Calculating a Land Carbon Accounting Factor in the United States: an Example and Implications

Companies that produce and use wood for products and energy find it increasingly important to communicate the carbon balance and potential climate effects of these activities. Computing forest carbon stocks and stock changes, and emissions from operations, are often part of institutional reporting for environmental, social, and governance purposes. This article describes an example methodology to assess forest carbon changes associated with the harvesting of wood products and proposes metrics that could be used to allocate harvesting effects to individual organizations for their reporting purposes. We discuss boundaries (types of forests and carbon pools to include), spatially appropriate evaluations given uncertainty, temporal considerations, risk of reversals, and allocation of net sequestration to products sourced from the region. We also discuss the complex nature of the biogenic carbon cycle and warn about the appropriate interpretation of this methodology. Study Implications: Purchasers of wood products are increasingly interested in the carbon effects of the wood they purchase. For example, are the forests from which this wood was harvested continuing to sequester carbon or are they in decline? One means of communicating this information would be a carbon accounting factor that expresses the net forest carbon change per unit of wood consumed. We describe an approach to develop such a factor and report results for regions of the conterminous United States. However, any single metric is unlikely to fully capture the carbon dynamics of wood sourcing, as illustrated by the carbon stock declines in the Rocky Mountain regions that cannot be attributed to forest harvesting or the very high factors for the Great Plains due to low harvest levels. We discuss several other metrics that can shed additional light on land carbon resiliency and land-use efficiency and could be considered in conjunction with net carbon stock change.