Assessment of New Production at the Upwelling Center at Point

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 102, NO. C4, PAGES 8573-8585, APRIL 15, 1997

Assessmentof new production at the upwelling center at Point Conception, California, using nitrate estimated from remotely sensed sea surface temperature Richard C. Dugdale• Department of BiologicalSciences, University of SouthernCalifornia, Los Angeles Jet PropulsionLaboratory, California Institute of Technology,Pasadena

Curtiss O. Davis 2 Jet PropulsionLaboratory, California Institute of Technology,Pasadena

Frances P. Wilkerson • Department of BiologicalSciences, University of SouthernCalifornia, Los Angeles

Abstract. Remotely sensedsea surfacetemperature (SST) and a model originally developedfor Cap Blanc, northwestAfrica [Dugdaleet al., 1989], are used to estimatenew production(i.e., nitrate uptake, in the senseof Dugdaleand Goering[1967]) for the persistentcoastal upwelling feature at Point Conception,California. Parametersrequired to initialize the model and remotelysensed SST (from advancedvery high resolution radiometerimages) were availablefor spring1983, from data collectedas part of the Organizationof PersistentUpwelling Structures(OPUS) study.Some examplesof the spatial extent of new productionare illustratedusing false color images,and temporal variability is shownby the time seriesof depth- and area-integratednitrate uptake obtained from eight images.The model resultsare comparedwith shipboarddata for three different upwellingconditions that occurred during OPUS-83, along with the model resultsand data publishedfor Cap Blanc. These two regions,Point Conception and Cap Blanc, representtwo ends of a spectrumof coastalupwelling performance, with low new productionat Point Conceptionand lesseffective conversion of availablenitrate into particulatenitrogen biomass in contrastto the high levels of new productionat Cap Blanc. The daily new productionat the Point Conceptionupwelling center is about 10% of the Cap Blanc new production,both estimatedfrom the remote-sensingmodel and satellite- derived SSTs. The model is shown to work well for both extremes and should therefore be suitablefor intermediate situations.The long-term objectiveis to produce a model which can be used for coastalupwelling systems globally to provide a estimate of new production from remotely senseddata in theseimportant areas and to assistin understandingthe role of thesecoastal systems in the air-sea exchangeof biogeochemicalelements.

1. Introduction due to downwardparticle flux [Eppleyand Peterson,1979] or into biomassyield [Dugdaleand Goering,1967]. Rates of new The major reservoir of nitrogen in the ocean is nitrate in production are also important to the global CO2 cycle since subthermoclinewater which is incorporatedinto phytoplank- they set the rate at which CO2 upwelledwith nitrate is taken up ton throughprimary productionprocesses when advectedinto and incorporatedinto biologicalparticles. the upper euphoticzone. The fraction of primary production Coastalregions are sitesof relatively intensenew production resultingfrom this or other new nitrogen sourcesrather from compared to open ocean systems.[Dugdale and Wilkerson, i989] and are consideredby someto be of equal importanceto and Goering,1967]. New productionis a key componentin the the open seain global new productionprocesses and the flux of global oceanic nitrogen and carbon budgets since nitrogen biogenic elements to the deep ocean [Jahnke et al., 1990; must be incorporatedfor carbon fixation (photosynthesis)to Walsh,1991]. Although many more measurementsof new pro- occur, and the input of nitrate largely determines the maxi- ductionwith the lSN techniquehave been made in nearshore mum export of nitrogen and carbon from the phytoplankton regionscompared to open ocean systems,the annual new production of coastal systemsremains poorly known, in part a •Nowat RombergTiburon Centers, San Francisco State University, result of the high spatial and temporal variability characteristic Tiburon, California. of suchareas, particularly easternboundary current upwelling 2Nowat NavalResearch Laboratory, Washington, D.C. regions [Woosterand Reid, 1963]. Copyright 1997 by the American GeophysicalUnion. Eastern boundary coasts typically exhibit a series of up- Paper number 96JC02136. welling plumes,e.g., alongthe coastof SouthAfrica [Taunton- 0148-0227/97/96J C-02136509.00 Clark, 1985] and the coast of South America [Guillen and

8573 8574 DUGDALE ET AL.: NEW PRODUCTION AT POINT CONCEPTION FROM SST

34o45 '

1000 rn

34o30 ' S77 100rn "•"'"•"•

STUDY AREA

34ø15'

I 121o0 ' 120o30 ' 120o0 ' Figure 1. Map of the OPUS-83study site showing G-1 (the stationnearest the upwellingcenter) and drifter S77 deployedat G-1 during period 1.

Calienes,1981] and along the California coastwhere offshore dissolvedorganic matter originating from land sourcesare filamentsand jets also occur [Brink and Cowles,1991; Abbott more important then phytoplanktonpigment in determining and Barksdale,1991]. Production cyclesin these coastal sys- the total absorptionof incidentradiation. Also, in this studywe temshave temporal scales that are relatedto the frequencyof usemodel and shipboarddata combinedto obtain a best esti- local winds as well as to remotely forced seasonalchanges in mate of new productionover the area of the upwellingplume. the depth of the thermocline.As a result, coastalsystems are New productionfor Point Conceptionis estimatedfor each of strongly undersampled due to practical limitations of re- eight images available over a period of 34 days. The area- sources,especially ship time. Correctionof the undersampling integratednew productionis then further integratedover the problem in coastalupwelling systemsis only feasibleusing time encompassedby the satellitedata to providea total area remotesensing [e.g., Platt and Sathyendranath,1988; Platt et al., and approximatelymonthlong total plume new production. 1989],in conjunctionwith data from mooredinstruments and shipboardobservations. Models for evaluationof new production from remotelysensed ocean color data are underwayin a 2. Study Site: Point Conception, California numberof laboratories[e.g., Pribble et al., 1994;Campbell and The upwellingcenter at Point Conceptionhas the general Aarup, 1992]. Remotely sensedtemperatures from the ad- appearanceand characteristicsof other easternboundary up- vancedvery high resolutionradiometer (AVHRR) sensorcan wellingcenters (15øS, Peru [MacIsaacet al., 1985]; Punta San be usedto constructthe surfacenitrate field [Motin et al., 1993] Hipolito, Baja California [Walshet al., 1974]; the Benguela and to determineproductivity as proposedby Traganzaet al. upwellingin SouthAfrica [e.g.,Shillington et al., 1990;Probyn, [1983] for the upwellingplume off Point Sur, California.Dug- 1992]), with cold, nutrient-rich,low-chlorophyll water being dale et al. [1989] and Sathyendranathet al. [1991] have used upwelledclose to the coastand movingoffshore as a surface AVHRR-SST in conjunctionwith coastalzone color scanner plume with increasingchlorophyll concentration and temper- (CZCS) data to estimatenew productionat Cap Blanc,north- ature and decreasingnutrient concentrations[Jones et al., west Africa, and GeorgesBank, respectively. 1988]. The upwellingplume has its origin betweenPoints Ar- Our present study uses the Dugdaleet al. [1989] remote- guelloand Conception[Jones et al., 1983]as the shorelineturns sensingmodel to estimatenew productionfor the upwelling abruptlyfrom a north-southorientation to east-west,becoming center at Point Conception,California. It differsfrom the pre- the north shore of the Santa Barbara Channel (Figure 1). vious applicationin that the input parameterswere obtained Details of physicalprocesses in the Point Conception up- duringthe sametime period asthe satelliteimages and a series wellingcenter have been describedby Barth and Brink [1987], of AVHRR imageswere used to get a range of values over Brink et al. [1984], Davis and Regier [1984], and Brink and different upwelling and nonupwellingconditions, whereas a Muench [1986]. This upwellingsystem was intensivelyinvesti- singleimage was used in the previouscase study and shipboard gated during the OPUS-83 field studyfrom April 4 to May 11, data from a different year were used to initialize the model. 1983 [Atkinsonet al., 1986],with ships,moorings, instrumented For the Point Conceptionstudy we use the temperature-only aircraft,and satelliteAVHRR imagery(images obtained, Ta- mode (no CZCS) of the model primarilyas a resultof unre- ble 1). Atkinsonet al. [1986] identifiedthree upwellingevents solvedproblems in identifyingthe case2 watersin whichchlo- and two downwellingevents from the wind stressdata obtained rophyll concentrationsare overestimated,especially in the duringOPUS-83, whichwere then organizedinto three periods nearshoreregions. In thesecase 2 waters,inorganic particles or (Table 1) by Dugdaleand Wilkerson[1989], characterizedby DUGDALE ET AL.: NEW PRODUCTION AT POINT CONCEPTION FROM SST 8575

Table 1. UpwellingState and PlumeAppearance at Point ConceptionDuring the AcquisitionPeriod of AVHRR ImagesUsed in This Study

Julian Day 1983 Date Period UpwellingState Image Features

104 April 14 1 moderateupwelling classicalcold water plume 107 April 17 2 early relaxation shrinkingplume 114 April 24 2 relaxed plume absent 115 April 25 2 earlyweak upwelling coolwater advectedaway 121 May 1 2 major downwelling plume absent 126 May 6 3 major upwelling developingplume 129 May 9 3 major upwelling plume and eddy to north 138 May 18 3 major upwelling large plumesat Point Conception and to the north

moderateupwelling, relaxation/downwelling, and strongup- Observationsof increasingVNO3 in shipboardenclosure ex- welling,respectively. During the first period (April 2-16 1983, perimentswere used to developa quantitativemodel of phy- Table 1), there was an upwellingevent on April 2-4, just toplanktonresponse to upwellingby Zimmermanet al. [1987a], beforeshipboard data were collected,and a secondupwelling who showedan accelerationterm for nitrate uptake to be eventfrom April 12 to 14.Then followeda calmperiod (period necessaryto reproduce the upwelling productivityat Point 2, April 17 to May 1) with relaxation and two downwelling Conceptionsuccessfully and to be a function of initial NO 3 events(April 17-20, April 27-30) and, finally,during period 3 concentration.The accelerationmodel was further developed (May 2-19, 1983), a prolongedupwelling event from May 2 to to predictproductivity changes along a simulateddrifter track 18 with sustainedstrong upwelling favorable winds. from the site of upwellingby incorporatingdeceleration as the The shipboardobservations of the spatialand temporaldis- nutrientswere usedup and depleted[Dugdale et al., 1990].This tributionsof nutrient and biologicalvariables during OPUS-83 model of phytoplanktonphysiological adaptation was intro- were describedby Dugdaleand Wilkerson[1989]. Besides time duced into a remote-sensingmodel to predict NO 3 uptake seriesstations at the upwellingcenter and mappingtransect (new production)in a casestudy using a singlesatellite image data, the biological observationsincluded Lagrangian style from the upwelling region of Cap Blanc, northwest Africa studiesfollowing near-surface (1-m depth) driftersplaced at [Dugdaleet al., 1989]. It is usedhere to estimatenew produc- the upwellingcenter, location G-1 (Figure1). Large shipboard tion at Point Conception,California, from AVHRR-derived enclosures(380-1 Nalgene barrels) were filled with surfacewa- SSTs. ter at G-1 as the drifterswere launched,and biologicalactivity The output (Figure 2) of the remote-sensingmodel (de- in the enclosureswas comparedwith that occurringin the scribedby Dugdaleet al. [1989]) is an estimateof the biomass- vicinity of the drifters [Wilkersonand Dugdale,1987]. During specificnitrate uptake rate FNO3 for each pixel which is then upwelling events, both along drifters and within the barrels, multipliedby an estimateof the particulatenitrogen biomass phytoplanktonactivity increasedwith elapsedtime from the PON to obtain the absolutenitrate uptake rate: upwellingcenter, whereas when drifterswere deployedduring downwelling,phytoplankton processes along the drifter track pNO3 = I/NO3 X PON (1) and in barrels filled with the samewater showeddecreasing Absolute NO 3 uptake rate oNO 3 with units of mg-atoms activitywith time as the low levelsof initially availablenutri- NO3-N m-3 h-I is the measureof new productionof the entswere depleted [Wilkersonand Dugdale, 1987]. system,and use of it eliminates any errors associatedwith detrital nitrogenin the system[Dugdale and Wilkerson,1986]. For each pixel the maximum rate of biomass-specificnitrate 3. New Production Model uptake(VmaxNO3(t)) is computedfrom a knowninitial rate of These changesin phytoplanktonprocesses were also ob- nitrateuptake, VNO3(,), acceleration of nitrateuptake.4 with servedat the upwellingcenter at 15øS,Peru, and form the basis timewith units of t-2, andthe elapsed time (t) sinceupwelling of a physiologicalmodel [Dugdaleet al., 1990] and the core of for the pixel (computed from the differencein SST between the remote-sensingmodel [Dugdaleet al., 1989] for new pro- the pixel and the SST at the upwellingcenter during active ductionin coastalupwelling areas. The increaseand decrease upwelling): in specificnitrate uptake (VNO3) activity with time can be describedwith a bell-shapedcurve extendingfor 8-10 days VmaxNO3(t):VNO3(t) -[- .4t (2) [Dugdalee! al., 1990]. High specificnitrate uptake rates and The pixel SST is used to estimate the nitrate concentration accumulatedbiomass occur at the peak followed by a steep from a regressionof surfacenitrate versusSST for the area and declinein VNO 3 and loss of biomass,as upwellednutrients consequentlydelineates the outer plume boundary beyond becomedepleted. The rate of increaseof VNO3 with time, like whichno significantnew productionoccurs. The pixelvalue for observationsmade in higherplants and bacteria(reviewed by NO3 isthen used to limitNO 3 uptake (VmaxNO3(t)) according Dugdaleand Wilkerson[1992]), has been confirmedin simu- to the Michaelis-Menten equation and calculatesthe realized lated upwellingexperiments in the laboratorywith the diatom nitrateuptake (VNO3(t)): Skeletonemacostaturn, using molecular methods [Smith et al., 1992],and been describedby othersin other upwellingstudies VNO3(t)= VmaxNO3(t)XNO3/(Ks + NO3) (3) [e.g.,Dortch and Postel,1989; Probyn, 1992; Yang, 1992; Kudela where Ks is the half-saturationconstant, with units mg-atoms and Dugdale, 1996; Gabtic et al., 1993, 1996; Vezina, 1994]. NO3-N m-3 for nitrate uptake.This allowsfor the effectof 8576 DUGDALE ET AL.' NEW PRODUCTION AT POINT CONCEPTION FROM SST

than 10 rn (from the hydrographicstations occupied by the R/V I AVHRRi VeleroIV during OPUS-83) gave the followinglinear regres- sion:

ITEMPERATUREI. NO3 = 88.56- (6.53 x T) r = 0.86, n = 20 The equation is nearly identical to that given in Table 1 of Time Dugdaleet al. [1989],for Point Conception,that usedthe larger Temp . Temp data set collectedaboard the R/V New Horizon during OPUS- 83, which includedall data from the upper 30 rn (B. H. Jones personalcommunication, 1988): NO3 = 89.54- (6.74 x T) r = 0.87, n = 268

This larger data set was obtained over an area encompassing essentiallythe entire image. No NO3 was ever detected at temperaturesabove 13.3øC.We have chosento use the shal- lower data set as it more closelyrepresents the surfacecondi- tions.The initial temperaturefor the upwellingsite usedwas VNO3 10.9øC,the measuredvalue at the upwellingcenter (G-l, Fig- VNO3•E ure 1) duringdeployment of a drifter (S77) duringperiod 1 E NO3 (April 5, 1983), and is a typicalvalue for the upwellingcenter I VN03(t) during activeupwelling [Atkinson et al., 1986]. The rate of changeof surfacetemperature with time during drift awayfrom the upwellingcenter was obtained from surface temperaturechanges observed while following the same sur- zone NO3 Z • IZel facedrifter (S77). The valueobtained, 0.49øC d -• [Dugdaleet 0 al., 1989,Table 2], is similarto that obtainedfor the upwelling systemsat 15øS,Peru (0.5øC d-I), andCap Blanc,northwest Africa(0.56øC d- 1). VNO3 4.2. Acceleration and Ks for Nitrate Uptake zoneeuph •PNO 3 In the applicationof this model to Cap Blanc [Dugdaleet al., oJ 1989]these parameters were obtainedby plottingVNO3 (from the upper three light depths)against surface temperature as a NEW measureof time sinceupwelling. The upwardlinear part of the PRODUCTION curveenveloping these points was fitted by calculatinga value Figure 2. The analyticalsteps taken by the model to estimate ofA (0.001h -2) thatagreed with the apparentinitial slope of new productionfrom remotely sensedtemperature. the pointsat coldertemperatures (i.e., smallvalues of elapsed time) and a Ks of 3 mg-atomsm -3 to obtainthe declining slope.The procedureamounted to placing an envelopeover decliningnitrate concentration.The verticalirradiance field is the data points,and its use was equivalentto the assumption computed from the characteristicdiffuse attenuationcoeffi- that all pointsin the imagewere the resultof a singleupwelling cient(Kd) for the areawith unitsm -1, obtainedfrom the eventwith the initial conditionsused to constructthe envelope depthto which1% of the surfaceirradiance penetrates. This is curve. However, the measureddata points showedconsider- then used to compute the vertical pattern of nitrate uptake able scatterunder the curve.This scattercan be simulatedby from another Michaelis-Mentenequation relating specificni- consideringthe temperature,NO3, and VNO3 fieldsto be the trate uptake (VNO3) to irradianceE: result of multiple upwellingevents with differentinitial condi- VNO3(tz)= VNO3(t)x E/(K• + E) (4) tions and different subsequenthistories of temperature,NO3, and VNO3 propagatedoffshore. The theoreticaleffects of vary- where KE is the light intensityrequired for half the maximal ing initial NO3 concentrationson the time courseof VNO3 are uptakerate. The pixelvalue of VNOz(tz)is multipliedby the shownin Figure 3a. If all the pixels in the upwellingplume mean value of PON for the area to obtain the absolute nitrate portionof an imagewere the resultof one upwellingevent with uptake rate pNO3, and the verticallyintegrated value of pNO3 some initial NO3, one of these curves should be followed. is calculatedfor the pixel. Finally, the valuesof pNO3 for all However,each upwelling event with differentinitial NO3 con- pixelswith NO3 concentrationsgreater than zero are summed centrationswill have a differentinitial temperature;that is, the to givethe plume-widenew productionrate. The development origin of lower NO3 curveswill be displacedto highertemper- of the model parametersfor Point Conceptionis givenbelow. atures,as shownin Figure 3b. Consequently,any plot of shipboard data including data encompassingmultiple upwelling eventswill certainlyshow suchscatter, and the use of an en- 4. Model Input Parameters velopewill generallyresult in an overestimateof NO3 uptake. 4.1. Temperature-Based Inputs In this applicationof the model to Point Conceptionwe have Temperature is used to estimatenitrate concentrationand obtained these parametersusing a more objectiveapproach, elapsedtime sinceupwelling. Shipboard data from depthsless even though when the shipboardVNO3 data for period 1 DUGDALE ET AL.' NEW PRODUCTION AT POINT CONCEPTION FROM SST 8577

0.07- 0.07- a 3o b o• 25 0.06- 0.06-

0.05- 0.05-

0.04- o'• 0.04- o z 0.03- :> 0.03-

0.02- 0.02-

0.01 -

o ..... I 0 o•j 1•0 2103•0 4•0 50 60 70 0.01-8 Time, h Temperature, oC Figure 3. Plots of modeled VNO3 (a) versustime and (b) versustemperature to simulate data in a temperature-VNO3field resultingfrom multiple upwelling eventsof varying strengths.Different lines or symbolsrepresent different initial nitrate concentrations at start of upwelling(5-30 mg-atomsm-3).

(when there was moderateupwelling) were plotted against to the mean shipboardvalues than the envelopecurve. We surfacetemperature (Figure 4), theywere encompassedby an usedthe calculatedvalue of A, 0.00064h -2, and the statisti- upper envelopecomputed from Cap Blancparameters (A = callyfitted value of K s = 4 mg-atomsm -3 in the modelruns. 0.001 h-2 andK s = 3 mg-atomsm-3). A wascalculated accordingto 4.3. Obtaining Values of KE and Ka ,4: 4 x 10-sx [NOD]+ 4 x 10-5 (5) Using periods 1 and 3 data from OPUS-83, the mean values of VNO3 at each light penetrationdepth (LPD) for the up- [fromZimmerman et al., 1987a,equation (2)] as 0.00064h -2 wellingcenter G-1 (lessthan 50 m deep) and for a numberof usingan initialNO 3 concentration(15 mg-atomsm -3) at G-1 other stationswith a bottom depth greater than 50 m were at deploymentof drifter S77.The valueof K s for NO 3 uptake, plotted againstthe LPD at which they were sampledand in- requiredto reducethe calculated VNO3(,) (equation (3)), was cubated.Mean (_+ standarddeviation) values for KE (half- obtainedby leastsquares, i.e., by obtainingthe sumof squared saturation constantfor the Michaelis-Menten expressionfor differencesbetween the valuesof measuredVNOs and values nitrate uptakeas a functionof incidentirradiance) and appar- of VNOs(,)calculated from (2) and(3), withthe value ofA = ent V ..... (maximaluptake assumingsaturation kinetics) were 13 130064 h 2 (frnm nhnve• fnr the rnnoe nf K_ vnllleq frnm 1 cnlc•lnted (Tnhle ?) using n Mn.ncl m•rve-fittingnrnornm to 6 mg-atomsm -s NOs.The minimumsum of squareswas [Zimmermanet aJ.,1987b]. A meanvalue for KE of 12.1% LPD obtainedwith a Kx of 4 mg-atomsm -s NOs,giving a standard was used in the model simulation. deviationof 0.013h -l (n - 30). The VNOsversus temper- The measureddepths of the euphoticzone (defined as the ature curve using these parametersand the observedrate of depth to which 1% of the surfaceirradiance penetrated) for changeof temperaturewith time, 0.49øCd •, is shownin periods 1 and 3 of OPUS-83 were used to calculateK a, the Figure4, and its usewill clearlyprovide a closerapproximation mean diffuse attenuation coefficient for mean downwelling

0.08

0.06

"• 0.04 o z >

0.02

10.5 11 I 1.5 12 12.5 13 13.5 Temperature, oC

Figure4. Shipboardsurface values (upper three LPDs) of biomass-specificnitrate uptake, VNO3 (h-I), versustemperature (øC) at Point Conceptionfor period 1. The upper envelopewas fitted usingacceleration A of 0.001h -2 andK s of 3 mg-atomsN m-s. Thelower line drawn through the data was calculated usingA of 0.0064h 2 anda Ks of 4 mg-atomsN m-3 obtainedusing a leastsquares fit. 8578 DUGDALE ET AL.: NEW PRODUCTION AT POINT CONCEPTION FROM SST

Table 2. Nitrate Uptake VersusIrradiance Parameters (Mean + StandardDeviation) Obtained During OPUS-83 at Point Conception

Period 1 Period 3

KE V .... rE Vmax, Region %LPD h-• %LPD h-•

G-1 stations 17.3 _+ 2.1 0.037 _+ 0.001 11.8 +_ 6.1 0.006 +__0.001 (<50-m depth) Offshore 12.0 _+ 4.0 0.040 _+ 0.004 7.2 _+ 7.7 0.007 _+ 0.002 ( > 50-m depth)

PAR (meanphotosynthetically available radiation valid for the 5. Results whole euphoticlayer), accordingto BeersLaw: 5.1. Spatial Extent of New Production E(Z) = E(O)e-s:•z (6) at Point Conception Plates 1 and 2 showthe sea surfacetemperatures for 256 x Mean valuesof K a were calculatedfor stationssampled during 256 pixel AVHRR images(282 x 282 km at nadir) for the periods1 and 3, separatedinto inshore(<50-m depth) and OPUS-83area, with surfacenitrate concentrations and uptake offshore(>50-m depth) stationsand all stationscombined (pNO3) computedfrom the temperatureimages. The features (Table3). Sincethe values of •a for thetwo different depth capturedby theseimages show the same eventsdescribed by groupswere not significantlydifferent, the overallmean of 0.3 Atkinsonet al. [1986], an upwellingevent (Julian Day 104), m- • wasused in subsequentcalculations. followedby a declinein total nitrate uptake as a major relaxation and downwellingevent occurred(Julian Day 107, 121), 4.4. Vertical Integration of VNO3 and Calculation of pNO 3 punctuatedby a short upwellingpulse (Julian Day 115), and the "spin-up"of new productionas winds and upwellingre- Equation (19) from Dugdaleet al. [1989]was usedto inte- commencedat the beginningof May (Julian Days 126-138). grate VNO3 verticallysince the equationfor VNO3 versusE The large area with warmer temperaturesand, consequently, (i.e., percentLPD) for Point Conceptionfits Michaelis Menten no nitrate or new productionis evidenton the westernside of kinetics(Table 2): all the images.Pixels with temperatureslow enoughfor new productionestimates are closeto the coastor in the upwelling plume. AN=PON Vm•xf e-KZ(KE +e-•:z) -•dZ (7) The AVHRR sea surfacetemperature field (Plate la) for JulianDay 104 (April 14, 1983, period 1) compareswell with The relationshipof pNO3 versusVNO3 is linear for Point the seasurface temperature map obtainedby aircraft[Atkinson Conception(Figure 5), indicatinga constantvalue of PON, as et al., 1986] at the sametime, and the nitrate field (Plate lb) was also observedfor Cap Blanc [Dugdaleet al., 1989]. The compareswell with the shipboardnitrate map [Joneset al., slopeof 2.12_+ 0.10 mg-atoms m -3 (n = 115) obtainedfrom 1988]. Surfacemaps of nitrate uptake were not made during Figure 5 was used in computingintegrated pNO 3 (i.e., new the OPUS-83 studyas sampling was focused on linear transects production)from valuesof integratedVNO3 from AVHRR images(equation (1)).

4.5. Summary of Input Parameters 0.25 The model input parameters developedabove for Point Conception,California, and usedto estimatenitrate and new D D productionfrom AVHRR SST,from shipboarddata, and from 0.20 combinedAVHRR and shipboarddata are summarizedin

Table 4. D *"E0.15 D D D E D D D D D D D Table 3. Mean Diffuse Attenuation Coefficient for PAR to C•• 0.10 D D z D the 1% Light Level During OPUS-83 at Point Conception D D D D D Period 1 Period 3 0.05 - D D •g D DDD D D D D D Region Mean __+s.d. n Mean _+s.d. n

Inshore 0.377 _+ 0.177 7 0.278 + 0.597 6 0.00 I I I 0.00 0.02 0.04 0.06 0.08 0.10 (Z < 50 m) Offshore 0.308 _+ 0.137 6 0.232 _+ 0.080 13 VN03,h'1 (Z > 50 m) All stations 0.345 _+ 0.241 13 0.246 +_ 0.078 19 Figure 5. Volume-specificnitrate uptake, pNO3 (mg-atoms N m-3 h-•), versusbiomass specific nitrate uptake, VNO3 Lightlevel is given in m-•. Valuesare calculated from the depths of (h-•), for the upperthree light depths (100, 50, 30%LPD) at 1% surfacelight penetration. Point Conception. DUGDALE ET AL.: NEW PRODUCTION AT POINT CONCEPTION FROM SST 8579

Table 4. Input ParametersUsed in the Model

Parameter Value Used Source

Temperature-Nitrate Slope,mg-atoms N m-3 øC-• -6.53 <10-m data, R/V Velero, OPUS-83 X intercept, øC 13.56 <10-m data, R/V Velero, OPUS-83 Y intercept,mg-atoms N m-3 88.56 <10-m data, R/V Velero, OPUS-83 Heatingrate, d øC-• 2.04 drifter S77, period 1, Dugdaleet al. [1989] Initial temperature,øC !0.9 G-l, Start of drifter S77, period 1, Dugdaleand Wilkerson[1989]

Other Parameters VNO3(,), h -• 0.0068 G-l, start of drifter S77, period 1, Dugdaleand Wilkerson[1989] Acceleration,h -2 x 104 6.4 G-l, start of drifter S77, period 1, calculatedfrom NO3(/) : 15 mg-atomsN m -3 Dugdaleand Wilkerson[1989] Ks, mg-atomsN m-3 4.0 Fitted by least squaresto period 1, OPUS-83 data (Figure 3) K_E,% surface 12.1 Table 2, periods 1 and 3 mean for G-1 and >50 m Ka, m-1 0.3 Table 3, periods 1 and 3 mean for G-1 and >50 m pNO3-VNO 3 slope 2.12 Figure4, linearreg. r 2 = 0.8, OPUS-83

(C lines), a time seriesat the upwellingcenter, and along pNO3 are 1.98and 5.02 mg-atoms m -2 d-•, respec.tively.An drifter trajectories[Dugdale and Wilkerson,1989]. However, alternative estimate of the mean depth-integratedpNO 3 was the patternof computednitrate uptakeshown in Plate lc for madeby averaging the shipboard •SNO 3 uptakemeasurements day 104,i.e., low uptakeat the upwellingcenter increasing with from the upper three light depths to obtain a mean surface distanceaway from the center, is consistentwith the drifter VNO3 for each period. The mean surfaceVNO3 valueswere results [Wilkersonand Dugdale, 1987; Dugdale et al., 1990; then depthintegrated using (7) and the requiredinput param- Dugdaleand Wilkerson,1992]. etersfrom Table 4. The relativelygood agreement between the The imagefor JulianDay 107 (Platesld-lf) illustrateswell shipboardmean VNO3 plusmodel with depth-integratedship- the relaxationor downwellingphase of OPUS-83 duringpe- board data for periods1 and 2 (Table 6) supportsthe vertical riod 2 as a result of slackenedwinds. Although the area of integrationprocedure used. For period 3 theseshipboard plus waterwith high nitrate concentrationshas decreased (Plate le) model valuescompare more favorablywith the shipboarddata comparedto the previousimage (Plate lb), a relativelybroad than the elevated value given by the satellite data. Possible band of water showine strone nitrate uptake aDDears near the explanationsfor the period 3 discrepancyare provided in the cnaqt nreq•mahlv the req•lt of qhoreward movement of water already well adapted physiologicallyto the available nitrate The daily total areal uptake for each AVHRR image was (i.e., in an advancedstate of shift-upfor nitrate uptake). The averagedto obtaina simplemean of 1.53 _ 0.97 x 10"kg N full impact of relaxation and downwellingcan be seen for d-• (Table5). Two additionalvalues of totalareal uptake for Julian Days 114 and 121 (Plates lg-li and 2a-2c) where tem- each image (Figure 6) were alsocalculated using mean depth- perature,nitrate, and nitrate uptakeplumes are absent.Some integratednew productionvalues obtained from shipboardand nitrate uptake appearsin Plate 11 as a result of a short-lived combinedshipboard plus model data (Table 6). Mean values pulse of weak upwelling (Plate lj). The May 2-18 upwelling from Table 6 for the appropriatetime period were multiplied event of period 3 is shown developingon Julian Day 126 by the area with NO3 > 0 (from Table 5). The valuesobtained (Plates2d-2f) with the appearanceof an offshoreeddy show- usingthe three methods(AVHRR, shipboard,and combined) ing nitrate at the surfaceand accompanyingnew productionon agree well except for period 3 when the AVHRR values give JulianDays 129 and 138 (Plates2g-2i and 2j-21). greater values than the shipboardor shipboardplus model

5.2. Time Series of New Production values (Figure 6). These area-integratednitrate uptake rates showfirst a decline(Julian Days 104, 107, 114) at the end of an Table 5 shows the calculated mean values of depth- upwellingevent followed by an increase(Julian Day 115) and integratedpNO 3 (i.e., mean of all pixelswith new production) decrease(Julian Day 121) reflectinga weak upwellingevent and total pNO3 (sumof all pixelswith new production)for the followed by downwelling(Figure 6). Finally, new production AVHRR images.Mean NO 3 uptake was computedonly for increasesas strongupwelling sets in (Julian Day 138), follow- those pixels with surface NO 3 > 0, and consequentlythe ing the end of the OPUS-83 field program. spreadin valuesis quite small, from 4.75 to 5.54mg-atoms m -2 Another approachto obtaining an overall new production d-1. Thevalue of meanpNO 3 fromall images(5.06 _+ 0.29 value for the plume and the studyperiod of 34 dayswas made mg-atomsm -2 d-1, n = 8) compareswell with the mean by integratingthe area under the curveof areal uptake versus integratedpNO3 from shipboard lsNO 3 uptakemeasurements time (Figure 6). The total new production(depth-, area-, and (4.03 +_ 6.2 mg-atomsm -2 d-1 n = 60) made during time-integrated)values are estimated to be 26.88-32.18x 106 OPUS-83 [Dugdaleand Wilkerson,1989] (Table 6). However, kg N after rejecting the temperature-only estimate. The when the resultsare divided into the three OPUS-83 periods, weighted daily mean for the study period was from 0.79 to a discrepancyappears between the shipboardand satellite- 0.95 x 10' kg N d-1 (Table7), lowerthan the simplemean, derived measurementsfor period 3 when the mean valuesof obtained from the AVHRR values from the eight images 8580 DUGDALE ET AL.' NEW PRODUCTION AT POINT CONCEPTION FROM SST

Table 5. New ProductionFrom AVHRR SST at Point Conception

No. of Pixels With Julian Day Mean pNO3, Total pNO3, New Production No. of Pixels No. of Pixels 1983 mg-atomsN m-2 d-• 106 kg N d-• NO3 > 0 Cloud/Land NO 3 = 0

104 4.98 3.01 35,649 3,548 26,339 107 5.54 1.53 16,310 34,564 14,662 114 4.83 0.49 5,892 17,623 42,008 115 5.38 2.17 23,766 6,628 35,141 121 4.75 0.72 8,932 2,943 53,659 126 4.84 0.53 6,364 883 58,267 129 4.93 1.20 14,379 847 50,253 138 5.24 2.57 28,987 3,274 33,275 5.06 _+ 0.29* 1.53 +_ 0.97*

*Mean +_ s.d.

1.53x 106 kg N d- • (Table5). A rangeof 0.79to 1.53x 106 tion averagedfrom eight AVHRR imageswas 26% higher kg N d-• maybe usedas a provisionalestimate for the mean thanshipboard lsNO 3 uptakedata acquired at the sametime daily new productionin the Point Conceptionupwelling center (Table 6). The satellite-derived values of mean depth- during the period of active upwelling, integrated pNO3 were also greater than shipboardvalues at Cap Blanc [Dugdaleet al., 1989]. Nitrate uptake for a single AVHRR imagefrom 1983 for CapBlanc compared to •sNO3 6. Discussion uptakeobtained on shipboardin 1974was 20% higher(Table Applicationof our remote-sensingmodel developedfor Cap 6). Platt andHarrison [1985] showed that undersamplingof the Blanc to the upwellingsystem at Point Conception,California, oceanicsystems will result in underestimationof production has reproducedboth qualitativeand quantitativeelements of values.The shipboardmeasurements of pNO3 made duringthe new production in the upwelling center observedduring the Point Conceptionand Cap Blanc studiesfall readily into the 1983 OPUS study.In the qualitative realm the model repro- undersamplingsituation, as the high standard deviationsin duces the spatial distributionsobserved in other upwelling Table 6 for shipboardmeasurements during OPUS-83 indi- studies;that is, in the region of maximumupwelling, there are cate. The differencesin the mean NO 3 uptakevalues between high nitrate concentrations,low temperatures,and low nitrate shipboardand satellite-deriveddata probablyalso result from uptake rates,with the opposite(decreased nitrate, increased differencesin the distributionof temperaturesampled by the temperature, and increasednitrate uptake rate) occurring two methods.The shipboarddata in OPUS-83 were more or downstreamfrom the upwellingsource. lessevenly distributed across the temperaturerange, while the The mean valuesof pNO3 obtainedfrom AVHRR and ship- satellitetemperatures were not normallydistributed. Pearson's board measurementsare comparedin Table 6 for the different first coefficientof skewness(mean-mode/standard deviation) OPUS periods that representdifferent upwellingstrengths at showedskewness toward the left, i.e., with the peak frequency Point Conception.The input parametersused for the model to the right of the mean [Spiegel,1961]. For example,for Julian were from period 1 when moderate upwellingoccurred. Con- Day 104 the Pearson'sfirst coefficient of skewnessfor the sequently,there was good agreementbetween oNO3 derived temperaturedata was -0.436, resultingin more pixelswith from AVHRR SSTs and the shipboarddata for that period. high uptake being includedin the calculationof the mean. The remote-sensingmodel agreeswell with shipboarddata for period 2, eventhough the input parametersused were not from relaxationor downwellingperiods. This is in part becauseof Table 6. Comparisonof Depth-IntegratedpNO 3 Estimated the high variabilityof the shipboarddata collectedat this time Using AVHRR Data, ShipboardData, and Combined that included some data collected from outside the OPUS-83 ShipboardData and Model studyarea (to the north of Point Arguello) and someoveres- timates that resulted from additions of saturatingconcentra- Mean Depth-IntegratedpNO3, mg- atoms m -2 d -• _+ s.d. tionsof •sNO3 to waterwith no measurablenitrate. There was also the contributionof a short but moderateupwelling pulse UpwellingCenter and AVHRR Shipboard Shipboardand (Plate lj) on JulianDay 115 that resultedin increasednitrate Period Data Data Model concentrationsand uptake. However, for period 3 the remote- Point Conception, California sensingmodel overestimatesmean nitrate uptake comparedto Spring 1983 5.06 _+0.29 4.03 _+6.20 that measuredon shipboard.The measuredrate of accelera- (n = 8) (n = 60) tionwas low during period 3, e.g.,0.0001 h -2 for drifterS239, Period 1 4.98 4.76 _+ 4.98 5.79 versus0.00038 h -2 duringperiod 1, for drifterS77, as were (JulianDays 92-106) (n = 1) (n = 16) initialnitrate concentrations (e.g., 9.9 mg-atoms m-3 for S239 Period 2 4.98 _+ 0.28 4.88 _+ 8.27 4.09 (JulianDays 107-121) (n = 4) (n = 27) versus15 mg-atomsm -3 for S77) [Dugdaleand Wilkerson, Period 3 5.02 + 0.21 1.98 + 1.60 1.24 1989]. Very likely the high windswith resultantstrong mixing (JulianDays 122-139) (n = 3) (n : 17) and reduced irradiance were the cause. However, some reduc- Cap Blanc, NW Africa tion in accelerationwould be expectedalso from the reduced northern sector 1983 18.22 nitrate concentration. 1974 15.20 The satellite estimateof new productionfor Point Concep- 'DUGDALE ET AL.: NEW PRODUCTION AT POINT CONCEPTION FROM SST 8581

JULIAN DAY TEMPERATURE(øC) [NO3](mg-at m '3) PNo3(mg'atN m-2d 4)

104

(b)

%. 107 "' aq , ." o,• , .,;• ...,•..,,

. '4' .

(d)

(g) (h) (i)

115

Plate 1. Each image is offshorefrom Point Conceptionand is 256 x 256 pixelsin size. (a) SST for day 104, (b) NO3(mg-atoms m -3) for day104, (c) pNO3(mg-atoms m -2 d-q) for day104, (d) SSTfor day107, (e) NO3 (mg-atomsm -3) for day 107,(f) pNO3(mg-atoms m -2 d-l) for day 107, (g) SST for day 114,(h) NO3 (mg-atomsm -3) for day 114,(i) pNO3 (mg-atomsm -2 d-l) for day 114,(j) SST for day 115, (k) NO3 (mg-atomsm -3) for day115, (1) pNO3 (mg-atomsm -2 d-l) for day115. 8582 DUGDALE ET AL.' NEW PRODUCTION AT POINT CONCEPTION FROM SST

JULIAN DAY TEMPERATURE(øC) [HO3] (mg-at m '3) PNoa(mg'atNm '2d'1 )

121

(a) (b) (c)

126

(d) (e) (f)

129

(g) (h) (i)

138

(j) (k) (I) Plate 2. Eachimage is offshorefrom PointConception and is 256 x 256pixels in size.(a) SSTfor day 121, (b)NO3 (mg-atoms m-3) forday 121, (c) pNO3 (mg-atoms m-2 d-x) forday 121, (d) SSTfor day 126, (e) NO3 (mg-atomsm-3) for day126, (f) pNO3(mg-atoms m-2 d-x) for day126, (g) SSTfor day129, (h) NO3 (mg-atomsm-3) for day129, (i) pNO3(mg-atoms m-2 d-•) for day129, (j) SSTfor day138, (k) NO3 (mg-atomsm-3) for day138, (1) pNO3 (mg-atoms m-2 d-x) for day138. DUGDALE ET AL.: NEW PRODUCTION AT POINT CONCEPTION FROM SST 8583

3.5

2.5

1.5

0.5

0 I I I I I I I I IO0 105 110 115 120 125 130 135 14o

Julian Day 1983 Figure6. Timeseries of total(area- and depth-integrated) 9NO3 (X 106 kg N d-•) estimatedusing AVHRR (triangles),shipboard (circles), and combinedAVHRR and shipboarddata (squares)from OPUS-83.

The mean nitrate uptake rate calculatedfor pixelswith mea- tire study area for the relaxation/downwellingperiod except surable NO 3 (i.e., greater than zero) is constrainedby the duringthe shortupwelling pulse on April 25 (Julian Day 115) mean variables that are used as initial conditions, e.g., the when the proportion of cloudsplus land pixelswas reduced to initial uptakerate, VNO3(i), nitrateconcentration, NO3(i) , 10% (Table 5). and Ka. The sensitivityof the model to parameter variability In principle,a more direct, nonphysiologicalestimate of new was testedduring the Cap Blanc analysis[Dugdale et al., 1989] productioncould be obtained for the surfaceregion by com- and showedthat the parametersrelated to the temperature- paring the temperature-derived NO 3 concentration from the nitrate regressionequation and the pNO 3 versusVNO3 regres- initial upwelled NO3 concentration.This estimate by disap- sion had a minor effect on the mean integratednitrate uptake, pearancewould, however, be an overestimatesince some ni- whereasvarying the heating rate by +_95% confidencelimits trate disappearsfrom an upwelled water massby mixing pro- gave differencesof 43%. In this studywe have examined the cessesand perhapsby other unaccountedfor effects.In drifter effectof changingA fromthe calculated value of 0.0064h-2 to experiments,for example,we have observedsuch larger disap- .4 ----0.0010 h-2, whichis the uppervalue obtained from the pearancerates for NO3 comparedto •sNO•measured uptake maximum envelope shownin Figure 4, and to A = 0.00038 rates (unpublished).Since the presentmodel giveshigher new h-2, the measuredvalue obtained using •SN during S77 [Dug- production estimatesthan shipboardvalues in some cases, daleand Wilkerson,1989]. The highervalue increasesthe mean there would be little to be gainedby usingthe nonphysiological integrated 9NO3 by 49%' the lower value reducesit by 35% temperature-NO3 model to estimate new production. (Table 8). Alternativevalues of K s (Table 8) were also tried. When remotely sensedocean color again becomesavailable WhenK s wasdecreased by 1 rag-atomsm 3,mean integrated and algorithms appropriate to coastal upwelling systemsare 9NO3 increased14%; a 10% decreaseoccurred when Ks was developed,considerable improvement in model resultsshould increasedfrom 4 to 5 mg-atomsm -3 be possible.CZCS data were not used in this study,primarily The effect of clouds obscuring the sea surface could be because the optical characteristicsof the Point Conception important in biasing the estimatesof new production by this region have not been well characterizedand becausethe case technique[Michaelsen et al., 1988]. Only during the relaxation 2 waters(i.e., waterswith high concentrationsof nonbiological and downwellingevent (JulianDays 107 and 114) were clouds particles)have not been delineatedfor the CZCS imagesavail- presentin a relatively large number of pixels (Table 5). Typi- able for the OPUS-83 study.In the model usedfor Cap Blanc, cally, they never exceeded5% of the pixels in the rest of the image.The possibilityof cloudsmasking cold water during the relaxation/downwellingperiod was eliminated in this studyby Table 8. Effect of ChangingInput Values of A and K s on examiningthe correspondingaircraft surfacemaps [Atkinsonet the Estimate of New Production Using Image for al., 1986] which showedthat temperaturesgreater than the Julian Day 104 interceptfor zero nitrate (i.e., 13.56øC)occurred over the en- A, Ks, Mean Integrated pNO3, h-2 mg-atomsm 3 mg-atomsm-2 d- • % Change*

Table 7. Depth-, Area-, and Time-Integrated pNO 3 0.00038 4 3.21 -35.5 0.00064 4 4.98 ... Obtained From the Areas Under the Curves in Figure 6 0.00100 4 7.44 +49.3 AVHRR Shipboard Shipboard and 0.00064 1 8.18 +64.3 Data Data Model 0.00064 2 6.65 +33.5 0.00064 3 5.68 +14.1 Total, 46.57 32.18 26.88 0.00064 5 4.46 -10.4 x 10•) kg N 0.00064 6 4.05 -18.7 Weighted mean, 1.37 0.95 0.79 x 10•'kgNd-• *Percent change in mean integrated 9NO3 calculated when A = 0.00064 h-2 andK s = 4 mg-atomsm 3. 8584 DUGDALE ET AL.: NEW PRODUCTION AT POINT CONCEPTION FROM SST

CZCS data were used to obtain chlorophyllvalues and the istics.Ideally, imagesfrom an annual cycleshould be used in depth of the euphotic zone. However, when the OPUS-83 thismanner to estimateannual new production.Here we offer chlorophyllversus irradiance data were plottedto obtainKd, a the first step toward the goal of estimatingthe annual new large overestimateof euphotic zone depth resulted, in part productionof a singleupwelling plume by providingareal and becauseof the large numberof G-1 (Figure 1) stationsthat are depth-integratednew productionfrom imagescollected during most likely case 2 waters. If these euphoticzone depthshad an upwellingmonth from Point Conception,California. been calculated from CZCS chlorophyll estimates,a similar overestimatein integrated new productionwould have re- Acknowledgments. This study was supported by NASA grant sulted.Instead, we usedthe temperature-onlymodel and ship- NRA-91-OSSA-07 and NSF OCE 85-05400to R. C. Dugdale and F. P. board measured1% light level depths.CZCS data were also Wilkerson and NASA grant 161-30-33 to C. O. Davis. We thank used to estimateprimary productionin the Cap Blanc study M. Hamilton (for producingthe false-colorimages) and Jet Propulsion and, with the new productionvalues to estimatef, the fraction Laboratory,California Institute of Technology,for making this study of new to total production.However, the Bricaudet al. [1987] possible. model usedto estimateprimary productivity is for case1 waters [Morel, 1988] only, so estimatesof primary production References couldnot be calculatedfrom the Point Conception,California, Abbott, M. R., and B. Barksdale,Phytoplankton pigment patterns and CZCS images.We suspectthat strongwinds and deep mixing wind forcing off central California, J. Geophys.Res., 96, 14,649- were the causesof the low accelerationand new production 14,669, 1991. rate observedin the OPUS 83 period 3. If sucha relationship Atkinson, L. P., K. H. Brink, R. E. Davis, B. H. Jones, T. Palusz- can be established,it shouldbe possibleto use scatterometer- kiewicz, and D. Stuart, Mesoscalehydrographic variability in the vicinityof PointsConception and Arguello duringApril-May 1983: derivedwinds to improvemodel estimates. The OPUS 1983 experiment,J. Geophys.Res., 91, 12,899-12,918, The robustnessof the model in predictingvalues of mean 1986. depthintegrated nitrate uptakethat agreewith shipboarddata Barth, J. A., and K. H. Brink, Shipboard acousticdoppler profiler suggeststhat reliable area-widenew productionrates can be velocityobservations near Point Conception:Spring 1983,J. Geo- phys.Res., 92, 3925-3943, 1987. obtained by combining mean AVHRR-derived values with Bricaud,A., A. Morel, and J. M. Andre, Spatial/temporalvariability of AVHRR predictedsize of area with detectablenitrate. The algal biomassand potential productivityin the Mauritanian up- major variablein total area-widenew productionis the sizeof wellingzone, as estimatedfrom CZCS data,Adv. SpaceRes., 7(2), the new production area sincethe mean pixel nitrate uptake 53-62, 1987. showsonly a smallrange, 4.75-5.54 mg-atoms N m-2 d-1 for Brink, K. H., and T. J. Cowles,The coastaltransition zone program,J. Geophys.Res., 96, 14,637-14,649,1991. the imagesused here, while the total area-widenew production Brink, K. H., and R. D. Muench,Circulation in the Point Conception- variedby a factorof 6, from0.49 to 3.01 x 106 kg N d-•. The SantaBarbara Channel region, J. Geophys.Res., 91,877-895, 1986. area with detectablenitrate is mosteasily determined from the Brink, K. H., D. W. Stuart, and J. C. VanLeer, Observations of the model and verified by shipboardsampling. coastalupwelling region near 34ø off California:Spring 1981, J. Phys. Oceanogr.,14, 378-391, 1984. This studyindicates that a rangeof 0.79-1.53x 106 kg N Campbell,J. W., and T. Aarup, New productionin the North Atlantic d- • maybe usedas a provisionalestimate for the meandaily derivedfrom seasonalpatterns of surfacechlorophyll, Deep Sea Res., areal depth-integratednew productionin the Point Concep- Part A, 39, 1689-1694, 1992. tion upwelling center during the period of active upwelling. Davis, R. E., and L. Regier, Current following drifters in OPUS-83, This is much lower than the valuescalculated using the same Ref. 84-12, 40 pp., ScrippsOceanogr. Inst., Univ. of California, San Diego, La Jolla, 1984. remote-sensingmodel for thenorthern (11.94 • 106 kgN d-•) Dortch, Q., and J. R. Postel,Biochemical indicators of N utilizationby andsouthern (2.85 x 106 kgN d-x) sectorsof the CapBlanc, phytoplanktonduring upwellingoff the Washingtoncoast, Limnol. northwestAfrica, upwellingregion [Dugdaleet al., 1989].Point Oceanogr.,34, 758-773, 1989. Conceptionhas been identifiedas an area characterizedby a Dugdale, R. C., and J. J. Goering, Uptake of new and regenerated forms of nitrogen in primary productivity,Limnol. Oceanogr.,12, low realization value (r = 0.2, i.e., does not realize the full 196-206, 1967. potential of available NO 3 in accumulatedPON biomass), Dugdale,R. C., and F. P. Wilkerson,The use of lSN to measure comparedto the northwestAfrica upwellingarea (r -- 1.2) nitrogen uptake in eutrophic oceans;experimental considerations, [Dugdaleet al., 1990]. If the Point Conceptionupwelling center Limnol. Oceanogr.,31,673-689, 1986. were to perform at full theoreticalrealization, 1.0, the daily Dugdale,R. C., and F. P. Wilkerson,New productionin the upwelling center at Point Conception,California: Temporal and spatial pat- area depth-integratednew productionrate would be 5 times terns, Deep Sea Res., Part A, 36, 985-1007, 1989. higherthan the valuesgiven above, i.e., 4.0-7.5 x 106 kg N Dugdale, R. C., and F. P. Wilkerson,Nutrient limitation of new pro- d-•, wellwithin the rangeof the twonorthwest Africa sectors. duction,in PrimaryProductivity and Biogeochemical Cycles in theSea, This analysisfocuses attention on the factors holding Point edited by P. Falkowski and A.D. Woodhead, pp. 107-122, BrookhavenNatl. Lab., Upton, N.Y., 1992. Conceptionnew productionto low levels.The primaryreason Dugdale, R. C., A. Morel, A. Bricaud, and F. P. Wilkerson, Modeling appearsto be the lack of accumulationof biomassin the Point new productionin upwellingcenters, 2, A casestudy of modeling Conception ecosystem.The ecosystemfactors causing this new production from remotely sensedtemperature and color, J. problem are unknown at presentbut are probablyrelated to Geophys.Res., 94, 18,119-18,133, 1989. the kind of grazingpressure exerted by higher levelsof the Dugdale, R. C., F. P. Wilkerson, and A. Morel, Realization of new productionin coastalupwelling areas: A meansto comparerelative food chain. performance,Limnol. Oceanogr.,35, 822-830, 1990. The two applicationsof the remote-sensingmodel have been Eppley, R. W., and B. J. Peterson,Particulate organic matter flux and to easternboundary centers at oppositeends of a continuumof new productionin the deep ocean,Nature, 282, 677-680, 1979. upwelling efficienciesand realizations[Dugdale et al., 1990], Gabric, A. J., L. Garcia, L. Van Camp, L. Nkyjaer, W. Eifler, and W. Schrimpf,Offshore export of shelfproduction in the Cape Blanc Cap Blanc, northwestAfrica, and Point Conception,Califor- (Mauritania) giantfilament as derivedfrom coastalzone colorscan- nia. Consequently,the model should be applicableover the ner imagery,J. Geophys.Res., 98, 4697-4712, 1993. wide range of upwellingcenters with intermediatecharacter- Gabric, A. J., W. Eifler, and W. Schrimpf,A lagrangianmodel of DUGDALE ET AL.: NEW PRODUCTION AT POINT CONCEPTION FROM SST 8585

phytoplanktondynamics in the northwestAfrican coastalupwelling southwestAfrica: Preliminarymeasurements of physicaland biolog- zone,Adv. SpaceRes., 18(7), 68-80, 1996. ical features,Deep Sea Res., Part A, 37, 1753-17.72,1990. Guillen, O., and R. Calienes, Upwelling off Chimbote, in Coastal Smith, G. J., R. C. Zimmerman, and R. S. Alberte, Molecular and Upwelling,Coastal Estuarine Sci., vol. 1, editedby F. A. Richards,pp. physiologicalresponses of diatoms to variable levels of irradiance 312-327, AGU, WashingtonD.C., 1981. and nitrogen availability:Growth of Skeletonemacostatum in simu- Jahnke, R. A., C. E. Reimers, and D. B. Craven, Intensification of lated upwellingconditions, Limnol. Oceanogr.,37, 989-1007, 1992. recyclingof organic matter at the seafloor near ocean margins, Spiegel,M. R., Statistics,359 pp., McGraw-Hill, New York, 1961. Nature, 348, 50-54, 1990. Taunton-Clark,K. J., The formation, growth and decayof upwelling Jones,B. H., K. H. Brink, R. C. Dugdale,D. W. Stuart, J. C. VanLeer, tonguesin responseto the mesoscalewind field during summer,in D. Blasco,and J. C. Kelley, Observationsof a persistentupwelling SouthAfrican Ocean Colour and UpwellingExperiment, edited by centeroff Point Conception,California, in CoastalUpwelling, part A, L. V. Shannon,pp. 47-61, Sea Fish. Res. Inst., Cape Town, South edited by E. Suessand J. Thiede, pp. 37-60, Plenum, New York, Africa, 1985. 1983. Traganza, E. D., V. M. Silva, D. M. Austin, W. L. Hanson, and S. H. Jones, B. H., L. P. Atkinson, D. Blasco, K. H. Brink, and S. L. Smith, Bronsink,Nutrient mapping and recurrenceof coastal upwelling The asymmetricdistribution of chlorophyllassociated with a coastal centersby satellite remote sensing:Its implicationto primary pro- upwellingcenter, Cont. ShelfRes.,8, 1155-1170, 1988. ductionand the sedimentrecord, in CoastalUpwelling: Its Sediment Kudela, R. M., and R. C. Dugdale,Estimation of new productionfrom Record, part A, Responsesof the SedimentaryRegime to Present remotely-senseddata in a coastalupwelling regime, Adv. SpaceRes., Coastal Upwelling,edited by E. Suessand J. Thiede, pp. 61-85, 18(7), 91-97, 1996. Plenum, New York, 1983. MacIsaac, J. J., R. C. Dugdale, R. T. Barber, D. Blasco, and T. T. Vezina, A. F., Mesoscalevariability in nitrogen uptake rates and the Packard,Primary productioncycle in an upwellingcenter, Deep Sea f-ratio duringa coastalphytoplankton bloom, Limnol. Oceanogr.,39, Res., Part A, 32, 503-529, 1985. 854-868, 1994. Michaelsen,J., X. Zhang, and R. C. Smith, Variability of pigment Walsh, J. J., The importance of continental margins in the marine biomassin the California Current systemas determined by satellite biogeochemicalcycling of carbon and nitrogen,Nature, 350, 53-55, imagery,2, Temporal variability,J. Geophys.Res., 93, 10,883-10,896, 1991. 1988. Walsh, J. J., J. C. Kelley, T. E. Whitledge, and J. J. MacIsaac, Spin-up Morel, A., Optical modeling of the upper ocean in relation to its of the Baja California upwelling ecosystem,Limnol. Oceanogr.,19, biogenousmatter content (case 1 waters), J. Geophys.Res., 93, 553-572, 1974. 10,749-10,768, 1988. Wilkerson, F. P., and R. C. Dugdale, The use of large shipboard Morin, P., M. V. M. Wafar, and P. Le Corre, Estimation of nitrate flux barrels and drifters to study the effects of coastal upwelling on in a tidal front from satellite-derivedtemperature data, J. Geophys. phytoplanktonnutrient dynamics,Limnol. Oceanogr.,32, 368-382, Res., 98, 4689-4695, 1993. 1987. Platt, T., and W. G. Harrison, Biogenicfluxes of carbonand oxygenin Wooster, W. S., and J. L. Reid, Eastern boundarycurrents, in The Sea, the ocean, Nature, 318, 55-58, 1985. vol. 2, pp. 253-280, Wiley-Interscience,New York, 1963. Platt, T., and S. Sathyendranath,Ocean primary production:Estima- Yang, S. R., Factors controlling new production in the equatorial tion by remote sensingat local and regional scales,Science, 241, Pacific, Ph.D. thesis,Univ. of South. Calif., Los Angeles, 1992. 1613-1620, 1988. Zimmerman, R. C., J. N. Kremer, and R. C. Dugdale, Acceleration of Platt, T., W. G. Harrison, M. R. Lewis, W. K. Li, S. Sathyendranath, nutrient uptake by phytoplanktonin a coastalupwelling ecosystem: R. E. Smith, and A. F. Vezina, Biologicalproduction of the oceans: A modeling analysis,Limnol. Oceanogr.,32, 359-367, 1987a. The case for a consensus,Mar. Biol. Ecol. Set., 52, 77-88, 1989. Zimmerman, R. C., J. Soo-Hoo, J. N. Kremer, and D. N. D'Argenio, Pribble, J. R., J. J. Walsh, D. A. Dieterie, and F. E. Muller-Karger, A Evaluation of variance approximation techniques for non-linear numericalanalysis of shipboardand coastalzone color scannertime photosynthesis-irradiancemodels, Mar. Biol. Berlin, 95, 209-215, seriesof new productionwithin Gulf Stream cycloniceddies in the 1987b. South Atlantic Bight, J. Geophys.Res., 99, 7513-7538, 1994. Probyn, T. A., The inorganicnitrogen nutrition of phytoplanktonin C. O. Davis, Naval Research Laboratory, Code 7212, Washington, the southern Benguela: New production, phytoplankton size and DC 20375. implicationsfor pelagic foodwebs,S. Aft. J. Mar. Sci., 12, 411-420, R. C. Dugdale and F. P. Wilkerson, Romberg Tiburon Centers,San 1992. FranciscoState University, P.O. Box 855, 3150 Paradise Drive, Tibu- Sathyendranath, S., T. Platt, E. P. W. Horne, W. G. Harrison, ron, CA 94920. O. Ulloa, R. Outerbridge, and N. Hoepffner, Estimation of new productionon the oceanby compoundremote sensing,Nature, 353, 129-133, 1991. Shillington, F. A., W. T. Peterson,L. Hutchings, T. A. Probyn, H. N. (ReceivedJanuary 5, 1996; revisedJune 7, 1996; Waldron, and J. J. Agenbag,A cool upwellingfilament off Namibia, acceptedJuly 2, 1996.)