Simultaneous time-domain deconvolution with application to the computation of receiver functions

Receiver functions are produced by deconvolving the horizontal components of a seismogram by the vertical component. Typically this is performed by spectral division of recordings of single events. Receiver functions from events with similar backazimuth and ray parameter are then stacked to improve the signal-to-noise ratio. To avoid the subjective and time-consuming method of prewhitening typically employed with spectral division, we have cast the deconvolution in the time domain. By performing the time-domain deconvolution as a regularized simultaneous inversion for a group of events that would normally have been stacked after deconvolution, we find that side lobes are reduced and resolution is improved. Furthermore, the regularized inversion allows the user to choose among a variety of objective model norms. In this paper, we present results from inversions employing the L₂ norm and lower-bounded least squares.