Regions of freshwater influence (ROFIs) - Introduction to the Physical and Biological Oceanography of Shelf Seas - page 279

Withdraw gate Withdraw gate

(a)

(d)

Air

(e)

(b)

(f)

(c)

Air off Air on

Figure B9.1 Illustration of a 'lock exchange' experiment without (left) and with (right) turbulent
mixing generated by bubbling air through the base of the tank. (a, d) Initial laboratory tank
set-up, with high density water (grey) separated from low density water (white) by a barrier. (b, e)
The barrier is removed, allowing two-way exchange. (c, f) As the experiment progresses, for the
case of no mixing the fluid eventually reaches a resting state with stable stratification.

colours added) as seen in Fig. B9.1a . When the gate is removed, a density-driven flow
develops with the light fluid moving over the heavier fluid to create a stratified 2-layer
flow, shown in Fig. B9.1b . The leading edge of the light water at the surface (termed
the 'gravity head') and that of the heavy water at the bed both advance at a speed of

p

g 0 D where g 0 ¼

D

r/r 0 ) is the reduced gravity and D is the layer depth. When the
two gravity heads reach opposite ends of the tank, the flow becomes more complicated
as momentum is reflected back into the tank at the boundaries. Eventually, the motion
dies out due to frictional damping, leaving a stably stratified water column ( Fig.
B9.1c ) with the two layers separated by the small amount of water of intermediate
density produced by the mixing which has occurred during the experiment.
In the modified version of the experiment, the tank is fitted with a false bottom
which allows air bubbles to be blown into the tank, illustrated in Fig. B9.1d . When
the air supply is switched on, the rising air bubbles induce vertical mixing which
opposes the development of the density current and greatly reduces the horizontal
transport of buoyancy ( Fig. 9.8e , f). If the airflow is switched off, the density
gradients accelerate the fluid and re-establish stratified flow. Although the experi-
ments are conducted without rotation, there is a clear analogy with buoyancy stirring
competition in the ROFI regime and they serve as a useful model, helping us to
visualise the processes operating in the full-scale system. One version of the experi-
ment has revealed the interesting effect of varying the modulation period of the
stirring. Switching between bubbles on for T seconds and then off for the same

g(