Microseismic Activity Research Articles

Machine Learning‐Assisted Microearthquake Location Workflow for Monitoring the Newberry Enhanced Geothermal System

AbstractEnhanced geothermal systems (EGS) offer a sustainable energy source but face challenges in accurately locating microearthquakes induced during reservoir stimulation. Locating these microearthquakes provides reliable feedback on the stimulation progress. Current deep learning methods for locating earthquakes require extensive data sets for training, which is problematic as detected microearthquakes are often limited. To address the scarcity of training data, we propose a practical workflow using probabilistic multilayer perceptron (PMLP) which predicts microearthquake locations from cross‐correlation time lags in waveforms. Utilizing a 3D velocity model of Newberry site derived from ambient noise interferometry, we generate numerous synthetic microearthquakes and 3D acoustic waveforms for PMLP training. Accurate synthetic tests prompt us to apply the trained network to the 2012 and 2014 stimulation field waveforms. To enhance the accuracy of source localization, we carefully handpick the P‐arrival times. Predictions on the 2012 stimulation data set show major microseismic activity at depths of 0.5–1.2 km, correlating with a known casing leakage scenario. In the 2014 data set, the majority of predictions concentrate at 2.0–2.9 km depths, consistent with results obtained from conventional physics‐based inversion, and align with the presence of natural fractures from 2.0 to 2.7 km. We validate our findings by comparing the synthetic and field picks, demonstrating a satisfactory match for the first arrivals. By combining the benefits of quick inference speeds and accurate location predictions, we demonstrate the feasibility of using realistic synthetic data set to locate microseismicity for EGS monitoring.

Characterization of Micro-seismic Activity in Northern Cyprus Using Complexity and Corner Frequency Methods

Cyprus is an island country located in the eastern Mediterranean, to the south of Türkiye and the western of Syria and Lebanon, and is a popular tourist destination. Due to being surrounded by seas on all four sides, meticulous planning of rescue, assistance, and evacuation plans is necessary in the face of disasters such as earthquakes and tsunamis. Tectonically, the southern part of the island is controlled by the Cyprus Arc, while the northern part is dominated by the Kyrenia Range. The demand for raw materials for construction and industry is met through controlled quarry blasting operations carried out by open-pit quarry companies in the districts of Kyrenia and Nicosia. As a result, both natural and artificial seismic events occur in the region, and these quakes are documented in seismic catalogs by seismology centers. However, due to the low energy content of micro-seismic events and the inadequacy of seismic stations on the island, the source types of these seismic events can be misidentified in the catalogs. In this context, the study focuses on 122 seismic events with magnitudes between 0.9≤Ml≤2.7 that occurred in Northern Cyprus during the January 2018 - December 2021 period (4 years). The seismic events recorded by the station LFK, operated by Boğaziçi University Kandilli Observatory and Earthquake Research Institute Regional Earthquake-Tsunami Monitoring Center (KOERI-RETMC), were classified using Linear and Quadratic Discriminant Functions based on complexity and corner frequency methods. According to the results obtained, 10 of the 122 seismic events were identified as natural, and 96 were determined to be artificial, resulting in a general success rate of 86.89%. However, classification results for 16 seismic events were inconclusive with the methods used. As a result, more detailed secondary analyses should be conducted to accurately determine the source types of micro-seismic events, and the seismic catalogs should be updated accordingly.

Mechanism of radial chain rockbursts in a deep granite TBM tunnel under the influence of weak band

Radial chain rockbursts (RCRs) in deep-buried TBM tunnels have prolonged durations and can pose significant safety risks. Further research is needed to deepen our understanding of the mechanisms of recurrence of rockbursts under the influence of structure plane. In this study, the geological characteristics, occurrence characteristics, microseismic (MS) activity laws, and rock mass fracture mechanism of several RCRs influenced by weak band were summarized to reveal the mechanism of recurrence of RCRs. Research shows that under the influence of weak band, (1) RCRs tends to occur in the area where the dip angle is large, and the strike intersects with the tunnel direction at a large angle and forms a small angle with the maximum principal stress. The width of the weak band typically ranges from 0.01 to 1 m, which is positively correlated with the final depth of rockburst pit. The strength of the weak band is relatively “soft” compared to the surrounding bedrock. (2) The depth of each rockburst pit in RCRs gradually decreases, the time interval between adjacent rockbursts gradually increases, and the maximum thickness of rock block in each rockburst gradually increases. (3) The maximum MS energy of MS events and the maximum apparent stress gradually decreased in the development process of RCRs. The region of stress concentration gradually extended to deeper zone, and the distance difference between stress concentration region and sidewall during each rockburst occurrence stage gradually decreased. (4) The effect of weak band on RCRs attributes to the excavation induced local stress concentration near the weak band, and the interface between the weak band and surrounding rock controls the boundary of the rockburst body. The “chain effect” of RCRs mainly shows that the promotion effect of the previous rockburst on the subsequent rockburst, by promoting the stress and rock fracture transferring to the deep zone, and the proportion of shear fractures increases.

Microseismic activity characteristics and range evaluation of hydraulic fracturing in coal seam

Accurately evaluating hydraulic fracturing (HF) activity characteristics and impact range is critical to optimizing the HF scheme design. This study proposes an adaptive threshold peak detection method and a wavelet transform-support vector machine (WT-SVM) polarity analysis method based on microseismic (MS) monitoring data from a coal mine in Guizhou, China. The results show that the adaptive threshold peak detection method effectively improves the accuracy and stability of the P-wave initial arrival pickup for low signal-to-noise ratio (SNR) signals. The main frequency band of the MS signal of HF is focused on 150–300 Hz. The improved particle swarm optimization algorithm was used by introducing a random-weight strategy to determine the high-density impact areas of the HF within approximately 20 m on both sides of the fracturing borehole. Density cloud map analysis revealed a small area of the blank zone at the junction of the fracturing section, which provided a basis for fracturing parameter design and scheme optimization. The directional characteristics of the MS waveform were quantitatively described by polarity analysis, where the dip angle distribution ranged from 39° to 49°, the strike angle ranged from 61° to 74°, the slip angle was concentrated near 44°, and the stretch angle ranged from 61° to 74°. The WT-SVM proposed in this study overcame the effects of low SNR signals and the limitations of a small-area station network; 62% of the shear ruptures and 38% of the tensile ruptures of the HF were obtained. Shear cracks were the primary expression of the HF rupture type.

Salinization of groundwater in shale gas extraction area in the Sichuan Basin, China: Implications for water protection in shale regions with well-developed faults

The expanding growth of shale gas development has sparked global concern over water-related environmental issues. However, research on groundwater contamination in shale gas areas in China remains limited, impeding environmentally friendly industry practices. To address this gap, we investigated the Wufeng-Longmaxi shale region in the Sichuan Basin, encompassing both operational and prospective shale gas extraction sites, to assess the effects of shale gas operations on shallow groundwater quality. We found there was no significant correlation between groundwater quality and the minimum distance from the shale gas well pads, and some groundwater samples located far from shale gas well pads, rather than those close to pads, were salinized. These findings suggest minimal impacts from shale gas drilling and hydraulic fracturing. The salinized groundwater samples are characterized by high salinity levels and ion concentrations, and are located near fault zones. The primary source of shallow groundwater salinization was derived from the Triassic formation brines confirmed through the assessment of the sensitivity and conservative mixing models. Faults in the study area were identified as pathways for the upward migration of Triassic brines, evidenced by the proximity of salinized samples to fault zones. However, further investigation is required to ascertain whether shale gas extraction activities have induced the migration of formation brines. The occurrence and reactivation of faults, induced by microseismic activities, may pose an increased risk of groundwater contamination in tectonically complex fault zones during shale gas extraction. Therefore, it is imperative to enhance extraction strategies and technologies, particularly in shale regions with well-developed faults, such as optimizing well placement regulation, controlling hydraulic fracturing scale, and strengthening environmental monitoring. By shedding light on potential environmental ramifications of shale gas extraction, especially in fault-rich regions, our study informs water protection strategies and the sustainable advancement of the shale gas industry.

Construction and Application of Fuzzy Comprehensive Evaluation Model for Rockburst Based on Microseismic Monitoring

Based on the relationship between rockburst and microseismic event indicators, this study proposes that the risk of rockburst in mine working faces, roadways, and even the entire mine should be studied through the “double high” risk evaluation of microseismic events. The 43 evaluation indexes of rockburst were optimized into eight indexes by using the expert scoring method. Considering the eight indexes as the basic events and the “double high” risk of microseismic events as the top event, the “double high” accident tree of microseismic events was established. According to the qualitative analysis results of the accident tree, the microseismic activity evaluation index was determined, and the “double high” risk evaluation index system was constructed for microseismic events. The system included three first-level indicators and eight second-level indicators. The fuzzy hierarchical comprehensive evaluation model was used to evaluate the “double high” risk of microseismic events in the Yanbei Coal Mine. In this paper, a microseismic monitoring and evaluation index model is constructed to simplify the existing evaluation system, which is convenient to effectively establish the connection between microseismic monitoring data and rockburst index and provide important theoretical support for underground monitoring and rockburst prevention.

Research on feature vector construction and classification recognition of typical mine microseismic waveforms

In the realm of mining seismology, the identification of microseismic waveforms is a critical step for enhancing the precision of microseismic event localization. Addressing the complexities and diverse sources of noise in subterranean mining environments, this study has developed a novel methodology for the automated categorization and recognition of characteristic microseismic waveforms. Initially, a comprehensive database for mining-induced microseismic waveforms was established, catering to their non-stationary and nonlinear attributes. Based on the distinct features of different microseismic waveforms, this study introduces an approach for classification and identification that relies on waveform characteristics. A feature selection model, KNN-LOO (K-Nearest Neighbor - Leave-One-Out), was employed to optimize the feature vectors, in addition to the utilization of Principal Component Analysis (PCA) for dimensionality reduction of the feature vectors. Ultimately, an enhanced BT-SVM (Binary Tree-Support Vector Machine) model was developed to accomplish the automated classification and identification of mining-induced microseismic waveforms. Experimental results indicate that, compared to conventional SVM classification methods, the proposed methodology not only exhibits theoretical and mechanistic advancements but also significantly boosts performance in practical applications. Specifically, the average misclassification rate for each category of waveforms is reduced to less than 5%, and the overall accuracy has increased by 20%. This considerably elevates both the accuracy and speed of microseismic waveform classification and recognition, thereby providing valuable insights for monitoring and analyzing microseismic activities within mining operations as well as enhancing the safety protocols for mineral extraction.

The Statistical Fingerprint of Fluid-Injection Operations on Microseismic Activity at the Val d’Agri Oil Field (Southern Italy)

In this paper, we examined the dynamical properties of the fluid-injection microseismicity at the Val d’Agri oil field (southern Italy) by applying different statistical methods to find correlations and common periodicities with injection parameters, such as injected volumes and injection pressure. Two periods of observation were analyzed: (1) from 2006 to 2015 (the first 10 years after the beginning of injection operations), the seismicity was recorded by the seismic network of the ENI company that manages the exploitation of the oilfield; (2) from 2016 to 2018, the seismicity was recorded by a denser seismic network capable of significantly reducing the completeness magnitude. If a significant correlation between seismicity and fluid-injection variables was found in the first period, in the second period, the seismic activity and injection variables were characterized by common periodicities after the reservoir acidification and for injection rates above 1900 m3/day. Finally, we applied and compared two different approaches proposed in the literature to forecast the maximum expected magnitude. The results showed that one of the approaches yielded an estimated maximum magnitude of Mmax = 1.7 ± 0.4, which is consistent with the maximum observed magnitude.

A Study on Disasters Induced by Head-On Ejection in High-Speed Driving under the Influence of Roof Drainage

During the gob-side entry driving of the Jurassic coal seam in Western China, ejection disasters occur under the influence of roof drainage, which rarely appear in the eastern mining area. To address this problem, a method combining theoretical analysis, numerical simulation, and field monitoring was used to study the disaster induced by head-on ejection during speedy driving under the influence of roof drainage in the context of gob-side entry driving of the 2202 auxiliary haulage roadway in a mine. A calculation model for the critical energy conditions for disasters induced by head-on ejection was established. The relationships between the driving velocity and the dynamic and static loads on the driving face and the disasters induced by ejection were clarified under the influence of roof drainage. The results indicate that the energy threshold for ejection-induced disaster is 12.23 kJ, and the elastic energy of the driving face induced only by static load fails to reach the energy threshold. When the driving velocity exceeds 5 m/d, microseismic activity in front of the driving face increases in a stepwise manner, and the influence of dynamic load intensifies. The superposition of accumulated elastic energy induced by static load and the energy transmitted by microearthquake to the driving face exceeds the energy threshold, resulting in the risk of ejection-induced disaster. Based on this, measures such as microseismic monitoring in front of the driving face, advanced pulverized coal monitoring, driving speed adjustment, and advanced pressure relief of large-diameter boreholes are adopted to ensure safe driving.

Coupled Hydromechanical Modeling and Assessment of Induced Seismicity at FWU: Utilizing Time-Lapse VSP and Microseismic Data

The objective of this work is to utilize integrated geomechanics, field vertical seismic profile (VSP) and microseismic data to characterize the complex subsurface stress conditions at the Farnsworth Unit (FWU). The model is based on a five-spot sector model extracted from a primary geomechanical model. The five-spot well injection pattern is characterized by extensive reservoir characterization data, such well logs, extracted cores and borehole geophone data, to facilitate the detailed examination of stress changes and microseismic event occurrences. The study utilizes field vertical seismic volumes acquired from the injection well 13-10A. The seismic volumes successfully provided snapshots of the behavior of the reservoir at distinct times. The use of VSP and microseismic data provided direct and indirect estimates of the dynamic stress changes occurring in the overburden, reservoir and underburden rock formations. In order to illuminate the stress regions and identify rocks that have undergone inelastic failure, microseismic event occurrences were utilized. Microseismic activity has been detected at the FWU; further study of its locations, timing, and magnitude was needed to deduce the nature of the changing stress state. The results of the study revealed that microseismic events were successfully modeled within the Morrow B formation. Moment magnitudes of seismic events were within the same magnitudes for events in the reservoir, suggesting the suitability of the model. The results of the study showed that the computed moment magnitudes for seismic events were insignificant to warrant safety concerns. The study findings showed the usefulness of coupled hydromechanical models in predicting the subsurface stress changes associated with CO2 injection. The knowledge gained from this study will serve as a guideline for industries planning to undertake underground CO2 storage, and characterize the subsurface stress changes.

Study on Microseismic Activity Characteristics of Composite Coal Pillar in 21201 Coal Face of Hulusu Mine

Hulusu coal mine adopts “121 method” for back mining, and the coal pillar between adjacent workings is 30m. 21102 working have been mined for more than 3 years, leaving the auxiliary return airway has not been repaired for a long time which results in the serious deformation. If employs the adjacent 21201 workings to mine, the cost of reusing the 21102 auxiliary return airway is high. Therefore, the mine dug a new return airway 10m away from the mine, and forming a 45m composite coal column with double coal pillars and double return airways. Based on the ARAMIS M\E microseismic monitoring system, the spatial distribution characteristics of microseismic accidents near the composite coal pillar of 21201 coal face are studied, and explore the quantitative relationship among the retrieval rate of 21201 coal face, the frequency, and energy of microseismic accidents, which provided a theoretical basis and technical support for determining the scientific and reasonable retrieval rate of 21201 coal face.

Experimental study on the microseismic response characteristics of coal and gas outbursts

A microseismic monitoring simulation experiment of coal and gas outbursts was carried out to study the microseismic response characteristics of the coal and gas outburst process. The microseismic response characteristics were analyzed from the microseismic waveform and time frequency at each stage of the outburst process. The quantitative description of the outburst process was achieved by the statistical analysis of microseismic signals and the determination of the microseismic energy index of outburst hazards. The results showed that the coal and gas outburst process included four stages: preparation, start, development, and stop, and the outburst moment is at the development stage. The protruding hole in the experiment was an ellipsoid with a small mouth and large cavity. Frequency components were complex and extensive and tended to high frequencies at the prominent stage. The frequency of the microseismic response waveform at the moment of prominence was higher than that before and after prominence, but energy at the moment of prominence was not maximum. Sudden changes in amplitudes at the preparation and development stages had more outbursts, and the maximum equivalent energy at the moment of prominence was as high as 45,000 Hz. Compared with the local high amplitude at the preparation stage, the development stage presents an overall high amplitude. Although the start stage is calmer, it is a critical period for predicting outbursts, and the microseismic activity characteristics of the outburst stage differed significantly from those of the other three stages. Quantitative indices reflecting the risk strength of coal and gas outbursts at the breeding stage were defined from microseismic energy, and the outburst risk indices under experimental conditions were established. When L1 < 3.64, L2 < 0.16, and L3 < 5.74, the coal mass had a weak outburst risk; when L1 > 3.64, L2 > 2.33, and L3 > 5.74, it had strong outburst risks. The research results provide an idea for coal and gas outburst precursor information and dynamic early warning.

Automated classification of seismic signals recorded on the Åknes rock slope, Western Norway, using a convolutional neural network

Abstract. A convolutional neural network (CNN) was implemented to automatically classify 15 years of seismic signals recorded by an eight-geophone network installed around the back scarp of the Åknes rock slope in Norway. Eight event classes could be identified and are adapted from the typology proposed by Provost et al. (2018), of which five could be directly related to movements on the slope. Almost 60 000 events were classified automatically based on their spectrogram images. The performance of the classifier is estimated to be near 80 %. The statistical analysis of the results shows a strong seasonality of the microseismic activity at Åknes with an annual increase in springtime when snow melts and the temperature oscillates around the freezing point, mainly caused by events within classes of low-frequency slope quakes and tremors. The clear link between annual temperature variations and microseismic activity could be confirmed, supporting thawing and freezing processes as the origins. Other events such as high-frequency and successive slope quakes occur throughout the year and are potentially related to the steady creep of the sliding plane. The huge variability in the annual event number cannot be solely explained by average temperatures or varying detectability of the network. Groundwater recharge processes and their response to precipitation episodes are known to be a major factor of sliding at Åknes, but the relationship with microseismic activity is less obvious and could not be demonstrated.

Analysis of Thermally Induced Strain Effects on a Jointed Rock Mass through Microseismic Monitoring at the Acuto Field Laboratory (Italy)

The study of the deformation of rock masses in response to near-surface thermal stresses is nowadays considered crucial in the field of geological risk mitigation. The superposition of heating and cooling cycles can influence the mechanical behavior of rock masses by inducing inelastic deformations that can trigger shallow slope instabilities, such as rockfalls and rock topples. This study reports the main outcomes obtained from the analysis of 20 month long microseismic monitoring at the Acuto field laboratory (Central Italy), where an integrated geotechnical and geophysical monitoring system has been operating since 2015. A preliminary event classification was performed through the analysis of time- and frequency-domain characteristic features of the extracted waveforms. Furthermore, the evolution of the local microseismicity was explored as a function of environmental factors (i.e., rock and air temperature, thermal gradients and ranges, and rainfalls) to highlight potential correlations. The here presented results highlight nontrivial insights into the role played by continuous near-surface temperature fluctuations and extreme thermal transients in influencing the stability of rock masses. In particular, the comparison of monitoring periods characterized by the most intense microseismic activity highlights a peculiar distribution of microseismicity during the heating and cooling phases of the rock mass in relation to different environmental conditions. These behaviors can be interpreted as the consequence of different driving mechanisms at the base of local failures.

Focal Mechanisms and Stress Field Characteristics of Microearthquakes in Baihetan Reservoir in the Downstream Area of Jinsha River

The Baihetan Reservoir was impounded on 6 April 2021, after which the water level rose significantly. Notably, after one week of impoundment, microseismic activities were prominent around the reservoir area, which was highly associated with the water level change. From 6 April 2021 to 31 December 2021, over 7000 microearthquakes were recorded by the seismic stations in the vicinity of the reservoir, including 12 ML &gt; 3 events. The maximum was the 21 December 2021 ML3.9 earthquake in Qiluogou town, Sichuan. The post-impoundment seismic events were clustered in Hulukou town in the Qiaojia Basin, with an overall “Y-shaped” pattern. In this study, taking advantage of the high-frequency waveform matching approach, the pre- and post-impoundment focal mechanism solutions totaling 207 ML &gt; 2 earthquakes are successfully obtained. The impoundment-induced stress change is analyzed, and the iterative joint inversion method is used to invert the stress field. Major results and conclusions include the following: (1) After impoundment, the number of normal fault earthquakes remarkably increased in the reservoir area; (2) Impoundment has led to a vertical compressive stress field and horizontal tensile stress field in the area where microearthquakes occurred. It is necessary to pay close attention to possible moderate-to-strong earthquakes in the future.

Hydrogeochemistry of Fault-Related Hot Springs in the Qaidam Basin, China

Hydrogeochemical characterization studies are regarded as an important method for determining the origin of hot springs. The major elements, trace elements, and stable isotopes of four groups of hot spring water samples and two groups of gas samples collected from the intersection of the Altyn Tagh fault zone and the East Kunlun fault belt were investigated in this study. The hot spring water temperature ranged between 6 °C and 14 °C. The water chemistry types of the hot springs were Na·Mg-Cl, Mg·Na-Cl·HCO3, Na-Cl·SO4, and Na-Cl·HCO3. The δD values ranged from −50.00% to −68.60%, while the δ18O values ranged from −6.90% to −8.60%. The hot spring water was recharged mainly by infiltrating precipitation, with a recharge elevation of 3390~3676 m. The heat storage temperature ranged from 66.7 to 164.9 °C. The circulation depth was estimated to range between 1043 and 2679 m. The strontium isotopic composition of the water samples in response to the main weathering sources comprised carbonate and sulfate. CO2 was the main component in the hot spring gas in the study region, and its content was over 95%. The 3He/4He-R/Ra relationship diagram revealed that the mantle-sourced helium from the Yitunbulake spring was 3.06%. In comparison, that from the Aiken spring was 7.38%, which indicated an intrusion of mantle-source material mixed into the hot springs in the study region. The crustal marine limestone contributed significantly to the carbon inventory of the hot spring gas samples (&gt;75%). The dissolution of the marl aquifer resulted in the release of CO2. Carbon was primarily obtained through metamorphism and hydrothermal reactions in the basement lithologies. The Yitunbulake and Aiken hot springs are found near the intersection of the Arjin and East Kunlun fractures, where the water–rock response is relatively strong and the depth of circulation and thermal storage temperature are both high. This causes relatively high ambient pressure to be released from the deep fluid, resulting in microseismic activity in this region. The continuous observation of Aiken spring water chemistry allows for the monitoring of fracture activity in the region. The results of the study could serve as a foundation for further exploration of the relationship between geothermal water and deep faults, shallow geological formations, hydrogeological conditions, and geothermal resource development in the region.

Rockburst and microseismic activity in a lagging tunnel as the spacing between twin TBM excavated tunnels changes: A case from the Neelum-Jhelum hydropower project

A reasonable spacing between tunnels is very important in deeply buried tunnels, as it may mitigate the risk of rockburst occurrence. This spacing is particularly important for tunnels excavated using a tunnel boring machine (TBM), as the methods available for controlling rockburst occurrence are inflexible and limited using these machines (compared to tunnels excavated using the drilling and blasting (D&B) method). In this paper, a rare example was studied in situ in which the spacing between twin tunnels was changed to mitigate the rockburst risk during TBM excavation. First, the characteristics of the rockburst and microseismic activity were investigated in three sections of the lagging tunnel before, during, and after the spacing between the tunnels was changed (we concentrated on the lagging TBM696 tunnel, which is part of the Neelum-Jhelum hydropower project). We observed a significant occurrence of rockburst and microseismic activity in the lagging tunnel at the original spacing of the twin tunnels in both the sandstone and siltstone lithologies, and thus a consideration of enlarging the spacing between the twin tunnels is reasonable. A more detailed investigation was then conducted to explore the effects of tunnel spacing on rockburst occurrence and microseismicity. The rockburst frequency and microseismic radiated energy due to excavation unloading generally decrease as the spacing between the twin tunnels increases (substantially at first and then more slowly as separation increases). In particular, the microseismic radiated energy exhibits a power-decreasing type in both the sandstone and siltstone. Increasing the spacing beyond a certain point, therefore, exerts a very limited effect on the microseismic radiated energy and occurrence of rockbursts (increments of 16–21 m and 15–20 m in sandstone and siltstone, respectively). Consequently, in practice, a very reasonable approach would be to enlarge the spacing of twin tunnels by 22.48 m. Thus, microseismic monitoring is a useful technique for assessing the rockburst risk and deriving a reasonable value for tunnel spacing. This study can be used as a guideline to derive reasonable tunnel spacings when using not only TBMs but also D&B.

Towards fast machine-learning-assisted Bayesian posterior inference of microseismic event location and source mechanism

SUMMARY Bayesian inference applied to microseismic activity monitoring allows the accurate location of microseismic events from recorded seismograms and the estimation of the associated uncertainties. However, the forward modelling of these microseismic events, which is necessary to perform Bayesian source inversion, can be prohibitively expensive in terms of computational resources. A viable solution is to train a surrogate model based on machine learning techniques to emulate the forward model and thus accelerate Bayesian inference. In this paper, we substantially enhance previous work, which considered only sources with isotropic moment tensors. We train a machine learning algorithm on the power spectrum of the recorded pressure wave and show that the trained emulator allows complete and fast event locations for any source mechanism. Moreover, we show that our approach is computationally inexpensive, as it can be run in less than 1 hr on a commercial laptop, while yielding accurate results using less than 104 training seismograms. We additionally demonstrate how the trained emulators can be used to identify the source mechanism through the estimation of the Bayesian evidence. Finally, we demonstrate that our approach is robust to real noise as measured in field data. This work lays the foundations for efficient, accurate future joint determinations of event location and moment tensor, and associated uncertainties, which are ultimately key for accurately characterizing human-induced and natural earthquakes, and for enhanced quantitative seismic hazard assessments.

Seismic activities before and after the impoundment of the Xiangjiaba and Xiluodu reservoirs in the lower Jinsha River

The lower Jinsha River basin is located at the junction of Sichuan and Yunnan provinces in Southwest China, a region with intense tectonic movements and frequent moderate to strong seismic activities. Cascade hydropower stations have been constructed along the lower Jinsha River since 2012. However, research on the effect of the impoundment of large-scale cascade reservoirs in a river basin on local seismic activities is currently lacking. Accurately identifying earthquake locations is essential for studying reservoir-induced earthquakes. Analyzing the spatiotemporal migration process of seismic activities based on complete and precise earthquake relocation is fundamental for determining the fluid diffusion coefficient, constructing fault models for reservoir areas, identifying earthquake types, exploring earthquake mechanisms, and evaluating seismic hazards. The seismicity pattern in the Xiangjiaba and Xiluodu reservoir areas, where seismic activities had been weak for a long time, has changed with the successive impoundment of the two reservoirs, showing microseismic events and seismic clusters. We investigated the spatiotemporal characteristics of seismic activities in the Xiangjiaba and Xiluodu reservoir areas using the waveform cross-correlation-based double-difference relocation technique and the b-value analysis method. We discovered that seismic events after the impoundment of these two reservoirs exhibited different characteristics in different regions. The seismic activities at the Xiluodu dam quickly responded to the rising water level, with the seismic intensity decaying rapidly afterward. These events were concentrated in the limestone strata along both sides of the Jinsha River, with a shallow focal depth, generally within 5 km, and a high b-value of approximately 1.2. Such features are close to those of karst-type earthquakes. Microseismic activities frequent occur on the eastern bank of the Yongshan reservoir section downstream of the Xiluodu dam, with two parallel NW-trending earthquake strips visible after precise earthquake relocation. The MS5.2 earthquake near Wuji town on August 17, 2014, had prominent foreshocks and aftershocks distributed in a clear NW-trending 20-km-long strip, perpendicular to the riverbank. These seismic events had a low b-value of approximately 0.7. The orientation of the node plane revealed by the strike-slip focal mechanism of the mainshock is consistent with that of the strip formed by the foreshock-mainshock-aftershock sequence, indicating the existence of a NW-striking concealed fault. Seismic activities near the Yanjin-Mabian fault upstream of the Xiangjiaba reservoir area since 2013 were concentrated in a NW-trending strip, with several near EW-trending seismic clusters on its western side, and with the largest event having a magnitude of ML3.7. So far, the impoundment of the Xiangjiaba and Xiluodu reservoirs has not triggered seismic activities on the large Jinyang-Ebian and Yanjin-Mabian faults nearby.

Long-term triggered seismicity on the Mosha fault by Damavand volcano, Iran: Implications on the seismic hazard of Tehran metropolis

The seismic history of the Mosha fault, the largest and most active fault of Eastern Tehran metropolis, and its relation to the Damavand active volcano, the highest mountain in the Middle East, is investigated. We deduce that the central Mosha, near the Damavand, has a higher seismicity than either its western or eastern segments. On 7 May 2020, an Mw 5.1 earthquake occurred on the central Mosha, about 40 km east of Tehran and 10 km southwest of the Damavand crest, and it was felt intensely in Tehran. Its rupture is imaged and located in a region that presented a relative seismic quiescence compared to its eastern and western parts, during the last 14 years, suggesting its partial locking and heterogeneous distribution of fault frictional strength on this segment of Mosha. Its significant directivity to the west is confirmed by the mainshock rupture model, its PGA distribution, and distribution of early aftershocks. The rupture model suggests a relatively small stress drop of 2.6 bar, which is consistent with the comparatively high rupture dimension of 9 km for a Mw 5.1 earthquake, and indicates the easy rupture expansion on the central Mosha near the Damavand Volcano. The central Mosha experienced earthquakes in 1930, 1955, and 1983, as well as high microseismic activity and the 2020 seismic sequence, all of which strongly point to a possible influence of the Damavand Volcano on the seismicity of the central Mosha. This is corroborated by the observation of hydrothermal zones on the Mosha fault and the extension of a sill-like Damavand young magma chamber until central Mosha in tomography studies. We propose that the existing heat may increase the pore pressure on the fault, which lowers the effective normal stress, facilitates the nucleation-expansion of the rupture, and unclamps the fault. Damavand could act as a fuse and nucleate earthquakes, and if the rupture extends toward the west, it could have a significant directivity effect on low-frequency seismic waves that reach Tehran without attenuation and affect tall structures. In addition, high site amplification for frequencies up to 16 Hz due to the deep sedimentary basin, mainly in the mid-city of Tehran, will be remarkable for short buildings.