Wind-induced exchange in semi-enclosed basins

The wind-induced circulation over laterally varying bathymetry was investigated in homogeneous and in stratified systems using the three-dimensional Regional Ocean Model (ROMS). For homogeneous systems, the focus was to describe the influence of the earth's rotation on the lateral distribution of the flow with particular emphasis on the transverse circulation. Along-basin wind-stress with no rotation caused a circulation dominated by an axially symmetric transverse structure consisting of downwind flow over the shoals and upwind flow in the channel along the whole domain. Transverse circulation was important only at the head of the system where the water sank and reversed direction to move toward the mouth. The wind-induced flow pattern under the effects of the earth's rotation depended on the ratio of Ekman depth (d) to the maximum basin's depth (h). The solution tended to that described in a non-rotating system as h/d remained equal to or below one. For higher values of h/d, the longitudinal flow was axially asymmetric. Maximum downwind flow was located over the right shoal (looking downwind). The transverse component of velocity described three gyres. The main gyre was clockwise (looking downwind) and occupied the entire basin cross-section, as expected from the earth's rotation and the presence of channel walls. The other two gyres were small and localized, and were linked to the lateral distribution of the along-channel velocity component, which in turn, was dictated by bathymetry. In stratified systems, the main focus was to study the interaction between the wind-driven and buoyancy-induced flow over laterally varying bathymetry. In particular the influence of the earth's rotation and the transverse circulation were examined. The interaction between the wind-induced and buoyancy-driven flow was characterized by, the Wedderburn number ( W_e) which compares wind stress accelerations to baroclinic pressure gradient accelerations. The influence of the earth's rotation was characterized by the inverse of h/d, i.e., the Ekman number. For both rotating and non-rotating systems under strong winds (W_e ≥ 1), the wind-induced pattern of downwind flow over the shoals and up-wind flow in the channel masked any effects of buoyancy driven flows as the water column remained nearly vertically homogeneous. For weak up-estuary winds (W_e $\ll$ 1), the gravitational circulation (with or without rotation) remained almost unaltered. Only the upper part of the water column was modified by the wind-stress, showing an increased surface mixed layer and reduction (increment) of the seaward flow with up-estuary (down-estuary) wind. The non-rotating experiments showed axially symmetric distribution of flow and salinity fields. Rotating cases under weak wind conditions showed low Ekman number, such that rotation effects translated into axial asymmetries of salinity and flow fields. The transverse circulation and the salinity field showed a distribution expected from the balance of the lateral density gradient force per unit mass and Coriolis accelerations. Rotating cases under strong wind conditions exhibited high Ekman numbers and tended to be axially symmetric, responding to a transverse balance between lateral pressure gradient and friction. Transverse flows showed two gyres located over the shoals in response to the lateral density gradient. These results compared favorably with a limited set of observations and are expected to motivate future measurements.