Expanding Access to Science Participation: A FAIR Framework for Petascale Data Visualization and Analytics.

Area of Science:

• Data Science
• Climate Science
• Scientific Computing

Background:

• Massive scientific data generation presents accessibility challenges.
• Petascale datasets are often difficult for researchers to access and analyze.
• Democratizing access to large scientific datasets is crucial for innovation.

Purpose of the Study:

• To introduce a novel data fabric abstraction layer for democratizing access to petascale scientific data.
• To enable user-friendly querying and analysis of complex datasets, hiding underlying infrastructure.
• To facilitate FAIR (Findable, Accessible, Interoperable, Reusable) data access for the scientific community.

Main Methods:

• Developed a data fabric abstraction layer to simplify data access and querying.
• Utilized progressive compression algorithms and machine learning for scalable data visualization.
• Created interactive, browser-based dashboards for managing, visualizing, and analyzing petabytes of data.
• Enabled data analysis on diverse hardware, from supercomputers to laptops.

Main Results:

• Achieved user-friendly, FAIR access to NASA's petascale climate datasets.
• Enabled dynamic identification of extreme events and trends in large datasets.
• Improved climate scientists' ability to visually explore data through interactive dashboards.
• Successfully deployed dashboards and training in an academic setting and for the general public.

Conclusions:

• The novel data fabric effectively removes barriers to accessing and utilizing petascale scientific data.
• The approach democratizes scientific data, enabling broader discovery and educational applications.
• Interactive dashboards enhance data exploration capabilities for climate scientists and the public.