A submerged structure the size of a small town sits 120 feet under the Arabian Sea, off Gujarat’s coast. Two decades on, nobody has proven it’s man-made — or that it isn’t.The story doesn’t start with archaeologists. The site, located about 36 metres underwater off India’s western coast, was identified in December 2000 by the National Institute of Ocean Technology during a routine pollution survey — not an excavation. Sonar operators scanning the seabed for contamination data instead picked up something nobody expected: geometric patterns that looked less like seafloor and more like the outline of a city.That accidental discovery, off the coast near Bhavnagar in the Gulf of Khambhat, has spent 25 years bouncing between “biggest archaeological find in Asia” and “sand doing what sand does.”
How big are we talking?
Big enough that if it’s real, it rewrites the opening chapter of Indian history. The sonar data revealed formations stretching roughly 8 kilometres in length and 3 kilometres in width, with shapes that researchers interpreted as grid-like layouts and large rectangular structures roughly the footprint of a small city, sitting quietly under the sea since long before anyone was around to record it.A 2002 BBC News report noted that debris recovered from the site included pottery, beads, sculptures, sections of walls, and human remains which is exactly the kind of haul you’d expect from a settlement — if it hadn’t been sitting underwater for thousands of years.
The number that changes everything
Here’s the detail that turned a curious sonar anomaly into a genuine controversy: a piece of wood recovered from the site was carbon dated to around 9,500 years old.For context, that would place it well before the Indus Valley Civilisation, which historians generally date to around 3000 BCE. If the dating holds up, this wouldn’t just be an older Harappan outpost — it would be a civilisation nobody had accounted for at all, submerged by rising seas at the end of the last Ice Age. One geologist on the original NIOT team went as far as suggesting the site points to an advanced culture predating the Harappans entirely.
Not everyone is convinced and some serious names are skeptical
This is where the story stops being a clean “lost city” headline and becomes a real scientific argument.Finnish Indus-script scholar Asko Parpola has publicly questioned whether the dating methods used can be trusted at these depths, and pointed out that the Gulf of Khambhat’s heavy tidal currents constantly shift sand around — which could flatten or reshape natural rock into something that looks like a wall or a street on sonar, without a single human hand involved.Archaeologist Justin Morris has been blunter still, arguing that culturally speaking, there’s no evidence of civilisation in that part of the world before roughly 2,500 BCE — only small village settlements.Even researchers who find the sonar shapes hard to dismiss as purely natural, like Indus-script specialist Iravatham Mahadevan, have cautioned that artefacts recovered from the site may simply have drifted there on strong currents, rather than originating from a structure at that exact spot.
So what actually happens next?
Nothing conclusive, yet. No full excavation has ever confirmed intact walls or inscriptions in place, which is the bar most archaeologists say would be needed to call this a settlement rather than a geological coincidence. The debate has sat frozen for over two decades, revived every so often by fresh sonar passes and renewed media interest — this year included.What keeps it alive isn’t hard evidence either way. It’s the fact that a stretch of seabed off Gujarat’s own coastline could be hiding — or simply resembling — something older than every textbook date for civilisation on the subcontinent. Until someone sends a submersible down for a definitive look, the Gulf of Khambhat gets to stay exactly what it’s been since 2000: one of the most argued-about “maybes” in Indian archaeology.
