Comparison of map-making algorithms for CMB experiments

¹ Helsinki Institute of Physics, PO Box 64, 00014 Helsinki, Finland e-mail: torsti.poutanen@helsinki.fi
² University of Helsinki, Department of Physical Sciences, PO Box 64, 00014 Helsinki, Finland
³ Dipartimento di Fisica, Università di Roma “Tor Vergata”, via della Ricerca Scientifica 1, 00133 Roma, Italy
⁴ IPAC, MS 100-22, Caltech, Pasadena, CA 91125, USA
⁵ INFN, Sezione di Roma II, via della Ricerca Scientifica 1, 00133 Roma, Italy
⁶ Computational Research Division, Lawrence Berkeley National Laboratory, Berkeley CA 94720, USA
⁷ Space Sciences Laboratory, University of California Berkeley, Berkeley CA 94720, USA
⁸ Department of Astrophysical Sciences, Princeton University, Princeton, NJ 08540, USA
⁹ Jet Propulsion Laboratory, 4800 Oak Grove Drive, Mailstop 169-327, Pasadena CA 91109, USA
¹⁰ Dipartimento di Fisica, Università di Milano, Via Celoria 16, 20131 Milano, Italy
¹¹ Institut d'Astrophysique de Paris, 98 bis boulevard Arago, 75104 Paris, France
¹² Service d'Astrophysique, DAPNIA, Centre d'Études de Saclay, 91191 Gif-sur-Yvette, France

Received: 8 February 2005
Accepted: 16 September 2005

Abstract

We have compared the cosmic microwave background (CMB) temperature anisotropy maps made from one-year time ordered data (TOD) streams that simulated observations of the originally planned 100 GHz Planck Low Frequency Instrument (LFI). The maps were made with three different codes. Two of these, ROMA and MapCUMBA, were implementations of maximum-likelihood (ML) map-making, whereas the third was an implementation of the destriping algorithm. The purpose of this paper is to compare these two methods, ML and destriping, in terms of the maps they produce and the angular power spectrum estimates derived from these maps. The difference in the maps produced by the two ML codes was found to be negligible. As expected, ML was found to produce maps with lower residual noise than destriping. In addition to residual noise, the maps also contain an error which is due to the effect of subpixel structure in the signal on the map-making method. This error is larger for ML than for destriping. If this error is not corrected a bias will be introduced in the power spectrum estimates. This study is related to Planck activities.