POLDER and the Age of Space Earth Sciences : a Study of Technological Satellite Data Practices.

Summary This essay focuses on the "technological practices" of collecting, producing, storing, disseminating and using satellite data applied to the Earth sciences in France. Through the description of a process of "conciliation-normalization", we study the efforts made by the promoters of space technologies (at CNES, CNRS and universities) to enlist a large scientific community in the execution of space experiments, and in particular to attract researchers from various fields of Earth sciences. On the one hand, space technologies must adapt to the practices and representations of these disciplines. On the other hand, the appropriation of satellite data requires researchers to learn, because the types of data are foreign to their expertise. It is during this process, taking place between 1980 and 2000, that research objects, tools and methods, as well as scientific communities were forged, communities sharing the belief that satellite data (in many forms) are relevant tools for scientific investigation. Our case study focuses on the processing of POLDER radiometer data. The socio-technical configuration at stake, borrowed from NASA, operates a number of framings. It gives a dominant place to data and processes with an explicit geophysical meaning, compared to other types of variables. It puts forward a specific epistemic community. It privileges certain technological practices, as well as particular forms of data production, storage and delivery. It also influences the architecture and planning of missions. Finally, this dynamic guides the ways in which data are used and makes them difficult to use by other researchers. We have thus analyzed the specific case of certain climate modelers whose work requires data produced according to other representations, sometimes opposed to the "norm".