Attribute analyses of GPR data for heavy minerals exploration

Abstract

This study is a continuation for our previous work [1] depicting soil mineralogy using Texture Analysis (TA) of Ground Penetrating Radar (GPR) data. In addition to TA, Complex Trace Analysis (CTA), and Center Frequency Destitution (CFD) were applied to GPR data to predict the existence of buried heavy mineral deposits. CFD and CTA attribute were also used to determine the concentration of the buried heavy mineral deposits. The features of CTA are useful in showing changes of the potential energy components such as instantaneous energy. τ-parameter and Normal Distribution of Amplitude Spectra (NDoAS) were calculated from CTA to inspect the concentration of the buried samples and CFD was used to reveal energy allocations using spectral content of GPR data in time and frequency domain. GPR data collected from laboratory experiments using 1.5 GHz antenna were used in the study. The experiments were conducted using various heavy mineral samples with different concentrations. Our previous study showed that buried minerals produced high entropy, contrast, correlation, standard deviation, and cluster, but these samples produced low energy, and homogeneity. Variance measure signifies edges of buried samples within host material. This study indicates that first and second derivatives of the envelope calculated from CTA emphasize the variation of the reflected energy and sharpen the reflection boundaries in the data. Instantaneous measures (energy and power) of envelope data reveal the existence of buried samples, while the frequency distribution of the data enables locating the contact of buried mineral. We found τ-parameter, NDoAS, and center-frequency proportionally increase with increased concentration of the mineral samples. The results from the three analyses, although in agreement with the previous work, they substantially improve the detection as well as quantifying the mineral concentration.