NavList:

   

Reply

  Last year I had the opportunity to cross the Atlantic on the Royal 
Clipper, a 437 foot long five masted full rigged ship as I reported at:

http://www.fer3.com/arc/m2.aspx?i=110827&y=200911

One of my goals on that sixteen day voyage was to evaluate the accuracy 
of celestial navigation so I brought my Tamaya sextant along. I got up 
early almost every morning to take the morning stars and took 
observations of the sun during the day and then more stars at evening 
twilight. Altogether I took 78 sights. I also had two GPS's with me and 
I recorded GPS fixes within just several seconds of each celestial 
observation so that I could use the GPS fixes as the assumed positions 
and so could determine the accuracy of each shot. I did the sight 
reductions with my Bygrave slide rule and then repeated the computations 
with a calculator for maximum accuracy. The shots were taken from 
several locations on deck and the height of eye varied from 24 feet to 
33 feet. I had measured these heights by dropping a weighted line over 
the side (actually a hotel complimentary small shampoo bottle) until it 
hit the water and floated. I then marked the line off and measured it 
off with my Ikea tape measure.

One one occasion I took 15 shots of Venus in 39 minutes and the 
intercepts had a standard deviation ( σ ) of 0.876 NM. On another 
occasion I took twenty sun shots in a 54 minute period and and the σ was 
0.923. Combining these two series, comprising 35 shots, results in a σ 
of 0.997.

I took another 31 normal sights. Leaving out the two worst shots in this 
group for the moments results in a σ of 1.371 for 29 sights. Adding in a 
3.8 and a 4.4 intercept (the two worst shots) brings the σ to 1.756 for 
the 31 sights. Combining these 29 shots (again leaving out the two bad 
ones for the moment) with the previous 35 shots produces a σ of 1.252 
for the 64 shots. Adding in the two bad shots raises the σ to 1.433 for 
66 shots.

Of these 66 shots 4 had intercepts of zero. Another 20 had intercepts of 
0.1 to 0.5 NM making 24 out of 66 shots (more than one third) having 
intercepts of 0.5 or less. Another 13 shots were in the range of 0.6 to 
1.0 NM making a total of 37 out of 66 (more than half) with intercepts 
of 1.0 or less. There were 11 in the range of 1.1 to 1.5 making 48 out 
of 66 (72%) having intercepts of 1.5 NM or less. There were 9 more in 
the range of 1.6 to 2.0 making a total of 57 out of 66 (86%) having 
intercepts of 2.0 or less. There were 7 in the range of 2.1 to 2.8 and 
then the two outliers of 3.8 and 4.4 NM.

In addition to the normal shots, I also took a series of shots in the 
middle of the night with the horizon illuminated by moonlight. On 
November 3, 2009 at 18° 39' north, 40° 36' west between 0024 Z and 0039 
Z I shot Fomalhaut, Deneb and Vega three times each. The intercepts were 
0.3T; 0.3A; 1.3T; 1.5A; 1.5T; 1.6A; 2.5A; 2.6T; and 3.8A making the σ 
2.08 for these 9 shots taken in the middle of the night! Have others 
tried this?

I also took 3 observations of the upper limb of the sun at sunrise 
without a sextant, just watching the top of the sun appear to pop up on 
the horizon with just my Mark one eyeballs. The intercepts for these 
observations were 3.0A; 4.7 T; and 0.6A, a σ of 3.93 NM, not as good as 
with a sextant at higher altitudes but still a useful level of accuracy.

I am curious what others have found when they have done this type of 
investigation.

gl

   

Reply