The ancient Romans looked to the heavens for signs of what might happen on Earth. According to historical documents, an aurora in the shape of horses and soldiers was said to appear in the sky in 44 BC, foretelling the fall of Julius Caesar. Another celestial event 114 years later presaged the sack of Jerusalem by Titus Vespasian.
Because auroras, comets, and meteors were omens, the Romans and other ancient peoples observed the sky fastidiously, recording heavenly observations alongside detailed accounts of earthly events and military triumphs. Twenty centuries later, those historical records have become scientific data for researchers such as WHOI Senior Scientist Andy Solow.
Auroras, or “northern lights,” are caused by eruptions and flares from the sun that stir up Earth’s magnetic field and atmosphere. These heavenly fireworks coincide with the appearance of sunspots on the face of our star. Three centuries of modern observations have shown that these spots and storms occur in a rhythmic pattern, with the number of sunspots waxing and waning in an 11-year cycle. But with only 300 years of observations, scientists can’t say whether this cycle is recent or abiding in the life of a 5-billion-year-old sun.
In the late 1970s, astronomer Richard Stothers of NASA’s Goddard Institute for Space Studies analyzed classical writings from 467 BC to 333 AD, particularly Livy’s history of Rome. He found dates and descriptions of celestial events that sound much like “great auroras”—northern lights that extend to middle latitudes (including Rome) during intense periods of solar activity. Stothers uncovered an 11-year cycle of solar activity at work in ancient times.
Some years later, Solow read Stothers’ paper and felt compelled to do another statistical analysis. He filed the problem away for another day, hoping to come back to it. In late 2003, he found the paper in his office, rekindling the urge.
“To me, data are data,” said Solow, who is trained in mathematics and statistics and leads the WHOI Marine Policy Center. “Research is driven by curiosity, and interesting statistical problems come up in all kinds of fields.”
Solow applied some new statistical methods to the dates Stothers gathered from ancient Rome. In a paper published in Earth and Planetary Science Letters in April 2005, Solow confirmed the 11-year solar activity cycle seen by Stothers. But he went further, finding that the cycle of solar activity and resulting auroras was bimodal—that is, the ancient cycle had two peaks. Solow’s finding jibes with modern observations that auroras occur in an 11-year cycle, but the peaks of activity tend to lag behind the peaks of the sunspot cycle (see third image at right).
For space physicists and earth scientists, establishing the stability of the 11-year pattern is important because solar variability has been linked to climate changes in the past. By finding both the solar and auroral cycles, Solow has provided twice as much confidence that the ancient solar cycle mirrored the modern one.