Evaluation of Different Types of Aerosols From Wildfires in South America Using Ground‐Based Observations, Reanalysis and CMIP6 Models

The validation of aerosol optical depth (AOD) data is crucial for understanding the impact of wildfires on climate, particularly in regions such as South America, where this phenomenon is recurrent. This study assesses the performance of two global re-analysis products (MERRA- 2 and CAMS) against the AERONET observation network for 2003–2023 and, for the overlapping period 2003–2014, also compares both reanalyses and CMIP6 global climate models. Results show that both reanalyses capture the spatiotemporal variability of AOD, with CAMS performing better in representing fire- derived aerosol types, including Black Carbon AOD (BC AOD) and Organic Carbon AOD (OC AOD). Seasonal patterns indicate peak AOD during the dry season (August–October), coinciding with increased fire activity, while systematic overestimation occurs in regions with complex topog-raphy. CMIP6 models reproduce the seasonal cycle of total AOD but underestimate BC AOD and OC AOD concentrations when compared with CAMS. These results underline the need for improved model parameterizations and for expanding and maintain-ing ground- based observation networks such as AERONET to enhance the reliability of aerosol simulations over South America.