Photochemical grid models such as the Community Multiscale Air Quality Model (CMAQ) are used to estimate local to continental scale O3, PM, and haze for scientific and regulatory assessments. Field data from specific and well characterized wildland fires is critically important to improve wildland fire emissions estimation approaches, plume transport, and plume chemical evolution in photochemical transport models to further confidence in predictive capabilities to support future scientific and regulatory assessments related to wildland fire impacts. The Fire and Smoke Model Evaluation Experiment (FASMEE) field campaign provides a unique opportunity to obtain surface, canopy, and upper air measurements of specific fire events to better constrain the dynamic nature of smoke emissions and the physical and chemical evolution of smoke plumes from fire events by combustion regime (flaming to residual smoldering). It is anticipated that measurements from this field study will lead to improved 1) fire characterization (size, location, etc.), 2) emission rates for different fuel types and combustion component; 3) PM, VOC and nitrogen speciation by fire type and phase of combustion; 4) allocation of plumes spatially and temporally; 5) near‐fire and downwind plume chemical evolution; and 6) optical properties of plumes. This field study will also provide valuable information to improve other less anticipated aspects of fire emissions and air quality modeling as work intensifies in this research area.

Burn units at Fishlake NF and Fort Stewart, GA planned for inclusion in FASMEE phase II were modeled to illustrate potential impacts on primary and secondarily formed pollutants. Additionally, a prescribed fire at Monument Peak in Fishlake NF from June 2, 2016 was replicated to illustrate how similar the planned burn unit at Manning Creek may be in terms of local to regional scale smoke transport and chemistry. The Manning Creek burn unit showed impacts similar in magnitude to the Monument Peak burn unit but smoke was transported to a different area downwind. Applying CMAQ with finer horizontal grid resolution (1 km compared to 4 km) resulted in higher predicted smoke impacts locally and downwind. The model predicted that O3 formation was inhibited at the fire location due to large amounts of fresh NO emissions but was produced further downwind at both the Fort Stewart and Fishlake burns. The Fort Stewart burn unit was modeled to burn on every day for an entire year to understand potential seasonal differences in smoke levels and composition. The region has enough solar radiation and temperatures are warm enough for O3 to form in all months of the year, although O3 formation was lowest in November and December. This suggests a field study in the southeast U.S. with the intention of examining photochemical changes in smoke should be done outside of those months. Colder weather with typically lower surface layer mixing tend to result in higher concentrations of primarily emitted PM2.5 than warmer seasons based on the Fort Stewart annual simulation. These real and hypothetical burns provide case studies to illustrate the need to evaluate and constrain each component of the modeling system.