Evaluation of Wildfire Emission Inventories Using Airborne Flux Measurements

Wildfires impact air quality, atmospheric composition, and the climate system. Emissions estimates of biomass burning (BB) are thus fundamental inputs to chemical transport and atmospheric circulation models used to predict the air quality and climate impacts due to wildfires. Several emissions models have been developed to estimate BB emissions using either satellite observations of fire radiative power (FRP) or burned area. Predictions between inventories often vary by several orders of magnitude, even for inert species such as CO, complicating the interpretation of the BB impacts. The evaluation of emission inventories is, therefore, an essential first step to constrain emissions and improve model predictions. Until recently, opportunities to evaluate emissions estimates have been limited, particularly at the fire-event level.

The University of Colorado Airborne Solar Occultation Flux (CU AirSOF) instrument was deployed in the summer of 2018 as part of the Biomass Burning Flux Measurements of Trace Gases and Aerosols (BB-FLUX) field campaign in the northwest United States. CU AirSOF is highly innovative, in that the instrument can determine highly time-resolved mass fluxes for many chemical species (incl. CO) from measurements on the scale of actual wildfires. Over 150 mass flux measurements were performed on 18 separate fires spanning different fuel types, fuel loadings, and fire intensities. This rich dataset provides a unique opportunity to evaluate and constrain emission inventories.

Here, we used measured mass fluxes of CO to evaluate eight BB emission inventory datasets at the fire-event scale. The eight datasets differ in their spatial coverage, resolution, and estimation methods. The datasets are: (1) the Blended Global Biomass Burning Emissions Product GBBEPx; (2) the Quick Fire Emissions Dataset QFED v2.5; (3) the Global Fire Assimilation System GFAS v1.2; (4) the emissions-product used in NOAA's High-Resolution Rapid Refresh with Smoke (HRRR-Smoke) model; (5) the Fire INventory NCAR FINN v1.5; (6) the National Emissions Inventory (2017 -NEI); (7) the non-official Western Wildfire Campaign emissions product (Hudak, Ottmar & Restaino); and (8) the non-official emissions-product for the 2018 fire season (O'Neill et al.).