earthquake activity before eruptions, scientists expect the Hawaiian volcanoes to continue to be more predictable than others. Monitoring of ground movements such as tilting, swelling, opening of cracks, or changes in the water level of lakes on or near a volcano has become a useful tool for recognizing change that might indicate a coming eruption. Today, satellite based radar and a network of Global Positioning System (GPS) receivers can be used to monitor change in volcanoes, including surface deformation, without sending people into a hazardous area.
Monitoring Volcanic Gas Emissions
The primary objective of monitoring volcanic gas emissions is to recognize changes in the chemical composition of the gases. Changes in both gas composition that is, the relative amounts of gases such as steam, carbon dioxide, and sulphur dioxide and gas emission rates are thought to be correlated with changes in subsurface volcanic processes. These factors may indicate movement of magma toward the surface. This technique was useful in studying eruptions at Mount St. Helens and Mt. Pinatubo. Two weeks before the explosive eruptions at Mt. Pinatubo, the emissions of sulphur dioxide increased by a factor of about 10.
Geologic History
Understanding the geologic history of a volcano or volcanic system is useful in predicting the types of eruptions likely to occur in the future. The primary tool used to establish the geologic history of a volcano is geologic mapping of volcanic rocks and deposits. Attempts are made to date lava flows and pyroclastic activity to determine when they occurred. These are the primary data necessary to produce maps depicting volcanic hazards at a particular site. Geologic mapping, in conjunction with the dating of volcanic deposits at Kilauea, Hawaii, led to the discovery that more than 90 percent of the land surface of the volcano has been covered by lava in only the past 1500 years. The town of Kalapana, destroyed by lava flows in 1990, might never have been built if this information had been known before development, because the risk might have been thought too great. The real value of geologic mapping and dating of volcanic events is that they allow development of hazard maps to assist in land-use planning and preparation for future eruptions. Such maps are now available for a number of volcanoes around the world.
Volcanic Alert or Warning
Geologic behavior, color-coded condition, and response: Volcanic Hazards Response Plan; Long Valley Caldera, California. |
At what point should the public be alerted or warned that a volcanic eruption may occur? This is an important question being addressed by volcanologists. At present, there is no standard code, but one being used with various modifications has been developed by the U.S. Geological Survey. The system is colour coded by condition; each colour green, yellow, orange, and red denotes increasing concern. This table was created specifically for the Long Valley caldera in California. Similar systems have been or are being developed for other volcanic areas, including Alaska and the Cascade Mountains of the Pacific Northwest. The colour-coded system is a good start; however, the hard questions remain: When should evacuation begin? When is it safe for people to return? Evacuation is definitely necessary before condition red, but when, during conditions yellow or orange, should it begin?