A Weather Driven Wildfire Emission Forecasting System and Its Impacts on Air Quality Prediction

Changing climate has caused an increase in wildland fire activity which is a growing concern for air quality and smoke managers. Wildfires emit trace gases such as NOx, CO and volatile compounds and aerosols which can often result in poor air quality downwind of the fires.  To protect human health from fire impacts, an accurate air quality prediction system is required.  Usually, air quality modeling requires emission processing, which relies on satellite data to provide necessary information for wildfire sources, such as fire area and location.  In forecasting applications, a persistence assumption is commonly used whereby satellite data is used to project emissions for an upcoming day based on the previous day’s satellite data and applying a default, non-dynamic temporal profile. This can result in poor smoke emissions estimates, errors in plume injection height and consequently, errors in the prediction of wildfire effects on air quality.

In this study, we use an alternative method where a numerical coupled fire-atmosphere model (WRF-SFIRE) is used to explicitly resolve fire progression and plume behavior. Results from nested WRF-SFIRE simulations including burn area, heat flux and plume height will be used to improve emissions estimation in AIRPACT, a WRF-SMOKE-CMAQ based regional air quality forecasting system.  This will replace the persistence assumption with more realistic, dynamically determined fire spread and emission estimates. We will share preliminary results from simulations of selected wildfires in the Pacific Northwest, including the Cougar Creek fire of August 2015, and discuss the impacts on smoke dispersion prediction achieved by the coupled fire behavior and air quality modeling.