Recording Earthquakes in the Oceans for Global Seismic Tomography by Freely-Drifting Robots

Abstract

This dissertation describes the treatment of data recorded by a novel

instrument named MERMAID, short for Mobile Earthquake Recording

in Marine Areas by Independent Divers. MERMAID is a passively-drifting float

that autonomously records and transmits seismograms of teleseismic earthquakes

from the global oceans. The primary goal of the instrument is to fill a seismic

data gap. The primary goal of this dissertation is to describe how MERMAID data are

analyzed and interpreted to achieve this end. We begin by developing a new

method to pick multiscale arrival times of seismic phases in its noisy

seismograms and further to assign uncertainty estimates and confidence

intervals to those times. Our method is first applied to data collected in

the Indian Ocean and the Mediterranean Sea. It is able to able to identify

multiple phases that traveled different routes through the Earth with high

accuracy and low uncertainty. Next we describe an ongoing deployment of

50 MERMAIDs in the South Pacific. There we study the seismicity rates yet

reported and project for the total volume of data we expect returned over the

lifetime of the deployment. We again use our method of arrival-time and

uncertainty estimation to add to our ever-growing catalog of MERMAID

travel-time residuals, and compute a similar catalog using seismograms from

traditional seismometers installed on islands proximal to our floats. Their

agreement proves that MERMAID records tomographically-useful data. We use

these residuals to provide the first analyses of travel-time anomalies for the

new ray paths sampling the mantle under the South Pacific. Finally, we end with

a description of the open-source software developed over the course of this

research and note that it may be used to reproduce the results of this

dissertation.