Geophysical Journal International

SummaryClay minerals are extensively used in a wide range of applications. In particular, clay-bearing formations are considered as suitable radioactive waste repository. Electrical resistivity tomography is an appropriate tool to monitor the properties of clay-bearing locations. However, an inherent drawback of a conventional resistivity survey is its ambiguity in distinguishing between the effects of groundwater salinity, clay content, and porosity. A discrimination can be achieved on the basis of the induced polarization method that provides a complex conductivity. The main purpose of this study is the investigation of the complex conductivity of clay samples with a special focus on the contribution of surface conductivity produced by an excess of ions in the electrical double layer coating the solid particles. Six clay mixtures were selected that include an almost pure kaolinite sample, a sample consisting of a mixture of kaolinite, illite, and smectite, a crushed saponite breccia, a Ca-bentonite sample, and two illite clay samples. Besides the enriched kaolinite, the other samples are natural geomaterials that contain more than 40 weight per cent clay minerals. The mineralogical compositions of the samples were determined by quantitative X-ray diffraction analysis. The clay powder was mixed with a varying volume of sodium chloride solution to get plastic state clay samples with varying water content. The samples were investigated by the spectral induced polarization method in a frequency range between 1 mHz and 1 kHz. The resulting complex conductivity spectra indicate a decrease of the real part of the electrical conductivity with rising water content for the illite, bentonite and saponite breccia samples. The overall conductivity of these clay samples is dominated by their surface conductivity. In contrast, the electrical conductivity of kaolinite and kaolinite-illite mixture does not show any significant changes with the water content. For all samples, the imaginary part of electrical conductivity increases at low water content. The real part of the surface conductivity indicates a linear dependence on the volumetric clay content. The slope of this linear relationship can be used to distinguish the types of clay. The ratio between imaginary conductivity and surface conductivity, which decreases with increasing clay content, proves to be a suitable parameter that characterizes the connectivity of clay aggregates in the sample. The surface conductivity of the pure kaolinite sample has been determined in an additional multi-salinity experiment. The resulting surface conductivity is in good agreement with the experiment of varying water content. The multi-salinity experiment has shown that the resulting petrophysical parameters depend on the procedure of sample packing, which may lead to anisotropy. The effect of anisotropy is attributed to the alignment of the plate-like kaolinite particles in the course of the packing and consolidation procedure.

SummaryThe aim of this study is to investigate the aftershock sequence data recorded by a dense temporary seismological network deployed in the epicentral area of the 2013 April 9 Shonbeh (Kaki) earthquake, located in the south of the Simply Folded Belt of the Zagros (Iran). For a comprehensive understanding, coseismic displacements of the Shonbeh earthquake have been investigated using Interferometric Synthetic Aperture Radar (InSAR) data. The epicentral distribution of high-resolution relocated aftershocks shows NW-SE and N-S trending of seismicity. The aftershocks are confined between ∼3 and ∼14 km depth, which implies that the rupture occurred mostly within the sedimentary cover beside the fault parameters retrieved from InSAR modeling. Projection of precisely located aftershocks on NE-oriented section and InSAR ground displacement data are consistent with both NW-trending NE- and SW-dipping fault segments. We observe a NNW-SSE right-lateral strike-slip motions that accommodate oblique convergence and differential motion between the North and Central Zagros. The spatial pattern and focal mechanisms of aftershocks are consistent with a distributed deformation between NW-SE trending reverse and N-S trending right-lateral strike-slip fault segments in the south of the Kazerun transition zone that accommodates a wide shear zone.

SummaryThe solution of the remote reference method, a frequently used technique in magnetotelluric data processing, can be viewed as a product of the two-input-multiple-output relationship between the local electromagnetic field and the reference field at a remote station. By applying a robust estimator to the two-input-multiple-output system, one can suppress the influence of outliers in the local magnetic field as well as those in the local electric field based on regression residuals. Therefore, this study develops a new robust remote reference estimator with the aid of robust multivariate linear regression. By applying the robust multivariate regression S-estimator to the multiple-output system, the present work derives a set of equations for robust estimates of the transfer function, noise variances, and the scale of the Mahalanobis distance simultaneously. The noise variances are necessary for the multivariate analysis to normalize the residuals of dependent variables. The Mahalanobis distance, a distance measure for multivariate data, is a commonly-used indicator of outliers in multivariate statistics. By updating those robust estimates iteratively, the new robust remote reference estimator seeks the transfer function that minimizes the robust scale estimate of the Mahalanobis distance. The developed estimator can avoid bias in the magnetotelluric transfer function even if there are significant noises in the reference magnetic field and handle outlying data more robustly than previously proposed robust remote reference estimators. The authors applied the developed method to a synthetic dataset and real-world data. The test results demonstrate that the developed method downweights outliers in the local electric and magnetic fields and gives an unbiased transfer function.

SummaryAtmospheric pressure changes on Earth’s surface can deform the solid Earth. Sorrells derived analytical formulas for displacement in a homogeneous, elastic half-space, generated by a moving surface pressure source with speed c. Ben-Menahem and Singh derived formulas when an atmospheric P-wave impinges on Earth’s surface. For a P-wave with an incident angle close to the grazing angle, which essentially meant a slow apparent velocity ca in comparison to P-wave (α′) and S-wave velocities (β′) in the Earth (ca ≪ β′ < α′), they showed that their formulas for solid-earth deformations become identical with Sorrells’ formulas if ca is replaced by c. But this agreement was only for the asymptotic cases (ca ≪ β′). The first point of this paper is that the agreement of the two solutions extends to non-asymptotic cases, or when ca/β′ is not small. The second point is that the angle of incidence in Ben-Menahem and Singh’s problem does not have to be the grazing angle. As long as the incident angle exceeds the critical angle of refraction from the P-wave in the atmosphere to the S-wave in the solid Earth, the formulas for Ben-Menahem and Singh’s solution become identical to Sorrell’s formulas. The third point is that this solution has two different domains depending on the speed c (or ca) on the surface. When c/β′ is small, deformations consist of the evanescent waves. When c approaches Rayleigh-wave phase velocity, the driven oscillation in the solid Earth turns into a free oscillation due to resonance and dominates the wave field. The non-asymptotic analytical solutions may be useful for the initial modeling of seismic deformations by fast-moving sources, such as those generated by shock waves from meteoroids and volcanic eruptions because the condition c/β′ ≪ 1 may be violated for such fast-moving sources.

SummaryApplications of machine learning in seismology have greatly improved our capability of detecting earthquakes in large seismic data archives. Most of these efforts have been focused on continental shallow earthquakes, but here we introduce an integrated deep-learning-based workflow to detect deep earthquakes recorded by a temporary array of ocean-bottom seismographs (OBSs) and land-based stations in the Tonga subduction zone. We develop a new phase picker, PhaseNet-TF, to detect and pick P- and S-wave arrivals in the time-frequency domain. The frequency-domain information is critical for analyzing OBS data, particularly the horizontal components, because they are contaminated by signals of ocean-bottom currents and other noise sources in certain frequency bands. PhaseNet-TF shows a much better performance in picking S waves at OBSs and land stations compared to its predecessor PhaseNet. The predicted phases are associated using an improved Gaussian Mixture Model Associator GaMMA-1D and then relocated with a double-difference package teletomoDD. We further enhance the model performance with a semi-supervised learning approach by iteratively refining labelled data and retraining PhaseNet-TF. This approach effectively suppresses false picks and significantly improves the detection of small earthquakes. The new catalogue of Tonga deep earthquakes contains more than 10 times more events compared to the reference catalogue that was analyzed manually. This deep-learning-enhanced catalogue reveals Tonga seismicity in unprecedented detail, and better defines the lateral extent of the double-seismic zone at intermediate depths and the location of 4 large deep-focus earthquakes relative to background seismicity. It also offers new potential for deciphering deep earthquake mechanisms, refining tomographic models, and understanding of subduction processes.

SummaryBased on both forced oscillation and ultrasonic pulse transmission methods, we investigated solid pore infill influences on rock elastic moduli in a broad frequency range $[ {1 - 3000,\ {{10}}^6} ]$ Hz for different differential pressures. For a Berea sandstone sample, filled sequentially by solid (${22}^{\rm{o}}{\rm{C}}$), quasi-solid (${26}^{\rm{o}}{\rm{C}}$) and liquid (${34}^{\rm{o}}{\rm{C}}$) octadecane, a frequency-dependence was found for the Poisson's ratio, Young's modulus and bulk modulus, nevertheless, these elastic parameters were strongly suppressed by increasing pressures. Experimental measurements showed that shear wave velocity and modulus of solid-octadecane-filled samples are significantly larger than those of the dry and liquid-octadecane-filled ones, implying the potential stiffening effects related to solid infill in compliant pores. A three porosity structure model, which describes the solid stiffening effects related to equant, compliant and the intermediate pores with aspect ratios larger than those of compliant pores but much less than those of stiff pores, was used to compare against the experimentally measured elastic properties for octadecane pore infill, together with several other fluid/solid substitution theories. The agreement between experimental measurements and theoretical predictions is reasonably good for the sandstone tested, providing that the three porosity model can be applied for pressure- and frequency-dependent elastic moduli estimations for a viscoelastic pore-infill-saturated sandstone. Evaluating the combined squirt flow mechanism responsible for the observed moduli dispersion and attenuation is of great importance to reduce potential errors in seismic AVO inversion and 4D seismic monitoring of gas-hydrate or bitumen-saturated reservoir, especially for reservoir rocks with complex microstructures and heterogeneous pore types.

SummaryMagnetic susceptibility behaviour around the Verwey transition of magnetite (≈ 125 K) is known to be sensitive to stress, composition and oxidation. From the isotropic point (≈ 130 K) to room temperature, decreasing magnetic susceptibility indicates an increase in magnetocrystalline anisotropy. In this study, we present a model which numerically analyses low-temperature magnetic susceptibility curves (80 to 280 K) of an experimentally shocked (up to 30 GPa) and later heated (973 K) magnetite ore. To quantify variations of the transition shape caused by both shock and heating, the model statistically describes local variations in the Verwey transition temperature within bulk magnetite. For the description, Voigt profiles are used, which indicate variations between a Gaussian and a Lorentzian character. These changes are generally interpreted as variations in the degree of correlation between observed events, i.e. between local transition temperatures in the model. Shock pressures exceeding the Hugoniot elastic limit of magnetite ($ \ge $ 5 GPa) cause an increase in transition width and Verwey transition temperature, which is partially recovered by heat treatment. Above the Verwey transition temperature, susceptibility variations related to the magnetocrystalline anisotropy are described with an exponential approach. The room temperature magnetic susceptibility relative to the maximum near the isotropic point is reduced after shock, which is related to grain size reduction. Since significant oxidation and cation substitution can be excluded for the studied samples, variations are only attributed to changes in elastic strain associated with shock-induced deformation and annealing due to heat treatment. The shocked magnetite shows a high correlation between local transition temperatures which is reduced by heat treatment. The model allows a quantitative description of low-temperature magnetic susceptibility curves of experimentally shocked and subsequently heat-treated polycrystalline magnetite around the Verwey transition temperature. The curves are accurately reproduced within the experimental uncertainties. Further applications for analysing magnetite-bearing rocks seem possible if model parameters, such as for oxidation are included into the model.

SummaryFor accurate modeling of groundwater flow and transport processes within an aquifer, precise knowledge about hydraulic conductivity K and its small-scale heterogeneities is fundamental. Methods based on pumping tests, such as hydraulic tomography (HT), allow for retrieving reliable K-estimates, but are limited in their ability to image structural features with high resolution, since the data from time-consuming hydraulic tests are commonly sparse. In contrast, geophysical methods like induced polarization (IP) can potentially yield structural images of much higher resolution, but depend on empirical petrophysical laws that may introduce significant uncertainties to the K-estimation. Therefore, this paper presents a joint inversion procedure for both HT and IP data, which allows for combining the complementary abilities of both methods. Within this approach, a travel time inversion is applied to the HT data, while the IP inversion is based on a full-decay time-domain forward response, as well as a re-parameterization of the Cole-Cole model to invert for K directly. The joint inversion is tested on a synthetic model mimicking horizontally layered sediments, and the results are compared with the individual HT and IP inversions. It is shown that jointly inverting both data sets consistently improves the results by combining the complementary sensitivities of the two methods, and that the inversion is more robust against changes in the experimental setups. Furthermore, we illustrate how a joint inversion approach can correct biases within the petrophysical laws by including reliable K-information from hydraulic tests and still preserving the high-resolution structural information from IP. The different inversion results are compared based on the structural similarity index (SSIM), which underlines the robustness of the joint inversion compared to using the data individually. Hence, the combined application of HT and IP within field surveys and a subsequent joint inversion of both data sets may improve our understanding of hydraulically relevant subsurface structures, and thus the reliability of groundwater modeling results.

SummaryTaking advantage of the simultaneous recording during 471 days between 2019 and 2021 by two superconducting gravimeters installed at the surface and 520 m under the surface at the Low Noise Underground Laboratory (LSBB) in Rustrel, France, we investigate whether a difference between the tidal gravity signals at the two locations can be detected. First, we model the periodical variations of the Earth’s gravity owing to the tidal influence from the Sun and Moon, at the Earth’s surface and at shallow depths. We provide analytical formulas for the Love numbers, gravimetric factor and gravity variation of simple spherical planetary models. We also numerically compute those parameters and function for a realistic spherical Earth model. We find that the fractional difference between the semi-diurnal tidal gravity variations at the surface and 520 m below is as small as 8.5 10−5. We next evaluate the effect on the amplitude of the recorded gravity signal due to the calibration factors of the two superconducting gravimeters at LSBB. Finally, we compute the spectra of the difference between the gravity variations measured on and under the surface in the semi-diurnal band of the M2 tidal wave. We find that the uncertainties associated to the calibration factors are larger than the theoretical or observational difference between the tidal gravity variations on the surface and at a 520-m depth.

SummaryWe define double (S1S2) and triple (PS1S2) singularity points and their degeneracy classes in triclinic anisotropic media. We derive equations for the group velocity image for all these cases. The degeneracy classes are defined by factorization of quadratic (double singularity point) and cubic (triple singularity point) forms with three variables.