11:00 AM - 11:15 AM
[MIS15-26] Use of oxygen isotope ratios of tree rings to improve the continuity of drought and long rain chronologies in historical documentary records
Keywords:Tree-ring, Cellulose oxygen isotope ratio, Drought, Long rain, Historical documentary records
JPGU2023_English
Use of oxygen isotope ratios of tree rings to improve the continuity of drought and long rain chronologies in historical documentary records
Introduction
Historical documentary record is one of the most useful tools for climate reconstruction of the past. Since the Edo period (1603-1868), clans and merchants throughout Japan have recorded diaries, which are effective tools for reconstructing climates before the start of modern meteorological observation. However, those records are often qualitative and fragmentary data affected by the writer subject.
On the other hand, tree rings have been actively studied for quantitative climate reconstruction. In particular, measurement of oxygen isotope ratios (δ18O) contained in tree ring cellulose have been adopted worldwide because of its relatively good availability and high temporal resolution.
In this study, we investigate a method to improve the temporal continuity of drought and long rain chronologies reconstructed from historical documentary records in the Tono region of Gifu Prefecture, Japan, using tree-ring δ18O and observation data from a meteorological observatory.
Materials and method
Tono region, which is located in the eastern part of Gifu Prefecture, was developed along the Nakasendo road in the Edo period (1603-1867). Therefore, there are many diaries of clans and merchants in this area. We investigated 21 documents of municipal and town history published by compiling historical documentary records in the Tono region.
The Tree-ring cellulose δ18O is negatively correlated with atmospheric humidity. The sample tree of this study is a Japanese cedar (Cryptomeria japonica) that grew in Okute village in Mizunami City, Gifu Prefecture, Japan. Cellulose plates of transverse section were prepared from the disk samples, and after dividing each annual ring into 12~1 intra-ring segments. δ18O ratios were measured using an on-line pyrolysis elemental analyzer and isotope ratio mass spectrometer installed at the Graduate School of Environmental Studies, Nagoya University.
Results and Discussion
The measurement of tree-ring δ18O yielded continuous data for approximately 400 years. By comparing these measurements with the mean relative humidity observed at the Gifu Meteorological Observatory from 1883 to 2020, the growing season of the sample tree was estimated to be from early May to mid-July.
10 drought years and 6 long rain years were found from 1700-1900 period as the years for which descriptions were found in multiple documents with the season of occurrence. Descriptions on the season of occurrence of the events were mainly from May to September of the new (solar) calendar year. We calculated the average values of δ18O for the 10 drought years(D), the 6 long rain years(R) and all years through 1700-1900(A) for each intra-ring segment. By comparing D, R and A, it was found that D showed higher values than A for the fourth and fifth segments and R showed lower values than A for the first to fifth segments (Figure 1). This result is consistent with the historical records on the season of occurrence of drought and long rain.
Using this result, we set threshold of +0.4‰ to A for fourth segment and +0.8‰ to A for fifth segment to extract drought years objectively using δ18O data. Using this threshold, years of 1914,1919,1920 and 2018 were extracted from 1883-2020 period as drought years. Comparing the monthly relative humidity averages for these four years with the averages for the period 1883-2020, the drought years showed similar or lower values throughout the tree growing season (Figure 2a).
Similarly, we attempted to extract long rainfall years. Using the result, we set threshold of -1.75‰ to A for third segment and -2.0‰ to A for forth segment to extract long rain years objectively using δ18O data. Using this threshold, years of 1929,1931 and 1946 were extracted from 1883-2020 period as long rain years. Comparing the monthly relative humidity averages for these three years with the averages for the period 1883-2020, the long rain years showed higher values in April and May (Figure 2b).
Conclusion
In this study, we compared the variation in tree-ring δ18O with descriptions of drought and long rain in historical documents in the Tono region, and examined methods to improve the continuity of chronologies of drought and long rain by applying threshold values of δ18O to extract drought and long rain years. Extracted drought years were found to be consistent with observed humidity data. However, further research is needed to determine the optimum threshold value and its application to other regions.
Use of oxygen isotope ratios of tree rings to improve the continuity of drought and long rain chronologies in historical documentary records
Introduction
Historical documentary record is one of the most useful tools for climate reconstruction of the past. Since the Edo period (1603-1868), clans and merchants throughout Japan have recorded diaries, which are effective tools for reconstructing climates before the start of modern meteorological observation. However, those records are often qualitative and fragmentary data affected by the writer subject.
On the other hand, tree rings have been actively studied for quantitative climate reconstruction. In particular, measurement of oxygen isotope ratios (δ18O) contained in tree ring cellulose have been adopted worldwide because of its relatively good availability and high temporal resolution.
In this study, we investigate a method to improve the temporal continuity of drought and long rain chronologies reconstructed from historical documentary records in the Tono region of Gifu Prefecture, Japan, using tree-ring δ18O and observation data from a meteorological observatory.
Materials and method
Tono region, which is located in the eastern part of Gifu Prefecture, was developed along the Nakasendo road in the Edo period (1603-1867). Therefore, there are many diaries of clans and merchants in this area. We investigated 21 documents of municipal and town history published by compiling historical documentary records in the Tono region.
The Tree-ring cellulose δ18O is negatively correlated with atmospheric humidity. The sample tree of this study is a Japanese cedar (Cryptomeria japonica) that grew in Okute village in Mizunami City, Gifu Prefecture, Japan. Cellulose plates of transverse section were prepared from the disk samples, and after dividing each annual ring into 12~1 intra-ring segments. δ18O ratios were measured using an on-line pyrolysis elemental analyzer and isotope ratio mass spectrometer installed at the Graduate School of Environmental Studies, Nagoya University.
Results and Discussion
The measurement of tree-ring δ18O yielded continuous data for approximately 400 years. By comparing these measurements with the mean relative humidity observed at the Gifu Meteorological Observatory from 1883 to 2020, the growing season of the sample tree was estimated to be from early May to mid-July.
10 drought years and 6 long rain years were found from 1700-1900 period as the years for which descriptions were found in multiple documents with the season of occurrence. Descriptions on the season of occurrence of the events were mainly from May to September of the new (solar) calendar year. We calculated the average values of δ18O for the 10 drought years(D), the 6 long rain years(R) and all years through 1700-1900(A) for each intra-ring segment. By comparing D, R and A, it was found that D showed higher values than A for the fourth and fifth segments and R showed lower values than A for the first to fifth segments (Figure 1). This result is consistent with the historical records on the season of occurrence of drought and long rain.
Using this result, we set threshold of +0.4‰ to A for fourth segment and +0.8‰ to A for fifth segment to extract drought years objectively using δ18O data. Using this threshold, years of 1914,1919,1920 and 2018 were extracted from 1883-2020 period as drought years. Comparing the monthly relative humidity averages for these four years with the averages for the period 1883-2020, the drought years showed similar or lower values throughout the tree growing season (Figure 2a).
Similarly, we attempted to extract long rainfall years. Using the result, we set threshold of -1.75‰ to A for third segment and -2.0‰ to A for forth segment to extract long rain years objectively using δ18O data. Using this threshold, years of 1929,1931 and 1946 were extracted from 1883-2020 period as long rain years. Comparing the monthly relative humidity averages for these three years with the averages for the period 1883-2020, the long rain years showed higher values in April and May (Figure 2b).
Conclusion
In this study, we compared the variation in tree-ring δ18O with descriptions of drought and long rain in historical documents in the Tono region, and examined methods to improve the continuity of chronologies of drought and long rain by applying threshold values of δ18O to extract drought and long rain years. Extracted drought years were found to be consistent with observed humidity data. However, further research is needed to determine the optimum threshold value and its application to other regions.