Developing a 3-D crustal velocity model for central Oklahoma

Shuo Zhang, & Hejun Zhu

Published August 2, 2019, SCEC Contribution #9315, 2019 SCEC Annual Meeting Poster #006 (PDF)

During the last decade, there are increasing numbers of earthquakes recorded in Oklahoma, most of which are small earthquakes. In order to study these earthquakes, a lot of local station networks have been deployed in this region, such as GS, YW, OK, etc. The USArray also covers the area during 2010-2011. With this large database, we attempt to estimate a 3-D crust velocity model for the state of Oklahoma by using full waveform inversion. We select 226 earthquakes from 2011 to 2018, recorded by 182 station. Additional 23 earthquakes are used as an independent dataset to test the inverted velocity model. The magnitude of these earthquakes ranges from 3 to 5, and the depth ranges from 1 to 15km. The initial model is a 3-D model built by using adjoint tomography with ambient noise cross-correlation functions. This initial model has a good constraint for the Anadarko and Arkoma basins. To avoid the cycle-skipping problem, we use a multiscale strategy and select three frequency bands, 10-30s, 5-30s, 2-30s. The final velocity model is the result of 20 preconditioned conjugate-gradient iterations. This new model enables us to better delineate geological features in central Oklahoma. It can also be utilized to characterize strong ground motions, locate earthquakes and determine moment tensor solutions in central Oklahoma.

Citation
Zhang, S., & Zhu, H. (2019, 08). Developing a 3-D crustal velocity model for central Oklahoma. Poster Presentation at 2019 SCEC Annual Meeting.

Related Projects & Working Groups
Ground Motions