Here we present high spatiotemporal resolution simulations of meteorology, PM and photo
hemistry over Utah from January 13 to February 1, 2017 using the WRF/Chem regional p hotochemical model. Correctly characterizing the meteorology is difficult due to the complex terrain and shallow inversion layers. We discuss the approach and limitations of the simulated meteorology, and evaluate low-level pollutant mixing using vertical profiles from missed airport approaches by the NOAA Twin-Otter, which were performed routinely for each flight.
Full photochemical simulations are calculated using NOx, ammonia and VOC emissions from the U.S. EPA NEI-2011 emissions inventory. Comparisons of the observed vertical column amounts of NOx, ammonia, aerosol nitrate and ammonium with model results show that the inventory estimates for ammonia emissions are low by a factor of four and NOx emissions are low by nearly a factor of two. The partitioning of both nitrate and ammonia between gas and particle phase depends strongly on these ammonia and NOx emissions to the model, and the model calculated NOx to nitrate conversion rates. These rates are underestimated by gas-phase chemistry alone, even though surface snow albedo increases model photolysis rates by nearly a factor of two. Several additional conversion mechanisms are added and evaluated in the model, including: heterogeneous nitrate to aerosol formation, catalytic conversion of NO2 to HONO and HNO3 at the snow surface, and direct HONO emissions from vehicles. By including the emission changes and additional NOx to nitrate conversion mechanisms, the model is able to match the observed NOx, total nitrate, and total NH3/NH4 levels within 25% for median statistics over the 15 Twin-Otter flights of the model study period.