When archaeologist Stéphen Rostain first started doing fieldwork in the Amazon rainforest about 40 years ago, there was very little interest in the region.
Archaeologists studying the history of the Americas beftore European colonisation were largely focused on the remains of pyramids and temples built by the ancient Mayas who lived in Mexico and Central America.
It was believed that for the past 2,000 years or more, small, nomadic tribes had lived in the dense forests of the Amazon, but that civilisation, as we know it, had never reached them.
“We were less than 10 archaeologists working in the Amazon,” says Rostain, now director of research at the French National Centre for Scientific Research (CNRS) in Paris. “Everybody said there’s nothing to find.”
For four decades, Rostain carried out excavations of the ground in the Upano Valley, a part of the Amazon rainforest in Ecuador, painstakingly collecting evidence that seemed to prove otherwise. But his breakthrough finally came in 2021, thanks to imagery captured from the air using light detection and ranging (lidar) technology.
Lidar is like radar, but uses lasers instead of radio signals. By firing pulses of laser light, a lidar device can detect how long it takes for the pulses to be reflected back and uses that information to gauge range. Fire enough pulses at something and you can use the range information to build up a 3D picture, mapping anything the pulses are hitting.

In Rostain's case, a lidar device was attached to the belly of an aircraft that would then fly over the Upano Valley, bouncing laser pulses off the ground below.
Using the information gathered during the various flights, he was able to produce high-resolution 3D maps of the area that revealed a vast network of lost cities.
"Lidar gives this magic result where you see the original landscape of the first inhabitants," he describes. "It's really impacting."
When Rostain's paper announcing his findings was published in 2024, it was hailed as a major discovery.
Archaeology from the air
In the last few years, airborne lidar has increasingly been used by archaeologists. Although the technology dates back about 60 years, it has only recently become more accessible, with the first commercial systems appearing in the mid-1990s.
One reason for this is decreasing cost, partly due to improvements in sensor technology. Systems have also become more compact, lightweight and less power-hungry, so they can now be attached to drones – a more cost-effective option than using manned aircraft for mapping a small area.
The pulse density of airborne lidar has also improved, meaning it can produce much higher resolution 3D maps – which naturally are of interest to archaeologists.
There are also more companies making lidar machines and they can now be used without expert knowledge.
“They’re getting much easier to operate,” says Mark Robinson, an associate professor of archaeology at the University of Exeter in the UK. “I don’t have a background in physics or in science, yet I could operate a lidar system and attach it to a helicopter, run it and process the data.”
Ease of use aside, what’s really driving the use of lidar among archaeologists is what it can reveal.
In certain landscapes – forests, for example, or areas with thick vegetation – it can be difficult to detect features on the surface, such as remains of settlements, by doing a ground survey (walking around with compasses, charts and a GPS device).
“I’ve hiked a lot of jungles and done surveys with a machete, and it’s embarrassing how close I’ve been to quite large stone pyramids and not seen them,” Robinson admits. “I couldn’t find any structures, but when I get a lidar image back, [I can see that] I was so close.”
Using lidar, archaeologists can also peer inside a forest without having to chop a single branch and potentially harm the environment.
As millions of laser pulses are fired towards the ground, bouncing off any surfaces they encounter before being reflected back to the lidar machine, each one generates a mapping point that forms part of the image. And because there are so many pulses, a few of them can penetrate gaps between leaves and hit the ground.

“For mapping, that’s still pretty good for us, to actually have that many points on the ground,” says Robinson. “When we get back in the lab and on the computer, we can remove all the vegetation at the top [of the lidar image] and interpolate between the ground points to fill in the gaps.”
Robinson and his colleagues were recently studying the pre-Hispanic history of the Bolivian Amazon, an area covered by a dense, evergreen tropical forest. By doing an airborne lidar scan, they discovered a network of connected settlements in the area, made up of larger cities and smaller towns with complex architecture.
Similar to Rostain’s work, this provides evidence that the Amazon was more inhabited than previously assumed and that settlements in this area had a level of complexity comparable to Europe and many other parts of the world in the same time period.
“This is nothing wild or out of the realm of what we expect for human settlements, but it's never really been shown in Amazonia. Certainly not on this scale,” says Robinson.
Robinson was surprised by the water management infrastructure revealed by the lidar imagery. It had been developed to deal with floods, which often occurred annually and would inundate about 100km (approx 60 miles) of savannah.
Gaps were built within the embankments, behind which pools of water could form where fish could be trapped and easily collected. There were also canals and causeways to drain excess water into nearby rivers.
"The level of engineering that went into this planned landscape is incredible," says Robinson.
Read more:
- Scientists uncover two gigantic hidden ancient cities
- 3D scanning gives new insight into 275-year-old Jacobite battlefield
- How scientists are using cosmic radiation to peek inside the pyramids
So much more to see
Lidar’s ability to reveal hidden features in the landscape was also transformative for Rostain. When he first surveyed the Ecuadorian Amazon on foot in 1995, he identified hundreds of mounds that typically suggest past settlements and places where buildings were erected.
But his recent lidar analysis showed that his expeditions had only identified a fraction of what was there.
“Thirty years after, I saw that there were not hundreds of mounds, but thousands of mounds,” Rostain says.
Some of the mounds were easy to miss due to their size: they were just 1m (approx 3ft) high. Rostain and his colleagues also uncovered other smaller details in the landscape from the lidar scan, such as furrows in terraced agricultural fields surrounding the lost cities.
“They’re 20cm [about 8in] high,” says Rostain. “They appear very clearly with lidar, but you have to use different filters to understand what you’re seeing.”

Thanks to lidar, Rostain was also able to gain new information about how different areas were connected. The technology allowed him and his team to generate a detailed map of an area covering roughly 600km² (approx 230 sq miles).
Although they had seen evidence of dirt roads from their ground surveys, the vast aerial view revealed a large street that was up to 15m (49ft) wide and 5m (approx 16ft) deep.
A closer look showed that it connected all the different settlements and that there was also a dense aggregation of roads everywhere.
“They were obviously trade roads, but also probably ceremonial roads connecting a complete territorial system,” says Rostain.
Lidar isn’t just providing archaeologists with more insights, it’s also speeding up their fieldwork. Pedro Guillermo Ramón Celis, a postdoctoral researcher at McGill University in Montreal, Canada, recently used a lidar system installed on a Cessna aircraft to survey the archaeological site of Guiengola in Tehuantepec, Mexico.
The 4km 2 (1.5 sq mile) site is home to a settlement built by the Zapotecs between 1350 and 1521, but is situated within a thick broadleaved forest.
Celis says that if his team had tried to map it using traditional methods, it would have required a crew of people equipped with machetes, GPS, compasses and charts, and taken a lot longer than it did. “Instead of doing two years of research, it took us two hours,” he says.
The lidar scan was also groundbreaking for Celis since it overturned previous thinking about the nature of the settlement at Guiengola. Historical documents also suggested that it had been a significant fortress occupied by the elite.
When Celis had previously surveyed the area on foot, he had identified about 450 well preserved buildings that weren’t covered with soil. However, an analysis of the high-resolution lidar map produced during his team’s airborne surveys, in which one pixel on the screen represented about 20cm (about 7in) of the actual area, revealed about 1,100 structures.
Among them was a palace, temples and plazas of various sizes, which suggest that it was a city with different neighbourhoods inhabited by a range of social classes.

Between the 13th and 15th centuries, the Zapotec people migrated towards the coast, and Celis thinks that Guiengola was a base where they would have stayed for a while before continuing towards the Pacific.
“Fundamentally, we realised that we were dealing with one of the Zapotec capitals of that time, which is transforming our idea of the political landscape,” says Celis.
Where we go from here
The spatial information provided by lidar is just one piece of the puzzle, though. For many archaeologists, it serves only as a starting point for further investigation. Other technologies are often needed to complete the detailed maps of ancient settlements that lidar can produce.
Rostain, for example, is now following up his surveys with radiocarbon dating to get a more precise idea of when people lived in the part of the Ecuadorian Amazon that he studied.
Likewise, Celis plans to return to Guiengola later in 2025 and survey the site again, this time using a magnetometer and ground-penetrating radar. "Now I want to see deep into the site," he says. "To find out how much construction there is below what we can see."
But it’s not just the structures the archaeologists are concerned with. The lidar surveys also prompt them to try and answer deeper questions about the societies that lived at these sites.
Celis, for example, is planning to investigate the chemical composition of pottery at Guiengola using X-ray fluorescence, a technique that can identify materials by examining them with low-level radiation so that they emit characteristic frequencies.
“I would like to know if [the locals] still had strong ties or commercial connections to their homeland, or if they were starting to create their own pots and material culture,” he says.
Robinson and his team also want to better understand specific features they uncovered in the Bolivian Amazon. To gain a greater understanding of the hydrologic infrastructure they found, for example, they would like to simulate how water would flow through it in different conditions, such as during a period of intense rainfall.
“What would that look like? How would it function and when would it break?” Robinson wonders.
The past informs the present

Whereas archaeology has typically focused on better understanding the past, Robinson thinks that it can also provide insight into the present, specifically the global challenges we face today.
Finding out more about how these past societies dealt with floods, for example, could inspire new solutions for the devastating floods that, due to climate change, are occurring more frequently.
Robinson is also interested in the long-term impact of human settlements on the landscape, which we know little about. His team's current work in Belize is using lidar data to investigate whether past urbanisation had a lasting effect on the biodiversity and carbon storage of the country's forests.
“We’re working with a lot of the lidar data to characterise where and what the disturbance was, its intensity, and the type of site,” says Robinson. “Was it a big marketplace, a big temple or was it a farm that may have enriched the soil?”
In recent work, Robinson and Sarah Eshleman, one of his post-docs, examined how Maya structures built 1,000 years ago in northwestern Belize correlated with the current forest structure.
They found that in areas where the Maya people had constructed the tallest buildings, there are now huge trees. They suspect it’s because the limestone blocks and plaster used in the buildings, as well as other residue from human activities, enriched the soil.
“Ecology and resource managers don’t have this kind of data on these long timescales,” says Robinson. “I think it’s an exciting way to expand the impact of archaeology.”
About our experts
Stéphen Rostain is an archaeologist and the director of research at the French National Centre for Scientific Research (CNRS) in Paris. He has authored, co-authored and appeared in 12 archaeology books, including Islands in the Rainforest: Landscape Management in Pre-Columbian Amazonia.
Dr Mark Robinson is an associate professor of archaeology at the University of Exeter in the UK. His work has been published in various journals including Journal of Archaeology, The Journal of Island and Coastal Archaeology and Quaternary Science Reviews.
Pedro Guillermo Ramón Celis is a postdoctoral researcher at McGill University in Montreal, Canada. His work has been published in journals such as Arqueología Mexicana, Ancient Mesoamerica and Reference Module in Social Sciences.
Read more: