La magnéto-tellurique (MT) (Cagniard 1953, Tikhonov 1950) est une technique d'exploration de la Terre basée sur des mesures de champs électromagnétiques (EM). Une source naturelle (non artificielle) harmonique en temps et située dans l'ionosphère (Weaver 1994) produit un champ EM régi par les équations de Maxwell. Les champs électromagnétiques sont enregistrés par plusieurs récepteurs placés sur la surface de la Terre. Ces mesures sont utilisées pour produire une image du sous-sol à partir d'un procédé d'inversion utilisant des méthodes numériques. Nous utilisons la méthode hp-FEM résultant d'une extension du travail de Demkowicz 2005. Nous avons développé un logiciel qui résout, pour la première fois, le problème MT avec des éléments finis auto-adaptatifs. La méthode hp-FEM permet des raffinements locaux, à la fois en taille h et en ordre p sur les éléments, ce qui est un avantage notoire puisque la combinaison de ces deux types de critères permet de mieux capter la présence de singularités, fournissant ainsi des erreurs de discrétisation faible. C'est donc une méthode très précise dont la convergence est exponentielle (Gui and Babuska 1986, Babuska and Guo 1996). En raison des défis d'implémentation encore non résolus (Demkowicz et al. 2002) et de la complexité technique des calculs hp-FEM en 3D, nous nous limitons, dans ce travail, à des calculs en 1D et 2D.Le domaine de calcul est tronqué par un matériau absorbant (Perfectly Matched Layer PML, Berenger 1994), qui est conçu pour s'adapter automatiquement aux propriétés physiques des matériaux. En particulier, il s'ajuste efficacement à l'interface air-sol, où le contraste entre la conductivité des matériaux atteint jusqu'à seize ordres de grandeur. Dans cette thèse, nous présentons également des résultats préliminaires pour la mise en place d'une technique dimensionnelle adaptative plus connue sous le nom de DAM (Dimensionally Adaptive Method (DAM)). Lorsque la distribution de la résistivité du sous-sol dépend de multiples variables spatiales, une analyse correcte de la dimensionnalité (Ledo 2005, Martí et al. 2009, Weaver and Agarwal 2000) rend parfois possible de considérer les différentes régions avec des dimensions spatiales différentes. Par exemple, il est parfois possible d’interpréter la distribution comme une formation unidimensionnelle plus quelques hétérogénéités en 2D (ou 3D). Basée sur cette interprétation, la DAM tire profit d’une telle situation. Ainsi, l'idée principale de cette méthode est d'effectuer l'adaptativité sur la dimension spatiale en commençant par un problème de faible dimension et en utilisant les résultats obtenus pour minimiser le coût des problèmes de dimension supérieure. Nous commençons l'inversion avec un modèle 1D. Les résultats de ce problème d'inversion 1D sont utilisés comme information a priori sur les modèles de dimension supérieure. Un avantage fondamental de cette approche est que nous pouvons utiliser les solutions des problèmes de dimension inférieure précédemment calculées comme composantes du terme de régularisation associé à un problème de dimension supérieure afin d'augmenter la robustesse de l'inversion. Cette thèse propose également une analyse numérique rigoureuse de divers aspects des problèmes MT. En particulier, nous avons: (a) étudié l'effet de la source, (b) effectué une analyse fréquentielle de sensibilité, (c) illustré l'augmentation du taux de convergence lorsque l'adaptativité hp est employée, (d) séparé les effets 1D et 2D dans la solution numérique et (e) exploré l'intérêt de considérer différentes variables pour effectuer l'inversion. / The magnetotelluric (MT) method is a passive exploration technique that aims at estimating the resistivity distribution of the Earth's subsurface, and therefore at providing an image of it. This process is divided into two different steps. The first one consists in recording the data. In a second step, recorded measurements are analyzed by employing numerical methods. This dissertation focuses in this second task. We provide a rigorous mathematical setting in the context of the Finite Element Method (FEM) that helps to understand the MT problem and its inversion process. In order to recover a map of the subsurface based on 2D MT measurements, we employ for the first time in Mts a multi-goal oriented self adaptive hp-Finite Element Method (FEM). We accurately solve both the full formulation as well as a secondary field formulation where the primary field is given by the solution of a 1D layered media. To truncate the computational domain, we design a Perfectly Matched Layer (PML) that automatically adapts to high-contrast material properties that appear within the subsurface and on the air-ground interface. For the inversion process, we develop a first step of a Dimensionally Adaptive Method (DAM) by considering the dimension of the problem as a variable in the inversion. Additionally, this dissertation supplies a rigorous numerical analysis for the forward and inverse problems. Regarding the forward modelization, we perform a frequency sensitivity analysis, we study the effect of the source, the convergence of the hp-adaptivity, or the effect of the PML in the computation of the electromagnetic fields and impedance. As far as the inversion is concerned, we study the impact of the selected variable for the inversion process, the different information that each mode provides,and the gains of the DAM approach.

Ce travail de thèse est consacré à l’étude de la structure électrique de la croûte terrestre dans deux régions du Vietnam, au Nord dans le delta du Fleuve Rouge et au Sud dans la région de la faille de la Rivière Saigon, et en Afrique de l’Ouest, où les données utilisées sont celles acquises durant l’Année Internationale de l’Electrojet Equatorial (AIEE). La technique utilisée est celle du sondage magnétotellurique. Deux des régions étudiées (le sud du Vietnam et l’Afrique de l’Ouest) sont situées au voisinage de l’équateur magnétique. Du fait de la présence de l’électrojet équatorial, l’hypothèse de l’onde plane, sur laquelle est basée la magnétotellurique, n’y est vérifiée que pour une gamme de périodes limitée. La modélisation de l’induction par un électrojet gaussien nous a permis de caractériser l’effet de source associé à l’électrojet équatorial, et de déterminer pour chacune des zones étudiées les gammes de périodes pour lesquelles cet effet de source est négligeable. Nous avons par ailleurs montré que l’effet de source associé à l’électrojet équatorial dépend des principales caractéristiques de la structure électrique du sous-sol, et de la distance entre le centre de l’électrojet équatorial et la position examinée. Dans le delta du fleuve Rouge, le profil magnétotellurique est orienté SO-NE. Long d’environ 32 km, il recoupe les failles majeures du système de failles du Fleuve Rouge. Les données magnétotelluriques ont été inversées pour obtenir une coupe transversale de résistivité électrique pour les trois premiers kilomètres. Cette coupe de résistivité électrique et une coupe de densité obtenue par modélisation des données gravimétriques acquises le long du même profil, ainsi que des informations déduites de forages effectués au voisinage du profil ont été utilisées pour proposer une coupe géologique de la région étudiée. Dans la région de la faille de la Rivière Saigon, le profil magnétotellurique, long d’environ 15 km, recoupe le trajet supposé de la faille. La structure électrique des premiers kilomètres de la croûte a été déterminée par inversion 2-D des données pour les gammes de périodes qui ne sont pas affectées par l’électrojet équatorial (< 0,7 s). La coupe transversale de résistivité électrique ainsi obtenue nous a permis de proposer une nouvelle localisation de la faille de la Rivière Saigon, à la limite entre un domaine résistant, le bloc de Da Lat au nord-est, et un domaine moins résistant, le bloc de Can Tho au sud-ouest. Cette nouvelle localisation est à environ 2,2 km au sud-ouest de la localisation généralement supposée pour cette faille. En Afrique de l’Ouest, nous avons montré que les données de jour sont affectées par l’électrojet équatorial pour toute la gamme de période étudiée (>120 s); ces données n’ont donc pas été utilisées pour déterminer la structure électrique de la croûte et du manteau supérieur. L’inversion 2-D des données de nuit, montre la présence d’une structure subverticale, relativement conductrice, qui sépare le Craton Ouest-africain très résistant des terrains situés sous le bassin du fleuve Sénégal. / This thesis is devoted to the study of the electrical structure of the Earth’s crust in two regions of Vietnam (the Red River Delta in the north and the region of the Saigon River fault in the south) and in West Africa. The data were those acquired during the International Year of the Equatorial Electrojet (IYEE). The technique used is the magnetotelluric (MT) sounding method. Two of the regions studied (southern Vietnam and West Africa) are located near the magnetic equator. Due to the presence of the equatorial electrojet in these regions, the assumption of plane wave, on which MT relies, is valid for a limited range of periods. Modeling of the induction by a Gaussian electrojet allowed us to characterize the source effect associated with the equatorial electrojet, and to determine for each studied region the period ranges for which the source effect is negligible. We have also shown that the source effect associated with the equatorial electrojet depends on the main features of the electrical structure of the basement, and the distance between the center of the equatorial electrojet and the location of the considered station. In the Red River Delta at the north, the MT profile is oriented SW-NE. It is about 32 km long, and goes across the major faults of the Red River fault system. MT data were applied 2-D inversion to obtain a electrical cross-section for the first three kilometers of the crust. This electrical cross-section, a density cross-section obtained by modeling of gravity data collected along the same profile, and information derived from holes drilled in the vicinity of the profile, both were used to construct a geological section of the study area. In the region of the Saigon River fault, the magnetotelluric profile is about 15 km long. It goes across the supposed location of the fault. The electrical structure of the first three kilometers of the crust was determined by 2-D inversion of data for periods which are not affected by the equatorial electrojet (< 0.7 s). The electrical cross-section obtained allowed us to propose a new location for the Saigon River fault; it corresponds to the boundary between a resistant domain, the block of Da Lat in the north-east, and a less resistant domain, the block of Can Tho in the south-west. This new location is about 2.2 km southwest of the location currently assumed for this fault. In West Africa, we have shown that electromagnetic impedances derived using day time data are affected by the equatorial electrojet for the whole range of period studied (> 120 s), so these data have not been used to determine the electrical structure of the crust and the upper mantle. The 2-D inversion of nighttime data shows the presence of a subvertical structure, relatively conductive, which separates the West African Craton which is very resistant from the geological structures beneath the Senegal River basin.

The towed streamer controlled source electromagnetic (CSEM) system collects data faster than the conventional static node-based CSEM system. However, the towed streamer CSEM is typically much noisier than the conventional static node-based CSEM. Identifying and quantifying various sources of noise is important for the development of future robust electromagnetic streamer system. This is the problem I address in this thesis. I achieve this in three parts. First, I examine the idea that the towed streamer suffers from noise induced by its motion through the Earth’s magnetic field according to Faraday’s law of induction. I derive expressions for the motionally-induced noise for the cases of a horizontal streamer parallel to the acquisition vessel’s path and a curved streamer caused by a constant cross-current. These expressions demonstrate that the motionally-induced noise is sensitive to the magnitude of the feather angle at the head and at the tail of the streamer, and to the vertical and lateral motion of the streamer. The key finding is that no motionally-induced noise is generated when the streamer is horizontal and moving in a constant magnetic field. By contrast, when the streamer shape is curved because of cross-currents, motionally-induced noise is generated if the velocity of the streamer varies over time. Second, I analyse and compare the noise recorded using the first generation of towed streamer with the noise recorded using a static ocean bottom cable (OBC) CSEM. I find out that within the frequency range of interest, 0.01–1 Hz the towed streamer noise is 20 dB greater (factor of 10) than the noise recorded with the OBC CSEM. I show also that the motion of the telluric cable between the pair of electrodes in the towed streamer is responsible for this difference in amplitude between the two systems. In the frequency ranges, 0.03–0.1 Hz and 0.03–0.2 Hz, the motionally-induced noise is shown to be uncorrelated across all channels. However, within the frequency band 0.1–0.3 Hz, the motionally-induced noise correlation gradually increases and becomes well correlated at about 0.2 Hz. This correlated noise could be caused by ocean swell from surface waves, water flowing around the streamer or cross-currents. Finally, to identify and quantify the contribution of several distinct sources of noise, and to describe the mechanisms generating each source of noise, I co-designed a prototype towed streamer CSEM. I carried out an experiment with the prototype streamer suspended 1 m below the water surface in the controlled environment of the Edinburgh wave tank located in King’s building campus (the University of Edinburgh). I then subjected the streamer to flow running at velocities of 0–1ms−1 along its length and to waves propagating in the same direction, at 45°, and perpendicular relative to the streamer direction.

Recent advances in computational power, as well as the hard work of a handful of brilliant scientists, have made Bayesian inversion of geophysical observations possible. This development is highly significant, as it permits the quantification of uncertainty, not only on the inverted model parameters, but also on related properties of interest. This dissertation focuses on the application of a particular kind of Bayesian inversion – trans-dimensional Markov chain Monte Carlo – to electromagnetic data, specifically airborne transient electromagnetic, magnetotelluric, and surface-towed controlled source electromagnetic data. In chapters 2-4, these data, both real and synthetic, are inverted for 1D models of subsurface electrical resistivity. In chapter 5, magnetotelluric data are inverted for 2D models of resistivity – the first time, to the best of my knowledge, that trans-dimensional Bayesian inversion of magnetotelluric data for 2D models has been achieved. In each instance, the uncertainty on bulk resistivity provided by the Bayesian inversion is used to estimate uncertainty on related subsurface properties, including pore fluid resistivity and salinity, porosity, melt fraction, melt volatile content, and bulk mantle volatile inventory. Chapter 1 introduces the topic of Bayesian inversion of electromagnetic data. Chapter 2 concerns trans-dimensional Bayesian inversion of airborne transient electromagnetic data. These data were collected above Taylor Glacier in the McMurdo Dry Valleys region of Antarctica in 2011, and were inverted using deterministic inverse methods to image a conductive channel beneath the glacier, interpreted as a package of brine-saturated sediments. The Bayesian inversion of these data confirms the existence of a conductive channel and provides quantitative uncertainties on the resistivity as a function of depth. These uncertainties are used in conjunction with Archie’s Law to estimate uncertainty on the resistivity of the pore fluids in the sediments. Additionally, the Kullback-Leibler divergence – a statistical measure of the dissimilarity of two distributions – is introduced as a measure of how much influence the observations have on the model parameters as a function of depth. The utility of Bayesian inversion in estimating the noise floor necessary to effectively resolve model structure is demonstrated. In chapter 3, a joint Bayesian framework for inverting electromagnetic data is introduced. A modified version of the algorithm utilized in chapter 2 is applied to jointly invert marine magnetotelluric and surface-towed controlled source electromagnetic data. These data were collected offshore New Jersey in 2015 to image a freshwater aquifer in the continental shelf. Deterministic inversions of this data clearly image a resistive body at depths consistent with low salinity from bore hole measurements collocated with the electromagnetic survey. The Bayesian inversion of this data set again confirms the existence of the resistive region while further providing uncertainty on the inverted resistivity with depth. In some instances, bimodality in the posterior distribution is found, demonstrating the importance of Bayesian inverse methods for fully exploring the model space. The uncertainty on bulk resistivity is used in conjunction with Archie’s Law and the porosity from bore hole measurements in a Monte Carlo framework to estimate uncertainty in the salinity of the pore water as a function of depth for three well locations. These estimates match well with measured salinities at these locations, validating the use of the Bayesian posterior in the context of a Monte Carlo framework to estimate uncertainty on related physical properties. In chapter 4, seafloor magnetotelluric data are again inverted for 1D models of subsurface resistivity, this time to image a conductive channel at the base of the lithosphere. The data are a subset of a deployment of 50 Broadband MT instruments on the seafloor above the Cocos plate offshore Nicaragua. Deterministic inversions of this data revealed a conductive structure at 45-70 km depth, beneath the Cocos plate. This earlier analysis concluded that melt was required at the lithosphere-asthenosphere boundary (LAB) to explain the inverted resistivity, but the deterministic inverse tools available at the time did not permit quantitative uncertainties – on the conductive anomaly itself, the requirement for partial melt, the degree of partial melt, or the degree of mantle hydration. Bayesian inversion of data from two magnetotelluric sites confirm that the conductor is indeed robust, and that melt is required by nearly 100% of the models that fit the data. Further, the resistivity uncertainty from the Bayesian inversion is used in conjunction with petrological modeling of partial melting in the mantle and an estimated probability distribution for temperature to place constraints on the degree of partial melt and mantle volatile (water and carbon) inventory over the depth range 45-63 km. This analysis concludes that large melt fractions and either high temperatures or a high degree of mantle hydration are likely needed to explain the resistivities produced by the Bayesian inversion, potentially explaining the mechanism for plate sliding that enables plate tectonics. Finally, chapter 5 introduces 2D trans-dimensional Bayesian inversion of magnetotelluric data, for the first time to my knowledge. A Gaussian Process-parametrized, trans-dimensional Markov chain Monte Carlo algorithm is used with MARE2DEM to invert synthetic data as well as field data from the Gemini data set from the Gulf of Mexico. For Bayesian inversion to be computationally feasible beyond inverting for 1D models, the cost of forward modeling must be reduced, as well as the number of model parameters that the algorithm must sample over. The first challenge is addressed through high performance computing. The forward modeling is performed on a cluster. In addition, we implement parallel tempering, where multiple Markov chains are run in parallel and swap models at each iteration, vastly increasing the rate at which the model space is explored and sampled. The curse of dimensionality is addressed by utilizing a Machine Learning technique known as a Gaussian Process to represent the model with far fewer parameters than required in a typical discrete finite difference or finite element representation of the subsurface. The Bayesian inversion of the Gemini data successfully recovers the model structure obtained by deterministic inversion of the same data, but additionally provides uncertainty on bulk resistivity. This thesis demonstrates the power and utility of Bayesian inversion to move beyond single estimates of subsurface resistivity. Not only does the work in this dissertation show that Bayesian inversion can provide uncertainty on inverted resistivity, it shows that these inverted uncertainties can be used to place quantitative constraints on parameters related to bulk resistivity. This is crucial to rendering the information obtained from inversion of electromagnetic data useful to disciplines far beyond electromagnetic geophysics.

Untersuchungen zur internen Architektur von großen Störungszonen beschränken sich üblicherweise auf die, an der Erdoberfläche aufgeschlossene, störungsbezogene Deformation. Eine Methode, die es ermöglicht, Informationen über die Tiefenfortsetzung einer Störung zu erhalten, ist die Abbildung der elektrischen Leitfähigkeit des Untergrundes.<br /> <br /> Die vorliegende Arbeit beschäftigt sich mit der kombinierten strukturgeologischen und magnetotellurischen Untersuchung eines Segmentes der 'West Fault'-Störung in den nordchilenischen Anden. Die West Fault ist ein Abschnitt des über 2000 km langen Präkordilleren-Störungssystem, welches im Zusammenhang mit der Subduktion vor der südamerikanischen Westküste entstanden ist. Die Aktivität dieses Störungssystems reichte vom Eozän bis in das Quartär. Der Verlauf der West Fault ist im Untersuchungsgebiet (22&#176;04'S, 68&#176;53'W) an der Oberfläche klar definiert und weist über viele zehner Kilometer eine konstante Streichrichtung auf. Die Aufschlussbedingungen und die Morphologie des Arbeitsgebietes sind ideal für kombinierte Untersuchungen der störungsbezogenen Deformation und der elektrischen Leitfähigkeit des Untergrundes mit Hilfe magnetotellurischer Experimente (MT) und der erdmagnetischen Tiefensondierung (GDS). Ziel der Untersuchungen war es, eine mögliche Korrelation der beiden Meßmethoden herauszuarbeiten, und die interne Störungsarchitektur der West Fault umfassend zu beschreiben.<br /> <br /> Die Interpretation von Sprödbruch-Strukturen (kleinmaßstäbliche Störungen sowie Störungsflächen mit/ohne Bewegungslineationen) im Untersuchungsgebiet weist auf überwiegend seitenverschiebende Deformation entlang von subvertikal orientierten Scherflächen hin. Dextrale und sinistrale Bewegungsrichtungen können innerhalb der Störungszone bestätigt werden, was auf Reaktivierungen des Störungssystems schliessen läßt. Die jüngsten Deformationen im Arbeitsgebiet haben dehnenden Charakter, wobei die kinematische Analyse eine unterschiedliche Orientierung der Extensionsrichtung beiderseits der Störung andeutet. Die Bruchflächendichte nimmt mit Annäherung an die Störung zu und zeichnet einen etwa 1000 m breiten Bereich erhöhter Deformationsintensität um die Störungsspur aus (damage zone). Im Zentrum dieser Zone weist das Gestein eine intensive Alteration und Brekzierung auf, die sich über eine Breite von etwa 400 m erstreckt. Kleine Störungen und Scherflächen in diesem zentralen Abschnitt der Störung fallen überwiegend steil nach Osten ein (70-80&#176;).<br /> <br /> Innerhalb desselben Arbeitsgebietes wurde ein 4 km langes MT/GDS Profil vermessen, welches senkrecht zum Streichen der West Fault verläuft. Für die zentralen 2 km dieses Hauptprofils beträgt der Abstand der Meßstationen jeweils 100 m. Ein weiteres Profil, bestehend aus 9 Stationen mit einem Abstand von 300 m zueinander, quert die Störung einige Kilometer entfernt vom eigentlichen Arbeitsgebiet. Die Aufzeichnung der Daten erfolgte mit vier S.P.A.M MkIII Apparaturen in einem Frequenzbereich von 1000 Hz bis 0.001 Hz.<br /> <br /> In den GDS Daten beider Profile ist die Störung für Frequenzen >1 Hz deutlich abgebildet: Die Induktionspfeile kennzeichnen eine mehrere hundert Meter breite Zone erhöhter Leitfähigkeit, welche sich entlang der West Fault erstreckt. Die Dimensionalitätsanalyse der MT Daten rechtfertigt die Anpassung der gemessenen Daten mit einem zwei-dimensionalen Modell für einen Frequenzbereich von 1000 Hz bis 0.1 Hz. In diesem Frequenzbereich, der eine Auflösung der Leitfähigkeitsstruktur bis mindestens 5 km Tiefe ermöglicht, läßt sich eine regionale geoelektrische Streichrichtung parallel zum Verlauf der West Fault nachweisen.<br /> <br /> Die Modellierung der MT Daten beruht auf einem Inversionsalgorithmus von Mackie et al. (1997). Leitfähigkeitsanomalien, die sich aus der Inversions-Modellierung ergeben, werden anhand von empirischen Sensitivitätsstudien auf ihre Robustheit überprüft. Dabei werden die Eigenschaften (Geometrie, Leitfähigkeit) der Strukturen systematisch variiert und sowohl Vorwärts- als auch Inversionsrechnungen der modifizierten Modelle durchgeführt. Die jeweiligen Modellergebnisse werden auf ihre Konsistenz mit dem Ausgangsdatensatz überprüft. Entlang beider MT Profile wird ein guter elektrischer Leiter im zentralen Abschnitt der West Fault aufgelöst, wobei die Bereiche erhöhter Leitfähigkeit östlich der Störungsspur liegen. Für das dicht vermessene MT Profil ergibt sich eine Breite des Störungsleiters von etwa 300 m sowie ein steiles Einfallen der Anomalie nach Osten (70&#176;). Der Störungsleiter reicht bis in eine Tiefe von mindestens 1100 m, während die Modellierungsstudien auf eine maximale Tiefenerstreckung <2000 m hinweisen. Das Profil zeigt weitere leitfähige Anomalien, deren Geometrie aber weniger genau aufgelöst ist.<br /> <br /> Die Störungsleiter der beiden MT Profile stimmen in ihrer Position mit der Alterationszone überein. Im zentralen Bereich des Hauptprofils korreliert darüber hinaus das Einfallen der Sprödbruch-Strukturen und der Leitfähigkeitsanomalie. Dies weist darauf hin, daß die Erhöhung der Leitfähigkeit im Zusammenhang mit einem Netzwerk von Bruchstrukturen steht, welches mögliche Wegsamkeiten für Fluide bietet. Der miteinander in Verbindung stehende Gesteins-Porenraum, der benötigt wird, um die gemessene Erhöhung der Leitfähigkeit durch Fluide im Gestein zu erklären, kann anhand der Salinität einiger Grundwasserproben abgeschätzt werden (Archies Gesetz). Wasserproben aus größerer Tiefe, weisen aufgrund intensiverer Fluid-Gesteins-Wechselwirkung eine höhere Salinität, und damit eine verbesserte Leitfähigkeit, auf. Für eine Probe aus einer Tiefe von 200 m ergibt sich demnach eine benötigte Porosität im Bereich von 0.8% - 4%. Dies legt nahe, daß Wässer, die von der Oberfläche in die Bruchzone der Störung eindringen, ausreichen, um die beobachtete Leitfähigkeitserhöhung zu erklären. Diese Deutung wird von der geochemischen Signatur von Gesteinsproben aus dem Alterationsbereich bestätigt, wonach die Alteration in einem Regime niedriger Temperatur (<95&#176;C) stattfand. Der Einfluß von aufsteigenden Tiefenwässern wurde hier nicht nachgewiesen. Die geringe Tiefenerstreckung des Störungsleiters geht wahrscheinlich auf Verheilungs- und Zementationsprozesse der Bruchstrukturen zurück, die aufgrund der Lösung und Fällung von Mineralen in größerer Tiefe, und damit bei erhöhter Temperatur, aktiv sind.<br /> <br /> Der Vergleich der Untersuchungsergebnisse der zur Zeit seismisch inaktiven West Fault mit veröffentlichten Studien zur elektrischen Leitfähigkeitsstruktur der aktiven San Andreas Störung, deutet darauf hin, daß die Tiefenerstreckung und die Leitfähigkeit von Störungsleitern eine Funktion der Störungsaktivität ist. Befindet sich eine Störung in einem Stadium der Deformation, so bleibt das Bruchnetzwerk für Fluide permeabel und verhindert die Versiegelung desselben. / The characterisation of the internal architecture of large-scale fault zones is usually restricted to the outcrop-based investigation of fault-related structural damage on the Earth's surface. A method to obtain information on the downward continuation of a fault is to image the subsurface electrical conductivity structure.<br /> <br /> This work deals with such a combined investigation of a segment of the West Fault, which itself is a part of the more than 2000 km long trench-linked Precordilleran Fault System in the northern Chilean Andes. Activity on the fault system lasted from Eocene to Quaternary times. In the working area (22&#176;04'S, 68&#176;53'W), the West Fault exhibits a clearly defined surface trace with a constant strike over many tens of kilometers. Outcrop condition and morphology of the study area allow ideally for a combination of structural geology investigation and magnetotelluric (MT) / geomagnetic depth sounding (GDS) experiments. The aim was to achieve an understanding of the correlation of the two methods and to obtain a comprehensive view of the West Fault's internal architecture.<br /> <br /> Fault-related brittle damage elements (minor faults and slip-surfaces with or without striation) record prevalent strike-slip deformation on subvertically oriented shear planes. Dextral and sinistral slip events occurred within the fault zone and indicate reactivation of the fault system. Youngest deformation increments mapped in the working area are extensional and the findings suggest a different orientation of the extension axes on either side of the fault. Damage element density increases with approach to the fault trace and marks an approximately 1000 m wide damage zone around the fault. A region of profound alteration and comminution of rocks, about 400 m wide, is centered in the damage zone. Damage elements in this central part are predominantly dipping steeply towards the east (70-80&#176;).<br /> <br /> Within the same study area, the electrical conductivity image of the subsurface was measured along a 4 km long MT/GDS profile. This main profile trends perpendicular to the West Fault trace. The MT stations of the central 2 km were 100 m apart from each other. A second profile with 300 m site spacing and 9 recording sites crosses the fault a few kilometers away from the main study area. Data were recorded in the frequency range from 1000 Hz to 0.001 Hz with four real time instruments S.P.A.M. MkIII.<br /> <br /> The GDS data reveal the fault zone for both profiles at frequencies above 1 Hz. Induction arrows indicate a zone of enhanced conductivity several hundred meters wide, that aligns along the WF strike and lies mainly on the eastern side of the surface trace. A dimensionality analysis of the MT data justifies a two dimensional model approximation of the data for the frequency range from 1000 Hz to 0.1 Hz. For this frequency range a regional geoelectric strike parallel to the West Fault trace could be recovered. The data subset allows for a resolution of the conductivity structure of the uppermost crust down to at least 5 km.<br /> <br /> Modelling of the MT data is based on an inversion algorithm developed by Mackie et al. (1997). The features of the resulting resistivity models are tested for their robustness using empirical sensitivity studies. This involves variation of the properties (geometry, conductivity) of the anomalies, the subsequent calculation of forward or constrained inversion models and check for consistency of the obtained model results with the data. A fault zone conductor is resolved on both MT profiles. The zones of enhanced conductivity are located to the east of the West Fault surface trace. On the dense MT profile, the conductive zone is confined to a width of about 300 m and the anomaly exhibits a steep dip towards the east (about 70&#176;). Modelling implies that the conductivity increase reaches to a depth of at least 1100 m and indicates a depth extent of less than 2000 m. Further conductive features are imaged but their geometry is less well constrained.<br /> <br /> The fault zone conductors of both MT profiles coincide in position with the alteration zone. For the dense profile, the dip of the conductive anomaly and the dip of the damage elements of the central part of the fault zone correlate. This suggests that the electrical conductivity enhancement is causally related to a mesh of minor faults and fractures, which is a likely pathway for fluids. The interconnected rock-porosity that is necessary to explain the observed conductivity enhancement by means of fluids is estimated on the basis of the salinity of several ground water samples (Archie's Law). The deeper the source of the water sample, the more saline it is due to longer exposure to fluid-rock interaction and the lower is the fluid's resistivity. A rock porosity in the range of 0.8% - 4% would be required at a depth of 200 m. That indicates that fluids penetrating the damaged fault zone from close to the surface are sufficient to explain the conductivity anomalies. This is as well supported by the preserved geochemical signature of rock samples in the alteration zone. Late stage alteration processes were active in a low temperature regime (<95&#176;C) and the involvement of ascending brines from greater depth is not indicated. The limited depth extent of the fault zone conductors is a likely result of sealing and cementation of the fault fracture mesh due to dissolution and precipitation of minerals at greater depth and increased temperature.<br /> <br /> Comparison of the results of the apparently inactive West Fault with published studies on the electrical conductivity structure of the currently active San Andreas Fault, suggests that the depth extent and conductivity of the fault zone conductor may be correlated to fault activity. Ongoing deformation will keep the fault/fracture mesh permeable for fluids and impede cementation and sealing of fluid pathways.

This thesis is a compendium of three papers related to the magnetotelluric (MT) method and its application to the characterization of lithospheric-scale geoelectrical structures. The thesis begins with an Introduction where the motivation and the aims are presented as well as the geological context of the Pyrenees, the previous geophysical studies and the used data set. The Introduction also contains a description of the MT method with basic concepts useful to understand the MT analysis developed in the three papers. Then the three papers, which are the main part of the thesis, are exposed. After the papers, there is a Discussion where the results are commented and compared, extending some of the ideas proposed in the articles and including a new geoelectrical crustal model of the Western Pyrenees. Then there are the Conclusions and proposals of future works related to the MT sites acquired in the Pyrenees and the proposed method for processing MT data. At the end of the thesis there are the references of the cited works and an annex containing a copy of the two published papers and a copy of the submitted manuscript of the third paper. The first published paper determines the geoelectrical structures below the Central Pyrenees using new BBMT and LMT data acquired close to the ECORS-Pyrenees deep seismic profile. The aim of this study has been to improve the results obtained by Ledo (1996) and Pous et al. (1995a), taking advantage of the progress occurred in the last fifteen years associated with the MT method and the acquisition of MT data. The most important improvements that have been achieved are the better definition of the top of the electrical resistivity structure associated with the Iberian subducted lower crust (IBSLC) and the determination of the lithosphere-asthenosphere boundary (LAB) below the Iberian and the European plates close to the collision zone. Complementary, an analysis of the geological properties of the lower-crust and the dehydration point of the muscovite and biotite rocks provides more arguments in favour of the hypothesis of partial melting to explain the geophysical observations associated with the IBSLC. In the second paper the MT method has been applied to characterize the geoelectrical structures below the West-Central Pyrenees in a new MT profile across the Pyrenean mountain range. This new MT profile is 130 km long and is composed by twenty BBMT sites acquired in 2010, seven of them also with LMT data. Results have been compared with the ones obtained in the Central Pyrenees characterizing the main geoelectrical variations between the two MT profiles. An analysis of the thermal re-equilibration of the IBSLC has been developed completing the analysis of the previous paper about the presence or absence of partial melting in the IBSLC. This analysis corroborates partial melting as the best explanation for the low electrical resistivity values of the IBSLC. The LAB below the West-Central Pyrenees has been determined not only below the Iberian and the European plates, as in the Central Pyrenees, but also below the collision zone. Additionally, an unexpected geoelectrical structure completely buried below the Ebro basin has been associated with a Variscan boundary. In the third paper a new method is proposed for processing MT data. The proposed method uses the horizontal magnetic fields recorded at a neighbouring site to process the MT data of a local site from inter-station tensor relationships. Results obtained in the experiments validate the method. A parameter has been defined in order to determine, in each specific study for each used period, if the proposed method can be applied. / Aquesta tesi es composa de tres articles relacionats amb el mètode Magnetotelluric (MT) i la seva aplicació a la caracterització d’estructures geoelèctriques de la litosfera. La tesi comença amb una introducció on es presenten les motivacions d’aquesta i els objectius plantejats, així com el context geològic dels Pirineus, els estudis previs que s’han realitzat i les dades fetes servir pel desenvolupament de la tesis. La introducció conté també una descripció del mètode MT amb l’objectiu d’ajudar a entendre l’anàlisi que es desenvolupa en els articles. Tot seguit s’exposen els tres articles, es discuteixen els resultats obtinguts i s’exposen les conclusions a les que s’ha arribat. El primer article mostra les estructures geoelèctriques sota el Pirineu Central. L’objectiu d’aquest estudi ha estat millorar els resultats de Ledo (1996) i Pous et al. (1995a), aplicant els progressos que hi ha hagut en el camp de la MT en els darrers quinze anys. Els nous resultats han permès caracteritzar millor la zona de fusió parcial associada amb la subducció de l’escorça Ibèrica Inferior (IBSLC) i el límit litosfera astenosfera (LAB). De forma complementaria, s’ha dut a terme un anàlisi de les propietats de l’escorça inferior que ha permès aportar més arguments a favor de la hipòtesis de la fusió parcial. En el segon article el mètode MT s’ha fet servir per caracteritzar les estructures geoelèctriques sota el Pirineu Occidental-Central en un nou perfil de MT. En aquest estudi també s’han observat estructures geoelèctriques associades amb la fusió parcial de la IBSLC i amb el LAB. Sota la conca de l’Ebre s’ha observat per primer cop una anomalia geoelèctrica que s’ha associat amb la presència d’un límit Hercinià. Els resultats han estat comparats amb els del Pirineu Central per tal de caracteritzar l’evolució de les principals estructures geoelèctriques al llarg de la serralada. En el tercer article s’ha proposat un nou mètode per processar dades MT. Aquest mètode utilitza els camps magnètics horitzontals enregistrats en una estació veïna per processar les dades de la zona d’estudi únicament a partir de relacions tensorials entre estacions. Els experiments duts a terme corroboren la validesa del mètode.

No description available.