The Development and Application of a Harmonized Burned Area Data Set for North America to Assess the Effects of Fire Disturbance on the Continental Carbon Budget

Tuesday, 16 December 2014
Guangsheng Chen and Daniel J Hayes, Oak Ridge National Laboratory, Oak Ridge, TN, United States
Fires burn an annual average of about 40,000 km2 in Canada and the U.S., making it an important feature of North American ecosystems through renewing ecosystem conditions and vegetation dynamics. Fire disturbances substantially modify ecosystem carbon dynamics both temporally and spatially. Ecosystems generally lose carbon for several years to decades following fire disturbance, but our understanding of the duration and dynamics of post-disturbance carbon fluxes remains limited. Owing to the prevailing collection of inventory data for fire burn area, intensity, distribution, and associated carbon-related parameters in North America, we are able to more accurately estimate carbon dynamics following fire disturbances. In our study, we integrated four major fire datasets (i.e., U.S. Monitoring Trends in Burn Severity dataset, Bureau of Land Management Alaska Fire Service dataset, and Canadian National Fire Database, and GFEDv3.1 fire dataset) and other auxiliary data to generate a comprehensive and continuous burned area history dataset, which covers the 1920 to 2012 time period and is gridded at quarter-degree resolution for the North American continent. Driven by this new dataset, we used the Terrestrial Ecosystem Model (TEM6.0) to simulate the impacts of fire disturbance on carbon dynamics across North American ecosystems. The results indicate that large amount of carbon was emitted due to fire disturbances during the study period, especially for the boreal ecosystems with slow recovery. The modeling results were also evaluated with the field measurements along a fire chronosequence and compared to estimates from other approaches.