From resistivity to clay thickness - An inversion approach

We present a concept using geostatistical estimation and non-linear inversion to optimize a function translating geophysical resistivity models to geophysical clay thickness. The concept minimizes the difference between reported clay thicknesses in boreholes and calculated clay thicknesses based on geophysical resistivity models. The concept is called geoStatistical estimation of Structural Vulnerability (SSV). For sedimentary areas the cumulated clay thickness in the upper part of the subsurface is an important factor for the water infiltration speed and thereby the vulnerability of underlying aquifers to pesticides, nitrate etc. Borehole information contains the most detailed information on the clay thickness, but most often borehole information is too sparse for the detail level required in actual mapping situations. However, the amounts of clay are also reflected in the resistivity of the sediments and thereby in geophysical data having resistivity as the physical measuring parameter. The geophysical data often has the desired spatial coverage (Christensen and Sørensen, 1998, Thomsen et al., 2004). In short, the concept incorporates: 1. Clay thicknesses in boreholes cumulated for some interval and accompanying uncertainties. 2. Layered models obtained from inversion of geophysical data including the parameter uncertainties. 3. A spatial interpolation (kriging) from the positions of the geophysical models to the positions of the boreholes. The uncertainties on the geophysical models are carried through the interpolation together with the uncertainty on the interpolation itself. 4. A non-linear inversion scheme minimizing the difference between observed clay thicknesses and calculated clay thicknesses The concept has been employed in a couple of vulnerability mapping campaigns in Denmark with promising results. We will show an example in which SSV was used to produce an optimized geophysical clay thickness map. The optimized map greatly improved the overview of the data and it improved the basis for decisions regarding the area planning.