← Back to Sessions

Bayesian Optimal Experimental Design for Seismic Monitoring Networks

Session: Network Seismology: Keeping the Network Running While Integrating New Technologies II

Type: Oral

Date: 4/22/2021

Presentation Time: 02:30 PM Pacific

Description: 

The Bayesian optimal experimental design (OED) problem seeks to identify data, sensor configurations, or experiments which can optimally reduce uncertainty about quantities of interest (e.g. event location). The goal of OED is to find an experiment that maximizes the expected information gain (EIG) given prior knowledge and models of expected data. Therefore, within the context of seismic monitoring, we can use Bayesian OED to configure sensor networks by choosing sensor locations, types, and levels of fidelity in order to improve our ability to identify and locate seismic events. By developing Bayesian OED tools for analyzing and designing monitoring networks, we can explore many questions relevant to monitoring such as: how and what data phenomenologies can be used to optimally reduce uncertainty, how much is gained by reducing sensor noise or earth model uncertainty, and how do sensor types, number, and locations influence uncertainty?

We will discuss the general Bayesian OED framework for seismic monitoring and also the application of this framework for optimizing a local seismic monitoring network to improve its ability to locate seismic events from arrival time data. With this example, we will show how we develop the four basic elements for solving this Bayesian OED problem: 1) A likelihood function that describes the uncertainty of seismic phase detections and arrival times, 2) A Bayesian solver that takes a prior and likelihood to identify the posterior distribution of likely event locations, 3) an algorithm to compute EIG, and 4) An optimizer that finds a sensor network which maximizes EIG. Finally, we will present the results of this analysis and discuss how different modeling assumptions such as travel time error spatial correlations and different levels of measurement noise influence network sensitivity and the optimal placement of new monitoring stations.

SNL is managed and operated by NTESS under DOE NNSA contract DE-NA0003525.

Presenting Author: Thomas A. Catanach

Student Presenter: No

Authors