NavList:

Jim, you wrote:

"but can any navigational star or planet be used to work a lunar distance, as
long as the altitudes of the body and moon are within the window for a given
method?"

 

Theoretically, any star or planet can be used. But you wouldn't want to use one that is too far off the ecliptic since then the distance won't change rapidly enough to measure. Let's do an example. If you go to my lunars site and select "predicted lunars", you can choose "Include Navigational Stars" for the "Body". This will generate lunar distances for the Sun, planets, and all the navigational stars (which includes the traditional lunars stars). Try it now.

 

You'll notice that the rate of change of the distance between the Sun and Moon today is about 31 minutes of arc per hour. So if we can measure angles to 0.1 minute of arc accuracy, we can get GMT to an accuracy of about 12 seconds. Now scroll down to Aldebaran, a star which is nearly on the ecliptic and was one of the traditional lunars stars. The rate of change here is 32.8 minutes of arc per hour (the mean rate is 32.9 so we can conclude that the Moon must be about halfway between apogee, where it's slowest, and perigee, where it's fastest). In principle, this means that we should be able to get GMT to an accuracy of about 11 seconds (that's 3600/328). Since the Sun moves, too, and in the same direction as the Moon, the rate of change for Sun-Moon lunars is a little lower than star lunars, but it's nothing to worry about (knowing GMT to +/-11 seconds is not materially different from knowing it to +/-12 seconds). Also, we don't have the option of choosing whether the Moon is at perigee or apogee so the rate of change for stars right on the ecliptic can be anywhere between about 28' and 38' per hour.

 

Next, scroll down to Polaris. It's a long way from the ecliptic, roughly 23 degrees from the North Ecliptic Pole. The rate of change of the distance between Polaris and the Moon today is about 15 minutes of arc per hour. If we shoot a lunar using Polaris and achieve an accuracy of 0.1 minutes of arc in the measured distance, this will determine GMT to an accuracy of only +/-24 seconds of time, substantially worse than for a star near the ecliptic. Finally, take a look at Canopus...

 

In the old lunars tables in the almanacs, the stars Altair and Markab are included. These are rather a long way off the ecliptic and they might seem inappropriate for lunars. But it's really not so bad as long as those stars are a good distance from the Moon. Imagine an arrow pointing out from the Moon's center more or less along the ecliptic indicating its direction of motion (+/- 5 degrees because the Moon's orbit is inclined). If that arrow points directly at a star, then we get the maximal rate of change in the angular distance to that star. Let's use the mean value of 33 minutes per hour. But suppose it's not lined up exactly. How much will the rate be reduced from 33? We can draw a great circle arc from the star back to the Moon's center. It will make an angle relative to that arrow indicating the Moon's direction of motion. Even if that angle is as large as 45 degrees, it will only reduce the rate of change of the lunar distance by about 29% which isn't all that bad. The rate of change in the lunar distance would be 23.3 minutes of arc implying that we could measure GMT to +/-15 seconds of time. This appears to be the standard that was used in the Nautical Almanac tables. We can turn this into a nice visual picture for finding acceptable stars for lunar distance sights: draw two perpendicular straight lines through the Moon's center. Calling the Moon's north pole N, south pole S, etc., these lines would be aligned along NE/SW and NW/SE. Extend those lines across the whole sky. They divide the sky into four sectors relative to the Moon. There are two sectors containing the Moon's poles, and two sectors pointing forward and back respectively along the Moon's orbit. Any star within those latter two sectors, no matter how far away, will have a rate of change of distance that is at least 71% of the maximal rate of change and would be acceptable for lunar distance sights, and this can easily include stars that are a long, long way from the ecliptic.

 

And:

"I know the old Almanacs had lunar distance data for only a few
selected stars, and I think that Frank's online calculator has solutions for
some selected stars"

 

Right. The present calculator does the Sun, navigational planets, and the traditional lunars stars. I've done some preliminary work to add the remaining navigational stars to the "Lunars Clearing" tool. I can understand that some people enjoy the concept of using a modern selection of stars. And I know that Steve Wepster's lunars tables include Nunki and maybe some other non-traditional lunars stars. Myself, I prefer to use the traditional lunars stars if I use stars for lunars at all. This little selection of nine stars was one of the longest lasting elements of celestial navigation.

 

And:

"but can I just go outside, shoot a set of sights
between the moon and a convenient body, and then come back indoors to reduce
the sight?  Or do I have to select from a short list of nagivational bodies?"

 

What's your goal? If you want to experience the 18th/19th century tradition, then perhaps you should stick to the nine lunars stars. Better yet, use the Sun since this appears to have been the prefered object (something like 80% of the time) in the 19th century. Of course, this is entirely personal preference.

 

And concluded:
"I have a run of sights on a star from last fall, taken with chilled but not
frozen fingers, but I recall that when I tried to do a preliminary reduction
using Frank's online calculator, that star was not available."

 

Which star?