699
1 INTRODUCTION
SatelliteEarthObservations(EO)forover40yearsare
usedinenvironmentalmonitoring.Itisestimatedthat
currently about 14% of satellite systems in orbit are
designedtoprovideEOdata,whichgivesabout225
systems in orbit [15]. Each year, new systems with
variousparametersandfunctionsarepla
cedinorbit.
The vast quantity of earth observation data brings
enormousopportunities,butalsochallengesinterms
ofchoosingandusingthemostappropriatedata.For
purposesrelatedtothemarinedomain,thebenefitsof
EOdatadonot onlyresultfrom thevolumeof data
available, but from the int
elligent use of data in a
targetedmanner.
Astheoceancoversapproximately360millionsq.
km. of the Earth’s surface, it is highly important
having a tool which provide information from all
around the world. This is a favourable time for
satellitemonitoringofbothoceansandlands,tha
nks
to the Copernicus Program, a European Union
flagship initiative. Copernicus is coordinated and
managed by the European Commission. This long
termprogramaimstoprovideanoperationalsatellite
monitoring capability and related services for the
environment and security. The information services
providedarefreelyandopenlyaccessibletoitsusers.
Cha
nges in the oceans affect not only everyday
weatherandlocalchangesinfisheries,butalsohave
global effects in the form of large‐scale atmospheric
circulation, climate change and marine ecosystems.
TypicalEOproductsgivesdataonoceanparameters,
suchasseasurfacetemperature,seaice,chlorophyll‐a
concentration (phyt
oplanktonʹs photosynthetic
pigment) and photo synthetically active radiation
(sunlight effective for the photosynthesis of plants),
basedonvariousobservationdatasets.Thecontinued
advancementofsatellitetechnologyisexpandingthe
range of oceanparameterswhich can be monitored,
but also gives opportunity to detect and identify
objects at sea, like oil spill detection or ship
movementinharbours.
Earth Observation Opportunities to Enhance Maritime
Safety
B.Weintrit
WarsawUniversityofTechnology,Warsaw,Poland
A
striPolska,Warsaw,Poland
ABSTRACT:Inthispaper,authorpresentscapabilitiesofusingsatellitedata,bothradarandopticalimages,to
enhance maritime safety. The concept of using satellite data includes a periodic data acquisition from the
European satellite systems, both commercial and public. The aim of this paper is to show modern space
technologies and the adv
antages of using them in the maritime domain. Advantages and disadvantages of
earthobservationareidentifiedanddescribed.Authoroutlinesopportunitiesinmaritimesafetydomain.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 11
Number 4
December 2017
DOI:10.12716/1001.11.04.17
700
2 SATELLITESYSTEMS
The basic parameters characterizing satellite images
are: the number of spectral bands in which the
recordingwasmade,therangeandresolutionofeach
band,aswellasthespatialandtemporalresolution.
Due to the type of registered electromagnetic
radiation, satellites can be divided into optical and
microwa
vesystems[12].Anotherdivision,duetothe
resolution of the obtained image, allows to
distinguishveryhighresolution(VHR)systemswith
apixelsizesmallerthan1m,highresolutionsystems
from1to4mpixel,mediumresolutionfrom4 to12
m, rather low resolution from 12 to 50 m and low
resolutionab
ove50m[5].
2.1 Copernicusdata
Copernicus gives a unique set of observations from
Sentinel satellites, which provide unmatched spatial
and temporal coverage with different sensors and
resolutions.
TheSentinelmissions supportmarinemonitoring
by providing timely, continuous and independent
dataonthebeha
viour,use,andhealthoftheoceans
and the associated coastal zones. Each Sentinel
missionisbasedonaconstellationofmorethanone
satellite, to fulfil revisit and coverage requirements
and has specific features, like a certain resolution,
revisittimeandsensors.
Figure1.Sentinelmissionandstatus(source:[4])
Sentinels provide datasets for Copernicus
Services:
− Sentinel‐1 is a Synthetic Aperture Radar (SAR)
with the prime objectives of Land and Ocean
monitoring.Thegoalofthemissionistoprovide
C‐Band all‐weather, day‐and‐night radar data
continuity following the end of the ERS‐2 and
Envisat mission. Its data is very valuable for the
observation and study of: management and
monitoringof the European marine environment,
seaicemeasurement,surfacewindsandwaves,oil
spillsdetection,landsurfacemovements,etc.
Sentinel‐2isaMultiSpectralImager(MSI)whose
databenefitsinmarineapplications,amongothers
theobservationandstudyofcoastalconditions.
Sentinel‐3includesthreemaininstruments:aradar
altimeter, an Ocean and Land Colour Instrument
(OLCI)andaSeaandLandSurfaceTemperature
Radiometer (SLSTR). Its main mission is to
provide a multi‐instrument capability to support
the accurate parameterization of such topics as
OceanColour(OC),SeaSurfaceHeight(SSH)and
Sea Surface Temperature (SST). Its data products
address almost all domains of Earth Sciences,
includingOceanography.
Sentinel‐4 and Sentinel‐5 are in preparation and
will be missions dedicated to the atmosphere
composition.
2.2 Thirdpartymissiondata
Around30fromESA,theirMemberStates,Eumetsat
and other European and international third party
missionoperatorsmakesomeoftheirdata available
for Copernicus Services. In majority, these data are
characterized by higher spatial resolution but also
smaller scene size. This means, that those high
resolution satellite data cannot be so easily used for
bigareamonitoring,butrathertoconfinesstudiesto
individualsites.
2.2.1 Radarsatellites
TerraSAR‐X satellite is the most accurate,
commercial satellite for radar observations of the
Earth built in a public‐private partnership: German
Space Agency, DLR and EADS Astrium (currently
AirbusDefense&Space).TerraSAR ‐Xusesradarwith
active antenna working in frequency 9.65 GHz (X
band, wavelength 3.1 cm) [1]. It is the length of the
emittedwaveandexcellentknowledgeofthelocation
allow, among others, extremely accurate detect
objectsonseasurface.
Table1. TerraSAR‐X available modes characteristic
(source:[1])
_______________________________________________
ModeCoverageAzimuthResolution
xRange(km
2
) Class(m)
_______________________________________________
ScanSARWide(SCW) 200x(194–266) 40
ScanSAR(SC) 150x10018
StripMap(SM)50x30 3
Spotlight(SL) 10x10 1.7‐3.5
High‐Resolution 5x10 1.4‐3.5
Spotlight(HS) 5x(5‐10)1.1‐1.8
300MHzHigh‐Resolution
Spotlight(HS300)
StaringSpotlight(ST) (2.5–2.8)x~6 0.24azimuth,
1.0range
_______________________________________________
Differentmodes giveopportunitytoacquire data
with smaller resolution for large area monitoring,
detect shoreline and recognize and identify objects
and infrastructure with very high resolution. Wide
ScanSARmodewithalargecoverage(upto270x200
km
2
)andaspatialresolutionof40misauniqueratio
between coverage and resolution, offering a unique
potential.
2.2.2 Opticalsatellites
For better understanding of radar images, it is
helpful to use optical data for photointerpretation.
Someveryhighresolutionopticalsatelliteswithmain
parametersarepresentedbelow.
701
Table2.Veryhighresolutionopticalsatellitessystems
______________________________________________________________________________________
Satellite/Sensor From To Spatialresolution[m]Numberofbands Swath[km]
______________________________________________________________________________________
IKONOS1999 2015 0,82(PAN) 3,28(MS) PAN+4MS11
QUICKBIRD 2001 2015 0,61(PAN) 2,44(MS) PAN+4MS16,5
GEOEYE‐120080,41(PAN) 1,65(MS) PAN+4MS15,2
WORLDVIEW‐2 20090,5(PAN) 2(MS)PAN+8MS16,4
WORLDVIEW‐3/4 2014/2016
0,31(PAN) 1,24/3,7(MS) PAN+8MS+8SWIR 13,1
PLEIADES‐1A/‐1B 20110,5(PAN) 2(MS)PAN+4MS20
______________________________________________________________________________________
Spatialresolutionindicatesthesizeofthesmallest
objectthat can be detected in the image. Comparing
thesesystemswithSentineldata,thetrendisclearly
visible,thebettertheresolution, the smallerthesize
ofthescene.
The Pléiades satellite system is a constellation of
veryhighdefinitionsatellitesand
atthesametimethe
first European system of this class. The Pléiades
system is in fact a constellation of the two twin
satellites Pléiades‐1A and Pléiades‐1B. The first of
thesewasplacedinorbitonDecember17,2011,and
the second on December 4, 2012. The constellation
belongs to Airbus Defense & Space‐the largest
European space technology company. The Pléiades
satellites are equipped with multispectral sensors R,
G,B,NIRwithapixelof2.0mandpanchromaticwith
apixelof0.5m.
WhatdistinguishesthePléiadessystemfromother
VHR systems is a strong sensitization
to the blue
range.Sensitizationofthesensorstartsfrom430nm,
whileusuallyitisonly450nm.Thepositiveeffectof
this is the increase in possibilities of studying water
bodies. In addition, images are recorded in 12‐bit
what gives a very large tonal distinction between
objects.
Thisfeatureallowsyouto analyseobjectsin
shadows and low contrast background in detail. No
othersystemcurrentlyregisters12‐bitimages.
3 MONITORING
As it was well recognized in [10], marine transport
development causes the increase of the intensity of
ships traffic and ships dimensions. The navigational
riskincrease
requiresmoresophisticatedmethodsof
itsassessment.
TheEuropeanMaritimeSafetyAgency(EMSA)is
theEntrustedEntityresponsibleforimplementingthe
CopernicusMaritimeSurveillance(CMS)service.The
CMS service is available for European Union (EU),
European Free Trade Association (EFTA) national
administrations with responsibilities at sea, and
relevant EU bodies
and institutions [3]. The CMS
servicesupportsmonitoringofhumanactivityatsea
for a range of functions, including: maritime safety
and security, marine pollution monitoring and law
enforcement. It supports authorities by value added
productsextractingparticularlyvaluableinformation
fromthebasicimagedata,analysisofobjects,features
or activities
at sea and allowing them to undertake
morequicklyandefficientlyactivitiesanddecisions.
ThroughCleanSeaNetService,EMSAprovidesan
oilspillmonitoringservice.CleanSeaNetsupplements
existing surveillance systems at national or regional
level, strengthens member state responses to illegal
discharges, and supports response operations to
accidental spills [6]. The
service is based on radar
imagesobtainedfromsyntheticapertureradar(SAR)
satellites. Operators evaluate the satellite images,
together with supporting meteorological,
oceanographic and other available data. Aim is to
detect and identify possible sea pollutions, or to
assesstheprobabilityofthepresenceofoilonthesea
surface,help
identifythesourceofcontaminationand
model the oil drift. When a spill is detected, a
pollutionalertissenttonationalauthorities.
4 OPPORTUNITIES
Thevalueaddedproductsbasedonsatellitedatafor
marinesafetyinclude:
vesseldetection,
oilspilldetection,
activitydetection,
SARwind
andSARwave.
Opticalsensorsprovideavarietyofinformationin
differentspectralbands.Imagesprovideeasierimage
interpretation due to the colour combination of red,
green,blue(RGB)bands.Opticalradiometerscannot
acquire images during the night or in cloud cover
conditions.
Synthetic‐aperture radar sensors use microwave
frequencies to
retrieve backscatter measurements
fromthedetectedsurface,includingseasurface.The
images can be acquired independent of the weather
condition and cloud cover, at any time of day or
night. SAR images of the ocean greatly depend on
surface roughness caused by wind stress at the sea
surface. By measuring
the roughness of the sea
surface,resultingimagesdisplayfeatureswhichstand
outagainstthebackground.
4.1 Vesseldetection
SARimages results of vessel detection analysis, and
vessel traffic information and other man‐made
structuresappearasbrightspots.
On Figure 2 ship detection is made based on
TerraSAR‐X very
high resolution data. In vessel
detection higher resolution SAR images are more
efficient.Keybenefitsare:
Quickavailabilitythrough
cloudandday‐lightindependence
(semi‐)automaticprocessing
nearreal‐timedelivery
Highsensitivitytoalsosmallerships
Highgeolocationaccuracy
Highmonitoringfrequency
702
Figure2. TerraSAR‐X Staring Spotlight ship detection
(source:AirbusDS)
Figure3. Pleiades ship detection and identification (source:
AirbusDS)
Shipdetectionresultscanbeusedtofurtherutilize
alsoopticalsatellitesensorsinordertoidentifyships
detectedusingSAR data.ExampleofPleiadesimage
interpretation is presented on Figure 3. With 0.5 m
spatial resolution, it is possible to recognize vessel’s
elementslikechimney,topofgangway,rail,etc.
and
measure the exact dimension of this elements from
image. Having those detailed information it is
possibletoidentifytheshipmodelandevennameof
thevessel.
Whatʹs more, the shipʹs accurate position is also
identifiable.OnthepictureEnricoIveoliispresented.
Itscoordinatesandmetric
information:
Latitude: 071329N
Longitude 0493511E
Length138m
Width 22m
Oriented 053
Radar and optical data complement each other
verywell.However,theopticaldataisdependenton
weather conditions, therefore their availability is
somewhatlimited.
4.2
Oilspilldetection
Inthe caseofoilspills or areaswithvery low wind
appearasdarkshapes.Thesurfaceroughnessofthe
seahasalargeimpactonthedetectionofoilspotson
thebasisofSARimages.Thisimpactismuchgreater
thanwhendetectingbrightobjects.

On Figure 4 the example of oil spill recognition
during different weather conditions is presented.
PictureshowsoilspilldetectionbasedonTerraSAR‐X
radar data. On the first image, on the top, it is not
possible to detect oil spills. Weather conditions are
defined as rough weather. On the
second one seeps
arealsonotvisible.Becauseofcalmweathertheseeps
areindistinguishable.Lastpicturepresenttheperfect
weatherconditionsandoilisclearlyvisible,bydark
shape.
Theidealweatherconditionsmeans:
Windspeed:between6to16knots,
Astablewinddirection,
Swellheight:
<1.5m,
Waveheight:<1.0m.
Figure4.Oilspilldetectiondependingonsurfaceroughness
(source:AirbusDS)
703
4.3 Dailymonitoring
Daily revisit of Pleiades satellites gives opportunity
forfrequentmonitoringthetrafficintheharbour.On
Figure5changesintheSuezHarbourarevisible.
Figure5.DailymonitoringbasedonPleiadesimages,Suez
harbor, Egypt, 2013, February 20 (upper) and February 21
(bottom)(source:AirbusDS)
Especiallyforshipowners itcan be aninteresting
source of information about the current situation in
theport,andonthesea.
5 CONCLUSION
Mixing radar and optical satellite data gives more
opportunity in monitoring and detecting objects.
Radar can see where optical sensors cannot. On the
otherhandopticalsensors
candetectsmallerobjects.
Example of combining radar data for detection and
opticalforidentificationwaspresentedinthispaper.
Anotheradvantageisthemorefrequentleveragingall
sensorstoafforddailyandevenintra‐dailyrevisit.
Moreover combining freely available Copernicus
data with available on request commercial data
allows
to increase capabilities its usage in maritime
domain.Copernicushasfullfreeandopenpolicyfor
data and information and provide continuous
environmental information. Detection and
identificationmightbedoneusingcommercialimages
withhigherresolution.
Theadvancementofthesatellitebasedtechnology
andtheabilitytomonitoroceansalinity,temperature
andmanyotherfactorswillshapethedevelopmentof
future satellite sensors. These satellite observations,
used in conjunction with in situ measurements and
models,willbecomeakeyelementinunderstanding
andassessingoceans.Whileimportantforadvancing
ocean monitoring capabilities, satellite Earth
observation will by no means replace in
situ
observations as they are needed to evaluate satellite
data.
REFERENCES
[1]Airbus DS 2014. TerraSAR‐X Image Product Guide.
http://www.intelligence‐airbusds.com/files/pmedia/
public/r459_9_201408_tsxx‐itd‐ma‐0009_tsx‐
productguide_i2.00.pdf;access:8.03.2017
[2]ColeuK.2015.SatelliteMonitoringOceanmeasurements
from space. https://www.oceanfdn.org/sites/.../files/
Satellite%20Monitoring.pdf,access:14.07.2017
[3]EuropeanMaritimeSafetyAgency:CopernicusMaritime
Surveillance Product Catalogue
http://marine.copernicus.eu/services‐portfolio/access‐to‐
products/,access21.06.2017
[4]Copernicus Programme
webpage
http://www.copernicus.eu/,access20.06.2017
[5]DermanisA.,BiagiL.2013.Satellites,SensorsandData
Formats. http://geomatica.como.polimi.it/corsi/rs_ia/
C03Satellites.pptx.pdf;access:6.02.2017
[6]
EMSA Earth Observation Services webpage:
https://portal.emsa.europa.eu, access: 14.07.2017
[7]ESA webpage https://earth.esa.int, access20.06.2017
[8]Flores A.P. 2014. Remote sensing applied to ground
water resources. http://aulavirtual.ig.conae.gov.ar/
moodle/pluginfile.php/513/mod_page/content/109/flores
_seminario.pdf;access:2.02.2017
[9]GoldsteinR.M.,ZebkerH.A.,WernerC.L.1988.Satellite
radar interferometry: Two‐dimensional phase
unwrapping.RadioScienceVol.23,No.4,pp.713–720.
[10]Gucma L., Marcjan K. 2012. Incidents Analysis on the
Basis
of Traffic Monitoring Data in Pomeranian Bay.
TransNav, the International Journal on Marine
NavigationandSafetyofSeaTransportation,Vol.6,No.
3,pp.377‐380.
[11]GyslingD.L.,BostickF.X.2002.
Changing paradigms in
oil and gas reservoir monitoring - the introduction and
commercialization of in-well optical sensing systems.
The 15th Optical Fiber Sensors Conference Technical
Digest (OFS-15), Portland, OR, pp.43‐46
[12]KurczyńskiZ.2006.Lotniczeisatelitarneobrazowanie
Ziemi. Warszawa, Oficyna Wydawnicza Politechniki
Warszawskiej.
[13]Land, P.E., Shutler J.D., Findlay H.S., et. al. 2015.
SalinityfromSpaceUnlocksSatellite‐BasedAssessment
of Ocean Acidifi cation. Environ. Sci. Technol., 49 (4),
pp.1987–1994DOI:10.1021/es504849s.
[14]Lillesand T.M.,
Kiefer R.W., Chipman J. 2000. Remote
Sensing and Image Interpretation, fourth edition. New
York,JohnWiley&Sons,Inc.
[15]Report2016.StateoftheSatelliteIndustryReport.The
Tauri Group. http://www.sia.org/wp‐content/uploads/
2016/06/ SSIR16‐Pdf‐Copy‐for‐Website‐Compressed.pdf;
access:10.02.2017