Geophysical Journal International

SummaryReconstructing subsurface structures with high resolution is one of the main goals and potentials of full waveform inversion (FWI). However, FWI is a highly nonlinear and ill-posed problem. Conventional physics-based FWI methods, which rely on gradient-based optimization to minimize the difference between observed and synthetic data face cycle-skipping challenge. Although numerous deep-learning inversion approaches have shown promise, they typically focus on latent representations of time-domain seismic data. This often causes an unstable inversion process due to waveform mismatches. To overcome these limitations, we introduce FFT-InversionGAN, an unsupervised seismic inversion framework that integrates physics-based forward modeling with adversarial learning of the frequency-domain data based on Wasserstein generative adversarial network with gradient penalty (WGAN-GP). Fast Fourier transformer (FFT) is employed to transfer the time and phase information of time-domain seismic data into the spectrum and amplitude distributions to modify the feature space and sensitivity of the adversarial loss to different types of mismatches. By leveraging Wasserstein distance constraints, this method can naturally operate on the spectral distributions of seismic data. Compared with L2 norm, Wasserstein distance is far less sensitive to the linear variations in the phase spectrum. And our proposed method eliminates the need for network pre-training while improving stability and flexibility. FFT-InversionGAN demonstrates enhanced accuracy and resilience in numerical experiments on noise-free, noisy and missing low-frequency benchmarks. This was observed when applied to the Marmousi and overthrust models, where it consistently outperformed conventional FWI and FWIGAN. These findings highlight that FFT-InversionGAN has superior inversion effectiveness.

SummaryThe 2020 MS 5.0 Qiaojia earthquake occurred in a tectonically complex region near the Xiaojiang fault in southwestern China. We investigated the rupture process of this moderate-sized earthquake using a multi-empirical Green’s function (EGF) inversion method that integrated waveforms from multiple EGF events. Synthetic tests demonstrated that the multi-EGF inversion method recovered the input model more robustly than any individual EGF inversion. The resolved spatiotemporal rupture model of this earthquake indicated a compact rupture lasting approximately 2.9 s, dominated by a major asperity near the hypocenter and characterized by predominantly eastward rupture propagation. Bootstrap resampling analyses further confirmed the robustness of the resolved major coseismic slip distribution and the overall moment release pattern. We also observed a spatial complementarity between the coseismic slip and aftershock distributions, with most aftershocks clustering around the periphery of the major asperity. This study not only elucidates the source complexity of the 2020 MS 5.0 Qiaojia earthquake, but also validates the robustness and effectiveness of the multi-EGF inversion method in resolving the rupture processes of moderate-sized earthquakes. Our results provide new insights into the rupture kinematics of moderate-sized earthquakes and the heterogeneity of fault strength and stress within the Xiaojiang fault zone and its surrounding regions.

SummaryThe Wasatch Front Community Velocity Model (WFCVM) is the most complete and detailed Earth model for the Wasatch Front region in north-central Utah (USA). Until recently, it had not been well evaluated with strong ground motion observations due to a lack of local earthquakes with magnitude M5+. The 18 March 2020 Mw 5.7 Magna, Utah, earthquake generated excellent strong ground motion data at dozens of stations along the Wasatch Front, with peak ground accelerations up to 0.54 g. Here, we use the forward finite-difference code SW4 to simulate waveforms of the 2020 Magna mainshock in the WFCVM up to 3 Hz and compare its predictions to observations from 35 nearby stations at epicentral distances of 4–46 km. We use a finite fault source model with a semi-stochastic slip distribution and overlay stochastic velocity perturbations (S) and surface topography (T) on the WFCVM, which we refer to as the 3D+S+T model. Observed-predicted amplitude ratios and Goodness-of-Fit (GOF) scores for PGA, PGV, PGD, Arias intensity and duration, cumulative energy and duration are calculated. Our 3D+S+T model performed fairly, matching the general character of the observations with an average GOF score of 5.20 (out of a maximum of 10), slightly better than the unaltered WFCVM score (GOF=4.97). Stochastic velocity perturbations mostly affect peak ground motions at the closest sites (< 20 km), while surface topography improves durations for basin sites and generates more realistic signals at higher frequencies. Neither addition resolves underprediction of basin amplification in the eastern Salt Lake Basin and overprediction of ground motion at basin-edge sites, which likely reflect inaccurate representations of basin structure in the WFCVM. Based on these results, we recommend including stochastic velocity perturbations and topography in future simulations but conclude that updating deterministic models of basin structure will lead to the biggest improvement in forecasting ground motion for future large (M6.75+) earthquakes in the Wasatch Front region.

SummarySpectral Induced Polarization (SIP) is a geophysical technique which measures the frequency dependent electrical properties of geologic materials which can, in turn, be linked to underlying petrophysical parameters. Four-electrode SIP measurements exhibit errors above 100 Hz related to parasitic capacitive coupling (PCC) inside of the instrumentation and to the impedance of the potential electrodes. These errors can easily mask the true sample response. Existing techniques to correct SIP data infected with these errors can be complex and prone to operational error. Here we present a simple procedure that utilizes joint two- and four-electrode measurements using the same sample holder to validate high frequency SIP data. We tested the practicality of this approach by performing a series of two electrode SIP measurements on a known NaCl solution using conventional coiled current electrodes composed of different metals. We compared this procedure with both theoretical values and against a four-electrode correction procedure (referred to as the Wang correction), which utilizes four impedance measurements to directly calculate high frequency phase errors in instruments with differential amplifiers. We found that two electrode measurements conducted with coiled Ag-AgCl electrodes performed well for resistive samples and for highly polarizable samples above 100 Hz, and for conductive samples above 1 kHz. The use of joint two- and four-electrode measurements on the same sample holder is simpler than existing correction techniques and presents a straightforward alternative to the validation of high-frequency four-electrode data.

SummaryMagnetization vector inversion is an effective method for analyzing magnetic anomaly data influenced by significant remanent magnetization. However, the multi-dimensional parameters of the magnetization vector increase both the non-uniqueness of the solutions and the computational burden. We propose a magnetization vector inversion method based on Gaussian radial basis function which the magnetization vector parameters are represented by the functional node parameters. By leveraging the inherent smoothness and local support characteristics of Gaussian radial basis function, the method suppresses spurious divergence in magnetization direction during the inversion process, thereby enhancing both the accuracy and computational efficiency of the inversion results. The proposed method is applied to interpret magnetic data in Xiangshan area for revealing the magnetization characteristics of magma-hydrothermal structures. The region of non-uniform magnetization vectors, which can be interpreted as lithological contacts and alteration fronts, may indicate multiple phases of magmatic intrusion. The distinct magnetization directions between shallow mineralized bodies and underlying magma conduits facilitates the identification of potential mineralized rocks and magma conduits that are undetectable by conventional magnetic intensity analysis. Drilling in the study area confirms the presence of Cu-Ni mineralization in the shallow mafic-ultramafic intrusions. Results demonstrate that the magnetization vector inversion could capture complex geological information, providing a promising tool for understanding volcanic and magmatic systems.

SummaryOn August 27th, 2024, at approximately 19:30 UTC, the Starlink-2382 satellite entered the Earth’s atmosphere following an uncontrolled re-entry manoeuvre over Central Europe. This event resulted in a relatively low-angle re-entry of the satellite into the atmosphere, which might have provided sufficient time to burn up the satellite before reaching the Earth’s surface. This study employs acoustic-seismic (A-S) data from 226 recording stations to analyse the trajectory of Starlink-2382’s re-entry, utilizing 3-D atmosphere models including wind data and acoustic ray tracing methods. To identify signals emitted by the falling satellite, we process A-S recordings of Austrian, French, German, Italian, Slovenian, and Swiss regional seismic networks. We compute the satellite trajectory with a novel ray-based direct-search optimization method and find an azimuth angle of 120.5°±0.4° from North and an initial elevation angle of 1.5° ±0.7°, together with an entry velocity of approximately 8.9 ±0.7 km s−1. Our findings indicate that this acoustic-seismic approach, including travel time effects due to wind, achieves a better fit to our large dataset compared to the trajectory solutions from optical methods in this specific context. Furthermore, we calculate an effective ablation coefficient of 0.11 ±0.02 s2 km−2 for the main satellite fragment. Within the limits of this estimate, this is consistent with a scenario in which the main fragment, with a mass of c. 100 kg could have experienced near-complete ablation during atmospheric descent.Finite-difference modelling illustrates the complex acoustic wavefield resulting from the satellite’s deceleration and shows the expected widening of the Mach Cone. This highlights the importance of accounting for trajectory curvature and time-varying Mach angles when modelling acoustic wave propagation from low-angle re-entering objects. For recording sites with both, acoustic (infrasound) and seismic sensors, the acoustic-to-seismic ground coupling coefficients are determined. These vary up to three orders of magnitude, from 4.31 $\times $ 10−10 m s−1 Pa−1 to 5.86 $\times $ 10−7 m s−1 Pa−1 across our station sites, which is primarily explained by differences in stiffness of surface rocks.

SummaryPrevious work has demonstrated a significant correlation between the pattern of sea level change computed from an altimeter-based inference of Greenland ice mass flux from 1993-2019 and sea surface height (SSH) observations adjacent to the island. However, a key question is unanswered in this detection; namely, what constraints on ice mass flux do the SSH observations provide? To address this issue, we perform a series of inversions of the available SSH data offshore Greenland. Our results indicate that such inversions are highly non-unique. However, we also demonstrate that robust inferences can be obtained by incorporating reasonable a-priori constraints, in our case limiting the ice model to a small set of discs associated with the major drainage basins of the ice sheet that are proximal to the SSH observations. Our inversions in this case yield estimates of average ice mass loss in the range 0.62-0.70 mm/yr in units of equivalent global mean sea level change over the period 1993-2019, when the observations are corrected for the signal of dynamic sea level change. This inference agrees with independent ice altimeter-based estimates of Greenland ice sheet mass flux rates, showing broadly consistent relative ice mass loss rates across southern Greenland basins. Our analysis is the first to directly invert SSH observations for ice mass changes and we conclude that the consideration of such data, particularly in combination with other data sets (e.g., GRACE gravity, ice altimeter measurements, GNSS observations) has the potential to improve constraints on ice sheet mass changes in a warming world.

SummaryAccurate traveltime computation is fundamental to high-accuracy 3-D seismic imaging and inversion. In anisotropic media, finite-difference schemes and conventional iterative fast sweeping methods (FSM) for the eikonal equation often suffer from numerical instability or convergence difficulties when the monotonicity of the slowness surface breaks down. Thus, we propose a traveltime computation method for 3-D tilted transversely isotropic (TTI) media that embeds a semi-analytical solver into the FSM framework. The proposed semi-analytical solver employs a lower triangular–diagonal–lower triangular transpose (LDLT) decomposition together with a resolvent cubic equation to robustly factorize the local quartic traveltime equation. Combined with a Newton-Raphson-based coefficient refinement strategy and a group-velocity-based causality check, the method directly and accurately identifies the physical root corresponding to the quasi-P (qP) wave. Numerical experiments show that the semi-analytical solver has better numerical stability than existing quartic solvers. For weakly anisotropic models, the proposed method achieves an accuracy comparable to that of Newton-based local solvers. Its main advantage lies in improved robustness in strongly anisotropic media or more complicated local quartic behavior, where admissible-root selection becomes more challenging.

SummaryThe collision between the European plate and the Adria microplate during the Cenozoic led to the formation and uplift of key mountain belts, including the Alps, Apennines, and Dinarides. This convergence also resulted in a highly complex assemblage of tectonic units, each characterized by distinct geological and geophysical properties within the accreted crustal domains. A comprehensive understanding of the geodynamic evolution of this region requires integrated imaging of both the crust and upper mantle. To achieve this goal, we apply Teleseismic Full Waveform Inversion (TFWI) to P-wave seismic data recorded by permanent European broadband stations, supplemented by the dense temporary deployments of the AlpArray initiative, SWATH-D, and CIFALPS-2 projects. Leveraging this unprecedented seismological coverage, our study aims to design a suitable TFWI workflow to develop a multiparameter model defined by P- and S-wave velocities and density of the Alpine orogen down to 500-km depth. The critical importance of high-quality data for ensuring the reliability of TFWI results first prompts us to develop a semi-automated workflow for data selection and quality control, from which we select 84 teleseismic events for inversion. The seismograms were filtered within the 5-to-25-s period band, and a 30-s time window from the first arrival was used for inversion. Other critical aspects are the assessment of the resolution power of TFWI provided by the field acquisition geometry, as well as potential sources of artefacts. We review the key theoretical factors controlling resolution and imaging artefacts, and further illustrate these issues with numerical experiments designed with the field acquisition geometry to provide the necessary guidelines for sound geological interpretation of the TFWI models. The reconstructed TFWI models effectively capture key crustal features, including low-velocity sedimentary basins, high-velocity anomalies like the Ivrea Body, deep mountain roots beneath the Alpine and Apennine chains, and the signature of the continental subduction. The TFWI models also reveal small-scale anomalies previously identified by local tomography studies. Then, we extend the analysis at upper-mantle depths by comparing the footprint of the subducting slabs in the P and S velocity TFWI models with previous ones obtained by surface wave tomography and teleseismic body-wave traveltime tomography. These comparative analyses highlight the incomparable power of TFWI to resolve multiparameter models of the Earth’s interior from the surface down to the upper mantle. From this first critical analysis of the TFWI results, a comprehensive geological survey of the reconstructed structures will be presented in a companion paper.

SummaryThe complexity of earthquake rupture dynamics and the diversity of observed seismic behaviors are fundamentally governed by the frictional properties of faults and their response to tectonic stress. Grounded in the rate-and-state constitutive law derived from laboratory experiments on rock friction at slow slip velocities, we employ a fully dynamic model to investigate how frictional conditions give rise to a diverse range of rupture modes and influence their propagation dynamics. Under uniform background stress and nucleation conditions, the rupture type, whether supershear, sub‑Rayleigh, self‑arresting or slow self‑arresting rupture (SSAR), is governed by the relative contributions of the direct effect and the evolution effect, expressed as $R = 1 - \frac{a}{b}$, together with the normalized characteristic slip distance D. Their respective regimes are summarized in a phase diagram. We demonstrate that the friction parameters R and D significantly influence the rupture process, with R primarily enhancing stress release and slip during rupture, while D predominantly controls the rupture speed. For varying values of R, there exists an optimal intermediate D that maximizes rupture velocity. Furthermore, simulations suggest that when frictional parameters approach the boundaries between different rupture types regimes, the earthquake may not be confined to a single mode. Instead, a single rupture event can exhibit complex, continuous, yet rapid transitions between distinct types under a single triggering without interruption. These transitions can occur smoothly among various rupture types, including transitions from SSAR to sub-Rayleigh rupture and subsequently to supershear rupture. This study indicates the key role of frictional properties in governing rupture dynamics, offering new perspectives on the inherent complexity of earthquake processes.

SummaryThe Gravity Recovery and Climate Experiment Follow-On (GRACE-FO) mission continues the legacy of satellite gravimetry in monitoring Earth’s mass redistribution. Equipped with dual-frequency Global Positioning System (GPS) receivers and a K-Band Ranging (KBR) system, it enables precise orbit determination and high-resolution gravity field recovery. While integer ambiguity resolution (IAR) has proven effective for GPS-based orbit determination, its impact on time-variable gravity field recovery remains unclear. Here we develop a dynamic framework that jointly estimates GRACE-FO satellite orbits and monthly gravity fields by integrating GPS and KBR observations, in which single-differenced integer ambiguities are fixed and constrained into the normal equations as pseudo-observations with micrometer-level constraint precision. Using GRACE-FO onboard data from July to December 2019, we compare ambiguity-fixed and ambiguity-float solutions in terms of post-fit residuals, orbit accuracy, and gravity field quality. IAR improves three-dimensional orbit precision to ∼1.2 cm RMS, with along- and cross-track components enhanced by up to 52 per cent and 71 per cent, respectively. Satellite Laser Ranging validation confirms ∼1.2 cm agreement. Gravity field solutions from float ambiguities agree closely with official Science Data System (SDS) RL06.1 models up to degree 96, whereas IAR-based solutions maintain consistency only to about degree 40 and exhibit irregular oscillations beyond this range, particularly near orbital resonance conditions around order 45. At higher degrees, these oscillations are accompanied by intensified north–south striping in equivalent water height maps. Covariance diagnostics reveal increased off-diagonal correlations between spherical harmonic coefficients under IAR, indicating weakened spectral orthogonality and potential leakage of high-degree noise. These results indicate that ambiguity-fixed gravity solutions do not consistently outperform float-based solutions beyond spherical harmonic degree 40 in the near-polar orbiting GRACE-FO constellation.

SummaryMicroseismic source location is essential for seismic monitoring and subsurface resource exploitation. Both traveltime inversion and waveform stacking methods suffer from limited accuracy when processing low signal-to-noise ratio (SNR) data under complex velocity models. Existing deep learning approaches mainly employ purely data-driven strategies without physical constraints, exhibiting limited capability to suppress large and unexpected location errors. We propose a physics-constrained deep learning method for microseismic source location that integrates the physical principles of cross-correlation stacking (CCS) imaging into network training. The method incorporates a joint loss function combining source imaging quality loss and traveltime consistency loss, with a Pareto dynamic weighting strategy to balance different loss components. Synthetic experiments on the Marmousi velocity model demonstrate that the joint-constrained method reduces the mean absolute error (MAE) from 34.09 m to 27.91 m compared to the purely data-driven approach. The maximum error decreases from 280.18 m to 130.38 m, a 53.5% reduction, demonstrating effective suppression of large location errors. The trained network achieves single-event imaging prediction in 0.04 s, providing a 75-fold speedup over the 3 s required by conventional CCS. The proposed method shows great potential in near-real-time microseismic monitoring with dense arrays.

SummaryWe present a new method, ODFTEX, for calculating evolving crystal preferred orientation (CPO) in deforming aggregates of olivine plus orthopyroxene undergoing dynamic recrystallization. The model is based on a continuum description of texture in terms of the orientation distribution function (ODF), which satisfies an evolution equation that we solve numerically. The model thus delivers the ODF directly, rather than a collection of grain orientations like most alternative models. Recrystallization is represented by a source term in the evolution equation, defined in such a way that crystals poorly oriented for slip recrystallize most rapidly. The model has only a single free parameter, the recrystallization rate, which we calibrate against a laboratory experiment on an olivine aggregate deformed in simple shear. We illustrate the predictive power of ODFTEX by using it to calculate evolving CPO along pathlines in a two-dimensional convective flow and a three-dimensional subduction zone flow. ODFTEX is computationally about 6-7 times faster than the D-Rex model of Kaminski et al.(2004).

SummaryWe present GEMMIE, a new open-source magnetotelluric (MT) three-dimensional inversion solver designed to handle large datasets with respect to survey area, period range, and number of observations. The main methodological innovations introduced in GEMMIE are: 1) a novel model parametrization strategy – the so-called non-local parametrization – which helps suppress artefacts near observation sites and significantly accelerates convergence to the inverse solution (by up to several times); 2) utilization of the recently developed version of the quasi-Newton iterative optimization method that exploits the structure of the regularized inverse problem and effectively eliminates the need for the additional iterations during line search; 3) the modelled fields interpolation technique that enables proper inverting data across a large number of periods preserving the integrity of the observed responses and their associated error estimates. The forward problem engine is inherited from the authors’ prior work and is based on a modern implementation of the volume integral equation approach, demonstrating near-linear scalability across thousands of computational cores. The workability of the presented solver and the efficacy of the proposed techniques are confirmed by validation on a synthetic dataset and benchmarked against results from inverting real data, obtained using a fundamentally different inverse solver based on the finite element method.

SummaryHigh-rate GNSS is a valuable seismogeodetic tool for real-time monitoring of seismic displacements. The BeiDou PPP-B2b service provides precise point positioning (PPP) augmentation corrections via the B2b signal for BDS-3 and GPS satellites, while the Galileo High Accuracy Service (HAS) delivers such corrections for GPS and Galileo satellites via the E6B signal. However, PPP-B2b and HAS services are designed for different constellations, which makes the simultaneous use of BDS-3, Galileo, and GPS impossible through a single service. To fully exploit the capabilities of multi-GNSS, we propose a method to integrate the PPP-B2b and HAS products with the emphasis on the positioning performance gain of the integrated products under high-rate and short-time seismogeodetic environment. In the simulated seismic wave experiments, PPP-B2b and HAS achieve positioning accuracies of 4.2 cm and 5.2 cm, respectively. For the 2024 Mw 7.0 Wushi earthquake, PPP-B2b attains accuracies of 0.50 cm, 0.94 cm, and 1.10 cm, while HAS achieves comparable accuracies of 0.83 cm, 0.67 cm, and 1.55 cm in the east, north, and up components, respectively. Compared with standalone product, the integrated solution improves positioning accuracy by an average of 21% in the single-axis shake table experiment and by 42%, 29%, and 45% during the Wushi earthquake, achieving average accuracy of 0.37 cm, 0.58 cm, and 0.69 cm in the east, north, and up components, respectively. Magnitudes for the Wushi earthquake derived from the peak ground displacement of GNSS seismic waveforms using the PPP-B2b, HAS, and integrated products show good agreement with the WUM product. These results confirm the feasibility of PPP-B2b, HAS, and particularly their combined use for high-precision positioning, highlighting their great potential for real-time seismogeodetic applications in regions with limited communication infrastructure.

SummaryHigh-resolution seismic tomography performed on rock samples at the laboratory scale is a key ingredient for subsurface rock characterization from seismic imaging. We investigate the performance of first-arrival travel-time tomography on data obtained from a 2D acquisition on a slice of a selected carbonate core using a well-controlled experimental prototype, which involves a point-like pulsed-laser (PL) or a piezoelectric transducer (PZT) as seismic source and a single-point Laser Doppler Vibrometer (LDV) as a receiver which can be shifted during a single experiment. Wave propagation simulations are run on a realistic synthetic 2D slice. Tomography trials on synthetic records establish an optimal inversion strategy, from handling first-arrival travel-time picking to building velocity models by first-break times tomography. The velocity image obtained from the PL-LDV dataset displays similar patterns compared to the X-ray CT-scan image, although the latter is a tomographic image of attenuation. In contrast, the velocity reconstructed from the PZT-LDV dataset shows substantial differences. We therefore recommend the PL-LDV protocol as a reference tool for experimental characterization of core samples based on seismic wave propagation. Adding quantitative core velocity reconstruction to crustal seismic imaging and well-log information will potentially improve the quantitative characterization of the complex subsurface composition. The possible extension to a 3D configuration should be even more fruitful when considering later phases for multi-physics interpretation.

SummaryShear-wave velocity imaging is key for near-surface geotechnical characterization of foundation soils, for subsurface anomaly detection, and for process monitoring in urban environments. In this study, we propose a workflow for characterizing metropolitan areas using ambient and anthropogenic noise records from distributed acoustic sensing (DAS) data collected on telecommunications fiber. We developed a finite-element model that computes shear-wave dispersion curves for fundamental and higher-order surface-wave modes from a layered velocity profile. The model parameters allow intralayer linear and power-law variations of velocity with depth to provide realistic near-surface soil behavior. The forward model is coupled with an inversion algorithm that minimizes the model’s misfit to dispersion curves derived from the DAS data. We demonstrate the workflow using DAS data collected on the University of Wisconsin-Madison’s campus network. The results are compared with the multichannel analysis of surface waves (MASW) method using a line of geophones co-located with one of the DAS sites and a nearby borehole log. Both solutions indicated a sharp change in shear-wave velocity at approximately 20 meters, corresponding to the depth of a sandstone formation at the site, and that a physically-based, power-law function better fits the velocity profiles of unconsolidated sediments. These results demonstrate that our method provides reliable near-surface imaging while suggesting likely soil types. The proposed measurements and modeling techniques can be extended to image the near surface wherever telecommunications fiber is available.

SummaryAttenuation and scattering play a key role in the propagation of seismic waves through complex media, yet their specific contribution to site–city interactions remains poorly understood. Numerical studies have recently introduced the concepts of urban attenuation and urban mean free path to describe these processes. However, observational evidence based on real data is still lacking. In this study, we experimentally investigate these processes using a forest as a natural analogue of an urban environment, where trees, organized as a resonant metamaterial (the METAFORET experiment), act as distributed resonant scatterers analogous to buildings. Using a dense nodal array that combines ambient noise and active-source measurements, we evaluate several key indicators including Arias duration, H/V spectral ratios, coherence, and spatial variability—alongside parameters describing wave attenuation and scattering. The observations reveal a redistribution of seismic energy, characterized by reduced direct-wave amplitudes and prolonged coda-wave durations. These results provide the first experimental evidence of urban-like attenuation and scattering derived from real seismic data, and they pave the way for a passive, noise-based approach to characterize seismic attenuation effects in urban environments.

SummaryThe updated earthquake catalog for Iraq covers the period from 1900 to the end of 2021 and includes over 37 000 recorded earthquakes. To create this comprehensive catalog, five key steps were taken: compiling bulletins, calculating moment magnitudes, harmonizing magnitudes, establishing empirical conversion relations, and evaluating the completeness of the catalog. A notable enhancement in this update is the direct calculation of moment magnitudes for approximately 2 800 earthquakes, achieved through the coda envelope technique and waveform data from the Mesopotamian Seismological Network (MPSN) in Iraq. This updated catalog serves as a valuable resource for examining the spatiotemporal distribution of earthquakes, with respect to earthquake density, maximum moment magnitude, and seismogenic depths. Additionally, the Gutenberg-Richter relationship was applied to calculate the a- and b-values specific to Iraq. The findings show that the Zagros Fold-Thrust Belt has a seismogenic layer (source) that ranges from 2 to 33 km deep and experiences high seismic activity. In contrast, the Mesopotamian Foredeep has a seismogenic layer ranging from 1 to 25 km deep and has lower seismic activity. The greatest seismic activity is concentrated around the Mandili-Badra-Teeb fault, which has experienced significant ruptures over time. The Outer Arabian Platform is identified as the main area of seismic activity, while additional activity occurs on the Inner Arabian Platform. Three major tectonic boundaries define the distribution of earthquakes in the northeastern Arabian Plate. These boundaries are defined by the Main Zagros Reverse Fault, the Zagros Foredeep Fault, and the Anah Graben and Abu Jir-Euphrates Fault Zone. These boundaries highlight variations in seismicity levels and the spatial distribution of deformation in the region. The updated earthquake catalog presented in this study is expected to play a vital role in regional seismicity assessments and seismic hazard analyses for Iraq and its surrounding areas.

SummaryThree-dimensional (3D) gravity inversion for subsurface density reconstruction is a highly ill-posed and non-unique problem due to the intrinsic ambiguity of gravity data and the high dimensionality of the model space. Conventional inversion methods typically rely on iterative optimization with explicit regularization, which can be computationally expensive and sensitive to parameter selection, often leading to smoothed models and limited resolution. In this study, we propose a physics-guided multi-scale encoder-decoder convolutional neural network (CNN) for efficient and physically consistent 3D density gravity inversion. The network establishes a direct nonlinear mapping from two-dimensional gravity anomaly or gravity gradient data to 3D subsurface density models by integrating hierarchical multi-scale feature extraction, dense skip connections, and deep supervision. To enforce geophysical consistency, a gravity forward operator is embedded into the loss function, constraining the inversion results to honor the governing physical laws. Numerical experiments using progressively complex synthetic models-including single-prism, double-prism, inclined staircase structures and irregular complex model-demonstrate that the proposed method accurately recovers density magnitudes, spatial geometry, sharp boundaries, depth discontinuities, and opposing density polarities, while producing gravity responses that closely match the observations. Application to airborne gravity gradient data over the Vinton salt dome further validates the method under realistic conditions, yielding geologically plausible density models compared with previously published conventional regularized and Bayesian inversion approaches. These results indicate that the proposed physics-guided CNN provides a robust, accurate, and computationally efficient alternative for large-scale and complex 3D gravity inversion problems in real-world geophysical applications