The classification of seismic events is crucial for monitoring underground nuclear explosions and regional unnatural seismic events. To classify tectonic earthquakes, explosions, and mining-induced earthquakes, we established 36- and 60-dimensional network-averaged datasets and single-station datasets through feature extraction and spectral amplitude analysis. Using different artificial intelligence (AI) methods, including the support vector machine (SVM), extreme gradient boosting (XGBoost), long short-term memory network (LSTM), residual neural network, and long short-term memory fully convolutional network (LSTM-FCN), we constructed two-class and three-class models, analyzed the change in the classification with epicentral distances, and evaluated the generalizability of different classifiers. The results showed that the accuracies of different AI models with the feature extraction dataset were higher than those achieved with the spectral amplitude dataset, indicating that the feature extraction method can more clearly highlight the differences between different types of seismic events. The accuracies with the network-averaged dataset were 5%–8% higher than that achieved using the single-station dataset. The earthquake and mining-induced earthquake classifiers constructed by different AI methods had the best performance, followed by the earthquake and explosion classifier, and the explosion and mining-induced earthquake classifier, with average accuracies of 97.4%–98.4%, 96.5%–97.6%, and 88.8%–90.6%, respectively. In the model generalization evaluation, the test accuracies and F1-Scores of the two-class models with the 36-dimensional network-averaged dataset exceeded 90%. Among the five AI methods, XGBoost and LSTM both performed well in classification of different datasets, indicating that these models have good application prospects for seismic event classifications.

Classification of low‐magnitude seismic events is a challenging issue for the Comprehensive Nuclear‐Test‐Ban Treaty. Path correction of the P/S amplitude ratio is the key to identifying earthquakes and explosions. In this article, the Bayesian Kriging interpolation method is used to conduct the path correction of P/S amplitude ratios and recognition of low‐magnitude seismic events. Based on a total of 5677 small earthquakes and 1769 explosions in Beijing and its adjacent areas, the Bayesian Kriging method is used to establish the path correction surface and uncertainty surface of Pg/Lg amplitude ratios measured within different frequency bands at five seismic stations, and path correction of amplitude ratios is conducted for all events. The results show that the correction surface is consistent with the observed amplitude ratios, which can reflect the differences in their propagation paths to a certain extent. The root mean square variation of the amplitude ratio is reduced by a maximum of 30% and the misclassification probability is reduced by a maximum of 8.5% after the Kriging correction. The high‐frequency Pg/Lg amplitude ratios can effectively classify low‐magnitude events, and the misclassification probability at each station is less than 15% and 10% based on high‐frequency Pg/Lg of $>7$ and $>9Hz$⁠, respectively. Of the five stations, BJT (Baijiatuan, Beijing) has the best recognition, with the misclassification probability being lower than 5% after Kriging correction based on high‐frequency Pg/Lg (⁠ $>9Hz$⁠). The classification ability of high‐frequency amplitude ratios (⁠ $>15Hz$⁠) is weakened due to high‐frequency noises. Bayesian Kriging correction can reduce the variance in the amplitude ratio of low‐magnitude seismic events and hence effectively improve the ability to classify small‐magnitude events, which has an important reference value for regional seismic monitoring and identification.

On 23 February 2023, an Mw 6.8 earthquake struck the border of Tajikistan and Xinjiang China, the source mechanism remains controversial according to different seismic inversions. To better comprehend the source characteristics and the surface deformation pattern, we used the ascending and descending orbital Sentinel-1A SAR data to obtain the coseismic deformation of this earthquake based on the traditional two-pass differential interferometric synthetic aperture radar (InSAR). The source model is inverted from the InSAR coseismic deformation results. The possible Coulomb Failure Stress (CFS) transfer is analyzed based on the preferred source model. The results illustrate that the earthquake ruptured a blind left-lateral strike-slip fault of strike 28.1° with a maximum slip of 1.53 m and the total geodetic moment is 1.99 × 10¹⁹ N·m (Mw 6.83). The strike direction and the fault characteristics suggest the Seismogenic fault is a secondary fault of the Sarez–Karakul Fault System. The 2015 Mw 7.2 Sarez Earthquake plays a triggering role in the occurrence of the 2023 Tajikistan earthquake. Earthquake hazard on Sarez–Karakul Fault System and Sarez–Murghab Thrust System is enhanced due to the Coulomb stress loaded by the Tajikistan earthquake.

The 3 September 2017 nuclear test conducted by North Korea was followed by a series of small to moderate seismic events, which poses a challenge for the discrimination of potential future nuclear tests. This study analyzes a data set comprising 773 earthquakes and explosions recorded in North Korea and adjacent regions from 2009 to 2023. Various methods, including the P/S amplitude ratio, high- and low-frequency energy ratio, corner frequency, waveform duration, complexity, zero-crossing rate, cepstral complexity, and instantaneous frequency complexity are employed to extract 36D feature data sets in the vertical, north-south, and east-west components. Support vector machine and gradient boosting decision tree methods are used to construct earthquake and explosion classifiers. The classifiers are applied to six underground nuclear tests and seismic events that occurred around the North Korean test site (NKTS) and compared with the east-central China classifiers. The results show that the accuracies of earthquake and explosion classifiers in North Korea and adjacent regions can reach 93%. All six underground nuclear tests are classified as explosions, and the probability of classifying the seismic events around NKTS as earthquakes is only 71%. Finally, a comprehensive analysis of the seismic events around NKTS is conducted. These events exhibit distinctive regional seismic signature features, differing from those of typical local earthquakes and explosions. Further investigation into these events has significant implications for the monitoring of small-scale underground nuclear tests.

The $Ms$ 5.1 Tangshan earthquake that occurred on 11 July 2020 was recorded by the infrasound array DQS at a source‐to‐receiver distance of 196 km and an azimuth of 278°. Relatively high‐apparent velocity, celerity of epicenter infrasound, and propagation modeling indicate a stratospheric arrival. The backprojection of the epicentral infrasound detection delineates a northeast direction‐extended infrasound radiation region with a long axis of about 57 km and a short axis of about 28 km, which is in good agreement with the fault rupture direction. The robustness of the infrasound radiation source location is also tested. The consistency between the spatial distribution of the radiation source and the epicenter confirms that the extended radiating zone around the epicenter excited infrasound signals lasting 2.5 min at a distance of 196 km. A comparison between the backprojected sound pressure level and peak surface pressure derived from measured seismic ground motions shows some disagreement, which may involve propagation path and/or topography around the epicenter. Therefore, these factors should be taken into account when locating the sources of infrasound and estimating the distribution of ground motion. This study shows the ability of infrasound arrays to detect, locate, and research moderate earthquakes and help reconstruct the seismoacoustic coupling in the source region and fault rupture. Moreover, the complexity of the infrasound radiation source and seismoacoustic coupling mechanism is recognized. It is expected that infrasonic observations of moderate earthquakes will increase with the number of infrasound arrays deployed worldwide. Our study is also helpful to the scientific design and optimal layout of infrasound monitoring systems, which will allow for the detection of earthquakes in addition to explosive sources.