Correcting the analysis of "IR Anisotropies in Spitzer Goods Images..." by Cooray et al (2006) (astro-ph/0612609)

Abstract

We point out that in their analysis of the deep Spitzer images, Cooray et al (2006) perform Fourier transform on maps which have very few pixels left (only 20 to 30 percent). For such deeply cut maps one cannot reliably compute large-scale map properties using Fourier transforms. Instead the maps must be analyzed via the correlation function, C(\theta), which is immune to mask effects. We find, when computing C(\theta) for their maps, that removing ACS/HST galaxies does not lead to appreciable change in the correlation properties of the remaining diffuse emission. We then demonstrate with simulations that the power spectrum of CIB fluctuations prior to removal of the ACS galaxies reproduces C(\theta) in the maps from which the ACS galaxies have been removed. This implies that these galaxies cannot be responsible for the CIB fluctuations detected in Kashlinsky et al (2005, 2007), contrary to the claims of Cooray et al (2006).