An Evaluation of WRF‐Chem's Performance in Simulating PM2.5 in West African Cities

Abstract West Africa experiences some of the highest particulate pollution levels globally due to Saharan dust, biomass burning and urban emissions. Despite the severe health burden, air quality monitoring remains sparse. This study evaluates WRF‐Chem's performance in simulating PM2.5 levels across seven West African cities in 2021 using newly available AirNow observational data sets. Results show that while the model effectively captures seasonal PM2.5 variability, it largely overestimates concentrations in northern landlocked cities during the dry season (November–April), including N’Djamena (mean bias (MB) = 86.2 ) and Bamako (MB = 102.3 ) due to excess dust, and underestimates PM2.5 in Lagos during the wet season (MB = −19.5 ), when dust contributions are minimal. Daily PM2.5 correlations are highest during the dry season at sites with dense observational coverage, including N’Djamena ( R = 0.63), Abidjan ( R = 0.63), and Accra ( R = 0.64), but correlations mostly decline during the wet season when non‐dust PM2.5 emissions become more prominent. Evaluations of AERONET aerosol optical depth further indicates widespread dry‐season overestimation at northern sites, supporting overestimated dust emissions in the model. Meteorological evaluation shows good representation of seasonal monsoon transitions but reduced skill in coastal wind fields, which may contribute to PM2.5 errors in coastal cities. Additionally, positive model biases in meridional wind across northern cities may contribute to enhanced simulated dust emissions and transport. Overall, these findings call for improved dust emissions, non‐dust PM2.5 evaluations, meteorological validations, and expanded monitoring for reliable air‐quality forecasting and public‐health protection.