A hybrid numerical approach was developed to study the dispersion of passive/reactive tracers in the Gulf of Naples (GON). To this end, an Eulerian and a Lagrangian scheme were implemented in the barotropic form of the Princeton Ocean Model (POM) and applied to the dispersion of zoo- and phytoplankton in the GON. The hybrid technique was first validated by comparing the tracer concentration patterns from the Eulerian model and maps of particle positions from the Lagrangian model. Excellent agreement in both spatial distribution and temporal evolution of these quantities was found between the two models. Second, the circulation in the GON was simulated using the POM model. While using simplified forcing fields, the simulated circulation patterns in the GON reproduce many observed features. These include the flushing of the GON waters typically occurring in spring and the formation of a close cyclonic gyre (trapping and homogenizing tracers in the GON) in autumn. The circulation patterns are strongly influenced by both the surface wind stresses and bathymetry and only "remotely" by the Tyrrhenian circulation. For the biological application, the spatial and temporal evolution of passive tracers (e.g., nutrients) was simulated using the Eulerian approach and that of the zoo- and phytoplankton using the Lagrangian approach. These populations were assumed to follow a prey-predator relationship and were studied using a grid resolution of 1.5 km. At these scales, the biological and physical processes (e.g., grazing, phyto- and zooplankton growth rate, mesoscale eddies, horizontal turbulent diffusion), influence plankton heterogeneity and patchiness. In particular, the model results show that phytoplankton variability have spatial and temporal scales similar than those of the carrying capacity (considered here as the effect of a limiting nutrient), yet bigger than the flow turbulence due to diffusion processes. The zooplankton population on the other hand develops on smaller scales due to its longer time taken to mature. The temporal evolution of the two populations shows that they follow a cyclical pattern which is smooth at time when the magnitude of the mean flow is much larger than the amplitude of the turbulent component of the ocean current and "noisy" at time when the turbulent component is larger than the mean flow. The high frequency oscillations are caused by the turbulence of the background ocean current through which the parcels move. A time lag of 20 days between phytoplankton and zooplankton concentrations was simulated, with the zooplankton temporal variability smaller than that of the phytoplankton. These results show that the hybrid Eulerian-Lagrangian technique here implemented is an appropriate tool to investigate the complex interactions between physical and biological dynamics. (c) 2005 Elsevier Ltd. All rights reserved.
905DWTimes Cited:1Cited References Count:43