Distribution Of Ground Motion Research Articles

Small intermediate fault segments can either aid or hinder rupture propagation at stepovers

Large‐scale geometrical complexities along faults are known to be likely endpoints for coseismic rupture, as suggested by analysis of historic ruptures and corroborated by models of rupture on bent or discontinuous faults. However, natural faults also include smaller‐scale complexities. We use the 3D finite element method to model dynamic ruptures on strike‐slip fault stepovers with a smaller intermediate fault between the main strands. We find that such small faults can have a controlling effect on rupture behavior and ground motion intensity and distribution. In particular, the intermediate fault can either aid or prevent rupture propagation across the stepover, depending on its length and basal depth. The results have important implications for hazard assessment of faults with large‐ and small‐scale geometrical complexities, and also suggest that more site‐specific modeling studies may be necessary to develop realistic rupture scenarios for individual complex fault systems.

Seismically induced boulder displacement in the Port Hills, New Zealand during the 2010 Darfield (Canterbury) earthquake

An analysis of boulders displaced during the September 2010 M W 7.1 Darfield (Canterbury) earthquake provides non-instrumental constraints on the variability, distribution and origin of strong ground motion during major earthquakes. Boulders ranging in mass from 10 to 5000 kg were displaced 8–970 cm laterally from hosting soil sockets of <1 cm to 50 cm depth at several sites in the Port Hills, roughly 35 km southeast of the earthquake epicentre. Boulder displacement was observed on N-striking (000–015°) ridges above c. 400 m elevation but not on NE-, NW- and SE-striking ridges. The prevailing boulder horizontal displacement azimuth of 250±20° is subparallel with the direction of instrumentally recorded transient peak ground horizontal displacements. Boulder displacement distance has no correlation with displacement azimuth, boulder mass or soil socket depth and has a partial correlation with slope angle. The lateral displacement of many boulders from low slope (<10°) ground surfaces on ridge crests exceeds nearby instrumentally recorded peak ground displacements at lower elevations by up to an order of magnitude, implying that seismic waves were amplified at the study sites. Preliminary 2-D FLAC modelling suggests that topographic amplification may explain this observation. The co-existence of displaced and non-displaced boulders at proximal (<1 m spacing) sites also suggests small-scale ground motion variability and/or varying boulder-ground dynamic interactions relating to shallow phenomena such as variability in soil depth, bedrock fracture density and/or microtopography on the bedrock–soil interface. Remapping of boulders following the February 2011 M W 6.2 Christchurch earthquake reveals no subsequent relocation despite locally recorded horizontal and vertical ground accelerations well in excess of the Darfield earthquake and pervasive rockfalls and landslides elsewhere. This study successfully identifies some of the major controls on spatial ground motion variability at non-instrumented locations and highlights the complexity of ground response at different spatial scales and for different earthquake characteristics.

Three-Dimensional Elastic-Plastic Dynamic Interaction of Earth Dam and Intermountain Basin

A new method that based on an interface program and secondary development of the analysis software FLAC3D was used to present the elastic-plastic dynamic interaction of earth dam and intermountain basin in 3D. By this method, we developed a complex 3D numerical model for a typical earth dam which was damaged seriously in the Wenchuan earthquake, along with its surrounding sites, considering the pore water pressure and the hysteretic damping of the soils, inputting the revised ground motion obtained from the station near the dam. The simulation reappeared the dam disaster process, gave the spatial and temporal distribution of PGA (peak ground acceleration) and residual deformation and the ratio of vertical and horizontal PGA. Our data demonstrate the mechanisms of the landslides, collapses and cracks on the earth dams. The present work shows that the existence of the dam has a significant influence on the distribution of ground motion in the intermountain basin.

Estimating two-dimensional static stabilities and geomorphic settings of precariously balanced rocks from unconstrained digital photographs

The need to accurately document the spatiotemporal distribution of earthquake-generated strong ground motions is essential for evaluating the seismic vulnerability of sites of critical infrastructure. Understanding the threshold for maximum earthquake-induced ground motions at such sites provides valuable information to seismologists, earthquake engineers, local agencies, and policymakers when determining ground motion hazards of seismically sensitive infrastructures. In this context, fragile geologic features such as precariously balanced rocks (PBRs) serve as negative evidence for earthquake-induced ground motions and provide important physical constraints on the upper limits of ground motions. The three-dimensional (3D) shape of a PBR is a critical factor in determining its static stability and thus susceptibility to toppling during strong ground shaking events. Furthermore, the geomorphic settings of PBRs provide important controls on PBR exhumation histories that are interpreted from surface exposure dating methods. In this paper, we present PBRslenderness, a MATLAB-based program that evaluates the two-dimensional (2D) static stabilities of PBRs from unconstrained digital photographs. The program’s graphical user interface allows users to interactively digitize a PBR and calculates the 2D geometric parameters that define its static stability. A reproducibility study showed that our 2D calculations compare well against their counterparts that were computed in 3D (R 2 = 0.77–0.98 for 22 samples). A sensitivity study for single-user and multiuser digitization routines further confirmed the reproducibility of PBRslenderness estimates (coefficients of variation c v = 4.3%–6.5% for 100 runs; R 2 = 0.87–0.99 for 20 PBRs). We used PBRslenderness to analyze 261 PBRs in a low-seismicity setting to investigate the local geomorphic controls on PBR stability and preservation. PBRslenderness showed that a PBR’s shape strongly controls its static stability and that there is no relationship between a PBR’s stability and its geomorphic location in a drainage basin. However, the geomorphic settings of PBRs control their preservation potential by restricting their formation to hillslope gradients

Near-source strong ground motions observed in the 22 February 2011 Christchurch earthquake

This manuscript provides a critical examination of the ground motions recorded in the near-source region resulting from the 22 February 2011 Christchurch earthquake. Particular attention is given to reconciling the observed spatial distribution of ground motions in terms of physical phenomena related to source, path and site effects. The large number of near-source observed strong ground motions show clear evidence of: forward-directivity, basin generated surface waves, liquefaction and other significant nonlinear site response. The pseudo-acceleration response spectra (SA) amplitudes and significant duration of strong motions agree well with empirical prediction models, except at long vibration periods where the influence of basin-generated surface waves and nonlinear site response are significant and not adequately accounted for in empirical SA models. Pseudo-acceleration response spectra are also compared with those observed in the 4 September 2010 Darfield earthquake and routine design response spectra used in order to emphasise the amplitude of ground shaking and elucidate the importance of local geotechnical characteristics on surface ground motions. The characteristics of the observed vertical component accelerations are shown to be strongly dependent on source-to-site distance and are comparable with those from the 4 September 2010 Darfield earthquake, implying the large amplitudes observed are simply a result of many observations at close distances rather than a peculiar source effect.

Near-source Strong Ground Motions Observed in the 22 February 2011 Christchurch Earthquake

On 22 February 2011 at 12:51 p.m. local time, a moment magnitude Mw 6.3 earthquake occurred beneath the city of Christchurch, New Zealand, causing an level of damage and human casualties unparalleled in the country's history. Compared to the preceding 4 September 2010 Mw 7.1 Darfield earthquake, which occurred approximately 30 km to the west of Christchurch, the close proximity of the 22 February event led to ground motions of significantly higher amplitude in the densely populated regions of Christchurch. As a result of these significantly larger ground motions, structures in general, and commercial structures in the central business district in particular, were subjected to severe seismic demands and, combined with the event timing, structural collapses accounted for the majority of the 181 casualties (New Zealand Police 2011). This manuscript provides a preliminary assessment of the near-source ground motions recorded in the Christchurch region. Particular attention is given to the observed spatial distribution of ground motions, which is interpreted based on source, path, and site effects. Comparison is also made of the observed ground motion response spectra with those of the 4 September 2010 Darfield earthquake and those used in seismic design in order to emphasize the amplitude of the ground shaking and also elucidate the importance of local geotechnical and deep geologic structure on surface ground motions. New Zealand resides on the boundary of the Pacific and Australian plates (Figure 1) and its active tectonics are dominated by: 1) oblique subduction of the Pacific plate beneath the Australian plate along the Hikurangi trough in the North Island; 2) oblique subduction of the Australian plate beneath the Pacific plate along the Puysegur trench in the southwest of the South Island; and 3) oblique, right-lateral slip along numerous crustal faults in the axial tectonic belt, of which the …

Ground motion for scenario earthquakes at Guwahati city

In this article, stochastic finite-fault simulation combined with site response analysis is used to understand the spatial distribution of ground motion in Guwahati city due to three damaging earthquakes. The rock level ground motion for the scenario earthquakes is generated based on the stochastic finite-fault methodology. These simulated motions are further amplified up to the surface by equivalent linear site response analyses using the available borelog data at 100 different locations in Guwahati city. A set of twenty simulated rock level time histories for each event, are used to compute the surface level ground motion. Response spectra are computed and the results are presented in the form of contour maps, at selected natural periods. The mean amplification due to local soil deposit is as high as 2.2 at most of the sites in Guwahati city. Based on these simulated motions, an average site correction factor is obtained for soil sites in Guwahati city. The standard error in the simulated response spectra is also reported. The contour maps obtained will be useful in identifying vulnerable places in Guwahati city.

Determination of 0 and Rock Site from Records of the 2008/2009 Earthquake Swarm in Western Bohemia

It is now generally accepted within the probabilistic seismic hazard analysis (PSHA) community that ground motion has to be treated as a random variable (Abrahamson 2000). As a consequence, the calculation of ground-motion exceedance rates for different levels of ground-motion intensity, plots of which are commonly referred to as hazard curves, requires an integration over the full ground-motion distribution. The spread of this distribution is known to strongly influence the estimated hazard levels, in particular at low exceedance rates. This is one reason why modern probabilistic seismic-hazard studies, in which this spread is usually accounted for via the standard deviation of the chosen ground-motion prediction equation, generally referred to as sigma ( σ ), lead to increased hazard estimates in comparison to studies in which the variability was ignored (Bommer and Abrahamson 2006). Assuming that the σ values in ground-motion models represent only irreducible aleatory variability is equivalent to the assumption of complete ergodicity, which may also not be realistic ( e.g. , Anderson and Brune 1999). Reducing the σ values in ground-motion models to their aleatory component, in other words trying to drop the ergodic assumption, is one of the main challenges in PSHA. In this context, the identification and possible isolation of different components of ground-motion variability are of critical importance. Because of the strict geometrical requirements, the number of datasets permitting such attempts is still rather limited (for a discussion see Al Atik et al. 2010). In particular, the determination of the standard deviation of the between-events, single source-region residual, which Al Atik et al. (2010) denoted as tau-zero ( τ ) puts very strong constraints on the datasets. It requires a large number of well-recorded ground motions at individual sites from a single source region, for example recordings of earthquake swarms. Earthquake swarms occur in many regions of the world. …

The 6 September 2009 Mw5.4 Earthquake in Eastern Albania – FYROM Border: Focal Mechanisms, Slip Model, ShakeMap

On 6 September 2009 (GMT 21:49) a moderate Mw5.4 earthquake sequence burst at the eastern border of Albania with the Former Yugoslav Republic of Macedonia (FYROM). The main shock was located ~6 km north of the epicentre of the 30 November 1967 Mw6.2 Dibra (or Debar) earthquake, which caused loss of life and considerable damage to buildings. We use broad band waveforms recorded by the Hellenic Unified Seismic Network (HUSN), which receives real-time waveforms from the neighbouring networks, to compute focal mechanisms, obtain the slip model and derive the Shake Map of the mainshock. The focal mechanisms of 18 of the stronger events of the sequence, obtained through time-domain moment tensor inversion, indicate that deformation is taken up by NNE-SSW-trending normal faults, in agreement with the ~E-W extension previously identified within the Albanian orogen. Our results show that the 2009 main shock ruptured a roughly 9 km normal fault at a depth of 6 km, which strikes at 194° and dips west at ~45°. The slip of the main shock was confined to a single patch of ~9 km × 6 km, the average slip was 5 cm and the peak slip was 18 cm. The slip model was incorporated in a forward modelling scheme to simulate the ground motion distribution in the near field. The Shake Map thus obtained, based on the distribution of Peak Ground Velocity at phantom stations, outlines the mesoseismal area within the Dibra and Bulqiza districts in Albania, in accordance with macroseismic observations. The region affected by the 2009 sequence, together with the seismogenic region of the 1967 Dibra event, form a roughly NNE-SSW-trending structure which is an active seismotectonic zone in eastern Albania constituting a threat for nearby urban areas.

Strong Ground-Motion Simulation of the 12 May 2008 Mw 7.9 Wenchuan Earthquake, Using Various Slip Models

Abstract On 12 May 2008, a devastating earthquake of M w 7.9 occurred in the Sichuan Province of China. The earthquake had disastrous consequences and cost more than 69,000 lives. The rupture occurred along a 300-km-long fault dipping northwest along the Longmen Shan fold-thrust belt, which separates the Longmen Shan of the Tibetan Plateau in the northwest from the Sichuan Basin in the southeast and is associated with a significant change in the crustal thickness. We simulate the ground shaking due to the Wenchuan earthquake by applying a hybrid broadband frequency strong ground-motion simulation technique that combines deterministic simulation of low frequencies (0.1–1.0 Hz) with semistochastic modeling of high frequencies (1.0–10 Hz). We use three available finite fault-slip models obtained from waveform inversion as input models for the earthquake scenarios. The resulting simulations reveal large variations in ground shaking due to the rupture complexity in terms of a variety of factors, including the location of asperities and the width of the fault plane. The applied methodology successfully reproduces the strong ground-motion distribution and frequency content of the seismic waves. The simulated ground motions, when calibrated with the recorded data, can also be used to verify the finite fault-slip models based on different teleseismic data and inversion schemes. Comparison with the damage distribution from reconnaissance field observations confirms the fault rupture complexity. The applied simulation methodology provides a promising platform for predictive studies.

The Variability of Ground-Motion Prediction Models and Its Components

Modern ground-motion prediction models use datasets of recorded ground-motion parameters at multiple stations during different earthquakes and in various source regions to generate equations that are later used to predict site-specific ground motions. These models describe the distribution of ground motion in terms of a median and a logarithmic standard deviation ( e.g. , Strasser et al. 2009). This standard deviation, generally referred to as sigma (σ), exerts a very strong influence on the results of probabilistic seismic hazard analysis (PSHA) ( e.g. , Bommer and Abrahamson 2006). Although there are numerous examples of sigma being neglected in seismic hazard, it is now generally accepted that integration over the full distribution of ground motions is an indispensable element of PSHA (Bommer and Abrahamson 2006). Attempts to justify, on a statistical basis, a truncation of the ground-motion distribution at a specified number of standard deviations above the median have proven unfeasible with current strong-motion datasets (Strasser et al. 2008). The most promising approach to reduce the overall impact of sigma on the results of PSHA is to find legitimate approaches to reduce the value of the standard deviation associated with ground-motion prediction equations (GMPEs). The present state-of-the-practice of seismic hazard studies applies the standard deviations from ground-motion models developed using a broad range of earthquakes, sites, and regions to analyze the hazard at a single site from a single small source region. Such practice assumes that the variability in ground motion at a single site-source combination is the same as the variability in ground motion observed in a more global dataset and is referred to as the ergodic assumption (Anderson and Brune 1999). In recent years, the availability of well recorded ground motions at single sites from multiple occurrences of earthquakes in the same regions allowed researchers to estimate the ground-motion variability …

Istanbul Earthquake Rapid Response System: Methods and practices

Potential impact of large earthquakes on urban societies can be reduced by timely and correct action after a disastrous earthquake. Modern technology permits measurements of strong ground shaking in near real-time for urban areas exposed to earthquake risk. The Istanbul Earthquake Rapid Response System equipped with 100 instruments and two data processing centers aims at the near real time estimation of earthquake damages using most recently developed methodologies and up-to-date structural and demographic inventories of Istanbul city. The methodology developed for near real time estimation of losses after a major earthquake consists of the following general steps: (1) rapid estimation of the ground motion distribution using the strong ground motion data gathered from the instruments; (2) improvement of the ground motion estimations as earthquake parameters become available and (3) estimation of building damage and casualties based on estimated ground motions and intensities. The present paper elaborates on the ground motion and damage estimation methodologies used by the Istanbul Earthquake Rapid Response System with a special emphasis on validation and verification of the different methods.

Statistical Analysis of Ground Motions Estimated on the Basis of a Recipe for Strong-motion Prediction: Approach to Quantitative Evaluation of Average and Standard Deviation of Ground Motion Distribution

For seismic hazard assessment, we study the variabilities of predicted ground motion on the basis of a “recipe for predicting strong ground motion” and propose approximations to evaluate spatial distributions of the standard deviation for PGV, R1.0, R2.0, and R5.0 in the estimated ground motions. For strong-motion prediction, we use a finite difference method for a long period range (>1.0 s). To estimate variabilities, a Monte Carlo simulation is used and we adopt the Latin Hypercube Sampling (LHS) technique to reduce computations. In this article, we consider only aleatory variabilities in source parameters among all possible variabilities, such as those in the source parameters, the propagation characteristics and site characteristics. Model sources are assumed for dip-slip fault and strike-slip fault, and the variabilities are considered for parameters such as asperity location, rupture starting point, average asperity slip contrast, stress drop and rupture velocity. On the target site, 100 instances of PGV, R1.0, R2.0 and R5.0 data are obtained for 100 sets of parameters and an average and a standard deviation of the log normal distribution, corresponding to the variability for ground motion estimation, are statistically analyzed. For all target sites uniformly distributed in the area around the faults, the average and the standard deviation are statistically analyzed and spread to spatial maps. It is found that the spatial distributions of standard deviation values for both the dip-slip and strike-slip faults are not uniform. Approximations are attempted to develop a quantitative evaluation for spatial distributions of the standard deviation of the log normal distribution for PGV, R1.0, R2.0, and R5.0. The spatial distributions by these approximations are considered to almost reconstruct the characteristics, which are statistically analyzed by the finite difference method.

Real-time Performance of the Virtual Seismologist Earthquake Early Warning Algorithm in Southern California

The Virtual Seismologist (VS) method is a Bayesian approach to regional network-based earthquake early warning (EEW) that estimates earthquake magnitude, location, and the distribution of peak ground motion using observed ground motion amplitudes, predefined prior information, and appropriate attenuation relationships (Cua 2005; Cua and Heaton 2007). The application of Bayes's theorem in earthquake early warning (Cua 2005) states that the most probable source estimate at any given time is a combination of contributions from prior information (possibilities include network topology or station health status, regional hazard maps, earthquake forecasts, the Gutenberg-Richter magnitude-frequency relationship) and a likelihood function, which takes into account observations from the ongoing earthquake. Prior information can be considered relatively static over the timescale of a given earthquake rupture. The changes in the source estimates and predicted peak ground motion distribution, which are updated each second, are due to changes in the likelihood function as additional arrival and amplitude data become available. The potential use of prior information differentiates the VS approach from other regional, network-based EEW algorithms, such as ElarmS (Allen and Kanamori 2003).

Sigma: Issues, Insights, and Challenges

The prediction of ground-motion levels at a site is one of the key elements of seismic hazard assessment. This prediction is commonly achieved using equations derived through regression analysis on selected sets of instrumentally recorded strong-motion data, hereafter referred to as empirical ground-motion prediction equations (GMPE). Reviews and compilations of equations published to date have been presented by, among others, Campbell (1985), Joyner and Boore (1988), and Douglas (2003, 2004, 2006). These equations relate a predicted variable ( Z pred) characterizing the level of shaking, most commonly the logarithm of a peak ground-motion parameter ( e.g. , PGA, PGV) or response spectral ordinate (SA, PSA, PSV, SD), to a set of explanatory variables { Xk }= X 1, X 2,... describing the earthquake source, wave propagation path, and site conditions: \batchmode \documentclass[fleqn,10pt,legalpaper]{article} \usepackage{amssymb} \usepackage{amsfonts} \usepackage{amsmath} \pagestyle{empty} \begin{document} \[ Z\_{\mathrm{pred}}=f(\{X\_{k}\})\] \end{document}(1) The explanatory variables { Xk } usually include the earthquake magnitude, M ; a factor describing the style-of-faulting of the causative event; a measure of the source-to-site distance, R ; and a parameter characterizing the site class. Recent equations sometimes also include additional terms to characterize the location of the site with respect to the rupture plane (hanging-wall factor), to distinguish between ground motions from surface-faulting events and buried ruptures, or to include the effects of sediment depth in the case of deep alluvial basins. Other factors that are known to influence the motion (and many others that are not yet known) are not included in the equation because the information is not readily available or not predictable in advance. For instance, anisotropy effects resulting from the dynamic propagation of rupture (including directivity effects) are currently not included in predictions, although back-analyses of ground motions from past earthquakes have shown that such effects may have a strong influence on the spatial distribution of ground motions. …

A Numerical Study on the Screening of Blast-Induced Waves for Reducing Ground Vibration

Blasting is often a necessary part of mining and construction operations, and is the most cost-effective way to break rock, but blasting generates both noise and ground vibration. In urban areas, noise and vibration have an environmental impact, and cause structural damage to nearby structures. Various wave-screening methods have been used for many years to reduce blast-induced ground vibration. However, these methods have not been quantitatively studied for their reduction effect of ground vibration. The present study focused on the quantitative assessment of the effectiveness in vibration reduction of line-drilling as a screening method using a numerical method. Two numerical methods were used to analyze the reduction effect toward ground vibration, namely, the “distinct element method” and the “non-linear hydrocode.” The distinct element method, by particle flow code in two dimensions (PFC 2D), was used for two-dimensional parametric analyses, and some cases of two-dimensional analyses were analyzed three-dimensionally using AUTODYN 3D, the program of the non-linear hydrocode. To analyze the screening effectiveness of line-drilling, parametric analyses were carried out under various conditions, with the spacing, diameter of drill holes, distance between the blasthole and line-drilling, and the number of rows of drill holes, including their arrangement, used as parameters. The screening effectiveness was assessed via a comparison of the vibration amplitude between cases both with and without screening. Also, the frequency distribution of ground motion of the two cases was investigated through fast Fourier transform (FFT), with the differences also examined. From our study, it was concluded that line-drilling as a screening method of blast-induced waves was considerably effective under certain design conditions. The design details for field application have also been proposed.

Seismic hazard estimation based on the distributed seismicity in northern China

In this paper, we have proposed an alternative seismic hazard modeling by using distributed seismicites. The distributed seismicity model does not need delineation of seismic source zones, and simplify the methodology of probabilistic seismic hazard analysis. Based on the devastating earthquake catalogue, we established three seismicity model, derived the distribution of a-value in northern China by using Gaussian smoothing function, and calculated peak ground acceleration distributions for this area with 2%, 5% and 10% probability of exceedance in a 50-year period by using three attenuation models, respectively. In general, the peak ground motion distribution patterns are consistent with current seismic hazard map of China, but in some specific seismic zones which include Shanxi Province and Shijiazhuang areas, our results indicated a little bit higher peak ground motions and zonation characters which are in agreement with seismicity distribution patterns in these areas. The hazard curves have been developed for Beijing, Tianjin, Taiyuan, Tangshan, and Ji’nan, the metropolitan cities in the northern China. The results showed that Tangshan, Taiyuan, Beijing has a higher seismic hazard than that of other cities mentioned above.

Truncation of the distribution of ground-motion residuals

Recent studies to assess very long-term seismic hazard in the USA and in Europe have highlighted the importance of the upper tail of the ground-motion distribution at the very low annual frequencies of exceedance required by these projects. In particular, the use of an unbounded lognormal distribution to represent the aleatory variability of ground motions leads to very high and potentially unphysical estimates of the expected level of shaking. Current practice in seismic hazard analysis consists of truncating the ground-motion distribution at a fixed number (e max) of standard deviations (σ). However, there is a general lack of consensus regarding the truncation level to adopt. This paper investigates whether a physical basis for choosing e max can be found, by examining records with large positive residuals from the dataset used to derive one of the ground-motion models of the Next Generation Attenuation (NGA) project. In particular, interpretations of the selected records in terms of causative physical mechanisms are reviewed. This leads to the conclusion that even in well-documented cases, it is not possible to establish a robust correlation between specific physical mechanisms and large values of the residuals, and thus obtain direct physical constraints on e max. Alternative approaches based on absolute levels of ground motion and numerical simulations are discussed. However, the choice of e max is likely to remain a matter of judgment for the foreseeable future, in view of the large epistemic uncertainties associated with these alternatives. Additional issues arise from the coupling between e max and σ, which causes the truncation level in terms of absolute ground motion to be dependent on the predictive equation used. Furthermore, the absolute truncation level implied by e max will also be affected if σ is reduced significantly. These factors contribute to rendering a truncation scheme based on a single e max value impractical.

Manuel Rocha Medal Recipient Wave interaction with underground openings in fractured rock

The objective of this work is to lead to improved models of seismic wave propagation around underground openings by studying the interaction of the waves with the fractured rock surrounding these openings. It demonstrates that seismic models can help in stability problems such as rockbursting in deep-level mining, or in the interpretation of micro-fracturing at waste storage sites. A significant emphasis is placed on comparing the models with observations from controlled experiments. These comparisons demonstrate that the wave propagation can be reliably and accurately modelled, and in so doing it motivates their application to the larger rock engineering problems. Seismic wave models are first applied to laboratory experiments on multiple fractures. Simulation through multiple displacement discontinuities yields strikingly similar waveforms to the experiments, while also identifying the need to build stress dependence into the fracture models, such as stress dependent fracture stiffness. The wave-fracture modelling is extended to in situ fractures in rock at the surface of a deep tunnel, using data collected during an acoustic emission experiment at the URL Mine-by tunnel. Waveforms from the velocity scans are compared against those from elastic models and various models of fracture, such as random assemblies of small open fractures (cracks) and larger fractures with fracture stiffness. Results indicate that it is possible to account for the wave-speeds and amplitudes using models with fractures. A generic method is then proposed for calculating the frequency variation of wave-speed and amplitude for any collection of cracks. The models of fracture are then applied to the rockburst problem, to investigate how the excavation affects the amplitude and the distribution of ground motion. The results provide important insights into the causes of the apparent amplification observed by researchers in this field. The thesis also covers the theory of the models used, including novel numerical work on dispersion and new grid schemes. The full detail of the work cannot be covered in this paper which instead seeks to summarize the main achievements.

Strong ground motion simulation for seismic hazard assessment in an urban area

Seismic wave propagation and strong ground motion in a number of urban settings are simulated by using the pseudospectral time domain method with staggered grid real fast Fourier transform differentiation. For urban seismic study and damage prevention, simulations of the seismic wave propagation and ground motion need a comprehensive consideration of earthquake type, building structure and three-dimensional velocity structures of the crust and especially the near-surface sediments. The distribution of ground motion in an urban area varies greatly in response to geological structures. Wave interferences between the body wave and the secondary surface wave may result in peak ground motions and high collapse ratio of buildings in urban regions far away from the seismogenic fault. The simulated waveforms using an impulse point source in a 3D velocity structure can be employed to synthesize seismograms of ground motion as Green's function for any prescribed rupture scenario and source functions. The simulated results show that the synthesized seismograms coincide well with the observed seismic waveforms for the 1995 Hyogo-ken Nanbu M7.2 earthquake, which occurred in a heavily urbanized area. The simulation Green's function method is useful for predicting the strong ground motion in an urban region without observed seismic waveforms to provide a useful strong ground motion and hazard scenario for urban planning and earthquake hazard prevention. The present technique is an important supplement to the empirical Green's function method that heavily depends on the availability of local and regional seismicity.