RESEARCH PAGES: Research by topic Faculty research pages World tour (large images) Research sites within the department

3-D Seismic Tomography at a Groundwater Contamination Site Investigator : Colin Zelt, Alan Levander, Aron Azaria, Diana Dana, Igor Morozov Funding Source: Department of Energy, Environmental Management Science Program

Final 2-D velocity model for a line across the channel from the 1998 pilot survey using a slide hammer source. Contour interval is 100 m/s; the 500, 1000 and 1500 m/s black contours are labeled. The known water table depth in the channel is indicated by the brown arrows. The pink dot indicates the depth (13.2 m) to the clay aquiclude from a well at 25 m distance.

Abstract As part of an ongoing environmental characterization project at Hill Air Force Base near Ogden, Utah, a 3-D seismic survey was performed over a contaminated aquifer in July and August 2000. The project was led by a team from the Rice Center for Computational Geophysics. This site contains significant amounts of dense non-aqueous phase liquids (DNAPLs) in a shallow aquifer less than about 15 m deep. The aquifer is bounded below by a clay aquiclude, in which a paleochannel acts as a trap for the contaminants. The overburden consists of Quaternary sands, gravels and clays. Imaging the structure of the paleochannel at depths up to 15 m is the main target of the survey. The four week experiment included 3-D reflection, 3-D refraction, checkshot surveys and vertical seismic profiles using wells up to 15 m deep. Here we present preliminary traveltime tomography results from the 3-D tomography and reflection surveys. The results of a 2-D pilot survey conducted in 1998 show that velocity in the near surfaceincreases by almost an order of magnitude in the upper 15 m at this site,from about 200 m/s to 1500 m/s. 3-D synthetic tests were developed using the 2-D results to design the field acquisition program. For the tomography experiment the entire site, roughly 95 m by 36 m in area, was surveyed with about 600 RefTek Texan recorders deployed in astationary rectangular grid having an inline spacing of 2.8 m and acrossline spacing of 2.1 m. A shot from a .223 caliber rifle wasfired at each receiver station, yielding a dataset with about 360,000traces, and offsets up to 100 m. The source provides a broad frequency range from 40 Hz to greater than 300Hz. The 2-D data and the synthetic tests suggest that the 3-Dwide-angle data will provide a well-resolved 3-D velocity model of the paleochannel using first-arrival tomography. Field and instrument support for the experiment was provided by IRIS/PASSCAL and the University of Texas at El Paso.

Relative difference between the final model for the slide hammer data from 1998 pilot survey and the 1-D starting model. The approximate position of the channel inferred from the reflection images is indicated above, and the depth to the water table is also indicated. There is a good correlation between the low-velocity region in the tomographic model and the extent of the channel and its depth. Note the large velocity perturbations from the starting model, up to 65% in magnitude.