Origin and Transport of Sediments in a Semi-Arid Environment

Abstract

Sediment transport in fluvial systems is governed by complex, non-linear interactions between water discharge and suspended-sediment concentration, giving rise to hysteresis loops that standard rating curves cannot capture. This thesis introduces an integrated, automated workflow for concentration–discharge hysteresis analysis. We develop the concept of a “hysteresis signature,” a systematic code that partitions each loop into linear, clockwise, and anticlockwise segments. We then quantify each segment both on a normalized scale (partial amplitudes) and in real discharge–concentration units via an asynchronous-load metric. Applied to a large multi-parameter dataset, our method automatically classified 51 distinct hysteresis forms—far beyond traditional five-type schemes—and resolved classification ambiguities inherent in existing indices. This approach provides a powerful tool for understanding watershed dynamics in terms of sediment production and for identifying sediment sources at the catchment scale. In a semi-arid case study on the Mekerra River (NW Algeria), we validated the workflow against successive flood events and found that bank-and-bed sources contribute roughly 84 % of total suspended sediment yield. The modular nature of the tool makes it readily adaptable to various climates and different watershed sizes, from large basins to small ones. The hysteresis signature method can also be applied to dissolved loads. The tool offers a robust, scalable approach for high-frequency monitoring, sediment-source apportionment, and watershed management.