ABSTRACT Tracer‐aided hydrological modelling (TAM) offers a robust framework for enhancing catchment characterisation and understanding of hydrological processes. However, the limited availability of long‐term and high‐resolution stable water isotope (SWI) observations in precipitation has constrained TAM applications. In recent years, the isotope‐enabled Global Spectral Model (IsoGSM) has emerged as a valuable alternative, providing global high‐temporal‐resolution simulations of the SWI composition of precipitation. Nevertheless, IsoGSM outputs have shown biases compared to observations and still require in situ data for adequate correction. In this study, we propose using accessible streamflow SWI observations, such as deuterium, to correct IsoGSM outputs (IsoGSM‐Discharge). We used IsoGSM‐Discharge to drive a spatially distributed tracer‐aided hydrological model (STARR) in a tropical mountainous inter‐Andean catchment to assess hydrological processes. Two additional inputs served as benchmarks: (1) IsoGSM outputs corrected with precipitation deuterium observations (IsoGSM‐Rain), and (2) an interpolated product (Interpolated‐Rain) based on long‐term (2014–2023) weekly precipitation deuterium. Our results show that IsoGSM‐Discharge performs well in simulating both streamflow (KGE: 0.40–0.60) and deuterium in streamflow (KGE: 0.28–0.73), achieving efficiencies similar to the other two products (streamflow KGE: 0.38–0.61; deuterium KGE: 0.27–0.75). During calibration, the IsoGSM‐Discharge approach required additional Monte Carlo simulations to assess parameter sensitivity, resulting in a 33% increase in computational time compared to the Interpolated‐Rain and IsoGSM‐Rain products. We also found that rather than relying on a single best simulation, ensemble modelling is essential for capturing the full range of uncertainty in hydrological processes. These findings suggest that correcting IsoGSM with streamflow isotope data represents a viable and efficient strategy for TAM, improving estimates of soil water movement and groundwater storage variability. This approach bridges global modelling systems and ground‐based observations, enhancing understanding in under‐monitored catchments.
Guamán et al. (2026) studied this question.