Fractal Characteristics Research On Microtopography At Loess Artificial Digging Tillage Slope In Water Erosion

Study on the regulation evolution of micro-topography on millimeter-scale as well as its influence on the runoff and sediment yield during the soil erosion process can provide a theoretical reference for the further understanding of the soil erosion process and for the establishing of a soil erosion model.The wide use of fractal theory makes it possible that the fractal dimension can be used to characterize the complex features of surface micro-topography.Predicting the future trend and primary cycle of runoff and sediment yield,it has an important value on monitoring the impact of continuous rainfall on soil & water conservation and soil erosion.Thus,artificial digging(AD)tillage treatment was chosen,which were commonly used in loess slope farmland,and then piecewise rainfall expriments were conducted under rainfall intensity(90 mm/h)and different slope gradients(5°、10°、15°、20°、25°).Obtaining the micro-topography relative elevation data based on high-precision laser scanning,semivariogram analysis、rescaled range analysis、box-counting method are adopted to investigate the influence of the surface micro-topography during the soil erosion process and calculating the fractal dimension.Simultaneously the changes of runoff and sediment yield during the development of water erosion were studied systematically.The main conclusions showed as follows:(1)The results of the semivariogram analysis showed that: The parameters of the semi variance function model could be used as point cloud data for the Kriging interpolation in Arcgis.The spatial autocorrelation of micro-topography was mainly affected by the structural factors such as slopes.In addition,the Caussian model is the best-fitting variogram model for the artificial digging sloping surface with a rainfall intensity of 90 mm/h.The mean codomain is 4.900 m that provide appropriate scale reference for future research under similar conditions.The artificial digging surface is a durable surface with a positive spatial autocorrelation characteristic.A clearly emerging trend is that the spatial fractal dimension and the spatial variability increases with the increases of slope in the same erosion stage.Precipitation at the SpE,ShE and RE stages initially decreases surface roughness,and there is an enhancement of surface roughness at the RE30 erosion stage.(2)Semivariogram analysis、rescaled range analysis、box-counting method are adopted to calculate the fractal dimension which the semivariance method is the best.Comparatively,results from the semivariance method are more specific and reliable.(3)The anisotrophy analysis results showed that: During the development process of water erosion,the micro-topography of slope farmland had obvious anisotrophy and has a clear spatial heterogeneity feature.The anisotrophy was influenced by slope,which was in、 creased with the increased slope.(4)The variation characteristics of runoff and sediment yield showed that: With the increased slope,the runoff speed of artificial digging slope is increased.Analysis runoff sequence by R/S analysis method,which indicated that the runoff sequences had positive long-term correlation under the five slopes.The runoff will be increased with the increase of rainfall,but the increased trend is weak.Morlet wavelet method was adopted to identify the primary cycle of runoff sequences,results showed that the primary cycle of runoff sequence was 29 min.Runoff and rainfall would be in a logarithmic function relationship,and the runoff prediction equation is S = alnt + b.Based on rescaled range analysis method analysis sediment sequences,which indicated that the sediment sequences had positive long-term correlation.The sdiment will be increased with the increase of rainfall and the increased trend is strong.Analysis sediment sequences by Morlet wavelet method,compared with the runoff sequence,there is no stable period,while part of the slope had the consistent cycle with 30 min which need a longer time series to verify.Sediment and rainfall would be in a logarithmic function relationship,and the sediment prediction equation is V = alnt + b.This work established a framework of the R/S analysis method integrated with Morlet wavelet method for predicting the future trend of runoff and sediment yield,it also has an important significance for further research on the agricultural production and soil & water conservation.In this paper,semivariogram analysis、rescaled range analysis、box-counting method are adopted to analyze the influence of rainfall erosion on the surface micro-terrain,and the semi-variance function method is the most excellent method.It is important to study the future trend and change period of runoff and sediment sedimentation by R/S analysis and Morlet wavelet analysis respectively,and it is very important to study the effect of continuous rainfall on soil and water conservation.