Natures of Conflagration in Japan

Abstract

Conflagrations are defined in this study as “fires extending over a considerable area and destroying numbers of houses-say more than 50 houses and 3,300m² in floor-area burnt down”.
Table Nos. 1 and 2 list 14 conflagrations occurring in Japan since March 1934, which involved 500-4,000 houses respectively, The tables show the conditions at the time the fires broke out, such as wind velocity, temperature, humidity, and duration of strong wind blowing through the fires.
The 14 conflagrations have been analyzed from three angles as follows :
A) Statistical approach
The conflagrations were disposed to begin in Kagoshima prefecture-the southern part of Japan-in the winter season, and to end in Hokkaido early in the summer season, gradually going northeastwards. Seasonally March and April had the highest fire frequency during the year.
A regional deviation of conflagration frequency was also found ; the prefectures in Tohoku facing the Japan Sea such as Aomori Akita, etc. were especially focused on with wide-spreading fires.
B) Weather conditions
1) Temperature : Atmospheric temperature had little effect on the occurrence of the conflagrations under review, but fires occurring in a foehn were always aggravated by the warm wind.
2) Humidity : Atmospheric humidity has a vital relation to the outbreak of a fire, but has relatively little effect upon developing it into a conflagration. Table 1 shows that ten of the conflagrations had a rather high humidity from 42～84 percent at the time the fires started, and that the remaining four were assisted by a low humidity from 23～33 per cent.
3) Atmospheric pressure : Atmospheric pressure is liable to cause the foehn in the districts facing the Japan Sea under certain conditions of typhoon in connection with the geographic features, promoting conflagration-potential of a fire happened to exist at such time.
4) Wind velocity : Wind velocity is a decisive factor in spread of fire. The duration of the high velocity wind which contributed to the fire spread varied from 2.5 to 26 hours in the 14 conflagrations listed. In the case of Iida City conflagration, a gust of wind due to the geographical features was considered one of the serious factors in developing the fire.
C) Spot-fires
If there were no spot-fires, there could be fewer expectation of conflagrations. The author tried to take actual examples of spot-fires which had effect upon fire spread, and made Fig. 2 showing the correlation of the distance of spot-fires with the wind velocity in the 14 conflagrations which were effectively developed by spot-fires.
Fig. 2 shows that when wind velocity was higher than 15 metres per second (Group “A” in the Fig.2), the areas within 99 metres from the main fire were most frequented by spot-fires, and that when the wind velocity was lower than 15 metres per second (Group “B” in the Fig.2), the spot-fires were mostly distributed over the areas from 100～299 metres distant from the main fire.