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

The Meteorology of the Arctic Region

THE METEOROLOGY OF THE ARCTIC REGION

By
Sverre Petterssen, W. C. Jacobs,
and B. C. Haynes

EA: Meteorology
[Sverre Petterssen, W.C. Jacobs and B.C. Haynes]

THE METEOROLOGY OF THE ARCTIC REGION
Page
Introduction 1
Composition and Structure 4
Inversions and Lapse Rates 19
Acoustic Phenomena 25
Optical Phenomena 37
Air Masses and Fronts 46
Cyclones and Anticyclones 52
Atmospheric Pressure 56
Surface Wind 66
Upper-Air Winds 98
Air Temperature 107
Precipitation, Snowfall, Thunderstorms 164
Humidity 211
Cloudiness and Ceilings 228
Fog and Visibility 264
Sunshine, Illumination 287
Information on Diagram 302
Reference to Literature 304
Legend to Diagrams 308
Note : The diagrams referred to in this paper are of a dimension too
large to be included in the present binding. These diagrams
may be consulted in the Stefansson Collection where they are
filed. (December, 1954)
INTRODUCTION
From the broadest point of view the climatic regions of the
world may be divided into five principal types, vis.,
A. Tropical Rainy Climates , which comprise the tropical rain
forests, the tropical monsoon systems, and the adjoining savannas.
B. Dry Climates , which comprise the deserts and steppes in
subtropical and adjacent latitudes.
C. Warm Temperate Climates , which comprise the part of the
midlatitude rainy belt that is not normally covered by snow in
winter.
D. Snow-Forest Climates , which comprise the mid and high
latitude belt with extensive forests and snow cover during the
winter.
E. Polar Climates , which comprise the tundra regions and the
fields of perpetual snow and ice.
Fig. 1 The distribution of these types of climatic regions in the nor–
thern hemisphere is shown in Fig. 1. Each of these regions may be
divided into sub-regions, depending upon the amount rainfall, seasonal variations, and other limiting factors that affect the
natural vegetation. In this respect the Polar Climate may be said
to be the simplest of all principal climatic types, inasmuch as it
suffices to divide it into two subtypes, namely the Tundra Climate
and the Frost Climate .
The Frost Climate occupies the regions of perpetual snow and
ice, while the Tundra Climate is characterized by bare ground during
the warm season. The vegetation typical of the tundra consists
largely of mosses, lichens and grasses with dwarf trees in sheltered
places. Along its equatorward border the tundra merges with the
vegetation of the snow-forest climatic zone of the northern hemi–
sphere, and the border between these two climatic zones has been
found to coincide very nearly with the line along which the mean
July temperature is 50°F (10°C). Using this isotherm as a criterion,
it is convenient to extend the border between the Polar Climate and
the adjoining regions across the oceans, as shown in Fig. 1.
From a meteorological point of view it is convenient to define the Arctic Region as the region around the North Pole occupied by
Polar Climates (Fig. 1.), excluding the isolated islands of such
climates that occur in certain mountainous regions in lower lati–
tudes. It should be noted, however, the weather conditions typical
of the arctic are normally encountered also in the regions occupied
by the snow-forest climate (i.e., regions D in Fig. 1).
Fig. 2 The normal distribution of the air temperature as a function
of latitude and as a mean for all meridians is shown in Fig. 2.
Using the 10°C July isotherm as the line of demarkation between the
arctic region and the adjacent climatic zones, it will be seen that
the mean position of this isotherm is about 66°N. The area of the
arctic region, as defined above, is therefore about one-twelvth of
the area of the northern hemisphere. Using the correspondence iso–
therm of the warmest month in the southern hemisphere as the border
of the antarctic region, it will be seen from Fig. 2 that this
isotherm is found about 48°S, indicating that the area of the
antarctic region is about three times as large as that of the
[: ] arctic region.
COMPOSITION AND STRUCTURE
Composition of Dry Air . - The air consists of mixture of a number
of gases. Most of these are present in a perfectly mixed state,
with the result that their relative amounts are constant all over
the world, at least up to 25-30 km. (80,000 - 100,000 ft.). The
most important of the constituents are given in Table I, which is
summarized from a recent publication of the International Meteoro–
logical Organization [ 21 ] . The amounts are expressed in terms
of mol. fractions, which for all practical purposes may be taken
to indicate the volume percentage occupied by each gas.
TABLE I. - Principal Constituents of Dry Air .
Nitrogen 78.09 per cent
Oxygen 20.95 per cent
Argon 0.93 per cent
Carbon Dioxide 0.03 per cent
In addition to these principal constituents, there are traces
of Neon, Helium, Krypton, Hydrogen, Xenon, Ozone and Radon, but their amounts are so small that they are of no practical importance.
The amount of carbon dioxide is not quite constant. The vege–
table world continuously consumes carbon dioxide which, again, is
produced by the animal world, through burning of fuels, volcanic
action, and various processes of decay in the soil. Although these
processes are not always balanced, the oceans, by dissolving the
excess of carbon dioxide, so effectively regulate it that no great
variations arise. In view of the absence of local sources, the
amount of carbon dioxide in the arctic is likely to be rather less
than the normal for the atmosphere as a whole.
Ozone, which is present in minute amounts in the atmosphere,
shows a considerable variation with season, latitude and height;
it also varies with the weather situation.
Extensive investigations by Dobson [ 11 ] , Tönsberg [ ], and
Langlo Olson [ 46 ] , Craig [ 10 ] and others have revealed the fol–
lowing broad features of the distribution and variation of the
amount of ozone.
(a) The amount of ozone per unit volume increases with elevation, reaches a maximum value somewhere between 20 and 30 km.
(65 - 100,000 ft.) and then decreases.
Fig. 3 (b) The total amount of ozone (in a vertical air column) has
a pronounced annual variation with a maximum in spring and a minumum
in late autumn (Fig. 3).
(c) The total amount of ozone in middle latitudes varies
aperiodically with the general weather situation, the amount being
larger when the air current is from a northerly direction than when
it is from a southerly direction.
From the foregoing discussion it follows that the composition of
the dry atmosphere in the arctic region is essentially the same as
elsewhere, except that the arctic region is particularly rich in
ozone, and probably slightly deficient in carbon dioxide content. Water Vapor . - The air also contains a variable amount of water
vapor. In many respects the water vapor is the most important
constituent of the atmosphere. The maximum amount of water vapor
that the air can absorb depends entirely upon the temperature; the
higher the temperature of the air the more water vapor can it hold,
the air being saturated with moisture when the maximum is reached.
The amount of water present in the air is conveniently
expressed by the pressure that it exerts. This pressure is usually
expressed in millibars, 1 mb. corresponding to 0.75 mm. or 0.029
inches of mercury under standard conditions.
The maximum amount of water vapor, or the saturation vapor
pressure , various temperatures is given in Table II. Comparing
these figures with the curves in Fig. 2, it will be seen that the
maximum vapor pressure corresponding to the mean temperatures in
the vicinity of the North Pole would be about 0.1 mb. in January,
6.0 mb. in July, and 2.8 mb. as a mean for the year. The amount of
water vapor in the arctic may, therefore, vary by several thousand
percent during the year.
TABLE II. Saturation Vapor Pressure (E, in millibars) at various
temperatures . E w refers to a water surface , and E i to ice surface .
T(°C) T(°F) E w T(°C) T(°F) E w E i
30 86.0 42.4 −5 23 4.21 4.02
28 82.4 37.8 −10 14 2.86 2.60
26 78.8 33.6 −15 5 1.91 1.65
24 75.2 29.8 −20 −4 1.25 1.03
20 68 23.4 −30 −22 0.51 0.38
15 59 17.0 −40 −40 0.19 0.13
10 50 12.3 −50 −68 0.06 0.04
5 41 8.7 −60 −86 0.01
0 32 6.1 −70 −104 0.003
Comparing the arctic region with the equatorial belt, it will
be seen from Fig. 2 Table II that the saturation vapor pressure
in the vicinity of the North Pole is about one-sixth in July, and
about one-four hundredth in January, of the saturation vapor pressure
near equator. Since the air normally is not quite saturated,
the contrasts of the actual amounts of water vapor will be somewhat
less. Nevertheless, the moisture content is expressed in ab–
solute amounts, the arctic region stands out as being excessively
dry during the cold season. This absolute dryness, together with
the low temperature, constitutes an environmental factor of great
importance.
Although the arctic air is dry, on an absolute scale, it is
not so in terms of relative humidity. Let e denote the actual
vapor pressure and E the saturation vapor pressure corresponding
to the air temperature. The relative humidity is then defined as
100 e/E, i.e., the actual vapor pressure expressed as a percentage
of the maximum value at the temperature in question. The distribu–
tion of the relative humidity, as a function of latitude and a Fig. 4 mean for all meridians, is shown in Fig. 4. It will be seen that
the relative humidity in the arctic is normally about 10 per cent
higher than in middle latitudes, and about 5 per cent lower than
in the equatorial belt.
Impurities. - Apart from the above-mentioned gaseous constituents,
the atmosphere contains a variety of impurities, such as dusts, soots
and salts.
The main source of dust is the dry climatic regions (Fig. 1).
The coarser material is never carried far from its source, but minute
dust particles are readily kept soaring by the turbulent motion and
carried long distances from their place of origin by the general
air currents. Before dusty air from the arid regions arrives in
the arctic, it will normally have been cooled so much that conden–
sation has occurred; the precipitation of water from the clouds
washes out the dust to a very large extent, with the result that
the air in the arctic region is particularly free of dust. This
is especially true in winter when the rate of cooling of the north–
ward moving air masses is largest.
The industrial regions and forest fires constitute the main
source of soot*, but since these sources are far removed from the
arctic, the impurities will normally have been removed, either
through precipitation or sedimentation, before the air arrives in
the arctic.
1
Observations show that the air normally contains considerable
amount of salts. Through the action of the winds, spray is whirled
up from the oceans, and when the spray droplets evaporate the salt
remains in the air. These minute salt particles constitute highly
effective nuclei of condensation, and are, therefore, washed out
of the air through the precipitation processes.
The arctic air masses are, therefore, characterized by ex–
tremely low values of turbidity, and this influences the visual
range very greatly. In the arctic, the sky, when clear, is
characterized by the brilliance of stars during periods of darkness, and by intense blueness during periods of light or dusk. Distant
objects (e.g., mountain ranges) stand out with great clarity in
shape and detail.
The purity of the arctic air is noticeable even in low and
middle latitudes when these regions are invaded by arctic air masses.
As has been shown by Bergeron [ 3 ] , the opalescent turbidity can
be used as a means of indentifying traveling air masses, and this
technique has been of importance in the development of the methods
of weather analysis and forecasting.
Troposphere and Stratosphere . - Although the state of the atmosphere
is subject to incessant variations, the mean (or normal) state indi–
cates a division of the atmosphere into fairly well-defined layers.
This stratification of the atmosphere is not immediately apparent
in the distribution along the vertical of atmospheric pressure and
density, but it stands out clearly in the distribution of tempera–
ture.
Fig. 5 Some typical examples of the distribution along the vertical
of the air temperature are shown in Fig. 5, which is reproduced from a recent publication of the Canadian Meteorological Service [ 27 ] .
Disregarding for the moment the conditions near the earth’s surface,
it will be seen that the temperature decreases with elevation at a
fairly regular rate of about 60°C per km. (10°F per 3,000 ft.) up to
about 8 km. (26,000 ft.) in winter and to about 10 km. (33,000 ft.)
in summer. At higher levels the temperature is either constant or
increases slightly with elevation.
The lower part of the atmosphere, in which the temperature de–
creases with elevation, is called the troposphere , and the upper
part, in which the temperature is constant or increases with eleva–
tion, is called the stratosphere . The transition from the tropo–
sphere to the stratosphere, which usually is quite distinct, is
called the tropopause .
The examples shown in Fig. 5 represent the conditions on the
fringe of the arctic. In the central part of the arctic region
still lower temperatures would be observed, particularly in the
lower half of the troposphere.
Fig. 6 Fig. 7 The mean thermal structure of the atmosphere up to 20 km.
(62,000 ft.) above sea level is shown in Figs. 6 and 7 for winter
and summer respectively. These diagrams represent the mean con–
ditions for 18 meridional sections at intervals of 20 degrees
starting from Greenwich meridian. The observational material used
for the construction is that contained in “The Normal Weather Maps”
[ 47 ] . The northernmost parts of the diagrams are largely based
upon extrapolations and tests for consistency. Although this may
have led to errors in detail, there can be little doubt that the
diagrams represent the essential features of the thermal structure.
Considering first the conditions in January, it will be seen
that the mean positions of tropopause, which is found at about
8 km. (26, 000 ft.) in the polar region, rises slowly southward to
about 50°N, and then rises at a rapid rate to about 25°N, where it
becomes horizontal at about 17 km. (56,000 ft.).
It will be seen from Fig. 6 that there are five regions in the
atmosphere (below 20 km.) which are characterized by extreme tempera–
tures. The coldest region is found at about 17 km. (56,000 ft.) above sea level in the equatorial belt, where the mean temperature
is about −75°C (−105°F). The next coldest region is found in the
vicinity of the arctic tropopause, where the mean temperature is
about −63°C (−81°F). The third coldest region is found at the sur–
face over the arctic fields of snow and ice, where the mean tempera–
ture varies between −25 and −41°C (−13 and −42°F). This lower cap
of cold air is separated from the upper cold region by a layer of
relatively warmer and fairly uniform air with temperatures in the
vicinity of −26°C (−13°F) at about 2 km. (6500 ft.) above the ice.
In contrast to these cold regions we find two warm regions,
one at low levels near the equator, and a second in the troposphere
in subpolar latitudes.
In summer (Fig. 7) the conditions are largely the same as in
winter, except that the cold regions in the arctic are less distinct.
In all seasons, the temperature in the troposphere decreases north–
ward, whereas in the stratosphere the temperature decreases, on the
whole, from the arctic toward the equator.
Fig. 8 The annual variation of temperature, as a mean for all meridians,
is shown in Fig. 8. It will be seen the maximum variation occurs
at low levels in the arctic. This variation decreases rapidly with
elevation and reaches a minimum of about 20°C (36°F) at about 3 km.
(10,000 ft.) above which level there is a slight increase up to
about 5-6 km. (16,000-20,000 ft.) and then a rapid decrease up to
10 km. (33,000 ft.) which is the mean summer position of the tropo–
pause. In the stratosphere the annual variation is relatively small.
Considering the conditions level for level, the annual variation of
temperature of the free atmosphere decreases from the pole to the
equator. The same is true of the conditions at the earth’s surface
if one considers the mean for all meridions It should be noted,
however, that the annual variation near the earth’s surface is
larger in Northern Russia, Siberia, and Canada than it is at the
pole (see p. ).
At heights greater than those shown in Figs. 6-8, ordinary
observations are so sparse that the meridional structure and annual variation cannot be evaluated with much confidence. From observa–
tions of meteors and sound waves and from a few direct observations
by rockets and sounding balloons it is possible to piece togather
a picture of the broad features of the uppermost atmosphere, and
these may be summarized as follows. The stratosphere is almost
isothermal up to about 35 km. (21 miles); above this level the
temperature increases rapidly and reaches a maximum of about 75°C
(167°F) at a height of about 60 km. (37 miles) whereafter it de–
creases to about −25°C (−13°F) at about 80 km. (50 miles). This
warm layer is sometimes called the mesosphere.
Above the mesosphere lies the ionosphere which extends up to
great heights and merges gradually with empty space. The ionosphere
is characterized by free electric charges. Some of the gaseous par–
ticles are broken down into ions and free electrons by absorption of
the ultraviolet radiation from the sun, which also causes a dissocia–
tion of oxygen and nitrogen molecules into their atomic forms and
causes very high temperature at extreme heights. The ionosphere is the abode of the aurora borealis; it is divided into several
layers that reflect radio waves in various wave lengths.
INVERSIONS AND LAPSE RATES
Fig. 9 Although the temperature normally decreases with height in the
troposphere as a whole, the lower part of the arctic region forms
an exception (see Fig. 6). Here the temperature normally increases
from the earth’s surface up to a distance which rarely exceeds 2 km.
(6,000 ft.) and sometimes may be as low as 200 m. (600 ft.) or less.
Some examples are shown in Fig. 9.
The rate at which the temperature decreases with elevation is
called the lapse rate . A layer through which the temperature in–
creases with elevation is called an inversion , and such layers are
characterized by counterlapse . The base of the inversion is the
level where the counterlapse commences, and the top of the inversion
is the level where the counterlapse changes into a lapse of tempera–
ture.
It is convenient to compare the observed lapse or counterlapse
with the adiabatic lapse rate, which is the rate at which a unit of
air would cool if it were thermally isolated and lifted against the
gravitational force. The adiabatic lapse rate, which is a critical value for many processes, is expressed by the formula
<formula>Γa = g/Cp</formula>
where g is the acceleration of gravity and Cp the specific heat
of air at constant pressure. Substituting the numerical values for
g and Cp, it is found that <formula>Γa =1°C per 100 m. = 5.5°F per 1000 ft.</formula> for nonsaturated
air.
Above the top of the inversion the lapse rate is normally about
1/2 to 2/3 of the adiabatic rate. Occasionally, the adiabatic rate
may be approached, but it is never exceeded by any appreciable
amount.
In the inversion layer, the counterlapse may be very large;
numerical values as high as 5°C per 100 m. are quite common, and
close to the snow surface values as high as 1°C per meter are not
uncommon, particularly in calm and cloudless conditions in winter.
In calm air or when the winds are light the base of the
inversion is found at the earth’s surface. However, when the wind Fig. 10 is sufficiently strong, friction along the earth’s surface causes
the lower layer to be mixed, and a normal lapse rate is established
in the lower layer while an inversion may be present at some distance
above the surface. These elevated inversions are usually less
intense than the ground inversions. Sverdrup [ 43 ] investigated
the occurrence of inversions by the aid of kites carrying instruments.
Since kites could be used only when the wind speed was sufficiently
high, his results (Fig. 10) apply to elevated inversions. It will
be seen that the inversion is lower in winter than in summer.
Fig. 11 The intensity of the inversion increases with decreasing could
cover, and the most intense inversions occur after spells of calm
and clear weather. An example of the dependence of the inversion
on wind speed and cloud cover is shown in Fig. 11, which is reproduced
from a recent publication by the Canadian Meteorological Service
[ 27 ] .
The inversions are most strongly developed over land and ice.
When the arctic air streams over open water of appreciably higher
temperature (e.g., in winter), the inversion is destroyed through heating of the surface layer. In such cases, an adiabatic, or
even superadiabatic, lapse rate develops above the water surface.
The same is true of arctic air that invades warm continents in
summer.
From the foregoing discussion it follows that the central
arctic is characterized by a well - developed inversion layer, and
that along the fringe of the arctic extreme variations occur, with
changes from large counterlapses to the adiabatic or superadiabatic
lapses.
The processes leading to the formation and maintenance of the
arctic winter inversions have been investigated by Petterssen [ 32 ]
and Wexler [ 53 ] . These involve the general circulation of the
atmosphere and the radiative and eddy flux of heat.
The snow surface, being an efficient radiator, will lose heat
toward space, and the air in contact with the snow will cool faster
than the air aloft. The snow surface will, therefore, act as a cold
source relative to the overlying layer of air. Since the atmospheric pressure over the arctic is higher than over the adjacent oceans
(heat sources), the distribution of heat and cold sources is such
as to constitute a hindrance to the circulation, and a layer of
stagnant air develops over the arctic snow and ice fields. Since
the air is stagnant, it becomes subjected to continued cooling from
below. As the surface layer becomes very much colder then the over–
lying air, downward radiative flux of heat will tend to balance
the cooling of the ground. The maximum difference in temperature
between the top and the base of the inversion, which depends upon
the contents of moisture and carbon dioxide of the air, has been
determined by Wexler to be about 30 °C ( 54 °F), and this
value agrees well with observation.
In summer the conditions are largely similar to those in winter,
except that the cold source is due mainly to the melting of snow,
while the air at higher levels is heated by radiation.
If the wind is sufficiently strong, the radiative cooling of
the surface layer will be offset by the downward eddy flux of heat, and the inversions become weaker, or may disappear temporarily and
locally. If the sky is cloudy, the back - radiation from the clouds
will have a similar effect.
The arctic inversions are of great importance in many ways,
notably in connection with propagation of sound and light.
ACOUSTIC PHENOMENA
No one who has lived in the arctic can have failed to observe
the frequent occurrence of supernormal audibility and the wide
variation in the audible range. For example, Captain Perry [ 35 ] ,
on his third voyage, noted a case where conversation was carried on
over a distance of 1.2 miles, and Collinson [ 9 ] reported on a
case where spoken words were heard at a distance of 2 miles. The
most extraordinary case of abnormal sound effects in the arctic is,
perhaps, the one described by Wegener [ 52 ] . On the Danish Green–
land expedition, 1907-08, observers at Pustervig, on the northeast
coast of Greenland, heard a tone of deep pitch (estimated at about
30 c.p.s.) which lasted for several hours and appeared to emanate
from a closed fjord called Dove Bay. This sound was heard on several
occasions when the fjord was filled with cold stagnant air.
These abnormal sound effects can readily be explained by
reference to the structure of the arctic atmosphere and the properties
of the snow and ice.
The range at which sound can be heard depends upon the temperature of the air, the speed and direction of the wind, and the rate at
which sound energy is absorbed by the earth’s surface.
1. Influence of Snow and Ice . - It is well known that soft
snow falling through the air absorbs sound energy very effectively.
The same is true of soft snow on the ground. On the other hand,
a hard crusted snow surface absorbs but little energy, and a smooth
ice surface is an almost ideal reflector of sound. The rate at
which sound energy is absorbed depends upon the pitch. Kaye and
Evans [ 22 ] measured the absorbtion coefficient of newly fallen
snow in England and found the values reproduced in Table III. It
TABLE III. Absorbtion Coefficient of Newly Fallen Snow .
Snow depth
inches
Frequency (c.p.s.)
125 250 500 1000 4000
1 0.15 0.40 0.65 0.75 0.85
4 0.45 0.75 0.90 0.95 0.95
will be seen that for a pitch higher than 500 cycles per second, a
snow cover 4 inches, or more, deep absorbs almost all sound energy.
Although comparable figures for hard snow surfaces are not available, it is evident that the absorbtion coefficient decreases rapidly
with the hardness, and is almost negligible for a smooth ice sur–
face. The audible range will, therefore, be short over a soft snow
surface, relatively large over hard snow, and excessively large
over ice fields.
2. Influence of air temperature . - One of the major causes of
the supernormal audible range in the arctic is due to the distribution
of temperature, and in particular to the inversion layers described
in the foregoing section. Let C denote the speed of propagation
of the sound, and T the absolute temperature of the air. In still
air, the velocity of sound is proportional to the square root of the
absolute temperature. We may, therefore, write
<formula>C = A√(T)</formula>
where A is a constant for any given composition of the air. The
minor variations in composition, discussed in a foregoing section,
are too small to have any noticeable effect on the speed of propa–
gation
Although the air temperature may very vertically as well as
horizontally, the latter variation is usually negligible in com–
parison with the former, and as shall here be concerned to discuss
only the influence due to the variation along the vertical.
We consider first the idealized case when the temperature is
uniform in all directions (isothermal conditions). The speed of the
sound would then be uniform, and the “sound front” would be a
spherical shell expanding with a constant speed.
Fig. 12 Instead of the “sound front” it is more convenient to consider
the “sound beams” or “sound rays”. These are represented by lines
originating in the sound source and being everywhere perpendicular
to the sound front. The sound rays in an atmosphere of uniform
temperature are shown in Fig. 12A, where the sound source is at
the earth’s surface. The rays are straight lines through the source.
Since the energy of a sound impulse is distributed uniformly on a
spherical surface, it is evident that the sound intensity must be
inversely proportional to the square of the distance from the
source, or <formula>I = I1/R2</formula>
where I 1 is the intensity at unit distance from the source, and
I is the intensity at the distance R from the source. In the fol–
lowing, we shall refer to eq. (2) as the inverse square law.
Let us now consider the case when the temperature decreases
along the vertical, as it normally does in middle and low latitudes.
The sound will travel faster in the horizontal than in the vertical
direction. The sound front will no longer be spherical, and the
sound rays will be curved upward as shown in Fig. 12B. The beams
that leave the source horizontally will lose contact with the earth’s
surface, and in the space below these beams, a sound shadow will be
found. This shadow refers to the beams, or the direct sound. A
certain amount of sound is, however, diffracted across the beams
into the shadow, but the intensity of this sound is small and it
decreases at rate which exceeds the inverse square law.
The conditions represented in Fig. 12B being typical of middle
and low latitudes, it is evident that most people’s experience about sound from distant sources is based upon the rather faint
sound which is diffracted into the beam shadow. Above the beam
shadow, the beams are more concentrated in Fig. 12B than they are
in Fig. 12A, with the result the intensity of the sound is corre–
spondingly increased. It will, thus, be seen that when the tem–
perature decreases along the vertical, the sound tends to escape
upward, and but little energy is transmitted along the earth’s
surface.
As was shown in the foregoing section, inversion layers are
almost always present in the arctic. We shall, therefore, consider
this case in some detail. Since the temperature increases upward
through the inversion layer, the sound will travel faster vertically
than horizontally; the wave front will now be elongated upward, and
the sound beams will be curved downward. Fig. 12C shows the rays
from a sound source ( ) at the earth’s surface when tempera–
ture distribution is as shown to the right of the beams. It can
easily be shown that a beam that leaves the source at certain
critical angle will become tangent to the top of the inversion layer where it splits, one branch (b) being curved downward and
the other branch (c) being curved upward. This critical angle
depends entirely upon the temperature difference between the top
and the base of the inversion, and is independent of the depth of
the inversion layer. The space between the beams b and c in
Fig. 12C is silent as far as direct sound is concerned.
The beams that leave the source at angles less than the critical
value, will be refracted toward the earth’s surface. A considerable
portion of this sound is, again, reflected from the earth’s surface,
and this together with sound that is diffracted across the beams
will penetrate into the part of the shadow that is below the top
of the inversion. On the other hand, beams that leave the source
at angles greater than the critical value, will penetrate the in–
version and escape into space.
Referring again to Fig. 12C, it is of interest to note that
the concentration of the beams is larger in the inversion layer and
less above this layer than in the radial case shown in Fig. 12A. From this it follows that the sound intensity below the top of the
inversion decreases more slowly than indicated by the inverse square
law; above the top of the inversion, the reverse is true.
We shall next consider Fig. 12D which illustrates the con–
ditions when the sound source is above the top of the inversion.
The inversion layer will now act as a hindrance to the propagation
of sound toward the earth’s surface. Except where the sound source
is directly overhead, or nearly so, very little sound energy reaches
the earth’s surface. Thus, an aircraft flying above the top of
the inversion is not readily detected by acoustic means.
From the foregoing discussion it follows that an inversion
layer acts as a duct for sound emanating from sources below its top,
and as a cushion against sound that emanates from sources above its
top. Neither the duct nor the cushion is perfect, and their
efficiency (in still air) depends upon the intensity of the inver–
sion.
The sound intensity may become greatly supernormal when the
sound source is situated below an inversion in a fjord (or valley) surrounded by steep walls. If the fjord is frozen and the mountain
sides covered by hard snow, an almost ideal sound channel is estab–
lished, sound being reflected from the ice, the mountain sides and
the top of the inversion. If the fjord has a local contraction, a
basin is formed which, when the dimensions are suitable, may form
a resonant box. The case described by Wegener (loc. cit.) apparently
belonged to this category of sound effects.
Although the mean state of the lower arctic atmosphere is
characterized by one inversion layer (see Fig. 6), multiple inver–
sions occur quite frequently, particularly over and near arctic
land masses (e.g. Greenland). The sound effects associated with
multiple inversions are extremely complex, and several zones of
shadow and zones of maximum intensity may occur, depending upon the
position of the source. Fig.12E shows, as an example, the sound
pattern of a source situated between two inversions. It will be
seen that the sound tends to become trapped between the top of the
lower and the base of the upper inversion, and that several shadow
zones may result.
3. Influence of wind . - If V denotes the speed of the wind,
the velocity of sound can be expressed by the formula
<formula>C = A√(T) + V</formula>
which is the same as the velocity in still air plus the velocity of the
medium through which the sound travels. Now the former of these
velocities is of the order of 300 m/sec. (700 mph) while the latter
is of the order of 10 m/sec. (20 mph). The direct influence of the
wind is, therefore, very small if the wind is uniform in all direc–
tions.
Owing to friction along the earth’s surface, the wind increases
with elevation up to about 500-1000 m. (i.e., 1500-3000 ft.). Al–
though the increase varies with the roughness of the ground, the
wind speed over a snow surface will normally be twice as large at
about 600 m. as is at 10 m. above the ground. Above this layer,
which is called the friction layer, the wind may increase or de–
crease with elevation depending upon the horizontal temperature
gradient.
The variation along the vertical of the wind has a marked influence on the propagation of sound. To demonstrate the nature
of this influence, we consider Fig. 12F, in which it is assumed
that the temperature is uniform along the vertical, while the wind
distribution is as indicated to the left. The beams that go down–
wind will be curved toward the earth’s surface. A beam that leaves
the source at a certain critical angle, will just touch the level
where the wind becomes uniform, and at greater distance from the
source, a sound shadow will be found below this level. The beams
that go upwind will be curved away from the earth’s surface, above
which another sound shadow is found. The greatest concentration of
sound beams is found in the downwind direction in the layer where
the wind increases, and it is here that the supernormal audibility
is observed.
In the arctic both the wind and the temperature will normally
increase with elevation through the friction layer, with the result
that both effects combine to give supernormal sound intensity down–
wind. In the upwind direction, the temperature effect is counter–
acted by the wind effect, and except when the wind is very light, the wind effect predominates.
4. Sound ranging . - From the foregoing discussion it follows
that for any given source intensity the audible range depends upon
the curvature of the sound beams in the vertical plane, and this
curvature is determined by the distribution along the vertical of
temperature and wind. Provided that soundings of temperature and
wind are available, the path of the sound beams can be reconstructed
and the position of the sound source identified. A convenient
method of sound ranging has been developed by Bedient [ ] .
For further information on propagation of sound in the atmosphere,
reference is made to the works of Wa e lchen [ 50 ] , Rothwell [ 36 ] ,
Whipple [ 55 ] , Gutenberg [ 15 ] , and Saby and Nyborg [ 37 ] .
OPTICAL PHENOMENA
In addition to the aurora borealis, the abode of which is in
the ionosphere (see pp ), the sojourner in the arctic
will observe a number of optical phenomena of great beauty and in–
tensity. Some of these, such as the rainbow, the corona and the
halo, are not essentially different from those observed in middle
latitudes and will not be described here. The optical phenomena
which are most typical of the arctic and of some importance to the
arctic traveler are the mirages which are due to abnormal bending
of the light rays, and the ice blinks and the water sky which are
due to reflection of light from ice and water surfaces by the lower
face of a cloud layer.
The mechanism of the formation of mirages is readily explained
by reference to the fact that light travels slightly faster in thin
air than it does in denser air. Thus, since the air density decreases
with elevation (except in very rare cases), a slant beam of light
will be curved downward, and this curvature depends upon the rate
at which the density decreases across the beam. In the following we shall be concerned to discuss the bending of light beams between
points on the surface. Since these beams are quasi-horizontal it
suffices to consider the lapse of density along the vertical.
Using the equation of state and the hydrostatic relationship,
it is readily shown that the rate of decrease of density (p )
with height (z) is expressed by
<formula>-(∂p/∂z) = (p/RT2)((g/R) – Γ)</formula>
where p denotes pressure, T absolute temperature, Γ lapse rate
of temperature, R the gas constant, and g the acceleration of
gravity.
Since g and R are physical constant and p varies but
little in any given place, it will be seen that the lapse rate of
density (and the refractive index) is determined almost exclusively
by the temperature conditions.
Travelers in the arctic have noticed a marked annual variation
in the optical phenomena, and this can readily be explained by
reference to equation (1). Let us assume for the moment that the lapse rate of temperature is the same in winter as in summer.
Under typical arctic conditions the absolute temperature would be
about 275°A in summer and about 230°A in winter. It is then readily
seen that the refractive index is normally about 40 per cent greater
in winter than in summer. In addition to this effect of the annual
variation of temperature there is a large annual variation in the
lapse rate of temperature (see Fig. ), with the result that
the refractive index may vary several hundred per cent during the
annual cycle. In fact, the largest variations are due to the change
in lapse rate of temperature.
As was shown in a foregoing section (p. ) the temperature
of the troposphere normally decreases with elevation such that
= 0.6°C per 100 meters (or 3.3°F per 1000 ft.), and this
together with the first term within the parentheses of eq. (1)
accounts for the normal refraction of light in the atmosphere. In
the arctic, however, the lapse rate may vary within very wide limits,
thus giving rise to abnormal bendings of the light beams.
1. The superior mirage occurs in connection with temperature
inversions (p. ). In the inversion layer the temperature in–
creases with elevation, and the lapse rate of temperature (i.e., )
is negative. It will then be seen from formula (1) that the den–
sity decreases along the vertical at an abnormally fast rate, with
a consequent abnormal downward refraction of the light beam. An ob–
ject seen through the inversion layer will become distorted so that
it appears elongated in the vertical direction. For example, a
relatively flat strip of coast land may appear as an erect strip
and give the false impression of being a steep cliff; irregulari–
ties in the coast line will appear like columns, and the distor–
tions produce a picture which resembles architectural pseudo–
prostyle. In pronounced cases, the erect image is surmounted by
an inverted image, and in rare cases the inverted image is, again,
surmounted by second erect image.
Fig. 13 An example of a superior mirage is shown in Fig. 13. The
upper picture shows the natural shape of Gundahl’s Knold while the
lower picture shows strong vertical distortion due to the presence of an intense inversion.
The superior mirage may occur without inverted image and dis–
tortion, in which case objects which are actually below the obser–
ver’s true horizon will appear above his apparent horizon. This
phenomenon is called looming. This, together with the pronounced
purity of the arctic air (p. ), probably accounts for many
instances of erroneous estimates of distances and reports of dis–
covered land masses and mountains in places where none exist. In
1818 Captain John Ross saw snow-covered peaks in Lancaster Sound
(74°N, 85°W), at an estimated distance of thirty miles, which appeared
to bar his way into the Northwest Passage. Subsequent explorations
have made it evident that the peaks seen by Ross were those of
North Somerset Islands (73°N, 93°W) at a distance of about 200
miles. Pearly and his companions clearly saw on two occasions in
1906 an extensive mountainous snow-covered land northwest of Cape
Colgate (82°N, 91°W) in Grant Land at an estimated distance of 130
miles, and named it Crocker Land. In 1914 MacMilland and Green sighted Crocker Land and sledged 130 miles in its direction, seeing
the mountains once on the journey but never reaching them. The
exact location of Crocker Land is still a mystery, and it appears
certain that the sightings were caused by looming of unusual in–
tensity.
The distance over which an object may loom depends upon the
height of the inversion and the difference in temperature between
the top and the base of the inversion. In calm and cold weather
the inversions are likely to be deep and intense; distances esti–
mated in such conditions are likely to be much in error.
It is of interest to note that looming is a local phenomenon;
at some point closer to the loomed object than the observer, the
object will not be visible. The effect may be likened to the
“skipping” of radio waves which permits reception close to the
transmitting station and at considerable distance from it, but not
at intermediate distances. As a consequence of this, it is readily
seen that for an aircraft that sees a loomed sun, there must be
points both above and below the aircraft for which the sun will Fig. 14 have set, and twilight will prevail both above and below an
aircraft illuminated by the loomed sun. These conditions are shown
diagrammatically in Fig. 14.
2. The inferior mirage forms when the temperature decreases
with elevation at an excessive rate. These mirage, which are
common occurrences is southern deserts, may be observed locally in
the arctic, particularly where cold air from the ice fields is
heated by streaming over open water, or where bare land adjacent
to ice, is heated in sunshine. The superheated layer is usually
quite shallow (4 to 10 ft.), and within it the lapse rate of tem–
perature is large and positive. It will be seen from formula (1)
above that the density decreases with elevation at a subnormal rate;
in extreme cases when [Math Formula] is greater than g/R, the lapse of
density is reversed, and the light beams may be curved away from
the earth’s surface. In such cases, the true horizon disappears,
and an apparent horizon is formed below the true horizon leaving
a gap between the apparent horizon and the inferior mirage of objects Fig. 15 above the true horizon. An example of an inferior mirage is shown
in Fig. 15.
As with the superior mirage, the inferior mirage results in
a change in the apparent distance of objects. In the case of in–
ferior mirage, the effect is that of disappearance over the apparent
horizon of previously seen objects which are known to be above the
true horizon. This phenomenon is called sinking.
Fig. 16 Fig. 17 3. The Fata Morgana is a combination of superior and inferior
mirage occurring when a superheated layer is surmounted by a
single or multiple inversion. Two examples of these extraordinary
optical distortions are shown in Figs. 16 and 17. A most vivid
account of such deformations has been rendered by Koch [ 23 ] in
the description of his journey across Greenland on 12 April 1913.
4. Optical haze , or shimmer, occurs in a layer of air next
to the ground within which the lapse rate of temperature is ex–
cessive. Within this layer small-scale convective currents develop
with the warmer lumps of air ascending and the colder descending. The differences in the refractive index of these lumps cause a
blurring of objects seen through the layer. Optical haze occurs
quite frequently in the arctic in the same meteorological condi–
tions as the inferior mirage; it makes it difficult to identify
details in the landscape and is annoying for telescopic observa–
tions, particularly range finding by the aid of coincidence range–
finder.
5. Ice blink and water sky . - In the summer season (when the
sun is above the horizon at a small angle of elevation) light is
reflected and scattered between the ice surface and the base of
low layers of clouds. The whitish glare that is often seen on low
clouds is due to refection of light from distant ice fields. Some–
times, these reflections are quite intense and are thus called ice
blink. Conversely, if there are patches or lanes of open water in
the ice, dark patches or lanes will be seen on the base of cloud
layers. This is called water sky. Observations on water sky and
ice blink are extremely useful for navigation on the ice, for they
indicate, as if seen in a mirror, what lies beyond the horizon.
AIR MASSES AND FRONTS
The concept of air masses, introduced by Bergeron [ 3 ] ,
is much used in modern meteorology and denotes a vast body of air
whose physical properties are more or less uniform in the horizontal
direction. The air, being almost transparent relative to high–
temperature radiation, absorbs only a small portion of the direct
solar radiation, and the earth’s surface, which is an efficient
absorber, takes up a large portion of this radiation, converts it
into sensible heat and gives it back to the atmosphere, partly
through low-temperature radiation but mostly through eddy motion,
or mixing. Consequently, the physical properties of the earth’s
surface constitute a predominant factor in the formation of the air
masses.
The air masses typical of the arctic region in winter are
characterized by low temperature at all levels, extreme absolute
dryness and excessively stable stratification. Some examples of
typical winter conditions are shown in Table IV. It will be seen
that the relative humidity at high levels is very low; this condition
TABLE IV. Typical Temperature (T) and Relative Humidity (R) of
Arctic Air Masses in Winter
Eureka Sound
(8 ft. above MSL)
Jan. 19, 1949.
Fairbanks
(440 ft. above MSL)
Jan. 14, 1949
International Falls
(1112 ft above MSL)
Jan. 28, 1949.
Height T(°C) R (%) T(°C) R(%) T(°C) R(%)
Station level −45.0 25 −32.8 42 −17.2 68
2,000 ft. −25.4 50 −18.1 68 −17.2 67
4,000 ft. −24.8 54 −17.5 75 −15.7 67
6,000 ft. −23.4 35 −17.0 67 −14.0 32
8,000 ft. −21.8 24 −20.6 53 −13.7 40
10,000 ft. −24.9 - −22.6 28 −14.0 31
15,000 ft. −35.1 - −31.4 - −19.3 32
20,000 ft. −44.7 - −40.5 - 29.3 -
is due to the circumstance that the air, on account of radiative
cooling aloft, takes part in a sinking, or settling motion. As
a result of this prevailing sinking motion and relative dryness,
high clouds are extremely rare over the central arctic in winter
(see p. ).
Fig. 18 The source region of arctic air masses in winter is shown in
Fig. 18. On the North American and on the Eurasian sides it borders
onto the source regions of polar continental air masses. These
latter air masses are in many respects similar to the arctic regions,
except that the air masses are more shallow and have less extreme
properties.
It will be seen from Fig. 18 that warm air from the North
Atlantic (polar maritime air) normally invades the arctic in the
region between Iceland and Norway as far east as Novaya Zemlya.
Less frequently, warm air from the Pacific invades the arctic along
the west coast of Alaska. On the other hand arctic air invades the
midlatitude belt most frequently over the eastern parts of North
America and Siberia. On the whole, more arctic air is shed southwards than warm air northwards at low levels, the differences
being made up by an excess of northward transport of warm air at
greater heights in the troposphere.
The southward flow of arctic air is by no means a steady one;
it appears to occur in outbursts of considerable strength, at in–
tervals of 3 to 10 days, the outburst being associated with intense
traveling cyclones. These outbreaks of arctic air may sometimes
reach as far south as 25°N, and are the main cause of cold spells
in low latitudes. The preferred regions for these outbreaks of
arctic air are the eastern part of North America and the western
part of the North Atlantic, and the eastern part of Siberia and the
adjoining part of the North Pacific.
Fig. 19 In summer the arctic source region is less effective, owing
to the sun’s being above the horizon, and the contrast between the
arctic and neighboring air masses is less extreme. The source region
of arctic air masses in summer is shown in Fig. 19 in relation to
neighboring sources. Some examples of typical arctic air masses in summer are given in Table V. On the whole, the relative humidity
aloft is higher in summer than in winter, and, as a result, the
amount of high clouds reaches a maximum in the warm season.
Fig. 20 Fig. 21 Fig. 22 The transition from the arctic to the neighboring air masses
is usually not continuous. The mean circulation of the atmosphere
is such that there is a tendency for the air masses from neighboring
source regions to be brought together along zones of convergence,
Along these zones of convergence, which are called fronts , or
frontal zones, more or less abrupt transition in wind, temperature,
humidity, and weather will be found. The mean positions of these
principal frontal zones are shown in Figs. 20 and 21 for winter
and summer respectively. It will be seen that, on the average,
the arctic front is not continuous around the pole; it is normally
absent in the preferred regions of outbreaks of arctic air. A
schematic meridional cross-section of the principal air mass sources
and frontal zones is shown in Fig. 22.
TABLE V. Typical Temperature (T) and Relative Humidity (R) of
Arctic Air Masses in Summer
Eureka Sound
(8 ft. above MSL)
July 1, 1950.
Fairbanks
(440 ft. above MSL)
July 20, 1946
International Falls
(1112 ft. above MSL)
July 12, 1946.
Height T(°C) R (%) T(°C) R(%) T(°C) R(%)
Station level 7.2 61 14.0 88 17.0 70
2,000 ft. 3.5 62 10.2 88 17.6 59
4,000 ft. −0.3 64 5.3 92 12.5 59
6,000 ft. −4.2 59 1.8 92 10.5 22
8,000 ft. −7.9 63 −0.6 98 8.3 -
10,000 ft. −11.0 66 −2.9 100 5.3 -
15,000 ft. −18.3 86 −10.8 16 −2.0 20
20,000 ft. −28.1 62 −20.5 58 −11.6 -
CYCLONES AND ANTICYCLONES
The frontal zones discussed in the foregoing section are
rarely stable. On account of the contrasts in energy stored along
frontal zones, perturbations (known as cyclones, depressions, or
lows) develop and travel along the frontal zones, generally from
the east to the west with a component toward the north. In the
areas between the cyclones, regions of high pressure, or anti–
cyclones, develop and travel, generally eastward with component
toward the south. Most of these traveling cyclones remain in the
sup-polar belt and affect the fringe of the arctic region; some
of them, however, move into and cross the arctic.
Fig. 23 An example of such a chain of fronts, cyclones and anticyclones
around the arctic is shown in Fig. 23. It will be seen that, on
this occasion, the arctic front is well developed over North
America and over northern Siberia. The polar front, too, is well
developed, more or less in its normal position. A series of cy–
clones is associated with the frontal systems, with anticyclones
in between.
Fig. 24 As compared with middle and low latitudes, the arctic region
is a relatively quiet area as far as traveling disturbances are
concerned. The mean meridional distribution of frequencies of
cyclogenesis (formation of cyclones), cyclones, anticyclogenesis
(formation of anticyclones), and anticyclones is shown in Fig. 24
as a mean for all longitudes.
In summer most anticyclones form about 50°N and move southward
such that their mean position if about 38°N. There is, however,
a secondary maximum of anticyclogenesis at about 75°N and a rela–
tively large maximum north of this latitude. It will further be
seen that most cyclones form about in latitude 50°N and move such
that their mean latitude is about 60°N. It has been shown by
Petterssen [ 32 ] that this poleward tendency of cyclone move–
ment is due to the thermal structure of the atmosphere and the
rotation of the earth.
In winter the frequency distribution is, in principle, the
same as in summer. In all seasons the arctic region as a whole is characterized by a low frequency of cyclones and a relatively
high frequency of anticyclones.
The low frequency of cyclones is not typical of the entire
arctic region. As has been shown by Petterssen [ 32 ] , general
dynamical principles require that cyclonic circulation (vorticity)
must be produced in the cold sources above cold land and ice fields
and exported along isentropic surfaces downward to sea level along
the arctic coast. Hence all the bays of open water along the
fringe of the arctic will be characterized by cyclonic activity.
These areas of maximum cyclonic activity are also regions of
generally bad weather.
Fig. 25 Fig. 26 Figures 25 and 26 show the geographical distribution of
cyclone centers in winter and summer, respectively. The following
regions in arctic and subarctic latitudes are characterized by high
frequency of cyclones: The Gulf of Alaska and the Aleutian Chain
(winter and summer), the Baffin Bay and Davis Strait (winter and
summer), the waters south and west of Iceland (winter and summer),
the Norwegian Sea (winter and summer), the Barentz Sea (mostly winter).
Fig. 27 Fig. 28 The corresponding frequencies of anticyclones are shown in Figs.
27 and 28. It will be seen that winter anticyclones are quite fre–
quent over the central arctic, with a maximum over the north coast
of Alaska. In summer (Fig. 28) anticyclones are rare on the out–
skirts of the arctic, but quite frequent over the ice pack.
ATMOSPHERIC PRESSURE
Until seventy years ago, all theoretical considerations concerning
the general atmospheric circulation postulated a zonal system of westerly
winds circulating about a low-pressure area centered at the North Pole.
Subsequent expeditionary data on winds and pressures within the Arctic,
however, failed to verify this simple concept. Furthermore, as early as
1888, Helmholtz [ 18 ] had deduced from hydrostatic considerations that
the prevailing low Arctic temperatures should produce a shallow surface
anticyclone in polar regions. As a result of the mounting evidence from
Arctic observations, together with the general acceptance of the Helmholtz
theory, it became popular during the next four decades to consider the
existence of a permanent polar anticyclone even though the details of the
Arctic pressure distribution remained essentially undetermined.
However, from charts published in 1929 depicting the monthly
averages of temperature, pressure, and cloudiness, Baur [ 1 ] was able to
show that the centers of high atmospheric pressure tended to correspond
closely to the regions of minimum hemispherical temperature. He traced
the movement of the principal Arctic center of maximum pressure (and
minimum temperature) from a winter position in Eastern Siberia to a
position north of the Canadian Archipelago in spring, i.e., at a time
when the sub-Arctic continental regions become relatively warm in comparison
to the Arctic Ocean. He then made note of a continued easterly movement
of the centers to a position northeast of Greenland and Spitzbergen in
early summer. In early autumn Baur found a secondary maximum in pressure
over the Polar Sea which he ascribed to the effects of rapid cooling in
the Canadian Archipelago and adjoining ice pack, but on his series of
charts this center is soon superceded by the Siberian high and a weaker
counterpart over the Yukon Territory.
The next published series of charts showing the distribution of
pressure over the Arctic was prepared by Sverdrup, Peterson and Loewe
[ 42 ] and drew heavily upon the analyses of Baur, Birkeland and Føyn
[ 1 ] . Dorsey [ 12 ] has more recently revised Sverdrup’s charts upon the
basis of modern observational data. Dorsey also has considered the charts
prepared by Dzerdzeyevski [ 13 ] in 1945 - a series which had made use of
the pressure data obtained near the pole by the Russian North pole Expedition.
The charts prepared by Dorsey are apparently the most up-to-date and
figs 29-32 here are the ones presented here as Figures 29 to 32 . While the accuracy of
the pressure field indicated for the Arctic Ocean, Greenland, and the
Canadian Archipelago may be open to some question, it is, nevertheless,
probable that the charts do contain the essential characteristics of the
true pressure field at the surface. The prevailing wind directions usually
agree with the isobars and the locations of the principal cyclone tracks
and frontal systems are in accord with the resulting wind distribution.
In winter (January, Figure 29 ), the Arctic pressure field is
dominated by the extensive anticyclonic system centered over the Asiatic
Continent and extending as a ridge toward the Chukchi Peninsula and two
very large low-pressure areas - one centered southwest of Iceland and
extending northeastward over the Barents Sea and northwestward over Davis
Strait, and a second centered over the Aleutians and occupying the entire
Bering Sea and Gulf of Alaska. A secondary anticyclone with a central
pressure in excess of 1020 mb is centered over the Mackenzie River Valley.
In spring (April, Figure 30 ), high pressure exists over the greatest
part of the Arctic Ocean and it is during this season that pressures reach
their annual maxima over the Canadian Archipelago and northern Greenland.
Meanwhile, the Siberian anticyclone has so weakened in intensity that it
can no longer be discerned in the mean isobaric pattern. The Aleutian
low weakens somewhat during this season but retains approximately the same
position as in winter. The Icelandic low, on the other hand, is very much
less intense and occupies its southernmost position of the year (south
of Greenland).
During summer (July, Figure 31 ), the pressure gradients at
their weakest over the Arctic. This is to be expected when one considers
that summer is also the season of minimum thermal contrast between polar
and temperate zones. It is during this season that the so-called
circumpolar belt of low pressure is located at its highest latitude. The
Polar Sea during this season is occupied by a true anticyclone as a result
of the contrast between the low surface temperatures over the cold sea
and the relatively higher temperatures which exist at the same time over
surrounding coastal and inland regions. (see Figure 45 .)
In autumn (October, Figure 32 ), there is a return to a pressure
distribution which is more typical of winter conditions. The Icelandic
and Aleutian lows increase in intensity as the deep cyclonic disturbances
over the northern North Atlantic and North pacific become more frequent.
Meanwhile, the Siberian high has begun to make its appearance, although
a closed anticyclone still remains over the Arctic Ocean north of Greenland
and the Canadian Archipelago as pointed out in a previous paragraph.
At this point it should, perhaps, be mentioned that the details
of the pressure distribution over interior Greenland have little meaning
as shown on the sea-level pressure charts. It is possible, on the basis
of theoretical considerations and actual pressure data, to construct an
idealized picture of the pressure field which would obtain over the region
were Greenland not in existence. However, such a picture has no physical
or meteorological significance. Greenland is not merely a positive
topographic feature but it is, in itself, an important air-mass source
because of the altitude and extent of the Ice Cap and because of the
abnormally steep horizontal (and vertical) temperature gradients which
are produced by the temperature contrasts between ice fields and coasts
[ 42 ] . It is because of these facts that there is found to be so little
connection between the pressure field as charted and the wind speeds and
prevailing directions which are recorded for various points over Greenland
or along its coastline.
Pressure Fluctuations . - The pressure variabilities observed from
day to day at any Arctic location are directly related to the number
and intensity of migratory cyclones and anticyclones which pass near
enough to affect the area (see pages to ). The extreme ranges
occur during winter in connection with intense cyclonic activity and at
most Arctic meteorological stations both the absolute maximum pressures
and the absolute minimum pressures have been recorded during one of the
colder months. For example, the absolute maximum pressure for a Kara Sea
location occurred at Ostrov Domashnii on February 17, 1933 (1058.5 mb),
while the absolute minimum pressure for the same area occurred at Ostrov
Belyi on November 19, 1933 (950.1 mb) [ 45 ] . The average daily variability
of pressure shows a characteristic annual course with the maximum values
in winter (8 to 10 mb) and the minimum values in summer. In some Arctic
areas there appears to be an additional secondary maximum in April. The
monthly pressure variations also show the same general trend as the daily
variations, as is illustrated by the following typical Arctic data [ 43 ] :
Monthly Pressure Variation (in mbs)
Yrs
Location Rec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Arctic Ocean
Fram and Maud
5 52.9 53.0 41.0 40.5 32.6 28.1 32.9 27.5 42.8 41.0 44.1 49.4
Central Canadian
Archipelago
15 42.7 45.7 39.6 35.5 30.8 26.8 26.1 27.7 33.2 34.1 37.6 42.0
Extreme pressure changes of as much as 50 mb in 24 hours have been recorded
at some stations on the coasts of the Arctic Ocean in connection with the
passage of an intense cyclone or anticyclone.
The mean monthly pressure values in individual years and even the
annual pressures for different years may deviate considerably from long–
term averages [ 45 ] . For example, the mean annual pressure at Yugossky
Shar was 1031.1 mb in 1933 and 994.3 mb in 1914. These differences are
completely accounted for by differences in the frequencies of cyclonic
activity between the individual months or years.
The preceding discussion applies only to true Arctic conditions
In lower latitudes around the periphery of the Arctic, the non-periodic
variations in pressure are of greater magnitude.
Pressures in the Upper Atmosphere Pressures in the Upper Atmosphere . - When Helmholtz [ 19 ] deduced the
existence of a polar anticyclone upon the basis of hydrostatic considera–
tions, by the same reasoning he also concluded that low pressure must exit
aloft over the Arctic to compensate for high pressure at the surface. Modern
information concerning pressures aloft over the Arctic verify the basic
concept of Helmholz but the relations between the surface pressure and the
pressure aloft are not as simple as was at first believed. The Siberian
high in winter appears to be a relatively shallow phenomenon and is probably
superceded by relatively low pressures at a comparatively low altitude.
Details concerning the vertical structure of the atmosphere at low levels
over this region are lacking, however. The essentials of the vertical
distribution of pressure elsewhere within the Arctic can probably best be
described by examining the pressure field at the 700-mb level.
Namias [ 30 ] has recently prepared upper - level charts illustrating
the height of the 700-mb surface over the Northern Hemisphere. His (slightly revised)
January and July charts are given as Figures and , with the altitudes
Figs. 33 and 34 here of the pressure surface represented by isolines plotted for 100-ft
intervals. The January chart (Figure ) shows the pressure field to have
a rather simple structure with two closed low-pressure centers - one over
the southeast portion of the Canadian Archipelago and a second in the
sub-Arctic over the Kamchatka Peninsula. The July chart (Figure 34 ),
on the other hand, shows much weaker pressure gradients than the winter
chart, and the pressures are higher as would be expected on the basis of
the higher surface temperatures. Three closed low-pressure areas are
shown within the Arctic during this season, the most important of which
is centered very nearly over the Pole. This is also an expected condition
when it is considered that the Arctic Ocean is essentially a “cold sources”
during warmer months in the Northern Hemisphere.
SURFACE WIND
It is extremely difficult to generalize upon the surface wind
field in the Arctic because of several circumstances which are of
peculiar importance at high latitudes. In the first place, the
character of the wind regime at most coastal points and at many
inland locations is largely determined by local factors which are
not amenable to regional generalization. Secondly, the periodic
changes and spa e t ial deformations in the general wind field are no
less variable than are the highly irresolute Arctic pressure distri–
butions which produce the winds in the first place. (See page .)
The third, but not the least important, difficulty is occasioned by
the fact that the scanty observational data available for analysis
do not represent a homogeneous period of record at all points of
observation. For this reason it is often difficult to ascertain
whether the differences in the wind conditions between two weather
stations represent true regional differences in the circulation or
whether they merely indicate that the observations were recorded
during different years — a difficulty which is particularly serious
when wind observations obtained on shipboard within the Polar Sea are
The General Wind Circulation . - The Arctic circulation is, of course,
dominated by the polar anticyclone which, during all seasons except
winter, is centered somewhere over the Arctic Ocean. So far as can
be determined from the scanty observational materials available, the
prevailing wind directions over the Arctic Ocean and surrounding
coasts appear to correspond to the mean pressure distribution. It is
apparent that easterly winds prevail over the well-explored portions
of the Arctic Ocean, over Iceland, the northern portions of Greenland,
and Alaska, and that northeasterly winds prevail in interior Alaska
and Greenland, all of which fits the prescribed mean pressure field.
(See Figs. 29 to 32 .)
Wind conditions over the interior and coastal portions of Arctic
Eurasia, however, appear to be less well-defined. The Siberian anti–
cyclone dominates the interior and coastal circulations during winter,
but during other seasons the winds are regionally highly irregular.
Along the Siberian coasts of the Arctic Ocean in summer there appears
to exist a large onshore wind component resulting from summer heating
over the interior.
In the more southerly portions of the Arctic, and particularly
in the peripheral maritime regions of frequent cyclone activity, the
surface winds are highly variable and do not exhibit a pronounced
“prevailing” direction. At several stations, for example, the
frequency data show that during the course of a year the winds tend
to blow almost as often from one direction as from any other. In
such regions the non-periodic features of the circulation far outweigh
any periodic or permanent characteristics.
The preceding generalization of the Arctic surface circulation
appears to be about as complete a description as possible of the
large-scale aspects of Arctic winds. The remaining periodic and
regional differences in the surface circulation are the result of
local factors which will be described in greater detail.
Surface Wind Speeds . - A large proportion of the description given
by polar explorers have stressed the prevalence of high winds and have
almost invariably given the impression that the Arctic is indeed a
stormy and inhospitable place. One fairly recent publication on
Arctic weather conditions, for example, presents a 3 1/2 page
discussion of winds and storms and, of this discussion, at least 3
pages describe extreme winds reported by various Arctic expeditions
since 1836. It is true that in some restricted areas within the
Arctic the almost continuous high winds are the most noticeable
feature of the climate. It is also true that excessive winds have at
one time or another been reported from nearly every Arctic observing
station. These circumstances, however, do not suffice to ascribe an
unusual severity to Arctic wind conditions.
According to Sverdrup / [ 42 ] / , relatively low wind velocities are
characteristic of the Arctic Ocean and Canadian Archipelago. The annual
mean wind speed as recorded over the Polar Sea was only 10 mph during
the Fram Expedition from 1893 to 1896, and 9 mph during the Maud
Expedition of 1922 to 1924. Summarizing the observed conditions,
Sverdrup states, “It is remarkable that very high wind velocities are
so rare.” The highest wind speed observed on Fram was 40 mph,
and on the Maud Expedition, 34 mph; the latter, however, is an hourly
mean value. The Russian North Pole Expedition of 1937 / [ 5 ] ] / found
that high winds occasionally do occur near the North Pole since they
reported from the Pole on June 8 and 9 to the effect that gusts had
attained speeds of 60 feet per second (41 mph).
The relative frequencies of high wind speeds at various Arctic
points can be judged from the data on the average monthly number of
days with winds of gale, force, which are presented in Table VI . Table VI here
Table VI. Mean number of days with “gales”*
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Oceanic:
004 5 4 4 3 2 1 ** ** 2 3 5 4 33 ..
006 5 4 3 3 2 4 1 2 4 3 4 4 39 7
007 5 3 3 5 2 3 1 2 7 8 5 7 50 5-6
Alaska, Coastal and Insular:
100 2 1 ** 1 1 0 1 1 2 5 3 2 19 5
101 8 9 8 1 1 .. 1 4 5 11 18 14 .. 0-2
103 14 19 13 7 4 9 9 7 7 10 13 12 121 2
104 4 2 3 3 1 ** ** 1 1 2 1 4 21 10
106 1 2 4 4 7 6 4 6 3 6 4 2 50 6-7
Alaska, Inland:
155 0 2 2 0 0 0 0 0 1 0 1 1 7 1-3
156 1 0 ** ** ** ** 0 0 0 0 0 0 1 10
Canada, Coastal and Insular:
222 4 2 2 4 3 1 2 1 2 6 4 7 38 2-3
224 3 1 2 2 1 ** 1 1 1 1 4 3 17 2-3
225 8 8 7 5 4 3 3 3 4 5 8 8 66 5-6
228 3 4 5 4 3 1 1 3 5 8 5 7 50 3
Greenland, Iceland, Coastal and Insular:
301 1 1 1 ** 1 2 1 2 2 1 1 ** 12 5-6
302 4 4 1 1 3 2 3 3 4 2 2 ** 28 2-3
304 1 2 1 1 2 1 2 1 1 2 2 ** 15 10
305 10 11 10 6 8 2 5 3 6 10 7 5 82 3-4
306 2 2 1 1 ** 1 ** 1 1 3 1 1 14 5
308 10 6 9 10 6 4 6 3 8 10 8 6 84 2
2 4 6 8 10 3 5 7 9 11
Table VI. Mean number of days with “gales”* (cont.)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Greenland, Iceland, Coastal and Insular (cont.):
309 4 ** 2 ** ** ** ** 1 1 2 2 2 15 4-5
310 4 5 6 3 2 ** 1 ** 5 5 5 8 43 2-3
311 4 2 1 1 1 ** ** ** ** 1 2 1 11 2-3
312 4 2 2 1 2 2 1 1 2 2 2 2 22 6-7
313 21 22 24 17 9 5 5 5 8 15 15 20 167 3-4
314 4 1 2 3 ** 5 2 2 3 2 2 2 29 4-5
315 6 2 2 2 1 1 ** 1 ** 1 3 5 23 4-5
316 10 15 15 9 10 4 3 3 13 5 12 21 121 2
317 7 5 4 2 1 1 ** 1 3 2 4 5 36 10
318 4 5 1 3 1 1 1 1 1 2 3 2 22 8
319 9 7 6 6 1 2 1 1 3 3 6 10 54 2-3
320 12 10 8 7 4 4 3 4 6 9 10 8 84 10
321 19 17 19 15 12 11 4 8 8 10 12 19 153 4-5
330 2 2 1 1 1 1 ** 1 1 1 2 2 14 16
331 3 2 1 1 ** ** 0 ** ** 1 2 3 13 16
332 3 2 1 2 1 1 ** 1 2 2 1 1 17 9
334 3 2 2 1 ** ** 0 ** 1 2 2 2 15 14
335 1 1 1 ** ** 0 0 ** ** ** 1 1 5 15
337 ** 0 0 ** 0 0 0 0 0 0 ** ** 1 11
339 2 1 1 ** ** ** 0 ** 1 1 1 ** 9 10
340 0 0 1 0 0 0 0 0 1 0 0 0 2 1
Greenland, Iceland, Inland:
361 ** ** 0 ** 0 0 0 0 ** ** 0 ** 1 16
2 4 6 8 10 3 5 7 9 11
Table VI. Mean number of days with “gales”* (cont.)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Europe, Coastal and Insular:
400 110 9 7 5 3 3 2 2 4 6 8 11 69 10
401 11 8 7 9 3 3 2 2 6 6 11 11 79 1
406 7 7 6 4 3 2 1 2 3 4 6 7 52 19
408 5 5 4 3 2 1 1 2 2 3 5 5 38 31
412 2 2 2 1 0 ** 0 ** 1 1 2 2 13 28
414 2 2 2 1 ** ** ** 0 ** ** 1 2 1 9 26
415 4 4 4 2 3 2 2 1 2 3 5 4 36 18
417 6 3 3 6 2 2 0 1 3 4 5 4 37 ..
420 3 3 2 1 ** ** ** ** 1 2 2 3 17 26
421 ** 0 ** ** 0 0 0 ** ** 0 ** ** ** 8
423 0 0 0 ** 0 0 0 0 0 0 0 0 ** 7
425 2 1 2 1 2 2 1 1 2 2 1 1 19 18
426 ** ** ** 0 0 0 0 0 ** 0 ** 0 1 17
428 2 2 2 1 2 2 1 1 2 2 2 2 20 18
429 1 1 1 1 1 1 1 ** 1 1 1 1 8 18
Europe, Inland:
450 ** ** ** 1 0 1 0 0 ** ** ** 1 3 7
451 ** ** ** ** 1 ** 0 ** 1 1 1 ** 4 14
455 0 0 0 0 0 0 ** 0 0 0 0 0 ** 8
456 ** ** 0 ** ** ** ** 0 1 ** 0 0 1 7
Asia, Coastal and Insular:
500 14 8 11 11 7 11 10 10 8 9 13 14 126 7-8
502 9 8 6 8 6 6 6 5 4 3 8 8 76 4-5
2 4 6 8 10 3 5 7 9 11
Table VI. Mean number of days with “gales”* (cont.)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Asia, Coastal and Insular (cont.):
503 8 4 4 2 2 4 3 2 3 3 6 6 47 7-8
506 6 5 5 5 5 3 1 2 3 6 7 5 53 24
507 11 10 13 13 10 11 8 6 8 6 8 11 114 17
508 4 3 5 4 3 2 1 2 4 5 5 4 41 6
509 5 6 6 6 7 2 2 4 4 8 10 5 65 6-7
510 11 19 8 7 5 3 1 3 5 6 9 8 75 19
513 1 1 0 1 ** 2 3 2 2 2 1 2 18 4
515 ** ** 2 1 1 2 2 2 1 2 1 1 15 5
517 8 3 6 2 3 3 1 2 8 7 8 4 55 3
518 10 9 8 8 7 5 3 2 5 7 10 9 82 26
519 6 6 5 6 6 2 1 2 5 7 8 8 60 22
521 4 4 5 4 5 2 1 3 4 5 4 4 46 13
522 1 ** 1 1 1 1 ** 2 1 0 0 1 9 5
523 10 6 8 5 5 4 2 4 5 7 4 8 68 3
525 5 5 5 5 5 3 6 6 6 6 8 8 68 9
526 3 2 3 2 1 1 1 1 1 2 2 2 21 14
527 4 4 4 2 1 ** 2 1 2 2 4 3 29 12
531 1 ** 2 ** 0 1 0 0 0 2 2 3 11 4
Asia, Inland:
550 4 4 2 2 1 0 1 ** 1 1 2 3 19 5
551 6 6 2 2 2 0 2 4 0 1 5 6 35 1
553 5 6 7 5 3 2 1 2 3 4 5 4 48 18
554 7 10 8 11 10 7 6 7 6 7 12 9 100 4
556 ** ** ** ** ** 2 2 1 1 ** ** 0 7 17
2 4 6 8 10 3 5 7 9 11
Table VI. Mean number of days with “gales”* (cont.)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Asia, Inland (cont.):
558 0 0 ** 0 ** 1 1 1 ** 1 0 0 4 8
560 ** 2 3 2 2 2 2 1 2 1 2 1 20 8
561 1 1 1 1 1 ** ** ** ** ** ** ** 5 11
562 ** 2 3 1 3 2 ** 1 2 1 3 1 18 10
563 2 2 1 2 1 1 1 1 1 2 2 2 17 6
566 ** ** 1 ** 1 1 1 ** 1 1 1 0 7 13
568 ** ** 1 1 2 2 1 1 1 1 ** ** 10 18
571 0 0 0 0 0 0 0 1 0 4 0 1 6 1
572 1 1 1 ** ** 1 ** ** 1 ** 1 1 7 16
573 2 2 2 3 3 2 2 1 3 2 2 2 25 9
574 1 ** ** ** 1 ** 0 ** ** ** 1 ** 5 10
575 1 ** 0 0 0 0 0 ** ** ** ** ** 2 18
576 ** 1 1 2 1 2 1 1 ** ** 1 1 11 13
2 4 6 8 10 3 5 7 9 11
Data from coastal stations along the Arctic Ocean indicate that
average wind speeds at most points are higher than over the ocean
areas, but even here the wind speeds are low except where strongly
influenced by local factors. In general, the average wind speeds at
coastal points are of the order of 10 to 15 mph except at more exposed
locations where averages of 15 to 20 mph are fairly common. (See
Table VII .) However, the coastal areas may also experience excessively Table VII here
high winds at times. For example, on February 8, 1909, Note: Possibility that this may be an incorrect date See Reference title, J.P.J. a temporary
weather station s at Winter Harbor, Melville Island, recorded a 1-hour
average wind speed of 86 mph and a speed of over 100 mph for a
20-minute period. The average speed for the 24-hour period was in
excess of 60 mph / [ 6 ] / .
Wind speeds over inland areas at low elevations within the Arctic
are usually much lower than those over either the Polar Sea or its
coasts. The mean annual wind speed at Verkhoyansk, for example, is
only 3.2 mph and at Yakutsk, 3.9 mph.
Table VII. Average specified wind speed (mph)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Oceanic:
001 11 9 10 11 13 14 12 11 13 14 14 .. .. 1
002 11 9 8 8 11 12 12 11 11 10 9 10 10 1
004 21 21 18 17 13 14 12 14 16 18 19 18 17 ..
006 14 14 11 11 13 13 11 12 16 14 14 12 13 5-6
007 15 13 13 13 14 14 13 15 17 18 17 16 15 5-6
008 11 .. .. .. .. .. 8 11 7 8 9 10 .. 0-1
Alaska, Coastal and Insular:
100 10 10 11 12 11 11 13 13 14 15 12 10 12 3-17
101 15 18 18 11 11 12 13 15 15 17 22 19 16 1
103 22 24 21 17 15 16 18 15 17 17 22 20 19 1
104 9 9 9 9 7 7 8 8 9 9 9 9 8 18-30
106 4 5 6 6 6 5 4 4 4 5 5 5 5 8
Alaska, Inland:
155 6 8 8 8 7 7 8 7 7 7 7 6 7 3
156 3 4 5 6 7 6 6 6 5 5 4 4 5 8
Canada, Coastal and Insular:
206 12 11 10 13 14 12 9 13 12 15 15 14 13 6
207 3 4 2 3 4 4 3 3 3 4 3 4 3 4
209 5 4 3 3 5 3 2 2 1 3 3 3 3 6
211 4 4 4 4 6 8 9 8 7 7 7 3 5 4
216 8 8 7 7 7 7 8 9 10 10 10 8 8 8
221 15 15 14 14 14 12 8 13 15 17 15 16 14 8
222 3 3 2 3 3 2 3 4 4 4 3 2 3 1-2
Table VII. Average specified wind speed (mph) (cont.)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Canada Coastal and Insular (cont.):
223 11 11 11 13 12 12 10 11 12 13 15 13 12 15
224 11 7 9 9 11 8 9 8 9 9 11 10 9 5-6
227 18 18 17 17 14 13 13 13 15 18 19 18 16 15
228 15 15 14 14 13 12 11 13 16 17 17 15 14 12
Canada, Inland:
251 3 3 3 3 4 4 5 3 3 3 3 3 3 3-7
253 9 9 10 8 9 8 8 9 9 11 8 8 9 5
254 5 4 7 7 10 10 10 8 6 7 5 4 7 6-21
255 2 3 6 6 6 6 6 6 6 6 3 2 5 7
Greenland, Iceland, Coastal and Insular:
301 2 2 3 1 .. 6 3 2 4 6 2 2 .. 5
302 7 10 10 9 6 6 6 7 10 10 9 8 8 2
304 5 4 4 4 5 6 4 4 5 6 6 6 5 30
307 10 8 6 5 5 6 7 6 8 9 11 9 7 ..
308 5 5 5 4 3 3 3 3 3 4 4 4 4 12
314 14 14 13 11 9 9 8 9 4 11 12 12 11 30
317 6 6 5 3 3 3 3 3 3 4 5 5 4 30
318 4 4 3 2 2 2 1 1 2 2 3 2 2 30
320 13 12 12 11 8 10 8 7 9 10 11 11 10 23
330 12 11 10 9 6 6 5 6 8 9 10 10 8 16
331 14 14 13 12 11 11 9 9 10 12 12 13 12 4
338 9 10 9 9 9 8 7 7 8 8 9 9 7 15
340 6 6 4 4 4 4 3 3 4 4 5 5 5 19
Greenland, Iceland, Inland:
351 11 9 13 12 9 9 9 8 11 10 9 14 10 2
Table VII. Average specified wind speed (mph) (cont.)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Europe, Coastal and Insular:
400 16 15 12 12 11 10 8 9 13 15 15 17 13 10
401 20 17 14 15 13 11 8 10 13 15 18 18 14 1
407 10 10 9 7 5 6 5 4 5 7 8 9 7 10
408 22 21 21 19 16 16 13 14 17 19 21 21 18 28
412 17 16 16 13 11 11 10 10 13 13 14 16 13 28
415 10 9 9 8 9 10 9 7 8 9 10 9 9 10
417 18 17 15 15 16 16 14 14 15 15 16 17 16 ..
421 8 8 8 8 9 9 7 8 8 7 7 7 8 8
423 10 11 11 11 12 12 10 10 11 11 11 11 11 8
424 10 9 9 8 8 8 8 8 9 10 10 10 9 26
425 11 9 10 9 11 10 9 10 11 10 12 10 10 10
426 10 10 10 9 9 9 9 9 10 10 10 11 10 10
428 12 10 11 9 10 9 10 10 11 11 12 10 10 25
429 8 9 9 9 10 9 9 8 8 8 8 8 9 6
Europe, Inland:
451 3 2 3 3 5 8 4 2 2 1 2 3 3 10
452 10 11 11 10 11 13 11 10 11 11 9 9 11 25
453 8 6 7 6 7 7 6 5 6 7 6 7 7 10
454 14 13 14 14 13 13 13 13 12 13 14 13 13 11
455 6 7 7 7 8 8 6 6 7 7 6 6 7 8
456 4 5 6 8 9 8 7 7 9 6 5 4 7 8
457 7 7 8 6 7 6 5 5 6 7 7 7 7 36
458 6 6 7 6 7 7 7 6 6 6 6 6 6 30
Table VII. Average specified wind speed (mph) (cont.)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Asia, Coastal and Insular:
500 21 16 18 19 17 18 18 18 19 19 21 21 19 7-8
501 16 14 12 12 13 13 12 14 17 15 17 16 14 5-6
502 15 16 14 15 16 15 13 17 18 14 15 15 15 4-5
503 16 13 13 12 13 14 13 13 14 13 14 15 14 7-8
505 26 25 23 21 18 17 16 13 17 20 22 20 20 13
506 17 16 14 15 14 13 12 12 15 16 19 17 15 24
507 15 16 16 18 16 17 14 13 15 13 15 17 15 17
509 14 16 15 17 17 15 14 15 15 17 18 13 15 6-7
510 19 19 16 17 16 16 15 16 17 16 18 17 17 19
511 12 9 8 11 14 14 17 14 15 15 13 10 13 4
512 14 11 8 8 11 10 9 11 12 11 11 9 10 7
513 6 6 6 8 9 11 12 10 10 9 7 9 8 4
514 12 10 10 11 12 13 11 11 13 12 12 12 12 7
515 8 8 9 9 9 12 12 12 11 9 9 9 10 9
517 12 11 11 9 9 7 9 8 13 15 15 12 11 10
518 18 18 16 16 15 14 12 12 15 17 20 18 16 26
519 17 18 16 16 16 15 12 13 16 17 19 17 16 22
520 9 10 11 13 13 15 12 12 12 11 12 8 11 12
521 15 14 14 16 17 15 14 14 15 16 17 15 15 13
523 13 11 12 11 11 11 10 13 14 17 14 12 12 2
525 11 11 12 13 11 10 12 12 11 13 15 16 12 7
527 13 14 12 11 9 10 11 11 11 12 13 13 12 14
528 12 14 11 11 10 6 6 7 13 15 12 13 11 3
530 23 25 22 19 17 11 12 12 14 16 20 18 17 10
531 7 5 6 4 4 4 4 4 4 6 7 8 5 4
Table VII. Average specified wind speed (mph) (cont.)
Station Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Ann Yrs Rec
Asia, Inland:
550 9 9 8 9 10 9 9 9 8 8 7 8 8 5
552 5 6 7 10 11 11 10 9 10 10 7 6 8 18
553 15 16 15 15 15 13 11 12 13 15 14 13 14 18
555 8 9 10 10 11 10 8 9 10 10 9 7 9 4
556 1 2 2 4 6 7 6 5 4 3 2 1 3 21
566 2 3 3 4 5 5 4 4 4 5 3 3 4 24
567 4 4 5 6 7 6 5 5 5 6 5 4 5 12
568 3 3 3 5 6 5 5 5 5 4 3 3 4 21
576 1 1 2 2 3 3 2 2 2 2 2 1 2 13
In a discussion of the frequency distribution of various wind
speeds, Simpson / [ 39 ] / has pointed out that at a given locality the
frequency with which winds of different velocities occur is closely
associated with the type of pressure distribution characteristic of
the region, i.e., whether cyclonic or anticyclonic. He says:
“In one type the relative frequency increases as the velocity
decreases right down to calms; this type is associated with
anticyclonic pressure conditions. In the other / [ type / ] the
frequency increases as the wind decreases down to a certain
velocity after which the frequency decreases as the wind
decreases and calms may have a very small frequency; this type
is associated with cyclonic pressure distribution.”
This seems to be a reasonably useful method for classifying wind–
speed distributions at Arctic locations. It is not established,
however, that one is justified in assuming that a “cyclonic” type of
frequency distribution indicates a predominance of cyclonic curvature
to isobars over the region (or vice-versa).
Sverdrup / [ 43 ] / has performed the Simpson type of analysis upon
the wind data obtained at several coastal points on the Arctic Ocean
as well as upon those recorded from the Maud over the Ocean proper.
He found the “cyclonic type” of velocity distribution most frequent
over the Polar Sea during all seasons, with a tendency toward an
“anticyclonic type” at coastal locations, particularly during winter.
Sverdrup / [ loc. cit. / ] also found a good agreement between the type of
wind-speed distribution and the annual variation of pressure, i.e.,
the higher the barometric pressure the more nearly the frequency
distribution approaches the “anticyclonic type.” A similar analysis
of data from inland Arctic locations has not been performed, but (from
the high frequency of calms and low wind speeds) it can be deduced
that the anticyclonic type would greatly predominate during the colder
months.
Vertical Wind Distribution Within the Surface Layar . - It has already
been pointed out that the existence of a surface temperature inversion
is characteristic of the Arctic during all seasons except over areas
with a continental type of climate during the warmer months. The
altitude and magnitude of the surface inversion varies systematically
with season, remaining fairly uniform during colder months and
becoming more variable during warmer months. A consideration of these
variations from place to place and from season to season is of consid–
erable importance to conclusions regarding the variations of winds
within the surface layers of the atmosphere over the Arctic. It is
for this reason that the relationships between the vertical temperature
structure of the surface layer of air and the local wind conditions
will be discussed in some detail.
It is a matter of common knowledge among meteorologists that
the atmospheric mixing rate (eddy conductivity) is small within a
surface inversion layer, and smallest when the temperature increase
within the inversion is greatest. A low inversion, therefore,
effectively “seals-off” the surface layer against frictional drag
by the atmospheric layer above the inversion. In other words, the
circulation below the surface inversion layer has a strong tendency
to act independently of the rest of the atmosphere. This circumstance
is one of the several reasons why it is so difficult to describe the
general Arctic surface circulation in broad regional terms. (See
page .)
The wind velocity near the surface depends partly upon the
transport of kinetic energy from above (which serves to move
the surface layer of air) and partly upon friction at the ground
(which serves to slow down the movement of air near the surface).
Assuming the friction at the ground to be constant, the lowest wind
velocities at the surface occur when a sharp surface temperature
inversion is present because it is under these conditions that the
transport of kinetic energy from above is least. By identical
reasoning it can be pointed out that the highest velocities at the
surface will occur when the surface inversion is weakest or entirely
absent. In the presence of a strong surface inversion the ratio
between the wind speed at the surface and that, say at 1500 feet, is
large. When the inversion is weak or non-existent, the ratio is
small [ 43 ] . This circumstance is clearly indicated by the following
data which show the vertical distribution of wind velocity as related
to altitude of the base of the inversion, and which have been averaged
by Sverdrup for a number of Arctic points [ loc. Cit. / ] :
Altitude of Base
of Inversion
Zero to
330 feet
330 to 660
feet
660 to 980
feet
Greater than
980 feet
Altitude of wind
Measurement (ft)
Wind speed
(mph)
Wind Speed
(mph)
Wind Speed
(mph)
Wind Speed
(mph)
1970 23.7 25.3 29.1 28.4
1640 24.2 26.2 30.6 28.2
1310 25.1 25.9 30.2 27.3
980 25.9 26.9 29.5 27.5
660 25.9 25.9 25.9 24.2
330 22.1 20.1 22.8 21.3
20 10.5 13.2 13.4 15.7
Over the pack-ice, Sverdrup found large ratios in winter
between the velocities at 1,600 feet and those on the surface, the
maximum ratio thus coinciding with the presence of a low and sharp
temperature inversion which is characteristic of this season. During
spring and autumn he found the ratio to be small because in these
seasons the inversion lies higher and is less pronounced. A secondary
maximum in the ratio was found to occur in summer over the pack-ice
but not at the coast. The reason for the difference here is that over
the pack-ice in summer the surface temperature cannot depart much from
freezing, whereas the temperatures can rise well above freezing at the
coast with a consequent destruction of the surface inversion. (See
page .)
Details of the vertical structure of the surface wind field are
similar over the interior Arctic. The existence of exceedingly
strong surface temperature inversions during winter accounts for the
prevalence of calms at interior stations during this season, even
though winter is the period during which the arctic circulation reaches
its greatest intensity. Calms are much less frequent everywhere during
the warmer months when the inversion is either weak or absent.
Sverdrup / [ loc. cit. / ] has also shown that under inversion conditions
there is a clockwise turning of the wind with altitude. The lower
(or stronger) the temperature inversion, the greater the degree of
turning. When the base of the inversion is at an altitude of less than
330 feet his data show an average deviation of 26 degrees between the
direction of the wind at the surface and the direction observed at
330 feet. When the base of the inversion is at an altitude greater
than 1,000 feed, the average turning is only 4 degrees.
It should, perhaps, be pointed out here that the velocity
profiles described in this subsection are typical of conditions
found over more-or-less level surfaces and do not apply to conditions
as they occur within a “fallwind” or “katabatic” flow. This latter
phenomenon will be described in a subsequent section.
Local Influences on Surface Winds . - No one can examine detailed
Arctic wind data from coastal and inland points without being
impressed by the fact that local surface wind speed and direction are
largely determined by exposure of the station (and wind instruments)
and by the location of the area with respect to land and water bodies
and to the regional orography. As an example of the influence of
exposure, Sverdrup [ 42 ] cites one case where a series of observations
made between 1900 and 1902 at a station on Kllesmere Island showed an
average wind speed of 11.2 mph, whereas a series of observations made
in the same area during the interval 1899 to 1900 gave an average
wind speed of only 2.1 mph. The reason for this marked difference
is that the winter lodgings were located in a more sheltered location
during the first year of observations than was the case during the
succeeding years. figs. 35 and 36 here
At many coastal points, particularly along the costs of
Greenland and Iceland, the direction and speed of the surface winds
are so local in character that they may bear little or no relation–
ship to winds in the offing. In Iceland, the numerous fjords
indenting the west, north, and east sides of the island contribute
their quota of gorge and channel winds. Similarly, at many places
along coastal Greenland sheltering bluffs or the trend of fjords
largely determine the local wind directions. For example, there are
very few northerly and northwesterly winds at Godhavn, a fact which
is due to the location of the observing station on the shore below
the southerly bluffs of Disko Island. In the same manner, the wholly
different frequencies of north, northeast, and east winds at Godhavn
and Jakobshavn, stations quite near each other, are entirely the
result of different topographic influences. In contrast, at Jakobshavn
and Holsteinsborg, the latter a station much farther from Jakobshavn
than Godhavn, the distribution of wind directions is quite similar
because of a similarity in topographic features / [ 42 ] / .
Regional Influences on Surface Winds . - All expeditions to the Ice
Cap region of Greenland have commented on the fact that the surface
winds ordinarily are directed from the interior toward the coast.
This wind regime is merely a large-scale example of a “gravitational”
or “katabatic” wind system. These downslope winds result from the
presence of the cold layer of stable air which forms over the Ice Cap
and which subsequently flows down the slopes of the Ice Cap under the
influence of gravity. The speed of the flow depends first upon the
steepness of the slope; secondly, upon the temperature (density)
contrast between the stable surface air and the warmer air above;
and, lastly, upon the pressure gradient between the inland ice and the
coast. Near the summits of the Ice Cap the winds are more variable,
since their direction and speed depend more upon the magnitude and
direction of the pressure gradient then upon the activity of the cold
air. These katabatic or downslope winds, while they are particularly
characteristic of the Greenland climate, are not limited to this
region. They are an important climatic feature in other Arctic regions of diverse topography, such as along the coasts of northern Norway, Spitzbergen,
at Wrangle Island, and in parts of Alaska. Winds of this character
can occur in any part of the Arctic where there is sufficient area
at high elevation to allow the accumulation and downslope flow of
air which is very cold relative to air in the free atmosphere at the
same elevation. Williwaws , also known as Takus or kniks , are a form
of this katabatic wind that occurs along parts of the Alaskan coast.
They are most frequently encountered below along precipitous coastlines.
In many instances the cold air supply which initiates the fallwind
is of limited supply and, for this reason, williwaws are often of
short duration since they cease as soon as the cold air is exhausted.
The depth of the katabatic wind is seldom more than five or
six hundred feet, even along the Greenland Ice Cap. These winds
are strongest and deepest when the temperature contrast between the
coast and the interior is greatest. For this reason they tend to
present a maximum frequency in the morning hours and a minimum in
the afternoon, although there are local exceptions to this general
rule, particularly nearer the centers of origin of the winds (as in
the interior of the Ice Cap). They are also somewhat stronger on
clear days than on cloudy days when interior temperatures are above
average. In contrast to the vertical distribution of wind speeds in
other types of Arctic wind systems (see page ), the maximum wind
speeds are found near the surface, i.e., winds are stronger below the
surface inversion than above.
Another regional factor affecting the wind regime along Arctic
coasts is exactly opposite in both cause and effect to the fallwinds
just described. In this case the wind is directed inland from a
relatively cool ocean toward a heated continental interior. This
phase of a monsoonal (seasonal reversal of circulation) wind regime
occurs during summer, as would be expected, and is well-developed on
a fairly large scale in portions of Siberia and Alaska. This component
of the surface wind is seldom very great and at no times do the wind
speeds approach the extreme values found in the katabatic flow. Thermal
lows develop during the warmest months in the interiors of Siberia and
Alaska, replacing winter anticyclones in each case. This substitution
results in a tendency toward a reversal of the circulation between
winter and summer. Actually there are few localities where a complete
180-degree reversal of the circulation takes place between the seasons.
In most cases the effect is merely one of several factors which
influence the circulation and may only serve to deviate the wind
slightly from the direction which would prevail were there no marked
temperature contrasts between land and sea.
Diurnal Variation of Wind . - The land- and sea-breeze effect is
seldom as well-developed in Arctic regions as in more southerly
latitudes — largely because there is no important diurnal variation
in surface heating and cooling. Nevertheless there is a tendency
toward a diurnal reversal in wind direction along Arctic coasts
during warmer months, particularly when pressure gradients are weak.
For example, in some parts of Greenland there is a tendency during
summer for winds to blow inward along the fjords during daytime and
outward during the night [ 42 ] .
The wind speed at most Arctic stations also shows a diurnal
variation, both inland and at the coast. At locations surrounded
by moderate elevations the chan g es in wind speed are directly related
to changes in stability, i.e., velocities are least during night hours
when stability is greatest and greatest during afternoon hours when
surface heating is at a maximum. Along the Arctic coast, Sverdrup
[ 43 ] found that the maximum wind speeds occur generally between 1200h
and 1400h and the minima between 0400h and 0600h. The diurnal variation
over the pack-ice is similar but of smaller magnitude, as would be
expected from the smaller diurnal range in stability. (See Fig 37 .) fig. 37 here
A diurnal period in the wind is also noticeable at higher
altitudes and on slopes (as along the Greenland Ice Cap) but in
these cases the phase is reversed, i.e., the minimum wind speeds
occur during the afternoon and the maximum speeds during night or
early morning hours. This type of diurnal variation in the case of
slope stations is related to the nighttime augmentation of the
katabatic-type flow through radiational cooling in the interior. At
locations on mountain ridges or peaks, a similar diurnal variation
takes place, but in these cases the cause is the lessened frictional
drag at the higher wind levels, lowered by increased stability at
the surface.
Annual Variation of wind speed . - Most of the factors involved in the
annual variation of winds over the Arctic have, of necessity, been
discussed in preceding paragraphs. All that remains is to describe
the more important factors affecting the seasonal variations of wind
speed as they apply to various types of Arctic topographies.
Wind speeds nearly everywhere along Arctic coasts and over the
polar sea are lowest during summer. At this time pressure gradients
are weakest because of the weakening of thermal contrasts between
high and middle latitudes and between continents and seas. However,
at locations within the continental interiors, winter is the period of
low wind speeds for the reasons that have already been pointed out on
page . The maximum wind speeds at interior stations tend to occur
in late spring or early summer (see data for Yakutsk and Verkhoyansk
in Table VII ), but, again, there are local exceptions. Elsewhere the
maximum wind speeds tend to occur during late autumn or winter at
times of maximum cyclonic activity.
UPPER-AIR WINDS
It was pointed out in the section on upper-air pressures that the
Arctic anticyclones are relatively shallow phenomena, and hence there
is a tendency for the direction of the pressure gradient aloft to reverse
itself at comparatively low altitudes. (Compare Figures 29 to 32 and
33 to 34 .) Pilot balloon observations over the Arctic Basin have, in
all cases, confirmed this tendency toward reversal. Sverdrup [ 43 ] has
shown that the predominance of easterly winds at the surface is confined
largely to the levels below 10,000 feet, and that they are replaced by
prevailing westerly winds at higher altitudes above the region of his
observations. This general turning of the wind to a westerly direction
is dynamically in accordance with the observed fact that the average
temperature within the lower atmospheric layer s of the west winds decreases toward the
north from the region represented by the upper-air soundings. It is
admitted, however, that the average wind aloft over other portions of
the Arctic may be from directions other than westerly. (See Fig. 39 .)
Variation of Wind with Altitude . — Some aspects of the vertical
distribution of winds in the lower layers of the Arctic atmosphere have
already been discussed in the section on surface winds. It was pointed
out that there exists a strong tendency for a clockwise turning of the
wind direction from the ground upward through a sharp surface temperature
inversion. The data show, however, that this turning is largely confined
to the lowest five or six hundred feet of the atmosphere. In the region
of Sverdrup’s observations [ 43 ] , it is found that for surface wind
directions of east, southeast, or south, a much less rapid clockwise
turning of the wind continues above the surface inversion layer. For
all other surface winds, the direction of turning is counterclockwise.
A very rapid turning between 26,000 and 30,000 feet is noted by Sverdrup
in the data which correspond to surface wind directions from southwest
(through west) to northeast. He concludes that this accelerated turning
marks the transition from a tropospheric to a stratospheric circulation.
These results, however, cannot in any way be taken as representative