In this study, the Weather Research and Forecasting (WRF) model, along with its coupled meteorology-chemistry model WRF Chem, is utilized to simulate Direct Normal Irradiance (DNI) over the United Arab Emirates (UAE). The research aims to evaluate the impact of Aeolian and anthropogenic aerosols on DNI forecasts, comparing the WRF and WRF Chem models in an arid region. The WRF-Chem model is currently being used operationally for photovoltaic rooftop systems. The simulation covers one month for each season of the year in 2022, representing distinct meteorological and climate regimes in the UAE. WRF Chem model simulations are configured with three nested domains over the Arabian Peninsula, focusing on the UAE. Hourly quality-checked pyrheliometric DNI measurements from DEWA's ground radiometric station are used for model performance evaluation. Results show that explicit modeling of aerosols in WRF Chem significantly enhances forecast accuracy for hourly DNI. The inclusion of aerosol emissions and chemistry leads to substantial improvements in normalized Mean Bias (NMB) and relative Root Mean Square Deviation (rRMSD) for all temporal forecast horizons and seasons. Notably, the incorporation of aerosol processes in WRF Chem improves rRMSD by 33.33% compared to the WRF reference model, with particularly high improvements (over 60%) in forecast horizons day 2 and day 3 for July (summer) and October (fall). The study concludes that explicit treatment of aerosol processes in numerical simulation and modeling significantly enhances forecast accuracy in aerosol-laden arid regions, providing valuable insights for the operational optimization of concentrated solar power plants.