Skip to main content
 Previous Next
  • Zoom In (+)
  • Zoom Out (-)
  • Rotate CW (r)
  • Rotate CCW (R)
  • Overview (h)
Arctic Sea Ice: Encyclopedia Arctica 7: Meteorology and Oceanography
Stefansson, Vilhjalmur, 1879-1962

Arctic Sea Ice

EA-Oceanography
[Harald U. Sverdrup]

ARCTIC SEA ICE

Introduction
The ice in the sea is classified basically as Sea Ice , River Ice , and Land
Ice . The sea ice is formed by the freezing of sea water, the river ice is, as
the name implies, frozen on rivers and is carried out to sea in spring or early
summer, and the land ice represents principally icebergs and pieces of glacier
ice which are set adrift by the calving of glaciers. Of these forms only sea ice
will be dealt with here, being the only form of major importance in the Arctic.
River ice disintegrates rapidly and is never encountered at any distance from
the river mouths, and icebergs are mainly important around Greenland, although
small icebergs also originate from glaciers on Spitsbergen, Franz Josef Land,
Novaya Zemlya, and Northern Land (Severnaya Zemlya).
A systematic discussion of the sea ice is difficult because processes of
freezing and melting, physical properties and appearance depend upon the life
history of the ice and vary within wide limits. A description which should cover
all the numerous aspects would become far too lengthy; certain generalizations
are, therefore, necessary.

EA-Oc. Sverdrup: Arctic Se Ice

Terminology
In recent years a number of attempts have been made to systematize the ice
terminology which during generations has been in use by arctic pilots, hunters,
and explorers and which pertains to the development stage and occurrence of the
ice and to features that characterize the appearance of the surface or the ar–
rangements of the ice.
Terms referring to development stage and features related to freezing and
melting:
Newly-Frozen Ice , subdivided in: Ice Crystals, separate needles or thin
plates of pure ice. Frazil , cinder like crystals formed in moving water. Slush ,
a thin layer of ice crystals which are not cemented together. Snow Slush , slush
formed by snow crystals falling on water at freezing point. Sludge , pieces of
soft ice mixed with slush. Ice Rind , a thin crust of hard ice. Pancake Ice ,
nearly circular pieces of a thin ice formed by the breaking up of a thin ice
crust and collis s ion of the pieces, giving them white, raised rims.
Young Ice , newly formed ice which is 5 to 20 cm. thick.
Winter Ice , ice which is not more than one winter old.
Polar Ice , ice of more than one winter’s growth.
Puddles , depressions in the ice filled by meltwater from a snow cover or
from the surface of the ice.
Meltwater Holes , generally round holes through which the meltwater runs off.
Terms referring to occurrence :
Fast Ice , ice which does not move horizontally, but moves vertically with
the tide. Subdivisions: Winter Fast Ice [: ] (Bay Ice), winter ice which remains
at the locality where it forms. Polar Fast Ice , polar ice which may be carried

EA-OC. Sverdrup: Arctic Sea Ice

toward the coast in the fall and remain there during winter.
Drift Ice , any type of sea ice which moves with wind and currents.
Grounded Ice , ice that is grounded in shallow water.
Anchor Ice (Bottom Ice), ice that is formed on the bottom, generally
around stones or on rocky bottom.
Terms referring to size and surface features, appearance :
Ice Field , compact ice, at least 2 to 3 nautical miles in extent. Ice Floes ,
ranging in dimensions from 30 feet to 2 or 3 nautical miles. Cakes, 6 to 30 feet
large. Bits , less than 6 feet. Floebergs , heavy, hummocked ice more than 30 feet
across and Growlers , small pieces of hummocked ice, lying low in the water.
Flat Ice (Level Ice), ice which never has been disturbed and therefore has
a flat surface, such as winter fast ice.
Disturbed Ice (Pressure Ice), ice which has been subjected to pressure
and may be further described by the following terms: Pressure Ridge , a long
series of raised ice blocks of various sizes. Hummocks , irregular or isolated
ice blocks. Hummocked Ice , ice with numerous irregular piles of ice blocks.
Ice Moutonnee, hummocked ice which has been subjected to melting during summer
whereby the hummocks have become low and rounded. Rafted Ice formed by floes
overriding each other. Tented Ice formed when ice blocks are raised on end,
resting against each other.
Terms referring to arrangement :
Open Drift Ice , small floes or cakes with sufficient open water between
to permit navigation. Close Drift Ice , floes or cakes with so little open water
between that navigation is possible , only by specially built vessels. Compact

EA-Oc. Sverdrup: Arctic Sea Ice

Drift Ice , no open water. Unbroken Ice , term applicable to winter ice only.
Terms like Bay , Tongue , Belt , Strip , Patches refer to features of distribution
and are self-explanatory.
Several of the above terms can be combined for complete description. One
can, for instance, speak of close winter drift ice, mainly composed of small floes
and cakes, or of a tongue of compact, hummocked polar ice.
Freezing of Sea Ice
The freezing point of sea water, t f , decreases nearly linearly with in–
creasing salinity and can be expressed approximately by the equation t f =−0.0545 S,
where S is the salinity in o/oo (parts per thousand). When sea water is being
cooled, certain processes that take place differ from those characteristic of
fresh water. Fresh water has its maximum density at 4°C. Therefore, when a large
body of fresh water of a temperature above 4° in being cooled, [: ]
[: ] the first stop is that the entire mass of water is
cooled to 4°. By further cooling of the surface the cooled water does not sink
because it becomes lighter than the deeper water, and ice freezes when the
surface temperature reaches 0°C. Stirring by wind and waves may produce a
surface layer of temperature 0° before freezing begins, but at greater depths
the water will have a temperature of 4°.
For sea water at salinity greater than 24.7 o/oo the density increase, on
the other hand, until freezing point is reached; consequently a body of sea water
of uniform salinity greater than 24.7 o/oo must be cooled to freezing point
before ice can form on the surface. In general the salinity increases with depth,

EA-Oc. Sverdrup: Arctic Sea Ice

but a surface layer of uniform salinity is often present, and this layer
must be cooled to the freezing point corresponding to its salinity before
ice forms at the x surface. Over the greater part of the Polar Sea the
salinity of the surface layer lies between 30 o/oo and 33 o/oo, and the
corresponding freezing points are about -1.63°C. and -1.80°C.
When freezing occurs, there is first formed a network of elongated or
platelike crystals of pure ice, and as the freezing progresses, this network
becomes more and more dense, whereby some sea water becomes trapped. When the
ice that is formed in this manner is cooled, part of the trapped water freezes
to pure ice, but a brine remains, the concentration of which depends upon the
temperature and always is such that the freezing point of the brine equals the
temperature of the surrounding ice. Thus, sea ice is formed which consists
of pure ice containing numerous small cavities filled by brine. At tempera–
tures of −23°C. or lower some of the salt in the brine crystallizes out.
At the underside of the ice freezing occurs in a similar manner during
winter. The upper surface of the ice will have about the same temperature as
the air, whereas the temperature of the underside remains at the freezing point
corresponding to the salinity of the water in contact with the ice. Because
of this temperature difference heat flows upward, and this heat is supplied by
freezing at the underside. The flux of heat equals the heat conductivity of
the ice multiplied by the temperature gradient in the ice (the rate of tempera–
ture increases with depth). With the same air temperature of temperature
gradient decreases with increasing thickness of the ice; consequently the ice
first increases fast in thickness and later on more and more slowly, other circum–
stances being equal. A snow cover [: ] protects the ice surface effectively,
and greatly reduces the freezing.

EA-Oc. Sverdrup: Arctic Sea Ice

In summer freezing takes place in a different manner in regions where
the ice does not melt completely. Meltwater from the ice that runs down through
cracks, has a temperature of about 0°C. and is practically fresh. It is, there–
fore, lighter than the sea water and spreads directly under the ice. By contact
with the still cold ice masses and the underlying cold sea water it freezes,
but in such a manner that a soft slush of ice forms with large spaces filled by
sea water between the ice crystals. Later in the season this slush slowly
solidifies by the freezing of the trapped sea water, and when this process
goes on, the ice does not appear to increase in thickness. The solidifying may
take several months, so that a noticeable change in the thickness of old ice
floes may not occur until after the middle of December.
Temperature of the Sea Ice
The temperature of ice in contact with sea water always equals the freez–
ing point that corresponds to the salinity of the water. The underside of ice
floes is therefore always at freezing point. The temperature on the top side of
the ice depends, on the other hand, on the air temperature and the snow cover.
Over the Polar Sea the winter precipitation is small and, furthermore, the snow
blows off the large floes and accumulates in drifts along pressure ridges and
hummocks. The large floes are therefore often free from snow or covered by 3
to 5 cm. of hard-packed snow, and in these circumstances the surface of ice floes
attains in winter a temperature that is only a few degrees above the air tempera–
ture. The salinity of the sea water in contact with the ice remains nearly
constant and consequently the temperature at the underside of a polar ice floe
remains practically constant during the entire year, whereas the temperature of

EA-Ic. Sverdrup: Arctic Sea Ice

the surface shows a large annual variation and at intermediate depths a
smaller one. Figure 1 shows the temperatures in an old ice floe throughout
the year and indicates also the changes in thickness of the ice. It is seen
that at the end of August the upper 50 cm. of the ice are at 0°C. and must
therefore be free from salt. Cooling of the surface begins in September and
penetrates rapidly to greater depths, but near the underside an appreciable
temperature gradient is not established before November. Near the surface the
lowest [: ] winter temperature, −30°C., is reached in February, but at a depth of
2 meters the lowest temperature, −10°C., is reached at the end of March. By
the middle of April the heating of the surface temperature rises more rapidly
than that at a greater depth, so that near the surface the temperature decreases
with increasing depth. The figure is based on measurements during the Maud
expedition, which gave results in general agreement with those of the Fram .
They can, therefore, be considered fairly representative of the temperature
conditions in old floes of polar ice.
Salt Content of Sea Ice
The salinity of sea ice can be defined in the same manner as that of sea
water and expressed in per mille (parts per thousand or gram of salt per
thousand gram of ice). The salinity of the ice varies within wide limits,
from 0 to 15 o/oo or more, and depends mainly upon the life history of the
ice, that is, upon the rapidity of freezing and upon the temperature changes
to which the ice has been subjected after freezing. In winter, when ice may
freeze at low air temperatures, the freezing takes place so rapidly that [: ]
much brine becomes trapped, and salinities as high as 15 o/oo may occur. Ice

EA-Oc. Sverdrup: Arctic Sea Ice

which freezes at the underside of a thick floe forms, on the other hand, slowly
and shows salinities of 4 to 6 o/oo. If the air temperature at freezing is
below about −30°C., salt crystallizes on the surface of the ice which then,
even at the lowest temperature, becomes moist and sticky. This ice is very
dark and the salt crystals on the surface look like rime. Travelers should
avoid such ice because the brine penetrates the best l e ather and sledge runners
stick.
When sea ice has been formed, its salinity does net remain constant, be–
cause the trapped brine moves slowly down. This downward motion which leads
to a ff freshening of the top layers has been explained by two processes. In
the first place the [: ] brine-filled cavities in the ice may be so numerous or
so large that the ice can be considered as a porous substance through which the
brine slowly sinks because of its higher specific gravity. This process is
probably of particular importance in spring and early summer when the temperature
of the surface layers of the ice is raised to 0°C. As the temperature approaches
0°C., the cavities in which the brine is enclosed must increase a great deal by
the melting of the surrounding ice because the concentration of the brine must
always be that which corresponds to the concentration at freezing point. When
the cavities become large enough, all brine trickles down, leaving the ice
completely free from salt. Ice blocks which have been raised as a result of
pressure become fresh in summer and render potable water when melted. Similarly,
ponds that late in the season form on the uneven surface of the floes, contain
fresh water.
A second explanation of the gradual downward movement of the brine has been
suggested by Dr. W. G. Whitman. He points out that where the temperature increases

EA-OC. Sverdrup: Arctic Sea Ice

with depth, the concentration of salt in a tiny bubble of brine must tend to
be greatest at the top of the bubble where the temperature is lowest. The
specific gravity will, therefore, tend to be greatest at the top of the bubble,
but such a tratification is unstable, and a slow overturn must take place,
whereby the concentration at the top becomes smaller than that corresponding
to the existing temperature, and at the bottom it becomes greater. To estab–
lish equilibrium some ice must freeze at the top and melt at the bottom, mean–
ing that the entire bubble must be displaced downward, and that the salt content
at any given depth must be altered.
It seems probable that both processes operate. The latter may be effective
during winter and would, for instance, account for the fact that during the
Maud expedition it was observed that at a depth of 125 cm. the salinity in a floe
of winter ice was 7.9 o/oo when the ice was 130 cm. thick, but two months later,
when the ice was 190 cm. thick, the salinity at a depth of 125 cm. had decreased
to 4.3 o/oo. Changes at other depths were probably also caused by the down–
ward movement of brine bubbles. The rapid spring and summer freshening of the
surface layers must, however, be caused by the first processes, the tricking
down of the brine through ice that has become porous. This conclusion is based
on the fact that by the middle of April the surface temperature of the ice has
risen so much that near the surface the temperature decreases with depth.
During late spring and early summer the temperature distribution should, there–
fore, bring about an upward movement of the brine, contrary to what actually
takes place, meaning that in this season the porosity of the ice is of dominating
importance.
When dealing with the temperature conditions it was pointed out that at the
end of the summer the ice was at a temperature of 0°C. down to a depth of about

EA-Oc. Sverdrup: Arctic Sea Ice

50 cm., and that therefore the upper 50 cm. of an old ice floe must be nearly
free from salt. This condition is supported by observations on board the
Maud of the salinity of the upper layers in an old ice floe which showed a
salinity of 0 o/oo at the surface and a salinity of 0.5 o/oo at 60 cm. These
values indicate that one can obtain potable water from the surface layers of
an old ice floe, but this cannot be [: ] recommended because a salt content of a
few g tenths of a per mille may be present, and in the long run even this small
amount is liable to cause trouble. The best procedure is to use ice from old
pressure ridges or ice that has frozen on freshwater ponds at the foot of pres–
sure ridges or large hummocks. Occasionally the water should be tested by adding
a few drops of a weak solution of silver nitrate to a sample. If the sample
turns milky, the salt content is too high.
The Melting of the Ice
From the preceding discussion of the salt content of the ice it is evident
that in a block of sea ice every lowering of the temperature is accompanied by
freezing, and every increase in temperature by melting. In spring, melting there–
fore begins from within when the temperature of the surface layers starts to rise.
The rise of the temperature is caused mainly by absorption of radiation, but
as long as there exists even a thin snow cover, about 75% of the radiation from
sun and sky is reflected, and only about 25% is absorbed. As soon as the snow
has melted and the ice has been exposed, the reflection from the surface is re–
duced to about 50%, and consequently the melting proceeds much faster. The melt–
ing is further accelerated when water collects on the ice in puddles, because
a water surface reflects only 8% of the radiation from an overcast sky. In addi–
tion small particles of mud, pebbles, and pieces of shell absorb most of the

EA-Oc. Sverdrup: Arctic S e a Ice

radiation and accelerate the melting around them. The result of these changes
in the reflective power of the surface is that in spite of the short summer
the thickness of the ice is greatly reduced by melting during the months of
July and August, when the air temperature remains close to 0°C. According to
the measurements during the Maud expedition, the average reduction in thickness
was about 120 cm., and it seems probable that even in the central part of the
Polar Sea an average of about 100 cm. of ice melts in summer. Some of the melt–
water freezes again and forms slush at the underside of the floes (see above), but
the increase in thickness by that process is not considered here.
The melting of the ice is very uneven. Pressure ridges are reduced in
height and assume rounded forms, and melting is rapid where particles occur
or where puddles of water have formed. In some instances puddles may grow to
such dimensions that they are called “lakes,” and may be up to 100 cm. deep.
From the Drifting Pole Station there is even a report of a lake that was 200 to
400 meters wide and 250 cm. deep. In other instances a small puddle may deepen,
a hole may develop clear through a thick floe, and part of the melt water runs
off through this hole. It stands to reason that dark particles and collections
of meltwater accelerate the melting because they reflect less of the incoming
radiation than the surface of the ice, but a closer examination of the processes
has not been carried out.
Physical Properties of Sea Ice
The physical properties of sea ice depend upon the salt content of the
ice and upon the air bubbles in the ice. On the Maud expedition Malmgren
carried out a large number of determinations of various of the characteristics

EA-Oc. Sverdrup: Arctic Sea Ice

of sea ice, and the following information is based mainly on his results.
Specific Gravity . The specific gravity of sea ice is reduced by air bubbles,
but increased if bubbles are filled by brine. It varies between 0.85 and 0.92
whereas the specific gravity of pure ice is 0.9168.
Latent Heat of Fusion . The freezing and melting of sea ice does not take
place at a fixed temperature because of its content of brine. It is therefore
not possible to designate the heat of fusion in the usual manner; instead one may
indicate the number of gram calories needed to melt 1 gram of ice of a given
salinity, assuming that at the beginning of the process the ice was at a tempera–
ture close to 0°C. Thus, if the initial temperature were −1°, the apparent
latent heat would be 80 g.cal., if the salinity were 0 (pure ice), 63 g.cal.,
if the salinity were 4 o/oo and 46 g.cal. if the salinity were 8 o/oo.
Specific Heat . The specific heat of pure ice is about 0.48 (g.cal./g/°C.),
and varies slightly with temperature. The specific heat of sea ice, on the
other hand, varies within wide limits with salinity and temperature, because
changing the temperature of sea ice involves freezing or melting of ice. At
temperatures close to 0° the amounts that freeze or melt at a slight change of
temperature are large if the salt content is higher, and the “specific heat” is
then quite anomalous. At low temperature the specific heat approaches that of
pure ice. These features are illustrated by a few [: ] numerical values:
Salinity, Specific heat at a °C. temperature of:
o/oo −2° −8° −14° −20°
0 .48 /49 .48 .47 .47
4 4.63 0.76 0.57 0.55
8 10.83 1.01 0.64 0.60

EA-Oc. Sverdrup: Arctic Sea Ice

Thermal Expansion . Pure ice behaves as most solids, expanding with
rising and contracting with falling temperature. The coefficient of expansion
is defined as the relative change in specific volume, [: ∝] , ( [: ∆∝] / [: ∝] ) per degree
change in temperature and has a value of 1.7 × 10 -4 . In the case of sea ice the
freezing and melting of the ice that surrounds the enclosed brine leads to
entirely different values, because freezing is accompanied by an increases of the
specific volume (ice has a lower density than water) and melting by a decrease.
Consequently, if the temperature of sea ice is lowered, it will expand if the
increase in specific volume due to freezing exceeds the decrease related to the
contraction of the pure ice. In general one finds that ice of high salinity
expands when cooled if the temperature is relatively high, but contracts at low
temperature. Expansion by cooling is indicated by a negative coefficient of
expansion. These features are illustrated by a few numerical values:
Salinity 10 4 × coefficient of thermal expansion
at a °C. temperature of
o/oo −2° −8° −14° −20°
0 1.7 1.7 1.7 1.7
4 −45.9 −1.4 0.8 1.1
8 −93.5 −4.4 −0.1 0.4
Thermal conductivity . The coefficient of thermal conductivity (g.cal./cm./
sec./°C.) of pure ice has a value of about 5 × 10 -3 . Malmgren’s measurements
show that the conductivity of sea ice is greatly reduced if the ice contains air
bubbles. In old ice floes there are generally more air bubbles near the surface
than at greater depths, and this is reflected in an increase of the thermal conduc–
tivity with depth. Near the ice surface the value was 1.5 × 10 -3 , at 0.5 meters
it was 4 × 10 -3 , and below 1 m. it was about 5 × 10 -3 .

EA-Oc. Sverdrup: Arctic S e a Ice

Vapor Pressure . The vapor pressure of sea ice has not been examined, but
is probably very nearly the same as that of pure ice.
Latent Heat of Evaporation . Under certain conditions ice can vaporize
directly, in which case the latent heat of [: ] evaporation is about 600 gram
calories per gram. In nature it seems that ice first melts and then vaporizes,
in which case the latent heat of evaporation is about 700 gram calories per gram.
Transformations and Seasonal Changes
Winter ice remains undisturbed only in protected areas like bays and fjords
(bay ice). When summer comes, this ice either melts completely, particularly
when fresh water flows out, or it melts near shore while the remainder breaks
up and drifts away.
Toward the end of the summer there are, all over the Arctic, many and large
openings in the polar ice. In the fall, ice freezes over these, but in the spring
one rarely comes across any expanses of winter ice except within protected regions.
The reason for this is that elsewhere the ice is always in motion under the in–
fluence of changing winds, and since the winds never blow with uniform velocity
and constant direction over large areas, the ice is in some areas pressed together
by converging winds, and in others it is term apart by diverging winds. Where
the winds converge the ice floes are squeezed together with such force that the
rims break off and pile up on top of each other, forming pressure ridges which
may rise 4 or 5 meters above the general level of the ice and below which ice may
accumulate to a total thickness of 20 to 25 meters. Where young ice of uniform
thickness is pressed together, the pressure ridges may be fairly regular (rafted
or tented ice), but where polar ice is subjected to pressure, more irregular

EA-Oc. Sverdrup: Arctic S e a Ice

ridges result.
Many [: ] arctic explorers have given vivid descriptions of the impressive
spectacle when huge pressure ridges are formed. They have also described the
many sounds, from creaking or groaning to a thunderous [: ] roar, that accompany
the jamming of the ice. In winter these sounds are often heard over distances of
many miles, particularly in calm weather. The reason for the excellent audibility
is in part that in calm weather there are no other disturbing sounds, and in
part that there always exists a marked temperature inversion which bends the
sound down and therefore makes it heard at long distances from the source. In
summer the inversion generally d si is appears and the sound range becomes much
shorter. If an inversion occurs in summer, it is accompanied by fog, and in fog
the sounds are therefore often little heard.
Where the winds diverge, the ice is term part, and by this process lanes
and leads are formed in any season of the year. In winter a lane that is open–
ing up can be observed from great distances (if the light permits) because of
the frost smoke (fog) from the open water. This frost smoke does not last long,
however, because in intense cold the lane freezes over very rapidly.
One might believe that, when pressure occurs at some later time, the
newly frozen ice over a lane may be squeezed together first, or that cracks will
first appear in the young ice if the ice opens up, but no hard and fast rules
apply. In many instances the ice is broken apart or pressed together along
lines of weakness, but in other cases an old floe may split or be compressed.
The lanes which open up in winter are often fairly straight, and similarly the
continuous pressure ridges may run over long distances, but they will vary in
height. In between there are relatively even ice floes which, however, rarely ,
are more than a fraction of a mile across, and are mostly covered by only a few

EA-Oc. Sverdrup: Arctic Sea Ice

inches of hard-blown snow. But these floes are not level because they are several
years old and may represent pressure fields which have been smoothed by summer
melting, or thick floes which have melted unevenly. Because of the random charac–
ter of the winds, the pressure ridges run in all directions, crossing each other
and forming an irregular network or forming pressure fields which represent masses
of irregular hummocks. On some parts of the continental shelves the tidal cur–
rents may contribute to the movement of the ice, but over the deep Polar Sea they
are too weak to be of any importance. Where the movement of the ice is impeded
by the coast and where the water is deep, enormous pressure ridges may be [: ]
built, reaching heights up to 20 meters.
There also exist regions far from land where prevailing conditions are re–
flected in special features of the ice. One may mention Peary’s “big lane” to
the north of Greenland and Zubov’s suggestion that to the north of Franz Josef
Land, beyond latitude 86°N. or 87°N., there exists a belt of hummocky ice which
[: ] separates the ice originated on the Siberian shelf areas from the ice of the
central Polar Basin.
The opening of lanes, the freezing of young ice, and the compression of
the ice lead to an increase in the thickness of the ice which is much greater than
that caused by freezing only, but no estimates are available as to how great the
increase may be. It will be shown that there exists a limit to the thickness
which relatively flat floes can attain by freezing only, but every flat floe is
surrounded by pressure ridges below which large quantities of ice are accumulated.
At the end of the winter the average thickness of the polar ice depends upon the
proportions of flat floes and pressure ridges, and upon the amount of ice under
the latter.

EA-Oc. Sverdrup: Arctic Sea Ice

The increase in the total ice cover and the increase [: ] in the thickness
of the ice due to freezing and compression do not give a complete picture of the
total winter production of ice in the Arctic. It must also be remembered that
large quantities of ice are carried out of the Polar Sea by the East Greenland
Current. The existing estimates of the amounts of ice that are carried out vary
from 3,000 km 3 . per year (Zubov) to 8,000 km 3 . per year (Wiese), and even 12,700 km 3 .
per year (Krümmel). Zubov states that his estimate is low, but the others are
undoubtedly high, for which reason the correct value may lie at about 5,000 km 3 .
per year. This amoung is probably more than 50% of the total amount that is
added to the ice cover of the Polar Sea by freezing. Assuming that at the end
of the summer an area of 7.5 million km 3 . is covered by ice of an average thickness
of 3.1 m. (taking pressure regions into account) and that at the beginning of the
summer the corresponding values are 8 [: ] million km 3 . And 4 m., respectively, one
finds that the amount of ice in the Polar Sea has been increased by 8,750 km 3 .
With a season of freezing of 9.5 months, the amount of ice dischar b g ed in that
time would have been 4,000 km 3 . And the total annual production would be about
12,750 km 3 .
In summer the lanes that open up do not freeze over because all over the
Polar Sea the air temperature remains at 0° or very slightly above from the middle
of June to the middle or end of August. The lanes which form follow irregular
lines, and the ice fields on both sides move apart and sidewise. Therefore,
when the ice is squeezed together, the saw teeth along the rims of the fields do
not fit into each other, but corner meets corner, leaving smaller or larger open
areas in between. As the season advances, more and more of such openings are
formed, and at the end of July they are so numerous and large that the areas of
openings in the polar ice may represent 5 to 10 per cent of the total areas.

EA-Oc. Sverdrup: Arctic Sea Ice

The reduction of the total ice cover may in part be caused by melting, but
is probably more related to the outflow of ice along the East Greenland coast.
During the 2-1/2 months when melting dominates, about 1,000 km 3 . of ice are
carried south by the East Greenland Current, meaning that the ice cover is re–
duced by an area of about 0.3 million km 3 ., representing about 4% of the total area
of the Polar Sea and adjacent seas.
The Thickness of the Ice. Transport of Material by the Ice.
Winter ice can freeze to a thickness of 150 to 250 cm., depending upon the
length and severity of the winter.
Polar ice attains a thickness which depends on freezing in winter and melt–
ing in summer, and on the compression and telescoping that takes place in winter.
For flat ice floes there exists a limit to the thickness which can be reached by
processes of freezing and melting only. The melting at the surface is rep [: ]
repeated each summer and depends on the length of the summer season, but is inde–
pendent of the thickness of the ice (provided that this is so great that the ice
does not melt completely). The increase in thickness by freezing at the under–
side of the ice depends, on the other hand, on the thickness of the ice and on
the length and severity of the cold season. If the meteorological conditions do
not vary much from one year to another, a balance will be reached, such that the
amount that freezes in winter equals the amount that thaws in summer. The ice
thickness will then show a regular annual variation with maximum thickness when sur–
face melting begins and minimum thickness when freezing temperatures become preva–
lent in the autumn, but the extremes will not vary much from one year to another.
The average thickness depends mosty mostly upon the amount of melting in summer,
and the average thickness decreases.

EA-Oc. Sverdrup: Arctic Sea Ice

In the East Siberian Sea the reduction in thickness by melting amounts to
about 120 cm. ( Maud ), but in the central part of the Arctic the summer melting is
probably less than 100 cm., and the corresponding maximum thickness is greater.
As examples the fallowing values may be quoted. During the Fram expedition
it was found that ice that started freezing over a lane in September 1893 attained
a thickness of 258 cm. on June 6, 1894. Subsequently the following values ware
found:
Date Thickness, cm. Remarks
June 20, 1894 258
July 7, 1894 276 Thickness increased by freezing of
meltwater at underside of ice
Sept. 1894 200
Nov. 11, 1894 208
Dec. 17, 1894 211
May 5, 1895 300
May 30, 1895 303
It is characteristic that a conspicuous increase in thickness did not start
before the middle of December, the reason being that from September to December
the loww of heat from the surface was used for lowering the temperature of the ice
and solidifying the slush that had formed at the under side in summer (see Fig. 1).
The same feature was observed on board the Sedov .
q The thickness recorded above was reached after two winters and does not
represent the maximum thickness that could have been reached. The measurements of
the thickness of that particular floe could not be continued because it was crushed,
but measurements of an old floe [: ] showed an increase from 336 cm. on Nov. 4,
1895, to 398 cm. on May 5, 1896. Therefore the maximum thickness of the relatively
even ice floes in the [: ] neighborhood of the Fram was about 4 meters.

EA-Oc. Sverdrup: Arctic Sea Ice

Measurements daring the drift of the Maud gave a maximum value of 334 cm.
and it was concluded that in the East Siberian Sea the thickness of old floes
varied from 200 to 250 cm. at the end of the summer to 300 to 340 cm. at the end
of the cold season.
On the Sedov , on the other hand, the maximum observed thickness of rela–
tively even ice was surprisingly small, only 218 cm. Zubov attributes this small
value to an increase of the length of the summer season compared to conditions
during the Fram expedition. The Drifting Pole Station was, on the other hand,
established on a floe more than 3 meters thick.
It seems possible that relatively flat ice floes of much greater thicknesses
than mentioned here may be found in other parts of the Arctic, but processes of
freezing and melting cannot be alone responsible for their formation. Extensive
hummocking must have taken place at some time. A field of badly hummocked ice
may attain an average thickness of 10 meters or more, but after it has been exposed
to melting during one or two summers, most of the hummocks have disappeared, leav–
ing a floe with a somewhat irregular surface, but with an average thickness of
about 8 meters. This floe will gradually decrease in thickness, but it will take
6 to 8 years before it is reduced to a thickness of 4 meters. Unusually heavy ice
floes formed in this manner may be found off the Canadian Archipelago and in the
central part of the Polar Sea, but not on the Siberian side.
From all that is known it seems that the ice under pressure ridges rarely
is as much as 20 meters thick. Off the Siberian coast the thickness is generally
lees because there grounded ice is most often found where the depth is 8 to 10
meters. From the region of the Canadian Archipelago a maximum grounding depth of
about 35 meters (20 fathoms) is reported.

EA-Oc. Sverdrup: Arctic Sea Ice

On the Siberian coast fast ice is generally found in winter inside of the
10-meter depth contour. On the open coast this fast ice is mainly composed of
smaller or larger floes of polar ice which ground and become cemented together
during the winter. When the water is shallow, mud, pebbles, and shell freeze
fast to the underside the floe. During summer, when melting takes place, the
floe rises and may be carried away from the coast bringing with it an embedded
layer of various debris. The floe may join the polar drift ice and may be subject
to freezing from below and melting in succeeding winters and summers. After each
winter more ice freezes below the layer of mud and after one, two, or three
years it reaches the surface of the floe some time during the summer. This
process accounts for the appearance of mud, stones, and shelle on the surface
of ice floes which may be located hundreds of miles from the nearest coast. When
the mud and so on become exposed the melting of the ice is greatly accelerated
and in many places holes are formed through which the particles are washed down.
The rafting of debris by the ice results therefore in depositions of near-shore
material at great distance from the source of the material.

EA-Oc. Sverdrup: Arctic Sea Ice

BIBLIOGRAPHY

Malmgren, F. “On the Properties of Sea Ice.” The Norw. North Polar Exped. with
the Maud , 1918-25. Sc.Res . vol. 1, no.5. Bergen, 1927.

Maurstad, A. Atlas of Sea Ice . Geof.Publ., vol. 10, no. 11, Oslo, 1935.

Mohn, H. “Meteorology,” The Norw. North Polar Exped. 1893-96. Sc.Res ., vol. VI.
Christiania, 1905.

Sverdrup, H.U. “The waters on the North Siberian Shelf.” The Norw. North Polar
Exped. with the Maud 1918-25. Sc.Res ., vol. 4, no.2. Bergen, 1905.

Whitman, W. G. “Elimination of Salt from Sea-Water Ice. (Mass. Institute of
Technology) Amer.Jrnl.Science , vol.11, 5th series, no.62, 1926.

Zubov, H. N. “The Drift of the Ice-Breaker Sedov.” Nature , vol. 145, p.533.
London, 1940.

----. The Center of the Arctic . Leningrad, 1940.

Harald U. Sverdrup [Figure]
HomeArctic Sea Ice : Encyclopedia Arctica 7: Meteorology and Oceanography
 Text Only
 Text & Inline Image
 Text & Image Viewer
 Image Viewer Only