A Canadian telescope pulled the universe’s hydrogen glow from 94 nights of data

Canada’s CHIME radio telescope has detected the faint 21-centimetre glow of cosmic hydrogen using 94 nights of observations recorded in 2019, a result published in The Astrophysical Journal in 2026. The detection reaches 12.5 sigma for the full signal and covers the era when the universe was roughly five billion years old. From it, the team infers that about 2 percent of the universe’s hydrogen was neutral atomic gas at that time.

Hydrogen’s 21-centimetre line and the foreground problem

Neutral hydrogen atoms emit a weak radio line at a wavelength of 21 centimetres. Because the universe expands, light from distant gas arrives stretched to longer wavelengths, so tuning a receiver to a lower frequency means listening further back in time. Instead of picking out single galaxies, the method maps the combined glow of all the hydrogen in each patch of sky.

The difficulty is that this glow is buried. Radio emission from the Milky Way and from bright extragalactic sources is orders of magnitude stronger than the cosmological signal, and human-made interference adds more. The team described new methods for flagging and masking interference, forming beams that behave consistently across frequency, and filtering foregrounds before averaging in time, a step the abstract says reduces spectral leakage that would otherwise smear bright contamination into the faint signal. CHIME’s detection used the band from 608.2 to 707.8 MHz, which the paper’s abstract converts to redshifts of about 1.01 to 1.34.

A stationary telescope that sees the northern sky daily

CHIME sits near Penticton, British Columbia, and is hosted by the National Research Council of Canada at the Dominion Radio Astrophysical Observatory. It looks nothing like a conventional dish. Its half-pipe reflectors stay fixed while the rotation of the Earth sweeps the sky overhead, and the instrument surveys the whole northern sky every day, according to the University of British Columbia release.

That daily repetition is what makes archival data valuable. Nearly seven years of recordings now exist, and the 94 nights from 2019 are a slice of them.

A 12.5 sigma auto-power detection, not a cross-correlation

The distinction matters. Earlier hydrogen intensity-mapping detections generally relied on cross-correlation, multiplying a radio map against a catalogue of galaxies whose positions were already known so that the shared structure stands out. CHIME’s result is an auto-correlation: the telescope’s own map is compared with itself to measure how the hydrogen signal varies across space. The preprint states that the detection does not depend on external galaxy surveys, and the university release calls it the first standalone detection of the signal by CHIME.

The measured quantity is a power spectrum, a statistical summary of how clumpy the hydrogen is at different scales, over wavenumbers from 0.4 to 1.5 h per megaparsec. It is not an image of the cosmic web. The full spectrum is detected at 12.5 sigma, and two independent sub-bands centred near redshifts 1.08 and 1.24 come in at 8.7 and 9.2 sigma, which guards against a single flawed stretch of spectrum producing the result.

Arnab Chakraborty, a postdoctoral fellow at the University of Toronto who first proposed the finding, said the team worked very hard to convince itself that the signal was not a false alarm, and the group then spent more than a year testing it. Mark Halpern, a University of British Columbia physics and astronomy professor and CHIME’s principal investigator, said: “This is a completely new technique for probing the cosmos.”

The two percent neutral-hydrogen fraction and dark energy

The two numbers in the release are different kinds of result. The signal strength is measured. The neutral fraction is inferred by comparing that strength with a theoretical model, which the first paper deferred to a companion, Interpretation of 21 cm Autopower Spectrum Measurement at z ~ 1. Shabbir Shaikh, a postdoctoral fellow at Arizona State University, said the data indicate that roughly two percent of the hydrogen in the universe was in neutral atomic form at this time, a value the team describes as broadly consistent with other measurements.

The detection itself is documented in the journal paper on 21-centimetre autocorrelation at redshift near 1. The dark energy link is about geometry. Hydrogen maps trace how matter is spread across space and how the expansion has changed, so they can test competing explanations for dark energy independently of other methods. Simon Foreman, an assistant professor at Arizona State University, said the result opens a new window on the universe.

The cost argument is the practical one. The release says the approach is a fraction of the price of galaxy-survey methods, which can run millions of dollars higher, because it reuses a telescope that is already scanning the sky. The team aims next to push the technique to an era when the universe was about three billion years old. Dark energy constraints from this measurement have not yet been published; that step depends on the longer data set and the interpretation work still to come.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview