It seems hard to lose track of an entire city. But that appears to be what’s taken place – and not just once, but over and over again. The infamous “ghost cities” of China have become a favourite internet meme of the past half-decade. These ghost cities are meant to be sprawling wastelands of empty streets and uninhabited megastructures, without a human being in sight. But for all the discussion, do these places really exist?

A Chinese ghost city should be easy to find. After all, it is not just a failed development, limited to one or two buildings in a quiet neighbourhood of an existing metropolis. It is an entire human settlement, built with government support, for a population of millions who – for whatever reason – have yet to arrive.

Rumour has it that China is full of these empty shells. But where they actually are poses a more complex question than might be expected. And the hunt for China’s ghost cities reveals things about urban development the world over.

Blind spots

Last month,researchers funded by Chinese search giant Baidu strode into the field. Their approach was surprisingly simple. By tracking the location data of Baidu’s hundreds of millions of daily users, they could identify the places with the heaviest internet traffic. This also revealed locations with no – or very little – online activity. Were these the ghost cities?

The researchers screened their results for seasonal variation. They omitted tourist towns that experience a population boom half of the year and fall silent for the rest, for example. In the blind spots of user data that remained were more than 50 vast empty regions in China that were also recent sites of significant development.

Previous investigations of habitation levels have been done by counting the number of homes with lights on at night, for example. Using internet data from hundreds of millions of people gives a more accurate picture. The same technique might even be used to survey the US rust belt. But none of these approaches address the question of why ghost cities exist in the first place.

This has not been the only attempt this year to seek out China’s ghost cities. A company called Orbital Insight in Palo Alto, California, made headlines a few months earlier for using high-resolution satellite imagery to do real-time economic forecasting. They offer what is in effect private space-age espionage: the company analyses satellite imagery in great detail and sells what they find to global investors.

This includes measuring the shadows cast by Chinese construction sites to verify whether or not a building project is still under way or if it has stalled. Measuring shadows from one week to the next can indicate whether or not the Chinese housing market has become saturated, for example.

Close to home

Orbital Insight’s analyses also suggest that we should be looking at ghost cities nearer to home – frozen construction sites, uninhabited neighbourhoods, failed building units. Other firms, including Spaceknow, Planet Labs and BlackSky Global, are also using hundreds of thousands of satellite images, taken daily, to look for details that could be turned into useful data. Even the presence of vehicles in car parks has economic implications, indicating that a metropolitan area is healthy and functioning.

Kai Caemmerer is an architectural photographer at Columbia College in Chicago. He has a different perspective on these questions. For the past year, Caemmerer has been documenting sites of underpopulated architectural development throughout China. He refers to these places as “unborn cities”. This is a deliberate flipping of the ghost cities metaphor: whereas “ghost city” implies a metropolis whose time has come and gone – an abandoned city – Caemmerer points out that these cities are yet to see life.

Caemmerer has visited the Kangbashi New Area in Inner Mongolia, the Meixi Lake area near Changsha, and Tianjin’s Yujiapu Financial District, all labelled as ghost cities. What he found are real estate developments simply waiting for people to arrive. The cities are deliberately built to the point of near completion before inhabitants move in, he says.

Different model

Calling them ghost cities reveals a misunderstanding, says Caemmerer. “It fails to recognise that they are built on an urban model, timeline, and scale that is simply unfamiliar to the methods of Western urbanisation.”

Nevertheless, Caemmerer’s photographs are eerie and astonishing. Ominous, unlit towers loom over empty business districts. Blank lakes reflect the orderly facades of uninhabited housing blocks in an orange glow of sunset. Concrete megastructures lurk in the polluted haze of construction sites as streetlights turn on in the distance, illuminating nothing. The effect is almost picturesque.

Caemmerer’s interest in China is as much about debunking the myth of ghost cities as it is about showing how ubiquitous such places are. “I’m interested in what happens to the urban landscape when those who it was built for are not present,” he says. This includes cities in the US, such as his hometown of Chicago, which also becomes “a ghost town from the right angle”.

For all the statistical analysis and computational firepower being unleashed by big data to find empty places on the far side of the world, we would do well to look a bit closer to home, at the unfinished landscapes of our own urban centres and the economic lessons we might draw from them.

Seeking out ghost cities is about looking ahead, says Caemmerer. These empty places are not built for the present. “I’m interested in what the architecture of these cities says about the future they’re expecting.”