Characterizing Radial Heterogeneity in the Mid Mantle

Revenaugh

The chemical partitioning of the Earth's mantle remains enigmatic despite significant advancements in geochemistry, geodynamics and seismic imaging in the past decade. Its importance to mantle dynamics is hard to overstate, and deciphering it, even in part, would undoubtedly improve our understanding of the chemical evolution and dynamics of our planet. In this proposal, the PI outlines a program of study to characterize the seismic velocity heterogeneity in the mid-mantle (i.e, between depths of ~1000 and 2500 km). He will pursue three complimentary approaches. The first extends ScS reverberation mapping to cope with strongly varying along-path structure. This work will offer constraints on mid-mantle shear-wave impedance variability and can be applied to paths with little interaction with present-day subduction. This is important inasmuch as most seismic imaging of mid-mantle scattering heterogeneity is of subduction zones, presenting a potentially skewed view of sub-transition zone structure. The second approach uses small and regional seismic 100 km aperture) array recordings of high-frequency body waves. S-to-P conversions, forward-scattered P, and P-to-S precursors to S will be investigated using source arrays in Tonga-Fiji, the Aleutians and South and Central America. Unlike most prior work, this will look extensively at the P-PcP and S-ScP lines of slowness-time space, investigating evidence of back-scattered P waves in the mid-mantle. The third approach is to adapt envelope-stacking methods for use with dual-arrays, combining this energy-centric approach with the statistical migration methods the PI has developed for examining crustal structure beneath regional arrays. By extensively characterizing mid-mantle heterogeneity using phases whose sensitivity to short-scale structure is strong and localized, the PI hopes to see what velocity tomography can't: evidence of chemical layering and/or mantle 'blobs.'

Broader impacts: The PI will disseminate the results, making available data sets that are not publicly archived at present, publishing codes of all new algorithms through IRIS, and presenting numerical results in raw and graphical form on the PI's webpages. This proposal would fund the Ph.D. research work for one graduate student and would employ several undergraduates for preliminary data analysis. The methods to be developed are intended to maximize the capabilities of large, irregular arrays like EarthScope and may have use in other fields, such as long-range sonar array sounding.