UKIDSS Newsletter No. 8

From: Steve Warren, Andy Lawrence

December 2003

Contents

Actions arising from the 21 Nov. 2003 meeting

Besides these actions against nominated individuals, there are several points in the minutes where comments are invited.

  1. Provide a brief document of observing 'best practice' (e.g. time between filter changes), to enable the surveys to design draft MSBs.
    [Jim Lewis, Steve Warren, by mid Jan.]
  2. Using the above document as a guide, produce draft MSBs for a small fraction (e.g. 10%) of the 2-year plan for each survey, including selection of pointings using the Survey Definition Tool.
    [Nominees of Survey Heads, by end Feb.]
  3. Comment on the WFCAM photometric calibration plan, including the article below, but especially the more detailed draft pdf document by Simon Hodgkin.
    [All members, and nominees of Survey Heads, by end Jan.]
  4. Test drive the SuperCOSMOS Science Archive Demo Version interface, as a prototype of the WFCAM Science Archive. Provide feedback to Nige Hambly on the functionality, documentation, and the database schema design (in the SSA and proposed for the WSA).
    [Volunteers, and nominees of Survey Heads, by mid Feb.]
  5. Survey working groups to prepare an outline plan for science exploitation (e.g. 1 to 2 pages).
    [Survey Heads, by end Feb.]

Minutes of 21 Nov. 2003 meeting, Imperial College

1. Updates

i. WFCAM current status and schedule

Steve Warren presented a talk prepared by Mark Casali. For fuller details download the ppt file [7 Mb].

The final components to be delivered are the secondary and tertiary mirrors, and the corrector plate (all due in Dec.), and the Z filters (due in Jan.). Recent achievements include completion of warm testing of the filter-paddle assembly, and the focus mechanism. High-level software now drives detector control and acquisition. The cooldown currently in progress will allow cold testing of the filter-paddle assembly, and the focus mechanism. Currently an engineering-grade device and two multiplexers are installed, allowing a test of the full detector wiring, and an assessment of autoguider crosstalk. The next important stage, over December and January, will be, successively, warm interferometric alignment of the tertiary, cold alignment of the tertiary, installation of the corrector plate, and then a warm test of the corrector plate + tertiary.

The other main item is installation, alignment, and characterisation of the science-grade arrays. Thermal cycling of the arrays represents one of the greatest remaining risks, following catastrophic failure by delamination of two engineering grade devices. Rockwell and the ATC have worked closely together to identify the cause of the failure, and the conclusion is that it is most probably caused by contamination of the bonding in manufacture. It is believed the science-grade devices are satisfactory. The thermal cycling procedure has been scrutinised by Rockwell, and they have accepted responsibility for ensuring WFCAM has four working devices.

The schedule anticipates delivery to JAC in April. The date for start of commissioning is not yet fixed, and must fit in with the overall engineering schedule for JAC for this year, and taking into consideration any long-term consequences of delay at this time. It is now expected that commissioning will commence either in April or June¹. The process of commissioning, on-sky characterisation, and initial setting up of standards is expected to last 3 months. The camera will then stay on the telescope through completion of the Autumn extraGalactic observing season.
¹The latter has since been confirmed by JAC.

ii. Update on pipeline

Brief summaries of three talks presented by CASU members, and the ensuing discussion, are provided below. The headings are links to fuller details of the talks. The CASU VDFS page contains the project documentation, and a link to the ppt presentations.

Overview of the CASU pipeline processing

Mike Irwin summarised the current status of the pipelines. These are the Standard pipeline (with the Summit pipeline as a subset), the Further Processing pipeline, and the Advanced pipeline. Version 1 of the Standard pipeline will be ready at the end of Dec 2003, but the final tuning is dependent on the results of the on-sky characterisation tests i.e. awaits commissioning. Due to the delay to WFCAM some of the elements of the later pipelines, e.g. difference imaging, stacking/mosaicing, list-driven photometry, PSF modelling and interpolation tests have been brought forward.

CASU software modules for the Summit pipeline were recently installed by JRL at JAC and tested on UFTI data under ORAC-DR. The FITS header definition (crucial for pipeline automation) has been finalised, as well as the data-flow procedures. The raw data from the Data Acquisition System will be converted to single extension FITS files using lossless Rice compression. With this compression, the LTO-I tapes (one per array) then hold about a week's worth of data. The tapes will be read at Cambridge, converted to Multi-Extension FITS format and archived online. The raw data archive will then be transferred to ESO over the internet.

NIR array tests with UFTI and ISAAC, and MSB design

Jim Lewis presented the results of his analysis of the issues involved in flat-fielding near-ir data and subtracting the sky. These issues affect how the data will be taken i.e. the MSB (Minimum Schedulable Block) design.

It is proposed to use twilight sky for flat-fielding, but this will depend on array stability. ISAAC data indicate that both dark and flat-field frame stability could be satisfactory i.e. stable over weeks. The fringing level is very variable between arrays, being significant (+-5%) in UFTI, and virtually undetectable in ISAAC. Steve Warren presented his own analysis of UFTI data. The pixel-to-pixel flat-field appears to be stable on a timescale of weeks, but dust spots produce significant (several percent) absorption/emission. Since this can vary from night to night, but flat-fields in the full filter set cannot be obtained from twilight sky every night, we need an analysis of the degree of focus of images of dust on the WFCAM field lens. The UFTI fringe pattern is stable (i.e. scales with sky level) over 30 mins, but is variable at the 1% level over a few hours, calling into question the GPS strategy for fringe removal (offset sky every 2 hours).

There followed a discussion on how best to make progress on preparing for the observations. Ideally all surveys would fill in the MSBs for the first two years' observations, and these could be fed into a simulator to ensure that the proposed scheduling of WFCAM will allow completion of the 2-year plans on time. However without knowing the characteristics of the instrument, the best data reduction strategy (and so best observing strategy) cannot be decided. The advantage of producing an observing strategy that is close to the final strategy, but could be modified dependent on the results of the on-sky characterisation tests, was recognised. To this end it was decided that CASU (Jim Lewis) and UKIDSS (Steve Warren) should produce a brief document, providing a guide to best observing practice (e.g. frequency of changing filters) in the light of their knowledge of the data reduction issues. Each survey should then produce draft MSBs for a fraction of the survey, which will be forwarded to CASU for comment.

WFCAM photometric calibration

Simon Hodgkin described the current status of the plan for calibrating the surveys. He provided a list of criteria for the selection of standard fields, and produced a draft target list. An analysis of UFTI data has been completed to try to understand what frequency of standard observations should be undertaken on any night. One issue to consider is whether all broadbands (ZYJHK) should be observed every night. This allows greater flexibility in scheduling the observations, for example if conditions vary. The Calibration Working Group is aiming at producing a concrete proposal for the calibration plan by end March. A draft is available, and everyone is invited to comment on this document in its current form.

It is planned to produce two UKIDSS papers on calibration, analogous to the SDSS papers by Fukugita et al., and Smith and Tucker. The first will present the WFCAM ZYJHK photometric system, on the basis of computed synthetic colours. The second will present the empirical system, based on actual measurements. Paul Hewett presented an outline of his plans for the first paper, that he will lead. Colours (g'r'i'z'ZYJHK) will be presented for stars (e.g. Gunn-Stryker atlas, Leggett MLT spectra, theoretical cooler brown and white dwarfs), galaxies (e.g. redshifted Coleman, Wu, and Weedman spectra, BC 2003 models), and quasars. The paper will include the zero-point differences Vega-AB for all the WFCAM bands. If you have any suggestions for the content of this paper, please contact Paul.

iii. Update on archive

All developments for the WFCAM Science Archive (WSA) are documented on the WSA web page.

Points arising from the Nov. 2002 meeting (Newsletter No. 7: the question of SDSS fluxes at IR source detection positions, inclusion of new external catalogue data, and general stacking options available at first light) have been addressed and are now incorporated in the design. A default, simple, stacking option will be available at first light. SDSS fluxes will be available one year later. Major milestones and deliverables achieved by WFAU since the last meeting are as follows:

WFAU now requires UKIDSS community input on the last item, specifically with a view to the following:

With this in mind, WFAU asks UKIDSS volunteers to test drive the SSA Demo Version interface. We are particularly interested in comments on the user-friendliness of the various access points which include traditional "cone search" pages, user-supplied catalogue cross-match, and fully flexible SQL interfaces.

Please also browse and comment on the analogous WSA database documentation, and send comments directly to Nige Hambly.

2. Survey Definition Tool

A beta version of the Survey Definition Tool (SDT) has been produced by Martin Folger. The SDT reads in definitions of areas of sky to be surveyed, and, choosing guide stars, tiles the area with WFCAM pointings. The results are then fed to the Observing Preparation Tool. Nige Hambly has tested the beta version of the SDT, and requested a number of adjustments. These will be fixed by the end of the year. In the New Year individuals from each survey should test out the SDT (not needed for UDS), in conjunction with producing draft MSBs (see above).

3. Other business

i. Opportunites for large public surveys with VST

The VLT Survey Telescope (VST) is expected to become operational in 2005. UKIDSS envisage applying to use the VST to provide complementary deep optical image for the GCS, LAS, and DXS surveys. Following a workshop in May this year, ESO have been developing the procedures to be followed for implementing large surveys with the VST. Alvio Renzini explained the proposed procedures. A brief summary follows, and the detailed proposal to the ESO STC is also available. Comments on this proposal are invited.

It is envisaged that 40% of VST time will be dedicated to public surveys. The procedures for implementing public surveys with the VST are designed to maximise the scientific productivity of the VLT, to ensure the most efficient use of VST nights, and to provide complementary optical coverage of surveys at other wavelengths. A set of surveys varying in depth, area, and Galactic latitude is expected.

The procedures proposed to the STC are broken into 12 stages. Following a "Call for Public Survey Concepts", the key element is the establishment of a Public Survey Panel (PSP), which will review the concepts, and use them as a starting point to design a comprehensive set of surveys, that optimally satisfy a set of criteria. The PSP will produce a document specifying the characteristics of each of the surveys designed (areas, depths, filters), as well as the detailed survey products, and a required delivery date. The ESO community will then be invited to form consortia to bid to undertake one or more surveys.

ii. Current status of VISTA

Will Sutherland presented a progress report on VISTA. All the major contracts were signed by early 2003, and during recent months many subsystem design reviews have been completed. The road and site levelling are complete, and digging foundations will start in December. The primary mirror blank is in Moscow awaiting gluing of the pads, and polishing will start next year. The telescope structure has passed PDR and is heading for FDR in March 04. The IR Camera consortium (RAL, UKATC, Durham) have issued the document pack for the Camera FDR on 18+19 Dec. The project is on track for science operations in late 2006.

The IR focal plane will be 16 Raytheon 20482 detectors at 0.34 arcsec/pixel, with 90% spacing in X and 42.5% spacing in Y, so six "pawprints" will give a 1.5 x 1.0 deg. tile with all pixels covered at least twice. Currently Y, J, H, Ks filters are on order, with an option for additional filters e.g. z and narrowbands at 27k pounds per band. With respect to detailed survey planning, this should to some extent depend on early results from WFCAM, but there is a need to plan some VST surveys quite soon to ensure availability of complementary visible data (see above notes on ESO Public Surveys).

iii. FMOS and CIRPASS

The meeting was reminded that FMOS (30 arcmin f.o.v., 400 fibres, 0.9-1.8µm, OH suppression) will be a powerful instrument for exploiting UKIDSS, and that a workshop will be held in Kyoto, Jan 13 and 14, 2004, to discuss science projects, and the possibilities for a large specroscopic survey (see Jul 15, 2003 email from Gavin Dalton, in the UKIDSS email archive). Enquries to Gavin Dalton or Kouji Ohta are still welcome.

The schedule for FMOS has the instrument on Subaru towards the end of 2005. In the meantime CIRPASS has had successful runs at the AAT and Gemini-S, and provides an indication of the potential of FMOS. CIRPASS is a private instrument, but there are opportunities for collaborations, as described on the CIRPASS web page.

iv. SCUBA2

The SCUBA2 instrument will be capable of mapping the sky up to three orders of magnitude faster than SCUBA. The idea of planning a set of large surveys is currently under consideration. The five elements of UKIDSS are natural targets. If you are interested in advancing this possibility get in touch with: Melvin Hoare (GPS), Ian Smail (DXS, UDS), or Steve Serjeant (LAS).

4. Debate on science exploitation of UKIDSS

Andy Lawrence led a debate on the science exploitation phase of UKIDSS. The issue was, to what extent should UKIDSS organise the exploitation of the surveys. Three models were discussed:

  1. It should be a free for all. UKIDSS' only role is to ensure that the surveys, as designed, are completed.
  2. UKIDSS provides the focal point during the exploitation phase, fostering collaborations, and organising meetings to share knowledge, and nurture scientific debate.
  3. Topic teams should be formed, and specific papers planned, with named lead authors.
Option 3. was universally considered unhealthily rigid, and, with one voluble Durham exception (identified only by the initials IS to give a false sense of anonymity), everyone agreed that 2. was the preferred option. Since the data are public, there was no concern that UKIDSS could be open to the criticism that they had ring-fenced the best science. It was pointed out that it was inevitable that the majority of the science would be undertaken by members of the consortium, and it would be foolish not to exploit the established UKIDSS network, in pursuing the goal of maximising the scientific return from WFCAM. Furthermore in applying for post-doc support it will be useful to be able to point to the consortium's coordinated approach to scientific exploitation.

It is assumed that collaborations will aggregate naturally. For example a brown dwarf mailing list is already in existence. It was considered useful to create a public UKIDSS 'Science exploitation' web page to share information. Collaborations can advertise their existence, and plans there. This will provide a forum for promoting further collaborations. Suggestions for the format of this page are welcomed.

Everyone agreed that a workshop, as early as feasible, to present the first science results, would be a valuable catalyst. A date about 4 months after the surveys start in earnest was considered best (e.g. Jan 05, for Sept start). Members are invited to volunteer to host the event. A larger, full conference in the summer of 2005 was also suggested.

To focus the debate, AL then sketched out a list of UKIDSS papers that could provide the technical references for the science papers. After modification, the draft list, with possible lead authors, looks like:

Comments on this scheme are invited.

The survey working groups were asked each to prepare an outline plan for science exploitation. These would be placed on the web page.