An integrated watershed modelling framework to explore the covariation between sediment connectivity and soil erosion

Abstract Quantitative identification of the covariation between sediment connectivity and soil erosion can contribute to provide the key information for watershed sediment management. However, this covariation and its spatiotemporal response mechanisms are still unclear, especially whether this covariation can be used as a basis for identifying critical source areas of sediment in large‐scale ecological restored watersheds. In this study, an integrated methodology framework by the revised universal soil loss equation (RUSLE), index of connectivity (IC) and sediment delivery ratio (SDR) was proposed to visually assess the spatiotemporal characteristics of erosion and sediment yield processes in the Yanhe Watershed with large‐scale ecological restoration from 1985 to 2020 and to identify the covariation between sediment connectivity and erosion in subbasins. The soil erosion estimated by RUSLE has decreased by over 80% since 1985 owing to increased vegetation cover and the effective implementation of soil conservation measures, but the upper reaches still have high erosion intensity due to differences in specific controlling factors such as topographic conditions and land cover, requiring focused soil conservation practice. The IC results showed that as the vegetation restoration and soil conservation measures in the Yanhe Watershed varied from year to year, their spatial and temporal patterns had a strong influence on the distribution of sediment connectivity, and some local areas in the middle reaches showed local minima of IC in 1995, 1998 and 2010 mainly due to the implementation of long‐term ecological restoration project. The developed IC‐Erosion maps indicated that areas with high connectivity but low erosion accounted for over 60% of the total watershed area from 1985 to 1999, demonstrating a reverse correlation between sediment connectivity and erosion. Meanwhile, over 40% of the erosion occurred in a few areas (approximately 20%) with high connectivity and high erosion from 2000 to 2004, which can be characterized as the critical areas of erosion. The methods and results of this study provide ideas for separately defining both erosion and connectivity and quantifying bi‐variable erosion–connectivity classification, which can be easily viewed on a scatterplot.