This article investigates the very large ground-motion amplitudes of peak ground acceleration (PGA) and short vibration periods on soft soil sites in high-seismicity regions from the 2022 New Zealand National Seismic Hazard Model (NZ NSHM). For example, a PGA of 1.36 g is predicted for the 2% in 50 year exceedance probability at a site with V-S30 = 225 m/s in Wellington, New Zealand. Such values are in the realm of the largest globally-recorded ground motions on soft-soil sites. In this context, the article examines the amplitudes of these NSHM-based ground-motion amplitudes against historical observations, and scrutinizes the treatment of nonlinear site amplification adopted in common ground-motion models (GMMs). Through such an examination, we form the opinion that, while the design ground motions can physically occur, there are multiple aspects that likely lead to overestimations, particularly for PGA. The most notable is the use of the equivalent-linear method to constrain nonlinear site response for high-intensity ground motions and associated shear strains, well beyond those for which this method is suitable. We propose adjusted nonlinear functions for PGA, for soft-soil sites with V-S30 < 300 m/s, based on nonlinear simulations for New Zealand sites, which more rigorously consider effects of nonlinear soil behavior. The adopted nonlinear functions were subsequently implemented into GMMs used in the NZ NSHM, and it was illustrated that they result in a decrease in PGA hazard, which becomes more pronounced with increasing ground-motion intensity. For example, a reduction in PGA of approximately 25% is predicted for the highest-hazard regions of New Zealand, such as Wellington, at the 2% in 50 year exceedance probability. This study highlights the important impacts of soil nonlinearity on PGA for soft-soil sites in high seismic hazard regions, and the need to more rigorously consider this modeling aspect in future GMM development.

Accurate ground motion (GM) estimates are essential for forensic analysis of structural damage following major earthquakes when direct recordings at the location(s) of interest are unavailable. Contemporary post-event GM estimation methods often leverage nearby observations to constrain estimates of intensity measures (IMs); however, existing approaches rely on empirical ground-motion models with well-known limitations in capturing spatial dependencies. This study introduces a graph neural network (GNN) approach for estimating ground-motion IMs, leveraging a graph-based representation to naturally encode spatial dependencies and allow for different observation types. Applied to a New Zealand (NZ) case study, the GNN achieves performance in line with the established multivariate normal conditional IM method, while learning spatial correlations directly from the data. These case study results illustrate the viability of GNNs for post-event GM estimation, while the data- and graph-based approach offers inherent advantages, such as support for any IM type and straightforward extensibility to additional observation types, for example, macroseismic intensity. Continued improvements in model architecture and increased data availability are expected to further enhance GNN performance and applicability for this problem.

This paper illustrates opportunities for the use of simulated ground motions in seismic performance assessment - from seismic hazard, ground-motion selection, response history analysis, and seismic demand hazard calculation. The illustration is through a comparative analysis with the conventional alternative of observed ground motions. Three different approaches for using simulated/observed ground motions are identified, and the impact of these alternative approaches on seismic risk is examined via the response of a 12-story structure (located at a site in the South Island, New Zealand) subjected to 20 ground motions selected for seven intensity levels. Structural response is quantified via peak floor acceleration and peak inter-story drift ratio and seismic risk is compared and contrasted in terms of demand hazard and collapse risk. The results indicate that the manner in which the seismic hazard is computed is the defining factor in the seismic risk results. That is, the largest differences were obtained between the cases where seismic hazard was obtained using conventional empirical ground-motion models as compared to ground-motion simulation-based seismic hazard analysis. There was a relatively smaller effect of ground-motion selection using either simulated or observed ground motions, when the target intensity measures that the ground motions were selected to match are based on seismic hazard using empirical ground motion models. This highlights that both (i) simulated ground-motions are a simple substitute for observed ground motions for scenarios which are poorly represented in observational ground motion databases; and (ii) seismic hazard analysis using simulation- and empirical-based ground-motion models can exhibit appreciable differences, and improvements in both types of methods is a valuable opportunity for further research.

This paper compares the responses of two 3D building models subjected to recorded and simulated ground motions following New Zealand (NZ) code-based ground-motion selection and scaling provisions to examine the applicability of simulated ground motions for use in conventional engineering practice in NZ. The buildings were designed according to NZ building codes and physically constructed in Christchurch prior to the 2010-2011 Canterbury earthquakes. 40 recorded ground motions from the 22 February 2011 Christchurch earthquake, along with previously published simulated ground motions for this event were considered. The seismic responses of the structures are principally quantified via the peak floor acceleration and peak inter-storey drift ratio. The peak floor accelerations of both buildings, and peak drifts of the 13 storey in-plan regular RC wall structure, were statistically consistent; whereas the peak drifts for the seven storey in-plan irregular mixed system are approximately 30% different. Overall, the results indicate a general agreement in seismic demands obtained using the recorded and simulated ground motion ensembles, and hence provide further evidence that state-of-the-art simulated ground motions can be used in code-based structural performance assessments in place of, or in combination with, ensembles of historically recorded ground motions.

ABSTRACT The Wellington central business district (CBD) in New Zealand overlies a complex sedimentary basin that significantly amplifies surface ground motions, as observed during events like the 2016 Mw 7.8 Kaikōura earthquake. This article presents a study on site amplification in Wellington predicted through 3D physics-based ground-motion simulations. Eight alternative models of the Wellington basin geometry are used to quantify predicted site amplification and the between-model epistemic uncertainty associated with basin representation. Response spectral amplification factors (for periods 0.5–10 s) are shown at several locations across the Wellington CBD, along 2D transects, and in spatial maps. When compared with observed site amplification, the simulations capture broad spectral amplification peaks and provide comparable levels of site amplification at some deeper basin sites. At shallower basin sites, the simulations have discrepancies in the predicted level and period dependence of site amplification, partially due to model spatial resolution and minimum S-wave velocity (500 m/s for this study) limitations. Linear simulated spectral amplification factors are as high as 8 in some locations. These results illustrate future opportunities for improving the simulations, which include better constrained basin-depth geometry and velocity characterization of the sediments within the basin; higher spatial-resolution simulations; decreasing the minimum S-wave velocity in the simulations; and near-surface nonlinear site response modeling using physics-based approaches.

This study presents the computational components and workflow developed to conduct probabilistic seismic hazard analysis (PSHA) in New Zealand (NZ) using physics‐based ground‐motion simulations, referred to as “CyberShake NZ.” Semi‐automated procedures generate multiple realizations of kinematic rupture models, create rupture‐specific velocity models that optimize the simulation domain dimensions, submit and monitor forward wave propagation jobs on high‐performance computing facilities, and postprocess the results. A nationwide grid of 25,948 surface receivers, with non‐uniform geographic density reflecting local subsurface conditions and population density, is used to extract simulation results at consistent locations and aggregate them into site‐specific hazard results. We summarize 18 executions of the workflow over an 8‐year period and provide example results from a nationwide execution adopting the Graves and Pitarka (2010, 2015, 2016) hybrid broadband ground‐motion simulation approach with a 0.2 km computational grid (giving a 0.5 Hz transition frequency between the low‐ and high‐frequency simulations) and an empirically calibrated local site response model. A Monte–Carlo scheme samples hypocenter location, rupture magnitude, and kinematic rupture realizations to partially account for ground‐motion variability, with the number of realizations for each fault scaled by rupture magnitude. A total of 16,043 finite‐fault simulations are undertaken; distributed seismicity is treated via conventional empirical ground‐motion models in the current execution. Results are presented as uniform hazard ground‐motion maps, seismic hazard curves, and disaggregation. Ongoing developments and planned improvements toward greater explicit use of physics‐based simulations in PSHA are also outlined. We emphasize that these simulation‐based results are a proof‐of‐concept: Continued validation against historical observations and comprehensive treatment of modeling uncertainties are required before nonzero logic tree weights can be assigned in practice.

V S 30 ‐based site amplification factors are commonly used to account for local site effects in hybrid broadband ground‐motion simulation when site‐specific data are limited. Validation against observations allows for testing alternative semi‐empirical V S 30 ‐based models and establishing proper protocols for their application in forward analyses. This article validates alternative implementations of V S 30 ‐based site factors using 5218 ground motions from 479 small‐magnitude events recorded at 212 sites in New Zealand, which represent a wide range of site conditions. The investigation of (Fourier) effective amplitude spectra (EAS) residuals allows for a direct assessment of the low‐frequency (LF; f < 1 Hz) and high‐frequency (HF; f > 1 Hz) components of the hybrid broadband simulation approach, and the corresponding effect of the site “adjustment.” Before applying the site factor, the LF simulation exhibits systematic underprediction and the HF simulation displays systematic overprediction. The use of site factors based on semi‐empirical models for EAS, developed in the last decade, results in comparable, or slightly better, prediction performance compared with response spectra‐based models commonly used in previous validation studies. A greater LF underprediction was identified at sites located within sedimentary basins that are less constrained by site data or not properly modeled in the LF simulation due to limitations in spatial discretization. A simple protocol is proposed for the application of the LF site factor based on a basin‐type and geomorphic site categorization, which reduces the variability of EAS site‐to‐site prediction residuals by up to 0.1 natural log units. The explicit incorporation of the parameter Z 1.0 , or alternatively, the use of a host‐to‐target correction factor demonstrates that a significant portion of the HF systematic overamplification in the frequency range 1.0–3.0 Hz can be explained by V S profile features not accounted for by V S 30 alone. These findings can inform the application of site adjustments in future studies and guide advancements in the LF and HF simulation components.

V-S30-based site amplification factors are commonly used to account for local site effects in hybrid broadband ground-motion simulation when site-specific data are limited. Validation against observations allows for testing alternative semi-empirical V-S30-based models and establishing proper protocols for their application in forward analyses. This article validates alternative implementations of V-S30-based site factors using 5218 ground motions from 479 small-magnitude events recorded at 212 sites in New Zealand, which represent a wide range of site conditions. The investigation of (Fourier) effective amplitude spectra (EAS) residuals allows for a direct assessment of the low-frequency (LF; f < 1 Hz) and high-frequency (HF; f > 1 Hz) components of the hybrid broadband simulation approach, and the corresponding effect of the site "adjustment." Before applying the site factor, the LF simulation exhibits systematic underprediction and the HF simulation displays systematic overprediction. The use of site factors based on semi-empirical models for EAS, developed in the last decade, results in comparable, or slightly better, prediction performance compared with response spectra-based models commonly used in previous validation studies. A greater LF underprediction was identified at sites located within sedimentary basins that are less constrained by site data or not properly modeled in the LF simulation due to limitations in spatial discretization. A simple protocol is proposed for the application of the LF site factor based on a basin-type and geomorphic site categorization, which reduces the variability of EAS site-to-site prediction residuals by up to 0.1 natural log units. The explicit incorporation of the parameter Z(1.0), or alternatively, the use of a host-to-target correction factor demonstrates that a significant portion of the HF systematic overamplification in the frequency range 1.0-3.0 Hz can be explained by V-S profile features not accounted for by V-S30 alone. These findings can inform the application of site adjustments in future studies and guide advancements in the LF and HF simulation components.

This study quantifies the influence of multidimensional effects on site response in the Wellington region of New Zealand by comparing results from different simulation approaches. Response-spectral amplification functions from 1D and 2D site response analyses are compared to quantify the magnitude and spatial distribution of multidimensional effects, which are significant at some locations. The 1D and 2D results are also compared to observations at strong motion stations. Simulated surface-to-borehole transfer functions are extracted from 1D and 2D site response models to show their sensitivity to subsurface stratigraphy and surface topography. Comparisons with observed spectral amplification indicate that the 1D and 2D models capture key spatial trends and, on average, are within 20-30% of the simulated values over the full period range. While results from the two approaches are not significantly different at most instrumented sites, the 2D analyses suggest that the high observed amplification at moderate to long periods is likely influenced by multidimensional effects at some sites. Topographic and stratigraphic features that influence site response are identified, and these features are shown to cause significant differences between 1D and 2D transfer functions over a large spatial extent, consistent with recent global studies demonstrating that many sites do not exhibit surface-to-borehole transfer functions consistent with 1D assumptions. In contrast, the differences in response-spectral amplification between 1D and 2D analyses are smaller than those for within transfer functions, suggesting that care should be taken to avoid over-inferring the performance of 1D site response modeling techniques using only transfer functions from vertical arrays.

Proper incorporation of shallow site effects in hybrid broadband ground-motion simulation is essential to improving predictions at soil sites. This paper validates and compares four methods to account for these effects in the predominantly linear regime, using 1446 ground motions from 213 small-magnitude earthquakes ( 3 . 5 <= M W <= 5 . 0 ) recorded at 38 sites in New Zealand representative of a wide range of soil conditions. These methods require V S 30 or a shear-wave velocity profile, with the second approach relying on either the square-root-impedance method or the theoretical one-dimensional (SH1D) transfer function. Incorporating shallow site effects with any of the considered methods significantly improves accuracy and precision for several intensity measures, including pseudo-spectral acceleration at short vibration periods. Although the reduction in prediction variability was comparable between methods, the SH1D-based approach showed particular benefits in the case of (1) stiff sites for which the simulation overestimates attenuation effects; (2) relatively stiff sites with high-frequency resonances; (3) relatively soft sites with strong resonances in the intermediate frequency range; and (4) sites with a near-surface predominant frequency lower than the frequency range considered for the application of the V S 30 -based adjustment. This study shows that the potential prediction improvements obtained by using more advanced site-adjustment methods that incorporate additional site-characterization data are highly dependent on the site characteristics and the quality of the ground-motion simulation in a given region. These findings can help to inform site response method selection in forward applications.

This paper revisits the sentiments expressed in a 2015 paper published by the author regarding site-specific seismic hazard analysis in New Zealand (NZ) [1]. While many of the general principles expressed remain the same, the completion of the 2022 NZ National Seismic Hazard Model (NSHM), and accompanying Draft Technical Specification TS1170.5:2024 have significantly altered how such analyses are performed in NZ, and the incremental value that they can provide beyond a code-based approach. This paper identifies instances where site-specific analyses remain valuable (and where they do not), where they will likely depart from the 'baseline' 2022 NZ NSHM (and TS1170.5:2024) results to a practically significant degree, and past practices for seismic source and ground-motion modelling choices which are now considered unviable. Lastly, challenges for practitioners and researchers are briefly addressed in order to further advance the practice of site-specific seismic hazard analysis over the next decade.

This paper examines the manner in which near-fault ground-motion phenomena are considered in the probabilistic seismic hazard analysis underpinning the 2022 New Zealand National Seismic Hazard Model (NZ NSHM), and its subsequent codification in the draft Technical Specification TS1170.5:2004. Directivity is already implicitly considered in the 2022 NSHM, and thus the ‘baseline’ draft TS1170.5 spectra, before any additional near-fault factor is applied. Specific studies in NZ and California suggest that explicit modelling of directivity could result in increases of up to 15-20% for 2475-year return period SA(T = 3:0s) values for ‘directivity-prone locations’, but are more likely to be on the order of 10% when a weighted average of multiple directivity models is considered. In contrast, the NZS1170.5:2004 near-fault factor, N(T;D), results in a 36% increase for small source-to-site distances. Hence, either: (1) the NZS1170.5:2004 near-fault factor should be removed so that directivity is implicitly considered in the hazard, which is consistent with other international codes, and also consistent with the implicit treatment of deep sedimentary basin effects in the 2022 NSHM, or (2) a parametric revision of the near-fault factor is needed that results in approximately a four-fold reduction in its size. Obtaining a more precise quantification of directivity effects is complicated by multi-segment ruptures in contemporary seismic source models, and the large uncertainty in predicted directivity modification factors from the alternative existing models.

Shallow site effects are indirectly considered in conventional hybrid broadband ground-motion simulations, and their proper incorporation may be key to improving ground-motion predictions at soil sites. This article presents and examines five methods to adjust hybrid simulations to account for these effects. These methods use different approaches to modeling site response and require different amounts of site-characterization data: Methods 1 and 2 use only proxy parameters (e.g. [Formula: see text], [Formula: see text]) to describe the site conditions, with Method 1 relying solely on proposed site response scaling factors in existing semi-empirical ground-motion models, and Method 2 incorporating a host-to-target velocity-profile adjustment; Methods 3 and 4 use a shear-wave velocity profile along with two different frequency-domain approaches to predict the linear site response, coupled with the nonlinear component of Method 1; and Method 5 uses 1D time-domain nonlinear site-response analysis and generally requires additional data to constrain nonlinear constitutive-model input parameters. The five methods are applied to four sites subjected to two levels of ground-motion intensity to illustrate the challenges involved in their implementation and to compare and contrast the resulting adjustments in terms of site amplification. The suitability and performance of each method will depend on multiple factors, including the availability of site-characterization data and complexity of the site, and hence, their relative advantages and disadvantages are discussed considering a broad range of scenarios.

This paper discusses the development of an adjustment factor for PGA from the 2022 update of the New Zealand (NZ) National Seismic Hazard Model (NSHM2022) and its implementation in the NZ technical specifications TS1170.5:2024 (TS1170). The study focuses on soft soil sites with VS30 ≤ 300 m/s (i.e., Site Classes IV, V, and VI in TS1170). The adjustment factor is based on nonlinear site-response simulations of NZ characteristic soft soil sites and an examination of observations from extensive national and global ground-motion databases. These simulations treat soil nonlinearity more rigorously than the approximations used in the empirical ground-motion models employed in NSHM2022. The scientific background and details of the analyses used to develop the PGA adjustment factors are documented in de la Torre et al. [1], while the focus of this paper is the parametrisation of the proposed adjustment factor for implementation into TS1170. The adjusted PGAs are compared to the PGAs obtained directly from NSHM2022, and the PGAs from the 2004 NZ seismic loading standard NZS1170.5:2004 for many cities. Application of the adjustment factor results in a reduction to PGA for all three site classes. The amount of reduction increases with increasing intensity of the NSHM2022 predicted ground motion (i.e., PGA), resulting in approximately 15-25 % reduction to the 2500-year return period PGA in the highest hazard regions of NZ. However, even with the proposed reduction factors, compared with NZS1170.5:2004, the adjusted PGAs in these high-hazard regions are still 40-50 % higher for the 500-year return-period ground motion.

Physics-based simulation of subduction earthquake ground motions remains less comprehensively validated than for shallow crustal earthquakes, despite subduction events contributing significantly to global seismic hazard. In this study, subduction-specific simulation models were developed and validated for small-magnitude ( M w 3.5-5) interface and slab earthquakes using hybrid broadband ground-motion simulation. Simulation models were constrained by (1) global empirical ground-motion models, (2) a global database of finite-fault rupture models, and (3) global ground-motion simulation studies. The simulation models include subduction source-specific representations for stress parameter and rupture velocity, with a significant depth dependence of the stress parameter for slab earthquakes. Volcanic backarc effects on anelastic path attenuation are also included through path attenuation-based scaling of the rock quality factors in the simulations. The simulations leverage models for site and basin effects which have been validated using crustal earthquakes, for which there are many recorded ground motions, and the subduction-specific model modification mainly affect the high-frequency component of the ground-motion simulations above 1Hz. Simulation predictions were validated against a compiled dataset of observed subduction earthquake ground-motion records in New Zealand. The subduction-specific modifications significantly improve predictive performance compared to the use of simulation parameter values for active shallow crustal earthquakes and provide comparable accuracy to prior simulations for crustal earthquakes in New Zealand. Despite these promising results, further studies are needed to address prediction residuals at low frequencies (f <= 1 Hz), as well as extending validation to larger magnitude earthquakes.

This paper outlines the consideration of deterministic limits on maximum ground motion levels within seismic design codes and standards. The specific motivation is to outline the basis for the exclusion of such a limit in the 2024 draft Technical Specification for NZS1170.5 [1], despite the presence of such limits in NZS1170.5:2004 [2]. An overview of the historical consideration of so-called ‘deterministic’ and probabilistic seismic hazard analysis methods is provided, as well as how they have translated into contemporary seismic design codes and standards in New Zealand (NZ) and internationally. The fundamental issues with deterministic maximum limits are outlined through the use of examples in a NZ-specific context. The underlying reason ‘well above average’ ground-motion intensity levels (for a given earthquake scenario) are prevalent in regions of high seismicity is discussed, as well as other common misconceptions that lead to the use of deterministic limits to achieve apparently realistic design ground motion intensities. Finally, in the vein of the hazard-risk separation principle, sentiments are expressed for achieving economic and resilient seismic design in regions of high seismicity without resorting to implementing deterministic limits.

Anterior cruciate ligament (ACL) injuries are a high injury burden in football, with a notably higher incidence rate among women compared to men. However, epidemiological studies are limited in women’s football so is presently no statistically rigorous or fully quantified context for understanding this difference. This study utilizes public news, blog, and social media data to create a statistical framework for analysing ACL injury rates in women’s football to enable better comparisons and understanding of the probability of the different numbers of injuries per team per season. Employing Poisson distributions, which are widely employed to assess the incidence of discrete events whose occurrence is relatively rare, we are able to quantify ACL injury incidence and compare it across different top women’s football leagues from various counties, and against the English Premier League for men where other longer-term data was available. The analysis includes calculating the raw mean and median injury rates, constructing confidence intervals, and evaluating the number of additional injury-free games and fewer injuries per season (for a league) required for women’s football to match the men’s injury rates. The results reveal significant differences in ACL injury rates between men’s and women’s football with an overall difference of 2.5x (p < 0.05), as expected, with unexpected variability of almost 4x in rates observed across the 12 women’s elite or premier leagues. The Swedish, Netherlands, and Mexican women’s leagues show no statistically significant difference in incidence rates to the men in this comparison though the raw data incidence is slightly higher. These insights offer an empirical foundation for targeted research into causes of these differences. Study limitations highlight the need for improved data collection and reporting practices. The overall study results underscore the importance of applying more rigorous statistical methods to this problem to better target inquiry and to assess future interventions.

The Next Generation Attenuation (NGA)-West3 Program database builds upon that of NGA-West2 for shallow crustal earthquakes in active tectonic regimes to provide a robust data set to develop the next iteration of NGA ground motion models (GMMs). Researchers from Italy, Japan, New Zealand, Taiwan, United Arab Emirates, and the United States, amongst others, have collaborated to develop consistently processed data with uniform metadata from the respective regions, with data for other regions drawn from literature (e.g., Greece, Japan, and Turkiye). Over 80,000 three-component ground motions from 632 events with magnitudes generally greater than 4.0 across the Western United States (mostly in California) have been newly added, and the total database size is over 175,000 ground motions (generally three-component except for two-component records from KiK-Net stations in Japan). The database is being used by several teams of NGA GMM developers and will be publicly released sometime in 2025 when the NGA-West3 GMMs have been drafted.

ABSTRACT We provide an overview of the treatment of site effects in the New Zealand National Seismic Hazard Model (NZ NSHM), including a case study of basin effects in central Wellington. The NZ NSHM 2022 includes a change in site parameter from subsoil class (NZS class) to VS30. Poor NZ VS30 characterization is a major source of uncertainty in the NSHM; however, advanced site characterization in Wellington allows for in-depth study. First, we construct a regional 3D shear-wave velocity model and maps of site parameters (T0, NZS class, and VS30) for central Wellington. At central city soil sites, we find the ratios of NZ NSHM 2022 hazard spectra with respect to the current equivalent design spectra range from factors of ∼0.8–2.6 (median ∼1.5), depending on local site conditions and spectral period. Strong amplification peaks at 0.5–2 s are observed in central Wellington. Linear site-specific amplifications from multiple methods are compared at 13 stations and are well-defined by both site-to-site residuals and response spectral ratios relative to station POTS. At many deeper soft sites (VS30<300 m/s), strong amplification peaks occur around T0 that are underpredicted by mean ergodic ground-motion model (GMM) predictions. This underprediction is slightly enhanced when using basin-specific Z1.0 as an additional site parameter. Our study highlights outstanding challenges in modeling strong basin response within shallow basins in NSHMs, including the need to consider region- or basin-specific modeling approaches as well as nonlinear effects at high shaking intensities that dominate the hazard. For New Zealand, in general, as illustrated in the Wellington case study, a priority is the further characterization of VS30 (and VS) for the seismic network to better isolate and quantify uncertainties in seismic hazard and allow useful exploration of regional–GMM adjustments and partially nonergodic approaches.