Dust polarization and ISM turbulence

Abstract

Perhaps the most intriguing result of Planck's dust-polarization measurements is the observation that the power in the E-mode polarization is twice that in the B mode, as opposed to pre-Planck expectations of roughly equal dust powers in E and B modes. Here we show how the E- and B-mode powers depend on the detailed properties of the fluctuations in the magnetized interstellar medium. These fluctuations are classified into the slow, fast, and Alfv\'en magnetohydrodynamic (MHD) waves, which are determined once the ratio of gas to magnetic-field pressures is specified. We also parametrize models in terms of the power amplitudes and power anisotropies for the three types of waves. We find that the observed EE/BB ratio (and its scale invariance) and positive TE correlation cannot be easily explained in terms of favored models for MHD turbulence. The observed power-law index for temperature/polarization fluctuations also disfavors MHD turbulence. We thus speculate that the 0.1--30 pc length scales probed by these dust-polarization measurements are not described by MHD turbulence but, rather, probe the large-scale physics that drives ISM turbulence. We develop a simple phenomenological model, based on random displacements of the magnetized fluid, that produces EE/BB $\simeq2$ and a positive TE cross-correlation. According to this model, the EE/BB and TE signals are due to longitudinal, rather than transverse, modes in the random-displacement field, providing, perhaps, some clue to the mechanism that stirs the ISM. Future investigations involving the spatial dependence of the EE/BB ratio, TE correlation, and local departures from statistical isotropy in dust-polarization maps, as well as further tests of some of the assumptions in this analysis, are outlined. This work may also aid in the improvement of foreground-separation techniques for studies of CMB polarization.