Tsuneji Rikitake

Recurrence of great earthquakes at subduction zones

Abstract Statistics of the recurrence times of great earthquakes at the Pacific subduction margins are made. The mean return period of great earthquakes is different from zone to zone, ranging from 27 to 117 years. The standard deviation of the return period proves to be very small, several years say, in some cases. The probabilities of a great earthquake recurring in each zone are estimated on the basis of Weibull distribution analysis. The mean return periods thus estimated are combined with the relative plate velocities at respective zones as obtained in the plate tectonics in order to estimate the ultimate displacement to rupture at the interface of the continental plate and the downgoing oceanic plate. It is presumed that great earthquakes at subduction zones occur as a result of a rebound of the continental plate at the time of rupture. The ultimate displacement thus estimated ranges from 2 to 8 m, and seems somewhat larger than that estimated on the basis of seismic observations, although the value of ultimate displacement seems to harmonize roughly with estimates based on geodetic observations on land. However, the ultimate displacement at the Aleutian—Alaska zone as estimated here seems much smaller than that estimated from actual observations. The ultimate strains, which are deduced from the displacements obtained on the assumption that the logarithmic extent of the deformed area is proportional to earthquake magnitude, are then calculated, and compared with those estimated for large inland earthquakes as revealed by repetition of geodetic surveys. The mean ultimate strain is estimated as 4.3 · 10−5 for subduction-zone earthquakes while that for inland earthquakes has been estimated as 4.7 · 10−5. As the agreement between both the ultimate strains is fairly good, it is tentatively concluded that the strength of the plate interface under the sea bottom is more or less the same as that in the crust on land.

Classification of earthquake precursors

Abstract In view of the fact that fairly many data on earthquake precursors have been reported in recent years, some statistics concerning precursor times are again presented on the basis of the previous and newly-added data. The previous conclusion that the precursors can be classified into three kinds is reconfirmed very positively. The precursor times of the first kind indicate a clear dependence on earthquake magnitude, while those of the second or imminent kind have no magnitude-dependence, with a mean value of 2.4 h. The nature of the third kind of precursors, those mostly represented by foreshocks and changes in tilt and strain, is not quite clear, but the mean precursor time is estimated as 4.5 days.

Biosystem behaviour as an earthquake precursor

Abstract An analysis of 157 existing items of data concerning abnormal animal behaviour prior to an earthquake brings out apparent precursors of two kinds: moderately short-range and extremely short-range precursors; these have precursor times of the order of days and of hours, respectively. The moderately short-range precursor, if it is a precursor at all, covers the time range for which very few geophysical precursors are observed, and in such a case the animal precursor supplements the geophysical one. The nature of extremely shortrange animal precursors is similar to that of geophysical precursors for the same time range. Although it is hard to establish the validity of animal precursors from the present statistics, modern reports from China, Italy and other countries seem to support the existence of such precursors even if the reason why animals are sensitive to usually unmeasurable stimulations preceding an earthquake is not very well understood.

Probability of a great earthquake to recur off the pacific coast of central Japan

Abstract Probabilities of a great earthquake recurring in the Tokai district. Central Japan, are estimated on the basis of the statistical analysis of recurrence tendency and strain accumulation as monitored by geodetic surveys. Although probabilities thus estimated vary in a fairly wide range depending on data, attention should be paid to the fact that the probability of having a great earthquake within 10 years time from 1977 is estimated as high as 30–50%.

Classification of earthquake prediction information for practical use

Abstract Analyzing precursor times of existing earthquake precursor data, time windows for dangerous periods, during which an earthquake of magnitude 6 or over is likely to occur, are defined by a time range during which the cumulative probability of earthquake occurrence increases from 30 to 80%. It appears that the time windows can be classified into three groups i.e. A (an earthquake would occur within several hours time), B (within a few days time) and C (within a period that is not exactly known but not a very long one), depending upon precursors having extremely and moderately short-range precursor times, respectively. Reliability of the above information is rated into three ranks i.e. X (highly reliable), Y (slightly less reliable) and Z (possible), by means of a combination of percentage of mutually independent prediction elements for which a premonitory effect is observed and that of observation stations at which that effect is registered. It is hoped that earthquake prediction information specified by one of the nine segments as defined by the combination of time window and reliability would provide some criteria for emergency control of the public in case a damaging earthquake is believed to be impending.

Physical Parameters in the Earthquake Source Region and Their Temporal Changes

Short-term and imminent earthquake prediction depends on temporal changes of physical parameters in the earthquake source region. In this paper, characteristics of such changes or earthquake precursors of various disciplines such as land deformation, change in sea level, tilt, strain,crustal stress, foreshock, anomalous seismic activity, seismic gap, growth and decay of seismicity, b value, source mechanism, hypocentral migration, seismic wave velocity, earth tide, geomagnetic and geoelectric field, earth resistivity, gravity, underground water, geochemical factor such as radon concentration and the like are summarized on the basis of recent data taken approximately after 1975.