Geology Reference
In-Depth Information
using a laboratory experiment setup. They froze simu-
lated seawater in very thin plastic cells (3 mm thick by
350 mm wide and 500 mm deep), when cooling was
applied only from above. They followed the oscillations
for 8 h and noted that the duration of the downward flow
was much longer than that of the upward flow. The most
regular oscillation period was roughly 1 h, with an
8-15 min inflow of saline water and an approximately
45 min outflow of ejected brine. It should be noted, how-
ever, that in the
Eide and Martin
[1975] experiment, the
cooling was applied from the top against the gravity,
which plays a big role in the desalination process.
Moreover, the experiment was carried out by observing
water between two closely spaced plastic plates to simu-
late the grooves. In this case, surface tension between the
plastic and the water plays a vital role in the shape of the
meniscus and complicates the whole situation.
Very recently,
Turner et al.
[2013] presented a new one‐
dimensional parameterization of gravity drainage based
on mushy layer theory. They solved a set of coupled, non-
linear equations for seaice temperature (enthalpy) and
salinity using an implicit Jacobian‐free Newton‐Krylov
method. Parameterizations were designed to represent
simultaneous rapid drainage occurring in a thin region
just above the ice‐water interface and slower desalination
taking place throughout ice above the skeletal layer. The
rapid desalination is convectively driven and is parame-
terized based on a consideration of flow driven upward
within the mush and downward in chimneys, modified
by the Rayleigh number. The slow desalination is repre-
sented as a simple relaxation of bulk salinity to a value
based on a critical porosity for sea ice permeability. They
showed that the model was capable of reproducing obser-
vations from laboratory and field experiments. This
model, however, does not give any considerations to the
microstructural aspects of the skeletal layer, and conse-
quently cannot be used in making a positive link between
desalination processes and microstructural evolution
including brine‐air porosity.
pocket migration (also called brine diffusion), (2) brine
expulsion, (3) gravity drainage, (4) brine flushing, and (5)
brine mobility through subgrain boundaries. While all
these processes cause brine to drain downward to the
bottom of the ice sheet, the first two mechanisms are
known to cause some brine to go upward to the ice sur-
face as well. The fifth mechanism, as will be seen later,
plays important yet not recognized roles in conjunction
with all the other known mechanisms.
The first four mechanisms are well recognized and ade-
quately described in the literature, based on the prevalent
and widely held notions on sea ice as a cryogenic material
[
Untersteiner,
1968;
Weeks and Ackley,
1982;
Notz and
Worster,
2009;
Weeks
2010;
Hunke et al.,
2011]. The fifth
mechanism has not been recognized by the sea ice research
community, associated with geography and remote sens-
ing, because of two logical yet difficult to comprehend
philosophies. The first and the foremost aspect is the high‐
temperature state of ice in nature. It is not easy to believe
that “ice is hot.” The second is related to the microstruc-
tural details of sea ice beyond the macroscopic views seen
in thick sections in naked (unaided) eyes or traditionally
made thin sections using hot-plate technique and examined
under polarized light. Subsequently, desirable attention
was not paid to the roles played by intergranular and intra-
granular, and hence transgranular, activities. Nonetheless,
physics and mechanics of sea ice related to the substructure
of grain boundaries (intergranular) and subgrains and
their boundaries (intragranular) are well developed as
briefly presented earlier using Figures 2.26 and 2.27. A
brief synopsis of each mechanism is presented next.
1. Brine Pocket Migration
Historically, brine pocket
migration was believed to be responsible for the desalina-
tion of sea ice. This concept was introduced by
Whitman
[1926] and later led to the development of the theory of
liquid zone migration described by
Weeks and Ackley
[1982]. It implies “marching” of pockets through the ice
mass by diffusion. This is caused by the temperature gra-
dient through the ice sheet, which is colder at the top and
warmer at the bottom in winter. Since the brine pockets
are usually elongated in the vertical planes along the
length of columnar grains, the temperature gradient is
also manifested within the brine pockets. It lends itself
into colder and higher salt concentration in the upper end
of the pocket and warmer with lower salt concentration
near the lower end. Consequently, the solute diffuses
from the upper to the lower end. This causes simultane-
ous freezing at the upper end of the pockets and melting
at the lower ends. Hence, all the brine pockets essentially
migrate downward. This trend may reverse in the early
spring when ice temperature may become warmer at the
top of the ice sheet.
Based on the theory of the liquid zone migration
through solid crystals,
Seidensticker
[1965] obtained the
2.3.3.2. Subsequent Slow Salt Rejection from the
Bulk Ice
The brine entrapped within the bulk of ice sheet con-
tinues to be rejected slowly throughout the entire grow-
ing season of FY ice sheet.
Cox and Weeks
[1975]
developed a simple empirical equation for describing
the rate of salinity loss, Δ
S
i
/Δ
t
, describing the rate of
salinity loss, Δ
T
/Δ
z
, given by
S
t
T
z
T
z
5
7
i
16810
.
3 37 10
.
V
(2.37)
b
The slow desalination process occurs as a result of one
or more of the following five mechanisms: (1) brine
