The main goal of the present project is to improve drastically the ′real time features′ of current applications in order to provide services under time constraints typical of natural disaster management (i.e. hours and not days).
Events such as floods, earthquakes, forest fires and oil spills etc. can then be covered through space observation, measuring the impact of these disasters on the environment and allowing for better support to the impacted people or environment.
The targeted system will couple meteorological and environmental data in real time to provide either early warning or relief systems within a decision support architecture compatible with the time frame of natural disasters. The project initiated by a consortium of service providers and RTD providers is aimed at developing and implementing two generic functions and two specific service applications within the existing portal: The generic features include the fast delivery through the Internet of large, high definition, satellite-generated images as well as the geo-referencing of satellite images within very short deadlines (less than three hours as compared to about a week today). The service applications comprise an image processing capability designed to analyse storm damage in forests and to implement object oriented techniques for image enhancement and segmentation in land-use classification and optimisation as well as an image processing capability for assessing flood damage.
Once operational, such real time capabilities will also help service providers and their client end users to build a set of data which describe past events related to natural disasters in appropriate detail and which, therefore, allow the identification of good practices for the management of similar natural disasters in the future.