A start-up company has raised seed money and collaborated with a top-notch university to launch a satellite this summer aimed at detecting signs of earthquake activity in California with the help of ground-based sensors.

The 10-pound micro-satellite is scheduled to be shipped soon from California to Canada and then on to Russia for launch between June 30 and July 10. The satellite was designed and built at a low cost by Stanford University and Palo Alto, Calif.-based QuakeFinder.

The QuakeSat I satellite is scheduled to lift off aboard a rocket from the Plesetsk launch facility in Russia, said Jeannie Seelbach, president and CEO of QuakeFinder, a spin- off company of Stellar Solutions. The parent company, also of Palo Alto, is involved in aerospace engineering services and is among the venture’s financial supporters.

QuakeFinder’s goal is to provide earthquake-forecasting data that will offer the same kinds of advanced warning that meteorologists are able to provide before hurricanes and tornadoes strike, Seelbach said.

QuakeSat I’s small size still allows room for a one-foot long magnetometer that extends on a telescoping boom, company officials said.

To save costs and provide educational outreach, QuakeFinder asked high school students to deploy 50 sensors along major faults in California that would work in tandem with the satellite. The students would monitor data from the sensors on a Web site, company officials said.

The satellite would be combined with technology that relies on extremely low frequency magnetic field fluctuations to detect early signs of earthquakes. This magnetic field noise has been observed before past earthquakes and could be an untapped resource for detecting future ones, said Seelbach, who holds a PhD in mathematics and offers a background in the signal-processing technology used on the satellite.

Creative Financing

QuakeFinder secured early-stage financing in creative ways. The satellite probably would have cost at least $1 million to build under normal circumstances, but QuakeFinder collaborated with Stanford University students to help cut the construction costs in half. The $500,000 needed was raised from private investors interested in the business, Seelbach said.

The sensor kits provided to the high school students for deployment will be used to detect indications of potential earthquakes. The sensors were developed with a $50,000 grant from the state of California and $10,000 from Stellar Solutions, Seelbach said.

Roger Rusch, president of the TelAstra satellite consultancy in Palos Verdes, Calif., complimented QuakeFinder on finding cost-effective ways to provide a “very interesting” satellite application.

“People have been looking for precursors to earthquakes,” Rusch said. “This has been a very important field for years. Even a little bit of advance warning can save a lot of lives.”

But can QuakeFinder become a viable business without the government being either a major financial backer or a customer?, Rusch asked. A potential revenue source could be companies that would subscribe to an advanced warning system to protect their property and employees, he said.

“This is one of the best examples I’ve ever seen of a cost-effective technology being developed that could succeed because it is so economical and the potential benefit is so high. It is quite a contrast to the enormously expensive satellite systems that have been built on the assumption of a huge market. This project is so cost-effective that it can be tried out without a large financial risk,” Rusch said.

Kevin Leclaire, an associate with the venture capital firm SpaceVest, of Reston, Va., said he was impressed with QuakeFinder’s ability to produce a satellite system so inexpensively.

Corporate Challenges

At the same time, Leclaire sees a number of challenges that QuakeFinder will have to overcome, including establishing the efficacy of the system in predicting significant earthquakes in advance, determining whether the system would compete with or augment existing networks of seismic sensors, and importantly, finding paying customers if the system proves to be effective.

One technical issue is that a low-earth-orbit (LEO) satellite like QuakeSat I is not able to provide coverage 24-hours, seven days a week, Leclaire said. To the extent that full coverage is necessary to guarantee early warning about a quake, more LEO satellites would be required for the system, he added.

The satellite would only be able to pick up magnetic field noise fluctuations in real time when it orbits directly over California, said Marshall Kaplan, a Bethesda, Md.-based satellite consultant. QuakeFinder explained that if the satellite detected noise fluctuations ten to 14 days before an expected earthquake, the company could rush portable ground sensors to monitor a high-risk site.

QuakeFinder’s chief mission would be to detect earthquakes days in advance to provide warnings that could save lives. At the same time, unless property can be moved quickly from an area where an earthquake is expected to hit, much of the damage caused to physical assets will be unavoidable. That limitation may reduce the information’s commercial value.

Quake Losses

The Federal Emergency Management Agency (FEMA) has estimated that annual losses caused by earthquakes in the United States approach $4.4 billion, with 84 percent of that damage occurring in California, Oregon and Washington, Leclaire said. Any system that could aid in the prediction of earthquakes would be of value not only to FEMA and local governments, but also to certain businesses, he added.

David Logsdon, manager of space operations at the Aerospace Industries Association of America, said that satellites already help in earthquake detection and damage assessment through the use of the Global Positioning System (GPS). Real-time GPS networks can help to detect earthquakes by using techniques to map horizontal movement of the Earth’s surface by as little as a few millimeters. This data is “critical” in understanding future movement of the Earth’s surface, Logsdon said.

That system was an integral part in developing a database of damage caused by the earthquake that hit Kobe, Japan, on Jan. 17, 1995, Logsdon said. The Kobe earthquake, assigned a magnitude of 7.2 on the Richter scale by the Japan Meteorological Agency, claimed nearly 5,400 lives and injured tens of thousands of people. Direct damage caused by the quake topped Y13 trillion (U.S.$150 billion). However, those losses do not include indirect economic effects from loss of life, production and business revenues, according to the National Information Service for Earthquake Engineering at the University of California, Berkley.

“Since there is precedent for this type of service, the company could use lessons learned from the GPS community to make their product more attractive to potential investors and customers,” Logsdon said.

Part of QuakeFinder’s challenge is to obtain sufficient data from its sensors and satellite to show convincingly that fluctuations in magnetic fields are “reliable precursors” to earthquakes, Seelbach said.

The Loma Prieta earthquake in the Santa Cruz Mountains that rocked San Francisco on Oct. 17, 1989, had the unexpected effect of demonstrating that magnetic field noise fluctuations could help to detect early signs of an impending earthquake. Antony Fraser-Smith, a geophysicist at Stanford University, placed an antenna three miles from what later became the earthquake’s epicenter in Corralitos, Calif. His objective was to measure the Earth’s natural magnetic field noise. Almost two weeks before the magnitude 7.0 earthquake, his antenna identified an abrupt increase in such noise.

Magnetic field noise grows prior to the earthquake but still fluctuates up and down, said Tom Bleier, QuakeFinder’s chief technology officer. In the Loma Prieta quake, the magnetic field noise spiked at 20 times the normal background noise level, then fell to between 10 and 15 times normal before soaring to 60 times the norm hours before the earthquake struck, he explained.

That data was collected and stored on a local computer, so it was not available in real-time to allow scientists to issue any warnings, Bleier said. Only after the earthquake was the data seen and analyzed, he added.

Saving Lives?

The Loma Prieta earthquake, described as “moderate” by scientists, caused 63 deaths and 13,757 injuries. Property losses amounted to 1,018 homes destroyed, 23,408 homes damaged, 366 businesses destroyed, and 3,530 businesses damaged. The total estimated loss was valued at more than $5.9 billion in public and private property damage, according to University of California statistics.

The research strongly suggests that magnetic field data could be used to predict earthquakes. However, additional data is needed to convince investors that the business is based on sound scientific reasoning and that the combined satellite-terrestrial service is worth expanding to other parts of the world, Seelbach said.

“The earthquake signals follow a unique path from the ground to the satellite that would allow us to detect whether an earthquake will occur,” Bleier said. “We think there is a signal hours to days before a significant earthquake occurs.”

Each ground sensor would operate 24 hours a day, but has the limitation of only monitoring magnetic field noise fluctuations within a diameter of 20 miles, Bleier said. The satellite uses a sensor similar to the ones in the ground units, but collects information over much wider geographic areas.

This is a huge advantage, Bleier said. At the same time, the satellite’s usefulness is limited because it would pass over California only about every seven to 14 days, he added.

“A combination of satellites and ground-based sensors could be very powerful,” Seelbach said. If the technology proves to be a reliable predictor, the company would attempt to offer its earthquake detection equipment or data for sale, she added.

“We would be happy if the government ends up taking over the research,” Seelbach said. “We’d like to be part of a public/private partnership that could take many forms.”

Seelbach said the use of technology to predict hurricanes was developed gradually with government help.

–Paul Dykewicz

(Jeannie Seelbach, QuakeFinder, 650/838-0949; Tom Bleier, QuakeFinder, 650/473-9870; Roger Rusch, TelAstra Inc., 310/373-1925; David Logsdon, Aerospace Industries Association, 703/358-1096; Kevin Leclaire, SpaceVest, 703/904-9800; Marshall Kaplan, 301/573-1498)