CASU meeting Wednesday March 23rd 11:30-13:00 CASU meeting area. Present: MJI, AKY, EGS, JPE, JRL, MR, NAW, SC, STH Apologies: RGM, JCR Agenda: 1. Actions from last meeting 2. Comments on WFAU minutes 3. Hardware update 4. Meetings and telecons 5. Data archives update 6. Optical/NIR processing 7. WFCAM update 8. VISTA update 9. AOB Minutes ======= 1. Actions from the last meeting -------------------------------- AKY - has now obtained a deep stacked UltraVISTA image from HJM, comparison with CASU equivalent ongoing <<<< STH - to put the i-band vs 2MASS colour relations on to our web pages <<<< JPE commented that he has implemented them in the SADTool. AKY - has been investigating flatfield behaviour with time and will turn this into a report with additional input from MJI. Issues for consideration are: do bad pixels change over time ? how do the detector-to-detector offsets vary ? is there any evidence for changing spatially dependent features ? how do the means, medians, noise properties, of the pixel values change <<<< EGS - completed the v1.0-v1.1 migration for data up until the end of February AKY and flagged it as ready to go MJI - computed the remaining monthly detector-level zeropoint offsets and illumination corrections for VISTA. He analysed the run of zeropoint variations from October 2009 to December 2010 and comented that Z and Y seem to be more variable (at the few percent level), while JHKs effects seem to be pretty small. EGS has made the illumination correction information available through the web pages. EGS - checked the EVALSO test file internet transfers from ESO Garching to AKY Cambridge and found that for the month of data under consideration a total of 8 files were missing from the internet transferred data where there was overlap with the disk transfers. MJI - still needs to put the examples and more information about the MJD variation across a tile on the web <<<< MJI - also still needs to add the note about QC information derived from tiles cf. to the more detailed QC information available from the pawprints <<<< JRL - made a speedy recovery and dusted off his {} brackets 2. Comments on WFAU Minutes --------------------------- No comments 3. Hardware Update ------------------ The Supermicro 45-bay JBOD expansion module is still sat on the floor though its host does at least have the updated 3ware card installed. Malheureusement the kernel now needs some TLC to get it to communicate with the card. <<<< The new fibre-channel RAID array has been installed and initialised (more than once) it nows needs adding to one of the fibre channel loops. Though not quite hardware the IoA web pages have been drupal'd and for once everyone agreed are actually an improvement. Fortunately this had less than minimal impact on our standalone CASU web pages. AKY was asked to fix the *&^%$* script which kept filling up the homespace as a stuffed homespace is an unhappy homespace. <<<< 4. Meetings and Telecons ------------------------ MJI noted that the next VISTA IoT is scheduled for April 14th, roughly a week after the anticipated restart of VISTA observing after the aluminising of M1 and M2. MJI reported that there was nothing much to report from the VISTA VMC Skypecon held on 24th February. Interestingly they didn't know about the VISTA recoating, nor indeed did any of the other survey PIs. MJI will be attending the NAM in Llandudno to talk about M31 surveys, and also reminded us about the upcoming Herts VVV/VPHAS/EGAPS meeting in July. 5. Data archives update ------------------------ All still good. 6. Optical/NIR processing -------------------------- Some INT WFC data from Sophia Feltzing via, Gerry Gilmore, is now in the to- process queue. The Kepler region observations will be later June (interestingly Panstarrs-I will be releasing observations of the Kepler region shortly). We also heard news that the VST test camera is on and delivering Image Quality of around 0.6 arcseconds. EGE has been asked by one of the GTO teams about processing VST SNe survey data. MJI has been talking to ESO and KK about processing all of the VST data, to ensure better calibration. He also noted the VPHAS+ Halpha filter is still in Garching and will shortly be repeat- tested to check for degradation in performance. 7. WFCAM update ---------------- The weather continues in its recalcitrant vein. MR noted a problem with missing headers items on detector #2 for a small subset of data taken on 16th,17th February. This showed up initially as a WCS problem. MR was not yet sure if this propagated to later nights (pending more processing). JAC noted that the problem correlates with frames that have a rather large -ve pixel value (~ -10**34) present and probably causes the grumpy SDF -> FITS conversion to grump. As its only one of the SHDUs it was easy enough to fix and MR will add a few more checks to the assorted checks already carried out at ingest and will note and fix any obvious missing header items on-the-fly. Apart from this minor hiccup, processing, checking and release is up to date. WFCAM data to end of February was flagged as ok to go to WFAU and duly slurped its way North. As ever the usual multiple queries about early access to the processed data were received from vigilant PIs and dealt with. The resubmission of the extended UKIDSS survey to the UKIRT Board was discussed. If successful this together with other WFCAM PI programmes would require the usual CASU processing. Still todo: some photometric recalibration of GPS in regions of heavy extinction; and the long overdue UDS 05B-07B H-band reprocessing. 8. VISTA update ---------------- Ingestion and processing all done apart from the first 3 days of March. V1.1 products are released to end of January. The Febuary data is processed and awaits detector zeropoint offsets et al. A few of the per-directory <<<< Semaphore flag files were missing at one point but have been fixed. PIs transfer most of the data over the internet but for VVV we are making a copy of the catalogues available on disk and ferrying them via ESO. The usual discussion about the network transfer of files between Garching and here ensued. EGS commented that the median transfer speed using 8 threads is around 12 MB/s and that a night of data typically takes 1-2 hours to transfer. MJI asked if we ever heard back from Matt Jarvis re: his narrow-band 975/985 data - no ! MJI asked about the ESO OB status and grade parameters. V1.1 now has the revised ESO values in the headers, in particular the OB overall grade now allows for interrupted and restarted OBs. There was also the usual discussion about ESO PHASE III delivery. Apparently each Institution is limited to 8 threads for uploads to ESO. Since CASU are delivering the majority of the bulk data this limit is a potential bottleneck. MJI and EGS will investigate whether this limit is negotiable. <<<< EGS downloaded a month of VHS data as a test and broke their digestive system for the best part of a week. The asymmetric window centred on 1st April is looking ever more skewed. The main deliverables are tiles and tile catalogues and VHS will also deliver pawprint images + catalogues. MJI commented that for both science use and provenance issues all surveys really need to deliver the equivalent pawprint products too and wondered how long it would take the EDP group to figure this out. JRL was asked if he ever replied to a software request to trigger on N images rather than an OB when the OB takes a long time to execute, and therefore the QC info lags behind at the telescope. His succint reply included the words "not my problem" and politely (sic!) noted whose problem it was. MJI was asked by Richard Hook about yet more colourful pictures, as were the survey PIs. If anyone has any favourites let MJI know, otherwise we'll be going with the M15 and Carina suggestions already offered. RGM has announced the formation of the VHSUCDSWG - honest ! STH, EGS and JPE will (and did) sit (sat) in the afternoon, and are they any the wiser is the key question they can answer next time. MJI noted the first of the potential V1.1->V1.2 upgrades. The illumination correction tables could probably do with an extra x-axis column at either end (thanks to Eckhard for pointing this out). MJI will investigate and make it so if needed. <<<< The images reported in the web page is a wee bit off for the night of 20101225 "It" says 25, but its more like 600 odd. Can somebody fix it ? <<<< JRL offered a modest bet that we still get at least one set of empty disks from ESO before VISTA starts up again. STH naively took him up on this. ESO - are you listening - he'll split it the profit with you. MJI reminded JRL about adding photometric zeropoint and calibration comments to the fits headers using the ELIXIR (CFH) example as a template. Mind you he still reckoned nobody would still RTFHDU and we would still be asked what does it all mean, how much does it cost and where can I get one. 9. CASUHELPs ------------- No time for a savoury selection this month. There was a brief discussion about CASUTOOLS and also the VISTA CPL pipelines which doesn't build on Darwin (Mac OSX) due to Darwin's lack of evolution of a sensible C complier. AOB --- Nothing of note. Continuing Actions: ------------------- AKY - compare Henry's deep stack with our equivalent STH - put 2MASS JHKs -> i-band predictor information on web pages AKY - continue flat-field properties investigations and report at next meeting MJI - add MJD tile catalogue info to web pages and example pic MJI - update QC/grade note on web pages about pawprint QC -v- tile QC New Actions: ------------ SC - apply TLC kernel TLC to enable it to communicate with 9750 3ware card AKY - fix (or destroy) home space killing script MJI - generate the February detector zpt offsets plus illum corrs EGS - investigate whether 8 thread limit is negotiable MJI MJI - see if fatter tables needed ??? - fix the 20101225 # of images faux pas