Understanding late Cenozoic western United States mantle dynamics and surface tectonics using forward-adjoint data assimilation models

mantle convection;western United States geology and tectonics;data assimilation models;intra-plate volcanism;western United States anisotropy

Zhou, Quan

关键词: mantle convection;    western United States geology and tectonics;    data assimilation models;    intra-plate volcanism;    western United States anisotropy;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/101166/ZHOU-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 】

Understanding mantle evolution is essential for investigating many processes of the Earth’s surface and interior. A straightforward way to study mantle evolution is through mantle convection simulation. However, a self-consistent and fully dynamic representation of past mantle convection for the real earth is still not feasible. Different data assimilation methods for mantle simulation therefore have been proposed, all of which have pros and cons. By combining the advantages of the forward and adjoint assimilation algorithms, we created a hybrid data assimilation approach. We evaluate the effectiveness of this new approach by comparing its recovered mantle evolution with results from several other existing methods. We show that the new approach is able to better capture both the fine-scale and large-scale dynamics of subduction and mantle convection.We apply the hybrid approach to the reconstruction of mantle dynamics below the western United States (U.S.) since 20 million years ago (Ma). In our model, the mantle dynamics is only driven by the mantle density structure. The model results reveal a new explanation for the formation of intra-plate volcanism in the western U.S. In particular, the Columbia River flood basalt, the Yellowstone and Newberry hotspot tracks, diffusive basaltic volcanisms in the Basin and Range, and the circum-Colorado Plateau volcanism can all be explained by the landward intrusion of the hot Pacific asthenosphere since the mid-Miocene. The hot mantle initially enters the continental mantle through tears within the Juan de Fuca slab and is subsequently advected landward by the sinking Farallon slab below the central-eastern U.S. The migration direction of the hot mantle is further regulated by the varying thickness of the continental lithosphere and the roll back of the Juan de Fuca slab. The temporal-spatial evolution of the hot anomalies at lithospheric depth matches the temporal-spatial distribution of the extensive intra-plate volcanisms.The flexibility of data assimilation models also allows us to quantitatively investigate the origin of mantle anisotropy observations. The distribution of shear wave splitting (SKS) in the western U.S. displays several unique patterns including the fast E-W anisotropy from Oregon to Yellowstone, the largely swirl pattern centered in Nevada, and the complex anisotropy around the Rockies. These unique patterns cannot be explained by the traditional toroidal or poloidal mantle flow near a subduction zone. Instead, the same mantle flow driving the intra-plate volcanism also best reproduces the observed SKS. In particular, the toroidal flow related to the Juan de Fuca slab is only restricted to its southern edge. The sinking of the former Farallon slab extends the toroidal flow further eastward. The lithospheric thickness variations help to form a circular deformation pattern centered in western Nevada. The prominent E-W oriented SKS in the Pacific Northwest corresponds to the persistent intrusion of hot anomalies underneath.