Answers are grounded in this paper and cite their sources.
The hydraulic connection between groundwater and surface water, resulting from localized leakage in large-scale water conservancy projects, is an urgent issue that requires immediate attention in relevant fields. Using the leakage and tracing of water inside and outside the main canal of the South‑to‑North Water Diversion Project as a case study, this study proposes an intelligent matching algorithm that integrates the Pearson correlation coefficient and Euclidean distance, and employs pre‑screened hydrochemical, isotopic, and ion‑ratio indicators as input features. This algorithm can accurately determine the matching score of the characteristic indexes of different waters. Combined with the results of hydrogeological survey and geophysical scanning, a comprehensive method of water source tracing with high sensitivity is developed. The results indicate that (i) the intelligent matching algorithm can accurately calculate the matching score between different waters, on a scale of 0% to 100%. Using the 95% confidence interval as the threshold, qualitative classification of hydraulic connectivity between water bodies can be rapidly obtained; (ii) The matching results are further substantiated through the combined analysis of hydrogeological survey data and geophysical scanning profiles; (iii) Based on the qualitative results, the mixing ratio between different water bodies can be calculated using end‑member mixing analysis. Taking the North China section of the South‑to‑North Water Diversion Project as a case study, this study applies the proposed integrated method and rapidly identifies three suspected leakage points. The matching scores between the suspected leakage points and ambient groundwater were 83.12% (N-288), 82.7% (N-294), and 82.12% (N-289), respectively. End‑member mixing analysis indicated that the proportions of main‑canal water mixed into groundwater were 88.3%, 38.3%, and 21.7% at the three suspected leakage points, respectively.
No takes yet. Share an insight, caveat, or question.
王伟强 et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: