Sensitivity of Regional WRF-Chem Air Quality and Climate Simulations to Biomass Burning Emission Datasets: a Case Study of the Impact of Canadian Wildfire to the US

Modelling the impact of remote wildfires on Urban air qaulity has gained importance with more frequent and more extensive wildfires observed in North America, Autralia and rest of the world. Our study focuses on the period 26 to 29 June 2023, when record high Canadian wildfires significantly impacted air quality in the Midwest and East Coast of the US. To investigate aerosol transport from Canadian wildfires to the US and evaluate the performance of a coupled weather-chemistry model in estimating PM2.5, O3, CO and NO2 concentrations and climate feedback from aerosols, we used three different biomass burning datasets (FINN, QFED and RAVE) in the Weather Research and Forecasting Model coupled with Chemistry (WRF-Chem). We compared the simulations with ground-based and remote sensing observations, including Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data, Local Climatological Data (LCD), EPA AirNow observations and Community Research on Climate and Urban Science measurement (CROCUS) measurements. Our findings indicate that a circulation caused by a low-pressure system moving from the Great Lakes to the Atlantic and the high-pressure system over the Atlantic was responsible for the aerosol transport from Canadian wildfires to the US. Both reanalysis dataset is able to capture the circulation. During the wildfire impact period the WRF-Chem underestimated aerosol optical depth (AOD) by more than 50%, PM2.5 concentrations by 65 –- 90%, peak O3 concentrations by 50 –- 55% for all three biomass burning datasets. CO and NO2 are also underpredicted in WRF-Chem. The underestimation of PM2.5 concentrations leads to a temperature overestimation of up to 3.6K due to overestimation of downward short-wave radiation mostly from the underestimation of direct effect of aerosols and the subsequent increase in sensible heat flux. Among the three biomass burning datasets, WRF-Chem with QFED shared the least underestimation of AOD and PM2.5 concentrations due to a more accurate estimation of PM2.5 in wildfire emissions, reducing temperature underprediction by 1.0 –- 1.5K during daytime. We emphasize the need for more accurate predictions of emissions of trace gases and aerosols from wildfire datasets to reduce uncertainties in estimating air pollutant concentrations and the climate feedback from aerosols.