NavList:

   

Reply

On Apr 21, 2005, at 9:43 AM, Alexandre Eremenko wrote:

> ...verifying ... arc precision
> is not a trivial matter.
>
> If after a year of trials I will say that the arc error is 0.4'
> rather than, for example 0.2' as advertized, it can be always
> blamed on my poor skill, etc.
>

The crux of the matter, in my opinion, is to develop your skills before
initiating calibration.  In my opinion, proficiency has been
accomplished when the standard deviation of dry land altitude shots
using an artificial horizon is consistently (but not always) under 0.2'
of arc, and frequently under 0.1', for three to five observations
(increasing the number of observations also can decrease the the
standard deviation, up to a point, but numbers of observations cannot
substitute for proficiency).  Only when proficiency is attained is it
appropriate to actually use the observations for calibration, in my
opinion. It took me more than a year to become proficient.  Now I'm
sure there are some out there who can pick up a sextant and start
getting good observations in a few weeks, but not me!

Embarking on a program of calibration is a good way to develop
proficiency.  It's just that one has to discard the first few months
(or years) of data, prior to the development of proficiency.

Here are data for the string of sights when I finally felt I had
achieved proficiency on dry land.  These are altitude shots using an
artificial horizon, showing the mean Hc in decimal degrees and the mean
and standard deviation of Ho-Hc in minutes of arc.

Date          Object            mean
                    n   Hc(dd)         std

09/21/2004      Moon    3       19.070  0.1     0.23
09/22/2004      Sun     3       50.266  -0.1    0.30
09/23/2004      Sun     3       22.184  0.0     0.11
09/23/2004      Moon    3       28.744  -0.5    0.19
09/25/2004      Moon    3       33.045  -0.3    0.07
09/26/2004      Sun     2       30.092  -0.8    0.08
09/26/2004      Sun     5       30.867  -0.4    0.02
09/26/2004      Sun     4       24.766  0.0     0.14
09/29/2004      Sun     3       14.903  -0.5    0.16
09/30/2004      Moon    5       30.222  -0.4    0.15
09/30/2004      Sun     6       21.581  0.0     0.11
10/01/2004      Moon    6       32.747  0.0     0.14
10/01/2004      Sun     6       29.658  0.0     0.10
10/02/2004      Sun     6       29.541  0.0     0.15
10/03/2004      Sun     6       48.553  0.9     0.21
01/26/2004      Sun     3       23.419  -0.9    0.22

As a disclaimer, the only reason to pursue this extreme precision is to
calibrate a sextant, and that is useful primarily for lunars, which are
as dead as the Dodo, as was said more than 100 years ago.  If you can
get the standard deviation of Ho-Hc under 0.5' of arc on dry land,
that's good enough for use at sea, especially on small vessels.

   

Reply