Gravity interpretation of bedrock topography: the case of the Oak Ridges Moraine, southern Ontario, Canada
Introduction
Since April 1993, the Oak Ridges Moraine area has been the subject of a regional hydrogeological study conducted by the Geological Survey of Canada (GSC) (Sharpe et al., 1996). An interdisciplinary approach was adopted to determine the geological framework and to study groundwater flow in the Greater Toronto area Sharpe et al., 1996, Russell et al., 1996, Pullan et al., 1994. In particular, the aim of the study is to map major Quaternary aquifers and aquitards in the region and to understand their influence on groundwater flow. Various factors may control groundwater flow within the moraine and across the moraine area. One factor, which may influence groundwater flow across the moraine on a regional scale, is bedrock topography. Coarse sediments deposited in deep channels in the bedrock form important aquifers that are often exploited for municipal water supply. The largest and deepest channel in the Oak Ridges Moraine area is the Laurentian Channel (Fig. 1(b)), which extends from Georgian Bay (Fig. 2(a)) to Lake Ontario and is estimated to be 30-km wide Spencer, 1881, Spencer, 1907, Brennand et al., 1998. An economical method of obtaining a regional image of bedrock topography is sought. The gravity method in applied geophysics is the most promising means to this end.
Traditionally, mapping bedrock channels has not been a customary application of the gravity method. Some isolated cases exist where the gravity method was used to map buried valleys Thomsen et al., 1999, Lennox and Carlson, 1967, Hall and Hajnal, 1962. Leão et al. (1996) developed an algorithm to constrain inverted basement relief using a few known depths. The ease of data acquisition on a regional scale and 2D coverage in an urban setting have been its main advantage over other geophysical methods employed by the GSC in its investigation of the moraine Pullan et al., 1994, Todd and Lewis, 1993, Todd et al., 1993, Pilon et al., 1994, Pugin et al., 1999. Seismic profiles in the region are at most 8 km in length. With its high acoustic impedance, the dense layer of Newmarket till contained in the moraine is a limitation in the seismic method in imaging the bedrock/sediment contact below. Electromagnetic soundings were limited by the low conductivities of the limestone bedrock north and east of the moraine (Pullan et al., 1994). In western areas, the shale/sediment contact was poorly resolved.
The Oak Ridges Moraine (ORM) is situated in southern Ontario, Canada, north of Toronto and east of the Niagara Escarpment (Fig. 1). It covers an area roughly 160×20 km and is composed of discontinuous layers of glaciogenic Pleistocene sediments Sharpe et al., 1994, Sharpe et al., 1996. The sediments include clays, tills, sands and gravel. The moraine is underlain by Ordovician shale and limestone bedrock (Johnson et al., 1992). Beneath Ordovician rocks are Precambrian rocks of the Grenville Province. A literature review of ORM-type sediments and bedrock provided estimates of moraine and bedrock densities Berman et al., 1942, Jakosky, 1950, Daly et al., 1966. The average sediment density (saturated) is estimated to be within a range of 1.8–2.0 g/cm3. Average bedrock density (dry) is estimated to be between 2.5 and 2.6 g/cm3. The sediment/bedrock density contrast is, thus, estimated to be between 0.5 and 0.8 g/cm3.
Over 25 000 boreholes reaching bedrock exist in the area (Fig. 2(a)). The GSC used the borehole data to construct a grid of sediment thickness, which was subtracted from a grid of surface topography to produce a map of bedrock topography (Brennand et al., 1998). The borehole data indicate sediment thickness ranges from 0 to 250 m, averaging 57 m (Russell et al., 1998). Surface topography ranges from 32 m above sea level (m asl) to 520 m asl (average 236 m asl) on the Niagara Escarpment. Bedrock topography, as described by the borehole data, ranges from 16 m below sea level (m bsl) to 495 m asl (average 177 m asl). Little borehole control on the depth to bedrock exists in the area of the moraine itself (kriged area in Fig. 2) where sediment thickness reaches greater values. As a result, there is poor horizontal resolution of bedrock topography. The gravity data acquired over the moraine would in theory improve the definition of bedrock channels in those areas where few boreholes exist.
Fig. 1(a) shows the gravity survey area. The distribution of the gravity stations is shown in Fig. 1(b). Four gravity surveys were completed between 1994 and 1996 covering the moraine and the Laurentian Channel Belisle, 1995, Gill, 1996, Gill, 1997, Jobin, 1997. The gravity readings were initially taken at 500-m intervals in the southwestern part of the moraine. Subsequently, the sampling interval was increased to 1 km. Gravity readings were taken with Scintrex CG-3 Autograv and Lacoste and Romberg G790 gravimeters. Station altitude was measured with an altimeter and psychrometer. In the final survey covering the Laurentian Channel west of Lake Simcoe (Fig. 1(b)), a GPS Sokkia system was used for coordinate (including altitude) measurements. Together with the national regional database, a total of 5681 observations were made available by the GSC for interpretation. The maximum errors on the measurements of horizontal coordinates, altitude, and gravity average 40.12 m, 2.04 m, and 0.066 mGal, respectively. The average error in altitude is about 1 m.
Section snippets
Background
The raw gravity data were corrected for station elevation and mass (together known as the Bouguer correction) via the formula defining the Bouguer anomaly, gBA:where g0 is observed absolute gravity (Gal); gt is theoretical gravity calculated on the reference ellipsoid (Gal); ∂g/∂z is the vertical gravity gradient at mean sea level (0.3086×10−5 Gal/cm); G is the gravitational constant (6.672×10−8 cm3/g s2); h is station elevation with respect to mean sea level (cm); ρ is
Model responses
Fig. 3 shows the gravity response of bedrock topography deduced from borehole information. The response was calculated in the Fourier domain using Parker's (1973) algorithm. The bedrock is modelled as a slab of infinite lateral extent with a flat bottom and topographic relief. The algorithm assumes a constant density contrast between the slab and the material overlying it. An additional parameter to be included in the calculation is the reference level, z0. The reference level is the average
Uniform density contrast
A given gravity signal is a convolution of the gravity effect of the source's density distribution with the effect of its geometry or structure or distance from the plane of observation (Blakely, 1996). In the Fourier domain:where g is the gravity signal, ρ is the density distribution, and ξ, the geometry. Assuming a uniform density contrast, it is possible to calculate from the signal the geometry of the causative structure. This calculation is known as inversion
Uniform density contrast
In Fig. 6, the structure west of 640 km UTM can be correlated with the Laurentian Channel. West of 580 km UTM, part of the bedrock high consisting of the Niagara Escarpment is retrieved. The inversion indicates the presence of three northwest trending channels around 10-km wide within the Laurentian Channel. The existence of the first channel from the west is less plausible since the borehole distribution is dense in this zone and the borehole model indicates that bedrock elevation rises toward
Modelling
Two profiles were extracted from the residual gravity field for 2.5D modelling (Fig. 4). The modelling process involves the calculation of the gravity response of a model and the modification thereof so that the response fit the observed signal. A 2.5D modelling program was used. The causative structure is modelled as a polygon with limited lateral extent. The polygon has a uniform density contrast with the surrounding medium. The code is based on equations assuming the gravity observations are
Conclusions
To conclude, a residual gravity field which contains the signal of bedrock topography but which also contains the remnants of signals from bodies located below the bedrock surface, was obtained. This is most evident with the northeast alignment, which originates in the Precambrian basement. The residual was inverted using two inversion codes. The first assumes a uniform density contrast and retrieves the geometry. The second retrieves a density distribution. From the density distribution, the
Acknowledgements
The author, Maria Annecchione, wishes to thank Charles Logan of the Geological Survey of Canada (GSC) for his incredible patience in providing the borehole data and regional topographic features to exact specifications repeatedly while she figured out what was required. The authors thank Susan Pullan (GSC) for the ORM geophysical logs made readily available and for reviewing this article. David Sharpe (GSC) and Marc Hinton (GSC) also reviewed the article making valuable adjustments to its
References (32)
- Belisle, J., 1995. The effectiveness of the gravity method for mapping bedrock topography beneath the Oak Ridges...
- et al.
Density at room temperature and 1 atmosphere
Potential Theory in Gravity and Magnetic Applications
(1996)- et al.
Constraints in 3D gravity inversion
Geophysical Prospecting
(2001) - Brennand, T.A., Moore, A., Logan, C., Kenny, F., Russell, H.A.J., Sharpe, D.R., Barnett, P.J., 1998. Bedrock Topography...
- et al.
Density of rocks
- Gill, B., 1996. Gravity Survey of the Oak Ridges Moraine (A Hydrogeological Investigation in the Bolton Area North of...
- Gill, B., 1997. Gravity Survey of the Oak Ridges Moraine (Toronto Harbour). Internal Report, Gravity Project 96102,...
- et al.
The gravimeter in studies of buried valleys
Geophysics
(1962) A case for upward continuation as a standard separation filter for potential field maps
Geophysics
(1987)