Each claim below is graded on what the evidence supports, not on whether the person making it is sincere. Some of these are made in good faith by people who deserve respect. Sincerity and accuracy are different variables.
“The structure is not geological — it shows man-made corridors and chambers.”
Methodologically flawedClaimed by: Khosrow Bakhtar, reported by GlobeNewswire, 1 October 2026, and repeated in the project’s own release.63
What the evidence shows: The conclusion is produced by an acquisition setting that suppresses geological returns by design, so “not geological” cannot emerge as a finding.3 The instrument’s predecessor failed independent evaluation, with output attributed largely to equipment self-noise.18 The 3D “heat maps” the public sees are processed interpretations in which colour and geometry respond to operator-tuned parameters — not raw data.18
Also unaddressed: folded and jointed limestone produces orthogonal fracture sets naturally. Right angles at depth are expected in this geology, not anomalous.21
“There is no bedrock beneath the formation — so the syncline explanation collapses.”
Unverified — and unverifiable as presentedClaimed by: Andrew Jones, in multiple interviews.1327
The argument is coherent. All three mainstream explanations — mudflow around bedrock, glacial shaping of a limestone boulder, and the syncline — do assume solid rock underneath.27 If you genuinely found none, that would be interesting.
But nothing has been shown. No borehole logs have been published: no hole IDs, core runs, start and end depths, recovery percentages, material descriptions, or sample IDs. Without those, “we found no bedrock” is an assertion about data nobody can inspect.18 The published statement is that deep radar “could not identify a clear bedrock boundary” — which is a statement about instrument resolution, not about rock.28
“The bit shattered on something harder than limestone — possibly petrified wood.”
Unsupported inferenceClaimed by: Andrew Jones; announced in the project’s 23 September headline.2414
What it actually indicates: The bit snapped at four or five metres.24 The site is a lahar deposit on a stratovolcano. Such deposits contain cobbles and boulders of limestone, basalt, chert and greywacke. A high-torque auger hitting a buried boulder will shear its coupling — this is a known failure mode, not a discovery.18
And the precedent is bad: the material previously identified at this site as petrified wood was analysed and found to be basalt.24
“Charred wood was recovered from about 30 feet down. It is under radiocarbon testing.”
Genuinely testable — result pendingReported by: the project, first on 27 October 2026 and again on 10 November 2026, when it released the first public photographs and said the samples had gone to four laboratories in four countries. It also states that one piece shows marks “consistent with tool work”, shaped and flattened “using an axe or a knife”. Carbon-14 dating and species identification are still underway; the team states no conclusions should be drawn until results are published.331
Why this is the strongest claim they have: it is falsifiable, it is physical, and it is being handled by third-party laboratories.7 Unlike radar imagery, a carbon date and a wood anatomy report cannot be tuned.
Why you should still wait: the project’s own founder pre-empted the sceptical reading. Jones said it is possible they have found the charred remains of the Ark — “but it’s also possible that we have found the charred remains of a shepherd’s campfire that he unwittingly decided to start while using the ark for shelter centuries ago.”3 That is an honest sentence, and it should be quoted more often than the headline.
What would settle it: a date consistent with a mid-Holocene vessel, a species identification matching a known shipbuilding timber rather than local scrub, and — critically — evidence of working: cut marks, jointing, or tool traces. A burnt branch proves a fire. Only worked timber proves a ship.
“The radar found three layers — matching the three decks described in scripture.”
Circular reasoningClaimed by: Prof. Dr. Cenker Atila at the Doğubayazıt symposium: “In our Holy Books, Noah’s Ark had three levels, in our radar studies, we also identified a formation with three levels.”3
The problem: any folded, stratified or multiply-deposited formation has layers. Finding three of them and then matching the count to a text is not a test of the text — it is a coincidence being promoted to evidence after the fact. Had the radar returned four layers, the argument would have been abandoned, which means it was never an argument.
“At 515 feet, the formation is exactly 300 cubits — matching Genesis.”
Sincere, but not soundWhere it comes from: Ron Wyatt’s measurement against the royal Egyptian cubit of about 20.6 inches.12
Why it does not hold: the cubit was not one number. Mesopotamian, Egyptian, Hebrew and later rabbinic cubits all differ, and the Genesis text specifies no standard. Choosing the one cubit that makes the arithmetic land is a selection effect, not a measurement. A formation 515 feet long will be “exactly” some historical unit under some convention — that is what large numbers of units guarantee.
Credit where due: this is the least cynical claim on the list. It is the kind of pattern-matching every human brain does. It is still not evidence.
“Samples inside the formation contain three times more organic material — consistent with decayed wood.”
Ambiguous — by the project’s own admissionSource: the project’s own 2024 soil testing, unpublished.12
Read their own sentence: the project’s website states that the evidence “suggests either a natural geological fold or a buried decayed wooden ship.”12 That is a remarkably candid line, and it is the correct reading. A depression in a hillside collects vegetation, holds moisture, and accumulates organic matter. Higher organic content in a topographic low is the default expectation, not an anomaly.
“The enormous pierced stones at Arzap are the Ark’s anchor stones.”
Contested — the project concedes the problemThe claim: stones over ten feet tall, several tons, each with a hole bored through the upper end, identified by Wyatt as drogues trailed behind the vessel to steady it.12
The counter-argument, printed on their own site: “skeptics offer alternative explanations — that the stones are Armenian memorial markers (khachkars) or repurposed pagan stelae — and their precise age and original purpose remain debated.”12 Many carry carved crosses, which the project attributes to later Armenian veneration. But a monument tradition that carves crosses on standing stones is a fully sufficient explanation for standing stones with crosses on them.
“Sea shells and coral high in the mountains prove the Flood.”
Misreads the evidenceThe observation is real. Marine fossils are genuinely present in these mountains, and the project cites them.9
The interpretation is backwards. The formation itself is Miocene limestone — a rock laid down in a shallow sea roughly 15 to 20 million years ago, then uplifted by the same tectonic compression that folded it.20 The shells are evidence that this ground was once under water. They are evidence of millions of years, not of a single year a few thousand years ago. The fossils are the geology’s best argument against the timeline, not for it.
Where the sourcing itself is the story
Every claim above reached the public through a press release, a symposium announcement, a Substack post, or a tabloid headline. None has appeared in a peer-reviewed journal. The project’s own release acknowledges this limitation.24 Sample results are indicated for “late spring 2027,” and Bakhtar’s technical report was said to be possibly available in November 2026.67
That is the sequence to watch. Not the next headline — the next journal article. Until one exists, there is nothing here for a reasonable person to accept or reject, only to wait on.
Sources: see Chapter 9 — Sources & Reading for the full numbered list.