Decoding bed armoring: A predictive method for grain size distribution in sand–gravel bed rivers

Abstract

Sand–gravel riverbeds downstream of hydraulic projects frequently undergo significant bed armoring. Accurately predicting the grain-size distribution of the armored layer remains a critical yet complex challenge within the field of fluvial dynamics. This study focuses on the response mechanism that connects the progressive reduction in fine sediment content with the evolution of hiding–exposure effects, the consequent alterations in sediment state transition probability, and the subsequent adjustment of the armoring rate during the process of riverbed armoring. An expression for sediment entrainment probability is derived under dynamically varying conditions, which include armoring intensity and the hiding–exposure effects of non-uniform sediments. A three-state exchange mechanism among suspended load, bed load, and bed surface material is integrated into the armoring calculation process. A sediment mass conservation relation under depositional conditions is also formulated, leading to the proposal of a novel stability criterion for bed armoring. Building on these advancements, a predictive method for determining the grain-size distribution of armored layers is developed, explicitly considering sediment exchange processes. Application of this method to the downstream reaches of the Danjiangkou and Three Gorges Reservoirs in China reveals that the predicted grain-size distribution of the armored layer deviates from field measurements by only 2.9 % on average, and from flume experiments by 4.1 %. The findings indicate that the proposed method effectively replicates gradation trends, grain-size distributions of armor layer, and scour depths under clear-water conditions. Furthermore, within the research framework of multi-stage armoring, characterized by the cyclic “formation–destruction–reformation” of armor layers, the method demonstrates considerable potential for applicability.