Autoregressive determination of the complex frequencies of the Earth's normal modes

Majstorović, Josipa (2016) Autoregressive determination of the complex frequencies of the Earth's normal modes. Diploma thesis, Faculty of Science > Department of Geophysics.

[img]
Preview
Text
Language: Croatian

Download (32MB) | Preview

Abstract

Based upon superconducting gravimeter and seismometer records, the autoregressive method is used to determine the frequencies and quality factors of five spheroidal modes: four fundamental modes 0S0, 0S2, 0S3, 0S4 and one higher mode 2S1. The data used are 113 records from five major earthquakes (Sumatra-Andaman earthquake 2004, Chile earthquake 2010, Tohoku earthquake 2011, Achet earthquake 2012 and Illapel earthquake 2015) with the moment magnitude above 8 recorded at 46 stations worldwide. Estimated weighted average values of the eigenfrequencies and the quality factors of the mode’s singlets are very close to the results of the previous studies as well as the PREM predictions. However, some results show deviations from theoretical PREM values, supporting the fact that this model is not appropriate to explain all processes in the Earth, because of Earth’s complexity. Additionally, standard deviation estimated by the bootstrap method are realistic and consistent. While this autoregressive method gave satisfactory results, one should be aware of the fact that this method relies heavily on the high quality data, because non-linearity destroys autoregressive nature of time series and hence introduces bias in the estimate.

Item Type: Thesis (Diploma thesis)
Keywords: Earth's normal modes, complex frequency, autoregressive method
Supervisor: Herak, Davorka
Date: 2016
Subjects: NATURAL SCIENCES > Geophysics > Seismology and Physics of the Earth's Interior
Divisions: Faculty of Science > Department of Geophysics
Project code: 119-1193086-1314
Depositing User: Iva Vrkic
Date Deposited: 06 Apr 2016 10:10
Last Modified: 06 Apr 2016 10:10
URI: http://digre.pmf.unizg.hr/id/eprint/4668

Actions (login required)

View Item View Item