基于Logistic回归模型的新市区耕地数量变化景观立地要素驱动力分析  

Driving Forces about Xinshi Distric Landscape Elements of Farmland Quantity Changes Based on Logistic Regression Model

作　　者：马翼翔 吴兆福[1] 王兆[1] 

机构地区：[1]合肥工业大学土木与水利工程学院,合肥230009

出　　处：《湖北农业科学》2016年第11期2987-2992,共6页Hubei Agricultural Sciences

基　　金：水利部公益性行业科研专项经费项目(201401063-02);安徽省国土资源科技项目(2014-K-13);河北省社科基金项目(HB10EYJ172)

摘　　要：以保定市新市区为研究区域,采用统计学的研究方法进行单因素相关性分析,利用流失耕地与未流失耕地的累积概率曲线和不同范围内的流失耕地比例来分析耕地数量变化与自变量之间的相关性,以找出对耕地数量变化有贡献的因素。最后运用Logistic回归模型对耕地数量变化状况进行回归分析,定量化研究导致变化发生的景观立地要素驱动力因子,并对耕地变化的风险进行了预测分析。研究表明,耕地流失过程主要受到城镇、主要道路和主要河流三个景观立地要素的影响,离城镇用地和主要道路的距离越近,耕地流失的风险越高,而离主要河流越近,耕地流失的风险越低;利用回归结果进行耕地变化风险预测,能够模拟未来耕地数量和土壤质量变化情况,为耕地保护提供科学指导和依据。Taking Xinshi District as the study area,integrated index method was applied to single factor correlation analysis,Logistic regression model was employed to quantitatively analyze the driving force about landscape elements of farmland quantity and soil quality changes,and evaluate the risk of farmland change. Research show that three landscape elements,including of urban,main road and rivers, affected farmland loss. The shorter distance to urban land, mining land and main road was,the higher loss risk was, while the effect of unutilized land was opposite;Risk of farmland change predicted by Logistic regression model could simulate quantity changes of farmlands in the future,which will provide advises for farmland protection.

关 键 词：LOGISTIC回归 景观立地要素 耕地数量变化 新市区 

分 类 号：F323.211[经济管理—产业经济]