NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Real accuracy of the method of lunar distances
From: Richard M Pisko
Date: 2004 Jan 12, 01:16 -0700
From: Richard M Pisko
Date: 2004 Jan 12, 01:16 -0700
On Sat, 10 Jan 2004 00:45:09 +0000, George Huxtable wrote: >This effect of parallax is superimposed on top of the normal motion of the >Moon against the starry sky, and as that motion is always toward the East, >you can see straightaway that the effect of the rocking due to parallax on >the true motion of the Moon is always to make it appear less. Sometimes >significantly so. In an extreme case, when the Moon passes overhead, it can >roughly halve the true motion, so it's a very important effect. > George, forgive my presumption, but I wonder if you would please indicate where my interpretation may have gone off the track. I think of the problem as being in two parts also. As an analogy in the absence of a blackboard, think of a person on one of those now banned bits of playground equipment that is a big, level, circular platform that pivots in the center and goes around and around. Part I Put yourself on the platform in a swivel chair. Look up at a lamp standard about thirty platform diameters away; or closer, to exaggerate the effect. Have someone impart a gentle, constant motion to the platform. Notice that the lamp moves in an ellipsoidal pattern against a distant background ... mostly side to side. Both these movements (up and down as well as side to side) are roughly harmonic or sinusoidal if plotted against time. For a platform constantly turning counterclockwise as seen from above, also note: (1) the lamp will turn clockwise against a distant background if it is above the plane of the eye of the observer, (2) the lamp will trace a straight side to side line if it is the plane of the observer's eye, (3) the lamp will appear to trace a counterclockwise elliptical path against a distant stationary background if it is below the plane of the eye of the observer. Incidentally, I have heard the term "parallax scroll" used in computer games for the movement of a foreground against a background to give a three dimensional effect to the older games. To avoid confusion with the correct term, I will use "parallax scroll" for this apparent motion. If I understand what you say, the actual (or true ) location of the lamp (or moon) against the distant background (or stars) is stationary in my example so far (or known and listed in the almanac) for any given instant as calculated to the center of the platform (or the center of the earth). From surveying practice, the apparent position of the street lamp's elliptical "parallax scroll" against the background (say a big brick wall with a few spot lamps on it) can best be fixed against that background with angular distance measurements from three equidistant objects radiating out at 120 degrees from the street lamp. (The stars are not that accommodating for the moon ... at least not in the last two months of overcast I have had over here ... would be difficult for a moon position close to the horizon, and may not show well against an illuminated limb of the moon.) Locating the Part I or "parallax scroll" position with a fixed precision and accuracy of observation would give the angular position of the platform to a varying degree of accuracy, greatest when nearest the street lamp, and least at the far left or the far right. Greatest when the vector arrow length (arc distance divided by time) of the parallax is longest. Translating this to Celestial navigation, it would seem the accuracy of position finding would be greatest when the observed moon crosses the meridian. Part II Superimposed on this "parallax scroll" of the lamp or moon motion is a movement of the whole background, in this example, it would not be moving very much, about one degree per revolution of the platform, about five feet at one hundred yards. Any observation of the background around the position of the street lamp would not be inherently much more difficult at any position of the platform ... but refraction would make the far east or far west observations more suspect. Finally, there is the movement of the lamp. It moves from right to left at the rate of about twelve degrees per revolution of the platform, and usually up or down a bit. If we take away the parallax scroll of Part I (by shifting to the center of the platform), there is no difference in the accuracy of locating the lamp against the background no matter where it is ... and the motion of the street lamp during the time of observation would describe an almost straight line against the background. For the moon, there is the problem of refraction near the horizons which spoils the observational accuracy in spite of a constant observational precision. Argument If you want a good fix of the lamp against the background, or the moon against the star field, take it when the apparent motion of the lamp (moon) is greatest; in other words, when the vector length per second of observation, as determined by the greatest vector sum of the *two* "parallax scroll motions", the lamp (moon) orbital motion, and the slow background (or celestial) motion. Don't worry about how fast the vector sum arrow is rotating or changing direction, only about how long it is, because that will determine the smallest error in position of the platform (longitude) for a given precision of an observation set of the angular distance between the street lamp (moon) and three distinct bricks (stars). Again, I am defining the vector sum as angular distance over time. Working backward, if the angular distance over time is zero, the position is effectively unknowable because the potential error is so great. Unfortunately, the vectors for the horizontal motion of the orbiting moon and the rotating earth oppose each other most at the meridian crossing, and depending on your latitude and the time of the year there may not be much vertical motion. Too bad atmospheric refraction degrades the accuracy of an observation close to the rising or setting moon by an unknown amount, otherwise that should be the best time to take a lunar. Also, the likelihood of having three stars to check against the moon is not good, because at least half are below the horizon. Conclusion All in all, it might be best to take the lunar whenever you can get the greatest number of well positioned stars for observation, as the apparent motion is still appreciable even at the meridian crossing. If the earth were transparent, the apparent motion would be greatest when the moon is on the meridian 180 degrees from the observer (anti-meridian crossing?), and *that* would be the best time, as the vector lengths of the individual major motions would combine to make the longest arrow. The moon could also be seen to be making the complete epicyclic path against the background stars instead of the "bob" you describe, which is the top half. Question Since I am so weak on the celestial math and the standard formulae, I don't know how accurately the stars above or below the moon's path can be used to help determine it's absolute location in the celestial sphere from the observed sextant angles ... and from there determine the location of the observer on the actual earth. It seems to me the stars most in line with the vector of the moon's *orbital* motion would give the most accurate results from the tables, not the ones most in line with the longest vector of the apparent motion, but I am not at all sure about that. (Not that there is ever the perfect choice of stars in the first place.) Have I made a reasonably good stab at an analogy and its conceptual analysis? I hope I have made my thoughts clear enough to critique. -- Richard ...