Quantifying sediment complexity and its implications in the Yanhe watershed using refined composite multiscale fuzzy entropy

Abstract The Yanhe watershed, located in the Loess Plateau of China, is a representative loess hilly-gully and ecologically fragile region, posing a critical challenge for sustainable sediment management. Regarding the watershed as a vibrant complex system, this study investigated sediment concentration dynamics in the Yanhe watershed from the perspectives of multiscale entropy and complexity. The refined composite multiscale fuzzy entropy (RCMFE) method was employed to quantify the sediment complexity and explore its variations and driving mechanisms. Three characteristics of the RCMFE curve were evaluated: local maxima, local minima, and volatility. The results revealed significant increases in sediment instability during 2000–2004 and after 2008, and the richness of variation information and complexity of sediment in the past two decades have notably decreased compared to the baseline period before 1970. Although complexity decline has been controlled since 2005, it remained at a low level. The notable decrease in the multi-scale structure and function will weaken the watershed's resistance to external interference and environmental changes, highlighting the need for widespread attention. Major factors contributing to complexity changes are China's Grain for Green Program (GFGP) and vegetation restoration, alongside the growing impact of erratic precipitation in the past decade. This research contributes to a multiscale understanding of sediment concentration dynamics, and provides new insights into sediment processes and sustainable strategies on the Loess Plateau under the GFGP and climate change.