Remote Sensing for Marine Spatial Planning andAquaculture

Sonia Cristina, Bruno Fragoso, John IcelySagremarisco

April 23, 2018

Abstract

This text is part of the introduction to aMasters-level course in ‘Planning and Man-aging the Use of Space for Aquaculture’made by the AquasSpace project. It intro-duces the use of Remote Sensing in Ma-rine Spatial Planning and site selection foraquaculture. Brief case studies illustrate theaquacultural utility of satellite remote sens-ing of the coastal waters of the Algarve, thenorthern Adriatic Sea, and a bay in north-ern France; a case study for Nova Scotiaillustrates the use of aerial drones. An ex-ercise involves the use of the ESA SentinelApplication PLatform (SNAP) software,

This document may be cited as: Cristina,S., Fragosa, B. & Icely, J.(2018) Remote Sens-ing for Marine Spatial Planning and Aquacul-ture, v.3. Aquaspace project document, 13 pp.Sagaremisco, Portugal.

Contents

1 Unit Study Guide 2

2 Remote Sensing 2

3 Missions and sensors 3

4 Optically Active Constituents 3

5 Remote Sensing and Aquaculture 4

6 Case Study: Algarve 4

7 Case Study: Adriatic 5

8 Case Study: Mont St-Michel bay 7

9 Case Study: Nova Scotia 8

10 Exercises and SAQ 8

A Use of SNAP with Sentinel-3 data 9A.1 Install SNAP . . . . . . . . . . 9A.2 Get Sentinel-3 data . . . . . . . 9A.3 Open data in SNAP . . . . . . 10

B NASA OceanColor site 12

1

6: Remote Sensing 2

1 Unit Study Guide

This text was written during the H2020 Aquas-pace project (2015-2018, contract no. 633476)for a Masters-level course in ‘Planning andManaging the Use of Space for Aquaculture’.The course consists of a number of units; thisunit provides an introduction to the use of Re-mote Sensing in marine spatial planning foraquaculture.

The unit comprises a text (the documentyou are now reading), a set of slides, requiredfurther reading, and some exercises. The textcomplements the slides, which can be used asthe basis for a class-room lecture. The readingprovides details of case studies. The exercisesillustrate some of the methods used in process-ing remotely sensed data

The learning outcomes for this unit are tobe able to

• demonstrate an understanding of RemoteSensing of the marine environment and itsuse for the spatial planning of Aquacul-ture;

• explain and critically discuss at least oneapplication of Remote Sensing for aqua-cultural site selection or management;

• apply skills in the processing of remotelysensed data, using at least one freely avail-able software package

2 Remote Sensing

Remote sensing is the science of obtaining in-formation about objects or areas from a dis-tance, typically from aircraft or satellites. Re-mote sensors collect data by detecting eitherpassive or active energy that is reflected fromEarth.

Passive sensors respond to external stimuliin the form of natural energy that is re-

flected or emitted from the Earth’s sur-face. The most common source of radi-ation detected by passive sensors is re-flected sunlight.

Active sensors use internal stimuli to collectdata about Earth. For example, a laser re-mote sensing system projects a laser ontothe surface of Earth and measures thetime that it takes for the laser to reflectback to its sensor.

Early passive sensors were essentially cam-eras, and these are still used for example toprovide the basis of the maps used by Googleetc. Sensors for sea-surface temperature orocean colour mostly use a scanning mechanism,whereby visible and near-IR light from strip ofplanetary surface is imaged along an array ofphotodiodes, usually after splitting into severalcolours plus infra-red. A picture of the land orsea is built up as the satellite advances alongits track.

There are numerous technical challenges ininterpreting data provided by sensors onboardsatellites. These include compensating: forthe signal absorbed or added by the atmo-sphere; for the influence on upwelling lightleaving the sea surface by the reflection fromthe bottom in shallow waters; for the absorp-tion and scattering effects of various suspendedand dissolved water components; for the influ-ence of the water itself (IOCCG, 2000; Pozd-nyakov and Graßl, 2003); for the rapid changesin signal at the land-sea boundary; and for theroughness of the water surface. These influ-ences on water-leaving light have to be cor-rected by ‘sea-truthing’ of the remote sensingmeasurements against observations from shipsor moorings before they can used to derive thewater products from remote sensing measure-ments (e.g. chlorophyll a concentration, sus-pended particulate matter) (IOCCG, 2000).

6: Remote Sensing 3

3 Missions and sensors

Airborne remote sensing of ocean colour hasa long history. In his book, ‘The Open Sea’,Hardy (1956) refers to “a sharp line sepa-rating the green water of the English Chan-nel from the deep blue of the Atlantic” thathe observed from the air in 1923. In theearly 1970s oceanographers used sensing of sea-surface temperature, based on infra-red emis-sion measured by low-flying aircraft, to fix thelocations of tidal mixing fronts in coastal wa-ters (Simpson and Hunter, 1974). Aircraftsensing was soon extended by images (Simp-son et al., 1978) from satellites flown by theUS agencies NASA and NOAA. The discov-ery of these fronts changed understanding ofthe physical and biological dynamics of season continental shelves (Holligan, 1981).

The first ocean colour sensor, the CZCS, wasflow by NASA between 1978 and 1986, and al-lowed mapping of chlorophyll in coastal waters,such as the North Sea (Holligan et al., 1989).It was followed by the SeaWIFS sensor car-ried on the SeaStar satellite (1997-2010), andthen by the MODIS sensor carried on severalNASA satellites, including EOS PM-1 (nowAqua) launched in 2002. Increasing numbers ofspectral bands allows better estimation of the‘optically active constituents’ in seawater, es-pecially chlorophyll in nearshore waters wherethe optical signal is made complicated by thatfrom suspended solids and dissolved ‘yellow-substances’ from freshwater.

A large number and variety of Earth obser-vation satellites are currently in orbit, placedthere by government agencies and private en-terprise, and serving a variety of purposes: in-cluding: military; weather forecasting; obser-vation of land and sea. We focus here on Eu-ropean sensors measuring properties of coastalwaters relevant to aquaculture.

The European Space Agency (ESA) was

founded in 1975. Envisat was launched 2002into polar orbit, and ceased functioning in2012. Sensors included MERIS (MEdiumResolution Imaging Spectrometer, measuringEarth reflectance in 15 bands of visible andnear-infrared light). The subsequent Coperni-cus programme comprises a sequence of Sen-tinel missions, each with paired satellites in avariety of orbits.1

Sentinel-1 is a polar-orbiting radar imagingmission, with the first satellite launchedon the 3rd April 2014 (Sentinel-1A)and the second on the 25th April 2016(Sentinel1-B);

Sentinel-2 is a polar-orbiting multispectralimaging mission covering mission cover-ing land as well as coastal waters, the firstsatellite was launched on 22nd June 2015and the second on the 7th May 2017;

Sentinel-3 is a polar-orbiting multispectralimaging mission with observations fo-cusing on the oceans to include sea-surface topography, temperature andocean colour; the latter uses OLCI (theOcean and Land Colour Instrument).The 3A satellite was launched on the 16thFebruary 2016, with Sentinel-3B plannedfor launching in April 2018.

4 Optically Active Con-

stituents

The Optically Active Constituents (OAC) ofseawater are the components that scatter, re-flect, or absorb underwater light, thus con-tributing to the amount and colour of sea-leaving radiance. They are:

1 Notice the distinction between programme,mission, platform (satellite or aircraft) and sensor.

6: Remote Sensing 4

seawater itself

phytoplankton pigments especiallychlorophyll but including carotenoids andphycobilins

SPM or Suspended Particulate Matter, bothliving and non-living

CDOM or Colour Dissolved Organic Matteror ‘gelbstoff’

These interact in complex ways to changethe colour, amount and angular distributionof light, as photons penetrate into the ocean(e.g. Bowers et al., 2000; Devlin et al., 2009).They also affect the small fraction of photonsthat is diverted from a downwards course andso leaves the sea in an upwards direction. Itis by comparing the colour and amount of sea-leaving radiance with that of sunlight falling onthe sea-surface, that it is possible to estimatethe concentrations of the OAC.

Doing so requires complex calculations,combining measurements of light at severalwavelengths, after correction for absorptionand scattering in the atmosphere throughwhich a satellite views the sea.

5 Remote Sensing and

Aquaculture

Remote sensing can provide information aboutthe distribution in space and time of manyproperties relevant to aquaculture. Theseproperties include

• chlorophyll concentration, allowing esti-mation of food for shellfish and obser-vation of the occurrence and spread ofHarmful Algal Blooms;

• sea-surface temperature, which controlsthe growth rate of organisms

Although the costs of launching and main-taining satellites is high, they can be spreadover many users. In addition, some data fromsatellite remote sensing is in the public domain,providing a low-cost way for developers andplanners to plan, monitor, manage and ratio-nally exploit marine resources.

The remainder of this text, and the corre-sponding slides, exemplify some of the usesof remotely sensed data for the managementand spatial planning of aquaculture and themonitoring of environmental conditions rele-vant both to aquaculture and to marine ecosys-tem health in general. The exercises suggestedin section 10 and appendix A provide an in-troduction to obtaining and processing suchdata.

6 Case Study: Algarve

The expansion of aquaculture is a strategicpriority for the Portuguese government to im-prove the use of the extensive exclusive eco-nomic zone occupied by Portugal (GovernoPortugal, 2013). Marine Spatial Planning, in-volving the linking of Remote Sensing imageswith Geographical Information Services (GIS),has been used to define areas off the Algarvecoast that are available for developing aquacul-ture (Fig. 1). This development has to complywith national and EU legal requirements forcoastal areas. These include

• the Marine Strategy Framework Directive(MSFD), aiming to prevent the deteriora-tion of the coastal and oceanic ecosystemsby requiring monitoring for, and main-tenance of, Good Environmental Status(GES);

• the Maritime Spatial Planning Frame-work Directive (MSPFD), aiming to reg-ulate the human uses of these waters.

6: Remote Sensing 5

Figure 1: Image showing the location ofareas available for offshore aquaculture inblue; offshore areas currently under use ingreen; offshore artificial reefs in brown andwhite; and fish ponds on land in brown andwhite (Source: www.eaquicultura.pt).

Christina et al. (2015) used Algarve coastalwaters, near Cape Saint Vincent and the townof Sagres, as a case study to test how MERISAlgal Pigment Index 1 (API1) could contributeto the monitoring for the Good EnvironmentalStatus(GES). The MERIS product extractedwas the API1 that corresponds to the totalconcentration of chlorophyll a and its degra-dation products.

The satellite data were obtained fromMERIS Level 2 Full Resolution (FR) and Re-duced Resolution (RR) satellite images, witha spatial resolution of 290 m × 260 m and1.2 km × 1.04 km, respectively. They wereanalysed, and API1 extracted, with the Ba-sic ERS and ENVISAT (A) ATSR and MERISToolbox (BEAM version 4.9; www.brockmann-consult.de/cms/web/beam).

The results shows that MERIS API1 prod-uct decreased from inshore to offshore, withhigher values occurring mainly between earlyspring and the end of the summer, and withlower values during winter. By using the satel-lite images for API1, it was possible to detectand track the development of algal blooms in

Figure 2: MERIS Reduced Resolution im-ages of Algal Pigment Index 1 (API 1) show-ing the development of an algal bloom betweenFebruary and March 2012 in the coastal watersof southern Portugal. Higher algal pigmentconcentrations shown in red. The inset zoomsin on Cape St Vincent. Source: Christina et al.(2015).

coastal and marine waters (Fig. 2) that arefood for shellfish but also have the potentialto produce toxins from Harmful Algal Blooms.The study demonstrated the usefulness of re-mote sensing for supporting the aquacultureeconomy that is developing in this study areaand for the implementation of the MSFD andMSPFD.

7 Case Study: Adriatic

Valentini et al. (2016) created a map of thesuitability of different parts of the northernAdriatic Sea for the farming of sea-bass andsea-bream. They used remote sensing, in-seameasurements, and models to provide infor-mation on the spatial distribution, and sea-sonal variation, of water characteristics impor-

6: Remote Sensing 6

tant for fish-farming. The data obtained fromsatellite sensors were sea surface temperature(NASA MODIS and ESA Copernicus prod-ucts) and chlorophyll and other OAC - fromESA MERIS).

The paper does not record the softwareused for the mapping, but the computa-tional method seems to be that of a pixel-by-pixel binary analysis. For each pixel inthe map, it was determined whether it showedspace available for aquaculture or already as-signed to another activity; whether the an-nual temperature range was suitable for fishgrowth; whether concentrations of OAC ex-ceeded thresholds; whether water currentswere above threshold values.2

The resulting map (Fig. 3) shows that themost suitable areas for fish farming were off-shore and in the eastern part of the northernAdriatic. The western Adriatic was stronglyinfluenced by the discharge of the Po river,with nutrients that stimulated phytoplanktongrowth and so resulted in chlorophyll concen-trations above acceptable limits for sea-bassand sea-bream. These high concentrations,however, make the western Adriatic particu-lar suitable for the cultivation of filter-feedingshellfish, as explored by Brigolin et al. (2017).

Brigolin et al. (2017) studied the potentialarea for aquaculture along the coast of the ofthe Emilia-Romagna Italian region (NorthernAdriatic Sea) particularly for offshore shellfishculture of the Mediterranean mussel (Mytilusgalloprovincialis). The study assessed: (i) ifthe establishment of new farms in deeper areas,beyond the 3 nautical miles would be benefi-cial for the activity, and if suitable to introducenew longline technology; (ii) the introductionof a new species, such the Pacific oyster (Cras-sostrea gigas); and (iii) the risks for a farm

2 Unit 5, ‘Introduction to AquaSpace integrat-ing tools’ explains pixel-based analysis in more de-tail.

Figure 3: Feasibility analysis for fish aqua-culture in northern Adriatic Sea. Green re-gion is suitable and free of other activities.Source: Valentini et al. (2016).

because of storm events, particularly, in lesssheltered off-shore areas.

Taking these three points into consideration,the methodology used to assess the suitabilityof the site for shellfish was the Spatial Multi-Criteria Evaluation (SMCE), which providedthe framework to combine mathematical mod-els and operational oceanography products.Intermediate level criteria (ILC) considered inthe analysis included optimal growth condi-tions, environmental interactions, and socio-economic evaluation (Fig 4).3

Making use of resources provided by re-mote sensing and operational oceanography insite selection overcame the scarcity of data formodel application. The use of environmen-tal parameters provided by the remote sensingsensor MODIS (Moderate Resolution ImagingSpectroradiometer) between 2003-2012 (e.g.Sea Surface Temperature (SST) and concen-

3 The ‘BlueFarm’ tool developed and applied byBrigolin et al. is similar to the ‘AquaSpace Tool’described by Gimpel et al. (2018) and introducedin unit 5.

6: Remote Sensing 7

tration of Chlorophyll a) were used in the dy-namic models of the present study. The non-linear effects of the different environmental pa-rameters were integrated over the duration ofthe farming cycle.

The results of the SMCE showed that thewhole coastal area between 0 and 3 nauticalmiles was highly suitable for farming of mus-sel, while the area comprised between 3 and 12nautical miles was divided between the highlysuitable northern part, and the less suitablesouthern part.

In addition, seven different scenarios of de-velopment of shellfish aquaculture industrywere explored. The introduction of oysterfarming did not change dramatically the de-gree of suitability for shellfish in the area. Theresults highlighted that: (i) the growth poten-tial in this area was high; (ii) the space withsuitability index > 0.5 increased when prior-itizing the optimal growth condition criteria,and (iii) the socio-economic criterion was themost restrictive of the ILC.

The results from Brigolin et al. (2017) showthat remote sensing can help identify zones forsustainable aquaculture in the MediterraneanSea, where the space for these activities is be-coming increasingly limited in coastal waters.

8 Case Study: Mont St-

Michel bay

Mont Saint-Michel Bay (North Brittany) wasused as a case study by Thomas et al. (2011)to assess the potential of this marine ecosys-tem for shellfish aquaculture and to evaluateits carrying capacity. This study clarified theresponse of exploited species to environmentalvariations using robust ecophysiological mod-els and available environmental data. It sim-ulated the response of blue mussel (Mytilusedulis L.) to the spatio-temporal fluctuations

Figure 4: Information flow, and frameworkadopted in the SMCE. Colours mark thedifferent Intermediate Level Criteria (ILC):(i) ‘optimal growth’ (orange); (ii) ‘environ-mental interactions’ (red); (iii) ‘socio eco-nomic evaluation’ (green). Constraints areshown in blue. Source: Brigolin et al.(2017).

of the environment in this study area by forcinga generic growth model based on Dynamic En-ergy Budgets (DEB) with satellite derived en-vironmental data (i.e. temperature and food).

The sea-surface temperatures were takenfrom nightime infra-red emission measure-ments by the NOAA-18 satellite, and thechlorophyll concentrations were calculated us-ing the OC4 algorithm from the ocean colour(sea-surface reflectance) measurements madeby the NASA SeaWiFS.

The model was applied over nine yearson a large area covering the entire bay andthe simulations provided an evaluation of thespatio-temporal variability in mussel growthand showed the capacity of the DEB modelto integrate satellite-derived data and to pre-dict the variability of spatial and temporalgrowth of mussels. In addition, seasonal, inter-annual and spatial growth variations were alsosimulated. The results showed a strong linkbetween food (satellite-estimated chlorophyll)and mussel growth.

6: Remote Sensing 8

9 Case Study: Nova Sco-

tia

This case study, provided by Professor JonGrant of the AquaSpace partner, DalhousieUniversity, concerns the use of low-altitudeaerial photography from balloons and dronesto visualise and map features in the vicinity offarms.

TEXT AWAITED

10 Exercises and SAQ

Exercise

The exercise involve the use of SNAP (SentinelApplication Platform) software to examine aSentinel-3 image obtained from EUMETSAT.It is detailed in Appendix A.

Further reading

Study one of the (open-access) papers associ-ated with one of the Case Studies:

Algarve : Christina et al. (2015);

Adriatic : Valentini et al. (2016) or Brigolinet al. (2017);

Mont St Michel : Thomas et al. (2011).

Self Assessment Questions

The SAQs that follow test your achievementof the learning outcomes and help you thinkactively about the points raised in this lec-ture. No answers are given. The (parenthet-ical) numbers refer to sections and the lettersto appendices.

1. What satellite data would help an aqua-cultural planner or developer to find afavourable environment for the cultivationof mussels? (7)

2. How are optical measurements made by asatellite converted to maps of chlorophyllconcentration? (4)

3. What useful information was gained byusing remote sensing to monitoring aqua-culture concessions in the coastal watersof the Algarve coast and in the bays ofNova Scotia? (6, 9)

6: Remote Sensing 9

4. What were the contributions of remotesensing data to the case studies in theNorth Adriatic and Mont Saint-Michelbay? (7, 8)

5. Describe the main steps of the methodfor obtaining a processed image ofChlorophyll-a concentrations. (A)

Appendices

A Use of SNAP with

Sentinel-3 data

SNAP is the ESA Sentinel ApplicationPlatform that provides analysis of the scien-tific data from the ERS-ENVISAT missions,the Sentinels 1/2/3 missions and a range ofnational and third party missions. Remotelysensed images are stored as arrays of data re-lating to pixels in maps. The exercise involvesdata that have already been subjected to someprocessing, to correct for atmospheric changesto the sea-leaving signal and to convert the lat-ter into variables of interest, such as sea surfacetemperature (SST) or – in the present exercise– the phytoplankton photosynthetic pigmentChlorophyll a (Chla). SNAP allows such dataarrays to be explored and manipulated at dif-ferent processing levels.

The exercise will have been completed whenthe student has obtained at least one chloro-phyll image for a coastal region of their choice.

A.1 Install SNAP

Install the SNAP software from the ESA tool-box site at step.esa.int/main/toolboxes/snap,which provides guidance. Chose the ‘SentinelToolboxes’ and then follow the normal prac-tice for your computer’s OS to install the pro-gramme. Note where it was placed.4

A.2 Get Sentinel-3 data

Although the SNAP app includes the abilityto access ESA data as one of its functions, itis suggested, here, that the student visits the

4In the case of Mac OS X, it is most likely inApplications/snap/bin.

6: Remote Sensing 10

EUMETSAT site and downloads data indepen-dently of SNAP. In some cases this will requirea user name and password.5 A directory orfolder should be set up to hold the (large) filesbefore they are loaded into SNAP.

Tutorials are available from the SNAP web-site at step.esa.int/main/doc/tutorials. Or, godirectly to YouTube for How to Visualise Sen-tinel 3 Data. As shown in this video,

• log in to the EUMETSAT site atcoda.eumetsat.int/#/home;

• select region of interest on the map thatis presented, and in the search crite-ria menu select OL 2 WFR from ‘ProductType’, OLCI from ‘Instrument’, and L2

from ‘Product Level’, to get Sentinel-3Level 2 Full Resolution water and atmo-sphere geophysical products from OLCI;

• you may be presented with many datasets, each from one pass of the satellite;identify the data set you want;

• download the data, which will be providedas a zip file: decompress it and store theresulting folder in the location you haveidentified for these data.

A.3 Open data in SNAP

Sentinel-3 products are provided as a collec-tion of files contained within a folder. Thefolder name is the actual product name, end-ing in .SEN3. Each folder contains a metadatafile named xfdumanifest.xml and at least onenetcdf-file. Each netcdf-file contains a subset ofa Sentinel-3 product’s content.

There are several ways to open a Sentinel-3product within SNAP. Use this one:

• Choose File → Open Product,

5 In the case of EUMETSAT, a new user can setup an account when first visitng the web site.

• navigate to the xfdumanifest.xml file inyour downloaded folder, and click Open ;

• the name of the folder will appear in theProduct Explorer window, with sub-folders that include Bands;

• ctrl-click or right click on the foldername to bring up a menu;

• select Open RGB Image Window , ac-

cept Tristimulus profile, and click OKto bring up an image in SNAP’s right-hand window.

The result may look different from that inthe video; a cloud-covered image, for example,will be mainly white.6 Continue as follows tooutput a chlorophyll image:

• in the Product Explorer window, →Bands → CHL → CHL OC4ME

• ctrl-click or right click on the file nameto bring up a menu;

• select Open Image Window to displaythe chlorophyll image;

• select (in main menus) Layer → Layer

Manager;

• in the Layer Manager window selectMasks → WQSF Isb CLOUD; close LM win-dow;

• select (in main menus) View → Tool

Windows → Colour Manipulation;

• In Colour window, select Table and clickon colours to change the colour scale dis-played for chlorophyll.

By this time, your SNAP screen should looksomething like Figure 5. The chorophyll imagecan be exported using the menu generated byctrl-click or right click .

6If the image is cloud-covered, it will be desir-able to return to EUMETSAT for another data set.

6: Remote Sensing 11

Figure 5: Screen grab from SNAP run-ning under Mac OS X, showing an imagecentered on the Strait of Gibraltar on 16November 2017

6: Remote Sensing 12

B NASA OceanColor

site

The NASA OceanColor Web site provides ac-cess to NASA data from a variety of mis-sions. If you use that link to enter thesite’s front page, chose Data → Level 1 and

2 browsers. Or follow the instructions belowto see global maps of chlorophyll concentra-tion or sea-surface temperature for a selectedmonth and year, derived from MODIS.

• Go to NASA Ocean Color site pageat oceancolor.gsfc.nasa.gov/cgi/browse.pl?sen=am

• set options to to CHL , MODIS aqua

and day ;

• select a month (in this case, August) andyear (in this case, 2017), to display aglobal map for that month;

• set option to SST for a temperaturemap;

• if you want to do more, use the help forguidance.

This NASA web page also provides accessto older data, including those from SeaWiFS(select MLAC) and CZCS.

References

Bowers, D., Harker, G., Smith, P., and Tett,P. (2000). Optical properties of a region offreshwater influence (the Clyde Sea). Es-tuarine, Coastal and Shelf Science, 50:717–726.

Brigolin, D., Porporato, E. M. D., Prioli,G., and Pastres, R. (2017). Making spacefor shellfish farming along the Adriaticcoast. ICES Journal Of Marine Science,74(6):1540–1551.

Christina, S., Icely, J., Goela, P. C., DelValls,T. A., and Newton, A. (2015). Using remotesensing as a support to the implementationof the European Marine Strategy Frame-work Directive in SW Portugal. ContinentalShelf Research, 108:169–177.

Devlin, M. J., Barry, J., Mills, D. K., Gowen,R. J., Foden, J., Sivyer, D., and Tett, P.(2009). Estimating the diffuse attenuationcoefficient from optically active constituentsin UK marine waters. Estuarine, Coastaland Shelf Science, 82:73–83.

Gimpel, A., Stelzenmuller, V., Topsch, S., Gal-parsoro, I., Gubbins, M., Miller, D., Muril-las, A., Murray, A. G., Pınarbası, K., Roca,G., and Watret, R. (2018). A GIS-based toolfor an integrated assessment of spatial plan-ning tradeoffs with aquaculture. Science ofthe Total Environment, in press:12. DOI:10.1016/j.scitotenv.2018.01.133.

Governo Portugal (2013). Estrategia Na-cional para o Mar 2013-2020. Technicalreport, Government of Portugal, Lisbon.Available from: www.dgpm.mam.gov.pt/Documents/ENM.pdf.

Grant, J. (unknown). The importance of spa-tial perspective in aquaculture management.

6: Remote Sensing 13

NSERC-Cooke Industrial Research Chairin Sustainable Aquaculture CHECK. Re-port, Department of Oceanography, Dal-housie University, Halifax, Nova Scotia.

Hardy, A. C. (1956). The Open Sea: its NaturalHistory. Part I: the World of Plankton. NewNaturalist 34. Collins, London.

Holligan, P. (1981). Biological implications offronts on the northwest european continen-tal shelf. Philosophical Transactions of theRoyal Society of London, A 302:547–562.

Holligan, P., Aarup, T., and Groom, S. (1989).The North Sea: Satellite colour atlas. Con-tinental Shelf Research, 9:667–765.

IOCCG (2000). Remote Sensing ofOcean Colour in coastal, and otheroptically-complex, waters. Report,S. Sathyendranath ed., vol. 3, Inter-national Ocean Colour CoordinatingGroup, Dartmouth, Canada. Available at:www.ioccg.org/reports/report3.pdf.

Pozdnyakov, D. and Graßl, H. (2003). Colourof Inland and Coastal Waters - A Method-ology for Its Interpretation. Marine Sci-ence and Coastal Management series, ed.Lang, A. Springer-Verlag, Berlin, Heidel-berg. xxii+170 pages.

Simpson, J. and Hunter, J. (1974). Fronts inthe Irish Sea. Nature, London, 250:404–406.

Simpson, J. H., Allen, C. M., and Morris, N.C. G. (1978). Fronts on the continental shelf.Journal of Geophysical Research: Oceans,83(C9):4607–4614.

Thomas, Y., J., M., Alunno-Bruscia, M., C.,B., Bouget, J.-F., Gohin, F., Pouvreau, S.,and Struski, C. (2011). Modelling spatio-temporal variability of Mytilus edulis (L.)growth by forcing a dynamic energy budget

model with satellite-derived environmentaldata. Journal of Sea Research, 66(4):308–317.

Valentini, E., Filipponi, F., Xuan, A. N., Pas-sarelli, F. M., and Taramelli, A. (2016).Earth Observation for Maritime SpatialPlanning: Measuring, observing and mod-eling marine environment to assess poten-tial aquaculture sites. Sustainability, 8(6 no519):24.