What Genghis Khan Could Teach Us About The Ghost Boat
Using satellite imagery to find the unfindable.
Six years ago, a group of researchers from the University of California, San Diego, shared an obsession. They were desperate to find the tomb of Genghis Khan — the creator of the largest Empire in history, and, in their words, the “most accomplished man to have walked the Earth”.
According to legend, when he died the Mongolian warlord wanted his burial site to be kept secret, so when he died and was interred, everybody who knew the location was summarily executed: First every person on the road was killed by his soldiers, then they murdered everyone who built it, and then they were themselves put to death. (Seems he was also laying claim to being the cruellest man to have walked the Earth.) Since then, many people have looked for it, but his final resting place has remained a mystery.
The UCSD researchers were confident that the tomb was somewhere in a Mongolian valley, but they also knew it was far too big a job for any team of archeologists — even one that comprised hundreds of experts — to realistically find an answer. The valley was gigantic (more than 6,000 square kilometers) and there were, unsurprisingly, significant protests from Mongolians whenever groups of would-be Indiana Jones went gallumphing around digging up the country’s most sacred sites.
So they decided to take a new, more hands-off approach to looking for the tomb: crowdsourcing. By taking satellite photography of the valley, and asking members of the public to examine small swatches and look for geographic features that warranted further attention, they essentially turned the hunt into a game that anyone could join in. If somebody spotted a set of clues from the air that looked promising, a team on the ground could go and — carefully — take a closer look.
Launched in 2010, the search saw more than 10,000 people sift through 84,000 map tiles, contributing the same amount of work in a few days that would have taken a single archaeologist more than three years to complete. They eventually found 55 sites to look more closely at: The whole thing is explored in a National Geographic documentary, and written up in the scientific journal PLoS One. Note: They still haven’t found the tomb of Genghis Khan.
The group turned the system it built into the basis of a new company, Tomnod (it means “Big Eye” in Mongolian), which specializes in creating large datasets and allowing people to sift through the images for clues. Recently, for example, they monitored damage from an earthquake in Chile and are currently trying to do a forensic examination of destruction caused by bush fires in Australia.
Could this approach be helpful for our own search for a boat of missing refugees?
Last week I spent some time talking with Tomnod about the possibility of using satellite photography in a similar way for our own quest. And the answer right now is… maybe.
There are significant challenges. First of all — and most importantly — there is the imagery itself. Tomnod uses photography from a small group of 11 satellites that snapshot any given area. The case we’re looking at is historical, so we’re reliant on whatever photography has already been taken. Unfortunately we can’t just send the satellites to look at what we need, as happens in most of their open projects.
In addition the area we are looking at, the coast of Libya and the open Mediterranean, is not one that is heavily photographed. The battery of satellites do not generally watch open sea —very little happens there, it would be a waste of resources—and what is there is not necessarily from the time that the boat disappeared. Unlike the high resolution shipping data we found on maritime movements in the area, there are not minute-by-minute records that could help us spot and track a vessel.
Still, we’re hopeful. Tomnod was used in the search for the wreckage of flight MH370, and we can take some pointers from that larger project — which, you’ll remember, still hasn’t found anything (despite being an international effort that has received millions of dollars in funding.)
If we keep our search closer to the coast, and we look at images in the weeks after the boat disappeared, there is data to be looked at. It means we’ll have to tweak what we’re looking for, and where.
I’m talking with Tomnod again later this week to see if we can make more progress.
In the meantime, we are — as always — asking for you to contribute. If you know anything about maritime movements, satellite photography, or accessing the data that could help us, this could be your moment.
Onward.
