At his laboratory in a wooded grove in northern Kyoto, Takeshi Nakatsuka holds up a vacuum sealed bag. Inside, bobbing in a bath of brown water, is a glistening disk the size of a dinner plate and the color of rich gravy. This soggy circle is the remnants of a 2,800-3,000-year-old tree, recovered from a wetland – water included, so the spongy wood does not deform – in Japan’s Shimane Prefecture, just north of Hiroshima. Within this ancient trunk lie secrets that can help us prepare for the future.
Nakatsuka, a palaeoclimatologist at Japan’s Research Institute for Humanity and Nature, along with a diverse team of 68 collaborators, has spent the last decade developing a novel method to reveal bygone precipitation patterns and interpret their effect on society. The results offer unprecedented insight into 2,600 years of Japanese rainfall patterns. By teasing out information locked inside the preserved wood of ancient forests, they are able to reveal just how much rain fell around the country over the past two and half millennia. It is an extraordinary record.
About every 400 years, the researchers found, the amount of rain falling on Japan would suddenly become extremely variable for a period. The nation would toggle between multi-decadal bouts of flood-inducing wetness and warmer, drier years that were favorable for rice cultivation. As the rains came and went, Japanese society prospered or suffered accordingly.
Palaeoclimatologist Takeshi Nakatsuka is using information preserved inside ancient tree stumps to learn about Japan’s climate in the past (Credit: Rachel Nuwer)
“Multi-decadal variability provides us with the chance to transform as well as the chance to collapse,” Nakatsuka says. Regardless of the outcome, he emphasises that such change caused large amounts of stress for the people who lived through it.
As weather patterns today increasingly defy expectations, this window into past climate variability hints at what may be in store for us in the coming years
As weather patterns today increasingly defy expectations and extreme events become more frequent and severe, this window into past climate variability hints at what may be in store for us in the coming years. “Today is not different than 1,000 or 2,000 years ago,” Nakatsuka says. “We still have the same lifespans and we are still facing large, stressful multi-decadal variation.”
Nakatsuka builds a picture of what happened in the past using a number of proxies, including tree rings, corals, stalagmites, ice cores and sediment. But his latest findings, which he and his colleagues are currently preparing for publication, primarily rely on a new method that uses isotope ratios contained within wood to estimate precipitation patterns.
Central Japan is a perfect location for such a study because of the multitude of hinoki, a type of long-lived cypress. Nakatsuka’s study includes data from 68 hinoki, whose samples he sourced from living trees, buried logs, wooden temples, coffin boards and more. All of the wood ranged in age from 100 to 1,000 years.
The ratio of oxygen isotopes in the tree rings within the wood help to link it to environmental conditions in which it grew. On dry days, leaves lose more water and are left with a higher isotope ratio than on wetter ones, helping to give information about the relative humidity in the atmosphere.
“This is a very simple but very strict relationship,” Nakatsuka says. Modern meteorological databases confirmed that the isotope ratios of the most recently-lived trees in his dataset did indeed provide an accurate read on summer precipitation.
Hinoki, a type of long-lived cypress that grows in many parts of central Japan, record the yearly changes in rainfall (Credit: Getty Images)
Isotope signatures, it turns out, also serve as time’s fingerprints: they are unique to the year in which they were created. Nakatsuka worked backwards, starting from a living tree whose age he knew. He used archeological and historic clues to approximate the centuries in which new tree samples lived. He then lined up their individual isotope signatures with other trees in his database that lived around the same time until he found the matching, overlapping pattern they shared. In this way, he stitched together a cohesive timeline from 600 BC to 2000 AD, creating a master chronology.
“Every tree in the master chronology is connected to the present,” Nakatsuka says. “It’s very accurate but time consuming and extensive work compared to traditional tree ring studies.”
While his timeline was able to reveal the erratic rise and fall of precipitation levels every 400 years or so, it didn’t tell Nakatsuka anything about what caused these oscillating patterns. The changes in rainfall he saw occurring every few decades closely matched previous data from conventional tree ring studies, though, and the multi-centurial and millennial patterns also lined up well with many previous reconstructions of past temperature fluctuations in East Asia and the world.
Collaborating with archaeologists and historians, Nakatsuka has been able to unravel what effect these changes in rainfall had on the people who lived at the time. Rainfall patterns over shorter and longer timescales corresponded, for example, to medieval ceremonies led by celebrity priests who prayed for rain. The development of irrigation systems and cooperative groundwater technologies meant to protect against drought also occurred at times when his record showed rainfall was low. As did the creation of government policies designed to rescue subjects from starvation during periods of famine. Most importantly, multi-decadal rainfall fluctuation neatly bookended major epochs in Japanese and Chinese history.
We archaeologists thought of the state formation process mainly in terms of social change, but now we can understand that floods are the background of such social change – Kunihiko Wakabayashi
“Before Nakatsuka’s analysis, we archaeologists thought of the state formation process mainly in terms of social change,” says Kunihiko Wakabayashi, a prehistorical archeologist at Doshisha University in Kyoto, who studies distributions of ancient human habitats around Osaka. “But now we can understand that floods are the background of such social changes.”
During the Yayoi period (1000 BC to 350 AD), for example, most human settlements near the Yoda River in Central Japan occurred in lowland delta areas. Rice cultivation began at that time and became a central part of life. People built peat homes alongside small rice paddies and tended to their plots individually. If waters shifted, people simply moved their homes to a nearby site, avoiding any large-scale upheaval.
By slicing the wood into thin samples, the researchers can extract cellulose for isotope analysis (Credit: Rachel Nuwer)