Th. doct.--Géodynamique--Lyon 1, 2000. / Bibliogr. p. 301-313. Résumé en français et en anglais.

Thèse de doctorat--Géodynamique--Montpellier 2, 2005. / Bibliogr. Titre provenant de l'étiquette du support.

Thèse de doctorat--Sciences de la terre--Rennes 1, 2007. / Notes bibliogr. Résumé en français et en anglais.

Reproduction de : Thèse de doctorat : Magmatologie expérimentale : Clermont-Ferrand 2 : 2005. / Thèse avec annexes. Titre provenant de l'écran-titre. Bibliogr. p. 252-277.

Thesis (Ph. D.)--Lehigh University, 2000. / Includes vita. Includes bibliographical references (leaves 152-153).

No description available.

The high abundance of fluid-mobile elements (FMEs) in arc magmas relative to the mantle attests to their recycling in subduction zones. Aqueous fluids released from the subducting slab transport these elements to the overlying mantle wedge for arc magma generation. However, some FME may be retained in the slab to the deeper mantle, incorporating into melts that form back-arc basalts, intraplate ocean island basalts (OIBs), carbonatites and kimberlites. This thesis reports the abundance and distribution of halogens (F, Cl, Br, I) and other FME (e.g., B, N) in various subducted lithologies to evaluate the behaviour of these elements during subduction, their partitioning among coexisting high pressure (HP) minerals and the role of these rocks and minerals in transporting these elements to the deeper mantle. Abyssal lizardite-serpentinites, such as those from the Dominican Republic Tertiary subduction complex, are enriched in halogens and B from hydration by seawater and sediment-modified fluids. During the subduction and subsequent lizardite-antigorite transition of serpentinites, F and B are retained or replenished, while Cl, Br and I are expelled. Halogen compositions of shallow (<25 km) forearc mantle lizardite-serpentinites from the Dominican Republic suggest their hydration by fluids released from subducting altered oceanic crust, with only minor sediment contribution. This is in contrast with deep mantle wedge antigorite-serpentinites associated with the Tso Morari UHP unit, NW Himalaya where elevated F (avg. 185 ppm), Br (avg. 0.13 ppm), I (avg. 0.16 ppm), B (avg. 51 ppm) and N (avg. 23 ppm) contents and their ratios demonstrate significant contributions from subducted shallow water sediments. The latter results are significant because they demonstrate serpentinites can host FME to depths of at least 100 km in the mantle wedge, and possibly to 200 km (limit of antigorite stability) in the slab of cold subduction zones or to 300 km in phase A after antigorite decomposition. A similar decoupling of F from Cl during subduction is also observed in the metamorphosed crust. Blueschists from the Tavsanli zone, NW Turkey and eclogites from the Tso Morari UHP unit and the Sulu UHP belt, China contain low bulk Cl and Br contents relative to their presumed protoliths, where as their F contents are similar or elevated. Iodine is also enriched in the blueschists and Tso Morari eclogites, but depleted in the Sulu eclogites relative to their protolith rocks. Enrichments of F and I in the metamorphosed crust are attributed to fluids released from subducting marine sediments, suggesting geochemical exchange among subducting lithologies. Shallow water trench sediments would be especially enriched in these elements given the close proximity of continents to these subduction zones. Apatite and phengite are the major host of halogens in these rocks as demonstrated by high F in these minerals; blueschists contain up to 3.5 wt% and 500 ppm F, respectively, and eclogites contain up to 2.3 wt% and 1600 ppm F, respectively. Other halogen-bearing minerals include lawsonite, Na-amphibole +/- chlorite in the blueschists and possibly titanite +/- Na-amphibole in the eclogites. Liberated F may be subsequently incorporated into antigorite-serpentinites of the overlying mantle wedge. Bulk B contents are generally low in the subducted crust, however phengite can host up to 100 ppm B to at least blueschists facies conditions. Overall, Cl, Br and I are lost from the altered oceanic crust and lizardite-serpentinites during shallow (< 50 km) subduction, while F and some B are retained. Given the stability of minerals such as antigorite, phase A (after antigorite decomposition), apatite, lawsonite and phengite to depths of 200-300 km in cold subduction zones and their accommodation of FME, they may transport significant F, some B and N and minor Cl, Br and I beyond subarc depths (~ 100 km) and contribute to the their abundances in magmas derived from the deeper mantle.

Thesis (M.Sc.) - Carleton University, 2002. / Includes bibliographical references (p. 170-177). Also available in electronic format on the Internet.

Ce travail vise à mieux comprendre les interactions entre couplage, glissement lent et rupture sismique en contexte de subduction. L’objet d’étude est la subduction Nord Chili–Sud Pérou, qui a été reconnue comme une lacune sismique et qui a fait l’objet d’un important effort international d’instrumentation géophysique. Cette zone a été affectée par tremblements de terre qui ont été bien enregistrés, ce qui en fait une cible bien adaptée pour étudier les mécanismes de préparation des grands séismes de subduction, et le lien entre couplage, glissements lents et rupture sismique.Les 65 stations GPS installées dans la zone ont été traitées en doubles différences avec le logiciel GAMIT-GLOBK sur la période 2000-2014. Les séries temporelles de position obtenues ont été analysées et les déplacements associés aux différentes phases du cycle sismique et aux mouvements saisonniers ont été modélisés.L’analyse des tendances dans les séries temporelles GPS ont permis de mettre en évidence un changement de vitesse intersismique avant et après le séisme en slab-pull de Tarapacà de juin 2005, dans la région qui a été rompue par le séisme d’interface d’Iquique en 2014. Ce changement de vitesse est associé à un changement de taux de sismicité superficielle et profonde. Le déclustering du catalogue sismique indique que ce changement de taux, affecte aussi la sismicité de fond, caractéristique du taux de chargement. Nous avons pu mettre en évidence que des interactions existent entre sismicité profonde et superficielle, et pourraient jouer un rôle important dans la préparation des grands séismes d’interface. A plus courte échelle de temps, les séries temporelles montre un autre changement de vitesse 8 mois avant le séisme d’Iquique. La modélisation indique que ce changement correspond à un glissement lent de Mw 6.5, essentiellement asismique, correspondant à la phase de nucléation long-terme de ce tremblement de terre.Nous avons également analysé les variations court terme dans les signaux non-modélisés des séries temporelles GPS, qui ont permis d’identifier, grâce à une méthode de Template Matching 48 petits événements de glissement lent pendant la période de chargement intersismique. Ces événements sont basés pour leur grande majorité en dessous de la zone sismogénique dans des zones de couplage très faible, ou de couplage intermédiaire, indiquant que le glissement se fait par relâchements successifs. Souvent ces événements transitoires sont corrélés à des pics d’activité sismique, notamment profonde.La cinématique de la source du séisme d’Iquique et de sa plus grosse réplique a été étudiée en combinant données GPS à haute fréquence et données accélérométriques. Les déplacements statiques ont été inversés pour caractériser la répartition du moment géométrique. Cette source statique a ensuite été utilisée comme prior pour une séquence d’inversions cinématiques en fréquence. Les résultats montre que ces deux séismes présentent une distribution de glissement bimodale, segmentée selon la profondeur. L’extension latérale du choc principal correspond à celle d’un bassin submergé d’avant arc associé à une anomalie de gravité, et pourrait être contrôlée par les structures tectoniques de la croûte supérieure. La réplique principale est située dans une zone de changement de contrainte de Coulomb induit par le choc principal, ce qui suggère qu’elle a été déclenchée par celui ci.Grâce à une combinaison de données géodésiques et sismologiques, ce travail offre donc une vue détaillée des processus en jeu au cours du séisme d’Iquique et des décades qui l’ont précédé. Les perspectives de recherches soulevées par ce travail sont nombreuses, notamment sur les possibilités d’observations raffinées des phénomènes associés au cycle sismique et à la préparation des grands tremblements de terre de subduction. / The aim of this Ph.D. thesis is to have a better comprehension of the interactions between coupling, slow slip events (SSEs) and the seismic ruptures in subduction zones. This work focuses on the North Chile - South Peru subduction zone that has been recognized as a mature seismic gap. Thus, the region has been the target of an important international effort in geophysical instrumentation (GPS and seismological stations), since the mid-2000s. The region has been affected by several well-registered earthquakes, that makes it good case to study the earthquake preparation phase and the relationship between coupling, SSE and seismic rupture.The 65 stations available in the region have been processed in double differences using the GAMIT-GLOBK software in the period 2000-2014. The GPS displacement time series have been analyzed and the associated displacements to the different stages of the seismic cycle as well as seasonal signals have been modeled.The analysis of the tendencies in the GPS time-series evidences a change in the velocity field before and after the Tarapaca slab-pull earthquake occurrence in June 2005, in the range of latitude affected by Iquique earthquake in 2014. This velocity change is associated with a change in the shallow and deep seismicity rates. The analysis of the declustered catalog shows that the velocity change observed affects the background seismicity as well, that often seen as a proxy for the tectonic loading. Finally, we find interactions between shallow and deep seismicity, that may play an important role in the interface earthquakes preparation phase. At a shorter time scale, the time series show another change in velocity 8 months before Iquique earthquake. Models indicate that the velocity change corresponds to an SSE Mw 6.5, mainly aseismic, corresponding to the long preparation phase of the earthquake.The short-term GPS velocity variations have also been analyzed on residual signal. It allows identifying, thanks to a matched-filter, 48 small SSEs during the interseismic period. These events are localized mostly in the deeper part of the seismogenic zone, in areas where the coupling is low or with intermediate values. It suggests that the slip in these regions is produced in burst way. Some of those events are correlated with peaks of seismic activity, especially at intermediate depths.Finally, the kinematic rupture process of Iquique earthquake process and its biggest aftershock are studied, employing a combination of high-rate GPS and strong motion data. The static displacements are inverted to characterize the slip. This static solution is used as apriori information for a sequence of kinematic inversions in the frequency domain. The results show that both events have a bimodal slip distribution, segmented along dip. The lateral extension of the mainshock is centered on a forearc offshore basin associated with a gravity anomaly, and may be controlled by tectonic structures of the upper crust. The aftershock is located in an area with strong Coulomb Stress Changes induced by the mainshock, suggesting that it was triggered by the mainshock.Thanks to the combination of geodetic and seismological data, this thesis provides a detailed vision of the processes involved during the Iquique earthquake and the previous decades. The research prospects raised by this work are numerous, particularly on the possibilities of refined observations of the phenomena associated with the seismic cycle and the preparation of large subduction earthquakes.

The Betic Cordillera of southern Spain is an orogen formed in response to convergence between Africa and Iberia, from the late Mesozoic to the present. The orogen consists of three main tectonic complexes, two of which have been subducted to depth, then exhumed back to the surface over short timescales. Subduction in the structurally higher of these complexes is relatively well constrained to the Eocene, but the timing of high-pressure metamorphism in the structurally lower complex, known as the Nevado-Filabride Complex, has been a topic of debate for several years due to conflicting geochronological data. Several proposed tectonic models for the Nevado-Filabride Complex are based on ages of single mineral phases. For example, models based primarily on 40Ar/ 39Ar dating on white mica in high-pressure schists require that the Nevado-Filabride and the overlying tectonic unit, the Alpujarride Complex, were coevally subjected to high-pressure metamorphism in the Eocene, and subsequently exhumed at different rates. More recent models, based on Lu-Hf dating on prograde garnets in eclogites, separate the timing of high-pressure metamorphism of the Nevado-Filabride Complex from the Alpujarride Complex by at least 10 m.y. We examine the viability of these models using multimineral Rb-Sr dating of blueschist and eclogite facies rocks in the Nevado-Filabride Complex. The multimineral isochron method uses the whole high-pressure mineral assemblage rather than a single phase, which allows testing for isotopic disequilibrium. Statistically valid Rb-Sr ages of two schists and one eclogite from the Nevado-Filabride Complex yield ages of 15.78+/-0.47, 15.8+/-1.1, and 17.6+/-1.1 Ma, respectively. The early Miocene Rb-Sr ages are in agreement with garnet Lu-Hf ages and zircon U-Pb ages for high-pressure conditions in the Nevado-Filabride Complex. The new ages imply that two episodes of subduction, punctuated by a period of extension and exhumation, occurred in the Western Mediterranean. / text