La mer de Marmara est la clef de voûte d’un système hydrique complexe entre la Méditerranée la mer Noire où l’équilibre des courants a souvent été rompu, lorsque le niveau global des océans passait sous celui des détroits. L’alternance de sédiments lacustres et marins témoigne de ces variations. La mer de Marmara est aussi une structure tectonique active traversée par la branche nord de la faille Nord-Anatolienne. Cette faille décrochante continentale, l’une des plus longues du monde, est régulièrement à l’origine de forts tremblements de terre. Les études scientifiques sont nombreuses à vouloir caractériser le fonctionnement actuel et passé de la faille pour préciser l’aléa sismique auquel Istanbul est exposé. Cette thèse est constituée de deux études qui s’attaquent à cette question.La première est une étude de la géométrie de la croûte basée sur l’analyse d’un modèle 3D construit par inversion de données gravimétriques corrigées de l’influence des bassins sédimentaires. Le modèle permet de quantifier le rôle des contraintes extensives et révèle le rôle de flux ductiles dans l’amincissement de la croûte inférieur étalé au-delà des limites des bassins. La deuxième étude visait à confirmer les modèles d’âges proposés actuellement en mer de Marmara par l’étude direct des sédiments à l’origine des premiers réflecteurs sismiques interprétés. Les méthodes employées vont de la géophysique à la caractérisation paléo-environnementale des dépôts en passant par le paléomagnétisme et la téphrochronologie. Les résultats ont permis de dater précisément le premier réflecteur ainsi que d’en comprendre l’origine. L’âge que nous proposons est plus jeune que ceux déterminés jusqu’alors. / The Marmara Sea is the keystone of a complex hydrological system between the Mediterranean and the Black Sea. There, the balance between inputs and outputs has been disturbed during glacial periods when the global ocean level dropped below the sill depths, isolating the Marmara Sea. The alternation of marine and lacustrine sediments reflects these variations. The Sea of Marmara is also an active tectonic structure cut by the North-Anatolian fault northern branch. The ruptures of this continental dextral transform fault, one of the longest in the world, are regularly causing massive earthquakes. Many scientific studies are aiming at characterizing the fault structure and its functioning in the hope of being able to foresee the next earthquake close to Istanbul. This thesis presents two studies tackling this subject.The first one presents an interpretation of the crust geometry based on the analysis of a 3D crustal model built from the inversion of gravity data corrected from the influence of sedimentary basins. From the model, we quantify the role of extension in the basin opening and show the existence of ductile flows, in the lower crust, that spread the thinning away from the basins. The second study purpose was to confirm or reverse the sedimentary age models available in the Marmara Sea through the direct study of the sediments constituting the first main reflector interpreted in the models. We used a broad range of methods to build a new age model for one core such as geophysics, paleo-magnetism, tephrochronology and a paleo-environmental characterization. The results give a date for the first main reflector that is younger than the one assessed by previous studies.

Tsunamis, as the catastrophic disasters, can cause loss of live and property when they come to the shores. Preparation of emergency plans is essential to reduce the damage. Consequently, any initiative in tsunami modeling and inundation mapping is of vital importance for progressing safety surveillance and maintenance. In an effort to achieve a thorough analysis of effect of tsunami, it is critical to estimate the geographical extent of possibly affected area and to predict tsunami impacts. The inundation mapping system also must serve to manage the simulation data in a scalable environment to reach end-users in the time of event. For this purpose, in this study, the generation of a Web based Geographic Information System (GIS) to serve inundation maps through web. The research methodology consists of four main stages: (i) simulating tsunamis based on six different scenarios (ii) processing simulation data through a GIS application / (iii) development of web interfaces and implementation of the developed model for Web-GIS application / (iv) verification of the created model for Marmara Sea Region. The proposed system is expected to be an efficient tool for improving inundation mapping efforts for expected tsunamis in Turkey.

<p>Magnetic surveys are one of the principal geophysical methods employed to map the structure of basement rocks deeply buried below cover strata. In resource studies, aeromagnetic surveys are commonly acquired at regional scales (100-1000’s km2) while very few studies have attempted to resolve basement structures at site-scale (<10 >km2). In this study, high-resolution lake-based magnetic survey methods were evaluated for mapping of deeply-buried basement faults (Küçükçekmece Lagoon, Turkey; 6 km2) and a suspected meteorite impact crater (Charity Shoal, Lake Ontario; 9 km2). Total magnetic intensity (TMI) surveys were acquired using a single Overhauser magnetometer with 50-150 m line spacing. Interpretation of the magnetic data was aided by forward modelling of TMI data and depth to basement estimates using Euler and analytic signal methods. Total magnetic intensity (TMI) maps of Küçükçekmece Lagoon identify several north-northwest trending (340-350°) magnetic lineaments that are aligned with strike-slip faults mapped from offshore seismic data. The pattern of magnetic anomalies in the lagoon is consistent with extensional normal faulting of bedrock and lake sediments. Magnetic results from Charity Shoal reveal a large (>1400 nT) parabolic-shaped magnetic anomaly centered over the crater basin and an annular magnetic high (40-50 nT) corresponding with the basin rim. Modelling results exclude the origin of the CSS as a shallow glacial erosional or karst sinkhole feature and are most consistent with a pre-Paleozoic meteorite impact in the Precambrian basement.</p> / Master of Science (MSc)

Tsunamis are characterized as shallow water waves, with long periods and wavelengths. They occur by a sudden water volume displacement. Earthquake is one of the main reasons of a tsunami development. Historical data for an observation period of 3500 years starting from 1500 B.C. indicates that approximately 100 tsunamis occurred in the seas neighboring Turkey. Historical earthquake and tsunami data were collected and used to develop two artificial neural network models to forecast tsunami characteristics for future occurrences and to estimate the tsunami return period. Artificial Neural Network (ANN) is a system simulating the human brain learning and thinking behavior by experiencing measured or observed data. A set of artificial neural network is used to estimate the future earthquakes that may create a tsunami and their magnitudes. A second set is designed for the estimation of tsunami inundation with relation with the tsunami intensity, the earthquake depth and the earthquake magnitude that are predicted by the first set of neural networks. In the case study, Marmara and Aegean regions are taken into consideration for the estimation process. Return periods including the last occurred earthquake in the Turkish seas, which was the izmit (Kocaeli) Earthquake in 1999, were utilized together with the average earthquake depths calculated for Marmara and Aegean regions for the prediction of the earthquake magnitude that may create a tsunami in the stated regions for various return periods of 1-100 years starting from the year of 2004. The obtained earthquake magnitudes were used together with tsunami intensities and earthquake depth to forecast tsunami wave height at the coast. It is concluded that, Neural Networks predictions were a satisfactory first step to implement earthquake parameters such as depth and magnitude, for the average tsunami height on the shore calculations.

Ce travail de thèse est divisé en deux parties. La première porte sur l'acquisition d'un nouvel enregistrement paléosismologique en Mer de Marmara. Différentes méthodes ont été utilisées afin d'identifier et de caractériser les perturbations sédimentaires associées aux tremblements de terre dans les carottes étudiées. Nous avons cherché à acquérir une compréhension spatio-temporelle des distributions des séismes en mer pour aboutir à une meilleure compréhension du comportement de la Faille Nord Anatolienne sur le long terme. Nous avons par ailleurs cherché à corréler les données de sismicité historique avec de nouvelles données paléosismologiques. Les événements sédimentaires associés aux séismes ont été caractérisés en combinant l'imagerie aux rayons X, des mesures de susceptibilité magnétique, de granulométrie et de composition géochimique. Les données des compositions élémentaires nous ont permis de tracer au travers des différents bassins les changements environnementaux et anthropiques ayant lieu dans la région. L’obtention d’une chronologie robuste dans les carottes a également été recherchée en combinant des datations carbone 14 et des données de 210Pb et 137Cs afin de relier les sismoturbidites à la sismicité historique. Les variations d'aimantation au travers des carottes ont été mesurées dans le but initial de contraindre par une méthode indépendante l'âge des sédiments échantillonnés. Les variations des propriétés magnétiques ont mis en évidence une chute d'aimantation importante que nous avons reliée à des dépôts sapropéliques.La deuxième partie porte sur l’étude géomorphologique du décalage de la rivière Kizilirmak le long du segment central de la Faille Nord Anatolienne. À cet endroit, trois terrasses préservées le long de deux bassins en pull-apart incisés par la rivière Kizilirmak ont été cartographiées. Les résultats principaux de cette étude ont été de contraindre par la méthode des isotopes cosmogéniques 10Be, 26Al et 36Cl l'âge de ces terrasses. Les datations montrent que la terrasse la plus basse est âgée de 6 ka, que la deuxième terrasses est âgée de 50 ka et que la troisième a un âge de 80 ka. Cette dernière montre une contribution importante d'âges jeunes liée à l'érosion du bassin versant situé au-dessus d'elle. Les résultats montrent une origine climatique des terrasses et ont permis d'estimer une vitesse d'incision de la rivière de l'ordre de 3 mm/an depuis le début de l'Holocène. / This PhD work is divided in two parts. The first part focuses on obtaining a new paleoseismological record of earthquakes in the Marmara Sea (West of Turkey) using different analysis to pinpoint and characterize earthquake-related sedimentary disturbance in the studied cores. We seek to provide a spatio-temporal understanding of earthquakes in the Marmara Sea allowing greater insight into long-term fault behaviour and seismic interaction by integrating historical and new paleosismological data (recurrence rate in the Sea of Marmara). We characterized earthquake-related sedimentary events by combining X-ray imagery, magnetic susceptibility, granulometry and XRF (chemical analyses) measurements. Geochemistry data also allowed us to trace between basin anthropogenic and environmental changes occurring in the Marmara Sea. We also aim at establishing a reliable and robust chronology of the cores combining radiocarbon dating (bulk sedimentation, foraminifers, shelves), 210Pb and 137Cs data to connect seismoturbidites to historical seismicity. Magnetic variations were also recorded in the cores to obtain a time constrain for the cores. The down-core changes in magnetic properties suggest taking into account possible delays in the acquisition of magnetization and have been correlated to lower sapropelic layers in the Marmara Sea.The second part deals with geomorphology and tectonic in the central part of the North Anatolian Fault situated in Turkey. There, three terraces preserved in two pull-apart and incised by the longest river in Turkey (the Kizilirmak River) are mapped. The main results of this work are several constraints on the ages of the terraces using 10Be, 26Al and 36Cl cosmogenic dating methods. The obtained in situ cosmogenic 36Cl exposure ages calculated are 6 ka for the lowest terrace, 50ka for the middle terrace, and 80ka for the highest terrace in the areas preserved from erosion. The highest terrace shows a contribution of younger ages due to erosion of the nearby limestone catchment. The obtained results imply a climatic origin of the terraces, and a mean incision rate of about 3 mm/yr since the early Holocene along the Kizilirmak River.