부산대학교 도서관

We analyse deep multi-colour Advanced Camera images of the largest known gravitational lens, A1689. Radial and tangential arcs delineate the critical curves in unprecedented detail and many small counter-images are found near the center of mass. We construct a flexible light deflection field to predict the appearance and positions of counter-images. The model is refined as new counter-images are identified and incorporated to improve the model, yielding a total of 106 images of 30 multiply lensed background galaxies, spanning a wide redshift range, 1.0$<$z$<$5.5. The resulting mass map is more circular in projection than the clumpy distribution of cluster galaxies and the light is more concentrated than the mass within $r<50kpc/h$. The projected mass profile flattens steadily towards the center with a shallow mean slope of $d\log\Sigma/d\log r \simeq -0.55\pm0.1$, over the observed range, r$<250kpc/h$, matching well an NFW profile, but with a relatively high concentration, $C_{vir}=8.2^{+2.1}_{-1.8}$. A softened isothermal profile ($r_{core}=20\pm2$\arcs) is not conclusively excluded, illustrating that lensing constrains only projected quantities. Regarding cosmology, we clearly detect the purely geometric increase of bend-angles with redshift. The dependence on the cosmological parameters is weak due to the proximity of A1689, $z=0.18$, constraining the locus, $\Omega_M+\Omega_{\Lambda} \leq 1.2$. This consistency with standard cosmology provides independent support for our model, because the redshift information is not required to derive an accurate mass map. Similarly, the relative fluxes of the multiple images are reproduced well by our best fitting lens model.
Comment: Accepted by ApJ. For high quality figures see http://wise-obs.tau.ac.il/~kerens/A1689