Effects of topography on upper mantle discontinuities for array detections of PP precursors

PP underside reflections off upper mantle discontinuities are frequently used to map discontinuity topography, impedance contrasts and to interpret these with respect to thermal and/or mineralogical variations. While the seismic discontinuities at 410 km and 660 km depth should be a global feature, several events show no detections or reduced amplitudes of the precursors. In this study, we investigate effects of topography on upper mantle discontinuities on array detections of PP precursors. Using the 2.5-D axisymmetric finite difference technique PSVaxi, we compute P-SV synthetic seismograms for two-dimensional model geometries with correct 3-D geometrical spreading. Retaining dominant periods of ~ 2 s, we investigate Gaussian-shaped upward or downward deflections of the 660 km discontinuity with varying lateral dimensions. Furthermore, we investigate effects of double discontinuities at ~660 km depth which are due to phase transformations in the non-olivine component of the mantle at subduction zone temperatures. Analyses of travel residuals indicate that topography of downward deflections of discontinuities is underestimated by 10 to 20 km while upward deflections are recovered within 10 km. Amplitude measurements show focussing and defocussing of PP precursor amplitudes by 60-70%. Reduced amplitudes of PP precursors are close or below the average noise level at seismic arrays and could result in non-detections. Double discontinuities at 660 km depth are recovered for lateral dimensions larger than 10° and if the two discontinuities are separated by at least 20 km. The observed effects in synthetic seismograms raise caution for interpretation of PP precursors in terms of impedance contrasts as well as thermal and mineralogical variations in the mantle.