|
|
|
|
|||||
|
||||||
From: Alexandre Eremenko (no email)
Date: Sun Apr 17 2005 - 21:34:52 EDT
Bill,
> In the revised paragraphs he does mention that
> refraction must be accounted
> for, but his method is simplistic.
> Take the difference between the two
> object's refraction and subtract from calculated true separation.
This method is incorrect, and this was discussed on the list
last October. The formula which produces the correct
refraction correction is essentially the same as for
the lunar distances, and it takes some time to compute by hand.
One of my Russian books has the following tables, specially
designed for determination of the instrumental correction.
It lists uncorrected distances from the Polar star to
6 other stars at distances 16, 34, 43, 51, 60 and 81 degrees.
(These stars are Kochab, Alioth, Capella, Vega, Alpheratz,
and Altair).
And then it gives 6 tables (one for each star) for
the refraction correction. The entry in these tables uses
altitude of the star and latitude.
These tables permit you to determine your instrumental error
(in the Northern hemisphere) without any calculations.
Just measure the distances and altitudes.
Altitudes are needed with only 5 degrees presision,
so this can be done with a cardboard sextant which has
artificial horizon, or with a star globe or with Rude
starfinder. So you don't need the horizon to do it.
I think it was quite a clever idea to design such tables:
the Polar star is always visible in our latitudes:-)
I can make you a copy if you wish; I hope you will
not confuse the Russian names of the stars:-)
However, one of the tables (the table of non-corrected
distances) has to be recomputed because the table
in my book uses the ephemerides of 1960.
For the refraction tables this is irrelevant.
Alex.
P.S. I have to say that so far, with all my efforts,
I could not determine reliably my own sextant's arc error,
if there is any. The problem is that I get contradictive
results. According to the same book the mean quadratic
error of a single star-to-star distance with SNO-T is
less than 0.2' if the lower star is on the right, but
for some reasons I cannot reach this precision consistently.
|