You are here

Radar stratigraphy – a method for analysing 3D GPR data in sedimentary environments as exemplified by fluvial sediments

Geographical Review, 2015, 87, 3, s. 439-456

Radar stratigraphy is a method of interpreting ground penetrating radar (GPR) data collected to image the sedimentary architecture of clastic deposits. It can be used to interpret the data objectively and present the results quantitatively. GPR data provides high-resolution (decimetre-scale) information on sedimentary architecture over large areas, offering a basis upon which processes at the time of deposition can be inferred. Such data thus afford high-resolution detail on the vertical and lateral characteristics of deposits that is not available using traditional techniques, such as vibrocoring, shallow trenches, exposures, or other geophysical methods. GPR has often been applied in 456 Tomasz Żuk, Gregory H. Sambrook Smith the study of fl uvial and fl uvioglacial systems, given their complexity, and the fact that they commonly form groundwater and hydrocarbon reservoirs. The aeolian deposits also studied frequently using GPR are associated with the best-quality data, due to low signal attenuation. Coastal environments have also been studied, though here the signal may be attenuated where a shallow saline/brackish groundwater table is present. Lacustrine deposits (and some floodplain deposits along lowland rivers) may contain high proportions of clay and organic matter that limit penetration of the electromagnetic signal – and hence suitability for GPR surveying – considerably. As sedimentary architecture is of an inherently three-dimensional character, an important advantage of 3D GPR is the classifi cation and environmental interpretation of facies and bounding surfaces on the basis of three-dimensional (3D) datasets. The latter are usually collected as a series of parallel lines characterised by close spacing (of 0.5 m or less, depending on radar frequency). Radar stratigraphy was formalised as a discipline at the beginning of the 1990s, when GPR was fi rst used to study the sedimentary architecture of fl uvioglacial aquifers. The principles of seismic stratigraphy were adapted to higher-resolution data, and used to distinguish the main bounding surfaces and facies. In order to include the spatial character of architectural units, Neal (2004) recommended including radar packages in data interpretation. However, when 3D data are available, radar facies and surfaces can also be classifi ed in line with their appearance in three directions. Although radar stratigraphy facilitates a quantitative and systematic approach to the study of sedimentary architecture, and has been in use for over 20 years now, radar facies and surfaces have not been described in full in line with their 3D appearance. 3D datasets from fl uvioglacial deposits have been presented for aquifer characterisation (Beres et al., 1995, 1999; Peretti et al. 1999; Asprion and Aigner, 1999; Heinz and Aigner, 2003), ancient fl uvial systems as reservoir rock analogues (McMechan et al., 1997; Corbeanu et al., 2002), contemporary and ancient aeolian systems (Jol et al., 2003; Adetunji et al., 2008, Bristow, 2009; González-Villanueva et al. 2011), as well as lagoon deposits (Moldoveanu-Constantinescu and Stewart, 2004). An attempt to include horizontal slices in the classifi cation of radar facies was made by Beres et al. (1999), while Hickin et al. (2009) used a 3D facies classifi cation based on information available from individual (2D) GPR profi les. Neal et al. (2008) identifi ed radar surfaces and radar packages in 3D data, and determined dip directions on horizontal slices in order to study the sedimentary architecture of an oolitic limestone formation. After a brief discussion of the history and principles of radar stratigraphy, this article presents the example of a 3D dataset collected from the South Saskatchewan River, Canada, as displayed using seismic interpretation software. Classifi cation of radar facies and surfaces on the basis of 2D data is compared with that using a 3D dataset. This demonstrates that reliance on individual profi les, even if closely-spaced, can be misleading. For example, the bar-top hollows described by Best et al. (2006) could be interpreted as channel deposits if only 2D data were used. Additionally, 3D data can offer greater certainty as regarding the orientation of sedimentary structures, while horizontal time slices reveal subtle (20°–30°) changes in palaeocurrent directions. It is concluded that the same principles of data interpretation as are presented here from a fluvial deposit canalso be applied to datasets collected from other sedimentary environments.