The search for a solution to the climate crisis demonstrates that we need to know a lot more about what lies beneath us. Perhaps ironically, the deepest such knowledge resides in the fossil fuel industries themselves, which for many years have injected CO2 routinely into oil and gas wells to push hydrocarbons to the surface. Doing this, however, has not necessarily required the CO2 to stay buried for the lengths of time required to slow the greenhouse effect. The Department of Energy's 2006 Carbon Sequestration Roadmap sets a goal of less than a 1 percent escape after 100 years. An earlier escape would nullify the benefits of the sequestered CO2, not to mention that CO2 would be emitted during the process of injection, McGrail says.

Although generally considered nontoxic, CO2 can harm vegetation and subsurface organisms. Soil organisms are adapted to the naturally higher CO2 content of their environment, but they can be killed by higher-than-normal concentrations. In large volumes, of course, CO2 can be lethal to large organisms, too. In 1986, a massive natural CO2 release from the bottom of Lake Nyos in Cameroon suffocated about 1,800 people.

Explosive releases of CO2 from sequestration sites are highly unlikely, provided the target formation is properly characterized and monitored, Peridas says. The main concern is that leaks would return CO2 to the atmosphere, reversing the benefits of the process.

Another worry about injecting CO2 involves the contamination of shallow groundwater used for drinking and irrigation. The very mineral leaching that is desirable for sequestering CO2 deep underground would make drinking water unpalatable, so it's imperative to determine the risk of communication between aquifers of different character. Most injection will be into deep aquifers that are already too brackish or saline for human use due to eons of chemical reactions with the rocks. But if underground equilibrium is disturbed, pressure from injected CO2 might "add the energy that could allow the mineralized water to migrate into one of the shallow aquifers," Stormon says.

Or, in standard English: Carbon dioxide injection could cause undrinkable, salty aquifers to contaminate groundwater consumed by humans.

And earthquakes may also pose problems: The increased pressure of intruding CO2 on rock formations can trigger an earthquake, or a natural earthquake could enlarge existing cracks and faults or create new ones in sequestration zones. Most geosequestration researchers, including McGrail, consider earthquakes a minor risk in the Columbia basalt. But there have been instances of induced quakes elsewhere. In the 1960s, the U.S. Army injected about 165 million gallons of liquid toxic waste from munitions manufacture into a formation under the Denver basin. From 1962 to 1967, there were some 1,500 seismic "events" centered on the injection area, including three earthquakes at or above Richter magnitude 5.

As to natural seismicity, McGrail says, extensive monitoring at Hanford shows that most earthquakes in the region are weak, sparse and random. The Columbia basalt is a few hundred miles from the complex subduction zone along the Pacific Coast, where the North American plate, the Juan de Fuca plate and the Pacific plate are sumo wrestling. Dunning, who is involved with Hanford safety issues, says when that zone experiences its next magnitude 9 earthquake - due any time - the Hanford area would be rattled by a "5-plus" shock. It's not clear what effect this would have on the Columbia basalt.

McGrail's confidence in the safety of basalt sequestration has been strengthened by tests showing that air and water have remained trapped in basalt pores for millions of years without mixing between formations. McGrail says this is good evidence that the basalt has been "undisrupted by seismicity," and that the injected CO2 will not migrate out of its target formation.

Still, many environmentalists are concerned about potential side effects. They are even more worried about other aspects of carbon sequestration.

"If (sequestration) were a simple solution to a complicated problem," says Sierra Club spokesman Josh Dorner, "people would already be using it." Still, he says, "we would love it if it turned out to work." The Sierra Club opposes new coal-fired power plants unless they sequester 100 percent of their emissions, Dorner says.

"It will take so much time to test it," Friends of the Earth spokesman Nick Berning notes. "There's an urgency about addressing global warming that demands that we take steps that can make a difference now, moving to cleaner energy, wind, solar, and conservation. We know wind power can generate energy with zero carbon emissions. We know solar can."

Regardless of any given site's geology, nobody knows how well the earth's crust will tolerate being saturated with CO2 in the volumes necessary to slow global warming. For his part, McGrail stresses that many basic questions must be satisfactorily answered before any large-scale CO2 sequestration in basalt can begin. His data, he says, "are compelling, but I would not regard it as sufficient proof of analogous isolation for CO2 storage until we have some field data to support it."

Today, Richland looks much like many of the other Western cities that have been transformed by rapid growth in the last 20 years. Along the Columbia in the middle of town, joggers traverse a pleasant esplanade, past gaggles of Canada geese and their goslings. Young people drift along downtown sidewalks wearing Goth markers: black lipstick, neck chains, body piercings. Traffic is congested. Every summer there are hydroplane races on the river.