Artificial Intelligence for Science

The Office of the Chief Science Data Officer (OCSDO) within NASA's Science Mission Directorate (SMD) researches how AI can help people use NASA's science data more effectively. Current projects include AI foundation models, which are trained on vast amounts of data and can be adapted to specific science studies with minimal resources, and large language models for enhancing different steps of research and data lifecycles.

ai for science data Strategy

5

AI foundation models

1

Large language model

AI for Science Focus Areas

  1. 01

    Prediction and Forecasting

    Develop data driven predictive modeling and simulation to allow scientists to forecast phenomena such as weather, space weather, and cosmic events.

  2. 02

    Open Science

    Streamline scientific workflows by improving data search, discovery, analysis, and visualization, thus enhancing research efficiency and promoting open science practices.

  3. 03

    Collaboration Across Sciences

    Facilitate broader scientific collaboration by integrating and analyzing data from diverse sources and missions, and overcome siloed mission science to move towards systems science.

  4. 04

    Infrastructure

    Establish and maintain a secure, scalable AI infrastructure to support model development and deployment within NASA's operational systems.

AI for Science Stories

Learn about the latest AI research for making the most of science data.

Hurricane Idalia as photographed by NASA's Terra satellite in August 2023. The swirling mass of the hurricane passes over some land masses and the ocean.

NASA, IBM Research Release New AI Model for Weather, Climate

NASA teamed up with IBM Research to create an artificial intelligence foundation model for a variety of weather and climate data applications. The model is now released to the public.

Members of the NASA Impact and IBM Research teams accept their certificate after winning the NASA Marshall Group Achievement Award.

Geospatial AI Foundation Model Team Receives NASA Marshall Group Achievement Award

A team of researchers from NASA and IBM received recognition for their work on an open-source AI geospatial foundation model for Earth scientists.

Five orange stars are connected with blue lines in a V-like shape, like a diagram of the constellation of Indus. Each of the stars is labeled with one of the NASA Science Mission Directorate divisions: astrophysics, biological and physical sciences, Earth science, heliophysics, and planetary science.

NASA-IBM Collaboration Develops INDUS Large Language Models for Advanced Science Research

A new large language model (LLM) created by NASA and IBM will power intelligent search tools for more effective science data discovery.

Headshot of Rahul Ramachandran in front of a blue background. He wears a light blue shirt, dark gray jacket, and glasses.

Marshall Research Scientist Enables Large-Scale Open Science

Meet senior researcher Rahul Ramachandran, who works on AI foundation models and LLMs to transform the use of NASA's science data.

Artificial Intelligence Science Data Projects

  • Prithvi Geospatial Foundation Model

    Pre-trained on Harmonized Landsat and Sentinel-2 (HLS) satellite data, the Prithvi Geospatial Foundation Model helps Earth science researchers conduct AI-powered studies such as flood mapping, fire-scar identification, and crop classification. The model is released openly on the Hugging Face platform.

    Access the Model

  • Prithvi Weather and Climate Foundation Model

    The Prithvi Weather-Climate Foundation Model enables Earth science projects like hurricane tracking, severe weather risk mapping, and renewable energy forecasting. It is pre-trained on Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA2) satellite data. The model is released openly on the Hugging Face platform.

    Access the Model

  • INDUS Large Language Models

    Named for the southern sky constellation, INDUS (stylized in all caps) is a suite of large language models tailored for astrophysics, biological and physical sciences, Earth science, heliophysics, and planetary science. The model is released openly on the Hugging Face platform.

    Access the Model

Explore More AI at NASA

NASA uses artificial intelligence for a wide variety of applications to plan missions, analyze data, and explore our universe.

Learn More
The far side of the Moon is depicted in colors of blue, purple, green, yellow, and red. Blue darkening to purple is centered in a large portion of the middle of the Moon, showing the extent of the South Pole-Aitken Basin. Green and yellow surrounds it, with blue usually highlighting craters which are in turn surrounded by green. Red indicates high elevations and is primarily found toward the top of the Moon image.