Transition zone structure of California from Ps converted waves

We present new estimates of the structure of seismic discontinuities in the upper mantle of California and the northern Gulf of California from the analysis of stacks of 10,000 TriNet, Berkeley Digital Seismic Network, and NARS-Baja waveforms. Throughout the region, Ps conversions from the 410-km and 660-km conversions arrive 2-3 s later than predicted by IASP91, due to low seismic velocities in the uppermost mantle of western North America. In addition, the differential traveltime of P660s and P410s is up to 1 s shorter than predicted by IASP91, which is among the lowest values observed worldwide, in accord with earlier observations of single-station stacks. We can explain these observations using the physical reference model of Lithgow-Bertelloni and Stixrude [2005] which incorporates a homogeneous pyrolytic mantle with an adiabatic temperature of 1350 C. While the transition zone beneath California is as thin as beneath various hotspot regions, it is not necessary to invoke an active upwelling to explain our data.