三峡库区耕地变化研究

Research of Changes in Cultivated Land Around the Reservoir Area of Three Gorges

三峡工程是世界性的伟大工程,三峡工程的兴建推动了三峡库区的经济发展。在三峡库区经济发展的过程中,耕地成为经济发展的重要资源。本文利用三峡库区1990-2004年的耕地面积和社会经济资料,采用统计分析与实证研究的方法,分析了三峡库区耕地变化的过程,了解库区耕地变化的驱动力,探寻了耕地数量模拟预测的新方法,为库区耕地的保护和经济可持续发展提供参考。结果表明:(1)三峡库区15年来耕地面积及人均耕地面积呈减少的趋势,不同的库区段也如此;(2)三峡库区历年耕地面积与对应年人均GDP的关系曲线呈倒“U”型,并且相关年份耕地的变化与GDP的变化不同步,呈现相对脱钩的状态,三峡库区耕地的变化与城市化的发展及退耕政策休戚相关;(3)在库区耕地模拟预测研究中表明,传统的多元线性回归分析方法的模拟精度不及BP神经网络的模拟精度,并利用BP神经网络预测了三峡库区2010年和2015年的耕地面积,其预测结果具有一定的合理性。

Three Gorges Project is the great project in the world, which accelerates economic development in the reservoir area of Three Gorges Project. In the process of development in the reservoir area of Three Gorges Project, cultivated land has become an important resource, and a lot of cultivated land has been occupied and become constructing land. At the same time, a lot of cultivated land has been flooded because of the rising of the water level. This paper uses the cultivated land areas and social economic indicators of reservoir area of Three Gorges for 1990- 2004, takes the statistic analyses and example research in order to analyze the process of cultivated land, gets the driving forces of cultivated land change, finds the new methods to stimulate and forecast the cultivated land areas in the future, and serves for the cultivated land protection and sustainable development in reservoir area of Three Gorges. The results indicate that, firstly, in the past 15 years, the cultivated land area has decreased by 200 142 hm^2, and the decreasing quantity per year is 13343 hm^2. The whole reservoir area is divided into three different areas： upper reach area, belly area and lower reach area. The trends of cultivated land change in different reservoir areas are similar to the whole reservoir area. Secondly, the curve of cultivated land areas and per capita GDP takes on a reverse U, and the paces between the change rate of cultivated land and the change rate of GDP are different in some years, which indicates that change of cultivated land and change of GDP is decoupled; besides that, change of cultivated land is in connection with the development of urbanization and the policy of returning forestry greatly. Lastly, the precision of multi-regression is lower than the BP neural network in the simulation of cultivated land. We then use the BP neural network to forecast the cultivated land areas for 2005, 2010 and 2015, and the forecasting results are reasonable.