基于灰色关联分析和遥感信息模型的新疆典型农区非耕地系数研究  

Influence Factors of Non-cultivated Land Coefficients in Typical Agriculture Area in Xinjiang based on Grey Correlation Analysis and Remote Sensing Information Modelling

作　　者：冯雪力[1,2] 吴世新[1] 陈红[1,2] 

机构地区：[1]中国科学院新疆生态与地理研究所,新疆乌鲁木齐830011 [2]中国科学院研究生院,北京100049

出　　处：《中国沙漠》2011年第1期156-161,共6页Journal of Desert Research

基　　金：中国科学院知识创新重大项目(KZCX1-YW-09-10)资助

摘　　要：非耕地系数反映了人类活动对耕地的影响状况。以2008年QuickBird数据为主要数据源,以伊宁县农耕区为研究区,采用灰色关联分析法研究与非耕地系数有关的因素,并利用相似性准则、量纲分析及多元回归方法建立遥感信息模型,构建单位面积内非耕地数量与其相关因素之间的数学关系。研究结果表明,在地形、土壤类型一定时,影响非耕地系数的主要因素有农区土地利用的斑块破碎度、水网密度、土地的垦殖指数等。遥感信息模型表明斑块破碎度、水网密度、人均生产率、农业人口密度与单位面积内非耕地数量呈正比,且它们对单位面积内非耕地数量的影响由强到弱依次是水网密度、农业人口密度、土地利用的斑块破碎度。耕地生产率、人均耕地与单位面积内非耕地数量呈反比。该模型反映人类活动对农区耕地的干扰状况,对土地利用和整理具有一定的指导意义。Non-cultivated land coefficient is a key factor in monitoring of cultivated land area by use of remote sensing technology on a large spatial scale.Taking a typical agriculture area of Yining county in Xinjiang as a study area,30 sampling quadrats are selected on the QuickBird images in 2008 to measure the non-cultivated land coefficients.Grey correlation analysis and remote sensing information modelling are used to study the relationship between non-cultivated land coefficient and influence factors.Grey correlation analysis shows that the main influence factors of non-cultivated land coefficient are land use fragmentation,water network density and land cultivation index.The remote sensing information modeling shows that non-cultivated land area is negatively correlated to land use fragmentation,water network density,per capita productivity,and agricultural population density,and is positively correlated to cultivated land productivity and per capita cultivated land area.The model reflects the effect of human activities on land use,so is useful in guiding land use and land management.

关 键 词：非耕地系数 灰色关联分析 遥感信息模型 地理相似准则 

分 类 号：F301.24[经济管理—产业经济]