Bringing technology to bear on critical issues

When it comes to hurricanes, Mother Nature still rules, as Katrina so vividly reminded us last year. That doesn’t mean humans are powerless in their struggles with major coastal storms. High-performance computing experts at RENCI are working with marine and atmospheric scientists and North Carolina disaster response agencies to use advanced technologies to improve storm prediction, modeling, and mitigation.

RENCI is a partner in the Southeastern Universities Research Association’s (SURA) Southeastern Coastal Ocean Observing and Prediction Program (SCOOP), an effort that involves SURA universities from Maryland to Texas and works to establish an integrated ocean observing system supported by high-performance computing and a comprehensive information infrastructure. SCOOP researchers from the Chesapeake Bay to Florida to the Texas Gulf Coast are creating models of storm-driven water levels, and RENCI has developed a grid-enabled system for real-time computing of these models.

The system is still in its infancy and will be tested for the first time this hurricane season. Long term, what researchers hope to put in place is a system that gathers data from a variety of sources—including stored data and real-time data from satellites and sensors—distributes it to various SURA resources for data processing and visualization, and then outputs accurate, high-resolution storm surge model data which scientists up and down the coast can share.

“SCOOP is a research project, which means its infrastructure won’t necessarily be for the front line emergency workers, but what we can provide is the needed research base and the baseline data that is the foundation of good emergency response,” said Rick Luettich, director of UNC-Chapel Hill’s Institute for Marine Sciences and a SCOOP scientist.

Luettich, who is based in Morehead City on the North Carolina coast, is working with RENCI to process data describing coastline position, the topography of the ocean bottom and low lying coastal regions, coastal land type and use, and atmospheric storm conditions. This data will be used to create a highly detailed version of his Advanced Circulation Model (ADCIRC) that is specific to the North Carolina coast.

Luettich helped set up a version of the ADCIRC model for Southern Louisiana and New Orleans that was used by Louisiana State University scientists to accurately forecast storm surges in advance of hurricanes Katrina and Rita last year.  Since then, the model has been used extensively to analyze conditions contributing to levee failure and to prepare designs for enhanced storm protection for the region.  The model also was used to assess possible movements of hazardous materials, such as chemical spills, that were initiated by Katrina’s storm surge inundation. RENCI helped Luettich secure the computing power necessary to do these models quickly and is supporting the effort to create the North Carolina-specific version of ADCIRC.

“SCOOP is about using high-performance computing resources to create more accurate high-resolution models and creating a distributed infrastructure to support this kind of work,” explains Luettich. “RENCI has a mission to use high-performance computing technologies to address critical problems in North Carolina, and those two goals compliment each other very well.”

RENCI is working to help both SCOOP scientists and storm prediction and mitigation in North Carolina by making the process of sharing data, using compute resources and creating models as seamless and simple to use as possible, says Lavanya Ramakrishnan, a senior grid software developer at RENCI.

“We want to create a robust system that requires little human intervention,” she explains. “If a computing system at one site goes down, a computing job should switch to a backup site and the scientist shouldn’t need to worry about it. We want a system that lets them concentrate on the science.”

Toward comprehensive disaster models
In addition to its work with SCOOP, RENCI has launched a comprehensive effort to work with the state of North Carolina on using high-end technologies and IT infrastructure to improve disaster response and mitigation and thereby lessen the human and economic costs of hurricanes and other natural disasters. RENCI’s Ken Galluppi, formerly with the Carolina Environmental Program at UNC-Chapel Hill, works with state agencies through the North Carolina Division of Emergency Management and with faculty from the UNC System to provide research and development in disaster studies and comprehensive tools for preparing for and responding to natural and man-made disasters. RENCI and its state partners plan to create comprehensive models of disaster situations that incorporate data from atmospheric models, storm surge models and transportation networks.

“A multidimensional model that incorporates atmospheric conditions, the movement of water along the coast and in flooded areas and the latest information on road conditions and possible escape routes would be an invaluable tool for evacuation and rescue missions and also for planning,” says Galluppi. “It will require a lot of computing power, a lot of work merging disparate data sets and a lot of integration of networking and communications systems that currently can’t talk to each other, but we think we are up to the task.”

For more information :
North Carolina Division of Emergency Management