Geology Reports⌕ Search

Geology topics

J.A. Power

Publications and source records attributed to J.A. Power.

At least 19 recordsLinked to original sources
USGS2014

Surface deformation associated with the March 1996 earthquake swarm at Akutan Island, Alaska, revealed by C-band ERS and L-band JERS radar interferometry

In March 1996, an intense earthquake swarm beneath Akutan Island, Alaska, was accompanied by extensive ground cracking but no eruption of Akutan volcano. Radar interferograms produced from L-band JERS-1 and C-band ERS-1/2 images show uplift associated with the swarm by as much as 60 cm on the western part of the island. The JERS-1 interferogram has greater coherence, especially in areas with loose surface material or thick vegetation. It also shows subsidence of similar magnitude on the eastern part of the island and displacements along faults reactivated during the swarm. The axis of uplift and subsidence strikes about N70°W, which is roughly parallel to a zone of fresh cracks on the northwest flank of the volcano, to normal faults that cut the island and to the inferred maximum compressive stress direction. A common feature of models that fit the deformation is the emplacement of a shallow dike along this trend beneath the northwest flank of the volcano. Both before and after the swarm, the northwest flank was uplifted 5–20 mm/year relative to the southwest flank, probably by magma intrusion. The zone of fresh cracks subsided about 20 mm during 1996–1997 and at lesser rates thereafter, possibly because of cooling and degassing of the intrusion.

Alaska↗
USGS2006

Ground deformation associated with the precursory unrest and early phases of the January 2006 eruption of Augustine volcano, Alaska

On January 11, 2006 Augustine Volcano erupted after nearly 20 years of quiescence. Global Positioning System (GPS) instrumentation at Augustine, consisting of six continuously recording, telemetered receivers, measured clear precursory deformation consistent with a source of inflation or pressurization beneath the volcano's summit at a depth of around sea level. Deformation began in early summer 2005, and was preceded by a subtle, but distinct, increase in seismicity, which began in May 2005. After remaining more or less constant, deformation rates accelerated on at least three stations beginning in late November 2005. After this date, GPS data suggest the upward propagation of a small dike into the edifice, which, based on the style of deformation and high levels of gas emission, appears to have ascended to shallow levels by mid-December 2005, about four weeks before the eruption began.

Alaska↗
USGS2004

Observations of deep long-period (DLP) seismic events beneath Aleutian arc volcanoes; 1989-2002

Between October 12, 1989 and December 31, 2002, the Alaska Volcano Observatory (AVO) located 162 deep long-period (DLP) events beneath 11 volcanic centers in the Aleutian arc. These events generally occur at mid- to lower-crustal depths (10-45 km) and are characterized by emergent phases, extended codas, and a strong spectral peak between 1.0 and 3.0 Hz. Observed wave velocities and particle motions indicate that the dominant phases are P- and S-waves. DLP epicenters often extend over broad areas (5-20 km) surrounding the active volcanoes. The average reduced displacement of Aleutian DLPs is 26.5 cm 2 and the largest event has a reduced displacement of 589 cm 2 (or M L 2.5). Aleutian DLP events occur both as solitary events and as sequences of events with several occurring over a period of 1-30 min. Within the sequences, individual DLPs are often separated by lower-amplitude volcanic tremor with a similar spectral character. Occasionally, volcano-tectonic earthquakes that locate at similar depths are contained within the DLP sequences. At most, Aleutian volcanoes DLPs appear to loosely surround the main volcanic vent and occur as part of background seismicity. A likely explanation is that they reflect a relatively steady-state process of magma ascent over broad areas in the lower and middle portions of the crust. At Mount Spurr, DLP seismicity was initiated by the 1992 eruptions and then slowly declined until 1997. At Shishaldin Volcano, a short-lived increase in DLP seismicity occurred about 10 months prior to the April 19, 1999 eruption. These observations suggest a link between eruptive activity and magma flux in the mid- to lower-crust and uppermost mantle.

Alaska↗
USGS2004

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. ?? 2003 Elsevier B.V. All rights reserved.

Journal of Volcanology and Geothermal Research↗
USGS2004

Spatial variations in the frequency-magnitude distribution of earthquakes at Mount Pinatubo volcano

The frequency-magnitude distribution of earthquakes measured by the b-value is mapped in two and three dimensions at Mount Pinatubo, Philippines, to a depth of 14 km below the summit. We analyzed 1406 well-located earthquakes with magnitudes MD ???0.73, recorded from late June through August 1991, using the maximum likelihood method. We found that b-values are higher than normal (b = 1.0) and range between b = 1.0 and b = 1.8. The computed b-values are lower in the areas adjacent to and west-southwest of the vent, whereas two prominent regions of anomalously high b-values (b ??? 1.7) are resolved, one located 2 km northeast of the vent between 0 and 4 km depth and a second located 5 km southeast of the vent below 8 km depth. The statistical differences between selected regions of low and high b-values are established at the 99% confidence level. The high b-value anomalies are spatially well correlated with low-velocity anomalies derived from earlier P-wave travel-time tomography studies. Our dataset was not suitable for analyzing changes in b-values as a function of time. We infer that the high b-value anomalies around Mount Pinatubo are regions of increased crack density, and/or high pore pressure, related to the presence of nearby magma bodies.

Bulletin of the Seismological Society of America↗
USGS2004

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 ??1/ ??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.

Bulletin of the Seismological Society of America↗
USGS2004

Earthquake triggering at Alaskan volcanoes following the 3 November 2002 Denali fault earthquake

The 3 November 2002 Mw 7.9 Denali fault earthquake provided an excellent opportunity to investigate triggered earthquakes at Alaskan volcanoes. The Alaska Volcano Observatory operates short-period seismic networks on 24 historically active volcanoes in Alaska, 247-2159 km distant from the mainshock epicenter. We searched for evidence of triggered seismicity by examining the unfiltered waveforms for all stations in each volcano network for ∼ 1 hr after the Mw 7.9 arrival time at each network and for significant increases in located earthquakes in the hours after the mainshock. We found compelling evidence for triggering only at the Katmai volcanic cluster (KVC, 720-755 km southwest of the epicenter), where small earthquakes with distinct P and 5 arrivals appeared within the mainshock coda at one station and a small increase in located earthquakes occurred for several hours after the mainshock. Peak dynamic stresses of ∼0.1 MPa at Augustine Volcano (560 km southwest of the epicenter) are significantly lower than those recorded in Yellowstone and Utah (>3000 km southeast of the epicenter), suggesting that strong directivity effects were at least partly responsible for the lack of triggering at Alaskan volcanoes. We describe other incidents of earthquake-induced triggering in the KVC, and outline a qualitative magnitude/distance-dependent triggering threshold. We argue that triggering results from the perturbation of magmatic-hydrothermal systems in the KVC and suggest that the comparative lack of triggering at other Alaskan volcanoes could be a result of differences in the nature of magmatic-hydrothermal systems.

Alaska↗
USGS2002

A tectonic earthquake sequence preceding the April-May 1999 eruption of Shishaldin Volcano, Alaska

On 4 March 1999, a shallow ML 5.2 earthquake occurred beneath Unimak Island in the Aleutian Arc. This earthquake was located 10-15 km west of Shishaldin Volcano, a large, frequently active basaltic-andesite stratovolcano. A Strombolian eruption began at Shishaldin roughly 1 month after the mainshock, culminating in a large explosive eruption on 19 April. We address the question of whether or not the eruption caused the mainshock by computing the Coulomb stress change caused by an inflating dike on fault planes oriented parallel to the mainshock focal mechanism. We found Coulomb stress increases of ???0.1 MPa in the region of the mainshock, suggesting that magma intrusion prior to the eruption could have caused the mainshock. Satellite and seismic data indicate that magma was moving upwards beneath Shishaldin well before the mainshock. indicating that, in an overall sense, the mainshock cannot be said to have caused the eruption. However, observations of changes at the volcano following the mainshock and several large aftershocks suggest that the earthquakes may, in turn, have influenced the course of the eruption.

Bulletin of Volcanology↗
USGS2002

Preeruptive inflation and surface interferometric coherence characteristics revealed by satellite radar interferometry at Makushin Volcano, Alaska: 1993-2000

Pilot reports in January 1995 and geologic field observations from the summer of 1996 indicate that a relatively small explosive eruption of Makushin, one of the more frequently active volcanoes in the Aleutian arc of Alaska, occurred on 30 January 1995. Several independent radar interferograms that each span the time period from October 1993 to September 1995 show evidence of ???7 cm of uplift centered on the volcano's east flank, which we interpret as preeruptive inflation of a ???7-km-deep magma source (??V = 0.022 km3). Subsequent interferograms for 1995-2000, a period that included no reported eruptive activity, show no evidence of additional ground deformation. Interferometric coherence at C band is found to persist for 3 years or more on lava flow and other rocky surfaces covered with short grass and sparsely distributed tall grass and for at least 1 year on most pyroclastic deposits. On lava flow and rocky surfaces with dense tall grass and on alluvium, coherence lasts for a few months. Snow and ice surfaces lose coherence within a few days. This extended timeframe of coherence over a variety of surface materials makes C band radar interferometry an effective tool for studying volcano deformation in Alaska and other similar high-latitude regions.

Journal of Geophysical Research B: Solid Earth↗
USGS2001

Seismic response of the katmai volcanoes to the 6 December 1999 magnitude 7.0 Karluk Lake earthquake, Alaska

A sudden increase in earthquake activity was observed beneath volcanoes in the Katmai area on the Alaska Peninsula immediately following the 6 December 1999 magnitude (Mw) 7.0 Karluk Lake earthquake beneath southern Kodiak Island, Alaska. The observed increase in earthquake activity consisted of small (ML < 1.3), shallow (Z < 5.0 km) events. These earthquakes were located beneath Mount Martin, Mount Mageik, Trident Volcano, and the Katmai caldera and began within the coda of the Karluk Lake mainshock. All of these earthquakes occurred in areas and magnitude ranges that are typical for the background seismicity observed in the Katmai area. Seismicity rates returned to background levels 8 to 13 hours after the Karluk Lake mainshock. The close temporal relationship with the Karluk Lake mainshock, the onset of activity within the mainshock coda, and the simultaneous increase beneath four separate volcanic centers all suggest these earthquakes were remotely triggered. Modeling of the Coulomb stress changes from the mainshock for optimally oriented faults suggests negligible change in static stress beneath the Katmai volcanoes. This result favors models that involve dynamic stresses as the mechanism for triggered seismicity at Katmai.

Bulletin of the Seismological Society of America↗
USGS2000

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 volcano's 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. Copyright 2000 by the American Geophysical Union.

Journal of Geophysical Research B: Solid Earth↗
USGS1998

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. George's 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 &lsquo;hybrid&rsquo; 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.

Geophysical Research Letters↗
USGS1998

Seismic image of the Mount Spurr magmatic system

The three-dimenisonal P-wave velocity structure of Moutn Spurr is determined to depths of 10km by tomographic inversion of 3754 first-arriving P-wave times from local earthquakes recorded by a permanent network of 11 seismographs. Results show a prominent low-velocity zone extending from the surface to 3-4km below sea level beneath the southeastern flank of Crater Peak, spatially coincident with a geothermal system. P-wave velocities in this low-velocity zone are approximately 20% slower than those in the shallow crystalline basement rocks. Beneath Crater Peak an approximately 3km-wide zone of relative low velocities correlates with a near-vertical band of seismicity, suggestive of a magmatic conduit. No large low-velocity zone indicative of a magma chamber occurs within the upper 10km of the crust. These observations are consistent with petrologic and geochemical studies suggesting that Crater Peak magmas originate in the lower crust or upper mantle and have s short residence time in the shallow crust. Earthquakes relocaetd using the three-dimensional velocity structure correlate well with surface geology and other geophysical observations; thus, they provide additional constraints on the kinematics of the Mount Spurr magmatic system.

Bulletin of Volcanology↗
USGS1998

Spatial variations in the frequency-magnitude distribution of earthquakes at Soufriere Hills Volcano, Montserrat, West Indies

The frequency-magnitude distribution of earthquakes measured by the b -value is determined as a function of space beneath Soufriere Hills Volcano, Montserrat, from data recorded between August 1, 1995 and March 31, 1996. A volume of anomalously high b -values ( b > 3.0) with a 1.5 km radius is imaged at depths of 0 and 1.5 km beneath English's Crater and Chance's Peak. This high b -value anomaly extends southwest to Gage's Soufriere. At depths greater than 2.5 km volumes of comparatively low b -values ( b ∼1) are found beneath St. George's Hill, Windy Hill, and below 2.5 km depth and to the south of English's Crater. We speculate the depth of high b -value anomalies under volcanoes may be a function of silica content, modified by some additional factors, with the most siliceous having these volumes that are highly fractured or contain high pore pressure at the shallowest depths.

Geophysical Research Letters↗
USGS1994

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

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.

Alaska↗
USGS1994

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

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. ?? 1994.

Journal of Volcanology and Geothermal Research↗
USGS1994

Statistical forecasting of repetitious dome failures during the waning eruption of Redoubt Volcano, Alaska, February-April 1990

The waning phase of the 1989-1990 eruption of Redoubt Volcano in the Cook Inlet region of south-central Alaska comprised a quasi-regular pattern of repetitious dome growth and destruction that lasted from February 15 to late April 1990. The dome failures produced ash plumes hazardous to airline traffic. In response to this hazard, the Alaska Volcano Observatory sought to forecast these ash-producing events using two approaches. One approach built on early successes in issuing warnings before major eruptions on December 14, 1989 and January 2, 1990. These warnings were based largely on changes in seismic activity related to the occurrence of precursory swarms of long-period seismic events. The search for precursory swarms of long-period seismicity was continued through the waning phase of the eruption and led to warnings before tephra eruptions on March 23 and April 6. The observed regularity of dome failures after February 15 suggested that a statistical forecasting method based on a constant-rate failure model might also be successful. The first statistical forecast was issued on March 16 after seven events had occurred, at an average interval of 4.5 days. At this time, the interval between dome failures abruptly lengthened. Accordingly, the forecast was unsuccessful and further forecasting was suspended until the regularity of subsequent failures could be confirmed. Statistical forecasting resumed on April 12, after four dome failure episodes separated by an average of 7.8 days. One dome failure (April 15) was successfully forecast using a 70% confidence window, and a second event (April 21) was narrowly missed before the end of the activity. The cessation of dome failures after April 21 resulted in a concluding false alarm. Although forecasting success during the eruption was limited, retrospective analysis shows that early and consistent application of the statistical method using a constant-rate failure model and a 90% confidence window could have yielded five successful forecasts and two false alarms; no events would have been missed. On closer examination, the intervals between successive dome failures are not uniform but tend to increase with time. This increase attests to the continuous, slowly decreasing supply of magma to the surface vent during the waning phase of the eruption. The domes formed in a precarious position in a breach in the summit crater rim where they were susceptible to gravitational collapse. The instability of the February 15-April 21 domes relative to the earlier domes is attributed to reaming the lip of the vent by a laterally directed explosion during the major dome-destroying eruption of February 15, a process which would leave a less secure foundation for subsequent domes. ?? 1994.

Journal of Volcanology and Geothermal Research↗