Abstract
During the analysis and management of flood events, the issue of time is crucial. In fact, such events require accurate knowledge of the time of their occurrence and their temporal relationships. However, in most cases, temporal information about flood events are uncertain. In this paper, we propose an intelligent system for managing the temporal uncertainty of these events. The modeling of this uncertainty is based on the belief functions theory and Allen's Interval Algebra. In fact, an uncertain flood event is represented using temporal intervals. Each interval is associated with a belief mass. If there is different uncertain information provided from different sources, a reasoning phase is carried out on these data in order to obtain more reliable information.