BATON ROUGE, La., Sept. 29 -- Hurricanes are an annual dangerous threat to Louisiana and other coastal states. But this year, many coastal residents wondered whether the Gulf of Mexico oil spill would worsen the potential impact of storm surge on the Louisiana coast.

A group of researchers with the LSU Center for Computation & Technology, or CCT, received complementary grant awards from the Louisiana Optical Network Initiative, or LONI, to model and visualize how severe storms would affect movement of the spilled oil particles. Gabrielle Allen and Bijaya Karki, who both are professors with the LSU Department of Computer Science, each received $10,000 grants and 500,000 CPU cycles on LONI to conduct a two-month study. Their awards also tie in to research with LSU's Q. Jim Chen, a professor with the Department of Civil and Environmental Engineering and CCT who received grant awards from LONI and TeraGrid, a national cyberinfrastructure for scientific research, to use those resources in conducting modeling toward oil spill research.

These awards involve joint work among the Coastal Modeling, Scientific Visualization and Computational Frameworks research groups at the CCT, and includes Sumanta Acharya from Department of Mechanical Engineering and Carola Kaiser with the School of the Coast and Environment, CCT researchers Jian Tao, Peter Diener, Werner Benger, and Marcel Ritter, post-doctoral researchers Kelin Hu and Haihong Zhao, LSU Department of Civil and Environmental Engineering, computer science graduate student Bidur Bohara, mechanical engineering graduate student Somnath Roy, and undergraduate mechanical engineering student Edwin Mathews.

"This is preliminary research into an area of great concern, and we plan to generate visualizations from our models that give scientists more insight into how severe storms would affect the path and spread of such an oil spill," Allen said. "Ultimately, we want to create comprehensive and quickly deployable environmental models that incorporate multiple elements such as hurricane winds, storm surge, oil spill and fish populations along with advanced visualization and analysis to better understand the coastal environment."
 
Most hurricane and storm surge models are not designed to include the three-dimensional transport of oil. To use such models, scientists need to combine other models using different physics. This research project will create a preliminary hurricane model that incorporates oil as an element.

"This project addresses some interesting challenges for computational science researchers, in terms of understanding the evolution of the oil spill as emerging from its underground source into the seawater," Karki said. "A wide variety of data related to the oil spill in the Gulf of Mexico is currently available and will be generated in the future. We plan to visualize some of these data to extract important information about the nature and extent of the oil spill."

For the first part of the project, the researchers who are part of Allen's grant will create an oil spill model that treats the spilled oil as individual particles using the Cactus Computational Framework, an open-source environment that allows scientists and engineers to use high-performance computing resources more effectively. Once a model for oil particles is in place, the researchers will input data on different wind speeds and water current movements to see how the oil particles behave in extreme weather conditions.

For initial examples, the research group will use wind speed and water current data they already have available from Hurricane Katrina (2005) and Hurricane Gustav (2008), expanding on data from these storms to simulate how and where oil from the spill would move under hurricane-force winds and storm surge from hurricanes on similar tracks.

For the second part of the project, the researchers working on Karki's grant through the CCT Scientific Visualization research group will conduct the final step of creating images of the data that describe the plume of oil in water and where the oil dissipates up to the surface because of immense upstream pressure, high temperature and buoyancy. They will produce movies and still images that will show researchers how spilled oil would move during a hurricane under different scenarios.

"Being from southern Louisiana, it is very exciting to be a part of a project that can directly affect my home and the entire Gulf South," said Edwin Mathews, an undergraduate mechanical engineering student working on the project. "On top of that, getting to work with a research group at a high-end facility like the CCT is a unique and eye-opening experience for an undergraduate student. Visualization is essential to the interpretation of the computational oil spill data, and being exposed to the inter-disciplinary workings of the project is something that will be of great value to me in the future."

The researchers on this project had experience creating multi-element, data-driven models for hurricanes through work on Cybertools, a National Science Foundation-funded project to develop tools and applications that allow scientists to use modern cyberinfrastructure to its full potential. All of the researchers on the LONI oil spill project had worked on the Cybertools project.

The research groups will use LONI's high-performance computing and advanced networking capabilities to make data input easier for the numerical model, and create visualizations using VISH, a software program developed at the CCT in collaboration with colleagues in Austria and the United Kingdom. VISH stands for "visualization shell," meaning it is an environment that can incorporate multiple elements of the visualization process while implementing newly developed algorithms and deploying them to researchers easily.

They plan to present the initial results of their research during the annual Supercomputing Conference (SC10), which will take place in New Orleans, Nov. 13-19, 2010.

-----

Source: LSU Center for Computation & Technology