Imaging Transition Zone Thickness Beneath South America from SS Precursors

We image detailed upper mantle discontinuity structure beneath a number of geologically active regions, including the South American subduction zone, the Scotia plate subduction zone, and several volcanic hotspots (e.g., the Galapagos Islands), in a region ~10,000 km by 10,000 km wide, spanning 70° S to 20° N and 20° W to 110° W. Precursors to the seismic phase SS are analyzed, which form as a result of underside reflections off seismic discontinuities beneath the midpoint of the SS path and are highly sensitive to discontinuity depth and sharpness. Our SS dataset consists of over 15,000 high-quality transverse component broadband displacement seismograms collected from the Incorporated Research Institutions for Seismology (IRIS), the Canadian National Seismic Network (CNSN), as well as data from EarthScope seismic stations, and from the Canadian Northwest Experiment (CANOE) temporary broadband array deployment. This dataset densely samples several regions in our study area and significantly improves the sampling for this area compared to previous precursor studies. Data with common central SS bouncepoints are stacked to enhance precursory phases. Solution discontinuity structure depends on a number of factors, including dominant seismic period, crustal correction, signal-to-noise ratio threshold, and tomography model used for mantle heterogeneity correction. We exclude precursor data predicted to interfere with other seismic phases, such as topside reflections (e.g., s670sS), which have been demonstrated to contaminate final stacks. Solution transition zone thickness is at least 20 km thicker than global average estimates of 242 km along the northwestern portion of the South American subduction complex (Peru, Ecuador, and Columbia); this thickening extends 1000-1500 km to the east beneath the continent, but does not appear to continue south of -20° latitude along the convergent margin. A minimum of 10 km of thickening is imaged to the west of the Scotia subduction zone, consistent with colder material reaching the transition zone, and 5-10 km of thinning off the southern coast of Brazil, possibly associated with the Martin/Trinidade and/or St. Helena hotspots. Our results are consistent with tectonics at much smaller scales and hence support the SS precursor approach as a valuable probe of transition zone structure and processes when abundant broadband data are available.