Hurricane Katrina water-level models and observations created using OpenIOOS and SCOOP technology. Boxes show coastal water level observations, while colors in the ocean indicate model predictions. Image Courtesy OpenIOOS

Coastal researchers can now harness only a limited amount of up-to-date monitoring information and computing power for their predictions. The Southeastern Universities Research Association (SURA) has undertaken the SURA Coastal Observing and Prediction (SCOOP) program with the hope to change that, by creating the first distributed real-time environmental prediction system.

"We're creating a prototype distributed laboratory that's advancing the science of environmental prediction and hazard planning," says SCOOP Program Director Philip Bogden. The SCOOP cyberinfrastructure will initially be focused around the southeastern coast of the United States, first integrating diverse data flows from a variety of already established coastal ocean observing efforts and then incorporating the data flows into an open-access, scalable environmental prediction system.

Eventually SCOOP will impact hazard planning for the nation and perhaps the world by providing the data integration and management capacity for a national initiative called the Integrated Ocean Observing System, the ocean component of the Global Earth Observation System of Systems.

The project will also revolutionize environmental prediction by making real-time ensemble prediction of coastal impacts a standardized tool for the first time. Estimating future environmental impacts is much more difficult than simply creating the single best forecast for a certain weather event. Scientists must also estimate the uncertainty of their best forecast, and calculating the associated probability distributions requires computer simulations to be run not once, but many hundreds or thousands of times—once for each plausible outcome—creating huge computational demands.

"As a scientist, you try to increase the probability that your prediction will have a bearing on the future," explains Bogden. "You not only try to predict the single most likely scenario, but also how reality may deviate from that scenario. The more data you can have to start your prediction, the closer to reality that prediction will be." SCOOP will provide more of more types of data and the computational power to carry out ensemble predictions.

Coastal scientists haven't studied such problems in the past because of the difficulty in setting up the computations. Computer scientists, best suited for the computational part of the research, didn't have the understanding of the science necessary to attack the problem.

"Many of the issues with ensemble prediction are not interesting to coastal researchers but are compelling and pertinent to the computer science and grid community," adds Bogden. "SCOOP is bringing together the two communities to address complementary issues that are critically important for real-time environmental prediction."

The same system that will support ongoing forecasts for use in day-to-day operations will also be used for research on past events. During the 2005 hurricane season, parts of the SCOOP architecture were already used to create predictions of the storm surges from Hurricane Katrina. Biweekly simulations of Katrina are now being run to iron out problems and upgrade the system in preparation for the 2006 hurricane season.

"Instead of coastal researchers spending 90% of their time setting up a problem and 10% doing the science, we hope to someday get to the point where they spend 10% of their time plugging in to the infrastructure and 90% of their time doing science," says Bogden.

SCOOP represents a collaboration of seven SURA universities, and is funded by the Office of Naval Research and the National Oceanic and Atmospheric Administration.

Learn more at the SCOOP Web site.