CHIME picks out distant hydrogen's radio glow from 94 nights of its own data

Cross-correlation had a control built in. Only the part of CHIME's signal that lined up with galaxies catalogued by other telescopes survived the comparison, so noise in CHIME's own data had nothing to match ^[3]. Working alone removes that check. The team spent more than a year testing the result before concluding it came from the universe at about 5 billion years old ^[7]. The paper appears in The Astrophysical Journal ^[2]. "We worked very hard to convince ourselves that this wasn't a false alarm," said Arnab Chakraborty, a University of Toronto postdoctoral fellow and co-author who first proposed the finding ^[9]. "After all the tests, the signal remained," he said ^[9].

The galaxy route has limits of its own. Surveys see only the regions hot and dense enough to form stars, and the release says they cost millions of dollars more ^[4]. Hydrogen's glow comes from the gas itself. "Its faint radio emission acts like a cosmic tracer, revealing how matter is distributed across space," Chakraborty said ^[11]. According to the release, mapping that combined glow lets CHIME work at larger scales and further back in time, without depending on anyone else's results ^[5].

CHIME was built to map hydrogen so astronomers could calculate the universe's expansion and investigate dark energy ^[17]. The release says this result paves the way for a faster, less expensive study of dark energy and lets CHIME test conflicting theories independently ^[12]. It does not report a detection significance, a cost for CHIME's approach, or any constraint on the expansion rate. What the team has measured so far is hydrogen. "By measuring how that hydrogen is distributed and clustered, CHIME gives us a new way to test our understanding of how galaxies form and evolve," said Shabbir Shaikh, an Arizona State University postdoctoral fellow and co-author ^[10].

The sample is small. CHIME, near Penticton, British Columbia, maps the whole northern sky every day ^[13]. This signal comes from 94 nights observed in 2019 ^[8]. Seven years of nightly mapping would be about 2,555 nights, and 94 nights is roughly 3.7% of that, if the telescope observed every night ^[1].

I'd expect the follow-up to be judged on whether the signal strengthens as more of those nights are added. Simon Foreman, an assistant professor at Arizona State University and co-author, put the weight on the method. "By actually showing that the technique works in practice, we've opened up a whole new window on the universe," he said ^[15].