Publication Detail

A New Model System Aimed to Simulate Interdisciplinary Multi-Scale Oceanic Processes: A Tool for Great Lakes’ Ecosystem

Changsheng Chen, Robert C Beardsley
2009
41 pp.
MITSG 09-34
$5.50 DOM / $7.50 INT
ORDER HARDCOPY / DOWNLOAD

A state-of-art coastal ocean model requires 1) grid flexibility to resolve the complex coastline and steep continental slopes as well as multi-scale (global-basin-coastal region-estuarine) physical processes; 2) accurate numerical methods that conserve mass, heat and salt; 3) proper parameterization of vertical and lateral mixing, 4) advanced data assimilation methods to integrate observations with simulation results, 5) modular design to facilitate selection and/or addition of essential model components needed in both scientific or management applications. This model should be a robust and open-source system with a flexible user interface and supported by an expanding base of users that continue to improve it. A major step towards such a system has been taken by a team of University of Massachusetts-Dartmouth and Woods Hole Oceanographic Institution researchers who have developed a new prognostic, free-surface, three-dimensional primitive equations-based unstructured-grid, Finite-Volume Coastal Ocean Model (FVCOM). The finite-volume method used in this model combines the advantage of finite-element methods for geometric flexibility and finite-difference methods for simple discrete computation. FVCOM has been successfully applied in estuarine, continental shelf, regional, basin and global ocean studies involving realistic model domains (for more information go here). Several examples for multi-scale applications will be presented, including Lake Superior.

The present version of FVCOM includes a number of options and components, including: (1) choice of Cartesian or spherical coordinate system, (2) a mass-conservative wet/dry point treatment to simulate the flooding/drying process, (3) the General Ocean Turbulent Model (GOTM) modules for optional vertical turbulent mixing schemes, (4) several water-quality modules to simulate dissolved oxygen and other environmental indicators, (5) four-dimensional nudging, optimal interpolation, and Reduced/Ensemble Kalman Filters for data assimilation, (6) a fully nonlinear ice model for Arctic Ocean studies, (7) a three-dimensional sediment transport module for estuarine and near-shore applications, (8) a generalized biological module (GBM) for food-web dynamics, and (9) unstructured grid surface wave model modified from SWAN. GBM allows users to either select a prebuilt biological model or to build their own biological model using the pre-defined pool of biological variables and parameterization functions. FVCOM has been upgraded to the semi-implicit version, which allows users to choose either mode-split or semi-implicit schemes. Both schemes have included the non-hydrostatic dynamics, which can be used to resolve the small-scale internal waves and vertical convection. An automatic configuration of nesting module has been built into FVCOM, which allows to run multiple computational domain experiments through mass conservative nesting approach.

For hindcast and forecast applications, an integrated coastal ocean model system has developed. An example is the Northeast Coastal Ocean Forecast System (NECOFS), which has been used by the US National Weather Stations, local government agencies, and private companies. NECOFS is an integrated atmosphere-ocean model system in which the ocean model domain covers the northeast US coastal region (the New England Shelf, Georges Bank, the Gulf of Maine, and the Scotian Shelf) with a horizontal resolution of 10-15 km in the open ocean, 1-5 km on the shelf, and down to 20 m in estuaries, inner bays, inlets and harbors. The system includes: 1) two community atmospheric mesoscale models, WRF (Weather Research and Forecasting model) and MM5 (fifth generation NCAR/Penn State model), modified to incorporate the COARE 2.6 air-sea flux algorithm); 2) the unstructured-grid Finite-Volume Coastal Ocean Model configured for this region (FVCOM-GOM) with a nested higher resolution FVCOM configured for Massachusetts coastal waters (FVCOM-MASS); 3) the unstructured-grid surface wave model (FVCOM-SWAVES); and 4) the FVCOM-based unstructured-grid sediment model. In its present initial stage, the forecast system is built based on WRF, MM5 and FVCOM-GOM/FVCOM-MASS. Both meteorological and ocean models have been tested through comparison with field data in hindcast experiments covering the period 1979 to present. The system produces 3-day forecast fields of surface weather, surface waves, water temperature, salinity, and currents, with daily updating using hindcast data assimilated fields whenever field data are available. FVCOM-GOM and FVCOM-MASS are being upgraded with a new semi-implicit FVCOM code, which will allow regional and coastal as well as estuarine model runs with significant reduction in computational power. A brief description of NECOFS and some critical issues in the model application to the forecast operation are discussed in this presentation. This system is a potential candidate that can be set up for Lake Superior.

type: Presentations

Parent Project

Project No.: 2006-R/RC-103
Title: Development of a Management Model System for the New England Shelf

Other publications by this author (as Lead)