Two-year Cosmology Large Angular Scale Surveyor (CLASS) Observations: A First Detection of Atmospheric Circular Polarization at Q Band

TL;DR

This paper reports the first detection of atmospheric circular polarization caused by Earth's magnetic field at Q band, using CLASS data, matching theoretical models and providing insights into polarized atmospheric emission for CMB observations.

Contribution

The study provides the first observational evidence of atmospheric circular polarization at Q band and validates atmospheric emission models for ground-based CMB experiments.

Findings

Detected atmospheric circular polarization at Q band.

Measured dipole pattern consistent with models.

Achieved amplitude agreement within 20%.

Abstract

The Earth's magnetic field induces Zeeman splitting of the magnetic dipole transitions of molecular oxygen in the atmosphere, which produces polarized emission in the millimeter-wave regime. This polarized emission is primarily circularly polarized and manifests as a foreground with a dipole-shaped sky pattern for polarization-sensitive ground-based cosmic microwave background experiments, such as the Cosmology Large Angular Scale Surveyor (CLASS), which is capable of measuring large angular scale circular polarization. Using atmospheric emission theory and radiative transfer formalisms, we model the expected amplitude and spatial distribution of this signal and evaluate the model for the CLASS observing site in the Atacama Desert of northern Chile. Then, using two years of observations at 32.3 GHz to 43.7 GHz from the CLASS Q-band telescope, we present a detection of this signal and compare the observed signal to that predicted by the model. We recover an angle between magnetic north and true north of $(-5.5 \pm 0.6)^\circ$, which is consistent with the expectation of $-5.9^\circ$ for the CLASS observing site. When comparing dipole sky patterns fit to both simulated and data-derived sky maps, the dipole directions match to within a degree, and the measured amplitudes match to within ${\sim}20\%$.