74 



OCEAN AND MARITIME ATMOSPHERIC MODELING 



For a number of years the CNR has continuously funded, as a major 

 S&T effort, research to enable the Navy and the nation to numerically 

 model ocean and atmosphere environmental conditions world-wide. ONR 

 efforts are spearheading the DoD efforts in Global Change. Major field 

 experiments have been conducted in cooperation with NSF, NOAA, NASA, 

 SERDP and DOE in such Global Change Program efforts as TOGA/TOGA 

 COARE, WOCE, JGOFS, and RIDGE. 



One specific focus for these oceanographic and atmospheric 

 modeling efforts is the improvement of tropical cyclone forecasts made 

 by the Navy-led Joint Typhoon Warning Center in Guam. The center has 

 responsibility for tropical cyclone forecasts for the mid-west Pacific and 

 Indian Ocean basins, just as NOAA has it for hurricanes in the Atlantic and 

 East Pacific. Navy S&T efforts are seeking to improve our ability to 

 understand and model the complex interactions which control the path and 

 intensity of these tropical cyclones. 



Air-sea interaction is a fundamental but complex process, the 

 understanding of which is crucial to predicting the evolution of both the 

 ocean and the atmosphere. The CNR has major S&T programs focussed on 

 improving our understanding of these interactions as they are a critical 

 component of two Navy operational forecast models developed by NRL. 

 These are the Navy's Operational Global (NOGAPS) and Regional (NORAPS) 

 Atmospheric Prediction Systems. 



In the area of tropical cyclone forecasting and other atmospheric 

 forecasts, there is a long history of cooperation between NOAA and 

 ONR/NRL. A new area of NOAA/NRL cooperation began this year with an 

 effort to transform global numerical models from reliance on vector super 

 computers to new-generation computers which use Massively Parallel 

 architectures. We expect the first results to be in by the end of next year. 



Massively parallel techniques offer potential improvements in 

 capability that we have only begun to explore. Expected breakthroughs in 

 this area will not only dramatically improve forecast speed and accuracy, 

 but will affect the computer industry as well. The technologies and 

 applications currently under development are potentially applicable to 

 many complex numerical problems. 



