Plan for completion of UKIDSS in 2012

Steve Warren, Imperial College London, on behalf of the UKIDSS Consortium

At the November 2009 meeting the UKIRT Board approved the plan
submitted by the UKIDSS Consortium for completion of the surveys,
assuming an end date of May 2012. It has now been confirmed that UKIRT
will remain operational until 2012, and we are implementing the
approved plan. The surveys are progressing well, but our original
operational model for the time needed was optimistic by perhaps 25%,
so each of the surveys has had to be trimmed. The main discrepancies
are that the overheads were underestimated, as well as the quality
control losses. Both have improved with time, nevertheless.

In Table 1, column 2, we compare the revised goal compared to the
original goal. The revised final goal involves a major descope of the
UDS, but only a modest descope of the other surveys. Although the UDS
is now observed most of the time when visible, previous operational
issues have meant that it lags behind the other surveys.

Survey  Revised   Completeness
         goal   at end 09A (DR7)
LAS      0.85        0.53
GCS      0.77        0.41
GPS      0.85        0.41
DXS      0.87        0.55
UDS      0.43        0.28

Table 1. The figure in the second column specifies for each survey the
completeness of the revised final goal relative to the original goal,
in terms of the product of area and integration time, summed over
filters. The third column quantifies the completeness relative to the
revised goal, counting observations taken up to the end of 09A (end
July 2009).

In preparing the case for the Board we assessed progress to date, and
revisited the original science goals, which date from 2001. In most
cases the original science goals remain valid, which is not too
surprising as the parameter space covered by UKIDSS remains
unique. Here we summarise the current status of the surveys, discuss
the science goals, and detail the revised final plan. The full
resubmission case is available from the UKIDSS web site at
www.ukidss.org/science/science.html

CURRENT STATUS

The third column in Table 1 summarises the status of the surveys at
the end of semester 09A. The 09A data have been released as DR7 for
the LAS, GCS, and DXS. The latest GPS release is DR6, and DR7 is not
expected until the summer. The current UDS release is DR5,
corresponding to the 07B data. The next release will include both 08B
annd 09B data (note that the UDS is not observable in the A semester).

REVIEW OF SCIENCE 

LAS: The original goals included studies of SDSS galaxies in the
near-infrared, and the discovery of the highest redshift quasars, and
the coolest brown dwarfs. The LAS has been used to define the target
list for GAMA, the successor to the 2dF and SDSS galaxy redshift
surveys. So far five new z=6 quasars have been found, but none above
z=6.4, indicating that the decline in space density accelerates beyond
z=6. A a number of T dwarfs cooler than T(eff)=750K have been
discovered, extending the T dwarf sequence to cooler luminosities by a
factor five. Second-epoch J observations will extend the depth of the
cool T dwarf searches.

GPS: The main aims of the GPS are: (1) to provide a legacy database of
1-2 billion stars for all aspects of Galactic astronomy; and (2) to
advance our knowledge of star formation and Galactic structure by
detecting pre-main sequence clusters and mapping structure right
across the Milky Way. A large sample of newly discovered stellar
clusters is about to appear, reaching much larger distances than
samples from Spitzer and 2MASS, due to the superior spatial resolution
of UKIDSS. The first small area of second epoch K imaging has proved
very useful, with the discovery of several FU Orionis stars. Because
of the high spatial resolution, and the small extinction in the K
band, the GPS provides the primary reference catalogue within the
broader world-wide multiwavelength campaign to survey the plane of the
Milky Way.

GCS: The GCS goal is to measure the sub-stellar initial mass function
in several star clusters of a variety of ages and metallicities. The
shape of the IMF is beginning to emerge from these studies. A
compilation of results from the Pleiades, Upper Sco, and sigma-Orionis
clusters indicates that a mass function of log-normal form provides a
good functional fit. Cluster membership will be made more secure once
second epoch K imaging has been obtained (for proper motions), and the
completion of more clusters will allow a test of the universality of
the substellar IMF.

DXS: The DXS is particularly good for the detection of the 4000A break
in galaxies in the redshift interval 1<z<1.5. The space density and
clustering signal of old galaxies in this redshift interval provide a
strong test of theories of galaxy formation and evolution. The DXS has
obtained new results in particular on the clustering properties of the
extremely red object (ERO) and distant red galaxy (DRG)
populations. Many of the DXS fields have planned multiwavelength
coverage by e.g. Herschel, SCUBA-2, AMI, and LOFAR. The DXS will play
an important role in identifying accurate positions for sources
detected in these surveys.

UDS: The UDS has become the most important very deep survey in the
near infrared. Using photometric redshifts it has provided a
measurement of the evolution of the galaxy luminosity function over
the entire redshift range 0<z<4. The form of the luminosity function
at the upper redshift end becomes sketchy, and improving the overall
depth by 1 mag. will solidify these results. Thanks to the combination
of depth and area, it has produced the largest sample of galaxies at
redshift 5<z<6, and provided the best estimates of the bright end of
the galaxy luminosity function at these redshifts, crucial for
discerning evolution. Complementary multi-wavelength coverage at
X-ray, optical, mid-infrared, sub-mm, and radio, exists or is being
gathered.

REVISED SURVEY GOALS

LAS: The original goal comprised 4028 sq. degs in YJHK, with a second
pass in J, to be added after a minimum 2 year interval. In the
redesign we have decided to retain as much area as possible in YJHK,
but to reduce the second-epoch area to 2000 sq degs. The planned final
YJHK coverage is 3792 sq degs. The footprint of the LAS has changed
substantially from the original plan, and is plotted in Fig. 1. The
large equatorial spring block L1 remains. The large northern spring
block was abandoned after completion of only two small regions, L2a,
L2b. These cover 265 sq. degs and have already been completed in YJHK
and second epoch J. For scheduling reasons the northern spring block
was replaced by an equatorial autumn block, L3. Finally a new block
L4, has been added (200 sq degs, currently being surveyed), which
encompasses the Herschel-ATLAS northern field.

Fig. 1. Revised LAS footprint. The definition of the five blocks is
given in the text.

GPS: The original goals comprised JHK coverage of 1868 sq degs, with
observations in K at three epochs. A small area of 17 sq degs near the
Galactic centre has been trimmed, and the third epoch K coverage has
now been cut. In addition it was planned to image a 300 sq deg region
in Taurus-Auriga-Perseus with the narrow-band H2 filter, at three
epochs. Some 200 sq deg of H2 imaging has been completed, in a single
pass, and now no further H2 imaging is planned under the auspices of
the GPS.

To summarise, disregarding the H2 imaging, the revised GPS plan
comprises JHK imaging, plus a second epoch in K, of 1851 sq degs of
the Galactic Plane,as follows: (i)l=358 to l=10 at |b|<1.3, (ii) l=10
to 15 at |b|<2, (iii) l=15 to 107 at |b|<5.

GCS: The original GCS plan involved observations of ten clusters and a
total of 1069 sq degs coverage in 6 filter-passes, except for
Orion, which does not require second epoch K. The summed total of
filtersXarea is then 6260 sq.deg.  Following the 2006 review, further
work on the Hyades was canceled. The revised plan is then very simple:
to continue with the rest as planned. We will complete the necessary
multi-colour coverage, and carry out the second epoch K-band
measurements. This produces a revised total of 4835 sq.deg, which
amounts to 77% of the original 2001 design.

DXS: The original goal was to obtain deep J, K imaging over a region
of 12 tiles in each of the four DXS fields. Coverage to date has been
uneven: the SA22 field has proceeded the fastest, as there is little
competition near RA=22h, while the XMM-LSS field has proceeded the slowest
due to competition with the UDS. The revised plan is to complete all
12 tiles in the SA22 and ELAIS-N1 fields, 10 tiles in the Lockman
Hole, and 8 tiles in the XMM-LSS field. The original plan of covering
some 5 sq degs in H has been dropped.

UDS: For a variety of operational and scheduling issues the UDS has
proceeded the slowest of the surveys. Nevertheless in semester 09B
over 300h of observations (some 230h integration) were secured,
providing a very significant boost. By obtaining a further 700h
observations over 10B and 11B it is estimated that the final depths
reached (Vega, 5sigma, 2" aperture) will be J=24.3, H=23.2,
K=22.8. This is close to the original target goal in K, but falls
0.5mag short in H and K.