| Abstract: | Rapidly acquired non-invasive geophysical data is key to reducing the risk inherent in subsurface investigations. It achieves this risk reduction by provision of spatiotemporally dense datasets and new methods to measure the efficacy of acquisition, analysis, and modeling. In a first example, I use two geophysical methods⁰́₄electrical resistivity tomography and time-domain electromagnetics⁰́₄to investigate the subsurface in a rapidly urbanizing alluvial floodplain setting. Specifically I focus on the geologic structure of a shallow alluvial aquifer in the Brazos River floodplain of Texas, characterizing dynamic hydrological interactions between the aquifer and the adjacent river. Based on new geophysical insights, I determine how the sedimentary architecture of the shallow alluvial aquifer acts as a control on its recharge and discharge and how bidirectional preferential flow pathways establish hydrologic communication between the aquifer and the river at human and geologic time scales. In a second example, I develop a protocol to improve identification of unmarked graves in a historic African-American cemetery. I show that a geophysicist⁰́₉s detection proficiency, expressed in terms of true-positive, true-negative, false-positive, and false-negative percentages, can be improved using radar signatures of nearby known targets as a proxy for ground-truth. The electronic version of this dissertation is accessible from https://hdl.handle.net/1969.1/197265 |
|---|