Frequent public transport coverage of the population in Latin American capital cities

INTRODUCTION

Frequent public transport is the backbone of daily accessibility in large Latin American cities, where a substantial share of trips is made by bus, metro and informal services rather than by private car. A stop is conventionally treated as frequent when a vehicle arrives at least every 15 minutes, and a resident is considered covered when that stop lies within a short walk of home, most often 400 to 800 metres.( 1)

The two thresholds, service frequency and walking distance, jointly define the population that can rely on public transport without consulting a timetable, and they are the operational core of policies that promise a frequent network for all residents.( 2) Walking distance is not a minor detail in this definition, because reported walking distances to stops concentrate within 500 to 800 m and decline steeply beyond that range.( 3,4)

The share of residents who live within reach of frequent service is a direct measure of whether the network reaches the people it is meant to serve, and it underlies the international monitoring of access to public transport as part of the sustainable development agenda.( 5)

Several of the region's capitals pioneered bus rapid transit and integrated fare systems during the 2000s, so the question of how much of the population these networks actually reach is of long-standing policy interest. Measuring this share requires combining the geometry of the network with a spatially explicit population layer, a task that open schedules and open population grids have now made reproducible.

The spread of the General Transit Feed Specification has turned this measurement into a routine computation, because schedules, stop locations and frequencies are published as open, machine-readable files in many cities.( 6,7)

Researchers have used these files to quantify job accessibility and its social distribution in São Paulo, Rio de Janeiro, Curitiba and Recife,( 8) to show how bus rapid transit changed access to opportunities for the poor in Lima,( 9) and to document the uneven geography of accessibility in Montevideo.( 10,11)

Slovic and colleagues linked accessibility restrictions to overlapping social inequalities in São Paulo,( 12) and Moreno-Monroy and colleagues connected public transport provision to school access in the same city.( 13) City-level GIS assessments in Mexico and Brazil show that the spatial reach of stops and frequent corridors varies widely within the region, even among cities that appear similar in size.( 14,15)

Capital cities are the natural unit for such a comparison because they concentrate both population and the institutional capacity to publish open schedules, and because their networks set the standard that secondary cities later follow. Informal and semiformal services, which carry a large share of trips in many of these cities, are usually absent from open schedules and therefore remain outside most coverage measures.( 16)

Methodological work has matured in parallel: network-based measures, the choice of accessibility indicator and GTFS-based gap metrics now allow coverage and equity to be compared across cities and over time.( 17-19)

Most of this regional literature measures travel times to destinations rather than the simpler supply-side question of whether a resident lives within walking distance of a frequent stop, and it concentrates on the largest and best-documented metropolitan areas.( 20)

The coverage of the population by frequent service, computed consistently across the capital cities of the region, has received less attention.( 21,22)

The objective of this study was to quantify the share of the population living within walking distance of a frequent public transport stop in Latin American capital cities, using only open, downloadable schedules and an open population grid. Three research questions guided the analysis. P1: which Latin American capital cities publish an open GTFS schedule feed that can be downloaded without login, and what network does each feed describe? P2: what share of the population in each capital lives within 800 m of a stop with weekday service at least every 15 minutes? P3: how much does that share change when the walking distance is shortened to 500 m, the frequency threshold is tightened to 10 minutes, or the study area is enlarged to a 15 km radius?

METHODS

The reporting follows the STROBE statement for observational studies. The design is cross-sectional and ecological: the unit of analysis is the capital city, and the outcome is a single coverage share computed for each city from two open data sources, without any individual-level data. Because the metric is computed over the complete grid and the complete stop set of each city, it is a census of the study area rather than an estimate from a sample.

RESULTS

The analysis comprised five capitals, seven GTFS feeds and a combined study-area population of 11,710,484 people distributed over the 2020 population grid. The three questions are reported in turn, each subsection opening with the main figure and closing with a single-number summary.

Frequent service covered most of the population in four capitals

With respect to the second question, frequent service covered most of the population in four capitals and about one third in La Paz (Table 2; Figure 1; Figure 2).

Capital	Population in study area	Stops (n)	Frequent stops (n)	Covered population	Coverage share (%)
Argentina - Buenos Aires	2364701	44550	30811	2344741	99.2
Chile - Santiago	2334527	12877	11974	2329000	99.8
Colombia - Bogota	4361920	8309	7717	4347231	99.7
Nicaragua - Managua	1209169	2247	1406	1075391	88.9
Bolivia - La Paz	1440168	76	50	494678	34.3

Note: Study area: 10 km disc around the city central reference point. Frequent stop: average weekday headway (07:00-20:00) of 15 minutes or less.

The population-weighted coverage across the five cities was 90.4%. By city, the share was 99.8% in Santiago, 99.7% in Bogotá, 99.2% in Buenos Aires, 88.9% in Managua and 34.3% in La Paz.

The median city coverage was 99.2%, with a range from 34.3% to 99.8%. Buenos Aires combined three feeds (bus, metro and commuter rail), whereas La Paz contributed a single cable car feed, and this difference in network completeness is visible in the number of stops reported in the table. In the three high-coverage capitals the share stayed above 97 percent even under the strictest combination of thresholds, which indicates that the walking buffers of frequent stops overlap extensively in the densest areas. The number of frequent stops ranged from 50 in La Paz to 30,811 in Buenos Aires, and the study-area population ranged from 1,209,169 in Managua to 4,361,920 in Bogotá. The single figure that summarises this block is the population-weighted coverage of 90.4%.

Coverage was robust to stricter thresholds except in Managua and La Paz

As for the third question, the coverage estimates were robust to stricter thresholds in the three largest networks but not in Managua and La Paz (Table 3; Figure 3). Under the 500 m walk and the 10-minute headway, coverage fell to 73.5% in Managua and to 18.7% in La Paz, whereas Buenos Aires, Santiago and Bogotá remained between 97.7% and 97.9%.

Capital	800 m, 15 min	800 m, 10 min	500 m, 15 min	500 m, 10 min
Argentina - Buenos Aires	99.2	99.0	98.2	97.9
Chile - Santiago	99.8	99.7	98.7	98.2
Colombia - Bogota	99.7	99.6	98.1	97.7
Nicaragua - Managua	88.9	82.0	80.8	73.5
Bolivia - La Paz	34.3	34.3	18.7	18.7

Note: Columns combine walking distance (w, metres) and frequency threshold (f, minutes).

The 10-minute threshold alone reduced Managua to 82.0% but left the other capitals nearly unchanged. The 500 m distance with the 15-minute headway produced an intermediate pattern, with Managua at 80.8% and La Paz at 18.7%. Enlarging the study area to a 15 km radius lowered coverage to 83.2% in Managua and to 25.4% in La Paz, while the three consolidated networks stayed above 98.3%. The summary figure of this block is the 18.7% reached by La Paz under the strictest combination.

DISCUSSION

The present study asked what share of the population of Latin American capital cities lives within walking distance of a frequent public transport stop. The answer has two parts: four of the five capitals that publish an open urban schedule reach near-universal coverage within 800 m of a stop with a 15-minute headway, while La Paz, where the open feed covers only the cable car network, reaches a markedly lower share.

Regarding P1, the finding that only five of twenty capitals offer a downloadable urban GTFS feed is itself informative about the region. Giesen and colleagues described public transport data as a research priority for Latin America and the Caribbean precisely because open schedules remain scarce.( 20)

Hidalgo and colleagues, mapping universal access experiences in the region, documented that the absence of standardised open data is one of the barriers to measuring and improving access.( 31) This pattern matches the wider literature on transport data availability, in which open and proprietary sources coexist unevenly and the Global South lags behind.( 7)

Conversely, the exclusion of Mexico City and San José reflects transient server failures rather than a structural absence of data, and these cities could enter future updates of the same protocol without any change in method. The detection of coverage gaps, which motivated several recent city studies, depends on the same schedules and is therefore impossible where no feed exists.( 22)

As an alternative to schedule files, Haamer and colleagues reconstructed accessibility from GPS vehicle traces, which could mitigate the data gap where feeds are absent.( 32)

The coverage values should be read as a snapshot of the networks as published at the extraction date, and the differing feed vintages mean that the comparison captures the current state of each network rather than a common calendar year. The asymmetry is consequential for equity monitoring, because the cities that are hardest to measure are often those where access is most contested, and the absence of a feed removes them from regional comparisons altogether.

Regarding P2, the near-universal coverage measured in Buenos Aires, Santiago and Bogotá is consistent with evidence that these cities operate dense, high-frequency networks. Boisjoly and colleagues reported that the Brazilian metropolises with formal networks concentrate most of their population within reach of frequent service when such service exists,( 8) and de Almeida Correia and colleagues found that Curitiba's accessibility is high but varies across modes and times of day.( 15)

Verbavatz and Barthelemy, using a global metric of access to mass rapid transit, showed that coverage close to saturation is attainable in cities that combine metros with frequent surface networks.( 33)

Oviedo and colleagues showed that a single bus rapid transit corridor in Lima improved access for the poor, which supports the interpretation that sustained trunk investment explains the high values in Bogotá and Santiago.( 9)

The lower value in Managua is closer to the levels reported by Falchetta and colleagues for paratransit-dominated African cities, where formal scheduled networks cover only part of the built-up area,( 34) and to the partial coverage that Mwesigwa and colleagues described for a bus rapid transit corridor in Dar es Salaam.( 35)

The La Paz result reflects a measurement limitation as much as a service one: the city's bus and minibus network is not published as an open schedule, so the coverage estimate captures only the cable car system, which serves a linear corridor rather than the whole urban area. Heinrichs and Bernet described the same mechanism for Medellín, where aerial cable cars improved access in informal hillside settlements but did not substitute a full urban network,( 36) and Stiller and colleagues showed that semiformal transport provides much of the residual accessibility in that city.( 16)

Comparable city-level GIS assessments in Mexico support the view that coverage varies with network completeness rather than with city size alone.( 14,37) It is worth stressing that coverage measures access to the network, not access to destinations, so a resident within walking distance of a frequent stop may still face long or inconvenient journeys once aboard.( 2)

Regarding P3, the robustness of the estimates under stricter thresholds distinguishes the cities. In Buenos Aires, Santiago and Bogotá the 10-minute and 500 m definitions barely changed coverage, which indicates that the frequent network is dense enough that most residents are served by several nearby frequent lines. Kim and Lee argued that frequency, rather than travel time, is what distinguishes usable service,( 2) and Wang and Loo showed that a highly transit-oriented city can still leave specific groups disadvantaged once fine-grained service frequency is considered.( 38)

The sensitivity of Managua and La Paz to the thresholds is consistent with networks in which frequency is concentrated on fewer corridors, the same pattern that a transit-gap framework revealed in other contexts.( 21,39)

The 15 km sensitivity confirms that this lower coverage is not an artefact of the 10 km boundary, because the frequent network itself does not extend into the wider periphery in those cities. The results also illustrate a practical point for indicator design: a single threshold applied uniformly will classify near-saturated networks as similar even when their service quality differs, so frequency-based coverage is best reported together with the underlying headway distribution.

The walking-distance dimension behaves as expected from the literature: van Soest and colleagues and Taylor and Somenahalli reported that the share of users declines steeply beyond roughly 500 to 800 m,( 3,4) which explains why the 500 m buffer, and not the frequency threshold, produced the largest drop in La Paz. Accessibility inequality measured at fine resolution also concentrates at such pedestrian thresholds, and the indicator is therefore sensitive to how coverage is defined.( 25,40,41)

The regional evidence on equity reinforces this interpretation. Hernandez showed that the social distribution of accessibility in Montevideo is uneven across income groups,( 10) and Slovic and colleagues linked accessibility shortfalls to overlapping inequalities in São Paulo.( 12)

Ribeiro and colleagues proposed accessibility as an indicator of social exclusion precisely because coverage gaps concentrate in the same neighbourhoods where other disadvantages accumulate.(41) Liu and colleagues, working with high-resolution transit data, showed that the same gaps affect people with mobility disabilities more sharply.( 40)

The coverage metric computed here is therefore a first screen rather than a substitute for distributional analysis within each city, and it identifies the cities where such analysis is most urgent.

Several methodological strengths support the internal validity of the findings. The input data are fully public, the computation is deterministic and reproducible, and the coverage metric is a complete enumeration of the population grid rather than an approximation. The use of a single population layer and a single frequency definition across cities removes two common sources of incomparability that earlier comparative work identified as consequential.( 17,18)

The sensitivity analyses make the role of each threshold explicit, which addresses the criticism that accessibility rankings change when the indicator is changed. These properties matter because coverage indicators for transport are increasingly used in monitoring and planning, and a metric whose assumptions are hidden cannot be audited. Because the study relies on schedule files and a population grid that are both released under open licences, the results can be recomputed and challenged by third parties, which is a precondition for trust in the indicator.

Several limitations point to concrete extensions. The sample is restricted to capitals with an open urban GTFS feed, so the results cannot be extrapolated to the fifteen excluded capitals, several of which are large, and the uneven data availability documented by Mahajan and colleagues suggests that this restriction will relax slowly unless open data are mandated.( 7)

The circular study area is a standardised analytical zone and not an administrative boundary, so it may include peripheral areas outside municipal limits or exclude parts of the metropolitan region. The population grid resolves about one kilometre, which is coarse relative to the 500 and 800 m buffers, although the area-weighted intersection mitigates this mismatch, and the choice between centroid and area weighting is itself a known source of variation in walking-access estimates.( 24,28)

The La Paz estimate covers only the cable car network and therefore understates the true coverage of that city. The population layer refers to 2020 while the schedules are recent, and the schedules describe planned rather than observed service, so actual reliability and crowding are not captured. Each of these limitations points to a concrete extension: integrating semiformal transport, using finer grids or building footprints, and validating scheduled headways against real-time vehicle positions.

These results also speak to comparability beyond the region. The coverage shares measured here sit within the range reported for formal networks in OECD cities( 33) and above the levels typical of paratransit-dominated African cities,( 34) which suggests that the frequent-network coverage metric behaves consistently across contexts when the same open-data pipeline is applied.

Because the pipeline uses only open schedules and an open population grid, it can be extended to any city that publishes a feed, and the main barrier to extension is the availability of schedules rather than the cost of computation. The extension would also allow within-region rankings to be updated as new feeds appear, which is a practical benefit for agencies and researchers alike.

The implications for policy are direct. Cities that already publish open schedules can be monitored continuously, and the five included capitals show that the population within reach of frequent service is now a computable indicator rather than an assumption, which aligns with the research priorities set for the region.( 20)

The comparison also exposes the asymmetric burden of data: residents of capitals without an open feed remain invisible to this kind of measurement, which is itself an argument for open data mandates. What remains open is whether the excluded capitals, once their schedules are published, will show coverage closer to the consolidated networks of Santiago and Bogotá or to the partial networks of Managua and La Paz, and whether the semiformal services that carry much of the region's demand can be brought into the same measurement frame.( 16,34)

CONCLUSIONS

Open urban GTFS schedules were identified and analysed for five of the twenty Latin American capital cities screened. Frequent service within 800 m covered the great majority of the combined population, with near-universal values in Santiago, Bogotá and Buenos Aires and lower values in Managua and La Paz. Coverage was robust to stricter thresholds in the consolidated networks but fell in Managua and La Paz, where the published network is smaller or only partial.

References

1. 1 El-Geneidy A, Grimsrud M, Wasfi R, Tétreault P, Surprenant-Legault J. New evidence on walking distances to transit stops: identifying redundancies and gaps using variable service areas. Transportation. 2014;41(1):193-210. doi: 10.1007/s11116-013-9508-z
2. 2 Kim J, Lee B. More than travel time: New accessibility index capturing the connectivity of transit services. Journal of Transport Geography. 2019;78:8-18. doi: 10.1016/j.jtrangeo.2019.05.008
3. 3 van Soest D, Tight MR, Rogers CDF. Exploring the distances people walk to access public transport. Transport Reviews. 2020;40(2):160-182. doi: 10.1080/01441647.2019.1575491
4. 4 Taylor MAP, Somenahalli S. Distributions of walking access to public transport in Melbourne, Australia – Evidence on acceptable and tolerable walking distances. International Journal of Sustainable Transportation. 2024;18(7):576-588. doi: 10.1080/15568318.2024.2385641
5. 5 Dabagh S, Miristice LMB, Gentile G. Accessibility Via Public Transport Through Gravity Models Based on Open Data. Transport and Telecommunication Journal. 2024;25(4):359-369. doi: 10.2478/ttj-2024-0026
6. 6 MobilityData / Google. GTFS Static Overview - General Transit Feed Specification [Internet]. 2025. Available from: https://gtfs.org/documentation/schedule/reference/
7. 7 Mahajan V, Kuehnel N, Intzevidou A, Cantelmo G, Moeckel R, Antoniou C. Data to the people: a review of public and proprietary data for transport models. Transport Reviews. 2022;42(4):415-440. doi: 10.1080/01441647.2021.1977414
8. 8 Boisjoly G, Serra B, Oliveira GT, El-Geneidy A. Accessibility measurements in São Paulo, Rio de Janeiro, Curitiba and Recife, Brazil. Journal of Transport Geography. 2020;82:102551. doi: 10.1016/j.jtrangeo.2019.102551
9. 9 Oviedo D, Scholl L, Innao M, Pedraza L. Do Bus Rapid Transit Systems Improve Accessibility to Job Opportunities for the Poor? The Case of Lima, Peru. Sustainability. 2019;11(10):2795. doi: 10.3390/su11102795
10. 10 Hernandez D. Uneven mobilities, uneven opportunities: Social distribution of public transport accessibility to jobs and education in Montevideo. Journal of Transport Geography. 2018;67:119-125. doi: 10.1016/j.jtrangeo.2017.08.017
11. 11 Hernandez D, Hansz M, Massobrio R. Job accessibility through public transport and unemployment in Latin America: The case of Montevideo (Uruguay). Journal of Transport Geography. 2020;85:102742. doi: 10.1016/j.jtrangeo.2020.102742
12. 12 Slovic AD, Tomasiello DB, Giannotti M, Andrade MDF, Nardocci AC. The long road to achieving equity: Job accessibility restrictions and overlapping inequalities in the city of São Paulo. Journal of Transport Geography. 2019;78:181-193. doi: 10.1016/j.jtrangeo.2019.06.003
13. 13 Moreno-Monroy AI, Lovelace R, Ramos FR. Public transport and school location impacts on educational inequalities: Insights from São Paulo. Journal of Transport Geography. 2018;67:110-118. doi: 10.1016/j.jtrangeo.2017.08.012
14. 14 Zazueta-Terrazas AG, Plata-Rocha W, Monjardín-Armenta SA, Íñiguez-Ayón YP, Rivas-Báez VM. Sustainable development goals (SDGs): spatial impact of public transport stops on the population of Culiacan, Sinaloa, Mexico. GeoJournal. 2025;90(6):304. doi: 10.1007/s10708-025-11547-7
15. 15 de Almeida Correia P, Carpio-Pinedo J, Moya-Gómez B, Condeço-Mehorado A. Modal Accessibility Gap in Curitiba (Brazil). Dynamic Analysis Considering Time and Spatial Variations. Appl Spatial Analysis. 2025;18(1):34. doi: 10.1007/s12061-025-09639-5
16. 16 Stiller D, Wurm M, Sapena M, Nieland S, Dech S, Taubenböck H. Does formal public transport serve the city well? The importance of semiformal transport for the accessibility in Medellín, Colombia. PLoS ONE. 2025;20(4):e0321691. doi: 10.1371/journal.pone.0321691
17. 17 Luo D, Cats O, van Lint H, Currie G. Integrating network science and public transport accessibility analysis for comparative assessment. Journal of Transport Geography. 2019;80:102505. doi: 10.1016/j.jtrangeo.2019.102505
18. 18 Klar B, Lee J, Long JA, Diab E. The impacts of accessibility measure choice on public transit project evaluation: A comparative study of cumulative, gravity-based, and hybrid approaches. Journal of Transport Geography. 2023;106:103508. doi: 10.1016/j.jtrangeo.2022.103508
19. 19 Li Y, Fan W“. Modeling and Evaluating Public Transit Equity and Accessibility by Integrating General Transit Feed Specification Data: Case Study of the City of Charlotte. J Transp Eng, Part A: Systems. 2020;146(10):04020112. doi: 10.1061/jtepbs.0000426
20. 20 Giesen R, Hidalgo D, Ramos R. Public transport: Status and research priorities for Latin America and the Caribbean. Latin American Transport Studies. 2025;3:100025. doi: 10.1016/j.latran.2025.100025
21. 21 Yadav M, Mepparambath RM, Patil GR. An enhanced transit accessibility evaluation framework by integrating Public Transport Accessibility Levels (PTAL) and transit gap. Journal of Transport Geography. 2024;121:104013. doi: 10.1016/j.jtrangeo.2024.104013
22. 22 Tiwari A. Detecting Transit Deserts Through a Blend of Machine Learning (ML) Approaches, Including Decision Trees (DTs), Logistic Regression (LR), and Random Forest (RF) in Lucknow. Future Transportation. 2025;5(2):70. doi: 10.3390/futuretransp5020070
23. 23 MobilityData. Mobility Database catalogs [Internet]. 2025. Available from: https://github.com/MobilityData/mobility-database-catalogs
24. 24 Schiavina M, Freire S, MacManus K, European Commission Joint Research Centre. GHS-POP R2023A - GHS population grid multitemporal (1975-2030) [Internet]. 2023. Available from: https://ghsl.jrc.ec.europa.eu/
25. 25 Liu D, Kan Z, Wang J, Kwan MP, Song J, Wei J. Using spatially explicit high-granularity 3D geospatial data for quantifying public transport walking accessibility inequality and vulnerability in the x-minute city. Cities. 2025;166:106245. doi: 10.1016/j.cities.2025.106245
26. 26 Hasan MM, Islam GMT, Matsumoto PSS, Novak M, Rahaman KR. Employing Geographic Information Systems in Analyzing Pedestrian Accessibility to Public Bus Stops in Halifax. J Urban Plann Dev. 2023;149(3):04023016. doi: 10.1061/jupddm.upeng-4268
27. 27 Taplin JH, Sun Y. Optimizing bus stop locations for walking access: Stops-first design of a feeder route to enhance a residential plan. Environment and Planning B: Urban Analytics and City Science. 2020;47(7):1237-1259. doi: 10.1177/2399808318824108
28. 28 Grisé E, Wasfi R, Ross NA, El-Geneidy A. Evaluating methods for measuring daily walking to public transport: Balancing accuracy and data availability. Journal of Transport & Health. 2019;15:100638. doi: 10.1016/j.jth.2019.100638
29. 29 Naro G, Biraghi CA, Lenzi E. City Transport Analyzer: A Powerful Qgis Plugin For Public Transport Accessibility And Intermodality Analysis. Int Arch Photogramm Remote Sens Spatial Inf Sci. 2024;XLVIII-4/W12-2024(4/W12-2024):113-119. doi: 10.5194/isprs-archives-xlviii-4-w12-2024-113-2024
30. 30 Goliszek S. Spatio‐Temporal Accessibility Modeling With Mobile Phone and GTFS Data: Insights for Urban Transport Planning in Helsinki. Transactions in GIS. 2025;29(8):e70163. doi: 10.1111/tgis.70163
31. 31 Hidalgo D, Urbano C, Olivares C, Tinjacá N, Pérez JM, Pardo CF, et al. Mapping Universal Access Experiences for Public Transport in Latin America. Transportation Research Record: Journal of the Transportation Research Board. 2020;2674(12):79-90. doi: 10.1177/0361198120949536
32. 32 Haamer M, Aasa A, Poom A. Assessing public transport accessibility using GPS data. Eur Transp Res Rev. 2025;17(1):36. doi: 10.1186/s12544-025-00733-w
33. 33 Verbavatz V, Barthelemy M. Access to mass rapid transit in OECD urban areas. Sci Data. 2020;7(1):301. doi: 10.1038/s41597-020-00639-3
34. 34 Falchetta G, Noussan M, Hammad AT. Comparing paratransit in seven major African cities: An accessibility and network analysis. Journal of Transport Geography. 2021;94:103131. doi: 10.1016/j.jtrangeo.2021.103131
35. 35 Mwesigwa L, Yin Z, Farber S. Evaluating the level of access and equity of the bus rapid transit (BRT) system: The case of Dar-Es-Salaam, Tanzania. Journal of Transport Geography. 2024;119:103953. doi: 10.1016/j.jtrangeo.2024.103953
36. 36 Heinrichs D, Bernet JS. Public Transport and Accessibility in Informal Settlements: Aerial Cable Cars in Medellín, Colombia. Transportation Research Procedia. 2014;4:55-67. doi: 10.1016/j.trpro.2014.11.005
37. 37 Calderón-Ramírez J, Gutiérrez-Moreno M, Mungaray-Moctezuma A, Sánchez-Atondo A, García-Gómez L, Montoya-Alcaraz M, et al. Spatial Analysis of Public Transport and Urban Mobility in Mexicali, B.C., Mexico: Towards Sustainable Solutions in Developing Cities. Sustainability. 2025;17(17):7802. doi: 10.3390/su17177802
38. 38 Wang H, Loo BPY. The public transport disadvantaged in a highly transit-oriented city: An analytical framework, key challenges and opportunities. Journal of Transport Geography. 2024;120:103983. doi: 10.1016/j.jtrangeo.2024.103983
39. 39 Li Y, “David” Fan W. Optimizing Transit Equity and Accessibility of the City of Charlotte, North Carolina, by Integrating Transit Gap Index, a General Transit Feed Specification Data-Relevant Performance Metric. J Transp Eng, Part A: Systems. 2021;147(4):04021005. doi: 10.1061/jtepbs.0000503
40. 40 Liu L, Kar A, Tokey AI, Le HTK, Miller HJ. Disparities in public transit accessibility and usage by people with mobility disabilities: An evaluation using high-resolution transit data. Journal of Transport Geography. 2023;109:103589. doi: 10.1016/j.jtrangeo.2023.103589
41. 41 Ribeiro J, Fontes T, Soares C, Borges JL. Accessibility as an indicator to estimate social exclusion in public transport. Transportation Research Procedia. 2021;52:740-747. doi: 10.1016/j.trpro.2021.01.019