Access to mass rapid transit in OECD urban areas
AAccess to mass rapid transit in OECD urbanareas
Vincent Verbavatz , Marc Barthelemy
September 9, 2020
1. Université Paris-Saclay, CNRS, CEA, Institut de physique théorique, 91191,Gif-sur-Yvette, France.2. Ecole des Ponts ParisTech, Champs-sur-Marne, France.3. Centre d’Etude et de Mathématique Sociales, CNRS/EHESS, 54 BoulevardRaspail, 75006 Paris, France.*corresponding author: Marc Barthelemy ([email protected])
Abstract
As mitigating car traffic in cities has become paramount to abate cli-mate change effects, fostering public transport in cities appears ever-moreappealing. A key ingredient in that purpose is easy access to mass rapidtransit (MRT) systems. So far, we have however few empirical estimatesof the coverage of MRT in urban areas, computed as the share of peopleliving in MRT catchment areas, say for instance within walking distance.In this work, we clarify a universal definition of such a metrics - PeopleNear Transit (PNT) - and present measures of this quantity for 85 urbanareas in OECD countries – the largest dataset of such a quantity so far.By suggesting a standardized protocol, we make our dataset sound andexpandable to other countries and cities in the world, which grounds ourwork into solid basis for multiple reuses in transport, environmental oreconomic studies.
Background & Summary
Motorized transport currently accounts for more than 15% of world greenhousegas emissions . As most humans live in urban areas and two-thirds of worldpopulation will live in cities by 2050 , mitigating car traffic in cities has be-come crucial for limiting climate change effects . Daily mitigating is the maindriver for passenger car use - about 75% of American commuters drive daily - while alternative transport modes such as public transportation networks areunevenly developed among countries and cities . a r X i v : . [ phy s i c s . s o c - ph ] S e p ver the last decades, various attempts to assess the environmental impactof car use in cities have emerged from multiple fields, ranging from econometricstudies to physics or urban studies . A seminal result of transport the-ory, by Newman and Kenworthy , correlated transport-related emissions witha determinant spatial criterion: urban density. Alternatively, Duranton andTurner claimed that public transport services were to unsuccessful in reduc-ing traffic, as transit riders lured off the roads are replaced by new drivers onthe released roads. Such results, however, crucially lack both theoretical andempirical foundations and new research shows that the two maincritical factors that control car traffic in cities are urban sprawl and access tomass rapid transit (MRT).More generally, understanding mobility in urban areas is fundamental, notonly for transport planning, but also for understanding many processes in cities,such as congestion problems, or epidemic spread for example. But what isa good measure of access to transit? Studies have mainly focused on the num-ber of lines or stops , length of the network or graph analysis . Fewworks , however, have considered investigating catchment areas of MRTstations, i.e. looking at the share of population living close to MRT stations,for instance within walking distance. Such conditions have however proved tobe essential in explaining commuting behaviours and mobility patterns .The most detailed definition of such catchment metrics is the People NearTransit (PNT), and originates from a 2016 publication from the Institute forTransportation and Development Policy (IDTP) . It produces a rigorousdataset of the share of population living close to transit (less than 1 km) for 25cities in the world (12 in OECD countries). However, definitions of urban areasand rapid transit systems in that dataset are multiple and need to be refinedwhile the number of cities must be expanded.Hence, in order to expand our global knowledge of urban mobility, we needa common, unified and universal definition of access to public transit as well assound measures of such a quantity. In this paper, we clarify its definition andpropose what is to our knowledge the largest dataset of PNT globally.Our analysis uses functional urban areas (FUA) in OECD countries, a con-sistent definition of cities across several countries . We restrict our measuresto mass rapid transit, usually referring to high-capacity heavy rail public trans-port, to which we added light rails and trams. In our sense, mass rapid transitthus encompasses: • Tram, streetcar or light rail services. • Subway, Metro or any underground service. • Suburban rail services.uses are not comprised in that definition. In contrast with , we do not excludeany form of commuting trains based on station spacing or schedule criteria. Aswe detail it in the Method section, we identify services and corresponding stopswith the General Transit Feed Specification (GTFS), a common format for pub-lic transportation schedules and associated geographic information .Crossing open-access information from public transport agencies in OECDurban areas with population-grid estimates of world population , we publishhere a list of 85 OECD cities (see Fig. 1) for which we were able to computethe People Near Transit (PNT) levels defined as the share of urban populationliving at geometric distances of 500 m, 1,000 m and 1,500 m from any MRTstation in the agglomeration:PNT ( d ) = population s. t. cartesian minimum distance < d total population (1)where d = 500 , , .We display on Tables 1 and 2 the 5 cities with easiest access to MRT (largestPNT) and the 5 cities with scarcest access to MRT (smallest PNT).We also provide for each city the population grid-maps with correspondingMRT access level, i.e. grid-maps of MRT catchment areas at different distanceswith population in each grid. As an example, Fig. 2 shows the 1000 m catch-ment area of MRT stations in Paris. Methods
Residential populations for FUA
Our analysis relies on the 2015 residential population estimates mapped intothe global Human Settlement Population (GHS-POP) project . This spatialraster dataset depicts the distribution of population expressed as the numberof individuals per cell on a grid of cells 250 m long. Residential populationestimates for the target year 2015 are provided by CIESIN GPWv4.10 andwere disaggregated from census or administrative units to grid cells.We downloaded population tiles that cover land on the globe in the Moll-weide projection (EPSG:54009) and in raster format (.tif files). These rasterdata are made of pixels of width 250 m with associated value the number of peo-ple living in the cell. We processed the downloaded tiles with Python and package gdal to convert the raster files into vectorized shapefiles.The resulting shapefiles are comprised of polygons with field value the popula-tion in each polygon. Since the polygonization process merges adjacent pixelswith common value into single polygons, populations for each polygon must beecomputed from polygon area and density through the simple following rule:Pop polygon = Pop pixel × Area polygon
Area pixel (2)where Area pixel = 250 ×
250 = 62 500 m . This leaves us with a list of 224shapefiles of population that cover land area on earth.By intersecting the resulting shapefiles with OECD shapefiles delineatingFunctional Urban Areas (FUA) in OECD countries (reprojected into Moll-weide projection), we can build a population-grided dataset of cities in OECDcountries.These resulting files are the population substrates used for measuring pop-ulation living close to MRT stations. Extracts of MRT stations from GTFS files
A common and de facto standard format for public transportation schedulesand associated geographic information is the General Transit Feed Specification(GTFS) .A GTFS feed is a collection of at least six CSV files (with extension .txt)contained within a .zip file. It encompasses general information about transitagencies and routes in the network, schedule information such as trips and stoptimes and geographic information for stops (geographic coordinates).The three main objects we require are: • Routes: distinct routes in the network of a certain type. A route is a(one-direction) regular line, for instance a metro or bus line. The routetypes we use are : – Tram, Streetcar, Light rail. Any light rail or street level systemwithin a metropolitan area. – Subway, Metro. Any underground rail system within a metropolitanarea. – Rail. Used for intercity or long-distance travel. – Cable tram. Used for street-level rail cars where the cable runs be-neath the vehicle, e.g., cable car in San Francisco.Our definition of MRT excludes bus and ferry types: – Bus. Used for short- and long-distance bus routes. – Ferry. Used for short- and long-distance boat service.
Trips: trips are associated to a route and define a particular and scheduledtrip between specific stations. For instance, the first train of the day is atrip. • Stops: stops are geographic locations of the stops, stations and theiramenities within the transit system. Stops are organized into a parentstation and their amenities (e.g. platforms or exits).Joining in this order the four tables routes.txt , trips.txt , stop_times.txt and stops.txt lets us bind stops with their associated route types. We canthus discriminate between bus stops and metro stops and thereby limit to ourdefinition of MRT.In a nutshell each GTFS file can be processed to produce localized androute-typed stops. Measure of People Near Transit (PNT)
In order to measure PNT within urban areas, we must bind transit systems withtheir respective FUAs. We need to retrieve - and merge - all available GTFSfiles pertaining to a specific urban area and make sure that no rapid transitagency is excluded in the process.Most GTFS files for cities in the world are collected by the OpenMobility-Data platform . For each city in our dataset, we cross-checked the OpenMobil-ityData with Wikipedia local network information to ensure that we consideredall agencies of rapid transit within the urban area.For some European countries (Germany, France), GTFS files were not availaibleon OpenMobilityData and have been retrieved from other sources . We alsonote that GTFS format is not common in South Korea, Japan and in the UnitedKingdom where we only found GTFS data for Manchester area on OpenMobil-ityData while we directly used station coordinates for London .We were thus left with a list of 85 urban areas in the world for which wehad complete, reliable and extensive data. From route-typed stop coordinateswithin that dataset, we can extract MRT stops (excluding buses and ferries) andbuffer - still using gdal - catchment areas for several distance thresholds: 500 m,1 000 m and 1 500 m. Intersecting the resulting buffers with the population-grided shapefiles gives us the total population living within catchment areas,that can be expressed as a share of the total urban area population resulting inthe value of the PNT metric. Our results are shown in Online-only Table 3. Code availability
Detailed code generating the database can be accessed from the source codehosted via
Gitlab . ata Records The Data Record of PNT in OECD urban areas is available online on
Figshare .PNT levels at distance thresholds: 500 m, 1 000 m and 1 500 m for the 85Functional Urban Areas are shown on the Online-only Table 3. The list of transitagencies for each city is online along with PNT statistics ( mrt_access.csv ) .We also provide, for each city, grid-maps of population at different distancesfrom MRT( pops_close_to_MRT.zip ) .The Tables read as follows: Basel urban area has 528811 inhabitants, ofwhich 57.78% live within 500 m of a MRT station, 80.15% within 1 000 m and86.96% within 1 500 m. Technical Validation
The most thorough and exhaustive measure of PNT in urban areas in existingliterature is a 2016 report from the Institute for Transportation and Develop-ment Policy . To validate our results and our methodology, we compared themwith those results.Out of the 12 OECD cities considered in , 11 are in our dataset: 5 in theUnited States, 2 in Spain, 1 in Canada, 1 in France, 1 in the United Kingdomand 1 in the Netherlands (see Table 4). Unfortunately, we found no data in theremaining city: Seoul.Out of these 11 cities, we had at first glance similar results for only twocities: Chicago (13% for both) and Vancouver (19% vs 23%). The discrepan-cies observed for the other cases stem from different definitions of cities andfrom the different transit systems that were taken into account. While we workwith Functional Urban Areas (FUA) only, the authors of mix two differentdefinitions of cities: FUA and urban cores. By applying our method to urbancores and not functional urban areas, we found the same or similar results forBarcelona, Madrid, Rotterdam and Washington (see Table 4).Also, the authors of considered a definition of the LRT (Light Rail Tran-sit) and Suburban Rail that depends on station spacing and schedule criteria.We didn’t choose this definition and for Boston and New York, we had thereforeto exclude suburban trains - while keeping the definition of FUA - in order toretrieve results similar to those of Table 4. In contrast, the study took intoaccount the Bus Rapid Transit for Los Angeles, that we decided to exclude.inally, in Paris the authors of considered that the so-called RER trains werecomprised in Suburban Rail, but not Transilien trains, while we included bothsystems in our analysis.The conclusion here is that for similar definitions for cities and transit sys-tems, we obtain similar results, validating our method and calculations. Tofacilitate comparison across future studies, we would recommend using the defi-nition of cities given by Functional Urban Areas since it is very commonly usedand already unified for OECD countries. Concerning transit systems, we thinkthat it is more relevant and also verifiable to consider transit systems based ontheir types (Rail versus Road) rather that on spacing and schedule criteria thatare specious and less universal. Hence, in comparing our results with resultsfrom the IDTP report and after checking on Table 4 that our methodology iscorrect, we decided to keep our unmodified estimations for the considered cities,despite the discrepancies with .For other cities in the dataset we have unfortunately found no existing datato compare with. Thus, we hope for future research to test and expand ourestimations and results. Usage Notes
Easy code and hints are given on
Gitlab .We strongly recommand using GDAL to handle geographic data with Python. Acknowledgements
VV thanks the Ecole nationale des ponts et chaussées for their financial support.This material is based upon work supported by the Complex Systems Instituteof Paris Ile-de-France (ISC-PIF).
Author contributions
VV and MB designed the study, VV acquired the data, VV analyzed and inter-preted the data, VV and MB and wrote the manuscript.
Competing interests
The authors have no competing interests. eferences [1] Herzog, T. World greenhouse gas emissions in 2005.
World Resources Insti-tute (2009).[2]
United Nations, Department of Economic and Social Affairs, PopulationDivision.
World Urbanization Prospects: The 2014 Revision. HighlightsST/ESA/SER.A/352 (2014).[3] Dodman D. Blaming cities for climate change? An analysis of urban green-house gas emissions inventories.
Environment and Urbanization , 185-201(2009).[4] Glaeser E. L. & Kahn M. E. The greenness of cities: Carbon dioxide emis-sions and urban development. Journal of Urban Economics , 404-418(2010).[5] Oliveira E. A., Andrade Jr. J. S. & Makse H. A. Large cities are less green. Scientific Reports , 13-21 (2014).[6] Newman P.G. The environmental impact of cities, Environment and Urban-ization , 275-295 (2006).[7] U.S. Department of Transportation, Bureau of Transportation Statistics,National Transportation Statistics. Table 1-41 at (2016).[8] Wikipedia contributors. List of Metro Systems, Wikimedia Foundation https://en.wikipedia.org/wiki/List_of_metro_systems (2020).[9] Creutzig F. et al . Global typology of urban energy use and potentials foran urbanization mitigation wedge.
Proceedings of the National Academy ofSciences , 6283-6288 (2015).[10] Pumain D. Scaling laws and urban systems (2004).[11] Barthelemy M. The structure and dynamics of cities.
Cambridge UniversityPress (2016).[12] Newman P. G. & Kenworthy J. R. Cities and automobile dependence: Aninternational sourcebook (1989).[13] Duranton G. & Turner M. A. The fundamental law of road congestion: Ev-idence from US cities.
American Economic Review , , 2616-52 (2011).[14] Buchanan M. The benefits of public transport. Nat. Phys. , 876 (2019).[15] Anderson M. L. Subways, strikes, and slowdowns: The impacts of publictransit on traffic congestion. American Economic Review , 2763-96(2014).16] Litman T. Evaluating rail transit benefits: A comment.
Transport Policy , 94-97 (2007).[17] Baum-Snow N., Kahn M. E. & Voith R. Effects of urban rail transit expan-sions: Evidence from sixteen cities, 1970-2000.
Brookings-Wharton paperson urban affairs , 147-206 (2005).[18] Verbavatz V. & Barthelemy M. Critical factors for mitigating car traffic incities.
PLoS one (2019).[19] Dalziel B.D., Pourbohloul B. & Ellner S.P. Human mobility patterns pre-dict divergent epidemic dynamics among cities.
Proceedings of the RoyalSociety B: Biological Sciences , 20130763 (2013).[20] Balcan D., Colizza V., Gonçalves B., Hu H., Ramasco J.J. & Vespignani A.Multiscale mobility networks and the spatial spreading of infectious diseases.
Proceedings of the National Academy of Sciences , , 21484-21489 (2009).[21] Fouracre P., Dunkerley C. & Gardner G. Mass rapid transit systems forcities in the developing world. Transport Reviews , 299-310 (2003).[22] Gallotti R. & Barthelemy M. Anatomy and efficiency of urban multimodalmobility.
Scientific reports , 1-9 (2014).[23] Gallotti R. & Barthelemy M. The multilayer temporal network of publictransport in Great Britain. Scientific Data , 1-8 (2015).[24] Musso A. & Vuchic V. R. Characteristics of metro networks and methodol-ogy for their evaluation. National Research Council, Transportation ResearchBoard (1988).[25] Gattuso D. & Miriello E. Compared analysis of metro networks supportedby graph theory.
Networks and Spatial Economics , 395-414 (2005).[26] Derrible S. & Kennedy C. The complexity and robustness of metro net-works.
Physica A: Statistical Mechanics and its Applications , 3678-3691 (2010).[27] Marks M., Mason J. & Oliveira, G. People near transit: Improving accessi-bility and rapid transit coverage in large cities,
Institute for Transportationand Development Policy (2016).[28] Singer G. & Burda C. Fast Cities: A comparison of rapid transit in majorCanadian cities (2014).[29] Dijkstra L., Poelman H. & Veneri P. The EU-OECD definition of a func-tional urban area.
OECD Regional Development Working Papers, 2019/11,Éditions OCDE, Paris (2019).[30] GTFS Static Overview. https://developers.google.com/transit/gtfs (2020).31] Florczyk A. J. et al . GHSL Data Package 2019.
Publications Office of theEuropean Union, Luxembourg , ISBN 978-92-76-13186-1 (2019).[32]
OpenMobilityData https://transitfeeds.com (2020).[33]
GTFS für Deutschland https://gtfs.de/ (2020).[34]
Open platform for French public data (2020).[35] Verbavatz, V. Source code.
Gitlab https://gitlab.iscpif.fr/vverbavatz/mrt-access-project (2020).[36] Verbavatz, V. & Barthelemy, M.: People Near Transit (PNT).
Figshare
Dataset https://doi.org/10.6084/m9.figshare.12013020.v4 (2020).[37]
Python Software Foundation . Python Language Reference, version 3.7.6available at .[38] GDAL/OGR contributors. GDAL/OGR Geospatial Data Abstractionsoftware Library,
Open Source Geospatial Foundation https://gdal.org (2020).[39]
Center for International Earth Science Information Network(CIESIN)—Columbia University . Gridded population of the world,version 4 (GPWv4): population density. (2016).[40]
Transport for London . TFL Station Locations available at https://data.london.gov.uk/dataset/tfl-station-locations (2020).[41]
MapTiler, OpenStreetMap contributors . MapTiler Basic and MapTilerTopo. (2020).igure 1: The 85 OECD cities for which we found data are mostly found inEurope and in North America .igure 2: 1000 m catchment areas of MRT stations (in orange) in Paris func-tional urban area (boundaries are in black) . ity Country Population 500 mPNT (%) 1000 mPNT (%) 1500 mPNT (%) Basel Switzerland 528811 57.78 80.15 86.96Bilbao Spain 986042 56.84 76.79 83.52Geneva Switzerland 592893 50.44 74.68 85.07London United Kingdom 11754700 43.09 72.56 85.8Zurich Switzerland 1329898 42.7 68.18 82.09Table 1: Population Near Transit values: Share of population living withincatchment area from a MRT station at thresholds 500 m, 1 000 m and 1 500 m.Top 5 cities with easiest (1000 m) access to MRT. ity Country Population 500 mPNT (%) 1000 mPNT (%) 1500 mPNT (%)
Winnipeg Canada 846133 0 0 0Detroit United States 4263202 0.1 0.18 0.31Houston United States 6706227 0.98 2.28 3.54Miami United Sates 5964846 1.26 3.51 5.65Dallas United States 7294931 1.18 4.05 7.64Table 2: Population Near Transit values: Share of population living withincatchment area from a MRT station at thresholds 500 m, 1 000 m and 1 500 m.5 cities with poorest (1000 m) access to MRT. ity Country Population 500 mPNT (%) 1000 mPNT (%) 1500 mPNT (%)
Adelaide Australia 1368481 11.43 27.9 40.79Amsterdam Netherlands 2766282 21.4 39.14 52.66Athens Greece 3667934 41.73 63.26 74.43Barcelona Spain 4838161 41.65 65.24 77.82Basel Switzerland 528811 57.78 80.15 86.96Berlin Germany 4953645 38.39 63.02 76.13Bilbao Spain 986042 56.84 76.79 83.52Bordeaux France 1176238 21.94 41.5 53.34Boston United States 4167892 13.4 30.69 44.69Bremen Germany 1253514 21.64 38.37 50.62Brisbane Australia 2307430 9.27 25.63 37.82Brussels Belgium 2632048 34.35 43.43 46.6Budapest Hungary 2972657 34.3 48.55 55.79Calgary Canada 1492971 5.7 18.93 31.15Chicago United States 9608320 5.91 13.34 18.45Cologne Germany 1960557 30.43 55.07 68.99Cracow Poland 1392519 22.36 34.41 41.95Dallas United States 7294931 1.18 4.05 7.64Denver United States 2738183 3.03 9.53 17.91Detroit United States 4263202 0.1 0.18 0.31Dresden Germany 1317454 35.47 56.14 66.95Dublin Ireland 1866112 15.96 35.23 50.17Dusseldorf Germany 1541332 33.01 55.86 69.01Edmonton Canada 1324949 3.6 10.8 17.26Florence Italy 770710 15.22 25.25 31.03Frankfurt am Main Germany 2579579 28.35 56.11 72.33Geneva Switzerland 592893 50.44 74.68 85.07Genoa Italy 699462 11.3 26.02 36.9Hamburg Germany 3191585 16.6 39.5 55.71Hanover Germany 1272611 30.89 54.79 65.58Helsinki Finland 1451912 24.39 45.22 56.88Houston United States 6706227 0.98 2.28 3.54Kaunas Lithuania 380048 37.79 51.5 58.29Lausanne Switzerland 410089 32.63 65.0 80.17Leipzig Germany 972864 42.76 60.57 69.03Lille France 1360801 26.07 48.25 61.46Lisbon Portugal 2831367 18.17 41.66 55.96London United Kingdom 11754700 43.09 72.56 85.80Los Angeles United States 17712325 2.26 8.56 15.91Luxembourg Luxembourg 577309 16.1 39.06 55.64Lyon France 1963944 29.56 50.85 63.87Madrid Spain 6615767 33.94 59.62 70.89 ity Country Population 500 mPNT (%) 1000 mPNT (%) 1500 mPNT (%)
Manchester United Kingdom 3298781 16.34 43.57 64.61Marseille France 1779703 17.58 31.23 42.23Melbourne Australia 4466894 23.27 40.75 54.65Mexico City Mexico 20578866 8.8 20.6 28.89Miami United States 5964846 1.26 3.51 5.65Milan Italy 4966888 28.26 48.99 63.0Montreal Canada 4478991 10.88 25.82 37.75Munich Germany 2825789 36.74 62.44 75.31Nancy France 477056 16.42 36.16 52.43Nantes France 908423 21.41 40.08 50.82New York United States 19694439 27.02 45.35 56.68Nice France 848591 20.2 39.87 53.29Oslo Norway 1332133 30.16 47.59 53.67Ottawa Canada 1500455 2.18 7.46 12.03Paris France 12012223 37.16 62.7 77.56Perth Australia 1930198 4.88 15.66 27.01Philadelphia United States 6432106 4.55 14.34 23.81Portland United States 2262652 6.64 15.61 24.75Prague Czech Republic 2251032 35.21 59.46 73.52Rennes France 720142 10.18 23.61 35.69Rome Italy 4161006 25.93 46.97 58.33Rotterdam Netherlands 1823101 20.8 41.13 56.04San Francisco United States 6273368 10.92 25.15 37.74Santiago Chile 7182609 13.0 34.01 49.49Seattle United States 3620117 2.83 6.45 10.12Stockholm Sweden 2221640 34.96 60.65 72.73Strasbourg France 779704 27.56 51.27 63.81Stuttgart Germany 2662983 27.57 51.06 64.64Sydney Australia 4903571 13.91 34.98 50.88Thessaloniki Greece 1076231 1.38 5.62 11.54Toronto Canada 7123826 13.21 23.06 33.72Toulouse France 1330243 14.27 30.02 42.65Turin Italy 1741546 38.38 52.55 61.3Utrecht Netherlands 882821 16.21 38.68 54.86Valencia Spain 1686890 31.04 56.66 71.73Vancouver Canada 2539976 8.52 22.55 34.37Venice Italy 557955 13.98 22.21 27.76Vienna Austria 2779253 44.63 68.02 78.94Vilnius Lithuania 691221 28.55 38.33 43.84Warsaw Poland 3099687 26.7 42.78 50.95Washington United States 8899517 3.12 8.42 12.89Winnipeg Canada 846133 0.0 0.0 0.0Zurich Switzerland 1329898 42.7 68.18 82.09Online-only Table 3: Population Near Transit values: Share of population livingwithin catchment area from a MRT station at thresholds 500 m, 1 000 m and1 500 m for 85 OECD Functional Urban Areas. ity Country Population in Typesin PNTShare(%)in OurPNTShare( % ) OurPNTShare( % )with crite-ria Commentsabout Barcelona Spain 3200000 Metro +LRT 76 65 74 Urban core andnot FUA;suburban trainsare de facto included in Boston US 4650000 Metro +LRT 15 31 17 Excludessuburban trainsChicago US 9500000 Metro 14 13 13 /London UK 10000000 Metro +LRT +SuburbanRail 61 73 / Some suburbantrains areexcluded in Los Angeles US 13000000 Metro +LRT +BRT 11 9 / We exclude BusRapid TransitMadrid Spain 5500000 Metro +LRT 76 60 72 Urban core andnot FUA;suburban trainsare de facto included in New York US 19800000 Metro +LRT 35 45 34 Excludessuburban trainsParis France 12000000 Metro +Tram +SuburbanRail 50 63 / Some tramlinesare excludedin Rotterdam Netherlands 1200000 Metro +LRT 55 41 50 Urban core andnot FUA;suburban trainsare de facto included in Vancouver Canada 2300000 Metro 19 23 23 /Washington US 5800000 Metro 12 8 12 Urban core andnot FUATable 4: Comparison of MRT Share from the IDTP report27