摘要
基于全国1996-2005年期间耕地面积与相关的社会经济数据为基础资料,分析研究期间耕地变化的过程,探讨耕地与社会经济因子之间的关系,模拟预测未来的耕地面积。研究结果表明在研究期内,耕地面积持续减少,尤其是以生态退耕为主;通过耕地动态度与耕地相对变化率的空间化,发现耕地数量变化与程度变化存在空间差异;通过相关性分析获得耕地变化的社会经济驱动因子,采用三层BP神经网络建立了耕地面积与相关社会经济因子的模型,模拟平均误差小;预测2010年、2015年耕地面积分别为11 195.82×10~4hm^2、9 889.30×10~4 hm^2,根据耕地的预测结果,可以判断现有的发展模式是否利于确保耕地红线,对促进我国社会经济发展有重要作用。
Cultivated land is the most important land use type, which shoulders the food safeguard for the human. This paper analyzes the process of cultivated land change in the research time, discusses the relationship between the cultivated land area and social economic factors, simulates and forecasts the cultivated land area in the future basing on the cultivated land areas and social economic datum in 1996 - 2005. The theories of land dynamic degree and land relative change ratio and BP neural network are used for illustrating the spatial difference and cultivated land simulating and forecasting separately with GIS and BP neural network methods. The results are as follows： （ 1 ） the area of cultivated land is decreasing continuously, which results from construction occupation, agriculture structure adjustment, ecologic restoring and disaster destroying, especially the ecologic restoring by 686.16×10^4 hm^2; （2） there is spatial difference in the quantitative change and extent change of cultivated land through the spacing the cultivated land dynamic degree and relative change ratio; （3） with the main influential social economic factors, the model about the relationship between cultivated land area and social economic factors is constructed using three BP networks and the simulating average errors are small; （4） the cultivated land areas in 2010 and 2015 are forecasted as 11 195.82×10^4 hm^2 and 9 889.30×10^4 hm^2, respectively. According to the forecasting results, if China keeps on the present social economic development model in the future, the protection aim for cultivated land could not be realized.
出处
《干旱区地理》
CSCD
北大核心
2008年第5期765-771,共7页
Arid Land Geography
基金
中国地质大学(北京)校内基金"土地规划环境影响评价"(51900961049)
国土资源大调查研究项目"土地资源调查
监测数据综合分析"
关键词
耕地
模拟
地理信息系统
BP神经网络
cultivated land
simulation
driving forces
BP neural network
