Scientists find evidence the Earth's crust ripped apart to create a 7km deep abyss
Huge parallel grooves in rock 7km under the sea are providing clues to the formation of the deepest point in the ocean that isn't within a trench.
Scientists have known about the massive chasm, known as the Weber Deep, beneath the Banda Sea in eastern Indonesia for 90 years but have been puzzled by how it got to be so deep.
Now researchers believe the abyss was formed when a piece of the Earth's crust measuring about 60,000sq km - around half the size of the North Island - was ripped apart by 120km of extension along a low angle crack, or detachment fault.
The fault plane itself became exposed, lead researcher Dr Jonathan Pownall from The Australian National University (ANU) said.
The fault, being called the Banda Detachment, has enabled the crust to be pulled apart and the deep hole to form. It represented a rip in the ocean floor exposed over 60,000sq km.
"We think it's been ripped apart over maybe two million years," Pownall said.
High-resolution maps of the Banda Sea floor revealed the grooves in the rock at the bottom of the abyss.
"We can demonstrate that because everything is parallel ... they are all in response to the same external force," Pownall said.
The grooves were somewhat like corrugations, between about 2km and 10km apart. The longest continuous groove was about 60km long, and the steep cliff sides of some grooves were hundreds of metres high.
Some of the rocks exposed in islands around the Weber Deep had come from the upper mantle - the layer in the Earth below the crust.
"This fault has to have thinned the crust a lot to bring rocks from the inner part of the Earth to the surface," Pownall said. In some places the oceanic crust was thinned, "in some places to zero".
It was not known whether the feature posed earthquake or tsunami risks. No earthquake from the past century could be attributed to the fault.
The fault might be dead, or it might operate by gently slipping apart, Pownall said. Another possibility was that it could cause large earthquakes but not often.
"We need to stress that's just a possibility. We're not in a position to assess how this fault works yet."
As for tsunami, huge landslides - one 100km wide - of loose rock and sediment had collapsed down the western side of the fault. That was one way a tsunami could potentially be caused.
"There are historical tsunami recorded in this region. Many of these things we can't attribute yet to a cause. There's a possibility this could be it, but it's a stretch to say with any certainty," Pownall said.