North Brazil Current Ring Generation and Evolution ...asl.whoi.edu/publications/documents/papers/seawifs paper.pdfNorth Brazil Current Ring Generation and Evolution Observed with ...

1058 VOLUME 32J O U R N A L O F P H Y S I C A L O C E A N O G R A P H Y

q 2002 American Meteorological Society

North Brazil Current Ring Generation and Evolution Observed with SeaWiFS*

DAVID M. FRATANTONI AND DEBORAH A. GLICKSON

Department of Physical Oceanography, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts

(Manuscript received 25 April 2001, in final form 22 August 2001)

ABSTRACT

The earth’s largest oceanic rings are formed by the retroflecting North Brazil Current (NBC) near 88N in thewestern tropical Atlantic. The NBC flows northward across the equator and past the mouth of the Amazon Riverentraining river-influenced shelf water along its nearshore edge. Enhanced phytoplankton production associatedwith the nutrient-rich Amazon discharge results in near-surface chlorophyll gradients that delineate the trajectoryof the retroflecting NBC. These large-scale gradients, visible from space using Sea-viewing Wide Field-of-viewSensor (SeaWiFS) ocean color imagery, enable visualization of NBC rings during the initial phases of theirevolution and northwestward translation. Observations of 18 NBC rings identified between September 1997 andSeptember 2000 are summarized. Six rings formed each year. Although nearly circular at formation the ringsquickly deformed as they translated at speeds near 15 cm s21 toward the Caribbean Sea. Typical core radii ofrings near 558W were 100 km and 150 km in the across- and alongshore dimensions, respectively. The contributionof each ring to intergyre mass transport (1.0 6 0.4 Sv) was estimated using SeaWiFS derived surface areas andan estimate of vertical penetration (600 m) based on in situ tracer observations. Several rings were observed(using satellite-tracked surface drifters in combination with SeaWiFS imagery) to violently collide with theLesser Antilles. At least one ring maintained an organized circulation while passing directly over the island ofBarbados.

1. Introduction

The North Brazil Current (NBC) is an intense westernboundary current and the dominant surface circulationfeature in the western tropical Atlantic Ocean. The NBCseparates from the South American coastline at 68–88Nand curves back on itself (retroflects) to feed the east-ward North Equatorial Countercurrent (NECC) andclose the anticyclonic (clockwise) wind-driven equa-torial gyre (Fig. 1). The retroflection of the NBC isdynamically similar to the Agulhas Current south ofAfrica (e.g., Lutjeharms 1996) and, like the Agulhas,the NBC occasionally curves back upon itself so far asto pinch off large warm-core vortices (Johns et al. 1990).These NBC rings, which can exceed 450 km in overalldiameter and 2000 m in vertical extent, swirl anticy-clonically at speeds approaching 100 cm s21 while trans-lating northwestward toward the Caribbean on a courseparallel to the South American coastline (Didden andSchott 1993; Richardson et al. 1994; Fratantoni et al.1995, 1999). After translating for 3–4 months the rings

* Contribution Number 10462 of the Woods Hole OceanographicInstitution.

Corresponding author address: Dr. David M. Fratantoni, WoodsHole Oceanographic Institution, Dept. of Physical Oceanography,Woods Hole, MA 02543.E-mail: [email protected]

are destroyed through interaction with abrupt topogra-phy in the vicinity of the Lesser Antilles. Locally, NBCrings and their filamentary remains episodically disruptsurface circulation patterns in the eastern Caribbean,impact the distributions of salinity and icthyoplankton(e.g., Kelly et al. 2000; Cowen and Castro 1994; Borstad1982), and pose a physical threat to expanding offshoreoil and gas exploration on the South American conti-nental slope. Globally, the five to six rings generatedannually by the equator-crossing NBC are responsiblefor up to one-third of the equatorial-to-subtropical masstransport associated with the upper limb of the Atlanticmeridional overturning circulation (MOC), a funda-mental component of the earth climate system (Fratan-toni et al. 1995, 1999, 2000; Goni and Johns 2001).

Due to their geographic location NBC rings are par-ticularly difficult to investigate using satellite remotesensing techniques. Rings shed from subtropical bound-ary currents (e.g., the Gulf Stream, Kuroshio, AgulhasCurrent, East Australian Current) are often discernablein satellite infrared imagery due to the contrast betweentheir (warm or cold) core temperature and that of thesurrounding environment (e.g., Brown et al. 1983). Rel-atively weak surface temperature gradients in the west-ern tropical Atlantic warm pool (e.g., Servain and Lukas1990) make sea surface temperature (SST) imagery in-effective for NBC ring identification. Similarly, the rel-atively small sea surface height (SSH) signature asso-

MARCH 2002 1059F R A T A N T O N I A N D G L I C K S O N

FIG. 1. Cartoon depicting the major upper-ocean circulation features in the western tropical Atlantic. The South Equatorial Current (SEC),North Brazil Current (NBC), and North Equatorial Countercurrent (NECC) close the wind-driven equatorial gyre. NBC rings provide amechanism for mass and tracer transport from this gyre through the tropical gyre and into the southern extent of the North Atlantic subtropicalgyre.

ciated with the azimuthal velocities of the low-latitudeNBC rings are difficult (but not impossible: see Diddenand Schott 1993) to distinguish from an energetic back-ground eddy field using satellite altimetry. Recently,Goni and Johns (2001) assembled a census of NBC ringsusing TOPEX/Poseidon SSH anomaly data merged witha mean field derived from historical hydrography. Whilethis approach appears promising, the relatively low tem-poral (10 days) and spatial (28–38 longitude) measure-ment resolution inhibits detailed altimetric study ofNBC ring structure and evolution from altimetry alone.Pauluhn and Chao (1999) combined TOPEX altimetrywith an eddy-resolving numerical simulation and wereable to identify a number of translating NBC rings.

Because of the difficulty of remote observation, NBCrings were not ‘‘discovered’’ until the late 1980s. Priorto this time it was generally understood that the westerntropical Atlantic is an energetic and eddy-rich environ-ment (e.g., Bruce et al. 1985). However, the essentialevidence that discrete, westward-translating rings areshed by the retroflecting NBC resulted from the analysisof Coastal Zone Color Scanner imagery (Johns et al.1990). As reported earlier by Muller-Karger et al.(1988), ocean color imagery of the western tropical At-lantic reveals a unique tracer of ocean circulation. TheAmazon River discharges onto the continental shelf ofequatorial Brazil resulting in elevated nutrient concen-

trations and enhanced biological productivity. A plumeof phytoplankton-rich, high-chlorophyll water is ad-vected northwestward along the nearshore edge of theNBC and into the interior as the NBC retroflects intothe North Equatorial Counter Current (NECC).1 Re-motely sensed surface chlorophyll measurements de-rived from ocean color observations reveal filaments ofhighly productive Amazon-influenced water adjacent toand surrounding relatively lifeless midocean water. Thiscontrast permits visualization of the cyclic advance andretreat of the NBC retroflection and the accompanyingformation of pinched-off NBC rings (Fig. 2).

In this article we present new observations detailingthe generation and evolution of NBC rings using threeyears of Sea-viewing Wide Field-of-view Sensor(SeaWiFS) ocean color measurements. We also presentand make ancillary use of surface drifter trajectories andhydrographic observations resulting from a recent Na-tional Science Foundation–sponsored regional field pro-gram. The acquisition, processing, and analysis of theSeaWiFS data is described in section 2. Our results,including a summary of ring characteristics and casestudies illustrating the genesis and demise of three NBCrings, are presented in section 3. A brief discussion of

1 See Geyer and Beardsley (1995) and accompanying reports fora recent summary of Amazon River outflow plume observations.


FIG. 2. A SeaWiFS chlorophyll a image depicting the NBC retroflection and a recently separated NBC ring. Relatively lifeless midoceanwater (dark blue; within the retroflection and the ring) contrasts sharply with the highly productive waters influenced by the nutrient-richAmazon outflow. Note also the plume of high-chlorophyll water spreading northwestward from the mouth of the Orinoco River. The abilityto track NBC rings with SeaWiFS degrades significantly in this region.

these results is provided in section 4, and our conclu-sions are summarized in section 5.

2. Data and methods

SeaWiFS measures chlorophyll a and water leavingradiances at six wavelengths (Hooker et al. 1992;McClain et al. 1998). We obtained daily global fieldsof SeaWiFS level-3 chlorophyll a data from the GoddardDistributed Active Archive Center at the NASA/God-dard Space Flight Center for the period September1997–September 2000. The level-3 data product con-sists of calibrated, atmospherically corrected data cor-responding to a period of 1 day and stored in a global,equal-area grid with cells approximately 81 km2. Fromeach daily global field we extracted a regional subfieldencompassing the western tropical Atlantic and the east-ern Caribbean Sea. The sole use of SeaWiFS imageryin the present study is for identification of near-surfaceproperty gradients. Further details regarding theSeaWiFS instrument, calibration methodologies, anddata processing can be found in Hooker et al. (1992)and McClain et al. (1998).

Due to the intense convection associated with themigrating intertropical convergence zone (ITCZ) thewestern tropical Atlantic is often obscured by clouds.Individual daily images in the study region ranged fromcloud-free to 95% obscured with significant daily andseasonal variation. To minimize the impact of occasionalcloudiness on our ability to discern large-scale spatialstructures we generated a sequence of 265 overlapping

7-day composite images centered every fourth day. Thecompositing algorithm retained the lowest value fromeach grid cell of the component images.2 The 7-daycomposite period is short compared to the 50–60-dayperiod of NBC ring generation, but is approximatelyone-half the rotation period of a translating NBC ring.Based on examination of several representative com-posites and their component images we do not believethe smoothing and smearing of spatial features inherentin the compositing process significantly impacts the re-sults presented herein.

Each of the 265 composite images was manually an-alyzed to determine the position of the tracer frontsdelineating the onshore edge of the NBC, the southernedge of the NECC, and the circumference of any trans-lating NBC rings. The position of the northwest cornerof the NBC retroflection was recorded for each image,as were the center and major and minor diameters ofany NBC rings present. Time series were constructedof the ring center position and along- and across-shorediameters for each individual ring. The time differencebetween the observed separation of a ring from the ret-roflection and the passage of the ring center through the558W meridian was used to compute northward and

2 Unlike sea surface temperature imagery in which the coldest pix-els generally signify clouds, the SeaWiFS level 3 imagery is distrib-uted with cloud values set to an arbitrarily high flag value. Hence amultiday composite that selects the lowest chlorophyll value elimi-nates cloud-flagged pixels in much the same way a warmest-pixelSST compositing algorithm would function.


TABLE 1. Rings identified and tracked using SeaWiFS.

Ring Date formed Date passed 558WDate tracked

with SeaWiFSWest velocity

(cm s21)North velocity

(cm s21)Speed

(cm s21)

ABCDE

10 Oct 19972 Jan 19983 Mar 1998

16 Apr 19981 Jul 1998

23 Nov 199726 Jan 199825 Mar 19984 May 1998

27 Jul 1998

9644445064

8.813.414.621.414.8

0.93.80.66.42.5

8.813.914.622.415.0

FH*GIJ

1 Jan 199930 Mar 199910 Feb 199917 May 199918 Sep 1999

25 Jan 199925 Apr 199910 Mar 199929 May 19997 Dec 1999

5644602496

11.514.8

9.216.1

6.4

5.16.43.73.21.8

12.516.2

9.916.4

6.7KLMN

16 Jan 20004 Mar 2000

25 Apr 200012 Jun 2000

11 Feb 200010 Mar 200031 May 200012 Jul 2000

50402424

19.810.714.317.1

6.96.45.74.3

21.012.515.417.7

Ave (std dev) 13.8 (4.2) 4.1 (2.1) 14.5 (4.3)

* Ring H was not clearly observed at 558W due to cloud cover. Estimated values of position and velocity are based on the trajectories ofthree satellite-tracked surface drifters launched in the ring core shortly after formation (see Glickson et al. 2000).

westward translation speeds. East of 588W the rings aregenerally easy to visualize with well-defined cores ofmidocean (presumably South Atlantic/equatorial gyre)water surrounded by highly productive Amazon-influ-enced water. West of this longitude ring tracking is moredifficult due to additional sources of elevated chloro-phyll (the Orinoco River discharge and enhanced pro-ductivity near the Lesser Antilles) and smearing of sur-face gradients by wind and lateral mixing. Surface drift-er trajectories (discussed below) confirm that the ringscontinue to translate at least as far west as Barbados(618W) while their surface color expression graduallyerodes or is masked by competing processes.

The subjective manual interpretation of images is in-formative and relatively precise, but not particularly ef-ficient. As we hope to extend this analysis in the future,we developed a semiobjective method for inferring NBCring formation and translation statistics from ocean colorimagery. Two indices were defined:

1) Retroflection position anomaly (RPA): For each 7-day composite image we determined the distancefrom an arbitrary upstream reference point to thenorthwest corner of the NBC retroflection. This lo-cation was extracted manually from each image, al-though in hindsight a relatively simple automatedmethod could be devised. The record-length (3-yr)mean position was subtracted from each observationto form an anomaly describing the alongshore ex-tension/retraction of the NBC retroflection relativeto its mean position.

2) Chlorophyll concentration anomaly (CCA): For each7-day composite image we computed the area-av-eraged chlorophyll a concentration in a 38 latitudeby 18 longitude subregion centered at 9.58N, 558W.Rings that pass this meridian are completely sepa-rated from the NBC retroflection but have not begunto experience significant deformation due to topo-

graphic interaction. A centered, 90-day moving av-erage was subtracted from each observation to min-imize the contribution of seasonal concentration var-iations. The resulting anomaly time series describesthe mesoscale variability in ocean color at a site wellremoved from the NBC retroflection and near thecenter of the NBC ring translation ‘‘corridor.’’

3. Results

a. Ring generation

A total of 14 individual ring formation events werevisually identified during the 3-yr study period (Table1; Fig. 3). Displacement vectors depicting the overalltranslation of each ring during the observed time periodare shown in Fig. 4. Ring formation dates were deter-mined subjectively by examination of individual 7-daycomposite images. Formation was inferred when thelow-chlorophyll ring core was visibly separated fromthe low-chlorophyll interior of the NBC retroflection bya band or ridge of relatively high-chlorophyll water. Onseveral occasions a ring appeared to separate but wassubsequently observed to reattach to the retroflection.These abortive separation events were not included inthe statistics enumerated in Table 1. Rather, we cata-loged only those formation events that could be linkedto a separated, freely translating ring observed at ordownstream of 558W.

A majority of the 14 observed ring shedding eventsare correlated with sudden, large (.250 km) south-eastward retractions of the NBC retroflection and com-mensurate reduction in RPA (Fig. 5). Similarly, nearlyall of the identified rings are evident as 30–50-day pe-riods of negative CCA. These anomalous low periodscorrespond to passage of individual NBC ring cores


FIG. 3. (a) Repeating annual cycles of Amazon River discharge (solid) from Figueiredo et al. (1991) and NBC volume transport (dashed)as measured near 48N by Johns et al. (1998). (b) Repeating annual cycles of wind direction (solid) and stress (dashed) at 498W, 38N fromthe global climatology of Hellerman and Rosenstein (1983). Wind direction follows the oceanographic convention (i.e., the direction in whichthe wind is blowing). (c) A graphical depiction of the time period during which each identified ring formed, passed the 558W meridian, andwas lost to remote observation. Several of these rings were observed in situ during a series of research cruises conducted between Nov 1998and Feb 2000. The formation date of rings Q1–Q4 was approximated based on observed mean translation speeds.

(composed of relatively ‘‘dead’’ midocean water)through the 558W monitoring domain. The retroflectionretraction signaling ring generation precedes the arrivalof a low-chlorophyll event at 558W by approximately1 month. We were unable to identify a consistent re-lationship between the amplitude of the retroflection re-traction and the duration or the intensity of the subse-quent low-CCA event at 558W.

Four additional low-CCA events, two each in the fallof 1998 and 1999, could not be correlated with subjec-tive observations of NBC ring translation past 558W,nor could they be related to a retraction of the NBCretroflection. Yet, these low-chlorophyll events are al-most identical in duration and amplitude to the 14 vi-sually verified ring passage events. In Figs. 3 and 5 thesequestionable CCA events have been labeled Q1–Q4.One of these features (Q2) was carefully surveyed dur-ing a research cruise in December 1998 and found tohave the expected hydrographic and velocity character-istics of an NBC ring. Based on this in situ validationand the general similarity of the four events, we surmisethat Q1–Q4 are in fact NBC rings that were visuallyunrecognizable in the composite imagery. Given theseasonality of the Amazon discharge, the NBC trans-port, and the local wind field it is not altogether sur-prising that our ability to visually detect NBC rings viatheir ocean color signature should exhibit a seasonal

dependence. This interesting and unexpected compli-cation is discussed further in section 4.

Although NBC transport varies significantly duringthe year (Fig. 3a) we detect no particular seasonality inring generation (Fig. 6). This finding is generally con-sistent with a recent studies of NBC ring generationusing TOPEX altimetry3 (Pauluhn and Chao 1999; Goniand Johns 2001) and with moored observations (Fra-tantoni et al. 1995), but is markedly different from pre-vious numerical results and inferences from sparse La-grangian observations (e.g., Richardson et al. 1994).Molinari and Johns (1994) demonstrated that the NBCretroflection is a consistent, year-round feature in thewestern tropical Atlantic thermocline, thus dispellingprevious notions that NBC ring generation must be aseasonal phenomenon. The few existing in situ obser-vations of NBC rings indicate significant ring-to-ringstructural variability. At present we are unable to de-termine if this variability is purely random or a con-

3 Goni and Johns (2001) segregate their observations into ‘‘Rings’’and ‘‘Eddies’’ depending on the altimetric observability of the NBCretroflection at the time an anticyclonic feature was identified. Forcomparison with our SeaWiFS observations, a portion of which over-lap the TOPEX measurements of Goni and Johns, we have regroupedtheir census results neglecting this distinction.


FIG. 4. Displacement of the 14 NBC rings enumerated in Table 1during the period of observation with SeaWiFS imagery. The meanposition of the NBC retroflection is illustrated schematically for ref-erence.

FIG. 5. Retroflection position anomaly (RPA) and chlorophyll concentration anomaly (CCA) time series (see text forexplanation). Where possible, shaded lines indicate visual correlation between retraction of the NBC retroflection (indicatedby sharp decline in RPA) and subsequent passage of a ring across the 558W meridian (indicated by a local minimum inCCA). Individual measurements (circles) taken from SeaWiFS imagery are connected by a cubic spline. CCA events Q1–Q4 do not visually correspond to ring generation events. See text for details.

sequence of the underlying seasonality of the NBC andits environs.

Questions remain as to the degree of interannual var-iability in ring generation. Goni and Johns (2001) reportsignificant (factor of 2) differences in the number of

rings formed from one year to the next. While the pre-sent study found no year-to-year variation in ring for-mation, the 3-yr study period is hardly adequate to ad-dress interannual variability. It should also be noted thatthe Goni and Johns (2001) observations appear to sta-bilize in the late 1990s with approximately six ringsformed per year from 1996 to the present (G. Goni 2000,personal communication).

b. Ring translation

Once free of the retroflection the NBC rings wereobserved to move northwestward parallel to the conti-nental slope at a mean speed of 14.5 6 4.3 cm s21. Thisspeed is consistent with previous estimates from surfacedrifters (Richardson et al. 1994), satellite altimetry (Did-den and Schott 1993; Pauluhn and Chao 1999; Goni andJohns 2001), and numerical simulations (Fratantoni etal. 1995). The translation speed of individual ringsvaries from a minimum of 7 cm s21 to a maximum of22 cm s21 in the vicinity of 558W. As will be illustratedbelow in the form of case studies, the translation speedof a particular ring can vary considerably during its brieflifetime with speeds as high as 30 cm s21 occasionallynoted. A portion of the translation speed variability re-sults from simultaneous deformation of the overall ringgeometry (see below). The mean trajectory traced byNBC ring centers is generally parallel to the continental


FIG. 6. Histograms depicting the quarterly frequency of observedNBC ring generation. Note that Goni and Johns (2001) do not spe-cifically report the date of ring generation, but rather the date onwhich a ring is first identified using TOPEX altimetry (G. Goni 2000,personal communication). This difference in methodology is mostlikely responsible for the differences shown.

shelfbreak in water 3000–4000 m deep (Fig. 4; Diddenand Schott 1993; Fratantoni et al. 1995).

The relative importance of NBC ring self-propagationversus advection by a background flow was investigatedbriefly by Fratantoni et al. (1995) and in greater detailby Jacob (1997). The latter study concluded that NBCrings self-propagate via the b effect at a speed consistentwith analytical models (e.g., Cushman-Roisin et al.1990). There is therefore no requirement for a back-ground northwestward flow to explain their movement.This provides support for the notion first advanced byRichardson et al. (1994) that the northwestward ‘‘Guy-ana Current’’ historically depicted in regional circula-tion schematics and pilot charts may be an artifact ofnon-eddy-resolving measurements of NBC rings.4 Theamplitude and seasonality of northwestward flow overthe wide and shallow (,100 m) continental shelf re-mains uncertain, although moored measurements up-stream of the retroflection suggest an annual-mean flowof up to 5 Sv (Sv [ 106 m3 s21) (Johns et al. 1998).Lagrangian observations of a coastal current (Limebur-ner et al. 1995; Glickson et al. 2000) are generally in-conclusive as most surface drifters launched on the shelfupstream of the retroflection are quickly pulled offshoreinto the NBC.

4 Recent observations in the western Indian Ocean have similarlydemonstrated that the Mozambique Current is best described as asuccession of eddies rather than as a feature of the mean circulation(DeRuijter et al. 2002).

c. Ring geometry

A summary of geometric parameters correspondingto ring observations near 558W is shown in Table 2. Tosimplify the characterization of ring geometry we ap-proximated each ring as an ellipse. Mean major (minor)axis lengths of 304 (213) km yield a mean ring-coresurface area of approximately 52 000 km2 at this lon-gitude. Typical aspect ratios (ratio of minor to majoraxis length) near 0.7 suggest that the standard kinematicapproximation of an oceanic ring as a circular, axisym-metric vortex may not apply in this region. There is atendency for the major axis of the elliptical ring to alignitself parallel to the sloping western boundary, partic-ularly as the ring translates to the west of 558W andencounters topography that is increasingly perpendic-ular to the ring’s preferred westward translation direc-tion. Several rings were observed to dramatically changetheir orientation as they evolved (see case studies be-low). When cast in terms of cross- and alongshore di-mensions these observations suggest mean radii of 100and 150 km, respectively. The circulation associatedwith the NBC retroflection and separated NBC ringsextends beyond the low-chlorophyll core. In situ ve-locity measurements (e.g., Fratantoni et al. 1999; Wilsonet al. 1999) indicate the overall scale of circulation as-sociated with an NBC ring may exceed 450 km in di-ameter.

The observed ocean color gradients provide a meansto visualize the generation and evolution of NBC rings.But do the ocean color patterns we observe correspondin a meaningful way to the expected horizontal velocitystructure of an NBC ring? For example, may we assumethat the color front that visibly defines the ring boundaryis dynamically related to the radius of maximum ve-locity, typically regarded as the boundary between thering core and the external environment? As part of arecent in situ measurement program, synoptic surveysof hydrographic properties and absolute vector velocitywere obtained in and around the NBC retroflection andseveral NBC rings (Fratantoni et al. 1999). On severaloccasions these in situ surveys coincided with relativelycloud-free atmospheric conditions permitting simulta-neous representation of the surface circulation withSeaWiFS. In Fig. 7 we show velocity vectors at a depthof 15 m superimposed on a nearly simultaneousSeaWiFS image of the NBC retroflection. Note partic-ularly the line of stations near the northwestward extentof the retroflection (A) and across the southeastward firstmeander of the NECC (B). In both areas the maximumvelocity of the NBC/NECC jet is aligned (within theresolution of the survey) with the observed ocean colorfront. Velocities in the NBC upstream of the retroflec-tion (C) are at a local maximum near the 100-m isobath,suggesting that the edge of the continental shelf is boththe onshore boundary for NBC transport and the off-shore boundary for high-chlorophyll Amazon-influ-enced water. These comparisons indicate that dynami-


TABLE 2. Geometric parameters of NBC rings observed near 558W.a

RingMajor axis

(km)Minor axis

(km) Aspect ratioSurface area(km2 3 104)

Volumeb

(m3 3 1013)Transportc

(Sv)

ABCDE

345400311222256

222289233167189

0.640.720.750.750.74

6.019.085.702.913.79

3.615.453.421.752.28

1.141.731.080.550.72

FGIJKL

211367311278311333

156278167233245167

0.740.760.540.840.790.50

2.588.004.075.095.974.37

1.554.802.443.063.582.62

0.491.520.770.971.140.83

Ave (std dev) 304 (58) 213 (47) 0.71 (0.10) 5.23 (2.01) 3.14 (1.21) 1.00 (0.38)

a Rings H, M, and N are not included in this table as they were obscured by clouds in the vicinity of 558W.b Volume computed assuming ring is an elliptical cylinder with vertical height 600 m. This height is determined from consideration of

oxygen anomaly profiles obtained during in situ survey of three NBC rings (see Fig. 8).c Annualized per ring transport is computed by dividing the ring volume (m3) by the number of seconds in one year.

cally relevant geometric parameters (e.g., Table 2) maybe reasonably inferred from ocean color imagery.

d. Ring volume transport

The effective transport of South Atlantic water intothe subtropical North Atlantic by NBC rings has beenestimated at approximately 1 Sv per ring by severalauthors (Johns et al. 1990; Didden and Schott 1993;Richardson et al. 1994; Fratantoni et al. 1995). Basedon such estimates, the annual mass flux carried by 5–6 rings amounts to roughly one-third of the inter-hemispheric transport required to close the upper limbof the Atlantic MOC (e.g., Schmitz and McCartney1993; Schmitz 1995). The relative consistency of thepublished mass flux estimates tends to obscure the factthat they are based on very limited data and use dif-ferent assumptions about horizontal scale, vertical pen-etration, and number of rings that are formed in a typ-ical year.

The present dataset provides a new opportunity toestimate the transport potential of NBC rings usingmethods independent from those employed by previousinvestigators. To compute ring volume we assume eachring may be approximated as an elliptical cylinder witha known surface area (Table 2) and a vertical penetrationdepth estimated from recent in situ ring observations.Three different NBC rings were surveyed near 578Wduring regional research cruises conducted between No-vember 1998 and February 2000 (Fratantoni et al.1999). In Fig. 8 we show vertical profiles of tempera-ture, salinity, and dissolved oxygen anomalies computedat the core of each ring relative to the ring’s immediatebackground environment. As fundamental constituentsof the density field, temperature and salinity are inti-mately related to the velocity field, and hence the pro-files of temperature and salinity anomaly reflect the dis-parate azimuthal velocity structure of these three rings(Fratantoni et al. 1999). Dissolved oxygen, however, is

a passive tracer that, at depth, can only be advected bythe ring’s velocity field or diffused. The three oxygenanomaly profiles shown in Fig. 8 are similar in form,each tending toward zero near a depth of 600 m. Thissuggests that the vertical segment of the swirling ringcore that is hydrographically distinct from its surround-ing environment may be approximated5 by the depthinterval 0–600 m. This vertical interval was used incombination with the SeaWiFS surface area measure-ments to compute the ring volume and effective an-nualized per ring transports shown in Table 2.

The resulting 1.0 6 0.4 Sv per ring transport valuededuced from our SeaWiFS observations is strikinglysimilar to, yet completely independent of, earlier NBCring transport estimates. Necessary simplifications inthe present calculation suggest that the ubiquitous 1Sv per ring value we achieved is probably an upperbound on the effective annual transport. In particular,Richardson et al. (1994) and Fratantoni et al. (1995)noted that NBC rings observed with Lagrangian de-vices and a current meter mooring, respectively, in-dicated reduced core diameter at depth relative to atthe surface. This suggests a ring core geometry moresimilar to a truncated cone than to a cylinder, and there-fore smaller in volume. A significant contribution ofthe present study to the assessment of ring transport isour independent estimate of the number of ringsformed per year (6), a value that can only be obtainedthrough remote observation and/or extended in situmonitoring. Additional detailed study of NBC ring wa-ter mass and velocity structure is required to refine thiscoarse estimate of volume transport.

5 For this calculation we ignore the apparent reversal in oxygenanomaly in the 100-m-thick surface mixed layer. This feature is mostlikely due to nonconservative upper-ocean sources and sinks of ox-ygen rather than lateral mixing between the ring core and its envi-ronment.


FIG. 7. A composite SeaWiFS image (centered on 8 Dec 1998) of the NBC retroflection aug-mented by in situ absolute velocity measurements at depth 15 m. Note the correspondence betweenhigh gradients in ocean color and local maxima in the NBC jet.

e. Ring evolution: Case studies

In this section we present case studies detailing theevolution of three NBC rings observed with SeaWiFS.The three rings (F, G, and K) are typical of the largerpopulation we observed. These particular rings werechosen for detailed study primarily because they wereextensively surveyed in situ as part of a regional fieldprogram (Fratantoni et al. 1999) and seeded with sat-ellite-tracked surface drifters (Glickson et al. 2000). Amore detailed account of the demise of NBC rings uponcollision with the Lesser Antilles, utilizing both surfacedrifters and vertical arrays of acoustically tracked sub-surface RAFOS floats, will appear in a forthcoming ar-ticle. Note that the Glickson et al. (2000) drifter datareport and a combined SeaWiFS-drifter animation loopare publicly available online at

(http://science.whoi.edu/users/dfratantoni).

1) RINGS F AND G

A time series of SeaWiFS images depicting the gen-eration and evolution of Rings F and G is shown in Fig.

9. Just prior to the separation of Ring F in early January1999, the NBC retroflection was in an extended statereaching almost to 98N (Fig. 10a). An indentation or‘‘neck’’ had formed where the first meander of thesoutheastward NECC approached the northwestwardNBC. On approximately 1 January the retroflection re-tracted sharply (see Fig. 5) leaving behind a pinched-off ring centered over the northeast corner of the De-merara Rise. Ring F moved hesitantly toward the north-west for about two weeks before its translation speedstabilized near 12 cm s21 (Fig. 10d). This ring was oneof the smallest identified during the study, with a surfacearea approximately one-half the average value observedat 558W. While approximately round shortly after for-mation (aspect ratio near 1.0 on 10 January) Ring Fgradually became more eccentric with time and rotatedsuch that its major axis was oriented parallel to thecontinental slope. This alongshore stretching is partic-ularly evident in the trajectories of surface drifterslaunched in the ring in February 1999 (Fig. 9). Thedrifters in the core of Ring F abruptly stopped theirlooping behavior in early March as the center of cir-culation neared Trinidad and the ring became extremely


FIG. 8. Vertical profiles of temperature, salinity, and dissolved oxygen anomaly computed withinthe core relative to a profile on the ring’s periphery. Three profiles are shown, one each for ringsobserved near 578W in 1998 (solid), 1999 (dotted), and 2000 (dashed). The oxygen anomalyprofiles suggest that the strongest water mass contrast between the core of an NBC ring and itssurrounding environment occurs above a depth of 600 m.

elongated. Just prior to the apparent collapse of its vor-tical circulation, Ring F extended as far north as Mar-tinique (138N) with an overall length of almost 600 km.Though we find no evidence that Ring F penetrated theLesser Antilles as a coherent vortex, the distribution ofits fractured remnants (identified by surface drifters)indicate that some fraction of the ring’s core volumeentered the southeastern Caribbean through a series ofshallow (50 m) passages in the Grenadines.

Ring G was generated immediately following ring Fon approximately 10 February 1999. The retroflectiondid not extend as far north prior to the formation of thisring (Fig. 5) nor was any significant ‘‘necking’’ of theretroflection noted (Fig. 11). Nevertheless, in contrastto Ring F, Ring G is one of the largest rings observedduring this study with a surface area almost 50% largerthan average. This size discrepancy suggests the pos-sibility that the extreme elongation of Ring F near theend of its life cycle may have been due in part to in-teraction with its neighbor. One common feature of vor-tex–vortex interactions, studied extensively in the lab-oratory and in numerical simulations, is that of a large,strong vortex entraining filaments of fluid from a neigh-boring, smaller, weaker vortex (e.g., Provenzale 1999).The close proximity of Rings F and G (not atypical ofother observed NBC rings) suggests that such interac-tion may be a common and important component ofdownstream ring evolution. Such ring–ring interaction

could result in variability in the location at which ringcore water is released to the surrounding environment.As seen in Fig. 9, the trajectories of drifters intentionallylaunched near the ring center indicate a closed circu-lation considerably smaller than the radius of maximumvelocity. Over time the radius at which the drifters loopbecomes larger but remains within the assumed radiusof maximum velocity, which we associate with the low-chlorophyll ring core. By late February Rings F and Gare separated by only a narrow (150 km) band of high-chlorophyll water that, we assume, is experiencing con-siderable horizontal shear resulting from the opposingazimuthal velocities of the two rings. The demise ofRing G is somewhat different from its predecessor withvery few of the core drifters entering the Caribbean.Instead, the ring maintains its vortical circulation whilemoving northward parallel to and east of the LesserAntilles. Most of the drifters in the ring core accelerateas they pass through a deep but narrow channel betweenthe islands of Barbados and St. Lucia.

While no drifter in Ring G demonstrated a completeloop around Barbados (but see Ring K, below) themovement of the low-chlorophyll ring core stronglysuggests that it passed directly over the island, com-pletely engulfing it (see Fig. 9, 11–19 April). How isthis possible? At a depth of 1000 m (i.e., somewhatbelow the vertical center of the ring’s azimuthal cir-culation) the island of Barbados appears to the ring as


FIG. 9. Time series of composite SeaWiFS images depicting the evolution of Rings F and G. Color palette is identical to that used in Fig.2. Seven-day segments of satellite-tracked surface drifter trajectories corresponding to the SeaWiFS image are also shown. The length ofeach line segment reflects the speed of the drifter. The most recent drifter position (i.e., the head of each 7-day ‘‘worm’’) is depicted by asmall circle.


FIG. 10. Summary diagram depicting the generation and evolution of Ring F. (a) Simplified representation of the color front bounding theNBC retroflection and Ring F during various stages of its life cycle. Dashed contour corresponds to the last observed configuration of theretroflection prior to identification of a separated ring. (b) Movement of the ring center during the period of observation. Circles appearevery 4 days. Approximated positions (required due to cloud cover) are denoted by open circles. (c–e) Time series of latitude (solid),longitude (dashed), translation speed, and aspect ratio (ratio of the major and minor axes of an elliptically shaped ring).

an obstacle approximately 40 km wide and 100 km long.With a core diameter at least five times the width ofthis obstacle (and a total extent perhaps twice that size)the ring is able to maintain a (possibly distorted) vorticalcirculation. This does not appear to be the case whenan NBC ring meets the considerably larger ridge en-compassing the Lesser Antilles, as in the case of RingF. A similar problem, that of Mediterranean Water eddies(meddies) colliding with an isolated cylindrical sea-mount, has been studied in the laboratory by Cenedese(2002) for both advected and self-propagating vortices.Cenedese found that for ratios of seamount diameter tovortex diameter of less than 0.2 (the regime correspond-ing to NBC Ring–Barbados interaction) the vortexmoved past the topographic obstacle with minimal dis-turbance.

2) RING K

The evolution of Ring K is depicted in a time seriesof SeaWiFS images shown in Fig. 12. This ring wasformed in January 2000 in a manner quite similar tothat of Ring F, accompanied by a clear extension, ‘‘neck-ing,’’ and subsequent retraction of the retroflection(Figs. 5 and 13a). Ring K accelerated rapidly towardthe west. The mean translation speed over its first monthof existence was 21 cm s21, making it one of the fastestobserved during this study. This rapid translation wasaccompanied by significant deformation of its alreadyelliptical shape, with an aspect ratio gradually decreas-ing from 0.8 to near 0.5. By early March the major axisof the ring had assumed a position parallel to the con-tinental slope east of Tobago.


FIG. 11. As in Fig. 10 but for Ring G.

Near the middle of March the center of ring K passeddirectly over Barbados. Unlike Ring G, we have detailedverification of this event as a surface drifter in the ringcore completes a closed loop around the island and thencontinues looping as the ring moves northward (see Fig.10: 20 March). This anticyclonic circulation persisteduntil the ring center reached about 178N. From this pointthe ring gradually decayed without further translation(not shown; see Glickson et al. 2000). None of the drift-ers launched in Ring K entered the Caribbean, althoughthree drifters grounded while heading in that direction,one each on St. Lucia, Martinique, and Guadeloupe.

4. Discussion

We have demonstrated that it is possible, albeit dif-ficult, to extract useful information about NBC ring gen-eration and evolution from remotely sensed ocean colordata. One particular source of difficulty stems from arecurrent July–October minimum in surface chlorophyll

contrast and the resulting pattern of successive ‘‘missingrings’’ (Q1–Q4) apparent in Fig. 3. In addition to hor-izontal advection by wind and the NBC, surface chlo-rophyll concentration on the equatorial continental shelfis assumed to be related to the Amazon dischargethrough an unknown and probably nonlinear biologicalresponse to changing nutrient input as well as inherentseasonality in phytoplankton productivity. It is reason-able to hypothesize that the annually varying Amazondischarge (Oltman 1968; Figueiredo et al. 1991), theNBC transport (Johns et al. 1998), and the local windstress (Lentz 1995a,b) conspire to reduce the amount of‘‘tagged’’ chlorophyll-rich water on the outer continen-tal shelf that is entrained into the NBC. This in turnleads to a reduced ability to visually distinguish the NBCretroflection and any pinched-off rings in ocean colorimagery.

As shown in Fig. 3, events Q1–Q4 occur within afew months of the climatological annual minimum inAmazon River discharge, shortly after the NBC trans-


FIG. 12. As in Fig. 9 but for Ring K.

port begins to decline from its annual maximum, andduring a period of relatively light southwestward winds.Lentz (1995b) examined seasonal variations in the con-figuration of the Amazon plume using climatologicalhydrographic data and found that, on average, the Am-azon discharge was contained much closer to shore(200–300 km) during July–December than during Jan-uary–June (400–500 km). This difference in offshoreextent was attributed to a combination of high Amazondischarge and the more-frequent occurrence of south-westward winds during the first half of the year. Thisis consistent with synoptic observations of the Amazonplume by Lentz and Limeburner (1995) that clearly re-veal a minimum in the offshore extent of low-salinitywater during November 1991 coincident with minimumAmazon discharge. We surmise that the large (50%)reduction in local wind stress and reduced Amazon dis-

charge during boreal summer and fall (Fig. 3) couldsignificantly reduce the amount of Amazon-influencedwater on the outer continental shelf that is available tobecome entrained into the NBC. Further observation,analysis, and/or numerical simulation is required to fur-ther explore the relationship between the various forcingmechanisms described above and their relative impacton surface chlorophyll distributions in the western equa-torial Atlantic.

5. Summary and conclusions

We have presented unique observations of North Bra-zil Current ring generation and evolution using SeaWiFSocean color imagery and satellite-tracked surface drift-ers. These observations provide insight into the com-plete life cycle of an NBC ring from its genesis in the


FIG. 13. As in Fig. 10 but for Ring K.

low-latitude Atlantic to its demise near the Lesser An-tilles. In combination with recent in situ observationsour relatively small collection of remotely observedrings enabled a new and independent estimate of thecontribution of NBC rings to intergyre volume transportin the western tropical Atlantic. The main conclusionsof this investigation can be summarized as follows:

1) Ocean color imagery of the western tropical Atlanticreveals the dynamically relevant spatial and temporalscales associated with NBC ring generation and evo-lution. Strong horizontal gradients in surface chlo-rophyll can be used to infer the position of the NBCretroflection and the radius of maximum velocity ofNBC rings.

2) The effectiveness of SeaWiFS imagery for visualNBC ring identification appears to have a seasonallimitation associated with the annual cycle of Am-azon River discharge and/or wind-forced cross-shelftransport. However, detectable mesoscale variability

in chlorophyll concentration persists along the NBCring translation corridor during all seasons of theyear.

3) Based on SeaWiFS observations and in situ tracermeasurements, a reasonable upper bound on the ef-fective annualized transport of a single NBC ring isapproximately 1.0 6 0.4 Sv. This value is both con-sistent with and independent from previous esti-mates.

4) Approximately Six NBC rings are formed each year.Successive rings evolve in close proximity to oneother. Our observations suggest that neighboringrings may exert substantial influence on each other.Interaction between adjacent rings may control thetime and location at which trapped ring core wateris released to the tropical/subtropical gyre circula-tion.

5) Progressive deformation of the elliptical ring ge-ometry occurs rapidly relative to the rate of north-


westward ring translation. This implies that preciseprediction of high-velocity events at a particular lo-cation (e.g., an island, oil rig, etc.) depends onknowledge of both the ring’s translation and defor-mation histories.

The societal importance of a thorough understandingof NBC ring generation and evolution extends beyondtheir role as a link in the Atlantic MOC. For example,expanding deep water oil and gas exploration effortsoffshore of Venezuela and Trinidad depend on accuratecharacterization and prediction of regional circulationvariability in order to assure the safety and profitabilityof these endeavors. In the absence of an extensive insitu monitoring program, the ability to predict the move-ment and intensity of NBC rings on an operational basisdepends largely on our ability to collect and interpretremote observations. Individually, satellite altimetryand ocean color measurements provide important in-formation regarding the generation and evolution ofNBC rings, and each approach has its benefits and lim-itations. Altimetric measurements are immune to cloudcover and allow direct estimation of circulation inten-sity, but at marginally useful spatial and temporal res-olution. Ocean color observations provide high-reso-lution, near-synoptic views of surface circulation pat-terns, but provide no intensity information and are sus-ceptible to cloud interference and seasonal changes indischarge-related productivity. A logical next step in thedevelopment of an operational ring tracking systemmight involve combination of the best features of thesetwo observational methods, perhaps using an assimi-lating numerical model to blend real-time surface ob-servations into a dynamically consistent three-dimen-sional picture of the tropical Atlantic circulation.

Acknowledgments. This work was supported by theNational Science Foundation through Grant OCE 97-29765. The ocean color data were provided by theSeaWiFS Project at Goddard Space Flight Center. Thedata were obtained from the Goddard Distributed ActiveArchive Center under the auspices of the National Aero-nautics and Space Administration. Use of this data isin accord with the SeaWiFS Research Data Use Termsand Conditions Agreement. Supplemental support forthe surface drifter measurements was provided by theNational Oceanic and Atmospheric Administration. Thein situ observations reported here result from the col-laborative data collection and analysis efforts of theNorth Brazil Current Rings Experiment coinvestigators(P. L. Richardson, D. M. Fratantoni, W. E. Johns, S.Garzoli, W. D. Wilson, and G. J. Goni). We acknowledgethe thoughtful comments made by P. L. Richardson andW. E. Johns on an early version of this manuscript. Wethank Semyon Grodsky and an anonymous reviewer fortheir insightful comments and suggestions for improve-ment.

REFERENCES

Borstad, G. A., 1982: The influence of the meandering Guiana Currentand Amazon River discharge on surface salinity near Barbados.J. Mar. Res., 40, 421–434.

Brown, O. B., D. B. Olson, J. W. Brown, and R. H. Evans, 1983:Satellite infrared observation of the kinematics of a warm-corering. Aust. J. Mar. Freshw. Res., 34, 535–545.

Bruce, J. G., J. L. Kerling, and W. H. Beatty, 1985: On the NorthBrazilian eddy field. Progress in Oceanography, Vol. 14, Per-gamon, 57–63.

Cenedese, C., 2002: Mesoscale vortices colliding with a seamount.J. Geophys. Res., in press.

Cowen, R. K., and L. R. Castro, 1994: Relation of coral reef fishlarval distributions to island scale circulation around Barbados,West Indies. Bull. Mar. Sci., 54, 228–244.

Cushman-Roisin, N., E. P. Chassignet, and B. Tang, 1990: Westwardmotion of mesoscale eddies. J. Phys. Oceanogr., 20, 758–768.

DeRuijter, W. P. M., H. Ridderinkhof, J. R. E. Lutjeharms, M. W.Schouten, and C. Veth, 2002: Observations of flow in the Moz-ambique Channel. Geophys. Res. Lett., in press.

Didden, N., and F. Schott, 1993: Eddies in the North Brazil Currentretroflection region observed by GEOSAT altimetry. J. Geophys.Res., 98, 20 121–20 131.

Figueiredo, A. G., M. Allison, and C. A. Nittrouer, 1991: Amazondischarge: Internal report for AMASSEDS researchers. Univer-sity Fed. Fluminense Tech. Rep., Niteroi, Brazil.

Fratantoni, D. M., W. E. Johns, and T. L. Townsend, 1995: Rings ofthe North Brazil Current: Their structure and behavior inferredfrom observations and a numerical simulation. J. Geophys. Res.,100, 10 633–10 654.

——, P. L. Richardson, W. E. Johns, C. I. Fleurant, R. H. Smith, S.L. Garzoli, W. D. Wilson, and G. J. Goni, 1999: The North BrazilCurrent Rings Experiment (abstract). Eos, Trans. Amer. Geo-phys. Union, 80, 43.

——, W. E. Johns, T. L. Townsend, and H. E. Hurlburt, 2000: Low-latitude circulation and mass transport pathways in a model ofthe tropical Atlantic Ocean. J. Phys. Oceanogr., 30, 1944–1966.

Geyer, W. R., and R. C. Beardsley, 1995: Introduction to specialsection on physical oceanography of the Amazon shelf. J. Geo-phys. Res., 100, 2281–2282.

Glickson, D. A., D. M. Fratantoni, C. M. Wooding, and P. L. Rich-ardson, 2000: North Brazil Current Rings Experiment: Surfacedrifter data report November 1998–June 2000. Woods HoleOceanographic Institution Tech. Rep. WHOI-2000-10, 129 pp.

Goni, G. J., and W. E. Johns, 2001: A census of North Brazil CurrentRings observed from T/P altimetry: 1992–1998. Geophys. Res.Lett., 28, 1–4.

Hellerman, S., and M. Rosenstein, 1983: Normal monthly wind stressover the World Ocean with error estimates. J. Phys. Oceanogr.,13, 1093–1104.

Hooker, S. B., W. E. Esaias, G. C. Feldman, W. W. Gregg, and C. R.McClain, 1992: An Overview of SeaWiFS and Ocean Color.NASA Tech. Memo. 104566, 24 pp.

Jacob, J. P., 1997: Influence of topography on the propagation ofisolated eddies. Masters thesis, Rosenstiel School of Marine andAtmospheric Science, University of Miami, 116 pp.

Johns, W. E., T. N. Lee, F. A. Schott, R. J. Zantopp, and R. H. Evans,1990: The North Brazil Current retroflection: Seasonal structureand eddy variability. J. Geophys. Res., 95, 22 103–22 120.

——, ——, R. C. Beardsley, J. Candela, R. Limeburner, and B. Cas-tro, 1998: Annual cycle and variability of the North Brazil Cur-rent. J. Phys. Oceangr., 28, 103–128.

Kelly, P. S., K. M. M. Lwiza, R. K. Cowen, and G. J. Goni, 2000:Low-salinity pools at Barbados, West Indies: Their origin, fre-quency, and variability. J. Geophys. Res., 105, 19 699–19 708.

Lentz, S. J., 1995a: The Amazon River Plume during AMASSEDS:Subtidal current variability and the importance of wind forcing.J. Geophys. Res., 100, 2377–2390.

——, 1995b: Seasonal variations in the horizontal structure of the


Amazon Plume inferred from historical hydrographic data. J.Geophys. Res., 100, 2391–2400.

——, and R. Limeburner, 1995: The Amazon River Plume duringAMASSEDS: Spatial characteristics and salinity variability. J.Geophys. Res., 100, 2344–2375.

Limeburner, R., R. C. Beardsley, I. D. Soares, S. J. Lentz, and J.Candela, 1995: Lagrangian flow observations of the AmazonRiver discharge into the North Atlantic. J. Geophys. Res., 100,2401–2416.

Lutjeharms, J. R. E., 1996: The exchange of water between the SouthIndian and South Atlantic Oceans. The South Atlantic: Presentand Past Circulation, G. Wefer, W. H. Berger, G. Siedler, andD. Webb, Eds., Springer-Verlag, 122–162.

McClain, C. R., M. L. Cleave, G. C. Feldman, W. W. Gregg, S. B.Hooker, and N. Kuring, 1998: Science quality SeaWiFS data forglobal biosphere research. Sea Technol., 39, 10–16.

Molinari, R. L., and E. Johns, 1994: Upper layer temperature structureof the western tropical Atlantic. J. Geophys. Res., 99, 18 225–18 233.

Muller-Karger, F. E., C. R. McClain, and P. L. Richardson, 1988: Thedispersal of the Amazon’s water. Nature, 333, 56–59.

Oltman, R. E., 1968: Reconnaissance investigations of the dischargeand water quality of the Amazon River. U.S. Geological SurveyCircular 552, 16 pp.

Pauluhn, A., and Y. Chao, 1999: Tracking eddies in the subtropicalnorth-western Atlantic Ocean. Phys. Chem. Earth, 24A, 415–421.

Provenzale, A., 1999: Transport by coherent barotropic vortices.Annu. Rev. Fluid. Mech., 31, 55–93.

Richardson, P. L., G. E. Hufford, R. Limeburner, and W. S. Brown,1994: North Brazil Current retroflection eddies. J. Geophys. Res.,99, 5081–5093.

Schmitz, W. J., Jr., 1995: On the interbasin-scale thermohaline cir-culation. Rev. Geophys., 33, 151–173.

——, and M. S. McCartney, 1993: On the North Atlantic circulation.Rev. Geophys., 31, 29–49.

Servain, J., and S. Lukas, 1990: Climatic Atlas of the Tropical AtlanticWind Stress and Sea Surface Temperature 1985–1989, IFRE-MER, 133 pp.

Wilson, W. D., S. L. Garzoli, G. J. Goni, W. E. Johns, R. H. Smith,C. I. Fleurant, P. L. Richardson, and D. M. Fratantoni, 1999:The North Brazil Current retroflection: Two recent surveys. Eos,Trans. Amer. Geophys. Union, 80, p. 43.

North Brazil Current Ring Generation and Evolution ...asl.whoi.edu/publications/documents/papers/seawifs paper.pdfNorth Brazil Current Ring Generation and Evolution Observed with ...

Documents