The Ultra Deep Survey (UDS) is a the deepest public survey to be conducted with the UKIRT Wide-Field Camera. This will reach depths of K=23, H=24 and J=25 over an unprecedented 0.77 sq degree region. The field chosen is the Subaru/XMM Deep Survey Field at J0218-05. A total of 15 nights of Subaru Suprimecam imaging was approved last month for B,R,I,z' to 5-sigma point source depths of of B=29, R=28, I=27, z'=26. This is also the proposed target for a major (0.5 sq deg) SCUBA sub-mm survey (the largest ever conducted). We propose to use the facilities available within ESO to complement these data with much-needed U and V band imaging to equivalent depths. This data will be vital for identifying Lyman-break galaxies at z~3, in addition to providing a wide baseline for photometric redshift determination. Further information can be found below.
The UKIRT Infrared Deep Sky Survey (UKIDSS) is scheduled to commence in late 2003, taking up 50% of the total time on the UK Infrared Telescope (UKIRT) to carry out a nested series of public surveys (from wide and shallow to very deep).
The survey instrument (WFCAM) consists of four 2048x2048 Rockwell devices at 90% spacing. With a pixel scale of 0.4 arcsec, with micro-stepping to sample the PSF, the exposed solid angle is an unprecedented 0.21 sq. deg per observation.
The deepest UKIDSS survey is the UDS (ultra-deep survey) which aims to reach an unprecedented depth of K=23, H=24, J=25 over 0.77 sq deg over the 5 year lifetime of UKIDSS.
Our current plan is to complete the K-band imaging over the first two years of operation. After much discussion, the Subaru Deep Field field has now been chosen as the location for this survey (02h, -05 dec).
The UKIDSS UDS will give a thousand-fold increase in area compared to any previous survey to this depth, which we anticipate will leave a lasting scientific legacy comparable to the impact made by the Hubble Deep Field. A major motivation is to determine the comoving number density of massive elliptical galaxies at high redshift (z>2) and hence discriminate between competing theories for their formation. Previous IR surveys to this depth, covering only several square arcminutes, have been unable to address this question. With the unique capabilities of WFCAM we can unambiguously test whether the most massive galaxies are already in place at z=3, or whether they form at lower redshifts through the mergers of less massive Lyman-break objects.
Another major goal is to study the links between (and clustering properties of) SCUBA galaxies, EROs, Lyman-break galaxies and X-ray selected quasars. Do some subset of these apparently disparate types of object represent a formation sequence (Archibald et al. 2001, Smail et al. 2001, Almaini et al. 2002)? Their evolution and clustering properties in particular will be provide a powerful test of hierarchical theories for galaxy formation.
The U-band is vital for identifying unobscured galaxies at z~3 by the Lyman drop out technique (Steidel et al. 1996). Such objects will then be ideal targets for future follow-up spectroscopy with VLT.
The V-band is also important, both for constraining possible B-band drop outs and for providing reliable UV continuum measurements of the SEDs of z~3 sources. In addition, combined with the very deep B,R,I,z' from Subaru and J,H,K from WFCAM we will have a very powerful baseline for photometric redshift determination.
With this data we can then compare the clustering and evolution of Lyman-break galaxies, X-ray selected quasars, sub-mm sources and faint radio sources in a single, exquisite deep field over an unprecedented 0.77 sq deg area.
Our aim is to obtain matching photometry for sources with precision IR-colours, i.e. a limit of K=22. Lyman drop-outs at K~22 corresponds to I~24-25 for typical field galaxies, or B~26. Using the selection criteria of U-B>1 for an LBG (Metcalfe et al. 2001) then we require depths of U~27. This will require ~4 nights on a 4m telescope, or ~20 nights on the 2.2m VST.
Likewise, the corresponding V-band limit (V~27) can be achieved on one night on a 4m telescope, or 2-3 nights on the 2.2m VST.
A summary of the multi-wavelength data available in this field (either observed or in preparation) is summarised below:
The addition of ESO U and V imaging to great depth will ensure that this survey remains unsurpassed in its scientific relevance for many years.
| Depth (5-sig) | U=27 V=27 (n.b. Sloan filters, Vega mags) |
|---|---|
| Area | 0.8 sq. degs |
| Field | Subaru/XMM Deep Field 02h -05deg |
| Clear nights | 25 nights on 2.2m VST (5 nights on 4m telescope) |