When Does the Sun Rise… Really?

It’s strange but true. We may not fully understand one of the simplest metrics in observational astronomy: just what time does the Sun rise… really?

It’s something so basic that we rarely thing about it. Every morning, the sunrise races at us from the east at up to (if you’re on the equator) over 1,000 miles per hour (1,600 kph), and will do so for tens of thousands of mornings throughout our lifetimes. If there’s one thing you think you could count on, it’s the morning sunrise.

Now, a Michigan Tech analysis by the U.S. Naval Observatory’s Teresa Wilson suggests that traditional methods and almanacs may put the quoted sunrise and sunset time off by as much as 5 minutes. Wilson announced the results of the fascinating study at the January 8th meeting of the American Astronomical Society in Seattle.

The problem is one of refraction. If we lived on an airless word, the calculated and observed moment of sunrise would be easy… but as air-breathing mammals, we’d have other problems to contend with. Air bends light, meaning we see the Sun slightly offset from its true position on the horizon due to the atmosphere. Along with the Moon, the Sun is one of the few celestial objects that is large and close enough to appear as more than a point of light to the naked eye. Also, like the Moon, the Sun’s apparent diameter is about half a degree across, meaning you could line the local horizon with 720 Suns end to end, or 180 Suns from horizon to zenith. This size also changes very slightly from perihelion in January to aphelion in July, as the Sun seems to grow then shrink from a value of 31.6′ 32.7′ arc minutes.

Most calculations assume local sunrise and sunset time as when the center of the Sun’s disk clears the horizon. Of course, your actual horizon is probably cluttered with foreground objects that the Sun needs to clear, unless you live on a remote mountaintop or are lucky enough to observe sunrise and sunset from the beach.

Most standard sunrise calculations assume a refraction angle of 34′ arcminutes, a little larger than the apparent diameter of the Sun. Wilson notes in the study that this value is cited as far back as 1865, and its use may go all the way back to that 17th century master of optics, Isaac Newton. However, this value is an approximation, and does not account for local meteorological conditions. Air behaves very differently, say, on a still January morning over the Great Lakes versus a hot dusty July morning off the west coast of Africa. Yet, simply using a standard value assumes the true conditions at these disparately different sites are the same.

Wilson’s study looked at historic records of 514 sunsets and 251 sunrises from 30 separate geographic locations. Most of these (about 600) came along with weather data for the site, which Wilson then fed into three separate refraction models.

Wilson found that while sunrise and sunset varied by season, wintertime predictions tended to run late, while summer predictions ran early. Viewing sunrise over water seemed to magnify the effect, though taking into account the observer’s altitude did dampen down the discrepancy.

Sunrise over Jimena de la Frontera, Spain. Dave Dickinson

Moreover, modeling the complex effect of the weather in the troposphere didn’t make the discrepancy go away. Wilson found that using the current 34′ standard, we can’t predict the actual sunrise time to better than within 2 minutes.

Why does it matter? Wilson notes that one minute of error measuring sunrise at sea using celestial navigation can lead to up to 15 nautical miles of error. This is crucial, as the U.S. Navy has resumed teaching cadets old-school celestial navigation, in the event a cyber attack blinds GPS capability. Also, for now, our time is set to astronomical time, though there has been calls to move away from this standard and abolish the deletion and insertion of leap seconds starting in 2023. I think the really fascinating story here, though, is the fact that the science surrounding this basic facet of astronomy is something that really anyone could have done, had they simply thought to do it.

The solution? Perhaps smart forecasts can work to account for local atmospheric conditions, delivering to observers a better sunrise and sunset prediction.

…and the Sun will continue to rise and set, every day.

One Reply to “When Does the Sun Rise… Really?”

  1. This is not new. I remember going around the Trevose lighthouse in Cornwall before it went automatic, and one of the keepers jobs was to record the sunset time. In the 1930’s, this was used to estimate the pressure over the Atlantic.

Comments are closed.