CMB LENSING POWER SPECTRUM BIASES FROM GALAXIES AND CLUSTERS USING HIGH-ANGULAR RESOLUTION TEMPERATURE MAPS

• Published in: The Astrophysical journal Vol. 786; no. 1; pp. 1 - 14
• Main Authors: van Engelen, A, Bhattacharya, S, Sehgal, N, Holder, G P, Zahn, O, Nagai, D
• Format: Journal Article
• Language: English
• Published: United States 01-05-2014
• Subjects: ACCURACY; ALGORITHMS; ASTROPHYSICS, COSMOLOGY AND ASTRONOMY; BACKGROUND RADIATION; Bias; Cosmic microwave background; COSMOLOGY; Galactic clusters; GALAXIES; GALAXY CLUSTERS; GHZ RANGE; GRAVITATIONAL LENSES; LIMITING VALUES; Masking; MASS; NEUTRINOS; Reconstruction; RELICT RADIATION; RESOLUTION; RESPIRATORS; SIMULATION; SPECTRA; TELESCOPES; UNIVERSE; Antarctica, South Pole
• ISSN: 0004-637X; 1538-4357
• DOI: 10.1088/0004-637X/786/1/13

The lensing power spectrum from cosmic microwave background (CMB) temperature maps will be measured with unprecedented precision with upcoming experiments, including upgrades to the Atacama Cosmology Telescope and the South Pole Telescope. Achieving significant improvements in cosmological parameter constraints, such as percent level errors on [sigma] sub(8) and an uncertainty on the total neutrino mass of ~50 meV, requires percent level measurements of the CMB lensing power. This necessitates tight control of systematic biases. We study several types of biases to the temperature-based lensing reconstruction signal from foreground sources such as radio and infrared galaxies and the thermal Sunyaev-Zel'ovich effect from galaxy clusters. These foregrounds bias the CMB lensing signal due to their non-Gaussian nature. Using simulations as well as some analytical models we find that these sources can substantially impact the measured signal if left untreated. However, these biases can be brought to the percent level if one masks galaxies with fluxes at 150 GHz above 1 mJy and galaxy clusters with masses above M sub(vir) = 10 super(14) M sub([middot in circle]) percent level bias, we find that only modes up to a maximum multipole of l sub(max) ~ 2500 should be included in the lensing reconstruction. We also discuss ways to minimize additional bias induced by such aggressive foreground masking by, for example, exploring a two-step masking and in-painting algorithm.