Abstract
1. In the previous study, the writer defined the extraordinarily heavy rainfall by day with its 24 hours amount exceeding 10% of the annual totals in the respective years. In this definition, relative quantity is adopted instead of the absolute one because the former is of more practical importance especially for flood damage. Loss or destruction of forests, cultivated lands etc. by the water mass from the falling rains does not necessarily depend upon its abso-lute quantity. All the plant and animal lives adapt themselves to their environment, especi-ally to the normal condition of climate. For example, plant communities on the mountain slope have been sustained by its sunshine, temperature, precipitation, winds etc. having pre-vailed in the long past ages until now. In a region favoured with not so much rains might rather injure their existence even though their amounts are not so remarkable. Therefore the critical limits of precipitation for the decay and destruction of plants and crops are not the same at the places with copious rainfalls against the areas with not so much rains. For such a reason, the relative quantity partakes of important meaning in the practical use.
2. However, the definition adopted in the previous paper may be too high for the rain-falls which caused the frequently experienced loss of human products and some objections or opinions against this definition were presented in the discussion at the symposium meeting held by the Meteorological Society of Japan in 1967.
For the better expression of an actual situation, therefore, limit value for the heavy rain-fall should be slightly lowered. Some one recommended, for this limit, the daily rainfall ex-ceeding 5%, instead of 10%, of the annual totals and this is considered to be more practi-cally significant. In this paper, therefore, this new standard is used. That's all very well in theory, however, it does not work in practice, because this method contains some disadvan-tages for the statistical work. In the “Report of Rainfall in Japan” or “Uryôhôkoku, ” monthly amounts of rain for every year together with the maximum daily amount in the year are ta-bulated. Therefore, although the daily rainfall exceeding 10% of the total amount for the respective years is easily picked up, the real number of rainfalls with more than 5 % of the annual totals cannot be found out because the latter occurs two or three times during a year. In this paper, it is applied only to the maximum daily rainfall for the respective years and hence the number is generally fewer than the real one. The number of heavy rainfalls is reckoned for fifty years from 1911 to 1960, but the number of observed years was different in each other at about 1400 places. For this reason, all the number was converted to the value for 100 years.
3. Fig. 1 is a distribution map of heavy rainfall frequency estimated for 100 years ac-cording to the new definition which shows a fairly singular pattern. The most important results are:
1) There exist 36 places where the heavy rainfalls occurred 100 times for 100 years, situating in southern Kyushu, Shikoku, Kantc Mountainland etc.
2) High frequency of heavy rainfalls occurred at the places in the districts of Kyushu, Nankai, Inland Sea or Setouchi and Kii Peninsula followed by Tôkai, Kantô, eastern part of TQhoku and Hokkaido facing the Pacific Ocean.
3) Hokuriku and San'in districts facing the Japan Sea are characterized by the least occurrence of heavy rains and really none at Sabae (Fukui Pref.) and Awashima (Niigata Pref.)
Fig. 2 obtained by using the previous definition is shown for the comparison with Fig. 1. There does not exist a very large difference between them and the general character or pat-tern is fairly similar to each other.
