Improving the Planck sky maps through Bayesian component separation

The Planck satellite has provided a multitude of data images of the full microwave sky, or sky maps, since it saw first light in 2009. Its observations of the Cosmic Microwave Background (CMB) data has been paramount in the estimation of cosmological parameters vital in many branches of physics and astronomy. Although the processing of this information is coming to an end, the data still exhibit significant systematic effects coupled with foreground contamination, which impairs our ability to determine these parameters accurately. In this thesis, we provide an overview of the Planck data analysis process with emphasis on Bayesian component separation methods for foreground removal. Furthermore, we seek to improve upon a new set of sky maps provided by Reijo Keskitalo at Lawrence Berkeley Laboratories. We do this by applying component separation in order to reveal systematic effects, which are subsequently corrected during map-making. This process is then repeated until systematic errors and foregrounds are suppressed to a satisfactory extent. Lastly, we present the results of our analysis with the application of the latest generation sky maps. These maps were produced as a result of the efforts described in this thesis. We conclude that the new sky maps exhibit a significant reduction in instrumental errors, in comparison to the current state-of-the-art.