Sharing Environmental Data through GEOSS.

Full text

Page 1

International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011 1
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
Keywords: Capacity Building, Data Sharing, GEOSS, Grid Computing, Interoperability, SDI, Service
Chaining
INTRODUCTION
Today we are living in a globalized world with
rapidly evolving processes including climate
change, population growth or environmental
degradation. In parallel, means of communi-
cation have expanded to take on a remarkable
place in our society, allowing us to access an
enormous and continuous flow of information.
In the last 30 years, the availability of geo-
spatial data has grown dramatically following
Sharing Environmental
Data through GEOSS
Gregory Giuliani, University of Geneva and UNEP, Switzerland
Nicolas Ray, University of Geneva and UNEP, Switzerland
Stefan Schwarzer, UNEP, Switzerland
Andrea De Bono, University of Geneva and UNEP, Switzerland
Pascal Peduzzi, UNEP, Switzerland
Hy Dao, University of Geneva and UNEP, Switzerland
Jaap Van Woerden, UNEP, Switzerland
Ron Witt, UNEP, Switzerland
Martin Beniston, University of Geneva, Switzerland
Anthony Lehmann, University of Geneva and UNEP, Switzerland
AbSTRACT
Understanding the complexity of earth-system processes is crucial to convey improved information on the
environment to decision-makers and the general public. Addressing this need by sharing environmental data
is challenging because it requires a common agreed framework that allows easy and seamless integration of
data from different sources. In this regard, the Global Earth Observation System of Systems (GEOSS) portends
major benefits through various sharing mechanisms and by giving access to services that could be linked
together to process and generate new understandable knowledge and information. Various United Nations
projects could greatly benefit from the GEOSS approach.
DOI: 10.4018/jagr.2011010101

Page 2

2 International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
the evolution of communication technologies
supported by the rapid development of spatial
data capture means such as remote sensing im-
agery, sensors and GPS (Philips, Williamson, &
Ezigbalike, 1999). One of the challenges we are
facing today is to make sense of this vast amount
of data in order to turn them into understandable
knowledge (Gore, 1998). Concrete actions can
be taken only on the basis of knowledge and
understanding, but often we know too little
about the state of our planet’s environment to
take informed and sound decisions about how
it should be managed.
Our planet is a multi-dimensional system
made of complex interactions highly intercon-
nected and continuously evolving at many
spatial and temporal scales (GEO secretariat,
2007b). This means that to understand these
interactions, we need to gather and integrate
different sets of data about physical, chemical
and biological systems. Altogether, these sets
of data constitute environmental data sets or
data related to the environment. These data are
often georeferenced, describing a geographical
location through a set of attributes and thus could
be understood as being part of geospatial data.
An environmental data set is seldom interesting
in itself, but rather displays its full information
potential when used in conjunction with other
data sets, allowing one to monitor and assess
the actual status of the global, regional or local
environments, to discover complex relation-
ships between them and to model future changes.
In 1998, the former vice-president of the
United States, Al Gore, presented his vision-
ary concept of a Digital Earth (Gore, 1998), a
representation of the Earth embedding a vast
amount of geospatial data and allowing to make
better sense of it. To achieve this vision, Gore
highlighted the need for a collaborative effort
(from government, industry, academia and
citizens) and pointed out the different tech-
nologies required: computational power, mass
storage, satellite imagery, broadband network,
interoperability and metadata.
Despite the fact that administrations and
governments are recognizing that geospatial
data are an important component of an informa-
tion infrastructure (such as e-governement) that
needs to be efficiently coordinated and managed
for the interest of all citizens (Ryttersgaard,
2001), this huge amount of geospatial data is
stored in different places, by different organi-
zations and the vast majority of these data are
not being used as effectively as they should.
In consequence, a framework allowing one to
discover, access, publish, share, maintain and
integrate geospatial data appears to be essential.
Such a framework is commonly known as a
Spatial Data Infrastructure (SDI).
Different initiatives at the regional and
global levels are influencing and promoting
the creation of SDIs allowing data providers to
share and publish their data in an interoperable
manner. These initiatives coordinate actions
that promote awareness and implementation
of complementary policies, common standards
and effective mechanisms for the development
and availability of interoperable geospatial data
and technologies to support decision making at
all scales for multiple purposes. These initia-
tives are related to data access, harmonization,
standardization, interoperability, seamless in-
tegration and services. Such an initiative is the
Global Earth Observation System of Systems
(GEOSS) which is a worldwide voluntary effort,
coordinated by the Group on Earth Observa-
tion secretariat, to connect already existing
SDIs and Earth Observation infrastructures.
GEOSS is foreseen to act as a gateway be-
tween producers of geospatial data and end
users, with the aim of enhancing the relevance
of Earth observations for the global issues
and offering public access to comprehensive
information and analyses on the environment
(GEO secretariat, 2005, 2007a). The GEOSS
Common Infrastructure (GCI) provides core
capabilities that allow users to search, access
and use data, information, tools and services, and
is made of five components: GEO portal (web
portal to access GEOSS and search registries),
GEOSS clearinghouse (connects the different
components), GEOSS components and services
registry (catalogue of services and components),
GEOSS standards and interoperability registry
(catalogue of standards to use allowing users to

Page 3

International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011 3
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
set up and configure an interoperable system),
and a best practices wiki (offers a single space
to share, discuss, propose and exchange ideas
and best practices within the community). These
components are dependent on the voluntary
contributions of members and participating
organizations. To support the nine defined So-
cietal Benefit Areas (SBAs) (disasters, health,
energy, climate, water, weather, ecosystems,
agriculture, biodiversity), the mechanisms for
data sharing and dissemination are presented
in a 10-year Implementation Plan Reference
Document (GEO secretariat, 2005) provid-
ing data sharing principles that any volunteer
member must endorse. The key element o share
data through GEOSS is to agree on “interoper-
ability arrangements” (GEO secretariat, 2007a)
allowing different components of the system to
communicate with each other.
Turning data into understandable knowl-
edge requires that data coming from different
sources be easily and seamlessly integrated.
With the capabilities offered by standards
like the one proposed by the Open Geospatial
Consortium (OGC), geospatial community can
not only discover, access and publish interop-
erable geospatial data but also services that
can be linked together, in chains of services,
to process data and generate new information.
Moreover, by registering services into GEOSS,
these different resources are now accessible in
a standardized way and are reusable for many
different purposes.
The aim of this paper is to present experi-
ences gathered through different United Nations
(UN) and European research projects and to
discuss promises and challenges envisioned
in participating to an initiative like GEOSS,
both in term of building chains of services and
sharing data.
THE NEED FOR DATA
SHARING AND INTEGRATION
Until very recently, the different systems used
to acquire environmental data were mostly
operating in isolation, which made it difficult to
easily discover, access and use the data content
of these systems due to incompatibilities and
inconsistencies of formats and data models
(Bernard & Craglia, 2005). In addition, there
is typically insufficient data exchange among
different stakeholders, which is partially due
to differing data policies. Other important
impediments to the flow of data are the delays
in accessing data that prevent timely use of
information, duplication and redundancy of data
acquisition, potential high costs associated with
data creation and access, and unclear access
rights and licensing policies (GEO secretariat,
2005). Altogether, these difficulties lead to a
fragmentation of data sources, impeding their
effective and efficient use, requiring much
more time than necessary for data collection
(Open Geospatial Consortium, 2004). All the
previous considerations highlight the growing
need to share data in an interoperable way and
to ensure that data are easily accessible and
discoverable, so that they can be used as often
and widely as possible (Arzberger et al., 2004).
Moreover, the adoption of the Agenda 21 reso-
lution, a United Nations initiative proposing a
set of actions to be taken at different scales to
promote a sustainable development, fostered
the importance of geospatial data to support
decision-making and management related to
degradation and threats affecting the environ-
ment (Nebert, 2005). Availability and access
to appropriate information, and the related
development of interoperable databases, are
the necessary conditions for creating the basis
for supporting the information management
needs of implementing and monitoring sustain-
able development policies and goals, such as
the United Nations Millennium Development
Goals (MDGs) (Henricksen, 2007). The MDGs
are eight development objectives (eradicate
extreme poverty and hunger, achieve universal
primary education, promote gender equality,
reduce child mortality, improve maternal health,
combat different diseases, ensure environmental
sustainability, and develop and global partner-
ship for development) that all UN members
have agreed to achieve by 2015.

Page 4

4 International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
Over the last twenty years, the emergence
and evolution of Geographic Information
Systems (GIS) technology and the advent of
applications such as Google Earth or Open-
StreetMap (Craglia et al., 2008), allowed for a
clear change on how geospatial data are handled
and incorporated into regular workflows of
organizations and agencies in the governmen-
tal, private and public sectors (Booz, Allen, &
Hamilton, 2005). Highlighting these changes,
Masser (2007) stated that to realize the full
potential and benefits of geospatial data, access
must be maximized with the help of Spatial
Data Infrastructures (SDIs), that allow users to
share, discover, visualize, evaluate and retrieve
geospatial data. Moreover, the vast amount of
data needed to run a complex model (e.g., in
climatology or ecology), and the recognition that
organizations and/or agencies need more data
than they can afford financially (Rajabifard &
Williamson, 2001), reinforce the concept that
once a particular set of geospatial data has been
created, it should be accessible to potential
users in both the public and private sectors
(Ryttersgaard, 2001). This reinforces the need
to store such data in databases that are made
widely accessible for various purposes (Phil-
ips et al., 1999). As a consequence, geospatial
data can be seen as a shared resource which is
maintained continuously.
To remove the barriers that block and im-
pede a wide use of geospatial data and related
information, Masser (2005, 2007) identified
different needs such as eliminating or reducing
restrictions on data access and availability (but
protecting intellectual property rights), promot-
ing interoperability between different data sets
and different systems, and disseminating the
information about data (metadata). Altogether
these objectives are designed to create an envi-
ronment that fosters activities for using, manag-
ing, producing and sharing geospatial data in
which all stakeholders can cooperate with each
other and interact with technology, to better
achieve their objectives at different political/
institutional levels (Rajabifard & Williamson,
2004). In this sense, interoperability appears
to be a key element enhancing data sharing,
communication and efficiency.
The great advantage of interoperability is
that it describes the ability of locally managed
and distributed heterogeneous systems (dif-
ferent operating systems, different databases,
different data formats) to exchange data in real
time to provide a service (OGC, 2004). The
shift towards a processed-based infrastructure
offering reusable and standardized components
responsive to user needs and requests is sup-
ported by the Service Oriented Architecture
(SOA) concept. In a SOA, services are the
elementary components representing a set
of operations that could be invoked by users
allowing them to access, in the case of the
geospatial community, distributed geospatial
data as well as geoprocessing services. To
implement and deploy geo-enabled services,
the OGC proposes a suite of standards that
use services over the Internet, so-called web
services, giving access to distributed data and
services through Uniform Resource Locators
(URLs). This allows data providers to publish
standardized services independently on how
it is implemented and on which platform it is
executed. This emphasizes the full potential
of interoperability allowing an organization to
maximize the value and reusability of data under
its control and giving the ability to exchange
these data with other interoperable systems.
Using such OGC web services offers the pos-
sibility to seamlessly couple and reuse them in
a variety of applications. By chaining together
a series of web services, users can perform a
set of operations to process data whereby new
knowledge emerges from relationships that
were not envisioned before (Open Geospatial
Consortium, 2004). Granell et al. (2009) define
service chaining as a mechanism for combining
individual geospatial web services to create
customized web applications. Although current
SDIs mostly offer the abilities to search, view
and access data, with the support of interoper-
able services and SOA related concepts it is
now possible to build new applications based
on distributed services (Friis-Christensen, et

Page 5

International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011 5
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
al., 2007; Diaz et al., 2008). When services are
organized through a coherent chain, combined
services can achieve a larger task (Di, 2004). The
International Organization for Standardization
(ISO) through its ISO 19119 standard (ISO,
2005) defines three types of chaining:
• Transparent (user-defined): the workflow
is defined and managed by users.
Translucent (workflow-managed): users
invoke a service that manages the chain.
Users are aware of atomic services that
constitute the chain.
Opaque (aggregated): users invoke an ag-
gregated service that carries out the chain.
Users have no awareness of the atomic
services that constitute the chain.
•
•
In this paper, we will focus on the trans-
parent chaining either by hard coding or by
using OGC Web Processing Service (WPS)
specification (OGC, 2007).
Through its online catalogue of registered
services, GEOSS is an interesting and promis-
ing entry-point to discover and access services
that could be integrated into service chaining
process. It offers a framework to share data,
expose them through interoperable services
and allow the production and dissemination
of timely and accurate data needed by decision
makers and the public (GEO secretariat, 2005).
SERVING DATA INTO GEOSS
In 1985, the United Nations Environment Pro-
gramme (UNEP)/Division of Early Warning
and Assessments/GRID-Europe was founded
as one of the first two centres of the Global
Resource Information Database (GRID) net-
work to support environmental decision-making
within UNEP and the UN system as a whole,
by generating and disseminating information
about the state of the world’s environment in a
timely and understandable manner. To provide
reliable environmental assessments and early
warnings, GRID-Europe specialized in han-
dling and analyzing spatial and statistical data
on environmental and natural resource issues
through computerized GIS and remotely-sensed
imagery. Over the years, GRID-Europe has
compiled an archive of global, European and
other geospatial databases as part of its informa-
tion management function. The experience and
in-house capabilities of GRID-Europe offer a
great potential to make geospatial and tabular
databases compiled over the years available to
a large array of users. Since its foundation, the
Geneva office has received considerable support
from Swiss and local authorities as well. This
supporting was significantly reinforced, and
GRID-Europe’s institutional base broadened,
with the signing of a “Partnership Agreement”
between UNEP, the Federal Office for the
Environment (FOEN) and the University of
Geneva in June 1998.
GRID-Europe closely monitors develop-
ments in information technologies and examines
their utility for environmental monitoring and
policy formulation and thus is extending and
developing its field of activities using SDIs.
Moreover, the “Partnership Agreement” pro-
vides a major opportunity to work at different
geographic scales ranging from global, to
regional (Europe) and national (Swiss) and
finally local (Geneva). Such a specificity al-
lows GRID-Europe to participate to different
applied research projects funded either by the
United Nations or the European Commission.
A common ground for these projects is to serve
and share data through the European Directive
on Infrastructure for Spatial Information in the
European Community (INSPIRE) (European
Commission, 2007), the United Nations Spa-
tial Data Infrastructure (UNSDI) (Henricksen,
2007), as well as GEOSS.
PREVIEW Global Risk
Data Platform
The PREVIEW (Project of Risk Evaluation,
Vulnerability, Information, and Early Warning)
Global Risk Data Platform (http://preview.grid.
unep.ch) is a collaborative effort of UNEP,
United Nations Development Programme
(UNDP/BCPR), United Nations International

Page 6

6 International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
Strategy for Disaster Reduction (UNISDR)
and the World Bank to share geospatial data
on global risk from natural hazards. Users can
freely visualize, download or extract data on
past hazardous events, human and economical
hazard exposure and risk from natural hazards.
The platform covers nine types of natural haz-
ard: tropical cyclones and related storm surges,
drought, earthquakes, biomass fires, floods,
landslides, tsunamis and volcanic eruptions.
The collection of data is made via a wide range
of partners. This geoportal was developed as a
support to the 2009 Global Assessment Report
on Disaster Risk Reduction (United Nations
International Strategy for Disaster Reduction
Secretariat, 2009), replacing the previous PRE-
VIEW platform initially designed by UNEP/
GRID-Europe and already available since 2000.
The new PREVIEW platform is fully compliant
with the OGC Web Services (OWS) to access
data using Web Map Service (WMS), Web
Feature Service (WFS), Web Coverage Service
(WCS), geo-enabled Really Simple Syndica-
tion (GeoRSS) or Keyhole Markup Language
(KML) as well as metadata using Catalogue
Service for the Web (CS-W).
GEO Data Portal
The GEO Data Portal (http://geodata.grid.unep.
ch) is the authoritative source for data sets
used by UNEP and its partners in the Global
Environment Outlook (GEO) report and other
integrated environment assessments. Its online
database holds more than 550 different vari-
ables, as national, sub-regional, regional and
global statistics or as geospatial data sets (maps),
covering themes such as Freshwater, Popula-
tion, Forests, Emissions, Climate, Disasters,
Health and Gross Domestic Product (GDP). The
data can be displayed and explored on-the-fly
through maps, graphs, data tables, downloaded
in various popular formats, or copied and pasted
into word processors. All information products
in the GEO Data Portal can be accessed and used
as web services as well. The retrieval of sta-
tistical and country-wide information has been
enabled via a Simple Object Access Protocol
(SOAP) connection; data from the database can
be retrieved as maps via WMS or WFS; graphs
can be displayed via a direct Uniform Resource
Locator (URL) usage.
enviroGRIDS
EnviroGRIDS (http://www.envirogrids.net) is
a European research project that will last from
2009 until 2013 and is funded under the seventh
framework programme (FP7). The Black Sea
Catchment is largely following an ecologically
unsustainable pathway based on inadequate re-
source management that could lead to severe en-
vironmental, social and economical problems,
especially in a changing climate (WWF, 2008).
The aim of the project is to build capacities in
the Black Sea region to use new international
standards to gather, store, distribute, analyze,
visualize and disseminate crucial information
on past, present and future states of this region,
in order to assess its sustainability and vulner-
ability. EnviroGRIDS objective is to federate
and strengthen existing Observation Systems
to address several GEOSS Societal Benefit
Areas within a changing climate framework.
The expected result will be a shared informa-
tion system that operates on the boundary of
scientific/technical partners, stakeholders and
the public. It will contain early warning systems
able to inform in advance decision-makers and
the public about risks to human health, biodi-
versity and ecosystems integrity, agriculture
production or energy supply caused by climatic,
demographic and land cover changes on a 50-
year time horizon. To achieve and support the
enviroGRIDS vision and objectives, a grid-
enabled Spatial Data Infrastructure (gSDI) is
under construction. The aim of the gSDI is to
host and analyze the data for the assessment of
GEOSS Societal Benefit Areas, as well as the
data produced within the project. These data
must be gathered and stored in an organized
form and accessible in an interoperable way
on the grid infrastructure in order to provide a
high performance and reliable access through
standardized interfaces.

Page 7

International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011 7
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
ACQWA
ACQWA (http://www.acqwa.ch) stands for
Assessing Climate impacts on the Quantity
and quality of Water. It is also a FP7 European
research project lasting from 2008 until 2013. As
the evidence for human induced climate change
becomes clearer, so does the realization that its
effects will have impacts on natural environment
and socio-economic systems. Some regions are
more vulnerable than others, both to physical
changes and to the consequences for ways of
life. According to the description of work, the
project will assess the impacts of a changing
climate on the quantity and quality of water
in mountain regions which are particularly af-
fected by rapidly rising temperatures, prolonged
droughts and extreme precipitation. Modeling
techniques will be used to project the influence
of climatic change on the major determinants
of river discharge at various time and space
scales. Regional climate models will provide
the essential information on shifting precipita-
tion and temperature patterns. Snow, ice, and
biosphere models will feed into hydrological
models in order to assess the changes in season-
ality, amount, and incidence of extreme events
in various catchment areas. Environmental
and socio-economic responses to changes in
hydrological regimes will be analyzed in terms
of hazards, aquatic ecosystems, hydropower,
tourism, agriculture, and the health implications
of changing water quality. Attention will also
be devoted to the interactions between land
use/cover changes, and changing or conflict-
ing water resource demands. Adaptation and
policy options will be elaborated on the basis
of the results. The chain of processes involved
in climatic, cryospheric and hydrologic models
is complex because each process impacts on
different compartments of human and natural
systems. Different types of data covering various
geographical regions are therefore necessary to
build different sets of scenarios, which translates
into substantial amount of data.
TECHNICAL COMPARISON
AND COMMON GROUNDS
All these projects have in common that they
already share (or will share in a near future)
their data and metadata into the GCI. As a pre-
requisite all the registered services have to be
interoperable using mainly standards proposed
by the OGC, but also other protocols like the
Simple Access Object Protocol (SOAP). A
short comparison of these different projects is
Table 1. Technical comparison of enviroGRIDS, ACQWA, GEO Data Portal and PREVIEW projects
Project name enviroGRIDSACQWA GEO Data PortalPREVIEW
Services
WMS, WFS, WCS,
CS-W, KML,
GeoRSS, WPS, grid
services
WMS, WFS, WCS,
CS-W, WPS, KML,
GeoRSS
WMS, WFS, WCS,
CS-W, SOAP
WMS, WFS, WCS,
CS-W, KML,
GeoRSS
Software
GeoServer, ArcGIS
Server, PyWPS,
GeoNetwork, gLite
GeoServer, GeoNet-
work, PyWPS
GeoServer, GeoNet-
work, MapServer
GeoServer, GeoNet-
work, MapServer
Type of models
- Hydrological
models
- Snow cover map-
ping
- providing base lay-
ers (socio-economic,
...)
- providing base lay-
ers (events, risk, ...)
Challenges &
difficulties
- linking SDI and grid
infrastructure
- capacity building
- authorization/
authentication
- portal integration
- capacity building
- data integration
- data integration
- data/metadata
harmonization
- different standard
implementation
- capacity building
- data integration
- data/metadata
harmonization
- capacity building

Page 8

8 International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
given in Table 1, indicating which services are
available, which software are used to publish
these services, what are the types of models
used to chain these layers, and finally what
are the challenges and difficulties raised while
integrating these services.
Most of these projects make use of free
and open source software (PostgreSQL/Post-
GIS, MapServer, GeoServer, GeoNetwork and
PyWPS) because it can ease the portability and
replicability of tools developed. Indeed, many
countries with low to moderate incomes are
often also affected by natural hazards, environ-
mental threats or degradation, and these coun-
tries are especially interested to manage and
share their geospatial data using free and open
sources software. Having tools readily available
to be deployed in these countries is a strong
incentive for capacity building, knowledge
transfer, and sharing of expertise.
These projects are also strongly related to
capacity building in order to enhance an “open
and sharing spirit”. It is necessary to show
and prove the benefits of data sharing through
appropriate examples, to communicate best
practices as much as possible and to develop
guidelines and policies. Altogether this will
help to reach agreement and endorsement on
the use of new standards. Such a participative
approach will certainly stimulate data provid-
ers to be more “open” and in consequence to
share their data. The different projects presented
before will organize different workshops and
develop various teaching material allowing
participants, ranging from students to members
of government, to learn how to use the specific
applications to share large amount of data.
Rajabifard and Williamson (2004) believe that
building capacities is an important challenge for
SDIs concepts to be accepted and adopted at
a large extent. For these authors, the best way
to reach this objective is to establish a long-
term commitment to education and research:
otherwise the SDI vision will remain unclear
and unachievable. Through these projects, the
objective is to build the capacity of scientists
to share and document their data in order to
strengthen existing observation systems, the
capacity of decision-makers to use it, and the
capacity of the general public to understand the
important environmental, social and economic
issues at stake.
Through simple data integrating scenarios
(integration into other web portals or applica-
tions) GEO Data Portal and PREVIEW have
pointed out different issues. Integrating some
socio-economic data sets coming from the GEO
Data Portal with natural hazards maps of the
PREVIEW project to compute, for example,
economical exposition of a country to a specific
hazard, was impossible. This problem comes
from the different implementations of OGC
specifications between Mapserver (used by
the GEO Data Portal) and Geoserver (used by
PREVIEW). Indeed, it appears that Mapserver
use an argument “MAP” that is not standardized
and not recognized by all clients. This problem
will be solved by migrating to Geoserver all
the data services of the GEO Data Portal. Thus
implementation of a same specification can
differ from one software to another and can
impend a consistent integration of services.
Another issue raised by data integration
process was raised by the United Nations High
Commissioner for Refugees (UNHCR) while
trying to integrate WMS data coming from
the PREVIEW project in order to identify
area that are not suitable to install refugees
camps. It appears that the only projection
available was EPSG:4326 (Geographic) whilst
UNHCR geoportal makes use of Google maps
in EPSG:900913 (Spherical Mercator). This
experience showed us that it is important, while
publishing data services, to support at least
the most frequent projection types. Geoserver
supports natively all types of projections and
it is easy to reproject on-the-fly data stored in
another projection so that it can be integrated
with data with other projections. In addition,
following the size of the data set, an important
processing overload has been observed caused
by the on-the-fly reprojection process. This can
slow the service chain and impend and efficient
data integration.
In the ACQWA project, a specific constraint
is the important number of partners involved and

Page 9

International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011 9
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
their different scientific backgrounds (climatol-
ogists, hydrologists, glaciologists, ecologists). It
is quite challenging to raise awareness on new
tools and way of gathering and exchanging data
without strongly influencing the way these dif-
ferent communities are working with geospatial
data. For that reason, the aim is to concentrate
on the promotion of GEOSS as an interesting
and useful framework to handle and discover
scientific data. Obviously, a dedicated geoportal
is under development to register the main out-
puts of the ACQWA project into GEOSS using
OGC web services. Nevertheless to show the
benefits of working with interoperable services,
we are currently developing a scenario to make
estimation of snow cover from remote sensing
imagery using data coming from the Moderate
Resolution Imaging Spectrometer (MODIS) and
Shuttle Radar Topography Mission (SRTM).
Project partners that are currently working to
produce such estimations are working with PCI
Geomatica, doing all the process chain manu-
ally. Our objective is to help our partners by
publishing a WPS geoprocessing service that
allows them to automatize this analysis (Figure
1). Once retrieved by FTP, MODIS images are
saved on a server that store also SRTM tiles.
All data are in EPSG:4326 and will be available
using WCS standard published by Geoserver.
Finally, the WPS service, currently under de-
velopment using PyWPS, will implement the
different steps to process the data. A major dif-
ficulty encountered until now is to “translate” the
PCI functionalities by finding the equivalent in
Geographic Resources Analysis Support System
(GRASS) software. Indeed, PyWPS does not
process data by itself and instead uses GRASS
as a backend to access all the geoprocessing
functionalities.
Once the snow cover process is success-
fully achieved, our hope is to convince other
communities within the project to benefit from
such an approach and to develop other sce-
narios especially making use of climate data.
In the process of turning data into un-
derstandable information and knowledge by
chaining data services a new challenge has
emerged. The ever-increasing spatial and tem-
poral resolution of geospatial data are causing
a tremendous increase in term of data volumes
and the limits of the processing capacities of
traditional GIS and SDI are being reached. With
the advent of grid computing and the progres-
sive deployment of large grid infrastructure
projects (e.g., Enabling Grids for E-sciencE)
many scientific disciplines now have access to
sizable computing resources and new opportuni-
ties are emerging. For Foster et al. (2008) grid
aims to federate resource sharing in a dynamic
and distributed environment across a network
allowing to access unused CPUs and storage
space to all participating computers. Currently,
SDIs are lacking processing power and should
therefore be made interoperable with grid infra-
Figure 1. Data sources and processing steps for a geoprocessing service estimating snow cover

Page 10

10 International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
structures, which are offering large storage and
computing capacities. Recent studies (Muresan
et al., 2008; Di, Chen, Yang, & Zhao, 2003)
applied a successful approach to extend grid
computing to the remote sensing community
and to make OGC web services grid-enabled.
Both studies considered that the grid has a great
potential for the geospatial disciplines. Padeberg
and Greve (2009) have identified several differ-
ences between OGC-compliant SDIs and grid
infrastructures concerning service description,
service interface, service state and security. In
particular, grid infrastructures are based on
SOAP messaging protocol to invoke opera-
tions and Web Service Description Language
(WSDL) to describe services. OGC-compliant
does not support neither SOAP nor WSDL,
except WPS, and thus chaining geospatial ser-
vices with grid services could be problematic.
In addition, OGC standards do not provide any
security mechanisms (authentication, encrypted
communication between resources) which is a
major concern in grid infrastructures. Finally,
Di et al. (2003) showed that the current grid
metadata catalog system is not good enough to
answer the needs of the geospatial community,
especially the requirements of the ISO19115
standard. All these differences must be over-
come in order to allow traditional SDIs to
benefit from the power of grid computing, and
consequently to offer new services to GEOSS.
The main scientific and technological
challenge of the enviroGRIDS project will
be to link an SDI with a grid infrastructure to
benefit from the processing capacities offered
by grids. Indeed, WPS appears to be an adequate
candidate to be grid-enabled because, first, it
supports SOAP protocol and, second, geospatial
community has a growing processing need that
current SDIs cannot deliver. A grid-enabled
SDI will allow users to model high resolution
hydrological models (e.g., Soil and Water As-
sessment Tool) of the Black Sea catchment under
various climate, land cover and demographic
scenarios. In order to develop such a gSDI to
support the development of Black Sea portal
functionalities, the different components of the
enviroGRIDS architecture are currently being
defined to highlight the main issues emerging
from different conceptual and technological
solutions (Figure 2).
These issues concern the choices of soft-
ware components, data repositories, data man-
agement, grid-oriented processing, grid portal,
and interoperability between SDIs and grid
infrastructures. Although the use of grid-enabled
web services to access data sets stored in the
SDI will also be explored (Maué & Kiehle,
2009), bridging architectural gaps between grids
and SDIs remains very challenging (Padberg
& Greve, 2009) without extensions and cus-
tomizations. For example, an important question
concerns the location of geospatial data re-
positories: inside or outside the grid? The answer
is not trivial and will greatly influence services
and in particular chains of services to process
data. In the one hand, being outside the grid,
all OGC-compliant services functionalities
remain the same and grid services are only used
to process the data. On the other hand, being
inside the grid, all OGC web services have to
be modified to support grid environment, be-
coming grid-enabled. The latter would allow
benefiting from all the advantages of the grid
(security, replicability, scalability, storage and
processing capacities) but would obviously
require a lot of developments for adapting al-
ready existing SDIs. In consequence, an incre-
mental development and implementation
strategy will be developed taking into account
different integration scenarios aiming to hide
the complexity of the grid while preserving
OGC interfaces.
CHALLENGES AND PROMISES
From the experience acquired, or being ac-
quired, through these different projects, it is
obvious that many challenges remain both
tangible (e.g., technology) or less tangible
(e.g., culture, behavior). Nevertheless, it is
critical to overcome them in order to improve
our knowledge, share our experience and at-
tempt to strive towards a society that is better
informed. Achieving the goal of sustainable

Page 11

International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011 11
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
development requires the integration of a large
number of different data from various sources.
Through agreed common standards and a clear
political will, these data can be integrated in an
interoperable way, leading to a new collabora-
tive approach to decision-making.
Having environmental data in digital form
allows easy storage and dissemination, facilitate
data exchange and sharing, faster and easier
update and corrections, ability to integrate
data from multiple source (see Figure 1), and
customization of products and services (Hen-
ricksen, 2007). In this sense SDIs appear to be a
good choice to encompass the sources, systems,
network linkage standards and institutional
issues involved in delivering geospatial data
from many data sources to the widest possible
group of potential users (Coleman, McLaugh-
lin, & Nichols, 1997). The fact that, during the
last years, multiple SDIs initiatives have been
developed all around the world, ranging from
local to regional levels, is a good sign. It ap-
pears that there is a growing recognition that
geospatial data is a critical element underpinning
decision making in many disciplines (Rajabifard
& Williamson, 2001) and as such needs to be
effectively managed.
The SDI hierarchy model proposed by Ra-
jabifard (2002) is composed of inter-connected
SDIs developed at different levels (from local
to global). Each SDI of a higher level is formed
by the integration of data developed and made
available by the lower level. Such a hierarchy
can be approached though two views: on one
hand, it is an umbrella in which the SDI at a
higher level encompasses all SDI components
from lower levels. On the other hand, it can
be seen as the building blocks supporting the
access of data needed by SDIs at higher levels.
This hierarchy allows creating an environment
in which users working at any level can rely on
data from other levels and integrate data from
Figure 2. EnviroGRIDS grid-enabled SDI components supporting Black Sea portal

Page 12

12 International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
different sources (Mohammadi & Rajabifard,
2009). Such a hierarchy is clearly envisioned in
the concept of the system of systems on which
GEOSS relies, integrating systems together
into an information highway which both links
together environmental, socio-economic and
institutional databases and provides a movement
of information from local to global levels. For
Masser (2006), the SDI hierarchy poses the
challenge of multi-stakeholder participation
in SDI implementation, because the bottom-up
approach differs a lot from the top-down ap-
proach. The top-down vision, common in the
SDI literature, emphasizes the need for stan-
dardization and uniformity while the bottom-up
view stresses the importance of diversity and
heterogeneity caused by the different needs of
the various stakeholders. As a consequence, it
is necessary to find a consensus ensuring suf-
ficient standardization and uniformity while
recognizing the diversity and heterogeneity of
the different stakeholders acting at different
levels. In particular, building a system of sys-
tems like GEOSS is highly dependent on a clear
governance structure that is understandable
and acceptable by the volunteer participants in
order to develop a shared vision of the system
and to allow users to feel a common sense of
ownership (Masser, 2007). As it is reminded
in the Strategic Guidance document (GEO
secretariat, 2007a), the success of GEOSS will
depend on interoperability arrangements that
data providers agree to endorse.
As a provider of environmental data, GRID-
Europe is continuously facing the challenge
of encouraging data providers to go “open”
and to share their data in an interoperable and
OGC-compliant way. At present, technology is
no longer a problem because solutions based
on a variety of software can be proposed and/
or developed depending on the requirements
and the technical capabilities available. The
most difficult task is to create an environment
allowing wide agreement on data sharing prin-
ciples. In this particular regard, the GEOSS
“best practices wiki” could be of great benefit
to help people promote sharing principles. A
lesson learned from our experience is that once
users can discover data they need, their most
important preoccupation is to know what is
the quality of the data they are going to access
and whether they can trust this data. We are
convinced that sharing data is an efficient way
to eventually recognize whether this data is of
sufficient quality. By submitting/exposing the
data to the judgment of the broader community,
one can know if it is useful or not. Through data
sharing, one can also benefit from the interac-
tion with end users by receiving feedbacks and
then improve the data sets accordingly. Sharing
data and participating to GEOSS can therefore
contribute to the improvement of data, which in
turn allows better information and eventually
better decisions.
In the current climate of economic con-
straints, interoperability and standardization
have never been so important because a non-
interoperable system impedes the sharing of
data, information and resources, which increase
the risk for a system to fail in delivering its
expected benefits and to remain unused (Open
Geospatial Consortium, 2004). Geospatial
data can be an expensive and time consuming
resource to produce, and for this reason, it is
of high importance to improve accessibility
and availability and promote its reuse. Many
decisions that organizations need to make
depend on good quality and consistent data,
readily available and accessible (Rajabifard
& Williamson, 2001). The process of reuse
does not only concern the data itself, but also
encompasses the capabilities, skills developed,
invested effort and capital. This process allows
an organization to share the costs of data, people,
and technology, which helps realize more rapid
returns on investment. By reusing data, one can
avoid duplication of efforts and expenses and
enable users to save resources, time and effort
when trying to acquire or maintain data sets
(Rajabifard & Williamson, 2001).
Percivall (2006) claimed that in a distrib-
uted environment, the help of open standards
such as OGC can help scientists to rapidly
find and evaluate a lot of different data sets
and processing approaches, providing a flex-
ible and cooperative environment that foster

Page 13

International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011 13
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
collaboration in the different scientific com-
munities that work with geospatial data. Thus,
organizing the workflows using standard-based
web services could provide a great benefit in
term of productivity to address the nine SBAs
of GEOSS. OGC standards provide a solid
ground for interoperability between services
within distributed geoprocessing environment
offered by SDIs (Friis-Christensen et al., 2007).
In particular, the fact that these services can be
reused and chained within other applications is
a very useful aspect offering the opportunity to
solve specific problems in a more flexible way
than with stand-alone applications. Neverthe-
less, some performance issues can appear with
services that need to access and move large
amount of data. This can negatively impact
the execution time of this service (e.g., huge
overload in gathering necessary data) especially
if this service is chained with other services.
Consequently, GEOSS represents a very
promising and potentially powerful framework
to share and expose data. In particular, the fact
that a good governance structure is already in
place allows a clear vision that can be easily
shared and endorsed by the participants. The fact
that participating to GEOSS is on a voluntary
basis could be seen either as a great opportunity
or as a risk. Indeed, the voluntary aspect poses
the threat that only a few data providers join
such an initiative and, as a consequence, the
system could miss its objectives. Nevertheless,
the growing number of components and services
registered through GEOSS is a good sign for
optimism. In particular, we think that interna-
tional organizations such as UNEP could play a
major role by paving the way toward a broader
acceptance by similar organizations. The fact
that GEOSS is based on distributed systems that
can operate, evolve and be managed in a relative
independence appears to be a good choice to find
a consensus ensuring sufficient standardization
and uniformity, while recognizing the diversity
and heterogeneity of the different stakeholders.
Finally, GEOSS offers a unique characteristic
that justifies by itself its existence, which is
the possibility to see emergent properties. For
Béjar et al. (2009), this emergence is the main
objective of a system of systems, where users
perform functions that cannot be made with
any single component. This means that such
a system is more than the sum of its parts and
offers the possibility to better understand the
complex relationships between the different
components of the Earth system.
CONCLUSIONS
Geospatial data is a critical element underpin-
ning decision-making for many disciplines and
is indispensable to make sound decisions at all
levels, from global to local. Experiences from
developed countries show that more than two-
thirds of human decision-making are affected by
spatially-referenced data (Ryttersgaard, 2001).
Even if the technology exists, organizations and
agencies around the world are still spending
billions of dollars every year to produce, man-
age and use geospatial data, but they still do
not have the information they need to answer
the challenges our world is facing (Rajabifard
& Williamson, 2001).
The web service model proposed by the
OGC appears to be suitable to allow users to
combine different services to solve a specific
problem in a scalable and flexible way. Nev-
ertheless, through simple examples of services
chaining, we have highlighted different issues
that could potentially impede an easy integra-
tion: problems with different implementation
of a same specification, problems regarding
different projections used in different web
applications, overload caused by on-the-fly
reprojection using large data sets. Moreover,
working with different communities that are not
necessarily aware of the possibilities offered by
OGC web services could limit the diffusion of
such approach outside the geospatial commu-
nity. These communities need to be convinced,
through simple examples, which working with
chained services can bring benefits in their own
working flows. Finally, grid computing appears
to be a promising complement of traditional
SDIs capabilities to build WPS services for
processing large data sets. To achieve this

Page 14

14 International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
objective, implementation of SOAP protocol
into OGC specifications is a pre-requisite in
order to allow the two types of infrastructures
to communicate (interoperability) and to ease
the combination of OGC and grid services in
efficient chains.
Ten years after, GEOSS could be seen as an
initial step to achieve Gore’s vision, because the
relevant technologies are available and there is
growing recognition that countries can benefit
both economically and environmentally from
better access to data. GEOSS has the potential
to support the achievement of sustainable de-
velopment initiatives such as the UN Millen-
nium Development Goals and to offer a unique
framework to share data and collaborate for a
better society. In this sense, organizations such
as UNEP can act as a “catalyst”, contributing to
GEOSS, building capacities and ensuring that
environmental data are easily accessible. This
is a necessary step to ensure better-informed
decision-making for the more sustainable de-
velopment of our planet.
ACKNOWLEDGMENTS
The authors would like to acknowledge the
European Commission “seventh framework
programme” that funded the enviroGRIDS
(Grant Agreement n° 226740) and ACQWA
(Grant Agreement n°212250) projects, and
UNEP for its support. The views expressed in
the paper are those of the authors and do not
necessarily reflect the views of the institutions
they belong to.
REFERENCES
Arzberger, P., Schroeder, P., Beaulieu, A., Bowker,
G., Casey, K., & Laaksonen, L. (2004). Promot-
ing Access to Public Research Data for Scientific,
Economic and Social Development. Data Science
Journal, 3, 17. doi:10.2481/dsj.3.135
Béjar, R., Latre, M. A., Nogueras-Iso, J., Muro-
Medrano, P. R., & Zarazaga-Soria, F. J. (2009).
Systems of Systems as a Conceptual Framework for
Spatial Data Infrastructures. International Journal of
Spatial Data Infrastructures Research, 4, 17.
Bernard, L., & Craglia, M. (2005). SDI - From Spatial
Data Infrastructure to Service Driven Infrastructure.
Paper presented at the 1st Research Workshop on
Cross-learning on Spatial Data Infrastructures (SDI)
and Information Infrastructures (II), Enschende, The
Netherlands.
Booz, A. (2005). Geospatial Interoperability Return
on Investment. Hamilton: NASA.
Coleman, D. J., McLaughlin, J. D., & Nichols, S.
(1997). Building a Spatial Data Infrastructure. Paper
presented at the 64th Permanent Congress Meeting
of the Fédération Internationale des Géomètres
(FIG), Singapore.
Craglia, M., Goodchild, M. F., Annoni, A., Camara,
G., Gould, M., & Kuhn, W. (2008). Next-Generation
Digital Earth: A position paper from the Vespucci
Initiative for the Advancement of Geographic In-
formation Science. International Journal of Spatial
Data Infrastructures Research, 3, 22.
Di, L. (2004). Distributed Geospatial Information
Services – Architectures, Standards, and Research
Issues. In Proceedings of the International Archives
of Photogrammetry, Remote Sensing, and Spatial
Information Sciences, Istanbul, Turkey (Vol. 35,
Part 2, Commission II).
Di, L., Chen, A., Yang, W., & Zhao, P. (2003). The
Integration of Grid Technology with OGC Web Ser-
vices (OWS) in NWGISS for NASA EOS Data. Paper
presented at the GGF8 & HPFC12, Seattle, WA.
Diaz, L., Granell, C., & Gould, M. (2008). Case
Study: Geospatial Processing Services for Web-based
Hydrological Applications. Geospatial Services and
Applications for the Internet, 31-47.
European Commission. (2007). Proceedings of the
Directive 2007/2/EC of the European Parliament
and the Council of 14 March 2007 establishing an
Infrastructure for Spatial Information in the Euro-
pean Community (INSPIRE), Brussels, Belgium.
Foster, I., Yong, Z., Raicu, I., & Lu, S. (2008). Cloud
computing and grid computing 360-degree compared.
Paper presented at the 2008 Grid Computing Envi-
ronment Workshop, Austin, TX.

Page 15

International Journal of Applied Geospatial Research, 2(1), 1-17, January-March 2011 15
Copyright © 2011, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global
is prohibited.
Friis-Christensen, A., Lutz, M., Ostländer, N., & Ber-
nard, L. (2007). Designing Service Architecture for
Distributed Geoprocessing: Challenges and Future
Directions. Transactions in GIS, 11(6), 799–818.
doi:10.1111/j.1467-9671.2007.01075.x
GEO secretariat. (2005). Global Earth Observation
System of Systems 10-Year Implementation Plan
Reference Document. Geneva, Switzerland: GEO.
GEO secretariat. (2007a). Strategic Guidance for
Current and Potential Contributors to GEOSS.
Geneva, Switzerland: GEO.
GEO secretariat. (2007b). The Full Picture. Geneva,
Switzerland: GEO.
Gore, A. (1998). The Digital Earth: Understanding
our planet in the 21st Century (p. 4).
Gorgan, D., Bacu, V., Ray, N., & Maier, A. (2009).
EnviroGRIDS data storage guideline (p. 67).
Granell, C., Gould, M., & Esbri, M. A. (2009).
Geospatial Web Service Chaining. In Karimi, H. A.
(Ed.), Handbook of Research on Geoinformatics (pp.
189–195). Hershey, PA: IGI Global.
Henricksen, B. (2007). UNSDI Compendium: A UN-
SDI Vision, Implementation Strategy, and Reference
Architecture. UNGIWG.
ISO. (2005). ISO 19119: Geographic Information
– Services. Gland, Switzerland: ISO.
Masser, I. (2005). The Future of Spatial Data Infra-
structures. Paper presented at the ISPRS Workshop
on Service and Application of Spatial Data Infra-
structure, Hangzhou, China.
Masser, I. (2006). Multi-level Implementation of
SDIs. GIM International, 20, 4.
Masser, I. (2007). Building European Spatial Data
Infrastructure. Redlands, CA: ESRI Press.
Maué, P., & Kiehle, C. (2009). Grid technologies
for geospatial applications: an overview. GIS Sci-
ence, 3, 65–67.
Mohammadi, H., & Rajabifard, A. (2009). Multi-
source Spatial Data Integration within the Context
of SDI Initiatives. International Journal of Spatial
Data Infrastructures Research, 4, 18.
Muresan, O., Pop, F., Gorgan, D., & Cristea, V.
(2008). Satellite Image Processing Applications in
MedioGRID. Paper presented at the 5th International
Symposium on Parallel and Distributed Computing.
Nebert, D. D. (2005). Developing Spatial Data
Infrastructure: The SDI Cookbook.
(2004). Open Geospatial Consortium (p. 7). The
Havoc of Non-Interoperability.
Open Geospatial Consortium. (2007). OpenGIS Web
Processing Service (p. 87).
Padberg, A., & Greve, K. (2009). Gridification of
OGC web services: challenges and potential. GIS
Science, 3, 77–81.
Percivall, G. (2006). GEOSS to Benefit from “Ser-
vice Chaining” Based on OGC® Standards (p. 5).
Geoinformatics.
Philips, A., Williamson, I., & Ezigbalike, C. (1999).
Spatial Data Infrastructure concepts. The Australian
Surveyor, 44(1), 8.
Rajabifard, A. (2002). Proceedings of Diffusion of
Regional Spatial Data Infrastructures: with Particu-
lar Reference to Asia and the Pacific, Melbourne,
Australia.
Rajabifard, A., & Williamson, I. P. (2001). Spatial
Data Infrastructures: Concept, SDI Hierarchy and
Future Directions. Paper presented at the Geomat-
ics’80, Tehran, Iran.
Rajabifard, A., & Williamson, I. P. (2004). SDI De-
velopment and Capacity Building. Paper presented
at the GSDI-7, Bangalore, India.
Ryttersgaard, J. (2001). Spatial Data Infrastructure:
Developing Trends and Challenges. Paper presented
at the International Conference on Spatial Information
for Sustainable Development, Nairoibi.
United Nations International Strategy for Disaster
Reduction Secretariat (UNISDR). (2009). Global
Assessment Report on Disaster Risk Reduction 2009.
Geneva, Switzerland: UN.
World Wide Fund. (2008). Greening the Black Sea
Synergy. Brussels, Belgium: WWF.