Imaging VTEM Data: Mapping Contamination Plumes In Tarlton, South Africa M. Combrinck Geotech Airborne Limited Summary VTEM data were acquired during a test survey flown over the Tarlton region in South Africa, 2009. This data were successfully used to map highly conductive plumes that are likely to indicate zones of contaminated ground water. Due to the very conductive nature of the target, it was imperative to use a high-power, large bandwidth TDEM system in order to achieve reasonable depths of penetration as well as high resolution. Introduction A test survey was flown for the Department of Water Affairs and Forestry (DWAF) to demonstrate the value that Time Domain Electromagnetic (TDEM) data can add to environmental projects, and contaminated groundwater studies in particular. The area was chosen as a test site in conjunction with the Department of Water Affairs and Forestry (SA) as a high probability zone for groundwater contamination due to worked out gold mines in the catchment region. The objectives of this survey were: To map highly conductive subsurface features that may be indicative of polluted groundwater To map dykes that can compartmentalize groundwater resources To map fracture zones that can influence groundwater flow and distribution General geology and location The Tarlton survey area is located 35 km W-NW of Johannesburg, South Africa. It falls within the Transvaal Supergroup which consists of clastic and chemical sediments (Fig. 1). It just borders on the Witwatersrand Supergroup, consisting of quartzite, shale and conglomerate with banded iron formation and basaltic lavas. The survey area itself is believed to be dominated by Chuniespoort group dolomite.
Figure 1: Tarlton survey area superimposed on a simplified geological map of South Africa. Adapted from: http://www.geoscience.org.za/images/stories/rsageology.gif Methodology Physical properties of the expected lithologies indicate that the magnetic responses would be dominated by lavas or dolerite dykes, whereas water content, and specifically the amount of contamination related salts ( or ions ) in solution in the ground water is expected to cause the most significant response in TDEM data. A total of 484 km of total field magnetic and TDEM data were acquired at 200 m line spacing, using standard VTEM acquisition parameters. It is important to note that the survey area falls within a developed area and there are numerous sources of cultural noise, including high voltage power lines. The distribution of these was mapped by the 50Hz power line monitor (Fig. 2). Despite the dense distribution of power lines in the area, good quality data were obtained, except at close proximity to the dominant SW-NE trending high voltage line. The magnetic data were filtered to produce analytical signal (Fig. 3) and 1 st vertical derivative grids in addition to the total magnetic field amplitude. Euler deconvolution was performed using an algorithm developed by the USGS to determine an approximate depth to the top of magnetic sources. Two main structural directions can be identified (015 and 095 ) from the data with 115 and 160 directions also present but of secondary significance. Apart from the dominant power line, the northern half of the survey is magnetically quiet and interpreted to consist of dolomite, at and below surface. This is supported by the lack of magnetic sources in the north resulting from Euler deconvolution. The magnetic anomalies in the southern half of the grid predominantly originate from depths between 100m to 350m, and are thought to be due to dolerite intrusions; a subsurface northern extension of the Witwatersrand Supergroup or more likely the Ventersdorp Supergroup lavas. The interpreted structural features (faults and/or lineaments) are superimposed on Fig. 3.
Figure 2: Grid of the 50Hz power line monitor amplitude. Figure 3: Grid of the Magnetic Analytical Signal with delineation of structural features based on magnetic data and derived products. TDEM data can be presented in presented in many different ways. For the purpose of this study the B- field conductivity contour maps are shown (Figure 4). The conductivity-depth distributions were calculated with EMFlow and contour maps were gridded in Geosoft. On these maps the 115 structural direction dominates, with a major feature cutting across the survey area. Comparing this feature with the lineaments on the simplified geological map of South Africa (Fig. 1), shows a good correlation with the only major lineament mapped in the survey area. This feature does not correspond to any of the magnetic structures or inferred lava/dolerites at depth. It is therefore interpreted as a fault or shear zone and increased weathering as well as groundwater flow can be inferred along this structure; in fact, these are exactly the mechanisms that will cause higher conductivities to be measured. In three dimensions its lateral extent (NE-SW) increases with depth and along strike it dips to the NW. This behaviour indicates higher complexities than associated with a simple fault and the outlines are mapped on Fig. 4 as conductive plumes. This feature also seems to act as a partial boundary on the eastern side, as the plumes extend more readily to the west than the
east. The exception is where a fault mapped from magnetic data (striking at 015 ) cuts the major structure and the conductive material leaks through to the NE, where it follows a structural weakness. Finally, the location of a known sinkhole is indicated on Fig.4. A small stream flows into this sinkhole and disappears into the subsurface. This sinkhole is located at the centre of this conductive zone and it is conceivable that surface water, with higher than normal conductivity, (and therefore total dissolved solids) penetrates into the aquifer at this point. It is a scenario that will explain the high amplitude and distribution of the subsurface conductivities. It is not likely to be a purely lithological conductor, as it does not follow the general geological trends as outlined by the magnetic data. Figure 4: Conductivity contour maps derived from B-field VTEM data at 50m depth intervals with interpretation results superimposed on the 250m depth grid. Lineaments based on EM data are shown in black; lineaments from magnetic data are in yellow; thick blue lines indicate large scale lineaments from the simplified RSA geological map; red polygons outline conductive plumes assumed to be groundwater with higher than normal conductivity; a sinkhole into which a small stream disappears are shown as a black circle. Conclusions In conclusion, a number of structural features were mapped with airborne magnetic and electromagnetic data. Highly conductive plumes that are structurally controlled were delineated in three dimensions up to depths of 250m. Although it is not possible to determine the exact cause of these plumes without follow-up investigations, the data seems to indicate that these highly conductive features are associated with groundwater (containing excess TDS) rather than lithological units.
Acknowledgments The author would like to thank all Geotech personnel involved in this project.