S51A-4429:
Investigation of Seismic Events Associated with the Sinkhole at Napoleonville Salt Dome, Louisiana

Friday, 19 December 2014
Avinash Nayak, Berkeley Seismological Lab, Berkeley, CA, United States and Douglas Scott Dreger, University of California Berkeley, Berkeley, CA, United States
Abstract:
This study describes the ongoing efforts in analysis of the intense sequence of complex seismic events associated with the formation of a large sinkhole at Napoleonville Salt Dome, Assumption Parish, Louisiana in August 2012. We investigate source mechanisms of these events, represented by a general 2nd order point source centroid seismic moment tensor (MT) using data from a temporary network of broadband stations established by the United States Geological Survey. We have implemented a grid-search technique to detect and compute the centroid hypocenter and MT solution of events using low frequency (0.1-0.3 Hz) 25-second duration displacement waveforms and 1D velocity models for the salt dome and the surrounding sediment sequence. Application of this technique to data extending up to 17 days before and after the appearance of the sinkhole yields a catalog of > 1200 events. We describe the evolution of the seismicity in time with respect to hypocenters, source-types, and magnitudes. We also examine the Gutenberg-Richter statistics and relationship between inter-event times and magnitudes. Since fewer stations were operational during the time period preceding the appearance of the sinkhole, we perform various sensitivity tests to examine the effects of decrease in station coverage from a 5-station network to a 3-station network on location, source-type and magnitude uncertainties. We also apply a finite source MT inversion approach using grid-search to isolate specific larger-magnitude events that occur in rapid succession resulting in overlapping waveforms. The MT analysis is supplemented by a separate waveform cross-correlation analysis to study smaller magnitude events that have poor signal-to-noise ratios at low frequencies. We also explore the role of fluids, possible a natural gas-water mixture, in the spectral peaks observed in the seismic signals.