artificial horizons, can anyone help"

[alan2]

In latest edition, 1994 edition of Mary Blewitt’s Celestial Navigation for Yachtsmen, page 53 under Practice Sights, one finds the following, re the Davis Type Artificial Horizon, essentially a dish of water or other liquid, with it‘s own sun shields, side panels and clear wind shields. “Corrections for index error and refraction must be applied to arrive at observed altitude.”
Fair enough until one looks at page 22 where one finds the following. "Nautical Almanac tables already contain refraction correction". Also, index correction is obtained from “zeroing” one’s sextant, then viewing the natural or sea horizon through the sextant’s optics. If there is a “break” in the horizon line, turn the micrometer wheel until it’s leveled out, and read the vernier. The number of minutes shown, if any, is the index correction. “If it’s on, take it off, if it’s off, put it on”, as the saying goes. Unless one has previously taken shots with this sextant, how does one arrive at the referenced index correction, or might this be a printing or editing error that got by?

Re the Celestaire Practice Bubble Horizon, page 13 of Celestaire’s 2009 catalog, see page 12 for Davis unit, one finds the following instruction for determining the BC (Bubble Correction). Using the Known Position Method, “take several observations from a known geographic position (GPS does serve a purpose), and compute the lines of position normally. You may attribute the average error to the bubble, and subtract as a correction (BC) for future use”. My questions follow.

Re the several observations mentioned, I assume the following. One takes several sets of sun shots. Reduce each set of shots, and plot them. You have already plotted your Known Position. Measure the displacement between calculated position plots and the plot of Known Position. Say you took three sets of shots, the average error, displacement between their plotted positions and that of Known Position being say 12 nautical miles, the BC would be 12 minutes of arc, 1 minute of arc equaling 1 nautical mile. Next question is, re “subtract this as a correction (BC)", subtract the obtained BC from what number. I use the USPS SR 96a form for reducing sights.

If it turns out that I’m way off base here, having seriously misunderstood the directions, someone please explain, so that I might correct the error of my ways. Thanks for reading through the above, and for the benefit of your wisdom, whomever you might be.

[Paul Elliott]

You can measure the index error by observing any distant object, such as a far hill, with the index zeroed then noting the index error as described in Blewitt.

[maxingout]

Index correction is easy to measure as long as you have an unobstructed horizon at sea. You simply look at the horizon and adjust the sextant so that there is a single line across the horizon, and then read the vernier setting on the sextant.

I used the Navy Mark V bubble sextant for navigation in the deserts of Arabia when I did expeditionary travel in the deserts. I never applied a bubble correction to my sites, and I found that I was genereally two to three miles off my gps position using star sights at twilight.

After a 110 km desert run, I generally found that my DR position was off by about 10 kilometers since we were not travelling in a straight line. I used my DR position as an assumed position to compute my sights at twilight, and I was usually 2-3 miles off the GPS position using the bubble sextant without a bubble correction being applied.

From my experience, I would ignore the bubble correction and not worry about it. The important thing to me was to be able to work the site and get within a couple of miles of my true position. Unless you are surveying with a transit on a tripod, it's going to be difficult to know your position within more than a couple of miles using celestial navigation. And when you get on a yacht at sea, you will be very lucky to be within two miles of your gps position in my experience. Of course, if you are on the deck of a freighter taking sight, you may be able to get closer than two miles.

I would pay attention to the index error, but I would not worry about the bubble correction.

[Lodesman]

Originally Posted by alan2

Unless one has previously taken shots with this sextant, how does one arrive at the referenced index correction, or might this be a printing or editing error that got by?

You should periodically check for index error, as has been described. If the error is gross you can fix it with the adjustment screw on the mirror. Once the error is less than (iirc) 3 min, then it can just be applied as a correction.

Re the Celestaire Practice Bubble Horizon, page 13 of Celestaire’s 2009 catalog, see page 12 for Davis unit, one finds the following instruction for determining the BC (Bubble Correction). Using the Known Position Method, “take several observations from a known geographic position (GPS does serve a purpose), and compute the lines of position normally. You may attribute the average error to the bubble, and subtract as a correction (BC) for future use”. My questions follow.

Re the several observations mentioned, I assume the following. One takes several sets of sun shots. Reduce each set of shots, and plot them. You have already plotted your Known Position. Measure the displacement between calculated position plots and the plot of Known Position. Say you took three sets of shots, the average error, displacement between their plotted positions and that of Known Position being say 12 nautical miles, the BC would be 12 minutes of arc, 1 minute of arc equaling 1 nautical mile. Next question is, re “subtract this as a correction (BC)", subtract the obtained BC from what number. I use the USPS SR 96a form for reducing sights.

I assume you're just using the bubble horizon for land-locked practice - if that's the case then I'd ignore it as Dave has suggested. If you still want to correct for it, then I would suggest charting the errors between your practice LOPs and your known position. The bubble error will be like index error - it will always be on or always be off. You can apply it in your calcs with the IE or you add or subtract from your run when plotting your LOP. Obviously the error between GPS and LOP is going to vary widely, so you will have to take a lot of shots to determine an error from the averaging method. Assuming all other errors will cancel each other out the only remaining error will be a close approximation of the bubble error.

Bowditch has a pretty good section on the sextant, but I don't recall if he talks about artificial horizons.

Kevin

[alan2]

maxingout, Lodesman and Paul Elliott:

Thanks to you all.

Some of the things suggested, like regularly checking index error, I do without fail, routinely. My Astra 111B seems quite stable, less than 1 minute "on" when last I was near the ocean in October.

As for other aspects, I will look closely at what you all were kind enough to offer. BTW, thanks for bothering to read through my post and questions.