Imaging the mantle using receiver functions beneath Southern California



Journal Title

Journal ISSN

Volume Title


Texas Tech University


Mantle convection in generally accepted as the driving mechanism for lithospheric plate motion. There is, however, a debate as to weather convection occurs in large cells involving the whole mantle or if there is layered convection in chemically stratified mantle. Thermal relationships and chemical composition hypothesized for the upper mantle based on these models will affect the depth to olivine and garnet phase changes differently. This study will use receiver functions to map short wavelength variations in the depths to these phase changes. In the past, earthquake seismology has lacked the data coverage to employ high resolution 3-D seismic reflection techniques typically employed by the oil industry. Southern California has twenty years of data from 22 permanent broadband seismic stations and, therefore, provides the richest data set for such analysis.

As waves pass through seismic discontinuities P to S conversions occur and Pds where d indicates the depth of conversion. For example the Pds conversion fix)m the 410 km discontinuity (the hypothesized olivine to spinel transition) will be labeled P410s. These receiver functions are produced by deconvolving radial component seismograms with the vertical (assumed to be the P-wave signal) to isolate P-to-s conversion amplitudes as a function of time. To increase the signal to noise ratio and provide a 3-D image of the subsurface these receiver functions are back projected through a 3-D model and stacked by common conversion point (CCP) similar to the CMP method used in reflection seismology. Earlier work conducted by these investigators revealed horizons near the 670 km depth that were anti correlated in depth in a way that supports the presence of phase changes in both the olivine and garnet mineral systems. While the depth variations between the 410 and 670 were consistent with the phase change in the olivine mineral system no evidence for the garnet system were observed at the 410 km discontinuity. This work revisits this earlier work using more than twice the volume of data to provide a higher resolution model of these upper mantle seismic boundaries.