Research On Forest Land Change Based On GIS And RS

This study mainly through the combination of RS and GIS methods to study the change of forest land in Tongzhou of Nantong in eight years, the purpose is to provide reference for the sustainable development of forestry in the region. Using ENVI and ArcGIS to extract the land use information of the three phases of the study area. Using the dynamic degree model of land use,the relative change rate of land use,the land use index and the change model,and the transfer rate matrix to analyze the dynamic change of land use types, so as to get the quantity changes of the forest land. Using the Lorenz curve to analyze the spatial distribution change of forest land area in the district. Using the Markov prediction model to predict land use pattern in 2014, then compared with the results of the classification of 2014. Correlation analysis and principal component analysis were carried out to analyze the driving factors affecting the change of forest land, and then the driving factors of forest land change were determined. Finally, according to the current situation of forest resources in Tongzhou, some suggestions are put forward for the sustainable development of forestry, which provides reference for the development of Tongzhou forestry. The main contents are as follows:(1) Using ENVI5.1 to realize the radiation correction and geometric correction of three periods of remote sensing image. Cut out the remote sensing image of study area based on the vector boundary of it. Using Gram-Schmidt fusion method to fusion image. Calculating vegetation index, water index of remote sensing image. Through the principal component analysis, the first principal component with the largest amount of information is extracted. Extracting texture information of the first principal component. Combine all bands of the fused image, vegetation index, water index and texture information in order to create new remote sensing image. Using support vector machine classification method to classify remote sensing images. Using Majority/Minority method and artificial visual interpretation processing the classified image. The kappa coefficients of the obtained classification results are 0.87, 0.88 and 0.77. Combine with ArcGIS10.2 to analyze and statistic the classification results. Extraction the land use information in the three periods of the study area.(2) According to the quantitative analysis of land use information, from 2006 to 2014 in Tongzhou, the area of forest land and construction land continued to increase, while waters, cultivated land and unused land showed a decreasing trend. The amplitude of variation of the area of various kinds of land is: cultivated land > construction land > forest land > water area > unused land. Among them, the area of forest land increased by 10166.93 hm2. Using the dynamic degree model of land use to analysis dynamic change of land use, the results showed that, the change of forest land is the most intense during the whole research period, and its growth rate in the second half of the study area is about two times of that in the first half of the study period. The dynamic degree of comprehensive land use in the first half of the study period was 0.94%. In the second half of the study period, the dynamic degree of comprehensive land use was 1.65%, the change of the land use type is more violent. Using the transfer rate matrix to analyze the transfer of the area between land types. The results shows that, in the second half of the study period, forest land transfer rate in the turn out side slightly reduced compared to the first half of the study period. But, forest land transfer rate in the turn in side greatly improved in the second half of the study period. The main land type which the forest land transfer out and transfer in is cultivated land. Analyzing the change of the spatial distribution of forest land by using the Lorenz curve, the study area's distribution of forest land gradually tends to be balanced from the year of 2006 to 2014. Variation range of spatial distribution of forest land in the first half of the study period was larger than that in the second half of the study period.(3)Analysis by township, the differences of the land use information is obvious between different towns. In the first half of the study period, increase quantity of forest land in Sanyu is the biggest, it has inceased 864.71 hm2. In the second half of the study period, the area of forest land of Jinsha increased the most, it has inceased 930.45 hm2. In the whole research period, Jinsha, which located in the center of the district, has the largest increased in the area of forest land, inceased 1383.34 hm2. By analyzing the relative change rate, during the first half of the study period, Dongshe, Xiting and Sanyu's relative change rate of forest land was relatively large. In the second half of the study period, the larger relative change rate of forest land was Dongshe, Erjia, and Wujie, the values were greater than 1. The changes in the area of forest land in these towns are larger than that in the whole area. By analyzing the land use degree index, the land use degree index of most towns is increasing, it is becanse a large number of lands which has a lower degree of land use classification index transfer to construction land which has a higher degree of land use classification index. Land use degree index of Liuqiao decreased slightly, it has reduced 0.01. Binhai's land use degree index is the largest, and Wu Jie's is the smallest. Most towns which has faster increasing of the area of forest land, their land use degree index change rate is lower.(4)The land use transfer matrix of 2006 to 2010 was used as a template to predict the land use pattern in 2014 by using Markov forecasting model. Compared with the prediction result and the classification result of 2014, the result shows that, the water area and unused land of the prediction result are more accurate. Compared the percentage of the predicted result and the actual result, there are only have differences of-0.06% and 0.02%. The forecast result of construction land is bigger than the classification result. The prediction results of the area of forest land and construction land are smaller than the classification result. This is related to the speed of construction of Tongzhou in four years, the implementation of forestry policy and other uncertain factors.(5)Analyze the driving forces of forest land change by correlation analysis and principal component analysis. The result shows that, two aspects of population and social economic construction were significantly related to the change of the area of forest land.The key drivers include average income and expenditure of urban residents, average income and expenditure of rural residents, GDP, primary industry, secondary industry, tertiary industry, total imports and exports, total retail sales of social consumer goods, gross industrial output value, fixed assets investment, fiscal revenue, rural investment, building industry and population.