Abstract:

This paper takes 30 provincial capitals as research objects and integrating the transport superiority degree index system from the big data level such as journey time, cost, train, running distance and speed with the traditional transport superiority degree system, using the entropy method to calculate the transport superiority degree of the provincial capitals in 2019 and explore its impact factors using GeoDetector and Geographically Weighted Regression models. The conclusions are as follows: (1) Significant spatial differences in transport superiority degree. The circle layer central cities of various levels of transport superiority degree show an uneven distribution pattern of space based on their railway connection intensity. At the same time, it presents a core- half periphery structure that decreases from the east and middle nodes along with the railway network and forms a traffic hierarchy network system with multi-level node city interaction. (2) Local general public budget expenditures and the number of the high-speed railway are the dominant detection factors of transport superiority degree, multi-factor interactions have significantly improved the interpretation of their spatial differentiation. (3) Except for the average slope and average elevation, the influence degree of each factor has greater spatial heterogeneity and all are positively affected; disposable income per capita, average elevation, and regression coefficients of ordinary college students show a trend of transition from the west to the east and the growth to the north; The regression coefficients of the average slope, number of the motor car, and number of ordinary trains gradually decrease from west to east; The number of train stations and the number of high-speed railways has a greater impact on the eastern, western and northern regions, respectively.

Key words: transport superiority degree, GeoDetector, Geographically Weighted Regression, spatial heterogeneity, provincial capital cities

摘要：

以30个省会城市为研究对象，将旅程时间、费用、列车、运行里程、速度等大数据融入交通优势度指标体系，采用熵值法计算2019年省会城市交通优势度并利用地理探测器和地理加权回归模型探究其影响因素。研究发现：（1）交通优势度空间差异显著，交通优势度各等级圈层中心城市在空间中呈现不均衡分布格局，呈现由华中、华东地区节点沿铁路网向外递减的“核心-半边缘”结构，形成多等级节点城市交互作用的交通等级网络体系；（2）地方一般公共预算支出、高铁数目为交通优势度的主导探测因子，多因子交互作用均使其空间分异的解释力显著提升；（3）各因素影响程度具有较大空间异质性，除平均坡度和平均高程外其他因素均为正向影响，人均可支配收入、平均高程和普通本专科学生的回归系数呈现由华西地区向华东地区过渡转而向北增长的趋势，平均坡度、动车数目、普通列车数目的回归系数由西向东分层递减，火车站数目和高铁数目分别对华东地区、华西地区和华北地区的影响较大。

关键词: 交通优势度, 地理探测器, 地理加权回归, 空间分异, 省会城市