Reassessment of the dipole in the distribution of quasars on the sky

Abstract

We investigate recent claims by Secrest et al. of an anomalously large amplitude of the dipole in the distribution of CatWISE-selected quasars on the sky. Two main issues indicate that the systematic uncertainties in the derived quasar-density dipole are underestimated. Firstly, the spatial distribution of the quasars is not a pure dipole, possessing low-order multipoles of comparable size to the dipole. These multipoles are unexpected and presumably caused by unknown systematic effects; we cannot be confident that the dipole amplitude is not also affected by the same systematics until the origin of these fluctuations is understood. Secondly, the 50 percent sky cut associated with the quasar catalogue strongly couples the multipoles, meaning that the power estimate at ell=1 contains significant contributions from ell>1. In particular, the dominant quadrupole mode in the Galactic mask strongly couples the dipole with the octupole, leading to a large uncertainty in the dipole amplitude. Together these issues mean that the dipole in the quasar catalogue has an uncertainty large enough that consistency with the cosmic microwave background (CMB) dipole cannot be ruled out. More generally, current data sets are insufficiently clean to robustly measure the quasar dipole and future studies will require samples that are larger (preferably covering more of the sky) and free of systematic effects to make strong claims regarding their consistency with the CMB dipole.