Image of Enceladus captured by Cassini. Credit: NASA/JPL/Space Science Institute

The water plumes erupting from the south pole of Saturn's moon Enceladus could be caused by a liquid ocean lurking many kilometres underground rather than by geysers erupting from a salty ocean just beneath the moon's surface — as early theories suggested.

Evidence that Enceladus could hold a subsurface ocean would be exciting because liquid water elsewhere in our solar system is the most promising place to look for signs of life.

The absence of sodium in the plumes pouring out of vents at Enceladus's south pole rules out geysers from water just below the surface. Nick Schneider at the University of Boulder, Colorado, and his colleagues used the Keck telescope on Mauna Kea, Hawaii, and the Anglo-Australian Telescope in Siding Spring in Australia to look for atomic or molecular sodium both in the plumes and in one of Saturn's rings thought to be fed by these plumes, the E ring.

If geysers caused Enceladus's plumes, then they would throw up salt-rich jets giving out a strong sodium signal that could be spotted by ground-based telescopes. But Schneider had trouble finding the familiar orange emission: "We stared at Enceladus for hours and hours and we saw nothing," he says. The team reports its findings in Nature1. "I'm a little disappointed," he says, "it would have been really cool if it had verified the near-surface ocean geyser theory."

Warm core

An alternative explanation is that a salty liquid ocean exists deep underground and is evaporating — releasing pure water as a jet of steam and leaving the salty residue behind.

Frank Postberg at the Max Planck Institute for Nuclear Physics in Heidelberg, Germany, likes the ocean idea. He and his colleagues used cosmic-dust data from the Cassini spacecraft currently flying around Saturn to study grains in the E ring. They found that a small percentage of those ice grains, up to 2%, were sodium rich, or salty2.

Salt-rich grains are directly frozen ocean water dragged up by strong vapour flow. Frank Postberg , Max Planck Institute for Nuclear Physics

These salty grains can only exist if there is a liquid ocean close to Enceladus's mineral-rich rocky core, deep beneath the thick icy crust, says Postberg. The crust could be around 50 kilometres thick. Postberg says the vapour evaporating from the ocean will also contain other gases, and bubbles of those gases carry salty water droplets through vents in Enceladus's crust, to be suddenly frozen once they get there. These grains travel out into space in the plumes along with salt-poor ice grains that are formed like snowflakes from pure water vapour.

"Salt-rich grains are directly frozen ocean water dragged up by strong vapour flow," says Postberg. The absence of sodium in the plume does not contradict his theory because the small amounts of sodium he would expect to see would not be detectable using the Keck telescope.

Cracking up

But other models exist to explain Enceladus's plumes apart from oceans or geysers. One of these suggests that reservoirs of clathrates — gassy molecules locked up in the lattice of another molecule — exist below the surface. As tectonic plates in the crust move and collide, the crust fractures and these clathrates release gases, which carry up ice particles with them to form the icy plumes. These ice particles could carry up salt as well, says Susan Kieffer a geologist at the University of Illinois at Urbana-Champaign. "Sodium isn't the proof of a liquid ocean," she says. "Sodium can be locked up in the ice of an icy clathrate model."

Kieffer also says that the other gases present in the plume — including methane, carbon dioxide and nitrogen — can only be accounted for with the clathrate model.

Francis Nimmo, a planetary scientist at the University of California, Santa Cruz, had suggested that the plumes came directly from the solid ice shell, heated by sheets of ice rubbing against one another. He has now been persuaded to think again. "I'm ready to accept there's an ocean and move on," he says.

But there are still other questions to answer, says Nimmo, such as the absence of sodium in the plumes. "There really ought to be some there," he says. Nimmo also doubts that Enceladus is warm enough to keep water liquid for so long. "That ocean should not have survived over the length of the Solar System," he says.

Schneider says that although he agrees that deep oceans explain much of the data in the two papers, he is keeping an open mind. The clathrate model also has "a lot of viable features", he says. "I'm still a little more sceptical about how firm our conclusions are" that a liquid ocean exists, he says. "I'm not convinced we have one theory that explains everything clearly."