ACADEMIA

To quench society's relentless thirst for accurate forecasts, meteorologists the world over are seeking ever more powerful tools to help them predict the behavior of weather systems, oceans and climates. To meet their rapidly escalating demands, they are turning to high-performance computing (HPC) solutions. After a year that saw some of the most devastating weather systems in recent memory, many of the world's top meteorologists will convene here Jan. 29 through Feb. 2 for the annual meeting of the American Meteorological Society (AMS). Visitors to the SGI booth (Nos. 354/455) will see powerful new technologies designed to accelerate the large-scale simulations and analyses that meteorologists need to run to quickly and accurately pinpoint weather trends. "Timeliness has always been an essential factor in numerical weather prediction. Combined with today's high-resolution and compute-intensive weather models, this puts an extraordinary burden on HPC," said Mike Clancy, Technical and Scientific Director of the U.S. Navy's Fleet Numerical Meteorology and Oceanography Center, in Monterey, CA. "At Fleet Numerical, SGI HPC systems allow us to meet the very stringent timeliness requirements that naval operations place on our models." At the SGI booth, AMS attendees will be able to view a range of NOAA animations, a demonstration of Viz5D visualizations, and a selection of tsunami and hurricane models from Florida State University. As a first for AMS attendees, SGI will spotlight the new SGI Altix 4700 blade system, revolutionary blade server that delivers higher density while enabling new levels of performance and flexibility. SGI will also demonstrate its Media Fusion technology powered by the Silicon Graphics Prism visualization system. Media Fusion allows users to link and interpret massive amounts of data acquired through various data sources into one common operating picture to make effective decisions. The past year's unparalleled sequence of major weather-related disasters has pushed storm forecasts into the forefront of the public's consciousness. "It is clear that more accurate forecasts are needed to instill a level of confidence that will allow lives to be saved due to improved evacuation and disaster response activities," said Ilene Carpenter, Ph.D., weather and climate applications technical lead, SGI. "As leading meteorologists convene to discuss their innovations at the AMS meeting, SGI is proud to see its solutions enable vital breakthroughs in meteorological sciences." SGI drives meteorology innovation The AMS annual meeting arrives as more meteorological innovators than ever are tackling their most complex forecasting and modeling problems with SGI solutions. More than half of the world's major meteorological research and forecasting sites use SGI compute, storage and visualization solutions. For instance, to meet the rapidly increasing requirements of ultra-short-range and short-range atmospheric forecasting, the Hungarian Meteorological Service (HMS) recently selected SGI Altix as its new weather forecasting supercomputer. SGI won the HMS contract over HP and IBM due to the SGI Altix platform's blistering performance on the ALADIN numerical weather prediction model and on MM5. With the help of the new Altix system, HMS scientists can compute forecasts in just a few minutes, instead of the hour required on the institute's current hardware. Deployed in two phases (the first expected in February and the second by May 2006), the new system will also allow HMS to exploit even higher resolution numerical models with more complex dynamical and physical computations. Brazil's National Institute of Meteorology (INMET) also recently turned to SGI to equip its Division of Numerical Modeling with SGI compute, visualization and SAN solutions. With the new solutions, INMET will run a wide variety of complex meteorological sciences models and applications, including the High Resolution Model, the High Resolution Brazilian Model, the Grid Analysis and Display System, and several in-house codes. The extraordinary memory demands of INMET's source code and the need for simple code maintenance and updates were key drivers in the selection of SGI to enhance the forecasting and simulation environment at INMET. The new SGI solutions are expected to be installed in February. HMS and INMET join a wide array of innovative SGI customers devoted to weather forecasting and climate change research. These include: -- Baron Advanced Meteorological Systems (BAMS). SGI Altix systems and SGI InfiniteStorage solutions help BAMS meteorologists run HPC atmospheric modeling programs to predict the course of weather threats such as hurricanes, severe thunderstorms, and unhealthy air quality. BAMS supplies more than 200 broadcast television customers in the US with real-time, site-specific forecast data, modeled using technology from SGI. -- Catalan Meteorological Service (Meteocat). Based in Barcelona, Spain, Meteocat uses SGI Altix servers as a platform for MM5 models to alert civil defense agencies of rapidly changing weather conditions. Meteocat serves the Catalonia region, which includes the Pyrenees mountains, a region that is vulnerable to flash flooding. -- Cooperative Institute for Meteorological Satellite Studies (CIMSS). An SGI Altix supercomputer helps CIMSS, as part of the Space Science & Engineering Center at the University of Wisconsin-Madison, to model the state of the atmosphere over extremely large domains. These model results are then used to simulate potential observations from the next generation of geo-stationary weather satellites. -- Desert Research Institute (DRI). Based in Reno, Nev., DRI uses SGI server and storage technologies to drive and manage mesoscale meteorology models that help the institute deliver twice-daily forecasts of fire weather and smoke information to meteorologists and fire and air quality managers in California and Nevada. -- German Climate Research Center (DKRZ). Because Earth system modeling puts enormous burdens on traditional visualization solutions, DKRZ has driven 3D visualizations on a Silicon Graphics Prism visualization system, and recently boosted its graphics capabilities. DKRZ serves as a "laboratory" for scientists studying the climate system with coupled numerical models of the most important physical, chemical and biological processes. -- Jacob Blaustein Institutes for Desert Research. Located at the Sede Boqer campus of Ben Gurion University of the Negev in Israel, the institute uses an SGI Altix supercomputer to identify new mechanisms of change in the biological diversity of arid areas, while examining possible scenarios for the loss of animal and vegetation species resulting from climate and environmental changes. -- Nanjing University of Information Science & Technology. SGI Altix and InfiniteStoragesolutions are helping this Chinese facility develop a new, more accurate weather forecasting model. -- NASA Advanced Supercomputing facility. Located at Ames Research Center in Mountain View, Calif., NASA's 10,240-processor Columbia Altix supercomputer allows researchers to simulate decades of ocean temperature changes in only days. -- Space Pole. This alliance of the Royal Meteorological Institute, the Belgian Institute for Space Aeronomy, and the Royal Observatory of Belgium deployed an SGI supercomputing and storage solution to predict the changing weather conditions for Belgium's 10.3 million citizens, study the upper atmosphere, and identify climate trends. -- U.S. Department of Energy's Oak Ridge National Laboratory (ORNL). ORNL researchers use industry-standard SGI servers and storage solutions to project the impact of increased carbon emissions and to study mitigation strategies. -- U.S. Navy's Fleet Numerical Meteorology and Oceanography Center. Fleet Numerical leverages SGI solutions to run its Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS) model to predict weather and ocean conditions in complex coastal areas of the world. SGI delivers innovative technology to researchers and scientists employed in the study of such phenomena as climate change, tsunami, ozone depletion, environmental impacts in urban planning, groundwater pollution, and the improvement of severe event and climate forecasting. Beyond weather prediction modeling, SGI solutions support satellite and radar image processing and data management solutions for the terabytes of data managed in this industry as well as the air and water quality modeling arena.