On the Persistence of Weather at Tyosi

Abstract

The investigation of the persistence of weather using various kinds of meteorological elements was made by Prof. Fujiwhara and others. By the same method the author investigated the persistence of weather at Tyosi from 1926 to 1930, for the four seasons separately. It was assumed that the weather for both 30 minutes before and after the observed time is represented by that at the time.
The following results were found,
i) f₁p and f₂(1-p) increase with n. where p: the probability of one rainy hour.
f₁(f₂): the coefficient to increase the probability of the occurrence of the second rainy (no rainy) hour.
n: continued hours of _??_ weather
ii) f₁p is large in spring and winter, small in autumn and summer; f₁p and f₂(1-p) are large in summer and winter, small in spring and autumn; p and f₂ are large in autumn and spring, small in winter and summer.
p, f₁, f₂, f₁p and f₂(1-p) are respectively 0.130, 6.50, 1.12, 0.847 and 0.976, if they are calculated by the values of 1_??_24 rainy hours in year.
iii) Mean continued hours of no rain are 26_??_37 hours, those of rain 4_??_5 hours.
iv) Wn, t for no rain increase rapidly until n=ca. 25, then slowly, Those for rain have the maximum values at n=4_??_15. Where Wn, t: The probability of (t+0.5) rainy (or no rainy) hours after (n-0.5) rainy (or no rainy) hours.
v) The distributions of continued hours of weather in year are shown by next formulas. F_Nn=500.0n^-1.91……no rain F_Rn=257.0n^-1.11……rain
vi) The sum of frequencies in year if they are calculated per minute, are 1.5 times of the sum of the former. Maximum frequency for rain is observed at 30^m, no rain at 10^m. The sum of actual continued hours are the same to the former.
vii) (a) f_D2=f₂
(b) f_D1=f₂ when rain is observed perio dically every day.
(c) P=0.437 if it's calculated by the values of 1_??_300 no rainy hours in year, where, f_D1(f_D2): the coefficient to increase the probability of the occurrence of the second rainy (no rainy) day. P=the probability of one rainy day.
vii) It seems that an intermittent rainfall must be considered as a continuous rainfall for the above calculation.