Jump to content

Expanded AI Model with Global Data Enhances Earth Science Applications 


Recommended Posts

  • Publishers
Posted

4 min read

Expanded AI Model with Global Data Enhances Earth Science Applications 

A false-color image of the East Peak fire burning in southern Colorado near Trinidad, captured by the Operational Land Imager on the Landsat 8 satellite. Dark green forests and light green grasslands cover most of the image, but a red patch in the middle represents a burn scar, and some orange spots around it represent actively burning areas.
On June 22, 2013, the Operational Land Imager (OLI) on Landsat 8 captured this false-color image of the East Peak fire burning in southern Colorado near Trinidad. Burned areas appear dark red, while actively burning areas look orange. Dark green areas are forests; light green areas are grasslands. Data from Landsat 8 were used to train the Prithvi artificial intelligence model, which can help detect burn scars.
NASA Earth Observatory

NASA, IBM, and Forschungszentrum Jülich have released an expanded version of the open-source Prithvi Geospatial artificial intelligence (AI) foundation model to support a broader range of geographical applications. Now, with the inclusion of global data, the foundation model can support tracking changes in land use, monitoring disasters, and predicting crop yields worldwide. 

The Prithvi Geospatial foundation model, first released in August 2023 by NASA and IBM, is pre-trained on NASA’s Harmonized Landsat and Sentinel-2 (HLS) dataset and learns by filling in masked information. The model is available on Hugging Face, a data science platform where machine learning developers openly build, train, deploy, and share models. Because NASA releases data, products, and research in the open, businesses and commercial entities can take these models and transform them into marketable products and services that generate economic value. 

“We’re excited about the downstream applications that are made possible with the addition of global HLS data to the Prithvi Geospatial foundation model. We’ve embedded NASA’s scientific expertise directly into these foundation models, enabling them to quickly translate petabytes of data into actionable insights,” said Kevin Murphy, NASA chief science data officer. “It’s like having a powerful assistant that leverages NASA’s knowledge to help make faster, more informed decisions, leading to economic and societal benefits.”

AI foundation models are pre-trained on large datasets with self-supervised learning techniques, providing flexible base models that can be fine-tuned for domain-specific downstream tasks.

Three images show the process of crop classification with NASA and IBM’s open-source Prithvi Geospatial artificial intelligence model. The first shows a true color composite image of a cropland area. The second shows a Ground Truth Mask of the types of land cover in the image — natural vegetation is colored red, forest is orange, corn is yellow, soybeans are light green, wetlands are mid-green, developed or barren land is darker green, open water is light blue, winter wheat is mid-blue, alfalfa is dark blue, fallow/idle cropland is dark purple, cotton is pink, and sorghum is dark pink. The third image shows a Predicted Mask that closely matches the Ground Truth Mask.
Crop classification prediction generated by NASA and IBM’s open-source Prithvi Geospatial artificial intelligence model.

Focusing on diverse land use and ecosystems, researchers selected HLS satellite images that represented various landscapes while avoiding lower-quality data caused by clouds or gaps. Urban areas were emphasized to ensure better coverage, and strict quality controls were applied to create a large, well-balanced dataset. The final dataset is significantly larger than previous versions, offering improved global representation and reliability for environmental analysis. These methods created a robust and representative dataset, ideal for reliable model training and analysis. 

The Prithvi Geospatial foundation model has already proven valuable in several applications, including post-disaster flood mapping and detecting burn scars caused by fires.

One application, the Multi-Temporal Cloud Gap Imputation, leverages the foundation model to reconstruct the gaps in satellite imagery caused by cloud cover, enabling a clearer view of Earth’s surface over time. This approach supports a variety of applications, including environmental monitoring and agricultural planning.  

Another application, Multi-Temporal Crop Segmentation, uses satellite imagery to classify and map different crop types and land cover across the United States. By analyzing time-sequenced data and layering U.S. Department of Agriculture’s Crop Data, Prithvi Geospatial can accurately identify crop patterns, which in turn could improve agricultural monitoring and resource management on a large scale. 

The flood mapping dataset can classify flood water and permanent water across diverse biomes and ecosystems, supporting flood management by training models to detect surface water. 

Wildfire scar mapping combines satellite imagery with wildfire data to capture detailed views of wildfire scars shortly after fires occurred. This approach provides valuable data for training models to map fire-affected areas, aiding in wildfire management and recovery efforts.

Three images show the process of burn scar mapping with NASA and IBM’s open-source Prithvi Geospatial artificial intelligence model.The first shows a true color composite satellite image, which contains shades of green and purple. The second shows a Ground Truth Mask, which shows the true extent of a burn scar on the land by blacking out the area around the burn scar. The third shows a Predicted Mask, which almost exactly matches the Ground Truth Mask.
Burn scar mapping generated by NASA and IBM’s open-source Prithvi Geospatial artificial intelligence model.

This model has also been tested with additional downstream applications including estimation of gross primary productivity, above ground biomass estimation, landslide detection, and burn intensity estimations. 

“The updates to this Prithvi Geospatial model have been driven by valuable feedback from users of the initial version,” said Rahul Ramachandran, AI foundation model for science lead and senior data science strategist at NASA’s Marshall Space Flight Center in Huntsville, Alabama. “This enhanced model has also undergone rigorous testing across a broader range of downstream use cases, ensuring improved versatility and performance, resulting in a version of the model that will empower diverse environmental monitoring applications, delivering significant societal benefits.”

The Prithvi Geospatial Foundation Model was developed as part of an initiative of NASA’s Office of the Chief Science Data Officer to unlock the value of NASA’s vast collection of science data using AI. NASA’s Interagency Implementation and Advanced Concepts Team (IMPACT), based at Marshall, IBM Research, and the Jülich Supercomputing Centre, Forschungszentrum, Jülich, designed the foundation model on the supercomputer Jülich Wizard for European Leadership Science (JUWELS), operated by Jülich Supercomputing Centre. This collaboration was facilitated by IEEE Geoscience and Remote Sensing Society.  

For more information about NASA’s strategy of developing foundation models for science, visit https://science.nasa.gov/artificial-intelligence-science.

Share

Details

Last Updated
Dec 04, 2024

Related Terms

View the full article

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.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Similar Topics

    • By Space Force
      The two-week event, held at Vandenberg Space Force Base, focuses on strengthening international partnerships, enhancing operational collaboration and promoting responsible behavior in space.

      View the full article
    • By NASA
      4 Min Read NASA Expands SPHEREx Science Return Through Commercial Partnership
      A sectional rendering of NASA's SPHEREx (Spectro-Photometer for the History of the Universe, Epoch of Reionization and Ices Explorer). Credits: NASA NASA is partnering with commercial industry to expand our knowledge of Earth, our solar system, and beyond. Recently, NASA collaborated with Kongsberg Satellite Services (KSAT) to support data transfer for the agency’s SPHEREx (Spectro-Photometer for the History of the Universe, Epoch of Reionization and Ices Explorer) mission to explore the origins of the universe. 
      “Not only is NASA moving toward commercialization, the agency is making technological advancements to existing systems and saving millions of dollars in the process — all while expanding human knowledge through science and exploration missions,” said Kevin Coggins, associate administrator for NASA’s SCaN (Space Communications and Navigation) program.
      To receive data from missions in space, NASA relies on the Near Space Network and Deep Space Network, a collection of antennas around the globe.
      In preparation for the recently-launched SPHEREx observatory, NASA needed to upgrade an antenna on the world’s most remote continent: Antarctica.
      Transmitted via NASA’s Near Space Network, this video shows SPHEREx scanning a region of the Large Magellanic Cloud. The shifting colors represent different infrared wavelengths detected by the telescope’s two arrays. Credit: NASA/JPL-Caltech NASA’s SCaN program took a novel approach by leveraging its established commercial partnership with KSAT. While upgraded KSAT antennas were added to the Near Space Network in 2023, SPHEREx required an additional Antarctic antenna that could link to online data storage.
      To support SPHEREx’s polar orbit, KSAT upgraded its Troll, Antarctica antenna and incorporated their own cloud storage system. NASA then connected KSAT’s cloud to the NASA cloud, DAPHNE+ (Data Acquisition Process and Handling Environment).
      As the Near Space Network’s operational cloud services system, DAPHNE+ enables science missions to transmit their data to the network for virtual file storage, processing, and management. 
      “By connecting the Troll antenna to DAPHNE+, we eliminated the need for large, undersea fiberoptic cables by virtually connecting private and government-owned cloud systems, reducing the project’s cost and complexity,” said Matt Vincent, the SPHEREx mission manager for the Near Space Network at NASA’s Goddard Space Flight Center in Greenbelt, Maryland.
      Each day, SPHEREx downlinks a portion of its 20 gigabits of science data through the Troll antenna, which transfers the files across KSAT’s network of relay satellites to the DAPHNE+ cloud. The cloud system combines and centralizes the data from each antenna, allowing access to all of SPHEREx’s health and science data in one convenient place. 
      The SPHEREx mission data is transmitted from space to the Troll Satellite Station, relayed through a network of satellites, and stored in the Near Space Network’s cloud system for easily-accessible analysis by scientists around the world.NASA/Dave Ryan With coverage throughout its orbit, SPHEREx transmits its 3D maps of the celestial sky, offering new insight into what happened a fraction of a second after the big bang. 
      “Missions like SPHEREx use the Near Space Network’s combination of commercial and government antennas,” explained Michael Skube, DAPHNE+ manager at NASA Goddard. “And that is the benefit of DAPHNE+ — it enables the network to pull different sources of information into one central location. The DAPHNE+ system treats government and commercial antennas as part of the same network.” 
      The partnership is mutually beneficial. NASA’s Near Space Network maintains a data connection with SPHEREx as it traverses both poles and KSAT benefits from its antennas’ integration into a robust global network – no new cables required. 
       “We were able to find a networking solution with KSAT that did not require us to put additional hardware in Antarctica,” said Vincent. “Now we are operating with the highest data rate we have ever downlinked from that location.” 
      The upgraded ground station antenna at Troll Satellite Station supports cloud-based space communications, enabling NASA’s Near Space Network to support scientific missions via a wireless cloud network.Kongsberg Satellite Services For NASA, its commercial partners, and other global space agencies, this expansion means more reliable space communications with fewer expenses. 
      Troll’s successful integration into the Near Space Network is a case study for future private and government partnerships. As SPHEREx measures the collective glow of over 450 million galaxies as far as 10 billion light-years away, SCaN continues to innovate how its discoveries safely return to Earth. 
      The SPHEREx mission is managed by NASA’s Jet Propulsion Laboratory in Southern California for the agency’s Astrophysics Division within the Science Mission Directorate at NASA Headquarters. Data will be processed and archived at IPAC at Caltech. The SPHEREx dataset will be publicly available at the NASA-IPAC Infrared Science Archive. Funding and oversight for DAPHNE+ and the Near Space Network come from the SCaN program office at NASA Headquarters and operate out of NASA’s Goddard Space Flight Center. The Troll Satellite Station is owned and operated by Kongsberg Satellite Services and located in Queen Maud Land, Antarctica. 
      About the Author
      Korine Powers
      Lead Writer and Communications StrategistKorine Powers, Ph.D. is a writer for NASA's Space Communications and Navigation (SCaN) program office and covers emerging technologies, commercialization efforts, exploration activities, and more.
      Share
      Details
      Last Updated May 06, 2025 Related Terms
      Communicating and Navigating with Missions Commercial Space Space Communications & Navigation Program SPHEREx (Spectro-Photometer for the History of the Universe and Ices Explorer) View the full article
    • By NASA
      6 min read
      NASA Data Helps Map Tiny Plankton That Feed Giant Right Whales
      This North Atlantic right whale, named “Bowtie,” was spotted feeding in southern Maine waters in January 2025. A new technique aims to use NASA satellite data to see the plankton these whales depend on from space. Credit: New England Aquarium, taken under NMFS permit # 25739 In the waters off New England, one of Earth’s rarest mammals swims slowly, mouth agape. The North Atlantic right whale filters clouds of tiny reddish zooplankton — called Calanus finmarchicus — from the sea. These zooplankton, no bigger than grains of rice, are the whale’s lifeline. Only about 370 of these massive creatures remain.
      For decades, tracking the tiny plankton meant sending research vessels out in the ocean, towing nets and counting samples by hand. Now, scientists are looking from above instead.
      Using NASA satellite data, researchers found a way to detect Calanus swarms at the ocean surface in the Gulf of Maine, picking up on the animals’ natural red pigment. This early-stage approach, described in a new study, may help researchers better estimate where the copepods gather, and where whales might follow.
      Tracking the zooplankton from space could aid both the whales and maritime industries. By predicting where these mammals are likely to feed, researchers and marine resource managers hope to reduce deadly vessel strikes and fishing gear entanglements — two major threats to the species. Knowing the feeding patterns could also help shipping and fishing industries operate more efficiently.
      Calanus finmarchicus, a tiny zooplankton powering North Atlantic food webs, fuels right whale populations with its energy-rich lipid reserves. Credit: Cameron Thompson “NASA invests in this kind of research because it connects space-based observation with real-world challenges,” said Cynthia Hall, a support scientist at NASA headquarters in Washington. She works with the Early Career Research Program, which partly funded the work. “It’s yet another a way to put NASA satellite data to work for science, communities, and ecosystems.”
      Revealing the Ocean’s Hidden Patterns
      The new approach uses data from the Moderate Resolution Imaging Spectroradiometer (MODIS) aboard NASA’s Aqua satellite. The MODIS instrument doesn’t directly see the copepods themselves. Instead, it reads how the spectrum of sunlight reflected from the ocean surface changes in response to what’s in the water.
      When large numbers of the zooplankton rise to the surface, their reddish pigment — astaxanthin, the same compound that gives salmon its pink color — subtly alters how photons, or particles of light, from the sun are absorbed or scattered in the water. The fate of these photons in the ocean depends on the mix of living and non-living matter in seawater, creating a slight shift in color that MODIS can detect.
      “We didn’t know to look for Calanus before in this way,” said Catherine Mitchell, a satellite oceanographer at Bigelow Laboratory for Ocean Sciences in East Boothbay, Maine. “Remote sensing has typically focused on smaller things like phytoplankton. But recent research suggested that larger, millimeter-sized organisms like zooplankton can also influence ocean color.”
      A few years ago, researchers piloted a satellite method for detecting copepods in Norwegian waters. Now, some of those same scientists — along with Mitchell’s team — have refined the approach and applied it to the Gulf of Maine, a crucial feeding ground for right whales during their northern migration. By combining satellite data, a model, and field measurements, they produced enhanced images that revealed Calanus swarms at the sea surface, and were able to estimate numbers of the tiny animals.
      “We know the right whales are using habitats we don’t fully understand,” said Rebekah Shunmugapandi, also a satellite oceanographer at Bigelow and the study’s lead author. “This satellite-based Calanus information could eventually help identify unknown feeding grounds or better anticipate where whales might travel.”
      Tracking Elusive Giants
      Despite decades of study, North Atlantic right whales remain remarkably enigmatic to scientists. Once fairly predictable in their movements along the Eastern Seaboard of North America, these massive mammals began abandoning some traditional feeding grounds in 2010-2011. Their sudden shift to unexpected areas like the Gulf of Saint Lawrence caught people off guard, with deadly consequences.
      “We’ve had whales getting hit by ships and whales getting stuck in fishing gear,” said Laura Ganley, a research scientist in the Anderson Cabot Center for Ocean Life at the New England Aquarium in Boston, which conducts aerial and boat surveys of the whales.  
      In 2017, the National Oceanic and Atmospheric Administration designated the situation as an “unusual mortality event” in an effort to address the whales’ decline. Since then, 80 North Atlantic right whales have been killed or sustained serious injuries, according to NOAA.
      NASA satellite imagery from June 2009 was used to test a new method for detecting the copepod Calanus finmarchicus in the Gulf of Maine and estimating their numbers from space. Credit: NASA Earth Observatory image by Wanmei Liang, using data from Shunmugapandi, R., et al. (2025) In the Gulf of Maine, there’s less shipping activity, but there can be a complex patchwork of lobster fishing gear, said Sarah Leiter, a scientist with the Maine Department of Marine Resources. “Each fisherman has 800 traps or so,” Leiter explained. “If a larger number of whales shows up suddenly, like they just did in January 2025, it is challenging. Fishermen need time and good weather to adjust that gear.”
      What excites Leiter the most about the satellite data is the potential to use it in a forecasting tool to help predict where the whales could go. “That would be incredibly useful in giving us that crucial lead time,” she said.
      PACE: The Next Generation of Ocean Observer
      For now, the Calanus-tracking method has limitations. Because MODIS detects the copepods’ red pigment, not the animals themselves, that means other small, reddish organisms can be mistaken for the zooplankton. And cloud cover, rough seas, or deeper swarms all limit what satellites can spot.
      MODIS is also nearing the end of its operational life. But NASA’s next-generation PACE (Plankton, Aerosol, Cloud, ocean Ecosystem) satellite — launched in 2024 — is poised to make dramatic improvements in the detection of zooplankton and phytoplankton.
      NASA’s Ocean Color Instrument on the PACE satellite captured these swirling green phytoplankton blooms in the Gulf of Maine in April 2024. Such blooms fuel zooplankton like Calanus finmarchicus. Credit: NASA “The PACE satellite will definitely be able to do this, and maybe even something better,” said Bridget Seegers, an oceanographer and mission scientist with the PACE team at NASA’s Goddard Space Flight Center in Greenbelt, Maryland.
      The PACE mission includes the Ocean Color Instrument, which detects more than 280 wavelengths of light. That’s a big jump from the 10 wavelengths seen by MODIS. More wavelengths mean finer detail and better insights into ocean color and the type of plankton that the satellite can spot.
      Local knowledge of seasonal plankton patterns will still be essential to interpret the data correctly. But the goal isn’t perfect detection, the scientists say, but rather to provide another tool to inform decision-making, especially when time or resources are limited.
      By Emily DeMarco
      NASA Headquarters
      Share








      Details
      Last Updated May 05, 2025 Editor Emily DeMarco Related Terms
      Earth Moderate Resolution Imaging Spectroradiometer (MODIS) Oceans PACE (Plankton, Aerosol, Cloud, Ocean Ecosystem) Explore More
      3 min read NASA Tracks Snowmelt to Improve Water Management
      As part of a science mission tracking one of Earth’s most precious resources – water…


      Article


      2 weeks ago
      5 min read NASA Airborne Sensor’s Wildfire Data Helps Firefighters Take Action
      Data from the AVIRIS-3 sensor was recently used to create detailed fire maps in minutes,…


      Article


      2 weeks ago
      3 min read Celebrating Earth as Only NASA Can
      Lee esta historia en español aquí. From the iconic image of Earthrise taken by Apollo 8…


      Article


      2 weeks ago
      View the full article
    • By NASA
      NASA/Jonny Kim The Soyuz MS-26 spacecraft is pictured backing away from the International Space Station shortly after undocking on April 19, 2025. Three hours later, the spacecraft landed in Kazakhstan, returning astronaut Don Pettit and cosmonauts Alexey Ovchinin and Ivan Vagner to Earth.
      While aboard the International Space Station, Pettit conducted hundreds of hours of scientific investigations, including research to enhance on-orbit metal 3D printing capabilities, advance water sanitization technologies, explore plant growth under varying water conditions, and investigate fire behavior in microgravity, all contributing to future space missions.
      Image credit: NASA/Jonny Kim
      View the full article
    • By European Space Agency
      Image: The Ocean and Land Colour Instrument on Copernicus Sentinel-3 captured this image of Earth’s biggest iceberg, A23a, on 5 April 2025. View the full article
  • Check out these Videos

×
×
  • Create New...