Joint inversions of three types of electromagnetic data explicitly constrained by seismic observations: results from the central Okavango Delta, Botswana

Kalscheuer, Thomas ; Blake, Sarah ; Podgorski, Joel E. ; Wagner, Frederic ; Green, Alan G. ; Maurer, Hansruedi ; Jones, Alan G. ; Muller, Mark ; Ntibinyane, Ongkopotse ; Tshoso, Gomotsang

In: Geophysical Journal International, 2015, vol. 202, no. 3, p. 1429-1452

Ajouter à la liste personnelle
    Summary
    The Okavango Delta of northern Botswana is one of the world's largest inland deltas or megafans. To obtain information on the character of sediments and basement depths, audiomagnetotelluric (AMT), controlled-source audiomagnetotelluric (CSAMT) and central-loop transient electromagnetic (TEM) data were collected on the largest island within the delta. The data were inverted individually and jointly for 1-D models of electric resistivity. Distortion effects in the AMT and CSAMT data were accounted for by including galvanic distortion tensors as free parameters in the inversions. By employing Marquardt-Levenberg inversion, we found that a 3-layer model comprising a resistive layer overlying sequentially a conductive layer and a deeper resistive layer was sufficient to explain all of the electromagnetic data. However, the top of the basal resistive layer from electromagnetic-only inversions was much shallower than the well-determined basement depth observed in high-quality seismic reflection images and seismic refraction velocity tomograms. To resolve this discrepancy, we jointly inverted the electromagnetic data for 4-layer models by including seismic depths to an interface between sedimentary units and to basement as explicit a priori constraints. We have also estimated the interconnected porosities, clay contents and pore-fluid resistivities of the sedimentary units from their electrical resistivities and seismic P-wave velocities using appropriate petrophysical models. In the interpretation of our preferred model, a shallow∼40 m thick freshwater sandy aquifer with 85-100 Ωm resistivity, 10-32 per cent interconnected porosity and <13 per cent clay content overlies a 105-115m thick conductive sequence of clay and intercalated salt-water-saturated sands with 15-20 Ωm total resistivity, 1−27 per cent interconnected porosity and 15-60 per cent clay content. A third∼60 m thick sandy layer with 40-50 Ωm resistivity, 10-33 per cent interconnected porosity and <15 per cent clay content is underlain by the basement with 3200-4000 Ωm total resistivity. According to an interpretation of helicopter TEM data that cover the entire Okavango Delta and borehole logs, the second and third layers may represent lacustrine sediments from Paleo Lake Makgadikgadi and a moderately resistive freshwater aquifer comprising sediments of the recently proposed Paleo Okavango Megafan, respectively