CFHTLS-SL2S Overview


The sky contains hundreds of thousands of foreground massive objects, either massive galaxies, Dark Halos, galaxy clusters (Miralda-Escude 1992, 1993a, 1993b and Marshall et al. 2005), which, when intercepting the line-of-sight of a galaxy in the background, produce multiple virtual images of that galaxy. They are, then, called Strong Lenses. Historically, strong lenses are classified into three varieties; giant arcs (actually multiple images of an extended object merged together; Soucail et al. 1987, Fort et al. 1988), multiple arclets (e.g. radial pairs; Fort et al. 1992) and gravitational rings (a rare occurence when both the foreground lens and the background source share the same line-of-sight; Warren et al. 1996 observed the first optical ring; also Cabanac et al. 2005).

Examples of giant arcs and rings in massive clusters, galaxy groups, and isolated ellipticals from the release T003 of CFHTLS (CADC restricted to CFHTLS community)

Because the optical depth of geometrical alignment events span 10-5 to 10-3 (Oguri et al. 2004, and reference therein), future spaceborne missions like DUNE (cf. pdf presentation by A. Refregier on DUNE) and SNAP will discover about a dozen of lenses per square degree. This field of studies will thrive in the era of large space telescopes (JWST). In this perspective a great deal of theoretical works have surged recently (see next section below).
The CFHTL-SL2S main objective is to prepare us before the avalanche of data, using the best datasets currently available from the ground; The DEEP and WIDE components of the CFHT Legacy Survey (CFHTLS). The CFHTLS-SL2S allows us to study central issues concerning Strong Lensing surveys, in particular a major hurdle for the use of scientific Strong Lensing samples: homogeneity of the sample and completeness.

Science rationale

Strong Gravitational lenses give a unique tool to probe directly the projected mass profile of any type of objects, such as galaxies (Kochanek et al 2001, Keeton 1998, 2001, McLeod 2001), compact groups or galaxy clusters (Pello et al. 1991, Mellier et al 1993, Fort & Mellier 1994, Kneib et al 1993, 1994, , 2000, 2003). Strong lensing has already led to remarkable premieres; first constraint on the upper value of the cosmological constant (Λ<0.9) for a flat universe (Kochaneck et al. 1992); first discovery of very high mass concentration in the center of galaxy clusters (Mellier et al. 93); The direct measure of the galaxy mass within the Einstein radius; The detection of the farthest objects in the universe (Pello et al. 2005, 2003, Kneib et al., 2004), to mention only the most spectacular. As impressive as past works might seem, it only represent a small token of the potential of Strong Gravitational Lensing, mostly because past samples (principally made of arcs in clusters and multiply imaged quasars) were very small. In addition, most discoveries were serendipitous in a wide range of wavelengths, resolutions, seeings, fields, hence sorting systematics and completeness of that sample was a formidable task. For instance, up to 2005, the largest sample homogeneous sample the Cosmic Lens All-Sky Survey (CLASS) comprising multiple images quasars, had only 22 strong lenses. A major survey of compact lenses detected spectroscopically on the SDSS is currently ongoing; The Sloan Lens ACS Survey (SLACS) already contains 28 targets and shall reach nearly 100. Due to the selection on the SDSS spectroscopic survey, the SLACS sample is restricted to deflectors at low redshift (z<0.3). In comparison, a preliminar scan of the first fields of the CFHTLS WIDE (172 deg2) shows that the final sample shall approach 200 ±30 giant arcs and at least twice that number of small rings and multiple arclets around massive galaxies and groups of galaxies. Hence, the CFHTLS potentially represents the largest database of Strong Lensing events in the coming 5 years.

The science realm the CFHTLS-SL2S can probe is vast, and our first priorities will be to characterize the deflectors (lenses) and the lensed sources. Constraints on cosmological parameters can certainly be approached by the CFHTLS-SL2S, but strong constraints down to levels of a few percents, which is required for post SN-WMAP dark energy cosmography, demand larger samples only available through space missions (DUNE or SNAP). However, our detection techniques are very generic and shall allows us to discover new types of Strong Lenses, such as dark lenses. In the short term, we are interested in the following axes of research:

  • The study of the masses of the collapsed CDM overdensities (halos) in the non-linear regime. The traditional methods; radial velocity dispersions, rotation curves and X-ray observations are all based on luminosities, wherease the General Relativity theory gives an elegant and simple interpretation to Strong lensing. The sole geometric configuration of a Strong Lens probes directly the total mass, the halo mass profiles (William, Navarro & Bartelmann 99) and ellipticities for ringlets ( De Filippis, Sereno & Bautz 2005, Gavazzi 2005, Oguri 2005, Broadhurst et al. 2005, Mandelbaum et al. 2005; sub-structure perturbative effect for conjugate images of arcs; Keeton,Gaudi & Petters 2005, 2003, Koopmans 2005, Rozo et al. 2005, Hagan, Ma, Kravtsov 2005, An 2005, Dobler & Keeton 2005). For galaxy clusters and groups, the interpretation of radial arcs coupled to weak lensing (Gavazzi et al. 2003) yields a good sampling of the central mass concentrations to be compared with numerical simulations. This is currently a hot topic, because the rare measures up to now indicate higher concentrations (2 to 3 times) than predicted, implying that the initial conditions of halo formations might not be fully understood (Lin & Ostriker 2002, 2003).
  • Rings, in turn, probe smaller gravitational potential, isolated halos of galaxies, within their optical radius. A sample of rings spanning a large redshift range z~0-1 shall give a unique tools to study the halo mass evolution over time (both Dark Matter Masses and the Mass-to-Light ratio; Ofek, Riz et Maoz 2003, Ofek et al. 2005, Chae 2005). A high-resolution follow-up with HST would allow us to find counter images near the lens center, and probe the very central mass distribution and, although more challenging, the super massive black hole influence (Rusin, Keeton, Winn 2005). Similarly, systems of arcs found around edge-on spirals would allow us to charaterize the shape of their halos up to the optical radii of the rings, a follow-up with HST would also be necessary in that case. New observations of M31 point towards a halo much more extended than previously thought (up to 150 kpc), constraints from strong lensing would gives a independent check on the actual total mass of such extended halos.
  • Remote galaxies may be magnified by more than an order of magnitude by a foreground massive deflectors (40 times for cB58, and 13 times for FORJ0332-3557), making them bright enough for spectroscopic follow-up to high redshifts. All redshift-record objects have been discovered on caustic lines of massive clusters ( Pello et al. 2005, 2003, Kneib et al., 2004), and gravitational magnification is the only way we can study the physical processes in situ at the highest redshifts, such as kinematics (e.g. simulations by Möller & Noordermeer 2005).
  • Last, the number of giant arcs is a robust statistical standard and is not well constrained by numerical simulations (Oguri 2002, Bartelmann et al., 2003, Dalal et al. 2003, Ho & White 2005, Li, Mao et al. 2005, Chen 2005). Interestingly enough, the preliminar number found in the CFHTLS per square degree greatly exceed the numerical predictions (by 3 to 10 times; Mellier priv. com. 2005). This needs to be robustly established by a better knowledge of completeness and systematic biases of the sample.