Coincident conductive and reflective middle and lower crust in southern British Columbia

Abstract

S U M M A R YProcessing and interpretation of magnetotelluric data, recorded as part of the LITHOPROBE Southern Cordillera transect studies, across the boundary of the Intermontane and Omineca morphogeological belts reveals: (a) high electrical conductivity in the middle and lower parts of the crust everywhere, and (b) a depth dependency of geoelectric strike. The data have been modelled using two different inversion algorithms and different methods for correcting 'static shifts'. The two different approaches gave similar results: the depth to the top of a conductive layer decreases from 15-17km in the west across the Intermontane Belt to 8-10km across the transition to the Omineca Belt. The top of this conductive layer is closely coincident with a layer of increased seismic reflectivity as shown by reprocessing of collocated LITHOPROBE seismic-reflection data. The eastward shallowing is associated with an increase in heat flow such that the top of the conductive and reflective zones remains at 400-450 "C. This coincidence suggests that the increased reflectivity and the high electrical conductivity observed in the middle crust may have a common cause, and that their presence is limited to where the present temperature exceeds a critical value. One explanation that meets these conditions is that both the conductivity and reflectivity are produced by a small amount of aqueous fluid porosity. We propose that fluids are trapped in the middle crust by a ductile shear zone, previously interpreted from the seismic sections as the Okanagan Valley Fault to the west of Okanagan lake. The geoelectrical strike varies from N25"W for the first 5-10 km of the crust, to N20"E for the middle/lower crust, and to N60"E for the upper mantle. This variation indicates that the exotic terrane material is concentrated in the uppermost part of the crust and that the remainder of the crust is composed of ancestral North American rocks.

Electric and Magnetic Field Galvanic Distortion Decomposition of BC87 Data.

“…Relevant studies include Chakridi et al (1992), Jones and Dumas (1993), Eisel and Bahr (1993), , Kurtz et al (1993), Ogawa et al (1994), Marquis et al (1995), Boerner et al (1995), and Gupta and Jones (1995). Physically, the 3D galvanic distortion is caused by the presence of electric charges along discontinuities or gradients in electrical conductivity associated with small-scale (relative to the background inductive scale length) surface structures.…”

Analysis of magnetotelluric data along the Lithoprobe seismic line 21 in the Blake River Group, Abitibi, Canada

“…Correlation between transparent and resistive zone has been previously mentioned in Gough (1986) and Jones (1987) and illustrated in many cases (e.g. Kellett et al, 1994;Marquis et al, 1995;Bitri et al, 1997). Variation in depth estimates between the seismic image and the MT model may be due to some error in the estimation of the MT-static shift parameters, in the approximate correlation between reflectivity and resistivity contrast or in the lower resolution of the MT method compared with the seismic reflection method.…”

Two-Dimensional Inversion of Papua New Guinea Data Using 'Least-Blocked' Models.

“…We have therefore preferred to take the simple approach suggested by Berdichevsky et al (1989) and used recently by Marquis et al (1995) in their analysis of the British Columbia data. Following this method we first displaced the TE apparent resistivity curves at all sites so that they assumed the value of 30 f2m at the long period of 682.7 s, on the assumption that an approximate onedimensional structure exists at great depths because the TE and TM phases approach each other at longer periods (Fig.…”

“…A maximum likelihood estimation of these parameters is conducted by minimizing a misfit of the response function to the measured data. The strategy employed in the method has been briefly described in connection with its application to some synthetic models , to the COPROD2 data , and to data from southern British Columbia (Marquis et al, 1995). It involves the automatic generation of a series of optimized models, based on starting models with successively 3,4,5.... columns, which stops when the inclusion of an extra column is no longer justified by an improvement of the response fit.…”