This is the power spectral density (PSD) of a state-space model for WMAP temperature data sampled along the prime meridian. Samples are spaced uniformly every 0.1 deg. Each sampled temperature value is the temperature of the HEALPix pixel that is closest to the sample point on the meridian. The modeled data are 5-year V-band temperatures. The kq75 mask is used to exclude known galactic-foreground and point-source pixels. The state-space model is based on 2,925 temperature samples that survive the kq75 mask. The white-noise floor is consistent with the model’s autocorrelation function, shown here.

If we model the WMAP data as samples from an isotropic homogeneous process, then we can also use the autocorrelation function to compute a radial power spectrum, which is the Hankel transform of the autocorrelation function. Once we have the radial power spectrum in hand, we compute a rigorous estimate of the angular spectrum associated with the spherical harmonic analysis of the temperature field. The estimate is based on the flat-sky approximation to the spherical analysis at sufficiently large multipole indices, where we set spatial frequency f [cycles/degree] = L/360, with L the multipole index. The resulting angular power spectrum CL leads to the cosmology power spectrum, which is shown here.