Geophysical Journal International

SummaryModerate magnitude earthquakes and seismic sequences frequently develop on fault systems, but whether they are linked to future major ruptures is always ambiguous. In this study, we investigated a seismic sequence that has developed within a portion of the stretching region of the Apennines in Italy where moderate to large earthquakes are likely to occur. We captured a total of 2 039 aftershocks of the September 18, 2023, Mw4.9 earthquake occurred during the first week, by using ML-based algorithms. Aftershocks align on two 5-7 km long parallel faults, from a length that exceeds what is expected from the mainshock magnitude. The segments are ramping at about 6 km depth on closely spaced N100 striking 70 N dipping planes, at a distance of some kilometers from the mainshock hypocenter. Our results indicate that even moderate magnitude events trigger seismicity on a spread set of fault segments around the mainshock hypocenter, revealing processes of interaction within the crustal layer. The possibility that larger earthquakes develop during seismicity spread is favored by pore pressure diffusion, in relation with the closeness to criticality of fault segments. Based on the very rapid activation of seismicity on the entire system and a back-front signal from the hypocenter of the main event, we infer that fluid pressure, initially high within the crustal layer, rapidly dropped after the mainshock. Our study reinforces the importance of timely extracting information on fault geometry and seismicity distribution on faults. ML-based methods represent a viable tool for semi-real-time application, yielding constraints on short-time forecasts.

SummaryLarge-scale ocean-bottom node (OBN) arrays of 1000s of multi-component instruments deployed over 1000s of square kilometers have been used successfully for active-source seismic exploration activities including full waveform inversion (FWI) at exploration frequencies above about 2.0 Hz. The analysis of concurrently recorded lower-frequency ambient wavefield data, though, is only just beginning. A key long-term objective of such ambient wavefield analyses is to exploit the sensitivity of sub-2.0 Hz energy to build long-wavelength initial elastic models, thus facilitating FWI applications. However, doing so requires a more detailed understanding of ambient wavefield information recorded on the seafloor, the types, frequency structure and effective source distribution of recorded surface-wave modes, the near-seafloor elastic model structure, and the sensitivity of recorded wave modes to subsurface model structure. To this end, we present a wavefield analysis of low- and ultra-low-frequency ambient data (defined as <1.0 Hz and <0.1 Hz, respectively) acquired on 2712 OBN stations in the Amendment Phase 1 survey covering 2750 km2 of the Gulf of Mexico. After applying ambient data conditioning prior to cross-correlation and seismic cross-coherence interferometry workflows, we demonstrate that the resulting virtual shot gather (VSG) volumes contain evidence for surface-wave and guided P-wave mode propagation between the 0.01-1.0 Hz that remains coherent to distances of at least 80 km. Evidence for surface-wave scattering from near-surface salt-body structure between 0.35-0.85 Hz is also present in a wide spatial distribution of VSG data. Finally, the interferometric VSG volumes clearly show waveform repetition at 20 s intervals in sub-0.3 Hz surface-wave arrivals, a periodicity consistent with the mean active-source shot interval. This suggests that the dominant contribution of surface-wave energy acquired in this VSG frequency band is likely predominantly related to air-gun excitation rather than by naturally occurring energy sources. Overall, these observations may have important consequences for the early stages of initial model building for elastic FWI analysis.

SummaryEarth’s plate tectonic behavior arises from lithospheric ductile weakening and shear-localization. The ubiquity of mylonites at lithospheric shear zones is evidence that localization is caused by mineral grain-size reduction. Most lithospheric mylonites are polymineralic, suggesting that the interaction between mineral phases by Zener pinning promotes grain-size reduction and weakening. Yet this interaction only occurs where mineral phases mix at the grain scale. Phase mixing and its effect on microstructure and strength have been shown in deformation experiments and natural field samples. Our theory for the interaction between phase mixing (treated as a stress driven diffusion) with two-phase grain damage has been compared to lab experiments. But using processes at the tiny grain-scale embedded within the small hand-sample and lab scales to model large-scale lithospheric processes, requires an upscaling scheme that captures the physics from micro- to macro-structures. For example, weakening from grain-damage in zones of mixing can lead to banded viscosity structure at the small scale that manifests as viscous anisotropy at the large scale. Here we provide a new framework for self-consistently upscaling from microscopic (grain) scales, to mesoscopic (petrological heterogeneity) scales to macroscopic (tectonic) scales. The first upscaling step models phase mixing and grain-size evolution in a small “mesoscopic” lab-scale volume or “patch”, which is equivalent to a point in the macroscopic space. Within this mesoscale patch, stress driven diffusive mixing is described by an analytical solution for mineral phase fraction, provided a minimalist Fourier representation of phase fraction, and a transformation to the patch frame of reference as well as to the principal stress directions at that point. The orientation and volume fraction of mixed-phase regions can then be extracted from the analytical solution for phase fraction. The grain-size and viscosity in the mixed bands are determined by two-phase grain-damage theory; the unmixed zone properties follow from mono-phase grain damage theory. The mesoscale banded viscosity field leads to a macroscale anisotropic viscosity at that point in space. But, the evolution of properties at each macroscale point involves tracking only a few quantities (phase fraction, grain sizes) rather than modeling each patch of mesoscale space as its own 2-D or 3-D system. For the final upscaling, the anisotropic viscosity field is used in a macroscale lithosphere flow model. We show an example of this scheme for a lithospheric Rayleigh-Taylor drip driven by ridge-push compressive stress, which can cause anisotropic weakening via grain mixing and damage that may help initiate subduction and passive margin collapse.

SummaryThe crust of the South American platform recorded imprints of dynamic processes related with the opening of the Central and South Atlantic but has not been well measured. Crustal structure can be retrieved from teleseismic receiver functions using H-κ stacking, but nearly-parallel stripes of high stacking values existing in stacking images for seismic stations in sedimentary area cause difficulties in identifying solutions. We show that some seemingly spurious stripes that do not point to any layer solution are helpful in the identification of the solution position. With the aid of the auxiliary stripes, we retrieved thicknesses and Vp/Vs of sedimentary and crystalline crust for 65 permanent stations of the Brazilian Seismographic Network and 6 new portable seismic stations in Brazil and Uruguay. The resulted sedimentary thickness and Vp/Vs exhibit a good correlation with the Phanerozoic sediments in the South American basins. The crust of Paraná–Etendeka Large Igneous Province (LIP) had been expected to be more mafic since it had ever been penetrated by mantle magma in the Cretaceous related to the south Atlantic opening. However, we found very low Vp/Vs (1.67) in the crystalline crust beneath the LIP, implying a more felsic crust and that no significant mafic intruding/underplating has occurred in the region. The more felsic crust may be formed in a special evolution early than the magmatic event, or during the magmatic event by releasing crustal volatiles. The resulted sedimentary thickness and Vp/Vs ratios exhibit a good correlation with the Phanerozoic sediments in the South American basins, which implies that Triassic–Jurassic and Cretaceous magmatism did not cause significant metamorphism in sediments formed before the magmatic events.

SummaryThe Sumatran Fault Zone (SFZ) of the Indonesian island of Sumatra, which is broken up into 19 fault segments, accommodates much of the trench-parallel component of the oblique convergence between the Indo-Australian and Sunda plates. To understand the potential hazard of SFZ earthquakes to the local population, we investigate slip rates and locking depths of three SFZ segments in southern Sumatra using previously unpublished data from our Sumatran Fault Monitoring (SuMo) campaign Global Positioning System (GPS) network. We model the GPS data using a two-dimensional interseismic dislocation model optimized using a Bayesian approach. For the Musi segment of the SFZ, we find that slip rates ranging from 10 to 22 mm/year and locking depths from 1 to 20 km fit the data similarly well, suggesting a lack of resolution for the SuMo network in this segment. For the Manna and Kumering segments where the resolution is better, the estimated slip rates are 18 [12–22, 95 per cent confidence intervals] mm/year and 12 [9–15] mm/year, respectively, while the estimated locking depths are 29 [15–47] km and 5 [3–16] km, respectively. The deep locking depth estimated for the Manna segment can be explained by the large station gap in this segment. Considering the uncertainty, all the estimated slip rates from our study remains aligned with the SFZ's average slip rate of ∼15 mm/year, which was previously derived using updated geological slip rates and geodetic block modelling of the entire SFZ. Our results support the idea that the forearc sliver west of the SFZ behaves as a rigid microplate.

SummaryThe conventional transient electromagnetic inversion method has a low calculation speed and precision and is susceptible to falling into local minima, which does not meet the fine detection requirements of urban underground space. In this study, we proposed a novel inversion method based on convolutional bidirectional long short-term memory neural networks for shallow subsurface transient electromagnetic inversion. This network structure possessed strong spatial feature extraction capabilities and a proficient understanding of sequential data, thereby addressing the issues of slow conventional inversion computations and inadequate inversion accuracy. Utilizing the apparent resistivity from a three-layer model as the sample input and the real model as the target, the network was trained using batch normalization and dropout techniques to accelerate the convergence rate. The resulting model achieved real-time inversion speeds and high accuracy, with robust generalization capabilities and adaptability to new data. To assess the inversion performance, we used a novel one-dimensional inversion error calculation index, the correlation area loss error, for a more accurate measurement. Numerical simulation experiments showed that the proposed method required only 2.121 ss to invert data from 100 observation points. The inversion efficiency was significantly superior to the conventional methods, maintaining excellent accuracy while effectively discerning subsurface electrical stratification in geophysics. Applying convolutional bidirectional long short-term memory neural networks to multi-dimensional and field data yielded results superior to those of conventional inversion, demonstrating the promising applicability and generalization of this approach. This study offers an efficient solution for shallow subsurface transient electromagnetic exploration and holds potential for application in other areas.

SummaryMany intracontinental basins form as broad depressions through prolonged, slow subsidence of the continental lithosphere. Such long-lived basins can record lithospheric processes over hundreds of millions of years, serving as important archives of lithospheric evolution. Since continental amalgamation in the Mesoproterozoic, the lithosphere beneath the intracontinental Canning Basin has been subject to several tectonic events, with extensive crustal reworking evidenced through different upper crust datasets. However, knowledge of the structure of the subcontinental lithospheric mantle is lacking. As a consequence, understanding the coupled evolution between surface and deep lithospheric processes, crucial to resolving basin formation, development, and survival, remains problematic. Here, we combine geochemical, geophysical, and petrophysical data within a thermodynamic modelling framework to determine the thermochemical properties, rheology, density, and seismic structure of the lithospheric and sublithospheric mantle beneath the Canning Basin. The results indicate a thick, rigid lithosphere with a maximum thickness of 185 km and strength of ca. 1$\times$1013 Pa m, and an anomalously Fe-enriched subcontinental lithospheric mantle with a Mg# of 88.6. This mantle structure is not consistent with pre-collisional fragments or a Precambrian collisional setting and may reflect magmatic refertilisation during high-volume mafic magmatic events. Potential candidate events are the ∼1070 Ma Warakurna, ∼825 Ma Gairdner, and ∼510 Ma Kalkarindji Large Igneous Provinces. The youngest of these is temporally and spatially correlated with and therefore interpreted to have influenced the Canning Basin formation. We propose that refertilisation caused a negatively buoyant subcontinental lithospheric mantle and prolonged subsidence and preservation of the basin, while the strong lithosphere ensured lithospheric stability and longevity.

SummaryCross-correlations of seismic ambient noise are frequently used to image Earth structure. Usually, tomographic studies assume that noise sources are uniformly distributed and interpret noise correlations as empirical Green’s functions. However, previous research suggests that this assumption can introduce errors in the estimated models, especially when noise correlation waveforms are inverted. In this paper, we investigate changes in subsurface models inferred from noise correlation waveforms depending on whether the noise source distribution is considered to be uniform. To this end, we set up numerical experiments that mimic a tomographic study in Southern California exploiting ambient noise generated in the Pacific Ocean. Our results show that if the distribution of noise sources is deemed uniform instead of being numerically represented in the wave simulations, the misfit of the estimated models increases. In our experiments, the model misfit increase ranges between 5 % and 21 %, depending on the heterogeneity of the noise source distribution. This indicates that assuming uniform noise sources introduces source-dependent model errors. Since the location of noise sources may change over time, these errors are also time-dependent. In order to mitigate these errors, it is necessary to account for the noise source distribution. The spatial extent to which noise sources must be considered depends on the propagation distance of the ambient noise wavefield. If only sources near the study area are considered, model errors may arise.

SummaryMoment tensor (MT) inversion is a classical geophysical inverse problem that infers a force-equivalent model of a seismic source from seismological observations. Like other inverse problems, the accuracy of the inversion depends on the reliability of the forward problem simulating waveforms from the source location through an Earth structural model. Apart from errors in data, the error in forward waveform simulation, also known as theory error, is a significant source of error contributing to the misfit function between the predicted and observed waveforms. Here, we set up numerical experiments to comprehensively probe the sensitivity of the linearized MT inversion to 3D regional Earth model errors, a known predominant factor of the theory error. Using the Monte Carlo method, we estimate the empirical structural covariance matrices to characterize the waveform mismatch due to the imperfect knowledge of Earth's structure. Firstly, although the inversion accuracy deteriorates with increasing model errors, incorporating the structural covariance matrices into the misfit function improves the accuracy of inversion results for all theorized error distributions. Secondly, we propose a slightly modified form of the structural covariance matrix, which further enhances the inversion outcome. Lastly, as the true structural errors are likely spatially correlated, we highlight the importance of adequately treating the correlation into the MT inversion because of its significant impact on inversion. Overall, as a preliminary effort in quantifying 3D structural errors on MT inversion, this study proves the computational feasibility by means of numerical experiments and will hopefully provide a way forward for future work on this topic.

SummaryThis study aims to expand on existing connections between magnetic minerals and hydrocarbons within petroleum systems. Previous studies have focussed on single-source petroleum systems whereas this study, for the first time, analyses a multi-source petroleum system to investigate potential correlations between different kerogen type source rocks and magnetic minerals. To do this, the study investigates the magnetic mineral characteristics of the Inner Moray Firth (IMF), UK North Sea, through room-, low-, and high-temperature techniques, and correlates this to published basin and petroleum systems modelling results that show a three-source hydrocarbon mix. Magnetic mineral analysis identifies extensive evidence for magnetite, goethite, and siderite, alongside more minor lepidocrocite and iron sulphides. Although we find that magnetite is ubiquitous within the IMF, its abundance is relatively low, and, in contrast, the relatively magnetically weak goethite is more likely the most abundant magnetic mineral throughout the IMF. In agreement with previous studies, we find magnetic enhancement at oil-water contacts (OWCs); however, here, we identify two different magnetic enhancement processes at OWCs in wells, which are dependent on the amount of sulphur available in the local environment. Wells with low levels of sulphur have increasing levels of magnetite towards the OWC, with the magnetic enhancement occurring at the top of the water-saturated section. Sulphur-rich environments display an increase in iron sulphides near the OWC at the bottom of the oil-saturated sediments. Additionally, we confirm the presence of siderite as indicator of upward vertical migration. Combining with petroleum system model predictions, we find direct links between iron hydroxide presence and Type I and II-III kerogen source rocks, and iron sulphide presence with Type II kerogen source rocks. This study shows the potential for further utilisation of magnetic mineral analysis within hydrocarbon exploration and petroleum system definition.

SummaryMagnetotelluric data are sometimes accompanied by ‘anomalous’ impedance phases ($\phi $xy and $\phi $yx) in the off-diagonal components deviating from the first (0º < $\phi $xy < 90º) or third (−180º < $\phi $yx < −90º) quadrant, especially in long-period bands. This phenomenon is called the phases out-of-quadrant (POQ). The POQ poses a challenge in Magnetotelluric modeling because simple one- or two-dimensional models cannot explain it. Previous studies have reported that strong inhomogeneity, anisotropy, or particular three-dimensional structures, such as the L-shaped or cross-shaped conductors, could explain the POQ. Aside from these models, we have discovered that a slanted columnar conductor also generates the POQ. Our systematic investigation through the synthetic forward modeling of an inclined conductive column with a varying geometry showed that the inclination angle and the column length may affect the POQ appearance. We investigated herein the behavior of the electric currents around the inclined conductive column embedded in a resistive half space. We found that the induced electric field in the region with the POQ tends to point in the opposite direction to the surrounding vectors. This result can reasonably explain the inverted phase in long-period bands. Furthermore, we confirmed that current is sucked into one end of the column, but discharged from the other end, suggesting that the column works as a current channel. The localized reverse vectors are associated with the current channeling along the inclined conductor, which generates the POQ. A volcanic conduit within a resistive host rock is one of the typical field examples of such an inclined channel. Our study suggests that the POQ is a helpful clue in imaging the geometry of a volcanic magma plumbing system through Magnetotelluric surveys.

SummaryObservations of the seafloor Rayleigh ellipticity contribute to seismic imaging in the ocean. To extract such observables from the arbitrarily oriented ocean-bottom seismometer (OBS) data, we develop an orthogonal-regression based approach to measure the waveform amplitude ratios of the unoriented horizontal and vertical components. The amplitude ratios are then used to calculate the Rayleigh ellipticity (and the sensor orientation angle). The robustness of our method is verified by applications to both the unoriented OBS data and the well oriented on-land seismic data. As we propose to calculate the Rayleigh ellipticity directly from the unoriented three-component data, the measurement process avoids the complexity arising from the surface wave non-great circle effects and uncertainties of the OBS sensor orientation angles. Overall the Rayleigh ellipticity measurements from our method are systematically higher than those by conventional analysis and show less uncertainties. Our analyses suggest that the Rayleigh ellipticity curve (14-60 s), which could be retrieved from the raw broadband OBS data, is effective to constrain the oceanic lithosphere structure, and the accurate measurement of Rayleigh ellipticity curve is important. The potential of seafloor Rayleigh ellipticity for seismic imaging in the ocean is evidenced by a case study of the Japan Basin, the Sea of Japan. Considering the insufficient station coverage in the ocean, the single-station measurement of seafloor Rayleigh ellipticity is of significance for OBS community.

SummaryInverse problems occur in many fields of geophysics, wherein surface observations are used to infer the internal structure of the Earth. Given the non-linearity and non-uniqueness inherent in these problems, a standard strategy is to incorporate a priori information regarding the unknown model. Sometimes a solution is obtained by imposing that the inverted model remains close to a reference model and with smooth lateral variations (e.g., a correlation length or a minimal wavelength are imposed). This approach forbids the presence of strong gradients or discontinuities in the recovered model. Admittedly, discontinuities, such as interfaces between layers, or shapes of geological provinces or of geological objects such as slabs can be a priori imposed or even suggested by the data themselves. This is however limited to a small set of possible constraints. For example, it would be very challenging and computationally expensive to perform a tomographic inversion where the subducting slabs would have possible top discontinuities with unknown shapes. The problem seems formidable because one cannot even imagine how to sample the prior space: is each specific slab continuous or broken into different portions having their own interfaces? No continuous set of parameters seems to describe all the possible interfaces that we could consider. To circumvent these questions, we propose to train a Generative Adversarial neural Network (GAN) to generate models from a geologically plausible prior distribution obtained from geodynamical simulations. In a Bayesian framework, a Markov chain Monte Carlo algorithm is used to sample the low-dimensional model space depicting the ensemble of potential geological models. This enables the integration of intricate a priori information, parametrized within a low-dimensional model space conducive to efficient sampling. The application of this approach is demonstrated in the context of a downscaling problem, where the objective is to infer small-scale geological structures from a smooth seismic tomographic image.

SummaryThis study examines error propagation from data space to model space during three-dimensional inversion of controlled-source electromagnetic data using a Gauss-Newton based algorithm. An expression for model parameter correction is obtained using higher-order generalised singular value decomposition for various regularisation strategies. Inverse modelling is performed for different types of noise employing distinct regularisation schemes to investigate the impact of error. Data corrupted with random noise suggests that the random noise mainly propagates when regularisation parameters are small, owing to the high-frequency nature of random noise. Furthermore, the random noise predominantly causes artefacts in the shallower part of the inverted model. However, it has little impact on the estimation of major anomalies because the anomaly primarily depends on the smoothly varying parts of data. These observations are valid for both isotropic and anisotropic inversions. Resistive geological anomalies, like vertical dyke or vertical fractures, may pose a significant challenge for isotropic inversion in terms of convergence and data fit, even if the subsurface is isotropic. On the other hand, anisotropic inversion performs remarkably well in such cases, showing faster convergence and better data fit than isotropic inversion. Anisotropic inversion is indispensable in the case of an anisotropic host medium, as isotropic inversion produces significant artefacts and poorer data fit. Numerical experiments suggest that, in general, anisotropic inversion produces relatively better data fit and faster convergence, even in the case of isotropic subsurface. However, due to the varying degree of sensitivity of CSEM data on thin resistive bodies, caution is required in interpreting an anisotropy obtained using anisotropic inversion. An investigation of field data also supports the observations obtained using synthetic experiments.

SummaryWe consider application of full waveform inversion (FWI) to radio-frequency electromagnetic (EM) data. Radio-frequency imaging (RIM) is a cross-borehole technique to image electromagnetic subsurface properties from measurements of transmitted radio-frequency waves. It is used in coal seam imaging, ore exploration and various engineering and civil engineering applications. RIM operates at frequencies from 50 kHz to several tens of MHz. It differs from other geophysical EM methods, because the frequency band includes the transition between the wave propagation and diffusion regimes. RIM data are acquired in two-dimensional cross-hole sections in a reciprocal manner. Traditionally, radio-frequency data are inverted by straight ray tomography because it is inexpensive and easy to implement. It is argued that due to attenuation, the sensitivity of the transmitted electric field is the strongest within the first Fresnel zone of the ray connecting the transmitter and receiver. While straight ray tomography is a simple to implement and fast method, the non-linearity in the relationship between model parameters and data is often strong enough to warrant non-linear inversion techniques. FWI is an iterative high resolution technique, in which the physical properties are updated to minimize the misfit between the measured and modelled wavefields. Full waveform techniques have been used and extensively studied for the inversion of seismic data, and more recently, they have been applied to the inversion of GPR data. Non-linear inversion methods for RIM data are less advanced. Their use has been hindered by the high cost of full wave modelling and the high conductivity contrasts of many RIM targets, and, to some extent, by the limitations of the measuring instruments. We present the first application of this methodology to simultaneous conductivity and permittivity inversion of RIM data. We implement the inversion in the frequency domain in two dimensions using L-BFGS optimization. We analyze the sensitivity of the data to the model parameters and the parameter trade-off and validate the proposed methodology on a synthetic example with moderate conductivity variations and localized highly conductive targets. We then apply the FWI methodology to a field data set from Sudbury, Canada. For the field data set, we determine the most appropriate preprocessing steps that take into account specific peculiarities of RIM: the insufficient prior information about the subsurface and the limitations of the measuring equipment. We show that FWI is applicable under the conditions of RIM and is robust to imperfect prior knowledge: we obtain satisfactory model recoveries starting from homogeneous initial models in all of our examples. Just as other methods, FWI underestimates large conductivity contrasts due to the loss of sensitivity of the transmitted electric field to the conductivity variations as the conductivity increases above a certain level. The permittivity inside high conductors can not be recovered, however, recovering permittivity variations in the resistive zones helps obtain better focused conductivity images with fewer artifacts. Overall, FWI produces cleaner, less noisy and higher resolution reconstructions than the methods currently used in practice.

SummaryWe present a computational technique to model hydroacoustic waveforms from teleseismic earthquakes recorded by mid-column Mermaid floats deployed in the Pacific, taking into consideration bathymetric effects that modify seismo-acoustic conversions at the ocean bottom and acoustic wave propagation in the ocean layer, including reverberations. Our approach couples axisymmetric spectral-element simulations performed for moment-tensor earthquakes in a one-dimensional solid Earth to a two-dimensional Cartesian fluid-solid coupled spectral-element simulation that captures the conversion from displacement to acoustic pressure at an ocean-bottom interface with accurate bathymetry. We applied our workflow to 1,129 seismograms for 682 earthquakes from 16 Mermaids owned by Princeton University that were deployed in the Southern Pacific as part of the South Pacific Plume Imaging and Modeling (SPPIM) project. We compare the modeled synthetic waveforms to the observed records in individually selected frequency bands aimed at reducing local noise levels while maximizing earthquake-generated signal content. The modeled waveforms match the observations very well, with a median correlation coefficient of 0.72, and some as high as 0.95. We compare our correlation-based travel-time measurements to measurements made on the same data sets determined by automated arrival-time picking and ray-traced travel-time predictions, with the aim of opening up the use of Mermaid records for global seismic tomography via full-waveform inversion.

SummaryGround penetrating radar (GPR) is becoming an increasingly important tool for understanding the shallow electrical structures of the earth and planets due to its adaptability to harsh detection environments, efficient data acquisition and accurate detection results. GPR full-waveform can simultaneously constrain the permittivity and resistivity of the medium, providing more comprehensive geophysical information and reducing the non-uniqueness of inversion. However, given the highly non-linear inverse problem and the massive data resulted from high temporal and spatial samplings, traditional full-waveform inversion algorithms are prohibitively costly. Inspired by Google's vision semantic segmentation system, we develop a robust deep learning-guided network that integrates geology and geophysics knowledge to support the real-time translation of zero-offset GPR data into dual-parameter electrical structures. We test our proposed network using synthetic data, which demonstrates that the algorithm can provide an accurate dual-parameter electrical model from a GPR sounding in milliseconds on a common laptop PC, exhibiting high robustness and adaptability to noise interference and extreme values of model parameters. We also apply our network to field data gathered for pollutant investigation in the US. The resulting dual-parameter structure provides a more comprehensive and realistic depiction of subsurface electrical properties and reveals the migration and aging of pollutants. Our algorithm's real-time and accurate advantages are expected to further unleash the potential of GPR technology and enable it to play a more significant role in earth and planetary exploration.

SummaryBackazimuthal variations in the shear wave splitting of core-refracted shear waves (SKS, SKKS, and PKS phases, jointly referred to as XKS) at the Black Forest Observatory (BFO, Southwest Germany) indicate small-scale lateral and partly vertical variations of the seismic anisotropy. However, existing anisotropy studies and models for the nearby Upper Rhine Graben (URG) area in the northern Alpine foreland are mostly based on short-term recordings and by this suffer from a limited backazimuthal coverage and averaging over a wide or the whole backazimuth range. To identify and delimit laterally confined anisotropy regimes in this region, we carry out XKS splitting measurements at six neighbouring (semi-)permanent broadband seismological recording stations (inter-station distance 10-80 km). We manually analyse long-term (partly > 20 yr) recordings to achieve a sufficient backazimuthal coverage to resolve complex anisotropy. The splitting parameters (fast polarization direction $\phi $, delay time $\delta t$) are determined in a single- and multi-event analysis. We test structural anisotropy models with one layer with horizontal or tilted symmetry axis and with two layers with horizontal symmetry axes (transverse isotropy). To account for lateral variations around a single recording site, modelling is compared for the whole and for limited backazimuth ranges. Based on this, we provide a 3-D block model with spatial variation of anisotropic properties. Based on delay times > 0.3 s and missing discrepancies between SKS and SKKS phases, which do not support lower mantle anisotropy, the found anisotropy is placed in the lithosphere and asthenosphere. The spatial distribution as well as the lateral and backazimuthal variations of the splitting parameters confirm lateral and partly vertical variations in anisotropy. On the east side of the URG, we suggest two anisotropic layers in the Moldanubian Zone (south) and one anisotropic layer in the Saxothuringian Zone (north). In the Moldanubian Zone, a change of the fast polarization directions is observed between the east and the west side of the URG, indicating different textures. At the boundary between the two terranes, an inclined anisotropy is modelled which may be related with deformation during Variscan subduction. Regarding the observation of numerous null measurements and inconsistent splitting parameters, especially (southwest of BFO) in the southern URG, different hypothesis are tested: scattering of the seismic wavefield due to small-scale lateral heterogeneities, a vertical a-axis due to a vertical mantle flow related to the Kaiserstuhl Volcanic Complex, as well as a different preferred orientation of the olivine crystals (not A-type, but C-type) due to specific ambient conditions (high temperature, water content).

SummaryWe developed a short-period Pn magnitude scale mb(Pn) for earthquakes along the equatorial Mid-Atlantic Ridge. Due to low signal-to-noise ratios, teleseismic body wave magnitude and long-period surface wave magnitude cannot be confidently determined for small earthquakes of mb < 4. Local magnitude scales are also not useful for these events because the oceanic environment does not allow the propagation of crustal phases. However, regional high-frequency Pn waves from these small- to moderate-size (mb 3–6) earthquakes are well recorded in the equatorial Atlantic region and can be used to assign magnitudes. We measured over 2 041 Pn peak amplitudes on vertical records from about 20 stations in northeastern Brazil and 11 stations in western Africa in the distance range of 700–3,700 km. We analyzed data from 189 events from the global centroid moment tensor catalog to tie our mb(Pn) scale to MW so that seismic moments can be readily estimated. Pn arrivals show apparent group velocity between 7.9 km/s at short ranges (∼1,000 km) and up to 9.1 km/s at 3,500 km. The measured peak amplitudes have a frequency between 0.8 and 3 Hz at 1 000–1,800 km, but at greater distances, 1 800–3,700 km, they show a remarkably consistent frequency of about 0.8 Hz. The peak amplitude attenuates at a higher rate at short distances (∼0.65 magnitude units between 700–2,000 km) but attenuates at a lower rate at long distances (∼0.35 magnitude units between 2 000 and 3,700 km). The low rate of amplitude decay with distance and nearly constant frequency content of the peak amplitudes suggest that Pn waves propagate efficiently in the lower part of the upper mantle in the equatorial Atlantic Ocean basins. These are important attributes of oceanic Pn waves that can be used to assign magnitude for small- to moderate-size earthquakes in the equatorial mid-Atlantic region. The estimated station corrections correlate well with upper mantle low-velocity anomalies, especially in Brazil.

SummaryThe Earth's subsurface structure provides critical insights into sustainable resource management and geologic evolution. The airborne electromagnetic (AEM) method is an efficient data acquisition technique and can be used to image the underground resistivity structure with high spatial resolution. However, inversion of the increasingly huge volume of AEM data poses a heavy computational burden. In this study, we develop a hybrid deep learning-based approach by employing the physics-guided neural network (PGNN) which incorporates the governing physical laws into the loss function to solve the AEM inverse problem. The PGNN integrates the strength of data-driven method for representation learning with electromagnetic laws and allows for the underlying physical constraints to be strictly satisfied. We validate the effectiveness of our approach using both synthetic and field datasets. Compared with the classic Gauss-Newton method, our PGNN inversion system shows strong robustness against multiple noise sources and reduces the risk of being trapped in local extrema. Moreover, the PGNN-inverted results are physically more consistent with the AEM observations compared to the purely data-driven approach. Application to the field AEM data from Northern Australia demonstrates that the PGNN-based inversion framework effectively estimates the subsurface electrical properties with considerable lateral continuity and significantly higher efficiency, completing the inversion of more than 2734000 AEM soundings taking only minutes on a common PC. Our proposed PGNN-based method shows great promise for large-scale underground resistivity imaging, and the well-identified subsurface resistivity structure can effectively improve our understanding of resource distributions and geological hazards.