Seismic reflection method

Sinkholes are one of the major geohazards in karst areas, causing significant economic damage and even occasional loss of life. Sinkhole formation mechanisms vary depending on geological conditions but are initiated by dissolution of bedrock (generally carbonates or evaporites) below the ground resulting in ground surface deformation and eventual collapse. The process may be accelerated by natural events like storms or heavy rains and droughts, or human activities like water pumping or loading of the land surface. In Florida, limestone dissolution leading to sinkhole development often results in the formation of surface depressions that are often water-filled and develop into depressional wetlands.
Previous studies using near-surface geophysical imaging techniques (including seismic refraction) in Central Florida have shown the correspondence between depressional wetlands and sinkholes originated in deep-seated interstratal karst with a variable overburden. However, these geophysical techniques are often unable to reach the karst interface which may typically be positioned at depths exceeding < 50-60 m. This research investigates the use of ground-based seismic reflection techniques to image deep paleokarst relief and better understand sinkhole development and extent below the overburden. This approach follows earlier studies by others using seismic reflection methods to identify sinkholes under lakes in Central Florida. While these previous studies deployed the method over water, the approach here investigates how land-based near-surface seismic reflection surveys may provide similar results below depressional wetlands. A total of three different locations with depressional wetlands under similar geological conditions (but somewhat variable depth to the karst interface) are investigated, including the Disney Wilderness Preserve near Poinciana (FL), the Allapattah Flats Wildlife Management Area near Palm City (FL); and the J.W. Corbett Wildlife Management Area in Palm Beach County (FL).

Interaction of normal incidence, wideband acoustic pulses with seabed is investigated to determine the acoustic frequency ranges that provide the most information about the sediment structure. An exact numerical model is developed for calculating the frequency response and impulse response of the seabed from an impedance profile of a sediment core. A database of impedance profiles from several ocean environments were studied to describe the shapes of commonly found impedance changes. The impulse response of the seabed is convolved with acoustic pulses to generate synthetic acoustic returns. The synthetic profiles are studied to determine the effect of operating frequency and bandwidth on resolution and on the accuracy of measuring impedance changes. This thesis explains why inversion procedures have failed to generate vertical impedance profiles of the seabed from normal incidence reflection data. The results of this work provide guidelines for selecting subbottom profiler array sizes and operating frequencies for quantitative sediment studies, and for subsampling cores.