If solar radiation modification (SRM) were ever deployed, how would we know that it worked? Answering this requires more than detecting a change: the change must also be attributed to the intervention rather than to natural variability or to other forcings. In this ongoing work, I develop a framework for SRM detectability that integrates detection and attribution into a single measure, and apply it to stratospheric aerosol injection (SAI) scenarios simulated by three Earth system models. It extends the detectability thinking I applied to marine cloud brightening (MCB) through the inadvertent aerosol perturbation of the IMO2020 shipping fuel regulation.

The Framework Detection asks whether a perturbed climate differs from the background state that preceded it; attribution asks whether that difference is due to the intervention, judged against a counterfactual simulation without SRM. Scoring both together, as a true detection rate against a false detection rate, yields a detectability score that penalizes false alarms as well as missed detections. I apply this to top-of-atmosphere shortwave reflection (monthly, 1° resolution) in CESM2-WACCM, E3SMv3, and UKESM1-1, using two GeoMIP-style scenarios branched from SSP2-4.5 in 2035: injection in the subtropics (30°N/S at 21 km) and at high latitudes (60°N/S at 15 km).

The Science Less than half of the globe is detectable, and where detection is possible it takes about 10 years on average, given a decade-long background record. The high-latitude injection scenario both enlarges the detectable area and shortens the average time to detectability. Spread across models is non-negligible, particularly at regional scales, which means that the confidence with which an intervention could be verified depends on the model used to assess it. These results echo the difficulty of detecting even a large, abrupt, real-world radiative forcing, as found for IMO2020, and underscore that monitoring requirements are an integral part of evaluating any climate intervention proposal.

References:

  • J. Zhang and G. Feingold: A framework for assessing SRM detectability: integrating detection and attribution. In preparation.
  • J. Zhang, Y.-S. Chen, E. Gryspeerdt, T. Yamaguchi, and G. Feingold (2025): Radiative forcing from the 2020 shipping fuel regulation is large but hard to detect. Commun. Earth Environ., 6(18), 1–11. doi:10.1038/s43247-024-01911-9