[vsnet-chat 1085] (fwd) Re: helio(bary)centeric correction (Thorstensen)

[Taichi Kato <tkato>]

From thorsten@dartmouth.edu Mon Jul 27 01:06 JST 1998
Date: Sun, 26 Jul 1998 12:04:46 -0400
From: "John R. Thorstensen" <thorsten@dartmouth.edu>
Message-ID: <vsnet-chat1085@hoge.baba.hajime.jp>
To: tkato@kusastro.kyoto-u.ac.jp
Subject: Re: helio(bary)centeric correction
Content-Type: text
Content-Length: 2198

Dear Dr. Kato,

   The JPL-based program that I have only has the dates 1980-2020 in
it, and the sample calculations you sent are for a JD2440404, which is
in 1969.  However, I do have a program based on the earth and planetary
ephemerides adopted from Meeus' book on astronomical formulae for
calculators, which does go to these earlier dates.  When I ran the 
calculation the barycentric JDs I produced matched your to 5 decimal
digits with occasional disagreements in the 6th place.  So this accuracy
is far below 0.1 sec.  

When I first checked your numbers using a Meeus-only correction, I got 
a very different number from what you had, but that was because
of the difference between TDT and UT.  My general calculator-type
program 'skycalc', which uses the Meeus ephemeris, assumes the input
is in UT and inserts a delta-T = TDT - UT correction before 
referring to the ephemerides, so in order to do the ephemeris
calculation for the same actual time I had to change my input 
time to be earlier by delta-T; since the barycentric correction is 
then applied to the input (UT) time, the barycentrically corrected 
time is also on the UT system, which is that much earlier than your numbers.

Of course the correct thing to do in critical cases is to state
the ephemerides in TDT, or TAI, or some other uniform time scale
rather than UT.  I haven't noticed much attention to this,
but I don't work on eclipsing systems that frequently so it may
have escaped my attention.

I did also run a calculation for you to crosscheck ... 

Object at J2000 coordinates 0h 0m 0s, +0d 0' 0''  .... 

TDT = JD 2449961.875000 (or approximately 2449961.874291 UT)

Barycentric correction derived from an early Meeus theory = 467.3 sec
                       derived from JPL DE200             = 467.340 sec

So this should come out as barycentric JD (TDT) 2449961.880409 in
your program.

I believe the JPL calculation used in this test-bed program does not
take into account the topocentric time correction (+- 21 msec or so).  

All this stuff is fairly trivial, but it's nice for both of us
to be able to get independent confirmation that our software is 
correct!


				Best,

					John Thorstensen

VSNET Home Page

Return to Daisaku Nogami

---

vsnet-adm@kusastro.kyoto-u.ac.jp