MERIDIONAL HEAT TRANSPORT
BY WIND-DRIVEN GYRES
IN THE OCEANIC MIXED LAYER
by
FRANCIS BRUNO TOURNEUR
Eng. Degree in Marine Environment
Ecole Nationale Supdrieure de Techniques Avancies
1986
DEA in Meteorology and Oceanography
Universit6 Paris VI
1986
SUBMITTED TO THE DEPARTMENT OF METEOROLOGY
IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE
OF
MASTER OF SCIENCE
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
September 1989
Signature of
Author
Departmetf
i
Earth, Atmospheric, and Planetary Sciences
July 30, 1989
Certified by
Regiti~ld E. Newell, Professor of Meteorology, Thesis Supervisor
Accepted by
Tom Jordan, Chairman, Departmental Committee on Graduate Students
MIT LIBdARIES
LUn~
This thesis is dedicated
to the memory
of
Susan Ary
"Nothing living should ever be treated with contempt.
Whatever it is that lives, a person, a tree, or a bird,
should be touched gently,
because the time is short."
Elizabeth Goudge
-3-
MERIDIONAL HEAT TRANSPORT
BY WIND-DRIVEN GYRES
IN THE OCEANIC MIXED LAYER
by
Francis Bruno Tourneur
Submitted to the Department of Meteorology
in June 1989, in partial fulfillment of
the requirements for the degree
of
Master of Science
ABSTRACT
Oceanic temperature profiles covering the North Atlantic and North Pacific Oceans from
10N to 56 N for the period 1949 - 1979 are processed and a seasonal climatology of the
mixed-layer depth of the ocean is determined. It is subsequently used to see how much of
the seasonal cycle of the total meridional heat transport within the top 300 meter of the
ocean could be explained by a horizontal zonal overturning cell in the oceanic mixed
layer. It is shown that the seasonal signature of this transport is partially retrieved and we
identify the regions and processes that are dominant in this approach.
Thesis Supervisor: Reginald E. Newell
: Professor of Meteorology
Title
-4-
ACKNOWLEDGEMENTS:
I would like to take this opportunity to sincerely thank my advisor, Professor Reginald
Newell, whose broad interest and unflagging enthusiasm for science have been of great
inspiration throughout my stay at MIT. Without his encouragements, care, constant
availability, this thesis would have never been started. My thanks also go to many friends,
colleagues and great people I met during this year and a half at MIT. They were there in
the good times and the bad. Their moral support and friendship were precious to me. I
would like to name here : Jane Hsiung, Ron Miller, Jane Mc Nabb, Nilton Renno, Jim
Rydock, Steve Walker, Shuntai Zhou, Wu Zhongxiang, and many others. I really
appreciated the long distance phoned friendship from Jim. Special thanks go to Mrs
Dorothea Franzosa for opening her home to me and giving me the heartwarming
atmosphere of a family.
Finally the person I wish to thank most is Susan Ary, to whom this thesis is dedicated.
-5-
Table of Content:
Abstract
Acknowledgements
Table of content
List of figures
I. Introduction
II. The Data Sets
2.1) COADS
2. 1.1) Data density
2.1.2) Seasonal fields of SST
2.1.3) Wind system over the North Atlantic and North Pacific
2.2) MOODS
2.2.1) Data density
III. Computation of the mixed-layer depth
3.1) Preliminary treatment
3.2) Filtering procedure
3.3) Interpolation
3.4) Results and discussion
3.4.1) Basic physics of the mixed layer
3.4.2) Results
3
4
5
6
9
12
12
13
14
15
15
17
19
24
26
27
27
29
IV. Meridional heat flux calculation
4.1) Introduction
4.2) Computation of the standing gyre oceanic heat transport
4.3) Derivation of the current speed from the wind data
4.4) Water density calculation __35
4.5) Preliminary results
4.6) Mass Balance conservation_32
32
33
34
4.7) Adjusted results
43
4.8) Other possible contributors
46
V. Conclusion
VI. Figures
VII. References
VIII. Annexes
35
47
49
83
6
-6-
List of figures :
Figures la, 1b, 1c:
Number of reports per month for respectively the Atlantic, the Pacific and global oceans.
(from COADS documentation).
Figure 2
Total number of individual SST measurements in each grid square for January over the 1854 1948 and 1949 - 1979 periods. (Atlantic and Pacific).
Figure 3
Number of years with SST observation for January over the 1854 - 1948 and 1949 - 1979
periods. (Atlantic and Pacific).
Figure 4
Sea surface temperature over the Atlantic Ocean : winter, spring, summer and fall.
Long-term means over the 1949 -1979 period.
Figure 5
Sea surface temperature over the Pacific Ocean : winter, spring, summer and fall. Long-term
means over the 1949 -1979 period.
Figure 6
Amplitude of the SST seasonal cycle, based on long-term means over the 1949 - 1979 period.
(Atlantic and Pacific Oceans).
Figure 7
Long-term averaged seasonal wind over the North Atlantic Ocean. Winter, spring, summer
and fall.
Figure 8
Long-term averaged seasonal wind over the North Pacific Ocean. Winter, spring, summer
and fall.
:
Figure9
Number of vertical soundings in the North Atlantic basin in January and July 1949, 1964 and
1973.
-7-
Figure 10
Number of vertical soundings in the North Pacific basin in January and July 1949, 1964 and
1978.
Figure 11
A sampling-based rejection criterion.
Figure 12
Mixed-layer depth climatology for the North Atlantic Ocean. Winter, spring, summer
and fall.
Figure 13
Mixed-layer depth climatology for the North Pacific Ocean. Winter, spring, summer
and fall.
Figure 14
Amplitude of the mixed-layer depth seasonal cycle, based on long-term means over the 1949 1979 period. (Atlantic and Pacific Oceans).
Figure 15
Long-term standing gyre heat transport contribution and standard deviation for the Atlantic
Ocean.
Figure 16
Long-term standing gyre heat transport contribution and standard deviation for the Pacific
Ocean.
Figure 17
Zonal averages of meridional transport, 0-300m for the Atlantic and Pacific Oceans. (from
Hsiung et al 1988).
Figure 18
Distribution of the standing gyre contribution for selected months over the North Atlantic
Ocean. (Climatologies).
Figure 19
Distribution of the standing gyre contribution for selected months over the North Pacific
Ocean. (Climatologies).
-8-
Figure 20
Idealized seasonal cycle of the mixed-layer depth and the northeasterly winds in the upwelling
areas off Africa and South California.
Figure 21
Long term mass imbalances and standard deviation for the Atlantic Ocean.
Figure 22
Long term mass imbalances and standard deviation for the Pacific Ocean.
Figure 23
Main path of the Kuroshio and Gulf-stream systems. From Hideo Kawai 1972.
Figure 24
Boundary current and other zones used for mass balance correction. ( Atlantic and Pacific).
Figure 25
Long-term adjusted standing gyre heat transport contribution and standard deviation for the
Atlantic Ocean.
Figure 26
Long-term adjusted standing gyre heat transport contribution and standard deviation for the
Pacific Ocean.
Figure 27
Long-term sea surface temperature anomalies for the North Atlantic and North Pacific Ocean.
-9-
) INTRODUCTION
Recent concerns about the Earth's climate have brought out the necessity to better understand the role that the
ocean has in carrying from the equator to the pole the heat required to balance the ocean-atmosphere system.
At the present time, although many estimations have been made, using various types of methods (Bryan
1962, Budyko 1963, Oort and Vonder Haar 1976, Hsiung et al 1989), much remains to be determined about
the magnitude and variability of its contribution. In this study, we will investigate the seasonal cycle of the
heat transport by the oceans.
The ocean heat transport is a transfer of thermal energy and its seasonal or interannual variations are likely to
depend on features that are closely related to the oceanic mixed layer, since it is the region of the ocean the
most subject to smaller time-scale fluctuations. The net heat transport is basically the result of an exchange
of water masses, having different temperatures, between latitudes. A positive flux will then be driven by a
northward flow of water at one temperature and by a southward flow at a lower temperature. The question to
be addressed is at which depths those exchanges occur. Do they all take place within the surface mixed layer,
as part of the general surface circulation imposed by the large scale horizontal gyre forcing, a kind of
horizontal zonal overturning, or do the poleward or equatorward flows necessary to account for the observed
seasonal cycle occur at different depths ?
It is well known that strong cooling coupled with enhanced evaporation forces at higher latitudes in the
Atlantic ocean for the Northern Hemisphere an equatorward flow of colder, denser water. However, it is
clearly improbable that this deep water circulation may affect the magnitude of the heat fluxes on a seasonal
or interannual basis. A decadal timescale would be more appropriate. As a matter of fact, any flow occurring
below the so-called seasonal thermocline is unlikely to produce any changes over a season or so. Hence we
are dealing with only the top 300 meters of the ocean. What are now the water mass circulations to invoke
-10-
within that 0-300 meter layer in order to explain our seasonal cycle ?
Undoubtedly any features within the mixed layer of the ocean are of major importance for time-scales of the
order of a season. The mixed layer is the interface between the atmosphere and the interior ocean and any
forcing, of mechanical or thermodynamical nature, has to go first through that layer, before it can act upon
the water at deeper levels. The mixed layer removes in the meantime a significant part of the energy supplied
by the atmosphere and only that which remains can be transferred further down with more or less delay. It is
unclear, however, how much of the seasonal variability can be explained by contributions within the mixed
layer only. This will be the topic of the present study.
We will tackle this question by using an approach classical in meteorology, where it is known as the standing
eddy formulation. In the present case, it will consist in computing the transport of heat due to correlations in
the horizontal plane between the temperature and meridional mass transport fields within the mixed layer of
the ocean. The mass transport field will be determined from surface wind speed data, by means of an
observations based relationship. Those wind data are drawn from the Comprehensive Ocean Atmosphere
Data Set, which also provides the sea surface temperature data. As part of the mass transport calculation,
mixed-layer depths will be determined by processing vertical oceanographic profiles from the Master
Oceanographic Observation Data Set. The North Atlantic and North Pacific oceans were selected for this
study and more precisely the latitude band between 10'N and 56'N. The reasons for choosing those limits are
various. The first one is a better coverage over a sufficient long period of time, namely 1949 - 1979. The
second one is that the forcing of the gyre is a subtropical feature. Besides the relationship mentioned above
does not hold at latitudes within ± 10' from the equator, and finally beyond 56'N the criterion that we applied
for our mixed-layer depth determination is less valid.
Chapter II of this thesis gives a description of the history, content and quality of the data sets. Long term
seasonal means of the sea surface temperature and wind fields are also presented as support for further
-11-
references. Chapter III details the procedure used to derive the mixed-layer depth data and discusses the
main mechanisms that control its evolution. The associated long-term means are provided and the reader
may find in annexe 2 the corresponding standard deviation as well as some other informations useful to get a
feel of the quality of the derivation. Chapter IV deals with the meridional heat flux computation outlined
before. A description of the method and assumptions is given and a specific treatment of the western
boundary current contribution is made. Finally the conclusion summarizes the principal results.
-12-
II) THE DATA SETS
In this study, data from the Master Oceanographic Observation Data Set, hereafter referred to as MOODS,
were used for computing the mixed- layer depth for the North Atlantic and North Pacific Oceans. Data on
sea surface temperature and wind speed were derived from the Comprehensive Ocean Atmosphere Data Set
(COADS). We shall describe in this chapter the most important features of these data sets and present the
preliminary processing procedures that lead to the final quantities used as input in our "standing gyre"
approach of the meridional oceanic heat transport.
2.1) The COADS data set:
About seventy million of weather reports taken near the ocean surface primarily from merchant ships have
been included in building this data base . It covers a 126 year period, from 1854 to 1979. The basic data
sources which were integrated into COADS are listed below :
Input
Atlas
HSST
(Historical SeaSurface
Temperature Data Project)
Old TDF-l1 (SupL B & C)
Monterey Telecommumcation
OceanStation Vessels, and
Supplements
Marsden Square 486 Pre-1940
Marsden Square 105Post-1928
National Oceanographic Data Center
(NODC) Surface, and Supplement
Australian Ship Data (file 1)
Japanese
Ship Data
DIMPC (intemational Exchange)
South African Whaling
Eltanin
'70s Decade
IMMPC (international Exchange)*
Ocean Station Vessel Z*
Australian Ship Data (file 2)*
Buoy Data*
'70s Decade Mislocated Data*
Million
Source
38.6
NCDC
25.2
7
4
NCDC(Germany)
NCDC
NCDC
0.9
0.07
0.1
NCDC
NCDC
NCDC
2
0.2
0.13
3
0.1
0.001
18
0.9
0.004
0.2
0.3
0.003
NCDC
Australia
NCDC
NCAR
NCDC
NCDC
NCDC
NCDC
Australia
NCDC
NCDC
100.**
Additions solely to 1970- 1979decade.
The approximate total includes 26.58 million relatively certain
duplicates, and someseriously defective or mis-sorted reports,
which, were removed byintial processing steps.
(From Coads Documentation)
-13-
Within each report up to eight observed quantities have been summarized on a monthly scale in 2 degree
latitude by 2 degree longitude squares. They are : sea surface temperature (SST), air temperature, sea level
pressure, total cloudiness and specific humidity, the eastward and northward components of the wind speed
taken at the typical level of 10 m above the ocean surface. For each of these monthly averaged values, the
standard deviation and number of observations available were provided.
2.1.1)Data density :
The number of reports per month over the 1854-1979 period is shown on figure 1, for respectively the
Atlantic Ocean (a), the Pacific Ocean (b) and the global ocean (c). Large fluctuations appear, with for
instance a peak at about 230 thousands of reports per month in 1968 and a stabilization in the 1970's at about
150 thousands observations per months if we look at the global ocean curve. The large increase in the mid
60's is due to the input of data from countries linked by a World Meteorological Organization agreement to
collect marine reports. The decrease in 1970 might be due in part to the time lag for entering the up-to-date
TDF- 11 data in this data set. Besides the input from the Soviet Union seems to have been stopped at that
time also.
In terms of degree of coverage in space, the 1949-1979 period is overwhelmingly better described than the
1854-1948 one, with the bulk of the data coinciding with the well traveled North Atlantic and North Pacific
Oceans. Beyond 65'N or south of 20'N the data are concentrated along narrow ship routes with large data
gaps in between.
Figure 2 shows the total number of SST measurements in January at each grid square for both periods.
January is generally the month with the least number of observations. Bearing in mind that the values for the
first period are actually sums over 95 years as opposed to the 31 years of the second one, the density in space
is in average about 20 times better during the latter period for the Atlantic Ocean and even more for the
-14-
Pacific Ocean. A maximum number is found off the coasts of Virginia during the 1949 -1979 period, while
its analogue for the previous period is located at the mouth of the French Channel, where different ship routes
are converging. In the Pacific Ocean, the maxima occur off Nagasaki (Japan) with about 2900 measurements
during the first period and off San Diego with approximately 8000 observations for the second period.
In the average the Atlantic Ocean is slightly better described, with gradients for the spatial density less sharp
than for the Pacific Ocean. These differences are strengthened if we consider the first period.
We also present on figure 3 the number of years with observations at each grid square for the month of
January as an indicator of the quality of the temporal coverage at that location. In the Atlantic, about 75% of
the basin has a full temporal coverage for the latter period, while this percentage drops to nothing if we
considered the former period, and only one quarter enjoys a 50% temporal coverage. In the Pacific, if we
look at the years 1854 to 1948, no square presents a 75% coverage (i.e 72 years with data) and only a few
percent reach the 50% level. About 45% of the boxes have a full coverage during the next period, but with
an averaged number of observations per year less than for the Atlantic.
In view of the previous remarks, it was decided to limit the study to the 1949-1979 period, provided that this
period also presents a reasonably good density of data both in time and space for the oceanographic data set
(MOODS), which will be the case but to a lesser extent as we shall see nexL
2.1.2)Seasonal fields of SST:
We present in figure 4 and 5 the climatologies associated with the SST fields for each season and basin, as a
support of our heat flux investigation.
-15-
The influence of latitude, currents and upwelling is evident. Along the African and Western American
coasts, intense upwellings give -anomalously cold waters compared to the zonal means. In addition,
advection of cold water within the surface layer may play a role also. The thermal signature of the Gulf
Stream and Kuroshio is clearly apparent, with a tilting of the isotherms towards the North East sharper than
that which one would expect otherwise.
Without temperature advection through a current or upwelling, the distribution of the SST would be zonal as
is the case in all the regions that are not altered by the above factors.
Strong seasonal variations are encountered in the northwestern side of both oceans, with amplitudes of more
than 12'C, as illustrated in figure 6. This feature can be ascribed to strong continental influences. Besides,
the annual variation of SST in the upwelling regions follows the variation in wind intensity as we shall see
next.
2.1.3) Wind systems over the North Atlantic and Pacific : (figures 7 and 8)
Both oceans are in summer under the influence of a basinwide anticyclonic wind circulation. In the Pacific
this gyre extends from the American coast to about 140'E where it joins the northward wind system
associated with the Asian summer monsoon. It is centered at 150'W, 37'N. Its Atlantic homologue covers
the whole width of the basin with a center around (35'W, 35'N). In winter, this anticyclonic circulation
weakens and its influence on the basin is replaced by that of a now stronger cyclonic system located above 30
N and centered at about 52N, 182'W in the Pacific Ocean and around Iceland for the Atlantic Ocean.
2.2) The MOODS data set:
The Master Oceanographic Data Set is the result of a four year program of gathering and processing
temperature and sound velocity depth profile observations, undertaken under the monitoring of the Fleet
Numerical Oceanographic Center (FNOC). Under its auspices more than 3.5 million of MBT, XBT, AXBT,
-16-
acronyms and others can be
hydrocast and STD records have been collected. The meaning of the previous
update levels that
found in appendix I. We summarize in the table below all the sources and corresponding
subsurface
have contributed to make out of MOODS the most complete and up-to-date set of oceanographic
data until 1981 or so.
SOURCES
NUMBER OF OBS.
YEARS - UPDATING
--------------------------------------- - - --
MBT100
CSJ30M
FNOC MESSAGE
NODC MBT
NODC SDII
FNOC XBT
NODC STD
BRITISH MBT/XBT
NODC CARD MBT(WHOI)
AUST OLD
NEW AUST MBT DATA
NODC XBT
NEW AUST XBT
JAPAN MBT/XBT
LAMONT XBT
ARGENTINA
JAPAN
NODC SDI
SOUTH AFRICA
JAPAN
NEW FRENCH MBT/XBT
KOREAN HYDRO
JAPANESE PUBLISHED DATA
FOSTER STD
OLD FRENCH MBT
NORWAY HYDRO
NODC KRUNCH STD
161,000
1,602
1,108,749
963,994
623,067
211,735
1,715
132,611
14,752
3,284
58,746
338,739
20,316
56,801
121
797
34
17,621
12,900
25
35,358
18,817
200
89
8,903
1,858
8,286
3,802,120
1942-1968
1958-19791965-1981/2
1942-1980
1980
1963-1981/7
1967-1977
1978/4
1943-1959
1960-1972
1943-1979/10
1980 1967-1979.
1970-1978
1976
1977-1978:
1964-19751923-1978
1959-1969
1969-1973
1952-1975/10
1961-1977
1961-1966
1975-1976
1952-1967
1948-1957
1972-1978
-.
-17-
2.2. 1)Data density:
Although it is the most complete data set of its kind available up to 1981, MOODS presents some serious
weaknesses due to the lack of data. The bulk of the soundings is concentrated in the Northern Hemisphere
with about 90% of the total number of observations. The maps on figure 9 to 10 are intended to give an idea
of the distribution of the data both in time and space. We selected the years 1949 and 1973 as typical of the
least covered, and the best covered years in the Atlantic respectively. 1964 stands here as representative of a
mean coverage. 1978 was the best described year for the Pacific Ocean.
A number of interesting features emerge from these maps. There is a significant increase with time of the
number of observations gathered. In addition, more soundings are made during the summer than during the
winter. The best covered period spreads over June, July and August, which we have taken to be the summer
season. On the contrary, January is the least favoured month of the year. The decrease in the number of
observations in the later years results primarily from the delay with which new collected data are
incorporated into the data set. The Atlantic Ocean is generally better described than the Pacific Ocean.
This data set presents in addition some very distinct aspects as opposed to a collection of surface
observations. The latter can generally be made by merchant ship in an extensive and routine manner and
present an upward trend with regards to the number of observations with time, generally everywhere, except
for particular locations and over limited periods of time (e.g. wars...), while the former shows highly
fluctuating data distributions both in time and space, with, however, still an upward trend, but a very jagged
one, even on a monthly scale. Certain areas, even large ones, could be observed for the purpose of some
localized scientific interest only during a given month over, say, 5 years and never again afterwards.
Generally we have observed a significant correlation between the time of the year a sounding was performed
and its location. This makes any attempt of studying annual cycles on a monthly basis strongly biased.
-18-
Therefore at the most only a seasonal basis could be considered, providing that the quantities to be derived do
not vary a lot from a month to an other.
Dealing with this type of data set presents nevertheless some advantages. One of those is an improved
accuracy. Our temperatures at depth were given with an accuracy of about 0.06'C, as opposed to typically a
few tenths of a degree centigrade for COADS for which data were collected by ships of opportunity on their
way to their usual business of commerce. On the other hand, it should be noted that the mixture of soundings
from different origins, countries or scientific institutions, the various methods or instrumentations used yield
potentially inhomogeneous data, not to mention the large discrepancies observed in the vertical samplings.
-19-
III) COMPUTATION OF THE MIXED-LAYER DEPTH
3.1) Preliminary treatment:
Altogether about two millions of profiles have been analyzed for both the Atlantic and the Pacific Ocean.
Any measurement made within the top one meter of the ocean was considered as a surface temperature data.
For each of those soundings, the mixed-layer depth was taken to be the level at which the temperature was
0.5'C less than the surface temperature. This criterion seems appropriate to describe the temperature of an
isothermal oceanic upper layer to which the mixed layer could be identified. This does not hold at high
latitudes for instance, as we shall see later on.
To determine the depth of the mixed layer, we linearly interpolated the temperature between the levels
available. By using this simplified vertical description, we strengthen the effects that the sampling could
introduce on the accuracy of our depth determination, making necessary some adapted data quality controls.
First of all we asked the profiles to meet certain basic requirements, such as to supply surface temperature
data, or to extend to a level at which the temperature is at least 0.5 C less than its surface value. Besides, we
retained only the profiles that presented at least three levels, one of those having to lie within the 0-300 m
layer. Those objective criteria are by no means self-sufficient. They yield, however, the elimination of an
unexpectedly large number of soundings, as evidenced by the following tables for each of the oceans under
study :
-20-
a) Atlantic Ocean:
Month
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
Number of data processed
per month
Number of data
discarded
Rejection
rate (%)
45,791
55,470
67,153
65,274
85,241
87,807
89,864
76,213
74,987
70,103
67,884
42,152
15,919
20,426
25,100
18,725
11,771
3,970
2,358
2,366
4,657
6,057
9,433
9,170
34.8
36.8
37.8
28.7
13.8
4.5
2.6
3.1
6.2
8.6
13.9
21.8
827,939
129,952
15.7
b) Pacific Ocean:
Month
Number of data processed
per month
Number of data
discarded
Rejection
rate (%)
----------------------------------------------------------------
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
68,410
79,599
91,303
105,382
110,472
110,934
106,932
106,891
101,468
95,582
81,640
55,991
11,533
14,973
13,876
9,736
4,286
1,955
1,423
1,550
2,964
4,513
5,947
7,845
16.9
18.8
15.2
9.2
3.9
1.8
1.3
1.5
2.9
4.7
7.3
14.0
1,114,415
80,601
7.2
-21-
The next constraint we applied was directly related to the sampling interval. A larger or smaller sampling
distance may, indeed, lead to serious uncertainties in the depth determination. In the case shown on figure
11, data are provided at the surface as well as at about 26 m. The depth of the mixed layer is necessarily
above the 26 m level where the decrement in temperature from the surface value is larger than 0.5 C. But it
entirely relies upon whether or not an intermediate level is provided and the accuracy with which it would be
described . The solid line illustrates the case where no intermediate datum is known and yields a mixed-layer
depth of 6.10 meter. Now, depending on which intermediate sampling may be available our determination
could give a priori any value between 0 and 26 meter. The constructions shown yield 3.52 or 24.62 meter.
Therefore it appears necessary to limit the sampling interval to a degree that should depend on the level at
which the mixed-layer depth derived from our 0.5'C criterion is lying.
We will allow a more coarse sampling as the depth increases.
50. x Depth
We chose the simple function:
90. + Depth
For a mixed layer at 6.10 meter, this distance should not exceed 3.17 meter. This index was crudely set to
ascertain a reasonable accuracy and resolution for the expected range of depths to be determined, namely 0 to
300 meters. This choice ensures an accuracy of about 12.5% at around 300 meters (about 38 m), but only
50% at 10 meters (5 m). It is thought to provide a good trade-off between the accuracy and the concern to
keep in the meantime sufficient data.
The rejection rate associated with this additional criterion are given in the following table :
-22-
a) Atlantic Ocean:
Month
Number of data processed
per month
Number of data
discarded
Rejection
rate (%)
--------------------------------------------------------------------
JANUARY
FEBRUARY
MARCH
APRIL
MAY
TUNE
JULY
AUGUST
SEPTEMBER
45,791
55,470
67,153
65,274
85,241
87,807
89,864
76,213
74,987
2,784
4,046
5,776
6,906
10,699
12,841
12,851
9,418
5,446
OCTOBER
70,103
2,993
- 4.3
NOVEMBER
DECEMBER
67,884
42,152
2,331
1,969
3.4
4.7
827,939
78,060
9.4
6.1
7.3
8.6
10.6
12.6
14.6
14.3
12.4
7.3
b) Pacific Ocean:
Month
JANUARY
FEBRUARY
MARCH
APRIL
MAY
JUNE
JULY
AUGUST
SEPTEMBER
OCTOBER
NOVEMBER
DECEMBER
Number of data processed
per month
Number of data
discarded
Rejection
rate (%)
68,410
79,599
91,303
105,382
110,472
110,934
106,932
106,891
101,468
95,582
81,640
55,991
4,537
2,156
3,173
7,073
8,426
13,482
16,391
18,766
17,181
9,630
4,330
2,850
5.7
3.9
4.6
7.8
8.0
12.2
14.8
17.5
16.1
9.5
4.5
3.5
1,114,415
107,995
9.7
-23-
Finally, we estimated that a temperature based criterion would not be appropriate to determine bases deeper
than 300 meters, for it is probable that in the cases of an isothermal thickness of 300 meters or more for the
regions under study (North Pacific and North Atlantic between 10 N and 56 N) we would be in the very
northern parts of the basins where salinity is likely to become an important factor in controlling the
convective stirring process.
The temperature is there typically
uniform over thicker and thicker layer,
Temperature (degree'C)
while a strong halocline may lie near the
surface, as a result of river runoff or
excess precipitation. On the right is
given an illustration of the above
remark.
It is obvious then, that a
temperature based criterion would break
down and a cutoff at 300 meters allows
us to stay within the domain where
55. 19N
48. 46W
1/9/1954
24Hoo
temperature can be considered as a good
index of the depth of mixing.
The next step was to organize the data using 2'latitude by 2'longitude boxes for each season of each year of
the 1949-1979 period. Once this was done, we proceeded to an extensive filtering procedure intended to
eliminate bad data, followed by a 1-step interpolation in order to fill some missing or removed data, while
still avoiding the propagation of "neighborhood effects" or contamination. Both these phases will be
developed next.
-24-
3.2) The filtering procedure:
The first action was to calculate for each season of each year the mean mixed-layer depth at each grid
square. If only one value was available, the square was given the missing value flag. We then computed the
standard deviation from the aforementioned mean.
If the number of data for a given box was less than 10, each of those data was compared to a weighted
average of the mean values obtained for the neighboring boxes. Each adjacent square received a different
weight according to its distance from the central square and its number of mixed-layer depth data. Our
hypothesis was that if that number is large, then the corresponding mean may be more reliable and should be
given a larger weight.
Let Ni be the number of data at box i , Mi the mean value of these data, and Wi the weighted average
attached to this box. This weighted average was computed as follows :
M1
N1
M2
N2
M3
N3
for corner boxes:
Na
No
N4
M7
M6
N6
M5
N5
N7
Wk =
k =1. 3.5,7
-N
',/i
>
k
NY
i1,m
_
Nt
Wt
for others
W=
2.4.6.8
E
i=1,m
weighted average
in
WM,
m
M : number of adjacent boxes with an available mean value.
Ni
The test was performed even if m was equal to 1. If one of the mixed-layer depths for a given box was
different from the weighted averaged mixed-layer depth of the surrounding by more than 15 meters, it was
discarded. If not, it was integrated into the computation of the mean mixed-layer depth for that square.
If now, the number of data for a box was more than 10, but the relevant standard deviation for that box was
greater than 5 m, an additional test was performed. We first assume that the mixed-layer depth data within
each square were normally distributed. Then we computed for each datum of each box a standardized
variable, which measured its deviation from the box mean in units of the appropriate standard deviation,
namely:
Mid -( M d),,
(Std)b,
The standard deviation we are using throughout this study is defined as:
Std =
/N) Ei(ai-
a)
where N is the number of values (ai) taken by the variable a and li their mean.
If #was found to be larger than 1.96 (which should occur for at the most 5%of the data under the Gaussian
assumption) and if in addition that datum fails the weighted average test described before, it was rejected.
The reasons for using two tests are that all the data for a given square may be too far from the surrounding
mean, without being necessarily bad data. They could possibly be ascribed to some smaller scale features at
the lower limit of our resolution capability. On the other hand, even if it is less probable, a particular datum
for a given box could be anomalous for that square but, however,close to the surrounding mean. This would
happen if all the other data for that square were bad or in case again of a smaller scale system. Being unable
to decide whether the data are bad or not for the above cases, it was decided to keep them.
Each time an update was made necessary, the statistical parameters (mean and standard deviation) were
accordingly corrected for further use in the filtering procedure.
All the aforementioned conditions were set after extensive trial analyses and hand checkings to make sure
that all the corrections were justified and that the filtering did not result in a too drastic reduction of the
number of data. We estimate that about 5% of the data were removed.
3.3) Interpolation :
Each missing data was replaced by the weighted averaged mixed-layer depth value computed for the
surrounding boxes, under the condition that at least three adjacent boxes had data.
-27-
3.4) Discussion and Results :
In this section we present the results of the mixed-layer depth determinations, whose procedure has just been
detailed. Long term means for each season and both oceans are displayed. It is beyond the scope of this
study to give a complete and extensive description of the mixed-layer physics and the present treatment has
to be seen within the framework of an investigation of the standing gyre contribution for the upper layer
oceanic heat transport. We will, therefore, provide only a brief outline of the main factors that may control
the mixed-layer behavior and explain some of the features of the mixed layer patterns, which are displayed in
figures 12 and 13.
3.4.1) Basic physics of the mixed layer:
The mixed layer is basically maintained by turbulence.
The ocean may be represented ideally as a two-layer fluid, with denser water in the lower layer and lighter
water above.(This simple picture is associated with the thermohaline circulation induced by a cooling at
higher latitude and a large scale upwelling in the ocean interior, which gives closure for mass balance.) This
structure is particularly stable and turbulent kinetic energy is necessary to locally break down this stability
and mix waters from each layer.
Kinetic energy may arise from wind forcing or convection related to buoyancy forcing. As opposed to the
atmosphere where convection is induced by surface heating, convection in the ocean is driving by surface
cooling, which increases the water density and allows water to subsequently sink with respect to lighter
water. Evaporation, by increasing the salinity, also increases the density. Note that in the atmosphere ,
evaporation also acts as a booster for convection, for it lightens ascending particles, making further ascending
motions easier.
-28-
Surface cooling results from a negative balance between radiational loss by infrared emission, latent heat loss
by surface evaporation and absorption of solar energy, the latter taking place within a very thin layer near the
ocean surface. Therefore convection is enhanced at night since the solar input is lacking.
The action of the surface cooling is to lower the center of gravity of the water column by compressing the
fluid it contains. When this center is lowered, the concomitant potential energy released is converted into
turbulent kinetic energy.
The other factor, we said, is the wind which drives a flux of momentum into the ocean. The air blowing over
the ocean surface is slowed down by friction, while the surface oceanic layer is speeded up. This velocity
shear is used to generate turbulent kinetic energy.
Waves and currents actually make this idealized view less simple. Surface waves are generated by the wind
through the absorption of a fraction of its momentum. This fraction of momentum is thus lost for local
conversion into turbulent kinetic energy, and will be recovered for that purpose only when the waves break,
which may be a long distance further. When the waves break, they also mix air bubbles, which will enhance
convection and turbulent kinetic energy creation. Internal waves affect also the local distribution of the
turbulent kinetic energy, but have no impact on a global scale. In addition, horizontal advection is usually
associated with a deeper mixed layer, since it strengthens the water stirring.
The turbulence generated develops a well mixed upper layer, whose thickening may be inhibited, however,
by upwelling, driven by the curl of the wind stress. At the bottom of the mixed layer, water from the denser
layer of the ocean is stirred with lighter water above, the potential energy of the water column rises. This
increase is made possible by consumption of turbulent kinetic energy. When the supply of turbulent kinetic
energy goes down, a lesser amount of potential energy is available to bring upward water particles from
below the mixed-layer bottom and therefore the stirring that occurs at that interface (which is not a material
-29-
surface) involves lighter water particles. This means that this interface has to rise and the thickness of the
layer diminishes.
This fairly basic treatment of the mixed layer physics will allow us to explain now some of the major features
evidenced on figure 12 and 13.
3.4.2) Results:
The mixed layer presents a uniform or quasi-uniform vertical structure with respect to various parameters. It
is highly fluctuating, with possible time scales as small as a day or less, and therefore it is necessary to use a
fair amount of data to be able to reach an estimation of its vertical extension that is representative of , say, a
seasonal cycle. We believe that for most of our selected areas this level of data was obtained, or the problem
was overcome by use of intercomparison criteria between adjacent 2'x 2squares. The reader will find in
annexe 2 a statistical description of our data under the form of number of observations and standard
deviation, that should give a feel of the locations where the level of seasonal representativeness may not be
met.
The mixed layer does vary with latitude, longitude and season, from around 200 m for the North Atlantic
winter to less than 15 m for the North West Atlantic and Pacific summer. The maximum thickness is reached
in winter and decreased then rapidly to the summertime minimum value. This change is an illustration of the
stabilizing effect of the solar heating. Our seasonal study can not show evidence of the intermonthly
changes, which are, however, very important. The erosion of the mixed layer between the maximum value at
the end of the wintertime cooling and the summertime minimum needs only 3 months (April to June) to
complete. Outside this sharp variation period, the evolution is slow. The amplitude of the seasonal cycle
(figure 14) may reach more than 130 m in both basins, with the areas of maximum variation coinciding
generally with those of maximum wintertime depths. This is a consequence of a fairly uniform small
-30mixed-layer depth in the summer (25 m for latitudes above 20'N). If we compare figure 14 with figure 6, it
appears that areas of maximum annual mixed-layer depth variations are not areas of maximum annual SST
changes, for surface cooling or warming is not the only factor in controlling the mixed layer thickness.
The Atlantic maps show, in addition, a band of higher mixed layer values spreading accross the basin from
the North East to the South West, above 307N. This is consistent with the study made by Lamb (1984). This
pattern can be explained in part by the fact that this area is located at the confrontation line between polar air
and warmer equatorial air masses. Southward incursion of cold air, also dry for it blows over continental
areas, yields strong surface cooling and subsequent mixed layer deepening. Note that the Gulf Stream region
has as a matter of fact the highest evaporative loss in the world ocean (Hsiung 1983). In addition this water
belongs to the high salinity water masses of the Sargasso sea. Therefore the variations of density with
temperature will be sharper, leading to more powerful vertical displacements. An.other favoring element in
driving convection within the Sargasso sea is its slow motion, which will lead to enhanced evaporation and
buoyancy fluxes.
On the northern side of this band, lower mixed- layer depth values are due to an increased stratification that
partly inhibits convection. This stratification results from the presence of relatively fresh water in the upper
layer in consequence of river runoffs or excess of precipitation over evaporation.
In both oceans the areas of upwelling are clearly identified by patterns of minimum mixed-layer depth values,
illustrating the inhibiting effect of vertical transports on the depth of mixing. They hug the coasts of Africa
and Western United States. Again we find a minimum in mixed-layer depth concomitant with the peaking of
the upwelling activity in summer, with typical values of less than 25 m at that time.
The region of maximum mixed-layer depth in the Northern Pacific Ocean in winter and spring centered
around 180'W, 43'N coincides with stronger winds and is far enough from the coasts so that the inhibiting
effect of fres, water is not felt.
-31-
Comparisons with previous studies are made difficult by the fact that usually each author has his own
definition of the mixed layer and hence applies to its determination some specific criterion. Levitus (1982)
based his mixed-layer depth computation on a AT=0.5 criterion. However his results are given for March
and September, which for our study are pivot months and hence their particularities are somehow swallowed
into the averaged features of the seasons that they belong to. This is a problem especially for March, after
which the mixed-layer depth is expected to decrease very rapidly. Note that Levitus (1982) had in his
disposal only half the amount of data we used in the present investigation. However, with the above
restrictions, both studies bear some resemblance, with the difference that his values are generally larger than
ours. It might be due only to the length of the period over which the temporal averages where computed.
This raises again the question on how many data one needs to be able to determine the mixed layer depth on a
monthly basis, without being too much affected by the impact of local and sudden storms that may induce a
strong deepening of the mixed layer for a short time.
-32-
IV) MERIDIONAL HEAT FLUX COMPUTATION:
4.1) Introduction
Our purpose in this chapter is to investigate the impact that the atmospheric gyre scale circulation exerts on
the meridional oceanic heat transport and to go towards a better understanding of the seasonal cycle of this
transport as described in Hsiung et al (1989).
The idea underneath our study is fairly simple. The oceanic circulation and vertical temperature profiles do
not undergo seasonal changes below a depth where the so-called seasonal thermocline lies. Above, changes
in wind forcing or solar heating induce marked seasonal fluctuations. The Ekman theory is then of major
importance in parameterizing the oceanic transport that results from the wind stress distribution. However,
this classical theory fails in giving a good representation of the depth to which the Ekman based transport
takes place, believed to be in general much smaller than the actual mixed-layer depth, not to mention the
much too exagerated decay of the current with depth predicted by that theory. Thus any attempt to estimate
meridional heat transports in the upper layer of the ocean from an Ekman point of view would yield a large
underestimation of their magnitude and encourage us to disregard their contribution in oceanic heat budget
calculation (De Szoeke et al. 1981).
We therefore decided to use mixed-layer depth data, surface winds speeds and surface temperatures to
compute what is felt to hold one of the strongest relative seasonal cycles among all the terms that work
together to build the oceanic heat budget, namely a correlation between the meridional mass transport and
temperature fields within the mixed layer of the oceans as a result of zonal circulation. An empirically
deduced formula (Pickard and Pond 1983) was used to compute the surface current velocities, assumed to be
uniform within the mixed layer.
-33-
4.2) Computation of the standing gyre oceanic heat transport:
We define by standing gyre heat transport the term that in the heat budget represents the correlation in the
horizontal plane between the temperature and the northward mass transport fields within the mixed layer,
namely:
/
Xeastf
Hg =jC
(1)
Xwest
-T* ( PVdz )*dx
-10)
where o is the depth of the mixed layer, p the density of seawater, v the current velocity,
T
the temperature, z
the vertical coordinate, Xthe longitude, 0 the latitude, and c the specific heat of seawater taken as 4218
J/KgK.
(.)denotes the deviation from the zonal average.
At this point several assumptions are made. First we suppose that the temperature is uniform or
quasi-uniform within the mixed layer. This was indeed the criterion on which we based our mixed-layer
depth computation. In addition, we assume that the density is independent of depth within this layer. It is
therefore inferred that the salinity has to be quasi-constant also. Since we derived the mixed-layer depth for
only those locations that were not controlled by salinity, the above requirement is implicitly satisfied.
In addition we made the hypothesis that the mixed layer does not present any vertical velocity shear. In a
first approximation, this is justified and many authors have been using it (Schopf and Cane, 1983). However,
stirring mechanical quantities, like momentum, is different from mixing thermodynamics properties. Some
authors have as a matter of fact shown evidence of significant velocityshears within the well mixed layer.
Niiler (1977) and Pollard (1977) found that the assumption of vertical uniform horizontal velocity does not
-1
-2
shold for shear in excess of 10 - 10 sec , which is not unusual in the ocean ( Halpern (1980)). Nevertheless,
-34-
because of turbulent momentum transport, shears are likely to be smaller. To view the mixed layer as a slab
appears, however, as a better approximation than the classical picture of an e-folding velocity decay over a
depth corresponding to the Ekman layer depth.
Finally T and V were taken as the surface values.
With these assumptions, expression (1) can be reduced to:
H,
C-
T (PVD)*dX
Xwest
4.3) Derivation of the current speed from the wind speed data:
Observational studies of surface current speed have shown that to a reasonable approximation the current
velocity Yis related to the wind speed k by the following relationship: (Pond and Pickard 1983).
0.0127
(sin|$[)1/
2
outside 1 10 latitude from the equator.
We furthermore assume that the water transport crossing a zonal surface elementPVdz dXwas directed 30 to
the right of the surface wind direction. 45 is the angle given by the Ekman theory. However , due to the
limitations of the theory, the expected 45 angle is rarely observed in nature and the rotation direction is likely
to be somewhat smaller. We took 30 for an estimated averaged most probable direction. A trial analysis
showed that the angle was of minor importance on our results, but this point deserves further study.
Thus, the meridional elementary water transport became:
PVdz dX =
0.0127
(sin(4[V
where ' is the angle associated with the wind direction.
- /w/
. sji
( 4-30-) dz dX
-35-
4.4) Water density calculation:
For each square Pwas computed using the surface temperature available at that box. If no temperature was
present, the box was given the mean density of its zonal belt. Since no data on salinity were ready to be used,
we took for each box a salinity value that was an estimation of its mean over the whole basin, namely 34.5%
for the Pacific and 36% for the Atlantic.
The formula applied for its computation is given below:
9= 999.842594
0.06793952
0.009095290
0.0001001685
0.000001120083
0.000000006536332
0.824493
0.0040899
0.000076438
0.00000082467
0.0000000053875
0.00572466
0.00010227
0.0000016546
0.00048314
T
T2
T3
T4
T5
S
S x T
S x T2
S x T3
S x T4
S:Y
Sc T2
S2
From Pond & Pickard 1983.
4.5) Preliminary results:
The discussion and results presented in this section are understood as incomplete, since they have been drawn
without the necessary mass transport conservation principle to be satisfied. However, it already provides a
good insight at some of the physics involved in this standing gyre approach for the meridional heat transport.
Most of the features we shall present here will not be denied later on by application of the zonal mass balance
criterion. In the next section, that mass balance will be forced by introducing into our model a western
boundary current. Since then, the velocities for this current are likely to be quite large compared to interior
-36-
ocean surface velocities, and the anomaly of temperature positive, we shall generate a rather strong possible
correlation between our temperature and mass transport fields, due to the occurrence of positive temperature
anomalies in that western boundary area. This will have a strong impact on our initial results and hence any
features which appear within the belt having a boundary current correction should not be considered as
conclusive at this stage but may be used as a reference to assess the ultimate impact of the mass balance
adjustement.
Figure 15 and figure 16 contain respectively the results for the North Atlantic and North Pacific, in terms of
long term means over the 1949 - 1979 period. In addition the corresponding standard deviations are
provided. They show a maximum at high latitudes in wintertime due to sparse data especially during that
season. This yields fairly uncertain heat transport values for those regions, which are the only locations
where negative values are obtained : 30 N - 38 N in winter, above 38 N the rest of the year for the Atlantic,
30'N to 37'N in winter, 40'N to 48N in spring and summer for the Pacific. Below those latitudes, values are
positive and show a maximum in May for the Atlantic. For the Pacific, and bearing in mind the uncertainties
on our data, the maximum is found at around 30 N somewhere between April and September, without a
particular month to present a significant peaking value.
The computation of heat flux from Hsiung et al (1989) reproduced on figure 17 show a relative maximum in
September for the Pacific. Note that data on figure 17 represent the total meridional heat transport as
opposed to ours . They differ as a matter of fact by a factor of 10 or so. From just wind or temperature
consideration, we would expect a stronger correlation to occur in summer since as shown on figure 4 the
isotherms are very much tilted and the development of the gyre is peaking. We have to keep in mind that it is
actually the tilt of the isotherms on the edges of the basin , where the strongest winds are expected with
opposite directions on each side, that only matters. Remember that a positive correlation means colder water
moving southwards and warmer water moving northwards. However, the depth of the mixed layer is reduced
by a factor of 4 between winter and summer, which yields a modulation of the same rate or so on the heat
-37-
transport. Therefore maxima somewhat shifted in time-from a summer occurrence do make sense.
The transports accross latitude belts have about the same size in the Atlantic and Pacific. However about
three times less heat per square meter is carried northwards in the Pacific than in the Atlantic through
correlation between the temperature and mass transport fields, since with these units our results are
normalized by the the width of the oceans, the Pacific being in the average three times broader than the
Atlantic. Again, this should be confirmed by the adjusted results to really be conclusive.
At this point, it is interesting to look more carefully at the spatial distribution of the correlation term, in order
to get a more precise feeling of the regions that count in each month and of the processes that may explain the
observed distribution. Figures 18 and 19 show those regions and how their importance is modulated with the
time of the year. The most significant result is the overwhelming influence of the upwelling areas, which
actually contribute
more
than 75% to the total value of this heat component. The intensity of their
contribution follows the annual cycle of the strength of the northeasterlies over the upwelling regions, more
powerful in summer or so. This fact would be consistent with a peaking value for the correlation term
occuring in late spring-summer. However the impact from the mixed-layer depth is likely to be the main
controlling factor. Figure 20 shows an idealized picture of the concomitant variations of the surface current
speeds over the upwelling areas and the mixed-layer depth during the year. The mixed-layer depth and
surface speed curves are qualitatively drawn from figures 12, 13 and 8, 9 respectively. For each parameter,
only the month to month variations matter here and the curves are therefore, for conveniency, expressed in
arbitrary units. We note that, while the differences between the currents are not very large between spring
and fall off southern California or off Africa, the differences in mixed-layer depths are indeed clearly larger.
-38--
FIGURE
20.
\
mixed-layer
\depth
surface current
speed
~M
A
I
4
A
a
Note that figures 18 and 19 show also the importance of the western side of the oceans, where continental
influence may drive a strong winter cooling both by latent heat or sensible heat release, since the air blowing
from those land areas is very dry and cold. This feature is more obvious in the Pacific ocean than in the
Atlantic and is consistent with the heat loss investigation of Hsiung (1986).
All those key areas do affect the standing gyre heat component through the tilt of the isotherms that they
drive. The depth of the mixed layer constitutes a modulating factor. The positive peaks obtained are due
primarily to the upwellings, while the negative ones are driven by negative net heat losses on northwards
moving flows. All these processes are as a matter of fact closely related to coastlines proximity and do not
extend over the open ocean over more than 1/4 of its width. Hence most of the open ocean may be neglected
for an exploration of the standing gyre contribution or at least a deep knowledge of the winds, temperature or
mixed-layer depth is not crucial. This is a very important result, since it allows us to concentrate our efforts
of collecting data only on a much smaller area. We did for the present investigation, however, a computation
over the whole basin.
-39-
4-6) Mass balance conservation:
All the previous results, as we mentioned, have been derived while omitting the necessary mass balance
requirement. In order to prevent any accumulation of water in the ocean, we need to ensure a constant mass
flux across every latitude belt. This constant value has to equal the assumed deep water transport, generated
at the surface at high latitudes, when strong cooling coupled with enhanced evaporation initiates sinking of
dense water in the Atlantic. As we saw, the salinity in the Pacific ocean at high latitudes is not large enough
to support a deep convective activity and most of the water that sinks there is unlikely to reach a level below
1000 meters. Thus we will not consider any deep water return flow in the Pacific. However there is a flow
from the Pacific into the Atlantic through the Bering Straits, which is estimated at about 1 Sverdrup.
Therefore we will force a 1 Sv mass transport across each latitude for the Pacific. For the Atlantic, the
strength of the deep flow is more controversal. The range of the estimates goes from a few Sverdrups (L. V.
Worthington) up to 40 Sv (Riser et al. 1978). In this study we decided to follow Worthington's value of 7
Sv, for it was based on basinwide considerations on water masses. Estimates that are made only with local
observations do not include possible deep water recirculation phenomena and may, therefore, be exaggerated
One of the 7 Sv assumed by Worthington for the Atlantic Deep Water comes from the Pacific. Another one
balances the mid-thermocline northward flow, so that the remaining 5 should move northwards within the
mixed layer. We will therefore force a 5 Sverdrup mass transport across each latitude for the Atlantic.
4-6.1) Mass Imbalances - Results:
Figures 21 and 22 show the net mass transport across each latitude and month for respectively the Atlantic
and the Pacific. Note that they are long term means over the 1949 - 1979 period and that the results do not
incorporate the 1 or 5 Sverdrups developed above. We give also the corresponding standard deviations.
They represent basically half the value of the mean.
-40-
The mass imbalance is due to two different phenomena . The first one is inherent to our computation and is
related to the noise of the data. We estimated at 50 %its contribution to the total computed imbalance, by
reference to the standard deviation level. We partitioned these noise imbalances over all the boxes along
their corresponding latitude. The remaining 50 %are related to the net southerly drift of water which results
from the part of the Sverdrup circulation that applies to the mixed layer. Negative values on the figures mean
southward flow. To those values we have to :subtract: 5 Sverdrups, that is 50 10+8 kg/sec or so to get the
actual imbalance. After those corrections, we can see that there is basically no net northward flow in the
Atlantic at any of the latitudes considered, and only below 20'N in winter or 26N in summer in the Pacific.
For both the Atlantic and Pacific, the bulk of the imbalances occurs in winter, at respectively 370N and 33'N.
They represent there a flow of 18 sverdrups in the Atlantic and 32 sverdrups in the Pacific. The summertime
values are very small and account only for the escape flow of water. This seasonal cycle is partly consistent
with the observations of the western boundary currents stronger in winter in their northern parts and weaker
in fall in all regions (Fuglister 1951), although we do not really get minimum values in fall but rather in
summer. For southern sections we get lower values in summer than in winter, which is consistent with
Fuglister's observations. However due to the noise level, especially in winter at those latitudes, the relative
variations may not be meaningful. Besides, we noted a downstream increasing mass transport consistent with
the observed downstream pressure gradient that acts to accelerate the flow. For the Pacific, features are quite
similar with the difference that everything is shifted southwards by about 4 or 5 degrees. The Kuroshio does
as a matter of fact separate from the coast and moves eastwards at a lower latitude than the Gulf Stream (see
for example figure 23).
The assumption we made then was that the negative mass imbalances, meaning net southward flow, are
balanced by a northward flow in a western boundary current. The latitude limits we set for these boundary
currents are 18'N to 40'N for the Atlantic and 14'N to 38'N for the Pacific . Outside these areas the negative
net southward flow was balanced by a northerly drift of water over the whole width of the ocean. We
assumed that in that case the entire belt acts to reduce the imbalance, which is also decreasing as we move
northwards. If a positive imbalance occurred, the corresponding southward flow correction was in a similar
way split over the entire width, since no boundary current, that should implicitly be on the eastern side of the
basin, could make up for such imbalances.
We show on figure 24 the areas over which the corrections were computed. In the narrowest regions of the
stream, the width is about 120 kIn, and increased as we go closer to the outer limits of the domain in order to
take into account a gradually zonal orientation.
The depth we took for the current was not the depth of the mixed layer given by our earlier computation. The
main reason is that the net southward drift within the mixed layer of the open ocean is unlikely to return
northwards with exactly the same physical properties, temperature and salinity, although it is improbable,
however, that those properties will change drastically under evaporative processes and give a return flow
much deeper than the mixed layer depth, already increased in the western boundary current areas, because of
a stronger advection. The depth of the current was taken to be about 50 %larger than the depth of the mixed
layer. In case the boundary current was occupying only one 2 degree grid square, the mixed-layer depth that
we took was the mean value calculated with the value of the next adjacent box to its east, in order to diminish
the dependency on only one data that is potentially affected by noise. The current speed was taken to be
uniform over this depth.
-42-
For each box new velocities were computed by adding the initial ones to those required for mass
conservation, namely : (Zones AB,C are defined on figure 24)
1) zone C and A : (for noise correction)
(gPD2 i-(Is t as
d(D (e1D.)i=wset
de
in rn/sec
where dD is the width of a grid square at latitude@.The velocity was the same for each square.
2) zone C : (for mass balance correction)
VeI(ZC)
V
((D)
I
=
- (1/2-50 )
n ((D) - [ ', Dint
n (() is the number of squares considered for the boundary current at that latitude.
50 is the value we took for the deep water flow transport in kg/sec.
is the mean of ( o) over those squares. The underlying idea is to replace in each square the actual
Dn
value,
by the mean over the squares involved in the current. This is just to avoid a fictitious speed-up of
the current in case of missing data, if the summation F, oi is made only where data is available. Again the
velocity is taken as uniform across the current.
3) zone B : (noise and mass balance corrections)
Vel ()
(eas -50&~
(west
west
i Di
in rn/sec
d
-43-
Finally each initial speed V was corrected by the above velocities, and the final velocities x9are:
zone A + C:
O
zone B:
o
=
V
+ Vel
V
+ Vel
The size of the corrections were of the order of 1 m/sec in the narrowest regions of the stream and a few
centimeters per second elsewhere.
The last step was to recompute the long-term standing gyre heat contribution. The procedure used in the
earlier stage was kept unchanged, only 0 was substituted in place of V.
4-7) Final results:
The long-term averaged adjusted standing gyre contribution over the 1949 - 1979 period is shown on figures
25 and 26 as well as the associated standard deviation, respectively for the Atlantic and Pacific. Compared to
what we obtained previously, the new standard deviations are generally smaller relatively to the means and
the largest values are found in winter at high latitudes, because of sparse data and bigger interannual
variability . We also provide on figure 27 a picture of the evolution of the mean anomalies of temperature
over the boxes that are included into zones C and B (figure 24) at each latitude, in order to give to the reader
a better understanding of the changes that have affected our initial results. As a matter of fact, even if the
boundary current generated as part of the requirements for mass balance conservation is strong, the correction
it will drive is unlikely to modify the original situation in case the temperature anomalies are small in that
region. This fact is fairly well illustrated for the Pacific ocean where large mass imbalances, obtained
-44-
between 22'N and 48'N in wintertime, do not generate overwhelming heat fluxes since the anomalies of
temperature are quite low there, only around 1.35 'C. Briefly our mass balance correction acted to globally
raise our heat fluxes by about 2 to 3 10+13 Watts from their initial values in that ocean.
As regards the Atlantic ocean, the old picture has been much more altered. The boundary correction took
place in a region of higher temperature anomalies which resulted in a 10 10+13 Watts increase for the heat
flux at 30 'N, and gave a second peak between January and May at that latitude. In addition, the meridional
gradient of heat flux there has been enhanced reflecting the large gradient in temperature anomalies
encountered as we move northwards beyond 30'N.
It is interesting to note that for both oceans the maximum temperature anomalies
in the boundary current
region occur at about 30'N. The peak is centered for the Atlantic in April and in June - July for the Pacific.
The anomalies are about 2 degree C smaller in the Pacific than in the Atlantic and hence, even if the mass
imbalances are larger in the former, the situation in terms of heat flux is reversed, with the largest values
occurring in the latter ocean. Above 40 N, in winter, however, the Pacific experiences a larger generation of
positive correlation than the Atlantic. This difference already appeared in our earlier results but it has been
reduced now by the mass adjustment made, which acted to diminish the southward flow of colder water.
4-7.1) Comparison with the total heat transport within the top 300 meters of the ocean: (Hsiung et al 1989)
figure 17.
Hsiung et al calculated the divergence of the total heat transport within the top 300 meters of the ocean as a
residual between the net surface heat flux and the rate of change of heat storage. They subsequently
integrated this divergence from North to South and obtained the meridional heat transport every 5 degrees.
A quick look at our results shows that our heat term based on an horizontal circulation cell cannot be
-45-
responsible for the seasonal mean northward flux of heat evidenced in figure 17. Other features are likely,
indeed, to produce a significant transport. This is the case for instance with the mid-ocean geostrophic flow,
even if we consider its contribution at depths below the mixed layer. Our present purpose was only to see
how well the seasonal cycle of the total meridional heat transport correlates with the seasonal variation of our
standing gyre term. However, the strength of the deep water flow that we assumed is of great importance in
our calculations. As a matter of fact, to take 20 Sv instead of 7 Sv would lead to values for our northward
Atlantic heat flux up to roughly 3 times larger.
a) Atlantic Ocean:
Qualitatively the agreement between our results and Hsiung and al 1989 is fairly good below 20 N. We do
get a peak around May and a slow decrease afterwards. This peak is however centered around 18"N in our
computation and around 10'N in Hsiung's results. We showed that this peak may be explained by the
contribution of the upwelling region off Africa. After looking at the standard deviations given in the Hsiung
et al study, around 5 or 6 10+14 Watts in the 5'N - 20"N area in May - June , we do not disagree
significantly. The main discrepancy is found in fact at the latitude of the boundary current. Further North,
i.e. beyond 350N our results are not anymore significant, because of large error bars. We still get a decrease
with the time of the year, but our maximum in winter and spring do not appear on Hsiung's estimates.
b) Pacific Ocean:
Our positive maximum centered in winter at 430N corresponds to an area of presumed negative flux in the
Hsiung and al study. However their values, there, are given with a standard deviation of 3 or 4 10+14 Watts.
The relative maxima centered at 30"N or so in May and July do not disagree much from their estimations.
Nevertheless there is still a lot of features in the seasonal cycle of the total meridional net heat transport that
our approach cannot explain.
-46-
4-8) Other possible factors :
The zonal circulation cell mechanism, though being an elegant way of studying the seasonal cycle of the
meridional oceanic heat transport is not the only process that can generate a positive correlation between the
mass transport and temperature fields within the 0 - 300 meter layer. It certainly represents the main
mechanism in our mixed layer, but thepicture is incomplete. If the assumption of an uniform horizontal
velocity is good within the mixed layer it is by no means obvious that the same remains true below.
Observations actually show,in many places, current or net drift with opposite direction at depth. In the same
way, the temperature anomalies at depth could possibly have an opposite sign also. Hall and Bryden (1981)
report, indeed, that below 100 meter the Florida Straits water presents a negative temperature anomaly, while
the flow is still northward. This would generate a negative correlation and hence a southward heat flow.
Hall and Bryden's direct estimates of ocean heat transport at 25-N in the Atlantic Ocean actually show that it
is primarily due to a vertical circulation cell or more likely a diagonal one, since they do not rule out the
importance of the horizontal component.
Although the returning flow between the mixed layer and a depth of 300 meter is relatively weak in terms of
mass transport, its seasonal variations could possibly account for some of the unexplained features mentioned
above .
We should also point out in this section that horizontal heat flux may not only proceed from organized
differential drift but also from horizontal turbulence. Bortkovskii (1961) actually showed that the turbulent
heat flow should not be neglected except for the upper 50 meter layer. Between 100 and 250 meters, its
contribution is actually dominant by more than a factor 10. Although due to large errors arising from his
method based on finite difference schemes with second derivatives, this mechanism is potentially important,
all the more because it appears that the water within that 50 to 300 meter layer is an important element also to
understand the seasonal cycle of the northward heat transport.
-47-
V. CONCLUSION:
This thesis was based on extensive data processing and analysis work. We first derived the mixed-layer
depth from about 2 million vertical oceanographic profiles. This depth was assumed to be the level at which
the temperature was 0.5'C less than at the surface. Not only did this first step provide more realistic data for
further integrated vertical mass transport computations than the Ekman theory would have predicted, but also
it constituted a good opportunity to update previous mixed-layer depth seasonal climatologies by using a
significantly increased number of profiles. However, a lack of data still prevented us from refining the
temporal resolution and describing a quantity that does vary sensibly on a monthly basis as well. Secondly,
we wanted to see how much of the seasonal cycle of the total meridional heat transport within the 0 - 300
meter oceanic layer could be explained by a horizontal zonal overturning within the mixed layer of the
ocean. We did not quite suceed in getting a full agreement but qualitatively some of the seasonal features of
the total heat flux were resolved, for instance the maxima found at around 10 - 15'N in May in the Atlantic
and 25 N in September in the Pacific, although some discrepancies may appear in the latitudes at which we
found those maxima, respectively 18-N and 30 N.
One of our main results is to have shown that only very limited areas actually control the seasonal cycle
through a standing gyre formulation , namely the upwelling regions off Africa and southern California in late
spring and summer, and the North Western corner of the basins in winter especially for the Pacific Ocean.
Thus, collecting data on smaller regions, while still allow to fully describe the phenomena. Indeed, this
description will have the potential to yield more accurate results.
We also found that the contribution of the horizontal zonal overturning cell in driving northwards heat flux
was larger in the Atlantic than in the Pacific, although the differences are rigourously not really significant
owing to the relevant error bars. However, since our Atlantic deep flow may have been underestimated,
-48-
this difference could become significant. In the Atlantic this contribution may, indeed, reach up to 50 %of
the total heat transport found by Hsiung et al. 1979. This stresses the importance of a better ~knowledge of
the thermohaline circulation in heat budget studies.
It is suspected that a good deal of the unexplained features encountered in the seasonal cycle of the
meridional heat transport within the top 300 meters of the ocean could be removed if slightly more
sophisticated vertical profiles of temperature and current were used. This refined description of the upper
ocean was unfortunately not available. However by using the data at our disposal, it would be interesting to
look at a situation where the meridional mass transport would be described by use of the surface current to a
depth corresponding to the mixed layer bottom and by residual from the Sverdrup relation predicted transport
below. Though there is an uncertainty about the reality of a level of no motion which affects the Sverdrup
theory, for our purpose this level would be equivalent to a level below which the anomalies of temperature
from a zonal mean would be small enough to be zero. For the upper layer we would still use our surface
temperature data, and for the lower layer, vertically averaged temperature anomalies, given at each depth by
the MOODS profiles. This possible development will be left at that for now.
-49-
aE
L
coo
-)
0n
L
0
aL
0
-o
C
0,9
X
1850
1880 18'70
10
18190
1900 1910
1920
1930O 1910
19150 1980
1970
1980
gjear-month
Figure 1-a. Basin 1 ATLANTIC reports after duplicate elimination.
L.4L
-C
0
L
0~
0
CL
-9
1801860
1070
1680
1890
1960 19'10
1920
19'30
1940
1950
9'0
170
16
gear-month
Figure 1-b. Basin 3 PACIFIC reports after duplicate elimination.
C
0
L
0)
L.
0
L
_0
C
0
(From Coads Docunesnadon)
OR].
gear-month
Figure 1-c. Global reports after duplicate elimination.
I
-50-
Total number of individual sst measurements
in each grid square
for January, over the 1854 - 1948 period
38W
48W
18W
28W
8W
50N
M ama1
33
a1
1
4l 3m
ii a
A3III
ia m
a
*amsse
a 4
333m11
xM 2a
40N
m11 M M
113171
A amams
53
1mm
in
a
I I 1i
i
m i l a I -, 1s I04i 2na
a #
a 1Ila11 11a0
23 is
B H
*
14 3 3 3 1I 183 II I
2421 3 3 1 3 II lIl 2
331 1 2 4 IS 23 3 2 3 12 U ie
23 3 Imxl 2
Ia3 ,4
a if
I 1 13 3111 123l
a I is I3'I3
3
a
30N
i Is inm
ra o
0
t 1 2
I I 4 1 2 1 1 3
*5 565741 33 13
20N
3 3413N 1 13
3 N If
IN i Is w In it
: a a
I
3 4 33
e s
43145 342N 13
a 4 4 3 15
A4I 4 74.54 1<~
78W
11.14
a
.1..ll
I.I 3 64 I
7,16
68W
58W
4,
38W
48W
18W
28W
8W
2E
8W
2E
ION.
Total number of individual sst measurements
in each grid square
for January, over the 1949 - 1979 period
38W
48W
58W
hassme2 WI 3n 33.
18W
28W
2a1ISO2!331133
m2a2s2
M1 .34 WM3 S22 3 43 03
2 3
52.
x6
32in
13
10 MIaM1M112
* a i m is a s
wilM
ai 11IISNI
ax IM w
nI IN1313
1i1no1 I11111231363
an "4II
al1a14
IS 633333311411 23l3m131333323
* 3 3ni lI
0111431
333333214333
2012823i
3241621 21 IN II 1152333lN1
113 W22311
l
DII W
Ml
40N
IM
13 O5llie M33111MM
a MM0 w a a
IsM1 Is3 IN M3IN1 2
IN IR
IM
3113311115
l
i
R 0251143311313
3I I23 11434 1 4 41 4 I
fi
3 3633 31m
2W310 10 OI 3 3M
MMM iGM3m 3 1 1104M4 3MM1 me III 0413
SIB1IJVVAa MD 3633133 931 IIIIt4 is 1111 HN2sw232
1M" 364133oM6bl
Ma 2 1333 N11331
MMNM NNM1%MD
M Xx le BMM0
ls
30N
2on3
3am Me
1M
3
l
Is a
-I 312 I
e~~D~
a1
I
.
3 1 In In a 1 91 2 13 93 31
o isof~t 1316233M6142313 I
IIIa 1
H331.1
rI3
1M.3 "1093.33im
111
78W
FIGURE 2.
ADM1 13
Ta 3431
aseansa~ax AVA rim Ll
N3
1
A 21 1 141
If
a
ma i 23 2
0 1i1321
1423 23132a15is 3 1
'7
30 il 133415 1 34 46
0 a 140aj 9 N V 4 a
a 31 1
'
.I
I3 MlwA0
3
41
312 3.11
2 263 3.3 0 13 23 2315. I
,5
a115
28W
38W
68W
58W
48W
2ON
ION
-51FIGURE 2. (cont.)
Total number of individual sst measurements
in each grid square
for January, over the 1854 - 1948 period
110E120E
140
130E
1501E
160E
1701
54N
16RW
1
120W
BOW
90W
100W
sow
WX 541
V..
~ 1155411
52N
140W
150W
160W
170W
BefB
311151121of 111
2
SON
iii~~BI~B~~iSlI151
~~~~~~~~~~~
~~~~.~
..... ......
48N *9
R1a11I ftl45i15......
11
.
l~J'..c
4614
46N -uM
. .. .
..
aI
1 14
1
iZ
6
sI 4 I I I i
1 1.
.
44N
x
44N
42N42
3 1,Bil l s a I$ v i
U11411332111 .mnI.:*a
~~.~u1ts
..
11111isM
I
i Ituu6FIImmj
8
a 3U 1 4 11 1511312 1114111
xIII I551111 11 1 11151115 II 1
1183I1421151 151 5 114111121141 15111131V%,l1.)....~
I IN.-tau immix 11150135
111111111111 S111 1 1115131
.531141 if a 3 4J533 5333I I Iii
M
38N
36N36
40N
34N24
1814
2014o
32N12
1 11fl1113111
..
124
01.1 1301111
670
011 o
111
~~
>
It.
2
ION
36
'
22N
00[61S1
4.
149 zg'197 period 13 111
-
1sIWI
3 1 1 1 5 uwuanonaw
160W
21 0
W 11
1 5 W 11
U
10WMw
3
13121
144f
O
90
W
1W
Wt
4
32N3
1W
20
90
0W
ix
I
9
1
II1391i
1. 1__11
.....
26NtN5
ml14OW
-1
4~h
-
BAlIfAl3311
DU
%
~ ~ ~
Sl
]I05N
Is~~~~~~~~
5151
5
1151
411,19,111
101% 120L 130L 140H 150.
1193131Ila
164i
311110
1:a
5
56w 70w
101:
'
555
188W
-j
150
Isailia
a1
3W
'
11,11
11w
Aa1,11
Ilw
I
~
0114
o
2414583
24N114
O
-2245
331111111115151840511515111311116150111114133
I I
'
.
52
O
12N380E
12
24N
a17011 110
21
I) I13S
14N
m ag lgll
11111
3 1 IO 11 11 11121
13149 2 41111111
3
11uh11 Bill 1u1I 11 141h 11 1 1
l
1 9 3
I o f j nis5 I' l l]] li l1 15 i
a 1. 1370
W
1853111191111124
26N
2414
4111
1
120
Nu1310
1 1111313111 5Nall1 III la15
a ~
3I 311~1
oa i
11111
441111 13310
.I..
.jn
13 11
ero
11311111
1z
26N
26N.
40II
for Jauroehstill31113
16N -NI
42N -2
111 fIIa IIIIsI11 1931
1631W11 110 Is
overlth 199497
for Janu7ary,5
.. IV
54a IN oilIa
48N
11111111)13
ION1101 1200130 14013 500 160I1700oi 11118
4N
3oil
1
20.014
N-
11 14#1R1411
is004i155JI 4
32ME
n111
MINORii~ie a I
1f1103 131311355211151111F
41 13JB 41a01141 all 1152X
Rf Oil111311191314131 t
5)R4NiN
511511J711513112111IB11
3 )1
.,....5
22N;0
360-14% -1- M
M 11113551135 11 flags41I Isit 111511111 1ItlO l~
S3IB1S513
24N
.5.....1
1
114J4)21411
a1
il3I)I~
glass25 1u14*Asia
t31sahN1m~m40J~
1 I fJ a31131110111131
1
111
Iis
4
iuuutsius lh551)5)411114151541141 311111l1131443111112111
1 13310111II Io a is111 a aIIf 16111 1125
31
..
24N
1I45,N1!31flh15,1,8683155,E13311153!31I5151111I5fhlIII'l
16411
'141
2014
Z=
$
a
5if1aIisUIUo
1 4is53l3a5is Ifi55Is55of3a1
1-:333c5183 of I1as11
:1
1am11
2014
NIaI
IIIIIIAIAII
1aIIIIII
1414
oalisN@Is am 1I~l4011415511 411 B4J17111411424l1 Na Ilas1J3101H15111511'K
ION~~~I11 1200in 130 141111m Ki0ll 16(111 111111
10W
1
711
530 a
50
140W3Ills
120
10W
1
9W
OW
IO
1
s1
I
-52-
Number of years with observation
FIGURE 3.
Max. Nr. = 95
for January, over the 1854 - 1948 period
4RW
2RW
3RW
18W
8W
5 5
321113
11313
43
5ON
.59112
1 4 461
115 4 41
1 52121i
1h
4l
5
0'
3
ji
40N
i 4all,
5
1 II IS
336514
112
i
13 9
3 1 22 II
I
I II 2
2
H1121
21 3 II
1
30N
20N
2
3 14 1
48W
38W
ION
18W
28W
Number of years with observation
Max. Nr.= 31
for January, over the 1949 - 1979 period
..
48RW
W
38W
'AW
18W
8
28RW
2
I1 I
313 3
13
3
1 3
13
131
1131
1 11
~11
1 11
131
1
31
1
1
1 11
3 3
I
3
1
1 I
t
I
3 33
1 1 1 1 1 1
1 3
133
1121
1
1111111
1 3
ii 1 1 1 1 1
1
3 1 1 1 1g1&3Mj 3 1 3 1 1
I 1 1 1 A~I I I I
10
22 3 3
II3 14I
3
1 13 1 3 1 1 1 1
1 3 1
1 ii 3
3I 2
31
1 I1 N~
Ij
-
InXT
lUL'q
13
3 113
88W
78W
1.
68W
3
111113 3
1 11113
1
111
1
35
23 3
1233
3111
31N3
3 31
.331113
111113 R
31
1
N1 1
13 1,1 3
58W
3
3.1 1
48W
38W
28W
18W
i
Not A%09 A%06 moot MolI MOZI MOtc
I Elf
EIE RE1E R1i
illiI
l lNEEEI
MOtI
Mfs,
M091
Mol
MOI
IlOLI
3051
1091
1H00'
MI
3
IZEDEBE1
I S RI i
ilil III
I"K
lI
RRENER
9 I
IIEURKE
'
i
~~KKKKEE1KKU2SiiU11UEUI'hIU
1 1
ChUIuiU
al19111119'111W
1111
92115
WIlUUIEEIIIEEELKEKUKIE
I[[E III ER R 2114111E 11l11IlllCl a 1K E 9 EEE
11ElhERE
E E
IK
291
S
Q l9KIR1E
I NE ONEs KI ll NOI Nil ER aE
ra ilt SllE R ERE EEgEal
[K El E1
EER a all
9 lRoanE 9 1E19491E
E 11HN
1199 E
11
I
Holi
I I I '[Ra
R EEEERK
I
I
NOT
a EII
E .1 E1111 V
E1E E114
E 11E
1
N91 -N9
Nol
Nel
NO?
-
KEEEIKIIIIEEE
i ER1EIKR11
1EEI(K 1EEE IIKIEII I IEE(I
IEfi11EER EEEEEEEKEEKKKK EEE
E
EEK EK
EREEE EI II IIIEE E EI KEEEIaIEE
5EERIEEKKEEEEEEEEKEKKIKKIIKEEEKKKEEEKKEKRKItEEIEKEEKEEIIE
REIEEERRE1EEERR1REEEEEEIEEEEEEEEEEEEEKEEEEEEEE1EEIIEIIIIIII
aZ a11gEl
E
NAH
Ni?
Net
IEKEEEEEEI
IEEEEEIIEEEEKEEEiEEEEEEEKKKIKIIII
IIEE E I iE E 11 1 IEE 1EKKEEEII
Kl K I
K 11E IKEII E KIIIE1N
us
EEEEEIIIIEKIIEEEEEEEEEIIEEEEEEEEIIKEEEEEEKIIIIIIIEIIE
IEE IEE KEEEEIIKIIlKK
IWEKEE EEI K KEEIE IEK KEKEE
III I
EtI
.
149Z
N'
5111
WE
NVE
IIIE
EEEIKIEK
MOEE
IEEEIIKI111Kl IKKEK(IIL!i1gEEEi
I IKIIKIEEEII
11 I O! Mlt M?
I MItI
11
1
fill
II
E E E
III0.1
'IIEE
E0EEEI
GtI
J-1104
w
E
1EEIIKEE
IKEEIEEK
NZI
IEEIKIIEEEKEIIEIEEEEEI'lltEEEEEE
N
NNPE
M~~~~~~~i~K
[II
II
K
N"Ns
MII MIC MCI MIP IC MIQ IC MI IO 109 1
IC IDC
IIll
EEE
IItE
1ENMQallR
EEREKERNEKIEIf
t9SiSit K I
ilIl
Sill
EE9 fill
K0098I
I
IC)I IC CI
SorI DI
l
085I
KK
~~M~~K~~i11111111 i~tISI l~k i fV t [fi MH11151 K If15 1 tIIlllltuifuIIagi1(Kgj
~Slkft11lf1Ilt1ElL II [I I II itIKI[iIIi it 11111111 1111111111 1111tIl I 11ulK KKll1 l5I t
Joi
I
91 11111[[ liI Mlill KKK111
UtIIlfhll
1111i
11
h10111Ildl~0KE1IK~l
MS
N9;
Nip;
K -d
IA
A0
Al I
-A
AM
# E
NI'il
R DIElx0D
4M
NXE
t
A%0
E E
OI
A E
taE
MOB MCI
.
li 511111111
Elli
ti t
lu~ilU11ilikitl[iil
W
A I I 1 11 1 19 1 1 15 N t I 1
1tii
E E E1 1 1 9
0lNE
E[EERR1'Ifki ?'11t
R I 011
ill
R1ftl
i 0EE11HE
I
EEilKE
R
9
EE B
'4
1111
l I 00l
KR0 11 1 IKiK
It ifE OilE1lf1i
1,1
1R
9R 11
11 11
E
s!Ei lDDl
B
li
l09
il
I SOP Ml00i li
ET MZ EM I E I
lDD
i
t
Hl
E 1I ll
51R
fl II I H I
51kNI Ill
III Ii H !! il l 3
1ol
1 D aEE'E
i R8H11l 1HRH11DIkk
11
11 Dfl9
kgg Oil
I ) 0hRl RE1
1
E
! 50
90100 of0
111
I 111111111111111111111111111 1E11 f i li E
II.i
II
~
KllI Rit llEH EEtoMIDI DEK
Eo II
Dll MIUBIB
i
I
Ot'
I 9 aIolI N Kof 9i 10 a Doic I N I up a1905|9 9 ill n AI
EIEczR
i0
[21 I5510
iERE1
R DE N 010 0 HH~lR R
ERR
J11099119
1 5511901 !!
Lil1
tpim
sn
R
jo joqrnN~
T n I
EE E E E
I
DEM
O
EN1iflR
I
NOT5
IG 1 4.
1 g
E
D11
i [ I C E I'I
no I cog I
DIB
my oI I161Ki l
I
* NE
111I-gt
m
E
N
s g n i g
1 9 l )91
11 011, 91 Ri a! Na!i pitllN NN595 D m of al
uOTwVAJ~sqo
E
"'D
5111115111111
DERE~t50
111hi
iII [il
M I I M MI E
00 91H11E9L1991298111
I I It
it IIIH
MCLI RIIMgaI
H I
Ei
-
A%5 A%9 MOL AMI
-
h
NEE
I
'K I 1
11 a I
l i S11 1 5 I R 1 N lr lHII 51 3 1E50
ffm MO I
ml
WE Mot H 5 % ii
M IT R most li H H 09 115 ant iiE i
M !ED ET M E
ilA%0iDii
Elf
1 VICE
i i I 1 llAi09
51 li
il 11 51 1 lEQua~u
k1 EugRtil
E9
iill
lE11i
ll kI11111
I15tiltk1i 5111111
15kIlill~iu
li111l1
5lil 1 tlS lDl1fllll
D ~
.N)'
ASI EI I 111111111
I A
1t t
ND
I
:ll014
(*iuo:,),c 3wn9ij
It
E
-54-
a
a-e
F
a
-~
-55-
0
K
0
z
0
0
'.4
a'*
0
FIGURE7.
Long term averaged wind
Long term averaged wind
Imh.Obu..315t1.I)
smaan
for Summer, 1949 - 1979
for Winter, 1949 - 1979
Long term averaged wind
uI.. Oh.-SIV(.n
Long term averaged wind
p.
RA
Y
- -- -
-
-~~
-
-
-
-
-~
-..-.-
- --
:7
t
4KWL
3W
8W
79
-
-~
-----
'2~~~~~~
-
-
-
- -
f;
. .. ..---------.--.
-
------
-
-
-
-.--.-
-
-
- .
- - - -~
ii
-
-I
~~ -- ~- ----
-~
--
ION
88W
-71W
68W
M
for Fall, 1949 - 1979
for Spring, 1949 - 1979
cf.
lN5Oh.
o -39.1I)
Smoded
53W
49W
39W
.
ate/I
t
- --- --
--------------
--.
-
.--
U
h.o
-S
FIGURE8.
iim
s
i INS
l
lm
~
~
92K
Long lenn averaged wind
Long enn averaged wind
for Winter, 1949 -1979
for Summer, 1949 -1979
i.
low
I-I
'~
E
Is
5
..
do-avon
- l---
.
\%%I
-~-
*
4
~-
-
5.
..
-
I
=S
u
n
n
n
--
.2./
'
-''"'
eoe------
.................-
1
waIwo
////////.
ww'a4n"
K,
Ii"
-
I"
to
low
I'
--
Long term averaged wind
Ne 1.-s.>n
onbL
31
c4
~....
......
...................
......................
~'/I~N
--
: :Js
lio
13.
e1.
.. .. .. . . .. .. .. .. .. . . .. .. ..
..
--
I=
no
Im
IIW
1
'-
-
,
,
,,,"
-
-
I
-
Long term averaged wind
[or Fall, 1949 - 1979
for Spring, 1949 - 1979
2". ;~j
---
e
-
1
a
rM,-
.
-
MK.. bL
..
WN
ks
a
I'm
-59Number of soundings for
January, 1949
59W
49W
39W
29W
9W
19W
54N
49N
-~
i
44N
39N
34N
129N
24N
19N
14N
9N
vow
7Qw
69W
49W
59W
39W
29W
19W
29W
19W
Number of soundings for
July,1949
59W
Ii
49W
39W
9W
$4N
149N
44N
39N
S34N
29N
24N
19N
14N
70W
FIGURE 9..
6 W
59W
49W
39W
29W
19W
9N
-60-
Number of soundings for
January, 1964
59W
1
4I
49W
39W
29W
-
12
71331
II
lI
7
II
I
I 451443443
4311114
29
4
19W
I
9W
14
33
11I
I
Il
II
111441324
12 1 1 2
1
14
12I
121
1
2
1
1
4
2 117
3
I
1
2
-9
24
2
2
1
3
-
1
1
-.
13
1
4
4
1
11
1
22
1 7
1 t
1l
1
1I
1
11
i1111
23!
2
4 1S12 11 1
4 13 1
1
13
11
nxt
7111
89W
1
1
11
14
79W
1
1
1
131
1
11
69W
1
232
2l2
33
11 1
2212 1 112 1211
-
75
2
3
1 1
1 1
I
1
2
s
1 11
1
473
4
1
1
1
3
1
1
1
IS
1
1
1
I
12
2 'I4
1 2
j14
I111 2
1
59W
49W
39W
29W
19W
29W
19W
Number of soundings for
July,1964
39W
49W
S 41 ,
1233
32
I
I
I
59W
22
SI
its
31
12
II
9W
54N
1
I 131
1
4
49N
21
63
II
3
1 11
M! 1
11 1
4111
1 11
414331
5471411111
1 111 1 13 1 1 li1 t1 II 1
111111II I l II
2 I
13
-
1111
11
43
II
I
N9
IS,
v1Il
112
739N
I
11
3
I
11 3 21
.11111
11,1
113lil
31114
11
41 111
11
41 42
1
44N
i
4
i
34N
2
111
fill
29N
521
I3
2t
I
l
24N
4
2
[9N
32
111
i
19N
I
2
33
2.
14N
r
i
89W
FIGURE 9. (cont.)
79W
69W
59W
49W
39W
29W
19W
99
-61Number of soundings for
January, 1973
39W
49W
29W
1Qw
5 12
132l1
3 4 1
114
9W
OW
19W
29W
2
931
4
113
54N
2
II
1
I?
II
1
39N
III
IN,
2
II
I ll5
12
2
21
I II
1
i
14
34NI11
2
21
21
S 41141
11 2V I
2I
2Nl1
AM
I 14
4112
12
1
11
12
3
24 1 2
3 4
1 3
4
52
1
1
1
I-
4 3 11
1
2I
2
39
I
lil
2
1
21
19N
I 1
1 12
1
1
sI I
14N
I 12
21
712112
79W for
89W
69W
59W
49W
39W
29W
19W
Number of soundings for
July, 1973
49W
S11
29W
1
11
19W
1
9W
il 1111 11
1 1417
17
141
1
31
41
Ils
3 4N
14 63
1
1
1
3i1.1
1
11 3171
21
11l
1
122 ]1
11 1 II 41
42
I 2 11
2M7 41 141 IlS
1
5114
li1t
11
1 1 3
I
213
2 1 1111
2
122414215
1 4
124116
21
11 1
141
I 11
1
4 421 49
H41
4
54
Ill15
1 il 411
111611
4431
1
111
3
1
111
1
41
1
71111
11421
712
1115
71112
ilia,
12 1 3Ill
til I 1 7
1
Ill
HI
14 2
1 311
2 1
fill116 13241
1
1k
M,
I~aL.I
1
2
1
11
111
1111
7
li
11
l 1
1
3
1
24
1
334
15 1
1
It I.l
1
li
29N
ifII
1211
1l 1
1
2
1
2
11
1
1 24N
I
1
7711-
1
21
1
21
1
1
2 1
1612
14
II 3..
1q
79W
FIGURE 9. (cont.)
39N
9
I II
I4
44N
4
13141
1
1
2
1 ItsI
121 2 il
i I
11 112 1
1
32
11
4$
1 1
1
49N
714 11
q41 s L4 11 7 18 31
32
1
1 11
4
14 ! 111
$111
111
11
1
12
21
1111111211
1
1 1 I ts
1
1 1 4 441
1339
69W
59W
49W
39W
29W
19W
9W
z
'.0
z
z zz
z
z
4N0
z z
z
-62-
z
z
z
z
04
0
0
Number of soundings for
FIGURE10.fcent.i
July,1964
S119E
1 59E
139E
179E
161iW
1 51
~'
134241i III
1i1W
I I
54N
21(1 I
3 1132
3'il
II
II
I
49N
44N
121W
A1W
35055
)lil.4t
I I g
I
II$
54N
E161
49N
........
.
~~...............
~ ~ f1
21.
I iI
39N
'V.~i 223
s~~~*
I
NME:
gl
.~
34N
29N
2
iti
1
1 13
5i
I
I ill
111332 II
......M
I f In
11111
21li
3
3
II I
I11
II
Ili
I
....
)
34N
111
IlI
I
|I
132!
33
II
II
Jill
I
I Il
33 i l 1 221
i ll
I i 23
I II1
13223 I
41 !!4)
I
121132
el21
1lilIII
J2344211
222)4l121
12
3ilt
4
S S1 5
24N
I42 48 1 1 2
1 13 1
1223 lil3 i 4l I3NI Mi
. 1132211
SI
,2124
4
I 1I
119E
139E
159E
179E
161W
29N
1l
4 I) 14
133341
I 11151
1 4 S
J31111
19N
lit
I
39N
II
ii
1132 I
2.
14N
I
I
2
S112 !!1
I111111 21
2!
I 111314 i I
III I
1
I Jil
11
II 22
24N
41
I
~
141W
19N
14N
Number of soundings for
FIGURE 10. ictnt.j
July, 1978
119E
139E
159E
179E
161W
141W
54N
54N
*
*.
.
....
"*
fA
'
I
i
l
21
I
l
49N
1
211
l
Miss'
44N
II
111111111
"
111111011
'-11
MON.
I IH l
1111
11 1
Il
I~
11421
I
II
I
I
I
I
I .4II5212 I II IlI I Il
I I 1
IIl
1 21 1111122 43113233
I
I I
1
t1IIM0Ill 21
X.. : :: x : . .
29N
II I 2
.~~
ll 2
I
1 W4ii 1l
ll ) 1I1
1Nill
i Ill13
l) 11 1 11 1 1
II
112
Il
19N
42
44N
2 I 13411
I
39N
tIll 11311
IM
I
III 1
22)|1
t111ill
I 1 1 1
1 1 1
11
1
1
9231
141 33
4 i1
1 11
I1 U31
1
13354131
211
1 5 ~~~'I
1
1
31
1
211i
Jil
i 4 3il
. . .111
. . ...
.1II
.).. . . . . . . . . II
IS431
1
II
I 1~
2 43 4 I
IiI
2435.....
I 12
11
1
344
1 ItU 1flI I
I I
1
441
I
24N
I
I ill
111 Ill
Is?I-nl
I
34N
49N
11
39N
M
11
all,
Ji~Il
i l
I 121f i 2l
1 41
1 31
11
34N
2
|
29N
II
24N
2IIII
19N
fillli
311133 1
I 14
11 1111 Il I
14N
1
I2
14N
11 211I
119E
139E
4
159E
IllI
:
179E
161W
141W
9N
-65-
A sampling-based rejection criterion
Temperature (degree C)
SST - 0.5C
25
MLD =
MLD=
3.52 m
6.10m
Profile given
------
-125
FIGURE
11
MLD = 24.62 m
E-.
I ZZZZl
o
1-
,'
7I
000
0.
'00
4
4"
00
too
oIII
01In0MMMMONMNMwMNNMW4"-.
-00
00 00
01*
11
0
noMMNN
0.v
'"",ZLz~zzzzzzzzzzzz
0
0on
0 44
FIGURE12 (COnt.
-67-
)
SUMMERTIME MIXED LAYER DEPTH
80W
62W
44W
FALLTIME MIXED LAYER DEPTH
sow
62W
44W
(1949-1979)
26W
(1949-1979)
28W
8W
-68FIGURE13
WINTERTIUE MIXED LAYER DEPTH
131E
182E
167W
SPRINGTIME MIXED LAYER DEPTH
131E
162E
167W
(1949-1979)
136W
105W
(1949-1979)
136W
105W
-69FIGURE13. <cont.
SUMMERTIME MDXD LAYER DEPTH
131E
162E
167W
FALLTIME MDED LAYER DEPTH
162E
167W
(1949-1979)
138W
105W
(1949-1979)
136W
105W
-70FIGURE 14.
AMPLrTUDE OF THE SEASONAL CYCLE MLD (1949-1979)
SoW
62W
26W
44W
ATLANTIC
OCEAN
136W
PACIFIC
oW
OCEAN
106W
-71-
ATLANTIC OCEAN -
Sam
NX
1
2
3
STANDING GYRE CONTRIBUTION
4
5
6
V*T*
7
(1949-1979)
8
10
9
I
I
54N
i
11
I
12
I
~
54N
S0
IN 10+13 WATTS
52N
52N
50N
5ON
,-,..
-0
48N
48N
A
*
46N
I
--
--
44N
44N
100
.
42N
42N
40N
38N
46N
40N
'-- --- -- -
38N
-
3ON
36N
34N
34N
32N
32N
30N
V..
30N
-C
28N
28N
26N
b.10
24N
p
28N
0
.
24N
22N
22N
20N
20N
18N
16N
1 .10
14N
14N
12N
tom
18N
.10
ION
*
1
I
,
12M
. to
2
3
4
ATLANTIC OCEAN 1
2
3
5
a
7
a
STANDING GYRE CONTRIBUTION
4
6
a
7
10
9
STD(V*T*)
6
11
12
11
12
LON
(1949-1979)
9
10
FON
.r mA,
54N
IN 10+1;
52N
50N
48N
48N
46N
46N
44N
44N
42NI
42N
40N
40N
38N
38N
30N
SON
34N
34N
32N
32N
3ON
3ON
2N
28N
2ON
26N
24N
24N
22N
22N
20N
2ON
5.50
ION
ION
ION
L-3
14N
3.50
s1.50
ION
-
14N
12N
1o
12K
'
1
FIGURE 15.
'
2
'
a
3
4
5
a
7
8
9
10
11
12
ION
-72-
PACIFIC OCEAN 3
2
1
6
5
4
V*T* (1949-1979)
9
8
7
10
12
11
-
-
-
DUN
STANDING GYRE CONTRIBUTION
54N
54N -
52N
IN 10+13 WATTS
52N ,
5ON
5ON --
48N
48N
46N -
-
-46N
-
c9
*
44N
42N --
44N
42N
.0
-
40N
*
40N
38N 36N -0
34N
.50
O0
.5s
38N
-O
.501N*.4.
I
34N
1
I-00
6
,
-
32N
12N
30N -
30N
S
.0028
28o,-7o
20N -O
22N --.
5
20N -.
0
.00
-Z O2N
S
.50
22N
00
II
ION -
7.00
su.Z
So
2ON
c2N
00
22N
so20 so
ION
SoONa
50
ION
0s
.
14N -~
1EN inm
012
PACIFIC OCEAN 1
2
ESN
3
54N IN 10+13 WATk
5
4
3
2
1
52N .
I04
lE
STANDING GYRE CONTRIBUTION
4
5
00
00
6
9
8
7
a
10
11
12
10
11
12
STD (1949-1979)
7
a
9
54N
9
52N
5ON
-
48N
5ON
1.00
48N
40N .-
40N
44N -
44N
42N -
-42N
40N -
40N
38N -
38N
36N -
36N
34N -
34N
32N -
32N
30N -
30N
28N -
28N
26N -
26N
24
24N
.00
22N
2ON
ISN
18N -
18N
-0
ION
100-
14N *
-----------
-00
.
1.00
.00
12N
ION
0
00
1.e
14N
12N
ION1
FIGURE 16.
2
3
4
I
I2
7
I
8
10
11
12
ION
-73S5N
IW)
i
I
45N --
I
i
I
I
I
I
I
I
55N
7(W
45N
0
7.(W)
35N8
(
(W)
25N -
35N
C3
e
g
25N
11.(BH)
10. (W)
14. (R)
15N
15N
5N S
i
55/ o
11\i
8'O' o
*5
es
15S
DEC
JAN
FEB
SN
MAR
e 6. (B)
APR
MAY
s
6.1 (R)o
JUN
JUL RUG
SEP
OCT
NOV
DEC
JAN
Zonal averages of meridional transport, 0-300m, 10"W, for the Atlantic Ocean. Positive values
denote northward transport. Estimates from direct observations are shown with letters representing authors:
B = Bryan (1962); BH = Bryden and Hall (1980), Hall and Bryden (1982); R = Roemmich (1983); W =
Wunsch (1980).
55N
SEN
-C,
45N
45N
P
35N -
35N
25N
25N
15N
15N
5N
SN
5.",
5S
-0
SS5
15S
DEC
JAN
FEB
MAR
APR
15S
MAY
JUN
JUL
for the Pacific Ocean.
FIGURE 17.
From HSIUNG et al 1988
AUG
SEP
OCT
NOV
DEC
JAN
-74-
ATCOEN- STAM 88 TR COFMU~88188
-
JCAtT(1949-t$?*)
Oy" CONTNM Om- rrNDCHO,
OCEAN
ATLAIMC
saw
GYM8
8801W1808
ATLW80OCEAN-BTAND040
-
ILAR
(1949-9079)
a"
ATLwm88 OCEAN .
ATI~xm8 08888 -
FIGUREIS.
44W
wr1momme
6188
Amy ct946-60101
*am
-
cow18*8U81
ovMUW8 0160x coww8U1106
-
s
rzx
maOYD3U css08-tvre)
(1040-tW78)
-75-
PL.n
CA
-V~MC
V
OV CO
-
(24-1
l
- 11.MIUSI
@138
COSMUJTION
PAMCM-DT3NlDVH
0138CO303332fl0- 33832(190-11110)
30808- STANDING
3-02120
PACMCOCxKANk3
FIGURE 19.
tall
3
C2232OW GRZC3T3232fO
t3."
-
mew8
300m331308 -
Why(1033-1*7l)
382,.
131
cr838233138003132.umm20
le2ow33w
-
133-213
N81W.023(A242-38)
l
-76-
ATLANTIC OCEAN -
~
1.0
1
I
2
I
3
FIGURE 21.
10
9
8
7
11
12
I
-...
*
4
ATLANTIC OCEAN -
(1949-1979)
o.
0
- - 0 .I
6
5
4
3
2
MASS BALANCE RESIDUAL
A
MASS BALANCE RESIDUAL
A
a
7
STD (1949-1979)
8
9
10
11
12
-77-
PACIFIC OCEAN 1
2
3
MASS BALANCE RESIDUAL
4
5
a
SON
7
(1949-1979)
8
9
10
11
12
I3
6N
54N 2N
- 54N
IN
52No
/SEC
10+8
-
-20-SE
52N
-
5ON
48N -
.---
40N
44N
.---
-
5ON
0-0
48N
50.0
---
--
--
.- 46N
-N
42N -
'
-42N
40N
40N
38N -
n38N
36N
36N
-,
--
-
-.
34N -
\
0------~
%%
{
S
34N
c
32N *
24N
28N -
-30-2N
-
12N
4
G.~~'/
0
-
--
-. a
...-.
-
-
12N
-- 4a00.0
--
21
-
10N
N
52NO.
0.0
4ON--
10+ "
/
-EC
5.00-ON
-
1
61N 12N 512
ION
N
1
---
33
'10'0r
-r
--
--
-r-
-
3
- -
-
i
--
9
0.
40.078
--
--
4
T
~
0
06
9
8
11
1
10
11
12
10
11
12
48N
4N
12N
11
ON
50N -0.0
PACIIC OCEAN
1
a
S
4
BALANCE RESUAL
5
6
7
STD (1949-1979)
9
a
-MASS
5ON
30N
52N
-
/SEC
- 2N
2ON -
SON
0.0
48N -
48N
8N -
441N
42N
54N -
SON
32N
--54N
42N
'
-
4.'--4.O50-- .2N
.:
32N
3011
92~
U.
24N
-
24
2 0N1
-GURE
-S
________
10N
ION
4N
30.0
0-.
5.
1ION
0N
4.
ION
ION41
14N1-
1211
12N
0
300
1
FIGURE 22.
2
3
4
6
7
a
g
10
11
1ON
12
-78FIGURE 23.
MAIN' PATH- QF THE KUROSHIO AND GULF-rSTREAM SYSTEMS
50
I
--
-..
-E
-4-
,
Grand Banks.
NORTH
ATLANTIC
CURRENT
40*
GULF STREAM
-
-
30
N*
. 40 0
From Hideo Kawai 1972
-
LI
-79-
FIGURE 24.
BOUNDARY CURRENT AND OTHER ZONES USED FOR MASS BALANCE CORRECTION
SOW
62W
44W
25W
BOUNDARY CURRENT AND OTHER ZONES USED FOR MASS BALANCE CORRECTION
131
162
167W
136W
106W
-80-
ATLANTIC OCEAN -
ADJUSTED
4
ATLANTIC OCEAN 2
FIGURE 2
3
STANDING GYRE CONTRIBUTION VT* (1949-1979)
5
a
7
9
ADJUSTED STANDING GYRE CONTRIBUTION
4
5
6
7
8
10
11
12
11
12
STD (1949-1979)
9
10
-81PACIFIC OCEAN 2
3
PACIFIC OCEAN 1
2
ADJUSTED STANDING GYRE CONTRIBUTION
3
5
6
7
8
ADJUSTED STANDING GYRE CONTRIBUTION
4
5
6
7
8
V*T* (1949-1979)
9
10
STD (1949-1979)
9
10
11
12
E8N
5 4N
54N
52N
52N
50N
SON
4 a'N
48N
46N
48N
44VN
4 4N
4210
42N
46N
4 0N
40N
38r
38N
36N
38N
34N
34iN
3 2N
32N
30N
30N
28N
28N
26N
26N
24N
24N
22N
22N
2oN
20N
18SN
ION
1.2a
14N
12hN
taxN
FIGURE 26
12N
1
2
3
4
5
6
7
8
9
10
11
12
-82-
ATLANTIC OCEAN -
%V
1
2
11
1
3
T* IN THE BOUNDARY CURRENT AREA (1949-1979)
4
5
6
7
9
8
1
10
1
11
12
- .AnmJ
38N
36N
1.353
1.35
34N -
34N
32N -
.05~
32N
2 7-2707
2.7
3ON
v-s
30N
28N
28N
70
26N -
.700
26N
24N
24N
1. 353
22N-
22N
2ON
-
1.35
1
2
1
2
1.35
3
4
PACIC OCEAN -
5
IN DEGREE C
20N
135S
7
8
9
10
11
12
10
11
12
18N,
T* IN THE BOUNDARY CURRENT AREA (1949-1979)
5
4
3
-1. 35
9
5
7
6
.500
36N
.51.50
1.*00
3aNi
2.0
00
34N
.00
2
3.0
32N
30N
28N
TL
I
24N
2N
'3
.
24N
4SI
.50
22N -
6
p~o
22N
.00
1=
20N
20N[
-
__
I
iI
I
1SN
1SN
ION
1
.m
FIGURE 27
1
2
3
4
5
6
7
8
9
10
11
12
14N'
-83-
References :
Bauer R. (1982)
Functional description of the Master Oceanographic Observation Data Set (MOODS).
Compass Systems, Inc., San Diego, CA.
Bortkovskii R. S. (1968)
Quantitative relations between water, heat and salt transport in the ocean.
Problems in Dynamical Oceanography
A. I. Fel'zenbaum Editor, Akademiya Nauk SSSR.
Budyko M. I. (1963)
Atlas of the heat balance of the Earth.
Moscow, Globnaia Geofiz. Observ., 69 pp
J. Dana Thompson and W. J. Schmitz, jr (1989)
A Limited-Area Model of the Gulf-Stream : Design, Initial Experiments, and Model-Data
Intercomparison.
J. Phys. Oceanog., 19 (6), 791 - 814.
De Szoeke, Levine M. D. (1981)
The advective flux of heat by mean geostrophic motions in the southern ocean.
Deep Sea Research 28A 10, 1057-1085.
Fuglister F. C. (1951)
Annual variations in current speeds in the Gulf Stream system.
Journal of Marine Research 10, 119 - 127.
Halpern David (1980)
Variability of near-surface currents in the Atlantic North Equatorial countercurrent during GATE.
J. Phys. Oceanog., 10, 1213 - 1220.
Hall M. M. and Bryden H. L. (1982)
Direct estimates and mechanisms of ocean heat transport.
Deep Sea Research, Vol. 29, No 3A, 339-359.
-84-
Hideo Kawai (1972)
Kuroshio : its physical aspects. Henry Stommel and Kozo Yoshida Editors. University of Tokyo
press.
235
-
352.
Rui Xin Huang (1984)
The thermocline and current structure in subtropical/subpolar basins.
Ph. D. thesis MIT.
Hsiung J. C. (1983)
Large scale sea-air energy fluxes and global sea surface temperature fluctuations.
Ph. D. thesis MIT.
Hsiung J. C. (1986)
Mean surface energy fluxes over the global ocean.
J. Geophys. Res., 91, 10,585 - 10,606.
Hsiung J,Newell R. E., Houghtby T. (1989)
The annual cycle of oceanic heat storage and oceanic meridional heat transport.
Quart. J. Roy. Meteor. Soc 115, 1-28.
Kraus E. B. (1987)
Les couches melangees : energetique, modelisation, et role dans le ssysteme ocean-atmosphere.
La Meteorologie, Vol 17.
Levitus S. (1987)
Meridional Ekman Fluxes for the World Ocean and Individual Ocean Basins.
J. Phys. Oceanog., 17, 1484 - 1492
Niiler P. P. (1977)
One dimensional models of the seasonal thermocline. The Sea,
E. D. Goldberg, I. N. Mc Cave, J. J. O'Brien and J.H. Steele Eds.
Wiley, 97 - 115.
-85-
Oort A. H. and T. H. Vonder Haar (1976)
On the observed annual cycle in the ocean-atmosphere heat balance over the Northern
Hemisphere.
J. Phys. Oceanog., 6, 781-800
Riser S. C., Freeland H. and Rossby H. T. (1978)
Mesoscale motions near the deep western boundary current of the North Atlantic circulation.
Deep Sea Research 25, 1179 - 1191.
Schopf P. S. and Cane M. A.(1983)
On Equatorial Dynamics Mixed-Layer Physics and SST.
J. Phys. Oceanog., 13, 917 - 935.
Slutz R. J. and seven others (1985)
COADS, Comprehensive Ocean - Atmosphere Data Set
Release 1, CIRES-ERL-NCAR-NCDC, Boulder, CO, 39 pp.
Worthington L. V. (1976)
On the North Atlantic Circulation.
The John Hopkins Oceanographic Studies 6.
John Hopkins University Press, 110 pp.
-86-
Annexe I: Acronyms
MBT :Mechanical Bathythermograph
XBT :Expendable Bathythermograph
AXBT : Aircraft dropped XBT
STD : Salinity temperature density probe
-87-
Annexe II :
1) Total number of soundings at each grid square integrated into the computation of the
seasonal climatology of the mixed-layer depth. Results are shown for each season. Atlantic
Ocean.
2) Number of years with observation at each grid square for the mixed-layer depth seasonal
climatology. Atlantic Ocean.
3) Standard deviation for the mixed-layer depth seasonal climatology. Atlantic Ocean.
4) Same as 1,2,3 but for the Pacific Ocean.
Total number of soundings at each grid square
integrated into the MLD computation
for Fall, 1949 - 1979
88W
78W
48W
58
68W
58W
12I1
.1
.
..
50N
#
io
..............
S3
125 4 5
...
? - e.5 M2 131INI
N4
16 68
241II
38W
28W
15
4
14
11 3
52 4614 3 2
5 1 S 174
22 3 4 5 I726 11 19 12
II U I i II 4
I i
I9 2141M 4I11
1mm41M 31 i 1 n 92Is I
i3
S33
1 2114413
o 2N2 10 173 M3
152Is IB5 12
1 I 1 15 I 17 III 119
113a 5
4951 6
4 33 13
3!3
" 1" in 91 i a 3 1 e
1 3 2
' 1
30N
in a
gi
1 19 5 1
11 4 12
351 US314
5 il5
1 1 3I I 1
40
5I
6 3
52 24 4 1
l 14 3 5
4
41 7 4
11 2 2 5 3 33 25 4 4
63
6 1611513
336
23
1 94M1
1 113919521
78W
68W
58W
451
10 2224 1 1914 16413
16 1
94j7 M
1 IN 1
ggg
50N
1
1 2
3421A
I 2
2411
U
6 5I
40N
49i 11 12
16 I5 63
21 14 14 9
%14
a3 Ill 1151 152 12is 22109
13
34 215 3 6 5 6316 1211~dEs~smN
Is 1 5 4
e
14123164111192
m14
B
1
10Eggggggg0gggg
30N
14 17 3 4 A
U 29 24 6 14 13 14
II 4
21 23 7 4
12
3 17 19 17
23
A
U
A....
.......
20N
3 24 2217 31341
1 32 3
24 7
10...14 A 14,
16, Rgl
47
141536,6
88W
11
414
6 1525 11 IS
31 II
19
ION
M
1 21 14 4 4 1 in 1
1 0 1 to I 10 ju
24
5 9920 1 3 R413
. 1
22 19 1655 1 7111913 46N51 79
.
415465 4412 5199
3 2 214431 313 313
41111225211 V 56 511
20N
I
.
I
1 a 2 1N33 341
3 24 292123 53
4 44445
35
515 343 1 11 111 59 2 5 M
-43
239
2131491 1M1a2 24 1621 3
262 If310M14901
isM2w14512 a
1 3 U 21 16 545
1644
3611 11
8w
14 A11
1 2219 22 14
51 24 17 111 i 1
.. 16/05 Ml5
40N
1612 213
18W
323
21 I1 14 II I 12 14
II
11 5 11 7 14129i
2, 4 137 5
48W
38W
28W
13219 1 F,
.
......................
13 1
18W
8W -
2E
ION
M
I
-M
Total number of soundings at each grid square
integrated into the MLD computation
for Summer, 1949 - 1979
88W
78W
68W
.........
SEl iN
i
....... .......
40N
22 p25
4 12953
21A
4l6144613 39 1 A 4 M
149213
137
1 13 I 31 11141I 13
el 7v
a17I
211
1a
1
is i
2 12439 5 92512
31 a 1 9 4 3 11 3
2jt
14 21
2
2 1114 3124 3
41 5 10 3121 4 414
&.
it 3
6193 5 19
12 1
26 15 411I,
415
Is 19
2223
111113 161
35 13
9l64
14#1
13 10461255
l 14 1
III 141515324 16153
13# 11
M4
1Ia41
511
as 24 IV363 1231 31
16 lI4
171
26416
14
II a l 2431 a
\ 4 19 1 24. 219 11 211
23
1 45'
1'4
3 21 23
521
63644
29 16
S17 Ii I
lI 21
1341 1. 52 EME
II 17
95 11 le
'Uli
II 29 I
ION
88W
I
i1
56 2 3
78W
.1
68W
58W
4 1
11 3 6 47
3 7365 113ll 1 4 43 112334339252
I I 1 HI xS M1,2 141 IH 33 3 1
295534113
529 2943 I
14
48W
14
.a
40N
2124222
%MNSE
30N
1 1
77 liiiim
11
9 4 s a 0 a a i 1is 74 8 a RI
51529
63 140
123 II115
39444 14134 iP #W
i136
1& 14 11 41 641
3
911
5456
1 13
6 MI 1 3 29 21 1 5 I if 4 1 4
5
452393 1 112279REMSkfl
16121
141161
6 M4443 1 12IQ4
84
0 I 1 % v a a M Ill 2
11 161
35
3
50N
341N139 Ill 16 I 6 4
4 3396 1610 163111
4 2 14 11146 0 14 6712 IN 16 123 3463 21661 24 12 54 1 3 1 2329 12a Is 14 14 1416 1313 2 3 6 1 7 i
20N
16
159 17Is
lai)
....5.11.
~I IMis 5
iy~
1 I35
174
8W
191i
10 4 2 3 4 31%61
5ON
30N
S 5113
18W
28W
38W
38W
48W
58W
. .......
~ **
48W
58W
9 1 2 I a 2 it 5 4
13 1 4 3
20N
1 14 9
142I 1214
1 1 Q 1A1 111 3 12 a
13
19 1
19
1 42
38W
28W
18W
8W
10N
.
.
Total number of soundings at each grid square
integrated into the MLD computation
for Spring, 1949 - 1979
78W
88W
58W
68W
48W
I
1
I
1
1
46.4
74.
.~...........
6
N
24
24I
E 51W
24
2
1445 345
-8-
8W
18W
1 5
1 3 aI
5113941
21 II 44
Elii
50N
28W
38W
48
I 41
421
I I i121
2 5
2 42
N I
95249 23)7
13)
6992 24 I 11
4 S
2227 lN N 14
214
.
.
3 6 S 4
67
40N
gi
1
1
50N
.....
........
211
S17
N 41i 5144 R6N9
'"""~4l1I1617 14
na2
1314192211
It$i9$24610 6T2 123 536 0
7149
43
49
4 4145 31 5 61
111594
11 31 16 344131
4
b
161 2 6I
125 3
12 1 7SIl I
1643 11
1
30N
410 l 7
1II
551655161 1 i2 6 21 93 6 1 3 914 13 1433
51 411% Il
2
22 39 2124
1513
2212 2 12 3 4 3
5H
.11 41
2191
2 6 21 242 2 5
3111672539 17 41212
m
j IS 243
M1
U523 14
11 1i 3I a 1 2 24 21 1 2
16
~ 15 16 0 26 113 G 3
k21M
593 1 . 29 439 1 M31156
112311311l
3611291 II
55I 1
76"75 1
20N
Is9
I
1n 19 I35I9 E15is1 4193
4 2465 4 1is
552
4
SI 173
0
32 3 7 I 2
ION
88W
31
46 21 5,3
78W
7
58W
1394
19 9
14 I II
4
48W
-
z.
.... -.. .w
40N
BmseEl
N 21 2114 24
x 31 4 441! 1 l A6 .: II i
II
43
4 4
3211
3
13)4
1
II 1 3 53 1
11
30N
31
7 4
2 41
.........................
. ...
...
...
...
91I7
..
. .
. .
.
1 27
7 7
21
I
1,
9
20N
2
5 lI 2
6, 24 153 3, 3 . V., 11 1 1 7 , 7
~ ~ . i IS
19 24 2511
68W
18
-~:~ 39
i 3 g4i1 6
I7 4
4 451
16 lI 21 9
2
1 2n IN15 11 31 4l I 11443 5117
1
3
--
16~~145
6 45
.441911 2 Il 1121311114
4
IN e 2
19 lie 22 30a
4 1511
194
21 Il 711 m 14 21 1 147 0 5
.:a
2
,3)2
38W
13
1
9, 2
28W
.44
.....
.
1
ION
18W
8W
2E
Total number of soundings at each grid square
integrated into the MLD computation
for Winter, 1949 - 1979
38W
48W
58W
6W
78W
8W
18W
28W
48W
I I
'31"
maI
50N
8W
11i 1
4
.......
6
1
24
3
211
1
2'
'
4
50N
Ii
$ 21 3 21 1
-
5I
... ..... ..
40N
S 1
..
3 33
2
I
11
...
533U92 53 16413
16
i11113 V 5W 2412 3
IM
2 6
44#21
3 11
3 2 2 11i 3 12
3 4 1119 3
6 3 3 211 11
1 2 3
II
2
31
15 6
211
A 5 19 29 19IS II
112153ii421 5"N 1 3321 2111 13 mS4 4 45 43
15 19
5 ill32 24 21 IM I Q5 4 1 3 4 29 29171
)mi
6 121 14Il
% 3 54 9510 1
14 21 3 I 64 46# 1E
1
.
.
15
-3 j!i.ihE, , i . I i
il!
alRN!E!
3 262315 13
30N
33 19 2411 9
1 250 5N 411
1, 2111131354114 li
16in
1I410
031
4 17 4 0 4 11593 i2g1
20N
5 1a14
l
141 24 11 13 11
l
I 423 I
a
17
3 212 Ii 13
II21 26 122 23 1
3 i s 24
1565 23)2 4 5
425244133110315311214352
# 6 2 14 D in
M14i
6 1 1, 1 a
11 1212I 2 2 2 14 15
14 16
a 34 N.,121
112 2 4 1 11i
3 16 4
1 12 21 16 1211
IA
I26262 131415 1 24I3 33625 116
13 53 11 1 .23
JA,U 1,3 4 5
34 3 3 3,l~.II U 3 4 4gij 4' 45
,
,
4
4
9323149I113
88W
78W
5i
IN41
68W
58W
48W
38W
30N
,
.....
5.....X
5 41
1 24 I 1, 11
116 19 5 I 1117
1 . 6f4
3
I I
16 2115 3 12
23112 3 a I 2 1411
i R
ION
21
1 1
40N
59i
S3
1 14
14 9
417
2. 1
,a
28W
19
3
20N
Mi
E
41142
I
5:
..
....
18W
8W
2E
10N
MLD
Number of years with observation
at each grid square
Max. Nr. =31
for Fall, 1949 - 1979
88W
78W
28W
38W
58W
68W
18W
22 21
48
1520 25
....
..
21 212518
I5 11 16
2122
2225 21 27dlI
$
14 11 16
2324
25212726IS
20
19 12 Is
21 25
24 262519 16
292926 26
21 13 15
t 23
25 24
30 30
I
2qN0
40N
1 7
~Z N0
1
% 25
29
....
20
2124
20 21 21
7 921 24
50N
' I512ire1
2E
8W
-
QI 7
192122 262
40N
22 23 2421
30 Q 3 3Q
50N
26 252120 29 7 26
23252129
2.29 29
3 3429 25
30N
22 24 24 26
26 21 26 26 26 24 26 24 23 21 22 22 21 22 22
18 21 19 1
12 13 1 18 19 17 15
iZ 27 26 27 26 26 23 22 19 2018IS
18 I716 1919
1519 19 2 021 21
101315 1518 20
10 1213
N
1
8 20
C
15 17 18 19
ION
13 14 i 13
88W
16 17 1
16 14 17 17 15 15 13
17 17 A-i
24X 26 25222423211617 18 IS 20 Is
12 12 12 11 10 1012 II
I5 Ii
1
13 13 11 1 12
-26242224 24 22 17 16 15 15 17 16
3 1613162329 5 1 22 2 25 211916161516 14
13 14 16 22 2320172224 2522..21 17 16 15 13 14
11161
10 to
2 r 282724262322 19 13 1518 17 20 21
2 I
23 20201 1411 13
13182120192 212223
II II
68W
58W
I 1515 14IS
II
101 1011 11 17 1816
12 1211 Il 141518
10 10 10 IE
, 12 22 23 19 17 17 1921 22 2120 20 18 13M 10
21 193
78W
30N
17 19IS.15
2 24 28 27 25 2 2522 16 18 15 17 15 17 17 19
13 19 19 19 20 22 26 2
20N
2528 24 24 23 23 21 23 24
48W
38W
28W
10
161
20N
1918
22 I
20 4
ISO
Is
I8
IS
18W
8W
2E
ION
MLD
Number of years with observation
at each grid square
Max. Nr.
31
=
for Summer, 1949 - 1979
88W
78W
68W
58W
120 26 20 25
1$~ t ~223
14 I
50N
.112
t6 26
il
1 6 1
26 a212
50N
272 21
~262029iiii-j-21 27
292929230
~~2
30 29 30 29 29 29
30 30 303029 29
30 30 30 30 30 30 30 30 30 30
40N
40N
30 29 29 29 29 30 29 27 21 27
29 29 29 28 20 20 26 26 25 25
20 21 272820 20 26 26 26 26
25 2 2 20 21 27 28 21 21 24 25 24 21
30N
8 19
2 28
22 24 24
30N
25 26 25 25 25 25 3 20 17 15 14
28 2. 2 2%24 25 23 25 23 19 I8 17 12 13
12 20 19 22 24 25 21
12 16 16 21 23 24 21 2
28 21 21 25 24 24 24 23
1818 14 14 15
Il 14 17 IS 21 21 24
13 131
27621 21 26 24 24 24 22 18 17 15 16 15
26 25 25 25 23 23 16 17 14 14
9 20 22 18I19
I
19 19 15B
IQ16 17 2D19 22 24 24 22 19 17 17 15
18 20 20 17
11 16 17 15 17 19 21 19 11 16 15 15 15
13.12J*
id J).. 7 . . f 12 9 11
58W
68W
78W
f
15 I7 11816,10
20N
23 B is 15 14 13
161416 2 29
11 1214 19 21 22 19 2 23 23 22.20 17 16 13
48W
28W
18W
2
8W
2E
**
10N
MLD
Number of years with observation
at each grid square
Max. Nr. =31
for Spring, 1949 - 1979
88W
78W
68W
48W
58W
68W
58W
38W
28W
8W
18W
M4 12 1427321
.....................
.
XXX
.
Xl:
50N
21 21
29 29
4" 29
4UN
2427
14 21 20 18 19 15 1
20 3
21 23
22 25
15 20 21
21 25 27
14 15 16
21
19 21
I Is
22
29 2
25
19 16
13 12
20 21
17
23
21 20
'
12 II1
10 10 10
2 IS I IS 0 2j 3
'
16
17
19161614227 20 22221 23 26
15I11217~
1117
1615
15
23 20
27 2
26 25
25 22
1 191 169
9 1I9 1920 220 2 32204219
1 2 02521 24
22 23 25 2 0
17 21 11
28 25 23
14
78W
14 13
13
1310
27 13317 18 15
2519 16 20 21 26 25.22 20 17
16
14 14
16 1
68W
22
17
30N
13 1(
7
12
14
5
J. 59 15
5 1
16 17
18 21
J
58W
74 80 11
MIZ .. 4
E
48W
3 1
1 122
s2
20N
1
1 9 1 3 1
5 7M
19 18 16
2
1 9 1
P
12
17 21 18 16 19 21 21 24 20 19 16
1IQ 17 17 16
1719 21 16
16 15
27 13sZ 1 2110 15
15 1211 412
2
16
2 23 2 20
171
24 132 12 17
IS2 Is 16 16 7 14 15 2 14 1I 13 1I 10 E 10Is 16
Is 21
24 22
40N
2 2
20 S 7 13 I 2 21 22 22 1
3 21 26 26 23 17 12 12 13
19 27
12
88W
50N
24 25 25
2f
19 19 22
20 I
1821 2220
25 25
26
6 29 29
1012 13 16
2
21 27
14 16 16 10
313 1019 1217 16
13 1221215101
11
1416 18
14~1 1412 12 101610
20 1313
18 10
27 21 20 26 25 23 24 25 13
17 14 1152 14 11 13 611 !P12
1 17 20 17
2
27
26 24 24 24 2211 14
18111416111013
1513 11151 21 2 61 5
7 6
5 t 10 12 11 16
11661
25 2
10 ID 1I
20N
2 21
21 27 21 21 27
S 27
i
21 29
25 24
27
28 29 21 25 24 22 19
2. 29 29
30N
'il
I2
6291
......
30 30 29
32
9 30
I10
29 A 24 23 20
30 30 A 21 27
ION
I
17
211
10
1 1
, lII
12
38W
28W
18W
8W
2E
ION
MLD
Number of years with observation
at each grid square
Max. Nr. =31
for Winter, 1949 - 1979
78W
58W
48W
2E
28W
z-Tr.50N
50N
11
1r
7
14
US
19191612E
3 22IsIs11
* 12 3*-
252624
24 23Is14 12
1211 14II191
2226 2126
25 I815 14 12
1616
2527 2425
2318 16 1617
is1616 1619211
1 2324 2524 25
2623 111920
161616 1I
2 24243 26 262219 17 19
3 V
isI5 5 15-
23 2425252525
22 17 15 1616
14161 15
Is2020 1821 19 1I2 23 2324 252424
18 1412 1313
13
14 17 2321 2019 24
23 Is12 11 11
14
232323 2123
22201214II
16
242224
1114 II I
10
0I 10 21217 ~ 2
2522 16 14 14
13
2522 2217 11
17
14 17 202323
2525..
2219 15
20
1012 11 131121
24 232319 14
a3
n
20232016 15
23
Lil
L
I
'1
Ii1614
24 8 22232325
k ull131116 16.1:A::0
20N
17 15Iii
ii
ii
5 222 19
ION
121
23 2
40N
30N
11 I4
19 2
7
10 12 16
0L
13fl
78W
68W
58W
15||55||5|1
.
ON
17
18
2ON
6
1614
*.0$ 1512,II1
88W
-
4 ON
48W
38W
28W
18W
8W
2E
1ON
MLD
Standard deviation at each grid square
In meter
for Fall, 1949 - 1979
68W
78W
88W
58W
48W
48W
17 13 15 5
7
28W
28W
38W
38W
9
12 3
50N
RE2
x9
@MiME~i
....... N
22222%~.
l. ...
..
...........
....
R
1
10 ~~
~
20 1122 1116 14I
10 14
16 39
17 Is 19
19 19 16 1117Is 11 1
Il44
It 18
11
15 12 111616 14 10 9 91I..........
19 11
17
9
)
17 16 15
1412
13 Is
13 14
21 23
I
121 13 11 11 9 9 14 10 9
910 ,
2
0 1
9,
10D 10
9 1 1
5.I61-g
S1 15 1 11 31
I
3B14
B 9 7
9 11 1 1313
14 13
15 16 13
II101210 8 I
10 IQ12 10 9
10 I 12 ID910...........
*
1212 13 11 9 8
12 12
1214 12
11IIII11 10 9 I00
10 12
12 11
10
10 1012 1I 10 10lH
II 14
8
9
13
15
Isis
7
4 6 11 2 13 13 12
8 12
9 7 12
7
10
8
12
1
79191 ~ 1 13 13 IS 1O 7, 1 6 1 1 9 1 9 1 12 11 I09 10
20N
61
6 1 1 1 11 1211
6
1 6 1 9 7
88W
9 6
13 11
10 7 7
1.1 9 13 12 10 9 7
I II1
10 14 I013 12 9 11 9 9 11 1 12 9 1
9 8
1
IQ5 1 13 I0 13 11 a 9 10 9 9 1 11 9 7, 5
11 11
7
11
'61
13 12 11 13 13 9 12 12 1 9176
1 1
109
7
9 10
9 1512121
1 9
78W
.~
68W
10,
102 10 6
58W
5 4
48W
11 7
10
10 1
38W
40N
a!EE.EMSE
MEN HEE.Ea
9 I 12II10 1
30N
1010 9
11 1 2
1012II Z
1 12
9 11
50N
12 91 12 12 12 11 1Q 9 0Q
13
Is 12
6 561011
.9
1~. 6~
19 17 14 1316 12 131Ie
10 11
10 11 13 12 12
.... ~ 12 .b:...12 6 12 fiiii4 13 12 1I 12 11 11 11 1 9 11 13
70 9 10 9 11 16 14 12 10 1
ION
15
5
18 19 19
14 19
12 91911
.8 %7 01 2
9i9 13151414 7113 10
13111 12 12
108 I 10 1 11
9 96 7
1s 16
0 1
24 13
11 11
30N
1921 16 16a 26
14 Is
7 13 13
1 1 13 1 8 11 18 IS 14 13 14Is16
1 4 11 14 1 2 ISI6II12
14 13 15 11 14 7 I5 14 14 16 149 14
12
12II
:6
8
7I
20
Is 15
14 15
1 12 11 11
22 1
40N
8W
12 16
I 13
ii3 t U 5 * 14 1612 142
1112119
10
35
1 ERIEEESN
S/EE 1
18W
7
9
9 1Q.
1
101
.
20N
5 ............
9
2
5S
iX5
6 5nj
4 65. 14W
:
.
...
i
Zl%~lli!iisilllilliiil@@~!i
5~~~~~~~~---.::.
6 5
7
.'
5
1 8
28W
3
18W
8W
2E
ION
MLD
Standard deviation at each grid square
Inmeter
for Summer, 1949 - 1979
78W
7RW
88W
RRW
38W
28W
18W
7 6 7 8
9 10 1
10 9 7 7 9
7 10 1
6 4 6
9 10 7
S10I 11
1 6 11
1 4 5
9 12 1
S
6 5 5
17 a 7 8
71
1 9 7 7 7 6 6 5
4 9 11 9
7 1 1
6 5 5 5 7 1 6 6 6
7 1428 12
9 15 10 1
151
13 9 10 6
7 7 9
5 4 4 6 7 7 7 6 5
3 5 7 4 5 13 9 13 10 13 21 13
is II 1 5
5 a 6
6 6 4 4 7
19
7 1 6
5 5 1 1 9
9 7 6 6
7 5 5
6 7 8 9 6
6 5 5
7 1 5
6 8 9 a
3
S...eC..-..4
-
48W
58W
68W
6RW
l
4 6
4 8
3
50N
R'
332
.l2
4
.1
g3
2 3 5 8
40N
..........
..............
$$3
S8
7 6
2
7
4
4
7 9 11
10 11 11
8 7 9 9 7 7 12 9 1 6
7 10 8 7 8 6 6 6 8 7 6
1
9
7 5 A
7 9 9 9 7 6
8 8 5 6 6 6 4 4 3 5 6 6 6
6 5 6
7 7 6
7 6 9
10 1 1 10 9
9 11 11
7 9 10 9 1I
12 11 13
....
6 9 101010.:i*
19 4 5 4 8 7 6
6 7 5 7 5 5 6 7 7 7 7 5 4 6 9 8 7 9
12
14 10 7
9 1I1
10 1 7 4 4 1 7 6
7
7 7 6 1 7 6 8 7 4 6 7 7 10 11 a 8
10
14 13 4
9 12 ~
37 29 12 10
8 1 7
31 21 26 10
7 11 7
9I I 9
7 8 6 a 9 7 6 1 I 1 9
4 4 11
,656
7 7 4
9
6 9 7
is 23 33 21
9 9 10
6 6 9 9 10 2
20 20 24 20
7 16 12
6 7
15 7 13 is
9 6 5
6
9 19 1
4 2 4
,I8
9 9 I6
28W
8 9:7 7 7 8 9 1 7 10 1 1 1
9 10 9 1 t
1, 6 6 7
17
1
9 11 12 10 10 1 771
8 10 8 1211 9 13161011 9 7 7.1110
6 9
71
.
6
:
6 6 6 1 1 9
88W
10 1 11 14 16 11 12 10 1 1 1 1 1 1 10 12 10 7 11 14
15 16 14 13 8 19 9 12 9 911 1616 11 8 9 11
6 I5 12
78W
1 I
68W
.I Q6 6 12 16 17 4
58W
48W
1 I II
IQ11
6
38W
6
50N
40N
6 5 6
10
8 6 5 6 5 5 9*
6
5 4 6
6 6 6 5 4 6 4 5 6 6 5 4 6 8 9 7 7 7
6 5 6
8 7
7 6 5 6 5 7 7 7 5 5 6 5
6
6
76 6 5 4
1 8 7 7 6 4
Q 1 8 6 4 4 5 4 6 5 5 4 4 6 1 17 6 6 5
30N
ION
4
4 1 7 10
4 5 6
.116565111..6.
.
....
20N
3
!"i'l~
8W
30N
.U8
j
.
.
...........
20N
<
55
18W
8W
2E
ION
M
MLD
Standard deviation at each grid square
In meter
for Spring, 1949 - 1979
7RW
88W
68W
48W
58W
58W
RRW
22
15 16
36 31 33 2623021530 20
a 26 25 21 14 10 16 13 12 l]
17 17
18 30 27 25 19 21 27131 30
a a 27 24 31 21 15 1211 12
19 25
25 238 21 16 21 30 36 43
27 37 3 34 31 35 19 14 15 10 7
it 3
3 403
23 33
34 41 33 41 42 2 3 19 17
42 38 24 38 39 22 2 Ii
25 23
26 3
25 21
2 35 36 36 54 30 3 36 47
38 353 26 I3 24 26 3436
EIM24 42 13 18 20 16 23 31 39 39 38 23
31 33
4546 22 19 24 22 30 3 32
3 32 38 28 3823 2622 4
-'0' 3 21 39 25 32 29 34 3 40 49 31 25
28 31
23 20 18 23 19 20 24 29 31
35 34 2926 29 32 25 328 2
12 12 19 26 34 35 34 21 31 48 45 45 23 21
25 26
21 21 I7 16 26 26 1 25 30
4S 2 23 19 28 3532
21 32
25 22 19 17 2 33 23 26 43
5823 2323 24 3 32
34 21
18 20 30 35 21 3 11 37 54
16 17 11 13 12 18 11 1* 14 18 16 12 16 16 24 19 16 22 23 14 14 3
30 31
23 51 32 34 12 12 40 31 44
4 36 21 20 21 25
43 19 19 19
18 IS 23 16 13 26 16 12 6 19 34-19 13 14 14 15 16 16 21 21 14
25 23
37 30
20 5
220 14 19
22 2 21 10 15 17 18 2m11 24 28 M 12 15 13 17 17 1 22 22 21
14 25
36 24
23 29
12 14 12 I17.
1814 I5 18 19 17 20 23 21 25
16 17
2626
21 26
10 9 14 1
1 18 is 21 20 21
16 32
25
20 16
12 9 1 10
13 15 17 17 18 26 16 19 16 17 17 17 16 16 19 21
Is 1
10 16
12 11
8 11 8 8~
1 6 2 17 19 1 17 18 17 19 15 17 18
13 14
IQ 8
11 11
10 9 6 .
20 14 21115 13 Is
12 Is
18 7
13 12
It 7 7
7, 11
,9
16 II
12 6 6 6
28W
18W
jq 7 2
50N
%
..................
..... .....................
~.........
1 S.9 12 Il~
9
9 15211
9 15 22 17 14
..............
.. g1. . . . ..
..
22 2 4 4 2 3
1223
323
43 3521 1
4N
14 15 16 20 21 19 21 30 2
f
1 4 11 12 11 I
9
81412 9
11 17 17 17 18 9
1
IS
61 5 4
1615 1316 18 A
4 6 3 6 6 , 99
11 7 12,14 S | gN ,JJa1,
88W
30
3744
2 31 36 35 26 23
3 31
32
12 16 1I 16 18 23 21 21 11 22 21 26 39 39
16 16 21 20 1
4:6
ION
2E
8 7 6 3 !
42 31 29 25 2423
t
1 0132
20N
18W
8W
I
23
II
Ja
31 2220 11
14 20
g
30N
38W
28W
a
18 17 9 25 31 4 41
78W
68W
.9 6 111
58W
48W
26 37
41 22 21 20 2D 18
38W
,
2 44 38 9 16 11 14
7
50N
9 21 30 40 38 3 21 3 19 16 8 A
1l 24 35
I ii 4 .
..
37
40N
2 25 31 35 40 M2 3
ER
30N
..
.....
....................
20N
8W
2E
ION
MLD
Standard deviation at each grid square
In meter
for Winter, 1949 - 1979
78W
88W
68W
48W
58W
27
28W
28W
38W
38W
48W
39 36
%4
-
14
1
0
3 3 33 23 32 32 23 2626
31 29 22 23 34 b 29 20 30
3 22 31 17 47 42
39
21 20 Is 14
26 23 19 27 31 49 31 39 40 55
27 30
3 24 20 17 253
33 22
26 24
23 25 11 20 22
37 23 24 22 2 26 22 21 19 21 36 20
20 17
a 24 18 20 17 1
21 24 17 20 17 18
17 15
2224 19 15 17 17
.
32 34 30 2
..............
27 31 24 29 27
16 30 31 28 322
40N
S1726 27 24 29 19 20 25 24 24 1 22 22
28 22
24 25
22 20
22 20
7 13 22 21 2 22 25 24 2 25 23 19 21 19 23
11 22
23 25
23 19
20 1930 30 23 1
16 2 27 19 18 17 18 117 16 25 19 3 15
12 16
15 23
20 22
14 21 30 27 1
17 15 1512 10 14 11 13 14 12 12 14 18
11 14
12 11
18 21
21 27 24
1 26 31 11 I 11 10 12 13 11 13 11 11 15 IS
11 25 24 18 14 21 26 212 26 21 14 10 16 12 11 13 13 I 15 19 17 11
19 19 21 21
15 IS16
1111 10 12 12 14 14 14 19 18 18 10
Is 17
10 13
35 2
24 25
19 17
Is 21
14 15
26
12 22"
It
12
19
30N
16 19 1
21 21 18 20 24 21 7 16 It
1724 22 11 20 21 19 3
17 16 12 18 13 10 11
K
It
911
14 If
6 6 9 9 14 3
88W
23
30N
x.2
i
...
...
.
.
- --
amenessemammmmmmmmman
201
11 16 12 18 18 14 13 13 12 7 9 9 12 17 7 12 16 18 19 15 14 16 14 16 7
13 2 14 12 10 9 15 16 13 I 10 10 12.10 11 I 13 11 10 13 14 9 9 18 13 18 12 11 9 10 10 11 9 13 1
7 6
50N
14 18
46
............ 19 23
ION
A-
5122
..........................
12I I 14 1
20N
22 51 55 86
1
14 14 13
40N
8W
8W
23 39 63
32
24 47 463 29 3126 3 1
50N
18W
18W
43 3
11
8
5 10 10 16 19 17 14 9 11 10 10 1 8 14 14 10 12 15 1714 16 16 14 12 13 1 7 9 8 1 12 9 6
12 5 14 I 25 23 5 1114 11 11 10 13 1412 9 1 14 17 12 13 12 12 6 4 7 10 6 8 7 7 6
17 13 10 j{
78W
24 JS :6 9,9 13 14 IQ 10 9 6
68W
58W
48W
II1 1 3 8 1 3 9 6 7 4 6
38W
28W
18W
8W
2E
0IN
M M
Ala
3U 0 N 41WB
$$
NVI -
tBf
11 311't
MNL'9 1011'
I
I
II Mos
ffIN111- It Ki ll I 1Be 0311KB 1ffll 19 i t1 S
AA
An I
If0
AAA) I
Ii
1
NI -
N8t1
91DBt
ii U1 3118111311111D111 KillBI0as1i
1 11
L:4i1
ig i
KI I
E £ IU Bis
lII
I NI I I 11i0 010 3 150ILtBE £
3a31310 IeII
11in5 3 1l311
aIWIK
MI I I
50w1 M
a I1Ig
IHUWN~i haaat
NZP
N9E
Ia I I
urI'l
Il9
£
0 1111 1 KBu I I f EBI
ta
li009hl
Eli 111111D111l1 f11 t EIIIIII
KIiBE K!K IE I St08lNItN
milo al
IN NIII 13I 1111KUI
101i I!IN
I illRIto F,1
I
LIENE
EKIl
I a10MMRa
-
l5h1011
E
..............
l
s11.1
ll11111 121i
DIII 131109
Ni6
DDE EE311
311111l
fll2
QE
41 ...
:
IE
lB E I
331100
ffIll(
DD 09KE
1101KB
BI 91
10133 KUil
DIl 10
Vi
REl1 I
i...l 1 11 1 1 1
......
.N
I rI
11111111Il lillsomill
01 fB 1lEKE
NNIN lI
ID
oK EB ilu
It I KI II
il ll!
a ti130 9 Bll INSit
El £
HE li N
01011l
N9ON
il
I
11110 EKE
13(115
le1 11
L
il 9(113 K13 BE II
liti
nI
II
01111 1
Nolsi
Unf
.A7
ME I
go10
l
r!E
1 ff11 I 11 11 1l
U~A
NoN If101100
IliI3
I I1I11I9
II11EID BlE1
E
KNUN
UKi
NOk
IlD Ill 1f
8
lg t111 im
iEII ItI KI
Dil
'lo
NB'
101011LULD
%ilaff
STAbI
lNI
allI
i
UnC I
U1o I
~~~'~~~'fLINI~~
BIK g 0#~LD11
I I0lit
115 l-l 1119(119
IIU
N9E.
NZE .
f
lii
KiDIE gill
aD 0
No
SI
1 1 '9 l
li'l
lIll
ff I it
N91 -
MAQI
............ .
l
. ..
p 1.
......
Li0
...
.
. ..
131KiLt
..
NO
MOS
AM06
MOO I
MO!!I
MOMI
MOE! MOt'!
MOOI
aMI34
M091
1.11 1 ILt lI
MOLT
MOB!I
I
HIOLI
!1091
U051
6L61 - 6V61 'II'Ud JOJ
uopuindwuoo fIpq[ oq} olu polwJoli
aj1?nbs pIp2
wu s2ulpunos jo ioquinu pjol
ipiQ
Hoi'
HOUI
HOZI
flOIT
..
*..
.
NOT
M02
M08
M06
M06
9
MOOT
moot
iNU
MOlT
mot I
MOU
i I'll I
MOT'!
MOPT
MOST
MOCI
MOZI
1iE U 21if1K
IK
II K IIil il Ii
I
l I
I
alU3i2KigNiiBKIEU1 11
Oi
NSI
MOC!
most
MO9T
M091
MOfRI
MOLT
MOM
90nLI
I
1i 1 1
T091
11051
K'K I
LI KI KI
lill
I 1i.al 11U 1 a I
1311 11 11 It U Ill iI
stilliilitillip
I 1113li
~K
HOLT
HOEI
HOI
ai gi ll
ifU EK ItIUN
I ig2aI
i 111
tillI 1
ttitil#.12K111 013l3iKlliii
I
[ iL
HOI I
NU
UI I -
KEUN
ulIttI
1I UKK IE fiKaIu I I IE I I10 iU%*1*I11923111
Ili
N9
N9
z
AI I151III
K9B II
IliII I I
UK KUt 35
IK 1 1 2 1 K10 K K I I I
NPI
NOt
N91
NZt
N8E
SI Ili K
K11K11111011111a111 ttIl NEEK
illI Ki
11 M t a 1111111 1 111 91t
9
1.519 11111 Ki
lotK 31 I I II 111111111l311 30K
KI 319 111
3
K
NOVk
1111 E1ill IM
KU of K
gg m
NzI'
~EE~-llis-
Not,
N191?
poma as Ra
Moot
MOlt
MOLT
MOU
NOIt
E
101 U11. 111111U
!II KK
~~
$$1
11 O
EU
KitKE
UN LEO
IU UKl1i5K I%1Il0
N
N.......
NO
N9
ili
it 33 Kui Kili191
8k'
301 KIl K
No
ES'
13Ul11111
353 113K 1
1
2* U 33 K K N 111191
UKi I
i ltil
AUSI
IM091
MULI
AI
I
UULI
U091
3051
6L61 - 6V61 Inmn joj
uoppindnoo (111N~ oqi owi pwiujlu
oiirnbs p!pj qoio 1P s~uipunos jo ioqainu imio
K
*;..
II
1931 ~l
moot
*II
Iz
N'
IMI
M06
NZ
N
z
N9
NV5S
MOS
2K1El15K
Ko K DR1 9I K 9I1K111K11
NSI'
NOS
K
a
1141 KK13K11
D I IEElI I 13111
*
I
116 ,1 ill 1iituNlIE
111 180EK9
iNINE
flu IUliE l
KE
UKU
Ell 1118
UKlK
31159~~I
K
1111111 KU L 321
1111311gI Ia I B 11211K KK KU U KN 13133311
K111151111K1
#6 iEER g 1it I
N9C
I I l
till,N5g KU0EUil
IKI11lKU13IlilI I I Ism
N9E
Mv2p
I
5 i K KIIB
HOPI
NZ
I
Nt
80E0
a0tl
3,011
Total number of soundings at each grid square
integrated into the MLD computation
for Spring, 1949 - 1979
1I0E
120E
13E
56N
U
14
i
54N
2l2
52N
lI 011
5ON
....
46N*:**:*'**:.....
46N
51
...
lil
44N
1
lIl
41 10A 31
.fl4
54N
52N
50N
113
1 31142 lII
3234 I3I a
I11
IN
Ia
111
NNE NR EN48N
34 112112
44N
01 II11I
Ill |11
li iil 1111
142 I
I 1I 110 151 111
40N
38N
-
36N
Ii1 012333322 012
-
34N
0112
32N
-
30N
-
CINI~ li
0
111
I IlN
28N
1120123 11liii
183 31105
II U3 13 11
I3 I 1|401
11 M132131111
M11 11 330al
I1 4 1116431
6113 511021#123113
K I1E II 3113 1110 iI
11
UI 1M213 BRI I1 I
9112221 344312 22 il311
34N
1131113 nill 3111111121
01621e
1130 1333311) (11
18N
4 III15
16N
133
322i 11341112ll IIlI
111Ill
13 41
A11
14N
511
12N
ION4
31114 01
110E3
1201E
1 30,1
13013
4,01
140E
.
150E
21i1l 222l
1112315 10
170L3
180W
30N
1112 1 1..1.011
%.....
32311 131422
0311
11631321511113 013 1231
121111013l132111112
162
120
12311 1 11113 10 2m
i3l811313211
41
11133#131 II1
1131
2113110111233l5
110
1121403033
1111116500111
170W
160W
150W
1
320521112 31 133soIa
384213
1223
13631111
113341011111063111111II
Ill
I, 3
1*,45 344, 111 I,51
160E
32N
.il.l..........
il601N21
I1191145111110
512)
111333 1111
132131631121
M# l 011331112t O1141
1122
38N
36N
22N
20N
- - 40N
211012181111111311i
24N
-
42N
011
12 III1311 111210436
30111033111
1411i l222l321 11l11
-,-11 101112131120
26N
%3333110I Il
3ifil 1311142 3 115
2U533 61
46N
121il 11 0611133162
4 5 1 1111131311 36
112 4232401 211
42N
80W
M332I~5611
II5
1l32
121111 121
-.-
48N
11
90W
l
IW
i
140W
11,
130W
231 2,23
120W
11,231i,31121
IloW
100W
90W
80W
Total number of soundings at each grid square
integrated into the MLD computation
for Winter, 1949 - 1979
56N
IU0E
IS
13Eu
20E
, .
::---
IU
IJUM
-
I
IIUI W I0
f
5 W
1UV
buT
130W
I0W
120W
iflW
Ilow
A
80W
90W
-
I
I
I
i
i:iI
i
52N
54N
54
.......
v
.....
....
......
52N
5ON
I
.
50N
4141134
48N
48N
I' *9
8i;
46N
2
I
44N
2111
4
44N
13214135
-n?
42N
114
a
32N
-
-o.--
-
--
-
51351
-
311131i
32323313
-
BR~oli1
8
333333
N
28N
31313
I I ONf
421111412141 1311111
26N
J I IIn3MI433132IaN I
24N
NI414
113
1523
12151413 111
141
551
ill II I4I22go315I
-
iI151
&
1413111 43,23
.1N1351311435911
14N
..
110
120E
130E
1401
15011
213515
811
51214
1941
11
12N
I
8888
888888
88
88
8
88
32N
8
.
li
12i
3,1355 I.,
160E
170
..
. -
-1
180W
I 15.,
26N
l88888
0151 iiiii
24N
3
313131434133379
I I 14 3 0
IA
21111111
131
*k
111111St
ips ill
8
34N
28N
11213111131331 t 5421ill32
1453112
4111
38N
30N
n
16N
s
H 151
22N
18N
- 40N
36N
I
1522 11313343156
121
30"
20N
-
42N
88 8 8 8 888 8 8 ees
Il
H
36N
34N
-4oo.:~~
-
40N
N38N
46N
§
211
~-
22N
-. IV
113132313
44141351
112331111
20N
18N
16N
14N
IP
I 2 I11 11
S121 4 1 14335 i)12221
12N
-
170W
A
160W
150W
140W
130W
12UW how
IloUw
IUUW
yulw
IUOw yvw
ouw
5uw
ION
MLD
MaxNr.=31
Number of years with observation
at each grid square
for Spring, 1949 - 1979
130E
120E
110E
56N
140E
160E
150E
170E
170W
180W
150W
160W
140W
130W
110W
120W
90W
100W
80W
.~56N
11 5 12 14,14 112
I1if911
.-
54N
SON
I l2R I
o
I
151
*
..............
54N
w
11
..
iil
5
a ..
-
50N
50N
48N
46N
48N
II115314111532324221212121a3II3
I
46N
44N
44N
1.......1112..11214161.2133. .....
44N
if9 29
I
If
A I a N 2411 111212 1613s
199
1517212 23
42N
42N
175litH
ill
40N
s2525211514110
36N
9 i
12131.3
32N
av32ll
1211116 l
3
B523 1115151,
30N
N
m
5
26N
24N.
24
*21
a
21
24
911
20N
11
18N
141614 115
f11 111111162It1a21A2a3It1a3a21 11 l 1111 1
10N
IOL
120E
33
Il 1411171121 M an mi l )I M"
9.
31~
%1,1 Is1 14131519O1111232131321213031311
.
32N
. . . . . 30N2
31 31 3
28N
2
111313 Ilf
a6 2 2 2525 3 1611141
12312
1313101414
AA2242
2151613911111 19 112212245212921a21212523232526212129313212131124 149 19o
101131i 151715516111119111611211212521292a21292a29292929331a24161
a 211112
1910 I
l1215
I161 192121213 31 3111
16
121
I101411
1111
I5
141212
1113
112
111i
2s2215
525
4 132 i5A 2 a2 14111 119i 1 51714 111
9.2132111129
36N
6 ii1119151415131415 1611112152632623 3131
lii I1 alRi12
616
11 I9It 11ill 1 014
15
5191114
13011
14
Is Ii 1li 1321
111 16
4161
11
190
11
111 III
012
51
17 14
-
1
i
0. 150H
1601
170H
181W
1
17OW
15121111
16OW
22N
11
23a 5
111 1 Ulf 141511
i i
11iii
24N
lllI 114112421322121
111115 I 11
I1
1S
1 A lll Ila11 21 a 17 I5 0 U211 1 If I
-11
26N
al30 1
111116232l
315
a11
1314
11
62 1
1409
11
9-
21140111 ,
321
loll
0112
I24
il a I
B 14 is 141li
21IS2l11|
232D aB25 a111111 3 161119
12N
38N
1 113224222233311112141112421331
I111429N3 16laNB5275212121212v1716 1511
212 111141312919 1912
12 21: 21
9 11117
141 7716
3 R 2 24254 21
14N
40N1
14161613I1214 141418325233333(
111412111 111121282425B6252331914111316161514512111925s 2321262243331131
1,11141a13121
121312
12
D112225tl 1111211
16N
%
-
12131331116131111111if1a1313113a21252a29212a2a
191 9111215161111341
131253111413 12131611i21141
34N
22N
111112111514111111211141241211712424241
9i a191215I
141
24215
a22
38N
1111
,,i1
1 14
1
1
i 1 f 15
16 I
16N
X
lil it B 23 i'
H ll 13 0 113 1131i
11191
150W
18N
I 11 1 21 32231 --
12
, ,
20N
2253
2
X
"is -14N
#N
12N
1 1316117
,
140W
130W
12WW
I l 00W
90W
80W
ION
t
Intiv 9
MUL, 9
i 1 il I Hll
InWo &
CTI I
v~
ItU
I
IIC I
1MI
W71
1nl
VALJ
1111II Iti1t 1 U 11 I If6 I 11 if ( 1 I
i
11i
5in
6II6I
a if If9 ff It i U91 i UU U II If611 11 If
0
(1H 6
LU111110 11 H611ll 11
ai if 9 9I
a f111 0 I 1 11 1 I if 1 9I a a g Ii u 16
l l 1
9 91 9 Uif t It UU11111111 H11 aUfAHi f 11 OR11
E I li
f
If& i If
519
[I KU
N1 Uit9
111 911 H 1H11 0 1I6 It1 I 91
KU1.UI91111
S 1UKK
111116191
9'au
1 1I
It:, i til U If1
N91
91 11
11lU
N1
N21
NOZ
KKK U il 19la1 II NII 11 K 11 I aa
LIN
1 91UU616 a1U 6
1111
I UiUUI 6111
I 6111
96K KIIf
Ntt
K KK 1 KKEi11 U1 Il Uh RU 11 R 11 11 0 111 N II I R6 aI 1 Ull u NRt Ka 61 1
16il
U
OKL19191
t
UU119 1lUUUU I U 11 61119IR I USUIZIiU1 ilUfI UU1 1111LIN619161
I i1I 6111
ItK1 It UUUi 111
KIKItK K K1
N9z
9 9 K"it16 if
U
1 U it It U Iit Ua
IfU1110 (111 u Ui 61 61 UUt
NOI
...
U UU 9it 99 9 K IK IIfII KK Ia
'llE
99
f 2 U 919
IIN 91 11uflI I I IE119 1il
RIta
I KKIIK KK IKU U U A
auLINaK KIRIII
lfKKA
91911119991 LIUEU1EER
I KK1 60I HAE11il I I A1
,&. iji~j4
5f1Iu KK IEKZI
N U It1l tll 1191 LIl K11 9191110 HA91U
UE
KEK U 111111
ENIt t K
I KItKKKUIl Uif N9if Ii U1 K1 11 6If Ii 161 11 91 H11111111A10U Ua11 U
. ~8
0 flR KUKE IM x a UI
I I 11l I11 U1
KKKRRR2ENR
IfU11111 1 LIl 11 91111
IK
KK
...... N8ZNOE
61til 1il1911H1t I UI It Uif Ll L I Uit if
61991LI
IKK IfKKKIfK14 ,1U2U911
NOE
NSE
xtIWtFelnl
fl
KKRR
lE111
IN111111111P111 1 11111116
EE.
I1ItK K11 UU
u., if UIf
I6P1 H I1111
RIl U691 II HIf 9111LfI 111111
UIl 111
9191
K KfII I K K K K K i U U i if UI 1611119111i11111111 11
111a1I
11
KKKKKKKK it K aUi UUUEi il Ii It N UU aU KR i
iS1191 I1 A
Ii ~
a (I i
~11 itU I1
~
*
"~~
...
.i
*i
NZ&p
Il9111ILq
p
Nok
I
NO5
UUUL1
111161
N
UUU
61U IU6
E~ll
I
)U
-N
a-lw
I LI
6
I
6L61
NZ5
It lKi
i 1 N1I 6
K # 6111 O61UR
M091
-
....
MOLI
MOSI
HOLI
6V61 'joLwanl KJo
uo1jBAJ~sqo qIJA smoo
jo joqwnNl
H091
10P1
HOEI
HOtt
::.
oxe
Holl
-
N9.
flIUVI
JJUL
IIJL
II I
I I' I In all 91n 11
1116I
I
L5I7
SIL1
1111111
a1111111
Ilp11611 1
IIUif 11
UUY0na
IIJ~
flhIIYI
0I~
l9
Up
Noll"i
t
6 l L.Ig Ki a1I
a1ft911
11111 af aO aILRE
IIi
2 12It 19iU t1 9912 If 1111t1i6 LI I11119 It2111111IN91
11ti i 6 111I lt 1191N LII1If ilt 222 K KU K9 111I LII 911 i i I i a111.912 % Ia1111
R111211
B f il 11Bl 11 11It
i
iPfU
ii
l1ilt
Oil2
li
Ki lS1
9Lil112221L1
Lill 211222222l9919 lilt21EiLii it
I 91211LIII
K 1111
I
0LII0 I U11
N11I aK 9 K9aIE
al lip111 R11111221122111111ERif11i1f1111111
f 1ilI lSIO11fluis
tItK1611111
KIl 116111191K
figKEtERKKK
E4
I II1I166
I f 9fli
1|1
EE
NOZ
NIZZ
E Il91912L
2IR 1 1 1 LI 918 U1111 £1 Pi119N11 LI 1 i a i I
21 1191E
I161119
I £K116 LLi191
£
N9Z
I ai61
alLl a1
4119
K Il
21K 91K
II
1 Kita it £ 61
Il
it111aIt9t2 £1
a a1 It I LIf Hil £ 1£15191 LI 111 IlAl5E
116 ili~
iititl Il. LI I
6If I 291
LIII a2
611 K£1i £12 il2i m91
l Ei 616£
1I1f
11
ELi
m
it12
111119 LIEE
16££191121 E11 I1111221il
~K 11611E2E29l 21ERRR111l~l1l
K R 919, 11USII 91 fll11101H II I 11 1 1 0(1
1ill6 1£1 6l1L6
N a
I 9 1
H 19
NOE
1(
1[11111
li1111111
1R111111LI1191111111
1111
ill 6111116111 111LISI L l*NE2LI
1RRE4L
fS9
LKa
1110011111L9
I99ili11111111 ti I 1101 1N1161 i 111 051H UN1i
NZE
R
. . . . . ..
liif It LI
K6
1f11K1211
1111 i t 11 i221 911 at1t9 L! 19if1111 lia a1111a it K19 Li a1 Ll Liaa K16 I 191
j2I I91 L 11111 1222KZU
1111 LSIII LI £L11611111IN illLI
Novlli
9
NOE'
N11iI
iLI 11 i61 2 D
N9C'
NSI'
W 1111
n91,111
22291
91Wlilt11
O tI&SlI II(til
N
Ea
NZP
11 1 I I 1 10111
OI'il 0199111N LLI filill'
1111
662161611111111
Li2
EEEEEEELill
EEK EEEaElipEpill
S%,u
if
A1
i I
ill 1
B
2ilill2llii11Illl1iI
IOilOill1il1 6 6i 111i 111
aL 9
11
MIt
JOj
6L61 - 6V61 '1?EE
1v
owrnbs pu
IN X1A
UliAI
UO1jPAJsqo 1p!M SmuoX
jo i~qain
H9
l9E EEii
L'
l lI
1
E EE
N9P
H01
H1I
2011
MILD
Number of years with observation
at each grid square
Max Nr.=31
for Winter, 1949 - 1979
56N
110E
150H
140E
130E
120E
170E
160E
170W
180W
100W
110W
120W
130W
140W
150W
160W
90W
80W
54N
...
.. ..
...l-ii
-.
52N
50N
8i
48N
.......
46N
.
M11
.
192242423211
..................
...
44N
911911 12141 52 21252
1|1119
42N
40
38N
9
34N
3
-......... |1
181111
f 152 9 3211 16 16 141 1412 1111 242
if is
3121h i 17a Ill? 01114 B11B11215
171
11
32Isis1
II
3121
21j1424
36N
-432N
99,''
lI 0i 2263511 141 11
11
131I
1si 11iti l 111111111
211
5 1)4 2 12
III,
i 151413119
1 1 9121252121
1611292423 31323
231 1 3133,
19i1113151 141614131 14 151615 24
1 a11 11121411111
16 1111
111,3V29 212 293 33313
l 21 2 2 21 1111 Is 141 R is4 15 17 11 111 Alis
1A19311 175121119 1 Il
Ia l
9
2
1
1
1 U is15919 i I I111215
5 114 i5
9141
517
iiS 8811|i~stSMiisssSM i
.....
1 7 I a A 3 Vl a a 2 1 3 1 1
21 12 121
) I ? to 1
IIEEEEEEE
if
24 213
II
II1213
l 16
IRBB111ME
, 3 1 5 16 I5
I 12141214111
161411 1I
I 12 12
'BIE
BEE
MEIB
BEil
14I
NOT
MOB
M06
MOE!
MOZI
MOTI
Moo!
NZI
9 1j 9 1 I L I I
N!'!
L
~LL9L
a
NZZ
-
I 111111 1111I IlI
N9Z
NRZ
U 11l1
SUR
1
11l1111
n II
i 111111
I11
111111
S
I
9 L
PSEn
......
.
;*§L
11l I
~~i
N9P
ilLi
Il
LI a
I 111111111
1 0lIll lil
U
IllI
plllI
QI1
If11i
IiI1111111
IlilfIin
11 f..
m11
ul
011A
M06
MOOT
Mot
I
M&I
MOEI
l
11 L
I£a11i911 Al I1 11 t1K1V11111191
il 11BLiO111911 I1111111 i 1LI111N1IfItLI1 gin
n II 1 91 91111a fp11p
111 1111
11Uif
9!11 919111
Igua lit
tiMU
l
U119
11111
a1111111111ill
111111
NO!
z
1
NPI
N91
NB!
ii a::005
H
I
0
99
M0O9I MOLT
NtZZ
I th9 P5 4
P II I 11 11
N9Z
pil 11 I
.
.
.
NOE
NZE
1 L ii.-i,*i
SN9E
till
ti
L 11
E11S1
Ifs
4
N8E
Mell11II I LI
Li 1 it
I .0.1
1 4.
.. NOV
.
. .
NZp
RXI
2L09 N1O1 ilS
1191
it n LIH1111
'
of
11 11
T
..............-
Ht %
N 9P
00N,
S
NZS
..
1111
110I 1111111191 11.1 LiI 11 1111
91211 K1191 11 91
MO51
NOZ
NVE
119 LI 110
a
L1LIliLIlit9111 nIl 111II I£10 Li till H1111
al11N
II 01
n iIf
11 Ia IfL 911
L1I11 U 11911 HI IIfIf 11 1111 Ii U1,N
BI1 191 11 11 i 91191 I L
Li U1111i 191
Mop'I
30!
1 115 1 L LI BIll I i t
p a 11 1 III i 11 1 NI
-- .
MOB
HOZI
6|01 1 I B 1 Ill I I I I Il II
I I AI
*
&
HOEI
illII I aI101 1 aall11*4a1
II
N I
NBS
il anl
II
ElII
III
NEP
I pa
fil III9
£ iI 11
la11i Uof11 U a I ti1111 11 ill 9111
11191 IUMfilH~lIill9
1111111
Ul1111 11 Itn A I 11UI110 Il
111111
1111 Il
i11 1£9
Iif all Ifa
pi 1 Ml 1190 )191N
I lli
N9E .
HOW!
if.
ti l 111 1 1111191 if11ll
I1191
IfIf11
II
lit 1 l i11a
t111 1 Un ll 11111 II
UAlIi
NAi Al 1 11(1£111U
1111 1 all 11111ll
IIII 1il
If a a a al tIl NIf l (I a I 91 ti #..I
111111
111l
1II n1111 II 111
Eill1 L
119111l 111U 1i1l
IIin1111 it111111911(1
NOE
1.
HOST
Itl111111111111111
II(I i
ifififtlui I HH
1 11 111 1U III tI
2i II
91 011Ii
l 119 1U
rii
i19 11911
U 1i11 I LI1111W
1191 91119 n
111 11#11
iI
11All LIL n1 9111 1191911
NVZ
H091
'a Ali1
a
UI11 II111 1111111111
11 1
L
fill LI 11111 11
it 11 0111(1111101 1111
111 1111l 11111
1111 1 1111111 111111 itI Ifilli
I I t111!11 V4
1 IflI I111 p a 011aNifII I I1I
11111
LLII9
L L
ljj%
11
ii
I
1
19I
I 1 111 Ii1f11191191
1111
I I1If11111 11111a 11 11
II I L I£L I
NOt
HOLT
MOBI
111I (I Ilitl1111 fill19 K
til NIII I
1t::L I 19 L L L L 9Il11 119
Ns
MOLT
L1 1'11119
I IL L
I I I11
I ILL L I L I L L9I9I
N91
M091
MOst
Moot
99I1 £II L I'1
L I tI L £1911I I'91U1L a ti'L
MOBI
-3LI
6L61 - 6-V61 'II'J JOJ
aunbs p!J2 IplUo wUOIIU!AQp pJUpuUIS
N9
MeLD
Standard deviation at each grid square
In meter
for Summer, 1949 - 1979
11013
120E
130E
140E
150E
160H
170E
180W
170W
-
I
-
33
.. . ...
'.....
.
a
2
...... I
.5555
.........
.......
it f il432I
if1 23a33%*44
1i
5
232
44
443314571
ION
91194i79
III.
"'1
4
991
i4s
48N
17111
""11
"'"111
1444i
111 122
"''1
"""11
941
19
4411l
Iil19II
5 61
99i
4411
54655
564e
91
11
S141
11111 1
19
195111
.
46N
4
5
::
36N
5454
4441
64177
4111
711191744
119
4411
1447
S111111741 11
111
13111
199111 119
19131A1514 1214141151
f
111
lIII
lli
1II,,
32N
30N
28N
95$
26N
1 11
24N
1117112 11
S111454
12
15141
1314
91911544
9151
l
1 12111111129I1
9119 1211ll 113
172111112113 11111l
i
34N
4111n51
112 1l
171117 I1l
I 11 17 199
i
40N
......-.
491
1115
--
38N
5119111111i
i 56 11
44N
:j:2 : :
42N
I
4555
711l195 91i194
4
I:.:
1411
3341 14l 111111312127119919l
12N
4
1.234444444154
18N
14N
50N
114
511
3 334611111111114115i
16N
11111911
!Il
9 9I
17111911711
14 97
20N
~lig
52N
111
5549
24N
22N
54N
41146
1151455161166
26N
1114
7114
S~~
135 5571 546514
1E1E1Eit Iif" A 4 Iia
11
~14I61444456
28N
I I I
ai"
554
233
80W
90W
100W
11119
1 541
-14333
10
120W
5465556
5
445451171
--*iE*i
MIE'iiEil
130W
140W
150W
160W
----
----
I
-
191911 14831141715IS1111
7 7
22N
20N
1I
18N
19121
119121410
451949l11191114111i9i 111191119111191li
14
5 11 1171311119i
199
4 1 1141212
1191111 1211
1199
131 a1415I 1711
1,91
13U 11,151 9if 11 11111 12,14
1414
12 1a,
111515
I9.15 1 13111124 4 I
1117 1211II I4,14
11112121115 449
41111191114
16N
19111!2111291117
14N
4111 119011412 411 S44311651117974
,*15
12N
111.716 16
114431,7171,115112,77411
_
110H
120E
130E
140E
150E
160
170E
180W
170W
160W
150W
140W
130W
120W
110V
100W
90W
80W
ION
MLD
In meter
Standard deviation at each grid square
for Spring, 1949 - 1979
JU'JJ'I
1101
120E
130E
140E
150E
16011
170E
I
180W
I
170W
150W
160W
140W
130W
100W
110W
120W
90W
80W
56N
259101617 114 14151251121969311
..........
...........
%
125 115 2 111113119i2119 3412
17215
54N
11~
52N
24 212 12 53
521 92 7 2521
21 V2221152 2122 S262324
II
44N
........
136111I
1524
29219
2x
38N
5 21 a116252 2411
1
........
s4i
3125321
34121
2523
3...1.. 2 1 21 A a11
2611
2 11
1I
32N
14925 16534 415
33311 3123
4 433 1311
31 I
24
26313
133 4221 2231
Q5 1
2052 214
211l111l13A4452%B a 2126
11923
221922
46N
44N
42N
40N
38N
'I
36N
MM5323 216 12I |
324251125
222325222224
3RB33
3 24112 16217 1421
f11 H6142A2R222
9152 16
2 114 252129111
if 121 22 133511 3 12155511216511221
28N
48N
i 211 aa 11 jjiii 11 l
1 H1
n24
33226121131
233351
Hi
44 2237
312127
13 1
43645 413V352S
356111
3 2 1551
114 2 3 2 2131321155 2 1 232
242 21 2311
30N
50N
1
24 352 2 3161642 I1
5411 61215544422% 1 2I 54536 3 41 24
5691, 152 511
40N
34N
12
15.........
42N
36N
15'14
6 11431
3
2
I i1la 21193 1123243515515 33513 3
423l26112 9511135 I.
I
14211133l 4 4333 2691641 4237 35
223 3244 22 254 532 121
252 2112225 11
IS 36 352 22B3 50 5 M1 2221
413311 111
x3 353
5
173 24 31 24 225I
5517
34N
3 3 24191411175
32N
32121 15 I 1I :D
21
231 4136
30N
222131 3 4136226
2224
232559
6 1114151 2222111
211 9 3165
i22
2124
232
23 2211212
131 1116 2111161313 2121
2231614191 151711
1214
5 14 16 14 13 1215 16 5511
28N
1262411
21313n
26N
26N
1613
121215141214 1111 5 1419116 I16 11 3 16 3 221311
112319111123 14151514
1511
24N
2 1611
541511
1614
16141
14262411611615
22N
V21
51415116
11
212 212324
13135 1612
3 1 61 1291522 12 1112a
l1231752
51 1414.19
a 1121
20 51 51 12 1 5Q
1212
1419
1316
A 1613131i5161
29 31
6516
I 1214s1123 1519163 541
5141 142
141 243 161414115 25191616 11 11 11 I Ile
1 11 1151213 is2 a 15 16 IS192122
19 21561215 51415 1411
1
21 43 16135 1125 2 11 2119
2122l 16151613 23 35 2411
11111 1513 111III
17 159 1 1131112429211763,1 1115
S1111 22232
111213 1411111 7 oil
565614
1411 2 19215152314
13151192169111514.12
5415Is2 1
172 1 10 14 1115,1211
11013
12011
130E
1401E
150E
160E
1701
180W
11114 1,13
4,13
2516015,16
170W
160W
150W
24N
25216 3 355111 I 1 1.1
22N
91214 15 1114 1 95 1
20N
12 1 A I II
151215
111 514
18N
115 131111195511556
3 15111 l
14115511 5l21 9 15D121
4 21 25 556
13
11516121 253 321 2219 11 12 14 15 161213
32516111 115213 241 14 113
25
124 5621
51456141
S0135 51115116
ii5541
16N
14N
4
42111253
S191,6
1l414 5, 112 9,917 21160111
110W
120W
130W
140W
12N
76,15
100W
636,69
--
90W
80W
ION
MLD
In meter
Standard deviation at each grid square
for Winter, 1949 - 1979
6N.
110E
120E
130E
150W
140E
160E120W
412 4314
4N
-
31313
IN
ll
33
22353194
8 2 2 253421
44131404
3213
335501I3
21
ZN
24 234
431
40N
46N
21 2311193 Is
.
M
..
41N
23 91513450
I5a11131
311 1
........
42N
31514
4 31110434
11114 5133
5 414
1112519111 16131
5 35 33a219151034
41
D4 262
3922 251SNI
IN
)N
-
a
R221
1721
312536 24
1021
123
61 2621
x U121
22113 19
15112113if11
ti1l12429
25311 22
1111
21219 22211101 14
1 1314 14
212H
A62D224111ll
3 1125s21
3isI
121111 1 14111113
36N
13
10I 10
1111
11
147
n i 5212 11.0 1 11 11
34N
16I1511 13 13
32N
1
i
1
30N
1211
10
28N
11411
131514 13 1 11 1414 12111916II
241514I
13 161 1 5 11 I1,19.
1:69
26N
56 14
lilt
1Is15 14 17 11 16
B4 1114
14
is
11 91ll11
214 II121 141313716
I2422
14Ila 1315
115 1413
15141411
101 12142119a IQ13
611313
111 1 il 11129
loll
11110i1921 11211
12
11911191111111 1111119 1 9 3 1511243 13114
11
121111
14161414
511131
K72521103
111141 13
11591 lol 12
1213111141616 Is 171114
1111 a21a
ZN
I
24N
I
152213I19121011
22N
1121
111
9 1
11 1 7
20N
17 5121
n 1 9161211131116 161411151a 1113
111 4 12111 i 121412199i12 15111911iii7
135 13131 12
s1113211215 111111111f14 a051121
1512l
111 1411216152432115 14If 125I0 7161 14 16 51 I 11 1312
121 12
10 16 191
I
7
Ili
I I I
i
I12 1 61 12 a
11 91 14 1631111114ls111 111141221
2B14114 9 2436
144 431
101 in
1211 114 16 13 1 If11111 11717
1 51
a 12 I if1 a I I a I I I I I I I II
t9
1910C11115111,11 144121512,113 121a
19a 1217 1
ZN
11011
12I9lot 1o: IsI lo121a12a114149
12011
13011
140E
14M
12aI14 115 11 14 9
1 53 a
1 106 7 3 n 123
12111415s ,1 1 121 1,213151114, 3 1 211H,101231,13 115
1506
1601
40N
38N
111919tI
f|ilN
a3 291
34 57UI21
2Q
3 111514
2l2#1
13 1A
92 l 25
1111
1119121313
13a 1161111
IlIs5Is16
351li2221512121
011 2 1121 161419B25
1124
1 5122 IQ ! 3S21929111111111il111123
IN
''
2114
1511416151
0Ii1 111
31 2
2125
1 24 119 19161
402
1029
211
16 Il1 12 99111 233
1212111 B1 131416
15H13111212 1
l 1. 02a9213l6222
15163111293 M2
ll355333319
l 41294
251521115912211 13
14111
111415
I 1214
32623 1911l17
1l12 1614
111111 141115
1511
16151415
2N
)N
5N
I
52N
12111 1113
15 213371611
19 153
50342441
IN
tinw
.
...
4ON
6N
RN
..
912010251119 221 11011 131515
1424219 til39 31
IN
)N
IAnw
56N
54N
ON
5N
igfw
80W
5
243533311
II113162512
2 391 12
2N
ZN
iRnW
100W 11w 90W i$Jw
170P
1low
1701
180W
170W
160W
I,1aI2
2242,1411
150W
140W
fill
11i
13 13N111
l
I
I I
11 I
99
I18N
16N
14N
1412 14
12N
1
13,911il121 1313 10317I69143
.
130W
120W
110W
100W
90W
80W
I ON
.