COMPARISON OF DIFFERENT HYPERSPECTRAL REMOTE SENSING METHODS FOR CENTRAL TEXAS OUTCROPS

Date

2021-12

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Remote sensing to characterize and classify outcrops has grown exponentially to become a vital tool in studying the surface. Remote sensing technology provides a new outlook on the methods of acquiring and analyzing spatial, spectral, and temporal resolutions. Hyperspectral imaging can identify objects and materials with a high spectral resolution by collecting reflectance values in wavelengths in visible infrared and shortwave infrared portions of the spectrum. In this study, ground-based hyperspectral data for four geological outcrops in the Central Texas are processed and analyzed to compare and determine which method of classification yields the most accurate results. Data from Lion Mountain Sandstone, Llanite, Packsaddle Schist, and Coal Creek Serpentine are analyzed to answer specific questions unique to these central Texas localities. Methods including spectral angle mapper (SAM), support vector machine (SVM) and artificial neural network (ANC) are used for classifying the data. Three to four samples are collected from each outcrop and ground truth is established by spectra collected on samples using an ASD spectroradiometer in the laboratory as well as a geochemical analysis conducted using a portable X-Ray Fluorescence instrument (pXRF). This serves to collect and study ground-truth data to ensure the results from the classifications are upheld. Spectra collected from samples obtained from the four geologic outcrops, identified the weathering effects of Lion Mountain on the electromagnetic spectrum, helped determine the petrogenesis of Packsaddle Schist and origin of blue quartz in Llanite to be caused by Rayleigh scattering.

Description

Keywords

Geology, Remote sensing, Central texas, Texas geology, Hyperspectral data

Citation