Keith Priestley - Cambridge University, UK

The coda of teleseismic P-waves contains information about both the earthquake source and the propagation characteristics of the Earth. In particular, S-waves in the P-wave coda arise from P-to-S phase conversions at boundaries within the crust and upper mantle beneath the seismograph site. The analysis of these converted S-waves has proved extremely useful in determining crust and upper mantle. The time series of the P-wave coda with just the S-wave conversions and the source effects removed has been termed the receiver function. This method was first proposed as a frequency domain technique by Phinney in 1964 but the real utility of the technique was not apparent until after Burdick and Langston recast it in the time domain in the late 1970's. Since that time numerous studies have employed receiver functions to determine the crustal structure beneath broadband seismograph sites. The most common method for removing the source contribution to the P-wave coda has involved either damped spectral division or time-domain deconvolution. The resulting time series can then be inverted directly for the velocity structure beneath the site or if the recordings are sufficiently dense,interpreted in much the same manner as reflection seismology. In the past five years there have been a dramatic increase in the use of receiver function techniques for determining crustal structure and the method has been extended to image interfaces in the upper mantle. New procedures have been proposed to accomplish the source deconvolution minimizing the numerical instabilities in damped spectral division or timedomain deconvolution, and the technique has been extended to regional P-waves and teleseismic S-waves. In this talk I concentrate on recent developments in using the receiver function technique