HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lokupitiya, E. Articles by Paustian, K. Search for Related Content PubMed PubMed Citation Articles by Lokupitiya, E. Articles by Paustian, K. Agricola Articles by Lokupitiya, E. Articles by Paustian, K. Related Collections Biogeochemical Processes Global Change Agricultural Systems Soil Systems Air Pollution

Agricultural Soil Greenhouse Gas Emissions

A Review of National Inventory Methods

Department of Soil and Crop Sciences and Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO 80523

^* Corresponding author (elokupit{at}nrel.colostate.edu)

Received for publication May 1, 2005.

	ABSTRACT

TOP ABSTRACT INTRODUCTION IMPORTANCE OF AGRICULTURAL SOILS... METHODOLOGIES FOR ESTIMATING... DISCUSSION AND CONCLUSIONS REFERENCES

 
Parties to the United Nations Framework Convention on Climate Change (UNFCCC) are required to submit national greenhouse gas (GHG) inventories, together with information on methods used in estimating their emissions. Currently agricultural activities contribute a significant portion (approximately 20%) of global anthropogenic GHG emissions, and agricultural soils have been identified as one of the main GHG source categories within the agricultural sector. However, compared to many other GHG sources, inventory methods for soils are relatively more complex and have been implemented only to varying degrees among member countries. This review summarizes and evaluates the methods used by Annex 1 countries in estimating CO₂ and N₂O emissions in agricultural soils. While most countries utilize the Intergovernmental Panel on Climate Change (IPCC) default methodology, several Annex 1 countries are developing more advanced methods that are tailored for specific country circumstances. Based on the latest national inventory reporting, about 56% of the Annex 1 countries use IPCC Tier 1 methods, about 26% use Tier 2 methods, and about 18% do not estimate or report N₂O emissions from agricultural soils. More than 65% of the countries do not report CO₂ emissions from the cultivation of mineral soils, organic soils, or liming, and only a handful of countries have used country-specific, Tier 3 methods. Tier 3 methods usually involve process-based models and detailed, geographically specific activity data. Such methods can provide more robust, accurate estimates of emissions and removals but require greater diligence in documentation, transparency, and uncertainty assessment to ensure comparability between countries. Availability of detailed, spatially explicit activity data is a major constraint to implementing higher tiered methods in many countries.

Abbreviations: CRF, Common Reporting Format • GHG, greenhouse gas • IPCC, Intergovernmental Panel on Climate Change • IPCC-GL, Intergovernmental Panel on Climate Change Guidelines • IPCC-GPG, Intergovernmental Panel on Climate Change Good Practice Guidance • NIR, National Inventory Report • UNFCCC, United Nations Framework Convention on Climate Change

	INTRODUCTION

TOP ABSTRACT INTRODUCTION IMPORTANCE OF AGRICULTURAL SOILS... METHODOLOGIES FOR ESTIMATING... DISCUSSION AND CONCLUSIONS REFERENCES

 
EFFORTS TO MITIGATE increasing greenhouse gas (GHG) concentrations through international cooperation have accelerated since the beginning of the 1990s, with the adoption of the United Nations Framework Convention on Climate Change (UNFCCC) in 1992 and the Kyoto Protocol in 1997. The UNFCCC entered into force in 1994 and so far over 180 countries have ratified it. The prime objective of the UNFCCC is to stabilize atmospheric GHG emissions at a level that would not further damage the climate system from human interference.

Under the UNFCCC, member countries are expected to submit national GHG inventories. Estimating the sources and sinks of GHG emissions at a national level is needed to quantify the sources and sinks from individual countries and to assess compliance with international agreements to reduce emissions. Accurate inventories also improve understanding of the relative importance of different sinks and sources and their spatial distribution. Developed countries that are listed in Annex 1 of the UNFCCC (referred to as "Annex 1 countries") are expected to submit an annual National Inventory Report (NIR) containing detailed and complete information on their emissions for all years from a base year (which is 1990 for the majority of the countries) to the current year for the annual inventory submission (United Nations Framework Convention on Climate Change, 1999).

Annex 1 countries have submitted national GHG inventories since 1994, and since 1996 they have made annual submissions. The quality, information content, and transparency differ substantially among the reporting countries (Acosta et al., 2002) and only some countries have published detailed method descriptions. According to Acosta et al. (2002), methodological problems associated with GHG inventories include the use of outdated emission factors, insufficiently robust assumptions, calculation errors, and use of methods, parameters, and emission factors that deviate substantially from IPCC default method and values.

Inventory methods for land use and agriculture related activities, in general, and for soils in particular, are arguably among the more complex and least developed inventory sectors, yet soils are a significant sink–source category for many countries. This paper reviews the methodologies used by member parties to the UNFCCC to prepare national GHG inventories of soil emissions and sinks. Our objective is to assess the current state, recent developments, and future trends in soil emission inventory methods. Since many non-Annex 1 countries have not yet reported soil emission inventories, our focus is on methodologies used by Annex 1 countries.

	IMPORTANCE OF AGRICULTURAL SOILS AS A SOURCE AND SINK OF GREENHOUSE GAS EMISSIONS

TOP ABSTRACT INTRODUCTION IMPORTANCE OF AGRICULTURAL SOILS... METHODOLOGIES FOR ESTIMATING... DISCUSSION AND CONCLUSIONS REFERENCES

 
Currently, agricultural activities contribute a significant portion (approximately 20%) of global GHG emissions (Fig. 1), making it imperative for many countries to inventory GHG emissions from agricultural soils. Nitrous oxide (N₂O) is the predominant GHG added to the atmosphere from agricultural soils, in terms of global warming potential. Nitrous oxide is produced through both nitrification and denitrification, which are microbial-driven soil processes that naturally occur in soils; these processes are stimulated by increased N inputs. Currently about 60% of total N₂O emissions in Annex 1 countries are derived from agriculture (Fig. 2). Anthropogenic increases in N₂O emissions in agricultural soils occur mainly as a result of fertilizer and manure use. Total N₂O emissions from all Annex 1 countries (Table 1), as well as for the United States alone, remained fairly constant during the 1990s (Fig. 3).

View larger version (52K): [in this window] [in a new window]   Fig. 1. Percent contribution of agricultural sector greenhouse gas emissions to increased radiative forcing attributed to anthropogenic enhancement of the greenhouse effect (Intergovernmental Panel on Climate Change, 1996).

View larger version (14K): [in this window] [in a new window]   Fig. 2. Nitrous oxide emission from agricultural soils as a fraction of cumulative N2O emissions by Annex 1 countries from 1990–2000.

View larger version (14K): [in this window] [in a new window]   Fig. 3. Nitrous oxide emission from agricultural soils by the United States and all Annex 1 countries from 1990–2000.

Although methane (CH₄) emissions from agricultural soils, specifically in paddy rice, are an important GHG source globally, CH₄ from rice cultivation is of minor importance for Annex 1 countries. Hence, only methods for estimation of CO₂ emissions and removals and N₂O emissions from agricultural soils are dealt with in this paper.

	METHODOLOGIES FOR ESTIMATING GREENHOUSE GAS EMISSIONS AND REMOVALS IN AGRICULTURAL SOILS

TOP ABSTRACT INTRODUCTION IMPORTANCE OF AGRICULTURAL SOILS... METHODOLOGIES FOR ESTIMATING... DISCUSSION AND CONCLUSIONS REFERENCES

 
According to Articles 4 and 12 of the UNFCCC, national inventories need to use comparable methodologies agreed on by the Conference of Parties (COP), the UNFCCC's principal policy-making body. Currently, parties to the UNFCCC use the 1996 Revised IPCC Guidelines (Intergovernmental Panel on Climate Change, 1997a, 1997b, 1997c), as elaborated by the IPCC Good Practice Guidance (Intergovernmental Panel on Climate Change, 2000b, 2003), and the UNFCCC Reporting Guidelines on Annual Inventories (United Nations Framework Convention on Climate Change, 1999). The IPCC Guidelines provide general guidance, default calculation methods, and reporting formats for national inventories to promote transparency and comparability among countries. The IPCC Guidelines (IPCC-GL) and Good Practice Guidance (IPCC-GPG) provide a default methodology (Tier 1) for estimating GHG emissions as well as guidance in developing and using country-specific data and methods (referred to as Tier 2 and Tier 3 methods—see below). Some Annex 1 countries have begun to develop and implement these country-specific "higher-tiered" methodologies. The methodologies currently used by different member countries to the UNFCCC in estimation of CO₂ and N₂O are summarized in Table 2. A description of the 1996 IPCC default methodology and methods developed by some Annex 1 countries for agricultural soils follow.

Nitrous Oxide Emissions from Agricultural Soils
The IPCC-GL and IPCC-GPG provides methodology for estimating (i) direct N₂O emissions from soils; (ii) indirect emissions of N₂O (i.e., due to losses from N volatilization and leaching that are subsequently deposited in non-agricultural environments and are subject to loss as N₂O); and (iii) direct soil emissions of N₂O from animal production (during waste storage for confined livestock and from livestock grazing). Direct N₂O emissions from soil are estimated using an equation (Eq. [1]) that incorporates emissions from all major N inputs and does not discriminate among different N sources (Intergovernmental Panel on Climate Change, 1997b, 1997c):

[1]

where F_SN is the N input from synthetic fertilizer use (kg N yr^–1); F_AW is the N from livestock manure applied to soil (kg N yr^–1); F_BN is the total N input in N-fixing crops (kg N yr^–1); F_CR is the N input from crop residues (kg N yr^–1); F_PRP is the N input from animal excretion on pasture, range, and/or paddock (kg N yr^–1); F_OS is the area of cultivated organic soils (ha); EF₁ is the emission factor for direct N application (kg N₂O-N kg^–1 N added); EF₂ is the emission factor for cultivated organic (e.g., peat) soils (kg N₂O-N ha^–1 yr^–1); and EF₃ is the emission factor for excretion on pasture, range, and/or paddock (kg N₂O-N kg^–1 N added).

Indirect soil N₂O emissions are defined as deriving from the volatilization (as NH₃ and NO_x) and subsequent deposition of previously applied N, as well as emissions from applied N that has been transported to riparian or aquatic environments through leaching and runoff. Thus the actual emissions may be occurring in non-agricultural environments, but the original source of the applied N was from agricultural soils. Other N₂O emissions from animal waste management systems are included in the agricultural sector but are not related to soil management.

According to the latest National Inventory Report (NIR) submissions, a majority of reporting countries use the IPCC default method (about 70% of the countries) and default emission factors (about 75% of the countries) in estimating N₂O emissions from agricultural soils.

Emissions and Removals of Carbon Dioxide from Agricultural Soils
Depending on the management practices being used, and their relative effect on C inputs from residues vs. C losses from decomposition, agricultural soils can be either a net source or a net sink for C (Lal, 2004; Paustian et al., 1997, 2000; Smith, 2004). The IPCC methodology estimates net CO₂ emissions (sinks and sources) from: (i) changes in C stocks of mineral soils due to changes in land-use practices; (ii) CO₂ emissions from organic soils converted to agriculture or plantation forestry; and (iii) liming of agricultural soils (Intergovernmental Panel on Climate Change, 1997b, 1997c, 2003).

For CO₂ emission from mineral soils, the net change in soil organic C is estimated for lands under different categories of land use and management, stratified by climate and soil type, over a specified time period; the default time period is 20 yr. Estimates of soil C stocks are for the top 30 cm of the soil profile, where the impacts of changes in land use and management are greatest and where most field measurements have been reported (Ogle et al., 2003a). Hence, changes in soil C stocks that may occur deeper in the profile are not captured by the method.

The default methodology uses a set of coefficients (stock change factors) based on soil type, climate, disturbance history, productivity, and management practices. Climate is divided into nine categories based on average annual temperature and precipitation. Soils are defined by taxonomic characteristics based on broadly defined soil properties, including texture, clay mineralogy, morphology, and drainage, that influence the ability of a soil to store organic matter. Default values for reference C stocks and stock change factors are stratified according to climate and soil type. Reference C stocks represent values found under native, unmanaged ecosystems.

The basic method combines the reference C stock, stock change factors, and activity data for land use and management changes over time. Mineral soil C stock change is estimated as shown in Eq. [2] (Intergovernmental Panel on Climate Change, 1997c, 2003):

[2]

where:

where SC is the annual soil carbon stock change (Mg C yr^–1); SC₀ is the soil organic carbon stock in the inventory period end year, for current land use and management (Mg C ha^–1); SC_(0-T) is the soil organic carbon stock T years before the end year of the inventory period (Mg C ha^–1); c represents climate zone and s soil type, and i is the set of land management systems defined for the country; A is the land area of each parcel (ha); SC_R is the reference carbon stock (Mg C ha^–1); F_LU is the stock change factor for land use type (dimensionless); F_MG is the stock change factor for management/disturbance regime (dimensionless); F_I is the stock change factor for carbon input level (dimensionless); and D is the time period for transition between equilibrium C stocks, as represented by stock change factors (default is 20 yr).

While the formulation is designed to be generic for all soils, the interpretation and values of these factors vary according to the type of ecosystem (i.e., cropland, grassland, forest) and the changes in land use and management that are being represented. For example, the land use factor provides a baseline level for C stocks in permanent cropland, short- (20 yr) and long-term (>20 yr) cropland set-aside (to perennial grasses or trees), shifting cultivation, and managed grassland and forest, relative to stock levels in native unmanaged ecosystems (where F_LU = 1). For cropland, the stock change factor for management regime (F_MG) specifies relative C stock values for different tillage regimes. For cropland and managed grassland (pasture, hayland), the input factor (F_I) relates to management practices that affect the relative amount of C returned to the soil (as plant-derived residues or exogenous additions like animal manure) for a particular land use type. Hence for cropland, it is dependent on the type of crops grown, whether residues are removed or retained, and whether manures are added. For grassland, F_I depends on management practices that influence primary productivity, such as fertilization, species improvement, and grazing regime. More detailed definitions and default values for stock change factors are given in the Good Practice Guidance for LULUCF (Intergovernmental Panel on Climate Change, 2003).

The default inventory period, to which land use and management activity are applied, is 20 yr, that is, default values for relative stock change factors are estimated for a 20-yr period (Intergovernmental Panel on Climate Change, 2003). However, the method can be applied for an inventory period of 20 yr or less, using the default factors. In calculating inventories for year t, land areas are stratified by climate, soil type, and "initial" land use and management for year t – T (where T is the length of the inventory period), and then land use and management conditions are specified for the same areas in year t. Changes in C stocks from one land use–management system to another are assumed to be linear over time. Implicit in the method is that if there are no changes in land use or management (i.e., F_LU, F_MG, and F_I are unchanged) for a given land area over the inventory period, then soil C stocks remain constant and there is no net emission or removal of CO₂.

The CO₂ emissions from cultivated organic soils and from liming of soils are handled differently from C stock changes in mineral soils. Both are considered as only sources of CO₂ and are estimated using simple emission factors (i.e., annual loss per unit area) multiplied by the areas of each activity. Net C loss from organic soils is calculated using the land area and annual loss rates that vary by broad climate divisions and land-use. The default values for annual loss rates of C given in this method are derived from a global survey on published literature. Emissions from liming are estimated assuming that the carbonate-C added (i.e., limestone or dolomite) is emitted as CO₂ in the year of application.

CORINAIR Methodology
Countries of the European Union (EU) prepare their national inventories according to the EU emission inventory program known as CORINAIR (CORe INventory of AIR emissions in Europe). It was initiated in 1985 to assist in the development of consistent, comparable, and transparent national inventories for "conventional" air pollutants such as SO_x, NO_x, and volatile organic compounds (VOCs). This system has evolved over time and guidelines for preparing atmospheric emission inventories by the EU member countries are included in the EMEP (i.e., Co-operative Programme for Monitoring and Evaluation of the Long-Range Transmission of Air Pollutants in Europe)/CORINAIR Atmospheric Emission Inventory Guidebook, first published in 1996. The source categories covered under CORINAIR 1990 include 11 main source sectors, including agriculture.

CORINAIR methodology involves more disaggregated source categories and spatial detail, and CORINAIR-based estimates can be transformed appropriately for different reporting purposes, including the IPCC format. EMEP/CORINAIR (2002, 2004) has classified GHG emissions from agricultural soils [in accordance with Common Reporting Format (CRF)/IPCC classification] under the categories: (i) Cultures with fertilizers [where emissions of NH₃, N₂O and NO_x, CO₂, CH₄, and non-CH₄ volatile organic compounds (NMVOCs) are estimated]; and (ii) Cultures without fertilizers (where emissions of NH₃, N₂O, NO_x, and VOCs are estimated). The methodology suggested for estimating N₂O emissions follows the IPCC default methodology; IPCC Direct and Indirect N₂O subsource categories in agricultural soils have been reported as CORINAIR subsectors for cultures with or without fertilizers (EMEP/CORINAIR, 2004).

In addition to the simpler methodology based on the IPCC default method, an improved methodology for N₂O emissions is given; methods have been developed based on multivariate regression analyses that incorporate important factors such as climate, weather, and soil conditions that control N₂O emissions. Mechanistic simulation models such as DNDC (Li, 2000) have also been applied at regional scale to inventory N-trace gas emissions (EMEP/CORINAIR, 2004).

No alternative methodology from the IPCC has been suggested for estimating CO₂ emissions and removals under Cultures with fertilizers. The IPCC land use changes considered under Cultures with fertilizers include: (i) conversion of woodland to grassland and cropland, (ii) conversion of grassland to cropland and vice versa, and (iii) other land use change activities that include drainage of wetlands and cultivation of organic soils. Although this latter category of land use changes is a significant source of CO₂ in certain countries (especially in northern Europe), such CO₂ emissions are rarely reported in national inventories, as no default methodology has been provided under the CORINAIR methodology. However, higher emission factors compared to IPCC for CO₂ released from cultivated organic soils have been suggested based on recent measurements made in Europe (EMEP/CORINAIR, 2004).

National Greenhouse Gas Accounting Systems Developed by Certain Annex 1 Countries
Currently the IPCC default methodology dominates among the methods used by Annex 1 countries in estimating national GHG emissions in agricultural soils. In recognition of the limitations inherent in using global and regional default values, the IPCC-GL and IPCC-GPG encourage countries to develop methodologies more appropriate for their national circumstances. However, development of such methods requires considerable time and resources, including testing and validation before implementation. Consequently, relatively few country-specific systems have been fully implemented to date. In this section, country-specific methods developed by certain Annex 1 countries to estimate agricultural soil CO₂ and/or N₂O emissions and removals are discussed. In each of these countries, emissions and/or removals from soils are a major component of the total impact of GHG in the agricultural sector (Fig. 4). Some of these methods are still in the development process and some are not fully utilized for agricultural soils at the national level and have been used and validated mostly at regional or project level. An overview of the methods used by all Annex 1 countries in estimating CO₂ and N₂O emissions from agricultural soils is given in Table 2.

View larger version (24K): [in this window] [in a new window]   Fig. 4. Relative impact of agricultural soils to the overall agricultural sector, in terms of global warming potential (computed as absolute values for both sinks and sources combined for CO2, N2O, and CH4), as a percentage of the total agricultural sector emissions for some selected Annex 1 countries. "Other" emissions represent combined N2O and CH4 emissions from burning of savannas and crop residues and CH4 from rice.

Australia has developed a National C Accounting System (NCAS) based on resource inventories, field studies, modeling, and remote sensing. NCAS involves a verified model-based accounting system operating at highly disaggregated spatial and temporal scales (25 m, monthly time steps; Australian Greenhouse Office, 2002b). Several submodels comprise the Full C Accounting Model (FullCAM) for estimating land use change emissions. FullCAM has components that incorporate C exchange between the atmosphere and agriculture- and forestry-related activities. Emissions and removals are estimated for biomass as well as soil C pools.

The five submodels of the FullCAM include a physiological growth model for forests (CAMFor), a C accounting model for cropping and grazing system (CAMAg), a residue decomposition model (GENDEC), and the Rothamstead soil C model (Roth-C) (Richards, 2001; Australian Greenhouse Office, 2002b). The FullCAM model provides a linkage between these submodels.

CAMAg reflects the management impacts on C accumulation and allocates crop biomass to various plant product pools and to decomposable and resistant organic residues. Change in agricultural soil C is estimated using Roth-C model (Coleman and Jenkinson, 1995), based on soil type, land use and management history, and residue inputs from different cropping systems. This model has been calibrated against long-term field measurements and verified using paired sites (undisturbed vs. cleared sites) in areas of major land use change; although further measurements from long-term field experiments are needed to refine the model (G. Richards, personal communication, 2004). Since CAMFor and CAMAg are both included within the FullCAM, transitional activities such as deforestation, afforestation, reforestation, and mixed systems such as agroforestry can be represented (Australian Greenhouse Office, 2002b). In addition, a mathematical framework has been developed to incorporate the ability to estimate non-CO₂ GHGs within the FullCAM model. This will enable the estimation of non-CO₂ emissions from forestry (i.e., CH₄ from decomposition and burning, N from decomposition, burning, fertilization, and soil preparation) and agriculture (i.e., N from fertilizer application, animal excrement, soil management, decomposition, and burning; CH₄ from decomposition and burning) (Australian Greenhouse Office, 2002b).

Austria
In Austria, agricultural area is about 41%, and forests occupy about 46% of the total land area. Agricultural area includes arable land, grassland, as well as vineyards and orchards. Natural reforestation in former agricultural land, and aforestation activities have contributed to the further extension of the country's forest area within the last few decades (United Nations Framework Convention on Climate Change, 2004).

The Austrian Carbon Balance Model (ACBM) is being developed for full C accounting of C stocks in Austria, considering 1990 as the baseline. The overall model covers five national subsystems: agriculture, forestry, energy, production, and waste. For each module, C stocks, flows, processes, and control variables have been identified. The model is formulated to estimate the full national C balance, including intersystem C flows to show the impact of one action on the other components of the model, and subsequent net flux to the atmosphere (Orthofer et al., 2000; Gerzabek et al., 2003).

Agricultural soil is one main component within the module for agriculture. Soil C dynamics are modeled using a simple approach that divides organic matter into three pools based on residence times. Factors for simulation and calculation of C stocks are derived using official soil survey data (Orthofer et al., 2000). Net emissions (expressed in CO₂ equivalents) from agricultural soil estimated using the ACBM were 13% lower than emission estimates made using the IPCC-GL. This difference was mainly due to the effect of C sequestration in agricultural products and forest soils and of the net effect from C import and export (Gerzabek et al., 2003).

Soil C estimates from ACBM have not been so far reported in the NIR, although C stock changes in Austrian forest soils during 1990–2010 were estimated using this modeling approach (Weiss and Schlamadinger, 2000).

Canada
In Canada about 42% of the country is covered by the forest, and about 7% of the land is under agriculture. About two-thirds of the farmland is used for crops and improved pasture (Environment Canada, 2001).

At present, Canada is using the IPCC Tier 1 method to estimate agricultural soil N₂O emissions, and CO₂ emissions and removals from agricultural soils are estimated using the Century model that has been calibrated for Canadian conditions (Smith et al., 1997; C. Liang, personal communication, 2004). However, a new National C and Greenhouse-gas emission Accounting and Verification System for agriculture (NCGAVS) is being developed to estimate soil C change and direct N₂O emissions from agricultural soils (McConkey et al., 2003). It is a model-based system that uses integrated databases of information on land, land management, and land use change.

NCGAVS will provide a more detailed country-specific methodology for estimating N₂O and CO₂ emissions and removals from Canadian agricultural soils. The basic geographic units used in NCGAVS are the Soil Landscapes of Canada (SLC) polygons, which are mapped at a scale of 1:1 000 000, yielding around 5000 polygons for the entire country. Soil C changes and N₂O emissions will be estimated for the components within a SLC, so that important attributes such as soil organic matter, texture, and topographic effects will be addressed. Inclusion of topographic effects is an improvement compared to the existing inventory methodology.

The initial version of NCGAVS will concentrate on estimating sources–sinks of soil C and direct N₂O emissions from land-use change, crop- and grazing-land management, and revegetation under primary agriculture over a 5-yr inventory period. Since CH₄ emissions from agricultural soils are not a significant source in Canada, it will not be included initially. It is expected that NCGAVS will be used for the inventory submission in 2006 (C. Liang, personal communication, 2005). Future versions are planned to have more extensive coverage, accounting for all GHG sources and sinks in Canadian agriculture. Since N₂O represents the largest GHG source for Canadian soils, further efforts to quantify N₂O emissions will be a major focus (McConkey et al., 2003).

Germany
In Germany, about 54% of land was under agriculture in 1997, with two-thirds of the area used for annual crops. About 3% of agricultural land is organically farmed. Fertilizer use and livestock are the major sources of agricultural sector GHG emissions of N₂O and CH₄ (Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit, 2002).

Germany has developed a system using two mechanistic models, Denitrification and Decomposition (DNDC) (Li et al., 2000) and Photosynthesis and Evapotranspiration–Nitrification–Denitrification and Decomposition (PnET-N-DNDC; based on the PnET model by Aber et al., 1996) for estimating N₂O emissions from agricultural and forest soils, respectively (Butterbach-Bahl et al., 2001, 2002, 2004). These process-based models integrate complex interactions among primary drivers, soil environmental factors, and biogeochemical reactions. Input parameters include daily climate data, soil properties (organic matter, texture, pH, bulk density), vegetation (crop or forest type, age), and land management practices.

This approach has been tested regionally, and so far modeled N₂O emissions have been comparable with the estimates derived from using IPCC guidelines. The DNDC-based estimates at the regional scale were slightly higher (about 10%) compared to the estimates based on IPCC default method (Butterbach-Bahl et al., 2002). Model validation is difficult, however, due to the scarcity of field measurements over entire years or entire regions, as well as the high spatial variability of N₂O emissions (Butterbach-Bahl et al., 2004). Although the IPCC methodology has been used in the current and past inventory submissions, a national inventory of N₂O emissions using the simulation modeling approach will be used in future German national inventory submissions (K. Butterbach-Bahl, personal communication, 2004).

Methodology for estimating CO₂ emissions from German agricultural soils is still in development. Carbon dioxide emission from cultivated organic soils is a key category. Soil C stocks for annual and perennial cropland, vineyards, grassland, and fallow land have been estimated using remotely sensed data and soil data; emission factors have been derived using a review of about 200 national and international studies (A. Gensior, personal communication, 2004).

New Zealand
Agriculture is the principal industry in New Zealand, predominantly as intensive and extensive pastoral systems. Only about 1% of the land area of New Zealand is devoted to annual cropland. Forty-nine percent of total GHG emissions in 2002 were from agriculture, and N₂O emissions from agricultural soils accounted for 34% of agricultural sector emissions (New Zealand Ministry for the Environment, 2001; United Nations Framework Convention on Climate Change, 2004).

Although soil C stock changes are not currently reported (United Nations Framework Convention on Climate Change, 2004), a soil Carbon Monitoring System (CMS) is being developed to account for changes in soil C stocks due to land cover changes occurring in the recent past (i.e., conversion of grazing land to plantation forestry and to native woody vegetation). The soil CMS of New Zealand is based on a simple empirical model, similar in concept to the IPCC Tier 1 approach. This accounting system uses three data layers (i.e., soil, climate, and land use), for which steady-state soil C stocks are assigned using georeferenced soil C measurements. The six IPCC recommended soil classes, supplemented by a separate class for podzol soils, are defined and a more detailed breakdown of climatic zones compared to the IPCC approach is used (Scott et al., 2002; Tate et al., 2003). To incorporate erosion impacts on soil C, an erosivity index (i.e., the product of slope and mean annual precipitation) has been included with soil type, climate, and land use as the major determinants of soil C stocks. Soil C values for different land cover–land use categories have been estimated for each of 18 soil–climate classes (Scott et al., 2002; Tate et al., 2003).

Advantages over the IPCC default method include a New Zealand–specific representation of soil and climate conditions (e.g., Table 3), measured soil C stocks for climate–soil–land use conditions in New Zealand, and factoring in effects of erosion, which is high in part of the country. The CMS-based estimates for soil C values were slightly higher for high clay-activity soils in cold, temperate, dry climates and slightly lower for high clay-activity soils in cold, temperate, moist soils, compared to the default IPCC soil C values for native vegetation in similar soil–climate categories (Table 3). For estimation of soil C changes, periodic update of national land cover–land use data is essential. Currently, uncertainty about changes in areas under different land cover and limited data on the effects of key land-use changes on soil C stocks are the major sources of uncertainties in estimating soil C emissions and removals.

View this table: [in this window] [in a new window]   Table 3. Organic C values derived for arable soils in New Zealand using the soil Carbon Monitoring System (CMS), compared with Intergovernmental Panel on Climate Change Good Practice Guidance (IPCC GPG)/default soil C values (0–30 cm) for native vegetation in similar soil and climate categories.Scott et al. (2002), Intergovernmental Panel on Climate Change (2003).

Sweden
In Sweden, more than two-thirds of the country is covered by forests, and agricultural land comprises 8% of land area, concentrated in the southern part of the country. A significant proportion (9%) of annual cropland is on cultivated peat soils (histosols) (Swedish Ministry of the Environment, 2002; United Nations Framework Convention on Climate Change, 2004).

For estimating soil C budgets, Sweden is developing a simulation-based approach, using the Introductory Carbon Balance Model (ICBM) (Andrén and Kätterer, 1997, 2001). This two-pool model has been calibrated using long-term field data and has been incorporated into a regional framework to enable estimates of soil C emissions and removals for national reporting (Andrén et al., 2004). The model is conceptually simple and, with suitable input data (i.e., annual agricultural statistics, daily weather data, climate region, soil type, and crop type, etc.), it can be run and optimized in a conventional spreadsheet program (Andrén et al., 2003, 2004). This model approach is still in the testing phase, and currently only the emissions from organic soils are reported in the NIR.

United Kingdom
In the United Kingdom, about 47% of the land area is used for intensive crop and pasture production while about 30% comprises less intensively managed grazing systems (Defra, 2001).

Soil C changes are estimated using a matrix of land use change (derived from land surveys), linked to a dynamic empirical model of C gain or loss. The model is conceptually similar to the IPCC approach, with the important difference that changes in C stocks over time are modeled as nonlinear, using an exponential function. Soil C changes with time, for a particular land use transition, are estimated as shown in Eq. [3]:

[3]

where C_t is the C stock at time t (Mg C ha^–1); C₀ is the initial C stock (Mg C ha^–1); C_f is the equilibrium C stock under the new land use (Mg C ha^–1); k is the specific rate of C change (yr^–1); and t is the time period (year).

For example, if the inventory year is 1990 and a base year of 1930 is chosen to represent the initial C₀ stock, then the total soil C loss or gain from 1930–1990 is estimated using Eq. [4].

[4]

where A_T is the area under the particular land use transition (ha).

Negative values of X₁₉₉₀ indicate removals (gains) in C, while positive values represent C loss. Similarly, the calculation can be made over the interval 1930–1989, as shown in Eq. [5].

[5]

The net change in C in 1990 is the difference between the estimated values from above Eq. [4] and [5] (i.e., F₁₉₉₀ = X₁₉₉₀ – X₁₉₈₉).

To apply the model, data is required to estimate the change in equilibrium C stocks from the initial to the final land use during a transition. These are calculated for each land use category, as area-weighted averages by major soil types, by countries (i.e., Scotland, England, and Wales) within the United Kingdom. Mean changes in equilibrium soil C stocks are calculated as shown in Eq. [6]:

[6]

where i is the initial land use; j is the new land use; c is the country; s is the soil group; C_sijc is the change in equilibrium soil C for a specific land use transition; and L_sijc is the land area by soil type for a specific land use transition.

The rate of C change depends on the type of land use transition. For a transition where C is lost, a "fast" specific rate constant (i.e., k) is applied, and for transitions where C is gained, a "slow" specific rate is applied; time ranges relevant to complete transitions were selected using a literature search on measured rates of soil C, and expert judgment (United Nations Framework Convention on Climate Change, 2004; R. Milne, personal communication, 2004). Land use change matrices for the periods 1947–1980 and 1984–1990 have been used in applications to date (United Nations Framework Convention on Climate Change, 2004). The C stock changes reported for the NIR (United Nations Framework Convention on Climate Change, 2004) using this methodology include means and estimates of uncertainty based on a Monte Carlo approach, computed separately for England, Scotland, and Wales. For Northern Ireland, C stock estimates have been made using an IPCC-based method, as currently no land-use change matrix is available for the country (United Nations Framework Convention on Climate Change, 2004).

The DNDC model has also been used to estimate N₂O emissions from United Kingdom agricultural soils. United Kingdom–specific, county-level soil characteristics, daily climate, crops, livestock, and farming practices had been used as input data. Model validation had been done using available, but limited, field data. To be consistent with the IPCC approach, emission factors calculated using model estimates had been used along with the activity data for different source categories to estimate N₂O emissions. The DNDC-based N₂O estimates from indirect emissions and agricultural practices (excluding animal waste during storage) were about 40% lower than the estimates made using the IPCC default method (Brown et al., 2002).

United States
In the United States, arable land covers 19% of total land area with an additional 6% in intensive pasture and 21% in rangeland. Although cropland area has remained relatively stable over the last century (Lal et al., 1998), recent trends show a 12% decline in cropland over the past 20 yr and other significant changes in land use and management continue to occur (National Resources Inventory, 2002). Important trends affecting agricultural soils include set-aside of marginal lands in conservation reserves, reductions in tillage intensity, and increases in cropping intensity (Ogle et al., 2003a).

Currently, the U.S. estimates soil C stock changes using a modified version of the IPCC default methodology with U.S.-specific reference C stocks and stock change factors (Ogle et al., 2003a). Activity data are stratified by IPCC-defined climate and soil types. A comprehensive national database, the National Resources Inventory (NRI), is the primary source of land use and management data. NRI records land use, crop type, and other information (e.g., irrigation, pasture improvement, soil type) on more than 400 000 permanent inventory points on agricultural land. Surveys have been conducted on 5-yr intervals (1982–1997), although currently the NRI is transitioning to an annual collection of data on a subset of inventory points. Supplemental data including county-level tillage practices (Conservation Technology Information Center, 1998), fertilizer use (USDA Economic Research Service, 2003), and manure production (Edmonds et al., 2003) are included in the inventory. A Monte Carlo approach is used to estimate 95% confidence intervals of stock changes for each climate–soil combination (Ogle et al., 2003a, 2003b).

A more advanced simulation approach using the Century model is being developed to estimate soil C emissions and removals. Annual changes are computed dynamically as a function of inputs of C to soil (e.g., crop residues, manure) and C emissions from organic matter decomposition, which are governed by climate and soil factors as well as management practices. The model simulates all major field crops (maize, wheat, and other small grains, soybean, sorghum, cotton) as well as hay and pasture (grass, alfalfa, clover). The same sources of input data as in the IPCC-based methodology are used for management variables, included tillage, fertilization, irrigation, drainage, and manure addition. Preliminary results predict that cropland mineral soils are a net sink of about 21 Tg yr^–1, which is higher than the estimates (11 Tg yr^–1) using the IPCC approach (USEPA, 2004). Both methods attribute C gains to conservation set-aside and reduced tillage, but the simulation approach also accounts for the long-term trend of increasing crop productivity, which is not captured by the IPCC method.

Soil N₂O emissions were previously estimated using the IPCC Tier 1 methodology, with activity data derived from county-level databases on mineral fertilizer use, animal manure use, crop residues and N-fixing crops, sewage sludge application, and grazing animals (USEPA, 2004). However, the United States has developed a simulation-based approach using the DAYCENT model (Parton et al., 1998; DelGrosso et al., 2001) to estimate N₂O emissions from agricultural soils. This approach has advantages over the empirical IPCC method in that it can better capture the interaction between different management conditions, including fertilization and manuring practices, soils, and varying climate. A major challenge is deriving activity data (such as synthetic fertilizer and manure nitrogen inputs) at a suitable spatial scale, since existing fertilizer use databases are aggregated at the country-level. Preliminary estimates using the dynamic method are 10 to 15% lower than with the IPCC method and with greater interannual variability, due to the inclusion of weather effects in the simulation approach (S. DelGrosso, personal communication, 2005). The largest factor accounting for the difference is lower emissions from N-fixing crops estimated by DAYCENT compared with the IPCC default method.

	DISCUSSION AND CONCLUSIONS

TOP ABSTRACT INTRODUCTION IMPORTANCE OF AGRICULTURAL SOILS... METHODOLOGIES FOR ESTIMATING... DISCUSSION AND CONCLUSIONS REFERENCES

 
Agricultural soils are a significant source of N₂O emissions for all Annex 1 countries and an important CO₂ source–sink category for many of them. However, compared to some other sectors of country GHG reporting, reporting of soil emissions and removals is, at present, generally less comprehensive and less uniformly applied across countries. Even for Annex 1 countries, Tier 1 methods and default emission factors are the most widely used alternatives (Fig. 5 and 6).

View larger version (10K): [in this window] [in a new window]   Fig. 5. Methods used for estimating N2O emissions by Annex 1 countries based on the National Inventory Reports (NIRs) and Common Reporting Formats (CRFs) submitted in 2004. T1, Tier 1; T2, Tier 2, NR/NO, not reported or not occurring.

View larger version (13K): [in this window] [in a new window]   Fig. 6. Methods used for estimating CO2 emissions by Annex 1 countries based on the National Inventory Reports (NIRs) and Common Reporting Formats (CRFs) submitted in 2004. Mineral, stock changes in C from mineral soils; organic, emissions from cultivated organic soils; liming, emissions due to liming; T1, Tier 1; T2, Tier 2; T3, Tier 3; NR/NO, not reported or not occurring.

Despite these challenges, a number of countries have successfully implemented soil emission inventories and several are in the process of developing advanced, computational and data intensive methods, tailored to national circumstances (Tier 3).

The IPCC default approach has the advantages of having a relatively simple structure as well as providing default emission and stock change factors, so that the main requirement for individual countries is to obtain suitable activity data. To facilitate this, the methods were designed to work with globally available data sets, at a minimum. However, with simplification there are tradeoffs in the form of increased uncertainty, particularly with the application of global default values. Inherently, globally averaged emission factors will be subject to error when applied to a particular country or region having conditions different than the global mean. In addition, global defaults for soil processes are likely biased in that most of the data used in their derivation are from temperate locations, where the preponderance of field research has been done. Hence, tropical conditions are often underrepresented. Finally, the use of highly aggregated data (e.g., national totals for soil N input sources) results in a loss of information about subregional (within country) differences in sources and sinks of greenhouse gases.

For both soil N₂O and CO₂, the processes controlling emissions and removals are highly influenced by spatially and temporally varying conditions such as temperature, soil moisture, and soil chemical and physical properties. These differences are unlikely to be adequately captured when national-level aggregate data are used. Additional limitations to the IPCC method for mineral soil C changes include the lack of inclusion of soil erosion and transport and restriction of C stock changes to the top 30 cm of soil. For organic soils, the data available for estimation of emission factors is quite limited and hence uncertainty is higher than for the analogous stock change factors for mineral soils (Intergovernmental Panel on Climate Change, 2003).

Country-specific approaches are being rapidly developed to overcome some of these deficiencies and improve estimates of soil-derived emissions. For soil C estimation, elaboration of the IPCC method using country-specific stock change and reference C stocks, along with other modifications, have been implemented in New Zealand and the United States. The United Kingdom approach is functionally similar, but includes a nonlinear change with time from an initial to a new equilibrium C stock (whereas the IPCC default assumes a linear transition). Fully dynamic approaches employing simulation models and detailed activity data are being developed in Australia, Austria, Canada, Sweden, and the United States.

For N₂O, nearly all countries continue to utilize the IPCC base methodology, although dynamic simulation approaches have been implemented in Germany and the United States, and are under development in Australia (Australian Greenhouse Office, 2002b) and Canada.

At present, few countries have estimated inventories using both advanced Tier 3 methods and simpler (Tier 1 or Tier 2) IPCC-default methods. Hence it is difficult to make a general assessment of the adequacy of the more simple IPCC approach compared to more sophisticated inventories. Austria reported a difference of 13% in soil C emissions and removals between simple and advanced inventory methods, and N₂O emissions in the U.S. inventory differed by 10 to 15% comparing IPCC default and simulation model-based approaches. Soil C values in New Zealand estimated using CMS were slightly higher or lower compared to IPCC default C values, for different soil–climate categories. In the United Kingdom, DNDC-based N₂O estimates were 40% lower compared to the estimates from IPCC default method. Although advanced methodologies have been developed for inventorying soil C stock changes in Australia, Canada, Sweden, and United Kingdom, so far no study has been done on how those estimates compare with the estimates from the IPCC method.

The higher-tier methods developed by certain Annex 1 countries are designed to be more representative of country-specific soil, climate, and management conditions, and in most cases, these methods are applied at a finer spatial scale than is used for a Tier 1 approach, Hence, the inventory results obtained are expected to be more accurate and with lesser uncertainty than when using the default method. However, at present there have been few instances of rigorous uncertainty assessments applied to either IPCC default (Tier 1) inventories or Tier 3 country-specific methods. Few countries have measurement networks, which can provide independent validation inventory estimates (Ogle and Paustian, 2005). Hence a "head-to-head" comparison of different inventory methods and tier levels, with regard to accuracy and uncertainty, is not currently possible. Hopefully as more countries develop and implement more advanced soil GHG emission and removal inventories, supplementary estimates will be made using IPCC methods to provide an evaluation of the reliability of continuing the more general IPCC approaches, which are likely to remain the dominant method for most developing countries, at least for the near future.

	ACKNOWLEDGMENTS

 
We are grateful for assistance provided by the inventory specialists and national greenhouse offices from many countries in preparing this review. Support for this work was provided by the Consortium for Agricultural Soil Mitigation of Greenhouse Gases (CASMGS) funded by the Cooperative State Research, Education, and Extension Service, USDA, under Agreement no. 2001-38700-1109.

	REFERENCES

TOP ABSTRACT INTRODUCTION IMPORTANCE OF AGRICULTURAL SOILS... METHODOLOGIES FOR ESTIMATING... DISCUSSION AND CONCLUSIONS REFERENCES

 

• Aber, J.D., P.B. Reich, and M.L. Goulden. 1996. Extrapolating leaf CO₂ exchange to the canopy: A generalized model of forest photosynthesis compared with measurements by eddy correlation. Oecologia 106:257–265.[CrossRef]
• Acosta, R., S. Pesmajoglou, J. Grabert, and R. Lichte. 2002. Technical review of national greenhouse gas inventories. Paper presented at the 11th International Emission Inventory Conference, Emission Inventories—Partnering for the Future, Atlanta. 15–18 Apr. 2002. To obtain a copy, contact the USEPA Technology Transfer Network, Clearinghouse for Inventories & Emission Factors (Chief_Info{at}epamail.epa.gov).
• Andrén, O., and T. Kätterer. 1997. ICBM—The Introductory Carbon Balance Model for exploration of soil carbon balances. Ecol. Appl. 7(4):1226–1236.
• Andrén, O., and T. Kätterer. 2001. Basic principles of soil carbon sequestration and calculating dynamic country-level balances including future scenarios. p. 495–511. In R. Lal et al. (ed.) Assessment methods for soil carbon. Lewis Publ., Boca Raton, FL.
• Andrén, O., T. Kätterer, and T. Karlsson. 2003. Carbon balances in Swedish agricultural soils: Improving IPCC methodology with limited resources. p. 127–132. In C.A.S. Smith (ed.) Soil organic carbon and agriculture: Developing indicators for policy analyses. Proc. OECD Expert Meeting, Ottawa, ON, Canada. Agriculture and Agri-Food Canada, Ottawa, and Organisation for Economic Co-operation and Development, Paris.
• Andrén, O., T. Kätterer, and T. Karlsson. 2004. ICBM regional model for estimations of dynamics of agricultural soil carbon pools. Nutr. Cycling Agroecosyst. 70:231–239.[CrossRef]
• Australian Greenhouse Office. 2002a. Australia's third national communication on climate change: A report under the United Nations Framework Convention on Climate Change. Australian Greenhouse Office, Canberra.
• Australian Greenhouse Office. 2002b. Greenhouse gas emissions from land use change in Australia: An integrated application of the national carbon accounting system. National Carbon Accounting System technical report. Australian Greenhouse Office, Canberra.
• Brown, L., B. Syed, S.C. Jarvis, R.W. Sneath, V.R. Phillips, K.W.T. Goulding, and C. Li. 2002. Development and application of a mechanistic model to estimate emission of nitrous oxide from UK agriculture. Atmos. Environ. 36:917–928.[CrossRef]
• Bundesministerium für Umwelt, Naturschutz und Reaktorsicherheit. 2002. Third report by the Government of the Federal Republic of Germany in accordance with the Framework Convention of the United Nations. Federal Ministry for the Environment, Nature Conservation and Nuclear Safety, Berlin.
• Butterbach-Bahl, K., M. Kesik, P. Miehle, and H. Papen. 2002. Calculation of regional budgets of N-trace gas emissions from soils using mechanistic models: Results from a case study for Saxony, Germany. p. 119–124. In J. Van Ham et al. (ed.) Non-CO₂ greenhouse gases. Mill Press, Rotterdam, the Netherlands.
• Butterbach-Bahl, K., M. Kesik, P. Miehle, H. Papen, and C. Li. 2004. Quantifying the regional source strength of N-trace gases across agricultural and forest ecosystems with process based models. Plant Soil 260:311–329.[CrossRef]
• Butterbach-Bahl, K., F. Stange, and H. Papen. 2001. Regional inventory of nitric oxide and nitrous oxide emissions for forest soils of southeast Germany using the biogeochemical model PnET-N-DNDC. J. Geophys. Res. [Atmos.] 106(D24):34155–34166.[CrossRef]
• Cole, V., C. Cerri, K. Minami, A. Mosier, N. Rosenberg, D. Sauerbeck, J. Dumanski, J. Duxbury, J. Freney, R. Gupta, O. Heinemeyer, T. Kolchugina, J. Lee, K. Paustian, D. Powlson, N. Sampson, H. Tiessen, M. Van Noordwijk, and Q. Zhao. 1996. Agricultural options for mitigation of greenhouse gas emissions. p. 745–771. In R.T. Watson et al. (ed.) Climate Change 1995. Impacts, adaptations and mitigation of climate change: Scientific–technical analyses. IPCC Working Group II. Cambridge Univ. Press, Cambridge.
• Coleman, K., and D.S. Jenkinson. 1995. RothC-26.3—A model for the turnover of carbon in soil: Model description and users guide. Lawes Agricultural Trust, Harpenden, UK.
• Conservation Technology Information Center. 1998. National Crop Residue Management Survey. Available at http://www.ctic.purdue.edu/CTIC/CRM.html (verified 3 Jan. 2006). CTIC, West Lafayette, IN.
• Defra. 2001. The UK's Third National Communication under the United Nations Framework Convention on Climate Change. Department for Environment, Food and Rural Affairs, London.
• Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel. 2001. Simulated interaction of carbon dynamics and nitrogen trace gas fluxes using the DAYCENT model. p. 303–330. In M.J. Schaffer et al. (ed.) Modeling carbon and nitrogen dynamics for soil management. Lewis Publ., Boca Raton, FL.
• Edmonds, L., R.L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J. Schaefer. 2003. Costs associated with development and implementation of Comprehensive Nutrient Management Plans. Part I—Nutrient management, land treatment, manure and wastewater handling and storage, and recordkeeping. Available at www.nrcs.usda.gov/technical/land/pubs/cnmp1.html (verified 3 Jan. 2006). USDA-NRCS, Washington, DC.
• EMEP/CORINAIR. 2002. EMEP/CORINAIR atmospheric emissions inventory guidebook. Tech. Rep. 30. European Environment Agency, Copenhagen.
• EMEP/CORINAIR. 2004. EMEP/CORINAIR emissions inventory guidebook. Tech. Rep. 30. European Environment Agency, Copenhagen.
• Environment Canada. 2001. Canada's Third National Report on Climate Change: Actions to meet commitments under the United Nations Framework Convention on Climate Change. Government of Canada, Ottawa, ON.
• Gerzabek, M.H., F. Strebl, and M. Tulipan, and S. Schwarz. 2003. Quantification of carbon pools in agriculturally used soils of Austria by use of a soil information system as basis for the Austrian Carbon Balance Model. p. 73–78. In C.A.S. Smith (ed.) Soil organic carbon and agriculture: Developing indicators for policy analyses. Proc. OECD Expert Meeting, Ottawa, ON, Canada. Agriculture and Agri-Food Canada, Ottawa, and Organisation for Economic Co-operation and Development, Paris.
• Intergovernmental Panel on Climate Change. 1996. Climate Change 1995. Impacts, adaptations, and mitigation of climate change: Scientific–technical analyses. Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge Univ. Press., Cambridge.
• Intergovernmental Panel on Climate Change. 1997a. Revised 1996 IPCC guidelines for national greenhouse gas inventories: Reporting instructions (Volume 1). IPCC, Meteorological Office, Bracknell, UK.
• Intergovernmental Panel on Climate Change. 1997b. Revised 1996 IPCC guidelines for national greenhouse gas inventories: Workbook (Volume 2). IPCC, Meteorological Office, Bracknell, UK.
• Intergovernmental Panel on Climate Change. 1997c. Revised 1996 IPCC guidelines for national greenhouse gas inventories: Reference manual (Volume 3). IPCC, Meteorological Office, Bracknell, UK.
• Intergovernmental Panel on Climate Change. 2000a. Land use, land-use change, and forestry: A special report. Cambridge Univ. Press., Cambridge.
• Intergovernmental Panel on Climate Change. 2000b. Good practice guidance and uncertainty management in national greenhouse gas inventories. IPCC/IGES, Hayama, Japan.
• Intergovernmental Panel on Climate Change. 2003. Good practice guidance for land use, land-use change and forestry. IPCC/IGES, Hayama, Japan.
• Kroeze, C. 1994. Nitrous oxide (N₂O). Emission inventory and options for control in the Netherlands. RIVM Rep. 773001 004. RIVM, Natl. Inst. of Public Health and the Environ., Bilthoven, the Netherlands.
• Lal, R. 2004. Agricultural activities and the global carbon cycle. Nutr. Cycling Agroecosyst. 70:103–116.[CrossRef]
• Lal, R., J. Kimble, R.F. Follet, and C.V. Cole. 1998. The potential for U.S. cropland to sequester carbon and mitigate greenhouse effect. Ann Arbor Press, Chelsea, MI.
• Li, C.S. 2000. Modeling trace gas emissions from agricultural ecosystems. Nutr. Cycling Agroecosyst. 58:259–276.
• Li, C.S., J. Aber, F. Stange, K. Butterbach-Bahl, and H. Papen. 2000. A process-oriented model of N₂O and NO emissions from forest soils: 1. Model development. J. Geophys. Res. [Atmos.] 105(D4):4369–4384.[CrossRef]
• McConkey, B.G., C.M. Monreal, T. Huffman, G.T. Patterson, J.A. Brierley, R.L. Desjardins, P. Rochette, M.R. Carter, M. Bentham, and S. Gameda. 2003. National carbon and greenhouse-gas emission accounting and verification system for agriculture (NCGAVS). p. 79–92. In C.A.S. Smith (ed.) Soil organic carbon and agriculture: Developing indicators for policy analyses. Proc. OECD Expert Meeting, Ottawa, ON, Canada. Agriculture and Agri-Food Canada, Ottawa, and Organisation for Economic Co-operation and Development, Paris.
• National Resources Inventory. 2002. National Resources Inventory 2002 annual NRI. Available at www.nrcs.usda.gov/technical/land/nri02/nri02lu.html (verified 3 Jan. 2006). USDA-NRCS, Washington, DC.
• New Zealand Ministry for the Environment. 2001. New Zealand's Third National Communication under the Framework Convention on Climate Change. Ministry for the Environment, Wellington.
• Ogle, S.M., J. Breidt, M.D. Eve, and K. Paustian. 2003a. Uncertainty in estimating land use and management impacts on soil organic carbon storage for US agricultural lands between 1982 and 1997. Glob. Change Biol. 9:1521–1542.[CrossRef]
• Ogle, S.M., J. Breidt, M.D. Eve, and K. Paustian. 2003b. Uncertainty in soil organic carbon storage for US agroecosystems due to land use and management between 1990 and 1997. p. 133–143. In C.A.S. Smith (ed.) Soil organic carbon and agriculture: Developing indicators for policy analyses. Proc. OECD Expert Meeting, Ottawa, ON, Canada. Agriculture and Agri-Food Canada, Ottawa, and Organisation for Economic Co-operation and Development, Paris.
• Ogle, S.M., F.J. Breidt, and K. Paustian. 2005. Agricultural management impacts on soil organic carbon storage under moist and dry climatic conditions of temperate and tropical regions. Biogeochemistry 72:87–121.[CrossRef]
• Ogle, S.M., and K. Paustian. 2005. Soil organic carbon as an indicator of environmental quality at the national scale: Inventory monitoring methods and policy relevance. Can. J. Soil Sci. 85:531–540.
• Orthofer, R., E. Gebetsroither, F. Strebl, H. Schwaiger, G. Jungmaier, J. Spitzer, L. Canella, G. Hintermeier, A. Windsperger, and F. Schneider. 2000. The Austrian Carbon Balance Model (ACBM) final project report. Austrian Research Centers.
• Parton, W.J., M. Hartman, D. Ojima, and D. Schimel. 1998. DAYCENT and its land surface submodel: Description and testing. Global Planet. Change 19:35–48.
• Paustian, K., H.P. Collins, and E.A. Paul. 1997. Management controls on soil carbon. p. 15–49. In E.A. Paul et al. (ed.) Soil organic matter in temperate agroecosystems. CRC Press, New York.
• Paustian, K., J. Six, E. Elliott, and H.W. Hunt. 2000. Management options for reducing CO₂ emissions from agricultural soils. Biogeochemistry 48:147–163.[CrossRef]
• Richards, G.P. 2001. The FullCAM Carbon Accounting Model: Development, calibration and implementation for the National Carbon Accounting System. Natl. Carbon Accounting System Tech. Rep. 28. Australian Greenhouse Office, Canberra.
• RIVM. 2002. Greenhouse gas emissions in the Netherlands 1990–2000. National Inventory Report 2002. RIVM Rep. 773201 006. National Institute of Public Health and the Environment (RIVM), Bilthoven, the Netherlands.
• Schlesinger, W.H. 1997. Biogeochemistry: An analysis of global change. Academic Press, San Diego, CA.
• Scott, N.A., K.A. Tate, D.J. Giltrap, C.T. Smith, R.H. Wilde, P.F.J. Newsome, and M.R. Davis. 2002. Monitoring land-use change effects on soil carbon in New Zealand: Quantifying baseline soil carbon stocks. Environ. Pollut. 116:S167–S186.
• Smith, P. 2004. Carbon sequestration in croplands: The potential in Europe and the global context. Eur. J. Agron. 20:229–236.