1. Projecting future changes in U.S. forest fuel and fire conditions using NARCCAP regional climate change scenarios Yongqiang Liu Center for Forest Disturbance Science USDA Forest Service, Athens, GA 2012 NARCCAP Users’ Meeting April 10 – 11, 2012. NCAR, Boulder, CO

2. Wildfire and impacts Climate and wildfire Impacts of climate change Application of NACCAP data Results Conclusions and discussion

3. Wildfire in the U.S. 4 million acres (16,000 km 2 ) burned annually in past 50 years. 6 million acres in past decade. Zybach et al. (2009)

4. Fire Regimes Low frequency and high severity in the west, high frequency and low severity in the east

5. 750,000 acres (3,000 km 2 ) burned 22 human lives lost 4,000 homes destroyed Billions of dollars in damage 12,000 firefighters in suppression Half of the cost for heat and drought 2005 Southern California Fires Social and Economic Impacts

6. Environmental Impacts 2.0 Pg C yr −1 during 1997–2009, 1/3 of total carbon emissions (van der Werf et al. 2010) Biomass burning contributes to about 40% of total BC emissions, which play a key role in the smoke- snow feedback mechanism. The radiative forcing of smoke particles reduces surface temperature, cloud and precipitation. Fire events could enhance climate anomalies such as droughts. Wildfires emit large amounts of PM 2.5 and O 3 precursors that lead to regional haze, smog, and visibility degradation MODIS images of the east Amazon basin on August 11, 2002 (Koren et al. 2004).

7. Climate and Wildfire Climate and Wildfire Weather /climate Wind, temperature, relative humidity, precipitation High temperature, low relative humidity, and lack of precipitation can make fire more likely by drying out the fuels. Wind can push fire spread Fuels Light or heavy, arrangement, fuel Moisture The dryer and lighter the fuels the more easily they will ignite. A continuous layer of fuels on the forest floor can aid in the spread of a fire. Topography Flat or slopes, Flat or slopes, aspect aspect A fire moves more rapidly up hills. A fire is more likely on southern and western aspects which are dryer. Atmospheric condition is one of environmental factors for wildfire together with fuel and topography

8. Drought and Mega-fires Big fires – Very large burned areas Big impacts – smoke transported to major metro areas Big efforts – forest management options for risk prevention and impact mitigation prevention and impact mitigation Occurring mostly under drought conditions -1988 Yellowstone fires. About 800,000 acres affected - The northern U.S. drought was among the driest of the 20th century -2011 TX and SE fires. About 4.3 million acres burned. -Worst ever one-year drought -2011 GA/FL Okefenokee fires. About 600,000 acres burned. -Worst drought in Georgia in a century

9. Climate Change and Future Wildfire Climate models have projected overall increase in temperature and more frequent droughts in many mid-latitude regions due to the greenhouse effect. Climate models have projected overall increase in temperature and more frequent droughts in many mid-latitude regions due to the greenhouse effect. It is likely wildfires will increase in frequency and intensity in these regions including the U.S. (IPCC 2007). It is likely wildfires will increase in frequency and intensity in these regions including the U.S. (IPCC 2007). The impact of climate change on fire is already occurring (Westerling et al., 2006) and will become more remarkable by middle of this century in the western U.S. The impact of climate change on fire is already occurring (Westerling et al., 2006) and will become more remarkable by middle of this century in the western U.S. (Spracklen et al. 2009)

10. Future U.S. mega-fires and air quality impacts Project future mega-fire activity under changing climateProject future mega-fire activity under changing climate Assess impacts on air quality in major metropolitan areasAssess impacts on air quality in major metropolitan areas Help mangers develop forest management options for mitigationHelp mangers develop forest management options for mitigation A key component is application of the NARCCAP regional climate change scenariosA key component is application of the NARCCAP regional climate change scenarios - A research project supported by the USDA and USDOI Joint Fire Science Program (JFSP)

11. Applications of NARCCAP Data 1. Calculating fire indices such as Keetch-Byram Drought Index (KBDI) 2007 Okefenokee fires in GA/FL (USFS Wildland Fire Assessment System) KBDI classification 0-200 (low fire potential), 200-400 (moderate), 400-600 (high), 600-800 (extreme) Meteorological variables: maximum temperature, daily rainfall, average annual rainfall. Dependent on historic conditions, more suitable for long-term impact.

12. 2. Driving Dynamic Land Ecosystem Model DLEM - a process-based terrestrial ecosystem model (Tian et al., 2010) Project fuel change in fuel loading, which is a factor for fire emissions

13. 3. Driving smoke and air quality models Simulation of 2007 GA fires with CMAQ

14. 4. Calculating windows for prescribed burning Prescribed burning is a management tool for reducing wildfire risk by removing the accumulating dead fuels. One of management tools for mitigation of future wildfire increase However, there is increasing risk for fire escaping (a control burning becomes a wildfire) due to global warming. Preferred weather conditions for prescribed burning in the southern U.S.: - wind speed at 20-foot above the ground of 6-20 mph; - relative humidity of 30-55%; - temperature of <60oF in winter; - fine fuel (1-hour) moisture of 10-20%, and - KBDI of 250-400.

15. Large values of over 600 (extreme fire potential) all seasons in the inter-Mountains KBDI increases from winter to fall in the East, up to 500 Using HadCM3-HRM3 projection Using HadCM3-HRM3 projection

16. RegionPresentFuture winterspringsummerfallwinterspringsummerfall PS 0.50 0.50 2.13 2.13 2.24 2.240.95 1.97 1.97 1.49 1.49 0.01 0.01 0.07 0.07 PW 0.43 0.43 0.06 0.06 0.80 0.800.17 0.56 0.56 1.25 1.25 1.24 1.24 1.57 1.57 SW 4.15 4.15 4.32 4.32 4.48 4.484.00 -1.44 -1.44-1.83-0.37-0.09 NW 2.58 2.58 2.83 2.83 2.39 2.392.68 2.61 2.61 2.03 2.03 2.87 2.87 3.34 3.34 SC 2.17 2.17 0.79 0.79 4.57 4.574.04 -3.56 -3.56-1.18 1.13 1.13 0.03 0.03 NC 2.64 2.64 1.86 1.86 1.42 1.422.98 0.34 0.34-0.22 1.90 1.90 3.37 3.37 SE -0.30 -0.30 -0.25 -0.25 2.71 2.713.20 -2.20 -2.20 0.61 0.61 1.61 1.61 1.76 1.76 NE 1.09 1.09 0.35 0.35 1.48 1.482.92 -1.37 -1.37 0.03 0.03 2.61 2.61 1.73 1.73 US 1.95 1.95 1.64 1.64 2.77 2.773.03 1.11 1.11 0.40 0.40 1.38 1.38 1.49 1.49 Slopes of fitting linear lines of KBDI variation curves over 30-year periods

17. Increase in Rocky Mountains all seasons Increase in Southeast and Pacific coast in Summer and Fall Decrease in the inter-Mountains all seasons in Winter and Spring

18. Also most remarkable during summer and fall seasons, but with different spatial patterns. Change in KBDI calculated using HadCM3 projection

19. Fire potential increases by one level, from low to moderate or from moderate to high in southern eco-regions

20. Future change in southern fuel loading (Zhang et al., 2009)

21. Future change in fuel moisture

22. The portion reduces in most areas. Largest reduction in southeast by up to 0.3 during summer. Increase slightly in the west coast and inter-mountains. Change in burning window for prescribed burning

23. Largest decrease in during summer and fall and smallest during spring. Relative decrease rate is about 30% during winter and 10% during spring, the two major prescribed burning seasons in the South. Seasonal change in the South

24. Conclusions and Discussion Climate is the most important environmental factor for long-term wildfire variability. Large change in future U.S. wildfire is expected under a changing climate. Climate is the most important environmental factor for long-term wildfire variability. Large change in future U.S. wildfire is expected under a changing climate. High-resolution climate change scenarios are necessary for projection of future wildfire trends. Dynamical downscaling is adding values to traditionally used statistical downscaling by providing unique information for integrated fire projection and impact research. High-resolution climate change scenarios are necessary for projection of future wildfire trends. Dynamical downscaling is adding values to traditionally used statistical downscaling by providing unique information for integrated fire projection and impact research. With the application of NARCCAP regional climate change scenarios, we were able to understand the impacts of climate change on U.S. fire potential, forest fuel conditions, and forest management. The data will be further applied to understanding the air quality impacts of changing wildfire. With the application of NARCCAP regional climate change scenarios, we were able to understand the impacts of climate change on U.S. fire potential, forest fuel conditions, and forest management. The data will be further applied to understanding the air quality impacts of changing wildfire. It would help improve our fire projection and impact research by further understanding the differences between GCM and RCM projections, A2 and other emissions scenarios, and dynamical and statistical downscaling. Transient climate simulation and projection would be useful for projection of forest fuel type changes which usually occur at decadal or longer scales. It would help improve our fire projection and impact research by further understanding the differences between GCM and RCM projections, A2 and other emissions scenarios, and dynamical and statistical downscaling. Transient climate simulation and projection would be useful for projection of forest fuel type changes which usually occur at decadal or longer scales.

25. Thanks!