Spatial Patterns and Influencing Factors of Hong Kong and Macao Vehicle Flows in Guangdong Province - estudo
jul26
Spatial Patterns and Influencing Factors of Hong Kong and Macao Vehicle Flows in Guangdong Province
投稿的翻译标题: 港澳车辆在粤流动的空间格局与影响因素研究- Mingrui Huang
- , Qitao Wu
- , Yuanjun Li
- , Kaihuai Liao
- , Jiaxuan Chen
- , Chao Cui
- Guangzhou Institute of Geography
- Guangdong University of Technology
科研成果: 期刊稿件 › 文章 › 同行评审
Cross-border transportation integration is a critical foundation for implementing the Guangdong-Hong Kong-Macao Greater Bay Area (GBA) strategy and promoting regional functional integration. To address limitations in existing research—particularly the insufficient characterization of cross-border vehicle flows and the limited interpretability of complex driving mechanisms—this study utilizes a comprehensive dataset comprising 623,133 valid origin-destination (O-D) records of Hong Kong- and Macao-licensed vehicles obtained from Guangdong's expressway toll system in September 2023. By integrating network analysis, kriging interpolation, and an Optuna-optimized XGBoost model combined with SHAP (Shapley Additive Explanations), this study systematically explores the spatial flow patterns and multidimensional driving mechanisms underlying these vehicle movements. The results indicate the following: 1. Spatial Distribution and Network Structure The spatial distribution of Hong Kong and Macao vehicles exhibits a pronounced "core-periphery" structure characterized by strong polarization. The Pearl River Delta (PRD) functions as a high-intensity agglomeration core, whereas Eastern, Western, and Northern Guangdong constitute peripheral zones. Cities such as Shenzhen, Dongguan, and Zhuhai have emerged as key hubs with the highest betweenness centrality, effectively linking the PRD with non-core regions. 2. Driving Mechanisms and SHAP Analysis SHAP dependency analysis reveals significant nonlinear threshold effects and spatial heterogeneity among the driving factors. Spatiotemporal Distance Effects Spatiotemporal distance functions as a universal inhibitory factor, with a critical threshold occurring at approximately 130 minutes (or 110 km). Within this threshold, flow intensity declines sharply as distance increases. Beyond this point, the marginal inhibitory effect of distance gradually weakens, producing a "long-tail" effect for medium-to-long-distance travel. Promoting Factors The number of residents with Hong Kong, Macao, or overseas backgrounds five years earlier represents the primary promoting factor across the province, exhibiting a distinct "spatial gate effect," whereby positive influence increases exponentially only after surpassing a certain concentration threshold. Regional Variations Flows within the PRD are predominantly driven by industrial clusters. Western Guangdong is primarily influenced by scenic tourism resources. Northern and Eastern Guangdong are more strongly shaped by per capita GDP and ancestral social networks. 3. Interaction Analysis Interaction analysis reveals differentiated coupling pathways: The PRD relies on the synergistic effects of industrial and economic drivers. Non-PRD regions are primarily influenced by the combined effects of tourism resources and commercial facilities. This study contributes an interpretable machine-learning framework that effectively decodes the "black box" of complex spatial interactions underlying cross-border mobility flows. From a practical perspective, the identification of a 130 minutes non-linear threshold provides a quantitative basis for optimizing transportation planning within the GBA. Accordingly, the implementation of "cross-border business commuter lanes" within the 130 minutes core zone and the enhancement of "ancestral tourism corridors" in peripheral regions are recommended to mitigate distance-related friction and support the GBA's transition from "physical connectivity" to "functional integration".
https://gdut.cnpereading.com/zh/publications/spatial-patterns-and-influencing-factors-of-hong-kong-and-macao-v/
Comments
Post a Comment