Teleseismic P and S Delay Times within Tectonically Active and Stable North America

We have measured teleseismic P and S relative delay times within 1) Stable North America (SNA) using waveforms from IRIS PASSCAL seismic arrays MOMA (Fischer et al., 1995), ABBA (Roecker and Beavan, 1995), Abitibi (Hearn and Mareschal, 1996), and FLED (Wysession and Fischer, 2001), and 2) Tectonically-active North America (TNA) using Earthscope's Transportable Array (TA). To study the contribution of mantle structure to these delays we subtracted delays predicted for topography and crustal structure, using CRUST 2.0 (Bassin et al., 2000). Preliminary analyses of delay times from earthquakes with Mw>=6.5 show surprising differences between the heterogeneity of the mantle beneath SNA and TNA. While the range of delay times is expectedly small for an intra-shield array such as Abitibi, the range of delay times from Proterozoic basement in the midwest to Paleozoic margin in New England is much larger and slightly exceeds that for the TA in TNA. This suggests that that the mantle of SNA is slightly more heterogeneous than TNA, despite there being relatively little surface expression of this heterogeneity. Patterns of P and S relative delay times measured in TNA correlate better with surface tectonics, suggesting that the mantle in TNA has a greater effect on the surface geology than in SNA. The central and southern Basin and Range are characterized by positive delays. As shown in previous studies, the Snake River Plain is also well delineated by positive delays. These delays exhibit a significant peak at station H17A in Yellowstone National Park. Teleseismic P and S waves arriving at stations in the Rocky Mountains are much faster, including in northern Idaho and western Washington, but not in western Oregon. For both SNA and TNA, the measured S and P delay times have a significant linear correlation, with S delays at approximately 3 times the P delays, which confirms the dominant effect of mantle temperature on mantle velocity structure. However, the slope of this correlation appears slightly higher for SNA than for TNA, which could indicate a compositional effect that has a relatively larger effect on S waves, such as significant lateral variation in water or iron content.