Listening to the Stars: Predicting Massive Star Properties with Machine Learning
Rachel Zhang and collaborators tested whether machine learning can estimate properties of massive O-type stars from TESS light curves. Using spectroscopic data from the IACOB project, they compared two approaches: neural networks trained on simple “red noise” parameters versus convolutional networks trained on full periodograms. The latter performed much better, showing that light curves contain enough information to predict stellar temperatures and luminosities, a valuable tool for future large surveys.
Counting the Milky Way’s Hidden Satellites: The DELVE Census
Tan and collaborators present the DELVE Milky Way Satellite Census, combining DES, DELVE, and PS1 data to search for faint companion galaxies. Using strict detection methods and efficiency tests, they recovered 49 known satellites and predicted about 265 in total. The results show satellites cluster near the Large Magellanic Cloud and broadly match cosmological predictions, offering key insights into galaxy formation and dark matter.
A Thick Blanket for Early Mars: Evidence of a Massive Primordial Atmosphere
Sarah Joiret and colleagues show that Mars once had a massive primordial atmosphere, captured from the solar nebula. By modeling comet impacts and comparing expected xenon delivery with today’s measurements, they find Mars’s atmosphere must have been at least 2.9–14.5 bar. This thick hydrogen-rich blanket may have warmed early Mars and shaped its volatile history.
Tracing the Milky Way’s Past with HDBSCAN: Finding the Ghosts of Ancient Galaxies
Andrea Sante and collaborators test the HDBSCAN clustering algorithm to trace the Milky Way’s merger history using Auriga simulations. By optimizing parameters and using a 12-dimensional feature space, they show HDBSCAN reliably identifies recent stellar streams but struggles with older, well-mixed debris. Contamination from stars formed inside the Milky Way further limits recovery, though cluster purity remains high.
Following the Tides: Stellar Streams in Open Clusters with Gaia DR3
Ira Sharma and collaborators used Gaia DR3 data and machine learning to detect tidal tails in five open clusters. These stellar streams, spanning 40–100 parsecs and containing up to 200 stars, lacked massive stars but showed higher binary fractions. The team also detected rotation in M67 and NGC 2281, estimating cluster masses with Plummer models. Their methods expand tidal tail studies to more distant clusters, improving our understanding of cluster evolution.
How the Milky Way’s Disc Survived a Cosmic Collision
This paper explores how the Milky Way’s disc formed and survived an ancient collision with the Gaia-Sausage-Enceladus galaxy. Using simulations and Gaia data, Orkney and colleagues show the disc was already spinning about 11 billion years ago and that the merger was likely minor, not major. Despite disruption, the disc reformed, with starbursts and globular clusters marking the event’s impact.
Tracing Stellar Mergers with Chemistry: Carbon Isotopes Reveal Clues to Mysterious Stars
Astronomers studied massive α-enriched (MAE) stars, which look chemically old but appear too massive to be ancient. By measuring carbon isotope ratios (¹²C/¹³C), Zachary Maas and collaborators found most MAE stars resemble thick disk stars, while a few show evidence of mass transfer or mergers. The results suggest MAE stars form through multiple pathways, with carbon isotopes serving as key clues to their hidden histories.
Makemake’s Hidden Activity: JWST Finds Methane Gas and Hydrocarbon Ices
JWST observations show that Makemake’s surface holds methane, ethane, acetylene, and possibly methanol, arranged in layered ices. The telescope also detected methane gas, either from a thin atmosphere or plume-like outgassing. These findings suggest Makemake is more active and chemically complex than once thought, challenging its image as a frozen, inactive world.
Exploring the Coldest Brown Dwarfs with Near-Infrared Colors
Leggett and collaborators use JWST data to study extremely cold Y dwarfs, comparing their near-infrared colors across JWST, Euclid, and Roman filters. They show that mid-infrared brightness at 4.6 microns reliably tracks temperature, while near-infrared colors vary with metallicity and gravity. The work highlights both the promise of upcoming surveys and the challenges of incomplete atmospheric models.
A Closer Look at Carbon-Chain Depleted Comets
Allison Bair and David Schleicher analyzed 17 strongly carbon-chain depleted comets using Lowell Observatory’s long-term photometry database. These comets, mostly Jupiter-family objects, show far lower levels of C₂ and C₃ (and sometimes NH) compared to typical comets, though their dust-to-gas ratios are similar. The authors argue this depletion reflects primordial formation conditions in the Kuiper Belt, rather than later heating in the inner Solar System.
Carbon-Enhanced Dwarf Stars: Clues from the Galactic Halo
This study analyzed over 1,000 dwarf carbon stars using SDSS and Gaia data, providing the first reliable distances for such a large sample. The results show that about 60% belong to the Milky Way’s halo and 30% to the thick disc, confirming they are mostly old, metal-poor stars. These findings establish dwarf carbon stars as valuable tracers of the Galaxy’s early history and stellar evolution.
Catching Makemake’s Shadow: A New Look at Its Mysterious Moon
Daniel Bamberger reanalyzed Hubble images of Makemake and its moon MK2, finding an 18-day orbit nearly edge-on to Earth. This alignment could mean eclipses and transits are happening now, offering a rare chance to study the system’s size and surface features. Preliminary results also suggest Makemake is slightly less dense than earlier estimates.
Faint Streams Hidden in Plain Sight: What the Mass–Metallicity Relation Tells Us About Tidal Disruption
Alexander Riley and collaborators use the Auriga simulations to test whether the mass–metallicity relation of galaxies rules out tidal disruption. They find that even heavily stripped satellites still follow the relation with little scatter, matching what’s seen in the Milky Way and Andromeda. This suggests many Local Group satellites have lost large fractions of their stars, and faint tidal streams may be revealed by future surveys.
The Universe’s Hidden Patterns: Fractals in the Cosmic Web
The paper by Jaan Einasto reviews how the Universe’s large-scale structure, the cosmic web, shows fractal-like patterns. Using the ΛCDM model, it explains that galaxies form clusters, filaments, and voids whose distribution follows power laws across scales. While small and medium scales reveal fractal behavior, larger scales smooth out into homogeneity, supporting the cosmological principle.
Untangling the Milky Way’s Halo with Aluminum
This study by Ernandes and collaborators shows that aluminum abundances ([Al/Fe]) are a powerful way to distinguish between stars formed inside the Milky Way and those accreted from dwarf galaxies. Using high-resolution spectra, they demonstrate that aluminum provides a cleaner separation than other elements, even at low metallicities. Their results refine previous classifications and highlight aluminum as a key tracer for unraveling the Galaxy’s merger history.
Spotting Satellite Streaks with Deep Learning
Hua-Jian Yu and colleagues developed ASA-U-Net, a deep learning model to detect satellite trails in telescope images. Using data from the Mephisto Telescope, the model combines channel attention and multi-scale pooling to spot faint and complex streaks. Tests show ASA-U-Net outperforms standard methods, reducing false detections and improving recall, making it a promising tool for cleaning astronomical survey data.
Building Worlds: How Protoplanetary Disk Chemistry Shapes Rocky Planets
Spaargaren and colleagues show that rocky planet compositions depend strongly on the chemistry of their birth disks. Using simulations of condensation for 1,000 stellar compositions, they find that Earth-like planets form in low carbon-to-oxygen disks, while higher ratios yield graphite-rich or metal-heavy planets. Their results suggest rocky exoplanets are far more chemically diverse than previously assumed.
How Supernova Explosions May Have Stopped Star Formation Near the Sun
Leonard Romano’s study revises the history of the Local Bubble, finding it to be only 3.5–5.5 million years old and powered by about 19–30 supernovae, not 14 million years and fewer explosions as once thought. Using 3D dust maps and simulations, the work shows the bubble’s rapid expansion likely quenched star formation near the Sun, challenging earlier claims that it triggered new stars.
Mapping Jupiter’s Skies: A Full-Atmosphere Model
Antonín Knížek and colleagues built the first full-atmosphere model of Jupiter, combining deep thermochemistry with upper-atmosphere photochemistry. The model predicts a mixed ammonia–ammonium hydrosulfide cloud layer, stable nitrogen levels from quenching, and a stratospheric region where hydrogen cyanide forms at detectable levels. These results bridge gaps between earlier models and make new, testable predictions for future missions.
Tracking Potassium in the Oldest Stars: What It Tells Us About Stellar Explosions
Miho Ishigaki and collaborators measured potassium in extremely metal-poor stars using the Subaru Telescope. They found that potassium-to-iron and potassium-to-calcium ratios were consistently enhanced with little scatter, unlike sodium-to-magnesium ratios, which varied widely. These results suggest potassium is produced through stable processes in massive stars and supernovae, making it a valuable tracer of how the earliest stars ended their lives.