Isotropy Test with Quasars Using Method of Smoothed Residuals

Abstract

To assess the significance and scale dependence of anomalous large scale modes in the CatWISE quasar data, we generate smoothed number density fields on the sphere and study their extreme values -- maximum, minimum, maximum antipodal difference. By comparing these summary statistics to those obtained from random isotropic realisations of the data, we determine the statistical significance of large scale modes as a function of smoothing scale. We perform our analysis using five different versions of the data -- the original quasar map, the maps after separately subtracting the ecliptic bias and the CMB dipole, the map obtained after subtracting both, and the map after subtracting the ecliptic bias and anomalous dipole inferred in \cite{Secrest2021}. We find that the ecliptic-corrected, CMB dipole-removed map exhibits large scale modes that are in tension with random realisations of the data (p-values $p \lesssim 10^{-4}$), over a wide range of smoothing scales $π/8 \leq δ\leq π/2$. The most prominent feature in the data is an under-density in the southern galactic plane at $(b,\ell) = (-31^\circ,78^\circ)$, which reaches its highest statistical significance when smoothed on scales $δ= π/6$ ($p = 1.2 \times 10^{-6}$). Notably, the minima statistics align with the maximum antipodal difference statistics, whereas the maxima do not. This suggests that the observed dipole-like behavior in the data is primarily driven by the under-density in the southern sky. The ecliptic corrected, anomalous dipole subtracted map reduces the significance of any residual anisotropic features, but an under-density in the south sky persists with p-value $p =0.0018$.