Limitations and potential applicability of mixing models and transit time estimates for mine-affected watersheds based on stable isotopes

Stable isotopes are a widely used tool in defining mixing of distinct waters and identifying solute sources on mine sites. Although the utility of stable isotopes in these approaches is well-acknowledged, analytical uncertainties and lack of endmember definition may limit the application of results in providing unambiguous information to facilitate remediation. In this context, this study used a multi-year high-frequency (monthly to weekly) dataset of δ²H, δ¹⁸O, and δ³⁴S collected from streams surrounding a mine site with a draining adit, the Captain Jack Mill superfund site near Boulder, Colorado, USA, to evaluate mixing between multiple distinctive potential solute and water sources. Iterative endmember mixing analysis was conducted to evaluate if stable isotopes can discriminate between mine water that is evaporatively fractionated in a surface pond versus diffuse discharge of mine water to the stream through the streambed. Timeseries analysis of paired estimated precipitation and stream isotopes was also conducted to determine if mine water discharge affects watershed-scale transit time metrics. Results indicate that model uncertainty propagated from analytical uncertainty overwhelms endmember mixing results using δ²H and δ¹⁸O because all water on the site is sourced from meteoric precipitation, limiting the definition of distinct endmembers. In contrast, endmember mixing results using δ³⁴S are highly sensitive to the simulated mixing fractions and results indicate that approximately half of the sulfate in the downstream stream water is derived from the mine adit and half is sourced from upstream watershed sources such as surface mine waste piles or diffuse groundwater discharge. Transit time metrics are statistically indistinguishable from upstream to downstream of the mine, meaning that mine water discharge does not cause a shift in transit times at the watershed scale. This study outlines notable limitations in the application of stable isotopes for informing remedial decisions at mine sites.