Testing the hypothesis of the slab-plume interaction on the formation of the Yellowstone hotspot system

Origin of the Yellowstone hotspot system remains debated. Proposed hypotheses fall into two main categories that involve either a deep mantle plume or only upper mantle processes. Recent seismic tomography images suggest the existence of hot mantle upwelling beneath the present-day Yellowstone, lending support to the plume hypothesis. However, the effect of the Farallon slab on the temporal evolution of the hypothesized plume remains unclear. We use 4-D geodynamic models to investigate the temporal evolution of slab-plume interaction on the formation of the intra- plate volcanic province that includes the mid-Miocene Columbia River flood basalt (CRFB) and subsequent Yellowstone and Newberry hotspot tracks. We find that the sinking slabs dominate the buoyancy and dynamics of the system, and that evolution of the mantle upwelling is subject to that of the slabs. Our best-fit model with a hot upwelling starting at 35 Ma below the present-day Yellowstone can match both the mid-Miocene flood basalt event and the present-day lower mantle seismic image, suggesting a possible contribution of deep mantle to the CFRB. However, all models predict very little upper mantle residual hot anomaly, in contrast to the voluminous slow seismic anomalies, especially beneath the Snake River Plain. These models also fail to generate a vertical plume conduit below Yellowstone as suggested by seismic tomography. We conclude that additional mechanisms must be responsible for adding more heat to the western U.S. upper mantle