The US road network of the is one of the largest human-made constructions on Earth. The 4 million mapped roads take up about 1 per cent of the nation's land area – about the size of South Carolina. So what would a country look like if the only information you had about it was its road network – no coastal outlines or mountain ranges allowed?
This is the kind of question that fascinates data manipulator and computational designer Ben Fry, who is based in Cambridge, Massachusetts. Using 2006 US census data, he built such a map.
He found out that you can't suppress the geographical features: they appear as the roads avoid them, and you can get a more immediate idea of the population density of a region just by looking at its roads than by studying its population data. He says the roads also give insights into the history of a place: "You can see an incredible range, from the highly regular roads around Kansas City, for instance, to the haphazard, like immediately outside the Bay Area on the West Coast," which reflects how the network was built up over time.
It wasn't just physical insights Fry gained from his map, but also some "interesting titbits" about the American psyche. For example, the top 10 street names are pretty much the same in every state, and roads are often named after trees even when no specimens exist for hundreds of miles: "'Magnolia' is such a nice-sounding word, town developers don't mind that it has no local relevance," he says.
The US Geological Survey (USGS) has interpreted similar information in another way. Using census data from 2000, they have created a map of the country, coloured according to the distance from fixed points to the nearest road, with data points every 30 metres. The green patches indicate areas where the nearest road is 10 to 20 kilometres away, and the yellow spots where a road can be found less than 100 metres away, with a spectrum of blue hues in between.
On both representations the areas with the most roads tend to appear in cities, but exceptions include the oil and gas fields of west Texas, where dense road networks have been built to export the raw materials. The areas with the least roads correspond to inaccessible landscapes or places where there is little economic motive to build them: for example, the sand hills of Nevada, the deserts of Arizona, New Mexico and California, the swamps of southern Florida and the steep slopes of the Rocky Mountains.
Fry's map was put together mainly out of curiosity, but the USGS says that its map will be used as a resource for "comparing landscape patterns manipulated by humans with natural patterns".