567
1 INTRODUCTION
In the 1980s, computer software for geospatial data
(better known as GIS software) had been on the
market for more than a decade, but hardware,
especially “highresolution” graphic screens were
veryexpensiveandmostsoftwareveryspecialized.It
was still early days in the marine geospatial world,
and much would happen before geospatial data
wouldbecomemainstream.
Around the mid 1980s several national
hydrographicofficesacquiredgeospatialsoftwarefor
chart production and internationally there was talk
about geospatial standards. Under IHO was a
working group responsible for defining new feature
classesbasedondatafoundonnavigational
charts.In
1987wasthefirstworkreleasedas“DX87”(Digital
eXchange87).Itdidn’tgetmuchattentionaroundthe
world,butthedevelopmentwas(veryslowly)rolling.
Electronic navigational charts was coming, but
manywouldarguethatitwouldhavetakenforeverif
itwasn’tforonespecial“event”:
OnMarch24,1989,
theoiltankerExxonValdezranagroundontheBligh
Reef,PrinceWilliamSound,Alaska,USA,resultingin
anoilspillestimatedtohavebeenaround40,000tons
[4].Thereisprobablynotapersontodayinvolvedin
electronicchartsthathaven’theardthisstory.
Whether
(oil)accidentsactedasa catalystfor the
developmentofelectronicchartsandchartsystemsor
not is maybe less important! Many people in the
industry and government organizations agreed that
electronic charts and electronic navigational systems
were the future. In 1990 DX87 had become DX90
and included in
a new IHO standard named S57
‘IHO Transfer Standard for Digital Hydrographic
Data’.Atthistimesomearguedthatelectroniccharts
would replace paper navigational charts within 5
Towards a Universal Hydrographic Data Model
H.Astle&P.Schwarzberg
CARIS,Fredericton,NB,Canada&CARIS,Heeswijk,TheNetherlands
ABSTRACT: The International Hydrographic Organization’s (IHO) [1] Transfer Standard for Digital
HydrographicDataS57[2]standardhasbeeninforceformorethanadecade,andhassuccessfullybeenused
for official ENCs adopted by Hydrographic Offices around the world
and by navigation equipment
manufacturers.AdditionallyS57hasbeenusedformanyadditionalpurposes.HoweverS57,andespecially
theadministrationofthestandard,hasalsoexperiencedlimitations.In2010,IHOreleasedthenextgeneration
hydrographicstandard calledS100 Universal HydrographicData Model [3]. A movethat will open
up the
door to new possibilities to existing S57 users and potentially broaden the use of IHO standards in the
hydrographiccommunity.
ThispaperwilltrytoexplainwhyanS57replacementwasneededandgiveexamplesonsomepossibilities
withS100anditsderivedproductspecifications
suchasS101.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 7
Number 4
December 2013
DOI:10.12716/1001.07.04.12
568
years. More than 20 years later that has still not
happened,butit’scertainlycloserthanever.
2 S57TAKINGOFF
With the introduction of the IHO S57 standard in
1990 the world had probably received its first truly
international recognised data exchange standard.
Thereweretwomajor
purposeswiththestandard:
1 1Exchangeofdigital paperchartbetweennational
hydrographicoffices
2 2Exchange of electronic charts from national
hydrographic offices to mariners’ navigational
systemsonboardships.
Thefirstpurposenevertookoff.Themajorreason
for this is probably that paper charts are graphical
products and a geospatial standard focusing on
feature encoding of “real world features” like S57
did, was not suitable for this purpose. Converting
graphical datafrom allthe various systems used by
hydrographic office around the world is far more
complex than most people expected at that time. In
fact
paperchartstodaymostly,ifnotentirely,arestill
exchangedasrasterdataandnotasvectororfeature
data.
ThesecondpurposeusingS57forelectronicchart
exchangehadmore progress! Therewasmuch tobe
learnedandduringthefirstyearsaftertherelease,the
IHO DataBase
Working Group that maintained the
standard met multiple times per year, and the
standardwas constantly under revision. However in
themid1990sthestandardhadmatured,andby1996
edition 3.0 was released. It was decidedthis edition
would remain frozen for 4 years to ensure stability,
intended to help
hydrographic offices and, maybe
especially, navigational equipment manufactures to
finalize and release date plus equipment for the
market.
Only one year earlier, in 1995, had the
International Maritime Organization (IMO [5])
adopted the socalled ECDIS Performance Standard
[6], meaning that Electronic Chart Display &
InformationSystems(ECDIS)nowwas
allowedtobe
usedfornavigation,insteadofpapercharts.Herealso
were many lessons to be learned, and amendments
were frequent in the early years. Work on this
standardhadtakenplaceinternationallyforaboutas
longasthedataexchangestandardandthetwohave
closeties.TheECDIS
standardrequiredthatchartsfor
ECDIS systems must be Electronic Navigational
Charts(ENCs)accordingtotheS57standard.
In 2000 edition 3.1 of the S57 standard was
released and this time the standard was frozen
“forever”. The production of ENCs still lacking
behindformany partsof the
world,and it wasalso
expensive for the ECDIS manufactures to keep
updating their systems. Freezing the standard was
intendedtohelpthat.Additionalitwasbelievedthat
onceanECDISwasdeliveredtoavesseltherewasno
guaranteethatsuchsystemwouldbeupdated,e.g.to
support new versions
of the standard. Neither the
ECDISstandard,northeENCstandard,hadmeansto
update ECDIS systems to accept new feature
classification or new symbology etc. This worked
relatively well for about half a decade, but by mid
2000s there were new requirements from IMO to
include new (environmental) features
on navigation
charts.InJanuary2007thesocalled‘Supplement1’to
S57 was released, and it was followed by another
supplementtwo years later.The evidencewas clear:
A frozen standard would not be able to support
futureneeds.
ItwasnotreallyaproblemwiththeS57
standard,
which originally included (technical) possibilities to
defineandencodenewfeaturetypes.Itevenincluded
away toencodeclass definitions ofnew features so
that navigational systems could “learn” about such
new features and present these features to the
mariner.Howeverthispartofthestandardwasnever
utilised.
Additionally S57 was intended to be used for
multiple product types, and had room for various
productdefinitionstobedefined. TheENCstandard
is actually “just” an appendix to the S57 standard.
Howeverthatpossibilitywasn’tused;partlybecause
ofthefrozenstateofthestandard.At
leastnotunder
IHOandofficiallynotpartofthestandard,butmore
about that below! The S57 standard and the ENC
productspecificationhadbecomealmostinseparable
andsomewhatsynonymous!
3 MULTIPLEPURPOSES
The frozen and somewhat limited S57 and ENC
specification did not stop other S
57 based
implementations.One of the earlier ideas for S57
wasforbathymetricsurveydataexchange,andeven
thoughthishasbeendonetosomedegree,therewas
much more use of the standard for other purposes.
Howevernoneofthesepurposesaredevelopedunder
IHO,butsomeofthem
startedusingS57asearlyas
themid1990s.
SomeoftheotherusesofS57are:
InlandENCs[7]
Electronic Charts for Inland ECDIS, which are
usedonrivers.
In 2001 the Economic Commission for Europe of
theUnitedNations(UNECE)adoptedtheInland
ECDIS
Standard as a recommendation for the
European inland waterway system; using Inland
Electronic Navigational Chart (IENC) data.
Outside Europe, other countries also looked at
Inland ECDIS and the U.S. Army Corps of
Engineers [8] developed the Inland Electronic
Navigation Charts. While the European Inland
ENCstandardextendedtheoriginalS57
standard
with new features, symbology and rules, the US
InlandENCstandardusedS57moreorlessasit
was. To align the standards the International
InlandENCHarmonizationGroup(IEHG)formed
in2003andthestandardsarenowmaintainedas
one standard. Other countries adopted this
standardand
InlandENCscoveringthousandsof
riverkilometresexiststoday.Countriesusingthis
standard include: Austria, Belgium, Brazil,
Bulgaria, China, Croatia, Czech Republic, France,
Germany, Hungary, Italy, Netherlands, Peru,
Poland, Russia, Serbia, Slovakia, Switzerland,
569
Ukraine, South Korea, USA, Venezuela (plus
others).
NotethatIENCsarenotoverlays;theywillnotbe
used at the same time as ENCs. They cover
different geographic areas and are made for
differentvessels.
IceENCs[9]
UsingS57 forencodingof Icefeatures started in
the
mid 1990s. These datasets are overlays with
additional dynamic (ice) information
supplementing ENC data.. Countries around the
North Pole, followed by countries around the
BalticSeaareamongtheplayers.
AdditionalMilitaryLayers(AMLs)[10]
Situation awareness layers of data supporting
military operations, which as the name indicates
are
overlay of additional data displayed on a
warship ECDIS (called WECDIS). Date layers
include bathymetric contours, routes, areas &
limitsfordangerandexerciseareas,fullwreckand
major bottom object information, detailed beach
andseabedenvironmentaldata,etc.
TherearemoretypeslikeBathymetricChartsand
specialised Pilot Charts. Many
more possibilities
using S57 for various purposes exist [11] and are
often generically referred to as Marine Information
Overlays(MIOs).Eachtypicallyhastheirownfeature
definitions,encodingrules,etc.Howeverthefactthat
theyareallbasedonS57makesitrelativelyeasyto
support more of
them. For instance a widely used
ENCproductiontool,CARISS57Composer[12],not
only allows users to create IHO ENCs but also the
otherS57producttypesmentionedabove.Userscan
evendefinetheirownS57producttypes.
4 WHAT’SNEXT?
ThereisnoquestionS57
willbeusedformanyyears
tocome.Thereissimplytoomuchdataandsomany
systemsusingS57preventingitfromdyinganytime
soon. ECDIS systems will be using S57 ENCs for
manyyearstocome.
Howeverthegeospatial world hasbeen evolving.In
the1980s
andeveninthe1990selectronicgeospatial
data was not mainstream, but today almost every
mobilephonehas not onlyelectronicmaps, butalso
navigation.ImaginingaworldwithoutGoogleMaps
in2013isveryhard!
The S57 standard was one of the first geospatial
standards, but it focused entirely
on marine
information,and(atleast)twoothermajorinitiatives
arenow moreknown thanS57 (atleast in the non
marinedomain):
GML[13]
Geography Markup Language (GML) is the
flavour of XML defined by the Open Geospatial
Consortium (OGC) to encode geospatial
information.Itisvery
generalandcanbeusedfor
manypurposes.Forinstancecandatatypesvarya
lot betweentwo different data sets both encoded
in GML.Meaning there is no guarantee that two
systemsbothsupportingGMLcanexchangedata.
Google’sKMLformatisoftencomparedto GML,
and can loosely be
described as a GML specific
flavour.
GML haswith the latestmajor version moved to
ISOconformity.
ISO/TC211[14]
The International Standards Organization’s
Standards Committee 2011 (ICO/TC 211) is
responsible for ISO’s standards on geospatial
information and many other organizations have
beenengagedinthisworktoo.OGC
andDGIWG
(Defence Geospatial Working Group) are among
these, and IHO have also been in close contact
with ISO in this area for decades. In 2012 a
Memorandum of Understanding (MoU) to
increasetheircooperationwassignedbyISOand
IHO.
ISO/TC211 hasreleased aset ofstandards know
as
the ISO 191xx series, each covering different
scopes.Forinstanceis19136thestandardforGML
and19115,whichmaybeisthebestknownofthese
standards,isthestandardformetadata.
5 NOTTHENEWS57
To resolve the issue described above (among other
things) IHO has developed
the S100 standard!
Howeverit willnot immediatelyreplace S57 and it
wouldbeamistaketocallitanewS57.
The work on S100 was started by the IHO
Transfer Standard and Maintenance Working Group
(TSMAD), which previously was known as the
DBWG,andthis
wasthegroupthatmaintainedtheS
57 standard. The initialversion of S100 was indeed
called S57 Edition 4.0! However since this new
versionshouldresolvesomeoftheissues/limitations
withS57(bothtechnicallyandadministrative)andit
isbasedonISO/TC211withnewterms
andmodelsit
was decided to give it a new name. Hence S100
which is following IHO’s naming convention for its
standards.
With S100 product specifications are kept
completely separate, meaning that feature classes,
encoding, etc. are not a part of S100. S100 doesn’t
evendictatewhat
fileformattousefortheencoding.
In line with ISO/TC 211 is GML for instance not
mandatoryformat,butanoptiondependingonwhat
needs to be encoded. It will be the product
specifications, which are separately maintained
standards, named S101, S102, etc. that will contain
the
encodingandotherproductimplementationrules.
Productspecificationsaredevelopedunderasetof
rules defined by S100, saying for instance that a
productspecificationmustconsistof(thusdefine)the
followingparts:
productidentification
datacontentandstructure
coordinatereferencesystem
dataquality
data
capture
datamaintenance
portrayal
encoding
productdelivery
S100 introduces some new data types and
structures for modeling and disseminating the data.