Journal of Seismology最新文献

Accurately predicting the timing, location, and magnitude of earthquakes is a major global scientific challenge. While machine learning methods have been widely used in seismic time series analysis and outperform traditional statistical models, their predictive capability for strong earthquakes (Mw ≥ 6.5) remains limited. This study develops a new prediction method by constructing multi-timescale seismic activity features and integrating them with machine learning models.The novelty of this approach is a parameterized framework based on a historical seismic window (T_history), forecast window (T_forecast), and magnitude threshold (M_threshold). It transforms earthquake prediction into an optimization problem, using machine learning to model the relationship between these parameters and performance metrics (e.g., AUC(Area Under the (ROC) Curve)), and searches the parameter space to maximize AUC. To improve spatial prediction, clustering algorithms divide the study area into subregions with consistent seismic characteristics, and an optimization algorithm determines the optimal time-magnitude parameters for each subregion.Using earthquake catalogs from Japan and New Zealand (1973–2024), K-means +  + was selected to partition Japan into 4 and New Zealand into 3 subregions. With T_forecast set to 365 days and time-series cross-validation applied to reduce overfitting, the method achieved average AUC values of 0.75 and 0.81 for predicting strong earthquakes within one year in Japan and New Zealand, respectively, outperforming existing models.The main contribution is a flexible and extensible prediction framework that overcomes the limitations of fixed regions, time windows, and magnitude thresholds. Future work will focus on extensible improvements, overfitting control, and integrating multi-source features to enhance generalization.

The Bucaramanga Seismic Nest (BSN), located in northwestern Colombia, is one of the world’s three most active volumes of concentrated intermediate-depth seismicity. Owing to its proximity to major urban centers, earthquakes with magnitudes (ge 4.5) in the BSN have repeatedly raised concern and caused infrastructure damage. Continuous monitoring has been carried out for decades, yet the prospect of anticipating the occurrence of moderate to large earthquakes in this region remains an open challenge. To address this problem, we develop a supervised learning approach that frames the task as a binary classification: a neural network issues a positive alert if at least one event with local magnitude (Ml) (ge 4.5) is likely to occur within the next six days. As inputs, the network uses daily event counts binned by magnitude over the preceding 30, 60, 180 or 365 days. The dataset comprises 148,910 earthquakes recorded between January 1994 and February 2024 by the Servicio Geológico Colombiano. We established a robust baseline using three classical classifiers—Linear Discriminant Analysis, Logistic Regression, and Random Forests. These models provided interpretable reference points but consistently failed to achieve ROC–AUC values above 0.60 and were strongly biased toward the majority class. In contrast, neural network models trained across multiple configurations consistently outperformed these baselines, with the best achieving ROC–AUC values (>0.60), non-negligible true-positive rates, and correct positive classifications in advance of observed events. Our results demonstrate that seismicity in the Bucaramanga Nest is not entirely random but contains learnable patterns, which will serve as a basis for future research efforts aimed at seismicity forecasting in the BSN.

Physics-based simulations (PBS) have found vast application in the generation of non-ergodic ground motion. However, due to their computational requirement, they are combined with other methodologies in a hybrid framework to expand the effect to a broadband frequency range. The present work proposes a new Artificial Neural Network (ANN) model developed using strong motion records from the Engineering Strong Motion (ESM) database, which predicts Fourier Amplitude Spectra (FAS) values in the high-frequency range. The developed model performed well in statistical comparisons and could capture various earthquake scaling effects. Further, the work includes the 2023 Elbistan, Türkiye earthquake as the case study for model validation. It was found that the model-predicted values correspond well with the recorded data for both spectra and ground motion Ground Motion Intensity Measures (GMIMs). The developed model is useful in performing non-ergodic seismic hazard analysis and obtaining broadband time histories necessary for non-linear time history analysis of structures.

We investigated the body waves (P & S) attenuation (({Q}_{P,S}^{-1})) at the well-known reservoir-triggered seismicity (RTS) site of the Koyna-Warna region. The study region has a history of intense seismic activity for more than six decades. Knowledge of body wave attenuation in such a tectonic environment is imperative for seismic hazard analysis and comprehension of the source processes, as well as for accurately assessing the material and physical properties of the media. In this study, we analyzed 113 earthquakes of M_L1.9–3.7 archived through a local seismic network of five stations. The attenuation of P (({Q}_{P}^{-1})) and S waves (({Q}_{S}^{-1})) are derived employing an extended Coda-Normalization method centered at seven central frequencies (1.5, 3.0, 4.5, 6.0, 9.0, 12.0, and 18.0 Hz).The findings of ({Q}_{P}^{-1}) and ({Q}_{S}^{-1}), indicates an apparent frequency dependence, which can be described by a power law equations: ({Q}_{P}^{-1}left(fright)=left(0.017pm 0.002right){f}^{left(-0.787pm 0.057right)}), and ({Q}_{S}^{-1}left(fright)=left(0.016pm 0.0004right){f}^{left(-0.682pm 0.059right)}) respectively. The results indicate that P-wave attenuation is more pronounced than S-wave at the entire range of frequencies, and both waves types have undergone significant dissipation along the ray path in the region. The obtained ({Q}_{P}^{-1}/{Q}_{S}^{-1}) values exceed unity across the span of frequency bandwidth, which may be potentially attributed to the influence of partially saturated basement rocks in the region. The increased values (({Q}^{-1})) in the region could be due to the progressive generation of heterogeneity in the subsurface media as a result of the repetitive occurrence of earthquakes. Our observations of ({Q}_{P}^{-1}) and ({Q}_{S}^{-1}) are comparable with those of other seismically active regions.

Egypt has adopted long-term strategies for urbanizing new cities across the country for resolving the overwhelming problem of overcrowded cities. New Mallawi is an example of such new cities. Egypt is occasionally being shaken by big-sized earthquakes such as the destructive 1754 local Cairo event and the strong ( 1995) Gulf of Aqaba earthquake; therefore, the buildings need to be designed to withstand the shaking of future damaging events. The hazard model of this research is based on an up-to-date catalogue of earthquakes. To obtain realistic results, lateral changes in soil characteristics are considered through using the parameter of average shear-wave velocity in near-surface layers. In addition, different state-of-the-art formulations of declustering temporal and spatial windows are used to represent the epistemic uncertainty in earthquake recurrence functions. Moreover, a set of ground motion models is used for better consideration of the uncertainty of the empirical estimates, and uncertainty in maximum earthquake size is treated. The resulting earthquake shaking is rather low, and the observed variability is affected by lateral changes in soil properties. Various hazard levels are investigated for reaching a better earthquake-resistant design that accounts for damaging earthquakes of hypothetically long return periods. This is done in an attempt to examine low-probability hazards after what was observed in the 2023 Turkey earthquake. The obtained results can be used for urban planning and risk mitigation at the New Mallawi city.

This study presents the calibration of a local magnitude scale ((M_L)) for Brazil, based on seismic data recorded between 2014 and 2023 by the Brazilian Seismographic Network (RSBR) and local networks operated by the Seismological Observatory of the University of Brasília (SIS/UnB). The dataset comprises 1083 seismic events and 155 stations, totaling 7193 vertical-component seismograms. Events were selected at hypocentral distances from 15 to 2400 km and magnitudes ranging from 1.0 to 4.3 (m_R). A fixed depth of 15 km was assumed, consistent with typical intraplate earthquake depths observed in stable continental regions and commonly adopted in previous magnitude scale studies in Brazil. Only events recorded by at least four stations were considered to ensure reliability. All waveforms were standardized by applying the Wood-Anderson instrument response, allowing for the measurement of maximum amplitudes of the S–Lg wave complex. An inversion was performed using the logarithmic and linear dependence of amplitude on distance, resulting in an empirical local magnitude equation (M_{L} = log _{10}(A_{text {WA,mm}}) + 1.0188 log _{10}(R/100) + 0.00062 (R - 100) + C pm S). The derived attenuation curve aligns with those from regions with similar tectonic settings. Station-specific correction terms were also estimated. These corrections enhance the model’s applicability for routine seismic monitoring across Brazil.

The velocity and attenuation structures of near-surface soils play a critical role in controlling ground motions at the Earth’s surface, often exerting an influence disproportionate to their thinness. While numerous passive, in-situ techniques exist for estimating velocity structure, constraining attenuation remains challenging. In this study, we revised the theoretical foundation of the spatial autocorrelation (SPAC) method, originally developed for velocity structure estimation, and formulated an attenuation-aware SPAC coefficient as a theoretical basis for attenuation estimation. Calculations for a representative near-surface model revealed that attenuation effects are expressed in the imaginary part of the attenuation-aware SPAC coefficient. Synthetic data analyses further demonstrated that, while the imaginary part is sensitive to directional aliasing, it remains robust against stochastic and incoherent noise. The effect of directional aliasing was found to be mitigated by increasing the number of observation points, as in the conventional SPAC method.

Seismological studies and historical reports indicate that the Aqaba Gulf region is characterized by significant earthquake activity, impacting surrounding areas such as northern Egypt and northwestern Saudi Arabia. Significant seismic events, including the 22 November 1995 earthquake with local magnitude of 7.3 as reported by the Egyptian National Seismic Network (ENSN), have affected these regions. This study aims to examine the faulting mechanisms and seismic source characteristics through full waveform moment tensor inversion, as well as to assess the ground motion effects using a stochastic simulation model for two moderate-magnitude earthquakes in the Aqaba Gulf region: the 27 June 2015 (Mw = 5.5) and 16 May 2016 (Mw = 5.4) events, occurring in the Aragonese and Dakar basins, respectively. The moment tensor inversion for the Aragonese earthquake revealed a strike-slip fault mechanism, with nodal planes-oriented NNE-SSW and ESE-WNW. In contrast, the Dakar earthquake displayed strike-slip movement with minor dip-slip along NNE-SSW fault planes. The full decomposition of the moment tensors into Double-Couple (DC), Compensated Linear Vector Dipole (CLVD), and Isotropic (ISO) components confirmed a tectonic origin for the earthquakes, with a tensile source type for the Aqaba events. Finally, the strong ground motion parameters are predicted, including the acceleration time history, Peak Ground Acceleration (PGA), Peak Ground Velocity (PGV), Peak Ground Displacement (PGD), and the Pseud Spectral Acceleration (PSA) at the investigated regions (Magna, AlBad’, Tayyib Al-ism, and the Assilah), northwest Saudi Arabia. The ground motion analysis of the 2015 and 2016 earthquakes indicates that shaking intensities were moderate, with PGA and PSA values insufficient to cause major structural damage but capable of producing minor property impacts. These results highlight the localized nature of seismic effects, with higher ground motion recorded closer to the epicenters, emphasizing the importance of site proximity and local conditions in assessing seismic hazard in northwestern Saudi Arabia.

The northeastern region of Algeria is renowned for its significant seismic activity within the North African context. The high seismicity rate in this area can generate a high level of earthquake hazard, emphasizing the pressing need for a reliable seismic hazard assessment to design structures capable of withstanding such catastrophic events. In fact, conducting a seismic hazard assessment, achievable through deterministic or probabilistic methods, requires a comprehensive, updated, unified earthquake dataset. This dataset serves as the primary and essential information source for delineating seismic source zones and comprehensively characterizing them. Accordingly, the main objective of this study is to conduct an improved Probabilistic Seismic Hazard Assessment (PSHA) for the northeast region of Algeria, spanning 35°–37° N latitude and 4°–8,5° E longitude, while establishing a unified earthquake catalog for the region in terms of moment magnitude scale (M_w). New data from national and international agencies and findings from recent scientific publications inform the development of a unified seismicity catalog with the least amount of uncertainties. Subsequently, the catalog is declustered to filter out foreshocks and aftershocks from the earthquake data. This refined catalog is used to delineate five seismic zones and to estimate the seismicity parameters for each seismic source zone. For the seismic hazard calculations, three suitable Ground Motion Prediction Equations (GMPEs) are used for the area under study. The code (REASSESS V2.0), a new software for single and multi-site PSHA, was used to carry out the seismic hazard analysis for the area under study. The PSHA was performed using a standard logic tree methodology, allowing for a systematic consideration of model-based uncertainties and their effects on the estimated ground motion parameters. The hazard is quantified in terms of peak ground acceleration (PGA), short-period (0.2 s), and long-period (1s) spectral acceleration maps, as well as uniform hazard spectra for return periods of 100 and 475 years. The outcomes gained from this study are anticipated to significantly contribute to future land-use planning and the development of facilities and infrastructure in the region.