Environmental pressures, such as overexploitation of natural resources and pollution, are urgent concerns affecting the Earth’s global system. Both experts and non-expert stakeholders need access to meaningful Open Data to analyze the environmental impact of global warming in an area of interest, and thus implement effective environmental policies. A large and free collection of Earth observation (EO) data from satellites is currently available. However, EOs are sensor data whose interpretation is reserved for specialists due to the lack of semantic contextualization (e.g., definition of satellite-derived data, metadata, etc.). Moreover, although several resources are already accessible on the Web, most EO data remain isolated, whereas linking them together would provide a comprehensive understanding of environmental changes over time. In this paper, we present the Linked Earth Observation Data Series (LEODS) framework that leverages Semantic Web (SW) technologies for the integration and publication of EO data in the Linked Open Data (LOD) Cloud. LEODS relies on a spatio-temporal modeling approach which complies with SW standards and ensures future semantic enrichment of EO data. Precisely, LEODS is the first step towards the main objective of the French-Swiss collaborative project called TRACES, that is, to build a Knowledge Graph (KG) integrating EO data with various data sources (socio-economic, urban, legislative texts, etc.) to monitor the environmental evolution of areas of interest. To highlight the advantages of our proposal, we present and explore, through SPARQL queries and visualizations, the results of implementing LEODS with data relevant to the TRACES project.