“…Among the range of machine‐learning algorithms, Random Forests (RF) [ Breiman , ] stands out for its ability to deal with complex nonlinear relationships between variables while minimizing problems with overfitting. Due to its simplicity and capabilities, RF has been used in a wide range of hydrological‐related applications, for example, high‐resolution soil type classification over the contiguous United States [ Chaney et al ., ], seasonal streamflow forecasting [ Zhao et al ., ; He et al ., ], natural flow regime alternation [ Carlisle et al ., ], vegetation‐type distribution [ Peters et al ., ], temperature [ Eccel et al ., ], and wind [ Davy et al ., ] downscaling, and satellite rainfall estimation from cloud physical properties [ Kühnlein et al ., ]. RF also has great potential for statistical precipitation downscaling although there are few studies that have addressed this issue.…”