A NASA-funded air pollution monitoring network has provided one of the most detailed long-term views yet of the role of black carbon, or soot produced by fires, diesel vehicles, and other combustion sources, in Ethiopia’s capital, Addis Ababa. The detailed measurements show how pollution changes by time of day and season, including increases associated with rush-hour traffic and holiday celebrations. The findings are relevant to cities around the world, including in the United States.

In a new paper published in ES&T: Air, scientists analyzed data collected throughout Addis Ababa between 2022 and 2025 from 10 air-quality monitoring sites deployed by NASA’s Multi-Angle Imager for Aerosols (MAIA) project.

The research comes as Ethiopia is taking steps aimed at improving air quality. In 2024, the country became the first in the world to ban the import of internal combustion engine vehicles, while cities have been adding bike lanes and electric vehicle infrastructure. The MAIA project’s measurements provide researchers with a baseline for understanding how air quality changes over time as Addis Ababa continues to grow and evolve.

The study focuses on particulate matter that is 2.5 micrometers or less in diameter, also known as PM2.5. The 2025 State of Global Air Report, cited in the paper, estimates that exposure to PM2.5 is associated with approximately 4.9 million deaths globally each year. Among the many kinds of PM2.5, black carbon has been has been studied for its potential effects on human health.

The paper found that Addis Ababa’s three-year average PM2.5 concentration was 30 micrograms per cubic meter, which is more than three times the level of the U.S. Environmental Protection Agency’s health-based annual PM2.5 standard. The new paper cites data from MAIA’s ground sensors indicating that average black carbon levels in Addis Ababa were approximately four to nine times higher than those measured in the three U.S. metropolitan areas the mission is monitoring.

“To our knowledge, this is the first long-term, multisite study of continuous PM2.5 and black carbon measurements in Ethiopia,” said Sina Hasheminassab, a coauthor of the paper and MAIA’s deputy principal investigator at NASA’s Jet Propulsion Laboratory in Southern California. “Many rapidly growing cities have limited long-term monitoring, so these measurements provide an important baseline for understanding how pollution changes across space and time.”

The composition and sources of PM2.5 can differ substantially between cities, depending on their local geography, traffic, industries, and more. Desert cities, for example, may have more dust, while those near coal-fired power plants may have higher concentrations of sulfate. Long-term surface measurements remain limited in many parts of the world.

NASA is supporting MAIA’s air pollution research in a dozen metropolitan areas around the globe, including three in the U.S.: Los Angeles, Atlanta, and Boston. The mission consists of a ground-based network of sensors already in operation as well as a space observatory, which uses a JPL-built camera that will be launched by the Italian Space Agency (ASI) on an ASI satellite no earlier than late 2027.

The camera is designed to identify different types of PM2.5 aerosols based on how they reflect light, making it possible to map particle concentrations over each city that the mission studies. Mounted on a gimbal, the camera captures data from multiple angles using JPL-pioneered technologies that make particles stand out more prominently against the surface background to provide valuable information about their shape and size.

The MAIA mission is the first NASA project to include public health researchers among a space mission’s team. These researchers will use MAIA’s PM2.5 concentration maps alongside health data to study potential relationships between different particle types and health outcomes. By developing a better understanding of particulate matter pollution, researchers can potentially advance how air quality is studied and managed.

“This paper shows how valuable the air sensor data is on its own, but combining the sensor network and satellite observations will be a game-changer,” said, Kyan Shlipak, the paper’s lead author, who worked on the research while interning at JPL.

Tracking black carbon

The greater Addis Ababa urban area is home to nearly 6 million people, and according to United Nations projections, that figure is expected to surpass 10 million by 2050.

“It’s a cosmopolitan city with many international communities,” said Araya Asfaw of Addis Ababa University, a coauthor of the paper and the MAIA project’s lead Ethiopian collaborator. “Think of it as Africa’s version of Brussels, where the European Union is based.”

“Even at night, when traffic dies down, you see high emissions from the burning of charcoal and other fuels,” Asfaw said.

The MAIA sensor network detected increases in black carbon during two major holidays in Addis Ababa that involve bonfires and was able to distinguish between particles originating from the fires and those from fossil fuel combustion. The findings demonstrate how detailed measurements can help researchers identify different sources of particulate matter and better understand how air quality varies across a city and over time.

To learn more about MAIA, visit:

https://science.nasa.gov/mission/maia/

2026-056

The NASA Science Mission Directorate (SMD) Community of Practice for Education (SCoPE) – part of the NASA Science Activation (SciAct) Program portfolio – enables Earth and Space Science and Engineering Subject Matter Experts (SMEs) – especially NASA-funded SMEs – to efficiently and effectively share their science with support from SciAct education experts. 

In Summer 2026, NASA SCoPE partnered with Arizona State University’s Facility for Open Research in a Compressed Environment (FORCE) Summer School to help seven undergraduate student interns build the science communication skills needed to share their research with a variety of audiences. FORCE is a world-class laboratory that uses high-pressure experimental equipment to recreate the extreme conditions found deep within Earth and other planetary bodies, enabling researchers to better understand how planets form, evolve, and behave under immense pressures.

As part of the Summer School, SCoPE facilitated two hands-on workshops on June 25 and 26, followed by office hours the following week, to help interns translate their technical research into compelling stories for non-expert audiences. The training focused on identifying the central themes of their work, developing clear and engaging messages, planning effective visitor interactions, and thinking through the logistics of public engagement. Interns also received guidance on preparing both their research posters and individual outreach stations.

The training culminated in two complementary outreach experiences. The first was the FORCE Open House, which welcomed approximately 50 members of the general public for an inside look at the laboratory. Visitors toured the facility, met the research team, explored the specialized equipment used to simulate the interiors of Earth and other planets, and interacted with interns at themed outreach stations designed to explain the science behind the experiments in accessible, engaging ways.

At the second event, a poster session for ASU faculty, staff, and students, the interns presented their research, providing an opportunity to discuss their scientific findings with members of the university community and receive feedback on their presentations.

By integrating science communication training into the Summer School experience, NASA SCoPE helped equip emerging planetary scientists with practical skills for engaging both scientific peers and public audiences. The poster session and Open House demonstrated how thoughtful communication training can strengthen researchers’ confidence while building stronger connections between cutting-edge planetary science and the communities it serves.
NASA SCoPE is supported by NASA cooperative agreement award number 80NSSC21M0006 and helps enrich and enhance the impact of the NASA Science Activation Program portfolio, which connects learners with authentic NASA science experiences through partnerships with educators and community organizations.

Possono trovarsi davanti a due possibilità equivalenti, senza alcun motivo per preferirne una all’altra, eppure finiscono per scegliere la stessa. Non perché qualcuno dica loro quale sia la risposta giusta, ma perché guardano cosa hanno scelto gli altri e si adeguano alla maggioranza. È il comportamento osservato in un nuovo studio sulle Intelligenze Artificiali, che mostra come gruppi di IA possano sviluppare spontaneamente una forma di coordinamento collettivo. La ricerca, pubblicata sulla rivista Science Advances, è stata guidata dall’Università di Costanza, in Germania, e ha coinvolto anche l’Istituto dei Sistemi Complessi del Consiglio Nazionale delle Ricerche e il Centro Ricerche Enrico Fermi di Roma. A coordinare il lavoro è stato l’italiano Giordano De Marzo, ricercatore dell’ateneo tedesco.

Gli esperimenti hanno messo alla prova gruppi di Large Language Model, i grandi modelli linguistici alla base dei più noti sistemi di IA. ChatGPT è uno degli esempi più conosciuti. Ai modelli venivano proposte due opzioni tra loro equivalenti, in una situazione quindi in cui non esisteva una ragione oggettiva per scegliere l’una o l’altra. La variabile decisiva era un’altra: ogni modello poteva conoscere le scelte effettuate dagli altri componenti del gruppo. E, osservando quelle preferenze, le IA tendevano progressivamente a convergere verso l’opzione che aveva ricevuto più voti. In altre parole, la maggioranza diventava una sorta di punto di riferimento. I modelli sceglievano l’opzione più “votata”, sviluppando così un comportamento di coordinamento collettivo che ricorda meccanismi osservabili anche in natura, dagli stormi di uccelli ai banchi di pesci fino ai comportamenti degli esseri umani.

L’aspetto più sorprendente è che il fenomeno non si limita a piccoli gruppi. Gli esperimenti hanno mostrato forme di coordinamento anche in gruppi estremamente numerosi, composti da circa mille IA. Tra i modelli coinvolti c’erano anche sistemi avanzati come Claude 3.5 Sonnet di Anthropic e GPT-4 Turbo di OpenAI. Il meccanismo, spiegano i ricercatori, non richiede necessariamente un coordinatore umano che stabilisca le regole del comportamento collettivo. Il consenso può emergere dal semplice confronto tra le decisioni dei singoli modelli.

È proprio questo elemento a rendere il risultato interessante, ma anche delicato. Se la capacità di coordinarsi può rendere più efficienti sistemi composti da molte IA che devono collaborare, allo stesso tempo può produrre dinamiche difficili da prevedere o controllare. “Il coordinamento spontaneo potrebbe rivelarsi vantaggioso”, osservano gli autori dello studio, “ma potrebbe anche porre sfide legate alla sicurezza“. La ricerca apre quindi una questione che va oltre la semplice capacità di un singolo modello di fornire una risposta. Se sempre più sistemi di IA saranno chiamati a lavorare insieme, infatti, sarà importante capire non solo come ragiona ciascun modello, ma anche cosa accade quando molti modelli osservano le decisioni degli altri e iniziano ad adattarsi reciprocamente. Il risultato dello studio suggerisce che, in determinate condizioni, una forma di consenso può emergere senza che nessuno lo abbia imposto. Le IA, semplicemente, guardano cosa fa la maggioranza. E la seguono.

Lo studio

L'articolo Le IA si mettono d’accordo da sole: seguono la maggioranza anche in gruppi di mille. Lo studio su Science proviene da Il Fatto Quotidiano.

Phys.org reports: Future fusion power plants aim to recreate the heart of a star here on Earth to power our future energy needs. While the core fusion plasma will burn at hundreds of millions of degrees, the surrounding structural components must handle sudden, punishing heat loads that rival the extreme temperatures faced by spacecraft upon reentry into Earth's atmosphere. Copper and its alloys are primary candidates for handling these intense heat fluctuations, making it vital to understand exactly how the metal behaves when pushed to its melting point. Now, researchers at the Department of Energy's SLAC National Accelerator Laboratory and collaborators have captured a detailed, step-by-step look at copper atoms as they underwent extreme heating. Published in Nature Communications, the results revealed a key parameter that allowed copper's crystal lattice to melt steadily rather than collapse instantaneously, as earlier simulations predicted. "These results greatly improve the simulations we use to predict which materials have the best shot at surviving the extreme conditions of future fusion reaction chambers," said Mianzhen Mo, a SLAC staff scientist who led the research. "They also demonstrate the incredible, atomic-scale resolution imaging we can achieve at SLAC's electron camera...." Researchers use computer simulations, aided by AI and machine learning, to sift through innumerable combinations of elements and identify promising candidate materials for real-world testing. "Whether the copper melts slowly or suddenly collapses, by the time the researchers look, the sample resembles nothing more than a metallic brown puddle," the article points out. But SLAC's powerful electron camera captures atomic and molecular movements down to the femtosecond — a millionth of a billionth of a second — and revealed that at around 1,424 degreesC (2,595 degreesF) there was still gradual melting as the temperature rose beyond the superheating limit, with real-world conditions showing the atoms shifted and retained some order. "It's a straightforward solution," said Mianzhen Mo, a SLAC staff scientist who led the research. "But molecular dynamics simulations had been overlooking it for years. When you have complex simulations attempting to capture every aspect of reality, down to individual atoms, it takes real-world data to show you what's missing from the calculations."

Read more of this story at Slashdot.

Certain kinds of clouds are misbehaving – appearing more often and lower in the sky than they used to. To help identify the factors influencing these changes (e.g. shifts in Earth’s long-term weather patterns), scientists have asked people around the world with cameras to submit fresh images of these clouds as a part of the NASA-supported Space Cloud Watch project. Now, one volunteer has developed a new tool to help other Space Cloud Watch volunteers work more efficiently. 

The misbehaving clouds are “noctilucent”  or “night-shining” clouds (NLCs). These clouds scatter light from the Sun long after sunset and long before sunrise, giving them a silvery glow. But despite this glow, it can be hard to differentiate NLCs from lower-altitude look-alikes. That confusion has meant extra work for project leaders.

Volunteer Namai Chandra shared, “I noticed that NLC images were being manually verified by the project leaders. It felt like a task well-suited for a human-in-the-loop machine learning pipeline, one that could handle the repetitive screening automatically, while keeping human judgment central for the images that matter most.” In other words, Namai found a way to help observers verify when they are indeed seeing NLCs and when they’re not. 

Namai reached out to the Space Cloud Watch scientists Drs. Chihoko Cullens and Brentha Thurairajah, who were delighted with his idea. Namai soon developed a machine learning pipeline, training it on a variety of cloud images, including both the NLCs and the lower altitude look-alikes that are often submitted to Space Cloud Watch. The pipeline combines image pre-screening, cloud classification, and confidence-based review routing. After several rounds of development, testing, and refinement, he released his NLC identification tool to the project. This tool is now being used by cloud contributors who are unsure whether they have observed NLCs, as well as project scientists that want to flag images for review. 

Grab a camera and join the Space Cloud Watch project today! If you’ve hesitated to contribute to Space Cloud Watch because you were not certain if what you were seeing was a noctilucent cloud, you now have a way to check before you share – thanks to Namai.