Gail M. Atkinson

Ground-Motion Prediction Equations for the Average Horizontal Component of PGA, PGV, and 5%-Damped PSA at Spectral Periods between 0.01 s and 10.0 s

This paper contains ground-motion prediction equations (GMPEs) for average horizontal-component ground motions as a function of earthquake magnitude, distance from source to site, local average shear-wave velocity, and fault type. Our equations are for peak ground acceleration (PGA), peak ground velocity (PGV), and 5%-damped pseudo-absolute-acceleration spectra (PSA) at periods between 0.01 s and 10 s. They were derived by empirical regression of an extensive strong-motion database compiled by the “PEER NGA” (Pacific Earthquake Engineering Research Centers Next Generation Attenuation) project. For periods less than 1 s, the analysis used 1,574 records from 58 mainshocks in the distance range from 0 km to 400 km (the number of available data decreased as period increased). The primary predictor variables are moment magnitude (M), closest horizontal distance to the surface projection of the fault plane ( RJB ), and the time-averaged shear-wave velocity from the surface to 30 m (VS30). The equations are applicable for M=5–8, RJB <200 km, and VS30=180–1300 m/s.

Empirical Ground-Motion Relations for Subduction-Zone Earthquakes and Their Application to Cascadia and Other Regions

Ground-motion relations for earthquakes that occur in subduction zones are an important input to seismic-hazard analyses in many parts of the world. In the Cascadia region (Washington, Oregon, northern California, and British Columbia), for example, there is a significant hazard from megathrust earthquakes along the subduction interface and from large events within the subducting slab. These hazards are in addition to the hazard from shallow earthquakes in the overlying crust. We have compiled a response spectra database from thousands of strong-motion record- ings from events of moment magnitude (M) 5-8.3 occurring in subduction zones around the world, including both interface and in-slab events. The 2001 M 6.8 Nis- qually and 1999 M 5.9 Satsop earthquakes are included in the database, as are many records from subduction zones in Japan (Kyoshin-Net data), Mexico (Guerrero data), and Central America. The size of the database is four times larger than that available for previous empirical regressions to determine ground-motion relations for subduc- tion-zone earthquakes. The large dataset enables improved determination of attenu- ation parameters and magnitude scaling, for both interface and in-slab events. Soil response parameters are also better determined by the data. We use the database to develop global ground-motion relations for interface and in-slab earthquakes, using a maximum likelihood regression method. We analyze regional variability of ground-motion amplitudes across the global database and find that there are significant regional differences. In particular, amplitudes in Cascadia differ by more than a factor of 2 from those in Japan for the same magnitude, distance, event type, and National Earthquake Hazards Reduction Program (NEHRP) soil class. This is believed to be due to regional differences in the depth of the soil profile, which are not captured by the NEHRP site classification scheme. Regional correction factors to account for these differences are proposed for Cascadia and Japan. The results of this study differ significantly from previous analyses based on more limited data and have important implications for seismic-hazard analysis. The ground-motion relations predict that a great megathrust earthquake (M 8) at a fault distance of about 100 km would produce pseudoacceleration (PSA), 5% damped, horizontal component on soil sites of about 110 cm/sec 2 at 0.5 Hz, 660 cm/sec 2 at 2.5 Hz, and 410 cm/sec 2 at 5 Hz, with a peak ground acceleration of about 180 cm/ sec 2 . These damaging levels of motion would be experienced over a very large area, corresponding to a rectangular area about 300 km wide by 500 km long. Large in- slab events (M 7.5) would produce even higher PSA values within 100 km of the fault, but the in-slab motions attenuate much more rapidly with distance. Thus the hazard posed by moderate to large in-slab events such as the 2001 Nisqually earth- quake is modest compared to that of a Cascadia megathrust earthquake of M 8, in terms of the area that would experience damaging levels of ground motion.

NGA-West2 Equations for Predicting PGA, PGV, and 5% Damped PSA for Shallow Crustal Earthquakes

We provide ground motion prediction equations for computing medians and standard deviations of average horizontal component intensity measures (IMs) for shallow crustal earthquakes in active tectonic regions. The equations were derived from a global database with M 3.0–7.9 events. We derived equations for the primary M- and distance-dependence of the IMs after fixing the VS30-based nonlinear site term from a parallel NGA-West2 study. We then evaluated additional effects using mixed effects residuals analysis, which revealed no trends with source depth over the M range of interest, indistinct Class 1 and 2 event IMs, and basin depth effects that increase and decrease long-period IMs for depths larger and smaller, respectively, than means from regional VS30-depth relations. Our aleatory variability model captures decreasing between-event variability with M, as well as within-event variability that increases or decreases with M depending on period, increases with distance, and decreases for soft sites.

Stochastic Modeling of California Ground Motions

Ground-motion relations are developed for California using a stochastic simulation method that exploits the equivalence between finite-fault models and a two-corner point-source model of the earthquake spectrum. First, stochastic simu- lations are generated for finite-fault ruptures, in order to define the average shape and amplitude level of the radiated spectrum at near-source distances as a function of earthquake size. The length and width of the fault plane are defined based on the moment magnitude of the earthquake and modeled with an array of subfaults. The radiation from each subfault is modeled as a Brune point source using the stochastic model approach; the subfault spectrum has a single-corner frequency. An earthquake rupture initiates at a randomly chosen subfault (hypocenter), and propagates in all directions along the fault plane. A subfault is triggered when rupture propagation reaches its center. Simulations are generated for an observation point by summing the subfault time series, appropriately lagged in time. Fourier spectra are computed for records simulated at many azimuths, placed at equidistant observation points around the fault. The mean Fourier spectrum for each magnitude, at a reference near- source distance, is used to define the shape and amplitude levels of an equivalent point-source spectrum that mimics the salient finite-fault effects. The functional form for the equivalent point-source spectrum contains two corner frequencies. Stochastic point-source simulations, using the derived two-corner source spectrum, are then performed to predict peak-ground-motion parameters and response spectra for a wide range of magnitudes and distances, for generic California sites. The sto- chastic ground-motion relations, given in the Appendix for rock and soil sites, are in good agreement with the empirical strong-motion database for California; the average ratio of observed to simulated amplitudes is near unity over all frequencies from 0.2 to 12 Hz. The stochastic relations agree well with empirical regression equations (e.g., Abrahamson and Silva, 1997; Boore et al., 1997; Sadigh et al., 1997) in the magnitude-distance ranges well represented by the data, but are better con- strained at large distances, due to the use of attenuation parameters based on regional seismographic data. The stochastic ground-motion relations provide a sound basis for estimation of ground motions for earthquakes of magnitude 4 through 8, at dis- tances from 1 to 200 km.

Comparisons of the NGA ground-motion relations

The data sets, model parameterizations, and results from the five NGA models for shallow crustal earthquakes in active tectonic regions are compared. A key difference in the data sets is the inclusion or exclusion of aftershocks. A comparison of the median spectral values for strike-slip earthquakes shows that they are within a factor of 1.5 for magnitudes between 6.0 and 7.0 for distances less than 100 km. The differences increase to a factor of 2 for M5 and M8 earthquakes, for buried ruptures, and for distances greater than 100 km. For soil sites, the differences in the modeling of soil/sediment depth effects increase the range in the median long-period spectral values for M7 strike-slip earthquakes to a factor of 3. The five models have similar standard deviations for M6.5-M7.5 earthquakes for rock sites and for soil sites at distances greater than 50 km. Differences in the standard deviations of up to 0.2 natural log units for moderate magnitudes at all distances and for large magnitudes at short distances result from the treatment of the magnitude dependence and the effects of nonlinear site response on the standard deviation.

NGA-West2 Research Project

The NGA-West2 project is a large multidisciplinary, multi-year research program on the Next Generation Attenuation (NGA) models for shallow crustal earthquakes in active tectonic regions. The research project has been coordinated by the Pacific Earthquake Engineering Research Center (PEER), with extensive technical interactions among many individuals and organizations. NGA-West2 addresses several key issues in ground-motion seismic hazard, including updating the NGA database for a magnitude range of 3.0–7.9; updating NGA ground-motion prediction equations (GMPEs) for the “average” horizontal component; scaling response spectra for damping values other than 5%; quantifying the effects of directivity and directionality for horizontal ground motion; resolving discrepancies between the NGA and the National Earthquake Hazards Reduction Program (NEHRP) site amplification factors; analysis of epistemic uncertainty for NGA GMPEs; and developing GMPEs for vertical ground motion. This paper presents an overview of the NGA-West2 research program and its subprojects.

Probabilistic characterization of spatially-correlated response spectra for earthquakes in Japan

Seismic hazard and risk assessments of spatially distributed infrastructural systems require seismic demand models that capture random but correlated simultaneous seismic effects at multiple sites. This study characterizes spatially correlated ground-motion parameters probabilistically using comprehensive databases of the K-NET and KiK-net strong-motion networks in Japan by developing a ground-motion prediction equation and then investigating the correlation structure of regression residuals from the prediction equation. Analysis results indicate that (1) interevent residuals of ground-motion parameters at different vibration periods are more strongly correlated than intraevent and total residuals with zero separation distance; and (2) intraevent spatial correlation coefficients can be described as a simple exponential decay function that is independent of the way the event-based intraevent standard deviation is calculated, of the earthquake type, and of the vibration period. The developed overall correlation model of spatially correlated ground-motion parameters may be used for seismic hazard and risk assessments in a subduction environment.

Hydraulic Fracturing and Seismicity in the Western Canada Sedimentary Basin

The development of most unconventional oil and gas resources relies upon subsurface injection of very large volumes of fluids, which can induce earthquakes by activating slip on a nearby fault. During the last 5 years, accelerated oilfield fluid injection has led to a sharp increase in the rate of earthquakes in some parts of North America. In the central United States, most induced seismicity is linked to deep disposal of coproduced wastewater from oil and gas extraction. In contrast, in western Canada most recent cases of induced seismicity are highly correlated in time and space with hydraulic fracturing, during which fluids are injected under high pressure during well completion to induce localized fracturing of rock. Furthermore, it appears that the maximum-observed magnitude of events associated with hydraulic fracturing may exceed the predictions of an often-cited relationship between the volume of injected fluid and the maximum expected magnitude. These findings have far-reaching implications for assessment of inducedseismicity hazards.

Simple Computation of Liquefaction Probability for Seismic Hazard Applications

A computer program has been written for the routine calculation of the probability of seismically induced liquefaction for level sand or silty sand sites. The method combines Seeds simplified method of assessing liquefaction potential with the Cornell method of assessing seismic hazard, by modifying the latter to consider the joint probability of magnitude and acceleration. The input data to the program are simple geotechnical data from SPT, CPT or laboratory tests, and readily obtainable regional seismicity data. The program is simple to use.

The 26 January 2001 M 7.6 Bhuj, India, Earthquake: Observed and Predicted Ground Motions

Although local and regional instrumental recordings of the devastating 26, January 2001, Bhuj earthquake are sparse, the distribution of macroseismic effects can provide important constraints on the mainshock ground motions. We compiled available news accounts describing damage and other effects and interpreted them to obtain modified Mercalli intensities (MMIs) at >200 locations throughout the Indian subcontinent. These values are then used to map the intensity distribution throughout the subcontinent using a simple mathematical interpolation method. Although preliminary, the maps reveal several interesting features. Within the Kachchh region, the most heavily damaged villages are concentrated toward the western edge of the inferred fault, consistent with western directivity. Significant sediment-induced amplification is also suggested at a number of locations around the Gulf of Kachchh to the south of the epicenter. Away from the Kachchh region, intensities were clearly amplified significantly in areas that are along rivers, within deltas, or on coastal alluvium, such as mudflats and salt pans. In addition, we use fault-rupture parameters inferred from teleseismic data to predict shaking intensity at distances of 0–1000 km. We then convert the predicted hard-rock ground-motion parameters to MMI by using a relationship (derived from Internet-based intensity surveys) that assigns MMI based on the average effects in a region. The predicted MMIs are typically lower by 1–3 units than those estimated from news accounts, although they do predict near-field ground motions of approximately 80%g and potentially damaging ground motions on hard-rock sites to distances of approximately 300 km. For the most part, this discrepancy is consistent with the expected effect of sediment response, but it could also reflect other factors, such as unusually high building vulnerability in the Bhuj region and a tendency for media accounts to focus on the most dramatic damage, rather than the average effects. The discrepancy may also be partly attributable to the inadequacy of the empirical relationship between MMI and peak ground acceleration (PGA), when applied to India. The MMI–PGA relationship was developed using data from California earthquakes, which might have a systematically different stress drop and therefore, a different frequency content than intraplate events. When a relationship between response spectra and MMI is used, we obtain larger predicted MMI values, in better agreement with the observations.