Evaluation of high-resolution gridded climate products in reproducing spatial and temporal variation in precipitation in central Panama

Tropical forests vary widely in their precipitation regimes and seasonal water availability, but high-quality in-situ (ground-based) meteorological data are rare, and few studies have evaluated the performance of global gridded climate products in the tropics. We compared the performance of eleven high-resolution gridded climate products against in-situ datasets spanning high rainfall variation in central Panama. The gridded products differed widely in their performance, although most captured the broad trends of spatial and seasonal rainfall variation in central Panama, and all underestimated precipitation in the wettest sites, especially in the dry season. Spatial, seasonal and interannual variation were best captured by CHIRPSv2; downscaled finer-resolution products performed similarly well in reproducing spatial variation. Our ability to quantify performance was constrained by limited data availability, even in this region with relatively many in-situ datasets, highlighting the need for more investment in ground-based data collection in the tropics.