Free access article

Issue A&A Volume 439, Number 3, September I 2005 Page(s) 901 - 912 Section Cosmology (including clusters of galaxies) DOI 10.1051/0004-6361:20040322

A&A 439, 901-912 (2005)
DOI: 10.1051/0004-6361:20040322

Trouble for cluster parameter estimation from blind SZ surveys?

N. Aghanim¹, S. H. Hansen² and G. Lagache<sup>1

1</sup>  IAS - CNRS, Bâtiment 121, Université Paris Sud, 91405 Orsay, France
    e-mail: aghanim@ias.u-psud.fr
²  University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland

(Received 24 February 2004 / Accepted 9 March 2005)

Abstract
The Sunyaev-Zel'dovich (SZ) effect of galaxy clusters is a tool to measure three quantities: Compton parameter, electron temperature, and cluster peculiar velocity. However, a major problem is non-removed contamination by astrophysical sources that emit in the SZ frequencies. This includes interstellar dust emission, infra-red (IR) galaxies, and radio sources in addition to primary Cosmic Microwave Background (CMB) anisotropies. The three former contaminations induce systematic shifts in the three SZ parameters. In this study, we carefully estimated, both for a large beam experiment (namely Planck Surveyor) and a small beam experiment (ACT-like), the systematic errors that result if a fraction of the expected levels of emission from dust, IR galaxies, and radio sources remains non-removed. We found that the interstellar dust emission is not a major contaminant for the SZ measurement. Unfortunately, the IR and radio source-induced systematic errors may be extremely large. In particular the intra-cluster temperature and peculiar velocity will be determined inaccurately for Planck and ACT-like experiments, if only the frequency dependences are used for the cleaning. The Compton parameter is also affected by the astrophysical contaminations. The systematic errors in this case were a factor of 2 to 5 times larger than the expected statistical error-bar for Planck. For the ACT-like experiment, the statistical error-bars were larger than in the case of Planck by a factor of about 5, and therefore the systematic shifts remain within about  of the statistical errors. We have thus shown that the systematic errors due to contaminating astrophysical emissions can be significantly larger than the statistical errors, which implies that future SZ surveys aiming at measuring cluster temperatures and peculiar velocities will not be able to do so on their own without including additional information like cluster shapes or follow-up observations.

Key words: cosmology: cosmic microwave background -- galaxies: cluster: general

© ESO 2005

What is OpenURL?