Résumé

Using a selection of globes from the previous corpus, the aim here is to analyze the spatial distribution and typology of the data used to produce the digital and interactive images of the global environment (3.1). A method is developed to reveal the data deserts (3.2). It should also enable interrogating the ghost data that fill in the blanks via the monitoring of the contamination of the databases by others.

Méthodologie

Data genealogy is carried out on a selection of globes, chosen according to the specialties of the team. Spatial analysis are used to visualize their distribution using an innovative method of identification of data deserts which is designed in part as an extension of the GÉOBS project and as part of a thesis in GIScience. A minima, the following three virtual globes are identified:

• Biodiversity: the globes produced by the GLOBAÏA company for the Mapping Biodiversity project are analyzed by spatializing the distribution of the 1.8 billion (as of August 1, 2021) observations by the GBIF that serve as the basis for such global biodiversity images [35].
• Deforestation: Global Forest Watch provides weekly indicators [36] that drape many globes. Heterogeneous data, spatio-temporal coverage and resolutions are studied.
• Marine ecosystems: since 2012, Google Ocean has been used to explore marine ecosystems. The aim is to analyze how a heterogeneous data set produces a vision of the global ocean [37].

Résultats attendus

Besides the catalog of virtual globes (WP2), WP3 feeds the website with targeted case studies with:

• Indicators to qualify the distribution of the data
• A method and a geovisualization interface to explore data deserts
• Scientific articles, but also storymaps that can be used as teaching aids (III.3).

The data and codes produced are disseminated on institutional repositories under a open license.