2,500-Year-Old Chronicle Is Oldest Known Record Of A Total Solar Eclipse And Reveals Some Surprises

Chinese records of a total solar eclipse in 709 BCE have provided insight into two unrelated scientific questions we might not be able to resolve any other way, and revealed a mistake no one suspected.

The rest of this article is behind a paywall. Please sign in or subscribe to access the full content.

The orbit of the Moon is so regular that we can calculate the timing of solar eclipses back a very long way. However, knowing an eclipse would occur and knowing where it would be visible are quite different things. Many factors, not all of them well understood, speed up and slow down the Earth’s rotation, which is why we sometimes have leap seconds. Marginal changes to this speed will, over time, affect where the Moon’s shadow will fall on the planet.

For decades, astronomers have used ancient eclipse records to calculate changes in Earth’s rotation before the time when we could measure it directly. The oldest such record describes the eclipse of July 17, 709 BCE, and was made by observers in Qufu, then part of the Chinese state of Lu Duchy.

We’re very lucky to know this, since no reports from the time itself survive. However, 200-300 years after the eclipse, a chronicle known as the “Spring and Autumn Annals” makes reference to the event, based on descriptions from the time.

When a team led by Dr Hisashi Hayakawa of Nagoya University attempted to verify this event, they were initially puzzled. No matter how they adjusted the planet’s spin rate, they could not place the path of totality over Qufu. That was until they realized that previous research had misidentified Qufu’s coordinates. 

A reexamination of archaeological reports led Hayakawa and co-authors to realize Qufu had been reported as being in the wrong place. Correcting these moved the city 8 kilometers (5 miles), enough to place it inside the zone of totality for certain values of the Earth’s rotation rates during centuries where these have been uncertain.

“This correction allowed us to accurately measure the Earth's rotation during the total eclipse, calculate the orientation of the Sun’s rotation axis, and simulate the corona’s appearance,” said Hayakawa in a statement.

“This new dataset fixes coordinate errors in previous Earth rotation studies,” added coauthor Dr Mitsuru Sôma of the National Astronomical Observatory of Japan. “Additionally, it improves the accuracy of dating and reconstructing historical astronomical events.”

Having determined the reports could indeed be accurate, rewritten the archaeological record, and updated planetary science, the team had one more use for the ancient accounts.

“What makes this record special isn’t just its age, but also a later addendum in the ‘Hanshu’ (Book of Han) based on a quote written seven centuries after the eclipse. It describes the eclipsed Sun as ‘completely yellow above and below.’ This addendum has been traditionally associated with a record of a solar corona. If this is truly the case, it represents one of the earliest surviving written descriptions of the solar corona,” Hayakawa said.

The authors acknowledge accounts written centuries after the event aren’t fully reliable; the corona reference could be an addition made by someone who had seen a much later total eclipse. However, getting the date right centuries later does add to the report’s credibility, and if the corona was indeed highly visible during the 709 BCE eclipse, it would indicate normal to high solar activity, such as sunspots and solar flares.

That would be an important finding, because there is evidence that such activity largely stopped during a period known as the “Neo Assyrian Grand Minimum” (or the “Homer Grand Minimum” depending on your side in ancient culture wars). Our best estimate is that this period lasted from 808 to 717 BCE, so a visible coronal streamer six years later, around the peak of an 11-year solar cycle, would serve as powerful confirmation that we’ve got the timing right.

Solar activity blocks cosmic rays, affecting the production of carbon-14. We estimate the strength of past solar cycles by looking at the proportion of carbon-14 taken up by trees, but the further back we go, the more uncertain the dates become, so other methods that can test these conclusions are scientifically valuable.

“This unique historical addendum for the possible solar coronal structure is critical for providing a spot reference on solar activity reconstructions from tree rings and ice cores, as well as providing independent validation of solar activity models,” said coauthor Professor Mathew Owens of the University of Reading.

The traditional Chinese belief that unexplained astronomical phenomena were a reflection of the failure of emperors inspired detailed record-keeping, much earlier than in other parts of the world. These records were often referenced long afterwards, so that even when the original was lost, the information has sometimes survived.

“Some of our ancestors were very skilled observers,” Dr Meng Jin, coauthor from the Lockheed Martin Solar and Astrophysics Laboratory, noted. “When we combine their careful records with modern computational methods and historical evidence, we can potentially find new information about our planet and our star from thousands of years ago.”  

Work like this relies on teams with varying specialties, but unusually, Hayakawa was able to fill two roles, having PhDs in the normally disconnected fields of oriental history and solar physics.

The study is published in The Astrophysical Journal Letters.