Ivan G. Wong

New Evidence on the State of Stress of the San Andreas Fault System

Contemporary in situ tectonic stress indicators along the San Andreas fault system in central California show northeast-directed horizontal compression that is nearly perpendicular to the strike of the fault. Such compression explains recent uplift of the Coast Ranges and the numerous active reverse faults and folds that trend nearly parallel to the San Andreas and that are otherwise unexplainable in terms of strike-slip deformation. Fault-normal crustal compression in central California is proposed to result from the extremely low shear strength of the San Andreas and the slightly convergent relative motion between the Pacific and North American plates. Preliminary in situ stress data from the Cajon Pass scientific drill hole (located 3.6 kilometers northeast of the San Andreas in southern California near San Bernardino, California) are also consistent with a weak fault, as they show no right-lateral shear stress at ∼2-kilometer depth on planes parallel to the San Andreas fault.

Ground-Motion Attenuation Relationships for Cascadia Subduction Zone Megathrust Earthquakes Based on a Stochastic Finite-Fault Model

The number of strong ground motion recordings available for regression analysis in developing empirical attenuation relationships has rapidly grown in the last 10 years. However, the dearth of strong-motion data from the Cascadia subduc- tion zone has limited this development of relationships for the Cascadia subduction zone megathrust, which can be used in the calculation of design spectra for engi- neered structures. A stochastic finite-fault ground-motion model has been used to simulate ground motions for moment magnitude (M) 8.0, 8.5, and 9.0 megathrust earthquakes along the Cascadia subduction zone for both rock-and soil-site condi- tions. The stochastic finite-fault model was validated against the 1985 M 8.0 Mi- choacan, Mexico, and the 1985 M 8.0 Valpariso, Chile, earthquakes. These two subduction zone megathrust earthquakes were recorded at several rock sites located near the fault rupture. For the Cascadia megathrust earthquakes, three different rup- ture geometries were used to model the M 8.0, 8.5, and 9.0 events. The geometries only differ in their respective fault lengths. A fault dip of 9 to the east with a rupture width of 90 km was selected to represent average properties of the Cascadia sub- duction zone geometry. A regional crustal damping and velocity model was used with the stochastic finite-fault model simulations. Ground motions were computed for 16 site locations. The parametric uncertainties associated with the variation in source, path, and site effects were included in the development of the ground motions. A functional form was fit to the ground-motion model simulations to develop region- specific attenuation relationships for the Cascadia megathrust rupture zone for both rock and soil site conditions. The total uncertainty was based on a combination of the modeling and parametric uncertainties (sigmas). These newly developed atten- uation relationships for Cascadia subduction zone megathrust earthquakes can be used in both the probabilistic and deterministic seismic-hazard studies for engineer- ing design for the Pacific Northwest.

Contemporary seismicity, faulting, and the state of stress in the Colorado Plateau

The contemporary seismicity of the Colorado Plateau based on seismic monitoring in the past 30 yr can be characterized as being of small to moderate magnitude, and contrary to earlier views, of a low to moderate rate of occurrence with earthquakes widely distributed. Concentrations of earthquakes have been observed in a few areas of the plateau. The most seismically active area of the Colorado Plateau is the eastern Wasatch Plateau-Book Cliffs, where abundant small-magnitude seismicity is induced by coal mining. The largest earthquakes observed to date, of estimated Richter magnitude (M L ) 5-6, have generally occurred in northern Arizona. Although very few earthquakes can be associated with known geologic structures or tectonic features in the Colorado Plateau, seismicity appears to be the result of the reactivation of pre-existing faults lacking surficial expression but favorably oriented to the tectonic stress field. The small to moderate size of the earthquakes and their widespread distribution are consistent with a highly faulted Precambrian basement and upper crust, and a moderate level of differential tectonic stress. Earthquakes in the plateau generally occur in the upper crust from the near-surface to a depth of 15-20 km, although events have been observed in both the lower crust and uppermost mantle in areas of low to normal heat flow. The latter suggests that temperatures are sufficiently low at these depths that brittle failure and hence earthquakes are still possible. The predominant mode of tectonic deformation within the plateau appears to be normal faulting on northwest- to north-northwest-striking faults with some localized occurrences of strike-slip faulting on north-west- or northeast-striking planes at shallow depths. The contemporary state of stress within the plateau is characterized by approximate northeast-trending extension in contrast to the previous belief that the plateau was being subjected to east-west tectonic compression. One area of the plateau, the eastern Wasatch Plateau and Book Cliffs, may still be characterized by compressive stresses; however, the nature of these stresses is not well understood.

Observations of mine seismicity in the eastern Wasatch Plateau, Utah, U.S.A.: A possible case of implosional failure

In the summer of 1984, a three-dimensional, high-resolution microearthquake network was operated in the vicinity of two coal mines beneath Gentry Mountain in the eastern Wasatch Plateau, Utah. During a six-week period, approximately 3,000 seismic events were observed of which the majority were impulsive, higher frequency (>10 Hz), short duration (<2–3 sec) events probably associated with the caving of the roof from a longwall operation. In contrast, 234 of the largest located events appeared to occur predominantlybeneath the mines to a depth of 2 to 3 km consistent with previous studies. The magnitudes of these events ranged from less thanMc0 to 1.6. In addition to the unusual depths of these latter events, an anomalous aspect displayed by the events was an apparent dilatational focal mechanism suggesting a non-double-couple, possibly implosional source. Implosional events have been observed in other studies of mine seismicity; however, the generally inadequate instrumental coverage of the focal sphere has cast some doubt on the validity of such mechanisms. Previously suggested source mechanisms for such implosional events have included tensional failure through strata collapse, and a shear-implosional displacement mechanism. Shear failure must be involved in the failure process of the Gentry Mountain implosional events as evidenced by well-defined shear waves in the observed seismograms. Simultaneous monitoring in the East Mountain coal mining area to the south by the University of Utah revealed typical shear failure events mixed with implosional events. The observed double-couple, reverse focal mechanisms at East Mountain were similar to mechanisms determined in previous studies and a composite focal mechanism determined in this study for a sequence outside the mining areas. This suggested that the shear events within the mining areas are being influenced by the regional tectonic stress field. Thus in addition to the seismic events associated with caving of the roof from the longwall operation, there appear to be at least two other types of mining-induced seismic events occurring in the eastern Wasatch Plateau, both submine in origin: (1) events characterized by apparent non-double-couple possibly implosional focal mechanisms and well-defined shear waves; and (2) shear events, which are indistinguishable from tectonic earthquakes and may be considered mining “triggered” earthquakes. The small mining-induced stress changes that occur beyond a few hundred meters from the mine workings suggest both types of seismic events are occurring on critically stressed, pre-existing zones of weakness. Topography, overburden, method of mining, and mine configuration also appear to be significant factors influencing the occurrence of the implosional submine events.

Potential Losses in a Repeat of the 1886 Charleston, South Carolina, Earthquake

A comprehensive earthquake loss assessment for the state of South Carolina using HAZUS was performed considering four different earthquake scenarios: a moment magnitude (M) 7.3 “1886 Charleston-like” earthquake, M 6.3 and M 5.3 events also from the Charleston seismic source, and an M 5.0 earthquake in Columbia. Primary objectives of this study were (1) to generate credible earthquake losses to provide a baseline for coordination, capability development, training, and strategic planning for the South Carolina Emergency Management Division, and (2) to raise public awareness of the significant earthquake risk in the state. Ground shaking, liquefaction, and earthquake-induced landsliding hazards were characterized using region-specific inputs on seismic source, path, and site effects, and ground motion numerical modeling. Default inventory data on buildings and facilities in HAZUS were either substantially enhanced or replaced. Losses were estimated using a high resolution 2-km×2-km grid rather than the census tract approach used in HAZUS. The results of the loss assessment indicate that a future repeat of the 1886 earthquake would be catastrophic, resulting in possibly 900 deaths, more than 44,000 injuries, and a total economic loss of

Statistical Analyses of Great Earthquake Recurrence along the Cascadia Subduction Zone

20 billion in South Carolina alone. Schools, hospitals, fire stations, ordinary buildings, and bridges will suffer significant damage due to the general lack of seismic design in the state. Lesser damage and losses will be sustained in the other earthquake scenarios although even the smallest event could result in significant losses.

Ground Motion and Liquefaction Simulation of the 1886 Charleston, South Carolina, Earthquake

Goldfinger et al. (2012) interpreted a 10,000 year old sequence of deep sea turbidites at the Cascadia subduction zone (CSZ) as a record of clusters of plate-boundary great earthquakes separated by gaps of many hundreds of years. We performed statistical analyses on this inferred earthquake record to test the temporal clustering model and to calculate time-dependent recurrence intervals and probabil- ities. We used a Monte Carlo simulation to determine if the turbidite recurrence in- tervals follow an exponential distribution consistent with a Poisson (memoryless) process. The latter was rejected at a statistical significance level of 0.05. We performed a cluster analysis on 20 randomly simulated catalogs of 18 events (event T2 excluded), using ages with uncertainties from the turbidite dataset. Results indicate that 13 cata- logs exhibit statistically significant clustering behavior, yielding a probability of clus- tering of 13=20 or 0.65. Most (70%) of the 20 catalogs contain two or three closed clusters (a sequence that contains the same or nearly the same number of events) and the current cluster T1-T5 appears consistently in all catalogs. Analysis of the 13 cat- alogs that manifest clustering indicates that the probability that at least one more event will occur in the current cluster is 0.82. Given that the current cluster may not be closed yet, the probabilities of an M 9 earthquake during the next 50 and 100 years were estimated to be 0.17 and 0.25, respectively. We also analyzed the sensitivity of results to including event T2, whose status as a full-length rupture event is in doubt. The inclusion of T2 did not change the probability of clustering behavior in the CSZ turbidite data, but did significantly reduce the probability that the current cluster would extend to one more event. Based on the statistical analysis, time-independent and time-dependent recurrence intervals were calculated.

Incorporating Descriptive Metadata into Seismic Source Zone Models for Seismic-Hazard Assessment: A Case Study of the Azores-West Iberian Region

As part of a comprehensive earthquake loss and vulnerability evaluation of the state of South Carolina, ground motions were simulated for a moment magnitude ( M ) 7.3 “1886 Charleston-like” earthquake using finite-fault and point-source stochastic numerical modeling. The probability for liquefaction was also predicted based on factors of safety computed from average cyclic stress and shear-wave-velocity-based cyclic resistance ratios, clay content, and saturation. Because there is considerable uncertainty regarding the 1886 source, the rupture plane of the 1886 event was modeled as both a 100-km-long, 20-km-wide fault (static stress drop of 27 bars) and a 50-km-long, 16-km-wide fault (107-bar stress drop). The source was assumed to be a north-northeast-striking, strike-slip fault coincident with the Woodstock fault. Based on comparing the computed and observed 1886 liquefaction areas and ground motions for both the low and high stress drop events, the two cases were weighted 0.8 and 0.2, respectively. To accommodate epistemic uncertainty in eastern U.S. earthquake source processes, three region-specific point-source attenuation models were also developed and used: a single-corner frequency model with both a constant stress drop and a magnitude-dependent stress drop, and the double-corner frequency model. The finite-fault and point-source models were weighted 0.8 and 0.2, respectively. To incorporate site effects into the ground-motion estimates, an extensive effort was made to characterize the thicknesses, shear-wave velocities ( V S), and dynamic material properties of unconsolidated sediments. Characteristic V S profiles were developed using the available subsurface information, which incorporated a wide range of soil and rock conditions. Amplification factors were computed for four site response categories, each of which were a function of soil thickness, input hard-rock motion, and spectral frequency. From the five weighted stochastic ground-motion models (two finite fault and three point source) and amplification factors, rock and soil ground motions were computed to produce statewide ground-motion maps for the M 7.3 scenario event. Weighting of the Charleston source and ground-motion models was implemented so that the resulting liquefaction areas matched the 1886 areas of liquefaction. Peak horizontal ground acceleration (PGA) values as high as 0.6 g -0.7 g were estimated in the vicinity of the modeled rupture. PGAs in the range of 0.3 g -0.4 g were estimated for Charleston consistent with the observed building damage and liquefaction. Significant ground shaking (PGA > 0.2 g ) extends out to distances of 50-60 km. Strong long-period (≥1.0 sec) ground motions are predicted throughout the state. The probabilities for liquefaction were highest in the epicentral region (>50%), consistent with the observed occurrences of liquefaction in 1886. Manuscript received 5 February 2003.

How big, how bad, how often: are extreme events accounted for in modern seismic hazard analyses?

In probabilistic seismic-hazard analysis (PSHA), seismic source zone (SSZ) models are widely used to account for the contribution to the hazard from earth- quakes not directly correlated with geological structures. Notwithstanding the impact of SSZ models in PSHA, the theoretical framework underlying SSZ models and the criteria used to delineate the SSZs are seldom explicitly stated and suitably docu- mented. In this paper, we propose a methodological framework to develop and docu- ment SSZ models, which includes (1) an assessment of the appropriate scale and degree of stationarity, (2) an assessment of seismicity catalog completeness-related issues, and (3) an evaluation and credibility ranking of physical criteria used to delin- eate the boundaries of the SSZs. We also emphasize the need for SSZ models to be supported by a comprehensive set of metadata documenting both the unique character- istics of each SSZ and the criteria used to delineate its boundaries. This procedure ensures that the uncertainties in the model can be properly addressed in the PSHA and that the model can be easily updated whenever new data are available. The pro- posed methodology is illustrated using the SSZ model developed for the Azores-West Iberian region in the context of the Seismic Hazard Harmonization in Europe project (project SHARE) and some of the most relevant SSZs are discussed in detail. Online Material: Tables describing characteristics and boundaries of the seismic source zones.

THE PEER-LIFELINES VALIDATION OF SOFTWARE USED IN PROBABILISTIC SEISMIC HAZARD ANALYSIS

The occurrence of several recent “extreme” earthquakes with their significant loss of life and the apparent failure to have been prepared for such disasters has raised the question of whether such events are accounted for in modern seismic hazard analyses. In light of the great 2011 Tohoku-Oki earthquake, were the questions of “how big, how bad, and how often” addressed in probabilistic seismic hazard analyses (PSHA) in Japan, one of the most earthquake-prone but most earthquake-prepared countries in the world? The guidance on how to properly perform PSHAs exists but may not be followed for a whole range of reasons, not all technical. One of the major emphases of these guidelines is that it must be recognized that there are significant uncertainties in our knowledge of earthquake processes and these uncertainties need to be fully incorporated into PSHAs. If such uncertainties are properly accounted for in PSHA, extreme events can be accounted for more often than not. This is not to say that no surprises will occur. That is the nature of trying to characterize a natural process such as earthquake generation whose properties also have random (aleatory) uncertainties. It must be stressed that no PSHA is ever final because new information and data need to be continuously monitored and addressed, often requiring an updated PSHA.