Utilizing Regional Spectral Model Outputs as Forcing Data for Isotope-enabled Eco-Hydrology Models

Most hydrological models rely on forcing data derived from in situ observations or reanalysis products. However, the scarcity of observational data, particularly isotope measurements, makes long-term hydrological simulations at the catchment scale or larger still challenging. With the ongoing development of global and regional circulation models, it has become feasible and increasingly common to simulate atmospheric variables at high spatial and temporal resolution. At the same time, precipitation isotopes, which serve as important tracers of water sources and physical processes, can now be simulated by several isotope-enabled circulation models. This study uses IsoRSM, an isotope-enabled regional spectral model, to simulate precipitation and its isotopic composition in Central America. The resulting dataset provides a valuable potential input for isotopic eco-hydrological models.

A 14-year (2010-2023) simulation was carried out using IsoRSM at a spatial resolution of 5 km and a temporal resolution of 6 hours. The model domain covered a 6°×6° region encompassing Costa Rica and surrounding areas. The outputs were validated against observations from 58 precipitation-amount sites and 28 precipitation-isotope data sites, and were also compared with a previous IsoRSM simulation at 10 km resolution. After applying quantile mapping (QM) bias correction, the simulated precipitation and isotope fields successfully reproduced the spatial distribution and seasonal patterns across Costa Rica. The average KGE for precipitation δD reached 0.44. Compared with the 10 km simulation, the 5 km resolution produced higher KGEs for precipitation δD, indicating that better representation of topography enhances the simulation of precipitation isotopes. Regarding interannual variability, most sites exhibited a positive relationship between annual mean precipitation δD and the Oceanic Niño Index (ONI), with 12 sites showing correlation coefficients above 0.4. In addition, potential evapotranspiration (PET) could also be calculated from IsoRSM outputs with the Penman-Monteith equation. Overall, the bias-corrected IsoRSM atmospheric variables provide a useful and applicable source of forcing data for isotope-enabled eco-hydrological models.