|
|
Subject: Re: Captain Schufeldt's report on marine sextants
From: George Huxtable (george@XXX.XXX)
Date: Fri Dec 27 2002 - 11:28:10 EST
Marc Bernstein asked-
>Do you in fact get more accurate results from computer sight reduction and
>almanac than from the printed versions? Assuming of course that the accuracy
>of the sextant sight isn't an order of magnitude worse.
=====================
Yes, that appears to be the case.
In the Nautical Almanac we are informed that the tabulated quantities are
generally correct to the nearest 0.1', except for the Sun which has been
adjusted by up to 0.15' to simplify interpolations. When interpolating to
an intermediate time between the tabulated values, errors increase, and the
largest error that can occur in the interpolated GHA or Dec of any body
except for the Sun or Moon is less than 0.2'; it may reach 0.25 for the GHA
of the Sun and 0.3' for that of the Moon.
The most demanding requirement a navigator is likely to make of the
Nautical Almanac is in calculating the Lunar Distance between two bodies,
say the Sun and the Moon, in which the errors in both bodies will to some
extent combine. Because the lunar distance itself does not involve the
horizon (except for minor correction purposes), then measurements of lunar
distance can be made to great accuracy, maybe significantly greater
accuracy than the lunar distance can be calculated using the Nautical
Almanac.
For all other sextant measurements, in which the horizon plays a part, the
position of the horizon adds sufficient uncertainty that it almost
certainly swamps any error in the Almanac predictions.
On the other hand, computations of the positions of most celestial bodies,
without such interpolation, can be made to an accuracy suitable for
astronomers (MUCH more demanding than us navigators) using the methods
described in Meeus, Astronomical Algorithms. His Moon predictions, however,
are quoted only as being to "approximastely 10"" or 0.17', which appear to
be adequate for calculating lunar distances. Maximum errors are not given.
If higher accuracies are needed, Meeus provides references for obtaining
the data.
Her Majesty's Nautical Almanac Office issues a PC program on CD, "AstroNav
PC and Compact Data 2001 - 2005", also published in the UK by The
Stationery Office as "NavPac and Compact Data 2001 - 2005".
As well as the CD data, polynomials are printed so that positions can be
obtained using a calculator..
Claimed maximum errors (in GHA) are
Sun 0.03'
Planets 0.04'
Moon 0.21'
Stars 0.05'
This accuracy is probably sufficient for calculating lunar distances.
The conclusion is that the Nautical Almanac provides predictions of
sufficient accuracy, but only marginally so for calculating lunar
distances. Computed data can improve significantly on that accuracy.
There are many astronomical prediction programs now available for PCs, and
it would be interesting for owners to supply claimed accuracy for the
predictions of those programs, particularly as it applies to geocentric
position of the Moon.
George Huxtable
------------------------------
george@XXX.XXX
George Huxtable, 1 Sandy Lane, Southmoor, Abingdon, Oxon OX13 5HX, UK.
Tel. 01865 820222 or (int.) +44 1865 820222.
------------------------------
|