Abstract: Previous multi-index bank-collapse prediction models are built on a river bank stability evaluation method that relies on the analytic hierarchy developed in recent years. As a key factor affecting the stability of river banks, determining the threshold values of hydrodynamic factors responding to different collapse risk levels is often empirical and lacks theoretical, systematic or practical research support. In this study, the thresholds of the key hydrodynamic factors－such as dominant discharge, flow velocity, water level variation, and channel diversion ratio change rate－are determined by integrating data statistics, theoretical derivation, numerical model calculation, and the previous results in literature. Determination of the dominant discharge threshold of a river and its water level variation threshold takes into account river-tide interactions in its tidal reach; a flow velocity factor is introduced in a formula for calculating riverbed stability. For a branched reach, the threshold of the diversion ratio change rate is calculated to distinguish its major branch channel from the minor ones, and we conclude the bank of the major branch is unstable at the diversion ratio change rate greater than 8%. The results support the development and application of an analytical bank-collapse-prediction hierarchy model for the lower Yangtze, and help prevent bank collapse disasters.

Key words: lower Yangtze River, bank collapse, analytical hierarchy model, dominant discharge, channel diversion ratio change rate, river-tide interactions