Impact of Dust on Spectral Distortion Measurements of the Cosmic Microwave Background

TL;DR

This paper assesses how interstellar dust and other foregrounds impact the measurement of spectral distortions in the cosmic microwave background by the PIXIE mission, highlighting challenges and sensitivities.

Contribution

It quantifies the effects of dust, synchrotron, free-free, and CIB on CMB spectral distortion measurements, emphasizing the importance of foreground modeling and ancillary data.

Findings

PIXIE can marginalize over dust parameters to recover y and μ with increased uncertainty.

Including additional foregrounds degrades the measurement accuracy of y and μ.

Modeling choices for the CIB significantly bias the spectral distortion estimates.

Abstract

Spectral distortions of the cosmic microwave background (CMB) are sensitive to energy injection by exotic physics in the early universe. The proposed Primordial Inflation Explorer (PIXIE) mission has the raw sensitivity to provide meaningful limits on new physics, but only if foreground emission can be adequately modeled. We quantify the impact of interstellar dust on Compton $y$ and $\mu$ measurements by considering a range of grain size distributions and compositions constrained by theoretical and observational priors (Zelko & Finkbeiner 2020). We find that PIXIE can marginalize over a modest number of dust parameters and still recover $y$ and $\mu$ estimates, though with increased uncertainty. As more foreground components are included (synchrotron, free-free), the estimates of $y$ degrade, and measurement of $\mu$ in the range sometimes considered for the standard $\Lambda$CDM of $2\times10^{-8}$ becomes infeasible without ancillary low-frequency foreground information. An additional concern is dust absorption of the CMB monopole, a subtle effect that must be included. We quantify one form of model discrepancy error, finding that the error introduced by fitting our interstellar medium dust model with a modified blackbody is too large for CMB spectral distortions to be detectable. The greatest challenge may be the cosmic infrared background (CIB). We find that $\mu$ and $y$ are extremely sensitive to modeling choices for the CIB, and quantify biases expected for a range of assumptions.