The image distortions of high-redshift galaxies caused by gravitational light deflection of foreground clusters of galaxies can be used to reconstruct the two-dimensional surface mass density of these clusters. We apply an unbiased parameter-free reconstruction technique to the cluster CL0939+4713 (Abell 851), observed with the WFPC2 on board of the HST. We demonstrate that a single deep WFPC2 observation can be used for cluster mass reconstruction despite its small field of view and the irregular shape of the data field (especially for distant clusters). For CL0939, we find a strong correlation between the reconstructed mass distribution and the bright cluster galaxies indicating that mass follows light on average. The detected anti-correlation between the faint galaxies and the reconstructed mass is most likely an effect of the magnification (anti) bias, which was detected previously in the cluster A1689. Because of the high redshift of CL0939 (z_d=0.41), the redshift distribution of the lensed, faint galaxies has to be accounted for in the reconstruction technique. We derive an approximate global transformation for the surface mass density which leaves the mean image ellipticities invariant, resulting in an uncertainty in the normalization of the mass. From the non-negativity of the surface mass density, we derive lower limits on the mass inside the observed field. We can break the invariance transformation for the mass using the magnification effect on the observed number density of the background galaxies. Assuming a mean redshift of 0.8 and a fraction of 15 % (20 %) of cluster galaxies in the observed galaxy sample with 23<R<25.5 we obtain for the mass inside the field approximately 5 (7) x 10**{14}/ h_{50} solar masses which corresponds to M/L ~ 100 (140) h_{50}.

In a series of previous papers we have considered the reconstruction of the surface mass density of a cluster of galaxies from images of lensed faint background galaxies. We showed that the reconstructed surface mass density is not uniquely determined, but that there exists a global invariance transformation that leaves the shape of the images of the lensed galaxies unchanged. Because of this, only lower limits on the total mass of a cluster can be derived if no further informations besides image ellipticities are used. Throughout these papers we used the simplifying assumption that all sources are at the same redshift. In this paper we account for a redshift distribution of the faint galaxies, and in particular, some of these galaxies can lie in front of the cluster or can be cluster members. We show how the mass distribution of a cluster of galaxies can be obtained from images of these faint galaxies, if the redshift distribution of these galaxies is known. We demonstrate that for the reconstruction of non-critical clusters we need less information on the redshift distribution of the galaxies, i.e., we only need to know two or three moments of the distribution. We show that the mean mass density across the data field is still a free variable, i.e., there remains a global invariance transformation of the resulting mass density field. For non-critical clusters we can derive the transformation explicitly and it is similar to that derived previously for the case of a single redshift of the sources. We discuss several theoretical ideas to break the mass degeneracy; of those considered, we find that only the magnification effect on the number density of galaxy images can be used successfully in practice.

The weak distortions of high-redshift galaxies caused by gravitational light deflection near clusters of galaxies can be used to reconstruct the projected (two-dimensional) surface mass density of intermediate redshift clusters. This technique, pioneered by Tyson, Valdes & Wenk, and Kaiser & Squires, is reconsidered in the present paper, where we generalize the inversion equation found by Kaiser & Squires (KS) in several respect. Adopting a different smoothing procedure for the discreetly sampled data (individual galaxy images), we effectively reduce the shot noise in the KS procedure. In particular we show that the best density reconstructions are obtained if the smoothing scale is adopted to the `strength of the signal', which yields a better resolution near the center of the cluster where the distortions are strongest. Furthermore, we point out the importance of boundary effects and demonstrate their disastrous impact on rectangular data fields (CCDs) with large side ratio. Most important, however, is the generalization of the KS method to critical clusters, i.e., to such clusters which are capable of producing multiple images and giant luminous arcs. The corresponding modifications of the inversion procedure are severe; in particular, the resulting inversion equation is much more difficult to solve. As we pointed out in a previous paper (Schneider & Seitz), there exists a local degeneracy if the cluster is critical. We have developed an iteration procedure to solve the inversion equation, which we demonstrate to yield a very accurate

The distortion of images of faint, high-redshift galaxies by light deflection at foreground clusters of galaxies can be used to determine the (projected) mass distribution of the clusters. In the case of strong distortions, which lead to arcs in clusters, the position of the arc and/or its radius of curvature yields an estimate for the total mass inside a circle traced out by the arc. Weak distortions, which can be observed to much larger angular separations, can be used to determine the mass profile. In the case of weak distortions, an approximation which identifies the observed distortion with the shear produced by the lens can be made; this (linear) approximation breaks down, however, if one wants to probe the center of the cluster, i.e., approach the region within which giant arcs can be formed. The methods developed hitherto, the most advanced of which is due to Kaiser \& Squires, rely on the linear approximation and thus cannot yield reliable results for regions close to the center of the cluster. The purpose of this paper is to provide a theoretical basis to extend the Kaiser \& Squires method into the nonlinear regime, thereby making it more powerful, since the nonlinear distortions provide strong constraints on the mass profile, which in combination with the weak distortions should yield more reliable cluster inversions. We discuss the statistical properties of the observable image ellipticities and provide several methods to determine the local distortion by the lens from observed galaxy images, some of which rely