Study On Landscape Pattern Evolution And Simulation In Pei Bei Mining Area Based On Multi Source Remote Sensing

This study With the remote sensing technology, multi-scale effects of landscape pattern in Pei Bei mining area have been studied. Synergistic relationships between landscape pattern evolution and coal resources exploitation in mining area have been analyzed. We discuss the relation between coal mining intensity and destruction of cultivated land. The landscape evolutions in different conditions have been simulated in this study. Research results can be found as follows:(1) Based on random forests classification method, Landscapes in the North Pei mining area are divided into following types on the IDL platform: paddy field, dry land, woodland, garden plot, traffic land, waters, construction land, industrial land and unused land. The waters is divided into the natural waters and the collapsed waters based on the land surface collapsing information gained from D-InSar technology. The total classification accuracies of remote sensing images are above 91%, and the Kappa accuracy is above 0.90.(2) The landscape revolution in mining area is discussed based on the model of Pattern – Process – Scale. And the multi-scale effect of landscape pattern in mining area using Ripley-L and O-ring functions. The landscape pattern index is calculated by the characteristic scale of landscape type. While the mutation points are detected with Mann-Kendall method. Synergistic relationships between landscape pattern evolution and coal resources exploitation in mining area are studied by excluding natural fluctuations. Synergistic relationships between cultivated land evolution and coal resources exploitation in mining area are analyzed by SD model. The results indicated that the disturbance influence value of coal mining achieve the maximum in 2006 while landscape pattern is recovered in 2010. The best index range for coal mining are between 0.8 and 1.2.(3) The inversions of NPP value of crops and soil moisture in mining area are carried out based on method of treatment of active and passive remote sensing. A ANN-CLUE-S model is built to simulated the landscape pattern in mining area under different scenarios in 2025. The simulation results show that under the scenario of natural development, building land growth is much slower, while subsidence waters area decreased and damage of cultivated land situation eased which mean that the majority of mine come into the stable production period and economy development are rather steady in mining area. In the context of sustainable development, the main activities aim to protect land resources instead of economy development. This leads to the significant recovery for landscape. On the contrary, economic situation in the mining area have greatly restricted. In the context of economic development, the mining construction land increased significantly while subsidence waters area increase rapidly and conditions of regional cultivated land become worst. Therefore, in the current coal mines to stable production stage land management pattern, both economic development and ecological protection can be considered in mining area land effectively.