CC BY-NC-ND
0ARTICLE IN PRESS
ENEECO-03040; No of Pages 10
Energy Economics xxx (2015) xxx-xxx
Contents lists available at ScienceDirect
Energy Economics
journal homepage: www.elsevier.com/locate/eneco
Agriculture, forestry, and other land-use emissions in Latin America
Katherine V. Calvin a,*< Robert Beach b, Angelo Gurgelc, Maryse Labrietd, Ana Maria Loboguerrero Rodriguez e
a Pacific Northwest National Laboratories, Joint Global Change Research Institute, College Park, MD, USA b Research Triangle Institute, Research Triangle Park, NC, USA
c Sao Paulo School of Economics, Fundaçao Getûlio Vargas (EESP/FGV), Brazil/MITJoint Program on the Science and Policy of Climate Change, USA d Eneris Environment Energy Consultants, Madrid, Spain
e CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS ),International Center for Tropical Agriculture, Cali, Colombia
ARTICLE INFO
ABSTRACT
Article history:
Received 29 August 2014
Received in revised form 15 March 2015
Accepted 30 March 2015
Available online xxxx
JEL classification:
Nearly 40% of greenhouse gas (GHG) emissions in Latin America were from agriculture, forestry, and other land use (AFOLU) in 2008, more than double the global fraction of AFOLU emissions. In this article, we investigate the future trajectory of AFOLU GHG emissions in Latin America, with and without efforts to mitigate, using a multimodel comparison approach. We find significant uncertainty in future emissions with and without climate policy. This uncertainty is due to differences in a variety of assumptions including (1) the role of bioenergy, (2) where and how bioenergy is produced, (3) the availability of afforestation options in climate mitigation policy, and (4) N2O and CH4 emission intensity. With climate policy, these differences in assumptions can lead to significant variance in mitigation potential, with three models indicating reductions in AFOLU GHG emissions and one model indicating modest increases in AFOLU GHG emissions.
© 2015 Battelle Memorial Institute and The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords:
Agriculture and land use Greenhouse gas emissions Latin America
1. Introduction
Globally, 42.6 PgCO2e of greenhouse gas (GHG) emissions were emitted in 2008; 81% of these emissions were from energy combustion and industrial processes (Fig. S1; EC, 2011).1 Latin America accounted for a mere 7% of global GHG emissions. However, nearly 40% of GHG emissions in Latin America (Fig. S2) were from agriculture, forestry, and other land use (AFOLU) in 2008, more than double the global fraction of AFOLU emissions. From 2005 to 2008, AFOLU emissions in Latin America declined dramatically due to a reduction in AFOLU CO2 emissions in Brazil (Fig. S3) as a result of stringent policies to reduce deforestation. An open question remains as to whether these declines will continue or if emissions will begin to rise again.
* Corresponding author.
E-mail address: katherine.calvin@pnnl.gov (K.V. Calvin).
1 The numbers quoted in this paragraph and in Figs. S1-S3 are from the EDGAR data set (EC, 2011). There is significant uncertainty in historical emissions, particularly of AFOLU CO2. For more information on emissions uncertainty, see Blanco et al. (2014).
In this article, we investigate the future trajectory of AFOLU GHG emissions in Latin America, with and without efforts to mitigate, using a multi-model comparison approach. This work builds on the work of Rose et al. (2012), which examines mitigation potential at a global level in a multi-model framework, and on the work of Smith et al. (2014a), which compiles bottom-up estimates of mitigation potential. While many recent papers examine the role of land in mitigation regimes, they have focused their attention on land transitions (e.g., Popp et al., 2014), bioenergy (e.g., Calvin et al., 2013; Rose et al., 2014), or trade-offs between different land policy schemes (e.g., Calvin et al., 2014; Reilly et al., 2012; Wise et al., 2009) at the global level. This paper expands on these efforts by examining the emission consequences and mitigation potential of land transitions in a particular region (Latin America).
Section 2 discusses the models and scenarios included in this article. Section 3 examines the AFOLU GHG emissions in Latin America absent any climate mitigation efforts. Section 4 discusses potential mitigation options and how they influence emissions under climate policy in Latin America. Section 5 provides some discussion, concluding remarks, and areas for future research.
http: //dx.doi.org/10.1016/j.eneco.2015.03.020
0140-9883/© 2015 Battelle Memorial Institute and The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
ARTICLE IN PRESS
K.V. Calvin et al. /Energy Economics xxx (2015) xxx-xxx
2. Methodology
This article utilizes the models and scenarios developed for the CLIMACAP-LAMP project2 to assess AFOLU GHG emissions in Latin America. For more information on the project and its scenarios, we refer the reader to the introductory article of this special issue (van der Zwaan et al., this issue). In this section, we describe the models and scenarios included in this study.
2.1. Models
Several approaches to modeling AFOLU in economic and integrated assessment models exist. Some models exclude the sector entirely, either explicitly or implicitly assuming that AFOLU GHG emissions are zero. Some models include AFOLU by parameterizing functions (e.g., bioenergy supply curves and AFOLU GHG marginal abatement cost curves) to other offline models or studies. These models often include limited feedbacks. For example, an expansion in bioenergy consumption in these models may not change GHG emissions, if both elements are included through separate, non-interacting functions, or if bioenergy is assumed to be sustainably grown and therefore carbonfree. A third type of model includes a structural representation of the agriculture and land sector, ensuring consistency between production, consumption, and emissions. In this article, we focus our analysis on the second and third types of models (see Table 1). The model descriptions included in this paper are focused on the treatment of AFOLU and AFOLU GHGs. For more information on these models, we refer to the reader to publications developed by their respective modeling teams: ADAGE (Ross, 2009); EPPA (Paltsev et al., 2005); GCAM (Calvin et al., 2011) and TIAM-WORLD (Loulou, 2008; Loulou and Labriet, 2008).
22. ADAGE
The Applied Dynamic Analysis of the Global Economy (ADAGE) model is a multi-region, multi-sector dynamic computable general equilibrium (CGE) model (Ross, 2009). The version of ADAGE used for the current study is a recursive dynamic version focused on the agricultural sector. It includes disaggregation of individual major agricultural crops and bioenergy feedstocks as well as incorporation of land as a factor of production with tracking of land cover and land use in terms of physical area (Beach et al., 2011). Land cover categories included are cropland, pasture, managed forests, unmanaged forests, natural grassland, and other land. Land conversion is modeled using a nested constant elasticity of substitution (CES) function and explicitly incorporates costs of land conversion as well as land supply elasticities. Marginal conversion costs are assumed to be equal to the difference in value between land types while land supply elasticities are based on historically observed rates of land conversion. The key database used in this study is the Global Trade Analysis Project (GTAP) data base version 7.1 (Narayanan and Walmsley, 2008) which comprises 57 sectors and 112 regions, corresponding to the global economy in 2004. Because there are no explicit sectors for biofuels and their respective feedstock crops and by-products in the GTAP database, we incorporated these sectors by splitting the relevant existing sectors. The final database includes disaggregated sectors such as corn, soybeans, rapeseed-mustard, palm-kernel, sugarcane, and sugar beets; biofuels categories such as corn ethanol, wheat ethanol, sugarcane ethanol, sugar beet ethanol, soy biodiesel, rapeseed biodiesel, palm oil biodiesel, and major byproducts of biofuels production such as dried distillers' grains with
2 The Integrated Climate Modelling and Capacity Building Project in Latin America (CLIMACAP) is a European Commission funded effort focused on analyzing the effects of mitigation strategies in key Latin American Countries. The Latin American Modeling Project (LAMP) is a similar effort funded by the U.S. Environmental Protection Agency and the U.S. Agency for International Development. The projects are collaborating to develop a multi-model comparison project focused on mitigation in Latin America. More information on the projects is available at: https://tntcat.iiasa.ac.at/CLIMACAP-LAMPDB/.
solubles (DDGS) and oilseed meals. The modified GTAP data base was aggregated to 8 regions and 36 sectors and updated to the model baseline year 2010 using secondary data on energy, biofuels, agriculture, and livestock sectors from secondary data sources including the Food and Agricultural Organization (FAO), International Energy Agency (IEA), U.S. Department of Agriculture (USDA), Energy Information Administration (EIA), U.S. Department of Energy (DOE), and others. GHG emissions from all sources are included in ADAGE, along with opportunities for GHG mitigation.
CO2 emissions from fuel use are tied directly to the quantity of each category of fossil fuel combusted. Options for fuel substitution in production and household energy consumption are controlled by the model's CES nesting structure and substitution elasticities. Non-CO2 emissions enter the production function as an input. Sector-specific abatement cost curves are implemented through elasticities of substitution between each GHG and all other inputs calibrated based on marginal abatement cost curves (EPA, 2006, 2013). Emissions from land use change are calculated by multiplying the area of land conversion by the difference in carbon sequestration (above and below ground vegetative carbon and soil carbon) provided by the two land types multiplied by IPCC default emissions factors for land use change (IPCC, 2006). As a result, changes in carbon stock occur immediately following a land conversion. ADAGE calculates projected global and regional economic production, energy use, agriculture activity, biofuel production, land use change and greenhouse gas emissions from all sources from 2010 to 2050 at 5-year time steps. Latin America is represented within the current ADAGE model by Brazil and an aggregated region of all other countries in Latin America.
23. EPPA
EPPA is a multi-region, multi-sector recursive-dynamic CGE model of the global economy (Paltsev et al., 2005). Latin America is represented in EPPA by Mexico, Brazil, and an aggregated region of all the other countries in Latin America. The model calculates emissions of greenhouse gases (CO2, CH4, N2O, HFCs, PFCs and SF6) and other pollutants, and also represents abatement and mitigation policies, including gas-specific control measures (Hyman et al., 2003). The agriculture activities in the model are crops, livestock and forestry, plus regional specific biofuels crops. Land use categories are cropland, pasture, managed forest, natural forest and natural grass. Natural vegetation is incorporated explicitly considering their "non-use" value in the utility function. EPPA considers competition among land use categories by considering that farmers can transform one land category to others if they are able to cover explicitly the costs of conversion. This approach implies that intensively managed land can be "produced" from less intensively or unmanaged land, and also that farmland can be abandoned. The conversion of natural vegetation in EPPA is limited by the observed land supply response in the last two decades (Melillo et al., 2009). It mimics the increasing costs associated to larger deforestation in a single period and the additional institutional costs, as environmental legislation and consumer pressures to conservationism. Land use changes in EPPA operates on a per country level, but it is connected with the Terrestrial Ecosystem Model - TEM (Felzer et al., 2004) to distribute EPPA's land-use predictions by 0.5° grid cell level based on climate, soil and economic information.
2.4. GCAM
GCAM is a global integrated assessment model, coupling representations of the economy, energy system, agriculture and land use system, and climate system. The model operates in five-year time steps from 1990 to 2100. GCAM disaggregates Latin America into seven regions (Argentina, Brazil, Colombia, Mexico, Central America and the Caribbean, Northern South America, and Southern South America). The agriculture and land use component of GCAM further disaggregates these
ARTICLE IN PRESS
K.V. Calvin et al. / Energy Economics xxx (2015) xxx-xxx
Table 1
Comparison of model methodology.
TIAM-WORLD
Land model
AFOLU emissions CO2 CH4
Policy options Afforestation
Reduced deforestation Bioenergy constraints/taxes
Land restrictions
Structural, using transformation functions with observed land supply responses
Changes in above & below ground carbon Drivers explicitly represented; additional mitigation available through MAC curves
Drivers explicitly represented; additional mitigation available through MAC curves
Structural, using transformation Structural, assuming a
Available, but not used Included
Available, but not used
Only through reduced deforestation policies
functions with observed land supply responses
Changes in above & below ground carbon
Drivers explicitly represented
Drivers explicitly represented
Available, but not used Included
Available, but not used
Only through reduced deforestation policies
market equilibrium with profit maximizing farmers
Changes in above & below ground carbon Drivers explicitly represented; additional mitigation available through MAC curves
Drivers explicitly represented; additional mitigation available through MAC curves
Available, but not used Included
Available, but not used
Only through reduced deforestation policies
Parameterized Functions, based on results from a structural land model
Exogenous, based on results from the MagPIE model (Labrietet al., 2013) Exogenous in Reference Case, based on results from MAgPIE model; endogenous mitigation available through MAC curves (up to 20% of annual CH4 emissions from AFOLU can be reduced at a cost of up to 500 US$/tCO2-eq) Exogenous in the Reference Case based on results from MAgPIE model; endogenous mitigation available through MAC curves (up to 20% of annual N2O emissions from AFOLU can be reduced at a cost of up to 500 US$/tCO2)
Based on results from MAgPIE model
(Labriet et al., 2013)
Based on results from MAgPIE model
(Labriet et al., 2013)
Bioenergy is constrained through the
bioenergy supply curve.
Only through changes in bioenergy
supply curves or changes in afforestation
and reduced deforestation MAC curves
regions into as many as 18 sub-regions based on agro-ecological zones (Monfreda et al., 2009), resulting in 283 land supply regions. GCAM is a market-equilibrium model, adjusting prices until supplies and demands of all products are balanced. Land allocation in GCAM is determined based on relative profitability and agricultural supply is determined based on exogenously specified yields and endogenously calculated land allocation (Wise et al., 2014). The model includes several types of bioenergy, including first generation, cellulosic, agricultural and forestry residues, and municipal solid waste. GCAM computes anthropogenic emissions of 16 greenhouse gases and short-lived species, including CO2 from land-use change, N2O from fertilizer use, and CH4 from the production of agriculture and livestock. Land-use change CO2 emissions are computed assuming that increases in vegetation carbon and changes in soil carbon occur over several years to decades after a land use change occurs, with the exact time depending on the region and land type (see Calvin et al., 2011 for more information). More information on the GCAM model is available at wiki.umd.edu/gcam.
2.5. TIAM-WORLD
TIAM-WORLD is a global and technology rich integrated assessment model of the TIMES family (Loulou, 2008; Loulou and Labriet, 2008), computing an inter-temporal dynamic partial equilibrium on energy and emission markets based on the maximization of total surplus, defined as the sum of suppliers and consumers surpluses. In TIAM-WORLD (Labriet etal., 2013), the World is divided in 16 regions, including Mexico and Central and South America. In the current applications, the model is set-up to explore the World energy system in 10-year time steps (slightly shorter from 2005 to 2020) from 2005 to 2100. Bioenergy included in the model includes dedicated energy crops, agricultural and forestry residues and waste and biomass from forest growth. The supply curve of energy crops is defined as a ten step supply curve (each step characterized by a specific price) calibrated to the global land-use MAgPIE model (Klein et al., 2014). The approach, described in Leimbach et al. (2013) can be summarized as follows: a series of 10 scenarios was run with MAgPIE, representing 10 different levels of possible supply by each region. Prices obtained from MAgPIE were used to feed the new 10-step bioenergy supply-curve of TIAM-WORLD. The same approach
was applied in several climate variants, including Reference case (no climate constraint), climate constraint equivalent to 550 ppm and to 450 ppm (in the cases with climate constraint, carbon prices obtained in TIAM-WORLD were used in MAgPIE). Klein et al. (2014) followed a similar approach to explore the price of bioenergy under different climate scenarios. CO2, N2O and CH4 from all anthropogenic sources (energy, industrial processes, land-use, agriculture, enteric fermentation, waste, etc.) are included. Land-use emissions of TIAM-WORLD were calibrated to the results obtained with REMIND/MAgPIE models in the ERMITAGE project (Labriet et al., 2013). Options for GHG emission reductions available in the model cover numerous fuel and technology switching options in each sector, specific CH4 and N2O destruction (leakages, adipic acid industry, etc.), mitigation of emissions from agriculture representing the implementation of advanced agriculture practices, CO2 capture (upstream, power plants, biofuel refineries, hydrogen generation) and sequestration (in geological sinks), and biological sequestration via reforestation/afforestation.
2.6. Scenarios
In this paper, we focus on three scenarios: a reference scenario and two mitigation scenarios (see Table 2). The reference scenario includes no climate policies beyond those currently enacted; proposed policies (e.g., the Copenhagen pledges) are excluded. The results of this scenario are described in Section 3. The two mitigation policy scenarios impose a globally harmonized carbon price on all GHG emissions, regardless of source (i.e., where and what flexibility). These carbon prices start at $10/tCO2e and $50/tCO2e in 2020 and rise at 4% per year, reaching $32
Table 2
Scenario description.
Scenario Scenario description
Core baseline Business-as-usual scenario including climate and energy
policies enacted prior to 2010.
Low CO2 price A global carbon tax is levied of $10/tCO2e in 2020, growing
at 4%/year to reach $32/tCO2e in 2050.
High CO2 price A global carbon tax is levied of $50/tCO2e in 2020, growing
at 4%/year to reach $162/tCO2e in 2050.
ARTICLE IN PRESS
K.V. Calvin et al./Energy Economics xxx (2015) xxx-xxx
1750 1500 1250 1000 750 500
300 250 200 150 100
il X^-
y/^ IV
1980 2000 2020 2040
1980 2000 2020 2040
Mexico
ADAGE EDGAR EPPA GCAM
TIAM-WORLD
1980 2000 2020 2040 Fig. 1. AFOLU GHG emissions in Brazil, Mexico, and Latin America without climate policy. Historical data from EDGAR (EC, 2011).
and $162/tCO2e in 2050, respectively. The results of these scenarios are discussed in Section 4.
3. Reference case results
Fig. 1 shows the evolution of AFOLU GHG emissions in Brazil, Mexico and Latin America across four models, without climate policy. As is evident in the figure, there is significant uncertainty in both historical and future emissions. For example, 2010 emissions in Latin America range from 2.3 GtCO2e/year (GCAM) to 3.3 GtCO2e/year (TIAM-WORLD) across models, due to differences in underlying source data and calibration procedures (see Table 3). In 2050, the range in emissions is similar, varying from 2.1 GtCO2e/year (TIAM-WORLD and GCAM) to 3.0 GtCO2e/year (EPPA). The transition from 2010 to 2050 shows some models with declines in AFOLU GHG emissions and some models with increases in AFOLU GHG emissions.
To understand the differences, we decompose GHG emissions into three gases: CO2, CH4, and N2O (Fig. 2a-c) and discuss each separately as different drivers affect the different gases, focusing on Latin America. More detail on Brazil and Mexico is provided in the Supplementary Material (Figs. S4 and S5). The largest uncertainty in current and future GHG emissions comes from uncertainty in CO2 (Fig. 2b). The models included in this study focus on land-use change (LUC) CO2 emissions, and thus, the largest driver for differences in future emissions are differences in future land cover (Fig. 3). Models have different realizations of future land cover, leading to different LUC CO2 emissions. For example, cropland and forestland in GCAM are virtually constant in area between 2010 and 2050, leading to LUC CO2 emissions that decline towards zero. EPPA, on the other hand, has increases in cropland at the expense of secondary and managed forests, leading to positive LUC CO2 emissions from 2010 to 2050. These emissions are declining over time, however, as forest conversion slows from the historical rate reflected in 2010 emissions. Land conversion and thus LUC CO2 emissions in ADAGE fall somewhere in between the EPPA and GCAM values, with net movement of land from cropland to forests and increasing LUC CO2 emissions after 2020.
To understand the differences in future land cover across models, it is necessary to examine the drivers of land cover, including population, income, food consumption, agricultural production, and yield. Fig. 4 depicts differences in agricultural production and yield across models. All else equal models with higher agricultural production will require more cropland; models with higher yield will require less cropland. As shown in this figure, EPPA has more growth in agricultural production
in Latin America than either ADAGE or GCAM.3 This growth in production more than offsets the increases in yield growth, resulting in more cropland area needed in EPPA than in either GCAM or ADAGE.
There is more agreement among models for CH4 and N2O than CO2, with all four models showing increases in both CH4 and N2O. Increases in CH4 emissions in these models are largely due to increases in the production and consumption of livestock (see Fig. S8). Increases in N2O in these models are largely due to increases in fertilizer use; either due to an expansion of crop production or a shift in management to more intensive fertilizer use.
4. Land-based mitigation in Latin America
4.1. Mitigation options
A number of land-based mitigation options are currently available, including reduced deforestation, afforestation, sustainable bioenergy, agricultural yield improvements, and reductions in CH4 and N2O from livestock and agriculture. The models included in this study incorporate some or all of these options in their analysis. In this section, we briefly describe each option and its potential effects on GHG emissions.
Reduced deforestation and afforestation are means of slowing the decline in or increasing terrestrial carbon stock through increased forest cover. These options are implemented in models through either command and control policies (e.g., imposing a constraint on the amount of forests) or through price policies. Both cases result in reduced land-use change CO2 emissions as compared to a reference scenario or an alternative policy scenario. However, these policies may have trade-offs in terms of increased agricultural prices (see Calvin et al., 2014; Reilly et al., 2012, and Wise et al., 2009) and competition for water resources.
Bioenergy is often deployed as a mitigation option in the energy system. It is a versatile fuel and can be used to produce electricity, liquid fuels, gas, hydrogen, or combusted directly. Additionally, bioenergy results in lower CO2 emissions than conventional fossil fuels because its carbon was removed from the atmosphere more recently. However, bioenergy can result in increased land-use change CO2 emissions if forests are cleared to grow the bioenergy (see Calvin et al., 2014; Reilly et al., 2012, and Wise et al., 2009). Bioenergy production can also result in increased N2O emissions if the bioenergy feedstock is
3 Differences in agricultural production across models are due to differences in population, income, and diet (see Supplementary Material).
ARTICLE IN PRESS
K.V. Calvin et al. / Energy Economics xxx (2015) xxx-xxx
Table 3
AFOLU emissions data sources.
TIAM-WORLD
CO2 Base year World Resources Institute CAIT 2.0 emissions database
MAC N/A curve
CH4 Base year EPA (2012) emissions
MAC Uses EPA MAC curves (EPA, 2006, Curve 2013) to calibrate substitution elasticities between inputs and emissions N2O Base year EPA (2012) emissions
MAC Uses EPA MAC curves (EPA, 2006, curve 2013) to calibrate substitution elasticities between inputs and emissions
Emissions are calculated from carbon densities computed using the Terrestrial Ecosystem Model (TEM). N/A
EDGAR, EPA, and country level inventories. See Waugh et al. (2011) for more details. Uses EPA MAC curves to calibrate substitution elasticities between inputs and emissions. See Hyman et al. (2003) for more details.
EDGAR, EPA, and country level inventories. See Waugh et al. (2011) for more details. Uses EPA MAC curves to calibrate substitution elasticities between inputs and emissions. See Hyman et al. (2003) for more details.
Emissions calculated from carbon densities and mature ages based on Houghton et al. (1999). See Kyle et al. (2011) for more details. N/A
EDGAR (2011) EPA (2006)
EDGAR (2011) EPA (2006)
Calibrated to MagPIE model
Sathaye et al. (2005)
Calibrated to MagPIE model
EPA (2006), with some updates from the ERMITAGE project
Calibrated to MagPIE model
EPA (2006), with some updates from the ERMITAGE project
fertilized (Beach and McCarl, 2010). Finally, when competing with agriculture lands, bioenergy can have implications for agricultural prices and the cost of mitigation (see Calvin et al., 2014; Reilly et al., 2012, and Wise et al., 2009).
Agricultural yield improvements are a means of mitigation because such improvements reduce the amount of cropland needed (Smith et al., 2014a, 2014b). Such a reduction in cropland area can result in reduced deforestation or even afforestation, if agricultural land is abandoned on previously forested land. These changes in land cover can reduce land-use change CO2 emissions. However, the means of achieving the yield improvements matters. Increasing yields through additional fertilizer application will lead to increases in N2O emissions, potentially offsetting the CO2 emissions reductions.
The final means of mitigating AFOLU GHG emissions discussed in this paper are reductions in CH4 and N2O from agriculture and livestock production. Two mechanisms exist in models for these reductions. First, emissions mitigation can occur through a reduction in emissions drivers, reflecting changes in lifestyle. For example, a decline in the consumption of meat and therefore livestock will result in a reduction in CH4 emissions. Second, emissions mitigation can occur through a reduction in emissions factors (i.e., emissions per unit of production), reflecting changes in agricultural and livestock production practices.
For example, a reduction in the fertilizer application rates will reduce N2O emissions, even if the same amount of food is produced.
4.2. Model-specific results: ADAGE
The version of ADAGE used in this study includes bioenergy production that can reduce GHG emissions by displacing fossil fuels, although there are potentially emissions associated with land use change that may at least partially offset the reductions in energy emissions. In addition, there are opportunities for abatement of both CO2 and non-CO2 emissions from crop and livestock production. Reallocation of agricultural inputs away from energy use as it becomes more expensive due to climate policy results in reduced CO2 emissions. A combination of reductions in output levels and emissions intensity can reduce CH4 emissions from rice cultivation and livestock production and N2O emissions from fertilizer application and livestock production.
The ADAGE model baseline includes large increases in cropland area in Latin America to meet growing demand for agricultural commodities, with an increase of 44.6% simulated for Brazil and an increase of 87.4% for the Rest of Latin America region between 2010 and 2050. This increase, while large, is comparable to historic growth in cropland area in Latin America (1.1% growth per year from 1990-2012 in FAO versus
3000 2000 -
1500 -1250 -1000 -750 -500 -
2000 1500 -1000 -500
llw^V.
1980 2000 2020 2040 1980 2000 2020 2040
ADAGE EDGAR EPPA GCAM
TIAM-WORLD
1980 2000 2020 2040 1980 2000 2020 2040
Fig. 2. AFOLU GHG emissions in Latin America without climate policy. Historical data from EDGAR (EC, 2011).
ARTICLE IN PRESS
K.V. Calvin et al./Energy Economics xxx (2015) xxx-xxx
1000 750 500
i. 250 1?
= 1000 £
750 500 250 0
Cropland
— EPPA
— FAO
— GCAM
2000 2020 2040 Fig. 3. Cropland, pastureland and forestland in Latin America without climate policy. Historical data from FAO (FAO, 2013).
1.3% per year from 2010-2050 in ADAGE; see Fig. 3). There is also a 6.4% increase in pasture in Brazil under the baseline, though pasture declines by 6.2% in the Rest of Latin America. Because of restrictions on forest conversion, cropland expansion in Brazil is coming primarily from other arable land (i.e., cropland pasture) though there is also a 2.3% decline in forest area (12.5 million hectares) between 2010 and 2050. In the Rest of Latin America, a much larger share (almost two thirds) of the land converted to cropland is coming from forest with smaller shares coming from pasture and other arable land. Although there is continued movement of other land uses to cropland in the baseline, the rate of natural land conversion to cropland slows greatly in Brazil after the 2010 period. Thus, the projected CO2 emissions from land use change fall considerably after 2010, though they remain positive. This result is consistent with recent trends in Brazil. Brazil has decreased its deforestation rates sharply from 2004 to 2010 due mostly to strong anti-deforestation policies based on: creation of new reserves in the Amazon region; command and control policies based on real time satellite monitoring and quick punishment of illegal deforestation; and targeting the municipalities with higher deforestation rates with penalties related to reduced agriculture credit to farmers (Assun^ao et al., 2012, 2013). The command and control policy was able to effectively change behavior in the country toward lower deforestation rates.
Under the mitigation scenarios analyzed, carbon price incentives result in reduced land conversion to cropland and lower crop and livestock production relative to the baseline. In addition, there is some expansion of bioenergy production, but this mitigation option is fairly limited in ADAGE due to restrictions on land conversion in Brazil. In addition, there is not currently an option for biomass electricity with carbon capture and storage, which reduces the attractiveness of mitigation through bioelectricity.
Under a scenario with a carbon price of $10/tCO2e imposed in 2020 rising at 4% per year, ADAGE results show substantial initial reductions in CO2 emissions from land use in Brazil. In 2020, these emissions are 21.5% lower than baseline levels. However, these emissions are only reduced by 2.5% in the Rest of Latin America in 2020. The reductions in CO2 emissions decline over time in Brazil, while becoming a bit larger in the Rest of Latin America. By 2050, these emissions are 3.2% lower than baseline levels in Brazil and 6.0% lower for the Rest of Latin America. Percentage reductions in emissions of non-CO2 GHG associated with AFOLU are considerably larger and are more similar across Latin America, with reductions of 24.8% and 20.7% for N2O and CH4, respectively, from Brazil and 22.1% and 16.4%, respectively, for the Rest of Latin America. Overall, net AFOLU GHG emissions from Latin America decline by about 15.4% in 2020 and 27.9% in 2050 under this relatively
— EPPA
— GCAM
— TIAM-WORLD
2010 2020 2030 2040 205(2010 2020 2030 2040 2050
Fig. 4. Agricultural production and yield in Latin America, without climate policy.
ARTICLE IN PRESS
K.V. Calvin et al. / Energy Economics xxx (2015) xxx-xxx
modest carbon incentive. With a carbon price of $50/tCO2e imposed in 2020 rising at 4% per year, emissions reductions are substantially higher, reaching reductions of 11.2%, 43.2%, and 39.1%, respectively, for CO2, N2O, and CH4 emissions from AFOLU in Brazil by 2050. Analogous values for Rest of Latin America are 4.4%, 23.5%, and 18.8% by 2050. The higher carbon price results in an overall reduction in net AFOLU GHG emissions of 34.8% across Latin America. In general, emissions reductions are considerably greater than reductions in output. For instance, 2050 production of crop commodities in Latin America falls by 3.3% for a carbon price of $10/tCO2e imposed in 2020 rising at 4% per year and 14.8% for the carbon price of $50/tCO2e imposed in 2020 rising at 4% per year. This indicates that reductions in emissions intensity within the agriculture sector associated with changing input use and production practices in response to the carbon price incentive are playing a major role in the ADAGE results.
4.3. Model-specific results: EPPA
EPPA considers abatement of CH4 and N2O as mitigation options in crop and livestock production. Energy use in agriculture is also subject to GHG mitigation policies. The agricultural sectors need to pay the tax or the carbon price considering the GHG equivalents. EPPA allows the deployment of lower emissions technologies in agriculture considering small elasticities of substitution between carbon allowances and inputs, calibrated to represent MAC curves for these sectors (Hyman et al., 2003). In the case of deforestation and land use changes, carbon policies and command and control policies may both be employed. In the current implementation, we assume command and control policies only, since it has been the only large-scale anti-deforestation policy used in the region in recent years. The model also represents biofuels and bioelectricity as mitigation options, properly accounting for their emissions.
For land use and agriculture emissions projections in Latin America, EPPA considers the observed sharp decrease in emissions from deforestation in the last ten years, possible mostly due to strong command and control policies in Brazil, the historically largest land use emitting country in the region. The current anti-deforestation policy in Brazil has almost achieved its goals, and there is no clear picture about future tighter targets. Other countries in the region are taking the Brazilian policy as an example to be followed. All these are considered in EPPA, which explains the almost constant emissions trajectory during the simulation horizon (Fig. 1). The slight increase in emissions after 2025 is due to CH4 and N2O emissions from agriculture and livestock activities, which keep growing following the economic growth in the region (Fig. S7).
Land use changes in EPPA indicate that cropland areas will double in the region from 2005 to 2050, mostly at the expense of the decrease of forest areas although there is some conversion of other arable land (e.g., cerrado) (Fig. 3). EPPA has two different categories of forestland: natural forest and a combination of harvested and secondary forests. The second type includes managed forest plantations and deforested areas under regeneration today. The conversion to cropland in EPPA occurs, in its majority, from the managed and secondary forest category, due to the constraints of the command and control policies against deforestation of natural vegetation. As the managed and secondary forest areas have much lower carbon content than the areas of natural forest, the land use CO2 emissions are kept under control at relatively low levels.
As EPPA considers command and control policies on deforestation in the core baseline scenario, carbon prices have a negligible effect on AFOLU emissions. A carbon price of $10/t of CO2 imposed in 2020, rising at 4% per year, reduces AFOLU emissions by 1.3% in 2050. Under a $50/t of CO2, emissions reductions are 2.2%. These reductions are due to the decrease of CH4 and N2O from crop and livestock production only, since the carbon policy does not affect the effectiveness of the anti-deforestation policy in EPPA and the CO2 emissions from deforestation accounts for 93% of AFOLU emissions in Latin America in 2050 in the baseline.
The carbon price increases the biofuel production in the region in the near term compared to the reference scenario, as the energy system shifts from freely-venting fossil fuels toward low carbon options like bioenergy. However, the constraint imposed in the land expansion due to the anti-deforestation policy avoids the biofuels production to expand beyond the production observed in the baseline scenario in the long run (Fig. 5). As biofuels are competitive relative to gasoline in the region, there is a sharp increase in the area devoted to it in 2030, even in the core baseline scenario. The carbon price accelerates the increase in biofuel production. Under a $50/t of CO2 the land area to grow biofuel crops reaches 12 million ha, twice the area in the reference scenario in that year. But, as the carbon price increases at 4% per year, in 2050 the area dedicated to biofuels is slightly smaller than in the reference, due to the costs related to CH4 and N2O emissions in the production of biomass crops and also the overall decrease in the economic activity due to the carbon price.
4.4. Model-specific results: GCAM
The version of GCAM used for this analysis includes bioenergy, reductions in CH4, and reductions in N2O as mitigation options that affect the agriculture and land use system. Bioenergy is deployed extensively in the energy system to reduce fossil fuel and industrial CO2 emissions, both within Latin America and elsewhere in the world. However, the use of this mitigation option can have adverse implications for AFOLU emissions (Calvin et al., 2014; Wise et al., 2009). Most reductions in CH4 and N2O are achieved through MAC curves; that is, GCAM assumes the same agriculture and livestock production levels can be achieved with lower emissions. Some reduction in these emissions is achievable through dietary shifts away from emissions-intensive livestock; however, the price elasticity of demand on food in GCAM is relatively low preventing significant dietary shifts and their associated reductions in CH4 and N2O emissions.
As a result of the mitigation options included, increasing carbon prices result in increases in AFOLU GHG emissions in GCAM. For low carbon prices, the effect is small, as the carbon price is not high enough to induce either a significant deployment of bioenergy or a significant reduction in CH4 or N2O as a result of the MAC curves. However, under a high carbon price, large changes in AFOLU GHG emissions are observed, with total GHG emissions 25% higher in 2030 in the high carbon price scenario than in the reference scenario. Fig. S10 decomposes the change in emissions into sector and gas for the year 2050. This figure shows that increases in carbon prices induce increases in land use change CO2 and N2O emissions for bioenergy production, as bioenergy production displaces higher carbon content ecosystems (e.g., forests4) because policies to prevent such deforestation are excluded from this analysis. However, increasing carbon prices also result in declines in CH4 from ruminant animals. The combination of effects leads to small changes in total AFOLU emissions.
The increase in land use change CO2 and N2O emissions for bioenergy production is the result of a large increase in purpose-grown bioenergy production (Fig. 5). This bioenergy is used in combination with CO2 capture and storage in the electricity and refined liquids sectors, leading to significant reductions in fossil fuel and industrial CO2 emissions.
The decline in CH4 emissions from ruminant animals is due to a combination of declines in livestock production (Fig. S8) and decreases in CH4 emissions factors (tonne of CH4 emitted per tonne of livestock produced). The decline in livestock production is due to price-induced shifts in diet. An increasing carbon price in GCAM leads to an increase in the price of crops due to increasing competition for land between food and bioenergy. This increase in crop price, in turn, leads to an increase in the price of livestock and a decline in livestock consumption.
4 The specific dynamics of bioenergy expansion differ by country. For example,
bioenergy expansion in Argentina does not result in reduced forest cover. Instead, bioenergy in Argentina comes at the expense of non-forest land (e.g., pasture).
ARTICLE IN PRESS
K.V. Calvin et al./Energy Economics xxx (2015) xxx-xxx
î x c o
Core baseline High C02 Price Low C02 Price
— EPPA GCAM
2010 2020 2030 2040 2050
Fig. 5. Area devoted to energy crop production in Latin America.
4.5. Model-specific results: TIAM-WORLD
Mitigation of CH4 and N2O from agriculture remains negligible in Latin America until 2050 in the low tax scenario, while in the higher tax scenario, mitigation starts in 2030 and remains stable until 2050 (5% of emissions are reduced).
Modern bioenergy production in Latin America satisfies two markets: local consumption and exports. While the low tax is not high enough to stimulate the supply of bioenergy before 2050, for local consumption or exports, the high carbon tax creates an incentive for higher bioenergy supply and local use from 2020 (supply +5%) to 2050 ( + 40%). Uses are diverse (production of liquid biofuels, direct use in industry, production of electricity) but dominated by liquid biofuels in all scenarios. Production of liquid biofuels increases by more than 50% in 2030 and 75% in 2050 in the high tax scenario compared to the Reference case; capture of CO2 at biorefinery levels starts in 2040 in the high tax scenario.
It is interesting to note that exports of bioenergy commodities do not increase in climate mitigation scenarios compared to the reference case. They even decrease in the mid-term (2030-2050). This trend is observed in several regions of the World in TIAM-WORLD's results, showing a preference for local consumption of bioenergy in climate scenarios. However, Labriet et al. (2013) point out that the assumptions related to the trade of bioenergy (cost, availability) may have a crucial impact on the localization of the supply of bioenergy, and could radically change the results of the big suppliers of bioenergy such as Latin America.
In TIAM-WORLD, crops for energy use are considered as carbon free over their lifecycle, and thus the corresponding bioenergy is assumed to be grown in a sustainable manner. Consistent with this assumption, first-generation biofuels are constrained (around 20 EJ at global level, 4.4 EJ in Latin America); this constraint reflects the amount of bioenergy that could be produced on surplus agricultural land without irrigation (Smeets et al., 2004,2007). This upper limit represents a default preference for second-generation over first generation biofuels, given the sus-tainability debates associated with the latter. CO2 emissions resulting from land-use become negative in tax scenarios because the CO2 MAC curve implemented in TIAM-WORLD assumes that afforestation measures are implemented as a cost-effective mitigation option; in other words, Latin America becomes a net sequester of CO2 in both climate scenarios.5 However, due to limited mitigation of CH4 and N2O, AFOLU GHG emissions are still positive even under the high CO2 price scenario.
4.6. Comparing across models
Fig. 6 depicts the relationship between carbon price and abatement for Latin American AFOLU GHG, CO2, CH4, and N2O emissions across the four models included in this study for 2050 (Fig. S11 shows this for 2020). These figures clearly depict the divergence of model results, with GCAM showing small increases in emissions as a result of climate policy, EPPA and ADAGE showing small decreases in emissions, and TIAM-WORLD showing large decreases in emissions. The large difference in response to policy is primarily due to differences in mitigation options. TIAM-WORLD includes afforestation as a mitigation option, which can lead to sequestration of CO2 in the terrestrial system. Such an option is deployed widely resulting in large emissions reductions. ADAGE, EPPA, and GCAM exclude such an option in their models for this study and thus have much more limited (or even negative) abatement. Previous analyses with EPPA (Reilly et al., 2012) and GCAM (Calvin et al., 2014) show that when afforestation is included as a mitigation option it is deployed widely as in TIAM-WORLD in this study.
Differences in the response of emissions to policy between the models are also influenced by differences in the effect of climate policy on bioenergy and the mitigation potential from N2O and CH4. In ADAGE and EPPA, total bioenergy consumption in Latin America in 2050 declines slightly when a climate policy is imposed.6 In GCAM and TIAM-WORLD, bioenergy consumption increases in 2050 as a result of a climate policy. TIAM-WORLD assumes that this increase in bioenergy is produced sustainably, resulting in no land-use change CO2 emissions. In GCAM, this increase comes at the expense of forests resulting in a noticeable increase in land-use change CO2 emissions. The CO2 emissions resulting from this deforestation overwhelm any mitigation from N2O and CH4 resulting in a modest increase in AFOLU GHG emissions. Since ADAGE and EPPA present declines in bioenergy consumption the effect of climate policy on land-use change CO2 emissions is negligible. In these models, the mitigation of N2O and CH4 dominates and results in modest decreases in total AFOLU GHG emissions.
The mitigation potential estimated in ADAGE, EPPA, and TIAM-WORLD is fairly consistent with estimates from the Intergovernmental Panel on Climate Change's (IPCC) fifth assessment report (Smith et al., 2014a). The IPCC estimates mitigation potential in Latin America between 0.25 GtCO2e (at $20/tCO2) and 0.6 GtCO2e (at $100/tCO2) in 2030 excluding forestry. In comparison, EPPA estimates
5 Note that several other studies have found afforestation to be a cost effective mitigation strategy (see Clarke et al., 2014 for a summary).
6 In 2020, bioenergy consumption is higher in EPPA under the mitigation policy, but the reverse is true in 2050.
ARTICLE IN PRESS
A) GHGs
K.V. Calvin et al. / Energy Economics xxx (2015) xxx-xxx
B) CO2
« 100 o
% Reduction from Reference
C) ch4
ADAGE EPPA g
GCAM g
TIAM-WORLD c3
« 100 O
200 400
% Reduction from Reference
D) n2o
ADAGE EPPA g
GCAM g
TIAM-WORLD c3
TIAM-WORLD
% Reduction from Reference
TIAM-WORLD
10 20 30
% Reduction from Reference
Fig. 6. AFOLU GHG emission abatement as a function of carbon price in Latin America for 2050.
mitigation between 0.3 GtCO2e (at $15/tCO2) and 0.5 GtCO2e in 2030 (at $75/tCO2). ADAGE estimates 0.3 GtCO2e of mitigation at $15/tCO2. With forestry, the IPCC mitigation potential increases to between 0.75 GtCO2e (at $20/tCO2) and 2.3 GtCO2e (at $100/tCO2). In comparison, TIAM-WORLD estimates mitigation between 1.2 GtCO2e (at $15/tCO2) and 2.1 GtCO2e in 2030 (at $75/tCO2). The GCAM model indicates the possibility of negative abatement, i.e., an increase in emissions, if bioenergy is not produced sustainably. The IPCC estimates are from bottom-up sector-specific models and thus the possibility of negative feedbacks from the energy system are excluded.
5. Discussion and conclusions
The work detailed in this paper has focused on examining the future evolution of AFOLU GHG emissions in Latin America. We find that there are large uncertainties in both present and future emissions, with and without climate policies. Without climate policy, differences in future AFOLU GHG emissions across models are largely driven by differences in the extent to which cropland displaces forest cover, reflecting a number of uncertainties including the effectiveness of deforestation policies. Models with more expansion of crops onto forested area have higher land-use change CO2 emissions and vice versa. Additionally, some differences exist in the future emissions factors for N2O and CH4, with one model projecting declines in emissions intensity leading to declines in N2O and CH4 emissions.
With climate policy, differences in future AFOLU GHG emissions across models are largely driven by differences in mitigation options.
Including afforestation as a policy option results in significant emissions abatement. Excluding this option results in more modest abatement or even increases in emissions. Models also differ in the extent to which bioenergy is deployed as a mitigation option and the effect of bioenergy production on land-use change CO2 emissions.
There are some important caveats to the work presented here. First, this paper has focused on climate implications of AFOLU and thus, we have focused on GHG emissions. We exclude the effects of agriculture and land use on other emissions (e.g., aerosols, ozone precursors). We also ignore the effects of AFOLU GHG mitigation efforts on the economy, ecosystem services, and biodiversity. Land-based mitigation, however, could have substantial effects on food prices and thus consumer welfare. We leave further discussion of these dynamics to other studies (see Lotze-Campen et al., 2014 for example). Next, the results included in this analysis ignore institutional barriers and transaction costs associated with implementing the policies discussed. Large barriers or costs would have implications for emissions and emissions mitigation.
Finally, the results included in this paper only harmonized carbon prices across models. Other dimensions of the model (e.g., population, GDP, mitigation options, bioenergy and crop trade assumptions, etc.) were determined by individual modeling teams and vary substantially across the models included in this study. While such a study design allows us to explore the uncertainty in the future evolution of AFOLU GHG emissions in Latin America with and without climate policy, it makes it difficult to isolate the effect of an individual mitigation measure. Future studies may seek to redress this problem with further model harmonization; however, this is outside the scope of this study.
ARTICLE IN PRESS
K.V. Calvin et al. /Energy Economics xxx (2015) xxx-xxx
Acknowledgments
The research that allowed the publication of this paper has been produced with the financial assistance of the European Union in the context of the CLIMACAP project (EuropeAid/131944/C/SER/Multi) and of the U.S. Agency for International Development and U.S. Environmental Protection Agency in the context of the LAMP project (under Interagency Agreements DW89923040 and DW89923951US). The contents of this publication are the sole responsibility of the authors and can in no way be taken to reflect the views of the European Union or the U.S. government. The authors would like to thank the feedback and efforts from all CLIMACAP and LAMP project partners for enabling the research results reported in this article.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.eneco.2015.03.020.
References
Assun^ao, J., Gandour, C.C., Rocha, R., 2012. Deforestation slowdown in the Legal Amazon:
prices or policies. Climate Policy Initiative Working Paper. Assun^ao, J., Gandour, C., Rocha, R., 2013. DETERring deforestation in the Brazilian Amazon: environmental monitoring and Law enforcement. Climate Policy Initiative Report, PUC-Rio, May.
Beach, R.H., Birur, D.K., Davis, L., Ross, M.T., 2011. A dynamic general equilibrium analysis of U.S. biofuels production. GTAP Resource #3652, 2011 Conference paper, 14th Annual Conference on Global Economic Analysis, Venice, Italy. Beach, R., McCarl, BA, 2010. US Agricultural and forestry impacts of the energy independence and security act: FASOM results and model description. http://yosemite.epa. gov/SAB/SABPRODUCT.nsf/962FFB6750050099852577820072DFDE/$File/ FASOM+Report_EISA_FR.pdf. Blanco, G., Gerlagh, R., Suh, S., Barrett, J., de Coninck, H., Diaz Morejon, C.F., Mathur, R., Nakicenovic, N., Ofusu Ahenkorah, A., Pan, J., Pathak, H., Rice, J., Richels, R., Smith, S.J., Stern, D., Toth, F., Zhou, P., 2014. Drivers, trends, and mitigation. In: Edenhofer, O., et al. (Eds.), Climate Change 2014: Mitigation of Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Calvin, K., Clarke, L., Edmonds, J., Eom, J., Hejazi, M., Kim, S., Kyle, P., Link, R., Luckow, P.,
Patel, P., 2011. GCAM Wiki documentation. from. https://wikik.umd.edu/gcam/. Calvin, K., Wise, M., Klein, D., McCollum, D., Tavoni, M., van der Zwaan, B., van Vuuren, D.P., 2013. A multi-model analysis of the regional and sectoral roles of bioenergy in near- and long-term CO2 emissions reduction. Clim. Chang. Econ. 4 (4), 32. Calvin, K., Wise, M., Kyle, P., Patel, P., Clarke, L., Edmonds, J., 2014. Trade-offs of different land and bioenergy policies on the path to achieving climate targets. Clim Chang. 123 (3-4), 691-704.
Clarke, L., Jiang, K., Akimoto, K., Babiker, M., Blanford, G., Fisher-Vanden, K., Hourcade , J.-C., Krey, V., Kriegler, E., Loschel, A., McCollum, D., Paltsev, S., Rose, S., Shukla, P.R., Tavoni, M., van der Zwaan, B., van Vuuren, D.P., 2014. Assessing Transformation Pathways. In: Edenhofer, O., Pichs-Madruga, R., Sokona, Y., et al. (Eds.), Climate Change 2014, Mitigation of Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. EC-JRC/PBL, 2011. EDGAR version 4.2. http://edgar.jrc.ec.europa.eu/). EPA, 2006. Global anthropogenic non-CO2 greenhouse Gas emissions: 1990-2020. http://www.epa.gov/climatechange/Downloads/EPAactivities/ GlobalAnthroEmissionsReport.pdfhttp://www.epa.gov/climatechange/Downloads/ EPAactivities/GlobalAnthroEmissionsReport.pdf. EPA, 2012. Non-CO2 greenhouse gases: international emissions and projections: 1990-2030. http://www.epa.gov/climatechange/Downloads/EPAactivities/EPA_ Global_NonCO2_Projections_Dec2012.pdf. EPA, 2013. Global mitigation of non-CO2 greenhouse gases: 2010-2030. EPA Report 430R13011 (http://www.epa.gov/climatechange/Downloads/EPAactivities/MAC_ Report_2013.pdf).
European Commission, Joint Research Centre (JRC)/Netherlands Environmental Assessment Agency (PBL), 2011. Emission Database for Global Atmospheric Research (EDGAR), release version 4.2. http://edgar.jrc.ec.europa.eu. FAO, 2013. FAOSTAT, Food and Agriculture Organization of the United Nations.
http://faostat.fao.org/site/291/default.aspx. Felzer, B.S.F., Kicklighter, D.W., Melillo, J.M., Wang, C., Zhuang, Q., Prinn, R.G., 2004. Ozone effects on net primary production and carbon sequestration in the conterminous United States using a biogeochemistry model. Tellus 56B, 230-248. Houghton, R.A., 1999. The annual net flux of carbon to the atmosphere from changes in
land use 1850-1990. Tellus 51B, 298-313. Hyman, R.C., Reilly, J.M., Babiker, M.H., De Masin, A.,Jacoby, H.D., 2003. Modeling non-
CO2 greenhouse gas abatement. Environ. Model. Assess. 8,175-186. Intergovernmental Panel on Climate Change (IPCC), 2006. IPCC guidelines for national greenhouse gas inventories. In: Eggleston, H.S., Buendia, L., Miwa, K., Ngara, T.,
Tanabe, K. (Eds.), Agricultural, Forestry and other Land Use, National Greenhouse Gas Inventories Program. vol. 4. IGES, Japan.
Klein, D., Humpenöder, F., Bauer, N., Dietrich, J.P., Popp, A., Bodirsky, B.L., Bonsch, M., Lotze-Campen, H., 2014. The global economic long-term potential of modern biomass in a climate constraint world. Environ. Res. Lett. 9.
Kyle, P., Luckow, P., Calvin, K., Emanuel, W., Nathan, M., Zhou, Y., 2011. GCAM 3.0 Agriculture and Land Use: Data Sources and Methods. National Laboratory, Richland, WA, Pacific Northwest (http://wiki.umd.edu/gcam/images/2/25/GCAM_AgLU_Data_ Documentation.pdf).
Labriet, M., Kanudia, A., Loulou, R., Biberacher, M., Edwards, N., Holden, P., Pizzileo, B., Joshi-Ram, S., Vielle, M., Dietrich, J., Leimbach, M., Babonneau, F., 2013. Deliverable 8.1. Uncertainty analyses in TIAM-WORLD. EU-FP7 ERMITAGE Project.
Leimbach, M., Dietrich, J., Popp, A., Klein, A., Labriet, M., Kanudia, A., Slavin, P., Bauer, N., Biewald, A., Bodirsky, B., Humpenöder, F., Lenz, T., Schultes, A., 2013. Deliverable 7.1. Macro-economy agriculture link. EU-FP7 ERMITAGE Project.
Lotze-Campen, et al., 2014. Agricultural Economics.
Loulou, R., 2008. ETSAP-TIAM: the TIMES integrated assessment model part II: mathematical formulation. Computational management science, special issue. Manag. Energy Environ. 5 (1), 41-66.
Loulou, R., Labriet, M., 2008. ETSAP-TIAM: the TIMES integrated assessment model Part I: model structure. Computational management science, special issue. Manag. Energy Environ. 5(1), 7-40.
Melillo, J.M., Reilly, J.M., Kicklighter, D.W., Gurgel, A.C., Cronin, T.W., Paltsev, S., Felzer, B.S., Wang, X., Sokolov, A.P., Schlosser, C.A., 2009. Indirect emissions from biofuels: how important? Science 326,1397-1399.
Monfreda, C., Ramankutty, N., Hertel, T., 2009. Global Agricultural Land Use Data for Climate Change Analysis. In: Hertel, T., Rose, S., Tol, R. (Eds.), Economic Analysis of Land Use in Global Climate Change Policy. Routledge.
Narayanan, B.G., Walmsley, T.L. (Eds.), 2008. Global Trade, Assistance, and Production: The GTAP 7 Data Base. Center for Global Trade Analysis, Purdue University, West Lafayette, IN.
Paltsev, S., Reilly, J., Jacoby, H., Eckaus, R., Mcfarland, J., Sarofim, M., Asadoorian, M., Babiker, M., 2005. The MIT Emissions Prediction and Policy Analysis (EPPA) Model: version 4. MIT Joint Program on the Science and Policy of Global Change, Report 125. MIT, Cambridge, MA Available in: http://web.mit.edu/globalchange/www/ MITJPSPGC Rpt125.pdf.
Popp, A., Rose, S., Calvin, K., Van Vuuren, D., Dietrich, J., Wise, M., Stehfest, E., Humpenöder, F., Kyle, P., Van Vliet, J., Bauer, N., Lotze-Campen, H., Klein, D., Kriegler, E., 2014. Land-use transition for bioenergy and climate stabilization: model comparison of drivers, impacts and interactions with other land use based mitigation options. Clim. Chang. 123 (3-4), 495-509.
Reilly, J., Mellilo, J., Cai, Y., Kicklighter, D., Gurgel, A., Paltsev, S., Cronin, T., Sokolov, A., Schlosser, A., 2012. Using land to mitigate climate change: Hitting the target, recognizing the trade-offs. Environ. Sci. Technol. 46 (11), 5672-5679.
Rose, S.K., Ahammad, H., Eickhout, B., Fisher, B., Kurosawa, A., Rao, S., Riahi, K., van Vuuren, D.P., 2012. Land-based mitigation in climate stabilization. Energy Econ. 34 (1), 365-380.
Rose, S., Kriegler, E., Bibas, R., Calvin, K., Popp, A., van Vuuren, D.P., Weyant, J., 2014. Bioenergy in energy transformation and climate management. Clim. Chang. 123 (3-4), 477-493.
Ross, M., 2009. Documentation of the Applied Dynamic Analysis of the Global Economy (ADAGE) model. Working Paper 09_01, April 2009, RTI International, Research Triangle Park, NC.
Sathaye, J., Makundi, W., Dale, L., Chan, P., Andrasko, K., 2005. Estimating Global Forestry GHG Mitigation Potential and Costs: A Dynamic Partial Equilibrium Approach. LBNL-55743.
Smeets, E., Faaij, A., Lewandowski, I., 2004. A quickscan of global bio-energy potentials to 2050. An analysis of the regional availability of biomass resources for export in relation to the underlying factors. Report NWS-E-2004-109.
Smeets, E.M., Faaij, A.P.C., Lewandowski, I.M., Turkenburg, W.C., 2007. A bottom-up assessment and review of global bio-energy potentials to 2050. Prog. Energy Combust. Sci. 33, 56-106.
Smith, P., Bustamante, M., Ahammad, H., Clark, H., Dong, H., Elsiddig, E., Harberl, H., Harper, R., House, J., Jafari, M., Masera, O., Mbow, C., Ravindranath, N., Rice, C., Robledo Abad, C., Romanovskaya, A., Sperling, F., Tubiello, F., 2014a. Agriculture, forestry and other land use (AFOLU). In: Edenhofer, O., et al. (Eds.), Climate change 2014: mitigation of climate change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.
Smith, P., et al., 2014b. Global Change Biology.
B. van der Zwaan, L. Clarke, K. Calvin (Guest Editors), "Climate Policy in Latin America: implications for energy and land use", Overview of a Special Issue on the findings of the CLIMACAP-LAMP project, Energy Economics, 2015
Waugh, Caleb, et al., 2011. Emission inventory for non-CO2 greenhouse gases and air pollutants in EPPA 5.
Wise, M., Calvin, K., Thomson, A., Clarke, L., Bond-Lamberty, B., Sands, R., Smith, S.J., Janetos, A., Edmonds, J., 2009. Implications of limiting CO2 concentrations for land use and energy. Science 324 (5931), 1183-1186.
Wise, M., Calvin, K., Kyle, P., Luckow, P., Edmonds, J., 2014. Economic and Physical Modeling of Land Use in GCAM 3.0 and an Application to Agricultural Productivity, Land, and Terrestrial Carbon. Climate Change Economics 5 (2).
