A generalised approach of accounting for biospheric carbon stock changes under the Kyoto Protocol

The Kyoto Protocol aims to reduce net emissions of greenhouse gases to the atmosphere by various measures including through management of the biosphere. However, the wording that has been adopted may be difficult and costly to implement, and may ultimately make it impossible to cost-effectively include biosphere management to reduce net greenhouse gas emissions. An alternative scheme is proposed here, especially for the second and subsequent commitment periods, to more effectively deal with the anthropogenic component of carbon stock changes in the biosphere. It would categorise the terrestrial biosphere into different land-use types, with each one having a characteristic average carbon density determined by land-use and environmental factors. Each transition from one land-use type to another, or a change in average carbon density within a specified type due to changed management would be defined as anthropogenic and credited or debited to the responsible nation. To calculate annual credits and/or debits, the change in average carbon stocks must be divided by a time constant which would either be a characteristic of each possible land-use conversion, or applicable to the sum of changes to a nation's biospheric carbon stocks. We believe that this scheme would be simpler and less expensive to implement than one based on the measurement of actual carbon changes from all specified areas of land. It would also avoid undue credits or debits, because they would only accrue as a result of identified anthropogenic components of biospheric carbon changes whereas carbon fluxes that are due to natural variation would not be credited or debited.     

Developing Canada's National Forest Carbon Monitoring, Accounting and Reporting System to Meet the Reporting Requirements of the Kyoto Protocol

The rate of carbon accumulation in the atmosphere can be reduced by decreasing emissions from the burning of fossil fuels and by increasing the net uptake (or reducing the net loss) of carbon in terrestrial (and aquatic) ecosystems. The Kyoto Protocol addresses both the release and uptake of carbon. Canada is developing a National Forest Carbon Monitoring, Accounting and Reporting System in support of its international obligations to report greenhouse gas sources and sinks. This system employs forest-inventory data, growth and yield information, and statistics on natural disturbances, management actions and land-use change to estimate forest carbon stocks, changes in carbon stocks, and emissions of non-CO₂ greenhouse gases. A key component of the system is the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS). The model is undergoing extensive revisions to enable analyses at four spatial scales (national, provincial, forest management unit and stand) and in annual time steps. The model and the supporting databases can be used to assess carbon-stock changes between 1990 and the present, and to predict future carbon-stock changes based on scenarios of future disturbance rates and management actions.     

Accounting of forest carbon sinks and sources under a future climate protocol—factoring out past disturbance and management effects on age–class structure

Today, forests in the northern hemisphere are a sink for carbon dioxide (CO₂) from the atmosphere, partly due to changes in forest management practice and intensity. Parties of the Kyoto Protocol had the option to elect to account for direct human-induced carbon (C) sources and sinks from land management activities since 1990. The effect of age–class structure of a forest landscape resulting from past practices and disturbances before the reference year 1990 should be excluded, but methods for “factoring out” the effects of this age–class legacy on carbon emissions and removals are lacking. The legacy effect can be strong and can even overwhelm effects of post-1990 management. It therefore needs to be “factored out”, i.e., removed from the direct human-induced post-1990 effects. In this study we examine how the contributions to forest biomass carbon stock changes of (1) past (pre-1990) disturbances and harvest and (2) recent (post-1990) changes in forest management can be differentiated in present and future observable carbon dynamics in managed forest ecosystems. We also calculate the consequences of different accounting rules for the magnitude and direction of accountable C stock changes in European countries in the period 2013–2017.Different accounting approaches are compared in terms of applicability and their ability to provide incentives for management changes to increase carbon sinks and reduce carbon sources. We demonstrate implications of the various ways of accounting for a sample of European countries with different initial age–class structures. The current forest age–class distribution in countries determines whether and how many credits can be created by the various accounting approaches. We suggest an approach that includes a dynamic, forward-looking baseline as reference and list options to define such a baseline. Accounting of recent management change against such a baseline factors out the contribution of the legacy effect to accounting results and only rewards the effect of recent changes in forest management practices in support of climate change mitigation. We demonstrate that relatively simple, state-of-the-art forest models can factor out effects of past practices and past disturbances on present and future carbon stock changes. Factoring out of past practice effects is thus technically feasible but the numerical results are highly dependent on the choice of baseline which will be subject to negotiation. It is possible, however, to select a dynamic baseline that represents “business-as-usual”, and to isolate and account for only the changes in management. Changes in accounting rules will always be advantageous for some countries and disadvantageous for others, but using a dynamic “business-as-usual” baseline effectively removes the legacy of pre-1990 age–class effects, and thus overcomes one of the acknowledged shortcomings of the current accounting approach.     

The ‘hidden’ social costs of forestry offsets

The article quantifies the size of ‘hidden’ social costs that are incurred by forestry offsets in the voluntary market that promise to offset present emissions sometime in the future. It does this by estimating the difference between the social costs of carbon (C) emitted and of costs offset by removal of C from the atmosphere by reforestation/afforestation. All current attempts to make forestry offsets more reliable focus on quality control rather than the mismatch of the timing of emissions and their offset. Recommendations that follow from the analysis are twofold. First, that markets for carbon dioxide equivalent (CO2e) removals by voluntary offsets should be confined to the annual incremental removals actually achieved. Second, the promoters of voluntary offsets projects should declare the annual stream of carbon credits and debits expected so that buyers can place a present value on such projects.     

Accounting for sequestered carbon: the question of permanence

In its attempt to provide quantitative limits on greenhouse gas emissions, the Kyoto protocol accepts the principle that sequestration of carbon in the terrestrial biosphere can be used to offset emissions of carbon from fossil-fuel combustion. Whether or not the Kyoto protocol ever comes into force, it is worthwhile to understand how carbon sequestration might be treated in any mitigation plan that provides a tax or ration on carbon emissions. Emission credits, as proposed for the energy sector, are based on the idea that a prevented emission is prevented forever, and emission credits might be traded among parties. In the event that sequestered carbon is subsequently released to the atmosphere, it would be advantageous to agree what the liability is and who assumes that liability. We describe a system whereby emissions credits could be rented, rather than sold, when carbon is sequestered but permanence of sequestration is either not certain or not desired. Our proposal is similar to that offered by the government of Colombia except that it casts these temporary emissions credits into the traditional concepts of rental agreements and it clarifies the opportunities for secondary transactions. A rental contract for emissions credits would establish continuous responsibility for sequestered carbon; credit would be assigned when carbon is sequestered and debits would accrue when carbon is emitted.     

Equal Opportunity for Biomass in Greenhouse Gas Accounting of CO2 Capture and Storage: A Step Towards More Cost-Effective Climate Change Mitigation Regimes

Carbon dioxide capture and permanent storage (CCS) is one of the most frequently discussed technologies with the potential to mitigate climate change. The natural target for CCS has been the carbon dioxide (CO₂) emissions from fossil energy sources. However, CCS has also been suggested in combination with biomass during recent years. Given that the impact on the earth's radiative balance is the same whether CO₂ emissions of a fossil or a biomass origin are captured and stored away from the atmosphere, we argue that an equal reward should be given for the CCS, independent of the origin of the CO₂. The guidelines that provide assistance for the national greenhouse gas (GHG) accounting under the Kyoto Protocol have not considered CCS from biomass (biotic CCS) and it appears that it is not possible to receive emission credits for biotic CCS under the first commitment period of the Kyoto Protocol, i.e., 2008–2012. We argue that it would be unwise to exclude this GHG mitigation alternative from the competition with other GHG mitigation options. We also propose a feasible approach as to how emission credits for biotic CCS could be included within a future accounting framework.     

Including land use,land-use change,and forestry in future climate change,agreements: thinking outside the box

This paper presents a framework that encompasses a full range of options for including land use, land-use change, and forestry (LULUCF) within future agreements under the United Nations Convention on Climate Change (UNFCCC). The intent is to provide options that can address the broad range of greenhouse gas (GHG) emissions and removals as well as to bring the broadest possible range of nations into undertaking mitigation efforts. We suggest that the approach taken for the Kyoto Protocol's first commitment period is only one within a much larger universe of possible approaches. This larger universe includes partially or completely “de-linking” LULUCF commitments from those in other sectors, and allowing commitments specified in terms other than tonnes of greenhouse gases. Such approaches may provide clarity and transparency concerning the role of the various sectors in the agreements and encourage participation in agreements by a more inclusive, diverse set of countries, resulting in a more effective use of LULUCF in addressing climate change.     

Forest carbon accounting methods and the consequences of forest bioenergy for national greenhouse gas emissions inventories

While bioenergy plays a key role in strategies for increasing renewable energy deployment, studies assessing greenhouse gas (GHG) emissions from forest bioenergy systems have identified a potential trade-off of the system with forest carbon stocks. Of particular importance to national GHG inventories is how trade-offs between forest carbon stocks and bioenergy production are accounted for within the Agriculture, Forestry and Other Land Use (AFOLU) sector under current and future international climate change mitigation agreements. Through a case study of electricity produced using wood pellets from harvested forest stands in Ontario, Canada, this study assesses the implications of forest carbon accounting approaches on net emissions attributable to pellets produced for domestic use or export. Particular emphasis is placed on the forest management reference level (FMRL) method, as it will be employed by most Annex I nations in the next Kyoto Protocol Commitment Period. While bioenergy production is found to reduce forest carbon sequestration, under the FMRL approach this trade-off may not be accounted for and thus not incur an accountable AFOLU-related emission, provided that total forest harvest remains at or below that defined under the FMRL baseline. In contrast, accounting for forest carbon trade-offs associated with harvest for bioenergy results in an increase in net GHG emissions (AFOLU and life cycle emissions) lasting 37 or 90 years (if displacing coal or natural gas combined cycle generation, respectively). AFOLU emissions calculated using the Gross-Net approach are dominated by legacy effects of past management and natural disturbance, indicating near-term net forest carbon increase but longer-term reduction in forest carbon stocks. Export of wood pellets to EU markets does not greatly affect the total life cycle GHG emissions of wood pellets. However, pellet exporting countries risk creating a considerable GHG emissions burden, as they are responsible for AFOLU and bioenergy production emissions but do not receive credit for pellets displacing fossil fuel-related GHG emissions. Countries producing bioenergy from forest biomass, whether for domestic use or for export, should carefully consider potential implications of alternate forest carbon accounting methods to ensure that potential bioenergy pathways can contribute to GHG emissions reduction targets.     

Carbon sequestration in wood products: a method for attribution to multiple parties

When forest is harvested some of the forest carbon ends up in wood products. If the forest is managed so that the standing stock of the forest remains constant over time, and the stock of wood products is increasing, then carbon dioxide is being removed from the atmosphere in net and this should be reflected in accounting for greenhouse gas emissions. We suggest that carbon sequestration in wood products requires cooperation of multiple parties; from the forest owner to the product manufacturer to the product user, and perhaps others. Credit for sequestering carbon away from the atmosphere could acknowledge the contributions of these multiple parties. Accounting under a cap-and-trade or tax system is not necessarily an inventory system, it is a system designed to motivate and/or reward an environmental objective. We describe a system of attribution whereby credits for carbon sequestration would be shared among multiple, contributing parties. It is hoped that the methodology outlined herein proves attractive enough to parties concerned to spur them to address the details of such a system. The system of incentives one would choose for limiting or controlling greenhouse gas emissions could be quite different, depending on how the attribution for emissions and sequestration is chosen.     

Clearing the way for reducing emissions from tropical deforestation

Carbon emissions from tropical deforestation account for about 25% of all anthropogenic carbon dioxide emissions but cannot be credited under current climate change agreements. In the discussions around the architecture of the post-2012 climate regime, the possibility of including credits for reduced emissions from deforestation arises. The paper reviews two approaches for this, compensated reductions (CR) as proposed by Santilli et al. and the Joint Research Centre proposal that combine voluntary commitments by non-Annex I countries to reduce emissions from deforestation with carbon market financing. Both approaches have the clear advantages of simplicity and the possibility of fitting to an evolving greenhouse gas emission reduction regime. The authors consider the strengths and limitations of each proposal and build upon them to address several implementation challenges and options for improvement. Given the urgency of avoiding dangerous climate change, the timely development of technically sound, politically acceptable, cost-effective and practicable measures to reduce emissions from deforestation and forest degradation is essential. These two approaches take us a step closer to this goal, but they need to be refined rapidly to enable this goal to be realised.     

Sensitivity of amounts and distribution of tropical forest carbon credits depending on baseline rules

One of the largest sources of global greenhouse gas emissions can be addressed through conservation of tropical forests by channeling funds to developing countries at a cost-savings for developed countries. However, questions remain to be resolved in negotiating a system for including reduced emissions from deforestation and forest degradation (REDD) in a post-Kyoto climate treaty. The approach to determine national baselines, or reference levels, for quantifying REDD has emerged as central to negotiations over a REDD mechanism in a post-Kyoto policy framework. The baseline approach is critical to the success of a REDD mechanism because it affects the quantity, credibility, and equity of credits generated from efforts to reduce forest carbon emissions. We compared outcomes of seven proposed baseline approaches as a function of country circumstances, using a retrospective analysis of FAO-FRA data on forest carbon emissions from deforestation. Depending upon the baseline approach used, the total credited emissions avoided ranged over two orders of magnitude for the same quantity of actual emissions reductions. There was also a wide range in the relative distribution of credits generated among the five country types we identified. Outcomes were especially variable for countries with high remaining forest and low rates of deforestation (HFLD). We suggest that the most credible approaches measure emissions avoided with respect to a business-as-usual baseline scenario linked to historic emissions data, and allow limited adjustments based on forest carbon stocks.     

Estimating baseline carbon emissions for the Eastern Panama Canal watershed

To participate in the potential market for carbon credits based on changes in the use and management of the land, one needs to identify opportunities and implement land-use based emissions reductions or sequestration projects. A key requirement of land-based carbon (C) projects is that any activity developed for generating C benefits must be additional to business-as-usual. A rule-based model was developed and used that estimates changes in land-use and subsequent carbon emissions over the next twenty years using the Eastern Panama Canal Watershed (EPCW) as a case study. These projections of changes in C stocks serve as a baseline to identify where opportunities exist for implementing projects to generate potential C credits and to position Panama to be able to participate in the emerging C market by developing a baseline under scenarios of business-as-usual and new-road development. The projections show that the highest percent change in land use for the new-road scenario compared to the business-as-usual scenario is for urban areas, and the greatest cause of C emission is from deforestation. Thus, the most effective way to reduce C emissions to the atmosphere in the EPCW is by reducing deforestation. In addition to affecting C emissions, reducing deforestation would also protect the soil and water resources of the EPCW. Yet, under the current framework of the Clean Development Mechanism (CDM), only credits arising from reforestation are allowed, which after 20 years of plantation establishment are not enough to offset the C emissions from the ongoing, albeit small, rate of deforestation in the EPCW. The study demonstrates the value of spatial regional projections of changes in land cover and C stocks: The approach helps a country identify its potential greenhouse gas (GHG) emission liabilities into the future and provides opportunity for the country to plan alternative development pathways. It could be used by potential project developers to identify which types of projects will generate the largest C benefits and provide the needed baseline against which a project is then evaluated. Spatial baselines, such as those presented here, can be used by governments to help identify development goals. The development of such a baseline, and its expansion to other vulnerable areas, well positions Panama to respond to the future market demand for C offsets. It is useful to compare the projected change in land cover under the business-as-usual scenario to the goals set by Law 21 for the year 2020. Suggested next steps for analysis includeusing the modeling approach to exploreland-use, C dynamics and management ofsecondary forests and plantations, soilC gains or losses, sources ofvariability in the land use and Cstock projections, and other ecologicalimplications and feedbacks resulting fromprojected changes in land cover.     

Global implications of biomass and biofuel use in Germany – Recent trends and future scenarios for domestic and foreign agricultural land use and resulting GHG emissions

The global land area required to meet the German consumption of agricultural products for food and non-food use was quantified, and the related greenhouse gas (GHG) emissions, particularly those induced by land-use changes in tropical countries, were estimated. Two comprehensive business-as-usual scenarios describe the development corridor of biomass for non-food use in terms of energetic and non-energetic purposes. In terms of land use, Germany was already a net importer of agricultural land in 2004, and the net additional land required by 2030 is estimated to comprise 2.5–3.4 Mha. This is mainly due to biofuel demand driven by current policy targets. Meeting the required biodiesel import demand would result in an additional GWP of 23–37 Tg of CO₂ equivalents through direct and indirect land-use changes. Alternative scenario elements outline the potential options for reducing Germany's land requirement, which reflect future global per capita availability.     

Emission trading in agriculture: a study of design options using an agent-based approach

The Scottish Government has proposed reducing Scotland’s greenhouse gas (GHG) emissions by at least 80% by 2050, compared to the 1990 baseline level. It is not yet clear how these reductions will be achieved, but it is likely that all sectors will be expected to make some contribution. Depending on their farm activities, farmers have different sets of abatement alternatives—the challenge facing them, however, is in finding strategies that help to meet reduction targets while maintaining their income. In this paper, we use an agent-based modelling approach to study the implications of carbon trading design options aimed at reducing GHG emissions in the agricultural sector, such as auctions, fixed carbon prices, or carbon credit banking. The feasibility of carbon trading scheme options is assessed regarding their ability to ensure that farmers obtain carbon credits at an affordable and adequate price, since low prices would reward farmers not adopting on-farm abatement options and high prices would encourage non-compliance to targets, thus increasing enforcement costs. Assuming a closed market within the agricultural sector, this study shows that farmers may face up to 50% loss of income to achieve a 30% reduction target if this requires a cut in production. However, market design options such as credits banking may allow farmers to progressively adapt to the scheme constraints. At an individual level, the rate of on-farm compliance and the mandated emission reduction target will determine which farmer strategy is the most efficient to cope with a trading scheme.     

Emissions trading in Hong Kong and the Pearl River Delta region—A modeling approach to trade decisions in Hong Kong's electricity industry

In 2002, the Hong Kong government and the Guangdong provincial government agreed to reduce emissions of sulfur dioxide, nitrogen oxides, respirable suspended particulates, and volatile organic compounds by 40%, 20%, 55%, and 55%, respectively. There was strong public demand for the power stations in Hong Kong to reduce emissions. Emission caps were introduced, with allowances for the trading of emission credits. However, local power stations were using equipment built in the 1980s and 1990s, making it difficult for them to meet the new emissions requirements. The situation presented a new challenge, which involved a choice of either improving the existing equipment, or using emissions trading to meet the emission caps. This study reviews the background on emissions in Hong Kong and the surrounding regions, the “cap and trade” system, and the technologies used for power generation and emission reduction. A modeling approach is adopted to simulate the equipment, the electricity dispatching requirements, and the costs of either reducing emissions or trading emission credits. Data from a power station in Hong Kong was chosen for the simulation. Different options were simulated in the model to identify the optimal strategy. The results were then compared with the plan for emission reduction. This study demonstrates that a modeling approach using linear programming can analyze the complicated options involving emission reduction and investments to achieve an optimized business solution.     

The Kyoto Protocol could make a difference for the optimal forest-based CO2 mitigation strategy: some results from GORCAM

The Kyoto Protocol was agreed on by more than 150 nations in December, 1997 and (if and when ratified) will establish international commitments to reduce emissions of greenhouse gases to the atmosphere. Under the Kyoto Protocol, some of the carbon emissions and removals within the land-use change and forestry sector can be counted toward a country's commitments for greenhouse gas emissions reductions. In addition to the impacts that land-use practices have on CO₂ emissions from fossil-fuel combustion, changes in the carbon stocks of forests (possibly including forest soils) caused by the direct human activities afforestation, reforestation and deforestation and taking place in the `first commitment period' (2008–2012), are to be accounted for under the Kyoto Protocol. Credits for carbon sinks in the biosphere are limited to projects initiated since 1990. A modified version of the model GORCAM has been used to assess eligible emission-reduction credits under the Kyoto regime and to illustrate how the optimal forest-based strategy for carbon dioxide mitigation might change under the provisions of the Kyoto Protocol. The Kyoto Protocol offers rewards for only some of the changes in carbon stocks that might occur and hence the forestry project that produces the most emission reduction credits under the Kyoto Protocol is not necessarily the same project that produces the greatest benefit for net emissions of carbon dioxide to the atmosphere. Supplementing the Protocol with appropriate definitions, interpretations and agreements could help to make sure that it does not provide incentive for activities that run counter to the objectives of the Framework Convention on Climate Change.     

Forgotten carbon: indirect CO2 in greenhouse gas emission inventories

National governments that are Parties to the United Nations Framework Convention on Climate Change (UNFCCC) are required to submit greenhouse gas (GHG) inventories accounting for the emissions and removals occurring within their geographic territories. The Intergovernmental Panel on Climate Change (IPCC) provides inventory methodology guidance to the Parties of the UNFCCC. This methodology guidance, and national inventories based on it, omits carbon dioxide (CO₂) from the atmospheric oxidation of methane, carbon monoxide, and non-methane volatile organic compounds emissions that result from several source categories. The inclusion of this category of “indirect” CO₂ in GHG inventories increases global anthropogenic emissions (excluding land use and forestry) between 0.5 and 0.7%. However, the effect of inclusion on aggregate UNFCCC Annex I Party GHG emissions would be to reduce the growth of total emissions, from 1990 to 2004, by 0.2% points. The effect on the GHG emissions and emission trends of individual countries varies. The paper includes a methodology for calculating these emissions and discusses uncertainties. Indirect CO₂ is equally relevant for GHG inventories at other scales, such as global, regional, organizational, and facility. Similarly, project-based methodologies, such as those used under the Clean Development Mechanism, may need revising to account for indirect CO₂.     

Factoring out natural and indirect human effects on terrestrial carbon sources and sinks

The capacity to partition natural, indirect, and direct human-induced effects on terrestrial carbon (C) sources and sinks is necessary to be able to predict future terrestrial C dynamics and thus their influence on atmospheric CO₂ growth. However, it will take a number of years before we can better attribute quantitative estimates of the contribution of various C processes to the net C balance. In a policy context, factoring out natural and indirect human-induced effects on C sources and sinks from the direct human-induced influences, is seen as a requirement of a C accounting approach that establishes a clear and unambiguous connection between human activities and the assignment of C credits and debits. We present options for factoring out various groups of influences including climate variability, CO₂ and N fertilization, and legacies from forest management. These are: (i) selecting longer accounting or measurement periods to reduce the effects of inter-annual variability; (ii) correction of national inventories for inter-annual variability; (iii) use of activity-based accounting and C response curves; (iv) use of baseline scenarios or benchmarks at the national level; (v) stratification of the landscape into units with distinct average C stocks. Other, more sophisticated modeling approaches (e.g., demographic models in combination with forest inventories; process-based models) are possible options for future C accounting systems but their complexity and data requirements make their present adoption more difficult in an inclusive international C accounting system.     

Carbon accounting for forest harvesting and wood products: review and evaluation of different approaches

Under the United Nations Framework Convention on Climate Change, more than 160 countries are required to report their national greenhouse gas inventories. To help countries meet this requirement, the Intergovernmental Panel on Climate Change (IPCC) prepared guidelines for inventorying greenhouse gases. These guidelines are regularly reviewed to ensure that they are based on the best scientific knowledge. In May 1998, in Dakar, Senegal, an IPCC expert meeting reviewed and evaluated three approaches for accounting for carbon from forest harvesting and wood products. They are the atmospheric-flow, stock-change and production approaches. In the future, governments may decide to include one of the three approaches in the land-use change and forestry module of the IPCC Guidelines for National Greenhouse Gas Inventories. Here, we demonstrate how such approaches can be evaluated using technical, scientific and policy criteria. The purpose of this evaluation is to help policy-makers potentially choose an approach for the IPCC Guidelines. This paper presents the framework of the evaluation by separating the technical and policy issues of each approach, but it does not make policy recommendations. On technical and scientific grounds, a group of experts found that the three approaches gave similar results at a global level. Data availability is not a critical factor in choosing between the approaches. However, at the national level, the approaches can differ significantly, for example, in terms of their system boundaries. Depending on technical features of each approach, credits and debits for CO₂ flows or changes in carbon stock in wood products are accounted for differently among countries that produce or consume wood. This leads to differing incentives for conserving or enhancing carbon stocks in forests, the use of imported wood products and woodfuels and waste minimisation strategies. Each approach has different implications.     

Regulating carbon emissions from indirect land use change (ILUC): U.S. and California case studies

As lifecycle emissions accounting becomes more widely used in policy, it is important to understand how it has been applied. This paper analyses policy-making for two U.S. fuel regulations—the federal Renewable Fuel Standard (RFS) and the California Low Carbon Fuel Standard (LCFS)—that were pioneering not only in using life cycle assessment (LCA) in performance-based environmental regulations, but especially for including emissions from indirect land use change (ILUC). The case studies in this paper focus, in particular, on the decision to include ILUC in lifecycle emissions accounting. Tracing the development of these policies shows the key role of environmental policy entrepreneurs in advocating for ILUC emissions accounting during policy formulation. Moreover, it highlights a paradox in the use of science: although ILUC policy proponents were motivated by best available research, they were also politically enabled by scientific uncertainty and lack of understanding. Understanding this political dimension of decision-making is valuable for scholars as well as practitioners facing similar decisions.