[1004.5602] Large angle anomalies in the CMB
Posted: May 04 2010
This is a very clear review of the CMB large angle anomalies. In particular, the authors reply to the WMAP7 paper 1001.4758 claiming that there are no anomalies. The WMAP7 paper quoted the work 0909.2495 by Francis and Peacock arguing that the low quadrupole power and quadrupole-octopole alignment is due to the ISW effect of local structures and the paper 0911.5399 by Efstathiou, Ma and Hanson arguing that the large-angle correlation function is not anomalous.
The authors argue that if the local ISW explanation of Francis and Peacock is correct, this implies two new anomalies, namely an unusually low amplitude for the Sachs-Wolfe effect so that the ISW effect can dominate, and accidental cancellation between the SW and ISW anisotropy patterns.
Efstathiou, Ma and Hanson claimed to determine the large-angle correlationsfrom the cut sky without assuming statistical isotropy. The present authors argue that the method implicitly assumes statistical isotropy by way of independence of low and high multipoles, and note that their method which works directly in angular (as opposed to multipole) space does not involve such an assumption and is therefore more robust. (In fact, I cannot understand how one could possibly reconstruct large-angle statistics from a sky where 25% has been cut without making assumptions about about what is missing.)
The authors argue that if the local ISW explanation of Francis and Peacock is correct, this implies two new anomalies, namely an unusually low amplitude for the Sachs-Wolfe effect so that the ISW effect can dominate, and accidental cancellation between the SW and ISW anisotropy patterns.
Efstathiou, Ma and Hanson claimed to determine the large-angle correlationsfrom the cut sky without assuming statistical isotropy. The present authors argue that the method implicitly assumes statistical isotropy by way of independence of low and high multipoles, and note that their method which works directly in angular (as opposed to multipole) space does not involve such an assumption and is therefore more robust. (In fact, I cannot understand how one could possibly reconstruct large-angle statistics from a sky where 25% has been cut without making assumptions about about what is missing.)