ALBERT

Description: Frequency-domain loop–loop electromagnetic (EM) methods are sensitive to the magnetic susceptibility of the Earth as well as its resistivity. Thus, inversion techniques have been used to simultaneously reconstruct both resistivity and susceptibility models from EM data. However, to take full advantage of inversion methods, calibration errors must be assessed and removed because ignoring them can result in misleading models. We present a multidimensional inversion method that jointly inverts EM and direct current (DC) resistivity data to derive offset errors as well as resistivity and susceptibility models, assuming that calibration errors can be represented by in-phase and quadrature offsets at each frequency. Addition of independent data such as DC data is effective for more accurately estimating the offsets, resulting in more reliable subsurface models. Synthetic examples involving small-loop EM data show that simultaneous inversion for resistivity and susceptibility is not stable, because of strong correlations between in-phase offset parameters and background susceptibility, but that the offsets are well determined when the data misfit is reduced rapidly in the early iteration step. Improvements achieved by joint inversion are mainly on the resistivity model. For airborne electromagnetic (AEM) data, the inversion process is stable, because AEM data are acquired using more loop–loop geometries and a wider range of frequencies. As a result, both the resistivity and susceptibility models are significantly improved by joint inversion.

Description: Seismic reflections from the oceanic water column contain information about ocean temperature and salinity. Even though seismic waveform inversion is effective for studying oceanic structure, its application is limited in the absence of sufficient direct temperature/velocity measurements. Here, two methods are developed to invert pre-stack seismic waveform data for temperature and salinity when independent temperature/velocity data are sparse or unavailable, allowing estimation of water-column temperature/salinity from any marine seismic reflection data set. The first method combines a genetic algorithm (GA) with non-linear least squares inversion, and the second method is a parallel implementation of a GA. Both methods produce results to an accuracy between 0 and 0.1 °C in estimating temperature when applied to a field data set from the South China Sea. Although the second approach is superior, it is computationally demanding and requires large parallel computers. The first approach runs extremely fast on parallel computers and can even be run on much smaller machines to provide results in a reasonable runtime. While both methods are viable choices for estimating temperature and salinity, the choice of one over the other will largely depend upon the available computational resources and the time frame within which the inversion needs to be completed.

Description: We present the first inversion of geomagnetic Sq data in a framework of 3-D conductivity models. This problem has been considered as immensely difficult due to the complex spatial structure of the Sq source which, in addition, varies with season and solar activity. Recently, we developed a 3-D electromagnetic (EM) inversion solution that allows one to work in a consistent manner with data that originates from sources, irrespective of their spatial complexity. In this paper, we apply our 3-D EM inversion scheme to Sq data collected during the Australian Wide Array of Geomagnetic Stations project. Within this project, three components of the geomagnetic field were recorded between 1989 November and 1990 December with the use of 53 portable vector magnetometers. The instruments were distributed over the Australian mainland with an average spacing of 275 km between sites. Inverting this unique—in a sense of its spatial regularity, density and long operational time—data set, we recovered the 3-D conductivity distribution beneath Australia at upper mantle depths (100–520 km). This depth range was justified in the paper from resolution studies using checkerboard tests. In addition, we performed extensive modelling to estimate quantitatively the influence of various factors on Sq signals, namely from hypothetical anomalies, inaccuracy in the source, ocean, and model discretization. As expected, the ocean (coastal) effect appeared to be the largest so that it has to be accounted for during 3-D inversion as accurately as possible. Our 3-D inversions—of data from either single or multiple days—revealed a strong offshore conductor near the south-east coast of Australia, which persists at all considered depths. Varying in details, this anomaly is remarkably robust irrespective of the considered day(s). We compared our results to those obtained from a different inversion scheme and an independent induction data set, and observed encouraging similarity. Combination of the two results suggests, that this conductor continues to the base of the mantle transition zone at 660 km. The nature of this anomaly is not fully understood but one possible explanation is that it is attributed to a reservoir responsible for three hotspots in the region.

Description: The activation of Late Quaternary faults in the Central Apennines (Italy) could generate earthquakes with magnitude of about 6.5, and the Monte Vettore fault system probably belongs to the same category of seismogenetic faults. Such structure has been defined ‘silent’, because of its geological and geomorphological evidences of past activation, but the absence of historical records in the seismic catalogues to be associated with its activation. The ‘Piano di Castelluccio’ intramountain basin, resulting from the Quaternary activity of normal faults, is characterized by a secondary fault strand highlighted by a NW–SE fault scarp: it has been already studied through palaeoseismological trenches, which highlighted evidences of Quaternary shallow faulting due to strong earthquakes, and through a 2-D ground penetrating radar (GPR) survey, showing the first geophysical signature of faulting for this site. Within the same place, a 3-D GPR volume over a 20 20 m area has been collected. The collection of radar echoes in three dimensions allows to map both the vertical and lateral continuity of shallow geometries of the fault zone (Fz), imaging features with high resolution, ranging from few metres to centimetres and therefore imaging also local variations at the microscale. Several geophysical markers of faulting, already highlighted on this site, have been taken as reference to plan the 3-D survey. In this paper, we provide the first 3-D subsurface imaging of an active shallow fault belonging to the Umbria-Marche Apennine highlighting the subsurface fault geometry and the stratigraphic sequence up to a depth of about 5 m. From our data, geophysical faulting signatures are clearly visible in three dimensions: diffraction hyperbolas, truncations of layers, local attenuated zones and varying dip of the layers have been detected within the Fz. The interpretation of the 3-D data set provided qualitative and quantitative geological information in addition to the fault location, like its geometry, boundaries and an estimation of the fault throw.

Description: A 3-D magnetotelluric (MT) inversion code using unstructured tetrahedral elements has been developed in order to correct the topographic effect by directly incorporating it into computational grids. The electromagnetic field and response functions get distorted at the observation sites of MT surveys because of the undulating surface topography, and without correcting this distortion, the subsurface structure can be misinterpreted. Of the two methods proposed to correct the topographic effect, the method incorporating topography explicitly in the inversion is applicable to a wider range of surveys. For forward problems, it has been shown that the finite element method using unstructured tetrahedral elements is useful for the incorporation of topography. Therefore, this paper shows the applicability of unstructured tetrahedral elements in MT inversion using the newly developed code. The inversion code is capable of using the impedance tensor, the vertical magnetic transfer function (VMTF), and the phase tensor as observational data, and it estimates the subsurface resistivity values and the distortion tensor of each observation site. The forward part of the code was verified using two test models, one incorporating topographic effect and one without, and the verifications showed that the results were almost the same as those of previous works. The developed inversion code was then applied to synthetic data from a MT survey, and was verified as being able to recover the resistivity structure as well as other inversion codes. Finally, to confirm its applicability to the data affected by topography, inversion was performed using the synthetic data of the model that included two overlapping mountains. In each of the cases using the impedance tensor, the VMTF and the phase tensor, by including the topography in the mesh, the subsurface resistivity was determined more proficiently than in the case using the flat-surface mesh. Although the locations of the anomalies were not accurately estimated by the inversion using distorted impedance tensors due to the slightly undervalued gain, these locations were correctly estimated by using undistorted impedance tensors or adding VMTFs in the data. Therefore, it can be concluded that the inversion using the unstructured tetrahedral element effectively prevents the misinterpretation of subsurface resistivity and recovers subsurface resistivity proficiently by representing the topography in the computational mesh.

Description: With advancing computational resources, 3-D inversion techniques have become feasible in recent years and are now a more widely used tool for magnetotelluric (MT) data interpretation. Galvanic distortion caused by small-scale near-surface inhomogeneities remains an obstacle for 3-D MT inversion which so far has experienced little attention. If not considered properly, the effect on 3-D inversion can be immense and result in erroneous subsurface models and interpretations. To tackle the problem we implemented inversion of the distortion-free phase tensor into the ModEM inversion package. The dimensionless phase tensor components describe only variations of the conductivity structure. When inverting these data, particular care has to be taken of the conductivity structure in the a priori model, which provides the reference frame when transferring the information from phase tensors into absolute conductivity values. Our results obtained with synthetic data show that phase tensor inversion can recover the regional conductivity structure in presence of galvanic distortion if the a priori model provides a reasonable assumption for the regional resistivity average. Joint inversion of phase tensor data and vertical magnetic transfer functions improves recovery of the absolute resistivity structure and is less dependent on the prior model. We also used phase tensor inversion for a data set of more than 250 MT sites from the central San Andreas fault, California, where a number of sites showed significant galvanic distortion. We find the regional structure of the phase tensor inversion results compatible with previously obtained models from impedance inversion. In the vicinity of distorted sites, phase tensor inversion models exhibit more homogeneous/smoother conductivity structures.

Description: We have developed an algorithm, which we call HexMT, for 3-D simulation and inversion of magnetotelluric (MT) responses using deformable hexahedral finite elements that permit incorporation of topography. Direct solvers parallelized on symmetric multiprocessor (SMP), single-chassis workstations with large RAM are used throughout, including the forward solution, parameter Jacobians and model parameter update. In Part I, the forward simulator and Jacobian calculations are presented. We use first-order edge elements to represent the secondary electric field ( E ), yielding accuracy O ( h ) for E and its curl (magnetic field). For very low frequencies or small material admittivities, the E -field requires divergence correction. With the help of Hodge decomposition, the correction may be applied in one step after the forward solution is calculated. This allows accurate E -field solutions in dielectric air. The system matrix factorization and source vector solutions are computed using the MKL PARDISO library, which shows good scalability through 24 processor cores. The factorized matrix is used to calculate the forward response as well as the Jacobians of electromagnetic (EM) field and MT responses using the reciprocity theorem. Comparison with other codes demonstrates accuracy of our forward calculations. We consider a popular conductive/resistive double brick structure, several synthetic topographic models and the natural topography of Mount Erebus in Antarctica. In particular, the ability of finite elements to represent smooth topographic slopes permits accurate simulation of refraction of EM waves normal to the slopes at high frequencies. Run-time tests of the parallelized algorithm indicate that for meshes as large as 176 x 176 x 70 elements, MT forward responses and Jacobians can be calculated in ~1.5 hr per frequency. Together with an efficient inversion parameter step described in Part II, MT inversion problems of 200–300 stations are computable with total run times of several days on such workstations.

Description: Following the creation described in Part I of a deformable edge finite-element simulator for 3-D magnetotelluric (MT) responses using direct solvers, in Part II we develop an algorithm named HexMT for 3-D regularized inversion of MT data including topography. Direct solvers parallelized on large-RAM, symmetric multiprocessor (SMP) workstations are used also for the Gauss–Newton model update. By exploiting the data-space approach, the computational cost of the model update becomes much less in both time and computer memory than the cost of the forward simulation. In order to regularize using the second norm of the gradient, we factor the matrix related to the regularization term and apply its inverse to the Jacobian, which is done using the MKL PARDISO library. For dense matrix multiplication and factorization related to the model update, we use the PLASMA library which shows very good scalability across processor cores. A synthetic test inversion using a simple hill model shows that including topography can be important; in this case depression of the electric field by the hill can cause false conductors at depth or mask the presence of resistive structure. With a simple model of two buried bricks, a uniform spatial weighting for the norm of model smoothing recovered more accurate locations for the tomographic images compared to weightings which were a function of parameter Jacobians. We implement joint inversion for static distortion matrices tested using the Dublin secret model 2, for which we are able to reduce nRMS to ~1.1 while avoiding oscillatory convergence. Finally we test the code on field data by inverting full impedance and tipper MT responses collected around Mount St Helens in the Cascade volcanic chain. Among several prominent structures, the north–south trending, eruption-controlling shear zone is clearly imaged in the inversion.

Description: New methods are required to combine the information contained in the passive electrical and seismic signals to detect, localize and monitor hydromechanical disturbances in porous media. We propose a field experiment showing how passive seismic and electrical data can be combined together to detect a preferential flow path associated with internal erosion in a Earth dam. Continuous passive seismic and electrical (self-potential) monitoring data were recorded during a 7-d full-scale levee (earthen embankment) failure test, conducted in Booneschans, Netherlands in 2012. Spatially coherent acoustic emissions events and the development of a self-potential anomaly, associated with induced concentrated seepage and internal erosion phenomena, were identified and imaged near the downstream toe of the embankment, in an area that subsequently developed a series of concentrated water flows and sand boils, and where liquefaction of the embankment toe eventually developed. We present a new 4-D grid-search algorithm for acoustic emissions localization in both time and space, and the application of the localization results to add spatially varying constraints to time-lapse 3-D modelling of self-potential data in the terms of source current localization. Seismic signal localization results are utilized to build a set of time-invariant yet spatially varying model weights used for the inversion of the self-potential data. Results from the combination of these two passive techniques show results that are more consistent in terms of focused ground water flow with respect to visual observation on the embankment. This approach to geophysical monitoring of earthen embankments provides an improved approach for early detection and imaging of the development of embankment defects associated with concentrated seepage and internal erosion phenomena. The same approach can be used to detect various types of hydromechanical disturbances at larger scales.

Description: The Rogaland Igneous Complex (RIC) in southern Norway intruded into post-Sveconorwegian granulite facies crust ~930 Ma. It includes three massif anorthosites, several small leuconorite bodies and the ~7 km thick norite-quartz mangerite layered Bjerkreim-Sokndal (BKS) intrusion. The intrusion consists of five rhythmic megaunits created by repeated magma influxes topped by a transition zone and more evolved mangerites and quartz mangerites. Over 70 palaeomagnetic sites have been collected in the BKS, sampling all the megacyclic subunits and overlying mangerites. Remanence within the BKS is held in hemo-ilmenite-only rocks (lower parts of the megacyclic units), mixed hemo-ilmenite and magnetite rocks (upper parts of the lower megacyclic units) and magnetite only rocks in the upper highest megacyclic unit and overlying mangerites. Due to the different oxides present magnetic susceptibility varies over four orders of magnitude with a bimodal distribution (mean susceptibility of 6.4 x 10 –3 SI for hemo-ilmenite rocks, and 8.7 x 10 –2 SI for magnetite rocks). NRM values do not show a strong bimodal distribution as many of the rocks lacking magnetite have hemo-ilmenite with strong lamellar magnetism; average NRM for the entire suite is 8.83 A m –1 . All sites within the cyclic part of the intrusion have stable remanence and produce well-clustered site means. Samples from the upper mangerite rocks, dominated by MD magnetite, are commonly unstable and not all sites provide acceptable data. Mean directions for 66 sites spanning the entire intrusion are I = –73.5°, D = 303.4°, with α 95 = 3.7° and k = 24. The resulting pole position is at 35.9°S and 217.9°E, with a palaeolatitude for this part of Baltica of –59.4°. Examination of the magnetic mineralogy combined with geochronology for RIC rocks and cooling rates for the region yields an age of magnetization of 916 Ma. Metamorphic country rocks yield similar directions at least 10 km from the contact, confirming the presence of a contact aureole around the intrusion. Comparison to other early Neoproterozoic palaeomagnetic poles from southern Scandinavia confirms the high southern latitude position of Baltica at this time, and combined with the few other ~900 Ma poles from Baltica affirm the presence of the Rodinia supercontinent at this time.