As Wrong as the Night is Long

moonAh, the danger in assumption. A lifetime of watching moon phases came back to haunt me last Friday last week as I realized my “close enough” estimate of moon rise and set isn’t, at least not for the latitude of my work in progress. It’s the exact same error as the writer who didn’t realize the sky in a medieval fantasy world would look different than his own simply because there was no widespread outdoor lighting. And I, having lived mostly at the lower latitudes, assumed that moon rise and set always correspond to sunset and sunrise as described. For countless times I’ve watched the full moon rise at sunset, and never gave thought that it wasn’t the same everywhere in the world.

That it’s not didn’t hit me until I was thinking about the length of day and night in conjunction with the book to be released. We all know that days are longer in the summer and shorter in the winter, and this effect becomes greater at the higher latitudes. Above the arctic and antarctic circles, you don’t see the sun at all at the winter solstice, and it doesn’t set during the summer solstice. That’s why it’s sometimes called “The Land of the Midnight Sun.” Then the obvious hit me: This means a full moon can rise other than roughly at sunset, which means my “close enough” spreadsheet can be way off. At high enough latitudes, it can rise and set twice in the same day.

Obviously, if the sun doesn’t set, the moon isn’t perpetually full. Though my description of why there are phases of the moon is correct, they are not tied to sunrise and sunset. I treated it as though it was a local event, which means my quick and dirty spreadsheet could be very wrong.

I went back and put the errata notice at the top of the original post, lest others make the same mistake, though I was already thinking about alternative ways of estimating sunrise and sunset for the work in progress. I wasn’t happy with any of them. One of them included setting up a spreadsheet, as I did for sunrise and sunset, and calculating moon rise and set. That means going through my calculations to see how I did them. Though I half remember how, calculating this for the moon the moon isn’t quite as simple as the sun, and odds are I’d have some sort of blunder in my calculations. It looked like I’d have to be fiddling around with an open source astronomy program, a workable solution, but one I’d have to do for every night scene.

Fortunately, an easier quick and dirty solution came to mind while commenting about this on another blog. There’s minute variations in sunrise and sunset, but not so much that I couldn’t just get rise and set times for a specific location and use them. The only two things to worry about is if the two years match – both a common year or leap year – and, if a story is set in an era where they reckoned dates with the Julian Calendar, to shift the dates to modern Gregorian Calendar. You can do that with Julian Calendar + integer portion(year/100) – integer portion(year/400) – 2. So, if you want to know about when the sun rose and set over Hastings on October 14, 1066, you get the rise and set data for that location for a modern year; calculate the date correction, as 1066 was under the Julian Calendar, and look it up on the modern calendar. The date correction is Integer Portion (Year/100) – Integer Portion(Year/400) – 2. For 1066, that works out to 6 days. So, to get an idea when the sun rose and set on October 14, 1066, you add 6 days and look up October 20 in the modern era. This is not precise and would make any astronomer or navigator cringe, but, for our purposes, it gives a reasonable idea of when the sun was up and the length of the day.

What does this have to do with moon rise and set for a given year? A good deal. It’s possible to look up the moon phases for different years. We can tell that there was a full moon on October 6, 1066, Julian Calendar. Correcting for the Gregorian, it would correspond to October 12 in a modern calendar. A web search for years with full moons on October 12 turns up the year 2011. Both 1066 and 1979 were common years, meaning they weren’t leap years, so looking up the moon rise and set for Hastings on October 12, 2011, should give some idea when it rose on October 6, 1066. This is not precise, not by a long shot, but it gives a passable approximation for fictional purposes.

By the way, on October 14, 1066, the date of the Battle of Hastings, the moon was its Last Quarter. Just tossing that out there.

Searching the web for full moons occurring on October 12, Coordinated Universal Time (UTC) works, but it doesn’t sit well with me. I thought I had it worked out with the Metonic Cycle, the 19 year period where the same phase of the moon falls on the same date. This isn’t an exact fit, and creeps forward 1 full day in 219 years. But tinkering around with the Metonic Cycle, I don’t quite have a fix; it worked in once case but not the other, and that won’t do at all. So, for now, I’m stuck with doing a web search.


What lead to all of this was the realization that I’m also guilty of assuming twilight occurs the same way in every place on earth. The problem is that it isn’t. You don’t have to be beyond the Arctic or Antarctic circles to observe this, either. Even London, England, doesn’t experience Astronomical Twilight, when the sun is between 12° to 18° degrees below the horizon, from about the third week in May to almost the third week in July. Most of us consider it dark long before that point, but what it means is the higher the latitude, the brighter the sky around the Summer Solstice, until you get to the Arctic and Antarctic circles and the sun doesn’t set at all. Even when the sun sets, if you’re at a high enough latitude, you can have enough light to see by. Basically, if you’re at a high enough latitude, the sky can get no darker than Civil Twilight.

Civil Twilight is that time after sunset and before sunrise when the center of the sun is no more than 6° below the horizon. That varies depending on location and time of year. Sometimes it’s officially defined by a set number of minutes. My introduction to Civil Twilight was in hunting, and a set number of minutes before sunrise and after sunset. Beyond that, it’s considered too dark to really make out things well.

Yet the sky itself may not be what we consider dark, at least not dark enough for anything but the brightest stars and planets to become visible. For that you have to wait for beyond Nautical Twilight. That’s when the center of the sun is 12° below the horizon, and it’s no longer possible to make out the horizon at sea.

What this means is that it’s not truly dark. During nautical twilight, it may be too dark for a character to see well, but there is still some light in the sky. The same during astronomical twilight, though it may get to what anyone but an astronomer would call dark. This affects what the character sees.

This had me worrying about that book to be released. In late summer, several chapters occur at night, including one where the characters are up the entire time. What do they see? Will night be long enough for all the events to take place? As it turns out, yes, it will. Some may say this is excessive for a work of pure fiction, and perhaps they are right. But anyone who lives at that latitude – and a significant portion of the world’s English speaking population does – it’s going to stick out like a sore thumb if it’s wrong. If I had set it around the Summer Solstice, it wouldn’t have been possible at all.

All this means is that we have to be just as careful in describing the sky in our stories, even if it’s only incidentally, such as whether the day or night is long enough for events in a particular season. It’s another point of research, just like climate and flora and fauna and all the other minute things we use to create a believable fictional world that won’t through our readers out of the narrative. Even if our characters only pause a moment to admire a sunset or look at the stars, it would be good to know if the sun even sets, or if the stars are visible at all.