NavList:

In my experience using traditional techniques (ie eyeball and sextant), the "fuzziness" of the horizon can cause relatively large variations in the accuracy of my LOP's.  I wouldn't say that my sextant was inaccurate if I found a larger than expected error when I averaged all of my sights, some of which were taken in less than perfect environmental conditions.  My test has always been how accurate can I fix my position given good shooting conditions.

 

Jeremy

 

Just to give an idea on the sort of errors and their propagation for
this photometric observation:

Using my techniques and skill, I obtain:
UL: 3600, 3601, 3600, 3600, 3600, 3600 (mean: 3600.2 +/-0.5)
LL: 3515, 3515, 3515, 3515, 3515, 3515 (mean: 3515.0 +/-0.3)
Horizon: 507, 507, 505, 507, 506, 506 (mean: 506.3 +/-0.9)
As it happens very often, it's the horizon (and not of the sun) which
scatters most.

Resulting in:
UL-Horizon: 3093.8 pixels (+/- 1 pixel, corresponding to +/-0.4 moa)
LL-Horizon: 3008.7 pixels (+/- 0.9 pixels, corresponding to +/-0.3 moa)

Estimated HS using 2 parameter arc-tan calibration formula having a
StdDev of +/-0.2 moa, considering further the observational errors
(H.E. and timing) of +/-0.6 moa results in:
HSest (UL): 1160.2 +/- 0.7 moa, compared to the calculated 1159.3 moa
HSest (LL): 1128.7 +/- 0.7 moa, compared to the calculated 1127.9 moa

There can be several reasons for this observation to be slightly an
outlier like e.g. the measurements not made the same way as the
calibration was done or differences in refraction (inversion in the
morning, whereas calibration was done during the afternoon).

Marcel