Improving Air Quality Simulations in California's San Joaquin Valley Using Land Surface Remote Sensing

Abstract Accurate regional air quality modeling is crucial for the development of effective air pollution mitigation plans. Precisely simulating meteorological conditions, especially planetary boundary layer (PBL) properties, is challenging due to complex land surface conditions. In this study, we used the NASA‐Unified (NU) Weather Research and Forecasting (WRF)/Community Multiscale Air Quality (CMAQ) modeling system to assess the benefits of land surface remote sensing in simulating PBL properties and air quality in California's San Joaquin Valley (SJV), a nonattainment region for both PM 2.5 and O 3 in 2018. NASA‐Unified Weather Research and Forecasting Model (NU‐WRF) runs that assimilated remote sensing of land use, leaf area index (LAI), green vegetation fraction, and soil moisture data predicted lower PBL heights, reduced near‐surface wind speeds and temperatures, and higher relative humidity as compared to simulations lacking these data inputs. Air quality simulations without land surface remote sensing underestimated both PM 2.5 and O 3 in the SJV. Air quality simulations including land surface remote sensing showed higher PM 2.5 concentrations in spring, summer, and fall, reducing daily PM 2.5 concentration biases by 10%–72% from March to August. This improvement is primarily attributed to the weaker advection and diffusion driven by lower PBL height and slower wind speed, and enhanced secondary inorganic aerosol formation driven by elevated humidity and lower temperature in the more data‐informed simulations. Annual and seasonal O 3 concentrations were comparable in simulations with and without land surface remote sensing, likely due to the counteracting effects of weaker advection and diffusion and reduced photochemical O 3 formation. These findings highlight the potential for improving air quality simulations by more data‐informed land surface simulations.