Soil erosion and sediment yield in the upper Yangtze, China

Abstract

Soil erosion and sedimentation are key environmental problems in the Upper Yangtze because of the ongoing Three Gorges Project (TGP), the largest hydro-power project in the world. There is growing concern about the rapid increase of soil erosion over the last few decades and its consequence for potential sedimentation in the reservoir. The study aims to examine controls on the spatial and temporal distributions of sediment transfer within the Upper Yangtze and the hydrological consequences of land use changes, using varied approaches at different catchment scales. First, soil erosion and sedimentation are examined using the radionuclide Cs-137 as a tracer within a small reservoir catchment in the Three Gorges Area. The results indicates that soil erosion on sloping arable land and the rates of reservoir sedimentation have been severe during the past 40 years, mainly due to cultivation on steep slopes. Changes in reservoir sedimentation rates are mainly attributed to land use changes. The suitability of the Cs-137 techniques for investigating soil erosion and sedimentation in intensely cultivated subtropical environments is also considered. The use of the technique for erosion investigation may have limitations due to the abundance of coarse soil textures, uncertainty about fallout deposition rates and the high incidence of human disturbance, but the technique shows promising perspectives for sedimentation investigation since a few dating horizons might be identified. Second, sediment and runoff measurement data for around 30 years from over 250 hydrological stations within the Upper Yangtze have been examined within a GIS framework. The dataset has been integrated with catchment characteristics derived from a variety of environmental datasets and manipulated with Arc/Info GIS. The analysis of the sediment load data has permitted identification of the most important locations of sediment sources, the shifting pattern of source areas in relation to land use change and sub-catchments exhibiting trending sediment yields corrected for hydrological variability. The study demonstrates the importance of scale dependency of sediment yield in both the identification of temporal change and the modelling of relationships between sediment yield and environmental variables, suggesting that the treatment of the scale problem is crucial for temporal-spatial studies of sediment yield.