| Laws of water erosion during construction period in a section of subgrade slope at Zhun-Shuo Railway located in the Loess Plateau were comprehensively and systematically studied in this dissertation and the main work was as follows:(1) A roadbed which located in MaZha station (ZhunShuo Railway DK148, connecting the ChenJialiang bridge) was selected as the sample test region in this paper. And the method of field survey on erosion gully was used for study.The catchment area was measured and calculated during the observation period and the amount of soil erosion was55708.09t/km2,which is belong to the severe erosion.(2) The total rainfall erosivity is137.26(m-t-cm/hm2-h) according to the formula R=2.35(P·I30/100)-0.523.(3) The relationship between rainfall, I10,I30, PI30,R and the amount of soil erosion was fitted respectively. The results revealed that rainfall and the amount of soil erosion was the most relevant, and the fitted equation is:y=0.0014x2+0.3268x3.1348, with the relevance factor R2=0.9922.(4) Several common forecasting method of soil erosion were introduced and the USLE equation was used for prediction based on the analysis of various factors contributing to the result W=R·K·L·S·C·P=58644.91t/km2, which is similar to the observed value55708.09t/km2with the relative error5.3%. Hence it is appropriate to use the USLE equation in the prediction of soil erosion for the section of railway subgrade slope. In addition, the statistical software SPSS was used in this paper and the soil erosion prediction model for the railroad subgrade slope was deduced as follows:Y=221.611X1+1676.033X2+13635.068X3-4808.67,R=0.991.X1, X2, X3denotes the rainfall, the average rainfall intensity, and rainfall intensity of maximum30min, respectively, and Y is the erosion. Relevance of the regression equation is extremely significant, so the model can be applied to the prediction of subgrade slope (θ=38.7°)soil erosion in the experimental zone depending on the similar rainfall condition. |