This stunning image comes from Lumina. It shows part of the cosmic web about 650 million light years across. The fine filaments represent dark matter, and the warm orange and yellow tones highlight where gas is moving the fastest between galaxy clusters. Lumina is the most detailed cosmological simulation ever created, even bigger than Illustris TNG 300.

Image Credit: Oliver Zier, Center for Astrophysics | Harvard & Smithsonian

Modern astronomy features massive telescopes with segmented mirrors, huge digital cameras, sophisticated data pipelines, and suites of powerful instruments with different specific purposes. But that's the observational side. Making sense of the exquisitely-detailed data these telescopes deliver requires its own approach.

This is where simulations come in. We live in an era of increasingly detailed cosmological simulations, all run on the world's fastest supercomputers. Cosmologists use them to understand everything from star formation to the expansion of the Universe.

The James Webb Space Telescope (JWST) and the Nancy Grace Roman Space Telescope (NGRST) are the two most capable space telescopes in existence. They're especially good at illuminating the Universe's very early days with their infrared capabilities. While the NGRST is just getting going, the JWST has been busy for a few years and has already delivered some surprises. The NGRST was only launched two months ago, and its mission hasn't really started yet. But once it does, it may deliver some surprises of its own.

This is where a new simulation from the Center for Astrophysics | Harvard & Smithsonian (CfA) comes in. Its name is Lumina, and it's one of the largest and most detailed simulations yet.

Lumina is presented in new research in the Monthly Notices of the Royal Astronomical Society titled "Introducing the lumina project: large-volume radiation-hydrodynamic simulations of the epochs of hydrogen and helium reionization." The lead author Oliver Zier, an Institute for Theory and Computation Fellow at the Center for Astrophysics | Harvard & Smithsonian.

"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," the authors write. "We present Lumina, a large-volume radiation-hydrodynamic simulation thatself-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."

Redshift z=3 corresponds to a lookback time of about 11.7 billion years, or about 2.1 billion years after the Big Bang. This is in the Cosmic Noon, which is when star formation peaked in the Universe, and when quasars were most active. It's also at the end of Helium II reionization, which is usually considered the last phase transition in the IGM, after neutral hydrogen had already reionized.

"Its galaxy population remains consistent with the original IllustrisTNG project, while the larger volume improves statistics for rare objects, large-scale environments, and cosmic variance, enabling forward modelling of observables linking H i and He ii topologies to the evolving galaxy and AGN populations," the researchers explain.

Exactly how the Universe's neutral hydrogen and helium reionized is still not fully understood. The contribution from the various sources—stars and supermassive black holes and their AGN—is unclear. The timeline on which it all played out is also something cosmologists are still puzzling over. Did it end all at once, in a smooth denouement? Or was it patchy? The JWST's observations are helping drive a new effort to understand it all.

“With JWST, the early universe has become an exciting new frontier in astrophysics,” lead author Zier said in a press release. “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.”

Lumina has a larger volume than another well-known simulation, Illustris TNG. TNG 300 was its largest run, with a box side of about 3.3 megaparsecs. But Lumina has a box side of 500 Mpc. (Simulator nerds are right to point out that these are co-moving parsecs.) So Lumina simulates an area about 1.5 billion light years on each side, which is roughly the same as astronomers can see with our most powerful telescopes.

This image shows the density of baryonic matter, meaning gas and stars, in zoom-in boxes with side lengths of 500 cMpc , 50 cMpc, 5 cMpc, and 500 ckpc. The figure illustrates the embedding of the halo in the cosmic web, its filamentary gas supply, and the compact stellar structure at its centre.

Image Credit: Zier et al. 2026 MNRAS

Though hydrogen reionization has been studied extensively, since it's the most abundant element in the Universe, helium reionization hasn't been. Lumina captures helium reionization and feeds those results into the rest of the simulation. This sheds light on helium reionization's effect on the cosmos.

While hydrogen was reionized predominantly by lower-energy ultraviolet photons from star-forming dwarf galaxies, they weren't energetic enough to reionize helium, which made up about 25% of the IGM. It had to wait for a massive population of quasars and extremely luminous active galactic nuclei, and those populations peaked during Cosmic Noon.

But extending a simulation to cover helium reionization wasn't easy. "Modelling He ii reionization introduces significant computational challenges," the authors explain. "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."

To accomplish that, Lumina runs on more than 15,000 graphics processing units on the Oak Ridge National Laboratory's Frontier supercomputer, the third fastest supercomputer in the world, after El Capitan and Line Shine. Frontier had the power to compute a massive volume of space at the required level of detail.

Lumina's first results on He II reionization are a good match with observations.

"He ii reionization begins at z (about equal to) 6 and is driven by the hard radiation field from AGNs. We find a midpoint at z (about equal to) 4 and near completion by z = 3, with a residual He ii volume filling fraction below 1 per cent, in agreement with current observational constraints," the authors write. " This demonstrates that Lumina can self-consistently follow the onset, topology, and completion of helium reionization alongside hydrogen reionization."

Simulations and observations together drive progress in cosmology, and it looks like theorists now have another powerful tool in their toolbox.

“We feel incredibly fortunate to be building simulations like Lumina,” said co-author Aaron Smith, an assistant professor of physics at the University of Texas at Dallas. “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.”

Now, a team of researchers led by the University of Warwick might have introduced a new way to literally look at stars, as they revealed new evidence for white dwarfs to exhibit colors other than what their name implies. These findings were recently published in RAS Techniques and Instruments and could provide new insights into the formation and evolution of stars throughout the cosmos. This is because stars are referred to as “perfect radiators”, also called blackbody objects, meaning they absorb all electromagnetic radiation and reflect nothing. In a nutshell, their glow is entirely dependent on temperature.

“As an object gets hotter it will emit more light at optical wavelengths, changing from red to orange (think of a blacksmith working on a piece of hot metal),” Dr. Mark Hollands, who is a postdoctoral researcher at the University of Warwick and the study’s lead author, tells Universe Today. “Eventually these orange hues will become paler, and by 6000K a black body will appear will just appear white as most of its radiation will spread across optical wavelengths. For even hotter temperatures an object will start to glow pale blue as the majority of its light will become emitted in the ultraviolet. This set of red/orange/white/blue are the only colors that are possible for a true blackbody.”

For the study, the researchers used a series of computer models to study 151 white dwarfs with 20 parsecs (65 light-years) from Earth and how white dwarfs might challenge the longstanding understanding of stars being blackbodies. The reason white dwarfs were chosen was due to their extreme characteristics compared to main-sequence stars like our Sun, most notably white dwarfs having much denser atmospheres. The researchers aspired to learn if the presence of certain elements like carbon, calcium, or sodium in their dense atmospheres could influence the glow of white dwarfs.

In the end, the researchers found that a total of five white dwarfs of the 151 sampled showed that a combination of a white dwarf’s dense atmosphere with certain elements could cause these five white dwarfs to glow pink and yellow, specifically two yellow and three pink. This indicates about 3.3 percent of white dwarfs in the cosmos could exhibit different colors aside from white.

A white dwarf whose atmosphere is comprise of carbon appearing pink.

(Credit: Mark A. Garlick/University of Warwick)

“One in thirty might not sound like a huge fraction, but across the entire population of white dwarfs in the Universe, that is still a lot of oddly colored stars out there,” said Dr. Boris Gänsicke, who is a professor in the Department of Physics at the University of Warwick and a co-author on the study. “Yet if we take the median across a representative population of white dwarfs, we found that the average white dwarf really is white."

Regarding the specific elements responsible for altering the white dwarf’s colors, the researchers found that carbon causes white dwarfs to appear pink or purple while cooler white dwarfs with hydrogen and helium tend to be cyan or blue. Additionally, the researchers found that white dwarfs whose atmospheres have debris from their original planetary system are typically green.

As noted, star colors are typically determined by their surface temperatures, with hotter stars to cooler stars ranging in several colors, including blue, blue-white, white, yellowish white, yellow (our Sun), orange, and red. The larger the star, the hotter and bluer it is. But while this study demonstrates there could be exemptions to this longstanding notion, could other stars also exhibit unexpected colors?

“The other class of object that can disobey the rule are brown dwarfs," Dr. Hollands tells Universe Today. "These are substellar objects that are too cool to fuse hydrogen in their cores, but can fuse deuterium instead. Because these are quite cold objects (compared to normal stars) most of their radiation is in the infrared, but some still comes out at optical wavelengths. And because brown dwarfs also can have dense atmospheres, absorption from sodium, potassium, and molecules like titanium oxide can also alter their colors away from blackbody colors. Some astronomers who study brown dwarfs have looked into their appearance in the past and concluded that they can appear violet in color!”

What new insights into white dwarfs and other star colors will researchers make in the coming years and decades? Only time will tell, and this is why we science!

As always, keep doing science & keep looking up!

The disk is further divided into a thin disk and a thick disk. They're not only structurally different, but their stellar populations and other characteristics are different.

The thin disk contains young stars, and middle-aged stars like our Sun. It holds the spiral arms, where our Solar resides and also holds most of the Milky Way's gas and dust and is where young stars form. It also has higher metallicity, generated by previous generations of stars that have lived and died.

The thick disk is more puffy and diffuse than the thin disk. It holds ancient stars, and almost no new stars form there. It has much lower metallicity than the thin disk, and astronomers sometimes refer to it as a kind of stellar fossil record.

Astronomers wrestle with understanding the formation of these disks, and the Milky Way as a whole. Observations have led to a multiple explanations. In one, stars are born in the thin disk, and over time, encounters with gas clouds, the spiral arms, or even the central bar, perturb some of these stars. Dynamical scattering drives them away from the plane of the thin disk.

Another explanation says that as stars form in the inner disk, the gas they form from settles down into an even thinner disk. In this scenario, each successive generation of stars forms in an increasingly thin disk.

Astronomers have difficulty distinguishing between these competing explanations. It relies on accurate measurements of the ages of individual stars. That's difficult to achieve, especially for the oldest stars.

New research in The Astrophysical Journal Letters used Gaia data to tackle the issue. It's titled "A Two-stage Settling of the Milky Way Disk Revealed by Precise ChronoGal Ages," and the lead author is David Mirabal. Mirabal is a pre-doctoral researcher at the Instituto de Astrofisica de Canarias (IAC).

"The spatial distribution of Galactic disk stars of different ages and metallicities encodes its dynamical evolution history and formation pathways," the authors write in their research letter. "The lack of precise ages for large, representative samples of stars has traditionally been a drawback for securely differentiating among formation scenarios proposed to explain the vertical properties of the Milky Way disk."

This work is based on data in Gaia Data Release 3 (DR3), released in June, 2022. This research is one of multiple papers released by a collaboration of researchers, aimed at deciphering stellar ages in the populations of the thin and thick disks. In this one, the authors employed colour-magnitude diagram fitting of the Gaia data. From that, they derived age-metallicity distributions for the stellar populations. They also studied their vertical profile, their position in either disk.

Earlier in the Universe's history, the gas in the Milky Way was more turbulent and the galaxy was more puffed up vertically, as if it was one thick disk. Stars were born orbiting in the thick disk. As the gas calmed over time and became less turbulent, it cooled down and shrank down, creating a thinner disk inside the thick disk. There was a significant change about 10.8 billion years ago, a couple of billion years after the Big Bang, as the thin disk formed, with the older stars maintaining their orbits in the thick disk.

The existence of a clear break in the process about 10.8 billion years ago is the main argument in this research. The rate of vertical settling changed drastically at around this time.

"We find evidence that the scale height decrease rate changed drastically (~10.85) Gyr ago: at older ages, the evolution is very steep, with hZ (vertical scale height) dropping from over 1.26 to ∼0.5 kpc in merely 2 Gyr; subsequently, the thinning is very gradual, with hZ ∼ 0.1 kpc for the youngest populations," the authors write.

So before about 10.85 Gyr ago, the disk flattened rapidly. In only about 2 billion years, it thinned from about 4,100 light years thick down to only about 1,600 light years thick. After that, the thinning was much more gradual, reaching a thickness of only about 400 light years.

This figure from the research is packed with information that wrinkly astrophysicist brains can decipher. For the rest of us, the important part is the sharp drop in hZ, or vertical scale height, about 10.8 billion years ago. This is when the disk stopped thinning so drastically, leaving ancient stars on orbits in the thick disk, well above or below the remaining thin disk.

Image Credit: David Mirabal et al 2026 ApJL 1010 L38

This means the vertical structure of the disk components grew from the thick layer inward, an upside-down scenario. Ancient stars in the thick disk weren't scattered there by interactions with gas clouds, spiral arms, or other things; they were born there when the disk was much thicker overall. As time passed, generations of stars were born in a successively thinner disk.

"The extremely rapid early vertical settling of the disk points to a fundamentally upside-down formation process at early times, when the thick disk component is basically formed," the researchers explain. "Enabled by precise ages and high completeness, our results exhibit two clear, time-dependent evolutionary phases for the Galactic disk, which allow piecing together a comprehensive picture of the physical processes responsible for a two-phase assembly history of the Galactic disk."

This created what we see today. The thick disk has older, metal-poor stars and almost no gas for new star formation. The thin disk has younger stars, is the site of most star formation, and holds most of the gas for new stars.

"We thus strongly support an evolutionary scenario defined by two distinct temporal phases: an initial, highly turbulent epoch that gave birth to the thick disk, and a subsequent more quiescent period of secular evolution and gradual gas settling, which formed the thin disk," the authors write.

“This result resolves a classic problem in astrophysics,” lead author Mirabal said in a press release. “In the simulations, both disc formation models were possible, but there was a lack of observational evidence to determine which one occurred in the Milky Way. This is the first clear observational evidence favouring one of the two models for the Galaxy’s early stage. And it has been possible thanks to the high precision we achieved for the older ages,” he said.