High‐Resolution Regional Atmospheric Simulation of Precipitation and Isotopes in Central America

Abstract With advances in global and regional circulation models, the high‐resolution simulation of atmospheric variables has become increasingly feasible. In this study, we use IsoRSM, an isotope‐enabled regional spectral model, to improve the simulation of precipitation and its isotopic composition over tropical Costa Rica, which is marked by a pronounced topography and is surrounded by two oceans. A 14‐year (2010–2023) simulation was performed at a relatively high spatial (5 km) and temporal (6‐hourly) resolution over a 6° × 6° domain covering Costa Rica and surrounding regions. Model outputs were evaluated using observations from 58 precipitation‐amount sites and 27 isotope‐monitoring sites and were compared with a 10 km IsoRSM simulation from a previous study over a larger spatial domain. After applying quantile mapping (QM) bias correction, the model generally captured the spatial patterns and seasonal cycles of precipitation inputs and precipitation isotopes across Costa Rica. The mean Kling‐Gupta efficiency (KGE) for bias‐corrected precipitation δ 2 H reached 0.43, and the 5 km simulation outperformed the previous 10 km run, highlighting the benefits of improved topographic representation. IsoRSM also clearly simulated the δ 2 H‐δ 18 O relationship and the interannual variability along the topographical gradient of Costa Rica. Thus, the bias‐corrected IsoRSM outputs offer a promising source of simulated data for further application, such as the forcing data for isotope‐enabled eco‐hydrological models.