Assessing the impact of aerosol optical depth on solar radiation forecasts using WRF-Solar over South Korea

Abstract

This study investigates the impact of aerosol optical depth (AOD) inputs on global horizontal irradiance (GHI) forecasts using the WRF-Solar model during spring in South Korea, when aerosol loading is typically higher than in other seasons. Three aerosol datasets are employed, including climatological aerosol data and hourly AOD forecasts from the Global Earth Observing System Version 5 (GEOS-5) and the Copernicus Atmosphere Monitoring Service (CAMS), which define the WRF-Solar experiments as CLM-AERO, GEOS5-AOD, and CAMS-AOD, respectively. GEOS5-AOD and CAMS-AOD outperformed CLM-AERO, reducing the root mean square error by up to 25.5% under overcast conditions and more accurately capturing the temporal variation and spatial distribution of GHI. Under clear-sky conditions, differences in forecast GHI among the experiments increased with AOD, and the AOD-forced experiments provided more accurate GHI forecasts, indicating the dominant role of the direct effect of aerosols. Additionally, under cloudy and overcast conditions, improvements in the AOD-forced experiments became more evident, reflecting aerosol-induced changes in cloud microphysics. These findings highlight that incorporating operational AOD forecasts into WRF-Solar improves solar radiation forecast accuracy, which is critical for day-ahead energy planning and grid reliability, as the expansion of renewable energy increases the demand for accurate solar forecasts. © 2026 Published by Elsevier Ltd.