Incorporating HF-Radar and Lagrangian Drifters into a Coastal and Mesoscale Model Validation Framework in the Eastern Caribbean Sea

The geographical location of Puerto Rico and the U.S. Virgin Islands poses a challenge in the numerical modeling of coastal and regional currents due to the range of hydrodynamical scales and forcing mechanisms. The sharp tidal gradient between the Atlantic Ocean and the Caribbean Sea generates strong North-South currents across the shelves, while the mesoscale features and the Eastern Caribbean regional circulation dominates the East-West circulation along the Southern coasts on the region. As the need of numerical models for support of maritime operations has increased, so has increased the need for a thorough model validation and development framework which takes advantage of the wealth of observational data now available in the region. In particular the combination of HF-Radar, Lagrangian drifters, and satellite-derived mesoscale and regional sea surface height and geostrophic currents provide a multi scale cross-validation to support the evaluation and improvement of the numerical models, from the coastal to the shelf and regional scales.

The CARICOOS HF-Radar network is a valuable asset that provides quasi real-time surface currents along the West and South coasts of Puerto Rico. This radar network has a range between 120 km in the West and 250 km in the South. Such coverage allows for the observation of currents over two different flow regimes: tidal-dominated flows in the West and mesoscale-driven flows in the South. Currently the HF-Radar fields are ingested into the U.S. Coast Guard SAROPS system, making it readily available for use as part of their Search and Rescue operations. To incorporate HF-Radar as an additional tool for coastal numerical model validation and development a field experiment was designed to cross-validate the HF-Radar measured currents with current estimates from surface and mixed-layer drifters. A set of six surface (CARTHE-like) and three mixed-layer (NOAA AOML SVP) drifters were deployed 30 km South of Puerto Rico. An additional mixed-layer drifter was deployed 130 km South of Puerto Rico with the aim of it being ingested into a large mesoscale feature and serve as a proxy for mixed-layer regional circulation. The experiment thus allowed to quantify the accuracy of the HF-Radar current fields over a duration of more than a month and over a spatial extent equal to the path lengths of each drifter. In addition, the combination of surface and mixed-layer drifters allowed to determine how surface and mixed-layer currents compare in the region, which is useful for the validation of the vertical structure of three dimensional circulation models and how regional mesoscale features such as eddies affect the coastal waters of Puerto Rico.

Results of the HF-Radar and Lagrangian drifters velocity cross-validation show in general a good agreement between the velocities measured by both instruments (Figure 1). These results lead to the use of the HF-Radar surface currents as a main component of the multi scale validation framework for the CARICOOS circulation models. As an example of the use of the HF-Radar we show the identification and solution of a surface mean flow discrepancy between the ROMS ocean model and the observed HF-Radar currents, as well as an artificial 5-7 day oscillation generated by ROMS but not observed in the HF-Radar observations. As a second example we show the comparison of the mixed-layer drifters tracks as they leave the ROMS domain with the NOAA AOML/CoastWatch satellite-derived mesoscale and regional gridded geostrophic current products (Figure 2). The results give insight as to how mesoscale eddies affect the flow into and out of the main CARICOOS region, and the implications of this regarding numerical model’s domains and needs for initial and boundary conditions and climatologies.