It has been said that the number of occurrence of aftershocks in a prescribed time interval decreases according to the following formulae,
n(
t)=
A/
t+
c (1) or
n(
t)=
kt-p (2)
where
t means the time elapsed from the main shock occurrence, and
A,
c,
k,
p are constants.
On the other hand, it has also been known that there exists a number of big fluctuations on the observed number of aftershocks compared to the number calculated by means of the upper formulae.
In order to investigate this fluctuation, the time interval of the consecutive aftershocks has been adopted as a main variable.
In this paper, the writer described the frequency distribution of the time interval between consecutive aftershocks, and tried to illustrate it by means of the stochastic process.
The results obtained are as follows.
(1) Time interval (τ) distribution is empirically expressed by the following formula at the prescribed time interval
T,
φ(τ)
dτ=μ
e-μτdτ (3)
where τ is not so small.
This means that the occurrence of aftershocks, at a suitable time interval
T, is random, i. e., an aftershock does not exert its influence on the consecutive one, namely, μ(τ), which means the non-conditional probability of aftershock occurrence at unit time interval, has no relation with τ,
μ(τ)=μ.
(2) In case of small τ, the time interval distribution is expressed by the formula
φ(τ)
dτ=
aτ
e-1/2aτ2dτ. (4)
This means that μ(τ) increases according to the time elapsed (τ) from the previous aftershock occurrence,
m(τ)
dτ=ατ
-βdτ.
This is known as TOMODA's distribution of the time interval using the name of the first investigator of this problem.
The writer showed that TOMODA's distribution can be derived using formula (3) or (4) and (2).
View full abstract