基于ESDA-GWR的浙江省土地城镇化空间特征及影响因素分析

摘要

The purpose of this study is to explore the spatial characteristics and influence factors of land urbanization, taking 76 counties in Zhejiang Province as the study areas. The methods used include exploratory spatial data analysis and geographically weighted regression model. The main results could be summarized as follows: 1)It is obvious that the land urbanization level is different in Zhejing Province with certain positive spatial correlation. Land urbanization of Zhejiang Province has“spot-corridor-patch” structure characteristics. Hangzhou City, Ningbo City and Wenzhou City, are the high level “spots”; and the southwest of Zhejiang Province is the low level “corridor”; and the counties belong to Shaoxing City, Ningbo City and Taizhou City are the low level “patch”. 2)The influence factors present spatial heterogeneity. The domestic factor mainly shows a decreasing trend from north part to south part while the foreign factor increasing from northwest part to southeast part. 3)GWR model is more suitable for analyzing influence factors of land urbanization than OLS model.%研究目的：以浙江省76个县级单元为例，探索土地城镇化水平的空间分布特征及其影响因素，为进一步揭示土地城镇化驱动机制提供参考。研究方法：探索性空间数据分析（ESDA）和地理加权回归（GWR）模型。研究结果：（1）浙江省土地城镇化具有“点—带—块”状结构特征。土地城镇化水平较高的杭州主城区、宁波主城区、温州市区等是土地城镇化水平较高的集聚“点”；浙西南区域的衢州和丽水所辖县市形成了土地城镇化水平较低的分布“带”；绍兴、宁波和台州所辖县市形成了土地城镇化水平低值“块”。（2）由社会经济指标归纳的国内动力和国外动力与土地城镇化水平都呈正相关关系，国内动力的回归系数在空间上主要是从北向南呈递减趋势，而国外动力的回归系数由西北向东南呈递增的趋势。研究结论：（1）浙江省土地城镇化水平空间差异明显，具有一定的正的空间自相关性；（2）浙江省土地城镇化影响因子具有明显的空间分异性；（3）GWR模型相比于OLS模型更适用于土地城镇化水平影响因素分析。