Title:Variance of dust temperature and spectral index in Planck polarization data using spin-moment expansion

Abstract:Thermal dust is the major polarized foreground hindering the detection of primordial CMB B-modes. Its signal presents a complex behavior in frequency space inherited from the joint variation in our Galaxy of the orientation of magnetic fields and of the spectral properties of dust grains aligned with magnetic field lines. In this work, we present a new framework for analyzing the thermal dust signal using polarized microwave data. We introduce residual maps, represented as complex quantities, that capture deviations of the local polarized SED from the mean complex SED averaged over the sky mask. We make simple predictions that relate the values of the statistical correlation and covariances between the residual maps to the physical properties of the emitting aligned grains. The validation or falsification of these predictions provide valuable information on the nature of the dust signal. We confront our predictions with Planck data over a 97% mask excluding the inner Galactic plane. Despite its simplicity, our model captures a significant part of the statistical properties of the data. The spectral dependence of the covariances between residual maps is, for the Sroll2 version of the data, compatible with a dust model assuming only temperature variations rather than spectral index variations, while it is incompatible with either of them for the PR4 Planck official release. Our methodology can be used to analyze future high precision polarization data and build more accurate dust models to be used by the CMB community.