Considerable progress has been made in mapping seismic variation in the Earth. Many different types of seismic data are brought together in this attack with shallow structure determined by dense regional arrays and global structure determined by using IRIS data in long-period tomographic imaging. An intercomparison of these Earth models derived from different methods and datasets fortunately show some agreement. For example, the fast lower mantle circum-Pacific belt, when examined at high-frequency with regional data, displays small triplications near the CMB indicative of phase changes and/or slab buckling. Similarly, the blurred tomographic image of the African Superplume first recognized in tomographic images appears as a distinct structure when viewed by dense broadband arrays where jumps in SKS travel times demonstrate wall sharpness. Unfortunately, the Pacific Superplume is enclosed by the Pacific Ocean making it very difficult to study. Moreover, the Pacific is surrounded by subduction zones which are not well modeled by global methods and the starting reference model is PREM which has a 20 km crust and a major upper mantle discontinuity at 220 km; neither are realistic.
Fortunately, it appears that the upper mantle beneath the Pacific is relatively simple in terms of geologic history and easily explained with 1D models, if we allow some variation in the lithospheric thickness and LVZ. Recent studies of running stacks of TriNet data (120 stations) allows the identification of secondary phases associated with upper mantle reflected multiples (triplications) S, SS, SSS, etc. Models produced from this data predict the various "pegleg arrivals", upper mantle reflections associated with multiple ScS remarkably well although the relative timing of the multiple ScS and sScS groups display rapid changes with position, at least along the Fiji-Tonga to TriNet array. Because of simplicity of the shallow structure, it appears that the multiples of ScS can provide the ideal data for mapping the fine structure associated with sharp walls of the Pacific Superplume if they exist and provide constrains on the Scd phase boundary especially along the rapidly varying margin structures.
In this study, waveform modeling analyses is applied to the excellent collection of waveform data (Baja NARS Array, PASSCAL experiments, IRIS, CISN, TriNet, and the new Mexican Array (MASE)). The latter contains 200 stations along a transect (Acapulco-Tampico) with roughly 5 km spacing. These efforts are producing refined images of the upper and lower mantle structure beneath the Pacific and eastern Margins where LVZs have been discovered near the 410 and along subducted slabs.
Broader Impact This grant supports a graduate student who will gain experience in the latest wave propagation techniques as well as inversion methods. The detailed images of the upper mantle that we plan to develop will also be of great use to the entire earth science community, including geodynamicists, geochemists, and geologists interested in the tectonics of plate boundaries, and especially mineral physicists.
