A modified isotopic mixing model for estimating the contribution of advected moisture to precipitation: A case study in southwest China

Chongqing, located in the upper reaches of the Yangtze River. We develop an RT-constrained, single-isotope (δ18 O) mixing model that uses leaf area index (LAI) to partition summer precipitation moisture into remote advection, local evaporation, and transpiration, addressing limitations from dual-isotope covariance in traditional mixing approaches. For 1981–2017 summers, the modified model tends to yield higher estimates in advected fraction than traditional model, although the inter-model difference is often comparable to propagated uncertainty and should be interpreted cautiously. The shift is physically consistent with differences in the imposed transpiration-to-ET ratio (R T ), which alters the effective isotopic composition of ET vapor and therefore the inferred source fractions under isotopic mass balance. Monte Carlo–Sobol analysis indicates that precipitating vapour isotopic composition dominates uncertainty in the advected fraction, while LAI has a small main effect but a non-negligible interaction effect (10%). This is because precipitating vapor reflects the combined influence of both advected and ET moisture, and its isotopic signature therefore accumulates uncertainties from multiple moisture sources. In contrast, LAI indirectly influences evapotranspiration partitioning and its effect can be further minimized by seasonal aggregation. Overall, the proposed R T -constrained single-isotope framework provides a complementary tool for diagnosing precipitation moisture sources and quantifying uncertainty in inland hydroclimate studies. • The modified isotopic model introduces leaf area index to identify precipitation sources. • The modified model tends to yield higher estimates in advected fraction than traditional model. • The leaf area index contributes the least uncertainty in overall estimation.