Editor’s Note: This week we’re at the 248th AAS meeting in Pasadena, CA. Along with a team of authors from Astrobites, we will be writing updates on selected events at the meeting and posting each day. Follow along here or at astrobites.com for daily summaries, or follow @astrobites.bsky.social on Bluesky for more coverage. The usual posting schedule for AAS Nova will resume on 22 June.

Table of Contents:

• Fred Kavli Plenary Lecture: Richard Teague, Revealing the Dynamics of Planet Formation
• Press Conference: The Science of Stellar Remnants
• Plenary Lecture: Sanmi (Oluwasanmi) Koyejo, The Measurement Gap: What AI Can Get Wrong and Why Astronomers Are the Fix
• Press Conference: Variable, Windy, and Disruptive: The Behavior of Supermassive Black Holes
• Plenary Lecture: Mario Juric, Early Results from the Rubin Observatory: Mapping the Solar System and Beyond
• Plenary Lecture: Esra Bulbul, Cosmology and Baryons Across Scales: Probing Halo Mass Function to Cosmic Filaments with eROSITA

---

Fred Kavli Plenary Lecture: Richard Teague, Revealing the Dynamics of Planet Formation (by Lexi Gault)

Richard Teague (Massachusetts Institute of Technology) gave the Fred Kavli Plenary Lecture on his groundbreaking work in the dynamics of planet formation. The study of planet formation, as Teague introduced, has advanced significantly in the last two decades with the discovery of thousands of exoplanets and the launch of cutting-edge instruments. While planets are found everywhere orbiting a wide variety of stars, not all planetary systems look the same. Our solar system looks very different from other exoplanetary systems, challenging our understanding of planet formation and requiring us to build a picture that works across the board.

Where do we look to build a better understanding of planet formation? Planets are inherently the byproduct of star formation, forming in protoplanetary disks around new stars. These disks are filled with gas and dust, concentrating mass into rings that then eventually form planets. Exactly how this process proceeds is variable, and the resulting planetary system depends on how the material in the protoplanetary disk mixes and combines into planets. Teague compared planets to bread: we know the simple list of ingredients that it takes to make them, but what you end up with depends on how you bring those same ingredients together. We need to study what is happening in the planetary mixing bowls to understand how planets form.

Leveraging the high spatial and spectral resolution of the Atacama Large Millimeter/submillimeter Array (ALMA), Teague and collaborators have developed exoALMA, a planet-hunting campaign focusing on the still-forming planets of protoplanetary disks. This project has focused on 15 sources, obtaining and analyzing high-resolution observations to explore the kinematics, 3D velocity structures, and mass distributions within these protoplanetary disks.

The first science results from exoALMA have started to reveal that the planet formation process is highly dynamical — precise kinematic measurements uncover dynamical perturbations within protoplanetary disks across their sample. These perturbations point to the impacts protoplanets have on disk kinematics and mass distributions as they continue to form, and the high resolution of ALMA observations unlocks the inner workings of protoplanetary disks.

Looking forward, Teague emphasized that the future is broad for planet formation science. The upcoming Wideband Sensitivity Upgrade on ALMA will enable even more detailed maps of protoplanetary disks, unlocking the mysteries of planet formation while it happens.

Return to Table of Contents.

---

Press Conference: The Science of Stellar Remnants (Briefing video) (by Lucas Brown)

In AAS 248’s very first press conference session, we heard from a variety of scientists and research teams on all sorts of stellar remnants: white dwarfs, supernovae, pulsars, and beyond. Basically, if it came from a dying star, it was fair game for this session.

Boron and Beryllium as Unlikely Probes of Extrasolar Planets

First up was Ben Zuckerman (University of California, Los Angeles), who updated us on ongoing research into the presence of boron and beryllium in certain white dwarf systems. For some context, there exist many white dwarf systems in which a companion object — sometimes a planet — can have its material accreted onto the surface of the white dwarf due to tidal disruption during close approaches in its orbit. This deposition of material can then be detected via spectral signatures in observations of the white dwarf.

In 2021, evidence was first presented for two white dwarfs being “polluted” with trace amounts of beryllium, which is a particularly rare element, cosmically speaking. This raises the question of how so much beryllium ended up in these systems. In the hunt for answers, however, another mystery has emerged: missing boron. As Zuckerman went on to describe, boron was also recently discovered in some of these white dwarf systems, but at an abundance roughly equal to that of the beryllium. This is despite boron being about a hundred times more common in the universe, further muddying the waters of what exactly is causing such a strange pattern of heavy metals in these systems. Work remains ongoing to find a physical model that can reconcile these observations.

White Dwarf Kicks via Episodic Mass Ejection from Red Giant Stars

Next we heard from Jim Fuller (Caltech). Fuller filled us in on some recent studies aiming to explain why white dwarfs often obtain such high velocities when formed. Just like our own Sun occasionally ejects plumes of plasma from its surface, the red giant stars that eventually form white dwarfs are expected to periodically undergo mass-ejection events. But unlike with our own Sun, these ejections can shear off a significant amount of the star’s total mass — up to ~0.01% per ejection. Each time such an ejection occurs, Newton’s 3rd law ensures that the star itself will also receive a random velocity boost in the opposite direction. Over many thousands of ejections, assuming they occur randomly in all directions, this can significantly “kick” a meandering red giant up to speeds exceeding half a kilometer per second. Once such a star collapses to form a remnant white dwarf, the white dwarf will have this excess velocity too.

Additionally, this random ejection process predicts that red giants in binary systems will experience a sort of “random walk” in their orbital parameters, not just their overall velocity. This can lead to the complete dissociation of binaries, or to the direct collision of a red giant with its companion — leading to predictions about how many widely spaced white dwarf binary systems should exist. | Press release

Probing Intrabinary Material in a New Spider Millisecond Pulsar

Rebecca Kyer (Michigan State University) gave an update on research into so-called “spider” millisecond pulsars. These rapidly rotating neutron stars are a sub-class of pulsar characterized by their very high spin rate and the presence of a companion from which the pulsar is accreting material. Such systems are further sub-classified as either “Black Widows,” which have very low-mass companions (~3% solar mass), “Redbacks,” which have larger companions at ~10-80% solar mass, and “Huntsmans,” which have more massive red-giant companions. These Huntsman pulsars are the newest category, with only two known cases — until now!

Kyer presented evidence collected by their team that may indicate the discovery of a third Huntsman pulsar. The evidence involves radial velocity measurements of the companion star, which shows an orbital period of 3.64 days and variable H-alpha emission. In these systems, the radio emission typically associated with the pulsar can be blocked from our view for unpredictable amounts of time due to clouds of hot gas and dust that emerges from the complex interaction between the pulsar and red giant winds. Despite this, Kyer hopes that with enough targeted observations with the Green Bank Telescope, a radio signal will eventually be detected, confirming the pulsar’s presence.

NASA’s Chandra Finds Unexpected Fireworks in Aftermath of Stellar Explosions

Finally, we heard from a team of three researchers — Andrea Prestwich (Catholic University of America / NASA Goddard SFC), Roy Kilgard (Wesleyan University), and Zoe Hoiland (Vassar College) — on an unexpected supernova discovery. Traditionally, supernovae are modeled as having a relatively constant or slightly decreasing luminosity, at least after their initial explosion. As the stellar material is blasted away from the progenitor’s core, it interacts with the surrounding interstellar medium and a blob of hot, shocked gas is left over. This blob will eventually cool and dim, but often over a long time period, leading to the aforementioned prediction of constant or slightly decreasing brightness.

However, in 14-year Chandra observations of a supernova remnant in Messier 83, Prestwich and her collaborators actually found significant X-ray variability, with the remnant occasionally even increasing in brightness. In a follow-up study, it was found that supernova remnants in Messier 51 likewise show this up-and-down variability. In contrast, an analysis of remnants in Messier 101 showed zero variable remnants. Notably, the former two galaxies are very actively star-forming, while the latter is not. Some possible theoretical models to explain this variability include the possibility of stellar material falling back on the compact object leftover by the supernova, creating an accretion disk that can generate variable emission. Alternatively, it has been proposed that there could be a neighboring star to the remnant object, forming an accreting X-ray binary system that could also induce variability. | Press release

Return to Table of Contents.

---

Plenary Lecture: Sanmi (Oluwasanmi) Koyejo, The Measurement Gap: What AI Can Get Wrong and Why Astronomers Are the Fix (by Lexi Gault)

Sanmi (Oluwasanmi) Koyejo, a professor of computer science at Stanford University, focused this plenary talk on the limitations of AI as it becomes more integrated with science and how astronomers can help solve these problems. Koyejo began by addressing the question of why an AI researcher is talking to astronomers. His research explores how to set AI standards and how to best assess the performance of an AI model, which becomes particularly important when these models are used to do science.

Evaluating the performance of an AI model relies on benchmarks, a fixed set of questions with known answers that are used to score the AI. These benchmarks are simple, reproducible, and comparable, but passing a benchmark does not equate to getting a meaningful scientific result. Benchmarks assume cheap checks, fast feedback, and fixed questions, but science is none of those things; answering astronomy questions relies on telescope time, and ground truth cannot be found quickly.

Much like telescopes, AI benchmarks are measuring instruments with noise that can compound. If you run an AI model many times, the goal is often convergence on an agreed-upon answer, but getting to an agreed-upon answer does not mean that the answer is correct. Koyejo’s recent research found that AI models more often agreed with each other than they did with the actual truth. This is made worse when AI is used to evaluate itself. For example, Koyejo notes the rising number of academic paper submissions due to AI-written papers greatly exceeds the capacity of human reviewers. If AI tools are used to evaluate AI-generated paper submissions, the AI will tend to favor AI-written materials, ignoring the actual scientific merit. Making evaluation of AI an AI problem compounds its failures, straying further from a correct answer.

This is where astronomers come in. Having the expert knowledge of what “correct” looks like, being able to judge which errors matter, access to expensive ground truth (telescope time), and a culture of blind analysis (not knowing exactly what the answer should be), astronomers are well equipped to build standards for AI evaluation. We can define what good AI is for astronomy and use our expertise to help build the standards. Koyejo left the audience with the charge to engage with AI (engagement does not mean endorsement), so that astronomers can help set the standards for appropriate AI use in science.

Read Astrobites’s interview with Sanmi Koyejo.

Return to Table of Contents.

---

Press Conference: Variable, Windy, and Disruptive: The Behavior of Supermassive Black Holes (Briefing video) (by Niloofar Sharei)

This press conference focused on what supermassive black holes do to the world around them, from relativistic jets of the most famous black hole, to the multi-phase winds launched from active galactic nuclei, to temperature-shifting coronae, and even the radio burps that are produced years after a black hole tears apart stars.

Chandra Tracks the Evolving Jet from Messier 87’s Black Hole: 

Camille Poitras (Laval University) presented the largest Chandra X-ray movie of jet evolution in the famous supermassive black hole Messier 87 (M87). M87 has been studied across the electromagnetic spectrum, but X-ray observations of its jet have always suffered from blur because high-energy photons are hard to focus. Using improved Chandra calibrations and deconvolution techniques, the team produced a 13-year X-ray movie of M87’s jet from 2012 to 2025. Tracking a bright knot within the jet, Poitras’s team saw it split into two components in 2023, with one component moving at an apparent five times the speed of light (which is a known illusion called superluminal motion). That allowed the team to look into where and how the particles are being accelerated to extreme energies along the jet. | Press release

Watching a Black Hole Wind Grow: Chandra and Hubble Reveal the Early Stages of Galaxy

Anna Trindade Falcao (NASA Goddard SFC) presented combined Chandra and Hubble observations that may capture a supermassive black hole wind in its earliest stages of growth. Active galactic nucleus (AGN) feedback is thought to happen, but how it begins remains poorly understood. Studying two nearby quasars in which Chandra (tracing hot X-ray wind) and Hubble (tracing warm ionized gas) data are aligned, the team found that the hot and warm phases extend together and not independently. This would suggest a single multi-phase outflow. Their team proposed a four-stage sequence for AGN feedback and also argued that their two systems capture the earliest stages of that process. XRISM Reveals Changing Accretion-Driven Winds in a Nearby Active Galaxy:

Xin Xiang (University of Michigan) presented the most detailed X-ray portrait of the winds flowing out of NGC 4151, an active galaxy roughly 16 million light-years away with a 30-million-solar-mass black hole. Every major X-ray observatory has observed this galaxy before, but earlier instruments could not separate its multiple wind components. With the high-resolution X-Ray Imaging and Spectroscopy Mission (XRISM) spectrometer, Xiang resolved three wind populations by velocity: warm absorbers (100–1,000 km/s), very fast outflows (1,000–10,000 km/s), and ultra-fast outflows (over 10,000 km/s). Across 14 XRISM observations over a year, they found that the fast winds turn on three hours after each bright X-ray flare, matching the predicted timescale for magnetically driven wind. | Press release

NuSTAR Reveals a Significantly Variable Active Galactic Nucleus Corona

Xiurui Zhao (Caltech) presented the first observational evidence of significant temperature variation in the corona of an AGN. The corona produces about 90% of all X-ray photons from an AGN, but its temperature can only be measured with X-ray missions sensitive to high-energy spectral curvature like NASA’s NuSTAR. They compared NuSTAR’s observations across multiple years and found that despite having the same luminosity in 2015 and 2024, the 2024 corona was significantly cooler than the 2015 corona. In follow-up observations in 2025, it was back at its earlier temperature. This is the first clear observation of coronal temperature variability in an AGN. | Press release

New Surprises from Radio Observations of Tidal Disruption Event Outflows

Kate Alexander (University of Arizona) presented new radio observations of tidal disruption events (TDEs). TDEs occur roughly once every 100,000 years in a given galaxy. It happens when a star drifts too close to the black hole, is stretched and squeezed into a long stream of debris (a process astronomers call spaghettification) and some of that debris falls onto the black hole while the rest is thrown outward and lights up in radio. Alexander and her team used six years of Very Large Array monitoring of several dozen nearby TDEs. They found that many TDEs unexpectedly turn on in the radio years after the original event, even long after their optical light has faded. Their team combined radio data with X-ray and optical accretion-rate measurements and found that these late-time radio flares appear when the black hole is eating either very quickly or very slowly, but not at moderate rates! This pattern matched what is seen in stellar-mass black holes and this suggests that accretion physics may operate similarly across many orders of magnitude in black hole mass. | Press release

Return to Table of Contents.

---

Plenary Lecture: Mario Jurić, Early Results from the Rubin Observatory: Mapping the Solar System and Beyond (by Kerry Hensley)

Mario Jurić (University of Washington; PI of the University of Washington Rubin operations element) began by remarking on how our collective concept of what an astronomer is has been transformed over the past 25 years. While a Google image search for “astronomer” yields an array of people pointing small telescopes at the night sky, today’s version of professional astronomy looks quite different: huge telescopes with enormous cameras covering vast swaths of the sky, with analysis aided by software and data deposited in massive databases.

This rapid transition was first spurred by enterprises like the Sloan Digital Sky Survey, which collected 20 terabytes of raw data from 1.2 billion observations. Now, digital sky surveys are common, and they’re becoming ever larger. The next massive digital sky survey will be the Legacy Survey of Space and Time from the NSF–DOE Vera C. Rubin Observatory — an observatory that can “download the universe” and is poised to transform our understanding of everything from asteroids to dark energy.

Rubin released its first images on 23 June 2025, and among these stunning images is a broad view that contains thousands of galaxies in the Virgo cluster, a golden tangle of four colliding galaxies, a nearby galaxy cluster, a distant galaxy cluster, and a handful of nearby spiral galaxies bursting with star formation. As fantastic as this image is, Jurić emphasized that it’s just one still image in a movie: as Rubin returns to the same fields of view time and time again, it’ll spot things that change brightness (like variable stars and transients) and things that move (like asteroids).

Jurić highlighted four areas in which Rubin has already contributed greatly, focusing on discoveries within our solar system:

1. 1. Asteroid variability: Rubin has already collected hundreds of thousands of observations of asteroids, resulting in the discovery of extremely fast rotators like 2025 MN45, which must have an extremely high tensile strength not to fly apart.
   2. Observations of interstellar object 3I/ATLAS: During the observatory’s commissioning survey, the Rubin team realized that they had observed 3I/ATLAS before it was discovered — but in observations that were technically under embargo. Recognizing the high impact of the data, the team found a way to share their findings, leading the 150-person collaboration to write a paper on their findings in a mere three days.
   3. High-precision asteroid photometry: While analyzing the commissioning data, the Rubin team discovered systematic offsets in the positions of asteroids — but no other types of objects. They found that this effect varied across the sky and was worse in the south. They eventually discovered that because most asteroids are observed from the north, that catalogs of asteroid orbits are systematically biased at a level of 20-40 milliarcseconds. Luckily, Rubin will provide the much-needed southerly data necessary to clean up these catalogs within about a year.
   4. Outer solar system studies: Researchers discovered 380 new trans-Neptunian objects in Rubin’s commissioning data. (Compare that to just 5,000 known trans-Neptunian objects in total!) In the time since the commissioning data were collected, hundreds more candidate objects were discovered.

With such enormous data yields from Rubin, how do we efficiently turn data into science? Jurić outlined strategies for handling massive datasets using well-crafted data formats, computing tools, and AI. In particular, Jurić highlighted the Astronomical Catalog Inference Driver, or ACID, which he built in four weeks on his phone using Claude. The tool is comparable to “a product developed in >3 years by extremely capable human professionals” and can rapidly handle immense data tables, showing how AI can accelerate the development of tools to handle tomorrow’s datasets.

Read Astrobites’s interview with Mario Jurić.

Return to Table of Contents.

---

Plenary Lecture: Esra Bulbul, Cosmology and Baryons Across Scales: Probing Halo Mass Function to Cosmic Filaments with eROSITA (by Lexi Gault)

Esra Bulbul (Max Planck Institute for Extraterrestrial Physics) began her plenary talk at the beginning of time — the Big Bang marked the start of the universe, and from there we attempt to constrain cosmological parameters in order to understand how the universe was born and how it has evolved. Bulbul notes that a lot of work has been done to constrain cosmology from the cosmic microwave background (CMB), and we must compare those early time measurements to those from later times to build a full picture. In cosmology, tensions have arisen between early-time and late-time probes. CMB measurements of the Hubble constant and S8, the “clumpiness” parameter of the universe, do not agree with late time probes that find that the universe is accelerating faster with smaller fluctuations than the CMB predictions. How do we work toward resolving this tension?

The extended ROentgen Survey with an Imaging Telescope Array (eROSITA) is a space telescope that has performed the first imaging all-sky survey in the medium X-ray range, seeking to map the large-scale structures of the universe to better constrain cosmological constants. eROSITA focuses on finding galaxy groups and clusters via the hot X-ray gas within and between them across the last 5-7 billion years of the universe. Since its launch in 2019, eROSITA has completed more than four full-sky surveys. Of the million X-ray sources detected by eROSITA, only 12,247 (1%) are galaxy clusters and groups.

From the observed X-ray emission from the galaxy clusters and groups, the eROSITA team has been able to measure the temperature of the intercluster medium within the sample. Working with weak gravitational lensing surveys, they have been able to measure the masses of the newly discovered clusters and groups. These measurements show that eROSITA has detected sources across three orders of magnitude in mass, reaching from the galaxy level to large-scale clusters.

The galaxy cluster and group observations allow them to constrain a few cosmological parameters including S8 (density fluctuations), Ωm (matter density), and w0 (dark energy). From the impressive eROSITA data, Bulbul showed that they could constrain S8 and Ωm to five times better precision than ever before! These late-time results are fully consistent with what has been measured from the early-time CMB measurements. Additionally, the dark energy measurements agree with what has been measured thus far from the Dark Energy Spectroscopic Instrument survey.

Bulbul then switched us from the exciting cosmological results to those regarding baryon detection. A striking result from eROSITA was the detection of hot gas well beyond what is typically considered the gravitational extent of a galaxy cluster — reaching out to 16 million light-years, often larger than the galaxy cluster itself. This underscores the importance of X-ray surveys in detecting the very extended gas encapsulating the largest-scale structures of the universe.

Looking forward, Bulbul discussed the future work planned for eROSITA. Transitioning now to working with the deepest data of the survey, the team expects to find about 40,000 galaxy clusters with about half of those being used for cosmology. They anticipate obtaining even higher precision on cosmological parameter measurements, comparable to that obtained from CMB observations. The work of eROSITA is revolutionizing cosmology at later times in the universe, and the future is exciting.

Read Astrobites’s interview with Esra Bulbul.

Return to Table of Contents.

The post AAS 248: Day 1 appeared first on AAS Nova.

The Transient Artifact and Continuous Learning System (TACLS) leverages data from continuously operating satellite networks coupled with machine learning models to help meteorologists at the National Weather Service forecast flash floods more efficiently. This new software is the result of a collaboration between NASA’s Jet Propulsion Laboratory, the University of California, San Diego (UCSD), and the National Oceanic and Atmospheric Administration (NOAA) National Weather Service (NWS).

Created with support from NASA’s Earth Science Technology Office (ESTO), TACLS leverages machine learning to automatically locate evidence (unusual increases in atmospheric moisture) of impending flash flooding that meteorologists may otherwise miss as they analyze large amounts of data. TACLS flags that evidence, indicates where flash flooding could likely occur, and displays that information via a user-friendly visualization for human analysts to interpret. Those analysts can then decide whether to issue a flash flood warning or weather advisory.

This novel framework for tracking extreme weather events and predicting imminent flash floods operates in near real-time, producing forecasts in as little as fifteen minutes.

“That’s really what we wanted to do, to give meteorologists a tool to help decision making for flash flood warnings,” said Yehuda Bock, Distinguished Researcher at the UCSD Scripps Institution of Oceanography and principal investigator for TACLS.

In simulations testing, TACLS used data from diverse severe weather events—including atmospheric rivers, monsoonal convection, and tropical cyclone remnants—between 2017 and 2023 and successfully captured 93% of the issued flash-flood warnings. Meteorologists from the National Weather Service are currently working to incorporate TACLS into their existing systems for forecasting flash floods in Southern California.

This learning system has two main components. First, an analytic back-end software suite uses machine learning algorithms to process satellite data and determine areas at risk for flooding. Second, user-friendly visualization software highlights those areas for further analysis by humans.

The ACLS back-end software analyzes data from satellites in the Global Navigation Satellite System (GNSS), a constellation of satellite networks that drive navigation services around the world. Water vapor in the troposphere delays signals from these satellites as they travel to Earth. This signal delay can be analyzed to calculate the amount of water vapor in the atmosphere over a particular location on Earth.

The TACLS analytic back-end software suite features a machine learning model trained using more than 30 years of past GNSS data. This model is an anomaly detector that tracks unusual increases in atmospheric moisture. The model then carefully examines that atmospheric moisture data and determines whether it’s either an artifact (a false feature or distortion in the data) or a transient (a time-sensitive physical event, like heavy precipitation) that requires interpretation by human analysts.

If TACLS determines the data represents a transient, such as an extreme weather event that warrants a flash flood warning, it will forward that data to the TACLS visualization software (MGViz) for further evaluation by humans. The analysts use their judgement and experience to interpret these events and determine whether the flagged data indicates a flash flood is likely, and, if necessary, issue a flash flood warning.

Several past innovations developed at JPL are leveraged by TACLS to process GNSS data and present the results. The analytic back-end software suite incorporates elements from JPL’s Domain-agnostic Outlier Ranking Algorithms program and the Time-series Forecasting, Evaluation, and Deployment program. The TACLS visualizer is based on the Multi-Mission Geographic Information System, originally developed at JPL for NASA’s Mars missions.

The TACLS software binds all these components within a novel system that enhances existing methods to reduce the amount of time it takes for a human analyst to determine whether to issue a flash flood warning.

Both the TACLS software and the data used to train it will be open-source, allowing scientists to either tailor this model in response to their unique research needs or create their own model from scratch.

For additional details, see the entry for this project on NASA TechPort. 

Project Lead: Dr. Yehuda Bock, University of California, San Diego. 

Sponsoring Organization(s): NASA’s Earth Science Technology Office Advanced Information Systems Technology Program; JPL; NOAA; National Weather Service. 

The Transient Artifact and Continuous Learning System (TACLS) leverages data from continuously operating satellite networks coupled with machine learning models to help meteorologists at the National Weather Service forecast flash floods more efficiently. This new software is the result of a collaboration between NASA’s Jet Propulsion Laboratory, the University of California, San Diego (UCSD), and the National Oceanic and Atmospheric Administration (NOAA) National Weather Service (NWS).

Created with support from NASA’s Earth Science Technology Office (ESTO), TACLS leverages machine learning to automatically locate evidence (unusual increases in atmospheric moisture) of impending flash flooding that meteorologists may otherwise miss as they analyze large amounts of data. TACLS flags that evidence, indicates where flash flooding could likely occur, and displays that information via a user-friendly visualization for human analysts to interpret. Those analysts can then decide whether to issue a flash flood warning or weather advisory.

This novel framework for tracking extreme weather events and predicting imminent flash floods operates in near real-time, producing forecasts in as little as fifteen minutes.

“That’s really what we wanted to do, to give meteorologists a tool to help decision making for flash flood warnings,” said Yehuda Bock, Distinguished Researcher at the UCSD Scripps Institution of Oceanography and principal investigator for TACLS.

In simulations testing, TACLS used data from diverse severe weather events—including atmospheric rivers, monsoonal convection, and tropical cyclone remnants—between 2017 and 2023 and successfully captured 93% of the issued flash-flood warnings. Meteorologists from the National Weather Service are currently working to incorporate TACLS into their existing systems for forecasting flash floods in Southern California.

This learning system has two main components. First, an analytic back-end software suite uses machine learning algorithms to process satellite data and determine areas at risk for flooding. Second, user-friendly visualization software highlights those areas for further analysis by humans.

The ACLS back-end software analyzes data from satellites in the Global Navigation Satellite System (GNSS), a constellation of satellite networks that drive navigation services around the world. Water vapor in the troposphere delays signals from these satellites as they travel to Earth. This signal delay can be analyzed to calculate the amount of water vapor in the atmosphere over a particular location on Earth.

The TACLS analytic back-end software suite features a machine learning model trained using more than 30 years of past GNSS data. This model is an anomaly detector that tracks unusual increases in atmospheric moisture. The model then carefully examines that atmospheric moisture data and determines whether it’s either an artifact (a false feature or distortion in the data) or a transient (a time-sensitive physical event, like heavy precipitation) that requires interpretation by human analysts.

If TACLS determines the data represents a transient, such as an extreme weather event that warrants a flash flood warning, it will forward that data to the TACLS visualization software (MGViz) for further evaluation by humans. The analysts use their judgement and experience to interpret these events and determine whether the flagged data indicates a flash flood is likely, and, if necessary, issue a flash flood warning.

Several past innovations developed at JPL are leveraged by TACLS to process GNSS data and present the results. The analytic back-end software suite incorporates elements from JPL’s Domain-agnostic Outlier Ranking Algorithms program and the Time-series Forecasting, Evaluation, and Deployment program. The TACLS visualizer is based on the Multi-Mission Geographic Information System, originally developed at JPL for NASA’s Mars missions.

The TACLS software binds all these components within a novel system that enhances existing methods to reduce the amount of time it takes for a human analyst to determine whether to issue a flash flood warning.

Both the TACLS software and the data used to train it will be open-source, allowing scientists to either tailor this model in response to their unique research needs or create their own model from scratch.

For additional details, see the entry for this project on NASA TechPort. 

Project Lead: Dr. Yehuda Bock, University of California, San Diego. 

Sponsoring Organization(s): NASA’s Earth Science Technology Office Advanced Information Systems Technology Program; JPL; NOAA; National Weather Service. 

The post NASA Uses Machine Learning to Enhance Flash Flood Warnings appeared first on NASA Science.

This week, AAS Nova and Astrobites are attending the American Astronomical Society (AAS) summer meeting in Pasadena, CA.

AAS Nova Editors Kerry Hensley and Susanna Kohler and AAS Media Fellow Lexi Gault will join Astrobites Media Intern Niloofar Sharei and Astrobiter Lucas Brown to live-blog the meeting for all those who aren’t attending or can’t make it to all the sessions they’d like. We plan to cover all of the plenaries and press conferences, so follow along here on aasnova.org or on astrobites.org! You can also follow Astrobites on Bluesky at astrobites.bsky.social for more meeting content.

Where can you find us during the meeting? We’ll be at the Astrobites booth in the Exhibit Hall all week — stop by and say hello! You can also find Susanna, Kerry, and Lexi at the press conferences in Ballroom H Monday and Tuesday at 10:15 am and 2:15 pm PDT and Wednesday at 10:15 am PDT. AAS press conferences are open to all attendees, and they can also be found on the AAS Press Office YouTube channel for anyone not attending the meeting.

In addition to the usual press conferences, the AAS Press team has collaborated with Los Angeles’s own Griffith Observatory to arrange a special live recording of the Observatory’s acclaimed astronomy program, All Space Considered. This event is open to all AAS 248 attendees and will take place Wednesday, 17 June, at 2:15 pm PDT in Ballroom H. We hope to see you there!

Finally, you can read the currently published AAS 248 keynote speaker interviews, which were conducted by Astrobiters Niloofar Sharei, Amaya Sinha, Sowkhya Shanbhog, Katya Gozman, and Natalie Price. Be sure to check back all week as the remainder are released!

---

Education, Outreach, and More at AAS 248

AAS 248 features a broad range of sessions related to education and outreach. For a comprehensive and curated list of sessions, check out this post from AAS Director of Education Programs Tom Rice. Here are a few highlights (all times PDT; must be logged in to aas.org for links to work correctly); note that times and locations are subject to change:

On Monday, kick off the meeting with the Astronomy Education, Outreach, and Night-Sky Stewardship iPoster session, 9:00–10:00 am in Exhibit Hall A. Wrap up your day by attending the Reception for Astronomy Educators from 6:30 to 8:00 pm in Room 107.

Tuesday morning, hear about pressing issues relating to climate change at the Special Session What Astronomers Can Do About Climate Change: Infrastructure, Education, and Communication, from 10:00 to 11:30 am in Ballroom G. That afternoon from 2:00 to 3:30 pm, tune in to the Pathways into Modern Space Science oral session in Ballroom C.

On Wednesday, learn how to Make Your Voice Heard as an Advocate for Science in a Special Session co-sponsored by the AAS and the American Institute of Physics from 10:00 to 11:30 am in Ballroom G. Throughout the week, check in with the League of Women Voters in the Exhibit Hall — they’ll be conducting a voter registration drive during the meeting.

Thursday, the final day of the meeting, closes with a look at a local institution in the 40 Years of Community Engagement at IPAC: Lessons Learned Special Session from 2:00 to 3:30 pm in Ballroom F.

If you’re sticking around Pasadena after the meeting, consider heading to Griffith Observatory for their open house on Saturday, 20 June, 2:00–10:00 pm, which will showcase accessible observing techniques like tactile images, 3D models, live camera feeds, and sonification.

And remember, you can find a longer curated list of education-related sessions here!

The post AAS 248: Welcome! appeared first on AAS Nova.

The 248th meeting of the American Astronomical Society, to be held in Pasadena, CA, 14–18 June, is nearly here. The AAS Publishing team looks forward to connecting with meeting attendees, and you can find representatives from the publishing and journals’ editorial staff at the AAS booth in the Exhibit Hall in the Pasadena Convention Center. AAS Director of Scholarly Publishing Kerry Kroffe will be available at the AAS Publishing booth, so be sure to stop by to say hello, chat about the journals, and pick up some swag!

AAS Nova Editors Kerry Hensley and Susanna Kohler, AAS Media Fellow Lexi Gault, Astrobites Media Intern Niloofar Sharei, and the rest of the Astrobites team will also be available periodically at the Astrobites booth in the Exhibit Hall. We look forward to seeing you there!

---

AAS Peer Review Workshop

This workshop led by the scientific editors of the AAS journals will teach participants about the peer review process, give them the opportunity to see both poor and exemplary referee reports, and provide them with hands-on experience in writing a peer review report. Participants will receive a graduation certificate. While signups for the AAS 248 peer review workshop are closed, more workshops are planned for future AAS meetings, so keep an eye out for more information about upcoming opportunities!

---

Open Science, Data, and AI in Astronomy: Sessions to Look for at AAS 248

Note: The links in this section take you to the corresponding entries in the AAS 248 block schedule. You must be logged in for the links to work correctly; otherwise, they will take you to the main block schedule page. Note that times and locations are subject to change.

The upcoming AAS meeting has a number of sessions, workshops, and lectures for those interested in open-source tools, data management strategies, and AI’s role in astronomy. Sunday, 14 June, will feature a trio of data-related workshops. (Note that these workshops require an additional fee and can be added to your registration.) The “Euclid Data in the Cloud: Access, Analysis, and Science Opportunities” workshop (10:00 am – 11:30 pm PDT in Room 101) will introduce participants to accessing and analyzing data from Euclid, a European Space Agency mission with NASA participation. The overlapping workshop “High Energy Science Analysis with HEASARC Services” (10:00 am – 12:00 pm PDT in Room 102), organized by the High Energy Astrophysics Science Archive Research Center (HEASARC), is targeted toward early-career scientists and anyone curious about HEASARC services. The workshop will begin with an overview of available services before transitioning into hands-on experimentation with these tools. Finally, “An Introduction to Fornax: Scalable Data and Compute for Scientific Analysis” (2:00–4:00 pm PDT in Room 101) will introduce attendees to NASA’s new Astrophysics Science Platform through live demonstrations and guided exercises.

Of particular interest regarding AI in astronomy is Monday morning’s plenary lecture from Sanmi (Oluwasanmi) Koyejo titled “The Measurement Gap: What AI Can Get Wrong and Why Astronomers Are the Fix.” This session will take place from 11:40 am to 12:30 pm PDT in Ballroom DE.

The “AI-Driven Science in the Survey Era” Special Session will take place on Tuesday, 16 June, from 2:00 pm to 3:30 pm in Ballroom F. This session will highlight recent breakthroughs in a broad range of astronomical sub-fields that have greatly benefited from the use of AI and machine-learning techniques. Also on Tuesday, check out recent work on imaging and data-handling methods at the “Observational Imaging Methods and Data Pipelines” iPoster session from 5:00 pm to 6:00 pm in the Exhibit Hall.

The post AAS Publishing at AAS 248 in Pasadena appeared first on AAS Nova.

Zooming through the galaxy faster than they should, hypervelocity stars are a curious population of stars whose origins are often difficult to determine. A recent study presents the first evidence of an old, low-mass hypervelocity star launched from the Milky Way’s center, providing clues into this elusive set of fast-moving stars.

Flinging Fast Stars

Astronomers have caught a number of hypervelocity stars zipping through the Milky Way since their first discovery in 2005. With speeds fast enough to escape the galaxy, hypervelocity stars were launched off course by some strong dynamical interaction. One such method, known as the Hills mechanism, predicts that when a binary star system approaches the Milky Way’s central supermassive black hole, one binary member is captured by the black hole and the other is flung away with high velocity.

However, constraining a hypervelocity star’s origin to the galactic center is limited by uncertainties in distances and proper motions, muddling how well researchers have been able to trace their orbits backward. Among a number of candidate galactic center–origin hypervelocity stars, only one has been confidently identified as having been kicked from the galaxy’s center. Intriguingly, this hypervelocity star and most galactic center–origin candidates are young, massive stars — old, low-mass hypervelocity stars from the galactic center have yet to be identified.

What does this dearth of old, low-mass zooming stars mean? Either stars in the galactic center are preferentially young and massive, with fewer than expected old and small stars, or low-mass hypervelocity stars are missed by selection effects, challenging to detect in current surveys. Searching for hypervelocity stars with higher precision and sensitivity may allow astronomers to tease out this missing population and trace back more zooming stars to a Hills mechanism origin.

Discovering An Old Hypervelocity Star

Leveraging high-quality spectra from the Dark Energy Spectroscopic Instrument (DESI) survey and precise astrometry from Gaia, Shunhong Deng (University of Chinese Academy of Sciences; China West Normal University) and collaborators reported the discovery of the hypervelocity star DESI-HVS1. From the spectroscopic and astrometric analyses, the authors determined that DESI-HSV1 is a ~14-billion-year-old low-metallicity star with a mass about 80% that of the Sun’s, moving through the galaxy at 523 km/s.

Exploring the possible origins of this hypervelocity star, the authors integrated the star’s orbit backward in time for one billion years. The backward orbit integrations show the star approaching the galactic center where it then turns around, traveling away faster than the local escape velocity of the galaxy. Even when varying the galactic potential and constants, all orbit integrations suggest DESI-HVS1’s motion through the galaxy is most naturally explained by the Hills mechanism.

While more refined galaxy models and comparison to other candidate galactic center–origin hypervelocity stars are necessary to fully confirm its origins, DESI-HVS1 provides the first compelling evidence for an old, low-mass hypervelocity star candidate consistent with a galactic center launch. This suggests that the observed lack of such stars is likely not a reflection of the Milky Way’s central stellar population, but rather a consequence of observational constraints making such small hypervelocity stars particularly difficult to detect. As large spectroscopic and astrometric surveys continue to advance, the sample of galactic-center hypervelocity star candidates will grow, enabling further exploration of the stellar environment and dynamical processes at the Milky Way’s core.

Citation

“An Old, Low-mass, Metal-poor Hypervelocity Star Candidate Consistent with Galactic Center Origin,” Shunhong Deng et al 2026 ApJL 1003 L9. doi:10.3847/2041-8213/ae6505

The post Shot from the Heart of the Milky Way: Discovery of an Old Hypervelocity Star appeared first on AAS Nova.

A compact, multifrequency radar built by a team at NASA’s Jet Propulsion Laboratory will make it easier to collect information about dynamic cloud systems. Called CloudCube, this new instrument simultaneously probes the atmosphere with three radar signals, spanning 36 to 240 GHz, for optimized sensitivity to a wide range of water droplet and ice particle sizes. 

Built with funding from NASA’s Earth Science Technology Office Instrument Incubator Program, CloudCube transmits and receives Ka-, W-, and G-band signals, making it the first compact radar system capable of simultaneously probing meteorological targets at wavelengths spanning approximately one to ten millimeters. Researchers will be able to combine information from the three signals to learn more about the initiation and evolution of precipitation, as well as cloud microphysics and radiative properties.

“We’re making a low-power, low-mass instrument to facilitate new cost-efficient missions for atmospheric observations. Building a multi-frequency radar, especially at G-band, is very novel,” said Raquel Rodriguez Monje, a systems engineer at JPL and principal investigator for CloudCube.

Each of CloudCube’s three signals observes a different element of cloud physics. Ka-band radar signals are ideal for collecting precipitation profiles; W-band radar signals are preferred for measuring cloud particles that give rise to precipitation; and G-band radar signals, which have never been collected from a space-based instrument, are ideal for measuring ice and liquid water content inside very light clouds (a paper describing this measurement can be found here).

Probing the atmosphere simultaneously with three signals allows researchers to collect data on all these cloud features at once, which is valuable for improving weather forecasts and especially climate modeling. CloudCube leverages innovations in millimeter-wave hardware to pack three radar modules–one for each signal–within a single compact system.

One CloudCube innovation concerns the specialized components required to transmit G-band power from a compact, low-power instrument. The detection of cloud signals requires high transmit power, which CloudCube achieves by combining the outputs of multiple high-efficiency frequency-multiplication devices that allow the instrument to generate hundreds of milliWatts at 240 GHz. Another innovation of CloudCube is that it was designed to use as few radio frequency components as possible to reduce its mass and power consumption, which could lower the cost of future Earth-observing orbital instruments.

Flying an instrument equipped with G-band radar in space will be a new capability and will allow researchers to achieve greater spatial resolution and sensitivity in the study of cloud microphysical processes.

“Basically, we’re weighing clouds using these combinations of frequencies in a way that we couldn’t do before we had the G-band,” said Matt Lebsock, a researcher at JPL and co-investigator for CloudCube.

The instrument has been tested in the field. A ground-based prototype of CloudCube’s G-band channel operated continuously for 11 months during the Department of Energy’s Cloud and Precipitation Experiment at Kennaook (CAPE-K) campaign. CloudCube also participated in the Eastern Pacific Cloud Aerosol Precipitation Experiment, a ground campaign sponsored by the Department of Energy. A paper describing the results of that experiment can be found here.

Most recently, CloudCube successfully operated all three frequency bands from NASA’s Gulfstream III aircraft and collected its first airborne observations of snowfall as part of the North American Upstream Feature-Resolving and Tropopause Uncertainty Reconnaissance Experiment campaign—a NASA-funded campaign designed to improve forecasts of high-impact winter weather. The CloudCube team is currently calibrating and processing the data for public release.

For additional details, see the entry for this project on NASA TechPort.

Project Lead: Dr. Raquel Rodriguez Monje, NASA’s Jet Propulsion Laboratory

Sponsoring Organization: NASA’s Earth Science Technology Office Instrument Incubation Program

A compact, multifrequency radar built by a team at NASA’s Jet Propulsion Laboratory will make it easier to collect information about dynamic cloud systems. Called CloudCube, this new instrument simultaneously probes the atmosphere with three radar signals, spanning 36 to 240 GHz, for optimized sensitivity to a wide range of water droplet and ice particle sizes. 

Built with funding from NASA’s Earth Science Technology Office Instrument Incubator Program, CloudCube transmits and receives Ka-, W-, and G-band signals, making it the first compact radar system capable of simultaneously probing meteorological targets at wavelengths spanning approximately one to ten millimeters. Researchers will be able to combine information from the three signals to learn more about the initiation and evolution of precipitation, as well as cloud microphysics and radiative properties.

“We’re making a low-power, low-mass instrument to facilitate new cost-efficient missions for atmospheric observations. Building a multi-frequency radar, especially at G-band, is very novel,” said Raquel Rodriguez Monje, a systems engineer at JPL and principal investigator for CloudCube.

Each of CloudCube’s three signals observes a different element of cloud physics. Ka-band radar signals are ideal for collecting precipitation profiles; W-band radar signals are preferred for measuring cloud particles that give rise to precipitation; and G-band radar signals, which have never been collected from a space-based instrument, are ideal for measuring ice and liquid water content inside very light clouds (a paper describing this measurement can be found here).

Probing the atmosphere simultaneously with three signals allows researchers to collect data on all these cloud features at once, which is valuable for improving weather forecasts and especially climate modeling. CloudCube leverages innovations in millimeter-wave hardware to pack three radar modules–one for each signal–within a single compact system.

One CloudCube innovation concerns the specialized components required to transmit G-band power from a compact, low-power instrument. The detection of cloud signals requires high transmit power, which CloudCube achieves by combining the outputs of multiple high-efficiency frequency-multiplication devices that allow the instrument to generate hundreds of milliWatts at 240 GHz. Another innovation of CloudCube is that it was designed to use as few radio frequency components as possible to reduce its mass and power consumption, which could lower the cost of future Earth-observing orbital instruments.

Flying an instrument equipped with G-band radar in space will be a new capability and will allow researchers to achieve greater spatial resolution and sensitivity in the study of cloud microphysical processes.

“Basically, we’re weighing clouds using these combinations of frequencies in a way that we couldn’t do before we had the G-band,” said Matt Lebsock, a researcher at JPL and co-investigator for CloudCube.

The instrument has been tested in the field. A ground-based prototype of CloudCube’s G-band channel operated continuously for 11 months during the Department of Energy’s Cloud and Precipitation Experiment at Kennaook (CAPE-K) campaign. CloudCube also participated in the Eastern Pacific Cloud Aerosol Precipitation Experiment, a ground campaign sponsored by the Department of Energy. A paper describing the results of that experiment can be found here.

Most recently, CloudCube successfully operated all three frequency bands from NASA’s Gulfstream III aircraft and collected its first airborne observations of snowfall as part of the North American Upstream Feature-Resolving and Tropopause Uncertainty Reconnaissance Experiment campaign—a NASA-funded campaign designed to improve forecasts of high-impact winter weather. The CloudCube team is currently calibrating and processing the data for public release.

For additional details, see the entry for this project on NASA TechPort.

Project Lead: Dr. Raquel Rodriguez Monje, NASA’s Jet Propulsion Laboratory

Sponsoring Organization: NASA’s Earth Science Technology Office Instrument Incubation Program

The post NASA’s CloudCube Pioneers Miniaturized Radar to Study Clouds, Precipitation appeared first on NASA Science.

What’s the story behind the longest gamma-ray burst ever observed? Researchers explore how repeated interactions between a white dwarf and an intermediate-mass black hole could explain GRB 250702B’s unusual properties.

An Exceptional Gamma-Ray Burst

In 2025, the Fermi Gamma-ray Space Telescope spotted an intense flash of gamma rays from a distant, dusty galaxy, and a fleet of spacefaring and earthbound telescopes soon revealed the remarkable characteristics of the event. GRB 250702B featured multiple bursts spanning several hours, individual flares lasting roughly 100 seconds, and X-ray emission in the 24 hours leading up to the gamma-ray activity — something never seen before in a gamma-ray burst.

These characteristics have been challenging to explain with typical models of long gamma-ray burst formation, such as collapsing massive stars, but researchers have developed numerous promising alternatives. In a new research article, astronomers have shown how GRB 250702B’s behavior might be explained by repeated close encounters between a white dwarf and an intermediate-mass black hole.

Eccentric Close Encounters

Yuri Sato (Tohoku University) and collaborators modeled the source of GRB 250702B as a white dwarf that has been captured into orbit around an intermediate-mass black hole. In their model, the white dwarf travels on an extremely eccentric orbit, drawing close to the black hole roughly once an hour. Each time the white dwarf ventures close, the black hole’s tidal forces strip away part of the star until, after about 40 such encounters, the white dwarf is completely destroyed.

When the black hole peels material away from the white dwarf, some material swirls into an accretion disk, and some is cast off into space. Accretion from the disk powers a narrow relativistic jet: the source of the event’s powerful gamma-ray flares. This should mean that there is one jet produced each time the white dwarf is partially disrupted — 40 times — yet astronomers detected only a handful of flares from GRB 250702B.

Jet Precession and Model Predictions

The discrepancy can be resolved if the angle between the jet and the black hole’s spin axis precesses over time so that only some of the jets are pointed toward us. This can occur if the accretion disk is misaligned relative to the black hole’s spin. As an alternative, disk instabilities or other processes could keep all but a few of the expected jets from being launched.

Sato and coauthors found that these two scenarios have different long-term outcomes, providing a potential way to ascertain what’s going on at the source; in the jet precession scenario, the jets that are launched away from our line of sight produce an order of magnitude more radio emission at late times.

This work by Sato and collaborators demonstrates how the repeated partial tidal disruption of a white dwarf by an intermediate-mass black hole can match the properties of one of the most remarkable recent transient sources — one that has pushed astronomers to consider new ways of producing gamma-ray bursts.

Citation

“Successive Partial Disruptions with Orbital Precession in a White Dwarf–Black Hole System for Repeating GRB 250702B,” Yuri Sato et al 2026 ApJL 1003 L44. doi:10.3847/2041-8213/ae6a8f

The post Could a White Dwarf–Black Hole Pair Explain the Longest Gamma-Ray Burst Ever Seen? appeared first on AAS Nova.

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: Dynamics of Planetary Rings Under Thermal Forces
Authors: Wen-Han Zhou et al.
First Author’s Institution: The University of Tokyo
Status: Published in ApJL

If you ask anyone what their favourite planet is, the answer you’ll most likely hear is Saturn. Why? Why else than the beautiful and intricate ring system surrounding the gas giant. The other gas and ice giants in our solar system — Jupiter, Uranus, and Neptune — have ring systems themselves but none quite as striking as Saturn’s. Would it surprise you to learn that astronomers’ best models have not yet totally explained why Saturn’s rings look the way they do?

In contrast, it may not surprise you to hear that people have been trying to explain the rings for as long as we have seen them with telescopes. We now know that planetary rings are collections of relatively small particles (think micrometre up to metre sized), most likely having once been the material of a larger body that was disrupted either by collisions or tidal forces. This material, due to the gravity of its local planet, is sculpted into a flat disc where more subtle interactions then lead to substructure forming within the disc such as gaps and ringlets. Many of these substructures are explained by well-understood physics — for example gaps being carved by embedded moonlets or resonances sculpting ring edges — though there remain some outstanding problems in our understanding.

The authors of today’s article set their sights on the problematic inner edge of Saturn’s A ring (Figure 1). They describe some mechanisms — namely the collisions of micrometeoroids within the rings — by which a sharp ring edge can be maintained, but there exists a gap in the understanding of how such an edge can form in the first place. All hope is not lost, though, as today’s authors reintroduce a physical process they call the “eclipse–Yarkovsky” effect, which seems to explain these phenomena.

Billions of Rocket-Powered Bumper Cars

The idea behind this process revolves entirely around light. When sunlight hits a particle within a planetary ring, non-absorbed sunlight imparts a little “bump” onto a particle, ever so slightly altering its trajectory. At the same time, some of the sunlight is absorbed into the particle, which heats it up; light is soon re-emitted via thermal radiation, which, again, can slightly change the trajectory of these tiny particles. From a point on the surface of each ring particle (see rightmost side of Figure 2), the photons from this thermal radiation are all emitted in random directions; however, across the whole surface of the particle there is a net force! These tiny particles are spinning, and so the side just near “sunset” is hottest and emitting the most thermal photons. In this way, the pressure from the sunlight plus the thermal radiation of each particle imparts a net torque on the ring itself which changes the angular momentum of the ring.

What we described just then is essentially solar radiation pressure plus the Yarkovsky effect. It turns out that the net force from this process typically averages out to zero as the ring particles orbit around the planet. This assumes, though, that the sunlight is constantly shining on the ring. Stunning imagery of Saturn tells us that this isn’t the case, so what happens when we take into account the shadow cast on the ring from the planet? Today’s authors find that the net effect induces a positive change in angular momentum of the ring particles, which they call the eclipse–Yarkovsky effect.

After today’s authors detailed all of the math involved in this process (and there is a lot), they put it into practice to try to help explain Saturn’s curious rings. Including the eclipse–Yarkovsky effect together with other known effects that drive ring evolution allowed them to reproduce the optical depth profile (how thick the ring looks) of Saturn’s A ring better than ever before (notably the sharp inner edge in Figure 3). On top of this, the effect provides another avenue for moonlet formation in the outer edges of ring systems as the positive torque from the effect drives material out toward and away from the Roche limit.

Undeniably useful for Saturn, the eclipse–Yarkovsky effect may also explain some other conundra in the solar system. Mars may have once had rings, which are thought to have eventually clumped together to form the inner moon Phobos. This idea is problematic, though, in that current models suggest there should still be some residual ring system around Mars even after Phobos formed. Enter the eclipse–Yarkovsky effect: being 100 times stronger for Mars than Saturn (due to its more intense incident sunlight), the effect may have driven out and dispersed that residual ring altogether. While the authors are currently looking into this possibility, the reintroduction of the eclipse–Yarkovsky effect has already shown great promise for our most beautiful ringed planetary neighbour.

Original astrobite edited by Wasi Naqvi.

About the author, Ryan White:

I am a first-year PhD student at Macquarie University in Australia, working mainly on binary/multiple systems with massive stars (Wolf–Rayets in particular!). Outside of study, I’m probably drinking coffee, baking, reading, or going for a run. You can also find me procrastinating on Bluesky @astroryan.bsky.social.

The post Shining a Light on the Shadowy Movements in Saturn’s Rings appeared first on AAS Nova.

When the third known object from the galactic space beyond our Sun barreled through the solar system last year, Earth was in a terrible spot to view its flyby. Thankfully, however, a spacecraft that was previously busy mapping Mars took advantage of much better positioning for a closer look.

Cheap Seats

Astronomers have now observed three interstellar objects as they’ve flown through the solar system: 1I/ʻOumuamua, which shocked astronomers in 2017; 2I/Borisov, which followed in 2019 and behaved almost identically to a solar system comet; and now 3I/ATLAS, which was found buzzing toward the Sun in July 2025. As soon as 3I/ATLAS was spotted, pretty much every telescope on the planet took a look. The data swiftly revealed much that intrigued astronomers: while 3I/ATLAS was behaving essentially like a standard comet, its chemistry was nothing like its solar system counterparts, and its coma and dust outflow were strangely shaped. Among other oddities, it seemed to counterintuitively have a dust tail that pointed toward the Sun.

Unfortunately for astronomers, 3I/ATLAS didn’t make it easy for them to study the shape of its coma. Thanks to a cosmic coincidence, the interstellar visitor was traveling almost perfectly within the plane of our solar system, and Earthbound observers were often limited to an edge-on view of its tail. Even worse, 3I/ATLAS looped around the back of the Sun from our perspective, meaning no one on the planet could observe the comet for a few weeks in the fall. This led a group of astronomers to consider their assets that are farther from home.

Better View from Mars

The Tianwen-1 spacecraft, which entered Martian orbit in early 2021 and has been diligently mapping the Red Planet ever since, has a high-resolution camera that’s normally pointed at the planet’s surface. By spinning the spacecraft around, a team of researchers led by Xin Ren (National Astronomical Observatories, Chinese Academy of Sciences) repurposed the camera to act as a small telescope when 3I/ATLAS flew by. In another cosmic coincidence, 3I/ATLAS happened to pass very near Mars last fall, and this close approach gave Tianwen-1 a great out-of-plane view of the coma. The orbiter managed to capture 57 images of 3I/ATLAS over three days last fall, representing China’s first-ever observations of an object from an asset in deep space.

The team took full advantage of the great view and leveraged the spacecraft’s unique perspective to extract some intriguing results. By tracking how the gentle pressure of sunlight nudged the escaping grains, the researchers found that the coma is dominated by surprisingly large dust. With an average size of hundreds of microns across, the dust is more like coarse sand than smoke, and they found that it ambles away from the comet at just 3–10 meters per second. All told, 3I/ATLAS was shedding dust at roughly 1,000 kilograms every second. The big grains may be a fingerprint of a cold, distant birthplace: models suggest such grains collect in the frigid outer reaches of planet-forming disks, hinting that 3I/ATLAS was forged far from its parent star before being flung into the dark.

As new surveys begin to turn up interstellar interlopers more often, this improvised observation makes a tidy point: the spacecraft already scattered across the solar system can double as a fleet of opportunistic comet-watchers, each offering a view we simply can’t get from home.

Citation

“Interstellar Object 3I/ATLAS Observed from Mars by China’s Tianwen-1 Spacecraft,” Xin Ren et al 2026 ApJL 1003 L10. doi:10.3847/2041-8213/ae61b3

The post Courtside Seats on Mars appeared first on AAS Nova.

Planets that get too close to their host stars are liable to be engulfed and consumed. New research explores how to identify stars that have had a planetary snack.

It’s Snack Time

Being a planet that closely orbits its star is dangerous, and various dynamical processes can draw a planet inward past the point of no return. When a planet vanishes beneath the surface of its host, it eventually disintegrates within the star, its atoms mixing into the blazing stellar soup.

This outcome is likely extremely common across the universe, and researchers have begun to see evidence for it: stars that appear to have absorbed angular momentum from their planets, white dwarfs with uncanny metals in their spectra, and stars with chemical abundances that don’t match those of their siblings. Now, in a recent publication led by Kaitlyn Lane (Vanderbilt University), researchers seek to find observational signatures of planetary engulfment of rocky exoplanets by main-sequence stars.

Modeling Engulfment

Lane’s team focused on main-sequence stars in the range of 0.5 to 1.4 solar masses, using one-dimensional analytical models to determine which of these stars will most readily display signs of having engulfed a planet. They simulated the stars as they consumed either an Earth-like planet or a super-Earth with the same composition as Earth but 15 times its mass.

At the onset of the model, the engulfment is set to begin, the planet placed tangent to the star’s surface. The planets sink down and orbit just barely within their stars for 1–10 years, their solid surfaces slowly beginning to evaporate.

After that relatively lengthy phase, the process proceeds rapidly: as the planet inches deeper into its star, the drag forces grow stronger and the planet’s demise accelerates, with the planet disintegrating over the course of hours. In the last phase, the planet makes its final plunge and is destroyed in a half hour.

Signs of Planets Past

Lane’s team found that the planets tended to be destroyed fully within the outer convective layer of their stars, except in the case of the most massive stars studied here. For more massive stars, the outer convective zone is shallower than it is for the less massive stars, allowing a chunk of each planet to survive the passage through the convective zone to reach the radiative zone below.

What this translates to, for the purposes of identifying planet-eating stars from afar, is that stars near the top of the mass range studied — from 1.0 to 1.4 solar masses — make the best targets. This is because the tell-tale planetary metals are not overly diluted within the stars’ relatively small convective zones, making these elements easier to detect. In terms of specific metals, the authors found that aluminum, calcium, and vanadium are the best signposts of planetary engulfment, with lithium also making the list.

While the authors acknowledge that there’s more work to be done — they plan to investigate the impacts of planetary engulfment on stellar structure more closely, for example — this work represents an important foray into identifying targets for future studies of stars that have eaten their planets.

Citation

“Observable Metal Pollution in Main-Sequence Stars: Simulations of Rocky Planets Engulfed by Stars in the 0.5 to 1.4 M_⊙ Range,” Kaitlyn T. Lane et al 2026 ApJ 1003 67. doi:10.3847/1538-4357/ae5b9a

The post Now That’s Metal: How to Spot Stars That Have Swallowed Their Planets appeared first on AAS Nova.

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 Panchromatic JWST Spectrum of a Giant Starspot on the Fully Convective M Dwarf TOI-3884
Authors: C. A. Murray et al.
First Author’s Institution: University of Colorado Boulder
Status: Published in ApJ

We often picture stars as smooth, glowing spheres, as if they’ve been run through an Instagram filter. But real stars have spots: cooler, darker regions on a star’s surface caused by strong magnetic fields.

Annoyingly, these spots can seriously interfere with how we study exoplanet atmospheres.

How to Probe a Planetary Atmosphere

Planetary atmospheres are often probed through transmission spectroscopy. When a planet transits its star, some starlight passes through the planet’s atmosphere before reaching us. We can distinguish the light that has passed through the planet’s atmosphere by comparing the star’s light in transit versus out of transit. In doing so, we can isolate the atmospheric spectrum and look for absorption features of species like water or oxygen.

Easy, right?

How Starspots Get in the Way

Things get more complicated once we stop thinking of the star as if it’s been smoothed out by an Instagram filter. Starspots are cooler than the surrounding stellar surface, which means they emit a different spectrum. They are also constantly changing: new spots can form, old ones can disappear, and the star’s rotation carries them in and out of view. This changes things.

First of all, your star’s spectrum changes over time, which could mean that the star’s spectrum out of transit is not the same as the star’s spectrum during transit. On the other hand, the planet may transit across a starspot instead of a “normal” stellar region. In this case, the light we measure during the transit is affected by the spot, and it is no longer accurate to directly compare it with the overall stellar light outside the transit.

Ultimately, absorption features that we thought were from the planetary atmosphere could actually be coming from starspot contamination instead.

Starspot Models: Our Solution?

To deal with this, astronomers build starspot models, where they vary the spots in terms of the following:

• The surface covering fraction: this parameter tells us how much of the stellar surface is covered in spots
• The temperature contrast: this parameter tells us how much cooler (in ratio) the spot is compared to the rest of the stellar surface, and by proxy how much dimmer

These models are promising, but how do we know that we have chosen the right parameters?

A Unique Laboratory: TOI-3884

The TOI-3884 system is a unique laboratory for testing our starspot models. As seen in Figure 1, it has very convenient starspot geometry, with a large starspot located close to its pole. On top of that, we observe this star almost completely pole-on, which means that we always see this starspot, no matter how much the star rotates. To top this all off, the star hosts a close-in planet, which orbits the star from pole to pole.

The planet transits the star, and as it does so, it always passes over the polar starspot. This gives us a rare opportunity to probe the spectrum of a starspot. Similar to how we can use normal in/out of transit observations to probe a planetary atmosphere, we can now compare the observations during/after a “starspot transit” to probe the starspot region. Today’s authors do exactly this using six different transit observations from JWST.

So… How Good Are Our Models?

From the JWST observations, the authors extract a starspot spectrum for the first time. Figure 2 shows this as the spot contrast (how much dimmer the spot is than the surrounding stellar surface) plotted across different wavelengths. To see how well our models are doing, they compare this observed spectrum to two commonly used starspot models. The difference between the data and the models (the residual) is plotted in the lower panel.

At wavelengths longer than about 1 micron, in the near-infrared regime, things look good! The residuals stay small, meaning the models do a solid job reproducing the observations. But move left into shorter wavelengths, in the optical regime, and the agreement quickly falls apart. The residuals grow, and it becomes clear that the models are missing something.

There’s still work to be done before we can confidently probe planetary atmospheres at optical wavelengths without worrying about stellar contamination, and this observed starspot spectrum provides a unique benchmark to test future starspot models. For now, the near-infrared remains a safer and more reliable window for planet atmosphere studies.

Original astrobite edited by Natalie Price.

About the author, Elise Koo:

I’m a PhD student at the University of Amsterdam, working to detect magnetic interactions between stars and their planets using radio and spectroscopic observations. Outside of research, I like to try out a variety of sports.

The post The First Starspot Spectrum Revealed by JWST appeared first on AAS Nova.

The Antennae galaxies are a nearby pair of interacting galaxies, shown in optical light from the Hubble Space Telescope (left) and in infrared light from JWST (right) in the image above. Between the two galaxies’ bright nuclei, there is a vast and dusty molecular cloud complex that harbors a large number of massive young star clusters — and in a recent article led by Rupali Chandar (University of Toledo), researchers have worked to pin down exactly how many clusters are embedded within the dusty whorls of the Antennae galaxies. The team used JWST to search for emission from star clusters too thickly blanketed with dust to be visible in previous observations by Hubble. They ultimately identified 45 sources, 40 of which were seen for the first time in this study. These star clusters mostly inhabit the overlap region between the two galaxies, are less than 2.5 million years old, and each contain roughly 10⁴–10⁶ solar masses. To learn more about how JWST’s infrared capabilities unveiled young star clusters in these interacting galaxies, be sure to check out the full article linked below.

Citation

“Nowhere Left to Hide: Uncovering All of the Massive Young Embedded Star Clusters in the Antennae with JWST,” Rupali Chandar et al 2026 ApJ 1002 64. doi:10.3847/1538-4357/ae56e8

The post Featured Image: Massive Young Star Clusters, Revealed appeared first on AAS Nova.

The fastest spinning pulsars in the universe are often quite difficult to detect. Leveraging multiple observations from the largest single-dish telescope in the world, researchers have discovered six previously undetected pulsars.

Millisecond Pulsars in Globular Clusters

Globular clusters, compact collections of tens of thousands to millions of stars, are highly dense and rife with dynamical interactions. These active environments are ideal factories for the formation of millisecond pulsars — the extremely fast-spinning cores of once-massive stars, shining beams of radio emission from their poles. One way millisecond pulsars are thought to form is when a pulsar in a binary system steals mass from its companion star, transferring angular momentum and causing the pulsar to spin up. The densely packed environment in a globular cluster provides ample opportunity for a pulsar to acquire a companion to siphon.

However, millisecond pulsars in distant globular clusters are notoriously difficult to detect. Traditional pulsar searches that rely on a single observation tend to miss millisecond pulsars, many of which have signals too faint to be distinguished from noise or radio interference. Instead, researchers must rely on a different method to find these fast-spinning pulsars.

Stack Search Success

To find faint millisecond pulsars, researchers have developed the stack search method, in which multiple radio observations are combined to tease out faint signals and decrease noise. Reliably stacking observations across epochs requires stability; isolated millisecond pulsars, now no longer spinning up after losing their companion, have very stable rotation periods, making them ideal targets for the stack search method. To date, the stack search method has successfully identified nine isolated globular cluster millisecond pulsars using data from the Arecibo, Parkes, and Green Bank radio telescopes.

In hopes of recovering more of these faint fast-spinning pulsars, Yinfeng Dai (Beijing Normal University) and collaborators used archival data obtained with the Five-hundred-meter Aperture Spherical radio Telescope (FAST), the world’s largest single-dish telescope. Targeting two specific globular clusters, NGC 6517 and M15, the team applied the stack search method to 23 and 19 observations of the globular clusters, respectively. From their search, the authors discovered six previously unidentified isolated millisecond pulsars: four in NGC 6517 and two in M15.

While previous studies identified brighter pulsars in the targeted globular clusters, the authors confirmed that all six newly discovered pulsars are too weak to be reliably recovered from single-epoch searches, underscoring the importance of the stack search method. In particular, this search increased the known pulsar populations of NGC 6517 and M15 by 27% and 18%, respectively, suggesting that a meaningful fraction of globular cluster pulsars are easily missed in traditional pulsar searches. Future studies employing this stack search method will continue to detect faint millisecond pulsars, unlocking this elusive population for further exploration.

Citation

“The Stack Search Tests on FAST Data: Discovery of Six Faint Isolated Millisecond Pulsars in NGC 6517 and NGC 7078 (M15),” Yinfeng Dai et al 2026 ApJL 1002 L31. doi:10.3847/2041-8213/ae5dbb

The post Stacking Up the Discovery of Six Millisecond Pulsars appeared first on AAS Nova.

Where do black hole mergers happen? Recent research finds evidence that most black hole mergers occur in triple systems containing a close inner binary and a more distant third party.

Assembling Merging Black Holes

Thanks to the combined efforts of the LIGO, Virgo, and KAGRA gravitational wave detectors, we’ve now collected the subtle signals from several hundred pairs of merging stellar-mass black holes. As this collection of spacetime ripples continues to grow, researchers are seeking to understand how these merging black hole pairs were assembled.

In the simplest scenario, pairs of high-mass stars independently evolve into black holes, lose momentum by emitting gravitational waves, and meld into one another. But the population of merging black holes likely contains contributions from multiple sources, including multiple-star systems, the disks surrounding accreting supermassive black holes, and black holes that have already experienced a merger. Which of these merging populations is reflected in our catalog of gravitational wave events?

Modeling Challenges

In theory, this question can be answered by extracting the properties of merging black holes — mass, spin, and the like — from gravitational wave observations and comparing these properties against model predictions from different formation pathways. In reality, this is exceedingly difficult, as certain merger pathways can produce a wide variety of results, depending on assumptions or fine tuning.

So far, the increasingly large pool of gravitational wave detections tentatively points to a feature that could illuminate the sites of black hole mergers: a peak in the spin–orbit tilt distribution around 90 degrees. This feature has been most confidently extracted from nonparametric analyses, which do not make assumptions about the population of merging black holes, but it has also been hinted at in parametric analyses, which do make population-level assumptions.

Triple-System Feature

To investigate this feature further, a team led by Jakob Stegmann (Max Planck Institute for Astrophysics) performed a new, astrophysically motivated parametric analysis of the most recent catalog of gravitational wave detections from LIGO, Virgo, and KAGRA.

Their modeling tests combinations of multiple populations, including mergers in isolated binary systems, triple-star systems, and dense environments like star clusters. They also include a potential contribution from higher-mass mergers involving black holes that have merged previously. Using this framework, Stegmann’s team found that the observed spin–orbit tilt distribution is best matched by a model dominated by mergers in triple-star systems, with some contribution from other sources. Specifically, this finding suggests that most mergers occur in hierarchical triple systems, which naturally predict an excess of 90-degree spin–orbit tilts without any fine tuning. Hierarchical triples contain an inner close binary with a distant outer companion; in these systems, relativistic spin precession, gravitational wave emission, and oscillations in eccentricity and inclination work to tilt the black holes in the inner binary by 90 degrees.

Stegmann and collaborators pointed out that the triple-system scenario aligns well with other properties of the observed population of black hole mergers, such as the mass and effective spin distributions. This scenario may also be necessary to explain certain merging pairs that might have had slightly eccentric orbits at the moment of merging. If future gravitational wave observations strengthen the evidence for a peak in the spin–orbit tilt distribution near 90 degrees, it will have profound implications for how and where black holes merge in our universe.

Citation

“Gravitational-Wave Observations Suggest Most Black Hole Mergers Form in Triples,” Jakob Stegmann et al 2026 ApJL 1000 L59. doi:10.3847/2041-8213/ae52ec

The post Third Star’s the Charm: Most Merging Black Holes Might Be in Triple Systems appeared first on AAS Nova.

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: Retrieving the Red Edge on Earth-Like Planets with Heterogeneous Clouds and Surfaces
Authors: Zachary Burr et al.
First Author’s Institution: Jet Propulsion Laboratory; ETH Zurich
Status: Published in ApJ

In the next 25 years, astronomers could find signs of life on other planets with the launch of the Habitable Worlds Observatory. Atmospheric signs of life, known as atmospheric biosignatures, are how astronomers currently search for life on other planets. Figure 1 shows examples of atmospheric biosignatures with pictures of each chemical’s origin on Earth. Before E.T. can phone home, though, we have to understand what chemicals make up a planet’s atmosphere. To do this, astronomers can use spectroscopy to observe a planet’s atmosphere in different wavelengths and measure the types and quantities of chemicals present.

There are many ways to observe the spectrum of a planet, but the one you’re probably most familiar with is transmission spectroscopy. You can learn more about transmission spectroscopy of exoplanet atmospheres and how to model transmission spectra with these bites: 1, 2, 3. However, all of the current techniques for exoplanet spectroscopy (including transmission spectroscopy) measure the star’s light and how the planet affects the star’s light. This means it’s heavily dependent on how we model the star’s light to indirectly observe the planet’s light. The only way to take the spectrum of a planet directly is with direct imaging: the technique that the future Habitable Worlds Observatory will use to find signs of life in planetary spectra.

But are atmospheric biosignatures the only way to tell if life exists on other planets? Short answer: absolutely not. Let’s take Earth as an example. If we imagine Earth as a directly imaged exoplanet and take spectra of the light it reflects, you’ll find that around near-infrared wavelengths (~700 nanometers or just past the red part of the visible light spectrum), the light reflected off of Earth increases sharply. This sharp increase is called the vegetation red edge and is caused by plant and ocean life on Earth’s surface as shown in Figure 2. This is made possible because chlorophyll (the thing that makes plants green) absorbs nearly all visible light but reflects near-infrared light. The vegetation red edge is an example of a surface biosignature. In today’s article, the authors determine if the vegetation red edge will be visible in spectra of Earth-like exoplanets!

Snapshots

Planets rotate; that’s how we have morning, noon, and night in 24 hours guaranteed. This means that different parts of the surface will be visible at different times when we take spectra. Sometimes, we may see more ocean than land, other times not. To determine if the vegetation red edge would be visible at all points in Earth’s day, the authors simulated nine different spectra of modern-day Earth representing nine snapshots of Earth’s rotation. Additionally, they also simulated a time-averaged spectrum that shows an entire Earth day in one snapshot without the time variability. These spectra are shown in Figure 3.

The authors then run retrieval models on all nine spectra to see if the vegetation red edge would be detectable in each spectrum, including the time-averaged one. Retrieval models take into account characteristics of the planet that would affect how light moves through its atmosphere, such as the atmospheric chemicals present and the surface gravity. They then use this information to generate possible spectral models that could fit the data. The best-fit model allows astronomers to infer what the atmosphere is made of and properties of the planet such as surface gravity. Because we are looking at Earth in visible light, much of the light from the planet is reflected. Typically, models of Earth-like planets in visible wavelengths only include atmospheric albedo: the measure of how reflective a planet’s atmosphere is. However, surface biosignatures can also be reflective as shown by the vegetation red edge. The authors use this to their advantage and introduce a surface albedo into their model that changes with the amount of desert, vegetation, and ocean visible on Earth’s surface at that time. They then further divide the nine spectra into three different types based upon the visible surface: majority land; 50% land, 50% ocean; and majority ocean. Check out the article online to see a cool animation of the Earth rotating and how the spectrum looks different at each time!

Arr! There Be… Vegetables?

After running the models, the authors were able to detect a sharp increase in the surface albedo value in every spectrum, including the time-averaged one. This result is shown in Figure 4. The authors take this result one step further in order to validate that the vegetation red edge is actually a result of surface variation of vegetation. They re-run models on each of their spectra but instead include a constant surface albedo instead of a time-varying one. This resulted in incorrectly measured radii, chemicals, and surface gravity in nearly all of the models. Without the vegetation red edge and surface features, the models could not accurately determine if life was present on Earth.

For the first time (to the author’s knowledge), the vegetation red edge has been shown to be a promising and observable biosignature in Earth-like atmospheres. This study has also laid the groundwork for future work on other surface biosignatures and their potential impact on spectra of habitable planets. Confirming signs of life in multiple different ways, both on the surface and in the atmosphere, will finally allow astronomers to say we are not alone in the universe. In fact, life likes to live on the (red) edge!

Original astrobite edited by Madison VanWyngarden.

About the author, Kaz Gary:

I am a fourth-year PhD candidate at The Ohio State University with a passion for planets. My current work focuses on modeling exoplanet observations for the Habitable Worlds Observatory and understanding planetary atmospheres. Outside of research, I help develop planetarium shows and love all forms of science communication. In my free time, I enjoy playing tabletop RPGs, painting, watching terrible reality TV, and hanging out with my pet hedgehog.

The post Life on the (Red) Edge appeared first on AAS Nova.

Each Jupiter-size planet in the galaxy falls into one of three distinct categories: hot, warm, or cold. A new study suggests that despite the apparent differences between these populations, they may have all formed from the same underlying dynamical process: a game of pool played at planetary scales.

A Diversity of Jupiters

Though our solar system has only one Jupiter-size planet, elsewhere in the galaxy we have found three different species of these massive gas giants. Cold Jupiters closely resemble their namesake and orbit far from their host stars; hot Jupiters are the opposite and are found whipping around their stars on extremely close-in orbits. In between these are the warm Jupiters, which tend to orbit in the intermediate space between 0.1 and 1.0 au.

Though these three populations are defined by their orbital distances, they differ from each other in other ways as well. For example, hot Jupiters almost never have nearby companions; if there are any other planets circling the same star, they’re usually far-out cold Jupiters. They can also orbit in pretty much any direction, including opposite the direction of their star’s spin, and are usually on perfectly circular orbits. In contrast, warm Jupiters often have friends nearby, are much more aligned with their stars’ spins, and can have modest eccentricities.

Given these differences, it’s often thought that each of these populations arrived at its current location through different dynamical processes and that the history of the warm Jupiters is likely quite different from the history of the hot Jupiters. However, a new study led by Julia Esposito (Georgia Institute of Technology) has proposed an alternative view. Maybe these populations, though they appear different now, were all created by the same process: planet–planet scattering.

Virtual Planetary Billiards

Esposito and collaborators set up 1,500 virtual planetary systems with three massive planets each, then simulated how the orbits evolved before probing the final configurations. In crucial contrast to previous simulation studies, the team initialized their Jupiters across a range of different distances and included the effects of tides sapping energy from orbits of planets that got too close to their host stars.

At the end of the simulation, the team surveyed the digital carnage. Almost every virtual system ended with only two planets after either ejecting one away from the star or having two planets collide. But, remarkably, the remaining two-planet systems looked tantalizingly similar to what we actually observe, with a mix of hot, warm, and cold Jupiters. Even more exciting, the end populations were highly correlated to where the violent scattering event took place.

For example, the warm Jupiters were almost all produced by “warm scattering” simulations, where the scattering took place between 0.1 and 1.0 au. The planets that survived the simulations and ended up as warm Jupiters matched all of the properties of the real warm Jupiter population: they had nearby companions, were moderately eccentric, and were mostly aligned with their stars. The hot Jupiters, meanwhile, were almost all produced by “cold scattering” events where the flybys happened far from the star and resulted in one planet hurtling inwards. These also matched all of the observed properties of real hot Jupiters.

The researchers concluded that planet–planet scattering can produce both the warm and hot Jupiter populations so long as you let the planets scatter from a variety of different distances. This exciting theoretical insight, if correct, would mean that astronomers could stop searching for different pathways to create each population. Happily, this model also provides testable predictions, and the authors lay out how the theory could be supported or disproven with additional data. Through virtual experiments like these, astronomers continue to build up an understanding of how the wide range of planetary architectures observed across the galaxy came to be.

Citation

“Unified Formation Channel of Hot and Warm Jupiters via Planet–Planet Scattering,” Julia Esposito et al 2026 ApJL 1003 L3. doi:10.3847/2041-8213/ae61b0

The post Playing Pool with Planets appeared first on AAS Nova.

Discovered in December 2024, the roughly 60-meter-wide asteroid 2024 YR4 was initially thought to have a slim chance of striking Earth in 2032. Further observations ruled out an Earth impact but left open the possibility of a Moon impact. New JWST observations show that 2024 YR4 will speed past both Earth and the Moon without a collision.

Another Update on 2024 YR4

In April 2025, we featured a Research Note that described March 2025 JWST observations of 2024 YR4 and characterized the asteroid’s size and albedo. At the close of that article, the authors suggested that further JWST observations in early 2026 could help to refine our understanding of the asteroid’s trajectory and determine whether it would strike the Moon.

On 18 and 26 February 2026, JWST carried out those critical observations. In this dataset, the asteroid is bright enough to be detected confidently, and the background is peppered with a sufficient number of reference stars — two factors necessary to determine the precise track of the asteroid as it zipped through space.

Trajectory Refinement

Julien de Wit (Massachusetts Institute of Technology) and collaborators analyzed the new JWST data and determined the asteroid’s position. The team used multiple methods to reduce the data and showed that their results were consistent regardless of the reduction method, ultimately yielding a measurement with an uncertainty of just 50 milliarcseconds. This is a huge improvement over the previous trajectory estimate, which had an uncertainty of 700 milliarcseconds.

Combining the latest JWST observations with existing data stretching back to the asteroid’s discovery, de Wit and coauthors determined a new, highly precise orbit for the object — one that definitively shows that the asteroid will safely pass by the Moon at a distance of 22,900 ± 800 km in 2032. In addition to ruling out the upcoming lunar impact, the team also ruled out any collisions with Earth in the next 100 years. Phew!

High-Impact Science

This work highlights JWST’s potential as a tool for planetary defense. The February 2026 observations detailed here represent the faintest-ever detection of a near-Earth object, and JWST can detect objects far fainter than ground-based telescopes can. This ability allows JWST to observe potentially hazardous asteroids over a larger swath of their orbits, which translates to earlier trajectory estimation — JWST pinned down 2024 YR4’s orbit two years before the asteroid would have become bright enough for ground-based telescopes to do so.

JWST’s role may be even more important for smaller objects that rapidly become undetectable after discovery. De Wit’s team demonstrated this for a 10-meter-diameter 2024 YR4 clone on the same trajectory; for an object of this size, JWST would be able to rule out an Earth impact before the asteroid became undetectable, but ground-based telescopes could not. With surveys from facilities like Vera C. Rubin Observatory and Near-Earth Object Surveyor set to discover hundreds of thousands of new near-Earth objects, JWST’s role as a hazard assessor will likely be in high demand.

Citation

“JWST Observations of Asteroid 2024 YR4 Rule Out a 2032 Lunar Impact and Demonstrate a New Regime for Planetary Defense Follow-Up,” Julien de Wit et al 2026 ApJL 1003 L21. doi:10.3847/2041-8213/ae6a95

The post Asteroid 2024 YR4 Won’t Hit the Moon, New JWST Observations Show appeared first on AAS Nova.

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: Shocks, Winds, and a Torus: The Large Binocular Telescope Interferometer (LBTI) Resolves the Active Nucleus of NGC 4151
Authors: Jacob W. Isbell et al.
First Author’s Institution: University of Arizona
Status: Published in ApJ

At the centre of nearly every galaxy lies a supermassive black hole that dominates this innermost region. Not only is there this millions-of-solar-masses dark beast, but usually too a whole bunch of stuff — stars, gas, dust, and more — which is often quite bright! When there is plenty of this material quite close to the black hole, we believe physics takes control to flatten it into a family of disk and toroidal structures in what we call the unified model of active galactic nuclei (AGNs; see Figure 1).

When astronomers look at different active galaxies (read: galaxies with active supermassive black holes at their core), we see a range of phenomenologies related to their brightness, spectra, morphologies, and more. The unified model of AGNs seeks to explain these different AGN appearances simultaneously by positing that they all have the same physical structure, and we are just viewing them from different angles, hence seeing different features.

The cores of AGNs are often imaged at the smallest scales (e.g., their accretion disks, viewed with interferometers like the Very Large Telescope Interferometer) and the largest scales (e.g., their relativistic jets, viewed with radio interferometers), but comparatively less effort has gone to directly observing the predicted dusty tori in the mid-infrared. That is exactly what today’s authors set out to do using the Large Binocular Telescope Interferometer (LBTI) — a pair of 8.4-metre-aperture mirrors separated by just over 14 metres and combined to simulate a telescope effectively 29 metres wide. This lets astronomers take direct images at a much higher resolution than a smaller-aperture telescope, and hence directly peer into the region around AGNs where this dusty torus should lie.

Today’s authors turn the LBTI towards NGC 4151, a medium-luminosity AGN. With the large effective aperture of the LBTI, they were able to resolve scales in the AGN region as small as 4.4–9.1 pc, about 14–30 light-years depending on the wavelength (see Figure 2), in the mid-infrared. These observations revealed warm dust emission in a complex structure around the central supermassive black hole. The authors note a central bar at all wavelengths, with a significant extension of cool dust arcing to the west (right in the image) and warmer dust localised to the centre (as evident by 3.7- and 4.8-micron emission only nearest to the supermassive black hole and its accretion disk).

To explain the observed morphology in Figure 1, the authors compare three different interpretations based on this new high-resolution imagery together with the results of previous studies looking at other scales and the spectrum of the AGN core. Each interpretation is illustrated and briefed in Figure 3. The first interpretation is in keeping with the unified model of AGNs: a geometrically and optically thick torus of dust surrounds the inner region. In this interpretation, the surface of the dusty torus re-radiates light from the accretion disk to produce most of the mid-infrared emission. Additional mid-infrared emission could come from localised concentrations of gas or interactions between the outflow and the jet. The authors immediately disfavour this explanation, as other studies have shown bright ionised emission at the location of the would-be torus, which is not expected if an optically thick torus were to be present. Hence, these new observations somewhat challenge the one “flavour” of the unified model of AGNs.

The second interpretation aligns with a different version of the unified model: one in which a geometrically thin disk replaces a thick torus around the AGN core. Provided the disk is optically thin too, the authors favour this approach as it is consistent with the geometry of ionised emission that worked against the first interpretation. While other studies suggest that this thin disk should be optically thick, this morphology is at least better aligned with what we see in other AGNs.

The third interpretation suggests that the emission comes from only the radiation pressure–driven wind emanating from the AGN core. The authors disfavour this explanation, citing previous radiative transfer simulations that show that the flux should fall off with distance from the core too quickly to be consistent with the observations.

No matter the interpretation, these LBTI observations are an important glimpse into the future of mid-infrared AGN studies that will be done with 30-metre-class telescopes (such as the European Southern Observatory’s Extremely Large Telescope). This article, together with the group’s similar study on NGC 1068, is challenging and refining an accepted view of AGN morphology — the consequences of which apply to galaxies near and far — and poses new questions perfect for sophisticated hydrodynamic and radiative simulations.

Original astrobite edited by Margaret Verrico.

About the author, Ryan White:

I am a first-year PhD student at Macquarie University in Australia, working mainly on binary/multiple systems with massive stars (Wolf–Rayets in particular!). Outside of study, I’m probably drinking coffee, baking, reading, or going for a run. You can also find me procrastinating on Bluesky @astroryan.bsky.social.

The post I Spy with My Large Binocular Eyes: A Dusty Torus Around a Supermassive Black Hole? appeared first on AAS Nova.