Members Can Post Anonymously On This Site
Weather and Climate Artificial Intelligence (AI) Foundation Model Applications Presented at IBM Think in Boston
-
Similar Topics
-
By NASA
Download PDF: Contact Dynamics Predictions Utilizing theNESC Parameterless Contact Model
Modeling the capture of the Mars Sample Return (MSR) Orbiting Sample (OS) involves understanding complex dynamic behavior, which includes the OS making contact against the interior of the capture enclosure. The MSR Program required numerical verification of the contact dynamics’ predictions produced using their commercial software tools. This commercial software used “free” parameters to set up the contact modeling. Free parameters (also known as free variables) are not based on contact physics. The commercial contact model used by MSR
required seven free parameters including a Hertzian contact stiffness, surface penetration, stiffening exponent, penetration velocity, contact damping, maximum penetration depth for the contact damping value, and a smoothing function. An example of a parameter that is not free is coefficient of friction, which is a physics-based parameter. Consider the free parameter, contact stiffness. Contact stiffness is already present in the finite element model’s (FEM) stiffness matrix where the bodies come into contact, and surface penetration is disallowed in a physically realizable contact model, as FEM meshes should not penetrate one another during contact (i.e., the zero-contact limit penetration constraint condition).
As such, with each set of selected free parameters generating a different contact force signature, additional numerical verification is required to guide setting these parameters. Contact modeling is nonlinear. This means that the stiffness matrices of contacting bodies are continuously updated as the bodies come into contact, potentially recontact (due to vibrations), and disengage. The modal properties of contacting bodies continuously change with state transitions (e.g., stick-to-slip). Some contact models have been proposed and incorporated in commercial finite element analysis solvers, and most involve static loading. A relatively smaller number involve dynamics, which has historically proven challenging.
In 2005, NASA conducted a study testing several commercial contact solvers in predicting contact forces in transient dynamic environments. This was necessitated by the Space Shuttle Program (SSP)—after the February 2003 Columbia accident— deciding to include contact dynamics in the Space Shuttle transient coupled loads analysis (CLA) to capture the impact of contact nonlinearities. This rendered the entire CLA nonlinear. The study found major difficulties executing nonlinear CLAs in commercial software. A nonlinear solver developed by the NESC and Applied Structural Dynamics (ASD) that was able to produce physically realizable results was numerically verified by NASA and later experimentally validated as well. This nonlinear solver was subsequently utilized to execute all NASA SSP CLAs (i.e., crewed space flights) from 2005 to the final flight in 2011, as well as currently supporting the SLS Program.
The objective of the MSR contact verification work was to provide data that could be used by the MSR team to help define the free parameters listed above for the commercial tool contact model. The NESC/ASD solver was used to model contact between simple cantilever and free beams, deriving contact forces and relative displacements. These resulting data can be used to determine parameter values for more complex structures. Two of the modeled configurations, one for axial contact (Figure 1) and the other for stick/friction (Figure 2), and sample results from the NESC nonlinear dynamic analyses are presented in Figures 1 and 2.
For information, contact:
Dr. Dexter Johnson dexter.johnson@nasa.gov
Dr. Arya Majed arya.majed@nasa.gov
View the full article
-
By NASA
At Goddard Space Flight Center, the GSFC Data Science Group has completed the testing for their SatVision Top-of-Atmosphere (TOA) Foundation Model, a geospatial foundation model for coarse-resolution all-sky remote sensing imagery. The team, comprised of Mark Carroll, Caleb Spradlin, Jordan Caraballo-Vega, Jian Li, Jie Gong, and Paul Montesano, has now released their model for wide application in science investigations.
Foundation models can transform the landscape of remote sensing (RS) data analysis by enabling the pre-training of large computer-vision models on vast amounts of remote sensing data. These models can be fine-tuned with small amounts of labeled training and applied to various mapping and monitoring applications. Because most existing foundation models are trained solely on cloud-free satellite imagery, they are limited to applications of land surface or require atmospheric corrections. SatVision-TOA is trained on all-sky conditions which enables applications involving atmospheric variables (e.g., cloud or aerosol).
SatVision TOA is a 3 billion parameter model trained on 100 million images from Moderate Resolution Imaging Spectroradiometer (MODIS). This is, to our knowledge, the largest foundation model trained solely on satellite remote sensing imagery. By including “all-sky” conditions during pre-training, the team incorporated a range of cloud conditions often excluded in traditional modeling. This enables 3D cloud reconstruction and cloud modeling in support of Earth and climate science, offering significant enhancement for large-scale earth observation workflows.
With an adaptable and scalable model design, SatVision-TOA can unify diverse Earth observation datasets and reduce dependency on task-specific models. SatVision-TOA leverages one of the largest public datasets to capture global contexts and robust features. The model could have broad applications for investigating spectrometer data, including MODIS, VIIRS, and GOES-ABI. The team believes this will enable transformative advancements in atmospheric science, cloud structure analysis, and Earth system modeling.
The model architecture and model weights are available on GitHub and Hugging Face, respectively. For more information, including a detailed user guide, see the associated white paper: SatVision-TOA: A Geospatial Foundation Model for Coarse-Resolution All-Sky Remote Sensing Imagery.
Examples of image reconstruction by SatVision-TOA. Left: MOD021KM v6.1 cropped image chip using MODIS bands [1, 3, 2]. Middle: The same images with randomly applied 8×8 mask patches, masking 60% of the original image. Right: The reconstructed images produced by the model, along with their respective Structural Similarity Index Measure (SSIM) scores. These examples illustrate the model’s ability to preserve structural detail and reconstruct heterogeneous features, such as cloud textures and land-cover transitions, with high fidelity.NASAView the full article
-
By NASA
Credit: NASA NASA, on behalf of the National Oceanic and Atmospheric Administration (NOAA), has selected Southwest Research Institute of San Antonio to build the Next-Generation Space Weather Magnetometer for the Lagrange 1 Series project as a part of NOAA’s Space Weather Next program.
This cost-plus-fixed-fee contract is valued at approximately $26.1 million and includes the development of two magnetometer instruments. The anticipated period of performance is from December 2024 through January 2034. The work will take place at the awardee’s facility in San Antonio, NASA’s Goddard Space Flight Center in Maryland, and Kennedy Space Center in Florida.
The contract scope includes design, analysis, development, fabrication, integration, test, verification, and evaluation of the magnetometer instruments; launch support; supply and maintenance of ground support equipment; and support of post-launch mission operations at the NOAA Satellite Operations Facility.
These instruments will measure the interplanetary magnetic field carried by the solar wind. The instruments provide critical data to NOAA’s Space Weather Prediction Center, which issues forecasts, warnings and alerts that help mitigate space weather impacts, including electric power outages and interruption to communications and navigation systems.
NASA and NOAA oversee the development, launch, testing, and operation of all the satellites in the Lagrange 1 Series project. NOAA is the program owner providing the requirements and funding along with managing the program, operations, data products, and dissemination to users. NASA and its commercial partners develop and build the instruments, spacecraft, and provide launch services on behalf of NOAA.
For information about NASA and agency programs, visit:
https://www.nasa.gov
-end-
Tiernan Doyle
Headquarters, Washington
202-358-1600
tiernan.doyle@nasa.gov
Jeremy Eggers
Goddard Space Flight Center, Greenbelt, Md.
757-824-2958
jeremy.l.eggers@nasa.gov
Share
Details
Last Updated Dec 09, 2024 LocationNASA Headquarters Related Terms
Science Mission Directorate Goddard Space Flight Center Heliophysics Joint Agency Satellite Division Kennedy Space Center NOAA (National Oceanic and Atmospheric Administration) Space Weather View the full article
-
By NASA
4 min read
Preparations for Next Moonwalk Simulations Underway (and Underwater)
The UAVSAR underbelly pod is in clear view as NASA’s Gulfstream-III research aircraft banks away over Edwards AFBNASA On a changing planet, where phenomena like severe hurricanes, landslides, and wildfires are becoming more severe, scientists need data to assess and model disaster impacts and to potentially make predictions about hazards. NASA’s C-20A aircraft is a significant asset that can carry key instruments for understanding the science behind these phenomena.
Based at NASA’s Armstrong Flight Research Center in Edwards, California, the C-20A is a military version of the Gulfstream III business jet and operates as an airborne science aircraft for a variety of Earth science research missions.
In October, the plane was deployed to fly over areas affected by Hurricane Milton. With winds of up to 120 miles per hour, the hurricane hit the Florida coast as a category 3 storm, and produced lightning, heavy rainfall, and a series of tornadoes. In the aftermath of the storm, the C-20A was outfitted with the Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) instrument to collect detailed data about the affected flood areas.
“Our team focused specifically on inland river flooding near dense populations, collecting data that could help inform disaster response and preparation in the future,” said Starr Ginn, C-20A aircraft project manager. “By all indications, this rapid response to support Hurricane Milton recovery efforts was a successful coordination of efforts by science and aircraft teams.”
The Uninhabited Aerial Vehicle Synthetic Aperture Radar, UAVSAR, is prepared for installation onto NASA’s C-20A aircraft. THE UAVSAR uses a technique called interferometry to detect and measure very subtle deformations in the Earth’s surface, and the pod is specially designed to be interoperable with unmanned aircraft in the future. It will gather data from Gabon, Africa in September of 2023.NASA/Steve Freeman The UAVSAR was developed by NASA’s Jet Propulsion Laboratory in Southern California, and uses a technique called interferometry to detect subtle changes to Earth’s surface. Interferometry uses the intersection of multiple wavelengths to make precise measurements. This detection system effectively measures the terrain changes or impacts before and after an extreme natural event.
When flown onboard an aircraft, radars like the UAVSAR can also provide more detail than satellite radars. “Where satellite instruments might only get a measurement every one to two weeks, the UAVSAR can fill in points between satellite passes to calibrate ground-based instruments,” Ginn said. “It takes data at faster rates and with more precision. We can design overlapping flights in three or more directions to detect more textures and motions on the Earth’s surface. This is a big advantage over the one-dimensional line-of-sight measurement provided by a single flight.”
The C-20A team also used the UAVSAR in October to investigate the Portuguese Bend landslide in Rancho Pales Verdes. The Portuguese Bend Landslide began in the mid- to late-Pleistocene period over 11,000 years ago. Though inactive for thousands of years, the landslide was reactivated in 1956 when a road construction project added weight to the top of it. Recently, the landslide has been moving at increasing rates during dry seasons.
NASA’s JPL scientists, Xiang Li, Alexander Handwerger, Gilles Peltzer, and Eric Fielding have been researching this landslide progression using satellite-based instruments. “The high-resolution capability of UAVSAR is ideal for landslides since they have relatively small features,” said Ginn. “This helps us understand the different characteristics of the landslide body.”
NASA flew an aircraft equipped with Uninhabited Aerial Vehicle Synthetic Aperture Radar (UAVSAR) flew above California fires on Sept. 3 and 10, 2020.NASA/JPL-Caltech The C-20A airborne observatory also provided crucial insight for studies of wildfire. The Fire and Smoke Model Evaluation Experiment (FASMEE), a multi-agency experiment led by the U.S. Department of Agriculture’s Forest Service Pacific Northwest Research Station, included flights of the C-20A. This experiment studied fire behavior and smoke.
“The airborne perspective allows FASMEE researchers to better understand fire behavior and smoke production,” said Michael Falkowski, program manager for NASA’s Applied Sciences Wildland Fire program. “Hopefully this data will help mitigate fire risk, restore degraded ecosystems, and protect human communities from catastrophic fire.”
Airborne data can inform how scientists and experts understand extreme phenomena on the ground. Researchers on the FASMEE experiment will use the data collected from the UAVSAR instrument to map the forest’s composition and moisture to track areas impacted by the fire, and to study how the fire progressed.
“We can explore how fire managers can use airborne data to help make decisions about fires,” added Jacquelyn Shuman, FireSense project scientist at NASA’s Ames Research Center in California’s Silicon Valley.
Share
Details
Last Updated Dec 04, 2024 EditorDede DiniusContactErica HeimLocationArmstrong Flight Research Center Related Terms
Armstrong Flight Research Center C-20A Earth Science Earth's Atmosphere Jet Propulsion Laboratory Explore More
4 min read 2024 AGU Fall Meeting Hyperwall Schedule
NASA Science at AGU Fall Meeting Hyperwall Schedule, December 9-12, 2024 Join NASA in the…
Article 41 mins ago 3 min read NASA Experts Share Inspiring Stories of Perseverance to Students
Article 2 days ago 2 min read This Thanksgiving, We’re Grateful for NASA’s Volunteer Scientists!
This year, we’re giving thanks to you for Doing NASA Science! You and the millions…
Article 1 week ago Keep Exploring Discover More Topics From NASA
Armstrong Flight Research Center
Earth Science
Climate Change
NASA is a global leader in studying Earth’s changing climate.
Jet Propulsion Laboratory
View the full article
-
By NASA
4 min read
Expanded AI Model with Global Data Enhances Earth Science Applications
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.
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.
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
Earth Science & Research Explore More
9 min read Towards Autonomous Surface Missions on Ocean Worlds
Article
23 hours ago
5 min read NASA-Led Team Links Comet Water to Earth’s Oceans
Scientists find that cometary dust affects interpretation of spacecraft measurements, reopening the case for comets…
Article
23 hours ago
1 min read Coming Spring 2025: Planetary Defenders Documentary
ow would humanity respond if we discovered an asteroid headed for Earth? NASA’s Planetary Defenders…
Article
23 hours ago
Keep Exploring Discover Related Topics
Missions
Humans in Space
Climate Change
Solar System
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.