From rgm@ast.cam.ac.uk Wed Oct 15 16:06:34 1997 Return-Path: Received: from cass49.ast.cam.ac.uk by ast.cam.ac.uk (SMI-8.6/SMI-SVR4) id QAA21587; Wed, 15 Oct 1997 16:06:33 +0100 Received: by cass49.ast.cam.ac.uk (SMI-8.6/SMI-SVR4) id QAA02278; Wed, 15 Oct 1997 16:06:32 +0100 Date: Wed, 15 Oct 1997 16:06:32 +0100 From: mike@ast.cam.ac.uk (Mike Irwin) Message-Id: <199710151506.QAA02278@cass49.ast.cam.ac.uk> To: rgm@cass41 Subject: JSC submission Content-Length: 19156 Status: RO X-Status: A For your info ..... Mike RGO/ING Deep Optical Sky Survey (The RIDOS Survey) ================================================== A proposal submitted to the JSC of the ING by Drs N A Walton and M J Irwin on behalf of the ING, RGO and UK Wide Field Astronomy Panel. 1. Summary ---------- The Wide Field Camera (WFC) at the 2.5-m INT provides the UK and Dutch astronomical communities with a unique opportunity to undertake a high spatial resolution, deep, broad band imaging survey of selected areas of sky at high Galactic latitudes. It is vital that we exploit this facility, with its excellent sensitivity and large area coverage, in a timely fashion for the benefit of the whole Anglo-Dutch community, in order that we maintain our excellence as world leaders in wide field optical survey work. Indeed, The UK Wide Field Astronomy Panel, at its meeting in June, stressed the urgency of a deep sky survey with the INT+WFC. Deep optical surveys combined with the renaissance of interest in large area surveys at all wavelengths are fundamental in searching for rare types of object, discovering new categories of object, and doing accurate large volume statistical surveys. This ability to upgrade our survey capabilities is an essential ingredient for successful future exploitation of the next generation of 8-m class telescopes such as Gemini. We propose to undertake a survey programme in the UBVRIZ broadband filters, and [OIII] 5007A narrow band filter, to initially map some 100+ square degrees of high Galactic latitude sky in various selected areas (e.g. Coma, Virgo clusters, an equatorial strip). Such a dataset will inevitably alter not only our approach to certain astronomical problems, but also the very questions we ask. In order to ensure maximum impact, the RIDOS survey should commence in Semester 1998B making full use of the newly commissioned large EEV CCDs in the WFC, and requires a preliminary allocation of time in 1998A to develop the pipeline processing and queue scheduling aspects. The key technology and expertise needed to rapidly process this sky survey exists within the ING and UK. In order to maximise the scientific returns, the data will immediately be made available to the UK and Dutch research community, in semi-processed form (in realtime), and later in processed form (ie. astrometrically and photometrically calibrated) via an on-line archive. The scientific returns will be enormous: ranging from Solar System studies, such as searches for primordial comets, out to searches for the most distant objects in the Universe, such as high redshift quasars and primordial galaxies. The data will be especially powerfull when used in conjunction with survey data in other wavebands, e.g. the 2 micron 2MASS survey, the deep VLA radio surveys FIRST and NVSS, the ROSAT all sky survey, and selected deep X-ray and radio survey fields. It is suggested that the project be reviewed after twelve months, with the expectation being that it would be continued for a further two year period. Hence in total the project life-span would be 4 years, with a survey area of 100-200 square degrees in up to seven colour bands. ** This proposal has the full support of the UK Wide Field Astronomy Panel and the Directors of the ING and RGO. ** ACTION POINTS FOR THE JSC: -------------------------- A: Approve the allocation of a 5 night allocation of the WFC in Semester 1998A (May or June) B: Approve and authorise the RIDOS survey, to commence in 1998B 2. Scientific Rational ---------------------- Despite the considerable advances of detector technology within astronomy, very little improvement has been made in surveys beyond those available in the 1950's when the Palomar Sky Survey was carried out. A photographic plate taken on a 1.2-m Schmidt is sky limited in about 1 hour, covers a sky area of some 40 square degrees, but is only 1-2% efficient. The current best sky surveys therefore amount to no more that a 60-120 second glance at the Universe using a 1.2-m telescope. Although the existing photographic surveys cover the whole sky there is no requirement that a fainter CCD-based survey should do the same. Significant scientific progress will be made by: - exploiting the large gains in efficiency, particularly in the U, I, and Z-bands - by going 10 times fainter - by exploiting the factor of 5 or so higher accuracy due to the far superior dyanmic range of CCDs and the superior image quality with seeing ~1 arcsec compared to 2-3 arcsec of current survey material. The RIDOS survey is designed to address a wide range of astronomical topics including: - cosmological parameters from redshift surveys and supernovae - identify and study quasars and galaxies at high redshift - large scale structure at z>1 - reddened quasars and hidden AGN - links between AGN and starburst galaxies - supernovae to z=1 - galaxy formation - galactic evolution at z<1 - halos of galaxies - Galactic structure and the faint star luminosity function - intergalactic stars in nearby clusters - microlensing toward nearby galaxies - extreme solar system objects It will also be a powerful tool in conjunction with other frequency surveys, in searching for objects: - Xray, radio, optical BL Lac searches - radio, optical searches for high-z RG and QSR's - high-z Xray and optical clusters - Dark clusters from S-Z searches More generally it will provide a uniform multi-colour database from which samples may be selected for other studies. In the context of the 8-m era, and to support the UK's strategic investment in the Gemini telescopes, the RIDOS survey will provide many objects that can be studied spectroscopically with 8-m class telescopes. A large fraction of the INT WFC programmes submitted as PATT applications from the UK and Dutch communities 97B and 98A would benefit from the RIDOS survey data. The RIDOS survey would operationally be an extremely efficient use of scarce observing time, especially when folded into the queue service mode of operation to which the INT+WFC combination is evolving to. 3. Other Deep Optical Sky Surveys ---------------------------------- Several other organisations have recognised the urgent need of and potential for deep optical sky surveys and are planning or have already started CCD-based surveys. However RIDOS will have substantial advantages over these as Table 1. shows. Table 1. Comparison of RIDOS with other Optical Surveys Survey Limiting Mag Sky Area Resolution Availability (R) (sq deg) arcsec (to community) ----------------------------------------------------------------------- RIDOS 24.5 100-200 < 1 from mid 1988 ----------------------------------------------------------------------- SDSS 23 > 10000 1 > 2007 NOAO 25 18 1 n.a. EIS 23.5 24 < 1 from early 1998 POSS 21 all sky 2-3 now+ HDF 28.5 < 0.002 0.04 now ----------------------------------------------------------------------- As can be seen important aspects of the RIDOS survey are its rapid availability (c.f. SDSS which will not be released to the general community for some 10 years after start date), its order of magnitude better sky coverage compared to NOAO and EIS (which will mainly concentrate on the south) and good resolution and limiting magnitude compared to the POSS. A brief summary of these and other significant survey instruments follows: The NOAO Deep Wide-Field Survey is sampling the sky in two 9 square degree patches in selected fields near the NGC and SGC, in both broadband optical and NIR passbands. It would have similar detection limits to the RIDOS survey, but much reduced sky coverage. The ESO Imaging Survey (EIS) is a 12 month program of deep multicolour broadband optical imaging using the 3.5-m NTT, to support observations with the VLT. This survey is limited to the Southern hemisphere, has similar sensitivities to RIDOS, but reduced survey area. The Big Throughput Camera (BTC) is an optical CCD mosaic camera operating at the prime focus of the CTIO 4-m with a 30x30 arcmin field of view. A further 38x38 arcmin imager will be shared between CTIO and KPNO. This is currently carrying out a survey for a private group and is also being used in the NOAO survey. Megacam on the CFHT, a project to put a 1 square degree mosaic imager of 4kx2k thinned CCDs at prime focus. No survey is yet planned for this camera. The most ambitious survey will be The Sloan Digital Sky Survey (SDSS). This will employ a purpose built 2.5-m telescope with a large area imaging camera (giving a 3 x 3 degree field). Additionally it is equipped with two 600 fibre units feeding intermediate disperion spectrographs. Some 10,000 square degrees of the sky above a declination of -20 degrees will be surveyed, with an additional 225^2 degrees on the SGP. The northern survey will reach a limiting magnitude in R of 23, some 2 magnitudes brighter than RIDOS. This survey will take some 5 years plus to complete, it is currently behind schedule, the initial year of testing not yet begun. Furthermore, the data is not being made public for a period of 10 years from the starting date of the survey. The RIDOS survey has advantages over all of these surveys, either in terms of limiting magnitude, number of passbands surveyed, or area of sky surveyed, as shown in Table 1. 4. The RIDOS survey ------------------- a. Field Locations: The Wide Field Astronomy Panel carried out a recent community survey on the need for faint wide field optical imaging. The survey showed a strong requirement for access to digital wide field imaging data at both optical and NIR wavelengths. A few years prior to this an INT CCD Working Group had consulted the community to identify several regions of sky of particular interest including: - the Coma and Virgo clusters - fields at the North Galactic Pole (for Apr - May) - fields at the South Galactic Pole (for Oct - Nov) - equatorial survey strips The more general survey fields will be selected to have low IRAS cirrus emission, low N(HI), the availability of radio data from FIRST, deep VLA fields and deep ROSAT survey regions, and at least one field close to the Ecliptic allowing a search for faint Solar System objects. b. Planned Depth and Sensitivity: All telescopes have a natural survey time constrained by slewing, aquisition and readout overheads (two minutes) and by a tradeoff between area coverage and depth. For a 2.5-m telescope on a good site (median seeing of less than 1 arcsec) the natural survey timescale is typically 10-15 min per field to simultaneously maximise time on sky, area covered and volume surveyed. All exposures are sky-noise limited, therefore to go 1 magnitude deeper than this natural limit requires 6x the exposure and would only be justified if dN(m)/dm > 6 at the limit. Given these generic constraints the INT WFC produces the following survey detection limits in dark sky, using PSF profile fitting in median seeing of 0.7 arcsec. Table 2. RIDOS Survey Detection Limits ----------------------------- Bands 5 sigma limit 1 sigma SB limits Exposure Time (mag) (mag/sq arcsec) U 24.75 26.6 700s B 25.75 27.6 600s V 25.25 27.1 600s R 25.00 26.8 775s I 24.25 26.1 850s Z 22.50 24.3 600s [For 1 arcsec seeing subtract 0.4 mag, 1.5 arcsec seeing subtract 0.8 mag.] The survey limits will allow: - detection of brightest cluster member galaxies out to redshifts beyond z>2 - detect Galactic Halo stellar tracers such as HB stars out to 300 kpc - cool white dwarfs to 1 kpc - distant supernovae to z=1 - Solar System objects beyond 50AU - probe the visible edge of the Universe down to absolute luminosities M_I = -23 c. Sky Coverage of INT WFC: The WFC mosaic imaging area with three 4096x2048 3-edge buttable EEV CCDs is 33.7 arcmin x 22.8 arcmin (i.e. 0.21 square degrees). In order to maximise the amount of sky surveyed we would not generally dither the mosaic to fill in the small spaces between the three individual CCDs in the mosaic. The EEV CCDs have 13.5 micron pixels resulting in sampling at 0.33 arcsec/pixel which is well suited to exploiting the median ING site seeing of 0.7 arcsec. The readout time of the mosaic (after soon to be implemented controller upgrades) will be some 120 seconds, hence on average about one square degree could be imaged per hour. Assuming 8-10 hour nights some *3* square degrees of sky could be surveyed per clear night, in three passbands. If necessary survey data obtained in non-photometric but otherwise good conditions, could be calibrated using service data obtained on the JKT. For the 'micro lensing' survey images in B of selected fields in M31 would be obtained at the begining and end of each 5 day period. For the searches of Kuiper belt objects, fields near the ecliptic would be observed twice (interspersed by 2 hours) in the V and R bands. d. Observing Plan: We propose to observe, initally over four semesters, each selected region to within 0.5 magnitudes or better of the limits specified in Table 2. This will require seeing conditions better than 1.2 arcsec During conditions poorer than this alternative service or PATT queue scheduled programmes would be carried out. * Full observations would commence in Semester 1998B (from September 1998) * * Five nights of preparatory time would be required in semester 1998A * after the commissioning of the EEV's in the WFC. In the bright parts of any grey nights allocated we would propose to image in the OIII, Z and possibly I-bands. The remaining broad-band imaging would need to be carried out during DARK time. This proposal would be ideally suited for `Queue' observing, and therefore could be scheduled at the 'grey' ends of a 'dark' queue period. In semester 98A (and 99A) we request TWENTY nights to be allocated to this project, to be split into FOUR periods of FIVE nights each in Mar, April, May, June. In the following semester 98B (and 99B) a similar allocation would be required, ie. four periods of five nights in Sep, Oct, Nov, Dec. We would also utilise the JKT to carry out photometric calibration as needed. By April 1998, the JKT control system will be completely re-engineered (TCS, ICS, DAS, Autoguider), making it possible for it to be remotely operated from the INT control room. Further the large 2048x2048 SITe CCD will equip its CCD camera giving it a field of 12x12 arcmin. The JKT would be used during some of each night of the WFC survey nights to carry out calibrations of some of the fields observered by the INT. Other fields would then be calibrated with respect to these. Some INT WFC photometric fields would be observed for calibration consistency check purposes. (The rest of the JKT nights could either be used for service, or for scheduled observing - it is assumed that there would be a standard survey filter wheel available at the JKT that could be quickly swapped for the scheduled observers filter wheel). Data from the CAMC would be used to give an indication of the grey extinction values for each night. Data from the WFC autoguider would be logged to track changes in image quality and relative transparency through the night. In total we request TWELVE weeks spread over FOUR semesters. This would enable approximately 100 square degrees of sky to be covered in four colour bands (with some additional fields surveyed in [O III] and the remaining bands). It is suggested that the project be reviewed after twelve months, with the expectation being that it would be continued for a further two year period. Hence in total the project life-span would be 4 years, with a survey area of 100-200 square degrees in up to seven passbands 5. Archive and Data Dissemination --------------------------------- We are committed to maximising the scientific return from this project by involving the rest of the community as much as possible. We will therefore use pipeline processing techniques to ensure that the data obtained is available to the community in the most timely manner possible. The raw data will be available on-line for a 6 month period (available via the ING archive but with access restricted to accredited UK and Dutch users) and subsequently accessible in both raw and reduced form via the `RGO/ING Deep Optical Sky Survey Site (RIDOS)' which will be hosted at the ING archive web site (http://www.ast.cam.ac.uk/ingarch/) which is presently located at the RGO. The reduced form will contain both digital atlasses and object catalogues astrometrically and photometrically calibrated. Astrometric calibration will be carried out using on-line photographic sky catalogues and incorporated in both the digital maps via FITS header extensions and in any object catalogues produced from the data. Photometric calibration frames will be obtained at intervals throughout each photometric night of the survey from the INT or JKT, or from JKT service data for non-photometric nights. Quality control will be assured by a rigorous and mainly automatic data handling system making use of the extensive improvements provided by the new INT DAS. It is anticipated that an international workshop will be organised in the winter of 1998, at the ING, to address the scientific exploitation of the survey data. This will enable the coming together of interested astronomers, both observational and theorists, to further the science programmes based on this data set. 6. Staffing and Equipment requirements -------------------------------------- The data acquisition will be carried out in 'queue' service mode. Local support astronomers and later, observing assistants (suitably trained telescope operators) will be required. From semster 98B the JKT will be operable remotely from the INT control room, hence it is foreseen that that data will also be obtained by the observers at the INT. Raw data will automatically be archived on the ING Engineering CD tower. Initial pipeline pre-processing will be performed to bias subtract and flat field the data, this being written to CD-ROM for inclusion in the archive (hence access to the user community). Copies of the pre-processed data will be made nightly and sent on CD-ROM to the RGO Data Centre in Cambridge. Staff there (with expertise from the APM and archiving/data handling eg. Drs. Mike Irwin, Jim Lewis and Peter Bunclark) will be responsible for further processing. It is expected that aproximately one staff year of effort will be required in the UK for the initial two year period of this survey. A flux calibrated product will be made available from the ING data science archive. Final products will include a catalogue of all objects, astrometrically sorted. The complete survey data will be made available as a CD-ROM or DVD-ROM set.