Carbon accounting of ponderosa pine forests across the interior western U. S. based on tree-ring and forest inventory data: Drivers of carbon stock and flux and their uncertainties

Kelly Heilman, Jacob Aragon, Andrew T. Gray and Margaret Evans, Laboratory of Tree-Ring Research, University of Arizona, Tucson, AZ, Michael C. Dietze, Earth and Environment, Boston University, Boston, MA, Andrew O. Finley, Department of Forestry, Michigan State University, East Lansing, MI, Alexis H. Arizpe, School of Natural Resource and the Environment, University of Arizona, Tucson, AZ, John D. Shaw, Forest Inventory and Analysis Program, USDA Forest Service, Ogden, UT, Stefan Klesse, Swiss Federal Research Institute WSL, Birmensdorf, AZ, Switzerland, R. Justin DeRose, Department of Wildland Resources, Utah State University, Logan, UT

Background/Question/Methods
Forest responses to climate change are highly uncertain, but critical for forecasting and managing carbon dynamics. Large-scale, robust carbon accounting of the forest carbon sink is needed to guide efforts to drawdown atmospheric CO2 concentration and repair the climate system. The US Forest Service Forest Inventory and Analysis (FIA) program provides estimates of standing carbon stocks, but lack detail about how annual climate variation affects carbon uptake. Tree-ring time series data can fill this gap, providing annually resolved growth responses to climate. Here we fused tree-ring time series and tree diameter data from FIA plots in the interior west US, within a Bayesian state-space model, to forecast tree size and growth at a regional scale. We included the effects of interannually varying climate, density-dependant competition, and site quality on tree growth. Hindcasted and forecasted tree diameter and diameter increments were then used to estimate aboveground biomass (stock) and biomass increment (flux), and their associated uncertainties. Finally, we integrated biomass estimates across trees to estimate forest stand aboveground biomass and biomass increments.
Results/Conclusions
Information from >900 ponderosa pine with tree-ring plus repeated tree diameter measurements allowed us to forecast tree growth and aboveground biomass. Assimilation allowed for forecasts of thousands of additional ponderosa pine across the intermountain western U.S. that lacked tree-ring data but had repeat diameter measurements. For a set of >800 ponderosa pine-dominated forests, tree-level estimates of aboveground biomass were used to estimate plot-level aboveground biomass and biomass increment, constrained by both standing tree stocks (diameters) and fluxes (tree rings). Propagating and parsing uncertainty in these forecasts indicated that the primary causes of uncertainty differ between stocks (diameter and biomass) versus fluxes (annual increments). We also found that while tree diameter uncertainty increases over time, allometric uncertainty drives much of the increase in aboveground biomass uncertainty. The ecological forecasting approach presented here lays the groundwork for fusing tree-ring and forest inventory data to forecast aboveground biomass and perform robust carbon accounting at tree, plot, and regional scales. Here we present the first example of regional-scale forest carbon accounting constrained by data assimilation of tree rings and other forest inventory measurements in a state-space model, demonstrating that tree-ring constrained forest carbon accounting can be scaled up to a large regional, hemi-continental scale.