China is one of the large coal mining countries in the world. Coal mining accelerates economic prosperity, as well as engenders a series of environment problems either. One of the most obvious problems is that coal mining changes the landforms around the mining areas. Abundant arable area, garden area, forest area and construction area have been changed under the drive of this dynamic landform. The law that other environment elements change resulting from transformation of one element can be analysed by location theory---Concentric-Circle Mode (or Circle Layer Mode) proposed by professor E. W. Burgess of Chicago University. For the case of Longkou coal mining subsidence area in Shandong province, based on the ground measurement elevation data of the years of 1978, 1989, 1995 and 2004, firstly, this paper considers the DEM data of 1978 before subsidence as standard elevation, and calculates the difference value DEM data of three periods through the difference operation of the other later three-period DEM data and the standard elevation. The coal mining subsidence region and area can be figured out, which is grid region and the overall sum of the grid area with z<0. Secondly, by choosing the digital remote sensing images which are the same period with the later three-period DEM data, with operation of the classifier of BP Artificial Nerve Network (BPNN), the author classifies these images by combining spectral information, texture information of remote sensing image with terrain index. Thirdly, under the guidance of location theory, the author uses location index to make "location image". Lastly, with spatial superposition of location image, three-period DEM data and land use classification result, the author figures out the area and proportion of all the land use types in different locations and the transfer matrix of land use types, and analyses the rule of space-time change of land use in different locations, in order to explain the location effect that coal mining subsidence affect land-use change.