基于Logistic回归模型的城市建设用地扩张影响因素分析——以深圳市为例

Analysis of the Influencing Factors of Urban Sprawl in Shenzhen City Based on a Logistic Regression Model

分析城市建设用地变化影响因素对于探究城市内在演化规律具有十分重要的意义。本文以深圳市为例，通过建立2002-2007年和2007-2012年两个阶段建设用地变化的Logistic回归模型，从空间角度分析了建设用地变化的规律及原因。研究结果表明，在第一阶段（2002-2007年），深圳市处于快速城市化时期，建设用地变化以新增建设用地为主，主要的解释变量是固定资产投资变化，第二产业产值变化，坡度和高程，而在第二阶段（2007-2012年），深圳市进入高度城市化阶段，建设用地变化表现为存量建设用地空间布局的变化，主要解释变量是固定资产投资变化，第二产业产值变化，人均GDP变化和人口密度变化。

For the research of the urban internal evolution rules, it is of great significance to analyze the affecting factors of the construction land change. This paper, taking Shenzhen city as an example, builds logistic regression models on the change of con-struction land in the first period （ 2002~2007） and the second period （2007~2012） respectively, and then analyzes the rules and causes of the change of construction land. The result indicates that in the first period （2002~2007） when Shenzhen was increas-ingly urbanized, the increase of construction land predominated in the change of construction land. The main explanatory variables are“change of the investment in fixed assets”,“change of the sec-ond industry output value”,“slope” and“elevation”, while in the second period （2007~2012） when Shenzhen was highly urbanized, the change of remained construction land spatial arrangement predominated in the change of construction. The main explanatory variables are“change of the investment in fixed assets”,“change of the second industry output value”, “GDP per capita” and“population density”.