🔭 astrophysics

The THESAN project: Lyman-alpha emitters as probes of ionized bubble sizes

Using THESAN radiation-hydrodynamics simulations, this study demonstrates that Lyman-alpha emitters serve as effective statistical tracers of ionized bubble sizes during the Epoch of Reionization, particularly before the midpoint, by revealing strong correlations between Lyman-alpha properties and bubble dimensions that can be used to interpret current and future JWST and narrow-band surveys.

Original authors: Meredith Neyer (MIT), Aaron Smith (UT Dallas), Mark Vogelsberger (MIT), Luz Ángela García (ECCI), Rahul Kannan (York), Enrico Garaldi (IPMU), Laura Keating (Edinburgh)

Published 2026-04-17

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Meredith Neyer (MIT), Aaron Smith (UT Dallas), Mark Vogelsberger (MIT), Luz Ángela García (ECCI), Rahul Kannan (York), Enrico Garaldi (IPMU), Laura Keating (Edinburgh)

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the early universe as a giant, dark room filled with thick, invisible fog. This fog is made of neutral hydrogen gas. For a long time, this fog was so dense that light couldn't get through it. Then, the first stars and galaxies turned on like lightbulbs. Their intense energy started to burn holes in the fog, creating clear, ionized "bubbles" of space around them. Over time, these bubbles grew, merged, and eventually cleared the entire room. This process is called the Epoch of Reionization.

The problem for astronomers is that we can't see the bubbles directly. We can only see the lightbulbs (the galaxies) and try to guess how big the clear space around them is.

This paper, written by Meredith Neyer and her team using a super-computer simulation called TheSan, is like a massive "virtual reality" experiment. They built a digital universe to figure out how to use a specific type of light signal—Lyman-alpha (Lyα)—to measure the size of those invisible bubbles.

Here is the breakdown of their discovery using simple analogies:

1. The "Foggy Window" Problem

Imagine you are standing in a room with a window. If the window is clean, you can see a bright light outside clearly. But if the window is covered in thick fog, the light looks dim, fuzzy, or might not be visible at all.

The Galaxy: The lightbulb.
The Bubble: The clear space around the lightbulb.
The IGM (Intergalactic Medium): The fog in the rest of the universe.
Lyα Light: The specific color of light the galaxy emits.

If a galaxy is inside a small bubble, its Lyα light hits the fog immediately and gets scattered or absorbed. It's like trying to shout through a thick blanket; the sound gets muffled.
If a galaxy is inside a huge bubble, the light travels through clear air for a long time before hitting any fog. It arrives at our telescopes bright and clear.

2. The "Dirty Glass" Complication

The team realized that just looking at how bright a galaxy is isn't enough. There's a second problem: Dust.

Think of a galaxy not just as a lightbulb, but as a lightbulb wrapped in a dirty, dusty scarf. Even if the bubble around it is huge and clear, the dust inside the galaxy itself can soak up the Lyα light before it even escapes.

Massive galaxies tend to be dustier (like a heavy, thick scarf), making their light dimmer.
Small galaxies are often cleaner (like a thin scarf), letting more light through.

The authors had to build a complex "mathematical filter" to account for this dust. They calibrated their computer model to match real-world telescope data, essentially teaching the simulation how to "clean the lens" so they could see the true relationship between the galaxy and its bubble.

3. The Big Discovery: "The Brighter the Signal, the Bigger the Bubble"

Once they fixed the dust issue, they found a very useful rule of thumb, especially in the early days of the universe (when the fog was still thick):

If you see a galaxy with a very strong Lyα signal (high brightness or a specific "color" called Equivalent Width), it is almost certainly sitting inside a giant, clear bubble.

Analogy: Imagine you are at a party in a crowded, noisy room. If you hear someone shouting clearly from across the room, you know they must be standing in a quiet corner (a big bubble). If the room is packed tight with people (small bubbles), no one's voice carries far.
The Finding: Before the universe was halfway cleared (around 700 million years after the Big Bang), galaxies with the strongest Lyα signals were living in the biggest bubbles. The size of the bubble was the main reason they were visible.

4. The "Fog Clears" Effect

However, the team found that this rule changes as time goes on.

Early Universe: The bubbles are small and scattered. Only galaxies in the biggest bubbles can be seen. The connection between "bright signal" and "big bubble" is very strong.
Late Universe: As the bubbles merge and the fog clears up, everyone can be heard, even those in smaller bubbles. The connection weakens. A faint galaxy might be visible just because the whole room is finally clear, not because it has a special big bubble.

5. Why This Matters for JWST

The James Webb Space Telescope (JWST) is currently taking pictures of these ancient galaxies. Astronomers are trying to figure out: "How big were the bubbles when this galaxy existed?"

This paper provides a translation guide.

Old Way: Scientists used simple, idealized math (like assuming bubbles are perfect spheres) to guess bubble sizes. This is like guessing the size of a room by listening to a single echo; it's often inaccurate.
New Way: This paper uses a realistic, messy, 3D simulation to say, "If you see a galaxy with these specific Lyα traits, it is statistically likely to be in a bubble of this size."

Summary

The team used a super-powerful computer simulation to create a "Rosetta Stone" for the early universe. They showed that by carefully measuring the specific light coming from ancient galaxies, we can map out the invisible bubbles of clear space that surrounded them.

They found that early on, the brightest, most distinct galaxies were the best signposts for the largest clear bubbles. As the universe cleared up, this signal became less distinct, but the data they generated will help astronomers interpret the flood of new images coming from the JWST, helping us finally "see" the shape of the universe's first dawn.

1. Problem Statement

The Epoch of Reionization (EoR, $z \approx 5-20$ ) marks the transition of the intergalactic medium (IGM) from neutral to ionized. While the global ionization history is constrained, the topology and size distribution of ionized bubbles remain difficult to map directly.

The Challenge: Lyman- $\alpha$ (Ly $\alpha$ ) emission is a primary probe for high-redshift galaxies, but its observability is complex. It depends on intrinsic production, resonant scattering, dust absorption, galactic outflows, and crucially, the size of the surrounding ionized bubble.
The Gap: Current interpretations often rely on idealized spherical bubble models to infer bubble sizes from Ly $\alpha$ visibility. These analytic approaches suffer from degeneracies and oversimplifications, failing to account for the clumpy, inhomogeneous nature of the IGM and the specific radiative transfer effects within galaxies. There is a need for a robust, simulation-based framework to calibrate Ly $\alpha$ observables against realistic bubble sizes.

2. Methodology

The authors utilize the Thesan suite of cosmological radiation-hydrodynamics (RHD) simulations, which combine the IllustrisTNG galaxy formation model with on-the-fly ionizing radiation transport using the AREPO-RT code.

Key Methodological Steps:

Simulation Setup: The flagship run (Thesan-1) uses a $95.5$ cMpc box with high resolution ( $m_{DM} \approx 3.12 \times 10^6 M_\odot$ ), resolving atomic cooling halos down to $z=5.5$ .
Intrinsic Emission & Transmission:
- Intrinsic Ly $\alpha$ luminosities are calculated from recombination, collisional excitation, and stellar emission (using BPASS models).
- IGM transmission ( $T_{IGM}$ ) is computed via ray-tracing (768 directions per galaxy) through the native Voronoi mesh, accounting for frequency-dependent scattering and damping-wing absorption.
Empirical Calibration (The "Best-Fit" Model):
- Since large-scale simulations cannot fully resolve small-scale ISM/CGM physics (dust, winds), the authors apply an empirical calibration model to match observed Ly $\alpha$ Luminosity Functions (LFs).
- Spectral Profile: Modeled as an asymmetric double-peaked profile from a point source in a spherical cloud, parameterized by peak velocity offset ( $\Delta v_{peak}$ ) and wind strength ( $\beta$ ).
- Dust & Outflows: An empirical dust attenuation model (based on the IRX–UV relation) is applied. Crucially, the Ly $\alpha$ escape fraction is constrained to be $\le$ the continuum escape fraction.
- Calibration: Parameters ( $a_{peak}, b_{peak}, a_{wind}, b_{wind}, a_{dust}, b_{dust}$ ) are optimized via a Latin hypercube search to minimize the penalty function between simulated LFs and observational data (Ouchi et al., Konno et al.) at $z=5.5$ and $z=6.6$ .
Bubble Size Measurement: Ionized bubble sizes are measured using the Mean Free Path (MFP) method, extending rays from the virial radius of central galaxies to determine the distance to the nearest neutral region.

3. Key Contributions

Updated Calibration Framework: The paper presents an improved, empirically calibrated model for Thesan galaxies that simultaneously accounts for dust, galactic winds, and IGM transmission, reproducing observed Ly $\alpha$ LFs more accurately than previous iterations.
Statistical Link to Bubble Sizes: It establishes a direct statistical connection between observable Ly $\alpha$ properties (Luminosity, Equivalent Width) and the physical size of the host ionized bubble, moving beyond idealized spherical assumptions.
Public Catalogs: The authors generate and will release public catalogs of calibrated LAEs, enabling the interpretation of current and future JWST and narrow-band surveys in terms of reionization topology.

4. Key Results

A. Ly $\alpha$ Luminosity Functions and Evolution

The calibrated model successfully reproduces the observed Ly $\alpha$ LFs at $z=5.5$ and $z=6.6$ .
The model captures the rapid decline in Ly $\alpha$ visibility at higher redshifts ( $z > 7$ ) due to increasing IGM neutrality.
Tension: The model slightly overpredicts the LAE fraction at lower redshifts ( $z < 7$ ) compared to recent JWST data, suggesting the Thesan simulation may have a slightly early reionization history or that empirical dust modeling biases continuum baselines.

B. Dependence on Galaxy Properties

UV Brightness vs. Ly $\alpha$ Visibility: Brighter UV galaxies ( $M_{1500} < -19$ ) generally exhibit lower Ly $\alpha$ escape fractions and Equivalent Widths (EW) due to higher dust content and stronger winds.
Selection Bias: Applying a cut on observed Ly $\alpha$ luminosity ( $L_{Ly\alpha} > 10^{41.5}$ erg s $^{-1}$ ) biases the sample toward galaxies with higher escape fractions and EWs, particularly at high redshifts.

C. Correlation with Ionized Bubble Sizes

Pre-Midpoint Reionization ( $z \gtrsim 7$ ): There is a strong correlation between Ly $\alpha$ $α$ properties and bubble size.
- Galaxies within larger bubbles ( $R \gtrsim 10$ cMpc) show significantly higher observed Ly $\alpha$ luminosity and EW compared to those in small bubbles.
- Ly $\alpha$ Luminosity ( $L_{Ly\alpha}$ ) and Equivalent Width (EW) are stronger tracers of bubble size than UV magnitude ( $M_{1500}$ ), especially at $z \gtrsim 7$ .
- The LAE fraction in large bubbles is significantly higher than in small bubbles during this phase.
Post-Midpoint Reionization ( $z \lesssim 7$ ): As bubbles merge and the IGM becomes highly ionized, the correlation weakens. The distinction between galaxies in large vs. small bubbles diminishes, and the "bubble size" becomes less of a limiting factor for Ly $\alpha$ visibility.

D. Statistical Tracers

The LAE fraction serves as a practical statistical tracer for bubble size distribution before the midpoint of reionization.
The ratio of LAE density in large bubbles to the total LAE density tracks the global reionization history closely.

5. Significance

Bridging Theory and Observation: This work provides a critical "translation layer" between theoretical bubble size distributions and observable Ly $\alpha$ statistics. It clarifies that while Ly $\alpha$ is a powerful probe, its interpretation requires accounting for the evolving topology of the IGM, not just global neutral fractions.
JWST Interpretation: The results offer a framework for interpreting JWST observations of high-redshift galaxies. Specifically, the finding that Ly $\alpha$ luminosity and EW are better bubble-size tracers than UV magnitude helps refine constraints on the reionization timeline from upcoming narrow-band surveys.
Model Validation: By identifying tensions (e.g., the overprediction of LAE fractions at low $z$ ), the study highlights areas where simulation physics (reionization history, dust modeling) or observational analysis (continuum extrapolation) may need refinement.
Future Directions: The paper sets the stage for using these catalogs to study 21-cm/LAE cross-correlations and damping-wing absorption, offering a path to map the 3D topology of reionization.