Special Issue of the Journal of Web Semantics on Geospatial Semantics

Geospatial reasoning has grown larger scope in the semantic web. Increasinglymore information is geolocated,moremobile devices produce geocoded records, and more web mashups are created to convey geospatial information. Semantics can support the integration of geospatial information, track the provenance and quality of the data shown to an end user, and improve visualizations and querying of geospatial data. This special issue highlights recent trends in research and practice in geospatial semantics. The availability of geospatial data in semantic web formats is constantly increasing. Examples include geonames.org, with geographical information of all countries in many languages, linkedgeodata.org, which offers OpenStreetMap data in RDF queried through SPARQL, data.ordnancesurvey.co.uk, which publishes all kinds of products from the UK’s mapping agency, and geo.linkeddata.es, which contains data from Spain in RDF. A diversity of standards have emerged in this area, including GeoSPARQL, GeoJSON, GML, and KML to name a few, in addition to other popular ISO standards that are used in Geographic Information Systems (GIS). Their merits and level of adoption vary, as the community puts them to the test in a wide range of application types and scales. The research community in this area is very diverse and well beyond semantic web researchers. Within computer science, it includes researchers in spatio-temporal databases, computational geometry, machine learning, and crowdsourcing of spatial information. Beyond computer science, it includes geographers, cartographers, and social scientists. There is strong interest in the commercial world and from governments, as well as in scientific research and social applications of geospatial data. This paints a very heterogeneous landscape of requirements and priorities, all centered around space and, inevitably, time. This special issue includes four papers with contributions in both research and practice in the area of geospatial semantics. Harbelot, Arenas, and Cruz present a model to track the evolution of geospatial entities over time, and apply it to track changes in land use and land cover in a region. They introduce the LC3 model, which uses GeoSPARQL and a model of time based on Allen’s classic temporal calculus. Here, semantic web technologies enable the system to reason about the geospatial information for two purposes. One purpose is to check the consistency of the observations and validate the geospatial information coming into the system. A second purpose is to make inferences to generate new knowledge from those observations. Their approach is used with satellite images of the European landscape, and identify for example when an area evolves from a coniferous forest to become urban fabric and eventually industrial units. A taxonomy of land use and land cover types enables the detection of temporal