For a century and a half, the study of Angkor was a study of temples — the stone survived, so the stone got the attention. Everything else the Khmer built was made of wood, thatch, and earth, and the tropics erased it. Archaeologists knew a city must have surrounded the monuments. They could not see it.
Then, in April 2012, a helicopter carrying a laser scanner spent about twenty hours flying grid lines over the Cambodian forest, and the picture of Angkor changed permanently.
How You Find a City With Lasers
LiDAR — Light Detection and Ranging — works by firing hundreds of thousands of laser pulses per second at the ground and timing their return. In a forest, most pulses hit leaves and branches, but a small fraction slip through gaps in the canopy, strike the actual soil, and bounce back. Discard everything except those ground returns, and you are left with a bare-earth model of the terrain: every mound, ditch, embankment, and pond, with the jungle digitally stripped away.
For Angkor this was transformative, because the missing city was not made of buildings that had vanished without trace — it was made of earthworks. House platforms raised above flood level, ponds dug beside them, canals, road embankments, field boundaries. All still there, all a few tens of centimetres of relief under a thousand years of forest, all invisible to anyone standing in it.
2012: The First Flight
The 2012 campaign, organised by a consortium and led by archaeologist Damian Evans, covered about 370 km² across Angkor itself, the Phnom Kulen plateau, and Koh Ker. Two results stood out.
First, at Angkor: the space between the famous temples was not empty parkland. It was densely inhabited urban fabric — grids of occupation mounds and ponds, streets, and water infrastructure, showing that Angkor Wat and its neighbours sat inside a living city rather than standing alone in the manner of the postcards.
Second, up on the Kulen: the survey traced the outline of Mahendraparvata, the 9th-century capital where Jayavarman II was consecrated in 802 AD and which existing scholarship knew only from inscriptions. The data showed raised embankments running close to the cardinal directions, forming a deliberate grid, with temples and structures inside the blocks. As Evans noted at the time, that degree of formal planning implies centralised control — in contrast to other Khmer settlements of the period, which grew organically. Mahendraparvata predates Angkor Wat by roughly 350 years.
2015: Going Much Bigger
A European Research Council grant funded Evans's Cambodian Archaeological LiDAR Initiative (CALI), and a second campaign in 2015 mapped more than 3,000 km² — at the time the largest archaeological LiDAR survey ever conducted in Asia.
The larger dataset confirmed and extended the picture: an entire previously undocumented, formally planned urban landscape into which the great temples were integrated, spreading across a region far larger than the walled enclosures. Angkor emerges as a low-density agrarian city — not dense in the manner of medieval Paris, but sprawling over an area plausibly making it the most extensive settlement complex of the pre-industrial world, sustained by the water-management system that gave the "hydraulic city" its name.
That reading also sharpens the collapse question. If the city was its water infrastructure, then the decades-long droughts and violent monsoon years that tree-ring records place in the 14th and 15th centuries were not a background inconvenience — they attacked the thing the whole settlement was built on.
What This Changes for a Visitor
Three things, and they are worth carrying into the park:
- You are standing in a city, not a temple complex. The forest between monuments is not wilderness that swallowed ruins; it is the ghost of neighbourhoods. Look at the low mounds and rectangular ponds beside the road on the way to Preah Khan — many are house platforms and household ponds, and once you know that, the drive stops being dead time.
- The temples were embedded in ordinary life. The Bayon's bas-reliefs of markets, cooking, gambling and childbirth are not quaint decoration; they are documentary evidence of the population the lasers found. See our Bayon guide for which galleries to walk.
- "Lost city" is a headline, not a description. Mahendraparvata was never lost to Cambodians and had been referenced in inscriptions for a very long time — what LiDAR provided was a map, not a discovery in the Indiana Jones sense. The same correction applies to the myth that Angkor Wat itself was ever abandoned; more on that in our facts guide.
Where to Go and See the Evidence
Phnom Kulen — Mahendraparvata itself
The plateau is a day trip (car recommended, separate ~$20 ticket) best known for its waterfall and reclining Buddha, but you are also driving across the footprint of a 9th-century capital. Ask a good guide to point out the linear embankments and reservoir traces near the temple sites; the Kulen guide has the logistics.
The barays — the system, at scale
Drive along the West Baray, or stand on East Mebon, which once sat on an island in a reservoir that is now dry farmland. Nothing conveys the ambition of Khmer hydrology faster than realising the "field" below you was a lake dug by hand.
Beng Mealea and the outer sites
The unrestored temple at Beng Mealea, an hour east, shows what the surveyed landscape looks like when nobody has cleared it — and sits on the old royal road that the LiDAR work helped trace.
The museum first
An hour at the Angkor National Museum before your first temple day gives you the chronology to hang all of this on.
If you want the landscape narrated rather than guessed at, a guide is the difference — the classic small-group day covers the icons with the context attached:
Longer trips get closer to the archaeology: the four-day itinerary has room for Kulen and Beng Mealea alongside the circuits.
The Work Isn't Finished
LiDAR shows shape, not date or function: a mound is a mound until someone digs it. The surveys have handed archaeologists decades of excavation targets, and interpretations of population size, city extent, and chronology are still actively debated. That is the honest state of play — and it is also why Angkor remains one of the most interesting places on Earth to be an archaeologist. Twelve hundred years in, the jungle is still handing things back.