One in Five Early Galaxies Leaked 87% of the Light That Lit Up the Universe

For a few hundred million years after the Big Bang, the universe was filled with neutral hydrogen gas and was effectively opaque. Then something switched it on. Ultraviolet radiation from the first stars and galaxies stripped electrons off that hydrogen, converting it into the ionized state it remains in today — an event called cosmic reionization, the last major phase transition the universe has undergone. Cosmic microwave background and quasar measurements place it around redshift 6 to 8, roughly 600 million to 1 billion years after the Big Bang.

The long-running argument has been about who did it: a small number of galaxies leaking ionizing light very efficiently, or an enormous number of faint galaxies each leaking a trickle. A new analysis led by Emma Giovinazzo of the University of Geneva, posted to the arXiv preprint server on July 24, comes down hard on the first answer.

The team compiled spectra for 1,428 galaxies observed with the James Webb Space Telescope's NIRSpec instrument, spanning redshift 5 to 10 — about 500 million to 1 billion years after the Big Bang. For each galaxy they calculated two things: how many ionizing photons it produces, and how many of those actually escape. That second number is the hard one. The relevant radiation is Lyman continuum light, with wavelengths shorter than 912 angstroms, and most of it never gets out; it is absorbed internally by the galaxy's own gas and dust. Only the escaping fraction reaches the intergalactic medium and does any ionizing.

The population split cleanly in two. About 20% of the galaxies qualified as "strong leakers," with escape fractions above 10%. The remaining 80% were weak leakers. But that minority produced roughly 87% of all the ionizing photons that made it into intergalactic space — enough, on its own, to drive nearly the entire reionization of the universe. "This indicates that it is necessary to identify highly leaking galaxies to find the drivers of reionization," the team writes.

Which galaxies carry the load also changes with time. At lower redshift, around z = 5 to 6, bright galaxies dominate the ionizing budget. At higher redshift, above z = 6, faint galaxies contribute roughly as much as bright ones. Stacking the whole picture together, the team estimates reionization finished around redshift 5.8 — consistent with independent measurements from other methods, which is a meaningful check on the analysis.

The authors are unusually direct about their limits. Their numbers depend on modeling choices, and the spectra come from a single survey covering a small patch of sky. The step they call their "biggest caveat" is the extrapolation down to faint galaxies, which the data do not directly verify. Fixing that, they write, "would require many more extremely deep field observations with JWST or the upcoming Extremely Large Telescope."

If the result holds, it changes the search strategy. Instead of counting up every faint smudge in the early universe and summing their contributions, the work is to find and characterize the rare galaxies that were punching holes in the fog — and to figure out what made those particular objects so leaky when their neighbors were not.