Beyond best-case SRM: Scenarios for a messy reality

Earth System models are increasingly used to explore physical uncertainties to possible Solar Radiation Modification.  However, the primary risks of SRM lie in potential human ability to maintain long-term deployment without interruption, conflict or reduction in carbon mitigation ambtion.  As such, physical and societal SRM risks could be significantly increased in an era of geopolitical fragmentation and institutional volatility

Mitigation scenarios are generally constructed by optimising mitigation costs over decades to limit temperature increase on a century timescale - ignoring political processes like conflicts, or policy reversals.  However, for SRM, these 'fast' human processes can change the climate on a timescale of months, allowing a direct coupling of climate responses and political dynamics.  This makes governance instability a first-order driver of risk.

We present the Solar Radiation Modification Pathway (SRMP) framework, which defines parameters for SRM deployment, including interruption probability, detecability and efficacy of action, regional heterogeneity, and mitigation coupling (moral hazard). This enables a structured exploration of non-ideal futures, including governance failure, geopolitical conflict, and coalition-driven deployments creating unequal outcomes for vulnerable regions.

We illustrate these dynamics with the FaIR simple climate model, coupled to a stochastic SRM algorithm with SRMP-defined failures and feedbacks.  Results show SRM deployment under non-optimal conditions can increase climate damages relative to a non-SRM baseline: termination shocks produce warming rates far exceeding conventional scenarios, while high moral hazard parameters can increase long-term damages and overshoot commitment.

These results demonstrate that 'peak shaving' experiments imply optimistic assumptions of high international cooperation with robust mitigation ambition. The SRMP framework provides a common structure to encourage diverse modeling approaches toward assessment that confronts the full risk space of deployment in both cooperative and uncooperative futures.