Earthquake Likelihood Research Articles

Mitigation strategies and policies recommendation for the economic impact of the Sunda Strait Megathrust: Seismic risk probability assessment and cost loss estimates

This study investigates seismic risk and potential impacts of future earthquakes in the Sunda Strait region, known for its susceptibility to significant seismic events due to the subduction of the Indo-Australian Plate beneath the Eurasian Plate. The aim is to assess the likelihood of major earthquakes, estimate their impact, and propose strategies to mitigate associated risks. The research uses historical seismic data and probabilistic models to forecast earthquakes with magnitudes ranging from 6.0 to 8.2 Mw. The Gutenberg-Richter model helps project potential earthquake occurrences and their impacts. The findings suggest that the probability of a major earthquake could occur as early as 2026–2027, with a more significant event estimated to likely occur around 2031. Economic estimates for a 7.8–8.2 Mw earthquake suggest potential damage of up to USD 1.255 billion with significant loss of life. The study identifies key vulnerabilities, such as inadequate building foundations and ineffective disaster management infrastructure, which could worsen the impact of future seismic events. In conclusion, the research highlights the urgent need for comprehensive seismic risk mitigation strategies. Recommendations include reinforcing infrastructure to comply with seismic standards, implementing advanced early warning systems, and enhancing public education on earthquake preparedness. Additionally, government policies must address these issues by increasing funding for disaster management, enforcing building regulations, and incorporating traditional knowledge into construction practices. These measures are essential to reducing future earthquake impacts and improving community resilience.

Predicting the Magnitude of Possible Earthquakes in the Shahrood District of Khalkhal County, Ardabil, Iran, Using Artificial Neural Networks

Background and objective Earthquakes are natural disasters that can cause high damage. Prediction of earthquakes is necessary to reduce damages, develop resistant infrastructure, and prevent the loss of life in human societies. This research aims to predict the magnitude of possible earthquakes in the Shahrood district of Khalkhal County using artificial neural networks (ANN). Method A multi-layer perceptron neural network was used to predict earthquake magnitude. Results The results showed that in the fault areas, especially the Kalur fault, earthquakes with a magnitude of 1-3 on the Richter scale are more likely to occur (70%). The likelihood of earthquakes with a magnitude of 4-6 on the Richter scale is moderate (26%) and the earthquakes with a magnitude of 7-10 on the Richter scale have a very low probability (4%). Conclusion In future studies, it is recommended to use machine learning models to better and more accurately predict earthquakes in the studied area.

Analysis of emergency assembly points for post-earthquake disaster management: a case study of Erzincan, Türkiye

The selection of emergency assembly points within the urban area holds significant importance for ensuring swift and effective intervention in the aftermath of a potential disaster until temporary shelter areas are prepared. In both historical and instrumental periods, Erzincan (Türkiye) has experienced major earthquakes resulting in significant loss of life and property. The likelihood of similar earthquakes occurring in Erzincan in the future remains high. Therefore, it is necessary to identify safe areas where people can assemble after an earthquake, moving away from hazardous zones. The aim of this study is to analyse the capacities and adequacies of emergency assembly points, which constitute a step in disaster management and enhance the earthquake resilience of the city, using Geographic Information Systems (GIS) based on selected criteria (accessibility, spatial distribution, proximity to healthcare facilities, size, capacity adequacy, and proximity to fault avoidance zones), and to propose solutions. The assessment revealed that the available assembly points are not homogeneously distributed throughout the city and are inadequate to meet the available building and population density. Therefore, in addition to the available 38 emergency assembly points, 67 new alternative emergency assembly points have been identified. The proposed emergency assembly points (in the 75–100% range) increased the number of neighborhoods within the 250 m service area by 43.75% and 42.5% for 500 m.The findings underscore the necessity for more comprehensive and effective planning for Erzincan in the event of a potential disaster or emergency.

Spatial Distribution of Earthquake Occurrence for the New Zealand National Seismic Hazard Model 2022

ABSTRACT We develop candidate hybrid models representing the spatial distribution of earthquake occurrence in New Zealand over the next 100 yr. These models are used within the onshore/near-shore, shallow component of the distributed seismicity model within the New Zealand National Seismic Hazard Model 2022. They combine a variety of spatially gridded covariates based on smoothed seismicity, strain rates, and proximity to mapped faults and plate boundaries in both multiplicative and additive hybrids. They were optimized against a standardized catalog of New Zealand earthquakes with magnitude M ≥ 4.95 and hypocentral depth ≤40 km from 1951 to 2020. We extract smoothed seismicity covariates using three different methods. The additive models are linear combinations of earthquake likelihood models derived from individual covariates. We choose three preferred hybrid models based on the information gain statistics, consideration of the ongoing Canterbury sequence and regions of low seismicity, and inclusion of the most informative covariates. Since the hazard model is designed for the next 100 yr, the preferred hybrid models are also combined with 20-year earthquake forecasts from the “Every Earthquake a Precursor According to Scale” model. Thus, in total, six hybrid spatial distribution candidates are advanced for sensitivity analyses and expert elicitation for inclusion in the final logic tree for the New Zealand National Seismic Hazard Model.

Analysis of Bridge Dynamics and Response Characteristics Under The Influence of Axle Coupling Vibration

To ensure the safety and stability of high-speed rail lines and reduce external interference, it is essential to build a large number of elevated bridges. These elevated bridges account for a considerable proportion of the total length of high-speed rail lines. However, when high-speed rail lines pass through earthquake prone areas, the likelihood of earthquakes occurring when trains pass through bridges increases significantly. Therefore, it is necessary to study the dynamic response of bridge structures under earthquake action to ensure the safety of bridges during train operation and operation. The optimization scheme proposed in this article has undergone moderate impact tests, and the results show that the maximum lateral displacement of the bridge can reach 124 mm, while the maximum vertical acceleration is 5.16 m/s2, Exceeded the safety limit of 0.35 g. Through the analysis of train derailment coefficient, wheel load reduction rate, lateral sway force, lateral and vertical acceleration, and Spelling comfort index, we have come to the conclusion that bridges can ensure the safety of train operation in the absence of earthquakes and small earthquakes, and can also maintain certain stability under medium and large earthquakes. These research results have important guiding significance for the design and construction of high-speed rail lines. By optimizing the bridge structure and adopting relevant technical measures, the seismic disaster resistance of high-speed rail lines can be further improved, ensuring the safety and comfort of passengers during travel. At the same time, these research results also provide useful reference and inspiration for the construction and improvement of future high-speed rail lines.

Impacts of Rate of Change in Effective Stress and Inertial Effects on Fault Slip Behavior: New Insights Into Injection‐Induced Earthquakes

AbstractUnderstanding the physical mechanisms which link fluid injection with triggered earthquakes is critical in minimizing hazard in subsurface fluid‐injection operations. Currently, injection‐induced changes in effective stress on faults are considered as the main criterion in triggering seismic fault slip. However, rate of change in effective stress, together with inertial effects, are also be implicated in this criterion. We present a modified critical stiffness criterion to investigate the relative likelihood of triggering earthquakes during injection for different injection rate schedules (constant‐vs‐cycled‐vs‐increasing). A stability analysis of fault stress is used to define a critical stiffness as a function of magnitudes and rate of change in effective stresses. The relative potential for triggering earthquakes due to fluid injection is investigated using a coupled fluid‐flow‐deformation model. Polarities of change in critical stiffness are employed as an index to define the tendency for a transition from aseismic to seismic reactivation. During constant rate injection and self‐equilibration stages, the absolute magnitude of effective stress controls the transition. Conversely, the rate of change in effective stress dominates this transition when injection suddenly starts or stops, and inertial effects suppress the transformation to seismic slip. Cycling injection rates into a given fault is the most stable, followed by constant injection, with linear injection the least stable for the same total volume injected. High permeability reservoirs and strike‐slip faulting regimes reduce the potential of inducing seismicity. This work provides both new insights into assessing the seismic risks associated with injection and guidance for mitigation.

Time-Independent Grid-Based Forecast Model for M ≥6.0 Earthquakes in Southeastern Tibetan Plateau Using GNSS Strain Rates and Seismicity

ABSTRACT Earthquake forecasting models play a vital role in earthquake occurrence assessment. Despite improved availability of seismic and geodetic data and processing techniques to produce high-resolution catalogs and deformation history, the implementation of earthquake forecasting models with seismic and geodetic data remains a challenge. In this study, we utilize seismicity and Global Navigation Satellite Systems (GNSS) data to propose time-independent grid-based regional earthquake likelihood models for southeastern Tibetan plateau (RELM-TibetSE). First, we solve the GNSS velocity field and strain rates from 1999 to 2017, deriving geodetic moment rates and introducing empirical correction coefficients to balance them with historical seismic moment rate. Subsequently, we employ a truncated Gutenberg–Richter law and Poisson process to calculate time-independent probabilities for M ≥ 6 earthquakes in 0.2° × 0.2° cells. The grid-based forecasting models indicate that the 30-year probability for M ≥ 6 earthquakes exceeds 1% in more than one-third of the entire study area, highlighting prominently higher earthquake occurrence in these regions. Probability distribution exhibits significant spatial variations. Finally, the predictive performance of the forecasting models is validated based on historical seismicity. The validation indicates that all RELM-TibetSE exhibit good predictive capability relative to a spatially uniform model. The RELM-TibetSE incorporating principal strain rates outperforms those involving maximum shear strain rate in forecasting seismicity. And the differences in forecasting performance between the RELM-TibetSE accounting for spatially varied seismogenic thickness and rigidity and those with uniform thickness and rigidity are not significant. The forecasting models also exhibit better predictive performance for seismic source areas than for epicenters. Moreover, the optimal model highlights zones with higher earthquake occurrence, including the zones about 50 km wide across the Ninglang fault, the zones across the southwestern segment of the Lijiang–Xiaojinhe fault, the China–Myanmar borderland north of the Nantinghe fault, and so on. Therefore, it is justified to conduct multidisciplinary rigorous observations to capture the potential nucleation process of future large earthquakes in these zones.

Embedding extreme events to mine project planning: Implications on cost, time, and disclosure standards

Extreme events, characterized as infrequent and catastrophic occurrences, can present significant challenges in various industries. These events can be associated with earthquakes, landslides, hurricanes, floods, rainfalls, and wildfires. The mining industry is also highly vulnerable to extreme events because it threats operation sustainability, employees' safety, and environmental compliance. This paper presents a comprehensive analysis of earthquakes in a mining area and evaluates their impacts on a mining operation through a case study based on a mine construction project. To this end, Extreme Value Theory (EVT) using both the Block Maxima (BM) and Peaks Over Threshold (POT) methods was employed to analyze the historical earthquake data and estimate the return levels corresponding to the planned mine project completion time. Furthermore, a Discrete Event Simulation (DES) model was developed to perform a comprehensive risk assessment, allowing for a detailed examination of the potential consequences that a specified magnitude earthquake could have on the project's cost and duration. 1000 earthquake scenarios for the period covering the project life were generated for a thorough risk assessment. The case study focused on a hypothetical open pit mine construction project located in the province of British Columbia, Canada. Historical earthquake data specific to the selected region were analyzed, and the potential consequences of earthquakes on the project were evaluated. According to the results, the return level of 4.7 ML earthquake was obtained corresponding to the return period defined as the planned project completion time. In the event of an earthquake with this specified magnitude, simulation results revealed that there is a 95% probability that the project will experience a delay of at least 6.28% of the project completion time, ultimately leading to project costs exceeding the normal project cost by a minimum of 4.85%. The study highlights the importance of considering extreme natural events, such as earthquakes, in mining project planning and risk mitigation. The results provide valuable insights into the likelihood and potential consequences of earthquakes, contributing to better risk assessment, strategic planning, informed decision-making, and improved project management practices in the mining industry.

A Spatially Self-Adaptive Multiparametric Anomaly Identification Scheme Based on Global Strong Earthquakes

Earthquake forecasting aims to determine the likelihood of a damaging earthquake occurring in a particular area within a period of days to months. This provides ample preparation time for potential seismic hazards, resulting in significant socioeconomic benefits. Surface and atmospheric parameters derived from satellite thermal infrared observations have been utilized to identify pre-earthquake anomalies that may serve as potential precursors for earthquake forecasting. However, the correlation between these anomalies and impending earthquakes remains a significant challenge due to high false alarm and missed detection rates. To address this issue, we propose a spatially self-adaptive multiparametric anomaly identification scheme based on global strong earthquakes to establish the optimal recognition criteria. Each optimal parameter exhibits significant spatial variability within the seismically active region and indicates transient and subtle anomaly signals with a limited frequency of occurrences (<10 for most regions). In comparison to the fixed criterion for identifying anomalies, this new scheme significantly improves the positive Matthew’s correlation coefficient (MCC) values from ~0.03 to 0.122–0.152. Additionally, we have developed a multi-parameter anomaly synthesis method based on the best MCC value of each parameter anomaly. On average, the MCC increased from 0.143 to 0.186, and there are now more earthquake-prone regions with MCC values > 0.5. Our research emphasizes the critical importance of a multiparametric system in earthquake forecasting, where each geophysical parameter can be assigned a distinct weight, and the findings specifically identify OLR, including all-sky and clear-sky ones, as the most influential parameter on a global scale, highlighting the potential significance of OLR anomalies for seismic forecasting. Encouraging results imply the effectiveness of utilizing multiparametric anomalies and provide some confidence in advancing our knowledge of operational earthquake forecasting with a more quantitative approach.

Geo-hazards assessment and land suitability estimation for spatial planning using multi-criteria analysis

The fundamentals contribution of multi-hazard assessment lies in its ability to guide and identify directions for territorial development, aiming to reduce vulnerability through the implementation of appropriate measures. In the present study, Multi-Criteria Decision Making supported by Geographic Information System was utilized to solve spatial problems related to territorial sprawl. Considering this perspective, an analysis was conducted on the susceptibility of land to the occurrence of geo hazards in the Moldova catchment, situated in the northeastern region of Romania, within a transitional zone between mountains and plateaus. The analysis focused on assessing the likelihood of geo-hazards like floods, landslides, and earthquakes, utilizing the weights obtained through the implementation of the Analytical Hierarchy Process (AHP). Subsequently, the obtained results were utilized to generate a multi-hazard map, which facilitated the identification of areas that are suitable for territorial development. The results were validated in two ways: by sensitivity analysis in which two minimum and maximum scenarios were proposed concerning the result obtained and by validation using the receiver operating characteristic (ROC) curve method. The areas with high susceptibility to geo-hazards triggers are located in the southeastern part of the region, in the proximity of the hydrographic network. Flood risk is the hazard with the highest recurrence. Localities with high suitability for spatial development have a central-western distribution. According to the applied model, sites for each locality included in the study area can be identified. The high and very high suitability classes account for 54% of the total area, while the unsuitable classes represent 15% of the area. However, the vulnerability of the area and the need for the study are generated by 35% of the settlements being located in areas with high susceptibility.

Small Earthquakes Can Help Predict Large Earthquakes: A Machine Learning Perspective

Earthquake prediction is a long-standing problem in seismology that has garnered attention from the scientific community and the public. Despite ongoing efforts to understand the physical mechanisms of earthquake occurrence, there is no convincing physical or statistical model for predicting large earthquakes. Machine learning methods, such as random forest and long short-term memory (LSTM) neural networks, excel at identifying patterns in large-scale databases and offer a potential means to improve earthquake prediction performance. Differing from physical and statistical approaches to earthquake prediction, we explore whether small earthquakes can be used to predict large earthquakes within the framework of machine learning. Specifically, we attempt to answer two questions for a given region: (1) Is there a likelihood of a large earthquake (e.g., M ≥ 6.0) occurring within the next year? (2) What is the maximum magnitude of an earthquake expected to occur within the next year? Our results show that the random forest method performs best in classifying large earthquake occurrences, while the LSTM method provides a rough estimation of earthquake magnitude. We conclude that small earthquakes contain information relevant to predicting future large earthquakes and that machine learning provides a promising avenue for improving the prediction of earthquake occurrences.

Probabilistic seismic hazard function based on spatiotemporal earthquake likelihood simulation and Akaike information criterion: The PSHF study around off the west coast of Sumatra Island before large earthquake events

The probabilistic seismic hazard function (PSHF) before large earthquake events based on the hypothesis earthquake forecast algorithm using the Akaike information criterion (AIC) is performed in this study. The motivation for using the AIC is to better understand the reliability model used to construct the PSHF. The PSHF as the function of the b-value is calculated based on a 5-year window length with a 1-year moving window (instantaneous PSHF) before a large earthquake event. The AIC is calculated based on the likelihood of success and failure using shallow earthquake catalog data around the west coast of Sumatra Island. The probability of occurrence defines the success criteria as more significant than the average probability of greater than or equal to the given magnitude; otherwise, it is defined as failure. Seismic potency has been determined based on the likelihood of an earthquake occurring in several decades or a hundred years. The seismicity rate model is developed based on the integrated data of pre-seismic shallow crustal movement data and the shallow crustal earthquake catalog data. Furthermore, the AIC is calculated based on the likelihood of success and failure as a function of b(t). The b(t) is the change in the b-value as a time function estimated based on shallow earthquake data from 1963 to 2016. In addition, the AIC before M7.9 of 2000, M8.5 of 2007, and M7.8 of 2010 is assessed. The δAIC is then introduced as a function of (AICmodel–AICreference) during the observation time. The positive δAIC implies that the likelihood of having a large earthquake is more significant; otherwise, it is smaller. By plotting the time of observation versus δAIC and the PSHF estimated as the function of b(t), we could identify a large positive gradient and increase the PSHF at each certain probability exceedance (PE) level before the great earthquake event. It consistently happened for the three events that were evaluated. It suggested that the results of this study might be very beneficial for probabilistic seismic hazard analysis (PSHA) and seismic mitigation realization.

New Tectonic Plate Model Could Improve Earthquake Risk Assessment

A new model of tectonic plates in New Zealand may identify areas of increased earthquake likelihood.

Towards ground motion prediction for potential large earthquakes from interseismic locking models

Interseismic locking models derived from geodetic data describe slip deficit accumulation on faults, hence indicating the likelihood of future earthquakes. In recent years, locking models have been used to develop dynamic scenarios for potential large earthquakes. However, whether those scenarios are effective in representing the rupture process and ground motion intensity in future earthquakes remains unclear. In this study, we examine the ground motion predictions in locking-based dynamic rupture scenarios by comparing them with the predictions from ground motion models (GMMs) and the observations in a real earthquake. We utilize two locking models as constraints on the stress heterogeneity and obtain synthetic ground motion measures, including the peak ground velocity (PGV) and the peak ground displacement (PGD), through conducting dynamic rupture simulations for M7 earthquakes on the Nicoya megathrust. The predictions are generally consistent with GMMs with some differences in attenuation rate and amplitude in the near field (< 30 km). However, the spatial patterns differ a lot from GMM-based predictions, mainly due to the rupture directivity effect. By comparing with scenarios under homogenous stress conditions, we observe the dependency of earthquake magnitude and rupture directivity on stress heterogeneity. The coseismic slip appears to negatively correlate with stress roughness. Furthermore, the predictions from one locking model capture most of the measurements on the local network during the 2012 Nicoya Mw 7.6 earthquake. Our results underline the necessity of involving rupture dynamics and stress heterogeneities in prescribing the earthquake source and highlight the potential application of locking models in seismic hazard assessment.

Small Magnitude Events Highlight the Correlation Between Hydraulic Fracturing Injection Parameters, Geological Factors, and Earthquake Occurrence

AbstractHydraulic fracturing (HF) operations are widely associated with induced seismicity in the Western Canadian Sedimentary Basin. This study correlates injection parameters of 12,903 HF stages in the Kiskatinaw area in northeast British Columbia with an enhanced catalog containing 40,046 earthquakes using a supervised machine learning approach. It identifies relevant combinations of geological and operational parameters related to individual HF stages in efforts to decipher fault activation mechanisms. Our results suggest that stages targeting specific geological units (here, the Lower Montney formation) are more likely to induce an earthquake. Additional parameters positively correlated with earthquake likelihood include target formation thickness, injection volume, and completion date. Furthermore, the COVID‐19 lockdown may have reduced the potential cumulative effect of HF operations. Our results demonstrate the value of machine learning approaches for implementation as guidance tools that help facilitate safe development of unconventional energy technologies.

Inherited Heterogeneities Can Control Viscous Subduction Zone Deformation of Carbonates at Seismogenic Depths.

This work links mineral-scale deformation mechanisms with structural evolution during subduction, providing examples showing how grain-scale heterogeneities facilitated viscous creep in calcite at nominally seismogenic temperatures. Carbonates commonly enter subduction zones, either highly concentrated in irregularly distributed sediments or as more distributed precipitates in seafloor volcanics. We present shear zones, localized in calcite veins formed during shallow subduction of calcareous sediment and seafloor volcanics, with viscous shear strains of ≥5. Shear strain localized because secondary phases and chemical variations maintained fine grain sizes in calcite aggregates, activating relatively rapid grain size-sensitive and frictional-viscous creep at temperatures (260±10°C), cooler than predicted from extrapolation of experimental data. Creep at increased strain rates may limit elastic strain accumulation during interseismic periods, reducing the likelihood of large megathrust earthquakes. As shown here for calcite, common inherited natural heterogeneities may induce weakening of viscous mechanisms in other rocks, or at larger scales in the lithosphere.

Pseudo-prospective testing of 5-year earthquake forecasts for California using inlabru

Abstract. Probabilistic earthquake forecasts estimate the likelihood of future earthquakes within a specified time-space-magnitude window and are important because they inform planning of hazard mitigation activities on different time scales. The spatial component of such forecasts, expressed as seismicity models, generally relies upon some combination of past event locations and underlying factors which might affect spatial intensity, such as strain rate, fault location and slip rate or past seismicity. For the first time, we extend previously reported spatial seismicity models, generated using the open source inlabru package, to time-independent earthquake forecasts using California as a case study. The inlabru approach allows the rapid evaluation of point process models which integrate different spatial datasets. We explore how well various candidate forecasts perform compared to observed activity over three contiguous 5-year time periods using the same training window for the input seismicity data. In each case we compare models constructed from both full and declustered earthquake catalogues. In doing this, we compare the use of synthetic catalogue forecasts to the more widely used grid-based approach of previous forecast testing experiments. The simulated catalogue approach uses the full model posteriors to create Bayesian earthquake forecasts, not just the mean. We show that simulated catalogue based forecasts perform better than the grid-based equivalents due to (a) their ability to capture more uncertainty in the model components and (b) the associated relaxation of the Poisson assumption in testing. We demonstrate that the inlabru models perform well overall over various time periods: The full catalogue models perform favourably in the first testing period (2006–2011) while the declustered catalogue models perform better in the 2011–2016 testing period, with both sets of models performing less well in the most recent (2016–2021) testing period. Together, these findings demonstrate a significant improvement in earthquake forecasting is possible although this has yet to be tested and proven in true prospective mode.

Discrete Element Modeling of a Subduction Zone with a Seafloor Irregularity and its Impact on the Seismic Cycle

AbstractSeafloor irregularities influence rupture behavior along the subducting slab and in the overriding plate, thus affecting earthquake cycles. Whether seafloor irregularities increase the likelihood of large earthquakes in a subduction zone remains contested, partially due to focus put either on fault development or on rupture pattern. Here, we simulate a subducting slab with a seafloor irregularity and the resulting deformation pattern of the overriding plate using the discrete element method. Our simulations illustrate the rupture along three major fault systems: megathrust, splay and backthrust faults. Our results show different rupture dimensions of earthquake events varying from tens to ca. 140 km. Our results suggest that the recurrence interval of megathrust events with rupture length of ca. 100 km is ca. 140 years, which is overall comparable to the paleoseismic records at the Mentawai area of the Sumatran zone. We further propose the coseismic slip amounts decrease and interseismic slip amounts increase from the surface downwards gradually. We conclude that the presence of seafloor irregularities significantly affects rupture events along the slab as well as fault patterns in the overriding plate.

Prospective evaluation of multiplicative hybrid earthquake forecasting models in California

SUMMARYThe Regional Earthquake Likelihood Models (RELM) experiment, conducted within the Collaboratory for the Study of Earthquake Predictability (CSEP), showed that the smoothed seismicity (HKJ) model by Helmstetter et al. was the most informative time-independent earthquake model in California during the 2006–2010 evaluation period. The diversity of competing forecast hypotheses and geophysical data sets used in RELM was suitable for combining multiple models that could provide more informative earthquake forecasts than HKJ. Thus, Rhoades et al. created multiplicative hybrid models that involve the HKJ model as a baseline and one or more conjugate models. In retrospective evaluations, some hybrid models showed significant information gains over the HKJ forecast. Here, we prospectively assess the predictive skills of 16 hybrids and 6 original RELM forecasts at a 0.05 significance level, using a suite of traditional and new CSEP tests that rely on a Poisson and a binary likelihood function. In addition, we include consistency test results at a Bonferroni-adjusted significance level of 0.025 to address the problem of multiple tests. Furthermore, we compare the performance of each forecast to that of HKJ. The evaluation data set contains 40 target events recorded within the CSEP California testing region from 2011 January 1 to 2020 December 31, including the 2016 Hawthorne earthquake swarm in southwestern Nevada and the 2019 Ridgecrest sequence. Consistency test results show that most forecasting models overestimate the number of earthquakes and struggle to explain the spatial distribution of epicenters, especially in the case of seismicity clusters. The binary likelihood function significantly reduces the sensitivity of spatial log-likelihood scores to clustering, however; most models still fail to adequately describe spatial earthquake patterns. Contrary to retrospective analyses, our prospective test results show that none of the models are significantly more informative than the HKJ benchmark forecast, which we interpret to be due to temporal instabilities in the fit that forms hybrids. These results suggest that smoothing high-resolution, small earthquake data remains a robust method for forecasting moderate-to-large earthquakes over a period of 5–15 yr in California.

Anomalies Prediction in Radon Time Series for Earthquake Likelihood Using Machine Learning-Based Ensemble Model

The ability to predict the radioactive soil radon gas concentration is important for human beings because it serves as a precursor to earthquakes. Several studies have been conducted across the globe to confirm the correlation of radon emission dynamics and earthquakes, and concluded that the soil radon gas is the witness of anomalous behaviour before the occurrences of several earthquakes. This anomalous behavior can help to construct a better prediction model for earthquake forecasting. This paper aims at employing different ensemble and individual machine learning methods on real time radon time series data with different scenarios to predict anomalies in data caused by the seismic activities.The ensemble methods include boosted tree, bagged cart and boosted linear model while standalone machine learning methods include support vector machine with linear and radial kernels and k-nearest neighbors ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${K}$ </tex-math></inline-formula> -NN). We tested the methods on a dataset recorded on the fault line located in Muzaffarabad. Time series data was collected over a period ranging from March 1, 2017 to May 11, 2018 including nine(09) earthquakes. The methods are tested in four different settings with 10 times 10 folds cross validation procedure over the time window of 1 to 4. The repeated 10 fold cross validation is performed to reduce the noise in the model performance estimation by replicating the 10 fold cross validation procedure 10 times. Statistical performance evaluation measures viz. root mean square error (RMSE), root mean squared log error (RMSLE), mean absolute percentage error (MAPE), percentage bias (PB), and mean squared error (MSE) have been calculated for the assessment of performance. In setting 1, the support vector machine with radial kernel performs better with the minimum RMSE score of 1381.023 when compared to other prediction models. In setting 3, it can be observed through different performance metrics such as RMSE, the value in the range [1262.864, 1409.616] which is minimum when other prediction models for predicting soil radon gas concentration dataset. For setting 4, the boosted tree model yielded the minimum RMSE and MAPE scores of 1573.174 and 0.056 respectively. Findings of the study shows that boosted tree and support vector machine with radial kernel proved to be better regression models for the prediction of anomalies in soil radon gas concentration during seismic activities. An important finding of this study suggests that by employing boosted tree ensemble method make us able to accurately predict soil radon gas concentration automatically from environmental parameters.