Estuary Processes and Managment - River Dynamics and Integrated River Management - page 517

such integral type jet models. On the other hand, 3D shallow water circulation models are run on grids on
the order of 100 m. Predictions tend to be overly optimistic as the impact is averaged over a region at
least of the order of 100 m. And yet, predictions are often needed in the transition from the near to far
field, or the intermediate field (Fig. 8.52). For example, the impact of a chlorinated wastewater discharge
on a nearby beach located 3 km down current needs to be assessed. In general, the use of either the near
field model or the far field model alone is highly unsatisfactory. Traditionally the near-far field coupling
methods involve either "one-way coupling" or weak “two-way coupling", so the dynamic effects of the
plume mixing cannot be satisfactorily represented in the far field model (Lee and Choi, 2008).
In the near-far field transition, the dynamics depends on the interaction of the near-field plume mixing
and the ambient flow, gravitational spreading, and modifications in ambient stratification. In order to
have a true two-way coupling for effective modeling of mixing and transport in the intermediate field, a
dynamic coupling of the near field model and a 3D water circulation model can be achieved using a
recently developed Distributed Entrainment Sink Approach (DESA, Choi and Lee, 2007). From the
viewpoint of the surrounding water, the near-field flows are the bulk sink flows (“loss”) due to the turbulent
jet entrainment and the bulk source flows (“gain”) due to the diluted discharges (mixed effluent). Therefore,
the plume mixing can be represented as a series of distributed sinks along the jet trajectory. A source term
representing the diluted source (total dilution) flow and the discharged pollution loading (tracer mass
flux) is introduced at the terminal level.
In this coupling method, at any instant time, the ambient current U a ( z ), salinity S a ( z ), temperature T a ( z ),
density ȡ a ( z ), and tracer concentration C a ( z ) upstream of the outfall discharge is provided by the solution
at the grid cells upstream of the source. Based on the ambient conditions, the embedded near-field jet
model (e.g., JETLAG) can be run to compute the evolution of the average properties of the plume elements
along the jet trajectory: average jet velocity and width, tracer concentration, location, density, and the
total fluid mass entrained into the plume element as a result of shear and vortex entrainment, ǻ M k . Based
on the location of the centre of the plume elements, the distributed entrainment sinks, Q e , for each far field
model grid cell can then be computed by summing the entrainment flows corresponding to all the plume
elements within that cell:

§ ·

' M

¦

e

Q

k

(8.15)

¨ ¸

U ' t

© ¹ a

The diluted source flow at terminal height of rise, Q d , is then given by:

¦

d e

QQ Q (8.16)

o

which is applied as a source term to the corresponding far field model grid cell, where Q o is the effluent
discharge flow. Also, a source term representing the discharged pollution loading (tracer mass flux) is
introduced at the terminal height of rise (i.e. the near field mixing is represented as fluid mass source in
the continuity equation and tracer mass source in the tracer transport equation). Thus, a true two-way
dynamic link can be established at the grid cell level between the near and far field models. The coupling
captures the key physical mechanisms of ķ turbulent jet entrainment; and ĸ 3D hydrodynamics with
the hydrostatic pressure approximation in the intermediate field.
The accuracy of this method has been demonstrated for a number of complex 3D near-far field
interaction problems including the interaction of a confined rising plume with ambient stratification,
mixing of a line plume in cross-flow, and impact of an outfall discharge on a nearby beach. For example,
the practically important problem of a finite “line plume” in a cross flow can be accurately predicted by
using DESA. It is clearly shown in Fig. 8.60 that the DESA-predicted surface spreading is reasonably
close to the observation especially in the near-to-intermediate field region where the effect of the rosette