Final Report Summary - ANISOTROPIC UNIVERSE (The anisotropic universe -- a reality or fluke?) In 2004 early reports based on WMAP (NASA's second-generation CMB satellite) observations were published by several groups, suggesting that the universe may exhibit tantalizing hints of large-scale anisotropy in violation with the Cosmological Principle. If correct and cosmological in origin, these reports held the promise of sheding new light on the very birth-moment of the universe itself, and they therefore received a great amount of interest in the field, both from observationalists and theorists. However, several critical voices were heard, questioning the origin of the reported signals: Could they be caused by instrumental systematics in WMAP? Or could they be due to contamination from the Milky Way, for instance in the form of radiation from vibrating interstellar dust particles? Or was is simply a statistical fluke? The primary goal of the current project was to address these questions quantitatively, employing a newly developed statistical framework called CMB Gibbs sampling. In particular, this framework was to be applied to the upcoming Planck observations, which would provide independent measurements of the same effect, and with much better foreground control than WMAP. And this work was a resounding success: During the course of this ERC-funded project, our group came to lead the entire astrohysical component separation effort in Planck, and produce the cleanest CMB maps ever produced. In the process, we also improved the overall understanding of the Planck instrument itself, and identified and corrected a wide range of systematic errors -- errors that otherwise might have compromised studies of large-scale cosmological anisotropy -- and we produced the most complete representation of the astrophysical sky to date, including the first spinning dust and polarized thermal dust maps. A very direct recognition of our successes was provided in the title of the first Planck 2013 press release, namely "Planck discovers an *almost* perfect universe": Through the course of this project, we successfully and convincingly managed to show that the initial reports of statistical anisotropy in the WMAP observations were *not* due to systematics or astrophysical foregrounds, but rather were due to real structures in the cosmological CMB sky. On the other hand, the pure statistical significance of the effect, after accounting for these uncertainties, remains at the tantalizing 2.5-4 sigma level, insufficient to claim the detection of new physics, but certainly most intriguing. As such, the hunt for new observables continues. While Planck polarization measurements are too noisy to provide useful input, next-generation experiments such as CoRE+, LiteBIRD and S4 all have the required power, and we are playing a central role in the component separation efforts of each of these experiments. This leadership role would not have been possible without the funding provided by the current ERC Starting Grant.
