It’s possible that I shall make an ass of myself. But in that case one can always get out of it with a little dialectic. I have, of course, so worded my proposition as to be right either way (K.Marx, Letter to F.Engels on the Indian Mutiny)
Friday, October 16, 2026
Not all early birds evolved toward better flight, Jurassic fossil suggests
For early birds, evolution wasn’t always about becoming better flyers. This conclusion is supported by the recent discovery of a remarkably well-preserved, 150-million-year-old bird fossil from Fujian Province, China, which is challenging long-held assumptions about early bird evolution.
A study published in Nature discusses the newly discovered species, Yuezhuornis yangyuanensis, which lived during the Late Jurassic period (150–148 million years ago). The research suggests that the evolution of the avian body plan was not a straightforward progression toward improved flight capabilities, but a complex process involving the exploration of different ecological niches, including some that did not require flight.
Discovered in the recently recognized Zhenghe Fauna of southeastern China by researchers from the Institute of Vertebrate Paleontology and Paleoanthropology (IVPP) of the Chinese Academy of Sciences (CAS) and the Fujian Institute of Geological Sciences, the crow-sized fossil represents the earliest known example of flight loss in the evolutionary history of birds.
Yuezhuornis challenges the traditional view that the avian body plan evolved almost exclusively under selective pressures associated with flight. Instead, it exhibits a striking mosaic of primitive and derived features, combining characteristic avian traits with features inherited from earlier dinosaur ancestors.
The fossil preserves closely packed, serrated teeth, which are typically found in deinonychosaurs but not birds. Most notably, Yuezhuornis possessed a hyperextensible second toe with a sickle-shaped claw, a characteristic feature of deinonychosaurs, including dromaeosaurids and troodontids.
“When we examined Yuezhuornis, we were struck by its extraordinary mix of features,” said Dr. WANG Min, corresponding author of the study. “It possesses bird-like pelvic and hindlimb structures yet retains the specialized predatory foot and teeth of a small meat-eating dinosaur. It shows us that early bird evolution was far more complicated than we previously realized.”
To determine where this enigmatic species fits on the evolutionary tree, the research team conducted comprehensive phylogenetic analyses using both maximum parsimony and Bayesian tip-dating methods.
“The results consistently placed Yuezhuornis firmly within Avialae (the group containing modern birds and their closest extinct relatives), just crownward of Archaeopteryx and as a sister taxon to Zhengheornis,” said Dr. ZHOU Zhonghe, a co-author of the study.
Given its unique anatomy, the team investigated the ecological niche of Yuezhuornis using ecomorphometric and multivariate statistical analyses. The analyses revealed that the limb proportions of Yuezhuornis were similar to those of living terrestrial birds and some non-avialan paravians, suggesting a terrestrial, flightless lifestyle and shedding new light on the ecological diversification of early-diverging avialans.
The discovery of a flightless, ground-dwelling bird from the Late Jurassic challenges the notion that aerodynamic selection was the dominant force shaping early avialan evolution. Instead, it suggests that some early birds followed evolutionary paths that did not favor flight, highlighting the remarkable ecological and morphological diversity that emerged during the early evolution of birds.
The ongoing discovery of early avialans also highlights the potential of the Zhenghe Fauna to shed light on a poorly documented stage of Jurassic bird evolution.
Life reconstruction of the Late Jurassic bird Yuezhuornis yangyuanensis.
Eco-morphometric disparity of Yuezhuornis and other Jurassic paravians.
Evolutionary tree showing the position of Yuezhuornis as one of the earliest branching of birds. The right panes show the fore limb evolution across dinosaur-bird transition.
Credit
Image by Min Wang.
New ‘Reversing Diabetes’ documentary reveals the revolutionary approach that is saving lives
WASHINGTON, D.C. — More than 40 million Americans are living with diabetes, but counting carbohydrates, portion control, and other standard dietary recommendations aren’t easing the epidemic. Reversing Diabetes, a groundbreaking new documentary, reveals the new nutritional method that is helping people dramatically improve—and even reverse—type 2 diabetes. The film will be available to rent or buy beginning Nov. 10 on Amazon Prime Video, Apple TV, Fandango, and Google Play. The trailer is available at DiabetesFilm.org.
Marc Ramirez, one of several people who share their struggle with diabetes in the documentary, was diagnosed with type 2 diabetes in 2002 and thought his genes were the cause.
“I tried to count carbs, count calories, portion control, exercise more, eat less,” says Ramirez. “Nothing really worked. I wasn’t able to eliminate my medications. I was still sick.”
But just months after making a revolutionary diet change, Ramirez reversed his diabetes and was able to stop taking his medications, along with other health improvements.
In the film, produced by the Physicians Committee for Responsible Medicine, nearly a dozen leading physicians and nutrition researchers, including Neal Barnard, MD, FACC, diabetes researcher and president of the Physicians Committee; Dean Ornish, MD; Kim Williams, MD, past president of the American College of Cardiology; and Kitt Petersen, MD, professor of endocrinology at the Yale School of Medicine, share their research and the scientific evidence demonstrating the diet that is helping people like Ramirez reverse their diabetes.
“What’s missing from typical diabetes treatment is addressing the cause of the disease,” says Dr. Barnard, whose 2003 National Institutes of Health-funded study first showed the diet change that actually helped people lower blood sugar, reduce medications, and in some cases actually reverse diabetes.
It is an approach that also worked for Stephanie Goettge, Andrew Storck, and Donna Jackson, who all share their stories of overcoming their diabetes. The film also highlights how diet can benefit people with type 1 diabetes, like David Marsh, who reduced the amount of insulin he needed.
To help viewers who are interested in putting the diet into action for themselves, dietitians share ideas for easy and delicious meals, tips for eating out, and more.
“Reversing Diabetes provides a lifesaving message that can revolutionize the way physicians and people with diabetes treat the disease,” says Dr. Barnard.
Founded in 1985, the Physicians Committee for Responsible Medicine is a nonprofit organization that promotes preventive medicine, conducts clinical research, and encourages higher standards for ethics and effectiveness in education and research.
Researchers improve weather forecasts looking two to six weeks out
A new machine-learning approach broke records for weather forecasting for an important time frame in emergency planning.
Researchers developed a machine-learning model that learns to correct errors in existing forecasting models to make their output more accurate over the two-to-six-week time frame, a key window for emergency planning.
When applied to a gold-standard forecasting model, the new method doubled forecasting accuracy over that time frame, placing first in a prominent international forecasting competition.
The new method could provide energy operators, farmers and emergency response officials around the world more time to better prepare for storms.
Cambridge, Mass. -- In 2021, Texas suffered the worst energy infrastructure failure in state history, following three severe winter storms in February. At least 246 people were killed as a result of the disruptions, which led to food, water, heat and electricity shortages.
If energy officials had had more time to prepare for the storms, for example by bringing in more trucks of natural gas to disrupted areas, damage from the event could have been mitigated. Unfortunately, weather forecasts beyond two weeks are notoriously unreliable, making for a tighter window for energy operators, farmers, and emergency response officials around the globe.
A new technique by a team including researchers at MIT significantly improves weather forecasting beyond two weeks. The researchers developed a machine-learning model that learns to correct errors in existing forecasting models to improve their output over the two-to-six-week time frame. When applied to a gold-standard forecasting model, it doubled forecasting accuracy over that time frame, placing first in a prominent international forecasting competition.
In a new paper in Nature, the researchers describe their approach and show how it enables forecasts that for the first time beat climatologic averages between two and six weeks.
“Not only does this do better on time averaged forecasts for this range, which is already very useful for farmers and members of the energy industry trying to match supply with demand, but it also improves predictions of extreme weather events, whether it’s flooding or cold spells or something else,” co-author and MIT Research Scientist Judah Cohen says. “It’s really important to predict both, but extreme events might be the most impactful part of this for society.”
Joining Cohen on the paper are lead authors Hannah Guan, a Harvard College undergraduate; Soukayna Mouatadid, a University of Toronto postdoc; Genevieve Flaspohler SM ’18, PhD ’22; Paulo Orenstein, an associate professor at Brazil’s National Institute for Pure and Applied Mathematics; Jeremy Berman of the data company DTN; Jakob Schloer and Joshua Talib, researchers with the European Center for Medium-Range Weather Forecasts; and Alex Lu, Haiyu Dong, Jonathan Weyn, Zekun Ni, and Lester Mackey, all researchers at Microsoft.
Longer forecasts
In the 1970s, longtime MIT Professor Edward Lorenz came up with the idea of chaos theory to describe how tiny changes in initial weather conditions lead to completely different outcomes.
“He came up with this theoretical two-week limit where beyond that, we just can’t predict weather because the noise in the system grows so much that it overwhelms any signal you start with,” Cohen says. “That had a strong influence on the field that may have held people back from exploring forecasting beyond the two-week range. I never really believed that completely. In theory the atmosphere can create an infinite number of patterns, but it likes to fall into these favorite patterns over and over again. Once one thing happens, I think you can predict the domino that will fall after that, and the domino that will fall after that.”
Today, the best weather forecasts come from either dynamical models, which numerically simulate how the atmosphere and oceans evolve over time, or artificial intelligence models, which learn to make predictions from historical weather data. AI models have shown improvements in predicting weather at short ranges, but they’re even worse than dynamical models at longer time ranges.
Cohen, who is based in the MIT Department of Civil and Environmental Engineering, focuses on forecasting weather two to six week in advance, which is known as subseasonal forecasting.
“That range is referred to as the ‘predictability desert,’ because in the short range you can capitalize on the initial conditions in the climate system and the atmosphere, and on the longer term you have these slow, varying boundary forces, like the super El Niño everyone’s focused on this winter,” Cohen says. “The subseasonal range can’t take advantage of either of those. In most cases forecasting in the subseasonal ranges is no different than going to a casino and rolling the dice.”
Cohen credits his co-author Lester Mackey, whom he has been collaborating with for more than a decade, with the initial idea of using machine learning to reduce errors in forecasting models.
“We built this wrapper — Lester calls it postprocessing,” Cohen says. “We know AI models have a lot of errors, so we thought maybe we can quantify those errors and remove them.”
While other forecasting systems do more simplistic error correction, the research team’s postprocessing layer uses AI to learn from a forecasting system’s mistakes based on historic data and shifts its forecasts. In that time range, forecasters use so-called probabilistic forecasts, where the likelihood of a range of possibilities if given.
“As more forecasts and data become available, our model adaptively improves its correction,” Cohen says.
To build the model, the researchers used MIT Engaging, a supercomputing cluster hosted by MIT’s Office of Research Computing and Data, but the resulting model does not require a supercomputer to run.
To put their model to the test, last year the research team submitted three models that included their post-processing layer to the European Center for Medium-Range Weather Forecasts (ECMWF), which is considered the best forecasting center in the world. Each season ECMWF holds forecasting competitions between models.
The submission that won first place combined ECMWF’s forecasting model with another, and then applied the postprocessing correction on top. That approach ended up doubling the original model’s predictive power in the subseasonal range. For the first time, the model offered an improvement over relying on climatologic averages.
“Up until now, no one had been able to show a model predicts weather better than long term averages for these subseasonal forecasts,” Cohen says. “They’ve pretty much all been worse than that — you could literally flip a coin and do better. People have questioned if it’s even worth investing in models at all for this time range. We consistently beat climatology averages with our models.”
The model also showed a statistically significant improvement in the ability to predict extreme weather.
“That’s one of the more important findings, because it’s so difficult to accomplish,” Cohen says. “When the ECMWF developed their own, non-AI, error-correction to their model, it got worse at predicting extreme weather. I think that’s because traditional error, or bias, correction is always trying to get to the mean, whereas extreme weather is on the fringe of probabilities.”
Preparing for disaster
In the Nature paper, the team showed their technique could be used to improve land forecasts in most regions across the globe. They also applied their bias correction technique to other models and showed it improved those as well, putting even lesser models on par with the ECMWF’s forecasts.
“What’s nice about our model is it’s an add-on wrapper,” Cohen explains. “We used it with the ECMWF’s dynamical model because that was considered the best, but it can be used with any model. It even improves models that are already bias corrected using another AI algorithm. So it shows a robust improvement. And it’s open source and low cost.”
Cohen says the meteorological office for the Republic of Kenya has already reached out about implementing the model to improve forecasting of the annual monsoon season to improve agricultural yields this winter. He hopes other groups will also use the model.
“This could help you decide when to fertilize around rain, when there’s a freeze coming to protect crops during winter harvests, or help you avoid droughts,” Cohen says. “It could also be useful for energy and emergency management. I’ve talked to energy officials who say, ‘If I’m told two days before that an ice storm’s coming, I don’t have enough time, but if I had two weeks, that could be enough to make adequate preparations.”
The work was supported, in part, by the U.S. National Science Foundation, the Carlos Chagas Filho Foundation for Research Support, and the European Union.
“Advancing Subseasonal Forecasting with Machine Learning”
Satellites accurately detect methane emissions in large-scale test
Researchers analyzed spaceborne instrument observations of controlled methane releases in a yearlong experiment. The results can inform efforts to identify major methane sources, eliminate leaks, and reduce emissions
Researchers have conducted the largest ever controlled test of space-based methane monitoring, finding that dedicated methane sensors correctly detected nearly all methane plumes released on clear days.
Published October 14 in Nature, the results provide a benchmark for satellite systems increasingly used by regulators, environmental groups, and companies to detect leaks and identify super-emitters of the planet-warming gas.
“There has been significant skepticism over the years about whether methane tracking really works because it’s a challenging technical problem to detect the gas from space,” said senior study author Adam Brandt, a professor of energy science and engineering in the Stanford Doerr School of Sustainability. “Our results should give regulators, policymakers, producers of methane, and the general public a lot of confidence moving forward.”
Testing satellite observations and algorithms
Together with PhD student and lead author Frances Reuland, Brandt and collaborators timed 481 methane releases at a Stanford research site in Casa Grande, Arizona, to overpasses by 11 government, nonprofit, and commercial satellites throughout 2025.
Fifteen research teams from corporations, nonprofits, government agencies, and academia across nine countries independently analyzed the observations, assessing whether each satellite image showed a methane plume and, if so, how much methane was released.
The study authors compared the resulting estimates with the known releases. Instruments designed specifically for methane monitoring produced the best detection rates, correctly identifying the presence of 92-100% of test plumes.
Most estimates based on these observations were up to 50% higher or lower than actual emission amounts, but the over- and underestimates roughly balanced out on average. That makes the estimates reliable enough to flag large leaks for follow-up, inform governments’ regional greenhouse gas emissions inventories, and assess the relative emissions intensity of different oil and gas regions, the authors said.
“In many cases, the detection is the important part,” said Brandt. “As long as you know approximately how big it is, then you still know you need to go fix it.”
Under ideal conditions, with few clouds and moderate wind, these instruments reliably detected plumes when methane flowed at rates down to about 50-200 kilograms per hour, depending on the system. This emission rate is close to the threshold of 100 kilograms per hour that triggers reviews under some state and national “super-emitter” programs.
“These results were great to see, and we hope they build confidence in this field overall,” said Reuland, who is a member of Brandt’s lab and a PhD student in energy science and engineering in the Doerr School of Sustainability.
Satellite-based monitoring systems detect methane using infrared light reflected off Earth’s surface. That reliance on reflected sunlight means even the dedicated instruments aboard GHGSat and Tanager-1 satellites struggled to detect methane plumes through clouds.
“This poses problems in some key methane-emitting regions,” the authors write, “including tropical oil and gas producers Nigeria and Venezuela, as well as cold, cloudy climates like Canada and Russia.” More research is needed, they added, to understand how these satellite systems and algorithms work over non-desert terrains, across various weather conditions, and over open water.
Opportunity to improve AI-powered methane monitoring
When skies over the Casa Grande test site were clear, instruments repurposed for methane sensing, such as Germany’s Environmental Monitoring and Analysis Program, or EnMAP, detected 62-88% of methane plumes. General-purpose land-scanning systems aboard U.S. and European government satellites detected just 29-37% and proved especially error prone for releases below 1,000 kilograms per hour.
The land-scanning systems gather less of the spectral data required to isolate methane’s unique signature. Their extensive coverage and historical archives remain valuable, however, for documenting long-term trends and identifying large leaks, the researchers said.
The authors gave analytics teams observations from 265 overpasses when no methane was released to test for false positives, which can reduce public trust in results and waste resources on efforts to address non-existent leaks.
These proved relatively rare with most methods, except when teams analyzed data with emerging artificial intelligence approaches. According to the study, 79% of false positives originated from commercial teams applying AI-powered methods to data from land-scanning government satellites.
The researchers said the AI methods tested need improvement, and that the new data from the Casa Grande experiments could help refine them. “With this study, we have orders of magnitude more controlled-source data points than we’ve ever had before for testing methane detection from space,” said Reuland.
Approximating real-world emitters
At the research site, researchers piped compressed natural gas, which is mostly methane, through a release stack, setting the emission rates to approximate point sources such as landfills and oil and gas facilities.
Controlled through an online interface, the releases lasted 8 to 10 minutes and ranged from about 10 to 1,200 kilograms of methane per hour. Over the full year of the experiment, the team released less than 50 metric tons of methane, equivalent to the annual emissions from a herd of about 500 dairy cows or about 13 seconds of methane emissions from global fossil fuel energy operations.
“By establishing these findings,” Reuland and co-authors write, “we hope this moves the conversation beyond if these systems work, toward discussions of how, given algorithmic trade-offs, and where they can be most effectively applied to drive global methane mitigation.”
‘Multiple motivations’ for fixing methane leaks
Rapidly reducing the world’s methane emissions, the biggest driver of climate change after carbon dioxide, is a major focus of international efforts to slow global temperature increases. Carbon dioxide lasts longer, but methane is 80 times more potent at trapping heat during its first 20 years in Earth’s atmosphere, giving it an outsized impact on near-term temperatures.
If monitoring and detection of leaks and intentional venting lead to repairs and overall emission reductions, this can deliver multiple benefits in addition to mitigating climate change. Some relatively low-cost repairs to oil and gas infrastructure, one of the largest sources of human-caused methane emissions, can yield financial gains for operators by keeping saleable gas from escaping, the authors said.
“Plus, there are safety issues associated with releasing methane because it is flammable and it contributes to ground-level ozone, a health hazard,” said Brandt. “There are multiple motivations here for supporting satellite monitoring where it’s win-win-win for the environment, the producers, and the general public.”
Co-authors of the study are affiliated with the University of Michigan, Ann Arbor; University of California, Berkeley; University of Iowa; and the Max Planck Institute for Chemistry.
Brandt is also a senior fellow at the Precourt Institute for Energy, part of the Doerr School of Sustainability. He is a member of Carbon Mapper’s Science and Measurement Committee, which co-author Eric Kort chairs.
Analysis teams included Carbon Mapper; the International Methane Emissions Observatory (IMEO); NASA JPL; and Space Research Organization Netherlands (SRON); Universitat Politècnica de València; Nanjing University; University of Bremen, Germany; GHGSat; Google; Kayrros; Orbio Earth; Arka; Orbital Sidekick; Geolabe; and Momentick.
Funding was provided by the Global Methane Hub, the Grantham Foundation, Saudi Aramco, and the Stanford Knight-Hennessy Scholars program, where Reuland was a scholar in the 2023 cohort.
This work was also supported by the Natural Gas Initiative, part of the Center for Fuels of the Future industrial affiliates program. Stanford industrial affiliates programs are funded by membership fees from companies. View current Stanford Doerr School of Sustainability affiliates members.
Plastics are supposed to be chemically inert. It’s the reason they’re used so frequently to package food, bottle water, and create children’s toys: They’re safe.
But something important happens when a large piece of plastic fragments into micro- and nanoparticles. Its surface area increases dramatically relative to its volume. Vastly more of the material is now exposed to its surroundings.
“As plastic fragments into smaller and smaller particles, its surface area increases enormously. Then microplastics change the water that surrounds them, generating high-energy, reactive radicals (reactive oxygen species, ROS) that very likely contribute to biological damage that we see in animals and potentially in humans as a result of microplastic exposure,” said Wei Min, a Columbia professor of Chemistry and Biomedical Engineering.
The work challenges the common view of microplastics as largely passive fragments of inert material, and contributes to a growing understanding of why the large-scale presence of microplastics is a cause for concern. The research was published today in the journal Nature.
This is particularly noteworthy for human health because human blood is largely made up of water. The new research suggests that microplastics that enter the bloodstream or cells could remain chemically active in this water-rich biological environment.
Min’s group at Columbia University also led a 2024 study that developed a laser based microscope technology and found that on average, a liter of bottled water contained some 240,000 detectable plastic fragments (mostly nanometer-sized plastic, nanoplastics)—10 to 100 times more fragments than had previously been estimated. That study led to a growing recognition that micro and nanoplastics exist throughout the food system.
The group’s 2024 research on nanoplastics garnered widespread media attention, and combating microplastic has become a public health priority. This year, the U.S federal government launched a $144 million program to combat toxic microplastics in the human body. The U.S. Environmental Protection Agency also recently added microplastics to a federal list of contaminants that may pose a concern for drinking water, the first time microplastics have appeared on the list.
Subsequent research in prominent journals including New England Journal of Medicine,Nature Medicine, and Nature Health have found links between micro and nanoplastic exposure and heart attack, stroke, dementia, and brain tumor. But, given its supposed inertness, the mechanism by which microplastics could contribute to these illnesses had been unclear.
This new study specifies a previously overlooked mechanism that might help bridge this gap.
“Our research found that when microplastics interact with water, as they do when they’re in bottled water, or when they’re absorbed into our bloodstream, they spontaneously generate what we call reactive oxygen species (ROS),” said Naixin Qian, a postdoctoral researcher in Wei Min’s lab who helped lead the research. “ROS are textbook examples of the kinds of molecules that can damage animal and human tissue.”
“Water is a key active player here, rather than just a spectator. It’s integral to making these microplastics reactive. Water molecules become chemically active when they come into contact with plastic, which is hydrophobic. Contact between the water and plastic creates an environment that facilitates electron transfer, and the formation of reactive radicals. When plastics fragment into smaller particles, their increasing surface-to-volume ratio creates more of these chemically active areas,” Min said.
The Columbia team conducted the chemical and biological experiments, while collaborators in the lab of Professor Teresa Head-Gordon at the University of California, Berkeley, created theoretical calculations and dynamical simulations to identify how the microplastic interacted with water.
Head-Gordon’s lab led the theory that showed that hydroxide ions that are promoted to the immediate water layers near the hydrophobic microplastic interface can more easily lose their electron to create hydroxyl radicals.
“Our fundamental work that the most hydrophobic microplastics interacting with water can drive chemical reactivity could ultimately lead to new engineering solutions by redesigning the chemistry of these commodity plastics to reduce the toxic outcomes on health and the environment,” said Head-Gordon.
This work has significant implications for environmental and human health,” said co-author Beizhan Yan of Lamont-Doherty Earth Observatory. “In certain environments, such as discharges from plastic manufacturers and laundromats, microplastics can form and enter aquatic environments. This work also indicates that radicals can form on hydrophobic plastic surfaces in water, which could affect ecosystems and human health following exposure. In addition, various uses and processes in our daily lives can generate microplastics, including microwave heating, sunlight exposure of bottled water during storage. This foundational work provides a potential pathway for understanding their health impacts.”
Journal
Nature
Method of Research
Experimental study
Subject of Research
Not applicable
Article Title
Intrinsic physicochemical activity at microplastic–water interfaces
