Development of an Operational Expert System for Flood Risk Management Considering Prediction Uncertainty

The aim of this project is to develop an efficient management tool for floods in meso-scale watersheds. The flood warning system should thereby be faster and more reliable. The basis of the investigation is the analysis of the whole "flood chain": "forecast of rainfall events - rainfall-runoff-model - hydraulic model".


Quantification and minimisation of uncertainties

In contrast to conventional methods, the individual uncertainties of the different models and parameters are being quantified. The consideration of uncertainties time-dependent, as a short forecasting horizon increases the reliability of the forecast. As a result, an adequate warning system can be established and the number of false alarms can be minimized.


Effective and operationally applicable system

Based on the results of the investigation a fuzzy-based expert system will be developed, which is operationally applicable and easy to use, so that an early warning of the population will be possible. Flood alert systems will be activated for each warning level, which is defined by specified threshold values. Furthermore it will be possible to assess the flood risk by analysing the likelihood of danger. Finally, damages will be able to be determined based on (dynamic) flood maps.

Prinzipskizze zum strukturellen Aufbau des Forschungsvorhabens

Partners and Co-Workers


Download > HORIX detailed description

Download > Project HORIX Poster in English

Download > Poster (sub-project 1) project management, model compound and development of the internet-based expert system

Download > Poster (sub-project 2) methods for the determination of the uncertainty of N-A-models

Download > Poster (sub-project 4) rainfall forecast and expert system

Download > Poster (sub-project 5) uncertainty of hydraulic models

Download > Poster (sub-project 6) quantification of the uncertainty and dependence of model parameters