Conference PaperPDF Available

The GEOSUD remote sensing data and services infrastructure

Authors:

Abstract and Figures

Highlights: This paper describes the GEOSUD project which aims to implement a national data and services infrastructure in order to facilitate the use of satellite imagery by the French scientific community and public institutions. This ecosystem of innovation is part of the THEIA Land Data Centre.
Content may be subject to copyright.
The GEOSUD remote sensing data and services infrastructure
Pierre Maurel (1), Jean-François Faure (2), Jean-Philippe Cantou (3), Jean-Christophe Desconnets (2),
Maguelonne Teisseire (1), Isabelle Mougenot (4), Cécile Martignac (5), Eric Bappel (1)
(1) Irstea, Maison de la Télédétection, Montpellier
(2) IRD, Maison de la Télédétection, Montpellier
(3) IGN Espace, Toulouse
(4) Université de Montpellier, Montpellier
(5) CIRAD, Maison de la Télédétection, Montpellier
Highlights: This paper describes the GEOSUD project which aims to implement a national data and
services infrastructure in order to facilitate the use of satellite imagery by the French scientific
community and public institutions. This ecosystem of innovation is part of the THEIA Land Data
Centre.
Key words: Spatial Data Infrastructure, Satellite Images, GEOSUD project, THEIA
Background: the satellite images, a high potential clearly underexploited by the public sector
The knowledge and the understanding of territories and their dynamics, the dialogue between the actors of
these territories, the definition, the implementation and the evaluation of public policies and strategies of
management, imply to mobilize and to share relevant, reliable and regularly updated spatial information. Beside
the reference data produced by IGN, the French national mapping agency, and the ones generated by institutional
actors (State decentralized services, public institutions, local authorities), associations and private operators,
satellite images allow observing in a homogeneous, repetitive and long-lasting way wide territories.
If the potential of this source of data was abundantly illustrated by numerous case studies, we have to admit
that it remains little used by the public sector in charge of the management of the environment, the agriculture
and the territories. Even if some domains have integrated this source of data (meteorology, oceanography, major
hazards, land cover at the European scale, Common Agricultural Policy), it remains absent in most of the French
operational policies and is still considered as a domain of specialists.
The reasons are known and are primarily due to the economic model of data access, the complexity of image
processing, the skills required for their valuation, and the lack of dissemination of methods tailored to
operational management needs. Over the last 20 years, satellite imagery has been largely seen as a sophisticated
tool that was left for specialists and that the majority of public actors could not appropriate. One consequence of
this weak demand is that the market was limited to companies offering services in this area.
Objectives of the GEOSUD project
In response to the context exposed above, the GEOSUD project is the result of a long-term strategic vision
which aims to develop the use of satellite imagery within the scientific community and public actors involved in
environmental management and territorial development. The strategy consists in removing, or at least mitigating,
the various obstacles to the operational use of satellite imagery. This project is situated at the interface between
“Information and Communication Sciences and Techniques” and “Environment and Society”
Initiated by the founding members of the Remote Sensing Centre in Montpellier (AgroParisTech, CIRAD,
IRD and Irstea), it relies on two major funding:
- The first one, registered in the State-Languedoc-Roussillon Region Project Program (CPER) for the 2007-
2013 period, includes the extension of the building of the Remote Sensing Centre and the installation of a
satellite receiving antenna.
- The second one, the Equipex GEOSUD, funded under the “Investments for the Future” Programme (2011-
2019), focuses on satellite imagery acquisition, the development of accessible products and services from a
Web Remote Sensing Data Infrastructure, training and networking of actors. It brings together 13 partners
from research, higher education, public sector, as well as the professional world of geomatics: AFIGEO,
AgroParisTech, CEREMA, CINES, CIRAD, CNES, Geomatys IGN, IRD, Irstea, the Universities of
Montpellier, Antilles-Guyane and Reunion. The French National Space Agency (CNES), the Ministries of
Agriculture and Sustainable Development are closely involved in steering the project.
In 2012, the holders of the GEOSUD project have strongly cooperated with the CNES to create the THEIA
Land Data Centre, which is described in depth in a companion paper [1].
Technical content of GEOSUD SDI
Main components
The GEOSUD project includes four main components:
Satellite data acquisition with single multi-user licence agreement: 5 years of annual high-resolution satellite
coverage of the French national territory (RapidEye, SPOT 4 and 5 images); very high-resolution images
(Pleiades satellites) covering main conurbations and areas with specific issues.
Data and computation infrastructure: data administration, broadcasting, computation capacities.
Research and methods: upstream research on image processing methods, applied research on the structure
and dynamics of ecosystems and territories
Communities and capacities building: networking between the scientific community and public policies
stakeholders; animation and training
Procedure for joining the GEOSUD community and to access its products and services
GEOSUD products and services are available to most of the French public institutions (scientific community,
State Departments, government agencies, regional and local authorities, private sub-contractors, foreign partners)
and small private companies for research and development.
Access to GEOSUD community and its products and services is structured in a simple two-step procedure:
- Opening a GEOSUD account (step 1) which consists of the signing by the director of the public institution
of two documents (a membership charter and an act of commitment to the respect of licenses) available on
line http://geosud.teledetection.fr and the designation of a contact person within the corresponding
organization;
- Accessing images (step 2) that are available on the site: the authorized user chooses the type of product and
the geographical area of interest. Once received the signed request, the GEOSUD team makes the images
available to the user through an FTP site.
IT architecture of GEOSUD SDI
The design of GEOSUD SDI was led by the functional requirements described above and those relating to
the interconnection between different image providers of the project including catalogs and images (i.e. the
Remote Sensing Centre in Montpellier and IGN Geoportal in Paris). It has also to ensure interoperability with
THEIA Land Data Centre SDI whose GEOSUD project is a major component (Figure 1), to contribute to
INSPIRE Directive for data sharing at European level and to the international community of Earth Observation
through the GEOSS initiative.
Figure 1 GEOSUD SDI within Theia’s Service & Data Infrastructure
The software component of GEOSUD SDI V1.0 addresses the needs for identifying and retrieving the
images delivered by GEOSUD. Functionalities to display images at full resolution as well as for metadata
consultation have been developed to accurately assess the suitability of the images for the intended use. Finally,
a download function has also been added to access the selected image.
Administration functionalities are also available on the current version. They allow the harmonization and
annotation of metadata, the pyramiding of images for fast display at different zoom. They also enable the
management of users and their membership to GEOSUD as well as their rights regarding the use of images.
The implementation of the interoperability relies on a Web services-based infrastructure. Thus, GEOSUD
SDI allows accessing images through interoperable services for image discovery, display and downloading.
Image discovery services are based on the OGC CS-W standard (Catalog Service for the Web) and take into
account the recommendations of the INSPIRE Directive. Moreover, this service is based on a semantic
referential that enriches the description of the images in order to facilitate their selection by non-expert users [2].
The viewing and download services are implemented from OGC standard services WMS (Web Mapping
Service) and WMTS (Web Mapping and Tiling Service). Finally, a single identification and authentication
mechanism (SSO) is shared between GEOSUD image providers and those of THEIA Land Data Centre. It is
built on the OAuth2 protocol that is widely used on the web. Using a single user account to access collections of
images spread among different image providers should greatly facilitate access to these data.
The provision of high and very high resolution image collections represents a significant volume, about a
hundred Terabytes. The collections of high resolution images (SPOT5, Rapid Eye) and very high-resolution
images (SPOT 6-7) represent respectively a volume of about 2 Tbytes and 12 Tbytes per annual coverage.
The next version of the SDI that will be available mid-2016 will enlarge its functional range to on-line image
processing capabilities. For that, several new functionalities will be developed in order to discover these
processing services, to assess their compatibility with the selected images, to make available enough computing
resources according to the complexity of the algorithms and the volume of data.
Satellite images portfolio
The GEOSUD web portal was opened in May 2011 with some regional image collections with a spatial
resolution of 5 to 10 m (Languedoc-Roussillon 2009, Saone Basin 2009, Massif des Landes Winter 2010). A
first annual coverage of the French territory (summer 2010) with Rapid Eye images (5 m) was acquired
following a competitive tender pursuant to the Public Procurement Code and put online in September 2011. As a
partner of the GEOSUD project, IGN then ortho-rectified these images and produced a mosaic of the whole
territory in pseudo-natural colors and published it on the national Geoportal.
This first attempt proved the legal and technical feasibility and the interest of acquiring a national annual
high resolution satellite coverage with a single multi-users licence. This approach was then renewed from 2011
to 2014.
In May 2015, the image collections acquired by GEOSUD and made available through its Web portal are
presented in Table 1.
Satellites
Spatial
resolution
Area
Period
Pleiades
0,5 m
Urban areas and areas with high issues (France
including overseas territories, southern countries)
From 2012 to 2015
SPOT 6-7
1,5 m
Whole French territories
2014
SPOT 5
5 m
Whole French territories, some areas abroad
2011/2012, 2013
SPOT 5
5 m
France
2005
Rapid Eye
5 m
Whole French territories, some areas abroad
2010, 2011
CosmoSkyMed
TerraSar X
Up to 1 m
Several areas in France and abroad
From 2013 to 2015
Table 1: Major GEOSUD image collections available in May 2015
Towards satellite direct receiving capabilities
To anticipate the end of the funding of satellite images on GEOSUD budget from 2016, a satellite direct
receiving station (DRS) has been installed in Montpellier in 2014 and 2015 by mobilizing several regional and
national funding sources. The GEOSUD DRS is currently composed of (1) an antenna of 5.5 meters in diameter
(Zodiac Data Systems) attached to an hexapod based on a platform of 8 meters in height and (2) a multi-mission
terminal (Airbus Defence and Space) able to receive and process SPOT 6 -7 telemetry, and potentially, Pleiades
and TerraSAR-X.
The choice of the first terminal of the SRD GEOSUD has been SPOT 6-7 because these satellites offer an
excellent compromise between high spatial resolution, ability to cover large areas and agility for custom
programming of new images. They are also very complementary of Pleiades, Sentinel 2 and Landsat 8 satellites
whose images will also be made available by the THEIA Land Data Centre.
Finally, six members of THEIA Land Data Centre have decided to pool their resources in a consortium for a
public contract of SPOT 6-7 telemetry with Airbus Defence and Space for a period of 5 years (2015-2019).
There are plans to expand the consortium to new partners and to acquire additional volumes of telemetry
depending on users needs and the availability of extra financial resources.
These direct receiving capabilities through GEOSUD DRS will be expanded in the next coming years with
additional types of images that will be chosen to complete the ones already accessible via THEIA.
User and uses
The dynamic of the number of GEOSUD members is shown in Figure 2 (temporal dynamic) and Figure 3
(geographic distribution). End of May 2015, 335 public bodies have joined the project, distributed between
research (85), State departments and agencies at different levels from national to local (124), local authorities
and territorial institutions (85) and 40 various actors (urban planning agencies, high schools, recognized
associations of public utility ...). It can be outlined that these public actors can temporarily provide GEOSUD
images to private sub-contractors, in the strict framework of non commercial activities. Under given conditions,
small private companies may also access these resources for research and development purposes.
Figure 2: Dynamics of GEOSUD public members (335 in May 2015)
Figure 3: Geographical distribution of GEOSUD members at the department level (May 2015)
Since the project began in 2011, 830 requests of images have been processed (311 from the scientific
community, 311 by the government, 117 by local authorities and 100 by other members) for a total of about
6700 images.
Total :
334
Guadeloupe
Martinique
La Réunion
Guyane
Mayotte
Nouvelle
Calédonie
St Pierre et Miquelon
The Rapid Eye 2010 national coverage, after 12 months of broadcasting, was valued by a factor of 6.7. The
SPOT 6-7 2014 national coverage has already been valued by a factor 9 after 5 months of broadcasting and 122
requests (more than 5 million sq km2 have been downloaded for a national territory of 550 000 sq km2).
Some examples of applications
The fields of application most frequently mentioned by the users are land-use dynamics, forests, agriculture
and biodiversity.
A first example of application deals with the spatialized monitoring and quantification of nitrogen pollution
pressures in western France to assess their impact on ground water quality. The November-December time
period, with heavy rainfall and low vegetation cover is a critical period for nitrogen leaching. Installing
interseasonal catch crops to ensure nitrogen fixation during the rainy autumn season is a major action
contributing to the European Nitrogen Directive.
Satellite images are used since 2012 by several governmental institutions to quantify interseasonal crop density
indicators, taking into account the dynamics of land use and agricultural practice. These data enable the public
servants of the Ministry of Agriculture to better spot the areas with no catch crop cover and to priorise site
inspections. Results contribute to assess the impact of sound agricultural practice on water quality, to orientate
public policies and to develop stakeholders’ awareness on the role and relevance of these practices.
Another example is the use of GEOSUD images for monitoring and controlling every year forest
clearcutting. A whole package of products and services has already been developed and is currently used in an
operational way by several regional services of the Ministry of Agriculture.
This package includes: images freely available; an algorithm for clearcutting detection implemented as a
QGIS plugin (And in 2016 also in the form of an online processing web service); a user manual prepared by the
experts of this application; a training kit (pedagogical material and 2 days training session); technical assistance
to the users. An online forum for sharing experiences among users and experts will be added by mid-2016.
Economic model in the context of THEIA
The ambition of GEOSUD is to boost the development of remote sensing by contributing, within the broader
framework of THEIA Land Data Centre, to create an ecosystem favorable to innovation. Besides access to the
technological platform, this ecosystem is betting a virtuous articulation between public institutions, research
teams (organized nationally in THEIA Centres of Scientific Expertise), higher education as well as private
players, to develop products and value-added services derived for all or part from remote sensing imagery.
New business models around this ecosystem under development at the national level must be i mplemented
within the next two years. The ideal solution would be to obtain sufficient funding from major public institutions
which benefit from this ecosystem to provide to the end users a free access to the products and services of the
platform. Otherwise it will be necessary to invent other models capable to charge to end users a part of the
operating costs of the THEIA / GEOSUD platform.
Acknowledgement
The authors of this work would like to acknowledge the financial support of the French Government handled by
the National Research Agency in the framework of the program "Investissements d'Avenir" for the GEOSUD
project (ANR-10-EQPX-20)
References
[1] Mathieu Kazmierski, Jean-Chistophe Desconnets, Bertrand Guerrero, and Dominique Briand. GEOSUD
SDI: Accessing Earth Observation data collections with semantic-based services. In Proceedings of the 17th
AGILE Conference on Geographic Information Science, Connecting a Digital Europe through Location and
Place, Castellon, Spain, June 2014.
See www.agile-online.org/Conference_Paper/cds/agile_2014/agile2014_138.pdf
[2] Marc Leroy et al, 2015. The THEIA Land Data Centre. RSDI Workshop, 1st October 2015, La Grande Motte,
France.
GEOSUD Web site
http://www.geosud.teledetection.fr/
... The EOAP was constructed in the context of the GEOSUD project (Maurel et al., 2015). The GEOSUD project was triggered by the finding that public stakeholders working in the field of environmental management and public policies underuse satellite images. ...
Chapter
Full-text available
The satellite images have become an essential source of information to address and analyze environmental issues quickly, repeatedly and in a reliable way. The increasing number of remotely sensed images are the first impediments for data discovery, access and processing. In this context, it is critical to simplify efficient multi-sensors image-based data access and query processing to provide accessibility to a variety of users in remote sensing. Describing satellite images through a metadata application profile may leverage capabilities to promote easy use of satellite image for environmental analysis. Accordingly, an application profile conforming to the Dublin Core Application Profile guidelines and designed for Earth observations data have been developed. The aim is to provide insights of key methodological considerations in relation to the design of this profile called EOAP (Earth Observation Application Profile).
Chapter
Full-text available
Spatial information linked to the observation of the Earth is increasingly present in natural resource management. Since the American Landsat program launched in 1972, many nations have set out to observe the Earth. More than 100 operational satellites are currently in orbit around the Earth; they allow access to geographical information from anywhere in the world, with increasing regularity. These satellite systems allow the monitoring of continental surfaces from local (submetric resolution sensors) to global (hectometric and decametric resolution) scales, with frequencies of observation going from daily to multiannual, and with different wavelengths going from the visible spectrum to microwaves. The growing expansion and complexity of the systems allow the development of increasingly specialized applications. The arrival of very high spatial resolution (VHSR), in the 2000s, expanded the scope of mapping applications that were previously reserved for aerial photography. In the near future, the Sentinel (S1 and S2) family of satellites from the European Copernicus program, providing free access to weekly or decadal decametric data, will allow the development of new services, in particular concerning real-time monitoring of the evolution of the landscape and use of resources.
Conference Paper
Full-text available
Highlights: This paper describes the Theia Land Data Centre, its service and data infrastructure, network of scientific expertise centres, product portfolio and main activities. The paper emphasizes in its conclusion the connection between Theia and the Copernicus programme at European level, and the forthcoming Remote Sensing Data Infrastructure network at international level.
Conference Paper
Full-text available
Ecosystems and territories are complex systems involving multidisciplinary approaches on different time scales and different locations. Yet, mastering the spatial information on these systems is critical to lead a relevant environmental research, as well as addressing efficient public policies. Although the volume of Earth Observation ( EO) data generated these last years greatly increased, their usages are still too limited when it comes to environmental issues. So as to remedy to this underutilisation, the GEOSUD (GEOinformation for Sustainable Development) project had undertaken to deploy a national Spatial Data Infrastructure (SDI) to ease access to high resolution and very high resolution satellite images for public stakeholders and scientists. This paper gives an overview of this infrastructure and places emphasis on the innovative components that fit the specific needs in terms of images discovery of the GEOSUD community. These components relies on domain controlled vocabularies, which can assist both experts and non-experts in the field of remote sensing in their search of the appropriate material to suit their needs
The THEIA Land Data Centre
  • Marc Leroy
Marc Leroy et al, 2015. The THEIA Land Data Centre. RSDI Workshop, 1 st October 2015, La Grande Motte, France.