How Much Do Design Choices Matter for Stratospheric Aerosol Intervention

Abstract:

The spatiotemporal climate impact of Stratospheric Aerosol Intervention (SAI) depends on how SAI is implemented. Recent research has shown that the latitude and season of injection are two major design variables that affect the spatial and seasonal distribution of stratospheric aerosol optical depth (AOD). Other design variables, such as altitude, also influence the AOD. While these design variables provide a great number of degrees of freedom for designing SAI, they also produce a high-dimensional design space. Various SAI design choices may not all contribute to the diversity of the resulting climate impacts, as the uniqueness of the climate impacts of a particular strategy depends not only on the cooling objectives, but also effectively on natural variability. If one can’t tell the difference between two different strategies, then those strategies are effectively equivalent. This leads to an important question: how much do design choices matter for SAI?

Their importance would depend on the chosen objectives of SAI. The main one is to reduce the global temperature while maintaining reasonable surface climate in different regions across the planet. The ability to distinguish different surface climate responses from different SAI design choices depends on the desired amount of cooling. With more cooling, the signal-to-noise ratio increases and the difference between various SAI design choices tends to become more noticeable.

Here, we introduce a novel approach to systematically evaluate the importance of design choices under different scenarios to address a fundamental question regarding SAI: how “big” is the design space? For a given level of cooling, how many usefully-independent degrees of freedom are available in tailoring outcomes to meet different objectives? This approach not only helps understand how much design choices matter for SAI, but also helps reduce the dimensionality of the design space, which enables the identification of alternative optimal SAI designs with different trade-off considerations. This will ultimately help support future informed decisions regarding climate intervention.