The Cromwell Current on the East Side of the Galapagos Islands

VOL. 78, NO. 33 3OURNAL OF GEOPHYSICAL RESEARCH NOVEMBER 20, 1973

TheCromwell Current on the East Side of the Galapagos Islands

H•soN• P•: •NI• J. R. V. ZANEVELD

School o] Oceanography,Oregon State University, Corvallis, Oregon 97331

Observationsmade during October and December1971 on the Yaloc 71 Cruise of Oregon State University indicate the presence of the Cromwell Current on the east side of the Galapagos Islands. Light scattering, particle size dist,ribution,nutrients, and standard hydrographic parameters were measured in water samples collected at !54 stations. The distribution of water properties shows that. the extension of the current to the east side of the islands derives primarily, from water flowing around the north side of the islands. This flow consistsof two branchesthat t6nd to merge into one at about 84øW. There was no clear indication of a branch around the south side of the islands within the area of the observations.

What happens to the Cromwell Current as Islands [Knauss, 1966; Stevenson and Ta]t, it passes the Galapagos Islands has been a 1971]. White [1969] describedthe circulation topic of great interest ever since the discovery on •he basis of a network of stations extending of the Cromwell Current itself. Montgomery to the east of 91ø00'W; however, there were [1962], in a review of the Cromwell Current, few stations close to the islands. concludedthat the Cromwell Current is usually In the present paper, the Cromwell Current weak or absent on the east side of the Galapa- circulationpattern will be describedir• the vicin- gos Islands. Wyrtki [1966] said that the Crom- ity of the GalapagosIslands on the basisof the well Current disintegrateswest of the Galapa- observationsmade on RV Yaquina during the gos Islands and it splits into north and south Yaloc 71 cruise. branches,with a part of the water extending to the upwellingregion off Peru. YALOC 71 CavISg Knauss [1966], in his report on the Swan The Yaloc 71 cruise, from October 16, 1971, Song Expedition, the specificpurpose of which to December7, 1971, coveredthe area between was •o study what happens to the Cromwell 3øN and 3øS and between 93øW and 83ø50'W Current as it moves eastward and beyond the and included observationsat 152 hydrographic Gala]•agosIslands, showedweak eastwardflow stationsshown in Figure 1. Temperature,salin- at 87øW centered north of the equator. Using ity, oxygen,light scattering,particle size distri- Alaminoscruise data, Whit• [1969] demon- .bution,and concentrationsof four kinds of strated that the southern branch of the Crom- nutrients (silicate, phosphate,nitrate, and well Current was present at 84øW at latitudes nitrite) weremeasured at 13 pointsin the upper from 2øS to 4øS on isanosteric surfaces of 120, 600 meters of water at each station. Nine para- 160, 200, and 280 cl/ton. The undercurrentwas chutedrogues set 100 metersdeep were traced inferred from the distribution of acceleratiQn either by radar rangingfrom the ship'sposition potential on the isanostericsurface, and was fixedby a satellitenavigation set or by main- also accompaniedby high values of oxygen. tainingthe ship'sposition close to the drogueso Stevensonand Ta•t [1971] measured an east- that the ship'spositions determined by satellite ward velocity of 37 cm/sec at the maximum navigationcould be usedas the drogueposi- salinity level at 84øW and near the equatorby tions. The water samples,obtained by hydro- parachutedrogue tracked relative to another graphiccasts with plastic NIO bottles,were parachutedrogue 310 metersdeep. analyzedfor temperature,salinity, and oxygen. The Cromwell Current has been observed at In addition,light scatteringwas measured with a few points on the east side of the Galapagos a Brice-Phoenix light-scattering photometer Copyright'¸ 1973 by the American GeophysicalUnion. [Briceet al., 1950;Spilhaus, 1965; Pal•, 1970].

7845 7846 PAl{ AND ZANEVELD' CROMWELL CURRENT

86' 84' 92' 90' 88' 4 ß i •

075 o,o.o o % o%/ 2/

077 ø A44 -- 0•0• 0• • 0•o • A45 07• 0•o 0• 0• 0•o• • O• • ".2 A46

o 0• 0•o A $$

2 ß • 091 0•0• 0•4 Oz• 0•• %• A 092 033 0 93 03002902e 027 01 A32

I I I I 94' 92' 90 ø 88 ø 86 ø •ig. 1. ¾•]oc 71 c•u•e stations with b•thymetric contours i• kilometers.

Particle size distributions were determined with were launched. In spite of the uncertainty of • Coulter Counter equipped with a 100-pm geostrophiccalculations in the equatorialregion, aperture [Carder, 1970], and nutrients were measured current directionsare in good agree- measuredwit h an autoanalyzer. ment with dynamic topography, except at one station just south of Isabela Island (the largest I•ESULTS of the GalapagosIslands). This agreementlends Observed results are presented to describe credenceto the complicatedflow pattern shown the circulationpattern of the Cromwell Current in dynamictopography. The 100-dbsurface was in the regionof the GalapagosIslands and to chosenbecause drogues were placedat this level. the east of the islandsby (1•) dynamic topog- According to Figure 2, the general flow pat- raphy and drogueobservations, (2) distribution tern on the east side of the GalapagosIslands of hydrographicproperties, and (3) distribution showseastward zonal flow separatedby trains of suspendedparticles. of eddies approximalely 50 km in diameter. Dynamic topography at 100 db and 600 db This patternof flow has a resemblanceto a and drogue data. In order to present the Von Karmancompound vortex street like that Cromwell Current in the region of the Gala- describedby White [1973]. A detailed flow pagos Islands and to the east of the islands, pattern of the Cromwell Current is described dynamic topographyof the 100-db surfacerela- accordingto the dynamictopography, because tive to the 600 db is shownin Figure 2. Nine we find it presentsthe basicpicture of the flow pa.rachute droguesset 100 metersdeep are also pattern. All the other data are presentedand indicated in the same figure by arrows corre- described in reference to that framework. A spondingto their velocity vectors,and the tails general picture of the flow pattern was neces- of the arrows are at the stationswhere drogues sary to describesmall-scale flows (branches) PAX AND ZANEVELD' CROMWELL CURRENT 7847 in terms of several different parameters. The the southerncoast of Isabela Island to merge velocity field is not completelydescribed by into the branch C on the east side of the same the geostrophicapproximation. Other lines of island. There is another eastward flow at about evidencetend to conflictwith the picture result- 03ø0ifs starting from about 90ø0ifs that may ing from the geostrophiccalculations, and these be fed by branch S after branch S extendsto are discussed later. the south of 03ø00'S. The location of this branch On the basis of the dynamic topography of (S-1) is similar to that of the 'southern branch the 100-db surface relative to 600 db, broken of the equatorial undercurrent' reported by streamlines are drawn to indicate the flow White [1969.] We cannot, however, verify pattern without any considerationof speedof White's south branch by our data, since our the flows (Figure 3). At about 92ø0ffW, just observationsextended only to 03ø00'S. west of Isabela Island, the Cromwell Current In between the zonal streams of eastward apparentlysplits into three branches;one head- flow, we find zonal bands flowing toward the ing north (N), the secondextending east around west; one is along the equator (SEQ-1) be- the northern coast of Isabela Island (C), and tween the branchesN-1 and C, and the other the third heading south off the west coast (SEQ-2) between the branchesC and S-1. The of the same island (S). From branch N another branch SEQ-1 is much wider than the other branch (N-l) splits off and turns at about branches,and observedcurrent velocity there 02ø0fiN, 91ø3ffW. Branch C extends to the is about 10-30 cm/secto the west at 100-meter eastalong about 02ø00'S after it passesthrough depth. Observedcurrent velocity of the branch the islands. SEQ-2 is about 20 cm/sec to the west at the Branch S heads to the south on the west side same depth. of Isabela Island beyond 03ø0ifS, the southern Unfortunately, we do not have any direct limit of our observations. There is a small current measurements of the Cromwell Current branch that turns to the east by going around on the east side of the GalapagosIslands. On

4. N! •dloW, o,s 4ON o ' "sped c/sec

0---' --0

2' -i

I , , i i I , i, ,, , 4•S 4'S 92 90 88 86 84øW Fig. 2. Dynamic topography in dynamic centimeters relative to 600 db and velocity of current determined by parachute droguesset out 100meters deep. 7848 PAK AND ZANEVELD: CROMWELLCURRENT the west side of the islands (station 84 at Islands. The core of the Cromwell Current 00ø00', 93'00'W), an eastwardvelocity of 50 could not be defined becausethere were no cm//secwas observedat 100-meterdepth, and direct measurementsof the velocity profile. distributionof temperature(Figure 4)in the However, we can qualitatively estimate the meridional section at 93ø00'W shows that the depth of the core by the shape and location of core of the Cromwell Current is below 100- the well-mixed layer indicated in the various meter depth. parametersof water properties.Such a crude Both branchesSEQ-1 and SEQ-2 turn to- approximation may not be of any value in ward the poles: the first toward the north and determining the absolute depth, but it still the second toward the south, near Isabela gives some indication of vertical variation of Island (at about 92ø00'W). It can be seen the current from one meridian to another. clearlyin Figures3 and 5 that a part of SEQ-2 The CromwellCurrent, with its high-velocity goes around the southwestcoast of Isabela core,is characterizedby intensevertical mixing. Island and then extendsto the north along the For example, above the Cromwell Current the west coastof the island. There is a droguemea- vertical distribution of temperature is char- surementat station 5-42 (southwestof Isabela acterizedby a minimum and below the Crom- Island) which supportsthe northwardflow on well Current it is characterizedby a maximum. the west side of Isabela Island. The core is located at the thermocline. Accord- Distributions of hydrographic properties. ingly, the lines of constant(rt and isothermsin Results of the Yaloc 71 cruise data, primarily the meridionalsections (Figure 4) bulgeupward the, distribution of water properties, are pre- (ridge) above the thermoclineand bend down- sented to show the extensionof the Cromwell ward (trough) below the thermocline, and the Current to the east side of the Galapagos core of the Cromwell Current coincideswith the

4ON! 92 90 8t6 86 84.øW 4ON

q.•-- •------..(-..%

2:-

4o$ 4os94øW 92I 90I 88t 86t 840WI Fig. 3. Streamlines 100 meters deep inferred from dynamic topography (Figure 2) with the axes of the Cromwell Current (heavy lines) determined by the equatorial thermostad. PAl< AND ZANEVELD'CROMWELL CURRENT 7849 _• • q.,

) i - .)

{w} HJ.d•Q

(w) HJ.d•a

(w) HJ.d3a (Ell)K,Ld•Q 785O PAK AND ZANEVELD.' CROMWELL CURRENT

4ON'4øW 92 90 Ot• 86 4øN

4øS94ow 92I 90, 88• 86• 84øW, ,4os Fig. 5. Thickness (meters) between two isanosteric surfaces; from i• --- 170 to 190 cl/ton with the axes of the Cromwell Current (heavy lines) determined by the dynamic topography. level isanostereor isotherm [Cromwell et Owing to the intensemixing within the Crom- 1954; Woosterand Jennings,1955; Montgomery well Current, the current is also identified by a and Stroup, 1962]. relatively thicker layer of water with tempera- The changesin the Cromwell Current as it ture or density closeto that of the core of the passesthe islands and extends to the east are current [Wooster and Cromwell, 1958; Knauss, shownin distributionof hydrographicproper- 1960; Montgomery and Stroup, 1962], $eitz ties in the meridional sections. Vertical distribu- [1967] proposedthe use of the word 'thermo- tion of temperature in eight meridionalsections stad' to indicate a layer of water with a mini- from 93ø00'Wto 85ø30'Ware presentedin Fig- mum vertical temperaturegradient, and White ure 4. The Cromwell Current is indicatedby [1969] showeda map of the thicknessbetween the layer of relatively well-mixedwater (maxi- isanosteric surfaces of 170 to 190 cl/ton to mum spreading of isotherms), and we can demonstrate that the equatorial thermostad identify the extent of the current, changesin correspondsto the equatorial undercurrent. depth, and width of the current as it extends The distribution of the equatorialthermostad east from the distributionof the mixed layer. determinedby the Yaloc 71 cruisedata is shown The eastward extension of the Cromwell Cur- in Figure 5. Arrows indicating the Cromwell rent is clear as far as 86ø30'W. The depth of Current (from Figure 3), as determinedby the the Cromwell Current core, estimatedby the dynamic topography,are superimposedon the depth of the most level isotherm, increases equatorialthermostad to indicatethe agreement slightly as it approachesthe islandsand then between the two sets of data. The axis of the decreasesas it extendseastward, contrary to Cromwell Current determined by connecting the report by Knauss [1966]. The width of the the equatorialthermostad is drawn in Figure 3 current is also decreased as it extends to the for the samepurpose. Locations of the branch east from the islands. N-1 and C in Figure 5 show a fair agreement, PAKAND ZANEVELD; CRO1?IWELL CURRENT 7851 but theyare not in exactagreement. For ex- strophiccalculations near the equator, the agree- ample,the thermostadis slightlysouth of mentbetween Figures 3 and 5 seemsto suggest branchN-1 and slightlynorth of branchC. that the CromwellCurrent can be approxi- Fromlocations of branches N-1 and C in Fig- matedby geostrophicflow as was reported by ures3 and 5, we cansee good agreement be- Montgomeryand $troup [1962] and Knauss tween geostrophic flow and the equatorial [1960]. thermostad.Considering the limitations of geo- The CromwellCurrent is alsoclearly indi-

4.N ! ,,.w • • • a? •'w 4.,, 4.,•-w •,2 • 0.• o.• ,•.'w •..

. . '• . • . /z.• 'ß, 'Nd;...... : ' ( ' •,••' •- ß ß • ß ...... _ : H::.•••. • •,,•' •,• I ' : ::....

4'$94,W • T•.

4'N'

Oxygen 4'S94,W 912 •0 88 86 84'W 4øS

4'NI•'W • • • I•l,•i I•,,w 4øN

2 ,,.•....", ./• •

....

2 • 2

. . . •• . Silicate 4., ,w • • •e • •.w 4.s Fi•. 6. Horizontaldistribution •t 250 me•ers depth of •emper•ture(de•rees Centigrade), s•]inity(per mi]), oxygen(mi]]i]iters per liter), ]i•h• sca•e•n• (1/m s•er),•nd silica• (microgramsper liter). 7852 :PAKAND ZANEVELD'CROMWELL CURRENT cated in Figure 6 by a maximum in tempera- B and C in the same figures. For example, ture, salinity,oxygen, an d light scatteringand branch S-1 is not supportedby oxygen:and by a minimum in silicatein the 250-meter sur- thermostad distributions, but it is indicated face. These define the Cromwell Current, at somewhatby temperature, salinity, and silicate least its lower part, very clearly on the east distributions. Since branch S-1 is on the border side'of the GalapagosIslands in two ,branches, of our observations,our confidencein it may be N-1 and C. These indications Of the Cromwell low. Current at a depth of 250 meters showa better Distribution o/s•spen.ded particles. Vertical agreementwith the equatorialthermostad than distributionsof suspendedparticles in the tropi- with the geostrophicflow. The two branches cal waters normally show a relative maximum separate just north of Isabela•!sland. The at the sea surface and another maximum, nor- separationmay occursimply from blockingof mally larger than the surfacemaximum, at the the Cromwell Current by 'the islands Isabela, thermocline.Below the thermocline,particle Pinta, and Marchena (Pinta and Marchena are concentration rapidly decreases with depth about 40 miles northeast of the northern coast to a broad minimum that extends over of Isabela). Since the w•dth of the Cromwell most of the water column. Near the bottom Current is larger than the longitudinalspan of the particle concentrationoften increases.The the blocking islands, such separation is likely two maximums• in the surface layer can be to occur. While branches N and C are indicated merged into one broad maximum and become clearly in Figures 5 and 6 and are in good indistinguishablewhen the thermoclineis shal- agreementwith the earlier interpretationsof low and mixing is intense in the surfacelayer. the Cromwell Current made on the basis of the The main source of the suspendedparticles in dyna•mic topography (Figures 2 and 3), indica- the open ocean is biologicalproduction, which tions of branch S are not as clear as branches is dependenton the penetrationof the sunlight.

,4oW 92 90

:,_

0--' 125

I I , I I I ,4'$ 4"S94"W 92 90 88 86 84øW Fig. 7. Total particle volume contained in the layer from 100 to 300 m in ppm/cm •. PAK AND ZANEVELD' CROMWELL CURRENT 7853 The high density gradient associatedwith the 3 and 5, passesby the islandswhere the par- thermoclineslows the particles settling from ticle distribution developsa pronouncedmaxi- the surfacelayer, causingthe particle maximum mum (Figure 7). It is located to the south of usually observed at this depth. The surface the particle maximum zone, on the east side of layer thus providesa sourceof suspendedpar- San ChristobalIsland (easternmostisland). We ticles. The intense vertical mixing associated cannot find any explanation for this devia.tion, with the Cromwell Current leads to an increased but we suggestthat it is associatedwith a small- downwardtransport of particles,similar to the scale phenomenonin time and space not well downward transport of other properties (e.g., defined by our observations.The increase in heat, dissolvedoxygen). The effects of this suspendedparticles is probably causedboth by downward transport are evident at 250 meters terrigenousparticles injected into the water by which is below the core depth of the current. the interaction of currents with the islands and Data on light scattering (Figure 6d), which by an increasein biogenousmaterial contributed is related to the suspendedparticles, show by the increasedsupply of nutrients resulting branch C of the Cromwell Current by bands from the upwelling processassociated with the of a slight relative maximum250 meters deep. Cromwell Current. Such upwelling is clearly The same results are observed in the total indicated in the horizontal distribution of tem- suspendedparticle volume in the layer between perature at 50-meter depth (Figure 9). 100 and 300 meters (Figure 7). Introduction of The index of refraction of suspendedparticles particlesfrom the islandsand subsequentdown- is calculated by light scattering and particle stream transport are alsoindicated by the maxi- size distribution as describedby Zaneveld and mum light scattering in a meridional section Pak [1973]. The method appliesan approxima- through Isabela Island and 90ø00'W (Figure 8). tion of a constant linear relation between the In Figure 8, we can identify the branchesN-1 total scatteringcoefficient and volume scattering by another maximum in light scattering, and function at 45ø and resultsin a singlevalue of SEQ-1 and SEQ-2 by minimum light scattering. index of refraction for a given water sample. These maximums in suspendedparticles are It is assumedthat terrigenousparticles have a probably related to the Cromwell Current higher index of refraction than biogenouspar- through both intense mixing and turbulent in- ticles. High values of index of refraction were teraction with the islands,resulting in entrain- observed in the Cromwell Current (branch C) ment of particles.Branch C, shown in Figures and in the general area surroundingthe islands

90- O0 W 91-15 W

I00 IOO •ioo !i••* ioo • 20 200 200 - 200

•oc :500 t S $ 2_ I LATITUBE LAT TUBE Fig. 8. Vertical distribution of fi(45) (1/m ster) at 91ø15'W a.nd 90øO(FW meridians. 7854 PAI< AND ZANEVELD' CROMWELL CURRENT

92 90 88 86 840W 4ON94øW I I I I 4ON

O•

I i I i i 4os 4øS94oW 92 90 88 86 840W :Fig.9. Temperature distribution at 50 meters depth.

(Figure 10). Such a distribution of index of south with respectto the kind and amount of refraction suggeststhat at least a part of the particlesthey supply to the surroundingwaters. increase in suspendedparticles in the Crom- In either case,the distribution of suspended well Current is contributedby terrigenouspar- particles showssome indicationsof the Crom- ticles, and that the GalapagosIslands are the well Current. We made an estimate of the total most likely sourceof theseparticles. mass of suspendedparticles contained in the On the other hand, high values of index of layer between 100 and 300 meters deep by refraction were not observed in the other assumingan average density of 2.0 g/cc for all branches of the Cromwell Current at 250-meter the particles counted (Figure 7). If the lowest depth (Figure 10). High values were, however, observedvalue of particle mass is assumedto observedin thosebranches at shallowerdepths: be the background value, then the mass of 50, 75, and 100 meters. Such vertical and hori- suspendedparticles introduced by the Cromwell zontal variations may be regarded as a result Current is of the order of I g/m 2 in the layer of the complex nature of spatial and temporal between 100 and 300 meters deep and 1.6 g/m • variations in the interactions between the Crom- in the layer between 50 and 250 meters deep. well Current and the GalapagosIslands. The DISCUSSION GalapagosIslands platform extendsfarther to the south of the equator than to the north. The Cromwell Current, on the west side of Thus one would expect the flow patterns of the GalapagosIslands, has been reported to the CromwellCurrent, whichis symmetricabout be located at the e%uator[Knauss, 1960, 1966]. the equator as it approachesthe islands,to be Isabela Island extendsroughly from the equator different to the north and south as the current to 1ø30'S latitude. Becauseof the asymmetric branches and flows around the islands. In addi- location of Isabela Island relative to the equa- tion, the island group may vary from north to tor, the major part of the Cromwell Current PAK AND ZANEVELD' CROMWELL CURRENT 7855

94øW 92 4*N 90I 8,8 86 8,4-W 40N

2- -2 ß L

ß H ß

O' --0

2- -2

I I I I I 40S 4øS9•PW 92 90 88 86 84øW Fig. 10. Distribution of index of refraction at 250 meters depth. appearst.o pass around the northern side of it (Figures5 and 6). A similarresult was reported by Knauss [1966]. 93-00 w Knauss [1960] reported that, based on the ! , , , , , , Dolphin Expedition, the Cromwell Current can be accountedfor by geostrophicflow. In his later report, K•auss [1966] found that the Cromwell Current was not in geostrophicbal- IOO IOO ance based on the data from the Swan Song Expedition. According to the data presented here, the geostrophicflow pattern was in goodagreement g with the flow determined by droguesand the 200 200 distribution of tracers. We found many suspi- ciousvalues in geostrophiccurrent speed,and we thus feel that thesedata may not be reliable. . The distribution of temperature at 93ø00'W,

92ø00'W, and 91ø40'W (Figure 4) and the dis- 500 500 tribution of oxygen at 93ø00'W (Figure 11) indicate that the axis of the Cromwell Current on the west side of the GalapagosIslands may •N be slightlydisplaced to the southof the equator. LATITUDE On the other hand,the fact that the Cromwell Fig. 11. Vertical distribution of oxygen at Current doesnot seemto extend directly around 93ø00'W meridian. 7856 PArc AND ZANEVELD.'CROMWELL CURRENT the southernside of Isabela Island, located near the equator, and it is quite possiblethat the 01ø00'S, seemsto contradict such displacement. locationsof the brancheseither vary seasonally In regard to this problem, we may have to or as a result of other large-scaledynamic ,con- consider the idea that the Cromwell Current is ditions. We could not detect the Cromwell causingstrong mixing to generatesuch a well- Current with six parachutedrogues set 100- mixed layer, but the mixed water can be de- metersdeep on the east side of the Galapagos flected away from the main stream when a Islands (Figure 2). Had we not made observa- proper deflectoris available.There is actually a tions over a grid of hydrographicstations as proper deflector,Isabela Island, and the Crom- well, we wouldhave reportedthat the Cromwell well Current (branch S) is deflected to the Current does not extend to the east side of the south of it accordingly.Before branch S reaches GalapagosIslands at all. the southern coast of Isabela Island, it loses Distributionsof water propertiesin the level the eastwardvelocity, and it runs into the south surfacesor isanostericsurfaces in depthsless equatorial current (branch SEQ-2), which flows than about 100 meters show a somewhat more to the west. All the observed data indicate that complicatedpicture than in the deeperwater. the south equatorial current (SEQ-2) is present Sucha complicateddistribution of water prop- as a zonal band to the south of about 01ø30'S erties is believed to be the result of intense and to the east of 91ø00'W. The northwestern mixing caused by the strong internal inter- edge of the SEQ-2 reachesthe southern coast action between the surface water and the Crom- of Isabela Island, which effectively blocks off well Current water flowing in the opposite eastward turning of the Cromwell Current direction,thus creatinga large velocityshear, (Figures 2, 3, 5, and 7). A branch of westward and the interruption of the flow by the Gala- flow just south of Isabela Island was shownin pagos Islands causing eddies of various sizes. White's [1969] data, although his picture was This intense mixing takes place in an area based on only a few observationsand he did where there is a convergenceof the surface not specificallydescribe the flow. In this case water from both hemispheres,which have dis- we can still observe a layer well mixed in tinctive water mass characteristics and often temperature and oxygen as indicated in Figure form a front. 11, even though the Cromwell Current is not The equatorial front, the boundary between displacedto the south of the equator. An am- the surface water of the south equatorial cur- biguity of this kind may be one of the weak- rent characterizedby low temperatureand high n.esses of studying a current by the distribution salinity (temperature less than 20øC and of water properties without direct measure- salinity greater than 34.0%) and the surface ment of velocity. water from the northern hemisphere(probably As the CromwellCurrent is blockedby Isa- from the equatorial counter current) charac- bela Island and by the southern equatorial terized by higher temperature and lower salin- current (the branch SEQ-2), the water carried ity, is shown in Figure 12. The front contains by the CromwellCurrent tendsto pile up on perturbationsof various sizes.These perturba- the west side of the island. The Cromwell Cur- tions can be described as intrusions of one water rent then may be affectedin a way that causes type into the other. This perturbation effect it to split into northern and southern branches will be superimposedon the effectsof the Crom- at some point farther upstream. If this is the well Current, and it may be dit•cult to resolve case,the southernbranch might extend to the the two. east through the region to the south of 03ø00•S Water below the core of the Cromwell Cur- until it reaches to about 90ø00'W, where it rent, on the other hand, is more homogeneous. shows as branch S-1. In this case branch The Cromwell Current is reflected in the dis- will conform with White's [1969] report of a tribution of water properties either as a hori- southern branch of the Cromwell Current be- zontal maximum in temperature, salinity, oxy- tween 02000 ' and 04ø00'S as far east as the gen, and light scatteringor minimum in silicate coast of Peru. with little confusion from other effects (Fig- On the east side of the GalapagosIslands, ure 6). the current was observed in two narrow chan- The interaction between the islands and the nels. Their locationsare not symmetricabout currentsmay createisland wakesof terrigenous PAX A:NDZA:NEVEL]:)' CROMWELL CURRENT 7857

40N:=-'---• l•,'W 4'N

0 "'- ø

ß ß ß

ß ß

ß ß ß ß ß

Salinity I I I I I ,4es 405940w 92 90 88 86 84'W

ß ß ß ß ß ß ß ß ß ß

ß ß -e ß ß ß ß ß _ •

ß ß

ß •' '" '-- ß ß • 0

ß ß

ß ß ß '-- 2,

Temp. 4-S94.W • 92I 90! 88I 86• 84'W' 4'S Fig. 12. Horizontaldistribution of salinity(top) and temperature(bottom) at 25 meters depth. 7858 PAK AND ZAI•'EVELD' CROMWELL CURRENT

94øW 92 90 88 86 84øW 4ON ,4ON

4os 4os 94øW

84øW 4*N94'W 92 90 8,8 86 4øN

oo 98 96 98 0o ß 2 ,- ß

ß ß ---'0

ß ß ß

ß ß -2

ß ß

S, lOOm 4os I I i ! 4os 94øW 92 90 88 8_6 840W Fig. 13. Horizontal distribution of salinity at 50 meters (top) and 100 meters (bottom) depth. PAx ANDZANEVELD: CROMWELL CURRENT 7859 particles suspendedin the water on the lee side REFERENCES of the island. We observeda high concentra- Brice, B. A., M. Halwer, and R. Speiser, Photo- tion of particulate matter and high values in electric light scattering photometer for de- light scattering in the Cromwell Current on termining high molecular weights, J. Opt. Soc. the east side of the GalapagosIslands. High Amer., •0, 768-778, 1950. valuesof index of refractionin the regionwhere Carder, K. L., Particles in the Eastern Pacific Ocean: Their distribution and effect upon op- the Cromwell Current is located support the tical parameters, Ph.D. thesis, Oregon State idea of islandwakes of particles,since the index University, Corvallis, 140 pp., 1970. of refraction of the terrigenous particles is Cromwell, T., R. B. Montgomery, and E. D. larger than that of biologicalparticles. Stroup, Equatorial under-current in Pacific While the Cromwell Current is identified by Ocean revealed by new methods, Science, 119, 648-649, 1954. a temperature and salinity maximum on..the Knauss, J. A., Measurements of the Cromwell 250-meter surface (Figure 6), the salinity dis- Current, Deep Sea Res., 6, 265-286, 1960. tributions at 50 and 100 meters (Figur• 13) do Knauss, J. A., Further measurements and obser- not seemto indicate the Cromwell Current by vations on the Cromwell Current, J. Mar. Res., 2•, 206-239, 1966. a salinity maximumas was suggestedby Steven- Montgomery, R. B., Equatorial undercurrent in son and Ta[t [1971]. Instead it is indicatedby review, J. Oceanogr.Soc. Jap., 20th Ann. Vol., a temperature minimum in the 100-meter sur- 487-498, 1962. face (not shown), which suggeststhe Cromwell Montgomery, R. B., and E. D. Stroup, Equatorial Current core is located below 100 meters. waters and currents at 150øW in July-August 1952, Oceanogr. Stud., vol. 1, 68 pp., Johns Hopkins Univ., Baltimore, Md., 1962. Pak, H., The Columbia River as a source of CONCLUSIONS marine light scattering particles, Ph.D. thesis, 1. The Cromwell Current was observed as OregonState University, Corvallis,110 pp., 1970. Seitz, R. C., Thermostad,the antonym of ther- far as 86ø30'W and was split aroundthe islands. mocline, J. Mar. Res., 25, 203, 1967. More water seemedto be flowingon the north- Spilhaus,A. F., Observationsof light scattering ern side of the islands in two branches (N-1 in seawater,Ph.D. thesis, MassachusettsInsti- and C), and a southern branch flowed to the tute of Technology, Cambridge, 242 pp., 1965. Stevenson,M. R., and B. A. Taft, New evidence east after a considerable detour to the south of the equatorial undercurrent east of the (probably beyond3ø00'S). GalapagosIslands, J. Mar. Res., 29(2), 103- 2. The Cromwell Current deepenedslightly 115, 1971. as it approachedthe GalapagosIslands, and it White, W. B., The equatorialundercurrent, the south equatorialcountercurrent, and their ex- roseto a shallowerdepth on the east sideof the tensions in the South Pacific Ocean east of the islands. GalapagosIslands during February-March, 3. The flow pattern determinedby dynamic 1967,Re•. 69-•-T, 74 pp., Texas A&M Univ., topographyat 100 db relative to 600 db is in College Station, 1969. good agreementwith that determinedby the White, W. B., An oceanicwake in the equatorial undercurrent downstream from the Galapagos distribution of the distance between 170 and Archipelago,J. Phys.Oceanogr., 3, 156-161,1973. 190 cl/ton isanostericsurfaces, with the move- Wooster, W. S., and T. Cromwell, An ocean- ment of nine droguesset 100 meters deep, and ographic description of the eastern tropical with the distribution of other properties (light pacific,Bull. ScrippsInst. Oceanogr.Univ. Calq., scattering, temperature, salinity, oxygen, and 7 (3), 169-282, 1958. Wooster, W. S., and F. Jennings, Exploratory silicate). oceanographic observations in the eastern 4. Island wakesof suspendedparticles were tropical Pacific,January to March 1953,Calif. observedin the vicinity of the Cromwell Cur- Fish Game, 41, 79-90, 1955. rent on the east side of the GalapagosIslands. Wyrtki, K., Oceanographyof the eastern equatorial Pacific,Oceanogr. Mar. Biol. Annu. Rev., J, 33-68, 1966. Acknowledgment. The nutrients and chemical Zaneveld, J. R. V., and H. Pak, A method for data were analyzed by Mr. J. Anderson and Mr. determination of index of refraction of particles R. Olund, University of Washington personnel suspendedin the ocean,J. Opt. Soc. Amer., in supervisedby J. J. Anderson. press, 1973. This research was supported by the Office of Naval Research through contract N000 14-67-A- (Received January 23, 1973; 0369-0007 under project NR 083-102. revised July 12, 1973.)