Triggering Processes of Oklahoma Earthquakes: Look Beyond Fluid Diffusion
A well-established injection experiment in the 1960's demonstrates that humans can influence earthquake occurrence by pumping water in the subsurface [Raleigh et al. 1976]. Decades later, the central United States has seemingly replicated the 1960's experiment by accident: rapid increase in seismicity rates since 2009 that cannot be explained by steady tectonic stress loading [Ellsworth 2013, Llenos and Michael 2013]. The mechanism is generally explained by the classic Coulomb stress view, where increased pore pressure reduces effective normal stress, and brings faults closer to failure. Induced earthquakes often occur in the form of "swarm-type" clustering with no clear main shocks, and often exhibit spatio-temporal migration that is regarded as the manifestation of fluid propagation [Shapiro et al. 2002]. Most of the previous studies focused on single injection well or wells that are directly linked to a localized sequence, where clear spatio-temporal correlation can be established (e.g., [Kim 2013, Horton 2012]). However, recent studies suggest that induced earthquakes can occur as far as 20 km away from the injection zone with long-term fluid diffusion [Keranen et al. 2014]. The larger spatial scale of pressure propagation makes it more challenging to understand the mechanism driving earthquake occurrence patterns, especially for large damaging earthquakes, as the expected pressure increase is much smaller due to the spatial decay [Goertz-Allmann and Wiemer 2013, Keranen et al. 2014].
While it could be explained by a critically stressed fault being re-activated by a small amplitude of stress perturbation (e.g., kPa), the roles of earthquake-to-earthquake interaction and aseismic slip are often omitted. Llenos and Michael  applied statistical modeling to an induced sequence in Arkansas and a natural swarm nearby, concluded that both background seismicity rates and aftershock productivity needed to be increased to match observations for an induced sequence. The increase in background seismicity is naturally expected due to increased external forcing (e.g., pore pressure or aseismic slip) [Hainzl 2004], while the increase in aftershock productivity suggests that inter-event triggering cannot be ignored. In Oklahoma, the Mw5.7 Prague earthquake is triggered by increased Coulomb stress from the M5.0 foreshock one day before [Sumy et al. 2014], suggesting that smaller earthquakes can trigger larger events.
In this study, we seek to better understand the generation processes of large earthquakes in Oklahoma by considering a full range of triggering processes. We begin by an overview of occurrence patterns of earthquakes in Oklahoma, and their relationship with injection zones. Then, we focus on the Fairview/Woodward in western Oklahoma and the Pawnee county with detailed analysis of the relationship between injection and background seismicity activities; as well as stress interaction between microearthquakes and large M4 earthquakes, where clustering of microearthquakes illuminates locations of M>=4 earthquake. Detailed analysis of the precursory activities before the Pawnee earthquake suggests that the mainshock was eventually triggered by fluid injection, aseismic slip and cascading failure from foreshocks.
The study is made possible with data provided by the Oklahoma Geological Survey, and the Oklahoma Corporation Commission, with results from hard-working graduate students and discussion with OU colleagues at MCEE.
Xiaowei Chen grew up in Liaocheng, Shandong province in China. She went to local public schools from kindergarten through high school. After graduation from high school in 2003, she went to the University of Science and Technology of China in Hefei, Anhui province for undergraduate education, majoring in Geophysics. During her undergraduate study, she received several awards for excellence in academics and research. Upon graduation in 2007, she went to the University of California, San Diego for her Ph.D study, where she studied earthquakes in Antarctica and California with Dr. Peter Shearer, a member of the National Academy of Science. After graduation in 2013, she visited the University of Tokyo in Japan for four months, where she worked at the Earthquake Research Institute and studied earthquakes in Japan with Dr. Aitaro Kato. After returning to the US, she focused on subduction earthquakes at the Woods Hole Oceanographic Institution as a postdoctoral researcher, working with Dr. Jeff McGuire. In January 2015, she started as an assistant professor at the ConocoPhillips School of Geology and Geophysics at the University of Oklahoma.