Stratospheric aerosol injection (SAI) has moved from a speculative thought experiment into an active modeling problem. The proposed intervention would place reflective particles, most plausibly sulfur-containing aerosols, in the stratosphere to scatter a fraction of incoming sunlight. The physical premise is straightforward: less absorbed solar energy would cool the surface. The regional climate response is not.
Recent coordinated simulations and increasingly detailed Earth-system models are clarifying why. SAI does not simply subtract a uniform number of degrees from global warming. Aerosols spread around much of the globe, interact with radiation and ozone chemistry, and alter atmospheric circulation. The resulting pattern depends on injection latitude, season, altitude, particle properties, and the emissions pathway being offset.
Cooling is global; consequences are regional
Model experiments generally find that SAI could lower global mean temperature and reduce some land and ocean heat extremes. In scenarios where aerosol forcing is tuned to offset a specified level of greenhouse-gas warming, global temperature metrics can be brought closer to present-day conditions. That apparent success hides substantial geographic variation.
Land areas often cool faster than oceans, while high latitudes can respond differently from the tropics. Seasonal patterns also matter: a reduction in annual mean temperature may coexist with altered hot-season extremes or shifts in the timing of snow and ice melt. Models further suggest that SAI could reduce some precipitation increases driven by greenhouse gases, but it would not restore historical rainfall patterns.
Rainfall and circulation are the hard problem
Greenhouse gases warm the surface and increase the atmosphere’s capacity to hold water vapor, generally intensifying the hydrological cycle. Solar dimming counters part of the energy imbalance but does not reverse the atmospheric effects of elevated carbon dioxide. This mismatch can weaken global mean precipitation responses while leaving regional changes in monsoons, storm tracks, and tropical rainfall.
Simulations commonly project that SAI could suppress or redistribute precipitation in some monsoon regions, with outcomes sensitive to the aerosol deployment strategy. Interhemispheric differences in aerosol loading may shift the position of the Intertropical Convergence Zone—the belt of rising air and heavy rainfall near the equator. Small changes in that circulation can have large consequences for food production and water security.
These are ensemble results, not precise forecasts. Internal variability, model disagreement, and uncertain aerosol microphysics can be as important as the nominal intervention. A strategy optimized to limit global temperature might therefore produce a different regional risk profile from one optimized for precipitation, sea ice, or extreme heat.
Ozone, termination, and the limits of optimization
Sulfate aerosols can accelerate chemical reactions that affect stratospheric ozone, although the magnitude and sign of the response depend on particle composition, altitude, latitude, and background chlorine levels. Newer modeling increasingly treats aerosol chemistry and microphysics explicitly rather than representing aerosols as a simple radiative veil.
Another persistent concern is termination shock. If SAI were deployed for decades and then stopped abruptly, greenhouse-gas concentrations would remain elevated while the masking aerosol burden disappeared. Surface temperatures could rise rapidly, challenging ecosystems and societies. Avoiding that outcome would require sustained deployment and reliable governance over generations.
The current scientific consensus is therefore narrower than either advocates or critics sometimes imply: SAI appears physically capable of reducing global mean warming, but no model demonstrates that it can do so without creating consequential regional winners and losers. It also cannot address ocean acidification, and it would not replace emissions reductions.
What the models can—and cannot—decide
Research is now shifting toward coupled assessments that combine climate physics with impacts on agriculture, health, ecosystems, and inequality. The central question is not whether a chosen aerosol injection rate can hit a global temperature target. It is whether any feasible strategy can reduce overall harm across regions, seasons, and generations—and who would decide what counts as harm.
For now, regional climate modeling is best understood as a boundary-setting exercise. It can identify mechanisms, expose trade-offs, and test the sensitivity of outcomes to deployment assumptions. It cannot establish a safe operating protocol for the atmosphere. That distinction is becoming more important as geoengineering research becomes more technically sophisticated and politically salient.
