USGS models predict severity of man-made earthquakes

WASHINGTON, April 23 (UPI) -- Man-made earthquakes are on the rise in Middle America, where industrial practices like fracking and deep-well wastewater storage are increasingly common.

In report released this week by the U.S. Geological Survey, scientists detailed a number analytical models they hope will better predict the severity of ground-shaking caused by man-made earthquakes.

The report highlights 17 areas within eight states -- Alabama, Arkansas, Colorado, Kansas, New Mexico, Ohio, Oklahoma and Texas -- which have experienced increased seismicity over the last several years. Researchers constructed the new prediction models by looking at earthquakes in these areas and plotting their size and power relative to nearby industrial triggers.

"This new report describes for the first time how injection-induced earthquakes can be incorporated into U.S. seismic hazard maps," explained Mark Petersen, who is heading up the USGS National Seismic Hazard Modeling Project.

"These earthquakes are occurring at a higher rate than ever before and pose a much greater risk to people living nearby," Petersen said. "The USGS is developing methods that overcome the challenges in assessing seismic hazards in these regions in order to support decisions that help keep communities safe from ground shaking."

The study pins most of the blame for a rise in man-made ground-shaking on wastewater injection practices. Though fracking has been linked with slight tremors (noticed only by seismic needles), USGS say it is only occasionally responsible for felt earthquakes.

Wastewater injection is a practice used to store toxic water leftover from a variety of industrial practices, including energy production. The dirty water is injected deep underground where it presumably cannot contaminate fresh water closer to the surface.

Researchers believe some of the injected wastewater may serve to lubricate fault lines, making earthquakes more likely.

Scientists at USGS say their new models aren't yet ready to inform decision making, but are experimental in nature. They hope to perfect the models in the coming months and deliver a more precise hazard model by the end of the year.