From the web site: http://adsabs.harvard.edu/abs/2003AGUFM.S21E0347S
A persistent earthquake sequence in northeast Ohio includes many distinct fore--main--aftershock sub--sequences, illuminates two faults, and was triggered by fluid injection. The first known earthquake from within 30~km of Ashtabula was an mb(Lg)~3.8 mainshock that shook the downtown area in 1987. Seismicity has continued at an average of about one felt event per year. The largest magnitude so far, mb(Lg)~4.3, caused slight damage (MMI VI) 26 Jan. 2001. The latest subsequence started July 2003 with mb(Lg)~2.6. Accurate hypocenters and focal mechanisms are available from three local seismograph deployments in 1987, 2001, 2003 and from regional broadband seismograms. These hypocenters are in the Precambrian basement, 0--2 km below the 1.8~km deep Paleozoic unconformity, and illuminate two distinct planar E-W striking sources zones 4 km apart, one in 1987 about 1.5~km long, the other in 2001 and 2003 about 5 km long. We interpret them as steep sub--parallel faults slipping left--laterally in the current regime. Like many of the faults that ruptured in hazardous SCR earthquakes, these faults were previously unknown and probably have small post--Precambrian displacements. The 1987 source was active a year after onset of class 1 fluid injection only 0.7~km north of the fault. The second fault, 5 km south of the injection well, became active in 2000, while the 1987 source was inactive. The well injected about 164 m3/day of waste fluid into the 1.8 km-deep basal sandstone with about 100 bars of well head pressure from May 1986 to June 1994. An annular high pore--pressure anomaly is expected to expand along this hydraulically confined horizon at the top of the basement even after injection ends and pressure drops near the well. Over 16 years, seismicity has shifted southward from ˜1 to 5--8~km from the point of injection. It seems to initiate when and where a significant pore pressure rise intersects pre--existing faults close to failure and to be turned off when pressure starts dropping back. The largest earthquakes postdated the end of injection at both Ashtabula and at the Rocky Mountain Arsenal near Denver, Colorado. Anthropogenic earthquake hazard may thus persist after the causative activity has ceased but can generally be closely monitored. High--stress and low strain rates in the eastern US and other SCRs can account for a larger proportion of triggered earthquakes in these regions than in active ones. Unlike hazard from natural SCR earthquakes, hazard from potential sources of anthropogenic earthquakes could generally be precisely identified in time and space. Anthropogenic triggering may have raised significantly the overall level of SCR seismicity during the last half century. Models that assume constant seismicity through the historic period may thus underestimate the overall hazard.
Keywords: 7209 Earthquake dynamics and mechanics, 7223 Seismic hazard assessment and prediction
