Fisheries Applications of Satellite Datain the Eastern North Pacific

PAUL C. FIEDLER, GARY B. SMITH, and R. MICHAEL LAURS

Introduction

Remote sensing of the ocean isplaying an increasingly importantrole in fishery research and fishharvesting along the Pacific coast ofthe United States and Canada.Satellite sensors make synopticmeasurements of water temperatureand color, winds, ice cover, waveheight, and surface currents overlarge areas of the ocean surface.Variations in these ocean conditionsplay key roles in natural fluctua­tions of fish stocks and in theirvulnerability to harvesting.

The promise of remote sensingtechniques for fisheries research,management, and exploitation hasbeen recognized since the early1960's when the first visible and in­frared images of the earth's surfacewere obtained from orbit. However,successful applications have onlyrecently been realized with the ad­vent of advanced and sensitiveradiometers, high-speed data pro-

ABSTRACT-Satellite sensors pro­vide extensive and detailed images ofseasurface temperature and color. Synopticdaily sampling by satellites gives aunique view of the ocean surface thatcan be extremely useful when used inconjunction with conventional ship­board data. Current and potential ap­plications of satellite data off the U.S.Pacific coast and Alaska include inter­pretation of ship survey data, explana­tion of pelagic fisheries distributions,prediction of stock recruitment,spatial/temporal monitoring of thecoastal zone and sea ice, studies ofmigration routes and timing, and pro­duction of charts to aid commercialfisheries.

46(3)

cessing, and the availability of im­agery to fishery scientists andfishermen.

Gower (1982) provides a usefuloverview of the different kinds ofremote sensing data relevant tofisheries science and oceanography.Laurs and Brucks (In press) reviewliving marine resources applicationsin the United States. Yamanaka(1982) gives examples of some usesof satellite data for fisheries applica­tions off Japan. Other potential,ocean-related uses of remote sensingdata were discussed by Montgomery(1981). This paper will review thesatellite sensors currently measuringsea surface temperature and oceancolor, data processing and availa­bility, and several examples of re­cent and potential applications toeastern North Pacific fisheries.

Satellites provide a unique view ofthe ocean by covering large areassynoptically. Coverage of historicallydata-poor areas is particularly useful.However, satellite measurements areusually limited to the surface or near­surface layers in cloud-free areas.Therefore, satellite data complementconventional shipboard observationsbut cannot replace them. The bestresearch approach often requires closecoordination of the two sources of in­formation. In this way, the evolvingcapabilities of satellite remote sensingare providing a powerful tool toenhance the efficient use of livingmarine resources.

Satellite Sensors

A variety of instruments measureradiance from the earth's surface invisible, thermal infrared (IR), and

microwave wavelength bands (TableI). The most readily available anduseful dat~ come from the AdvancedVery High Resolution Radiometer(AVHRR) on meteorological satellitesoperated by the National Oceanic andAtmospheric Administration(NOAA) and the Coastal Zone ColorScanner (CZCS) on the experimentalNimbus-7 satellite operated by theNational Aeronautics and Space Ad­ministration (NASA). These ad­vanced sensors are characterized byhigh sensitivity in narrow wavelengthbands, fine ground resolution, andextensive data archival.

Satellites receive electromagneticradiation emitted from the sea surface(the IR temperature signal) and backscattered from below the surface (thevisible ocean color signal). Thesesignals are contaminated by reflectionfrom the sea surface and clouds, andby absorption, emission, and scatterby atmospheric particulates andmolecules. Some of these errors canbe eliminated or minimized very sim­ply. For example, sunglint and "limbdarkening" (by long atmosphere pathsat oblique viewing angles) are avoidedby constraining the viewing geometryof the sensor. Corrections of someother errors, however, require ad­vanced image processing methods.

Dense clouds may so completelyabsorb visible and IR radiation fromthe sea surface that no type of data

Paul C. Fiedler and R. Michael Laurs arewith the Southwest Fisheries Center, Na­tional Marine Fisheries Service, NOAA, 8604La Jolla Shores Drive, La Jolla, CA 92038.Gary B. Smith is with the Northwest andAlaska Fisheries Center, NMFS, NOAA,2725 Montlake Boulevard East, Seattle, WA98112.

1

Sea surface tem·perature, sea ice

1.1

0.825 Phytoplankton pig·ments, turbidity, seasurface temperature

1,566

3,0000.63

0.440520.550.670.75

11.5

0.913.74

10.8'12.0

VisibleNear-

infraredInfraredInfraredInfrared

VisibleVisibleVisibleVisibleNear-

infraredInfrared

12345

6

34

'5

SpectralSatellite Channel band

Nimbus-7

TIROS·NNOAA""

7,8

AVHRR

Sensor'

CZCS

Table 1.-Some satellite sensors measuring visible, infrared, or microwave radiance for oceanographicmeasurements. WaYelengths are band midpoints. Ground resolution dimensions are directly beneath the satellite.

GroundWavelength Scan width resolution Primary

(pm) (km) (km) measurements

JFMAMJJASOND

MONTH

'Sensors: AVHRR = Advanced Very High Resolution Radiometer; CZCS = Coastal Zone Color Scanner;MSS = Multispectral Scanner; VISSR = Visible and Infrared Spin Scan Radiometer; SMMR = Scanning Multichan·nel Microwave Radiometer.'Channel 5 on NOAA·7 and NOAA·8 satellites only.

Figure I. - Mean monthly totalcloud amount over coastal waters,from surface marine weather obser­vations, 1921-72 (Nelson andHusby, 1983): Baja California(27°N, 115°W), Southern Califor­nia Bight (33°N, 119°W), CapeMendocino (41°N, 125°W), Van­couver Island (48°N, 126°W). Gulfof Alaska data (56°N, 138°W) arefrom satellite observations, 1%7-72(Sadler et al., 1976), intercalibratedagainst data from Nelson andHusby (1983).

MSS

VISSR

SMMR

LANDSAT

GOES

Nimbus-7SEASAT

123

12

12345

VisibleVisibleNear·

infraredNear­

infrared

VisibleInfrared

MicrowaveMicrowaveMicrowaveMicrowaveMicrowave

055 1850.650.75

0.95

0.62 Earth disk11.5

4.54 x 10' 6002.8 x 10'1.66x10'1.36x10'0.81 x 10'

0.079 Water color, tur·bidity, sea ice

7 x 3 Cloud cover, sea sur­face temperature

20-100 Sea surface tem­perature, sea ice,near-surface winds

processing can retrieve a useful signal.Clouds severely limit satellite coverageof the sea surface in some regions ofthe eastern North Pacific, particularlynorth of Jat. 40 0 N (Fig. 1). South oflat. 27°N, off Baja California, meanmonthly cloud cover is consistentlyless than 50 percent, due to persistentoffshore flow of dry continental air.From lat. 30 0 N to 38°N, coastalwaters are covered by a dense layer oflow stratus clouds during the summerupwelling season (Nelson and Husby,1983). The most favorable conditionsfor remote sensing at these latitudesare found from October throughMarch or April, especially during oc­casional brief periods when SantaAna winds blow warm, dry desert airoffshore and produce cloud-free con­ditions up to 1,000 km from the coast.

In contrast, the most cloud-freeconditions at lat. 4Oo -50oN are foundin late summer, from August to Oc­tober. Mean monthly cloud cover in­creases to the north and is consistently70 percent or greater in the Gulf of

Alaska, although March and Aprilmay be relatively clear. As a generalrule, percent cloud cover increases byat least 10 percent from the coast to adistance on the order of 200 km off­shore in the eastern North Pacific(Nelson and Husby, 1983).

The probability of cloud-free con­ditions in a region of interest duringregular satellite passes is looselyreflected in these monthly cloud coverstatistics. Cloud cover restricts the ex­tent to which satellite data can be an­ticipated, or depended upon, to beavailable. Whereas it may be possibleto obtain regular daily or weeklycoverage of sea surface features in thesouth, such as in the SouthernCalifornia Bight, in the north thereare generally more frequent andlonger data gaps. Satellite coverage ofocean features in the Alaska region ischaracteristically limited to occasionalcloud-free scenes that, despite beinginfrequent, can be rich in informa­tion. Microwave radiometers canmeasure sea surface temperature

through clouds, but with a lower sen­sitivity and much coarser resolutionthan IR radiometers.

Data Processing

The AVHRR measures thermal in­frared radiant energy in threewavelength bands (Table 1).Temperature calibration data are ob­tained by scanning deep space and in­ternal blackbody targets. Accurate seasurface temperatures are calculatedfrom empirical regressions of ship andbuoy temperatures on satellitetemperatures in two or three bands.Such multispectral corrections arebased on the different response ofeach band to the cold bias caused byatmospheric water vapor (McClain etaJ., 1983). Pixels (samples) containingclouds even smaller than the sensor'sfield of view are screened in daytimepasses using the near-infrared albedodata from channel 2 (Bernstein,1982).

The CZCS measures visible lightin five narrow wavelength bands

2 Marine Fisheries Review

Figure 2. - Satellite data collection and processing network utilized by the Na­tional Marine Fisheries Service on the west coast.

mm.rcial.hlng

V....I.

tammg environmental satel1ite dataare given in Cornillon I. In general,photographic copies of raw data areof limited value and a user must haveaccess to a computer-based, imageprocessing system to extract useful in­formation from satel1ite data ondigital tapes (this situation maychange in the future if products deriv­ed from satel1ite data are offered bycommercial processing enterprises).

Applications

We briefly review here examples ofgeneral applications of satellite datato fisheries research and managementproblems along the Pacific coast be­tween California and Alaska. Someof these examples have been publish­ed elsewhere, but none of the applica­tions has as yet been fully realized.

I Comillon, P. 1982. A guide to environmentalsatellite data. Univ. R. I. Mar. Tech. Rep. 79,469 p. Available for $20.00 from the Universityof Rhode Island, Marine Advisory Service,Publication Unit, Narragansett, RI 02882.

Nlmbua-7CZCS~

,:\"i'l'.1,.,"I' '"',:.

Ocean Environment/Fisheries Relationships

\ /Scrlppa

SataUlt. Oc.anography 510Facility Vialblllty Laboratory

Data Received CZCS Dataand Archived

Data Processing~ AVHRR Data Ocean Color

Data Processing Imagery

Sea Surface TemperatureCoand Chlorophyll Imagery

FI

I HState Fishery ~NMFS Southw.at Agencies

FI.h.rl•• C.nt.r

UFish Catch Data

~~.... NOAA-7.~ AVHRR

reception capabilities, have beenestablished at the University of Miamiand by the Northeast Area RemoteSensing System, which is a regionalassociation of university, industry,and government organizations thatwas recently formed in the north­eastern United States.

Global satellite data are archived bythe Satellite Data Services Division ofNOAA and are available to thepublic, although acquisition may re­quire several months and a largebacklog of CZCS data is unprocessed.AVHRR and CZCS data are availablein various forms including photo­graphic prints and negatives, digitaldata on magnetic tapes, and maps ofderived sea surface temperature andice cover. These products and acatalog of available CZCS data areavailable from the National En­vironmental Satellite Data and Infor­mation Service, Satellite Data ServicesDivision, NOAA, Room 100, WorldWeather Building, Washington, DC20233. Detailed procedures for ob-

selected for estimating phytoplanktonpigments, suspended sediments, anddissolved organic matter. Up to 90percent of the visible radiance re­ceived at the satellite is skylightreflected and scattered within the at­mosphere. Corrections are based onassumptions that the red light emittedby the ocean and measured by chan­nel 4 is either negligible or can be ac­curately predicted from radiances inother bands. The atmospheric ra­diance measured by channel 4 canthen be related to atmospheric ra­diances in other channels usingknown spectral properties of at­mospheric scattering (Gordon et aI.,1983). Corrected ratios of blue toyel1ow-green (channell/channel 3) orgreen to yellow-green (channel2/channeI3) radiance are then used tocalculate phytoplankton pigment con­centration from empirical regressionrelationships (Smith and Baker,1982).

Satel1ite data, properly correctedfor the various errors describedabove, have been validated by seatruth data from ships to ± < laC(Bernstein, 1982) and ± 0.4 logch).0fophyl1 concentration (Smith andBaker, 1982). While important sub­surface features such as chlorophyl1maximum layers and some cold-coreeddies or oceanic fronts may not bedetected by satel1ites, the measuredparameters are, in general, closelyrelated to properties such as mixed­layer temperature and integratedchlorophyll or primary productivity inthe euphotic zone (Smith, 1981).

Data Availability

Applications of satellite data areultimately limited by their availability.Ideally, a user would have immediateaccess to data received directly fromsatel1ites and conveniently archived,with data processing facilities at hisfingertips. This ideal has been ap­proached by an arrangement at theSatellite Oceanography Facility ofScripps Institution of Oceanographyin La Jol1a, Calif., for most of thework reported here (Fig. 2). Similararrangements, lacking direct data

46(3) 3

Figure 3. -Northern anchovy egg distribution, 20 March-IO April 1980; • =0, 1= 1-4,2 = 5-17, 3= 18-245 eggs/0.05m2• Top photo (A) is sea surface temperature (0C) from NOAA-6 AVHRR, 7 April 1980; Photo Bis of phytoplanktonpigments (mg m- 3) from Nimbus-7 CZCS, 8 April 1980.

Interpretation of Ship Survey DataA large stock of northern anchovy,

Engraulis mordax, is found offsouthern California, where intensiveegg surveys have been conducted since1980 to estimate spawning biomass.Satellite images of sea surfacetemperature and phytoplankton pig­ment concentration, obtained duringan egg survey in April 1980, depict en­vironmental limits on the range ofspawning anchovy (Fig. 3). North ofSan Diego, no eggs were found inwater colder than 14°C. To the south,spawning northern anchovy were con­fined along the coast to a narrowband of water with high pigment con­centrations. Similar distribution pat­terns were observed in 1981 and 1982(Lasker et aI., 1981; Fiedler, 1983).

The distribution of spawning adultsmay reflect the critical need by first­feeding larvae for aggregations ofsuitable food organisms in a stablewater column. The environmentallimits revealed by the satellite imageswould not have been revealed by theegg survey data alone, becausephytoplankton pigments were notmeasured and spatial coverage was

4

limited.Satellites data may be used to inter­

pret the large-scale patterns and possi­ble causes of spatial/temporal vari­ability in other types of ship surveydata. Most measurements of fish eggsand larvae, plankton concentrations,chlorophyll, nutrients, temperature,and salinity are made at point loca­tions at regular intervals alongtransects or in grid patterns. Remotesensing data collected concurrently, ifavailable, may provide valuable infor­mation on factors contributing to pat­terns in the shipboard data and onconditions beyond the area surveyed.Besides the insight that usually resultsfrom a synoptic view, this informa­tion is important in evaluating thevalidity of interpolation and ex­trapolation of ship survey results.

Pelagic Fisheries Distribution

Albacore, Thunnus ala lunga, is amigratory oceanic tuna that is an im­portant target species for jigfishing,live bait, and recreational fishingalong the U.S. Pacific coast from Julyto October. Fishermen search for ag­gregations of these fish in warm, blue

oceanic waters near temperature orwater color fronts at the offshore edgeof productive coastal waters. Dailycatch records during peak fishing inlate summer 1981 were taken fromlogbooks submitted to west coast statefisheries agencies. Concentrations offishing activity and large catches in­dicate sites of albacore aggregations.

Satellite images of sea surfacetemperature and phytoplankton pig­ment concentrations clearly showfavorable sites for aggregations inpockets of warm, blue oceanic waterintruding into the colder and moreturbid coastal water mass (Fig. 4;Laurs et aI., In press). Albacore arevisual predators and may feed mostefficiently in the clear water adjacentto coastal waters where prey densitiesare higher. Albacore fishermen, run­ning several days out of port in smallboats to reach offshore fishinggrounds, can benefit from timelymaps of sea surface temperature andocean color gradients derived fromsatellite data. Potential benefits in­clude decreased search time, lowerfuel use, and increased catches.

Satellite data may be applied to

Marine Fisheries Review

Figure 4. - Daily albacore catches off central California, 19-24 September 1981: I = 0-27,2=33-63,3=64-81,4=87-140,5= 148-750 fish/boat. A is sea surface temperature (0C)from NOAA-7 AVHRR, 21 September 1981; B is phytoplankton pigments (mg m- 3) fromNimbus-7 CZCS, 22 September 1981.

other pelagic fisheries, both to directfishing effort and for fisheriesmanagement. In the Pacific North­west and off Alaska, five species ofPacific salmon - sockeye salmon,Oncorhynchus nerka; chum salmon,0. keta; pink salmon, O. gorbuscha;chinook salmon, 0. tshawytscha; andcoho salmon, O. kisutch-are impor­tant resources for a large fishing in­dustry. During their oceanic lifehistory phase, these salmon arecaught by power or hand-trollingboats using methods similar toalbacore fishing. The study describedby Borstad et al. (1982) is an exampleof a use of remote sensing techniquesto study relationships between salmondistributions, fishing, and oceano­graphic conditions. Other pelagicresources for which there may beanalogous applications includePacific herring, C/upea harenguspallasi, off Alaska; jack mackerel,Trachurus symmetricus, and Pacificmackerel, Scomber japonicus, offCalifornia; and bluefin tuna, Thun­nus thynnus orienta/is, off BajaCalifornia.

Stock Recruitment Predictions

An application of satellite data that

46(3)

merits investigation is the improvedunderstanding and prediction ofvariations in recruitment to coastalfish stocks. In the eastern NorthPacific, offshore transport and tur­bulent mixing may have important ef­fects on the larval survival and subse­quent recruitment of some species(Bakun and Parrish, 1982). Bothprocesses are driven by surface windstress. Offshore transport removeseggs and larvae from productive near­shore nursery grounds. This may bemost important where strong seasonalupwelling occurs over a narrow con­tinental shelf, as it does along theU.S. coast north of lat. 34°N. Tur­bulent mixing reduces stratification ofthe near-surface water column anddisrupts aggregations of foodorganisms required for successful firstfeeding by newly hatched larvae(Lasker, 1981).

Variations in annual recruitment ofPacific whiting, Mer/uccius produc­tus, have been related to monthlymean Ekman transport estimatedfrom equatorward wind stress offcentral California (Bailey, 1981).Fiedler (1984) has used AVHRR im­agery to document interannualchanges in the sea surface temperature

field related to vanatlOns in coastalupwelling off the U.S. Pacific coast.Satellite imagery has also revealedthat offshore transport is a highly ir­regular process in both time andspace, manifested as meandering jetsand eddies of cold water (e.g. Fig. 5).Offshore transport could perhaps bequantified by measuring the areal ex­tent of relatively cold water nearshoreor the offshore displacement of fron­tal features between daily satellitepasses. This application may be ex­tended to other species for whichthere is evidence of similar en­vironmental effects (e.g. Hayman andTyler, 1980; Bakun and Parrish,1982).

The seasonal timing and spatiallocation of northern anchovy spawn­ing has been related to turbulent mix­ing, as indexed by the cube of meansurface wind speed in historicalmarine weather observations, byHusby and Nelson (1982). They sug­gested that strong year classes may de­pend upon the occurrence of "time­space windows" with sufficiently lowturbulence to allow development oflayers of food organisms.

The magnitude and direction ofsurface wind stress are directly

5

Figure 5. - Localized offshoretransport of cold upwelled water(whiter shades) off northernCalifornia and southern Oregon.The three major coastal landforms,from top to bottom, are Cape Blan­co, Cape Mendocino, and PointArena. White features in the top leftcorner are clouds. Image fromNOAA-7 AVHRR channel 4, 14June 1982.

measurable by microwave sensorssuch as the scatterometer which flewon the short-lived SEASAT satellite in1978. The U.S. government has notyet committed funds to orbit anothersuch sensor, for which there are manypotential scientific and commercialapplications (NASA Satellite WindStress Working Group, 1982).Fisheries applications might includethe real-time planning of fishingoperations to avoid adverse sea stateconditions, as well as the measure­ment of surface winds driving en­vironmental processes affectingrecruitment into coastal fish stocks.These benefits would be particularlysignificant in northern areas, such asthe Pacific Northwest and Alaskaregions, where present remote sensingcoverage by visible and infrared sen­sors is usually limited by cloud cover.

6

Coastal Zone MonitoringMesoscale processes, such as

upwelling and eddy formation, arevery important components of thespatial and temporal variability ofeastern North Pacific coastal waters,especially in the complex CaliforniaCurrent System. Extensive and high­resolution satellite data are particular­ly valuable for studies of these proc­esses (Bernstein et aI., 1977).Anomalous ocean conditions alongthe U.S. Pacific coast were monitoredduring the 1982-83 El Nino event us­ing AVHRR and CZCS data (Fiedler,1984). Localized sea surfacetemperature anomalies up to +6°Cwere observed. Reduced coastalupwelling during the first half of theyear was detected in patterns of cold,nearshore surface waters. CZCS im­ages of the Southern California Bightindicated reduced phytoplankton pro­ductivity. This event affected manycommercial and sport fisheries alongthe coast (Anonymous, 1984).

Monitoring coastal zone variabilitywithin a year or season with satellitedata may be useful for developingbetter release strategies for hatchery­reared juvenile salmonids by enablingmanagement decisions to be moreadaptive and based on conditions innearshore nursery areas. It might bepossible to improve the growth andsurvival of hatchery-reared salmon byscheduling releases during periods ofmost favorable ocean conditions.AVHRR and CZCS imagery could beused as sources of information on seasurface temperature, river plume tra­jectory, upwelling intensity, primaryproduction, and the oceanic frontalareas associated with each of thesefactors. Small improvements in thesurvival of juveniles released fromhatcheries could result in substantialincreases in the numbers of returningadults. If successful, this applicationwould be particularly important in thePacific Northwest and Alaska region,where hatchery operations are exten­sive and salmon resources have highvalue.

Salmon enhancement programs onthe Columbia River, the largest riveron the Pacific Coast of NorthAmerica, provide an example of thepotential. The National Marine

Fisheries Service funds 22 salmonhatcheries and 7 rearing ponds on theColumbia River at an annual cost ex­ceeding $4.5 million. Each year, thesefacilities culture and release 80-100million salmon smolts in late Apriland early May. The present releaseschedule is decided largely on thebasis of administrative criteria, tradi­tion, and attempts to emulate nature.However, survival rates for thesehatchery-reared salmon are relativelylow, only about 3.7 percent (range,0.23 to 12.35 percent) for cohosalmon in their first year (Mathews,1980).

There is increasing evidence thatjuvenile salmon survival, and subse­quent adult returns, are critically in­fluenced by conditions in the oceanenvironment within the first 6 monthsafter hatchery release (Hartt, 1980;Wahle and Zaugg, 1982). Importantfactors may be food availability andthe extent that food limitations occurin the nursery habitats as a result ofcrowding. On the Columbia River,although hatchery-reared salmon passdownstream and through the lowerriver and estuary relatively rapidly,significant losses occur in each ofthese migratory stages and areas.After reaching the ocean, juvenilesalmon typically migrate along arelatively narrow coastal belt (Hartt,1980). In purse seine studies off theColumbia River, juvenile salmonids(< 50 cm) have been found to bedistributed mainly within 28 km ofshore; chinook salmon and steelhead,Salmo gairdneri, were distributedalmost entirely in the river plume(Miller et aI., 1983). Upwelling and itseffects on food production have beenrelated to juvenile survival and adultsalmon production (Gunsolus2

). Abetter understanding of these en­vironmental relationships is a goal ofcurrent research.

Figure 6 shows three satellite im­ages of sea surface temperatures along

'Gunsolus, R. T. 1978. The status of Oregoncoho and recommendations for managing theproduction, harvest, and escapement of wildand hatchery-reared stocks. Oreg. Dep. FishWild\., Columbia Reg. Off., 17330 S.E. EvelynSt., Clackamas, OR 97015. Unpub\. manuscr.,59 p.

Marine Fisheries Review

Figure 6. - Variations in the intensity and pattern of coastal upwelling off thePacific Northwest between lat. 39°00N and 48°3<YN, July-August 1982. Seasurface temperature (0C) from NOAA-? AVHRR channels 4 and 5. Clouds(white) screened with channel 2 data.

the Pacific Northwest coast duringsummer 1982, illustrating changes incoastal conditions that can occur overrelatively short (2-3 week) time inter­vals. On 5 July, coastal upwelling wasvery weak, although signs of inten­sification were developing at CapeMendocino. By 23 July, upwellinghad intensified considerably, especial­ly to the north. By 18 August, how­ever, upwelling was again weaker.Changes in the mesoscale pattern ofcold-water jets and meanders can beseen. In the latter two images, theColumbia River plume is indicated bya dark, warm-water break in the bandof cold, upwelled water along thecoast.

Monitoring Sea lee

Sea ice is an important seasonalfeature of the Alaskan environmentthat can be monitored by satellitecoverage (Weeks, 1981; McNutt,1981). Ice affects commercial fishingactivities by threatening vessel safety,limiting navigation and access, anddamaging fishing gear. Yet sometimeswaters near the edge of pack ice canoffer shelter or rich fishing grounds.Ice is also important because of itssignificance in the ecology of, and as ahabitat for, northern marine mam­mals.

In the eastern North Pacific, winterice cover occurs in inlets along thecoasts of southeastern Alaska, theGulf of Alaska, and the AleutianIslands. Its formation is oftenassociated with freshwater runoff,such as in Cook Inlet (Poole and Huf­ford, 1982). Freshwater ice forms onstream and river deltas, then sea iceforms in lower parts of the em­bayments. However, much greater icecover occurs in the extended area tothe north. Ice covers the Chukchi Sea,Bering Strait, and nearly the entirenortheast half of the eastern BeringSea during winter and spring. Move­ments of pack ice in the Bering Seaare influenced by winds and watercurrents, and are highly dynamic.

Ice off the North Slope, along thesouthern coast of Alaska, and in theBering Sea, is monitored routinely bythe National Weather Service. Whencloud cover permits, ice and openwater can be discriminated with a I

46(3)

km resolution using AVHRR visibleand infrared data (Fig. 7). Combinedwith data from the Nimbus-7 Scan­ning Multichannel Microwave Radi­ometer, which are not influenced byclouds but have a resolution of 60 km,large-scale maps of ice coverage orconcentration are produced. Some in­formation about ice type (age andthickness) can be gleaned fromsatellite data, but the analysis is sup­plemented by aerial reconnaissanceand reports from ships and shore sta­tions. Ice analysis maps are distrib­uted to users in the fishing and oil in­dustries three times a week by radiofacsimile and once a week by mail.

Winter fisheries affected by icecover in the eastern Bering Sea in­clude: Foreign groundfish fisheries,involving 200-300 vessels per year andan annual catch of about 1.3 millionmetric tons (Bakkala et aI., 1979);highly-valued U.S. fisheries for kingcrab, Para/ithodes spp., and snow

(Tanner) crab, Chionoecetes spp.(Otto, 1981); and U.S. fisheries forPacific herring that take place alongthe north shore of Bristol Bay. Vesselsoperating in these fisheries often needto work around sea ice, can havestability problems, and are sometimestrapped in shifting pack ice and lost.

Ice is an important part of thehabitat of 25 species of marine mam­mals that occur in the Bering Sea(Fay, 1974; Burns et aI., 1981). It mayserve as a substrate, barrier, or toforce migrations. Species that regular­ly come in contact with ice include thefollowing: Polar bear, Ursus mari­timus; walrus, Odobenus rosmarus;harbor seal, Phoca vitu/ina; ringedseal, P. hispida; ribbon seal, P.jasciata; bearded seal, Erignathusbarbatus; narwhal, Monodonmonoceros; beluga whale, Delphinap­terus leucas; and bowhead whale,Balaena mysticetus. For these andother species, the wide views of ice

7

.'

Figure 7. -Ice cover in the eastern Bering Sea, 18 February 1983, NOAA-7 AVHRR channel 4. Ice, clouds, and snow­covered land all appear whiter (colder) than open water. Pack ice extends from the Bering Strait, in the top left corner,past St. Matthew and Nunivak Islands, but does not reach the Alaska Peninsula to the southeast. Photograph courtesyof G. L. Hufford, NOAA/NESDIS, Satellite Field Services Station, Anchorage, Alaska.

characteristics that can be obtainedfrom satellite imagery provide uniqueinformation for use in conservationand management.

Migration Routes and Timing

Satellite imagery can provide usefulinformation on environmental condi­tions related to the routes and timingof long-distance migrations by marinemammals and fish. In the easternNorth Pacific and in waters offAlaska, many important species oc­cupy large habitat areas and make ex­tensive seasonal migrations. These

8

large activity areas are often in remoteand data-poor regions. Satellitecoverage, because of its wide arealviews and long-term repeated obser­vations, provides data with time andspace characteristics that are ap­propriate for interpreting large-scalemigratory phenomena. Applicationsinclude planning research, experimen­tal design, and evaluating strategiesfor population enumeration.

The North Pacific albacore, Thun­nus alalunga, performs transpacificmigrations and supports importantcommercial fisheries in the western,

central, and eastern North Pacific.Seasonal migration into NorthAmerican coastal waters is associatedwith the Transition Zone betweenPacific central and subarctic watermasses (Laurs and Lynn, 1977). Thefrontal structure of these waters mayaffect both the timing and location ofthe arrival of albacore into the sum­mer fishery along the coast. Figure 8illustrates CZCS color frontal pat­terns in the central Pacific that couldinfluence the course of the migration.

The potential for using both colorand infrared temperature imagery in

Marine Fisheries Review

Figure 8. - Phytoplankton pigment concentration in a 600 X 600 km area of thecentral North Pacific, measured by the Coastal Zone Color Scanner, 30 June1980. Black = clouds, darkest gray =0.08 mg m- 3, white=5.7 mg m- 3•

1800 WI

I1800 W

fishery operations in these waters hasbeen explored on a limited basis withpromising results. Although cloudcover allows only infrequent views ofthe sea surface, the surface patternsseem to be less dynamic than incoastal waters. Color boundaries aredetectable from satellites during allseasons, but pronounced sea surfacetemperature fronts are not presentduring summer-autumn due toseasonal warming. However, infraredsatellite imagery has been used tolocate the subtropical front duringwinter (Van Woert, 1982) and springmonths3

•

In Arctic and subarctic areas, suchas the Beaufort Sea and northern Ber­ing Sea, the movements and distribu­tion of sea ice have important in­fluences on the movements of marinemammals. Bowhead and beluga

3R. Lynn. 1984. Southwest Fisheries Center,National Marine Fisheries Service, NOAA,P.O. Box 271, La lolia, Ca 92038. Pers. com­mun.

46(3)

whales make regular seasonal migra­tions through the shear zones and leadsystems that develop in the Arcticpack ice (Braham et aI., 1980). In thecase of the bowhead whale, theprediction and evaluation of migra­tion paths is important because it is anendangered species and the westernArctic population requires careful an­nual censusing. Satellite imageryshowing ice cover and open-watercorridors is used to spot check thesampling design for summer whalecounts.

There is potential for similar ap­plications of satellite data in otheroceanic and coastal areas, althoughthe relationships between environ­mental characteristics and marinemammal migrations are not as ob­vious. For example, the easternPacific stock of gray whale,Eschrichtius robustus, undergosregular annual migration betweensummer feeding grounds in theChukchi Sea and winter breedinggrounds in Mexico (Pike, 1962).Satellite coverage of the onset of win-

ter ice cover in the Bering Strait mayhelp explain the environmental signalsthat initiate the southern migration.Other marine mammals that are im­portant in the region, and thatundergo long-distance movements,include: Northern fur seal, Cal/o­rhinus ursinus, migrating betweensummer breeding grounds on thePribilof Islands in the eastern BeringSea (and on other islands around theNorthern Pacific rim) and winteringareas in the Gulf of Alaska and offthe U.S. Pacific coast (Fiscus, 1978);humpback whale, Megaptera nova­eangliae, an endangered species ofwhich the eastern Pacific stockmigrates between breeding and calv­ing areas in the Hawaiian Islands andsummer feeding grounds in the inlandwaters of southeastern Alaska(Wolman, 1978); and a number ofspecies of dolphins and porpoise.

Allocation of Sampling Effort

Annual egg and larva surveys ofnorthern anchovy are a regular activi­ty of the California CooperativeOceanic Fisheries Investigations(CaICOFI). Each survey requires 4-5weeks of ship time to make verticalnet tows at 800-900 stations. The ac­curacy of the estimate of spawningbiomass derived from each surveydepends on complete coverage of thegeographic range of the spawningstock. Without a priori knowledge ofthis distribution, many extra stationsbeyond the range limits must besampled to ensure adequate coverage.Weare now investigating the use ofsatellite imagery to plan egg surveysbased on environmental limitscharacterizing the spawning habitatand detectable from satellites, asdescribed above.

The 1983 survey plan wasmodified during the cruise using in­formation from satellites. When thecruise began north of Point Concep­tion in February, eggs were foundmuch farther north and offshore thanin recent years. Upon examination ofAVHRR temperature imagery, werealize that this extension of thespawning range was due to the1982-83 El Nino event. The cold­water boundary which normally limits

9

J"~',ii."";,."~~.,

IO

Figure 9. - Northern anchovy eggdistribution, 9 February-29 March1983: -=0, 1=1-3, 2=4-12,3 = 13-229 eggs/O.OS m2 • Sea sur­face temperature (DC) fromNOAA-7 AYHRR, 10 February1983.

spawning to the southeast of PointConception had shifted to the northand offshore (Fig. 9, compare withFig. 3). As a result, the lines of sam­pling stations were extended fartheroffshore than originally planned, toensure coverage of the entire spawn­ing stock.

In another study conducted in June1980 on the feeding biology of larvaljack mackerel, satellite data were usedto locate an intensive sampling grid ona temperature front. The front wasobserved southwest of San Diego in aNOAA-6 AVHRR image obtainedover 1 week prior to the cruise, and itpersisted during the sampling period(Fig. 10). The front was subsequentlyfound to be related to an importantgradient in food availability 4.

These two examples demonstratehow a single satellite image, from datareceived and processed in a matter ofhours, can save days of ship time bylocating significant environmentalfeatures and permitting efficientallocation of sampling effort. Thepotential cost savings are obvious, butthe real-time use of satellite data re­quires facilities for direct receptionand processing.

Fisheries Aid Charts

Operational applications of satellitedata to commercial fishing activities

4 R. Hewitt. 1983. Southwest Fisheries Center,National Marine Fisheries Service, NOAA,P.O. Box 271, La Jolla, CA 92038. Pers. com­mun.

Figure 10. - Uncalibrated NOAA-6A YHRR channel 4 image offsouthern California, S June 1980.Lighter shades represent cold seasurface temperatures, clouds appearwhite. Box encloses a 30 x 67 kmgrid of 41 stations centered at lat.31 0 N, long. 1200 30' W.

Marine Fisheries Review

i~ .

~~j1... ;!;

iio~

~$~-g

j~"II! 'i

~~

J1~"~--l~ ~

.~ f gJ1.~S::~LC - ~

<,.m: II6~hw,===:i~~"==~~'::~;=~~NA~T~'O;:;;N~AL~WE~A~T~HE~R~S~E';RSVI~CE:5DATE ISSUED: "lA,4,1 8,. ~

A.. ·".'rA SEATTLE OCEANSATELLITE IMAGE: ....'10 ~"".'

" =-n... ~. SERVICES UNITSHIP OBSERVATIONS: AM, f-?,83 _

began along the Pacific coast in 1975(Breaker, 1981) and have now beenextended to Alaska. Charts showingsea surface temperature fronts and seaice visible in satellite AVHRR images,and surface isotherms mapped fromsatellite and ship data, are producedroutinely by NOAA (Fig. 11). TheNorthwest Ocean Services Center inSeattle produces a chart coveringnorthern California, Oregon, Wash­ington, and southern British Colum­bia, from lat. 400 N to lat. 52°N. TheSatellite Field Services Station in Red­wood City, Calif., produces a chartcovering central and southern Califor­nia and northern Baja California,from lat. 28°N to lat. 40o N. Coverageof both charts extends offshore tolong. 135°W. They are produced onceor twice weekly year-round and aredistributed primarily by radio fac­simile from the U.S. Coast Guard

Figure II. - Sea surfacetemperature charts producedby the National WeatherService.

Lorna

SEA SURFACETHERMALANALYSIS

NAT'L EARTH SATELLITESERVICE, 660 PRICE AVE.,REDWOOD CITY, CA 94063Date: CJ60 I

F MAR 4N

___liM' Of (QU.l I(M'lUlu.

",.-" f~~::~~ ~~~:~

•.••.• TlI':'~~.~C'~~~l

C=COlD\11= ....

2.0--+---

11_-""",-

Sur Comments:f---~...t-----t---+-~r--t----+---Tfi,--------f\'\Piedras fJA,A'UN(r UitS

lancas :bECIt54SEDTemp.Conversion°C OF °C OF24 . 75.2 14 - 57222 - 71.6 12- 53.620' 68.0 10 - 50.01 8 - 64.4 8 - 46.416 ·60.8

46(3) /I

Figure 12. -Ocean color boundary chart off northern California and Oregon(R. Wittenberg, Scripps Visibility Laboratory).

radio station at Point Reyes, Calif.Fishermen use these charts to savetime in searching for productivefishing areas associated with frontalfeatures.

The Alaska Ocean Services Unit,located in Anchorage, producescharts covering British Columbia, theGulf of Alaska, Aleutian Islands,eastern Bering Sea, Chukchi Sea, andBeaufort Sea, from lat. 48°N to lat.75°N. Charts showing sea surfacetemperature and sea ice are distrib­uted three times a week; these are in­tended to aid the safety and efficiencyof fishing. Other charts showing3-and 5-day sea ice forecasts are alsodistributed daily in winter. Both typesof charts are issued by radio facsimilefrom the U.S. Coast Guard Station atKodiak, Alaska.

Ocean color boundary charts havebeen produced experimentally fromCZCS data since 1981 in a NASA/JetPropulsion Laboratory program(Montgomery, 1981). The chartsdelineate strong gradients in theblue/green color ratio (channellIchannel3 radiance). In 1983, chartswere produced at almost weekly inter­vals from May to October (18 charts

12

OCeAN C-OIOR160\..1\'\ t>AR.IE S

1'4 Sept IqB">

if-I - Gree" Coa.,1.\ lIJ,{e.

~;l. - Gree""k ~\~e

tl3- "t,l,u'h. G-ree,,-

1<~ Deep 1):.<0" ~luE

"Do.sk.eJ I,,,e.::. r'whr-Ii \c.<;<"

i)ef"\e& i3i>l.l.>\.d.Ar,e So

J.?l

'::.t.p+ IS 1'83

Vlslc:o.'b - Sr,oQw.

total), covering coastal areas up to700,000 km2 between VancouverIsland, B.C., and Guadalupe Island,Mex. (Fig. 12). A chart was producedonly when a large, cloud-free area waslocated in a Nimbus-7 pass. The chartwas then broadcast on the same orfollowing day to fishing boats byradio facsimile from Point Reyes andLa Jolla, Calif. Color photographs ofthe satellite images were distributedby express mail to various fishingports and to Sea Grant marine ad­visors in daily contact with fishermen.The color boundary charts andphotographs are used by albacore andsalmon fishermen. These fish aresometimes found aggregated alongcolor fronts which do not correspondto temperature fronts.

Conclusions

Satellites have altered our percep­tions of the ocean environmentthrough the extensive spatialcoverage, temporal continuity, andhigh resolution of the data they pro­vide. Limited, but useful, applicationsto several types of problems infisheries research and operations havebeen demonstrated. Continued

development of these and other ap­plications will depend on inexpensiveand convenient access to data anddata-processing facilities. Recenttrends in the federal budget have notbeen encouraging. For instance, thereare no current prospects for a newocean color scanner to replace the ag­ing CZCS on Nimbus-7. The com­mercial utility of satellite datadepends largely on near real-timeavailability to fishermen and othermaritime users. In the future, this de­mand may be met by processedsatellite data products tailored moreto particular user needs.

Acknowledgments

This work was supported jointly bythe Southwest Fisheries Center, LaJolla, Calif., and the Northwest andAlaska Fisheries Center, Seattle,Wash., as a cooperative researchproject. We thank the many personswho contributed information andideas for various applications, as wellas those who provided technical sup­port at the Scripps SatelliteOceanography Facility and theScripps Visibility Laboratory. Pro­cessing of CZCS data was supportedby NASA order WI5,334.

Literature Cited

Anonymous. 1984. 1983 in review. Pac. Fish.1984 Yearb.

Bailey, K. M. 1981. Larval transport andrecruitment of Pacific hake, Merluccius pro­ductus. Mar. Eco!. Prog. Ser. 6:1-9.

Bakkala, R., W. Hirschberger, and K. King.1979. The groundfish resources of the e<,sternBering Sea and Aleutian Island regions. Mar.Fish. Rev. 41(11):1-24.

Bakun, A., and R. H. Parrish. 1982. Turbu­lence, transport, and pelagic fish in theCalifornia and Peru Current systems.CalCOFI Rep. 23:99-112.

Bernstein, R. L. 1982. Sea surface temperatureestimation using the NOAA-6 satellite Ad­vanced Very High Resolution Radiometer. J.Geophys. Res. 87:9455-9465.

---=~,----' L. Breaker, and R. Whritner. 1977.California Current eddy formation: Ship, air,and satellite results. Science 195:353-359.

Borstad, G. A., R. M. Brown, D. Truax,T. R. Mulligan, and J. F. R. Gower. 1982.Remote sensing techniques for fisheriesoceanography: Examples from British Co­lumbia. NAFO Sci. Counc. Studies 4:69-76.

Braham, H. W., M. A. Fraker, and B. D.Krogman. 1980. Spring migration of thewestern Arctic population of bowheadwhales. Mar. Fish. Rev. 42(9-10):36-46.

Marine Fisheries Review

Breaker, L.c. 1981. The application of satelliteremote sensing to west coast fisheries. J. Mar.Tech. Soc. 15:32-40.

Burns, J. J., L. H. Shapiro, and F. H. Fay.1981. Ice as marine mammal habitat in theBering Sea, p. 781-797. In D. W. Hood andJ. A. Calder (editors), The eastern Bering Seashelf: Oceanography and resources, Vol. 2.Univ. Wash. Press, Seattle.

Fay, F. H. 1974. The role of ice in the ecologyof marine mammals of the Bering Sea, p.383-399. In D. W. Hood and E. J. Kelly(editors), Oceanography of the Bering Seawith emphasis on renewable resources, Univ.Alaska, Fairbanks, Inst. Mar. Sci. Occas.Publ. 2.

Fiedler, P. C. 1983. Satellite remote sensingof the habitat of spawning anchovy in theSouthern California Bight. (CaICOFI Rep.24:202-209.

-:-::-::-::-C=-' 1984. Satellite observations of the1982-83 EI Nino along the U.S. Pacific coast.Science 224: 1251-1254.

Fiscus, C. H. 1978. Northern fur seal,p. 153-159. In D. Haley (editor), Marinemammals of eastern North Pacific and Arcticwaters. Pac. Search Press, Seattle.

Gordon, H. R., D. K. Clark, J. W. Brown,O. B. Brown, R. H. Evans, and W. W.Broenkow. 1983. Phytoplankton pigmentconcentrations in the Middle Atlantic Bight:Comparison of ship determinations andCZCS estimates. Appl. Opt. 22:20-36.

Gower, J. F. R. 1982. General overview of thenature and use of satellite remote sensing datafor fisheries application. Northwest Atl. Fish.Organ., Sci. Counc. Studies 4:7-19.

Hartt, A. C. 1980. Juvenile salmonids in theoceanic ecosystem-the critical first summer,p. 25-57. In W. J. McNeil and D. C.Himsworth (editors), Salmonid ecosystems ofthe North Pacific. Oreg. State Univ. Press,Corvallis.

Hayman, R. A., and A. V. Tyler. 1980.Environment and cohort strength of Doversole and English sole. Trans. Am. Fish. Soc.109:54-70.

Husby, D. M., and C. S. Nelson. 1982.Turbulence and vertical stability in theCalifornia Current. CalCOFI Rep.

46(3)

23: 113-129.Lasker, R. 1981. Factors contributing to

variable recruitment of the northern anchovy(Engraulis mordax) in the California Current:Contrasting years, 1975-1978. Rapp. P.-v.Reun. Cons. int. Explor. Mer 178: 375-388.

-----==-__, J. Pelaez, and R. M. Laurs. 1981.The use of satellite infrared imagery fordescribing ocean processes in relation tospawning of the northern anchovy (Engraulismordax). Remote Sensing Environ.II :439-453.

Laurs, R. M., and R. J. Lynn. 1977. Seasonalmigration of North Pacific albacore, Thun­nus alalunga, into North American coastalwaters: Distribution, relative abundance, andassociation with Transition Zone waters.Fish. Bull., U.S. 75:795-822.

____, and J. T. Brucks. In Press. Livingmarine resources applications of satellite im­agery. In B. Saltzman (editor), SatelliteRemote Sensing in Oceanography. Advancesin Geophysics, Acad. Press, N.Y.

____, P. C. Fiedler, and D. R. Mont­gomery. In Press. Albacore tuna catchdistributions relative to environmentalfeatures observed from satellites. Deep-SeaRes.

McClain, E. P., W. G. Pichel, C. C. Walton,Z. Ahmad, and J. Sutton. 1983. Multi­channel improvements to satellite-derivedglobal sea surface temperatures. Adv. SpaceRes. 2:43-47.

McNutt, S. L. 1981. Remote sensing analysisof ice growth and distribution in the easternBering Sea, p. 141-165. In D. W. Hood andJ. A. Calder (editors), The eastern Bering Seashelf: Oceanography and resources, Vol. I.U.S. Gov. Print. Off., Wash., D.C.

Mathews, S. B. 1980. Trends in Puget Soundand Columbia River salmon, p. 133-145. InW. J. McNeil and D. C. Himsworth (editors),Salmonid ecosystems of the North Pacific,Oreg. State Univ. Press, Corvallis.

Miller, D. R., J. G. Williams, and C. W. Sims.1983. Distribution, abundance and growth ofjuvenile salmonids off the coast of Oregonand Washington, summer 1980. Fish. Res.2:1-17.

Montgomery, D. R. 1981. Commercial appli-

cations of satellite oceanography. Oceanus24(3):56-65.

NASA Satellite Wind Stress Working Group.1982. Scientific opportunities using satellitewind stress measurements over the ocean.Nova Univ.lN.Y.I.T. Press, Fort Lauder­dale, Fla., 153 p.

Nelson, C. S., and D. M. Husby. 1983.Climatology of surface heat fluxes over theCalifornia Current region. NOAA Tech.Rep. NMFS SSRF-763, 155 p.

Otto, R. S. 1981. Eastern Bering Sea crabfisheries, p. 1037-1066. In D. W. Hood andJ. A. Calder (editors), The eastern Bering Seashelf: Oceanography and resources, Vol. 2.Univ. Wash. Press, Seattle.

Pike, G. C. 1%2. Migration and feeding of thegray whale (Eschrichtius gibbosus). J. Fish.Res. Board Can. 19:815-838.

Poole, F. W., and G. L. Hufford. 1982.Meteorological and oceanographic factors af­fecting sea ice in Cook Inlet. J. Geophys.Res. 87(C3):2061-2070.

Sadler, J. c., L. Oda, and B. J. Kilonsky.1976. Pacific Ocean cloudiness from satelliteobservations. Univ. Hawaii, Honolulu, 137p.

Smith, R. C. 1981. Remote sensing and depthdistribution of ocean chlorophyll. Mar. Ecol.Prog. Ser. 5:359-361.

_-:-:-_--:-' and K. S. Baker. 1982. OceaniccWorophyll concentrations as determined bysatellite (Nimbus-7 Coastal Zone Color Scan­ner). Mar. BioI. 66:269-279.

Van Woert, M. 1982. The subtropical front:Satellite observations during FRONTS 80. J.Geophys. Res. 87:9523-9536.

WaWe, R. J., and W. S. Zaugg. 1982. Adultcoho salmon recoveries and their Na + - K+ATPase activity at release. Mar. Fish. Rev.44(11):11-13.

Weeks, W. F. 1981. Sea ice: the potential ofremote sensing. Oceanus 24(3):39-48.

Wolman, A. A. 1978. Humpback whale,p. 47-53. In D. Haley (editor), Marine mam­mals of eastern North Pacific and Arcticwaters. Pac. Search Press, Seattle.

Yamanaka, I. 1982. Application of satelliteremote sensing to fishery studies in Japan.NAFO Sci. Counc. Studies 4:41-50.

13