Evaluación de los resultados de los modelos climáticos globales del proyecto CMIP5 y CMPI6 para la precipitación y temperatura para Colombia

Abstract

Climate change has significantly altered global temperature and precipitation patterns, with particularly severe impacts on vulnerable regions such as Colombia. In this context, Global Climate Models (GCMs) from the CMIP5 and CMIP6 projects serve as valuable tools for projecting and understanding future climate trends. This study aimed to evaluate the performance of these models in representing the annual cycle of air temperature and precipitation in Colombia, using observational data from IDEAM and simulations from selected global models. The methodology was structured in three phases: (1) collection and processing of observational and reanalysis data; (2) spatial interpolation using the Inverse Distance Weighting (IDW) method to construct multi-annual climatological maps of temperature and precipitation; and (3) statistical validation of model outputs using metrics such as RMSE and Pearson correlation, compared against data from meteorological stations located at the country’s main airports. The results revealed a sustained increase in average monthly temperatures across most of the national territory between the periods 1971–2000 and 1991–2020, with a more pronounced rise in the Caribbean and Orinoco regions. In terms of precipitation, spatial redistribution and a reduction in peak values were identified, particularly in highly humid areas such as the Pacific coast. The evaluated models displayed varying levels of accuracy depending on the region and variable, with FGOALS-gl (CMIP5) emerging as the best-performing model overall. It is concluded that, while global models have limitations in accurately simulating Colombia’s complex climate, their integration with local observational data provides valuable insights for environmental management and territorial planning. The use of downscaling techniques and continuous local validation is recommended to enhance applicability, as well as incorporating these findings into climate change adaptation and mitigation policies.