Is the lack of power anomaly in the CMB correlated with the orientation of the Galactic plane?

Abstract: The lack of power at large angular scales in the CMB temperature anisotropy pattern is a feature known to depend on the size of the Galactic mask. Not only the large scale anisotropy power in the CMB is lower than the best-fit $\Lambda$CDM model predicts, but most of the power seems to be localised close to the Galactic plane, making high-Galactic latitude regions more anomalous. We assess how likely the latter behaviour is in a $\Lambda$CDM model by extracting simulations from the {\it Planck} 2018 fiducial model. By comparing the former to {\it Planck} data in different Galactic masks, we reproduce the anomaly found in previous works, at a statistical significance of $\sim 3 \, \sigma$. This result suggests the existence of a bizzarre correlation between the particular orientation of the Galaxy and the lack of power anomaly. To test this hypothesis, we perform random rotations of the {\it Planck} 2018 data and compare these to similarly rotated $\Lambda$CDM realisations. We find that, among all possible rotations, the lower-tail probability of the observed high-Galactic latitude data variance is still low at the level of $2.8 \, \sigma$. Furthermore, the lowering trend of the variance when moving from low- to high-Galactic latitude is anomalous in the data at $\sim 3\,\sigma$ when comparing to $\Lambda$CDM rotated realisations. This shows that the lack of power at high Galactic latitude is substantially stable against the "look elsewhere" effect induced by random rotations of the Galaxy orientation. Moreover, this analysis turns out to be substantially stable if we employ, in place of generic $\Lambda$CDM simulations, a specific set whose variance is constrained to reproduce the observed data variance.