The following report describes the workflow for calculation of indicator 11.2.1 within the framework of the GEOSTAT 3 project, work package 2, by Statistics Portugal. Due to the lack of public transportation data for the whole country, the calculation of the 11.2.1 indicator focuses only on the Lisbon Metropoltan Area (AML). AML is constituted by18 municipalities that together represent around 3,3% of the national territory and has about 3 million inhabitants (approximately 25% of the Portuguese popuation).
Figure 1: Portuguese Mainland Municipalities and Lisbon Metropolitan Area (AML)
The following data sources have been used:
- Census 2011 population data.
- Urban centers, urban clusters and rural areas from the Global Human Settlement Layer (GHSL) 2011.
- Urban centres, urban clusters and rural areas delimited by Statistics Portugal from national data, the 2011 census localities, and following the GHSL methodology and criteria – aggregation of contiguous localities based on their population density and built area.
- Data on officially recognized public transportation for the Lisbon Metropolitan Area (AML). The public transportation data includes the location of stops and the routes for all transports means (bus, train, metro/subway and tagus river boats) within AML. Transportation data was made available to Statistics Portugal by the AML, the transportation and mobility authority for the metropolitan area.
The steps to calculate the 11.2.1 indicator are described in detail below:
a) Prepare geocoded population data
To calculate Indicator 11.2.1 the Census 2011 population was used. All Census 2011 data was collected through a point based framework – geocoded to Statistics Portugal buildings points geodataset (BGE). BGE is linked, by a common id, to the Statistics Portugal Buildings and Dwellings register (FNA) and to the Census 2011 microdata database.
BGE is a point based geodatabase that was created at the Census 2011. BGE contains the X and Y coordinates of all the buildings that, at the Census 2011 reference date (21/03/2011), had at least one residential dwelling. FNA is Statistics Portugal Buildings and Dwellings register and contains, among other characteristics, the address of all the buildings and dwellings collected at Census 2011. Census 2011 microdata database contains all the data collected at the census operation. Since 2011 BGE and FNA are being continually updated by Statistical Portugal surveys and by external sources like the new buildings permits and concluded building permits that are licensed by municipalities and reported to INE, including the X and Y coordinates of each new and demolished building, on a monthly basis through Indicators System of Urban Operations (SIOU).
To calculate the 11.2.1 Indicator, AML population data from the census 2011 microdata database, by gender and age groups, was obtained through the spatial intersection between BGE buildings points (restricted to the existing buildings at the Census 2011 reference date) and AML municipalities from Portugal Official Administrative Boundaries Map (CAOP).
b) The Urban Centers, Urban Clusters and Rural areas typology of Statistics Portugal:
In order to calculate the 11.2.1 indicator, and compare the results with the results obtained from GHSL Layers, Statistics Portugal also delimited urban centres, urban clusters and rural areas from national data, specifically from the Statistics Portugal Census 2011 localities.
To this purpose the GHSL methodology and criteria where applied to the Portuguese localities and the urban centers and urban clusters where created by the aggregation of contiguous localities based on their population density and built area.
After testing and analysing the definition of a urban-rural typology delimited from national data by selecting individual localities based only on their population number without any aggregation – the “nº of population method” (> 50.000 inhabitants for urban centers and > 5.000 inhabitants for urban clusters) – or by applying to the same national data the GHSL methodology.
AML, like many other regions from Mainland Portugal (especially coastal regions) and despite their urban continuum, are retailed in several different localities that, for themselves individually, don’t reach the number of inhabitants required to properly compare them with the GHSL layers. For this reason the urban-rural typology delimited from national data following the GHSL methodology was the selected one.
The following images show the urban centers and urban clusters of AML delimited from the Portuguese localities with the GHSL method (in red) and from the plain selection of Portuguese localities over 50.000 inhabitants and 5,000 inhabitants (yellow area).
Figure 2 – AML – Urban Centers (Portuguese Localities) – nº of population method Vs GHSL method
Figure 3 – AML – Urban Clusters (Portuguese Localities) – nºof population method Vs GHSL method
The following images show the urban centers and urban clusters from the GHSL layers (green area) and from the Portuguese localities applying the GHSL method (in red).
The GHSL method applied to the national data reflects a high degree of accuracy between Statistics Portugal urban-rural typology and the GHSL layers and allows a more reliable comparison of results between these two data sources.
Figure 4 – AML – Urban Center – GHSL layers and Portuguese Localities GHSL Method
Figure 5 – AML – Urban Cluster – GHSL layers and Portuguese Localities GHSL Method
c) Selection and preparation of public transportation stops:
The public transports dataset that was used was made available to Statistics Portugal by the AML – the public transport and mobility authority for the metropolitan area. The transports dataset includes the location of stops and the routes for all transports means (bus, train, metro subway and tagus river boats) within the AML.
The public transport dataset metadata and the transports timetables were not made available so all the stops from all means of transport where considered with the exception of the stops from the Western train line and Trafaria- belém peers from the Tagus river boat connection who’s lack of offer are publicly known.
For stops with access to persons with disabilities there were only considered:
- Bus stops from Lisbon municipality. Lisbon’s transport company (CARRIS) has fully adapted buses for disabled persons, with lowered doors and places for weelchairs;
- Metro/Subway stations with lifts;
- All the trains stations (the train company assures access to all disabled persons);
- All the Tagus river boats peers.
d) Computation of service areas
The service areas were computed, for each mean of transport individually, using a Euclidian distance buffering operation. The 500 meters buffers around each stop/station/peer were dissolved creating a single feature and removing existing overlaps.
Buffers for stop/station/peer with access for the disabled were also created separately for each mean of transport and the same dissolve method was applied.
e) Calculation of the population within urban-rural typology and public transport service areas:
After obtaining AML population by building point of residence, after the delimitation of the different area polygons from the urban-rural typology based on Portuguese localities (with the GSHL method) and after the creation of the different 500 meters buffer polygons of public transport stops/stations/peers (including for the disabled) a spatial join between the buildings points and each area polygon was applied. As a result of this geoprocessing process each building was classified, in a master table, with a simple yes/no boolean attribute indicating it’s spatial relationship with the different polygons from the urban-rural typology and public transports buffer areas.
The Master table was then used to calculate the final results for 11.2.1 Indicator.
Table 5: master table with the population resulting from the spatial relation between buildings (buildings codes) and the different urban-rural areas typology and tranport buffer areas
As a result from the processes described above the table in below show the results of Indicator 11.2.1, the number and share of population with access to public transport, by sex, age groups and access to persons with disabilities for the both, GHSL and Portuguese localities, urban-rural typology.
Analysing the results we can conclude that the share of the urban population with access to public transportation in the AML is high and reaches 98,30%.
The values for urban centers or urban clusters from the GHSL layers or the urban-rural typology delimited from national data (using the GHSL method) are also very high (almost 100%) and similar between the two datasets. For urban centers, the values are around 99,80% and for urban clusters around 99,10%. At the rural areas the differences between the two datasets slightly differ. For the GHSL layers the value is 81,26% and for the Portuguese localities the value rises to 85,09%.
The share the different genders, male or female, and for all age group is quite similar to the overall population for all urban-rural areas.
As expected, transports that are accessible for disabled persons are only available for around 33% of the AML population. Acessability to public transports by disabled persons rises to 40% in urban centers and to 34% in urban clusters but in the rural areas it almost doesn’t exist and less than 1% of the population has access to public transport that are suitable for disabled persons.
Table 6: Number and Share of population with convenient access to public transportation, disaggregated by sex, age and persons with disabilities
For Statistics Portugal, the general conclusions that arise from the production of 11.2.1 indicator are:
- The availability, at statistics Portugal of a point based location of statistical units for geocoding data enables the establishment of spatial relations between statistical data and thematic territorial units that allow the production of new statistical territorial indicators;
- Due to the unavailability of public transport data with national coverage, from a official or trusted provider, it was not possible to produce 11.2.1 indicator for the whole of Portugal;
- Due to the unavailability of public transport timetables it was not possible to fully comply with GEOSTAT 3 requirements for the production of 11.2.1 indicator. By this reason the 11.2.1 indicator produced by statistics Portugal has to be understood more as a potentially convenient access to public transports;
- Due to the inexistence of a streets network, it was not possible to calculate 11.2.1 indicator by the method of Network distance measurement;
- The results reflect the implementation level of GSGF in Portugal namely at the bottom of the framework – principle 1 and principle 2.