2
National Inventories of Greenhouse Gas Emissions

All countries that are party to the United Nations Framework Convention on Climate Change (UNFCCC) are required to provide national inventories of emissions and removals of greenhouse gases due to human activities. These inventories form the basis for monitoring the progress of individual countries in reducing emissions and for assessing the collective effort of countries to mitigate climate change. The inventories provide self-reported estimates of selected anthropogenic greenhouse gases for four sectors: energy; industrial processes and product use; agriculture, forestry, and other land use (AFOLU); and waste. Countries prepare the estimates using methods developed by the Intergovernmental Panel on Climate Change (IPCC) and approved by the UNFCCC. The methods generally involve multiplying national data on an emissions-generating activity, such as cement production, by an emission factor that specifies greenhouse gas emissions per unit of activity. This chapter describes current practices for developing greenhouse gas inventories, summarizes the accuracy of estimates for the gases and activities responsible for most of the emissions, and identifies improvements that would facilitate monitoring of emissions for an international climate treaty.

DEVELOPING AND REPORTING NATIONAL INVENTORIES

UNFCCC National Inventory Reporting and Review

UNFCCC reporting and review requirements for national inventories differ for developed (Annex I) and developing (non-Annex I) countries. As a result, the scope and quality of national inventories vary greatly. Developed countries annually report calendar-year estimates for all sources and sinks of the six greenhouse gases specified by the UNFCCC (carbon dioxide [CO2], methane [CH4], nitrous oxide [N2O], sulfur hexafluoride [SF6], perfluorocarbons [PFCs], and hydrofluorocarbons [HFCs]) going back to 1990. The estimates are broken down by sector and into categories within a sector (e.g., aluminum production within the industrial sector). The national inventories, along with detailed documentation of the methods and data sources used to calculate emissions and removals, are submitted electronically in a standard format to facilitate data analysis and comparison. Similar regulations govern reporting of chlorofluorocarbons (CFCs), but these fall under agreements other than the UNFCCC.

The national inventories of developed countries are subject to international review by teams of greenhouse gas inventory experts. These reviews do not attempt to reconstruct the inventory or verify estimates with independent data, but rather assess whether correct methods and appropriate data sources were used to produce the inventory. Statistical analysis of reported data is also performed to identify inconsistencies within a report or with previously submitted reports. In addition, data are analyzed across countries to determine a range of expected levels of emissions per unit of output or activity (implied emission factors) and to identify deviations from these values. Where possible, data submitted by countries are compared with data compiled by international organizations. For example, national statistics used to estimate energy emissions are



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 21
2 National Inventories of Greenhouse Gas Emissions A ll countries that are party to the United and developing (non-Annex I) countries. As a result, Nations Framework Convention on Climate the scope and quality of national inventories vary C hange (UNFCCC) are required to pro - greatly. Developed countries annually report calendar- vide national inventories of emissions and removals year estimates for all sources and sinks of the six of greenhouse gases due to human activities. These greenhouse gases specified by the UNFCCC (carbon inventories form the basis for monitoring the progress dioxide [CO2], methane [CH4], nitrous oxide [N2O], of individual countries in reducing emissions and for sulfur hexafluoride [SF6], perfluorocarbons [PFCs], assessing the collective effort of countries to mitigate and hydrofluorocarbons [HFCs]) going back to 1990. climate change. The inventories provide self-reported The estimates are broken down by sector and into estimates of selected anthropogenic greenhouse gases categories within a sector (e.g., aluminum production f or four sectors: energy; industrial processes and within the industrial sector). The national inventories, product use; agriculture, forestry, and other land use along with detailed documentation of the methods and (AFOLU); and waste. Countries prepare the estimates data sources used to calculate emissions and remov- using methods developed by the Intergovernmental als, are submitted electronically in a standard format Panel on Climate Change (IPCC) and approved by the to facilitate data analysis and comparison. Similar UNFCCC. The methods generally involve multiply- regulations govern reporting of chlorofluorocarbons ing national data on an emissions-generating activity, (CFCs), but these fall under agreements other than such as cement production, by an emission factor that the UNFCCC. specifies greenhouse gas emissions per unit of activity. The national inventories of developed countries are This chapter describes current practices for developing subject to international review by teams of greenhouse greenhouse gas inventories, summarizes the accuracy gas inventory experts. These reviews do not attempt of estimates for the gases and activities responsible for to reconstruct the inventory or verify estimates with most of the emissions, and identifies improvements independent data, but rather assess whether correct that would facilitate monitoring of emissions for an methods and appropriate data sources were used to international climate treaty. produce the inventory. Statistical analysis of reported data is also performed to identify inconsistencies within a report or with previously submitted reports. In addi- DEVELOPING AND REPORTING NATIONAL tion, data are analyzed across countries to determine INVENTORIES a range of expected levels of emissions per unit of output or activity (implied emission factors) and to UNFCCC National Inventory Reporting and identify deviations from these values. Where possible, Review data submitted by countries are compared with data UNFCCC reporting and review requirements for compiled by international organizations. For example, national inventories differ for developed (Annex I) national statistics used to estimate energy emissions are 

OCR for page 21
 VERIFYING GREENHOUSE GAS EMISSIONS compared to International Energy Agency (IEA) data, ent, complete, accurate, consistent over time, and com- and statistics used to estimate agriculture emissions are parable across countries. Because different countries compared to UN Food and Agriculture Organization have different capacities to produce inventories, the (FAO) data. If anomalies are identified by this analysis, guidelines lay out tiers of methods (typically three) for the review teams dig deeper into that country’s methods each emissions source, with higher tiers (Tier 3 is nor- and data. Unless otherwise indicated, the inventory mally the highest) being more complex and/or resource methods discussed in this chapter pertain to developed intensive than lower tiers. The higher-tier methods countries. usually incorporate country-specific conditions, data, Reporting requirements are much less rigorous and emission factors and are thus considered more accu- for developing countries. Emission inventories are rate than the lower-tier methods. For example, the Tier reported only periodically in conjunction with a broader 1 method for calculating CO2 emissions from stationary national report of climate change programs and activi- combustion uses default emission factors for each fuel ties. There is no set frequency for these national reports type, whereas the Tier 2 method requires each country and their submission often depends on the provision to develop and use country-specific emission factors for of international funding. As a result, most developing each fuel type (see detailed guidance in Gómez et al., countries have submitted only one national inventory to 2006). The Tier 3 method uses emission factors that date. Reporting of only CO2, CH4, and N2O is required are not only country-specific, but also differentiated and only at the sector level, not for categories within by technology and operating conditions. The choice each sector. Developing countries are not required to of method used for a particular source in a particular provide emissions trends over time or to document country depends on (1) the importance of that source methods and data sources, and their inventories are to the level and trend of emissions in that country and not reviewed. (2) the resources available to prepare the inventory. Countries are encouraged to use country-specific data and emission factors to the extent possible. However, IPCC Methodologies they are not expected to use higher-tier methods if The IPCC’s National Greenhouse Gas Inventory doing so would jeopardize their ability to estimate other Program is responsible for developing methods for important emissions sources. The scope of the effort to creating national inventories of greenhouse gas emis- prepare the U.S. inventory is described in Box 2.1. sions. The IPCC guidelines describe how to estimate national emissions of CO2, CH4, N2O, SF6, PFCs, and Implications for Monitoring and Verification HFCs from anthropogenic sources and sinks using national statistics (activity data) and activity-based Although multiple greenhouse gases are emit- emission factors for the four sectors. Guidance is also ted from multiple activities in multiple sectors, the provided on data sources, data collection methods, monitoring and verification problem is comparatively quantification of uncertainties, management of inven- simple because only a few activities and greenhouse tories, quality assurance and control, documentation, gases are responsible for the large majority of emis- and data archiving. The guidelines have evolved over sions. Table 2.1 summarizes emissions by sector for time to include more emissions sources and to improve Annex I countries as a group, and Figure 2.1 compares and standardize the methodologies. The first edition of emissions across sectors for Annex I and non-Annex I the IPCC guidelines was completed and approved in countries. The most important simplifying message is 1994; the most recent (2006) edition has not yet been that well over 90 percent of global greenhouse gas emis- endorsed by the UNFCCC and is thus not yet used sions are in the energy and AFOLU sectors, making for reporting purposes. However, the 2006 guidelines these sectors an obvious focus for monitoring. Energy are expected to be adopted as the basis for reporting alone is responsible for almost 90 percent of total net national inventories beginning in 2015. greenhouse gas emissions from Annex I countries and The IPCC methodologies are intended to yield more than 40 percent of net emissions from develop- national greenhouse gas inventories that are transpar- ing countries. In both groups, CO2 from fossil-fuel

OCR for page 21
 NATIONAL INVENTORIES OF GREENHOUSE GAS EMISSIONS of emissions in developing countries than in developed BOX 2.1 Preparation of the U.S. Greenhouse countries because of tropical deforestation and lower Gas Emissions Inventory levels of industrial development. Agriculture contrib- utes 15 percent of emissions in the developing world Preparation of a national greenhouse gas emissions inven- tory is a major undertaking, involving many people and agencies and deforestation contributes 35 percent. The figures in a typical Annex I country. The current version of the IPCC are lower for Annex I countries, with agriculture emis- guidelines for inventory preparation is a five-volume set of books sions of less than 10 percent of the total and forests with several hundred pages of formulas, supporting data, sample representing a net sink for CO2 (note the negative calculations, reporting formats, and technical references. In the sign for the Annex I forestry emission in Table 2.1). United States, the Environmental Protection Agency is ultimately responsible for planning and preparing the national inventory, but In contrast, emissions from the industrial and waste it relies on many other agencies and individuals for data, scientific sectors in both Annex I and non-Annex I countries are information, analysis, and review. Contributors to the inventory an order of magnitude lower than emissions from the include the Department of Energy, U.S. Department of Agriculture energy and AFOLU sectors. (U.S. Forest Service, Agricultural Research Service, National Agri- cultural Statistics Service), U.S. Geological Survey, Department of A second simplifying message is that fewer than Transportation (including the Bureau of Transportation Statistics), one quarter of the 185 countries in the World Resources Department of Commerce, Federal Aviation Administration, De- Institute database were responsible for more than 80 partment of Defense, and Colorado State University. Many private percent of global emissions in 2000 (Figure 2.2). This sector contractors and academic and research institutions from concentration of sources has increased over the last all sectors of the economy also contribute data and analysis to the inventory. The majority of emissions in the U.S. inventory are decade with the surge in fossil-fuel emissions from a estimated using the highest-tier methods, and its preparation has few rapidly growing, developing economies. Because stimulated some original research to improve basic understanding developed countries and the largest emitting develop- and emissions coefficients. The U.S. inventory also undergoes ing countries are likely to be the focus of mitigation separate expert and public review before publication and submis- efforts under a new climate treaty, their emissions will sion to the UNFCCC. be of particular importance for monitoring. The many activities and gases in the two most important sectors—energy and AFOLU—are cata- loged in the next section. The most important gases in combustion comprises the bulk (90 percent) of energy these sectors are CO2, CH4, and N2O. Corresponding emissions. material for the industrial and waste sectors, including a The AFOLU sector is responsible for approxi- discussion of CO2, HFCs, N2O, and CH4, is contained mately 30 percent of global greenhouse gas emissions in Appendix A. (Figure 2.1) but represents a much greater component TABLE 2.1 Sectoral Emissions from Annex I Countries for 2007 Fraction of Sectoral Emissions by Gas Fraction of Sector Total Emissionsa CO2 CH4 N 2O HFCs PFCs SF6 Total 1.00 0.81 0.12 0.06 0.01 0.00 0.00 Energy 0.91 0.94 0.05 0.01 0.00 0.00 0.00 Industrial processes 0.08 0.69 0.00 0.07 0.18 0.03 0.03 Solvent and other product use 0.00 0.48 0.00 0.52 0.00 0.00 0.00 Agriculture 0.08 0.00 0.47 0.53 0.00 0.00 0.00 Land use, land-use change, and forestry (LULUCF)b –0.09 1.04 –0.03 –0.01 0.00 0.00 0.00 Waste 0.03 0.07 0.90 0.06 0.00 0.00 0.00 Other 0.00 0.95 0.02 0.02 0.01 0.00 0.00 aThe disparate gases were added assuming their global warming potential with a 100-year time horizon. Each number in the table represents the sum of values from the national reports of all Annex I countries. Each value has its own confidence level, which varies with the gas and the national and sectoral source of emissions. No composite uncertainty calculations have been attempted. bFor Annex I countries, this sector overall is a net sink due to sequestration of CO . In national inventories, this is shown as negative emissions. Because 2 emissions of CH4 and N2O in this sector are positive, the fraction of LULUCF that they represent is shown as a negative. SOURCE: .

OCR for page 21
 VERIFYING GREENHOUSE GAS EMISSIONS sions. The CO2 emissions from fossil-fuel combustion Emissions ( MT CO2e) 45,0 00 accounted for 80 percent of total greenhouse gas emis- 40,0 00 non-Annex I 35,0 00 sions (on a CO2-equivalent basis) in the United States Annex I 30,0 00 25,00 0 in 2006 (EPA, 2008). Other emissions from the energy 20,0 00 sector include CO2 from the non-energy use of fossil 15,000 10,000 fuels (e.g., as petrochemicals, solvents, lubricants), CH4 5,000 0 from fuel production and transport systems (e.g., coal Total Waste Energy Forestry Industrial Land-Use Change & Processes Agriculture mines, gas pipelines), and N2O from transportation systems. Carbon Dioxide. Most estimates of CO2 emissions IPCC Sector from energy systems are based on self-reporting of FIGURE 2.1  Greenhouse  gas  emissions  by  sector  in  2000  fuel consumption. Emissions are estimated from the Fig 2.1.eps for Annex I and non-Annex I countries; 2000 is the most recent  year for which comprehensive data on the greenhouse gases are  amount of fuel burned, the carbon content of the fuel, available. SOURCE: Data compiled from the Climate Analysis  and the efficiency of combustion (i.e., the fraction of Indicators Tool, Version 6.0, World Resources Institute, . the point of combustion as, for example, carbon mon- oxide or ash). The fraction left unoxidized is small in modern combustion systems, and the IPCC now sug- SECTOR-BASED REPORTING gests using the default assumption that 100 percent of the carbon in a fuel is fully oxidized (IPCC, 2006). A Energy challenge is that the amount of fuel burned is gener- In most Annex I countries, CO2 from energy ally measured in mass or volume units and the carbon use dominates anthropogenic greenhouse gas emis- content is not generally measured. There is a good cor- Cumulative Contribution to Global Emissions 7,000 0.9 0.8 National Emissions ( MT CO2e) 6,000 0.7 5,000 0.6 4,000 0.5 Emissions 0.4 3,000 Percent of Global emissions 0.3 2,000 0.2 1,000 0.1 0 0 Ve Afr lic Cn n Ua B r ia N i ne ia a C ys i a ( S ar na M da So e n al n st ) re an o de z i l a U lia ne i c a o tio an r i c Au u t h hb di . R Ir a M pa Ko My ic el s er ni am ut p u Fe a In h i a ra ne a In ra ex pe e zu ig an Ja a o kr ro Am do Eu of em an es ,D at si go us St R on d te C ni U Country FIGURE 2.2  National greenhouse gas emissions from all IPCC sectors of the top 20 emitters in 2000. Note that the 27 countries  in the European Union are treated as one. SOURCE: Data compiled from the Climate Analysis Indicators Tool, Version 6.0, World  Resources Institute. Fig 2.2.eps

OCR for page 21
 NATIONAL INVENTORIES OF GREENHOUSE GAS EMISSIONS relation between the carbon content of fuels and their feedstock production, for example, would be included energy content, so conversion from mass or volume in the AFOLU inventory. This can lead to incom- units to energy content permits the carbon content plete accounting of emissions if burning of biomass to be estimated, but with greater uncertainty (see, for fuels occurs in a reporting country but the decrease example, IPCC, 2006; Marland et al., 2007; EPA, 2008, in biospheric carbon stocks occurs in a nonreporting particularly Annex 2). country (e.g., see Marland and Schlamadinger, 1997; Direct data on fuel consumption are not always Searchinger et al., 2009). available. In some cases, CO2 emissions can be esti- Fossil fuels are used for a variety of applications. mated using models that represent the major fuel- For example, asphalt is used for roads and roofs, oils consuming processes, such as the amount of CO2 are used as lubricants and solvents, and a variety of emitted per vehicle-mile of travel. At national scales, it petrochemicals are used in plastics and fibers. In the is sometimes appropriate to estimate consumption from United States, more than 6 percent of the carbon in the amount of fuel produced and the net of imports and fossil fuels used in 2006 ended up in non-fuel applica- exports. However, fuel may not be consumed where tions (EPA, 2008), and this number can be higher in and when it is made available. For example, fuel sold to countries with large petrochemical industries. Some of ships, planes, or even road vehicles may be carried out the carbon used in these applications will be oxidized of the country and burned elsewhere.1 Data on fuel pro- over time, often at slow rates (e.g., products in land- duction or consumption will generally record when and fills). The IPCC provides guidance on estimating the where the fuel passed some point in the distribution lifetime, fate, and greenhouse gas emissions from these chain. Models based on parameters such as vehicle- products. However, emissions are diffuse in both time miles traveled or flight patterns can approximate when and space, and only approximate values can be assigned and where consumption takes place, but they are not to product fates and lifetimes and to changes in stocks widely used. Moreover, it is difficult to capture the over time. demonstrated tendency for travelers to purchase fuel where it is cheapest (e.g., Banfi et al., 2003). Methane and Nitrous Oxide. Emissions of CH4 and Combustion of biomass fuels is reported to the N2O depend on fuel characteristics, combustion tech- UNFCCC, but the associated CO2 emissions are not. nology and maintenance, pollution-control equipment, In theory, biomass removes CO2 from the atmosphere system leakage, and prevailing current practice. Emis- when growing and releases CO2 back to the atmosphere sions occur as a result of both combustion processes when burned, so a sustainably managed system should and leaks from production and transport facilities such have no net CO2 emissions. Fossil fuels used in the as coal mines and gas pipelines. Emissions vary widely, production, harvest, and transport of the biomass are and emissions estimates are generally based on broad counted in the fossil-fuel emission inventories. When indicators and aggregate emission factors. Uncertainty sustainably produced ethanol is combined with gasoline is greater than for CO2 emissions from combustion, as a fuel, the emissions counted are only those from but the sources are generally small and emissions esti- combustion of the gasoline fraction. Similarly, if waste mates for Annex I countries have been improving with is used as a fuel and the waste includes both biomass increasing interest in mitigation. and fossil-fuel-derived materials, only the CO2 emit- ted from the fossil-fuel fraction is generally counted. Agriculture, Forestry, and Other Land Use Any net emissions of CO2 from unsustainable use of biomass fuels should be captured as a decrease in the The AFOLU sector is responsible for about 30 amount of biomass in the AFOLU sector. Thus, CO2 percent of global anthropogenic emissions, predomi- emitted from the conversion of forested land to biofuels nantly in the form of CO2 emissions from land-use change (dominated by tropical deforestation) and CH4 and N2O emissions from farming and animal hus- 1There is no internationally agreed-upon method for allocating bandry (Smith et al., 2007). Emissions vary over space fuels used in international commerce, so emissions from bunker and time, depending on how the land is used and on fuels are currently reported, but not attributed.

OCR for page 21
 VERIFYING GREENHOUSE GAS EMISSIONS the local climate, topography, and soil and vegetation Dominant sources and sinks for the AFOLU sec- properties. Currently, greenhouse gas emissions from tor include CO2 emissions and removals from woody land-use change are highest in tropical areas of South biomass and soils; CO2, CH4, and N2O emissions from America, Southeast Asia, and to lesser extent, Africa fires; CH4 emissions from livestock and manure man- (Houghton, 2003; Achard et al., 2004; DeFries et al., agement and from rice cultivation; and N2O emissions 2007).2 In contrast, extratropical regions in the north- from soil management (especially nitrogen fertiliza- ern hemisphere have recently produced carbon sinks tion) and manure management. Other emission sources (Figure 2.3 bottom) because of net forest regrowth that may be locally important but are less significant from earlier harvesting or encroachment on aban- globally include soil liming, organic soil cultivation, and doned agricultural land and other processes, such as non-rice managed wetlands or flooded lands. Long- sequestration of carbon in landfills and water reservoirs lived harvested wood products are a potential CO2 and woody encroachment into pastures. These sinks sink, although the average lifetime of wood products are thought to absorb roughly 5-20 percent of global is relatively short (20 years) and UNFCCC accounting fossil-fuel emissions (CCSP, 2007). Methane and N2O rules for them have not yet been agreed upon. emissions are highest in regions with intensive, high- input agriculture (Figure 2.3 top), predominantly in Carbon Dioxide. Net CO2 release or uptake in managed Europe, North America, China, and India. Globally, lands is estimated from changes over time in five carbon agricultural emissions of CH4 and N2O have increased stocks: above- and belowground biomass, dead organic by about 1 percent per year since 1990 (Smith et al., matter (coarse woody debris and litter), and soil organic 2007). matter. Declines in ecosystem organic carbon stocks Estimating emissions from land-use activities is represent net CO2 emissions to the atmosphere, and challenging due to the distributed nature of emission increases in stocks represent net CO2 removals from sources, the multitude of different processes involved, the atmosphere. Carbon stock changes are calculated and the fact that much of the carbon is belowground for six major land categories (forestland, cropland, in soils. Many countries simply use aggregate land-use grassland, settlements, wetlands, and other land) and statistics (i.e., activity data) and default emission factors for land-use change between categories (IPCC, 2006). (Tier 1 methods), but a number of Annex I countries Predominant carbon stock changes (and hence CO2 use more advanced methods, employing a mixture of fluxes) are associated with changes in forestland area, remote sensing, ground measurements and surveys, and land use (deforestation and afforestation), and the process-based models. Ground surveys, where avail- relative carbon balance determined by growth versus able, provide complementary information on land use harvest and natural mortality and decay. Studies of CO2 and management, and ground-based measurements of emitted from land-use change have adopted different carbon stock change are used to calibrate models and variables to keep track of changes in land and biomass, provide estimates of uncertainty. Models can also be making it difficult to compare results directly with calibrated with CO2 and non-CO2 flux measurements IPCC reporting (Ito et al., 2008). and related biological data. Remote sensing provides spatial information on land cover and surface charac- Agricultural Methane. The largest sources of agricul- teristics (e.g., vegetation type, species, forest age; EPA, tural CH4 are enteric fermentation in the digestive sys- 2008) and is sometimes used to directly estimate emis- tem of ruminants and other livestock, livestock waste, sions associated with land-use change (e.g., Richards, and emissions from paddy (flooded) rice. The simplest 2001). methods (Tier 1) for estimating enteric fermentation are based on emissions per animal for each major species. Countries that have additional information 2 As a result of high tropical deforestation rates, the Conference on livestock demographics (e.g., sex and age classes), of Parties to the UNFCCC has been considering offering eco- feed quality, and intake rates can incorporate energy nomic incentives to avoid greenhouse gas emissions by changing balance calculations to make more accurate estimates forest management practices. See details of the COP 11 and COP 13 meetings at .

OCR for page 21
 NATIONAL INVENTORIES OF GREENHOUSE GAS EMISSIONS 500 Reported Agriculture Emissions (2006) 400 Tg CO2-eq/ yr 300 200 100 0 Canada Poland Ireland US France Australia Germany Spain NZ Italy Ukraine Japan Romania Netherlands Turkey Belarus Greece Belgium Russia UK Reported LULUCF Emissions (2006)s Fig 2.3 top.ep 400 Germany Romania Bulgaria Sweden Ukraine Belarus Norway Finland Austria Poland France 200 Turkey Japan Spain Italy US NZ 0 Canada Australia Russia Tg CO2-eq/ yr –200 –400 –600 –800 –1000 FIGURE 2.3  Annex I countries with the highest reported emissions or removals of greenhouse gases from agricultural sources (top)  and from forestry and other land-use sources (bottom) in 2006. Greenhouse gases are reported as CO2 equivalents. Negative emis- sions  represent  removals  of  CO2  from  the  atmosphere.  SOURCE:  Data  compiled  from  national  greenhouse  gas  inventory  reports;  Fig 2.3 bottom.eps . statistics kept in most Annex I and some non-Annex I management (i.e., crop residues, manure; Wassmann et countries provide accurate estimates of the number of al., 1996; van der Gon, 1999), and cultivar type. animals by major species, but not always detailed infor- mation on animal nutrition. Emission factors for CH4 Agricultural Nitrous Oxide. Approximately 3-5 per- from livestock waste are estimated by using manure cent of nitrogen added through fertilization and, to production rates from livestock statistics, along with a lesser extent, fossil-fuel combustion is subsequently information on how manure is stored and managed emitted as N2O (Crutzen et al., 2008; Galloway et (e.g., waste lagoons, dry lot, composted). The primary al., 2008) through microbial processes in soils. Both controlling factors for emissions from flooded rice are direct emissions (i.e., those occurring at the location water management (length and periodicity of flooding; of nitrogen addition) and indirect emissions (i.e., those Li et al., 2002; Frolking et al., 2004), organic matter stemming from nitrogen that is volatilized or leached

OCR for page 21
 VERIFYING GREENHOUSE GAS EMISSIONS and redeposited elsewhere, often in nonagricultural be most severe for a treaty that includes commitments soils, and then emitted as N2O) are considered in from developing countries, many of which lack the AFOLU inventories. Most countries use aggregate infrastructure and capacity to collect, analyze, and activity data on the amount of nitrogen added to soils manage inventory data consistently. In addition, our (from synthetic fertilizer, manure, biological nitrogen ability to compare inventories with data derived from fixation from leguminous plants, and crop residues) other methods is limited by incomplete coverage of and emission factors to estimate fluxes, although the greenhouse gases, the time lag between the occurrence United States and some other Annex I countries are of emissions and the completion of inventories, and the employing more disaggregated simulation model-based lack of spatial information. approaches to reduce uncertainty. At the country level, synthetic fertilizer nitrogen production and use are well Infrequent, Incomplete, and Unreliable Reporting quantified from industry production and trade statis- for Non-Annex I Countries tics (approximately 25 percent of global production is internationally traded) maintained by trade organiza- The establishment of rigorous reporting and review tions (e.g., International Plant Nutrition Institute) guidelines and requirements through the UNFCCC and FAO. Other nitrogen addition sources such as has led to the creation and steady improvement of biological nitrogen fixation and manure additions are national greenhouse gas inventories in Annex I coun- more uncertain. Quantities of other nitrogen losses (as tries over the past decade (Breidenich and Bodansky, ammonia, nitrogen oxides, and leached nitrogen) and 2009). There are still high levels of uncertainty in the their contribution to indirect N2O emissions are also estimates for biogenic sources, gases other than CO2, highly uncertain. and some sectors and activities with low total emis- sions (see below). However, uncertainties in national Emissions from Fires. Under IPCC guidelines, CO2 totals are relatively low for Annex I countries due to fluxes from forest fires on managed lands are incorpo- their well-developed statistical systems and capacity rated in estimates of ecosystem carbon stock changes. to use higher-tier methods. The fact that most of their Emissions of CH4, N2O, and greenhouse gas precur- emissions are from fossil-fuel combustion, which has sors (e.g., carbon monoxide, nitrogen oxides, volatile low uncertainty, also reduces the overall uncertainty of organic compounds) from incomplete combustion are these inventories. Current uncertainties in annual emis- inventoried separately for forests, grasslands, and crop- sions estimates for Annex I countries are of comparable lands as a function of the area burned, pre-fire carbon magnitude to emissions reductions commitments, and stocks, and fire seasonality. The areal extent of fires this suggests that multiyear trends could be verified if can be estimated from satellite imagery (Giglio et al., the remaining uncertainty is considered properly ( Jonas 2006, 2009), although many short-duration fires and and Nilsson, 2007; Swart et al., 2007). smaller (<500 m) individual fires are difficult to detect In contrast, national inventories of many develop- via satellite (Al-Saadi et al., 2008). ing countries generally have greater uncertainty and are not sufficiently rigorous to enable monitoring of emissions. The low quality of national inventories in LIMITATIONS OF NATIONAL INVENTORIES developing countries largely reflects a lack of financial, FOR MONITORING technical, and institutional capacity. Funding to prepare National greenhouse gas inventory reporting cur- national inventories is provided by the Global Envi- ronment Facility,3 but it is sporadic and insufficient to rently has a number of limitations for a new climate treaty. These include poor reporting by developing countries, uncertainties in reported data (which are 3The Global Environment Facility is a partnership among 178 countries, international institutions, nongovernmental organiza- especially high for some sources and some greenhouse tions, and private companies that provides financial assistance to gases), and lack of data to independently verify the help countries meet their obligations under international agree- activity data or the condition-specific emission factors ments and conventions, including the UNFCCC. See .

OCR for page 21
 NATIONAL INVENTORIES OF GREENHOUSE GAS EMISSIONS enable consistent collection and processing of activity ments (see Chapter 4), and proxy measures that have data or to maintain institutional capacity for inventory some correlation with emissions (e.g., trade statistics; preparation over time. For example, even when inter- see below). national donors provide assistance in data collection and processing, the questionnaires on energy produc- Uncertainty in Fossil-Fuel CO2 Emissions. UNFCCC tion and consumption that are distributed annually inventories in Annex I countries produce reasonably by the United Nations and IEA statistical offices are accurate estimates for the most important emissions returned incomplete or not at all by many countries. category: fossil-fuel CO2. Uncertainties for Tier 1 Finally, land-use changes and agriculture frequently methods are estimated to be on the order of ±5 per- comprise a substantial source of emissions in develop- cent, and some countries believe they can estimate ing countries, and reliable data in these sectors are often emissions with even lower uncertainties using higher- not available. tier methods (IPCC, 2006). For example, Rypdal and Winiwarter (2001) suggested that the 2σ uncertainty in annual fossil-fuel CO2 emissions for countries with Uncertainty in Self-Reported Data “well-developed energy statistics and inventories” may The uncertainty in estimates of greenhouse gas be as low as 2-4 percent. The uncertainty in estimates emissions from self-reported data depends on the of fossil-fuel CO2 emissions from the United States uncertainty in both activity data and methods used has been estimated to be –1 percent to 6 percent (95 to calculate the inventory. Uncertainty in activity data percent confidence level; EPA, 2008). Analysis of the depends largely on a nation’s commitment to data sum of Annex I reported emissions as well as some collection and processing but also on its measure- independent estimates and inverse modeling results found a 1σ uncertainty of 6 percent for fossil-fuel CO2 ment capabilities. Uncertainty in methods depends on knowledge of parameters such as the heating value and (Prather et al., 2009). carbon content of fuels used nationally, which is used to In contrast, uncertainty in CO2 emissions from derive emission factors, and on knowledge of biochemi- developing countries is considered to be significantly cal processes, such as denitrification in agricultural higher. For countries with “less well-developed energy soils. In general, estimates of greenhouse gas emissions data systems,” the uncertainty may be on the order of ±10 percent (IPCC, 2006). The 2σ uncertainty in from fossil-fuel consumption have less uncertainty than emissions from biogenic processes, such as land-use annual estimates of fossil-fuel CO2 emissions from change. Uncertainties tend to be lower for emissions China may be as high as 15-20 percent (Gregg et al., trends than for emissions values for a given year. 2008). Table 2.2 shows a representative range of uncer- Uncertainty in CO2 emissions from fossil-fuel tainties for the various sources and gases as calculated consumption arises largely from uncertainty in activ- and reported by Annex I countries in their national ity data. A comparison of annual CO2 emission values inventories. It also provides estimates of uncertainties calculated from national energy data reported to the for the emissions estimates, where available, and appor- United Nations and IEA by different analysts showed tions them into contributions from uncertain activity significant differences for individual countries, but no levels and from uncertain emission factors. Estimates systematic bias and similar global totals (Marland et of uncertainty, including those in Table 2.2, have tradi- al., 1999). The mean difference was on the order of 3 tionally been made through expert judgment about the percent for 19 Western European countries and 7 per- quality of the data used in the calculations. Indications cent for 52 African countries, with differences ranging of uncertainty can also be derived by comparing (1) from 0.9 percent for the United States to 10 percent estimates from different sources at a specified time, (2) for India to 52 percent for North Korea, suggesting estimates from a single source over time, (3) estimates that the quality of activity data is a greater concern for made by different methods, and (4) estimates with developing countries. In a separate analysis, Gregg et al. model predictions, remotely sensed data (Marland et (2008) found that revisions in energy data by the China al., 2009; see also Chapter 3), atmospheric measure- National Bureau of Statistics, which became available

OCR for page 21
0 VERIFYING GREENHOUSE GAS EMISSIONS TABLE 2.2 Magnitude and Uncertainty Associated with Emission Factors, Activity Data, and Annual Emissions Estimates for the Principal Emission Source Categories in Annex I Countries Range of Uncertainty for 7 Annex I Countriesb Anthropogenic Emissions Emission Source of All Annex I Countriesa Emission Factor Activity Data Emission Estimate 1 CO2—total anthropogenic CO2—largest sources Energy Fuel combustion 82.9% 1 1 1 Fugitive emissions from fuels 0.6% 1-5 1-2 1-5 Industrial processes Mineral products 2.3% 1 1 1-3 Metal production 2.0% 1 1 1-2 Chemical industry 0.6% 1-2 1 1-3 AFOLU Forestlands –9.5% 2-4 1-2 2-4 Croplands 2.3% 1-4 1-2 1-4 2-3 CH4—total anthropogenic CH4—largest sources Energy Fugitive emissions from fuels 4.6% 2-5 1-2 1-5 AFOLU Enteric fermentation 2.7% 1-3 1-3 1-3 Manure management 0.7% 1-4 1 1-5 Waste Solid waste disposal on land 2.1% 1-3 1 1-5 2-5 N2O—total anthropogenic N2O—largest sources Energy Fuel combustion 0.7% 2-5 1 1-5 Industrial processes Chemical industry 0.5% 3-4 1 2-4 AFOLU Agricultural soils 4.0% 2-5 1-2 2-5 Manure management 0.5% 2-4 1-3 2-5 2-5 HFCs, PFCs, and SF6—total anthropogenic HFCs—largest source Consumption of halocarbons 1.1% 1-5 1 1-5 PFCs—largest source Aluminum production 0.2% 1-2 1-3 1-3 SF6—largest source Use in electrical equipment 0.1% 1-3 1-3 2-4 98.4% Total % emissions covered NOTES: 1 = 100% (i.e., for the last category, we cannot be certain if the actual emissions value is a source or a sink). aReported2006 data are from the UNFCCC’s online greenhouse gas database (UNFCCC, 2008). bPercentagesfor the largest sources are based on the 2006 greenhouse emissions data reported by seven Annex I countries—Australia, Denmark, Germany, Greece, Poland, Portugal, and the United States—selected to represent a range of institutional capabilities for compiling inventories. Uncertainty ranges were derived from the uncertainty estimates reported by the Annex I countries as part of their 2008 greenhouse gas inventory submissions for 1990 through 2006 to the UNFCCC. Each Annex I party is required to quantitatively assess and report the uncertainty of its inventories in accordance with IPCC (2000) good practice guidance. The uncertainty estimates for Denmark, Germany, Greece, and Poland were prepared for the majority of emissions source categories using the Tier 1 method, while the uncertainty estimates for Australia, Portugal, and the United States were developed using the Tier 2 method (i.e., the Monte Carlo method). The uncertainties associated with total anthropogenic emissions of each gas were available only for Denmark, Greece, Poland, Portugal, and the United States. Combined uncertainties for the emissions source categories were calculated from the uncertainty estimates given for the subcategories for each country according to the following equation (IPCC, 2000): Uncertainty(combined) = √([Uncertainty(1) × Emissions(1)]2 + [Uncertainty(2) × Emis- sions(2)]2 + … + [Uncertainty(n) × Emissions (n)]2) / (Emissions(1) + Emissions(2) + … + Emissions(n)). Values are reported at the 95% confidence level. See Australian Government Department of Climate Change (2008); Danish National Environmental Re - search Institute (2008); EPA (2008, Annex 7); German Federal Environmental Agency (2008); Greek Ministry for the Environment (2008); Poland National Administration of the Emissions Trading Scheme (2008); Portuguese Environmental Agency (2008).

OCR for page 21
 NATIONAL INVENTORIES OF GREENHOUSE GAS EMISSIONS in 2006, raised the initially reported 2000 values by 23 Finally, emissions from fires are highly uncertain (>100 percent. In contrast, Austria’s self-reported estimates percent) because of uncertainty in the amount of fuel have been refined every year but have remained within actually combusted and in the trace gas emissions per a range of 2-3 percent. unit of fuel burned (Campbell et al., 2007). Methods also make a difference in fossil-fuel CO2 emissions estimates. For example, the IEA examined Limited Availability of Independent Data Sources the effect of replacing the 1996 IPCC guidelines with for Validation and Verification the 2006 IPCC guidelines for fossil-fuel CO2 and found that only the inventories that rely on default There are no truly independent sources of activ- values (i.e., use Tier 1 methods) for factors such as heat- ity data, such as fuel use, against which data used in ing values and carbon contents of fuels were affected, national greenhouse gas inventories can be compared. but inventories that used higher-tier methods were not The United Nations, International Energy Agency, (IEA, 2009). Department of Energy (DOE) Energy Information Administration, and BP Corporation create large Uncertainty in AFOLU Emissions. Reported uncer- international datasets on energy production and con- tainties for annual CO2 emissions from the AFOLU sumption, but all of these datasets rely primarily on sector in Annex I countries range from less than 10 the same self-reported national statistics. Data on fuel percent to 100 percent (Table 2.2). Independent esti- production and trade are sometimes available from mates in the United States and other countries typically corporate sources, but the most complete energy data yield uncertainties in excess of 50 percent (Pacala et al., are self-reported by countries. In many developing 2001; CCSP, 2007; Ito et al., 2008). The true uncer- countries, these data are not complete or accurate or tainty probably exceeds 100 percent in many or most are not consistently reported. non-Annex I countries. Methodological uncertainty is more important for Validation of Fossil-Fuel CO2 Emissions. Direct mea- AFOLU emissions than for energy CO2 emissions. surements of CO2 emissions are not currently used in Uncertainties for agricultural methane emissions are preparing national-level inventories but could be used greater than 50 percent for rice cultivation due to vari- to validate portions of national inventories (Ackerman ability in irrigation and other management practices, and Sundquist, 2008). For example, at several hundred as well as inherent spatial and temporal variability power plants in the United States, continuous monitor- in soil CH4 production and consumption rates. The ing equipment measures the amount of CO2 and other situation is better for CH4 emissions associated with gases discharged from the smokestack (EPA, 2005). animal husbandry, because uncertainties in CH4 emit- This technique is both complex and expensive and ted per animal are on the order of 30 percent for Tier it is applied only at large point sources. Continuous 1 methods and 20 percent for Tier 2 methods (IPCC, monitoring of emissions is particularly useful where 2006), and the numbers of animals are reasonably well the fuel is heterogeneous or its delivery rate is difficult known (Table 2.1). Methane emissions from manure to measure. Ackerman and Sundquist (2008) compared are uncertain primarily because of poor documentation CO2 emissions estimates from fuel-based calculations of manure management practices by small farms. N2O and direct stack measurements at 828 U.S. power plants emissions have high temporal and spatial variability due that used conventional fuels in 2004. The average abso- to the dynamic nature and variability of the concentra- lute difference between the two sets of values was 16.6 tion of mineral nitrogen species (e.g., NH4+, NO3–), percent, with the stack measurements giving higher labile organic carbon, and oxygen that largely govern values on average. However, because the stack measure- N2O emissions from soils. Consequently, uncertain- ments are both higher and lower than the fuel-based ties for N2O emissions from anthropogenic nitrogen calculations, the two types of estimates differed by only additions are high—greater than 50 percent—because 1.4 percent for total conterminous U.S. CO2 emissions. of the variability in the direct and indirect emission of Because emissions from power plants are such a large N2O per unit of nitrogen added as well as uncertainty fraction of the U.S. total, resolving the differences in in nitrogen addition rates and management practices. these two methods may be an efficient way to reduce

OCR for page 21
 VERIFYING GREENHOUSE GAS EMISSIONS uncertainty in national inventories (Ackerman and and afforestation with reasonable accuracy, and ground Sundquist, 2008). surveys can be used to verify carbon stock changes Similarly, facility data collected to support emis- resulting from alterations in management. Verifying sion trading programs could also be used to validate interventions to reduce enteric CH4 emissions (e.g., and improve national inventories. For example, the genetic improvement of livestock, use of methane European Emissions Trading System created a new, inhibitors) and N2O soil emissions (e.g., altered timing, independent data source, which European countries amount, and placement of fertilizer; use of nitrifica- use to identify and fill gaps in national inventories and tion inhibitors) depends on ground survey informa- to improve country-specific emission factors (Herold, tion and self-reporting. Water management (i.e., 2007). Data collected under EPA’s recently adopted extent and periodicity of flooding of rice fields) could greenhouse gas reporting rule4 can play a similar role be monitored via remote sensing, although assessing in improving the U.S. inventory. other methane mitigation practices (e.g., improving Atmospheric measurements may also be useful cultivars, field additives to suppress CH4 production) for verifying values or trends in CO2 emissions (see would depend on ground survey data. As outlined in Chapter 4). For example, problems with the energy Chapter 4, atmospheric sampling networks and trans- data from China (see “Uncertainty in Fossil-Fuel CO2 port modeling for agriculture-intensive regions might Emissions” above) had been identified earlier based on help to constrain emission estimates of CH4 and N2O satellite measurements of trends in NO2 column abun- from agricultural sources. dance (Stinton, 2001; Akimoto et al., 2006; Zhang et Proxy data could also be used to validate AFOLU al., 2007; Gregg et al., 2008). e missions. For example, fertilizer production and In the absence of physical measurements of CO2 import-export statistics could be used to verify overall emissions, proxy data could be used to assess some reductions in fertilizer nitrogen use at the country scale trends. However, a climate treaty would change the and, together with statistics on agricultural production, historic relationships between emissions and proxies. to track changes in nitrogen fertilizer use efficiency as For example, gross domestic product (GDP) is strongly an indicator of improved management practices. correlated with CO2 emissions, but this relationship changes through time, particularly after an energy price Limited Comparability with Data Derived from shock, and it differs by a factor of 2 or more among Other Monitoring Methods countries (see, for example, Raupach et al., 2007). Emissions reduction policies are designed specifically UNFCCC inventories are difficult to compare to decrease the trend in emissions, while having mini- with physical measurements because (1) they do not mal impact on the trend in GDP. Independent data on provide complete accounting of greenhouse gas sources world trade might indicate trends in CO2 emissions in and sinks, (2) geographically and temporally resolved some countries and sectors. For example, 32 countries emissions data are not reported, and (3) final emissions in the UN energy database burn only imported liquid values are commonly not available for 2 years or more fuels, and the import and export of fuel are captured after they occur. National inventories do not include in trade statistics. all emissions of greenhouse gases, primarily because the UNFCCC addresses only anthropogenic emis- Validation of AFOLU Emissions. Independent verifi- sions and removals. Thus, for the AFOLU sector, only cation of changes in emission rates is difficult for all anthropogenic emissions and removals on managed lands are required to be counted.5 Because unmanaged but the largest AFOLU sources because field mea- surements are scarce and expensive. As discussed in Chapter 3, remote sensing can be used to quantify sig- 5The IPCC defines managed land as “land where human inter- ventions and practices have been applied to perform production, nificant changes in land use and rates of deforestation ecological or social functions” (IPCC, 2006). Individual countries are permitted to define what constitutes managed lands, as long 4 S ee . consistently over time.

OCR for page 21
 NATIONAL INVENTORIES OF GREENHOUSE GAS EMISSIONS land is not considered, the UNFCCC inventories miss revisions of earlier-year estimates when more accurate emissions and removals from natural disturbance and or more complete data become available. recovery as well as emission increases induced by cli- mate change (e.g., increased CO2 and CH4 emissions NEAR-TERM CAPABILITIES FOR in high-latitude tundra and boreal ecosystems). As a IMPROVING NATIONAL GREENHOUSE result, there is an inherent mismatch in the land-use GAS INVENTORIES component between UNFCCC inventories, process studies (e.g., Ito et al., 2008), and the atmospheric Self-reporting has been, and is likely to continue methods described in Chapter 4. In addition, some to be, the primary means of monitoring greenhouse anthropogenic sources are not reported, either because gas emissions and reductions under an international methods to measure them have not been developed climate treaty. Tier 1 IPCC methods deliver national or approved (e.g., CH4 from reservoirs) or because fossil-fuel CO2 emissions estimates that are sufficiently UNFCCC guidelines do not require their reporting accurate to document national multiyear trends of the (e.g., harvested wood products). dominant greenhouse gas source for the energy sector as The lack of geographically resolved data is impor- a whole, although higher-tier methods are required to tant for tracer-transport inversions, which depend on accurately estimate emissions by subsector. In contrast, assumptions about the initial spatial and temporal Tier 1 methods do not yet produce sufficiently accurate pattern and magnitude of emissions. An inverse model emissions estimates from the next-largest sources: CO2 evaluates how atmospheric observations can alter these from deforestation and CH4 and N2O from agricul- prior estimates while maintaining the prior pattern ture. However, even where more accurate Tier 2 and (e.g., Gloor et al., 2000; Rayner and O’Brien, 2001; 3 methods exist, many developing countries lack the Law et al., 2003). The initially estimated pattern of resources and infrastructure to use them. Improving the emissions at the surface is critical for the analysis (e.g., inventories of both developed and developing countries L auvaux et al., 2008). The availability of accurate and enhancing self-reported data to facilitate indepen- geographically and temporally resolved inventory data dent verification are discussed below. Chapters 3 and would provide better prior estimates of emissions. If 4 describe research needed to strengthen knowledge of we know where and when the emissions occurred, as the poorly constrained minor sectors and gases, as well well as how much was emitted, the tracer-transport as independent data on emissions that could improve modeling can tell us the extent to which atmospheric confidence in reported inventories. m easurements are compatible with the emissions estimates. Although many developed countries collect Building Inventory Capacity in Developing some spatially resolved greenhouse gas data to support Countries mitigation or air quality programs, this information is not included in national inventory reports. Although technical and financial assistance has The delay in reporting final emissions values frus- helped many developing countries complete a national trates efforts to independently validate national inven- inventory, the challenge is to build the capacity for tories with real-time physical measurements, although these countries to create complete and accurate inven- it does not compromise the use of national inventories tories regularly and for all years. Some international for treaty purposes. Much of the activity data on which efforts to improve energy statistics include a compo- emissions estimates are based is collected on survey nent to increase participation by developing countries. forms from industrial firms and energy producers and Examples include the Oslo Working Group on Energy consumers. It may take several years to collect and Statistics, which is working to refine international stan- process data after the close of an accounting year. As dards and definitions (e.g., United Nations, 2008a,b,c), a result, many emissions estimates are published 1 to and IEA initiatives to harmonize energy statistics 2 years after the emissions actually occur, and annual among international organizations (Karen Treanton, updates include not just another year of data but also IEA, personal communication, July 2009).

OCR for page 21
 VERIFYING GREENHOUSE GAS EMISSIONS Improving estimates of emissions from defores- energy systems and land use. Full and open access to tation and agriculture, which are primary sources of international data would make it possible for any party greenhouse gases in developing countries, is challeng- to make estimates, compare data, or consider proxy ing because many countries do not have the technical indications. The United Nations, IEA, and FAO have capacity to monitor these sources. International efforts established standards to ensure that the collection and to develop comprehensive global land category maps reporting of publicly available data in international data and to disseminate affordable satellite imagery could compilations are uniform across countries and over greatly improve emissions estimates in developing time. These international compilations include some countries by providing country-level activity data on of the same data used by countries to estimate their land use for preparation of national inventories (see emissions (e.g., production, trade, and consumption Chapter 3). of energy; fuel characteristics; land use; agricultural In many cases, external funding and training will production), but they provide supporting information be required to strengthen government and research that allows comparisons and correlations and they are institutions in developing countries to build and retain publicly accessible. In contrast, data in some countries expertise in greenhouse gas inventories. The cost of have limited availability (e.g., hard to find, incomplete, creating an ongoing capability for Tier 1 reporting in on paper only) or are available only at cost. the largest emitting developing countries using exist- ing data is relatively modest. For example, an initial Facilitating Independent Verification of Self- investment of $450,000 per country6 for data collection, Reported Emissions Data training, and software and inventory tools could greatly improve the capacity of those countries to use higher- Complete accounting of all greenhouse gas sources tier methods for inventory preparation. Additional and sinks would enable more accurate comparisons resources on the order of $200,000 annually would of country totals with global and atmospheric data, be needed in countries without existing institutional facilitate regional estimates, and improve confidence in capacity to maintain a permanent team of experts for national estimates. Countries should continue to move inventory preparation (Mausami Desai, Environmental toward more complete reporting of anthropogenic Protection Agency, personal communication, Septem- sources and sinks within the UNFCCC inventories. ber 24, 2009). If only half of the 20 highest-emitting Initiatives needed to develop estimates of naturally developing countries require support for institutional occurring biogenic emissions and removals, such as capacity, the cost to obtain annual estimates of their the production of global land maps and research on emissions would be $11 million over 5 years. biogeochemical cycles, are discussed in Chapter 3. The development of spatially and temporally gridded emission datasets is also critical for improving Improving Access to Data in Developed Countries atmospheric models, and several initiatives are currently under way. For example, the Vulcan project7 estimates Developed countries typically have the infrastruc- ture needed to generate reasonably accurate inventories hourly emissions based on facility-level data and geo- of emissions from many significant sources and to carry graphic patterns of fuel use, but so far it covers only out research on sources that are poorly understood fossil-fuel CO2 emissions from the continental United (e.g., CH4 and N2O in the AFOLU sector). However, States and only for a year. Annual gridded greenhouse developed-country inventories would be improved gas estimates for the world are available from the by providing full and open access to basic data on Emissions Database for Global Atmospheric Research ( EDGAR), 8 b ut they are based on disaggregated 6This is generally consistent with Global Environment Facility national, annual data and indicators of local emission- funding levels ($450,000 per country for national communications). producing activities. These initiatives are useful begin- However, inventory preparation is only one element of the national communication and only one year of data is currently required. See 7 See . 8 See pdf>. .

OCR for page 21
 NATIONAL INVENTORIES OF GREENHOUSE GAS EMISSIONS nings, but they need to be significantly expanded. An sions of CH4 could be as low as 10 percent in countries international effort among developed countries to where emissions are dominated by energy, and as high prepare and publish spatially and temporally gridded as 50 percent in countries where emissions are domi- estimates of national emissions would provide critical nated by AFOLU. initial data for the independent verification of national Extending regular UNFCCC inventories to other emissions, and for this effort, the disaggregated data countries, pushing inventories to higher tiers, carrying may be more important than just the total magnitudes out research to improve emission factors, and encourag- of emissions. ing the expansion of inventories to estimate emissions Governments also have a critical role in developing at finer spatial and temporal scales should not only standardized methods to ensure that the gridded esti- reduce uncertainty, but also improve opportunities for mates are comparable and that they are produced at the independent verification of traditional, national emis- appropriate spatial and temporal resolution to support sions inventories. Consequently, UNFCCC parties comparison with the models. The horizontal resolution should strengthen self-reported national emissions necessary is likely to be 50 to 100 km for global models, inventories in the following manner: 8 km for regional models (Lauvaux et al., 2008), and 1 km for point sources. All of these methods would yield 1. Extend regular inventory reporting and review improved emissions estimates with any good initial to all countries. data, even if they were not at the optimal resolution for the method. Temporal resolution, particularly the • Where necessary, provide sustained techni- typical diurnal, weekly, and/or seasonal cycle of emis- cal and financial support to develop and maintain sions, is also a critical piece of initial data. Because of institutional capacity and tools and training for day-night patterns in meteorology, greenhouse gases preparing inventories in developing countries. emitted at different times of the day can end up at • Create a central land-use database to improve different locations. For near-instantaneous measure- AFOLU estimates in national inventories from ments near point sources, the emissions measured are developing countries (see Chapter 3). only from the last hour and depend on the duty cycle of the power plant. 2. Continue to improve methods used by all countries. RECOMMENDATIONS • Support basic research on greenhouse gas The uncertainties in current estimates of emissions emissions processes and corresponding improve- for the various greenhouse gases and major sources, ments in IPCC methodologies, particularly for as evaluated by the countries submitting the national b iogenic sources and the AFOLU sector (see reports, are summarized in Table 2.2. Uncertainties Chapter 3). in self-reported CO2 emission estimates are low (<10 • Extend top-tier (most stringent) reporting percent) in many developed countries and could be to the most important greenhouse gas sources in lowered to similar levels in others by using the most each country. accurate IPCC methods. Reducing uncertainties for N2O and CH4 emission estimates will also require 3. Facilitate cross-comparisons of self-reported improved activity data and emission factors, which will data with data derived from other monitoring methods in some cases require research. Uncertainties in total and develop independent data sources. anthropogenic N2O are driven by the AFOLU com- ponent and are likely to remain high (10-100 percent) • Support the development of inventories of in the near term. For CH4, the relative importance of naturally occurring, land-based emissions and sinks energy and AFOLU sources will determine the extent for all lands, not just managed lands. to which uncertainty can be reduced. If improvements • Promote free and open access to relevant are made, uncertainties for total anthropogenic emis- national and international statistics.

OCR for page 21
 VERIFYING GREENHOUSE GAS EMISSIONS • Support IPCC, United Nations, and IEA ods for preparing and publishing Annex I country efforts to improve energy statistics and FAO efforts inventories that are gridded at spatial and temporal to improve land-use and forestry statistics. resolutions appropriate for the particular green- • Develop and implement standardized meth- house gas and source.