On the Accuracy of Hypocenter Determination of the Hokuriku Microearthquake Observatory

Abstract

The accuracy of a method of hypocenter determination which was carried out as a routine work at the Hokuriku Microearthquake Observatory was investigated. A telemetering observation system with seven satelite stations was completed in 1976 at the Observatory, with a time accuracy of 20 milliseconds. In this article, the accuracy of hypocenter determination is discussed principally from the viewpoint of errors of reading seismograms and the propriety of analysis methods. The accuracy of reading seismograms was examined by comparing the results read independently by two persons on a pen-recording seismogram with a paper-speed of 8 mm/sec. The results show that the reading errors caused by the reader's personality are estimated as, at most, 0.1 s for P-wave and 0.5 s for origin time converted from S-P time. Variations of hypocenters were calculated by giving observational errors Δt in various ways, in which Δt is 0.1 s for P arrival times or 0.5 s for the origin times. The results are as follows : Variations of epicenters are less than 2 km in the neighbourhood of the network, and are about 5 km apart from it. In the middle distant area, the operation of adding Δt mainly affects the depths of hypocenters. The variation of hypocenters caused by the differences of combinations of stations and of the sampling of the assumed P-wave velocity data were also examined. These results show that the variations are small enough to discuss seismic activity. From these results, it seems important for the accurate determination of hypocenters to know crustal structure as correctly as possible.