JOURNAL OF JOURNAL OF GEOPHYSICAL GEOPHYSICAL RESEARCH, RESEARCH, VOL. VOL. 99, 99, NO. NO. C4, C4, PAGES PAGES 7355-7370, 7355-7370, APRIL APRIL 15, 15, 1994 1994 Comparison of of the the seasonal seasonal and and interannual interannual variability Comparison variability of phytoplankton pigment concentrations in the of phytoplanktonpigmentconcentrationsin the Peru Peru and and California California Current Current systems systems A. A. C. C. Thomas Thomas and and F. F. Huang Huang Atlantic Atlantic Centre Centre for for Remote RemoteSensing Sensingof of the the Oceans, Oceans,Halifax, Halifax, Nova Nova Scotia, Scotia,Canada Canada P. P. T. T. Strub Strub and and C. C. James James College of Oceanic Oceanic and and Atmospheric Collegeof AtmosphericSciences, Sciences,Oregon OregonState StateUniversity, University,Corvallis Corvallis Monthly composite composite images images from from the the global global coastal coastal zone zone color color scanner Abstract. Abstract. Monthly scanner (CZCS) data set are used to provide an initial illustration and comparison (CZCS) data set are used to provide an initial illustrationand comparisonof of seasonal seasonal and concentration along and interannual interannualvariability variability of of phytoplankton phytoplanktonpigment pigment concentration along the the western western coasts coastsof of South South and and North North America America in in the the Peru PeruCurrent Current system system(PCS) (PCS) and andCalifornia California Current the entire data Currentsystem system(CCS). (CCS). The The analysis analysisutilizes utilizesthe entire time time series seriesof of available availabledata (November 1978 to June 1986) to form a mean annual cycle and an index of interannual (November 1978to June 1986) to form a mean annual cycle and an index of interannual variability for aa series and cross-shelf regions within within each each current current variability for seriesof of both both latitudinal latitudinaland cross-shelfregions system. Within km of of the the coast, coast, the the strongest system. Within 100 100km strongestseasonal seasonalcycles cyclesin in the the CCS CCS are are in in two and the the second and 29°N, each two regions, regions,one onebetween between34° 34øand and45°N 45øNand secondbetween between24° 24ø and 29øN,each with maximumconcentrations concentrations(>3.0 (>3.0mg mgmm3) in May-June. MayJune. Weaker seasonal withmaximum -3) in Weaker seasonal variability California variability is is present presentnorth northof of45°N 45øNand andin inthe theSouthern Southern CaliforniaBight Bightregion region(32°N). (32øN). Within the PCS, coastal region, region, highest highest (>45øS) (>45°S) and and lowest lowest Within the PCS, in in the the same same100-km-wide 100-km-wide coastal (<20°S) latitude latitude regions regions have have aa similar seasonal cycle cycle with maximum concentrations concentrations (<20øS) similar seasonal with maximum _3) during the austral spring, summer, and fall, matching that (>1.5 mg m (> 1.5mgm-3) duringtheaustralspring, summer, andfall,matching thatevident evident throughout throughoutthe the CCS. CCS. Between Betweenthese theseregions, regions,off off northern northernand andcentral centralChile, Chile, the the seasonal maximum maximum occurs occurs during duringJuly-August JulyAugust (austral winter), contrary contrary to to the seasonal (austral winter), the influence of of upwelling upwellingfavorable favorablewinds. winds. Within Withinthe the CCS, CCS, the the dominant dominant feature feature of of influence interannual variability in the 8-year time series is a strong negative concentration interannualvariability in the 8-year time seriesis a strongnegativeconcentration anomaly in in 1983, an E1 El Nifio Niflo year. year. The The relative anomaly 1983, an relative value value of of this thisnegative negativeanomaly anomalyis is strongest central California and is is followed followed by by an an even strongestoff off central Californiaand even stronger strongernegative negativeanomaly anomalyin in 1984off offBaja BajaCalifornia. California.In Inthe the PCS, PCS, strong strong negative negative anomalies during the the 1982-1983 El 1984 anomaliesduring 1982-1983E1 Niflo period are evident only off the Peruvian coast and are evident there only in the Nifio period are evidentonly off the Peruviancoastand are evidentthere only in the regions 100 km or or more more from from the the coast. regions 100km coast.Although Althoughnegative negativeanomalies anomaliesassociated associatedwith with the El 20°S) the E1Niflo Nifio were were not notpresent presentat athigher higherlatitudes latitudes(more (morethan thanapproximately approximately 20øS)in in the the PCS, the extremely sparse sampling weakens our confidence in the results of the PCS, the extremelysparsesamplingweakensour confidencein the resultsof the interannual analysis analysis in in this this region. region. An An upper upper estimate the systematic interannual estimateof of the systematicwinter winter bias bias remaining in in the the global global CZCS CZCS data data after after reprocessing with remaining reprocessing with the the multiple multiplescattering scattering algorithm is given given in in the algorithm is the appendix. appendix. 1. 1. Coastal Transition Zone have provided CoastalTransition Zone programs programshave provideddetails detailsof of Introduction Introduction both both physical physical and andphysiological physiologicalprocesses processeson on relatively relatively space and A major oceanographic oceanographic characteristic characteristic of localized space and time time scales. scales. A major of eastern easternboundary boundary localized Previous between the currents Previouscomparisons comparisonsbetween the dynamics dynamicsof of different different currents (EBCs) (EBCs) is is the thepresence presenceofofrecurring recurringwind-driven wind-driven coastal easternboundary boundarycurrents currents[e.g., [e.g.,Smith, Smith,1981; 1981;Schulz, Schulz,1982; 1982; coastalupwelling upwelling.Physical Physicalforcing forcingintroduces introducesnutrient-rich nutrient-rich eastern subsurface water water into into the the upper in Minas et al., which relied relied on on in in situ situ data data were, subsurface upper water water column, column, resulting resulting in Minas et al., 1986] 1986] which were, by by increased increased phytoplankton phytoplankton production production and and concentrations. concentrations. necessity, necessity,restricted restrictedin in the the time time and andspace spacescales scalesthey they could could Large, and multiyear programs such such as data Large, multidisciplinary multidisciplinary and multiyear programs as address addressand and rarely rarely had had access accessto to temporally temporallycoincident coincidentdata Coastal Upwelling Upwelling Ecosystems Ecosystems Analysis Analysis (CUEA) (CUEA) and and CoopCoop- from the different synopCoastal from the differentregions. regions.The The repetitive, repetitive,large-scale large-scale synoperative Investigation Investigationof of the the Northern Northern Part Part of erative of the the Eastern Eastern tic offered by by satellite satellite observations observations are tic views views offered are well well suited suitedfor for Central Atlantic Atlantic (CINECA) (CINECA) demonstrated demonstrated the the strong tempo- estimating Central strong tempothe temporal and spatial estimatingthe temporaland spatialscales scalesof of variability variabilityof of ral and ral and spatial spatial variability variability of of both both physical physical dynamics dynamics and and biological patterns patterns in in EBCs. EBCs. More More recently, recently, programs biological programssuch such as the as the Coastal CoastalOcean OceanDynamics DynamicsExperiment Experiment(CODE), (CODE), OrgaOrganization of nization of Persistent PersistentUpwelling Upwelling Structures Structures (OPUS) (OPUS) and and Copyright 1994 by the the American Copyright 1994 by American Geophysical GeophysicalUnion. Union. Paper Paper number number93JCO2146. 93JC02146. 0148-0227/94/93JC-02146$05.O0 0148-0227/94/93JC-02146505.00 near-surface oceanographic oceanographic characteristics. characteristics. Over Over the the last near-surface last decade, satellite of ocean ocean color color from from the the coastal coastal zone decade, satelliteimages imagesof zone color scanner scanner (CZCS) have revealed revealed many the larger-scale color (CZCS) have many of of the larger-scale time and space features of biological variability time and space features of biological variability of of the the California Current system (CCS) [Abbott and Zion, California Current system (CCS) [Abbott and Zion, 1987; 1987; Pelaez Smith et et al., Pelaez and and McGowan, McGowan, 1986; 1986; Smith al., 1988] 1988] and and some some details of of structure details structure within within the the Peru Peru Current Current system system(PCS) (PCS) 7355 7355 7356 7356 THOMAS ET AL.: CURRENT PIGMENT VARIABILITY THOMAS ET AL ßPERU PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT VARIABILITY [e.g., Uribe [e.g., Uribe and and Neshyba, Neshyba, 1983; 1983; Espinoza Espinoza et et al., al., 1983]. 1983]. Substantially fewer fewer CZCS Substantially CZCS data data were were collected collectedover overthe thePCS. PCS. Sufficientimages imagesare are available, available,however, however, to to obtain obtain aa first Sufficient first approximation of the the major approximation of major seasonal seasonalfeatures features that that are are evievident dent in in the theavailable available satellite satelliteocean oceancolor colordata. data.Although Although biased biased by by missing missing data data toward toward certain certain years years and andmonths, months, they temporal they allow allow aa general generalcomparison comparisonof of the the larger-scale larger-scaletemporal and and spatial spatialvariability variability of of phytoplankton phytoplanktonbiomass biomassin in these thesetwo two Pacific eastern boundary Pacific eastern boundarycurrents. currents. In In this this paper paper we we present presentand andcompare comparethe the mean mean seasonal seasonal and and interannual interannualvariability variability of of phytoplankton phytoplanktonpigment pigmentconconcentrations centrationsalong along the the west west coasts coastsof of North North and andSouth SouthAmerAmerica Current systems. ica in in the the Peru Peru and and the the California California Current systems.The The time time series of a period series of CZCS CZCS data data includes includes1982-1983, 1982-1983, a period with with aa particularly Barber and and particularly strong strong El E1 Niflo Nifio event event [Cane, [Cane, 1983; 1983; Barber Chavez, 1986]. The analysis is offered as a general summary Chavez, 1986]. The analysisis offered as a generalsummary and synthesis and synthesisof of the the larger-scale larger-scaletemporal temporal and and spatial spatialfeafeatures tures which which can can be be deduced deducedfrom from the the available available CZCS CZCS data, data, flawed though thoughthat that data data set set may and flawed may be be by by poor poor sampling sampling and potential bias. A more complete analysis of features on these potential bias. A more completeanalysisof featureson these time and and space time spacescales scalesawaits awaits the the arrival arrival of of additional additionaland and hopefully more more complete complete data data from hopefully from future future color color sensors, sensors, such as such as the the sea-viewing sea-viewingwide-field wide-field sensor sensor(Sea (Sea WiFS), WiFS), supsupported by in situ measurements. ported by in situ measurements. Major Major features featuresof of the the temporal temporalvariability variabilityin in the the California California Current have have previously been shown shown by by Pelaez Current previously been Pelaez and and estimatedthe the accuracy accuracy of Gordon et al. Gordon et al. [1983a] [1983a] estimated of the the CZCS-derived to CZCS-derivedpigment pigmentmeasurements measurements tobe belog(chlorophyll) log(chlorophyll) ±0.5 --.0.5 (i.e., (i.e., within within aa factor factor of of 22 in in concentration). concentration).A A comparcompar- ison ison of of CZCS CZCS data data with with ship shipmeasurements measurementsoff off southern southern California [Smith et et al., al., 1988] showed the the satellite California [Smith 1988] showed satellite data data to to be be ±40% within the the chlorophyll ---40%within chlorophyll concentration concentrationrange range0.05-10.0 0.05-10.0 Although the the multiple-scattering algorithm used mg mgm m-3. Although multiple-scattering algorithm usedin in processing the global global data data set processingthe set improves improves the the concentration concentration estimates, at higher estimates, especially especially at higher latitudes latitudes during during winter winter months and Castano, Gordon et et al., months [Gordon [Gordon and Castano, 1987; 1987; Gordon al., 1988], 1988], aa systematic bias may may still still be be present. systematicbias present. This This bias, bias, as aswell well as as other and random other systematic systematic and random errors errors associated associated with with the the CZCS data set, set, is by Strub CZCS data is discussed discussedin in more more detail detail by Strub et et al. al. [1990]and andisisnot not repeated repeated here. here. An An upper upper estimate estimate of of the the [1990] magnitude of the the possible possible winter winter bias bias in in the the data data is in magnitudeof is given given in the appendix. Efforts were made in the reprocessing of the the appendix. Efforts were made in the reprocessingof the data to account data set set [Feldman [Feldman et et al., al., 1989] 1989] to account for for sensor sensor degradegra- dation over the the life of the et al., dation over life of the sensor sensor [Gordon [Gordon et al., l983b; 1983b; Mueller, Mueller, 1985]. 1985]. The The mean mean seasonal seasonal cycle cycle in in pigment pigment concentration concentration over over the extent of of the the PCS the entire entire latitudinal latitudinal extent PCS and and the the CCS CCS was was calculated. calculated. An An overall overall mean mean image image for for each each calendar calendarmonth month was formed by compositing all available images was formed by compositingall available imagesfor for aa given given month series. These These images are used month to to form form aa 12-month 12-month series. images are used to to provide provide a a qualitative qualitative description description of of spatial spatial and andtemporal temporal patterns patternsover over the the year year as aswell well as asregional regionaldifferences differenceswithin within each EBC. Subsampling each of the mean each EBC. Subsamplingeach of the meanmonthly monthly images images by within 100 100 xx 100 km boxes boxes oriented by averaging averagingwithin 100 km oriented by by the the coast a quantitative of these coast allows allows a quantitative analysis analysis of these seasonal seasonalpatpatterns regions of of missing missing data data due due to to clouds. clouds. ternsand andaids aidsin in filling filling regions The location of these boxes is shown in Figures la and lb. The location of these boxes is shown in Figures la and lb. with any any data data set, set, quantitative quantitative interpretation interpretation is to CZCS data data set set [Feldman et al., CZCS [Feldman et al., 1989], 1989], is is made made and and prepre- As As with is subject subjectto sented. the errors errors and and bias bias of of the the measurements measurements(outlined (outlined above). above). In In sented.These Thesedata datawere were processed processedutilizing utilizing aa multiple multiple scatscat- the tering algorithm in the the atmospheric the vicinity vicinity of of the the highly highly complex complex coastline coastlineoff off southern southern tering algorithm in atmosphericcorrection correction [Gordon [Gordon et et the Chile, a smoothed coastline oriented along the western al., 1988], which reduces the errors caused by large solar side al., 1988], which reduces the errors caused by large solar Chile, a smoothed coastline oriented along the western side zenith angles (high latitudes latitudes in in winter winter months). months). In In section of the the islands islands was was used. used. Despite Despite these thesespatial spatialand and temporal temporal zenith angles(high section 2 2 of we summarize characteristics of the CZCS averaging procedures, some some regions regions of of the the PCS we summarize characteristics CZCS data data and and details details averaging procedures, PCS still still had had data in in some some months. months. Missing data within within aa box of the the analysis methods. In missing data Missing data box in in of analysismethods. In section section3 3 we we present presentthe the mean mean missing seasonal variabilityinin the the two two EBCs any month month were were supplied suppliedby by spatially spatiallyaveraging averagingthe the concenconcenseasonal variability EBCs and and an an analysis analysis of of any interannual variability variability in in the the PCS PCS and and CCS over the from the the boxes from the the coast trations from boxes equidistant equidistant from coast and and interannual CCS over the 8-year 8-year trations to the the north north and extent of the CZCS CZCS data data set. set. A comparison and and discussion discussion of of immediately immediately to and south south for for the thesame samemonth. month. extent of the A comparison patterns evident in section section 4, 4, and and aa Total data supplied in this this manner manner were were less Total missing missingdata suppliedin lessthan than 5% 5% patterns evidentin in the the two two EBCs EBCs is is made made in of the total matrix of 1836 data points (12 months summary and conclusions follow in section 5. The appendix of the total matrix of 1836 data points (12 months xx 33 summaryand conclusionsfollow in section5. The appendix zones xx 51 presents an offshore zones 51alorigshore alongshoreboxes). boxes). presents an empirical empiricalestimate estimateof of magnitude magnitudeof of the thesystemsystem- offshore Previous atic errors Previousanalyses analysesof of interannual interannualvariability variability in in the theCaliforCaliforatic errors which which might might remain remain in in the the global global CZCS CZCS data data set set owing to to uncorrected at high nia Current Current using using CZCS CZCS data data [Pelaez [Pelaez and and McGowan, McGowan, 1986; 1986; owing uncorrectedatmospheric atmosphericscattering scatteringat high latilati- nia Thomas and and Strub, Strub et tudes in winter. tudes winter. Thomas Strub, 1989, 1989, 1990; 1990; Strub et al., al., 1990] 1990] have have shown that that the to resolve shown the coverage coverageis is generally generally sufficient sufficient to resolve 2. Data 2. Data and and Methods Methods features features on on time time scales scales of of a month month or less. less. The The CZCS CZCS data data set for for the the Peru Peru Current less complete. complete. To To produce produce The ocean pigment Current is is much much less The ocean pigment concentration concentrationmeasurements measurementsprepre- set aa time sented sented here here are are from from the the coastal coastal zone zone color color scanner scanner instruinstrutime series seriesallowing allowing examination examination of of interannual interannualvariability variability ment of the ment aboard aboard the the Nimbus Nimbus 7 7 satellite. satellite. The The entire entire CZCS CZCS data data of the PCS PCS over over the thelifetime lifetime of of the theCZCS, CZCS, large-scale large-scale temporal and and spatial is necessary. necessary. This set spatial averaging averaging is This averages averages set has has been been reprocessed reprocessed[Feldman [Feldman et et al., al., 1989] 1989] using using a a temporal over gaps in the data set due to clouds and/or standard epsilon value and a multiple Rayleigh scattering lack of of standard epsilon value and a multiple Rayleigh scattering over gaps in the data set due to clouds and/or lack collection effort effort but but obviously reduces the the amplitude algorithm in the the atmospheric atmospheric correction. obviously reduces amplitude of of algorithmin correction.One Oneof of the theproducts products collection McGowan [1986] and Thomas Thomas and and Strub and McGowan [1986] and Strub [1989, [1989, 1990] 1990] and compared to to wind by Strub A detailed detailed compared wind forcing forcing by $trub et et al. al. [1990]. [1990]. A description and analysis will will not not be be repeated repeated here. here. However, However, description and analysis a recalculation a recalculation of of this this variability variability in in the the CCS, CCS, warranted warranted by by the recent the recentavailability availability of of the the complete, complete,global, global,reprocessed reprocessed shorter time of is aa set time scale scale events. events. To To accomplish accomplishthis, this,8-month 8-month of this this reprocessing reprocessingis setofofglobal globalfields fieldsof ofpigment pigment shorter concentration km, composited composited concentrationwith with aa spatial spatialresolution resolutionof of 20 20 km, within calendar months. This image set was subsampled to within calendar months. This image set was subsampledto provide two time series of 92 monthly images each, extendprovide two time seriesof 92 monthly imageseach, extending ing over over the theentire entireCZCS CZCS lifetime, lifetime,from fromNovember November1978 1978to to June One series series covers June1986. 1986. One coversthe the Peru PeruCurrent Currentregion regionfrom from6° 6ø to and the the second to 51°S, 51øS,and secondcovers coversthe the California California Current Current region region from from 21° 21 ø to 48°N. 48øN. composite images images temporally temporally centered centered on on the the midpoint composite midpoint of of summer in each hemisphere were formed within each year of of summerin each hemispherewere formed within each year data coverage. data coverage. For For the the California California Current, Current, the the8-month 8-month summer composite image image in in each each year year is summer composite is of of months months March March through October. October. For For the through the Peru Peru Current, Current, the the8-month 8-month comcomposite image was formed posite image was formed from from the the months months September September through April. April. The The resultant resultant image image time time series series extends extends over over through THOMAS ET AL.' AL.: PERU CURRENT PIGMENT VARIABILITY THOMAS ET PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT VARIABILITY I [ I ] j[ ] I 7357 7357 I a 45N 45N - PACIFIC REGION ClFICNORTHWEST NORTHWEST REGION 40N 40N CENTRAL CALIFORNIA REGION - '..... NORTHERN CHILE tillZSS 4 35N - 35N 1 SOUTHERN CALIFORNIA BIGHT SOUTHERN CALIFORNIA BIGHT• _ •NTRAL C• 30N J30N - , _ 25N - BAJA CALIFORNIA REGION BAJA CALIFORN• REGION 125W 120W 115W - SOUTHERN CHILE-' 110W 125W 120W 115W 110• I I I I (a) The The California California Current Current system system and and (b) (b)the the Peru Peru Current Current system system study study areas areas showing the Figure Figure 1. 1. (a) showingthe regions within which which image image data data were were spatially averaged to to extract regionswithin spatially averaged extract seasonal seasonaland andinterannual interannualvariability. variability. The lines show show the the three three 100-km-wide stripsparallel parallelto tothe the coasts coasts extending extending offshore offshore and and the the extents extents The solid solid lines 100-km-widestrips of four regions regions used used to to illustrate illustrate interannual interannual variability. of four variability. The The dashed dashedlines lines indicate indicate the the system systemof of 100 100 xx 100 100 km boxes boxes used used to to spatially average the the image image data data for for presentation presentation of of mean mean seasonal seasonal cycles. cycles. km spatially average 8 years 8 years in in each each study study region. region. For For the the CCS, CCS, the the 1986 1986 compositeisis from from March Marchtoto June, June, and and for for the composite the PCS PCS the the 1978-1979 composite is is from from November November to 1978-1979 composite to April. April. Quantitative analysis Quantitative analysisof of interannual interannualvariability variability in in pigment pigment concentration was carried out by subsampling and concentration was carried out by subsampling and spatially spatially averaging within within these these 8-month composite images images to to derive averaging 8-month composite derive mean concentrations. concentrations. These These were were calculated calculated over over each each of mean of four latitudinal latitudinal zones four zones extending extending 400 400 km km or or more more alongshore. alongshore. An annual anomaly was formed by subtracting An annual anomaly was formed by subtracting the the overall overall mean concentration concentration (the mean (the mean mean of of the the 8-month 8-month period period over over all all 8 years) for for each each zone zone from from the the concentrations 8 years) concentrationsfound found in in each each year. The year. The sample sample zones zones are are oriented oriented by by the the coastline coastline in in the the same manner as as the same manner the smaller smaller boxes boxes used used above above for for the the seasonal seasonalcycle. cycle. To To illustrate illustrate cross-shelf cross-shelfdifferences differencesin in interinterannual variability, variability, each annual each latitudinal latitudinal zone zone was was subdivided subdividedinto into three 100-km-wide areas, and and aa separate three 100-km-wide areas, separate mean mean was was calcucalcu- lated for for each. in aa time each. This This resulted resulted in time series series for for each each lated latitudinal zone zone and latitudinal and each each of of the the three three offshore offshore bands bands (0-100, (0-100, 100-200,and and 200-300 200-300km) km)shown shownininFigures Figuresl la and lb. lb. 100-200, a and is substantially less. less. In is substantially In many many months, months, some someareas areasof of the thePCS PCS are to 33 years years of of data, data, and few areas areas are represented representedby by only only 2 2 to and a a few in Lack of in some some months months remain remain completely completely unsampled. unsampled. Lack of coverage is most coverage is most severe severe during during months months of of austral austral winter winter (June, (June, July). July). These These images images are are composed composedof of data data from from only only 1-2 years. 1-2 years. -3 Plate shows that that concentrations Plate 11 shows concentrationsgreater greaterthan than2.0 2.0 mg mgm m are are found found within within approximately approximately 40 40 km km of of the the coast coastthroughthroughout out the the year year and and throughout throughoutthe thelatitudinal latitudinalrange rangeof of the theCCS CCS study area. High concentrations (>1.0 mg m3) studyarea. concentrations (>1.0 mg m-3) extend extend farthest farthest from from shore shore off off northern northern and and central central California California (ap(approximately 340 to 43øN) 43°N)and and to to aa lesser proximately 34 ø to lesserdegree degreeoff off Baja Baja California (south of California (south of 30°N). 30øN). Higher Higher pigment pigment concentrations concentrations remain most closely associated with the coast throughout throughout the remain most closely associatedwith the coast the year and southern year in in the the California California Bight Bight and southern California California regions. regions. High concentrations(> (>1.0 mg m m3) farHigh concentrations 1.0 mg -3) appear appearto toextend extend far- thest thest from from shore shore in in November, November, December, December, and and January, January, forming diffuse pattern pattern off off northern northern and and central forming aa diffuse central California California with these values with no no strong strong zonal zonal gradients. gradients. Although Although these values from from late and early early winter winter may may continue continue to to be be contaminated contaminated by late fall fall and by incompletely removed atmospheric effects [Strub et al., incompletely removed atmospheric effects [Strub et al., 3. Results 3. Results 1990], in the the appendix appendix we we estimate estimate the 1990], in the maximum maximum magnitude magnitude 3.1. 3.1. Seasonal SeasonalCycles Cycles of m3-3south ofof40°N, of such sucherrors errorsto to be beless lessthan than0.15 0.15 mg mg m south 40øN, The multiyear to 0.4 mg m m3 risingto -3 or orless lessatat48°N. 48øN.In InFebruary Februarythese these The multiyear means means of of pigment pigment concentration concentration for for each each rising month the calendar highest concentrations concentrations are are closer closer to to shore this region, month of of the calendar year year are are shown shown in in Plates Plates 11 and and 22 for for highest shore in in this region, but but the California Current and in March March they they appear appear to the California Current and Peru Peru Current Current regions, regions, respecrespec- in to extend extendup up to to 700 700 km km offshore offshore again, again, tively. Spatial resolution in in these these composites composites is is the the 20 20 xx 20 as shown and Strub [1989]. In In April April the shown by by Thomas Thomas and $trub [1989]. the diffuse diffuse tively. Spatial resolution 20 as pattern of concentrations appears km of the the original global data data set. km pixel pixel dimensions dimensions of original global set. A A pattern of offshore offshorehigher higher concentrations appearsto toconsolconsolpresentation idate closer closer to to the the coast. coast. By the highest concentrations By May May the highest concentrations presentation of of the the number number of of years years(samples) (samples)which which make make idate up each pixel in each monthly mean is given in Figures 2 and are within approximately 200 km of the coast, up each pixel in each monthly mean is given in Figures 2 and are within approximately 200 km of the coast, and andaarelarelatively sharp sharp frontal 3. frontal gradient gradient separating separatinghigher higher pigment pigment concon3. In In general, general, pixel pixel values valueswithin within the themonthly monthly composite composite tively images of of the the CCS CCS are are from from 55 or or more more years years of centrations from from very very low concentrations is low offshore offshore concentrations is estabestabimages of data. data. As As was was centrations previously stated, stated, Figure lished. In In June June this takes on this gradient gradient takes on aa scalloped scalloped shape shape off off previously Figure 33 shows showsthat that coverage coverageof of the the PCS PCS lished. THOMAS ET AL.' AL.: PERU CURRENT PIGMENT VARIABILITY THOMAS ET PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT VARIABILITY 7358 7358 3.13 3.13 •.. [CHL] ß . - 45N - 40I " 1.04 mgm-3 1.04 0.31 0.31 ß I - 35N 35N - 0.06 - 30N 30N - - 25N 25W •IAN FEB MAR APR MAY [CHL] mgm-3 JUN I 3.13 3.13 1.04 1.04 0.31 0.31 - I 35N ø•"-'•' "k, O.06 30N - 30N - -- 25N 25W JUL AUG SEP OCT NOV DEC pigment concentration concentration and and spatial spatial distribution distribution along along the the west west coast coast Plate Plate 1. 1. The The mean mean seasonal seasonalcycle cycle of of pigment of North (the entire entire CZCS global data data set) of North America. America. The The images images are are 8-year 8-year composites composites (the CZCS global set) of of monthly monthly satellite-measured pigment pigment concentration concentration in satellite-measured in each eachcalendar calendarmonth monthfrom from the theCZCS CZCS 20-km-resolution 20-km-resolutionglobal global data set. set. data central and This zonal zonal gradient gradient remains remains central and northern northern California. California. This relatively steep steep until until September. From May relatively September. From May through through SepSeptember, patterns tember, patterns off off the the Oregon Oregon and and Washington Washington coasts coasts appear relatively stable, stable, with with highest highest concentrations concentrations extendextendappear relatively ing approximately approximately 100 km offshore. offshore. Off Baja, relatively relatively high ing 100 km Off Baja, high concentrations extend extend farthest concentrations farthest from from shore shore(approximately (approximately 200 km) km) in inMay Mayand andJune. June. The The gradient gradient between between low low offshore offshore 200 concentrations and higher coastal concentrations is concentrations and higher coastal concentrations is stronstrongest from from February February until gest until July. July. Inspection of the Inspection of the individual individual monthly monthly images images (not (not prepresented which make make up up these these multiyear multiyear means means indicates indicates sentedhere) here) which that that the the relative relative strengths strengthsof of the the zonal zonal gradients gradients described described above above are are determined determinedprimarily primarily by by interannual interannual variability variability in in the concentration gradient gradient than the position position of of aa larger larger pigment pigment concentration than that that evident evident in in these these figures. figures. The The consistently consistently lower lower zonal zonal concentration concentrationgradients gradientswhich which appear appearin in the the multiyear multiyear means means does does not not represent represent the the strength strengthof of the the gradient gradient on on any any given given year. year. A general general mean mean seasonal seasonal cycle cycle of of pigment pigment patterns patterns in in the the A Peru Current Current (Plate to discern Peru (Plate 2) 2) is is more more difficult difficult to discern owing owing to to lack of of data. data. The The mean mean images images over over the the latitudinal latitudinal extent extent of of lack the PCS PCS for for each each month month indicate indicate that that there the there are are two two regions regions THOMAS ET AL.: CURRENT PIGMENT VARIABILITY THOMAS ET AL.: PERU PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT VARIABILITY 7359 7359 •:-_::::-:-: -•-•<•:.:-.---.o•<.... -....................• ::::-'.-:::::::::::8:::.?' ::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ::::::::::::::::::::::::::::::::::::::::::::::::::::::: ...... ........ _• =:-----------:::.---•-• :::::::::::::::::::::::::::::::::::::::::::::::::::::: ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: :::::::::::::::::::::::::::::::::::::::::::::::::::::: :::::::::::::::::: ============================================================= .... ====================================================== .... ========================================================= .... ========================================================= .... .... .... ....... ........ .............. .... ========================================================== -....: ............ :.:.:.:.:-:.:.:.:.:.:.:.:.. '•::::::::: :::::::::::::::::::::::::::::::::::::::::::::: ============================================================ ===================== ::::::::5::::::::: ::::::::: ::::::.========================================================= ============================================== :::::::::::.. ============================================= =========================================================== ================================================ ....::::::::::::::::::::::::::::::::::::::::::::::::::::::: ============================================================== ...======================================================= =================================================================== .--.:-:.:.:.:.:.:-:.:-:.:.:.:.:.:.:.:.:.:.:.:.:.:-:-:.:..... --.-.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.........:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:..... =============================================================== =========================================== .......................................................... ......................................_................... ....:.:.:.:.:.:.:...:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:... .......................................................... .................................................. ..::::::...:... ::::::::::::::::::::::::::::::::::::::::: ====================================== ====================================================== :::::::::::::::::::::::::::::::::::::::::::::::::::: ================================================ •:::.::::::'. R1-2 3-4 •::: :::::::'•:•:::::::::::::::::::::::::::::::::::::::: ========================================================= -::: ::::::::::::::::::::::::: ::::'::::::::::::::: ::::::::::::: ::::::: ::::: ::: ::::':'":'•i•::•:'.•:::::":•:: :::::":.: '•:::i::' -•'• ...-: ...... >4 ....... .:::::::::::::::::::::::::.•::::::::::::::::::•::::::::::.:.:.:.:.::::::•:::::::::::::::::::::•:•:•::::::::::::::::::::::::::.:::::.:.::::::::::::::::::::::::::::::::::•:::::::::::::::::::::::::..::......:::::::•::.::::::•:•:::•:::::::::::::: ..... JLtATJG SEP OCT NOV DEC of data data points which make make up up each each of in Figure Figure2. 2. The Thenumber numberof points(years) (years)which of the themean meanmonthly monthlycomposites composites in Plate Plate 1. 1. where where zones zones of of higher higher pigment pigment concentrations concentrations (>1.0 (>1.0 mg mg m -3) 3) closely closely associated associated with the coast coast are throughm withthe arepresent present throughout the out the year. year. The The first first region region is is north north of of 18°S 18øSalong along the the Peruvian coast, and Peruvian coast, and the the second secondis is south southof of 30°S 30øSalong along the the central and central and southern southern Chilean Chilean coast. coast. Along Along the the Peruvian Peruvian coast, highest and the coast, highest concentrations concentrations and the farthest farthest offshore offshore exextension of of these these concentrations tension concentrations appear appear to to occur occur during during the the austral summer summer (December-March), (DecemberMarch),with withfall fall(i.e., (i.e.,AprilApril austral km offshore. offshore. ItIt is km is not not possible possible to to determine determine from from the the availavailable data set patch or or able data set whether whether this this is is aa spatially spatially isolated isolated patch simply the cloud-free cloud-free portion portion of of aa more more extensive simply the extensive region region of of high concentrations. Extreme Extreme cloud cover and high concentrations. cloud cover and lack lack of of data, data, especially during austral austral winter, winter, prevent prevent aa better especially during better description description of in this this most most equatorial portion of of the the study of seasonality seasonality in equatorial portion study area. to area. Along Along the the coast coastof of northern northernChile Chile (approximately (approximately18° 18øto 30°S), 30øS), aa region region of of relatively relatively low low pigment pigment concentrations concentrations June) June) concentrations concentrationsmore more closely closelyassociated associatedwith with the the coast. coast. While lack lack of of data data makes makes the the winter While winter concentrations concentrations difficult difficult to interpret, an area to interpret, from from May May through through August August an area of of increased increased concentrations centered concentrations centered at at 15°S 15øSextends extends approximately approximately 200 200 winter (JulySeptember). Concentrations winter (July-September). Concentrations in in this this region region do do not farther north not reach reach those those found found either either farther north or or farther farther south. south. (often throughout the (often<0.3 mg mg m3) m-3)persists persists throughout theyear, year,with with slightly elevated concentrations concentrations appearing appearing in in the slightly elevated the austral austral THOMAS ET ET AL.' AL.: PERU CURRENT THOMAS PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT PIGMENT VARIABILITY VARIABILITY 7360 7360 3.13 3.13 [CHL] - -3 10S mc3m 3. 04 1.04 - 20S - 30S 0.31 0.31 0.06 0.06 - 40S _ 50S APR '•' JAN MAY"'" .JUN I lOS - 20S - 30S - 40S _ 50S 3.13 3.13 1.04 1.04 0.31 0.31 I Plate 2. Plate 2. UL • A .• SEP OCT ' 0.06 0.06 NO The of pigment pigment concentration concentration and and spatial spatial distribution distribution along along the the west west coast coast The mean mean seasonal seasonalcycle cycle of of South South America. of America. Very low are present Very low pigment pigment concentrations, concentrations,similar similar to to open-ocean open-ocean nearshore nearshore concentrations concentrations are present throughout throughout the the year. year. offshore extension extension (up (up to concentrations far far offshore in the concentrations offshore in the southern southern Pacific, Pacific, often often Maximum Maximum offshore to 400 400 km) km) of of these theseconconintrude to within 40 40 km km of of the the coast coast during during all all months months except except centrations centrationsappears appearsto to occur occurduring duringaustral australsummer summerininJanJanintrude to within July, July, August, August, and and September. September.Along Along the the southern southernChilean Chilean uary uary and and March. March. Offshore Offshoreextension extensionappears appearsto to be beminimum minimum coast (south higher pigment pigment concentrations concentrations persispersis- in July, and coast (south of of 30°S), 30øS),higher in winter winter (May, (May, July, and August), August), with with a a second secondextension extension tently extend extend up in aa region region centered centered at at offshore evident in in September. tently up to to 200 200 km km offshore offshorein offshore evident September. approximately 37øS 37°S during during austral austral fall, fall, winter, winter, and A more quantitative quantitative presentation presentation of of the the seasonal seasonal cycles approximately and spring. spring. A more cycles in in The in Figures Figures 44 and and 5. 5. Variability The July July image, image,showing showingrelatively relatively high highconcentrations concentrations the the CCS CCS and and the the PCS PCS is is given given in Variability extending north north of of 37°S and up up to to 400 km offshore, offshore, results results is as contours concentration in in time time and and extending 37øSand 400 km is shown shown as contours of of pigment pigment concentration from over the the extent extent of of each latitude over each EBC EBC for for each each of of three three from only only 2 2 years years of of data data(1979 (1979and and1981), 1981), making making it it latitude impossible to determine how well 100-km-widezones zonesextending extendingoffshore. offshore. Averaged Averaged over over the the impossibleto determine how well this this pattern patternrepresents represents 100-km-wide mean July July conditions. conditions. Between and 51°S, relatively high high 100-km-'wide 100-km-widezone zoneclosest closestto to shore shore (Figure (Figure 4a), 4a), two two regions mean Between370 37ø and 51øS,relatively regions THOMAS ET ET AL.: AL.: PERU CURRENT PIGMENT VARIABILITY THOMAS PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT VARIABILITY 7361 l-2 1-2 3-4 •,::::•i•:i:i•:ii::i•,:ii 3-4 ß '•-::::•;.-'--..-•ii: :'::.:':?-:•:..::7-•:':•.'• ß :.:.::::::::::::.: >4 ':'"i::•:i.:.•: .•i.•'" ....- .................. . ::::::::::::::::::::::::::::::::::::: .. ß Figure Figure 3. 3. The The number number of of data data points points(years) (years) which which make make up up each each of of the the monthly monthly composites compositesin in Plate Plate2. 2. Calculations are the the same same as as those those for for Plate Calculations are Plate II and and Figure Figure 2. 2. of cycles, and of the the CCS CCS have have relatively relatively strong strong seasonal seasonalcycles, and two two regions regionsshow show weaker weaker seasonal seasonalvariability. variability. A A region regionof of strong strong variability to 30°N variability reaching reachingfrom from approximately approximately 22° 22ø to 30øN is is cencentered the Baja tered at at 26°N 26øN along along the Baja coast, coast, and and aa second, second,reaching reaching from to 45øN, 45°N, is is centered centered at from approximately approximately 340 34ø to at 40°N 40øN along along the the central central and and northern northernCalifornia California and and southern southernOregon Oregon concentrations increase, beginning beginningearliest earliestinin the the year concentrations increase, year at at 37°N. By May, May, concentrations concentrationsgreater greaterthan than3.0 3.0mg mgmm3 37øN.By -3 are are found throughout this this region. region. Highest Highest concentrations concentrations appear found throughout appear to at 37°N in summer. North of to remain remain longest longest at 37øN in summer. North of 45°N, 45øN, the the relatively weak, with CZCS-estimated seasonal cycle is seasonal cycle is relatively weak, with CZCS-estimated mean pigment pigment concentrations concentrations within within 100 100km km of of the the coast mean coast -3 coasts. Baja, mean mean concentrations concentrationsare are <1.5 <1.5 mg mg m m3 remaining above3.0 3.0mg m3 the coasts. Off Off Baja, remaining above mgm -3 throughout throughout theyear. year.Maximum Maximum from through the the winter (>4.0 mg mg m m3) from August August through winter into into March. March. By By April, April, concentrations concentrations (>4.0 -3)within withinthis thisregion, region,however, however, concentrations are 2.0 mg m and increase to a maximum also occur during summer months. Within concentrations are2.0 mgm-3 andincrease to a maximum also occur during summer months. Within the the Southern Southern of >3.0 mg Between central California Bight region region (centered (centered at mg m3 m-3ininMayJune. May-June. Between central California California California Bight at approximately approximately 32°N), 32øN), and and central central Oregon, Oregon, concentrations concentrationsare are approximately approximately 2.0 2.0 the the seasonal seasonalcycle cycle is is also alsorelatively relatively weak. weak. Concentrations Concentrations mg September until February. In boxes do averaged within within the the 100-km-wide 100-km-wide boxes do not not fall fall below below 0.5 0.5 mgm m-3 from fromSeptember untilFebruary. InMarch Marchand andApril, April, averaged 7362 THOMAS ET AL.: CURRENT PIGMENT THOMASET AL.' PERU PERUAND ANDCALIFORNIA CALIFORNIA CURRENT PIGMENTVARIABILITY VARIABILITY 7362 48 48 48 46 48 46 44 44 44 44 42 42 42 40 40 40 .0 38 ' 40 36 3[5 i 36 36 34 •i 34 32 32 - 38 36 ' 34 32 ;,;,,,,•. 5•. . i ' 34 5• i . . 32 30 3O 3030 28 28 28 26 26 26 26 24 24 24 22 22 22 30 ,, . NDJ a(] FM AMJ J AS OND JF NDJ F b b 0 - 100 KM OFFSHORE 100 -- •oo 200 n• KM on, OFFSHORE •oo s•o•. . NDJ FM AMJ I AS OND IF c OFFSHORE 200 - •oo 300 KM •oo • o•s•o•s C Figure 4. The mean seasonal cycle of surface pigment concentration in the California Current system. Figure4. Theare mean seasonal cycle ofsurface pigment concentration intheCalifornia Current system. Concentrations of pigment per within Concentrations aremilligrams milligrams ofsatellite-measured satellite-measured pigment percubic cubicmeter meterand andare aremeans means within 100-km-wide zonesoriented orientedparallel parallelto tothe the coast coast and and contoured contoured in in time time and and latitude. latitude. Data Data were sampled 100-km-wide zones were sampled from shown in 11and Figure 22by within 100 X 100 km areas oriented along the fromthe theimages images shown inPlate Plate and Figure byaveraging averaging within 100x 100kmareas oriented along the coast coast (see (see Figure Figure 1). 1). mg m the late fall, and and winter winter and and rise rise to to a scattering mg m-3 in the latesummer, summer, fall, scattering algorithm; algorithm; and and differences differences in in the the land landmasking masking maximum of less less than than 2.0mg in May May and and June. June. Of the processing of the two maximum of 2.0mgm m-3 in Of note note used usedin in the processing of the two data datasets, sets,which whicheliminated eliminated is is the the fact fact that that no nowinter wintermaximum maximumis is found found in the the Southern Southern some concentrations next next to the coast some high pigment concentrations coastin in the the California Bight in in the the reprocessed CZCS CaliforniaBight reprocessed CZCSglobal globaldata dataset. set. previous previousstudy. study. This which This is is in in contrast contrastto toprevious previousanalyses, analyses, whichused useddata data In In the the regions regions100-200 100-200 km km and and200-300 200-300 km kmoffshore offshore processed with atmospheric 4b and and 4c), 4c), seasonal seasonal variability variability is is minimal minimal south south of of processed withthe thesingle singleRayleigh Rayleighscattering scattering atmospheric (Figures (Figures4b correction et al., al., 32°N. In this this southern southern area, area, concentrations greater correctionand andfound foundaa winter wintermaximum maximum[Michaelsen [Michaelsen et 32øN.In concentrations greaterthan than0.5 0.5 1988]. 1988]. mg m m3 offshore into zone mg -3do donot notextend extend offshore intothe the200-to 200-to300-km 300-kmzone Concentrations over range (Figure 4c). 4c). Within Within the the 100to 200-km zone they develop Concentrations overthe thelatitudinal latitudinal rangeof ofthe theCalifornia California (Figure 100-to 200-kmzonetheydevelop Current shown approximately initially in in April, April, centered centered at at 27øN. 27°N. In In May and and June June these these Current shownin in Figure Figure4a 4a are, are,on onaverage, average, approximatelyinitially 0.5 than those reported by rub etal. concentrations are are present present throughout this portion 0.5mg mgm m-3higher higher than those reported bySt $trub etal.[1990]. [1990]. concentrations throughout thissouthern southern portion This due principally principally to to three three factors: factors: the the use use of Thisdifference differenceis due of of until only of the theCCS, CCS,but butthey theypersist persist untilNovember November onlyatat27°N. 27øN. the the entire entire CZCS CZCS data data set setto tocalculate calculateconcentrations concentrationsshown shown Maximum offshore concentrations (>1.5 and >1.0 Maximumoffshore concentrations (> 1.5and> 1.0mg mgm m-3 here as previously; zones) hererather ratherthan thanthe theuse useof ofaasubset subset aswas wasdone done previously;in in the thetwo twocross-shelf cross-shelf zones)are arepresent presentnorth northof of32°N 32øNin in the global CZCS and California in theuse useof ofthe thereprocessed reprocessed global CZCSdata datawith withaamultiple multiple NovemberJanuary November-January andoff offcentral central California in the thespring spring 6 -6 c -11 -11 -11 -11 -16 -16 -16 -21 -21L -26 -2 -31 -31 a -2 -31 ß -36 -38 -41 -41 -41 -48 -46 -46 NDJ FM AMJ JJ AS JF NDJ FM AMJ AS OND OND JF OFFSHORE 0 KM 0 -- 100 100 KMOFFSHORE -51 b ( -21 , -36 -51 '.L.)( ) 51-L)I, NDJ AMJ JJ AS JF ND J FM FM AMJ AS OND 0 ND JF 100 100 - 200 200 KM OFFSHORE OFFSHORE NDJ AMJ IJ AS JF ND J FM FM AMJ AS OND 0ND JF C 200 200 - 300 300 KM OFFSHORE Figure seasonal cycle pigment concentration concentration in in the the Peru Peru Current Current system. Details Figure5. 5. The Themean mean seasonal cycleof ofsurface surface pigment system. Details are are the thesame sameas asfor for Figure Figure4. 4. THOMAS ET AL.: CURRENT THOMAS ET AL.' PERU PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT PIGMENT VARIABILITY VARIABILITY 7363 7363 (beginning in March). March). Lowest (<1.0 and the same and latitudinal patterns described (beginning in Lowest concentrations concentrations(<1.0 and the same seasonal seasonaland latitudinal patterns describedabove, above, <0.5mg 4b are with concentrations approximately 0.5 0.5 mg mg m-3 less less than <0.5 mgm3, m-3in , inFigures Figures 4band and4c, 4c,respectively) respectively) arepresent present with concentrations approximately than initially in May May in in the the northern portion of of the the study those in in the the zone zone 100-200 100-200 km km offshore. offshore. initially in northern portion studyarea area those (north and extend extend south south to (north of of approximately approximately 41°N) 41øN) and to include include the the whole whole study studyarea areanorth northof of32°N 32øNby by late late summer. summer. Contours of pigment in time Contours of pigment concentration concentration in time and and latitude latitude within the first km of of the the South South American American coast coast (Figure within the first 100 100 km (Figure Sa) 5a) indicate indicate four four regions regionsof of differing differingseasonal seasonalvariability; variability; however, lack data (Figure 3) for for austral austral winter winter presents presents aa however, lack of of data (Figure 3) 3.2. 3.2. Interannual Variability Interannual Variability While details of of interannual on time While details interannual variability variability on time scales scalesas as short as short as 15 15 days days are are sometimes sometimespossible possible to to resolve resolve in in the the California Current system system [Thomas [Thomas and and Strub, California Current Strub, 1989], 1989], the the sparse CZCS CZCS data data set set over over the the Peru Peru Current Current prevent prevent all all but but sparse severe for interpretation. The isolated peaks of severe problem problem for interpretation. The isolated peaks of the the most most coarse coarseestimate estimateof ofinterannual interannualvariability variability (even (even concentrations greater during July July (austral but they concentrations greaterthan than1.5 1.5mg mgm m-3 during (austral these these are are suspect, suspect, but they represent represent all all that that can can be be done done winter) in in aa relatively band centered centered at at with winter) relatively narrow narrow latitudinal latitudinal band with the the CZCS CZCS data data set). set). The The concentration concentration anomalies anomalies are are 12°Sand andin inOctober October over over aa wider wider latitudinal latitudinal range range come come from from for temporal means means within within each each year, year, centered 12øS for the the 8-month 8-month temporal centered images for the the July-October JulyOctober period within each each of imagesfor periodwith withvery veryfew few data datain in this this in in midsummer, midsummer, within of the the four four alongshore alongshoreoriented oriented region. Coverage in Figures Figures llaa and and lb, region. Coverageimproves improvesduring during November November through throughMay May zones zones shown shown in lb, as asdescribed describedin in section section2. 2. (austral summer, and and fall), provide an interannual differences (austral spring, spring, summer, fall), and and the the data data show showrelarela- These These anomalies anomaliesprovide an index index of of interannual differences tively high high values values in in spring spring and and summer summer and and aa decrease decrease in fall in the relative pigment concentrations concentrations over over the the time time span span of of tively in fall in the relative pigment data set (AprilMay). (April-May). Since Sinceupwelling-favorable upwelling-favorable winds winds are are most most inin- the the CZCS CZCS data set for for the the summer summer period. period. Because Because two two tense off zones within within the the PCS latitudinal zones PCS show show maximum maximum seasonal seasonal tense off Peru Peru in in winter winter [Bakun [Bakun and and Nelson, Nelson, 1991] 1991] and and latitudinal surface indicate strongest strongest upa second values in in midwinter, midwinter, a secondanomaly anomaly was was calculated calculatedfor for surface temperature temperatureclimatologies climatologiesindicate up- values welling (coldest (coldest surface surface temperatures) temperatures) in welling in winter winter [Weare [Weare et et the the PCS PCS using using an an 8-month 8-month mean mean centered centered in in (austral) (austral) midmidal., 1980; Halpern et al., 1980; Halpern et al., al., 1992], 1992], it it seems seemsmost most likely likely that that winter. winter. These These means means are are therefore therefore over over aa period period identical identical to to chlorophyll concentrations remain high during winter, that for the CCS anomaly calculations. chlorophyll concentrations remain high during winter, that for the CCS anomaly calculations. spring, and summer, although the the data data here here document Figure 6 shows Figure 6 shows the the pigment pigment concentration concentration anomalies anomalies for for spring, and summer, although document only only the spring four latitudinal regions within within the the CCS Northwest, the springand and summer summerperiods periodswell. well. four latitudinal regions CCS (Pacific (Pacific Northwest, the Southern Along the the northern Along northern Chilean Chilean coast coastbetween between18° 18øand and33°S, 33øS, central central California, California, the Southern California California Bight, Bight, and and Baja Baja concentrations are are less less than than 0.5 for each each of of the the three zones extendCalifornia) for three 100-km-wide 100-kin-wide zones extendconcentrations 0.5 mg mgm m-3 from fromNovember November California) until April offshore. Within Within these these latitudinal latitudinal regions regions the the patterns patterns of ing offshore. of until April (austral (austral spring springthrough throughfall). fall). Concentrations Concentrations ing greater than are present present only July and and August variabilityare are similar, similar, but but the greater than1.0 1.0mg mgm m-3 are onlyin in July August interannual interannual variability the amplitude amplitude dede(winter), and creases increasing distance distance of of the the zones zones from from shore. shore. A creaseswith with increasing A (winter), centered centeredat at 28°S. 28øS.In In aa third third region, region, between between33° 33øand 45°S, concentrations less than 1.0 mg m3 are present from second general feature is the reduced relative amplitude 45øS,concentrations lessthan1.0mgm-3 arepresentfrom second general feature is the reduced relative amplitude of of January through (summer). Maximum Maximum concentrations concentrations anomalies in the January through March March (summer). anomalies in the two two lowest lowest latitude latitude regions regions (Southern (Southern (>2.0 mg m3) latitudinal region are in Bight and and Baja Baja California). California). Common Common to to all regions California Bight all regions (>2.0 mgm -3)ininthis this latitudinal region arereached reached inMay May California (fall) and again again in in July July (winter), (winter), with with concentrations concentrations greater greater is is aa negative negativeanomaly anomaly in in 1983, 1983, in in agreement agreementwith with previous previous (fall) and persisting (except in at from Thomas and and Strub, Strub, 1990; Strub et than 1.5 studies[Ffedler, [Fiedler, 1984; 1984; Thomas 1990; Strub et al., al., than 1.5mg mgm m-3 persisting (except inOctober) October) at37°S 37øS from studies 1990; Abbott and and Barksdale, Barksdale, 1991] that show 1990; Abbott 1991] that show relatively relatively low low pigment concentrations during (the El pigment concentrations during 1983 1983 (the El Nino Nifio period). period). anomaly in in 1983 is strongest strongest in in the with minimum concentrations(<1.0 (<1.0mg mgmm3) minimumconcentrations -3) in in austral austral The The negative negative anomaly 1983 is the central central winter (MayJuly) and region and and is is accompanied accompanied by California region by aa negative negativeanomaly anomaly winter (May-July) andconcentrations concentrationsgreater greater than than 1.5 1.5 mg mg California m3 spring, summer and in 1984 off Baja Baja and and in in the of similar similar magnitude magnitude in 1984 off the Southern Southern m-3during during spring, summer andfall. fall.In Inthe thezone zone100-200 100-200km km of June until June until December December (winter (winter and andspring). spring).South Southof of45°S 45øSthe the seasonal appears similar to that that at seasonalcycle cycle appears similar to at the the lowest lowest latitudes, latitudes, offshore offshore(Figure (Figure 5b), 5b), differing differingseasonal seasonalcycles cycleswithin withindifferdiffer- California Bight(in (in the the 0California Bight 0- to to 100-km 100-km zone). zone). This This latter latter ent regions are are also contrasts to to the anomaly contrasts the situation situation in in the the Pacific Pacific Northwest Northwest ent latitudinal latitudinal regions also evident. evident. At At lowest lowest latitudes latitudes anomaly (equatorward concentrations are are highest region, where where the the largest largest positive positive anomaly anomaly occurs occursduring during (equatorward of of 12°S), 12øS),concentrations highest(>1.0 (> 1.0mg mg region, The largest largest positive positive anomalies in central m3) until and and 1984. The anomalies in central California California m-3)from fromOctober October untilMarch March(spring (spring andsummer) summer) and 1984. lowest from April April until until September, September, although occur in in 1981 1981 and and 1986. 1986. lowest(<0.5 (<0.5mg mgm m_3) -3) from although occur lack of data in winter may obscure the true signal during The interannual variability pigment concentration concentration along lack of data in winter may obscure the true signal during The interannualvariability of of pigment along coast is 7 for for the the four four JulySeptember. the South South American American coast is shown shown in in Figure Figure 7 July-September.InInaaregion regioncentered centeredatatapproximately approximately15°S, 15øS, the maximum annual concentrations concentrations (>1.0 are reached reached latitudinal regions and and three three cross-shelf zones shown latitudinal regions cross-shelf zones shown in in maximum annual (> 1.0mg mgm m_3) -3) are in Figure lb. lb. In In discussing discussingthe the interannual interannualvariability variability off off Chile, Chile, in JuneAugust June-August(winter), (winter),and andconcentrations concentrationsremain remainabove above Figure 0.5 mg m m3 the zone is important important to to note note that that the the anomalies anomaliesduring during the the 1983 1983 El El 0.5 mg -3throughout throughout theyear. year.In Inthis thisoffshore offshore zoneany any itit is Niflo are are based based on on only only one one data data point within the the seasonal cycle cycle has point (1 (1 month month within seasonal has aa relatively relatively small small amplitude amplitudebetween between17° 17ø Nifio and 23øS. 23°S. Between Between 23 23° and41øS, 41°S,however, however,there there is is an an austral austral 8-month 8-month averaging averagingperiod). period). Given Given the the scatter and øand scatter of of monthly monthly data points points during during periods periods when when more more than than one one are are available, available, winter maximum maximum with with concentrations concentrations greater greater than than 1.0 winter 1.0 mg mg data present in July. concentrations (>0.5 mg the most conclusion isis that that the the available data in most reasonable reasonable conclusion available data in m -3 present m July.Increased Increased concentrations mg the associated with with this this maximum begin earlier earlier and and persist persist the the global CZCS data data set to document m _3) global CZCS set are are insufficient insufficient to document the the m -3) associated maximum begin response to increasingly longer longer at at higher higher latitudes latitudes within to the the1983 1983El El Niflo Nifio off off Chile. Chile. The The negative negativeanomanomincreasingly within this this region. region.In In response off Peru, Peru, for for which which more more data data points points are are available, are the region south of the seasonal seasonal cycle cycle appears appears similar alies off available, are the region south of 43°S, 43øS,the similar alies more credible. Off the Peruvian coast, within the first 100 to that at the lowest latitudes, with highest concentrations more credible. Off the Peruvian coast, within the first 100 to that at the lowest latitudes, with highest concentrations km, maximum positive anomalies anomalies occur occur in in (the present in in early early spring spring and and summer summer(August-October, (AugustOctober, Janumaximum positive (the summer summer present Janu- km, period ending aryMarch) and concentrations (<0.5 (<0.5 period ending in) in) 1980 1980 and and 1982, 1982, and and maximum maximum negative negative ary-March) andan anindication indicationof of lower lower concentrations during austral austral fall fall and and winter July), occur in and 1984. Farther from mg m anomalies occur in 1979 1979 and 1984. Farther from shore shore (100(100mg m_3) -3) during winter(April (Aprilthrough through July), anomalies to 200and 200to 300-km zones) in this lowest latitude although the data for most of this latter period are few. The to 200and 200to 300-km zones) in this lowest latitude althoughthe data for most of this latter period are few. The offshore zone zone 200-300 km from from the the coast region, the the strongest strongestsignals signalsare are negative negativeanomalies anomaliesin in 1983 1983 offshore 200-300 km coast (Figure (Figure Sc) 5c) shows shows region, THOMAS ET AL.: THOMAS ET AL.' PERU PERU AND AND CALIFORNIA CALIFORNIA CURRENT CURRENT PIGMENT PIGMENT VARIABILITY VARIABILITY 7364 7364 a 0-100 0-100 KM OFFSHORE OFFSHORE 4.0 4.0 2.0 t 0 2.0 0.0 -2.0 - * o * 79 79 80 80 4.0 40 2.0 o- 0 o 0 0.0 0 0 • o Ii 82 83 84 82 83 84 Pac i;fic f I c Northwest Pac Northwest 01 81 0.0 85 85 -2.0 -20 ...... • 79 79 I 80 80 I I II 82 82 I I 83 83 2.0- 20 -20 79 I 80 8 I 81 80 81 : I 83 83 ?. 84 84 I I B E 79 79 3.0 3 0 - 1.5 15 - -1.5 -15 I 85 85 86 83 83 82 82 84 84 0 79 79 0o o ° o I I I 81 82 82 83 83 86 0 o 0 o 0 o • I 79 79 I 80 80 0o a__._•_.• .......,_. 81 I 83 83 o 1.0 I 84 84 I 85 85 8 86 I 79 79 calculated for each each year year calculatedfrom from the the 8-month 8-monthcomposite compositeimage imagefor formed formed from from all all 20-km 20-km pixels pixels available availablewithin within the the 8-month 8-month period. period. Individual Individual months months with with large large amounts amountsof of missing missing data data would would therefore therefore contribute contribute little little to to the the 8-month 8-month anomanomaly, points in in the the aly, although althoughthey they would would appear appearas asindividual individual points figure for the the year year in figure for in question.) question.) 1.0 10 - 85 85 86 0 o o ............... 0 I 80 80 81 83 83 82 82 0 o 0 o 8o o -..... ..- S-0 6 -1 .0 I E • -1 0 79 79 2.0 2 0 84 84 Pacific Pac; fic Northwe,t Northwest 0.0 • E 0 0 0.0 -1 -1.0 .0 ' o 0 e the the monthly monthly values values presented presented in in the the figure. figure. They They were were central centralChile, Chile, the thepositive positiveanomaly anomalyin in1983 1983becomes becomesincreasincreasingly ingly strong strongwith with distance distancefrom from shore. shore. I 82 82 2.0 anomalies of anomaliesshown shownby by the the solid solidcurve curveare are not not aa simple simplemean mean of from from shore. shore. At At the the highest highestlatitude latitude region region of of the the PCS, PCS, off off southern Chile, Chile, the the strongest strongest signal signal in in all all three three zones zones is is aa southern positive positive anomaly anomalyin in 1981. 1981. This This is is accompanied accompaniedby by positive positive anomalies in 1982 and 1983. 1983.InIn aa trend trend similar to that that off anomalies in 1982 and similar to off 886 200-300 KM 200-300 KM OFFSHORE OFFSHORE C c 2.0 2.0 - anomaly anomalyin in 1982, 1982, with with aa positive positiveanomaly anomalyin in1983 1983becoming becoming increasingly increasinglysimilar similar to to that thatof of1982 1982with with increasing increasingdistance distance 85 85 Baja Ba3a California California indicates one or or more indicates one more values values off off scale. scale. (Note (Note that that the the and and 1984 1984 and and positive positive anomalies anomalies in in 1982. 1982. Anomalies Anomalies off off northern northern Chile Chile are are the the weakest weakest in in the thefour four regions regionsand and are are generally generally positive positive in in 1981, 1981, 1982, 1982, and and1983 1983and and negative negativein in other years. Off the central Chilean coast the other years. Off the central Chilean coast thestrongest strongest relative in all relative signal signal in all three three cross-shelf cross-shelfzones zones is is aa positive positive 0.... Southern California Bight Southern California Bight Baja Baja California Cal ifornia Eight-year time time series Figure Figure 6. 6. Eight-year series(solid (solidcurve) curve)of of anomalies anomalies of of pigment pigment concentration concentrationwithin within three three offshore offshorezones zones and and four four latitudinal latitudinal regions regionsof of the the California California Current Current system system(see (see Figure Figure 1). 1). Units Units of of the theanomalies anomaliesare aremilligrams milligramsof of pigment pigment per per cubic cubic meter meter and and are arean an8-month 8-monthmean meanconcentration concentration (centered (centeredon on midsummer) midsummer)minus minusthe the 8-year 8-year average averageof of these these concentrations for that that zone. concentrationsfor zone. Note Note the thechange changeof of scale scalefor for the the anomalies anomaliesas as distance distancefrom from shore shoreincreases. increases.Data Data points points indicate indicate the the range range and and number numberof ofactual actualmonthly monthlymeans means available the 8-month 8-month period period in in each each year. year. An An asterisk asterisk availablewithin the o I I 84 84 00•.....•....•.....o .'' 85 0 õ 80 80 1.5 15 -1 .5 -15 84 e86 0 0.0 83 85 85 Central California Central Cal iforn;a 3.0 - 0 82 8 86 6 00 84 84 I 85 85 0 E -1 .5 0 8 0 0 o.ek.4 0 o o o 0 0.00 Southern California Bight Southern California Oigh• 4 I 82 82 0 ..... • --9•L•$ ..... 81 I 81 Pacific Northwest Pacific Northwest 15 o 0.0 I 80 80 - 1.5 0 o Central Cen•nai California Cal ;fornia 2.0 20 79 79 3.0 30 82 83 84 85 -2 0 79 79 80 80 0181 82 83 84 85 86 4.0 40 -1 -15 .5 8 86 0 a -2.0 15 o •o 0 0 0 0 6õ 3.0 3 0 0 0 100-200 KM OFFSHORE 100-200 OFFSHORE b - I I I I Central Cal Centra• Cal ifornia ;fornia 0 o 88 o 0 I 80 81 85 8 80 81 82 82 83 83 84 84 85 86 0 •' 1.0610o I 8 0.0.-.....0....i o 0 ß 0 8 -p 0 1 00 .....•....o ...q....H.• -1.0 -lO o 6 79 79 I 80 II 81 I I 82 82 84 84 83 83 85 85 8 86 Southern California Bight Southern California Bight 2.0 1.0 0 0.0 0 I-.' -•I -....... -1 -1 .0 o 79 79 I 80 80 0o •' '•' ..... I 81 a w I 82 - -I--- .•. ,.. I 83 83 ' ' '' C ..... I 84 84 Baja Ba j a California Ca t i forn i a Figure Figure 6. 6. (continued) (continued) ., I 85 85 - - - -u 86 86 7365 7365 THOMAS ET ET AL.' AL.: PERU THOMAS PERU AND AND CALIFORNIA CALIFORNIA CURRENT CURRENT PIGMENT PIGMENT VARIABILITY VARIABILITY 0-100 0-100 KM KM OFFSHORE OFFSHORE a 4.0 4.0 2.0 * :0 oO 8 0 2o o.• 0 79 79 80 80 81 8 8 9 ' I 82 82 83 83 84 84 Peru 4.0 4 0 - Peru - , E -2.0 -2 0 .... 79 I 80 80 I I 81 •- ........ 85 85 I I 82 82 -1.55 • II II I 84 84 85 85 8 -1.55 -1 I 86 B E E 0o 0.0 0.0 -2 0 -2.0 79 79 80 80 81 4.0 0 4 , 02 82 Centrat Central 85 81 82 ----., - • I - I 80 80 . I 79 7g 80 80 81 86 I I I I 84 84 03 83 1 .---,',_.•... Northern Northern Chile Chi[e I I 85 85 T 886 .... 80 80 82 82 81 - 83 83 Central , 04 84 85 85 86 86 85 85 86 86 Chile Oh i le 0 0.0 o o 0 I 82 82 I I 1.55 0 • 82 82 81 81 1 I 85 _ ... .... 0 -2.0 -2 0 84 - 79 79 3.0 3 0 0 0.0 0 0 83 ..... 8 86 o 2.0 2 0 -: Peru 79 79 1.5 I 83 84 83 84 Chile Chile 80 bO 3.0 1.5 1 5 I 79 79 0.0 -___ 2 2.00 °o -0 -_.r00 , _._,., • - • ..•. o 0 83 83 8 0o 0 3.0 30 1.5 - NortI,ern Chile Nor"•:h,),ern Chi [ e 4.0 - 1.51 5 -1 886 * ß -c-.---- * 0.0 -o 2.0 0 0 0 3.0 - 3 0 o 0.0 0 0 -2.0 -2.0 100-200 KM 100-200 KM OFFSHORE OFFSHORE b -1 .55 I -1 83 83 84 84 85 85 86 86 79 79 80 I 84 84 Southern Chile Southern Chi l e Southern Chile Sou f. hern Ch i[e Eight-year time time series FIgure Figure 7. 7. Eight-year seriesof of anomalies anomaliesof of pigment pigment concentration concentrationwithin within four fourlatitudinal latitudinalregions regionsfor for each eachof of three three cross-shelf zones in Peru Current system (see Figure 1). cross-shelf zones inthe the Peru Current system (see Figure 1).CC The The solid solid curves curvesrepresent representanomalies anomaliescalculated calculated from from the the 8-month period period (September-April) (September-April) centered centered on on austral austral sum8-month sum- mer. The mer. The dashed dashedcurves curvesrepresent representanomalies anomalies from from the the 8-month period period centered centered on on austral austral winter winter (March-October). (March-October). 8-month Note the concentration between Note the change changeof scale scalefor pigment pigmentconcentration between the zone the zone closest closest to to shore shore (0-100 (0-100 km) km) and and those those farther farther offshore. Data Data points points are are for the austral offshore. for the australsummer summermonths monthsand and are are as as described describedfor for Figure Figure 6. 6. I 83 83 I 82 82 81 81 200-300 200-300 2 0 1 0 0 0 KM KM OFFSHORE OFFSHORE o -1 0 79 2 80 I I 81 82 82 o 83 83 84 84 85 85 8 86 Peru P er u 0 1.0 1 0 Anomalies in the the PCS Anomalies in PCS formed formed during during the the 8-month 8-month winter winter period (Figure 7, curve) do differ markedly from period (Figure 7,dashed dashed curve) donot not differ markedly from• 0.0 ½½g those formed from averaging period. Two those formed fromthe thesummer summer averaging period. Twodifdif- ? -1.0 79 79 ferences appear most notable. Off the Peru coast, negative E anomalies in 1983 and 1984 are not as strong as those from 2.0 E summer period period. In In the the the summer the northern northernand andcentral centralChilean Chilean where seasonal occur in in winter, winter, the regions, regions,where seasonalmaxima maxima occur the stronstron•1.0o gest positive anomalies occur in 1981 rather than 1982 or ferences appear most notable. Off the Peru coast, negative • -• 0 anomalies in1983 and 1984 are not asstrong asthose from• 20 - --' ----- -.----.g ----•-• ..... .-..-_, e----• I 80 80 ß gest positive anomalies occur in1981 rather than 1982 or 1983. 1983. 0.0 oo -1.0o -1 4. Comparisons and Comparisons and Discussion Discussion between latitudinal latitudinal patterns patterns of The The strongest strongestsimilarity similarity between of in the the CCS seasonal seasonalvariability variability in CCS and and the the PCS PCS evident evident in in Figures Figures 4 4 and and 55isisthat thatminimum minimumconcentrations concentrationsand and aa low-amplitude seasonal cycle cycle are are present low-amplitude seasonal presentat atmidlatitude midlatitude locations within these these EBCs. EBCs. This This is is counter counter to to the the standard locations within standard biological patterns expected expected in in open biologicalpatterns open oceans, oceans,where whereminiminimum seasonal variability is is expected mum seasonal variability expected at at lower lower latitudes. latitudes. Previous Previouswork work in in the theCCS CCS [Lynn [Lynn and andSimpson, Simpson,1987; 1987;Strub Strub 2.0 20 79 79 I I I 83 84 I I 83I 84I Northern Chile Nor •hernOh i [e 81 82 82 I 85 85 886 0 • ......•• --•:g-,•-.......... I 60 80 - I • 81 81 I I 82 82 I I 83 83 I 84 84 85 85 86 86 85 85 86 86 ,1,Central Chile Central Ch i le 10 * 1.0 0 0 0.0 o 0 -1 .0 I 79 79 80 80 I 81 81 I 82 82 I 83 83 84 84 Southern Chi Sou•:hern Chi l e e Figure 7. (continued) Figure 7. (continued) THOMAS ET CURRENT PIGMENT VARIABILITY THOMAS ET AL.: AL.' PERU PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT VARIABILITY 7366 7366 Summer Summer Winter Winter PaII Fall Spring Spring 55 60 75 70 -85 -80 -75 -70 • 10 -10 2 10 -15 15 15 -2o 20 20 -25 25 25 30 30 -3o 35 -55 40 -40 . -t t/; . ,, %. ,,1. "7' 35 -, 40 -. *. %. 1-,o 45 -45 45 50 -50 50 65 60 75 70 -85 -80 -75 -70 65 60 75 70 -85 -80 -75 -70 -10 10 10 -15 15 15 -2O 20 20 25 -25 25 30 -3O 30 35 -35 35 40 40 45 -45 45 50 -5O 65 60 75 70 -85 -80 -75 50 -70 Figure Filure 8. 8. The The 10-year 10-year (1979-1988) (1979-1988) mean mean wind wind seasonal seasonal cycle cycle over over the the PCS PCS calculated calculatedfrom from ECMWF ECMWF wind wind fields. fields. (top) The 10-year mean field fieldfrom fromsummer summer(January-March), (JanuaryMarch), (top) The 10-year mean fall (AprilJune), winter (Ocfall (April-June), winter(JulySeptember), (July-September),and andspring spring (OctoberDecember); tober-December); (bottom) (bottom) the the same sameperiods periods with with the the 1010year removed to to emphasize emphasize seasonal seasonal year overall overall mean mean wind wind field field removed differences. differences. large latitudinal latitudinal range range of of the the South west coast large South American American west coast during July. July. In In Plate during Plate 2 2 and and Figure Figure 3, 3, increased increasedconcentraconcentrations in the the northern region, both both nearshare nearshore and and in tions in northern Chilean Chilean region, in regions farther farther from km), are are evident evident in in other other regions from shore shore(100-300 (100-300 km), winter months months (May, (May, August, August, and and September). September). The The degree degree to to winter which the the average which average July July concentrations concentrationsavailable available from from the the CZCS data set CZCS data set shown shown here here represent representlong-term long-term mean mean July July conditions is unknown. within conditions is unknown. Concentrations Concentrations within the the middle middle latitudinal range range of of the the PCS PCS for for July July are are calculated calculated from from data data latitudinal from just just two from two years years (see (seeFigure Figure 3). 3). Examination Examination of of the the original CZCS monthly monthly images images from from each each year original CZCS year (not (not prepre- sented sented here) here) indicates indicates that that the thedata dataare arefrom from1979 1979and and1981. 1981. Furthermore, patterns and within these these 22 Furthermore, patterns and concentrations concentrations within years pigment values values extend extend years are are dramatically dramaticallydifferent. different. High High pigment both than is both farther farther north north and andfarther fartheroffshore offshorein in July July 1981 1981 than is evident in either or other evident in either July July 1979 1979 or other winter winter months monthsin in any any year year (June (June and and August). August). In In the the CCS, CCS, the thetime-latitude time-latitudeevolution evolutionof of the thepigment pigment concentration region closest closest to to the the coast coast concentrationin in the the 00- to to 100-km 100-kmregion (Figure 4) is is similar to the (Figure 4) similar to the time-latitude time-latitudeevolution evolutionof of alongalongshore wind stress stress curl curl discussed discussed by by shorewind wind stress stressand and offshore offshorewind Strub [1990].This Thisis isconsistent consistent with with the the hypothesis hypothesis that that Strubet et al. al. [1990]. higher higher surface surface pigment pigment concentrations concentrationsare are associated associatedwith with coastal [Strub et et al., al., 1990]. Based Based on coastaland andoffshore offshoreupwelling upwelling [Strub climatological wind data, data, this this does does not not appear appear to true over climatologicalwind to be be true over much much of of the the PCS. PCS. Figure Figure88shows showsthe the10-year 10-year(1979-1988) (1979-1988) mean meanwind wind stress stressseasonal seasonalvariability variabilitywithin within3-month 3-monthperiods periods over wind stress stress data data over the the PCS. PCS. These These data data are are based basedon on daily daily wind from the European Weather Forefrom the EuropeanCentre Centrefor for Medium-Range Medium-RangeWeather Forecasts Wind stress stress over casts(ECMWF). (ECMWF). Wind over the the PCS PCS is is persistently persistently upwelling favorable from from 66° to 32°S, upwelling favorable ø to 32øS, in in agreement agreementwith with Bakun and and similar similar to to the Bakun and Nelson Nelson [1991] [1991] and the CCS CCS equatorequatorward the wind wind ward of of approximately approximately36°N. 36øN.Between Between34° 34øand and 50°S, 50øS,the stress stressover over the the PCS PCS reverses reversesitself itself seasonally, seasonally,with with the the most most seasonal seasonalvariation variationbetween between34° 34øand and44°S, 44øS,again againsimilar similarto to the the CCS CCS north north of of 38°N. 38øN. Comparison of the the wind Comparison of wind and and pigment pigment seasonal seasonalcycles cycles in in the that over the PCS PCS shows shows that over Peru, Peru, northern northern Chile, Chile, and and central central Chile, pigment concentrations concentrations do Chile, pigment do not not have have an an obvious obviousrelarelationship to wind forcing such such as as that that seen tionship to wind forcing seen in in the the CCS. CCS. The The region where the region where the surface surfacepigment pigment concentrations concentrationsare are maximaximum in (austral) (austral) fall, fall, winter, winter, and mum in and spring spring is is either either the the region region (approximately 32° to 42øS) 42°S)where where winds winds have have aa strong (approximately 32 ø to strong seasonal seasonal cycle cycle (most (most upwelling upwelling favorable favorable in in summer summer and and etal., favorable in et al., 1987; 1987; Thomas Thomas and Strub, $trub, 1990; 1990;Abbott Abbott and and Barksdale, Barksdale, downwelling downwelling favorable in winter) winter) or orthe theregion region(20° (20øto to32°S) 32øS) 1991] suggests suggests that that current where winds 1991] current structure structure and/or and/or wind-driven wind-driven where winds are are persistently persistentlyupwelling upwellingfavorable favorablewith with little little physical forcing can can dominate patterns, at variation. Along Along the the northern physical forcing dominate seasonal seasonalpatterns, at least least seasonal seasonal variation. northern Chilean Chilean coast, coast, the the within this EBC, CZCS data data show within this EBC, and and that that seasonality seasonalityis is low low within within the the CZCS show that that in in aa region regionwhere wherewinds windsare areupwelling upwelling Southern favorable throughout throughout the the year, SouthernCalifornia California Bight. Bight. Figure Figure 55 suggests suggeststhat that such suchgengen- favorable year, pigment pigment concentrations concentrations rereeralities be extended extended to and that that seasonality is main main very very low low at at all eralitiesmight might be to the the PCS PCS and seasonalityis all times times along along the the coast. coast. This This pattern pattern also contrasts strongly also low low within within and and immediately immediately south south of of the the large largebight bight contrasts strongly with with that that evident evident in in the the CCS CCS even evenwithin within formed near near the Peru-Chile formed Peru-Chile border. border. the same latitudinal region (Baja California). Even the same latitudinal region (Baja California). Even within within 20 20 While a similar similar mean mean seasonal seasonal cycle, km of of the the coast coast off northern Chile, off northern Chile, concentrations concentrations in in Plate Plate 22 While a cycle,albeit albeitwith with aa strongly strongly km varying amplitude, is and Figure November through varying amplitude, is apparent apparentthroughout throughoutmost mostof of the thelatilati- and Figure 3 3 appear appear relatively relatively low low from from November through tudinal the CCS April. This This argues tudinalrange rangeof of the CCS studied studiedin in this thispaper paper(Figure (Figure4), 4), the the April. arguesagainst againstaa simple simplecoastal coastaltrapping trappingof of high high CZCS data indicate indicatethat that this this isis not pigment values which may be missed by the 100-km-wide CZCS data not the the case case over over the the pigment values which may be missed by the 100-km-wide latitudinal range range of of the the PCS averaging boxes. boxes. In In the latitudinal PCS (Figure (Figure 5). 5). The The lowest lowest and and averaging the northern northernregion regionoff off Peru Peru(equator(equatorhighest latitude regions ward of where pigment concentrations are are high highestlatitude regionsof of the the PCS PCS have havemaximum maximumpigment pigment ward of 18°S), 18øS), where pigment concentrations high in in concentrations during summer, summer, aa pattern pattern similar to that that in winter, spring, in fall fall concentrationsduring similar to in winter, spring, and and summer summer with with aa distinct distinctminimum minimum in the ofofthe (AprilJune), the the CCS. CCS. The Themiddle middlelatitude latituderegions regions thePCS PCS(17° (17øto to45°S) 45øS) (April-June), the winds winds are, are,again, again,upwelling upwellingfavorable favorable exhibit a mean out of of phase phase with throughout the the year. exhibit a mean seasonal seasonalcycle cycle 180° 180ø out with this. this. throughout year. A A wider wider analysis analysisof ofoceanographic oceanographic Plate variables in in this shows that that high Plate 22 and and Figures Figures33 and and55 show showthat thatthis thiscycle cycleis isprimarily primarily variables this region region by by Bakun Bakun [1987] [1987] shows high aa result pigment concentrations concentrations in in spring resultof of relatively relativelyhigh highpigment pigmentconcentrations concentrationsalong along aa pigment spring (Figure (Figure 5) 5) are are associated associated THOMAS ET AL.' AL.: PERU CURRENT PIGMENT VARIABILITY THOMAS ET PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT VARIABILITY 7367 with the with the climatological climatologicalseasonal seasonalmaximum maximumin in wind wind mixing, mixing, alongshore wind stress, stress, and alongshore wind and offshore offshore Ekman Ekman transport. transport. These data These data do do not not suggest suggestan an explanation explanationfor for the the pigment pigment cently CZCS data data confirm confirm the the conclusions conclusions of of cently released releasedglobal globalCZCS minimum minimum in fall. fall. cycle in in the the Southern Southern California California Bight. Bight. The The present present analysis analysis cycle extends this pattern to show a summer maximum off extends this pattern to show a summer maximum off Baja Baja California. This seasonal seasonal evolution evolution is is consistent California. This consistent with with the the hypothesis hypothesisthat that wind-driven wind-drivencoastal coastaland andoffshore offshoreupwellings upwellings are are responsible responsiblefor for the the summer summermaxima maxima in in surface surfacepigment pigment concentrations. concentrations. The The present presentanalysis analysisof of the the global global data data set setalso alsoincludes includes the for the the first the winter winter months months for first time. time. These These data data show show an an offshore extension of of pigment offshore extension pigment concentrations concentrationsgreater greater than than 0.5-1.0 mg pattern off central and 0.5-1.0 mgm m-3 in in aa diffuse diffusepattern offcentral andnorthern northern California and extending extending north. north. This This offshore offshore extension extension of of California and higher pigment concentrations creates a winter maximum in higherpigmentconcentrationscreatesa winter maximum in the seasonal km offshore at the seasonalcycle cycle for for regions regions more more than than 100 100 km offshore at these higher latitudes. A these higher latitudes. A winter winter maximum maximum is is consistent consistentwith with that seen and 40°N in an an that seenby by Lewis Lewis et et al. al.[1988] [1988]between between250 25ø and 40øN in analysis of of seasonal seasonal cycles cycles evident evident in in Secchi analysis Secchi disk disk depths depths Comparisons between the the time Comparisons between time series seriesof of concentration concentration anomalies in the the CCS anomaliesin CCS and and the the PCS PCS (Figures (Figures66 and and 7) 7) allow allow some general statements statements of somegeneral of differences differencesand and similarities. similarities.Most Most significantly, visual inspection inspection indicates indicates that significantly, visual that interannual interannual anomalies of pigment pigment concentration concentrationininthe the PCS PCS are anomalies of are not not correlated with with those correlated those in in the the CCS. CCS. Unlike Unlike the the CCS, CCS, interaninterannual anomalies anomalies in in the the four four regions regions along along the the latitudinal latitudinal extent extent nual of the the PCS PCS do do not not have have aa common commoncharacteristic. characteristic. in In addition, addition, of most of of the in the the PCS most the strongest strongest relative relative anomalies anomalies in PCS are are positive, positive,whereas whereasstrong, strong,and andoften oftenthe thestrongest, strongest,anomalies anomalies in all all regions regions of of the the CCS CCS are are negative. in the the highestin negative.Only Only in highestlatitude region of southern Chile and in the 100latitude region of southernChile and in the 100-to to 200-km200-kmoffshore zone zone off Peru do offshore off Peru do relatively relatively strong strongnegative negativeanomaanomalies occur. occur. lies A feature in in these these EBCs is present A common common feature EBCs is present in in the the PeruPeru- Strub et et al. there is in most Strub al. [1990]: [1990]' there is aa summer summer maximum maximum in most locations closest to to the the coast, locations closest coast, with with aa very very weak weak seasonal seasonal vian coast coast region. vian region. The The year year 1983 1983 appears appears as as aa negative negative along 180°W in the central North Pacific. It is also consistent along 180øWin the central North Pacific. It is also consistent anomaly (Figure (Figure 7), 7), especially especially strong offshore zones, zones, as anomaly strongin in offshore asitit does throughout throughout the does the CCS. CCS. Anomalous Anomalousoceanographic oceanographiccondiconditions with aa major tions associated associatedwith major El E1Niflo Nifio event event were were present present throughout the the CCS throughout CCS in in1982-1983 1982-1983 [e.g., [e.g., Simpson, Simpson,1983; 1983; Huyer Huyer and and Smith, Smith, 1985; 1985;Rienecker Rienecker and andMooers, Mooers,1986]. 1986]. Although the strongest strongest anomalies in sea sea surface surface temperature temperature Althoughthe anomaliesin (SST) along the South American coast associated with this this (SST) along the South American coast associatedwith El Niflo occurred north E1 Nifio occurred northof of approximately approximately20°S 20øS[Cane, [Cane, 1983; 1983; Rasmusson Rasmusson and and Wallace, Wallace, 1983], 1983], positive positive SST SST anomalies anomalies occurred at least occurred at least as as far far south southas as40°S 40øS[Rasmusson [Rasmusson and and Wallace, Results presented presented here Wallace, 1983]. 1983]. Results here (Figure (Figure 6) 6) agree agreewith with with large large sedimentation sedimentation rates rates recorded recorded 100-300 km offshore offshore with 100-300 km of Cape Cape Blanco in winter winter and and spring [Sancetta, of Blanco (43°N) (43øN) in spring 1988 1988 [Sancetta, 1992],which whichwere wereattributed attributed to to plankton plankton productivity productivity char1992], characteristic of of the the coastal acteristic coastal regions. regions. Both Both the the CZCS CZCS seasonal seasonal distribution and these offshore distribution and its its magnitude magnitude in in these offshoreregions regionsagree agree with chlorophyll chlorophyllconcentrations concentrationsgreater greaterthan than1.0 1.0mg mgmm3 in with -3 in FebruaryMay February-May reported reportedby byLandry Landryetetal. al.[1989] [1989] for for distances distances more than than 50 km from from the the coast coast of and greater greater more 50 km of Washington Washington and than 0.51 mg mg m m3 reported than -3 ininNovemberDecember November-December reportedby by Thomas and Emery for the the outer Thomas and Emery [1986] [1986] for outer southern southernBritish British Columbia continental shelf. Although the cause of the diffuse previous previous work work [Fiedler, [Fiedler, 1984; 1984; Thomas Thomas and and Strub, Strub,1990; 1990; Columbia continental shelf. Although the causeof the diffuse offshore maxima maxima in in winter winter remains remains to to be be determined, determined, winter offshore winter Strub Strub et et al., al., 1990] 1990]showing showingthat that pigment pigmentconcentrations concentrationsin in surface nutrient concentrations presented by Landry et surface nutrient concentrations presented by Landry et al. al. the CCS during 1983, especially in regions more than 50 km the CCS during 1983, especiallyin regionsmore than 50 km [1989] are high enough to support phytoplankton growth, [1989] are high enough to support phytoplankton growth, offshore, were anomalously low. In a time series of transects offshore,were anomalouslylow. In a time seriesof transects and previous previous analysis analysis by by Thomas indicates Thomas and and Strub Strub [1989] [1989] indicates off off the the Peruvian Peruvian coast, coast, Barber Barber and andChavez Chavez [1983] [1983] show show aa and similar response response to to the the El similar E1Niño, Nifio, with with maximum maximumreduction reductionin in chlorophyll concentrations and chlorophyllconcentrations andlongest longesttemporal temporalpersistence persistence of of this this condition condition in in areas areas more more than than 50 50 km km offshore. offshore. In In the the three Niflo year threehighest highestlatitude latituderegions regionsof of the the PCS, PCS, the the El E1Nifio year of of 1983 is slightly slightly positive positive and and does does not not appear 1983 is appear anomalous anomalousin in the 8-year time series the 8-year time series (for (for both both the the winter winter and and summer summer anomalies). This This result result must must be be treated treated with with caution caution because because anomalies). of the sparse data data over over the the PCS of the relatively relatively sparse PCS and and also also the the smoothing effect of of the the temporal temporal mean. smoothingeffect mean. The The 8-month 8-month mean mean for (September 1982 for 1983 1983(September 1982to to April April 1983) 1983)is is heavily heavilyweighted weighted toward September because of missing data, and this this is toward Septemberbecauseof missingdata, and is early early in the the E1 El Nifio Niño episode, episode, especially at higher higher latitudes latitudes [Barber [Barber in especiallyat and Chavez, Cane, 1983]. This is is not not true, true, however, and Chavez, 1983; 1983; Cane, 1983]. This however, for the the (austral) (austral) winter for winter 8-month 8-month mean mean anomaly anomaly also alsoshown shownin in Figure 7, where Figure 7, where the the entire entire period period MarchOctober March-Octoberisisrelarelatively well well represented. represented. This This period, period, however, tively however, is is obviously obviously weighted by by months weighted monthsafter after the themaximum maximumEl E1Niflo Nifio anomalies anomalies when near-surface when near-surfacenutrients nutrientsand and chlorophyll chlorophyll concentrations concentrations were recovering were recovering to to normality normality [Barber [Barber and and Chavez, Chavez, 1983]. 1983]. Thus it Thus it seems seemslikely likely that that the the CZCS CZCS missed missedthe the maximum maximum amplitude of the 1982-1983 El Niho signal signal at at higher higher latitudes latitudes amplitude of the 1982-1983E1Nifio in the in the PCS. PCS. Summary Summary and and Conclusions Conclusions In the California Current the seasonal In the California Current system, system, the seasonalcycles cycles of of surface surface pigment pigment concentrations concentrationsrecalculated recalculatedusing usingthe therere5. $. that light may not not be be limiting, limiting, at at least least in in late late winter, winter, at at these that light may these latitudes. latitudes. The The analysis analysis of of interannual interannual variability variability shows shows aa common common feature throughout throughoutthe the CCS CCS to to be feature be low low values values in in 1983, 1983, following the the strong strong 1982-1983 El Nifio. Niflo. These These low following 1982-1983 E1 low values values persist through off Baja Baja California California but but are are followed persist through1984 1984 off followed by by the major the major peak peak in in the the 8-year 8-year record record off off the the Pacific Pacific NorthNorthwest. west. Although lack lack of of data data for for the the Peru Peru Current Current system Although systemsomewhat somewhat reduces our our confidence in the the results results of of the reduces confidence in the analysis, analysis, the the available CZCS CZCS data data do do suggest suggest fundamental fundamental differences differences in in the the available seasonal cyclesof of the the two two EBCs seasonalcycles EBCs and and point point to to important important oceanographicquestions questionswhich whichneed need to to be oceanographic be addressed addressedwith with future data data sets. a future sets. The The seasonal seasonalcycle cycle south south of of 45°S 45øS shows shows a summer maximum, which which may may be be consistent summer maximum, consistentwith with wind wind forcforcing, providing one one of of the the few few similarities to the the CCS. ing, providing similaritiesto CCS. Between Between 17° and 45øS, 45°S, however, however, maximum maximum values values occur occur in 17 ø and in austral austral winter, despite the fact winds are are either winter, despite the fact that thatclimatological climatologicalwinds either persistently favorable (off northern Chile) persistentlyupwelling upwelling favorable (off northern Chile) or or cycle cycle between conditions favorable (austral (austral summer) and unfavorbetween conditions favorable summer) and unfavorable central Chile able (austral (australwinter) winter) to to wind-driven wind-drivenupwelling upwelling(off (off central Chile between and 40øS) 40°S) [Bakun [Bakun and and Nelson, Nelson, 1991]. northern between 30° 30ø and 1991]. Off Off northern Chile, even within 20-40 km km of of the Chile, even within 20-40 the coast, coast,relatively relatively low low pigment concentrationsare arepresent presentfor for much much of of the the year pigment concentrations year despite winds. Off Off Peru Peru between between despitepersistent persistentupwelling-favorable upwelling-favorablewinds. 6° and 17°S, there is is aa minimum in fall, fall, despite despite the the fact fact that that 6ø and 17øS,there minimum in winds favorable (the winds are are persistently persistentlyupwelling upwellingfavorable (the lack lack of of data datain in 7368 THOMAS ET CURRENT PIGMENT VARIABILITY THOMAS ET AL.: AL.' PERU PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT VARIABILITY winter scatterometersshould shouldclarify clarify the therelationships relationshipsglimpsed glimpsedonly only winter may may mask mask what what is is really really aa broad broadwinter, winter,spring, spring,and and scatterometers partially and imperfectly imperfectly with with the the CZCS summer maximum). partially and CZCS data data set. set. summer maximum). The The analysis analysisof of interannual interannualvariability variability off off South SouthAmerica America shows a decrease in relative pigment concentration shows a decrease in relative pigment concentration in in offoffAppendix shore shore zones zones off off the the Peru Peru coast coast(lowest (lowest latitudes) latitudes) after after the the Appendix The pattern of 1982-1983 El Nifio. Niño. This This parallels The diffuse diffuse pattern of surface surfacepigment pigment concentration concentrationvalval1982-1983 E1 parallels observations observationswithin within the the ues, shown stretching far offshore in Plate andFigure Figure4, 4, is is CCS which show CCS during during1983 1983which showthat thatmaximum maximumdifferences differencesdue due ues, shown stretching far offshore in Plate 11and similar to to the the pattern pattern caused caused by by the the single singleRayleigh Rayleigh scatterscatterto were offshore and that that very very nearshore nearshore similar to El E1Niflo Nifio conditions conditions were offshore and atmospheric correction correction used ing atmospheric used in in previous previous versions versionsof of the the pigment pigment concentrations concentrationsremained remainedhigh high[Fiedler, [Fiedler,1984; 1984;ThoTho- ing CZCS data, data, although although much much lower lower in in magnitude magnitude [Strub [Strub et et al., al., mas mas and and Strub, Strub,1990]. 1990].This This pattern patternis isalso alsoin in agreement agreementwith with CZCS 1990] (hereinafterreferred referred to to as as SJTA). 1990] (hereinafter SJTA). Although Although the the global global in in situ situ observations observations from from aa cross-shelf cross-shelf transect transect at at 5°S 5øSprepreCZCS data set uses the multiple Rayleigh scattering algoCZCS data set uses the multiple Rayleigh scattering algosented by Barber and Chile, the the historical historical sented by Barber andChavez Chavez[1983]. [1983]. Off Off Chile, rithm [Gordon [Gordon et etal., 1988; Feldman et al., 1989], itit is natural rithm al., 1988; Feldman et al., 1989], is natural CZCS data are too sparse to document interannual variabilCZCS data are too sparseto documentinterannualvariabilto wonder wonder how how much much systematic systematic error error may still be be caused caused by by to may still ity ity with with any any confidence. confidence. incompletely corrected atmospheric scattering at high sunThe The most most robust robust result result of of our ouranalysis analysis concerns concerns the the incompletely corrected atmospheric scattering at high sunsatellite angles (high latitude, winter). seasonal cycle where the seasonal cycle off off Chile Chile between between 30° 30ø and and 40°S, 40øS, where the satellite angles (high latitude, winter). SJTA developed developed aa technique for estimating this error SJTA technique for estimating this error spring-summer data data coverage coverage is is best best and and where where the the seasonal seasonal spring-summer empirically for for the the previous previous data. data. That That technique empirically techniqueis is repeated repeated increase in upwelling-favorable winds is greatest during the increasein upwelling-favorablewinds is greatestduring the here. The assumptions of the technique are that (1) the error here. The assumptionsof the technique are that (1) the error same period same period (OctoberMarch). (October-March). Here, Here, pigment pigmentconcentraconcentrawith latitude; latitude; (2) (2) the the error increases monotonically monotonically with error should should tions in the 100 km next to the coast only briefly exceed 1.5 increases tions in the 100 km next to the coast only briefly exceed 1.5 have the at times have the greatest greatest magnitude magnitude at times around around the the winter winter mg m m3, below This mg -3 dropping dropping below1.0 1.0 mg mg m3 m-3ininsummer. summer. This solstice; and (3) winter values of pigment concentration far solstice; and (3) winter values of pigment concentration far contrasts to values contrasts valuesgreater greaterthan than3.0-4.0 3.0-4.0mg mgm m-3 during during offshore are uniformly low, without a monotonic increase offshore are uniformly low, without a monotonic increase spring and summer km of of the the coast coast in the region springand summerwithin within 100 100 km in the region of of seasonality seasonalityin in upwelling-favorable upwelling-favorablewinds windsoff off California. California. Thus forces Thus forces other other than thanlarge-scale large-scaleseasonal seasonalwind wind forcing forcing appear appear to to dominate dominate the the seasonal seasonalcycles cyclesof of surface surfacepigment pigment with with higher higher latitude. latitude. The technique uses the The technique simply simply uses the CZCS CZCS pigment pigment concentraconcentra- tion in aa north-south tion values values in north-south band band far far from from the the coast coast in in an an empirical orthogonal function function (EOF) (EOF) analysis empirical orthogonal analysis and and examines examines concentration concentration over over large large portions portions of of the the PCS. PCS. the the space spaceand and time time series seriesfor for the the properties propertieslisted listed as asassumpassumpThe significance The significanceof of this this difference differencegoes goesbeyond beyondthe thephytophyto- tions tions 1 1 and and 2 above. above. Since Since SJTA SJTA used used data data from from the the West West plankton response response to plankton to local local wind wind forcing. forcing. It It raises raisesthe the question question Coast from the the coast, Coast time time series, series, which which extended extended only only 5° 5ø from coast, of whether the of whether the surface surfacesignature signatureof of the thephytoplankton phytoplanktonstanding standing they band separated separated from they formed formed aa 2.5°-wide 2.5ø-wide band from the the coast coastby by stock can be be used of 2.5° stock (as (as seen seenfrom from aa satellite) satellite) can used as as an an indicator indicator of of longitude. longitude. The The first first EOF EOF accounted 2.5ø of accountedfor for 71% 71% of of the the the of the the coastal ocean system system at at the the overall overall productivity productivity of coastal ocean the variance, variance, with a temporal temporal function with a function that that had hadmaximum maximum values values higher trophic levels. levels. The The productivity the coastal ocean off highertrophic productivityof of the coastalocean off of of 3.0 3.0 in in December December and and January January and andapproximately approximately 0.0 0.0 in in Chile, fish Chile, as as measured measuredby by the theharvest harvestof ofphytoplankton-eating phytoplankton-eating fish June Juneand and July. July. The The spatial spatialfunction function ramped ramped up up from from approxapproxspecies species(anchovies (anchoviesand and sardines), sardines),is is reportedly reportedly55 to to20 20times times imately imately 0 0 at at 25°N 25øN to to 1.3 1.3 between between40° 40øand and 50°N. 50øN. Thus Thus errors errors greater than occurred in in winter greater than that that off off California California (P. (P. Smith Smith and and R. R. Parrish, Parrish, of 4.0 of approximately approximately 4.0mg mgm m-3 occurred winternorth northof of personal communications, The harvest harvest of predator fish personal communications,1992). 1992). The of predator fish 40°N. 40øN. species jack mackerel Although the global global data data set set is not restricted speciessuch suchas asjack mackerelis is also alsolarge. large. Why, Why, then, then, do do the the Although the is not restrictedto to the theregion region satellite satellitecolor color data data suggest suggestmuch much lower lower surface surfaceconcentrations concentrations near near the the coast, coast, there there are are more more data data for for that that region. region. For For this this of of chlorophyll chlorophyll (and (and phytoplankton)? phytoplankton)?Are Are the thephytoplankton phytoplankton reason reason and and in in order order to to facilitate facilitate aa comparison comparison to to the the results results located can see? see? of data set located deeper deeper in in the the water water column column than than the the satellite satellite can of SJTA, SJTA, the the technique technique was was applied applied to to the the global global data set in in Is that higher concentrations of phytoplankton exist exist two bands: one one based Is it it possible possiblethat higher concentrations of phytoplankton two latitudinal latitudinal bands: based on on 1° 1ølatitude latitude by by 2.5° 2.5ø in to resolve blocks separated in aa very very narrow narrow nearshore nearshoreregion regiondifficult difficult to resolve in in this this longitude longitude blocks separatedfrom from the thecoast coastby by2.5° 2.5øofoflongilongirelatively relatively low low spatial spatialresolution resolutiondata dataset? set?Do Do very veryhigh highprimary primary productivity rates maintain the fishery, with biomass so productivity rates maintain the fishery, with biomassbeing being so heavily grazed that heavily grazed that higher higherconcentrations concentrationseither eitherdo donot notbuild build up up or or do do not not spread spreadoffshore? offshore?These These are are questions questionsthat that the the CZCS CZCS data data set setcan canonly only pose. pose.Answers Answersrely rely on onthe thecombination combinationof of future data and and more color data. data. future field field data more reliable reliable satellite satellite color Thus, the relative relative lack lack of Thus, acknowledging acknowledging the of data, data, the the time time series of series of CZCS CZCS data data suggest suggestthat that there thereare arefundamental fundamental differences in the differences in the seasonal seasonalvariabilities variabilitiesof of surface surfacepigment pigment concentrations concentrations in in the the Peru Peru and and California California Current Current systems, systems, even wind forcing is similar. even though though large-scale large-scale wind forcing is similar. A A more more complete understanding understanding of complete of this this variability variability and and its its relation relation to to wind wind forcing, forcing, upwelling, upwelling, and and horizontal horizontal transport transport within within these current a more these current systems systems will will be be possible possible only only when when a more complete data data set complete set becomes becomesavailable. available. As As always, always, the the need need for in for in situ situ observations observations over over aa wide wide range range of of spatial spatial and and temporal scales is tantamount. tantamount. In temporal scalesis In the the future, future, combinations combinationsof of pigment data data from pigment from SeaWiFS SeaWiFS (and (and other) other) sensors, sensors,current current data TOPEX and and other data from from TOPEX other altimeters, altimeters, and and wind wind data data from from tude, tude, and and another another based based on on1° 1ølatitude latitude by by3.0° 3.0ølongitude longitude blocks separated from the coast by 5.0° of longitude. blocks separatedfrom the coast by 5.0ø of longitude.The The data data are are from from the the first first 4 4 years years of of the the global global data data set set (1979-1982), whendata data coverage coverage was was better. better. The (1979-1982), when The results results are are shown in Figures Figures Ala Ala and shown in and Alb, Alb, respectively. respectively. The The average average seasonal seasonaltime time series seriesof of each eachfunction function show show low values (0.1-0.2) (0.1-0.2) in in spring spring and and summer summer and values low values and higher higher values (0.5-0.9) in fall (0.5-0.9) in fall and and winter. winter. The The spatial spatialfunctions functions show show localized maxima superimposed superimposed on on aa gradual gradual increase increase with localized maxima with latitude which isis more latitude which more evident evident farther farther from from shore. shore. An An identical dashed curve curve based identical dashed based on on the the increase increase in in minimum minimum values from the the more band (Figure Aib) values from more offshore offshore band (Figure A lb) has has been been drawn by hand in both The fact fact that drawn by hand in both figures. figures. The that it it fits fits the the minimum values in in Figure Figure A Ala that minimum values 1a gives gives us us some someconfidence confidencethat it it describes describesan an upper upper estimate estimate of of the the bias bias caused causedby by uncoruncorrected rected atmospheric atmospheric scattering scattering (or (or other other systematic systematicoptical optical effects). Using the the more band, where effects). Using more offshore offshore band, where we we expect expect lower lower true true pigment pigment values, values, the the maximum maximum value value for for the the averaged and occurs averagedseasonal seasonaltime time series seriesis is 0.4 0.4 and occursin in November November THOMAS ET AL.' AL.: PERU THOMAS ET PERU AND AND CALIFORNIA CALIFORNIA CURRENT CURRENT PIGMENT PIGMENT VARIABILITY VARIABILITY 7369 3.5 ;5.5 ! 3.0 3.0 1.0 1.0 '2.5 0.8 10.6 2.0 0.6 0.6 1.5 1.0 A A 0.4 0.4 0.2 0.2 0.5 0.5 0.0. 0.0 ....... , I I• Jon Feb Mar Apr Apr MoyJun May Jun Jul Jon FebMot JulAug AugSep SepOct OctNov NovDec Dec Time .rea Time sir;el bL2 l.2 .-.." ....... 20 20 30 50 40 40 / 3.5 3.0 1.0 1.0 50 50 LoUtud. Latitude('N) 2.5 ,•' 2.5 •, o.s 110.8 ,..,,• 2.0 0.6 l.5 1.5 8 0.4 I.0 0.2 0.2 0.5 0.5 0.0 0.0 00 JonFebMorAprMoyJunJulAugSepOctNovD.c JonFebMotAprMoyJunJul AUgSepOctNovDec Tim. Time seriI sir;el ......... 20 20 I ......... I 30 40 •0 40 Lotituds ("N) ('N) Lot;tude .... 50 50 Figure AI. Al. Latitudinal Latitudinal dependence dependence and in Figure and time time series seriesfor for monthly monthlypigment pigmentconcentrations concentrations in two twooffshore offshore zones of of the orthogonal function function (EOF) (EOF) decompositions. decompositions. (a) (a) zones the CCS CCS shown shownas asthe thefirst firstmodes modesof of empirical empiricalorthogonal The first first EOF EOF mode variability, accounting accounting for for 63.7% 63.7% of of the the total total variance, variance, of The modeof of pigment pigmentvariability, of aa zone zone2.5° 2.5øof of longitude wide wide and and 2.5 2.5°ø(approximately (approximately 300 300km) km)away awayfrom fromthe thecoast coast of of North North America. America. (b) longitude (b) The The first firstEOF EOF mode of of pigment pigment variability, variability, accounting accounting for for 62.9% 62.9% of of the the total total variance, variance, of mode of aa zone zone3.0° 3.0øof of longitude longitudewide wide and (approximately 600 600 km) km)away awayfrom from the the coast coast of of North North America. and 5° 5ø (approximately America. through January. The through January. The spatial spatialfunction function increases increasesto to 0.25 0.25 at at approximately and 0.75 which would approximately 38°N 38øN and 0.75 at at 48°N, 48øN, which would indicate indicate maximum errors ofof0.0.1-0.4 1-.0.4 mg maximum errorsin inNovember-January November-January mgm m-3 at at these latitudes. these latitudes. Whether these maxima are aa result Whetherthese maximaare resultof of advection advectionof of pigment pigment or byproducts from from closer closer to to the the coast coast or or represent represent or pigment pigmentbyproducts local production is local production is not not known. known. Others may draw their Others may draw their own own curves curvesand and conclusions, conclusions,but but Acknowledgments.This This work work was was supported supportedby by NASA NASA grant grant we estimate estimate any any error error caused caused by by these these systematic systematic processes processes Acknowledgments. we NAGW-2475. PeterBugden Bugdenand andSerge SergeL6vesque Lvesque assisted with NAGW-2475. Peter assisted with to be less than 0.1-0.4 mg m in winter winter at at latitudes latitudes between between aspects tobelessthan0.1-0.4mgm-3 in aspectsof of the the plotting plottingand andimage imageprocessing. processing.We We thank thankSatlantic Satlantic 40° and 50øN 50°Nand and negligible negligibleatatother other times times and and latitudes latitudes less less Inc. 40 ø and Inc. of of Halifax, Halifax, Nova Nova Scotia, Scotia, for for access accessto to their their archive archive of of CZCS CZCS global monthly monthly composites. composites. than 40°N. This is is supported supported by by data data from from the the Peru Peru Current Current global than 40øN. This system (Figure 5), 5),where wherevalues valuesremain remainless lessthan than0.5 0.5mg system (Figure mgmm3 -3 in winter winter at latitudes of in the the band band farthest farthest from of 45° 45 ø to 51°S 51 øS in from the coast. the coast. References References The The analysis analysisalso also provides provides another anotherlook look at at the the latitudinal latitudinal Abbott, Abbott, M. patterns M. R., R., and andB. B. Barksdale, Barksdale,Phytoplankton Phytoplanktonpigment pigmentpatterns structure structure of of the the offshore offshorewinter winter maximum maximum indicated indicated by by Figure 4. 4. Approximately 300 km km from from the the coast Figure Approximately 300 coast(Figure (Figure Ala), there Ala), thereare arelocalized localizedmaxima maxima at at32°, 32ø,37°, 37ø, 44-46°, 4a. •.6ø, and and 48°N. The The individual individual monthly monthly time time series series of of the the EOFs EOFs show 48øN. show that these were than in winter that were weaker weaker in in winter winter of of 1979 1979 than winter of and wind off central central California, California, J. J. Geophys. and wind forcing forcing off Geophys.Res., Res., 96, 96, 14,649-14,667, 1991. 1991. Abbott1M. M. R., R., and and P.M. P. M. Zion, of Abbott, Zion, Spatial Spatialand andtemporal temporalvariability variabilityof phytoplankton pigment pigment off off northern phytoplankton northernCalifornia California during duringCoastal Coastal Ocean Dynamics Dynamics Experiment Experiment 1, J. J. Geophys. Res., 92, Ocean Geophys.Res., 92, 1745-1756, 1745-1756, 1987. 1987. 1980-1982.Approximately Approximately600 600km kmfrom fromthe the coast coast (Figure A., Monthly in the the ocean Bakun, A., Monthly variability variability in ocean habitat habitat off off Peru Peru as as 1980-1982. (Figure Bakun, deduced from maritime observations, 1953-1984, in The Peru deduced from maritime observations, 1953-1984, in The Peru Aib), a single maximum is centered on 42°N. It was stronger Alb), a singlemaximumis centeredon 42øN.It was stronger Anchoveta and Its Anchoveta and Its Upwelling Upwelling Ecosystem: Ecosystem:Three Three Decades Decades of of in 1981 and 1982 1982than than in in 1979 1979and and 1980, 1980,with with an an exceptionally exceptionally in 1981and Change, edited by by D. Change, edited D. Paul Paul and and I.I. Tsukayama, Tsukayama,pp. pp.46-74, 46-74, high value value in in November November 1982, creating the the peak peak in high 1982,creating in NovemNovemICLARM Studies Studies and 1987. ICLARM andReviews, Reviews,Manila, Manila,Philippines, Philippines, 1987. ber A., and cycle ber in in the the mean meanseasonal seasonaltime time series seriesshown shownin in Figure Figure Alb. A lb. Bakun, Bakun,A., andC. C. S. S. Nelson, Nelson,The The seasonal seasonal cycleof of wind wind stress stresscurl curl 7370 THOMAS ET CURRENT VARIABILITY THOMAS ET AL.: AL.: PERU PERU AND AND CALIFORNIA CALIFORNIA CURRENT PIGMENT PIGMENT VARIABILITY in in sub-tropical sub-tropical eastern easternboundary boundarycurrent currentregions, regions,J. J. Phys. Phys.OceanOceanogr.,21, ogr., 21,1815-1834, 1815-1834,1991. 1991. Barber, Barber, R. R. T., T., and andF. F.P.P.Chavez, Chavez,Biological Biologicalconsequences consequencesof of El El Niflo, Nifio, Science, Science, 222, 222, 1203-1210, 1203-1210, 1983. 1983. Barber, Barber, R. R. T., T., and andF. F. P.P.Chavez, Chavez,Ocean Oceanvariability variability in in relation relation to to living resources during El Niflo, Nature, 319, living resources during the the 1982-83 1982-83 El Nifio, Nature, 319, 279279285, 285, 1986. 1986. Cane, Cane, M. M. A., A., Oceanographic Oceanographicevents eventsduring duringEl El Niflo, Nifio, Science, Science,222, 222, 1189-1194, 1189-1194, 1983. 1983. Espinoza, F. R., Espinoza, F. R., A. A. Maoxiang, Maoxiang, and and S. S. Neshyba, Neshyba, Surface Surface water water motion Chile revealed revealed in in satellite satellite images images of of surface surface chlorophyll chlorophyll motion off off Chile and and temperatures, temperatures, in in Proceedings, Proceedings,International International Conference Conferenceon on Marine Resources of of the the Pacific, edited by by P. M. Arana, Marine Resources Pacific, edited P.M. Arana, pp. pp. 41-57, 41-57, Universidad Universidad Catolica Catolica de de Valparaiso, Valparaiso, Valparaiso, Valparaiso, Chile, Chile, 1983. 1983. Feldman, G., N. J. Elrod, Feldman, G., N. Kuring, Kuring, C. C. Ng, Ng, W. W. Esaias, Esaias,C. C. McClain, McClain, J. Elrod, N. M. Carle, Cane, N. Maynard, Maynard, D. D. Endres, Endres, R. R. Evans, Evans, J. J.Brown, Brown, S. S. Walsh, Walsh, M. and data set, and 0. G. Podesta, Podesta,Ocean Oceancolor: color:Availability Availability of of the the global global data set, Eos Trans. Eos Trans. AGU, AGU, 70, 70, 634-641, 634-641, 1989. 1989. Fiedler, ofofthe along Fiedler, P. P. C., C., Satellite Satelliteobservations observations the1982-1983 1982-1983El El Niflo Nifio along the the U.S. U.S. Pacific Pacific coast, coast,Science, Science,224, 224,1251-1254, 1251-1254,1984. 1984. Gordon, H. R., Gordon, H. R., and and D. D. J. J. Castano, Castano, The The coastal coastal zone zone color color scanner scanner atmospheric atmospheric correction correction algorithm: algorithm: Multiple Multiple scattering scatteringeffects, effects, App!. Opt., Appl. Opt., 26, 26, 2111-2122, 2111-2122, 1987. 1987. Gordon, H. R., R., D. Gordon, H. D. K. K. Clark, Clark, J. J. W. W. Brown, Brown, 0. O. B. B. Brown, Brown, R. R. H. H. Evans, Evans, and andW. W. W. W. Broenkow, Broenkow, Phytoplankton Phytoplanktonpigment pigmentconcentraconcentrations tions in in the the Middle Middle Atlantic Atlantic Bight: Bight: Comparisons Comparisonsof of ship ship determideterminations nations and and CZCS CZCS estimates, estimates, App!. Appl. Opt., Opt., 22, 22, 20-36, 20-36, l983a. 1983a. Gordon, W. Brown, Brown, O. 0. B. Gordon, H. H. R., R., .1. J. W. B. Brown, Brown, R. R. H. H. Evans, Evans, and and D. D. K. K. Clark, Clark, Nimbus Nimbus 77 CZCS: CZCS: Reduction Reduction of of its itsradiometric radiometricsensitivity sensitivity with 1, 1983b. with time, time, App!. Appl. Opt., Opt.,22, 22,3929-393 3929-3931, 1983b. Gordon, H. R., Gordon, H. R., J. J. W. W. Brown, Brown, and andR. R. H. H. Evans, Evans,Exact ExactRayleigh Rayleigh scattering for use scattering calculations calculations for use with with the the Nimbus-7 Nimbus-7 coastal coastal zone zone color color scanner, scanner,App!. Appl. Opt., Opt.,27, 27,862-871, 862-871,1988. 1988. Halpern, D., W. W. Knauss, 0. Brown, of Halpern, D., Knauss, O. Brown, and and F. F. Wentz, Wentz, An An atlas atlas of monthly monthly mean mean distributions distributions of of SSMI SSMI surface surfacewind wind speed, speed,Argos Argos buoy buoy drift, drift, AVHRR/2 AVHRR/2 sea sea surface surfacetemperature, temperature,and andECMWF ECMWF surface wind JPL Pub!. pp., surface wind components componentsduring during1989, 1989, JPL Publ. 92-17, 92-17, 112 112 pp., Jet Jet Propul. Propul. Lab., Lab., Pasadena, Pasadena,Calif., Calif.,1992. 1992. Huyer, ofofEl Huyer, A. A. E., E., and andR. R.L. L.Smith, Smith,The Thesignature signature ElNiflo Nifio off off Oregon, Oregon, 1982-1983, J. Geophys. 1982-1983, J. Geophys. Res., Res., 90, 90,7133-7142, 7133-7142, 1985. 1985. Landry, M. R., Landry, M. R., J. J. R. R. Postel, Postel, W. W. K. K. Peterson, Peterson,and andJ.J.Newman, Newman, Broad-scale Broad-scaledistributional distributionalpatterns patternsof of hydrographic hydrographicvariables variableson on the the Washington/Oregon Washington/Oregonshelf, sheif,in in Coasta! CoastalOceanography Oceanographyof ofWashWashington ington and and Oregon, Oregon, edited edited by by M. M. R. R. Landry Landry and and B. B. M. M. Hickey, Hickey, pp. pp. 1-40, 1-40, Elsevier, Elsevier, New New York, York, 1989. 1989. Lewis, Lewis, M. M. R., R., N. N. Kuring, Kuring, and andC. C. Yentsch, Yentsch,Global Globalpatterns patternsof of ocean ocean transparency: for the the new transparency: Implications Implications for new production production of of the the open open ocean, ocean, J. J. Geophys. Geophys.Res., Res.,93, 93,6847-6856, 6847-6856,1988. 1988. Lynn, Lynn, R. R. J., J., and andJ.J.J.J.Simpson, Simpson,The TheCalifornia CaliforniaCurrent Currentsystem: system:The The seasonal of its seasonalvariability variability of its physical physical characteristics, characteristics,J. J. Geophys. Geophys. Res., Res., 92, 92, 12,947-12,966, 12,947-12,966, 1987. 1987. Michaelsen, J., X. Michaelsen, J., X. Zhang, Zhang,and andR. R.C. C.Smith, Smith,Variability Variabilityof ofpigment pigment biomass in the biomass in the California California Current Current system system as as determined determined by by satellite 2, Temporal J. Geophys. Res., 93, satellite imagery, imagery, 2, Temporal variability, variability, J. Geophys. Res., 93, 10,883-10,8%, 1988. 10,883-10,896, 1988. Minas, H. J., Minas, H. J., M. M. Minas, Minas, and andT. T.T.T.Packard, Packard,Productivity Productivity in in upwelling areas deduced upwelling areas deducedfrom from hydrographic hydrographicand andchemical chemicalfields, fields, Limno!. Oceanogr., Limnol. Oceanogr.,31, 31,1182-1206, 1182-1206,1986. 1986. Mueller, Mueller, J. J. L., L., Nimbus Nimbus77CZCS: CZCS: Confirmations Confirmationsof of its itsradiometric radiometric sensitivity sensitivity decay decay rate ratethrough through1982, 1982,App!. Appl. Opt., Opt.,24, 24,1043-1047, 1043-1047, 1985. 1985. Pelaez, patterns in in Pelaez, I., J., and andJ. J.A. A. McGowan, McGowan, Phytoplankton Phytoplanktonpigment pigmentpatterns the California Current as as determined the California Current determinedby by satellite, satellite,Limno!. Limnol. OceanOceanogr., 927-950, 1986. ogr., 31, 31,927-950, 1986. Rasmusson, E. E. M., M., and aspects of of Rasmusson, and J. J. M. M. Wallace, Wallace, Meteorological Meteorological aspects the El the El Niflo/Southern Nifio/Southern Oscillation, Oscillation,Science, Science,222, 222,1195-1202, 1195-1202,1983. 1983. Rienecker, M. M. M., M., and Rienecker, and C. C. N. N. K. K. Mooers, Mooers, The The1982-1983 1982-1983 El El Nib Nifio signal off northern northern California, J. Geophys. signal off California, J. Geophys.Res., Res.,91, 91,6597-6608, 6597-6608, 1986. Sancetta, Sancetta,C., C., Comparison Comparisonof of phytoplankton phytoplanktonin in sediment sedimenttrap trap time time series series and and surface surfacesediments sedimentsalong along aa productivity productivity gradient, gradient, PaPa!eoceanography, leoceanography,7, 7, 183-194, 183-194,1992. 1992. Schulz, Schulz, S., S., A A comparison comparisonof ofprimary primaryproduction productionininupwefflng upwelling regions off off northwest northwest and and southwest southwest Africa, Afnca, Rapp. Rapp. P. regions P. V. V. Reun. Reun. Cons. Cons. Perm. Petra. mt. Int. Explor. Explor.Mer, Met, 180, 180,202-204, 202-204,1982. 1982. Simpson, J. J., Simpson, J. J., Large-scale Large-scalethermal thermal anomalies anomaliesin in the theCalifornia California Current the 1982-1983 El Nifio, Niflo, Geophys. Geophys. Res. Res. Lett., Left., 10, Current during during the 1982-1983 El 10, 937-940, 1983. 937-940, 1983. Smith, Smith, R. R. C., C., X. X. Zhang, Zhang,and andJ.J.Michaelsen, Michaelsen,Variability Variabilityof ofpigment pigment biomass in the biomass in the California California Current Current system systemas asdetermined determinedby by satellite imagery, 1, I, Spatial J. Geophys. satellite imagery, Spatial variability, variability, J. Geophys. Res., Res., 93, 93, 10,863-10,882, 10,863-10,882, 1988. 1988. Smith, R. L., Smith, R. L., A A comparison comparisonof of the the structure structureand andvariability variability of of the the flow field in in three three coastal regions: Oregon, flow field coastalupwelling upwelling regions: Oregon,northwest northwest Africa and Peru, Coastal Estuarine Africa and Peru, in in Coasta! CoastalUpwe!!ing, Upwelling, Coastal EstuarineSci., Sci., vol. 18, AGU, vol. 1, 1, edited editedby by F. F. A. A. Richards, Richards,pp. pp.107-1 107-118, AGU, Washington, Washington, D. C., 1981. D.C., 1981. Strub, P. T., Strub, P. T., J. J. S. S. Allen, Allen, A. A. Huyer, Huyer, R. R. L. L. Smith, Smith, and and R. R. C. C. Beardsley, cycles of of currents, currents, temperatures, Beardsley,Seasonal Seasonalcycles temperatures,winds, winds,and and sea sea level level over over the the northeast northeast Pacific Pacific continental continental shelf: sheif: 35°N 35'N to to 48°N, J. Geophys. 48'N, J. Geophys.Res., Res.,92, 92,1507-1526, 1507-1526,1987. 1987. Strub, Strub, P. P. T., T., C. C. James, James,A. A. C. C. Thomas, Thomas,and andM. M. R. R. Abbott, Abbott, Seasonal Seasonal and plankton concentrations, concentrations, hydrography, and nonseasonal nonseasonalplankton hydrography,and and satsatellite-measured sea surface surface temperature, temperature, J. J. Geophys. ellite-measured sea Geophys. Res., Res., 95, 95, 11,501-11,530, 11,501-11,530, 1990. 1990. Thomas, Thomas, A. A. C., C., and and W. W. J. J. Emery, Emery, Winter Winter hydrography hydrographyand andplankton plankton distribution on the distribution on the southern southernBritish British Columbia Columbia continental continental shelf, sheif, Can. Can. J. J. Fish. Fish.Aquat. Aquat.Sci., Sci.,43, 43,1249-1258, 1249-1258,1986. 1986. Thomas, A. C., Thomas, A. (2., and and P. P. T. T. Strub, Strub,Interannual Interannualvariability variability in in phytophytoplankton distribution during during the the spring plankton pigment pigment distribution springtransition transition along along the J. Geophys. the west west coast coast of of North North America, America, J. Geophys. Res., Res., 94, 94, 18,09518,09518,117, 1989. 1989. Thomas, Thomas, A. A. C., C., and and P. P. T. T. Strub, Strub, Seasonal Seasonaland and interannual interannualvariability variability of of pigment pigment concentrations concentrationsacross acrossaaCalifornia California Current Currentfrontal frontal zone, zone, J. J. Geophys. Geophys.Res., Res.,95, 95,13,023-13,042, 13,023-13,042,1990. 1990. Uribe, Uribe, T., T., and andS.S.Neshyba, Neshyba,Phytoplankton Phytoplanktonpigments pigmentsfrom from the the Nimbus-7 coastal zone zone color Nimbus-7 coastal color scanner: scanner: Coastal Coastal waters waters of of Chile Chile from to 40°S, in Proceedings, from 18° 18' to 40'S, in Proceedings, International International Conference Conferenceon ott Marine Resources of of the the Pacific, Pactfic, edited edited by by P. M. Arana, Marine Resources P.M. Arana, pp. pp. 33-40, 33-40, Universidad Universidad Catolica Catolica de de Valparaiso, Valparaiso, Valparaiso, Valparaiso, Chile, Chile, 1983. 1983. Weare, C., P. D. Samuel, Weare, B. B.C., P. T. T. Strub, Strub,and andM. M.D. Samuel,Marine Marine climate climate atlas atlas of the tropical tropical Pacific Ocean, Contrib. Pacific Ocean, Contrib. in in Atmos. Atmos. Sci., Sci., 20, 20, 147 147pp., pp., Univ. Univ. of of Calif., Calif., Davis, Davis, 1980. 1980. F. F. Huang Huangand andA. A. C. C. Thomas, Thomas,Atlantic AtlanticCenter Centerfor forRemote RemoteSensing Sensing of the Oceans, North Street, Suite 301, Halifax, Halifax, Nova Nova Scotia, Oceans, 6155 North Canada Canada B3K B3K 3R3. 3R3. C. C. James Jamesand and P. P. T. T. Strub, Strub,College Collegeof ofOceanic Oceanicand andAtmospheric Atmospheric Sciences, Sciences,Oregon Oregon State State University, University, Corvallis, Corvallis,OR OR97331. 97331. (Received (Received September September14, 14, 1992; 1992;revised revisedApril April 28, 28, 1993; 1993; accepted acceptedAugust August 4, 4, 1993.) 1993.)