IAMs typically poorly represent resources and their uses or feature an aggregation of economic sectors that is too high to provide valuable systemic circularity insights, related to primary and resource-intensive manufacturing sectors, and they mostly focus on flows while disregarding material stocks. In contrast, resource flow models only consider material energy carriers without explicit representation of the energy system, also without representing climate feedbacks or climate-explicit policies. This cultivates the need for ambitious whole-system modellingto fully understand climate action in consideration of resource, material, monetary, and energy efficiency and stocks, technical progress, demand shifts, rebound effects, and inter-, intra-, and cross-sectoral flows. We address this research gap, by first understanding what circularity dimensions are covered by the project’s IAMs and determine the geographic and system boundariesthat will revolve around critical economic sectors (metals, cement, AFOLU, food, etc.). Acknowledging that no single IAM offers the required comprehensive representation of end-to-end supply chains and the profound understanding of the physical and monetary flows throughout the economy, we will create integrated frameworks that soft-link models with different degrees of representation of the anthroposphere, materiality of production and consumption, and measure of assessment in physical and monetary terms.
Three advanced IAMs will be linked with material flows. First, critical manufacturing sectors (steel, cement, aluminium) will be mapped onto key demand sectors (transport, buildings, agriculture, power generation, water, infrastructure, etc.), and mechanisms leading to emission reductions will be described considering limiting, slowing, regenerating, and recirculating material flows. To better represent production processes that are relevant for both circularity performance and climate change mitigation as well as to create the foundations for running scenarios on the macro impacts of different circular economy interventions, we will make use of the UPDATE & UPGRADE advancementsin terms of representation of critical “technologies” that are currently missing in IAMs, before eventually soft-linking the three IAMs with the ENGAGE-Materials model. ENGAGE is the most advanced global CGE model representing steel in a circularity context, which will be further expanded to cement and aluminium. These links will allow to assess economic implications of different decarbonisation pathways/net-zero targets in an integrated manner with circular economy policies, helping to highlight synergies, trade-offs, and rebound effects, as well as energy, emissions, and investment impacts of decarbonisation scenarios in the steel and cement industries. This will lead to new IAM capabilities, as economic activities related to circularity will be incorporated (material recycling, repair, share and re-use), eventually enabling to quantify how relevant interventions can contribute to energy and emissions cuts, while increasing the process resolution for key circular economy sectors.
Nature-based solutions (NbS) are identified as promising options to mitigate and adapt to climate change while helping address other global environmental change challenges. While IAMs deploy some NbS to achieve climate mitigation targets, the input parameters and model structures underpinning need improvement. For example, IAM decarbonisation scenarios have been criticised for over-reliance on negative emissions through bioenergy with carbon capture and storage (BECCS). The preference for BECCS is in large part explained by the structures of models, namely the options that are available in the models and, importantly, those that are not. Energy system and economic models usually have limited representation of land use dynamics and agriculture. General equilibrium models have an aggregated economic representation of the AFOLU sectors based on economic value of produced commodities and inputs (including land) and focus on CO2 emissions only. Linking such models to a land-use and agriculture model would allow for inclusion of non-CO2 emissions and bioenergy supply constraints in an explicit manner. While the spatial distribution of the activities is important at the grid cell level, most IAMs are not spatially explicit and work at a regional aggregation.
DIAMOND will create scenarios of bioenergy production and demand as well as land use emissions using the gridded land-use model MAgPIE at various regional configurations of IAMs in the project. In addition, many IAMs are in fact framework combinations of various sub-models, usually an energy-economic model coupled with a land use and agriculture model. These partial equilibrium integrated models are often soft-linked via emissions from AFOLU, carbon prices, and bioenergy demand and supply. The bioenergy is usually aggregated into a single total bioenergy category or is divided into 1st and 2nd generation and agroforestry residues. DIAMOND will instead create detailed links that differentiate between types of primary bioenergy that can be used inside partial equilibrium models that already have a bioenergy supply chain but no links to a spatially explicit land-use module. For models that differentiate between agricultural commodities, we will also provide MAgPIE projections for these commodities. Finally, while global IAMs are useful in providing insights into the structural changes that must happen at sectoral levels, their outputs need to be expanded to provide information at a higher granularity, including sectors that may not be represented. In line with other INTEGRATE & EXPAND activities, DIAMOND will link MAgPIE to bottom-up engineering models of agricultural production at farm-level to explore trade-offs/synergies that may arise in the transition. To maximise the added value of these links, we will also improve MAgPIE itself. Its economic parameters will be better informed by more detailed modelling that considers a wider range of options for agricultural production—e.g., land-sparing approaches that minimise land demand while delivering sustainable, affordable, and sufficient nutrition to the world’s growing population.
IAMs generally produce emission pathways, and other modellers use those pathways to, in turn, model the resulting climate response. This approach has four significant challenges. First, IAMs are often calibrated with different emissions data and need to be harmonised with agreed historical data before climate modelling can commence (e.g., Gidden et al., 2019). Second, not all IAMs consider all climate-relevant emissions and instead require infilling to undertake a climate assessment (e.g., Lamboll et al., 2020). Third, reduced-complexity climate models are typically used for climate assessment, but these need to be calibrated to reproduce the historical period and model outputs of more complex models (e.g., Nicholls et al., 2020). Fourth, undertaking the entire pipeline (harmonisation, infilling, and climate modelling) is an extensive process, which may exclude many smaller modelling groups that do not have the skills or resources, which ultimately reduces the diversity of modelling outcomes assessed in the literature, even if some of the necessary components of the pipeline are available as open-source software. Despite some of the recent efforts to improve the climate response modelling pipeline, there has been insufficient analysis of the sensitivity to different harmonisation and infilling assumptions. There are also challenges to reporting and assessing climate outcomes from IAMs that are heavily based on synthetic data (such as with infilling). Scenario databases used to generate the synthetic data also come with significant biases in terms of models available and scenario protocols used.
DIAMOND will allow a systematic analysis of the climate response pipeline from harmonisation, infilling, and climate modelling, with a more exhaustive sensitivity analysis than previously undertaken. It will conduct a series of analyses to systematically assess the effects of infilling and harmonisation in a perfect model framework; using existing tools (ANERIS – harmonisation, SILICONE – infilling, OpenSCM – climate response uncertainty), we will assess the potential for bias in existing pipelines where multi-gas information is known and explore the capacity to represent uncertainties in modified frameworks. We will deliver an approach to providing a set of ‘storyline’ scenarios that will span a plausible set of infilled solutions to a given sparse scenario description, allowing native assessment of uncertainties in climate outcome and impact assessments for a given scenario class, thus reflecting the uncertainties associated with infilling, harmonisation, and climate response.
Toward adequately modelling and thereby enhancing coherence and accelerating a transition towards climate neutrality, IAMs must be further developed in directions, in which important barriers, feedback loops, and opportunities for the climate transition exist. Apart from materials, recent literature has identified finance and labour as potential key constraints on the transition. As far as these two areas are concerned, the constraints are not so much quantitative, at least not in the sense that there is too little finance or labouravailable. Instead, the problem is of a qualitative mismatch nature as labour often does not have the skills required for the transition and finance is not directed towards clean technologies. This has hitherto been complicated to build into the existing model structures, since large, complex IAMs are not well versed to handle diversity and qualitative mismatches in capital and labour markets.The challenge, therefore, is to find the right balance between simplifying assumptions to keep models tractable and manageable, on the one hand, and bringing in realistic and important diversity and mismatch in labour and financial markets, on the other. The latter is critical to advancing our quantitative understanding of the requirements towards a net zero future, as qualitative mismatches in finance and skills can slow down the speed, or drive up the costs, of the transition considerably.
DIAMOND will develop methods for achieving this balance in these two areas. This enriched representation of financial markets will allow the advanced IAMs to trace qualitative shortages in the supply of certain types of finance across pathways to net zero. These enhanced model runs will inform where policymakers may intervene in cost-effective ways to keep the climate transition going forward in different stages of the process. Moreover, this richer representation of finance will allow for a more holistic analysis of the impacts of “green” monetary policy, including post-COVID recovery funds in EU and globally. Similarly, models of skill-biased technical change and endogenous human capital accumulation and education will enable linking the demand for workers in blue-collar and white-collar green, brown, and other jobs to workers with various skills. Keeping track of this diverse set of worker types, as well as linking the demand and supply for them endogenously to climate transition scenarios, will help policymakers to identify likely shortages in the labour market that may hinder or stall the transition, even if finance is not a constraint. Keeping a close eye on employment/shifting patterns of labour demand and required skills will allow to identify vulnerable worker groups that will require support and help design appropriate educational programmes and re-skilling strategies that ensure a just, inclusive, and broadly supported transition to net-zero.Finally, another aspect of DIAMOND is to incorporate behaviour change in modelling as well as to codesign model development and use with all stakeholder groups, including the civil society. Societal transformations are deemed to link mobilisation, network formation, and institution building for sustainability transitions, as well as to interact with state- and market-led transformations in many ways. Political agency is central to these mechanisms, as it enables to challenge assumptions and seriously consider alternatives that may be invisible to (or uncontested by) the mainstream view and modelling science, such as post-growth realities, and how climate action reconciles with these trajectories. In this vivid discussion on the need for—and/or impossibility of—continuous economic growth, DIAMOND will broaden the concept of welfare beyond GDP, disentangling the two in the light of climate change mitigation.
Growing attention has recently been paid to aspects of feasibilityin IAMs. This encompasses the feasibilities of ‘what’ can happen to mitigate climate change, but also of‘where’ and ‘when’ it can happen (under which circumstances), as well as for ‘whom’. The latter, however, is highly intertwined with largely overlooked aspects of preferences and desirability, which are increasingly viewed as a critical research gap for integrated assessment modelling science. IAM specialists have long acknowledged the predominant focus of modelling practice on supply-side transformations as well as the importance of end-use transformations for effective mitigation. While some behavioural aspects have been introduced in IAMs, these remain indirectly touched upon as simplified exogenous assumptions. Most IAMs are based on the core assumption that citizens are rational agents that aim to optimise individual utility based on all available information. This assumption has been challenged by a plethora of research from the psychological sciences, demonstrating that individuals are boundedly rational and that decision preferences vary substantially between individuals. Recent research has started to integrate individual heterogeneity in IAMs to better account for non-rational decision outcomes.
DIAMOND will contribute to these efforts by developing a comprehensive framework of climate-relevant decisions by integrating fine-grained decision preferences (e.g., risk preference) and other behavioural aspects into existing IAMs. Specifically, the approach will involve collecting cross-national data to capture heterogeneity in preferences and behaviour that can then be integrated into existing IAMs to model specific consumer/citizen groups. Given the future-oriented focus on IAMs, such a framework will not only include stable factors but also dynamic mechanisms of behaviour change. The goal of this contribution is to address the lack of consideration of behavioural aspects in IAMs by integrating insights from the social sciences and humanities.Further model improvements will further allow to gain more focused insights into the socioeconomic impacts of decarbonisation, which constitute a critical issue in the policy debate and can only be assessed with better model representation of household heterogeneity. The development of different income and household groups, which exhibit different reaction ‘parameters’, will enable a targeted quantification of the objective and subjective distributive effects of climate change and policies. Achieving a better representation of heterogeneity across households will yield policy-relevant outputs for energy poverty as well as adjustments in energy and non-energy consumption. Parameterisation and calibration of the advanced IAMs will be based on modules of the WILIAM model; hard linkages with the FIDELIO input-output model; soft linkages with the CHANCE computable equilibrium – microsimulation modelling framework to capture the heterogeneity of households in their expenditure and income patterns, allowing for detailed incidence analysis; soft links with agent-based modelling, to allow relevant heterogeneity to affect aggregate outcomes, validate and calibrate the parameterisation of the heterogeneous labour and finance matching modules, and thus trace in detail the underlying welfare and distributional implications.
Last but not least, IAMs typically consider a very high-level description of the electricity system, thereby missing an accurate picture of the entire network. This includes significant deployment of variable and intermittent renewables, storage, and new grid infrastructure to support interconnectivity and cross grid balancing. These transmission grid investments are critical for the system to work efficiently as a whole, avoiding congestions and profiting from the synergies arising from spatial correlation—or, better, decorrelation—of the availability of renewable sources.
DIAMOND will develop an interoperable electricity module based on the Open Generation and Transmission Operation and Expansion Planning Model with renewable energy sources (RES) and energy storage systems (ESS), the open-source version of the TEPES electricity system model that was developed as a part of H2020 openENTRANCE, including the main elements of the physics of power networks, and a comprehensive representation of uncertainty, as well as phase-shifting and power control technologies. It will link this to the advanced IAMsand assess the produced scenarios based on their network and infrastructure needs. The analysis will identify no-regret investments common among different scenarios and constitute the basic architecture of a European electricity expansion plan, making the scenarios more real-world representative of the European electricity dynamics.