Historically, assessments of the forest resource situation have focused on timber supply, and the data used to support the assessments came from traditional forest inventories designed to provide reliable estimates of timber volume, growth, removals, and mortality (U.S. Department of Agriculture, Forest Service 1982). The most recent assessment included data and analysis of forest resources other than timber, including wildlife, range, water, recreation, and other resources associated with the Nation's forest lands (U.S. Department of Agriculture, Forest Service 1989). Future forest resource assessments will include expanded analyses of environmental issues such as the effects of acid deposition on forest health, the prospective effects of global warming on forests, and the impacts of prospective strategies to mitigate or adapt to changing environmental conditions.

A key issue analyzed in the 1989 Resources Planning Act (RPA) Assessment is the impact of climate change on America's forests (Joyce and others 1990). Another issue undergoing intense analysis at this time but not included in the 1989 RPA Assessment is the evaluation of forestry opportunities for mitigating the effects of global warming. Analysis of forestry opportunities requires knowledge of carbon storage and accumulation in forest ecosystems. It is the purpose of this publication to provide estimates of carbon storage and accumulation for U.S. forests. Because it takes years to design and conduct detailed inventories of U.S. forest lands, the only way to satisfy current, expanding information needs is to integrate the best available data from the national inventory sample with data from special studies of selected forest ecosystems.

Forests and the Global Carbon Cycle

Carbon dioxide in the atmosphere has been increasing steadily since at least 1958 (Keeling 1984). Predictions of future climate change as a consequence of increasing atmospheric carbon dioxide vary widely. Under a scenario of equivalent doubling of atmospheric carbon dioxide by the middle of the next century, most predictions show an increase in average global temperature of between 2 and 5 degrees centigrade and an increase in average global precipitation of between 7 and 15 percent (Schneider 1989). These prospective changes have generated interest in strategies to reduce emissions of carbon dioxide to the atmosphere, or to offset emissions by storing additional carbon in forests.

The total amount of carbon in the atmosphere has been estimated at 720 billion metric tons, the total amount of carbon in terrestrial biomass is about 560 billion metric tons, and the total amount of carbon in terrestrial soils is about 1,500 billion metric tons (Solomon and others 1985). Although oceans store a far greater amount of carbon than terrestrial ecosystems, our ability to manage terrestrial ecosystems is greater and likely to have a greater mitigation effect.

Forest ecosystems are capable of storing large quantities of carbon in solid wood and other organic matter. Forests may add to the pool of carbon dioxide in the atmosphere through burning of forest lands, deforestation, or decomposition of wood products and byproducts. Forests may also reduce the amount of carbon dioxide in the atmosphere through increases in biomass and organic matter accumulation. Young, growing forests take up carbon at high rates, while carbon uptake in mature forests is balanced by carbon release from decaying vegetation. The end use of timber harvested from forests is an important factor in evaluating the contributions of forestry to the global carbon cycle. If the end uses of forest products are in long-term durable goods such as furniture or timber bridges, the carbon is stored in those materials. If the end use is for paper products that are rapidly used and discarded to decay, then the carbon is released to the atmosphere. Carbon in waste from the manufacturing process and discarded wood products may be sequestered in landfills for long periods of time. When forest biomass is burned for energy it may be substituted for fossil fuels, which is an effective way to reduce the depletion of nonrenewable fossil carbon.

Because of the relation between forests and atmospheric carbon dioxide, there are opportunities to manage forests in ways that would result in storage of additional carbon and thus reduce atmospheric carbon dioxide. Major forestry opportunities include increasing forest area, increasing the productivity of existing forest lands, reducing forest burning and deforestation, increasing biomass production and utilization, planting trees in urban environments, and increasing use of wood in durable products.

Estimation Methods

Carbon storage was estimated separately for several forest ecosystem components: trees, soil, forest floor, and understory vegetation. The definitions of these components were broad enough to include all sources of organic carbon:

Forest ComponentDefinition
TreesAll above- and below-ground portions of all live and dead trees including the merchantable stem; limbs,tops and cull sections; stump; foliage; bark and rootbark; and coarse tree roots (greater than 2 mm).

All organic carbon in mineral horizons to a depth of 1 meter excluding coarse tree roots.
Forest floorAll dead organic matter above the mineral Sod horizons including litter humus and coarse woody debris.
Understory vegetation

All live vegetation except that defined as live trees.

Carbon storage was estimated in a four-stage process corresponding to these four major forest ecosystem components. Separate estimates were generally made at the State level and for major forest types and plantation species in 8 geographic regions (fig. 1). The general approach was to estimate the volume of growing stock from forest inventories, to derive factors from biomass studies and other sources to convert the volume of growing stock to carbon, and to derive estimates for the other ecosystem components from models.

Several principal data sources were used to make estimates of carbon storage in forest trees. Statewide forest inventories, such as those conducted periodically by the USDA Forest Service, typically involve estimation of timber volume, growth, removals, mortality, and forest biomass for the purpose of analyzing current and prospective timber supplies. Data from these inventories were the basis for estimating carbon storage in forest trees. The data were supplemented by information from a special study to estimate the amount of carbon in tree roots and the conversion of volume to carbon (Koch 1989). Because regional forest inventories are based on a statistical sample designed to represent the broad range of forest conditions actually present, estimates of carbon storage in forest trees are representative of the true average values, subject to sampling errors, estimation errors, and errors in converting data from one reporting unit to another. Because of the complexity of making the estimates of tree carbon, the magnitude of the error has not been estimated, but it is likely quite small since the forest inventories used to derive the estimates have very small sampling errors over large areas.

Estimates of carbon storage in the soil, forest floor, and understory vegetation were developed through the use of models based on data from forest ecosystem studies. Although these studies include all of the key forest ecosystem components, they are valid only for the specific ecosystem studied. Uncertainty is introduced into the estimation process by assuming that the results of specific ecosystem studies are representative of regional or national averages without being part of a statistical sample that represents a large geographical area. Therefore, estimates of carbon storage in the soil, forest floor, and understory vegetation are subject to the following errors: bias from applying data from past studies that do not represent all forest conditions, modelling errors (imperfect assumptions), and errors in converting estimates from one reporting unit to another. No attempt has been made to estimate the magnitude of these errors.

Details of the modeling and estimation process for estimating carbon storage are presented in appendix 1.

Estimates of changes in carbon storage over time were limited to estimating carbon changes in live trees. One could assume that carbon changes in trees are correlated with changes in the whole forest ecosystem, since an increasing quantity of tree biomass is likely associated with an increase in soil and forest floor carbon because litter from trees is one of the main inputs of organic matter to the forest floor and soil (Raich and Nadelhoffer 1989; Vogt and others 1986). Carbon changes in live trees were estimated using the same procedures as in estimating carbon storage, but the starting estimates were volume growth, removals, and mortality rather than timber volume. The same conversion factors that were used to convert volume to carbon storage were used to convert volume growth, removals, and mortality to carbon accumulation, removals, and mortality.

Figure 1--Broad geographical regions used to report estimated carbon storage.

Estimates of carbon storage and accumulation for reserved forest land and other forest land (all forest land except that classified as timberland) were made in order to include the entire forest land base of the United States. Forest inventories are less comprehensive for these lands, and volume, growth, removals, and mortality estimates are not routinely aggregated to regional or national totals. Carbon storage in trees on reserved timberland was assumed to equal the average carbon storage on unreserved timberland for each State. Carbon storage in trees on other forest land was based on inventory statistics reported in a variety of Resource Bulletins issued by Forest Service experiment stations, using the conversion process outlined in appendix 1. Estimates for carbon storage in other forest ecosystem components were made using the procedures outlined in appendix 1.

[ Table of Contents | Next ]