Minutes of UKIDSS Consortium meeting, 11/11/05, Edinburgh

Author: Simon Dye

Those Present

Steve Warren, Andy Lawrence, Simon Dye, Melvin Hoare, Chris Simpson, Nigel Hambly, Sebastien Foucaud, Antony Smith, Mark Casali, Richard Jameson, Richard Mc Mahon, Simon Hodgkin, Mike Irwin, Jon Davies, Paul Hirst, Tim Kendall, Nick Lodieu, Omar Almaini, Alastair Edge + others (add your name if I've missed you off!)

Presentations

1. Mike Irwin, Pipeline status

• 05A release version 2 will fix stacking bug (causing non-linearitires in aperture fluxes), have improved sky subtraction, have improved 2MASS calibration, but no X-talk correction
• 05B version 1 will fix X-talk but will have no persistence corrections
• Inter-pixel capacitance causes 20% underestimate of pixel-pixel noise esitmates
• There was discussion about possible flagging of regions and/or sources potentially affected by X-talk. UKIDSS community strongly in favour of this.
• Pipeline astrometry cf. 2MASS has rms scatter of 80mas and most of this is error in 2MASS
• Internal astrometry scatter even smaller.
• Including r^5 term in distortion reduces this scatter by nearly 50% again.
• Dark sky background levels vary by +/- 4% about mean over all 4 arrays. This is not an error in gain since don't see corresponding error in photometry cf. 2MASS. Spatial variation of residuals wrt 2MASS stacked over many images over all arrays show scatter of 0.02-0.03 mag on average (sometimes worse, sometimes less depending on group of images selected). Pattern of residuals fairly consistent between different image groups and filters, southern-most half of array 1 is worst. Simon Dye suggested that CASU try repeating exercise with linear flatfields (a la 2MASS) as scattered light, not included in linear flat, may be the cause.
• Progress maps for whole survey by Eduardo and Marco (CASU) shown.

2. Nigel Hambly, Archive Status

• Takes 2-4 weeks from receiving last bit of data from CASU before it can be released in WSA
• WSA team want to know: max acceptable ellipticity, seeing and departure from nominal zero points.
• Rejected MSBs are still not flagged in FITS headers so archive and pipeline can't pick them up. Paul Hirst pointed out that mechanism exists already: CASU have local copy of OMP containing rejected MSB info.
• Data release schedule discussed (see later section)

3. Simon Dye, Science Verification

• Manual reduction of data has revealed: NDR mode (10s) reduces read noise by 30% wrt CDS mode, array 4 has 2x spread in pixel sensitivity compared to arrays 1&3 and array 2 is 1.5x, thermal emission/scattered light in K changes quickly throughout night, zero points computed using spectrophotometric standard agree very well with CASU zero points (Mike stated that the small offsets found have been reduced even further in SV2), not all LAS SV targets found by pipeline

ACTION: SD to verify that this is because they fall beneath the detection threshold.

• Comparing SExtracted magnitudes of manually reduced data with magnitudes of the same objects from the WSA (same data but flattened and dark subtracted with different flats/darks) shows unexpectedly large scatter in differences. Using SExtracted magnitude errors, this spread has sigma=1.75, considerably larger than it should be.

ACTION: SD to repeat analysis but running SExtractor on a stacked WSA frame instead.

• Highlights of SV from the other surveys: H band photometry is systematically off (corrected in SV2), GCS would like astrometric accuracy to be twice as good for propper motion studies (although SV data has poor image quality - need to retest with better data), star gal separation looks good on the whole but there are binary stellar objects that are still not properly classified.

4. Steve Warren, Survey Status

• Internal/external photometric accuracy is 0.03-0.04, spec is 0.02.
• Differing classification of the same objects between bands.
• LAS Y band obs reach Y=20.10 +/- 0.28 in 40s, spec is 20.50
• LAS has moved from 2x2 + 2x2 5s microstepped obs in every filter to 2 x 20s unmicrostepped in Y, 2x2 + 2x2 5s microstepped in J and 4 x 10s unmicrostepped in H & K.
• Guide star problem with LAS - lack of them => some MSBs (10% ish) are unguided. Faint guide stars around 16.5 affected heavily by moon and seeing. Paul Hirst suggested such faintly guided observations be put into separate MSBs with more stringent seeing and moon constraints.
• Survey speed is ~20% slower in 05A than expected. Not quantified for 05B yet.
• 45 institues have registered for UKIDSS data access. 2/3 are UK based.
• Complementary VST proposal has reached phase 2
• Publicity: press release June 24, photom paper now submitted, had 6 UKIDSS talks at Edinburgh workshop over last two days, idea of legacy gallery gaining acceptance.

5. Richard Jameson, LAS status

[missed first half of talk]

• Various brown dwarfs detected in SV data, possibly some Helium white dwarfs.

6. Nigel Hambly, GCS status

• Currently have: 05A SV; 6.2 sqdeg in IC4665, 6.2 sqdeg in Upper Sco, 4.6 sqdeg in Coma Ber. 05A May/June; 40 sqdeg in Upper Sco, 20 sqdeg in Coma Ber. 05B to date; 80 sqdeg in Pleiades, 50 sqdeg in Alpha per, 20 sqdeg in Tau-Aur & Orion, 290 sq deg in Hyades.

7. Melvin Hoare, GPS status

• Currently have 37 sqdeg of |b|<1 in inner galaxy in JHK, photometric conditions, and 76 sqdeg of 1<|b|<5 of inner gal in K, non-photom, 1st of 3 epoch.
• Av extinction map of M17 constructed
• GPS's own PSF fitting photometry package compares well to 2MASS photom and CASU's aperture mags.
• GPS have been allocated future SCUBA 2 time.

8. Alastair Edge, DXS status

• DXS has completed 185 MSBs to date.
• SV has a nearly complete tile in K, 1/2 complete tile in J in ELAIS N1 Depth not quite as expected in these - prob get better with better image quality.
• DXS have obtained 10 sqdeg of deep optical Subaru data from Toru Yamada to help with separating out 1<z<2 galaxies. This data has a counterpart for nearly every object detected in DXS SV data.
• ERO number count plot agrees well with literature.

9. Omar Almaini, UDS status

• 62 hours observed since Sep but only 39 hours on source (guide stars, obs efficiency and standard fields being responsible for loss in efficiency).
• Target for end of 05B is 200 hours on source. Realistic goal is 120-150 hours on source with K=22, J=23.2 (5sig pt src)
• UDS has new field centre => all guide stars have R<15
• Have settlted on 3x3 microstepping until end of survey. No noticeable difference cf 2x2 but doesn't degrade image quality and allows for future improvement in image quality.
• Having looked at 6 hours of data, reached K=20.3 with 50% completeness. This is very close to requirement.

Andy Lawrence, The future

• Possibility of having 05A extended release (see below)
• Need to make report to board May '06
• Nov '06 board: renew proposal if required.

[SD left meeting ... to be continued ....]

Discussion on data release

Disucssion on what to release when:

• UDS in favour of releasing as much as possible in Jan 31st release. They feel data is good enough already and need more than small amount contained in 05A for Spitzer proposal deadline in mid Feb 06. UDS are also in favour of incremental releases, rather than large releases every semester. WSA in favour of this approach.
• Andy Lawrence pointed out that a 'release' is really an announcement, that data can be released independently of these announcements.
• There was some concern over releasing too much data too soon. Don't want to give UKIDSS a bad name by having several re-releases. Also, releasing too much in Jan 06 will necessarily involve several different data versions. Better to keep releases simple.
• One compromise could be to call DR1 an EDR and have this in different parts (eg. EDR part 1 processed with v2 pipeline, EDR part 2 with v3. etc) Then DR1 would be released on Jun 31st 06.
• EDR could contain 05A + n months worth of data from 05B.
• Paul Hirst pointed out that UKIDSS doesn't officially get to make the decision anyway, that it goes through VDUC who communicate with the UKIRT board. Andy Lawrence agreed. UKIDSS will simply make suggestions via the official channels. AL pointed out however, that the UKIDSS consortium, in the eyes of the astronomical community are responsible for the data and its release.
• Owing to lack of general agreement, Andy Lawrence suggested that survey heads contact their respective working groups and write a proposal each with their data release requirements. This was agreed.

ACTION: Survey heads to contact their working groups and write a proposal concerning their data release requirements.