High Resolution Modeling of Coastal Circulation Using FVCOM in Puerto Rico and the U.S. Virgin Islands

Abstract

The Finite Volume Community Ocean Model (FV-COM) has been implemented in Puerto Rico and the US Virgin Islands (PRVI) by Caribbean Coastal Ocean Observing (CARI-COOS). Unlike the ocean models built on a structured grid, FVCOM is composed of staggered prismatic cells formed by the triangular elements in the horizontal to provide a more accurate geometric representation of complex coastlines. A variable spatial resolution of 100–3000 m is applied in PRVI, where the grid gets more delicate towards the coast. This coastal circulation model is initialized and forced at the boundaries by the U.S. Naval Oceanographic Office (NAVOCEANO) operational AmSeas forecasting system, which provides the sub-tidal elevations, temperature, and salinity at 3-hour intervals. Meteorological conditions are interpolated from the CARICOOS 2-km Weather Research and Forecasting Models (WRF). Tidal constituents are derived from the T_tide Matlab toolbox with harmonic constants of TPXO-7.2 tidal model. The daily SST and SSH from satellite products are used for data assimilation, as well as hourly surface currents monitored by High-Frequency Radars (HFR) along the southern and western coast of Puerto Rico. In the study domain, various observational data available from field observations (Acoustic Doppler Current Profilers (ADCPs), Buoys, tide stations, and HFRs) were analyzed to evaluate the proposed FVCOM model performance. The AmSeas model acts as a benchmark in this work and is compared to the FVCOM. Modeled results of sea surface temperature are validated with the observed data, indicating an obvious reduction of the root mean square error (RMSE). Compared statistically with ADCP and buoy observations, ocean currents show a significant improvement over the AmSeas model, especially at higher frequencies dominated by the tides, where the higher resolution coastal model can better represent the bathymetry.