A California company is planning to launch a satellite that will monitor, and hopefully one day predict, the state's shakiest feature - earthquakes.

Dubbed QuakeSat by its Palo Alto-based creators, the microsatellite is tasked with watching the world for extremely low frequency (ELF) signals in the Earth's magnetic field -- signals that could be a precursor to an oncoming temblor. But first, it has to show that the signals have value as an earthquake indicator in the first place.

"We're obviously very optimistic, but we won't know until we try it," said QuakeFinder president and CEO Jeannie Seelbach in a telephone interview. QuakeFinder is developing the satellite as well as a ground-based earthquake detection network. "However, we think that this is something that has enough promise to be explored."

Despite its amiable goal, the QuakeFinder project does have its skeptics, including researchers with the United States Geological Survey (USGS) that monitor earthquake activity across the country.

Kinerney told SPACE.com that some past earthquake prediction efforts, like that of New Zealand's Geoforecasters, have been reluctant to share their methodology, making verification all but impossible by geologists.

But for QuakeSat, Seelbach and her colleagues are banking on past observations as precedent.

One such observation is an experiment by Antony Fraser-Smith, a Stanford University geophysicist, who was conducting experiments near the epicenter of the Loma-Prieta earthquake when it rocked Northern California in October 1989.

"He happened to get lucky and have a large magnetometer on hand for the quake," said Tom Bleier, QuakeFinder's chief technology officer.

The instrument detected a 20-fold ELF signal increase in local magnetic field, he added, and subsequent studies with the Russian satellite Kosmos 1809 and the French Ariel 3 appear have detected similar signals, but only after earthquakes occurred.

But USGS officials are skeptical of those space-based reports because of the difficulty in detecting earthquake-related ELF signals from space.

Theories over the cause of ELF fluctuations range from fracturing crystal deep in the earth to the movement of subsurface water, but the signals do not always precipitate temblors.

"It's not a slam dunk," said Bleier. "But hopefully we'll be able to capture some of these ELF in conjunction with an earthquake."

When ready for launch, QuakeSat will measure about 1 foot (30 centimeters) long and 4 inches (10 centimeters) in width and height. In orbit, however, it unfolds to nearly four times its length, counting solar panels and the two-foot telescoping magnetometer boom that will be used to seek out extremely low frequency signals from 0 to 1,000 hertz.

"It should be sensitive enough to detect these signals," said Robert Twiggs, a Stanford consulting professor in aeronautics and astronautics, who led the satellite's construction.

QuakeSat should circle the earth for about a year in a 522-mile (840-kilometer) polar orbit. Its off-shelf parts will succumb to the rigors of space soon after, Twiggs told SPACE.com. Despite its small demeanor, the satellite is actually a larger version of what Twiggs calls picosatellites, small 4-inch boxes roughly the size of a Rubik's Cube.

The QuakeSat effort is the result of collaboration between QuakeFinder, Stanford University, Lockheed Martin and the company Stellar Solutions, from which QuakeFinder is a spin off.

The driving force behind QuakeSat, and the study of earthquakes in general, is to find a way to predict them. Such a feat could save lives by alerting a community when an especially large temblor will hit.

While the USGS is studying a system that could give a 30-second warning of an impending quake, one that officials say could be ready by next year, but substantial advance notices of quakes remain elusive.

"The idea is to bring earthquake forecasting in line with something like the weather," Seelbach said. "If you think about what satellite technology has meant for hurricane forecasting, you begin to get an idea."

Bleier said QuakeSat's alternating current magnetometer should be capable of detecting ELF fluctuations on the picotesla level. One tesla is a measurement of magnetic field intensity. Measurements for the Earth are typically made in nanoteslas, an order of magnitude up from picoteslas.

It's the sensitivity of QuakeSat that's key. Critics are skeptical that once in space, it will be able to sift through the electromagnetic noise in space and key in on the minute changes along Earth's fault lines.

"It's not a trivial problem," said Malcolm Johnston, a USGS research geophysicist based in Menlo Park, California. "The technology is fine, we've been using magnetometers to measure magnetic fluctuations for the last 30 years. But if you're trying to measure them from space, well that's a long ways away."

Johnston told SPACE.com that the enormous energy release that accompanies an earthquake causes fluctuations of about 1 nanotesla, a tiny fraction of the 50,000 nanoteslas averaged each day by the Earth's magnetic field. The precursors to an earthquake, which are what QuakeSat will be looking for, are even smaller but can be detected with ground instruments.

There are daily variations to the planet's magnetic field too, with a range of about 30 nanoteslas, all of which doesn't include the electronic noise caused by people living their electronics-enhanced lives on Earth.

"Can they do it, I don't know," Johnston said. "We would like to see them do it." After all, a viable earthquake prediction system would be invaluable not just for geologists, but for the public-at-large.

Earthquakes typically occur along the boundaries of tectonic plates, huge masses of Earth crust that skate over a layer of molten rock. There are 12 plates in all where 90 percent of the world's earthquakes appear, according to the USGS.

Geologists use the Richter scale, a mathematical formula that measures the magnitude of an earthquake depending on the readouts of earth-shaking sensors called seismographs. Statistics compiled by the USGS since 1900 show that on the average, a major temblor of 8.0 or higher occurs somewhere in the world at least once a year. Major quakes measuring 7.0 on the scale occur every month or so and literally thousands of minor shakers from 3.0 on down occur each day.

QuakeSat isn't the only earthquake-hunting satellite set to circle the Earth. The French satellite Demeter is expected to launch on a similar mission in 2004. The Russian craft Kompass was also tasked with the same goal, but failed to communicate to ground scientists after its 2001 launch.

If QuakeSat succeeds in proving itself a viable way to detect ELF signals for earthquakes, Seelbach and Bleier plan to put up QuakeSat 2, a larger version capable of scanning wider range of ELF frequencies. Even if the current satellite comes up flat, future incarnations may be in order.

"It would certainly be easier to get support for future projects [if QuakeSat succeeds], " Seelbach said. "But we don't intend to give up."

QuakeSat is expected to launch later this year aboard a Russian rocket carrying 10 other small satellites into orbit.