Cause and Effect Versus False Correlations in Subsurface Hydrochemistry.

Plots of element and isotope ratios are used extensively in many earth science disciplines. Variations in their values within a given system can provide the clues necessary towards understanding the geochemical evolution of unsaturated zone soil water and also groundwater. While element and isotope ratios are usually correctly plotted against other parameters such as depth and location along a flow path, many investigators have failed to recognise the self correlation that can occur when common variables are plotted against each other. Examples from hydrochemistry are discussed where self-correlation occurs because two parameters containing a common variable are compared. We highlight examples involving correlations between two different element ratios (a/b vs c/b), between element ratios and element concentrations (a/b vs b), and between element concentrations and the total sum of element concentrations (a vs (a + b + c, etc.)), and the potential implications of not accounting for the self correlation.