Receiver functions from three 30-station IRIS-PASSCAL small-aperture arrays (2–15 km station spacing) operated for 10 months each in the northern Rocky Mountains show a ubiquitous negative polarity P to S conversion just preceding the 410-km discontinuity arrival. Data from the three arrays were sorted into NW, SE, and SW back-azimuth quadrants and stacked to form nine quadrant stacks. Remarkably, the negative polarity arrival (NPA) is apparent in 8 of the 9 quadrant stacks, with 7 of the 8 having well-correlated waveforms. Each quadrant stack also contains clear P to S conversions from the 410- and 660-km discontinuities. Moveout analysis shows that all the major phases display the correct moveout for forward scattered P-S phases. The waveshapes for the seven similar NPA-410 km discontinuity arrivals are modeled with a five-parameter “double gradient slab” model that is parameterized as follows: a top gradient thickness and shear velocity decrease; a constant velocity layer; bottom gradient thickness; and shear velocity increase. Model misfit is assessed via a grid search over the model space using a reflectivity code to calculate synthetic seismograms. Model likelihood is determined by calculating 1- and 2-D marginal probability density functions (PDF) for the five parameters. The 1-D marginals display a range of peak values, although significant overlap is observed for the top gradient thickness and its associated velocity decrement. From the peak value of the summary PDF, we find the top velocity gradient to be sharp (s). Defining an effective thickness of the low-velocity layer as the mean layer thickness plus half the mean gradient thicknesses, the 410 low-velocity layer thickness is found to be 22 km. A review of changes in the physical state required to match our new 410-LVL constraints suggests that the water-filter model remains an operative hypothesis to test.
Available at: http://works.bepress.com/jjasbins/2/