Le système volcanique de La Réunion est formé de la coalescence des massifs du Piton des Neiges et du Piton de la Fournaise. Son édification, liée à l’activité d’un panache mantellique, est caractérisée par une phase de croissance sous-marine, puis subaérienne, suivie d’une période de dégénérescence, comme proposé classiquement pour les volcans boucliers océaniques d’Hawaii. De précédentes études ont montré que l’activité subaérienne de La Réunion aurait débuté il y a environ 2,2 Ma. Les analyses géochimiques menées jusqu’alors sur les produits associés à cette activité volcanique (subaérienne et sous-marine) ont montré un caractère chimique transitionnel avec une composition isotopique (87Sr/86Sr, 143Nd/144Nd et 176Hf/177Hf) particulièrement homogène pour un volcanisme de point chaud. Dans cette étude, nous montrons que des laves sous-marines récemment identifiées sous la série basique du Piton de la Fournaise (au sein de la sur Rift Zone Nord-Est) ainsi que des laves subaériennes affleurant à la base de canyons incisés dans le massif (Série Différenciée de la Rivière des Remparts), présentent des caractéristiques ne s’inscrivant pas dans le modèle d’évolution proposé précédemment pour ce volcan. Se pose donc la question de la nature et de l’origine de ces laves, ainsi que de leur place et appartenance dans l’édification du système volcanique réunionnais. De nouvelles investigations pétro-géochimiques et géochronologiques ont été menées sur la partie sous-marine de l’édifice de La Réunion. Elles révèlent l’existence d’un groupe de laves exceptionnel. Leur composition se distingue clairement de celle des échantillons communément analysés à La Réunion et montre (1) un enrichissement en éléments incompatibles couplé isotopiquement à (2) un rapprochement vers un pôle mantellique enrichi de type EM. Ces particularités géochimiques démontrent que le panache mantellique possède (1) une source hétérogène affectée par (2) des taux de fusion variables. De plus, deux échantillons de ce nouveau groupe datés à 3,77 (0,08) et à 3,34 (0,07) Ma étendent considérablement la période d’activité connue de l’édifice. Ces âges remarquables font de ces laves les plus anciennes jamais datées à La Réunion. La nature géochimique ainsi que la position stratigraphique des laves de cette étude (échantillons subaériens et sous-marins des massifs du Piton des Neiges et du Piton de la Fournaise) impliquent une réinterprétation du schéma d’évolution global de l’île. Contrairement aux études précédentes, nos résultats sur ces laves montrent que : (1) Les laves différenciées de la Rivière des Remparts, qui sont des laves subaériennes, de part leur position stratigraphique et géographique sous-jacente au Piton de la Fournaise et par analogie avec le Piton des Neiges, ne peuvent être reliées à l’activité de la Fournaise, (2) Le signal isotopique du groupe de laves sous-marines de la Rift Zone Nord-Est de la Fournaise, révèlent l’hétérogénéité de la source du panache mantellique sous La Réunion, source que nous identifions comme des enclaves d’éclogite contenues dans une lherzolite à spinelle, (3) L’âge plus ancien de ces laves sous-marines et leur localisation sous le flanc Est de la Fournaise remet en questions le schéma jusqu’alors admis pour l’édification de l’île de La Réunion reposant sur la construction du Piton des Neiges puis celle, adjacente, du Piton de la Fournaise. En conséquence, nos données géochimiques et géochronologiques sont de nouveaux arguments en faveur de l’élaboration d’un modèle d’évolution plus complexe, suggérant l’existence d’un troisième massif volcanique. Il est en accord avec les précédents travaux de pétrologie et de géophysique suggérant l’existence d’un troisième édifice à l’Est de La Réunion, communément appelé le Volcan des Alizés. / The volcanic system of La Reunion is made of the coalescent Piton des Neiges and Piton de la Fournaise edifices. Its formation, associated with the activity of a mantle plume, is characterized by phases of submarine and subaerial growth, followed by a period of destruction, as described in the classical model proposed for Hawaiian Island volcanoes. Previous studies showed that the subaerial activity of La Reunion would have started around 2.2 Ma ago. Geochemical analyses carried out on the products associated with subaerial and submarine volcanism revealed their transitional chemical nature and their peculiar homogeneous isotopic compositions (87Sr/86Sr, 143Nd/144Nd et 176Hf/177Hf) for a hotspot derived magmatism. In this study, we show that some unusual submarine lavas were recently identified below the basal series of Piton de la Fournaise volcano (dredged on the North-East Rift Zone of la Fournaise) and below the subaerial lavas outcropping at the base of the differentiated series of Rivière des Remparts, indicate that the evolution of La Reunion system is probably more complex than previously suggested. The question of the nature and origin of these lavas, together with their implication on the formation of La Reunion, is an issue of major interest in order to better constrain the global evolution of the volcanic system. As a consequence, new petrological, geochemical and geochronological investigations were conducted on the submarine part of La Reunion edifice. They revealed the existence of a geochemically exceptional group of lavas. Its composition is clearly different from common samples of La Reunion and presents (1) an enrichment in incompatible elements and (2) tends toward an enriched EM endmember. These chemical specificities reveal that the source of La Reunion magmatic products is (1) an heterogeneous source affected by (2) variable melting degrees. Moreover, two samples of this new group dated at 3.77 (0,08) and 3.34 (0,07) Ma extend considerably the period of activity of the island. These rocks are the oldest samples ever dated at La Reunion. The nature and stratigraphical location of subaerial and submarine samples from Piton des Neiges and Piton de la Fournaise imply a new interpretation of the global evolution of the island. Unlike previous studies, our results indicate that :(1) The differentiated subaerial lavas from Rivière des Remparts, due to their stratigraphical and geographical location underlying Piton de la Fournaise, and by analogy with the Piton des Neiges, cannot be linked to the activity of Piton de la Fournaise. (2) The isotope signature of the submarine lava group from the North-East Rift Zone of Piton de la Fournaise, reveals the heterogeneous character of the mantle source under La Reunion Island, that can be modelled as embedded eclogite in a matrix of spinel lherzolite, (3) The age of this submarine group and its location under the eastern flank of Piton de la Fournaise, imply a more complex model of evolution of La Reunion. Therefore, our geochemical and geochronological data are new arguments suggesting the existence of a third volcanic center on the island. This assumption confirms the previous petrological and geophysical evidence supporting the existence of this volcano at the East of La Reunion, commonly called Les Alizés volcano.

The Earth's climate system is driven by varying insolation from the Sun. The dominant variations in insolation are at 23 and 40 thousand year periods, yet for the past million years the Earth's climate has glacial cycles at approximately 100 kyr periodicity. These cycles are a coupled variation in temperature, ice volume, and atmospheric CO₂. Somehow the Earth system's collective response to 23 and 40 kyr insolation forcing produces 100 kyr glacial-interglacial cycles. Generally it has been assumed that the causative mechanisms are a combination of ice dynamics (high ice reflectivity controlling temperature) and ocean circulation (changing carbon partitioning between the deep ocean and the atmosphere, and heat transport to the poles). However, these proposed mechanisms have not yet resulted in a compelling theory for all three variations, particularly CO₂. This thesis explores the role of volcanic CO₂ emissions in glacial cycles. I calculate that glacial-driven sea level change alters the pressure on mid-ocean ridges (MORs), changing their CO₂ emissions by approximately 10%. This occurs because pressure affects the thermodynamics of melt generation. The delay between sea level change and the consequent change in MOR CO₂ emissions is several tens-of-thousands-of-years, conceptually consistent with a coupled non-linear oscillation that could disrupt glacial cycles from a 40 kyr mode to a multiple of that period. I develop an Earth system model to investigate this possibility, running for approximately one million years and explicitly calculating global temperatures, ice sheet configuration, and CO₂ concentration in the atmosphere. The model is driven by insolation, with all other components varying in response (and according to their own interactions). This model calculates that volcanism is capable of causing a transition to ̃100 kyr glacial cycles, however the required average volcanic CO₂ emissions are barely within the 95% confidence interval. Therefore it is possible for volcanic systems and glacial cycles to form a 100 kyr coupled oscillation.

FundaÃÃo Cearense de Apoio ao Desenvolvimento Cientifico e TecnolÃgico / A RegiÃo Metropolitana de Fortaleza, no Estado do Cearà à composta por uma diversificada paisagem geomorfolÃgica, estruturada e modificada ao longo de milhÃes de anos durante a era geolÃgica. Entre as diversas morfologias, as formas originadas a partir do Ãltimo evento vulcÃnico ocorrido no Nordeste brasileiro hà 30 milhÃes merecem total apreÃo devido sua singularidade morfogenÃtica. Tais relevos tiveram suas gÃneses ligadas a partir da aÃÃo de um hot spot que ocorreu nesta Ãrea e juntos recebem o nome de FormaÃÃo Messejana. Esta formaÃÃo compÃe-se de uma dezena de pequenos relevos que se dispÃe em setores da margem continental ao longo da regiÃo metropolitana de Fortaleza. Dentre estes estÃo os relevos vulcÃnicos do Caruru, do Ancuri e PÃo-de-aÃÃcar. Esta pesquisa tem como objetivo: estabelecer a origem, estruturaÃÃo, caracterizaÃÃo morfolÃgica e ambiental destes relevos, assim como, definir as etapas da evoluÃÃo geomorfolÃgica da paisagem local. Para compreender tais objetivos buscamos se aprofundar na bibliografia sobre o assunto, interpretaÃÃo de material cartogrÃfico de diversas escalas, dados fÃsicos-ambientais com apoio de ferramentas de geoprocessamento e visitas a ÃrgÃos pÃblicos e a campo. Como resultados podemos concluir que os trÃs relevos da FormaÃÃo Messejana analisados neste trabalho apresentam-se em formas de Necks arredondados (Caruru e PÃo-de-AÃÃcar) e elipsoidal (Ancuri) e sÃo compostos por rochas alcalinas. Tais relevos se caracterizam por apresentarem extensÃes e altitudes bem diferentes, vertentes Ãngremes, solos poucos desenvolvidos e vegetaÃÃo de pequeno porte. Por fim, o Caruru, apresenta-se parcialmente destruÃdo devido à extraÃÃo das rochas para a construÃÃo civil, enquanto que no Ancuri esta atividade cessou hà aproximadamente sete anos. / The Metropolitan Area of Fortaleza, capital of the Cearà State, North-east Brazil, is composed by a diversified geomorphologic landscape, which has been structured and modified during geological times. Among this diversity, there are forms originated by the last volcanic event occurred in the Brazilian northeast, 30 millions ago, which create a very singular local morphology. Such prominences were probably originated by the action of a âhot spotâ. The magma extrusion resulting of this action created a geological domain named  Formation Messejana Â. This formation is composed by a set of ten volcanic reliefs, as small prominences disposed in many sectors of the city of Fortaleza coastal area and adjacencies. Among these prominences are the Caruru, Ancuri and PÃo de AÃucar hills. The present research has as objective to establish the origin, the processes of morphological structuration, the morphological characterization and the environmental situation of these volcanic prominences, as well as define the phases of morphological evolution of the local landscape. For this purpose, the following stages have been developed: bibliographical research about the matter, interpretation of cartographical data, with the use of diverse scales and thematic maps and cartographical geoprocessing techniques, and field work. As results, we are able to conclude that the three prominences of the Formation Messejana analyzed in this work present form of dome-like necks (Caruru and PÃo de AÃucar) and ellipsoidal form (Ancuri), being composed by alkaline rocks. Such prominences are characterized for presenting different extensions and altitudes, steep slopes, poorly developed soils and sparse vegetation cover. In the present moment, the Caruru hill is a place of mining activity, being partially destroyed due to the extraction of the rocks for the civil construction. This situation creates environmental problems, which may be extended to other volcanic hills, fact that put in danger of eradication these singular elements of the Cearà geomorphic landscape.

In this collection of papers I have summarized my investigations into the field of evolutionary genetics and more specifically into patterns of biodiversity and evolutionary processes. The lizards (and frogs) studied here share common features in that they are largely present in unique environments, which are also regions that are biologically understudied. Most of these taxa show high degrees of endemism, interesting natural history characteristics, and each group manifests distinctive adaptations of general evolutionary interest. My work in the genus Telmatobius has been a progressive approach that began in my MS program, and it first focused on alpha taxonomy, morphological variation, and species boundaries. This work led to new studies initiated and completed at BYU involving further taxonomic revision (Formas et al., 2003; Chapter 1), and then revisiting and re-evaluating species boundaries established earlier (with allozyme markers) and this time with population level molecular (mitochondrial DNA) markers (Chapter 2). Our results indicate that the striking differences in size, coloration and general appearance in the various Lake Titicaca morphotypes are not genetically based. Further, there is evidence that these morphotypes have evolved very rapidly after demographic bottlenecks eroded present genetic variability. Telmatobius frogs of Lake Titicaca are listed by the International (IUCN) as critically endangered. We support this classification and further suggest studies to explore open questions like the possibility of adaptation along ecological resource gradients. Lizards of the genus Microlophus are interesting but for different reasons, and studies of this group constitutes the bulk of my dissertation work. The genus includes both Galapagos insular species, and continental taxa distributed in a linear gradient along > 4000 km of the western coast of South America. In studying Microlophus I first tackled the unresolved phylogenetic relationships within the genus (Chapter 3) and then pay attention to phylogeographic aspects of the most speciose lizard radiation in the Galapagos Archipelago (Chapter 4). Chapter 3 is a single manuscript provisionally accepted in the journal Systematic Biology. This paper introduces the lizard genus Microlophus (“lava lizards”) as a study system, and includes a large nuclear data set accompanied by an equally large mitochondrial data set (7877 characters in total). This paper explicitly differentiates among sequence alignments of gene regions that vary in tempo and class of mutational events. We show that this recognition is important and we suggest ways to appropriately deal with the alignment of multi-locus non-coding DNA data sets. A secondary finding in this study is that mtDNA and nDNA topologies are discordant with each other but that both are strongly supported, and that the nuclear topology is concordant with species distribution patterns along coastal South America. We hypothesize that in this particular region of the tree, the nuclear genome recovers a topology that is closer to the species tree, and conflicts occur due to likely secondary contact of distantly related taxa, suggesting that unique taxonomic relationships in the mtDNA gene tree are the result of hybridization. This last point highlights the value of dense taxonomic and character sampling for teasing apart different aspects of evolutionary processes. Chapter 4 is a manuscript to be submitted to the journal Evolution; in this study we further investigate the most speciose radiation of Microlophus in the Galapagos, based on an unparalleled sampling of most islands and small islets in the Archipelago. We use mtDNA sequences to both test hypothesized between-island colonization routes, as well as the expectation that within-island phylogeographic structure should be greater on older islands. Our mtDNA gene tree is strongly supported and allows rejection of previous alternatives, and we propose a novel sequence of between-island colonization events. Our results also reject the idea of phylogeographic structure been related solely to island age. Instead, we provide evidence to suggest that active volcanism as a major player in the generation of genetic diversity in within-island environments, and this is further compounded by the seemingly stochastic nature of within-island long-distance colonization routes mediated by ocean currents. We suggest that the direction and intensity of these currents, as currently understood, are insufficient to generate a priori hypotheses of oceanic colonization routes and their influence on gene flow. We do show that the standard stepping-stone model of migration, where genetic interchange is only possible among neighboring localities, does not explain much of the within-island population genetic structure unraveled by this study. From a biological conservation perspective the study of patterns of recent evolutionary history in the Galapagos provides with a window to evolutionary processes that have shaped and continue to impact the generation of biodiversity in the Galapagos Archipelago. Islands have long been viewed as natural laboratories of evolutionary change, and thus all island isolates are or could be distinctly important components of the larger, archipelago-wide processes. We provide working hypotheses for some of the demographic processes that might be generating within- and between-island biodiversity in this clade of lizards; confirmation of these explanations with independent data will have management implications for conserving the unique patterns observed in the Galapagos biota, but also the processes that generated these patterns.

Dans la partie Nord de la zone de subduction des Petites Antilles, la convergence est fortement oblique alors qu’au Sud elle est frontale. Cette étude vise à comprendre les conséquences de l’entrée en subduction du plateau des Bahamas fortement flottant par l’étude de l'évolution du champ de contrainte dans le NE de la plaque Caraïbes, depuis l’Eocène. Ce travail montre que sur les îles les plus anciennes, l’arc volcanique s'initie à l’Eocène. La migration de l’arc vers sa position actuelle se produit durant la période Miocène inférieur – Miocène supérieur. La période Oligo-Miocène est charnière : le champ de contrainte évolue d'une extension pure parallèle à la fosse de subduction, à une extension radiale. Après une restauration de la déformation régionale, j'attribue ce changement d’état de contrainte, à l’initiation du partitionnement de la déformation dans le NE de la plaque Caraïbe qui accommode la courbure de la fosse faisant suite à l’entrée en subduction du banc des Bahamas. Dans l’archipel Guadeloupéen, les analyses structurales à terre et en mer montrent des régimes en extension pure perpendiculaire à la fosse. Les orientations des failles similaires depuis l’Eocène confirment un fort héritage structural. À l’actuel, dans le coin NE des Petites Antilles, le régime tectonique est décrochant avec une direction d’extension principale orientée NO-SE soit parallèle à la fosse. Dans l’archipel Guadeloupéen le régime tectonique est purement extensif orienté NE-SO perpendiculairement à la fosse. Cette rotation est interprétée comme résultant de l’augmentation vers le Nord de l’obliquité de la convergence du fait de la courbure de la zone de subduction. / To the north of the Lesser Antilles subduction zone, from North to South, the obliquity of the subduction is decreasing. This study is focus on the consequences of the entrance of Bahamas bank buoyant plateau into the subduction by studying stress field evolution in the NE of the Caribbean plate since Eocene. This work show that the volcanic arc activity on Anguilla bank islands begin at the Eocene. The volcanic arc migration toward his actual localisation occurs during early Miocene – late Miocene period. The Oligo-Miocene period is transitional as a switch in the stress field from pure parallel-to-the-trench to radial extension occurs. A restoration of the regional deformation shows that this switch is related to strain partitioning initiation in the upper Caribbean Plate in response to trench bending that followed the entrance of the Bahamas Bank in the subduction zone. In the Guadeloupean archipelago, kinematic analyses onshore and offshore show a pure extension with a perpendicular-to-the-trench σ3. The similar faults orientations since Eocene confirm that inherited structures control strain localisation. At present day, in NE corner of the Lesser Antilles, the NW-SE main extensional direction of strike-slip stress regime is trench-parallel. In the Guadeloupean archipelago, the pure extensive stress regime is trench-perpendicular (NE-SW). These different orientations of the stress field are interpreted to be the result of increasing trench bending to the North responsible for a northern increase of subduction obliquity

Abstract not available

The interaction of magma with continental crust at convergent margins is fundamental to understanding if and how continents grow. Isotopic and elemental data constrain the progressive stages of development of the magmatic underpinnings of the long-lived Aucanquilcha Volcanic Cluster (AVC), situated atop the thick continental crust of the central Andes in northern Chile. Whole rock data are used in conjunction with mineral compositions to infer processes that gave rise to eleven million years of intermediate, dominantly dacite, arc volcanism. A pulse of volcanic activity at the AVC between ~5 and 2 Ma is bracketed by more sluggish rates. We document chemical changes in the lavas that accompany this eruptive evolution. Trace element data suggest that crystal fractionation and magma mixing were the dominant mechanisms generating the diversity observed in the AVC whole rock data. Fractionation was dominant during early and waning stages of magmatism, and magma mixing was an important process during the high flux period. Peak thermal maturity of the AVC underpinnings coincided with the high magma flux and likely promoted open system processes during this time. Mineral compositions from zircon, amphibole, pyroxene, and Fe-Ti oxides confirm the importance of material recycling in the production of evolved AVC rocks. Various geothermometers were employed to calculate the pre-eruptive conditions of AVC magma using mineral compositions. Pressure estimates from amphibole and two-pyroxene barometry indicate crystallization depths of 1 ��� 5 kb and 4 ��� 6 kb, respectively. Temperature estimates from zircon, Fe-Ti oxides, amphiboles, and pyroxenes indicate temperatures ranging from ~700��C to 1100��C. Zircon temperatures are always the lowest (700��C - 950��C), and pyroxene temperatures are always the highest (1000��C - 1100��C), with Fe-Ti oxide and amphiboles temperatures falling in between. U-Pb ages from zircons and thermometry from individual samples evidence the thermal maturation and consolidation of the underpinnings below the AVC, presumably culminating in a large, crystal-rich mush zone where magmas were trapped and processed. It is in these middle to upper crustal zones where magmatic diversity is attenuated and giant, relatively homogeneous batholiths are formed. Isotopes of AVC lavas are similar to values observed from other central Andes volcanic centers. Lead isotopes are consistent with the AVC's location within a Pb isotope transition zone between the Antofalla and Arequipa basement terranes. Oxygen and Sr isotopic ratios are high and Nd isotopic ratios low with respect to a depleted mantle. Through time, ������Sr/������Sr values of AVC lavas progressively increase from lows of ~0.70507 to ~0.70579 (upper values of 0.70526 to 0.70680), and ��Nd values decrease from highs of -1.0 to -4.6 (lows of -1.6 to -7.3). Similarly, O isotopes (�������O) show a slight increase in base level through time from lows of 6.5��� to 7.0��� (highs of 6.75��� ��� 7.5���). Dy/Yb and Sm/Yb ratios also increased systematically from highs of 2.11 to 3.45, and 2.76 to 6.67, respectively. Despite the temporal isotopic variation, there is little isotopic variation with indices of fractionation, suggesting this signal is the consequence of deep magmatic processing, here attributed to an expanding zone of melting, assimilation, storage, and homogenization (MASH) of mantle-derived magma in the deep crust. Upward expansion brought the MASH zone into contact with rocks that were increasingly evolved with respect to Sr and Nd isotopes, explaining the isotopic shifts. Downward expansion of the MASH zone enhanced garnet stability during basalt fractionation, explaining the increased Dy/Yb and Sm/Yb ratios. Mass balance calculations involving Sr, Nd, and O isotope modeling are consistent with a crustal component making up 10 - 30% of AVC lavas, implying that although the history of central Andean magmatism is replete with large scale crustal recycling, the current phase is largely a crust formation event. / Graduation date: 2012

Composition and texture of the Middle and Late Eocene volcanic, volcaniclastic, and volcanic-sedimentary rocks in the Bijgerd-Kuh e Kharchin area, in the Uromieh-Dokhtar zone northwest of Saveh, Iran, suggest the complexity of the magmatic system that involved multiple eruptions from one or more sources. Hydrated volcanic fragments in hyaloclastic rocks, and the presence of a sequence of shallow and intermediate-depth marine microfossils, suggest that the Middle Eocene units were erupted in a marine basin. The bimodal volcanism of the Late Eocene is distinguished by the presence of four alternating sequences of hyaloclastite lava and ignimbrite. The REE patterns show spatial homogeneity and temporal heterogeneity in the composition of all the Late Eocene sequences, suggesting origination of magma from varying sources that erupted at different times. The trace element distributions of the hyaloclastites and ignimbrites are compatible with those evolved through fractional crystallization of the lower and upper continental crust, respectively.

This dissertation analyzes the spatial distribution and kinematics of the Late Cenozoic Basin and Range (BR) and cross normal fault (CF) systems and their related graben basins around the Snake River Plain (SRP), and investigates the spatio-temporal patterns of lavas that were erupted by the migrating Yellowstone hotspot along the SRP, applying a diverse set of GIS-based spatial statistical techniques. The spatial distribution patterns of the normal fault systems, revealed by the Ripley's K-function, display clustered patterns that correlate with a high linear density, maximum azimuthal variation, and high box-counting fractal dimensions of the fault traces. The extension direction for normal faulting is determined along the major axis of the fractal dimension anisotropy ellipse measured by the modified Cantor dust method and the minor axis of the autocorrelation anisotropy ellipse measured by Ordinary Kriging, and across the linear directional mean (LDM) of the fault traces. Trajectories of the LDMs for the cross faults around each caldera define asymmetric sub-parabolic patterns similar to the reported parabolic distribution of the epicenters, and indicate sub-elliptical extension about each caldera that may mark the shape of hotspot’s thermal doming that formed each generation of cross faults. The decrease in the spatial density of the CFs as a function of distance from the axis of the track of the hotspot (SRP) also suggests the role of the hotspot for the formation of the cross faults. The parallelism of the trend of the exposures of the graben filling Sixmile Creek Formation with the LDM of their bounding cross faults indicates that the grabens were filled during or after the CF event. The global and local Moran’s I analyses of Neogene lava in each caldera along the SRP reveal a higher spatial autocorrelation and clustering of rhyolitic lava than the coeval basaltic lava in the same caldera. The alignment of the major axis of the standard deviational ellipses of lavas with the trend of the eastern SRP, and the successive spatial overlap of older lavas by progressively younger mafic lava, indicate the migration of the centers of eruption as the hotspot moved to the northeast.

The thermal structure of subduction zones depends on the age-controlled thermal state of the subducting slab and mantle wedge flow. Observations indicate that the shallow part of the forearc mantle wedge is stagnant and the slab-mantle interface is weakened. In this dissertation, the role of the interface strength in controlling mantle wedge flow, thermal structure, and a wide range of subduction zone processes is investigated through two-dimensional finite-element modelling and a global synthesis of geological and geophysical observations. The model reveals that the strong temperature-dependence of the mantle strength always results in full slab-mantle decoupling along the weakened part of the interface and hence complete stagnation of the overlying mantle. The interface immediately downdip of the zone of decoupling is fully coupled, and the overlying mantle is driven to flow at a rate compatible with the subduction rate. The sharpness of the transition from decoupling to coupling depends on the rheology assumed and increases with the nonlinearity of the flow system. This bimodal behaviour of the wedge flow gives rise to a strong thermal contrast between the cold stagnant and hot flowing parts of the mantle wedge. The maximum depth of decoupling (MDD) thus dictates the thermal regime of the forearc. Observed surface heat flow patterns and petrologically and geochemically estimated mantle wedge temperatures beneath the volcanic arc require an MDD of 70-80 km in most, if not all, subduction zones regardless of their thermal regime of the slab. The common MDD of 70-80 km explains the observed systematic variations of the petrologic, seismological, and volcanic processes with the thermal state of the slab and thus explains the rich diversity of subduction zones in a unified fashion. Models for warm-slab subduction zones such as Cascadia and Nankai predict shallow dehydration of the slab beneath the cold stagnant part of the mantle wedge, which provides ample fluid for mantle wedge serpentinization in the forearc but little fluid for melt generation beneath the arc. In contrast, models for colder-slab subduction zones such as NE Japan and Kamchatka predict deeper dehydration, which provides greater fluid supply for melt generation beneath the arc and allows deeper occurrence of intraslab earthquakes but less fluid for forearc mantle wedge serpentinization. The common MDD also explains the intriguing uniform configuration of subduction zones, that is, the volcanic arc always tends to be situated where the slab is at about 100 km depth. The sudden onset of mantle wedge flow downdip of the common MDD overshadows the thermal effect of the slab, and the resultant thermal field and slab dehydration control the location of the volcanic arc. The recognition of the fundamental importance of the MDD has important implications to the study of geodynamics and earthquake hazard in subduction zones.