Enhancements of urban ozone during wildfire events in the Pacific Northwest

The Pacific Northwest region experiences elevated ozone periods during summers, sometimes due to the emissions of VOCs and NOx (ozone precursors) from seasonal wildfires. AIRPACT5 is an air-quality forecasting system using the CMAQ photochemical grid model to forecast ozone, PM2.5 and related species in the Pacific Northwest, using explicit treatment of wildfire emissions. This modeling system has tended to significantly overestimate surface ozone during wildfire periods, compared to non-wildfire periods. AIRPACT5 ozone results were compared to measurements for an intensive measurement campaign at a site, near Boise, ID, which was strongly affected by wildfire smoke during August and September 2017. The main goals of this study are to explore the issues with simulating ozone during wildfires and to find ways to improve ozone forecasting during such events.

Two likely complications for urban area ozone simulations during wildfire periods are changes in the VOC-to-NOx abundance that impacts ozone photochemistry and aerosol effects on photolysis rates at high PM levels. A significant portion of NOx in a wildfire plume can be converted to peroxyacetyl nitrate (PAN); in turn, PAN can be transported downwind and NOx can be reconstituted, contributing to ozone production. So, correct treatment of NOx-PAN dynamics from emission, through plume evolution, downwind transport, and during ozone production downwind, appears essential to reducing errors in ozone prediction during such wildfire events.

Reduction in downward shortwave (including UV) radiation reaching the surface due to smoke leads to reduction of both photolysis and surface temperature, and as a result reduction in ozone production. Therefore, using corrected photolysis rates and temperature profiles may reduce ozone overestimation. This can be investigated using the off-line, decoupled approach in AIRPACT5, where CMAQ uses WRF forecasts that utilize recent aerosol data to attenuate radiation. Alternatively, this can be done using a fully coupled met-AQ modeling system, as is done using WRF-Chem or integrated WRF-CMAQ.