Measuring cosmological weak lensing using the Advanced Camera for Surveys on board the Hubble Space Telescope

Following from the theory of General Relativity, light-bundles are deflected and differentially distorted while passing through the gravitational potential of matter inhomogeneities. The gravitational lensing effect caused by the large-scale matter distribution in the Universe is termed cosmological weak lensing, and provides a powerful probe of cosmology. By studying the distortions which are imprinted onto the observed shapes of distant galaxies, the statistical properties of the foreground density field can be constrained free of assumptions on the relation between luminous and dark matter. Due to the weakness of the effect, it is challenging to measure and can only be detected statistically from large ensembles of coherently lensed galaxies. In addition, careful correction for systematic effects is required, first of all for the image point-spread-function (PSF). In this PhD thesis we present a detailed cosmological weak lensing analysis using deep high-resolution images from the Advanced Camera for Surveys (ACS) on board the Hubble Space Telescope (HST). Including data from the ACS Parallel Cosmic Shear Survey, the HST/GEMS Survey, and the HST/COSMOS Survey, this data set constitutes the largest survey used to measure cosmological weak lensing from space today. In order to achieve the high accuracy required for weak lensing studies, we developed several upgrades for the data reduction pipeline including careful image registration, improved bad pixel masks, and an optimised weighting scheme. We also perform a thorough investigation of the ACS PSF and develop a new correction scheme for its spatial and temporal variations, which are caused by thermal breathing of the telescope. We present numerous tests of our shear measurement pipeline using simulated images from the STEP Programme, and demonstrate that it achieves a relative shear-measurement accuracy better than 2% for ACS-like images. We perform the analysis of the ACS data in two steps, starting with a pilot study, in which we test the capabilities of ACS for cosmological weak lensing measurements with early parallel observations and the combined GEMS and GOODS ACS mosaic of the Chandra Deep Field South (CDFS, 0.22 deg2). We perform a number of diagnostic tests indicating that the remaining level of systematics is consistent with zero for the GEMS and GOODS data confirming the success of our PSF correction scheme. For the parallel data we detect a low level of remaining systematics which we interpret to be caused by a lack of sufficient dithering of the data. Combining our shear estimate of the GEMS and GOODS observations using 96 galaxies arcmin-2 with the photometric redshift catalogue of the GOODS-MUSIC sample, we determine a local single field estimate for the mass power spectrum normalisation σ8=0.59+0.13-0.17(stat)±0.07(sys) (68% confidence assuming Gaussian sampling variance) at a fixed matter density Ωm=0.24 for a ΛCDM cosmology, where we marginalise over the uncertainty of the Hubble constant and the redshift distribution. This estimate agrees only marginally with the WMAP-3 result of σ8=0.761+0.049-0.048 (Spergel et al. 2007) and is significantly below values found by recent ground-based surveys. From this discrepancy we conclude that the CDFS is subject to strong sampling variance with a significant under-density of compact foreground structures. This is consistent with a recent study by Phleps et al. (2007), who find a strong deficiency of red galaxies in this field. In a second step we perform a preliminary cosmological weak lensing analysis of the HST/COSMOS Survey (1.64 deg2). The significantly increased statistical accuracy reveals previously undetectable residual systematic errors indicated by a significant B-mode signal. So far we have not been able to unambiguously identify their origin, but note that similar indications for remaining systematics have been found in an independent analysis of the same data by Massey et al. (2007). Using only B-mode-free scales (>1' in the shear two-point correlation function), we find σ8 = 0.71±0.09 (68% confidence) from COSMOS for a flat ΛCDM cosmology and fixed Ωm=0.24, where the error includes the uncertainties in the redshift distribution, the Hubble constant, and the shear calibration, as well as a Gaussian estimate for sampling variance. This result is in excellent agreement with the WMAP-3 constraints, but is significantly below the estimates found by Massey et al. (2007). In addition to the cosmological weak lensing analysis we present a reconstruction of the projected mass in the COSMOS field, as well as first results from a weak lensing analysis of the HST/STAGES Survey targeting the galaxy super-cluster Abell 901/902. Furthermore, we briefly summarise ACS studies of galaxy clusters, which make use of the developed data reduction and weak lensing pipeline