Conceptual data models in Earth Sciences and GIS

As observational and model output datasets in the FES (Fluid Earth Sciences, mainly oceanography and atmospheric science) increase in resolution, there is an increasing demand for information systems that interoperate between the GIS and FES realms However, differences in the way the two communities think about their data can give rise to difficulties in integrated analysis and display of datasets from the two disciplines. Moreover, new kind of services (functionalities) are getting more and more important: interoperability services. They require data models suited for enabling web service to "understand" and easily exchange FES datasets. These data models must be particularly accurate on metadata and encoding model, which enable the effective sharing of data content and meaning, and hence the real system interoperability. XML encoding dialects, based on semistructured data models, seems to be an effective technology to facilitate FES and GIS interoperability. In our vision, these data models are extremely useful to achieve applications interoperability. They are useful to interconnect sibling services implemented on heterogeneous applications. A conceptual, abstract view of differences between the data models of the FES and GIS communities is presented, as well as a schematic description of where the data systems overlap and where they are distinct from each other. Well accepted data models (i.e. netCDF and OpenGIS/ISO coverage) are considered and presented. Examples of significant XML dialects (i.e. Markup Languages) for both communities are presented (i.e. GML and ncML), discussing an innovative mediation solution approach: the ncML-GML. This XML dialect is an extension of the netCDF Markup Language (ncML); it enables encoding of netCDF datasets in terms of the Geography Markup Language (GML) elements. It was conceived to facilitate interoperability between the FES and the GIS communities. NcML-GML leverages the ncML ability to encode multi-dimensional arrays, and the wide acceptance of GML for encoding geo-spatial metadata. NcML-GML is also particularly useful where both FES and GIS semantics and metadata content are required. For interoperability purpose, according to the Service Oriented Approach, it is possible to distinguish different service tiers; each tier contains systems and tools which implement its specific task. Such a framework for the FES and the GIS information communities is analyzed and discussed. The heterogeneity of existing FES protocols and data models is outlined. The interoperability opportunities laid by the OGC's specifications are briefly introduced. The need of achieving the two communities' frameworks interoperability, and its importance for enabling effective Spatial Data Infrastructures (SDIs) are introduced and discussed. The need for data model crosswalks as well as access and discovery protocol adapters is outlined. The mediation approach is a valuable technique to address this issue. A valuable framework, based on well-accepted protocol and data model interoperability standards is briefly presented and a main experimentation of its adopted mediation solution is reported..