Overdispersed radio source counts and excess radio dipole detection

The source count dipole from wide-area radio continuum surveys allows us to test the cosmological standard model. Many radio sources have multiple components, which can cause an overdispersion of the source counts distribution. We account for this effect via a new Bayesian estimator, based on the negative binomial distribution. Combining the two best understood wide-area surveys, NVSS and RACS-low, and the deepest wide-area survey, LoTSS-DR2, we find that the source count dipole exceeds its expected value as the kinematic dipole amplitude from standard cosmology by a factor of $3.67 \pm 0.49$ -- a $5.4\sigma$ discrepancy.