Study of the seismic velocity ratio for several regions of the South Island, New Zealand: evaluation of regional ts/tp and near source Vp/Vs values

Abstract

The regional variations in space and time of the travel-time ratio ts/tp over a 600-km long section of the Alpine fault zone in the South Island, New Zealand, are discussed. The analyses are based on readings from micro-earthquakes. A significant difference in this ratio with depth between shallow and intermediate focus earthquakes in Fiordland was found, the former having a value of 1.72 and the latter 1.68. A 7 per cent decrease in ts/tp preceded a magnitude 3.5 earthquake in the Arthur's Pass region by about 10 day. The entire region studied was found to be characterized by regional ts/tp values of 1.74–1.70 except a 75-km long region near Arthur's Pass at the junction of the Alpine and Hope faults which had a value of about 1.63. This low value was observed in both 1972 and 1973. Near-source Vp/Vs ratios of about 1.55 were estimated for two earthquake zones (sub-regions) within the Arthur's Pass region. Computed estimates of the acoustic velocities Vp and Vs for the Arthur's Pass region yielded values of about 5.48 and 3.40 km/s, respectively. When compared to ts/tp values from other parts of the world and with ultrasonic data on rock types similar to those present in the South Island, the regional ts/tp study indicates that the Arthur's Pass region possesses a significantly low value of ts/tp. This low ratio for the Arthur's Pass region may be interpreted in two different ways. Since the ts/tp values are higher in the regions to the north and south of the Arthur's Pass region, with all three regions having the same lithologic unit exposed on the surface, it may reflect the depth to the crystalline basement. In particular, it may be suggestive of a greater thickness of sediments existing within the Arthur's Pass region. Alternatively, this low ratio may be indicative of a temporal effect precursory to a future earthquake. On this assumption, a quantitative estimate of the size, location and time of occurrence of a possible future earthquake is deduced. Such an event would have a minimum magnitude of 6.5 and would occur in the next few years. Whatever the explanation, which from this analysis cannot be definitely decided, this study, by attempting to define lateral heterogeneities in the crust, warns that in earthquake prediction studies involving the interpretation of ts/tp anomalies, considerable lateral variations in this parameter may result from local variations in geology.