Page 232 - Changing Living Spaces

Satoshi Murayama, Hiroko Nakamura, Noboru Higashi and Toru Terao

even for a single village, there were drought-affected districts and oth-
er areas where irrigation could be used as usual; therefore, farmers were
able to successfully adjust to the administrative policy of the village and
minimize the damage. This shows that the rainfall and irrigation system
affected the micro level, as they could manage the agricultural bases of
the whole village in one unit, which enabled successful drought preven-
tion. Floods also occurred in 1817.
On the other hand, drought damage also occurred in the year when
prayer rituals for rain were frequent, often resulting in a bad harvest.
Typical examples of this were the years 1799 and 1818. On the other hand,
the year 1798 had the second highest number of days with rain rituals
among the records from the observed 21 years. Despite this fear of water
scarcity, there appears to have been no drought damage. This is considered
a matter related to the amount of rainfall or its timing. Unfortunately,
since there were no measurements of rainfall in Japan in the early mod-
ern period, we consider here the timing and frequency of rainy/cloudy or
sunny, i.e. wet or dry, days.
Natural disasters such as earthquakes and storms were also noted in
the diaries at the top of each day, as the village head briefly noted the day’s
weather after the date description, using terms such as ‘fine’, ‘cloudy’,
‘rainy’, ‘storm’, ‘north wind’, ‘south wind’, ‘severe wind’, and ‘earthquake
in the afternoon’. On 15 May 1797 (Kansei 9), for example, we find in the
lunar calendar the description ‘cloudy weather, southwest wind, calm
wave’. Similarly, on June 19, the description was ‘clear weather, north-
west wind’ In addition, on July 23, the description was ‘sunny, easterly
wind, daytime evening rain’. The description of 2 July 1803 (Kyowa 3), as
mentioned above, was high waves and ‘rainy, south-southward from the
southeast and big wave’.
To quantify these weather statements, we assigned 2 points for rain,
1 point for clouds, 0 points for sunny days, 2.5 points for heavy rain, and
1.5 points for weak rain. Figure 5 shows a drought year in 1799, a flood
year in 1803, a bad harvest year in 1816, and a normal harvest year in
1804. The difference in rainfall from May to August was determined us-
ing the monthly average points. Adequate rainfall and sunshine during
these months were critical factors for rice cultivation and production. All
lunar calendar data were converted to solar history data.
Figure 6 shows that the quantified, descriptive weather information
from the diaries corresponds excellently with the years of the 1803 flood
and the 1799 drought. It can also be seen that a normal rice harvest was

230