This study examines the performance of nonlinear total-stress wave-propagation site response analysis for modelling site effects in physics-based ground motion simulations of the 2010-2011 Canterbury, New Zealand earthquake sequence. This approach allows for explicit modeling of 3-dimensional ground motion phenomena at the regional scale, as well as detailed site effects and soil nonlinearity at the local scale. The approach is compared to a more commonly used empirical VS30 (30 m time-averaged shear wave velocity)-based method for computing site amplification as proposed by Graves and Pitarka (2010, 2015).
The 2010-2011 Canterbury earthquakes were recorded over a dense strong motion network in the near-source region, yielding significant observational evidence of seismic complexities, and a basis for interpretation of multi-disciplinary datasets and induced damage to the natural and built environment. This paper provides an overview of observed strong motions from these events and retrospective comparisons with both empirical and physics-based ground motion models. Both empirical and physics-based methods provide good predictions of observations at short vibration periods in an average sense. However, observed ground motion amplitudes at specific locations, such as Heathcote Valley, are seen to systematically depart from ‘average’ empirical predictions as a result of near surface stratigraphic and topographic features which are well modelled via sitespecific response analyses. Significant insight into the long period bias in empirical predictions is obtained from the use of hybrid broadband ground motion simulation. The comparison of both empirical and physics-based simulations against a set of 10 events in the sequence clearly illustrates the potential for simulations to improve ground motion and site response prediction, both at present, and further in the future.
This dissertation addresses a diverse range of topics in the physics-based broadband ground motion simulation, with a focus on New Zealand applications. In particular the following topics are addressed: the methodology and computational implementation of a New Zealand Velocity Model for broadband ground motion simulation; generalised parametric functions and spatial correlations for seismic velocities in the Canterbury, New Zealand region from surface-wave-based site characterisation; and ground motion simulations of Hope Fault earthquakes. The paragraphs below outline each contribution in more detail. A necessary component in physics-based ground motion simulation is a 3D model which details the seismic velocities in the region of interest. Here a velocity model construction methodology, its computational implementation, and application in the construction of a New Zealand velocity model for use in physics-based broadband ground motion simulation are presented. The methodology utilises multiple datasets spanning different length scales, which is enabled via the use of modular sub-regions, geologic surfaces, and parametric representations of crustal velocity. A number of efficiency-related workflows to decrease the overall computational construction time are employed, while maintaining the flexibility and extensibility to incorporate additional datasets and re- fined velocity parameterizations as they become available. The model comprises explicit representations of the Canterbury, Wellington, Nelson-Tasman, Kaikoura, Marlborough, Waiau, Hanmer and Cheviot sedimentary basins embedded within a regional travel-time tomography-based velocity model for the shallow crust and provides the means to conduct ground motion simulations throughout New Zealand for the first time. Recently developed deep shear-wave velocity profiles in Canterbury enabled models that better characterise the velocity structure within geologic layers of the Canterbury sedimentary basin to be developed. Here the development of depth- and Vs30-dependent para-metric velocity and spatial correlation models to characterise shear-wave velocities within the geologic layers of the Canterbury sedimentary basin are presented. The models utilise data from 22 shear-wave velocity profiles of up to 2.5km depth (derived from surface wave analysis) juxtaposed with models which detail the three-dimensional structure of the geologic formations in the Canterbury sedimentary basin. Parametric velocity equations are presented for Fine Grained Sediments, Gravels, and Tertiary layer groupings. Spatial correlations were developed and applied to generate three-dimensional stochastic velocity perturbations. Collectively, these models enable seismic velocities to be realistically represented for applications such as 3D ground motion and site response simulations. Lastly the New Zealand velocity model is applied to simulate ground motions for a Mw7.51 rupture of the Hope Fault using a physics-based simulation methodology and a 3D crustal velocity model of New Zealand. The simulation methodology was validated for use in the region through comparison with observations for a suite of historic small magnitude earthquakes located proximal to the Hope Fault. Simulations are compared with conventionally utilised empirical ground motion models, with simulated peak ground velocities being notably higher in regions with modelled sedimentary basins. A sensitivity analysis was undertaken where the source characteristics of magnitude, stress parameter, hypocentre location and kinematic slip distribution were varied and an analysis of their effect on ground motion intensities is presented. It was found that the magnitude and stress parameter strongly influenced long and short period ground motion amplitudes, respectively. Ground motion intensities for the Hope Fault scenario are compared with the 2016 Kaikoura Mw7.8 earthquake, it was found that the Kaikoura earthquake produced stronger motions along the eastern South Island, while the Hope Fault scenario resulted in stronger motions immediately West of the near-fault region. The simulated ground motions for this scenario complement prior empirically-based estimates and are informative for mitigation and emergency planning purposes.
The overarching goal of this dissertation is to improve predictive capabilities of geotechnical seismic site response analyses by incorporating additional salient physical phenomena that influence site effects. Specifically, multidimensional wave-propagation effects that are neglected in conventional 1D site response analyses are incorporated by: (1) combining results of 3D regional-scale simulations with 1D nonlinear wave-propagation site response analysis, and (2) modelling soil heterogeneity in 2D site response analyses using spatially-correlated random fields to perturb soil properties. A method to combine results from 3D hybrid physics-based ground motion simulations with site-specific nonlinear site response analyses was developed. The 3D simulations capture 3D ground motion phenomena on a regional scale, while the 1D nonlinear site response, which is informed by detailed site-specific soil characterization data, can capture site effects more rigorously. Simulations of 11 moderate-to-large earthquakes from the 2010-2011 Canterbury Earthquake Sequence (CES) at 20 strong motion stations (SMS) were used to validate simulations with observed ground motions. The predictions were compared to those from an empirically-based ground motion model (GMM), and from 3D simulations with simplified VS30- based site effects modelling. By comparing all predictions to observations at seismic recording stations, it was found that the 3D physics-based simulations can predict ground motions with comparable bias and uncertainty as the GMM, albeit, with significantly lower bias at long periods. Additionally, the explicit modelling of nonlinear site-response improves predictions significantly compared to the simplified VS30-based approach for soft-soil or atypical sites that exhibit exceptionally strong site effects. A method to account for the spatial variability of soils and wave scattering in 2D site response analyses was developed and validated against a database of vertical array sites in California. The inputs required to run the 2D analyses are nominally the same as those required for 1D analyses (except for spatial correlation parameters), enabling easier adoption in practice. The first step was to create the platform and workflow, and to perform a sensitivity study involving 5,400 2D model realizations to investigate the influence of random field input parameters on wave scattering and site response. Boundary conditions were carefully assessed to understand their effect on the modelled response and select appropriate assumptions for use on a 2D model with lateral heterogeneities. Multiple ground-motion intensity measures (IMs) were analyzed to quantify the influence from random field input parameters and boundary conditions. It was found that this method is capable of scattering seismic waves and creating spatially-varying ground motions at the ground surface. The redistribution of ground-motion energy across wider frequency bands, and the scattering attenuation of high-frequency waves in 2D analyses, resemble features observed in empirical transfer functions (ETFs) computed in other studies. The developed 2D method was subsequently extended to more complicated multi-layer soil profiles and applied to a database of 21 vertical array sites in California to test its appropriate- ness for future predictions. Again, different boundary condition and input motion assumptions were explored to extend the method to the in-situ conditions of a vertical array (with a sensor embedded in the soil). ETFs were compared to theoretical transfer functions (TTFs) from conventional 1D analyses and 2D analyses with heterogeneity. Residuals of transfer-function- based IMs, and IMs of surface ground motions, were also used as validation metrics. The spatial variability of transfer-function-based IMs was estimated from 2D models and compared to the event-to-event variability from ETFs. This method was found capable of significantly improving predictions of median ETF amplification factors, especially for sites that display higher event-to-event variability. For sites that are well represented by 1D methods, the 2D approach can underpredict amplification factors at higher modes, suggesting that the level of heterogeneity may be over-represented by the 2D random field models used in this study.
his poster presents the ongoing development of a 3D Canterbury seismic velocity model which will be used in physics-based hybrid broadband ground motion simulation of the 2010-2011 Canterbury earthquakes. Velocity models must sufficiently represent critical aspects of the crustal structure over multiple length scales which will influence the results of the simulations. As a result, numerous sources of data are utilized in order to provide adequate resolution where necessary. Figure 2: (a) Seismic reflection line showing P-wave velocities and significant geologic horizons (Barnes et al. 2011), and (b) Shear wave profiles at 10 locations (Stokoe et al. 2013). Figure 4: Cross sections of the current version of the Canterbury velocity model to depths of 10km as shown in Figure 1: (a) at a constant latitude value of -43.6˚, and (b) at a constant longitude value of 172.64˚. 3. Ground Surface and Geologic Horizon Models Figure 3: (a) Ground surface model derived from numerous available digital elevation models, and (b) Base of the Quaternary sediments derived from structural contours and seismic reflection line elevations. The Canterbury region has a unique and complex geology which likely has a significant impact on strong ground motions, in particular the deep and loose deposits of the Canterbury basin. The Canterbury basin has several implications on seismic wave phenomena such as long period ground motion amplification and wave guide effects. Using a realistic 3D seismic velocity model in physics-based ground motion simulation will implicitly account for such effects and the resultant simulated ground motions can be studied to gain a fundamental understanding of the salient ground motion phenomena which occurred during the Canterbury earthquakes, and the potential for repeat occurrences in the Canterbury region. Figure 1 shows the current model domain as a rectangular area between Lat=[-43.2˚,-44.0˚], and Lon=[171.5˚,173.0˚]. This essentially spans the area between the foot of the Southern Alps in the North West to Banks Peninsula in the East. Currently the model extends to a depth of 50km below sea level.
High-quality ground motion records are required for engineering applications including response history analysis, seismic hazard development, and validation of physics-based ground motion simulations. However, the determination of whether a ground motion record is high-quality is poorly handled by automation with mathematical functions and can become prohibitive if done manually. Machine learning applications are well-suited to this problem, and a previous feed-forward neural network was developed (Bellagamba et al. 2019) to determine high-quality records from small crustal events in the Canterbury and Wellington regions for simulation validation. This prior work was however limited by the omission of moderate-to-large magnitude events and those from other tectonic environments, as well as a lack of explicit determination of the minimum usable frequency of the ground motion. To address these shortcomings, an updated neural network was developed to predict the quality of ground motion records for all magnitudes and all tectonic sources—active shallow crustal, subduction intraslab, and subduction interface—in New Zealand. The predictive performance of the previous feed-forward neural network was matched by the neural network in the domain of small crustal records, and this level of predictive performance is now extended to all source magnitudes and types in New Zealand making the neural network applicable to global ground motion databases. Furthermore, the neural network provides quality and minimum usable frequency predictions for each of the three orthogonal components of a record which may then be mapped into a binary quality decision or otherwise applied as desired. This framework provides flexibility for the end user to predict high-quality records with various acceptability thresholds allowing for this neural network to be used in a range of applications.
This paper investigates the effects of variability in source rupture parameters on site-specific physics-based simulated ground motions, ascertained through the systematic analysis of ground motion intensity measures. As a preliminary study, we consider simulations of the 22 February 2011 Christchurch earthquake using the Graves and Pitarka (2015) methodology. The effects of source variability are considered via a sensitivity study in which parameters (hypocentre location, earthquake magnitude, average rupture velocity, fault geometry and the Brune stress parameter) are individually varied by one standard deviation. The sensitivity of simulated ground motion intensity measures are subsequently compared against observational data. The preliminary results from this study indicate that uncertainty in the stress parameter and the rupture velocity have the most significant effect on the high frequency amplitudes. Conversely, magnitude uncertainty was found to be most influential on the spectral acceleration amplitudes at low frequencies. Further work is required to extend this preliminary study to exhaustively consider more events and to include parameter covariance. The ultimate results of this research will assist in the validation of the overall simulation method’s accuracy in capturing various rupture parameters, which is essential for the use of simulated ground motion models in probabilistic seismic hazard analysis.
he strong motion station at Heathcote Valley School (HVSC) recorded unusually high peak ground accelerations (2.21g vertical and 1.41g horizontal) during the February 2011 Christchurch earthquake. Ground motions recorded at HVSC in numerous other events also exhibited consistently higher intensities compared with nearby strong motion stations. We investigated the underlying causes of such high intensity ground motions at HVSC by means of 2D dynamic finite element analyses, using recorded ground motions during the 2010-2011 Canterbury earthquake sequence. The model takes advantage of a LiDAR-based digital elevation model (DEM) to account for the surface topography, while the geometry and dynamic properties of the surficial soils are characterized by seismic cone penetration tests (sCPT) and Multi-Channel Analyses of Surface Waves (MASW). Comparisons of simulated and recorded ground motions suggests that our model performs well for distant events, while for near-field events, ground motions recorded at the adopted reference station at Lyttelton Port are not reasonable input motions for the simulation. The simulations suggest that Rayleigh waves generated at the inclined interface of the surficial colluvium and underlying volcanic rock strongly affect the ground motions recorded at HVSC, in particular, being the dominant contributor to the recorded vertical motions.
This dissertation addresses a diverse range of topics in the area of physics-based ground motion simulation with particular focus on the Canterbury, New Zealand region. The objectives achieved provide the means to perform hybrid broadband ground motion simulation and subsequently validates the simulation methodology employed. In particu- lar, the following topics are addressed: the development of a 3D seismic velocity model of the Canterbury region for broadband ground motion simulation; the development of a 3D geologic model of the interbedded Quaternary formations to provide insight on observed ground motions; and the investigation of systematic effects through ground motion sim- ulation of small-to-moderate magnitude earthquakes. The paragraphs below outline each contribution in more detail. As a means to perform hybrid broadband ground motion simulation, a 3D model of the geologic structure and associated seismic velocities in the Canterbury region is devel- oped utilising data from depth-converted seismic reflection lines, petroleum and water well logs, cone penetration tests, and implicitly guided by existing contour maps and geologic cross sections in data sparse subregions. The model explicitly characterises five significant and regionally recognisable geologic surfaces that mark the boundaries between geologic units with distinct lithology and age, including the Banks Peninsula volcanics, which are noted to strongly influence seismic wave propagation. The Basement surface represents the base of the Canterbury sedimentary basin, where a large impedance contrast exists re- sulting in basin-generated waves. Seismic velocities for the lithological units between the geologic surfaces are derived from well logs, seismic reflection surveys, root mean square stacking velocities, empirical correlations, and benchmarked against a regional crustal model, thus providing the necessary information for a Canterbury velocity model for use in broadband seismic wave propagation. A 3D high-resolution model of the Quaternary geologic stratigraphic sequence in the Canterbury region is also developed utilising datasets of 527 high-quality water well logs, and 377 near-surface cone penetration test records. The model, developed using geostatistical Kriging, represents the complex interbedded regional Quaternary geology by characterising the boundaries between significant interbedded geologic formations as 3D surfaces including explicit modelling of the formation unconformities resulting from the Banks Peninsula volcanics. The stratigraphic layering present can result in complex wave propagation. The most prevalent trend observed in the surfaces was the downward dip from inland to the eastern coastline as a result of the dominant fluvial depositional environment of the terrestrial gravel formations. The developed model provides a benefi- cial contribution towards developing a comprehensive understanding of recorded ground motions in the region and also providing the necessary information for future site char- acterisation and site response analyses. To highlight the practicality of the model, an example illustrating the role of the model in constraining surface wave analysis-based shear wave velocity profiling is illustrated along with the calculation of transfer functions to quantify the effect of the interbedded geology on wave propagation. Lastly, an investigation of systematic biases in the (Graves and Pitarka, 2010, 2015) ground motion simulation methodology and the specific inputs used for the Canterbury region is presented considering 144 small-to-moderate magnitude earthquakes. In the simulation of these earthquakes, the 3D Canterbury Velocity Model, developed as a part of this dissertation, is used for the low-frequency simulation, and a regional 1D velocity model for the high-frequency simulation. Representative results for individual earthquake sources are first presented to highlight the characteristics of the small-to-moderate mag- nitude earthquake simulations through waveforms, intensity measure scaling with source- to-site distance, and spectral bias of the individual events. Subsequently, a residual de- composition is performed to examine the between- and within-event residuals between observed data, and simulated and empirical predictions. By decomposing the residuals into between- and within-event residuals, the biases in source, path and site effects, and their causes, can be inferred. The residuals are comprehensively examined considering their aggregated characteristics, dependence on predictor variables, spatial distribution, and site-specific effects. The results of the simulation are also benchmarked against empir- ical ground motion models, where their similarities manifest from common components in their prediction. Ultimately, suggestions to improve the predictive capability of the simulations are presented as a result of the analysis.
This article presents a quantitative case study on the site amplification effect observed at Heathcote Valley, New Zealand, during the 2010-2011 Canterbury earthquake sequence for 10 events that produced notable ground acceleration amplitudes up to 1.4g and 2.2g in the horizontal and vertical directions, respectively. We performed finite element analyses of the dynamic response of the valley, accounting for the realistic basin geometry and the soil non-linear response. The site-specific simulations performed significantly better than both empirical ground motion models and physics based regional-scale ground motion simulations (which empirically accounts for the site effects), reducing the spectral acceleration prediction bias by a factor of two in short vibration periods. However, our validation exercise demonstrated that it was necessary to quantify the level of uncertainty in the estimated bedrock motion using multiple recorded events, to understand how much the simplistic model can over- or under-estimate the ground motion intensities. Inferences from the analyses suggest that the Rayleigh waves generated near the basin edge contributed significantly to the observed high frequency (f>3Hz) amplification, in addition to the amplification caused by the strong soil-rock impedance contrast at the site fundamental frequency. Models with and without considering soil non-linear response illustrate, as expected, that the linear elastic assumption severely overestimates ground motions in high frequencies for strong earthquakes, especially when the contribution of basin edge-generated Rayleigh waves becomes significant. Our analyses also demonstrate that the effect of pressure-dependent soil velocities on the high frequency ground motions is as significant as the amplification caused by the basin edge-generated Rayleigh waves.
This paper summarizes the development of a high-resolution surficial shear wave velocity model based on the combination of the large high-spatial-density database of cone penetration test (CPT) logs in and around Christchurch, New Zealand and a recently-developed Christchurch-specific empirical correlation between soil shear wave velocity and CPT. This near-surface shear wave velocity model has applications for site characterization efforts via the development of maps of time-averaged shear wave velocities over specific depths, as well as use in site response analysis and ground motion simulation.
This presentation summarizes the development of high-resolution surficial soil velocity models in the Canterbury, New Zealand basin. Shallow (<30m) shear wave velocities were primarily computed based on a combination of a large database of over 15,000 cone penetration test (CPT) logs in and around Christchurch, and a recently-developed Christchurch-specific empirical correlation between soil shear wave velocity and CPT. Large active-source testing at 22 locations and ambient-wavefield surface wave and H/V testing at over 80 locations were utilized in combination with 1700 water well logs to constrain the inter-bedded stratigraphy and velocity of Quaternary sediments up to depths of several hundred meters. Finally, seismic reflection profiles and the ambient-wavefield surface wave data provide constraint on velocities from several hundred meters to several kilometres. At all depths, the high resolution data illustrates the complexity of the soil conditions in the region, and the developed 3D models are presently being used in broadband ground motion simulations to further interpret the observed strong ground motions in the 2010-2011 Canterbury earthquake sequence.
This report summarizes the development of a region-wide surficial soil shear wave velocity (Vs ) model based on the unique combination of a large high-spatial-density database of cone penetration test (CPT) logs in the greater Christchurch urban area (> 15, 000 logs as of 1 February 2014) and the Christchurch-specific empirical correlation between soil Vs and CPT data developed by McGann et al. [1, 2]. This model has applications for site characterization efforts via maps of time-averaged Vs over specific depths (e.g. Vs30, Vs10), and for numerical modeling efforts via the identification of typical Vs profiles for different regions and soil behaviour types within Christchurch. In addition, the Vs model can be used to constrain the near-surface velocities for the 3D seismic velocity model of the Canterbury basin [3] currently being developed for the purpose of broadband ground motion simulation. The general development of these region-wide near-surface Vs models includes the following general phases, with each discussed in separate chapters of this report. • An evaluation of the available CPT dataset for suitability, and the definition of other datasets and assumptions necessary to characterize the surficial sediments of the region to 30 m depth. • The development of time-averaged shear wave velocity (Vsz) surfaces for the Christchurch area from the adopted CPT dataset (and supplementary data/assumptions) using spatial interpolation. The Vsz surfaces are used to explore the characteristics of the near-surface soils in the regions and are shown to correspond well with known features of the local geology, the historical ecosystems of the area, and observations made following the 2010- 2011 Canterbury earthquakes. • A detailed analysis of the Vs profiles in eight subregions of Christchurch is performed to assess the variablity in the soil profiles for regions with similar Vsz values and to assess Vsz as a predictive metric for local site response. It is shown that the distrubution of soil shear wave velocity in the Christchurch regions is highly variable both spatially (horizontally) and with depth (vertically) due to the varied geological histories for different parts of the area, and the highly stratified nature of the nearsurface deposits. This variability is not considered to be greatly significant in terms of current simplified site classification systems; based on computed Vs30 values, all considered regions can be categorized as NEHRP sites class D (180 < Vs < 360 m/s) or E (Vs < 180 m/s), however, detailed analysis of the shear wave velocity profiles in different subregions of Christchurch show that the expected surficial site response can vary quite a bit across the region despite the relative similarity in Vs30
