--- In
guide-user@yahoogroups.com, Bill J Gray <pluto@p...> wrote:
> I can confirm the time you got from Guide. Dunno why Occult
> would differ by more than a few seconds... perhaps some odd setting
> buried in it? (I'm not really familiar with the workings of the
> program.)
Here is an explanation from David Herald, author of the Occult
program, on defining deltaT and deltaAT that is used in the Occult
program:
An Occult user asks: As I understand it, current delta T is ~ 64.6
s but my Occult version shows 2003=65.5 and 2004=65.8 Does it MATTER?
Do I have the current (correct) delta T file for Occult? Am I correct
that current delta T is 64.6? Should I edit the delta T file?
You need to be careful to distinguish between 'deltaT' and 'deltaAT'.
Occult maintains files for both of these values, but uses deltaAT in
preference to deltaT whenever available. And perhaps I should try to
explain the differences here, so all can understand....
There are some time scale issues that need to be understood.
* 1. Universal time [UT] - is strictly determined by the rotation
of the Earth. And because the Earth's rotation is gradually slowing
down, it is not a strictly uniform time scale.
* 2. Ephemeris time [ET] was derived around the '60's when the
Earth's rotation was established as being non-uniform. It is based on
the motion of the major planets as representing a uniform time scale.
The term "ephemeris time' derives from the comparison of the positions
of the planets against the 'Ephemeris'.
* 3. Atomic time [AT] is defined by atomic clocks, and is strictly
uniform. This time scale is linked to the earlier 'Ephemeris Time',
with a fixed offset of 32.184 secs.
* 4. The quantity deltaT is the difference between UT and
Ephemeris time. As such, the value of deltaT varies on a daily basis.
* 5. In 1972, there was a fundamental change to the way time
signals were broadcast. Rather than trying to maintain a broadcast
signal that was consistent with the variable rotation of the Earth,
the time signal was maintained fixed with respect to Atomic time. To
allow for the slowing rotation of the Earth, the time signal is
adjusted every so often by the insertion of a leap second - so that it
is always within 0.7secs of true Universal Time. this time signal is
called Coordinated Universal Time, or UTC. And if you listen to WWV,
there is coding in the signal to tell you the current offset between
UTC and UT in tenths of a second.
* 6. The quantity deltaAT is the difference between UTC and Atomic
time. Add 32.184 secs, and you have the difference between UTC and
ephemeris time. IMPORTANTLY, deltaAT is a fixed quantity between the
insertions of leap seconds.
Pulling all this together. For computing the positions of astronomical
objects relative to each other at a time measured using the broadcast
time signal (eg WWV, or GPS) - you need to use deltaAT. The difference
between UTC and UT is only important for measurements involving the
location of the observer on the surface of the Earth - to correct for
the rotational displacement of the Earth about its axis over that
period of time. [This is generally not important for predictions, but
is allowed for in the reductions done by OCCULT.]
In OCCULT, there are several challenges in dealing with the 'deltaT'
issues. For historical purposes, the values of deltaT are more or less
known back to the 1700's. But note that before about 1800, time
measurement was not very accurate; Harrison's chronometer, which
revolutionised time keeping, was a late 18th century creation. Prior
to then, there are various standard expressions which approximate
deltaT as deduced from historical solar eclipses.
In the far future, deltaT is also used, using computational
expressions when there are no values listed in the deltaT table. But
these values are of low precision.
Since 1972, and up to 2010 (presently) OCCULT uses deltaAT. IT DOES
NOT USE deltaT. This means that the times generated by OCCULT are
directly referenced to the broadcast time signals (i.e. UTC) - which
gives maximal accuracy.
For the future, there is the challenge of predicting the Earth's
rotation accurately. And I'm not aware that anyone has that
capability. What stands out is that through the 70's, 80's to the
mid-90's, UTC was being corrected with leap seconds at least once a
year. But since then corrections have been reducing in frequency -
such that there hasn't been a leap second insertion since Jan 1999 - 5
years ago - and there won't be one this January.
For the near future, the challenge is to predict when the next leap
second will be inserted. It will occur at the end of either a December
31, or June 30. I have guessed leap seconds for June 2004, Jan 2007,
and Jan 2010. These are listed under the deltaAT table in the form for
editing the deltaT values.
So while this is a bit long-winded, for predictions of events prior to
2010, OCCULT will use deltaAT to generate the UTC of events. The value
of deltaT held in the table is not used at all. So do not be concerned
about the values of 2003 65.2, and 2004 65.8. Changing these values
will NOT change the prediction times, as the deltaAT values will be
used. And the correction derived using deltaAT is 32 +32.184 = 64.184
secs - which will remain constant until the next leap second is
inserted into UTC.
Finally, if a leap second is NOT inserted this coming June 30, then
the deltaAT table will need to be edited to specify the best guess at
when it will be inserted. Otherwise all predictions after July 1 will
be out by exactly 1.0 secs.
Hope this is of some assistance in explaining this critical but
difficult topic.
Dave Herald - December 28, 2003