Astrobites RSS

illustration of an exoplanet and exomoon around a star

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: A Deep Search for Exomoons Around WISE 0855 with JWST
Authors: Mikayla J. Wilson et al.
First Author’s Institution: University of California, Santa Cruz
Status: Published in AJ

The “Moon”-umental Question

The solar system hosts hundreds of moons, ranging from volcanic worlds like Io around Jupiter, to icy objects like Enceladus around Saturn, to captured objects like Neptune’s retrograde moon Triton. Moons are essential to our model of how the solar system formed and also offer some of the best chances we have for finding life beyond Earth.

Astronomers also expect exomoons, or moons orbiting planets outside the solar system, to be abundant around other giant exoplanets. But how common are exomoons? How do they compare to the moons in our solar system?

In order to begin answering those questions, we must first detect an exomoon, which has proved difficult despite decades of searching by astronomers. Fortunately, JWST presents a new opportunity to uncover the exomoon population by looking at lonely free-floating planets as they drift through space.

Why Free-Floating Planets?

One proposed method for searching for exomoons is by looking for their transits in front of their host planets, characterized by the dips in brightness of the planet as the moon passes in front, blocking the planet’s light. Looking for exomoon transits around planets orbiting stars is quite difficult, as the bright starlight can easily drown out the small signals of exomoon transits. Free-floating planets solve this issue by removing the star entirely, increasing our sensitivity to such detections. (See this bite for a good review.)

The authors of today’s article directed the exomoon hunt towards the free-floating WISE J085510.83-071442.5 (or WISE 0855). It has the prestige of being the coldest known brown dwarf (250–285K) while also sitting at a relatively low mass (3–10 Jupiter masses). Notably, it is also one of our closest neighbors at a distance of only 7.4 light-years, making it ideal for high-precision observations despite its faintness. Even though brown dwarfs are technically distinct from planets, the authors opt to refer to companions around WISE 0855 as moons given WISE 0855’s “planetary-mass” status. (It’s complicated…)

Repurposing JWST Data… for Moons!

The JWST observations used in this study contain 11 hours of near-infrared (2.87–5.27 microns) time-series spectra originally intended to study water clouds and weather on WISE 0855. Time-series brightness monitoring can also be used for transit searches, which the authors take advantage of.

One complication is that WISE 0855 is variable, meaning its intrinsic brightness changes over time. Variability is likely driven by clouds and other dynamic processes within its atmosphere. So how do the authors distinguish between a passing moon and a turbulent atmosphere? The key idea is that variability is wavelength dependent, meaning that the brightness of WISE 0855 will fluctuate differently depending on the observed wavelength. In contrast, transits are “gray,” meaning that the same amount of light is blocked at all wavelengths, producing a consistent feature across the entire spectrum.

Finding Moons with Statistics!

The authors apply this idea and pick out two wavelength regions of WISE 0855’s spectrum that contain two distinct variability patterns, which should both contain an identical moon transit signal (if present). They then generate a light curve (how brightness changes over time) for these two regions (see Fig. 1).

WISE 0855 light curves

Figure 1: (A) Light curves from two selected wavelength regions of WISE 0855’s spectrum with injected transit signals. Also plotted is the best-fit Gaussian processes + transit model for the two light curves. (B) Light curve data after subtracting the Gaussian processes portion of the best-fit model, revealing the example injected transit signals. [Wilson et al. 2025]

To appropriately model the variability, the authors employ Gaussian processes, a flexible tool that can model complex, quasi-periodic signals like atmospheric variability. They compare fits from two types of models:

  • Gaussian processes–only model: Assumes that all observed variability is intrinsic to the planet itself
  • Gaussian processes + transit model: Includes a simple trapezoidal exomoon transit signal that is simultaneously fit in both light curves

Using Bayesian evidence (a measure of how well each model explains the data), they determined which model was favored. So, what do they find?

The Bad News and the Good News

Based on Bayesian evidence, the authors conclude that there are no statistically significant detections of exomoons in the data. The results suggest very weak evidence for a ~0.53-Earth-radius moon at a wide separation from WISE 0855 — an unlikely scenario given that transit probability decreases at greater separations (and therefore longer orbital periods).

Yet, the study goes further: What kinds of moons is JWST able to detect, if any? To answer this, the authors performed injection and recovery tests, where they injected artificial transit signals of varying depths (exomoon sizes) into the data and tested how well their models were able to recover them (results shown in Fig. 2). They find that JWST is capable of detecting 96% of transits with depths ≥0.5%, equivalent to a Titan-like moon. Smaller Io-like moons were also detectable more than half of the time. This means that if a Titan analog had actually transited during these observations, we would almost certainly have seen it!

plot of successful detections of injected transit signals

Figure 2: Results showing the number of successful detections for the transit injection and recovery tests. Fifty transit injections are done for transit depths of 1%, 0.5%, 0.4%, 0.3%, 0.2%, and 0.1%. The transit depths represent different exomoon sizes, with the shaded regions representing Io-like and Titan-like moons. [Wilson et al. 2025]

JWST will continue to gather more time-series data of free-floating planets, brown dwarfs, and directly imaged exoplanets, each providing a new opportunity to help us better understand the moon population outside of our solar system. We’re still waiting for the first confirmed exomoon, but when that transit finally happens, we know that JWST will be ready.

Original astrobite edited by Kelsie Taylor.

About the author, Jared Bull:

I am a 2nd-year PhD student at Johns Hopkins University. I study brown dwarf variability and am interested in using time-series observations to uncover dynamic processes within their atmospheres. In my free time I like to read, cook, and do astrophotography.

galaxy cluster MACS J1149.5+2223

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: Metal-Poor Star-Forming Clumps in Cosmic Noon Galaxies: Evidence for Gas Inflow and Chemical Dilution Using JWST NIRISS
Authors: Vicente Estrada-Carpenter et al.
First Author’s Institution: Arizona State University
Status: Published in ApJ

If you want to reconstruct a galaxy’s life story, one of the best “fossil records” is its metallicity. In astronomy this refers to the abundance of elements heavier than helium, which are made in stars and returned to a galaxy’s gas through winds and supernovae. Over time, more star formation usually means more metals mixed into the gas.

Now zoom to Cosmic Noon (roughly when the universe was most actively forming stars). Many galaxies in this epoch look “clumpy”; their star formation is concentrated in several bright knots scattered across the disk. The big question is where those clumps come from. Do they form from the galaxy’s own gas via internal disk instabilities, or do they light up when fresh, metal-poor gas flows in and both fuels star formation and dilutes the local metallicity?

The authors try to answer that question with JWST by measuring the metallicity of each clump relative to its immediate surroundings, rather than comparing clumps to a single galaxy-wide number (which can be misleading if the galaxy has a metallicity gradient).

They use JWST/Near Infrared Imager and Slitless Spectrograph (NIRISS) slitless grism spectroscopy from the CAnadian NIRISS Unbiased Cluster Survey (CANUCS) to study 20 lensed galaxies at redshift 0.6 < z < 1.35. (Lensing effectively acts like a zoom lens, helping to resolve smaller structures.) They focus on emission lines that trace star-forming gas, especially Hα (a star formation tracer) and sulfur lines [SII] and [SIII] (needed for their metallicity method).

Slitless spectra come with a headache: because there is no slit, different parts of a galaxy can overlap in the dispersed image. To make reliable emission-line maps from slitless data, the authors use a forward-modeling code called Sleuth, which allows the continuum to vary across the galaxy.

The authors identify clumps using the Hα map together with rest-frame ultraviolet imaging because these tracers are sensitive to star formation on different timescales: Hα highlights gas ionized by the youngest massive stars, while ultraviolet light traces young stellar light over longer periods. As a result a clump can be bright in one and not the other, especially if dust is involved.

To estimate gas-phase metallicity, they use the “strong-line” method, which infers metallicity from ratios of bright emission lines calibrated using models and empirical samples. Their main diagnostic is S23 = ([SIII] + [SII]) / Hα. Because some line ratios also depend on the ionization state (how strongly the gas is being ionized by young stars), they also use the sulfur ratio S32 = [SIII]/[SII] as a check and iterate to a self-consistent solution.

So, Are Clumps Really Chemically Different from Their Surroundings?

For each clump, the authors measure the metallicity inside the clump and compare it to an annulus just outside the clump (masking neighboring clumps to avoid mixing). When they plot “clump metallicity” versus “local disk metallicity,” most points fall below the 1-to-1 line, meaning the clumps are more metal poor than their surroundings (Figure 1). The mean offset is about 0.1 dex, which corresponds to roughly 20% dilution in the clump gas.

plot of gas-phase metallicity in star-forming clumps compared against metallicity of nearby disk regions

Figure 1: Each point compares a star-forming clump’s gas-phase metallicity to the metallicity of the nearby disk region immediately surrounding it. If clumps and disks had the same metallicity, they would lie on the dashed 1-to-1 line. Instead, most clumps sit below it, showing a typical ∼0.1 dex metallicity deficit, consistent with local chemical dilution. (The solid red line shows a best-fit linear trend to the clump measurement.) [Adapted from Estrada-Carpenter et al. 2025]

An extra wrinkle is that the galaxy medians hint at two populations: some galaxies have clumps with small offsets (near the 1-to-1 line), while others show larger offsets. The authors suggest this could mean two formation pathways, one dominated by internal gas reservoirs (smaller offsets) and another where inflow of metal-poor gas plays a bigger role (larger offsets). They are careful, though, as the sample is still small.

If inflow is really the driver, you should expect a link: the more strongly a clump is forming stars relative to its surroundings, the more diluted its metallicity should be. That is exactly what the authors find. The clumps that are most boosted in star formation are also the most metal diluted.

The article also emphasizes that clumps are not chemically uniform blobs. In at least one detailed example, the peaks in Hα (highest star formation) coincide with local minima in metallicity along a cut through the galaxy (Figure 2), suggesting internal star formation rate and metallicity gradients that reinforce the same story, intense star formation goes hand in hand with lower metallicity in the clump regions.

galaxy image showing spatial variation of H alpha and metallicity

Figure 2: Example galaxy from this study illustrating the clump-scale link between star formation and metallicity. The color image shows the galaxy with a rectangular strip marking the region used for a 1D cut. In the bottom panel, the turquoise line shows the Hα flux (a tracer of recent star formation), while the white line shows the metallicity along the same cut. Peaks in Hα flux line up with local dips in metallicity, showing that the brightest star-forming clumps are also the most chemically diluted compared to nearby regions. [Adapted from Estrada-Carpenter et al. 2025]

Are These Clumps Really “In Situ,” or Could They Be Small Satellites?

A reasonable alternative is that some clumps are actually small companion galaxies projected onto the disk. These could also look metal poor, because low-mass galaxies tend to be low metallicity. The authors look for evidence using face-on systems and find that more massive clumps tend to sit closer to galaxy centers, which is consistent with in-situ clumps that form in the disk and migrate inward, though it does not rule out satellites. They argue that kinematics from JWST/NIRSpec IFU will be needed for a definitive separation.

Why This Matters

Gas inflow, star formation, and feedback, together known as the baryon cycle, are key drivers of how galaxies grow. What this article adds is a spatially resolved view that compares each clump to its local environment, showing that regions of elevated star formation also tend to be locally metal poor. That pairing is hard to explain as a simple metallicity gradient or a galaxy-wide averaging effect, and it is exactly what you would expect if at least some clumps are being fueled by relatively metal-poor inflows. In short, these clumps may be snapshots of galaxies refueling in real time.

Original astrobite edited by Ryan White.

About the author, Niloofar Sharei:

I’m an astronomy PhD candidate at UC Riverside studying how galaxies grow through star-forming clumps. I track how these clumps emerge, evolve, and sometimes survive long enough to reshape their galaxies. When I’m not thinking about cosmic blobs, I’m reading, hiking, or listening to Bach.

illustration of stars in the early universe

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: Hunting for the First Explosions at the High-Redshift Frontier
Authors: Junehyoung Jeon et al.
First Author’s Institution: The University of Texas at Austin
Status: Published in ApJ

Back in the 1920s, astronomers discovered that we live within just one of many, many galaxies in the big, wide universe. Since then, we’ve been racing to search for the most distant galaxy that can be observed — in other words, searching for the oldest starlight we can see, since the light from these distant sources has been travelling towards us for most of the age of the universe. (Remember: more distant = higher redshift = longer lookback time.)

This race to the redshift frontier has had a pretty eventful history (see a great overview video here), which became even more eventful with the launch of JWST. JWST rapidly smashed the previous redshift (z) record of z = 10.6 by discovering a galaxy at z = 13.2, and then it broke its own record twice more. The current title holder sits at z = 14.4, observed less than 300 million years after the Big Bang.

Several galaxy candidates (to date unconfirmed) have now even been proposed at z ~ 25–32 (e.g., Capotauro), only 100 million years after the Big Bang! If real, these sources would pose a serious challenge to our understanding of the formation of the first galaxies, as galaxies shouldn’t really be observable at such early times. In today’s article, the authors put forward an intriguing alternative: what if some of these ultra-high-redshift candidates aren’t galaxies at all, but transient explosions from the universe’s first stars?

The First Stars and Their Explosive Endings

The earliest generation of stars (Population III; see my previous bite on these here) formed from pristine hydrogen and helium gas. Without metals to cool the gas efficiently, theory predicts that these stars were extremely massive, often exceeding 100 solar masses. While such stars would be short-lived, their deaths could be spectacular.

Population III stars of sufficient mass are predicted to end their lives as hyper-energetic pair-instability supernovae (PISNe). This is a long-winded name for a rapid, intensely hot explosion that leaves no remnant behind — not even a trace of the pre-existing star. Whilst nothing would remain of the star, the light emitted in that explosion could be bright enough to masquerade as a high-redshift galaxy candidate in current JWST surveys, but only if three key conditions are met:

  1. JWST must observe a sufficiently overdense region, where lots of Population III stars can form very early.
  2. A PISN must occur while JWST is “watching.”
  3. The explosion must be bright enough to rise above JWST’s detection limits.

Simulating a (Biased) Universe

To address the likelihood of these conditions having already been met by existing JWST observations, the authors turn to cosmological simulations. Rather than simulating an “average” patch of the universe, they focus on an extremely overdense region (Fig. 1). This creates a rare but important environment where structures collapse earlier than usual. These regions are exactly where large numbers of Population III stars are expected to form at the highest redshifts.

projection of the gas density in the simulated overdense region at z = 30.4

Figure 1: A projection of the gas density in the simulated overdense region at z = 30.4. The densest structures stand out clearly, tracing the locations where the first stars are able to form. Black dots mark newly formed groups of stars, while the most massive dark matter halo in the region is highlighted with an orange circle. The figure illustrates that in such an unusually dense patch of the early universe, star formation can already be underway just 100 million years after the Big Bang, creating the conditions needed for early Population III stars and their explosive deaths. [Adapted from Jeon et al. 2026]

In their simulations, star formation begins as early as z ~ 30–40 (within the first hundred million years after the Big Bang), producing Population III stars and, by extension, potential PISNe very shortly afterwards. While such overdense regions are rare, the authors show that the total area already surveyed by JWST (including large surveys such as CEERS, JADES, PRIMER, and COSMOS-Web) is large enough that it is plausible JWST has already observed at least one such region.

Catching a Cosmic Explosion in the Act

So, we can tick off condition #1: it’s possible that a sufficiently overdense region has already been observed by JWST. Next up, how lucky do we have to be to catch an explosion in the act (condition #2), so to speak? For this condition, cosmic time dilation actually works in our favour. A PISN explosion at z > 20 that lasts only months in its own rest frame can last for decades in the observed frame. (This whole time-being-relative thing sounds wacky because it is — please join me, Neil deGrasse Tyson, and countless others in struggling to imagine this.)

But would any such explosion be bright enough (condition #3)? Using theoretical PISN spectra, the authors show that these explosions could reach observed magnitudes of ~28–29 at z ~ 30 — right at the depth of current JWST deep surveys (Fig. 2). In fact, the predicted brightness and colours are somewhat comparable to those of some proposed z ~ 30 candidates (Fig. 3), raising the possibility that these objects could be PISNe rather than galaxies.

plot of predicted PISN brightness

Figure 2: This plot compares the predicted brightness of PISNe originating from different types of extremely massive stars to the depth reached by existing JWST surveys, showing that these explosions could remain detectable for ~20 years at peak brightness in the observed frame. [Jeon et al. 2026]

Theoretical PISNe spectra compared to a proposed z = 32 source

Figure 3: A comparison of theoretical PISNe spectra with the observed photometry of one proposed z ≈ 32 source (Capotauro). The coloured curves show model spectra for PISNe originating from extremely massive stars, at different stages of the explosion, while the data points represent the observed brightness of the high-redshift candidate across multiple JWST filters. [Jeon et al. 2026]

So… Are We Seeing the First Stars Die?

It’s not time to throw a party just yet. The authors note there are several caveats and uncertainties. The nature of Population III stars is still highly uncertain, JWST does not continuously monitor the same patch of sky, and identifying a PISN at such high redshift would be extremely challenging. Thanks to time dilation, these explosions would fade very slowly, making them hard to distinguish from steady sources using photometry alone. Alternatively, there are other plausible explanations for these ultra-high-redshift candidates: lower-redshift interlopers (a rather infamous example is CEERS-93316), local brown dwarfs, or even nearby exoplanets.

Still, the idea is exciting. If JWST were to detect a genuine PISN at z > 20, it would represent a direct glimpse of the very first stars, pushing observational astronomy into truly uncharted territory. For now, the most distant explosions in the universe may already be hiding in JWST images; we just have to learn how to recognise them.

Original astrobite edited by Nathalie Korhonen Cuestas.

About the author, Lucie Rowland:

I’m a fourth (and final!) year PhD student at Leiden Observatory in the Netherlands, studying massive, star forming galaxies in the early universe with ALMA and JWST. It’s a really exciting time to be interested in astronomy, so I hope to make groundbreaking new research more accessible!

Population III stars

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: GLIMPSE: An Ultrafaint ≃105 M⊙ Pop III Galaxy Candidate and First Constraints on the Pop III UV Luminosity Function at z ≃ 6–7
Authors: Seiji Fujimoto et al.
First Author’s Institution: University of Toronto; The University of Texas at Austin
Status: Published in ApJ

You, me, your laptop, my $8 matcha, and just about everything else on Earth was forged in the fiery bellies of dying stars. Generations of stars had to live and die before the universe became enriched with any elements heavier than helium (what astronomers call “metals”). The first stars to undergo this cosmic cycle are known as Population III (Pop III) stars. Though their existence has been hypothesized since the 1960s, astronomers have failed to observe these distant metal-free stars or the faint, low-mass galaxies that host them.

The first Pop III stars likely formed around 100 million years after the Big Bang in pristine pockets of hydrogen gas. Although these are too distant for us to observe, we expect that as the universe started to become metal enriched, there were still existing pockets of gas introverted enough to survive unpolluted and form metal-free Pop III stars up to a redshift of z ~ 6–7 (when the universe was around 900 million years old)!

JWST is the perfect instrument to search for these systems. You can read other astrobites on the search for possible Pop III systems with JWST here and here. The authors of today’s article seek to develop the most efficient way of using JWST’s Near-Infrared Camera (NIRCam) to find the galaxies hosting Pop III stars. Using their selection method on existing NIRCam data, the authors identified one promising Pop III galaxy candidate.

I’m Not Like Other Galaxies

In order to find a Pop III galaxy, we need to take a look at galaxies’ spectral energy distributions (SEDs). These are graphs that show the energy emitted by a galaxy at different wavelengths of light. Pop III galaxies are expected to have SEDs that differ from your everyday, metal-enriched galaxy. NIRCam will be especially sensitive to three key spectral features that show up in the SEDs of Pop III galaxies: an absent [O III] line (light emitted by doubly ionized oxygen atoms), a strong H-alpha line (light emitted when a hydrogen atom transitions from its third to its second energy level), and a significant Balmer jump (light absorbed to ionize electrons in the second energy level of a hydrogen atom). To identify these key SED characteristics, the authors use SED fitting and color–color diagrams to execute an efficient Pop III search with NIRCam.

The first selection method involves SED fitting. Astronomers create template SEDs that represent different types of galaxies and then compare these templates to the observed SEDs to see which one matches best. In this work, the authors use metal-rich galaxy templates and Pop III templates to fit the galaxies observed with NIRCam. They then calculate the chi-squared χ2 (a statistical measure of best fit) between the data and all the SED templates. A galaxy is selected as a Pop III candidate if the Pop III model provides a good fit (χ2 < 10) to the photometry and is significantly better than any metal-rich model. It’s kind of like looking for Cinderella by making every woman in the kingdom try on the glass slipper.

A color-color diagram showing how Pop III models lie in a different parameter space than metal-rich galaxies.

Figure 1: Color–color diagram for selecting Pop III galaxies where the x and y-axes show the different NIRCam filters being subtracted. The cyan symbols are different Pop III models while the other colored dots are different metal-rich galaxy models. [Adapted from Fujimoto et al. 2025]

A color–color diagram plots the difference in magnitude between two filters on each axis. NIRCam filters are specially chosen to emphasize the SED characteristics above. When these filters are chosen, Pop III galaxies occupy a distinct region of this diagram as compared to metal-rich galaxies. For example, subtracting the F356W filter from the F277W filter is sensitive to the presence of the [O III] line and the Balmer jump. Figure 1 demonstrates how this color selection separates Pop III galaxies from typical galaxies.

O Pop III, Pop III, Wherefore Art Thou?

The authors apply their fresh new selection criteria to publicly available NIRCam data from large surveys. And (drum roll please) the slipper fits! The Pop III galaxy candidate GLIMPSE-16043 is an ultra-faint galaxy at z = 6.5. It was imaged in the GLIMPSE survey, which uses the technique of gravitational lensing to observe faint and distant galaxies.

The GLIMPSE survey targeted a massive galaxy cluster, Abell S1063. The cluster bends the light from distant galaxies and, like a giant lens, magnifies faraway objects, providing some of the deepest JWST imaging to date. The Pop III candidate passes both tests: it resides in the Pop III region of the color–color diagram, and its SED is best fit by a Pop III model, not a metal-rich galaxy model (see Figure 2). Next, spectroscopic follow-up is needed to ensure that this galaxy is truly metal free and not just extremely metal poor.

The spectral energy distribution of a Pop III galaxy candidate. This plot shows that the JWST data is best fit by a Pop III model rather than a metal-rich galaxy.

Figure 2: SED of GLIMPSE-16043 with the best-fit Pop III template (blue) and best-fit metal-enriched template (gray). The top panel is the galaxy imaged in different filters from NIRCam and the Hubble Space Telescope. [Fujimoto et al. 2025]

The authors conclude that our best shot at identifying additional Pop III galaxy candidates is using NIRCam to image large numbers of gravitationally lensed clusters. Without magnification from gravitational lensing, it may be impossible to see these ultra-faint Pop III galaxies. Once candidates have been identified, they can be followed up with deep spectroscopy to confirm their redshift and their lack of metals. Who knows? With these new methods, we may soon get a glimpse of the universe’s very first stars.

Original astrobite edited by Chris Layden and Margaret Verrico.

About the author, Madison VanWyngarden:

I am a first-year PhD student in astronomy and NSF Graduate Research Fellow at the University of Arizona. I study galaxy formation and evolution in the distant universe and am particularly interested in dusty star-forming galaxies. In my free time, I love reading, hiking, and baking bread!

galaxy transitioning from star forming to quiescent

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: Searching Within Galaxies for the Earliest Signs of Quenching With Spatially Resolved Star Formation Histories in UVCANDELS
Authors: Charlotte Olsen et al.
First Author’s Institution: New York City College of Technology
Status: Published in ApJ

How and Why Do Galaxies Stop Forming Stars?

Galaxies serve as important laboratories for many subfields of astrophysics. As such, astrophysicists are interested in the full galactic life cycle, from how galaxies are born to the end of their formation. Whereas young galaxies in the distant universe are actively forming stars, many nearby galaxies are much quieter, with little or no ongoing star formation. Astronomers like to describe these galaxies as “red and dead” because their stars are older and therefore appear redder in color. The process by which galaxies shut down their star formation is known as quenching, and it causes galaxies to become quiescent. Understanding the origin of quiescent galaxies is a fundamental question in galaxy evolution. Studying this process is challenging, however, because star formation is intertwined with many other factors that shape galaxies, including their environments and supermassive black holes.

Star formation occurs on scales that are much smaller and over timescales much shorter than the global properties and overall lifetime of a galaxy. By observing star formation on these smaller scales over time, astronomers can gain a clearer picture of how star formation ceases in a galaxy, and they can determine whether this process is driven from the inside out — starting with the supermassive black hole in the galactic center — or from the outside in, influenced by environmental effects. In today’s article, the authors use the Cosmic Assembly Near-Infrared Deep Extragalactic Legacy Survey with high-resolution ultraviolet coverage (UVCANDELS) to examine eight “golden” galaxies, shown in Figure 1. They then reconstruct the star formation histories within small regions of each galaxy to detect the “earliest signs of quenching.”

galaxy sample from Olsen et al.

Figure 1: Images of the eight galaxies in the “Golden Sample” of galaxies studied in this work (right). These galaxies were chosen because of their clear detection, and they show a range of edge-on to face-on views as well as shapes. The authors divided each galaxy into small, resolved regions in which they measured the star formation (example shown on left). [Adapted from Olsen et al. 2026]

Characterizing the Star Formation Histories

To trace a galaxy’s star formation over cosmic time, today’s authors model its integrated light — which encodes the many episodes of past star formation — using spectral energy distribution (SED) fitting. By comparing a galaxy’s light across many wavelengths to computer models, SED fitting can estimate the ages of its stars and when they formed. This approach allows astronomers to study how a galaxy’s star formation rate evolves as a function of its stellar mass, known as the SFR – Mstellar relation. Note that the article considers the star formation rate and stellar mass in individual regions of each galaxy, so it analyzes these properties as surface densities, ΣSFR and ΣMstellar, instead of the total values.

Using SED fitting, the authors determine the expected star formation rate and stellar mass of the regions from 250 million to 1 billion years before the time of observation. Figure 2 shows the line of best fit for the SFR – Mstellar relation of the regions for each galaxy, before and at the time of observation. The SFR – Mstellar relations are very similar across the sample 1 billion years before the observations; however, by the time of observation, the overall star formation rates have decreased slightly, indicating that the galaxies are beginning to quench. They also exhibit increased variation suggesting that each galaxy is quenching in a different way.

best-fit SFR – M stellar relation

Figure 2: The best-fit SFR – Mstellar relation of the regions for each galaxy, represented by the different colored lines, 1 billion years (1 Gyr) before and at the time of observation (left and right panels, respectively). At 1 billion years before the observation, the galaxies appear to have the same relation, but their variation at the time of observation suggests that the galaxies have begun quenching in different ways. [Olsen et al. 2026]

To further probe how these galaxies are starting to quench, the authors also measure the specific star formation rate (sSFR) of the regions as a function of their distance from the center of the galaxy. The sSFR can be thought of as the rate of new star formation relative to the amount of existing stellar mass: SFR /Mstellar. Figure 3 shows three examples from the galaxies studied in this work. Once again, the star formation rates of the regions decrease over the course of 1 billion years. However, the regions where the star formation rate has declined most rapidly differ from galaxy to galaxy — ranging from the central regions to the mid-disk to the outskirts — as each galaxy is undergoing a different quenching mechanism.

specific star formation rate of the regions as a function of their distance from the center of the galaxy

Figure 3: The specific star formation rate of the regions as a function of their distance from the center for three of the galaxies in the sample. Overall, the star formation rate decreases from 1 billion years before the observation (red) to the time of observation (blue). However, the regions where the star formation decreases most significantly (indicated with the yellow boxes) vary from galaxy to galaxy. [Olsen et al. 2026]

Detecting the Earliest Signs of Quenching

It should be noted that all of the galaxies in this study are still forming stars. However, they are on the verge of having their star formation suppressed, offering valuable insight into how quenching begins and how diverse the process can be. New telescopes are pushing the boundaries of what we can observe. On one hand, JWST can capture galaxies at exceptionally high resolution; on the other, the Vera C. Rubin Observatory and the Nancy Grace Roman Space Telescope will deliver vast datasets containing billions of galaxies. Future investigations leveraging these facilities could reveal more about the physical processes that quench galaxies on small scales, shedding light on the mechanisms that drive galaxy evolution.

Original astrobite edited by Catherine Slaughter.

About the author, Shalini Kurinchi-Vendhan:

After studying astrophysics and literature at Caltech, I moved onto a Fulbright Fellowship in Heidelberg, Germany. I’m passionate about using computer simulations to explore supermassive black holes and galaxy evolution — but I also love poetry and traveling.

X-ray dot

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: The X-Ray Dot: Exotic Dust or a Late-Stage Little Red Dot?
Authors: Raphael E. Hviding et al.
First Author’s Institution: Max Planck Institute for Astronomy
Status: Published in ApJL

What Are Little Red Dots?

One of the most intriguing results produced by JWST was the serendipitous discovery of a new class of object: “little red dots” (LRDs). These objects started popping up everywhere in our early universe observations. Surprisingly, they look like compact (little) red dots in imaging data (astronomers are an incredibly creative bunch). The number of LRDs we observe drops drastically at redshifts less than about z = 4 (about 12 billion years ago), implying that LRDs are likely evolving into something else entirely. The luminous, compact nature of LRDs, paired with their disappearance from the cosmic stage, has continuously puzzled astronomers in recent years.

At least part of the LRD population has routinely been explained as a new class of active galactic nuclei (AGN). AGN are the central engines of many galaxies — luminous regions powered by actively accreting supermassive black holes. Recent evidence has increasingly pointed towards LRDs being a new class of AGN: black hole stars, or black holes surrounded by dense cocoons of gas. Naturally, this raises questions surrounding the placement of LRDs in our understanding of supermassive black hole and galaxy evolution across cosmic time.

A hallmark of typical AGNs is their X-ray brightness. X-rays are one of the most energetic types of electromagnetic radiation, and they are primarily produced by the most extreme astrophysical situations (think neutron star mergers). One of the key challenges in explaining the nature of LRDs has been their lack of X-ray emission. Their lack of emission in this regime is a piece of evidence pointing towards a black hole star scenario — dense gas can block the X-rays being produced by the black hole. However, the lack of LRDs in the present-day universe implies that they likely shed their cocoons at some point. Catching an LRD in the act of shedding its cocoon would thus provide an important piece of evidence surrounding their nature and evolution. Thankfully, today’s authors may have done exactly that! They report the discovery of what they call the “X-ray dot” (XRD): an LRD-like object that is also X-ray luminous at a redshift of approximately z = 3.28.

New JWST Observations of the XRD

While archival observations of the XRD with the Hubble Space Telescope, the Canada France Hawaii Telescope, the Spitzer Space Telescope, and the Chandra X-ray Observatory existed, new spectral data from JWST were needed to study the system in depth. Spectral data have a huge advantage over photometric observations when it comes to modeling the system. With a good quality spectrum, astronomers can use sophisticated modeling tools to try and pin down the physical nature of the emission we’re observing, and thus understand its intrinsic nature.

Figure 1 showcases the archival data from Hubble, Spitzer, and Chandra, along with the XRD’s spectrum from JWST (black curve, bottom panel). Photometric observations of the source indicate that it is incredibly compact (radius ≲ 250 pc, about 100 times more compact than the Milky Way) and shows similarities to LRD spectra, but with some key differences. (An example LRD spectrum is shown as the red curve in Figure 1.) The impressive X-ray luminosity of the XRD — typical of standard AGNs — is another key difference between it and standard LRDs.

A figure showing image cutouts from HST, Spitzer, and Chandra, of the XRD. The bottom panels also shows its spectrum alongside a typical LRD spectrum.

Figure 1: Top: Image cutouts from various Hubble, Spitzer, and Chandra observations of the XRD. Bottom: The spectrum of the XRD (black) shown alongside a similar LRD spectrum (red) and two quasar spectra (blue and purple), one of which has been reddened due to the presence of dust (purple). [Hviding et al. 2026]

So…What’s the Deal with the XRD?

To understand the nature of the XRD, the authors fit a variety of models describing a wide range of physical systems to the available data. Surprisingly, their best-fit models indicate that if the XRD is simply a typical AGN heavily obscured by astrophysical dust, its dust properties are drastically different from those of typical galaxies and AGN. Instead, trying to explain the system as an AGN embedded in a cocoon of gas (the black hole star model) provides better results (more aligned with observed LRDs), but it still isn’t perfect.

Notably, the emission in the ultraviolet-to-optical regime of the electromagnetic spectrum differs greatly from that of LRDs. In LRDs, this emission is indicative of a single dense gas component around the supermassive black hole, while in the XRD the authors find evidence suggestive of a patchier, less uniform distribution of gas. However, in order for this explanation to match the data, physical conditions of the model must be finely tuned, suggesting that this model may need to be refined further. These factors seem to suggest that the XRD is poorly understood in the context of our current paradigm of models describing AGNs and LRDs. The potentially patchy nature of the XRD’s gas envelope could suggest that this object is an LRD in the process of shedding its outer envelope, evolving into a typical AGN.

Regardless of its true nature, the XRD opens up new doors in our understanding of AGNs and LRDs. It provides an exciting glimpse at AGN evolution in action — a transitional fossil for early universe black holes. With a new piece of the puzzle slotted into our picture of AGN evolution, it’s only a matter of time before astronomers fully contextualize the stubbornly enigmatic LRDs.

Original astrobite edited by Ansh R. Gupta.

About the author, Drew Lapeer:

Drew is a first-year PhD student at the University of Massachusetts Amherst. They are broadly interested in the evolution of galaxies, with a focus on the impact of cosmic feedback on the galactic ecosystem. In their free time, they enjoy reading, rock climbing, hiking, and baking!

Large Magellanic Cloud

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: Protostars at Subsolar Metallicity: First Detection of Large Solid-State Complex Organic Molecules in the Large Magellanic Cloud
Authors: Marta Sewiło et al.
First Author’s Institution: NASA Goddard Space Flight Center
Status: Published in ApJL

Stars and planets form when clouds of molecules collapse in on themselves due to the overwhelming force of gravity (which, relatable). This process is quite mysterious, as these clouds will only live for a couple million years before collapsing, and the newly forming protostars only live for about half a million years before becoming main-sequence stars.

But we are able to find some protostars in the Milky Way and in the Small and Large Magellanic Clouds near us, and observing the molecules around these protostars can help us understand how protostars form.

To look for molecules around protostars, astronomers take spectra of the protostars in the infrared, which is a wavelength region where molecules emit a lot of light. As better instruments are being built, astronomers are more often able to look for what are called complex organic molecules, which are carbon-bearing molecules with at least six atoms.

These complex organic molecules can be detected in different states: most detections have been of gaseous molecules in and around protostars or planet-forming disks, but they can also be found in a solid state on the surfaces of dust grains in the interstellar medium in a galaxy. In this state, they are called complex organic molecule “ices.” These “ices” were rarely found pre-JWST, but since its launch, detections have been more numerous, mainly in the Milky Way. Observing these complex molecules on the surfaces of dust grains tells us more about protostellar dust chemistry, which is an important but still mysterious aspect of star formation.

In this study, researchers look outside of our own galaxy to explore the complex organic molecules in the Large Magellanic Cloud (LMC). The LMC is an interesting place to look for these molecules because it has lower metal content (i.e., elements heavier than hydrogen and helium) than the Milky Way, as well as a harder radiation field. This means photons flying around in the LMC generally have higher energies than photons flying around in the Milky Way, and elements like carbon, oxygen, and nitrogen are less abundant. This is important because these conditions are more similar to what typical galaxies were like in the early universe. Thus the LMC is a better test bed to look at protostars than the Milky Way if we want to understand how stars formed in the early universe.

The authors of this study have found signatures of complex organic molecules in protostars forming in the LMC. They focus on the protostar ST6 in the LMC, shown in the right image in Figure 1. They used JWST’s Mid-Infrared Instrument to take spectra of the region around ST6.

Large Magellanic Cloud and N158

Figure 1: The LMC (left) and a zoomed-in view of region N158 (right) containing protostar ST6, the focus of this study. In the inset panel of the right picture, contours of carbon monoxide emission are shown. [Sewiło et al. 2025]

Spectroscopic Signatures of Molecules on Dust Grains

The researchers used the Python tool called dEcompositioN of Infrared Ice features using Genetic Modelling Algorithms (ENIIGMA) to fit the spectroscopy. This code compares the spectra to laboratory data of these complex organic molecules at the temperatures we see in the LMC. The authors fit for complex organic molecules they might expect to see, and they detect methanol (CH3OH), acetaldehyde (CH3CHO), ethanol (CH3CH2OH), methyl formate (HCOOCH3), and acetic acid (CH3COOH); of these, acetaldehyde, ethanol, and methyl formate have never before been detected outside of the Milky Way. Interestingly, this is the first time acetic acid has been definitively detected in any astrophysical environment. The spectral contribution from each molecule is shown in Figure 2; all the individual contributions added together make up the observed spectrum.

spectrum of the protostar ST6

Figure 2: Model fits to the spectrum of the protostar ST6 using the code ENIIGMA. Molecules and their fitted emission are given in the legend. [Sewiło et al. 2025]

These detections provide evidence that these complex organic molecules can be produced on the surfaces of dust grains. Further, their detection in the LMC provides evidence that these molecules can be produced in “harsh” environments like the LMC, with lower metal content and harder radiation fields.

When the authors compared the abundances of these ices to other protostars in the Milky Way galaxy, they found that the abundances are slightly lower around the LMC protostar. This may happen because the dust temperatures are generally higher in the LMC due to the higher-energy photons, which affect the chemistry happening on the dust. Some species they found, however, do have similar abundances to the Milky Way protostars, indicating that the harder radiation field of the LMC does not affect how these species form. Finding more complex organic molecules in the Magellanic Clouds will help us understand further how these compounds form on dust grains and how the galactic environment may affect their formation.

Original astrobite edited by Shalini Kurinchi-Vendhan.

About the author, Caroline von Raesfeld:

I’m a third-year PhD student at Northwestern University. My research explores how we can better understand high-redshift galaxy spectra using observations and modeling. In my free time, I love to read, write, and learn about history.

Illustration of two white dwarfs about to merge

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: Fast Radio Bursts from White Dwarf Binary Mergers: Isolated and Triple-Induced Channels
Authors: Cheyanne Shariat et al.
First Author’s Institution: California Institute of Technology
Status: Published in ApJL

The History of Fast Radio Bursts

Since the discovery of the Lorimer Burst in 2007, astronomers have studied fast radio bursts (FRBs) in extreme detail. However, these extragalactic millisecond-duration radio transients have remained an enigma. Like many astrophysical phenomena, their discovery was serendipitous. The most prolific FRB-discovery instrument, CHIME, was not originally built to study them; the acronym stands for “Canadian Hydrogen Intensity Mapping Experiment,” reflecting its original (and ongoing) mission to map the cosmological distribution of hydrogen in the primordial universe. Since its construction, CHIME has discovered several thousand FRBs and even identified a smaller fraction that are known to repeat, some with consistent periods and others that are sporadic and seemingly random. (Check out their live catalog of repeating sources.)

Despite having thousands of sources to study, the only thing we know fairly confidently about FRBs is that many of them are extragalactic, and some others are even at cosmological distances. If we don’t understand how FRBs are produced, how can we know their distance? FRBs display a unique signature in their radio spectra, where higher frequencies arrive at the telescope first and lower frequencies arrive later, delayed by an inverse frequency-squared relationship. In a “dynamic spectrum,” this makes the pulse actually appear like a sweep from high to low frequency that follows the shape of a parabola; for example, the brightest pixels in Figure 1 of this Bite display this sweeping shape. This effect results from the FRB propagating through an ionized plasma with a specific density. Since the FRB is ultimately an electromagnetic wave, this slows down the speed of the FRB as a function of frequency. Measuring this effect, called dispersion, allows us to infer that the FRB signal passed through more plasma than can be accounted for within our own Milky Way galaxy, indicating that the FRB must have originated from much farther away. So far, there is only one example of an FRB from within our own galaxy, which was identified as originating from a magnetar, a highly magnetic pulsar.

FRB Progenitor Candidates

Although we know that most FRBs are located at extragalactic or even cosmological distances, it’s still not clear what objects that can generate them. Since magnetars and other compact objects are often considered reliable progenitors of FRBs, understanding their populations and comparing them with observed FRBs may reveal the origin of FRBs. Candidate FRB progenitor systems are usually divided into “prompt” channels — like the collapse of a star into a magnetar during a core-collapse supernova — and “delayed” channels where the FRB engine is formed on longer timescales, like accretion-induced collapse of white dwarfs or compact-object mergers. Figure 1 displays several of these channels, particularly those considered in today’s article, in which stellar evolution results in the transfer of material from one star to another (accretion-induced collapse) or gravitational radiation allows two objects to inspiral and merge. These events could create highly magnetized white dwarfs or magnetars that may produce FRBs.

different pathways and their associated fractions for producing an FRB candidate

Figure 1: Three different pathways and their associated fractions for producing an FRB candidate. Depending on the mass and composition of the merging objects, an FRB candidate may be produced instead of causing the white dwarf to explode in a Type Ia supernova. These are the systems of interest in today’s article. [Shariat et al. 2026]

Simulating Hierarchical Triples

In today’s article, we’re considering the role of hierarchical triple star systems in producing FRB-emitting objects. A hierarchical triple consists of an inner binary and a distant, tertiary star; see the first panel of Figure 1. These systems may play an important role because many stars that become white dwarfs of at least 0.8 solar mass exist in hierarchical triples: 35% of 2-solar-mass stars and 50% of 8-solar-mass stars are in triples. Stars of both masses will end their lives as white dwarfs.

The tertiary star plays an essential role in increasing the likelihood of the inner binary interacting: the eccentric Kozai–Lidov (EKL) mechanism causes the inner orbit to oscillate between highly inclined (with respect to the orbit of the tertiary star) and highly eccentric. Unlike isolated binaries that aren’t excited in this way, the eccentric orbit phase causes the inner stars to pass much closer to each other, increasing the likelihood of interaction or even a merger that could produce an FRB-emitting remnant.

The authors of today’s article produced a realistic population of triple systems by sampling from empirically measured distributions of masses, periods, and orbital eccentricities. These distributions were derived from observations collected by the Gaia satellite, an all-sky optical telescope that has now observed more than one billion stars in the Milky Way. The authors also created a “control” sample of isolated binaries by removing the tertiary star from the same sample of systems. The systems were then evolved for an age randomly selected between 0 and 12.5 billion years to simulate “constant” star formation similar to that in the Milky Way. Finally, they reapply these simulations at three different “metallicities” — the fraction of elements other than hydrogen or helium relative to that of the Sun.

Figure 2 displays these two populations: the isolated binaries are shown as filled histograms, and the hierarchical triples are shown as histograms with dashed lines. Each column represents the three different progenitor classes considered in Figure 1. At solar metallicity, the delays to merging are greater in triples than in secular binaries, leading to more mergers overall (top left panel). At lower metallicities, there are fewer mergers in all progenitor pathways in triples than in isolated binaries due to changes in stellar evolution attributed to weaker winds and larger pre-white dwarf masses.

mergers as a function of time for different metallicities and formation pathways

Figure 2: The number of mergers as a function of time from the hierarchical triple’s formation, given different three metallicities (rows) and the three primary formation pathways for FRB-emitting merged remnants considered in Figure 1. Filled histograms are isolated binaries, dashed histograms are triples. [Shariat et al. 2026]

A Cosmic Population of FRBs

Since we can’t realistically track individual star systems from their formation to the end, we need to convert the delay-time distributions shown in Figure 2 to an observable. To do this, the authors combine the results of their simulations with measurements of the cosmic star formation rate — how many stars form per unit time as a function of the universe’s age — and the evolution of the universe’s metallicity as stars fuse hydrogen into heavier elements and return this material to the interstellar medium as these stars end their lives in shedding red giants or supernovae. This analysis produces the number of FRBs that should be observed as a function of distance or lookback time into the universe’s history, which is displayed in Figure 3.

number of FRBs expected as a function of cosmological redshift

Figure 3: The number of FRBs expected as a function of cosmological redshift. The present is to the left, the distant and early universe is on the right. The top panel shows the three formation channels, with binary and triple systems separated. The bottom panel displays the composite rate and compares it with empirical estimates from existing FRB observations. [Shariat et al. 2026]

This analysis reveals key measurable differences between lone binaries and hierarchical triples: most notably, binary white dwarf mergers peak later in the universe (redshift of z ~ 1), while single binaries peak earlier (redshift of z ~ 2). These results highlight the importance of including triples in FRB progenitors because they are more likely to contribute to populations at later times — making them closer and more easily detectable — in the total population of FRBs. Future instruments like CHIME with the new outriggers and the Deep Synoptic Array will expand the sample of FRBs to reach farther cosmological distances, with better localizations, enabling more statistically motivated tests of different formation pathways.

Original astrobite edited by Catherine Slaughter.

About the author, Will Golay:

I am a graduate student in the Department of Astronomy at Harvard University and the Center for Astrophysics | Harvard & Smithsonian, advised by Edo Berger. I study radio emission from transient astrophysical objects like tidal disruption events.

G272.2-3.2

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: Supernova Ia Remnants with M-Dwarf Surviving Companions
Authors: Kuo-Chuan Pan (潘國全), Pilar Ruiz-Lapuente, and Jonay I. González Hernández
First Author’s Institution: National Tsing Hua University
Status: Published in ApJ

Type Ia supernovae are important tools astronomers use to measure distances in space. Despite their important role as “standardizable candles,” their origins remain hotly debated. It’s long been theorized that a Type Ia supernova can develop from one of two pathways: the single-degenerate channel or the double-degenerate channel. In the single-degenerate channel, a white dwarf and a non-degenerate star (such as a red giant or main-sequence star like our Sun) orbit each other in a close binary system. The white dwarf accretes material from the outer surface of the non-degenerate star until it reaches a critical mass and explodes. The double-degenerate channel begins instead with two white dwarfs orbiting each other. Most explosion mechanisms from this channel result in the complete obliteration of the secondary, less-massive white dwarf. However, one theoretical explosion mechanism, called the dynamically driven double-degenerate double-detonation (D6) channel, might allow the secondary companion star to survive an explosion from a double-degenerate Type Ia supernova.

Generally, due to the immense energy from a supernova explosion, we expect these surviving companion stars to be “kicked” away from the blasts at high velocities. Fortunately, astronomers have found evidence of these surviving companion stars. As covered in a previous Astrobite, three hypervelocity white dwarfs have been found in Gaia data and traced back to the locations of double-degenerate Type Ia supernovae. Given the prevalence of Type Ia supernovae, however, astronomers are led to question why we haven’t seen many more of these candidates. It’s additionally important to wonder what similar “runaway” companions might look like for the single-degenerate channel.

Recent observations by the Gaia spacecraft near the supernova remnant G272.2-3.2 show an M-dwarf star, MV-G272, with an unusually high velocity around 240 km/s and a trajectory tracing back to the center of the supernova remnant. Is it possible for this M dwarf to be the runaway companion to the remnant’s original supernova? The authors of today’s article look at simulations of potential companion stars from single-degenerate Type Ia supernovae and compare their results to this observed potential companion candidate.

Because this work focuses on the potential for M-dwarf companions, the authors explored a range of simulated companion-star masses and radii characteristic of M dwarfs: 0.5, 0.6, and 0.7 solar mass and 0.396, 0.454, and 0.483 solar radius, respectively. To make these models, they used the stellar evolution code Modules for Experiments in Stellar Astrophysics (MESA) to construct and evolve a number of simulated stars with these specifications.

After setting up the models in MESA, they used the magnetohydrodynamics code FLASH to perform 2D and 3D simulations of the supernova explosion and the subsequent impact of the exploding star’s ejecta (or hot material) on its M-dwarf companion. After the explosion and impact were modeled, the companion star’s evolution was further followed in MESA.

Figure 1 shows the density of the supernova’s ejecta, or exploding material, as it impacts the companion star. (You can think of this figure as if you were looking at the companion star out in space.) After roughly 100 seconds, you can see a bow shock develop around the top of the companion (or the “front” of the star) as it wraps around the companion. This impact causes significant compression of the companion star and rips away upwards of 50% the companion’s mass. The impact of this shock is what gives the companion star its “kick,” sending it hurtling through space.

simulations of supernova ejecta colliding with a star

Figure 1: Visual of the density of the companion star (center of each plot) and the supernova ejecta over time. In the top row of plots and the bottom-left plot, you can see the higher-density bow shock bent around the top (or front) of the companion star. In the bottom-right plot, you can see some of the supernova ejecta and unbound mass from the companion star settling around the companion star. [Pan et al. 2025]

Because the authors explored a range of initial companion masses and distances between the primary and companion star, their results show some range. Figure 2 shows two important results: the evolution of the companion’s mass before and after the explosion (left plot), and the evolution of the companion’s kick velocity (right plot). The dip in mass around 100 seconds (102 s) indicates the impact of the bow shock. Even though the simulations model companions with a relatively narrow mass range (0.3–0.7 solar mass), there is considerable variation in the final masses after explosion. For example, the initial 0.5-solar-mass model ends up with anywhere between 0.07 solar mass (just 14% of its initial mass) and almost its full initial mass of 0.5 solar mass, depending on the distance between the companion and its exploding star.

plots of bound mass and kick velocity as a function of time

Figure 2: The different models shown in both plots indicate different initial masses and separation distances between the companion and the exploding star. Left: how the companion’s gravitationally bound mass evolves over time. The plot can be read left to right. The dip around 102 seconds indicates the impact of the supernova shock, which unbinds layers of the companion’s mass. Some of that mass becomes bound again to the companion after the shock passes. Right: how the companion star’s kick velocity changes based on initial mass and separation distance between companion and exploding star. The peak in values between 102 and 103 seconds indicates the impact, or “kick,” from the supernova. [Pan et al. 2025]

The results from this work generally show, predictably, that a companion closer to its primary star gets a bigger kick in velocity. These “kicks” range from about 50 km/s to 150 km/s. When added to the system’s initial velocity in space, it brings the total velocity of these companions into the 150–200 km/s range. This result is generally well aligned with the observed velocity of MV-G272, which was roughly 240 km/s.

Notably, runaway companion stars of these Type Ia supernovae seem to retain information from the initial explosion. Given the final velocities of these companions, astronomers might be able to figure out the companion’s initial mass and trace back its trajectory to the system it originated from. Overall, the fact that MV-G272 has a trajectory that tracks back to the center of a supernova remnant makes it a good companion candidate. These results support the feasibility of M-dwarf runaway companions from single-degenerate Type Ia supernova systems. Now that we know what these companions might look like — and how fast they can be going — astronomers can hopefully better search for these high-velocity stars in the future.

Original astrobite edited by Catherine Slaughter.

About the author, Mckenzie Ferrari:

I’m currently a PhD student in the geophysical sciences program at the University of Chicago. While I now study the atmosphere and oceans of Earth, most of my previous research focused on simulations of Type Ia supernovae and galaxy formation and evolution. In my free time, I foster cats for a local organization, enjoy cooking, and can often be found running along Lake Michigan.

illustration of an accreting black hole

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at astrobites.org.

Title: The z = 7.08 quasar ULAS J1120+0641 May Never Reach a “Normal” Black Hole to Stellar Mass Ratio
Authors: Meredith A. Stone et al.
First Author’s Institution: University of Arizona
Status: Published in ApJ

The first galaxies in the universe emerged out of an archaic age of darkness. Like luminous factories, they started rapidly churning out stars, spilling light into the cosmos. This process of star formation continues to the present day. In some galaxies, it occurred at astonishing rates, with the equivalent of thousands of times the mass of the Sun being assembled into stars in a single year. At other times, it came to a slow crawl, with all fresh reserves of star-building gas seemingly exhausted. Yet when the gears of star formation ground to a halt, certain events were able to send them spinning again.

One of the most spectacular of these is a merger, a train-wreck collision between two galaxies. Images of these events are often breathtaking, showing arms of light twisting together and vast stellar streams being violently cast away. Mergers directly increase the stellar mass (the mass of all stars) of galaxies by combining two into one. The intense gravitational interactions between the galaxies involved can also disrupt stagnant pools of gas, igniting a new burst of star formation. Thus, interactions between galaxies can be a significant driver of their growth.

Sometimes, a reservoir of matter is driven toward the center of a galaxy due to a “secular” process (one that occurs without external intervention) or from a merger. At the galactic nucleus, the infalling streams of matter encounter a supermassive black hole. The matter falls into a swirling accretion disk, heating to extreme temperatures and furiously glowing. If the black hole is massive enough and the gas falls in quickly enough, the accretion disk can outshine all of the stars in the galaxy, and we call such a source a quasar. The intense radiation produced by a quasar can impact the entire rest of the galaxy, either suppressing the formation of new stars or sometimes igniting it.

Thus, a galaxy has an important role in the growth of its central supermassive black hole, and vice versa. This connection manifests in complicated ways. Most strikingly, it is thought to be the cause of the correlation between a galaxy’s stellar mass and its central black hole mass. In the nearby universe, galaxies have stellar masses around a thousand times greater than their black hole masses. JWST, our most powerful tool for investigating early galaxies, has given us unprecedented insight into whether this holds up in the distant universe. Though the jury is still out, many researchers are convinced that the answer is a resounding no. We’ve found distant black holes that appear to be nearly equal in mass to their host galaxies, something that’s unfathomable in the local universe.

However, this raises a key question. Black holes and their galaxies grow together, and they started on nearly equal footing. So why do all of the galaxies in the nearby universe appear so much more massive than their black holes? Today’s article underscores this puzzle, pushing it to the breaking point. The authors examine a well-studied quasar, ULAS J1120+0641 (hereafter J1120), which we see as it appeared just 750 million years after the Big Bang. This source has a supermassive black hole containing an astounding 20% the mass of its host galaxy. The shocking part is that, even after carefully accounting for all of the ways this galaxy could grow in the future, there appears to be no way for J1120 ever to reach “normal” proportions.

Through the Looking Glass, ULAS J1120

First, the authors analyze the key observations of J1120 obtained to date. In addition to observations from JWST, this source has been targeted by some of the most powerful astronomical facilities, including the Hubble Space Telescope (optical and near-ultraviolet), the Chandra and XMM-Newton observatories (X-rays), the Atacama Large Millimeter/submillimeter Array (radio/millimeter), and numerous other ground-based telescopes. This incredible dataset provides a detailed picture of J1120 across the electromagnetic spectrum, spanning X-rays to radio emission.

The authors compare various measurements of the black hole’s mass and find that multiple methods yield a similar result of about a billion and a half times the mass of the Sun. Turning to the host galaxy, they estimate the mass of the visible stars, which is more complicated than it sounds — you can’t exactly count them, since the enormous distance between us and J1120 causes it to appear as just a fuzzy blob. They combine this mass with the mass of newly forming stars that are buried in dust and obscured from our view. The mass of a nearby galaxy that’s actively merging with J1120 is added to this sum. Finally, the authors compute how many stars could form in the future from all of the gas left in the galaxy, and they throw this into the total.

Even with all of these components, the host galaxy would only be 10 times more massive than the black hole at the center. This ratio is a far cry from the factor of a thousand we expect for normal galaxies in the local universe!

A Train Wreck of Train Wrecks

Well, if one merging galaxy won’t do it, what if J1120 merges with literally every single other nearby galaxy? The authors use new JWST data to count all sources near the quasar (Figure 1) and measure their stellar masses. Going further, they estimate how many galaxies their technique may have missed and add all of these to the pile. Not done yet, they finally conclude by making a rough approximation of how much loose gas is in these nearby galaxies, and compute how much extra mass could exist in stars born from this fresh material. Assuming that J1120 survives the full-frontal assault, being battered by mergers from each of its neighbors in turn, its total stellar mass would grow by quite a large factor.

J1120 and nearby galaxies

Figure 1: All of the galaxies found by JWST that are close enough to J1120 to potentially merge with it in the future. The circles mark confident detections, whereas the squares represent galaxies for which the exact distances couldn’t be measured confidently. What would happen if all of these galaxies crashed into J1120? [Stone et al. 2025]

What’s the final result? Shockingly, even after all of this growth, the stellar mass of J1120 would still only be about 40 times that of the central black hole. That’s 25 times less than we’d expect from a nearby galaxy! Even if the black hole never grows again, it appears that this galaxy will never reach “normal” proportions. So why is it that, when we look around at the local universe, there’s nothing like J1120? The authors point out that galaxies near J1120 would be missed by their methods if they are obscured by thick shrouds of gas and dust. Such sources would increase the estimated mass that could be contributed by mergers. On the other hand, maybe J1120 will adopt the strategy of the blue whale, filter feeding on the extremely sparse matter between galaxies, yet growing to enormous proportions.

A fascinating final possibility is that galaxies like this one do exist in the nearby universe. With their black hole switched off, all we would be able to see is a relatively small package of stars. Place such an object far enough away, and we might miss it entirely! Whether galaxies like J1120 achieved normal proportions through unknown mechanisms or are lurking in the far fringes of the local universe, studies like this one shine light on the amazing connection between black holes and their host galaxies. In the future, they will be one key piece in the puzzle of the endlessly complicated past of our own galaxy and the universe at large.

Original astrobite edited by Will Golay.

About the author, Ansh Gupta:

I’m an astronomy graduate student at the University of Texas at Austin working with Steven Finkelstein. I use data from JWST to study the formation and growth of the first galaxies and black holes in the universe. In my spare time, I enjoy playing piano, reading, and making YouTube videos.

1 2 3 … 48