> From: Robert A. Jacobson [Robert.A.Jacobson@jpl.nasa.gov]

> Sent: Tuesday, November 18, 2003 2:16 PM

> To: mgordon@mail.arc.nasa.gov

> Cc: luke@boulder.swri.edu; rfrench@ahab.wellesley.edu;

> c.d.murray@qmul.ac.uk; lissauer@ringside.arc.nasa.gov;

> showalter@ringside.arc.nasa.gov

> Subject: data archive

>

> Mitch,

>

> Overall the archive looks pretty good. I had no problem understanding it, and my spot check verifies that I am using the same data that is contained in tables. My comments are:

>

> 1. in the *.LBL files you should state that the observation times are UTC.

> The UTC time scale is mentioned in the original text files but not in

> the basic data description.

Done.

> Since the archive is designed for someone

> in the "future" it might be useful to provide some sort of reference,

> if it exists, describing how the HST clock times are related to UTC.

We were unable to find such a document.

> 2. Through out it is stated that the observations are in the "J2000"

> coordinate system. Again, since this is a long time archive, you might

> want to switch to calling the coordinate system the International

> Celestial Reference System (ICRF). The ICRF for all practical purposes

> is identical to the FK5/J2000 optical system, and most work in the

> future will be referring to the ICRF and not the J2000 system.

The current PDS standard is to give RA & DEC in the J2000 system. Your suggestion requires a formal change which is being initiated.

> 3. The data are given as offsets from the planet, i.e. the position of the

> satellite relative to the planet. It is implied that the position is

> as seen from HST. We currently process the data as if the positions are

> those seen from the center of the Earth which is not the case. It would

> be useful to either

> a. provide a link to a source of HST positions at the observation times

> or

We have included a pair of SPICE SPK files for the ephemeris of HST covering the period of this data.

> b. provide a "parallax" type correction much in the same way as classical

> topocentric observations were translated into geocentric

Included comments on this in the DATAINFO.TXT file. From the review my Mike Evans, “The differences (Earth centered vs. HST centered) in the derived orbital elements are much smaller than the associated errors.”

4. The format of the archive with the data for each satellite in a separate

> file is clearly applicable to the case of planet relative positions. It

> is also applicable to the case of absolute astrometric positions. If you

> have satellite-satellite relative positions, however, what format would

> propose? Should a totally different approach be taken or should your

> data files be extended to explicity include a designation for the

> "observed" satellite and the "reference" object (planet, satellite, or

> none for absolute positions)?

Clearly, it would depend on the nature and volume of data. In principle I believe we’d try to retain the current approach but expand the tables to indicate the "observed" satellite and the "reference" object explicitly.

> Bob Jacobson

>