DI53A-06
P- and S-wave Slowness Anomalies in the Mantle

Friday, 18 December 2015: 14:55
301 (Moscone South)
Yao Yao, University of Utah, Geology and Geophysics, Salt Lake City, UT, United States and Michael Scott Thorne, University of Utah, Salt Lake City, UT, United States
Abstract:
Anomalies in the slowness of teleseisms have been observed in numerous studies, with previous efforts focusing on crust and upper mantle sources for their origin. Little attention has been devoted to the global distribution of P- and S-wave slowness anomalies in the deep Earth. In this study, we use large aperture seismic array data to examine slowness anomalies as a function of depth in the lower mantle. We collected seismic recordings from all broadband seismic stations in North America for earthquakes between January 2004 and June 2015 with moment magnitudes between 5.8 and 7.5, event depths greater than 100 km, and epicentral distances from 40° to 90°. We chose the time range to coincide with the Earthscope seismic experiment. The epicentral distance range used in this study ensured the target phases, direct P and S wave arrivals, turned in the mantle at depths ranging from 1000 to 2800 km. The original data set contained 420 events with 171,696 seismograms. We inspected each seismogram manually and discarded traces without clear P or S arrivals. Our final data set consists of 278 events with 129,748 seismograms. For each event, we grouped the data into 3° radius geographic bins and calculated relative time shifts for each bin using the Automated and Interactive Measurement of Body-wave Arrival Times (AIMBAT) technique. AIMBAT is a python tool for measuring teleseismic arrival times based on the multi-channel cross-correlation (MCCC) method. For each bin, we plotted the relative time shifts as a function of epicentral distances and calculated the corresponding least-square regression line. The slowness (dT/dΔ) can be obtained as the slope of the regression line. The slowness values of all geographic bins were collected to build a slowness profile for each event. In order to identify slowness anomalies, these slowness profiles were compared with synthetic slowness profiles calculated using the 2.5-D axi-symmetric finite-difference methods PSVaxi for P waves and SHaxi for S waves. Seismic tomography models LLNL-G3Dv3 for P-waves and S40RTS for S-waves were used as the baseline for synthetic modeling in order to eliminate any anomaly caused by known structures. We present a global distribution of mantle P- and S-wave slowness anomalies as a function of depth in the mantle and explore their possible origins.