No abstract

The characteristics as well as the correlation between runoff and erosion was studied under 3 kinds of cover conditions of bare-slope, artificial grass slope and ecological-restoration slope with 20°slope gradient by runoff scouring intensities of 9l/min. The results showed that: (1) Sediment yield which sourced from artificial grass slope and ecological-restoration slope was separately decreased by 95% and 98% than bare slope, and the runoff reduced by 20% and 50% than bare slope respectively; The infiltration rates of the two kinds of grass slope were higher than the bare slope by 38.32% to 51.10% and 114%; (2) Sediment and runoff process showed that the stable state appeared respectively at 26min, 18min and 6min under bare slope, artificial grass slope and ecological-restoration slope, the amount of runoff was bare slope> artificial grass slope > ecological-restoration slope. (3) For bare slope and artificial grass slope, the relationship between sediment and runoff, sediment and the drag coefficient which both showed a negative correlation, but the correlation which from the ecological-restoration slope wasn’t obvious, so, further studies should carry out to promote runoff-sediment relations on ecological-restoration slope.

Sediment transport capacity of slope runoff is an important hydrodynamic parameter in the establishment of soil erosion prediction model. According to simulated runoff-scouring experiments, sediment transport capacity of slope runoff under different conditions is calculated. The impact factors of sediment transport capacity of slope runoff were analyzed by the method of Mean Impact Value, and then the input variables including dry bulk density, slope, Inlet flow, outlet flow, hydraulic radius, flow rate were determined. GRNN model was established and optimized by Adaboost algorithm to forecast Sediment transport capacity of slope runoff. The validation results showed that the GRNN model was applied to Sediment transport capacity forecasting of slope runoff. In conditions of experimental training samples, GRNN model had better computed results compared to BP Neural network model, and Adaboost algorithm could effectively decrease error of GRNN model.

No abstract

Runoff, sediment yield and infiltration process of shrub plots were studied under rainfall intensities of 45, 87 and 127 mm/h with 20° slope gradient using simulated rainfall experiment. The results showed that cumulative runoff and cumulative sediment yield of shrub plot had an obvious positive correlation with rainfall time. Under rainfall intensity of 45 mm/h, runoff and sediment yield of shrub plot kept a constant level. Under rainfall intensity of 87 mm/h, runoff kept a fluctuant increase, whereas sediment yield basically kept steady. Under rainfall intensity of 127 mm/h, runoff and sediment yield of shrub plot increased evidently due to the formation of erosion pits. Infiltration rate of shrub plot had a negative relation with runoff as well as sediment yield.