Teleconferences are currently every Monday at 11 AM EDT.
1. Dial Toll-Free Number: 866-740-1260 (U.S. & Canada)
2. International participants dial:
Toll Number: 303-248-0285
Or International Toll-Free Number: http://www.readytalk.com/intl
3. Enter your 7-digit access code, followed by “#”
Your access code: 7691719
Monday, 1 July 2013¶
- Basilio Santiago - exposure grading with galaxy counts
- Klaus Honscheid - pointing accuracy
- Aaron Roodman - IQ status
Monday, 10 June 2013¶
- Eric Huff - adding artificial galaxies to SV images
- Lucas Beaufore - updated IQ/telemetry catalogs
- Agenda / prep for Instrumental Signatures Workshop 17-18 June.
- Klaus - status of work on pointing errors and survey table feedback.
Monday, 20 May 2013¶
- General updates/status
Monday, 25 Mar 2013¶
- Progress on photometric calibrations (D. Tucker)
- Progress on DECal (W. Wester)
- Any issues arising from Directors' Review
Monday, 11 Mar 2013¶
- Left from last week: PSF size dependence on counts (Gary, Huan?)
- Image quality issues:
- News from the TCS analyses (Kevin)
- AOS/donut update (Aaron/Kevin)
- IH outputs update (Ken)
- Any other IQ items
- Docs for upcoming reviews (Gary)
Monday, 25 Feb 2013¶
- Review of observations, any pending observational tasks
- Discussion of approaches to completing and closing out Science Verification. What transitions to Acceptance Testing, what goes to DESDM, what needs some other structure? Here are general areas GB considers needing continuing effort:
- Image quality (including IH cross-validation, AOS, comparison with DIMMs, developing environmental control strategies)
- Pointing and astrometry (evaluating pointing performance; updated astrometric solutions & code; WL-level performance)
- Precision photometry: star flats, ubercal, evaluation of standards regimen
- DECal - making it possible to collect and use these data.
- Operations: Obstac upgrades and SN/Main survey switching; deciding staffing levels; review documentation & interfaces for next season, RASICam integration
- Camera nonlinearities (low-level, high-level, crosstalk): characterization and correction
- Hot/cold pixels, edge bleeds
- Flux-dependent PSF - characterize and correct
- DECam regression test suite; engineering-night test set.
- Exposure quality metrics and the Survey Table
- Catalog-level requirements (S/G separation, galaxy photometry, etc.)
Monday, 11 Feb 2013¶
- Paul presents Flat-field_structure. There is full correspondence now between 3 tape bump changes and changes of >=2 K on the focal plane temperature. Does not return to original flat field even when focal plane temperature comes back. No change in total flux, so these are electric-field rearrangements. Couple of percent changes maximum. A discussion follows on how to insure that Paul's very clear conclusions on monitoring of CCD focal plane and on changes to the flat procedures get properly implemented.
- Don will make it possible for Gary & Klaus to enter tickets for DESDM for identified processing steps.
- Tom will make sure that Run Managers and the CTIO staff have a focal-plane temperature check on their daily maintenance checklists.
- Eric can alter Night Summary such that a red warning appears at top if focal plane T changes by 2K. But note there are no Night Summaries on non-DES nights.
- Gary will send message to Ofer for distribution to WG's, telling them to be responsible for addressing problems with their science that might arise from the tape bumps. Don't expect any problems outside of WL (who will mask bumps) and SNe (who will see a few false positives here perhaps).
- Vinu showed initial results on brightness-dependence of PSF - see docdb #6964. A 1 or 2% change in R50 is found that is not an artifact of the way SExtractor assigns fluxes, supporting an earlier conclusion by Nicolas Regnault. A search for dependence of R50 on distance from serial register shows <0.3% change across the full CCD, suggesting the that larger flux-dependent change is not resulting from charge transfer problems.
- Huan points us to results from spot projector in SiDet tests that showed flux dependence in yy moments of spots, apparently attributable to a charge trail.
- _Vinu will look for flux and position dependence of the xx and yy moments of the PSF as well as R50.
- Ken: IH can now calculate R50's and second-moment whiskers. Using 3" radius apertures to define whiskers and total flux; typically 400 stars measured per exposure. Robust mean R50's and whiskers for full focal plane being calculated. Will install before next run: need to alter the database to accommodate the new fields.
- Final 5 half-nights of observing begin Sunday night.
- First 2 nights of Arjun Dey's COSMOS observations look ok (not great seeing, typical of our past weeks). If this keeps up we won't need to look at COSMOS.
- SNe visible only 1st hour of these nights, even SPT-E is below airmass cutoff by end of the half-nights so obstac will go crazy.
- Klaus is collecting the IQ data for the exposures so far, so we know what to re-observe.
- Convene a phone conference Thu or Fri to plan observations for the run, including possibly raising the max allowable airmass for obstac.
- These engineering/diagnostic images known to be desired for our or DD time next run:
- Star flats in a denser field
- On-sky linearity test
- Aaron & Kevin wavefront characterization
- Steve's filter-ghost hunting
- Douglas's eastern standards fields.
Monday, 4 Feb 2013¶
1) Update on progress/plans of the SV observations (Marisa, Phil Rooney, Brian on the mountain):
- TCS problems last night, no observations taken. Resolved.
- The half-chip offsets were completed for all but 2 SN fields. [Huan reports after meeting that pointings for all exposures are ok.]
- Seeing is always starting at ~1.5" and dropping to ~1.0" by 11 PM. Marcelle is looking into the status of mirror cooling [answer obtained later: daytime mirror cooling was only done properly starting Feb 4, according to Alistair]. Note that mirror cooling stops when the mirror covers open, usually done before sunset. Sucker is on when mirror covers are open.
- Marcelle: what to do when seeing is poor but SN fields have set? Consensus is to take SPT-E via obstac and then nearer end of run we will exclude worst exposures and retake them if telescope time is available. Take stock of this during the week off before our final run - topic for next week.
- Some discussion of defining cuts for the Survey Table. Will stick with something simple, hope to have during this week off the telescope a systematic data file giving seeing (and whisker?) for all SPT-E exposures.
2) A discussion of the Task List assembled at Punch_List
- Gary introduced task list and asked for additional tasks to add plus volunteers. I (Gary) will update the list with those who volunteered and was given the action of re-circulating the list to SV before going to the full collaboration seeking more people to take on tasks. Comments on the Tasks below:
- Eric N. notes that implementing automated SN/Main survey switch in obstac has 4 prereqs:
- Figure out mapping between IH seeing measure and whatever our desired cut will be.
- Model and/or fix dec dependence of seeing (we expect this to be gone by September!)
- Model wavelength dependence of PSF size
- Need to have pointing good enough that SN observing does not require manual repointing.
- It is believed that all of the above are either outside SV scope or already described in a task.
- It was suggested that Regression Test scripts would work better with HA/Dec or Alt/Az pointings in scripts, but these shut off tracking when sent to TCS. So Klaus will implement a SISPI option in which scripts specify HA or Alt/Az, but RA/Dec are sent to TCS after computing them on the fly at the time of exposure, so that tracking is still turned on.
- Tom: add a task to track down the "filter edge ghost." Also some other ops things that we might leave in the Ops domain instead of SV.
- Need to add a task to keep refining the AOS loop.
Monday, 28 January 2013¶
- Observing protocols for February runs, incl. SNe
- Results from WL analyses (Peter)
Monday, 21 January 2013¶
- Analysis updates:
- AOS (Aaron)
- Nonlinearities (Huan)
Monday, 14 January 2013¶
- Observing plan through end of the season: SN cadence, SPT-E field selection, AOS or other experiments
- Planning for image quality monitoring and improvement during the off-season
- To-do list of analyses to pursue before April DES meeting:
- Star flats
- Astrometric solution
- Understanding / controlling IQ
- Setting and simulating obstac strategy to get desired i-band depth
- Crosstalk nonlinearity
- Final cut and coadd outputs!
- Stability of zeropoints and appropriate strategy for standards
- ubercal results
- Analysis updates:
- AOS (Aaron)
- Depth (Eli)
- Nonlinearities (Huan)
Thursday, 27 December¶
- Plans for observing run that starts tomorrow: crate1 readout is shutting itself down, do we need to consider what observing can be done if it is still out Friday night? Since recovery is considered likely, we will think about possibilities and table until tomorrow. Thoughts of doing standards all the way through twilight to find a appropriate start time; and possibly Kevin & Aaron could use the time to compare science array donuts to F&A donuts. Tom says this could be the fan interlock issue he highlighted at TAMU. At 4 PM EST Klaus tells me that this has been fixed.
- Image quality data for December.
- Approx 1400 image id's are on Klaus's useful list. About 700 or 800 of these have First Cut at NCSA. Of those missing, perhaps 400 are community data, the rest are First Cut failures for some reason or other. Do not at this time know that the failures are biased in any particular way re image quality. PSM may be failing, but Bob says that First Cut catalogs should be produced anyway.
- Another 300 images from the last 3 nights before xmas shutdown are now being analysed by Jiangang and Eric. Klaus will collate these into another release of the IQ catalog.
- Aaron's donut quantities, plus the dome & mirror temperatures that he showed indicate strong evidence of mirror seeing, should be added also.
- Klaus requests that we examine the file format in current version to see if any changes are needed to the format before he cuts another version. Note -999 indicates lack of data in a field.
- Jiangang's plots show that image size (r50) and length of avg whisker are reasonably reproducible from different methods/people (still could use some examination). But both Mike & Jiangang see large RMS after removal of mean whisker, which is not expected and might be due to noise. Mike, Jiangang, & Gary to convene at UPenn next week to think about this, but meanwhile the image size and mean whisker measures are good enough for evaluation of the IQ to proceed.
- We will plan a call with the new observers at 4 PM CTIO call tomorrow. Some critical people may be busy w/family at this time, so CTIO may need to speak with them later if there are SISPI issues etc.
Monday, 17 December¶
- Diagnostic observations to be done this week: first time good conditions are available, 2 experiments to be done: first, turn on donut control of all 5 hexapod DOF to see that system stabilizes. Then, careful donut / PSF observations in a part of the sky where BCAM and donut LUT's disagree (alt=45, az = 200-230). (Aaron, Kevin, Tom @telescope)
- Action item list from TAMU conference: Discussion & volunteers
- Image quality diagnostics logging for December / January: All images after guider fixups of Dec 6 will be examined. Let's say >=30s exposure times only.
- Create vector mean ellipticity/whisker and then RMS of deviations from mean. Jiangang will do this using First Cut outputs where possible.
- New guider stats sensitive to jumps, oscillations. Jim will create such statistics from the guider logs. Email thread already exists on how to extract jump statistics in addition to stats already being put into the database.
- Determine ellipticity/whisker stats expected for well-tracked, nearly-aligned images. Jiangang, Steve to do this. Note we need to choose well-defined measures.
- Log for each exposure: HA, Dec, exptime, filter, R50, mean vector ellipticity, RMS variation, guider stats. Robert, Aaron, Eric, Klaus, Jim, Gary to converge on exactly what statistics need to be merged and most effective way to pull them together into a single table of some ASCII variety. Gary & Klaus will talk with Alistair about what information from the TCS telemetry should flow into this or just info to flow out to CTIO.
- Track progress of exposure quality through the month. Multiple people will start to dissect this table once it is assembled. Note we have 2 purposes: (1) Have clear measures of image quality (rate of jumps, seeing, as function of filter & dec) to use for gauging whether Survey should begin in January, and (2) diagnostic info to improve telescope performance.
- Check "interpolability" of whisker against DES Requirement R-20. Mike Jarvis will do this.
- Finish reconciliation of IH/QR stats (not discussed)
- QA and Survey Table loop:
- Help Robert set thresholds on the relevant quantities for Survey Table & related alarm levels in IH/QR, and also need to cull observations since Dec 6 to decide what should be tossed out of current Survey Table. Jim volunteered for last of these tasks, agreed that other work has lower priority than IQ tasks below.
- Need human examination of DESDM outputs for images with ghosts & scattered light: do we need to predict and exclude chips with bright ghosts? Marisa might want to do this but was not on the call.
- Image depth analyses. Eli is doing this and will draw on Eric Nielsen, Tom, Paul, Huan, Jim, Gary who have looked at sky brightness models and depth forecasts in the past.
- Star Flat Derby: Douglas is organizing efforts to make star flats with different codes and intercompare. Bob has made catalogs from one set of dithered images that we will start with. Likely to have teleconference this week with aim of getting results together (maybe face-to-face) mid-January. Expect Douglas, Gary, Vinu, Bob, Jim, Anne Bauer, Nicolas Regnault to be involved. Douglas will also report soon on repeatability of the zeropoints at 0.01 mag level from chip to chip and night-to-night. We need to consider how to test our star flats.
- Image quality diagnostics logging for December / January: All images after guider fixups of Dec 6 will be examined. Let's say >=30s exposure times only.
- Next Monday is Christmas Eve, next one is New Years' Eve, so no phone conferences. But we will need to report at end of month about the image quality. So we will probably want to schedule a call between xmas and New Year's. This week, the various subgroups listed above will have to be communicating by des-sv and phone calls.
Monday, 5 November¶
- Comments/suggestions on SV Observing plans
- Comments/issues with data analysis so far
- Update on photometry & flats results (Paul, Douglas)
- Update on anomalies / eyeball observations (Daniel)
- Other analyses of SV data so far
Monday, 29 October¶
Online: Klaus, Daniel, Jiangang, Andres, Eli, Paul, Bob, Marcelle, Ricardo, Angelo, Tom, Steve, Nacho, + ???
Klaus: there is now WIki page for SVOps, links for observers, tasks, and plans for each night. Send any comments upon looking at this.
Will have 4 PM CTIO time meeting each day, chaired by run manager (see who it is on WIki). [This is 3 PM EDT this week then 2 PM EST starting Sunday when daylight savings ends in US.] Each day, eyeball squad results will be discussed, plans for the following evening's observations made. Will want to have a specific plan for photometric or non-photometric night. Daily email reminders will be sent out, use this SV number. First meeting is Wednesday 31 Oct (day before SV starts) to get into routine. Marcelle & Darren will be there and able to report on open issues needing resolution before Thursday.
Data processing plan: still to use DESDM, more optimistic as of Friday that they will be able to do it. Will get master calibration frames and first-cut: not sure yet what turnaround time will be, maybe not next day esp at start. Raw images will be available however.
We are “nervously ready.”
Gary: Observations for SV. The Datasets page contains links to instructions for nearly all the observations that will be taken during SV. Bob has now posted a dither sequence to use for determination of astrometric and photometric internal solutions. Still need someone to post coords / instructions for the tests of completeness etc. under varying seeing conditions.
Paul will keep track of conditions under which sky brightness has been measured; will use daily phone conferences to tell us which new circumstances should be observed on a given night.
Klaus: need to zeropoint the TCS each evening, takes ~5 min. Do this at first standards field? There is a Kentool which figures out offset for a bright star. DS9 can download a catalog and overlap on image. Gary suggests that at least on SN fields, for which pointings must be adjusted to be accurate, that we can identify a standard star (and its desired pixel position) on the N4 CCD instead of using super-bright stars for Kentools.
Eyeball squad: we will want to record several pieces of info for every sky exposure during SV. See Eyeball examination page for proposed list. Ricardo reports that QR should be able to give PSF FWHM and sky brightness by analyzing one (or few) CCDs of every exposure. We also want to log pointing errors; if DESDM first cut is running, the true pointing will be derived automatically, but otherwise Huan points out the DS9 overlay of USNO stars should make this easy to do by eye. We also want all images examined for anomalies (guiding failures, excess-noise regions, etc.). Daniel Gruen + 2 students at Sussex have volunteered. Will discuss with them details; add your advice to the above Wiki page.
Astrometry: Gary notes that astrometric sol'ns with 20-30 mas RMS residuals have been derived. But during SV, will want to do this carefully in each filter, 2x to check for stability.
Pointing/guiding: Alistair emailed a summary of current status. Ken notes that the 3 guiding goals are not too affected by they ~arcmin pointing errors. 2 are ready to go, need some code to write for 3rd. But do not have guider info for 1st 20s, so cannot check settling of telescope. Klaus: Should try to get access to the TCS datastream (via Mike Werner) to check settling. Current recommendation is 10s after slew stops (others want 5 sec). Also need to figure out the 1” jumps.
Relative photometry: use std star fields to check on zpoints and stability; done only crudely during Commissioning so far. Paul will get zpoints and color terms from stds each photometric night.
Jim has run ubercal on S82 data; need to do more of this kind of thing to understand the flats.
A/B problem is right in our range of sky brightness for g & r bands. Esp. for std-star exposures. Add some differing exposure time std fields during early SV to check for linearity at low counts?
Absolute photometry - no news; Douglas on line, no one has heard of any results yet from extensive standard-field exposures taken at end of commissioning.
Characterization of IQ: Jiangang has looked at a few images in good focus. Marginally meet specification of whisker length. Have seen QR/IH FWHM’s track DIMM for one night, but we need to track this through all SV and test that the DECam/DIMM offset is acceptable. *Need a phone conference of IQ people, including Dara, to decide whether we need frequent focus frames, and how we will decide whether IQ is sufficient for survey.
Question: after SV starts, do we need Monday calls? Yes, maybe even more of them to discuss analyses, since daily calls w/CTIO will be more operational and not cover analyses.
Monday 15 October 11 AM EDT¶
- DC7: no final cut or coadds produced yet and will not be for at least another week. Preparations requiring profile fitting and coadds will need to proceed with DC6.
- Coordinate week: would like to collect on the wiki coordinate lists for all SV observations, especially to compile the mini-survey, SN, and cluster pointings before Peter disappears later this week.
- SV Calibration regimen: plan a separate discussion of what calibration data to take on a regular basis during SV.
- Analysis software updates:
- Marcelle (galaxy photometry)
- Eli (more galaxy photometry)
- Nacho (masks, s/g separation)
- Huan (photo-z)
Monday 8 October
Status reports on testing suites: Paul (signals/noise), Douglas (photometric calibrations), Steve/Jiangang (PSFs/IQ), Ken (pointing/settling), Bob/Gary (astrometric calibration)
Next meeting: Monday 1 October 11 AM EDT¶
- GB: SV is now more confidently schedule for nights of 1 Nov through 24 Nov. Then engineering, then Survey operations to begin 1 Dec. Gary will contact all those to confirm the schedule for observers during Nov & early Dec.
- Peter described the plan for "normal" observing targets, namely a 3-tier approach to the "Mini-survey". See MiniSurvey page for details. Basically, highest priority for the fields covering VVDS field (and one of the SN deep fields within it); then cover about 2/3 of our target; then expand to cover rest of desired area if time/weather permit.
- Peter hard to reach after 10/16, so best if Jim Annis & Eric Nielsen can get him final pointing scheme for main survey before then so we can finalize coords. Klaus suggests entering highest-priority pointings onto SISIP queue by hand instead of inventing another OBSTAC tactician to do so.
- Klaus notes currently 2 meanings of where the telescope is pointing. All hope that by November, the telescope coords will refer to axis point rather than fiducial pixel in middle of N4 chip that is 4' away.
- Image processing update: standalone pipeline taking ~40 minutes per exposure (single core) through cataloging. Running at CTIO, FNAL, OSU, UVa, limited runs at Penn so far. Because this will be a long time to process 200 images from a night, we will need to coordinate standalone reductions on the SV data that will not go through DESDM. Huan reports there is now a scriptable way to process multiple exposures.
- DC7 data not officially released for analysis yet but Huan says you can find some in the DTS directory above where the real DECAM images are being put. Look des/red directory (sorry Huan, I missed the details).
- Reminder: open ftp access to DESDM servers will go away soon. Get your data access passwords in line.
- GB wants each requirement/goal group to report on progress in coding their tests over the next 2 teleconferences. Emphasis is not on results with current/DC7 data, but on having the analysis tools ready. Schedule for this below - please plan on giving ~10 minute update on the teleconference, create Wiki page at Commissioning Tests for supporting descriptions.
- 8 Oct: Paul (signals/noise), Douglas (photometric calibrations), Steve/Jiangang (PSFs/IQ), Ken (pointing/settling), Bob/Gary (astrometric calibration)
- 15 Oct: Marcelle, Eli (galaxy photometry), Nacho (star/galaxy, masks), Huan (photo-z's),
- 15 Oct: also examine issue of which flats to take or other operations questions (Klaus); probably have a distinct phone meeting for the flats question.
- Observing Plans: nights, observers (Gary)
- Observing Plan: targets (Peter)
- News from the front (anyone):
- Data transfer and access
- Experience on the mountain (Klaus or Marcelle?)
- Schedule presentations next 2 weeks on trial analyses with real / DC7 data
Next meeting: Monday 24 September 11 AM EDT¶
Progress reports on work with DC-7 and real data. Seeking workers for tasks requested in message below from Alistair.
I hope we can hear from the following (if possible, given the short notice):
- Progress on the mountain (Gary, or anyone there)
- Standalone pipeline (Eric, Huan)
- Flats (Paul)
- Astrometry (Gary)
- Standards (Douglas)
- CR's, bad pixels (Daniel)
- Donuts, focus, alignment (Aaron or Steve or Jiangang)
- Field selections (Peter)
- Star/Galaxy (Nacho)
- Galaxy photometry (Eli, Marcelle?)
Next meeting: Monday 10 September 11 AM EDT¶
Determine extent to which DC7 and other not-the-real-sky data can be used to rehearse all SV tests.
Monday 13 August 11 AM EDT¶
1) Update on field selection from Peter Melchior
2) Update on specialized-observation inventory from Paul Martini
3) Software inventory: getting started & discussion (Gary)
4) Any other updates.
Monday 30 July¶
Online: Daniel, Steve, Eli, Huan, Tom, Brenna, Peter, Marcelle, Andrea, Ken, Eric, Jiangang, Bob, Paul, Nacho, John, Gary, Douglas, Klaus, +???
- Tom and Klaus reported on progress made defining operations procedures and documentation. Marcelle and Tom are on Tololo testing and refining our plans. Text to complete the OPS requirement in the SV wiki is being prepared (Tom)
- Training to become a run manager will take a lot of time. It is time to select/recruit the run managers for year 1 (including SV)
Action item Tom, Brenna, Klaus and other interested parties will meet this week to get this started.
- Some progress has been made on the stand alone (DES DM) software. The complete stack can now be compiled successfully at FNAL and OSU
- Peter reported on the current status of the observing plan. He requests that requirements/goals owners revisit their proposed procedures and additional details or specifics could be added to the description of required observations.
- Paul is working on a spreadsheet for non-obstac observations
- Standard calibrations, ie bias and flats, and special operations requirements, e.g. to test obstac, will be folded in when we map the observing plan to the available nights.
- Douglas reported on Cal-R2, R3 and R4
- The PSM code to do the matching is in hand but is implemented as a stored procedure in the database and hence can only run as part of DES DM
- Paul reported on Cal-R5. It was discussed that we might have to extend the fringing tests to the z-band as some z-band fringing was seen with precam.
- We need to understand where the DIMM instrument is pointing and be ready to make corrections for filter and airmass different from DECam's data.
- Currently it's not working reliably. What is the predicted status for SV?
- Steve and Ken reviewed the pointing requirements
- To test the residual vibrations of the telescope after the slew it was suggested to look into the TCS encoder information for some subset of the exposures.
- To determine whether dead time is consistently within requirements (instead of just checking a few cases), we will need to have some log entries by the observers whenever there are delays caused by operators (e.g. deciding what to do, falling asleep with empty SISPI queue...).
- Daniel presented the proposed procedures for Anom R1 through R6
- Huan suggested that some studies could be performed now using the data taken with the imager at CTIO
- Filter dependencies will be added to R4
- R1 will be refined to take into account the angular dependency of the cosmic ray flux.
- Gary clarified that R6 is a requirement for DES-DM - before we update the exposure table we need to reject exposures with airplane trails and similar artifacts that are severe enough to invalidate an entire exposure.
Monday 23 July¶
Online: Daniel, Steve, Eli, Huan, Tom, Brenna, Peter, Alistair, Ken, Eric, Jiangang, Bob, Paul, Nacho, Tesla, Klaus, John, Douglas, +???
- Last week's action items
- Peter reported on the first meeting of the SV observations group. He is preparing a draft and expects to have this available next week. He asked that all owners of requirements and goals add requests for specific observations to their procedure descriptions.
- Klaus informed the group that if we decide to attach a Fermilab based svn repository for the SV codes to the SV wiki page, each user will have to have a Fermilab computer account and a Kerberos principle. This is different from the fermilab service account required for (SWG) wiki access. Nobody had a problem with this. Klaus will post instructions to the SV wiki.
- Everybody was reminded that we need a detailed list of SV codes very soon
- Update from CTIO
Alistair reported that the crew is making good progress installing the cable trays and hoses on the telescope. A milestone was reached last Friday when the RA tray installation was completed.
Brenna informed us that they are ready to proceed with the imager cool down tomorrow as scheduled. Tom and Ken arrived to work on the filter changer and the northwest platform. Everything is on schedule for the technical First Light on September 28. So far very little of the 3 week schedule contingency has been used.
Check docDB #6583 for daily updates.
- Photo z (Huan)
- The photo-z goals require the observation of the VVDS 2hr field to full depth in all filters under standard survey conditions (good seeing, photometric)
- There would be significant benefits if the field could also be observed with the new u filter.
- The codes exist at DES DM and fermilab (from the dc6 photo-z challenge)
- The VVDS catalog with about 10,000 redshifts will be online
- Goals 2, 3, 4 have been combined
- Galaxies (Eli and Tesla)
- The galaxy tests will use the CFHTLS W1 field as "truth" (~35 sq. deg.). Need to observe this to full depth.
- Catalog is public and being prepared by Tesla
- Will use CFHTLS photo-z's for Gal-G8 and DES photo-z's for Gal-G10
- Action item on catalogs: Tesla, Eli, Marcelle and Huan will discuss how we can keep all catalogs required for SV in a central place for easy access by the SV squad and the collaboration.
- SN (John)
- Require and early observation of all sn fields (6 hours of telescope time)
- 2nd visit about a week later (6 hours of telescope time)
- Use obstac to orchestrate this (dither pattern and such)
- Need to work with DES DM and SISPI on manifest concept
- The group works with DES DM (weekly meetings) on software needs. template code exists at fermilab. sn pipeline in reasonable good shape.
- Next week
- All owners of the remaining goals and requirements agreed to review their procedures at the next phone meeting.
- Be aware that this might be a 2 hour meeting to give us enough time to complete the reviews.
Monday 16 July¶
Online: Bob, Brenna, Gary, Ofer, Tom, Huan, Douglas, Steve, Paul, John, Eli, Klaus, Eric, Ken, Angelo, Nacho, Peter, +???
- Discussion of where to collect codes needed for SV. Two suggestions: svn on the DESDM repository, or svn on the Fermilab Redmine site. Some codes are already in DESDM (e.g. PSM photometry) but Redmine has convenient ties to the wiki. Klaus will investigate and report back next week.
- Need to begin reviewing the observing plan for SV. Willing to work on this: Paul, Peter, Gary - would like to have Jim Annis and Eric Nielsen too (and will consult Douglas on standards). Gary will try to convene a phone conversation Tues or Wed
- Procedure reviews: next week, Huan, John, & Eli will review their goals.
- Milestones / schedules:
- Want all requirements and goals to have a procedure outline by 30 July
- We will need to identify all of the code that must be written to execute SV tests, then start tracking its progress. And identify which ones are needed for Commissioning.
- Working backwards: if 1st light is Sep 28, then should have some running version of relevant codes by 7 Sep, giving 3 weeks to test and correct.
- Aug 15 is nominal date to evaluate the reality of the DESDM and the standalone processing pipelines.
- Review of astrometry procedures:
- For Gary's Ast-R1, SISPI-generated WCS accurate to 1 arcsec, the following were noted:
- Steve: Need to see whether polar misalignment of telescope is large enough to cause trouble at 1" level.
- Huan: DESDM and SISPI tables currently have room for 2x10 PV coefficients per CCD (cubic order correction) but this is not sufficient if PV's must be the same (global polynomial) for all CCDs, since there is a significant 4th-order term. There is no requirement we know of in any code, though, that PV's are same in all CCDs, although there are places where same CRVALs for all CCDs is needed or desirable. Gary will check what the solutions might need, and alert Klaus about any needed extra FITS header elements. Also need to be in touch with NOAO about this.
- John: what's motivation for this requirement? Answer: 1" accuracy makes downstream matching (by other SV codes) much easier. Also non-DES users may very much appreciate being able to locate desired targets by simple cursor/ds9 identification. Tied in also with locating objects on guider CCDS - Gary must contact Eric Nielsen about the relevant codes.
- Bob's Ast-R2,R3:
- Klaus: procedure notes need for full-depth observations. Really needed? Bob: not the depth that's needed, just multiple dis-registered exposures to permit internal solution.s
- Gary notes that multiple dithered exposures needed to generate the astrometric solutions will be taken quickly & also used to generate the star flats through analogous photometric procedure. Then the exposures taken in "normal DES mode" can be used to validate this solution.
- For Gary's Ast-R1, SISPI-generated WCS accurate to 1 arcsec, the following were noted:
Monday 9 July¶
- Discussion of pipeline processing: GB outlined plan that DESDM will do "first cut" and supernova reductions during SV, and will run coadd pipeline on selected data at lower priority. Standalone pipeline will be implemented and used for all non-standard reductions needed for SV instead of creating special cases for DESDM. Don reports that first cut may not do any model fitting. Is it needed for any SV tests? Huan reports that several galaxy photometry goals will require model fitting on single exposures. We will plan on running these with standalone system if DESDM does not implement in first cut.
- Discussion of the signals/noise requirements with Paul.
- Jiangang has some code that will be useful for measuring read noise (and potentially looking for pattern noise).
- Who keeps the nominal response curves for DES elements? Huan has these, will need to transfer to Paul for the S/N tests.
- Eric Nielsen has the model that ObsTac is assuming for sky brightness. Paul will need to speak with him to see what moon angles / phases must be validated to insure that ObsTac has correct info in the intended observing regimes. Will need to keep track of what filters/moon phases/moon distances are observed incidentally, then fill in as needed.
- Mask goals discussion with Nacho.
- Mask code is likely to not be part of DESDM pipeline before SV, so it will have to be run "by hand" on images we want to test.
- Need to define the set of quality cuts that are eliminated "junk" detection.
- Gary had to leave conversation for a while so minutes are incomplete here - fill in if you remember!
h3. Monday 2 July
Online: Bob, Gary, Jiangang, Paul, Daniel, Steve, John, Eli, Nacho, Peter, Josh, Douglas. Several MIA at the SPIE meeting in Amsterdam and ICHEP.
- Requirements and Goals: Eli notes a Red-sequence test that has been missed in transition from Wiki pages to MSWord listings. Gary will chase this down.
No other issues raised, so the Req/Goal list is essentially complete, though some critical parameters remain TBD.
- Note that latest docdb-6549 version of SV Plan (v0.08) includes sections 4 & 5 giving list of pipeline processing requirements for SV and observations needed for SV, respectively. Everyone should look through these to see if we have been complete. Gary will convene discussion with Don Petravick, Brian Yanny, John Marriner (or designee), and Robert Gruendl on the pipeline processing priorities.
- Would like to convene a subgroup of SV that examines the collection of required/goal observations, and puts together a schedule and any needed alterations to the request to CTIO. Paul volunteers; we will need Jim Annis; Douglas can help with photometric standard fields; will want perhaps another person.
- Owners: Need to get an owner for each Requirement and Goal. Note that "owning" does not imply "do all the work" as we will be collaborating on many of these. Owner is just responsible for seeing that it's planned and done. Not everything has an owner yet. Trying to fill out the list:
- Ast-R1 will be Gary
- Douglas will help distribute ownership of Calibration goals
- IQ-R4, R5 to Jiangang. The rest of IQ and pointing will probably be combination of Aaron, Steve, Jiangang.
- Ops and Observers' tools we will foist on Klaus and Brenna
- John will make sure that SN tasks get delegated
- Eli and Peter will discuss filling out the Galaxy Detection and S/G tasks with Marcelle, Nacho, & Tom Diehl.
- Bob and Aaron likely suspects for the PSF modelling.
- Owners should get assigned this week and start writing the procedures for their reqs/goals. Further teleconferences will be largely focussed on reports from owners about progress and problems. Begin with Paul and Nacho next week.
- Schedule: current schedule has SV running 11/19 - 12/12. Should plan on having all needed SV tools in hand by end September (some will be needed during Commissioning; many will need refinement). Josh notes that late August is when DC7 will come out - should be useful for tests of many SV tools during September.
Monday 25 June¶
Agenda: Expect to be reviewing completed Req/Goals list, and discussing data processing resources required. Will start assigning jobs for additional software that is needed for SV activities.
Monday 18 June¶
Minutes: (Attending: Angelo, Gary, Jiangang, Brenna, Daniel, Klaus, Paul, Ofer, Bob, Peter, Eric, Huan, Marcelle, Luiz, Steve Ken, Douglas, John).
Action items are noted in italics
Looking through the SV Requirements / Goals posted on Requirement Categories. Items of note:
- S/N section: define signals & noise to be "close" to expectations if they're w/in 10%.
- Image Quality: need to convene discussion with Jiangang, Gary, Steve, Aaron
- Pointing: Pointing is in DES Reqs only as part of PSF budget, Steve will look this up. Eliminate 25s wording, push for 20s. Implementation issue: how to measure guiding contributions to PSF size?? Might require SISPI modifications to clock images during exposure to make a "streak camera"
- Anomalies: need to convene Daniel, Gary, Jim to fill this one out. Discussion of where bad pixel lists are kept.
- Supernovae: John points out that many SN goals are overlapped with photometry, etc., but may be distinct. Definitely want to demonstrate real-time SN detection in first season to establish credibility for 2013 telescope proposals. This means things need to be working by ~1 month after SV even if not fully functional in SV. Perhaps SV group can extend their time horizon a bit. Certainly want to at least have connectivity during SV if not full-quality image subtractions.
- Galaxy Photometry: Note that having truth-table data in hand is part of SV prerequisites so these analyses can be done. Marcelle & Liz already on this.
- Star/Galaxy separation: Peter notes that requirement for S/G homogeneity on 4-degree scales is not consistent with use of HST data as truth. Marcelle suggests approach given in Galaxy test 2-2, namely adding artificial objects to real images and testing recovery rates. Note there will be a code requirement here.
- Observers' Tools: Luiz will link to docs giving current QR / QA capabilities.
- Data processing abilities (list by function; we will decide later whether it's DESDM or otherwise that provides them). E.g. "we need astrometric solutions for images accurate to xxx. And we need stellar photometry for flat-fielded single exposures."
- Further post-processing code to execute the SV tests, e.g. code that returns detection efficiency given a truth table.
- External datasets required
- Observations required (exposure times, filters, timing or other conditions, fields)
- People needed on-site
- Data access needed for off-site people. Klaus notes to sign up for DM access at UIUC if you have not already.
Monday 11 June¶
Agenda: Review the list of SV requirements and goals that I will distribute tomorrow (Friday) and divide ourselves into groups who will concentrate on various subsets of them.
Minutes: (Attending: Brenna, Aaron, Gary, Marcelle, Douglas, Huan, Daniel, Jiangang, Paul, Alistair, Nacho, Jim, Ken, Dara, Peter, Eric).
Short conference to assign people to complete descriptions of SV goals & requirements in a list of categories. The resulting list of names is on the Requirement_Categories page. Discussion of what exactly is needed at this point: we want, by the next teleconference, to determine all the quantitative and qualitative measurements/accomplishments that would be needed (requirements) or wanted (goals) by the end of SV. We are not concerned yet with exactly what resources are needed to achieve them, also not worrying yet about whether they'll be done during Commissioning or not. You can either put your derived reqs on the wiki page, or edit the MSWord document that is attached to the SV wiki home page (please track changes if you do the latter).
It was noted that the DES SV is currently requesting 20 1/2-nights. Dara said there is a high likelihood that this would be granted.