Data analysis for high-sensitivity cosmic microwave background observations

In recent decades, the cosmic microwave background radiation (CMB) has been one of the most important tools in cosmology. Due to its primordial origin, the CMB holds information about the early universe and how the universe evolved with time. Inferring cosmological information from the CMB is therefore essential for learning more about the universe. Our abilities to produce high-precision CMB measurements have progressed immensely over the years, which helped to constrain the standard cosmological model with remarkable accuracy. As CMB measurements improve, efforts to improve our analysis methods continue with it. The main aim of the work presented in this thesis is to continue this endeavour for improving our ability to extract information from CMB measurements. We first explore several filtering methods for lensing reconstruction, and also devise a new filtering step. We show the benefits of using an optimal filter for upcoming ground-based CMB experiments. We adopt our lensing reconstruction method to test how instrumental systematics may affect lensing reconstruction results of an experiment similar to the Simons Observatory (SO), and show how some of the resulting lensing biases might be mitigated. We continue by using our lensing reconstruction pipeline to present new lensing results from a recent release of CMB maps from the Planck collaboration which are more accurate on large scales compared to the previous Planck analysis method. We show how the uncertainty of different cosmological parameters benefits from the improved reconstruction accuracy. We conclude by looking into a different CMB effect — the effect of Rayleigh scattering on the CMB anisotropies. We demonstrate a possible pipeline for extracting the Rayleigh signal from multi-frequency CMB measurements, and forecast the ability of detecting the Rayleigh signal from an SO-like experiment