Did the Universe Light Up Briefly in Its Dark Ages?
Did a brief burst of ionization light up the universe long before the first stars? This study finds today's oldest light mildly prefers one.
In one paragraph
The standard story says the universe's gas stayed neutral until the first stars switched on. This preprint adds a short, early burst of ionization to the standard model and fits it to cosmic microwave background, galaxy and supernova data. The data mildly prefer a burst, at about 2–2.5σ, adding about 0.012 to the optical depth and raising the total to about 0.073, but its timing depends on how Planck's data were processed. It also eases a puzzling preference for negative neutrino mass.
How did researchers test for an early burst of ionization?
When gas in the universe is ionized, free electrons scatter the cosmic microwave background, and the total effect is measured as the optical depth. In the standard model, nearly all of it comes from the first stars and galaxies at redshifts of about 10 to 20.
The authors asked whether a brief, earlier burst could also be present, for example from small dense regions that collapsed around redshift 100 to form stars or black holes. Without assuming a cause, they described the burst by how much optical depth it adds and when it switches off, anywhere from redshift 10 to 300.
They fitted this model to Planck, ACT DR6 and SPT-3G data, DESI galaxy measurements and Pantheon+ supernovae, using three alternative Planck processings.
What did the data reveal about the cosmic dark ages?
All data combinations mildly prefer a burst, with Δχ2 of 6.90 to 8.33, roughly 2–2.5σ. The preferred burst adds Δτburst ∼ 0.012 (0.001 < Δτburst < 0.022 at 95%), raising the total optical depth to τ ∼ 0.073.
The timing is less settled. HiLLiPoP favours z ∼ 45, CamSpec z ∼ 100 and plik z ∼ 165, a difference the authors trace to Planck polarization data on intermediate scales.
Removing the low-ℓ polarization data removes the preference, so it arises from a tension between those data and the rest. Allowing a burst also shifts the neutrino mass posterior median from −0.035 eV to −0.021 eV, easing but not removing the negative neutrino mass tension.
Why would early reionization matter for cosmology?
Several tensions trouble the standard cosmological model, including an apparent preference for a negative neutrino mass, which is physically impossible. A slightly larger optical depth has been proposed as a way to ease some of these tensions, and an early burst provides one way to get it while staying consistent with large-scale polarization data.
If such a burst were confirmed, it would point to unusual physics in the dark ages, such as early black holes or dark matter effects, at times no telescope can observe directly.
A fair reading is that this is a well-posed question with a weak but interesting answer, which upcoming experiments such as LiteBIRD and Simons Observatory should be able to test.
What are the limits of this early ionization study?
Read before you quote it
This is a preprint that has not been peer reviewed, and its limits are clear:
- Mild preference. About 2–2.5σ is far short of a detection.
- Processing-dependent timing. When the burst occurred depends on which Planck processing is used.
- No physical source. The model is phenomenological and does not identify a cause.
- Fixed shape. The burst width was fixed rather than fitted.
The study does not show that a burst of early ionization actually happened. It also does not identify what would cause one, whether early black holes, dark stars or dark matter, and it cannot rule out that the signal reflects quirks of data processing.
- It does not show that a burst of early ionization actually happened.
- It does not identify a source, such as early black holes or dark matter, for the extra ionization.
- The preference is mild, about 2–2.5σ, far short of a detection.
- When the burst happened is sensitive to which Planck processing is used, especially for multipoles 30 to 100.
- The model is phenomenological: it does not specify what physical source would cause the burst.
Takeaways
- saysCurrent cosmological data mildly prefer a brief burst of ionization in the dark ages, at about two to two and a half sigma.
- saysThe burst adds about point zero one two to the optical depth whatever the processing, but its timing depends on how Planck's data were processed.
- saysAllowing a burst eases the puzzling negative neutrino mass, but it's a hint, not a detection.
- explainsThe optical depth measures how much the cosmic microwave background was scattered by free electrons from ionized gas.
- interpretsA fair reading is that the dark ages may not have been completely dark, but better data are needed before drawing conclusions.
Ask the paper
In PaperSays you can ask this paper anything, by text or by saying “Hey Aiden”. Three questions listeners asked:
What were the cosmic dark ages?
The long period after the early universe cooled, before the first ordinary stars switched on and lit up the gas. Hear the full answer in the app →
Was there early reionization in the dark ages?
Current data mildly prefer a brief burst, at about 2–2.5σ, which is a hint rather than a detection. Ask in the app →
How could an early burst affect neutrino mass estimates?
Allowing a burst shifts the inferred neutrino mass toward less negative values, easing a known tension. Ask in the app →
Briefing reviewed by an editor before publication. The paper belongs to its authors; this page summarises it and links to the original. Paper licence: CC BY 4.0. Corrections: support@papersays.com.