Wind-d i en ci cula ion o he eas e n No h A lan ic Sub opical
Gy e om A go da a
Eugenio F aile-Nuez and Alonso He na´ndez-Gue a
Facul ad de Ciencias del Ma , Uni e sidad de Las Palmas de G an Cana ia, Cana y Islands, Spain
Recei ed 3 No embe 2005; e ised 4 Decembe 2005; accep ed 7 Decembe 2005; published 1 Feb ua y 2006.
[1] A ans-oceanic sec ion a 24.5N in he No h A lan ic
has been sampled a a decadal equency. This wo k
demons a es ha he wind-d i en componen o he
Me idional O e u ning Ci cula ion (MOC) may be
moni o ed using au onomous p o iling loa s deployed in
he eas e n No h A lan ic Sub opical Gy e. Mo e han 500
CTD e ical p o iles om he su ace o 2000 m dep h,
spanning one yea ( om Ap il 2002 o Ma ch 2003), a e
used o compu e he geos ophic anspo s eam unc ion a
24.5N. The ba oclinic anspo ob ained om he
au onomous p o iling loa s is no s a is ically di e en
han ha om h ee hyd og aphic c uises ca ied ou in
1957, 1981 and 1992. A good ag eemen is ound be ween
he geos ophic anspo s eam unc ion and he anspo
de i ed om he wind ield h ough he S e d up ela ion.
Ci a ion: F aile-Nuez, E., and A. He na´ndez-Gue a (2006),
Wind-d i en ci cula ion o he eas e n No h A lan ic Sub opical
Gy e om A go da a, Geophys. Res. Le .,33, L03601,
doi:10.1029/2005GL025122.
1. In oduc ion
[2] The polewa d hea luxes in he ocean and he
a mosphe e a e app oxima ely equal [B yden and Imawaki,
2001]. The Me idional O e u ning Ci cula ion (MOC) is
belie ed o ca y mos o he hea polewa d in he A lan ic
Ocean a 24.5N[Hall and B yden, 1982; Ganachaud and
Wunsch, 2000]. The esul is he mild clima e o wes e n
Eu ope compa ed o ha o eas e n No h Ame ica. A
24.5N, he MOC is composed o a wind-d i en con ibu-
ion and a he mohaline componen . Acco ding o Schmi z
e al. [1992], he app oxima e 30 S anspo ed no hwa d
by he Flo ida Cu en comes om a 17 S con ibu ion
om he wind-d i en anspo wes o 55–60W, and 13 S
om he he mohaline anspo .
[3] The moni o ing o he MOC is a key issue in
unde s anding clima e change [Ma o zke, 2000]. T ans-
oceanic sec ions based on ships gi e he mos exac es i-
ma es o he MOC. Those ha ha e been ca ied ou a e a
decadal scales, sho e imescales a e p ohibi i e due o he
cos and pe sonnel in ol ed. Di e en s a egies ha e been
desc ibed o ensu e he moni o ing o he MOC [Hi schi e
al., 2003; Baeh e al., 2004]. The p esen s udy deals wi h
he possibili y o moni o ing he wind-d i en componen o
he MOC using da a om he A go p og am. Fo his
pu pose, geos ophic anspo based on da a om A go
loa s deployed in he eas e n basin o he No h A lan ic
Sub opical Gy e is compa ed wi h he S e d up anspo
and geos ophic anspo s p e iously epo ed om ans-
oceanic c uises including he Oc obe 1957 DISCOVERY II
IGY [Lee maa e al., 1977], he Sep embe 1981 ATLAN-
TIS [Roemmich and Wunsch, 1985] and he Augus 1992
HESPERIDES [Ganachaud, 1999].
2. Da a and Me hods
[4] Unde he auspices o he Gy oscope p ojec , a
con ibu ion o he in e na ional A go p og am, 19 p o iling
Lag angian loa s we e deployed in Ma ch 2002 in he
eas e n basin o he No h A lan ic Sub opical Gy e, du ing
a c uise onboa d R.V. Vizconde de Eza. The loa s we e
ballas ed o d i a a pa king dep h o 1500 m and
p og ammed o di e o 2000 m e e y 10 days be o e ising
o he su ace, while measu ing p essu e, empe a u e and
salini y. The loa s spend a leas 4 hou s a he su ace
epea edly ansmi ing he e ical p o ile da a o he
ARGOS sa elli e o ensu e p ope ecep ion be o e descend-
ing o he pa king dep h o s a a new cycle.
[5] The loa s we e ini ially deployed a he la i udes
24.5N and 30N om he Cana y Islands o he Mid-
A lan ic Ridge wi h app oxima ely 3spacing in longi ude.
Addi ionally, wo loa s we e launched a oughly 26N,
18W and 27.5N, 42W (Figu e 1). Two di e en ypes o
loa s we e used, 5 APEX and 14 PROVOR loa s wi h
SeaBi d and Falmou h Scien i ic Ins umen (FSI) senso s,
espec i ely. The sampling dep hs consis ed o abou 60 and
90 le els o APEX and PROVOR, espec i ely. Floa s
ansmi ed da a du ing he whole yea , excep o he one
deployed a 24.5N, 20W, which s opped in Sep embe
2002. Mo e han 500 e ical p o iles we e ca ied ou in
one yea as seen in Figu e 1.
Figu e 1. Posi ions o he loa deploymen s (la ge
ci cles), loca ions o e e y e ical p o ile om each loa
om Ap il 2002 o Ma ch 2003 (small ci cles) and posi ions
whe e empe a u e and salini y a e es ima ed using an
objec i e analysis (s a s).
GEOPHYSICAL RESEARCH LETTERS, VOL. 33, L03601, doi:10.1029/2005GL025122, 2006
Copy igh 2006 by he Ame ican Geophysical Union.
0094-8276/06/2005GL025122$05.00
L03601 1o 4
[6] The CORIOLIS Da a Cen e collec s, quali y con-
ols, alida es, o ma s and dis ibu es he da a, which a e
a ailable h ough he Global Telecommunica ions Sys em
(GTS). Salini y da a om he loa s a e calib a ed using
Wong’s me hod [Wong e al., 2003], which is ecommended
by he in e na ional A go Science Team.
[7] An objec i e analysis is used o in e pola e he
i egula ly spaced obse a ions o a egula ly spaced g id
e e y 1in la i ude and longi ude as seen in Figu e 1. Each
da a poin is weigh ed, aking in o accoun i s dis ance om
he g id poin s. The objec i e analysis used is Successi e
Co ec ions, which in e s he a ay wi h an i e a i e p o-
cess [Pedde , 1993]. Annual clima ological empe a u e and
salini y da a om WOA94 (Wo ld Ocean A las, 1994) a e
used as a i s guess [Le i us and Boye , 1994; Le i us e al.,
1994]. The scale ac o and he wa eleng h o he il e a e
chosen as 0.81and 5.4in la i ude and longi ude as
sugges ed by [Daley, 1991] o la ge-scale phenomena.
[8] Ve ical sec ions o empe a u e and salini y a e e y
ansec a e shown in Figu e 2. The e ical sec ion o
neu al densi y (g
n
), which di ides he wa e column in o
di e en laye s, is shown in Figu e 3a. The i s ou laye s
(g
n
27.38 kg m
3
, oughly 700 m dep h) de ine he
he mocline wa e s ha a e a ec ed by No h A lan ic
Cen al Wa e (NACW), cha ac e ized by ela i ely small
slope isolines [Ha ey, 1982; Eme y and Meincke, 1986;
He na´ndez-Gue a e al., 2005]. The shallowes laye a he
sou hwes e n egion displays he highes salini y (>37.2)
due o an excess o e apo a ion and de ines he sub opical
high-salini y wa e . The nex h ee laye s (27.38 < g
n
<
27.922 kg m
3
, abou 700–1600 m dep h) de ine he
in e media e laye ha is a ec ed by wo wa e masses, a
eshe (<35.2) An a c ic In e media e Wa e (AAIW) laye
and a wa me and sal ie (>35.3) Medi e anean Wa e
(MW) laye [Wo hing on, 1976; Ka¨se e al., 1986; A han
e al., 1994]. Figu e 2 shows ha he MW and AAIW
signa u es a e s onge in he no heas e n and sou hwes e n
egions, espec i ely.
[9] The objec i e analysis allows he es ima ion o a
s a is ical e o o empe a u e and salini y by compa ing
he e ical p o iles om he A go da a wi h ha ob ained
Figu e 2. Ve ical sec ions o a) po en ial empe a u e and b) salini y as a esul o he objec i e analysis. No h, Wes ,
Sou h and Eas s ands o each ansec , and e ical dashed lines indica e he ends o each ansec . Top axis shows each
deg ee in la i ude o longi ude.
Figu e 3. a) Neu al densi y sec ion (see cap ion o
Figu e 2 o de ails). The dashed line co esponding o g
n
=
27.922 kg m
3
is used as a e e ence laye o in eg a e he
he mal wind equa ion, and b) in eg a ed mass anspo as
a unc ion o densi y laye o he no h (ci cles, solid
line), sou h (s a s, do ed line), eas (c osses, dashed line),
wes (diamonds, dashdo line) and he o al mass anspo
om he ou ansec s (hea y, solid line). Fo each
ansec , a posi i e/nega i e sign indica es no hwa d/
sou hwa d o eas wa d/wes wa d low. The sign o he
ne anspo is aken as posi i e/nega i e o di e gen /
con e gen low ou /in he box.
L03601 FRAILE-NUEZ AND HERNA
´NDEZ-GUERRA: WIND-DRIVEN CIRCULATION L03601
2o 4
om he analysis (Table 1). The la ges e o is ound in he
shallowes laye due o he seasonal hea ing and e apo a-
ion. The laye be ween 27.38–27.62 kg m
3
p esen s a
ela i ely high salini y e o due o he p esence o Medi-
e anean and An a c ic In e media e wa e s.
3. Mass T anspo
[10] The neu al densi y in e ace 27.922 kg m
3
is
used as he e e ence le el o in eg a e he he mal wind
equa ion. This le el sepa a es he in e media e wa e s
om he deep wa e s a ec ed by No h A lan ic Deep
Wa e (NADW) [Ganachaud, 2003; He na´ndez-Gue a e
al., 2005]. Figu e 3b shows he in eg a ed mass anspo pe
laye o each ansec and he o al mass anspo om he
ou ansec s. Ekman anspo , ob ained ac oss each an-
sec om es ima es based on QuikSca da a o he pe iod
Ap il 2002-Ma ch 2003, has been added o he shallowes
laye . The es ima es o Ekman anspo a e 0.5 S , 0.1 S ,
2.4 S and 0.7 S o he no he n, wes e n, sou he n and
eas e n ansec s, espec i ely. Posi i e/Nega i e sign
means no hwa d/sou hwa d o eas wa d/wes wa d low.
This pape uses he uni S o indica e bo h mass
anspo (1 S = 10
9
kg s
1
) and olume anspo (1 S =
10
6
m
3
s
1
). Figu e 3b shows a mainly sou hwa d low.
[11] A sou ce o s a is ical unce ain y in mass anspo
is ela ed o he s a is ical e o om he objec i e analysis.
A Mon e Ca lo me hod is used o es ima e his e o
[Hamme sley and Handscomb, 1964]. Fo each in e al o
g
n
, a sample o 2000 alues o x( empe a u e and salini y)
(x
i
,i=1,..., 2000) a e simula ed in he ange xdx<x
i
<
x+dx,xbeing he alue ob ained om he objec i e
analysis o each dep h and dxi s s anda d de ia ion
acco ding o Table 1. Wi h hese new alues, a sample o
2000 mass anspo s a e calcula ed and used o es ima e he
s a is ical unce ain y o each ansec and laye . The
s a is ical unce ain y o he ne mass anspo as seen
in Table 1, is in oduced as an e o ba in Figu e 3b.
Figu e 3b shows ha he ne mass imbalance is indis in-
guishable om ze o in e e y laye .
[12] Figu e 4a shows he ne mass anspo a he
he mocline laye in each ansec . The no he n ansec
shows a sou hwa d low a a a e o 9.1 ± 2.0 S , p esum-
ably due o he di e en b anches in o which he Azo es
Cu en di ides [S amma, 1984]. Posi ions o he b anches
a e no possible o iden i y due o he coa se spa ial
esolu ion o he A go da a. The Cana y Cu en is seen in
he eas e n ansec , anspo ing 2.2 ± 1.0 S o he wes ,
sou h o he Cana y Islands. This anspo is app oxima ely
1 S smalle han ha ob ained by Machı´n e al. [2006]
om an a e age ob ained om ou c uises o he no h o
he Cana y Islands. The di e ence is a ibu ed o a po ion
o he Cana y Cu en sepa a ing om he A ican coas ,
sou h o 24.5N[He na´ndez-Gue a e al.,2005].The
wes e n ansec shows a e y weak, and no signi ican ly
di e en om ze o, anspo o 0.8 ± 1.2 S . The sou he n
ansec shows a sou hwa d anspo a a a e o 11.7 ± 2.3 S .
[13] Assuming ha he wind-d i en low does no pene-
a e o he ocean loo and he e ical eloci y is negligible
a he lowe limi o in eg a ion, which Luy en e al. [1983]
poin ed ou as co esponding o s
q
= 27.4 (app oxima ely
g
n
= 27.38 kg m
3
), he S e d up anspo s eam
unc ion is:
1
b Zxw
xe
cu l dx ¼Zxw
xeZ0
H
Gdzdx 1
Zxw
xe
xdx
whe e x
e
and x
w
deno e he eas e n bounda y and he
wes e n end poin o in eg a ion, espec i ely, b=@ /@y( is
he Co iolis pa ame e ), densi y, he wind s ess ec o
(
x
he no h-sou h componen ) and
G
he geos ophic
eloci y.
Table 1. S a is ical E o F om he Objec i e Analysis and
T anspo Unce ain y F om he Mon e Ca lo Me hod
g
n
Tempe a u e, C Salini y DG,10
9
kg s
1
<26.44 0.4846 0.1007 2.1
26.44–26.85 0.2519 0.0413 2.0
26.85–27.162 0.2033 0.0293 1.5
27.162–27.38 0.1894 0.0277 1.1
27.38–27.62 0.1834 0.0336 1.0
27.62–27.82 0.1036 0.0176 0.9
27.82–27.922 0.0841 0.0148 0.4
>27.922 0.0395 0.0076 -
Figu e 4. a) Ne mass anspo o he he mocline wa e s a each ansec , and b) olume anspo o he he mocline
wa e s a 24.5N o 1957 IGY da a (blue, dashed line), 1981 da a (magen a, do ed line), 1992 da a (g een, dashdo line)
and gy oscope da a ( ed, solid line). The hea y black solid line is he S e d up anspo compu ed om mean annual wind
s ess.
L03601 FRAILE-NUEZ AND HERNA
´NDEZ-GUERRA: WIND-DRIVEN CIRCULATION L03601
3o 4
[14] Figu e 4b compa es he olume geos ophic ans-
po s eam unc ion o he he mocline laye a 24.5N
wi h hose ob ained p e iously in 1957, 1981 and 1992. The
obse ed olume geos ophic anspo om A go da a
exhibi s a simila pa e n o hose om 1981 and 1992.
Di e ences a e due o he mesoscale eddy ield ha is
smoo hed using da a spanning one yea . The di e en
pa e n obse ed in he in eg a ed geos ophic anspo
co esponding o 1957, is p esumably due o he ela i ely
coa se esolu ion o he da a ob ained om his c uise.
Figu e 4b also shows ha he compu ed geos ophic ans-
po is in good ag eemen wi h he es ima e om he wind
s ess (S e d up anspo minus Ekman anspo ).
[15] This s udy has demons a ed ha he con inuous
moni o ing o change in he wind-d i en componen o
he MOC is possible using da a om he A go p og am.
Howe e , hese da a a e no use ul o moni o ing changes
in he hea lux. Acco ding o Va gas-Ya´n˜ez e al. [2004],
he eas e n pa o he No h A lan ic a 24.5N has wa med
a a a e o 0.042C/y a 400 m dep h in he las en yea s.
This annual wa ming is much smalle han he s a is ical
unce ain y om he objec i e analysis (Table 1).
[16]Acknowledgmen s. This s udy has been unded by he Eu opean
Commission p ojec Gy oscope (EVK2-CT-2000-00087). The wind ield
we e ob ained om he Physical Oceanog aphy Dis ibu ed Ac i e A chi e
Cen e (PO.DAAC) a he NASA je P opulsion Labo a o y, Pasadena, CA
(h p://podaac.jpl.nasa.go ). We hank E an Mason o his help wi h he
English language and wo anonymous e iewe s o hei help ul commen s
on he manusc ip .
Re e ences
A han, M., A. Colin de Ve` die e, and L. Meme y (1994), The eas e n
bounda y o he sub opical No h A lan ic, J. Phys. Oceanog .,24,
1295–1316.
Baeh , J., J. Hi schi, J.-O. Beismann, and J. Ma o zke (2004), Moni o ing
he me idional o e u ning ci cula ion in he No h A lan ic: A model-
based a ay design s udy, J. Ma . Res.,62, 283– 312.
B yden, H., and S. Imawaki (2001), Ocean hea anspo , in Ocean Ci cu-
la ion and Clima e, edi ed by G. Siedle , J. Chu ch, and J. Gould, pp.
455–474, Else ie , New Yo k.
Daley, R. (1991), A mosphe ic Da a Analysis, 457 pp., Camb ige Uni .
P ess, New Yo k.
Eme y, W., and J. Meincke (1986), Global wa e masses: Summa y and
e iew, Oceanol. Ac a,9, 383–391.
Ganachaud, A. (1999), La ge scale oceanic ci cula ion and luxes o esh-
wa e , hea , nu ien and oxygen, Ph.D. hesis, Mass. Ins . o Technol./
Woods Hole Oceanog . Ins . J . P og am, Camb idge.
Ganachaud, A. (2003), La ge-scale mass anspo s, wa e mass o ma ion,
and di usi i ies es ima ed om Wo ld Ocean Ci cula ion Expe imen
(WOCE) hyd og aphic da a, J. Geophys. Res.,108(C7), 3213,
doi:10.1029/2002JC001565.
Ganachaud, A., and C. Wunsch (2000), The ocean me idional o e u ning
ci cula ion, mixing, bo om wa e o ma ion and hea anspo , Na u e,
408, 453–457.
Hall, M., and H. B yden (1982), Di ec es ima es and mechanisms o ocean
hea anspo , Deep Sea Res.,29, 339–359.
Hamme sley, J., and D. Handscomb (1964), Mon e Ca lo Me hods, Mono-
g aphs on S a is ics and Applied P obabili y, 178 pp., CRC P ess, Boca
Ra on, Fla.
Ha ey, J. (1982), q-S ela ionships and wa e masses in he eas e n No h
A lan ic, Deep Sea Res.,29, 1021–1033.
He na´ndez-Gue a, A., E. F aile-Nuez, F. Lo´pez-Laa zen, A. Ma ı´nez,
G. Pa illa, and P. Ve´lez-Belchı´ (2005), Cana y Cu en and No h
Equa o ial Cu en om an in e se box model, J. Geophys. Res.,
110, C12019, doi:10.1029/2005JC003032.
Hi schi, J., J. Baeh , J. Ma o zke, J. S a k, S. Cunningham, and J.-O.
Beismann (2003), A moni o ing design o he A lan ic me idional o e -
u ning ci cula ion, Geophys. Res. Le .,30(7), 1413, doi:10.1029/
2002GL016776.
Ka¨se, R., J. P ice, P. Richa dson, and W. Zenk (1986), A quasi-synop ic
su ey o he he mocline ci cula ion and wa e mass dis ibu ion wi hin
he Cana y Basin, J. Geophys. Res.,91, 9739–9748.
Lee maa, A., P. Niile , and H. S ommel (1977), Does he S e d up ela ion
accoun o he mid-A lan ic ci cula ion, J. Ma . Res.,35, 1–10.
Le i us, S., and T. Boye (1994), Wo ld Ocean A las 1994, ol.4,
Tempe a u e, NOAA A las NESDIS 4, 129 pp., NOAA, Sil e Sp ing,
Md.
Le i us, S., R. Bu ge , and T. Boye (1994), Wo ld Ocean A las 1994,
ol. 3, Salini y, NOAA A las NESDIS 3, 111 pp., NOAA, Sil e
Sp ing, Md.
Luy en, J., J. Pedlosky, and H. S ommel (1983), The en ila ed he mocline,
J. Phys. Oceanog .,13, 292–309.
Machı´n, F., A. He na´ndez-Gue a, and J. Peleg ı´ (2006), Mass luxes in he
Cana y Basin, P og. Oceanog ., in p ess.
Ma o zke, J. (2000), Ab up clima e change and he mohaline ci cula ion:
Mechanisms and p edic abili y, P oc. Na l. Acad. Sci. U. S. A.,97, 1347–
1350.
Pedde , M. (1993), In e pola ion and il e ing o spa ial obse a ions using
succesi e co ec ions and Gaussian il e s, Mon. Wea he Re .,121,
2889–2902.
Roemmich, D., and C. Wunsch (1985), Two ansa lan ic sec ions: Me i-
dional ci cula ion and hea lux in he sub opical No h A lan ic Ocean,
Deep Sea Res.,32, 619–664.
Schmi z, W., D. Thompson, and J. Luy en (1992), The S e d up ci cula ion
o he A lan ic along 24N, J. Geophys. Res.,97, 7251–7256.
S amma, L. (1984), Geos ophic anspo in he wa m wa e sphe e o he
eas e n sub opical No h A lan ic, J. Ma . Res.,42, 537–558.
Va gas-Ya´n˜ez, M., A. L. G. Pa illa, P. Ve´lez-Belchı´, and C. Gonza´lez-Pola
(2004), Tempe a u e and salini y inc ease in he eas e n No h A lan ic
along he 24.5N in he las en yea s, Geophys. Res. Le .,31, L06210,
doi:10.1029/2003GL019308.
Wong, A., G. Hohnson, and W. Owens (2003), Delayed-mode calib a ion
o au onomous CTD p o iling loa salini y da a by q–S clima ology,
J. A mos. Oceanic Technol.,20, 308–318.
Wo hing on, L. (1976), On he No h A lan ic Ci cula ion, 110 pp., Johns
Hopkins Uni . P ess, Bal imo e, Md.

E. F aile-Nuez and A. He na´ndez-Gue a, Facul ad de Ciencias del Ma ,
Uni e sidad de Las Palmas de G an Cana ia, E-35017 Las Palmas, Spain.
([email p o ec ed]; ahe nand[email p o ec ed])
L03601 FRAILE-NUEZ AND HERNA
´NDEZ-GUERRA: WIND-DRIVEN CIRCULATION L03601
4o 4