a) To map lens systems whose images have complex internal structure. The distortions of the intrinsically complex structures introduced by lensing depend strongly upon the mass distribution in the lens, and can be revealed by high-resolution radio maps. Various theoretical modelling techniques (see below) will be used to deduce the (smooth) radial mass profiles of the lenses. Additionally evidence will be sought for effects of any mass substructure of the type predicted by most CDM simulations. These sub-structures could manifest themselves, for examples as images of radio jets that cannot be transformed into each other by using matrices derived from mass models with a simple radial dependence. Vital to the success of this quest is to produce the best possible simulations with a resolution to predict reliably the sub- structure expected within galaxy DM haloes. The groups in Cambridge, Potsdam and Shanghai are all leading exponents of this art. An objective of the proposal is to make the best simulations and, for the first time, use these to predict the ranges of lensing effects such halos would produce. The recent evidence (Jing, 2002, ApJ, 574, 538) indicates that the haloes are triaxial and the observational consequences of this need to be explored.
b) To develop new techniques that make optimum use of radio data and optical pictures. Wucknitz (Potsdam) has developed a much- enhanced version of the LENCLEAN algorithm which makes explicit use of the fact that there are multiple images and simultaneously produces an optimum map of the lensed images and a reconstruction of the unlensed object. In an outer loop, different model mass distributions can be tried and the one that produces the minimum residual visibilities chosen. There are further enhancements to be made and data on more lens systems to be collected (see above) and analysed with this technique. Bradač(MPIfR) has successfully used an alternative method for the same kind of task. One of the advantages of these new methods is that they produce robust error estimates. Optical pictures, too, often contain vital information on the mass distributions. One of our other objectives is to extend some of the underlying philosophy of LENSCLEAN to optical deconvolution.
c) To use microlensing to explore the make-up of DM haloes. Galaxies contain stars (and compact dark objects) that, due to their motions, change their geometry with time and produce observable, short-timescale, microlensing effects on already macrolensed images. The astrophysical importance is that the microlensing light curves contain unique information about the mass function of the stellar-mass compact objects doing the lensing and which make up part of the dark halo. In this way the nature of the dark matter is being probed. It is proving a powerful technique in our Galaxy. The aim is to obtain radio and optical microlensing light curves and compare with numerical simulations.
d) To measure lens galaxy velocity dispersions. The first evidence for dark matter came from measuring rotation curves and velocity dispersions for galaxies. Lensing is the only other way to obtain such information. Our aim is to combine the techniques and use large optical telescopes to obtain spatially resolved velocity dispersions for lensing galaxies. The complementary information gained from both techniques will be used to address open questions about galaxy dynamics.