IMAGING THE CRUSTAL AND MOHO STRUCTURE OF CENTRAL OKLAHOMA WITH REGIONAL SEISMICITY AND RECEIVER FUNCTIONS

The Mesoproterozoic accretionary growth history of the Laurentian Supercontinent is recorded within the Southern Granite-Rhyolite province. Geochemical studies from drill cuttings have laid the basis of current evolutionary models for the lithosphere in this region, but these samples are representative only of the uppermost crust. Seismic studies focusing on the deep crustal structures can improve our understanding of the lithospheric evolution. We use the rich seismic catalog of local and teleseimic events recorded in Oklahoma, resulting in part from the increase in induced seismicity of the past decade in the state, to study the deeper crust in central Oklahoma. The dense deployment of local seismic monitoring networks also provides an opportunity to employ the receiver function technique with relatively high spatial resolution.

We use 27, 582 local earthquake events recorded between 2010-2017 across Oklahoma. We apply common-midpoint sorting, stacking and inversion to the Pg phases of local earthquakes to obtain localized 1-D Pg velocity-depth functions. These 1-D velocity curves are combined to obtain a 3-D P-wave velocity model for central Oklahoma. Our methodology provides a significant increase in the depth of investigation (up to ~ 40 km) as compared to conventional local earthquake tomography. Furthermore, we obtain a crustal thickness model and Vp/Vs ratios for central Oklahoma through receiver function analysis and H-k stacking performed on 212 teleseismic events recorded across Oklahoma.

Our models reveal a high velocity lower crust (V_Pg > 7 km/s) and an average V_P/V_S ratio greater than 1.8 which are indicative of a mafic lower crust. Velocity variations observed within the crust are closely related to gravity and magnetic variations. Crustal thickness varies between ~40-48 km in central Oklahoma. We interpret our results within the context of the prevailing lithospheric accretionary growth models.