Using Space Satellites to Spot Ancient Cities
Computer analysis of satellite imagery has revealed what could be a record number of archaeological sites
Archaeologists have traditionally found the remains of ancient settlements using a number of labor-intensive techniques: ground surveying, analysis of historic texts, surveys of local residents and the occasional stroke of dumb luck.
These methods have yielded a huge amount of information about ancient human societies, but the process of discovery has occurred in spurts, limited by financial resources and the sheer amount of time researchers had to spend on the ground. Before they could begin digging up a site, it could take months or years to even find one.
A new technique is rapidly turning this paradigm on its head: Instead of getting close to the ground, people are increasingly finding archaeological treasures by looking down from space. And in a new study, Harvard social scientist Jason Ur and MIT research affiliate Bjoern Menze announce the development of a computer program that systematically analyzes satellite images to identify likely locations of ancient artifacts. Their paper, published yesterday in the Proceedings of the National Academy of Sciences, included the analysis of a 23,000-square-kilometer area in Syria and turned up a remarkable 9,000 or so possible settlements.
The potential of this new approach to speed up archaeological discovery is massive. “I could do this on the ground,” says Ur, “but it would probably take me the rest of my life to survey an area this size. With these computer science techniques, however, we can immediately come up with an enormous map which is methodologically very interesting, but which also shows the staggering amount of human occupation over the last 7,000 or 8,000 years.”
The 9,000 possible sites are at least ten times the number of settlements identified previously. The surveyed area is in the northeastern part of Syria, representing part of the Fertile Crescent, home to some of the oldest permanent human settlements in existence, dating as far back as 7,000 B.C.
The analysis program found them by carefully scrutinizing satellite images. Some were detected via radiation from the infrared and near infrared parts of the spectrum, which can show lighter soils with more organic materials, resulting from ancient mud buildings and human settlements turned to dust. Others were identified via artificial mounds, known as Tells, which indicate a series of settlements built atop one another over time.
The researchers then used the volume of settlement sites in a given area as a proxy for continued occupation, and sought to find trends between the placement of settlements and natural resources, such as fresh water. They suggest that complete mapping of sites in the Fertile Crescent will uncover long-term patterns in human settlements during ancient history.
Ur predicts that the new technique will accelerate the pace of archaeological discovery—not just in the Fertile Crescent, but in many other areas around the world. ”Anyone who comes back to this area for any future survey would already know where to go,” he says. “There’s no need to do this sort of initial reconnaissance to find sites. This allows you to do targeted work, so it maximizes the time we have on the ground.”