During an earthquake,the energy released by vertical crustal movement propagates to ionospheric altitudes in the form of acoustic or gravity waves,triggering co-seismic iono-spheric disturbances(CID).Studying CIDs holds significant practical value,including enhan-cing earthquake early warning capabilities,deepening the understanding of lithosphere-atmo-sphere-ionosphere coupling mechanisms,and comprehensively assessing the scope and sever-ity of seismic hazards.GNSS-TEC(Global Navigation Satellite System-total electron content)is one of the key observational methods for investigating seismic ionospheric effects.Influ-enced by seismic parameters(e.g.,magnitude,focal mechanism)and non-seismic factors(e.g.,epicentral topography,solar activity,geomagnetic activity,atmospheric wind fields),CID-induced TEC disturbances(CID-TEC)exhibit notable variability and diversity in their dis-turbance patterns,morphology,and time-frequency characteristics.Accurate classification of CID-TEC is fundamental to precisely analyzing disturbance propagation velocity,direction,and source localization.However,under conditions of sparse observational data,data sparsity can prevent the natural formation of a linear time-epicentral distance relationship in travel-time diagrams when fitting CID propagation speeds.Additionally,the fitting process may suffer from the omission of similar disturbances or contamination by dissimilar disturbances,leading to deviations between the fitted and actual propagation velocities—or even generating spurious velocities in extreme cases. Based on the differential characteristics of CID-TEC,this paper proposes a CID-TEC dif-ferentiation and analysis method combining cross-correlation and continuous wavelet transform.Specifically,cross-correlation analysis quantifies the degree of differentiation between distur-bances,while wavelet analysis characterizes the time-frequency properties of CID-TEC,enabling their differentiation and classification.The method is validated through three case stud-ies.Results show that the cross-correlation coefficients between CID-TEC signals of the same propagation mode are significantly higher than those between different modes,indicating that cross-correlation analysis can quantitatively measure differentiation and qualitatively determine whether disturbances share the same propagation mode.The time-frequency features provided by continuous wavelet transform further depict the energy distribution of CID-TEC in the time and frequency domains,identifying the acoustic-gravity wave frequency range responsible for exciting CID-TEC. Applying this method,we investigate the CIDs triggered by the MW7.2 Alaska earthquake on July 16,2023.After differentiation analysis and classification,spurious propagation modes caused by spatiotemporal(epicentral distance)overlaps of partial CID-TEC signals are elimi-nated.Three acoustic wave propagation modes are identified,with corresponding velocities of 741.85 m/s,690.29 m/s,and 680.38 m/s,and the potential dominant propagation directions of CID-TEC are determined.
The SmS phase,generated by reflection of downgoing S waves at the Moho discon-tinuity,was initially identified in only a limited number of earthquakes.Its observability is con-strained by several factors,including source characteristics,propagation path,site response,and network density,so the distance range over which it can be clearly recorded is limited.With the continued densification of strong-motion networks and the accumulation of high-qual-ity records,it has gradually become clear that,although the SmS phase is a common seismic phase,a prominent high-amplitude SmS arrival is not observed in every earthquake.When well developed,SmS often exceeds the direct Sg phase in amplitude,so the peak value of a strong-motion record may be controlled more by the reflected phase than by the direct arrival.This pa-per reviews the current understanding of the effects of SmS on ground motion,with emphasis on reported observations,controlling factors,implications for ground-motion prediction,and recent progress in China. Reported observations from North America,Europe,and Asia indicate that SmS can sig-nificantly enhance recorded amplitudes within specific epicentral distance ranges.The 1989 Loma Prieta earthquake remains a classic example:strong-motion levels in parts of the San Francisco Bay area exceeded those predicted by standard attenuation relations and were later linked to a strongly developed SmS phase.Similar effects have since been reported in eastern North America,the Marmara region,the Taipei basin,the Po Plain of northern Italy,south-western China,and northeastern China.These cases show that the influence of SmS is wide-spread but strongly region dependent.Its manifestation varies among regions and records:in some cases,SmS can be recognized as a later phase,whereas in others it is reflected mainly in enhanced peak amplitudes and localized flattening of attenuation curves,even when the phase itself is not clearly separated. The development of SmS is controlled primarily by crustal structure and variations in the Moho discontinuity.When an earthquake source is located above the Moho,downgoing S-wave energy is partitioned into reflected and transmitted components at the crust-mantle bound-ary.The strength of SmS therefore depends fundamentally on crustal and uppermost mantle structure,because these parameters govern the reflection and transmission coefficients at the Moho.As epicentral distance increases,the incidence angle at the Moho changes,and once re-flection approaches or enters the critical regime,reflected energy can increase markedly,allow-ing SmS to develop into a strong and coherent phase.This explains why strong SmS is usually observed only within a limited distance range.Variations in crustal complexity and Moho geo-metry further influence this process.In simple and laterally homogeneous crust,less energy is lost through repeated reflection,transmission,and scattering at internal interfaces,so SmS is more likely to preserve coherent waveforms and strong amplitudes.In contrast,structurally complex crust and irregular Moho geometry can redistribute seismic energy into multiple wave components and shift the distance window for critical reflection,thereby weakening or obscur-ing SmS.Other factors mainly modulate,rather than fundamentally control,the observed char-acteristics of SmS.Source depth affects the incidence angle of the downgoing S wave at the Moho,so deeper crustal earthquakes are generally more favorable for strong SmS development over shorter critical-distance ranges.Focal mechanism may introduce azimuthal differences in amplitude through source radiation effects,and attenuation structure may alter the amplitude ra-tio between direct S and reflected SmS because the two phases follow different ray paths.Site response adds another layer of complexity at the recording stage,because the amplification as-sociated with SmS depends not only on local site conditions but also on the frequency relation-ship between the reflected phase and site resonance. Beyond its seismological significance,SmS is also important from an engineering per-spective because its impact ultimately depends on whether the reflected wavefield is strong enough to modify peak ground motion.When SmS develops into a strong phase,the reflected energy may equal or exceed that of the direct S-wave arrival over specific distance ranges,thereby altering the expected attenuation pattern and,in some cases,controlling the peak amp-litude of the record.This makes SmS directly relevant to earthquake engineering,because peak ground motion is a key parameter in seismic hazard assessment,ground-motion prediction,and performance-based design.In regions where SmS is strongly developed,neglecting its contri-bution may lead to systematic underestimation of actual shaking levels.This issue also reveals an important limitation of conventional ground-motion prediction equations(GMPEs).Most classical GMPEs represent path effects using simple distance-dependent attenuation functions,sometimes combined with magnitude scaling,and therefore cannot explicitly account for Moho reflections and related complex propagation effects.As a result,ground motions within the characteristic distance range of strong SmS development may be underestimated by standard models.This has motivated the introduction of segmented geometrical-spreading terms and oth-er path-sensitive modifications in several regional studies. The problem is particularly relevant in China.With the rapid expansion of the national strong-motion network,increasing observations have shown that SmS can exert strong control on attenuation behavior and peak ground motion in regions such as Sichuan-Yunnan and the Bo-hai Rim.These findings indicate that SmS should be considered not only in the physical inter-pretation of strong-motion records,but also in regional seismic hazard assessment,ground-mo-tion modeling,and engineering applications.Taken together,current studies suggest that SmS is not merely a phase-identification issue,but a regionally significant path effect that should be incorporated more explicitly into future ground-motion analyses.
The Sichuan-Yunnan block is one of the most active regions in the world in terms of tectonic and seismic activities.The frequent large earthquakes occurring on the boundary faults,whereas the activity and seismicity of the intra-block faults have not attracted enough attention.At the same time,these intra-block faults also have the potential to trigger destructive earth-quakes.In addition,intra-block fault activity causes cascading response of boundary faults,and intra-block faults can accommodate kinematic inconsistency between boundary fault motions,which affects regional crustal deformation patterns.Therefore,it is important to understand intra-block fault activity and associated surface-rupturing earthquake features for regional seismic hazards assessment and tectonic activity analysis. During an active fault investigation in the Liangshan area of the Sichuan-Yunnan block,a distinct lithological boundary named Caimashui fault was found.The NE-striking Caimashui fault is a intra-block fault within the Sichuan-Yunnan block,with a total length of about 80 km.Located 12 km to the southeast of Huidong County,the fault generally strikes NE.It lies on the west side of the northern segment of the Zemuhe fault zone and Xiaojiang fault zone,runs roughly parallel to the Ninghui fault,and represents a relatively continuous and well developed structure.However,due to the complex geological conditions and poor accessibil-ity,the formation age,active characteristics and slip rate of the Caimashui fault are still unknown.In this paper,we conduct comprehensive analyses including remote sensing inter-pretation,field investigation,unmanned aerial vehicle(UAV)topographic mapping,trench excavation and Quaternary dating,to clarify the geometric distribution of the Caimashui fault and quantitatively investigate its Late Quaternary active characteristics. Field geological investigation shows that the activity of the Caimashui fault varies signific-antly on both sides of Haiba township.In the northeastern segment on the east side of Haiba township,the Caimashui fault is an Early-Middle Pleistocene fault with a length of 24 km,while in the southwestern segment extending from Haiba township to Xinfa township,the Cai-mashui fault is a Holocene fault with a length of 58 km.Typical dextral strike-slip landforms are well developed along the southeastern segment of the Caimashui fault,including displaced river terraces,drainage systems,mountain ridges,and alluvial fans,which collectively indicate that the Caimashui fault has been dominated by dextral strike-slip since the Late Quaternary. Based on the high-resolution UAV mapping and dating results,the dextral strike-slip rate of the Caimashui fault can be quantitatively constrained.Field investigations show that,to the northwest of Xiaochacun,the Caimashui fault has produced obvious synchronized dextral dis-placement of ridges and gullies,and the morphology of the fault scarps inside the fault trough is clear,with faults scarps distributed in the form of geese.There is no trace of artificial modifica-tion,and the fault scarp ponds possess favorable depositional conditions and are rich in organic materials,allowing complete preservation of paleoearthquake records.Therefore we excavated the Xiaochacun trench perpendicularly to the fault scarp.The trench exposure reveals that the Caimashui fault has experienced at least three paleoearthquake events since 30 ka:event E1(29002-20312 BC),event E2(9774-6919 BC),and event E3(2730-1526 BC).The fault has the potential to generate earthquakes with M≥7.0.In the trench,the stratigraphic unit U1 is a black peat layer deposited within an abandoned plug pond.Using the age obtained from the top of stratigraphic unit U1 and the gully dextral dislocation of 22.8 m,the average dextral slip rate of the Caimashui fault since the Late Pleistocene is estimated to be(0.85±0.01)mm/a,although this value may be slightly higher than the actual long-term slip rate.The study of the active tectonics around the Sichuan-Yunnan block shows that the left-lateral strike-slip rate of the Xiaojiang fault zone is much greater than that of the Xigeda-Yuanmou fault,resulting in westward extrusion and counterclockwise rotation of the central Yunnan micro-block.To ac-commodate such rotational motion,intra-block regions develop compressional deformation at the block margin.In addition,the left-lateral shear deformation of the block causes the intra-block faults to undergo dextral slip and compressional motions,accommodating regional strain partitioning.
On December 18,2023,a significant MS6.2 earthquake struck Jishishan County,Gansu Province,China,resulting in heavy casualties and severe infrastructure damage.This earthquake occurred within a seismically active northeastern margin of Qinghai-Xizang Plat-eau,characterized by complex fault systems including the Lajishan south margin fault and the Lajishan north margin fault.The event provides a valuable case for studying coseismic deforma-tion,fault mechanics,and potential precursor signals using advanced remote sensing tech-niques.This study combines interferometric synthetic aperture radar(InSAR)and thermal infrared remote sensing technology to comprehensively analyze the coseismic deformation field,source parameters,fault slip distribution,Coulomb stress changes,and spatiotemporal characteristics of thermal infrared anomalies of the Jishishan earthquake. Coseismic deformation fields of the 2023 Jishishan earthquake were derived from Sentinel-1A synthetic aperture radar(SAR)data of both ascending and descending tracks.Differential InSAR(D-InSAR)processing results revealed a dominant uplift pattern with maximum line-of-sight(LOS)displacements of approximately 6.5 cm and 7.2 cm for the ascending and des-cending tracks,respectively.The deformation field,spanning approximately 18 km in the east-west direction and 22 km in the north-south direction,is mainly distributed between the Lajishan south and north margin faults,without clear surface rupture observed,indicating a blind thrust fault mechanism. To constrain the fault geometry and slip distribution,a Bayesian inversion framework was implemented using the Geodetic Bayesian Inversion Software(GBIS),incorporating quadtree-sampled InSAR data.Nonlinear inversion results indicate that the seismogenic fault is character-ized by a length of(12.85±0.35)km,awidth of(7.8±0.2)km,atop depth of(5.5±0.3)km,a strike of 324°±1°,and a dip angle of 32°±2°.The fault exhibits a predominantly thrust mechanism with a minor right-lateral strike-slip component.Finite fault slip inversion,constrained by the downsampled InSAR data,reveals that the rupture was confined to depths between 10 km and 20 km,with a maximum slip of 0.54 m occurring at approximately 14 km depth.The total seismic moment released is estimated at 1.29× 1018 N·m,corresponding to a moment magnitude MW6.0,consistent with solutions from international agencies such as USGS and GCMT. Coulomb stress change analysis was conducted at depths of 5,10,15,and 20 km using a layered viscoelastic crustal model.The results indicate significant stress loading(ΔCFS＞0)on several segments of the surrounding faults.Specifically,positive Coulomb stress changes are observed along the entire Lajishan south margin fault,the WNW-trending segment of the Lajishan north margin fault,and the NNW-trending segment to the south of the epicenter.This suggests an elevated seismic hazard potential in these regions,necessitating ongoing monitor-ing and risk assessment. Complementing the InSAR deformation analysis,this study investigated pre-seismic thermal infrared anomalies using brightness temperature data from the Fengyun-2G(FY-2G)geostationary meteorological satellite.The relative power spectrum method was applied to the nighttime brightness temperature data to extract thermal infrared anomalous signals.The results reveal significant pre-seismic thermal infrared anomalies in the epicentral region.These anom-alies exhibit notable spatiotemporal consistency with the deformation processes inferred from time-series InSAR,reflecting the process of strain energy accumulation and release.Spatially the thermal infrared anomalies were primarily concentrated along the Lajishan fault zone.We hypothesize that these thermal anomalies were primarily induced by stress-induced micro-fracturing,which enhances the release of deep subsurface gases or geothermal energy,fol-lowed by a closure of microfractures as the local stress field intensifies prior to the mainshock. Time-series InSAR analysis,utilizing both persistent scatterer InSAR(PS-InSAR)and small baseline subset InSAR(SBAS-InSAR)techniques,was performed on a stack of 21 Sentinel-lA images covering the period from April 2023 to January 2024.The results delineate the pre-,co-,and post-seismic deformation evolution.The pre-seismic period(January to September 2023)was characterized by minor and gradual deformation,with a maximum cumu-lative displacement of approximately 15 cm,interpreted as pre-seismic creep.A rapid and sig-nificant increase in deformation occurred around the epicenter from October 2023 to January 2024,with displacements exceeding 10 cm,which is coincident with the mainshock.Post-seismic deformation rates subsequently decreased,indicating a relaxation phase. The integration of InSAR-derived deformation fields and thermal infrared observations provides robust evidence for a deformation-thermal coupling mechanism associated with the Jishishan earthquake.The consistency between the spatial patterns of pre-seismic thermal in-frared anomalies and the coseismic deformation field,coupled with the calculated stress load-ing on adjacent fault segments,further demonstrates the application potential of multi-source remote sensing data in earthquake monitoring and earthquake hazard assessment. In conclusion,this study demonstrates that the 2023 Jishishan MS6.2 earthquake was a blind thrust event with a significant uplift component.The seismogenic fault did not rupture the surface,with slip concentrated at depths of 10-20 km.Pre-seismic thermal infrared anomalies were detected and show correlation with the deformation process.Coulomb stress transfer has increased the seismic risk on several segments of the Lajishan fault system.The synergistic use of InSAR and thermal infrared remote sensing proves to be a powerful approach for elucidating the complex processes of earthquake preparation,occurrence,and post-seismic adjustment,offering valuable insights for understanding seismic hazards in similar tectonic settings.
Basins represent a special type of site condition.Their complex assemblage of sedi-mentary soils and basin basement structure can produce substantial seismic amplification,known as basin effect.At the same time,the complex shape of the basin basement and the loose sediments within the basin can lead to ground motion focusing effects and the basin-edge effects,accompanied by a substantial increase in earthquake duration.Consequently,build-ings located within basins often suffer severe damage during earthquakes.However,many large and medium-sized cities worldwide are situated in basins,such as Beijing,Xi'an,Yin-chuan,Kunming,Chengdu,Fuzhou in China,Tokyo and Osaka in Japan,Los Angeles in the United State,Mexico city in Mexico,Umbria in Italy.Many scholars at home and abroad have emphasized that the study on earthquake damage mechanism of basins should be the focus in urban earthquake prevention and disaster reduction. This paper analyzed the basin effect on ground motion in Dayao basin using microtremor tests and numerical simulations.The study aims to reveal ground motion characteristics of small-scale basins and provide a theoretical basis for rational determination of ground motion parameters of engineering structures in Dayao basin.In the ground pulsation test,data denois-ing was performed using baseline correction,digital filtering,superposition average and other methods.The standard spectral ratio method(HS/HR)was adopted to analyze the spectral ratio of the ground pulsation records after processing at each observation point on soil sites within Dayao basin,from which the predominant frequency and site spectral ratio amplification coeffi-cient at each observation point were obtained.In the numerical simulation,the 2D model of Dayao basin was analyzed using the dynamic finite element method,and the dynamic amplific-ation coefficients and spectral characteristics at each characteristic point were obtained. In the ground pulsation observation test,the standard spectral ratio results for each obser-vation point show that the predominant frequencies of measurement points SP1-SP3 are mainly concentrated in the high frequency band 6-9 Hz,and amplification coefficients in E-W direc-tion are greater than those in N-S direction,with E-W values ranging from 2.5 to 4.2 and N-S values from 1.1 to 1.3.Measurement points SP4-SP6 exhibit similar characteristics:their pre-dominant frequency are also concentrated in the high frequency band of 6-9 Hz,and the E-W amplification coefficient(2.5-4.2)is greater than the N-S coefficients(1.1-1.3).For SP7 and SP8,the predominant frequencies are mainly distributed in the low frequency band of 1-3 Hz,and the E-W and N-S amplification coefficients are relatively close.In addition,the magnifica-tion coefficients differ somewhat between the E-W and N-S directions,and the spectral ratio curves show multiple peaks. In the numerical simulation analysis,the spectral ratio characteristics at each observation point indicate the dominant frequency of 1-3 Hz and 6-9 Hz,with the most prominent peak occurring at 6-9 Hz.The amplification coefficients vary among different observation points,with a maximum of 20.8.The magnification coefficients also differ between the two subbas-ins,with softer soil yielding larger values.Furthermore,the amplification coefficient of the steeper basin margin is greater than that at the gentler basin margin. The spectral ratio curves of each soil layer in Dayao basin show multi-peak characterist-ics,indicating that the site does not consist of a single soil layer.The predominant frequencies at most observation points within the basin are concentrated in 6-9 Hz,and certain differences exist among these frequencies,reflecting the variations in site stiffness at different observation points.The amplification coefficient showed a certain difference between the E-W and N-S dir-ections,with a maximum of 4.2,indicating the anisotropy of the site.In the numerical simula-tion analysis,the amplification coefficients reveal a distinct basin-edge effect and the ground-motion focusing effect.The different amplification coefficients of the two small basins reflect the significant influence of the soil layer structure on ground motion,and the amplification coefficient is larger than that of the soft soil layer.The amplification coefficient at the steeper basin margin is greater than that at the gentler margin,demonstrating that the slope gradient of the basin edge has a significant effect on the ground motion near the basin boundary.The res-ults indicate that the HS/HR method has good reliability in basin effect analysis,and the com-bined application of the microtremor observation and numerical simulation can reveal the mech-anism of basin effect more effectively. The different amplification coefficients between the two small basins reflect the significant influence of soil layer structure on ground motion,with thicker or softer soil layers correspond-ing to larger amplification.
Previous studies have shown that sedimentary basins significantly amplify seismic waves and prolong ground motion duration,thereby intensifying earthquake damage.The Kanto basin,as one of Japan's most densely populated urban regions,exhibits particularly strong ground motion amplification effects due to its thick sedimentary layers and complex geo-logical setting.This study applies the generalized inversion technique to strong motion observa-tion data from the Kanto Basin and its surrounding areas in Japan.By extracting and separating the S-wave Fourier amplitude spectra,we obtain seismic source parameters,quality factors,and site responses,with a focus on elucidating the differences in site response inside and out-side the basin as well as their relationships with site parameters such as vS30 and the sedimentary layer depth Z3.2.Furthermore,we compare the site amplification results derived from the gener-alized inversion method and the horizontal-to-vertical spectral ratio(HVSR)method to assess their applicability across different frequency ranges. A strong-motion dataset covering the period from 2010 to 2019 was constructed,including 1 194 earthquake events and 74 strong-motion stations,with moment magnitudes MJMA ranging from 3.0 to 6.5 and source-to-site distances less than 200 km.Using the two-step non-parametric generalized inversion method,we first separate the path attenuation term,and then apply a reference station constraint to extract source and site terms.The main findings are as follows: 1)The derived source spectra conform to the ω2 model in the 0.3-24 Hz frequency range.An empirical relationship between seismic moment M0 and comer frequency fc was established:lgM0=(22.817±0.797)-31gfc,corresponding to an average stress drop of 14.9 MPa.The es-timated frequency-dependent S-wave quality factor is Q(f)=67f1.19. 2)Correlation analysis of site response and vS30 at different frequencies shows that,for stations inside the basin,site response and vS30 are moderately to strongly correlated in the low-frequency range(0.5-2.0 Hz),while this correlation weakens significantly at higher frequen-cies.For stations outside the basin,the frequency range of moderate to strong correlation ex-tends up to 8 Hz.These results indicate that vS30 is limited in its ability to represent basin site ef-fects,whereas it performs better outside the basin.The sensitivity of site response to vS30 dif-fers between the two regions:the correlation peaks near 1 Hz inside the basin and around 4 Hz outside the basin. 3)Empirical relationships between site response and sediment depth Z3.2 were established.Stations outside the basin show better correlation in the low-frequency range,whereas stations inside the basin exhibit low correlation across all frequencies,with a sign reversal near 3 Hz.In the low-frequency range(0.3 Hz and 1 Hz),most basin stations have higher site responses than those outside.In both regions,site response generally increases with increasing depth Z3.2.However,at high frequencies(5.0 Hz and 10.0 Hz),site response inside the basin decreases with increasing depth Z3.2,especially at 10 Hz,reflecting the attenuation effect of thick sedi-ments on high-frequency ground motion. 4)The HVSR method systematically underestimates site amplification,particularly in the low-frequency range(＜1 Hz),and this underestimation is more pronounced for stations inside the basin than those outside. These findings reveal the complexity of site response within the basin,which limits the effectiveness of using single site parameters to construct empirical models.The study identifies distinct site amplification characteristics between intra-basin and extra-basin stations in the Kanto region,and offers insights that may inform future investigations of basin effects on ground motion in other regions such as China.
Ground motion prediction models,also known as ground motion prediction equa-tions(GMPEs)or attenuation relationships,utilize mathematical formulations to estimate the median values and variability of ground motion parameters.They serve as a critical component in engineering practices such as seismic zoning and seismic safety evaluations for major engin-eering sites.The attenuation characteristics of ground motion parameters exhibit regional spe-cificity and are closely related to the tectonic and geological context of the earthquake source and the affected area. In seismic hazard analysis,it is common to consider the probability of engineering struc-tures being subjected to ground motions exceeding a specified threshold.Since ground motions generated by small earthquakes typically do not significantly impact engineering structures,the primary focus of engineering seismology is often on relatively intense and destructive earth-quakes.Certain regions often occur induced seismicity due to human production activities,such as oil and gas extraction or geothermal field operations.Induced earthquakes often occur at shal-low depths.Even small to moderate seismic events can generate relatively high-frequency ground motions in near-field areas,which may cause destructive effects on low-rise residential buildings. The attenuation characteristics of ground motions induced by human activities differ signi-ficantly from those of typical shallow-crustal earthquakes.In China,research on ground mo-tion attenuation laws related to energy extraction-induced seismicity started late in progress.Most existing studies focus on reservoir-induced seismicity caused by water impoundment,while research on ground motion attenuation laws and prediction models for seismicity induced by fluid injection activities,such as shale gas hydraulic fracturing,remains young.Therefore,developing ground motion prediction models that account for shallow,small-to-moderate in-duced earthquakes holds significant engineering importance for regions frequently affected by industrial extraction-induced seismicity. Methods The prediction model presented in this study adopts surface-wave magnitude as the mag-nitude parameter and hypocentral distance as the distance parameter.It employs a linear mag-nitude term combined with a near-field saturation factor,which effectively captures near-field distance saturation effects.The model is capable of accounting for depth influence while main-taining consistency with commonly used engineering models and seismic hazard analysis frame-works in China. Taking the southern Sichuan region of China as an example,this area belongs to a moder-ate-to-strong seismic zone where small to moderate shallow crustal earthquakes predominantly occur.First,strong-motion data from the NGA-West2 database were collected and compiled.Using the western United States as a reference region,the intensity attenuation relationship for moderate-to-strong seismic zones in China was re-regressed based on recently supplemented in-tensity data.Subsequently,the intensity analogy method was employed to establish a baseline prediction model for peak ground acceleration suitable for moderate-to-strong seismic zones.By incorporating seismological records of small to moderate earthquakes in southern Sichuan,the baseline model was refined,resulting in the development of a horizontal peak bedrock accelera-tion prediction model for shallow small to moderate earthquakes in this region.Results and discussions Data from the high-density broadband seismic network around the industrial areas of south-ern Sichuan,where seismic activity has been relatively frequent in recent years,were utilized.The dataset covers earthquakes with surface-wave magnitudes ranging from 4.0 to 6.0,recor-ded between January 2015 and February 2021.A total of 93 free-field bedrock events and 2246 broadband records were selected,with hypocentral distances extending up to 200 km.These data were used to refine the baseline prediction model for moderate-to-strong seismic zones.The final revised model for horizontal peak bedrock acceleration in southern Sichuan is applic-able within the magnitude range of 4.0-6.0 and distance range of 0-200 km. The revised prediction model shows good agreement with the regression data.Comparison with the limited number of local bedrock strong-motion records indicates that the results of this study are reasonably conservative.When compared with internationally commonly used ground-motion models for oil and gas field induced seismicity,the results of this study are sim-ilar in the near-field for earthquakes below magnitude 5.0.However,given the broader mag-nitude-distance range covered by the dataset in this study,the proposed model is expected to of-fer greater reliability for higher-magnitude events. Conclusions This paper proposes a method for developing ground motion prediction models tailored to shallow,small-to-moderate earthquakes closely linked to industrial activities.Taking the indus-trial extraction region in southern Sichuan as a case study,a horizontal peak bedrock accelera-tion prediction model was recalibrated,applicable for surface-wave magnitudes ranging from 4.0 to 6.0 and distances up to 200 km.The new model maintains consistency in form with com-monly used engineering ground motion prediction models while incorporating depth effects.It effectively extends the lower magnitude limit of existing models and supports refined regional ground motion attenuation predictions. In the future,more dense and reliable arrays of strong-motion and broadband seismic sta-tions will be able to support further research.
Vertical seismic profiling(VSP)benefits imaging by recording along borehole depth,but reliable imaging requires separating upgoing and downgoing energy.We compare three classical approaches—f-κ filtering,median filtering,and Radon transform-on synthetic VSP data from the Marmousi model,with and without added noise.Radon achieves the clean-est separation in complex media and shows the best noise robustness,at the expense of higher computational cost.f-κ filtering is efficient and effective for simple stratification,but suffers from aliasing and edge artifacts when sampling is marginal or dips are complicated.Median fil-tering is practical and good for impulsive noise,yet depends strongly on first-arrival picking and window selection,which degrades performance under low SNR or wavefield overlap.We summarize trade-offs,provide quantitative metrics,and outline parameter choices to guide method selection under varying structural complexity and noise levels. In the VSP data processing workflow,wavefield separation is a crucial step.Since seis-mic waves generate multiple wave types during underground propagation,including upgoing waves and downgoing waves,the mixing of these wavefields seriously affects the final imaging quality.To further optimize the utilization efficiency of seismic data and improve imaging qual-ity,accurately separating upgoing waves and downgoing waves has become a key technical link.Effective wavefield separation not only improves the signal-to-noise ratio of seismic data but also provides high-quality input data for subsequent processing steps such as velocity analys-is and migration imaging. This article focuses on the comparative study of three classic wavefield separation meth-ods:f-κ filtering,median filtering,and Radon transform.These three methods each have their theoretical foundations and applicable ranges,showing different advantages and limitations un-der different geological conditions and data quality. The core idea of the f-κ filter is to utilize the different apparent velocity characteristics of upgoing waves and downgoing waves in the f-κ domain,separating wavefields propagating in different directions by designing appropriate filters.f-κ filters have the advantages of high com-putational efficiency and simple implementation.The median filter method mainly relies on ac-curate picking of first arrivals,achieving separation of upgoing and downgoing waves by ana-lyzing the arrival time characteristics of wavefields.Median filtering has significant advantages in processing impulse noise and can effectively suppress anomalous values in data.The Radon transform method is based on the sparse representation characteristics of wavefields in the Radon domain,achieving wavefield separation by transforming seismic data to the τ-p domain(time-ray parameter domain).This method can better handle complex wavefields,especially in the presence of multiples and noise interference,where the Radon transform shows strong ro-bustness.However,its computational process is relatively complex and requires more computa-tional resources. To evaluate the application effects of these methods under complex geological conditions,we used the Marmousi velocity model for finite-difference forward modeling to generate seis-mic data containing complex structural features.The Marmousi model is a widely recognized standard test model in the geophysical community,whose complex geological structures and velocity distributions can well simulate various complex situations in actual geological environ-ments. Through systematic comparative analysis,we found that the three wavefield separation methods each have their characteristics and applicable ranges:The Radon transform shows high-er separation accuracy when dealing with complex wavefields,especially in the presence of strong noise and multiple interference;its robustness is significantly superior to the other two methods.This benefit is mainly attributed to the Radon transform's ability to better distinguish wavefield components with different propagation directions and apparent velocities in the τ-p domain.The f-κ filter has significant advantages in computational efficiency,capable of quickly completing wavefield separation tasks and is suitable for large-scale data processing.However,in complex wavefields,f-κ filters are prone to wavefield aliasing phenomena,which affect separation effectiveness.This aliasing mainly occurs when wavefield components are complex and frequency content is rich,limiting its application effectiveness in complex geo-logical environments.Median filtering largely depends on accurate picking of first arrivals,which requires data to have a high signal-to-noise ratio and clear first arrival characteristics.When noise is strong or first arrivals are not obvious,the performance of median filters will sig-nificantly decline.Additionally,this method also has wavefield aliasing problems,resulting in generally poor overall separation performance.However,median filters perform excellently in processing impulse noise and can effectively suppress anomalous interference in data. For complex structures and strong noise environments,the Radon transform method shows stronger robustness,capable of maintaining high separation accuracy under complex condi-tions.The median filter method has significant effects in processing impulse noise and is suit-able for situations where VSP data has serious noise interference.In such application scenari-os,median filters can effectively suppress anomalous values and improve data quality. This study not only provides a comprehensive discussion and empirical analysis of wave-field separation methods in VSP technology but also provides important reference for selecting optimal separation strategies in complex geological environments in the future.Through sys-tematic comparative research,we have developed a thorough understanding of the advantages and limitations of various methods,providing a scientific basis for practical applications.
Deep-borehole seismic observation has become an increasingly important compon-ent of modern earthquake monitoring systems due to its outstanding capability for suppressing surface environmental noise,its proximity to earthquake source regions,and its suitability for long-term in situ observations of crustal processes.Conventional surface seismic stations are strongly affected by anthropogenic activities,meteorological disturbances,and near-surface heterogeneities,particularly in densely populated and highly urbanized regions.These factors significantly limit the detection of weak seismic signals,including microearthquakes and fore-shocks,and constrain the reliability of earthquake early warning and source characterization.By deploying high-sensitivity instruments at depths ranging from several hundred meters to sev-eral kilometers,deep-borehole observation effectively isolates sensors from near-surface noise sources and provides high signal-to-noise ratio(SNR)recordings across a broad frequency band,offering a unique observational window into earthquake generation processes and deep crustal dynamics. This paper presents a comprehensive review of the development history,international and domestic research progress,key technologies,and future directions of deep-borehole seismic observation.Representative international projects are systematically summarized,including the San Andreas Fault Observatory at Depth(SAFOD)in the United States,the International Ocean Discovery Program(IODP)long-term borehole monitoring systems,deep-borehole ob-servatories in Japanese subduction zones(e.g.,NanTroSEIZE),and the German Continental Deep Drilling Program(KTB).These projects demonstrate that deep-borehole observatories play a critical role in resolving fault-zone physical properties,monitoring microseismicity and slow-slip events,and constraining stress evolution and fluid-fault interactions under extreme temperature and pressure conditions.International experience highlights the importance of multi-parameter integration,long-term stability under harsh environments,and real-time data transmission for both scientific research and hazard mitigation. In contrast,deep-borehole seismic observation in China has developed rapidly over the past two decades,supported by major national initiatives such as the Chinese Continental Sci-entific Drilling(CCSD)project and the Wenchuan Fault Scientific Drilling(WFSD)project.Although a nationwide borehole seismic network with nearly 200 stations has been established,most existing stations are relatively shallow(typically less than 400 m),which limits their noise suppression capability.Recently,significant progress has been achieved through the construc-tion of 2 km-class deep-borehole observation platforms,including the Changde deep-borehole ultra-broadband seismic observation platform and the Leizhou(Zhanjiang)2000 m deep-bore-hole optoelectronic composite observation station.These platforms mark a major milestone in China's deep-borehole seismic monitoring capacity and provide critical testbeds for advanced instrumentation and engineering solutions. A key contribution of this review is the establishment of a four-dimensional technical and methodological framework for deep-borehole seismic observation,encompassing sensors,packaging and coupling,acquisition and timing,and data processing and analysis.Within this unified framework,the performance parameters of typical domestic and international projects-such as usable bandwidth,dynamic range,depth adaptability,and long-term stability-are summarized using consistent criteria,forming a comparable reference baseline.In particular,the successful engineering application of the domestically developed very-broadband seismo-meter TDE-120VB,with a response band from 120 s to 50 Hz,a dynamic range exceeding 145 dB,and proven long-term operation at depths greater than 2000 m,represents a critical breakthrough in the localization of deep-borehole seismic instrumentation in China. From an engineering perspective,this paper further proposes a recommendatory paradigm for repurposing abandoned energy wells into deep-borehole seismic observatories.Based on practical experience from the Changde platform,key procedures and acceptance criteria for full-chain sealing from wellhead to borehole bottom,grouting consolidation,sensor packaging and coupling,retrievable deployment,and operation and maintenance are systematically summar-ized.This paradigm provides a feasible pathway toward low-cost,reproducible construction of deep-borehole observation stations,addressing one of the major bottlenecks that currently limits large-scale deployment. In terms of data processing and analysis,deep-borehole observations offer unprecedented opportunities for detecting weak seismic events and resolving fine-scale source processes.However,traditional detection and phase-picking methods often struggle with the complex and time-varying noise characteristics of borehole environments.Recent advances in artificial intel-ligence and deep learning,such as PhaseNet,EQTransformer,and physics-informed neural networks,show great potential for improving weak-event detection,phase identification,and source parameter inversion when combined with high-quality deep-borehole data.The integra-tion of seismic observations with other physical fields,including strain,temperature,fluid pressure,and distributed fiber-optic sensing,further enhances the ability to characterize deep crustal processes in a holistic manner. Despite these advances,several challenges remain.Long-term stability of sensors under extreme temperature and pressure conditions,high drilling and operation costs,and the lack of standardized metadata and quality control frameworks continue to hinder the expansion of deep-borehole seismic networks.Addressing these challenges requires coordinated efforts in materi-als science,drilling engineering,system integration,and intelligent data processing.Looking forward,the future development of deep-borehole seismic observation is expected to emphas-ize retrievable and modular platform design,edge-computing data acquisition and processing,physics-constrained artificial intelligence methods,and unified standards for data sharing and interoperability. In summary,deep-borehole seismic observation represents a powerful and indispensable approach for advancing earthquake science and improving seismic hazard assessment.By syn-thesizing international experience and recent domestic progress,and by proposing unified tech-nical frameworks and engineering paradigms,this review aims to provide a reference baseline for future deep-borehole observatory construction and operation.The continued integration of advanced instrumentation,intelligent data analysis,and network-scale deployment will play a key role in promoting the large-scale application of deep-borehole seismic observation and fos-tering interdisciplinary collaboration in Earth system science.
This study uses the global ionospheric total electron content(TEC)grid data provided by the Jet Propulsion Laboratory(JPL)of the International GNSS Service(IGS)Cen-ter to conduct analysis of electromagnetic anomalies.Based on geographic coordinate informa-tion,the nearest TEC grid data to the MS5.7 earthquake in Songyuan,Jilin Province(45.27°N,124.71°E)in 2018 was selected for calculation and analysis.Since pre-seismic ionospheric anomaly regions are not all concentrated above the epicenter,they also occur in re-gions at a certain distance from the epicenter.Therefore,the research grid points were selected from the nine closest grid points around the epicenter. Through comparative calculation of the traditional quartile method and the sliding quartile method,we find that the sliding quartile method is less susceptible to special phenomena than the traditional quartile method in the calculation results.It can more objectively reflect the actu-al background of the observation points,yield more distinct results,and is consistent with pre-vious studies.By applying the sliding quartile method,we conduct time series analysis on the four study grid points closest to the Songyuan earthquake epicenter and find that the impact of polar substorms on ionospheric observations(on days-11 and-5 relative to the earthquake)can be eliminated.The ionospheric observations first increased from day-14 to 8 and then de-creased from Day-8 to 1 in the week before the earthquake,which is consistent with the study by Wang et al(2014).Meanwhile,the amplitude of ionospheric anomalies increases as the epi-central distance decreases.As Songyuan is located further north,it is more susceptible to polar magnetic substorms.After excluding the effects of space weather,there is a high probability that positive ionospheric anomalies persist for five days and negative ones for two days,which are associated with the earthquake. The analysis of seismic ionospheric spatial distribution anomalies shows that:① The closer the northern ionospheric disturbances are to the intense phase of polar substorms,the stronger they become;② Pre-seismic TEC observations first increased,then decreased,and rebounded in the post-seismic period,with the anomaly amplitude increasing as the epicentral distance decreases;③ In terms of ionospheric spatial distribution,positive and negative anom-alies are temporally clustered and exhibit regular characteristics,with an overall"N"-shaped pattern.Anomalies over and in the vicinity of the epicenter persisted for more than two hours,characterized by a small spatial coverage,low intensity and distinct localization. The lithosphere-atmosphere-ionosphere coupling(LAIC)mechanism governs energy propagation during seismic gestation and occurrence.By combining analyses of deep litho-spheric resistivity,atmospheric and ionospheric datasets for the Songyuan earthquake,we identify a close correlation among the three spheres during the seismogenic and coseismic peri-ods:fracturing of the lithospheric medium,energy release and electrical variations exert a coupling effect on the atmosphere and ionosphere.Negative ionospheric TEC anomalies are likely directly related to changes in the atmospheric electric field caused by negative charges re-leased during rock rupture and upward electron transport via the vertical atmospheric electric field.Positive TEC anomalies mainly result from the release of positive charges in rock masses,thereby accelerating collisional ionization of atmospheric ions and altering the ion density distribution in the ionosphere.The atmospheric thermosphere and ionosphere also ex-hibit mutual influence and synchronous variations—particularly,high-amplitude positive iono-spheric TEC anomalies show good synchronization with atmospheric thermal infrared anom-alies in terms of intensity enhancement and spatial distribution patterns. Currently,the application of pre-seismic ionospheric detection in China is mainly concen-trated in the western regions for earthquakes with MS≥6.0,with few studies conducted in east-ern China,where earthquakes are fewer and weaker.In recent years,Songyuan has become one of the most seismically active areas in Northeast China.This study analyzes the largest earthquake in Songyuan over the past five years:the MS5.7 Songyuan earthquake in Jilin Province on 28 May 2018(with an epicentral depth of 10 km).It presents the application of a short-term ionospheric TEC prediction method in Songyuan,supplementing the seismic data of eastern China for ionospheric anomaly-based seismic prediction and accumulating experience for the application of pre-seismic ionospheric detection.
Using the waveform data of the 2022 Luding MS6.8 earthquake sequence recorded by the mobile stations and the regional permanent network deployed by Sichuan Earthquake Agency,we performed computational analysis on S-wave splitting of local earthquakes,and investigated the spatial distribution and temporal evolution characteristics of upper crustal aniso-tropy parameters at each station,along with the spatiotemporal variations in fast S-wave polar-ization direction and slow S-wave time delay.The results show that the spatial distribution of the fast wave direction and time delay exhibits obvious zoning characteristics and a certain trend change over time,implying that the upper crust anisotropy is controlled by both spatial location and time-dependent stress field.On the Xianshuihe fault zone,the fast wave polarization direc-tions of the stations such as L5512,XXXJ and LDXX are all NW-SE,which is consistent with the strike of the Xianshuihe fault zone,indicating that the main compressive stress field and fault structure in the region may act together on the crustal anisotropy.At the same time,the fast wave direction of the stations AJW,NTW and L5137 located near the Daduhe fault zone is NE-SW,which is consistent with the strike of the Daduhe fault zone,indicating the significant impact of the fault structure on the anisotropy of the upper crust.For the stations in the Luding source area,the average slow S-wave time delays are higher than those in the surrounding regions,indicating that stress accumulation level in the source zone was far more substantial than in peripheral areas during the seismogenic period of the 2022 Luding MS6.8 event.This phenomenon reflects the differences in stress field changes and crustal response between the source area and the surrounding areas,further indicating that the stress concentration level in the source area is high,which may provide the driving force for seismic activity.The average post-seismic time delay of the stations L5512 and L5513 located in the north of the source area is(4.219±0.469)ms/km,the average post-seismic time delay of the stations CNXJ,AJW and NTW located in the central part of the source area is(6.118±1.878)ms/km,and(4.812±1.880)ms/km for the stations BLG,XMC and SMI located in the south of the source area.And the slow S-wave time delay of stations in the north,south and central parts of the source area was low before the main shock,and then increased rapidly after the main shock.This indicates that the aftershock activity of the Luding MS6.8 earthquake may continue,and the release and adjustment of the stress field may take some time.In addition,slow S-wave time delays at stations in aftershock-rich areas remain elevated above pre-seismic levels,implying that stress release and adjustment in these regions will persist for a period of time.
Active faults undergo creep deformation during the interseismic period.Fault creep is a slow but persistent process that lasts for hundreds or thousands of years,accumulating dis-placements ranging from several meters to tens of meters over hundreds of years.Fault creep can directly induce or exacerbate landslide hazards,particularly when the fault zone traverses a landslide body,as fault activity directly controls slope stability.Such hazards frequently occur in the Sichuan-Yunnan region,where the Xianshuihe fault zone,a typical active fault in the area,exhibits significant creep deformation,with a left-lateral slip rate reaching 9-11 mm/a.Creep-induced landslides in this region are closely correlated with fault activity.Furthermore,the long-term fault creep-induced accumulated displacement poses a significant threat to engin-eered structures that cross active faults,such as transportation tunnels and water conduits.For example,the Claremont water tunnel crossing the Hayward fault was offset by 0.33 m due to creep.The Berkeley Hills tunnel section of the San Francisco Bay Area Rapid Transit(BART)system,crossing the active Hayward fault,was observed to experience 8 cm of right-lateral off-set between 1971 and 1980.Scientifically fault creep deformation prediction is crucial for the seismic resilience of engineering structures and the mitigation of natural disasters. Numerous researchers have studied fault creep deformation,proposing various computa-tional methods such as the boundary element method(BEM)and three-dimensional boundary integral method.These methods,however,typically model the elastic response of fault slip within a half-space model,neglecting the viscoelastic effects of crustal materials.This omis-sion can lead to significant errors in long-term creep simulations.Employing viscoelastic mod-els is more appropriate for studying postseismic displacement.The semi-analytical three-dimen-sional elastic model for half-space deformation(hereafter referred to as the Smith3D model)can simulate both stress accumulation on the locked segment of a fault during the interseismic peri-od and the viscoelastic response of the deep crustal medium.Research results indicate that fault stiffness varies during the interseismic period,a phenomenon explainable by the rheological properties of fault-zone materials.While these analytical methods and computational models rely on the assumptions that the fault's elastic modulus remains constant,they conspicuously fail to capture modulus changes over extremely long timescales. Uniaxial rock creep experiments reveal a three-stage creep process:decelerating creep(stage I),steady-state creep(stage Ⅱ),and accelerating creep(stage Ⅲ).Throughout these stages,rock stiffness evolves with time.Using uniaxial rock creep experimental data and the Maxwell creep model,we propose a modified Maxwell creep model incorporating variable stiffness.The correlation coefficients between the theoretical model results and the experimen-tal data for mudstone and greenschist creep were 0.974 and 0.983,respectively,validating the model's effectiveness.Integrating this variable-stiffness model into the Smith3D model,we simulated the long-term creep deformation of the Xianshuihe fault zone from 1990 to 2015.For this simulation,we selected five representative horizontal creep observation sites along the fault zone,ordered from northwest to southeast:namely,Xialatuo,Qiajiao,Goupu,Longdengba,and Laoqianning.The correlation coefficients between the improved model's simulations and monitoring results were 0.995 6,0.882 1,0.954 3,0.982 9,and 0.931 5,respectively,demonstrating the efficacy of the variable-stiffness Smith3D model for analyzing long-term fault creep deformation. The influence of three nearby earthquakes,the 2001 Western Kunlun Mountains earth-quake(MS8.1),the 2008 Wenchuan earthquake(MS8.0),and the 2014 Kangding earthquake(MS6.3),on the creep displacement of the Xianshuihe fault zone was analyzed.Although the epicenter of the 2001 Western Kunlun Mountains earthquake was distant from the Xianshuihe fault zone,the induced plate motion still caused long-term reversed slip at the Qiajiao site.Cou-lomb stress changes on the fault zone resulting from the 2008 Wenchuan earthquake led to an in-crease in creep rate at the Goupu site.The 2014 Kangding earthquake,with its epicenter relat-ively close to the fault zone,produced co-seismic displacements causing short-term reversed slip at the Laoqianning site.The results from these three earthquakes indicate that fault creep deformation on the Xianshuihe fault zone is associated to both moderate-to-strong earthquakes in the near field and large earthquakes in the far field.
As the forefront of the plateau's intracontinental expansion,the eastern margin of the Qinghai-Xizang Plateau,marked by significant topographical and geomorphological contrasts,is characterized by complex tectonic deformation and high seismicity.This region is critical for investigating the formation processes of continental strong earthquakes and serves as an ideal location for studying subsurface structural evolution and geodynamics.Therefore,the high resolution 3D P-wave velocity structures of the crust and upper mantle on the eastern margin of the Qinghai-Xizang Plateau is very important for understanding the uplift and deformation mechanisms of the plateau,as well as the crust-mantle structural evolution.However,there are obvious differences in the results of previous studies due to the different seismic stations data and methods used in these studies.For instance,debates continue regarding the subduction morphology,front location of the Indian lithospheric slab and the formation mechanisms of the Tengchong volcano. In order to improve the accuracy of crust-mantle structure inversion in the eastern margin of the Qinghai-Xizang Plateau and its surrounding regions,the data used in this study are derived from two different sources,including 249 permanent broadband seismic stations of China Digital Seismograph Network,and 1 694 temporary broadband seismic stations deployed by international and domestic scientific expedition.Rigorous data screening was performed based on the rationality of arrival times,resulting in the selection of 200 737 P-wave arrival times from 4 377 local seismic events and 1 378 547 relative travel time residuals from 18 902 tele-seismic events.A joint inversion of these local and teleseismic data was conducted using three-dimensional P-wave tomography,yielding a detailed P-wave velocity structure within a depth range of 0 to 900 km beneath the eastern margin of the Qinghai-Xizang Plateau.With the check-board tests,the scale of anomalies 0.5°×0.5°×100 km could be recovered perfectly in most study area.The tomographic results with good resolution illuminate that the three-dimensional velocity structure is robust and reliable. The three-dimensional P-wave velocity model reveals distinct deep structures within the study region.A high-vP anomaly exists at 0-300 km in depth beneath the Sichuan Basin,rep-resenting the preserved cratonic lithosphere from early geodynamic processes of the Earth.Clear images of the Indian mantle lithosphere(IML)reveals the different subducted angles and northern limits between the west side and east side of the IML.The IML goes downdip to the Bangong-Nujiang suture zone at the west side,whereas the subduction of the IML extend much further at the east side,hinting the possible tearing of the IML,which corresponds to NS-trend-ing rifts.In contrast,beneath the Myanmar Arc,a continuous high-velocity anomaly is obser-ved in the upper mantle,extending significantly deeper than the Wadati-Benioff seismic zone and penetrating into the mantle transition zone.This anomaly reaches eastward up to approxim-ately 100°E,potentially representing the eastward subduction of the Indian lithospheric plate.The observed discrepancy between the termination depth of seismicity and the penetration depth of the slab beneath Myanmar is hypothesized to be due to the failure of the physical mechan-isms responsible for deep-focus earthquakes.A distinct high-velocity anomaly is identified at the base of the mantle transition zone beneath the southeastern margin of the Qinghai-Xizang Plateau.This anomaly is not connected with the high-velocity structure to the west,which rep-resents the subducting Indian slab,and may represent delaminated lithosphere in the southeast-ern margin of the Qinghai-Xizang Plateau.Additionally,a distinct low-vP anomaly exists beneath the Tengchong volcano at 0-300 km in depth,suggesting that the Tengchong volcano originates from slab dehydration and mantle convection induced by the subduction of the Indian plate.
The southwestern margin of the Sichuan basin,particularly the Changning region,is characterized by highly complex geological structures with intricate and variable geometries.The dominant tectonic unit is the Changning-Shuanghe anticline,and the direction of the max-imum principal stress is NWW-SEE.The structure of the anticline is broad in the east and nar-row in the west,and is further dissected by secondary folds and faults.Historically,back-ground seismicity in this region was relatively weak,with magnitude of most events below 3.0.However,in recent years,the area has experienced frequent moderate-to-strong earthquakes.For instance,on 16 December 2018,an MS5.7 earthquake struck Xingwen County,Sichuan,causing multiple injuries and widespread secondary disasters.Subsequently,on 17 June 2019,the region experienced the largest earthquake ever recorded in its seismic history—the Changn-ing MS6.0 event,which filled the long-standing absence of earthquakes exceeding magnitude 6.0 in this area and was followed by numerous aftershocks above magnitude 5.0.This se-quence has made Changning one of the most seismically active regions within the Sichuan basin,drawing broad attention from both the scientific community and the public. Despite these developments,most existing studies have emphasized single-parameter ana-lyses(e.g.,stress field or seismicity)and have seldom conducted integrated investigations of multiple factors.In particular,comprehensive studies jointly examining source parameters and seismic activity to elucidate the regional stress evolution remain limited.To address this gap,the present study investigates earthquakes with MS≥3.0 that occurred between October 2010 and July 2022 in the Changning area(104.6°E-105.2°E,28°N-28.5°N).We employ the CAP full-waveform inversion method to derive focal mechanism solutions and apply the damped least-squares inversion to resolve regional stress field characteristics.In addition,we extract key source parameters such as stress drop and corner frequency through a multi-taper spectral in-version approach.By combining these parameters with the spatiotemporal evolution of the b-value,we systematically explore the coupling mechanisms between seismic activity,industrial operations,and tectonic stress perturbations.The ultimate objective is to clarify the seismogen-ic processes and dynamic evolution of stress,thereby providing a scientific basis for seismic hazard assessment in southern Sichuan. The results demonstrate that the dominant focal mechanisms are thrust-faulting and strike-slip types,with epicenters mainly concentrated on the western margin of the tectonically active zone.The stress field inversion indicates that the maximum principal stress is oriented NWW-SEE,consistent with the extrusion regime of the eastern Qinghai-Xizang Plateau.Al-though the regional stress field is generally stable,a short-term disturbance occurred following the 2019 mainshock.Temporal variations in b-value and stress drop exhibit a stage-wise coup-ling,reflecting a three-phase stress evolution cycle of"loading-release-reloading."Specific-ally,the period of 2010-2014 corresponds to a loading stage;2015-2019 represents a release stage characterized by frequent ruptures;and the post-2019 period indicates reloading,during which stress gradually recovered and the b-value stabilized(Fig.2).The stress inversion res-ults(Fig.7)broadly agree with the background tectonic regime,further supporting the inter-pretation that most earthquakes(primarily thrust and strike-slip events)in the region are gov-erned by the regional stress field. Furthermore,clear correlations are identified between source parameters and earthquake magnitude:stress drop and source radius increase with magnitude,whereas corner frequency decreases,implying that larger earthquakes correspond to greater rupture dimensions and en-hanced energy releases.Integrated analysis suggests that the evolution of source parameters and seismicity in the Changning region is controlled primarily by two factors.First,the complex and densely distributed fault system,together with the intersection of anticline and syncline structures,produces heterogeneous stress fields that facilitate localized ruptures.Second,the seismic activity may be associated with large-scale industrial fluid injection during unconven-tional energy development,whereby fluids infiltrate hidden fault zones adjacent to reservoirs,altering effective fault stress conditions and triggering earthquakes.Additionally,strong earth-quakes themselves can induce local stress readjustments and promote concentrated energy re-lease. Overall,this study provides new insights into the coupling between tectonic stress,indus-trial activities,and seismic processes in Changning.The findings not only contribute to a more robust assessment of regional seismic hazards but also offer important implications for under-standing the potential impact of fluid injection on fault system evolution.
On 6 August 2023,an MS5.5 earthquake struck Pingyuan County,Shandong Province,China.This event represents the largest earthquake in Shandong Province since the MS5.9 Heze earthquake on 7 November 1983.Investigating its seismogenic structure and after-shock sequence is therefore of significant importance.Large earthquakes are commonly accom-panied by abundant aftershocks,and the rapid detection and accurate location of these events are essential for understanding seismogenic processes and assessing seismic hazards. In recent years,with the rapid growth of seismic observational data,artificial intelli-gence-based aftershock detection methods have been widely adopted to improve detection effi-ciency.In particular,deep learning-based phase picking approaches have demonstrated clear advantages over traditional techniques.Continuous waveform data recorded by 98 seismic sta-tions between 1 and 11 August 2023 were collected.Phase picking was performed using neural network models,followed by an evaluation of picking accuracy,which confirmed that the res-ults met the requirements for subsequent earthquake location.Using both preliminary and re-fined relocation procedures,an initial earthquake catalog containing 276 events was obtained. Phase identification-based detection methods can identify the majority of earthquake events.However,to further improve catalog completeness,previous studies have commonly employed template matching techniques to detect small-magnitude events with low signal-to-noise ratios or those obscured by coda-wave interference.Accordingly,the initial catalog was merged with the unified catalog of the China Earthquake Networks Center,yielding a com-bined dataset of 290 events.Using this unified catalog as templates,template matching was ap-plied to further expand the event set,ultimately increasing the number of identified earthquakes to 396,all of which were classified as aftershocks of the Pingyuan earthquake.The number of detected aftershocks is approximately 3.2 times that of the manually compiled catalog for the same period.Meanwhile,the magnitude of completeness was reduced from ML1.8 to ML1.5,resulting in improved conformity with the Gutenberg-Richter relation and a more reliable estim-ate of the b value. The source region of the Pingyuan earthquake is located within the Linqing Depression,where the sedimentary cover is relatively thick and seismic velocities are significantly lower than those in the surrounding uplifted regions.As a result,the shallow velocity structure is highly heterogeneous,making the selection of an appropriate velocity model critical.Based on travel-time inversion,a one-dimensional velocity model was constructed for the source region to better characterize the shallow low-velocity structure.The mainshock was located at a depth of 10.8 km,which is consistent with the typical focal depth range of earthquakes in the North China region. The relocated aftershocks are predominantly distributed along a north-northeast(NNE)trend,with focal depths mainly concentrated between 8 and 12 km.The mainshock did not oc-cur on any previously mapped faults.Most nearby faults strike southwest-northeast,whereas only a few exhibit east-west orientations.Based on the spatial distribution of the aftershocks and their consistency with regional fault orientations,we infer that the Pingyuan earthquake likely ruptured a blind fault segment located between the Lingxian Fault and the Guanxian fault,intersecting the Linnan Fault.In addition,a small-scale subparallel fault in the vicinity of the mainshock was also activated,further highlighting the structural complexity and multi-fault interactions in this region.
Deep learning methods,particularly Convolutional Neural Networks(CNNs),have demonstrated significant advantages in seismic event classification so far.However,the effects of data representation method and signal-to-noise ratio(SNR)on model performance remain unclear.This study evaluates four CNN architectures(ResNet18,VGG16,DenseNet121,and Inception V3)for the classification of natural mine earthquakes and blast events,using both waveform and spectrogram inputs under varying SNR conditions.Our research utilizes seismic data collected from a coal mine in Honggu,Lanzhou,Gansu Province,China,where frequent mining-induced seismicity poses significant safety challenges.The microseismic monitoring system,equipped with 24 short-period vertical-component seismic detectors sampling at 500 Hz,provides a comprehensive dataset of both natural mining earthquakes and controlled blasting events.We extracted 1 554 blast waveforms and 990 natural mine earthquake wave-forms for classification,with each sample containing 1 000 data points(2 seconds)starting from the earthquake origin time.We implemented three distinct data processing approaches:direct waveform input,waveform image conversion,and spectrogram generation.The direct waveform approach used preprocessed time series data as CNN input with appropriate padding to ensure consistent input size.The waveform image method involved plotting each prepro-cessed waveform as a 256×256 pixel image with black lines on a white background,removing all axes and ticks to eliminate extraneous information.For the spectrogram approach,we em-ployed short-time Fourier transform(STFT)with a 256-sample Hanning window and 50%over-lap,converting the resultant frequency domain data into grayscale images.Our experimental results reveal that data representation significantly impacts classification performance.For CNN-based architectural models,image-based inputs(both waveforms and spectrograms)sub-stantially outperformed direct waveform inputs,with classification accuracies exceeding 90%compared with less than 70%for direct waveforms.The superior performance of image-based inputs can be attributed to the intrinsic design of CNNs,which were originally developed for computer vision tasks and are particularly adept at extracting spatial features from images through local receptive fields and weight sharing characteristics.Among image-based represent-ations,spectrograms generally yielded better classification performance than waveform images for ResNet18,DenseNet121,and Inception V3.This superiority may stem from several factors:spectrograms simultaneously present frequency distribution and energy characteristics of signals,enabling better discrimination of key patterns associated with event types in a struc-tured time-frequency space;they achieve dimensionality reduction and feature concentration,providing more compact representation of key features in two-dimensional space;and different types of seismic events often exhibit distinctive frequency patterns and energy distribution char-acteristics that are more prominently displayed in spectrograms.The impact of SNR on classi-fication performance proved to be substantial.After dividing our dataset into high and low SNR groups,we observed that all models exhibited excellent performance on high SNR data,with waveform images surprisingly outperforming spectrograms—a reversal of the trend observed in mixed SNR conditions.Inception V3 achieved the best performance with waveform image in-puts under high SNR conditions(F1 score:0.974 9±0.014 5),suggesting that waveform im-ages better preserve critical temporal features when signal quality is high.Conversely,for low SNR data,all models showed marked performance degradation,but spectrogram inputs main-tained a performance advantage,with Inception V3 achieving the best results(F1 score:0.844 5±0.028 2).This indicates that spectral domain representation offers greater robustness in noisy environments,possibly because spectral transformation can partially separate signals from noise,making feature extraction more reliable.To elucidate the decision-making mechanisms of different CNN models,we employed gradient-weighted class activation mapping(Grad CAM)visualization.This technique precisely locates the regions most influential in classifica-tion decisions by calculating gradients of target class scores with respect to feature maps.Our analysis revealed that high-performing models typically focused on seismic phase arrival regions and their temporal context.ResNet18 demonstrated large attention areas directly targeting seis-mic waveform regions,particularly P-wave arrival zones,where natural mine earthquakes and blast events exhibit significant waveform characteristic differences.This effective weight distri-bution contributed to ResNet18's superior performance across both waveform and spectrogram representations.In contrast,VGG16 showed smaller focus areas concentrated in blank regions of the images,potentially explaining its relatively poorer classification performance. This study demonstrates that data representation choice significantly affects model perform-ance in seismic event classification.For CNN-based architectures,image-based inputs substan-tially outperform direct waveform inputs.Under mixed SNR conditions,spectrograms gener-ally outperform waveforms;however,this advantage reverses under high SNR conditions,where waveform representations better preserve critical temporal features.Model architecture plays a crucial role in classification robustness,with ResNet18 exhibiting excellent perform-ance across both data representations,likely due to its residual connection structure facilitating effective feature extraction.Finally,model visualization through Grad CAM provides intuitive understanding of decision processes,revealing that superior models typically focus on seismic phase arrival regions and surrounding temporal context—a finding consistent with traditional earthquake classification method and validating the models'learning strategies.
To obtain a detailed understanding of the rupture process of the MW7.6 earthquake that occurred in Japan on January 1,2024,this study integrates seismic data from multiple agencies and researchers.By analyzing the focal mechanism solutions provided by these institu-tions,we derived two nodal planes of the centroid moment tensor(CMT)solution for this event.Through an in-depth analysis of the depth distribution of aftershocks within 24 hours fol-lowing the mainshock,we concluded that the southeast-dipping nodal plane in the CMT solu-tion is more appropriate to be identified as the causative fault plane for this earthquake.This determination is based on the spatial alignment of aftershock depths and their correlation with the inferred fault geometry. To further refine the rupture characteristics,we employed broadband P-wave waveform data from 35 seismic stations within the Global Seismographic Network(GSN),located at dis-tances between 30° and 90° from the epicenter.Using the waveform inversion method for earth-quake rupture processes,we iteratively tested and optimized the parameters to reconstruct the rupture behavior of this earthquake.This methodology allowed us to achieve a reliable solution that delineates the spatiotemporal evolution of the rupture. The results of the inversion reveal several key characteristics of the earthquake's rupture process: 1)Rupture duration and sub-events:The total rupture duration of this earthquake was approximately 40 seconds.The source time function shows three distinct sub-events,each cor-responding to different stages of the rupture.The largest sub-event occurred around 30 seconds after the initial rupture,releasing a significant portion of the earthquake's total seismic moment.This pattern suggests a complex rupture process involving multiple fault segments or asperities. 2)Rupture zone and slip distribution:The primary area of concentrated slip is located southwest of the epicenter.This region experienced the most significant fault slip during the rupture process.The maximum slip on the fault plane reached 4.28 meters,while the maxim-um slip velocity was measured at 1.01 meters per second.The spatial distribution of slip indi-cates that the rupture propagated predominantly in a unilateral direction,with a notable concen-tration of energy release in the southwest quadrant of the fault plane. 3)Seismic moment and faulting mechanism:The moment magnitude of this earthquake was determined to be approximately MW7.54,consistent with previous estimates from global seismic networks.The rupture mechanism is characterized as a thrust faulting event,typical of subduction zone earthquakes.The fault geometry and slip distribution suggest that this earth-quake was associated with a reverse fault,dipping to the southeast,in alignment with regional tectonic stress orientations. This earthquake is noteworthy for its complex rupture dynamics and its ability to generate large-scale fault slip over a relatively short time span.The rupture process involved multiple stages,with significant energy release concentrated in a few key sub-events.The identification of the southeast-dipping fault plane as the causative structure for this earthquake provides valu-able insight into the tectonic behavior of the region.Additionally,the analysis of aftershock depth distributions supports the conclusion that this plane is the primary fault responsible for the mainshock. The implications of this study extend to the broader understanding of seismic hazard in Japan,particularly in regions prone to thrust faulting events.The detailed characterization of the rupture process and fault slip behavior contributes to the growing body of knowledge on earthquake mechanics in subduction zones.Furthermore,the findings underscore the import-ance of integrating diverse data sources,such as focal mechanism solutions,aftershock distri-butions,and seismic waveform inversions,to achieve a comprehensive understanding of earth-quake rupture processes. The results of this study may also have broader applications in seismic hazard assessments and earthquake forecasting efforts,as the detailed reconstruction of rupture processes can help inform models of future seismic activity in the region.By identifying key areas of concentrated slip and energy release,future studies can focus on the potential for aftershock activity or fur-ther seismic events along the same fault structure. In summary,the 2024 MW7.6 earthquake in Japan exhibits a complex rupture process with multiple sub-events and a concentrated area of slip southwest of the epicenter.The southeast-dipping fault plane is identified as the most likely causative fault,based on aftershock distribu-tion and focal mechanism analysis.This earthquake represents a significant seismic event in the region,and the detailed study of its rupture process provides important insights into the nature of earthquake mechanics in subduction zones.
The Izu-Bonin subduction zone,located in the northwest Pacific Ocean and south-east of Japan,is a typical subduction system formed by the westward subduction of the Pacific slab beneath the Philippine Sea Plate,and the strike direction is nearly N-S.This region is char-acterized by frequent seismic and volcanic activities,complex tectonic processes,and has long served as a key area for studying geodynamics and plate interactions.Seismicity within the sub-duction zone provides valuable insights into the movement and deformation of Earth's interior materials.The accompanying shear-waves record anisotropic characteristics along the propaga-tion paths,offering opportunities to understand plate dynamics,mantle flow mechanisms,and the interaction between subducting slabs and surrounding mantle structures.Compared to previ-ous studies using land-based stations,this study aims to investigate the anisotropic characterist-ics within the subducting Pacific Plate and the sub-slab mantle using ocean bottom seismo-meters. Based on the seismic data from 17 Japan's S-net seafloor seismometers deployed near the trench at the northern end of the Izu-Bonin subduction zone,eight shallow and deep seismic events are chosen from the central and southern parts of the subducted plate.Such station-event geometry excludes the influence of the overriding plate and mantle wedge in the subduction zone,allowing for an analysis of anisotropy within the subducting Pacific slab and the sub-slab mantle.Then two shear-wave splitting analysis methods,the minimum eigenvalue minimiza-tion method and waveform rotation cross-correlation method,are employed to investigate the anisotropic characteristics of the Pacific subducting slab and the sub-slab mantle in the Izu-Bonin subduction zone.By conducting shear-wave splitting analysis on the direct S phases and the seismic upward-propagating s phases,a total of 36 reliable shear-wave splitting measure-ments were obtained and the results were categorized into three types based on the ray paths reflecting anisotropic characteristics across different regions. TypeⅠ e vents,with focal depths of 61-67 km,primarily samples the shallow part of the subducting slab.The fast axis orientations are NNE-SSW,with an average orientation of 67.8°,intersecting the trench at a small angle.These results reflect the anisotropy associated with trench-parallel faults formed during subduction,later modified by local stress field,caus-ing the fast-axis to deviate from a trench-parallel orientation.Variations in the local stress field may enhance or reduce the regional anisotropy,resulting in significant variations in splitting time(0.14-0.82 s)along different seismic ray paths. TypesⅡ andⅢ events are deep earthquakes(399-464 km)with distinct ray path character-istics.TypeⅡ ray paths mainly propagate within the slab,with limited propagation in the sub-slab mantle,while type Ⅲ paths sample both the subducting slab and the sub-slab mantle,incorporating more sub-slab mantle anisotropy.The fast-axis orientations of typeⅡ events are NNW-SSE,with an average orientation of-80.7°,and splitting times ranging from 0.08-0.6 s.Similarly,typeⅢ events have fast-axis orientations of NNW-SSE,with an average ori-entation of-74°,and splitting times ranging from 0.12-0.86 s. Integrated analysis of typeⅡ and typeⅢ results reveal that the fast axis orientations in the deep Pacific slab aligns with the seafloor spreading direction,indicating the"fossil"aniso-tropy formed during plate solidification.In contrast,the anisotropy in the sub-slab mantle is primarily driven by the dynamics of the subducting slab.As the slab dip angle steepens from north to south,the slab undergoes back-rotation,inducing significant trench-parallel mantle flow in the sub-slab mantle.This flow results in lattice-preferred orientation(LPO)aligning along the trench direction,producing pronounced trench-parallel anisotropy.This dynamic interaction highlights the critical role of slab geometry and motion in shaping subduction zone anisotropic patterns. A comparison of splitting times indicates that the average splitting time for typeⅢ events is 0.34 s,significantly greater than the 0.226 s average for type Ⅱ events.This suggests that anisotropy intensity in the sub-slab mantle is markedly higher than the"fossil"anisotropy with-in the subducting slab.The sub-slab mantle anisotropy is driven by the dynamic processes of the subducting slab,with its intensity reflecting the impact of these processes on mantle flow.It is noteworthy that mantle anisotropy intensity generally varies with depth:shallow regions exhib-it enhanced anisotropy due to dislocation creep under high stress,while deeper regions show reduced anisotropy due to diffusion creep.Furthermore,the splitting times range from 0.08-0.86 s,indicating the heterogeneous nature of observed anisotropy,which may reflect the in-tensity variations of mantle anisotropy along the trench direction,influenced by slab geometry and associated mantle flow pathways. These results provide new insights into the spatial distribution and genesis of anisotropy within the subducting Pacific slab and sub-slab mantle in the Izu-Bonin subduction zone.They highlight the combined effects of plate formation history,local stress field variations,and sub-duction dynamics on anisotropy.Future research will integrate more seismic network data,par-ticularly data with longer time spans,to further validate and refine the anisotropic characterist-ics of the study area.Additionally,by incorporating more geological,geochemical,and tomo-graphic imaging results,a more comprehensive model can be built for the Izu-Bonin subduc-tion zone to show its flow field and deformation history.
Soil liquefaction is a major cause of foundation settlement and structural damage dur-ing earthquakes.Currently,there have been numerous studies on the use of gravel piles and overlying gravel layers separately for site liquefaction treatment,but relatively little research has been conducted on the mechanism and effect of combining the two to improve the liquefac-tion resistance of foundation soil layers.This study combines physical model experiments with FLAC3D numerical simulations to investigate the anti-liquefaction effects of using gravel piles in combination with overlying gravel layers in sandy soils. For the physical model testing part,different combinations of physical models involving overlying gravel layers and gravel piles were constructed in a model box.Using a small shaking table,the liquefaction process of the soil layer under seismic action was simulated,and the ex-cess pore water pressure under different working conditions was measured. In the numerical simulation part,based on the corresponding physical experimental mod-els,the responses of gravel piles,overlying gravel layers of different thicknesses,and the com-bination of gravel piles with overlying gravel layers of different thicknesses under earthquake action were simulated.Emphasis was placed on analyzing their impact on the excess pore water pressure ratio of the soil layer and the changes in the final surface settlement. The results show that the overlying gravel layer can effectively increase the effective stress of the liquefied sand,significantly reduce the peak value of the excess pore water pressure ra-tio,and thus is unfavorable to sand liquefaction.As the thickness of the overlying gravel layer increases,the anti-sand liquefaction effect in shallow areas becomes more pronounced. Gravel piles provide a rapid drainage path in the liquefied soil layer,reducing the accumu-lation of excess pore water pressure.The closer to the pile,the better the drainage effect.Their influence range shows a linear growth trend with the increase of burial depth.The cloud plot of the excess pore water pressure ratio in the numerical simulation also exhibits a trapezoidal distri-bution characteristic,being narrow at the top and wide at the bottom. The overlying gravel layer significantly reduces the excess pore water pressure ratio in the shallow layer of the foundation,while gravel piles mainly function in the deep layer.The com-bination of the two can improve the overall liquefaction resistance of the foundation. The significant difference in modulus between gravel piles and the surrounding soil can lead to uneven surface settlement of the site,especially in the initial stage of gravel pile applica-tion.Increasing the overlying gravel layer can effectively improve the overall compressive strength of the engineering site and reduce the uneven surface settlement caused by the modulus difference between gravel piles and the surrounding soil. The installation of overlying gravel layers and gravel piles at the construction site can signi-ficantly reduce the pore water pressure of the soil layer and increase the effective stress of the soil layer,thereby reducing the excess pore water pressure ratio and effectively preventing li-quefaction.Especially in the prevention and control of liquefaction in shallow and deep soil lay-ers,the combination of overlying gravel layers and gravel piles has shown a good synergistic ef-fect.Reasonable design of the thickness of the overlying gravel layer and the spacing of gravel piles can help improve the liquefaction resistance of the site and reduce the risk of earthquake disasters. Although gravel piles can effectively enhance liquefaction resistance,the uneven settle-ment they cause needs to be fully considered in the design.By appropriately increasing the thickness of the overlying gravel layer,this impact can be effectively avoided. Overall,this paper systematically studied the anti-liquefaction effect and interaction mech-anism of overlying gravel layers and gravel piles on liquefied soil layers through a combination of physical model experiments and numerical simulations.A reasonable design optimization scheme was proposed,providing valuable references for the liquefaction treatment of actual en-gineering sites.
The characteristics of near-fault ground motions and their impact on engineering structures are critical research topics for both the seismological and engineering communities.Variations in fault rupture propagation lead to the generation of ground motions with distinct dynamic characteristics.Extensive researches have been conducted on the effects of forward directivity and fling-step pulse-like ground motions on building structures,as these near-fault motions exhibit strong impulsive behavior. In the early morning of 21 September 1999,an earthquake(MW7.6)struck central Taiwan near Chi-Chi.Before this event,there were only a few of ground motion records worldwide for earthquakes greater than M7.0 within distance less than 20 km from causative fault.The Chi-Chi earthquake generated staggering 70 three-component recordings within 20 km of the fault.Additionally,a significant number of pulse-like ground motions exhibiting forward directivity and fling-step effects were observed.This earthquake provided an exceptional research oppor-tunity,significantly advancing the understanding of near-fault strong ground motions and mak-ing a profound contribution to the fields of seismology and earthquake engineering. Furthermore,observations from numerous earthquake damage cases reveal that ground motions comprise both translational and rotational components.The torsional effects induced by earthquakes have emerged as a significant factor in reducing the safety of engineering struc-tures.However,due to the limitations of current observational technologies,directly measur-ing the rotational components of ground motion remains a challenge.To address this,many researchers have attempted to derive rotational components using various mathematical models.With ongoing refinements by researchers,the frequency-domain method has achieved high accuracy in calculating rotational components and has been widely adopted in studies focusing on the rotational ground motion. In real earthquake events,the structure will suffer both the translational and rotational components of the ground motion.Especially for the near-fault pulse-like ground motion,there will be a significant influence on structural performance under the combined effects of the above two components.Thus,this paper selects forward directivity pulse-like ground motions,fling-step pulse-like ground motions and non-pulse-like ground motions as initial translational records.The frequency domain method is used to generate corresponding torsional components.First,the results of the shaking table tests on steel frame structures previously conducted by the research group were used as a benchmark to validate the accuracy and rationality of the finite element modeling method.Then,the seismic response of a five-story seismic-designed steel frame structure is analyzed under rare earthquake level by combining different types of pulse-like ground motions and torsion components.The results show that when only consider the in-fluence of pulse-like ground motion,the forward directivity pulse-like ground motion has the greatest effect on the interstory drift with an increase of 44%.When only consider the influence of torsion component,the interstory drift of the corner column increases by 24%,while the interstory drift of middle column almost unchanged.Under the combined effect of the pulse-like ground motions and the torsional component,the corner column is more affected by the torsion-al component,while the middle column is more affected by the velocity pulse-like ground mo-tion.Therefore,the seismic design should consider the plane position of the components and pay attention to the pulse-like ground motion and torsional component.And it is necessary to take the corresponding design measures to improve the overall seismic capacity of the structure.
