[fitswcs] [Fwd: SED data model v0.92]

[Steve Allen]

Greetings Jonathan,

On Thu 2004-11-18T12:22:38 -0500, William Pence hath forwarded:
> -------- Original Message --------
> Subject: SED data model v0.92
> Date: Wed, 17 Nov 2004 11:47:47 -0500 (EST)
> From: Jonathan McDowell <jcm at head.cfa.harvard.edu>
> Reply-To: Jonathan McDowell <jcm at head.cfa.harvard.edu>

> A group led by myself and Doug Tody have been working on
> the data model and data formats for the Simple Spectral Access Protocol,
> which represents 1D spectra, time series and spectral energy distributions.
> The document is still at a draft stage but has now reached a level
> of maturity that it may be interesting for people to look at.
> You can find it at
>  http://hea-www.harvard.edu/~jcm/vo/docs/spec.html
> and I'll put it on the twiki in the next few days.
>    Jonathan


I have an immediate comment about section 4.5

    The reference time should be specified by a date in ISO-8601
    format.  The default value of the reference time is
    -4713-11-24T11:59:27.81, corresponding to the origin of Julian Day
    Number on the TT (Terrestrial Time) timescale.  Using TT is
    preferable to UTC because it does not contain leap seconds, so the
    elapsed time in days is just equal to the difference in JD values.

Can you justify the use of ISO 8601 for dates prior to the existence
of the Gregorian Calendar?

Does the ISO 8601 standard explicitly admit that negative values for
the year are acceptable?

The phrasing of the first sentence is still ambiguous.
The reference time should have its timescale immediately adjacent to
its value.
I believe that you are trying to say that the origin epoch is
JD=0.0 (TT)
which corresponds to (assuming negative ISO year is acceptable)
-4713-11-24T11:59:27.81 (pTAI)
where by "pTAI" I mean some sort of proleptic TAI.

The concept of a proleptic TAI, however, is very problematic
given the change in rate of TAI by 1 part in 1.e12 in 1977.
Over the elapsed span of years since the origin of JD that
rate difference amounts to nearly a full second.  As such the
current TT - TAI difference of 32.184 seconds would not result in
a TAI of 11:59:27.81 so long ago.

I suggest that you make it explicit that the default reference origin
is JD = 0.0 (TT).  If you give the corresponding calendar date
you must first verify that ISO permits negative years.  If that is not so
it is far safer to express it as Julian proleptic date=-4713 January 0.5 (TT).

And this is to say nothing of the value of Universal Time that
corresponds to JD=0.0 (TT), because that date is several thousand
years prior to the earliest eclipse records found by Stephenson and
Morrison for determining earth rotation.

--
Steve Allen          UCO/Lick Observatory       Santa Cruz, CA 95064
sla at ucolick.org      Voice: +1 831 459 3046     http://www.ucolick.org/~sla
PGP: 1024/E46978C5   F6 78 D1 10 62 94 8F 2E    49 89 0E FE 26 B4 14 93