NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: More about lunar observations
From: Bill Noyce
Date: 2002 Mar 21, 15:29 -0500
From: Bill Noyce
Date: 2002 Mar 21, 15:29 -0500
I worked through Chuck's lunar observations, and my results lead to an even later time than Chuck's did. My calculations say the time of Chuck's middle observation (lunar dist = 37d 43.2) was at 00:20:30 UT, rather than at 00:15:27. I doubt if this is all observational error; I'm pretty sure I have a lingering error in here.\ -- Bill Here's the work. Almanac data: Venus Moon '0h' GHA dec '0h' GHA dec 19-Mar 00:00 UT 163d 28.7 +4d 26.2 125d 58.0 +15d 55.3 19-Mar 08:00 UT 122d 14.7 +17d 13.3 20-Mar 00:00 UT 163d 19.8 +4d 56.7 ------------------ ------------------- change in... (24h) -8.9 +30.5 (8h) -223.3 +78.0 change in 1 hour -0.4 +1.3 -27.9 +9.7 19-Mar 01:00 UT 163d 28.3 +4d 27.5 125d 30.1 +16d 05.0 The '0h' GHA ignores the 15 degrees/hour of the earth's rotation, to make the numbers smaller for interpolation. It's easy enough to add in separately, but it's not needed for this situation, since all we care about is the difference in GHA for Venus and the Moon. These results agree reasonably well with Chuck's -- the 01:00 moon position is off by a couple of tenths of a minute, probably because I had to interpolate where the moon's motion is changing rapidly. I computed the predicted distances using Ageton's method. First, at 00:00 163d 28.7 - 125d 58.0 ----------- LHA 37d 30.7 -> log csc .21544 dec +4d 26.2 -> log csc 1.11150 log sec .00130 add -------------- R 37d 22.8 log sec .09984 <- log csc .21674 sub --------------- K +5d 35.2 <- log csc 1.01166 Lat +15d 55.3 --------- dif -10d 20.1 -> log sec .00710 add log sec R --------------- Hc 51d 25.2 <- log csc .10694 dist = 90d - 51d 25.2 = 38d 34.8 Similarly at 01:00 163d 28.3 - 125d 30.1 ----------- LHA 37d 58.2 -> log csc .21096 dec +4d 27.5 -> log csc 1.10939 log sec .00131 add -------------- R 37d 50.1 log sec .10251 <- log csc .21227 sub --------------- K +5d 38.9 <- log csc 1.00688 Lat +16d 5.0 --------- dif -10d 26.1 -> log sec .00724 add log sec R --------------- Hc 50d 57.5 <- log sec .10975 dist = 90d - 50d 57.5 = 39d 02.5 Expressing Chuck's predicted distances in minutes, he had D1 (00:00) = 38d 34.82, D2 (01:00) = 39d 02.30 and I got 38d 34.8, 39d 02.5 Close enough here (though I had to correct some blunders to get this far). I noticed that a good check on this manual method would be to reverse the roles of the moon and the other body, and compute the distance again. Most of the intermediate results would be different, but of course the final result ought to match. I didn't go through this exercise this time, though. Now for the sights. I didn't have Chuck's Height of Eye, so I ignored dip. The 3' difference in altitudes isn't going to lead to a significant change in Refraction or Parallax, as far as I can see. The first step is to average the sights to bring them to a common time. For the distances, that's easy: 00:14:59 37d 43.4 00:15:27 37d 43.2 00:16:01 37d 43.2 ------------------ 00:15:29 37d 43.2 average distance Next we want the altitude of each body at 00:15:19. I averaged the two early sights of each body, and averaged the two late sights of each body, then used linear interpolation between them to find a 00:15:29 altitude. This appears to be different from what Chuck did when he "roughly averaged the sights". Venus Moon 00:11:41 7d 50.6 00:12:48 44d 52.6 00:14:02 7d 25.0 00:14:02 45d 25.6 ----------------- ------------------ 00:12:53 7d 37.8 00:13:25 44d 39.1 00:16:36 6d 43.0 00:17:31 43d 51.6 00:18:09 6d 24.0 00:19:09 43d 30.0 ----------------- ------------------ 00:17:22 6d 33.5 00:18:20 43d 40.8 269 seconds -64.3 295 seconds -58.3 146 seconds -34.9 124 seconds -24.5 00:15:29 7d 02.9 00:15:29 44d 14.6 These are the Hs of each body at the time we measured the lunar distance. We need to correct the moon's altitude and the lunar distance for semidiameter (and correct both altitudes for dip, which I'm ignoring) to get Ha. Then correct both altitudes for refraction and the moon's for parallax to get Ho. My tabulated SD and HP are for the middle of each day: 12:00 UT 18-Mar SD=14.9 HP=54.8 12:00 UT 19-Mar SD=15.1 HP=55.3 Interpolating to 00:15 19-Mar, I get SD=15.0 and HP=55.1. Because the Moon's altitude is between 30 and 55 degrees, we can add 0.2' augmentation to SD. Looking back at my work, I see I used 15.3' for SD - I think I just used the tabulated SD for the day, plus augmentation. Venus Hs 7d 02.9 Moon Hs 44d 14.6 distance 37d 43.2 SD +15.3 SD +15.3 s 7d 02.9 m 44d 29.9 d 37d 58.5 Refraction -7.3 Refraction -1.0 Parallax +38.7 S 6d 53.6 M 45d 07.6 Looking back over the numbers, it seems that Parallax should have been closer to 39.2 -- not sure where 38.7 came from. Now we can plug the above numbers into Borda's method. (Thanks to George Huxtable for writing it up.) It can use the same log sec & log csc tables as Ageton's method. m 44d 29.9 -> log sec .14675 s 7d 02.9 -> log sec .00330 d 37d 58.5 -------------- ------------ sum .15005 sum 89d 33.7 /2 44d 46.85-> log sec .14906 -d 6d 48.35-> log sec .00306 M 45d 07.6 -> log sec .15148 S 6d 53.6 -> log sec .00318 ------------ -------------- M+S 52d 05.6 sum .30678 /2 26d 02.8 - .15005 -------------- diff .15673 A 33d 23.6 <- /2 .078365 ------------ sum 59d 26.4 -> log csc .06495 diff 7d 20.8 -> log csc .89323 -------------- sum .95818 D/2 19d 22.85 <- /2 .47909 D 38d 45.7 Converting Chuck's cleared distance to minutes, he got 38d 43.49 -- 1.2 minutes less than mine. If I had used the larger value for Parallax, my cleared distance would be even larger. This difference looks larger than random noise, and I assume it's larger than Letcher's method would introduce, so probably I have made a mistake. But, pressing on, we have Predicted 00:00 38d 34.8 Predicted 00:00 38d 34.8 Predicted 01:00 39d 06.7 Observed 38d 45.7 ------------------------- ------------------------- change in 1 hour +31.9 observed change +10.9 10.9*60/31.9 = 20.5 minutes In other words, the observation occurred at 00:20:30 UT. Thus, my calculations say Chuck's clock is exactly 5 minutes slow -- even worse than his own conclusion.