Groundwater drawn from beneath California's San Joaquin Valley may have helped keep your grocery store stocked with fruits, nuts, and vegetables, but the water's loss also may be contributing to the risk of earthquakes along the central San Andreas fault.
Over the past 150 years, the valley has lost enough groundwater through pumping, irrigation, and evaporation to fill Lake Tahoe. Groundwater levels have dropped by more than 300 feet in some areas, while the Earth's crust underneath has been relieved of more than 176 billion tons of weight.
And therein lies the rub, according to a team of geophysicists from the US and Canada who were studying the evolution of the Sierra Nevada range. Relieved of that enormous weight, the crust in the region has bulged upwards, raising the mountains that flank the valley – the Sierra Nevada to the east and, in particular, the Coastal Range to the west, through which the San Andreas runs.
A study by the team appearing in Thursday’s issue of the journal Nature suggests that the uplift has served to ease stress across the fault that clamps the two sides together, a change that has the potential to increase the rate of earthquake activity along the central section of the fault.
The team is the first to show how withdrawing water in the valley can alter stress patterns on a large fault, writes Paul Lundgren, a seismologist at NASA's Jet Propulsion Laboratory in Pasadena, Calif., in an e-mail.
Up to now, he says, this form of human influence on seismic activity hasn't been incorporated into assessments of stresses on faults and earthquake hazards.
The study shows "that such effects make a significant contribution," adds Dr. Lundgren, who was not involved in the study.
The results appear at a time when researchers with the US Geological Survey are working on ways to add another human activity to its national earthquake hazards assessments – the underground disposal of waste fluids from oil and gas operations, including fracking. In that case, water is pumped back into the ground, and if faults are nearby, the buildup of water pressure within a fault can counteract the stresses across the fault that would keep the segment clamped.
The impact of groundwater withdrawal on earthquake patterns was not something the research team had in mind when it started out, says Colin Amos, a geophysicist at Western Washington University and lead author of the paper describing the study results in Nature.
Using a dense network of GPS receivers set up in California to help assess earthquake hazards, the team was trying to "see what the modern-day signal of motion is for the mountains – up or down," he says. The question is part of a larger quest to better understand when and how the Sierra Nevada formed, the forces driving the uplift, and how modern motion might be related to this history.
The GPS data suggested "that the modern-day picture perhaps doesn't have very much to do with the long-term geologic process," he said. The highest rates of uplift appeared in mountains flanking the intensely farmed San Joaquin Valley. That led the team to ask if the extraction of groundwater could play a role.
The team then gathered data on groundwater depletion and plugged it into a model to see what the impact would be on the mountains on either side of the valley, based on the deformation that would result as the weight of the water is lifted from the crust. When the team set the model results against the GPS measurements of uplift, "we got a good match that showed us that this really could be the dominant signal that we're seeing," Dr. Amos said.
But what initially alerted the researchers to the potential impact on the San Andreas from the bulging of the Earth’s crust was a closer look at seasonal changes they observed in the San Joaquin in the height both of the flanking mountain ranges and in the level of the valley floor. Those fluctuations correlated with seasonal changes in earthquake activity in Parkfield, Calif., a town that sits on the central San Andreas.
The GPS stations, which can measure changes in height as small as half a millimeter, recorded a seasonal rise and fall to the mountains of about a centimeter, corresponding to dry and wet periods. During winter and spring, the weight of snow or accumulated rain pushes them down, while the valley floor rises as the aquifer recharges somewhat. The reverse happens during the hot, dry season in summer and early fall. Relieved of snow and water that has flowed into the valley, the mountains rise, while farmers and thirsty residents are drawing the water down, causing the valley floor to subside.
Most of this is natural activity year to year, but it still causes the crust to flex. When the mountains are pushed up during the dry season, the “clamp” across the San Andreas eases, then tightens again during the wet season.
This seasonal variation appears in the earthquake record at Parkfield, where swarms of tiny earthquakes come and go. During the dry season, when the mountains rise, earthquake activity picks up, then falls off again during the wet season.
Compared with stress released during an earthquake, the stress changes during these seasonal shifts are small, but potentially consequential.
The increase or decrease accounts for about 10 percent of the average clamping stress across the fault, estimates David Sandwell, a geophysicist at the Scripps Institution of Oceanography in La Jolla, Calif., who didn't take part in the study.
JPL's Dr. Lundgren notes that the change is comparable to changes in stress that a large earthquake on one fault can impart to another fault nearby.
The current seasonal 1-centimeter rise and fall is greater than the long-term annual average increase in the mountains' height, which is from 1 to 3 millimeters a year, but the net direction over time is up. That means the San Andreas is subject to a net reduction in clamping stresses with time, Amos says.
The paper is the first to suggest the influence of groundwater extraction in the valley on regional earthquake activity, and as such, it will be subject to scrutiny by other geophysicists interested in seismic hazards in the Golden State.
Moreover, Lundgren and Dr. Sandwell suggest that more-sophisticated models of the crust's behavior would help give a better sense of stress changes that groundwater withdrawal imparts.
Still, the mechanism the team describes for today's uplift rates is plausible, says Dr. Sandwell. "Also I think they have a compelling story about the modulation of seismicity by annual variations in groundwater."
Looking ahead, this may be a factor that seismic-hazard specialist will need to take into account, especially with the increasing competition for increasingly scarce ground water that global warming is expected to bring to the region, Amos and Lundgren suggest.