Drone image of newly documented Nasca lines, along the edge of an archaeological site found by GlobalXplorer° citizen scientists. Image courtesy of Luis Jaime Castillo Butters.

GlobalXplorer° Completes Its First Expedition: What the Crowd Found in Peru

GlobalXplorer° is thrilled to announce the results of our inaugural expedition — the world’s first crowdsourced mapping of archaeological sites in Peru.

Here, we’ll fill you in on some of the insights we gleaned from this ambitious and unprecedented experiment. It was, we’ll admit, a huge risk. Could we demonstrate that our vision was possible? We could, and we did. GlobalXplorer° proved our theory that the crowd can seriously speed up the important but time-consuming task of mapping the world’s archaeological sites, taking the first step towards creating a tool that would help humanity protect and preserve our collective cultural heritage.

While we were full of hope, we weren’t entirely sure what to expect. As it turned out — as with the best experiments — the outcomes were both better than what we could have hoped for and opened up many new questions to be answered and explored.

In the beginning…

It all began on January 30th, 2017, when we launched the new GlobalXplorer° platform and put it in the hands of the crowd. As we recently reported, the GlobalXplorer° platform was built in conjunction with our partners DigitalGlobe, a satellite-image provider whose high-resolution images of Earth GlobalXplorerº users combed through for signs of archaeological sites.

Since the day of launch, over 70,000 citizen scientists from more than 100 countries joined the GlobalXplorer° mission, logged in and began looking at and voting on 14,620,932 individual satellite images of Peru, covering 20% of the country, in a region where we thought the search would be most fruitful. The search took in 150,000 square kilometers of the country, from Peru’s Pacific coast to the Sacred Valley.

The region within Peru that the GlobalXplorer° crowd examined. Image: Jennifer Wolfe

Why Peru?

For its first expedition, GlobalXplorer° had a tough choice to make: where to begin our mission. While we had the whole planet to choose from, we settled on Peru for a variety of important reasons. First, it is known as a country rich in archaeological heritage — from Machu Picchu to the Nasca lines and beyond — so we knew it would excite the interest of citizen scientists around the world. Second, we knew that Peru’s government and archaeological experts are open to the use of technology: the Ministry of Culture in Peru has been using drones to map sites for the last several years.

We also needed to test GlobalXplorer° on a country that has its archaeological sites out in the open — without too much forest coverage, for example — and Peru’s coastal desert landscape fit the bill. We wanted the archaeologists we partnered with to be safe — someplace with no threat of war or conflict. Finally, Peru is also a country whose sites have been vulnerable to looting for hundreds of years, an ongoing problem that authorities are constantly working to address. We knew that GlobalXplorer° would be able to help alleviate the challenge of finding looted sites.

Topping all those considerations off was that archaeologists had not yet mapped Peru’s sites in totality. We knew that Peru’s authorities and archaeological community would be happy to receive and make use of the data GlobalXplorer° would generate.

Luis Jaime Castillo Butters at work with a drone. Image courtesy of Sarah Parcak.

Linking sky to land

GlobalXplorer° partnered with National Geographic to help us make all the connections we needed to join up our crowdsourcing efforts with experts on the ground in Peru. Larry Coben, an archaeologist who founded Sustainable Preservation Initiative (SPI), our in-country partner for ground truthing and community development projects, and our NatGeo point person Matthew Piscitelli — an archaeologist specializing in Peru — connected GlobalXplorer° with archaeologist Luis Jaime Castillo Butters, vice president of SPI in Peru as well as a National Geographic Explorer, professor at Pontificia Universidad Católica del Perú (PUCP) and former Peruvian vice minister of culture.

Luis Jaime has deep knowledge of where archaeologists are working in Peru’s various regions, so he has been instrumental in helping GlobalXplorer° figure out which of the crowd-identified sites should be prioritized for investigation. He also helps us decide what incoming data would be most immediately useful to local archaeological research and protection. Finally, he’s an expert in documenting archaeological sites using drones — so he took the lead in confirming the first batch of sites found by our citizen scientists. We’ll tell you more about that in just a bit.

This satellite image reveals evidence of ancient connected buildings, found by the GlobalXplorer° crowd. Image courtesy of DigitalGlobe.

The big question: what did the crowd find?

Meanwhile, back on the GlobalXplorer° platform, the crowd began the monumental task of examining more than 14 million individual tiles and voting on those that appeared to have human-made structures on them. (Read more about the voting process in our inaugural post.) Over the course of the campaign, the votes rolled in. In the process, the crowd identified 19,084 features of archaeological interest, dating all the way back from Caral, the oldest civilization in Peru c. 3200 BCE, all the way to the fall of the Inca in 1572 CE.

The GlobalXplorer° team, led by founder Sarah Parcak and fellow satellite archaeologist and executive director Chase Childs, then sorted the 19,084 sites into categories, including 342 sites of very high interest because they may potentially be large and important settlements. The crowd also positively identified evidence of looting, which will be used by Peruvian authorities to monitor ongoing looting and for site protection.

Overall, the crowd found thousands of structures. While we can’t say for sure what they all are, there appear to be terraced lands, fortified hilltop structures, settlements and elite complexes. For some contextual clues, we know that the hilltop settlements often signify periods of stress (in Andean chronology, “Intermediate Periods”), while people settled on valley floors during so-called “horizons” (the Wari and Inca eras would count as horizons) — so potential sites of these types could belong to these cultures, for example. The Inca also built rectilinear structures, while the Wari might have built D-shaped structures and cellular/gridded settlements.

While we of course can’t yet know what role these structures might have played and what culture they belong to, we can report that the crowd found many groupings of features on hilltops and on mountain summits, easily defensible sites that might indicate settlement locations.

The sites the crowd found will undoubtedly add to what we know so far of ancient Peruvian cultures, and tell us more about ancient ordinary people’s daily lives. These are the sorts of questions that the data being gathered by GlobalXplorer° will help researchers answer. We’re very much looking forward to sharing results of future studies with you as they unfold in the years to come.

Boots on the ground, drones in the air

After the staff at GlobalXplorer° HQ sorted through and organized the crowd’s positively identified images, they handed the first batch of 400 confirmed sites of interest — located around the Palpa and Nasca Valley regions of Peru — to Luis Jaime. Of these, Luis Jaime selected 40 for an initial round of ground truthing, and assembled a team that included Johny Isla, an expert in the history of the Nasca region and the archaeologist from Peru’s Ministry of Culture responsible for the protection and conservation of the sites located in this area. His expertise was instrumental in helping the ground-truthing team access all the sites investigated. The team also included photographer and journalist Diego Ochoa and PUCP archaeology students Karla Patroni and Fabrizio Serván.

Luis Jaime’s job: to get as close as he could to the sites and map them using drones. This step further confirms what the crowd finds, and adds much more information about each site to peg to the satellite data. The imaging available using drone technology is astonishingly detailed: the best resolution available from satellite imagery is between 70 and 50 centimeters per pixel, whereas drones can take the resolution to one centimeter or less per pixel.

The ground-truthing team set out to map the 40 sites, which are located in mountainous areas, requiring a method they call “slope flying”: sending out drones to map one strip of land at a time at varying altitudes, to ensure the most consistent, best-quality imaging. The resulting data — which includes high-resolution physical representations of sites in the form of still photos, 3D models and videos — can be used to protect sites from such damage as rain and looting, and help future researchers consider sites without having to disrupt the natural environment. Once this imagery is handed over to the Peruvian authorities, protection boundaries for the site can be set up. The Ministry of Culture can then also legally register that site as an archaeological property under the protection of the state.

Using drones not only allowed exceptional imaging, but saved the ground-truthing team from having to hike into remote or inaccessible areas such as the tops of mountains — or sandy deserts that can strand vehicles. “There was a very interesting site in the Nasca region just where six or seven rivers meet to form one single river,” says Luis Jaime. “The problem is that it was far, in an area that was really sandy. Each of our three vehicles got stuck at some point. Some of the team members got lost, too — it was a mess. But sometimes you have to go to the end of the world to find these things.”

Drone images taken by Luis Jaime Castillo Butters in the region explored by the GlobalXplorer° crowd revealed geoglyphs that had never been previously documented. Images courtesy of Luis Jaime Castillo Butters.

Looking for lines in all the right places

In an exciting twist, ground truthing our Xplorer’s sites serendipitously gave Luis Jaime and his team a chance to record as-yet undocumented geoglyphs possibly related to Peru’s famous Nasca lines—those giant geometric and organic shapes etched in the landscape by the Nasca people between 1 and 700 CE. Local communities knew of these geoglyphs, as did Japanese and German research teams working in the region for several years, but Luis Jaime and his team had hitherto never seen them.

The new lines were not visible to the crowd via satellite imagery because of imagery resolution, the slopes on which the geoglyphs were situated and their color being very similar to their context. But the ground-truthing expedition made possible by GlobalXplorer° offered Luis Jaime and his team the opportunity to add to the knowledge about them, too. “We want to empower local experts with GlobalXplorer° crowdsourced data, and we are thrilled to be able to collaborate with Luis Jaime and his team,” says Sarah Parcak. “We cannot wait to see what other similar collaborations bring globally.”

Read more about the Nasca lines and the data Luis Jaime was able to capture at National Geographic.

SPI’s bicycle tour program in the Pachacamac archaeological site. Photo: SPI

Creating stewardship

GlobalXplorer° isn’t just about finding sites — it’s also about engaging local communities living in and around them — incentivizing them to participate in protecting their local heritage while increasing their income and creating a sense of stewardship.

For this aspect of the Peru expedition, we looked to our partner SPI, a New York–based organization dedicated to creating community development programs. We weren’t able to find a community in the immediate vicinity of the area we were ground truthing, so instead we chose a community living near Pachacamac, a huge archaeological site on the outskirts of Lima with a long history of occupation, ending with the Inca. Pachacamac is constantly under threat from human encroachment because it is so close to highly populated areas. So in partnership with SPI, GlobalXplorer° launched a program that enlists young people to give bicycle tours, called BiciTour — Peru’s first such enterprise.

Lessons learned — and what next?

As we write this, SPI in Peru is getting ready to ground truth some of the other sites that the crowd identified. This process will continue to unfold as resources and environmental conditions allow. Later this year, GlobalXplorer° will also write up the findings for the Peru expedition in an academic paper, and hand over its crowdsourced data to the Peruvian government so that they can add to a centralized database of all archaeological sites — which we hope will be useful to the country’s citizens, government and researchers. As all this continues to unfold, there’s much more ahead: we’re preparing to launch our next citizen expedition in Country Number 2 — to be revealed very soon.

As excited and proud as we are of GlobalXplorer°, our citizen scientists and our partners at the completion of our first expedition, it’s important to remember this really is just the very beginning. The work we’ve done in Peru is only the first phase in demonstrating what the crowd can contribute to the world’s archaeological efforts: the impact going forward will likely be immeasurable.

Having said that, the first and most important thing we learned is that the system works. GlobalXplorer° has started changing the face of archaeology by making ordinary citizens part of the process. The crowd’s eyes are accurate to 85%, looting sites were very easily and accurately spotted and thousands of sites of interest were found — and fast.

In fact, with some tweaks to how we sort and organize the satellite image tiles before they’re presented to GXº Xplorersº — deleting any images with too many visual obstructions, for instance, and searching one small area at a time — we should be able to cover more ground more quickly, and be able to organize ground-truthing expeditions more efficiently.

Remember GlobalXplorer’s original goal of mapping the entire planet’s archaeological sites in ten years? We’ve realized that, if the results of the Peru expedition are any indication, our dream just might be attainable.

Not that we want to get ahead of ourselves. It’s not all about technology. We still have to navigate the human challenges — organization, government collaboration, limited resources and so on. Regardless, the question we need to start thinking about now is: How will we best make use of all the data we’re about to gather, and organize and disseminate it in the interest of making the preservation of cultural heritage sites a global priority?

No small task. Watch this space.