Terrain–Infrastructure Interactions and Modeled Freight Trip Attraction in a Mountainous City: Evidence from Chongqing, China
Mountainous terrain fragments urban land, constrains network design, and raises the cost of linking firms across supply chains. This study examines how terrain and the built environment are associated with modeled freight trip attraction (FTA) in Chongqing, China. A route-level dataset for large and medium-sized enterprises is combined with measures of water and mountain obstacles, elevation variation, road and building density, and industrial-park configuration. XGBoost models and SHAP values are used to identify nonlinear relationships and interactions. The results show that water barriers, mountain obstacles, and elevation variation are generally associated with lower FTA. Road and building density can moderate these constraints, but the relationships are threshold-dependent and show diminishing or reversing contributions at high levels. Industrial-park configuration also matters: the same infrastructure characteristic has different associations depending on whether origin and destination firms are located inside or outside a park. Sector-specific models further indicate that agriculture-related chains respond more strongly to improved connectivity, whereas coal and other extractive activities remain sensitive to elevation. Our findings support terrain-sensitive land-use and freight planning that coordinates industrial location, road investment, and cross-barrier connections rather than relying on uniform infrastructure expansion.
Authors
- Chang Shen (ORCID: https://orcid.org/0000-0001-8540-2268)
- Jianping Gu (ORCID: https://orcid.org/0000-0002-1385-3112)
- Qing Lu (ORCID: https://orcid.org/0000-0003-2206-899X)
- Zhangjian Ding
Institutions
- Chongqing University (CN)
- Chinese Research Academy of Environmental Sciences (CN)
Publication Details
- Journal
- Land
- Published
- 2026-09-24
- DOI
- https://doi.org/10.3390/land15101796
- Primary Topic
- Urban and Freight Transport Logistics
- Type
- article
- Field-Weighted Citation Impact
- 0.00