Web service-based exploration of Earth Observation time-series data for analyzing environmental changes

The increasing amount of Earth observation (EO) data requires a tremendous change, in order to property handle the number of observations and storage size thereof. Due to open data strategies and the increasing size of data archives, a new market has been developed to provide analysis and application-ready data, services, and platforms. It is not only scientists and geospatial processing specialists who work with EO data; stakeholders, thematic experts, and software developers do too. There is thus a great demand for improving the discovery, access, and analysis of EO data in line with new possibilities of web-based infrastructures. With the aim of bridging the gap between users and EO data archives, various topics have been researched: 1) user requirements and their relation to web services and output formats; 2) technical requirements for the discovery and access of multi-source EO time-series data, and 3) management of EO time-series data focusing on application-ready data. Web services for EO data discovery and access, time-series data processing, and EO platforms have been reviewed and related to the requirements of users. The diversity of data providers and web services requires specific knowledge of systems and specifications. Although service specifications for the discovery of EO data exist, improvements are still necessary to meet the requirements of different user personas. For the processing of EO time-series data, various data formats and processing steps need to be handled. Still, there remains a gap between EO time-series data access and analysis tools, which needs to be addressed to simplify work with such data. Within this thesis, web services for the discovery, access, and analysis of EO time-series data have been described and evaluated based on different user requirements. Standardized web services specifications, output and data formats are proposed, introduced and described to meet the needs of the different user personas.