NOAA's Grand Challenge research in climate prediction and weather forecasting is critical to its mission to describe and predict changes in the Earth's environment, manage the Nation's ocean and coastal resources, and promote global stewardship of the world's oceans and atmosphere. This research depends on advances in high-end computing and on the collection and dissemination of environmental data.
Increased computing power will enable higher resolution in the current models of the Earth's atmosphere-ocean system. Increased resolution will enable accurate representation of key features such as weather fronts and ocean eddies, and eliminate distortions due to clouds. More accurate NOAA models will improve the understanding of the behavior of climate and weather systems, making possible better decision making by government and industry on issues that affect both the environment and the economy.
NOAA is the Federal agency responsible for archiving and disseminating the Nation's environmental data, and is the principal repository for the Nation's climatic data for the U.S. Global Change Research Program. By increasing NOAA connectivity to the Internet, researchers and other users will have better access to this growing collection of large data sets located at more than a score of sites.
Through HPCC efforts in climate modeling, NOAA will provide better simulations of atmosphere-ocean coupling and a first-ever direct attack on the regional climate change problem. More accurate and more timely assessment of the future impact of climate change will make it possible to avoid "false choices" between the economy and the global environment. In weather forecasting, finer resolution in global and regional models will result in better weather forecasting and warning services, especially for hazardous weather and flight safety.
NOAA participates in the NREN, ASTA, IITA, and BRHR components of the HPCC Program as follows.
By using the Internet, NOAA researchers can easily access massively parallel systems at distant locations. Geographically distributed researchers can collaborate and easily share NOAA computing resources and data by accessing these systems remotely from high performance workstations.
NOAA plans to make its vast environmental data archives more accessible to the scientific community, while preserving the integrity of operational NOAA systems. Greater NOAA Internet connectivity will substantially improve access to these data. To support this increased Internet use, a NOAA Internet Network Information center has been established. The center will provide information and assistance to all NOAA Internet users and to scientists at other agencies and in academia with whom they collaborate.
NOAA is developing advanced algorithms and redesigning climate prediction and weather forecasting models to use new parallel programming paradigms.
NOAA will conduct a phased acquisition of scalable parallel systems for use in climate prediction and weather forecast modeling. The new models will be installed and evaluated on these systems.
NOAA proposes to investigate environmental monitoring, prediction, and assessment applications, and to expand efforts to make its environmental data more accessible.
The number of NOAA users who can effectively use scalable systems will grow considerably. In addition, substantially more visiting scientists will work with NOAA researchers on Grand Challenge problems, encouraging both evolutionary improvements and creative breakthroughs in climate prediction and weather forecasting.
Princeton, New Jersey
Camp Springs, Maryland
NOAA's pilot Internet Network Information center has begun to provide assistance in using the Internet, network management for connected NOAA facilities, mail and other server capabilities, and protected access to selected NOAA operational data sets, such as NMC model output data.
In FY 1994, at least 10 more NOAA computational and data archiving facilities will be connected to the Internet, and connection bandwidth will be increased.
Over the past two years, NOAA's Geophysical Fluid Dynamics Laboratory (GFDL) has begun redesigning its most important atmospheric and oceanic models to make them modular and parallel in design. This activity includes a collaborative effort by GFDL with DOE/Los Alamos National Laboratory (LANL) scientists that has led to the successful design of a model shell version of the GFDL Modular Ocean Model (MOM) to a form that is amenable to highly parallel computers and that maintains its modular coding structure. High resolution experiments using this new code were performed on the 1,024-node Thinking Machines CM-5 at LANL.
GFDL's complete SKYHI global atmospheric grid-point model, which is used to study climate, middle atmosphere dynamics, and ppp atmospheric ozone, has been redesigned for parallel execution and ported to the LANL CM-5, as part of the GFDL-LANL collaboration.
The GFDL radiation physics package, which is a critical part of all GFDL atmospheric climate models, was redesigned to a modular form that is suited to highly parallel computer systems.
In FY 1994, high resolution atmospheric modeling experiments using a simplified-physics version of a parallel atmospheric grid model will be performed to test model sensitivity to grid resolution.
In FY 1994, a parallel version of the GFDL limited-area non- hydrostatic model will be developed for use in investigating the interaction between radiation and clouds.
An adiabatic shell for the NOAA National Meteorological center (NMC) spectral model has been developed and executed on a Thinking Machines CM-200 at DOD/Naval Research Laboratories (NRL) in collaboration with NRL.
A regional atmospheric model has been restructured for execution on highly parallel systems by NMC, in collaboration with NOAA's Forecast Systems Laboratory (FSL).
The initial restructuring and recoding of the regional/mesoscale Optimum Interpolation (OI) analysis code for execution on the CM- 200 at DOD/NRL has been completed.
FSL has developed a parallel version of the Mesoscale Analysis and Prediction System (MAPS) as a functional prototype system for both the Federal Aviation Administration and the National Weather Service.
MAPS is the first of several strategic weather models to be parallelized for the Aviation Weather Program. These models will provide high resolution forecasts on both the national and regional levels to support operational and aviation meteorology. The MAPS model was engineered using an FSL-specified layered software approach to provide portability among scalable parallel systems.
A computer model representation of a blizzard over the Colorado Front Range area, produced jointly by NOAA's Forecast Systems Laboratory (FSL) and the Colorado State University. The image represents a six hour forecast, looking over Colorado from the southeast. The white/gray area depicts the forecast region for significant clouds, and the red area within the clouds represents the region in which aircraft icing is likely to occur.
A scalable architecture system, a 208-node Intel Paragon, is being installed at NOAA/FSL in conjunction with the ARPA-sponsored National Consortium for High Performance Computing and as a part of the Boulder Front Range Consortium. Consortium members include NCAR (under NSF sponsorship), the University of Colorado (under ARPA sponsorship), and FSL.
FSL has developed a benchmark suite to evaluate parallel processors for use in weather analysis and prediction applications. FSL has also specified and is implementing a layered software approach to support the portability of such applications between various scalable systems and networks of workstations, and has taken the initial steps toward implementing a meteorological software library using this approach.
GFDL has solicited technical input from massively parallel computer vendors through a public comment process as part of its effort to redesign its primary atmospheric climate model to a suitable parallel design. The resulting technical collaborations are leading to a more effective code design for this and other models that are being rewritten for parallel systems.
Contours representing the wind flow in the middle atmosphere, produced by a weather forecast model running on the Cray Y-MP8 at NOAA's National Meteorological center, superimposed on a NOAA GOES satellite infrared cloud image.