Earthquake Measurement at Miyako

The results are collected in the Table at the end of this paper, to which the following remarks may- be prefixed. 36. The area of earthquake disturbance. Of the earthquakes whose exact dates are known, No. 4 and its fore-and after-shocks, namely, Nos. 3, 5, 6, 8, and 19, originated in the mountain regions forming the boundary between the provinces of Rikuchu and Ugo, while the remaining 19 earthquakes all originated under the Pacific Ocean at variable distances from the coast. The isoseismal lines and the areas of disturbance of these earthquakes, the typical ones of Which are illustrated in figs. 4, 5, 6, 7 and 8, were generally not circular, but more or less elliptical, with the longer axes parallel to the coast, that is to say, in a direction of N slightly E to S slightly W. In the last column of the Table, therefore, both the greatest and the least radii of propagation are given for each earthquake, the exceptions being Nos. 22, 26, 4, 5, 6 and 10, whose areas of disturbance were nearly circular. The smallest and the greatest among the 25 earthquakes here considered, were respectively No. 24 and No. 9, the radii of propagation of the former being 50 and 65 km. and those of the latter 430 and 700 km.In those cases in which the earthquakes originated under the Pacific Ocean, the shocks were generally propagated only to an insignificant amount to the western, coast of Japan. This is probably due to the seismic waves having been stopped by the heavy mountain masses running through the Main Island.Propagation along the coast is, however, evidently much favoured by the geological formations, which generally run parallel to the coast. On the other hand, it is also very probable that the epifoci of most of the earthquakes under consideration were not points, but were linear and parallel to the axis of the Island. Thus, the forms of the isoseismal lines in figs. 4 and 5, suggest very naturally that the earthquakes, Nos. 20 and 8 were probably caused by long fissures or faults in the direction of nearly NEW-SWS.37. The position of the earthquake origin.Of the 25 earthquakes above considered, six originated at distances of 100 to 130km. to the W or WSW of Miyako, while the remaining nineteen originated under the sea at distances between 40 and 210 km. to the ENE, E, or ESE of the same station. The positions of the nineteen suboceanic origins are included within a circle of about 2° radius, the centre being in long. 143° E, lat. 39° N. This point, which is at about 1° or 110 km. to the E of Miyako, is thus to be regarded as the most. active centre of earthquakes disturbing the eastern part of Northern Japan. As the majority of earthquakes disturbing Miyako, or more generally the eastern coast of Northern Japan, originate under the Pacific, it is to be expected that the shocks will sometimes, like No. 1 and No. 12, be accompanied by sea-waves. It may be that the disturbance of the sea occurs more or less whenever a sufficiently strong shock takes place, either at the sea-bottom itself or at a very small depth below it, so as to give commotion to the water. Of course, the same phenomena may be produced, when a comparatively sudden elevation or depression of the sea-bottom takes place.38. The duration of an earthquake.It is impossible to measure exactly the total duration of an earthquake, which is of course much longer when instrumentally recorded than when simply felt without such an aid. Again, the duration recorded by an instrument depends on its sensibility. Thus, for instance, the duration of motion in Eqke. No. 23 was, as given in the Table, only 120 seconds; but the duration of the same earthquake in Tokyo, recorded by Omori's Horizontal Pendulum, was 2 hours. The fact is that the ordinary seismographs of the Ewing and Milne and Gray types so widely used in Japan, are capable of registering only the quicker vibrations of an earthquake but insensitive to waves of long period, the latter being very easily obscured by the friction of the instruments.The total earthquake durations given in this paper, which varied between 8.5 and 200 seconds (see the 3rd column of the Table), are to be regarded as the durations of the sharper vibrations in different earthquakes, namely of those waves causing the greatest effects on our senses. Speaking generally, the duration of an earthquake will depend both on its magnitude and on the distance of the observing station from the origin. As a rough approximation, let us assume D=A/kπ, in which π=3.1416; k is a constant; D, the duration in seconds of an earthquake (less the duration of the preliminary tremor) at a given station; d the distance in kilometres between the origin and the observing station; and A the area of disturbance, in which the intensity of motion is sufficiently strong to be generally felt without any instrumental aid. The values of the constant k, given in the 6th column of the Table vary within rather wide limits, namely, 4 and 28, but it would be impossible to expect any very perfect agreement between several cases, as, besides the difficulty of accurately measuring the duration, D, the value of the area. of disturbance, A, is very apt to be subjected to accidental modifications, the origins being, in the present cases, mostly. submarine. The average value of k is 13.6, so that the above equation becomes, for Miyako, D=0.024×A/d. Further discussions of this question shall be made in the report on the Tokyo seismometrical observations to be soon published.39. The duration of the vertical motion. The relation of the duration of the vertical motion, measured in 14 cases, to that of the horizontal, was as follows. In 5 cases, the durations of the two components were equal to each other; in 1 case, the V. M. lasted longer than the H. M. in the ratio of 1.2:1; and in the remaining 8 cases, the duration of the V. M. was shorter than that of the H. M., the ratios being respectively 0.6:1, 0.8 : 1, 0.8: 1, 0.4 :1, 0,5: 1, 0.7: 1, 0.6: 1 and 0.6 : 1. Taking the average of these 14 cases, the durations of the vertical and the horizontal components were in the ratio of 0.8: 1.40. The duration of the principal portion of an earthquake, which is much shorter than that of the whole of it, varied in the present cases, as given in the 5th column of the Table, between 0.7 and 26.0 seconds. The principal portion is generally well-defined for comparatively near shocks, 'but often very indefinite for distant ones.41. The duration of the preliminary tremor, which has been measured in 24 cases, varied between 0 and 26 seconds. As already discussed by one of the present authors, the duration of the preliminary tremor of an earthquake, y (expressed in seconds), at a given station seems to be related to the distance of the latter from the origin, x (expressed in kilometres), by the following linear equation - 7.51y=x-24.9 km. (2) x being assumed to be sufficiently great, say, greater than 100 kilometres, or y greater than 10 seconds. The approximate distances of the epicentres from Miyako have been calculated by means of equation (2) in 11 cases, in which the duration of the preliminary tremor was greater than 10 seconds. These results are given in the 17th column of the Table, while the distances judged from the earthquake maps are given in the 16th column. It will be observed that the two sets of values given in columns 16 and 17 agree on the whole pretty well with each other. Thus, the mean value of the distances under consideration deduced from eight cases, namely, Nos. 2, 3, 12, 13, 14, 17, 19 and 23, is according to equation (2), 160 km., and according to the method of seismic maps, 180 km.42. The period of vibration. In the Table are given for different earthquakes the complete periods of maximum movements, as well as the average complete periods of vibrations in the preliminary tremor, the principal portion, and the end portion, with respect to both the ripples and the slow undulations. The period of the maximum H. M. varied, for slow undulations, between 0.53 and 1.7 seconds; and, for ripples, between 0.12 and 0.15 second. The corresponding period in the vertical motion varied, for slow undulations, between 0.53 and 1.7 seconds, the vertical and the horizontal components having, in most cases, the same period. With regard to ripples, the period, which was identical in all the three components of an earthquake, became generally a little longer in the principal portion than in the preliminary tremor. Thus, for the latter epoch, the mean value. of the average periods which varied between 0.04 and 0.087 second, was 0.08 second; while that of the average periods for the former epoch, which varied between 0.055 and 0.12 second, was 0.10 second. For the end portion, the mean value of the average periods of ripples,, which varied between 0.06 and 0.10 second, was 0.08 second, namely the same as in the case of the preliminary tremor.. The average period of ripples in the principal portion, which are characterized by their having larger amplitude, is thus seen to be longer than that in the two other portions of the earthquake motion,. the difference amounting to 0.02 second. For the principal portion, the mean value of the average periods of horizontal slow undulations, which varied between 0.56 and 1.7 seconds, was 1.3 seconds in the (EW) and 1.0 second in the (NS); while the mean of the corresponding quantities for the end portion, which varied between 0.9 and 1.7 seconds, was 1.3 seconds in (EW), and 0.94 second in the (NS). Thus; for the horizontal slow undulations, the average period seems to remain approximately constant in the principal and the end portions. These undulations existed more or less also in the preliminary tremor, the mean of the average period being 1.1 seconds in the (EW) and 1.0 second in the (NS). For the V. M., the mean of the average periods of slow undulations was about 0.8 second. It. Is remarkable that the average period of ripples, remained roughly constant in the different earthquakes here considered, never varying much. from 0.1 second. It is possible that these ripples were preceded by, or superposed upon, very minute vibrations of still shorter periods.43. The range of motion. The maximum ranges of motion in different earthquakes are given in the 9th column of the Table. The greatest ripple of 7.4 mm. occurred in Eqke. No. 11, the next greatest of about 5 and 6.4 mm. occurring respectively in Eqke. No. 4 and Eqke. No. 23. The greatest slow. undulations, which probably occurred in these two earthquakes, were unfortunately confused by the proper oscillations of the pendulums serving as the "steady masses" of the seismograph. As an example, in which this element of motion was satisfactorily measured, Eqke. No. 3 may be mentioned, the maximum range of motion in question being 9.1 mm. The range of V. M. was invariably smaller than. that of. the corresponding H.M., the maximum motions of the two components being, for ripples,, in the average ratio of 0.20: 1; and for slow undulations, in the average ratio of 0.21 : 1.44. The direction of motion. .The motion in the principal portion was, with the exception of one case or two, always greater in the (EW) than .in the (NS). For the preliminary tremor, however, no such peculiarity was found, the ratio of the ranges of motion in the two horizontal components during this epoch having generally no relation to the corresponding ratio for the principal portion. The directions of the maximum movements in different earthquakes are given in the 10th column of the Table, from which it will be observed that these were, with the exceptions of Eqkes. No. 5 and No. 16, nearly EW, ESE-WNW or ENE-WSW. (See fig. 9). It will also be seen that the majority of the suboceanic earthquake origins were situated towards the E, ESE and ENE of Miyako, so that we may, in a general way, say that the directions of the maximum earthquake movements were nearly coincident with the directions of the lines joining the observing station with the epicentres. The fact that the origins under consideration were chiefly in the directions of SE or ESE from Miyako, while the majority of the directions of the maximum movements were included between the directions EW and NE-SW, seems to indicate some peculiarity inherent in the region about the observing place, causing the direction of motion in different earthquakes to approach to a certain prevailing direction. In the cases of nine earthquakes, Nos. 1, 2, 3, 8, 10, 14, 16, 19 and 22, the directions of the maximum movements did exactly, or roughly, coinside with the respective directions of the origins from Miyako; and especially, in three earthquakes, Nos. 8, 11 and 20, the angles of emergence have been measured respectively to be 7°.2, 9° and 36°. Of these three last earthquakes, two, namely, Nos. 8 and 20, were the after-shocks of the Eqke. No. 4; while the remaining one, No. 11, had its origin under the Ocean, towards NE of Miyako. The value of 36° for Eqke. No. 20 was evidently too great, the error being probably due to its having been obtained, not from the principal portion, but from the preliminary tremor of the earthquake. Omitting this case, we obtain for the focal-depths of the two earthquakes Nos. 8 and 11, respectively the values of 15 and 9 km.