Abstract. An approach to improve an RCC-derived geospatial approximation is presented which makes use of concept inclusion axioms in OWL. The algorithm used to control the approximation combines hypothesis testing with consistency checking provided by a knowledge representation system based on description logics. Propositions about the consistency of the refined ABox w.r.t. the associated TBox when compared to baseline ABox and TBox are made. Formal proves of the divergent consistency results when checking either of both are provided. The application of the approach to a geospatial setting results in a roughly tenfold improved approximation when using the refined ABox and TBox. Ways to further improve the approximation and to automate the detection of falsely calculated relations are discussed.

Abstract. Environmental databases store a wide variety of data from heterogeneous sources which are described with domain-specific terminologies and refer to distinct locations. In order to make them accessible also to non-expert users, terminological concepts and spatial relations must be represented in a way that they can be exploited for searches. In this paper we propose a hybrid knowledge representation system architecture which integrates terminological and spatial aspects of the application domain and provides support for reasoning with RCC, a well-known calculus for spatial reasoning, and the Semantic Web's ontology language OWL. Our approach is motivated by the observation that RCC cannot be expressed in OWL without a major revision of the latter. Issues of upholding consistency of the knowledge base in view of an evolving ontology and of computational complexity are discussed.

Three severe hail-and windstorms, which occurred in Northern Switzerland, have been investigated. The storms produced hailswaths of 70-170 km in length and 12-25 km in width. Compact tracks of severe wind damages within the hailswaths are documented from public reports. These tracks have a length of 12-25 km. The damage was produced by straight-line winds rather than by tornadoes. Volume-scan Doppler radar data of the storms are available in time steps of 5 min. Radar signatures, such as low-level convergence and shear, mid-level vorticity, and high-level divergence, were attributed to the damage tracks at the ground. The low-level radar signatures allow the deduction of the time of occurrence of the damage tracks with a precision of some minutes.Striking similarities in the evolution of the three storms were found. The storms developed in the foothills of the Alps and the Jura mountains and propagated towards the plains of the Swiss midland. The storms can be classified as 'high-precipitation' supercell storms, known as producers of severe straight-line winds in the USA. Meso(anti)cyclonic vortex signatures were seen 40-55 min in advance of the heavy wind damage at the ground. The damage tracks were associated with explosive secondary cellular growth aloft. The physical explanation of this behaviour is that the first cells with mid-level rotation produced a gust front outflow that was accelerated to damaging strength at the time when the secondary cellular growth was initiated. The operational implication is that the nowcasting of severe and damaging winds can be improved by considering mid-level rotation in an early stage of the evolving storms.

No abstract

No abstract