Constraining a Tracer‐Aided Hydrological Model With Multi‐Objective Calibration and Young Water Fractions: Toward Improved Process‐Based Simulations

ABSTRACT Tracer‐aided hydrological models are powerful tools for revealing hydrological processes of a watershed of interest, but their performance is often limited by parameter uncertainty. This study aims to develop a transferable, process‐consistent parameter‐constraint framework by combining multi‐objective calibration strategies with observed young water fraction ( F obs ) constraints within the Two‐Reservoir StorAge Selection (TRSAS) model. The model was applied to a humid hilly watershed in southern China to assess how weighting different observational targets (i.e., discharge, streamflow δ 18 O, soil water δ 18 O, and groundwater δ 18 O) affects model performance and parameter identifiability. Calibrating solely to discharge yielded optimal discharge simulations but poor reproduction of streamflow δ 18 O, whereas prioritising streamflow δ 18 O led to consistently strong performance across variables (NSE > 0.70), except for groundwater δ 18 O (NSE < 0.55). The best‐performing scenario assigned the highest weights to discharge (0.4) and streamflow δ 18 O (0.3), achieving NSE values of 0.76 and 0.66 for the calibration and validation periods, respectively. Streamflow δ 18 O was especially effective in constraining both flow and transport module parameters, while soil and groundwater δ 18 O provided limited or localised constraint. Groundwater storage parameters remained poorly identifiable across all calibration strategies. In addition, incorporating F obs did not significantly improve performance metrics but enhanced internal consistency, such as simulated young water fraction ( F sim ) and runoff source apportionment. Overall, the proposed calibration–constraint framework strengthens the process realism and diagnostic capability of tracer‐aided hydrological models and shows strong potential for application in other humid watersheds.