Gravitational lensing in standard and non-standard frameworks as a probe for precision cosmology: Challenges and applications
Theses and Dissertations
Gravitational lensing has been identified as a critical cosmological tool in studying the evolution of large scale structure in the universe as well as the nature of dark matter and dark energy. One of the primary physical systematics of weak lensing due to large scale structure (cosmic shear) is the intrinsic alignment (IA) of galaxies, which poses a barrier to precision weak lensing measurements. Methods for identifying and removing its effects on cosmological information are key to the success of weak lensing survey science goals. We have expanded model-independent techniques to isolate and remove the IA contamination from the lensing signal. These self-calibration techniques take advantage of complementary survey information to self-calibrate the lensing signal, which along with unique lensing and IA geometry and separation dependencies, allow us to reconstruct the IA correlations at the level of the spectrum and bispectrum. We have demonstrated that the self-calibration approach can reduce the IA bias over most relevant scale and redshift ranges by up to a factor of 10 or more. This could reduce a potential 10-20% bias in some cosmological information down to the 1-2% level. The self-calibration techniques have the added benefit of preserving the IA signal, which itself provides additional information that can be used in studying the formation and evolution of large scale structure in the universe. We have also identified a new source of intrinsic alignment contamination in cross-correlations with cosmic microwave background lensing and proposed a method to calibrate it, and we explored the potential of future surveys to measure directly various 2- and 3-point intrinsic alignment correlations. Finally, we have investigated the use of exact anisotropic and inhomogeneous models in general relativity for large- and small-scale structures in the universe, developing the frameworks necessary to analyze gravitational lensing in such models, and have compared them to observations, identifying potential sources of bias. We have found, for example, that ignoring substructure level anisotropies in structures could bias the lensing convergence, shear, and kinematic mass estimates by up to 10% or more. We conclude by presenting a numerical code package for calculations in such exact anisotropic and inhomogeneous models.