Members Can Post Anonymously On This Site
Showcase your climate data visualisation talent
-
Similar Topics
-
By NASA
6 min read
Smarter Searching: NASA AI Makes Science Data Easier to Find
Image snapshot taken from NASA Worldview of NASA’s Global Precipitation Measurement (GPM) mission on March 15, 2025 showing heavy rain across the southeastern U.S. with an overlay of the GCMD Keyword Recommender for Earth Science, Atmosphere, Precipitation, Droplet Size. NASA Worldview Imagine shopping for a new pair of running shoes online. If each seller described them differently—one calling them “sneakers,” another “trainers,” and someone else “footwear for exercise”—you’d quickly feel lost in a sea of mismatched terminology. Fortunately, most online stores use standardized categories and filters, so you can click through a simple path: Women’s > Shoes > Running Shoes—and quickly find what you need.
Now, scale that problem to scientific research. Instead of sneakers, think “aerosol optical depth” or “sea surface temperature.” Instead of a handful of retailers, it is thousands of researchers, instruments, and data providers. Without a common language for describing data, finding relevant Earth science datasets would be like trying to locate a needle in a haystack, blindfolded.
That’s why NASA created the Global Change Master Directory (GCMD), a standardized vocabulary that helps scientists tag their datasets in a consistent and searchable way. But as science evolves, so does the challenge of keeping metadata organized and discoverable.
To meet that challenge, NASA’s Office of Data Science and Informatics (ODSI) at the agency’s Marshall Space Flight Center (MSFC) in Huntsville, Alabama, developed the GCMD Keyword Recommender (GKR): a smart tool designed to help data providers and curators assign the right keywords, automatically.
Smarter Tagging, Accelerated Discovery
The upgraded GKR model isn’t just a technical improvement; it’s a leap forward in how we organize and access scientific knowledge. By automatically recommending precise, standardized keywords, the model reduces the burden on human curators while ensuring metadata quality remains high. This makes it easier for researchers, students, and the public to find exactly the datasets they need.
It also sets the stage for broader applications. The techniques used in GKR, like applying focal loss to rare-label classification problems and adapting pre-trained transformers to specialized domains, can benefit fields well beyond Earth science.
Metadata Matchmaker
The newly upgraded GKR model tackles a massive challenge in information science known as extreme multi-label classification. That’s a mouthful, but the concept is straightforward: Instead of predicting just one label, the model must choose many, sometimes dozens, from a set of thousands. Each dataset may need to be tagged with multiple, nuanced descriptors pulled from a controlled vocabulary.
Think of it like trying to identify all the animals in a photograph. If there’s just a dog, it’s easy. But if there’s a dog, a bird, a raccoon hiding behind a bush, and a unicorn that only shows up in 0.1% of your training photos, the task becomes far more difficult. That’s what GKR is up against: tagging complex datasets with precision, even when examples of some keywords are scarce.
And the problem is only growing. The new version of GKR now considers more than 3,200 keywords, up from about 430 in its earlier iteration. That’s a sevenfold increase in vocabulary complexity, and a major leap in what the model needs to learn and predict.
To handle this scale, the GKR team didn’t just add more data; they built a more capable model from the ground up. At the heart of the upgrade is INDUS, an advanced language model trained on a staggering 66 billion words drawn from scientific literature across disciplines—Earth science, biological sciences, astronomy, and more.
NASA ODSI’s GCMD Keyword Recommender AI model automatically tags scientific datasets with the help of INDUS, a large language model trained on NASA scientific publications across the disciplines of astrophysics, biological and physical sciences, Earth science, heliophysics, and planetary science. NASA “We’re at the frontier of cutting-edge artificial intelligence and machine learning for science,” said Sajil Awale, a member of the NASA ODSI AI team at MSFC. “This problem domain is interesting, and challenging, because it’s an extreme classification problem where the model needs to differentiate even very similar keywords/tags based on small variations of context. It’s exciting to see how we have leveraged INDUS to build this GKR model because it is designed and trained for scientific domains. There are opportunities to improve INDUS for future uses.”
This means that the new GKR isn’t just guessing based on word similarities; it understands the context in which keywords appear. It’s the difference between a model knowing that “precipitation” might relate to weather versus recognizing when it means a climate variable in satellite data.
And while the older model was trained on only 2,000 metadata records, the new version had access to a much richer dataset of more than 43,000 records from NASA’s Common Metadata Repository. That increased exposure helps the model make more accurate predictions.
The Common Metadata Repository is the backend behind the following data search and discovery services:
Earthdata Search International Data Network Learning to Love Rare Words
One of the biggest hurdles in a task like this is class imbalance. Some keywords appear frequently; others might show up just a handful of times. Traditional machine learning approaches, like cross-entropy loss, which was used initially to train the model, tend to favor the easy, common labels, and neglect the rare ones.
To solve this, NASA’s team turned to focal loss, a strategy that reduces the model’s attention to obvious examples and shifts focus toward the harder, underrepresented cases.
The result? A model that performs better across the board, especially on the keywords that matter most to specialists searching for niche datasets.
From Metadata to Mission
Ultimately, science depends not only on collecting data, but on making that data usable and discoverable. The updated GKR tool is a quiet but critical part of that mission. By bringing powerful AI to the task of metadata tagging, it helps ensure that the flood of Earth observation data pouring in from satellites and instruments around the globe doesn’t get lost in translation.
In a world awash with data, tools like GKR help researchers find the signal in the noise and turn information into insight.
Beyond powering GKR, the INDUS large language model is also enabling innovation across other NASA SMD projects. For example, INDUS supports the Science Discovery Engine by helping automate metadata curation and improving the relevancy ranking of search results.The diverse applications reflect INDUS’s growing role as a foundational AI capability for SMD.
The INDUS large language model is funded by the Office of the Chief Science Data Officer within NASA’s Science Mission Directorate at NASA Headquarters in Washington. The Office of the Chief Science Data Officer advances scientific discovery through innovative applications and partnerships in data science, advanced analytics, and artificial intelligence.
Share
Details
Last Updated Jul 09, 2025 Related Terms
Science & Research Artificial Intelligence (AI) Explore More
2 min read Polar Tourists Give Positive Reviews to NASA Citizen Science in Antarctica
Article
6 hours ago
2 min read Hubble Observations Give “Missing” Globular Cluster Time to Shine
Article
6 days ago
5 min read How NASA’s SPHEREx Mission Will Share Its All-Sky Map With the World
Article
7 days ago
Keep Exploring Discover Related Topics
Missions
Humans in Space
Climate Change
Solar System
View the full article
-
By NASA
4 min read
Preparations for Next Moonwalk Simulations Underway (and Underwater)
A lot can change in a year for Earth’s forests and vegetation, as springtime and rainy seasons can bring new growth, while cooling temperatures and dry weather can bring a dieback of those green colors. And now, a novel type of NASA visualization illustrates those changes in a full complement of colors as seen from space.
Researchers have now gathered a complete year of PACE data to tell a story about the health of land vegetation by detecting slight variations in leaf colors. Previous missions allowed scientists to observe broad changes in chlorophyll, the pigment that gives plants their green color and also allows them to perform photosynthesis. But PACE now allows scientists to see three different pigments in vegetation: chlorophyll, anthocyanins, and carotenoids. The combination of these three pigments helps scientists pinpoint even more information about plant health. Credit: NASA’s Goddard Space Flight Center NASA’s Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) satellite is designed to view Earth’s microscopic ocean plants in a new lens, but researchers have proved its hyperspectral use over land, as well.
Previous missions measured broad changes in chlorophyll, the pigment that gives plants their green color and also allows them to perform photosynthesis. Now, for the first time, PACE measurements have allowed NASA scientists and visualizers to show a complete year of global vegetation data using three pigments: chlorophyll, anthocyanins, and carotenoids. That multicolor imagery tells a clearer story about the health of land vegetation by detecting the smallest of variations in leaf colors.
“Earth is amazing. It’s humbling, being able to see life pulsing in colors across the whole globe,” said Morgaine McKibben, PACE applications lead at NASA’s Goddard Space Flight Center in Greenbelt, Maryland. “It’s like the overview effect that astronauts describe when they look down at Earth, except we are looking through our technology and data.”
Anthocyanins, carotenoids, and chlorophyll data light up North America, highlighting vegetation and its health.Credit: NASA’s Scientific Visualization Studio Anthocyanins are the red pigments in leaves, while carotenoids are the yellow pigments – both of which we see when autumn changes the colors of trees. Plants use these pigments to protect themselves from fluctuations in the weather, adapting to the environment through chemical changes in their leaves. For example, leaves can turn more yellow when they have too much sunlight but not enough of the other necessities, like water and nutrients. If they didn’t adjust their color, it would damage the mechanisms they have to perform photosynthesis.
In the visualization, the data is highlighted in bright colors: magenta represents anthocyanins, green represents chlorophyll, and cyan represents carotenoids. The brighter the colors are, the more leaves there are in that area. The movement of these colors across the land areas show the seasonal changes over time.
In areas like the evergreen forests of the Pacific Northwest, plants undergo less seasonal change. The data highlights this, showing comparatively steadier colors as the year progresses.
The combination of these three pigments helps scientists pinpoint even more information about plant health.
“Shifts in these pigments, as detected by PACE, give novel information that may better describe vegetation growth, or when vegetation changes from flourishing to stressed,” said McKibben. “It’s just one of many ways the mission will drive increased understanding of our home planet and enable innovative, practical solutions that serve society.”
The Ocean Color Instrument on PACE collects hyperspectral data, which means it observes the planet in 100 different wavelengths of visible and near infrared light. It is the only instrument – in space or elsewhere – that provides hyperspectral coverage around the globe every one to two days. The PACE mission builds on the legacy of earlier missions, such as Landsat, which gathers higher resolution data but observes a fraction of those wavelengths.
In a paper recently published in Remote Sensing Letters, scientists introduced the mission’s first terrestrial data products.
“This PACE data provides a new view of Earth that will improve our understanding of ecosystem dynamics and function,” said Fred Huemmrich, research professor at the University of Maryland, Baltimore County, member of the PACE science and applications team, and first author of the paper. “With the PACE data, it’s like we’re looking at a whole new world of color. It allows us to describe pigment characteristics at the leaf level that we weren’t able to do before.”
As scientists continue to work with these new data, available on the PACE website, they’ll be able to incorporate it into future science applications, which may include forest monitoring or early detection of drought effects.
By Erica McNamee
NASA’s Goddard Space Flight Center, Greenbelt, Md.
Share
Details
Last Updated Jun 05, 2025 EditorKate D. RamsayerContactKate D. Ramsayerkate.d.ramsayer@nasa.gov Related Terms
Earth Goddard Space Flight Center PACE (Plankton, Aerosol, Cloud, Ocean Ecosystem) Explore More
4 min read Tundra Vegetation to Grow Taller, Greener Through 2100, NASA Study Finds
Article 10 months ago 8 min read NASA Researchers Study Coastal Wetlands, Champions of Carbon Capture
In the Florida Everglades, NASA’s BlueFlux Campaign investigates the relationship between tropical wetlands and greenhouse…
Article 3 months ago 5 min read NASA Takes to the Air to Study Wildflowers
Article 2 months ago View the full article
-
By NASA
5 Min Read 3 Black Holes Caught Eating Massive Stars in NASA Data
A disk of hot gas swirls around a black hole in this illustration. Some of the gas came from a star that was pulled apart by the black hole, forming the long stream of hot gas on the right, feeding into the disk. Credits:
NASA/JPL-Caltech Black holes are invisible to us unless they interact with something else. Some continuously eat gas and dust, and appear to glow brightly over time as matter falls in. But other black holes secretly lie in wait for years until a star comes close enough to snack on.
Scientists have recently identified three supermassive black holes at the centers of distant galaxies, each of which suddenly brightened when it destroyed a star and then stayed bright for several months. A new study using space and ground-based data from NASA, ESA (European Space Agency), and other institutions presents these rare occurrences as a new category of cosmic events called “extreme nuclear transients.”
Looking for more of these extreme nuclear transients could help unveil some of the most massive supermassive black holes in the universe that are usually quiet.
“These events are the only way we can have a spotlight that we can shine on otherwise inactive massive black holes,” said Jason Hinkle, graduate student at the University of Hawaii and lead author of a new study in the journal Science Advances describing this phenomenon.
The black holes in question seem to have eaten stars three to 10 times heavier than our Sun. Feasting on the stars resulted in some of the most energetic transient events ever recorded.
This illustration shows a glowing stream of material from a star as it is being devoured by a supermassive black hole. When a star passes within a certain distance of a black hole — close enough to be gravitationally disrupted — the stellar material gets stretched and compressed as it falls into the black hole. NASA/JPL-Caltech These events as unleash enormous amount of high-energy radiation on the central regions of their host galaxies. “That has implications for the environments in which these events are occurring,” Hinkle said. “If galaxies have these events, they’re important for the galaxies themselves.”
The stars’ destruction produces high-energy light that takes over 100 days to reach peak brightness, then more than 150 days to dim to half of its peak. The way the high-energy radiation affects the environment results in lower-energy emissions that telescopes can also detect.
One of these star-destroying events, nicknamed “Barbie” because of its catalog identifier ZTF20abrbeie, was discovered in 2020 by the Zwicky Transient Facility at Caltech’s Palomar Observatory in California, and documented in two 2023 studies. The other two black holes were detected by ESA’s Gaia mission in 2016 and 2018 and are studied in detail in the new paper.
NASA’s Neil Gehrels Swift Observatory was critical in confirming that these events must have been related to black holes, not stellar explosions or other phenomena. The way that the X-ray, ultraviolet, and optical light brightened and dimmed over time was like a fingerprint matching that of a black hole ripping a star apart.
Scientists also used data from NASA’s WISE spacecraft, which was operated from 2009 to 2011 and then was reactivated as NEOWISE and retired in 2024. Under the WISE mission the spacecraft mapped the sky at infrared wavelengths, finding many new distant objects and cosmic phenomena. In the new study, the spacecraft’s data helped researchers characterize dust in the environments of each black hole. Numerous ground-based observatories additionally contributed to this discovery, including the W. M. Keck Observatory telescopes through their NASA-funded archive and the NASA-supported Near-Earth Object surveys ATLAS, Pan-STARRS, and Catalina.
“What I think is so exciting about this work is that we’re pushing the upper bounds of what we understand to be the most energetic environments of the universe,” said Anna Payne, a staff scientist at the Space Telescope Science Institute and study co-author, who helped look for the chemical fingerprints of these events with the University of Hawaii 2.2-meter Telescope.
A Future Investigators in NASA Earth and Space Science and Technology (FINESST) grant from the agency helped enable Hinkle to search for these black hole events. “The FINESST grant gave Jason the freedom to track down and figure out what these events actually were,” said Ben Shappee, associate professor at the Institute for Astronomy at the University of Hawaii, a study coauthor and advisor to Hinkle.
Hinkle is set to follow up on these results as a postdoctoral fellow at the University of Illinois Urbana-Champaign through the NASA Hubble Fellowship Program. “One of the biggest questions in astronomy is how black holes grow throughout the universe,” Hinkle said.
The results complement recent observations from NASA’s James Webb Space Telescope showing how supermassive black holes feed and grow in the early universe. But since only 10% of early black holes are actively eating gas and dust, extreme nuclear transients — that is, catching a supermassive black hole in the act of eating a massive star — are a different way to find black holes in the early universe.
Events like these are so bright that they may be visible even in the distant, early universe. Swift showed that extreme nuclear transients emit most of their light in the ultraviolet. But as the universe expands, that light is stretched to longer wavelengths and shifts into the infrared — exactly the kind of light NASA’s upcoming Nancy Grace Roman Space Telescope was designed to detect.
With its powerful infrared sensitivity and wide field of view, Roman will be able to spot these rare explosions from more than 12 billion years ago, when the universe was just a tenth of its current age. Scheduled to launch by 2027, and potentially as early as fall 2026, Roman could uncover many more of these dramatic events and offer a new way to explore how stars, galaxies, and black holes formed and evolved over time.
“We can take these three objects as a blueprint to know what to look for in the future,” Payne said.
Explore More
5 min read NASA’s Webb Rounds Out Picture of Sombrero Galaxy’s Disk
Article
1 day ago
2 min read Hubble Filters a Barred Spiral
Article
1 day ago
5 min read Apocalypse When? Hubble Casts Doubt on Certainty of Galactic Collision
Article
2 days ago
View the full article
-
By European Space Agency
From its vantage point outside Earth’s atmosphere, more than 36 000 km above Earth’s surface, the Copernicus Sentinel-4 mission will detect major air pollutants over Europe in unprecedented detail. It will observe how they vary on an hourly basis – a real breakthrough for air quality forecasting.
View the full article
-
By European Space Agency
The deadline to apply for ESA’s Junior Professional Programme (JPP) is fast approaching, with applications closing on 5 June. The JPP is a unique opportunity for recent graduates and early-career professionals to gain hands-on experience at one of the world’s leading space organisations. Successful candidates will work on real ESA projects alongside experienced experts in fields ranging from engineering and science to business and administration. If you are ready to take the first step towards a future in space, don’t miss this opportunity. Submit your application before the 5 June deadline.
View the full article
-
-
Check out these Videos
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.