Cosmological parameters and the WMAP data

Abstract

I discuss whether the standard cosmological models fit the WMAP data well enough to justify parameter estimation with standard assumptions. The observed quadrupole is low (but has significant foreground uncertainty) and drives weak evidence for theoretical models predicting low values, such as models with a running spectral index. Other more seriously outlying points of the WMAP power spectrum appear not to fit the expectations of simple Gaussian models very well. The effective temperature chi-squared is however acceptable on large scales. There also appears to be evidence for an anisotropic distribution of power, which taken together with the other points may indicate that either there is a problem with the WMAP data or that standard cosmological models are incorrect. These issues should be clarified before cosmological parameter extraction for the usual standard models can be trusted, and hint that maybe the CMB is more interesting than we imagined. I also discuss various systematic and analysis issues, and comment on various oddities in the publicly available first year WMAP data and code.