Regionalized shear velocity models of the Pacific upper mantle from observed Love and Rayleigh wave dispersion

Abstract

Summary. Group and phase velocities for periods between 16 and 110 s have been determined using the single-station method along 33 Rayleigh-and 30 Love-wave paths across the Pacific. These dispersion data were used to detect variations in upper mantle structure beneath the Pacific. Taking both regional changes and azimuthal anisotropy into consideration, regionalized group and phase velocities were derived for four regions of differing age. The boundaries of the regions were chosen to correspond to isochrons of 20, 50 and 100 Myr, as inferred from geomagnetic lineations. A small, but significant degree of azimuthal anisotropy (less than 0.8 per cent) was found to occur throughout the Pacific if the degree and direction of maximum velocity were assumed to be uniform everywhere. The average direction of maximum velocity was found to be 85°± 6° as measured clockwise from north. Regionalized group and phase velocities for Rayleigh waves at the larger periods increase systematically with increasing age of the ocean floor. For Love waves, departures from a systematic increase occur in the youngest (0–20 Myr) and oldest (> 100 Myr) portions of the Pacific. The dispersion data were corrected for the effects of anelasticity and were inverted to obtain shear velocity models for each of the four regions using the stochastic inversion method. A comparison of the shear wave models derived from Rayleigh waves in the four regions indicates that average shear velocities at depths between 30 and 110 km increase rapidly with age. This increase occurs mainly because the depth to the base of the lithosphere increases; however, the shear velocities of the lithosphere also increase for the regions out to 100 Myr in age. Slightly lower velocities in the upper part of the lithosphere for greater ages may be due to a large influx of mid-plate volcanism during the Cretaceous. Separate inversions of Love-and Rayleigh-wave velocities indicate that polarization anisotropy can be resolved for the lithosphere. Polarization anisotropy cannot, be resolved for the low-velocity zone beneath the Pacific using these data, except possibly in the youngest regions, where lateral complexities of the East Pacific Rise are apt to affect the data adversely. Assuming that we can adequately invert our Love-and Rayleigh-wave data separately to obtain information on anisotropy, we find that average SV velocities in the lithosphere increase from about 4.3 to 4.6 km/s, and average SH velocities increase from about 4.5 to 4.7 km/s, from the youngest to oldest regions. Average velocities in the upper portion of the low-velocity zone increase from about 4.1 to about 4.25 km/s for the same regions. The existence of anisotropy in the lithosphere is constant with the results of seismic refraction experiments in the Pacific and supports peroxide as the predominant upper mantle material. The occurrence of anisotropy through-out the entire lithosphere, suggests a situation in which periodite crystallizes at the base of the lithosphere from a partially molten low-velocity zone and olivine crystals become aligned in response to the stress field within the Pacific plate.