2003 Seattle Annual Meeting (November 2–5, 2003)

Paper No. 1
Presentation Time: 1:30 PM-5:30 PM

IMAGING CLASTIC DIKES AT THE HANFORD SITE WITH GROUND PENETRATING RADAR


CLEMENT, William P., Center for Geophysical Investigation of the Shallow Subsurface, Boise State Univ, 1910 University Dr, Boise, ID 83725-1535 and MURRAY, Christopher J., Applied Geology and Geochemistry Group, Pacific Northwest National Lab, PO Box 999, K6-81, Richland, WA 99352-0999, billc@cgiss.boisestate.edu

The Hanford site contains numerous clastic dikes that are possible conduits for water flow through the vadose zone at Hanford. The near-surface sediments at the Hanford site predominantly consist of coarse to fine grained sands. The dikes are composed of a thin, clay or silt skin with sand and silt as the in-filling material. The dikes are easily located at the surface by vegetation changes. However, the subsurface character of the dikes is often poorly known. Furthermore, many dikes may not intersect the surface and thus are undetected.

To determine the spatial distribution and depth of clastic dikes present at the Hanford site, we collected 3-dimensional, 100 MHz GPR surface reflection data at two sites; the 216-S-16 Pond and the Army Loop Road sites. The 3-dimensional data consist of 90 or 60 m lines with a spacing between traces of 0.1 m. We collected lines every 2 m to cover the area. We acquired more widely spaced cross-lines to provide ties between the lines. We also collected a 6.9 km linear profile with trace spacing of about 0.1 m to understand the dikes at larger scales. The survey crossed desert grassland similar to the Army Loop Road area and traversed many sand dunes and blowouts. The sand dunes enable deeper signal penetration below the land surface, but the data may not image any deeper below a common reference elevation.

The dikes are easily distinguished in the GPR data by diffractions from the dike edge and a loss of coherence in underlying reflections. In general, the data quality is better at the Army Loop Road and Traverse sites than at the 216-S-16 Pond site. Cobbles and boulders are more common at the 216-S-16 Pond site and probably cause the poorer data quality. The signal penetration, and thus the ability to map the dikes in the subsurface, varies between 5 to 12 m below the land surface.