Two critical issues must be solved in the development of a numerical model for the Gulf of Maine/Georges Bank. One is the steep-slope bottom topography and another is the fitting of irregular geometric coastlines, island complexes, and barriers. We have successfully applied our new FVCOM to the Gulf of Maine/Georges Bank region. Compared with the ECOM-si model (set up by us for the Gulf of Maine/Georges Bank region), FVCOM shows promise in the simulation of tides, tidal mixing and also wind and buoyancy-driven flow.

This is the model that successfully resolves the near-resonance M2 tide in the Bay of Fundy. With high horizontal resolution, FVCOM also provides more accurate distribution of tidal-induced clockwise residual flow around Georges Bank and adjacent coastal regions and the buoyancy flow around the coast. Numerical experiments in the Gulf of Maine/Georges Bank have clearly shown that the sigma-error caused by the steep-slope bottom topography at the shelf break and northern flank would become a serious issue if the horizontal resolution were not high enough. A series of numerical tests have been conducted by FVCOM using the realistic geometry of the Gulf of Maine/Georges Bank/Northern Atlantic Ocean. As 60 non-uniform sigma levels are chosen, at least 1 km horizontal resolution is needed to keep the sigma-error smaller than 1 cm/s. A detailed discussion of this issue is found in C. Chen's recent manuscript on the application of FVCOM to the Gulf of Maine/Georges Bank region.*

FVCOM is an unstructured grid, finite-volume, 3D primitive equation, turbulent closure coastal ocean model developed recently by Chen et al. (2001e). As POM, ECOM-si and QUODDY, FVCOM is a prognostic model, incorporating free surface and the modified Mellor and Yamada level-2.5 (MY-2.5) and Smagorinsky turbulent closure for vertical and horizontal mixing (Mellor and Yamada, 1982; Galperin et al., 1988; Smagorinsky, 1963).

The model uses a transformation in the vertical to convert irregular bottom topography into a rectangular computational domain for a simple numerical approach. Like POM, FVCOM is composed of external and internal modes that are computed separately using two split steps. The distinct difference is that FVCOM is solved numerically by the flux calculation in the integral form of primitive equations over non-overlapping, unstructured triangular grids. This numerical approach combines the best of the finite-element method for accurate coastal geometric fit and the finite- difference method for simple structure of the model code and computational efficiency.

In addition, the flux calculation method with an integral form of equations provides a better representation of momentum, mass, salt, and heat conservation. The finite volume numerical method used in FVCOM was described in detail by Chen et al. (2001e). You can see an example describing how the velocity and surface elevation are calculated over each triangular cell of the model on the Marine Ecosystems Dynamics Modeling web site.

*Chen, C., and R. Beardsley. 2002. Cross-frontal water exchange on Georges Bank: modeling exploration of the US GLOBEC/Georges Bank phase III study. Journal of Oceanography, 58, 403-420.

Chen, C. H. Liu, R. C. Beardsley, 2002. An unstructured, finite-volume, three-dimensional, primitive equation ocean model: application to coastal ocean and estuaries. Journal of Atmospheric and Oceanic Technology,20, 159-186.