Exploring the discrepancy between Planck PR3 and ACT DR4

Abstract: We explore the scales and the extent of disagreement between $Planck$ PR3 and Atacama Cosmology Telescope (ACT) DR4 data. $Planck$ and ACT data have substantial overlap in the temperature anisotropy data between scales corresponding to multipoles $\ell\simeq 600-2500$ with complementing coverage of larger angular scales by $Planck$ and smaller angular scales by ACT. Since the same cosmology should govern the anisotropy spectrum at all scales, we probe this disagreement in the primordial power spectrum. We use a parametric form of power law primordial spectrum that allows changes in the spectral tilt. We also reconstruct the primordial spectrum with a non-parametric method from both $Planck$ and ACT temperature data. We find the disagreement exists within scales 0.08 - 0.16 ${\rm Mpc}^{-1}$ where ACT temperature data prefers a scale invariant/blue spectrum. At scales larger and smaller than this window, ACT data strongly prefers a red tilt, which is consistent with $Planck$. This change in the spectral tilt can be identified in the ACT data at 2$\sigma$ C.L. without using $Planck$ data, indicating that the tension is driven by different preferences for tilts within the ACT data. The addition of $Planck$ data up to intermediate scales ($\ell\le650$) increases this significance to 3$\sigma$. Given the large overlap between $Planck$ and ACT within 0.08 - 0.16 ${\rm Mpc}^{-1}$ and considering the internal consistency between different $Planck$ temperature and polarization spectra, the scope of new physics as a solution to the tension remains limited. Our results -- a strong preference for an intermediate transition in spectral tilt and the variation of this preference in different data combinations -- indicate that systematic effects can be misperceived as new physics emerging from different non-standard cosmological processes.