The Yellow River (YR) covers three climatic zones including arid region, semi-arid region and temperate monsoon region, with frequent appearance of flow intermittence in the Lower Yellow River (LYR) before 1999. Channel migration occurs frequently in braided rivers, which is a major focus of study in geomorphology and river dynamics. The braided reach in the LYR is featured by a complexly spatio-temporal variation in channel migration parameters owing to the varying condition of flow and sediment. It is crucial to investigate the migration characteristics of channel centerline for the sake of fully understanding channel evolution. A detailed calculation procedure is proposed to quantify migration rates and intensities of channel centerline at section- and reach-scales, using the measurements of remote sensing images and cross-sectional topography. Migration rates and intensities of channel centerline at section- and reach-scales from 1986 to 2016 were calculated, with the characteristics and key factors to control the migration intensity of channel centerline being identified quantitatively. Calculated results indicate that: (i) the mean probability of centerline migrating toward the left side was approximately equal to the probability of rightward migration from a long-term sequence; (ii) the mean reach-scale migration rate of channel centerline was reduced from 410 m/yr in 1986–1999 to 185 m/yr in 1999–2016, with a reduction of 55% owing to the Xiaolangdi Reservoir operation in 1999, and the mean reach-scale migration intensity of channel centerline was decreased from 0.28 to 0.16 m/(yr·m), with a reduction of 43%; (iii) the incoming flow-sediment regime was a dominant factor affecting the degree of channel migration, although the channel boundary conditions could influence the intensity of channel migration; and (iv) the reach-scale migration intensity of channel centerline can be written as a power function of the previous two-year average incoming sediment coefficient or fluvial erosion intensity, and the reach-scale migration intensities of channel centerline calculated using the proposed relations are generally in close agreement with the measurements over the period of 30 years.

Hyperconcentrated confluent floods in the Middle Yellow River and its tributaries are characterized by the phenomena of extremely high backwater stage, flood water intrusion, and local channel blockage in confluence areas. In order to study the propagation process of hyperconcentrated confluent floods, most existing channel network models need reinforcement in the computation of sediment transport and improvement in the treatment of interior boundary at channel junctions. A river network model for hyperconcentrated confluent floods is proposed based on the finite volume method scheme, in which the study domain is represented by ordinary 1‐D cells and junction cells. As a preprocess step, a bathymetry interpolation algorithm is developed to generate topography data and to extract the stage‐volume relationship for the junction cell. The storage change and sediment transport in the junction cell are calculated in each time step with no need to distinguish convergent or divergent flow pattern, and the solution in each reach can be executed simultaneously. The model was validated by simulating the hyperconcentrated confluent flood events in 1977 and 1966, with the maximum backwater stage and the occurrence time of flood water intrusion in the Lower Weihe River being well predicted. The response was investigated of the Middle Yellow River and Lower Weihe River to the arrangement of peak times of confluent floods, with the critical condition for flood water intrusion and the scenario that achieves the most effective scouring in the downstream channel being identified.