KGS Home PagePRS Home PageOnline Pubs Page

Kansas Geological Survey, Open File Report 96-49

GPR Study of The Drum Limestone, part 4 of 9


GPR Methods, Data Colllection, and Processing, continued

Equipment and Data Collection

A site map of the study area shows the relationship between the GPR lines and the outcrop (Fig. 2). Both uninterpreted and interpreted versions of the photomosaic and GPR data are shown in Figures 4-6. Line D1 (Fig. 5) was acquired from a northeast to southwest direction parallel to the highwall, and approximately 9 m (30 ft) southeast of it (Fig. 4). This orientation is sub-parallel to the primary and secondary dip directions of the cross bedding (Fig. 2). Line D2 (Fig. 6) was acquired perpendicular to line D1 from a northwest to southeast direction with the intersection point occurring at station 0 on both lines. This line is sub-parallel or highly oblique to the strike direction of the cross-beds (Fig. 2). Station locations on the photomosaic are indicated by survey crew members holding station number signs while standing on the stations (Fig. 4). The stations are indicated on the GPR data by vertical lines on the sections every 1.5 m (5 ft) and double vertical lines every 7.6 m (25 ft). The data are plotted to 50 ns below the surface. This corresponds to a depth of approximately 3.0 m (10 ft) assuming an average one-way velocity of 0.12 m/ns (0.4 ft/ns), which is appropriate for limestone (Davis and Annan, 1989).

Preparation of the study site included clearing the antenna path of major obstructions such as large rocks, flagging stations along the antenna path, collecting relative elevation information for the stations, and obtaining photomosaics of the outcrops in relation to the stations. The pathway remained as close to the quarry wall as possible to facilitate correlation with the exposed geology (Fig. 4). The lines were acquired directly on the limestone outcrop with either no soil cover or a very thin veneer of clay or sand. This part of the quarry was chosen because shale and soil cover was recently removed from the top. Other areas of the quarry above the highwall contained a relatively thick layer of soil and shale which would rapidly attenuate the GPR signal. Stations were flagged at a 1.5 m (5 ft) interval because it allowed correlation of the GPR data with specific ground locations via the photomosaics, and aided in control of antenna velocity during data collection to ensure even cdp coverage. The collection of relative elevation information from stations every 1.5 m (5 ft) allowed the data to be corrected for elevation differences during processing. These corrections aided the interpretation of reflection information and correlation with outcrop information. Elevations were obtained using a level and rod with an accuracy of plus or minus 3 cm. The photomosaics were gathered in increments of 15.2 m (50 ft) to allow correlation between the outcrop and GPR data.

The equipment used for the study was a GSSI SIR System-8 GPR unit, with a DT6000A tape unit and 500-MHz dominant-frequency antenna. A scan length of 60 ns was recorded at a rate of 12.8 scans/second as the antenna was pulled along the line. The equipment was placed within a large-wheeled garden cart, which allowed continuous profiling along the entire line for each scan length. The tape unit recorded coherent system noise beginning at approximately 40 ns on each trace, masking much of the reflection information below 40 ns. The data were gathered using a monostatic antenna, meaning that a single antenna was both the source and receiver, as opposed to a bistatic mode of operation where two antennas are separated. The monostatic mode of operation is quick, simple, and yields many traces in a short amount of time. It also resulted in a single trace approximately every 3 cm, depending on how fast the antenna was pulled along the ground. However, in this mode of operation with this equipment, vertical stacking (summing) of traces to reduce random noise was not possible. It was also not possible to gather velocity information because there were no cdps generated to determine stacking velocities. Thus, time to depth conversions are possible using only approximate velocities estimated for a given rock type.


Report Contents || Previous Page--Methods || Next Page--Data Processing
Kansas Geological Survey, Open-File Report 96-49
Placed online Jan. 1997
Comments to webadmin@kgs.ku.edu
URL=http://www.kgs.ku.edu/PRS/publication/OFR96_49/rep4.html