Earthquake Hazard Estimation Research Articles

Earthquake spatial probability and hazard estimation using various explainable AI (XAI) models at the Arabian peninsula

Earthquakes are the most destructive natural hazards because of their adversely severe impacts on urban areas. Earthquakes affect people's lives and properties, thus captivating the extensive attention of seismologists. Carrying out probability and hazard assessment for the prevention, and reduction of mega-events and recovery will be of great significance in affected areas. Given that limited studies have attempted to estimate earthquake Spatial Probability Assessment (SPA) in the Arabian Peninsula, this study aims to evaluate the SPA and Earthquake Hazard Assessment (EHA). This study implements and evaluates various machine learning and explainable-AI (XAI) techniques for the estimation of SPA and EHA in the Arabian Peninsula, explores the contribution and highlights the importance of different factors in the development of AI-based models. A total of twelve factors ranging from seismological to geophysical factors were evaluated. Two machine learning models namely Light Gradient Boosting Machine (LightGBM) and deep Recurrent Neural Networks (RNN) along with three XAI approaches (i.e, Smart predictor, Smart Explainer and Local Interpretable Model-Agnostic Explanation (LIME) model) were investigated. Results of the comparative earthquake SPA estimation demonstrated that the accuracy of 89% and 87% were achieved by LightGBM and RNN models. Moreover, the results of the XAI models show that the Smart Predictor provides better spatial outputs than the other evaluated XAI models. The stable factors identified by Smart Predictor were magnitude variation and earthquake frequency whereas the important factors were magnitude variation, earthquake frequency, depth variation, and seismic gap. Collectively, results of SPA show that, the Gulf of Aden, Red Sea, Iran, and Turkey are falling under a very-high SPA index (0.991–1). Correspondingly, Gulf areas, coastal areas of Saudi Arabia, and areas in the Zagros fault and Anatolian fault zone fall under a very-high hazard zone. This research could support planners, and decision-makers for emergency planning, infrastructure development, and reconstruction projects.

An Earthquake Early Warning Method Based on Bayesian Inference

Earthquake early warning (EEW) systems can give people warnings before damaging ground motions reach their site and can reduce earthquake-related losses. It has been shown that Bayesian reference can be used to estimate earthquake magnitude, location, and the distribution of peak ground motion using observed ground-motion amplitudes, predefined prior information, and appropriate attenuation relationships. In this article, we describe a Bayesian approach to earthquake early warning (EEW) systems for the estimation of magnitude (M) and peak ground-motion velocity (PGV) using the Gutenberg–Richter relation as a priori probability derived from statistical historical earthquake information from the Japanese KiK-net. Moreover, using the magnitude Mj = 4.5 and PGV = 0.4 cm/s as the warning threshold, an earthquake hazard discrimination model was established. Following this, the proposed approach was compared with the traditional fitting method to analyze the earthquake hazard discrimination using an Mj = 4.3 earthquake in the Mt. Fuji area and an Mj = 5.5 earthquake in the Ibaraki area of Japan as examples. The results are as follows: (1) The probabilistic algorithm founded on the predictive model of the magnitude from the average period (τc) and PGV from the displacement amplitude (Pd) from the initial 3 s of the P wave is able to provide a fast and accurate estimation of the final magnitude and location. (2) The Bayesian inference approach for the earthquake hazard discrimination model has a 2.94% miss rate and a 5.88% false alarm rate, which is lower than that of the traditional fitting method, thus increasing the accuracy and reliability of earthquake hazard estimation.

The role of fault interaction in earthquake migration in Central Sulawesi, Indonesia

Over the past 50 years, a sequence of destructive earthquakes, such as the 1968 (Mw 7.2), 1996 Tonimi (Mw 7.9), and 2018 Palu (Mw 7.5) earthquakes, occurred on the Palu-Koro strike-slip fault and the Minahassa Megathrust, which caused catastrophic disasters in Central Sulawesi, Indonesia. To explore the relationship between fault interaction and earthquake migration in this target area, we simulated the Coulomb stress change (ΔCFS) caused by 13 historical large earthquakes (M ≥ 6.0) since 1968. We show that the 1968 earthquake (Mw 7.2) on the Palu-Koro Fault promoted the 1996 earthquake sequence (Mw 7.9, Mw 6.6, and Mw 7.0) on the Minahassa Megathrust by transferring positive stress to its western edge with a maximum ΔCFS value of 124 kPa before 1996. Consequently, the 1996 earthquake (Mw 7.9) promoted six large earthquakes (M ≥ 6.0) including the 2018 Palu earthquake (Mw 7.5) on the Palu-Koro Fault between 1998 and 2020 via stress loading over the entire Palu-Koro Fault with a maximum ΔCFS value of 76 kPa. Via a positive feedback, these six earthquakes have increased stress at the western edge of the Minahassa Megathrust with a maximum ΔCFS value of 265 kPa up to 2020, resulting in increased seismic and consequent tsunami hazards in this region. Based on these risks, close attention should be paid to the future seismic hazards in Central Sulawesi. These findings suggest that the Palu-Koro Fault and the Minahassa Megathrust constitute a coupled fault system in which seismicity in either fault might bring the other closer to failure. Our results provide new insights into the causes of earthquake generation, migration, and hazard estimation in Central Sulawesi, and provide a basis for understanding stress transfer and seismic activity in strike-slip and thrust fault systems globally.

Nowcasting-Based Earthquake Hazard Estimation at Major Cities in New Zealand

Nowcasting-Based Earthquake Hazard Estimation at Major Cities in New Zealand

Uncertainty analysis of strong earthquake hazard estimation based on the generalized Pareto distribution model: A case study of the northeastern Tibetan Plateau

A statistical method to analyze the uncertainties of strong earthquake hazard estimation is proposed from Generalized Pareto distribution (GPD) model using the northeastern Tibetan Plateau earthquake catalogue (1885–2017) data. For magnitude threshold of 5.5, the magnitude return levels in 20, 50, 100, 200, and 500 years are 7.19, 7.70, 7.99, 8.22, and 8.45, respectively. The corresponding 95% confidence intervals are [6.77, 7.60], [7.27, 8.12], [7.53, 8.44], [7.71, 8.72], and [7.84, 9.06], respectively. The upper magnitude limit obtained from this GPD model is 9.07 and its 80% confidence interval is [8.16, 9.98]. The sensitivity analysis by the Morris method indicates that the input magnitude threshold has a relatively large influence on the estimation results. Thus, threshold selection is important for the GPD model construction. The sensitivity characteristic ranking of input factors become increasingly stable with the increasing of return period, which implies that GPD model is more suitable for estimating strong earthquakes magnitude return levels and upper magnitude limit. The GPD modeling approach and qualitative uncertainty analysis methods for strong earthquake hazard estimations proposed in this paper can be applied to seismic hazard analysis elsewhere.

Quantifying the current state of earthquake hazards in Nepal

Quantitative estimates of present-day earthquake hazard in major cities are essential for effective policymaking, community development, and seismic risk reduction. In this study, we develop a statistical analysis of natural times in Nepal to compute earthquake potential score (EPS) that describes the current level of seismic progression of a city through irregular repetitive cycle of regional earthquakes. The method, known as earthquake nowcasting (Rundle et al., 2016), uses a discrete time domain of natural times, cumulative counts of small interevent earthquakes, to characterize the present state of fault system by way of considering all earthquake events, including dependent, induced, or triggered seismicity. Data analysis and statistical inference of natural times corresponding to M ​≥ ​6 events assign EPS values between 59% and 99% to 24 major cities of Nepal, with the scores of metropolitan areas Kathmandu (95%), Pokhara (93%), Lalitpur (95%), Bharatpur (93%), Biratnagar (92%), and Birganj (93%). Physically, these nowcast scores, viewed as a way of tectonic stress accumulation since the last event, provide a realistic estimate on how far along is a city in its earthquake cycle of large sized events at current time. The proposed analysis and emanated results produce valuable information to the academia, industry, and public on the current dynamical state of seismic hazard in the highly earthquake prone Nepal region.

Documentation of Surface Fault Rupture and Ground-Deformation Features Produced by the 4 and 5 July 2019 Mw 6.4 and Mw 7.1 Ridgecrest Earthquake Sequence

Abstract The Mw 6.4 and Mw 7.1 Ridgecrest earthquake sequence occurred on 4 and 5 July 2019 within the eastern California shear zone of southern California. Both events produced extensive surface faulting and ground deformation within Indian Wells Valley and Searles Valley. In the weeks following the earthquakes, more than six dozen scientists from government, academia, and the private sector carefully documented the surface faulting and ground-deformation features. As of December 2019, we have compiled a total of more than 6000 ground observations; approximately 1500 of these simply note the presence or absence of fault rupture or ground failure, but the remainder include detailed descriptions and other documentation, including tens of thousands of photographs. More than 1100 of these observations also include quantitative field measurements of displacement sense and magnitude. These field observations were supplemented by mapping of fault rupture and ground-deformation features directly in the field as well as by interpreting the location and extent of surface faulting and ground deformation from optical imagery and geodetic image products. We identified greater than 68 km of fault rupture produced by both earthquakes as well as numerous sites of ground deformation resulting from liquefaction or slope failure. These observations comprise a dataset that is fundamental to understanding the processes that controlled this earthquake sequence and for improving earthquake hazard estimates in the region. This article documents the types of data collected during postearthquake field investigations, the compilation effort, and the digital data products resulting from these efforts.

The importance of surface ruptures and fault damage zones in earthquake hazard assessment: a review and new suggestions

Abstract Large earthquakes (M W > c. 6) generate surface ruptures and strong ground motions, resulting in extensive property damage and loss of lives. The earthquake damage and causality levels depend, for example, on the earthquake magnitude, seismic wave characteristics, the distance from the epicentre, the depth of earthquake and the foundation of related area . However, these factors cannot all be known for future earthquakes, because the exact times and locations of earthquakes cannot be predicted with the current technology. Thus, it is only possible to prepare for earthquakes by implementing anti-seismic design codes based on the local geological conditions. With accurate information about the locations of active faults or potential surface ruptures and deformation patterns, earthquake-prone areas could be avoided during site selection, especially for important facilities such as nuclear power plants and high-level radioactive waste disposal facilities. In this study, we investigated several factors for earthquake hazard estimation, focusing on the fault damage zones and respect distances. Fault damage zones around faults are mainly controlled by fault types, tip modes, exposed locations and slip senses along the fault. The characteristics of geological faults and earthquake ruptures show asymmetrical displacements and damage patterns, making it difficult to estimate respect distances and surface deformations. Therefore, both the respect distance and fault damage zone should be considered when identifying safe regions. This information will facilitate earthquake hazard assessment and appropriate site selection to reduce earthquake hazards in earthquake-prone areas.

Resolving rupture processes of great earthquakes: Reviews and perspective from fast response to joint inversion

Resolving rupture processes of great earthquakes has fundamental importance to the study of earthquake physics, rupture dynamics, fault zone structure, and evolving processes. It also plays an essential role in earthquake hazard estimation, emergency response and seismic hazard mitigation. This paper reviews the major progress of the earthquake rupture process studies in the last decades, with an emphasize on the research directions of the department geophysics of Peking University including real-time response, back-projection techniques, geodetic data analysis, joint inversion and inversion in complex earth medium. We discussed the advantages and limitations of tradition methods; proposed a systematic and integrated approach from fast-response to detailed study. We also raised perspectives of using source models for ground motion prediction and the possibility of full-dynamic inversion.

Capturing epistemic uncertainty in the Iranian strong-motion data on the basis of backbone ground motion models

In the current practice of probabilistic seismic hazard analysis (PSHA), the different estimates of ground motions predicted by ground motion models (GMMs) are attributed to epistemic uncertainty. The epistemic uncertainties arise from the lack of knowledge which is reflected in imperfect models and can be handled by either logic tree or backbone approaches. The use of backbone approach for PSHA provides a more robust estimation of the GMM contribution to the epistemic uncertainty. In this study, we quantify the epistemic uncertainty in the Iranian strong-motion data by a scale factor that can be calibrated to the recorded strong-motions. The scale factor is then added and subtracted from the backbone GMM to fairly cover the spread in the predictions from other GMMs. For this purpose, we used the Iranian strong-motion database that includes 865 records from 167 events up to 2013, with the moment magnitude range of 5.0 ≤ M ≤ 7.4, and distances up to 120 km including a variety of fault mechanisms. On the other hand, several candidate GMMs were selected from local, regional, and worldwide data. Then, we applied a data-driven method based on the deviance information criterion to rank the candidate GMMs and select the best GMM as the backbone model. The results of this study show that the epistemic uncertainty varies approximately from 0.1 to 0.3 in base-10 logarithmic units. It generally has minima in the magnitude range of prevalent data (M 5.5–6.5) and increases for small (M 4.5–5.5) and large earthquake magnitudes (M 6.5–7.5). The results also show that the scale factors generally grow with distance. Moreover, notable site effects are seen in the Iranian strong-motions. We conclude therefore that the proposed backbone GMMs along with the estimated scales factors of this study are promising for use in future earthquake hazard estimation in Iran, as they capture the recorded data and provide information on the upper and lower bounds of ground motion estimates.

Site Effect Assessment in Ulaanbaatar, Mongolia through Inversion Analysis of Microtremor H/V Spectral Ratios

Due to the population growth and urban sprawl in Ulaanbaatar city (UB), Mongolia, hazard and risk analysis for future earthquakes have become an important issue for disaster mitigation planning. Evaluation of a site effect is one of the essential parts of the earthquake hazard estimation in this area. The site effect can be evaluated by site amplifications calculated from shear-wave velocity (VS) models including from bedrock to surface layers. However, it is difficult to assess the pattern of the site effects in UB because shallow mostly up to 15 m and a small number of investigated VS models are available in previous studies. In this study, the VS models are estimated using microtremor data at 50 sites and inversion analysis is applied to the observed data in order to evaluate site amplifications in UB. In particular, the joint inversion technique based on a diffuse field approach is applied to estimate the VS structures at three sites using the observed horizontal-to-vertical (H/V) spectral ratios and surface wave phase velocities obtained by Odonbaatar (2011). The rest of the sites are estimated by the single inversion technique using the observed microtremor H/V spectral ratios considering the results of the joint inversions. The seismic microzoning in UB is performed based on the site amplifications computed from the inverted VS models to characterize the pattern of seismic hazard. The result shows the largest site amplification zone is computed along the Tuul river in the southeastern part of UB.

A generalized theory for full microtremor horizontal-to-vertical [H/V(z, f)] spectral ratio interpretation in offshore and onshore environments

Advances in the field of seismic interferometry have provided a basic theoretical interpretation to the full spectrum of the microtremor horizontal-to-vertical spectral ratio [H/V(f)]. The interpretation has been applied to ambient seismic noise data recorded both at the surface and at depth. The new algorithm, based on the diffuse wavefield assumption, has been used in inversion schemes to estimate seismic wave velocity profiles that are useful input information for engineering and exploration seismology both for earthquake hazard estimation and to characterize surficial sediments. However, until now, the developed algorithms are only suitable for on land environments with no offshore consideration. Here, the microtremor H/V(z, f) modeling is extended for applications to marine sedimentary environments for a 1D layered medium. The layer propagator matrix formulation is used for the computation of the required Green's functions. Therefore, in the presence of a water layer on top, the propagator matrix for the uppermost layer is defined to account for the properties of the water column. As an application example we analyze eight simple canonical layered earth models. Frequencies ranging from 0.2 to 50 Hz are considered as they cover a broad wavelength interval and aid in practice to investigate subsurface structures in the depth range from a few meters to a few hundreds of meters. Results show a marginal variation of 8 percent at most for the fundamental frequency when a water layer is present. The water layer leads to variations in H/V peak amplitude of up to 50 percent atop the solid layers.

ESTIMATION OF SHEAR-WAVE VELOCITY STRUCTURE USING MICROTREMOR MEASUREMENTS

Due to the recent construction boom in Ulaanbaatar city (UB), Mongolia, newly urban areas are rapidly expanded to the surrounded area of the city. According to the previous researches and reports, the thickness of the sedimentary basin in UB reaches 120m at the maximum in the area along the Tuul river. Therefore, the evaluation of site amplification is one of the essential parts of the estimation of earthquake hazard in this area. In this paper, the shear-wave velocity (Vs) structures in UB are estimated from single microtremor observations with the existing microtremor array observations. We carried out the microtremor observations at 50 sites and computed horizontal-to-vertical (H/V) spectral ratios. Inversion analysis is performed to the observed H/V spectral ratios based on the diffuse field approach (DFA) to determine the Vs structures. The site characterizations are evaluated from the amplification factors of the Vs structures.

Three‐Dimensional Basin and Fault Structure From a Detailed Seismic Velocity Model of Coachella Valley, Southern California

AbstractThe Coachella Valley in the northern Salton Trough is known to produce destructive earthquakes, making it a high seismic hazard area. Knowledge of the seismic velocity structure and geometry of the sedimentary basins and fault zones is required to improve earthquake hazard estimates in this region. We simultaneously inverted first P wave travel times from the Southern California Seismic Network (39,998 local earthquakes) and explosions (251 land/sea shots) from the 2011 Salton Seismic Imaging Project to obtain a 3‐D seismic velocity model. Earthquakes with focal depths ≤10 km were selected to focus on the upper crustal structure. Strong lateral velocity contrasts in the top ~3 km correlate well with the surface geology, including the low‐velocity (<5 km/s) sedimentary basin and the high‐velocity crystalline basement rocks outside the valley. Sediment thickness is ~4 km in the southeastern valley near the Salton Sea and decreases to <2 km at the northwestern end of the valley. Eastward thickening of sediments toward the San Andreas fault within the valley defines Coachella Valley basin asymmetry. In the Peninsular Ranges, zones of relatively high seismic velocities (~6.4 km/s) between 2‐ and 4‐km depth may be related to Late Cretaceous mylonite rocks or older inherited basement structures. Other high‐velocity domains exist in the model down to 9‐km depth and help define crustal heterogeneity. We identify a potential fault zone in Lost Horse Valley unassociated with mapped faults in Southern California from the combined interpretation of surface geology, seismicity, and lateral velocity changes in the model.

NOWCASTING EARTHQUAKES IN THE NORTHWEST HIMALAYA AND SURROUNDING REGIONS

Abstract. With the rapid increase and availability of seismic data, an automatic, transparent and regular way of earthquake hazard estimation strategy is highly desirable in many seismically active large geographical regions. In this paper, we implement a novel method of nowcasting (Rundle et al., 2016) that can indirectly assess the current progression of a region through its earthquake cycle of large events. Nowcasting differs from the method of forecasting in which future earthquake probabilities are calculated. Using statistics of natural times, counts of small earthquakes between large earthquakes in a defined region, nowcasting provides an earthquake potential score (EPS) to enable scientists and city planners a snapshot of the current level of earthquake hazard in the region. Applied to a number of selected major cities in the northwest Himalaya and surrounding regions, we found that the EPS values corresponding to M ≥ 6 events in New Delhi, Chandigarh, Dehradun and Shimla reach about 0.56, 0.87, 0.85 and 0.88, respectively. These estimated scores thus indicate that New Delhi is about half-way through its cycle for magnitude 6.0 or higher earthquakes, while Dehradun is about 85 percent of the way through its cycle. Towards the end, we discuss some implications and applications of these nowcast values to improve the present earthquake hazard assessment practice in the study region.

Shear Wave Velocities in the Estimation of Earthquake Hazard Over Alluvium in Seismically Active Region

It is well known that the potential hazard during an earthquake is mainly in alluvium or alluvium filled basins; shear wave velocity plays a significant role in estimating the possible hazard during an earth quake in such an area. This paper presents shear wave velocity profile from Jabalpur, Central India mainly dominated by alluvial soil that was moderately affected by an earthquake of magnitude 6.5 in May, 1997. The acquired shear wave velocity by Multichannel Analysis of Surface Wave (MASW) in as many as 36 sites over alluvial soil ranges from 200 m/sec to 400 m/sec and in a few sites marginally less than 200 m/sec corresponding to a sub-surface depth of 30–35 m. Further, the computed N values vary as low as near zero to less than 25. The study is substantiated by the estimation of P-wave velocity by refraction seismic method at the same locations of MASW which ranges from 350 m/sec to 2200 m/sec. The results suggest that the damage during an earth quake appears to be highly unlikely in view of the marginally high Vs up to depth of 30 m. This study on seismic hazard is substantiated by the estimation of frequency of the ground as well as amplification which is found to be a maximum of about 2.5 in the frequency band of 2–6 Hz in west and north western portion of the study area.

Application of High Performance Computing to Earthquake Hazard and Disaster Estimation in Urban Area

Integrated earthquake simulation (IES) is a seamless simulation of analyzing all processes of earthquake hazard and disaster. There are two difficulties in carrying out IES, namely, the requirement of large scale computation and the requirement of numerous analysis models for structures in an urban area, and they are solved by taking advantage of High Performance Computing (HPC) and by developing a system of automated model construction. HPC is a key element in developing IES, as it needs to analyze wave propagation and amplification processes in an underground structure; a model of high fidelity for the underground structure exceeds a degree-of-freedom larger than 100 billion. Examples of IES for Tokyo Metropolis are presented; the numerical computation is made by using K computer, the supercomputer of Japan. The estimation of earthquake hazard and disaster for a given earthquake scenario is made by the ground motion simulation and the urban area seismic response simulation, respectively, for the target area of 10,000 $\times$ 10,000 m.

Real-time acceleration monitoring-based rapid earthquake response for ports in Korea

ABSTRACTTo establish viable earthquake counterplans for ports in Korea, data regarding earthquake motion on the ground and in buildings must be collected using acceleration monitoring systems. Acceleration monitoring-based strategies for port facilities are useful not only for ensuring rapid responses during and after earthquakes but also for regional data collection to assist reliable seismic design. Acceleration monitoring systems were installed at coastal facility sites in target ports, including a soil site to represent the facility and a rock site as a reference. The systematic earthquake alert software was designed to help them in decision-making about a possible seismic hazard and its reporting. The earthquake alert system was composed of two sequential functional software systems sharing an acceleration monitoring database applied to the target ports. The earthquake response software system triggers an alarm based on the peak ground acceleration per second computed from the monitored data. Then, the earthquake hazard estimation software system evaluates possible earthquake-induced site-specific geotechnical hazards linked to the peak ground acceleration. The integrated system was successfully operated and was able to rapidly provide an emergency report containing event records and geotechnical earthquake hazards during the September 2016 Gyeongju earthquake, the largest recorded earthquake in Korea.

Seismic hazard in Central Italy and the 2016 Amatrice earthquake

<p><em>The Amatrice earthquake of August 24th, 2016 (Mw 6.0) struck an area that in the national reference seismic hazard model (MPS04) is characterized by expected horizontal peak ground acceleration (PGA) with 10% probability of exceedance in 50 years higher than 0.25 g. After the occurrence of moderate-to-large magnitude earthquakes with a strong impact on the population, such as the L’Aquila 2009 and Emilia 2012 ones (Mw 6.1 and 5.9, respectively), possible underestimations of the seismic hazard by MPS04 were investigated, in order to analyze and evaluate the possible need for its update. One of the most common misunderstanding is to compare recorded PGA only with PGA with 10% probability of exceedance in 50 years. Moreover, by definition, probabilistic models cannot be validated (or rejected) on the basis of a single event. However, comparisons of forecasted shakings with observed data are useful for understating the consistency of the model. It is then worth highlighting the importance of these comparisons. In fact, MPS04 is the basis for the current Italian building code to provide the effective design procedures and, thus, any modification to the seismic hazard would also affect the building code.</em></p><em>In this paper, comparisons between recorded ground motion during the Amatrice earthquake and seismic hazard estimates are performed, showing that the observed accelerations are consistent with the values expected by the MPS04 model.</em>

Allowing faults out of lock-up

Geophysics![Figure][1] Relict microatoll on southeastern island of Simeulue, Sumatra PHOTO: ARON J. MELTZNER/EARTH OBSERVATORY OF SINGAPORE The crust around a locked, earthquake-prone fault responds to the two sides pulling in opposite directions by bowing and bulging over time. Meltzner et al. and Wesson et al. challenge a common assumption that this deformation progresses mostly in a uniform, linear way. Deformation shows up in seafloor bathymetry, which changes abruptly for both Sumatra and Isla Santa Maria, Chile, over a seismic cycle. Periods during which the fault is weakly locked explains the non-uniform behavior. Quantifying the effect clarifies subduction zone mechanics, which may require updating estimates of earthquake hazard. Quat. Sci. Rev. 10.1016/j.quascirev.2015.06.003 (2015); Nat. Geosci. 10.1038/ngeo2468 (2015). [1]: pending:yes