The physical and biological impact of a small island wake in the deep ocean

A primitive equation numerical model is used to systematically investigate wake formations at Cato Island (155degrees32'E, 23degrees15'S) under a variety of realistic flow conditions. The model faithfully reproduces the key features of data obtained in the vicinity of the island under conditions of strong (similar to0.7 m s(-1)) and weak (similar to0.3 m s(-1)) incident currents. For the strong inflow study, a vortex shedding wake is indicated, with an eddy shedding period of approximately 36 h. Interaction between wake and free stream currents produces strong downwelling and upwelling in regions of flow convergence and divergence, respectively. For the weak inflow case, a Lagrangian analysis of wake currents shows strong particle retention properties and vertical pumping in the wake; these results are consistent with observations of nutrient uplift and biological enhancement (the island mass effect) in the vicinity of the island in February, 1993. Numerical sensitivity experiments demonstrate that incident flow speed, background rotation rate and coastal island geometry each have a strong controlling influence on wake formations. Increasing the background rotation rate reduces the frequency of eddy shedding, while disproportionately increasing the circulation strength within shed eddies. For the biologically important non-shedding flow scenario, Lagrangian wake characteristics are examined in detail using the float-tracking scheme of the numerical model. It is found that unsteadiness severely compromises wake retention of passively drifting particles. Coastal geometry also has a strong controlling influence on wake retention. The numerical experiments suggest that particle retention in island wakes has a hair trigger characteristic controlled by incident flow speed and direction. This simple but powerful observation is used as the basis for a new proposal to explain the long-standing recruitment problem of biological oceanography. Good overall agreement between field data and numerical predictions further establishes two-dimensional representations of island topography as a viable and computationally efficient alternative to full, three-dimensional modelling, when the modelled flows are dynamically deep. (C) 2002 Elsevier Science Ltd. All rights reserved.