What counts as a ‘city’ in the 2018 map?

The study used densely populated urban areas above a population threshold, derived from global settlement data, rather than administrative city limits. That is why small towns do not count as destinations.

How accurate are travel-time maps?

They are good for comparing regions and spotting patterns, but individual values can be off because travel speeds are assumed, road data may be incomplete and conditions change. They describe 2015 infrastructure.

Can I make my own travel-time map?

Yes, at a small scale. Several route planners and open-source tools can draw isochrones, areas reachable within a set time, for walking, cycling, driving or public transport from a chosen point.

Travel time to cities: accessibility maps explained

Q: Where can I find the original 2018 accessibility map?

The study is published in Nature under doi:10.1038/nature25181. The underlying data layers are distributed by the Malaria Atlas Project. This page is an independent explainer and does not host the original map.

Looking for the interactive accessibility map that used to be on this page? It is no longer hosted here. This page is an independent explainer written by the editors of Roadless Forest Guide. We are not connected to the researchers, institutions or funders behind the original map, and we link to their published sources below.

A travel-time map, also called an accessibility map, shows for every point on land how long it would take to reach the nearest city, using whatever is available: roads, railways, rivers and, where there is nothing else, walking across open land. It turns the vague idea of ‘the middle of nowhere’ into something you can measure. The best-known global version was published in Nature in 2018 and estimated that roughly four in five people on Earth lived within an hour of a city.

The 2018 study in brief

The study ‘A global map of travel time to cities to assess inequalities in accessibility in 2015’ (Weiss et al., 2018, Nature 553, doi:10.1038/nature25181) combined large road datasets, including OpenStreetMap and Google road data, with information on railways, navigable rivers, land cover, slope and borders. From this it calculated the travel time from every square kilometre of land to the nearest city in 2015.

Its headline findings:

Finding	Value (2015)
Share of world population within one hour of a city	about 80.7 %
In high-income countries	about 90.7 %
In low-income countries	about 50.9 %
Spatial resolution	about 1 km

The authors used the map to show how unequally access to cities, and with it access to markets, health care and education, is distributed. The data layers are distributed by the Malaria Atlas Project, which also hosts later related datasets, for example on travel time to health facilities.

How such maps are built

Most travel-time maps follow the same three steps:

A friction surface. Every grid cell gets a travel speed or ‘cost’: fast on motorways, slower on minor roads, slower still on tracks, and very slow across forest, swamp or mountains. Water, borders and steep slopes add further costs.
Destinations. The model needs a definition of what you are travelling to: in the 2018 map, densely populated urban centres above a population threshold, not every village.
Least-cost paths. An algorithm finds, for each cell, the fastest combination of routes to the nearest destination and records the time.

The result is a continuous surface. Dark or ‘hot’ colours typically mark the most remote areas: the interior of the Amazon, the Sahara, the Tibetan Plateau, northern Canada and Siberia.

What the maps do not tell you

Travel-time maps are models. Keep these limits in mind:

Speeds are assumptions. A gravel road may be passable at 60 km/h or at 10 km/h depending on season and vehicle.
Road data are incomplete. Missing roads make places look more remote than they are; abandoned roads make them look closer.
Snapshot in time. The 2018 map describes 2015. New roads, closed borders and seasonal conditions change the picture.
One mode, one destination. The maps measure time to the nearest city, not to a hospital, a train station or your home.
Remoteness is not wilderness. A place can be far from a city but still crossed by forestry tracks.

Europe through the lens of travel time

On a continental scale Europe is one of the most accessible regions of the world. Large areas where the nearest city is many hours away are concentrated in northern Scandinavia and Finland, the Icelandic interior, the Scottish Highlands and parts of the Balkans and Carpathians. Our guides to the most remote places in Europe and Germany’s most remote spots explain which criteria make sense on a smaller scale.

The link between access and forest is direct: where roads go, logging, hunting and settlement follow. We look at that in why roads drive forest loss.

Using travel-time thinking for your own trips

You do not need a global model to find quiet places. A few practical ideas:

Draw isochrones. Several route planners and open-source tools can show the area you can reach within, say, 30 or 60 minutes on foot from a car park or a station. The edges of these shapes are often the quietest places.
Look at the gaps in the road network. On a topographic map, areas without any roads or tracks are rare in Europe and worth a closer look.
Combine with public transport. A trailhead that is two hours by bus from the nearest town is, for most visitors, further away than one 20 minutes by car. Our guide to reaching wild forests without a car helps with planning.
Plan for no signal. Remote by any definition often means no mobile coverage. Download maps before you go: see offline maps and navigation.

Sources

Weiss et al. (2018): A global map of travel time to cities to assess inequalities in accessibility in 2015. Nature 553, 333–336. doi:10.1038/nature25181
Malaria Atlas Project: data and maps