The study is published in the journal Seismological Research Letters.
The earthquakes occurred in Lake Erie's central basin near Conneaut, Ohio; two of the quakes registered a magnitude of 4.0, while the rest were smaller. While minor earthquakes are not uncommon in the northeastern United States, this was the first swarm of quakes known to have occurred near Lake Erie, where such activity had been considered rare or non-existent.
The researchers initially reviewed more than 25 geological and geophysical data sets to investigate if the earthquakes could be attributed to pre-existing faults, salt structures or other subsurface formations beneath the lake. To test the likelihood of fluid-induced seismicity, they ran two types of modeling — one to understand the response of Earth materials near the quakes to various pressure changes, and another to simulate the stress changes in the area, with some of the simulations including pore pressure increases or decreases.
To better constrain their models, the researchers also installed six instruments on the lake bottom and recorded seismic activity for nearly nine months. Based on the data from the new instruments, the quakes appear to be concentrated near a north-south trending zone that extends 20 kilometers south from the vicinity of Elk Creek, Ontario.
The study found no evidence of any specific faults, salt structures or subsurface weakness zones at depths consistent with the quake locations that could have been reactivated by pressure changes.
"The earthquake sequence is still unexplained, but we can constrain some of its characteristics," said study lead author Yajing Liu, a research scientist in the University at Buffalo College of Arts and Sciences' Department of Geology. "Future studies would benefit from monitoring additional local seismicity and installing a seismic array in Lake Erie to better locate the earthquakes."
While there are no known active faults in the Lake Erie basin that could have caused the earthquakes, the researchers say it is possible that there are still-undiscovered faults capable of producing small quakes.
The scientists say their study demonstrates that fluid-related processes cannot be ruled out as a potential cause of the quakes, although proving the connection is challenging. One possibility, they say, is that the lake's water levels could be influential.
Lake Erie is a shallow lake with a maximum water depth of 64 meters. Its water levels fluctuate naturally, with lake level records going back to the early 1900s. During wet years, the water level rises, flooding the region's wetlands and increasing the weight of the water column in deep sections of the lake. This extra weight can increase fluid pressures in subsurface formations. Additionally, the researchers note that the timing of the quakes — between December and February — is consistent with the historical maximum lake levels that tend to occur during the late winter.
"Given the large volume change in Lake Erie, even a relatively small change in lake level can create a significant change in loading and pore fluid pressure," said co-author Hao Zhang, professor of geology in the UB College of Arts and Sciences.
Although the quakes were small, the researchers say the sequence was sufficiently large to be felt over a broad region. The 2012 quake reached an intensity level V, which is strong enough to cause damage to older or poorly constructed buildings, and the 2014 quake reached an intensity level IV. The researchers point out that these intensities may be higher if there are buildings founded on soft sediments on the lakeshore.
"In the vicinity of Lake Erie, there is the Perry nuclear power plant and, more densely, wind turbines constructed in the lake," Zhang said. "Although these infrastructure features are designed to accommodate larger earthquakes, future studies about seismicity in this region are warranted to mitigate any risk."
