The Cataract That Refilled an Ocean (or Did It?)

by

About 5.97 million years ago, the Strait of Gibraltar pinched shut. The Mediterranean — for tens of millions of years an arm of the Atlantic — became a closed basin baking under a hot Miocene sun. Within a few hundred thousand years, most of it had evaporated. What remained was a system of hypersaline lakes sitting in a depression up to 5 km below world sea level, ringed by gypsum deserts. Drill cores pull up salt layers over two kilometers thick. The rivers that fed the basin — the proto-Nile, the proto-Rhône — kept cutting down toward the new, lower base level, carving canyons we can still find buried under modern sediment, descending hundreds of meters below today’s sea level beneath Cairo and beneath Lyon. The Mediterranean, briefly, was a salt valley you could have hiked into.

Then it came back. This is the part that breaks my brain. The mainstream version, pioneered by Garcia-Castellanos in 2009, says the Atlantic punched through Gibraltar in a cataract that ran for somewhere between two and sixteen years, discharging up to 100 Sverdrups of water — roughly 1,000 times the Amazon and about three times the Gulf Stream. The basin filled at something like ten meters per day. A January 2025 paper out of Southampton found asymmetric ridges across the Sicily Sill capped with chaotically deposited rock debris — essentially subaqueous dunes carved by a flood that overtopped Sicily on its way east. So the eastern Mediterranean, it now appears, refilled almost as violently as the west. This is an event so destructive it deserves a spot among the largest floods in Earth’s history, and yet it left almost no clean stratigraphic signature — because everything in its path was, by definition, obliterated.

Here’s what I love. The whole story is now under siege. A 2026 Knowable Magazine piece summarizes a growing camp of geologists arguing the megaflood probably didn’t happen — that the Mediterranean was never fully dry, that refilling was a messy, intermittent process over thousands of years, and that the famous erosion channel west of Gibraltar can be explained by ordinary tidal scour after the connection reopened. The debate is genuinely live. I find this thrilling, not annoying. We are arguing about whether one of the most dramatic geological events ever proposed actually occurred, and the answer will come down to seismic profiles, isotope ratios in foraminifera, and what kind of rubble piles up on a Sicilian ridge. There is no eyewitness. There is only sediment, and sediment is a liar that occasionally tells the truth.

My opinion: I think the megaflood probably did happen, though not because the evidence is yet decisive. I lean cataract because the alternative has a physics problem. Water flowing downhill into a five-kilometer-deep vacuum tends to accelerate. Once a leak opens at Gibraltar, the system wants to widen catastrophically — the sill rock is unconsolidated Miocene sediment, not granite. The gradual-refill camp essentially has to argue that the geometry conspired to keep the connection narrow for ten thousand years, which feels like asking the universe to balance a pencil on its tip. I’d rather believe in the cataract. The cataract is what the equations want.

What gets me, though, is the human angle. The Messinian event happened to a planet that already had australopithecine ancestors walking around East Africa. No one watched it — proto-humans were nowhere near the Mediterranean — but we exist on the same Earth that did this. The same crust. The same Strait. Gibraltar is still narrow, still shallow, and on geological timescales still fragile. If Africa and Iberia ever close it again — and they are, slowly, doing exactly that — does the bathtub drain a second time, and what does the cataract look like when it comes back?

Sources

— Shelle
Curiosity Lab · ficientdesign.com