More than two decades ago, a survey crew working west of Cuba recorded sonar returns that appeared to form blocks, corridors and geometric patterns on the deep seafloor. The images triggered speculation about ruins, but the site has never received the sustained, published investigation needed to establish what the shapes represent.
The available record supports an unresolved sonar anomaly discovered around 2001. It does not support the idea that a current program is continually mapping the formation, which makes the present-tense description unreliable.
The original survey was searching for shipwrecks
The basic history summarized in the Cuban underwater formation record begins with marine engineer Paulina Zelitsky and a survey company using side-scan sonar near the Guanahacabibes Peninsula. The target was not an ancient city. The work was associated with deep-water exploration and shipwreck searches when the team noticed unusual high-contrast returns across an area of seafloor.
Reported depths were roughly 600 to 750 meters, or around 2,000 feet and deeper. At that depth, divers cannot inspect the site directly. A remotely operated vehicle later returned low-resolution footage of large stone-like features, but visibility, scale and limited coverage prevented a decisive geological or archaeological interpretation.
Side-scan sonar records echoes rather than photographs
NOAA Ocean Exploration explains that sonar sends sound into water and measures returning echoes. Side-scan systems are excellent for revealing texture and objects across broad strips of seabed. Their images encode acoustic intensity and shadow, not natural color or an ordinary camera perspective. Hard surfaces, steep slopes and edges can produce bright returns and dramatic dark shadows.
Geometry in a sonar mosaic can therefore be real without being architectural. Fractured bedrock, faulting, sediment scarps, volcanic formations and processing artifacts can all create lines and right-angle impressions. Researchers need bathymetry from several directions, sub-bottom profiles, close video, samples and a documented geological setting before deciding whether an apparent wall is cut stone, a natural joint or an artifact of viewing angle.
The proposed age creates a major archaeological problem
Some early speculation suggested that a structure now so deep would have been above water when sea level was much lower. Research on submerged prehistoric landscapes shows that ancient shorelines do preserve human sites, especially on continental shelves flooded after the last ice age. A depth near 600 meters is far beyond ordinary postglacial sea-level rise, however, so simple coastal flooding cannot explain it.
A human-built interpretation would require major tectonic subsidence or another geological mechanism capable of lowering the site hundreds of meters after construction. It would also require evidence of people with the organization and technology to build large stone works at the proposed time. No peer-reviewed excavation, diagnostic artifact assemblage or securely dated cultural layer has established those points.
Remoteness has allowed the mystery to outlive the data
Deep-ocean work is expensive. A capable vessel, mapping equipment, remotely operated vehicle, technicians and favorable weather can cost far more than a conventional land survey. Political and logistical constraints around Cuba add another layer. Those barriers help explain why an intriguing first pass did not automatically become a long research program.
They do not transform absence of follow-up into evidence for ruins. Scientific uncertainty means competing explanations remain open in proportion to the data. Natural geology is generally the starting expectation because blocky formations are common, while an archaeological claim requires artifacts, modified surfaces, coherent plans and dates that rule out natural processes.
A modern expedition could resolve much of the debate
A useful return would begin with high-resolution multibeam bathymetry and repeat side-scan lines collected from different directions. An ROV could then photograph selected targets with lasers for scale, inspect contacts between blocks and substrate, and recover samples under an approved research plan. Publishing raw maps, navigation data and methods would let independent specialists evaluate the geometry.
Until that happens, the site is best described as an old deep-water anomaly rather than a mapped city. Straight-looking edges are a reason to investigate, not proof of construction. The enduring story reflects how little has been measured since the first survey, not a continuing stream of new sonar discoveries.
Researchers would also need a transparent chain of custody for any recovered object. A pot fragment or shaped stone has little value if its coordinates, depth, surrounding layer and recovery method are missing. Deep currents and fishing activity can move recent material, so an isolated artifact would not automatically date the formation. Multiple objects in an undisturbed layer, tied to mapped structural features, would be much stronger.
The distinction protects both archaeology and public curiosity. Dismissing every anomaly from a distance would be premature, but promoting a city before baseline geology is known reverses the order of proof. The next meaningful update should follow new survey data, not recycle the dramatic interpretation of a low-resolution image collected more than twenty years ago.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
