The term debris avalanche is used to refer to the sudden and very rapid movement of an incoherent, unsorted mass of rock and soil mobilized by gravity (Schuster and Crandell, 1984). Movement is characterized by flowage in a dry or wet state, or both. Debris avalanches commonly originate in massive rockslides which, during their movement, disintegrate into fragments ranging in size from small particles to blocks hundreds of meters across. If the avalanche has a large water content, its matrix may continue to flow downslope as a lahar after its coarser parts have come to rest.
Volcanic-debris avalanches occur occasionally at large, steep-sided volcanoes and are among the most hazardous of volcanic events (Voight and others, 1981; Crandell and others, 1984. Such avalanches form when part of a volcanic edifice fails catastrophically and moves downslope. Disruption of a volcanic cone may be the result of intrusion of magma and earthquake shaking, as at Mount St. Helens in 1980 (Voight and others, 1981), or the result of a volcanic explosion as at Bezymianny in Kamchatka, U.S.S.R., in 1956 (Gorshkov, 1959; Bogoyavlenskaya and others, 1985). Steep-sided volcanoes may also fail from other causes, e.g., after gradual weakening by hydrothermal alteration, or after heavy rains which may saturate and weaken parts of the edifice.
Debris avalanches typically produce thick hummocky deposits that can extend tens of kilometers from a volcano and cover hundreds of square kilometers. A debris avalanche that occurred at Mount Shasta between about 300,000 and 360,000 yrs ago (Crandell and others, 1984) traveled more than 64 km from the summit of the volcano, covered more than 675 km2, and had a volume of at least 45 km3 (D. R. Crandell, personal commun., 1986).
Debris avalanches can destroy everything in their paths by impact or burial beneath tens of meters of debris. Because debris avalanches can occur with little or no warning and can travel at high speeds (Voight and others, 1981), areas that might be affected should be evacuated if an avalanche is anticipated.
Pyroclastic flows are high-density mixtures of hot, dry rock fragments and hot gases that move away from their source vents at high speeds. They may result from the explosive eruption of molten or solid rock fragments, or both, or from the collapse of vertical eruption columns of ash and larger rock fragments. Pyroclastic flows may also result from a laterally directed explosion, or the fall of hot rock debris from a dome or thick lava flow.
Rock fragments in pyroclastic flows range widely in grain size and consist of dense rock, pumice, or both. Individual pyroclastic flows, worldwide, range in length from less than one to more than 200 km, cover areas from less than one to more than 20,000 km2, and have volumes from less than 0.001 to more than 1000 km3 (Crandell and others, 1984). Pumiceous pyroclastic flows with volumes of 1-10 km3 can reach distances of several tens of kilometers from a vent and travel downslope at speeds of 50 to more than 150 km/hr (Crandell and Mullineaux, 1978), their velocity depending largely on their volume and on the steepness of slopes over which they travel. Pyroclastic flows and their deposits commonly contain rock debris and gases with temperatures of several hundred degrees Celsius (Banks and Hoblitt, 1981; Blong, 1984, p. 36).
Most pyroclastic flows consist of two parts: a basal flow of coarse fragments that moves along the ground, and a turbulent cloud of finer particles (ash cloud) that rises above the basal flow (Crandell and Mullineaux, 1978). Ash may fall from the cloud over a wide area downwind from the basal flow.
Pyroclastic flows generally follow valleys or other depressions, but can have enough momentum to overtop hills or ridges in their paths. The larger the mass of a flow and the faster it travels, the higher it will rise onto obstacles in its path. Some pumiceous pyroclastic flows erupted during the climactic eruptions of Mount Mazama (Crater Lake) about 6850 years ago moved 231 m upslope to cross a divide 17 km from the volcano (Crandell and others, 1984) and ultimately reached a downvalley distance of 60 km from the vent (Williams, 1942; Bacon, 1983).
Pyroclastic flows are extremely hazardous because of their high speeds and temperatures. Objects and structures in their paths are generally destroyed or swept away by the impact of debris or by accompanying hurricane-force winds (Blong, 1984). Wood and other combustible materials are commonly burned by the basal flow; people and animals may also be burned or killed beyond the margins of a pyroclastic flow by inhalation of hot ash and gases.
Pyroclastic flows have been erupted repeatedly at many volcanic centers in the Cascade Range during Holocene time. Moreover, large silicic magma chambers may exist at several volcanic centers in the Cascade Range that have had explosive eruptions of large volume (101 - 102 km3). Such eruptions can produce pyroclastic flows which could travel more than 50 km from a vent and could be extremely destructive over wide areas. Because pyroclastic flows move at such high speeds, escape from their paths is unlikely once they start to move; areas subject to pyroclastic flows must be evacuated before flows are formed.
Pyroclastic surges are turbulent, low-density clouds of rock debris and air or other gases that move over the ground surface at high speeds. They typically hug the ground and depending on their density and speed, may or may not be controlled by the underlying topography. Pyroclastic surges are of two types: "hot" pyroclastic surges that consist of "dry" clouds of rock debris and gases that have temperatures appreciably above 100o C, and "cold" pyroclastic surges, also called base surges, that consist of rock debris and steam or water at or below a temperature of 100o C (Crandell and others, 1984).
Both hot and cold pyroclastic surges damage or destroy structures and vegetation by impact of rock fragments moving at high speeds and may bury the ground surface with a layer of ash and coarser debris tens of centimeters or more thick (Crandell and others, 1984). Because of their high temperatures, hot pyroclastic surges may start fires and kill or burn people and animals. Both types of surges can extend as far as 10 km from their source vents and devastate life and property within their paths. During an eruption of Mont Pelee on Martinique in 1902, a cloud of hot ash and gases swept into the town of St. Pierre at an estimated speed of 160 km/hr or more (Macdonald, 1972). About 30,000 people died within minutes, most from inhalation of hot ash and gases. Pyroclastic surges have occurred at volcanoes in the Cascade Range in the past and can be expected to occur again. Future cold surges (base surges) are most likely to occur where magma can contact water at volcanic vents near lakes, those that have crater lakes, and at vents in areas with a shallow water table.
Volcanic blasts are explosions which may be directed vertically or at some lower angle. Vertically directed explosions may produce mixtures of rock debris and gases that flow, motivated chiefly by gravity, down one or more sides of a volcano. Such a blast at Mount Lamington, New Guinea, in 1952 produced pyroclastic surges that moved down all sides of the volcano, killing about 3,000 people and destroying nearly everything within an area of about 230 km2 (Taylor, 1958).
A volcanic explosion that has a significant low-angle component and is principally directed toward a sector of no more than 180o is referred to as a lateral blast (Crandell and Hoblitt, 1986). Such a blast may produce a mixture of rock debris and gases hundreds of meters thick that moves at high speed along the ground surface as a pyroclastic flow, a pyroclastic surge, or both. The high velocity of the mixture of rock debris and gases, which may be at least 100 m/s, is due both to the initial energy of the explosion and to gravity as the mixture moves downslope.
Lateral blasts may affect only narrow sectors or spread out from a volcano to cover a sector as broad as 180o, and they can reach distances of several tens of kilometers from a vent (Crandell and Hoblitt, 1986). The resulting deposits form a blanket of blocks, lapilli, and ash that thins from a few meters near the source to a few centimeters near the margin (Hoblitt and others, 1981; Waitt, 1981; Moore and Sisson, 1981). Because of they carry rock debris at high speeds, lateral blasts can devastate areas of tens to hundreds of square kilometers within a few minutes, and can destroy manmade structures and kill all living things by abrasion, impact, burial, and heat.
A lateral blast at Mount St. Helens in 1980 moved outward at a speed of at least 100 m/s (Malone and others, 1981), devastated an area of 600 km2 out to a distance of 28 km from the volcano, and killed more than 60 people (Christiansen and Peterson, 1981). A similar blast in 1956 at Bezymianny volcano, U.S.S.R., affected an area of about 500 km2 out to a distance of 30 km from the volcano (Gorshkov, 1959; Bogoyavlenskaya, and others, 1985). Both events were closely associated with debris avalanches.
Volcanic blasts are most likely at steep-sided stratovolcanoes and may occur when viscous gas-rich magma is emplaced at a shallow level within the volcano (Bogoyavlenskaya and others, 1985). For purposes of long-range land-use planning, Crandell and Hoblitt (1986) have suggested that circular hazard zones with a radius of 35 km be drawn around symmetrical volcanoes where lateral blasts are possible. The sector beyond the volcano that is most likely to be affected cannot be forecast unless and until precursory seismic activity and deformation suggest the possible site of a lateral blast (Gorshkov, 1963; Crandell and Hoblitt, 1986). Although short-term warnings suggested by such precursory activity obviously are not useful for determining safe locations for fixed structures, they may allow people to evacuate threatened areas (Crandell and Hoblitt, 1986).
Lava flows are streams of molten rock that erupt relatively nonexplosively from a volcano and move downslope. The distance traveled by a lava flow depends on such variables as the effusion rate, fluidity of the lava, volume erupted, steepness of the slope, channel geometry, and obstructions in the flows path. Basalt flows are characterized by relatively low viscosity and may reach more than 50 km from their sources; in fact, one Icelandic basalt flow reached 150 km (Williams and McBirney, 1979). Andesite flows have higher viscosity and few extend more than 15 km; however, one andesite flow of Pleistocene age in the Cascades is 80 km long (Warren, 1941). Because of their high viscosity, dacite and rhyolite lava extrusions typically form short, thick flows or domes.
Lava flows cause extensive damage or total destruction by burning, crushing, or burying everything in their paths. They seldom threaten human life, however, because of their typically slow rate of movement, which may be a few meters to a few hundred meters per hour. In addition, their paths of movement generally can be predicted. However, lava flows that move onto snow or ice can cause destructive lahars and floods, and those that move into forests can start fires. The flanks of moving lava flows typically are unstable and collapse repeatedly, occasionally producing small explosive blasts or small pyroclastic flows.
Lava flows have been erupted at many vents in the Cascade Range during Holocene time; their compositions range from basalt to rhyolite. The longest known basalt, andesite, and rhyolite lava flows erupted at Cascade volcanic centers during Holocene time are, respectively, the 45-km- long Giant Crater basalt flow at Medicine Lake volcano, the 12-km-long Schriebers Meadow andesite flow at Mount Baker, and the 2-km-long Rock Mesa rhyolite flow at Three Sisters. Lava flows of varied composition are likely to erupt again in the Cascade Range and will endanger all non-moveable objects in their paths.
Volcanic domes are mounds that form when viscous lava is erupted slowly and piles up over the vent, rather than moving away as a lava flow. The sides of most domes are very steep and typically are mantled with unstable rock debris formed during or shortly after dome emplacement. Most domes are composed of silica-rich lava which may contain enough pressurized gas to cause explosions during dome extrusion.
The direct effects of dome eruption include burial or disruption of the preexisting ground surface by the dome itself and burial of adjacent areas by rock debris shed from the dome. Because of their high temperatures, domes may start fires if they are erupted in forested areas. Domes are extruded so slowly that they can be avoided by people, but they may endanger man-made structures that cannot be moved. The principal hazard associated with domes is from pyroclastic flows produced by explosions or collapses. Such pyroclastic flows can occur without warning during active dome growth and can move very rapidly, endangering life and property up to 20 kilometers from their sources (Miller, 1978; 1980). Such pyroclastic flows can also cause lahars if they are erupted onto snow and ice or incorporate water during movement.
Domes ranging in composition from dacite to rhyolite have been erupted repeatedly during late Pleistocene and Holocene time in the Cascade Range (Appendix A). Domes at Mount Shasta, Mount St. Helens, Glacier Peak, Mount Hood, and near Lassen Peak have collapsed or exploded to produce hot pyroclastic flows, some extending as far as 20 km from their sources (Miller, 1980). Lines of domes erupted at Medicine Lake and South Sister volcanoes within the last several thousand years appear to have formed over short intervals of time when vertical dikelike magma bodies reached the surface (Fink and Pollard, 1983; Scott, 1987). Dome emplacement typically follows more explosive eruptions.
Lahars (also called volcanic debris flows or mudflows) are mixtures of water-saturated rock debris that flow downslope under the force of gravity. For simplicity in the discussions and compilations in this report, we have followed the usage of Crandell and others (1984) and used the term lahar to include both true lahars (Crandell, 1971), and downstream lahar-runout flows (Scott, 1985). Lahar-runout flows are hyperconcentrated streamflows that form by downstream transformation of lahars through loss of sediment and dilution by streamflow (Pierson and Scott, 1985; Scott; 1985, 1986). Additional dilution downstream may result in transformation of hyperconcentrated flows into normal streamflows, or floods.
Rock debris in lahars ranges in size from clay to blocks several tens of meters in maximum dimension. When moving, lahars resemble masses of wet concrete and tend to be channeled into stream valleys. Lahars are formed when loose masses of unconsolidated, wet debris become mobilized. Rocks within a volcano may already be saturated, or water may be supplied by rainfall, by rapid melting of snow or ice, or by a debris-dammed lake or crater lake. Lahars may be formed directly when pyroclastic flows or pyroclastic surges are erupted onto snow and ice, as apparently occurred in November 1985 at Nevado del Ruiz, in Columbia, where about 23,000 people lost their lives (Herd and Comite\" de Estudios Vulcanologicos, 1986). Lahars may be either hot or cold, depending on the temperature of the rock debris they carry.
Lahars can travel great distances down valleys, and lahar fronts can move at high speeds--as much as 100 km/hr. Lahars produced during an eruption of Cotopaxi volcano in Ecuador, in 1877, traveled more than 320 km down one valley at an average speed of 27 km/hr (Macdonald, 1972). Lahars that descended the southeast flank of Mount St. Helens in 1980 had initial flow velocities that exceeded 100 km/hr; average lahar flow velocities were about 67 km/hr over the 22.5 km traveled before the lahars entered a reservoir (Pierson, 1985). High-speed lahars may climb valley walls on the outside of bends, and their momentum may also carry them over obstacles. Lahars confined in narrow valleys, or dammed by constrictions in valleys, can temporarily thicken and fill valleys to heights of 100 m or more (Crandell, 1971).
The major hazard to human life from lahars is from burial and impact by boulders and other debris. Buildings and other property in the path of a lahar can be buried, smashed, or carried away. Because of their relatively high density and viscosity, lahars can move and carry away vehicles and other large objects such as bridges.
An inverse relation exists between the volume and length of lahars and their frequency; that is, large lahars are far less frequent than small ones (see lahar frequency plots in Chapter 4). For this reason, lahar hazard progressively decreases downvalley from a volcano, and at any point along the valley, hazard from lahars decreases with increasing height above the valley floor.
Lahars have occurred repeatedly during eruptions at snow-covered volcanoes in the northwestern U. S. during Holocene time (Appendix A). Large lahars originating in debris avalanches have occurred at Mounts Shasta, Hood, St. Helens, Rainier, and Baker, and some have been caused by the failure of debris- or moraine-dammed lakes. Small lahars are frequently generated at ice-covered volcanoes by climatic events such as heavy rainstorms and periods of rapid snowmelt due to hot weather (Miller, 1980).
Floods related to volcanism can be produced by melting of snow and ice during eruptions of ice-clad volcanoes, by heavy rains that may accompany eruptions, and by transformation of lahars to stream flow. Floods carrying unusually large amounts of rock debris can leave thick deposits at and beyond the mouths of canyons and on valley floors leading away from volcanoes. Eruption-caused floods can occur suddenly and can be of large volume; if rivers are already high because of heavy rainfall or snow melt, such floods can be far larger than normal.
Danger from eruption-caused floods is similar to that from floods having other origins, but floods caused by eruptions may be more damaging because of an unusually high content of sediment. The hydrology of river systems may be altered for decades following the rapid accumulation of great quantities of sediment (e.g., U.S. Army Corps of Engineers, 1984). Subsequent reworking of this sediment may lead to further channel aggradation, and aggravate overbank flooding during high river stages. Floods can also be generated by waves in lakes that overtop or destroy natural or man-made dams; such waves can be produced by large masses of volcanic material moving into the lake suddenly as a debris avalanche, lahar, or pyroclastic flow.
Tephra consists of fragments of lava or rock blasted into the air by explosions or carried upward by a convecting column of hot gases (e.g., Fisher and Schmincke, 1984; Shipley and Sarna-Wojcicki, 1983). These fragments fall back to earth on and downwind from their source volcano to form a tephra, pyroclastic-fall, or volcanic "ash" deposit. Large fragments fall close to the erupting vent, and progressively smaller ones are carried farther away by wind. Dust-size particles can be carried many hundreds of kilometers from the source. Tephra deposits blanket the ground with a layer that decreases in thickness and particle size away from the source. Near the vent, tephra deposits may be tens of meters thick. According to Blong (1984), rates of drift of clouds containing ash are usually in the range of 20-100 km/hr, but can be higher where wind speeds are higher.
Tephra deposits consist of combinations of pumice, glass shards, dense-rock, and crystals that range in size from ash (< 2 mm), through lapilli (2-64 mm), to blocks (> 64 mm). Eruptions that produce tephra range from those that eject debris only a few meters into the air, to cataclysmic explosions that throw debris to heights of several tens of kilometers. Explosive eruptions that produce voluminous tephra deposits also typically produce pyroclastic flows.
Effects of tephra are closely related to the amount of material deposited and its grain size. Thickness versus distance relationships for several well-known tephra deposits in the Cascade Range are shown in Figure 3-1. Figure 3-2 shows median particle diameter versus distance from source for various tephra deposits. The relationship generally approximates an exponential one, but shows wide scatter. Within about 100 km of a vent, the median particle diameter of a tephra deposit varies by several orders of magnitude depending on the intensity of the eruption, fall velocity of particles, and velocity of the wind. Beyond several hundred kilometers, the mean particle diameter typically is silt-size (about 0.063 mm) or less, but still shows considerable variation.
Tephras generally do not completely destroy facilities or kill people; instead they adversely affect both in many ways. Tephra can be carried to great distances and in all directions; no site in the Pacific Northwest is immune from tephra hazards. The magnitude of hazard from tephra varies directly with deposit thickness. In general, deposit thickness and grain size decrease with increasing distance from a vent. However, the tephra fall from the May 18, 1980, eruption of Mount St. Helens displays a secondary maximum of tephra thickness about 300 km from the volcano (Sarna-Wojcicki and others, 1981). Carey and Sigurdsson (1982) proposed that aggregation of very fine ash into larger particles caused premature fallout at the secondary thickness maximum; they suggested that the same process may accompany other tephra eruptions. The few data points for some of the larger tephra falls in Figure 3-1, and the problems ofdetermining original fall thicknesses from prehistoric deposits, leaves open the possibility of secondary thickness maxima in these layers. The tephra-thickness plot for Mazama tephra is a composite from many sources and suggests that a secondary thickness maximum may occur in the Mazama tephra at about 200- 500 km from its vent. Alternatively, it may reflect varying methods used by different workers to determine original fall thickness.
Close to an erupting vent, the main tephra hazards to man-made structures include high temperatures, burial, and impact of falling fragments. Large blocks thrown on ballistic trajectories from an erupting vent can damage structures and kill or injure unprotected people. Most blocks will fall within 5 km of the vent (Blong, 1984), but unusually powerful explosions may throw some blocks at least twice as far (Crandell and Hoblitt, 1986). Hot tephra may set fire to forests and flammable structures, but this is not likely to be a hazard beyond a distance of 15-20 km. Structural damage can also result from the weight of tephra, especially if it is wet. A tephra layer 10 cm thick may weigh 20-100 kg/m2 when dry, but 50-150 kg/m2 when wet (Crandell and others, 1984). Also, tephra is much more cohesive when wet than when dry, and can adhere to steeper surfaces and is much more difficult to remove. Tephra 10 cm or more thick may cause buildings to collapse (Blong, 1984, p.212). Drifting of tephra by winds can locally increase accumulations and loads on sloping structures far above that resulting from unmodified fall thicknesses.
At distances of tens to hundreds of kilometers, the chief hazards from tephra falls are the effects of ash on machinery and electrical equipment and on human and animal respiratory systems. Ash only 1 cm thick can impede the movement of most vehicles and disrupt transportation, communication, and utility systems (Schuster, 1981, 1983; Warrick and others, 1981). Machinery is especially susceptible to the abrasive and corrosive effects of ash (Schuster, 1981, 1983; Shipley and Sarna-Wojcicki, 1983).
Specific possible effects of airfall tephra on nuclear power plants have been outlined by Shipley and Sarna-Wojcicki (1983). They include (1) loading of structures, particularly by thick accumulations of wet tephra, (2) clogging of water and air filtering systems by influx of tephra, (3) abrasive effects of ash on machinery, (4) corrosion and shorting out of electrical systems by freshly fallen ash (Sarkinen, 1980), (5) effects of tephra accumulations in circulation of water-cooling systems, and (6) a variety of secondary or indirect effects on maintenance and emergency systems that may be impacted by factors 1-5. Shipley and Sarna-Wojcicki (1983) also pointed out the likelihood of "cascading effects", when the impact of tephra on one function or group of functions impairs additional dependent systems, each of which may produce further cascading effects.
In addition to the specific effects discussed above, the fall of tephra may severely decrease visibility or cause darkness, which could further disrupt transportation, disrupt outdoor activities, and possibly result in psychological stress and panic even among people whose lives are not threatened (Blong, 1984). These effects could impair the ability of personnel to perform even routine tasks in areas affected by tephra fall. A wide range of compositions and volumes of tephra have been erupted during the past 15,000 years from Cascade volcanoes (Appendix A). These tephra deposits range in volume from the 116-km3 Mazama tephra (Bacon, 1983; Druitt and Bacon, 1986) to those of only a few thousand cubic meters. The May 18, 1980 eruption of Mount St. Helens deposited an estimated minimum volume of 1.1 km3 of uncompacted tephra on areas east-northeast of the volcano (Sarna-Wojcicki and others, 1981). Most tephra eruptions in the Cascade Range have produced elongate lobe-shaped deposits that extend primarily into a broad sector northeast of the source volcano owing to prevailing wind directions.
Relatively small volumes (< 0.1 km3) of basaltic and basaltic-andesite tephra have been erupted at many vents during Holocene time. Such eruptions have been far less explosive than more silicic eruptions and have produced cinder cones and tephra deposits that are restricted chiefly to within a few tens of kilometers downwind. Similar small-volume eruptions of tephra are anticipated in the future at new vents within fields of basaltic volcanism in the Cascade Range.
All magmas contain dissolved gases that are released both during and between eruptive episodes. Volcanic gases generally consist predominantly of steam (H2O), followed in abundance by carbon dioxide and compounds of sulfur and chlorine (Wilcox, 1959; Thorarinsson, 1979). Minor amounts of carbon monoxide, fluorine and boron compounds, ammonia, and several other compounds are found in some volcanic gases.
The distribution of volcanic gases is mostly controlled by the wind; they may be concentrated near (1-10 km) a vent but become diluted rapidly downwind. Even very dilute gases can have a noticeable odor and can harm plants and some animals tens of kilometers downwind from a vent.
Within about 10 km of a vent, volcanic gases can endanger life and health as well as property. Acids, ammonia, and other compounds present in volcanic gases can damage eyes and respiratory systems of people and animals, and heavier-than-air gases, such as carbon dioxide, can accumulate in closed depressions and suffocate people or animals. Corrosion of metals and other susceptible materials can also be severe (Crandell and others, 1984; Blong, 1984).
Three main sources of earthquakes in the vicinity of volcanoes (Blong, 1984) are (1) those generated by the movement of magma or by formation of cracks through which magma can move, and those resulting from gas explosions within a conduit; (2) other earthquakes that result from readjustments of a volcanic edifice following eruption or movement of magma; and (3) tectonic earthquakes, which may also facilitate the rise of magma. Volcanic earthquakes belonging to the first category rarely have Richter magnitudes greater than 5.0 (Okada and others, 1981; Latter, 1981) and generally have foci at depths of less than 10 km. Damage from such earthquakes is limited to a relatively small area (Rittmann, 1962; Shimozuru, 1972).
The relationship between volcanic activity and earthquakes of categories 2 and 3 above is less well understood. Few quantitative data are available concerning the maximum magnitude of such earthquakes, although events larger than magnitude 5 have been described. A sequence of tectonic earthquakes that occurred near Mammoth Lakes, California, in 1980 included four events of magnitude 6+ (Urhammer and Ferguson, 1981); these may have been triggered by magmatic processes (Bailey, 1981). One of the largest earthquakes of possible magmatic origin occurred at Sakura-jima volcano, Japan, in 1914. The earthquake had a focal depth of 13 km, a magnitude of 6.7 (Shimozuru, 1972), and caused considerable damage and some loss of life in Kagoshima, 10 km from the volcano. Earthquakes at least as large as magnitude 7.2 have occurred on Kilauea volcano, Hawaii (Tilling and others, 1976); however, these earthquakes are related to displacements of large sectors of the volcanic edifice rather than to a specific volcanic event (Swanson and others, 1976) and thus resemble tectonic earthquakes.
In summary, earthquakes directly associated with movement or eruption of magma seldom exceed a magnitude of about 5.0, and structures at distances greater than a few tens of kilometers from the volcano are not likely to be damaged by such events. Structures situated outside of the proximal-hazard zone are not likely to be damaged by volcanic seismicity. Volcanoes located in geologic settings that are tectonically active are likely to be at risk from tectonic earthquakes that are far larger than volcanogenic ones. Structures sited and designed to withstand the maximum credible tectonic earthquake should not be threatened by volcanogenic seismicity.
Eruption induced atmospheric shock waves are strong compressive waves driven by rapidly moving volcanic ejecta. Although most volcanic eruptions are not associated with such waves, a number of examples are known. Some of the eruptions best known for this type of behavior are: Vesuvius, 1906 (Perret, 1912); Krakatau, 1883 (Verbeek, 1885, in Simkin and Fiske, 1983); Tambora, 1815 (Stewart, 1820); Sakura-jima, 1914 (Omori, 1916); and Asama, 1958 (Aramaki, 1956). Air-shock waves can be sufficiently energetic to damage structures far from their source. The 1815 eruption of Tambora, on the island of Sumbawa, produced a shock wave that broke windows at a distance of about 400 km (Stewart, 1820). In 1883, a barograph deflection of about 7 millibars (0.7 kPa) was recorded 150 km from Krakatau (Strachey, 1888). Air shocks can apparently couple to the ground strongly enough to cause damage to buildings at 100 km (Simkin and Howard, 1970).
Few quantitative observational data are available upon which to construct a model relating shock strength (overpressure and rate of compression), distance, and energy release. Considering the uncertainties, the simple theory of self-similar motion is adequate for a first approximation. This theory (Thompson, 1972; Landau and Lifshitz, 1959; Zeldovich and Razier, 1966) was developed for the motion of the atmosphere in response to nuclear blasts. The source pressures in volcanic explosions, however, are much lower than those in nuclear blasts.
Assume (1) the atmosphere is uniform in structure and (2) is at rest at the time of the eruption; (3) at time t = 0, a large energy, E, is released at the volcano; (4) the dimensions of the region over which E is released are small compared to the distances of interest here; (5) the resulting motion of the atmosphere is spherically symmetric. For shock pressures of 6 bars or more, the shock pressure will decay as 1/R3:
PS = E/R3,
(1 bar = 1 X 106 dyne cm-2)
(1 erg = 1 dyne-cm)
(1 cm = 1 X 10-5 km)
where PS is the pressure immediately behind the shock front, R is the radial distance from the source, and E is the energy of the shock wave. Volcanologists currently consider about 500 bars as an upper limit to the initial value of PS (Self and others, 1979; Kieffer, 1981). The value here assigned to E is 5 x 1024 ergs, the energy thought to have been dissipated in the atmosphere by the 1883 eruption of Krakatau (Press and Harkrider, 1966), perhaps the greatest explosion ever recorded. This eruption was of the same order of magnitude as the climactic eruption of Mount Mazama 6850 yr ago (Friedman and others, 1981; Simkin and others, 1981), taken here as the largest credible future eruption of a Cascade volcano. Using the Krakatau energy value,
PS = 5 x 1024/R3.This equation holds approximately between the source and the radial distance at which Ps decays to about 6 bars, about 9 km. Beyond this distance, strong shock theory is inappropriate, and the pressure decays approximately as 1/R:
PS/PSS = RSS/R,
where PSS is the pressure (6 bars) at the lower limit of the strong shock regime, and RSS is the distance at the limit of the strong shock approximation. For the case considered here (E = 5 x 1024 ergs, RSS = 9 km), PS is about 1 bar at 50 km, the radius of the proximal-hazard zone (Chapter 5), and is about 0.4 bars at 150 km (approximately 50 times the observed value at this range at Krakatau). The wave would be calculated to decay to 0.1 bar, the threshold for damage, at about 540 km. These overpressure estimates are maximum values for at least two reasons. First the energy we have used is probably an upper limit on the energy of Krakatau. Secondly, the density structure of the atmosphere, neglected in this formulation, tends to reduce the pressure by a factor of 2 to 3 in the region of a few tens to hundreds of kilometers.
A more empirical approach is to take the observed damage threshold distance, assume an overpressure of 0.1 bar, then calculate the overpressure at lesser distances. The 1883 eruption of Krakatau caused windows to break at 150 km from source (Verbeek, 1885, in Simkin and Fiske, 1983, p. 202). Accordingly, RSS = 2.5 km. Then, E = 9 x 1022 ergs, and PS = 0.3 bar at 50 km. Based on the preceding analyses, a reasonable worst-case overpressure range for large eruptions of Cascade volcanoes at 50 km, the margin of the proximal-hazard zone, is about 0.3-1.0 bars.
One of the only detailed calculations done of atmospheric response to an observed volcanic eruption event is that of Bannister (1984), in which he calculated the response of the atmosphere within 1000 km to the accelerations of the May 18 blast at Mount St. Helens. The calculated overpressures were in good agreement with barograph records observed in the range 50 to 400 km. The peak positive overpressure at 10 km was 1600 Pa (0.16 bar) and at 50 km was nearly 400 Pa (0.04 bars). These pressures are directly dependent on the initial velocity and time history of the ejecta. Since ejecta velocities substantially larger than the 147 m/s used by Bannister for the Mount St. Helens ejecta are plausible, higher overpressures for larger events are conceivable. These cannot be predicted without numerical modelling, but we believe that overpressures that could exceed the Mount St. Helens example by factors of 2, 3, or 5 are plausible. This reasoning supports the above estimates of worst case overpressures of several tenths of a bar. These estimates, however, are too poorly supported to be used as design criteria. If eruption-induced overpressures are to be considered in design, we recommend that additional research be undertaken to develop better-constrained overpressure estimates.
Return to CVO Home page