Theses and Dissertations

Date of Award

5-2021

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Physics

First Advisor

Dr. Soumya Mohanty

Second Advisor

Dr. Soma Mukherjee

Third Advisor

Dr. Malik Rakhmanov

Abstract

On the frontier of gravitational wave (GW) astronomy, the LIGO detectors record vast quantities of data that need to be analyzed constantly for rare and transient GW signals. A foundational problem in LIGO data analysis is the identification of spectral line features in the Power Spectral Density (PSD) of the data. Such line features correspond to high power terrestrial or instrumental signals that must be removed from the data before any search for GW signals can take place. In this study the method developed aims to automate the extraction of the frequencies and bandwidths of the lines, treated as sharp features in the PSD. For this purpose, we use a non-parametric curve fitting method recently developed at UTRGV called SHAPES. Our method automates the determination of line features through window smoothing, estimation through SHAPES and extremum analysis with a program called ALINE. In order to apply SHAPES, which uses B-spline curve-fitting with an adaptive knot sequence, we had to first overcome its limitation on the length of input data. This was achieved by applying SHAPES to overlapping segments of the PSD data along with an exponentially weighted averaging of the SHAPES estimates in the overlap regions. With the estimation groundwork laid, the second step is determining the locations of the lines through adaptive extrema differences in the SHAPES estimate of the PSD with ALINE. Our study shows that it is possible to expand SHAPES, an adaptive estimation tool, to (i) any sized data, thus impacting applications in all statistical endeavors, and (ii) use it to automate the identification of lines in the PSD of LIGO data with the ALINE algorithm.

Comments

Copyright 2021 Thomas A. Cruz. All Rights Reserved.

https://go.openathens.net/redirector/utrgv.edu?url=https://www.proquest.com/dissertations-theses/automated-identification-lines-data-gravitational/docview/2564965157/se-2?accountid=7119

Download

Share

COinS