Sensor Data Quality for Geospatial Monitoring Applications

Nowadays, emerging applications in the geographic domain, exploit more and more geospatial data coming from sensors like the environmental monitoring systems. Increasingly, this kind of systems reveals their concern about the quality of discovered data. Considering monitoring as a main component of environmental crisis management system, an early and reliable detection of critical events is crucial. Thus, in order to make meaningful decisions with sensor data, the quality of data exploited through the system must be ensured. In this paper, we address this issue with an approach oriented to define and manage the quality of sensor data for geospatial applications, especially in a monitoring context. More specifically, our approach proposes a study and modelling of quality properties with considering sensor data specificities and the collateral factors impacting its quality. In order to qualify data, we adopt a product-oriented view supported by the management of metadata in real time; we analyze sensor data from acquisition to discovery. We also propose a visualization interface for sensor data quality discovery.