Periodic Reporting for period 1 - BeyondPlanck (Beyond Planck -- delivering state-of-the-art observations of the microwave sky from 30 to 70 GHz for the next decade)
Berichtszeitraum: 2018-03-01 bis 2020-11-30
The detection of primordial gravity waves created during the Big Bang ranks among the greatest potential intellectual achievements in modern science. During the last decades, the instrumental progress has been nothing short of breath taking. However, from the latest ultra-sensitive experiments such as BICEP2 and Planck, it is clear that instrumental sensitivity alone will not be sufficient to make a robust detection of gravitational waves. Contamination in the form of astrophysical radiation from the Milky Way, for instance thermal dust and synchrotron radiation, obscures the cosmological signal by orders of magnitude, and second-order interactions between this radiation and the instrument characterization lead to a highly non-linear problem.
The BeyondPlanck collaboration has implemented a ground-breaking solution to this problem that allows for joint estimation of cosmological parameters, astrophysical components, and instrument characterization. The engine of this method is called Gibbs sampling, and this has now been applied to analysis of raw time-ordered observations observed directly by the instrument, opening up entirely new possibilities for modelling and understanding instrumental systematic effects. As a demonstration of the method, we have analyzed the Planck LFI observations, and delivered a new set of state-of-the-art Planck LFI maps that includes a completely new uncertainty description in the form of Monte Carlo samples. We have also combined these data with similar observations from WMAP (23-94 GHz), and produced the world’s best measurements of polarized synchrotron emission at CMB frequencies. These data products and mehods will play a central role in designing and optimizing future inflationary gravitational wave experiments.
The BeyondPlanck collaboration has implemented a ground-breaking solution to this problem that allows for joint estimation of cosmological parameters, astrophysical components, and instrument characterization. The engine of this method is called Gibbs sampling, and this has now been applied to analysis of raw time-ordered observations observed directly by the instrument, opening up entirely new possibilities for modelling and understanding instrumental systematic effects. As a demonstration of the method, we have analyzed the Planck LFI observations, and delivered a new set of state-of-the-art Planck LFI maps that includes a completely new uncertainty description in the form of Monte Carlo samples. We have also combined these data with similar observations from WMAP (23-94 GHz), and produced the world’s best measurements of polarized synchrotron emission at CMB frequencies. These data products and mehods will play a central role in designing and optimizing future inflationary gravitational wave experiments.
The main results that have been achieved during the course of this project include the following:
1) Development of a fully functional end-to-end Gibbs sampler for CMB observations, publicly released under an OpenSource GPL license.
2) Production of new state-of-the-art Planck LFI frequency maps at 30, 44 and 70 GHz, including a fundamentally new uncertainty characterization in the form of Monte Carlo samples.
3) A new model of polarized synchrotron emission between 30 and 70 GHz that includes both Planck and WMAP observations.
4) A new large-scale CMB likelihood that accounts for end-to-end error propagation.
These results have been presented to the cosmological community during a release conference on November 18-20, 2020, attended by more than 200 scientists. Both algorithms and results are described in a suite of 18 scientific papers, five of which have already been submitted, while the remaining 13 are still in preparation. All software and products are made publicly available through the project homepage (currently at http://beyondplanck.science; will be moved to https://www.beyondplanck.uio.no) and main products will also be made available through the Planck Legacy Archive.
The machinery developed during the course of this project will form the main algorithmic infrastructure for the ERC-funded Cosmoglobe project, which aims to combine a wide range of experiments into one global sky model.
1) Development of a fully functional end-to-end Gibbs sampler for CMB observations, publicly released under an OpenSource GPL license.
2) Production of new state-of-the-art Planck LFI frequency maps at 30, 44 and 70 GHz, including a fundamentally new uncertainty characterization in the form of Monte Carlo samples.
3) A new model of polarized synchrotron emission between 30 and 70 GHz that includes both Planck and WMAP observations.
4) A new large-scale CMB likelihood that accounts for end-to-end error propagation.
These results have been presented to the cosmological community during a release conference on November 18-20, 2020, attended by more than 200 scientists. Both algorithms and results are described in a suite of 18 scientific papers, five of which have already been submitted, while the remaining 13 are still in preparation. All software and products are made publicly available through the project homepage (currently at http://beyondplanck.science; will be moved to https://www.beyondplanck.uio.no) and main products will also be made available through the Planck Legacy Archive.
The machinery developed during the course of this project will form the main algorithmic infrastructure for the ERC-funded Cosmoglobe project, which aims to combine a wide range of experiments into one global sky model.
The single most important long-term legacy of BeyondPlanck will be the direct demonstration of a fundamentally new approach to CMB analysis, based on end-to-end Bayesian analysis. From a statistical point-of-view, this approach represents in many respects the ultimate method for analysing such experiments, and perhaps the most important reason for why it has never been done before, is simply because it was generally not believed to be technically feasible. After BeyondPlanck, a wide range of experiments will very likely explore similar approaches, both within and outside the CMB community.