Summary of the survey programme components

In this section we present an overview of each component of our proposed survey programme. Detailed science cases for each component are presented in subsequent chapters. Figure 1.2 shows the K-band depth and area for each component, and compares these to existing or imminent surveys. This figure makes the point that each of our proposed survey components is an improvement on the current situation by one or two orders of magnitude.


Figure 1.2. K-band limiting magnitude versus area for various near-IR surveys. In blue : the proposed UKIDSS survey components. In red : various other surveys, either completed or underway. Note that the envelope of blue points lies about 2 orders of magnitude to the right of the envelope of red points. In other words the UKIDSS surveys will cover approximately 100 times the area of current planned surveys at any particular depth. The dashed line shows the Euclidian number counts relation, which goes as 10-0.6K, i.e. the position of a survey relative to the line is a relative measure of the volume of space surveyed. This illustrates, for example, that for surveys for brown dwarfs 2MASS will detect many more than the KPNO survey, and the LAS is an order of magnitude bigger than 2MASS.


General considerations : shallow versus deep surveys. For many of our science goals it is essentially volume that matters, either in order to construct large samples of objects, or to maximise the chance of finding rare objects. In this case, surveying more area increases the volume in proportion to time t, whereas going deeper in a fixed area increase volume as t3/4. For such projects, then, a large shallow survey is the best. For some other projects, such as searching for high redshift quasars or clusters, the tradeoff between area and depth is more subtle, depending for example on redshift of interest, assumed cosmology and the shape of the luminosity function, but the net effect is that in these cases also large area is preferable. However too short an exposure leads to overheads dominating time. For an exposure of 40 seconds per sky position we expect an efficiency of 65% (see Figure 1.3, section 1.7), and this drops rapidly with shorter exposures. This leads to a natural basic shallow-survey unit, which gives a 5 sigma point source limit of K=18.4. Note that this is already ~25 times deeper than 2MASS. For shallow surveys then, the depth is effectively fixed, and the main decision comes in the area required to achieve the science goals, in competition with other components of the programme. Our high latitude Large Area Survey comes into this category. The proposed Galactic Cluster Survey is also effectively in this category as the clusters are so large.

In other cases, science goals have intrinsic depth requirements. For our Galactic Plane Survey we wish to map to large distances through the Milky Way, and to be able to detect specific types of object such as the bottom end of the IMF, or X-ray binary counterparts, to kpc distances. In addition, in a reasonable amount of time we can map the majority of the available Galactic Plane and so would like to go a little deeper. Some extragalactic goals have intrinsic depth requirements, for example to be able to detect typical galaxies at a given redshift. There are also intrinsic area requirements, for example to measure clustering on a given scale at a given redshift. In fact the crude depth required (K=21 for z=1 and K=23 for z=3) is reasonably clear, leading to the idea of Deep and Ultra-Deep surveys, but the time spent in depth versus multiple wavebands versus more area needs careful thought.

The Large Area Survey (UKIDSS-LAS). The Large Area Survey (LAS) aims to map as large a fraction of the Northern Sky as feasible (4000 square degrees) within a few hundred nights, which when combined with the SDSS, produces an atlas covering almost an order of magnitude in wavelength. Furthermore a huge number of objects will already have spectroscopic data from the SDSS project, making an unparalleled dataset. The basic shallow survey reaches J=19.6, H=18.8, K=18.4, but we also require a second pass in the J-band to detect proper motions of low mass objects to determine their kinematic distances, so that the final J depth is J=20.0. In addition we propose to use a new filter we have called Y, covering 0.97 to 1.07 microns, specifically designed to detect extremely high redshift (z=7) quasars, and to distinguish them from very low mass stars.

The Large Area Survey will produce a catalogue of a million galaxies with colours and spectra, and 10 million galaxies with photometric redshifts; will detect thousands of rich clusters out to z=1; will find fifty times more brown dwarfs than 2MASS, will probe to much fainter objects, and can get statistical ages and masses from kinematics; and will produce a complete sample of 10,000 bright quasars, including reddened quasars, using the K excess method.

We are particularly driven however by three especially exciting prospects. (i) A search for the nearest and smallest objects in the solar neighbourhood. Our survey is deep enough to detect brown dwarfs and free floating planets with as little as 5 Jupiter masses, and has a volume large enough to find of the order ten such objects at sub-parsec distances even for fairly pessimistic mass function extrapolations. (ii) The combination of IR and optical colours, and large expected proper motions, will allow us to find halo brown dwarfs if they exist, testing the universality of star formation processes, and the formation history of the Milky Way. (iii) We hope to find quasars at z=7 and to detect the epoch of re-ionisation. SDSS have found z=5-6 quasars by ''i' drop-out''. Beyond z=6 quasars become rapidly redder, indistinguishable from brown dwarfs in standard colours, and too faint to be in the SDSS z' survey. We therefore propose a survey in the new Y filter to match our JHK survey. Extrapolating popular evolution functions, we expect to find 10 quasars in the range z=6-7 and 4 in the range z=7-8.

The Galactic Plane Survey (UKIDSS-GPS). The Galactic Plane Survey (GPS) aims to map half of the Milky Way, to within a latitude of 5o. Given the declination constraints of UKIRT, we can survey l=15-107 and l=142-230. Confusion in the crowded star fields may prevent accurate photometry at l<40 or so (test data is being taken to assess this problem) but for now we assume we can map all the way in to l=15. The scientific goals require as large an area as possible, but in a reasonable time this gives an impressive depth - J=20, H=19.1, K=19.0. This is deep enough to see all the way down the IMF in distant star formation regions, to detect luminous objects such as OB stars and post-AGB stars across the whole Galaxy, and to detect G-M stars to several kpc. The K-band exposure is built up in three separate passes in order to detect variable objects. In addition, we propose to make a narrow-band molecular hydrogen survey in a smaller region (300 square degrees).

Like the high-latitude LAS, the GPS has its prime importance as a fundamental resource for future astronomy. We expect to detect 108-9 sources in total. However there are a number of immediately expected achievements. (i) We expect to increase the number of known Young Stellar Objects (YSOs) by an order of magnitude, and to measure the duration of the YSO phase versus mass and environment. (ii) Rare brief-duration YSO variables such as FU Orionis stars will be found in significant numbers for the first time. (iii) Thousands of evolved objects such as post-AGB stars and PNe will be found, a huge increase over previous samples. (iv) Star formation regions will be mapped throughout the Milky Way, measuring star formation efficiency versus Galactic radius, and estimating the overall star formation rate of the Galaxy. (v) Many stellar populations will be mapped to large distances through the Milky Way extinction, measuring scale height versus stellar type, and for the first time mapping the arms and warp, so far seen only in gas, in the stellar populations. (vi) Hundreds of X-ray binaries will be identified, and hundreds more such objects located in advance of X-ray measurements.

The Galactic Clusters Survey (UKIDSS-GCS). The Galactic Clusters Survey (GCS) aims to survey eleven large open star clusters and star formation associations, covering a total of 1600 sq. degs using the standard single pass depth plus a second pass in K for proper motions, giving a depth of J=19.6, H=18.8, K=18.8. These clusters are all relatively nearby and so are at intermediate latitudes and are several degrees across.

The GCS is the most targeted of our surveys, being aimed at the crucial question of the sub-stellar mass function. The stellar mass function is well determined down to the brown-dwarf boundary but more or less unknown below, and it is not known whether the IMF as a whole is universal or not. The mass limit reached varies somewhat from cluster to cluster, but is typically around ML~30 MJ. The number of objects expected to be detected in the range ML to ML+10MJ ranges from 100 to 3000 for the range of possible mass function models, showing how well we will constrain the IMF compared to current knowledge.

To find extreme objects - the very nearest examples, the lowest mass objects - the large area survey is better. But to measure the IMF, one wants to target the 30-100 MJ region, and to obtain masses one needs both a distance and an age, for which mapping clusters is ideal. This approach has of course already been started. Our survey improves on current studies not by going deeper but by collecting much larger numbers, and examining objects with a range of ages and metallicities, to examine the question of universality.

The Deep Extragalactic Survey (UKIDSS-DXS). The Deep Extragalactic Survey (DXS) aims to map 35 sq. degs of sky to depths of K=21, and J=22.5, in three separate large regions. (The sensitivities are for point sources; we expect the target galaxies to be roughly an arcsecond across and so have total magnitude detection sensitivities several tenths of a magnitude slower.) The theme of the DXS is a comparison of the properties of the Universe at 1.0<z<1.5 against the properties of the Universe today. The DXS will survey a similar volume at these redshifts to the 2dF and Sloan Digital Sky Survey (SDSS) volumes, and the near-infrared gives coverage of the same rest-frame wavelengths as SDSS. We will ensure that deep optical imaging will be collected for the same areas, giving photometric redshifts for nearly all the galaxies. As the final fields are selected over the next year or two, at the same time as XMM-Newton, GALEX, CFHLS, VIRMOS, and SIRTF plans are finalised, international multi-wavelength key areas will emerge.

There are four main scientific goals. (i) To measure the abundance of rich galaxy clusters at 1<z<1.5. This is the principal goal and defines the scope of the survey. The purpose is to obtain constraints on cosmological parameters that are complementary to measurements of the cosmic microwave background. Ultimately we hope to make an important contribution to reaching beyond the three-parameter Ho, omegam, lambda cosmology that describes the geometry and dynamics of the Universe, to obtain useful constraints on the dark energy equation of state parameter w=P/rho, and thereby to explore quintessence models. (ii) To measure galaxy clustering at z>1, and more specifically the evolution of bias, a key test of hierarchical models. The aim is to reach two magnitudes down the luminosity function at z=1, which requires a depth of K=21. To test the models we need to measure the correlation function on a series of scales, ideally out to 200 Mpc comoving, which is 3.5o at z=1.¹ To overcome cosmic variance, we need three such regions. (iii) To quantify the contribution from star formation to the cosmic energy budget, as a function of wavelength over the ultraviolet to far-infrared region. This will include studying a large sample of the rare but important Extremely Red Objects (EROs), and the precise measurement of their clustering properties. (iv) To measure the contribution of AGN to the cosmic energy budget, as a function of wavelength, by systematically identifying sources discovered in XMM X-ray and SIRTF mid-infrared surveys.

The Ultra Deep Survey (UKIDSS-UDS). The Ultra Deep Survey (UDS) aims to map 0.77 sq. degs of sky to a depth of J=25, H=24, and K=23, the first large volume map of the high redshift universe. The depth of K=23 is needed to reach galaxies at z=3 and greater. Many of the galaxies of interest, especially the reddened ones, will be undetectable in visible light until NGST, so that full JHK coverage is very important. The combination of depth and area will make this the most important existing archive of near-ir data for statistical studies of the early stages of galaxy formation. The inspiration for the UDS came from observing the success of the HDF as a public legacy database that has been used for all kinds of landmark science that had not been thought of at the time of conception.

There are three prime aims - the abundance of high-redshift ellipticals, the clustering of galaxies at z=3, and the relationship between EROs, ULIRGs, and AGN. We expect to detect 50,000 galaxies in total and 10,000 galaxies at z>2, mapping a region 100 Mpc comoving across and 2 Gpc deep (2<z<4), giving the first picture of large scale structure at z=3, and together with the DXS and local surveys, measuring the growth of structure with cosmic epoch. However as well as measuring the general field population, the key goal is simply to see whether giant ellipticals exist at z=3. The many successes of hierarchical models strongly suggests that no such large units existed at that time, but the ages of radio galaxies, and the local colour-magnitude relation strongly hint that they did. This then is one of the key questions of modern cosmology. If the density of giant ellipticals is similar to that of today, we should see of the order of 1000, whereas a traditional high density CDM model predicts about 30, and a lambda CDM model about 100, giving us a sensitive test of galaxy formation theories. It is tempting to identify EROs and/or SCUBA sources as spheroids in formation, in which case a matching post-starburst population of at least similar density should be present.

¹ Cosmological parameters omegam=0.3, lambda=0.7, Ho=70 are assumed throughout this proposal, unless otherwise stated.