Simulation Of Soil Erosion In The Yellow River Basin

Wang Guangqian,Li Tiejian

2010-01-01

Abstract:One of the key concerns on the Yellow River basin is the intensive (as high as 5,000 t/(km(2) . yr)) and extensive (the coarse sediment source region covering 78,600 km(2)) soil erosion in its middle reach, the Loess Plateau, where serious soil loss is the root cause of local environmental and ecological degradation as well as downstream sedimentation. The mechanism of soil erosion in the middle Yellow River is of great complexity mainly due to two reasons: 1) The earth surface in this region is highly incised by numerous streams and gullies, which lead to steep gully regions where gravitational erosion happens; 2) Soil erosion in this region is primarily caused by high - peak short - duration rainstorms, and then hyper concentrated sediment - laden flow in the streams is highly unsteady and non - equilibrium. Therefore, high demand for physically - based soil erosion simulation is raised for the Yellow River basin. In this paper, the structure and key techniques for physically based simulation of soil erosion is well developed. The structure of simulation is based on the classification of different sub processes of soil erosion according to different local topographies. The sub processes are comprised of hillslope soil erosion, gravitational erosion in gully region, and hyper - concentrated flow routing in channels. Based on existing work, sub models for each sub process have been developed and integrated into the whole modeling system, the Digital Yellow River Model (DYRIM). Furthermore, to make the DYRIM capable for high - resolution (demanded by the complex topography and unsteady flow) and large scale simulation, methodologies for drainage network coding and parallel computing have been developed and described in this paper. At last, the DYRIM is applied for continuous simulation of rainfall - runoff - soil erosion in a watershed in the Loess Plateau, the Chabagou Watershed. Satisfiable results are achieved and prove the effectiveness of the proposed model.