Stand age versus tree diameter as a driver of forest carbon inventory simulations in the northeastern U.S.
Abstract
Estimating the current status and future trends of carbon (C) stocks and stock changes in forests of the northeastern United States is desired by policy makers and managers as these forests can mitigate climate change through sequestration of atmospheric carbon dioxide (CO2). We developed C flux matrix models using tree and stand variables by tree diameter class and stand age class to compare size-structured models with age-structured models in their capacity to predict forest C dynamics that are central to policy decisions. The primary control variables for the C flux matrix models (diameter at breast height, stand basal area, stem density, and stand age) were all statistically significant at the α <= 0.05 level. Through comparing the simulation results and root mean square error of C flux matrix models by tree diameter class and stand age class, we found that tree diameter class more accurately predicted C stock change status across the broad compositional and structural conditions in the spatial and temporal domain. An uncertainty analysis revealed that predictions of aboveground C and soil C would be distinctively different whether using tree diameter class or stand age class with high certainty. Overall, this work may enable better integration of forest inventory data and remotely sensed data for the purpose of strategic-scale forest C dynamic simulations.