COS 151-6 - A new dynamic fire and forest model for diagnosing the causes and consequences of increased burning in the western United States

Background/Question/Methods

Fire is a dominant disturbance in the temperate and boreal biomes, and increasing burned area with climate change may fundamentally alter forest cover and structure. Improved information about how fire-induced changes to forests feedback to affect subsequent burning could inform mitigation and adaptation strategies to protect people and the provision of ecosystem services. However, fire is often simplistically represented or ignored in broad-scale vegetation models, and detailed fire models often assume static fuels. To address this challenge, we developed the Dynamic Temperate and Boreal Fire and Forest-Ecosystem Simulator (DYNAFFOREST). DYNAFFOREST represents the hierarchical structuring of forests, from the individual cohort to the continent, making it possible to explicitly capture dynamic feedbacks between fire and forests at broad scales over decades to centuries. We parameterized DYNAFFOREST for the data-rich forests of the western United States. We then ran DYNAFFOREST for 300 years forced with mid-20^th century climate and benchmarked model output with observations from the USDA Forest Inventory and Analysis plot network and remotely sensed fire perimeters. Finally, we conducted an experiment with DYNAFFOREST to ask; how do recent increases in fire across the western US compare to the historical range of variability in fire activity before Euro-American settlement?

Results/Conclusions

Modeled distributions of 12 PFTs closely matched current biogeographic patterns of dominant forest type across the western US after 300 years. Forest structure and downed fuels also corresponded well with observations. Modeled median tree heights and DBHs differed from observations by 16% and 20%, modeled median leaf and aboveground wood biomass were within 29% and 12% of observed medians, respectively, and median litter and coarse wood biomass were within 5% and 7%, of observed values. The current fire module produced fire frequencies, sizes, annual burned areas, and fire severities consistent with average 20^th-century fire seasons across a spatial domain with wide ranging fire regimes, but missed the extremely large fire events that were relatively rare in the last century. A more advanced fire module is in development. Across the western US, annual forest fire area has been increasing at astonishing rates. However, results of an experiment with DYNAFFOREST indicate that recent fire seasons are still within the range of past variability, indicating potential for continued acceleration of burned area over coming years to decades before fuel limitations emerge. This study highlights how DYNAFFOREST could be a powerful tool for informing mitigation and adaptation strategies.