Re-Evaluating Potential Uncertainty in Mechanistic Interpretations Using CSIA
Abstract
Dual stable isotope plots have been extensively used in contaminant hydrology to infer information about reaction mechanisms in the laboratory or in the field. However, the calculation and use of lambda values (Λ), the slope of dual isotope plots, is not standardized across the literature, resulting in inconsistent reporting of lambda values and its associated error. This lack of standardization becomes increasingly important when lambda values are used as evidence to distinguish between reaction mechanisms and have larger implications when interpreting field data. Using simulated as well as previously reported datasets, we evaluated several different regression techniques including ordinary linear regression, reduced major axis regression, and the generalized least squares cubic method - also known as the York method. Ordinary linear regression can underestimate the slope significantly when the error in the isotope measurement is sufficiently large or the measurement range is sufficiently small. The York method may best represent lambda as it accounts for uncertainty in both isotope measurements over a range of scenarios. We demonstrate that two or more lambda values can be more accurately distinguished on a dual isotope plot from rigorous hypothesis testing (z-test) and that assessing the overlap between 95% confidence intervals can lead to misinterpretation of the data. We propose that either the standard error be reported along with the value of lambda, or that the 85% confidence interval be used when comparing lambda values, as it better equates with the z-test at p<0.05. Findings from this study emphasize the importance of and provide recommendations for standardizing, calculating, and interpreting dual-isotope data.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFM.H23B..01O
- Keywords:
-
- 1829 Groundwater hydrology;
- HYDROLOGYDE: 1831 Groundwater quality;
- HYDROLOGYDE: 1832 Groundwater transport;
- HYDROLOGYDE: 1847 Modeling;
- HYDROLOGY