Logo tutelle
Logo tutelle Logo tutelle
Logo tutelle


Sur ce site

Sur le Web du CNRS

Theme 4.1 Observation

Before we collect data, in order to describe and qualify spatial configurations and their evolutions, we need to consider relevant variables and indicators, both quantitative and qualitative, as well as their selection, representativeness and sampling (Janelle, 2004 [1]). Data collection itself requires context-relevant protocols, either by using established ones or by developing new observation and measurement methods. Prodig will engage in methodological research (deployment, improvement and development of methods) in the fields of metrology (the science of measurement and its applications) and remote sensing (or teledetection).

By observing change, we aim to acquire new data, in particular within institutional systems or observation networks (for instance, the observation of permafrost), or as part of specific programmes (monitoring older people’s access to care services in urban areas; monitoring ground-level or sub-surface temperature in wine-growing areas; monitoring water quality in the Andean foothills; characterising heterogeneity in stakeholder perceptions, etc.). One first strand of reflection aims to identify relevant variables and indicators for drawing up timelines: robustness of indicators, comparability over time, etc. (Cossart, 2011 [2]). A second strand considers the spatial context, by factoring in the influence of local (or neighbouring) configurations on observation. These considerations converge with the issue of sampling (random vs. purposive) in the acquisition of data. We are then faced with a double question: can spot observation be significant on a small scale, and therefore replicable on a regional scale? And can the multiple effects of local context be fed into regionally collected data?

The importance of methodological developments in the field of remote sensing is now widely recognized in geography as in other disciplines: images generated through remote sensing can provide a qualitative (land use, etc.) and quantitative (leaf area, rainfall…) picture of the territory and/or of our study objects, covering a much vaster area that what can be achieved using field metrology. Because data capture is easily replicable, these images make it possible to deliver diachronic analyses at very diverse intervals of time, from minutes to decades, depending on the satellites’ orbit and on research requirements. For instance, we aim to develop new indicators using satellite imagery to detect the ground’s salinity and the distribution of water tables in arid areas. The diversification of sensor types has also multiplied the potential for observation. For instance, using optical satellite images with an average spatial resolution can enable us to quantify eco-systemic responses to climate change: changes in bush landscapes in Zimbabwe caused by draughts, evolution of the grassy or ligneous cover in Sahel and its impact on communities, or seasonal evolution of the boreal vegetation (Delbart et al., 2008 [3] ). By using high- or very high spatial resolution optical images, we can for instance capture and monitor the evolution over several years of watering spaces in the Hwange Park (“Savarid” programme) which is used both by the population and by wild animals. This tool will also enable us to map out urban development, metropolisation and urban sprawl, or deforestation. Hyperspectral imaging has the potential to help us distinguish between specific types of vegetation and even species, while radar imaging is used to quantify land shifts, floods or the forest biomass.

Remote sensing thus combines repetitiveness with synopticity and with a wide spectrum of applications. However, each new application requires new methodological developments, including a physical analysis of the signal and the creation of digital data processing methods. These development activities and their dissemination will by supported by the Campus Spatial (“Space Campus”) at Paris Diderot University. These activities are also integrated with the Master in Remote Sensing and Geomatics applied to the Environment (“Télédétection et Géomatique appliquées à l’Environnement”) where several members of Prodig play an active part.

[1] Janelle D.G., 2008, « Spatial Reorganization: A Model and Concept. In S. Hanson and M.-P. Kwan, eds. Transport: Critical Essays in Human Geography », Aldershot : Ashgate Publishing Ltd. Reprinted from 1969, Annals of the Association of American Geographers, 59, p. 348-364.

[2] COSSART E., 2011, « Mapping Glacier Variations at Regional Scale through Equilibrium Line Altitude Interpolation: GIS and Statistical Application in Massif des Écrins (French Alps) », Journal of GIS, 3 (3), p. 232-241.

[3] DELBART N., Picard G., Le Toan T., Kergoat L., Quegan S., Woodward I, Dye D., Fedotova V., 2008, « Spring phenology in boreal Eurasia in a nearly century time-scale », Global Change Biology, 14, (3), p. 603-614.