Rogue (freak) waves are rare events but they pose one of the greatest maritime risks. This Computational and Data-Enabled Science and Engineering (CDS&E) award would be used to develop a unified, data-driven framework and tools to advance the analysis and prediction of oceanic freak waves. The results of the research would provide key information for risk-based design of civil engineering and marine systems such as offshore platforms, wind turbines, and ships. The framework and tools would also benefit fundamental research on other complex systems that generate large amounts of data that must be processed-- ranging from manufacturing processes to disaster planning and response. The researchers' work on identifying signatures of rare events within big data sets could advance fundamental research in a variety of areas from medical diagnostics to airport security and the development of advanced materials. Collaborations with the National Oceanic and Atmospheric Administration (NOAA) and the US Navy will help to ensure that the work will be used to enhance weather forecasting and safety on the seas. The new framework is to be incorporated into a software package, which is used by NOAA, and is to be made available to researchers, students, and the public. Students will participate in the research and the results will be integrated into course materials and outreach activities for students in the University of Maryland, the US Naval Academy, and Women in Engineering. The researchers will also provide art-in-science displays on extreme wave phenomena for K-12 students.
This research will be used to integrate statistical learning, signal processing, and Koopman operator theory to develop software tools and processes that can be configured to identify energy localizations across a variety of domains--including ocean waves-- to provide a prediction capability. As rare events, there are few measurements of rogue waves but there is a wealth of information on storms, other atmospheric events and wave behavior. An interdisciplinary team will integrate existing NOAA databases, General-Purpose Graphics Processing Units (GPGPU) computing, and the application of machine learning. The investigators will advance our understanding of how to use parallel processing of large-scale computations done by combining the processing power of GPUs and Central Processing Units (CPUs). This enhanced understanding will advance fundamental research in dynamical systems by enabling more effective use of a suite of simulation and signal processing tools, such as wavelets, Fast Fourier Transforms, and Monte Carlo simulations.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
