PDS_VERSION_ID = PDS3 RECORD_TYPE = STREAM OBJECT = TEXT NOTE = "Comments on the Cassini SPICE Archive" PUBLICATION_DATE = 2015-09-23 END_OBJECT = TEXT END COMMENTS ON THE Cassini SPICE DATA ARCHIVE ====================================== This document contains comments and errata concerning all Cassini SPICE Data Archive Volumes. 1. Text Kernel File Format -------------------------- All SPICE text kernel files (LSK, PCK, SCLK, IK, FK) provided on these volumes have lines terminated by only (UNIX convention.) This is dictated by NAIF's archiving approach to include a consistent set of UNIX binary (IEEE architecture) and text ( only for end of line marker) kernels on the archive volumes. However, the format of the text kernels is specified in their associated labels as ASCII (INTERCHANGE_FORMAT = ASCII.) This is a violation of the PDS standard stating that lines in ASCII files must be delimited by the combination of . 2. Use of SPACECRAFT_NAME in the Labels --------------------------------------- PDS labels for SPICE kernels provided in this data set contain the deprecated keyword SPACECRAFT_NAME instead of the keyword INSTRUMENT_HOST_NAME. This is standard practice for SPICE volumes in the PDS and is dictated by NAIF's archiving approach to make SPICE archives for more recent missions consistent with SPICE archives for older missions, for which use of SPACECRAFT_NAME was legitimate. 3. Values of INSTRUMENT_NAME in the Labels ------------------------------------------ PDS labels for SPICE kernels provided in this data do not set the keyword INSTRUMENT_NAME to "SPICE KERNELS", that is the name of the instrument described by the instrument catalog file SPICE_INST.CAT provided in the data set. Instead this keyword is set to either "N/A" or the name(s) of the instrument(s) or structure(s), the geometric parameters of which a particular kernel file provides. While this error does prevent the individual data products from being associated with the "SPICE KERNELS" instrument, it does not affect usability of the SPICE data provided in this data set in any way because the SPICE toolkit does not use any information from the PDS label to access the data. The other side of this error is that the data set does not include instrument catalog files for the following instruments named in the INSTRUMENT_NAME keyword in the PDS labels: "CASSINI ORBITER" "CASSINI PLASMA SPECTROMETER" "CASSINI RADAR" "COMPOSITE INFRARED SPECTROMETER" "COSMIC DUST ANALYZER" "IMAGING SCIENCE SUBSYSTEM" "ION AND NEUTRAL MASS SPECTROMETER" "MAGNETOMETER" "MAGNETOSPHERIC IMAGING INSTRUMENT" "RADIO AND PLASMA WAVE SCIENCE" "RADIO SCIENCE SUBSYSTEM" "STELLAR REFERENCE UNIT" "ULTRAVIOLET IMAGING SPECTROGRAPH" "VISIBLE AND INFRARED MAPPING SPECTROMETER" 4. Missing FILE_RECORDS Keyword in the Binary Kernel Labels ----------------------------------------------------------- All labels for binary SPICE kernels provided in this data set are missing the FILE_RECORDS keyword, which is required because these labels have the RECORD_TYPE keyword set to FIXED_LENGTH. While this error violates the PDS standards, it does not affect usability of the SPICE data provided in this data set in any way because the SPICE toolkit does not use any information from the PDS labels to access the data. 5. Incorrect SCLK Values in the Comment Area of CK Files -------------------------------------------------------- The values of the SPACECRAFT_CLOCK_START_COUNT and/or SPACECRAFT_CLOCK_STOP_COUNT keywords in the copy of the PDS label embedded at the top of the comment area of CK files are incorrect for some of CK files containing more than one segment. The values of these keywords provided in the detached PDS labels are correct. 6. Bad Version 1 MIMI/LEMMS CKs ------------------------------- The version 1 MIMI/LEMMS CKs covering from June 2000 to February 2005, added to the data set in the release 20 provided erroneous data for a significant part of their coverage, especially during the period from April 2001 to October 2003. The version 2 MIMI/LEMMS CKs covering the same time interval, added to the data set in the release 26 fully supersede the version 1 CKs. They fix most of the errors found in the version 1 CKs, except for a small set of time intervals when instrument pointing could not be reliably determined. Data for these intervals was removed from the version 2 MIMI/LEMMS CKs and a special CK provided in the data set for the record. 7. Empty CK Comparison Reports Between 2010-306 and 2010-328 ------------------------------------------------------------ Because the Cassini S/C went into safing at 2010-306T21:59:54 in the sequence S64 and a background sequence was not running running until the beginning of the sequence S65 at 2010-328T19:35:00.000, this entire time period is masked in the CK comparison reports. 8. Incompatible Meta-kernels ------------------------------------------------------------ Because the following meta-kernels cas_1997_v11.tm cas_1998_v11.tm cas_1999_v11.tm cas_2000_v13.tm cas_2001_v13.tm cas_2001_v14.tm cas_2002_v13.tm cas_2003_v13.tm cas_2004_v13.tm cas_2005_v14.tm cas_2006_v13.tm cas_2007_v11.tm cas_2008_v09.tm cas_2009_v06.tm cas_2010_v05.tm cas_2011_v01.tm cas_2011_v02.tm cas_2011_v03.tm cas_2011_v04.tm added in releases 0028, 0029, 0030, and 0031 listed the smaller satellites SPK ``110120BP_IRRE_00256_25017.bsp'' that erroneously included data for Saturn relative to the Saturn Barycenter, they could be used only by themselves. If they were loaded simultaneously with each other or any other meta-kernels, the Saturn-to-Saturn Barycenter data from ``110120BP_IRRE_00256_25017.bsp'' masked the Saturn-to-Saturn Barycenter data from the spacecraft SPKs causing the Saturn-relative positions of the CASSINI spacecraft and major satellites computed by SPICE to be incorrect by as much as 750 km. On November 13, 2012, in between the 0031 and 0032 releases, these meta-kernels were supersedes by the following meta-kernels cas_1997_v12.tm cas_1998_v12.tm cas_1999_v12.tm cas_2000_v14.tm cas_2001_v15.tm cas_2002_v14.tm cas_2003_v14.tm cas_2004_v14.tm cas_2005_v15.tm cas_2006_v14.tm cas_2007_v12.tm cas_2008_v10.tm cas_2009_v07.tm cas_2010_v06.tm cas_2011_v05.tm listing the older smaller satellites SPK ``090202BP_IRRE_00256_50017.bsp'' that did not have data for Saturn relative to the Saturn Barycenter and did not pose a problem for simultaneous loading of multiple meta-kernels. 9. Inconsistent CK Coverages in MKs in Releases 0008-0042 ------------------------------------------------------------ In meta-kernels (MKs) for the years 2003-2014 provided in Releases 0008-0042 (April 2007 - June 2015), up to and including MKs cas_2003_v21.tm cas_2004_v21.tm cas_2005_v22.tm cas_2006_v21.tm cas_2007_v19.tm cas_2008_v17.tm cas_2009_v14.tm cas_2010_v13.tm cas_2011_v12.tm cas_2012_v09.tm cas_2013_v07.tm cas_2014_v03.tm for CKs included in a meta-kernel the combined predicted CK coverage extending outside of the year boundaries was in most cases greater than the combined reconstructed CK coverage. This is not a problem when using a single meta-kernel and asking for attitude data strictly within the year boundaries, -- if available, reconstructed attitude will always be returned "ahead" of predicted attitude in such cases. This does become a problem, though, when asking for data outside of the year boundaries, where reconstructed attitude is not available but predicted attitude is, or when loading multiple meta-kernels together, in which case predicted attitude data from a later loaded meta-kernel can mask reconstructed attitude data from an earlier loaded meta-kernel, usually at the start or at the end of a year and spanning days or even weeks. Specifically for the latest yearly meta-kernels as of release 0042 (June 2015) loaded in the year-increasing order, predicted CKs provided by meta-kernels shown in the first column would mask reconstructed CKs from earlier years' meta-kernels during the periods shown in the second and third columns: ------------------------------------------------------------- Meta-kernel Start UTC Stop UTC ------------------------------------------------------------- cas_2005_v22.tm 2004-12-16T13:21:59 2004-12-26T00:00:03 cas_2006_v21.tm 2005-12-17T14:21:00 2005-12-28T00:00:00 cas_2007_v19.tm 2006-11-24T16:29:59 2006-12-28T00:00:02 cas_2008_v17.tm 2007-12-14T15:59:59 2007-12-28T00:00:00 cas_2009_v14.tm 2008-11-26T17:55:00 2008-12-27T00:00:01 cas_2010_v13.tm 2009-12-22T23:26:00 2009-12-27T00:00:00 cas_2011_v12.tm 2010-11-24T19:34:59 2010-12-30T00:00:02 cas_2012_v09.tm 2011-11-16T03:02:00 2011-12-30T00:00:02 cas_2013_v07.tm 2012-11-02T14:30:00 2012-12-29T00:00:01 cas_2014_v03.tm 2013-12-28T01:46:59 2013-12-29T00:00:01 ------------------------------------------------------------- The same meta-kernels, loaded in the less likely year-decreasing order, would have predicted CKs mask reconstructed CKs from later years' meta-kernels during these periods at the start of each year: ------------------------------------------------------------- Meta-kernel Start UTC Stop UTC ------------------------------------------------------------- cas_2013_v07.tm 2014-01-02T23:59:58 2014-03-13T21:11:49 cas_2012_v09.tm 2013-01-02T23:59:58 2013-01-13T17:50:49 cas_2011_v12.tm 2012-01-03T23:59:56 2012-01-24T22:54:49 cas_2010_v13.tm 2011-01-03T23:59:59 2011-01-17T08:41:49 cas_2009_v14.tm 2009-12-31T23:59:57 2010-01-23T14:59:51 cas_2008_v17.tm 2008-12-31T23:59:58 2009-01-09T15:15:50 cas_2007_v19.tm 2008-01-01T23:59:57 2008-01-22T13:34:52 cas_2006_v21.tm 2007-01-01T23:59:56 2007-01-05T13:49:53 cas_2005_v22.tm 2006-01-01T23:59:57 2006-01-27T04:02:52 cas_2004_v21.tm 2005-01-01T23:59:59 2005-01-22T10:37:50 cas_2003_v21.tm 2004-01-02T11:59:57 2004-01-09T12:08:17 ------------------------------------------------------------- In release 0043 (October 2015), starting with MKs cas_2003_v22.tm cas_2004_v22.tm cas_2005_v23.tm cas_2006_v22.tm cas_2007_v20.tm cas_2008_v18.tm cas_2009_v15.tm cas_2010_v14.tm cas_2011_v13.tm cas_2012_v10.tm cas_2013_v08.tm cas_2014_v04.tm this was fixed by including enough reconstructed CKs in each MK to fully cover the span of all predicted CKs listed in it.