|
|
|
|
|||||
|
||||||
Subject: Re: September Equinox computation
From: Dan Allen (danallen46@XXX.XXX)
Date: Tue Sep 24 2002 - 12:29:16 EDT
On Tuesday, September 24, 2002, at 04:43 AM, Pierre Boucher wrote:
> Which method would you use to PRECISELY compute (hh-mm-ss) the
> September
> equinox?
This is actually a fairly hard problem. We usually know that the
equinox
has a declination of zero and a right ascension of 12:00:00 hours in
September.
My computation calculates the position of the sun using the Meeus
formulas and
iterates with a secant root finder to find when the declination is
zero, or when
the right ascension is 12:00:00. The hard part is that our
mathematical models
do not always have a time when these two quantities have these two
desired values.
I always seem to be a few minutes off of what the Naval Observatory
says is the
beginning of Fall and I have discussed this with a friend at the Jet
Propulsion
Laboratories (JPL) and he says that what I need to get the "right"
answer is the
full DE200 emphemeris package, which is huge and largely in Fortran.
Alas, I
am a C programmer.
Dan Allen
PS - My computations for this year put the equinox at 4:52:04 AM GMT on
23 Sep 2002.
|