THE PRECURSORY EARTHQUAKE SWARM IN NEW-ZEALAND - HYPOTHESIS TESTS II

A series of tests is being carried out with the object of determining whether the relations evident in New Zealand and Japan, between precur sory swarms and major earthquakes, are of value for long-range, synopt ic earthquake forecasting. The first New Zealand test (completed in 19 90) showed that clustering, both of swarms and of mainshock events, sh ould be allowed for, and the hypothesis was reformulated accordingly. The second New Zealand test, now completed, confirms the importance of clustering. It also reveals that the applicability of the swarm/mains hock relations is strongly affected by large-scale tectonics. Further, a simulation study shows that, at the end of the test, the performanc e of the swarm hypothesis relative to the Poisson model lay between wh at would be expected if the hypothesis and the Poisson model, respecti vely, were correct in general. These results support a further reformu lation of the hypothesis, which is now taken to be strongly applicable in the Hikurangi and Fiordland subduction zones, not at all in the re gion of Quaternary volcanism, and weakly elsewhere. This change and th e provision for clustering made previously bring the present New Zeala nd formulation close to that under test in the Japan region. The third New Zealand test commenced on 20 December 1995 and is still in progre ss.