05/14/2003 07:00:00

Caltech planetary scientist has "modest proposal" for sending probe to Earth's core

PASADENA, Calif. - Dave Stevenson has spent his career working on "swing-by" missions to the other planets. Now he has a modest proposal he'd like to swing by some government agency with a few billion dollars in available funding.

According to Stevenson's calculations, it should be possible to send a probe all the way to Earth's core by combining several proven technologies with a few well-grounded scientific assumptions about the workings of the planet. The probe would sink straight to the core in an envelope of molten iron, sending temperature readings, compositional information, and other data along the way.

"We've spent more than $10 billion in unmanned missions to the planets," says Stevenson, who is the Van Osdol Professor of Planetary Science at the California Institute of Technology. "But we've only been down about 10 kilometers into our own planet."

The benefits to science would be significant, Stevenson says, because so little has been directly observed about the inner workings of the planet. Scientists do not know, for example, the exact composition or even the temperature of the core, and what they do know is based on inferences about seismic data accumulated during earthquakes.

Stevenson says his proposal should be attractive to the scientific community because it is of the same scale, price-wise, as planetary exploration. To date, NASA has flown unmanned missions past all the planets except Pluto (if indeed Pluto is a planet at all), has made a few highly successful soft landings on Mars, has probed the clouds of Jupiter, is getting ready to probe the atmosphere of Titan, and has sent four spacecraft into interstellar space. Sending something into the earth, Stevenson believes, will have comparable payoffs in the quest for knowledge.

"When we fly to other worlds, we are often surprised by what we find, and I think the same will be the case if we go down."

Stevenson's plan calls for a crack to be opened in the earth, perhaps with some sort of explosion-probably a nuclear bomb. According to his figures, the crack will need to be several hundred meters in length and depth, and about 30 centimeters wide, to accommodate a volume of about 100 thousand to several million tons of molten iron.

The instant the crack opens, the entire volume of iron will be dropped in, completely filling the open space. Through the sheer force of its weight, the iron will create a continuing crack that will open all the way to the planet's core 3,000 kilometers below. Anything on a smaller scale may not work; anything larger will be even more expensive, so Stevenson thinks a crack of those dimensions is about right.

"Once you set that condition up, the crack is self-perpetuating," he explains; "it's fundamentally different from drilling, where it gets harder and harder-and eventually futile-the farther you go down."

The iron will continue to fall due to gravity because it is about twice the density of the surrounding material. Riding along in the mass of liquid iron will be one or more probes made of a material robust enough to withstand the heat and pressure. The probe will perhaps be the size of a grapefruit but definitely small enough to ride easily inside the 30-centimeter crack without getting wedged.

Inside the probe will be instrumentation for data collection, which will be relayed through low-intensity mechanical waves of some sort-probably through deformations of the ball itself to send out a sort of "Morse code" of data. Because radio waves cannot propagate through Earth, this is the only way to get the data transferred.

The probe will likely operate with about 10 watts of power, and it may even be possible to replenish energy and dispense with an on-board battery by harnessing mechanical energy from the force of the fall, just as electricity can be generated from falling water.

Such a low power rating will not make it possible to generate very strong shock waves for data transmission, but strong waves may not be necessary. In fact, Stevenson further suggests that the Laser Interferometer Gravitational-Wave Observatory (LIGO) might be recalibrated in its downtime to track the falling ball.

Based on the rate the molten iron would fall due to gravity, the ball would move downward into Earth at roughly human running pace (about 10 miles per hour), meaning that the entire mission would last a few weeks.

All this may sound to some like science fiction, but Stevenson says each of the principles involved is based on sound knowledge of crack propagation, fluid dynamics, mechanical-wave propagation, and "stress states." If these things didn't already work in nature, we would have no volcanoes and poorly performing bathroom plumbing, but little to fear from a pebble shattering our windshields.

"The biggest question is how to initially open the crack," says Stevenson. "Also, there's the technological challenge of having a probe that actually does what it's supposed to do."

Stevenson says he came up with part of the title "A Modest Proposal" for his paper, which is appearing in this week's journal Nature, to have a bit of fun but at the same time to issue a serious scientific proposal. He purposely took the title from Jonathan Swift's famous essay of the same name. The Swift essay suggests that Ireland's terrible economic circumstances could be solved by people eating their own children, thereby allowing England to continue pillaging the country's resources for its own one-sided benefit.

"My proposal is not as outrageous as suggesting one should eat his own children, but still combines a serious proposal with some levity," Stevenson says. "Ninety-five percent of the scientists who read the article may laugh at an enjoyable read, but if the other five percent seriously consider the goal of probing Earth's core, then I'll be happy."

"The biggest question should not be the cost, but whether we should pursue the goal of exploring Earth's interior," he says. "That said, I'd suggest we do it if we can keep the cost under $10 billion."

Contact: Robert Tindol (626) 395-3631