Lensing and e-MERLIN
Overview of eMerlin Lens programme
PIs: Steve Serjeant (OU), Neal Jackson (Manchester)
The UK gravitational lensing community have agreed a co-ordinated strategy for the next generation radio facility, eMerlin, which will make ground-breaking progress in lensing research. Our proposed Legacy Survey is two-pronged: a major lens discovery programme, using Herschel and SCUBA-2 surveys (ranked Medium-High and High by STFC respectively); and a deep lens mapping programme, fed in part by our new lens discoveries. This programme has been agreed by the UK lensing community to be a world-leading strategy using unique UK facilities. In the next two sections we describe the programme.
The UK-led Herschel ATLAS is the biggest Open Time Key Project on Herschel (ranked Medium-High by STFC), and will conduct a wide-field (>500 deg^2) survey with the PACS and SPIRE instruments at 110-500 microns. Gravitational lensing is one of the survey's six major science themes. The steep submm source counts make this survey extremely prone to gravitational lens magnification bias, and we expect several thousand gravitational lenses in this survey (e.g. [WWW] Negrello et al. 2007.). Our source count models predict that the lens selection efficiency is extremely high (>30%), and after excluding local galaxies and blazars using existing radio and optical catalogues, the lens selection efficiency approaches an astonishing 100%. (Compare the 0.14% lens fraction in radio surveys.) A similarly high lensing fraction is predicted for SASSy, the all-sky legacy survey with SCUBA-2 (ranked High by STFC). In total, many hundreds of lenses are predicted in Herschel ATLAS and SASSy.
This will be the primary lens discovery project for eMERLIN, generating the largest and most distant sample of lenses to date. The project will play the role for eMERLIN that the pioneering CLASS survey had for MERLIN. The eMERLIN array is of central importance to the lens exploitation of the Herschel ATLAS survey and SASSy, providing essential confirmation of multiple imaging in fields at declinations too high for ALMA, and in equatorial fields well in advance of ALMA. A one-minute eMERLIN snapshot is enough to verify the source image structure. Confirmed lenses will feed into the deep eMERLIN lens mapping programme.
The lens sample will allow evolution of galaxy mass profiles to be studied out to an unprecedented redshift of 1.5, back to when the Universe was a mere 30% of its current age. Using well-proven techniques, we will separate the dark halo and baryonic component of each lens galaxy to directly observe the build-up of stellar mass and its interplay with dark matter over this period.
We will also reconstruct the lensed sources to study the properties of high redshift dust sources well below the confusion limit, a technique first used in the SCUBA surveys. With lens magnifications often in excess of 10, we will be able to detect sources at S(250 microns) ~ 4mJy, similar to the flux limit of the deepest Herschel PACS guaranteed time surveys, but with the advantage that the Herschel ATLAS sources will be selected from a much larger volume and thus represent a fairer sample of the universe. In addition, we will reconstruct images of the surface brightness structure of highly magnified sources to a significantly higher resolution than is possible with eMERLIN alone to investigate evolution of the morphology of these very distant systems.
The detailed study of gravitational lenses is important because it allows us to look at matter distributions, both baryonic and dark, in galaxies at a cosmologically significant redshift. Such studies can allow direct confrontation of observations with CDM theories, which are just beginning to make testable predictions at galaxy scales. Radio-loud lenses are particularly important in this respect, partly because radio fluxes are unaffected by microlensing (providing clean constraints on mass distributions) and partly because the lens galaxy does not emit at radio wavelengths, allowing us to see the central lensed image. As part of the gravitational lens legacy project, we intend to undertake a campaign to image the ~30 known radio gravitational lenses to microJansky sensitivity. Many of these were discovered by the CLASS survey, based at Jodrell Bank.
Lens systems are predicted to have a central image which results from passage of light through the maximum in the time delay surface, very close to the centre of the lensing galaxy. This is an extremely important measurement, because it probes the mass distribution about 10-20pc from the centre of the lens. We can thus work out the degree of central concentration of galaxies at redshift ~0.5, and study both the central stellar cusp and the central black hole in a way that is otherwise difficult. The more singular the potential, the fainter the central image, and so the flux density is sensitive to the slope of the inner cusp and the mass of the black hole. With luck, it should be possible to detect more than one image produced by central black holes in some objects.
These observations are difficult because the flux densities we expect are low, of the order of a few microJy (Keeton 2005). Only one secure detection has been made so far, and resulted in a Nature paper (Winn, Rusin & Kochanek 2003). Efforts with current telescopes have been unsuccessful, including our observations (Zhang et al. 2007) with the HSA, consisting of all the VLBA telescopes, Green Bank, Arecibo and the phased VLA. e-MERLIN should comfortably be able to beat this combination of the world's largest telescopes by a factor of 10, and conduct a census of central image flux densities; its resolution, between EVLA and VLBI, is ideally matched to the problem which requires resolution of ~50mas in most cases. This will uniquely allow us to do statistics of the potential wells at the centres of galaxies at cosmologically significant redshifts, and is important for understanding how these regions gain mass with cosmic epoch. Because the central region mass correlates well with bulge properties (Ferrarese & Merritt 2000) this in turn has implications for the formation of elliptical galaxies.
The second scientific aim which will be realised with these observations will be the study of the extended structures produced by multiple components in the lensed objects of the lens systems. Dalal & Kochanek (2002) studied the existing data on image flux densities in radio-loud lens systems, and used these to argue that substructure, resulting from CDM subhaloes of the type predicted by numerical simulations, was present in the lens galaxies. This is inferred because the image flux densities observed cannot be reproduced by a smooth galaxy mass model; radio flux densities are required for this deduction because they are not affected by microlensing or extinction. Much more powerful tests, and hence more direct confrontation with CDM models, can be achieved by detection of more structure in the images. This will have low surface brightness and has hitherto been invisible; an increase of a factor of 10 in sensitivity is needed for such studies, exactly what e-Merlin will provide.