P1.48 FEDERATING EOS HETEROGENEOUS AND DISTRIBUTED INFORMATION RESOURCES

Environmental and geographical data describing atmospheric or terrestrial phenomena are collected in digital form since three decades. Geomatics, that is automatic management and processing of environmental and geographical information, is changing from a niche discipline to a horizontal application area, involving scientific as well as industrial and institutional bodies. Nowadays, in the “Internet Era”, it is reasonable to expect this wealth of data to be accessible and shareable in a simple and coherent manner among the different actors of the geo-science arena, integrating the existing information systems with the ones that future missions and research activities are set to provide. The key factor for these expectations to become reality is interoperability, which may be generically defined as “system cooperation for the sake of information and process sharing”. Guaranteeing interoperability means to solve such problems as: domain ontology heterogeneity; data encoding mismatch; software application distribution and heterogeneity; data protection and accessibility. In the geophysics sector, these questions are even more complex, because of the data inherent multidimensionality (e.g. spatial, temporal and sampling features). In this overall picture, the Earth Observation from the Space (EOS) community is a cross-sector community encompassing Meteorology, Oceanography, and Hydrology, with specific needs for on-line cataloguing and interactive information processing, due to: the enormous amount of existing data, the present, ever-increasing data growth rate (caused by remote sensing technologies evolution, namely in satellite imagery) and the intrinsic heterogeneity level among data producers. The EOS community has always faced problems such as acquisition, management and access to huge data amounts, implementing largely heterogeneous and autonomous solutions under the logical, the methodological and the technological point of view. Today, distributed computation and networking technologies prove mature enough to allow standard connectivity and access to multi-medial data, but the real problem is the integration of higher level information, that is data along with their ontological and operational semantic significance. There have been efforts in standardization of data types and methodologies, with various degrees of success and diffusion (see http://www.statkart.no/isotc211, http://www.opengis.org ), but their adoption on a large scale remains hindered by the actual difficulties to adapt any common model to the peculiarities of existing data collections.