Advancing holistic risk assessment for measures to address climate change

Stratospheric aerosol injection (SAI) has been the subject of increasing scrutiny as a potential climate measure, with most scientific attention focused on its efficacy in reducing net global warming. However, knowing that SAI would likely be effective does not answer the question of whether it is a “good idea” as a climate measure: whether the deployment of SAI would reduce the overall detrimental impacts of Earth heating relative to scenarios in which it is not deployed. Attempts to answer this question face two major obstacles. First, investigations of the physical response to SAI have mostly focused on its ability to reduce global mean temperatures; however for outcomes directly related to human impacts such as drought, air quality, or weather extremes, SAI is expected to produce different results than cooling through a reduction of greenhouse gases. Deeper engagement from Earth scientists is required to assess these impacts in relation to cooling from SAI versus GHG reduction. Second, producing ever more detailed scenarios of SAI deployment is not sufficient to produce effective risk analysis. It centralises SAI and its physical effects as the primary question driving future decisions, ignoring the social, political and economic dimensions. This can lead to studies which treat representative scenarios of SAI (or its absence) as comparable predictions of the future, rather than as indicative simulations which can help us to understand physical differences only. It also limits Earth scientists’ understanding of which physical risks are likely to be most – or least – consequential, and therefore which responses most urgently need more research. A research framework is needed which integrates risk analysis with Earth system modelling. This would allow risk analysts to more effectively explore the consequences of different climate measures while guiding researchers in Earth science towards the unanswered gaps in SAI modelling which most affect assessments of future risk. We therefore propose a new framework for physical modelling, designed to provide the information needed for cross-disciplinary risk analysis. Rather than continuing to build and refine scenarios, we instead choose to focus on a limited set of (mostly existing) scenarios, and to develop an understanding of the relationship between SAI and physical outcomes in ways which can inform risk analysis. We complement these scenarios with simplified sensitivity simulations to improve our understanding of the physical responses of critical systems to different climate measures (including SAI strategies, termination shock, SAI-based peak shaving, emergency SAI deployment, and different emissions mitigation levels). These responses are then translated into inputs for holistic risk assessment as a collaborative task with experts in risk analysis, based on the key factors which are expected to be relevant in future scenarios – centralizing the issue of how different risks are mitigated, exacerbated, or compounded by different climate measures, rather than starting from the question of SAI deployment. This project will set the standard for future assessments of all climate measures. By integrating risk analysis into our understanding of SAI, we ensure that the end results will be an improved understanding of the physical impacts of different climate measures which is guided by experts in risk analysis. This then enables those experts to provide holistic risk-risk analysis of different futures with and without SAI, and we anticipate our framework will continue to grow and provide policy-relevant information far beyond the end of this programme.