|
|
|
|
|||||
|
||||||
Subject: Re: Accuracy of position
From: Dan Allen (danallen@XXX.XXX)
Date: Tue Oct 19 1999 - 13:29:00 EDT
As I mentioned once several months ago, I have been able to determine my
position on land to a few hundred yards using metal sextants such as a
Tamaya Jupiter or C. Plath Navistar Professional.
My procedure for these measurements was:
1. Use my Garmin GPS III in averaging mode for 20 minutes to determine my
location to within 20 feet. This was my assumed position.
2. Measured the altitude of the sun using a Davis Artificial Horizon, so the
angles measured were double the actual values.
3. I then input the date and time and measurements into a command line tool
I've written in the programming language C, with the following invocations
of the tool (which contains a nautical almanac accurate to 1" of arc and all
of the nav triangle computations as well):
sun "7/18/1999" "2:41:19 p" -a 47~28.9 -b 121~47.9 -y 116~12.4 # Plath
sun "7/18/1999" "2:42:45 p" -a 47~28.9 -b 121~47.9 -y 115~52.0 # Plath
sun "7/18/1999" "2:44:55 p" -a 47~28.9 -b 121~47.9 -y 115~33.8 # Jupiter
sun "7/18/1999" "2:47:42 p" -a 47~28.9 -b 121~47.9 -y 114~43.8 # Plath
sun "7/18/1999" "2:49:08 p" -a 47~28.9 -b 121~47.9 -y 114~26.4 # Plath
The -a and -b options for my tool input the assumed position (as measured by
my GPS in step #1), and the -y option inputs the altitude that I measured in
step #2 via the artificial horizon, which is then halved by the program.
The output of these calculations are distances to the assumed position:
Results with default refraction:
-0.95 nmi
-1.49 nmi
4.23 nmi
-1.41 nmi
0.01 nmi
Results at 70?F with 29.45 inHg: (the measured conditions at my house)
-0.92 nmi
-1.46 nmi
4.26 nmi
-1.38 nmi
0.04 nmi
Mean error = 0.108 nmi ~ 656 feet ~ 219 yards
Not bad! Note that individual sights varied more. I want to repeat the
experiment but on the waves of the tossing sea.
Dan
Daniel K. Allen --> danallen@XXX.XXX
Navigate | Calculate | Set Sail!
-----Original Message-----
From: Navigation Mailing List
[mailto:NAVIGATION-L@XXX.XXX Geoffrey
Kolbe
Sent: Tuesday, October 19, 1999 12:10 AM
To: NAVIGATION-L@XXX.XXX
Subject: Accuracy of position
I would welcome some input on the accuracy of various aspects of celestial
navigation using a sextant. I apologise if this has all been discussed
before, (it almost certainly will have been).
The first question concerns the absolute accuracy obtainable using a
sextant to determine an altitude. I have briefly done some experiments (on
dry land) where I used an artificial horizon (Freiberg) carefully levelled
to better than 0.1'. The purpose of using an artificial horizon was to use
the doubled altitude angle to effectively halve the errors in the arc and
in reading the angle. I was able to obtain readings that were consistently
within 0.2' of what they should have been. Has anyone else any experience
of what kind of accuracy in the altitude is practically obtainable using a
sextant?
The next question is, what sort of practical working accuracy would you
expect using a marine sextant at sea? Would 1.0' be a fair estimate or
would you reasonably expect better (or worse) than this?
The final question is, in your experience, what sort of accuracy in the
computed position have you come to expect from the observation of two or
more celestial bodies? In other words, having computed your position, what
is the radius of the "comfort zone" circle that you would draw around that
position, such that you would want known dangers to be outside this circle?
(Note that this is a different question to what sort of theoretical
accuracy should you expect.)
Thankyou.
Geoffrey.
Dr Geoffrey Kolbe, Border Barrels Ltd., Newcastleton, TD9 0SN, Scotland
Tel: +44 (0)13873 76253 Fax: +44 (0)13873 76214
www.border-barrels.com
|