Date of Award

12-30-2013

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Applied Science

First Advisor

Haydar Al-Shukri

Abstract

Attribute analyses have been used successfully in seismic applications for many years. In this study, the application of the attribute analyses to Ground Penetrating Radar (GPR) data has been proposed to detect and identify heavy minerals within the Moon soil (regolith). Lunar samples are mostly composed of heavy minerals such as ilmenite, plagioclase, olivine and pyroxene, a characteristic that makes lunar soil a source for elements such as titanium, oxygen and iron. The main goal of this study is to demonstrate the use of GPR method for detecting and mapping heavy minerals concentrations. The attribute analyses used in this study are Attenuation Analysis (AA), Complex Trace Analysis (CTA), Texture Analysis (TA) and Center Frequency Destitution (CFD). Attribute analysis was applied to both synthetic models and prototype laboratory measurements to study its application to GPR data. The results indicate that the attribute analyses of GPR data can be useful to provide valuable subsurface information. The findings of AA show that attenuation values are function of mineralogy of the subsurface. This could be applicable to Moon and Mars in addition to Earth environment to explore their near-surface soils. CTA can successfully estimate the location of heavy mineral samples embedded inside host medium through the variation of reflected energy around buried sample and sharpen the reflecting interface. Results indicate that as the amount of the buried heavy minerals increases, the value of CTA parameters (Normal distribution of amplitude spectra `NDoAS' and τ,-parameter) proportionally increase. TA measures combined can be used as an enhanced interpretation tool. The texture results show that heavy mineral concentrations can be identified by the high contrast, entropy, autocorrelation, correlation, cluster, dissimilarity, standard deviation, variance and low energy, maximum probability, and homogeneity. The measures also help in highlighting the edge of the buried samples. The normalized information measure of correlation can depict the buried mineral sample with its zero value. The results of CFD analysis allows the detection and location of buried heavy minerals. Correlation coefficient measure is useful to inspect the relation between two statistical features of TA. T-test results indicate the rejection of the null-hypothesis. This concludes statistically significant differences between two statistical features of TA. Based on the findings, the test could reveal the weak reflections of GPR data that couldn't be seen with the naked-eye. There is a linear correlation between correlation coefficient and t-test scores. The spectral measures of CFD could be helpful to define the spectral characteristic of GPR reflections. Spectral variance and standard deviation measures are useful measures to define frequency variation about the center frequency. The results indicate a relationship between the frequency distribution and the amount of the heavy minerals. CFD is also capable of highlighting mineral layers that are very thin.

Download

Share

COinS