Seismic Thermometry of the North Pacific and Equatorial Indian Oceans

Citation

Peng, Shirui (2024) Seismic Thermometry of the North Pacific and Equatorial Indian Oceans. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/a814-nf75. https://resolver.caltech.edu/CaltechTHESIS:06012024-234504692

Abstract

The ocean absorbs the majority of excess heat in the climate system. Ocean mixing is also critical in setting Earth's thermal inertia. Over the course of the past few decades, conventional observations like Argo floats have drastically improved the coverage of the global ocean. However, their temporal and spatial resolutions are still limited. Resolving trends and patterns of temperature variations in the ocean under climate change remains a challenging sampling problem. This dissertation seeks to reduce such sampling errors by developing seismic thermometry. It is an acoustic method that measures large-scale ocean temperature changes using sound waves generated by repeating earthquakes. The chapters in this thesis attempt to combine physical understanding with statistical analysis to improve and implement seismic thermometry in several ways. First, acoustic waves generated by earthquakes along the Japan Trench and received at Wake Island are used to constrain temperature variation in the Kuroshio Extension region. An inversion that combines these measurements for the time and azimuth dependence of the range-averaged deep temperatures reveals lateral and temporal variations due to Kuroshio Extension meanders, mesoscale eddies, and decadal water mass rearrangements. Second, a comprehensive covariance structure is proposed to represent variabilities due to stochastic mesoscale, regional trend, and large-scale seasonality. It demonstrates statistical consistency between conventional float data and seismic measurements, and shows quantitatively that seismic thermometry reduces basin-scale temperature uncertainty when combined with conventional measurements. Finally, seismic data are compared with ocean models in the equatorial Indian Ocean to study the vertical structure of biweekly Yanai waves. The comparison indicates qualitative agreements in biweekly variations, and regression analysis confirms their origin as west-propagating Yanai waves. Yet quantitative differences in the biweekly variance magnitude demand further calibrations in both models and the seismic inversion.

Thesis Files