The ASCE/SEI 7‐16 standard contains prescriptive provisions for soil–structure interaction (SSI). However, SSI is rarely incorporated into nonlinear response history analyses, particularly in performance‐based seismic design. This paper examines the seismic response of the $4.3 billion University of California, San Francisco Health Helen Diller Hospital Building, explicitly considering SSI effects. The 258‐foot‐tall structure comprises 15 aboveground levels and two basement levels and is supported by a pile foundation system embedded in complex soil conditions. Due to its proximity to the San Andreas Fault, the design incorporates a series of viscous dampers within a structural steel special moment frame system to enhance seismic performance. This paper extends beyond the SSI requirements of the project. The study employs a parametric approach starting with a fixed‐base model as benchmark and then progressively refining the representation of the soil and massive foundation system with springs and dashpots. The study focuses on how the representation of SSI affects the design demands of the different elements in the foundation and basement, even when forces above ground remain mostly unaltered. Additionally, results demonstrate the important dynamic higher mode effects of a massive and large footprint pile foundation. The higher modes associated with SSI resulted in significant inertia forces of the basement and the foundation even for small lateral deformations. The response above ground indicates that while interstory drift remains largely unaffected, SSI significantly influences floor accelerations. Finally, the effect of different sources of energy dissipation associated with SSI such as the hysteretic behavior and the radiation damping of the pile foundations, the soil passive pressure of the retaining walls, as well as the friction between the soil and the slabs on grade are explicitly studied.
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.
This study presents a comprehensive set of earthquake ground motion cycle count datasets derived from the NGA‐West2 and NGA‐Sub databases. We applied peak counting, mean‐crossing, level‐crossing, and rainflow cycle counting methods to over 275,000 acceleration time series. Each method was applied to full and duration‐filtered time series, using thresholds based on accumulated Arias intensity (5%–95% and 5%–75%). We treated each component independently, which allows different component combination techniques to be applied. The datasets reveal that duration filtering significantly reduces the number of low‐amplitude cycles, while high‐amplitude cycles remain largely unaffected. Peak counting generally predicts the highest number of cycles, followed by level‐crossing, rainflow, and mean‐crossing the fewest, particularly in lower amplitude bins. The datasets are provided in both raw and binned formats, enabling flexible use in applications such as liquefaction triggering, structural fatigue assessment, and performance‐based earthquake engineering. The different datasets will allow researchers to select the one most appropriate for their application or site, or use multiple to account for epistemic uncertainty.
Building safety inspections can be an impeding determinant of post‐earthquake recovery, yet their durations remain inadequately quantified because fine‐grained operational data are rarely available. Here, we present a probabilistic Bayesian framework to quantify building‐level inspection time, calibrated against a comprehensive multiagency dataset from the 2010–2011 Canterbury Earthquake Sequence (CES) in Christchurch, New Zealand. Using Markov chain Monte Carlo (MCMC) inference, we represent inspection timeframes as stochastic processes and infer marginal posterior distributions that jointly characterize aleatory variability and epistemic uncertainty in post‐disaster operations. The resulting posteriors distinguish the temporal signatures of rapid building assessment (RBA) and detailed damage evaluation (DDE) protocols and reveal strongly upper‐tailed delays driven predominantly by institutional and logistical frictions beyond damage state alone. By providing empirically derived probability curves for inspection time, the framework enhances the fidelity of regional recovery simulations and offers emergency managers an evidential basis for capacity planning, resource mobilization, and inspection strategies designed to accelerate community recovery after future earthquakes.
This study investigates the effect of incorporating the Devil's Mountain Fault (DMF) and its interaction with the Leech River Valley Fault (LRVF) on the elastic surface deformation hazard in and near the city of Victoria, British Columbia, Canada. Previously, the surface deformation hazard of the region has been characterized by considering the LRVF rupture alone. Recent geological surveys show that the LRVF continues eastward and connects with the DMF. Considering only the LRVF as the main seismic source may underestimate the surface deformation hazard of the region. Hence, a probabilistic fault‐slip‐induced deformation hazard model is developed for the LRVF–DMF system. The new model incorporates a range of plausible fault lengths and uncertainties in earthquake occurrence frequencies and evaluates potential interactions between the two faults. Ground deformations are calculated using stochastic source modeling and analytical ground deformation equations. Hazard estimates are computed for the region with and without accounting for the multisegment rupture of the LRVF–DMF system to determine its influence on expected elastic surface deformation. The deformation metric considered represents the continuous elastic deformation field at the ground surface resulting from slip on the modeled fault plane. It is therefore distinct from the principal and distributed surface rupture displacements commonly considered in conventional probabilistic fault displacement hazard assessment (PFDHA) studies. The results indicate that the LRVF–DMF rupture scenario has the largest contribution to total hazard among all rupture scenarios at selected sites around Victoria. Its contribution is higher than 90% at large deformation values and increases the expected deformation by a maximum of 28 cm (from 1 cm) compared to the LRVF rupture at 0.5% probability of exceedance in 50 years.
Earthquakes continue to cause significant human and economic losses, often due to the failure of structurally vulnerable buildings that require retrofit. Performance‐Based Earthquake Engineering (PBEE) provides a rigorous framework for quantifying retrofit performance, while sustainability frameworks—life‐cycle assessment, life‐cycle cost, and emerging life‐cycle sustainability assessment—expand this evaluation to include environmental and social dimensions. However, the intersection of these frameworks has proven challenging, particularly when interacting with retrofit policy mechanisms. This review synthesizes recent advances across PBEE, sustainability assessment, retrofit technologies, and public policy, linking engineering decision‐making with governance strategies. Retrofit alternatives are benchmarked using seismic and life‐cycle criteria, while behavioral, financial, regulatory, and international policy mechanisms are analyzed according to their barriers, strengths, and applicability to engineered and nonengineered buildings. The review further distinguishes project‐level challenges faced by practitioners from policy‐level challenges faced by governments and financial institutions, showing how this multiscale perspective enables better retrofit strategies. By bridging these perspectives, the study identifies pathways for integrating PBEE, sustainability, and policy within seismic retrofit, and proposes directions for future research.
Postearthquake workflows currently support three distinct classes of decisions: safety clearance through rapid tagging, portfolio‐level economic damage estimation through probabilistic loss models, and field inspections based on locally defined survey forms, which may subsequently trigger project‐specific strengthening through retrofit design. However, commitment to executable repair work at the building level, what to repair, where, and in what quantities, remains largely ad hoc, despite being required early in recovery for the majority of lightly to moderately damaged buildings. This paper introduces the RELA Repair Module (RELA‐RM), a rapid, observation‐driven framework that translates visually observed damage into executable repair quantities through predefined decision logic and standardized repair norms. Damage is recorded at the level of individual building elements and transformed into a building‐level Repair Bill of Quantities (RBoQ) expressed in person‐hours and material quantities. Repair cost and duration are subsequently derived through externally supplied pricing and workforce assumptions, while optional qualitative escalation indications flag cases that may require detailed engineering assessment. Spatial localization of repair actions is retained throughout the workflow to preserve building‐level organization of repair scope. Uncertainty is treated as bounded variability arising from three identifiable sources within the workflow: photographic measurement effects, human annotation variability, and dispersion in standardized productivity norms. These contributions are propagated to produce an approximate combined dispersion of ≈26%, interpreted as an engineering bound on expected variability in executable repair quantities. The framework is implemented and demonstrated for masonry walls, while the underlying decision principle is intended to be extensible to other building elements and typologies.
Section 12.4.2 of the ASCE/SEI 7‐22 provisions offers two alternative methods for estimating the seismic load effects due to vertical ground motions: one based on horizontal spectral acceleration at a short horizontal period and the other uses the vertical spectral acceleration at a vertical structural period. This study examines similarities and discrepancies between the two methods for the contiguous U.S., covering Seismic Design Categories (SDCs) A through E, and vertical structural periods ranging from 0.01 to 10 s. Overall, 331,231 sites in the U.S. are analyzed in this study. A lognormal probabilistic model is developed to maintain approximate mathematical tractability of the distribution of the ratio of vertical seismic load effects computed by the two methods across the considered vertical period range for each SDC. Results indicate that, for sites located in the Western U.S. (WUS), and for vertical structural periods between 0.05 and 0.2 s, the approach based on vertical spectral acceleration generally results in higher vertical seismic load effects than the one based on short‐period horizontal spectral ordinates. For sites located in the Central and Eastern U.S. (CEUS), the short‐period horizontal spectral acceleration approach consistently yields comparable or higher estimates of the vertical seismic load effects than the one based on the vertical spectral accelerations.
Fragility and vulnerability models are key components of regional seismic risk analysis. They relate ground shaking intensity, expressed through an intensity measure (IM) such as spectral acceleration or peak ground velocity, to a probability of damage or loss. When performing a risk analysis for an entire portfolio of structures, these fragility models are often developed based on different IM types for each taxonomy, creating a problematic mismatch in IMs when looking to comparatively evaluate and collectively apply them regionally. Previous research has proposed various approaches for converting fragility functions to a common IM, including methods based on uniform hazard equivalence, total probability, and the use of proxy structures. Another uniform risk equivalence procedure is trialled here along with a more refined consideration of hazard. These methods are systematically compared via 2,260 conversion cases, assessing their ability to convert fragility functions defined with both traditional and next‐generation IMs, such as average spectral acceleration, which have yet to be explored. The conversion strategy is also extended and tested for vulnerability model conversion, which is still lacking in the literature. Results show that methods based on uniform hazard and risk equivalence are not consistently reliable, while total probability‐based approaches perform better but require detailed hazard information. A key finding is that conversions from suboptimal IMs, such as peak ground acceleration, do not yield improved predictive performance even when mapped to more optimal IMs. Practical guidance and accompanying tools are provided to support analysts in applying and testing these conversion methods for their own fragility and vulnerability models.
On June 6, 2025, an inslab earthquake of Mw 6.4 occurred in the north‐central region of Chile, near the city of Copiapó, within the Atacama seismic gap. Despite its moderate magnitude, the event produced strong ground motions with recorded accelerations exceeding 0.4 g and widespread nonstructural damage. Due to these unusual consequences, we conducted a multidisciplinary field study, collecting in situ data, which provides an opportunity to characterize both local site conditions and structural behavior in a mature seismic gap region. Shallow shear‐wave velocity profiles and predominant site periods were derived from ambient‐vibration measurements to evaluate site amplification effects in Copiapó. Damage surveys revealed widespread nonstructural damage in designated evacuation zones, significantly impacting the functionality of numerous facilities. Finally, using ambient‐vibration measurements, we estimated the structural properties of nine buildings (five damaged and four undamaged) of a residential housing complex located within the affected area. Despite their similar architecture and design, we identified systematic differences in the dynamic properties of the damaged buildings compared to the undamaged ones. On average, damaged buildings presented approximately 20% larger fundamental vibration periods and 10% higher damping ratios than undamaged buildings. The findings of this field survey provide insight into the vulnerability of buildings under moderate seismic events, whose seismic demand remains below that expected during a major earthquake.
The Italian national seismic hazard model, MPS04, was developed more than 20 years ago to provide a rational basis for regulatory applications. Since 2009, it directly provides seismic actions for structural design and assessment in the building code. MPS04 is still the official reference despite other authoritative models being developed in the meanwhile. It was also employed for research purposes and for large‐scale seismic risk analyses. Both the direct application of MPS04 for structural design and its longevity are somewhat peculiar and may thus warrant analysis. This short article retrospectively recounts: its development, the events leading to its enforcement, the implications it had on the earthquake engineering community, as well as the limitations and challenges it faced, especially due to observations following major seismic sequences that occurred in Italy since its release. The story of MPS04 is interpreted as a success one yet may have had some unintended consequences. In hindsight, on one hand, its grounding on the historical earthquake catalog, its source model based on seismotectonics, its rigorous scientific and probabilistic nature, its reproducibility in relation to its slim logic tree, and the consequent trackability, enabled gaining confidence in the main and side results. On the other hand, being the first model of its kind for Italy, and having been directly applied in codes for nearly two decades, it may have made it harder for more recent, and presumably more state‐of‐the‐art models, to take its place.
With the recent increases in measured ground motion records for large earthquakes, most notably the 2023 M7.8 Pazarcik earthquake in T & uuml;rkiye, the potential for pervasive high velocity fault rupture propagation over an extensive fault length has been observed. This shallow strike-slip event, with unprecedented near-fault strong motion instrumentation throughout a large segment of the fault rupture zone, exhibited intense fault-normal directivity pulses and fault-parallel fling step for over two hundred kilometers along the fault. Observations of near-shear (fault rupture velocity approaching the shear wave velocity) and supershear (fault rupture velocity exceeding the shear wave velocity) ruptures in recent major earthquakes, and the associated potential for highly damaging ground motions, motivate the need for deeper understanding of the relationship between high velocity ruptures and seismic risk to infrastructure systems. In this article, recent developments in high performance, regional-scale fault-to-structure simulations are applied to provide additional insight into how high velocity ruptures can impact ground motions and the demands on building structures. First, the characteristic nonlinear response of buildings to observed intense near-fault motions from the T & uuml;rkiye event are assessed, and second, fault-to-structure simulations are performed for M7 strike-slip events on the Hayward fault in the San Francisco Bay region for varying fault rupture velocities. These simulations illustrate the pronounced influence of fault rupture velocity and quantify the significant increase in regional seismic demand on structures for near-shear and supershear scenarios.
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.
Appropriately modeling seismic wave attenuation is critical for ground response analyses (GRAs), which aim to replicate local site effects observed in ground motions. Accurately predicting even small-strain site effects remains challenging because theoretical transfer functions (TTFs) from GRAs often overestimate empirical transfer functions (ETFs) when the small-strain damping ratio (D-min) is set equal to laboratory-obtained values. Previous studies addressed this by increasing D-min in 1D GRAs to account for apparent damping mechanisms that cannot be inherently modeled in 1D. These attempts have improved predictions of fundamental-mode amplitudes, but often result in overdamping at higher modes. This study investigates more direct modeling of apparent damping using 2D GRAs at four downhole array sites: Delaney Park (DPDA), I-15 (I15DA), Treasure Island (TIDA), and Garner Valley (GVDA). At each site, three numerical damping formulations were evaluated: Full Rayleigh, Maxwell, and Rayleigh Mass, each implemented with both a conventional D-min and an inflated D-min (m x D-min) obtained from site-specific calibration. Results show that the appropriate D-min multiplier (m) correlates well with the velocity contrast of a site. When applied with inflated D-min, Full Rayleigh and Maxwell damping systematically overdamped higher modes, and Maxwell damping also shifted modal peaks, particularly at higher frequencies. In contrast, Rayleigh Mass damping provided the closest match to ETFs at three sites, reducing transfer function misfit by 27%-44% relative to Full Rayleigh and 44%-49% relative to Maxwell. Rayleigh Mass damping was also computationally efficient, allowing average timesteps more than eight times larger than those with Full Rayleigh. These findings demonstrate that inflated D-min can account for unmodeled attenuation in 2D GRAs, particularly at sites with low velocity contrast, and that frequency-dependent formulations such as Rayleigh Mass damping can more effectively capture site response than traditional frequency-independent approaches, underscoring the need to re-evaluate the long-standing practice.
In many design procedures, a scalar response spectrum like RotD50 or RotD100 is interpreted as a component spectrum applied independently along two orthogonal structural directions. When a maximum‐direction measure such as RotD100 is used in this way, an implicit assumption is introduced: both structural axes experience the worst possible orientation of ground motion. This has contributed to the view that RotD100 is overly conservative for structural design. This paper clarifies that RotD100 is not inherently conservative but that it represents a different type of quantity than a typical component response. Unlike RotD50, which represents a typical directional component, RotD100 represents an instantaneous peak response occurring in some direction of the horizontal plane. The properties of different rotation‐invariant spectra are examined using a geometric interpretation in which the directional response field of ground motion is approximated by an ellipse. Selecting a scalar design spectrum can then be viewed as replacing this ellipse with a circle. Within this geometric framework, several common misconceptions regarding RotD100 can be clarified and avoided in seismic design. The paper concludes that the suitability of a scalar spectrum for design depends on how it is deployed in structural analysis. Ensuring consistency between the chosen intensity measure and the design‐action model is therefore essential for the meaningful use of rotation‐invariant spectra in structural design.
In 2015, the city of Los Angeles adopted one of the most encompassing seismic retrofitting ordinances in the United States, mandating retrofit of, among other structures, non‐ductile reinforced concrete buildings. This ordinance was heralded as a success in improving the seismic safety of building occupants and, hence, resilience for the greater Los Angeles community. However, to date, fewer than 15% of the non‐ductile buildings in Los Angeles have submitted their retrofit plans. The relatively slow progress of retrofit suggests that owners face barriers to retrofitting their buildings that warrant investigation. Accordingly, we conducted semi‐structured interviews to investigate stakeholder perceptions of enablers and barriers under the ordinance. Five primary themes emerged from the interviews: seismic safety, financial, technical, relocation, and governance. Interview participants perceived seismic safety as the primary enabler for retrofitting non‐ductile buildings; financing retrofits was perceived as the greatest barrier. Participants perceived technical issues as a barrier because engineering decisions directly influence retrofitting costs and impacts of resident relocation. Governance is also perceived as a barrier, as participants reported experiencing uncertainty while navigating the ordinance and expressed a desire for more financial incentives. The participants proposed modifications to the ordinance's design philosophy and implementation that they believe could reduce costs for building owners.
Ground velocity observations generated through a time difference of Global Navigation Satellite Systems (GNSS) phase observables and orbits have been shown to be comparable to seismic recordings without clipping during intense ground motions. Computing GNSS velocities is also computationally scalable and increases the density of ground motion observations during significant earthquakes, so their inclusion into ShakeMaps can potentially improve rapid assessments of strong ground motion. Here, we present ShakeMaps created using GNSS velocities as an additional instrumental recording in the existing interpolation scheme for a total of 16 unique earthquake events. We thoroughly analyze four of these earthquakes (2016 M6.6 Norcia, Italy; 2019 M7.1 Ridgecrest, California; 2016 M7.8 Kaikoura, New Zealand; and 2020 M6.5 Challis, Idaho) with the goal of showcasing the broad application for a variety of earthquake rupture characteristics and regional instrument coverage. We validate all ShakeMap instances against published results that currently use only seismic instrumental recordings in their interpolation scheme. For both our approach and published results, we compare the instrument-derived ground motions and the ShakeMap-inferred ground motions directly. In our approach, we find the geodetic stations exhibit a standard deviation of 0.36 natural log units while the seismic stations are not degraded in performance. Additionally, where collocated geodetic and seismic observations exist, we quantitatively assess any biases between the geodetically derived velocity waveforms and the seismically derived velocity waveforms. Lastly, we assess the optimal weighting within the interpolation for intensity observations produced with peak ground accelerations (PGA), peak ground velocities (PGVs) from seismic instruments and geodetic instruments. This article shows that including GNSS velocities can further constrain the interpolation schemes used in ShakeMaps and potentially lead to a better constraint on ground motions and shaking intensity realized for different earthquakes observed by real-world irregular sensor network geometries.
In this paper, the physics-based numerical simulation of the strongest instrumental earthquake in the recent seismic history of Italy, i.e., the Nov 23, 1980 Irpinia earthquake in Southern Italy, with magnitude M W 6.8, is presented and discussed. A 3D spectral element model covering an extended region (147 & times; 110 km2) was constructed, with a frequency resolution up to 2 Hz, encompassing a rather complex multi-segment kinematic fault rupture model and a crustal velocity model including the ground topography. Although a strict validation of the numerical model is not possible, owing to the limited data available for this historical earthquake, simulations are found to capture the main features of the spatial and temporal variability of ground motion realistically, as testified by the quantitative comparison with the available recordings and with the macroseismic intensity observations. The results are investigated to gain insights into the key features of near-source ground motion during normal-fault earthquakes, including the spatial variability of the permanent ground deformation field, as well as amplitude, period content, azimuth, and polarization (fault normal vs. fault parallel) of pulse-like signals caused by up-dip directivity effects.
Conventional earthquake impact assessments often rely on broad assumptions and simplified models, which introduce significant uncertainty into building damage predictions. Although recent advances in machine learning, detailed building inventories, and low-cost sensors have created opportunities for more refined analyses, these tools have not yet been comprehensively evaluated at the portfolio scale. To address this gap, this study presents a digital twin framework validated through comprehensive virtual experiments. The proposed system is designed to integrate ML-based damage data with real-time sensor data, thereby reducing the bias and uncertainty inherent in traditional methods. The approach is demonstrated through the development of a detailed digital twin of the Alvalade parish in Lisbon, Portugal, where advanced simulations were used to evaluate building responses under earthquake scenarios. Sensor placement strategies were also examined to further minimize uncertainty. Results from this computational validation demonstrate the theoretical potential for digital twins to substantially improve damage assessment accuracy, providing a framework for future real-world implementation and operational deployment.
Fire sprinkler systems are critical for fire safety, but their functionality can be impaired by seismic damage. This study develops new seismic fragility functions for acceleration-sensitive damage requiring monitoring, maintenance, and repair in code-compliant fire sprinkler piping systems, using experimental data from full-scale experimental tests. These new fragility functions benchmark existing FEMA P-58 fragility functions. Three estimation methods, namely, Method of Moments (MM), Sum of Squared Errors (SSE), and Maximum Likelihood Estimation (MLE) were applied, with MLE-derived parameters selected for the final fragility functions. Statistical analysis of no-leakage cases revealed that several of the FEMA P-58's Damage State 1 (minor leakage) fragility functions are conservative compared to experimental evidence. This study proposes updated fragility functions comprising: (a) a new experimental fragility function for the Minor Repair damage state; (b) data-informed adjustments to existing judgment-based FEMA P-58 functions for minor leakage; and (c) the original FEMA P-58 judgment-based functions. The proposed fire sprinkler fragility functions are expected to improve the accuracy of post-earthquake assessments of building re-occupancy and functional recovery times.
