Science

A Few Leaks Lit the Universe

By cpstestersofficial@gmail.com · August 5, 2026 · 4 min read

Title: Reionization driven by the few: the ionizing budget of galaxies at z=5-10 from JWST/NIRSpec

Authors: Emma Giovinazzo, Pascal A. Oesch, Anne Verhamme, Romain A. Meyer, Callum Witten, John Chisholm, Rui Marques-Chaves, Charlotte Simmonds

First Author’s Institution: Department of Astronomy, University of Geneva

Status: Submitted to Astronomy & Astrophysics [open access]

Intro to Reionization

Sometime during the Universe’s first billion years, hydrogen gradually went from being completely neutral to completely ionized (losing an electron). This transition is known as the epoch of reionization, beginning with the formation of the first stars and galaxies. Reionization is likely driven by star-forming galaxies and is expected to be “patchy”: bubbles of ionized gas form and grow around objects that leak the most photons.

The exact details of how reionization occurred are largely unknown. In the simplest patchy model, we can think of three dependencies that would affect how fast reionization happened:

1. The number of galaxies with the ability to produce high-energy ionizing photons (ρ_UV)
2. How efficiently galaxies can produce those photons (ξ_ion)
3. How many of those photons actually escape the galaxy and affect the intergalactic medium (f_esc)

**Figure 1:** A visual of what we know about reionization. The first stars and galaxies ionize the gas surrounding them, forming bubbles that expand. The bubbles grow until the entire universe is ionized (*Image Credit: NASA*)

Weak vs Strong Leakers

JWST has measured Quantity #1 and #2 decently well, but Quantity #3, called the escape fraction, is the wild card. It is impossible to measure directly since these photons are absorbed by neutral hydrogen and never reach us. Fortunately, all is not lost; when photons are absorbed, they create an emission line. This appears to us as galaxies with high escape fractions having weaker nebular emission lines than expected given the stars they contain. The authors of today’s paper examine 1428 sources observed with the James Webb Space Telescope (JWST) using the NIRSpec/PRISM instruments and place new constraints on the timeline of reionization.

Unsurprisingly, brighter galaxies tend to emit more ionizing photons overall since they contain more stars. The authors find a more interesting result at fixed brightness—galaxies with high escape fractions deliver ~30 times more ionizing photons into the intergalactic medium than galaxies with low escape fractions. The authors call these two groups strong leakers (f_esc > 10%) and weak leakers (f_esc < 10%), making up about 20% and 80% of the sample, respectively.

**Figure 2:** Plotting the sample by brightness (x-axis) and leakiness (y-axis) reveals two distinct populations. The large, dark cloud at the bottom contains most of the sample with weak leakers, and the small, light cloud at the top contains the rarer strong leakers. Notice that leakiness does not depend much on brightness, indicating that faint and bright galaxies are equally capable of being strong leakers. Also, there is a large vertical gap, meaning that strong leakers are more effective at delivering ionizing photons. (*Figure 4 in the paper*)

Who Won?

A natural question with the discovery of these two groups is which group actually reionized the universe … a small population of strong leakers or a large population of weak leakers? The authors resolve this by mapping the 1428 galaxies onto a complete population of early galaxies from other surveys. Then, they determine the neutral hydrogen fraction as a function of time and find that reionization ends at about z = 5.7 (~12.5 billion years ago). This result is consistent with measurements with Planck, which observed the Cosmic Microwave Background.

Giovinazzo et al. repeat this process, but instead test the strong and weak leakers separately. Ultimately, they find that strong leakers alone reproduce nearly the full reionization history, while weak leakers have barely any effect!

**Figure 3:** The fraction of hydrogen still neutral as the Universe ages (time runs right to left). The red line shows what the strong leaker minority could accomplish on their own; the blue line shows the same for the weak leaker majority. The red line fits much better with high-redshift observational data in grey (*Figure 9 in the paper*).

So reionization wasn’t a group effort. A small minority of galaxies (about 20%) supplied nearly 90% of the photons that reionized the Universe. This result could have impacts on how we observe reionization: perhaps we should look for rare leaky galaxies, not the many ordinary ones. To understand the drivers of reionization, we will have to measure f_esc with faint galaxies beyond current spectroscopic limits, eventually with the Extremely Large Telescope.

Astrobite edited by Munira Hoosain

Featured image credit: Jingchuan Yu, Beijing Planetarium

I am a Physics PhD student and NSF Graduate Research Fellow at Stanford University, working with Josh Frieman. I am interested in theoretical and observational cosmology, specifically in cross-correlations between the Cosmic Microwave Background and tracers of large-scale structure.

View all posts