When the solar eclipse happened in Pittsburgh last Monday, we knew all about it: the eclipse would begin at 2 p.m., end at 4:30 p.m., and 97 percent solar coverage would occur at precisely 3:17 p.m. How did we know this in advance? A lot of math and technology told us.
Astronomers have developed models of how the Sun, Earth, and Moon move around each other using their gravitational fields, Newton’s laws of motion, and the exact shapes of each of these celestial bodies. Since all space is relative, astronomers take into account where the Sun and Moon are relative to Earth’s location. Using the current positions of these bodies at a given time and several complex equations relating how they move, astronomers can program a computer to integrate these equations to find the relative positions of the Sun and Moon as seen from Earth at any time , says NASA.
With this, along with historical pattern recognition, NASA can calculate the details and timing of eclipses a thousand years in the future and thousands of years in the past.
But you don’t necessarily need this technology to predict eclipses.
Many historical astronomers have done well at predicting eclipses without modern supercomputers.
The first documented correct prediction of a total solar eclipse was made by the English astronomer Edmund Halley in 1715. Halley predicted the eclipse to within four minutes, and he plotted a path of totality across Great Britain that was only about 20 miles away. He did this by using many of Newton’s new theories, astronomical measurements and a lot of mathematics. This was in contrast to the fact that ancient civilizations could not predict the locations of total solar eclipses because there was no easy solar cycle to extrapolate from, and they could not accurately predict the location and movement of the Moon. Also, the path of a total solar eclipse is relatively narrow, meaning that even when a total solar eclipse occurs, it will only be visible from a patch of Earth.
Although correctly predicting total solar eclipses is a relatively new phenomenon in human history, humans have been predicting lunar eclipses for millennia. One massive help is the Saros cycle, an 18-year cycle for which the Sun, Earth, and Moon will be in roughly the same position at some point during each cycle.
When a lunar eclipse occurs on Earth, there is a chance that another lunar eclipse will occur there 18 years later. But the same cannot be said for total solar eclipses; if one total solar eclipse occurs at one time, another is likely to occur 18 years later.
The catch is that the aggregate will most likely be in a completely different place on Earth. (Lunar eclipses, on the other hand, are visible from about half the planet, making them easily predictable.)
After three Saros cycles, the sun usually returns to roughly the same location on Earth, but with enough errors to mean that total solar eclipses repeat in roughly the same locations every 54 years.
We don’t know how far eclipse prediction goes—the Mayans, Chinese, and Babylonians may have predicted eclipses using naked-eye observations and pattern recognition even before written records began. The Greeks had the Antikythera Mechanism, a mechanical device that predicted the positions of objects in the night sky. He could predict lunar eclipses using the Saros cycle, but he could only predict when a total solar eclipse would occur, not any details about where on Earth. The Mesopotamians also predicted eclipses using the Saros cycle, which was key because they associated total solar eclipses with the death of kings.
Eclipses are part of a shared human experience transcending space and time, and interestingly, predicting eclipses is also part of that shared human experience
Societies for most of history were unable to make accurate astronomical measurements and perform complex calculations. Today, these barriers are behind us. So if you’re dying to see another eclipse, astronomers can tell you exactly where to go and when.
