Viewing seismic velocity anomalies with 3‐D continuous Gaussian wavelets

Abstract

Seismic velocity anomalies (SVA) have traditionally been viewed as spatial objects. We present a new method for looking at SVA, based on a 3‐D continuous Gaussian wavelet transform. Local spectra of the seismic anomalies are calculated with the wavelet transforms. Two proxy quantities based on wavelets are used for viewing SVA. These proxy quantities are the 3‐D spatial distributions of (1.) the local maxima of the L2‐norm of the seismic anomalies, E‐max, and (2.) the associated local horizontal wavenumber k‐max. The P1200 tomographical model [Zhou 1996] has been used for this purpose. Geographical distributions of E‐max and k‐max yield information which are not obvious from direct visual inspection of SVA. Some examples are the depth extent of the tectonic boundaries and the inference of a plume‐like object beneath the transition zone under Iceland.