Different tracers result in different storage and transport parameters in transient storage models 

Abstract

Predicting in-stream transport and transient storage processes is crucial for determining biogeochemical turnover and protecting stream ecosystems. One common way to study these processes consists in analysing slug injections of artificial tracers into the stream. The explanatory power of processes inferred from these experiments depends on the quality and completeness of the recorded tracer signal, i.e., the breakthrough curve (BTC). The stream settings strongly influence the explanatory power of the BTC. For instance, an increase in transient storage zones can result in a more pronounced tailing of the BTC. It is well-known that different tracers such as sodium chloride or dye tracers exhibit different detection limits. However, limited guidance exists if and how the choice of the selected tracer or the stream settings biases the conclusion drawn from tracer experiments. To address this research gap, we carried out numerical experiments generating BTCs from slug injections through 10,000 randomly selected parameter combinations mimicking stream conditions. We employed these randomly selected parameter combinations from a predefined range for the dispersion parameter, flow velocity, cross-sectional area, and with and without consideration of transient storage exchange processes. The BTCs were truncated based on the detection limits for sodium chloride and the dye tracer uranine. We calculated transport metrics such as the temporal moments of BTCs and the transient storage index (TSI) to identify differences between the BTCs truncated based on the detection limits of both tracers. We found that the different tracers resulted in clearly different transport metrics. Specifically, the BTCs of the dye tracer exhibited higher TSI values compared to those resulting from BTCs derived from the salt tracer. The absolute and relative differences between the transport metrics of both tracers increased with higher values for the transient storage parameters, particularly for higher flow velocity and a higher dispersion coefficients representative for small streams. Our results revealed that analyzing BTCs from small streams is clearly biased when relying on sodium chloride in the experiments. These finding raise caution considering the importance of the choice of tracer, and we recommend the use of dye tracers over salt tracers for small streams where a high impact of transient storage processes is expected.