Pub Date : 2026-02-01DOI: 10.1016/j.jtrangeo.2026.104569
Julia Cardwell , Paul L. Delamater , Charles E. Konrad
Structural road network redundancy contributes to reliable mobility by providing alternative routes in caseof disruptions. Existing structural redundancy metrics tend to emphasize the criticality of links (infrastructure-focused) or the pairwise redundancy of origin-destination (OD) connections under equal weighting (demand-agnostic), overlooking place- and population-specific differences in redundancy. This study introduces an alternative population-based, demand-aware redundancy metric to evaluate structural redundancy at the census block group level by integrating origin-specific demand distributions derived from mobility data with a path-penalized Dijkstra's algorithm to evaluate redundancy separately for each census block group. This population-centered approach aligns with transportation justice principles and provides an accessible tool for assessing heterogeneities in local access to road network redundancy. Applied to North Carolina, the metric highlights geographic variation in redundancy across the state, especially considering urban-rural and regional divides, and identifies populations in western North Carolina as having comparatively lower access to redundancy. The findings of the case study underscore the necessity of considering road networks' unique structural characteristics in planning for equitable and resilient transportation systems.
{"title":"A population-based, demand-aware framework for measuring structural road network redundancy","authors":"Julia Cardwell , Paul L. Delamater , Charles E. Konrad","doi":"10.1016/j.jtrangeo.2026.104569","DOIUrl":"10.1016/j.jtrangeo.2026.104569","url":null,"abstract":"<div><div>Structural road network redundancy contributes to reliable mobility by providing alternative routes in caseof disruptions. Existing structural redundancy metrics tend to emphasize the criticality of links (infrastructure-focused) or the pairwise redundancy of origin-destination (O<img>D) connections under equal weighting (demand-agnostic), overlooking place- and population-specific differences in redundancy. This study introduces an alternative population-based, demand-aware redundancy metric to evaluate structural redundancy at the census block group level by integrating origin-specific demand distributions derived from mobility data with a path-penalized Dijkstra's algorithm to evaluate redundancy separately for each census block group. This population-centered approach aligns with transportation justice principles and provides an accessible tool for assessing heterogeneities in local access to road network redundancy. Applied to North Carolina, the metric highlights geographic variation in redundancy across the state, especially considering urban-rural and regional divides, and identifies populations in western North Carolina as having comparatively lower access to redundancy. The findings of the case study underscore the necessity of considering road networks' unique structural characteristics in planning for equitable and resilient transportation systems.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104569"},"PeriodicalIF":6.3,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146072723","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study examines how proximity to metro stations is associated with housing prices in Medellín, Colombia, and how this relationship varies across income groups and spatial scales. While the capitalization of transport infrastructure into real estate values is widely recognized, its social and spatial heterogeneity remains underexplored in Latin American cities. Using a dataset of georeferenced housing transactions and spatial information on urban amenities, we estimate hedonic pricing models—including Ordinary Least Squares, Spatial Autoregressive, and Spatial Error Models—with income-level interaction terms and stratified buffer thresholds optimized through an iterative procedure.
The results reveal that metro accessibility is not uniformly valued across the city. In low-income districts, proximity to a metro station is associated with a price premium of approximately 9.4%, whereas in high-income neighborhoods, it corresponds to a statistically significant price penalty of 9.8%. Moreover, the spatial reach of metro influence decreases with income level, ranging from 1.7 km in low-income areas to just 0.6 km in high-income zones. Among all urban amenities considered, metro stations exhibit the most pronounced variation in market valuation across socioeconomic groups.
These findings underscore the need for disaggregated, income-sensitive approaches in both spatial modeling and urban policy design. The buffer-based strategy proposed here offers a flexible and empirically grounded method to capture how accessibility is differentially perceived and valued in stratified urban contexts. In cities of the Global South—where inequality is spatially embedded—planning efforts must account for these disparities to ensure that transit investments promote inclusive urban development.
{"title":"Unequal benefits of metro station proximity: Income-level differences in housing price effects in Medellín","authors":"Hernán Darío Villada-Medina , Luisa Díez-Echavarría","doi":"10.1016/j.jtrangeo.2026.104571","DOIUrl":"10.1016/j.jtrangeo.2026.104571","url":null,"abstract":"<div><div>This study examines how proximity to metro stations is associated with housing prices in Medellín, Colombia, and how this relationship varies across income groups and spatial scales. While the capitalization of transport infrastructure into real estate values is widely recognized, its social and spatial heterogeneity remains underexplored in Latin American cities. Using a dataset of georeferenced housing transactions and spatial information on urban amenities, we estimate hedonic pricing models—including Ordinary Least Squares, Spatial Autoregressive, and Spatial Error Models—with income-level interaction terms and stratified buffer thresholds optimized through an iterative procedure.</div><div>The results reveal that metro accessibility is not uniformly valued across the city. In low-income districts, proximity to a metro station is associated with a price premium of approximately 9.4%, whereas in high-income neighborhoods, it corresponds to a statistically significant price penalty of 9.8%. Moreover, the spatial reach of metro influence decreases with income level, ranging from 1.7 km in low-income areas to just 0.6 km in high-income zones. Among all urban amenities considered, metro stations exhibit the most pronounced variation in market valuation across socioeconomic groups.</div><div>These findings underscore the need for disaggregated, income-sensitive approaches in both spatial modeling and urban policy design. The buffer-based strategy proposed here offers a flexible and empirically grounded method to capture how accessibility is differentially perceived and valued in stratified urban contexts. In cities of the Global South—where inequality is spatially embedded—planning efforts must account for these disparities to ensure that transit investments promote inclusive urban development.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104571"},"PeriodicalIF":6.3,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146077907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-01DOI: 10.1016/j.jtrangeo.2026.104570
Sui Ye , Jianchao Xi , Ziqiang Li
Food delivery services (FDS) are rapidly reshaping urban food environments, with contested impacts on spatial inequality. Here, we analyze walking-based versus delivery-based food service accessibility and spatial inequality across 43,504 neighbourhoods in 19 Chinese megacities (>10 M population), encompassing 430,600 service outlets. We find that while pronounced spatial clustering of service outlets drives significant baseline inequality, food delivery services substantially mitigate this disparity through three mechanisms: expanding service reach (55.91% coverage increase), vastly increasing choice (28.75-fold increase), and reducing spatial variation in access time (19.93% CV decrease). Critically, our cross-city comparative analysis reveals this equalizing effect is neither uniform nor universal, but exhibits strong scale-dependency and morphological contingency. Smaller cities (<1500 km2) achieve substantial Gini coefficient reductions of 45–53%, with FDS functioning as transformative spatial equalizers through a “Hump Effect” that maximally benefits suburban belts (10–20 km from centers). Conversely, larger megacities (>2000 km2) experience attenuated improvements of 13–21%, where FDS operate more as “core optimizers” reinforcing existing hierarchies. Socioeconomic stratification analysis reveals a dual equity structure: while 80.8% of lowest-income neighbourhoods show positive accessibility improvements, wealth-based disparities persist—particularly in large cities where the highest-income quintile achieves gains nearly three times greater than the lowest-income quintile. However, a “compensatory floor effect” emerges in smaller cities, where lower-income neighbourhoods attain accessibility improvements comparable to middle-income groups in larger cities. These findings challenge simplistic narratives of either technology-driven polarization or universal democratization, demonstrating that FDS impacts are conditional outcomes shaped by interactions between platform algorithms, urban morphology, and developmental stage.
{"title":"Food delivery services reduce inequality in urban food accessibility: Evidence from 19 Chinese megacities","authors":"Sui Ye , Jianchao Xi , Ziqiang Li","doi":"10.1016/j.jtrangeo.2026.104570","DOIUrl":"10.1016/j.jtrangeo.2026.104570","url":null,"abstract":"<div><div>Food delivery services (FDS) are rapidly reshaping urban food environments, with contested impacts on spatial inequality. Here, we analyze walking-based versus delivery-based food service accessibility and spatial inequality across 43,504 neighbourhoods in 19 Chinese megacities (>10 M population), encompassing 430,600 service outlets. We find that while pronounced spatial clustering of service outlets drives significant baseline inequality, food delivery services substantially mitigate this disparity through three mechanisms: expanding service reach (55.91% coverage increase), vastly increasing choice (28.75-fold increase), and reducing spatial variation in access time (19.93% CV decrease). Critically, our cross-city comparative analysis reveals this equalizing effect is neither uniform nor universal, but exhibits strong scale-dependency and morphological contingency. Smaller cities (<1500 km<sup>2</sup>) achieve substantial Gini coefficient reductions of 45–53%, with FDS functioning as transformative spatial equalizers through a “Hump Effect” that maximally benefits suburban belts (10–20 km from centers). Conversely, larger megacities (>2000 km<sup>2</sup>) experience attenuated improvements of 13–21%, where FDS operate more as “core optimizers” reinforcing existing hierarchies. Socioeconomic stratification analysis reveals a dual equity structure: while 80.8% of lowest-income neighbourhoods show positive accessibility improvements, wealth-based disparities persist—particularly in large cities where the highest-income quintile achieves gains nearly three times greater than the lowest-income quintile. However, a “compensatory floor effect” emerges in smaller cities, where lower-income neighbourhoods attain accessibility improvements comparable to middle-income groups in larger cities. These findings challenge simplistic narratives of either technology-driven polarization or universal democratization, demonstrating that FDS impacts are conditional outcomes shaped by interactions between platform algorithms, urban morphology, and developmental stage.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104570"},"PeriodicalIF":6.3,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146077908","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A metric of global maritime supply chain disruptions: The global supply chain stress index - maritime (GSCSI-M)","authors":"Jean-François Arvis, Jean-Paul Rodrigue, Daria Ulybina, Cordula Rastogi","doi":"10.1016/j.jtrangeo.2026.104575","DOIUrl":"https://doi.org/10.1016/j.jtrangeo.2026.104575","url":null,"abstract":"","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"158 1","pages":""},"PeriodicalIF":6.1,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146110669","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The transition to electric vehicles is critical for sustainable transportation, yet disparities in the distribution of electric vehicle charging stations (EVCS) pose challenges for equitable adoption. This study investigates the accessibility of EVCS in rural and urban areas, focusing on how socio-demographic factors influence accessibility metrics. Using geographic information systems (GIS), network analysis, and binary logistic regression, we analyzed spatial data on EVCS locations and socio-demographic attributes such as income levels, educational attainment, and minority representation across diverse regions. Accessibility metrics, such as average distance and travel time to nearest EVCS, were calculated to assess geographical and social inequities. The closest facility analysis revealed substantial disparities in EVCS accessibility between urban and rural areas, with urban areas having an average travel distance of 3 miles to the nearest EVCS compared to 9 miles in rural areas. Statistical analyses underscored the influence of factors such as income and educational attainment on the availability and accessibility of EVCS. The results highlight critical gaps in infrastructure planning and underscore the need for policies aimed at equitable distribution of EVCS to ensure inclusive participation in the electric vehicles (EV) transition. This research provides valuable insights into the socio-spatial dynamics of EVCS accessibility, offering actionable recommendations for policymakers and urban planners to address existing inequities and promote sustainable and inclusive transportation systems.
{"title":"Disparities in electric vehicle charging stations in rural and urban areas: Analyzing accessibility and socio-demographic influence","authors":"Dennis Bwire , Tumaini Sakaza , Thobias Sando , Emmanuel Kidando","doi":"10.1016/j.jtrangeo.2026.104567","DOIUrl":"10.1016/j.jtrangeo.2026.104567","url":null,"abstract":"<div><div>The transition to electric vehicles is critical for sustainable transportation, yet disparities in the distribution of electric vehicle charging stations (EVCS) pose challenges for equitable adoption. This study investigates the accessibility of EVCS in rural and urban areas, focusing on how socio-demographic factors influence accessibility metrics. Using geographic information systems (GIS), network analysis, and binary logistic regression, we analyzed spatial data on EVCS locations and socio-demographic attributes such as income levels, educational attainment, and minority representation across diverse regions. Accessibility metrics, such as average distance and travel time to nearest EVCS, were calculated to assess geographical and social inequities. The closest facility analysis revealed substantial disparities in EVCS accessibility between urban and rural areas, with urban areas having an average travel distance of 3 miles to the nearest EVCS compared to 9 miles in rural areas. Statistical analyses underscored the influence of factors such as income and educational attainment on the availability and accessibility of EVCS. The results highlight critical gaps in infrastructure planning and underscore the need for policies aimed at equitable distribution of EVCS to ensure inclusive participation in the electric vehicles (EV) transition. This research provides valuable insights into the socio-spatial dynamics of EVCS accessibility, offering actionable recommendations for policymakers and urban planners to address existing inequities and promote sustainable and inclusive transportation systems.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104567"},"PeriodicalIF":6.3,"publicationDate":"2026-01-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146038032","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-22DOI: 10.1016/j.jtrangeo.2026.104564
Jin Ki Eom, Kwang-Sub Lee, Sangpil Ko, Jin Hong Min
Shrinking cities face complex mobility issues driven by demographic shifts, spatial contraction, and declining activity levels. They also undermine the effectiveness of conventional transportation modeling approaches. This study investigates how urban shrinkage—characterized by population decline, aging, and spatial reorganization—affects activity patterns and travel demand in regional cities. We introduce the ABATA (Activity-BAsed Traveler Analyzer), a web-based modeling system, and apply it to the shrinking city of Iksan, South Korea. Using ABATA, we analyze four scenarios: a 2024 base case and three alternatives for 2035 incorporating demographic and land use changes. The model integrates mobile phone data, household travel surveys, and census-block land use information to produce high-resolution hourly estimates of activity profiles and origin-destination flows. Simulation results reveal scenario-dependent variations in activity intensity, travel distances, and the distribution of trip purposes. Particularly, scenarios involving a larger elderly population and central area contraction show sharp declines in long-distance commuting, while decentralized land use patterns lead to a redistribution of demand toward peripheral districts. These findings suggest that policy efforts in shrinking cities should prioritize local accessibility, right-sized infrastructure, and scenario-based planning approaches. The study underscores ABATA's value in high-resolution, activity-based mobility analysis and its potential to inform strategic and adaptive transportation planning under conditions of urban shrinkage.
{"title":"Scenario-based mobility simulation for shrinking cities using ABATA: A case study of Iksan, South Korea","authors":"Jin Ki Eom, Kwang-Sub Lee, Sangpil Ko, Jin Hong Min","doi":"10.1016/j.jtrangeo.2026.104564","DOIUrl":"10.1016/j.jtrangeo.2026.104564","url":null,"abstract":"<div><div>Shrinking cities face complex mobility issues driven by demographic shifts, spatial contraction, and declining activity levels. They also undermine the effectiveness of conventional transportation modeling approaches. This study investigates how urban shrinkage—characterized by population decline, aging, and spatial reorganization—affects activity patterns and travel demand in regional cities. We introduce the ABATA (Activity-BAsed Traveler Analyzer), a web-based modeling system, and apply it to the shrinking city of Iksan, South Korea. Using ABATA, we analyze four scenarios: a 2024 base case and three alternatives for 2035 incorporating demographic and land use changes. The model integrates mobile phone data, household travel surveys, and census-block land use information to produce high-resolution hourly estimates of activity profiles and origin-destination flows. Simulation results reveal scenario-dependent variations in activity intensity, travel distances, and the distribution of trip purposes. Particularly, scenarios involving a larger elderly population and central area contraction show sharp declines in long-distance commuting, while decentralized land use patterns lead to a redistribution of demand toward peripheral districts. These findings suggest that policy efforts in shrinking cities should prioritize local accessibility, right-sized infrastructure, and scenario-based planning approaches. The study underscores ABATA's value in high-resolution, activity-based mobility analysis and its potential to inform strategic and adaptive transportation planning under conditions of urban shrinkage.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104564"},"PeriodicalIF":6.3,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146033644","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-22DOI: 10.1016/j.jtrangeo.2026.104562
Mingyue Li , Xinghua Li , Yuntao Guo , Xinwu Qian , Haobing Liu , Lu Teng
Heavy-duty trucks (HDTs) form the backbone of urban freight transportation, facilitating the long-distance goods movement and sustaining economic activities. Despite their significance, the spatiotemporal characteristics of HDT operations remain insufficiently explored, particularly in complex metropolitans. This study proposes a Multi-Resolution Spatiotemporal Fusion (MRSF) framework to uncover latent HDT activity-trip chain patterns using large-scale GPS data, point-of-interest (POI) information and road network attributes. The framework comprises three key components: a Trip Chain Structuring Module that reconstructed GPS data into structured trip chains by detecting freight-related stops and inferring stop purposes (i.e., activity) based on the functional categories of nearby POIs; a Resolution Fusion Module that captures spatiotemporal dynamics across nested spatial scales by integrating adaptive grid zoning and hierarchical weighting, enabling more accurate representation of both localized and regional freight activities; and a Feature Aggregation Module that combines trip dynamics and activity attributes into a unified similarity matrix, enabling robust pattern identification and analysis across HDT operations. Using a one-week GPS dataset from over 20,000 HDTs in the Harbin metropolitan region, China, MRSF identifies 160,983 unique trip chains and uncovers five distinct activity-trip chain patterns, characterized by spatiotemporal attributes including service regions, trip length, temporal rhythms, activity purposes, and functional context. Validation results demonstrate that MRSF consistently outperforms conventional fixed-resolution and single-feature models in clustering validity, boundary separability, results interpretability, and stability and robustness. The findings support evidence-based freight zoning and routing strategies, offering practical value for freight operators and urban policymakers engaged in advancing sustainable and resilient logistics systems.
{"title":"Spatiotemporal analysis of heavy-duty truck activity-trip chains in Harbin, China: A multi-resolution GPS data fusion approach","authors":"Mingyue Li , Xinghua Li , Yuntao Guo , Xinwu Qian , Haobing Liu , Lu Teng","doi":"10.1016/j.jtrangeo.2026.104562","DOIUrl":"10.1016/j.jtrangeo.2026.104562","url":null,"abstract":"<div><div>Heavy-duty trucks (HDTs) form the backbone of urban freight transportation, facilitating the long-distance goods movement and sustaining economic activities. Despite their significance, the spatiotemporal characteristics of HDT operations remain insufficiently explored, particularly in complex metropolitans. This study proposes a Multi-Resolution Spatiotemporal Fusion (MRSF) framework to uncover latent HDT activity-trip chain patterns using large-scale GPS data, point-of-interest (POI) information and road network attributes. The framework comprises three key components: a Trip Chain Structuring Module that reconstructed GPS data into structured trip chains by detecting freight-related stops and inferring stop purposes (i.e., activity) based on the functional categories of nearby POIs; a Resolution Fusion Module that captures spatiotemporal dynamics across nested spatial scales by integrating adaptive grid zoning and hierarchical weighting, enabling more accurate representation of both localized and regional freight activities; and a Feature Aggregation Module that combines trip dynamics and activity attributes into a unified similarity matrix, enabling robust pattern identification and analysis across HDT operations. Using a one-week GPS dataset from over 20,000 HDTs in the Harbin metropolitan region, China, MRSF identifies 160,983 unique trip chains and uncovers five distinct activity-trip chain patterns, characterized by spatiotemporal attributes including service regions, trip length, temporal rhythms, activity purposes, and functional context. Validation results demonstrate that MRSF consistently outperforms conventional fixed-resolution and single-feature models in clustering validity, boundary separability, results interpretability, and stability and robustness. The findings support evidence-based freight zoning and routing strategies, offering practical value for freight operators and urban policymakers engaged in advancing sustainable and resilient logistics systems.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104562"},"PeriodicalIF":6.3,"publicationDate":"2026-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146033643","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-21DOI: 10.1016/j.jtrangeo.2026.104558
Bozhan Qin , Min Yang , Yucheng Wang , Fan Jiang , Boqing Wang , Mingye Zhang
Understanding the ground access travel behaviour of airport users is essential for improving airport services. While previous studies largely focus on unimodal travel, limited attention has been paid to door-to-airport intermodal access. This study examines ground intermodal access mode choice behaviour, using Beijing Daxing International Airport as a case study. We apply Extreme Gradient Boosting (XGBoost) to model mode choice behaviour, and utilise interpretable machine learning techniques including SHapley Additive exPlanation (SHAP) values and Accumulated Local Effects (ALE) plots to capture nonlinear behavioural patterns. Findings reveal that intermodal choices are strongly shaped by traveller characteristics and access/feeder travel time thresholds. The metro-private vehicle intermodal is attractive when metro travel time is 43–62 or 75–105 min, and feeder time exceeds 7 min. The airport coach-private vehicle intermodal is appealing when the feeder time is under 16 min and the coach line-haul time exceeds 79 min. Although the effect is modest, the high-speed rail–private vehicle intermodal is facilitated by a feeder time of 14–33 min. Key policy implications include time- and threshold-specific strategies, with integrated bundles and real-time coordination of feeder and line-haul services. The study advances understanding of threshold-sensitive intermodal decisions and provides insights for developing sustainable and traveller-oriented airport ground transport services.
{"title":"Examining airport intermodal access mode choice behaviour using interpretable machine learning","authors":"Bozhan Qin , Min Yang , Yucheng Wang , Fan Jiang , Boqing Wang , Mingye Zhang","doi":"10.1016/j.jtrangeo.2026.104558","DOIUrl":"10.1016/j.jtrangeo.2026.104558","url":null,"abstract":"<div><div>Understanding the ground access travel behaviour of airport users is essential for improving airport services. While previous studies largely focus on unimodal travel, limited attention has been paid to door-to-airport intermodal access. This study examines ground intermodal access mode choice behaviour, using Beijing Daxing International Airport as a case study. We apply Extreme Gradient Boosting (XGBoost) to model mode choice behaviour, and utilise interpretable machine learning techniques including SHapley Additive exPlanation (SHAP) values and Accumulated Local Effects (ALE) plots to capture nonlinear behavioural patterns. Findings reveal that intermodal choices are strongly shaped by traveller characteristics and access/feeder travel time thresholds. The metro-private vehicle intermodal is attractive when metro travel time is 43–62 or 75–105 min, and feeder time exceeds 7 min. The airport coach-private vehicle intermodal is appealing when the feeder time is under 16 min and the coach line-haul time exceeds 79 min. Although the effect is modest, the high-speed rail–private vehicle intermodal is facilitated by a feeder time of 14–33 min. Key policy implications include time- and threshold-specific strategies, with integrated bundles and real-time coordination of feeder and line-haul services. The study advances understanding of threshold-sensitive intermodal decisions and provides insights for developing sustainable and traveller-oriented airport ground transport services.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104558"},"PeriodicalIF":6.3,"publicationDate":"2026-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146015042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-21DOI: 10.1016/j.jtrangeo.2026.104555
Zhiqiang Zheng
Large-scale infrastructure projects along development corridors are widely expected to facilitate rural transformation, create new livelihood opportunities, and alleviate poverty in the Global South. Yet, these projects also have uneven impacts on mobility and risk, causing mobility injustice and deepening inequalities among rural populations. This study uses qualitative methods to examine how large-scale transport infrastructure in northern Laos reshapes rural mobility dynamics to impacts the livelihoods of various rural groups. This paper suggests mobility injustice as an approach to understanding the dynamic effects of infrastructure development on rural livelihoods. The findings reveal that large-scale transport infrastructure has facilitated travel convenience and tourism-related jobs for groups residing in economically visible and tourism-focused areas. However, the findings indicate that the large-scale infrastructure development has not effectively mitigated the spatial differences in mobility that shape rural livelihoods. The impacts suggest risks of mobility injustice for disadvantaged rural groups living far away from the main transport route and non-tourist areas. Moreover, within the newly large-scale infrastructure development, the future maintenance and development of transport networks may exacerbate challenges for the inclusive transformation of rural livelihoods and mobility justice in different regions. This research rethinks the relationship between rural livelihood transformation and infrastructure development by considering the realities and needs of diverse rural populations in the Global South.
{"title":"Encountering large-scale infrastructure development in the rural areas: Dynamic effects on rural livelihoods and mobility injustice in northern Laos","authors":"Zhiqiang Zheng","doi":"10.1016/j.jtrangeo.2026.104555","DOIUrl":"10.1016/j.jtrangeo.2026.104555","url":null,"abstract":"<div><div>Large-scale infrastructure projects along development corridors are widely expected to facilitate rural transformation, create new livelihood opportunities, and alleviate poverty in the Global South. Yet, these projects also have uneven impacts on mobility and risk, causing mobility injustice and deepening inequalities among rural populations. This study uses qualitative methods to examine how large-scale transport infrastructure in northern Laos reshapes rural mobility dynamics to impacts the livelihoods of various rural groups. This paper suggests mobility injustice as an approach to understanding the dynamic effects of infrastructure development on rural livelihoods. The findings reveal that large-scale transport infrastructure has facilitated travel convenience and tourism-related jobs for groups residing in economically visible and tourism-focused areas. However, the findings indicate that the large-scale infrastructure development has not effectively mitigated the spatial differences in mobility that shape rural livelihoods. The impacts suggest risks of mobility injustice for disadvantaged rural groups living far away from the main transport route and non-tourist areas. Moreover, within the newly large-scale infrastructure development, the future maintenance and development of transport networks may exacerbate challenges for the inclusive transformation of rural livelihoods and mobility justice in different regions. This research rethinks the relationship between rural livelihood transformation and infrastructure development by considering the realities and needs of diverse rural populations in the Global South.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104555"},"PeriodicalIF":6.3,"publicationDate":"2026-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146015043","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-20DOI: 10.1016/j.jtrangeo.2026.104559
Laura Antón-González, Israel Villarrasa-Sapiña, Luis-Millán González, José Devís-Devís
Bicycle use in urban environments is a feasible and beneficial alternative to motor vehicles. However, in Spain, bicycles are used less than other means of transport. This study aimed to predict cycling in urban settings through the presence of cycling facilities and car or public transport options around residences, as well as sociodemographic characteristics, perceived health, and physical activity levels of residents in the four most populated Spanish cities. A total of 957 individuals completed a questionnaire via the Maptionnaire platform. After processing the data, a chi-square test was applied to identify significant differences between bicycle use and sociodemographic variables, and a decision tree was built to predict urban cycling. Gender and socioeconomic status showed significant differences in bicycle use. The most important differentiating variable in the decision tree was leisure-time physical activity. The decision tree produced 20 individual profiles, five of which included the largest number of participants, all non-cyclists. The most relevant variables indicated that the availability of cycling infrastructure was the strongest predictor, followed by leisure-time physical activity, age, and income level. Being physically active during leisure time and having a medium-to-high socioeconomic status were associated with bicycle use for commuting, whereas individuals aged 41 years or older did not cycle. Furthermore, women and those with higher BMI exhibited lower levels of bicycle use. To effectively promote cycling in cities, transport-related environmental amenities should not be addressed in isolation but integrated with the characteristics of residents, considering especially gender differences.
{"title":"Urban cycling: A non-linear analysis of transport amenities and socio-demographic and health variables","authors":"Laura Antón-González, Israel Villarrasa-Sapiña, Luis-Millán González, José Devís-Devís","doi":"10.1016/j.jtrangeo.2026.104559","DOIUrl":"10.1016/j.jtrangeo.2026.104559","url":null,"abstract":"<div><div>Bicycle use in urban environments is a feasible and beneficial alternative to motor vehicles. However, in Spain, bicycles are used less than other means of transport. This study aimed to predict cycling in urban settings through the presence of cycling facilities and car or public transport options around residences, as well as sociodemographic characteristics, perceived health, and physical activity levels of residents in the four most populated Spanish cities. A total of 957 individuals completed a questionnaire via the Maptionnaire platform. After processing the data, a chi-square test was applied to identify significant differences between bicycle use and sociodemographic variables, and a decision tree was built to predict urban cycling. Gender and socioeconomic status showed significant differences in bicycle use. The most important differentiating variable in the decision tree was leisure-time physical activity. The decision tree produced 20 individual profiles, five of which included the largest number of participants, all non-cyclists. The most relevant variables indicated that the availability of cycling infrastructure was the strongest predictor, followed by leisure-time physical activity, age, and income level. Being physically active during leisure time and having a medium-to-high socioeconomic status were associated with bicycle use for commuting, whereas individuals aged 41 years or older did not cycle. Furthermore, women and those with higher BMI exhibited lower levels of bicycle use. To effectively promote cycling in cities, transport-related environmental amenities should not be addressed in isolation but integrated with the characteristics of residents, considering especially gender differences.</div></div>","PeriodicalId":48413,"journal":{"name":"Journal of Transport Geography","volume":"131 ","pages":"Article 104559"},"PeriodicalIF":6.3,"publicationDate":"2026-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146006411","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号