"Long before it's in the papers"
March 19, 2012

RETURN TO THE WORLD SCIENCE HOME PAGE

Spotting ancient sites from space

March 19, 2012
Courtesy of Harvard University
and World Science staff

An ar­chae­o­lo­gist and a com­put­er sci­ent­ist say they have greatly sim­pli­fied the pro­cess of find­ing early hu­man set­tle­ments, by har­ness­ing com­put­ers to scour sat­el­lite im­ages.

The re­search­ers claim to have there­by un­cov­ered thou­sands of new sites that might re­veal clues to the ear­li­est com­plex hu­man so­ci­eties.

Har­vard Uni­vers­ity ar­chae­o­lo­gist Ja­son Ur and Mas­sachus­sets In­sti­tute of Tech­nol­o­gy com­put­er sci­ent­ist Bjo­ern Menze de­vel­oped the sys­tem, which iden­ti­fies set­tle­ments based on an ar­ray of fac­tors in­clud­ing soil dis­col­ora­t­ions and a dis­tinc­tive mound­ing left be­hind af­ter mud-brick set­tle­ments col­lapse.

Ur used the tech­nique to ex­am­ine sat­el­lite im­ages of a 23,000 square-kilometer (9,000 square miles) ar­ea of north­east­ern Syr­ia. He said he turned up about 9,000 pos­si­ble set­tle­ments—an in­crease of at least ten­fold over what was pre­vi­ously known. The find­ings are de­scribed in a pa­per pub­lished March 19 in the jour­nal Pro­ceed­ings of the Na­t­ional Acad­e­my of Sci­ences.

With con­ven­tion­al meth­ods, Ur said, “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.

“What’s more, an­y­one who comes back to this ar­ea for any fu­ture sur­vey would al­ready know where to go,” he con­tin­ued. “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.”

* * *

An archaeologist and a computer scientist say they have greatly simplified the process of finding early human settlements, by harnessing computers to scour satellite images. The researchers claim to have thereby uncovered thousands of new sites that might reveal clues to the earliest complex human societies Harvard University archaeologist Jason Ur and Massachussets Institute of Technology computer scientist Bjoern Menze developed the system, which identifies settlements based on an array of factors including soil discolorations and a distinctive mounding left behind after mud-brick settlements collapse. Ur used the technique to examine satellite images of a 23,000 square-kilometer (9,000 square miles) area of northeastern Syria. He said he turned up about 9,000 possible settlements—an increase of at least tenfold over what was previously known. The findings are described in a paper published March 19 in the Proceedings of the National Academy of Sciences. With conventional methods, Ur said, “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. “What’s more, anyone who comes back to this area for any future survey would already know where to go,” he continued. “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.”