c
..
"
International Cowleil for the Exploration ofthe Sea
CM
11leme Session 0
by
Jens Meineke
Institut für Mcereskwlde, Troplowitzstr. 7, D-2252911amburg
11le Northem North Atlantie is the site of extreme interaction between the oeean aUlI the atmosphere. Changes in atmospheric forei!lg pattern are rapidly transferred by conveetion into the ocean's interior and affeet components ofthe North Atlantie Overturning Circulation. We present two examples of recent extremes in hydrographie conditions, which vary at decadal
. '
" time seales: 11le cessation of convective deep water renewal in the Greenland Sea and the rapid spreading of new Labrador Sea Water, which has extremely low temperatures due to
\ ; '
.
strong convection that started in 1988. 11le lesson from these observations are straightforward:
.
.
Any longer tenn project in the Northem North Atlantie needs to b~ accompanied by a monitoring scheme whieh enables a continuous re-definition of its basie hydrographie state.
,
'
,
Introduction nIe Northem North Atlantie is the site of extreme interaetion between the ocean and the atmosphere. Northward transport of subtropical waters into the high Arctie allows a suite of processes to convert wann surface waters into different species of cold and saline waters and into iee. Whereas the latter eventually melts and flo\vs southward as cold and fresh Surface
Waters, the cold and saline' species sink and spread southward as intennediate and deep waters
(Fig. I). 'Illis thennohaline driven circulation - 11le North Atlantie Overtuming Cireulation
NAGC - is central to the oceans role in the climate system. For the hydrographie conditions in
,
" the NOIth Atlantic this ituplies a rapid response of stratification a!ld circulation to changes in surface forcing. 111is has been demonstrated for the outflow of Polar Surfaee Water in a paper
,
" by Dickson ct al (1988)
the Great Salinity Anomaly. In the following we will denlOnstrate
1

the amplitüdes ami time-seales of such ehanges in the interior of the Northem Nortil Atlantic on
exalnpie,oftwo reeelit events:
'fhe eessation of deep reaehing eOllVeetion in ihe Greenland Sea arid ihe
,
of intermedi:ite waters in the North AtlanÜc originaiing froni enhanced eonvectioii in the Lalmidor sea.
Cessation of deeu convection in thc Grecniand Sea nie, Greeillaiid Sea is a component of the Nordic Seas I Arctic Oceall thermoilaline
",Ilere ollen ocean convection can reach down into deei> arId bouoni layers, providing the cold and relatively fresll Greenhlnd Sea Deep Water (GSD\V; T':: 1.25°C, S 34.892).
'l1uough Fram Strait the deep Greenland Sea is connected to the deejl Arctic Ocean, ",here the
\vann and rdatively saline Arctic Ocean Deep \Vater (AODW, T '=0.82, S 34.93) is fomled by slope convection (Rudels and
1991).
AccordiJig to l\1eiilcke ei ai (1992),
Meincke et ai (subm.) ami Dickson et al (SUblil.) the conditions for intense eonvectiOli in the
Greelli31id Sea itave becOllle illcreasillgly unfavourable since 1980.
l11crefore in the deep Grecnland Sea' the
bet\\'een convective rene\vals of GSDW arId advectiOll öf AODW lias clüinged in favour of AODW, iiicreasirig teinperatüre :iiid salinity for ihedeep layers arId certainly
water Ilmss composition. Fig. 2
th'e variaBöiis iil \\':lter iilas!; pröpeities and voh.imes foc' ihe perlod 1958 to 1993.
l11e large sigtials are fOUlld 3t dellths belo\v 1500 hi. '11le tinie 'vaTying balance between tite eOlivectively generated GSDW and the
Warnl arid saline deep ,vaters (\VSDW) adveeted frohi die Arctic Ocean are cleariy seen froni the :iiiii-pluise in their
nie ehanges in iile deeii wäter
ihe 'Grecül31id Sen radiate out into the Norwegiall Sea, ,vhich Dicksoll et' al (subm.) h:l've exaiiiillified on cOllpled deep' wah~r teniperaiure challges deinonStrated iti Fig.
3.
l11e conveciive renewal ofthe layers above
1500
i.e. thc fortuatoli of cold and fresh Arciic hitenllediate \Vater (CFAIW) prövides the
" , , ; . ; < ..
source waters fOf the overflo\v across the Greenland-ScotIand ridge. nIe volunle increase (Fig.
2)
ilidicates a healtlly renewal. IIO\VeVer, it should not bci surprislng ifthis ehanges, because ofthe cilanges iti thc deep themlOhalilie circulation intemal to the Nordic Seas I Arctic Ocean tuay weit affect the conditiollS for shallow alld intemiediate coilVectioit adversely..
Recerit changes of W:lter mass properties in the North Atl:lntic
the fraillework
World Oceali Circulatioll
the t\VÖ traiisatlaiitic sectiOlis sho\\11 in Fig.
4 \vere n~peatedly rün by Geniian groups, tjü~ section across the Labrador Sea is a Calladian standard section. Accordiiig to Fig.
the sectiOlis cut ac'ross all tile relev3lit Ueel>
2
.
.' I.

-
~-
-
~.-
.:' e •,'>.
NAOC-Ilaths, includirig the Ol'eil ocean convective system ofthe Läbrador sea. Tile low saiinity Labrador Sea Water (LSW) is spreadiJig .along the paths
iri Fig. 4. Dickson ci al
(subni) have ideiltiried a change iri the convective state of the Labrador Sea in \vinter 1988,
\vlteil it tümed frortl 10\\1 {O high and has stayed high since then. Luckily enough, the llOltherilmoSt seetion SI.O\\1l iri
Fi g
.4
iileets tile LSW branches several tinics.
It has been takell
iri
1991 and was since tllen repeated 5 times (the last coverage, was in August 1996). Fig.
5 sho\vs the distribution of saÜllity, with the major water masses reievaiii to the NAOC indicated.
ChallgeS of tllis distribution are giveri in Figs. 6-8, based on differences observed betwen 1991 and i995. iitis period was cilOsen, because it took "ntil 1995 thai tile by rar strongest sigrial
Oll the sectiOll - the chaiiges in
which had started 1988 in the Labrador Sea - has reached tire European contineütal slope.
.
,
Syet al (subnL) arid Rhein et al (subni) have analysed the LSW-sPreading in detail and concllide: (i) 11le LS\V-spreading across tile North Atiantic was 2-3 tinies faster in tile 90's than previous eStimates by Read and Gould (1992) for an earlief event in the 70's rl years
red (o 18 years). (ii) 111e dominant chaiiges iri the LSW were cooling and deepening, as evident from Figs. 7 arid 8.
(iii) 11lc LSW-chariges are a demonstration ofhow rapid a change of
regioiiai \\indstress curi over the Labrador Sea, which Dickson et ai (suimL) have found to initilite tile 1988 convection twent, alters the ,vater niass
in the ititerior North AtIan;, tie.
TIlere are fUrther significani changes seen from the repeat section, \Vilich relate to the NAOC:
Figs. 7 and 8 sho\v the üllper layers ofSubpolar, Mode Water tci be less satiile in 1995 than in
1991. 11lis is also tme fOf the Deimiark Strait Overrio\v alid for the Iceland Scotlalid Overriow. Uniike the changes of the LS\V \ve have nci iilfofmatioii Oll, the causes for thc obserVed
.
, changes in the upper layers aild the overflows. For the Ü{lper Jayers \Ve cannot distinguish between changes due to adveetiOli Of air-sea interaciiOli, for the overflo\vs \ve have lio further infoni13tion orl the relative signiiicailce of cluinges in source wate'i" ilfoperties versus entraillnieiii alollg the spreading paths.
Suillmary arid Condusions
Willt rcspcct to the NAOC we eari sUJiunarize tile foregoing exalnples as folIows:
In our llresellt decade ihe Noniic Seas experience aperiod of\veakenirig conveeiion, whiell has ecased to rcilew tile <lcCI) ,vaters, hut "hieh is still aciive enolJgh tci vcntilatc tlle uilper and
3

- - - _ . - - - - -
- - - ~ intenllediate layers for the supply of the overflows across the Greenland-Scotland sill.
In contrast to the Nordic Seas the Labrador Sea experiences aperiod of enhanced convection, suppl}ing colder and denser water to intemlediate levels and thus changing the water mass stmcture in the North Atlantic. 11le time scale ofthese changes is the order of 10 years or less.
Dickson et al (subnl) have discussed tlle timing ofthese changes in more detail and could relate it to regional atmospheric forcing, which was fOWld to be coupled to the North Atlantic oscillation index (the difference in atmospheric pressure between the Azores and Ice1and). 'Illis discussion is not repeated here, but its results are used for the conclllsions relevant to the topic ofthe tlleme session:
11le atmospheric forcing over the North Atlantic differs in space and time such, that - via the rapid convective downward transfer into the ocean's interior - different components ofthe NAOC are at different stages ofactivity in a particlilar period.
As a consequence, field work related to process stlldies in the Northem North
Atlantic has to be completed within aperiod mllch shorter than a decade to ensure a quasi-stationary physical environment.
From the experience ofprojects Iike WOCE in the North Atlantic any longer tenn stlldies like GLOBEC need to be supported by a monitoring scheme.
Ongoing planning for CLlVAR-GOOS should be drawn into the preparations and analysis ofGLOBEC.
4

,;
Bersch, M. (1995): On the eirculation ofthe northeastem North Atlantic. Deep-Sea Res., 42,
9, 1583-1607
Dickson, R.R., J. Lazier, J. Meincke, P. Rhines, lH. Swift: Long-tenn coordinated changes in the convective activity ofthe North Atlantic. Sub. to Ilrogr. Oceanogr.
Dickson, R.R., J. Meincke, S.-A Malmberg, Al Lee (1988): 'Ilie "Great Salinity Anomaly" in the Ilorthem North Atlantic 1968-1982. !>rogr. Oceanogr., 20, 103-151
Fahrbach, E., 1. Meineke : nle global oeean eireulation. In: Landolt-Bömstein, New Ser.,
V36, Springer 1986
Meincke, J., S. J6nsson, J.H. Swift (1992): Variability of conveetive eOllditions in the Greenland Seas, ICES mar. Sei. Symp., 195, 32-29
Meineke, 1.,
Rudels, 11. Friedrich: nIe ArcHe Ocean - Nordic Seas 111enllohaline System.
Subnt to ICES Joum. ofMar. Systems
0sterhus, S., T. Gammelsf0d and R. Hogstad: Ocean Weatller Ship Staion M (pers. eont)
Read, lF., W.J. Gould (1991): Cooling and freshening ofthe subpolar North Atlantic sinee the
1960's, Nature, 360, 55-57
Rhein, M., A Sy, A Putzka, 1. Lazier: Time seales of LSW-spreading in the subpolar North
Atlantic. Subm. to Nature
Rudels, B., D. Quadfasel (1991): Conveetion and deep water fonllation in the Aretic Oeean -
Greenland Sea System. J. Mar. Systems,
435-450
Sy, A, 1. Lazier, KP. Koltenllann, J. Meincke, U. Paul, M. Herseh: Suprisingly rapid eooling ofthe intermediate layer in the North Atlantic Ocean. Subm. to Nature
5

Legend to Figure
, I
Fig.1
Geographieal distribution of two components of the North Atlantic Overturning
Circulation: The convective regimes and the deep spreading paths. For abbreviations see text. From Fahrbach and Meincke, 1986 '
Fig. 2
of
(T), salinity (S) and volume (V) for water masses of the
Greeland Basin from winter '58 (W58) to winter '93 (W93). The water masses are
CFAIW:
WSDW:
GSDW:
Cold and Fresh Arctic Intermediate Water
Warm and Saline Deep Water
Greenland Sea Deep Water
Fig. 3 Time series of srnoothed rnonthly rnean temperature in 2000m depth at OWS"M" in the Norwegian Sea (upper curve) and mean temperature between 2000m and the bottoln in the central Greenland Sea (Iower curve). After 0sterhus, Gammelsr0d and
Hogstad.
Fig. 4 The loeation of repeat-sections and generalized spreading paths of Labrador Sea
Water. LeUers denote basin names. From Sy et al, subm..
, '
Fig. 5 Salinity distribution along the section between Greenland and Ireland (see Figure 4).
The major water masses are denoted
SPMW: Subpolar Mode Water "LSW: Labrador Sea Water.·.
.
.
ISOW: Iceland-Scotland Overflow Water DSOW: Denmark Strait Overflow Water
After Bersch, 1995
Fig.6 Salinity difference 1995-1991 along the northern section shown in Fig. 4.
• ' • • • ' I
Fig. 7 'Longitudinal distribution of differences 1995-1991 of LSW core properties depth, potential temperature, salinity and density. Longitudinal means inserted.
, \ .
Fig.8 Basin mean TS diagrams for Sept. 1991 and June1995::: a) Irminger Basin b) Basins east of Mid-Atlanlic Ridge
.',
,
6

V(km'. 10')
200
150
100
T("e)
-0.83
085
-.
-"-i---
-<l.87
50 -0.89
.s--.........CFAIW
. / v
• ".
~.</
..
i
:
~.
\ \
\ , :
.I'
'\ ,
"-:'
"
;;;;-J'----W-82 r-,
----W"T~-9--W9
-3-
I
S
34.875
34.870
34.865
34.860
400
34.914
300
34.910
200
34.906
100
300
200
100
-....---j'---r--.....---.-
W58 W82 W89 W93
·1.18
.1.20
.~G~SDW
••/
....
••••••••\ \
~
34.898
34.896
·1.22
___. _..;/
~~-....-----.----;-
W58 W82 W89 W93
34.894
3000m ZOOm o
Formotion oreos
-
IM Mixing of walers wilh 1010 salinities ollorge Oeplhs n.n
Mixing 01 wafers wilh 1010 ond high solinil ies
01 shollower deplhs
7

60'
·0.10
·us
·1\0
·1.l5
~
0
Q.
a l!
Q.
.
~
..
E • ...
E
·1.20
·1.25
·1.30
•US
-+rTT"rTT"rTT"rTT1rTTlrTTlrTTlrTTlrTTlrTTlrTTlrTTlrTTlrTTlrrnrrt-
.,50 '''0 "10 "10
0"
• WQCE SIH,'lIon AI
'1"
8

~
-
.
:
'
.
• I i
I
_ _ _ -
U)
U)
~ tD com 0
( 0 ( 0 " "
UI Ul 111 Ln Ul
, ,
.
,"
_ _ N - - -
. . , : p l n t D , . . . m m O N f 9 ) ; l U l UJ
,...,... ....
-
.....
-
CD
...
tn 0
,...,....,...,...IOIOCDQ)CD OJ fD Q) CO
\ntll "" Ul In Ln lfllll Vl Ln Ul
- - - - -
-U)~CQcn
n
-
_
-
N
-
M
-
~
- - - -
" , 0 1 0 _
-
N M
an
_ _
CD~t\I m DlOlcncnm Cl 0 0 0 0 0 0 _ _ -~l'\I
Ln,.'" In tll 1ft l.O
IIH,Uf)1f)\Il CD lO tD tD U) tDtDU) U) U) CD tD lDUlllnDDU)
. 1000
:, !
.1500
i:
2000
; : i!=.
~
2500
I '
3000
3500
. 4000
(pSU)
1995 - 1991
1
4S00.L.--r40-~:J)--T36-"""34--3""'2--rJ)--~28----26"'-----'-24----'-22--r20-'-1T"8---'16'---OV-L
9
I ,

LONGITlU. I u\lJ
.4...
_ _ '-t
OOOT
GOO
400
200
LONGITLD: (°111
-45 -40 -35 -30 -25 -20 '15
O.6t---+---+---+-~i=-~+---+--....:t
-10
88=
oe
..
..
0
-400
8Z= +155 m
-600
-BOOI+---+--+---+---t---t---I----l-
I -
~ -0.'
-0.6+---+---1f----t---+--+----+--~
LONGITlDE (0\11
·45 -40 -35 -30 -25 -20 -15 -10
O.OO+---+----I---I---+---1I---t---+
0.04
L(HjITtD: (OV I
-45 -40 -::5 -30 ·25 -20 ·15 -10
O.06-t----t---t---1f---+--....::t---t:------:t
&
&
0.04
0.02
-0.04
8S=
-O.OO+---t----I---I---+---1I---t----l--
&
-0.04
-0.06+.---+----1---1--_+_--1----+---+
'
..
34.!15
JS.OO
~
...:
~
>]
9
9
6
10
5
12
11
34.115
tI
13
34.00
34.!15
JS.oo
JS.l!i
JS.IO
JS.15
JS.20
35.Zi
2+---+=::.--+---+--;;-;;:~--+-----<---+--J
,
10