Integration of disturbance patterns into carbon (C) flux estimates to improve terrestrial-atmosphere C exchange is a critical priority for the North American Carbon Program. This project is built upon previous finding from The Chequamegon Ecosystem Atmospheric Study and aims to quantify uncertainty in C flux upscaling due to disturbance and seasonality, evaluate multiple disturbance stressors, and develop two-way communication channels between federal agencies and scientists. This project asks three main questions: (1) Does incorporation of variation in physiological model parameters improve seasonal and interannual CO2 flux hindcasts from eddy flux towers? (2) Does incorporation of stand-replacement and partial disturbance processes from remotely sensed observations improve yearly to decadal CO2 flux hindcasts from eddy flux towers? and (3) To what degree does model-data integration aid regional and landscape decision-making for forest C storage management?
Results/Conclusions
We show that parameter and prediction uncertainty in terrestrial C upscaling increases with increasing stand age with a slight decline at the end of the stand age spectrum (old stands). The maps of regional C fluxes and uncertainty under different management scenario are of interest to the park managers. The finding from this project directly contribute to national efforts to constrain uncertainty in terrestrial-atmospheric C exchange in several important ways. First, it utilizes a new disturbance algorithms using Landsat imagery to test whether inclusion of partial and stand-replacing disturbance reduces uncertainty in C flux upscaling. Second, it employs a computationally tractable but responsive photosynthetic model to evaluate whether a more sophisticated parameterization of plant physiology aids temporal diagnosis of C flux estimates. Third, by collaborating with regional and national Forest Service personnel, this project partially addresses the ‘end-to-end’ problem of C cycle science by helping managers to diagnose adaptive capacity of forested landscapes, target locations, and prioritize C management activities.