Bayesian array-based Receiver Function (RF): Towards stable, reliable, and easy-to-interpret RFs

Xin Wang, Zhong Minyan, & Zhan Zhongwen

Published August 15, 2019, SCEC Contribution #9811, 2019 SCEC Annual Meeting Poster #092

Receiver function (RF) has been an indispensable tool in structural seismology. Though a variety of RF deconvolution techniques have been developed, they are mostly performed at individual station-earthquake pairs, the results of which are adversely affected by data over-fitting and non-uniqueness of deconvolution. Here, we present a Bayesian array-based RF deconvolution method to improve the reliability of RF estimation. We take advantage of the coherency of dense-array wavefields and use a Markov chain Monte Carlo Bayesian algorithm to find an ensemble of RF solutions. The RF results are then presented in the probability format, which provides an objective way for deciding which features warrant geological interpretation. We test the algorithm on synthetic data, which suggests that the array-based RF method retrieves stable and reliable RFs even in the presence of high noise levels. We then apply the algorithm to real data collected by 3C nodal instrumentation deployed in the northern Los Angeles region. Despite the high noise level in the urban area and the short duration of the deployment, our results well reveal the shape of sedimentary basins. The highly stable, reliable and easy-to-interpret RFs obtained using array-based RF deconvolution provide the opportunity to take advantage of the emerging dense-arrays to better image the subsurface structures.

Key Words
dense array, receiver function, bayesian, basin

Wang, X., Minyan, Z., & Zhongwen, Z. (2019, 08). Bayesian array-based Receiver Function (RF): Towards stable, reliable, and easy-to-interpret RFs. Poster Presentation at 2019 SCEC Annual Meeting.

Related Projects & Working Groups