IEEE Geoscience and Remote Sensing Magazine - December 2020 - 18

processing (DSP) for a public-use delegation of those
from Pleiades. Through the DIAS pooling mechanism for the
Sentinel images, three usage levels are present. First, open
access to the satellite images is provided via the DIAS platform. Second, processing tools can be found to extract added
value and allow product and service development. Finally,
a third level groups these elements following a marketplace
logic. The DIAS infrastructure, the result of an industrial
consortium with European grants, raises several questions
about its sustainability and long-term economic model.
On the other hand, access to Pleiades images for public use
is conditioned on the establishment of a DSP agreement
between the CNES and ADS. Therefore, it seems necessary to
think about the place occupied by HR satellite images, that is,
between the free MR Sentinel and the VHR Pleiades images.
Furthermore, recent research confirms the benefits of having shared access to a multisensor and multiresolution image
bundle to cover a variety of application domains. For example, advances in artificial intelligence and machine learning
illustrate the use of VHR and HR imaging as a massive learning base for processing MR Sentinel-2 images [5], [6]. In fact,
the results of our study serve to highlight this issue through
examining the usefulness of this type of data, whose complementarity to MR and VHR images needs to be justified.
POOLING MECHANISMS
In general, the differences among satellite-imagery resolutions
are not trivial and can manifest as stark shifts, such as in landcover classifications, image sharpness, patch-level metrics,
pattern analyses, and so on. The data continuity, increased affordability, and improved access conditions are essential elements in the supply of HR satellite images, whose benefits
are numerous and could be found in several forms and applications (e.g., see [38] for tropical biodiversity studies, [10]
for biodiversity conservation, [35] for land-use classification,
[66] for geology, and [58] for precision farming). Furthermore, the cost of satellite imagery has a large impact on its use
and the resulting societal benefits [34]; if too expensive, it will
not be used extensively as originally intended.
In fact, efficient satellite-image pooling mechanisms, such
as the GEOSUD SDI, reinforce these facts. The access to and
use of a large library of HR satellite imagery play a role in
supporting the institutional services in the implementation
of their territorial planning missions, through the assistance
and integration of image-based space technologies into public policy systems [19], [61]. Furthermore, although satelliteimage users are increasingly close to their local and territorial
issues, the availability of a free HR satellite database allows
them to manage their day-to-day tasks in a more precise way.
In complementarity with MR and VHR images, HR images
offer a good compromise between spatial resolution and high
coverage capabilities. Thus, the GEOSUD SDI is allowing
many users who could not afford the price of the images to be
present in the satellite-imagery field. By adopting an upstream
financing strategy through public authorities and open access
downstream, the GEOSUD SDI is conserving a large base of
18

its subscribed users, despite their unwillingness to pay for the
images and services provided. Similarly, Amazon, Google,
and the ESA, with their new strategies for providing Landsat and Sentinel satellite images on their platforms, are also
heading toward expanding the use of and access to such data.
While pooling access to satellite imagery, they are creating a
networked community whose coordination and collaboration drives more innovation processes and the development
of added-value services and products.
Although free access and use provide great opportunities
for the community of satellite-imagery users, the lack of financial resources calls into question the sustainability of the
SDIs if this service continues to be offered at no cost. Hence,
the outcomes of this study could be used to secure public funds for SDIs by providing to public bodies all of the
impacts of and justifications for pooling-mechanism strategies. The study could also be used to develop appropriate business models to respond to the free supply of HR
satellite data. The WTP results make it possible to build
economic models based on the practices of the SDI's direct users. Moreover, they allow refinement of the existent
economic scenarios with the least negative impact on direct
users of the satellite imagery. Although the GEOSUD SDI
is evolving toward the Data Terra National Research Infrastructure, which will include the Dispositif Institutionnel
National d'Approvisionnement Mutualisé en Imagerie Satellitaire (DINAMIS) pooling mechanism, it will maintain
elements to guide and situate discussions with its partners
about current funding opportunities (annual membership,
price per image, premium beyond a certain number of free
images provided, and so on), thus allowing for better future
strategic choices based on factual bases. To cover the full
costs of DINAMIS, estimated at €3.5 million/year (including the access to HR SPOT 6/7 and VHR Pleiades imagery
for noncommercial use), the targeted economic model is
based on upstream financing provided by a consortium
of six public bodies carrying DINAMIS. It will be complemented by a financial contribution from the direct users of
the service with a differentiated pricing policy. The financial
commitments made at the beginning of 2019 by DINAMIS
holders and the simulations of contributions from direct
users shape the business model implemented in 2020.
ORGANIZATIONAL-ROUTINE CONCEPT
The GEOSUD/Theia initiatives have profoundly changed the
landscape in France for the supply and access of commercial HR satellite imagery to public and academic institutions
by bringing major innovations, such as the logic of sharing
and pooling through an all-public actor, licensing strategy,
and archive of images. These images and, more generally,
the remote sensing technology produce savings beyond their
direct use [54]. Integrating them into processes that include
other added-value products and are capable of generating resources [51] explains users' WTP despite the lack of budget.
This issue meets the organizational-routine concept [3], [47],
which states that a service becomes fully valuable only once
IEEE GEOSCIENCE AND REMOTE SENSING MAGAZINE

DECEMBER 2020
IEEE Geoscience and Remote Sensing Magazine - December 2020

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - December 2020
