Geoscience Reference
In-Depth Information
(b)
(a)
Salt water
source
S
*
Salt water
source
S
*
Narrow
slot
Fresh water
resevoir
at 20°C
Chamber
Fresh water
resevoir
at 20°C
Chamber
Bath temperature
T
Bath temperature
T
*+20
°
C
*+20
°
C
(c)
(d)
Cooling
20 -
Salt water
at 20°C
T
* °C
Q
1
Fresh water
at 20°C
Fresh water
at 20°C
Reservior
Chamber
Spill
Q
2
Salt water
at 20°C
Bath temperature
T
Copper
foam
Plexiglas
*+20
C
°
Q
3
Figure 13.1.
Four laboratory experimental configurations that have produced abrupt thermohaline transitions: (a) driving param-
eters
T
∗
,
S
∗
, and saltwater pumping rate (for a, b, and d) are indicated. (a) box experiment, (b) slot experiment, (c) layered
experiment, and (d) cavity experiment. (Figure adapted from
Whitehead
2009).
experiment is designed to limit the mixing so that layers
in the box can exist. The layered experiment seeks to con-
trol flow in three layers, and the cavity experiment has fully
three-dimensional flow.
As in theory and computer transitions, the abrupt ther-
mohaline transitions that are produced in the experiments
separate two flows that are distinctly different. This is most
clearly seen in shadowgraphs of two flows on either side
of an abrupt thermohaline transition in the slot experi-
ment (Figure 13.2). Different names for the various modes
are found in the literature, but for clarity in this review,
we adopt a single set of names. The first such mode is
flow driven primarily by temperature. The upper panel
shows such a flow, and here it is called the temperature
mode (called the T-mode in
Whitehead et al.
[2003]). This
mode are found if the temperature of the bath above the
reservoir temperature is greater in magnitude than a cer-
tain critical value
T
T
. Essentially, the flow is the same
as the flow when salinity forcing is absent. In the top
panel of Figure 13.2 this mode is shown by a shadow-
graph. Water from the reservoir flows into the chamber
at the bottom of the slot and hot salty water flows out
at the top as indicated by the arrows. The salty water is
dyed, and the shadowgraph indicates that there is a lot of
small-scale turbulent mixing. The mixing is provided by
convection at relatively high Rayleigh numbers. The salt
water injected by the tube is mixed and diluted to such
an extent that the salinity makes negligible contribution
to density. The hot plume rises to the top of the cham-
ber where it exits through the top of the slot. The lower
picture shows the salinity mode (called the S-mode by
Whitehead et al.
[2003]). This can be found if bath tem-
perature is below a second critical value
T
S
. Three layers
exist. The salty water sinks to the floor almost as though
heating were absent and it forms a hot layer of water that
flows out of the chamber at the bottom of the slot. Fresh
water from the reservoir flows into the chamber along the
top interface of the hot salty layer and forms the mid-
dle layer. It is heated by thermal conduction from the hot
salty layer and rises to the top of the chamber accompa-
nied by convection cells. The hot, fresh water then forms
the top layer what exits the chamber through the top
of the slot. Hysteresis happens because the experiments
show that
T
T
< T∗T
S
, consequently
either
flow is found for
