|
Abstract
|
The development of a decision support system dedicated to the study of phenomena taking place in the environment requires a good representation of spatial knowledge: geometrical objects and spatial processes have to be taken into account for the construction of such a system. The aim of these spatial representations is to apply one or more models. Several systems have been developed in France recently for different natural hazards: snow avalanches, slopes stability, forest fires, snowdrift and torrential floods. For each of these applications, the representation of space has been studied and implemented separately in different manners.
The aim of the ARSEN project is to take advantage of this experience to build a kernel of spatial objects and methods which could be used as a generic tool. It will be integrated in future systems to organize and manipulate spatial knowledge in order to get a simulation of natural hazards and face real-time crises. The main advantages of such a tool should be: 1. (a) to allow an easy integration of different models concerning natural phenomena, by supporting the construction of the spatial representations required by these models;2. (b) to make the access to spatial knowledge as easy as possible, whatever this knowledge is (geometrical and topological description).In the first part, this paper gives a few samples of objects and processing methods used in the already existing systems. We lay emphasis on the analogy between these different applications and we try to make an analysis of the spatial representation requirements. Then, we suggest a definition of the kernel of ARSEN. In this second part we describe the spatial knowledge organisation and the common methods. In a third part, we describe the way in which ARSEN is implemented. This part is the most technical one, describing the way the links between the spatial objects are organized and programmed. Finally, we present applications where the kernel of ARSEN is used to take into account rock falls and snow avalanches. |