6. Land-Use Issues

Overview

Land-use and forestry issues are important to national and global inventories of greenhouse gases in three ways:

• Vegetation can “sequester” or remove carbon dioxide from the atmosphere and store it for potentially long periods in above- and below-ground biomass, as well as in soils. Soils, trees, crops, and other vegetation may make significant contributions to reducing net greenhouse gas emissions by serving as carbon “sinks.”
• Harvested wood put into wood products, or eventually into landfills, can potentially sequester carbon dioxide from the atmosphere for decades before the carbon stored in the wood products decays and is released to the atmosphere.
• Human-induced land-use changes and forest management practices can alter the quantities of atmospheric and terrestrial carbon stocks, as well as the natural carbon flux among biomass, soils, and the atmosphere.¹²⁰

Land-use issues are of particular interest to U.S. policymakers, because U.S. forests and soils annually sequester large amounts of carbon dioxide. Much of the forest land in the United States was originally cleared for agriculture, lumber, or fuel in the hundred years before 1920. Since then, however, much of the agricultural and pasture land has reverted to forest land, increasing its ability to sequester atmospheric carbon dioxide.

The amount of carbon being sequestered annually is uncertain, in part because of an absence of data and difficulties in measuring carbon sequestration. Moreover, in addition to technical uncertainties, there are also policy and accounting questions about the aspects of the carbon cycle that should be included in national inventories as anthropogenic emissions and removals.

The 1996 revised guidelines for national emissions inventories, published in 1997 by the Intergovernmental Panel on Climate Change (IPCC), include methods for calculating carbon sequestration and net carbon dioxide flux to the atmosphere resulting from land-use changes and land-use activities, such as forestry.¹²¹ The U.S. Environmental Protection Agency (EPA) estimates annual U.S. carbon sequestration in 2003, based on data generated by the U.S. Department of Agriculture (USDA), at 828.0 million metric tons carbon dioxide equivalent (MMTCO₂e), a decline of approximately 21 percent from the 1,042.1 MMTCO₂e sequestered in 1990¹²² (Table 33). Land use, land-use change, and forestry practices offset approximately 16.9 percent of total U.S. anthropogenic carbon dioxide emissions in 1990 and 11.9 percent in 2003.¹²³

New IPCC Good Practice Guidance for Land Use, Land-Use Change, and Forestry

The estimates of carbon sequestration in this chapter involve several categorical and methodological changes from previous years, which have been implemented by the EPA in response to new land-use guidelines issued by the IPCC in 2003. The IPCC Good Practice Guidance for Land Use, Land-Use Change and Forestry¹²⁴ (LULUCF GPG) recommends reporting carbon stocks according to several land-use types and conversions—for example forest land remaining forest land, non-forest land becoming forest, and forest land becoming non-forest land. These categories of “land-use type remaining land-use type” and “land-use type becoming land-use type” are a new convention adopted in the LULUCF GPG.

Currently, there are no consistent datasets for the entire United States that would allow the results for forest lands, croplands, and settlements to be disaggregated in this fashion. Thus, the net changes in carbon sequestered are aggregated to one category for each land-use type: forests, croplands, and settlements. For example, in the case of forest lands, net changes in forest carbon stocks for “forest land remaining forest land” encompass all forest-related land, including non-forest land converted to forest and forest land converted to non-forest land.

The EPA report, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2003, upon which this chapter is based, is the first to follow the LULUCF GPG. The LULUCF GPG had to be consistent with the 1996 IPCC Guidelines, because there was an existing agreement among Parties to the United Nations Framework Convention on Climate Change (UNFCCC) to use the latter.¹²⁵ The IPCC defines this consistency as: (1) traceability of specific sources or sinks addressed by the LULUCF GPG to categories in the IPCC Guidelines; (2) use of the same functional forms (or their equivalent) for the equations in the LULUCF GPG that are used in the IPCC Guidelines; and (3) facilitation for the correction of errors or deficiencies in the IPCC Guidelines by application of the LULUCF GPG.

In its most recent inventory, the EPA followed the LULUCF GPG, reporting fluxes according to changes within and conversions between forest lands, croplands, settlements, grasslands, and wetlands. Due to the lack of consistent datasets, the EPA limited its estimates of greenhouse gas flux to: (1) forest land remaining forest land, (2) croplands remaining croplands, and (3) settlements remaining settlements. This new categorization provides additional sources of information on nitrous oxide emissions by major land-use type. Further, the EPA cautioned that other land-use and land-use change activities cause fluxes of greenhouse gases other than carbon dioxide that are not accounted for, including methane from managed forest soils and artificially flooded lands.

Significant differences between the sequestration values in this report and those in EIA’s previous reports on U.S. greenhouse gas emissions are broadly attributable to changes that have been made in order to conform with the LULUCF GPG, as well as a variety of differences in calculation methods. Definitional changes include the following:

• The forest soil pool is now termed “soil organic carbon.”
• The forest floor is now termed “litter.”
• Previously, the tree pool included the mass of standing dead trees; now, the mass of standing dead trees, together with down dead wood, is categorized as “dead wood.”
• Previously, the remainder of the tree pool, live biomass, and understory pool was divided into above- and below-ground portions; now, the above-ground tree and understory pools are summed into an above-ground biomass pool, and the below-ground portions are summed into a below-ground biomass pool.

Important differences in calculation methods include the following:

• The USDA’s State Soil Geographic (STATSGO) database¹²⁶ and its relationship with data from the USDA’s National Forest Inventory and Analysis (FIA) survey¹²⁷ are now interpreted differently for the estimation of soil organic carbon. For this report, soil organic carbon in the conterminous United States was calculated using the STATSGO database, and data gaps were filled with representative values for similar soils. Links to regions and forest types were developed with the assistance of the USDA’s FIA Geospatial Data Services, by overlaying FIA forest inventory plots with existing soil carbon maps.
• The newer USDA Forestry Inventory and Analysis Database (FIADB) datasets were considered for non-soil forest carbon estimates, along with the USDA Resources Planning Act (RPA) data.¹²⁸ For last year’s estimates, only RPA data were used when needed.
• Final values for all carbon pools were extrapolated from the last carbon stock change values calculated from the FIA survey data, because it was not possible to model final stocks in a manner consistent with available inventory data. Previously, the estimates of final carbon stocks were based on model results.

Forest Land Remaining Forest Land

The values for forest carbon dioxide fluxes reported in this chapter are based on estimates of changes in forest carbon stocks. The components analyzed are above-ground biomass, below-ground biomass, dead wood, litter, soil organic carbon, harvested wood products in use, and harvested wood products in landfills. The estimated carbon dioxide flux from each of these components was calculated using FIADB data and methodologies consistent with the LULUCF GPG and the Revised 1996 IPCC Guidelines. Nitrous oxide emissions from fertilized forest soils were calculated by using a default methodology consistent with the LULUCF GPG. Pine trees, being the dominant species planted for timber in the southeastern United States, were taken as representative of fertilized forests in the country, and the average reported fertilization rate of 150 pounds of nitrogen per acre was multiplied by the area of pine receiving fertilizer.

Croplands Remaining Croplands

Estimates of carbon fluxes from croplands include changes in agricultural soil carbon stocks on both croplands and grazing lands, because datasets necessary to separate the two were not available. Changes in agricultural soil carbon stocks result from the use and management of cropland and grazing land and emissions of carbon dioxide from the application of crushed limestone and dolomite. The estimation methods used for this report are consistent with the Revised 1996 IPCC Guidelines and the LULUCF GPG.

Settlements Remaining Settlements

Fluxes from settled lands include methane from landfilled yard trimmings and food scraps, carbon from urban trees, and nitrous oxide from fertilized soils. Changes in carbon flux were estimated by analyzing life-cycle emissions and sinks associated with waste management. Stock changes in urban trees were estimated on the basis of field measurements and data on national urban tree cover, using a methodology consistent with the LULUCF GPG to estimate carbon flux. Nitrous oxide emissions from nitrogen applied to turf grass were estimated as 10 percent of all synthetic fertilizer used in the United States.

Land-Use Change and Forestry Carbon Sequestration

The EPA’s estimates for carbon sequestration from land-use change and forestry in 2003 include three main sink categories: (1) changes in forest carbon stocks for forest land remaining forest land (752.7 MMTCO₂e or 91 percent of the total); (2) changes in agricultural soil carbon stocks for cropland remaining cropland (6.6 MMTCO₂e or 0.8 percent of the total); and (3) changes in settlements remaining settlements (68.7 MMTCO₂e or 8.3 percent, including 58.7 MMTCO₂e from urban trees and 10.1 MMTCO₂e from landfilled yard trimmings and food scraps).¹²⁹ For a discussion of worldwide trends in forest and croplands see discussion on "Millennium Ecosystem Assessment: Forest and Cultivated System".

Forest Land Remaining Forest Land: Changes in Forest Carbon Stocks

In the United States, the most significant pressures on the amount of carbon sequestered through forest lands are land management activities and the continuing effects of past changes in land use. These activities directly affect carbon flux by shifting the amount of carbon accumulated in forest ecosystems.¹³⁰ Land management activities affect both the stocks of carbon that can be stored in land-based carbon sinks, such as forests and soils, and the fluxes of carbon between land-based sinks and the atmosphere.

The components or “pools” of forest carbon analyzed by the EPA for its most recent inventory include above-ground biomass, below-ground biomass, dead wood, litter, soil organic carbon, harvested wood products in use, and harvested wood products in landfills. As a result of natural biogeochemical processes occurring in forests, as well as anthropogenic activities, carbon is constantly cycling through these components and between the forest and the atmosphere. The net change in overall forest carbon may not always be equal to the net flux between forests and the atmosphere, because timber harvests may not necessarily result in an instant return of carbon to the atmosphere. Timber harvesting transfers carbon from one of the five “forest carbon pools” to one of the two “wood products carbon pools.” Once carbon is transferred to a product pool, it is emitted over time as carbon dioxide or methane as the product decays or is combusted. Emission rates vary significantly, depending on the type of product pool that houses the carbon.¹³¹

In the United States, enhanced forest management, regeneration of formerly cleared forest areas, and timber harvesting have resulted in net annual sequestration of carbon throughout the past decade. Since the 1920s, deforestation for agricultural purposes has become a nearly defunct practice. More recently, managed growth practices have become common in eastern forests, greatly increasing their biomass density over the past 50 years. In the 1970s and 1980s, federally sponsored tree planting and soil conservation programs were embraced. These programs led to the reforestation of formerly harvested lands, improvement in timber management activities, soil erosion abatement, and the conversion of cropland to forests. Forest harvests have also affected carbon sequestration. The majority of harvested timber in the United States is used in wood products. The bulk of the discarded wood products is landfilled, and thus large quantities of the harvested carbon are relocated to long-term storage pools rather than to the atmosphere. The size of this long-term storage pool has increased over the past century.¹³²

According to the EPA, carbon sequestration by U.S. forests totaled 753 MMTCO₂e in 2003 (Table 34). Between 1990 and 2003, U.S. forests accounted for an average annual net sequestration of 832 MMTCO₂e, resulting from domestic forest growth and increases in forested land area; however, there was a decrease of approximately 21 percent in annual sequestration over the same period.¹³³

The overall decline in forest carbon sequestration was driven by a 27-percent reduction in the level of sequestration in the forest carbon pool (739 MMTCO₂e in 1990 versus 537 MMTCO₂e in 2003). The reduction in the forest carbon pool sequestration rate can be attributed primarily to a 110-percent decline in the estimated level of sequestration in forest soils. Not only was there no forest soil sequestration in 2003 (as compared with 125 MMTCO₂e sequestered in 1990), but forest soils became a source with average annual emissions of 12.0 MMTCO₂e during the period 1999-2003. The EPA explains that the decrease in sequestration in this pool is derived from forest inventory data and is a direct consequence of changes in total forest area or changes in forest type.¹³⁴

The EPA points out that net forest growth and increasing forest area, particularly before 1997, contributed to rising sequestration. Since 1997, forest land area has remained relatively constant, and the increase in carbon density (per area) has resulted in net forest carbon sequestration. National estimates of forest land are obtained by summing State surveys for the conterminous United States. Because the State surveys are not completed each year, interpolation between data points is used to provide estimates for years without surveys.

Overall annual sequestration levels in harvested wood carbon stocks increased slightly between 1990 and 2003. The trend in net sequestration amounts has been generally upward, from 210 MMTCO₂e in 1990 to 216 MMTCO₂e in 2003 (Table 34). Annual sequestration levels in landfilled wood declined from 162 MMTCO₂e in 1990 to 155 MMTCO₂e in 2003, but that decline was offset by an increase in carbon sequestration in harvested wood products, from 48 MMTCO₂e in 1990 to 60 MMTCO₂e in 2003.

The EPA has estimated carbon stocks in wood products in use and in landfills from 1910 onward, based on USDA Forest Service historical data and analyses using the North American Pulp and Paper (NAPAP) model,¹³⁵ the Timber Assessment Market Model (TAMM),¹³⁶ and the Aggregate Timberland Assessment System (ATLAS) model.¹³⁷ Carbon decay in harvested wood was analyzed by the EPA for the period 1910 through 2003, using data on annual wood and paper production. The analysis included changes in carbon stocks in wood products, changes in carbon in landfills, and the amount of carbon emitted to the atmosphere (carbon dioxide and methane) both with and without energy recovery. The EPA also followed the “production approach”; that is, carbon stored in imported wood products was not counted, but carbon stored in exports was counted, including logs processed in other countries¹³⁸ (see discussion on "Accounting for Harvested Wood Products in Future Greenhouse Gas Inventories").

Croplands Remaining Croplands: Changes in Agricultural Soil Carbon Stocks

The amount of organic carbon in soils depends on the balance between the addition of organic material and the loss of carbon through decomposition. The quantity and quality of organic matter within soils, as well as decomposition rates, are determined by the interaction of climate, soil properties, and land use. Agricultural practices—including clearing, drainage, tillage, planting, grazing, crop residue management, fertilization, and flooding—can alter organic matter inputs and decomposition, causing a net flux of carbon to or from soils.

The IPCC methodology, which is used by the EPA to estimate the net flux from agricultural soils (Table 35), is divided into three categories of land-use and land-management activities: (1) agricultural land use and land management activities on mineral soils;¹³⁹ (2) agricultural land-use and land-management activities on organic soils;¹⁴⁰ and (3) liming of soils. Of the three activities, the use and management of mineral soils is estimated to be the most significant contributor to total carbon sequestration from 1990 through 2003. Sequestration in mineral soils in 2003 was estimated to be 51.7 MMTCO₂e, while emissions from organic soils and liming were estimated at 35.6 and 9.5 MMTCO₂e, respectively. Together, these three activities resulted in a net 6.6 MMTCO₂e sequestered through agricultural soils in 2003.¹⁴¹

Settlements Remaining Settlements

Changes in Urban Tree Carbon Stocks

Urban forests make up a considerable portion of the total tree canopy cover in the United States. Urban areas, which cover 3.5 percent of the continental United States, are estimated to contain about 3.8 billion trees, accounting for approximately 3 percent of total tree cover in the United States. The EPA’s carbon sequestration estimates for urban trees are derived from estimates by Nowak and Crane,¹⁴² based on data collected throughout the 1990s and applied to the entire time series in this report. Net carbon dioxide sequestration from urban trees is estimated at 58.7 MMTCO₂e sequestered annually from 1990 through 2003 (Table 33).¹⁴³

Changes in Landfilled Yard Trimming and Food Scrap Carbon Stocks

Carbon stored in landfilled yard trimmings can remain sequestered indefinitely. In the United States, yard trimmings (grass clippings, leaves, and branches) and food scraps make up a considerable portion of the municipal waste stream, and significant amounts of the yard trimmings and food scraps collected are discarded in landfills. Both the amount collected annually and the percentage that is landfilled have declined over the past decade. Net carbon dioxide sequestration from landfilled yard trimmings and food scraps has declined accordingly, from 26.0 MMTCO₂e in 1990 to 10.1 MMTCO₂e in 2003 (Table 36).

Since 1990, municipal policies limiting pickup and disposal have led to a 20-percent decrease in yard trimmings collected. In addition, composting of yard trimmings in municipal facilities has increased significantly, reducing the percentage of total yard trimmings placed in landfills from 72 percent in 1990 to 34 percent in 2003. In contrast, the percentage of food scraps disposed of in landfills has decreased only slightly, from 81 percent in 1990 to 77 percent in 2003. The EPA’s methodology for estimating carbon storage relies on a life-cycle analysis of greenhouse gas emissions and sinks associated with solid waste management.¹⁴⁴

Land Use and International Climate Change Negotiations

In past international negotiations on climate change, the United States and many other countries have maintained that the inclusion of LULUCF activities in a binding agreement that limits greenhouse gas emissions is of the utmost importance; however, issues of whether and how terrestrial carbon sequestration could be accepted for meeting various commitments and targets have remained subjects of complex and difficult international negotiations.

Many of the countries involved in climate change negotiations have agreed that implementation of LULUCF activities under an international climate change agreement may be complicated by a lack of clear definitions of “reforestation” and “forest.” Further, implementation may be hindered by the lack of effective accounting rules. According to research published by the Pew Center on Global Climate Change,¹⁴⁵ implementation of LULUCF provisions in an international climate change agreement raises many issues, such as:

• What is a direct human-induced activity?
• What is a forest and what is reforestation?
• How will the issues of uncertainty and verifiability be addressed?
• How will the issues of (non) permanence and leakage be addressed?
• Which activities beyond afforestation, reforestation and deforestation (ARD), if any, should be included, and what accounting rules should apply?
• Which carbon pools and which greenhouse gases should be considered?

Uncertainties related to data issues have also slowed international negotiations on climate change.

The Ninth Session of the Conference of the Parties to the UN Framework Convention on Climate Change (COP-9) was held in Milan, Italy, in December 2003. The parties agreed on some of the rules for carbon sequestration projects under the Clean Development Mechanism (CDM), but the issue of how to treat the non-permanence of carbon sinks projects remained unresolved. Delegates at COP-9 decided to limit the duration of credits generated from carbon sequestration projects and addressed the topics of additionality, leakage, uncertainties, and socioeconomic and environmental impacts.¹⁴⁶

A year later in Buenos Aires, Argentina, delegates at the Tenth Conference of the Parties (COP-10) did address the issue of small-scale afforestation and reforestation project activities under the CDM. The following decisions were made at COP-10:¹⁴⁷

• Adopt simplified modalities and procedures for small-scale afforestation and reforestation project activities in the first commitment period.
• Limit the designation of small-scale afforestation and reforestation projects to those with net anthropogenic greenhouse gas removals by sinks that are less than 8,000 metric tons carbon dioxide equivalent per year. For projects that result in greenhouse gas removals of more than this quantity, the excess would be ineligible for temporary or long-term certified emissions reductions.
• Exclude funds obtained through small-scale project activities from the share of proceeds to be used to assist developing countries particularly vulnerable to the adverse impacts of climate change. Such countries shall be entitled to a reduced level of the non-reimbursable fee for requesting registration and a reduced rate of the proceeds to cover administrative expenses of the CDM.

Land-Use Data Issues

The EPA’s most recent inventory report discusses the uncertainty inherent in the methodology used to estimate forest carbon stocks.¹⁴⁸ The estimates of forest carbon in live biomass, dead wood, and litter are based on USDA forest survey data for the conterminous United States, because no survey data are available for Alaska, Hawaii, and the U.S. Territories. The survey data are statistical samples designed to represent vast areas of land. The USDA mandates that the survey data be accurate to within 3 percent, at a confidence level of 67 percent.¹⁴⁹ An analysis of this methodology for the southeastern United States showed that the uncertainty resulted from sampling errors and not from the regression equations used to calculate tree volume (and thus carbon) from survey statistics such as tree height and diameter. The standard errors of 1 to 2 percent for volumes of growing stock in individual States are insignificant; however those for changes in the volumes of growing stock are much higher, ranging from 12 percent to as much as 139 percent.¹⁵⁰

Additional uncertainty is associated with the estimates of carbon stocks in other carbon pools, which are based on extrapolations of the relationships among variables in site-specific studies to all forest land. Such extrapolation is needed in the absence of survey data on other carbon pools.¹⁵¹ The extrapolations bring in uncertainty from modeling errors and conversions between different reporting units. The effect of land-use change and forest management activities (such as harvest) on soil stocks is another large source of uncertainty, with little consensus in the literature.

Notes and Sources

Tables 33-36