Receiver function analysis of mantle discontinuities beneath the Colorado Plateau and Colorado Rockies using the RISTRA and CREST seismic arrays teleseismic data.

摘要

Converted seismic waves (e.g. P-to-S) are sensitive to rapid velocity change and have been extensively used to characterize seismic discontinuities in the crust and mantle. A receiver function is a time series representing the Earth's responses and is conventionally estimated by deconvolving radial (or SV) component from vertical (or P) component with water-level. Two primary seismic discontinuities in the mantle, the 410- and 660-km discontinuities, are due to olivine phase changes. The region in between is termed as the mantle transition zone which separates the upper mantle from lower mantle. Thickness and topography of the two discontinuities contribute to understanding of mantle convection process, thermal anomaly and chemical composition. A 410-km low velocity layer atop the 410-km discontinuity has been identified regionally and globally. The low velocity layer is interpreted as partial melting and coherent with the prediction of the transition zone water filter model. Using receiver function analysis, I first study the mantle transition zone in the Colorado Plateau by the RISTRA array and find the 410-km low velocity layer is absent in high velocity regions about 410 km depth in P wave tomogram, and is present in low velocity regions. Our finding is consistent with a simple interpretation of the transition zone water filter model which predicts the production of a hydrous melt layer only where upflow of sufficiently hydrated transition zone mantle occurs. The second study is seismic imaging of mantle discontinuities beneath the Colorado Rockies by the CREST array combined with TA stations. The stacking images show a thinner than global averaged transition zone of 241 km, and existence of the 410-km low velocity layer. Those two findings suggest that the mantle is warm and upwelling in the region, which may provide partially uplift to the excess topography of the Colorado Rocky Mountains. A third finding is a negative polarity phase below the 410-km discontinuity beneath the Colorado Rocky Mountains. In the third study, I provide seismic constraints on the two negative polarity phases above and below the 410-km discontinuity found in the CREST project via waveform fitting technique. The best-fitting velocity model finds that the olivine-wadsleyite interval is 21 km, which is larger than anhydrous mineral anhydrous mineral physics predictions (4-10 km). The negative polarity phase below the 410-km discontinuity may result from downward viscous entrainment zone of wet-wadsleyite. The results suggest that the mantle is hydrated beneath the Colorado Rocky Mountains.