NASA Logo, National Aeronautics and Space Administration
Intelligent Systems Division Banner

Data Science For Climate Networks Proposal Awarded

"Scalable Analysis of Earth System Data Using Parallelized Graph-Based Approaches" is one of seven selected proposals from among twenty-three submissions to the Computational Modeling Algorithms and Cyberinfrastructure (CMAC) program under the NASA ROSES 2014 call for proposals. This proposal has Dr. Vipin Kumar (University of Minnesota) as the PI, Dr. Kamalika Das (UARC, NASA Ames) and several colleagues from University of Minnesota as Co-Is. The project will run for two years.

The objective of this project is to develop graph-based algorithms for clustering (community discovery), anomaly detection, and dipole discovery for a systematic evaluation of climate model outputs. This will enable model inter-comparison, understanding different climate phenomena and their interactions, and discovering new long-range spatio-temporal dependencies. Since most graph algorithms are computationally intensive, our goal is to make them scalable and efficient for high-resolution data sets through the use of computation primitives that can take advantage of modern multi-core and high-performance computing architectures, such as the NASA Advanced Supercomputing facility.

BACKGROUND: Rapid gains in computational power and storage capacity have resulted in a data deluge for climate science in terms of observed and Global Climate Model (GCM) simulated climate data. This proposal aims to address the need for better understanding of interactions within the climate system and improved model evaluation by developing a set of tools for analyzing graphs that are constructed as a representation of observed or simulated climate data. The tools will then be integrated into the NASA Earth EXchange (NEX) for use and further refinement by the Earth science community.

The specific goals to be addressed as part of this proposal are: (1) enhance existing graph construction and network analysis algorithms to work on heterogeneous, high-resolution climate graphs with the goal of understanding various climate processes and projections better; (2) develop Open-Source software for these multi-core network analysis algorithms; and (3) integrate these new Open-Source complex network analytic modules into existing NEX data access and management infrastructure. Integration will provide exploratory capabilities to climate scientists for further investigations into a wide range of complex Earth-system processes and interactions at multiple spatial and temporal scales.

PROGRAM FUNDING: Computational Modeling Algorithms and Cyberinfrastructure (CMAC) program, Science Mission Directorate (SMD)

COLLABORATORS: PI: Vipin Kumar (UMN), Co-I’s: Kamalika Das (UARC, NASA), Gowtham Atluri, Stefan Leiss, Michael Steinbach, and Karsten Steinhaueser (UMN)

POC: Kamalika Das,

First Gov logo
NASA Logo -