Stacking Global Seismograms Revisited

Friday, 19 December 2014: 10:20 AM
Peter M Shearer1, Janine S Buehler1, Marine Denolle1, Wenyuan Fan1, Zhitu Ma1, Nicholas J Mancinelli1, Robin S Matoza1, Wei Wang1, Yongfei Wang2 and Zhongwen Zhan1, (1)University of California San Diego, La Jolla, CA, United States, (2)San Diego State University, San Diego, CA, United States
Over 20 years ago, stacks of global seismograms produced direct images of the global seismic wavefield highlighting the visibility, frequency content, and polarity of known seismic phases, and also identified a host of new phases associated with reflections and phase conversions from upper-mantle discontinuities. Two different stacking methods proved particularly useful: (1) STA/LTA-filtered stacks that describe the local signal-to-noise characteristics of the major seismic phases. These serve to image the entire wavefield in a uniform way for educational purposes and to show which phases are observed most clearly as a guide to future research. These stacks also resolve SH versus SV timing differences consistent with radial anisotropy. (2) Reference-phase stacks that preserve the polarity, amplitude, and timing of traces with respect to a specified target phase. These show a large number of top-side and bottom-side reflections and phase conversions from the 410- and 660-km discontinuities that create weak phases with a characteristic "railroad track" appearance both preceding and following many of the main seismic phases. Reference-phase stacking can also be used to produce coherent surface-wave stacks at very long periods, which directly show the dispersive character of the surface waves.

Here we revisit and update these stacks by exploiting the vastly increased data now available from the IRIS DMC to produce greatly improved wavefield images. We present several examples of the different stacking approaches and point out their various features, including promising targets for future research.