Seismicity in Pennsylvania

Author:

Homman, Kyle A

Graduate Program:

Geosciences

Degree:

Master of Science

Document Type:

Master Thesis

Date of Defense:

September 25, 2015

Committee Members:

Andrew Arnold Nyblade, Thesis Advisor/Co-Advisor

Keywords:

seismic
seismicity
induced seismicity
earthquakes
Pennsylvania
mining
cross correlation

Abstract:

In this thesis, data from several seismic networks have been used to develop a catalog of seismic events in Pennsylvania from February 2013 to December 2014. Using the Antelope Environmental Data Collection Software Suite, P-wave arrivals were picked using 1-5 Hz bandpass filtered vertical component seismograms. Initial hypocenters were determined using the IASP91 velocity model. P-wave arrival times were subsequently used with the HYPOELLIPSE code (Lahr, 1989) and a velocity model developed for Pennsylvania to relocate the events. The catalog consists of 1355 events in Pennsylvania and is complete to a M_L of 2.0. Magnitudes were computed using Richter’s (1935) method for determining local magnitude (M_L). Most events occurred between 12:00 and 22:00 UTC during week days. The Gutenberg-Richter plot yielded a b-value of 2.63 for the catalog. Events were classified based on criteria used by the United States Geological Survey and others for determining mining-related seismicity. The results of event classification showed that the catalog predominantly consists of mining-related seismicity. Through the classification process, 11 of the 1355 events in the catalog were found to be unrelated to mining activity. P-wave and S-wave arrival times for these events were repicked using 1-10 Hz bandpass filtered data and relocated using the Pennsylvania velocity model and HYPOELLIPSE. Event locations and origin times were examined to determine if there was a spatial and temporal correlation with natural gas extraction activity. No correlation was found and therefore there appears to be little, if any, evidence for seismic events in the catalog caused by hydraulic fracturing or wastewater injection. The 11 non-mining related events all occurred in areas of known faults. A cross correlation match and locate technique was applied to a cluster of events near Williamsport, PA using a template event that occurred on September 25, 2013. Upon locating the cross correlation detections relative to the template event, and then moving the locations to the centroid of the cluster, it was determined that the cluster is a result of mining-activity at the Thomas Coal Mine. The findings of this study are important for future monitoring of seismicity in Pennsylvania. The natural gas production region and the bituminous coal region in the Commonwealth overlap spatially. Accurate event locations, magnitude estimates, and origin times are required to discriminate between mining events and possible seismicity induced by natural gas production. To obtain that information, a large, permanent seismic network is required. The amount of mining-related seismicity generated from both coal and industrial mineral mines also creates challenges for detecting and locating small tectonic events in both known and possibly unknown areas of earthquake activity.