The Yellow River experienced many major floods in the past, often with catastrophic effects for the North China Plain. After establishing the Yellow River Conservancy Commission (YRCC) of the Ministry of Water Resources, a number of reservoirs were constructed for flood control and water resources management. For the mid and lower Yellow River, the operation of the multi-reservoir system plays an essential role in minimizing possible damage in relevant regions. In order to deal with changing reservoir storage conditions and variable river discharges due to climate change, adaptive management procedures are required.

At present the decision making process takes place in collective management meetings with the support of numerical simulations for flood simulation. There is a need to develop new software tools to achieve more effective decision support within precious lead time. Also, a special robust routing technique was needed as developed in this thesis for accurate and robust flood routing that can deal with multi-valued storage-release relations.

Nowadays advanced numerical flood simulation models are available with great level of detail and high computational efficiency. Also, appropriate software architectures are capable of combining model-based simulation systems with a data-centered approach at near real-time operational performance. Case study applications show that such approach is feasible for the next generation software systems for adaptive management and decision support of complex river systems.