Leaderboard

This comes up in conversations occasionally and I've been meaning to write up a concise and simple explanation for a while now. So here we go. But before I say anything else, I feel like I need to emphasize that I don't want to come across like I'm complaining about the books. Stormlight Archive is my absolute favorite series and this is such an obscure detail. Astronomy and orbital mechanics is a low-key passion for me (I blame Kerbal Space Program, mostly), so when I'm reading a book that I love and see information about this topic I can't help but dig into it. Brandon and his team do a LOT of work to make all of his fantastical worldbuilding make sense. They can't knock it out of the park every time. Maybe they'll look into this when doing the Stormlight leatherbounds and find a way to "fix" it. Maybe they won't--Roshar's moons are really imaginative, so why throw that out over some minor issues that 99% of readers will never even stop to consider? I'm just writing this up because I enjoy exploring the "science" of these books. Sometimes it doesn't work out nicely... and that's okay. It's fun to do the math regardless, and I'm just thankful Brandon put enough into the books that I can overthink it this much. The Situation I want this to be accessible for people who know very little about astronomy, so let's start with some quick basics. Below is a simple image of Roshar from "above". We're looking "down" on the North pole here.You can see which way it rotates relative to the Sun. You can see where the sun is setting and where it's rising. Note that the continent is actually in the southern hemisphere (I think it's one of the TWoK maps that shows the equator along the edges in a subtle way). For simplicity, in this post I'm just going to look at points along the equator. Also note that the continent spans about 120 degrees longitude, and Iri/Shinovar are about 90 degrees longitude behind the Shattered Plains. In other words, sunset on the Shattered plains is noon in Shinovar. Midnight in Shinovar is dawn on the Shattered Plains. We know that Roshar's moons orbit the planet once per day (every 20 Rosharan hours) because they rise and set the same time every night. This is something the books are never 100% explicit on, but it's heavily implied and has been confirmed. (Technically, they orbit a *bit* faster and precess slightly so that they keep their position relative to the Sun at all times of the year.) This nearly implies a geosynchronous orbit, where a satellite ends up over the same spot on the ground every day. But it can't be that or the moons would be up all day and night. Viewed from the ground they'd just sort of wobble back and forth around a fixed point in the sky. And that's not the case. They rise in the east, set in the west, and are only up for a few hours. (TWoK 2 & 23, among others) This means they are in a retrograde orbit (they go around clockwise) and it means they're in an elliptical orbit with the low point on the nighttime side of Roshar. This is what you need to get such behavior. Here's a very simplified approximation of one such moon: There are 3 of them of course. The Arcanum Unbounded star chart of Roshar suggests they are all at slightly different orientations relative to the sun. They are also inclined orbits, which is necessary to have the rise/set in the east/west for a viewer in the southern hemisphere. But for this post we're just going to keep it simple and pretend they're on the same plane as the equator. The order and timing of the moons as observed from the Shattered Plains is (and I'm using a 20-hour clock)... Salas rises just after sunset, around 15:00. Salas is up for about 2 Rosharan hours. Salas sets and we have an hour of darkness before Nomon rises. ("the hateful hour") Nomon rises around 18:00 and is up for 3-4 hours. Nomon sets as Mishim rises, around 1:00 or 2:00, and Mishim sets at sunrise (5:00). So why don't they make sense? Let's zero-in on the first moon, Salas. The moon should JUST become visible on the horizon at sunset, so the viewer's line of site looks something like this. Now we skip ahead 2 hours. Roshar turns about 36 degrees during that time (2 / 20 hours = 10%), so I've put an X at our viewer's new location. At this time, Salas is setting behind the horizon. So their line of sight looks like this. So consider what a person at points Y (2 hours behind the Shattered Plains) and Z (Shinovar) are seeing. It's pretty clear that they can't possibly see the moon rise at sunset. In Shinovar, Salas will be high in the sky already in their afternoon. Now, to be fair, the orbital path that I've drawn for Salas is somewhat arbitrary here. But the fundamental problem can't be fixed with a different orbit. If you want to slow the moon down, so that it's simply visible around sunset in Shinovar, that means it's going to be visible longer in the Shattered Plains' night sky. Solutions? I can't help but wonder what could be done differently. If you were going to try and "fix" the books, what approach would you take? (Aside from the option of just ignoring the problem and not thinking about it too much) The simple option is to just make the orbits work for the Shattered Plains and say that they ARE indeed not visible for some of western Roshar. Maybe they are visible during the day. The problem with this is that Szeth seems to suggest they don't do that. The Shin are actually where we get a name ("hateful hour") for the time between Salas and Nomon. But that could be changed. And there are some other timings that would need to be adjusted, but not too many. Alternatively, you might be able to push all of them back in their orbits a bit so that they all show up in the night sky for everyone across the continent. This would probably result in them moving faster across the sky. And it would probably mean that eastern Roshar has a few hours of darkness between sunset and Salas while western Roshar has a few hours of darkness between Mishim and sunrise. Another option is to have them move more slowly. It wouldn't be hard to have them set to all rise at roughly the same time, if you turn the orbit the right way. It simply means they're going to be relatively high in the sky for that portion, and that means they're going to be visible for much longer. Their movement would almost be more from Roshar turning than from the moons' own movement. (i.e. more like the way our own moon moves across the sky) This means their visibility will overlap and it means we have to toss out the hateful hour. Of course, if someone sees something I'm missing, please let me know. I HAVE simplified this by keeping things on Roshar's equatorial plane, but I'm fairly confident that you get more or less the same thing if we incline the orbits a little and view the moons from a point in the southern hemisphere.