Notes from the 2002 Comet Data Review

General Topics

• Mal notes that the MISCHA labels do not include COLUMN_NUMBER keywords, despite the fact they contain 54 columns! (Actually they do, as Ed showed.)

• Tim suggests that there should be more description about general organization, what tools are available and how to use them. ReadPDS failed on the IUE binary tables! He also notes that the IUE binary tables really should have been ASCII tables because the record size is relatively small.

Data Sets

DeVico Spectral Atlas (Anita Cochran)

• Mike D. is rather impressed by the data and organization.

• Lien 1: Add note to DATA_SET_DESC, something like "Note that the atlas presented here is for line identification only. Those wishing to further study the original digital data should contact the data provider, Anita Cochran, directly."

• Lien 2: Add the field of view description from the instrument file to the DATA_SET catalog file.

• Lien 3: Add references to the PNG files to the catalog and documentation.

• Lien 4: The PNG files in the context/ directory are all identical - which is incorrect.

• Lien 5: in the context.pdf file there is a table of contents that ought to be converted to an ASCII Table file for reference.

• NOTE: There is no way to ingest the data as is in a reasonable format currently supported by PDS. We'll have to work that one out with CN.

• Tyler's list:
  • INSTRUMENT_HOST_NAME, INSTRUMENT_NAME and TARGET_NAME vary between catalog and data files
  • There is no ASCII equivalent for the documents - but there can't be

Borrelly Spectra (Anita Cochran)

• Desire for the underlying observational data. Mike A. is in favor of actively pursuing the flux-calibrated data. Mal suggests not ingesting these until the flux-calibrated data are also available for ingest. Tony will ask about these data and data taken subsequently.

• Mike A. notes that it would be useful to have the current data, even without the flux-calibrated data, in view of the upcoming Borrelly data review.

• Mike D. notes that the ':' and '::' uncertainty flags are never quantified. They aren't quantified in the paper, either, but Tony will see if Anita can provide some additional precision for these.

• Lien 1: Add a note to the data set catalog file that we anticipate an update of this data set when the flux-calibrated data become available.

• Lien 2: Mike D. notes that the column label says "Hour angle", whereas the label says "Right Ascension". The label should be changed to agree with the published table, which uses HA.

• Lien 3: Mike D. has marked-up labels with typos and errors that we should fix. Things like:
  • Missing units in some places
  • Wrong wavelength - "500" rather than "5000"

• Lien 4: obsparam.lbl gives bad desciption for "START TIME" column.

• Lien 5: The headers in the data file are not declared in the label. They should probably be removed anyway.

• Lien 6: Mike D. would like to see times as well as dates. Tony will ask Anita about this as well.

• Lien 7: Tyler notes catalog files have not been validated. Also INSTRUMENT_NAME varies between catalog and data files. Further comments TBD

SL9 Photometry -(Laura Woodney)

• Dave Schliecher joins via telecon.

• Tim notes there is no ephemeris for Io covering the time period of the data. There really ought to be one. It is not easy to determine the exact observation times, although it is probably true that the instructions provided in the data set catalog file might be sufficient. Note that these data are completely raw, but it is probably true that they could be reduced, if someone was determined to do so.

• The paper Laura submitted to Icarus was rejected but is probably around somewhere. We should probably publish it here as part of this data set.

• Dave notes the unique nature of this data set as an argument for archiving.

• Lien 1: Laura's manuscript should be required for archiving.

• Lien 2: Note that these labels and catalog files need to be seriously verified. The DATA_SET_ID is wrong, for example.

SL9 Images - (Marc Buie)

• These images were an attempt to observe flashes. Jupiter and previous impact sites are visible in the images. Dave S. notes that the images give a good temporal sequence of events during that week, which should probably be emphasized over the flash-search which is the current focus of the aareadme.txt.

• Dave S. notes that calling the collection "Lowell" implies that the data were taken at Lowell observatory. The data directory name should probably be changed to some reasonable neutral acronym. Also, in the index directory, PDS keywords like "LINES" and "LINE_SAMPLES" is confusing to naïve users. Also in the index files, the image sizes change without explanation.

• Dave was unable to locate any processing history or statement about the reduction state, but there is a "Processing" section in the data set catalog file that seems to provide this information. Dave didn't find it, though, which is hardly surprising.

• Lien 1: The aareadme.txt and data set catalog files should clearly and quickly state the systems used and any associated parameters.

• Lien 2: In the ref.cat file, the name "Chodas" is misspelled in the refkeyid for Chodas & Yoemans (1996)

• Lien 3: Error in the aareadme.txt file, which claims the images are attempting to observe flashes on Jupiter, when in fact they are looking for flashes on the Galilean satellite

• Ed also noted several additional liens.

• Tyler has a list of typos TBD.

SL9 Hi/Low Dispersion Spectra (IUE MAST)

• Tim reports:
  • ReadPDS would not run on the binary tables.
  • There didn't seem to be any global targeting information - he had to wander through directories.

• Lyle notes:
  • There is no ephemeris information here. Shouldn't there be?
  • In the index directory, the labels are those for the CDB indices, not these indices.
  • The labels in the document/iuecal directory are also incorrect (wrong DSID).
  • Are the scripts available for archiving? (Ed will check)
  • There should be cumulative indices covering both volumes (13 and 14).

• Ed notes that three of the files are missing pointing information. Although it appears the pointing will never be available, reviewers agree the data should still be included in the archive. Tim notes some information might be in the original scripts. Ed will see if these are available and useful. Tim has offered his notebook for scanning.

• There was a brief discussion about what to do with those annoying VICAR headers in the IUE file labels. It's currently in the DESCRIPTION field, which is what we would display in an automated system - thus the problem. There is general agreement that the VICAR header should be preserved somewhere in the archive.

• NOTE: We should probably provide the VICAR standard as part of our holdings. Should it go in the document directory of these disks as well? We'll need to check copyright.

• Lien 1: Correct problems noted by Lyle.

• Lien 2: Vol 14 index directory labels START_BYTE values in the column definitions are wrong.

• Lien 3: Tyler's list of typos TBD.

• Lien 4: Observing scripts must be scanned from Tim's notebook and added to the archive, with the appropriate cross-indexing. These scripts have been digitized for the IUECDB Files. Tim will supply the notebook for the SL9 observations and then these will be Digitized before the SL9_0014 volume is released.

• Lien 5: Must add a PERSONNEL.CAT file.

IUE Comet Data Base (Paul Feldman)

• Tim notes this one has ephemerides, and found them very helpful. Also liked the organization (by MPC ID) and the cross-id table. ReadPDS doesn't read the merged spectra (as before).

• Lyle's notes:
  • Aareadme.txt has wrong name for ephemeris directory
  • DATA_SET_ID is truncated in the index.tab file

• The ephemerides were included as just labeled output from the Horizons system. This seems to be very readable and helpful - no need to turn them into separate TABLE files.

• Lien 1: Correct aareadme.txt as above.

• Lien 2: TARGET_NAME fields must be corrected in the data labels to contain the full name string.

• Lien 3: Correct DATA_SET_ID values in the index.tab files

4 December 2002 - Afternoon Session

Data Sets - Continued

VEGA 1/2 MISCHA (Magda Delva)

• Steve Joy, Joe Mafi, and Ray Walker join via telecon

• Mal thinks the dataset still needs a lot of work. Data quality is highly questionable. Specifically, he is concerned about the zero field offsets in various fields. He sees conflicts between offsets listed in the data files and the offset file. Documentation is not sufficient for archive purposes.

• Mal and Steve discussed the offset issue. Mal noted the different coordinate systems used between VEGA 1 & 2 and discovered an offset problem in the CSE coordinates. (VEGA2 data is only available in CSE.)

• An additional problem: The transformation matrix used to rotate coordinates from PSSO to CSE coordinate was not time-variable. They data producers considered this good for about 4 hours to either side of close approach, but Mal feels this is just sloppiness in data reduction. This exacerbates the situation in the VEGA2 data because SE coordinates are not available at all.

• Ed notes that the "original" files in the EXTRAS directory contain none of the faulty transformations. Theoretically you could work forward from this, but the original producers did not feel they could produce good offsets for the critical period. Mal is not sure anything meaningful could ever be extracted from these data.

• Steve notes that the instrument failed shortly after the end of the offsets listed in the file, although that fact is not noted prominently in the support files. It is possible the instrument was becoming unstable, making offsets impossible to calculate.

• Steve Joy is concerned that there is no mention of how saturated data are handled, plus the transformation applied masks the fact that saturation even occurred. Also, there is no mention of how the digitization/quantization was performed.

• Mal notes that the VEGA 2 data prior to Feb 1985 might be all right - it's after that point (during encounter) when it goes wonky.

• Recommendations:
  • Safe the VEGA 1&2 original data, currently located in the extras/ directory, with added documentation regarding the known problems. (PPI will provide the minimal labels and DSID for safing.) The safed data will be moved to a separate volume set.
  • VEGA 1 processed data requires additional review and documentation updates before archiving for the various caveats and problems discussed. The CONFIDENCE_LEVEL_NOTE should mention the saturation problem (as well as the aareadme.txt).

• SBN will upgrade the labels for the processed data.

• VEGA 2 (processed data) recommendations:
  • Drop processed data after Feb 21 (when it goes to 999 flags and becomes highly suspect). (Labels and catalog files will have to be updated.)
  • All documentation upgrades made to VEGA 1 should be made here.

Lowell Observatory Comet Data Base Observations and Production Rate Summaries

• Dave doesn't like the acronym "LOCDPR". Change it to "LOCDSUM".

• Lien 1: Change LOCDPR to LOCDSUM (data set catalog file).

• Lien 2: Dave's notes were distributed to all reviewers
  • In the host.cat the 21" telescope listed was never used in this program. Instead the 24" Morgan telescope (f13.5) was used. Check the 1981 Astronomical Almanac for information on this instrument. The plate scales listed are nominal, not measured. This should be noted.
  • The data set catalog files need to have the old comet designations added (as in the data files.
  • The data files have blank CHARACTER fields, which caused Dave a problem loading this table in to a spreadsheet. He requests a character filler. We need to discuss this with the subnodes, then act accordingly.
  • LOCDFLUX.TAB uses 1e32 for missing values. There are two situations: never measured and unsuccessful observation. If it is not practical to do so, then at least we need to indicate what situations the flag value might represent.
  • RESTRICT.TAB - could not review in advance because file was not ready. Dave will look it over and send us comments.

• Lien 3: Mike A.'s notes
  • In the catalog files, "A(q)frho" and it should be "A(theta)*f*rho"
  • Also in the catalog file, A(theta) is the observed phase angle, not the observed Bond albedo. See Mike's email for further typos and definition problems.

• Lien 4: Column headers will again be discussed with the full SBN. The interim solution is to remove the headers and supply a tool to remove (or replace ) the headers.

End Discussion

• Tony mentioned two other ground-based data sets: Dave Schleicher's and Nalin Samarasinha's. The problem is the state of the data and time for reduction.

• Also Hal Weaver's HST stuff (high-resolution, near-nucleus images of various comets). Mike suggests a index for finding HST comet observations as a first step. Mal suggests downloading the abstract catalog and searching for "comet" in order to get the program IDs, which will then lead to the observations.