Abstract:
During the last decades, private, governmental and non-profitable organizations have been
developing information systems to monitor, alert and manage environment-related emergencies
[1]. Platforms, such as flood emergency alert and warning systems [2], comprise an extensive
network of sensors, a bundle of forecast simulations models, and decision-support modules that
rely largely on a robust and reliable perception of the conditions of the physical monitoring
environment. Sensors provide this insight of the real world, where the notion of continuous time
and continuous values of implicated phenomena meet the computerized notion of discrete
model of time and discrete estimation of the real data.
Complex and powerful forecast systems are now able to predict water levels or to track
storm events with low errors, but they depend on a continuous confirmation with data. Realtime
monitoring data, such as surface water elevation, flow or precipitation depend solely on
the sensor hardware deployed at the water bodies (oceans, river, lakes, etc…). The goal of this
paper is to propose a framework to increase the validity of the information provided by these
sensors.
