Modeling Fire Emissions, Black Carbon Concentrations and Deposition Rates


Wildland fires are a major source of particulate emissions, including black carbon (BC). In combination with other emissions, these BC and particulate emissions can directly lead to air quality degradation. BC and other particulate matter (PM) can also affect climate in various ways, including by scattering and absorbing radiation, modifying clouds formation and properties, and changing snow albedo. Assessing the effects of fire BC and particulate emissions on the overall climate system requires placing fire in a larger context of different emissions sources and understanding the evolution and reactions that take place between the different chemical species.


Project Objectives:

  1. What is the contribution of BC from anthropogenic (fossil fuel and biofuel) versus fire emissions (wild and prescribed) to atmospheric BC concentrations in the US?
  2. What is the contribution of primary OC and SOA from fire emissions and how much do these affect the overall PM2.5 contribution from fire emissions?
  3. How do the contributions above vary by season?
  4. What is the contribution of BC from anthropogenic versus fire emissions in the US on BC deposition rates onto snowpack and glacial surfaces in western US?
  5. For all the questions above, what is projected change in 2050s versus present-day?



We use the WRF-BlueSky-CMAQ modeling system for modeling impact of fires on ambient BC and PM2.5 concentrations. Both WRF-CMAQ and BlueSky are commonly used for air quality management.

Figure 1: The modeling framework includes the WRF (Weather Research and Forecast)1 mesoscale meteorological model, BlueSky modeling framework for fire emissions2 (Fig. 2), MEGAN biogenic emission model3, the SMOKE emission processing tool4, and the CMAQ (Community Multiscale Air Quality) chemical transport model5. Figure 2: The BlueSky Modeling Framework for fire emissions2. Default modules used in this study are in bold and underlined: FCCS = Fuel Characteristic Classification System6,7, CONSUME8, FEPS = Fire Emissions Production Simulator9, and WRAP = Western Regional Air Partnership10; different configurations of BlueSky is current being explored. Because we use CMAQ for chemical transport modeling (Fig. 1), the dispersion component of BlueSky is not used in this project.


  1. Skamarock, W. C., et al., A Description of the Advanced Research WRF Version 3, 2008.
  2. Larkin, N. K. et al., The BlueSky smoke modeling framework, Intl J Wildland Fire, 2009;
  3. Guenther, A. et al., Estimates of global terrestrial isoprene emissions using MEGAN, Atmos. Chem. Phys., 2006.
  4. Houyoux, M., et al., Sparse Matrix Operator Kernel Emissions (SMOKE) modeling system v3.0 user manual, 2011.
  5. Byun, D. and Schere, K. L., Review of the governing equations, computational algorithms, and other components of the models-3 Community Multiscale Air Quality (CMAQ) modeling system, Applied Mechanics Reviews, 2006.
  6. McKenzie, D. et al., Mapping fuels at multiple scales: landscape application of the Fuel Characteristic Classification System, Canadian J. Forest Res., 2007.
  7. USDA Forest Service, The Fuel Characteristic Classification System (FCCS), 2008;
  8. USDA Forest Service, CONSUME, 2008;
  9. USDA Forest Service, Fire Emission Production Simulator (FEPS), 2008;
  10. Air Sciences, Inc., 2002 Fire emission inventory for the WRAP Region Phase II. Report prepared for the Western Governors Association/Western Regional Air Partnership, Denver, CO/Ronan, MT, by Air Sciences, Inc., Denver, CO/Portand, OR, Project No. 178-6, July, 2005; 2.pdf


Project Benefits:

In the research proposed here, we anticipate several specific benefits to the land management community by providing a suite of results that will:

  • determine the contribution of fire to BC concentrations relative to other sources
  • assess how these contributions vary by season relative to one another
  • provide one of the first modeling analysis of BC on snow-covered landscape in the US
  • investigate the contribution of BC in the context of total PM2.5
  • evaluate the uncertainty of modeled BC concentrations and deposition rates due to the variance in published BC emission factors.

Historic Fire Records

Historical Fire Records (Western U.S.)

Figure 1: Acres-Days burned in the western U.S. for June-October of 1996-2005. Records are from the Federal Fire History Internet Map Service


Figure 2: Monthly PM2.5 emissions simulated by BlueSky based on historical fire records shown in Fig 1.

Aug 2000 Results

Aug 2000 Results

August 2000 had the highest PM2.5 emissions from fires during the 1996-2005 period; therefore, we focused on August 2000 to perform model-observation comparison and evaluate the maximum fire impact. Observed data from the IMPROVE network were obtained from Overall the model tends to underpredict the impact of fires. Currently, we are investigating how different model treatments of plume height may improve model performance.

Figure 1: Model results for August 2000. Left: monthly total BC emissions from fires (g m-2); right: differences in monthly-mean BC concentrations (μg m-3) between the fire and no fire cases. Locations of sites where model-observation comparisons are shown in Figs. 2 and 3 below are indicated on the right panel.


Figure 2: Observed and modeled 24-hour mean BC (top) and PM2.5 (middle) concentrations at three selected IMPROVE sites for August 2000; also shown are the estimated excess non-soil potassium aerosol concentrations (ex ns-K) (bottom). Ex ns-K is an indicator of fire influence and is calculated from the IMPROVE iron and potassium aerosol data following Park et al, Atmos. Environ. (2007). Locations of the three sites are shown on the right panel of Fig 1.


Figure 3: Scatter plots of observed and modeled 24-hour mean BC (left) and PM2.5 (right) concentrations at 18 IMPROVE sites for August 2000. Locations of the sites are shown in Fig. 1.

Aug 1997-2005 Results

Aug 1997-2005 Results

Here, we present results for August of 1997 to 2005 to document the effects fires have on ambient BC concentrations. Because model results have a negative bias compared to observations from the IMPROVE sites, model results presented here are likely lower-bound estimates of fire influence on downwind concentrations.

Figure 1: Modeled August results averaged over 1997 to 2005. Left: average August-total BC emissions from fires (g m-2); right: average differences in August-mean BC concentrations (μg m-3) between the fire and no fire cases.


Figure 2: Modeled cell-by-cell maximum of August during 1997 to 2005. Left: maximum August-total BC emissions from fires (g m-2); right: maximum differences in August-mean BC concentrations (μg m-3) between the fire and no fire cases.


Figure 3: Comparison of modeled and IMPROVE 24-hour BC concentrations for August of 1997-2005 by region. The whisker plots indicate the median, 25%, 75%, minimum, and maximum values. Region designation is similar to that of Jaffe et al., Environ. Sci. & Techno., 2008.

Civil & Environmental Engineering, PO Box 642910, Washington State University, Pullman WA 99164-2910, 509-335-2576