Applicability Of Probability Integral Method In Prediction Of Coal Subsidence In The Loess Gentle Slope Hill

85% of China's coal resources are from underground mining.Subsidence of mined land,resulted from underground mining,will inevitably leads to the collapse of roads and houses,declining of crop production.Furthermore,people's lives and property are endangered.Additionally,coal mines are intensively distributed in Shanxi,Shan'xi and Inner Mongolia where ecology is fragile and is easily to be destroyed by mining.Therefore,it is urgent to study the characteristics of collapse and the accurate prediction of collapse to provide the basis suggestion for reclamation of mining areas in the Loess Plateau.Based on the morphological characteristics of collapse,mining subsidence theory,we selected probability integral method,a popular prediction method for ground subsidence in plain area,China,to predict the collapse of 903 working face,Pingshuo 3rd underground mine.Besides,a the subsidence with an area of 100×100 m plots were investigated with quantitative and qualitative measurement.The subsidence property values,namely the subsidence area,depth,crack number,were analyzed by spatial autocorrelation methods to reveal the characterizes of the subsidence.We also compared the subsidence survey values and values calculating from probability integral method to determine the applicability of this method in the hilly area of loess slope.The main conclusions and viewpoints of this paper are as follows:(1)To predict the collapse of the 903 working face,MASA system was applied.It is concluded that the maximum subsidence area is located at the center of the 903 working face with a range of 1-9m,and the maximum subsidence value is 9269 mm.The collapse interpreted from DEM of measured region showed that collapse degree were serious gradually from east to west within 100-meter scope and the subsidence depth was 1-3.5m.(2)The ground surface subsidence and ground fissures were the main modes that destroyed the mined land.The subsidence area of the working face of the 903 coal mine covers an area of 220.02 hm2,among which 200.18 hm2 of farmland and 19.82 hm2 of forest were destroyed.Field measurement of subsidence area showed the subsidence area and crack number were in uniform distribution,but not the depths.We also found that the central and southern of our study area were in serious subsidence,while northwest collapse was not serious,which could be due to mining and geological conditions.(3)The subsidence value of subsidence area is compared with the actual value,and the normalized root mean square error was calculated.The normalized root mean square error value is 2.18 times of the standard deviation.According to the definition of normalized root mean square error,it is shown that the error between the actual subsidence value and the predicted subsidence value is larger,which could be due to geological conditions and survey timing.