Challenges in the Application of Existing Process-Based Models to Predict the Effect of Climate Change on C Pools in Forest Ecosystems

Abstract

Process-based models used to investigate forest ecosystem response to climate change were not necessarily developed to include the effect of carbon dioxide (CO2) and temperature increases on physiological processes. Simulation of the impacts of climate change with such models may lead to questionable predictions. It is generally believed that significant shifts in the performance of black spruce (Picea mariana [Mill] B.S.P.) will occur under climate change. This species, which accounts for 64% of Ontario's coniferous growing stock and 80% of the annual allowable cut, represents important economic activity throughout the boreal forest region. Forest management planning requires relatively accurate productivity estimates. Thus, it is imperative to ensure that process-based models realistically predict the effect of climate change. In this study, CENTURY and FOREST-BGC models were calibrated for a productive, upland black spruce stand in northwestern Ontario. Even though both models predicted similar relative outcomes after 100 years of climate change, they disagreed on the impacts of temperature in combination with an increase in CO2. Also, absolute amounts of carbon sequestered varied with climate change scenarios. Comparison of both models indicated that the representation of critical processes in these two forest ecosystem models is incomplete. For instance, the interactive effects of CO2 and temperature increases on physiological processes at stand and soil levels are not well documented nor are they easily identifiable in the models. Their incorporation into models is therefore problematic. Practitioners must consequently be wary of assumptions about the inclusion of critical processes in models.