The Repository Stone Record

This Swamp Hid a Fortress. It Hid Something Else, Too.

September 1, 2026 · Runtime 9:28 · Watch on YouTube

In 2016, an airborne laser survey over a forest in eastern Poland flagged a ring under the tree canopy that matched nothing on any map. Officially, that's the story of how this fortress was found — a machine, from the air, in one afternoon. It isn't the whole story. Getting from that ring on a screen to 244 catalogued objects took nine years, three independent lines of evidence that never met in the field — and, buried in the same mud as the medieval fort, a second discovery nobody was looking for.

Chapters

  1. 0:00 A fortress "found" by a machine in 2016 — not quite
  2. 1:57 What the laser actually saw
  3. 2:51 The historian who got there first, on foot
  4. 4:13 A full survey without digging a single trench
  5. 6:03 244 objects, and what they say about the settlement
  6. 7:03 The same soil, seven centuries later
  7. 7:48 So who actually found it?

Transcript

A fortress "found" by a machine in 2016 — not quite 0:00

In twenty sixteen, a laser survey flying over a state forest in eastern Poland picked up something the maps did not have a name for. A ring, under the tree canopy, on an island in the middle of a marsh. Officially, that is the story of how this fortress was found — a machine, from the air, in one afternoon. It is not actually the whole story. Getting from that ring on a screen to two hundred and forty-four objects out of the ground took nine years. It took three separate people, working from three separate kinds of evidence. And, buried in the same mud as the medieval fort, there was a second discovery nobody was looking for.

The laser survey gets the credit, and it deserves some of it. But before any archaeologist set foot on that island, an independent historian had already worked out roughly where to look. Not from data — from a thirteenth-century chronicle — and he crossed the marsh alone to check. The professional dig that followed used almost no shovels at all. And what came out of the ground was not only early medieval. Pieces of a much more recent history were sitting in the same topsoil. That use of the island has nothing to do with a fortress, and everything to do with why it worked as one.

Start with the ground itself. Near the village of Kaplonosy-Kolonia, in the Lublin region of eastern Poland, there is an oval island of dry land inside a wetland forest. It sits close to the town of Włodawa — about zero point seven of a hectare, roughly the footprint of a football pitch. It rises just enough above the bog to stay dry through the spring floods, which is reason enough on its own to choose it. Reaching it on foot means crossing standing water, fallen trees and dense undergrowth. That difficulty is not incidental. It is the entire reason anything survived here at all.

What the laser actually saw 1:57

Two thousand and sixteen. An airborne laser survey fires pulses down through the tree canopy, timing them on their way back up. That builds a bare-earth model of the ground, with the trees digitally stripped away. On that model, over this stretch of forest, an anomaly appeared: a closed ring of raised earth, sitting on high ground inside a bog. It matched nothing on any historical map of the area. Grzegorz Mączka, of the regional heritage protection office in Lublin, flagged it. And then, officially, nothing happened for years.

A laser survey can find a shape. It cannot tell you what the shape is, how old it is, or whether it is worth the trouble of wading into a bog to check. An anomaly like this one sits in a queue behind hundreds of others, and for six years, that is exactly where it stayed.

The historian who got there first, on foot 2:51

Meanwhile, somebody was already looking, from a completely different direction. Dariusz Wołosiuk, an independent historian, had been reading the old chronicles of this borderland. They describe a network of fortified strongholds, mentioned in medieval Rus sources — some of them burned in raids recorded centuries ago. He reasoned that one of those lost sites should sit in this stretch of forest. In twenty twenty-two, without a survey team, without the laser data, he crossed the marsh himself and found the ring on foot. That is two years before any official fieldwork began. He later attributed the site's destruction to a raid recorded in a thirteenth-century chronicle. That is his own reading, not a conclusion the excavation itself has confirmed.

Lay the dates next to each other and the shape of the investigation becomes the story. Twenty sixteen: the anomaly is flagged from the air. Twenty twenty-two: an independent historian reaches it on foot, from a written source. Spring twenty twenty-four: the first official reconnaissance, and the first medieval pottery out of the topsoil. Autumn twenty twenty-five: a full non-invasive survey. Three different kinds of evidence, arriving nine years apart, agreeing on the same spot.

A full survey without digging a single trench 4:13

And when the professional survey finally came, in the autumn of twenty twenty-five, it went in almost without digging. Multispectral aerial photography, a refined laser terrain model, electrical resistivity readings pulled through the soil. And every recovered object was mapped by precise GPS, instead of a hand-drawn trench plan. That is not a budget shortcut. The site sits inside terrain now protected as a nature reserve, and the method matches the protection. Everything published this year about this fortress comes from reading the ground from above and below, without opening it up.

What that reading found: a rampart, broken into distinct segments in the resistivity data. That's consistent with a timber box construction — walls built as a frame and packed with earth, stone or clay, the standard method of this period. One clear gap in the eastern side of that rampart, in a position that reads as a gate. Buried pits and hearths beneath the interior, never opened, only located. A fortress described in outline, entirely from instruments, before a single test trench has been dug.

Dating a site nobody has excavated means comparing what turned up on the surface against objects from sites that have been dug. Here, that comparison — knife forms, pottery types — points to occupation from the tenth century through the thirteenth, then abandonment. Three hundred years of use, established without a single stratified layer. It comes purely from matching one unexcavated assemblage against two excavated ones nearby, at Horodyszcze and Strzeżewa. It is a weaker kind of proof than a dug section. It is also, right now, the only kind this site is going to get.

244 objects, and what they say about the settlement 6:03

The surface and near-surface collection alone produced two hundred and forty-four catalogued objects. Sixty-six of them are knives or knife fragments — more than a quarter of everything found, on its own. The rest spans an entire working settlement: sickles and an iron plough blade for farming the margins of the bog, fire-strikers and arrowheads. A silver earring, a horseshoe-shaped brooch, broken pottery. And one half-finished silver crescent pendant, a lunula, abandoned partway through being made. That last piece is not refuse. It is evidence that somebody was working silver, on this island, when whatever ended the settlement happened.

None of that explains everything the topsoil gave up. Because mixed into the same ground as the knives and the half-finished pendant, researchers found something else. Spent rifle cartridges, and the crushed case of a pocket watch. Those are not early medieval.

The same soil, seven centuries later 7:03

They are from the nineteen forties. Local accounts collected by the same research team connect them to the German occupation of the region. This same island of dry, defensible ground, reachable only by wading through a bog, sheltered people hiding from that occupation. By local memory, Jews from the Włodawa area. It is a single line in the published findings, not a chapter. The object record keeps it exactly that: a few cartridge cases and watch parts, in the same soil as ninth-century knives. The paper documenting all of it, fittingly, is published by the museum in Włodawa's former synagogue complex. It's a building with its own two histories in one address.

So who actually found it? 7:48

So: who found this fortress? A laser, in an afternoon, from the air, in twenty sixteen. An independent historian, on foot, from a medieval chronicle, in twenty twenty-two. And a survey team, without a shovel, in twenty twenty-five. None of those three was sufficient by itself. A shape on a scan, a guess from an old text, a scatter of objects with no context — each one, alone, is just a maybe. What makes this a fortress, on the record, is that all three landed on the same ring of earth, independently, by different methods, years apart.

That is what 'ground truth' means in practice, and it is worth carrying to the next unexcavated anomaly on somebody's laser scan. A remote-sensing hit is a hypothesis, not a find. What turns it into a find is agreement — a second, independent method reaching the same conclusion by a different route, ideally before anyone digs. The published verdict on this one calls it, in translation, an absolutely unique object on the archaeological map of the Lublin region. And an almost untouched relic of an early medieval defensive site. That verdict rests on three readings agreeing, not one team's certainty.

The excavation permit for Kaplonosy-Kolonia has not been issued yet. What stands published, for now, is a shape read from the air, a shape read from a thirteenth-century chronicle, and a shape read from resistivity meters. Three readings of one ring of earth in a Polish bog, agreeing with each other. And agreeing with nobody who has stood on that ground in the last seven hundred years, except twice.

Next: a repair mistaken for the original

Next time, a repair mistaken for the original. At Rochester Castle in England, one corner of the great stone keep is round while the other three are square. The round one is not an architectural flourish. It is a scar: proof that in the year twelve fifteen, an underground fire fed with the fat of forty pigs brought the real corner down.

Description and sources

This channel's first entry in Ground Truth follows the discovery of a previously unrecorded early medieval stronghold at Kaplonosy-Kolonia, in the Włodawa forests of eastern Poland — and the messier, more honest story behind the headline that a laser "found" it.

An airborne LiDAR survey flagged the anomaly in 2016. Officially, nothing happened for six years. Unofficially, an independent historian named Dariusz Wołosiuk had already worked out roughly where a lost stronghold from this borderland's network of fortifications should sit — not from remote-sensing data, but from a medieval chronicle — and crossed the marsh alone in 2022 to check, two years before any professional fieldwork began. When the full non-invasive survey finally came, in the autumn of 2025, it went in almost without digging: multispectral aerial photography, refined LiDAR terrain modelling, electrical resistivity, and every recovered object mapped by precise GPS instead of a trench plan. Three independent methods, nine years apart, converged on the same ring of raised earth in a bog.

What came out of the topsoil — 244 catalogued objects, from 66 knives to a half-finished silver pendant abandoned mid-craft — is entirely early medieval, 10th to 13th century, dated by comparison against two excavated reference sites. Except for a handful of objects that aren't: spent German rifle cartridges and the crushed case of a pocket watch, from the 1940s, in the same soil. Local accounts collected by the survey team connect them to the German occupation — this same dry, defensible, hard-to-reach island sheltering people in hiding seven hundred years after the fortress that made it worth building on was abandoned. The paper documenting all of it is published by the museum in Włodawa's former synagogue complex.

PRINT-READY, FROM THIS CHANNEL

How to Read a Castle, Vol. 01 — the method, the measured drawings and the documents behind the first two episodes

https://therepository.gumroad.com/l/read-a-castle

PRIMARY SOURCES

Stone Record reads castles as documents — what the surviving paperwork and the standing masonry actually say, against what gets repeated about them. Primary sources for this episode are linked above.

Every image in this video showing the Kaplonosy-Kolonia site is an original survey-style illustration made for this channel — atmospheric renderings of the wetland terrain, not a reconstruction of the fortress itself, since no surveyed plan of the site has been published. No photograph of this site is public domain; none is used. The diagrams, the timelines and the quoted assessment are original drawings and typography made for this channel.

#archaeology #lidar #medievalhistory #poland #history