Log-ratios and geochemical discrimination of Scottish Dalradian limestones: a case study

Abstract

Abstract Geochemical data are used widely to help correlate lithostratigraphical sequences, particularly where they are unfossiliferous and/or affected by metamorphism and deformation. In this study, geochemical data for variably impure metamorphosed limestones from the Neoproterozoic — Cambrian Dalradian Supergroup of Scotland have been used to aid lithostratigraphical discrimination and correlation in a region where apparently similar sequences crop out in widely separated regions affected by major deformation. The key problem in the statistical analysis of geochemical data is that the data are constrained to a constant sum and cannot, therefore, be analysed in raw form by conventional statistical techniques. To overcome this problem, the data have been transformed into log-ratios. Subsequent statistical analysis of the transformed data focuses on determining the simplest sub-composition that effectively discriminates between the limestones. The results show that the Fe 2 O 3 -MgO-CaO subcomposition is sufficient in this regard, with the additional benefit that it reflects the composition of the carbonate component in the limestones. Mann-Whitney tests of the log-ratio Fe 2 O 3 /MgO show that limestones from most different lithostratigraphical levels are statistically different from each other at high levels of significance. The results have clarified interpretations of the lithostratigraphy and subsequent interpretations of the tectonic history of the region.