azimuth by lunar

Aha, found John Hamilton's 2005 paper thereon.

I'm sure it can be done, but doubt the ultimate accuracy will be as good as solar.

I'm certain it will not as be good as using Polaris when conditions allow.

The closest thing I found to an answer is toward the end of this paper:
http://www.terrasurv.com/azimuths.pdf

The orbit of the moon is hard to predict with the accuracy of a solar ephemeris because the sun also influences it in addition to the earth. And your latitude and elevation have much more effect on lunar azimuth than for a solar azimuth because the moon is closer.

Then there is the problem that at most times the illumination is not uniform, which could bias you in determination of the edges of the disk.

The difficulty of having current ephemeris data kept the moon from being much used in classical navigation, although tables were published until the early 1900's to allow coarse determination of longitude by comparing angles of moon versus sun or a star.

A coarse computation can be obtained from this Matlab function
http://www.mathworks.com/matlabcentral/fileexchange/22992-lunar-azimuth-and-altitude-estimation-algorithm

Also, watch eBay. Every now and then they have these neat astronomic compasses like the used for navigation in B-17's during WWII. They are awesome. Most lately have been auctioned with the storage box and instruction manual and appear to be in great condition. They even would look slick sitting on your desk in the office between uses.
Enter "Military Compasses" in the search window on eBay. I have seen them go from $40 to $150, depending on if a bidding war breaks out or not.

I'm sorry

But that idea seems like sheer lunacy!

🙂

(Sorry, it just struck me!)

N

Also try searching for ASTRO COMPASS

If you're searching on eBay, use quotes ("astro compass") or eBay will assume you are a dolt who can't spell auto and will show you a whole bunch of crappy auto accessory junk.

For example:

Astro Compass MK-II w/Case-Military WW-2

Astro-Compass MkII

supposedly you can shoot Polaris during the day. I've tried to find it without any luck in the daytime.

I did try in June, it might be easier at this time of year with the sun down lower.

The advantage of Polaris is it doesn't require a filter and the calculations are fairly simple. Another advantage is it is always in the same direction so if you are in trees you just need a hole to the north; it won't go hide behind a tree before you get the observations completed. Also you don't need the time any more precise than a minute or so.

I've been told the Moon wobbles making the calculations complicated, to say the least.

I have tracked Polaris for about an hour after official sunrise, and likewise found it before sunset by already having azimuth to less than a degree. Conditions: Topcon GTS-2B, July, small city < 1000 ft elevation, clear sky.

Suggestions of a 11" black paper tube (exaggerated lens shade) made almost no noticeable difference for me.

To find Polaris when it is light you have to have very good focus at infinity - something a fraction of a mile away may not give you good enough preset focus.

With good optics at very high elevations and with clear sky (dry and cold) I've heard that it is possible during most of the day.

Lighting the crosshairs when it is dark remains a problem for me, since that function is not built into the instrument. External lights are awkward, hard to adjust, and if they block any of the lens aperture may alter the apparent azimuth. I'd like to hear good ideas for this or see a diagram of the manufacturer's external illuminator.

Although Polaris calculations are simple, don't forget the Laplace correction if it is significant in your area. That was not mentioned in older textbooks.

Not concerned about accuracy so much as just wanted to know how to do an azimuth by lunar observations. Looks like John Hamilton's paper is a good place to get started. As he points out, the calcs are tedious but that's what computers are for. I'm working on learning the Python language, so that might make a good little project.

Boredom is our worst enemy.

I have done both solars and Polaris. Daytime Polaris, not so much. Near sea level and with city smog it doesn't work too well. Best I was able to do was to "stake to" where Polaris should be and finally see it between sunset and dusk.

Those old compasses look interesting but are out of the budget for now.

Last time I was doing Polaris (with a Leica 1203 and a Wild T2) I was using a red LED light on a gooseneck that you could clip to the tripod and shine into the objective lense from the side. It worked pretty well.

Every once in a while I progress a little on my lunar azimuth software program. I have it now where it will automatically go to the USNO web site and download the polynomials needed to accurately compute the moons position. These are listed in a yearly file (17 per day): http://asa.usno.navy.mil/SecD/2011/lunarpoly_2011.txt

My belief is that the moon can be used for accurate azimuth determination IF you properly compute the ephemeris (using the polynomials) and compute the azimuth using TOPOCENTRIC rather than geocentric coordinates. In other words, you have to account for the fact that you are not observing from the center of the earth, which is where the ephemerides are referenced to.

As for the "illumination not being uniform", I am not sure about that causing much error. But, it should be possible to compute an "offset" to a particular feature on the moon, like a prominent crater. I believe all of the visible features have fairly accurate lunar coordinates.

http://www.fourmilab.ch/cgi-bin/Earth/action?opt=-m&img=MoonTopo.evif

Once I get my astro pillar built in my yard (the hole is dug, I have the form ready, just need a metal top plate), and my software program finished, I plan on doing some testing to really see what kind of accuracy is possible in the real world. I have a Trimble S6, Zeiss S10, Kern DKM2, Wild T2 (all 1 sec) and a Wild T3 to play with. Problem is, I never seem to have much free time.

Nautical almanac lists the moon, so within it's limited accuracy it can be used for ephemeris data. One thing to watch out for is the large amount of parallax, which is tabulated there also.

I have shot it a few times without too much effort, and probably didn't really understand what I was doing, came within a minute or so. The results using the NA are not going to be super accurate, but would give you some opportunity to play with the idea.

Nautical Almanac listings for a given object and day are available on the net, so you don't need to order one.

Usually you can see the dark limb through any reasonable powered scope.

- jlw