气候变化领域集成服务门户

Prescribed burning is practiced to benefit ecosystems but the resulting emissions can adversely affect air quality. A better understanding of the uncertainties in emission estimates and how these uncertainties affect smoke predictions is critical for model-based decision making. This study examined uncertainties associated with estimating fire emissions and how they affected smoke concentrations downwind from a prescribed burn that was conducted at Eglin Air Force Base in Florida, US. Estimated variables used in the modelled emission calculation were compared with field measurements. Fuel loadings, fuel consumption and emission factors were simulated using Photo Series, Consume, and previously published values. A plume dispersion model was used to study the effect of uncertainty in emissions on ground concentration prediction. The fire emission models predicted fuel loading, fuel consumption and emission factor within 15% of measurements. Approximately 18% uncertainty in field measurements of PM2.5 emissions and 36% uncertainty attributed to variability in emission estimating models resulted respectively in 20% and 42% ground level PM2.5 concentration uncertainties in dispersion modelling using Daysmoke. Uncertainty in input emissions influences the concentrations predicted by the smoke dispersion model to the same degree as does the model's inherent uncertainty due to turbulence.