A Huge and Powerful New Simulation Helps Explain the Universe's Reionization
Modern astronomy is a tale of two complementary disciplines. On one side stand the observational giants: massive telescopes with segmented mirrors, billion-pixel digital cameras, sophisticated data pipelines, and suites of powerful instruments each tuned to specific wavelengths or phenomena. But collecting exquisitely detailed data is only half the battle. Making sense of what these instruments reveal requires an equally powerful theoretical approach — and this is precisely where cosmological simulations become indispensable.
We live in a golden age of cosmological simulations, vast computational experiments run on the world's fastest supercomputers that allow researchers to model the universe from its earliest moments to the present day. These simulations help cosmologists understand everything from the birth of the first stars to the large-scale structure of the cosmic web, from the violent feeding of supermassive black holes to the slow expansion of spacetime itself. Now, a new and formidable entry into this field is poised to reshape our understanding of one of the most transformative eras in cosmic history.
The Observational Context: JWST, Roman, and the Early Universe
The James Webb Space Telescope (JWST) and the Nancy Grace Roman Space Telescope (NGRST) represent the current pinnacle of humanity's observational reach into the early universe. Both are particularly adept at probing the cosmos in the infrared, allowing them to peer through dust and detect the highly redshifted light of galaxies that existed when the universe was just a few hundred million years old. The JWST, now several years into its mission, has already delivered a series of stunning surprises — detecting unexpectedly massive and mature galaxies at extraordinary distances, challenging some existing models of early galaxy formation. The NGRST, launched just two months ago, has not yet begun its primary science mission, but astronomers anticipate it will deliver comparably revelatory data, particularly through its wide-field surveys.
This torrent of new observational data creates both an opportunity and a challenge. To extract the full scientific meaning from what JWST and Roman observe, theorists must provide detailed, physically consistent predictions to compare against. Raw data alone does not tell a complete story — it must be interpreted through the lens of well-constructed theoretical models. This is the intellectual space that a remarkable new simulation, Lumina, is designed to occupy.
"With JWST, the early universe has become an exciting new frontier in astrophysics. To understand what we are seeing with this telescope, we need to translate our theories into detailed predictions. After several years of work, with plenty of ups and downs, it is incredibly rewarding to see the simulation results and watch our predictions come to life in video." — Oliver Zier, Lead Author and Institute for Theory and Computation Fellow, Center for Astrophysics | Harvard & Smithsonian
Introducing Lumina: A New Benchmark in Cosmological Simulation
Developed by researchers at the Center for Astrophysics | Harvard & Smithsonian (CfA), Lumina is one of the largest and most detailed cosmological simulations ever constructed. It is presented in new research published in the Monthly Notices of the Royal Astronomical Society under the title "Introducing the Lumina Project: Large-Volume Radiation-Hydrodynamic Simulations of the Epochs of Hydrogen and Helium Reionization." The lead author is Oliver Zier, an Institute for Theory and Computation Fellow at the CfA.
What sets Lumina apart is both its ambition and its technical scope. It is a radiation-hydrodynamic simulation, meaning it simultaneously models the gravitational and hydrodynamic behavior of matter alongside the propagation and effects of radiation — a coupling that is computationally extremely demanding but physically essential when studying the reionization epoch. The simulation tracks the co-evolution of the intergalactic medium (IGM), galaxies, and active galactic nuclei (AGNs) across two major cosmic phase transitions: hydrogen reionization and helium reionization.
As the authors state in their paper:
"Understanding how galaxies and active galactic nuclei (AGNs) jointly drive the reionization of the intergalactic medium (IGM) across cosmic time remains a major challenge in cosmology. We present Lumina, a large-volume radiation-hydrodynamic simulation that self-consistently follows the coupled evolution of the IGM, galaxies, and AGNs through H i, He i, and He ii reionization down to redshift z = 3."
What Is Cosmic Reionization — and Why Does It Matter?
To appreciate what Lumina is modeling, it helps to understand the concept of cosmic reionization — one of the most profound phase transitions the universe has ever undergone. In the immediate aftermath of the Big Bang, the universe was an extraordinarily hot, dense plasma of particles and radiation. As it expanded and cooled over the first ~380,000 years, protons and electrons combined to form neutral hydrogen atoms — an event known as recombination. The universe then entered what cosmologists call the Cosmic Dark Ages: a period spanning hundreds of millions of years during which no stars or galaxies yet existed, and the universe was filled primarily with cold, neutral gas.
The Dark Ages ended with the formation of the first stars and galaxies, which began emitting ultraviolet radiation energetic enough to strip electrons from neutral hydrogen atoms — a process called ionization. This era, known as the Epoch of Reionization (EoR), transformed the universe's neutral IGM back into the ionized plasma we observe today. Understanding exactly how, when, and by what sources this transformation occurred is one of the defining open questions in modern cosmology.
- Hydrogen reionization is believed to have begun around redshift z ≈ 10–12 (roughly 300–500 million years after the Big Bang) and completed by z ≈ 6 (about 900 million years after the Big Bang).
- The primary drivers of hydrogen reionization are thought to be low-mass, star-forming dwarf galaxies that produced copious ultraviolet photons.
- Helium reionization is a distinct and later phase, driven by the harder (more energetic) radiation fields of quasars and luminous AGNs, which peaked in activity during Cosmic Noon.
- Helium II reionization (the ionization of singly-ionized helium to fully ionized helium) is generally considered the last major phase transition of the IGM, completing around redshift z ≈ 3.
Redshift z = 3 corresponds to a lookback time of approximately 11.7 billion years, placing us about 2.1 billion years after the Big Bang — well into the epoch astronomers call Cosmic Noon, when star formation across the universe peaked and quasars were at their most numerous and luminous. It is at this cosmic moment that Lumina's simulation concludes, covering both major reionization events in a single, self-consistent framework.
Scale and Resolution: Outpacing Its Predecessors
One of Lumina's most striking features is its sheer physical scale. The simulation operates within a computational "box" with a side length of 500 comoving megaparsecs (cMpc) — roughly 1.5 billion light-years on each side. For context, this is comparable to the volume of the observable universe that our most powerful telescopes can survey in detail. Compare this to the previously state-of-the-art IllustrisTNG simulation suite: its largest run, TNG300, had a box side of approximately 300 Mpc — already enormous, yet still significantly smaller than Lumina's volume.
This expanded volume is not merely a matter of scale for its own sake. A larger simulation box offers profound scientific advantages:
- Better statistics for rare objects: Massive galaxy clusters, the most luminous quasars, and giant cosmic voids are statistically rare. A larger volume captures more of these extreme objects, enabling more robust comparisons with observations.
- Reduced cosmic variance: In smaller simulations, random fluctuations in the initial conditions can skew results. Larger volumes average over more of the universe, producing more representative outcomes.
- Large-scale environmental effects: The properties of galaxies and the IGM are influenced by their large-scale surroundings — cosmic filaments, voids, and superclusters. A bigger box captures these environmental dependencies more faithfully.
- Forward modeling of observables: Lumina enables researchers to directly simulate what telescopes like JWST and Roman should observe, rather than working backward from observations alone.
Despite its enormous volume, Lumina maintains sufficient resolution to track individual galaxies and the physical processes driving reionization. Its galaxy population is kept consistent with the established physics of the IllustrisTNG project, providing a validated foundation on which the new, larger-volume simulation builds.
The Computational Frontier: Running Lumina
Simulations of Lumina's ambition do not run on ordinary hardware. The simulation is executed on more than 15,000 graphics processing units (GPUs) at the Oak Ridge National Laboratory's Frontier supercomputer, currently ranked among the world's three fastest supercomputers. Frontier's extraordinary parallel processing capability — measured in exaflops, or quintillions of floating-point operations per second — provides the computational muscle necessary to evolve the simulation across billions of years of cosmic time while maintaining the physical fidelity required to capture reionization processes accurately.
Extending the simulation to cover helium reionization presented particular computational challenges, as the authors candidly acknowledge:
"Modelling He ii reionization introduces significant computational challenges. To reach z = 3, the simulation must evolve for almost an additional gigayear after the end of H i reionization, during which the matter distribution becomes increasingly clustered."
As matter becomes more clustered over time, gas dynamics and radiative transfer become increasingly complex and computationally expensive. The fact that Lumina successfully navigates this challenge represents a genuine technical achievement, expanding the frontier of what radiation-hydrodynamic simulations can accomplish.
Key Scientific Results: Helium Reionization Comes Into Focus
Although Lumina is designed to address a broad array of cosmological questions over its scientific lifetime, its first results focus on the history and topology of helium reionization — a process that has received considerably less attention than hydrogen reionization, despite being essential to understanding the full thermal and chemical evolution of the IGM.
While hydrogen was reionized predominantly by lower-energy ultraviolet photons from star-forming dwarf galaxies, those photons lacked the energy to doubly ionize helium. Helium II reionization — the stripping of the remaining electron from singly-ionized helium (He⁺) — required the far harder radiation field produced by supermassive black holes accreting material at the centers of quasars and luminous AGNs. This population of extreme objects peaked in number and luminosity precisely during Cosmic Noon, around z ≈ 2–4, explaining the late timing of helium reionization relative to hydrogen.
Lumina's results on He II reionization show excellent agreement with current observational constraints:
- Onset: He II reionization begins at approximately z ≈ 6, driven by the hard radiation fields of AGNs.
- Midpoint: The simulation finds a midpoint at approximately z ≈ 4, consistent with observational inferences from quasar absorption spectra.
- Near-completion: Reionization is essentially complete by z = 3, with a residual He II volume filling fraction below 1%, matching current observational limits.
- Self-consistency: Crucially, Lumina achieves all of this while simultaneously and self-consistently tracking hydrogen reionization, the evolving galaxy population, and AGN activity — without treating any of these as separate, decoupled processes.
As the authors summarize: "This demonstrates that Lumina can self-consistently follow the onset, topology, and completion of helium reionization alongside hydrogen reionization." This self-consistency is scientifically critical — in reality, the reionization histories of hydrogen and helium are intertwined through the thermal and ionization state of the IGM, and a simulation that treats them in isolation risks missing important feedback effects.
Implications for Understanding Galaxy Formation and the IGM
Beyond its direct results on reionization, Lumina carries significant implications for our understanding of galaxy formation and evolution. The thermal history of the IGM — heavily influenced by reionization events — directly affects how gas cools and collapses to form stars, particularly in lower-mass galaxies. When UV radiation reionizes a region of the IGM, it also heats it, raising the pressure of gas in shallow gravitational potential wells and potentially suppressing star formation in small galaxies. Accurately modeling this radiative feedback is essential for explaining the observed properties of the faint galaxy population that JWST is now beginning to resolve at high redshifts.
Similarly, the topology of reionization — whether it proceeded smoothly and uniformly or in a patchy, bubble-driven fashion — leaves imprints on large-scale structure that can be detected through techniques such as the Lyman-alpha forest, 21-centimeter emission mapping, and cross-correlations with galaxy surveys. Lumina's large volume makes it ideally suited to generate predictions for these observational probes, enabling direct comparisons with data from facilities like the Square Kilometre Array (SKA) and upcoming spectroscopic surveys.
The simulation's compatibility with the IllustrisTNG galaxy model also means that its predictions for galaxy properties — stellar masses, star formation rates, metallicities, and morphologies — can be validated against the extensive body of observations that originally calibrated IllustrisTNG, providing confidence that Lumina's galaxy population is physically realistic even as it explores new territory in reionization physics.
A Powerful New Tool for the Cosmological Toolbox
The broader significance of Lumina lies not just in its first results, but in what it enables going forward. As NASA's Nancy Grace Roman Space Telescope begins its wide-field surveys and JWST continues to probe ever-earlier epochs of cosmic history, the demand for detailed, large-volume theoretical predictions will only intensify. Lumina is designed precisely to meet this demand — to serve as a theoretical laboratory where the consequences of different physical assumptions can be tested against a rich landscape of observational data.
"We feel incredibly fortunate to be building simulations like Lumina. For the first time, we can follow along as both hydrogen and helium changed into their modern form across such an enormous volume of the universe, while still resolving the galaxies and other luminous objects driving that transformation." —
Frequently Asked Questions
Quick answers to common questions about this article
1 What is cosmic reionization and why does it matter?
Cosmic reionization was a pivotal era roughly 400 million to 1 billion years after the Big Bang when energetic radiation from the first stars and galaxies stripped electrons from neutral hydrogen gas, fundamentally changing the universe's structure. Understanding it helps explain how the cosmos transitioned from a dark, foggy state to the transparent universe we observe today.2 What is the Lumina simulation and what makes it special?
Lumina is a cutting-edge cosmological simulation designed to model the early universe in unprecedented detail. Running on some of the world's fastest supercomputers, it generates physically consistent theoretical predictions that astronomers can directly compare against real observations from telescopes like JWST, helping scientists decode the behavior of the first stars and galaxies.3 How does the James Webb Space Telescope study the early universe?
JWST detects infrared light, which allows it to observe extremely distant galaxies whose light has been stretched to longer wavelengths by the universe's expansion — a phenomenon called redshift. This capability lets astronomers observe galaxies that existed just a few hundred million years after the Big Bang, some of which have turned out to be surprisingly large and mature.4 When did the Nancy Grace Roman Space Telescope launch and what will it do?
The Nancy Grace Roman Space Telescope launched approximately two months before this article's publication and has not yet begun its primary science mission. Once operational, it will conduct wide-field infrared surveys covering vast stretches of sky, expected to complement JWST's deeper, narrower observations and reveal large populations of early galaxies.5 Why do astronomers need computer simulations if they already have powerful telescopes?
Telescopes collect raw data, but data alone cannot explain the physical processes behind what we observe. Simulations model how gravity, gas, stars, and black holes interact over billions of years, producing testable predictions. Comparing simulated universes against real telescope observations allows scientists to confirm, refine, or challenge their theories about cosmic evolution.6 What surprising discoveries has JWST made about early galaxies?
JWST has detected galaxies at extraordinary distances that appear far more massive and structurally mature than existing models predicted for such an early cosmic era. These unexpected findings suggest that star formation and galaxy assembly in the early universe may have proceeded faster or more efficiently than scientists previously believed, prompting researchers to revisit fundamental theories.Related Articles
