A NUMERICAL INVESTIGATION INTO THE PHYSICAL PARAMETERS WHICH DETERMINE RESIDUAL DRIFT IN LONG ISLAND SOUND

A vertically averaged numerical model is applied to Long Island Sound (LIS) to investigate selected physical mechanisms which affect residual drift. Particular emphasis is placed on the steady circulation generated by nonlinear interactions of the M(,2) tide and seasonal variation of the wind stress.^ The model, based on the vertically averaged x- and y-momentum equations and the continuity equation, includes the effects of the earth\u27s rotation, the advection of momentum, horizontal friction, wind stress and bottom friction. At the open boundary the sea surface height is specified as a sinusoidally varying function of time simulating the reciprocation of the M(,2) tide in adjacent Block Island Sound. The time-varying output of the model is time averaged to determine the tidal residual flow field.^ In all, three sets of numerical experiments are conducted. In the first, comparison is made between model and analytical solutions for a system of simple geometry, that is, with constant depth and parallel sides. Wind stress, Coriolis accelerations and horizontal friction are neglected. The first-order tidal solutions agree well with analytical results. In the stationary state, the model underestimates the Eulerian residual circulation somewhat, compared with analytical results.^ In the second set of numerical experiments, the tidally driven residual circulation in LIS is investigated. In these experiments, wind stress is neglected and a basin of realistic geometry is modeled. The nonlinear advective terms in the momentum equations are largely responsible for a calculated tidally driven residual current of 5 to 10 cm s(\u27-1) in eastern LIS and 1 to 2 cm s(\u27-1) in central and western LIS. Further numerical experiments suggest that both the bottom topography and breadth variations in LIS significantly affect the residual flow field. The effect of the earth\u27s rotation, on the other hand, is relatively minor. Comparison between numerical results and available data supports the calculated residual circulation although no direct evidence exists to prove conclusively that it is nonlinearly driven.^ Finally, the tides are neglected and the barotropic response of LIS to a seasonal wind stress is investigated. Numerical solutions suggest that only during the winter is the wind-driven component of the residual flow likely to be significant. Vertically averaged wind-driven currents are about 5 cm s(\u27-1) in the shallow water along the Connecticut and New York coasts and 1 to 2 cm s(\u27-1) in the center of LIS. Available data do not strongly support the existence of a significant wind-driven barotropic component of the net circulation in LIS.