Spatial distributions of strain- and tilt-steps associated with two earthquakes which occurred in the region near Izu Peninsula, Japan at May 9, 1974 (
M=6.9) and Jan. 14, 1978 (
M=7.0) are investigated with respect to their source mechanisms.
In both cases, it is found that the observed step values do not harmonize well with the theoretically expected residual strain or tilt fields which are calculated based on an appropriate fault model. It is said that the amplitudes of steps may be considerably amplified by the effect of crustal layering and topography, whereas the polarities of them are not affected so much.
To see the circumstances of this discrepancy, critical review of similar investigations studied so far are attempted adding to the results obtained here. First, the dependence of the polarity consistency between the observed and the theoretical steps upon the epicentral distances are investigated for six earthquakes. The main results are as follows. (1) No established empirical relationships between them are found. (2) It does not always follow that the data from nearby stations are more reliable than those from distant stations. (3) In all cases, the total ratio of the polarity consistent data is about 60% for strain-steps and 50% for tilt-steps, which means a nearly random condition.
Next, the responces of the specified instruments at five stations to six earthquakes are examined only to find neither obvious peculiarities nor simple relationships between the observed and the theoretical step values.
As to strain or tilt measurement, there remains miscellaneous unknown factors which may seriously distort the signal such as local ground conditions or underground water pressures. Without the accurate knowledge about these effects, it seems to be difficult that we use strain- or tilt-step data as a material for precise argument about source processes.
抄録全体を表示