Editing IPCC:AR6/WGIII/Chapter-3 (section)

=== 3.2.2 The Utility of Integrated Assessment Models ===

<div id="h2-5-siblings" class="h2-siblings"></div>

Integrated Assessment Models (IAMs) are critical for understanding the implications of long-term climate objectives for the required near-term transition. For doing so, an integrated systems perspective including the representation of all sectors and GHGs is necessary. IAMs are used to explore the response of complex systems in a formal and consistent framework. They cover a broad range of modelling frameworks ( [[#Keppo--2021|Keppo et al. 2021]] ). Given the complexity of the systems under investigation, IAMs necessarily make simplifying assumptions and therefore results need to be interpreted in the context of these assumptions. IAMs can range from economic models that consider only carbon dioxide emissions through to detailed process-based representations of the global energy system, covering separate regions and sectors (such as energy, transport, and land use), all GHG emissions and air pollutants, interactions with land and water, and a reduced representation of the climate system. IAMs are generally driven by economics and can have a variety of characteristics such as partial-, general- or non-equilibrium; myopic or perfect foresight; be based on optimisation or simulation; have exogenous or endogenous technological change amongst many other characteristics. IAMs take as input socio-economic and technical variables and parameters to represent various systems. There is no unique way to integrate this knowledge into a model, and due to their complexity, various simplifications and omissions are made for tractability. IAMs therefore have various advantages and disadvantages which need to be weighed up when interpreting IAM outcomes. Annex III.I contains an overview of the different types of models and their key characteristics.

Most IAMs are necessarily broad as they capture long-term dynamics. IAMs are strong in showing the key characteristics of emission pathways and are most suited to questions related to short- versus long-term trade-offs, key interactions with non-climate objectives, long-term energy and land-use characteristics, and implications of different overarching technological and policy choices (Clarke et al. 2014; [[#Rogelj--2018|Rogelj et al. 2018]] a). While some IAMs have a high level of regional and sectoral detail, for questions that require higher levels of granularity (e.g., local policy implementation) specific region and sector models may be better suited. Utility of the IAM pathways increases when the quantitative results are contextualized through qualitative narratives or other additional types of knowledge to provide deeper insights ( [[#Geels--2016a|Geels et al. 2016a]] ; [[#Weyant--2017|Weyant 2017]] ; [[#Gambhir--2019|Gambhir et al. 2019]] ).

IAMs have a long history in addressing environmental problems, particularly in the IPCC assessment process ( [[#van%20Beek--2020|van Beek et al. 2020]] ). Many policy discussions have been guided by IAM-based quantifications, such as the required emission reduction rates, net zero years, or technology deployment rates required to meet certain climate outcomes. This has led to the discussion about whether IAM scenarios have become performative, meaning that they act upon, transform or bring into being the scenarios they describe ( [[#Beck--2017|Beck and Mahony 2017]] , 2018). Transparency of underlying data and methods is critical for scenario users to understand what drives different scenario results ( [[#Robertson--2020|Robertson 2020]] ). A number of community activities have thus focused on the provision of transparent and publicly accessible databases of both input and output data ( [[#Riahi--2012|Riahi et al. 2012]] ; [[#Huppmann--2018|Huppmann et al. 2018]] ; [[#Krey--2019|Krey et al. 2019]] ; [[#Daioglou--2020|Daioglou et al. 2020]] ), as well as the provision of open-source code, and increased documentation (Annex III.I.9). Transparency is needed to reveal conditionality of results on specific choices in terms of assumptions (e.g., discount rates) and model architecture. More detailed explanations of underlying model dynamics would be critical to increase the understanding of what drives results ( [[#Bistline--2020|Bistline et al. 2020]] ; [[#Butnar--2020|Butnar et al. 2020]] ; [[#Robertson--2020|Robertson 2020]] ).

Mitigation scenarios developed for a long-term climate constraint typically focus on cost-effective mitigation action towards a long-term climate goal. Results from IAM as well as sectoral models depend on model structure ( [[#Mercure--2019|Mercure et al. 2019]] ), economic assumptions ( [[#Emmerling--2019|Emmerling et al. 2019]] ), technology assumptions ( [[#Pye--2018|Pye et al. 2018]] ), climate/emissions target formulation ( [[#Johansson--2020|Johansson et al. 2020]] ), and the extent to which pre-existing market distortions are considered ( [[#Guivarch--2011|Guivarch et al. 2011]] ). The vast majority of IAM pathways do not consider climate impacts ( [[#Schultes--2021|Schultes et al. 2021]] ). Equity hinges upon ethical and normative choices. As most IAM pathways follow the cost-effectiveness approach, they do not make any additional equity assumptions. Notable exceptions include [[#Tavoni--2015|Tavoni et al. (2015)]] , [[#Pan--2017|Pan et al. (2017)]] , [[#van%20den%20Berg--2020|van den Berg et al. (2020)]] , and [[#Bauer--2020|Bauer et al. (2020)]] . Regional IAM results therefore need to be assessed with care, considering that emissions reductions are happening where it is most cost-effective, which needs to be separated from who is ultimately paying for the mitigation costs. Cost-effective pathways can provide a useful benchmark, but may not reflect real-world developments ( [[#Calvin--2014a|Calvin et al. 2014a]] ; [[#Trutnevyte--2016|Trutnevyte 2016]] ). Different modelling frameworks may lead to different outcomes ( [[#Mercure--2019|Mercure et al. 2019]] ). Recent studies have shown that other desirable outcomes can evolve with only minor deviations from cost-effective pathways ( [[#Bauer--2020|Bauer et al. 2020]] ; [[#Neumann--2021|Neumann and Brown 2021]] ). IAM and sectoral models represent social, political, and institutional factors only in a rudimentary way. This assessment is thus relying on new methods for the ''ex'' ''post'' assessment of feasibility concerns ( [[#Jewell--2020|Jewell and Cherp 2020]] ; [[#Brutschin--2021|Brutschin et al. 2021]] ). A literature is emerging that recognises and reflects on the diversity and strengths/weaknesses of model-based scenario analysis ( [[#Keppo--2021|Keppo et al. 2021]] ).

The climate constraint implementation can have a meaningful impact on model results. The literature so far includes many temperature overshoot scenarios with heavy reliance on long-term CDR and net negative CO 2 emissions to bring back temperatures after the peak ( [[#Rogelj--2019b|Rogelj et al. 2019b]] ; [[#Johansson--2020|Johansson et al. 2020]] ). New approaches have been developed to avoid temperature overshoot. The new generation of scenarios show that CDR is important beyond its ability to reduce temperature, but is essential also for offsetting residual emissions to reach net zero CO 2 emissions ( [[#Rogelj--2019b|Rogelj et al. 2019b]] ; [[#Johansson--2020|Johansson et al. 2020]] ; [[#Riahi--2021|Riahi et al. 2021]] ; [[#Strefler--2021b|Strefler et al. 2021b]] ).

Many factors influence the deployment of technologies in the IAMs. Since AR5, there has been fervent debate on the large-scale deployment of bioenergy with carbon capture and storage (BECCS) in scenarios ( [[#Fuss--2014|Fuss et al. 2014]] ; [[#Geden--2015|Geden 2015]] ; [[#Anderson--2016|Anderson and Peters 2016]] ; [[#Smith--2016|Smith et al. 2016]] ; [[#van%20Vuuren--2017|van Vuuren et al. 2017]] ; [[#Galik--2020|Galik 2020]] ; [[#Köberle--2019|Köberle 2019]] ). Hence, many recent studies explore mitigation pathways with limited BECCS deployment ( [[#Grubler--2018|Grubler et al. 2018]] ; [[#van%20Vuuren--2019|van Vuuren et al. 2019]] ; [[#Riahi--2021|Riahi et al. 2021]] ; [[#Soergel--2021a|Soergel et al. 2021a]] ). While some have argued that technology diffusion in IAMs occurs too rapidly ( [[#Gambhir--2019|Gambhir et al. 2019]] ), others argued that most models prefer large-scale solutions resulting in a relatively slow phase-out of fossil fuels ( [[#Carton--2019|Carton 2019]] ). While IAMs are particularly strong on supply-side representation, demand-side measures still lag in detail of representation despite progress since AR5 ( [[#Grubler--2018|Grubler et al. 2018]] ; [[#Lovins--2019|Lovins et al. 2019]] ; [[#van%20den%20Berg--2019|van den Berg et al. 2019]] ; [[#O’Neill--2020b|O’Neill et al. 2020b]] ; [[#Hickel--2021|Hickel et al. 2021]] ; [[#Keyßer--2021|Keyßer and Lenzen 2021]] ). The discount rate has a significant impact on the balance between near-term and long-term mitigation. Lower discount rates &lt;4% (than used in IAMs) may lead to more near-term emissions reductions – depending on the stringency of the target ( [[#Emmerling--2019|Emmerling et al. 2019]] ; [[#Riahi--2021|Riahi et al. 2021]] ). Models often use simplified policy assumptions ( [[#O’Neill--2020b|O’Neill et al. 2020b]] ) which can affect the deployment of technologies ( [[#Sognnaes--2021|Sognnaes et al. 2021]] ). Uncertainty in technologies can lead to more or less short-term mitigation ( [[#Grant--2021|Grant et al. 2021]] ; [[#Bednar--2021|Bednar et al. 2021]] ). There is also a recognition to put more emphasis on what drives the results of different IAMs ( [[#Gambhir--2019|Gambhir et al. 2019]] ) and suggestions to focus more on what is driving differences in result across IAMs ( [[#Nikas--2021|Nikas et al. 2021]] ). As noted by Weyant (2017, p. 131), ‘IAms can provide very useful information, but this information needs to be carefully interpreted and integrated with other quantitative and qualitative inputs in the decision-making process.’

<div id="3.2.3" class="h2-container"></div>

<span id="the-scenario-literature-and-scenario-databases"></span>