John A. Power

Earthquake classification, location, and error analysis in a volcanic environment: implications for the magmatic system of the 1989–1990 eruptions at redoubt volcano, Alaska

Abstract Determination of the precise locations of seismic events associated with the 1989–1990 eruptions of Redoubt Volcano posed a number of problems, including poorly known crustal velocities, a sparse station distribution, and an abundance of events with emergent phase onsets. In addition, the high relief of the volcano could not be incorporated into the hypoellipse earthquake location algorithm. This algorithm was modified to allow hypocenters to be located above the elevation of the seismic stations. The velocity model was calibrated on the basis of a posteruptive seismic survey, in which four chemical explosions were recorded by eight stations of the permanent network supplemented with 20 temporary seismographs deployed on and around the volcanic edifice. The model consists of a stack of homogeneous horizontal layers; setting the top of the model at the summit allows events to be located anywhere within the volcanic edifice. Detailed analysis of hypocentral errors shows that the long-period (LP) events constituting the vigorous 23-hour swarm that preceded the initial eruption on December 14 could have originated from a point 1.4 km below the crater floor. A similar analysis of LP events in the swarm preceding the major eruption on January 2 shows they also could have originated from a point, the location of which is shifted 0.8 km northwest and 0.7 km deeper than the source of the initial swarm. We suggest this shift in LP activity reflects a northward jump in the pathway for magmatic gases caused by the sealing of the initial pathway by magma extrusion during the last half of December. Volcano-tectonic (VT) earthquakes did not occur until after the initial 23-hour-long swarm. They began slowly just below the LP source and their rate of occurrence increased after the eruption of 01:52 AST on December 15, when they shifted to depths of 6 to 10 km. After January 2 the VT activity migrated gradually northward; this migration suggests northward propagating withdrawal of magma from a plexus of dikes and/or sills located in the 6 to 10 km depth range. Precise relocations of selected events prior to January 2 clearly resolve a narrow, steeply dipping, pencil-shaped concentration of activity in the depth range of 1–7 km, which illuminates the conduit along which magma was transported to the surface. A third event type, named hybrid, which blends the characteristics of both VT and LP events, originates just below the LP source, and may reflect brittle failure along a zone intersecting a fluid-filled crack. The distribution of hybrid events is elongated 0.2–0.4 km in an east-west direction. This distribution may offer constraints on the orientation and size of the fluid-filled crack inferred to be the source of the LP events.

Precursory swarms of long-period events at Redoubt Volcano (1989-1990), Alaska: Their origin and use as a forecasting tool

Abstract During the eruption of Redoubt Volcano from December 1989 through April 1990, the Alaska Volcano Observatory issued advance warnings of several tephra eruptions based on changes in seismic activity related to the occurrence of precursory swarms of long-period (LP) seismic events (dominant period of about 0.5 s). The initial eruption on December 14 occurred after 23 years of quiescence and was heralded by a 23-hour swarm of LP events that ended abruptly with the eruption. After a series of vent-clearing explosions over the next few days, dome growth began on December 21. Another swarm, with LP events similar to those of the first, began on the 26th and ended in a major tephra eruption on January 2. Eruptions continued over the next two weeks and then ceased until February 15, when a large eruption initiated a long phase of repetitive dome-building and dome-destroying episodes that continued into April. Warnings were issued before the major events on December 14 and January 2, but as the eruptive sequence continued after January 2, the energy of the swarms decreased and forecasting became more difficult. A significant but less intense swarm preceded the February 15 eruption, which was not forecast. This eruption destroyed the only seismograph on the volcanic edifice and stymied forecasting until March 4, when the first of three new stations was installed within 3 km of the active vent. From March 4 to the end of the sequence on April 21, there were eight eruptions, six of which were preceded by detectable swarms of LP events. Although weak, these swarms provided the basis for warnings issued before the eruptions on March 23 and April 6. The initial swarm on December 13 had the following features: (1) short duration (23 hours); (2) a rapidly accelerating rate of seismic energy release over the first 18 hours of the swarm, followed by a decline of activity during the 5 hours preceding the eruption; (3) a magnitude range from −0.4 to 1.6; (4) nearly identical LP signatures with a dominant period near 0.5 s; (5) dilatational first motions everywhere; and (6) a stationary source location at a depth of 1.4 km beneath the crater. This occurrence of long-period events suggests a model involving the interaction of magma with groundwater in which magmatic gases, steam and water drive a fixed conduit at a stationary point throughout the swarm. The initiation of that sequence of events is analogous to the failure of a pressure-relief valve connecting a lower, supercharged magma-dominated reservoir to a shallow hydrothermal system. A three-dimensional model of a vibrating fluid-filled crack recently developed by Chouet is found to be compatible with the seismic data and yields the following parameters for the LP source: crack length, 280–380 m; crack width, 140–190 m; crack thickness, 0.05–0.20 m; crack stiffness, 100–200; sound speed of fluid, 0.8–1.3 km/s; compressional-wave speed of rock, 5.1 km/s; density ratio of fluid to rock, ≈0.4; and ratio of bulk modulus of fluid to rigidity of rock, 0.03–0.07. The fluid-filled crack is excited intermittently by an impulsive pressure drop that varies in magnitude within the range of 0.4 to 40 bar. Such disturbance appears to be consistent with a triggering mechanism associated with choked flow conditions in the crack.

Soufrière Hills Eruption, Montserrat, 1995–1997: Volcanic earthquake locations and fault plane solutions

A total of 9242 seismic events, recorded since the start of the eruption on Montserrat in July 1995, have been uniformly relocated with station travel-time corrections. Early seismicity was generally diffuse under southern Montserrat, and mostly restricted to depths less than 7 km. However, a NE-SW alignment of epicentres beneath the NE flank of the volcano emerged in one swarm of volcano-tectonic earthquakes (VTs) and later nests of VT hypocentres developed beneath the volcano and at a separated location, under St. Georges Hill. The overall spatial distribution of hypocentres suggests a minimum depth of about 5 km for any substantial magma body. Activity associated with the opening of a conduit to the surface became increasingly shallow, with foci concentrated below the crater and, after dome building started in Fall 1995, VTs diminished and repetitive swarms of ‘hybrid’ seismic events became predominant. By late-1996, as magma effusion rates escalated, most seismic events were originating within a volume about 2 km diameter which extended up to the surface from only about 3 km depth - the diminution of shear failure earthquakes suggests the pathway for magma discharge had become effectively unconstricted. Individual and composite fault plane solutions have been determined for a few larger earthquakes. We postulate that localised extensional stress conditions near the linear VT activity, due to interaction with stresses in the overriding lithospheric plate, may encourage normal fault growth and promote sector weaknesses in the volcano.

Seismic evolution of the 1989-1990 eruption sequence of Redoubt Volcano, Alaska

Abstract Redoubt Volcano in south-central Alaska erupted between December 1989 and June 1990 in a sequence of events characterized by large tephra eruptions, pyroclastic flows, lahars and debris flows, and episodes of dome growth. The eruption was monitored by a network of five to nine seismic stations located 1 to 22 km from the summit crater. Notable features of the eruption seismicity include : (1) small long-period events beginning in September 1989 which increased slowly in number during November and early December; (2) an intense swarm of long-period events which preceded the initial eruptions on December 14 by 23 hours; (3) shallow swarms (0 to 3 km) of volcano-tectonic events following each eruption on December 15; (4) a persistent cluster of deep (6 to 10 km) volcano-tectonic earthquakes initiated by the eruptions on December 15, which continued throughout and beyond the eruption; (5) an intense swarm of long-period events which preceded the eruptions on January 2; and (6) nine additional intervals of increased long-period seismicity each of which preceded a tephra eruption. Hypocenters of volcano-tectonic earthquakes suggest the presence of a magma source region at 6–10 km depth. Earthquakes at these depths were initiated by the tephra eruptions on December 15 and likely represent the readjustment of stresses in the country rock associated with the removal of magma from these depths. The locations and time-history of these earthquakes coupled with the eruptive behavior of the volcano suggest this region was the source of most of the erupted material during the 1989–1990 eruption. This source region appears to be connected to the surface by a narrow pipe-like conduit as inferred from the hypocenters of volcano-tectonic earthquakes. Concentrations of shallow volcano-tectonic earthquakes followed each of the tephra eruptions on December 15; these shocks may represent stress readjustment in the wall rock related to the removal of magma and volatiles at these depths. This shallow zone was the source area of the majority of long-period seismicity through the remainder of the eruption. The long-period seismicity likely reflects the pressurization of the shallow portions of the magmatic system.

Ground deformation associated with the March 1996 earthquake swarm at Akutan volcano, Alaska, revealed by satellite radar interferometry

In March 1996 an intense swarm of volcano-tectonic earthquakes (∼3000 felt by local residents, Mmax = 5.1, cumulative moment of 2.7×1018 N m) beneath Akutan Island in the Aleutian volcanic arc, Alaska, produced extensive ground cracks but no eruption of Akutan volcano. Synthetic aperture radar interferograms that span the time of the swarm reveal complex island-wide deformation: the western part of the island including Akutan volcano moved upward, while the eastern part moved downward. The axis of the deformation approximately aligns with new ground cracks on the western part of the island and with Holocene normal faults that were reactivated during the swarm on the eastern part of the island. The axis is also roughly parallel to the direction of greatest compressional stress in the region. No ground movements greater than 2.83 cm were observed outside the volcanos summit caldera for periods of 4 years before or 2 years after the swarm. We modeled the deformation primarily as the emplacement of a shallow, east–west trending, north dipping dike plus inflation of a deep, Mogi-type magma body beneath the volcano. The pattern of subsidence on the eastern part of the island is poorly constrained. It might have been produced by extensional tectonic strain that both reactivated preexisting faults on the eastern part of the island and facilitated magma movement beneath the western part. Alternatively, magma intrusion beneath the volcano might have been the cause of extension and subsidence in the eastern part of the island. We attribute localized subsidence in an area of active fumaroles within the Akutan caldera, by as much as 10 cm during 1992–1993 and 1996–1998, to fluid withdrawal or depressurization of the shallow hydrothermal system.

Temporal and Spatial Variation of Local Stress Fields before and after the 1992 Eruptions of Crater Peak Vent, Mount Spurr Volcano, Alaska

We searched for changes in local stress-field orientation at Mount Spurr volcano, Alaska, between August 1991 and December 2001. This study focuses on the stress-field orientation beneath Crater Peak vent, the site of three eruptions in 1992, and beneath the summit of Mount Spurr. Local stress tensors were calculated by inverting subsets of 140 fault-plane solutions for earthquakes beneath Crater Peak and 96 fault-plane solutions for earthquakes beneath Mount Spurr. We also calculated an upper-crustal regional stress tensor by inverting fault-plane solutions for 66 intraplate earthquakes located near Mount Spurr during 1991–2001. Prior to the 1992 eruptions, and for 11 months beginning with a posteruption seismic swarm, the axis of maximum compressive stress beneath Crater Peak was subhorizontal and oriented N67–76° E, approximately perpendicular to the regional axis of maximum compressive stress (N43° W). The strong temporal correlation between this horizontal stress-field rotation (change in position of the σ 1  / σ 3 axes relative to regional stress) and magmatic activity indicates that the rotation was related to magmatic activity, and we suggest that the Crater Peak stress-field rotation resulted from pressurization of a network of dikes. During the entire study period, the stress field beneath the summit of Mount Spurr also differed from the regional stress tensor and was characterized by a vertical axis of maximum compressive stress. We suggest that slip beneath Mount Spurr’s summit occurs primarily on a major normal fault in response to a combination of gravitational loading, hydrothermal circulation, and magmatic processes beneath Crater Peak. Online material : Regional and local fault-plane solutions.

Observations of hybrid seismic events at Soufriere Hills Volcano, Montserrat: July 1995 to September 1996

Swarms of small repetitive events with similar waveforms and magnitudes are often observed during the emplacement of lava domes. Over 300,000 such events were recorded in association with the emplacement of the lava dome at Soufriere Hills Volcano, Montserrat, from August 1995 through August 1996. These events originated < 2–3 km deep. They exhibited energy ranging over ˜ 1.5–4.5 Hz and were broader band than typical long-period events. We term the events “hybrid” between long-period and volcano-tectonic. The events were more impulsive and broader band prior to, compared with during and after, periods of inferred increased magma flux rate. Individual swarms contained up to 10,000 events often exhibiting very similar magnitudes and waveforms throughout the swarm. Swarms lasted hours to weeks, during which inter-event intervals generally increased, then decreased, often several times. Long-duration swarms began about every two months starting in late September 1995. We speculate that the events were produced as the magma column degassed into adjacent cracks.

The reawakening of Alaska's Augustine volcano

Augustine volcano, in south central Alaska, ended a 20-year period of repose on 11 January 2006 with 13 explosive eruptions in 20 days. Explosive activity shifted to a quieter effusion of lava in early February, forming a new summit lava dome and two short, blocky lava flows by late March (Figure 1). The eruption was heralded by eight months of increasing seismicity, deformation, gas emission, and small phreatic eruptions, the latter consisting of explosions of steam and debris caused by heating and expansion of groundwater due to an underlying heat source.

Evidence for dike emplacement beneath Iliamna Volcano, Alaska in 1996

Two earthquake swarms, comprising 88 and 2833 locatable events, occurred beneath Iliamna Volcano, Alaska, in May and August of 1996. Swarm earthquakes ranged in magnitude from −0.9 to 3.3. Increases in SO2 and CO2 emissions detected during the fall of 1996 were coincident with the second swarm. No other physical changes were observed in or around the volcano during this time period. No eruption occurred, and seismicity and measured gas emissions have remained at background levels since mid-1997. Earthquake hypocenters recorded during the swarms form a cluster in a previously aseismic volume of crust located to the south of Iliamna’s summit at a depth of −1 to 4 km below sea level. This cluster is elongated to the NNW–SSE, parallel to the trend of the summit and southern vents at Iliamna and to the regional axis of maximum compressive stress determined through inversion of fault-plane solutions for regional earthquakes. Fault-plane solutions calculated for 24 swarm earthquakes located at the top of the new cluster suggest a heterogeneous stress field acting during the second swarm, characterized by normal faulting and strike-slip faulting with p-axes parallel to the axis of regional maximum compressive stress. The increase in earthquake rates, the appearance of a new seismic volume, and the elevated gas emissions at Iliamna Volcano indicate that new magma intruded beneath the volcano in 1996. The elongation of the 1996–1997 earthquake cluster parallel to the direction of regional maximum compressive stress and the accelerated occurrence of both normal and strike-slip faulting in a small volume of crust at the top of the new seismic volume may be explained by the emplacement and inflation of a subvertical planar dike beneath the summit of Iliamna and its southern satellite vents.

Seismological aspects of the 1989–1990 eruptions at redoubt volcano, Alaska: the SSAM perspective

Abstract SSAM is a simple and inexpensive tool for continuous monitoring of average seismic amplitudes within selected frequency bands in near real-time on a PC-based data acquisition system. During the 1989–1990 eruption sequence at Redoubt Volcano, the potential of SSAM to aid in rapid identification of precursory Long-Period (LP) event swarms was realized, and since this time SSAM has been incorporated in routine monitoring efforts of the Alaska Volcano Observatory. In particular, an eruption that occurred on April 6 was successfully forecast primarily on the basis of recognizing the precursory LP activity on SSAM. Of twenty-two significant eruptions that occurred between December 14 and April 21, eleven had precursory swarms longer than one hour in duration that could be detected on SSAM. For individual swarms, the patterns of relative spectral amplitudes are distinct at each station and remain largely stationary through time, thus indicating that one source may have been preferentially and repeatedly activated throughout the swarm. Typically, a single spectral band dominates the signal at each seismic station: for the vigorous one-day swarm that preceded the first eruption on December 14, signals were sharply peaked in the 1.9–2.7 Hz band at the closest station, located 4 km from the vent, but were dominated by 1.3–1.9 Hz energy at three more distant stations located 7.5–22 km from the vent. The tendency for the signals from different swarms recorded at the same station to be peaked in the same frequency band suggests that all of the sources are characterized by a predominant length scale. Signals from the precursory LP swarms became weaker as the eruption sequence progressed, and swarms that occurred in March and April could only be detected at seismographs on the volcanic edifice. Onset times of precursory LP swarms prior to eruptions ranged from a few hours to about one week, but after the initial vent-clearing phase that ended December 19 these intervals tended to become progressively shorter for successive swarms. These trends in the relative onset times and intensities of successive precursory LP swarms are consistent with an overall depressurization of the magmatic system through time. In general, each of the swarms had an emergent onset, but the intensities did not always increase steadily until the eruptions. Instead, as the time of an eruption approached the intensity usually increased more rapidly before peaking and then declining prior to the eruption; for three of the swarms, two distinct peaks in intensity were apparent. The time intervals between final peaks in swarm intensity and ensuing eruptions ranged from about 2 hours to almost 2 days, but the peaks always occurred closer to the eruptions than to the swarm onsets. Both the onset of LP swarm activity and a decline in intensity prior to an eruption may represent critical points in the process of pressurization that drives the flow of fluids and gas in a sealed magmatic system. A notable exception to this pattern is the eruption of March 9 which lacked a detectable precursory LP swarm, but was followed by an unusually long period of strong LP seismicity that may have been stimulated by a depressurization of the magmatic system resulting from dome failure. On both December 14 and January 2, the spectra of early syn-eruptive signals have peaked signatures much like those of the spectra of precursory LP activity from shortly before the eruptions; these similarities may indicate that the source of precursory seismicity continued to be active during at least the early part of each eruption. In syn-eruptive signals from March and April recorded at stations on the volcanic edifice, the dominant spectral energy progressively shifts with time during the eruption to lower frequencies; at least part of the energy in these signals may have been generated by the debris flows associated with dome failures.