Component separation for the Planck mission
For over 40 years the cosmic microwave background (CMB) has been the most important source of information on the global properties of the Universe and on its contents. In recent years, high accuracy experiments such as BOOMERanG and WMAP have provided a remarkably accurate determination of the key cosmological parameters. A further leap forward is expected from the Planck mission that will survey the whole sky with an unprecedented combination of frequency coverage, angular resolution and sensitivity. Planck will measure both temperature and polarization anisotropies of the CMB, providing also highly valuable data for a broad range of astrophysics, including solar system, Galactic, and extragalactic science.
G. De Zotti is the coordinator of the Component Separation activity for the Planck Low Frequency Instrument (LFI). This activity aims on one side at cleaning the CMB maps produced by Planck in the frequency range 30 - 860 GHz from the astrophysical signals superposed on it, and, on the other side, at reconstructing as accurately as possible each individual foreground component, which has its own astrophysical interest. The relevant components include Zodiacal light, Galactic free-free, synchrotron and dust emissions, extragalactic point sources (radio sources and dusty galaxies), and Sunyaev-Zeldovich effects from clusters of galaxies. Recently, increasing evidence has been accumulating that two kinds of Galactic dust may actually be present in the Planck frequency range: the usual thermal dust emission, dominating at wavelengths shorter than a few mm, and emission from very small fast spinning grains, peaking in the range 20 - 40 GHz. However, the physical origin of this additional component has not been unambiguously established yet, so it is frequently referred to as "anomalous emission". In the last two years our component separation work has been focused on the implementation of a new method, proposed by L. Bedini and E. Salerno and dubbed Correlated Component Analysis (CCA), exploiting a second order statistics to derive the spectral properties of each component present in the CMB maps and the relevant cross-covariances, taking also into account complementary information from other surveys. We have implemented and tested algorithms based on this technique and applied them to the Planck Sky Model, a realistic simulation of the sky developed by the Planck Working Group on Component separation. The application, performed in the framework of the so-called Planck Component Separation Challenge, aimed at assessing and comparing the performances of the different methods worked out within the Planck Consortium, demonstrated that the CCA is highly competitive for the separation of diffuse components. The technique was also applied to WMAP data. This study allowed us to assess the effect of uncertainties on foreground subtraction on recovered CMB maps and, in particular, on reported "anomalies" in the CMB power spectrum. It also allowed us to obtain the first, albeit preliminary, all-sky map of the "anomalous emission" (see Figure).
People: A. Bonaldi (Post-doc), G. De Zotti, M. Massardi (Post-doc)
Collaboration: the Planck consortium (see Link)
Publications: Bonaldi et al. (2007), MNRAS, 382,1791; Leach et al. (2008), A&A, 491,597
Link: Planck