Characterizing shallow aquifers with wave-propagation based geophysical methods: Imaging and attribute analysis
Sawyer, Dale S.; Talwani, Manik
Doctor of Philosophy
As the results of seismic reflection and ground penetrating radar (GPR) studies become more prevalent as input for quantitative groundwater and engineering studies, it is important to evaluate traditional approaches to data processing and analysis. Where conventional methods fail it is necessary to investigate and/or develop non-traditional approaches to data analysis. I present five stand-alone studies that are focused on characterizing shallow aquifers using seismic reflection and GPR data processing and analysis. Each of the projects involves a new approach to data analysis either through alternative processing strategies that are not widely applied in environmental studies or development of new processing methods and/or algorithms. The first two studies are focused on seismic reflection imaging problems that arise in the shallow environment. I first present a detailed discussion of the errors that can result from conventional normal-moveout (NMO) processing, and the application of pre-stack depth migration (PSDM) to improve image accuracy. Second, I apply dip-moveout (DMO) processing to a data set from the Puget Sound. DMO is rarely applied in environmental studies, but can improve velocity analysis and image quality where there are dipping layers or scattering events. The final three studies are focused on attribute analysis of GPR and seismic reflection data in direct detection studies. Over the past 10--15 years, direct detection has been used successfully in the energy industry to identify oil and gas reserves from exploration scale seismic reflection data, but is a new approach to the analysis of GPR and shallow seismic reflection data. Direct detection studies use reflected wave attributes such as amplitude, frequency content, and phase to estimate material properties. I first present a detailed discussion of GPR amplitude vs. offset (AVO) analysis for direct detection of free phase non-aqueous phase liquid contaminants (NAPLs). The analysis is not straightforward and requires careful consideration of electromagnetic dynamic properties. Second, I present a shallow seismic case study where a predictable AVO response was detected. Finally, I present a new wavelet decomposition and time-frequency representation, and illustrate applications in GPR attenuation analysis for NAPL detection and lithology characterization.