In the previous paper, the author deduced dispersive wave forms expected at various distances from the origin due to various initial pulses and dispersions. He investigated the validity of the conventional method of analysis of observed results for determining the crustal structure. In the present paper, he assumes a simplified dispersion curve which is similar to a rigorous one for some seismic surface waves and investigates the method of analysing observed results, that will give the best agreement to the true dispersion curve of surface waves in the crust under consideration. The results are as follows;
(1) At a certain epicentral distance, the duration of oscillations of seismic waves is independent of the thickness of the layer but the larger is the thickness of the layer, the longer become the periods of waves. If the thickness is the same, the duration becomes the longer, the larger is the epicentral distance. In the case of the same distance and thickness, the larger is the velocity- or rigidity-ratio of lower and upper media, the shorter are the periods of waves.
(2) The relation between group-velocity (
Un) and period (
Tn) should be taken as
Un=Δ/2[1/
tn+1/
tn+1],
Tn=2(
tn+1-
tn)(cf. Fig. 8).
(3) The relation between phase-velocity (
cn) and period should be taken as
cn=δ/2[1/
t2, n-
t1, n+1/
t2, n+1-
t1, n+1],
Tn=(
t1, n+1-
t1, n)+(
t2, n+1-
t2, n) (cf. Fig. 9).
In the case of tripartite array, we can obtain the corresponding relation by the equation (41) and (42) in this paper. In this case, we must take δ smaller than
cn×
Tn, in order to avoid misidentification of troughs and crests for different stations.
抄録全体を表示