Formed in the crucible of the early universe, the light that composes the cosmic microwave background (CMB) contains an abundance of information about the formation and composition of the universe in both its temperature and polarization. The faintest signal from the CMB is the B-mode polarization signal, which has two potential sources: the gravitational lensing by intervening matter on small angular scales or anisotropies induced by inflationary gravity waves on large angular scales.

Inflationary B-modes have a magnitude of less than 100 nK, so noise is a constant adversary. Noise can originate from the atmosphere, the instrument itself, and galactic dust emission. To reach the sensitivity necessary for measuring B-mode polarization, experiments not only need highly sensitive detectors but also cutting-edge instrumentation and a skilled team of experimentalists and analysts.

To curtail atmospheric contamination, ground-based CMB experiments are conducted in some of the highest and driest places in the world: the Atacama Desert in Chile and the South Pole. High elevations decrease the amount of atmosphere the CMB photons must travel through before detection, and dry locations minimize the amount of water vapor in the air, which is one of the main culprits because it absorbs and emits at microwave frequencies. While these locations are ideal for ground-based observations, their remoteness makes daily operation a monumental effort, which would not be possible without strong local support from skilled local engineers and constant fuel supplies.

Members of CMB experiments work under difficult conditions to ensure the success of years of development and observation. At the beginning of an experiment, they must assemble a telescope, install its systems, and deploy its detector arrays. During nominal observations, they must perform telescope maintenance; plan and implement a 24-hour observation schedule; and run calibration measurements to ensure that the instrument and the detectors are thoroughly characterized. Supplies are scarce, so observers must find imaginative and durable solutions to malfunctions. They must have a deep understanding of all the experimental systems, including the computers, detectors, readout and biasing systems, cryogenic systems, and motion control systems. I myself have worked many months over the past few years on the Atacama B-mode Search (ABS) at its site 17,000 ft. up in the Atacama Desert.

Even in these extreme locations, atmospheric noise can still be an issue due to its nonuniformity, especially for observations at large angular scales. Sections of the unpolarized atmosphere can change on the timescale of minutes, causing fluctuations of tens of mK in the signal. Pioneered by the balloon CMB experiment MAXIPOL, a continuously rotating half-wave plate (HWP) can further reduce atmospheric noise by modulating the incident polarization signal. The rapid polarization modulation acts as a lock-in amplifier for the polarized signal and also mitigates systematic effects. ABS, the first ground-based telescope to use a continuously rotating HWP, showed that using a HWP to decrease atmospheric noise was extremely effective. ABS also demonstrated that a HWP can recover CMB polarization data at large angular scales, which are typically obscured from the ground by atmospheric fluctuations. Now other ground-based CMB experiments, including the Advanced Atacama Cosmology Telescope polarization receiver (AdvACT), are adding HWPs to their optics.

Noise can also come from the instrument itself. The optical elements of telescopes that couple the detectors to the sky must be optimized and well characterized in the field to minimize cross-polarization and the creation of spurious polarized signals. Experimentalists must also design and deploy electromagnetic shielding for the electrical systems and cables to minimize electrical noise. Future experiments will increase the packing density and number of detectors to increase sensitivity by decreasing statistical variation. To maximize the signal, CMB polarization experiments employ transition-edge sensor (TES) bolometers as detectors and read them out with a series of inductors and superconducting quantum interference devices, which provide low-noise amplification of the signals. (See Sam Ciocys’ story on page 13 for more information about TES detectors.) Improvements to these readout systems are crucial for future applications as arrays gain more detectors and the noise requirements become more stringent.

Some of the largest sources of instrumental noise are thermal. Vibrations within the telescope during motion must be minimized, as they can cause excess thermal noise. To further reduce thermal noise, the critical temperatures of the bolometers must be tuned to 100–500 mK for most experiments, which requires advanced cryogenic and readout technologies. In the past decade, cryogenic systems large enough to house many full detector arrays and even entire telescope optics (as in the case of ABS) have made great technological strides and can now reliably reach base temperatures below 100 mK for extended periods of time. Additionally, readout and feedback lines can carry thermal energy from warmer cryogenic stages to the detectors. To minimize the number of cryogenic wires, the detectors are multiplexed so that many detectors can be read out on a single wire.

Dust from our galaxy is a major source of signal contamination, as mentioned in Colin Hill’s story on page 9. To remove this contamination, the spectrum of the polarized dust emission must be well characterized, which requires observations at several frequencies. Today’s experiments usually link two TES bolometers to orthogonal polarizations at a single frequency. Future experiments will employ multichroic pixels, which use on-chip filters that allow several frequency bands to be detected by a single pixel. Upgrading to multichroic detector arrays necessitates the development of efficient wide-band elements, including lenses, filters, antireflective coatings, and HWPs. The Atacama Cosmology Telescope Polarization receiver (ACTPol) has already successfully deployed a 90/150 GHz multichroic array, which is currently observing, and AdvACT will use low-, mid-, and high-frequency multichroic arrays to observe with five frequency bands.

Competition in the CMB field can be fierce, as  teams search  for evidence related to everything from inflation to the curvature and content of the universe, and the growth of structure to the sum of neutrino masses. However, the competition between projects dissolves in the field. We observers face harsh conditions, including high UV exposure, low temperatures, low oxygen levels, long hours, and even llama traffic. Overcoming those obstacles requires camaraderie, with groups working together to push the boundaries of our scientific understanding of the universe.