In the earliest moments of the Universe, the first photons were trapped in a sea of ionized gas. They scattered randomly with the hot nuclei and electrons of the cosmic fireball, like tiny boats in a stormy sea. Then, about 370,000 years after the big bang, the Universe cooled enough for the photons to be free. After one last scattering, they could finally ply interstellar space. Some of them traveled across 14 billion years of space and time to reach Earth, where we see them as part of the cosmic microwave background. The remnant first light of creation.
The CMB is a central point of evidence supporting the Big Bang and the standard model of cosmology. By observing the scale of fluctuations within the CMB, we can measure things such as the shape of space, the distribution of matter and energy, and the rate of cosmic expansion. It’s that last one that has been troubling astronomers, thanks to the Hubble tension problem.
Astronomers have several ways to measure the Hubble parameter, the value of which tells us the rate of cosmic expansion. The methods generally fall into two types: those based on observations of the CMB, and those based on astrophysical phenomena such as supernovae. The problem is that these two types of methods don’t agree on the value. They even contradict each other, leading some astronomers to argue there must be something wrong with the standard model.
Polarization fluctuations within the CMB. Credit: SPT-3G Collaboration
Of the two types, the CMB method is the one with the most limited data. The best CMB observations we have come from space telescopes such as Planck, which measured fluctuations in CMB intensity. One solution to the tension problem would be to argue that the CMB observations are somehow biased. But new observations gathered by the South Pole Telescope (SPT) throw that idea out of the water.
Rather than measuring intensity fluctuations in the cosmic microwave background, the SPT observed variations in its polarization. All the CMB light we observe comes from a moment of last scattering, when photons scattered off an ion one last time before making the billion-year journey to reach us. When light is scattered, it is polarized relative to the distribution of ionized gas. So these observations are a truly independent measure of cosmic expansion.
Different modes of CMB polarization. Credit: Sky and Telescope
One challenge in working with polarized CMB data is that as the first light traveled through space, it interacted with matter, space, and time. Not only is the light red-shifted due to cosmic expansion, it is gravitationally lensed by galaxies, which changes the polarization. Some of the light scatters off interstellar gas, which gives a false polarization. Even ripples of gravitational waves can affect the light’s orientation. So the team looked at not just the raw polarization of the CMB, but also what are known as E-mode and B-mode polarization. Each of these is sensitive to different kinds of bias. For example, the E-mode is more sensitive to secondary scattering, while the B-mode is more sensitive to cosmic inflation and gravitational waves.
By combining and contrasting these polarization modes, the team was able to calculate a new value for the Hubble parameter. Since it isn’t based on intensity fluctuations, it is free of any bias in the space-based CMB observations. Based on their data, the team got a value of H<sub>0</sub> at 66.0–67.6 (km/s)/Mpc. This agrees with the intensity-based observations of WMAP and Planck, which found a value of 67–68 (km/s)/Mpc. In comparison, the astrophysical methods find a value of 73–75 (km/s)/Mpc.
This study confirms that earlier CMB observations are not biased. The Hubble tension is very real, and we currently have no clear way to resolve it.
Reference: SPT-3G Collaboration. “Cosmology From CMB Lensing and Delensed EE Power Spectra Using 2019-2020 SPT-3G Polarization Data.” arXiv preprint arXiv:2411.06000 (2024).