Precursory and coseismic resistivity changes

More than 20 coseismic resistivity steps were observed by a resistivity variometer at a station about 60 km south of Tokyo during a six-year period of observation. The logarithmic magnitude of steps normalized to an epicentral distance (δ) of 100 km is linearly correlated to earthquake magnitude (M). The larger the earthquake magnitude is, the longer is the time required for completing a resistivity step. Earthquakes, that produced a step, as plotted on anM-δ graph are clearly separated by a straight line $$M = - 12.5 + 2.5\log _{10} \Delta $$ from those that did not produce a step, δ being measured in units of km. The straight line is approximately equivalent to a resolution of strain steps of the order of 10−9. As the smallest rate of resistivity change that can be observed by the variometer is 10−5 or so, it is apparent that mechanical strain is amplified by a factor of 104 or thereabouts when it is measured in terms of resistivity change. Some of the steps are preceded by a premonitory effect. When long-period changes, mostly due to tidal loadings, are eliminated by a numerical filtration technique, precursors are brought out very clearly in a number of cases. The precursor time ranges from 1 to 7 hours and appears to be independent of magnitude.