Publication Date

2025

Advisor(s) - Committee Chair

Nahid Gani, Michael May, M. Royhan Gani, Jun Yan

Degree Program

Department of Earth, Environmental, and Atmospheric Sciences

Degree Type

Master of Science

Abstract

The Olmstead Quadrangle (scale 1:24,000), in south central Kentucky, is a karst-rich region within the Pennyroyal Plateau Sinkhole Plain, serving as a primary groundwater recharge zone for the Mammoth Cave system. Composed of Mississippian-age sedimentary rocks, it contains numerous sinkholes, dry valleys, and sinking streams that shape its hydrology. While the Pennyroyal Fault System runs through the area, no faults have been documented within the quadrangle. Despite its significance for groundwater flow, karst development, and geohazards, its lithologic and structural features remain inadequately mapped. The existing geologic map by George E. Ulrich (1966) merged Renault Limestone, Bethel Sandstone, and Paint Creek Limestone into a single undifferentiated unit, overlooking key lithologic variations. Given advancements in remote sensing, LiDAR, and machine learning, an updated map is needed to refine stratigraphic boundaries and document geologic structures. This study integrates ASTER, LiDAR, and Sentinel- 2 satellite imagery with machine learning algorithms, field investigations, thin section petrography, and SEM analyses. Results confirm that remote sensing effectively distinguishes major stratigraphic units, particularly oolitic limestone, red sandstone, and minerals within stratigraphic units, including silica, kaolinite, calcite, and dolomite. Band math (Band Arithmetic) function performs arithmetic operations on the bands of a raster layer. and LiDAR analysis link sinkhole development to Paint Creek Limestone, highlighting its karst susceptibility. Vegetation and farmland affect remote sensing accuracy, with winter imagery yielding the best results. LiDAR and field validation confirmed the absence of major faulting within the quadrangle, aligning with previous assessments but differing from adjacent quadrangles where faults are documented. Thin section and SEM analyses revealed dolomitization, indicating diagenetic alterations that may influence unit classification. This study enhances geologic mapping accuracy and establishes a framework for future mapping in Kentucky by integrating remote sensing, machine learning, and field validation. Additionally, it demonstrates the effectiveness of advanced mapping techniques in heavily vegetated and farmland-covered terrains, for future societal benefits and resource management strategies. The updated map provides critical insights into groundwater management, hazard assessment, and urban planning.

Disciplines

Earth Sciences | Geochemistry | Geology | Physical Sciences and Mathematics | Sedimentology | Stratigraphy | Tectonics and Structure

Download

Available for download on Saturday, April 14, 2125

Share

COinS