By using the data of annunal rainfall in Korea for 153 years Mr. R. Sekiguchi investigated the Far Eastern seasonal correlation in connection with the solar activity (this Journal 37th year). Using the same data, the present author has investigated the correlation between the monthly amount of Korean rainfall and the solar activity. He calculated the range of average rainfall by the formula
where m is the average rainfall, m
1 and m
2 its extreme values, M is the mean value, Δ is the absolute deviation of the amount of rainfall in each year from the mean value M, and n is the number of years taken in the statistics. The result is as follows:
Then he classified the months into three classes of rich, moderate and poor precipitation. In rich months, as he understand, the amount of rainfall surpasses m
1; in moderate months it lies between m
1 and m
2; in poor months it sinks under m
2. By counting the number of months in each class, subclassified into three with respect to the sunspot number of the year, he obtained the following table.
In the above table, the winter precipitation represents the sum of those for December, January and February, because the amount for each month is too small. For summer rainfall the amounts for July August and September are dealt with separately so that the frequency is 3 times as greater in summer than in winter. The general summer tendency is that in the year of small sunspot number there is remarkable tendency to_??_cause poor rainfall (27/20=135% ei 7/20 in favour) and in the year with moderate sunspot number there is tendency (10%) against poor precipitation and in favour of (8/62=13%) moderate precipitation. Rich precipitation is not so much affected by the sunspot number as moderate or poor rainfall. It occurs, however, preferably in years of great sunspot-number in the ratio of 21/18 or3/18=17% in favour.
In winter the tendency becomes opposit. Small sunspot number rather corresponds to rich precipitation, moderate to moderate and great sunspot number to poor rainfall. In short, there is positive correlation of sunspot number with summer rainfall in Korea and negative one with winter precipitation.
The author has also noticed that in some years of small sunspot number extremely heavy rainfall occurred in July. Taking statistics he has found that extremely rich rainfall and extremely poor rainfall during the summer three months occurred just in the same frequency as shown in the following table.
The author then investigated the influence of the amplitude of the sunspot cycle on the Korean precipitation and found that in the greater sunspot cycle, the epoch of maximum and minimum rainfall shifts toward the year with greater sunspot number.
He then investigated the sunspot correlation with the spring and autumn precipitation in Korea and found the gradual shift of the correlation from winter to summer type and again from summer to winter type.
The author examined the correlation between the Korean summer rainfall and the summer temperature in Tohoku District in Japan. He has found that the sense of correlation has changed in 1910. Before that year, the correlation was negative and after that, the sense was positive. He then calculated the correlation between the summer temperature in Tohoku and summer precipitation of the preceeding year in Korea for years before 1910 and attained the positive high correlation. This must be a very significant result as to reveal the very important but curious phenomenon of the mutation of the sense of correlation or the rocking effect on correlation. (S. F.)
View full abstract