EM 1110-2-1100 (Part II)
30 Apr 02
current will approach a velocity equal to 2 to 3 percent of the wind speed. But, given the lengths of open
water found in most harbors, it is likely that the resulting current velocity will be less than this magnitude.
(b) The wind-induced surface current would cause a lateral or bottom return flow to develop, resulting
in some circulation in a harbor under sustained winds. The resulting current pattern would be very dependent
on the wind direction and the harbor geometry. Some flow in or out of the harbor entrance could develop.
Also, the surface current will skim floating materials to the down-wind areas of the harbor. Generally,
wind-induced circulation and flushing will be much less effective than circulation and flushing induced by
(3) River discharge.
(a) Fresh water may enter a harbor from the land side as surface runoff or as concentrated flow in rivers
that enter the harbor. Surface runoff will generally contribute more to the harbor pollution load because of
the agricultural or urban pollutants in the water than it will to alleviating pollution by contributing to harbor
(b) Likewise, the efficacy of river discharge in easing harbor pollution will depend on the quality of the
river water. If there are lower chemical and biological pollution loads in the river water than in the harbor,
the flushing effect will be positive. But this could be counterbalanced if the river has a suspended silt load
that will deposit in the slower moving harbor water. Silt suspension is a particular problem inside sections
of the harbor such as pier slips. The silt-laden water may set up a turbidity current that will force bottom
water having higher silt concentrations into a dead area, where it then deposits (Lin, Lott, and Mehta 1986).
The silt load in a river may be relatively low during periods of normal river flow, but the concentration of
suspended sediment may increase by orders of magnitude during storms.
(c) Some harbors are built inland from the coast along river/estuary systems. River flow past the harbor
entrance can have conflicting effects. The flow past the entrance can generate a rotary flow system at the
entrance that produces some flushing of the harbor. But, if the river silt load is significant, there can be a net
deposition of sediment in the harbor near the entrance.
c. Predicting of flushing/circulation. Harbor flushing rates and circulation patterns can be predicted
by numerical and physical models and by field studies. Often, some combination of these efforts is the most
effective approach. Numerical and physical models benefit from the collection of field data that is usually
required to calibrate and verify the models.
(1) Numerical models.
(a) A numerical model of the hydrodynamics of a harbor can be developed by employing finite
difference solutions of the equation of mass balance and the two horizontal component momentum equations.
With the appropriate boundary conditions at the fixed boundaries of the harbor and the tide as a forcing
function at the harbor entrance, one can compute the time-dependent water flow velocities and water surface
elevations at selected grid points in the harbor. If so desired, the surface wind stress on the water can also
be employed as a forcing function. And river flow into the harbor can be specified at points along the harbor
boundary. From the computed time-dependent grid of flow velocities, the resulting circulation patterns in
the harbor can be defined.
(b) Commonly, the two-dimensional depth-integrated forms of the equations are used. That is, the
horizontal flow velocities are averaged over the water depth and it is assumed that the water column is well
mixed so there is no vertical density stratification. Also, it is assumed that vertical flow accelerations are
small compared to the acceleration of gravity, so the pressure is hydrostatic and vertical components of flow