STAD: An Ontology Design Pattern and Ontology for the Semantic Representation of Aggregate Spatial and Temporal Data

Advances in data collection technologies have resulted in the availability of vast amounts of spatio-temporal data across environmental and scientific domains. However, storage constraints and privacy and security concerns demand dissemination of such data only in aggregated forms rather than as individual observations. While statistical aggregation helps summarizing and interpreting large-scale phenomena, most aggregated spatial and temporal data published on the Web lack detailed explicit semantic annotations, making it difficult to retrieve, integrate, and reuse the data correctly. This limitation hampers data interoperability and poses challenges for large-scale analysis. To address these challenges, this paper introduces a novel Ontology Design Pattern (ODP) for representing the semantics of statistically aggregated spatial and temporal data, enabling clear specification of aggregation characteristics such as spatial and temporal support, resolution, and transformation method. Building on this pattern, we present the STAD ontology as a concrete OWL 2 implementation that formally encodes these semantics in a machine-interpretable form. It supports reasoning over the spatial, temporal, and statistical dimensions involved in the aggregation process. We evaluate STAD through a set of competency questions and demonstrate its use with a case study involving temperature data from two SCAN sites in New Hampshire. Together, the pattern and its implementation offer a foundation for semantically consistent publication, integration, and analysis of aggregate spatio-temporal data.