Recorded presentationJ3.3
COAMPS-OS®: Transitioning coupled atmospheric and ocean prediction using the Earth System Modeling Framework (ESMF) into operations
Daniel A. Geiszler, SAIC, San Diego, CA; and J. Cook, S. Chen, M. D. Frost, P. Harasti, C. Hutchins, A. Zhao, R. A. Allard, T. J. Campbell, D. Martinez, P. May, L. McDermid, and J. E. Kent
COAMPS-OS is a suite of software supporting operational and research-oriented atmospheric and ocean prediction for Navy and non-Navy users. At the core of COAMPS-OS is the Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS) providing 24/7 regional weather prediction for Navy operations around the world. The “OS” in COAMPS-OS stands for “On-demand System” and symbolizes the capabilities allowing users to rapidly set up and run COAMPS and view the forecasts from web-based clients. During the last three years, the modeling capabilities of COAMPS-OS have been extended to include three-dimensional ocean prediction (NCOM; Navy's Coastal Ocean Model), near-shore wave prediction (SWAN; Simulating WAves Nearshore), and most recently, deep-water wave prediction (WW3; WAVEWATCH III). The relationship among the models in the current version of COAMPS-OS is best described as a one-way coupled system where COAMPS provides atmospheric forcing to each ocean model, but the ocean models do not influence the COAMPS forecast, nor do the ocean models influence each other's forecast.
This paper will describe the integration of ESMF (Earth System Modeling Framework) coupled versions of COAMPS, NCOM, SWAN, and WW3 into COAMPS-OS. Coupling using ESMF permits two-way interaction among select atmospheric and ocean parameters at the time steps and between the native computational grids of each model. Enhancements to other components of COAMPS-OS such as data management, the graphical user interface (GUI), visualization, and verification will also be described.
Joint Session 3, Advances in Modeling and Forecasting
Tuesday, 28 September 2010, 1:30 PM-3:15 PM, Capitol AB
Previous paper Next paper