Components of Energy Intensity. Since World War II the U.S. economy has been moving away from traditional “smokestack” industries toward more service-based or information-based enterprises. This has meant that over the second half of the 20th century economic growth was less tied to growth in energy demand than it was during the period of industrialization in the 19th and early 20th century. Other factors contributing to decreases in energy intensity include:

• Improvements in the energy efficiency of industrial equipment as new materials and methods improved performance in terms of energy inputs versus outputs
• Increased efficiency of transportation equipment as lighter materials and more efficient engines entered the marketplace
• Improvements in commercial and residential lighting, refrigeration, and heating and cooling equipment
• Developments in new electricity generating technologies, such as combined-cycle turbines.

Further reductions in energy intensity, which are projected to continue, will among other things promote deeper reductions in U.S. carbon intensity.

Components of the Carbon Intensity of Energy Supply. Changes in the carbon intensity of energy supply have been less dramatic than changes in energy intensity. There was a slow but steady decline from 1980 until about the mid-1990s, after which it has remained relatively unchanged. The primary reason for the decline has been the development of nuclear power, which is carbon-free and therefore weights the fuel mix toward lower carbon intensity. Other factors that can decrease the carbon intensity of the energy supply include:

• Development of new renewable resources, such as wind power, for electricity generation
• Substitution of natural gas for coal and oil in power generation
• Transportation fuels with a higher biogenic component, such as ethanol.

Energy-Related Carbon Dioxide Emissions in Manufacturing

Manufacturing is the single largest source of energy-related carbon dioxide emissions in the U.S. industrial sector, which also includes agriculture, forestry, fisheries, mining, and construction. The manufacturing subsector accounted for about 84 percent of energy-related carbon dioxide emissions and 90 percent of energy consumption in the industrial sector in 2002. The table below shows estimates of energy-related carbon dioxide emissions from manufacturing in 2002, based on end-use energy consumption statistics from EIA’s Manufacturing Energy Consumption Survey (MECS), which surveys more than 15,000 manufacturing plants every 4 years. The most recent MECS data available are from the 2002 survey. The table on "Carbon Diioxide Emissions from Manufacturing by Fuel" shows estimates of manufacturing emissions by fuel, based on statistics from the 1991, 1998, and 2002 surveys.

The 1991 MECS reported energy consumption (for fuel and nonfuel purposes) that yielded carbon dioxide emissions from the manufacturing subsector as a whole totaling 1,251.4 million metric tons. The corresponding estimate for 2002 is 1,401.2 million metric tons—an increase of 149.8 million metric tons, representing an average increase of 1.0 percent per year. Over the same interval, the demand for manufacturing products (as measured by gross output^a) increased by 1.3 percent per year. Therefore, the overall carbon intensity of U.S. manufacturing, measured as metric tons of carbon dioxide emitted per million chained 2000 dollars of gross output, was 420.4 in 1991 but had dropped to 358.4 by 2002, representing an average decrease of 1.4 percent per year.

The overall carbon intensity of the U.S. manufacturing subsector is the ratio of its total carbon dioxide emissions (C) to manufacturing output (Y), as measured by the gross output (in chained 2000 dollars). That ratio (C/Y) can be calculated as the product of the subsector’s aggregate carbon intensity of energy supply—carbon dioxide emissions (C) per unit of energy consumed (E)—and its energy intensity—energy consumed (E) per unit of gross output (Y). That is:

C/Y = (C/E) H (E/Y)   .

For the manufacturing subsector as a whole, energy intensity (the ratio E/Y) is a function primarily of the energy intensities of different production groups and their contributions to the total gross output mix in the subsector. The subsector’s carbon intensity of energy supply (the ratio C/E) is determined primarily by the mix of energy fuel inputs and the mix of fuel and nonfuel (sequestering) uses of the inputs. Thus, the overall carbon intensity of manufacturing (C/Y) is a combination of the energy intensity of manufacturing gross output and the carbon intensity of the energy consumed to meet manufacturing energy demand.^b

The manufacturing E/Y ratio fell by 1.2 percent per year from 1991 to 2002; however, the reduction was largely the result of a structural shift (i.e., a change in relative market shares in the subsector). The energy intensity for the “other manufacturing” category declined by 1.6 percent per year, and at the same time its gross output grew by 3.2 percent per year, from $2.0 trillion in 1991 to $2.9 trillion in 2002 (in chained 2000 dollars), as newer, less energy-intensive industries accounted for an increasing share of manufacturing activity. In 1991 the four most energy-intensive industries (petroleum, chemicals, primary metals, and paper) accounted for 29.0 percent of total manufacturing gross output, but by 2002 their share had fallen to 23.9 percent. For three of the six manufacturing categories, energy intensity increased from 1991 to 2002 (petroleum by 0.4 percent per year, chemicals 1.5 percent, and nonmetallic minerals 0.1 percent). For paper, primary metals, and other manufacturing, energy intensity declined by 0.4 percent, 0.9 percent, and 1.6 percent per year, respectively.

The mix and quantity of energy fuels consumed by manufacturers (for both fuel and nonfuel uses) affect the subsector’s aggregate carbon intensity of energy supply. Overall, manufacturing industries had C/E ratios equal to 50.9 million metric tons carbon dioxide equivalent per quadrillion Btu in 1991 and 49.5 million metric tons carbon dioxide equivalent per quadrillion Btu in 2002; however, the carbon dioxide factors of the various industries differed markedly.

The petroleum and chemical industries both transform some energy products into products that sequester carbon, such as petrochemical feedstocks, asphalt, and plastics. Because of that use, both the petroleum and chemical industries have lower aggregate C/E ratios than the manufacturing average (45.3 and 43.2 million metric tons carbon dioxide equivalent per quadrillion Btu for the petroleum industry and 45.8 and 41.5 for the chemicals industry in 1991 and 2002, respectively).

The paper industry makes extensive use of wood byproducts as an energy source. Carbon dioxide emissions from wood consumption are considered to be zero, because the carbon that is emitted has been sequestered recently, and the regrowing of trees will again sequester an equivalent amount of carbon dioxide. Consequently, the paper industry has a relatively low C/E ratio, at 37.4 and 36.5 million metric tons carbon dioxide equivalent per quadrillion Btu in 1991 and 2002, respectively. In contrast, the primary metals industry, which uses large amounts of coal and other carbon-intensive fuels (e.g., electricity), has a high C/E ratio: 68.2 in 1991 and 68.7 in 2002.

Between 1998 and 2002, manufacturing industries had decreases in carbon dioxide emissions associated with their use of electricity (20.9 million metric tons) and natural gas (49.0 million metric tons). Even so, electricity use continues to account for the largest share of manufacturers’ energy-related carbon dioxide emissions: 37.8 percent (561.6 million metric tons) in 1998 and 38.6 percent (540.7 million metric tons) in 2002.

As a result of the above changes in energy intensity, in combination with the structural shift in the subsector, the overall manufacturing energy intensity (E/Y) declined by 1.2 percent per year from 1991 to 2002. When the influence of the structural shift is removed, however, decomposition analysis suggests that the aggregate energy intensity of the manufacturing sector is virtually unchanged.^c

^aConsists of sales, or receipts, and other operating income, plus commodity taxes and changes in inventories.

^bThe ratios presented here are estimated as aggregations of several manufacturing industries. Specifically, 22 manufacturing industry groups were aggregated into 6 groups for calculations of industry-specified E/Y and C/Y ratios. Therefore, quantifying influences on the change in overall carbon intensity is valuable to extent that these groupings represent changes in the U.S. manufacturing sector. It should be noted, however, that these ratios are based on survey data that are subject to sampling errors and other uncertainties.

^cThere are several approaches that, based on index number theory, can be used to decompose aggregate values. The values reported here are based on a discrete approximation of the Divisia integral index.

Ethanol and Greenhouse Gas Emissions

Because the carbon in biogenic material is part of the natural carbon cycle, using ethanol in place of gasoline has the potential to reduce greenhouse gas (GHG) emissions. The nature of the impacts could vary greatly, however, depending on the fuels, feedstocks, and processes used to produce the ethanol.

For this report, Argonne National Laboratory produced a life-cycle (“well to wheels”) comparison of GHG emissions for conventional motor gasoline and ethanol per gallon of fuel consumed, on a Btu equivalent basis. As shown in the figure below, there is substantial variation in the potential GHG savings for ethanol as compared with motor gasoline. The analysis, based on the near future (2010), compared an outcome based on the current industry average with what could be technically feasible for 2010. Key inputs for the analysis included: corn yield (bushels per acre); nitrogen fertilizer application rate (pounds per acre); nitrogen fertilizer production (Btu per pound); corn ethanol conversion rate (gallons per bushel); ethanol conversion process (Btu per gallon); total energy use (Btu per gallon); and coproduct energy credits (Btu per gallon).

figure data

Among the simulations performed, the smallest savings in GHG emissions when ethanol is used are 7 percent (for an ethanol plant using coal as the input fuel, corn as the energy crop feedstock, and a dry mill production process). The comparison based on the projected industry average for ethanol production in 2010^b shows savings in GHG emissions of about 18 percent. When a dry mill process is assumed with 100 percent natural gas as the input fuel and corn as the energy crop, the potential savings are about 38 percent.

The higher GHG emissions savings are estimated to occur when the input fuel is renewable and the energy crop is cellulosic rather than corn. With a biogas fuel input and switchgrass as the energy crop, the potential savings are estimated at about 87 percent; with corn stover as the energy crop, the savings are estimated to be more than 90 percent.

The intent of this analysis was not to weigh in on a particular position with regard to the feasibility of the scenarios examined. It is clear, however, that input assumptions are significant in any examination of the potential for GHG emissions savings from the use of ethanol as a transportation fuel. The analysis examined neither economic feasibility nor issues of scale-up to meet a targeted market share, and the future technologies and crop inputs assumed in the analysis remain untested on a national scale.

^aThe industry average in 2010 is projected to be 30 percent wet and 70 percent dry process, with an input fuel mix of 72 percent natural gas, 18 percent coal, and 10 percent electricity for a dry mill plant and a fuel mix of 60 percent natural gas and 40 percent coal for a wet mill plant.

Methane Emissions from Abandoned Coal Mines

Thousands of coal mines in the United States have been closed and abandoned during the past 100 years. The U.S. Department of Labor’s Mine Safety and Health Administration (MSHA) estimates that since 1980 more than 7,500 coal mines have been abandoned, and many continue to emit methane. In an April 2004 report,^a the U.S. Environmental Protection Agency (EPA) estimated that methane emissions from abandoned coal mines ranged between 3.0 MMTCO₂e and 4.6 MMTCO₂e in 1990, and between 4.6 MMTCO₂e and 6.4 MMTCO₂e in 2002. More recently, the EPA estimated methane emissions of 7.1 MMTCO₂e from abandoned underground coal mines in 2004, up from 6.0 MMTCO₂e in 1990 but down from a peak of 8.7 MMTCO₂e in 2000 due to a decline in the number of gassy mines being abandoned.^b Because access to abandoned mines is limited and a systematic measurement program at those sites would be time-intensive and costly, the EPA estimates rely on actual emissions data from when the mines were operating and assume a decline function in emissions based on mine and coal-seam characteristics.

The most important variable in determining the amount of methane emissions from an abandoned mine is its post-mining status—whether the mine has been sealed, flooded, or continues to be vented. Sealed and flooded mines have much lower rates of emissions than vented mines. Another variable deemed important is whether the mine was gassy (emitting more than 100,000 cubic feet per day) when it was operating. Gassy mines are estimated to emit 98 percent of all methane emissions from operating coal mines, and the EPA assumes that abandoned mines which had been gassy when operating represent a similarly predominant portion of emissions from abandoned mines. The EPA’s 2004 study thus focuses on abandoned mines that had been gassy prior to closure. Of the 438 gassy mines abandoned since 1972, the EPA has data on the status (i.e., whether the mines were sealed, flooded, or continue to be vented) of 263 or 60 percent of them. From those data, the EPA calculates percentage shares of emissions by status, then assumes that the same shares apply to mines for which it does not have data.

For abandoned mines that have been vented, the EPA derives an emissions decline curve based on three primary factors: adsorption isotherms by coal basin, coal permeability estimates, and estimates of pressure at abandonment. For mines that are flooded, the EPA assumes a decline curve equation based on measurements taken from eight abandoned mines in two basins. The EPA treats sealed mines similarly to those vented, adjusting the initial emissions rate and length of time for emissions to dissipate, given the slower release rate from sealed vents.

The EPA sought to calibrate its estimation methodology to field measurements, but restricted access precluded measurement at all but seven mines. Although results from those mines suggested the general accuracy of the estimation method, the methodology had not yet been validated when this report was being prepared. EIA expects the new method to be included in the 2006 IPCC Guidelines for National Greenhouse Gas Inventories and anticipates the inclusion of estimates for this source in future annual reports.

^aU.S. Environmental Protection Agency, Coalbed Methane Outreach Program, Methane Emissions from Abandoned Coal Mines in the United States: Emission Inventory Methodology and 1990-2002 Emissions Estimates (Washington, DC, April 2004), web site www.epa.gov/ cmop/pdf/amm_final_report.pdf.

^bU.S. Environmental Protection Agency, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2004, EPA-430-R-06-002 (Washington, DC, April 2006), web site http://yosemite.epa.gov/oar/globalwarming.nsf/content/ResourceCenterPublicationsGHG EmissionsUSEmissionsInventory2006.html.

Methane Emissions from Industrial Wastewater Treatment

When wastewater containing large amounts of organic material is treated through anaerobic decomposition, methane is emitted. The best estimate of those emissions would be based on a systematic measurement of all point sources; however, the number and diversity of U.S. industrial wastewater sources make such an approach unaffordable and impractical. As an alternative, methane emissions from industrial wastewater treatment are calculated by the following equation:

M = P ´ O ´ COD ´ A ´ EF   ,

where M = methane emissions, P = product output, O = wastewater outflow per unit of product output, COD = organic loading in outflow, A = percentage of outflow treated anaerobically, and EF = emissions factor for anaerobically treated outflow.

The Intergovernmental Panel on Climate Change (IPCC), in its Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories,^a provides default data for wastewater generation and COD on an industry-specific basis. The default data, often based on one or two literature sources, are assumed to have an uncertainty range of minus 50 percent to plus 100 percent (although no justification for the range is provided). The IPCC also provides a single default factor of 0.25 kilograms methane per kilogram of COD, premised on a general approximation of the theoretical maximum for this emission factor, and identifies an uncertainty of plus or minus 30 percent for this estimate.

There are currently no specific U.S. data that could be used to improve on the IPCC defaults, and the uncertainties make it impossible for the Energy Information Administration (EIA) to provide a reliable estimate of emissions from this source. It can, however, be noted that—depending on the extent to which industrial wastewater from such industries as meat and poultry processing, pulp and paper manufacturing, and vegetable, fruit, and juice processing (which is likely to have a high content of organic material) is treated anaerobically—excluding the resulting methane emissions from the U.S. emissions total will tend to produce an underestimate. The U.S. Environmental Protection Agency estimates that U.S. methane emissions from industrial wastewater treatment could be as high as 16.9 MMTCO₂e in 2004.^b EIA anticipates that additional methodological guidance and data will be forthcoming in the 2006 IPCC Guidelines for National Greenhouse Gas Inventories that will support emissions estimates for this source in subsequent annual reports.

^aIntergovernmental Panel On Climate Change, Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories (Montreal, Canada, May 2000), web site www.ipcc-nggip.iges.or.jp/public/gp/english/.

^bU.S. Environmental Protection Agency, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2004, EPA-430-R-06-002 (Washington, DC, April 2006), web site http://yosemite.epa.gov/oar/globalwarming.nsf/content/ResourceCenterPublicationsGHG EmissionsUSEmissionsInventory2006.html.

Revisions in EPA Emissions Estimation Methodology

The primary source for the emissions estimates presented in this chapter is data obtained from the U.S. Environmental Protection Agency (EPA), Office of Air and Radiation. The Office of Air and Radiation also prepares an annual inventory of greenhouse gas emissions, which is published pursuant to U.S. commitments under the United Nations Framework Convention on Climate Change (UNFCCC). The UNFCCC encourages parties to revise methods regularly and to recalculate emissions affected by the revisions. The data supporting the EPA inventory, including the emissions estimates for 2005, incorporate a number of revisions to the data and estimation methodologies used for hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and sulfur hexafluoride (SF₆) in its most recent emissions inventory.^a Those changes are reflected in the EPA’s historical emissions estimates, as described below:

• Electricity Transmission and Distribution. Changes in the calculations of emissions from electricity transmission and distribution resulted in an average annual increase in estimated SF₆ emissions from electric power systems of 0.1 to 0.6 million metric tons carbon dioxide equivalent (MMTCO₂e) for the 1990-2003 period.
• Magnesium Production and Processing. Emissions estimates from the EPA have been revised to reflect more accurate data on emission factors for sand casting activities and updated historical secondary production data from the U.S. Geological Survey (USGS). The changes resulted in a decrease in estimated SF₆ emissions from magnesium production and processing of 0.1 MMTCO₂e (5 percent) for 2002.
• Substitution of Ozone-Depleting Substances. The EPA has updated assumptions for its Vintaging Model pertaining to trends in chemical substitutions, market size and growth rates, and amounts used. The changes resulted in an average annual net decrease in estimated HFC and PFC emissions of 2.0 MMTCO₂e (3 percent) for the 1990-2003 period.
• Aluminum Production. The EPA has revised smelter-specific emissions factors and aluminum production levels to reflect recently reported data on smelter operating parameters. The changes resulted in an average annual increase of less than 0.5 MMTCO₂e (0.4 percent) for the 1990-2003 period.

^aU.S. Environmental Protection Agency, Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2004, EPA-430-R-06-002 (Washington, DC, April 2006), web site http://yosemite.epa.gov/oar/globalwarming.nsf/content/ResourceCenterPublicationsGHG EmissionsUSEmissionsInventory2006.html.

The EPA Vintaging Model: Estimation Methods and Uncertainty

The U.S. Environmental Protection Agency (EPA) uses a detailed Vintaging Model for equipment and products containing ozone-depleting substances (ODS) and ODS substitutes to estimate actual versus potential emissions of various ODS substitutes, including hydrofluorocarbons (HFCs) and perfluorocarbons (PFCs). The model estimates the quantities of equipment and products sold each year that contain ODS and ODS substitutes, and the amounts of chemicals required for their manufacture and/or maintenance over time. Emissions from more than 50 different end uses are estimated by applying annual leak rates and release profiles, which account for the lag in emissions from equipment as it leaks over time.

For most products (refrigerators, air conditioners, fire extinguishers, etc.), emissions calculations are split into two categories: emissions during equipment lifetime, which arise from annual leakage and service losses plus emissions from manufacture; and disposal emissions, which occur when the equipment is discarded. By aggregating the data over different end uses, the model produces estimates of annual use and emissions of each compound.^a

The EPA periodically attempts to improve the model and reduce the uncertainty of emissions estimates by using more accurate data from emitting industries. The level of detail incorporated in the EPA Vintaging Model is higher than that of the default methodology used by the Intergovernmental Panel on Climate Change, although there still is some uncertainty about some of the model inputs, such as equipment characteristics and sales figures, and end-use emissions profiles.

^aU.S. Environmental Protection Agency, Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2003, EPA-430-R-05-003 (Washington, DC, April 2005), web site http://yosemite.epa.gov/oar/globalwarming.nsf/content/ResourceCenterPublicationsGHG EmissionsUSEmissionsInventory2005.html.

Methane Emissions From Vegetation: New Findings

For several decades, the conventional view of climate scientists has been that terrestrial vegetation produces methane only under anaerobic conditions, through the action of anaerobic bacteria on organic matter in rice paddies and wetlands; however, recent studies by Frank Keppler of the Max Planck Institute for Nuclear Physics, along with researchers at other European institutions, indicate that vegetation can emit methane under aerobic conditions.^a This discovery has prompted a lively discussion both among scientists and in the press, given the importance of methane as a greenhouse gas of concern (methane is second in importance only to carbon dioxide, with 23 times its global warming potential) and the role of afforestation and reforestation in national and international efforts to mitigate emissions of greenhouse gases.

Keppler and his fellow investigators measured methane emissions from vegetation exposed to methane-free air, using intact plants in plexiglass chambers and freshly collected tree and grass leaves in sealed vials. To rule out a possible role of anaerobic bacteria, gamma radiation was used to kill any such bacteria in the samples. First-estimate extrapolations from laboratory measurements to a global scale, based on net primary productivity, produced an estimate of annual methane emissions from terrestrial vegetation between 62 and 236 MMT. In comparison, previous estimates of global methane emissions from all known sources (wetlands, animals, rice cultivation, biomass burning, and fossil fuel production) have totaled approximately 600 MMT per year. Given the considerable uncertainty associated with methane emissions estimates for those sources, Keppler suggested that up to 50 MMT of methane from vegetation may already be included (erroneously) in some of the previous estimates attributed to sources such as wetlands or rice paddies.

Keppler’s estimate of methane emissions from vegetation, extrapolated to the global scale, was then used to estimate annual emissions from various sources, based on type of vegetation, ecosystem, and region. For tropical forests, the researchers estimated mean annual methane emissions of 78.2 MMT per year, and for tropical savannas and grasslands they estimated mean annual emissions of 29.2 MMT per year. The results are in conformity with recent field and satellite measurements, which indicate annual methane emissions in upland zones of the Brazilian Amazon in the range of 4 to 38 MMT^b and annual emissions in the northern part of the Guyana shield of Venezuela in the range of 30 to 60 MMT for the entire savanna.^c

Other recent research has compared emissions measured using the SCIAMACHY instrument of the European Space Agency’s ENVISAT satellite against modeled data, finding significant discrepancies over tropical forests.^d The measured values were consistently higher than the modeled values, with a discrepancy of 30 MMT methane per year. Adding the discrepancy to the modeled value of 45 MMT per year yields an estimate of 75 MMT for annual methane emissions from tropical forests, as compared with Keppler’s estimate of 78.2 MMT.

The extrapolation of emission rates from laboratory experiments to global emission rates in the study by Keppler was based on the rate of growth of terrestrial vegetation, or “net primary productivity.” Other researchers, however, have argued that the use of net primary productivity leads to an overestimate of methane emissions from vegetation, and that estimation methods based on leaf mass and photosynthesis would be more appropriate.^e Those methods yield global estimates of 10 to 60 MMT methane per year. A similar range, 0 to 46 MMT, has been estimated by researchers using a “top-down” approach based on ice core records.^f

When limits are placed on emissions of greenhouse gases—either through binding international commitments such as the Kyoto Protocol or through voluntary programs, such as those being instituted at the State and Federal levels in the United States—knowing how much methane is emitted from various sources will be of obvious importance. In particular, if tree planting is proposed as a means of mitigating greenhouse gas emissions through carbon sequestration, the possibility that the same trees could be a major source of methane emissions would affect calculations of their potential benefits, depending on the balance between carbon dioxide sequestration and methane emissions.

Writing in the same issue of Nature that contains the original article by Keppler et al., David Lowe hinted at some of the policy implications, suggesting that trees in reforestation projects might increase greenhouse warming through methane emissions.^g On the other hand, researchers in Australia have reported that, based on their own extrapolations of methane emissions from vegetation at the global level, the likely increase in methane emissions as a result of tree planting would offset only a small part (estimated at 0.1 to 1.1 percent) of the benefit resulting from increased carbon sequestration.^h

^aF. Keppler, J.T.G. Hamilton, M. Braß, and T. Röckmann, “Methane Emissions From Terrestrial Plants Under Aerobic Conditions,” Nature, Vol. 439 (January 2006), pp. 187-191, web site http://moab.colorado.edu/BRG/Methane.pdf.

^bJ.B. do Carmo, M. Keller, J.D. Dias, P.B. de Camargo, and P. Crill, “A Source of Methane From Upland Forests in the Brazilian Amazon,” Geophysical Research Letters, Vol. 33, No. 4 (2006), pp. 1-4, web site www.agu.org/pubs/crossref/2006/2005GL025436.shtml.

^cP. J. Crutzen, E. Sanhueza, and C. A. M. Brenninkmeijer, “Methane Production From Mixed Tropical Savanna and Forest Vegetation in Venezuela,” Atmospheric Chemistry and Physics Discussions, Vol. 6 (2006), pp. 3093-3097, web site www.copernicus.org/EGU/acp/ acpd/6/3093.

^dC. Frankenberg, J.F. Meirink, M. van Weele, U. Platt, and T. Wagner, “Assessing Methane Emissions From Global Space-Borne Observations,” Science, Vol. 308, No. 5724 (2005), pp. 1010-1014, web site www.sciencemag.org/cgi/content/abstract/1106644.

^eM.U.F. Kirschbaum, D. Bruhn, D.M. Etheridge, J.R. Evans, G.D. Farquhar, R.M. Gifford, K.I. Paul, and A.J. Winters, “A Comment on the Quantitative Significance of Aerobic Methane Release by Plants,” Functional Plant Biology, Vol. 33, No. 6 (2006), pp. 521-530, web site www.publish.csiro.au/nid/102/paper/FP06051.htm.

^fD.F. Ferretti, J.B. Miller, J.W.C. White, K.R. Lassey, D.C. Lowe, and D.M. Etheridge, “Stable Isotopes Provide Revised Global Limits of Aerobic Methane Emissions from Plants,” Atmospheric Chemistry and Physics Discussions, Vol. 6 (2006), pp. 5867-5875, web site www.copernicus.org/EGU/acp/acpd/6/5867.

^gD. Lowe, “Global Change: A Green Source of Surprise,” Nature, Vol. 439 (2006), pp. 148-149, web site www.nature.com/nature/ journal/v439/n7073/edsumm/e060112-09.html.

^hM.U.F. Kirschbaum et al., “A Comment on the Quantitative Significance of Aerobic Methane Release by Plants,” Functional Plant Biology, Vol. 33, No. 6 (2006), pp. 521-530, web site www.publish.csiro.au/nid/102/paper/FP06051.htm.

Global Forest Resources Assessment 2005

The Food and Agriculture Organization of the United Nations (FAO) is the main intergovernmental source of data on global forests. FAO’s global forest assessments date back to 1948, with the most recent assessment—Global Forest Resources Assessment 2005—published in 2005. The FAO’s 2000 assessment^a was the first to include a uniform definition of forests for all regions of the world—that is, areas with at least 10 percent of canopy cover (excluding stands of trees primarily used for agricultural production). The current report estimates the world’s forested area in 2005 at approximately 4 billion hectares or 30 percent of the Earth’s total land area.

The 2005 report points out that, while the rate of deforestation (mainly through conversion to cropland) continues at the high rate of about 13 million hectares per year, average net annual losses of forest have fallen from 8.9 million hectares per year over the period 1990-2000 to 7.3 million hectares per year over the period 2000-2005. Forest planting, landscape restoration, and the natural expansion of forests have significantly reduced the net loss of forest area.^b

The largest reported net loss of forest land from 2000 to 2005 was in South America, with 4.3 million hectares lost per year, followed by Africa, which lost 4.0 million hectares annually. North and Central America and Oceania each had a net loss of about 350,000 hectares per year, while Asia reported a net gain of 1 million hectares per year from 2000 to 2005, primarily from large-scale afforestation in China. Forest areas in Europe continued to expand, although at a slower rate than in the 1990s.

^aFood and Agriculture Organization of the United Nations, Global Forest Resources Assessment 2000, “Executive Summary,” web  site www.fao.org/DOCREP/004/Y1997E/y1997e05.htm#bm05.

^bFood and Agriculture Organization of the United Nations, Global Forest Resources Assessment 2005, “Executive Summary,” web site www.fao.org/docrep/008/a0400e/a0400e00.htm.