Presentation Information
[O12-P16]Reconstructing Edo period Temperatures by Comparing Temperature Observations from the Reikenkobo and the Hirosaki Domain Edo Diary
*Miyuna Nakano1, *Yuki Maki1 (1. IKEDA GAKUEN Ikeda Junior and Senior High School)
Keywords:
Reikenkobo,Hirosaki Domain Edo Diary,famine,Maunder Minimum
1.Motivation
This year, we analyzed the Reikenkobo records which are a 17year record of astronomical observations compiled by astronomer Shibukawa Kagesuke under the orders of the Edo Shogunate.(Fig.1)
2.Purpose of the study
The Reikekobo records contain temperature observations using a thermometer obtained from the Netherlands.We used temperature records from the Reikenkobo records to quantitatively reconstruct Edo period temperatures, and clarify the relationship between weather fluctuations and famines. (Fig.2)
3.Research methods
We would like to explain our database creation and weather classification.
We classified the weather in the historical documents to make the classifications closer to the definitions currently used by Japan Meteorological Agency.
The weather was classified as snowy, rainy, cloudy, or sunny.
4.Data processing
The acquired data is comprised of 5,788 days over 17 years for the Reikenkobo records, and 52,554 days over 201 years for the Edo Diary.
5.Monthly Average Temperature (℃) for noon in the corrected Reikenkobo records
(1)Temperature variations throughout the day
(2)Next, the location for weather observations in the Reikenkobo records was moved once. Altitude correction
(3)Finally, temperatures recorded in Fahrenheit were converted to Celsius.
6. Research and Discussion
(1)Research① and Discussion:Review of Previous Research" (Fig.4)
The correlation coefficient between Tokyo's monthly average daily maximum temperature and the number of sunny days was 0.68. (Fig.5)
(2)Research②・③and Discussion:Comparison of July sunny day occurrence rates in two diaries (1838-1855)
Comparing the frequency of sunny days in July between the Reikenkobo records and the Edo Diary yielded a correlation coefficient of 0.83. (Fig.6)
Therefore, when examining the relationship with the temperature at noon in July, we found that there is a stronger positive correlation between the frequency of sunny weather in the Edo Diary and the temperature in the Reikenkobo records than between the frequency of sunny days in the Reikenkobo records and the temperature.
(3)Research ④ and Discussion:The average temperature on July, reconstructed from the Edo Diary (1668–1868)
A strong correlation was observed between the frequency of sunny days in the Edo Diary
and temperatures recorded in the Reikenkobo, enabling us to reconstruct July temperatures for the noon hour.
Despite missing data for some years, the average temperature from 1668 to 1859 was 27.2℃, with a standard deviation of 1.5℃.
The highest temperature was 31.0℃ in 1693, identified as a major drought year by Dr. Kondo's research.
Reports of drought exist across Japan for 1693.The lowest temperature was 23.3℃ in 1836, the year of the Tenpo famine.
The average temperature in July during the Maunder Minimum was 28.2℃, which is 1.0℃ higher than the average for the entire Edo Diary.
Thus, while temperatures in July tended to be high during the Maunder Minimum, they dropped during the four famine years, which is consistent with historical records. (Fig.8)
This result is also consistent with our research findings from last year. (Fig.9)
According to the analysis by Dr.Yamaguchi, the oxygen isotope composition of tree rings during the Maunder Minimum indicates that Japan experienced wet conditions, and that Greenland and the northern hemisphere experienced cold conditions.
Thus, while temperatures in July tended to be high during the Maunder Minimum, they dropped during the four famine years, which is consistent with historical records.
(4)Research ⑤ and Discussion“Average August noon temperatures reconstructed from the Edo Diary (1668–1868)
We compared the August temperatures from the Reikenkobo records with the frequency of sunny days recorded in the Edo Diary. Using the interquartile range to identify and exclude outliers from the 1846 sunny days occurrence rate yielded a coefficient of determination of 0.66.
Applying this regression equation to reconstruct the entire period revealed temperature declines during the four famines of the Maunder Minimum, consistent with historical records. (Fig.10)
11.Summary
Years of temperature decline during the Maunder Minimum corresponded with the years of the four famines.
Quantitative temperature reconstruction using the Reikenkobo records has rarely been used in previous research.
12.Future Prospects
The Hirosaki Domain Edo Diary had missing years, so we attempted to reconstruct more accurate temperatures by comparing it with the Sakakibara Domain Diary and the Edo Shogunate Diary, both also written in Tokyo.
This year, we analyzed the Reikenkobo records which are a 17year record of astronomical observations compiled by astronomer Shibukawa Kagesuke under the orders of the Edo Shogunate.(Fig.1)
2.Purpose of the study
The Reikekobo records contain temperature observations using a thermometer obtained from the Netherlands.We used temperature records from the Reikenkobo records to quantitatively reconstruct Edo period temperatures, and clarify the relationship between weather fluctuations and famines. (Fig.2)
3.Research methods
We would like to explain our database creation and weather classification.
We classified the weather in the historical documents to make the classifications closer to the definitions currently used by Japan Meteorological Agency.
The weather was classified as snowy, rainy, cloudy, or sunny.
4.Data processing
The acquired data is comprised of 5,788 days over 17 years for the Reikenkobo records, and 52,554 days over 201 years for the Edo Diary.
5.Monthly Average Temperature (℃) for noon in the corrected Reikenkobo records
(1)Temperature variations throughout the day
(2)Next, the location for weather observations in the Reikenkobo records was moved once. Altitude correction
(3)Finally, temperatures recorded in Fahrenheit were converted to Celsius.
6. Research and Discussion
(1)Research① and Discussion:Review of Previous Research" (Fig.4)
The correlation coefficient between Tokyo's monthly average daily maximum temperature and the number of sunny days was 0.68. (Fig.5)
(2)Research②・③and Discussion:Comparison of July sunny day occurrence rates in two diaries (1838-1855)
Comparing the frequency of sunny days in July between the Reikenkobo records and the Edo Diary yielded a correlation coefficient of 0.83. (Fig.6)
Therefore, when examining the relationship with the temperature at noon in July, we found that there is a stronger positive correlation between the frequency of sunny weather in the Edo Diary and the temperature in the Reikenkobo records than between the frequency of sunny days in the Reikenkobo records and the temperature.
(3)Research ④ and Discussion:The average temperature on July, reconstructed from the Edo Diary (1668–1868)
A strong correlation was observed between the frequency of sunny days in the Edo Diary
and temperatures recorded in the Reikenkobo, enabling us to reconstruct July temperatures for the noon hour.
Despite missing data for some years, the average temperature from 1668 to 1859 was 27.2℃, with a standard deviation of 1.5℃.
The highest temperature was 31.0℃ in 1693, identified as a major drought year by Dr. Kondo's research.
Reports of drought exist across Japan for 1693.The lowest temperature was 23.3℃ in 1836, the year of the Tenpo famine.
The average temperature in July during the Maunder Minimum was 28.2℃, which is 1.0℃ higher than the average for the entire Edo Diary.
Thus, while temperatures in July tended to be high during the Maunder Minimum, they dropped during the four famine years, which is consistent with historical records. (Fig.8)
This result is also consistent with our research findings from last year. (Fig.9)
According to the analysis by Dr.Yamaguchi, the oxygen isotope composition of tree rings during the Maunder Minimum indicates that Japan experienced wet conditions, and that Greenland and the northern hemisphere experienced cold conditions.
Thus, while temperatures in July tended to be high during the Maunder Minimum, they dropped during the four famine years, which is consistent with historical records.
(4)Research ⑤ and Discussion“Average August noon temperatures reconstructed from the Edo Diary (1668–1868)
We compared the August temperatures from the Reikenkobo records with the frequency of sunny days recorded in the Edo Diary. Using the interquartile range to identify and exclude outliers from the 1846 sunny days occurrence rate yielded a coefficient of determination of 0.66.
Applying this regression equation to reconstruct the entire period revealed temperature declines during the four famines of the Maunder Minimum, consistent with historical records. (Fig.10)
11.Summary
Years of temperature decline during the Maunder Minimum corresponded with the years of the four famines.
Quantitative temperature reconstruction using the Reikenkobo records has rarely been used in previous research.
12.Future Prospects
The Hirosaki Domain Edo Diary had missing years, so we attempted to reconstruct more accurate temperatures by comparing it with the Sakakibara Domain Diary and the Edo Shogunate Diary, both also written in Tokyo.