July 9, 2008

image or animation

Tom Bettge. (©UCAR.)

NCAR Allocates Over 12 Million Processor Hours to Climate and Atmospheric Research

Climate Simulation Laboratory (CSL)

BOULDER—The Computational and Information Systems Laboratory (CISL) of the National Center for Atmospheric Research (NCAR) has allocated 12 million processor hours to 13 projects over the next 18 months for climate modeling in support of the multi-agency Climate Change Science Program.

These allocations were made through NCAR's Climate Simulation Laboratory (CSL) which is administered by the National Science Foundation. Proposals are peer-reviewed by the NSF-appointed CSL Allocation Panel. The science research for this year's awardees is sponsored by the National Science Foundation, U.S. Department of Energy, Environmental Protection Agency, National Oceanic and Atmospheric Administration, National Aeronautics and Space Administration, and the California Energy Commission.

Since 1995 the multi-agency-sponsored CSL has provided a sustained commitment to climatologists to study the impact of increasing greenhouse gases in the atmosphere. In addition to contributing to the reports of the Intergovernmental Panel on Climate Change, which was awarded the Nobel Peace Prize in 2007, CSL has advanced several significant, if less publicized, discoveries.

Projects that received allocations on NCAR's supercomputers started processing on NCAR's IBM POWER5 and IBM POWER5+ systems, and are moving to NCAR's new IBM POWER6, dubbed bluefire, in July 2008. Approximately half of the IBM POWER6 computer's 3,744 batch processors will be devoted to these 13 projects.

bluefire

Bluefire. (Photo by Carlye Calvin, ©UCAR.)

"The introduction of bluefire will allow scientists to simulate the impact of anthropogenic changes to Earth systems in much greater regional detail," says Tom Bettge, director of operations and services at CISL. "NCAR is committed to the success of the CSL by providing a secure, consistent, and sustained high performance computing environment in which scientists can develop, test, and apply the complex mathematical climate models necessary to advance these investigations."

The CSL computational facilities are housed, operated, and maintained by CISL at the NCAR's Mesa Laboratory, located in Boulder, Colorado. CSL was established in 1995 to provide high performance computing and data storage systems to support large-scale, long-running simulations of the Earth's climate system. These simulations incorporate atmosphere, oceans, land, and cryosphere, and associated biogeochemistry and ecology, on time scales of seasons to centuries.

As with many large efforts of this type, a wide range of institutions are involved in the CSL. The PIs and Co-Is for the 2008-09 projects are from the following universities and research institutions: Argonne National Laboratory, California Institute of Technology, Center for Ocean-Land-Atmosphere Studies, Columbia University, Duke University, Florida State University, George Mason University, Lawrence Berkeley National Laboratory, Massachusetts Institute of Technology, National Center for Atmospheric Research, National Oceanic and Atmospheric Administration, Pacific Northwest National Laboratory, Princeton University, Scripps Institution of Oceanography, University of Arizona, University of Bristol, University of California, Los Angeles, University of Colorado, University of Miami, University of Michigan, University of Virginia, University of Washington, University of Wisconsin–Madison, Washington State University, and Woods Hole Oceanographic Institution.

Projects include:

  • Significant development of the carbon cycle and atmospheric chemistry component for release with Community Climate System Model Version 4, and coupled runs to support this release
  • Seasonal and interannual prediction and predictability with coupled ocean-atmosphere-land models
  • Experiments examining the influence of chemistry using the Whole Atmosphere Community Climate Model, a chemistry-climate model
  • Ensemble data assimilation for climate model development
  • Production of a decadal coupled ocean/sea-ice state estimate of the Arctic
  • Cloud/aerosol parameterization in the application and improvement of general circulation modes
  • Dynamical downscaling of global simulations over California
  • Modeling climate change and climate variability in coupled climate-land vegetation models for present and past climates
  • Ensemble analyses of the impact and uncertainties of global change on regional air quality in the U.S.

University allocations in the atmospheric and related sciences

 More information: 

Contacts
Marijke Unger, External Relations,
NCAR Computational and Information Systems Laboratory
303-497-1285

Tom Bettge
NCAR Computational and Information Systems Laboratory
303-497-1371

On the Web
Bluefire main Website
Bluefire quick facts
Bluefire tech facts
Bluefire image gallery gallery

Every six months CISL awards large computing allocations to the university community for research in the atmospheric and related sciences. This spring CISL awarded 1.5 million processor hours to 20 projects whose research is supported by the National Science Foundation.

Each of the projects was successfully reviewed by the CISL HPC Advisory panel. The universities and non-profit research organizations being supported include: Center for Ocean-Land-Atmosphere Studies, Cornell University, Florida State University, George Mason University, Harvard University, Pennsylvania State University, Rutgers University, University of Alaska, Fairbanks, University of Colorado, University of California, Berkeley, University of California, Los Angeles, University of Maryland, University of Miami, and University of Utah.

Some of the largest projects being supported are:

  • Lateral Influences on Madden-Julian Oscillation Initiations
  • Understanding westerly wind burst modulation by the sea surface temperature
  • Fire at the Intersection of Global Carbon and Water Cycles
  • Semi-Stochastic Westerly Wind Bursts and ENSO Predictability
  • Development of New Statistical Learning Theory and Techniques for Improvement of Convection Parameterization in Climate Models
  • VAMOS Ocean-Cloud-Atmosphere-Land; climate simulation in the Southeast Pacific region using a regional earth system modeling framework
  • Understanding Observed Low Frequency Variability of Sea-Surface Temperature in the North Atlantic
  • Carbon-Climate Interactions with Increasing Water Demand

The Computational and Information Systems Laboratory (CISL) is part of the National Center for Atmospheric Research (NCAR) in Boulder, Colorado. NCAR is operated by the University Corporation for Atmospheric Research under the primary sponsorship of the National Science Foundation and other agencies. Opinions, findings, conclusions, or recommendations expressed in this publication do not necessarily reflect the views of any of UCAR's sponsors.