Scenarios serve as a critical tool in climate change analysis, enabling the exploration of future evolution of the climate system, climate impacts, and the human system (including mitigation and adaptation actions). This paper describes the scenario framework for ScenarioMIP as part of CMIP7. The design process has involved various rounds of interaction with the research community and user groups at large. The proposal covers a set of scenarios exploring high levels of climate change (to explore high-end climate risks), medium levels of climate change (anchored to current policy), and low levels of climate change (aligned with current international agreements). These scenarios follow very different trajectories in terms of emissions, with some likely to experience peaks and subsequent declines in greenhouse gas concentrations in this century. An important innovation is that most scenarios are intended to be run, if possible, in emission-driven mode, providing a better representation of the Earth system uncertainty space. The proposal also includes plans for long-term extensions (up to 2500 AD) to study long-term impacts, climate change-related processes on long timescales, and (ir)reversibility. This proposal forms the basis for further implementation of the framework in terms of the derivation of emissions and land use pathways for use by Earth system models and additional variants for adaptation and mitigation studies.

Scenarios serve as a critical tool in climate change analysis. Defined as plausible alternative descriptions of how the future may develop based on a coherent set of assumptions, they are used by different research communities to explore potential future avenues of socio-economic conditions, assess the effects of different drivers of climate change, characterize future climatic conditions, and assess impacts of climate change as well as adaptation and mitigation responses. As such, scenarios are also useful to bridge across different research communities. The World Climate Research Programme’s (WCRP) Coupled Model Intercomparison Project (CMIP) has organized several rounds of the development and use of such scenarios. The CMIP scenarios play not only a pivotal role in climate research but also act as integrating tools for scientific assessment processes and policy analysis. In the most recent phase 6 (CMIP6), CMIP delegated the organization of experiments targeting specific research areas and questions to the scientific community, and CMIP6-Endorsed Model Intercomparison Projects were established (Eyring et al., 2016). The Scenario Model Intercomparison Project (ScenarioMIP) formed a primary activity within CMIP6 that facilitated multi-model climate projections based on alternative plausible forcing scenarios that are directly relevant to societal concerns regarding climate change mitigation, adaptation, and future climate impacts and risks1 (O’Neill et al., 2016; Tebaldi et al., 2021).

In this continuing role, ScenarioMIP’s goal for CMIP7 is the design of a limited set of scenario-based experiments to be run by climate models, that (1) serve direct science questions, (2) function as input to other (science) communities and (3) support policy. With regard to the direct science contribution, the scenario information is used to study and understand climate processes and how their response to anthropogenic forcings emerges from internal variability and model structural uncertainties. To accomplish this aim, it is important that the set of forcing variables considered is internally consistent and varies over a plausible range of forcing levels, and that the set explores futures that may result in very different climate process dynamics (such as increasing and declining concentrations). With regard to the contribution to other (science) communities, ScenarioMIP ensures that data becomes available about future changes in climate variables (such as temperature, precipitation, humidity, etc.) together with information on human forcings/drivers (such as population, economic activity, land use, etc.) to a diverse set of user communities beyond the physical climate sciences. This aim is to facilitate further understanding of climate change, its impacts, risks and response options, including adaptation and mitigation choices. Targeted communities include, for instance, researchers on impacts and mitigation, but also practitioners, who might use this information for national or sub-national or local risk assessment, climate finance, mitigation policy or adaptation planning. Finally, regarding policy, ScenarioMIP aims to provide relevant outputs to support climate policy development, including through their use in IPCC assessments. Despite these important roles, climate modeling groups participating in ScenarioMIP can only run a limited set of scenarios. Computational expenses associated with setting up, running and archiving output from climate model experiments pose strict constraints on the number of scenarios that ScenarioMIP’s protocol can include. Therefore, a small set of scenarios needs to be selected such that the set satisfies the three critical goals but does not require resources that are beyond reach for current scientific and computational capacity. […]

We have proposed a limited set of scenarios that, modeled by IAMs, would provide pathways of emissions, concentrations, and land use for use as external forcings for ESMs participating in CMIP7. These scenarios with their climate system outcomes are intended to facilitate studies of climate processes; provide inputs to a wide range of studies of impacts, adaptation, and mitigation; and support climate policy development and communication. They cover a wide range of plausible outcomes, from a High scenario representing policy failure and strong drivers of greenhouse gas emissions, to a Very Low scenario with dramatic emissions reductions.

The implementation of these scenarios in IAMs and ESMs poses a number of challenges, not all of which can be addressed or fully explored within the ScenarioMIP process. As a result, in a number of cases it became necessary to make design choices on the basis of current literature. Throughout the article, we have indicated where further research could advance the field and summarize important directions here.
Produce wider set of IAM scenarios:

Produce idealized or counterfactual emissions and land use scenarios to complement the plausible scenarios proposed here. The ScenarioMIP design is limited to scenarios judged to be plausible, but other scenarios could also provide useful insight. For example, scenarios that are higher or lower than those proposed here, or that imagine counter-factual historical outcomes (Meinshausen et al., 2024), can be of scientific interest or policy relevant.
Develop alternative implementations of the scenarios requested here from the IAM community using different modeling paradigms and normative assumptions. There are a number of questions of wide interest to the scientific and policy communities that would be unlikely to imply forcing pathways different enough to substantially affect climate system outcomes, given the role of ScenarioMIP. They therefore fall outside the scope of ScenarioMIP but are worth pursuing through other means. For example, different normative bases for approaches to mitigation may lead to different regional or sectoral emissions pathways, within approximately the same global forcing pathway. Similarly, alternative socio-economic pathways such as sustainable development scenarios with rapid convergence of regional incomes and living standards are useful to pursue but, if coupled with a target of limiting warming to (for example) likely below 2 °C, would lead to similar global forcing pathways to other 2 °C scenarios. […]