"A generation ago, people thought we were right around the corner from being able to predict earthquakes," says Dr. Eliza Richardson, assistant professor of geosciences, Penn State. "But we do not understand how earthquakes work and we will not until we understand the little ones better. We need to know how they nucleate and how they propagate. We do believe this is the same for the big earthquakes and the little ones."
The researchers are investigating small earthquakes in the zero to magnitude 3 range.
"There is a break in knowledge about seismic activity," Richardson told attendees today (May 18) at the spring meeting of the American Geophysical Meeting. "What comes out of the laboratory and what is measured in the real world by surface seismographs leaves a gap. We cannot measure the very smallest earthquakes."
The researchers, who include Richardson, Andrew A. Nyblade, associate professor of geosciences from Penn State, and William R. Walter and Arthur J. Rodgers, seismologists, Lawrence Livermore National Laboratory, are using data collected in South African gold mines owned by AngloGold Ltd. The U.S. Department of Energy is funding this research as part of attempts to enforce the Nuclear Non-Proliferation Treaty.
Because of the deep mining in South Africa, many small earthquakes are recorded in South Africa. The DOE would like to have a profile of what mining seismic activity looks like so they can tell the difference between mining and other explosions. At the same time, the Penn State researchers want to know more about small earthquakes.
The AngloGold mines are about two miles deep and blasting takes place every day. The seismic detectors are placed in the mine tunnels to monitor both the activity of the explosions and any earthquakes that occur separate from the blasting. The detectors record thousands of events each day. While AngloGold uses the data only for safety purposes, they allow the seismologists to use the data. Ground level seismic arrays typically record earthquakes larger than about magnitude 3, but miss anything smaller. Laboratory experiments have suggested that earthquakes cannot be infinitely small, but that there has to be a smallest earthquake.
"From laboratory work, zero corresponds to the smallest seismic event expected," says Richardson. "That is just enough energy to get over the activation barrier and move a very small distance and then stop."
The earthquakes recorded in the mine represent two different types of events. The first group, recorded at the time of the blasting, generates negative magnitude numbers (something never seen in nature) and is caused by the actual breaking up of the rocks from the explosion. The second group, which occurs hours to weeks after the blast, produces earthquakes from magnitude zero to about 3. These earthquakes are caused by the build-up of stresses from the mining operations on the long dormant faults lying beneath the ground. "The nicest thing about using the mine data is that we know we are seeing all the events from zero to three because we are seeing the even smaller events caused directly by the blasting," says Richardson. "We know that it is not the limit of the equipment, but that the lower limit of earthquakes is around zero."
Because the researchers are confident they are recording all the earthquake activity, they can investigate another earthquake question. Large earthquakes happen infrequently compared to small earthquakes and, for earthquakes recorded above magnitude 3, the magnitude of the earthquake is related to the number of that magnitude that occur.
"The mine data shows fewer events than expected at the lower numbers," says Richardson. "There is a scaling breakdown that we see in the laboratory and now we see in the mine data."
In both the numbers of earthquakes and the critically small size of earthquakes, the researchers are using the mine data to bridge the gap between what can be seen in the laboratory and what is recorded by seismic arrays on the surface.
"We are getting to the intersection of where laboratory work and field work meet and provide a continuum," says Richardson.