SHELS: Testing Weak-Lensing Maps with Redshift Surveys

Weak-lensing surveys are emerging as an important tool for the construction of "mass-selected" clusters of galaxies. We evaluate both the efficiency and completeness of a weak-lensing selection by combining a dense, complete redshift survey, the Smithsonian Hectospec Lensing Survey (SHELS), with a weak-lensing map from the Deep Lens Survey (DLS). SHELS includes 11,692 redshifts for galaxies with R <= 20.6 in the 4 deg² DLS field; the survey is a solid basis for identifying massive clusters of galaxies with redshift z lsim 0.55. The range of sensitivity of the redshift survey is similar to the range for the DLS convergence map. Only four of the 12 convergence peaks with signal to noise >=3.5 correspond to clusters of galaxies with M gsim 1.7 × 10¹⁴ M _sun. Four of the eight massive clusters in SHELS are detected in the weak-lensing map yielding a completeness of ~50%. We examine the seven known extended cluster X-ray sources in the DLS field: three can be detected in the weak-lensing map, three should not be detected without boosting from superposed large-scale structure, and one is mysteriously undetected even though its optical properties suggest that it should produce a detectable lensing signal. Taken together, these results underscore the need for more extensive comparisons among different methods of massive cluster identification.