Artificial Intelligence (AI) refers to the simulation of human decision-making capabilities in machines. Machine learning (ML) is a subfield of AI that uses statistics and mathematical models to detect patterns in data. When applied to Big Data collections, such as NASA Earth observing data, AI and ML can be used to sift through years of data and imagery rapidly and efficiently to find relationships that would be impossible for a human to detect. NASA's Earth Science Data Systems (ESDS) Program is committed to the use of AI and recognizes its potential to significantly advance existing data systems capabilities, improve operations, and maximize the use of NASA Earth observing data.
ESDS AI/ML research is conducted primarily through NASA's Interagency Implementation and Advanced Concepts Team (IMPACT). Located at NASA's Marshall Space Flight Center in Huntsville, Alabama, IMPACT works to further the ESDS goal of maximizing the scientific return of NASA's missions and experiments for scientists, decision makers, and society. The IMPACT ML team consists of machine learning specialists, computer scientists, and Earth science data specialists and works to build tools and pipelines for applying ML algorithms to NASA Earth science datasets to improve data discovery.
Along with AI/ML work through IMPACT, teams at NASA Distributed Active Archive Centers (DAACs) are applying AI and ML to the data they archive and distribute. One example is the ongoing work at NASA's Goddard Earth Sciences Data and Information Services Center (GES DISC) to implement a machine learning framework using natural language processing (NLP) to make it easier for GES DISC data users to find appropriate datasets.
ESDS also fosters AI/ML research through NASA's Advancing Collaborative Connections for Earth System Science (ACCESS) program. This competitive program develops and implements technologies to effectively manage, discover, and utilize NASA's archive of Earth observations for scientific research and applications in support of NASA Earth science research goals. The ACCESS 2019 solicitation specifically sought technology developments for ML related to NASA Earth science data systems (including new training datasets for ML).
Another NASA-supported undertaking for fostering AI/ML research is the Frontier Development Lab (FDL). The FDL was created as an initiative through NASA’s Office of the Chief Technologist, and is an applied research accelerator based at NASA’s Ames Research Center in Silicon Valley, California. Through internal NASA collaborations as well as collaborations with academia and Silicon Valley companies, the FDL works to further NASA AI efforts.
NASA ESDS Program AI and ML Resources
AI in Action
Read about the many ways AI is being used in research and in the development of NASA-supported products and applications to improve the usefulness of NASA Earth observing data:
- Rise of the Machine (Learning)
- Counting Trees in Africa's Drylands
- Machine Learning Prototyping
- Deep Learning-based Hurricane Intensity Estimator
- ImageLabeler
ESDS AI/ML Research
Global research teams are finding new ways to apply AI to NASA Earth science data and data systems to enable faster, more efficient research. Read more about work that's taking NASA data to the next level:
- Advancing Machine Learning Tools for Earth Science: Workshop Report
- Advancing an Open-Access Repository for Earth Observation Training Data, and Machine Learning Models
- Developing Passive Satellite Cloud Remote Sensing Algorithms Using Collocated Observations, Numerical Simulation and Deep Learning
- GeoWeaver: Building An Open-Source Platform for Enabling Ad Hoc Management, Open Sharing, and Robust Reuse of NASA Earth Data-Driven Hybrid AI Workflows
- Machine Learning Datasets for Earth's Natural Microwave Emission
- Machine Learning Planet High Resolution Training Data for Medium Resolution Land Cover and Disturbance Mapping
- Pangeo ML: Open Source Tools and Pipelines for Scalable Machine Learning Using NASA Earth Observation Data
- Spatio-Temporal Machine Learning and Cloud Computing for Predicting Dynamics of Global Vegetation Structure from Active Satellite Sensors
- Training Data for Streamflow Estimation
- Artificial Intelligence and Machine Learning for Data Systems (Geoscience and Remote Sensing Society [GRSS] Webinar; January 12, 2021)
