Applications

Accessibility indicators computed with the methods in this wiki are used across several domains.

National, European and global network models

• Dutch national network model (PBL), nationwide travel times for car, bicycle, pedestrian and public transport, with published methodology. It underpins the PBL reports Toegang voor iedereen? (2022) and Beter bereikbaar? (2024).
• Worldwide network model, a global car network built per continent from OSM, combined with the GHSL population grid and settlement layer. It produces settlement hierarchies, catchment areas, travel times and distances to settlements by rank, functional rural areas, and traffic flow proxies, and feeds the CRISP model for land use downscaling.
• European network model, travel time analysis across European countries combined with a service allocation procedure, used to study school and service coverage, scale efficiency and network restructuring.
• Planmonitor NOVI, accessibility indicators for national spatial planning policy.

Land use modelling

Accessibility measures are calculated and used as driving factors in land use modelling, where the attractiveness of a location for a given activity depends on what it can reach.

Social domain

In our projects in the social domain, distances over the road network are calculated from dwellings to facilities such as supermarkets, care centres and public transport stops. These distances are suitability factors for housing for people with mobility impairments, where the difference between a straight line and the actual walking route is decisive. See Euclidean versus network distance.

Datasets

PC4 travel time matrix for the Netherlands

Car travel times in minutes between all Dutch PC4 postcode zones, roughly 4,000 zones and over 16 million origin destination pairs. It was computed for the VPRO multimedia project Nederland van Boven (2011) and updated in 2019, and has since been used in national accessibility analyses in the press.

Method notes worth knowing before using it:

• The 2011 version was built entirely on OSM. The 2019 update uses the NWB for the Dutch road network, which gives a more completely connected car network, supplemented with OSM roads in Belgium, Nordrhein-Westfalen and Niedersachsen, because those affect fastest routes between Dutch zones. For walking distances we would still recommend OSM, which includes footpaths.
• Ferry connections to the Wadden islands were added by hand, since without them the islands form separate subnetworks. Remaining disconnected fragments, such as service roads and isolated segments on industrial estates, were dropped.
• Speeds are assigned per road type, and passing a junction adds a penalty that grows with the number of segments meeting there. See Travel speeds per mode.
• Each PC4 zone is represented not by its geometric centroid but by the weighted centre of its BAG residential objects, which for accessibility questions is the more meaningful point. See Zoning and spatial units.

Downloads, both under CC BY-SA 3.0 NL:

• Matrix and table, 29 January 2019 (zip, 88 MB, two CSV files in matrix and table form)
• Original table, 6 December 2011 (zip, 55 MB, one DBF file)

Other

• Road distances from every BAG building to the nearest primary school, made available to EduGIS.

More

See objectvision.nl, accessibility for the full project and publication overview.