From Andrew Alden,
Your Guide to Geology.
FREE Newsletter. Sign Up Now!

Atlantis! No geologist can hear the name without mixed emotions. The ancient legend, forwarded to us by Plato from Egyptian sources, is difficult to translate and analyze, and a thick crust of speculation and outright hooey surrounds any truth in it. On the other hand, it describes the destruction of an island city by earthquake and tsunami.

A conference in July 2005 gathered researchers, including geologists, on the Greek island of Milos to discuss Atlantis. (Read the Atlantis 2005 abstracts for a wonderful lunchtime browse.) At least a dozen candidates for the site of Atlantis were presented, ranging from Ireland to India. But to my mind the most interesting place is right where Plato said it was, "near the Pillars of Hercules," in a locality that spawns great earthquakes and tsunamis. (See some photos of the area here.)

The Quake, the Place and the Time

In 2002 a team led by French researcher Marc-Andre Gutscher reported in Geology on the Gulf of Cadiz, the bay just west of the Strait of Gibraltar—which all agree is the "Pillars of Hercules" Plato referred to. A detailed seafloor survey of the gulf showed the unmistakable pattern of a subduction zone and a thick wedge of sediment, or accretionary arc, building up in front of it. Gutscher wrote that this subduction zone "must be considered as a possible source of the great M>8.5 earthquake and tsunami that ravaged Lisbon and the Gulf of Cadiz in 1755."

At the Atlantis 2005 conference, Gutscher presented more recent research into a large submerged island in the Gulf of Cadiz named Spartel Bank. As it happens, he has published the same work in Geology. He took four key pieces of data from Plato's account: a date 9000 years before 600 BCE or 11,600 years ago; an island about 15 kilometers across; a location in the sea beyond the Pillars of Hercules; the sinking of the Atlantean city beneath the waves in a single day leaving a muddy shoal behind.

These facts fit a scenario based on new geologic data about that time and place. At that time, sea level was more than 100 meters below its present elevation and Spartel Bank was an island. In fact, Gutscher's new mapping of the site shows that it would have been a rather small island at that time, smaller than previously thought. But things change when we add the effects of large subduction earthquakes. As we all know from the Sumatra quake of 2004, large areas of land sink by several meters and more during these events. If we restore the effects of great earthquakes, which Gutscher estimated as recurring every 2000 years or so, then the island would have been higher and larger.

The Atlantis Scenario

Gutscher proposed that an exceptionally large quake could have dropped Spartel/Atlantis by 10 meters at once, while tsunami waves of 10 meters or greater height would have obliterated any human structures and left the island unrecognizable. A few more subduction earthquakes would have sunk the remaining islets beneath the sea, leaving treacherous muddy shallows, well before Plato's time.

"Although the catastrophic destruction described by Plato is consistent with the geological and tectonic history of the Straits of Gibraltar," Gutscher writes, "this does not imply that Atlantis ever existed. It simply means the account is geologically plausible." Proof might come with close-up submersible searches of Spartel Bank—or not. Whether Atlantis is ready to yield to modern science or not, it is thrilling to see new light shed on the possibilities.

PS: Since I wrote this, Archaeology Guide Kris Hirst sank Gutscher's scheme based on facts outside geology: no advanced civilization of any kind existed 12,000 years ago. Gutscher wrote her to say he agreed.