摘要
北方农牧交错带作为我国传统农业区与畜牧业区交汇和过渡的地带,其土地集约利用有突出的特点。以内蒙古鄂尔多斯市为例,从北方农牧交错带土地集约利用的内涵出发,构建评价指标体系,分别运用熵值法和人工神经网络模型计算集约度。结果显示,2003-2007年鄂尔多斯土地集约度数值总体上升;利用效率对鄂尔多斯土地集约利用影响最大,其次是利用程度和投入强度,持续状况的影响最小。结果表明,熵值法和人工神经网络模型均适用于土地复杂系统的集约利用评价计算。相对于其它方法,熵值法在一定程度上避免了赋权的片面性和主观性,人工神经网络模型弥补了传统的多指标综合评价方法主观性强、缺乏自学能力的缺陷。
As the transition zone of traditional agricultural area and animal husbandry of China,agro-pastoral transitional zone is prominent in land use intensity.Taking Ordos City in Inner Mongolia as an example,this paper made a calculation of land use intensity of Ordos.On the basis of intensive land use,an index system was built to represent the status of land use intensity,which included 18 indexes in index layer like urban construction land per capita,population density of built up area,proportion of real estate development and so on.Neuron structure of BP network and entropy method was used to calculate the intensity value,and both of them got a similar conclusion.The result showed that,the intensity represented a rising trend during 2003-2007,which showed the positive impact of the strategy of industrialization,urbanization,industrialization of agriculture and animal husbandry especially after the eleventh five-year plan.Secondly,for the indexes of target layer,utilization efficiency had the most significant influence on land use intensity in Ordos.Utilization degree took the second place with input intensity following,and sustainable development played the least important role.Except for the year by year decreasing of utilization degree,other indexes of target layer increased year by year.The reverse change made a direct influence of the reinforcement of intensive land use in Ordos.For the indexes of index layer,most indexes got a remarkable ascendant trend.The significant fluctuation of total retail sales of consumer goods per unit indicated the importance of land market perfection and land management reinforcement.The decreasing of population density of built up area revealed that the redevelopment of commercial housing and the efficiently using of available land was effective measure for the intensification of the intensive land use.The decreasing of the proportion of agricultural labor reflected that the reduction of labor input restricted the further development of intensive land use in Ordos.Besides,the result proved that,both entropy method and ANN model were suitable for the intensive land use research in the complex system.Comparing to other methods,entropy method prevented the process of weighing calculation from one-sidedness and subjectivity.ANN model can make up the deficiency of the traditional multiple-indicators comprehensive evaluation like strong subjective and self-study lacking.The evaluation effect of the two methods depended on the structure of index system to some extent.The index system should be adjusted according to the subject investigated in the matter of space,and should also be adjusted according to the stage of urban development in the matter of time.What's more,the implement of GIS and RS can be helpful for index system amendment.
出处
《干旱区地理》
CSCD
北大核心
2011年第6期1017-1023,共7页
Arid Land Geography
基金
国家自然科学基金项目(40871048)
国家科技支撑计划课题(2006BAD20B07)资助
关键词
农牧交错带
土地集约利用
熵值法
人工神经网络
agro-pastoral transitional zone
land use intensity
entropy method
ANN model