Paleosol Characterization and Identification in the Subsurface Using Nuclear Logs: A Case Study from the Kansas Cretaceous
by John H. Doveton, Kansas Geological Survey; and Peer Hoth, Bundesanstalt fuer Geowissenschaften und Rohstoffe
Spectral gamma-ray and lithodensity-neutron logs of Lower Cretaceous formations in Kansas show distinctive signatures that characterize deeply weathered paleosols. An integrated paleosol analysis of core and logs allows predictions of paleosol occurrence to be made in boreholes that have been logged but not cored. Kaolinitic paleosols are readily differentiated on logs from illitic marine shales by relative thorium enrichment and potassium depletion. Uranium concentration appears to be closely linked with organic matter content, while the photoelectric absorption curve is a sensitive indicator of iron-rich zones. Two methods were applied to paleosol log analysis. In the first, statistical pattern recognition methods were applied to log measurements and their transforms to identify multivariate log signatures that differentiated paleosols from other facies. In the second, quantitative mineralogical analyses of kaolinite, illite, illite-smectite mixed-layers, quartz, feldspar, and minor accessory minerals as well as the geochemical composition of shales were determined by a combination of XRD, XRF, ICP-MS and SEM/EDX analysis and used in calibrating an inversion procedure to convert nuclear logs to compositional estimates of minerals. The calibration gave useful insights on the links between petrophysical measurements and mineral composition of paleosols as well as measures of reliability in estimating each component. Paleosol facies predictions from nuclear logs in uncored wells are useful aids to outline the complex lateral geometry of Cretaceous floodplain deposits and to track regional changes in mineral assemblages in both time and space.