Abstract
At the early decade of this century, the warming of world-wide climate with its peak in the fifth decade, especially in the middle and high latitude regions in the northern hemisphere, was reported by some climatologists and meteorologists. It was also cofirmed in Japan by H. Arakawa (1936) for the period of fifty years preceding to 1935. In this paper, the writer intends to make clear the temperature trend for thirty years since 1936 as related to the world-wide change, influence of urban development on the secular variation of its temperature and some other subjects. The main results obtained are summarized as follows:
1) As is shown in Fig. 1., annual temperature in Tokyo is approximately expressed by the following simple equation,
θ=a+bt
where θ is an annual temperature for the t th year, a and b the numerical constants. In this paper t is reckoned in such a way as 1, 2, 3 ……for 1951, 1952, 1953……and -1, -2, -3, ……for 1950, 1949, 1948……respectively.
The same equation can be adopted at all the places in Japan. The most probable values for a and b are determined for fifty seven stations in Japan by the method of least squares (Table 1.)
Excepting a few places, magnitude of a which means the average annual temperature for this thirty years becomes larger than that for the preceding fifty years which is shown by the numerical figures enclosed by brackets. In Tokyo, for example, it grew up to 14.8°C or 0.9°C higher than the preceding period. Also as for b which means the average amount of temperature rise per year, similar trend is recognized with a more striking tendency.
Namely it increased very rapidly since the preceding period in some giant cities as are given in the following examples:
Tôkyô from 0.009° to 0.032°C/year
Ôsaka 0.015 0.029
Kyôto 0.010 0.032
Hence the magnitude of b grew up two-or threefold and at Kanazawa it became eighteen times larger.
2) As already mentioned, rapid increase of temperature was experienced in the entire world. Mitchell, J. M. (1961) calculated the average rise of temperature for the zonal areas between every ten degree of latitude. Meanwhile, the average increase of temperature at fourty eight places in Japan from Kagoshima in the south and Akita in the north between 30° and 40° N. was 0.8°C, which is nearly equal to the results obtained by Mitchell. In the northern part of Japan, mainly occupied by Hokkaido, however, temperaure for this period still increased against the remarkable decrease of average temperature for the belt area between 40° and 50° N. This was probably due to the fact that the greater part of this latitude belt was invaded by the cold air-mass from the Arctic region, whereas the Japanese Islands were left alone being covered by the relatively warm air tongue as is shown in Fig. 2. More important fad t is that the subsequent fall of temperature since about 1940 in many countries in the middle and high latitudes was far delayed in Japan, faintly appearing in the latest years of this period. It is probably concluded in Fig. 2. that the warming in Japan was maintained until recent years, not influenced by the cold air invasion now extending to the east and south and therefore, the winter cold in Japan might become more severe in the near future.
3) The fact that the development of urban temperature relates to the size and population of the city was confirmed by Duckworth and Sandberg (1954) in three cities with a mutually different area and population near the San Francisco Bay in the U.S. In this paper the urban size is represented by its population for convenience sake and it is clearly recognized in Fig. 3. that the increasing rate of temperature is closely related to the square root of population for the cities with a population of less than 500, 000.
