Report Says New Systems Required To Better Exploit Satellite Data

WASHINGTON
— The increasing capabilities of Earth observing satellites and demand for improved weather forecasts is helping spur the need for high-end capability computing, according to a National Research Council report being prepared for the White House National Science and Technology Council.

The report, “The Potential Impact of High-End Capability Computing on Four Illustrative Fields of Science and Engineering,” which is in prepublication form and subject to final revision, examines the computer technologies needed to advance atmospheric sciences, astrophysics, evolutionary biology and chemical separations. All four subject areas rely upon high-end capability computing – advanced computer hardware, software and data – to simulate theories and forecasts that cannot be physically observed or tested.

The committee of 16 scientists who contributed to the report found that advances in computing would immediately extend the range, accuracy and utility of weather predictions. Such advances also would improve scientific understanding of pollution, climate change, the atmospheric forces associated with moisture and chemical exchange at the Earth’s surface, and make possible more timely predictions of severe weather, according to the report.

New satellites coming on line, such as the National Polar-orbiting Operational Environmental Satellite System (NPOESS) and the next-generation Geostationary Operational Satellites (GOES) will require a tenfold increase in computing capabilities to manage and analyze the complex data they generate, said meteorologist John Dutton, chief scientist of New York-based Storm Exchange and a member of the committee.

“Major weather services of the world ingest billions of data sets every day,” Dutton said during a Sept. 22 symposium, “Potential Impacts of High-End Capability Computing,” at the National Academies of Science in
Washington
. “The assimilation of NPOESS and GOES requires high-end capability computing.”

While weather forecasting is largely based on computer modeling, merging data that comes from different sources or is collected at different times can be a greater challenge than creating appropriate, reliable models, the report said. That is especially true for older data that requires reanalysis as capabilities improve, such as the recently completed North American Regional Reanalysis, which contains 32-kilometer-resolution atmospheric and surface data from 1979 to the present, the report said.

Data from NASA Moderate Resolution Imaging Spectroradiometers, which were launched in 1999 and 2002 aboard the Terra and Aqua satellites, respectively, also have been reanalyzed as computer technology has advanced.

“Algorithms to derive climate variables have improved and each improvement requires reprocessing of entire data sets,” the report said. “With current NASA computing capability, about four days of data can be processed in one day of computation.”

High-end capability computing also will help scientists seeking clues to the formation and evolution of stars, planets, galaxies, quasars and super massive black holes, said Lars Hernquist, professor of astronomy at
Harvard
University
and a member of the committee.

“The only way we’re going to be able to study these things is through the use of computer calculations,” Hernquist said at the symposium. “Analytical theory will only get us so far.”

NASA’s Wilkinson Microwave Anisotropy Probe and the Cosmic Background Explorer (COBE) missions provided data on initial conditions during the formation of the universe, but astrophysicists want to know how quasars, galaxies and super massive black holes have evolved since then, the report said.

Advanced computer capabilities also could lead to discoveries about what triggers high-energy supernovae and gamma-ray bursts, the report said.

The report warned that without computer technology advances for modeling astrophysics, data collected is likely to be “underexplored.”

“Inadequate support for high-end capability computing would lead to a failure to optimize investment in expensive experimental and observational facilities,” the report said.

President George W. Bush requested the report in his 2006 budget. The committee was formed under the National Research Council’s Division of Engineering and Physical Sciences and the Division on Earth and Life Sciences.