If you were to pop into a cosmology conference today, the chances are that you would see this image in at least one presentation. It is a striking snapshot of a cluster of galaxies acting as a gravitational lens: the cluster bends light from galaxies lying behind it and 'smears' the light to produce multiple images and giant arcs.

As pretty as their effects are, gravitational lenses are giving cosmologists a few headaches. For example, the observed incidence of giant arcs and their distance from the clusters' centres, which marks the size of features called Einstein rings, indicate that these clusters may have a stronger 'lensing' ability than expected in the framework of the currently accepted model of the cosmos. In a paper to appear in Astronomy & Astrophysics, Meneghetti et al. describe an analysis that advances our understanding of these systems (M. Meneghetti et al. Preprint at http://arXiv.org/abs/1103.0044; 2011).

The authors compared the lensing ability of a numerically simulated sample of clusters with that of a sample of well-characterized, X-ray-luminous clusters obtained by the MAssive Cluster Survey (MACS). In contrast to earlier studies, their simulations factor in elements known to affect lensing power — for example, the fact that the lenses are complex three-dimensional structures. They found that the simulated clusters produce 50% fewer arcs than do the observed MACS clusters, and that the median size of Einstein rings differs by 25% between the two samples. These are much smaller discrepancies between theory and observation than previously reported. But as the authors themselves concede, more data are needed to confirm their findings.

Credit: NASA, A. FRUCHTER & THE ERO TEAM [S. BAGGETT (STSCI), R. HOOK (ST-ECF), Z. LEVAY (STSCI)] (STSCI)