This uncertainty does not address questions about the likelihood of extended interruption, rather what the consequences would be. The main risk associated with interruption would be a rapid rate of temperature increase and associated climate impacts. Interruptions would be reversible with limited consequences, so long as injections are restarted in a reasonable amount of time, and the length of this time period depends on the rate at which warming occurs when deployment ceases.
Metric
Global mean temperature rises by more than 0.3°C, 2 years into deployment interruption, given interruption at the 10th year of a deployment cooling by 0.5C
Uncertainty
Around 50% of the equilibrium response to a forcing (including the loss of SAI forcing) is expected after 10 years (Parker & Irvine, 2018), but this fraction is dependent on how long SAI has been deployed for and at what magnitude. Using the emulator from Farley et al. (2024): If an interruption occurs after 0.5C cooling 10 years into deployment, around 0.2C of cooling is lost 2 years after deployment ceases.
Decision relevance
If warming due to interruptions occur at a more rapid pace, the direct impacts of this rate are harder to predict due to being faster than normal climate variables. For example, Pinatubo cooled the globe by 0.4°C in 2 years (McCormick et al., 1995). A faster rate of warming after interruption also makes the timeline for when injected must be restarted shorter.