DDDAS-TMRP: DynaCode: A General DDDAS Framework with Coast and Environment Modeling Applications

点击翻译按钮获取中文摘要

Focusing on the rich and very urgent scientific problems of ecological and coastal modeling of Lousiana's coastal region, DynaCode will create a powerful, generic, modern cyberinfrastructure for a new generation of complex DDDAS applications that will advance many areas of science and engineering. We will build upon and integrate well developed and widely used computational infrastructures and emerging standards, including Cactus, Triana, and grid services. By using very powerful application development tools that have already demonstrated advanced DDDAS capabilities, we are able to concentrate efforts on developing specific DDDAS algorithms and software modules that will apply not only to ecological and coastal modeling, but many other applications. The project will address scientific problems of particular interest to NOAA, involving scenarios that couple ocean circulation, storm surge, and wave generation models for the Gulf, as well as ecological, hydrodynamic, and sediment transport models of the Mississippi River Delta region of Louisiana, and will develop infrastructure and algorithms that enable these models to be coupled, both to each other and to external inputs from sensors, wind, and other data, and databases to optimize execution of complex workflows on grids, invoking appropriate models, meshes, and algorithms, depending on current conditions. The real time coupling of input data and complex workflows will allow, for example, dynamically invoking more accurate models and algorithms as a hurricane approaches the coast, choosing appropriate computing resources, comparing model results with actual observations, coupling freshwater control diversions with storm surge models to reduce hurricane damage and flooding, or guiding placement of future sensors and dams. In the course of this coastal modeling project, we will develop toolkits to facilitate DDDAS scenarios by enabling functionalities such as wrapping legacy codes, integrating framework support for advanced codes that allow dynamic composition of data as it becomes available, running multi scale simulations, and adapting to current conditions by dynamically recomposing if necessary.