Forests play an important role in regulating the Earth’s climate through global biogeochemical cycles. However, a complete understanding of how forests affect these biogeochemical cycles, especially the carbon cycle, is lacking. Tree-rings provide evidence of past aboveground carbon storage that can be used to inform and constrain ecosystem model predictions. However, formal state data assimilation requires the characterization of the uncertainty associated with the data being assimilated. In addition, aboveground biomass reconstructions from living trees typically suffer from the fading record problem- that trees that once contributed to the aboveground biomass pool may have died. In this work, we use tree-rings from a network of sites to estimate aboveground biomass and its uncertainty using a Bayesian hierarchical framework. Then, using relationships among biomass, mean stand diameter, and density from Forest Inventory Analysis data, we develop an approach to correct for the fading record problem. We test this fading record correction at multiple sites (Harvard Forest and Huron Mountain Club) for which we have long-term census and tree-ring data before applying it to sites that have only tree-ring data.
Results/Conclusions
Results from the assessment of the approach developed to correct for the fading record demonstrate both its potential and limitations. Aboveground biomass and biomass increment, both corrected for the fading record, and with characterized uncertainty, are estimated for six sites in the northeastern United States. Analysis of these patterns demonstrate the variability in biomass accumulation as a function of stand development stage and disturbance. Site histories confirm the impact of disturbance events on aboveground biomass accumulation rates. We find that sites that have experienced severe disturbance events show different patterns in aboveground biomass increment. In particular, sites that are in earlier developmental stages show longer-term sustained impacts from severe disturbance events. Results from this work are currently being assimilated into ecosystem models, providing empirical constraints to improve ecosystem forecasts.