Virtual Observatory Infrastructure for Earth Observation Data
Earth Observation (EO) data have increased considerably over the last decades with satellite sensors collecting and transmitting back to Earth several terabytes of data per day. This data acquisition rate is a major challenge to existing data management, exploitation and dissemination approaches used by various agencies such as ESA, NASA and European national space agencies. The availability of earth observation data now outstrips the availability of human intelligence to exploit it.
To make the available petabytes of EO data easily accessible by an even larger group of end user applications, the proposed project will design and implement a Virtual Observatory infrastructure for EO data. The Virtual Observatory will use state-of-the art scalable DBMS technology to store and query raw image data, image metadata and symbolic representations of images capturing their content and its semantics. We will develop efficient algorithms to integrate images with relevant geospatial data sets, and to discover knowledge which currently lies dormant in EO archives. Because satellite instruments transmit data continuously, we will pay special attention to the streaming nature of EO data and will make significant contributions to querying streams of EO images and relevant geospatial data sets.
Our advances to the state of the art are demonstrated in two use cases: "A Virtual Observatory for TerraSAR-X data" and "Real-time fire monitoring based on continuous acquisitions of satellite images in combination with geospatial data".
With the support of the European Space Agency, EO scientists will be involved in our project throughout its entire life-time, from requirements definition, through development to testing and validation, thus making sure that the our results have a lasting impact in the EO community.