Studying subduction zone dynamics and continental evolution in South America using data-oriented geodynamic models

Subduction;craton modification;flat slab;seismic anisotropy;mantle flow;geodynamic model;data assimilation;South America

Hu, Jiashun

关键词: Subduction;    craton modification;    flat slab;    seismic anisotropy;    mantle flow;    geodynamic model;    data assimilation;    South America;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/101529/HU-DISSERTATION-2018.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Among all continental plates, South America remains one of the least explored regarding its geological evolution. This continent possesses the longest subduction zone, the second highest continental plateau and tremendous amounts of volcanic eruption on its western coast. On the eastern coast, the Brazilian highlands that reach more than 1 km above the sea level were once submerged by shallow ocean in Mid-Cretaceous. All these magnificent landscapes and the associated dramatic geologic events are potentially linked to the underlying restless mantle convection that has persisted for billions of years.The purpose of my dissertation is to understand the causes and consequences of the major geological events in South America and evaluate the potential role of mantle convection in driving subduction zone dynamics and continental evolution. Yet, these geological events usually have distinct characteristics. Some of them are extreme events that occur within several minutes to days and have great impact on local societies, such as earthquakes and volcanic eruptions; while others may span a long geological period and exert gradual but significant influence in Earth evolution, such as mountain building and surface uplifts.Modern observations have contributed a lot to the study of the tectonics in South America, but they usually have limited spatial or temporal coverage. For example, seismic tomography reveals the present mantle structure which is only a snapshot of Earth's history. Thermochronology can tell us the exhumation history of a rock that may span tens to hundreds of millions of years, but the inference is spatially restricted to the sampling area. While numerical modeling can largely overcome these shortcomings, it is affected by uncertain model parameters. As a result, many tectonic questions remain open, with little consensus from decades' of dedicated research.One promising way to further investigate the tectonics of South America is through the cutting-edge multi-disciplinary research. In this dissertation, we apply numerical modeling with sequential data assimilation that progressively incorporates the paleo-reconstructions of plate kinematics and seafloor ages. The outputs are calibrated by multiple observations, including seismic tomography, earthquake source properties, distribution of volcanisms, seismic anisotropy, topography and gravity, mineral physics, as well as geological data such as mountain shortening and paleo-altimetry. The synergy among all these disciplines not only increases the spatial and temporal coverage of the research topics, but also narrows down the intrinsic uncertainties in each discipline.In practice, we tailor our multi-disciplinary models for two research topics, subduction zone dynamics and continental evolution.For the first topic, we use geodynamic models with plate kinematics and seafloor ages as boundary conditions to reproduce the history of South American subduction since the Late Cretaceous. With this model and the constraints mentioned above, we attempt to investigate the dynamic causes (Chapter 2) and consequences (Chapter 3) of flat subduction, as well as seismic anisotropy caused by subduction-induced mantle flow (Chapter 4). Our model reveals that the flat slabs in South America are caused by the synergy of dynamic suction from the overriding plate and the extra buoyancy provided by subducting oceanic plateau and aseismic ridges. The broken flat slab configuration due to the subduction of aseismic ridges better explains the abnormal distribution of intermediate-depth earthquakes and volcanisms as well as the intra-slab stress patterns than earlier models. This model also suggests that the mantle flow is controlled by the subducting slabs, with Poiseuille flow dominating the sub-continental region and Couette flow dominating the sub-oceanic region. Such a flow pattern best matches the observed patterns of surface wave anisotropy and shear wave splitting.For the second topic, we further combine this well-established geodynamic model with observations from other disciplines, including residual topography, residual gravity, reconstructed hotspot tracks, and surface geology, to study the temporal evolution of the cratonic lithosphere (Chapter 5). We find that lowermost cratonic lithosphere is compositionally denser than the asthenospheric mantle and can be episodically removed when perturbed by underlying mantle dynamics, while the shallower buoyant lithosphere helps to stabilize cratonic crust over billions of years. We further show that zones where the lithosphere was lost would take tens of millions of years to recover thermally, but the density of the new thermal root would remain less than that of the intact root. This new model challenges the traditional view on the density profile and tectonic stability of cratons, and thus have important implications on continental evolution.Overall, this dissertation shows that modern multi-disciplinary research that combines data-assimilation numerical modeling with various geological and geophysical observations, could greatly help us understand the tectonic driving forces in continents like South America.