With 0800 h convenient data and numerical products as input,the element field of a key region is expanded as the Chebyshev orthogonal polynomial,whose coefficients are taken as predictors to be trained with the aid of the B-P neural network,whereupon is established an arti- ficial rainfall fallout region forecasting network for categories 5,10 and 25 mm for 24 h in ad- vance.

At present, the USLE was usually used to predict the loss of soil and water along railway in the evaluation of the influences on the ecological environment due to railway construction, and there had been several regional prediction equations in our country.