Cylindrical anisotropy of Earth's inner core re-examined through robust parameter search

The Earth's inner core (IC) remains one of the most enigmatic parts of our planet. The structure and dynamics of the IC have been at the forefront of deep Earth research ever since the discovery of Earth's innermost shell in 1936 by Inge Lehmann. However, the formation, internal structure...

Full description

Bibliographic Details
Main Author: Stephenson, Jo
Format: Article in Journal/Newspaper
Language:English
Published: The Australian National University 2021
Subjects:
Online Access:https://dx.doi.org/10.25911/9w68-m248
https://openresearch-repository.anu.edu.au/handle/1885/247359
Description
Summary:The Earth's inner core (IC) remains one of the most enigmatic parts of our planet. The structure and dynamics of the IC have been at the forefront of deep Earth research ever since the discovery of Earth's innermost shell in 1936 by Inge Lehmann. However, the formation, internal structure and dynamics of the IC still remain a fundamental puzzle for researchers. The core of the Earth is responsible for the geodynamo, generating Earth's magnetic field, protecting life on Earth from solar radiation. Global seismology is an excellent tool to provide insights into the IC. Seismic waves travel at different velocities through media of varying densities and reflect off sharp discontinuities. This allows us to infer structure below the surface of the Earth. In particular, the seismic phase "PKIKP" traverses the IC and is the primary data used throughout this thesis. As the number of data increases and computational techniques improve, the variability and complexity of IC models have increased dramatically. Each model is unique, and interpretation is often dependent on data choice, processing and methodology employed. In particular, this thesis explores the hypothesis of IC cylindrical anisotropy - the idea that seismic rays traveling through the IC parallel to the Earth's rotation axis travel faster than those parallel to the equator. This thesis investigates the anomalously fast South Sandwich Islands (SSI) to Alaska seismic paths. On these paths, PKIKP differential travel times show a consistent deviation of 2-4 seconds from predicted travel times in the outer region of the IC - an observation arguably responsible for the proposal of IC anisotropy in the upper part of the IC. However, the origin of the anomaly is poorly understood. Research in this thesis focuses on a detailed analysis of the waveforms from SSI earthquakes. Novel methods are employed to pick a comprehensive new data set of differential travel times from 36 events. Results show that the significant travel time anomaly present in the data is unlikely to be a result of systematic error in picking of the core phases. Instead, analysis shows that the anomaly is likely a result of strong heterogenous structure elsewhere within the deep Earth. In addition to the new SSI dataset, this thesis contributes further data to the global PKP differential travel time dataset, focusing primarily on polar antipodes. Most significantly, this thesis investigates the radial dependence of cylindrical anisotropy in the Earth's IC. Varying models have been created for an innermost IC (IMIC), with radius anywhere between 300 and 800 km. To investigate the existence of the IMIC, the direction of fast/slow axes and its radial extent, global absolute PKIKP data from the International Seismological Centre (ISC) is used in conjunction with the parameter search method the "Neighbourhood Algorithm" (NA). The merit of the ISC catalogue is in the numerous available PKIKP travel times collected over many decades, along with improved location algorithms. The NA provides a robust means of testing this hypothesis where it produces ensembles of all models that satisfactorily fit the data. This method can be employed without regularisation or needing to employ any subjective choices such as binning of phase data. In addition to the NA, a complementary likelihood ratio approach is used to further constrain uncertainty on anisotropy parameters. This investigation suggests that spatial averaging methods are unnecessary to observe the IMIC and may be partially responsible for disagreements between models of IC anisotropy. Results further show that the IMIC is not defined by a change in strength of anisotropy, but by a significant deviation in the slow direction of propagation at 650 km. Such a result could be interpreted as a phase change in iron or mineralogical impurities at depth within the IC.