- Managing wildlands to protect species and ecosystem services in response to climate change is challenging. To develop effective long-term strategies, natural resource managers need to account for the projected effects of climate change as well as the uncertainty inherent in those projections. Vegetation models are one important source of projected climate impacts. Interpreting those model results can be difficult due to both uncertainty in results and model limitations. Factors contributing to uncertainty include embedded assumptions about atmospheric CO2 levels, uncertain climate projections driving models, and model algorithm selection. Limitations include processes excluded by models, such as mortality from maladaptation and succession, as well as algorithmic simplifications such as assumptions about wildfire ignitions.
To understand the potential impacts of climate change on vegetation and wildfire in 21st century, I used the MC2 dynamic global vegetation model (DGVM) to simulate vegetation for the Northwest conterminous United States using results from 20 different Climate Model Intercomparison Project Phase 5 (CMIP5) models downscaled using the MACA algorithm. Results were generated for representative concentration pathways (RCP) 4.5 and 8.5 under vegetation modeling scenarios with and without fire suppression for a total of 80 model runs for future projections. For analysis, results were aggregated by three subregions: the western Northwest (WNW), from the crest of the Cascade Mountains west; Northwest plains and plateau (NWPP), the non-mountainous areas east of the Cascade Mountains; and eastern Northwest mountains (ENWM), the mountainous areas east of the Cascade Mountains.
To understand MC2 sensitivity to model assumptions, I further explored results and associated uncertainties from the MC2 Dynamic Global Vegetation Model for the WNW subregion. I compared model results for vegetation cover and carbon dynamics over the period 1895-2100 assuming: 1) unlimited wildfire ignitions versus stochastic ignitions, 2) no fire, and 3) a moderate CO2 fertilization effect versus no CO2 fertilization effect.
Finally, I implemented an Environmental Evaluation Modeling System (EEMS) decision support model using MC2 DGVM results to characterize biomass loss risk for the WNW subregion. Risk was based on biomass present, fire occurrence and severity, and mortality of climate-maladapted vegetation as indicated by modeled vegetation type change. I characterized the uncertainty due to RCP, fire suppression, and climate projection choice, and I evaluated whether fire or climate maladaptation mortality was the dominant driver of risk.
In the 21st century, in the WNW, mean fire interval (MFI) averaged over all climate projections decreases by up to 48%. By the end of the 21st century, potential vegetation shifts from conifer to mixed forest under RCP 4.5 and 8.5 with and without fire suppression. In the NWPP, MFI averaged over all climate projections decreases by up to 82% without fire suppression and increases by up to 14% with fire suppression resulting in woodier vegetation cover. In the ENWM, MFI averaged across all climate projections decreases by up to 81%, subalpine communities are lost, but conifer forests continue to dominate the subregion in the future.
In evaluating the effects of ignition and CO2 fertilization assumptions, the greatest carbon stock loss in the WNW, approximately 23% of historical levels, occurs with unlimited ignitions and no CO2 fertilization effect. With stochastic ignitions and a CO2 fertilization effect, carbon stocks are more stable than with unlimited ignitions. For all scenarios, the dominant vegetation type shifts from pure conifer to mixed forest, indicating that vegetation cover change is driven solely by climate and that significant mortality due climate-maladapted vegetation as indicated by modeled vegetation shifts are likely through the 21st century regardless of fire regime changes.
The risk of biomass loss in the WNW generally increases in current high biomass areas within the study region through time. The pattern of increased risk is generally south to north and upslope into the Coast and Cascade mountain ranges and along the coast. Uncertainty from climate future choice is greater than that attributable to RCP or +/- fire suppression. Fire dominates as the driving factor for biomass loss risk in more model runs than mortality due to climate maladaptation. This method of interpreting DGVM results and the associated uncertainty provides managers with data in a form directly applicable to their concerns and could prove helpful in adaptive management planning at regional to local scales.