Résumé
One of the main missions of the Observatoire de Recherche Montpelliérain de l'Environnement (OSU OREME) is to maintain and disseminate data. This observatory brings together 18 Observation Services from 8 environmental research laboratories, producing a wide range of data (sensors, visual observations, genetic analyses, cartographic data, photos) on a variety of themes (hydrology, geophysics, ecology, etc.).The Information Systems department develops the tools needed to make observation data available1 in compliance with the norms and standards of the field, and in particular with the European Inspire directive. The wide diversity of data types and scientific themes raises many questions concerning data acquisition, database structuring, use of reference systems, distribution, interoperability, cataloguing, etc.We try to take as general an approach as possible in building up the information system, firstly in terms of methods and interaction with data producers, then in terms of tools, standards and norms, and finally in terms of developments.The raw data is structured in relational databases, as are the higher levels (corrected, calculated, validated data, etc.). Numerous descriptors enable the context of the data to be precisely defined. To ensure data interoperability, we use repositories in the data and their descriptions, such as taxonomic or geographic repositories, or environmental thesauri. This approach enables us to deduce new information, automatically enrich data and link them together.Data is catalogued in a standardized way, automatically re-using all the descriptions associated with the data. Similarly, DOIs can be generated on our datasets, although compliance with the rules imposed on a Datacite-referenced dataset requires more complex thinking and strategies than simply creating a DOI.Finally, our data is distributed in compliance with Open Geospatial Consortium standards, using WMS, WFS and, in the near future, SOS protocols.In this presentation, we will detail our approaches to database construction, data enrichment, data qualification, referencing and distribution, in relation to the 4 FAIR principles.