Weak lensing from space. II. Dark matter mapping

We study the accuracy with which weak-lensing measurements could be made from a future space-based survey, predicting the subsequent precision of three-dimensional dark matter maps, projected two-dimensional dark matter maps, and mass-selected cluster catalogs. As a baseline, we use the instrumental specifications of the SuperNova/Acceleration Probe ( SNAP) satellite. We first compute its sensitivity to weak lensing shear as a function of survey depth. Our predictions are based on detailed image simulations created using "shapelets,'' a complete and orthogonal parameterization of galaxy morphologies. We incorporate a realistic redshift distribution of source galaxies and calculate the average precision of photometric redshift recovery using the SNAP filter set to be Deltaz = 0.034. The high density of background galaxies resolved in a wide space-based survey allows projected dark matter maps with an rms sensitivity of 3% shear in 1 arcmin(2) cells. This will be further improved using a proposed deep space-based survey, which will be able to detect isolated clusters using a three-dimensional lensing inversion technique with a 1 sigma mass sensitivity of approximately 10(13) M. at z = 0.25. Weak-lensing measurements from space will thus be able to capture non-Gaussian features arising from gravitational instability and map out dark matter in the universe with unprecedented resolution.