Surface wave waveform inversions for local shear-wave velocities under eastern Australia

The waveform inversion method developed by Kushnir et al. [1989] and expanded by Passier and Snieder [1995b] yields estimates of structure along short interstation paths and of average velocity gradients. In contrast, 3D tomographic inversions yield local estimates of the earth structure by combining the information contained in a large set of long propagation paths. Both approaches lead to estimates of the local earth structure, but the two methods are subject to different types of artifact. As a test case of the consistency of the two approaches we consider the recent S-velocity model of the Australian region by Zielhuis and Van der Hilst [1996]. We invert 20 Rayleigh wave waveforms between 20 and 110 s recorded on the Australian continent, most of them by the mobile SKIPPY network. From the data, we directly extract measurements of the S-velocity structure at various short interstation paths in Australia and obtain estimates of the average velocity gradient along long paths. The estimates of the local interstation structure and the estimates of the velocity gradient are consistent with the 3D model of Zielhuis and Van der Hilst [1996]. These results provide illustrative examples of how specific features of a complex velocity structure can be determined using only a few waveforms.