An examination of prospective foreshock probabilities in New Zealand

Abstract

This thesis investigates aspects of prospective foreshock probabilities in New Zealand. An earthquake qualifies as a foreshock if it is followed by an event of at least the same magnitude within the next 5 days and an epicentral radius of 30 km. The process of identifying foreshocks is a Binomial situation. Foreshock probabilities are estimated and analysed both for the whole of New Zealand and on a regional basis. Foreshock probability maps resulting from the regional assessment are displayed and discussed. Probabilities obtained from simulated earthquake catalogues are also estimated. In the process of assessing these probabilities, an important aspect subsequent to the identification of foreshocks is the declustering algorithm used to remove aftershocks from the original data before estimation of probabilities. Different declustering algorithms and other improvements to the procedure of estimation of foreshock probabilities are discussed in the course of the thesis. Statistical analyses of regional, temporal and depth relationships to foreshock probabilities are provided. Regional variations in foreshock probabilities are found to be significant. Foreshock probabilities are found to be stationary in time. Probabilities estimated from the Taupo Volcanic Zone are larger than elsewhere in New Zealand.