This map has the extremely ambitious objective of representing the “accessibility” of cities. As its legend indicates, the chosen magnitude is, at any point on the Earth’s surface, the time of travel to the nearest city. I drew it in a projection designed by John Paul Goode in 1923, using colors reminiscent of relief, and shadows indicating changes in magnitude as if they were variations in altitude. The eye accustomed to topographic maps naturally associates the steeper or lower slope to the cost of “climbing” to reach this point from a city (& vice versa). Urban areas form the lowlands, dark green in color, and desert, inaccessible or very remote areas are represented by ever-higher mountains as one moves away from the cities.
The data do not come from actual measurements taken on site, but from a model, published in the scientific journal Nature by the research team of the Malaria Atlas Project, Oxford University, under the direction of Dr. Daniel Weiss.
This model is based on a large number of databases: for example, in order to assess roads and traffic possibilities, it uses the Open Street Map and Google Maps databases; the world’s population distribution is derived from estimates made by the European Commission’s Joint Research Centre, etc.
As this study argues:
travel time better captures the opportunity cost of travel than Euclidean or network distance, and ultimately reflects the information humans use to inform transport decisions.
The travel times estimated by this model are expressed in minutes. It is easy to understand that a place located twice as far from one city as another, or only reached by a road that is twice as slow, is half as accessible. Each place on the planet therefore has a “coefficient of friction”, which indicates the time it takes to cross 1 km, by land or sea: roads, trains, types of terrain are taken into account in a differentiated way. You obviously don’t travel at the same speed on a track, a road or in the forest. This global matrix of “transport costs”, expressed in minutes per kilometre, is then analysed by an optimization algorithm, which calculates the fastest route to one of the 13,840 urban areas taken into account (in addition to cities with a population of 50,000 inhabitants, the model includes areas of at least 1 km2 where the population density is estimated at more than 1500 inhabitants/km2).
This definition is of course arbitrary, but the method chosen by the authors of the model — available in open source —, makes it possible to set its own criteria, depending on the phenomena we want to highlight. If we want, for example, to have an idea of the level of accessibility of natural parks, maritime coasts, schools, hospitals, or certain markets, the points that we will find in the “lowlands” will be totally different. The calculation will result in another landscape, new plains and new peaks of inaccessibility.
The scientific objective for the Oxford team is to measure inequalities in access to health infrastructure. Their article details how they position the results of health surveys conducted in 52 countries and covering 1.77 million households on this global model to highlight statistical relationships between urbanization, transport networks, poverty distribution, etc., and public health issues. The model is far from being exhaustive, since social inequalities in access to transport must also be taken into account (the same distance of 50 km, for those who have to walk to the hospital, is a more formidable obstacle than for those who can afford a taxi). But it already makes it possible to establish, for example, that only “50.9% of the inhabitants of low-income regions (concentrated in sub-Saharan Africa) live within an hour’s transport from a city, compared to 90.7% of the inhabitants of rich regions”, with all the consequences this can have in terms of access to health care, education, etc.
Beyond this fundamental question, it seems to me that the images created using this tool offer a new way of showing the distribution of human populations throughout the planet. We can compare countries among themselves, observe unexpected cuts and surprising distributions in certain places. Accessibility can be seen as a somewhat abstract “terrain layer” that can be integrated into a geographic information system and overlayed with other information to better analyze it in context. But, like any terrain, it also represents a landscape in itself, which can be contemplated and walked for its own beauty and the daydreams that it brings forth.
↬ Philippe Rivière.