Sensitivity of Regional WRF‐Chem Air Quality and Weather Simulations to Biomass‐Burning Emission Data Sets: A Case Study of the Impact of Canadian Wildfire on the US

Abstract This study focuses on the period from June 26 to 29, 2023, when record‐breaking Canadian wildfires severely impacted air quality in the Midwest United States. Using the Weather Research and Forecasting Model with Chemistry (WRF‐Chem) and four biomass‐burning data sets (Fire Inventory from NCAR version 1, Fire Inventory from NCAR version 2.5, Quick Fire Emissions Data set [QFED], and Regional ABI‐VIIRS Emission), we analyzed aerosol transport from Canada to the US and assessed the model's accuracy in predicting , , and aerosol weather feedback. Model simulations were compared with ground‐based and remote sensing observations as well as field measurements from the Community Research on Climate and Urban Science (CROCUS) project. Our findings show that the movement of a low‐pressure system from the Great Lakes to the Atlantic, combined with the high‐pressure system over the Atlantic, caused the transport of aerosols from Canadian wildfires to the US. Results show WRF‐Chem significantly underestimated key atmospheric components: aerosol optical depth (AOD) by over 50%, by 65%–90% and peak concentrations by 50%–55% across four biomass burning data sets. Additionally, CO and concentrations were underpredicted. The substantial underestimation of led to an overestimation of temperature by up to 3.6C primarily due to excessive downward shortwave radiation, which resulted from the underestimation of direct aerosol effects and an increase in sensible heat flux. Among the biomass‐burning data sets, QFED produced the most accurate AOD and predictions due to improved wildfire emission estimates, leading to a 1.0 to 1.5C reduction in temperature overestimation during the daytime. These findings underscore the need for improving wildfire emission estimates for trace gases and aerosols to enhance air quality and weather feedback predictions.