Implementation of an infrastructure to support operation and evaluation of Gulf of Mexico models

Describes implementation of a modeling infrastructure that was necessary to the transition of a Gulf of Mexico model developed by Dynalysis of Princeton called Princeton Dynalysis Ocean Model with assimilation (PDOM-A) from a research mode to a model that produced twice daily 48-hour forecasts for the Gulf of Mexico at the Naval Oceanographic Office (NAVOCEANO) Major Shared Resource Center (MSRC). The MSRC provides high performance computing services and support to the U.S. Department of Defense. It is located within the NAVOCEANO headquarters complex at NASA´s John C. Stennis Space Center on the Mississippi Gulf Coast. PDOM-A model runs in a hands-off mode, and its implementation illustrates that the model can be run without the modeler. The model forecasts are provided to the oceanographic community via the Master Environmental Library (MEL) as data and depicted on the Northern Gulf of Mexico Littoral Initiative (NGLI) Web page at http://128.160.323.41. MEL is a Defense Modeling and Simulation Office (DMSO)-sponsored one-stop site for ordering environmental operation. NGLI is a multiagency program established through a partnership between the Commander, Naval Meteorology and Oceanography Command, and the Environmental Protection Agency´s Gulf of Mexico Program. Forecasts are also available at http://www.dynalysis.com/Projects/projects.html#oms-gom. As part of the NGLI Program, forecasts from the model are also used to specify outer boundary conditions for a regional model for the Mississippi Bight - called Estuarine and Coastal Ocean Model (ECOM). One of the objectives of the PDOM infrastructure was to implement a system that was model-independent. The model dependency would be contained in the software used to create the data need to run a particular model, such as PDOM-A. The modules that comprise the PDOM system and the connecting logs and data files provide a robust environment for operational models. Each module performs a series of functions. As needs change, additional functions can be added to that module. The modules in the infrastructure are the Gather Environmental Data Module; Environmental Data Processing Module; Expert Modeler Module; Model Data Preparation Module; Model Simulation Module; and Model Po- st-Processing and Skill Assessment.