Long- and short-term earthquake prediction in Kamchatka

Abstract

This paper presents the results of long- and short-term earthquake prediction obtained during 1971–1974. They can be summarized as follows: The map of long-term prediction for the Kurile—Kamchatka zone compiled in 1965 and supplemented in 1972 by S.A. Fedotov is in good agreement (in four of four possible cases) with recorded seismicity. The results obtained allow us to suppose that the areas for which the log (Ep/Es) of small earthquakes is low may be the areas of future large earthquakes. Prediction of active periods for the Kamchatka earthquakes with M > 7 has been made on the basis of studying the correlation of seismicity with the lunar tide with a 18.6-year period. A possibility has been found for using the phenomenon of “induced foreshocks” for earthquake prediction, i.e., when a large remote earthquake induces small preceding events in the zone of preparation of a large earthquake. The following three methods were used for operative short-term prediction of the time and place of future earthquakes with M > 5.5. 1.(1) Use of specific electrotelluric field anomalies, from 5 to 20 days in duration, which are recorded by a specially designed network of stations. 2.(2) Method of Vp/Vs anomalies. The anomalously high and low Vp/Vs values for a seismic station point to the possibility of large earthquakes near the latter. 3.(3) The earthquake statistics method described by Fedotov et al. in 1972. Short-term seismic prediction is being made twice a week in two versions: Forecast I (for the whole of Kamchatka) and Forecast II (for each of six overlapping segments of the Kamchatka seismic zone). This paper discusses the results of successful testing of short-term earthquake prediction during two years. During the “alarm” periods the probability of large earthquakes is double the average. Paper presented at the Symposium on Earthquake Forerunners Searching, Tashkent, May 26–June 1, 1974.