Understanding the role of mean state biases in the stratospheric response to SAI

The stratospheric circulation is driven by complex wave-mean flow interactions, and models and even reanalysis products have significant differences in their representation of the stratospheric circulation. Nevertheless, these models have been used to predict climate change due to Solar Radiation Management (SRM) strategies. The introduction of aerosol or aerosol precursors to create an artificial reflective stratospheric aerosol layer, known as stratospheric aerosol injection (SAI), leads to two major radiative effects: (1) a reduction in short wave forcing, and (2) an increase in lower stratospheric heating.

To date, very little work has been done to specifically examine the effects and uncertainties related to this lower stratospheric heating. To fill this gap, we analyze the outputs of a model intercomparison project in which a persistent, idealized heating rate is applied in the equatorial lower stratosphere of five GCMs. Despite an identical forcing, each model exhibits unique stratospheric and surface adjustments. We quantify inter-model differences in the stratospheric dynamical response to heating, providing a basis for understanding discrepancies in the corresponding tropospheric responses. In particular, we assess which features of the mean state stratosphere are most influential in controlling the response to heating, and leverage idealized modeling to explore test the state dependence of the stratospheric response to aerosol induced stratospheric heating. We hope to extend this work to provide a useful basis for future work employing emergent constraints to limit uncertainty across the response.