Calibration and Uncertainty Analysis of a Carbon Accounting Model to Stem Wood Density and Partitioning of Biomass for Eucalyptus globulus and Pinus Radiata

Author： Paul Keryn Polglase Phil Snowdon Peter Theiveyanathan Tivi Raison John Grove Tim Rance Stan

Publisher： Springer Publishing Company

ISSN： 0169-4286

Source： New Forests, Vol.31, Iss.3, 2006-05, pp. : 513-533

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

There is a need to calibrate models for carbon accounting in forest systems if they are to be applied for carbon trading and off-set schemes. One such model, Full Carbon Accounting Model (FullCAM), calculates stem mass by taking annual inputs of tree growth in stem volume and multiplying these by basic stem wood density. Stem mass is then multiplied by user-entered coefficients to determine the mass of other tree components. Using datasets of Pinus Radiata and Eucalyptus globulus that comprised of between 73 and 187 observations, we determined empirical relationships that can be used in FullCAM to relate basic stem wood density to stand age, and masses of bark, foliage or branches to mass of stem wood for these two species. All fitted relationships were highly significant (p < 0.001), explaining between 35 and 89% of the variance. These calibrations were then tested using three case studies where data on volume yield curves and repeated measures of biomass of stand components were available: one of P. radiata and two of E. globulus. Although accumulation of biomass in foliage and branches were not well predicted by the model, sensitivity analysis showed that this was relatively unimportant to total carbon storage because of the dominance of the stem, particularly once the stand is older than 5 years. Indeed, FullCAM accounted for 99% of the variance in measured above-ground biomass at all three sites because calibrations for the mass of stem was reasonably well constrained. Uncertainty analysis showed that despite the standard errors of parameters used in relationships for basic density and biomass partitioning, and for estimates of carbon contents of tree components, we can be 95% confident that sequestration of carbon in trees and debris of Pinus Radiata and Eucalyptus globulus plantations are, on average, within 13% of that predicted by FullCAM.