← Back to Sessions

Bayesian Analysis of Ambient Noise Cross Correlation Functions Using a Low-Quality Seismic Network

Session: Advances in Seismic Interferometry: Theory, Computation and Applications [Poster]

Type: Poster

Date: 4/22/2021

Presentation Time: 11:30 AM Pacific

Description: 

Although ambient noise dispersion measurements are being published on the African continent, how these measurements compare to previous results, affect updates to global lithospheric models, or should be used in joint inversion schemes, is not well understood or often rarely investigated. In this study, we apply a hierarchical Bayesian technique to extract ambient noise phase dispersion and its uncertainties using station-pairs within sub-arrays from different classes of networks on the continent. Unlike iterative schemes, this approach requires limited a-priori constraints, and employs a global search for constraining the measurement uncertainty during the inversion for phase and group velocity dispersion using the ambient noise cross correlation functions. Since data quality varies across the entire region, we report on how the uncertainties in phase dispersion measurements depend on structural and non-structural parameters: i.e., duration of operation, metadata quality, and sediment or low-velocity zones. A continent-wide reference dataset with uncertainties derived from data will be useful for updating lithospheric models during joint inversions with other seismic datasets. For example, global lithospheric models are heavily cited and widely used across the broader geoscience community, so ensuring that they fully incorporate short-wavelength, short-period ambient noise surface wave measurements, especially in regions that have only recently been instrumented, e.g., Africa, will be beneficial.

Presenting Author: Siyu Xue

Student Presenter: Yes

Authors