Onshore-Offshore Body Wave Tomography of the Cascadia Subduction Zone: Identifying Challenges and Solutions for Shore-Crossing Data

Session: Amphibious Seismic Studies of Plate Boundary Structure and Processes
Type: Oral
Date: 4/29/2020
Time: 05:30 PM
Room: 120 + 130
Description: 

Amphibious (onshore-offshore) studies of subduction zones advance our understanding of how the incoming oceanic mantle, subduction dynamics and surface processes are related. However, joint onshore-offshore imaging is challenging due to the structural contrasts between continental and oceanic lithosphere. These near-surface variations (<50 km depth) impact how we image and interpret targets in the mantle due to the inability of teleseismic raypaths to resolve structure of this scale. Here, we present teleseismic tomography results for the Cascadia subduction zone (CSZ), derived from S- and P-wave delay times. We characterize the problems inherent in shore-crossing teleseismic data by performing several synthetic tests. First, forward modeling suggests >1s of variance in onshore-offshore P delay times due to structure in the upper 50km from changes in elevation and crustal thickness. Likewise, sediment thickness variations can heavily impact offshore measurements, for S-waves up to 1 s. Next, we look at ways to account for this structure, running synthetic inversions with competing methodologies and identifying imaging artifacts. We find that using station statics reduces the amplitude of mantle targets and can create imaging artifacts, specifically in the subslab region. We prefer a method that uses an a priori starting model, derived from independent constraints, to account for near-surface structure. Our tomographic method allows us to explicitly include elevation, perform iterative 3D raytracing through the starting model and includes finite frequency kernels. We apply all of the methodologies explored in our synthetic tests to the observed delay time data, documenting how they influence the final images and identifying potential artifacts. Finally, we present our preferred P- and S-wave models, which utilize a 3D starting model derived from surface wave studies and heavily damped stations statics to account for local geology. We compare the P and S models noting regions which may be influenced by physical properties such as anisotropic fabric or melt content.

Presenting Author: Miles Bodmer

Authors