Integration of data and computing infrastructures for earth science: an image mosaicking use-case

This paper addresses the emerging issue of integrating data sharing and computing e-infrastructures for multidisciplinary applications. In the recent years several solutions have been proposed to implement digital infrastructures for sharing and processing scientific data and observations. Spatial data infrastructures currently enable effective and efficient geo-information data sharing in many disciplinary communities, and innovative solutions are under development to support new open data and linked data paradigms. In parallel, High Performance Computing systems, computing grids and more recently cloud services, enable fast processing of big data. However, the integration of data and computing e-infrastructures is a raising issue in multidisciplinary research. In the context of the Global Earth Observation System of Systems (GEOSS) initiative, an innovative approach has been proposed. Taking into account that the heterogeneity of data and computing e-infrastructures and related technologies cannot be reduced beyond a certain extent, since it is due to the need of supporting use cases and scenarios from different scientific communities, a brokering solution has been designed and developed. A Business Process Broker (BPB) is a component which takes a formal description of a scientific business process, and translates it in an executable process which can be run on multiple and remote processing and workflow services. In doing this it solves all the interoperability issues in a (semi-)automated way. It allows lowering the entry barrier for both computing service providers and users, decoupling the specification of the scientific process from the underlying enabling infrastructures. The paper presents and discusses a BPB use-case from the European project IASON, implementing an Earth Observation application involving satellite image mosaicking, HPC computing services and spatial data e-infrastructures.