Crop yield response

Because crop yields are central to food security and economic resilience, their sensitivity to SAI makes vegetation response one of the most consequential impacts of stratospheric aerosol injection (SAI). SAI alters crop yields primarily through changes in precipitation, diffuse radiation (sunlight scattered by aerosols rather than arriving as a direct beam), temperature, and nutrient availability. Current evidence suggests that SAI would likely have a ~10% or more increase in global crop yields relative to background warming scenarios (Clark et al., 2023; Fan et al., 2021). However, these responses carry significant uncertainty at the local and regional scales.

Uncertainty exists in the magnitude of global crop yield changes, but the localized precipitation changes that may decrease local crop yields matter most. Modeling uncertainty is high for these impacts, making them hard to predict accurately.

Other uncertain impacts include a plausibly large diffuse-light fertilization effect (the possibility that more scattered sunlight lets plants photosynthesize more efficiently). Models show diffuse radiation increasing by 11–50% under SAI (Xia et al., 2016; Yang et al., 2020). How much this mechanism actually enhances crop and vegetation growth remains unclear, since its magnitude and representation vary widely across models. Additionally, models might not fully capture the potential impact of nitrogen limitation on plant production – cooler soils have reduced nitrogen supply which may reduce fertilization benefits. (Duan et al., 2020; Parry et al., 2026). Mechanistically understanding these competing drivers as well as better modeling regional precipitation changes — and how they interact across different SAI strategies, regions, and crop systems — is essential for assessing whether SAI could create new food security risks.

This research activity will fund research over multiple years to develop mechanistic understanding of how SAI affects vegetation and crop growth and to identify which climate drivers dominate crop yield responses across different regions.

This work could involve a variety of approaches, but will most likely entail some combination of the following:

1. Simulating yields using offline crop models driven by the downscaled outputs of climate models. This allows for the translation of changes to surface climate on a local scale under SAI into estimates of changes to crops. 
2. Developing mechanistic understanding of vegetation drivers, specifically, changes in crop yields and the nutritional value of those crops under SAI. This includes research aimed at better understanding how a novel climate under SAI impacts plants, specifically their response to radiative fluxes and a novel combination of temperatures and carbon dioxide (CO₂) concentrations. Research toward this goal includes:
   1. Improved accuracy of vegetation and crop modeling that can be coupled with Earth system models (ESMs) to create more accurate vegetation response.
   2. Using existing observational data, such as natural experiments, to examine how plants and crops responded to past climate variations that resembled SAI-like conditions.
   3. Designing targeted controlled experiments that systematically change the above variables.

The climate modeling downscaling activity is an important complement to this work, which would provide updated simulations to feed into the offline modeling in (1), above. This area of study will be ongoing, with multiple impact studies over time. This research activity is focused on the initial scope of work needed to meaningfully advance scientific understanding of these processes and make future research faster and more robust. Reflective has funded the following projects in this area of study as of September 2026.

• The impact of SAI on primary productivity and global food supply. Dr. Nicole Lovenduski, University of Colorado Boulder and Dr. Lili Xia, Rutgers University
• Impact of stratospheric aerosol injection on agroclimatic extremes and crop suitability in West Africa. Dr. Vincent Olarenwaju Ajayi, Federal University of Technology, Akure and Dr. Temitope Samuel Egbebiyi, University of Cape Town