Rosetta Review, Tuesday 2016-02-16 Steve Joy Cesare Grava Ottaviano Reusch Tony Farnham Ludmilla Kolokolova Anne Raugh (recording) Tilden Barnes Mike Kelley Mike A'Hearn Emily Law Bryan Butler (joined ~10:50am) Sascha Kempf (joined ~1:15pm) Heather Franz (joined ~1:25pm) Lori Feaga (?) Joint Review with ESA Alice ===== 2 major RIDs will be addressed by improved documentation. Eric Quemirais' presentation... o There are no predicted line shapes or PSF computations included in the documentation. Because of the oddly shaped slit, these would be very useful. If possible, please add something to the documentation to povide some guidance here. o The "chameleon" noted in several spectra should be added to the documentation. OSIRIS ====== Alan Fitzsimmons presentation... o There is software in the EXTRAS/ directory that needs additional documentation to be usable. The team wil likely delete this software, though, as they are phasing it out in operations anyway (and presumably will not be supporting the software included as archive extras). o [Mike A'Hearn asked if, as the spacecraft approaches the sun, whether the temperature dependent calibration constants will be supplied with those data sets,as they are not in this one. OSIRIS responds "yes".] o Dark frames (RID 103) are no longer required because that information is rolled into the calibration data. CIVA ==== Colin Snodgrass' presentation... o A team voice notes there is a mistake in the EAICD for CIVA - there is no SDL data. But Colin indicates there should be something. Apparently there has been some discussion at PSA about including data that has no recognizable science use. It seems like a waste of significant effort to the team. Anne R. suggested we treat this as safed data. This will need to be worked offline to see what we can get with a minimal effort approach. (SBN may need to pick up some slack here.) VIRTIS ====== Frederic Schmidt... o Need a solar flux spectra in order to estimate reflectance. This will be provided in the calibration directory. o Units for wavelength and FWHM in the labels are wrong - this is a known issue already scheduled for correction. o There is no geo-index (with footprints for the spectra). The team says this requirement was dropped by PSA. Also, the team notes there is some geometry information included in the index in the form of lat/long extrema. Returning to Mike DiSanti's RIDs... o Mike A'Hearn took notes on the RID spreadsheet regarding particular RIDs. ROLIS ===== Laetitia le Deit's presentation... o No major RIDs. NAVCAM ====== Thomas Roatsch's presentation, but presented by someone else (?)... o NAVCAM seems reluctant to define their display direction. This is a major problem from a PDS point of view. (The reviewer does not seem to have a good grasp of how PDS3 software works with the data dictionary, which adds to the confusion.) Commenting UMD RIDS... o Documentation is being withheld as it is considered proprietary. o These are level 2 data, so they are considered "raw" and no calibration is provided. o Additional documentation, for example on sensitivity, is still being created for inclusion in the level 3 data set. o Suggested moving one copy of the data - either FITS or IMG - to the EXTRAS/ directory to avoid the duplication in the archive. The team would likely prefer to move FITS files to EXTRAS/. The team seems annoyed - but it seems to me (Raugh) this has come up before when Maud asked in a more general sense. I should research this, just in case I was having a stroke... --------------------------------------------------------------------------- UMD Reviews SESAME Data =========== Kevin Walsh's presentations contain notes that should be taken as liens except as noted or modified below. o Some directories indicated in the AAREADME structure image do not exist in the present data sets. Probably the AAREADME should note that not all directories should be present and why (modes not used, for example). o Although the calibration description is only in the DOCUMENT/ directory document, it does seem to be complete. It would be useful to have a reference to this document in the dataset.cat file. Sascha Kempf's presentation contains notes that should be taken as liens except as noted or modified below. o The point about the name of the SIS/EAICD is a valid one, but it's not clear what sort of controls can actually be applied here. This probably requires the archive (PDS, PSA) to develop a way to make documents visible regardless of designation according to what they actually contain (not easy). Sascha notes that even just a more descriptive, spelled-out name would be helpful. Certification: Not certified. Non-DIMS data needs substantial revision and re-review. COSIMA Data =========== Sascha Kempf's presentation contains notes that should be taken as liens except as noted or modified below. o The attached labels really are very problematic for these data for users who are not relying on the PDS3 OAL interface (and I can't guarantee that works perfectly). o The descriptions in the *.FMT files seem to be saying that the interpretation of what values represent varies from record to record based on (poorly identified and explained) flag fields. o It is extremely difficult to identify spectra of interest. o Displayed spectra all seem to look the same, and there's no explantion of this. NOTE: This is supposed to be a level 3 (calibrated) data set, but both reviewers interpreted it as a level 2 (uncalibrated) data set. What is the actual level of this data set? Please state it very, very clearly in the data set description. Eberhard Grun's presentation contains notes that should be taken as liens except as noted or modified below. o The note about Collector Aperture Geometry seems to be putting too great a burden on the dataset.cat description. This level of detail should be in a more complete document, but the dataset.cat should reference it if it is going to mention it at all. Certification: Not certified. Requires full review. GIADA Data ========== Sascha Kempf's presentation contains notes that should be taken as (minor) liens except as noted or modified below. o The name of the Interface Control Document is probably beyond the control of anyone to dictate at this point. Certification: Certified. ROSINA Data =========== Heather Franz provided RIDs in the standard spreadsheet format. Additional notes: o The Cheops reference frame is described by a document in the Shape model data set and in a published paper. One or the other or both of these should be mentioned in the data set documentation. o The BG/ and NG/ directories seem to contain extreme numbers of files (141000+ in the case examined). This is several orders of magnitude more files than other directories at the same level (the RG/ directory in the same data set has 258 entries, for example). It it is not possible to subdivide these directories, then please include a high-level and very visible warning to users that these directories contain that many files, so that they will realize the sudden lack of response is because of data volume, not connection or machine failure. Certification: Certified. Archive pending lien resolution. PTOLEMY Data ============ Heather Franz provided RIDs in the standard spreadsheet format. Certification: Not certified for want of documentation. Delta review sufficient for certification/archiving. COSAC Data ========== Heather Franz provided RIDs in the standard spreadsheet format. Certification: Not certified for want of documentation. Delta review sufficient for certification/archiving. MUPUS Data ========== Kevin Walsh's presentation contains notes that should be taken as liens. (He may have written up RIDs on these already.) The European reviewer provided a list of RIDs, many of which he marked as "minor" that PDS woul tend to view as "major" (lack of units, e.g.). Certification: Not certified. SD2 Data ======== Kevin Walsh's presentation... o Apparently the level 2 (uncalibrated) data are in binary files that have no structural information in the labels. This alone is cause for PDS rejection. The European reviewer listed mainly minor RIDs, except: o The UTC time tag is apparently repeated in groups of 4 rows, with different values in the other columns. Certification: Not certified. MIDAS Data ========== Andrew Steele's presentation contains issues that should be taken as liens except as noted or modified below. o The BCR-STM format used is a proprietary commercial format. PDS viewers cannot reproduce the image as displayed by the proprietary software. It is not clear why. This must be investigated and resolved. The data cannot be archived if it is not accurately described by the PDS label alone. [Note for myself: A reference point for what values should actually be physically stored in the file would solve this quickly, as would the same for scaled/offset modified values - as read in by software, i.e.] The European reviewer reported numerous minor RIDs. Certification: Not certified.