Deliverable D1.2

Report on integrating dataflows for mapping and assessing ES and linkages with eutrophication across realms

This deliverable provides a unified operational framework for harmonising diverse spatial, temporal, and thematic datasets required for modelling aquatic ecosystem services (ES) and eutrophication dynamics across terrestrial, freshwater, coastal, and marine realms. As ES-relevant datasets are produced for different purposes, with different resolutions, update cycles, and semantic structures, substantial discrepancies arise when they are combined for cross-realm analysis. Such inconsistencies, such as misaligned spatial coverage, different sampling frequencies, non-synchronous reference years, or different classification systems, compromise reproducibility, comparability, and model performance. To address these challenges, the report introduces a typology of common mismatch types, outlines process-appropriate integration strategies, and presents a catalogue of methods for spatial, temporal, cross-realm, and thematic harmonisation. Within the CREATE project, this deliverable complements D1.1 by translating identified data sources into modelling-ready inputs via structured workflows that support WPs 2 and 3. The guidance provided enables practitioners and researchers to overcome fragmentation in aquatic ES data flows, ensuring that existing monitoring data, Earth observation products, and model inputs can be integrated into coherent, interoperable datasets suitable for cross-realm assessments and nature-based solutions (NbS) design. Additionally, this material is suitable for use as a teaching and training resource in higher education, particularly within curricula related to geography, spatial data science, environmental modelling, and ES assessment. As CREATE progresses, these workflows will be iteratively refined. This deliverable is also intended as a living document and will be updated regularly. Future updates will expand the suite of integration methods, strengthen links with partner-contributed datasets, and incorporate user feedback from the modelling teams. An updated version is planned for February 2027.