Memoirs of the Kakioka Magnetic Observatory Vol.11 No.01， p.71, March, 1963

Results of Geomagnetic Routine Observations and Earthquake

Yoshimatsu, T.

Abstract

 Using the ordinary magnetograms at Kakioka (Geogr. Lat. 36°14′N, Long. 140°11′E) for the period 1959 to 1962 and for the period 1960 to 1961 at Kanoya (Geogr. Lat. 31°25′ N, Long. 130°53′ E),(ΔZ/ΔH)′s of short period changes are calculated through a crucial criterion for the selection of the changes. The time durations of them are within about a few minutes to several tens of minutes and their maximum amplitudes in the vertical and horizontal intensities are denoted by ΔZ′s and ΔH′s.
  The monthly mean values (ΔZ/ΔH)’s are examined to be correlated to the earthquake of which magnitude is M≧5.5 and epicentral distance Δ≦330km from each station.
  The following interesting results are found out at both stations.
 
  1). Monthly means (ΔZ/ΔH)’s show statistically significant changes throughout the whole periods over the limit of fluctuations of individual values of (ΔZ/ΔH)’s. (Fig.1
  and Table. 1)
 
  2). There is a tendency that large scale earthquakes are apt to occur in the months of larger values of (ΔZ/ΔH)’s. (Fig. 2)
 
  3). All maximum value of (ΔZ/ΔH)’s, which are appeared in both the month of earthquake and month of no earthquake, take nearly the constant value at each station, namely, 0.660 at Kakioka and 0.520 at Kanoya, respectively. (Fig. 1) It is desirable to check using more available data whether these maximum values are remained constant even in the period lasting several months without earthquake occurrences.
  4). The change of (ΔZ/ΔH)’s is distinguished as a whole in two types of mode, a short-interval change which lasts less than three months between two consecutive maximum values of (ΔZ/ΔH)’s and another a long-interval change lasting several months, namely, Oct. 1959-July 1960,Aug. 1961-Mar. 1962 and June 1962-0ct. 1962. (Fig. 1)
 
  5). Nearly each earthquake considered is preceded by a minimum value of (ΔZ/ΔH)′s, one or two months before the earthquake. The relation between the magnitude M and the preceding minimum value is given by a least square method as follows;
 
  (ΔZ/ΔH)・−1,2＝Aκa - Bκa・Mκa (1)
 
  (ΔZ/ΔH)・−1,2＝Aκy - Bκy・Mκy (2)
 
  where suffixes ka and ky represent Kakioka and Kanoya, respectively, and A’s and B’s are constants. (Figs. 3 and 4) It is desirable to confirm whether or not Bκa is equal to Bκy and also these formulae hold good for even minor earthquakes with M<5.5,using more available data. It should be noted that (1) and (2) contain no terms of Δ.
 
  6). It may be expected from (1) and (2) that even the greatest earthquake ever occurred in Japan with M=8.0〜8.5 does not yield the minimum value of (ΔZ/ΔH) deviating by 20% from the maximum value 0.660 at Kakioka, or 0.520 at Kanoya.
 
  7). A triangular mark in Fig. 3, which is slightly deviated from the regression line in the figure, represents (ΔZ/ΔH)κa just one month before a great earthquake with M=7.5, which occurred off the northeast coast of Honshu, Japan (Sanriku-oki). The epicentral distance Δ from Kakioka is far beyond the maximum boundary of Δ considered here. This earthquake is contained in the first long-interval change at Kakioka said above,while the second principal long-interval change, Aug. 1961-Mar.1962,also is fully considered in the formula (1) in respect to the earthquake in Apr. 1962. The third long-interval change of June 1962-0ct. 1962 contains an earthquake with intensity grade V, now M not yet determined, in Miyake Island where the great eruption of the volcano Oyama occurred in Aug. 1962. From these facts it may be suggested that both short-and long-interval changes may be discussed on the same basis about the mechanism.

[Full text (PDF; in Japanese; size:582KB)]