This report is a continuation of two reports [1, 2] published previously. We can adjust strictly an arbitrary geodetic network on the surface of a reference ellipsoid by using the Universal Program which is called TO16G4 in the code of electronic computer in GSI. Though the program has mainly been used, up to the present, for the purpose of finding transformation-vectors of geodetic stations, it has been improved so as to be able to find all kinds of strains since 1970. We use the coordinates system of longitude-latitude in the program, but it is not suitable for calculation of strains. Strains are computed on the plane-coordinates trans-formed through a simple formula (1) for individual triangle (Fig. 1). If the crust is distorted according to a linear transformation, a circle drawn on the earth is transformed into an ellipse after a long time as shown in Fig. 2. There are also relations (2) between old and new coordinates of three stations, and we can find six strain-constants xo, yo, a, b, c and d for every triangle. Rotation, dilatation, maximum shear, major and minor principal axes are computed by using these constants through formulas from (3) to (8).

View full abstract

We have two, old and new, surveys for the network of the first order triangulation in Japan. The old and new surveys have been carried out during the period from 1883 to 1909 and that from 1948 to 1968, respectively. Horizontal deformation of the earth's crust except the Hokkaido district was discussed mainly by means of displacement-vectors of triangulation stations in the previous papers [4, 5]. We had been observing in the Hokkaido district when the paper [4] was written. Fig. 1 shows the displacement-vectors in the Hokkaido and northern Tohoku districts found under the condition that sum of vectors is to be zero [7]. This condition gives us an easy figure to see without restraining observations artificially. The figure of displacement-vectors is convenient to see intuitively the movement of the crust. But it has a defect that it is never a unique figure because we can get a different figure of vectors by changing the fixed station. Though those figures look to express different deformations of the crust at a glance, of course they are essentially same about relative displacements among the stations. Consequently, it might be better for studying about crustal deformation to use strains. Strains for every geodetic triangle have been found by the improved Universal Program [6] on the assumption that the crust deforms according to linear functions of coordinates. Three kinds of strains; that is, dilatation, principal axes and maximum shear, are shown in Figs. 2, 3 and 4, respectively. It seems to be the most characteristic fact in these figures that there are no strains over 1/20, 000. Large strains usually correspond to the past great earthquakes. It seems to be very important that the remarkable strains in northeastern area of the Hokkaido district do not have a clear correspondence with a great earthquake.

View full abstract

The reliabilities of single observation by means of Worden, North American and LaCoste-Romberg gravity meters are checked by using many observations made at Tokyo and Chiba, where the gravity difference is small. The reliabilities of single observation for the former two and the latter one are obtained as ±6/100 mgal and ±2.5/100 mgal respectively. The secular change of scale factor makes it difficult to detect the gravity change of 1/100 mgal definitely. The ratio between the gravity change and the height change was obtained from the slope of regression line: Δg =α⋅Δh + b instead of Δg = α⋅Δh which passes through the origin.1) Tokyo, 1969 (LaCoste)-1971 (LaCoste), (Fig. 1)a = 0.36±0.09 mgal/mb = 0.01±0.01 mgals. d. =±0.03 mgal2) Matsushiro, 1966 (LaCoste)-1967 (LaCoste), (Fig. 2)a = 0.37±0.09 mgal/mb = 0.01±0.01 mgals. d. =± 0.02 mgal3) Niigata, 1959 (Worden)-1965 (LaCoste), (Fig. 3)a = 0.24±0.03 mgal/mb = 0.01±0.02 mgals. d. =±0.05 mgal ∂g/∂h(=a) can not be determined in high precision, if the number of stations and the change of height are small. As the constant term (b) means the gravity change without vertical displacement, the change of water in water-bearing layers may be detectable from the constant term. Standard deviations are not contradictory to the reliabilities mentioned above. Within the observational accuracy, dg/6h at Niigata corresponds to 3g/0h = 0.27 mgal expected from the consolidation theory, in which the infinite slab of water is taken off and the land subsides by thickness of the slab.

View full abstract

Leveling surveys between Tokyo datum station and the fixed point at Aburatsubo tidal station were carried out 18 times during the period from 1923 to 1970. The vertical movement of Aburatsubo relative to Tokyo datum station (the uppermost curve of Fig. 3) shows almost linear secular subsidence and considerably large irregular oscillation. The latter phenomenon has been studied by some authors from a standpoint of influence of change in mean sea level, but positive conclusion has not yet been obtained. In the present paper, reliability of leveling survey is re-examined in consideration of accumulating error of the survey. The leveling route in the area concerned was observed as a single line (west line) in 1923-1948, and as a closed loop after 1952. The closing error of the loop is a measure of quality of the survey, but this is not always correct, because it can be considered that there is a case in which accumulated errors of each half of loop are canceled in total. When change in observed height differences between successive two surveys is abnor-mally large, and also this change is confined in short range of time in regards with expected secular movement, it can be considered that survey of this time, or preceding survey, or both of them are erroneous. For example, when secular movement is actually very small, erroneous too large change in height difference will be proved by too large change o4., contrary direction in the next survey. In order to study the problem according to this criterion, changes in observed height difference of each part of the leveling routes in the area concerned are obtained, and are shown in Fig. 2. After examination of this figure and also of closing error, reduction of surveys are made as follows : surveys in 19231944 adopt the west route; survey in 1948 is rejected; surveys in 1952, 1955 and 1957 adopt the east route; surveys in 1965 and 1968 adopt the west route; surveys in 1960, 1963, 1970.1 and 1970.9 adopt net adjustment. The result is shown on the lowest curve of Fig. 3 as 35.1-F(A). BM 35.1 is considered to be on very stable ground. From this curve of secular movement, it is shown that the abnormal subsidence in about 1953 which is clear in D-F(A) curve, disappears. However, the subsidence of considerable large speed in the period from 1963 to 1968, and the upheaval of considerable large speed after 1968, are confirmed. Whether this phenomenon is a mere oscillation of noise in the linear subsidence of long range, or is related to some geophysical phenomena such as the coming earthquake will remain to be uncertain till future surveys will be carried out.

View full abstract

The second trial of precise gravity survey was accomplished in the Tokai and Kinki Districts in 1970. Three LaCoste & Romberg gravimeters were used in each district to get gravity values and to discuss an accuracy of the gravimeters. The survey was carried out under the predetermined routine along the levelling route. Results of the survey are summarized in the present report. The measuring accuracy of the present survey was estimated to be 0.02 mgal, and it was approximately equal to that obtained through the first precise gravity survey in the Miura and Boso Peninsulas.

View full abstract

Repetition of leveling survey is one of the most effective methods for detecting land deformations prior to earthquakes. On the basis of statistics of severe earthquakes having occurred in Japan since 1800, the optimum project of leveling survey is discussed by means of operations research.

View full abstract

Precise gravity survey was carried out at specially selected stations, where gravity difference from the Gravity Reference Station of Geophysical Institute of Kyoto University is smaller than about 4 mgal, distributed in the Kinki District, by means of 3 LaCoste & Romberg gravimeters (model G) in November 1970. Back and forth measurement was adopted, step by step, in order to determine the drift of each gravimeter as accurate as possible in the survey. But the results obtained were not more accurate than those by an ordinary gravity survey. On the other hand, an experiment for investigating the characteristics of LaCoste & Romberg gravimeters due to vibrations was made under the following conditions: (1) Gravimeters were shaken by a running motor-car. (2) Gravimeters were taken in and out their carrying cases. (3) Gravimeters were fixed at the station. The results obtained through the present investigations are as follows: (1) Reliability of single measurement by means of a single LaCoste & Romberg gravimeter is to be ±0.02-0.03 mgal. (2) Gravity survey should be made for stations located as distant as possible during a day so far as a LaCoste & Romberg gravimeter is available in gravity survey .

View full abstract

It is seen that the gravimetric values observed in the Tokai District are explicitly more accurate than in the Southern Kanto District. This fact is thought to be caused by differences of ground noise level as well as air temperature changes, which effect directly on drift of the gravimeter. The author makes an experiment in order to obtain the temperature-drift relation. Gravity measurements are continuously made, while thermisters automatically pick up temperatures in the air, on the gravimeter surface and inside the gravimeter. As a result, the author finds a tendency that gravity values observed in high air-temperature can be more accurately determined than in low air-temperature.

View full abstract

Results of precise gravity surveys obtained through cooperative measurements by the Working Group for Gravity Survey and Precise Levelling in Japan and also the Working Group for Comparing the Gravimeters in Japan are summarized in the present paper. It was ascertained, as a result, that reliability of a single field measurement by means of a LaCoste & Romberg gravimeter (model G) was to be ±0.02-0.03 mgal. On the basis of the obtained results, the following points are discussed in the present paper. (1) Accuracy of gravity survey by means of a LaCoste & Romberg gravimeter. (2) How to use a LaCoste & Romberg gravimeter in gravimetric field survey. (3) Some problems to be checked and investigated in near future in order to raise the accuracy of gravity survey.

View full abstract