From GCM to RCM: How Well Does Dynamical Downscaling Reproduce Southern Tropical South American Moisture Transport and Rainfall?

Simulating precipitation over South America remains difficult because rainfall is tightly controlled by interactions between large-scale circulation and the Andes, whose north–south barrier strongly shapes moisture transport from the Amazon toward southern South America. In the southern tropics, this circulation creates precipitation “hotspots” east of the mountains, needing high-resolution modeling to be correctly represented. This study assesses how well SENAMHI-Peru’s dynamically downscaled regional climate simulations reproduce the key atmospheric circulation patterns linked to continental precipitation, by comparing historical seasonal fields (1981–2014) from both the driving global models (GCMs) and the regional simulations (RCMs) against ERA5 reanalysis. The regional simulations were generated using the Weather Research and Forecasting (WRF) model forced by two CMIP6 global climate models: MPI-ESM1.2-LR-ENS, which includes bias correction prior to downscaling, and NorESM2-MM, which does not include bias correction but demonstrates skill in representing large-scale atmospheric patterns over South America. Using a broad set of diagnostics, the analysis evaluates the representation of major features such as the South American Low-Level Jet, Hadley cell structure, ITCZ position, and local circulation effects. Model–data differences are traced mainly to how topography is represented and to physical parameterizations, providing guidance for improving SENAMHI-Peru’s regionalization protocol ahead of future downscaling experiments.