Evidence of Possible Induced Seismicity in the Wabash Valley Seismic Zone from Improved Microearthquake Locations

We reanalyzed data from a temporary seismic network in southern Indiana and Illinois deployed in November 1995 through July 1996. We produced the first comprehensive catalog of the complete 211 days of that experiment. The results are dominated by a cluster of 534 microearthquakes that occurred near the town of New Harmony, Indiana. Waveform similarity among all of the events suggests a more closely spaced grouping than that found by independent earthquake locations. We relocated the earthquakes using time-domain cross-correlation of the vertical traces and complex cross-correlation of the horizontal traces to improve arrival times for P- and S-waves, respectively. Additional constraints on the absolute locations were applied using arrival time difference analysis of earthquakes recorded by two stations. Average source depths from standard location methods and waveform modeling with synthetic seismograms indicate, in contrast to larger events in the region, the earthquakes occurred at depths < 4 km, within the sedimentary section of the Illinois Basin. We propose that these earthquakes are likely related to oil production, in particular water-injection for the purpose of secondary recovery. The primary evidence for this includes: (1) tight spatial clustering of earthquakes; (2) unusually shallow earthquake depths; (3) good spatial correlation of the relocated hypocenters to the existing wells and oil fields; and (4) an unusually steep slope in the log-linear relationship of earthquake magnitude and frequency. It is therefore necessary to discard the events in the cluster before consideration of magnitude-frequency relations for earthquake hazards estimates.