LiDAR and Forgotten Cities

Core Concepts

What LiDAR actually does

LiDAR — Light Detection and Ranging — is an airborne laser-scanning technology. A sensor mounted to a plane or helicopter fires millions of laser pulses per second downward into the landscape, measures how long each pulse takes to bounce back, and constructs a precise 3D point-cloud of every surface it hits.

The critical capability for archaeology is canopy penetration. LiDAR does not see through vegetation — rather, it exploits the fact that no canopy is perfectly sealed. In dense tropical rainforests, 10–30% of laser pulses find gaps in foliage and reach the ground. That is enough. Processing software strips out returns from tree-tops and mid-canopy layers and retains only the ground-level returns, generating a "bare earth" digital elevation model. Features invisible to anyone standing in the forest — earthen platforms, roads, ditches, raised fields — appear on screen as clear topographic relief.

From site excavation to regional reading

The methodological effect runs deeper than simply "finding more sites." LiDAR shifted landscape archaeology from a site-by-site logic — dig here, survey there — to regional-scale landscape reading. Researchers can now examine the accumulated human modifications across thousands of square kilometres in a single survey campaign. Settlement patterns, road networks, and agricultural systems become legible as integrated wholes rather than isolated data points. As one review in the Annual Review of Anthropology put it, LiDAR enables archaeologists to read entire regional palimpsests — the layered record of human action written into the land over centuries.

Low-density urbanism

Many of the societies revealed by LiDAR do not fit the textbook image of a dense, nucleated city. Instead they practised low-density urbanism: dispersed residential platforms, gardens, and managed forests connected by roads and organised around ceremonial or civic centres. Population was spread across the landscape rather than compressed into it. This settlement logic — sometimes called garden urbanism — integrates agricultural infrastructure directly into the urban fabric. Understanding it requires abandoning the European city as the reference model.

Terra preta and the engineered Amazon

Terra preta do Indio — Amazonian Dark Earth — is an anthropogenic soil found across the Amazon basin. It is characterised by high concentrations of charcoal, organic matter from food remains, and pottery sherds, and contains two to three times the nutrient content of surrounding poor-quality soils. It was intentionally manufactured — not a byproduct of burning but a deliberate soil-engineering practice sustained over generations. Modern-day Kuikuro descendants continue the practice by spreading ash, charcoal, and cassava waste in agricultural fields.

Terra preta matters for the story of LiDAR not because laser light detects soil chemistry, but because its approximately 154,063 square kilometres of distribution across the basin is independent corroboration of long-term, large-scale human presence. The soil itself is a landscape archive.

The pristine myth and terra nullius

Colonial observers — and many twentieth-century ecologists — described Amazonia as a "pristine wilderness," an environment unshaped by human hands. This framing had a political function: if no one meaningfully occupied or managed the land, it was available for appropriation. The legal doctrine of terra nullius ("empty land") formalised this logic, holding that lands without Christian occupation were vacant and claimable by European states.

LiDAR evidence systematically refutes both the descriptive claim and the political inference it was used to support.

---

Narrative Arc

Before the laser: what we thought we knew

For most of the twentieth century, the dominant position in Amazonian anthropology held that the basin's poor, leached soils could not sustain large, dense populations. The rainforest was a "counterfeit paradise" — visually lush but nutritionally thin. Complex societies, it was argued, could only have existed on the rich floodplain margins, and even then at modest scale. Interior Amazonia was treated as essentially empty of pre-Columbian complexity.

This view was never purely academic. It supported colonial land claims and provided a scientific veneer for the argument that Indigenous peoples had not really used the land in any way that deserved legal recognition. The wilderness was available; it was just waiting.

The 1990s rupture

The first major crack appeared through archaeology, not technology. Researchers working in the Upper Xingu and in the Llanos de Mojos of Bolivia began finding earthworks — raised fields, causeways, ring-ditches — that demanded explanation. Anna Roosevelt's excavations at Marajó Island pushed back estimates of Amazonian social complexity significantly. Betty Meggers' "counterfeit paradise" thesis came under sustained empirical pressure.

Beginning in the 1990s and accelerating through the 2010s, this archaeological evidence combined with LiDAR technology enabled the systematic recovery of pre-Columbian urbanistic and landscape-engineering knowledge. The question shifted from whether complex societies existed in pre-Columbian Amazonia to how large and how widespread they were.

What changed in the 1990s was not just new sites — it was a new willingness to take Indigenous accounts of a dense, managed Amazon seriously as historical testimony rather than legend.

LiDAR arrives: the scale problem becomes visible

The technology's full impact arrived in a series of landmark surveys from the mid-2000s onward. In 2013, LiDAR was applied to Angkor. In 2018, PACUNAM LiDAR Initiative results for the Maya Biosphere Reserve were published in Science. The Amazon surveys followed in rapid succession through the late 2010s and early 2020s.

The scanning of just 5,315 square kilometres of publicly available Brazilian LiDAR data — less than 0.1% of the Amazon basin — identified more than 900 previously known earthworks and discovered 24 previously unreported sites, including suspected ceremonial centres, fortified villages, and megalithic structures. Projections from that coverage suggest 23,000 or more undiscovered earthworks remain beneath the canopy.

The cumulative effect was not incremental. It was a paradigm break.

The 2024 culmination: Upano Valley

The clearest statement of how far the field has moved came in a 2024 study published in Science. The Upano Valley in Ecuador contains the oldest documented pre-Columbian urban complex in Amazonia, dating to at least 2,500 years ago — around 500 B.C.E. — more than 1,000 years earlier than any other known complex Amazonian society. Radiocarbon dating confirms occupation spanning from 500 B.C.E. to between 300 and 600 C.E.

The valley contained five large settlements and ten smaller ones, distributed across 300 square kilometres in a geometric pattern connected by straight roads and intertwined with agricultural landscapes. More than 6,000 earthen platforms were arranged across the valley in a structured geometric plan. The core area of Kilamope was described as comparable in scale to the Giza Plateau or the main avenue of Teotihuacan.

This single case demolished what remained of the "counterfeit paradise" thesis in its strongest form.

---

Annotated Case Study

The Upper Xingu: galactic urbanism in the Brazilian Amazon

The Upper Xingu region of the Brazilian Amazon provides one of the best-documented and longest-studied examples of what LiDAR made legible.

Anthropologist Michael Heckenberger began excavations in the region in the 1990s, working closely with Kuikuro communities whose oral traditions describe a dense, sophisticated pre-Columbian landscape. His three decades of field work documented what he called "low-density garden cities" — a settlement form integrating ceremonial cores, residential platforms, roads, plazas, orchards, and managed forests.

The focal site is Kuhikugu, an ancient fortified settlement on the Upper Xingu River. LiDAR and archaeological documentation reveal a complex of more than twenty large villages spanning over 13 square kilometres, connected by straight planned roads, and featuring central plazas, a chief's residence fortified with a log palisade and defensive ditches, surrounding orchards, and managed forests. Dating evidence places occupation at least 1,500 years before present. Regional population estimates for the Xingu settlement cluster reach 30,000 to 50,000 or more people.

At regional scale, the Upper Xingu contained articulated settlement clusters representing small independent polities within a regional peer-polity network — what researchers describe as a "galactic" form of prehistoric urbanism. Each prehistoric cluster supported an estimated 2,500 to 5,000 people. The clusters share road alignments, architectural conventions, and ritual practices, indicating sustained coordination across the region.

What makes this case instructive is the convergence of independent evidence streams:

1. Oral traditions. Kuikuro communities had maintained accounts of a densely settled, sophisticated landscape. These were initially treated with scepticism.
2. Ground archaeology. Heckenberger's excavations confirmed earthworks, roads, and settlement sequences consistent with those accounts.
3. LiDAR. Aerial surveys made the regional scale visible at once, resolving arguments about whether individual sites were anomalies or part of a broader pattern.
4. Soil chemistry. The presence of terra preta at settlement sites confirmed sustained occupation and intentional landscape management.
5. Botanical evidence. A significant positive relationship exists between 35 domesticated tree species — including Brazil nut, Pará rubber tree, murumuru, and cacao — and the probability of finding pre-Columbian earthworks. The forest composition itself is a map of past habitation.

No single evidence stream was decisive. The case rests on their convergence.

The agricultural infrastructure underlying this urbanism is equally significant. Raised field systems in Amazonia supported estimated population densities of 50 to 100 inhabitants per square kilometre — figures incompatible with the "sparse hunter-gatherer" model. These raised fields were not natural features. They required coordinated labour, sustained maintenance, and spatial planning.

The political stakes. LiDAR evidence confirms that in 1492, human influence had spread throughout the Amazon basin with substantial settlements, engineered landscapes, and sophisticated land-use practices. The entire southern rim of the Amazon was populated by complex societies whose population may have reached 5 million. Ancient Amazon earthwork findings now play an explicit role in ongoing political debates about Indigenous land rights, with LiDAR results bolstering descendants' claims to ancestral lands by confirming historical presence and long-term occupation of contested regions.

A comparative note: Maya lowlands and Angkor

The Amazonian case is not isolated. LiDAR produced comparable paradigm breaks in two other major tropical civilisations.

In the Maya Biosphere Reserve of northern Guatemala, LiDAR surveys revealed over 61,000 ancient structures, including isolated houses, large palaces, ceremonial centres, causeways, agricultural terraces, reservoirs, and defensive walls. The raised causeways — sacbeob — extended hundreds of kilometres across the lowlands, functioning as the organising spines of an integrated regional settlement system. The impact was compared to what the Hubble Space Telescope did for astronomy: revealing that the subject of study is fundamentally larger and more complex than anyone had understood.

At Angkor, Cambodia, LiDAR surveys revealed that the iconic temples were merely the epicentre of an immense low-density settlement complex stretching over approximately 1,000 square kilometres. The complex featured elaborate water-management engineering designed to handle monsoon variability. Peak population reached an estimated 688,000–900,000 people, making Angkor one of the most extensive low-density urban complexes of the pre-industrial world.

Fig 1