The lowermost mantle beneath northern Asia-I. Multi-azimuth studies of a D″ heterogeneity

Abstract

Summary In most areas of the globe it is only possible to sample the lowermost mantle or D″ region with a limited range of azimuths. This is due to limitations in source–receiver geometries. An exception is the region beneath northern Siberia where we are able to sample D″ along orthogonal great-circle paths using northern-Pacific earthquakes recorded in Europe and Hindu-Kush earthquakes recorded in Canada. Receiver vespagrams (slant stacks of one source recorded at many receivers) for a Kurile event recorded at the German Regional Seismic Network (GRSN) and the Grafenberg-array and source vespagrams (many sources, one receiver) of Hindu-Kush events recorded at stations of the Canadian National Seismograph Network (CNSN) show reflections from a discontinuity at the top of D″. Vespagrams for real data are compared to vespagrams for synthetic data in order to guide their interpretation. Additional support for our interpretation comes from frequency-wavenumber (f−k) analysis which provides a measure of the slowness and azimuth of a given phase. The traveltimes of the D″ reflected phases are then used to provide estimates of D″ thickness. In regions where the roughly orthogonal paths overlap we estimate the thickness to be 310 (±15) km, which is in agreement with existing models for this region. We also estimate a thickness of 210 (±15) km in a previously unsampled region east of this area. Using the slowness, backazimuth and traveltime values for a third source–receiver combination, we find a reflection point at a depth of 2800 km (thickness of D″ of 91 km) in a region several degrees to the west. These results and those of previous studies suggest considerable topographic relief (100 km) over distances of less than 5° (approximately 300 km at the CMB). Comparing the results from this study with results from other seismic studies of D″ structure in this area, we find this to be a complicated region which exhibits heterogeneity on many scales. This region may be the lower-mantle graveyard of the subducted Izanagi plate and one interpretation of this complexity is that it is due to subduction related deformation at the base of the mantle.