========= DISP_1000 ========= Reviewers: S. Joy, D. Scheeres General ------- - All SPICE kernels load correctly as long as one reads the documentation. - Why aren't the interpolated CKs not included in this archive? These were generated to estimate pointing for gaps in actual CKs. - Center point of nucleus is not where it is predicted by SPICE for some pre-impact images. Documentation indicates there is a kernel that should produce the correct information. Could this be the integer issue with mounting angles? Could N. Mastrodemos help us here, to help resolve this issue with Tony? /data - All labels contain product_version_type of ACTUAL (some PDSDD values are ACTUAL, PREDICT, RECONSTRUCT, etc). Why use ACTUAL? Why not PRELIMINARY ACS RECONSTRUCTION. - Instrument name in data set label is not used else where. - All binary kernels, are fixed length but the required keyword for FILE_ECORDS is not included. /data/ek - Hard to validate (command dictionary and events log). Can NAIF provided EK files in a easily viewed format, e.g., ASCII? - Are these kernels ITAR-clearance? Ask NAIF to verify this. data/ik - No discussion of the optical distortion in the HRII kernel. There is a discussion for HRIV. Per MA, state optical distortion from HRI telescope that falls on the slit same as that for HRIV. Add to comments section of data/sclk - The ifo.txt file thoroughly documents how to use these data. Dataset.cat tells you how to load the appropriate SCLK kernels. data/spck - Again, info.txt file thoroughly documents how to use these data. data/fk - Why do HRII, HRIV, and ITS have integer rotation frame from primary s/c axes (i.e., why are all decimal places zero)? Are these predicted values? Only MRI has rotations stated as fractions of a degree. We do not expect integer degree rotations for the final values. - Consider making corrections recommended by S. Joy (appended below). These recommendations are discussed above. - D. Scheeres' did not have any corrections but his presentation for the review is appended below for completeness. ______________________________________________________________________________ ______________________________________________________________________________ Steven Joy's List of Recommendations ------------------------------------ Deep Impact - SPICE Review Steven Joy 3/30/06 Overview The SPICE data appear to be in very good shape despite the apparent volume of comments that follow. For the most part, the documentation is clear and detailed, and the data files all appear to be correctly formatted. The content of SPICE kernels is a little tricky to validate without extensive and detailed analysis which I did not perform. I did load all kernels into SPICE based applications and generated basic ephemeris and pointing data. In this process I was able to determine that the kernels are at least plausible (cameras point at the comet, Impactor hits the comet at or very near the published time of impact, Flyby spacecraft position wrt Sun and comet consistent with lighting conditions observed in images, etc.). Pointing errors of a few degrees or ephemeris errors of a few 10s of kilometers would have been undetected by my analysis. ======================================================================== ROOT ======================================================================== The AAREADME.TXT, AAREADME.LBL, and ERRATA.TXT files are acceptable. The AAREADME.HTM file is acceptable, but could be improved if it included a pointer to the "dataset.html" document. This subordinate document makes browsing the data set more convenient than the mechanism provided currently within the AAREADME.HTM file. The dataset.html file does not belong at the root level of the volume. The file should be moved to an EXTRAS directory and the paths to the various links corrected accordingly. There are a few links that are only valid for the SBN peer review web page (links to the .tar and .zip files). However, the file does add value to the disk and should be retained if possible. ======================================================================== CATALOG ======================================================================== DIF.CAT, DII.CAT, PERSON.CAT, RELEASE.CAT, REF.CAT =================================================================== Catalog files are acceptable as is. SPICE_HSK.CAT ============= The catalog directory and the .CAT file name extension are reserved for files PDS catalog objects (CATINFO.TXT file exempted). None of the PDS 3.5, 3.6, or 3.7 standards include a DATA_SET_HOUSEKEEPING object. Clearly this file is something requested by the PDS. Please put it in the EXTRAS directory (as is) or in the DOCUMENT directory as a labeled text file (change the extension to .TXT). CATINFO.TXT ============ See comment on SPICE_HSK.CAT and remove the line referencing the file from this file. 1867G1.CAT ========== File name is different than on the ITS archive volumes, and I prefer the name on the ITS volumes (9p_tempel_1_1867_g1.cat) The target description appears to be a bit sparse. I'm used to seeing things like: A_AXIS_RADIUS : 71492. B_AXIS_RADIUS : 71492. BOND_ALBEDO : UNK C_AXIS_RADIUS : 66854. FLATTENING : 0.0065 MAGNETIC_MOMENT : 1.55E+20 MASS : 1.8988E+27 MASS_DENSITY : 1.330 MINIMUM_SURFACE_TEMPERATURE : UNK MAXIMUM_SURFACE_TEMPERATURE : UNK MEAN_SURFACE_TEMPERATURE : UNK EQUATORIAL_RADIUS : 71492. MEAN_RADIUS : 69911. SURFACE_GRAVITY : 25.9 REVOLUTION_PERIOD : 4333.0 POLE_RIGHT_ASCENSION : 268.0 POLE_DECLINATION : 64.50 SIDEREAL_ROTATION_PERIOD : 0.410 MEAN_SOLAR_DAY : 0.410 OBLIQUITY : 3.100 ORBITAL_ECCENTRICITY : 0.048 ORBITAL_INCLINATION : 1.300 ORBITAL_SEMIMAJOR_AXIS : 778376719. ASCENDING_NODE_LONGITUDE : 100.50 PERIAPSIS_ARGUMENT_ANGLE : 275.20 (from the Jupiter target template) included in the description. Given that you are able to produce a 50 year SPK kernel and a PCK kernel, it seems as though you could provide information about the radii of the triaxial ellipsoid, pole orientation and rotation rate, and orbital parameters here. SPICEDS.CAT =========== In general, this document does an excellent job of describing the SPICE data set, its usage, and problems associated with its creation. There were numerous typos and grammatical errors which I found it easier to correct than describe. In addition, I added or modified a few sentences to make the descriptions a little more clear to readers with less PDS/SPICE experience. In particular, I found myself writing comments on various sections only to find that the comment was addressed later in the file. In these areas, I direct the reader to the appropriate section (or file) with a parenthetical statement. I've attached a new version of the DS.CAT file to my review comments. Confidence Level Note The section that describes the EKs states that the "Sequence EK" was generated from the PEF (not as-run PEF) and then goes on to explain why a full as-run PEF was not available. Fine. However, the main data set description section (under the Data Types heading) describes the sequence as being generated from the ISOE, which is a different sequence product. These two sections should be self-consistent. Despite my preference for the way you've done things (properly), the PDS way to tie the data set to multiple instrument hosts is to repeat the DATA_SET_HOST object, once for each host (DIF, DII). The data set should have some references. You cite the IAU 2000 report as the source of information for the PCK file so at a minimum you should add this as a reference. I would also like to see the "DI Time Issues Working Group" report included with the archive. If it can be added, it should be referenced. Lastly, NAIF should consider getting JPL document numbers for the SPICE toolkit required reading files. This would allow that material to be referenced as well. It might also be appropriate to cite the SSR spacecraft paper. SPICE_INST.CAT ============== Again, despite my preferences, the PDS way to tie the instrument to multiple instrument hosts is to repeat the INSTRUMENT object, once for each host (DIF, DII). ======================================================================== INDEX ======================================================================== The INDEX.CAT, INDEX.LBL, and INDXINFO.TXT file all are acceptable. ======================================================================== DOCUMENT ======================================================================== The DOCINFO.TXT, LBLINFO.TXT, and ONLABELS.TXT files are acceptable. Given that there were so many timing issues related to this mission, and the SPICE products in particular, it would be nice if this archive included the "DI Time Issues Working Group" report as a document in this directory. ======================================================================== SOFTWARE ======================================================================== The SOFTINFO.TXT file is acceptable. ======================================================================== DATA ======================================================================== General comments that apply to labels in all subdirectories What does PRODUCT_VERSION_TYPE = ACTUAL mean? The PDSDD is vague as to the meaning of the keyword and there are no standard values. All of the kernel labels examined (except the gimbal CK based on commands rather than observations) have the value ACTUAL. At the minimum, the value should be PREDICT for some of the SPKs and (*tempel1*, *preenc174*, etc.). The keyword could be useful to users if the values were more informative. For example: FINAL IMAGE NAVIGATED RECONSTRUCTION PRELIMINARY ACS RECONSTRUCTION Currently the PDSDD value length is set to short (20) to create values as described above, however, the value length can easily be changed in the PDSDD. INSTRUMENT_NAME = "HIGH GAIN ANTENNA" or "N/A" etc. The instrument name associated with this data set is "SPICE KERNELS". While it's understood that the point is to make clear that a particular CK file describes the HGA attitude, it's not clear that this is the appropriate keyword for the job. I think that you should propose a new keyword NAIF_INSTRUMENT_NAME, much the way that you have NAIF_INSTRUMENT_ID and NAIF_DATA_SET_ID keywords to store NAIF specific metadata. Alternatively, the REFERENCE_OBJECT_NAME keyword appears to be designed for this application. When RECORD_TYPE = FIXED_LENGTH, as it does in the binary kernel labels, FILE_RECORDS is a required keyword. It is missing from all of these labels. ======================================================================== DATA - CK ======================================================================== The CKINFO.TXT file is great! Kernels: dif_sc_050112_050809.bc dif_sc_050225_highrate.bc dif_sc_050704_highrate.bc dif_hga_050117_050801_vch.bc dif_hga_050118_050801_bch.bc dii_sc_050112_050703.bc dii_sc_050703_050704.bc are all acceptable. All load normally into SPICE toolkit applications and generate data that appear plausible. Labels are valid with the exceptions noted above. ======================================================================== DATA - EK ======================================================================== The EKINFO.TXT file is great! I was able to eventually able to get INSPEKT to load and display some data from both of the kernels, however the experience was painful. There is a lot of work that could be done to improve the help and documentation of the INSPEKT program. While the E-kernel provides the ability to code applications that use information in the PEF, the INSPEKT program is not user friendly. My feeling is that most users of the events data would prefer to the see the PEF itself. Can this file also be included in the archive? Both E-kernel files appear to be valid. Without the original PEF files or Command Dictionary, the content cannot be validated any further. Labels are valid with the exceptions noted above. ======================================================================== DATA - FK ======================================================================== The FKINFO.TXT file is acceptable. di_v16.tf The kernel file loads correctly into SPICE applications. However, I'm a bit concerned about it's content. The only frame that is not defined to be "perfectly aligned" with the spacecraft frame is the DIF_MRI. All of other frames (DIF_HRI_OPTICS, DIF_HGA, etc.) are given as rotations in integer degrees (to several decimal places) about the s/c frame. The DIF_MRI frame is a misaligned by a few 1/100ths of a degree, which is as good as can be legitimately expected. It appears as though the exact mounting angles of the various subsystems were never determined, or not captured in this kernel. Why not? The label is valid with the exceptions noted above. ======================================================================== DATA - IK ======================================================================== The IKINFO.TXT file is acceptable. dif_hri_v10.ti The kernel file loads correctly into SPICE applications. The file does not include a optical distortion model for the IR subsystem. Is this model currently being developed or is there no planned effort in this area? The comments imbedded in this file should mention if there is no plan to develop a model. If a model is being developed, then the user should be directed to look for a new version of this file internally, in the IKINFO.TXT file, and in the DS.CAT file. dif_mri_v10.ti is acceptable The kernel file loads correctly into SPICE applications and appears to be complete. dii_its_v10.ti is acceptable The kernel file loads correctly into SPICE applications and appears to be complete. The labels are valid with the exceptions noted above. ======================================================================== DATA - LSK ======================================================================== All files are acceptable. NAIF0008.TLS loads correctly into SPICE applications and contains records for all past leap seconds. The label is valid with the exceptions noted above. ======================================================================== DATA - PCK ======================================================================== All files are acceptable. Both PCK00008.TPC and DI_TEMPEL1_V01 load correctly into SPICE applications. The labels are valid with the exceptions noted above. ======================================================================== DATA - SCLK ======================================================================== All files are acceptable (SCLKINFO.TXT is great!). Kernels: dif_sclkscet_00015.tsc, dif_sclkscet_00015_science.tsc, dif_sclkscet_00010_scu_b.tsc, dii_sclkscet_00008.tsc, dii_sclkscet_00008_science all load correctly into SPICE applications. The labels are valid with the exceptions noted above. ======================================================================== DATA - SPK ======================================================================== All files are acceptable (SPKINFO.TXT is great!). Kernels: dif_sclkscet_00015.tsc, dif_sclkscet_00015_science.tsc, dif_sclkscet_00010_scu_b.tsc, dii_sclkscet_00008.tsc, dii_sclkscet_00008_science all load correctly into SPICE applications and generate trajectory data that are consistent with published results. The labels are valid with the exceptions noted above. ______________________________________________________________________________ ______________________________________________________________________________ Dan Scheeres's Presentation --------------------------- D.J. Scheeres The University of Michigan Ann Arbor, MI 48109-2140 scheeres@umich.edu Overview: - Database of all relevant geometric and dynamic variables and events for the spacecraft in question. - Database includes the following ÒkernelsÓ: - Attitude of spacecraft - Special events on the spacecraft - Proper coordinate frame definitions - Instrument mounting geometry on the bus - All leap seconds that occurred during the time spans - Planetary constants file - S/C to ephemeris time mapping - Spacecraft and planetary ephemeris Review of Spice Kernels: - Both datasets have identical format and structures. - Cursory checks show that the data content between the two datasets relate to the proper events - Mission descriptions for both appear adequate - Additionally, complete descriptions of the Spice Kernels are given, along with directions that point to the S/W necessary to use them Recommendations/Requests: - There are no concerns or issues with the data as delivered.