Statistical Analysis of Seismicity in Alberta

English Abstract:

In western Canada, hydraulic fracturing treatments as well as extended fluid injection operations have been linked to the generation of seismicity in some areas. An increase in seismographic stations over the last few years enables improved modeling of induced seismicity processes. Characterization of the statistics of these events is vital to developing a predictive understanding of induced seismicity, which is ultimately required for the mitigation of future induced seismicity.

Using a compiled earthquake catalog (www.inducedseismicity.ca), we perform a statistical analysis to characterize the frequency-magnitude distributions of seismicity across western Alberta, from a regional perspective. We also look more closely at the seismicity distributions within concentrated areas where clusters of activity in time and/or space are evident. The average b-value (the slope of the Gutenberg-Richter relation) is determined for each chosen area, using the Maximum Likelihood Method. We map the b-value variation spatio-temporally. We consider alternative statistical models to describe the observed distributions, including the Poisson process and an ETAS (Epidemic Type Aftershock Sequence) model. We seek preliminary correlations between the properties of the seismicity distributions and industrial activities that may have induced them.