NavList:

Zane, you wrote:
"Stellarium must assume a chart datum so I would guess all its altitudes will differ from what you actually read by how high you are above sea level? Does it make any assumptions for refraction? I would guess not!"

There's no "chart datum" in Stellarium. It does include a standard refraction model. In fact, it's the usual approximate Bennett formula which has been included for calculator work in the back of the Nautical Almanac for many years. There are better models. This one is cheap and easy to implement. The refraction model in Stellarium also incorporates temperature and pressure corrections. You can set them in one of the standard information dialogs. Note that standard temperature in Stellarium is 15°C while in most celestial navigation tables and tools, the standard is 10°C. One can simulate higher altitudes above sea level by lowering the pressure, but in Stellarium this is nothing more than a hack.

As I have already noted, Stellarium's apparent altitudes include refraction and parallax. They do not include dip. This is based on a particular definition of apparent altitude which references the true horizon. The true horizon is the set of points exactly 90° away from the zenith. For nearly all astronomical work, that's what is meant by the horizon. Also, as I have already noted, if you want to take an apparent altitude from Stellarium and convert it into one that is roughly comparable to an observed sextant altitude, you can just subtract 12' for a Sun LL sight. The more general rule is add 0.97sqrt[ht.eye(feet)] for any body for dip. Then subtract 16' for Sun LL. Or subtract 0.2724HP for a Moon LL. Add the same for a Moon UL. Gets messy pretty quick.

So why, then, can you enter an altitude in meters for the observer's position in Stellarium?? This sure sounds an awful lot like they would be using it for dip, right? Nope. This is strictly a refinement to parallax calculations. You can test this easily.

Pick any star in Stellarium, like Polaris. Pause the time evolution. Select the star (click on it) so that you can see its apparent altitude. Now for your current observing location, change the altitude from whatever is normal to 50000m. At that height the refraction is near zero, and the dip is over 7°. But you'll notice that the altitude of Polaris does not change at all when you vary your altitude above sea level. That's because its parallax in altitude is zero to high accuracy. Now try the same thing with the Moon. Pick a date and location when the Moon is low in the sky. Change your observing altitude from 0 to 50000m. Notice that there is a small change in the Moon's apparent altitude. This is a result of the small "topocentric" change in the Moon's parallax in altitude. So basically none of this is relevant to the celestial navigator in any way. You can leave the observer altitude in Stellarium at the default value and after that just ignore it. 

Frank Reed
Clockwork Mapping / ReedNavigation
Conanicut Island USA