A Maya City Was Hiding in Lidar Flown to Count Trees. Nobody Had Looked.

The survey that found a city was a search engine

In 2024, a team led by the Tulane doctoral student Luke Auld-Thomas and his advisor Marcello Canuto announced a large Maya city in the eastern Mexican state of Campeche. They named it Valeriana, after a nearby lagoon. It has pyramids, plazas, a ball court, and a reservoir; the survey around it mapped roughly 6,700 structures; at its Classic-period height, more than a thousand years ago, it may have held on the order of 50,000 people.

They did not find it by hacking a machete through the bush. Auld-Thomas found it on the internet. The laser data that revealed Valeriana had been flown in 2013 for an environmental project, monitoring forest and carbon, not ruins, and posted online, where it sat, unread by any archaeologist, until a graduate student went looking on the second page of a search.

What lidar actually does

Lidar, light detection and ranging, fires laser pulses from an aircraft and times their return. Enough of those pulses slip between the leaves to reach the ground that software can, in effect, delete the forest from the data and leave the bare earth behind. Under the canopy of the Maya lowlands, that bare earth is not empty. It is corrugated with the low mounds of collapsed buildings, the lines of causeways, the geometry of plazas, a built landscape that the trees had been hiding in plain sight.

That is why the find matters more than one more dot on the map. The 2013 flight was never meant to look for people. It found a city anyway, because in this part of the world there is no such thing as a lidar survey that does not.

The point of the paper is in its title: Running out of empty space.

The correction is about us, not them

For a long time the working assumption about the Maya lowlands was that dense cities were the exception and the jungle between them was mostly wilderness. Two decades of lidar have steadily dismantled that idea. The Campeche study pushed it further: even a patch of land that archaeologists had simply never chosen to survey, one close to a modern town and a highway, turned out to be crowded with settlement, second in built density in its region only to the great city of Calakmul.

So the honest headline is not that Valeriana was "lost." A city fifteen minutes from a road is not lost to the people who live beside it; it was unrecorded by the discipline that claims the right to find it. What the data corrected was a scholarly habit of treating unsurveyed ground as empty ground, the same habit this Report flagged in the Amazon, where "lost cities" were only ever lost to outsiders.

What it does not tell us

Lidar is a map of shapes, not a history. It can show that thousands of structures exist; it cannot, on its own, date them precisely, say how many stood at once, or count the people who lived in them. The 50,000 figure is a modeled estimate built on assumptions about how many structures were houses and how many people a house held, a reasonable number, and still a number with a range around it. The buildings are real and measured. The population is an inference, and should be read as one.

What is not in doubt is the scale of what remains unlooked-at. The survey covered about 130 square kilometres. The Maya lowlands are hundreds of thousands. If a city of this size can surface from a single environmental flight that nobody flew for archaeology, the safest assumption about the rest of the forest is the one the paper makes: it is not empty. It is unread.