全球变化知识资源中心

Integrated assessment models can help in quantifying the implications of international climate agreements and regional climate action. This paper reviews scenario results from model intercomparison projects to explore different possible outcomes of post-2020 climate negotiations, recently announced pledges and their relation to the 2 °C target. We provide key information for all the major economies, such as the year of emission peaking, regional carbon budgets and emissions allowances. We highlight the distributional consequences of climate policies, and discuss the role of carbon markets for financing clean energy investments, and achieving efficiency and equity.

So far, international climate policy has been ineffective in curbing the rise of global greenhouse gas (GHG) emissions. Still, ambitious climate targets such as the 2 °C target require a phase-out of global emissions by the end of the century, and an active participation of all world regions in climate policy¹. Given the many obstacles to global cooperative action on climate change, the question remains how diverse national climate policies can be coordinated and strengthened globally. Within the United Nations Framework Convention on Climate Change (UNFCCC), the Durban Platform for Enhanced Action² provides an important platform for a post-2020 international climate agreement. It contains several innovative elements, most notably a focus on the major economies that goes beyond the traditional divide between Annex I and non-Annex I countries. The Durban platform calls for a new climate treaty to be agreed in 2015 and implemented as early as 2020. The recently announced United States–China climate deal and the EU 2030 climate framework provide encouraging steps forwards, but aligning the incentives of the major emitters in pursuing stringent climate policies remains a challenge. In this Review, we aim at assessing the implications of post-2020 climate policies with specific reference to the major economies. We provide quantitative estimates of regional emission budgets, timing of emission peaking, and distribution of mitigation costs. We examine the role of carbon markets and different burden sharing schemes to alleviate distributional inequalities and finance the investment needs in low carbon mitigation technologies. In order to quantify these policy-relevant variables, we resort to global models.

Integrated assessment models (IAMs) are tools designed to investigate the implications of achieving climate and other objectives in an integrated and rigorous framework. They are numerical models that account for major interactions among energy, land-use, economic and climate systems. Models differ in the economic, technological and sectoral representation and in the way they are solved, with some models maximizing an intertemporal objective function (such as economic activity) and others simulating a set of equilibria (see the Supplementary Material for individual model description and references to documentation). Models generate global long-term scenarios for a number of regions or countries that can be used to inform climate and energy policies and to translate long-term climate objectives into potential medium-term courses of action^{3, 4, 5, 6, 7, 8, 9}. Scenarios from IAMs provide important input to scientific reviews such as the assessment reports of the Intergovernmental Panel on Climate Change (IPCC) and the United Nations Environment Programme (UNEP) Emissions Gap Report. Given the focus of this review on climate mitigation policies, the models reviewed are used to assess the implications of cost-effective policies to achieve a given climate goal (like in the IPCC), rather than to determine the appropriate ambition of such a goal in a cost–benefit setting. In other words, the potential damages from climate change costs are not considered explicitly here, setting our analysis outside the controversial discussion regarding climate impacts and the social cost of carbon.

In order to generate conclusions that are robust to different models' specifications, IAM teams have engaged in model intercomparison projects (MIPs), in which a variety of models implement a common study protocol. Although cross-model comparison literature has developed fast, it has so far mostly reported on global issues^{6, 10, 11, 12}. Information from a MIP regarding the regional impacts of post-2020 climate policies is limited. This Review aims at synthesizing insights from the most comprehensive MIP on this subject, the LIMITS project^{13, 14, 15}. Box 1 provides information about the policy dimensions we assess. Although other MIPs have explored the role of fragmented regional mitigation effort and staged accession to climate cooperation (EMF22¹⁶, AMPERE¹⁷, EMF27¹), globally delayed participation (RECIPE¹⁸, ROSE¹⁹, AMPERE¹¹) and burden-sharing schemes (RECIPE²⁰), none except for LIMITS has focused on potential outcomes of the Durban platform negotiations: that is, a period of fragmented moderate climate policy followed by global cooperative action under different assumptions about burden-sharing regimes. In addition, in LIMITS results are reported at a high regional resolution (for 10 regional aggregates that best match the native model regions), short-term climate and energy policies are well detailed, the likelihood of achieving the 2 °C target is relatively harmonized across model scenarios (using the MAGICC climate model) and a new burden-sharing scheme is introduced and evaluated. Although we will use LIMITS as guiding example throughout the paper, the insights are framed by and compared with all the relevant literature on climate policy modelling^{21, 22, 23, 24, 25, 26, 27}.

Box 1: International climate policy through the lens of IAMs.

International climate policy involves complicated negotiations among different parties over a wide range of activities. As international climate agreements are voluntary, they need to be self-enforcing. The formation of such deals can be studied by model-based analysis of the incentives for joining or leaving these agreements. This has led to a specific strand of literature based on game theory and strategic interaction^{85, 86, 87, 88, 89, 90}, which includes IAM applications^{91, 92, 93, 94, 95, 96}. More often, though, the formation mechanism of the policy agreement is taken as given. Models explore the implications of regional or global policies, comparing them, for instance, with a counterfactual world in which such policies are absent.

The LIMITS MIP can be used to illustrate how this is done in practice. A set of scenarios are implemented in the six participating models (GCAM, IMAGE, MESSAGE, REMIND, TIAM-ECN, WITCH). These include (1) the extent and date of implementation of climate and energy policies, (2) the stringency of the regional emission pledges, (3) the long-term climate objective, and (4) the way the climate policy burden is shared across regions (see Supplementary Table S2 for the scenario description). First of all, a counterfactual scenario with no climate policies is built ('No Policies'). Second, the study analyses a reference case representing the current situation of regionally fragmented mitigation efforts, based on extrapolation of the strengthened Copenhagen pledges throughout the whole century ('Pledges'; see Supplementary Table S3 for their exact definition). In addition, a successful outcome of the Durban Platform negotiations is modelled by global cooperation after 2020 on either a long-term CO₂-equivalent concentration objective of 450 ppm-eq or 500 ppm-eq. Given the uncertainty surrounding climate change, each of these concentration levels produces a probability distribution of temperature outcomes. By using the MAGICC climate model, 450 and 500 ppm-eq targets were found to correspond to a likely (>66%) and as-likely-as-not (>50%) chance of achieving the 2 °C target respectively.

The stabilization scenarios are implemented in a cost-effective way, with emissions reduced where it is cheapest to do so. Different burden-sharing regimes across regions have been considered, to allow regions to be compensated for their emission reductions. Thus, in addition to the case of a globally harmonized carbon tax (without allowing for transfers between regions, we considered the assignment and trade of emissions permits based on either convergence to equal per capita emissions or equalization of mitigation costs across regions (see Supplementary Table S2)).

One of the most valuable uses of integrated assessment models is in the translation of mitigation policies into climate outcomes, and conversely the translation of global climate objectives into regional commitments and timing of emission reductions. This allows the 'when' and 'where' questions that are key elements of climate policy considerations to be addressed.

Figure 1 provides insights into the 'when' question, reporting the year of peaking of greenhouse gas emissions in 10 major economies for different policies (see Supplementary Table S1 for a definition of the 10 regional aggregates). The emission peak year is an important indicator for policy, as it signals by when emissions should start to fall. Without explicit mitigation policies, models project emissions to increase over the century in essentially all regions, although with significant model variation. This result is based on the expectation of continued economic growth and availability of fossil fuels. Mitigation pledges, based on extrapolation of the currently discussed targets²⁸, would lead to differentiated peak years that depend on the stringency of the commitment and on the growth of baseline emissions²⁹. Industrialized economies are projected to keep emissions below current levels, but several developing country regions would see emissions rising until the second half of the century. Emissions in China would peak slightly later than 2030. It should be remarked that not all policy targets are included in the pledge scenario: for example, in the recently announced US–China deal, in addition to peaking emissions in 2030, China also pledged to meet 20% of energy demand with non-fossil sources. Depending on the metric used to convert nuclear and renewables into primary energy, this target is in line with what the LIMITS models foresee in the pledge or 2 °C scenarios.

Year of regional maximum emissions (Kyoto gases; markers show median across models, and lines show 10th–90th percentile ranges). '2100' denotes an increasing emissions trajectory throughout the twenty-first century until the end of the time horizon of the models. Models report information typically in 5-10 year steps. Full set of results by model is available in Supplementary Fig. S1.

A key consideration in climate policy is how to distribute the economic effort of GHG mitigation. Even if global mitigation costs were low, policymakers care and argue about the regional distribution of policy costs, as this affects economic development, competitiveness and even political stability. The scenarios indicate that the costs of mitigation will vary significantly across countries^{1, 14, 20, 54, 55, 56, 57, 58}. Figure 5 portrays this finding for the LIMITS models, showing that—in a cost-effective framework with uniform carbon pricing but without carbon trading or other compensatory transfers— mitigation costs in the OECD would be lower than global average, and the opposite would hold for developing economies, and especially for energy-exporting regions, which would face adverse terms-of-trade effects^{1, 20, 58, 59, 60}. This ranking is rather robust across climate targets, mitigation cost metrics and IAMs^{14, 61}, although the ranges are considerably larger for developing economies.

Relative mitigation costs (percentage reduction of GDP) in each region divided by global mitigation costs (again in percentage reduction of GDP), for 450 ppm-eq without carbon trading and transfers. The line at 100 indicates the case in which all regions would face the same relative policy costs, as in the case of an 'equal cost' burden-sharing scheme. GDP is discounted over the period 2010–2100 at 5%.

The challenge of achieving a comprehensive agreement to reduce emissions is often portrayed as either insurmountable or simply a matter of lack of sufficient political will. Rigorous analysis of the implications of implementing mitigation measures can help in characterizing the subtleties of this challenge, supporting a differentiated view on the future of global climate policy and providing useful insights for policy design and on the negotiation process. Such an analysis needs to focus on all the key emitting regions and account for the uncertainties characterizing emission reduction opportunities.

In this Review, we show that scenarios generated by energy–economy–climate models can help in this task, providing critical information to the ongoing policy debate on a post-2020 climate agreement. The use of MIPs can help to ensure that key uncertainties are taken into account by using a diversity of different models and model assumptions. Reviewing a recent MIP focused on international climate policy in the context of broader literature, we relate short- and long-term climate objectives to key regional indicators such as peaking of emissions, carbon budgets and abatement potentials. Our analysis highlights the main challenges in sharing the economic effort associated with reducing emissions equitably, while showing the importance of regional cooperation towards climate stabilization. The importance and limitation of markets is highlighted. Global carbon markets can alleviate some—but not necessarily all—of the distributional tensions in climate change mitigation. They can also provide much-needed revenues for filling investment gaps in clean energy, and if possible achieving other societal goals. Nonetheless, additional policy instruments will be needed to attain the technological and behavioural transformations to achieve climate stabilization.

The currently discussed targets, including those announced in China, the EU and the United States, are important steps forward; our analyses indicate that additional and more comprehensive efforts would be needed if we hope to keep temperature from exceeding critical thresholds. Still, expanding and strengthening climate cooperation while aligning national interests is by no means straightforward. The numerical estimates by the MIPs reviewed here highlight some critical areas of the climate policy process, which include the regional diversity of mitigation opportunities and costs, the institutional requirements for carbon markets, the best use of climate revenues, the linkages with national policy priorities, and the relevance of clean technology innovation and diffusion. Progress in all these key areas will be needed to motivate enhanced national ambition in reducing emissions in the next decades.

This Review has assessed mitigation challenges and opportunities without considering the regional benefits of reducing GHG emissions, mostly because a robust quantification of the latter is not yet available in the literature. Similarly, some potential additional strategies for dealing with climate change, such as mitigation of short-lived gases, adaptation and geo-engineering, have not been considered in these model exercises. We hope that these topics will also be examined in the near future, using similar common protocols and making use of a large number of integrated assessment models.

1. Blanford, G., Kriegler, E. & Tavoni, M. Harmonization vs. fragmentation: overview of climate policy scenarios in EMF27. Clim. Change 123, 383–396 (2014).
   • CAS
   • URL: