| With the study of coastal zone environmental changes in Longkou city on the main line, the paper, based on 9 periods of remote sensing images and other relevant data, analyzes the RS image-geography mapping and takes proper image processing techniques to enhance the thematic information of the coastal zone. On base of the spectral response characteristics of land, sea and transitional area in RS image, it separates the reach area into sea area, land area and coastal area, then according to their own geographical trait convert image form to geographical form using the optimal image classification technology. It was difficult to be drawn out the low tide line directly for being submerged by sea long time, so the paper calculates the low tide line through creating the RS-fathoming model with the land DEM data and the tide data. For abstracting the beach own geographical trait more accurately, besides of taking full advantage of multispectral data, it does K-T transformation and vegetation index abstraction to RS images and extracts the beach information by way of decision trees classification technology. Based on the unique geographic environment of the land area in coastal zone, it combines the information of spectrum, special texture and geographical auxiliary to extract the land use and land cover information by the way of BP Neural Network classification technology. The paper does a series of analysis on combination and transformation for the spectral data, the RS image DN data and the groundwater observation in the sea-water intrusion area and finds the relationship among them. Then it establishes the inverse calculation model for sea-water intrusion based on the spectral recognition mode and makes an inversion of its historical evolution. The paper, based on the DTM of 1978, obtains the topographic variation of 1984,1996,2000 and 2004 by way of minus operation image processing and calculates the direction of surface runoff in flow vector matrix on grid to separate the coal-mine subsidence and the sand excavation subsidence, then establishes the groundwater resource mathematical model and determines the subsidence times and the submerged area according to the hydrological characteristic. Based on the result, supported by GIS, the paper unites acquired environmental information and establishes the Geo-Information Tupu of Longkou coastal zone changes to 2004, then reveals its environmental change regularities of time and space, and researches the driving mechanism in natural and human points of view. |