Statistical and systematical errors in cosmic microwave background maps

Sky temperature map of the cosmic microwave background (CMB) is one of the premier probes of cosmology. To minimize instrumentally induced systematic errors, CMB anisotropy experiments measure temperature differences across the sky using paires of horn antennas with a fixed separation angle, temperature maps are recovered from temperature differences obtained in sky survey through a map-making procedure. The instrument noise, inhomogeneities of the sky coverage and sky temperature inevitably produce statistical and systematical errors in recovered temperature maps. We show in this paper that observation-dependent noise and systematic temperature distortion contained in released Wilkinson Microwave Anisotropy Probe (WMAP) CMB maps are remarkable. These errors can contribute to large-scale anomalies detected in WMAP maps and distort the angular power spectrum as well. It is needed to remake temperature maps from original WMAP differential data with modified map-making procedure to avoid observation-dependent noise and systematic distortion in recovered maps.