Compressional and shear velocities in the upper mantle beneath New Zealand determined using local crustal earthquakes

Abstract

Lateral inhomogeneity is to be expected in the upper mantle beneath a tectonically active region such as New Zealand. This study was undertaken to attempt to establish the extent of inhomogeneity at the top of the upper mantle by mapping the variation of Pn and Sn velocities. The velocities are found to vary markedly from region to region. The primary data for this study are derived travel times between stations. These were useful because they have standard errors that could be estimated and expectations that are directly related to the Pn and Sn velocities and crustal delay times, which measure travel times in the crust. The first attempts at estimating the velocities and crustal delay times, which cannot be validly calculated independently, employed reversal methods. Reliable estimates of average velocities and some crustal delay times were obtained but these methods were, in general, inadequate. Next, taking into account all the estimates available, the distribution of velocities at the top of the upper mantle was approximated using a simple model, which consisted of regions within which the velocities were assumed to be constant. A priori constraints on the values of the crustal delay times were used together with the model to determine further estimates of the velocities and crustal delay times. The residuals from this analysis were used to define the final model. Nine regions were needed to model the top of the upper mantle beneath New Zealand, and for completeness a tenth was added to cover the area between New Zealand and the Chatham Islands. The velocities calculated for these regions range from 7.4 km/sec to 8.7 km/sec for Pn and 3.95 km/sec to 4.8 km/sec for Sn. The regions and the calculated values of velocity correlate well with a number of other geological and geophysical phenomena, in particular the distribution of source properties of shallow earthquakes. The crustal delay times from the final analysis effectively depend on crustal thickness and wave velocities in the crust only. They are used to examine whether or not elevations above sea level are compensated beneath New Zealand as predicted by Airey's hypothesis and whether or not Bouguer gravity anomalies measure crustal thickness taken from sea level. It was found that the crustal delay times probably reflect regional geology. To conclude the thesis a petrological model is suggested to explain the calculated values of the Pn and Sn velocities.