Optimal selection of hydraulic indexes with classical test theory to compare hydraulic performance of constructed wetlands

Residence time distributions (RTDs), obtained from tracer experiments, are one of the main tools for investigating the hydraulic performance of constructed wetlands. However, the existence of various hydraulic indexes derived from RTDs has led to inconsistency in their application for comparing the hydraulic performance. In this work, eight hydraulic indexes were initially selected based on their popularity, and then divided into three categories: hydraulic efficiency indexes (HEIs) (lambda(m) lambda(e), lambda(p), and MI), short-circuiting indexes (SIs) (t(5) and t(10)), and mixing indexes (MIs) (sigma(2) and Morril index). Then, the hydraulic indexes were optimized for compatibility, discrimination, difficulty, and their mutual relationships between different categories. The results showed large inconsistencies among HEIs, and small inconsistencies among SIs and MIs. Among the four HEIs, lambda(e) performed best in terms of compatibility, discrimination, and difficulty. Among SIs, t(5) and t(10) differed little in the three aforementioned aspects. Among MIs, the Morril index performed better than sigma(2) in terms of discrimination and difficulty. The significant correlation between short-circuiting flow and hydraulic efficiency highlights the necessity of reducing short-circuiting flow. Within each category, lambda(e) t(10), and the Morril index are recommended to quantify hydraulic efficiency, short-circuiting flow, and mixing flow, respectively. This study brings clarity to the application of hydraulic indexes and provides uniform standards for the quantification of hydraulic performance.