Christopher D. Stephens

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.

Comments on Baseline Correction of Digital Strong-Motion Data: Examples from the 1999 Hector Mine, California, Earthquake

Residual displacements for large earthquakes can sometimes be determined from recordings on modern digital instruments, but baseline offsets of unknown origin make it difficult in many cases to do so. To recover the residual displacement, we suggest tailoring a correction scheme by studying the character of the velocity obtained by integration of zeroth-order-corrected acceleration and then seeing if the residual displacements are stable when the various parameters in the particular correction scheme are varied. For many seismological and engineering purposes, however, the residual displacements are of lesser importance than ground motions at periods less than about 20 sec. These ground motions are often recoverable with simple baseline correction and low-cut filtering. In this largely empirical study, we illustrate the consequences of various correction schemes, drawing primarily from digital recordings of the 1999 Hector Mine, California, earthquake. We show that with simple processing the displacement waveforms for this event are very similar for stations separated by as much as 20 km. We also show that a strong pulse on the transverse component was radiated from the Hector Mine earthquake and propagated with little distortion to distances exceeding 170 km; this pulse leads to large response spectral amplitudes around 10 sec.

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.

Evolution of the December 14, 1989 precursory long-period event swarm at Redoubt Volcano, Alaska

Abstract The intermittency pattern and evolution in waveforms of long-period (LP) seismic events during the intense, 23-h swarm that preceded the December 14, 1989 eruption of Redoubt volcano are investigated. Utilizing cross correlation to exploit the high degree of similarity among waveforms, a substantially more complete event catalog is generated than was available from near realtime detection based on short-term/long-term amplitude ratios, which was saturated by the high rate of activity. The temporal magnitude distribution of the predominant LP events is found to have an unusual banded structure in which the average magnitude of each band slowly increases and then decreases through time. A bifurcation that appears in the uppermost band shortly after the peak in magnitudes is characterized by a quasi-periodicity in intermittency and magnitude that is reminiscent of one of the classic routes to chaotic behavior in some non-linear systems. The waveforms of the predominant events evolve slowly but unsteadily through time. These gradual changes appear to result from variations in the relative amplitudes of spectral peaks that remain stable in frequency, which suggests that they are due to differential excitation of a single, resonant source. Two other previously unrecognized, repetitive waveforms are also identified, but the signals from these secondary events are not clearly recorded at distances beyond the closest station. Similarities among the spectra of the predominant and secondary events suggest that the signals from these events also could represent different modes of exciting the same source. Significant changes in the rates and the sizes of the largest of these secondary events appear to coincide with the peak in the size distribution of the predominant LPs. At least some of the non-repetitive LP waveforms in the swarm appear to be the result of the superposition of signals from the rapid repetition of predominant LP source, thus placing a constraint on the repeat time of the triggering mechanism for this source. A lone hybrid event, which has a waveform character intermediate between the predominant LP events and high-frequency volcano-tectonic events, was also identified in the swarm; the occurrence of this event provides important evidence that the low-frequency character of the LP events is a source rather than a path or site effect.

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.

Wrangell Benioff zone, southern Alaska

The first unequivocal evidence for the existence of a Benioff zone that may be related to the Quaternary Wrangell volcanoes comes from a set of 86 well-located hypocenters for earthquakes smaller than about magnitude 4 in the region between lat 61 and 62.5°N and between long 142 and 145.5°W. About half of the earthquakes occur at depths of 25 km or less. Below 40 km a clearly defined north-northeast–dipping zone of seismicity, here termed the “Wrangell Benioff zone,” extends to a depth of about 85 km and continues for about 115 km along strike, subparallel to the volcanic trend. The western end of the zone may be offset from the northern end of the much more active Aleutian Benioff zone. Where the Benioff zone shoals to 30 to 40 km, it becomes nearly horizontal and cannot be clearly distinguished from upper-plate seismicity. It is uncertain whether the subducted plate segment that contains the Wrangell Benioff zone is structurally part of the Pacific plate or the Yakutat block.

Recorded Motions of the 6 April 2009 Mw 6.3 L'Aquila, Italy, Earthquake and Implications for Building Structural Damage: Overview

The normal-faulting earthquake of 6 April 2009 in the Abruzzo Region of central Italy caused heavy losses of life and substantial damage to centuries-old buildings of significant cultural importance and to modern reinforced-concrete-framed buildings with hollow masonry infill walls. Although structural deficiencies were significant and widespread, the study of the characteristics of strong motion data from the heavily affected area indicated that the short duration of strong shaking may have spared many more damaged buildings from collapsing. It is recognized that, with this caveat of short-duration shaking, the infill walls may have played a very important role in preventing further deterioration or collapse of many buildings. It is concluded that better new or retrofit construction practices that include reinforced-concrete shear walls may prove helpful in reducing risks in such seismic areas of Italy, other Mediterranean countries, and even in United States, where there are large inventories of deficient structures.

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

Abstract 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.

PRISM Software: Processing and Review Interface for Strong‐Motion Data

ABSTRACT A continually increasing number of high‐quality digital strong‐motion records from stations of the National Strong Motion Project (NSMP) of the U.S. Geological Survey, as well as data from regional seismic networks within the United States, calls for automated processing of strong‐motion records with human review limited to selected significant or flagged records. The NSMP has developed the Processing and Review Interface for Strong Motion data (PRISM) software to meet this need. In combination with the Advanced National Seismic System Quake Monitoring System (AQMS), PRISM automates the processing of strong‐motion records. When used without AQMS, PRISM provides batch‐processing capabilities. The PRISM software is platform independent (coded in Java), open source, and does not depend on any closed‐source or proprietary software. The software consists of two major components: a record processing engine composed of modules for each processing step, and a review tool, which is a graphical user interface for manual review, edit, and processing. To facilitate use by non‐NSMP earthquake engineers and scientists, PRISM (both its processing engine and review tool) is easy to install and run as a stand‐alone system on common operating systems such as Linux, OS X, and Windows. PRISM was designed to be flexible and extensible to accommodate implementation of new processing techniques. All the computing features have been thoroughly tested.