Dissertations / Theses on the topic 'Inversion modelling'

In this dissertation an approach is presented for the three-dimensional electrical resistivity tomography (ERT) using unstructured discretizations. The geoelectrical forward problem is solved by the finite element method using tetrahedral meshes with linear and quadratic shape functions. Unstructured meshes are suitable for modelling domains of arbitrary geometry (e.g., complicated topography). Furthermore, the best trade-off between accuracy and numerical effort can be achieved due to the capability of problem-adapted mesh refinement. Unstructured discretizations also allow the consideration of spatial extended finite electrodes. Due to a corresponding extension of the forward operator using the complete electrode model, known from medical impedance tomography, a study about the influence of such electrodes to geoelectrical measurements is given. Based on the forward operator, the so-called triple-grid-technique is developed to solve the geoelectrical inverse problem. Due to unstructured discretization, the ERT can be applied by using a resolution dependent parametrization on arbitrarily shaped two-dimensional and three-dimensional domains. A~Gauss-Newton method is used with inexact line search to fit the data within error bounds. A global regularization scheme is applied using special smoothness constraints. Furthermore, an advanced regularization scheme for the ERT is presented based on unstructured meshes, which is able to include a-priori information into the inversion and significantly improves the resulting ERT images. Structural information such as material interfaces known from other geophysical techniques are incorporated as allowed sharp resistivity contrasts. Model weighting functions can define individually the allowed deviation of the final resistivity model from given start or reference values. As a consequent further development the region concept is presented where the parameter domain is subdivided into lithological or geological regions with individual inversion and regularization parameters. All used techniques and concepts are part of the open source C++ library GIMLi, which has been developed during this thesis as an advanced tool for the method-independent solution of the inverse problem.

The marine Controlled Source Electromagnetic (CSEM) method is an important and almost self-contained discipline in the toolkit of methods used by geophysicists for probing the earth. It has increasingly attracted attention from industry during the past decade due to its potential in detecting valuable natural resources such as oil and gas. A method for three-dimensional CSEM modelling in the frequency domain is presented. The electric field is decomposed in primary and secondary components, as this leads to a more stable solution near the source position. The primary field is computed using a resistivity model for which a closed form of solution exists, for example a homogeneous or layered resistivity model. The secondary electric field is computed by discretizing a second order partial differential equation for the electric field, also referred in the literature as the vector Helmholtz equation, using the edge finite element method. A range of methods for the solution of the linear system derived from the edge finite element discretization are investigated. The magnetic field is computed subsequently, from the solution for the electric field, using a local finite difference approximation of Faraday’s law and an interpolation method. Tests, that compare the solution obtained using the presented method with the solution computed using alternative codes for 1D and 3D synthetic models, show that the implemented approach is suitable for CSEM forward modelling and is an alternative to existing codes. An algorithm for 3D inversion of CSEM data in the frequency domain was developed and implemented. The inverse problem is solved using the L-BFGS method and is regularized with a smoothing constraint. The inversion algorithm uses the presented forward modelling scheme for the computation of the field responses and the adjoint field for the computation of the gradient of the misfit function. The presented algorithm was tested for a synthetic example, showing that it is capable of reconstructing a resistivity model which fits the synthetic data and is close to the original resistivity model in the least-squares sense. Inversion of CSEM data is known to lead to images with low spatial resolution. It is well known that integration with complementary data sets mitigates this problem. It is presented an algorithm for the integration of an acoustic velocity model, which is known a priori, in the inversion scheme. The algorithm was tested in a synthetic example and the results demonstrate that the presented methodology is promising for the improvement of resistivity models obtained from CSEM data.

La surface terrestre est affectée par de nombreux processus locaux tels que des événements volcaniques, des glissements de terrain ou des tremblements de terre. Parallèlement à ces processus naturels, les activités anthropiques, y compris l’extraction et le stockage des ressources profondes (par exemple, les minéraux ou les hydrocarbures) façonnent la Terre à différentes échelles spatiales et temporelles. Ces mécanismes produisent une déformation du sol qui peut être détectée par divers instruments et techniques géodésiques tel que le GNSS, l’InSAR, les inclinomètres. Le but de cette thèse est de développer un outil numérique permettant l’inversion conjointe de multiples données géodésiques associées à la déformation de la plaque ou au changement de contrainte volumique en profondeur. Quatre types d’applications sont ciblés: la déformation intersismiques des plaques, la déformation des volcans, l’exploitation minière profonde et l’extraction de pétrole et de gaz. Différentes complexités du modèle inverse ont été considérées: le niveau I considère un seul type de données géodésiques avec un processus indépendant du temps. Une application est réalisée avec l’inversion des données GPS à travers le sud de la Californie pour déterminer les variations latérales de la rigidité lithosphérique (Furst et al., 2017). Le niveau II représente également un seul type de données géodésiques mais avec un processus dépendant du temps. La détermination conjointe de l’historique des changements de contrainte et des paramètres de dérive d’un réseau d’inclinomètres est étudiée à l’aide d’un exemple synthétique (Furst et al., soumis). Le niveau III considère différents types de données géodésiques et un processus dépendant du temps. Un réseau fictif combinant des données GNSS, InSAR, inclinométriques et de nivellement est défini pour calculer le changement de volume dépendant du temps d’une source profonde de déformation. Une méthodologie pour implémenter ces différents niveaux de complexité est développée dans un seul logiciel. Parce que le problème inverse peut être mal posé, la minimisation de la fonctionnelle peut produire plusieurs minima. Par conséquent, un algorithme d’optimisation global est utilisé (Mohammadi and Saïac, 2003). Le problème direct est traité en utilisant un ensemble de modèles élastiques numériques et analytiques permettant de modéliser les processus de déformation en profondeur. Grâce à ces développements numériques, des avancées concernant les problèmes inverses en géodésie devraient être possibles telle que l’inversion jointe de différents types de données géodésiques acquises lors de la surveillance des volcans. Dans cette perspective, la possibilité de déterminer par inversion les paramètres de dérive des inclinomètres permettrait une détermination précise des sources de déformation profondes. En outre, la méthodologie développée peut être utilisée pour une surveillance précise de la déformation des réservoirs de pétrole et de gaz
The Earth’s surface is affected by numerous local processes like volcanic events, landslides or earthquakes. Along with these natural processes, anthropogenic activities including extraction and storage of deep resources (e.g. minerals, hydrocarbons) shape the Earth at different space and time scales. These mechanisms produce ground deformation that can be detected by various geodetic instruments like GNSS, InSAR, tiltmeters, for example. The purpose of the thesis is to develop a numerical tool to provide the joint inversion of multiple geodetic data associated to plate deformation or volume strain change at depth. Four kinds of applications are targeted: interseismic plate deformation, volcano deformation, deep mining, and oil & gas extraction. Different inverse model complexities were considered: the I-level considers a single type of geodetic data with a time independent process. An application is made with inverting GPS data across southern California to determine the lateral variations of lithospheric rigidity (Furst et al., 2017). The II-level also accounts for a single type of geodetic data but with a time-dependent process. The joint determination of strain change history and the drift parameters of a tiltmeter network is studied through a synthetic example (Furst et al., submitted). The III-level considers different types of geodetic data and a timedependent process. A fictitious network made by GNSS, InSAR, tiltmeters and levelling surveys is defined to compute the time dependent volume change of a deep source of strain. We develop a methodology to implement these different levels of complexity in a single software. Because the inverse problem is possibly ill-posed, the functional to minimize may display several minima. Therefore, a global optimization algorithm is used (Mohammadi and Saïac, 2003). The forward part of the problem is treated by using a collection of numerical and analytical elastic models allowing to model the deformation processes at depth. Thanks to these numerical developments, new advances for inverse geodetic problems should be possible like the joint inversion of various types of geodetic data acquired for volcano monitoring. In this perspective, the possibility to determine by inverse problem the tiltmeter drift parameters should allow for a precise determination of deep strain sources. Also, the developed methodology can be used for an accurate monitoring of oil & gas reservoir deformation

In Australia, as in many countries, there is an increasing requirement for spatial and temporal information on forest structure and particularly biomass (carbon). The use of remote sensing data is well suited for the provision of such data, in particular Synthetic Aperture Radar (SAR) data. However, the majority of studies have concentrated on empirical relationships between above ground biomass (AGB) and radar backscatter. Limitations of such approaches include saturation of the signal (e.g., above certain biomass levels), structural variations, and the influence of environmental conditions (e.g., surface moisture). Recognising these limitations, this study focused on the retrieval of parameters through inversion of a physically based backscatter model (that of Durden et al., 1989) using a non-linear estimation algorithm (Moghaddam and Saatchi, 1999). The study focused on the Brigalow Belt Bioregion of Queensland, Australia, and particularly regrowth forests dominated by Acacia harpophylla. A generalised method of parameterisation has been described and validated using relationships established through fieldwork for three Australian species, used to develop parametric scattering equations for use in a non-linear estimation algorithm for the retrieval of structural and dielectric parameters from AIRSAR and ALOS PALSAR data with Landsat derived Foliage Projective Cover (FPC). Four case studies were presented in which the sensitivity of the estimation algorithm to noise was tested for a variety of scenarios. Following sensitivity analysis the algorithm was applied to real data for the product of structural maps. Although currently only modest levels of accuracy have been obtained with the approach there is much scope for improvement (i.e., better backscatter modelling, more data channels) and enhancement (i.e., inclusion of LiDAR data) of the algorithm.

Northern Sweden is one of the largest hosts for mineral resources in Europe and always has been an interesting area for researchers from various disciplines of Earth sciences. This dissertation is a comprehensive summary of three case study papers on airborne VLF, gravity gradient and magnetic data in the area. In the first paper, tensor VLF data is extracted from an old data set which contains only the total and the vertical magnetic components. The anomalous part of the horizontal magnetic field components is computed by a Hilbert transform of the vertical magnetic field. The normal part of the horizontal magnetic field component is computed as a function of total, vertical and anomalous part of horizontal magnetic fields. The electric field is also calculated for TE mode and impedance tensor and apparent resistivity are computed. In addition tippers are calculated for two transmitters and inverted by a 3D inversion algorithm. Comparison of the estimated model and geology map of bedrock shows that lower resistivity zones are correlated with mineralizations. The second paper deals with the internal consistency of airborne gravity gradient data. The six components of the data are estimated from a common potential function. It is shown that the data is adequately consistent but at shorter land clearances the difference between the estimated data and the original data is larger. The technique is also used for computing the Bouguer anomaly from terrain corrected FTG data. Finally the data is inverted in 3D, which shows that the estimated density model in shallow depth is dominated by short wave length features. Inversion of TMI data is the topic of the third paper where a new type of reference model for 3D inversion of magnetic data is proposed by vertically extending the estimated magnetization of a 2D terrain magnetization model. The final estimated 3D result is compared with the magnetization model where no reference model is used. The comparison shows that using the reference model helps the high magnetization zones in the estimated model at shallow depths to be better correlated with measured high remanent magnetization from rock samples. The high magnetization zones are also correlated with gabbros and volcanic metasediments.

The aim of my work is to estimate viscoelastic parameters of rock samples from waveforms of ultrasonic waves propagating through these samples. To this end, I develop an automated Python modules in Finite Element Modelling software Abaqus, and tailored it specifically for a controlled transmission experiment using ultrasonic source and receiver. The approach is verified using test Aluminium samples, and then applied to real rocks to estimate ultrasonic attenuation using Prony formulation of viscoelasticity.

The main aim of this thesis is to gain new knowledge about the lithospherical structure and tectonics of the Carpathian-Pannonian Basin region. We applied three different methods: 1Dautomatic modelling, 2D integrated geophysical modelling and 3D inversion to achieve this goal.These methods are similar concerning the used databases but differ by used processing andinterpretation. At first we apply 1D automatic modelling to get a very first overview of thestudied region. Secondly, we apply 2D integrated modelling of the lithosphere which combines the interpretation of surface heat flow, geoid, gravity, and topography data in the Carpathian-Pannonian Basin region and surrounding areas. This approach is able to constrain the complicated lithospheric structures of the studied region better than interpreting each data set onits own. We present four 2D integrated models of the lithosphere in the Carpathian-PannonianBasin region and surrounding areas. Finally, based on the 3D Inversion algorithm, we present the geophysical models of the lithosphere in the Carpathian-Pannonian region. The algorithm returns the density structure of the lithosphere from joint inversion of free air gravity, geoid andtopography data based on a Bayesian approach. The models are based on different input data sets and constrained by different a priori data. Based on our modelling we cannot confirm theextreme thinning (less than 70 km) of the Pannonian Basin lithosphere proposed by other authors.On the other hand, the results show the increasing trend of the lithospherical thickness of theCarpathian Arc from the Western Carpathians toward the Eastern Carpathians which confirms theprevious theories about the propagation of subduction process. We got some controversial resultsin the area of the Southern Carpathians. The results based on 3D inversion show extremely thinlithosphere in the area; on the other hand, the results based on 2D integrated modelling do notsupport such thinning. However both methods indicate that it is probable that the MoesianPlatform is bend and underthrusted underneath the Southern Carpathians. The south-eastern edge of the Pannonian Basin based on 3D inversion shows unexpected and surprisingly thinlithosphere. Since the area is quite large, we could exclude an effect of flexure, therefore this area might be potentially interesting for further investigation.

Seismic surveys are well established in hydrocarbon prospecting, and technology for processing seismic data has been developed through decades. The first electromagnetic survey for hydrocarbon prospecting however, was performed in 2000, and as a consequence of the short time span the technology is not as well developed as in seismics. For instance, the need for efficient and robust forward modelling software and inversion schemes for collected data is urgent. In this thesis forward modelling using integral equations and the Contrast Source Inversion (CSI) method is investigated for forward and inverse 3D electromagnetic scattering experiments in hydrocarbon prospecting, respectively. The mathematical model is developed in an arbitrary isotropic, conductive medium, with contrast in electric permittivity and electric conductivity between the scattering object and the background, while in the numerical examples the background model is restricted to a horizontally layered medium with variations in the $z$-direction only and contrast in electric conductivity. The difference in electric conductivity is considered the backbone of electromagnetic hydrocarbon prospecting. The main result concerning forward modelling in this thesis is the establishment of a, to my knowledge, previously unpublished method for solving the electric Lippmann-Schwinger equation in a conductive medium by fixed point iteration. In the inversion part of this thesis the previously scalar CSI method is extended to a full vector valued method. A new CSI method for inversion with respect to all the electromagnetic parameters (the electric permittivity, electric conductivity and magnetic permeability) is also presented, which I have yet to find treated elsewhere. Only the former method is tested numerically, using synthetic data, due to the computational complexity of the latter. The numerical results from the forward modelling show the numerical validity of integral equation modelling and the fixed point iteration, whereas the results from the inversion show some promise. With several source and receiver lines present the lateral position of the scattering object is reconstructed well, whereas the vertical position causes problems. When textit{a priori} information about the position of the scattering object is introduced to further regularise the problem, the approximate position of it is successfully inverted, which illustrates the essential part additional regularisation plays in this inverse scattering problem. Thus the CSI method could be useful in petroleum geophysics, and should be developed further for the purpose of locating hydrocarbons in the subsurface. Several possibilities for further work is noted. This work was performed for StatoilHydro ASA.

The self-potential method responds to the electrokinetic phenomenon of streaming potential and has been applied in hydrogeologic and engineering investigations to aid in the evaluation of subsurface hydraulic conditions. Of specific interest is the application of the method to embankment dam seepage monitoring and detection. This demands a quantitative interpretation of seepage conditions from the geophysical data. To enable the study of variably saturated flow problems of complicated geometry, a three-dimensional finite volume algorithm is developed to evaluate the self-potential distribution resulting from subsurface fluid flow. The algorithm explicitly calculates the distribution of streaming current sources and solves for the self-potential given a model of hydraulic head and prescribed distributions of the streaming current cross-coupling conductivity and electrical resistivity. A new laboratory apparatus is developed to measure the streaming potential coupling coefficient and resistivity in unconsolidated soil samples. Measuring both of these parameters on the same sample under the same conditions enables us to properly characterize the streaming current cross-coupling conductivity coefficient. I present the results of a laboratory investigation to study the influence of soil and fluid parameters on the cross-coupling coefficient, and characterize this property for representative well-graded embankment soils. The streaming potential signals associated with preferential seepage through the core of a synthetic embankment dam model are studied using the forward modelling algorithm and measured electrical properties to assess the sensitivity of the self-potential method in detecting internal erosion. Maximum self-potential anomalies are shown to be linked to large localized hydraulic gradients that develop in response to piping, prior to any detectable increase in seepage flow through the dam. A linear inversion algorithm is developed to evaluate the three-dimensional distribution of hydraulic head from self-potential data, given a known distribution of the cross-coupling coefficient and electrical resistivity. The inverse problem is solved by minimizing an objective function, which consists of a data misfit that accounts for measurement error and a model objective function that incorporates a priori information. The algorithm is suitable for saturated flow problems or where the position of the phreatic surface is known.

Les glaces de l'Antarctique constituent un matériau d'archivage unique pour garder la trace de l'atmosphère du passé. La plus ancienne archive glaciaire disponible (800 ka) ne permet malheureusement pas d'étudier une transition climatique majeure qui s'est déroulé entre 1.2 Ma et 0.9 Ma (transition du mi-Pleistocène). Localiser un site ayant permis la conservation d'une glace aussi ancienne est donc un enjeu majeur au sein de la communauté des carottes de glace. Plusieurs régions autour du Dôme C présentent de bonnes caractéristiques d'un point de vue thermique et mécanique. Le but de ces travaux est d'identifier précisément des sites propices à un archivage de long terme, en utilisant différents modèles successifs. D'abord un modèle thermique 1D, qui permet de calculer le flux géothermique local, et d'affirmer que les hauteurs du socle sont exemptes de fusion depuis au moins 800 ka. Ensuite un modèle 3D d'écoulement, qui permet la description des trajectoires des particules de glace depuis la surface jusqu'au socle. Pour minimiser les perturbations stratigraphiques, il conviendra de placer le site de forage proche de la crête, et pour assurer une résolution suffisante le site ne devra pas se trouver à l'aplomb du sommet du relief basal, mais sur son flanc. Enfin, on utilise un modèle de datation 1D contraints par les isochrones, pour connaître la distribution des âges basaux et résolutions compatibles avec les isochrones observées, et estimer les risques de succès ou d'échec du forage. On propose deux sites de forage distants de 37 km de Dôme C, pour lesquels la présence de glace exploitable de 1.5 Ma et d'une résolution suffisante est hautement probable
Antarctic ice is an exceptional archive of ancient atmospheric air. The oldest available ice archive (800 ka) is not old enough to study a main climatic transition that occurred between 1.2 Ma and 0.9 Ma (mid-Pleistocene transition). Locating a future old-ice drill site is now one of the main goal of the ice-core community. In the region of Dome C, several old-ice targets have required characteristics, from a thermal and mechanical point of view, to hosts very old ice. The goal of this work is to precisely locate suitable coring sites, using several glaciological models. First a 1D heat model to compute the local geothermal flux, that make us confident in the fact that bedrock heights have been free of basal melting for at least the last 800 ka. It is very likely that such sites are able to host a 1.5 Ma ice archive. Second, a 3D ice flow model, to describe the ice trajectories from the ice surface to the bed. Drilling less than 10 km from the surface ridge mimimizes the risks of stratigraphic disturbance, and drilling on the flanks of the bed relief - and not on its top - is mandatory to ensure a sufficient age resolution. Third, a 1D dating model constrained by isochrones let us describe the distribution of basal ages and age resolutions compatible with observed isochrones, and assess the risk of success or failure of the drilling. With currently available data and modelling tools, two drilling sites are proposed that are located 37 km from Dome C, at which 1.5 Ma-old ice is very likely to be archived in required conditions and with sufficient resolution

Les procédés de traitement par ultrafiltration sont certes largement utilisés pour la production d’eau potable mais leurs performances environnementales restent néanmoins mal connues. Cetravail vise à une meilleure compréhension de l’empreinte environnementale des membranes, ainsi qu’à la proposition et mise en oeuvre de solutions innovantes et pertinentes afin de l’atténuer. Une approche multi-étapes d'écoconception est appliquée aux fibres creuses fabriquées par inversion de phase et utilisées en filtration frontale. Deux modèles paramétrés ontété développés pour évaluer les impacts environnementaux lors des étapes de fabrication et d'utilisation en fonction des conditions opératoires. L’analyse de cycle de vie a mis en évidenceque les conditions opératoires liées au glycérol étaient des leviers d’action intéressants pour l’étape de fabrication. La quasi-exclusivité des impacts environnementaux est toutefois liée à l’étape d’utilisation, pour laquelle la production d’électricité et la fabrication d’hypochlorite de sodium sont les deux principaux contributeurs. L'analyse montre également l'influence du flux de filtration. Des opportunités existent concernant l’utilisation de solvants biosourcés, d’où la stratégie d’amélioration qui consiste à remplacer les solvants d’origine pétrochimique lors de la fabrication membranaire. Un cadre méthodologique basé sur des indicateurs a été proposé afin de rationaliser la fabrication de membrane durable. Des membranes planes biosourcées ont ainsi été obtenues en utilisant le lactate de méthyle comme solvant. Les résultats environnementaux mitigés n'excluent pas pour autant le remplacement de solvant en tant que stratégied'amélioration. Les résultats soulignent la faisabilité et la pertinence de cette approche d'écoconception basée sur la modélisation de procédé et le travail expérimental, et posent lesbases des améliorations futures
Although ultrafiltration treatment processes are widely used for drinking water applications, their environmental performances remain poorly understood. The objectives of this work are to provide understanding of membrane environmental footprint, as well as to suggest and implement relevant innovative solutions for its mitigation. A multiphase ecodesign approach is applied to hollow fibres fabricated by phase inversion and operated in a dead-end mode. The two developed parameterized models evaluate environmental impacts during fabrication and operation as a function of operating conditions. The subsequent Life Cycle Assessment highlights glycerol-related operating conditions as interesting action levers for environmental mitigation of membrane fabrication. However, membrane operation accounts for nearly exclusively all environmental impacts, with electricity production and sodium hypochlorite fabrication as the two main contributors. The analysis also pinpoints the influence of filtration flux. Since opportunities for biosourced solvents exist, substitution of petrochemical solvents during membrane fabrication is chosen as an improvement strategy. A metrics-based methodological framework is suggested to rationalize sustainable membrane fabrication. Biosourced flat membranes using methyl lactate as a solvent have thus been prepared. The mixed environmental scores do not preclude solvent substitution as a relevant improvement strategy. Results highlight the feasibility and relevance of such an ecodesign approach based on process modelling and experimental work, and lay the foundations for further development

La problématique de l'imagerie radar pose comme objectif de caractériser la cible observée. Dans le contexte de la télédétection appliquée aux milieux forestiers, il s'agit d'extraire les paramètres physiques constitutifs de la végétation.Ce manuscrit traite donc de la problématique de l'inversion de données électromagnétiques pour les basses fréquences radar, de la bande VHF à la bande P. Dans ces bandes de fréquences la canopée (feuilles et petites branches) n'interagit quasiment pas avec l'onde radar et la forêt peut donc, en première approximation, être représentée par des cylindres verticaux (troncs) et inclinés (branches primaires).Dans ce contexte, les méthodes d'inversion numériques s'appuient sur des modèles qui calculent, à de multiples reprises, les champs rayonnés par ce type de diffuseurs. Il faut alors choisir un modèle physique dont le compromis rapidité/précision penche en faveur de la rapidité, donc un modèle "approché". La précision du modèle choisi est ensuite étudié puis validé par comparaison à un modèle dont le compromis rapidité/précision est inversé, donc un modèle "exact". Parmi les nombreuses méthodes d'inversion, celle de l'algorithme génétique (AG) a été choisie. Cet algorithme bien connu est ici analysé puis amélioré pour répondre aux exigences du problème électromagnétique posé. Il est ensuite validé sur des données simulées et mesurées. Cette validation est aussi l'occasion d'étudier l'impact du jeu de données utilisé sur la rapidité et la précision de l'inversion, en fonction des configurations d'antenne (fréquences, positions, polarisations)
The objective of radar imaging is to characterize the observed target. In the case of forest remote sensing, the objective is to extract the component of the vegetation physical parameters.This manuscript addresses the problem of the inversion of electromagnetic data for low frequency wave, VHF band to the P band. In these frequency bands, the canopy (leaves and small branches) does not almost interact with the radar wave and the forest can thus, in first approximation, be represented by vertical cylinders (trunks) and tilted (primary branches).In this context, the numerical electromagnetic methods of inversion lean on models which calculate, many times, the scattered fields radiated by such broadcasters. It is then necessary to choose a physical model which compromise speed/accuracy tilts in favor of the speed, thus a "approached" model. The exactness of the chosen model is then studied and then validated by comparison to a model which the compromise speed/accuracy is inverted, thus a "exact" model. Among the many numerical methods of inversion, that of the genetic algorithm (AG) was chosen. This algorithm, well known, is analyzed here then improved to meet the requirements of the composed electromagnetic problem. It is then validated on simulated and measured data. This validation is also an opportunity to study the impact of the set of data used on the speed and the precision of the inversion, according to the configurations of antenna (frequencies, positions, polarizations)

Bien que les modèles hydrologiques soient aujourd'hui utilisés pour la prévision des crues, leur mise en place sur un territoire particulier peut encore poser question. Classiquement, elle nécessite des chronologies observées de pluies et de débits. Cependant, ces données peuvent s'avérer incertaines, insuffisantes voire indisponibles. Cela impact alors l'identification des modèles ainsi que leurs prévisions en les rendant elles-mêmes incertaines et potentiellement inopérantes.Dans cette thèse, nous avons cherché à revisiter ces problématiques liées à la mise en place des modèles, en proposant des approches innovantes étalonnant les modèles sans certaines données pouvant s'avérer trop incertaines ou indisponibles. Ces dernières, habituellement exigées pour l'identification des modèles, sont de plus estimées au cours de l'étalonnage. Deux méthodes applicables quelque-soit la nature du modèle considéré sont proposées. La première est une méthode d'inversion des pluies et des paramètres. Elle permet d'étalonner un modèle hydrologique à partir de chronologies observées de débits et de cumuls de pluies seulement. Cette méthode identifie de plus simultanément les chronologies des pluies au pas de temps horaire. Une application particulière de cette méthode à des modèles déjà établis généralise par ailleurs à des modèles non-analytiques l'approche d'inversion des pluies proposée par Kirchner (2009). La seconde est une méthode de calibration pluie-hauteur d'eau. À partir de chronologies observées de pluies et de hauteurs d'eau seulement, elle permet de mettre en place un modèle hydrologique et d'identifier une courbe de tarage. Outre les analyses originales pouvant être menées sur les chronologies de pluies ou les courbes de tarage estimées par ces méthodes, ces dernières étendent à des contextes plus restreints en données, la possibilité d'identification de modèles pluie-débits et offrent des applications en opérationnel.En parallèle du développement de ces méthodes, cette thèse nous a aussi amené à construire une représentation conceptuelle appelée espace des connaissances. Celle-ci permet d'appréhender de manière commune les questions nouvelles ayant été étudiées ainsi que des approches plus classiques comme la calibration ou la simulation
Rainfall-runoff models are used for flood forecasting by warning authorities. However their implementation on a particular territory is still a challenge. Indeed, they generally need to be calibrated by using observed rainfall and discharge time series. These data may be subject to errors and uncertainties. They are not always available. Then, the model calibration relevancy may be affected and the forecasts may also be subject to significant uncertainties. This research would like to address such issues related to the rainfall-runoff models calibration, by proposing original methods which may set up a model by using less data than the ``classical'' calibration. The unused data might be either subject to uncertainties or not available. Moreover, these unused data may be estimated by the methods. Two model independant approach were suggested. Both are an heuristic inversion algorithm of rainfall-runoff models. The first method estimates simultaneously hourly rainfall time series and models parameters, by using only observed hourly discharge time series and total areal rainfall of flood events. A specific application of this method to set up models (with fixed parameters), generalises to models which are not invertible analyticaly, the ``hydrology backward'' approach proposed by Kirchner (2009).The second method estimates simultaneously models parameters and a rating-curve, by using only observed hourly rainfall and stage time series. Original analysis may be performed on the rainfall time series and the rating-curve estimated by the methods. Also, they extend the applicability rainfall-runoff models to hydrological context with restricted available data and offer promising operational applications. Yet, this research lead us to build a conceptual framework, denoted knowledge space. This framework unifies not only the original approaches which were proposed, but also some more ``classical'' approaches to hydrology as the calibration and the simulation

Master of Science
Department of Geology
Abdelmoneam Raef
Matthew W. Totten
This work focuses on the use of advanced seismically driven technologies to estimate the distribution of key reservoir properties which mainly includes porosity and hydrocarbon reservoir pay. These reservoir properties were estimated by using a multitude of seismic attributes derived from post-stack high resolution inversions, spectral imaging and volumetric curvature. A pay model of the reservoir in the Wierman field in Ness County, Kansas is proposed. The proposed geological model is validated based on comparison with findings of one blind well. The model will be useful in determining future drilling prospects, which should improve the drilling success over previous efforts, which resulted in only few of the 14 wells in the area being productive. The rock properties that were modeled were porosity and Gamma ray. Water saturation and permeability were considered, but the data needed were not available. Sequential geological modeling approach uses multiple seismic attributes as a building block to estimate in a sequential manner dependent petrophysical properties such as gamma ray, and porosity. The sequential modelling first determines the reservoir property that has the ability to be the primary property controlling most of the other subsequent reservoir properties. In this study, the gamma ray was chosen as the primary reservoir property. Hence, the first geologic model built using neural networks was a volume of gamma ray constrained by all the available seismic attributes. The geological modeling included post-stack seismic data and the five wells with available well logs. The post-stack seismic data was enhanced by spectral whitening to gain as much resolution as possible. Volumetric curvature was then calculated to determine where major faults were located. Several inversions for acoustic impedance were then applied to the post-stack seismic data to gain as much information as possible about the acoustic impedance. Spectral attributes were also extracted from the post-stack seismic data. After the most appropriate gamma ray and porosity models were chosen, pay zone maps were constructed, which were based on the overlap of a certain range of gamma ray values with a certain range of porosity values. These pay zone maps coupled with the porosity and gamma ray models explain the performance of previously drilled wells.

Au cours de ce travail, nous avons étudié deux types de défauts interfaciaux: domaines d’inversion (DI) et joints de grains (JG) dans des semiconducteurs de structure wurtzite (nitrures- d’éléments III, ZnO et l’hétérostructure ZnO/GaN) en utilisant le MET haute résolution et la modélisation ab initio. Dans le cas des DI, nos analyses théoriques montrent qu'une configuration tête-à-tête avec une séquence d'empilement à l’interface AaBbAa-AcCaA (H4) est la structure la plus stable dans les composés binaires (nitrures et ZnO wurtzites). De plus, un gaz d’électrons (2DEG) ou de trous (2DHG) à 2 dimensions est formé pour les configurations « tête-à-tête » ou queue-à-queue. A l’interface ZnO/GaN, l'observation de MET très haute résolution a confirmé la configuration H4 avec une interface -Zn-O-Ga-N. Notre modélisation théorique a mis en évidence la formation d’un gas de trous à 2 dimensions à cette hétérointerface. Nous avons aussi réalisé l’étude topologique, théorique et par MET des joints de grains de rotation autour de l’axe [0001] dans ces matériaux. Dans le GaN, nous avons trouvé que les plans du joint sont simplement formés par des dislocations de type a déjà connues pour le matériau en couche mince. Par contre, dans ZnO, la théorie topologique est complétement démontrée, et la dislocation [101 ̅0] est une brique de base dans la constitution des joints de grains avec des cycles d’atomes 6-8-4-
In this work, we investigated two kinds of interfacial defects: inversion domain boundaries (IDBs) and grain boundaries (GB) in wurtzite semiconductors (III-nitrides, ZnO and ZnO/GaN heterostructure) using high-resolution TEM and first-principle calculations. For IDBs, theoretical calculation indicated that a head-to-head IDB with an interfacial stacking sequence of AaBbAa-AcCaA (H4) is the most stable structure in wurtzite compounds. Moreover, 2-dimensional electron gas (2DEG) and 2-dimensional hole gas (2DHG) build up in head-to-head and tail-to-tail IDBs, respectively. Considering the IDB at the ZnO/GaN heterointerface, TEM observations unveiled the H4 configuration with a -Zn-O-Ga-N interface. Moreover the theoretical investigation also confirmed stability of this interface along with the corresponding formation of a 2DHG. A detailed topological, TEM and theoretical investigation of [0001] tilt Grain Boundaries (GBs) in wurtzite symmetry has also been carried out. In GaN, it is shown that the GBs are only made of separated a edge dislocations with 4, 5/7 and 8 atoms rings. For ZnO, a new structural unit: the [101 ̅0] edge dislocation made of connected 6-8-4-atom rings is reported for the first time, in agreement with an early theoretical report on dislocations and jogs in the wurtzite symmetry

The Blötberget mining area, the focus of this MSc project, is located about 230 km northwest ofStockholm and 12 km southwest of the city of Ludvika (central Sweden). The mining area has beenknown since 1600 for its various types of mineralization particularly iron-oxide deposits (magnetite andhematite) with the mining commenced in 1944. Previous geoscientific research in the area providesdetailed information about lithological variations and structure of the bedrock near the surface.However, knowledge of the depth extent of the mineral deposits and their host rocks is limited. To shedlights on these issues and support deep mineral exploration potential in the study area, within the recentlylaunched StartGeoDelineation project, new ground gravity data, 180 data points on average 150 m apart,were collected during two field campaigns in 2015 and 2016. Aeromagnetic data were obtained fromthe Geological Survey of Sweden (SGU) to complement the ground gravity measurement interpretationsand modelling. After a careful inspection of the field gravity data, they were reduced to completeBouguer anomaly with a maximum error estimate of about 0.6 mGal due to uncertainty in theinstrumental drift, slab density, geodetic surveying, diurnal variations and terrain (or topography)correction. The Bouguer gravity data after separation of regional field (second order polynomial at theend was used) were used (~ 8 mGal range) for interpretation and 3D inverse modelling. Clear anomalouszones are noticeable in the gravity data particularly due to mineralization and a major boundaryseparating a gravity low from gravity high in the southern part of the study area likely representing afault boundary separating two different lithological units. In my study, both forward and inversemodelling using rudimentary objects/shapes and voxel-type (mesh) approach were carried out. Effect ofinitial and reference models were tested on both gravity and magnetic datasets. While the constrainedmodels have still significant ambiguity, they help to suggest structural control on the location ofmineralization and may allow estimating an excess tonnage due to the presence of mineralization in thestudy area. Due to access limitations (e.g., unable to measure on the water-filled pit) the gravity modelis sensitive to the measuring positions and constraints using known shape of mineralization was not atthe end successful to overcome this. Collecting more gravity data on the target area and repeated test of3D inversion by adjusting the inversion parameters might help to improve the final result.
Gruvområdet Blötberget som denna MSc avhandling är fokuserat kring ligger 230 km från Stockholm,12 km från Ludvika, i Bergslagen. Mineralförekomster, framförallt järnmalm (magnetit och hematit)har varit kända i området sedan 1600-talet, och storskalig brytning inleddes år 1944. Tidigare geologiskaundersökningar i området har gett detaljerad information om fyndighetens ytnära litologi och struktur.Hur långt ner förekomsten och moderbergarten sträcker sig har dock varit okänt. Som del av detnystartade projektet StartGeoDelineation utfördes marknära gravimetrimätningar. Totalt 180 mätpunkter,med ett medelavstånd av 150 m, samlades in under två fältkampanjer under 2015 och 2016.Vid modellering komplementades gravimetridata med magnetometridata, insamlad under flygmätningarutförda av Sveriges geologiska undersökningar (SGU). Efter noggrann bearbetning av gravimetridatatogs den kompletta bougeranomalin fram. Det uppskattade felet är ca 0.6 mGal och är till följd avosäkerhet i korrigeringar för drift hos instrument, dygnsvariation, geodesi och topografi. Efter korrigeringav regional trend (uppskattad från 2:a ordningens pylonom, och med satt skala av 8 mGal somresultat) gjordes en 3D modell, via inversionsalgoritmer, samt en tolkning. Det står klart av framförallti gravimetridatan att det finns två avvikande zoner. Dessa indikerar mineraliseringen och en gräns i densödra delen av undersökningsområdet med gravimetridata i låg respektive höga värde. Detta återspeglartroligtvis också en förkastningszon mellan två lithologiska enheter. I denna studie har enkla geometriskaformer och voxlar (mesh) använts för bådadera forward modellering och inversionsalgoritmer. Deursprungliga och referensmodellerna testades på både dataset för gravitmetri och magnetometri. Trotsatt modellerna fortfarande visar tvetydiga resultat så kan de ändå användas för att ge förslag på struktureroch läge för mineraliseringen, och skall även kunna användas för att uppskatta tonnage. Det sistnämndakunde dock inte uppnås då punktäthet i mätdatan, till följd av att det numera vattenfyllda dagbrottet intekunde inkluderas i mätområdet, och att formen av mineraliseringen inte kunde avgränsar på etttillfredsställande sätt. För en förbättring av resultaten bör fler mätpunkter till gravimetridata samlas in iområdet så att 3D-modelleringen kan förbättras genom upprepade justeringar av inversionsparametrarna
StartGeoDelineation

L'imagerie électrique est un outil de plus en plus important pour un large domaine d'applications relatives à la caractérisation de la subsurface proche. D'importants développements ont été réalisés au cours des vingt dernières années pour l'amélioration des systèmes d'acquisitions et des algorithmes d'inversions. L'acquisition et le traitement de gros jeux de données reste toutefois une tâche délicate, en particulier en présence de topographie. Afin d'améliorer la gestion de la topographie, nous avons développé un nouvel algorithme d'inversion électrique 2.5D et 3D. Nous avons proposé deux nouvelles formulations pour supprimer la singularité à la source. Le problème direct est résolu en utilisant la méthode des Différences Finies Généralisées et des maillages non structurés, permettant une représentation précise de la topographie. Le code d'inversion utilise la méthode de l'état adjoint pour calculer le gradient de la fonction objective de manière économique. Cette approche a donné de bons résultats avec des données synthétiques. Les premiers résultats sur des données réelles ont permis de retrouver les principales structures de la subsurface, ainsi que plusieurs zones de faibles résistivités pouvant correspondre à des zones fracturées.

Abstract The lithospheric conductivity in the westernmost Fennoscandia has been studied using magnetotelluric (MT) data. The western margin of Fennoscandia was significantly affected in Paleozoic by the Caledonian orogeny and later by the rifting of Laurentia and the opening of the Atlantic Ocean c. 80 Ma ago. Magnetotelluric studies have been carried out in two target areas in southern Norway and in western Fennoscandia. The first study resulted in 2-D geoelectric models of two profiles stretching from Oslo to the Norwegian coast. The interpretation suggests that the basement is in general very resistive with a few conductive upper crustal layers, representing the alum shales, and middle crustal conductors possibly imaging the remnants of the closed ocean basins. A more extensive MT study was performed within the project "Magnetotellurics in the Scandes". Measurements were carried in summers of 2011 to 2013, resulting in an array of 279 MT sites. The data allowed us to derive 2-D geoelectric models for the crust and upper mantle as well as 3-D models for the crust. The inversions revealed a resistive upper crust and a conductive lower crust, two upper crustal conductors in the Skellefteå and Kittilä districts, highly conducting alum shales in the Caledonides and a conductive upper crust beneath the Lofoten peninsula. The thickness of the lithosphere is around 200 km in the north and 300 km in the south-west. The Palaeoproterozoic lithosphere is the thickest, not the Archaean, on contrary to a generally accepted hypothesis. A better image of the lithosphere will help to evaluate the proposed mechanisms of the exhumation of the Scandinavian Mountains. The theoretical part of this study is the development of a new multi-resolution approach to 3-D electromagnetic (EM) modelling. Three-dimensional modelling of MT data requires enormous computational resources because of the huge number of data and model parameters. The development of the multi-resolution forward solver is based on the fact that a finer grid resolution is often required near the surface. On the other hand, the EM fields propagate in a diffusive manner and can be sufficiently well described on a grid that becomes coarser with depth. Tests showed that the total run time can be reduced by five times and the memory requirements by three times compared with the standard staggered grid forward solver
Tiivistelmä Olemme tutkineet litosfäärin sähkönjohtavuutta Fennoskandian länsiosassa magnetotelluurisen (MT) menetelmän avulla. Fennoskandian länsireuna muokkautui merkittävästi paleotsooisena aikana Kaledonidien vuorijonopoimutuksessa sekä myöhemmin mesotsooisena aikana Laurentia-mantereen repeytyessä ja Atlantin valtameren syntyessä noin 80 miljoonaa vuotta sitten. MT-tutkimukset tehtiin Etelä-Norjassa ja Fennoskandian luoteisosassa. Ensimmäisessä tutkimuksessa kallioperän sähkönjohtavuutta kuvattiin kaksiulotteisilla (2-D) johtavuusmalleilla, jotka ulottuvat Oslosta Norjan rannikolle. Mallien tulkinta viittaa siihen, että maan kuori on pääosin hyvin eristävä lukuun ottamatta muutamaa kuoren ylä- ja keskiosassa olevaa johdekerrosta. Yläkuoren johteet edustavat alunaliuskeita ja keskikuoren johteet todennäköisesti suljetuissa merialtaissa syntyneitä hiilipitoisia sedimenttikerrostumia. Laajempi MT-tutkimus tehtiin ”Magnetotellurics in the Scandes” -hankkeessa. Mittauksia tehtiin 279 mittauspisteessä kesinä 2011–2013. Saadun aineiston avulla voitiin laatia 2-D inversiomallit kuoresta ja ylävaipasta sekä 3-D inversiomalli kuoresta. Tulosten mukaan täällä kuoren yläosa on eristävä kun taas kuoren alaosa on sähköä hyvin johtava. Edellisen lisäksi malleissa näkyy yläkuoren johtavat muodostumat Skellefteån ja Kittilän alueilla, korkean johtavuuden alunaliuskeet Kaledonidien alueella sekä johde Lofoottien alla. Litosfäärin paksuus on noin 200 km mittausverkon pohjoisosassa ja noin 300 km lounaassa. Tämän mukaan litosfääri on paksuin varhaisproterotsooisen litosfäärin alueella, ei arkeeisen litosfäärin alueella vastoin yleistä hypoteesia. Tutkimuksen teoreettisessa osassa kehitettiin sähkömagneettiseen mallinnukseen uusi monitasoiseen diskretisointiin perustuva menetelmä. MT-aineiston 3-D käänteisongelman ratkaisu ja siihen liittyvä suora mallintaminen vaativat suuren laskennallisen kapasiteetin, koska havaintojen ja mallin kuvaamiseen tarvittavien parametrien määrä on erittäin suuri. Moniresoluutio-algoritmi perustuu siihen, että mallin hienojakoisempaa diskretisointia tarvitaan yleensä lähellä maan pintaa kun taas syvemmälle edettäessä, sähkömagneettisen aallon diffuusin etenemisen vuoksi, malli voi olla karkeampi. Tietokonesimulaatioiden mukaan suoritusaika on viidennes ja muistitarve kolmannes verrattuna tavanomaiseen suoran laskennan ”staggered grid” -diskretisointiin

Cette thèse présente une plateforme d'assimilation de données issues de l'imagerie vidéo et intégrée au modèle numérique d'évolution de profil de plage 1DBEACH. Le manque de jeux de données bathymétriques haute-fréquence est un des problèmes récurrents pour la modélisation morphodynamique littorale. Pourtant, des relevés topographiques réguliers sont nécessaires non seulement pour la validation de nos modèles hydro-sédimentaires mais aussi dans une perspective de prévision d'évolution morphologique de nos plages sableuses et d'évolution de la dynamique des courants de baïnes en temps réel. Les récents progrès dans le domaine de l'imagerie vidéo littorale ont permis d'envisager un moyen de suivi morphologique quasi-quotidien et bien moins coûteux que les traditionnelles campagnes de mesure. En effet, les images dérivées de la vidéo de type timex ou timestack rendent possible l'extraction de proxys bathymétriques qui permettent de caractériser et de reconstruire la morphologie de plage sous-jacente. Cependant, ces méthodes d'inversion bathymétrique directes sont limitées au cas linéaire et nécessitent, selon les conditions hydrodynamiques ambiantes, l'acquisition de données vidéo sur plusieurs heures voire plusieurs jours pour caractériser un état de plage. En réponse à ces différents points bloquants, ces travaux de thèse proposaient l'implémentation puis la validation de méthodes d'inversion bathymétrique basées sur l'assimilation dans notre modèle de différentes sources d'observations vidéo disponibles et complémentaires. A partir d'informations hétérogènes et non redondantes, ces méthodes permettent la reconstruction rapide et précise d'une morphologie de plage dans son intégralité pour ainsi bénéficier de relevés bathymétriques haute fréquence réguliers.

Nous étudions le transport du magma dans la croûte superficielle par propagation d'une fracture planaire remplie de fluide sous pression. Deux écoles de pensée formalisent la modélisation de ce phénomène. Elles permettent d'interpréter, soit les aspects géométriques (forme, trajectoire) en négligeant le comportement visqueux du fluide, soit l'aspect temporel (la vitesse d'écoulemenemporalité et de discriminer, in situ et en laboratoire, les champs d'application des écoles de pensée. Nous combinons les données InSAR et GNSS dans une procédure d'inversion originale, tirant avantage, à la fois de la couverture spatiale de l'InSAR et de la résolution temporelle du GNSS. La méthode développée est appliquée à l'étude de la propagation complexe (changements de direction et évolution de la vitesse) ayant conduit à l'éruption du 26 Mai 2016 au Piton de la Fournaise. Ceci permet de valider la méthode et apporte des contraintes nouvelles sur l'alimentation et le déclenchement de cette éruption. En laboratoire, nous cherchons à étudier l'influence de la viscosité du fluide sur la vitesse et la trajectoire d'une fissure sous pression de fluide, isolée, remontant par flottabilité en présence d'un champ de contraintes hétérogène. Nous montrons que l'ajout de sel dans la gélatine utilisée comme analogue de l'encaissant élastique augmente sa résistance à la fracturation. Nous montrons aussi que la trajectoire est le résultat d'une compétition entre la pression interne, le champ de contrainte externe et la longueur de la fissure. Enfin, nous mettons en évidence l'influence des propriétés du milieu et celles du fluide injecté sur la vitesse de propagation et des variations de cette vitesse au cours de la remontée, en présence d'un champ de contrainte hétérogène
We study magma transport in the upper crust by propagation of a buoyancy-driven fluid-filled crack. Two schools of thought formalize the modelling of this phenomenon. They provide a framework to interpret either the geometrical aspects (shape, trajectory) when fluid viscosity is neglected, or the temporal aspect (flow velocity of the fluid) when, resistance to fracturation of the medium is neglected. We use two complementary approaches~: temporal in situ monitoring by inversion of displacement data and analogue modelling, in order to constrain both the geometry and the timing and to discriminate the field of application of each school of thought.We combine InSAR and GNSS data in an original inversion procedure, taking advantage of both the spatial coverage of InSAR and the temporal resolution of GNSS. The method is applied to study the complex propagation (changes of direction and velocity) that led to the eruption of 26 May 2016 at Piton de la Fournaise. This makes it possible to validate the method and provides new constraints on the supply and triggering of this eruption. In the laboratory, we are investigating the influence of fluid viscosity on the velocity and trajectory of a buoyancy-driven fluid-filled crack during ascent in the presence of a heterogeneous stress field. We used gelatine an analogue of elastic host-rock and we show that the addition of salt increases its resistance to fracture. We also show that the trajectory is the result of a competition between the internal pressure, the external stress field and the crack length. Finally, we highlight the influence of the properties of the medium, as well as those of the injected fluid on the propagation velocity and the velocity variations during ascent in the presence of a heterogeneous stress field

La propagation des ondes sismiques dans les milieux poreux multiphasiques présente des enjeux nombreux, tant sur le plan environnemental (risques naturels, géotechnique, pollutions de nappes...) que pour les réservoirs (aquifères, hydrocarbures, stockages de CO2...). L'utilisation des ondes sismiques pour étudier ces milieux se justifie par le fait qu'en se propageant, les ondes sont déformées par le milieu qu'elles traversent et contiennent ainsi des informations aux capteurs sur les phases fluides et solides et sur le squelette poreux. Ce travail de thèse s'intéresse aux caractéristiques des ondes sismiques dans les milieux multiphasiques (plusieurs phases fluides et solides), depuis la description physique jusqu'à la caractérisation des paramètres constitutifs par inversion, en passant par la modélisation numérique 2D de la propagation. La première partie du travail a consisté à décrire la physique des milieux multiphasiques (phase par phase et leurs intéractions dynamiques) en utilisant des méthodes d'homogénéisation pour se ramener à un milieu équivalent défini par sept paramètres. Ainsi, dans des milieux simple porosité saturés et dans des milieux plus complexes (double porosité, partiellement saturés ou visco-poroélastiques), je peux calculer la propagation des ondes sismiques sans approximation. En effet, j'utilise une méthode numérique dans le domaine fréquence-espace qui permet de prendre en compte tous les termes qui dépendent de la fréquence sans approximation. La discrétisation spatiale utilise une méthode d'éléments finis discontinus (Galerkin discontinu) qui permet de considérer des milieux hétérogènes.Je montre notamment que les attributs sismiques (vitesses et atténuations) des milieux poreux complexes sont fortement dispersifs et les formes d'ondes complètes, calculées sans approximation, sont fortement dépendantes de la description physique du milieu. La caractérisation des paramètres poroélastiques s'effectue par inversion. Une méthode en deux étapes a été proposée : la première consiste en une inversion ''classique'' (tomographie, inversion des formes d'ondes complètes) des données (sismogrammes) pour obtenir des paramètres macro-échelles (attributs sismiques). La seconde étape permet de reconstruire, à partir des paramètres macro-échelles, les paramètres poroélastiques micro-échelles. Cette étape d'inversion utilise une méthode d'optimisation semi-globale (algorithme de voisinage). Une analyse de sensibilité montre qu'en connaissant a-priori certains paramètres, on peut inverser avec précision les paramètres du squelette poroélastique ou retrouver la nature du fluide saturant, à partir des vitesses de propagation. En revanche, pour retrouver la saturation en fluide, il est préférable de connaître les atténuations. Deux applications réalistes (monitoring de réservoir et hydrogéophysique) mettent en oeuvre ce type d'inversion en deux étapes et démontrent qu'à partir de données estimées par des méthodes classiques d'imagerie, on peut remonter à certains paramètres poroélastiques constitutifs.

L'amélioration des images sismiques peut aider à mieux contraindre l'exploration deshydrocarbures. Les ondes élastiques qui se propagent dans la Terre peuvent être classifiéescomme ondes de volume et ondes de surface. Si ces dernières sont les plus énergétiques,seules les ondes de volume sont couramment considérées comme des signaux utiles.Cependant, les ondes de surface sont utiles pour caractériser la proche sub-surface.Classiquement, les ondes de surface sont analysées dans des contextes de propriétésélastiques localement 1D.Nous proposons une modification de l'inversion des formes d'onde classique pourreconstruire des profils de propriétés 2D (la windowed-Amplitude Waveform Inversion, w-AWI). La w-AWI est spécialement robuste en ce qui concerne le choix du modèle initial.Nous appliquons la w-AWI aux données synthétiques ainsi qu'aux données réelles, montrantque cette approche permet de récupérer des propriétés 2D.

This thesis presents an all-inclusive framework on how the current formant tracking and audio (and/or visual)-to-articulatory inversion algorithms can be improved. The possible improvements are summarized as follows: The first part of the thesis investigates the problem of the formant frequency estimation when the number of formants to be estimated fixed or variable respectively. The fixed number of formant tracking method is based on the assumption that the number of formant frequencies is fixed along the speech utterance. The proposed algorithm is based on the combination of a dynamic programming algorithm and Kalman filtering/smoothing. In this method, the speech signal is divided into voiced and unvoiced segments, and the formant candidates are associated via dynamic programming algorithm for each voiced and unvoiced part separately. Individual adaptive Kalman filtering/smoothing is used to perform the formant frequency estimation. The performance of the proposed algorithm is compared with some algorithms given in the literature. The variable number of formant tracking method considers those formant frequencies which are visible in the spectrogram. Therefore, the number of formant frequencies is not fixed and they can change along the speech waveform. In that case, it is also necessary to estimate the number of formants to track. For this purpose, the proposed algorithm uses extra logic (formant track start/end decision unit). The measurement update of each individual formant trajectories is handled via Kalman filters. The performance of the proposed algorithm is illustrated by some examples The second part of this thesis is concerned with improving audiovisual to articulatory inversion performance. The related studies can be examined in two parts
Gaussian mixture model (GMM) regression based inversion and Jump Markov Linear System (JMLS) based inversion. GMM regression based inversion method involves modeling audio (and /or visual) and articulatory data as a joint Gaussian mixture model. The conditional expectation of this distribution gives the desired articulatory estimate. In this method, we examine the usefulness of the combination of various acoustic features and effectiveness of various types of fusion techniques in combination with audiovisual features. Also, we propose dynamic smoothing methods to smooth articulatory trajectories. The performance of the proposed algorithm is illustrated and compared with conventional algorithms. JMLS inversion involves tying the acoustic (and/or visual) spaces and articulatory space via multiple state space representations. In this way, the articulatory inversion problem is converted into the state estimation problem where the audiovisual data are considered as measurements and articulatory positions are state variables. The proposed inversion method first learns the parameter set of the state space model via an expectation maximization (EM) based algorithm and the state estimation is handled via interactive multiple model (IMM) filter/smoother.

Future sea level rise associated to global warming is one of the greatest societal and environmental challenges of tomorrow. A large part of the contribution comes from glaciers and ice sheets discharging ice and meltwater into the ocean and the recent worldwide increase is worrying. Future predictions of sea level rise try to encompass the complex processes of ice dynamics through glacier modelling but there are still large uncertainties due to the lack of observations or too coarse parameterisation, particularly for processes occurring at the glacier interfaces with the bed (sliding) and with the ocean (calving). This thesis focuses on modelling these processes from two marine-terminating glaciers in Svalbard, Kronebreen and Tunabreen. By inverting three years of high temporal resolution time-series of surface velocities on Kronebreen, basal properties are retrieved with the ice flow model Elmer/Ice in Paper I. Results suggest that surface melt during the summer greatly influences the dynamics of the following season and that sliding laws for such glaciers should be adapted to local and global processes changing in space and time. The subglacial drainage system, fed by the surface melt, is modelled in Paper II during two melting seasons. Results show different configurations of efficient and inefficient drainage systems between years and the importance of using a sliding law dependent on spatio-temporal changes in effective pressure. The interaction with the ocean is incorporated in Paper III by combining a series of models, including an ice flow model, a plume model and a particle model for discrete calving and compares the output with observations. Results show the importance of glacier geometry, sliding and undercutting on calving rate and location. However, more observations and analytic methods are needed. Time-lapse imagery placed in front of Tunabreen have been deployed and a method of automatic detection for iceberg calving is presented in Paper IV. Results show the influence of the rising plume in calving and the front destabilisation of the local neighbourhood.

Les eaux souterraines constituent une réserve d’eau potable non négligeable, leur alimentation se fait en majeure partie par les précipitations, appelée recharge des nappes. Du fait de leur grande importance, la compréhension du fonctionnement de ces ressources en eau est plus que jamais indispensable. Celle-ci passe par l’élaboration de modèles mathématiques. Ces outils nous offrent une meilleure appréhension et une bonne prévision des phénomènes physiques. Les systèmes hydrogéologiques sont généralement très complexes et caractérisés par des dynamiques hydriques très variables dans le temps et dans l’espace. Cette complexité a attiré l’attention de nombreux hydrogéologues et un grand nombre de modèles très sophistiqués a été développé afin de décrire ces systèmes avec précision. Cependant, la prise en compte de la recharge de ces réservoirs reste toujours un défi dans la modélisation hydrogéologique. En effet, le plus souvent, les modèles hydrogéologiques simulent l’écoulement dans la nappe tout en considérant la recharge comme une constante sur le domaine et indépendante du système. De plus, elle est souvent calculée de façon simplifiée. Or, la recharge traduisant la quantité des précipitations atteignant les nappes est une composante hydrologique complexe et variable car elle interagit avec les nappes et dépend des conditions climatiques, du couvert végétal et du transfert de l’eau dans le sol. Ce présent travail vise à intégrer cette recharge variable et complexe aux modèles hydrogéologiques. À cet effet, un modèle couplé a été développé. Une première partie de ce modèle permet de calculer la recharge des nappes en modélisant les interactions précipitations-sol et l’hydrodynamique dans le sol. Cette modélisation a été effectuée en utilisant des modèles conceptuels simples basés sur des lois empiriques (Gardénia, Nash) et des modèles physiques résolvant l’équation de Richards. La recharge ainsi calculée est intégrée à la deuxième partie du modèle simulant l’hydrodynamique dans les nappes décrite par l’équation de diffusivité. Des méthodes numériques précises et robustes ont été utilisées pour résoudre les équations du modèle mathématique : les éléments finis non conformes ont été utilisés pour résoudre l’équation de diffusivité et l’équation de Richards est résolue sous sa forme mixte par une méthode itérative en temps. En somme, ce modèle couplé permet de décrire les variations de niveaux de nappe à partir des données météorologiques connaissant les paramètres caractéristiques de cet aquifère. [...]
Groundwater is the main available water resource for many countries; they are mainly replenished by water from precipitation, called groundwater recharge. Due to its great importance, management of groundwater resources is more essential than ever, and is achieved through mathematical models which offer us a better understanding of physical phenomena as well as their prediction. Hydrogeological systems are generally complex thus characterized by a highly variable dynamic over time and space. These complexities have attracted the attention of many hydrogeologists and many sophisticated models that can handle these issues and describe these systems accurately were developed. Unfortunately, modeling groundwater recharge is still a challenge in groundwater resource management. Generally, groundwater models are used to simulate aquifers flow without a good estimation of recharge and its spatial-temporal distribution. As groundwater recharge rates show spatial-temporal variability due to climatic conditions, land use, and hydrogeological heterogeneity, these methods have limitations in dealing with these characteristics. To overcome these limitations, a coupled model which simulates flow in the unsaturated zone and recharge as well as groundwater flow was developed. The flow in the unsaturated zone is solved either with resolution of Richards equation or with empirical models while the diffusivity equation governs flow in the saturated zone. Robust numerical methods were used to solve these equations: we apply non-conforming finite element to solve the diffusivity equation and we used an accurate and efficient method for solving the Richards equation. [...]

Les matériaux polymères sont aussi bien utilisés dans des applications de la vie courante que dans des domaines de haute technologie. Ces matériaux font intervenir des échelles spatiales et temporelles variées et étendues, rendant la modélisation de leurs propriétés inaccessible avec une seule méthode. Cette thèse propose le développement d’une stratégie multi-échelle, couplant ainsi plusieurs niveaux de représentation de la matière. Le but est d’accéder aux propriétés rhéologiques d’un polymère, faisant intervenir des temps de relaxation lents, tout en conservant les caractéristiques chimiques intrinsèques à sa microstructure de façon à pouvoir établir des relations structure-propriétés. Les potentiels d’interaction de l’échelle mésoscopique sont développés à partir des configurations atomistiques. Ils permettent ensuite une reproduction quantitative de plusieurs propriétés structurales du polymère, telles que la masse volumique ou la distance bout à bout. La transférabilité des potentiels mésoscopiques a été étudiée à travers la dépendance des propriétés thermomécaniques en température, en pression et en nature du polymère. À partir de ces potentiels, des simulations hors équilibre ont permis de déterminer des grandeurs caractéristiques comme la masse d’enchevêtrement ou le module élastique. L’approche multi-échelle est étendue à l’interaction polymère-silice, dans le but d’étudier l’impact des facteurs comme le degré de confinement ou la densité de greffage sur les propriétés dynamiques et structurales des chaînes au voisinage de la surface
Polymer materials are widely used, both for everyday applications and in high-technology products. These materials involves a wide range of time and length scales, making the modelling of their properties challenging by using only one method. This thesis focuses on the development of a multiscale strategy, combining different levels of description of the matter. The aim is to reach the rheological properties of a polymer over a large time scale, while retaining the chemical structure inherent of its microstructure. The investigation of structure-property relationships will then be facilitated. The mesoscopic potentials are developped from atomistic configurations. A quantitative reproduction of several structural properties of the polymer such as density or end to end distance is obtained. Then, the transferability of the potentials has been studied through the dependence of temperature, pressure or polymer structure on thermomechanical properties. By using these potentials, nonequilibrium simulations have been carried out to calculate the entanglement mass and the plateau modulus. The multicale approach has been extended to model the polymer-silica interaction in order to study the impact of the degree of confinement or the grafting density on the dynamical and structural properties of polymer chains close to the surface

Abstract The crustal and upper mantle structures of the Shield on the regional scale were investigated using the data of the POLENET/LAPNET passive seismic array and the previously published models of active and passive seismic experiments in the study area. This area is centred in northern Finland and it extends to surrounding areas in Sweden, Norway and northwestern Russia. The bedrock there is mostly of the Archaean origin and the lithosphere of the region was reworked by two orogenies during Palaeoproterozoic. One of the results of the thesis was a new map of the Moho depth of the study area, for which new estimates of the crustal thickness were obtained using receiver function method and complemented by published results of receiver function studies and controlled source seismic profiles. The map differs from the previously published maps in two locations, where we found significant deepening of the Moho. The 3D structure of the upper mantle was studied using teleseismic traveltime tomography method. The resulting model shows high seismic velocities below three cratonic units of the study area, which may correspond to non-reworked fragments of cratonic lithosphere and a low velocity anomaly separating these cratonic units from each other. The regional scale studies were complemented by two smaller scale studies in upper crust level using combined interpretation of seismic profiling and gravity data. These studies were centred on Archaean Kuhmo Greenstone Belt in eastern Finland and central Lapland in northern Finland located in the crust reworked during Palaeoproterozoic. Both areas are considered as prospective ones for mineral exploration. Both studies demonstrate the advantage of gravity data inversion in studying 3D density structure of geologically interesting formations, when the Bouguer anomaly data is combined with a priori information from petrophysical and seismic datasets.

Les premiers mètres à centaines de mètres de la proche surface terrestre sont le siège de processus naturels dont la compréhension requiert une caractérisation fine de la subsurface, via une estimation quantifiée de ses paramètres. Le géoradar est un outil de prospection indirecte à même d'ausculter les milieux naturels et d'en estimer les propriétés électriques (permittivité et conductivité). Basé sur la propagation d'ondes électromagnétiques à des fréquences allant du MHz à quelques GHz, le géoradar est utilisé à des échelles et pour des applications variées concernant la géologie, l'hydrologie ou le génie civil. Dans ce travail de thèse, je propose une méthode d'imagerie quantitative des propriétés électriques sur des sections 2D de la subsurface, à partir de données radar acquises à la surface du sol. La technique mise en oeuvre est l'inversion des formes d'ondes, qui utilise l'intégralité du champ d'ondes enregistré.Dans une première partie, je présente les principes physiques et l'outil de modélisation numérique utilisés pour simuler la propagation des ondes électromagnétiques dans les milieux hétérogènes à deux dimensions. Pour cela, un algorithme de différences finies en domaine fréquentiel développé dans le cadre des ondes visco-acoustiques est adapté au problème électromagnétique 2D grâce à une analogie mathématique.Dans une deuxième partie, le problème d'imagerie est formulé sous la forme d'une optimisation multi-paramètre puis résolu avec l'algorithme de quasi-Newton L-BFGS. Cet algorithme permet d'estimer l'effet de la matrice Hessienne, dont le rôle est crucial pour la reconstruction de paramètres de différents types comme la permittivité et la conductivité. Des tests numériques montrent toutefois que l'algorithme reste sensible aux échelles utilisées pour définir ces paramètres. Dans un exemple synthétique représentatif de la proche surface, il est cependant possible d'obtenir des cartes 2D de permittivité et de conductivité à partir de données de surface, en faisant intervenir des facteurs d'échelle et de régularisation visant à contraindre les paramètres auxquelles l'inversion est la moins sensible. Ces facteurs peuvent être déterminés en analysant la qualité de l'ajustement aux données, sans hypothèse a priori autre que la contrainte de lissage introduite par la régularisation.Dans une dernière partie, la méthode d'imagerie est confrontée à deux jeux de données réelles. Dans un premier temps, l'examen de données expérimentales permet de tester la précision des simulations numériques vis-à-vis de mesures effectuées en environnement contrôlé. La connaissance des cibles à imager permet en outre de valider la méthodologie proposée pour l'imagerie multiparamètre dans des conditions très favorables puisqu'il est possible de calibrer le signal source et de considérer l'espace libre environnant les cibles comme modèle initial pour l'inversion.Dans un deuxième temps, j'envisage le traitement d'un jeu de données radar multi-offsets acquises au sein d'un massif calcaire. L'interprétation de ces données est rendue beaucoup plus difficile par la complexité du milieu géologique environnant, ainsi que par la méconnaissance des caractéristiques précises des antennes utilisées. L'application de la méthode d'inversion des formes d'ondes à ces données requiert donc une étape préliminaire impliquant une analyse de vitesse plus classique, basée sur les arrivées directes et réfléchies, et des simulations numériques dans des modèles hypothétiques à même d'expliquer une partie des données. L'estimation du signal source est effectuée à partir d'arrivées sélectionnées, simultanément avec des valeurs moyennes de conductivité et de hauteur d'antennes de façon à reproduire au mieux les amplitudes observées. Un premier essai d'inversion montre que l'algorithme est capable d'expliquer les données dans la gamme de fréquences considérée et de reconstruire une ébauche des principaux réflecteurs
The quantitative characterization of the shallow subsurface of the Earth is a critical issue for many environmental and societal challenges. Ground penetrating radar (GPR) is a geophysical method based on the propagation of electromagnetic waves for the prospection of the near subsurface. With central frequencies between 10~MHz and a few GHz, GPR covers a wide range of applications in geology, hydrology and civil engineering. GPR data are sensitive to variations in the electrical properties of the medium which can be related, for instance, to its water content and bring valuable information on hydrological processes. In this work, I develop a quantitative imaging method for the reconstruction of 2D distributions of permittivity and conductivity from GPR data acquired from the ground surface. The method makes use of the full waveform inversion technique (FWI), originating from seismic exploration, which exploits the entire recorded radargrams and has been proved successful in crosshole GPR applications.In a first time, I present the numerical forward modelling used to simulate the propagation of electromagnetic waves in 2D heterogeneous media and generate the synthetic GPR data that are compared to the recorded radargrams in the inversion process. A frequency-domain finite-difference algorithm originally developed in the visco-acoustic approximation is adapted to the electromagnetic problem in 2D via an acoustic-electromagnetic mathematical analogy.In a second time, the inversion scheme is formulated as a fully multiparameter optimization problem which is solved with the quasi-Newton L-BFGS algorithm. In this formulation, the effect of an approximate inverse Hessian is expected to mitigate the trade-off between the impact of permittivity and conductivity on the data. However, numerical tests on a synthetic benchmark of the literature display a large sensitivity of the method with respect to parameter scaling, showing the limits of the L-BFGS approximation. On a realistic subsurface benchmark with surface-to-surface configuration, it has been shown possible to ally parameter scaling and regularization to reconstruct 2D images of permittivity and conductivity without a priori assumptions.Finally, the imaging method is confronted to two real data sets. The consideration of laboratory-controlled data validates the proposed workflow for multiparameter imaging, as well as the accuracy of the numerical forward solutions. The application to on-ground GPR data acquired in a limestone massif is more challenging and necessitates a thorough investigation involving classical processing techniques and forward simulations. Starting permittivity models are derived from the velocity analysis of the direct arrivals and of the reflected events. The estimation of the source signature is performed together with an evaluation of an average conductivity value and of the unknown antenna height. In spite of this procedure, synthetic data do not reproduce the observed amplitudes, suggesting an effect of the radiation pattern of the shielded antennae. In preliminary tests, the inversion succeeds in fitting the data in the considered frequency range and can reconstruct reflectors from a smooth starting model

Em razão da crescente preocupação com a proteção ao meio ambiente e preservação da qualidade das águas superficiais e subterrâneas, as áreas de disposição de resíduos, tanto domésticos como industriais, têm merecido especial atenção das instituições de pesquisa ligadas à proteção ambiental. Atualmente o uso dos métodos geoelétricos vem se destacando como metodologia indireta de investigação de áreas afetadas pela disposição inadequada de resíduos sólidos urbanos e industriais pois fornecem, desde que adequadamente escolhidas as técnicas e os parâmetros de aquisição, informações importantes com respeito à localização e geometria dos depósitos de resíduos, extensão da pluma de contaminação, posição da zona saturada e sentido do fluxo subterrâneo, características litológicas e presença de estruturas geológicas. A integração de ferramentas como simulações e métodos de interpretação qualitativa e quantitativa também auxiliam na caracterização e monitoramento ambiental de áreas de disposição de resíduos sólidos. O objetivo principal deste trabalho foi a avaliação de alguns parâmetros de aquisição 2D (arranjos eletródicos, espaçamento entre eletrodos e níveis de investigação) com os métodos da Eletrorresistividade e da Polarização Induzida, para caracterização e monitoramento ambiental. Esta avaliação foi realizada através do uso de modelagens diretas bidimensionais e ensaios de campo, tendo como área de estudos o antigo lixão do município de Ribeirão Preto-SP. O uso das modelagens diretas permitiu escolher os melhores parâmetros de aquisição de acordo com as respostas obtidas pelo modelo geológico-geofísico simulado. Os parâmetros de aquisição testados em campo refletiram os resultados obtidos através das modelagens diretas. O arranjo Polo-dipolo (a= 5 metros e n= 10) apresentou os melhores resultados (melhor resolução lateral e maior profundidade de investigação). Os resultados obtidos com as investigações de campo comparados com resultados anteriores (1997), mostraram que ainda há contaminação, marcada por baixos valores de resistividade elétrica (1 à 30 Ohm.m) e altos valores de cargabilidade (15 à 50 mV/V).
Due to the increasing concern with the environmental protection and preservation of the shallow water and groundwater aquifers quality, the areas of disposal of residues, have been special attention of the environmental protection agency. Currently the geoelectrical methods have been largely used as an investigation non-evasive methodology in affected areas by inadequate disposal of urban solid and industrial residues, because it provides information of the localization and geometrical characteristics of the waste disposal, contamination plume extent, saturated zone position, underground flow direction and geologic characteristic, if the techniques and the acquisition parameters are adequately chosen. The simulations and interpretation methods also assists during the characterization and environmental monitoring of landfill areas. The objective of this research was the evaluation of the acquisition field parameters (electrodes arrays, electrodes spacing and number of the investigations levels) and 2D investigations with the resistivity and induced polarization methods for characterization and environmental monitoring. This evaluation was carried through the use of 2D forward modeling and field investigations in old landfill site at Ribeirão Preto city, São Paulo, Brazil. The use of the forward modeling allowed to choose the best field acquisition parameters in accordance with the answers for the geologic-geophysical simulated model. The comparison between our results and previous results showed contamination marked for low electric resistivity values (1 ? 30 Ohm.m) and high chargeability values (15 ? 50 mV/V) and also that the tested parameters of acquisition in agreement with the results of the forward modeling. The Pole-dipole array (a= 5 meters and n= 10) show the best result (resolution image and investigation depth), forward modelling and field acquisition.

An integrated use of radio magnetotelluric (RMT) and controlled source tensor magnetotelluric (CSTMT) measurements, the so-called CSRMT method, has been employed in environmental and resource exploration studies. A number of case histories, including a groundwater investigation in glacial deposits, a study of fracture zones for geotechnical purposes and a mining exploration study of a copper deposit, are presented in this thesis in order to illustrate the usefulness and capability of the CSRMT method. The resolutions of the estimated models using various types of data are studied. Magnetotelluric transfer functions are used to analyze the dimensionality, the near surface resistivity distortions and the near field effects in the case of CSTMT data analysis. The near field effects in CSTMT data have also been identified by performing 2½D forward modelling. Data analysis, dimensionality tests and forward modelling show that at the lowest frequencies used the CSTMT transfer functions are generally distorted by source effects, except when the source-receiver distances are sufficient large compared with the penetration depth. Regarding CSTMT transfer functions, apparent resistivities are generally less distorted than phases. TM mode transfer functions are more affected by the sources than TE mode, while tipper vectors generally contain source signatures at all frequencies. Based on the analysis of dimensionality and source effects 2D inverse modelling of CSTMT and RMT data, as well as their combination, have been performed under the plane wave assumption. The RMT method proved to be a powerful tool for imaging the upper 50 m near-surface, but their penetration depth reduces as a conductive layer structures cover the targets at depth. The penetration depth can be increased by including the CSTMT data in the modelling if the measurements are in the far field range. The resolution of the deeper parts of the models may be improved by performing a joint inversion of TE and TM modes, if the strike direction is well-defined. Alternatively, inversion of determinant data can be performed, since the determinant data are less affected by 3D structures and source effects. However the resolution of the determinant models is somewhat degraded compared to the models inverted from combined TE and TM modes.

Le travail présenté dans ce mémoire de thèse s'inscrit dans le cadre de la commande non-linéaire et robuste des avions de transport. Le but de cette thèse est de coupler les propriétés de la commande non-linéaire (adaptation aux non-linéarités de l'avion, synthèse de correcteurs explicites, facilité de réglage une fois la synthèse réalisée, généricité des lois de commande obtenues) à des propriétés de robustesse indispensables à l'activité aéronautique. En effet, pour garantir la sécurité des vols tant en pilotage manuel qu'en pilotage automatique, les lois de commande doivent présenter des propriétés fortes de stabilité et de performances robustes. Après une introduction au contexte industriel et de recherche du sujet de thèse, une partie "techniques, méthodes et outils" nous permet de mettre en avant les contributions du travail de thèse dans les domaines de la commande non-linéaire robuste et de la modélisation automatique. La technique de commande non-linéaire robuste présentée, appelée commande RMI (Robust Multi-Inversion) s'appuie sur la technique désormais classique d'inversion de la dynamique, notamment étudiée à Airbus depuis quelques années (Fabrice VILLAUME, Jean DUPREZ) et qu'elle robustifie par l'adjonction d'une boucle supplémentaire d'observation. Nous présentons aussi un outil de génération automatique de modèles non-linéaires, multivariables et embarquables, ainsi que les méthodes afférentes basées sur les réseaux de neurones. Cet outil est nécessaire à l'industrialisation des lois de commandes non-linéaires basées modèles. La partie applicative de la thèse souligne ensuite les particularités du système "avion" et propose des architectures de lois de commande, des trajectoires de référence associées, et la validation avancée de l'ensemble par simulations sur simulateur certifié. Enfin, après une conclusion sur le bilan de la thèse et les perspectives envisageables, nous proposons des annexes permettant d'approfondir certains aspects de notre étude
The work presented in this PhD thesis report is situated within the framework of the nonlinear and robust control of transport aircrafts. The purpose of this thesis is to couple the properties of nonlinear controllers (adaptation to the aircraft nonlinearities, explicit controllers synthesis, easy and decoupled setting once the synthesis is achieved, genericity of the obtained control laws) with essential robustness properties. Indeed, to guarantee the flight safety, both in manual handling and in automatic control, the control laws have to present strong robust stability and performances properties. After an introduction to the industrial and research context, a "techniques, methods and tools" part allows us to point out the thesis contributions in the nonlinear robust control and automatic modelling domains. The nonlinear robust control technique presented, called RMI control (for "Robust Multi-Inversion") is based on the now classical Nonlinear Dynamic Inversion (NDI) technique, notably studied at Airbus for some years (Fabrice VILLAUME, Jean DUPREZ), and is robustified by adding a complementary observation loop. We also present an automatic tool creating nonlinear, multivariable and embeddable models, as well as neural networks correlated methods. This tool is mandatory for the industrialization of our model-based flight control laws. Then the applicative part of the thesis underlines the specificities of the "aircraft" system and proposes flight control laws architectures, associated reference trajectories, and the advanced validation of the whole system by simulations performed on Airbus' certified simulator. Finally, after a conclusion on the main results and perspectives linked to the thesis, we propose annexes allowing to go further into the details of certain parts of our study

La modélisation des structures géologiques est une étape cruciale de la construction de géomodèles. Elles conditionnent la répartition spatiale des propriétés du sous-sol et la connexion des différents volumes rocheux qui le constituent. Cependant, les données disponibles s'avèrent incomplètes et incertaines. A cela s'ajoute une incertitude sur les phénomènes géologiques ayant conduit à la formation des structures modélisées. De multiples interprétations sont donc généralement possibles. Les méthodes classiques de modélisation des structures reposent principalement sur une approche géométrique, statique et déterministe. En d'autres termes, l'approche consiste à déterminer la meilleure représentation possible des structures en se fondant sur des critères qui portent principalement sur la géométrie actuelle des structures, comme la minimisation de leur courbure, et sur leur compatibilité avec les données dans leur état actuel. L'histoire tectonique des structures n'est prise en compte que de manière indirecte par le modélisateur et la compatibilité cinématique et mécanique du modèle structural reste complexe à évaluer. Cette thèse explore différentes méthodes de modélisations permettant de mieux prendre en compte l'histoire tectonique des structures. Trois approches complémentaires sont développées. La première propose un opérateur cinématique représentant les déplacements associés aux failles. Il tire son originalité du système de coordonnées curvilinéaires utilisé pour décrire l'espace de la faille et de l'intégration progressive des déplacements permettant de contrôler leur évolution. La seconde approche propose un outil de déformation pseudo-mécanique permettant l'édition interactive des modèles structuraux et la simulation approchée de l'histoire tectonique qui leur est associée. Il repose sur un algorithme de déformation original, basé sur des éléments rigides, issu du domaine de l'infographie. Enfin, une troisième approche permet de répercuter l'historique de déformation sur le système de coordonnées utilisé pour la simulation des propriétés du sous-sol. Nous montrons comment ce système peut être construit grâce à la restauration des structures. Cette approche permet de modéliser les propriétés du sous-sol en s'appuyant sur des paléodistances compatibles avec l'ensemble des hypothèses structurales, cinématiques et mécaniques établies lors de la construction de l'histoire tectonique du géomodèle. Ces différentes approches ouvrent des portes prometteuses dans la prise en compte des aspects historiques, cinématiques et mécanique dans la modélisation des structures et des propriétés du sous-sol. Elles s'attachent à accroître la compatibilité des géomodèles et à simplifier la paramétrisation des déformations géologiques afin de faciliter la résolution de problèmes structuraux par des approches inverses
The main approaches to the modelling of geological structures are mainly geometrical, static and deterministic. In other terms, their geometry and connections are determined by applying criteria based on the compatibility with available data in their current state. The evolution of the geological structures is only integrated indirectly by the modeller and the kinematical and mechanical compatibility of the produced models remain difficult to assess. This thesis explores different methods which aim at better including the evolution of geological structures in the modelling process. Three complementary approaches are developed. First, a kinematical fault operator based on a 3D curvilinear fault frame is presented. It aims at progressively deforming the structures surrounding faults. The second approach is based on a pseudo-mechanical deformation tool inspired form computer graphics, based on rigid elements. It is used to interactively editing the structures and approximately simulate their deformation history. The last proposal is to compute the paleo-geographical coordinates from the restoration of geological structures. This way, the heterogeneities are characterised based on paleo-geographic distances which are compatible with the structural, kinematical and mechanical hypotheses specified when building the geological model. These different contributions open numerous perspectives to better take into account the evolution of the geological structures when modelling the subsurface and its heterogeneities. They help us to increase the compatibility of geomodels and simplify the parameterization of geological deformation to facilitate the characterisation of geological structures by inverse approaches

Basal Temperature of Snow (BTS) measurements were used as the primary inputs to a high resolution (30 x 30 m grid cells) empirical-statistical regional permafrost probability model for the southern and central Yukon, and northernmost British Columbia (59° - 65°N). Data from seven individual study areas distributed across the region were combined using a blended distance decay technique, with an eighth area used for validation. The model predictions are reasonably consistent with previous permafrost maps for the area with some notable differences and a much higher level of detail. The modelling gives an overall permafrost probability of 52%. North of 62°N, permafrost becomes more extensive in the lowland areas whereas farther south permafrost is typically common only above treeline. Significant differences exist between the mountain environments of the Yukon and the Swiss Alps where the BTS method originated and as a result different modelling approaches had to be developed. This work therefore: (1) develops additional explanatory variables for permafrost probability modelling, the most notable of which is equivalent elevation, (2) confirms the use of ground truthing as a requirement for empirical-statistical modelling in the Yukon and (3) uses a combination of models for the region in order to spatially predict between study areas. The results of this thesis will be of use to linear infrastructure route-planning, geohazard assessment and climate change adaptation strategies. Future work employing the model will allow the effects of scenario-based climate warming to be examined.

O presente trabalho teve como objetivo a modelagem de células solares MOS operando em regime de inversão controlado por centros positivamente carregados na interface SiNxOy/Si. Este tipo de célula solar foi recentemente fabricada pela primeira vez no âmbito dos trabalhos desenvolvidos no grupo de Superfícies, Interfaces e Deposição Eletroquímica (GSIDE) do LSI/PSI/EPUSP utilizando dielétricos de porta ultra-finos (~2nm). A receita de crescimento de dielétrico ultra-fino desenvolvida foi no sentido de assegurar reprodutibilidade e uniformidade da espessura do dielétrico ao longo de áreas extensas de alguns cm2. Baseado nas curvas experimentais CxVg, GxVg e IxVg das células solares fabricadas, foi mostrado para as células fabricadas em substrato tipo P que existem os centros K predominantemente preenchidos com cargas positivas em todos os regimes de operação (acumulação, depleção e inversão). A densidade de cargas positivas (Qiq) na interface SiNxOy/Si além de ter resultado positivo, apresentou um comportamento linear com o potencial de superfície (ys) ou com a tensão de porta Vg de acordo com os resultados obtidos através de um simulador numérico desenvolvido para esta aplicação específica. Tal comportamento consistiu no acomodamento das cargas positivas na interface de forma que uma região de depleção profunda (Wd) é formada sem a presença da camada de inversão na condição sem iluminação. Para as células MOS submetidas a diferentes níveis de iluminação, tanto para os dielétricos crescidos a 850oC como também para aqueles que foram crescidos a 700oC, foi constatado que os centros K na interface funcionam como uma região de armazenamento de cargas positivas a medida em que os elétrons tunelam em direção à porta metálica da estrutura MOS. Como resultado, este tipo de comportamento significa uma nova forma de implementar o efeito fotovoltáico.
The goal of the present work was the modeling of MOS solar cells operating in an inversion regime controlled by positively charged centers at the SiNxOy interface. This type of solar cell was recently manufactured for the first time in the activities developed in the group of Surfaces, Interfaces and Electrochemical Deposition (GSIDE) from LSI/PSI/EPUSP using ultra-thin gate dielectrics (~2nm). The recipe for the growth of ultra-thin dielectrics was developed to ensure reproducibility and uniformity of the dielectrics thickness over large areas of few square centimeters. Based on the experimental curves CxVg, GxVg e IxVg of the manufactured MOS solar cells, it was shown for cells manufactured in P-type substrate that there are K centers dominantly filled with positive charges in all operating regimes (accumulation, depletion and inversion). The positive charge density (Qiq) at the SiNxOy/Si interface, in addition to having a positive charge, presented a linear behaviour with the surface potential (ys) or with the gate voltage (Vg) according to the results obtained from a numerical simulator developed for this application. Such behavior consisted of accommodating the positive charges at the SiNxOy/Si interface so that a deep depletion region (Wd) is formed without the presence of the inversion layer in the condition without illumination. For MOS cells subjected to different levels of illumination, both for dielectrics grown at 850oC as well as for those grown at 700oC, it was found that the K centers at the SiNxOy/Si interface work as a region of positive charge storage as the electrons tunnel from the interface towards the metal gate of the MOS cells. As a result, this type of behaviour means a new way of implementing the photovoltaic effect.

Actuellement, la majorité des équipements électroniques prenant place dans nos véhicules s'interconnectent encore par des liens câblés. Or, l'établissement d'un réseau sans fil permet de remplacer ces liens existants. Afin de permettre cette connectivité et pour répondre aux exigences futures, une solution qui accepte suffisamment de débit doit être développée. Parmi les solutions actuelles possibles, l'Ultra-Large Bande est le candidat idéal. De manière générale, cette Thèse se consacre à l'étude de la propagation à l'intérieur d'une voiture. Afin d'évaluer les performances de ce type de système à l'intérieur d'un véhicule, un modèle de canal pouvant reproduire avec fidélité la propagation intra-véhiculaire est nécessaire. De par la similarité entre les environnements, le modèle de canal proposé est basé sur la théorie des chambres réverbérantes. Compte tenu des faibles dimensions de l'environnement et de la proximité des antennes par rapport aux passagers, l'idée est de réduire l'énergie absorbée par ces derniers tout en maximisant celle-ci à la position d'un récepteur visé. Parmi les solutions possibles, le retournement temporel apparait comme une technique idéale pour répondre à ces objectifs. Cette technique permet en effet de focaliser temporellement et spatialement le signal électromagnétique au niveau du récepteur. Dans cette Thèse, une évaluation des performances du RT à l'intérieure d'une voiture est proposée. Cette étude nous a amenés à porter notre attention sur les propriétés spatiales de l'énergie totale portée par un signal utilisant le retournement temporel et, de ce fait, nous proposons un nouveau formalisme de la distribution spatiale de cette énergie. / Current vehicles still use wired links to interconnect the embedded electronic equipment. However, advances in wireless technologies permit to replace these links with wireless personal area networks. To enable this connectivity and to meet the future requirements, a solution that provides sufficient data rate must be developed. One technological candidate that can be used to meet the data rate requirements is Ultra-Wide Band. This thesis investigates the ultra-wideband propagation environment in intra-vehicular environments. In order to evaluate the performances of UWB inside a vehicle, channel models that accurately reproduce the intra-vehicular propagation characteristics are required. Because of obvious similarities, the proposed model is based on reverberation chamber theory. Given the small size of the environment and the proximity of the antennas to the passengers, one goal of intra-vehicular communications is to reduce the energy absorbed by the passengers while maximizing the energy at the position of the intended receiver. Among the possible solutions, Time Reversal appears to be the ideal technique to meet this objective. Time Reversal permits to spatially and temporally focus the electromagnetic signal at the receiver. In this thesis, a performance evaluation of Time Reversal inside a car is proposed. Moreover, a model of the spatial properties of the total energy of a signal using Time Reversal is proposed.
Doctorat en Sciences de l'ingénieur
info:eu-repo/semantics/nonPublished

[ES] El principal objetivo de la presente tesis es el desarrollo de una completa metodología para el modelado numérico del UHPFRC desde el material hasta el elemento estructural. Se pretende contribuir al avance del conocimiento del comportamiento mecánico del UHPFRC obteniendo como resultado un procedimiento para la modelización numérica que permita el modelado y diseño estructural que permitiría hacer que este material fuera competitivo para ser utilizado en el mercado de la construcción. En la metodología de modelado propuesta, se considera un comportamiento constitutivo del UHPFRC optimizado por medio de un procedimiento directo y fiable con el que se aprovechan las ventajas del material, resultando en un diseño estructural eficiente desde el punto de vista mecánico y económico. ¿Es necesario producir SH-UHPFRC para conseguir grandes propiedades mecánicas? ¿Es posible generar SS-UHPFRC de manera que queden reducidos los costos iniciales y se mantengan unas propiedades mecánicas y de durabilidad competitivas que comporten un diseño estructural efectivo? El desarrollo de UHPFRC con bajo endurecimiento por deformación y de SS-UHPFRC puede reducir sus propiedades mecánicas, pero si son adecuadamente estudiadas y controladas, éstos podrían ser optimizados. La tesis aborda algunas de estas cuestiones a través del estudio del comportamiento a tracción que va desde SH-UHPFRC hasta SS-UHPFRC. Se pretende llevar a cabo una propuesta de procedimiento fiable para caracterizar el comportamiento constitutivo a tracción y definir un modelo numérico de elementos finitos fiable para modelar con precisión la respuesta de probetas y elementos estructurales armados de UHPFRC. Para definir el procedimiento directo para caracterizar a tracción tanto SH-UHPFRC como SS-UHPFRC, se ha llevado a cabo una campaña experimental y numérica en la que se ha analizado el resultado de ensayar 227 probetas sin armadura fabricadas con UHPFRC con cantidades de fibras cortas y lisas de acero de 120-130kg/m3 y 160kg/m3, ensayadas a flexión a través del ensayo a cuatro puntos (4PBT). El desarrollo y la validación de dicho proceso se respaldan mediante un modelo no lineal de elementos finitos (NLFEM) fiable. La validación numérica llevada a cabo ha sido decisiva para que este procedimiento sea preciso, simple y fiable. Utilizando esta campaña experimental, se ha desarrollado una aplicación predictiva para estimar los parámetros que definen el comportamiento constitutivo a tracción del UHPFRC. Esta aplicación es simple y directa y evita la posible variabilidad producida por malas interpretaciones en la aplicación del proceso. Además, se ha llevado a cabo una segunda campaña experimental constituida por vigas de UHPFRC armadas a flexión con diferentes escalas: 36 vigas cortas con 130 y 160kg/m3 de fibras y dos vigas largas. Esta campaña experimental se ha modelado con el NLFEM aquí desarrollado teniendo en cuenta efectos importantes debidos a la interacción del UHPFRC con las barras de armado. También se han modelado con el NLFEM tirantes de UHPFRC armados de una campaña experimental de otra investigación. El modelo considera efectos debidos a la retracción, al 3D y comportamiento tensión stiffening que generan resultados muy precisos cuando se comparan con los resultados experimentales. Como resultado de la presente tesis doctoral, se ha obtenido un modelo de elementos finitos capaz de modelar con precisión elementos estructurales de UHPFRC armados. Los resultados no sólo demuestran la fiabilidad del NLFEM llevado a cabo sino también la coherencia del procedimiento desarrollado para caracterizar el comportamiento constitutivo a tracción del UHPFRC para los dos casos, tanto SH-UHPFRC como SS-UHPFRC, tanto en elementos estructurales armados a flexión como en elementos estructurales armados a tracción directa. Consecuentemente se ha propuesto una metodología completa y efectiva para el modelado numérico del UHPFRC
[CA] El principal objectiu de la present tesi es el desenvolupament d'una completa metodologia per al modelat numèric de l'UHPFRC des del nivell material fins arribar als elements estructurals. Es pretén contribuir a l'avanç del coneixement del comportament mecànic de l'UHPFRC per mitjà d'un procediment per al modelat numèric útil per al modelat i disseny estructural que permeta fer que aquest material siga competitiu al mercat de la construcció. En la metodologia de modelat proposta, es considera un comportament constitutiu de l'UHPFRC optimitzat per mitjà d'un procediment directe i fiable amb el qual s'aprofiten els avantatges del material, resultant en un disseny estructural eficient des del punt de vista mecànic i econòmic. És necessari produir SH-UHPFRC per a aconseguir grans propietats mecàniques? És possible generar SS-UHPFRC amb el qual queden reduïts els costs inicials mantenint unes propietats mecàniques i de durabilitat competitives que comporten un disseny estructural efectiu? El desenvolupament d'UHPFRC amb baix enduriment per deformació i de SS-UHPFRC pot reduir les seues propietats mecàniques però, si són adequadament estudiades i controlades, aquests podrien ser optimitzats. La tesi aborda algunes d'aquestes qüestions per mitjà de l'estudi del comportament a tracció de l'UHPFRC que va des de SH-UHPFRC fins SS-UHPFRC. Es pretén dur a terme una proposta de procediment fiable per a caracteritzar el comportament constitutiu a tracció i definir un model numèric d'elements finits fiable per a modelar amb precisió la resposta de provetes i elements estructurals armats d'UHPFRC. Per a definir el procediment directe per a caracteritzar a tracció tant SH-UHPFRC com SS-UHPFRC, s'ha dut a terme una campanya experimental i numèrica en la que s'ha analitzat el resultat d'assajar 227 provetes sense armadura fabricades amb UHPFRC amb quantitats de fibres curtes i llises d'acer de 120-130kg/m3 i 160kg/m3, assajades a flexió per mitjà de l'assaig a quatre punts (4PBT). El desenvolupament i la validació de l'esmentat procés són assegurats per mitjà d'un model no lineal d'elements finits (NLFEM) fiable. La validació numèrica duta a terme ha estat decisiva per a que aquest procediment siga precís, simple i fiable. Utilitzant aquesta campanya experimental, s'ha desenvolupat una aplicació predictiva per a estimar els paràmetres que defineixen el comportament constitutiu a tracció de l'UHPFRC. Aquesta aplicació és simple i directa i evita la possible variabilitat produïda per males interpretacions en l'aplicació del procés. A més a més, també s'ha dut a terme una segon campanya experimental constituïda per bigues d'UHPFRC armades a flexió amb diferents escales: 36 bigues curtes amb 130 i 160kg/m3 de fibres i dos bigues llargues de gran escala. Aquesta campanya s'ha modelat amb el NLFEM ací desenvolupat incloent efectes importants deguts a la interacció de l'UHPFRC amb les barres d'armat. Addicionalment, també s'han modelat amb el NLFEM tirants d'UHPFRC armats a tracció provinents d'una campanya experimental d'altra investigació. El model considera efectes deguts a la retracció, al 3D i comportament tensió stiffening que generen resultats molt precisos quan es comparen amb els resultats experimentals. Per tant, com a resultat de la present tesi doctoral, s'ha obtingut un model d'elements finits capaç de modelar amb precisió elements estructurals d'UHPFRC armats. Els resultats del model comparats amb els resultats experimentals no sols demostren la fiabilitat del NLFEM dut a terme sinó que també la coherència del procediment directe desenvolupat per a caracteritzar el comportament constitutiu a tracció de l'UHPFRC als dos casos, tant per a SH-UHPFRC com SS-UHPFRC, tant en elements estructurals armats a flexió com amb elements estructurals armats a tracció directa. Conseqüentment, s'ha proposat una metodologia completa i efectiva per al modelat numèric de l'UHPFRC des del niv
[EN] The main objective of the present PhD thesis is to develop a complete methodology for the numerical modelling of UHPFRC from the material level to structural elements. It intends to contribute to advanced knowledge of mechanical UHPFRC behaviour to lead to a numerically modelling proposal that is useful for structural modelling and design that allows options for this material to be competitive in the construction market. Optimised UHPFRC material constitutive behaviour, characterised by a direct reliable defined procedure, is considered in the proposed modelling methodology to take advantage of these properties, and to lead to an efficient structural design from the mechanical and economical points of view. Is it necessary to produce SH-UHPFRC to obtain excellent properties? Is it possible to develop SS-UHPFRC that leads to lower initial costs and to maintain competitive mechanical and durability properties that result in an effective structural design? The development of low strain-hardening and SS-UHPFRC would lead to reduce its mechanical properties, but they can be optimised if they are studied and controlled. The thesis addresses some of these questions by studying tensile UHPFRC behaviour to cover a wide range of tensile constitutive behaviours from SH-UHPFRC to SS-UHPFRC. It intends to propose a reliable tensile characterisation process and a reliable finite element model capable of accurately simulating the response of UHPFRC specimens and reinforced structural elements. An extensive experimental and numerical campaign with 227 unreinforced four-point bending test (4PBT) specimens with amounts of smooth-straight (13/0.20) steel fibres of 1.53-1.66% (120-130kg/m3) in volume and with 2.00% (160kg/m3), which represents SS-UHPFRC and SH-UHPFRC tensile behaviours, was carried out to set up a direct tensile characterisation procedure involving SS-UHPFRC and SH-UHPFRC. The direct procedure's development and validity are ensured by a reliable non-linear finite element model (NLFEM). Numerical validation was carried out and is decisive for performing the direct procedure to characterise the tensile behaviour of both SS and SH-UHPFRC herein developed accurately, simply and reliably. With the experimental programme herein, a predictive application for estimating tensile UHPFRC parameters was developed. The prediction offers reliable results. The application is simple and direct, and avoids variability in the characterisation procedure due to possible misinterpretations in its application. In addition, a second experimental programme, which includes reinforced concrete flexural beams on different scales, with 36 UHPFRC reinforced short beams with 130 and 160kg/m3 of steel fibres and two full-scale long beams, was carried out and modelled with the NLFEM herein developed including major effects due to the interaction between UHPFRC and reinforcement bars. Additionally, reinforced UHPFRC tensile bars from a recent experimental campaign performed by other researchers were modelled with the NLFEM. The model considers shrinkage effects, tension stiffening behaviour and 3D effects due to the particularities of the test, which provide very accurate results compared to those obtained with the experimental tests. As a result of this PhD thesis, an accurate NLFEM was obtained to model reinforced UHPFRC structural elements. The results of the model compared to the experimental ones demonstrate not only the reliability of the developed NLFEM, but also the coherence of the developed direct procedure to characterise tensile UHPFRC behaviour in both strain-softening and strain-hardening in reinforced flexural and direct tensile structural elements. Consequently, a complete and effective methodology for numerical UHPFRC modelling from the material level to structural elements is proposed.
Mezquida Alcaraz, EJ. (2021). Numerical Modelling of UHPFRC: from the Material to the Structural Element [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/167017
TESIS