A harmonized, modular data quality framework facilitating cross-disciplinary usage and time-efficient evaluation of geospatial data

The increasing availability of spatial data coupled with the utilization of artificial intelligence, makes it essential to focus on the evaluation of data quality. At the same time, the fragmentation of existing quality frameworks hinders the attainment of comparable assessment results. We introduce a novel, modular framework for the evaluation of geospatial data quality with particular emphasis on FAIRness, transferability, reusability and spatial consistency. The framework thereby accommodates data of differing data processing levels, types and contexts. The hierarchical structure integrates common quality dimensions (e.g., completeness, accuracy, consistency) with new dimensions emphasizing upstream validity (metadata, traceability of input data, reproducibility) and downstream usability (applicability, transferability). Additionally, the framework enables the evaluation of two interlinked concepts: general data quality (DQ) and data adequacy (DA). The latter incorporates the relevance of data and the fit to use case-specific requirements. DQ and DA are measured through a combination of machine-evaluable metrics and structured expert judgment, aggregated as indicators on dimension and domain level. The assessment protocol is implemented in form of a spreadsheet and a web-based survey tool. The overall objectives of this development are (1) to achieve harmonization of existing quality concepts to facilitate cross-disciplinary data integration; (2) to support data selection processes in geospatial applications which involve multiple data sources and/or time-critical situations, through the reusability of evaluation results; and (3) to leverage the reflected data usage and integration into operational workflows through the consideration of spatial uncertainties and the implementation of aspects of FAIRness.