Understanding Global Change:
Data Gathering, Simulation, and Analysis

We affect our environment and it affects us. To find out what those effects might be, ORNL researchers are using high-performance computing to study global and regional systems.

Data Gathering

Recognizing the critical link between data and supercomputing, ORNL is a key player in gathering and providing the best possible data and analysis -- information needed to advance our understanding of global climate change. ORNL operates several national data centers and data gathering experiments:

• CDIAC
• NASA DAAC
• ARM Archive
• FACE
• The DOE Center for Research on Enhancing Carbon Sequestration in Terrestrial Ecosystems (CSiTE)

  A major goal of CSiTE is to determine how to increase the amount of carbon entering into "pools" that are stabilized in soil and protected against decomposition. The center will also research ways to measure, monitor, and verify sequestration so that it may eventually be recognized in international treaties as a mitigation strategy.

• NARSTO

5lab report : "..a vigorous national commitment to develop and deploy cost-effective energy-efficient and low-carbon technologies could reverse the trend toward increasing carbon emissions...(with) energy savings ...roughly equal to or greater than costs."

Simulation

1997 Presidential address to the UN: "The science is compelling and clear: we humans are changing the global climate. ... In order to reduce greenhouse gases and grow the economy, we must invest more in the technologies of the future."

ORNL researchers are collaborating with other government agencies and universities to advance understanding of global climate change by combining the best possible information with advanced computing techniques in complex simulations.

• atmosphere (CCM)
• land surface model
• ocean (POP)
• river runoff
• dynamic sea ice

PCM collaborators:

• National Center for Atmospheric Research
• Naval Postgraduate School
• U.S. Army Cold Regions Research and Engineering Laboratory
• Los Alamos National Laboratory
• Oak Ridge National Laboratory
• University of Texas

ORNL's tasks:

• running climate simulations
• mapping parallel codes to machine memory hierarchies
• breaking scalability barriers
• evaluating early systems/benchmarking for climate
• developing and evaluating numerical methods

Analysis

ORNL's work in regional climate prediction combines data gathering, analysis, and simulation techniques. Many climate effects can only be properly assessed at regional scales or below. Just as fine scale data must be aggregated for use in global models, global predictions must be downscaled for use in regional predictions.

Global prediction and Regional data combine to generate Trends and Anomalies then researchers apply Statistical methods to get Regional predictions.

11lab report (from summary): "...the United States should develop and pursue a detailed and comprehensive climate change technology strategy. Further, the planning process should begin immediately, and implementation of the strategy should occur quickly."