THE PRECURSORY EARTHQUAKE SWARM IN NEW-ZEALAND - HYPOTHESIS TESTS

Tests have been carried out on the simple precursory swarm hypothesis, which proposes a one-to-one relation between swarms and mainshock events, and tests are in progress on the generalised hypothesis, which allows for clustering. The performance of the simple hypothesis relative to the usual stationary Poisson model was favourable for several years after the start of the tests in 1977, then deteriorated, and fell to the proposed rejection level with the occurrence of the Doubtful Sound earthquake of 1989 May 31, M 6.1. During the tests on the simple hypothesis, the database changed substantially with the adoption of a revised magnitude scale by the New Zealand Seismological Observatory. This resulted in additional swarms appearing in the past catalogue, and also in swarm clusters being recognised as far back as 1968-72, including a cluster precursory to the 1989 Doubtful Sound earthquake. These changes gave further grounds for rejecting the simple hypothesis. Accordingly, the generalised hypothesis, which has been under test in Japan since 1983, was formulated for New Zealand conditions, and tests were started in July 1991.