Boreal forests contain over 30 percent of Earth’s terrestrial carbon and are currently an important land carbon sink. However, over the past century boreal forests have experienced significant warming and drying. It is projected that warming temperatures and hotter drought conditions will intensify, making the future role of the boreal forest in the terrestrial carbon cycle uncertain. Terrestrial biosphere models (TBMs) are important tools for diagnosing the current state of the carbon cycle and forecasting ecosystem responses to global change. Yet, TBMs often oversimplify environmental and biological controls on tree growth, recruitment, and mortality, and few studies assess how this affects predictions of decadal-scale forest biomass dynamics. In this study, we present the results of a case analysis of the biological and environmental controls on forest growth, recruitment, and mortality in the Alaskan boreal forest using forest inventory measurements. We then perform a regional-scale model evaluation using the boreal forest version of Ecosystem Demography model 2 (ED2-boreal). We evaluated ED2-boreal against forest inventory measurements in Canada and Alaska using simulations with different growth and mortality schemes to assess the relative impact on model performance against forest inventory biomass dynamics.
Results/Conclusions
The Cooperative Alaska Forest Inventory (CAFI) consists of field-gathered inventory measurements from 570 permanent sample plots distributed across interior and south-central Alaska including the Kenai Peninsula. Over 60 percent of the permanent sample plots have been remeasured and over 20 percent have been remeasured three times. Our analysis of the CAFI over the 1994-2013 period shows significant temporal trends in tree growth, recruitment, and mortality that vary with permafrost distribution, rainfall, biological disturbance events, and tree species. In ED2-boreal, we (1) parameterized new plant functional types representative of dominant forest inventory tree genra including spruce, poplar, pine, fir, and larch, (2) optimized tree mortality based on our CAFI analysis, and (3) optimized belowground tree growth parameterizations. We found that incorporating the new mortality scheme and belowground parameterizations offered improvements in predicting decadal-scale forest biomass dynamics over the traditional model. These results imply that the projected carbon budget of the boreal zone is sensitive biological and environmental controls on tree growth, mortality, and recruitment and that TMBs such as ED2-boreal can more successfully capture regional-scale variability in ecosystem carbon dynamics by including increasingly mechanistic representations of the processes behind individual tree growth and mortality.